CT and MRI of the Whole Body, Sixth Edition

CT AND MRI OF THE WHOLE BODY

VOLUME I

CT AND MRI OF THE WHOLE BODY Sixth Edition John R. Haaga, MD, FACR, FSIR, FSCBT, FSRS Gold Medalist AARS and SCBTMRI Professor of Radiology Case Western Reserve University School of Medicine Emeritus Chairman and Professor of Radiology University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio

Daniel T. Boll, MD, FSCBT Professor of Radiology Section Chief Abdominal and Oncologic Imaging University Hospital of Basel Basel, Switzerland

1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899

CT AND MRI OF THE WHOLE BODY, SIXTH EDITION ISBN: 978-0-323-11328-1 Volume 1 Part Number: 9996117383 Volume 2 Part Number: 9996117324 Copyright © 2017 by Elsevier, Inc. All rights reserved. The copyright for chapter 2 is owned by the author, Mark Patrick Supanich. Seth Kligerman retains right for his original images. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher, except that, until further notice, instructors requiring their students to purchase CT and MRI of the Whole Body by John R. Haaga, MD, may reproduce the contents or parts thereof for instructional purposes, provided each copy contains a proper copyright notice as follows: Copyright © 2017 by Elsevier Inc. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Previous editions copyrighted 2009, 2003, 1994, 1988, and 1983. Library of Congress Cataloging-in-Publication Data Names: Haaga, John R. (John Robert), 1945- , editor. | Boll, Daniel T., editor. Title: CT and MRI of the whole body / edited by John R. Haaga, and Daniel T. Boll. Description: Sixth edition. | Philadelphia, PA : Elsevier, [2017] | Includes bibliographical references and index. Identifiers: LCCN 2016001268 | ISBN 9780323113281 (hardcover : international   edition : alk. paper) | ISBN 9780801670572 Subjects: | MESH: Tomography, X-Ray Computed | Magnetic Resonance Imaging | Whole Body Imaging Classification: LCC RC78.7.T6 | NLM WN 206 | DDC 616.07/54—dc23 LC record available at http://lccn.loc .gov/2016001268 Executive Content Strategist: Robin Carter Senior Content Development Specialist: Ann Ruzycka Anderson Publishing Services Manager: Patricia Tannian Senior Project Manager: Cindy Thoms Design Direction: Amy Buxton Printed in China Last digit is the print number:  9  8  7  6  5  4  3  2  1

This book is dedicated to Elizabeth E. Haaga, daughter of John and Ellen Haaga, who was born on August 19, 1972, and died December 9, 1985. Beth had a disseminated neuroblastoma, which was diagnosed in 1984. She was treated with a bone marrow transplant and died from graft-versus-host disease and infection. As her parents, we loved her dearly and cherish the memory of her early years when she was well. After the onset of her illness, we came to know that her gentle and loving nature was accompanied by a remarkably strong character. She endured her pain and suffering without bitterness and never sought to hurt those who loved her. Indeed, most incredulously, she tried to lessen our emotional pain even while enduring her physical discomforts. Many authors have marveled at the qualities of children, and although Beth’s short life and premature death have left us saddened beyond comprehension, her remarkable courage and sweetness have given us a lasting pride and respect. We remember her lovingly. The book is also dedicated to my wonderful family: my wife and soul mate, Ellen, and our children, Matthew and Stacie Haaga, Timothy and Molly Haaga, and Rebecca Haaga. Although professional successes are important and rewarding, the joy and pride of a loving family far surpass such accomplishments. All my love to my current family (and any future additions).

John R. Haaga

I am proud that my efforts as a “narrative diagnostician” have helped young physicians familiarize themselves with the challenges of modern clinical radiology. Hence, professionally, I want to dedicate this book to the wonderful mentors with whom I was honored to work over the years. A hallmark of any medical science is its continuously evolving and expanding knowledge base which, in turn, requires regular reevaluation and revitalization of established workflows. I want to thank my mentors for never tiring in reminding me of this fact and inspiring me to find new approaches to improve patient care. Jonathan S. Lewin, MD (Emory University, Atlanta, GA), a gentleman scholar who guided me through the continuously evolving field of applied MRI physics and the parallel research methods; as a role model he influenced me greatly as I evolved into a scientist and mentor myself. Jeffrey L. Duerk, PhD (Case Western Reserve University, Cleveland, OH), an exceptional teacher and visionary inventor, never tiring of “translating” formulas, figures, and (magnetic) field theories, thereby bringing together physicists and physicians in order to broaden everyone’s horizon beyond expectation. Erik K. Paulson, MD (Duke University Medical Center, Durham, NC), an outstanding diagnostician, teacher and leader who inspired me not only with his expertise and devotion to medical science and clinical practice but also through his willingness to explore unbeaten paths, thereby assembling an outstanding group of like-minded physicians and clinicians who cross-inspire each other. And finally, my life-long friend, visionary academician and exceptional mentor, Elmar M. Merkle, MD (University Hospital of Basel, Basel, Switzerland), who has been there every step of the way and without whom my career would not have been possible. When we all met initially, we all were in young stages of our careers. Over the evolving decades, our friendships formed and matured, emphasizing this most important characteristic of mentorship. Even now, meeting together with the now Executive Vice President for Health Affairs, the Dean of Biomedical Engineering, and two Chairmen of visionary and fascinating radiology departments, respectively, helpful and honest advice is always obtained, for mentorship is seen by these outstanding individuals as a lifelong responsibility. Thank you to all. I also want to dedicate this book to my parents, brother, and wife, who, without having (too much of) a choice, have been there every step of the way and always encouraged me to live a life as a Weltbürger, finding friends and mentors all across the world. I will see you at home, wherever this will be; thank you for everything.

Daniel T. Boll

SECTION EDITORS Kristine A. Blackham, MD

Raj M. Paspulati, MD

Assistant Professor Divisions of Radiology and Neurosurgery University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio

Associate Professor Department of Radiology University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio

Mauricio Castillo, MD

Prabhakar Rajiah, MD

Distinguished Professor of Radiology Chief of Neuroradiology University of North Carolina Chapel Hill, North Carolina

Associate Professor of Radiology Cardiothoracic Imaging Associate Director of Cardiac CT and MRI University of Texas Southwestern Medical Center Dallas, Texas

Thorsten R. Fleiter, MD Associate Professor Department of Diagnostic Imaging and Nuclear Medicine University of Maryland Medical Center R Adams Cowley Shock Trauma Center Baltimore, Maryland

Robert C. Gilkeson, MD Director of Cardiothoracic Imaging Professor of Radiology University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio

Claudia M. Hillenbrand, PhD Associate Member Department of Diagnostic Imaging Division of Translational Imaging Research St. Jude Children’s Research Hospital Memphis, Tennessee

Richard A. Leder, MD Associate Professor of Radiology Clinical Associate in Surgery Duke University School of Medicine Durham, North Carolina

Mark R. Robbin, MD Professor Department of Radiology Chief, Musculoskeletal Imaging University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio

Ehsan Samei, PhD, DABR, FAAPM, FSPIE Professor of Radiology, Medical Physics, Biomedical Engineering, Physics, and Electrical and Computer Engineering Director, Carl E. Ravin Advanced Imaging Laboratories Founding Chief Clinical Imaging Physics Group Duke University Medical Center Durham, North Carolina

Jeffrey L. Sunshine, MD, PhD Professor of Radiology, Neurology, Neurosurgery Vice Chairman, Department of Radiology University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Chief Medical Information Officer, University Hospitals Cleveland, Ohio

Frank K. Wacker, MD Jeong Min Lee, MD Professor Department of Radiology Seoul National University Hospital Seoul, Korea

Suresh K. Mukherji, MD, MBA, FACR Professor and Chairman Department of Radiology Walter F. Patenge Endowed Chair Chief Medical Officer and Director of Health Care Delivery Michigan State University Health Team East Lansing, Michigan

viii

Institute of Diagnostic and Interventional Radiology University School of Medicine Hannover, Germany

CONTRIBUTORS James J. Abrahams, MD

Mustafa R. Bashir, MD

Donna G. Blankenbaker, MD

Professor, Diagnostic Radiology Diagnostic Radiology Program Fellowship Director, Neuroradiology Yale University School of Medicine New Haven, Connecticut Orbit

Assistant Professor Department of Radiology Duke University Medical Center Durham, North Carolina Tissue Characterization in Liver Imaging Using Advanced MR Techniques

Professor of Radiology Department of Radiology University of Wisconsin School of Medicine and Public Health Madison, Wisconsin Hip and Pelvis

Michael Abrahams, MD

Javier Beltran, MD, FACR

Uttam K. Bodanapally, MBBS

Department of Emergency Medicine University of Toledo Toledo, Ohio Orbit

Chairman Department of Radiology Maimonides Medical Center Brooklyn, New York Clinical Director of Musculoskeletal Radiology Radisphere National Radiology Group Beachwood, Ohio Knee

Assistant Professor Department of Diagnostic Radiology and Nuclear Medicine University of Maryland Medical Center R Adams Cowley Shock Trauma Center Baltimore, Maryland Traumatic Brain Injury

Federica Agosta, MD, PhD Neuroimaging Research Unit Institute of Experimental Neurology Division of Neuroscience San Raffaele Scientific Institute Vita-Salute San Raffaele University Milan, Italy Neurodegenerative Disorders

Muneeb Ahmed, MD, FSIR Chief Division of Vascular and Interventional Radiology Beth Israel Deaconess Medical Center Associate Professor of Radiology Harvard Medical School Boston, Massachusetts Image-Guided Ablation of Parenchymal Organs

Stuart Bentley-Hibbert, MD, PhD Associate Professor of Radiology Department of Radiology Columbia University Medical Center New York, New York Mesentery

Professor Radiodiagnosis Tata Memorial Hospital Mumbai, India Pharynx

Associate Professor Department of Imaging Sciences University of Rochester Rochester, New York Kidney

Department of Radiology University Hospitals/Case Medical Center Cleveland, Ohio Knee Musculoskeletal Tumors

Sotirios Bisdas, MD, PhD, MSc Yong Ho Auh, MD Professor of Radiology Weill Cornell Medical College of Cornell University Attending Radiologist New York-Presbyterian Hospital New York, New York Mesentery

Romulo Baltazar, MD Attending Radiologist Imaging Subspecialists of North Jersey St. Joseph’s Regional Medical Center Paterson, New Jersey Knee

Professor of Radiology Associate Dean for Academic and Clinical Affairs Harvard Medical School Boston, Massachusetts Airway

Daniel T. Boll, MD, FSCBT Shweta Bhatt, MD

Nicholas Bhojwani, MD Supreeta Arya, MD, DMRD, DNB

Phillip M. Boiselle, MD

Neuroradiology Consultant National Hospital for Neurology and Neurosurgery Honorary Senior Lecturer University College London London, United Kingdom Professor of Radiology Department of Neuroradiology Karls Eberhard University Tübingen, Germany Adenopathy and Neck Masses

Professor of Radiology Section Chief Abdominal and Oncologic Imaging University Hospital of Basel Basel, Switzerland Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases Tissue Characterization in Liver Imaging Using Advanced Magnetic Resonance Techniques

Eliana Bonfante, MD Assistant Professor of Radiology Department of Diagnostic and Interventional Imaging University of Texas Health Science Center Houston, Texas Spinal Trauma

Jeffrey R. Brace, MD Assistant Professor of Radiology University of Minnesota Medical Center, Fairview Hennepin County Medical Center Minneapolis Veterans Affairs Medical Center Minneapolis, Minnesota Intracranial Neoplasms

Kristine A. Blackham, MD Assistant Professor Divisions of Radiology and Neurosurgery University Hospitals/Case Medical Center Case Western Reserve University Cleveland, Ohio Intracranial Neoplasms

ix

x

Contributors

Miriam A. Bredella, MD

Onofrio Catalano, MD

Jared D. Christensen, MD

Associate Professor of Radiology Harvard Medical School Department of Radiology Musculoskeletal Imaging and Interventions Massachusetts General Hospital Boston, Massachusetts Shoulder

Assistant Professor of Radiology Division of Abdominal Surgery Harvard Medical School Boston, Massachusetts Pancreas

Assistant Professor Radiology Duke University Medical Center Durham, North Carolina Mediastinal Disease

Majid Chalian, MD

Michael Coffey, MD

Radiology House Staff, PGY3 Radiology University Hospitals Case Western Reserve University Cleveland, Ohio High Resolution 3T Magnetic Resonance Neurography: Applications, Techniques, Pitfalls

Assistant Professor Radiology Case Western Reserve University School of Medicine Cleveland, Ohio Demyelinating Disease and Leukoencephalopathies

Michael K. Brooks, MD, MPH Clinical Assistant Professor of Radiology SUNY-Stony Brook Department of Radiology Division of Musculoskeletal Imaging and Intervention Winthrop University Hospital Mineola, New York Degenerative Disease

Christopher Brown, MD Research Analyst Department of Diagnostic Radiology University of Maryland Medical Center Baltimore, Maryland Coronary Arteries, Heart, and Pericardium

Ji Y. Buethe, MD Chief Resident Department of Radiology University Hospitals/Case Medical Center Cleveland, Ohio Image-Guided Drainages

John A. Carrino, MD, MPH Vice Chairman of Radiology Department of Radiology and Imaging Hospital for Special Surgery New York, New York High-Resolution 3T Magnetic Resonance Neurography: Applications, Techniques, and Pitfalls

Brian Cox, MD Ronil V. Chandra, MBBS (Hons), MMed, FRANZCR Associate Professor Neuroradiology Service Monash Imaging Monash Health Stroke and Ageing Research Centre Monash University Melbourne, Victoria, Australia Stroke

Tushar Chandra, MD Clinical Instructor Radiology Medical College of Wisconsin Milwaukee, Wisconsin Pharynx

Apeksha Chaturvedi, MD Assistant Professor Radiology University of Rochester Medical Center Rochester, New York Chest Imaging in the Pediatric Patient

Francesca Caso, MD

Avneesh Chhabra, MD

Neuroimaging Research Unit Institute of Experimental Neurology Division of Neuroscience San Raffaele Scientific Institute Vita-Salute San Raffaele University Milan, Italy Neurodegenerative Disorders

Department of Radiology University of Texas Southwestern Medical Center Dallas, Texas High-Resolution 3T Magnetic Resonance Neurography: Applications, Techniques, and Pitfalls

Mauricio Castillo, MD

Dong-il Choi, MD

Distinguished Professor of Radiology Chief of Neuroradiology University of North Carolina Chapel Hill, North Carolina Brain Proton Magnetic Resonance Spectroscopy Cystic Lesions

Associate Professor Sungkyunkwan University School of Medicine Faculty, Department of Radiology Samsung Medical Center Seoul, Republic of Korea Biliary Tract and Gallbladder

Department of Radiology University of Texas Southwestern Medical Center Dallas, Texas High-Resolution 3T Magnetic Resonance Neurography: Applications, Techniques, and Pitfalls

Kelly L. Cox, DO Assistant Professor Radiology and Imaging Sciences Emory University Hospitals and School of Medicine Atlanta, Georgia Female Pelvis

Sheilah Curran-Melendez, MD Department of Diagnostic Radiology Allegheny General Hospital Pittsburgh, Pennsylvania Meningeal Processes

Hugh D. Curtin, MD Professor of Radiology Harvard Medical School Chief of Radiology Massachusetts Eye and Ear Boston, Massachusetts Larynx

Carlo N. De Cecco, MD, PhD Assistant Professor Department of Radiology and Radiological Sciences Division of Cardiovascular Imaging Medical University of South Carolina Charleston, South Carolina Advanced Cardiovascular CT Imaging

Contributors Alfred Delumpa, MD

Massimo Filippi, MD

Ritu R. Gill, MD, MPH

Fellow Department of Neuroradiology Baylor College of Medicine Houston, Texas Spinal Tumors

Professor Neuroimaging Research Unit Department of Neurology Division of Neuroscience Institute of Experimental Neurology San Raffaele Scientific Institute Vita-Salute San Raffaele University Milan, Italy Neurodegenerative Disorders

Associate Radiologist Assistant Professor of Radiology Brigham and Women’s Hospital Harvard Medical School Boston, Massachusetts Disease of the Pleura, Chest Wall, and Diaphragm

David R. Di Lorenzo, MD Diagnostic Radiology University Hospitals/Case Medical Center Cleveland, Ohio Peritoneum

Thorsten R. Fleiter, MD

Professor of Radiology and Urology Department of Imaging Sciences University of Rochester Rochester, New York Kidney

Associate Professor Department of Diagnostic Imaging and Nuclear Medicine University of Maryland Medical Center R Adams Cowley Shock Trauma Center Baltimore, Maryland Traumatic Brain Injury

David Dreizin, MD

Tomás Franquet, MD

Assistant Professor Department of Diagnostic Radiology and Nuclear Medicine University of Maryand Medical Center R Adams Cowley Shock Trauma Center Baltimore, Maryland Spinal Cord Injury Traumatic Brain Injury

Chief, Section of Thoracic Imaging Radiology Hospital de Sant Pau Associate Professor of Radiology Universitat Autonoma de Barcelona Barcelona, Spain Nonneoplastic Parenchymal Lung Disease

Vikram S. Dogra, MD

Melanie B. Fukui, MD Jeremy J. Erasmus, MBBCh Professor Diagnostic Radiology University of Texas MD Anderson Cancer Center Houston, Texas Neoplastic Disease of the Lung

Maryam Etesami, MD Radiology Case Western Reserve University University Hospitals Case Medical Center Cleveland, Ohio Chest Imaging in the Pediatric Patient

Steven Falcone, MD, MBA Professor Radiology, Neurological Surgery, Ophthalmology Executive Dean for Clinical Affairs Miller School of Medicine Chief Executive Officer University of Miami Medical Group Associate Vice President for Medical Affairs University of Miami Miami, Florida Systemic Diseases Affecting the Spine

Aurora Neuroscience Innovation Institute St. Luke’s Medical Center Milwaukee, Wisconsin Meningeal Processes

John L. Go, MD Director of Head and Neck Imaging Medical Director of the Imaging Science Center University of Southern California Keck School of Medicine Los Angeles, California Cerebral Infections and Inflammation

Joseph R. Grajo, MD Assistant Professor of Radiology University of Florida College of Medicine Gainesville, Florida Male Pelvis

Sachin K. Gujar, MBBS, MD Assistant Professor Neuroradiology Division Russell H. Morgan Department of Radiology and Radiological Science Johns Hopkins University School of Medicine Baltimore, Maryland Functional Magnetic Resonance Imaging

Junyu Guo, PhD Stephen R. Fuller, BS Department of Radiology and Radiological Sciences Division of Cardiovascular Imaging Medical University of South Carolina Charleston, South Carolina Advanced Cardiovascular CT Imaging

MR Scientist Diagnostic Imaging St. Jude Children’s Research Hospital Memphis, Tennessee Contrast-Enhanced Magnetic Resonance Imaging

Amit Gupta, MD Nicholas L. Fulton, MD Resident Radiology University Hospitals/Case Medical Center University Heights, Ohio Vasculogenesis

Fellow, Neuroradiology University Hospitals/Case Medical Center Cleveland, Ohio Demyelinating Disease and Leukoencephalopathies

Hyun Kwon Ha, MD Romyll Garcia, MD Voluntary Assistant Professor Radiology University of Miami Miami, Florida Systemic Diseases Affecting the Spine

Amilcare Gentili, MD Professor Radiology University of California–San Diego San Diego, California Musculoskeletal Tumors

Professor of Radiology University of Ulsan College of Medicine Professor and Chief of Gastrointestinal Section Department of Radiology Asan Medical Center Seoul, Republic of Korea Gastrointestinal Tract

xi

xii

Contributors

John R. Haaga, MD, FACR, FSIR, FSCBT, FSRS Gold Medalist AARS and SCBTMRI Professor of Radiology Case Western Reserve University School of Medicine Emeritus Chairman and Professor of Radiology University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio Image-Guided Aspirations and Biopsies Imaging Principles in Computed Tomography Peritoneum Vasculogenesis

Ellen Hoeffner, MD

Stephen A. Kieffer, MD

Professor of Radiology University of Michigan Ann Arbor, Michigan Pharynx Temporal Bone

Professor of Radiology (Neuroradiology) University of Minnesota Medical School Attending Radiologist University of Minnesota Medical Center, Fairview Hemepin County Medical Center Minneapolis Veterans Affairs Medical Center Minneapolis, Minnesota Intracranial Neoplasms

Alena Horská, PhD Assistant Professor Russell H. Morgan Department of Radiology and Radiological Science Division of Neuroradiology Johns Hopkins University School of Medicine Baltimore, Maryland Brain Proton Magnetic Resonance Spectroscopy

Timothy L. Haaga, MD Clinical Assistant Professor Radiology University Hospitals/Case Medical Center Case Western Reserve University School of Medicine Cleveland, Ohio Image-Guided Aspirations and Biopsies

Mukesh G. Harisinghani, MD Professor of Radiology Harvard Medical School Boston, Massachusetts Male Pelvis

Tobias Heye, MD Department of Radiology University Hospital of Basel Basel, Switzerland Tissue Characterization in Liver Imaging Using Advanced Magnetic Resonance Techniques

Claudia M. Hillenbrand, PhD Associate Member Department of Diagnostic Imaging Division of Translational Imaging Research St. Jude Children’s Research Hospital Adjunct Associate Professor Department of Biomedical Engineering University of Memphis Memphis, Tennessee Magnetic Resonance Imaging in the Pediatric Patient

Lisa M. Ho, MD Associate Professor of Radiology Department of Radiology Duke University Medical Center Durham, North Carolina Adrenal Glands

David W. Jordan, PhD Senior Medical Physicist, Radiology University Hospitals/Case Medical Center Assistant Professor, Radiology Case Western Reserve University Cleveland, Ohio Imaging Principles in Computed Tomography Imaging Principles in Magnetic Resonance Imaging

Ah Young Kim, PhD Associate Professor of Radiology University of Ulsan College of Medicine Associate Professor of Radiology Division of Abdominal Radiology Asan Medical Center Seoul, Republic of Korea Gastrointestinal Tract

Jung Hoon Kim, MD Assistant Professor Medicine-Geriatrics Baylor College of Medicine Houston, Texas Mesentery

Kyoung Won Kim, MD, PhD Varsha Joshi, DNB, DMRD Senior Consultant Division of CT and MRI Vijaya Diagnostic Centres Hyderabad, India Visiting Consultant Tata Medical Center Kolkata, India Paranasal Sinuses

Sue C. Kaste, DO Full Member Department of Diagnostic Imaging Department of Oncology St. Jude Children’s Research Hospital Full Professor, Radiology University of Tennessee School of Health Sciences Memphis, Tennessee Magnetic Resonance Imaging in the Pediatric Patient

Simon Rupe Khangure, MD, MBBS, BMdSc, FRANZCR Neuroradiology Service Monash Imaging Monash Health Melbourne, Victoria, Australia Stroke

Assistant Professor of Radiology University of Ulsan College of Medicine Faculty Member, Department of Radiology Asan Medical Center Seoul, Republic of Korea Biliary Tract and Gallbladder

Paul E. Kim, MD University of Southern California Keck School of Medicine Los Angeles, California Cerebral Infections and Inflammation

Seth Kligerman, MD Associate Professor of Radiology Diagnostic Radiology and Nuclear Medicine University of Maryland Baltimore, Maryland Coronary Arteries, Heart, and Pericardium

Satoshi Kobayashi, MD, PhD Professor Quantum Medical Technology Kanazawa University Graduate School of Medical Science Kanazawa, Japan Liver: Focal Hepatic Mass Lesions

Christos Kosmas, MD Assistant Professor of Radiology Division of Musculoskeletal Imaging University Hospitals/Case Medical Center Case Western Reserve University Cleveland, Ohio Foot and Ankle

Contributors

xiii

Gerdien Kramer, MD

Jae Hoon Lim, MD

Osamu Matsui, MD, PhD

Department of Thoracic Imaging Hopital Calmette Universite Lille Nord de France Lille, France Coronary Arteries, Heart, and Pericardium

Department of Radiology Myongji Hospital Goyang, Republic of Korea Biliary Tract and Gallbladder

Professor Emeritus Radiology Kanazawa University Graduate School of Medical Science Kanazawa, Japan Liver: Focal Hepatic Mass Lesions

Chia-Shang J. Liu, MD, PhD Timo Krings, MD, PhD, FRCPC The David Braley and Nancy Gordon Chair in Interventional Neuroradiology Chief of Diagnostic and Interventional Neuroradiology Toronto Western Hospital and University Health Network Professor of Radiology and Surgery University of Toronto Toronto, Ontario, Canada Vascular Lesions

Assistant Professor Radiology University of Southern California Keck School of Medicine Los Angeles, California Cerebral Infections and Inflammation

Minh Lu, MD Radiology University of Maryland Baltimore, Maryland Coronary Arteries, Heart, and Pericardium

Lester Kwock, PhD Professor of Radiology University of North Carolina School of Medicine Chapel Hill, North Carolina Brain Proton Magnetic Resonance Spectroscopy

Kenneth Lury, MD Assistant Professor (Retired) Radiology University of North Carolina Chapel Hill, North Carolina Noninfectious Inflammatory Diseases Affecting the Spinal Cord

Theodore C. Larson III, MD Director of Neurointervention Centura Neurosciences St. Anthony Hospital Lakewood, Colorado Cerebral Aneurysms and Cerebrovascular Malformations

Calvin T. Ma, MD

Leslie K. Lee, MD

Henry Ma, MBBS, FRACP

Instructor in Radiology Harvard Medical School Boston, Massachusetts Male Pelvis

Associate Professor Stroke Unit, Monash Health Stroke and Ageing Research Centre Monash University Melbourne, Victoria, Australia Stroke

Seung Soo Lee, MD Assistant Professor of Radiology University of Ulsan College of Medicine Seoul, Republic of Korea Gastrointestinal Tract

Young-Jun Lee, MD, PhD Associate Professor, Radiology Hanyang University Hospital Seoul, South Korea Spinal Vascular Diseases

Alexander Lerner, MD Assistant Professor of Clinical Radiology Radiology University of Southern California Keck School of Medicine Los Angeles, California Cerebral Infections and Inflammation

Staff Radiologist Sutter Imaging Sutter Health Sacramento Sierra Region Sacramento, California Knee

Bahar Mansoori, MD Department of Radiology University Hospitals Case Western Reserve University Cleveland, Ohio High-Resolution 3T Magnetic Resonance Neurography: Applications, Techniques, and Pitfalls

Joseph P. Mazzie, DO Clinical Assistant Professor of Radiology SUNY-Stony Brook Department of Radiology Division of Musculoskeletal Imaging and Intervention Winthrop University Hospital Mineola, New York Degenerative Disease

H. Page McAdams, MD Professor of Radiology Radiology Duke University Medical Center Durham, North Carolina Mediastinal Disease

Indu Rekha Meesa, MD, MS Neuroradiology and Pediatric Radiologist Summit Radiology Fort Wayne, Indiana Neuroradiology and Pediatric Radiology Fellow Department of Radiology University of Michigan Ann Arbor, Michigan Imaging of the Head and Neck in the Pediatric Patient

Carolyn Cidis Meltzer, MD William P. Timmie Professor and Chair Department of Radiology and Imaging Sciences Associate Dean for Research Emory University School of Medicine Atlanta, Georgia Meningeal Processes

Elmar M. Merkle, MD Department of Radiology University Hospital of Basel Basel, Switzerland Tissue Characterization in Liver Imaging Using Advanced MR Techniques

Daniel Mascarenhas, BS

Achille Mileto, MD

Department of Diagnostic Radiology and Nuclear Medicine University of Maryland Medical Center R Adams Cowley Shock Trauma Center Baltimore, Maryland Spinal Cord Injury Traumatic Brain Injury

Department of Radiology Duke University School of Medicine Durham, North Carolina Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

xiv

Contributors

Suyash Mohan, MD, PDCC

John P. Nazarian, MD

Manuel Patino, MD

Assistant Professor of Radiology Radiology, Neuroradiology Division University of Pennsylvania Philadelphia, Pennsylvania Pharynx Temporal Bone

Resident, Radiology University Hospitals/Case Medical Center Cleveland, Ohio Demyelinating Disease and Leukoencephalopathies

Postdoctoral Fellow Division of Abdominal Imaging Massachusetts General Hospital Boston, Massachusetts Pancreas

Manal Nicolas-Jilwan, MD

Thanh Phan, MD, MBBS, FRACP

Radiology University of Virginia Health System Charlottesville, Virginia Congenital Abnormalities

Professor Stroke Unit, Monash Health Stroke and Ageing Research Centre Monash University Melbourne, Australia Stroke

Jesse Montagnese, MD Assistant Professor Radiology University of Pittsburgh Pittsburgh, Pennsylvania Intracranial Neoplasms

Sameh K. Morcos, FRCS, FFRRCSI, FRCR Professor of Diagnostic Imaging University of Sheffield Consultant Radiologist Northern General Hospital Sheffield, United Kingdom Contrast Nephropathy and Its Management

Sherif G. Nour, MD, FRCR Director, Interventional MRI Program Associate Professor of Radiology and Imaging Sciences Department of Radiology and Imaging Sciences Emory University Hospitals and School of Medicine Atlanta, Georgia Female Pelvis MRI-Guided Interventions

Fanny E. Morón, MD Associate Professor of Diagnostic Radiology Baylor College of Medicine Houston, Texas Spinal Neoplasms

Suresh Mukherji, MD, MBA, FACR Professor and Chairman Department of Radiology Walter F. Patenge Endowed Chair Chief, Medical Officer and Director of Health Care Delivery Michigan State University Health Team East Lansing, Michigan Imaging of the Head and Neck in the Pediatric Patient

A. Orlando Ortiz, MD, MBA, FACR Clinical Professor of Radiology SUNY-Stony Brook Chairman, Department of Radiology Winthrop University Hospital Mineola, New York Degenerative Disease

Theeraphol Panyaping, MD Diagnostic Neuroradiology Division Department of Radiology Ramathibodi Hospital Mahidol University Ratchathewi, Bangkok, Thailand Spinal Infection

Dean A. Nakamoto, MD

Seong Ho Park, MD

Section Chief, Abdominal Imaging Radiology University Hospitals/Case Medical Center Cleveland, Ohio Computed Tomography—Guided Drainage Spleen

Associate Professor of Radiology University of Ulsan College of Medicine Asan Medical Center Seoul, Republic of Korea Gastrointestinal Tract

Raj M. Paspulati, MD Sadhna B. Nandwana, MD Assistant Professor Department of Radiology and Imaging Sciences Emory University Hospitals and School of Medicine Atlanta, Georgia Female Pelvis

Associate Professor Radiology University Hospitals/Case Medical Center Case Western Reserve University Cleveland, Ohio Liver Transplantation Rectum

Jay J. Pillai, MD Director of Functional MRI Associate Professor Neuroradiology Division Russell H. Morgan Department of Radiology and Radiological Science Johns Hopkins University School of Medicine Baltimore, Maryland Brain Proton Magnetic Resonance Spectroscopy Functional Magnetic Resonance Imaging

Colin S. Poon, MD, PhD, FRCPC Assistant Clinical Professor Department of Diagnostic Radiology Yale University School of Medicine New Haven, Connecticut Staff Radiologist Orillia Soldiers’ Memorial Hospital Orillia, Ontario, Canada Orbit

Andrea Prochowski, MD Postdoctoral Fellow Division of Abdominal Imaging Massachusetts General Hospital Boston, Massachusetts Pancreas

Anandh Rajamohan, MD Assistant Professor Radiology University of Southern California Keck School of Medicine Los Angeles, California Cerebral Infections and Inflammation

Prabhakar Rajiah, MD Associate Professor of Radiology Cardiothoracic Imaging Associate Director of Cardiac CT and MRI University of Texas Southwestern Medical Center Dallas, Texas Chest Imaging in the Pediatric Patient

Contributors Parvati Ramchandani, MD

Nisha Sainani, MD

Eric Schrauben, PhD

Professor of Radiology and Surgery, Section Chief GU Radiology Department of Radiology Hospital of the University of Pennsylvania Philadelphia, Pennsylvania Retroperitoneum

Assistant Professor of Radiology Harvard Medical School Staff Radiologist Abdominal Imaging and Intervention Brigham and Women’s Hospital Boston, Massachusetts Pancreas

Medical Physics University of Wisconsin–Madison Madison, Wisconsin Imaging Principles in Magnetic Resonance Angiography

Wilburn E. Reddick, PhD Member Faculty Diagnostic Imaging St. Jude Children’s Research Hospital, Memphis, Tennessee Contrast-Enhanced Magnetic Resonance Imaging

Matthias Renker, MD Visiting Scholar Department of Radiology and Radiological Services Division of Cardiovascular Imaging Medical University of South Carolina Charleston, South Carolina Resident, Department of Medicine I, Cardiology/Angiology University Hospital Giessen Giessen, Germany Advanced Cardiovascular CT Imaging

Roy Riascos, MD, DABR Associate Professor Chief of Neuroradiology Department of Diagnostic and Interventional Imaging University of Texas Health Science Center at Houston Houston, Texas Spinal Trauma

Mark R. Robbin, MD Professor Department of Radiology Chief, Division of Musculoskeletal Imaging University Hospitals/Case Medical Center Case Western Reserve School of Medicine Cleveland, Ohio Musculoskeletal Tumors

Haris I. Sair, MD Assistant Professor Neuroradiology Division Russell H. Morgan Department of Radiology and Radiological Science Johns Hopkins University School of Medicine Baltimore, Maryland Functional Magnetic Resonance Imaging

Dilan Samarawickrama, MD Neuroradiology Fellow Radiology University of Michigan Ann Arbor, Michigan Temporal Bone

Rima Sansi, DNB Clinical Associate Department of Radiology Sir H.N. Reliance Foundation Hospital and Research Centre Prarthana Samaj, Girgaum, Mumbai, India Paranasal Sinuses

Asha Sarma, MD Resident Radiology Brigham and Women’s Hospital Boston, Massachusetts Disease of the Pleura, Chest Wall, and Diaphragm

Piyush Saxena, DNB, MBBS Consultant Radiologist and Chief of MRI Radiodiagnosis Vishesh Hospital and Diagnostics Indore, Madhya Pradesh, India Paranasal Sinuses

Santiago E. Rossi, MD Centro de Diagnostico Dr Enrique Rossi Buenos Aires, Argentina Neoplastic Disease of the Lung

Dushyant Sahani, MD Associate Professor of Radiology Harvard Medical School Director of Computed Tomography Massachusetts General Hospital Division of Abdominal Imaging and Intervention Boston, Massachusetts Pancreas

Ken L. Schreibman, MD, PhD Professor of Radiology Division of Musculoskeletal Imaging University of Wisconsin–Madison University of Wisconsin Hospitals and Clinics Madison, Wisconsin Foot and Ankle

Danielle M. Seaman, MD Staff Radiologist Radiology Durham Veterans Affairs Medical Center Durham, North Carolina Mediastinal Disease

Nicole Seiberlich, PhD Assistant Professor Biomedical Engineering Case Western Reserve University Cleveland, Ohio Imaging Principles in Magnetic Resonance Imaging

Saugata Sen, MD Senior Consultant Radiology and Nuclear Medicine Tata Medical Center Kolkata, West Bengal, India Paranasal Sinuses

Rickin Shah, MD Department of Radiology University of Michigan Health System Ann Arbor, Michigan Temporal Bone

Steven Shankman, MD Vice-Chairman, Department of Radiology Maimonides Medical Center Brooklyn, New York Knee

U. Joseph Schoepf, MD Professor Department of Radiology and Radiological Sciences Division of Cardiovascular Imaging Director, CT Research and Development Medical University of South Carolina Charleston, South Carolina Advanced Cardiovascular CT Imaging

xv

Kenneth Sheah, MBBS FRCR MMed (Diagnostic Radiology) Consultant Radiologist Lifescan Imaging Pte Ltd Singapore Shoulder

xvi

Contributors

Haojun Shi, MD

Mark Patrick Supanich, PhD

Charles S. White, MD

Assistant Professor of Radiology Huazhong University of Science and Technology Department of Radiology Union Hospital Wuhan, Hubei Province, China Normal Anatomy

Associate Professor Diagnostic Radiology and Nuclear Medicine Rush University Medical Center Chicago, Illinois Computed Tomography Imaging Operation

Professor of Radiology and Medicine University of Maryland Medical Center Baltimore, Maryland Coronary Arteries, Heart, and Pericardium

George Shih, MD Adjunct Assistant Professor of Radiology Columbia University College of Physicians and Surgeons Associate Attending Radiologist New York-Presbyterian Hospital Associate Professor of Clinical Radiology Weill Cornell Medical College New York, New York Mesentery

Julian Lukas Wichmann, MD Katarina Surlan-Popovic, MD, PhD Associate Professor of Radiology Clinical Institute of Radiology University Medical Center Ljubljana Ljubljana, Slovenia Adenopathy and Neck Masses

Janio Szklaruk, PhD, MD Professor of Diagnostic Radiology University of Texas MD Anderson Cancer Center Houston, Texas Spinal Tumors

Mark S. Shiroishi, MD

Visiting Instructor Department of Radiology and Radiological Sciences Division of Cardiovascular Imaging Medical University of South Carolina Charleston, South Carolina Resident Department of Diagnostic and Interventional Radiology University Hospital Frankfurt Frankfurt, Germany Advanced Cardiovascular CT Imaging

Oliver Wieben, PhD

Assistant Professor Radiology University of Southern California Keck School of Medicine Los Angeles, California Cerebral Infections and Inflammation

Gustavo A. Tedesqui, MD Research Fellow Neuroradiology University of North Carolina Chapel Hill, North Carolina Cystic Lesions

Associate Professor Medical Physics and Radiology University of Wisconsin–Madison Madison, Wisconsin Imaging Principles in Magnetic Resonance Angiography

Stephanie Soriano, MD

Aaryani Tipirneni-Sajja, MS

Leo Wolansky, MD

Radiology University Hospitals/Case Medical Center Cleveland, Ohio Spleen

Research Assistant Division of Translational Imaging Research Department of Diagnostic Imaging St. Jude Children’s Research Hospital Research Assistant Department of Biomedical Engineering University of Memphis Memphis, Tennessee Magnetic Resonance Imaging in the Pediatric Patient

Professor, Radiology Case Western Reserve University School of Medicine Cleveland, Ohio Demyelinating Disease and Leukoencephalopathies

Cathy Soufan, MD Neuroradiology Service Monash Imaging Monash Health Melbourne, Victoria, Australia Stroke

Drew A. Torigian, MD, MA Jason W. Stephenson, MD Assistant Professor of Radiology Department of Radiology University of Wisconsin School of Medicine and Public Health Madison, Wisconsin Hip and Pelvis

Stephen L. Stuckey, MBBS, MD, MMed, FRANZCR, FAANMS, PGDA Director Monash Imaging Monash Health Department of Imaging Monash University Melbourne, Victoria, Australia Stroke

Associate Professor of Radiology Department of Radiology Hospital of the University of Pennsylvania Philadelphia, Pennsylvania Retroperitoneum

Erik M. Velez, MD Department of Radiology Beth Israel Deaconess Medical Center Boston, Massachusetts Image-Guided Ablation of Parenchymal Organs

Thomas J. Vogl, MD Professor of Radiology Chair Department of Diagnostic and Interventional Radiology University Hospital Frankfurt Frankfurt, Germany Adenopathy and Neck Masses

Hanping Wu, MD Resident Department of Radiology University Hospitals/Case Medical Center Cleveland, Ohio Image-Guided Aspirations and Biopsies Normal Anatomy

Haibo Xu, MD Chairman Department of Radiology Zhongnan Hospital of Wuhan University Wuhan, Hubei Province, China Normal Anatomy

Chi-Shing Zee, MD Professor of Radiology and Neurosurgery Department of Radiology Keck Hospital of University of Southern California Los Angeles, California Cerebral Infections and Inflammation

P R E FA C E This sixth edition of CT and MRI of the Whole Body represents a monumental accomplishment for Elsevier and our many contributing authors, editors, and section editors who have contributed past and present. The authors of the current chapters are due many, many kudos because of the changes that have occurred in healthcare with the emphasis on “productivity” and RVUs. Institutions have reduced physician support in the form of secretaries, assistants, etc., which makes it very difficult for academic physicians to contribute to projects such as this book. THANK YOU, THANK YOU! Being senior editor of this book has been a remarkable honor for me because it enabled me to collaborate with many radiologists and physicists. Everyone who has been involved for the past 40 years can take great pride in this book because it has formed the foundation for educating thousands of fellow radiologists in the United States and around the world. Over the years, radiologists from all over the world have expressed their gratitude for having this book available as a reference. As I reflect back, I become nostalgic just thinking about how exciting radiology has been for the past 40 years. So many innovations and advances have been made. Of the contributions I have made, I am the most proud of developing CT interventions such as biopsies, abscess drainages, nerve blocks, the Topogram/Scoutview, and radiation dose reduction. It is very flattering that the Smithsonian Institution asked for and accepted for curation several personal artifacts related to these innovations. Over the years I have met many pioneers in the field of imaging, but my two favorites were Sven Seldinger, MD (Fig. 1), and Paul Lauterbur, PhD (Fig. 2), as pictured below. Dr. Seldinger was a very dynamic person who developed the Seldinger Technique for angiography.

Unfortunately, he didn’t receive the recognition he deserved until late in his career. Few know that he did not receive an academic appointment until we awarded him with one at Case Western Reserve University. Paul Lauterbur was the genius who received the Nobel Prize for MRI. Strangely enough, he and I lived in the same small town, Troy, Ohio, for a number of years. His sense of humor as well as his science was of Nobel quality. I have taken the editor’s prerogative of referencing and publishing my own innovations in this book. My last innovation, which I am confident will improve the diagnosis and treatment of cancer, is the ALPHA concept, which is discussed in Chapter 70. The concept has been published in the journals Surgery and the American Journal of Roentgenology (AJR). We have just finished the experimental proof of ALPHA, which includes standard laboratory tests, but also the latest RNA sequencing tests. We will be submitting it for publication soon. This concept explains how aerobic metabolism has been overemphasized and that glycolysis is more important for the development and treatment of cancer. I hope and believe that we will be able to diagnose and cure many more cancers with this knowledge. Readers are encouraged to search the literature for future publications by my protégés who are continuing this work. There are numerous young physicians working on this concept who have accomplished much to date. I would like to acknowledge a number of young investigators, Hooman Yarmohammadi, MD; Luke Wilkins, MD; Steven Dreyer, MD; Dan Patel, MD; Nicholas Fulton, and Ji Buethe, for their hard work and commitment to continue work on this ALPHA concept.

FIG 1  Sven Seldinger (right) and John R. Haaga.

FIG 2  Paul Lauterbur (right) and John R. Haaga.

John R. Haaga

xvii

P R E FAC E TO T H E F I R S T E D I T I O N Since the introduction of computed tomography (CT) in 1974, there has been a remarkable revolution in the medical treatment of patients. The clinical use of CT has had a broad positive impact on patient management. Literally thousands of patients have been saved or their quality of life improved as a result of the expeditious and accurate diagnosis provided by CT. This improvement in diagnosis and management has occurred in all medical subspecialties, including neurological, pulmonary, cardiac, gastrointestinal, genitourinary, and neuromuscular medicine. Aside from the imaging advantages provided, the role of CT in planning and performing interventional procedures is now recognized. It is the most accurate method for guiding procedures to obtain cytological, histological, or bacteriological specimens and for performing a variety of therapeutic procedures. The evolution and refinement of CT equipment have been as remarkable as the development of patient diagnosis. When we wrote our first book on CT, the scanning unit used was a 2-minute translaterotate system. At the time of our second book the 18-second translaterotate scanning unit was in general use. Currently standard units in radiological practice are third and fourth generation scanners with scan times of less than 5 seconds. All modern systems are more reliable than the earlier generations of equipment. The contrast and spatial resolution of these systems are in the range of 0.5% and less than 1 mm, respectively. The sophistication of the computer programs that aid in the diagnosis is also remarkable. There are now programs for three-dimensional reconstructions, quantitation of blood flow, determination of organ volume, longitudinal scans (Scoutview, Deltaview, Synerview, and Topogram), and even triangulation programs for performing percutaneous biopsy procedures. CT units are now being installed in virtually every hospital of more than 200 beds throughout the United States. Most radiologists using these units are generalists who scan all portions of the anatomy. Because of the dissemination of this equipment and its use in general diagnosis, there exists a significant need for a general and complete textbook to cover all aspects of CT scanning. Our book is intended to partially supplement the knowledge of this group of physicians. We have attempted to completely and succinctly cover all portions of CT scanning to provide a complete general reference text. In planning the book, we chose to include the contributions of a large number of talented academicians with expertise different from and more complete than our own in their selected areas. By including contributors

xviii

from outside our own department, we have been able to produce an in-depth textbook that combines the academic strengths of numerous individuals and departments. The book is divided into chapters according to the organ systems, except for some special chapters on abscesses and interventional procedures. In each of the chapters the authors have organized the material into broad categories, such as congenital, benign, or neoplastic disease. Each author has tried to cover the major disease processes in each of the general categories in which CT diagnosis is applicable. Specific technical details, including the method of scanning, contrast material, collimation, and slice thickness, are covered in each chapter. The interventional chapter extensively covers the various biopsy and therapeutic procedures in all the organ systems. Finally, the last chapter presents an up-to-the-minute coverage of current and recent developments in the CT literature and also provides extensive information about nuclear magnetic resonance (NMR) imaging. At this time we have had moderate experience with the NMR superconducting magnetic device produced by the Technicare Corporation and have formulated some initial opinions as to its role relative to CT and other imaging modalities. A concise discussion of the physics of NMR and a current clinical status report of the new modality are provided. We would like to thank all those people who have worked so diligently and faithfully for the preparation of this book. First, we are very grateful to our many contributors. For photography work, we are deeply indebted to Mr. Joseph P. Molter. For secretarial and organizational skills, we are indebted especially to Mrs. Mary Ann Reid and Mrs. Rayna Lipscomb. The editorial skills of Ms. B. Hami were invaluable in preparing the manuscript. Our extremely competent technical staff included Mr. Joseph Agee, Ms. Ginger Haddad, Mrs. E. Martinelle, Mr. Mark Clampett, Mrs. Mary Kralik, and numerous others. Finally, we are, of course, very appreciative of the support, patience, and encouragement of our wonderful families. In the Haaga family this includes Ellen, Elizabeth, Matthew, and Timothy, who provided the positive motivation and support for this book. Warm gratitude for unswerving support is also due to Rose, Sue, Lisa, Chris, Katie, Mary, and John Alfidi.

John R. Haaga Ralph J. Alfidi

F O R E WO R D I am honored and pleased to write the foreword for the 40th anniversary edition of CT and MRI of the Whole Body, a textbook of radiology edited by John Haaga, MD. The esteemed Dr. Haaga has been its editor since its first edition in 1976. Professor Haaga has received a Gold Medal from the American Roentgen Ray Society (ARRS), which is the society’s highest honor for distinguished service to radiology. He received this award for his contributions in leadership, teaching, and research, work that he continues to this day. His major innovations include CT-guided interventions that guide biopsies and treatments for abscess, nerve blocks, and cancer. Artifacts from his early work are archived at the Smithsonian Institution. At the age of 36, Dr. Haaga became a tenured professor in the Department of Radiology at Case Western Reserve University School of Medicine and University Hospitals of Cleveland. He later served for 14 years as department chairman at that institution. He has mentored and taught many hundreds of residents, fellows, and junior faculty members, many of whom have advanced to become distinguished names in their own fields. I am one such beneficiary, having had the fortune of being mentored and guided by John in my early years in radiology. His values and wisdom have allowed me to progress in the field of radiology. I have been inspired to write many books, such as Genitourinary Radiology: Kidney, Bladder and Urethra, The Pathologic Basis, published by Springer. I also became Editor-inChief of the Journal of Clinical Imaging Science, published by Wolters Kluwer. John is bigger than a film celebrity in India, which I witnessed firsthand when I had the opportunity to accompany him. Everywhere he lectured in India the venue was at capacity. He was literally mobbed by radiology residents wanting to take pictures with him. This book is considered the bible of radiology in India. Every aspiring radiology resident, as well as established faculty, reads this book. I am sure the story is no different in other countries. CT and MRI of the Whole Body is the largest selling radiology textbook in the world and has been translated into Spanish, Portuguese, and Persian. More than 25,000 copies have been sold in the world. I have been an integral part of this book for the past many years,

xx

as I have contributed a complete chapter on Imaging of the Kidney and have been Associate Editor for Genito-Urinary Radiology for its fifth edition. Dr. Haaga, a world renowned radiologist, anchors this book. He has assembled an impressive team of coauthors who are experts in various subspecialties of radiology. The book is divided into four parts: PART I—Principles of Computed Tomography and Magnetic Resonance Imaging, which is further subdivided into sections on Computed Tomography and Magnetic Resonance Imaging; PART II—CT and MR Imaging of the Whole Body, which has seven sections: Neuroradiological Imaging of the Brain and Meninges, Neuroradiological Imaging of the Head and Neck, Neuroradiological Imaging of the Spine, Imaging of the Chest, Imaging of the Abdominal and Pelvic Organs, Imaging of the Cardiovascular System, and Imaging of the Musculoskeletal System; PART III—Image-Guided Procedures; and PART IV—Leading Edge Imaging Concepts. The sixth edition of Haaga’s CT and MRI of the Whole Body is a welcome addition. The book is well organized, the chapters are uniformly structured, and each section is richly illustrated. The quality of the paper, printing, illustrations, and figures is excellent. There is ample use of line drawings to explain difficult concepts. The sixth edition of Haaga’s CT and MRI of the Whole Body is a welcome addition. I am not aware of any other radiology book in the market that is as complete and uniquely illustrated. The cost is reasonable considering the number of illustrations and the quality of work. This book meets the needs of its intended audience. The current edition is completely up to date in the field of diagnostic radiology. It is a must have for radiology residents, radiology faculty, radiology libraries, and serious academic radiologists.

Vikram S. Dogra, MD, FSAR, FESUR, FAIUM, FSRU Professor of Radiology, Urology & Biomedical Engineering Associate Chair of Education and Research Department of Radiology University of Rochester Rochester, New York

F O R E WO R D Very rarely does a textbook fulfill the promise of being both comprehensive and readable. CT and MRI of the Whole Body, edited by Drs. John Haaga and Daniel Boll, not only fulfills that promise, but has continued a legacy of success that has been sustained over multiple editions. It is an honor to have the opportunity to write this foreword. Having trained under the expert eye of Dr. John Haaga, I studied the earliest editions of this book, providing my introduction into the intricacies of cross-sectional imaging. I then had the privilege of watching this true pioneer of CT and MRI take his personal experience and use it to create a resource that has evolved as the field has matured over the past four decades. The current edition of this text covers the wide breadth of CT and MRI anatomically, and the editors have enlisted experts in neuroradiology, head and neck imaging, abdominal imaging,

thoracic imaging, musculoskeletal radiology, physics, and a long list of other subspecialties in our field to create a textbook that is as deep as it is broad. With the addition of the energy, talent, and creativity that Dr. Daniel Boll brings to this edition, the result is truly remarkable. I have no doubt that you will agree as you dig into this text, and learn from the combined wisdom of a team of true expert authors, as guided by the steady hands of Drs. Haaga and Boll. Enjoy!

Jonathan S. Lewin, MD, FACR Executive Vice President for Health Affairs, Emory University Executive Director, Woodruff Health Sciences Center President, CEO, and Chairman of the Board, Emory Healthcare Atlanta, Georgia

xix

1  Imaging Principles in Computed Tomography David W. Jordan and John R. Haaga

Computed tomography (CT) has grown quickly from an innovative specialized tool to a mainstay of medicine in every healthcare setting. Hospitals, outpatient clinics, and physician offices find CT to be an essential tool for patient diagnosis and management. Every year brings new advances in CT technology and applications. This chapter is dedicated to the basic principles and physics of CT operation. The practicing radiologist who can evaluate new technology with a good understanding of these principles will be successful in choosing the best tools for his or her practice and in getting the greatest benefit from them.

THE CT IMAGE Radiography reduces a three-dimensional (3D) body part to a twodimensional (2D) image with limited contrast, because structures that lie on top of one another are projected onto a single image. The contrast can be improved by using exogenous agents to enhance certain structures or by taking extra projections from different angles to separate the structures in the image plane, but there are limits to how well this can work. The advantage of CT is the improvement in image contrast that comes from using a 2D image to show an almost-2D section of the patient without the effect of overlapping structures. The CT image is a cross-sectional view of the patient rather than an x-ray shadow of the beam passing through the body part (Fig. 1-1). An x-ray beam is used to collect information about the tissues, but the image is not an ordinary projection view from the perspective of the x-ray tube looking toward the film or detector. The image is a crosssectional map of the x-ray attenuation of different tissues within the patient. The typical CT scan generates a transaxial image oriented in the anatomic plane of the transverse dimension of the anatomy. Reconstruction of the final image can be reformatted to provide sagittal or coronal images; these are viewed from the same perspective as a digital radiograph, but they show thin slices of tissue rather than superimposed tissues and structures. The pixel values show how strongly the tissue attenuates the scanner’s x-ray beam compared to the attenuation of the same x-ray beam by water. The CT image is produced by the process of reconstruction: digitally combining information from x-ray projections through the patient from many different angles to produce the cross-sectional image. Because the image is digital, it is made up of a group of pixels (shortened from “picture elements”). Each pixel has a grayscale value that is displayed to the viewer. The image is 2D, but it represents a 3D volume of tissue with a finite thickness (usually a very small thickness compared to the field-of-view (FOV) size [≈2-5 mm]). Each pixel is the projection, or 2D representation, of the x-ray attenuation of a voxel (shortened from “volume element”) of physical tissue. The size of the pixels and the thickness of the voxels relate to some important image quality features, such as detail, noise, contrast, accuracy of the attenu-

ation measurement (CT number value), and artifacts. These will be discussed in more detail as they relate to the processes of acquiring and reconstructing CT data.

CT ACQUISITION OVERVIEW The basic process of collecting data in CT is illustrated in Figure 1-2. In a CT of a single section of tissue using a single detector, the x-ray beam is collimated to the desired image thickness. The detector array has a number of individual detector elements that each record the intensity of the beam passing through the tissue along the path from the x-ray tube to the element. The system captures a simple projection x-ray through the patient, consisting of a thin strip or row of pixels. It can be thought of as a one-dimensional (1D) radiograph. The scanner then rotates the source and detector to capture additional 1D “strip x-rays” through the same section of the patient, viewed from a number of angles. Each strip radiograph (projection) is stored in the computer memory for later reconstruction. In multislice CT (Fig. 1-3) this operation is performed simultaneously for many arrays of detectors stacked side by side along the z-axis (long axis) of the patient. The x-ray beam collimators can be opened so that a wider section of the patient is irradiated, and each row of detectors can measure a separate transmission signal for the tissue section that lies between the detector row and the tube. The width of tissue that is sampled by each detector row is determined by the physical width of the detector elements along the z-axis. CT images and individual rotations of the scanner gantry are often called slices because a single data acquisition and reconstruction produces an x-ray map of a thin section of the patient’s body. The tissue displayed in the image represents the same tissue as if a thin slice or section of the patient’s body were cut in a plane perpendicular to the long axis (superior-inferior) of the patient’s body and fixed for viewing. Early CT scanners collimated the width of the x-ray beam to the width of a slice (e.g., 5 mm), irradiated one slice at a time, and collected data for transmission through one slice at a time. Multislice scanners use a wider beam to cover more tissue with each pass of the tube, and the detectors contain arrays that are arranged to collect data for multiple individual adjacent slices at the same time. The detector element(s) for each slice store data separately and represent different physical sections of tissue. The term slice can be confusing because in common use it can refer to several different things. When a CT scanner is called a “16-slice” model, it usually means that the scanner can acquire up to 16 individual detector data sets at one time. It is more precise to call this a 16–detector row CT scanner. The individual data elements acquired should be referred to as acquired images or acquired projection data. In a multidetector-row scanner, the data are often acquired in thicknesses of 0.5 to 0.7 mm and reconstructed in image thicknesses of 3 to 5 mm.

3

4

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging Y

XRT

Z X

FIG 1-1  A CT image represents a cross section of the imaged subject rather than the x-ray shadow of the anatomy, as in a conventional radiograph. a

b

c

d

FIG 1-3  In multislice CT, several independent detectors arranged side by side sample data from unique locations within the x-ray beam.

l (x-rays/cm2)

l0 l l0 L l = l0e −µL

L (cm)

FIG 1-4  Exponential attenuation of x-ray beam intensity in an absorber, FIG 1-2  A simple CT scan produces a one-dimensional strip radiograph

such as tissue.

for each projection through the patient.

The images that are presented to the operator or physician for viewing should be called reconstructed images, although many people continue to refer to these as slices.

Linear Attenuation and Projections The fundamental x-ray physics of CT are the same as those for radiography. A source of ionizing radiation is transmitted through an object to recreate an image of the object based on its x-ray absorption (Fig. 1-4). The intensity of the transmitted radiation beam is given by the equation: I = I0e− µx where: I0 is the incident intensity of an x-ray beam on the surface of an object of thickness x

I is the transmitted intensity e is Euler’s constant (2.718) µ is the linear attenuation coefficient Attenuation of an x-ray beam with a particular spectrum depends on two distinct properties of the tissue: the atomic number and the density of the attenuating material. When collecting one projection, it is not possible to determine what combination of atomic number and density resulted in the measured attenuation. Thus both very dense materials (e.g., bone) and materials with a high atomic number (e.g., iodine contrast media) produce strong attenuation. Both would have a similar appearance on CT even though they have very different physical properties. The key advantage of dual-energy and spectral CT techniques (discussed later) is that they can be used to probe the attenuation arising from density and atomic number separately by making two different measurements of the same sample, object, or body part.

CHAPTER 1  Imaging Principles in Computed Tomography The transmission of the beam (the ratio I:I0) determines the signal that is collected by the CT scanner detectors at each projection angle, and the attenuation differences along different paths through the patient’s body result in the patterns that are formed in the data for each projection acquisition.

CT Reconstruction and Image Display Overview Once the 1D projection radiographs have been collected, they are transformed into the cross-sectional image to be displayed to the user. There are several approaches to image reconstruction, broadly classified as filtered backprojection (FBP) and iterative reconstruction (IR) techniques. All of these methods use computerized mathematical operations to combine the 1D projection attenuation information into 2D maps of the attenuation of the scanned subject.

CT Number Scale The reconstructed CT image is a map of tissue attenuation, and these maps are quantitative owing to the use of a scale called the Hounsfield unit (HU) scale. The scale is named for the inventor of the x-ray CT scanner, Sir Godfrey Hounsfield. The HU scale is a relative scale for linear attenuation of the tissues. It is valuable to quantify the pixel values so that the physician can compare the composition of one tissue with another. The HU scale is based on relating all measured attenuation values to the attenuation of water. A CT numbering system that relates a CT number to the linear attenuation coefficients of x-rays in different absorbers compared to water is given by: CT Number =

K (µ − µw ) µw

where: µw equals the attenuation coefficient of water µ is the attenuation coefficient of the pixel in question When K = 1000, the CT number scale is the HU scale. Other K values can be used to change the scaling of the entire system. When the HU scale is used, air has a value of −1000 HU, water has a value of 0 HU, and dense bone has a value of +1000 HU. The advantage of the HU scale is that density differences of 1 part in 1000 (0.1%) can be represented by distinct values. The inherent density resolution of CT scanners is about 0.5%, so the HU scale is sufficient to display all attenuation differences the scanner can measure. Increasing the value of K would not improve on the density resolution of the system. As a general rule, the human eye cannot appreciate contrast differences of less than about 10%, whereas CT scanners can easily demonstrate differences of less than 1%. Thus the small density-resolution difference measured by the CT scanner must be exaggerated for viewing. This is accomplished by enabling the user to select a small range of gray levels from the entire CT number scale and to reset the black and white limits. This is done automatically for many scanners by presetting an appropriate “window/level” setting to be used for display; this setting can be modified by the user, both on the scanner and when the image is displayed on other workstations and picture archiving and communication systems (PACS).

GENERATION OF A CT IMAGE Components of a Data Acquisition System Gantry.  Modern CT systems use slip ring technology to permit the scan frame to rotate continuously for spiral or helical CT scanning. Slip rings are used to transmit power and some control signals to the system components mounted on the gantry rotor. Systems with many

5

detector elements and data input channels use broadband wireless communication to transmit collected image data from the detector assembly to a receiver on the stationary base of the scanner. To enable high-speed scanning with good temporal resolution, CT system developers have focused on providing high gantry rotation speeds. For high-performance systems with mechanical roller gantry bearings, gantry rotation times of 0.33 seconds are achievable.63 Operating modes are available to acquire data over a partial arc (<360 degrees) to shorten the apparent rotation time further. Air bearing technology52 enables even faster rotation speeds. Scanners with air bearing gantry designs can achieve rotation times of 0.27 seconds.20 Such rotation speeds are desirable to achieve cardiac imaging without using β-blockers to slow patients’ heart rates. Together with largecoverage detector arrays, subsecond scan times permit rapid acquisition of contiguous anatomic areas for volume data acquisitions.

X-Ray Generators and Tubes.  High-frequency x-ray generators are required to produce intense x-ray beams for high-speed scanning. Typical operating frequencies range from 20 to 50 kilohertz (kHz).32 Generator power ratings of 80 to 120 kilowatts (kW) are common, as shown in the manufacturer specification charts for commercial systems available in the United States. Most generators support a wide range of exposure techniques from 80 to 140 kilovolts (kV) and 30 to 500 milliamperes (mA). Some generators have a lower tube potential option (70 kVp) available as an approach to radiation dose reduction. Generators are located on the rotating scan frame within the CT gantry, together with the x-ray tube and detector assembly, and are powered by a low-voltage power supply provided via the slip ring. Modern generators have a number of features to enable them to adjust and optimize radiation output and image characteristics in real time while scanning. The generator design is computerized and flexible to allow for computer-controlled x-ray techniques generated by the scanner’s master control computer. Modulation of tube current based on patient thickness is a standard feature. Scanners may have the ability to vary the tube current along the z-axis of the patient, in the x-y plane, or both. Some systems allow dynamic or automated selection of kVp based on patient size and/or study type.10,37,49,62 For dual-energy (spectral) imaging, one approach that relies on a high-performance generator is fast kV switching.29 The generators for these systems can toggle rapidly (≈150 msec) between two operating voltages (usually 80 kVp and 140 kVp). Modern CT scanner x-ray tubes have rotating anodes and use innovative cooling methods to continuous high x-ray output. Many tubes feature focal spot modulation or “flying focal spot” technology.14,15,35,57 These tubes, in addition to providing large and small focal spots, have the ability to rapidly shift the position of the selected focal spot during the scan. This has the effect of changing the sampling behavior of the scanner to enhance the resolution of the image data. Tube ratings for typical modern CT scanners can have 30 million heat units (30 MHU) of heat storage capability and a continuous output power capability of 120 kW.

Filters.  CT scanners have x-ray filters to shape both the geometry and spectrum of the beam. The standard filter is made of aluminum, like a radiographic tube filter, but there are important differences in the filter design. CT filters are generally thicker than radiographic filters and often combine other materials (e.g., Teflon, copper) with aluminum. The CT beam is much “harder” than a radiographic beam at the same kVp—it contains fewer low-energy photons and the average energy is much higher (Fig. 1-5). This is done to reduce patient radiation dose (because low-energy photons are more likely to deliver dose to the patient without contributing to the image). In radiography the

6

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

Blue = 100 keV Red = 50 keV Gray slab = 1 HVL thickness for this beam Red photons twice as likely to be absorbed as blue photons owing to energy difference

Incident: Intensity = 6 photons Average energy = 75 keV

Exiting: Intensity = 3 photons Average energy = 83 keV

FIG 1-5  Beam-hardening example. A beam composed of different photon energies passing through one half-value layer of absorber will have a lower intensity (fewer photons) and a higher average energy in the transmitted beam, because the low-energy photons are more easily absorbed and the high-energy photons are better able to penetrate the absorber.

lower-energy photons help provide contrast to the image, but CT has much higher inherent contrast sensitivity than radiography, so a useful image can be obtained without the lower-energy photons. The compensating filters used in CT are often called bowtie filters59 because of their shape. The filter is thinnest at the central ray of the beam and thicker toward the azimuthal edges of the fan beam. Thus the thinnest part of the filter is in line with the thickest path length through the patient’s body as the gantry rotates, and the thicker part of the filter is in line with the thinner peripheral anatomy (peripheral). The filter partially compensates for the patient’s body thickness contour. The result is that for any line across the FOV, the total attenuation and beam hardening along the x-ray path is much more uniform than if the beam only passed through the patient. Compensating for the attenuation differences allows the scanner to operate with a smaller dynamic range at the detectors, and compensating for the beam hardening improves the scanner’s quantitative accuracy in representing tissue attenuation as CT number values (HUs). A scanner will have at least two different bowtie filters to provide compensation for head and body scans, and some scanners have as many as five or six different filters to match a variety of patient body sizes. The bowtie filter is selected by the scanner and switched into the beam path automatically based on the scan setup chosen by the operator.

Detectors.  CT scanners use scintillation crystals to convert x-rays to visible light pulses, which in turn can be detected and counted by electronic circuits. Conventional detectors produce light pulses and voltage pulses, resulting in a simple count of x-ray photons detected. Research is underway to develop photon counting devices that can detect the number of x-ray photons as well as information about the energy of each individual photon. CT detectors use crystals composed of fast scintillator ceramics such as cadmium telluride28 and gadolinium oxysulfide.64 Development of new crystal materials is an active area of research to further improve scanner performance, and each CT manufacturer has newer proprietary crystal types available or in development. The crystals convert x-rays to light pulses whose intensity is proportional to the total energy of photons striking the crystal. The light is converted to current and hence a voltage across a resistance, which in turn is detected as digital data for use by the reconstruction system. This multistep process of analog signal conversion to digital data produces

a number of errors and inefficiencies related to quantization error, sampling and recovery speed, and electronic noise.21 Despite these inefficiencies, even primitive CT devices can measure low-level signals accurately over a wide dynamic range with great precision, permitting reliable imaging of diverse density extremes such as bone, air, and water. Detector arrays continue to grow as individual detector elements remain small, providing small pixel size with ever-greater anatomic coverage. Common detector arrays accommodate scanning of 80 to 160 mm of coverage in a single tube rotation, with resolution of 0.5 to 0.6 mm, so that entire sections of patient anatomy can be captured in a single revolution of the CT scanner’s rotating frame, taking less than 1 second. Because of the fixed position of the tube/detector combination, the associated detectors and electronics system must be stable, uniform, and capable of tracking over a wide dynamic range. Noncompensation for errors as small as a few thousandths of 1% can cause visible (ring) artifacts.34 In addition, spatial resolution, a key image quality parameter of CT (see later), can be limited by the data sampling. In this geometric relationship, the number of detectors in the arc or array determines data sampling; that is, spatial resolution depends on how closely spaced the rays are in each view. The typical number of detectors in the array ranges from about 600 to more than 900. These numbers result in limiting spatial resolution of approximately 5 to 10 line pairs per centimeter (lp/cm), depending on the reconstruction type (filter), reconstruction matrix size, and scan FOV. Sampling theory requires that to visualize a particular object, the sampling frequency must be at least twice the spatial frequency associated with the object.26 In practice even higher sampling rates are desirable. Axial resolution of greater than 5 to 10 lp/cm (0.5-1.0 mm) may be limiting for specific clinical applications.

Data Acquisition Systems Data acquisition systems (DAS) are the heart of CT scanners. These systems perform detector signal analog-to-digital conversion and data preprocessing for the scanner’s reconstruction system. Steps performed by the DAS include conversion of electric current (from the detector)  to voltage, digitization of voltage pulses, subtraction of background offset signal, logarithmic conversion of data, and transmission of data to the preprocessing system. In modern systems, these functions are 

CHAPTER 1  Imaging Principles in Computed Tomography

7

integrated and combine multiple steps on specialized hardware— boards, chips, or modules integrated into detector modules that can be housed on the detector assembly within the rotating scan frame. An ideal DAS would accurately and consistently digitize the input signal from the detectors with no noise. The inherent electronic noise produced within the DAS must be small enough that the total noise in the sampled signal depends mostly on the x-ray quantum noise. The DAS must also have a wide dynamic range and be as linear as possible across the range of possible input signals so that the x-ray beam intensity can be accurately determined for the range of available x-ray outputs as well as a wide variety of attenuation types (tissues) and thicknesses. A DAS will have some dark current that can vary with temperature and over time, affecting the precision (reproducibility) of the detector signal measurement. A low dark current that is stable with ambient temperature changes is desirable. Crosstalk limits the DAS performance when residual signal from one sample is collected as part of the next sample. A high-performance DAS must be capable of high sampling rates with low crosstalk.21

Scanner Operation and Data Acquisition

Rotate/Rotate Scanner Geometry.  During the development of CT, several combinations of x-ray tube and detector movement were used to collect data. The historical significance of each is described in previous editions of this book. The rotate/rotate geometry emerged as the best practical design choice for high-speed scanning in multidetectorrow systems and is the geometry used currently in modern multidetector CT systems. In a rotate/rotate system the x-ray tube and detector array are mounted opposite each other on the rotating scan frame as shown in Figure 1-2. A wide-angle fan-beam geometry (50-55 degrees) is used, and an arc of detectors and the x-ray tube rotate continuously around the patient for 360 degrees. Projection data (or data samples) are obtained at various angles, and for each angular sample taken at a position of the tube and detector, a view is created. In rotate/rotate geometry, the detectors are fixed radially and do not view the scan area uniformly. Only the center detectors in the array “see” the pixels at the center of the FOV. However, this fixed relationship allows the detectors to be highly collimated, which greatly reduces detection of scattered radiation; this maintains contrast in the image so that the viewer will be able to detect objects against a higher background of image noise (or mottle, or grainy background). To overcome the limitations of detector size on spatial resolution, manufacturers have developed various methods that exploit the scanner geometry. One method of increasing spatial sampling (and spatial resolution) is to offset the detector array relative to the center of rotation by a quarter-ray or eighth-ray. This allows data collected during the second half of a 360-degree rotation to be interleaved with that of the first half, effectively doubling the sampling and resolution. Patient motion can reduce this benefit. In addition, this method is limited in spiral CT scanning to certain modes in which data from projections acquired through 360 degrees are available for interleaving. Another effective method of increasing data sampling is to oscillate the x-ray tube focal spot a few millimeters during a scan (Fig. 1-6) to achieve interleaved data samples. The position of the focal spot is accurately controlled and synchronized with rotational speed and position of the rotating frame. Data acquired during the use of each focal spot position are stored separately and then interleaved to provide greater spatial sampling of the tissue. This approach can be used to increase sampling within the scan plane as shown in Figure 1-6, and can also be used to provide increased sampling along the z-axis.14,35 When focal spot modulation is done along the z-axis, the finer

FIG 1-6  The dynamic or “flying” focal spot technique is used to provide finer effective spatial sampling of tissue.

sampling can be used to produce sagittal and coronal reformatted images with smaller pixel sizes than would be possible with the physical detector elements alone. Systems with this capability are often marketed as having more “slices” than the number of actual detector rows available on the scanner; for example, a scanner with 32 detector rows may be marketed as “64 slice” or a scanner with 128 detector rows may be marketed as “256 slice.” Spatial resolution requires adequate view sampling to prevent aliasing artifacts. Views (angular projections) are limited by the electronic measurement capability of the scanner, which is typically not the gating factor for maximum resolution. These factors vary for every CT manufacturer and can dramatically affect the overall performance of a scanner.

Spiral (Helical) Scanning.  Until the late 1980s, CT scanners acquired data in discrete slices of patient anatomy in a method commonly called axial scanning. In axial (cross-sectional) CT, each revolution of the x-ray tube around the patient produces a single data set (slice). During data collection, the patient table is motionless. To create an additional slice, the table is translated by a given amount and the x-ray tube rotates around the patient again. Each rotation produces a single image. The advent of slip-ring technology and new data reconstruction techniques enabled spiral, or helical, CT. In spiral/helical CT, the patient table translates through the gantry while the scan frame rotates continuously around the patient, sampling data as a volume. This permits new options in reconstruction. For example, once a volume of data is collected, an image can be reconstructed at any point along the effective path traced by the x-ray tube. There are several key terms associated with helical scanning: Acquisition is the entire volume of data collected during a continuous spiral scan. Revolutions refer to the number of 360-degree rotations of the x-ray tube during a single acquisition. Pitch is the distance the couch travels during one 360-degree revolution of the x-ray tube divided by a length associated with the x-ray beam and/or data acquisition. The most common value

8

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

used (as in the International Electrotechnical Commission [IEC] definition) is the width of the collimated fan beam along the z-axis of the scanner. Interpolation is a reconstruction method (most accurately described as “deconvolution”) that permits the realignment of spiral/ helical scan data for reconstruction of an axial (cross-sectional) slice. Spiral CT has many advantages over conventional or axial CT, including: • The ability to minimize motion artifacts owing to faster/shorter acquisitions • Decreased incidence of misregistration between consecutive axial slices • Reduced patient dose • Improved spatial resolution in the z-axis • Enhanced multiplanar or 3D renderings Motion artifacts are minimized in spiral CT because of the faster scan times associated with the examination (versus step-and-shoot axial scanning); many spiral CT examinations can be completed in a single patient breath hold. The faster scan times associated with spiral CT also permit accurate slice-to-slice registration and eliminate the misregistration associated with inconsistent breath holds in conventional axial scanning. Reconstructions can be overlapped or ideally positioned for accurate visualization of small lesions. Spiral scanners achieve reduced patient dose by extending the pitch factor for a given study. For example, consider a typical thorax, abdomen, or pelvis examination. Typical slice widths range from 2 to 5 mm. With spiral CT, the same examination with the same slice thickness as axial CT can be completed with 50% greater table speed using a pitch of 1.5; this results in a 33% reduction in patient dose (Fig. 1-7). The downside to this approach is a slight broadening of the slice sensitivity profile (effective axial slice thickness) based on the type of spiral reconstruction chosen. This limits the spatial resolution along the z-axis, which can reduce the detail available in sagittal or coronal reformatted images. With continuous collection of data, spiral CT enables a reduction in image-to-image separation, thus improving spatial resolution in the z-axis if the data are oversampled (pitch < 1.0). Although axial CT might overcome this challenge by overlapping consecutive axial acquisitions, this would lengthen scan time needed to cover a given region of patient anatomy more than doing so in a helical acquisition. Spiral CT data acquisition systems must be capable of collecting large volumes of data at extremely fast data rates. Data sampling for

33% dose reduction

50% greater coverage

FIG 1-7  Increasing pitch decreases radiation dose and increases the scan coverage that can be achieved in a given scan time, using the same rotation speed and number of rotations.

most systems is in the range of 3 kHz and can provide up to nearly 2400 projections or views per rotation of the gantry. Data rates for many new systems exceed 200 megabits/second. This extremely fast sampling capability enables high spatial resolution and enhanced z-axis sampling of patient anatomy. The increased speed of data acquisition systems and a number of other technologic developments paved the way for data acquisition at subsecond rotation speeds of the rotating frame. Subsecond spiral data acquisition permits temporal resolution on the order of 220 to 300 milliseconds, depending on the scanner manufacturer. These temporal resolutions enable more examinations to be scanned within a single breath hold and increase the range of anatomy that can be covered in a single acquisition. This improves the performance of clinical applications such as CT angiography, which requires precise timing of a contrast bolus injection and rapid acquisition during the arterial phase. Fast data acquisition can also improve the z-axis resolution for a given study by covering the same amount of anatomy in the same time with thinner slices and/or with oversampling at low pitch.

Multidetector-Row Designs.  Multidetector-row scanners are motivated by the desire to cover large regions of anatomy in short times with fine sampling along the z-axis of the patient. In early scanners with only a single data acquisition channel (single slice), the beam width directly determined the coverage as well as the image thickness (spatial resolution). With the development of fast, stable, compact electronics, it became possible to build a DAS that could simultaneously acquire and store multiple images. By arranging multiple narrow detector rows side by side along the z-axis, connecting them to a DAS with multiple input channels, and increasing the x-ray collimation width to cover the entire assembly, scanner designers provided users with more flexibility to achieve faster scans or finer spatial detail. The first deployment of this technology included a dual-arc detector as shown in Figure 1-8. In this configuration, two parallel arcs of detectors are used to acquire data simultaneously during a single revolution of the scan frame, dividing the total x-ray beam into two equal beams defined by the detector aperture of each row of detectors. Considering small gaps between rows, which can be ignored from the perspective of data acquisition, the total x-ray beam collimation now becomes the sum of detector row collimations (Fig. 1-9). This concept can be extended to for systems with multiple rows of detectors. Early multidetector-row systems were commonly equipped with a four-channel DAS. Such systems were often referred to as “four-slice” CT scanners because they could collect up to four images simultaneously in a single rotation of the scan frame. A user upgrading to such a system from a single-detector system would have the flexibility when imaging 5-mm slices to increase acquisition speed or to use finer spatial sampling in the z direction. Configuring the scanner to simultaneously acquire four 5-mm images in each rotation using a 20-mm beam collimation would reduce the acquisition time by a factor of 4. Alternatively, sampling the data using four 1.25-mm detectors and reconstructing 5-mm images would accomplish the scan in the same period of time, but the data available for sagittal and coronal reconstructions would have finer sampling and thus a smaller pixel dimension in the z direction. A possible scheme of detector configurations for such a four-channel system is shown in Figure 1-10. This tradeoff between coverage and z-sampling affects systems with 2 to 16 detector rows, and such systems were designed with different available combinations of physical detector element sizes, DAS input configurations, and x-ray beam collimation widths to provide clinically useful choices. Two detector designs have been deployed by CT manufacturers, unequal width and equal width, referring to the sizes of the individual detector-row elements across the z dimension of the

CHAPTER 1  Imaging Principles in Computed Tomography

Intensity

Slice A

9

Slice B

5 mm

z

FIG 1-8  Early multislice detector system with dual-row detector. (Courtesy J. S. Arenson.)

X-ray tube focal spot Prepatient collimation

Postpatient collimation

A

B

Single-detector arc

Dual-detector arc

FIG 1-9  A, Arrangement of a dual-row detector array along the z-axis of the patient. B, Total x-ray beam width (source collimation) is the sum of the consecutive detector row widths along the z-axis.

detector array. In systems with unequal-width detector designs, the center of the detector array contains smaller detector elements that can be connected individually to the DAS for fine z-sampling with individual detector channel widths such as 0.5 mm, 0.625 mm, or 1.25 mm. The outer elements are larger and are used for wide coverage

applications; in such configurations, several small center elements are electronically combined into one DAS channel. For example, one DAS channel could connect to a single outer-row element with 2.5-mm width, while another DAS channel could collect data from four 0.625 elements acting as a single 2.5-mm element. In this arrangement, the

10

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging 4 x 1.25 1.25 1.25 1.25 1.25

Not used

Not used

4 x 2.5 Not used

2.5

2.5

2.5

2.5

Not used

4 x 3.75 Not used

Not used 4x5

FIG 1-10  Matrix detector with 4-channel DAS. Electronic combination of individual 1.25-mm elements produces slice width configurations of 4 × 1.25 mm, 4 × 2.5 mm, 4 × 3.75 mm, or 4 × 5 mm.

data from the four 0.625-mm elements cannot be recovered as individual z-samples, because they are combined at the input to the DAS, so the effective z-sampling is based on 2.5-mm channel widths for this scan. An alternative design is the equal-width design, where all of the detector elements have the same physical width and are connected to individual DAS channels or grouped in combination depending on the application and the needed resolution and coverage. For systems with more than 16 detector channels, it is possible for most applications to use an equal-width detector with submillimeter elements (0.5 mm or 0.625 mm are common) and provide isotropic z-sampling across the entire detector array with wide coverage. For example, a 64-detectorrow system with 0.625 mm detectors can sample 40 mm of coverage with each scanner rotation while providing a separate DAS channel for each individual detector row. Multirow detector arrays have rapidly progressed from having 2 or 4 detector rows to 64, 128, 256, and up to 320 rows in commercial CT systems. This is possible because the miniaturization of electronics enables more detector and DAS components to be installed in a detector assembly that can be practically attached to the rotating scan frame, given its size and weight. Systems with many detector rows also require advanced low-power-consumption electronics or larger power supplies to support the detector and DAS; practical limitations on these systems include providing adequate cooling for the electronics within the scanner housing. These systems use a wireless link to transfer data from the DAS on the rotating frame to the scanner’s reconstruction system, so capturing many channels of data simultaneously with high sampling rates requires a high-throughput wireless link. Advances in high-speed wireless data technology have enabled the 256- to 320channel configurations in use in current systems, and it is possible that still larger numbers of data channels will be used the future.

CT TECHNIQUE FACTORS CT technique factors are the settings that are selected by the operator at the scanner’s control panel. Each has a specific effect on radiation dose to the patient and the characteristics of the scanned images. The technique factors for a scan describe the x-ray beam parameters and the geometry of the scan. The x-ray technique factors in CT are familiar from conventional radiography. The kilovoltage (kV) applied to the x-ray tube and the tube current (mA) are selected by the operator. The exposure time is

usually chosen to be the time of one complete 360-degree rotation of the rotating scan frame, so the speed of the rotation is adjusted for each scan to achieve the desired exposure time. These typically range from 0.4 to 1.0 seconds, although shorter times and faster speeds are possible. The selection of a helical (spiral) or axial (step-and-shoot) scanning mode is a key technique factor for the scanning geometry. The collimation of the x-ray beam along the z-axis is normally not selected directly by the operator, but rather is chosen by the scanner to match the detector configuration. Detector configuration refers to the size (length along the z-axis) of individual data acquisition channels and the total number of such channels or rows along the z-axis. For example, 64 rows of 0.625-mm detectors (N = 64, T = 0.625 mm) would be used with a total collimated beam width of 40 mm. On some systems the user may select the detailed detector configuration to achieve a certain degree of z-sampling/spatial resolution for multiplanar reconstructions. The scanner software will often choose a detector configuration automatically based on the selected reconstructed image thickness. The pitch is a geometric technique factor in helical scanning. The coverage and scanning speed can be determined from N × T, rotation time, and pitch (for a helical scan). In this way, faster rotation speeds and wider beams covering greater z extent of the patient are used to achieve faster scans of a given anatomic region. The scan FOV is chosen to approximately match the body size being scanned. On some scanners there are only two modes (Head and Body); other scanners have an array of SFOV size choices to contour to various patient sizes and body parts more closely. Automatic exposure control and tube current modulation are particularly critical for radiation dose management in modern CT. These technologies usually have a few parameters and settings to control how they operate. Use of these tools and the settings used make up part of the technique factors. Although not technically part of technique factors, reconstruction, postprocessing, and display techniques all have a dramatic  impact on the quality and appearance of the images that are eventually produced.

Automatic Tube Current Selection and Modulation In an ideal scenario, dose and image quality would be optimized by delivering the same x-ray beam flux to the detector for every projection

CHAPTER 1  Imaging Principles in Computed Tomography through the patient. Of course the density and width of a patient’s body can vary greatly within a region of interest, especially when scanning an entire body region (or multiple regions). Using a single fixed mAs value does not achieve the needed optimization; some anatomy will necessarily be depicted with inadequate image quality or exposed to excess radiation in this approach. Automated tube current modulation (ATCM)30 is a tool to address this challenge. Ideally ATCM can be used to optimize the scanning parameters in real time so that each patient is exposed to just enough radiation to produce a diagnostic-quality image—and no more. Automating this process should remove variation among users and the possibility of user error. The technology can be subdivided into three related techniques: automatic tube current selection, varying the tube current along the z-axis of the patient, and varying the tube current in the x-y plane around the patient. Automatic tube current selection1,31,36,40,47,56 (ATCS) is usually done by analyzing the scout/topogram image or images to estimate  the attenuation of the patient and select an appropriate mAs value.  The specific implementation on any given scanner may require a posteroanterior/anteroposterior (PA/AP) scout image, a lateral scout image, or both. Given the differences in patient dimension and density along a frontal versus a lateral projection, using both images should produce the most accurate choice of the correct mAs. In real implementations, most single-scout tube current modulation schemes are accurate enough. The attenuation of the patient as measured from the scout image can be measured and expressed as the water-equivalent diameter (WED) of the patient.39 The WED is the diameter of a water-filled cylinder that would produce the same x-ray attenuation as the patient cross-section. This value can be very different from the physical dimension of the patient, especially in regions containing large areas of low attenuation (chest) or high attenuation (pelvis). Once the patient’s WED is determined by the scanner, the scanner chooses an mAs value based on the image quality parameter chosen by the user. In some cases this is done by specifying a target image noise value; the scanner uses calculation or lookup tables to determine what mAs value will produce the target noise for the chosen kVp and the estimated patient WED. In other cases, this is done by specifying a target or reference mAs value. In the reference mAs implementation, the reference mAs value actually represents the amount of image noise produced when a reference WED (e.g., 24 cm) is scanned with the specified mAs. Thus when a user specifies a reference mAs value of 200, what they are actually doing is asking the scanner to use an mAs value to scan the patient that will produce the same amount of noise that 200 mAs produces in the reference WED. When using ATCS it is important to understand the specific functions as implemented by the scanner manufacturer. There are differences among commercial implementations in how the scout images are used to determine the estimated patient WED. Some scanners use the PA scout, some use the lateral, some use both, some use the first scout acquired during the examination, and some default to a fixed mAs if the expected projection (i.e., PA) is not performed first. This can lead to important workflow considerations for the technologist performing the scan. It is also important to know when designing and setting up scan protocols and technique factors on the scanner. Z-axis tube current modulation (zTCM) is closely related to ATCS and is used to vary the tube current during each rotation of the tube, based on the different attenuation properties of different body regions. This is usually done using the estimated WED determined from the scout image. Once the baseline tube current has been established, either using automated selection or manually, zTCM is used to increase or decrease the tube current during tube rotations covering portions

11

of the body with higher or lower attenuation. The simplest example is a scan of the chest, abdomen, and pelvis. The manually or automatically selected baseline tube current would likely be the value used to scan the abdomen, whereas a reduced current would be used in the less-dense chest and a higher current would be used to achieve good beam penetration through the dense bones in the pelvis region. X-Y, or azimuthal, tube current modulation (aTCM) allows a further degree of fine-tuning of the radiation output to match the projection attenuation. Once the tube current is chosen to achieve a target noise level and varied based on the general density of the body region, there is still a large difference in attenuation between lateral projections and AP/PA projections. This difference is pronounced for most of the chest, abdomen, and pelvis when patients are imaged in supine or prone positions, and is especially pronounced in the shoulders and hips because of the dense bones that are superimposed in lateral projections. For scanners capable of varying the current within a single tube rotation, a modulation scheme may be available to reduce the current for AP and PA projections and increase it for lateral projections. In a variation of this approach, azimuthal modulation can be used to target exposure reduction to certain organs. For example, to reduce radiation dose to the eyes during a head scan or to the breasts during a chest scan, the azimuthal modulation can be modified to further reduce the mAs during the AP projection and compensate with increasing mAs for the lateral and/or PA projections. In practice these three techniques may be available as separate options or activated together as part of an automated scanning mode, depending on the manufacturer’s implementation. Some manufacturers have extended this concept further by providing organ-specific boost functions (to fine-tube target noise values in specific anatomic regions) and enabling the operator to automate the selection of tube kV based on the patient attenuation determined from the scout image(s).

Dual-Energy/Spectral Acquisition Transmission measurements in conventional CT are a physical measure of the linear attenuation properties of tissue, as described earlier. Linear attenuation depends on two underlying properties—density and atomic number—that cannot be measured separately in a single transmission measurement. However, if a sample’s transmission is measured using two different beam spectra, the two different attenuation measurements can be used to separate the two components of attenuation. This property can be useful to probe the chemical composition of a sample. For example, a blood vessel containing iodinated contrast attenuates x-ray much more strongly than a vessel with contrast-free blood, although both have about the same density. The k-edge of iodine (determined from its atomic structure) makes it a strong attenuator. If the two vessels were imaged in a dual-energy mode and information about the physical density were recovered (independent of the atomic structure effects), the two vessels would appear to have about the same attenuation. Similarly the chemical composition of solids such as renal calculi can be probed with multiple x-ray energies to assess the content and composition of calculi having similar physical densities. The most promising applications for dual-energy and spectral CT capabilities are virtual noncontrast (VNC) exams, iodine quantification, and calcium quantification. Dual-energy/spectral reconstructions can also prevent beam hardening artifacts and related errors in attenuation quantification (HU values). There are several commercially available CT scanner design implementations that provide options for dual-energy or spectral CT scanning.

12

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging µ2

µ1

Rotation direction

Z

L1

X

I2

I1

I0

L2

I L3

FIG 1-12  Attenuation of photon beam through successive thicknesses

I

of tissue.

I

26 cm

Det B

J

y

between the energy measurements. In current commercial systems, the x-ray generator is not able to provide tube current modulation in fast kV switching modes, so manual tube current selection must be used, forgoing some tailoring of radiation exposure to the patient.

Dual Detector

Det A

FIG 1-11  Configuration of a dual-source CT scanner. (Courtesy Siemens Healthcare.)

DUAL SOURCE The dual source scanner employs two complete imaging assemblies, each consisting of an x-ray tube and detector array mounted at right angles to one another and aligned to scan the same plane simultaneously (Fig. 1-11). This approach increases the amount of azimuthal detector coverage so that 360-degree sampling can be achieved with a partial revolution of the scanning frame; for very fast rotations, this greatly increases scanning speed. Each x-ray tube can be operated at a different kV and mAs setting, and the data from the two detector arrays can be stored separately, so it is straightforward to enable the dualsource scanner to operate in a dual-energy mode. A typical approach would be to operate one tube at 80 kV and the other tube at 140 kV. The dual-source approach has several limitations. First, the second detector covers a smaller arc than the main detector, so the FOV is limited for any dual-source scan (whether high-speed or dual-energy). Second, the data collection for each of the two detectors is offset  from the other by a quarter rotation (in space and time), so the two energy measurements are not simultaneous. There can be misregistration between the two energy images if there is patient motion during the scan.

Fast kV Switching Scanners that use fast kV switching to provide dual-energy scanning have special x-ray generators (discussed earlier) capable of alternating the kV to the x-ray tube approximately every 150 ms so that data for the two spectra are acquired for interleaved azimuthal projections during the tube rotation. A typical acquisition would involve toggling between 80 kV and 140 kV. The data acquisition is synchronized to the kV switching so that the high- and low-energy data are stored separately to produce two different attenuation measurements in the tissue. Fast kV switching does not provide truly simultaneous acquisition of the two spectral measurements, but there is less spatial and temporal offset than in dual source. A fast kV switching scan would thus be less prone to misregistration between the two energy scans due to motion. Fast kV switching dual-energy scanning requires that the detectors contain scintillators with very fast light decay to prevent crosstalk

The dual detector approach uses a dual-layer detector exposed to a single x-ray beam. Because the beam has a wide energy spectrum, the top detector layer detects the low-energy photons, while the higherenergy photons pass through the top layer and are detected in the bottom layer. The detector is characterized to determine which portion of the spectrum is detected by each layer. The signals from the two layers are stored separately to provide two distinct energy-related transmission measurements. Unlike dual source and fast kV switching approaches, there is no need to prospectively select dual-energy operation with the dual layer detector. Every scan is inherently acquired in two detector data sets. At reconstruction these data can be combined into a conventional transmission measurement or used separately to recover density and composition information. In the dual-layer detector, all of the energy measurements are simultaneous for each projection, so there is no misregistration between the spectral signals due to motion. The major drawback to the dual-layer detector approach is that there is a large overlap in the spectra to which the top and bottom layers are sensitive, which degrades the ability to decompose the acquired data into material properties and accurately quantify the atomic number and density of species in the sample.

RECONSTRUCTION Basic Principles Reconstruction produces an image from the digital data, representing the attenuation coefficients of the tissue voxels, provided by the data acquisition system. The acquired scan data are composed of a large number of individual ray attenuations. Figure 1-12 shows a simplified diagram of an x-ray beam passing through successive slabs of tissue. The exit intensity from the first slab is the entrance intensity of the second slab, and so forth. To simplify this example, we assume that each of the slabs has the same thickness (L1 = L2 = … Ln) and the same linear attenuation coefficient µ, so that the equation for the transmitted beam passing through the n slabs is written as: I = I0e− µL where: µL = µ1L1 + µ2L2 + … µnLn If we now take the natural logarithm (ln) of both sides of the transmitted beam equation and rearrange it, the result is: 1 I µ = (µ1 + µ 2 + … µn ) = ln 0 L I

CHAPTER 1  Imaging Principles in Computed Tomography This equation shows that if the transmission (ratio of incident intensity I0 to transmitted intensity I) and path length L are known, the sum of the attenuation coefficients along the path of the x-ray beam can be calculated. In a CT scan, the transmission is measured and the distance from the source (x-ray tube) to the detector is known. Each projection then provides a measurement of µ. In a real sample, each incremental thickness of tissue does not have the same attenuation as each other one, and a projection only measures the sum of many tissues along the path. By measuring other projection views along different angles (paths), the individual elements that make up a path can be deduced from the other paths that cross it when other view angles are sampled.

Filtered Backprojection The simplest and most common method of producing a cross-sectional image from CT data is called filtered backprojection (FBP).19 This method is relatively fast and can be used to reconstruct a large number of images in a few seconds using modest computing resources. In FBP, a mathematical operation is applied to the projection data to extend each 1D projection radiograph into a 2D image. This is the “backprojection” of the 1D data into a 2D map. The 2D images are superimposed over each other, each oriented in the direction of the projection that was used to produce it. This process is illustrated in Figure 1-13. Some blurring always occurs because when each view angle is backprojected, the operation assumes that the measured attenuating material is uniformly distributed across the entire FOV; in reality, the attenuating material is localized somewhere within the FOV.

Filtering is applied to the data to help reduce the blurring due to backprojection. The filtering procedure consists of operating on the ray data with a filter function or convolution kernel. The filter function is a complex one that depends on many parameters, including x-ray tube geometry and detectors. It can take on a number of forms depending on the desired result. For example, one type of filter function might enhance edges and thus sharpen the image, whereas another would blur edges for more gradual density changes. The edge-sharpening filter enhances spatial resolution but simultaneously decreases density resolution. Also, filters that preserve (or enhance) greater detail also result in a noisier reconstructed image. Filters that introduce some blurring also have the effect of smoothing, which reduces noise in the reconstructed image. FBP of data acquired in helical scanning modes requires special treatment because the patient is moved continuously through the scan plane during data acquisition. If the data over a 360-degree rotation were reconstructed as in conventional (axial) CT, motion artifacts due to the movement of the table or patient would be visible. To minimize these artifacts and reconstruct a planar image anywhere throughout the volume of collected data, some data must be synthesized or interpolated. The simplest form of interpolation is linear. Before reconstruction, projection data are weighted to produce individual sections of data at select locations. Spiral interpolators apply weighting factors to segments of the spiral data to estimate the data that would have been measured if an axial scan had occurred at a particular location.

Backprojected ray

Optical density

A

Field position

Projection profile

C

Optical density

B

Field position

D

13

E

FIG 1-13  Filtered backprojection concept. A, Object in center of field of view, backprojection of one view, and backprojection of three views sampled on different view angles. B, Backprojection of one view angle with filtering applied. C-E, Reconstructed image of the simulated object acquired using 18, 36, and 72 views, respectively.

14

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

FIG 1-14  Cone-beam geometry in multidetector-row CT. The x-ray beam diverges along the z-axis, and the source-detector length is significantly different for central and outer detectors.

In multidetector-row CT, the fan-beam reconstruction techniques used by conventional and spiral CT must be modified to account for the divergence of the fan beam along the longitudinal axis (z-axis), which creates a cone-beam shape (Fig. 1-14). The cone-beam effect results from the source-detector geometry; there is a significant difference between the distance from the x-ray focal spot to the center detector row and the distance from the focal spot to the outer detector rows. This effect is more severe for wider detector arrays. If the number of rows of detectors is limited to a small number (e.g., four rows) of narrow detector-row widths (e.g., 1.5 mm), the cone-beam divergence is very small relative to the resolution of the detector element. One of the weaknesses of FBP reconstruction in general is that there is an inverse relationship between image noise and radiation dose to the patient. As radiologists seek to lower the radiation dose delivered to patients without degrading image contrast and detail, new methods have been developed to work around these limitations. Iterative reconstruction techniques have been in common use in nuclear medicine (single-photon emission computed tomography [SPECT] and positron emission tomography [PET]) for some time, but the number and size of the image files are much smaller for those modalities. Iterative reconstruction was not viable for CT until recent advances in computing enabled iterative reconstruction of CT data sets in times short enough to be feasible for clinical use.

Iterative Reconstruction Iterative reconstruction (IR) is a name for a broad class of methods that have general similarities.2,11,12,44,58,60,61 There are many distinct implementations that have been developed; a detailed description of all of these methods would occupy an entire textbook by itself. Here we focus on the general principle of how IR works and the clinical utility of IR for CT imaging. The most basic approach to IR is very different from FBP. The projection data are stored but are not backprojected. Instead the com-

puter generates a “guess” image. This image is digitally forwardprojected—simulating the process of CT scanning—to produce a set of projection data from the guess image. Next the projection data are compared to the real projection data acquired during the scan (usually by subtraction) and a difference data set is generated. The difference data are used to update the guess image. Then the cycle starts again; the new guess image is forward-projected, and so forth. This process continues until a preset number of iterations are completed or until the reconstruction converges. This is determined by the difference data containing small enough values for the latest guess image to be considered “equal” or “close” to the physical sample represented by the acquired data. It is interesting to note that in this simplified theoretical implementation of IR, the sampled data are not directly reconstructed; they are used as a reference for comparison but are not backprojected. This is not true of all commercial implementations of IR. These methods introduce the sampled data into the process in some way, such as blending the IR results with FBP image data reconstructed from the acquired data. The noise in CT is contained in the acquired projection data; it is either a product of random statistical photon noise in the x-ray beam nor electronic noise introduced by the scanner’s electronics. Thus the noise can be reduced in methods that use the acquired data to guide formation of the image by means other than reconstructing it directly. Qualitative comparisons of FBP and IR images generally note that the image characteristics differ in ways other than the noise amplitude. IR processing introduces different spatial characteristics into the image than FBP. Quantitatively these are shown as differences in the noise power spectrum (NPS) of each reconstructed image type.45 Qualitatively these feature differences are sometimes described as differing image textures.16 Matching or comparing IR and FBP images based on characteristics such as noise, signal-to-noise ratio (SNR), contrast-tonoise ratio (CNR), and spatial resolution (detail/blurring) would not fully describe the observed differences in image appearance that arise from these reconstruction and NPS differences.

DATA DISPLAY AND MANAGEMENT Image display includes all the system components necessary to convert the digital data provided by the reconstruction system to electrical signals used by the display monitor, enabling a graphic display of individual CT numbers representing attenuation values of individual sections of anatomy.8 In addition the display system includes the ability to display patient information, scan protocol, and reconstruction parameters, and it provides the user with many graphic aids to assist in the interpretation of clinical images. In Figure 1-15, a wide window display setting at left shows all the tissues but minimal contrast within and between them. The image at the right uses a narrow window setting that emphasizes contrast within the liver while saturating other tissues to white or black pixels.

Multiplanar Reformatting Because a conventional CT study consists of several contiguous axial images perpendicular to the long axis of the body, coronal or sagittal images are not possible to acquire directly unless the gantry is tilted or the body is positioned to show the image in the desired plane. This limitation of CT can be overcome by image manipulation commonly referred to as multiplanar reformatting (MPR). In this process, image data are taken from several axial slices and reformatted to form images in another plane. In this viewing mode, the user defines the number of imaging planes and their position, orientation, thickness, and spacing, and the reformatted image is displayed in sagittal, coronal, or

CHAPTER 1  Imaging Principles in Computed Tomography

15

B

A

FIG 1-15  Liver scan. A, Wide window setting; some lesions are not readily visible. B, A more appropriate window setting improves the visibility of lesions. (Courtesy Dr. John R. Haaga.)

oblique planes. In the x-y plane direction, the pixel of the reformatted image has the same length as the axial image; in the z-axis direction, however, the pixel length is the same as the slice thickness. Because in most scans the pixel length is considerably smaller than the slice thickness, the reformatted scan can have an unusual appearance (Fig. 1-16). The resolution is good in one direction in the image plane but very poor in the other. This problem can be addressed by scanning with a smaller slice thickness; also, as described earlier, the image is best enhanced with the use of spiral/multislice CT scanning, in which longitudinal resolution permits a slice thickness resembling minimal pixel length in the x-y plane. Scanning in this mode is commonly referred to as isotropic imaging. Isotropic voxels permit increased visualization of complex anatomic shapes that do not run linearly along the z-axis (patient axis). Data from successive slices can also be reformatted to generate other viewing modes; 3D shaded surface (Fig. 1-17), volume rendering, and maximum intensity projection (MIP) are a few examples. Each of these rendering techniques will be described in detail. Figure 1-18A illustrates a small segment of eight voxels representing a large-volume data set for demonstration purposes. Each of the eight voxels is shown with its values displayed. A viewpoint is defined from where imaginary rays are projected. To simplify the example, we can assume that each ray intersects two voxels, one behind the other.

Three-Dimensional Shaded Surface For a surface rendition, the user selects a threshold range (see Fig. 1-18B). This allows the user to select only the tissue (e.g., bone) to be rendered. The voxels with Hounsfield values within the threshold range are set to the “on” state, whereas the rest of the voxels are set to the “off ” state. Depending on the number of bits used for display, “on” and “off ” voxels are assigned appropriate values. For an 8-bit display, “on” would be 256 and “off ” would be 0. The second step is to project rays through the entire volume. As the rays pass through the data, they stop when they identify the first “on” voxel. For that particular ray, this first “on” voxel is part of the surface; the other voxels are ignored. This is done for all the rays, and all of the “on” voxels are used to create the surface. For our example, the “on” voxels that are part of the surface would be yellow in Figure 1-18C. As the name suggests, only a surface is displayed in this method; there are

no details below the surface. Shading the voxels enhances the effect so that the volume image appears to be illuminated by light sources. Multiple light sources can be used, but a single light source is used in most medical applications.

Three-Dimensional Depth-Based Shading One type of shading technique is called depth-based shading. With  this method, those voxels that are closer to the viewer are illuminated at a greater intensity than those that are farther back. As described earlier, it is also possible for the operator to select a range of CT numbers to display. If an operator were interested in a 3D view of the skull, he or she would select the range of CT numbers of the skull. The computer would then assemble those selected voxels into a 3D image. The result is a reconstructed 3D image of the skull with the soft tissue removed. As with sagittal and coronal images, the quality of the image is improved by interpolating the data between slices to form a smooth continuous image. When various 3D views are selected sequentially in time, the image can appear to rotate. These images pose the same challenges as do sagittal and coronal images. The vertical resolution is the same as the slice thickness and not nearly as good as it is in other directions. Thus spiral/multislice CT dramatically improves 3D viewing because of its ability to cover the same anatomy as a conventional CT scan at much thinner slice thicknesses and with overlapped image reconstructions.

Volume Rendering Another possible viewing mode is compositing/volume rendering (VR). VR is an advanced rendering technique that displays an entire volume set with control of the opacity or translucency of selected tissue types. In this case, each voxel has an associated intensity in addition to an associated opacity value. Figure 1-19 shows the same 8-voxel block as Figure 1-18A, with each voxel having an opacity value associated with it. The user can define the opacity values for various HUs. Once the opacity values are assigned, the second step is to pass rays through the volume, accumulating values along the way. The value for each voxel is determined using the intensity value of the voxel and the opacity selected by the user. Figure 1-20 demonstrates the results of this technique for the simplified example.

16

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

FIG 1-17  Shaded-surface view of coronary arteries presented as a

A

“soap bubble” view. Such views permit better visualization of the overall paths, but as with any reconstructions, meticulously executed algorithms are essential to prevent artifacts. (Courtesy Philips Medical Systems.)

The obvious advantage of VR over 3D shaded-surface rendering is that it provides volume information. By using transparency, the operator can visualize information beyond the surface. For example, VR is the preferred viewing mode for stent evaluation, because stents can be made transparent to view the lumen of vessels. Other clinical examples include the ability to make plaque transparent for more accurate diagnosis of vessel stenosis.

Maximum Intensity Projection

B

Using maximum intensity projection (MIP) for visualization permits easy viewing of vascular structures or air-filled cavities.13 Compared with 3D or volume rendering techniques, MIP is a relatively simple method. Unlike 3D shaded-surface and VR displays, no preprocessing is required. The rays are cast throughout the volume, and depending on whether it is MIP or minimum intensity projection, maximum or minimum values along the rays are used in the final image display. On the basis of the 8-voxel example described earlier, Figure 1-21 shows the final pixel values for MIP and minimum intensity projection. MIP is a preferred method for many CT angiography applications for visualization of contrast-filled vasculature. MIP enables easy viewing of an entire vessel in one image. This is because voxels representing the contrast-filled vessels are most likely to be the ones with the highest values along the ray (assuming no bone along the ray). Similarly, minimum intensity projection can be used to visualize airfilled cavities.

IMAGE QUALITY Quantitative Measurements C FIG 1-16  Multiplanar reformatting of CT images of the cervical spine. A, Multiplanar reformatted images acquired with a slice thickness of 2.5 mm, showing stair-stepping artifacts. B and C, Multiplanar reformatted images acquired with slice thickness of 0.5 mm (isotropic resolution). (Images courtesy Dr. Jay Cinnamon.)

Several quantitative measurements are used to define a CT system’s image quality: spatial resolution, contrast resolution, image noise, and temporal resolution. Image quality is also related to the radiation dose used to produce the image.

Spatial Resolution.  Spatial resolution is measured by the ability of a CT system to distinguish two small high-contrast objects located very close to each other under noise-free conditions. Optimal spatial

CHAPTER 1  Imaging Principles in Computed Tomography 230

17

210

Threshold range 270 10

260

280

30

1

220

0 200

A

B

View point 230

300

Hounsfield units

210

270 10

260

280

30

220

C

View point

FIG 1-18  A, Sample of 8-voxel data set with displayed CT number values. B, User-defined threshold of CT number values for tissue definition. C, Shaded voxels represent defined tissue surface (shaded-surface display). (Courtesy Shalabh Chandra, Philips Medical Systems.)

230

210 115

435

250

135

270 10

260

220 280

30

View point Opacity values for the voxels are given by: HU < 50. α = 0.0: 50 ≤ HU < 200. α = 0.2: 200 ≤ HU. α = 0.5

FIG 1-19  Sample of 8-voxel data set with representative opacity values assigned to various CT number ranges. (Courtesy Shalabh Chandra, Philips Medical Systems.)

Volume rendered, 4-D angio, composited

FIG 1-20  Image results for volume-rendered image of Figure 1-19.

18

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

Relative profile optical density

Profile line

A

100

Ideal response

50

FWHM Measured response

B

Position along profile line

FIG 1-21  Full width at half-maximum (FWHM) demonstrated for a point source. A, Solid body in uniform medium. B, Profile across the image demonstrates the fading periphery of the image of the solid body. The rectangular profile represents an ideal response with perfectly sharp edges.

A

230

260

10

210

280

220

270

30

MIP—Maximum intensity projection

B

MIP—Minimum intensity projection

FIG 1-22  A, Maximum intensity projection of a sample voxel in which only maximum values are displayed. B, Minimum intensity projection in which only minimum values are displayed.

resolution is required for evaluating high-contrast areas of anatomy (e.g., inner ear, orbits, sinuses, bone) in general because of their complicated shapes. Spatial resolution can be specified by spatial frequencies, which indicate how efficiently the CT scanner represents different frequencies. Modulation transfer function (MTF) describes this property. A 2D Fourier transform of the point spread function taken with a very thin wire provides MTF values. These concepts are described next. Consider the image of a point source that can be obtained by placing a thin wire on end in water or plastic and then scanning across the wire. Ideally the resulting image should be a uniform background with a bright dot denoting the position of the wire. In fact the image does demonstrate the bright dot but with a fading periphery rather than an absolutely sharp edge. A profile across the image (i.e., a plot of the picture density along a line passing through the point source center) produces a bell-shaped distribution called the point response distribution. The height of this curve represents the maximum value of the density, and the width represents the uncertainty in the measurement of the exact boundaries of the wire. Full width at half-maximum (FWHM) is the width of the curve at the point where the attenuation values are 50% of the peak value (Fig. 1-22). The smaller the FWHM, the better the resolution. One can appreciate this fact by noting that a large FWHM reflects a larger fading periphery, which means a poorer reproduction of the actual object. Another method of examining spatial resolution is to consider line images. This is similar to examining point images, but the scan would be done parallel to a wire in a medium rather than across the axis. If a series of lines of different spacing or different sizes were used, the ability to distinguish separate lines in an image would measure the spatial resolution. With very fine lines a measure of the resolution can

be stated in terms of the number of lines that are still discernible, packed together in a given distance. Thus one would speak of the number of lines per centimeter that could be visualized in the image. Describing the spatial resolution in terms of lines per centimeter is said to describe the resolution in the frequency domain. The frequency in this case is the parameter lines per centimeter. A profile across the line image yields the line response curve, which is similar to the point response curve. Again, one can speak of FWHM and the analysis is similar to that done with the point source. Obviously the two response curves are related. Performing a mathematical operation, the Fourier transform, on these response curves yields the MTF. The physical interpretation of the MTF is that it defines the relative fidelity of the image compared with the actual object. Another description of the MTF is shown in Figure 1-23. The rectangular profile represents the density profile of an object composed of dense lines. The density profile of the reconstructed image is shown as rectangles with rounded corners. As the lines become closer together, the region between lines fills in because of the overlap of the response functions. The ratio of the height of the rectangles to the height of the valley between lines becomes smaller as the lines become closer in the reconstructed image, but obviously the ratio is fixed in the object. This change or decrease in ratio is seen as a decrease in contrast and is a manifestation of the decreasing ability of the system to resolve small objects (or small separation of objects). The MTF is basically the value of this ratio in the image divided by the value of the ratio in the object. As the frequency increases (i.e., more lines per centimeter) the ability of the system to reproduce the lines and valleys accurately is decreased and the fidelity also therefore decreases.

19

CHAPTER 1  Imaging Principles in Computed Tomography MTF

Line object 1

A

B

Actual density profile

Relative A–B = line A contrast

A1

B1

Image density profile

Relative A1 – B1 image = A1 contrast MTF =

(A1 – B1)/A1 (A – B)/A

A B

1

2

3

4

5

6

Frequency (lines/cm)

FIG 1-24  Modulation transfer function (MTF) curves for two theoretical systems. The system represented by curve A demonstrates a better spatial resolution than that of curve B.

FIG 1-23  Modulation transfer function (MTF) demonstrated in an array of line objects. As the line frequency increases, the relative contrast of the image of the lines decreases, which is reflected in the decreasing MTF.

With widely separated lines, the MTF is 1—the maximum value; as the frequency increases, the MTF drops. Thus the higher the MTF at a given frequency, the better the fidelity (i.e., the better the resolution). In systems with large values of MTF at high frequencies, edges are more clearly defined; such systems produce images that appear sharp. If high frequencies are exaggerated relative to low frequencies, the result is an edge enhancement. Techniques for such high-frequency enhancement are often used to advantage in radiology, and this is the main purpose of xeroradiographic methods or subtraction methods. In Figure 1-24, curve A denotes a system with an MTF higher than that shown by curve B at any given spatial frequency; in this case, system A would have better resolution than system B. The spatial resolution (composite MTF) is affected by many design parameters. The most important are: Choice of filter used in reconstruction Opening size of detector aperture Number of projection profiles obtained Matrix (or pixel) size Focal spot size of the x-ray tube Relative object-to-background contrast (density) The last two parameters in this listing are presented later under the heading “Contrast/Density Resolution and Image Noise.” Spatial sampling (based on number of rays, views, and number of detectors) is the limiting factor for the composite MTF for most rotate/rotate systems. Focal spot size and detector aperture size typically do not limit the composite MTF of these systems, although they have been important considerations in other historical scanner designs. Filter effects on resolution.  As discussed earlier, the major role of the convolution filter is to remove the image blurring created by the backprojection process. Various filters control the amount of image blurring created by accentuating high-frequency components found in the data. For a crisp image, the high spatial frequencies are accentuated, and this has the effect of sharpening the edges and improving resolution. If the high spatial frequencies are not accentuated, the image of the object appears more blurred.

A tradeoff is that image noise increases and the density resolution diminishes with accentuation of high spatial frequencies. One pays for a crisp picture with a decrease in density resolution. Similarly, by increasing density resolution, one pays with loss of some spatial resolution and image crispness. Number of projection profiles.  To reconstruct the image, data must be acquired for enough projection angles to provide independent measurements through each desired pixel. If one wishes to reconstruct an image with a matrix of n columns by n rows, then n2 measurements or ray sums must be obtained. With the large number of detectors used in present machines, this criterion is easily met. By itself, this criterion is not sufficient to guarantee the desired resolution, and the question of optimum number of angular measurements or views must be answered. Theoretically this optimum number depends on required resolution and object size.25,27,49,55 To obtain a resolution of 1 mm with small or average-sized bodies, 300 to 400 views may suffice; for larger bodies, 700 to 800 views may be needed. Modern CT scanners are capable of meeting these requirements. In many machines these needs are exceeded significantly, allowing spatial resolution down to the range of 0.25 mm. Matrix size.  It is obvious that the spatial resolution can be no greater than the size represented by the pixel length. In reality,  pixel size should be 1.5 to 2 times smaller than the desired resolution. Unless a matrix element exactly coincides with an object, the object representation will be averaged over two or more pixels and thus  may not be visualized. It must be realized that the pixel size refers  to the FOV (or body), not the viewing display. For example, in a field size 50 cm in diameter, a matrix size of 512 × 512 would indicate that each pixel represents a 0.98 × 0.98 mm area in the FOV. Typical matrix sizes available in current commercial scanners are 512 × 512 and 1024 × 1024. The imaging goal is to match the pixel size with the inherent reso­ lution of the CT system and scan protocol/filter selection for each examination. As discussed earlier, spatial resolution is measured in line pairs per centimeter. To convert this to millimeters, use the following formula: Spatial resolution (mm) =

10 lp × 2 cm

20

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

To illustrate the concept, consider a high-resolution study with a protocol and reconstruction filter designed to achieve a spatial resolution of 20 lp/cm and an FOV of 250 mm. Converting 20 lp/cm to spatial resolution in millimeters using the preceding formula results in a spatial resolution of 0.25 mm. To achieve a pixel size of 0.25 mm without zooming, the reconstructed FOV requires a matrix of 1024 × 1024 (250 mm/0.25 = 1000). To use a matrix of less than 1024 × 1024 would suggest a visual loss of spatial detail; the spatial information is present in the raw data but cannot be viewed in the reconstructed image.

Spiral and multislice spiral CT effects on spatial resolution.  Factors that affect in-plane spatial resolution for multislice spiral

CT are the same as for conventional CT, although spatial resolution greater than 10 lp/cm is limited to axial modes of scanning or to systems equipped with x-ray tube focal spot modulation techniques to increase the data sampling rate during spiral scanning. During spiral and multislice spiral CT, there is also a potential loss of z-dimension (through-plane) spatial resolution (for a given slice width). The loss of resolution depends on the spiral interpolation algorithm used during reconstruction and in some cases the pitch factor used. This issue does not cause problems for most clinical applications, but the potential exists for partial-volume averaging effects or degradation of MPR images when scanning at high pitch values.

Contrast/Density Resolution and Image Noise.  The second major factor affecting the ability of a scanner to measure and display anatomy accurately is contrast (or density) resolution, which is the ability to differentiate the attenuation coefficients of adjacent areas of tissue. Because most soft tissues have nearly identical densities, usually it is necessary to differentiate variations of a few percent or less. In the computation of any single pixel value, there is error in the form of statistical variation; it is this variation that limits the ultimate contrast resolution. This variation (called image noise) is manifested as a grainy background, or mottle. The parameter used to evaluate this variation is the standard deviation (SD, or σ), and the usual procedure for evaluating a system is to obtain a scan of a uniform substance (e.g., water) and to compute the SD of pixel values within a region of interest placed on the image within the scanned object. The meaning of the SD is that rescanning the same water bath and recalculating the same pixel CT numbers would yield values in a range equal to the previously calculated value ± SD in approximately two of three instances. The SD therefore describes the uncertainty in a measurement and is commonly expressed in terms of percentages. This percentage is obtained by dividing the computed SD in CT units by the range of the scale. For example, an SD equal to 5 HU on a Hounsfield scale of +1000 would translate to 5 HU/1000 HU = 0.005, or 0.5%. The interpretation is that two adjacent pixels are not significantly different if their CT numbers vary by 5 HU or less. Because noise is the ultimate limitation in the accuracy of density resolution, parameters that affect or induce noise need to be understood. The most important parameters are (1) photon flux, (2) x-ray scatter, (3) computation-induced error, (4) filter frequency response, and (5) voxel size. The effect of filters was discussed previously. In any ray measurement, the statistical variation associated with that measurement is directly proportional to the number of photons detected. If I is the number of detected photons, the SD is given by the square root of I, and the relative noise in a ray measurement is given by: I I Because the variation in a ray measurement is propagated into every pixel computation that involves the ray, the relative image noise is

affected inversely as the square root of the detected photon flux. Large photon fluxes, produced by high x-ray tube currents (milliamperes) set by the scanner operator or automated scheme, can reduce relative noise; unfortunately they also increase radiation dose to the patient. Therefore limits on CT image noise reduction are imposed by practical limitations on patient radiation dose. The radiologist has an obligation to accept the highest-noise image (lowest-dose protocol) that is consistent with a satisfactory diagnosis. The photon intensity that reaches the detector, not the intensity incident on the patient, must be taken into consideration. The efficiency of the detector, impinging photon flux, size of the patient, and presence of high-attenuation materials affect the detected intensity and in turn the uncertainty in the measurement. The image of a large patient thus has a more mottled appearance than the image of a small patient for the same x-ray tube radiation tube output. Similarly the presence of a high-contrast object, such as bone or a prosthesis, causes increased noise by reducing the transmitted photon flux. Another aspect of noise is x-ray scatter. No detector can discriminate between a primary photon directly from the source and a scattered photon arising from an area not in the line of the ray projection. In the energy range used in scanners, many photons are scattered through small angles. The effect is to alter the recorded values in adjacent detectors. This produces a decrease in differences between adjacent measurements, with a concomitant decrease in contrast resolution. The sharp interfaces between tissues also are diminished by this phenomenon, causing a reduction in spatial resolution. The amount of scattering that is detected is controlled by collimation of the detectors. Smaller apertures reduce scattering and therefore reduce the noise component due to scattered photons. The measured intensity is also affected by the voxel length (i.e., slice thickness). If the slice thickness is reduced by 50%, the detector collimator size must be reduced accordingly, which in turn reduces the intensity by the same amount. To retain the same accuracy, one must restore the detected photon intensity, which means doubling the dose. Similarly, decreasing pixel size increases the relative inaccuracy unless the photon flux is raised. If the pixel side length is reduced by half, the voxel volume represented by the pixel is reduced by a factor of 4. The SD of the resulting calculated CT number, shown earlier to vary as the square root, would double. The image would become more mottled in appearance, and the contrast resolution would deteriorate. To restore the uncertainty to its previous value, one would need to increase the radiation dose by a factor of 4. Low contrast sensitivity and density resolution (as reflected through the SD) are related to spatial resolution (as reflected through aperture opening or pixel size), and these two components of image quality cannot be treated completely separately. Spatial resolution is usually stated for objects of high contrast. In CT scanning, attenuation coefficient differences of approximately 12% or greater are considered high contrast. As contrast differences decrease, spatial resolution also decreases, as implied by MTF curves.

Spiral and multislice spiral CT effects on contrast resolution and image noise.  The same variables that affect image noise and

contrast resolution for conventional CT—slice thickness, kilovoltage, milliamperage, patient size, voxel size, and convolution filter—affect spiral and multislice spiral CT, with the added variables of table speed (pitch) and the spiral interpolation algorithm. For a given pitch, using more scan data to reconstruct an image (with linear interpolation) reduces noise and enhances contrast resolution at the expense of z-axis spatial resolution.18,27 Temporal resolution refers to the ability of a CT scanner to capture objects that change shape or position over time and depends primarily on the gantry rotation speed and the reconstruction method used. For

CHAPTER 1  Imaging Principles in Computed Tomography

21

180 LIA

Helical weight

2.5 2 1.5 1

FW/HM

0.5 0 –200

–100

0

100

200

View angle relative to center of slice

FIG 1-25  Weighting values of 180-degree linear spiral interpolator demonstrating temporal resolution (full width at half-maximum) of approximately 0.625 second for a 1-second, 360-degree scan.

FIG 1-26  Ring artifact demonstrated in the center of a uniform water phantom. Note streak artifacts are also pictured, caused by metal pins or feet supporting the bottom of the phantom.

conventional CT using a full 360 degrees of scan data, temporal resolution is equivalent to scan time per rotation. A 1-second scan has a 1-second temporal resolution. For spiral or multislice spiral CT, temporal resolution depends on the gantry rotation speed and the spiral interpolating algorithm used during reconstruction. Figure 1-25 depicts a 180-degree linear interpolator that demonstrates the relative spiral weighting values applied to projection data at the center of an axial slice. The central 180 degrees of data, including the fan angle (≈50-55 degrees), are acquired in approximately 0.625 second for a gantry rotating at a speed of one revolution per second. The temporal resolution (FWHM) for a 360-degree linear interpolator, which uses 720 degrees of scan data, is approximately 1.25 seconds, or double that of the 180-degree interpolator. Projection data can be reconstructed at prescribed intervals coincident with a patient’s heart rate (monitored as an electrocardiographic waveform) to reduce the effective temporal resolution to approximately 100 milliseconds or less for data acquired using 0.5-second gantry rotation speeds (or less). As described in the clinical chapters, these techniques are in common use to provide the required temporal resolution for applications such as cardiac and functional CT imaging.

IMAGE ARTIFACTS AND THEIR CAUSES There are many possible causes of artifacts in an image, and it is important for the radiologist to recognize all types so that they can be eliminated or minimized.

Ring Artifacts Ring artifacts are usually the result of detector calibration problems (Fig. 1-26). An incorrect response (high or low signal) from any one detector incorrectly backprojects the error in each view angle to produce the ring artifact. If a detector is not matched or not calibrated accurately, the backprojection for each data ring will be slightly different, causing multiple rings. Poor alignment of tube and detector causes displacement of calculated values because measured values are occurring in lines different from those assumed by the reconstruction algorithm. The result can be blurring edges if the misalignment is slight, or ring artifacts if the misalignment is large. Detectors in the center of the detector arc are most sensitive.

Streak Artifacts The presence of objects having an exceptionally high or low attenuation can create artifacts by forcing the detector to operate in a

nonlinear response region. Because this incorrect response occurs at specific directions of the beam through the object, incomplete cancellation of the backprojected rays during reconstruction occurs and yields streaking artifacts. Metallic pins, for example, can give rise to streaks, as can bone or gas (Fig. 1-27). The range of x-ray intensity values to which the scanner can accurately respond linearly is called the dynamic range. The larger the dynamic range, the less prone the instrument is to creating such artifacts.

Beam-Hardening Artifacts Beam-hardening artifacts result from the preferential absorption of low-energy photons from the beam. In the keV photon energy range used in CT scanners, a prominent mode of interaction is the photoelectric absorption process, and low-energy photons are more readily absorbed via photoelectric interactions than high-energy ones. Thus as a beam traverses the body, the average beam photon energy is progressively increased (see Fig. 1-5). Accordingly, toward the end of the x-ray path the attenuation is less than at the beginning because the attenuation coefficient is smaller for the remaining beam, which has higher average energy. The reconstruction program, however, assumes a monochromatic beam and attributes any change in beam intensity to a change in tissue composition rather than to the result of a shift in average photon energy. The assigned attenuation coefficients are thus in error, and the densities seen on the image are in error. The effect is most pronounced in regions of strong attenuation, such as bone. The artifact is seen as a shadow beneath the ribs, for example, or increased shadows in the mediastinum or skull (Fig. 1-28). This effect occurs throughout the image, but qualitatively it is not usually perceived except where there is a great deal of hardening (e.g., in the vicinity of bone). This effect can be compensated for by use of a correction method that consists of forward-projecting only the “bone” portion of an image and by determining the required correction using a polynomial.48

Partial-Volume Artifacts When tissues of widely varying absorption properties occupy the same voxel, the beam attenuation is proportional to the average value of the attenuation coefficient of the voxel. A volume average is computed for such voxels, leading to the partial-volume error. A common site of such an effect is the lung-diaphragm interface, where an increased density may be attributed to the lung base when in reality it does not exist. A

22

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C FIG 1-27  Streak artifacts. A, Uniform medium with metallic pin; note the large amount of streak artifact. B, A hip prosthesis. C, Streaks from metal clips in the heart.

A

B FIG 1-28  Beam-hardening artifacts. A, Postoperative CT image of the brain in an adult. B, Comparison of reductions in beam-hardening artifacts with use of special reconstruction techniques. (Courtesy Marconi Medical Systems.)

CHAPTER 1  Imaging Principles in Computed Tomography

A

23

B FIG 1-29  Liver scan showing partial-volume artifact. A, A 2.5-mm slice thickness shows good visibility of small liver lesions. B, A 10-mm slice thickness shows reduced lesion detectability with partial-volume averaging. (Images courtesy Dr. John Haaga.)

partial-volume artifact can be subtle and gradual, requiring care in interpretation. Partial-volume artifacts can obscure the appearance of small lesions that are averaged together with normal tissue lying nearby in the z direction (Fig. 1-29). Another common partial-volume artifact is observed in scans of the head in the posterior fossa region.17

Motion Artifacts A motion-induced artifact usually appears as a streak or starburst and occurs where there are high-density or low-density interfaces (e.g., gas in the bowel; Fig. 1-30). This occurs because movement during the measurement process causes structures to be in different positions when different views are taken, so the backprojections are not added correctly. Artifacts can also appear as “double imaging,” with the same outline shown twice.

Stair-Stepping Artifacts Stair-stepping artifacts (see also “Multiplanar Reformatting”) occur when in one direction the pixel of the reformatted image has the same length as the axial image but in the other direction the pixel length is the same as the slice thickness. Because pixel length in most scans is considerably smaller than slice thickness, the reformatted scan has an unusual appearance (see Fig. 1-16).

Spiral Pitch Artifacts Pitch artifacts appear primarily because larger pitch factors are used to attain maximum coverage. Spiral artifacts exhibit similar stepping as axial scans in reformatted images, except that the steps appear as a spiral groove (Fig. 1-31). Multislice pitch artifacts (again, for large pitch factors) have a unique appearance in axial scans; a star pattern is seen off of sharp edges, where the number of spokes in the star is directly related to the number of multislice detector rows (Fig. 1-32).22,41 This is because each row contributes only a portion of its projection data to the multislice reconstruction.

FIG 1-30  Streak artifact (open arrow) caused by motion of bowel gas in a slow abdominal scan on a conventional single-slice CT scanner. Machine-caused streaks are also seen. Streak artifact (solid arrow) is caused by an unstable rotor in the x-ray tube, producing unstable focal spot motion during the scan.

Cone-Beam Artifacts As described earlier, with increasing number of rows in a multislice scanner, the divergence of the cone beam along the z-axis becomes significant relative to the detector array width. For a small number of rows, this effect is small and noticeable only for sharp objects distant from the scan center. For scanners with eight or more rows, the conebeam effect becomes more significant, requiring the reconstruction process to account for the divergence effect (Fig. 1-33). These reconstructions have become highly effective, and because more of the x-ray

24

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C

D FIG 1-31  Lung scan demonstrating effects of increasing spiral pitch with single-slice spiral CT scanner. High spatial and z-axis resolution at pitch 1 (A) and pitch 2 (B). C, Same scan as B but windowed to demonstrate spiral pitch artifacts. D, Same scan as A but windowed to demonstrate pitch artifacts. Note the increased spiral pitch artifacts at pitch 2 (C) versus those at pitch 1 (D).

beam is used for information acquisition, radiation dose to the patient has been reduced. Special filtration of the outermost perimeter of the cone beam is now possible, further reducing the radiation dose to the patient.

BENEFIT AND RISK: RADIATION DOSE IN CT The total volume of CT scans, the share of CT scans as a proportion of medical imaging procedures, and the average number of scans per patient increased rapidly from 1980 to 2006.43 In 2009, several highprofile incidents involving large radiation exposures to CT patients raised concerns among patients, healthcare professionals, and the

public about risks from radiation exposures in CT.3-5 A number of recent research studies have estimated cancer risk for CT patients to be large.7,9,38,46,50,54 As a result there is now increased attention on radiation exposures in CT. The increased power of CT scanners means they have become capable of delivering large radiation doses very quickly. The growth of clinical uses of CT has allowed imaging to replace riskier or lesseffective approaches to diagnosis and treatment. In the balance of these factors, it is important to take seriously both the risks and benefits of CT imaging in modern health care. CT scans give benefits to the patient that outweigh the risks when used appropriately (justified) and when the equipment and technique are used in a way that provides the

CHAPTER 1  Imaging Principles in Computed Tomography

A

25

B FIG 1-32  Phantom study demonstrating reduced appearance of multislice spiral CT pitch artifacts at pitch 0.675 (A) versus pitch 1.25 (B). (Simulations courtesy D. J. Heuscher and M. Vembar, Marconi Medical Systems.)

A

B FIG 1-33  Phantom study demonstrating cone-beam artifacts. A, Four-slice CT scanner study. The patient’s ribs are positioned at a steep angle along the z-axis to demonstrate cone-beam artifact at pitch 1.25. B, Enhanced contribution of multislice CT cone-beam artifacts caused by divergence of the x-ray beam in the z-axis when the number of detector rings used to acquire a CT image is increased from four to eight. (Simulations courtesy D. J. Heuscher, M. Vembar, and C. Vrettos, Marconi Medical Systems.)

needed information for the lowest radiation dose (optimized). The details of justification and optimization and applying these principles to CT scanning protocols are discussed in detail in Chapter 2. It is useful to quantify the benefit of CT by determining the typical number of nonincidental findings per number of exams performed or patients examined. For example, in one hospital, brain CT for a specific presentation in the emergency department might yield a result of 1 injury per 10 patients who required CT for diagnosis and required immediate treatment. These figures can be compelling in discussions with patients and referring physicians, and they provide a useful metric for optimization and quality-improvement efforts discussed

below. There is probably no single target value for this ratio. In an ideal scenario it would be possible to avoid all imaging of patients without abnormalities, and every scan performed would give adequate image quality and information to make the diagnosis, leading to a ratio of 1 finding per exam. This is far from realistic, but the goal of optimization efforts should be to reduce this ratio to as close to 1 : 1 as practical. Radiation doses in CT are potentially large enough that they are worth managing carefully. They are not large enough to justify avoiding medically necessary scans owing to fear of the consequences of radiation exposure.

26

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

Radiation doses in CT are potentially large enough that they are worth managing carefully. Individual diagnostic CT exams performed correctly deliver typical effective doses of 3 to 9 mSv, which is the equivalent of 1 to 3 years’ worth of background radiation. For context, in the United States, people who work with radiation (e.g., radiologists, imaging technologists, and nuclear facility workers) are allowed to receive up to 50 mSv per year of effective dose from their work activities. In practice, workplaces limit exposures to much less than this value, but doses up to this level over time are not considered unsafe for workers by the regulatory authorities. The risks for the radiation exposure from CT are called stochastic risks, and in practical terms this means the risk that the radiation exposure will cause a genetic mutation that will lead to cancer later in the patient exposed. The best evidence for specific cancer risks is for effective doses above 100 mSv from a single exposure. The risks associated with effective doses less than 100 mSv, and for multiple exposures within a short period of time (1-2 months or less), are estimated from the higher-dose data using the linear-no threshold (LNT) model to obtain a worst-case estimate. In several high-profile incidents around 2009, CT patients suffered skin injuries and hair loss during head CT scans. These injuries are of the class of effects called deterministic effects; these normally only occur during procedures like complex fluoroscopically guided interventions and radiation therapy. For one specific type of CT exam—brain perfusion—the normal radiation exposure is around one eighth of the skin dose needed to produce skin injury. This is due to the fact that this study requires many repeated exposures of the same region of the head. Unexpected and unnoticed changes to the protocol settings can readily increase the dose by a factor of 8 or more, which can lead to skin injuries like those publicized for several hospitals and a group of a few hundred patients. However, this is not a concern for most CT exams and protocols; there is a much larger margin between the normal exam dose and the deterministic injury dose for all other CT exam types.

Radiation Dose Metrics

CT Dose Index.  The fundamental parameter used to quantify radiation exposure and dose in CT is the CT Dose Index (CTDI).42 This parameter is written in absorbed dose/air karma units (mGy, and in some references, mrad) and reflects a specific method for measuring the radiation output of a CT scanner. There are a number of different methods of measuring CTDI, and the various resulting indices are identified by their labels: CTDI100, CTDIw, and CTDIvol are just a few. Because the x-ray beam in CT is very intense and tightly collimated to expose only a relatively narrow slab or slice of tissue during each tube rotation, there is a large amount of scattered radiation that exposes the tissue surrounding the primary exposure region. In the course of scanning a particular bit of tissue as well as the neighboring tissue in adjacent slabs along the z-axis, the tissue of interest will receive a total dose much higher than the contribution of the primary beam that exposes it. At the same time, the patient is exposed from all sides owing to the rotating scan, and tissue in the body will receive different doses depending on their depth in the body, because internal organs and tissue are partially shielded by the superficial organs and tissues. The CTDI approach is a way to account for the radiation output of the scanner, the depth-dose variation in tissue, and the effect of large amounts of scatter alone the z-axis using a simplified and standard method. The basic CTDI value, CTDI100, is named for the measurement of scanner radiation exposure, which is made using a pencil-shaped ionization chamber that is 100 mm long. The chamber is placed inside a

drilled channel in a circular phantom made of solid acrylic; the phantom has a diameter of 16 cm to simulate a head or 32 cm to simulate a body and is about 15 cm thick. When a CT exposure is made with the fan beam passing through the center of the ion chamber and phantom, the phantom will produce scatter along the z-axis, and the ion chamber is long enough to collect the scattered radiation along its length. The total exposure to the ionization chamber in this setup reflects the exposure to the tissue in the center slab from the direct exposure to the CT beam, as well as the scatter that central slab would have received from performing additional scans with the same technique factors through the tissue immediately surrounding the central slab—in this case if the entire phantom were scanned end to end. The CTDI100 measurement can be made at any location within the phantom, but the standard positions are in the center of the phantom and at peripheral positions about 1 cm from the edge of the phantom. To account for the variation of dose with depth, the weighted CTDI (CTDIw) is used. To determine CTDIw, CTDI100 is measured at the center and peripheral positions in the phantom (described earlier). The CTDIw is equal to the weighted sum: CTDIw = 1 3 (CTDIcenter ) + 2 3 (CTDIperipheral ) This can be thought of as a spatial average of absorbed dose in the phantom, although the actual dose value at specific locations varies with depth. The volume CTDI (CTDIvol) is calculated from the CTDIw to account for the effect of pitch values not equal to 1.0 in helical scanning. For axial scanning, CTDIvol = CTDIw. As described earlier, helical scans with pitch greater than 1.0 leave some tissue unexposed between passes of the x-ray tube, whereas scans with pitch less than 1.0 expose some regions of tissue to the beam on two successive passes. The CTDIw is always measured using a single axial acquisition with the phantom, ion chamber, and patient table stationary during the measurement. To show the effect of pitch on a clinical protocol, the CTDIw is measured using an axial exposure with the same kV, mA, rotation time, beam collimation/geometry, and x-ray filter combination used in the helical scan, and the pitch value for the helical scan is noted. Once the CTDIw is determined, CTDIvol = CTDIw/pitch. It is important to note that all CTDI values are really indications of the radiation output of the CT scanner under specific measurement conditions. Although expressed in units of absorbed dose, they are related to but not the same as the absorbed dose to a patient or to any organ or tissue within the patient. To determine detailed patient dose values, the CTDI values may be used, but many additional details are needed. It is not normally necessary to determine the radiation dose to specific patients in detail in CT. CTDI values (especially CTDIvol) are useful for quantifying the relative dose changes caused by adjustments to the CT scanner protocol. As scanners determine the estimated CTDIvol value from the selected scan parameters and display it on the operator’s console, this is a useful parameter to monitor relative radiation exposure levels and a useful way to identify unusually high or low exposures for a given protocol. CTDI values are also useful for quality control of scanner performance because they can be measured directly. Commercial CT scanners determine the CTDIvol values for display by using lookup tables or formulas based on the selected x-ray operating factors, and the accuracy of the displayed value can be validated.

Dose Length Product.  Although CTDI provides a useful way to assess the relative impact of changes in scanner technique factors on dose, all forms of CTDI relate to the exposure from a single scanner rotation. Therefore the CTDI values do not increment for scans that

CHAPTER 1  Imaging Principles in Computed Tomography expose larger regions of anatomy. A scanner would report the same CTDIvol value of 15 mGy for a given set of scanning parameters regardless of whether those settings were used to scan a patient’s chest, abdomen, and pelvis or just the chest. Where a metric is needed to reflect the fact that a patient in the former case receives radiation dose to more organs and tissues and thus has a higher risk for radiationrelated effects, the dose length product (DLP) is a convenient and straightforward value to obtain.23,42 The DLP is simple to calculate for a scan of a region that includes more than one rotation of the scanner. The average CTDIvol value for all of the rotations is determined and then multiplied by the total length (along the z-axis) included in the data acquisition. The result is the product of an absorbed dose and a length, and it is usually labeled with units of mGy*cm. On commercial CT scanners, the DLP is displayed to the operator and written into data headers and exam summary files at the completion of a study. Interpretation of DLP values is not straightforward, but many technologists and radiologists have developed familiarity with the range of normal DLP values for specific scan protocols in their institutions and are readily able to spot outliers that deviate from the expected range.

Effective Dose.  Effective dose is not unique to CT, but there is a CT-specific method for determining estimated effective dose values for CT protocols or studies. Each general body region (e.g., head, thorax, abdomen) has a k-factor (reference), which is a coefficient that can be multiplied by the DLP value for a scan performed on that body region to arrive at an estimated effective dose value.24,53 This value is a very rough estimate because none of the underlying organ doses are calculated directly; the k-factors contain many assumptions and generalizations about which organs are exposed when scanning each body region, as well as assumptions regarding patient body size, density, location and depth of organs within the body, and so forth. Because of the imprecision of the approach, it is generally not advisable to calculate or report effective dose based on DLP and k-factors for individual patients or CT studies. Also, for the range of effective doses normally delivered by diagnostic CT scans (≈3-10 mSv), the stochastic risk of future cancer for an individual can be estimated but cannot be calculated precisely. There are limited data regarding the health risks of effective doses less than 100 mSv. Where relative risk comparisons are needed between CT and other types of exposure, other imaging studies, or CT of different parts of the body, effective dose (based on DLP and k-factor) can be a useful tool. For example, such a comparison might be made to show the magnitude of radiation-related risk for a CT scan of the chest versus a chest radiograph by determining the effective dose (in mSv) for both procedures.

Size-Specific Dose Estimate.  The size-specific dose estimate (SSDE) has been developed recently as a way to relate CTDIvol to patient dose based on body habitus.6 As detailed in American Association of Physicists in Medicine (AAPM) Report 204, the CTDIvol for the scan is scaled by a conversion factor based on the patient size to arrive at an SSDE, expressed in absorbed dose units (mGy). The SSDE conversion coefficient value can be chosen based on the lateral width or A/P thickness of the body part being imaged, or the sum of the two measurements (if available). The conversion values in AAPM Report 204 can be used to convert CTDIvol values referenced to the 32-cm CTDI phantom. SSDE is a useful tool for reviewing exam dose values; by finding the SSDE value, one can determine whether an abnormally high or low CTDIvol value is truly a cause for concern or whether it merely reflects

27

the body habitus of the patient.51 Used prospectively the SSDE value can be established as a target or reference dose value when constructing protocols; working backward, the scanner protocols can be defined for different patient body sizes using different CTDIvol targets. This is especially useful for pediatric protocols, given the large variation in body habitus from infants through teenagers.33 As of this writing, SSDE is a relatively new method, and some commercial software tools have been released that can collect CTDIvol information and CT images from the scanner and automatically compute SSDE values for individual exams. It is likely that in the near future, CT scanner manufacturers will develop the capability for the scanner software to calculate and display SSDE values and report them in the DICOM Radiation Dose Structured Report (RDSR). CT scanner software may also gain the ability to display a prospective SSDE before a scan, based on a patient size estimate from the scout or topogram image as well as the x-ray settings planned for the scan (which determine the CTDIvol).

Radiation Dose Management in CT Radiation dose management is accomplished through approaches that fall into three broad categories: appropriateness and utilization management, procedure and practice-based approaches, and technologybased approaches. All these approaches are aimed at improving the benefit/risk ratio for CT by increasing the benefit, reducing the risk, or (preferably) both. Appropriateness-based utilization management is a tool for reducing the number of unnecessary CT scans performed on patients. Here, “unnecessary” means CT scans that are unlikely to provide information that will affect the patient’s diagnosis or treatment. There are many reasons for such scans to be performed, and there is evidence that they are performed quite commonly in the United States. Some scans are ordered where the patient’s clinical presentation does not provide valid indication for CT. Some scans are ordered by clinicians who are unaware of or do not have access to recent (and useful) imaging of the same body part performed by another physician or institution. In all cases these unnecessary scans reduce the benefit/risk ratio of CT in two ways. First, they increase the radiation dose being delivered to the patient (thus increasing the risk). Second, they increase the number of scans that have to be performed to find one abnormal result or finding, decreasing the effective benefit of each scan. For example, if the rate (or probability) of head injury in patients presenting to the emergency department with specific complaints is constant but emergency department physicians become less selective about which of these patients are referred for head CT, a ratio of 1 finding per 10 patients scanned can quickly degrade to 1 in 20, halving the benefit for emergency head CT in the hospital. At the same time, many more patients are being exposed to the radiation dose associated with the head CT study. In this scenario, some of the scans are likely to be repeated exposures of individual patients who frequently visit the hospital for the same complaint. An effective approach to combat inappropriate utilization is to require the use of evidence-based guidelines for imaging appropriateness. This helps ensure that patients receive an adequate clinical exam and workup prior to being scanned and that the scan is more likely to yield information that will help in the patient’s diagnosis or treatment. There are extensive guidelines available for appropriate ordering of CT as well as other radiologic procedures to maximize benefit/risk for specific clinical scenarios. There are also a variety of software tools that integrate these guidelines into electronic medical information systems such as the Radiology Information System (RIS) or Electronic Medical Record (EMR). Some institutions have successfully implemented such

28

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

decision support tools, but in many places there is pressure from nonradiologists to delay or suspend use of such tools because they are perceived to limit physician autonomy and clinical judgment in ordering imaging exams. Procedure-based and practice-based approaches to optimization involve cyclical efforts to monitor and assess radiation dose and image quality and update imaging protocols. The details of these approaches are discussed in detail in Chapter 2. Technology-based approaches to CT optimization are an ongoing topic of interest and commercial development by CT equipment vendors. Examples of successful approaches include tube current modulation, automated selection of scanning parameters such as tube voltage, and advanced reconstruction methods such as iterative reconstruction to enable the user to achieve desired image quality at reduced radiation exposure. CT manufacturers have released several generations of products with progressively more sophisticated tools  to optimize radiation output from the scanner and match it to  the patient and the clinical imaging task. More recently the manu­ facturers, individually and through participation in the Medical Imaging Technology Alliance (MITA), have developed new tools for reporting and recording radiation exposure information in greater detail and helping the operator of the scanner remain aware of and manage radiation dose. The “Dose Check” feature allows scanners to be programmed with prospective alerts (globally and per protocol)  to notify the operator when an unusually high exposure is about to be performed due to automatic or manual changes from the default protocol settings. The DICOM RDSR standard provides a means  for manufacturers to report very detailed information about each activation of the x-ray tube during an exam. This supports the data collection component of continuous quality-improvement efforts in protocol optimization. A number of commercial products have been developed to automatically collect radiation dose information (e.g., CTDIvol and DLP values) from CT scanners, store it in a database, and provide user analytics to establish baselines and perform ongoing monitoring; these products take advantage of the availability of scanner data in the DICOM RDSR format. The American College of Radiology has introduced a Dose Index Registry for CT facilities to collect data from numerous hospitals and healthcare centers across the United States to be compiled into a master database from which benchmarks can be extracted for individual exam types by geographic region/location and facility type/size. With such a registry an academic medical center in New York could compare its CTDI and DLP values for any number of exams with the values for the same exam(s) performed by other hospitals in New York, for example, or with other academic U.S. medical centers across the country. Such benchmarks can harness the collective efforts and results of many radiologists and medical physicists who have worked to optimize their protocols locally, in isolation or perhaps in cooperation with a small number of collaborators, using conventional manual methods of incrementally modifying protocols and qualitatively verifying image quality for a number of patient exams at each step. A comprehensive approach to radiation dose management in CT includes all of the following essential elements: • Building and maintaining protocols that are clearly identified for specific clinical purposes • Establishing a team consisting of radiologists, technologists, and medical physicists to work collaboratively on efforts affecting CT radiation doses • Properly maintaining and calibrating all scanners in accordance with manufacturer requirements, regulatory requirements, and any applicable accreditation standards

• Understanding and correctly using all of the dose-management or dose-reduction tools available on each scanner • Identifying and reducing or eliminating unnecessary CT exams • Collecting and analyzing dose and image quality data from patient exams to identify improvement needs or opportunities • Educating patients and referrers about the benefit/risk for CT and the proactive efforts undertaken by the department to optimize imaging performance, such as accreditation, quality control, and protocol optimization.

SUMMARY Rapid improvements in the speed and spatial resolution of CT have driven the popularity of this imaging technique, resulting in a rapid increase in CT scan volume from around 1990 through 2015. This increase has fueled concerns about radiation dose to patients, and in response, CT manufacturers have developed dose-reduction technology while imaging professionals and other commercial entities have developed methods and tools for optimizing, monitoring, and managing CT performance in terms of dose and image quality. Scanners with fast high-resolution coverage of large areas of anatomy are expanding the roles of functional applications such as CT angiography and perfusion, while dual-energy and spectral techniques are also advancing the role of CT beyond anatomic evaluation. As the growing capabilities of CT encourage its adoption for more and new clinical applications, CT will also remain the tool of choice for routine cases that require crosssectional or 3D imaging information from a fast, economical diagnostic test that is relatively safe and noninvasive. Many more clinical applications are on the horizon; CT is both a well-established mainstay of diagnostic imaging and a platform for innovative approaches with a promising future. CT is perhaps the single most cost-effective, value-driven diagnostic imaging tool available. Examples of many of the new CT clinical applications are described in the chapters throughout this book.

REFERENCES 1. Althen JN: Automatic tube-current modulation in CT—A comparison between different solutions. Radiat Prot Dosimetry 114:308–312, 2005. 2. Beister M, Kolditz D, Kalender WA: Iterative reconstruction methods in x-ray CT. Phys Med 28:94–108, 2012. 3. Bogdanich W: After Stroke Scans, Patients Face Serious Health Risks. New York Times 2010. Newspaper article available online at: http://www .nytimes.com/2010/08/01/health/01radiation.html?_r=0. 4. Bogdanich W: West Virginia Hospital Overradiated Brain Scan Patients, Records Show. New York Times 2011. Newspaper article available online at: http://www.nytimes.com/2011/03/06/health/06radiation.html. 5. Bogdanich W, McGinty JC: Medicare Claims Show Overuse for CT Scanning. New York Times 2011. Newspaper article available online at: http://www.nytimes.com/2011/06/18/health/18radiation.html. 6. Boone JM, Strauss KJ, Cody DD, et al: Size-Specific Dose Estimates (SSDE) in Pediatric and Adult Body CT Examinations: Report of AAPM Task Group 204, College Park, MD, 2011, American Association of Physicists in Medicine. 7. Brenner DJ, Hall EJ: Computed tomography—An increasing source of radiation exposure. N Engl J Med 357:2277–2284, 2007. 8. Cody DD: AAPM/RSNA physics tutorial for residents: Topics in CT. Image processing in CT. Radiographics 22:1255–1268, 2002. 9. Cohen M: Cancer risks from CT radiation: Is there a dose threshold?   J Am Coll Radiol 10:817–819, 2013. 10. Eller A, May MS, Scharf M, et al: Attenuation-based automatic kilovolt selection in abdominal computed tomography. Invest Radiol 47:559–565, 2012.

CHAPTER 1  Imaging Principles in Computed Tomography 11. Fessler JA: Iterative methods for image reconstruction. In IEEE Int Symp Biomed, Arlington, VA, 2006, Imaging Tutorial. 12. Fessler, JA: Iterative image reconstruction for CT. Conference Proceedings: IEEE 2006 International Symposium on Biomedical Imaging, April 6–9, 2006, Arlington, Virginia, USA. 13. Fishman E, Ney D, Heath D: Volume rendering versus maximum intensity projection in CT angiography: What works best, when, and why. Radiographics 26:905–923, 2006. 14. Flohr T, Stierstorfer K, Raupach R, et al: Performance evaluation of a 64-slice CT system with z-flying focal spot, RöFo—Fortschritte Auf Dem Gebiet Der Röntgenstrahlen Und Der Bildgeb. Verfahren 176:1803–1810, 2004. 15. Flohr TG, Stierstorfer K, Ulzheimer S, et al: Image reconstruction and image quality evaluation for a 64-slice CT scanner with z-flying focal spot. Med Phys 32:2536–2547, 2005. 16. Funama Y, Taguchi K, Utsunomiya D, et al: Combination of a low-tubevoltage technique with hybrid iterative reconstruction (iDose) algorithm at coronary computed tomographic angiography. J Comput Assist Tomogr 35:480–485, 2011. 17. Glover GH: Nonlinear partial volume artifacts in x-ray computed tomography. Med Phys 7:238, 1980. 18. Heuscher DJ, Vembar M: Reduced partial volume artifacts using spiral computed tomography and an integrating interpolator. Med Phys 26:276, 1999. 19. Hiriyannaiah HP: X-ray computed tomography for medical imaging. IEEE Signal Process Mag 14:42–59, 1997. 20. Hsiao EM, Rybicki FJ, Steigner M: CT Coronary angiography: 256-slice and 320-detector row scanners. Curr Cardiol Rep 12:68–75, 2010. 21. Hsieh J: Computed Tomography Principles, Design, Artifacts, and Recent Advances, Hoboken, NJ, 2009, Wiley. 22. Hu H: Multi-slice helical CT: Scan and reconstruction. Med Phys 26:5, 1999. 23. Huda W, Mettler F: Volume CT dose index and dose-length product displayed during CT: What good are they? Radiology C:236–242, 2011. 24. Huda W, Ogden K, Khorasani M: Converting dose-length product to effective dose at CT. Radiology 248:995–1003, 2008. 25. Huesman RH: The effects of a finite number of projection angles and finite lateral sampling of projections on the propagation of statistical errors in transverse section reconstruction. Phys Med Biol 22:012, 1977. 26. Jerri AJ: The Shannon sampling theorem & its various extensions and applications: A tutorial review. Proc IEEE 65:1565–1596, 1977. 27. Kalender WA: Physical performance characteristics of spiral CT scanning. Med Phys 18:910, 1991. 28. Kalender WA: Computed Tomography: Fundamentals, System Technology, Image Quality, Applications, Erlangen, Germany, 2011, Publicis Publishing. 29. Kalender WA, Perman WH, Vetter JR, et al: Evaluation of a prototype dual-energy computed tomographic apparatus. I. Phantom studies. Med Phys 13:334–339, 1986. 30. Kalra MK, Maher MM, Toth TL, et al: Techniques and applications of automatic tube current modulation for CT. Radiology 233:649–657, 2004. 31. Kalra MK, Naz N, Rizzo SMR, et al: Computed tomography radiation dose optimization: Scanning protocols and clinical applications of automatic exposure control. Curr Probl Diagn Radiol 34:171–181, 2005. 32. Katcha, JS, Schmidt JR: X-ray generator and slip ring for a CT system, US6975698 B2. 2005. 33. Kleinman PL, Strauss KJ, Zurakowski D, et al: Patient size measured on CT images as a function of age at a tertiary care children’s hospital. AJR Am J Roentgenol 194:1611–1619, 2010. 34. Kowalski G: The influence of fixed errors of a detector array on the reconstruction of objects from their projections. IEEE Trans Nucl Sci 24:2006–2016, 1977. 35. Kyriakou Y, Kachelrieβ M, Knaup M, et al: Impact of the z-flying focal spot on resolution and artifact behavior for a 64-slice spiral CT scanner. Eur Radiol 16:1206–1215, 2006. 36. Lee CH, Goo JM, Ye HJ, et al: Radiation dose modulation techniques in the multidetector CT era: From basics to practice. Radiographics 28:1451–1459, 2008.

29

37. Lee KH, Lee JM, Moon SK, et al: Attenuation-based automatic tube voltage selection and tube current modulation for dose reduction at contrast-enhanced liver CT. Radiology 265:437–447, 2012. 38. Mathews JD, Forsythe AV, Brady Z, et al: Cancer risk in 680,000 people exposed to computed tomography scans in childhood or adolescence: Data linkage study of 11 million Australians. BMJ 346:f2360, 2013. 39. McCollough CH, Bakalyar DM, Bostani M, et al: Use of Water Equivalent Diameter for Calculating Patient Size and Size-Specific Dose Estimates (SSDE) in CT: The Report of AAPM Task Group 220, College Park, MD, 2014, American Association of Physicists in Medicine. 40. McCollough C, Bruesewitz M, Kofler J: CT dose reduction and dose management tools: Overview of available options1. Radiographics 503–513, 2006. 41. McCollough CH, Zink FE: Performance evaluation of a multi-slice CT system. Med Phys 26:2223, 1999. 42. McNitt-Gray MF: AAPM/RSNA Physics tutorial for residents: Topics in CT. Radiation dose in CT. Radiographics 22:1541–1553, 2002. 43. Measurements NC: on R.P. and, NCRP Report No. 160: Ionizing Radiation Exposure of the Population of the United States, Bethesda, MD, 2009, National Council on Radiation Protection and Measurements. 44. Mehta D, Thompson R, Morton T, et al: Iterative model reconstruction: Simultaneously lowered computed tomography radiation dose and improved image quality. Med Phys Int 1:147–155, 2013. 45. Miéville FA, Gudinchet F, Brunelle F, et al: Iterative reconstruction methods in two different MDCT scanners: Physical metrics and 4-alternative forced-choice detectability experiments—A phantom approach. Phys Medica 29:99–110, 2013. 46. Miglioretti DL: The Use of Computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk. JAMA Pediatr 1, 2013. 47. Mulkens TH, Bellinck P, Baeyaert M, et al: Use of an automatic   exposure control mechanism for dose optimization in multi–detector row CT examinations: Clinical evaluation 1. Radiology 237:213–223, 2005. 48. Newton TH, Potts DG: Radiology of the Skull and Brain: Technical Aspects of Computed Tomography, St. Louis, 1981, Mosby. 49. Park YJ, Kim YJ, Lee JW, et al: Automatic tube potential selection with tube current modulation (APSCM) in coronary CT angiography: Comparison of image quality and radiation dose with conventional body mass index-based protocol. J Cardiovasc Comput Tomogr 6:184–190, 2012. 50. Pearce MS, Salotti JA, Little MP, et al: Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: A retrospective cohort study. Lancet 6736:1–7, 2012. 51. Seibert JA, Boone JM, Wootton-Gorges SL, et al: Dose is not always what it seems: Where very misleading values can result from volume CT dose index and dose length product. J Am Coll Radiol 11:233–237, 2014. 52. Sharpless RB, Brunnett WC: Aerostatic rotor bearing, US6404845B1. 2002. 53. Shrimpton P, Wall B: Effective dose and dose-length product in CT. Radiology 250:321–322, 2009. 54. Smith-Bindman R, Lipson J, Marcus R, et al: Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med 169:2078–2086, 2009. 55. Snyder DL, Cox JRJ: An overview of reconstructive tomography and limitations imposed by a finite number of projections. In Ter-Pogossian MM, Phelps ME, Brownell GL, et al, editors: Reconstruction Tomography in Diagnostic Radiology and Nuclear Medicine, Baltimore, MD, 1977, University Park Press. 56. Söderberg M, Gunnarsson M: Automatic exposure control in computed tomography—An evaluation of systems from different manufacturers. Acta Radiol 51:625–634, 2010. 57. Sohval AR, Freundlich D: Plural source computerized tomography device with improved resolution, US 4637040 A, 1987. 58. Stayman JW: Iterative Reconstruction Methods in Computed Tomography. 1–24, 2012. American Association of Physicists in Medicine (AAPM) annual meeting, Charlotte, North Carolina, USA, August 2, 2012.

30

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

59. Walters RG, Carlson RW: Computerized tomographic scanner with shaped radiation filter, US 4288695 A. 1981. 60. Willemink MJ, de Jong PA, Leiner T, et al: Iterative reconstruction techniques for computed tomography. Part 1: Technical principles. Eur Radiol 23:1623–1631, 2013. 61. Willemink MJ, Leiner T, de Jong PA, et al: Iterative reconstruction techniques for computed tomography. Part 2: Initial results in   dose reduction and image quality. Eur Radiol 23:1632–1642, 2013.

62. Winklehner A, Goetti R, Baumueller S, et al: Automated attenuationbased tube potential selection for thoracoabdominal computed tomography angiography. Invest Radiol 46:767–773, 2011. 63. Wintersperger BJ, Nikolaou K, von Ziegler F, et al: Image quality, motion artifacts, and reconstruction timing of 64-slice coronary computed tomography angiography with 0.33-second rotation speed. Invest Radiol 41:436–442, 2006. 64. Yamada H, Suuzuki A, Uchida Y, et al: A scintillator Gd2O2S:Pr, Ce, F for x-ray computed tomography. J Electrochem Soc 136:2713, 1989.

2  Computed Tomography Imaging Operation Mark Patrick Supanich

PROTOCOL MANAGEMENT AND OPTIMIZATION Protocols are the turn-by-turn instructions that allow the computed tomography (CT) operator to get from point A—registering the patient on the scanner from the order in the radiology information system (RIS)—to point B—sending a complete set of images to the picture archiving and communication system (PACS) for a radiologist to interpret. To comply with the National Electrical Manufacturers Association (NEMA) XR-29 standard,27 all CT systems must be equipped with a set of reference adult and pediatric scan protocols. Although these default protocols are typically very reasonable for clinical use, many practices prefer to customize scanning protocols to best match radiologist preferences and scanning workflows. The term protocol can be used in many different contexts in radiology, so it is important to take a moment and offer several refinements to the broader term. These refinements are intended to be consistent with the anticipated definitions in the forthcoming (at time of  publication) Digital Imaging and Communications in Medicine (DICOM) Supplement 121 on CT Procedure Plan and Protocol Storage SOP Class. The procedure referred to in all the listed refinements is the imaging procedure that is indicated by the order placed by a physician: Defined procedure protocol: a description of the parameters and settings for CT image acquisition and reconstruction for a requested procedure. The defined procedure protocol is not patient specific and may not even be scanner specific insofar as a defined procedure protocol may include scanning and reconstruction instructions for a variety of scanners. The defined procedure protocol is equivalent to what some institutions may call their “master protocol.” Planned procedure protocol: contains patient-specific acquisition parameters and scan settings. The planned procedure protocol is developed by applying the defined procedure protocol to a specific patient and taking into account any patient or imaging task-specific requirements. Performed procedure protocol: the set of imaging parameters actually used to scan the patient. The performed procedure protocol is inextricably linked with the patient the exam was performed on. Scanner protocol: not part of the anticipated DICOM protocol SOP class, but a term useful to define. The scanner protocol is the protocol defined and named on a specific CT scanner that matches the defined procedure protocol. Scanners may have different naming conventions or requirements and different technical specifications, so not all scanner protocols may match in name or imaging parameters, but they should be within the specifications described in the matching defined procedure protocol.

Use of Protocols in an Imaging Operation Under normal clinical operations—excluding exams ordered and performed in an emergency setting—a referring physician (primary care or specialist) orders a radiology imaging procedure for a patient. This order may be printed out at the clinic and given to patients to take to a radiology service provider of their choice. Alternatively, if the referring physician has a relationship with a radiology group or is part of a hospital or healthcare network, the order may be entered into the patient’s electronic medical record (EMR), and the radiology practice may contact the patient directly to schedule the imaging procedure. An order may appear in one of many forms. It may be as sophisticated as an electronic order entered through a radiology order entry software interface with decision support incorporated or as low tech as a faxed handwritten order with the patient’s history and indications for the exam. The United States passed and enacted the Protecting Access to Medicare Act of 2014, which requires use of physician-developed appropriateness criteria when ordering advanced imaging exams (including CT) for full reimbursement of studies completed on patients covered by Medicare starting in 2017. The implementation and use of clinical decision support tools (including appropriateness criteria) by physicians ordering advanced imaging exams (including CT) is a key challenge hospitals and imaging centers will need to address. The goal of linking Medicare reimbursement rates to the use of appropriateness criteria is to increase adoption of decision support systems that aid clinicians in selecting the most appropriate exam for a patient’s medical history; this can reduce the instances of unnecessary or repeated scans.17,18 As shown in Figure 2-1, clinical decision support tools can guide physicians in selecting the appropriate imaging exam when integrated with an electronic order entry system.

Defined Procedure Protocol Selection.  The order placed by the referring physician for the patient to be imaged must be translated from the order to a procedure code as defined by DICOM and eventually to a defined procedure protocol specific to the radiology service provider. This translation can be done manually or by using lookup tables that are included in a software program. Radiology service providers maintain some sort of master procedure book that includes the practice-specific procedure code long and short names and the Current Procedural Terminology (CPT) codes associated with that procedure code. When an order comes in from a referring physician, it may or may not conform to the naming conventions of the master procedure book of the radiology practice. If the radiology practice uses, and the ordering physician has access to, an

31

32

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C FIG 2-1  Screen shots from an example software package providing clinical decision support. A, Radiology orders that may be selected by the ordering physician. B, Entry screen for the patient history and indications for the exam. C, Results for a sample clinical decision support scenario with guidance to the ordering physician if the ordered exam is not appropriate. (Used with permission from Ip IK et al: Adoption and meaningful use of computerized physician order entry with an integrated clinical decision support system for radiology: Ten-year analysis in an urban teaching hospital. J Am Coll Radiol 9:129–136, 2012.)

TABLE 2-1  Examples of Orderable Radiology Exams, Procedure Codes, and Nomenclature* Procedure Code

Procedure Name

Protocol Name

Playbook Long Name

Playbook Short Name

RPID

CTABDWO

CT ABDOMEN W/O CONTRAST CT ABDOMEN, PELVIS W CONTRAST CT MAXILLOFACIAL WO CONTRAST CT SINUSES WO CONTRAST CTA ABD AORTA W BILAT ILIOFEMORAL W CONTRAST

CT ABDOMEN

CT ABDOMEN WITHOUT IV CONTRAST

CT ABD WO IVCON

RPID3

ABDOMEN AND PELVIS (ENTEROGRAPHY) SINUS/MAXILLOFACIAL

CT ABDOMEN PELVIS ENTEROGRAPHY WITH IV CONTRAST CT MAXILLOFACIAL WO IV CONTRAST

RPID152

SINUS/MAXILLOFACIAL

CT HEAD PARANASAL SINUSES WITHOUT IV CONTRAST CT ABDOMINAL AORTA/ILIOFEMORAL RUNOFF ANGIO W AND WO IV CONTRAST BILATERAL

CT ABD PELVIS ENTERO W IVCON CT MAXILLOFACIAL WO CT HEAD SINUSES WO IVCON CT ABD RUNOFF ANGIO W/WO BILAT

CTENTERO CTFACWO CTSINWO CTAABDRUNW

CTA RUNOFF

RPID34 RPID62 RPID5995

*Examples of different naming conventions for radiology procedure codes, order names, and different names for the protocols are shown. Columns 1 and 2 present the radiology orderable in both short name and long name format. Column 3 is a defined procedure protocol associated with the radiology orderable in the same row. Columns 4 and 5 are the long and short Radlex Playbook names generated by converting the protocol name to a standardized nomenclature using the Radlex Playbook. The final column is the Radlex Playbook ID (RPID) assigned to the Playbook name.

electronic order entry system, then the received order will contain the practice’s procedure code. However, if the radiology practice does not use such software or the physician does not have access to it, the requested order may not match directly with the procedure codes  the practice uses. For example, a physician may order “CT Head” but

the radiology practice’s long name for this procedure code may be “CT Brain Without Contrast” with a short name of “CTHDWO.” Examples of radiology orderable exams and procedure codes along with translations to a standardized nomenclature (Radlex Playbook) are presented in Table 2-1. The Radlex Playbook is being used by the American

CHAPTER 2  Computed Tomography Imaging Operation College of Radiology (ACR) Dose Index Registry (DIR) as the common nomenclature for protocol naming for exams submitted to the registry.26 Using a common nomenclature and Radlex Playbook ID (RPID) facilitates comparison of dose indices for similar protocols across multiple institutions. If the order was placed without the use of electronic order entry, an individual must manually review the order and determine which of the practice’s procedure codes best matches the order. This step may be done by a customer service representative, a CT operator, or a radiologist. If there is any doubt whatsoever about which procedure code best matches the order, the referring physician should be contacted to verify the correct procedure code is selected so the required exam can be completed. Following identification of the correct radiology service provider– specific procedure code, the next step in translating the order is identification of the defined procedure protocol. Some practices may have a one-to-one correspondence between the procedure code and the defined procedure protocol. However, it is common for more advanced imaging studies to have multiple defined procedure protocols matched to a signal procedure code. Selection of the defined procedure protocol then will depend upon the indications for the exam and the patient history. For example, a physician’s order may be translated to the procedure code “CT Abdomen and Pelvis Without and With Contrast,” but the radiology practice may have multiple defined procedure protocols that can be used with that procedure code. One defined procedure protocol may require a precontrast CT of the entire abdomen and pelvis with a single delayed acquisition following injection of an iodinated contrast agent. Another defined procedure protocol matched to that procedure code may be a precontrast abdomen and pelvis study followed by a contrast injection with scans of just the abdomen at the arterial and venous and phases and a delayed abdomen and pelvis scan. Both the indications and information obtained for those two defined procedure protocols are very different, but they carry the same procedure code. If there is not a one-to-one match between the procedure codes used by a radiology service provider and the defined procedure protocols, a radiologist must review the patient history, the indications for the exam, and the requested procedure to assign a defined procedure protocol to the patient and the procedure code. This step is typically referred to as protocoling and is an important quality assurance check to ensure that the correct protocol with the correct parameters and timing of the acquisitions are used. Additionally, for radiology service providers operating in the United States and accredited by The Joint Commission, the 2015 revision to the Diagnostic Imaging Standards requires that the facility verify that the correct protocol is used to perform CT imaging exams. One way to demonstrate compliance for exams without a one-to-one correspondence between the procedure code and defined procedure protocol is to have a radiologist “protocol” the exam by selecting the appropriate defined procedure protocol.

Preimaging Protocol Verification.  Once the order for the imaging study arrives at the radiology service provider, a number of steps must be followed after the appropriate defined procedure protocol is assigned to the imaging study. The steps that follow may not be carried out in the particular order listed, but each one should be addressed before the patient arrives to have the imaging study performed. Prior to any of the steps, a “step zero” may occur in which a radiologist protocols the procedure requested for the patient. At this step, as described earlier, the radiologist reviews the order, patient history, indications for the exam, and appropriateness of the requested procedure and assigns a defined procedure protocol to the order. If a

33

radiologist does not protocol the order, there is typically a lookup table of some sort that matches the requested procedure to the defined procedure protocol of the radiology service provider: 1. Exam verification. The customer service representative or CT operator scheduling the exam must first verify that the practice is able to perform the requested service. For example, a physician may order a CT of the abdomen for a patient with a history of kidney stones and include in the comments for the order the request that the exam be performed using a dual-energy acquisition with postprocessing to assess the composition of the kidney stone. If the practice does not have a dual-energy CT scanner or only has it at a certain location, that information must be taken into account when scheduling the exam. 2. Insurance check. The radiology service provider will typically check with the patient’s insurer(s) to see if the imaging study needs preapproval or is automatically covered. 3. Scheduling. The imaging study is scheduled on the appropriate imaging modality and on a system that is able to perform the requested procedure. 4. Patient preparation. Instructions are relayed to the patient on how to prepare for the exam and could include how early to arrive for the exam or whether or not eating or drinking in a certain time period before the exam is allowed. 5. Patient screening. The patient may be asked questions and the patient’s medical record reviewed, depending upon the ordered exam, to ensure that performing the requested procedure is safe. Screening questions for CT exams are used to ensure that allergies or sensitivities to contrast agents are taken into account and to ensure that any medications the patient may require before the exam will be taken.

Patient Arrival and Preparation.  Many CT examinations require specific patient preparation steps before the study can proceed. Examples include consumption of contrast material or water several hours prior to an exam of the abdomen and pelvis to fill the bowels, placement of an intravenous (IV) line, or lab tests to confirm the patient is not pregnant or can tolerate the contrast agent used in the exam. Some of these preparations may require the patient to arrive at the clinic well before the scheduled scan start time or for the results of lab tests to be communicated in a timely manner to the facility performing the examination. Once the patient is at the clinic and has gone through the necessary preparation steps, they may be asked to change out of some or all of their clothes. A key requirement is that any metal objects the patient is wearing that will be in the scan field of view (FOV) be removed if possible. Objects including necklaces, earrings, belts, and even eyeglasses should be removed and placed nearby to return to the patient after the scan. There is some debate as to whether or not removal of brassieres is required for examinations of the thorax area. Image quality and breast tissue radiation dose may be improved by keeping the brassiere on—even if it has a metal underwire—especially on scanners using organ dose modulation techniques.31 An additional preparation requirement for studies using iodinated contrast agents is that the patient has a patent IV line of sufficient size to support the flow rate and pressure of the contrast injection. Alternatively, placement of an IV line can be skipped if the patient has a functioning port for the infusion of chemotherapy agents that can be used. Verification that the line used for the contrast injection is well positioned, has proper flow, and is large enough to tolerate the flow rate is key because poor placement or preparation of an IV line can result in contrast extravasation at the IV site, which can have significant negative patient outcomes.39

34

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

If the exam requires use of an iodinated contrast agent, the proper concentration of iodine must be loaded into the power injector by the CT operator, and the injection protocol for the indicated exam must be entered and confirmed at the power injector interface. Many institutions prescribe different amounts of contrast and contrast injection rates based on body habitus for adults.7 As with radiation dose, contrast injection protocols for pediatric patients should be correctly sized, and for infants may require use of a manual injection technique. Once the patient has gone through all the required preparation steps, including verification of the correct patient identity according to the policy and procedure at the facility performing the exam, the final preparation step of positioning the patient in the scanner may occur. The patient must either be transferred or place him/herself on the CT scan table in the correct orientation for the exam. Body examinations may require the patient to be positioned feet first toward the scanner, whereas head examinations may call for a head-first orientation. Most CT scans are performed with the patient in the supine position. However, some exams may call for the patient to be prone, such as certain acquisitions for virtual colonoscopies or for highresolution chest CTs. Positioning aids such as head holders, knee pillows, Velcro body fasteners, or breast straps may also be used to ensure patient comfort and optimal scan outcome. Additionally, if the examination calls for the patient’s arms to be above his/her head, an arm rest may be added for patient comfort and to limit movement during the exam. For examinations of the thorax and the abdomen, moving the arms to a position above the head ensures optimal image quality and radiation dose utilization. Some patients may not be able to move one or both arms above the head, and if that is the case, reconstruction artifacts and increased radiation dose should be expected. Limited patient arm movement may also affect proper vertical centering of the patient. For example, if the patient is only able to raise his or her arms so that the elbows are bent next to the ears, the table height may be limited because the elbows may touch the top of the gantry. Once positioned properly, the patient table is advanced toward the CT gantry, and laser lights to aid in the centering of the patient are turned on (Fig. 2-2). Correct vertical centering of the patient is vital to ensure that the anatomy of interest is as close to the scanner isocenter as possible. There are two main reasons proper vertical centering is key. The first is that by keeping the anatomy of interest as close to the

isocenter as possible, the reconstructed FOV can be kept as small as possible. This will ensure that the best possible in-plane spatial resolution is achieved. The second is that proper vertical centering is vital for the automatic exposure control (AEC) system of the CT scanner to work properly.14,15,37 Care must be taken to ensure that the patient is well centered laterally with respect to the gantry isocenter. The patient’s midline should be as close as possible to the middle of the scanner, because a patient mispositioned to the right or left of center will also interfere with the optimal functioning of the AEC. Mispositioning in this direction may also result in some anatomy at the extreme sides of the body not being included in the reconstructed image sets. The most extreme impact of lateral mispositioning is introduction of truncation artifacts if portions of the patient are outside the CT scan FOV.16,41 Proper functioning of the AEC is an important driver of proper patient positioning vertically and laterally. This is in part because AEC systems are designed with the assumption that the patient is well centered, and situations in which assumptions are not met in CT scanning typically lead either to nonoptimal image quality or nonoptimal radiation dose. A vertically mispositioned patient will result in a CT localizer radiograph that either magnifies (if positioned above isocenter) or shrinks (if positioned below isocenter) the apparent size of the patient (assuming the tube is positioned above the patient for the localizer radiograph; the resulting magnification/minification scenarios are reversed if the tube is below the patient). Figure 2-3 shows an example of the change in apparent size of a patient due to vertical mispositioning. Many CT scanners use the size of the patient as determined by the CT localizer radiograph to inform the operation of the AEC system.  If the patient is positioned below isocenter, the AEC system may adjust the planned radiation dose to fit the smaller apparent patient size of the localizer radiograph instead of matching the dose to the actual patient size. The result of this vertical centering error can be an increase in image of noise of 20% for a positioning error of 6 cm.14 A clinical example of the result of a patient positioned too low in the gantry is shown in Figure 2-4. Positioning the patient above isocenter can lead to significant increases in patient radiation dose37 because the AEC system may adjust the dose modulation pan to fit the larger apparent size of the patient from the radiograph. Figure 2-5 presents an example of the result of a phantom purposely positioned too high in the gantry; the radiation dose used for the scans increased by 50% when the phantom was positioned above the gantry isocenter. A final consideration in patient positioning in the scanner is whether or not gantry angulation will be used for the scan. The gantry may be angled to adjust the angle of incidence of the x-ray beam on the patient during the scan. This is most typically done for head exams to limit scanning through dental work (which causes artifacts) and to avoid irradiating the eye orbits in brain studies, because the eye lenses are susceptible to radiation-induced cataracts.40 If gantry angulation is not available or not possible for the exam and the patient is able to comply, the patient’s chin may be tucked down toward the chest to introduce a virtual angulation to limit eye irradiation and scanning through any metal dental implants.

Patient Registration.  In the course of normal CT imaging opera-

FIG 2-2  Gantry laser lights are used to properly position the patient vertically and horizontally in the CT scanner bore.

tion, the order for the patient’s study will be entered into the institution’s RIS. The RIS contains information not only about the order for the imaging exam but also typically links with the institution’s EMR software to include patient demographic information.23 Shortly before or after the patient enters the CT scan room the CT operator must open a patient registration window, an example of which is shown in Figure 2-6. If the order for the patient exists in the RIS, the patient demographics are typically automatically loaded into the appropriate

CHAPTER 2  Computed Tomography Imaging Operation

42 cm

35

47 cm

FIG 2-3  Correct (left) and incorrect (right) vertical positioning of the patient and the impact on the localizer radiograph. Correct positioning of the patient in the scanner results in an apparent lateral dimension of 42 cm as measured on the radiography. The patient diameter is measured to be 47 cm in the right image because the patient was positioned above the isocenter for an AP localizer radiograph exposure. The geometric magnification resulting from this miscentering results in a larger apparent patient size that may affect performance of the scanner AEC and result in a higher-than-required radiation dose. (Used with permission from William Pavliceck, Mayo Clinic, Scottsdale, Ariz.)

FIG 2-4  Two reconstructions of the same image acquisition. The image on the left shows the patient centered in the reconstructed FOV but demonstrates higher-than-acceptable image noise. The image on the right shows a reconstructed FOV centered at the scanner isocenter instead of on the patient and demonstrates that the patient was positioned too low in the gantry. The mispositioning of the patient resulted in a smaller apparent patient size on the localizer radiograph, which may affect the AEC operation and result in lower-than-required radiation dose for the desired image quality. (Used with permission from William Pavliceck, Mayo Clinic, Scottsdale, Ariz.)

fields in the registration window. After confirming all the information in the registration fields is correct and entering in any additional information (e.g., institution name, study type) the operator can complete the registration, and the workflow at the CT console moves onto the protocol selection.

Protocol Selection and Verification at the Scanner.  Once the patient is positioned and registered, the CT operator next must select

the scan protocol on the system that matches the defined procedure protocol. Selection of the protocol on the scanner is of vital importance to ensure the proper radiation dose, appropriate timing of contrastenhanced scans, and correct reconstructions are used. Operators should have access to a protocol book with a one-to-one correspondence between defined procedure protocols and scanner protocols to ensure the correct protocol for the diagnostic task at hand is selected. Matching scanner protocols across different CT brands and models is

36

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

FIG 2-5  The impact on the radiation dose for scans using AEC of improper patient positioning. The image on the left is from a patient positioned too high in the gantry, with a localizer radiography acquired in the AP direction. The CTDIvol required for this examination was 18 mGy. The image on the right is the phantom correctly positioned vertically in the gantry, with a CTDIvol of 12 mGy required. (Used with permission from William Pavliceck, Mayo Clinic, Scottsdale, Ariz.)

FIG 2-6  Screen shot of a typical patient registration screen from the scanner console. If a patient is scheduled in the RIS worklist, the patient may be selected from a list, and a number of the fields may be automatically populated.

CHAPTER 2  Computed Tomography Imaging Operation

37

FIG 2-7  The CT operator is presented with a protocol selection screen following patient registration. This screen shot shows an example of the protocol selection screen. The operator will typically be allowed to select from an adult or pediatric set of protocols and select one from a group of different body regions.

addressed later. Following selection of the proper protocol(s) as shown in Figure 2-7, the CT scan workflow moves to the next step of acquiring the localizer radiograph.

Performance of Localizer Radiograph.  A localizer radiograph (or a scanned projection radiograph) is performed before any CT acquisitions occur. Each CT manufacturer has a branded name for the localizer radiograph, and the CT Lexicon developed by a committee of the American Association of Physicists in Medicine (AAPM) contains the manufacturer-specific names for the localizer radiograph as well as many other common CT terms.1 The localizer radiograph is acquired in a fixed position as the table moves through the gantry. With the x-ray tube stationary and on while the patient moves through, a lineby-line x-ray image of the patient is acquired. Localizer radiographs are acquired either with a lateral view or with an anteroposterior (AP) or posteroanterior (PA) view. For some studies, both views may be obtained to ensure proper patient positioning. When lateral and AP/ PA localizer radiographs are both acquired, the lateral view is usually acquired first to ensure that the patient is well centered vertically in the scanner. If the patient is not well positioned, the table height may be adjusted before proceeding with the AP/PA localizer radiograph.

The localizer radiograph is used to plan the anatomic coverage of the CT exam and is used by the scanner’s AEC system to adjust the radiation dose used for the exam. For both of these reasons, the localizer radiograph must cover all the potential anatomy of interest and must use a technique to properly penetrate the patient. An under- or overexposed localizer radiograph will not allow visualization of all the patient’s anatomy, which can make planning the extent of the exam challenging. For example, if the localizer radiograph for a very large patient is acquired with the default technique, it may be impossible to identify the extent of the liver, which may lead to portions of the liver not being included in a CT exam focused on the liver or unnecessary irradiation of anatomy not relevant to the exam. Figure 2-8 demonstrates the difficulty in visualizing the anatomic structures of a patient acquired with an insufficient localizer radiograph technique. Another pitfall of incorrect localizer acquisition is demonstrated in Figure 2-9. In this case the localizer radiograph for a chest x-ray did not cover the apices of the lung, which would lead to the AEC system using more radiation than required, had the full extent of the scan range been covered. Following the localizer radiograph(s), the CT operator may proceed with prescribing the extent of the CT scan coverage. Common to all CT platforms is a transparent box that is shown over the CT localizer

38

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

FIG 2-8  The importance of choosing proper acquisition parameter settings for the localizer radiograph is demonstrated in these images. On the left an average-sized patient is irradiated with the default localizer radiograph technique, and the anatomy is well visualized. On the right the same technique settings are used to acquire a localizer radiograph for a larger patient, and visualization of the patient’s anatomy is limited.

FIG 2-9  A result of not including the entire area to be imaged in the localizer radiograph. The range for the CT scan is indicated by the closed brackets and the variation of the effective tube current over the imaged area is shown by the red line. Because the localizer does not extend below the dashed green line, no information about patient size or attenuation is available and the effective tube current defaults to the maximum available value.

radiograph. This box may be extended, shrunk, or moved to cover the required anatomy of interest for the protocol. This scan plan box defines the extent of the images to be reconstructed once the scan is completed and also sets the reconstructed FOV. If multiple acquisitions are required for the selected protocol, each irradiation event (or scan) has its own planning box. For example, a multiphase chest, abdomen, and pelvis study may have one planning box covering the entire torso, a second phase with the box covering the abdomen and pelvis, and a third acquisition planned only to cover the liver. Figure 2-10 shows two planning boxes for a chest and abdomen study on a phantom. Inconsistent or incorrect localizer radiograph acquisitions can have significant negative impacts on CT scan image quality. It is therefore recommended that acquisition of localizer radiographs be standardized as much as possible and CT operators be educated on the importance of the localizer radiograph to the proper functioning of the system AEC. Different localizer radiograph protocols may be configured for patients of different body habitus, because additional radiation may be required to adequately visualize the patient anatomy for large patients, and less radiation can be used for small patients. In the eyes of the operator, the localizer radiograph is first and foremost the means to visualize the patient anatomy and plan the scan to cover the required areas, but to the scanner software the localizer is the data it uses to plan the radiation dose used for the scan. Figure 2-10 shows this on the left, where the planned effective milliampere-seconds (mAs) for a scan that is yet to occur is shown alongside the localizer radiograph. As the attenuation of the area being scanned increases or decreases as measured by the localizer radiograph, the planned effective milliampere-seconds also changes. Also shown in Figure 2-10 is the prescan estimated average volume CT dose index (CTDIvol), which is displayed following the localizer and changes as the planned scan FOV is adjusted.

Review Planned Procedure Protocol.  Following acquisition of the localizer radiograph(s) and the definition of the start and end

CHAPTER 2  Computed Tomography Imaging Operation

39

FIG 2-10  Example of the scan planning screen a technologist sees before initiating the CT scan. The “patient” being scanned is a set of anthropomorphic phantoms. The planned scan regions for the Thorax scan is shown overlaying the phantom as the blue box, and the planned scan region for the Abdomen WO scan is displayed as the gray box. The operator may adjust the size and scope of the planned scan region to cover the required anatomy.

locations for the exam, the expected scan acquisition parameters will be available for the CT operator to review. Chief among these is the planned CTDIvol for the acquisition. Confirming that the planned CTDIvol is appropriate for the exam type and patient body habitus is an important practice to ensure that patients receive the appropriate radiation dose for the exam. Anomalous low or high values of the planned CTDIvol should result in a timeout to review the scan parameters and ensure the scan will proceed as expected. The dose-check standard that is one component of NEMA XR-2927 requires CT scanners to produce popup notifications if the planned CTDIvol or doselength product (DLP) exceeds a defined reference level. Configuring protocol-specific dose reference levels and communicating those values to CT operators is the most prudent way to ensure that patients are not exposed to unnecessarily high levels of radiation for the selected protocol and given the patient’s size. Of equal importance is educating CT operators to be aware of planned scans that have uncharacteristically low planned CTDIvol values, because these studies will result in poor reconstructed image quality that may render the exam nondiagnostic. In addition to reviewing the planned CTDIvol for the exam, it is also prudent for the CT operator to check other key scan acquisition

parameters such as the tube potential, scan time, and average effective milliampere-seconds. Confirmation of the correct scan parameters prior to initiating CT acquisition is an important timeout to verify that the appropriate radiation dose is used and the required image quality is obtained. CT operators may be assisted in confirming the correct scan parameters or planned CTDIvol by having reference tables of expected CTDIvol ranges for different patient sizes and different protocols. Alternatively, if a prescan estimate of the size-specific dose estimate (SSDE)9 is available, a table of expected SSDE vales for different scan protocols may be reviewed. SSDE is a dose index that incorporates patient size or attenuation. Correction factors based on patient size are applied to the average CTDIvol value for the scan. As described by Boone et al., patients with less attenuation have higher correction factors than larger patients because smaller patients absorb more radiation dose than larger patients exposed to the same radiation output from a CT scan. The use of SSDE as a dose index can help remove some of the significant variations in CTDIvol observed for patients of different body habitus. If any acquisition scan parameters require adjustment, it is important to confirm that the change does not result in an undesirable adjustment of another scan parameter. For example, when scanning a

40

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

large patient using AEC, it is possible that the maximum effective tube current time product may be reached. Adjusting the rotation time or pitch may allow for an increase in the maximum effective tube current time product, resulting in improved image quality. However, adjusting the pitch or rotation time will result in an increased total scan time, which may be undesirable for some angiography studies. If adjusting a scan parameter such as pitch, rotation time, and tube current while using AEC, other parameters may change automatically to maintain the desired image quality and radiation dose. If the image-quality reference parameter is changed prior to the scan acquisition, the radiation dose and image noise of the acquisition will change. When scanning using a manual technique with user-defined values for tube potential, tube current, rotation time, pitch, and so forth, a change in one parameter will result in a change in the radiation dose and image quality. The full list of technique parameters that may be adjusted by the operator and affect the image quality and radiation output for a scan are detailed in Table 2-2. The AAPM CT Dose Education Slides3 describe these parameters in detail as well as the relationship between a change in the scan parameter and the planned CTDIvol for a study.

TABLE 2-2  Adjustable Acquisition

Technique Parameters* Acquisition Parameter/Setting Target tube potential Dose range

Detector configuration options Pitch options Tube rotation time options Reconstruction kernel Reconstructed image thickness Reconstructed FOV

Scan region Multiplanar reformats

Use of iterative reconstruction

Use of AEC

Contrast amount and timing

Comment Target kV or range of acceptable kVs for the imaging task—may be size dependent Expected radiation dose range (CTDIvol, SSDE, or DLP) for the scan for at least averagesized patients and ideally for a variety of patient sizes Target minimum detector size and total collimation at the detector A range of acceptable pitches A range of acceptable tube rotation times Characteristics of the kernel for each reconstructed image series Target range of image thickness for each reconstructed image series Instructions for size of FOV to be reconstructed for each image series and any reformats Target start and end anatomic locations for each image series Description of the orientation and anatomic range of any multiplanar reformats for each image series Indication if iterative reconstruction techniques should be used on any image series and (if applicable) the strength of the iterative reconstruction setting Indication if AEC should be used for image acquisition and if so, what types of AEC may be used Type of IV contrast, amount (or habitus-based protocol to follow), and injection rate

*Representative examples of acquisition and reconstruction parameters and settings that should be included in a defined procedure protocol are shown. A description of the scope or intent of each parameter or setting is also included.

Performed Procedure Protocol Review.  After confirmation that the scan parameters of the planned procedure protocol are correct, the operator may initiate the CT acquisition(s) prescribed by the protocol. If multiple acquisitions covering the same body part are planned, it may be prudent to review the reconstructed images of each phase before the next phase is scanned to ensure that the image quality is sufficient and no artifacts are present. A delay of 5 minutes between the arterial and delayed phases of an abdominal exam is sufficient time for the operator to review the images and note if any adjustments to the scan protocol may be required. Following completion of all acquisitions, the operator should review all the acquired images to ensure they are not degraded by patient motion or artifacts. If artifacts are present or patient motion is visible in the reconstructed images, the images should be transmitted to the institution’s PACS and the radiology practice or section responsible for interpreting the exam should be contacted to determine whether the study needs to be repeated. If all images demonstrate acceptable quality, the patient can be informed that the exam is completed and any leads or IV lines may be removed to allow them to be transported or moved out of the scan room. Additional image sets other than the standard axial reconstructions may be called for by the defined procedure protocol. The additional reconstructions using alternate reconstruction kernels (e.g., bone, lung) or different orientations (sagittal, coronal, oblique) should be completed after the primary acquisitions are reviewed and confirmed to be free of artifacts. All reconstructed image sets should then be transmitted to the institution’s PACS system and/or image processing workstation. Following the transmission, the CT operator must review the images on PACS to ensure that all reconstructions and images are present. Following this step, the operator may then change the status of the exam in the RIS to verified so that it can be interpreted and dictated by the radiology service provider.

Protocols and Scanning in an Emergency Setting.  The preceding discussion of the typical workflow for a CT exam is typical for non-emergency settings. In an emergency setting, the workflow and CT scan setup will be significantly different. Patients may arrive at the CT scanner without any identifying information or without an order, or they may be unable to move their arms into the required position for the exam. This section discusses some of the unique aspects of the practice of scanning in an emergency department (ED) setting. An extreme example of the unique challenges of CT scanning in an emergency setting is that of a patient arriving in the ED unconscious and unable to move, without any identification and in need of an immediate CT scan. One such situation is a patient transported to the hospital in an ambulance following a traumatic car crash. In these cases the patient may be brought directly to a CT scanner to determine the extent of the injuries. A typical trauma CT protocol will cover the patient’s head, neck, chest, and abdomen. In a non-emergency setting the patient would be instructed to drink bowel-opacifying contrast or water in advance of the abdominal exam, which is not possible in the emergent setting. If there is any question of a possible neck or spine injury, the patient will be immobilized, requiring that the chest and abdomen be scanned with the arms down instead of in the preferred position above the head. Furthermore, for true emergencies the operator may be required to scan the patient as a Jane or John Doe if no identifying information is available, and the order for the exam may be given verbally by a physician accompanying the patient to the scan room. Another common example that may require substantial deviations from the standard CT workflow is when a patient arrives at an ED exhibiting symptoms of a stroke. In such situations a brain CT scan is required as quickly as possible to determine if there is any blood or hemorrhaging in the brain. The presence of a bleed in the brain is a

CHAPTER 2  Computed Tomography Imaging Operation contraindication for administration of tissue plasminogen activator (tPA), which has been shown to reduce the effects of ischemic strokes if administered in a short time frame following the stroke.6 In cases of stroke, “time is brain,” and rapid determination of the cause of the stroke and the best treatment option is vital to limiting the long-term effects.30 Like trauma patients, stroke patients may arrive at the ED unable to communicate or without identification and may need to be scanned under a dummy name. Additionally, the need for speed in assessing the patient may also require that the CT scan be completed following a verbal order for the examination from a physician. High throughput, patients unable to cooperate, and the emergent nature of work in ED settings often results in a very different workflow than scanning in an outpatient or inpatient setting. However, by identifying and acknowledging the differences in the environment, an efficient and effective emergency CT section can be established that meets the needs of both patients and ED physicians.

Developing and Maintaining CT Protocols Given the vital role protocols play in successful acquisition of diagnostic images at the lowest achievable radiation dose, development and maintenance of the protocols are tasks that must not be marginalized or ignored. A number of accrediting bodies require regular review of CT protocols to ensure that they are up to date and representative of best practices. Development and standardization of protocols across multiple scanner makes and models can help ensure that patients scanned on different systems within the radiology practice receive imaging exams with similar image quality. Additionally, standardized protocols can aid in the development of hanging protocols in PACS to present the series acquired for each exam in a consistent manner to the radiologist, regardless of the system on which the exam was acquired. Development and maintenance of CT protocols is not a oneperson job and requires a team effort to be completed efficiently and effectively. Team members should at the very least include a lead radiologist familiar with similar current protocols and the relevant literature, a qualified medical physicist (QMP),2 and a lead CT operator. The team may also include a representative from the CT scanner manufacturer or a referring physician championing the addition of the new protocol. When the team is considering adjusting current protocols or developing new ones, an iterative approach of small adjustments to the protocol is highly recommended. Large changes in the AEC settings or manual technique may result in either nondiagnostic images or significant overirradiation of the imaging subject. Changes to the effective tube current time product or image-quality reference parameter on the order of 10% are recommended. A change of 10%, observed for a week or a reasonable number of exams can then be reviewed by the team to ensure that the image quality and radiation dose are reasonable for a variety of patient body habitus. Further 10% increases or reductions may then be applied as needed. In this manner the required image quality can be obtained at the lowest achievable radiation dose in a systematic manner.34

Developing CT Protocols.  Unless starting a radiology practice from scratch, a radiology service provider will typically have a set of master protocols or defined procedure protocols. However, as new technologies emerge and new types of exams are requested by referring physicians, it may be necessary to develop new scanning protocols. The principles outlined here can also be used to generate an entirely new set of defined procedure protocols or scanner protocols. Development and optimization of scanner protocols on a new CT scanner is covered in the next section. Development of a new defined procedure protocol may be initiated for a number of reasons. Common examples include frequent requests

41

from referring physicians to modify an existing protocol, results in the literature that indicate the utility of a protocol not currently in use at the practice, or the desire to create a modified protocol for a subset of patients. For instance, a new defined procedure protocol may be developed for patients in need of a head CT who have ventricular-peritoneal shunts. The indications and image-quality needs for head CTs on these patients may be different from those of routine brain CTs, and a separate defined procedure protocol may called for.5,33,38 A new defined procedure protocol should be matched with an existing radiology orderable, or a new orderable should be created with which the new protocol will be related. The components of the defined procedure protocol, slice thickness, target kilovolts, target CTDIvol for an average-sized patient, number of phases, and timing (if contrast enhanced) can be adopted from an existing protocol or adapted from protocol descriptions in the literature. New scanner protocols may also be adopted from existing scanner protocols, often by copying and renaming existing protocols and then adjusting scan parameters as needed. Care should be taken to ensure that each acquisition parameter and each acquisition series and any relevant breathing instructions  and contrast injection timing, volume, and rates are confirmed to be correct. If the new protocol will share an order code with an existing protocol or if the new protocol is likely to be assigned the same hanging protocol as another protocol, it is important to match the series names in each scanner protocol as closely as possible with the series names of the existing protocol. A nomenclature scheme such as the Radlex lexicon28 can be used to match series names and descriptions. Matching series names across scanner makes and models may be challenging because some CT software versions limit the number of characters that may be used or require underscores to be used instead of spaces between words, or disallow use of underscores. A concerted effort to match the names as closely as possible will result in unambiguous series descriptions and the best possible performance of any defined hanging protocols. When developing a new scanner protocol, at a minimum the default manual technique for the scanner protocol should be used to scan the appropriate head or body CTDIvol phantom to ensure that the radiation dose for the scan is reasonable. Ideally an anthropomorphic phantom of the head or torso should be used to test the protocol as if scanning an actual patient. In this case, if AEC is used for the scan, it can be enabled and allowed to adjust the tube current or tube potential for the acquisition. The performance of the protocol may be further interrogated if a set of anthropomorphic phantoms of different sizes are available. Test scans using the new protocol on different-sized phantoms can help ensure that the protocol will perform as expected under different clinical conditions. The observed or expected CTDIvol and DLP values for the protocol should also be compared to similar protocols and to relevant reference levels. As the new protocol is used clinically, the radiologist involved in building the new protocol should take special care to note and review cases in which the protocol is used. The image quality, timing of phases (if applicable), and radiation dose should all be evaluated and compared to similar protocols or expected values. If image quality is insufficient to meet the needs of the diagnostic task, the manual technique or AEC settings should be adjusted to deliver more radiation. Conversely, if the image quality is sufficient or exceeds the diagnostic task and the radiation dose is higher than expected given the patient size, the manual technique or AEC settings should be adjusted to use less radiation dose.

Protocol Development and Optimization on a New CT Scanner.  Many of the principles and approaches described thus far can be used when developing or optimizing protocols on a newly acquired and

42

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

installed CT scanner. However, some differences from one-off protocol generation—aside from the scope of the task—do exist and deserve highlighting. Pursuant to NEMA XR-29, all new CT scanners are required to ship with a set of reference adult and pediatric protocols.27 These protocols should generally be reasonable for clinical use and should be able to be safely used to deliver reasonable image quality and radiation dose. However, many institutions may desire to adjust the default protocols or generate new ones. The default scanner protocols may not encompass the scope of practice at a site or may be configured differently than the scanner protocols on other systems that share a common defined procedure protocol. Adjustments to the reconstruction kernels or the strength of the iterative reconstruction may need to be applied to achieve an image quality and noise texture that are acceptable to the radiologists interpreting the images. When purchasing a new CT system, several days (if not a week or 2) of onsite applications training is typically included. During this time, an applications specialist(s) from the manufacturer will be present at the imaging service site to work with the protocol team (radiologist, QMP, and lead technologist) to adjust any default scanner protocols and build any required new protocols. Typically the default scan protocols will remain on the system unchanged and user or custom protocols will be built to reflect any modifications or the need for new protocols. The onsite applications specialist(s) is typically a former CT technologist with significant experience building and adjusting protocols specifically for the manufacturer of the new CT system, so taking advantage of their expertise in developing and modifying new protocols can be extremely beneficial. As the new scanner is used clinically over the first week or 2 after the installation, it is ideal to schedule as many different types of examinations on the scanner as possible so the protocol team may review the image quality and radiation dose of as many protocols as possible to be able to take advantage of the applications specialist’s expertise if any adjustments are needed. After the end of the training period, the protocol team should continue to monitor image quality and radiation dose to adjust on their own. Alternatively the site may choose to bring the applications specialist back in for another round of onsite training after several months of clinical use to further refine the protocols, or online or telephone support offerings from the CT manufacturer may be used to achieve the same goal.

Acquisition Parameters to Include in Defined Procedure Protocol.  Defined procedure protocols should be stored electronically or as a hard copy in a location that is easily accessible by both CT operators and radiologists. The defined procedure protocol may be matched with scanner protocols from a variety of CT manufacturers, so using general terms to describe the various acquisition parameters is recommended. The AAPM lexicon offers a lookup table of manufacturer-specific names for different CT terms matched to a generic manufacturer-neutral term.1 The defined procedure protocol should include general scanning instructions such as the orders associated with the exam, patient preparation and positioning, and instructions on which reconstructed image sets to send to PACS. In addition the defined procedure protocol should include ranges or settings for a number of scan acquisition parameters. General terms for parameters it may be useful to include in the defined procedure protocol are listed in Table 2-2.

Matching CT Protocols Across CT Scanners.  Applying a defined procedure protocol to a wide array of CT scanners can be a very challenging task. The ultimate goal of matching CT protocols across scanners is to ensure that consistent image quality for the same defined

procedure protocol is available regardless of the system a patient is scanned on. Radiology service providers may face challenges such as matching protocols across CT scanners from different scanners, or matching scanners that differ in age and capability by 10 or more years. Exactly replicating the image quality of a head CT from a modern 256-detector-row scanner with iterative reconstruction on an older 4-detector-row scanner may be challenging, but there are several steps that can be taken to standardize image quality for a defined procedure protocol across CT scanners. While undergoing the process of matching protocols across scanners, it may also be beneficial to match protocol names and series description to a standardized nomenclature to facilitate the use of hanging protocols in a PACS environment. Matching image quality across scanners.  As described by Favazza et al, there are three image-quality metrics that may be measured and compared to best match protocols across scanners.21 Ideally a reference or target scanner protocol should be identified that produces high-quality diagnostic images at a reasonable radiation dose. By evaluating these image-quality metrics and adjusting the acquisition or reconstruction parameters to closely match the values of the metrics on the reference scanner, similar image quality across scanner makes and models may be obtained. It should be noted that different scanner makes and models may use different radiation dose levels  to achieve similar image quality. If a scanner is unable to achieve  image quality equivalent or close to that of the reference scanner, with a radiation dose below the institution’s reference level for a given patient size, it may be prudent to consider not performing that protocol on the scanner or adjusting the image-quality expectations of the protocol. The image-quality metrics described in the following sections should evaluated by a QMP and can be evaluated using commercially available phantoms such as the Gammex 464 or the CATPHAN 500 or 600 series. To fully interrogate the performance of the system, especially when iterative reconstruction is used, more specialized phantoms and analysis software may be required. However, for routine clinical practice, matching the spatial resolution, tomographic image thickness, and standard deviation (SD) of the image noise are easily assessed image-quality metrics that can be matched across scanners (Table 2-3). Spatial resolution or modulation transfer function.  Spatial resolution is an indicator of the ability to visualize small objects or sharp edges in the reconstructed image. The simplest way to evaluate spatial resolution in a reconstructed image is to determine the highest

TABLE 2-3  Image-Quality Reference

Parameters* Manufacturer

Image-Quality Reference Parameter

Action to Increase Image Quality and Radiation Dose

GE Hitachi Neusoft Philips Siemens Toshiba

NI (Noise Index) SD # Dose Right Step Dose Right Index Quality Reference mAs SD

Decrease Decrease Increase Increase Increase Decrease

*A list of the image-quality reference parameters used to determine the desired image quality of a CT image when using AEC from major CT manufacturers is shown. The effect of increasing the imagequality reference value on the scanner output (for imaging the same object) is shown in the final column. The image-quality reference parameter name, and its behavior may change depending upon the software version used by a scanner.

CHAPTER 2  Computed Tomography Imaging Operation frequency of line pairs that can be resolved in an image.22 In CT, images reconstructed with “sharper” kernels—such as those for small objects in the temporal bones or lungs—typically have better spatial resolution than images reconstructed for visualization of soft tissue. The increased spatial resolution is reflected by the ability to resolve more line pairs per unit length in a phantom. Ensuring that the images reconstructed for the same defined procedure protocol on different scanners have equivalent performance as measured by the number of line pairs per unit length resolved is one way that image quality across scanners can be standardized. A more rigorous test of spatial resolution can be performed by evaluating the modulation transfer function (MTF) as described  in various publications.13,29 The MTF presents a spatial frequency– dependent evaluation of the spatial resolution of the system. By comparing the MTFs of different reconstruction kernels on different systems, scanner protocols with similar MTFs may be configured to ensure that the protocols are similar in their ability to resolve image objects across all spatial frequencies. Tomographic section thickness.  Matching the tomographic image thickness or slice thickness across protocols is important for a number of reasons. First, matching the thickness across scanners ensures that the influence of partial voluming on the appearance of anatomic structures is similar regardless of the scanner used.8 Partial voluming is reduction of the apparent CT number of objects with a smaller extent than the thickness of the image object. Matching image thickness will also ensure that similar radiation dose is required to achieve the desired image quality in the reconstructed image set. For scans using AEC, the image-quality reference parameter set in the scanner protocol defines the desired image quality for the target reconstructed image thickness. If a thinner or thicker image thickness is selected for the acquisition, the amount of radiation dose required to meet the desired image quality will increase or decrease.24

Standard deviation or noise power spectrum of image noise.  Configuring scanner protocols to ensure that similar levels of image noise are obtained for the same defined procedure protocol is perhaps the most important step in matching protocols across scanners. The ability of radiologists to interpret an image may change if image noise varies from scanner to scanner. Image noise may be evaluated by measuring either the SD of the CT number or by measuring the noise power spectrum (NPS) from a set of reconstructed images. The simplest approach to evaluating image noise is to scan a phantom object with a section of a uniform CT number with the protocol being evaluated. The SD of the CT number of the region of interest in the uniform region is a measure of the image noise and can typically be evaluated at the scanner console.10,22,32 Adjusting the manual technique for the protocol or the AEC settings to closely match image noise measurement across scanners will help ensure that the image quality for all scanner protocols associated with the defined procedure protocols is consistent. If the adjustment results in patient or phantom scans that would lead to radiation doses exceeding the institution’s reference levels for that protocol, consideration should be given as to whether or not that protocol should be used on that scanner. As with spatial resolution, a more rigorous and spatial frequency– dependent assessment of image noise may be evaluated. The NPS of a scanner protocol may be measured by analyzing a set of images acquired on a uniform phantom using the protocol to be matched.11,13,20 The NPS allows for quantitative evaluation of the image noise texture. Images with equivalent noise SDs may look very different because the spatial frequency distribution of the noise is not evaluated by the SD assessment. Using the NPS to match scanner protocols can help verify that both the magnitude and texture of the image noise are similar across scanners.

43

Quality Assurance of Protocols.  After CT defined procedure and scanner protocols are developed or modified, the protocol team’s work is not complete. The protocols must be the subject of ongoing quality assurance to ensure they continue to deliver appropriate image quality with the lowest diagnostically achievable radiation dose.25 Annual reviews of the imaging protocols are strongly recommended and are a requirement of the quality assurance components of the accreditation programs of the American College of Radiology.4 Hospitals accredited by The Joint Commission in the United States are also required to review CT imaging protocols in a time frame defined by the institution. Even if a radiology service provider is not accredited by one of the two bodies, an annual review of the most commonly used protocols should be part of the radiology service provider’s ongoing quality assurance process, and an annual review of all protocols is consistent with best practices. To help guide and inform the process of protocol quality assurance, the AAPM developed and published a CT protocol management and review practice guideline in 2013.12 The practice guideline describes the staffing requirements for the CT protocol review team and best practice for the review process. Several key points identified in the practice guideline are the utility of a consistent naming scheme, limiting access to the ability to edit scanner protocols, and use of dose index monitoring software to establish facility reference ranges. The practice guideline also calls out six high-volume or high-radiation-dose protocols that must be reviewed annually if performed at the institution. The guideline recommends annual review of head and abdomen protocols for both adult and pediatric patients, the high-resolution chest protocol, and the CT brain perfusion protocol.12 A number of recent publications have described institutional experiences complying with the recommendations of the practice guideline and on the protocol review process and offer a number of useful suggestions to facilitate the protocol review process. Szcykutowicz and Siegelman suggest that a semiautomated approach using a database  of protocol parameters may offer an efficient workflow for quality assurance.36 Kofler et al. offer a number of practical implementation suggestions including prioritizing any protocols with identified imagequality issues or ones exceeding identified reference levels.19 A detailed review of what was required at one institution to comply with the recommendations of the AAPM practice guideline is included in a 2015 article by Szcykutowicz et al.35 Regular review of all aspects of CT imaging operations and identification of areas for improvement or modernization is vital to ensure that the radiology service provider is meeting the industry’s best practices. Investment and participation in the quality assurance practice and optimization of workflow by administrators, radiologists, technologists, and physicists is required to ensure that the required image quality at the lowest diagnostically achievable radiation dose is achieved in normal CT imaging operation.

REFERENCES 1. AAPM, AAPM CT Terminology Lexicon, (n.d.). 2. AAPM: AAPM Definition of Qualified Medical Physicist. Am Assoc Phys Med n.d. 3. AAPM, AAPM CT Dose Education Slides, (2013). 4. ACR, ACR Accreditation Program Requirements, (n.d.). 5. Alhilali LM, Dohatcu AC, Fakhran S: Evaluation of a limited three-slice head CT protocol for monitoring patients with ventriculoperitoneal shunts. AJR Am J Roentgenol 201:400–405, 2013. 6. ASNR: ASNR Patient Info on Strokes. Am Society Neuroradiol n.d. 7. Bae KT: Intravenous contrast medium administration and scan   timing at CT: Considerations and approaches. Radiology 256:32–61, 2010.

44

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

8. Barrett JF, Keat N: Artifacts in CT: Recognition and avoidance. Radiographics 24:1679–1691, 2004. 9. Boone JM, Strauss KJ, Cody DD, et al: Size-Specific Dose Estimates (SSDE) in Pediatric and Adult Body CT Examinations (Task Group 204), College Park, MD, 2011. 10. Brisse HJ, Brenot J, Pierrat N, et al: The relevance of image quality indices for dose optimization in abdominal multi-detector row CT in children: Experimental assessment with pediatric phantoms. Phys Med Biol 54:1871, 2009. 11. Christianson O, Chen JJS, Yang Z, et al: An improved index of image quality for task-based performance of CT iterative reconstruction across three commercial implementations. Radiology 275:725–734, 2015. 12. Cody DD, Fisher TS, Gress DA, et al: AAPM Medical Physics Practice Guideline 1 . a: CT protocol management and review practice guideline. J Appl Clin Med Phys 14:3–12, 2013. 13. Friedman SN, Fung GSK, Siewerdsen JH, et al: A simple approach to measure computed tomography (CT) modulation transfer function (MTF) and noise-power spectrum (NPS) using the American College of Radiology (ACR) accreditation phantom. Med Phys 40:051907, 2013. 14. Habibzadeh MA, Ay MR, Asl ARK, et al: Impact of miscentering on patient dose and image noise in x-ray CT imaging: Phantom and clinical studies. Phys Med 28:191–199, 2012. 15. Harri PA, Moreno CC, Nelson RC, et al: Variability of MDCT dose due to technologist performance: Impact of posteroanterior versus anteroposterior localizer image and table height with use of automated tube current modulation. AJR Am J Roentgenol 203:377–386, 2014. 16. Hsieh J, Chao E, Thibault J, et al: A novel reconstruction algorithm to extend the CT scan field-of-view. Med Phys 31:2385, 2004. 17. Ip IK, Schneider LI, Hanson R, et al: Adoption and meaningful use of computerized physician order entry with an integrated clinical decision support system for radiology: Ten-year analysis in an urban teaching hospital. J Am Coll Radiol 9:129–136, 2012. 18. Khorasani R, Hentel K, Darer J, et al: Ten commandments for effective clinical decision support for imaging: Enabling evidence-based practice to improve quality and reduce waste. AJR Am J Roentgenol 203:945–951, 2014. 19. Kofler JM, Cody DD, Morin RL: CT protocol review and optimization.   J Am Coll Radiol 11:267–270, 2014. 20. Li K, Tang J, Chen G-H: Statistical model based iterative reconstruction (MBIR) in clinical CT systems: Experimental assessment of noise performance. Med Phys 41:041906, 2014. 21. Favazza CP, Duan X, Zhang Y, et al: A cross-platform survey of CT image quality and dose from routine abdomen protocols and a method to systematically standardize image quality. Phys Med Biol 60:8381–8397, 2015. 22. McCollough CH, Bruesewitz MR, McNitt-Gray MF, et al: The phantom portion of the American College of Radiology (ACR) computed tomography (CT) accreditation program: Practical tips, artifact examples, and pitfalls to avoid. Med Phys 31:2423–2442, 2004.

23. McEnrey KW, Radiology Information Systems and Electronic Medical Records, 2013. 24. McNitt-Gray MF: AAPM/RSNA physics tutorial for residents: Topics in CT. Radiation dose in CT. Radiographics 22:1541–1553, 2002. 25. Miller DL, Schauer D: The ALARA principle in medical imaging. AAPM Newsl 40:38–40, 2015. 26. Morin RL, Seibert JA, Boone JM: Radiation dose and safety: Informatics standards and tools. J Am Coll Radiol 11:1286–1297, 2014. 27. NEMA, NEMA XR 29: Standard Attributes on CT Equipment Related to Dose Optimization and Management Published by: National Electrical Manufacturers Association, 2013. 28. Radlex, Radlex Lexicon, (n.d.). 29. Richard S, Husarik DB, Yadava G, et al: Towards task-based assessment of CT performance: System and object MTF across different reconstruction algorithms. Med Phys 39:4115–4122, 2012. 30. Saver JL: Time is brain–quantified. Stroke 37:263–266, 2006. 31. Seidenfuss A, Mayr A, Schmid M, et al: Dose reduction of the   female breast in chest CT. AJR Am J Roentgenol 202:W447–W452, 2014. 32. Solomon JB, Christianson O, Samei E: Quantitative comparison of noise texture across CT scanners. Med Phys 39:6048–6055, 2012. 33. Strauss KJ, Goske MJ, Kaste SC, et al: Image gently: Ten steps you can take to optimize image quality and lower CT dose for pediatric patients. AJR Am J Roentgenol 194:868–873, 2010. 34. Supanich MP, Bevins N, Myers D: Reducing Radiation Dose in Abdominal CT Studies: ACR Dose Index Registry as an Impetus for Quality Improvement. In RSNA Qual. Story Boards, 2013. 35. Szczykutowicz TP, Bour RK, Pozniak M, et al, Compliance with AAPM Practice Guideline 1. a: CT Protocol Management and Review—from the perspective of a university hospital, 16 (2015) 443–457. 36. Szczykutowicz TP, Siegelman J: On the same page—Physicist and radiologist perspectives on protocol management and review. J Am Coll Radiol 1–7, 2015. 37. Toth T, Ge Z, Daly MP: The influence of patient centering on CT dose and image noise. Med Phys 34:3093, 2007. 38. Udayasankar UK, Braithwaite K, Arvaniti M, et al: Low-dose nonenhanced head CT protocol for follow-up evaluation of children with ventriculoperitoneal shunt: Reduction of radiation and effect on image quality. AJNR Am J Neuroradiol 29:802–806, 2008. 39. Wang CL, Cohan RH, Ellis JH, et al: Frequency, management, and outcome of extravasation of nonionic iodinated contrast medium in 69,657 intravenous injections. Radiology 243:80–87, 2007. 40. Yuan M-K, Tsai D-C, Chang S-C, et al: The risk of cataract associated with repeated head and neck CT studies: A nationwide population-based study. Am J Roentgenol 201:626–630, 2013. 41. Zamyatin AA, Nakanishi S: Extension of the reconstruction field of view and truncation correction using sinogram decomposition. Med Phys 34:1593, 2007.

3  Imaging Principles in Magnetic Resonance Imaging Nicole Seiberlich and David W. Jordan

Magnetic resonance imaging (MRI) is based on spatially encoding a nuclear magnetic resonance (NMR) signal. Therefore an understanding of the basics of NMR is critical for understanding MRI, and this information is given in the first section of this chapter. Magnetic field gradients are used in MRI to spatially encode the NMR signal to obtain the raw MRI data, or k-space. Images are reconstructed from k-space using mathematical algorithms, most commonly the Fourier transform. Image contrast is controlled by changing the timings of the encoding using specific pulse sequences. Thus in addition to a basic explanation of NMR signal generation, the process of spatial encoding and contrast generation are described. Additionally, commonly used rapid imaging techniques and frequently encountered artifacts are introduced.

PHYSICS PRINCIPLES Nuclear Spin All atoms are made up of electrons surrounding an atomic nucleus, which can contain protons and neutrons. The nucleus of an atom can have a quantum property known as nuclear spin.28 For instance, a hydrogen atom has a proton as its nucleus, and the magnetization from a macroscopic sample of proton spins is what is commonly measured in NMR and MRI. Many other atoms have nuclei with this property, but the discussion here will focus on hydrogen nuclei (i.e., protons) because hydrogen is a key component of biological tissue (e.g., water, fat) and most commonly visualized in clinical MRI. The behavior of each individual nucleus is described by quantum mechanics. However, when working with many nuclei in bulk, the picture can be simplified to make it more easily understandable. In this treatment the magnetization from a single proton can be thought of as an arrow pointing in a specific direction in three-dimensional (3D) x-y-z space. In the absence of an externally applied magnetic field, the nuclear magnetization is randomly oriented; in other words the magnetization from each proton in a group of protons (i.e., in a cup of water or the human body) is not aligned. If one were to try to measure the net magnetization or aggregate magnetization coming from the group of protons, it would be zero because each arrow would be pointing in a different direction, resulting in total cancellation. However, in the presence of an external magnetic field such as that used in NMR or MRI, the nuclei act like tiny bar magnets and align themselves with the field in one of two states—parallel or antiparallel.28 If an equal number of protons were in each state, the net magnetization would again be zero (the magnetic field from the nuclei pointing up would cancel the field from the nuclei pointing down). However, the state parallel to the external magnetic field has a slightly lower energy level than the antiparallel state, so slightly more nuclei end up in the parallel state than in the antiparallel state. This small difference in

46

parallel and antiparallel protons leads to the signal used for imaging in MRI. For clinical MRI field strengths, only one or two protons in a million can be detected; this number increases as the MRI field strength increases (which leads to an increase in the signal-to-noise ratio [SNR] at higher field strengths). The net magnetization, or magnetization from these excess protons, can be visualized as a larger arrow pointing in the direction of the external magnetic field, where the size of the arrow indicates the magnitude of the magnetization.

Nuclear Magnetic Resonance The net magnetization of protons in an external magnetic field aligns with the external magnetic field. This is called the equilibrium state of the magnetization and is usually defined as the z-direction (Fig. 3-1A [red arrow]). However, by applying energy in the form of electromagnetic waves at precisely the right frequency, the net magnetization can be made to tip away from the equilibrium position into the x-y plane.10,28,32,60,66 This frequency is known as the Larmor frequency: ω = γB0 where ω is the precession frequency, B0 is the strength of the external magnetic field, and γ is a constant that differs for each atomic nucleus, known as the gyromagnetic ratio). For instance, protons (with a gyromagnetic ratio of 42.577 MHz/T)38 in a magnetic field of 3.0 tesla (T) will be tipped when energy with a frequency of 127.731 MHz is applied. This energy is transmitted by so-called transmit coils and is usually referred to as a radiofrequency (RF) pulse.10,28,32,60,66 The angle by which the magnetization is tipped is called the tip angle. Once displaced from equilibrium, the net magnetization will precess around the external magnetic field. In other words, if the magnetization is tipped away from the z-axis to the x-axis using a 90-degree RF pulse (see Fig. 3-1A [orange arrow]), the end of the arrow describing the magnetization will move around the x-y plane, tracing a circle (yellow, green, and blue arrows). The frequency of this precession also depends directly on the external magnetic field according to the Larmor relationship shown above: the stronger the field, the faster the magnetization will precess. Signal can only be picked up from magnetization that is precessing in the x-y plane.10,28,32,60,66 An RF receiver coil is used to collect the signal after it is tipped into this plane using an RF pulse. In simple terms the receiver coil “sees” magnetization along one direction; in other words, voltage is induced in the receiver coil because of the change in direction of the magnetization vector that results from  the precession (similar to an electrical generator). To demonstrate  the relationship between the precession of the magnetization and the signal, the x-direction will be used as an example here. When the magnetization is pointed toward the x-axis, the signal will be at a maximum (see Fig. 3-1B [orange dot]). When the magnetization

47

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging precesses to be completely on the y-axis, there is no magnetization in the x-direction, so the signal picked up by the receiver coil is zero (see Fig. 3-1B [green dot]). When the magnetization precesses further into the −x direction, the signal will have a minimum (see Fig. 3-1B [blue dot]). The signal over time seen by precessing magnetization traces z

x

Signal

y t

FIG 3-1  A, Magnetization depicted on a 3D axis. Before excitation the

The net magnetization in an external static magnetic field aligns with the magnetic field, but an RF pulse can be used to tip this magnetization into the x-y plane as explained earlier. Once tipped and with no further perturbation, the magnetization returns to the equilibrium state along the z-axis. This return to the equilibrium state is known as T1 relaxation, otherwise known as spin-lattice relaxation or longitudinal relaxation.28 T1 relaxation follows an exponential recovery curve, as shown in Figure 3-2A, which is described by the following equation (assuming the signal started in the x-y plane after a 90-degree RF pulse):

where the signal at time t is written as s(t). Different tissues have different T1 times based on properties including the physical state of the components of the tissue (liquid, fibrous, solid) and the mobility of the protons in the tissue. Two example T1 relaxation curves for tissues with different T1 values are shown in Figure 3-2A. The tissue with the shorter T1 requires less time for the protons to return to the equilibrium state (i.e., aligned with the z-axis), whereas the tissue with a longer T1 time relaxes more slowly (see Fig. 3-2A). These differences in T1 times can be used to generate contrast in MRI images. By collecting data when the differences between the relaxation curves is largest, a maximum T1-weighting between two tissues can be achieved. To increase these differences in T1 and thus increase the contrast between tissues, an external contrast agent can be injected to shorten the T1 of one tissue. Gadolinium-based contrast agents are often injected into the blood pool to shorten the T1 time of blood and/or perfused tissues, which (given appropriate pulse sequence settings) can cause a large difference between signals from tissues in the proximity of the contrast agent and other tissues unaffected by the contrast agent.50

T1 = 200 ms

100% 80%

Transverse magnetization (Mxy)

Longitudinal magnetization (Mz)

magnetization is aligned along the z-axis (red arrow). After a 90-degree RF pulse the magnetization is tipped into the x-y plane. Here it is shown as being aligned with the x-axis (orange arrow). Once in the x-y plane the magnetization begins to precess in the x-y plane, first moving between the x and y axes (yellow arrow) and then aligning with the y-axis (green arrow) and after a longer time pointing in the negative direction along the x-axis (blue arrow). The magnetization will eventually precess around to the x-axis again and continue to move around in the x-y plane. Note that the effect of a 180-degree pulse would be to completely invert the magnetization (purple arrow). B, The signal collected from a receiver coil along the x-direction. Immediately after the RF pulse, the signal along the x-direction is maximal, leading to a large signal (orange dot). As the magnetization precesses, the signal decreases (yellow dot) until it reaches a minimum when the magnetization points along the y-direction (green dot). The magnetization continues to precess, leading to a negative signal when it is pointed in the negative x-direction (blue dot). Continued precession leads to the sinusoidal signal shape over time.

T1 = 1000 ms

60% 40% 20% 0% 0

A

Relaxation

s(t) = 1− e− t T1

B

A

out a sinusoid, which can be appreciated in Figure 3-1. This treatment ignores relaxation, which is discussed next.

1

2

3

Time (sec)

4

100% 80% 60% T2 = 200 ms

60% 20%

T2 = 50 ms 0% 0

5

B

0.2

0.4

0.6

Time (sec)

FIG 3-2  A, A plot showing the T1 relaxation of two tissues with different T1 values after a 90-degree RF pulse. Immediately after the pulse the longitudinal magnetization (Mz) is zero because all of the magnetization points along the x-y plane. Over time the magnetization slowly returns to equilibrium. The curve shown in black describes the relaxation of a tissue with a long T1 value (1000 ms); note that it takes a long time for the magnetization to return back to the equilibrium value. The red curve describes the relaxation of a tissue with a short T1 value (200 ms); this curve shows that the longitudinal magnetization is restored more quickly. B, A plot showing the T2 decay for two tissues with different T2 values. Immediately after a 90-degree RF pulse, all of the magnetization is found in the x-y plane. As the spins precess, they start to dephase, leading to a decrease in the net magnetization, or T2 decay. The red curve shows the signal time course for a tissue with a long T2 value (200 ms), and the black curve shows a tissue with a short T2 value. The signal coming from the tissue with the shorter T2 decays faster, and this effect is used to generate T2 contrast in MRIs.

0.8

1

48

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

In addition to T1 relaxation, a second relaxation mechanism is often used to impart contrast to MRI. This second relaxation parameter is T2 relaxation, otherwise known as spin-spin or transverse relaxation.28 The T2 parameter describes the behavior of the magnetization as it precesses in the x-y plane. Recall that the net magnetization is actually a sum of the magnetization of many protons. If all the protons precess at exactly the same frequency, the net magnetization would retain the same magnitude as it precesses. However, if some of these spins precess at slightly higher frequencies (i.e., a bit faster) and some at slightly lower frequencies (i.e., a bit slower), the magnetization will begin to “fan out” in the x-y plane. The net sum of this fanned out magnetization will have a smaller magnitude than when all the spins precess at exactly the same frequency. Over time the slight differences in precession frequency will lead to magnetization that is spread in all directions in the x-y plane, leading to a net magnetization of 0. This T2 relaxation process is described using the following equation: s(t) = 1− e

− t T2

The smaller the T2 value, the faster the magnetization fans out around the x-y plane, leading to a rapid decrease in the signal (see Fig. 3-2B). Like T1, different tissues have different T2 values depending on the makeup of the tissue. Small disturbances in the local magnetic field of the tissue can cause rapid T2 decay as well as differences in the chemical environments. In a broad sense, T2 provides information about how fluidlike a particular tissue is; solid tissues have shorter T2 values, whereas pure liquid water has a very long T2 value. It is important to note that the T1 and T2 times of tissues are not mathematically related, and MRIs are often weighted by one (or both) of these parameters to highlight desired structures. However, T2 is always shorter  than T1. Although the T2 value is an intrinsic tissue property, the transverse magnetization can also be affected by small differences in the magnetic field. Any inhomogeneities in the field will cause individual nuclei to have slightly different resonance frequencies and lead to a more rapid fanning out of the magnetization in the x-y plane. The parameter that describes the relaxation of magnetization in the transverse plane secondary to field inhomogeneities is T2*.28 The actual T2* depends on the local magnetic field, which can be affected by differences in magnetic susceptibility of adjacent materials (i.e., tissue-air interfaces) or the presence of iron. T2* is always shorter than T2, and although T2 relaxation cannot be avoided, the effects of T2* relaxation can be eliminated by using some specialized imaging techniques (e.g., spin echo sequences).

Chemical Shift So far it has been assumed that all protons have the same precession frequency, also called the resonance frequency. However, the Larmor equation actually only describes the resonance frequency of an isolated hydrogen nucleus. The proton is almost never found completely isolated; hydrogen is normally bound to another nucleus (in water, this is oxygen) through chemical bonds. These chemical bonds are made up of electrons moving around the nuclei of the bound atoms, and this cloud of electrons serves to “shield” the protons of the hydrogen atom from the externally applied main magnetic field. Thus the actual magnetic field seen by the protons is usually slightly less than the actual applied field, as described in the following equation: ωc = γ ⋅ (1− σ)B where ωc describes the true resonance frequency of the proton and σ describes the degree of shielding from the electrons.28 The value of σ is determined by the types of chemical bonds; for instance, the degree of shielding, and thus the resonance frequency, is different for water

RF

t

ACQ

t

FIG 3-3  A basic NMR pulse sequence. Both lines describe the timing of events in the pulse sequence. The top line shows the RF pulse. The bottom line shows when signal is collected (ACQ). Note that the signal shape is an exponentially decaying sinusoid owing to relaxation.

and fat. This effect, called chemical shift, is exploited for NMR spectroscopy, a technique used to characterize different chemical compounds. Chemical shift is also observed in MRI, and this effect will be described later in the chapter.

The Basic NMR Experiment The basic concepts described can be put together to understand how a simple hydrogen NMR experiment works. The timing of the events in the experiment is described using a pulse sequence diagram,4,7 where each line represents a different type of event. The basic NMR pulse sequence diagram in Figure 3-3 shows when RF pulses are applied  (top) and when data are collected (bottom). A sample made up of substances containing protons is placed in an NMR scanner with a strong external magnetic field, B0. This field causes the protons to align, leading to a net magnetization. An RF pulse is applied to tip the magnetization from the equilibrium z-axis into the x-y plane. Once in the x-y plane the magnetization begins to precess at a frequency determined by the chemical environment of the hydrogen nuclei. Signal from this precessing magnetization is collected using a receiver coil. During signal collection the magnetization experiences both T1 and T2 relaxation simultaneously. The precession causes the signal to oscillate, and the relaxation causes the signal to decay to zero, leading to an overall signal form that looks like an exponentially decaying sinusoid (this shape is shown on the data acquisition line labeled “ACQ” in Fig. 3-3). This process can be repeated several times and the signal averaged to increase the SNR (see “Imaging Time and Performance Tradeoffs”). Such an experiment can be used to determine how many and what type of chemical compounds are present in a sample by examining the resonance frequencies in the collected signal. However, because the receiver coil picks up signal from all parts of the sample, there is no information about the spatial distribution of the chemical compounds in the sample.

IMAGING Introduction NMR is traditionally used to determine the chemical structure of a sample and does not require spatial localization. However, MRI depends on being able to distinguish between spins at different physical locations, so spatial localization is performed by applying magnetic field gradients in each of the three spatial axes in addition to the main static magnetic field.44 A schematic of this idea is shown for onedimensional (1D) encoding in Figure 3-4. Although this picture does not fully explain how two-dimensional (2D) or 3D spatial localization is performed, this idea forms the basis for spatial encoding in MRI, which will be explained in detail in the following sections.

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging B=B0–xGx

B=B0

49

B=B0+xGx Gx Time FT

X Signal (t) or

A

Summed signal from all vials (t)

B

1D “Image”

Freq

x

FIG 3-4  A, Three vials are filled with a liquid (red) and placed in an MRI scanner with a main magnetic field strength of B0. A magnetic field gradient, Gx, which varies along the x-direction, is also applied; the net magnetic field at any position in space depends on the combined strength of both of these magnetic fields. After an RF pulse the magnetization in the left vial precesses slowly, the center vial moderately, and the right vial rapidly because of the different magnetic fields they experience. If signals in time were collected from each vial independently, these differences in precession rate would be seen, and differences in the amplitude of the signals would correspond to the amount of liquid at each of the spatial positions (middle). The summed signal (bottom) is what is actually collected by the MRI scanner. This type of encoding, where a gradient is played out during signal acquisition, is called frequency encoding. B, The summed MRI signal can be decomposed into separate frequencies using the Fourier transform. The time signal is converted into frequency units, which reflects the position of the vials. This can be represented as a 1D “image” where frequencies with high signal amplitude are bright and frequencies with low signal amplitude are dark.

In Figure 3-4A, three signal sources (liquid in tubes) are shown in red. Here a linearly varying magnetic field is applied in the x-direction; this field gradient is called Gx. At the center the amplitude of this additional magnetic field is 0, and the spins at that location precess at the Larmor frequency determined by the main magnetic field. However, the spins on the left side experience a total magnetic field that is slightly less than the main magnetic field owing to the gradient. These spins will precess with a slightly lower frequency. Similarly the spins on the right side experience a slightly higher magnetic field and precess at a slightly higher frequency. Thus the result of applying this magnetic field gradient is to differentiate spins at different locations by causing them to precess at different frequencies. In this way a one-to-one mapping can be made that links the location of the spins to their frequency. Additionally, note that the left-most vial contains less fluid and the signal coming from this vial has a lower amplitude than the center vial, which also has a lower amplitude than the right-most vial. In fact the signal amplitude at each frequency is directly proportional to the number of spins precessing at that frequency. The signals coming from the different locations, however, cannot be separated by the receiver coil. Only the sum of the three signals is actually collected. This signal is shown at the bottom of Figure 3-4A, and the next section describes how image information is derived from this composite signal.

Data to Image: Fourier Transform The schematic shown in Figure 3-4A and the legend explain how spatial encoding is performed but do not explain how spatial information is extracted from the signal to form an image in MRI. As described

earlier the MRI signal is collected from the entire volume using a receiver coil. Thus the challenge is to separate individual single frequency signals from the complex multifrequency data. Mathematics has a simple way to perform this separation: the Fourier transform,9 which decomposes a signal in time into its component frequencies— exactly the problem that must be solved in MRI.21 By finding the Fourier transform of the MRI signal, a “map” of the frequencies contained by the signal can be derived. Because the frequencies correspond directly to the spatial locations of the spins, this frequency map is also equivalent to the 1D “image” of the spins. Note that in the 1D image shown in Figure 3-4B, the vial on the left with little liquid has a low signal intensity, the center vial a medium intensity, and the vial on the right a high intensity. All frequencies that were not present in the signal correspond to zeros in the frequency map and black in the 1D image.

Frequency Encoding Nearly all MRI scans perform spatial localization in one dimension using frequency encoding, which is the closest form of encoding to the basic schematic shown in Figure 3-4. In frequency (or read) encoding, a magnetic field gradient is applied to the object in one direction while data are collected (and hence the alternative name, read encoding). The spins precess at a frequency dictated by their local magnetic field, and the resulting data are a complex combination of all of the different frequencies, as described earlier. If a 1D Fourier transform is applied to these data, a 1D projection through the object would result, as shown at the bottom of Figure 3-4B. Only one dimension can be resolved using frequency encoding, because the second dimension is not encoded and the spins in that direction are not distinguishable from one another.

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging Gy strength

50

Time At end o

f gradie

A

y

Gy strength

Gy strength

y

Gy strength

B

nt

y

At end of gradient

ky = 0 y

Time At end of gradient

ky = 1 y

Time At end of gradient

ky = 2 y

Time

FIG 3-5  A, A phase-encoding gradient is applied along the y-direction, which causes spins in different locations in the y-direction to precess with different frequencies. When the gradient is turned off, the spins have a phase that is dependent on their position. This phase is retained when frequency encoding is applied in a direction orthogonal to the phase encoding. B, To collect different phase-encoding lines, phase-encoding gradients with different strengths are used. For ky = 0, no phase-encoding gradient is applied, and no y-position–dependent phase is imparted. For ky = 1, a gradient is applied with a strength and duration that leads to one period of a sinusoid across the field of view (FOV) in the y-direction. For ky = 2, the gradient area is twice that for ky = 1, and the phase variation of the magnetization over the FOV covers two periods.

For this reason, another encoding mechanism must be employed in addition to frequency encoding, namely phase encoding.

Phase Encoding Phase encoding works through the same basic mechanism as frequency encoding. The only difference is that the magnetic field gradient used in phase encoding is not applied while the data are being collected. Instead a phase-encoding gradient is turned on in a direction orthogonal to the readout direction before data collection. While the phaseencoding gradient is turned on, the spins precess with a frequency that depends on their spatial location (Fig. 3-5A). Once the gradient is turned off the spins will stop precessing, but they will retain the phase they had due to the phase-encoding gradient (hence the name). A readout gradient is then applied in the second encoding direction, as described earlier, and data are collected. The application of one phase-encoding gradient would not help to make a 2D image, because just imparting a phase is not enough to spatially encode information. Thus many phase-encoding gradients with different strengths are applied, followed by a readout after each. The effect of applying phase-encoding gradients with different strengths is shown in Figure 3-5B. Here, schematic gradients are shown for ky = 0, ky = 1, and ky = 2. Note that the area of the gradient (time multiplied by amplitude) determines how much phase is accumulated across the field of view (FOV). If each readout line for different phaseencoding gradient magnitudes are collected and put into a 2D grid, the resulting data matrix is known as k-space.

k-Space k-Space is the commonly used name for the raw data collected in MRI.48,58,75 These data are a set of sampled points stored in a matrix of numbers, where each line of numbers corresponds to one value of the phase-encoding gradient. The units of k-space are time (s) or inverse meters (m−1); both are different ways of looking at the same data. The k-space data are encoded as previously described, with phase and frequency encoding using magnetic field gradients. The gradients control the path that is traversed through k-space. An example of how gradients are used to “drive” through k-space is shown in Figure 3-6. The phase-encoding gradient is used to move up and down in k-space, and the readout gradient is used to move to the left and right. Once the k-space data have been collected, a discrete Fourier transform is applied to mathematically transform the data to the image domain.75 The Fourier transform takes the complex multifrequency data and reports back which frequencies are present in the data. The result is a matrix of numbers that correspond to the image. The units of the image are the inverse of the k-space data, namely inverse seconds (s−1, or Hz) or meters (m), which matches with the idea that the image is a map of the frequencies generated using phase and frequency encoding as previously described.

Resolution and Field of View The FOV is defined as the total size of the image (in units of distance). The matrix size of the image is often a power of 2 (128, 256, 512) owing

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging

51

Trajectory through k-space

MRI pulse sequence RF Pulse

Gz

Gy

Gx ky

A

ACQ

B

kx

FIG 3-6  A, Schematic of an MRI pulse sequence, where the RF pulse and slice select gradients are shown in blue, the phase-encoding gradients in yellow and orange, the read dephasing gradient in green, the frequency encoding gradient in pink, and the acquisition window in gray. B, Each gradient is used to “move” around in k-space. After slice selection, no spatial encoding has yet been performed, and the location in k-space is denoted with the blue dot (center). The phase-encoding gradient enables motion up and down in k-space (orange and yellow arrows), the read dephasing gradient causes motion to the left (green arrow), and the frequency encoding gradient to the right (red arrow). If data are collected during the frequency encoding gradient, the result is a line of k-space data.

to historical restrictions on performing a fast Fourier transform. Given the FOV and matrix size for an image, resolution is defined as the FOV divided by the matrix size, resulting in the nominal pixel or voxel size. For instance, if a 2D image has an FOV of 200 mm × 200 mm with a matrix size of 128 × 96 in the read and phase-encoding directions, the spatial resolution (pixel size) would be 1.56 mm × 2.08 mm in these directions. The FOV and resolution of the image are directly related to the spacing and coverage of the data collected in k-space. The spacing between adjacent points in k-space determines the FOV in the image. The closer the phase-encoding lines or readout points are in k-space, the larger the FOV will be. In general the k-space spacing is chosen to be just large enough to cover the entire object. If the spacing between k-space lines becomes too large, the FOV may no longer cover the object, which can lead to aliasing artifacts (described in the “Parallel Imaging” and “Clinical Artifacts” sections). The resolution of the image is determined by the extent of   the k-space covered. The extent can be described by looking at the product of the spacing between k-space points (which determines  the FOV) and the number of points collected. The larger this product is, the higher the image resolution will be. This relationship can  be understood by thinking of the image as a frequency map. When more time-domain data (k-space) are collected, more frequencies can be resolved from that data, and the resolution of the image will be higher. The significance of the different parts of k-space can be examined by removing some portions of data and performing a Fourier transform to see what happens in the resulting image. It is often said that the center of k-space determines the contrast of the image and the outer portions of k-space provide resolution. Figure 3-7 explores this statement by manipulating the k-space. At the top, an example k-space and its corresponding image is shown. If the extent of the k-space is reduced by eliminating some of the outer k-space lines, the resulting image has a lower spatial resolution (it appears blurrier [middle]). If

only the outer portions of k-space are retained and the central portions set to zero, only the edges appear in the image (bottom). The effects of eliminating other areas of k-space are examined later in “Fast Imaging for Clinical Practice.”

2D vs. 3D: Slice and Volumetric Encoding Most clinical MRI images are collected in 2D, where the two spatial dimensions are encoded using read and phase encoding.28 However, before these dimensions can be encoded the desired 2D slice must be selected from the 3D volume of the body. This process is performed using slice-selection encoding. Simultaneously with the RF excitation pulse, a slice-selection magnetic field gradient is applied to the subject in the slice direction (perpendicular to read and phase-encoding direction). The RF pulse will only excite magnetization resonating at specific frequencies. Because the resonance frequency can be controlled by application of the sliceselect gradient, by setting the pulse sequence parameters correctly, only magnetization inside the desired slice will be excited by the RF pulse, and the magnetization outside the slice will remain mostly untouched. This process is shown schematically in Figure 3-8. The excited magnetization will then be subjected to read and phase encoding to resolve the spatial information in the 2D slice. As in read and phase encoding, a higher definition in the slice direction (or thinner slice) requires a stronger gradient in the slice direction. If the slice gradient is very weak, the slice will be very thick (light gray slice in Fig. 3-8). Such a thick slice can also be selected and then further encoded using a second, independent phase-encoding gradient in the slice direction. Such encoding is termed 3D encoding, and the second phase-encoding operation is often called partition encoding (to distinguish a 2D slice from a single 3D partition); 3D encoding results in a true 3D k-space, requiring a 3D Fourier transform to reconstruct the 3D image. Although 3D encoding is more time consuming, the image SNR is significantly higher because of the larger area covered. However, because the scan time is longer, there is a greater chance for image

52

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

Full k-space

FT

Center only

Image

FT

Outside only

k-space

FT

FIG 3-7  Top, An example of fully sampled k-space (left) and the corresponding image (right). Middle, If only the center portion of the k-space is collected, the resulting image has a lower resolution but shows the overall image contrast. Bottom, The outer portions of k-space correspond to the high-resolution details of the image.

artifacts due to motion, which is why 3D imaging is used clinically less frequently than 2D imaging.

Pulse Sequences for MRI Pulse sequences are methods for manipulating the magnetization of tissues and the signals they produce to generate a specific image contrast.4 An MRI system programmer uses pulse sequences to exploit the NMR principles that produce signal from tissue (and thus contrast among different tissues) and to perform spatial encoding. The pulse sequence is a series of events triggered over time by the MRI scanner to change the magnetization of tissue in some way. There are two basic ways of manipulating the magnetization of tissue: applying RF energy and applying magnetic field gradients (Gx, Gy, and Gz). The pulse sequence diagram is a schematic that describes the order and duration of RF and gradient fields being switched on and off over time to produce the type of image contrast weighting and spatial encoding the MRI user wants. Each family of pulse sequences has strengths and weaknesses for different clinical imaging applications. As described earlier the RF is switched on for a very short time with variable power and then switched off, and this short transmission is

called an RF pulse. When it is used to excite the tissue signal for imaging, it is often called an excitation or flip pulse. For a specific MR scan the pulse will be associated with a flip angle that corresponds with the resulting magnetization of the excited tissue (0 degrees is magnetization parallel to the B0 field, shown by the red arrow in Figure 3-1, 90 degrees is perpendicular to B0, shown by the orange arrow, and 180 degrees is antiparallel to B0, shown by the purple arrow). Additional RF pulses may be applied at different times, depending on the type of pulse sequence.

Pulse Sequence Diagrams.  The pulse sequence diagram is an illustration of the time relationship of the different events that make up the pulse sequence. It is not always drawn to scale. There are  some common symbols used in pulse sequence diagrams, but there  is no standard format. Events in the pulse sequence are represented  by symbols located on the respective line at the relative time of the event. There are five essential system hardware functions whose events are usually shown on a pulse sequence diagram: RF, Gslice, Gphase, Gread, and ACQ (Fig. 3-9). Note that the pulse sequence shown in this figure is a

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging

53

Gss

ω2=γ(B0+X2GSS) ω1=γ(B0+X1GSS)

B

X1

X2

X

FIG 3-8  A schematic of slice selection using using a magnetic field gradient. A gradient (Gss shown as the orange-to-green wedges) is applied along the direction of the slice to be selected (in this diagram, the x-direction). The total strength of the magnetic field is shown on the vertical axis. If an RF flip pulse is used that excites frequencies between ω1 and ω2 but not the frequencies outside of this range, only spins along the x-direction that are precessing at these frequencies will be selected. Thus in this diagram all spins between X1 and X2 would then be excited in the ideal case (the selected slice is shown in light gray). Note that if the strength of the slice select gradient is increased (where the new slice select gradient is shown as the light blue wedges), the same RF flip pulse exciting frequencies between ω1 and ω2 will result in a slice that is twice as thin as the original slice (shown in dark gray) because only these spins will be precessing at the appropriate frequencies with the increased magnetic field gradient.

RF

Gslice

Gphase

Gread

ACQ

TE

FIG 3-9  Gradient echo pulse sequence diagram showing symbols used for RF and gradient pulses.

gradient echo sequence, but the same basic elements appear in all pulse sequences. RF pulses are shown on an RF line to indicate any excitation and refocusing pulses. The three orthogonal magnetic field gradients, namely phase, slice, phase, and frequency (or read) encoding, are shown as separate lines. The activation of each of the gradients is indicated by a step change on the appropriate gradient’s line. The activation of the receiver and the data acquisition period (ACQ) is normally shown in context with the other elements of the pulse sequence. Sometimes the acquisition is shown in the same line as the RF pulses because this also represents an RF event. The phase-encoding gradient pulse (Gphase) is often drawn as a stacked series of pulses (hexagon with horizontal lines) to show that the gradient pulse has different amplitudes for each repetition of the

pulse sequence to accomplish phase encoding. For the complete imaging sequence, phase encoding is repeated with a variety of pulse amplitudes with both negative and positive polarities as shown. These different polarities allow lines both above and below the center of k-space to be collected. A pulse sequence diagram can be drawn to represent either functional (slice select, read, or phase gradients) or physical gradients (in the x, y, or z direction). In 2D imaging there are three spatial encoding functions that are performed by gradients: slice selection (Gslice), phase encoding (Gphase), and frequency encoding or readout (Gread). An MRI scanner has three orthogonal sets of electromagnetic gradient coils (Gx, Gy, and Gz) that produce the gradient fields. For imaging in any of the three orthogonal planes of the scanner, each of the physical gradient coils performs one of the three spatial encoding roles. However, in the case of oblique imaging, combinations of the physical gradients are used to create gradients within or through the oblique planes defined by the scanner operator. The time period over which the data collection is repeated is called the repetition time (TR) in an MRI scan. The initial RF flip pulse begins each cycle, so the TR value is typically the time from one flip pulse until the next flip pulse. The MRI operator chooses the TR value for each scan. The value of TR selected helps determine the contrast of the resulting image (see “Contrast in MRI”). The echo is the process used to recover magnetization signal in a sample after it has developed signal differences (contrast) between different tissues over time. Recall that the net magnetization lost to T2 relaxation cannot be recovered, but other components of T2* dephasing, or “fanning out,” can be recovered. The echo is the technique for recovering, or rephasing, the lost magnetization. In an MRI scan the echo time (TE) is the time period that elapses between an RF flip pulse and an echo. The MRI scanner operator can choose the TE for each scan, and data acquisition occurs at TE. Like the TR value, the TE value selected plays a role in determining the image contrast.

54

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging TE/2

RF

90

180

Gslice

Gphase

Gread

ACQ TE

FIG 3-10  Spin echo pulse sequence diagram.

Spin Echo Pulse Sequences.  Spin echo pulses sequences19,41 were some of the first to be introduced for imaging. A simple pulse sequence diagram for the spin echo is shown in Figure 3-10. The sequence begins with a 90-degree flip pulse with accompanying slice-select gradient followed by the phase-encoding gradient. Dephasing describes the fanning out of spins due to T2 or T2* effects (spin-spin interaction, chemical shift, and B0 field inhomogeneities). At a time equal to TE/2 (halfway between the RF pulse and the TE time), a 180-degree refocusing RF pulse is played. The slice-select gradient is applied at the same time as this pulse so that only the spins in the selected slice are effected. The effect of this RF pulse is to reverse the orientation of all the spins. During the time between the refocusing pulse and the echo, all the dephasing that occurred because of chemical shift and B0 field inhomogeneity will rephase. At the time TE, the spins will momentarily be in phase again and produce a strong signal. The signal from all the spins will be less intense than the original excitation signal owing to the T2 relaxation, which cannot be rephased. Just before the time TE, the frequency-encoding gradient is activated together with the signal receiver (often labeled ADC for analogto-digital converter). The frequency-encoding gradient is on while the RF signal from the tissue is collected by the receiver. The result of such a pulse sequence is an image that is weighted by T2 relaxation; this relaxation cannot be reversed. One drawback of spin echo imaging is that there must be a long gap, or TR, between the collection of different lines of k-space to let the magnetization fully relax according to the T1 of the tissues. The effect of a long TR (on the order of several seconds) is that the total scan time for a spin echo sequence is also long (on the order of several minutes), making spin echo images susceptible to artifacts due to motion.

Gradient Echo Pulse Sequences.  The gradient echo52 is one of the most commonly used pulse sequences in MRI. The pulse sequence diagram for the gradient echo is shown in Figure 3-9. The gradient echo does not use a 180-degree RF pulse to form the echo. Instead an opposite-polarity readout gradient pulse is applied. This additional gradient is shown on the Gread line in the pulse sequence diagram in Figure 3-9—it is the first trapezoid. This gradient creates a dephasing of protons in the frequency-encoding direction that is reversed (refocused) when the readout gradient is applied at TE. The rephased spins form an echo at TE that is detected by the receiver.

The gradient reversal technique does not refocus the dephasing caused by chemical shift and B0 inhomogeneity the way the 180-degree RF pulse does in the spin echo pulse sequence. Gradient echo images are usually weighted by the T1 values of the tissues because of the short TR values used. Additionally, gradient echo images can have T2* weighting rather than T2. When T2* weighting is acceptable or preferred to T2 weighting, gradient echo sequences can be used to perform very fast imaging. T2* gradient echo images can be obtained using flip angles of 20 to 30 degrees. This allows a shorter TR to be used because the T1 relaxation of the tissues to equilibrium takes much less time than when a full 90-degree flip angle is used. TE values must be very short to collect usable signals, because the T2* relaxation is very fast. The drawback of gradient echo imaging is that there is generally less signal available than in spin echo imaging, since dephasing due to B0 inhomogeneity and local magnetic susceptibility is not refocused.

Contrast in MRI The ability to detect contrast between tissues using differences in T1 and T2 relaxation has been important since the beginning of MRI in medicine. Modern clinical MRI also includes many other techniques to detect or manipulate contrast in images. MRI can be used to highlight or hide tissues and fluids based on their composition, position, or motion, depending on the application. To achieve clinically useful contrast between different tissues, the magnetization must be allowed to relax for some period of time to develop different contrast at the time the signal is collected. The problem with this approach is that the strength of the signals from the tissues decreases very quickly during the waiting period. The rate of signal decay is based on the T2* relaxation time, which is very fast for most tissues. Capturing the signal immediately after the flip pulse would produce a lot of signal but very little contrast between tissues, as shown for T1 and T2 relaxation in Figure 3-2. On the other hand, waiting a long time after the excitation pulse before collecting data would result in a lot of contrast but very weak signal (also shown in Fig. 3-2). The echo technique provides a way to resolve this difficulty. Thus the majority of imaging techniques are based on either the gradient echo or the spin echo sequences, with additional encoding mechanisms for other types of weighting.

T1 and T2 Weighting In clinical MRI it is usually of interest to produce images that show differences in tissue signal based mostly on a single physical property (e.g., T1, T2, diffusion, perfusion) of the tissues. The two most common approaches to contrast in clinical MRI are to select T1 weighting or T2 weighting. To achieve T1 weighting the MRI operator sets the scanner for short TR and short TE values. This means that TR should be short compared to the T1 values of the tissues that are expected to be in the volume of interest, and TE should be short compared to the T2 values of these tissues. For example, clinical T1-weighted spin echo imaging with 90-degree flip angle typically uses TR values of 400 to 600 milliseconds and TE values of 10 to 20 milliseconds. When the TR is selected to be short, tissues with shorter T1 relaxation times (faster relaxation) have higher signal intensities. A short time after the RF flip pulse, the tissues with short T1 will relax to a greater degree, which means there will be more longitudinal magnetization present to generate signal upon application of the next RF pulse (see Fig. 3-2A). Tissues that have longer T1 values and are only partially relaxed will not produce as much signal if another RF pulse is applied. It is helpful to think of the longitudinal magnetization of the tissue at the end of the TR as the amount of transverse magnetization that will be available once the next RF pulse is applied. Tissues with shorter T1 times have more available signal a short time after the RF pulse. Using

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging a short TE in T1 imaging ensures that the signal is collected soon after the RF pulse before any differences in signal due to T2 relaxation can develop. In T2-weighted imaging, a long TR (relative to the T1 of the  tissues) and a long TE are used (relative to the T2 of the tissues). The long TR is chosen so that all tissues will fully return to their equilibrium state along the longitudinal axis after each RF pulse before another RF pulse is applied. This ensures that there are relatively small signal differences based on differences of the T1 values in the tissues (and thus little T1 weighting). The long TE is used to allow tissues to develop different signal intensities shortly after the RF pulse due to T2 relaxation. Then the signal is collected at a time when tissues with different T2 times will have different signal intensities. In T2-weighted imaging the tissues with the longest T2 times will produce the highest signal intensity; tissues with shorter T2 times will dephase faster, so there will be less signal available to collect at time TE (see Fig. 3-2B). Clinical T2-weighted spin echo or fast-spin echo imaging typically uses TR values of 3000 to 4000 milliseconds and TE values of 80 to 120 milliseconds. Because gradient echo pulse sequences cannot refocus the dephasing caused by local inhomogeneity in B0 and local magnetic susceptibility, gradient echo scans with long TR and long TE are weighted by T2*—the composite result of these effects plus the T2 decay in the tissue—rather than by T2 of the tissue alone. The chosen TE for such scans is long with respect to the T2* time of the tissues of interest to produce contrast. Also, it is common to use flip angles less than 90 degrees for such applications (25 degrees is a typical value). A TR value that is long compared to the time for the tissue to undergo full T1 relaxation from the flip angle chosen must be used, but this requires less time than relaxation from a full 90-degree flip. A clinical T2*weighted gradient echo sequence might have a TE value of 10 milliseconds or less, a TR value of 300 to 500 milliseconds, and a flip angle of 20 to 30 degrees.

Inversion Recovery Inversion recovery (IR)11 is a technique that can be used to selectively remove signal from a tissue or fluid from the image. It uses the differences between characteristic T1 relaxation times in different tissues to do this. In an IR pulse sequence the cycle begins with a 180-degree RF inversion pulse, which flips all of the protons into an inverted longitudinal magnetization (purple arrow in Fig. 3-1A). The protons then begin to relax (according to their T1 values) back to the equilibrium longitudinal state, as shown in Figure 3-2A. At different times the protons in different tissues and fluids will pass through the transverse plane and exhibit no longitudinal magnetization. If an excitation RF pulse is applied at the time the protons in a specific tissue reach the transverse plane, those protons will not be sensitive to the RF pulse. All of the protons in the other tissues are excited into the transverse plane with synchronized phase. The excited tissues will produce either T1- or T2-weighted signals (depending on the TR and TE used), but the tissue that was saturated will produce no signal and thus will not be visible in the resulting image. The IR technique is simple and effective for reducing the signal from species that are normally very bright but may tend to mask other details of interest. It is most commonly used for fat in T1-weighted imaging and water/fluid in T2-weighted imaging. IR used for fat suppression is called short tau inversion recovery (STIR) for the short T1 time constant of fat.8 Fluid-attenuated inversion recovery (FLAIR)13,30 refers to sequences that are used to suppress water or fluid signals. The time from the 180-degree inversion pulse to the excitation pulse is called the inversion time and is often represented by the symbol TI on the MRI user interface. To suppress fully relaxed species, an IR sequence

55

should use TI = ln(2)* T1 for the T1 time of the species to be suppressed, where ln(2) ≈ 0.69. The IR technique reduces the overall SNR of the image. When the RF pulse is applied, the tissue or fluid chosen for suppression is fully saturated, but all of the other tissues of interest are partially saturated— they have some transverse magnetization but not the full amount available immediately after an RF pulse without inversion. As a result they are not able to absorb energy as fully from the excitation pulse, so they produce less signal than if they had been excited from a fully longitudinal magnetization state.

Spectral Techniques for Signal Selection or Suppression Because IR reduces signal from the tissues of interest, it is not ideal for fast or high-resolution imaging. Spectral fat suppression, often referred to as “fat sat,” is a family of approaches to overcome this problem. These methods are based on the fact that protons in fat molecules have a different resonant frequency than protons in water molecules (see “Chemical Shift”). The difference is 3.5 parts per million (ppm), so the frequency difference depends on the strength of the B0 magnetic field. In a 1.5-T magnetic field the resonant frequency is 63.86 MHz for water protons, and the 3.5 ppm shift is 223 Hz for fat protons. An RF pulse with a very narrow frequency bandwidth can selectively excite water or fat protons without exciting the other species. A water-selective pulse can be applied to excite only the water protons.33 Alternately a fat-selective pulse can be applied to excite the fat protons and saturate them; then an additional gradient known as a spoiler gradient can be applied to rapidly dephase the transverse magnetization in the saturated fat. Subsequently the water protons can be excited for imaging and the dephased fat protons do not contribute to the image signal. Clinical fat suppression techniques use a combination of spectral fat saturation and IR to achieve the best suppression of fat signal; spectral presaturation with inversion recovery (SPIR)42 and spectral attenuated inversion recovery (SPAIR)5 are two examples in common usage.

Diffusion Diffusion is the net position change over time of molecules in continuous random (thermal) motion. A good illustration of diffusion is adding a drop of dye to a bottle of water. The dye droplet is very concentrated at first, but eventually it disperses into the water. After some time the dye is uniformly distributed throughout the water; there is no location where the dye is dense or dark, and the water takes on a very faint uniform color from the dye. This is an example of isotropic diffusion—the dye disperses evenly in all directions at the same speed. The time it takes for the dye droplet to disperse can be described by a diffusion coefficient (mm3/sec) that is specific to the dye in water and the temperature. If the same dye were added to alcohol or oil, the time scale for the diffusion would be much faster or slower than the same process in water, and the diffusion coefficient would be much larger or smaller, respectively. Next consider a straw with small pinholes placed inside the water bottle. The dye droplet is placed into the end of the straw. In this example the dye will diffuse readily along the length of the straw but much more slowly into the rest of the bottle. This is an example of anisotropic diffusion, where the movement of the dye is constrained in at least one direction. For this arrangement there are different diffusion coefficients to describe diffusion in different directions. The dye has a large diffusion coefficient along the axis of the straw and a smaller diffusion coefficient for movement through the walls of the straw.

56

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging TE/2

RF

Free water

Off-resonance RF excitation 90

180

Gslice Bound water

Gphase f0

FIG 3-12  Magnetization transfer (MT). The RF pulse is resonant with

Gread

ACQ TE

the bound water and saturates the bound water pool. There is MT to the free water nearby, so the MT affects only free water pools close to bound water (in cells, interstitial spaces). MT does not affect free water protons in, for example, vessels or ventricles, where no bound water is nearby.

FIG 3-11  Diffusion weighting. A diffusion gradient pulse is added to the phase-encoding gradient before and after the 180-degree pulse; the effects of these pulses cancel each other for stationary protons but cause additional dephasing in diffusing protons undergoing random motion.

Diffusion of proton-containing molecules in tissue or fluid is a feature of clinical interest that can be detected by MRI.45,46 Diffusionweighted imaging (DWI) uses additional gradient pulses during the dephasing and rephasing periods of a sequence to induce extra dephasing of protons that are part of diffusing molecules. Because the diffusion motion is random, the diffusion gradient pulses will cause dephasing of these protons that cannot be recovered. The faster the diffusion motion, the more dephasing the spins undergo as they move in the diffusion gradient field. Meanwhile the diffusion gradients do not produce any dephasing of the stationary spins, because the dephasing from the gradient pulse is rephased by the second gradient pulse. Thus a diffusion-weighted image will have bright signal in areas of low diffusion and less signal from areas of high diffusion. Figure 3-11 shows an illustration of a pulse sequence used for DWI. In this example the extra gradient pulses are applied to the phaseencoding axis so that the resulting image will be weighted by how much diffusion occurs along the phase-encoding direction. In practice, this process is repeated several times using different gradient strengths called b values. The larger b values reflect stronger diffusion gradients; strong diffusion gradients produce more diffusion weighting (greater contrast between tissues based on their diffusivities) but less signal (and lower SNR) overall. Imaging is repeated using diffusion-weighting gradients along all three physical dimensions to measure diffusivity in different directions. The data from the individual scans are used to calculate images that show average diffusion weighting (non–directionspecific) and maps of the apparent diffusion coefficient (ADC). Diffusion tensor imaging (DTI) is a technique closely related  to DWI that is useful in situations where the diffusion is highly anisotropic, such as movement of water molecules within fibers such  as white matter of the brain. If these structures are imaged using diffusion-weighting gradients in several directions, the individual  data sets can be used to construct a 3D map of the pathways along which the diffusivity is highest. These maps are the basis of MR fiber tractography.47,54,55

Magnetization Transfer Magnetization transfer27,34,81 is a technique that allows the MRI user to saturate tissues containing water. Water molecules can be part of free

fluid (e.g., blood, cerebrospinal fluid [CSF]), or they can be bound (by hydrogen bonds) to macromolecules such as proteins. Protons in free water have long T2 values, and protons in bound water have very short T2 values. The center of the spectrum for bound water is at the same frequency as the spectrum for free water, but free water has a tall narrow peak and bound water has a shorter broad peak (Fig. 3-12). The protons in the bound water are able to readily interact with each other to transfer magnetization. If an RF pulse is applied at a frequency outside the free water peak but within the bound water range, the RF pulse can excite the bound water protons within that resonant frequency range, and those protons can in turn excite the rest of the bound water protons. In this way, tissue containing water adsorbed to macromolecules becomes saturated while fat and other species do not. Magnetization transfer occurs between the bound water and nearby free water, such as near membranes, where the signal from the free water will become saturated and thus strongly attenuated in the image. An RF pulse applied in this way can be used to excite and saturate the protons in tissue with bound water; this will suppress the MR signal from the tissue. In turn the free water can be excited and imaged using an MR pulse sequence. This method is used to suppress background tissue signal when imaging blood vessels in MR angiography (MRA)34 (see later). When magnetization transfer is used to generate contrast between tissues that contain different amounts of bound water, the technique is called magnetization transfer contrast (MTC).

Flow Effects/MR Angiography Flow weighting is an essential contrast for clinical MRI and can be accomplished in a number of ways.3,18,66,79 There are two principal effects in MRI that are used to image flowing blood without exogenous contrast media; these are the same properties that produce artifacts from flow in images of stationary tissue. The first is the time-of-flight effect. This term describes the phenomenon of blood or fluid changing position between the time an excitation RF pulse is applied and the time that signal is acquired from a (potentially different) slice or volume of tissue. The second is the fact that a proton that moves during a phase-encoding gradient pulse will acquire an additional phase shift, so that the phase of that proton’s signal will not correctly represent its position. The time-of-flight effect is important for flow in vessels that move perpendicularly through the imaging slice. In a typical MR experiment with standard slice excitation, the blood in the vessel will flow out of

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging the slice and be replaced by nonexcited blood before the image is obtained. This produces a “black blood” image where the vessels are seen as dark voids in the image (see “Clinical Artifacts”). It is also possible to excite a slice or slab at one location and then to acquire data at a different downstream location at a later time so that the excited blood travels to the location of interest at the time the image is collected. This is the technique that is used in time-of-flight MR angiography (TOF-MRA) to isolate the flowing blood as the subject of the MRI. TOF-MRA can be performed in a 2D or 3D acquisition. This technique provides viable images when there is simple flow and the direction and speed of the flow are known or can be estimated. Under these conditions, TOF-MRA can be performed quickly with high SNR and good separation of vessels from the background. The final step in reconstructing and displaying the MRA study is to prepare a maximum intensity projection (MIP) image from the 3D MRA images.2,64 The MIP technique together with the MRA technique and use of magnetization transfer results in a high degree of background signal suppression in the final image. Most MRIs are reconstructed to show the magnitude of the MR signal, but it is also possible to reconstruct and view the phase information in the MR data. Phase-contrast MRA (PC-MRA)79 studies use a special velocity-encoding gradient pulse to imprint phase shifts to the spins in flowing blood based on their velocity along the magnetic field gradient direction. Spins moving along the gradient with a particular speed accumulate a phase shift during the gradient pulse that is proportional to the speed and the gradient amplitude. By recovering the phase angle information from the data, the velocity of the blood flow in each voxel can be measured. A phase-contrast MRA study of a volume generally involves collecting four images. First, a scan is acquired that is designed to suppress fluid signal and show only the stationary tissue. This image is used as a subtraction mask. Next, the same tissue is scanned three more times, once with the velocity-encoding gradient along each of the three orthogonal directions. Finally, the data from the three velocity-encoded scans are combined to show the vessels in the desired reconstruction; the mask image is used to remove the background from the vessel images. A PC-MRA study takes a relatively long time to acquire, but the data set can produce a lot of different information, including flow velocity maps as well as the same MIP vessel images that are produced in other kinds of MRA studies. The simplest MRA technique to understand is 3D contrastenhanced MRA (CE-MRA).12,82 Unlike the MRA techniques already discussed, in CE-MRA the patient is given an intravenous bolus of gadolinium-based contrast agent. The scan timing is chosen so that the bolus will arrive at the vessel of interest while the scan is performed. The presence of gadolinium leads to a dramatic reduction in the T1 value of water protons, so scanning with a short TR results in high signal from the contrast-influenced blood while the background tissue is mostly saturated (very little T1 relaxation during each TR). The pulse sequence is designed so that the center portion of k-space is collected at the time the bolus arrives at the location of interest. Thus the contrast of the image is determined mostly by the contrast media filling the vessel at the proper time. Data collected before and after the bolus arrival are used in the outer lines of k-space to construct the edge detail of the image. A very large 3D volume can be scanned in a short time (6-7 seconds) using this technique. For patients who can safely be given gadolinium contrast agent, CE-MRA can be used to produce detailed images of large, fast-flowing vessels in a large volume of interest in a very short time. In CE-MRA, as in phase-contrast techniques, a mask image of the region is scanned before the injection. The mask is used to remove

57

background tissues when the MIP is reconstructed. Magnetization transfer techniques are also used in CE-MRA to help saturate the signal from stationary tissue. Advancements in scanning speed have made it possible to capture a time-resolved series of CE-MRA images rather than a snapshot at a single point in time.24 These images can show the filling of vessels like an x-ray angiogram. The general technique for this method, which is also used in dynamic contrast-enhanced (DCE) perfusion imaging (see later), is called keyhole sampling.77 The principle of keyhole sampling is that the full k-space data are collected once, then the center of k-space is scanned repetitively at very short time intervals. The center data require only a small fraction of the total scan time to acquire. For the final images the keyhole data from each time point are combined with the data sampled for the rest of k-space to provide a complete image for the full FOV at each time point. Each MRI manufacturer’s implementation (e.g., TRICKS, TWIST) has a different approach to the size of the keyhole and pulse sequence designs to manage the temporal resolution and frame rate.39

Perfusion Imaging (DSC, DCE, ASL) Perfusion MRI techniques are used to probe the transit of blood flow into and through the microvasculature of tissues.22 Perfusion MRI can be performed with or without exogenous contrast media. Dynamic susceptibility contrast (DSC) imaging40,71 uses gadolinium contrast media with a T2*-weighted echo planar imaging (EPI) technique (see “Multiecho Sequences”). Gadolinium has a strong T2* shortening effect due to its magnetic susceptibility, so the passage of contrast through the tissue of interest produces a drop in the MR signal intensity. A DSC imaging technique begins with acquiring a background or baseline image of the region of interest. The contrast agent is injected and after an appropriate delay for the bolus to reach the anatomy of interest, the T2* EPI data are acquired while the contrast washes into and out of the tissue. Data collection with the EPI sequence can be completed in about 1 second, so the change in signal intensity over time can be measured with a high temporal resolution. To improve the sensitivity of detecting the susceptibility change, a reference image is acquired and the DSC images are subtracted from it to highlight the changes that are due to the contrast agent. The time course of signal intensity changes in images with a high temporal resolution can be used to calculate blood volume (BV), blood flow (BF), and the perfusion parameters time to peak (TTP) and mean transit time (MTT). In brain imaging the relative cerebral blood flow (rCBF) and relative cerebral blood volume (rCBV) are important parameters that are derived from perfusion imaging. Further details of these parameters can be found in Chapter 5, Contrast-Enhanced Magnetic Resonance Imaging. DSC techniques are used to measure first-pass perfusion of tissues that do not have high permeability for the contrast agent. In tumors where the microvascular environment is highly permeable, the contrast agent can diffuse into the tissue as it passes through the capillary bed. This is the mechanism responsible for contrast enhancement of tumors and other lesions in CE-MRI. In conventional CE-MRI the tissues are scanned at a single time point with T1-weighted imaging. In dynamic contrast enhanced (DCE) MRI the wash-in of the contrast agent into the tumor is observed over time using fast T1-weighted imaging and keyhole sampling techniques as described earlier for MRA. The time course of the enhancement is used to calculate perfusion parameters for the tumor, such as ktrans and time to peak enhancement. Refer to Chapter 5 for additional details. As an alternative to approaches that require injection of gadolinium or other paramagnetic contrast agents, arterial spin labeling (ASL)16,80 offers an approach that is completely noninvasive. ASL uses a

58

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

TABLE 3-1  Effects of Changing MRI Parameters on Scan Time, Signal-to-Noise Ratio, and

Image Detail Increasing

Scan Time

SNR

Image Detail

Notes

TR

Increases

Increases*

No change

TE Frequency matrix

No change No change

Decreases Decreases

No change Increases

Phase matrix Field of view (frequency axis) Field of view (phase axis)

Increases No change

Decreases Increases

Increases Decreases

For short TR, increases the number of slices that can be scanned; alters T1 weighting of image Alters T2 weighting of image Decreases chemical shift artifact, truncation artifact, and geometric distortion Decreases truncation artifact Increases coverage

No change

Increases

Decreases

Receiver bandwidth

No change

Decreases*

No change

Flip angle Number of signal averages (NSA) or number of excitations (NEX)

No change Increases

Increases* Increases

No change No change

Increases coverage; Decreases aliasing artifacts (if present) Decreases chemical shift artifact and geometric distortion Alters weighting of gradient echo images

*The effect on signal-to-noise ratio (SNR) is complex, and SNR may increase or decrease depending on the starting value and how much the parameter is changed.

combination of IR and time-of-flight concepts to turn flowing blood into a perfusion contrast agent. First a reference image of the slice of interest is acquired. Upstream from the slice of interest, a slice containing flowing arterial blood is labeled with inversion RF pulses. The labeling slice or slab is chosen so that the blood will travel from the labeling region to the slice of interest in a time TI, so that it arrives as the longitudinal magnetization crosses zero and the labeled blood in the imaging slice yields no transverse magnetization when excited. The slice of interest is scanned as the labeled arterial blood arrives and the acquired image is subtracted from the reference image. The difference image shows the perfusion of the tissue by the labeled spins.1

Fast Imaging for Clinical Practice

Imaging Time and Performance Tradeoffs.  The time needed to acquire imaging data is one of the most significant constraints on clinical MRI. The total scan time for a single image can generally be described as: T = TR ⋅ PE ⋅ NA where T is the total scan time, TR is the repetition time, PE is the number of phase-encoding lines, and NA is the number of averages, or times each phase-encoding line is collected.4,10,28 Note that the TR is set by the user and depends on the type of sequence and target contrast, and PE is determined by the FOV and desired spatial resolution. Collecting each line of k-space once results in the shortest scan time, but in cases where the signal is low, a higher number of averages may be needed. Long scan times contribute to the likelihood of patient movement, resulting in artifacts. In addition, patients find it uncomfortable and unpleasant to remain still and confined in the scanner for long periods of time. For imaging in the chest and abdomen, image quality is also limited by breathing motion, cardiac motion, and other physiologic processes. However, it may sometimes be possible to adjust some acquisition parameters to reduce the scan time (usually at the expense of some aspect of image quality). It is therefore useful for the MRI user to be familiar with the impact various scanning parameters have on imaging time and image quality characteristics.

Table 3-1 describes the effects of changing each individual MRI parameter on scan time, SNR, and image detail, assuming all other parameters are kept the same. The first column lists the operatorcontrolled parameter that can be changed; the other columns list whether each of the outcome parameters increases or decreases. This is a general overview; the details of how much the parameters increase or decrease can be found in other references4,28 for readers who wish to calculate or balance the effects of multiple parameter changes when optimizing MRI protocols.

Accelerating Gradient Echo Sequences.  As described earlier, gradient echo sequences work by using magnetic field gradients to form an echo. Generation of the echo can be done relatively quickly (in a few milliseconds). However, traditional gradient echo sequences are time consuming because the TR is determined by the relaxation times. Any magnetization that remains in the transverse x-y plane after encoding could lead to image artifacts if it is not completely relaxed before the next excitation pulse. Thus TRs on the order of T1 or T2 (hundreds of milliseconds to seconds) must be used for standard gradient echo sequences. However, there are ways to avoid having to wait these long times between the collection of k-space lines. The first is to use additional spoiler gradients, not for spatial encoding but to destroy (or dephase) the magnetization in the x-y plane. If all this leftover magnetization can be rapidly eliminated, another excitation pulse can be applied quickly instead of having to wait for T2 relaxation to take place. This type of sequence is known as a spoiled gradient echo sequence29,31,76,83 (SPGR, T1-FFE, FLASH). Note that the SNR is generally low in these images because small flip angles must be used and T1 relaxation is not taken into account.29 Another way to avoid long TRs is to “reset” the magnetization so that it can be used for the next excitation. This “resetting” involves realigning the magnetization so that it always is pointed in the same direction right before encoding a k-space line. In terms of the pulse sequence, this means all gradients that are used for spatial encoding must be “undone,” or balanced in each TR. Note in Figure 3-13 that each positive gradient is balanced with a negative gradient in each of

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging the gradient directions. These types of sequences are called rewound or balanced gradient echo sequences6,17,57,67,68 (bSSFP, FISP/TRUFI, FIESTA, b-FFE). Because both T1 and T2 relaxation have to be balanced in these sequences, the resulting images have a mixed T1 and T2 contrast. However, unlike spoiled gradient echo sequences, more magnetization is involved in the imaging process and higher flip angles can be used, leading to a high SNR.

Multiecho Sequences.  The basic pulse sequences described work by encoding one line of k-space after each excitation, which can be time consuming if long TRs must be used to obtain the desired contrast. However, in some pulse sequences it is possible to collect multiple lines of k-space for each excitation pulse. If, for example, four k-space lines can be acquired after one excitation, the total scan time is reduced by a factor of 4. It is possible to accelerate both spin echo and gradient

TR

RF

Gslice

Gphase

Gread

ACQ

TE=TR/2

FIG 3-13  Schematic of the balanced steady state free precession (bSSFP) sequence. Note that all of the gradients are rewound, or balanced, in each TR in order to balance the T1 and T2 relaxation.

echo sequences in this way, and these sequences are often called turbo or fast pulse sequences. Multiecho spin echo sequences36,35,49 (RARE, FSE, TSE) work by collecting several spin echoes encoding different lines of k-space after one excitation pulse. The number of echoes collected is known as the echo train length or the turbo factor. Although using a large turbo factor can dramatically reduce the total scan time, it also increases the amount of time needed to collect the data after the excitation. Because relaxation also occurs while data are collected, this means that different lines of k-space can have different contrast weightings. If the contrast changes significantly from the first echo to the last echo, artifacts can appear in the images because of large signal intensity jumps in k-space. For this reason, k-space lines are usually collected in an interleaved manner when using multiecho spin echo sequences (Fig. 3-14). Similarly, if too much relaxation occurs while the data are being collected, the resulting image can have a low SNR. The extreme case is when all of the data for the image are collected after a single excitation pulse. These sequences are called single-shot sequences, and they are typically used to image tissues with long T2 values. For imaging tissues with shorter T2 times, the turbo factor is limited because of the SNR losses, and values of 8 or 16 are commonly used for clinical MRI scanning. However, even these limited turbo factors lead to large reductions in scan time, and multiecho spin echo sequences are essential for clinical spin echo scanning. Gradient echo sequences can also be accelerated by collecting multiple lines after each excitation pulse. The echo planar imaging (EPI) sequence51,73 works by collecting multiple gradient echoes after one excitation by switching the direction of readout gradient for each phase-encoding line. EPI sequences are very rapid, enabling data collection in hundreds of milliseconds for a single image. However, the data collection time is usually long and the signal can be adversely effected by T2* relaxation during data acquisition. This relaxation coupled with the long readout can lead to blurring, image distortions, and low SNR and signal dropouts in EPI images.61,65

Rapid Imaging: SAR Management and Patient Safety.  One of the most serious safety issues in MRI is the heating due to the RF pulses and the potential for thermal injuries to patients where this

RF 90

180

180

TE

TE

180

180

Gslice

Gphase

Gread

ACQ

A

59

B FIG 3-14  A, Schematic of a multiecho spin echo sequence. Here four k-space lines are collected after one excitation pulse at the beginning of the pulse sequence. B, The four k-space lines (solid lines) that were collected using the multiecho spin echo pulse sequence shown in A. Note that the collected lines are not adjacent to one another. The next set of k-space lines that would be collected are shown as dashed lines; this interleaving of lines avoids discontinuities in k-space due to relaxation.

60

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

FT

A

FT

B

FT

C FIG 3-15  A, A fully sampled k-space and its corresponding image. B, A partial Fourier acquisition where scan time is saved by asymmetrically collecting phase-encoding lines. Note the blurring in the resulting image. C, K-space undersampled by a factor of R = 2, where every second phase-encoding line is skipped. Note the aliasing artifacts in the image resulting from this undersampling.

heating is concentrated.14,20,43 The specific absorption rate (SAR) of energy by the tissue is used to give an estimate of the patient’s exposure to the RF energy and potential heating. Careful attention to patient positioning is critical to ensure that there are no points of concentrated heating, such as conductive loops formed by portions of the patient’s body or lines, tubing, and other accessories used during the exam, but it is also important to be aware of SAR in the selection of pulse sequences and parameters and the design of exam protocols. Many of the techniques that enable fast imaging using multiple echoes also increase SAR, so for situations where SAR must be carefully managed, these techniques increase the risk of excessive heating. Because the RF exposure is pulsed, SAR can be reduced by decreasing the amplitude of individual pulses, decreasing the total number of pulses applied in a given time period, and lengthening the time period over which pulses are delivered to allow for dissipation of heat. Large flip angles increase the peak power and RF amplitude of individual pulses. Long echo trains of spin echoes, use of spectral or spatial saturation, and IR techniques increase the number of pulses applied during each TR cycle. Short TR cycles limit the cooling or heat dissipation that can occur before further RF energy is delivered. Thus to avoid

excess SAR, rapid imaging options that do not rely on short TR values or multiple RF pulses are often preferred.

Partial Fourier Reconstructions.  Another way to shorten the total scan time is to reduce the number of phase-encoding lines collected48,74 (see earlier imaging time equation). As shown in Figure 3-15 the effect on the image of reducing the number of phase-encoding lines depends on which lines are skipped. Image resolution is directly related to the extent of the k-space covered. If fewer lines of data are collected by symmetrically leaving out some of the outer portions of k-space, the scan time is reduced along with the image resolution (see Fig. 3-7). Another option is to collect all the data on one side of k-space, using the fact that the k-space of a purely real image is symmetric, and reduce the number of lines collected in the other half. This type of data collection is known as partial Fourier acquisition because the entire Fourier domain (k-space) has not been acquired. As an example, if the k-space matrix is 256 × 256, it would be possible to collect only 192 lines, which are not centered in k-space. Because only three fourths of the lines have been collected, the scan time is reduced by 25%. This type of acquisition is called 6/8 partial Fourier.

61

Coil 1

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging

A Ideal image

Coil 2

E PILS combination

B

Coil sensitivities

C

Single coil images

D

Undersampled images

FIG 3-16  Schematic of the PILS reconstruction. The ideal image (A) is multiplied by the coil sensitivities (B) to yield the single coil images (C). In areas where the coil sensitivity is zero (black), no signal is collected. If the k-space data for these single coils is undersampled by a factor of R = 2, the field of view (FOV) is also reduced by a factor of 2 (D). However, because only half of the FOV contains signal, no signal aliasing will occur. The two half images can be stuck together, forming a complete image in half of the scan time (E).

There are several ways to reconstruct images from partial Fourier data. One is to simply replace the missing k-space lines with zeros (“zero-filling”). For small data reductions (6/8 or 7/8 partial Fourier) there is usually little difference between the images generated from fully collected data sets and those from zero-filled data sets. However, as the amount of data collected is reduced, blurring in the phaseencoding direction will appear (see Fig. 3-15B) because the number of k-space lines in this direction has been reduced. More sophisticated partial Fourier reconstruction approaches use information about the image phase in the center symmetric portion of k-space to improve image quality. These reconstruction techniques include the homodyne reconstruction56 as well as projections onto convex sets (POCS).53 In cases where significant phase differences occur in the images owing to motion or flow, these phase-based reconstructions are preferred to zero-filling. Nearly all clinical scans employ partial Fourier acquisitions to reduce total scan time. Note that the absolute maximum scan time reduction that could be achieved using these techniques is two, because at least half of the k-space data must be collected. In practice, significantly more than half of the data is acquired (6/8 or 7/8 partial Fourier), especially when using an approach like POCS, to accurately capture phase information and avoid SNR losses. An extreme use of partial Fourier is the HASTE sequence (half-Fourier acquisition single-shot turbo-spin echo),59 where only slightly more than half of the k-space lines are collected using a single-shot turbo-spin echo sequence. This sequence allows very rapid data collection at the expense of blurring due to relaxation.

Parallel Imaging.  Parallel imaging is the general name for the family of reconstruction techniques that help reduce scan time by using receiver coils to help localize signals.25,26,62,72 In parallel imaging the k-space data are undersampled, as shown in Figure 3-15C, to save scan time. These undersampled data sets are collected using a receiver array where each receiver has a different sensitivity profile. Nearly all modern MRI scanners use array coils that consist of multiple detectors that are

used to collect the MRI signal. These array coil detectors are placed around the part of the body being imaged; they are primarily sensitive to signals arising from areas near to them. An idealized cartoon of the sensitivity from two coils is shown in Figure 3-16. Each of the coils covers exactly half the FOV of the original image in Figure 3-16A. The top coil in Figure 3-16B is sensitive only to the top half of the image, shown in the coil sensitivity map as the white area. The bottom half of the FOV is not picked up by the upper coil at all, corresponding to the black area in the coil sensitivity map (zeros). Examples of the individual images that would be collected by each of these coils are also shown in Figure 3-16C. Note that the images are black (zero) in places where the coil sensitivities are zero. To create a complete image like that shown in Figure 3-16A, multiple receivers are placed around the area to be imaged, and the individual images are mathematically combined so that no gaps or areas of higher signal intensity appear.63,78 As stated earlier, scan time is reduced in parallel imaging by skipping phase-encoding lines, as shown in Figure 3-15C. If every other line is skipped in an acquisition, the scan time is accelerated by a factor of 2; this is known as an R = 2 acquisition. However, if the separation between the phase-encoding lines is then twice as large as in a standard scan, the FOV of the image will be halved (see “Resolution and Field of View”). If the FOV is smaller than the extent of the excited magnetization, foldover artifacts, also called aliasing or wraparound, occur. This phenomenon can be seen in Figure 3-15C. Thus the goal of parallel imaging is to use differences in receiver coil sensitivities to eliminate this aliasing. The simplest parallel imaging reconstruction algorithm is known as partially parallel imaging with localized sensitivities (PILS),26 and it requires coil sensitivity profiles that are nonoverlapping and homogeneous, as shown in Figure 3-16B. If the data collection is accelerated by a factor of R = 2, pixels from the top half of the image alias with pixels from the bottom half of the image. However, because the receiver coils are only sensitive to half the image, no signal aliasing will occur (see Fig. 3-16D). Each half of the FOV will contain only half of the image. These half-images can simply be concatenated to form a full

62

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

IA

CA1

IA × CA1

IB

CB1

IB × CB1 F1 = IA× CA1 + IB × CB1

IA

CA2

IA × CA2

IB

CB2

IB × CB2 F2 = IA× CA2 + IB × CB2

Ideal image

×

Coil sensitivities

=

Single coil images

Aliased images R=2

SENSE FIG 3-17  SENSE reconstruction. The full field-of-view (FOV) images seen by the individual receiver coils can be obtained by multiplying the initial image (I) with the sensitivity of each coil (C). If these images are undersampled with R = 2 (F), the signal of each pixel in the smaller FOV image is the sum of two pixels of the full images. As aliased images F are collected using multiple coils, the resulting set of linear equations can be solved to find the values of the unaliased pixels IA and IB in a SENSE reconstruction.

image. PILS is not typically used in practice; the requirements on the receiver coils are difficult to realize, but it is a useful method for describing the idea of parallel imaging. The two most commonly used parallel imaging methods are sensitivity encoding (SENSE)62 and generalized autocalibrating partially parallel acquisition (GRAPPA).25 SENSE works by unfolding aliased pixels in the multireceiver images, and GRAPPA works by reconstructing missing lines in k-space. Both are used clinically, but because GRAPPA involves a slightly more complex mathematical description, only SENSE will be described here. Figure 3-17 shows two aliased images collected with an acceleration factor of R = 2 and two receiver coils. Here the coil sensitivities are ranging from 0 to 1 and are shown in color. The red areas indicate that the coil is sensitive to signals coming from that area (coil sensitivity is 1), and the blue areas indicate that the coil is not sensitive to those areas (the receiver sensitivity is 0). In each aliased image, we can write down an equation for each aliased pixel, F1 and F2, in terms of the coil sensitivity values at each voxel location and the ideal unaliased image voxels: F1 = IA ⋅ CA1 + IB ⋅ CB1 F2 = IA ⋅ CA 2 + IB ⋅ CB2 The coil sensitivities can be obtained by performing a prescan of the patient, so that CA1, CB1, CA2, and CB2, are known. In this case the only unknowns are the unaliased voxel values IA and IB. Because there are two equations and two unknowns, the unknown ideal pixel values can be calculated. The same operation is performed for all pairs of aliased voxels, resulting in an unaliased image. SENSE and GRAPPA are limited by the number of receiver coils and their sensitivity profiles—in clinical practice, acceleration factors

of greater than R = 4 are rarely used.15 Note that SENSE requires the coil sensitivity profiles to be known. If the patient moves between the calibration of the coil sensitivities and the accelerated scan, residual aliasing artifacts may remain in the image (see “Clinical Artifacts”). Most images reconstructed using parallel imaging have a lower SNR than their unaccelerated counterparts, and this SNR loss is nonuniform across the image. Areas near the center of the image often have lower SNR values than areas near the edges of the image; this is due to overlapping and similar coil sensitivities in these areas. Despite these minor drawbacks, parallel imaging is used to accelerate nearly every type of image in clinical MRI.70

CLINICAL ARTIFACTS An artifact in medical imaging is any structure visible in an image, or any contrast or image characteristic, that does not actually reflect the anatomy or physiology being imaged. Image artifacts can be produced by hardware malfunction, physical properties of the structure producing the artifacts, limitations in the underlying imaging sequences,  or physiologic processes such as motion. Occasionally artifacts can have diagnostic value—the presence or absence of an artifact reveals something about the structure being imaged. Examples include susceptibility artifacts in metallic objects or blood products, flow artifacts, and chemical shift artifacts. It is important to have a working knowledge of artifacts (so they can be recognized and differentiated from true image characteristics) and their causes (so the problem can be resolved and the artifact eliminated). This section discusses common artifacts with clinical relevance and offers brief explanations of how they arise.

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging

the FOV, the phase imparted on this part of the object by the phaseencoding gradient will be the same as the phase on a part of the object within the FOV, and signals coming from these two areas will not be distinguishable. Thus these signals will appear on top of one another in the image, and the image appears to “wrap around” (Fig. 3-19) (also see Fig. 3-15 [bottom]). This artifact can be eliminated by expanding the FOV so that it contains all of the object. This comes at the expense of resolution if the number of phase-encoding steps is kept constant, or at the expense of time if the number of steps is increased to keep the resolution constant. Note that not all aliasing is problematic; if uninteresting portions of anatomy only fold over onto other unneeded anatomy, aliasing is tolerated and can reduce scan time. If the aliasing is too severe and obscures relevant anatomy, parallel imaging can also be applied to eliminate these foldover artifacts.

Susceptibility Artifact Different materials have different magnetic susceptibilities (ability to become magnetized in a magnetic field). At or near interfaces between materials of drastically different magnetic susceptibilities, there is a distortion of the magnetic field. This distortion can cause areas of bright and dark signal in MRIs, known as a susceptibility artifact. Susceptibility artifacts are accentuated in gradient echo sequences, particularly when long TEs are used (Fig. 3-18). Clinically these artifacts are seen in two important settings—surrounding metallic objects such as surgical clips, screws, shunts, aneurysm clips, stents, and spinal hardware, and in areas of aging blood products, which can contain ferromagnetic material. These artifacts have the effect of degrading the image; however, they can also be diagnostically useful in detecting small hemorrhages (which appear larger or “bloom” at longer TEs) or in detecting surgical clips that are not directly seen on MRI.

Parallel Imaging Artifacts A specific form of aliasing artifact can also occur when using parallel imaging. Recall that k-space lines are skipped in parallel imaging to reduce scan time, but the result is aliasing in the reduced images. If the parallel imaging reconstruction is completely successful, these foldover artifacts will be removed. However, if some of the conditions for parallel imaging are not met (e.g., appropriate calibration information, enough receiver coils with a proper arrangement), residual or faint

Aliasing Aliasing, also known as wraparound or foldover artifacts, occurs when the FOV is set to be smaller than the object being imaged (see “Parallel Imaging”). Technically, aliasing can occur in either the frequency encoding or phase-encoding direction, but it is almost always encountered in the phase-encoding direction. If part of the object lies outside

A

B

FIG 3-18  Susceptibility artifact. Dual gradient echo opposed-phase (A) and in-phase (B) images obtained with a TE of 2.3 and 4.6 msec, respectively. Susceptibility artifact manifests as increased signal void in B in regions of cholecystectomy surgical clips (straight arrow) and iron-containing Gamna-Gandy bodies in the spleen (curved arrow). (Courtesy Dr. Raj M. Paspulati, Case Western Reserve University and University Hospitals of Cleveland.)

FOV FOV/2

A

63

B FIG 3-19  A, Full image where the field of view (FOV) is chosen to be larger than the object in a T1-weighted gradient echo image. B, Aliasing (otherwise known as foldover or wraparound artifacts) observed because the anteroposterior FOV (phase-encoding direction) was selected to be too small. (Courtesy Dr. Vikas Gulani, Case Western Reserve University and University Hospitals of Cleveland.)

64

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B FIG 3-20  A, Parallel imaging artifact (arrow) from an R = 2 accelerated scan. Note that residual aliasing from

the chest wall is still visible following the parallel imaging reconstruction. B, Parallel imaging artifacts from an R = 2 accelerated scan similar to A, with additional SNR reduction near the center of the body. (Courtesy Dr. Vikas Gulani, Case Western Reserve University and University Hospitals of Cleveland.)

aliasing artifacts can still be present in the final images (Fig. 3-20A). These artifacts are typically worse when the reduction factor is high  or when there is a long time between collection of the reference  data needed for the parallel imaging reconstruction and the reduced acquisition. In addition to these residual aliasing artifacts, local increases in image noise can be present at similar locations in the image (see Fig. 3-20B). These nonuniform SNR losses, known as g-factor losses, can be significant if large reduction factors are used. To avoid residual aliasing artifacts and losses in SNR, the reduction factor should typically be kept below half the number of receiver coil elements used in the study.

Gibbs Ringing or Truncation Artifact In standard MRI, it is assumed that the relationship between the MRI and the k-space data collected by the scanner can be described by the discrete Fourier transform. However, real-life objects are continuous, but images have a finite spatial resolution. Because the spatial resolution in the image is related to the extent of k-space covered, only an infinitely large k-space would actually reflect the continuous object. In reality, all of k-space can never be collected because this would require an infinite scan time. Thus in practice, only a finite number of k-space lines are collected, which in signal processing is known as truncation or windowing. The result of collecting only a finite number of k-space lines is a “ringing,” or an oscillating over- and undershoot in image intensity, particularly at high-contrast interfaces (Fig. 3-21). This ringing is called Gibbs ringing or truncation artifact in signal processing and imaging and can be reduced by applying a filter to the k-space data before performing the Fourier transform.

Motion Artifacts and Ghosting Motion artifacts arise from object motion during the data collection process. Because of the Fourier relationship between k-space and the image, each data point in k-space can affect each point in the corresponding image. Thus a single corrupt data point in k-space can lead to artifacts covering the whole image. This effect is often seen in breath-hold failures (Fig. 3-22). Although only a single or a few lines of k-space are corrupted, the artifacts can spread over the entire image in the form of ghosting. Because the frequency encoding occurs over a few milliseconds, motion is rarely fast enough to cause problems in

FIG 3-21  Gibbs ringing or truncation artifact. Alternating bands of bright and dark signal (arrows) at a high-contrast interface between a renal cyst and perinephric fat on a fat-saturated T2-weighted turbo-spin echo image.

this direction. However, phase encoding occurs over a number of repeated steps and is generally accomplished over seconds to minutes, allowing considerable time for objects to move. The moving objects are misrepresented in the image, resulting in motion artifacts. If the motion is not periodic (e.g., patient movement, bowel peristalsis), the consequence for the image is blurring and general misrepresentation of signal across the phase-encoding direction (see Fig. 3-22). If the motion is periodic (e.g., blood vessel or CSF pulsation), the edges of the moving object can have discrete misrepresentations (ghosts) across the image, depending on how fast the object is moving in comparison to the phase-encoding steps; the ghosts are spaced farther apart if the rate of motion is high (Fig. 3-23). Occasionally these ghosts can be mischaracterized as tumors or other pathology, but careful examination of the image allows proper identification. If there is any doubt, swapping phase- and frequency-encoding directions can confirm that there is no true lesion.

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging Flow Artifacts The effect of flow was recognized early in the history of MRI, and flow can be encoded as part of the imaging process, as previously described. However, flow can also cause image artifacts. One type of flow artifact is ghosting due to periodic flow (see Fig. 3-23). Another type of flow artifact is the potential misrepresentation of the signal level owing to an increase or decrease in magnetization caused by flow. These artifacts can be categorized as black blood or bright blood, depending on the imaging appearance of flowing blood. Black blood artifacts arise if blood flowing into the plane of the image has experienced saturating RF or gradient pulses in another slice. If this encoded magnetization is subjected to additional encoding in its new position in the slice of interest, the flowing magnetization may not be refocused and remain dark on the image. These areas in the image, which contain blood, can appear dark (i.e., as flow voids) in the image (Fig. 3-24A). Similarly, bright blood artifacts are seen when unsaturated spins flow into the slice of interest from outside the volume being imaged, resulting in

FIG 3-22  Patient motion creates an image blur (arrows) in a fatsaturated T1-weighted 3D volume-interpolated breath-hold examination (VIBE; gradient echo) sequence.

A

65

bright signal in blood vessels (see Fig. 3-24B). Flow artifacts are classic examples of how artifacts can be consciously exploited by the imager for diagnostic purposes.

Radiofrequency Inhomogeneity The MRI process relies not only on a uniform main magnetic field B0 but also on uniform RF pulses, often called the transmitted B1 field. Any deviations in the strength of the RF pulse, or flip angle, over the FOV can lead to patches of higher or lower signal across the image. These inhomogeneities are more common at higher field strengths and are significantly more noticeable at 3.0 T than at 1.5 T. In addition, high-field images can have a dark periphery and a bright center. This phenomenon arises from a distortion of the transmitted RF field caused by the patient’s presence within the coil. These distortions lead to constructive and destructive interference in transmitted RF pulses, giving rise to signal inhomogeneities in the image. These effects have also been termed dielectric resonance or field focusing in the literature.37,69

FIG 3-23  Ghosts (arrows) of the abdominal aorta due to periodic aortic pulsations seen on an axial fat-saturated T1-weighted FLASH (gradient echo) image.

B FIG 3-24  A, Dark blood-flow artifacts (arrows) in a T2-weighted half-Fourier acquisition single-shot turbo-spin echo (HASTE) image. B, Bright blood-flow artifacts (arrows) in true fast imaging with steady-state precession (FISP) of the abdomen.

66

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B FIG 3-25  A, Chemical shift artifacts arising from a low bandwidth during data collection. Note that especially around the kidneys, where areas of fat and water interface, dark (purple arrow) and bright (orange arrow) edges appear owing to a shift between the apparent locations of fat and water signals. B, The “India ink” or “etching” artifact due to cancellation of signal from out-of-phase fat and water magnetization can be seen at the edges of all the solid organs and musculature.

Chemical Shift Artifacts As described earlier, protons in different chemical environments have different resonance frequencies. However, the MRI process uses frequency information to determine the spatial location of magnetization. If two different chemical species with different resonance frequencies are being imaged, spatial misregistration and thus image artifacts can result. For instance, the chemical shift of fat relative to water is approximately −223 Hz at 1.5 T, meaning that the fat resonance frequency is 223 Hz lower than that of water. Thus when forming an image, signal from a fat-containing voxel appears to arise from voxels that exist 223 Hz away from that position. The result is essentially two superimposed images, one representing water and one representing fat, which can have a shift of several pixels in the frequency encoding direction relative to each other. Because there is usually significantly more water in the images, the effects of this superposition are most noticeable at the borders of structures containing fat, where they form dark and bright bands at opposite edges (Fig. 3-25A). A second type of chemical shift artifact arises at interfaces with mixed-content voxels if signal is collected at particular TEs. Because of the different resonant frequencies, fat and water signals will sometimes be pointing in opposite directions in the x-y plane. When this happens the signals are said to be out of phase, which occurs at TEs of approximately 2.2 milliseconds, 6.6 milliseconds, and so forth, at 1.5 T. Voxels at fat-water interfaces can be expected to be nearly equally fat and water; thus at these specific TEs, fat and water signals nearly cancel each other out at these locations, leaving a dark band. This phase cancellation is sometimes termed the India ink or etching artifact (see Fig. 3-25B). This phenomenon is another example of an artifact that is often exploited for diagnostic purposes, and it is the underlying principle in in-phase and opposed-phase imaging of the body, with application in the liver, adrenal glands, and kidneys. For example, the phase-cancellation artifact is diagnostic of angiomyolipomas in the kidney or lipomas in water-containing solid organs. These fatcontaining lesions produce a phase-cancellation artifact at their interface with normal renal parenchyma.

ACKNOWLEDGMENTS The authors would like to thank Dr. Vikas Gulani and Dr. Mark Griswold for providing figures from the previous edition.

REFERENCES 1. Alsop DC, Detre JA, Golay X, et al: Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: A consensus of the ISMRM perfusion study group and the European Consortium for ASL in Dementia. Magn Reson Med 116:102–116, 2014. doi: 10.1002/ mrm.25197. 2. Anderson CM, Saloner D, Tsuruda JS, et al: Artifacts in maximumintensity-projection display of MR angiograms. Am J Roentgenol 154:623–629, 1990. doi: 10.2214/ajr.154.3.2106232. 3. Axel L: Blood flow effects in magnetic resonance imaging. AJR Am J Roentgenol 143:1157–1166, 1984. doi: 10.2214/ajr.143.6.1157. 4. Bernstein MA, King KF, Zhou XJ: Handbook of MRI Pulse Sequences, 1st ed, San Diego, 2004, Academic Press. 5. Bernstein MA, King KF, Zhou XJ: Adiabatic radiofrequency pulses. In Handbook of MRI Pulse Sequences, Burlington, Mass., 2004, Elsevier, pp 190–198. 6. Bieri O, Scheffler K: Fundamentals of balanced steady state free precession MRI. J Magn Reson Imaging 38:2–11, 2013. doi: 10.1002/ jmri.24163. 7. Bitar R, Leung G, Perng R, et al: MR pulse sequences: What every radiologist wants to know but is afraid to ask. Radiographics 26:513–537, 2006. doi: 10.1148/rg.262055063. 8. Bley TA, Wieben O, François CJ, et al: Fat and water magnetic   resonance imaging. J Magn Reson Imaging 31:4–18, 2010. doi: 10.1002/jmri.21895. 9. Bracewell RN: The Fourier Transform and its Applications, New York, 1978, McGraw-Hill. 10. Brown MA, Semelka RC: MRI: Basic Principles and Applications, 4th ed, Hoboken, 2010, Wiley-Blackwell. 11. Bydder GM, Young IR: MR imaging: Clinical use of the inversion recovery sequence. J Comput Assist Tomogr 9:659–675, 1985. 12. Chandra T, Pukenas B, Mohan S, et al: Contrast-enhanced magnetic resonance angiography. Magn Reson Imaging Clin N Am 2012. doi: 10.1016/j.mric.2012.08.007. 13. De Coene B, Hajnal JV, Gatehouse P, et al: MR of the brain using fluid-attenuated inversion recovery (FLAIR) pulse sequences. Am J Neuroradiol 13:1555–1564, 1992. 14. Dempsey MF, Condon B: Thermal injuries associated with MRI. Clin Radiol 56:457–465, 2001. doi: 10.1053/crad.2000.0688. 15. Deshmane A, Gulani V, Griswold MA, et al: Parallel MR imaging. J Magn Reson Imaging 36:55–72, 2012. doi: 10.1002/jmri.23639. 16. Detre JA, Leigh JS, Williams DS, et al: Perfusion imaging. Magn Reson Med 23:37–45, 1992. doi: 10.1002/mrm.1910230106.

CHAPTER 3  Imaging Principles in Magnetic Resonance Imaging 17. Duerk JL, Lewin JS, Wendt M, et al: Remember true FISP? A high SNR, near 1-second imaging method for T2-like contrast in interventional MRI at .2 T. J Magn Reson Imaging 8(1):203–208, 1998. 18. Dumoulin CL, Hart HR, Jr: Magnetic resonance angiography. Radiology 161:717–720, 1986. doi: 10.1148/radiology.161.3.3786721. 19. Edelstein WA, Hutchison JM, Johnson G, et al: Spin warp NMR imaging and applications to human whole-body imaging. Phys Med Biol 25:751–756, 1980. 20. Eising EG, Hughes J, Nolte F, et al: Burn injury by nuclear magnetic resonance imaging. Clin Imaging 34:293–297, 2010. doi: 10.1016/j. clinimag.2009.06.025. 21. Ernst RR: Application of Fourier transform spectroscopy to magnetic resonance. Rev Sci Instrum 37:93, 1966. 22. Essig M, Shiroishi MS, Nguyen TB, et al: Perfusion MRI: The five most frequently asked technical questions. AJR Am J Roentgenol 200:24–34, 2013. doi: 10.2214/AJR.12.9543. 23. Reference deleted in proofs. 24. Grist TM, Mistretta CA, Strother CM, et al: Time-resolved angiography: Past, present, and future. J Magn Reson Imaging 36:1273–1286, 2012. doi: 10.1002/jmri.23646. 25. Griswold MA, Jakob PM, Heidemann RM, et al: Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47(6):1202–1210, 2002. 26. Griswold MA, Jakob PM, Nittka M, et al: Partially parallel imaging with localized sensitivities (PILS). Magn Reson Med 44(4):602–609, 2000. 27. Grossman RI, Gomori JM, Ramer KN, et al: Magnetization transfer: Theory and clinical applications in neuroradiology. Radiographics 14:279–290, 1994. doi: 10.1148/radiographics.14.2.8190954. 28. Haacke EM, Brown RW, Thompson MR, et al: Magnetic Resonance Imaging: Physical Principles and Sequence Design, New York, 1999, Wiley-Liss. 29. Haase A, Frahm J, Matthaei D, et al: Flash imaging—Rapid NMR imaging using low flip-angle pulses. J Magn Reson 67(2):258–266, 1986. 30. Hajnal JV, De Coene B, Lewis PD, et al: High signal regions in normal white matter shown by heavily T2-weighted CSF nulled IR sequences.   J Comput Assist Tomogr 16:506–513, 1992. 31. Hargreaves BA: Rapid gradient-echo imaging. J Magn Reson Imaging 36:1300–1313, 2012. doi: 10.1002/jmri.23742. 32. Hashemi R, Bradley WG, Lisanti CJ: MRI: The Basics, Philadelphia, 2010, Lippincott Williams & Wilkins. 33. Hauger O, Dumont E, Chateil J-F, et al: Water excitation as an   alternative to fat saturation in MR imaging: Preliminary results in musculoskeletal imaging. Radiology 224:657–663, 2002. doi: 10.1148/ radiol.2243011227. 34. Henkelman RM, Stanisz GJ, Graham SJ: Magnetization transfer in MRI: A review. NMR Biomed 14:57–64, 2001. doi: 10.1002/nbm.683. 35. Hennig J: Multiecho sequences with low refocusing flip angles. J Magn Reson 78:397–407, 1988. 36. Hennig J, Nauerth A, Friedburg H: RARE imaging: A fast imaging method for clinical MR. Magn Reson Med 3(6):823–833, 1986. 37. Hoult DI, Phil D: Sensitivity and power deposition in a high-field imaging experiment. J Magn Reson Imaging 12(1):46–67, 2000 Jul. 38. http://physics.nist.gov/cgi-bin/cuu/Value?gammapbar. 39. Ivancevic MK, Geerts L, Weadock WJ, et al: Technical principles of MR angiography methods. Magn Reson Imaging Clin N Am 2009. doi: 10.1016/j.mric.2009.01.012. 40. Johnson G, Wetzel SG, Cha S, et al: Measuring blood volume and vascular transfer constant from dynamic, T2*-weighted contrastenhanced MRI. Magn Reson Med 51:961–968, 2004. doi: 10.1002/ mrm.20049. 41. Jung BA, Weigel M: Spin echo magnetic resonance imaging. J Magn Reson Imaging 37:805–817, 2013. doi: 10.1002/jmri.24068. 42. Kaldoudi E, Williams SC, Barker GJ, et al: A chemical shift selective inversion recovery sequence for fat-suppressed MRI: Theory and experimental validation. Magn Reson Imaging 11:341–355, 1993. doi: 10.1016/0730-725X(93)90067-N. 43. Kanal E, Barkovich AJ, Bell C, et al: ACR guidance document on MR safe practices: 2013. J Magn Reson Imaging 37:501–530, 2013. doi: 10.1002/ jmri.24011.

67

44. Lauterbur P: Image formation by induced local interactions:   Examples employing nuclear magnetic resonance. Nature 242:190–191, 1973. 45. Le Bihan D, Breton E, Lallemand D, et al: MR imaging of intravoxel incoherent motions: Application to diffusion and perfusion in neurologic disorders. Radiology 161:401–407, 1986. 46. Le Bihan D, Poupon C, Amadon A, et al: Artifacts and pitfalls in diffusion MRI. J Magn Reson Imaging 2006. doi: 10.1002/jmri.20683. 47. Lee S-K, Kim DI, Kim J, et al: Diffusion-tensor MR imaging and fiber tractography: A new method of describing aberrant fiber connections in developmental CNS anomalies. Radiographics 25:53–65, discussion 66–68, 2005. doi: 10.1148/rg.251045085. 48. Liang ZP, Lauterbur P: Principles of Magnetic Resonance Imaging: A Signal Processing Perspective, Piscataway, NJ, 1999, IEEE Press. 49. Listerud J, Einstein S, Outwater E, et al: First principles of fast spin echo. Magn Reson Q 8(4):199–244, 1992. 50. Maki JH, Chenevert TL, Prince MR: Contrast-enhanced MR angiography. Abdom Imaging 23(5):469–484, 1998. 51. Mansfield P: Multi-planar image formation using NMR spin echoes.   J Phys C: Solid State Phys 10:L55–L58, 1977. 52. Markl M, Leupold J: Gradient echo imaging. J Magn Reson Imaging 35:1274–1289, 2012. doi: 10.1002/jmri.23638. 53. McGibney G, Smith MR, Nichols ST, et al: Quantitative evaluation of several partial Fourier reconstruction algorithms used in MRI. Magn Reson Med 30(1):51–59, 1993 Jul. 54. Mukherjee P, Berman JI, Chung SW, et al: Diffusion tensor MR imaging and fiber tractography: Theoretic underpinnings. Am J Neuroradiol 2008. doi: 10.3174/ajnr.A1051. 55. Mukherjee P, Chung SW, Berman JI, et al: Diffusion tensor MR imaging and fiber tractography: Technical considerations. Am J Neuroradiol 29:843–852, 2008. doi: 10.3174/ajnr.A1052. 56. Noll DC, Nishimura DG, Macovski A: Homodyne detection in magnetic resonance imaging. IEEE Trans Med Imaging 10(2):154–163, 1991. 57. Oppelt A, Graumann R, Barfuss H, et al: FISP: A new fast MRI sequence. Electromedica 54:15–18, 1986. 58. Paschal CB, Morris HD: K-space in the clinic. J Magn Reson Imaging 19:145–159, 2004. doi: 10.1002/jmri.10451. 59. Patel MR, Klufas RA, Alberico RA, et al: Half-Fourier acquisition single-shot turbo spin-echo (HASTE) MR: Comparison with fast spin-echo MR in diseases of the brain. AJNR Am J Neuroradiol 18:1635–1640, 1997. 60. Plewes DB, Kucharczyk W: Physics of MRI: A primer. J Magn Reson Imaging 35:1038–1054, 2012. doi: 10.1002/jmri.23642. 61. Poustchi-Amin M, Mirowitz SA, Brown JJ, et al: Principles and applications of echo-planar imaging: A review for the general radiologist. Radiographics 21(3):767–779, 2001. 62. Pruessmann KP, Weiger M, Scheidegger MB, et al: SENSE: Sensitivity encoding for fast MRI. Magn Reson Med 42(5):952–962, 1999. 63. Roemer PB, Edelstein WA, Hayes CE, et al: The NMR phased array. Magn Reson Med 16:192–225, 1990. 64. Rossnick S, Laub G, Braeckle R, et al: Three dimensional display of blood vessels in MRI. In Proceedings of the IEEE Computers in Cardiology Conference, 1986, pp 193–196. 65. Rzedzian R, Chapman B, Mansfield P, et al: Real-time nuclear magnetic resonance clinical imaging in paediatrics. Lancet 2(8362):1281–1282, 1983. 66. Saloner D: The AAPM/RSNA physics tutorial for residents. An introduction to MR angiography. Radiographics 15:453–465, 1995. doi: 10.1148/radiographics.15.2.7761648. 67. Scheffler K, Hennig J: Is TrueFISP a gradient-echo or a spin-echo sequence? Magn Reson Med 49(2):395–397, 2003. 68. Scheffler K, Lehnhardt S: Principles and applications of balanced SSFP techniques. Eur Radiol 13(11):2409–2418, 2003. 69. Schick F: Whole-body MRI at high field: Technical limits and clinical potential. Eur Radiol 15(5):946–959, 2005. 70. Schönberg SO, Dietrich O, Reisner MF, editors: Parallel imaging in clinical MR applications, Berlin, 2007, Springer-Verlag. 71. Shiroishi MS, Castellazzi G, Boxerman JL, et al: Principles of T2*weighted dynamic susceptibility contrast MRI technique in brain tumor

68

72.

73. 74. 75.

76.

77.

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging imaging. J Magn Reson Imaging 00:296–313, 2014. doi: 10.1002/ jmri.24648. Sodickson DK, Manning WJ: Simultaneous acquisition of spatial harmonics (SMASH): Fast imaging with radiofrequency coil arrays. Magn Reson Med 38(4):591–603, 1997. Stehling ML, Turner R, Mansfield P: Echo-planar imaging: Magnetic resonance imaging in a fraction of a second. Science 245:43–50, 1991. Tsao J: Ultrafast imaging: Principles, pitfalls, solutions, and applications. J Magn Reson Imaging 32:252–266, 2010. doi: 10.1002/jmri.22239. Twieg D: The k-trajectory formulation of the NMR imaging process with applications in analysis and synthesis of imaging methods. Med Phys 10(5):610–621, 1983. van der Meulen P, Groen JP, Cuppen JJ: Very fast MR imaging by field echoes and small angle excitation. Magn Reson Imaging 3(3):297–299, 1985. Van Vaals JJ, Brummer ME, Dixon WT, et al: “Keyhole” method for accelerating imaging of contrast agent uptake. J Magn Reson Imaging 3:671–675, 1993. doi: 10.1002/jmri.1880030419.

78. Walsh DO, Gmitro AF, Marcellin MW: Adaptive reconstruction of phased array MR imagery. Magn Reson Med 43:682–690, 2000. 79. Wheaton AJ, Miyazaki M: Non-contrast enhanced MR angiography: Physical principles. J Magn Reson Imaging 36:286–304, 2012. doi: 10.1002/jmri.23641. 80. Williams DS, Detre JA, Leigh JS, et al: Magnetic resonance imaging of perfusion using spin inversion of arterial water. Proc Natl Acad Sci U S A 89:212–216, 1992. doi: 10.1073/pnas.89.9.4220e. 81. Wolff SD, Eng J, Balaban RS: Magnetization transfer contrast: Method for improving contrast in gradient-recalled-echo images. Radiology 179:133–137, 1991. doi: 10.1148/radiology.179.1.2006263. 82. Zhang H, Maki JH, Prince MR: 3D contrast-enhanced MR   angiography. J Magn Reson Imaging 25:13–25, 2007. doi: 10.1002/jmri.20767. 83. Zur Y, Wood ML, Neuringer LJ: Spoiling of transverse magnetization in steady-state sequences. Magn Reson Med 21(2):251–263, 1991.

4  Imaging Principles in Magnetic Resonance Angiography Eric Schrauben and Oliver Wieben

INTRODUCTION Magnetic resonance angiography (MRA) enables noninvasive assessment of vascular disease in both the arterial and venous circulation. Owing to the availability of volumetric data sets without the need for ionizing radiation exposure and facilitated by technologic and methodological advances in hardware, data sampling, and image reconstruction approaches, MRA has become a powerful tool for the diagnosis of a variety of vascular pathologies. MRA techniques explore different mechanisms for contrast generation and can provide clinical information on vessel luminography or function. The method of choice depends on the diagnostic task and the specific demands of the vascular territory. Time-of-flight (TOF) and phase-contrast (PC) MRA differentiate the signal in vessels from surrounding tissues based on inflow effects (TOF) or velocity encoding (PC) of the moving blood. The latter can be used to actually quantify velocities and blood flow. Balanced steadystate free precession (bSSFP) has been recently added to the repertoire of clinically used MRA techniques and generates vascular maps based on a T2/T1-weighted image contrast that can be enriched by using concepts of inflow encoding and signal saturation. TOF, PC, and bSSFP MRA are referred to as non–contrast enhanced (NCE) MRA techniques, since they do not require any external contrast agents. Yet another very commonly used approach is contrast-enhanced MRA (CE MRA), which uses an intravenous injection of a T1-  shortening, gadolinium (Gd)-based contrast bolus. The much higher signal in the blood pool combined with relatively short acquisition times results in a very robust acquisition that provides large anatomic coverage, low sensitivity to artifacts, and short scan times. This chapter will present the basic physical principles and imaging strategies related to the most common approaches for MRA. The benefits and disadvantages of the various approaches will be illustrated, and considerations for physiologic gating to the cardiac and respiratory cycle will be discussed.

BASIC GRADIENT ECHO SEQUENCE Ideally, all MR data are acquired while in a constant relaxation state, with high spatial resolution, a high signal-to-noise ratio (SNR), ultrashort imaging time, and no artifacts. The difficulty in achieving all these properties has led to the development of many acquisition methods that differ in image contrast, acquisition speed, SNR, susceptibility to and type of artifacts, energy deposited in the imaged patient, suppression of unwanted signal such as fat, and other aspects. Their corresponding images represent a combination of tissue-specific parameters T1, T2, and proton density ρ and scan-specific parameters

such as repetition time (TR), echo time (TE), flip angle, field of view (FOV), spatial resolution, magnetization preparation, and so forth. All MRA imaging strategies described here rely on gradient echo acquisitions26 for image generation, which have been introduced in Chapter 3. Gradient echoes allow for faster scanning compared to spin echo acquisitions, owing to the potential for short TEs and short TRs, formed using fast switching of gradients to facilitate a measurable echo formation. The pulse sequence diagram (PSD) for a basic gradient recalled echo (GRE) sequence without phase encoding is shown in Figure 4-1. After radiofrequency (RF) excitation of the slice (twodimensional [2D] acquisition) or volume (three-dimensional [3D] acquisition) of interest with a flip angle α less than 90 degrees, a negative readout gradient lobe dephases the transverse magnetization and then refocuses it with a second lobe of inverted polarity, also referred to as the readout gradient, to form a gradient echo at the echo time TE. Data are acquired during the data acquisition window while the readout gradient is played out. Unlike spin echo sequences, this echo is not characteristic of T2 relaxation but rather of T2* relaxation, because dephasing from chemical shift, magnetic susceptibility, and B0 inhomogeneity cannot be inverted by the gradient reversal. With a long repetition time (TR ≫ T2), there is no carryover of transverse magnetization between excitations because the transverse magnetization is completely dephased. In fast GRE imaging, the TR  is significantly reduced and generally less than the T2 values of biological tissues. Under this condition, the transverse magnetization of a preceding RF pulse is not completely dephased and generally results  in a complex superposition of free induction decays (FIDs), primary echoes, stimulated echoes, and secondary and higher-order echoes for a train of periodic RF pulses. Under certain conditions, a steady state can be reached from repetition to repetition for one or more components of the magnetization.44 It is necessary to enforce a steady-state condition for the transverse magnetization in order to acquire data with consistent signal from one excitation to the next. As we will see below, the spoiled gradient echo sequence and the bSSFP sequence are such implementations and form the basis of all MRA sequences discussed here.

CONTRAST-ENHANCED MRA In CE MRA,18,40,41 contrast between blood and background signal is achieved through the injection of a T1-shortening contrast agent. Although intraarterial injections similar to conventional x-ray angiography are possible,36 practically all CE MRA exams are conducted with intravenous injections, which dramatically decreases risks and costs as well as improves patient comfort. The agent is extracellular and is typically based on gadolinium, a paramagnetic metal that decreases both

69

70

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging α

α

SPOILED GRADIENT ECHO

α,Φ RF

t +2G

Greadout

Echo formation

–G

FID

t

T2*

α,Φ

RF

Spoiler

Gslice

Greadout

Signal

DAQ

t

TE

Signal TR

FIG 4-1  Basic components of a gradient recalled echo (GRE) pulse sequence diagram. A radiofrequency (RF) pulse with a flip angle α (<90 degrees) is applied to excite tissue within the volume/slice of interest (slice/volume select gradient not shown here for simplicity). A free induction decay (FID) follows as spins relax to their equilibrium state, but it is difficult to measure and typically not used for image generation. The negative lobe of the readout gradient (−G) purposefully dephases the spins in the imaging volume. Then the direction of the readout gradient is inverted, and when the gradient areas of the prewinder and rewinder are equal at the echo time t = TE, the spins are refocused and form an echo. The data acquisition window (DAQ) is centered around the echo time. Typically a single line of k-space is acquired for each excitation, and this process is repeated until all data are acquired.

T1 and T2 relaxation time. Pure Gd is highly toxic, so clinically used formulations must be chelated to ligands for safe delivery to human patients.15,17 Recently the use of Gd-based contrast agents in patients with severely compromised kidney function has been linked to the occurrence of nephrogenic systemic fibrosis (NSF).46,47 Proper screening of patients at risk through measures of glomerular filtration rate and serum creatinine level minimizes this risk and provides a high safety profile.4

Imaging Basics Image contrast in CE MRA is explored by the use of heavily T1-weighted sequences. The mixing of the contrast agent with the blood pool will generate high (bright) signal in the vasculature while most other tissues have a much longer T1 relaxation time and appear dark. The spoiled gradient echo sequence is well suited for this imaging task when played with a short TR. It is designed so that the transverse magnetization generated by an excitation pulse only contributes to the signal acquired in the echo period immediately following this pulse. Therefore these sequences are also referred to as longitudinal steadystate sequences, because only the longitudinal component of the magnetization is in a steady state.32 In case of ideal spoiling, the transverse magnetization Mxy would be completely destroyed immediately after each repetition. In practice this cannot be achieved, but it is possible to almost eliminate the signal contributions from all preceding pulses to the transverse magnetization Mxy for each TR. The phase ϕ of both the excitation pulse and the receiver is altered in a pseudorandom sequence13 to eliminate contributions caused by transversal magnetizations from previous excitations. In addition, a dephasing gradient is played out, usually along the slice-encoding direction after the acquisition of each echo to dephase the transverse magnetization. The dephasing gradients need to be strong enough to ensure a minimal dephasing of 2π of the spin ensembles in all voxels for a homogeneous signal

t

TE TR

FIG 4-2  The spoiled gradient echo sequence is the method of choice for CE MRA data acquisition. The transverse magnetization is forced into a steady state by a combination of gradient dephasing, also called spoiling, which is usually played as a strong gradient lobe along the slice-select and/or readout directions, and phase spoiling by adding a pseudorandom phase ϕ to the RF excitation.

across the images. Figure 4-2 shows the timing diagram for the 3D spoiled gradient-recalled echo (SPGR) pulse sequence. In spoiled gradient echo imaging for CE MRA, the TR values are reduced to less than 10 milliseconds. If the echo time is also very short (TE ≪ T2*), it can be shown that the signal only depends on TR, T1, and α.53 For large flip angles and a short TR, the sequence then becomes heavily T1 weighted as desired for this application. Vendor implementations and acronyms of this sequence include (Spoiled GRASS = Spoiled Gradient Refocused Acquisition in the Steady State, GE Healthcare, Waukesha, WI), FLASH (Fast Low Angle Shot, Siemens Medical Systems, Erlangen, Germany) and FFE (Fast Field Echo, Philips, Best, The Netherlands).16 Some publications have been dedicated to assist the interested reader in listing and comparing acronyms for MR pulse sequences.1 One tissue that can interfere with the CE vascular signal is fat, which also has a relatively short T1 time. In many parts of the body, high amounts of fat can be present in background tissue and may diminish vascular visualization from CE MRA, especially in the commonly used projection images that represent the whole image volume from various angles. Fat suppression pre-pulses or subtraction of precontrast image masks can be used to enhance the delineation of arteries in such cases.

Bolus Timing.  Intravenous delivery of the contrast can be performed manually, but the method of choice is use of a power injector with a controllable injection rate (in mL/sec) followed by a saline flush. Dosing is typically based on body weight (e.g., 0.2 mmol/kg) but may also be a fixed volume. The injection rate, dosing, and acquisition timing must all be tailored to the vascular territory of interest and to the suspected pathologic condition. An extremely important aspect of CE MRA lies in the timing of MR acquisition with what is known as the first-pass of contrast through the arteries of interest after the bolus has passed from the venous injection site to the heart and through the pulmonary circulation (Fig. 4-3). There will be a brief time period where the arterial signal is at its peak without any signal in the veins. After this arterial first-pass, contrast will circulate through the venous side, eventually dispersing into background tissue for extracellular Gd chelates. Timing of the acquisition

CHAPTER 4  Imaging Principles in Magnetic Resonance Angiography T1 (ms) 1000

2000

200

71

50

Signal intensity → Blood←Soft tissues→Fat

Fluid

Blood + contrast B

A Contrast injection Signal

Vein

Acquisition delay

Artery

Background Time

FIG 4-3  Timing diagram of a CE MRA exam. After intravenous contrast injection, the acquisition is delayed some time to allow for bolus arrival from the venous injection site to the arteries of interest. The mixing of the blood and the Gd-based contrast agent shortens the T1 of the blood considerably below the T1 relaxation times of the surrounding tissues. The bolus arrives in the arteries (red line), quickly followed by venous enhancement (blue line), and surrounding background tissue (green line). As shown during different CE MRA acquisitions in the thoracic descending aorta, short acquisition times and correct timing of the acquisition are vital in capturing the arterial window with peak enhancement (A). Delayed images (B) show decreased signal and contrast as well as venous signal. Note that both images have the same window/level.

to the first-pass window therefore will best delineate arterial signal from background signal and venous interference. Two main methods of determining this window of first-pass capture are typically used clinically. The first employs a test bolus of a small amount of contrast (2-3 mL) and a subsequent saline flush coupled with a dynamic series of a low-resolution single slice images in the main artery of interest.7 Saturation slabs above and below the slice of interest can be applied to remove inflow effects. This allows for manual measurement of the contrast arrival times in arteries, veins, and background tissue, so that the main acquisition can be timed accordingly. The second method uses a technique similar to fluoroscopy for either a manual or automated trigger to start the CE MRA acquisition. For example, manual visual inspection of contrast traversing vasculature in a single thick slice is monitored using continually updated images, usually with a temporal resolution of around 1 second.43 The acquisition is triggered by the operator, who watches the arrival of contrast. For cases with unusually fast or slow flow, this technique provides the advantage of giving the operator direct feedback on flow characteristics. The thick slice also provides dynamic information for large anatomic coverage. In the automatic triggering method, which is used less frequently, a single voxel or region of interest within the artery is monitored and the CE MRA scan is triggered when this value rises above some predetermined threshold value.11 The advantage with this method is the removal of user dependence; however, acquisitions for extremely slow flow will not be correctly delayed with this technique. CE MRA acquisitions in the abdomen are complicated by the additional need for a breath hold that is synchronized to the bolus arrival and data acquisition. The approaches are similar to the ones just described, but the time lost for the initiation of a breath hold can lead to nonoptimal acquisition windows past the peak enhancement or with enhancement in venous structures.

Scan Time Considerations.  Many technical aspects of CE MRA must be correctly prescribed and monitored for a successful CE MRA

scan completion. Imaging parameters are chosen to optimize the conflicting aims of high spatial resolution, short scan times, and bright signal of arterial arrival from contrast injection (Fig. 4-4). Because the primary target of a CE MRA exam is to capture enhancement during the arterial window, acquisition of the 3D volume should be performed as quickly as possible. In vascular regions of short turnaround between bolus arrival in the arteries and veins (e.g., kidneys, brain) or in pathologic conditions such as arteriovenous malformations (AVMs), this window may be as short as 5 seconds. Consider a 3D renal CE MRA acquisition with an in-plane matrix of 256 × 256 and 64 slices with a typical TR of 5 milliseconds. Without the use of any accelerated imaging, the scan time for such an acquisition would be T = TR × Nphase × Nslices = 5 milliseconds × 256 × 64 = 82 seconds. This obvious discrepancy between desirable (and clinically necessary) spatial resolutions and the corresponding scan times has led  to incorporation of many innovative rapid imaging approaches  to CE MRA, including partial Fourier acquisitions,5,27,34 parallel imaging,14,20,42,45 and others, many of which have been discussed in Chapter 3. Parallel imaging uses overlapping sensitivity regions from neighboring coils to undersample data, thus providing massive scantime savings that come at the cost of reduced SNR. However, CE MRA usually provides ample SNR so that these scan-time savings can be exploited and have proven to be very beneficial. In addition, k-space properties have been explored to further tailor acquisitions to the properties of CE MRA exams. This includes the use of elliptical centric view ordering,52 which is designed to capture the center of k-space during peak enhancement when the main contrast of the imaging scene is defined. The higher spatial frequencies that define the peripheral portions of k-space are acquired at time points where decreased arterial signal and potential venous enhancement has less of a detrimental effect on the MRA. State-of-the-art CE MRA sequences combine parallel imaging and high receiver counts with some of the other methods mentioned earlier to reduce scan time to within a breath hold or even shorter durations for dynamic imaging. Potential pitfalls of these CE MRA approaches include improper

72

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

C

B

FIG 4-4  CE MRA of the pelvis (A), thighs (B), and calves (C) in a single patient. The three stations are displayed as coronal maximum intensity projection (MIP) images and capture the bolus with high quality and high arterial signal. Some venous and renal parenchymal enhancement can be observed in the pelvic station, where the contrast agent is quickly passing through the arterial bed (A). (Courtesy Harald Kramer, MD, Dept. of Radiology, Ludwig-Maximilians University, Munich, Germany.)

timing when the acquisition is started too early (no contrast bolus yet) or too late (suboptimal arterial signal and venous overlap).

Time-Resolved CE MRA The goal of time-resolved CE MRA is to capture multiple time frames of the image volume before, during, and after the passage of the bolus, similar to dynamic series in conventional x-ray angiography, with the added benefit of a 3D acquisition. A great benefit of such an approach is to accurately capture the peak arterial window in cases where synchronization of bolus arrival and acquisition timing are difficult. Other benefits include the ability to visualize complex flow patterns that do not have a single optimal time frame, such as asynchronous enhancements in regions with poststenotic and collateral vascular filling, as well as signal enhancement in complex vascular pathologies such as large aneurysms, dissections, or AVMs. These acquisitions are necessarily performed in even shorter times than single-acquisition CE MRA, and thus spatial resolution must be further compromised or additional temporal constraints must be used in the reconstruction process to facilitate sufficiently high temporal resolution. Many innovative imaging approaches for rapid MRI have been  motivated by and were often first implemented in CE MRA, including keyhole imaging,49 time-resolved imaging of contrast kinetics (TRICKS),19 kt-acceleration,48 and compressed sensing.23 In keyhole imaging,49 the acquisition is separated into two phases. In the first phase, only a small central region of k-space is acquired repeatedly and rapidly to capture  the dynamic changes during contrast passage. In a later phase, all high spatial frequencies are acquired when the contrast is in the arteries and veins. Complete data sets are reconstructed for all dynamic phases by sharing the high spatial frequencies among all time frames. The assumption at work is that most of the signal evolution will be reflected only in the low spatial frequency components near the center of k-space. This approach can generate high temporal resolution but can hide dynamics or late enhancements in certain conditions.35 The 3D TRICKS approach19 and similar variants combines keyhole imaging with view sharing and variable rate sampling and partial Fourier imaging to overcome some of these pitfalls and are commonly used in clinical practice. Figure 4-5   shows an example for a dynamic CE MRA of a patient with an aortic dissection, illustrating the benefits of a data series when no single time frame is optimal to properly capture all details. Chapter 5 will further discuss dynamic imaging with CE MRA in the context of capturing  tracer kinetics.

NCE MRA NCE MRA techniques have been realized since the advent of the clinical use of MRI. There are several advantages and disadvantages when comparing NCE MRA with CE MRA. CE MRA usually provides higher signal and contrast and is considered more robust in many vascular territories. However, NCE MRA does not require injection of a contrast agent, thereby reducing costs and increasing patient safety, especially for patients with renal insufficiency at risk for NSF. In addition, scan times can be extended beyond an arterial window, thereby facilitating higher resolution scans or physiologic gating to the respiratory and/or cardiac cycle, which are usually time-prohibitive in CE MRA. As will be shown, PC MRI can also provide quantitative flow measurements for added value to the diagnostic process. Recent advances in MR hardware, such as multireceiver coil arrays and parallel imaging, as well as gradient performances that facilitate the use of bSSFP imaging in addition to the search for alternative MRA approaches for patients at risk for NSF,46,47 have renewed interest in NCE MRA techniques over the last 15 years.29,30,51 This, coupled with the development of more advanced gradient hardware and the introduction of higher-strength magnetic fields in clinical practice, has greatly increased the use of NCE MRA techniques clinically.29 This section will highlight the imaging principles behind the most common NCE MRA approaches: TOF, PC, and bSSFP MRA.

Time-of-Flight MRA TOF50 MRA is a commonly used technique, especially in the peripheral and intracranial vasculature. As in CE MRA, the pulse sequence employed is a spoiled gradient echo sequence, here with slightly longer TRs (<50 ms) as shown in Figure 4-6A. The repeated slice-selective RF pulses in a spoiled gradient echo sequence diminish signal from stationary tissue until equilibrium is reached from one excitation to the next. However, blood flowing into the imaging slice experiences no or very few of these RF pulses, which results in high blood signal relative to the saturated static tissue. Thus the contrast in TOF is dependent on the inflow of unsaturated (often called “fresh”) blood and the proper suppression of background tissues within the imaging slice. Consequently the slice or slab thickness as well as the speed of the blood flow play a crucial role for image contrast (see Fig. 4-6B). The slower the blood flow and the wider the imaging slab, the more  the contrast diminishes. Perpendicular orientation of the imaging slice 

CHAPTER 4  Imaging Principles in Magnetic Resonance Angiography

5

8

14

73

17

FIG 4-5  Selected time frames from a dynamic chest CE MRA acquired with the TRICKS method in a patient with an aortic dissection. Time frame 8 shows arrival of the contrast agent in the right side of the heart and slightly later in the pulmonary aorta (arrow in time frame 8). Early arterial enhancement of the aorta shows the true lumen of the aorta (short arrow in time frame 14). A later time frame (17) shows similar signal for both the true and the false lumen (upper right arrows) in the aorta. (Courtesy Chris Francois, MD, Dept. of Radiology, University of Wisconsin–Madison, WI.)

Presaturation pulse

Slow flow

α,Φ

Fast flow

RF/signal Stationary tissue imaging volume

Greadout

Gslice

Spoiler Sat selection

A

Sat spoiler

B

FIG 4-6  A, Pulse sequence diagram for time-of-flight (TOF) imaging. A typical spoiled gradient echo sequence is used to create bright signal from spins that flow into the imaging slab while the stationary tissue signal is in steady state and of low signal amplitude. A spatially selective presaturation pulse can be used to selectively suppress signal from vasculature entering the imaging slab from above or below. B, The inflow effect principle for TOF imaging. Fast-flowing blood that experiences fewer RF excitations generates higher signal in the imaging slice or volume as compared with slower-flowing blood. In a 3D acquisition, the signal decreases as the spins stay longer in the imaging volume.

to the flow direction generates the highest image contrast, whereas in-plane blood flow diminishes vessel contrast. As a result, TOF angiography is best suited for territories with mostly straight vessel paths. 2D TOF typically produces bright signal from both arterial and venous contributions as spins from these vessels enter the imaging slab. For a purely arterial or venous image, a saturation slab can be placed parallel to, above, or below the imaging slice to saturate the signal from arteries or veins as their blood enters the slice.8 Figure 4-7 illustrates the selective generation of an arterial and venous angiogram (no presat), an arterial angiogram (presat below the imaging slab), and a venous angiogram (presat above imaging slab) for an example in the neck. When imaging below the heart, the location of the presaturation slabs is inversed for the suppression of venous and arterial signal. TOF can be performed in both 2D and 3D acquisitions. 2D acquisitions benefit from a very high vascular signal, because the imaging slab

is thin and the inflow effect is pronounced. Large vascular territories can be covered with sequential 2D TOF acquisitions oriented perpendicular to the main vessel path. 3D acquisitions benefit from a higher SNR, but the vessel contrast decreases because the spins in the vessels have to travel longer in the thicker imaging slab and therefore experience more RF excitations, which reduces their signal. 3D acquisitions also allow for imaging with thinner slices and reduces the TE time, but a smaller tip angle must be used to avoid signal loss for vessels with slow blood flow. This again results in a lower signal amplitude. The superior vessel details from 2D TOF and the decreased noise level from 3D TOF have been combined to the 3D multislab technique MOTSA (multiple overlapping thin-slab acquisition). With MOTSA,39 thin overlapping slabs are acquired, which minimize saturation effects. In addition, magnetization-transfer is used to improve small-vessel contrast. Advantages of TOF include the short TE and imaging time.

74

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

Saturation

Saturation Vein

Artery

FIG 4-7  A, The use of saturation slabs in TOF MRA in an axial slice in the neck. The use of presaturation slabs can selectively visualize the arterial or the venous system. Top, With no saturation slab, venous and arterial inflow is observed for the carotid artery (red arrows) and the jugular vein (blue arrows). Middle, When placing a saturation slab superior to the imaging slice, the venous blood magnetization is saturated and these vessels are not visualized. Bottom, A saturation slab on the inferior side of the imaging volume will saturate arterial signal into the carotid and vertebral arteries.

A

B

FIG 4-8  A, Cranial MOTSA TOF image acquired with three overlapping slabs at 3T. The axial (A) and coronal (B) MIP images display excellent vessel conspicuity and detail. (Courtesy Howard Rowley, MD, Dept. of Radiology, University of Wisconsin–Madison, WI.)

Short TE times are particularly beneficial in the regions of complex blood flow. Disadvantages of this technique are the vessel saturation in large volume 3D acquisitions and in-plane saturation of spins.28 Regions of slow blood flow will produce no signal, and intravoxel dephasing in regions with disturbed flow will result in artifacts. Major clinical applications of TOF are in the cranial vasculature (Fig. 4-8), carotid artery, and peripheral vasculature, where straight vessel segments are prevalent and can be easily imaged.

Phase-Contrast MRA PC MRA uses bipolar gradient pairs to sensitize the acquisition to motion.31,33 Figure 4-9A illustrates the modification in the gradient echo sequence to incorporate velocity encoding along the sliceencoding axis. Application of the first half of the gradient causes a

change in spin phase. In stationary tissues, where no motion is present, the second half of the bipolar gradient inverses the phase shift caused by the first lobe and cancels out the accrued spin phase to a net phase of zero (see Fig. 4-9B [middle row]). Meanwhile, spins that move along the direction that the bipolar gradient is played out will accumulate a net phase that is directly proportional to the velocity of these spins (see Fig. 4-9B [bottom row]). The most common use of PC MRI is single-slice acquisition with through-plane velocity encoding. For this unidirectional flow assessment, velocity encoding has to be applied in the slice-encoding direction as shown in Figure 4-9A. In practice, there are many other sources that contribute to phase in an MRI, including eddy currents, measurement imperfections, magnetic susceptibilities, and others. To differentiate the velocity-induced phase contributions from these other

CHAPTER 4  Imaging Principles in Magnetic Resonance Angiography

75

BIPOLAR GRADIENT G

α t RF/Signal t Greadout Gslice

Spoiler

Static spins

ϕ =0

Moving spins

ϕ ~ν

Phase

Phase

t

Velocity encoding / Bipolar gradient

FIG 4-9  A, Phase-contrast (PC) MRA uses a bipolar gradient pair in a gradient echo sequence for a velocitysensitive acquisition. The bipolar gradient is inserted along the axis of motion encoding, here shown for the slice-encoding direction. B, Graphical representation of the effect of the bipolar gradient in static tissue and those with measurable velocity. Zero total phase is accrued in static spins, while in moving spins, the phase accrued during the bipolar gradients is proportional to their velocities in the encoded direction.

150

subcut. fat

100 50

AA

0

PA

90

Ascending aorta Descending aorta

40

-50

DA

-100

A

140 Flow (mL/s)

cm/s

B

-150

-10

C

40 210 380 550 710 880 050 1 Time in RR cycle (ms)

FIG 4-10  2D cine-PC MRA with through-plane velocity encoding for assessment of aortic flow. The data were acquired with prospective ECG gating during a 30-second breath hold. Images A and B show an endsystolic time frame. A, The magnitude images show details of anatomy including good vessel delineation of the ascending aorta (AA), the descending aorta (DA), and the pulmonary artery (PA). B, Phase images show voxel-by-voxel velocity maps with flows in opposing directions for the ascending (blue) and descending (red) aorta. Note zero velocity in static tissues and white noise in air. C, Instantaneous flow measurements over the cardiac cycle can be generated by integrating the measured velocities over the segmented vessel area.

sources, the acquisition of a reference image without velocity encoding is needed. This requirement effectively doubles the scan time. Once both acquisitions are completed, phase maps are calculated and the reference map is subtracted, so that the resulting velocity map contains only phase shifts that are related to velocity components. This subtraction comes at the expense of an increase in imaging noise by the factor of √2. In addition to the velocity map, the corresponding magnitude images are also generated for anatomic information. Usually the two magnitude images are averaged, with an effective gain in SNR of √2. Figure 4-10 shows a typical example of a 2D cine-PC MRI in the chest, where the acquisition was electrocardiographically (ECG) gated over multiple heart beats so that the resulting image series represents magnitude and velocity maps at various time points throughout the cardiac cycle. The magnitude (see Fig. 4-10A) and phase (see Fig. 4-10B) images represent an end-systolic time frame. The magnitude image displays the anatomic features, such as various vessels, lung (black because of air), muscle tissue, and subcutaneous fat, while the phase map displays the captured through-plane velocities on a voxelby-voxel basis. Blood flow toward the head is color coded in dark grayscale values, blood flow toward the feet in bright grayscale values,

and a midgrade tone, such as in the chest wall, representing no motion along the body axis. Instantaneous flow measures throughout the cardiac cycle can be obtained by integration of the measured velocities over the segmented vessel area. 2D PC MRI is frequently used in clinical practice—for example, for assessment of vascular flow in the chest and neck for measurements such as cardiac output and the ratio of pulmonary over systemic flow (Qp/Qs).12 An important aspect of PC MRA is the velocity-encoding (VENC) setting that is chosen prior to acquisition.22 The VENC setting is selected by the user as the highest velocity that can be uniquely encoded by the scan to properly sensitize the acquisition to the expected velocities in the interrogated vascular region. The lower the expected velocities are, the larger the area under the bipolar gradients is adjusted by the sequence. The velocity-to-noise ratio (VNR) is inversely proportional to the VENC. In other words, the lower the VENC is, the higher the VNR is in the velocity map. However, if the VENC is chosen too low, velocity aliasing occurs and the measured velocity can become erroneous. Because a measure of phase is inherently a cyclical entity, the dynamic range is limited to [0 180 degrees] for positive velocities and [−180 degrees 0] for negative velocities. If the VENC is chosen too

76

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C

FIG 4-11  Cranial 3D PC MRA of a patient with a dural arteriovenous fistula. The coronal MIP image (A) shows the presence of arteries and veins in the angiogram generated by the combination of the corresponding magnitude and phase images from 3-directional velocity encoding. Volume-rendered images (B and C) demonstrate the pathology from various angles, with good background suppression. (Courtesy Thorsten Bley, MD, Dept. of Radiology, University of Würzburg, Germany.)

low, phase wrapping occurs and different velocities will generate the same phase value; for example, the corresponding velocity measures for −90 degrees and +270 degrees as well as −450 degrees will all be mapped to the same phase value of −90 degrees on the unity circle. Therefore the VENC setting has to be carefully chosen to the expected velocity range, and in some instances repeated PC MR scans are needed for proper selection. Velocity-encoding gradients can also be added into one, two, or three spatial directions to generate (1) through-plane velocity maps as already discussed, or (2) in-plane velocity encoding, or (3) the generation of velocity vector maps with all three spatial directions. Any of these measures will need an independent acquisition with bipolar gradients played out along the desired spatial direction. In combination with the need for a single phase reference scan, the total scan duration requires two, three, or four acquisitions for 1-, 2-, and 3-directional velocity encoding, respectively. PC MR can also be extended to 3D acquisition, either by sequential 2D scanning or direct 3D encoding with the benefits of added SNR and thinner slices. 3D acquisitions with 3-directional velocity encoding are well suited to generate PC MRAs for the display of vascular anatomy in addition to the velocity maps.6 This is accomplished by the nonlinear combination of the phase map and the magnitude image averaged from the four acquisitions based on complex difference processing. The corresponding angiograms are typically characterized by excellent background suppression because the velocities of stationary tissues appear with a phase of zero. An example of such a PC MRA of the head is shown in Figure 4-11, where the presence of arteries and veins can be appreciated. In comparison to other MRA techniques, 3D PC MRA requires long scan times because of the repeated scans for the velocityencoding directions and the phase baseline scan, as well as the prolonged TRs resulting from the addition of bipolar gradients. Although these acquisitions also provide quantitative information on blood flow, their clinical use is limited because of the long scan durations. The combination of 3D PC MR acquisitions with cardiac gating was recognized early on as a powerful approach for the comprehensive assessment of vascular anatomy and hemodynamic function.9 However, the demands on extensively long scan times, adequate computational processing and visualization power for volumetric dynamic velocity vector fields, and lack of proper postprocessing algorithms have slowed the adaptation of this approach into mainstream applications. Recent

advances on these fronts have facilitated great research interest on the use of what is now frequently referred to as 4D flow MRI in various vascular territories.24 Similar to the techniques discussed earlier, scan times have been reduced to clinically acceptable durations by advances in hardware through faster gradients and improved coil technology, as well as accelerated approaches such as parallel imaging, view sharing, and non-Cartesian trajectories. 4D flow MRI can be used for velocity and flow measurements in various vessels and vessel segments after scan completion (Fig. 4-12), as well as the derivation of additional hemodynamic parameters such as local pulsatility index, pressure gradients, wall shear stress, pulse wave velocity, kinetic energy, and others.24,25

Balanced Steady-State Free Precession MRA In steady-state free precession (SSFP) pulse sequences, the phase coherence of the transverse magnetization is maintained from repetition to repetition, and both the transverse and the longitudinal magnetizations reach a steady state. In contrast the spoiled gradient echo sequence maintains a steady state by spoiling components of the magnetization vector, thereby somewhat reducing the signal. Although this phenomenon was first reported in 19562 and imaging sequences were proposed in the 1980s,37 its clinical use was delayed until recently because of severe banding artifacts from inhomogeneities in the static field. This imaging approach, also known as true fast imaging with steady-state precession (trueFISP), fast imaging employing steady-state acquisition (FIESTA), and balanced fast field echo (bFFE) has greatly benefitted from recent hardware improvements, particularly in gradient performance and field shimming to minimize image-degrading artifacts. A simple pulse sequence diagram for this sequence is shown in Figure 4-13. It is designed such that all gradients are fully refocused (or balanced) over one repetition interval and the spins are completely rephased. It can be shown that typical MRA acquisitions with the bSSFP sequence produce a T2/T1-weighted image contrast independent of the TR.21 This property explains its extensive use in the evaluation of cardiac function, with striking contrast between the blood pool (high T2/T1 ratio) and the myocardium (low T2/T1 ratio).3 Although arterial blood generates a high signal in bSSFP imaging, so do other tissues with a high T2/T1 ratio, like fat as well as other fluids such as CSF and venous blood. The major tasks for successful use of this sequence in MRA are (1) an implementation that is robust

CHAPTER 4  Imaging Principles in Magnetic Resonance Angiography Source 4D flow MR images

77

Anatomical (complex difference)

Vx

Magnitude

60cm/s

Vy

Vz

60cm/s

Visualization and quantification

Velocity visualization

Mask segmentation

corV SV

MPV

30 Velocity [cm/s] SMV

IMV

0

FIG 4-12  Abdominal 4D flow MRI of a 59-year-old man with portal hypertension. The workflow starts with source magnitude and velocity images (upper left, axial plane), which are combined into an anatomic PC angiogram (PCA) using complex difference processing (upper right, surface shaded display). The PCA is segmented into color-coded vascular territories (bottom left, surface shaded display), followed by streamline or particle trace visualization (bottom right). Note the reversed blood flow in the coronary vein and inferior mesenteric vein, which would be very hard and/or time consuming to identify with ultrasound or repeated 2D PC measurements.

α RF/signal Greadout Gslice Area

+ Area

=0

Area

+ Area

=0

FIG 4-13  Basic pulse sequence for balanced steady-state free precession (bSSFP) imaging. Gradients along all directions are balanced such that the total area of the gradient waveforms is equal to zero at the end of each TR. Phase coherence of the transverse magnetization is maintained from repetition to repetition, and both the transverse and the longitudinal magnetizations reach a steady state.

to off-resonance artifacts and (2) suppression of unwanted signal from sources other than arterial blood. Imperfect shimming and magnetic susceptibility cause off-resonances across the imaging plane or volume and can lead to signal voids, often referred to as banding artifacts. A short TR (<5 milliseconds) is required to reduce the sensitivity of

bSSFP imaging to these off-resonance artifacts, and this requirement becomes more stringent at higher field strengths. Various schemes have been applied to the suppression of unwanted signal contributions, especially in regard to fat. The use of selective saturation pulses for fat, selective excitation for protons without contributions from fat spins through RF excitation or so-called phase cycling of the bSSFP acquisition, selective excitation of arterial blood entering the imaging volume (arterial spin labeling), and saturation slabs over regions where venous bloods enters the imaging slab are examples of such signal manipulations for increased conspicuity of the arterial signal.29,30,51 Figure 4-14 shows an example of bSSFP imaging in a patient suspected of having a stenosis in a renal transplant.

CONSIDERATION OF PHYSIOLOGIC MOTION MR acquisitions including volumetric MRA are too long to be conducted in a real-time fashion similar to ultrasound or conventional x-ray angiography. Therefore the effects of and compensation for physiologic motion, particularly respiratory and cardiac, have to be considered.

Respiratory Motion The average duration of the respiratory cycle is on the order of 3 seconds but varies greatly among patients and from one breathing

78

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging CE MRA

bSSFP + ASL

DSA

FIG 4-14  Example of a bSSFP MRA in a patient suspected of having renal transplant stenosis. Corresponding CE MRA and DSA images are shown along with the bSSFP images acquired with arterial spin labeling (ASL) for venous suppression, indicating good agreement among those modalities in this case. Measurements of the transstenotic pressure gradient (TSPG) conducted under DSA guidance were indicative of treatment with angioplasty in the right common iliac artery (small arrow) (TSPG > 10 mm Hg) and no treatment of the renal transplant artery (TSPG < 10 mm Hg). (Courtesy Thorsten Bley, MD, Dept. of Radiology, University of Würzburg, Germany.)

cycle to the next. In the abdomen and thorax, uncompensated respiratory motion can cause severe image degradation. The two common approaches for minimizing those artifacts are to (1) reduce the imaging time to a single breath hold or (2) gate the acquisition during free breathing and limit the data sampling to the end-expiration plateau, which is the most quiescent period in the respiratory cycle. For respiratory gating, a stress transducer system in the form of respiratory bellows can be used as an external measure of the breathing cycle. However, the belt is worn around the patient’s abdomen, tracking abdominal circumferential rather than the targeted image area. Precision becomes a concern because these bellows can undergo baseline drifts or be undetectable in shallow-breathing individuals. Nonetheless this simple approach is often used to monitor whether a patient is successfully performing a breath hold and is commonly used for prospective respiratory gating in longer acquisitions. An alternative to system bellows is the tracking of the lung-liver tissue interface with imaging using navigator echoes.10 Although this technique is more precise than bellows, it is also more complex, and with sequences that require fast scanning can interrupt necessary steady signal states. The addition of navigator excitations to pulse sequences can increase the TR, resulting in increased scan time compared with traditional bellows respiratory gating. Typical gating efficiencies are 40% to 50% to compensate for the most dominant respiratory motion artifacts.

Cardiac Motion and Pulsatility The rhythmic beating of the heart introduces significant motion of the myocardium and directly attached vessels. It also creates pulsatile flow, which causes variations of velocities and inflow effects throughout the cardiac cycle, as well as vessel diameter changes in the arterial system. The most common approaches for synchronization of the MR data acquisition with the cardiac cycle are the use of an ECG or a peripheral plethysmograph waveform from a pulse oximeter sensor, which also varies throughout the cardiac cycle. The most precise method for cardiac gating is through R wave detection from the ECG signal by attaching MR-compatible sensors on the patient’s chest. However, the ECG signal may be compromised within the scanner bore, where magnetohydrodynamic effects manifest themselves as enlarged T waves, potentially causing mistaken trigger points or an incorrectly rapid heart rate. Also, the time-varying imaging gradients can cause noise in the ECG leads. These issues can be ameliorated with more advanced

hardware/software, such as processing multiple leads simultaneous with vector gating systems or the use of fiberoptic ECG cables for extended connections. Careful placement of ECG leads and shielding from gradient switching and RF transmission can also improve ECG signal. Alternatively, finger probes or reflectance probes implementing optically based arterial blood oxygen saturation monitors allow for a simple and reliable setup. In the case of subjects with poor circulation or weak pulses, issues can arise. This leads to a less distinct peak in the cardiac cycle, making it unsuitable for acquisitions requiring very precise synchronization. Although the setup for this approach is much simpler than the patient preparation for ECG gating, it also leads to shifted waveforms for systole and diastole because the peripheral plethysmograph signal is significantly delayed to the cardiac activity.

CLINICAL APPLICATIONS The following section will discuss the use of the various MRA approaches by vascular region.

Head/Neck TOF MRA is very commonly used for cranial MRA and as shown in Figure 4-8, high-quality angiograms can be obtained. The prolonged scan times over bolus chase acquisitions allow for high spatial resolutions, and the vessel alignment and high blood velocities are usually well suited for substantial inflow effects. Imaging at 3T further improves image quality in those scans because of the inherent SNR gains and improved contrast from T1 lengthening of the background tissues. In patients with carotid disease, 2D, 3D, and MOTSA TOF have been widely used. However, with the relatively long scan times, patient motion between TOF slices/slabs can become a problem. Additionally, for patients in whom arteries fill in a retrograde fashion owing to occlusions, signal dropout may occur from the distal saturation slab. Slow flow within the imaging slice will also cause unwanted signal dropout and may be mistaken as stenosis or occlusions. In the assessment of stenosis, TOF MRA is known to overestimate the degree of stenosis because of partial volume effects. CE MRA is the method of choice in the management of stroke patients, where dynamic data are beneficial and coverage from the aortic arch to the brain is desired. It is also commonly used in patients

CHAPTER 4  Imaging Principles in Magnetic Resonance Angiography

3D PCA

79

3D TOF

FIG 4-15  3D PC MRA and TOF MRA in a patient with arteriovenous malformation. The visibility of the arteries and veins makes 3D PC MRA a very suitable option for assessing cranial vascular malformations. (Courtesy Patrick Turski, MD, Dept. of Radiology, University of Wisconsin–Madison, WI.)

with carotid artery disease. PC MRA (Fig. 4-15) is used to assess veins, vascular malformations, and to compare flow measures before and after treatment.

Chest CE MRA is the preferred method for diagnostic tasks in the chest that include assessing the pulmonary arteries, aortic dissections, or aneurysms in the descending aorta. Often, dynamic acquisitions are preferred, as shown in Figure 4-5. Diagnoses that involve the aortic root, such as in congenital heart disease, the ascending aorta, or aneurysms, greatly benefit from cardiac gating to minimize motion artifacts from the vessel displacement throughout the cardiac cycle. This makes bSSFP sequences the method of choice. As it is a very rapid acquisition scheme, it can be gated to the cardiac cycle without excessive scan times. Imaging of the coronary arteries remains a goal for MRA, but image quality and robustness is inferior to state-of-the-art multidetector computed tomography angiography (CTA). The most common approach for coronary MRA uses gated bSSFP acquisitions as well.

Abdomen In abdominal MRA, various sequences are used. CE MRA is challenging because of the respiratory motion and the timing issues for coordinating bolus arrival and breath-hold initiation. Accelerated imaging approaches have facilitated the use of CE MRA in these challenging areas, and dedicated contrast agents are becoming popular for liver imaging.38 bSSFP-based imaging has also made great progress and can provide better background suppression for the renal parenchyma, as shown in Figure 4-14. TOF has been used for abdominal imaging as well, but robustness of image quality and scan times have been a concern. The high pulsatility of arterial flow in the abdomen can lead to substantial signal fluctuations in TOF acquisitions that rely heavily on inflow effects. Therefore abdominal TOF acquisitions need to be gated to the cardiac cycle to minimize these signal fluctuations. Data acquisition occurs in systole when the inflow effect is greatest. Image quality in TOF acquisitions outside the chest and abdomen would also benefit from cardiac gating for improved data consistency but is often ignored in practice for the sake of reduced scan times.

Peripheral Arteries MR assessment of peripheral artery disease is usually based on CE MRA (see Fig. 4-4) or 2D sequential TOF. CE MRA exams are either

conducted as a bolus chase with a single injection and rapid scanning of three stations to capture the peripheral arterial tree from the pelvis to the feet. Alternatively, three subsequent injections are used for three scans, which can lead to Gd accumulation in background tissues, thereby reducing the imaging contrast. TOF is a frequently used alternative but requires long scan times for the sequential 2D slice acquisitions along the length of the legs.

REFERENCES 1. Brown MA, Semelka RC: MR imaging abbreviations, definitions, and descriptions: A review. Radiology 213(3):647–662, 1999. 2. Carr HY: Steady-state free precession in nuclear magnetic resonance. Phys Rev 112:1693–1701, 1956. 3. Carr JC, Simonetti O, Bundy J, et al: Cine MR angiography of the heart with segmented true fast imaging with steady-state precession. Radiology 219(3):828–834, 2001. 4. Daftari Besheli L, Aran S, Shaqdan K, et al: Current status of nephrogenic systemic fibrosis. Clin Radiol 69(7):661–668, 2014. 5. Du YP, Parker DL, Davis WL, et al: Reduction of partial-volume artifacts with zero-filled interpolation in three-dimensional MR angiography.   J Magn Reson Imaging 4(5):733–741, 1994. 6. Dumoulin CL, Souza SP, Walker MF, et al: Three-dimensional phase contrast angiography. Magn Reson Med 9(1):139–149, 1989. 7. Earls JP, Rofsky NM, DeCorato DR, et al: Breath-hold single-dose gadolinium-enhanced three-dimensional MR aortography: Usefulness of a timing examination and MR power injector. Radiology 201(3):705–710, 1996. 8. Felmlee JP, Ehman RL: Spatial presaturation: A method for suppressing flow artifacts and improving depiction of vascular anatomy in MR imaging. Radiology 164(2):559–564, 1987. 9. Firmin DN, Gatehouse PD, Konrad JP, et al: Rapid 7-dimensional imaging of pulsatile flow. Paper presented at: Computer in Cardiology IEEE Computer Society, London, UK, 1993. 10. Firmin D, Keegan J: Navigator echoes in cardiac magnetic resonance.   J Cardiovasc Magn Reson 3(3):183–193, 2001. 11. Foo TK, Saranathan M, Prince MR, et al: Automated detection of bolus arrival and initiation of data acquisition in fast, three-dimensional, gadolinium-enhanced MR angiography. Radiology 203(1):275–280, 1997. 12. Gatehouse PD, Keegan J, Crowe LA, et al: Applications of phase-contrast flow and velocity imaging in cardiovascular MRI. Eur Radiol 15(10):2172–2184, 2005. 13. Glover GH: Gradient echo imaging. In Bronskill MJ, Sprawls P, editors: The Physics of MRI—1992 AAPM Summer School Proceedings, Woodbury, NY, 1992, American Institute of Physics, pp 188–205.

80

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

14. Griswold MA, Jakob PM, Heidemann RM, et al: Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47(6):1202–1210, 2002. 15. Guglielmo FF, Mitchell DG, Gupta S: Gadolinium contrast agent selection and optimal use for body MR imaging. Radiol Clin North Am 52(4):637–656, 2014. 16. Haase A, Frahm J, Matthaei D, et al: Flash imaging—Rapid NMR imaging using low flip-angle pulses. J Magn Reson 67(2):258–266, 1986. 17. Hao D, Ai T, Goerner F, et al: MRI contrast agents: Basic chemistry and safety. J Magn Reson Imaging 36(5):1060–1071, 2012. 18. Hartung MP, Grist TM, Francois CJ: Magnetic resonance angiography: Current status and future directions. J Cardiovasc Magn Reson 13:19, 2011. 19. Korosec FR, Frayne R, Grist TM, et al: Time-resolved contrast-enhanced 3D MR angiography. Magn Reson Med 36(3):345–351, 1996. 20. Larkman DJ, Nunes RG: Parallel magnetic resonance imaging. Phys Med Biol 52(7):R15–R55, 2007. 21. Liang Z-P, Lauterbur PC: Principles of Magnetic Resonance Imaging—A Signal Processing Perspective, New York, 2000, IEEE Press. 22. Lotz J, Meier C, Leppert A, et al: Cardiovascular flow measurement with phase-contrast MR imaging: Basic facts and implementation. Radiographics 22(3):651–671, 2002. 23. Lustig M, Donoho D, Pauly JM: Sparse MRI: The application of compressed sensing for rapid MR imaging. Magn Reson Med 58(6):1182–1195, 2007. 24. Markl M, Frydrychowicz A, Kozerke S, et al: 4D flow MRI. J Magn Reson Imaging 36(5):1015–1036, 2012. 25. Markl M, Kilner PJ, Ebbers T: Comprehensive 4D velocity mapping of the heart and great vessels by cardiovascular magnetic resonance.   J Cardiovasc Magn Reson 13:7, 2011. 26. Markl M, Leupold J: Gradient echo imaging. J Magn Reson Imaging 35(6):1274–1289, 2012. 27. McGibney G, Smith MR, Nichols ST, et al: Quantitative evaluation of several partial Fourier reconstruction algorithms used in MRI. Magn Reson Med 30(1):51–59, 1993. 28. Mistretta CA: Relative characteristics of MR angiography and competing vascular imaging modalities. J Magn Reson Imaging 3(5):685–698, 1993. 29. Miyazaki M, Akahane M: Non-contrast enhanced MR angiography: Established techniques. J Magn Reson Imaging 35(1):1–19, 2012. 30. Miyazaki M, Lee VS: Nonenhanced MR angiography. Radiology 248(1):20–43, 2008. 31. Moran PR: A flow velocity zeugmatographic interlace for NMR imaging in humans. Magn Reson Imaging 1(4):197–203, 1982. 32. Mugler JP, 3rd, Brookeman JR: Three-dimensional magnetizationprepared rapid gradient-echo imaging (3D MP RAGE). Magn Reson Med 15(1):152–157, 1990. 33. Nayler GL, Firmin DN, Longmore DB: Blood flow imaging by cine magnetic resonance. J Comput Assist Tomogr 10(5):715–722, 1986. 34. Noll DC, Nishimura DG, Macovski A: Homodyne detection in magnetic resonance imaging. IEEE Trans Med Imaging 10(2):154–163, 1991. 35. Oesterle C, Strohschein R, Kohler M, et al: Benefits and pitfalls of keyhole imaging, especially in first-pass perfusion studies. J Magn Reson Imaging 11(3):312–323, 2000.

36. Omary RA, Henseler KP, Unal O, et al: Comparison of intraarterial and IV gadolinium-enhanced MR angiography with digital subtraction angiography for the detection of renal artery stenosis in pigs. AJR Am J Roentgenol 178(1):119–123, 2002. 37. Oppelt A, Graumann R, Barfus H, et al: FISP—A new fast MRI sequence. Electromedica 54:15–18, 1986. 38. Palmucci S: Focal liver lesions detection and characterization: The advantages of gadoxetic acid-enhanced liver MRI. World J Hepatol 6(7):477–485, 2014. 39. Parker DL, Yuan C, Blatter DD: MR angiography by multiple thin slab 3D acquisition. Magn Reson Med 17(2):434–451, 1991. 40. Prince MR, Grist TM, Debatin JF: 3D Contrast MR Angiography, 2002, Springer. 41. Prince MR, Yucel EK, Kaufman JA, et al: Dynamic gadolinium-enhanced three-dimensional abdominal MR arteriography. J Magn Reson Imaging 3(6):877–881, 1993. 42. Pruessmann KP, Weiger M, Scheidegger MB, et al: SENSE: Sensitivity encoding for fast MRI. Magn Reson Med 42(5):952–962, 1999. 43. Riederer SJ, Tasciyan T, Farzaneh F, et al: MR fluoroscopy: Technical feasibility. Magn Reson Med 8(1):1–15, 1988. 44. Scheffler K: A pictorial description of steady-states in rapid magnetic resonance imaging. Concepts Magn Reson 11(5):291–304, 1999. 45. Sodickson DK, Manning WJ: Simultaneous acquisition of spatial harmonics (SMASH): Fast imaging with radiofrequency coil arrays. Magn Reson Med 38(4):591–603, 1997. 46. Thomsen HS: Nephrogenic systemic fibrosis: A serious late adverse reaction to gadodiamide. Eur Radiol 16(12):2619–2621, 2006. 47. Thomsen HS, Morcos SK, Dawson P: Is there a causal relation between the administration of gadolinium based contrast media and the development of nephrogenic systemic fibrosis (NSF)? Clin Radiol 61(11):905–906, 2006. 48. Tsao J, Kozerke S, Boesiger P, et al: Optimizing spatiotemporal sampling for k-t BLAST and k-t SENSE: Application to high-resolution real-time cardiac steady-state free precession. Magn Reson Med 53(6):1372–1382, 2005. 49. van Vaals JJ, Brummer ME, Dixon WT, et al: “Keyhole” method for accelerating imaging of contrast agent uptake. J Magn Reson Imaging 3(4):671–675, 1993. 50. Wehrli FW, Shimakawa A, Gullberg GT, et al: Time-of-flight MR flow imaging: Selective saturation recovery with gradient refocusing. Radiology 160(3):781–785, 1986. 51. Wheaton AJ, Miyazaki M: Non-contrast enhanced MR angiography: Physical principles. J Magn Reson Imaging 36(2):286–304, 2012. 52. Wilman AH, Riederer SJ: Performance of an elliptical centric view order for signal enhancement and motion artifact suppression in breath-hold three-dimensional gradient echo imaging. Magn Reson Med 38(5):793– 802, 1997. 53. Zur Y, Stokar S, Bendel P: An analysis of fast imaging sequences with steady-state transverse magnetization refocusing. Magn Reson Med 6(2):175–193, 1988.

5  Contrast-Enhanced Magnetic Resonance Imaging Junyu Guo and Wilburn E. Reddick

INTRODUCTION Contrast agents used in magnetic resonance imaging (MRI) can be either endogenous or exogenous. The most common type of endogenous contrast-enhanced MRI is arterial spin labeling (ASL) techniques,27,28,134 which use a dynamic labeling of internal water to measure brain perfusion. ASL is noninvasive, and the sequences needed to acquire this type of imaging are included on most major vendors’ platforms. However, ASL still suffers from poor signal-tonoise ratios (SNRs) and limited spatial coverage. Exogenous contrast agents for MRI, first introduced in the early 1980s, can substantially increase image contrast to facilitate diagnosis of lesions in both brain and body imaging.41,93,104 Beyond simply increasing the conspicuity of lesions, exogenous contrast agents are also commonly used to measure tissue properties such as blood perfusion and permeability.23,119,126 Approximately 30 million MRI examinations are performed in the United States annually, and exogenous contrast agents are used to improve image quality in about 30% of cases. Exogenous contrast agents for MRI are administered either by venous injection or orally and usually include gadolinium (Gd)-, iron-, or manganese-based compounds,78,98,104 with Gd-based contrast agents being the most common. The applications of iron- and manganese-based agents are still in early-stage development. In this chapter, we focus on Gd-based exogenous contrast agents and their applications. Gadolinium (Gd, atomic number 64) is a paramagnetic heavy metal. Although the Gd(III) ion is quite toxic to humans, chelated-  Gd(III) compounds are much less toxic and are referred to as Gd-based contrast agents (GBCA). GBCAs are manufactured by a chelating process in which large organic molecules encapsulate the gadolinium. This procedure reduces the chances of toxicity due to free Gd ions because these stable chelated compounds are predominantly eliminated via the kidneys before the compounds degrade and release  free Gd ions. The U.S. Food and Drug Administration (FDA) first approved a GBCA for use in MRI in 1988.37 Since then, eight additional GBCAs have been approved for use in clinical studies.37 FDA-approved GBCAs are as follows: Ablavar (gadofosveset trisodium), Dotarem (gadoterate meglumine, Gd-DTOA), Eovist (gadoxetate disodium,  Gd-EOB-DTPA), Gadavist (gadobutrol, Gd-DO3A-Butriol), Magnevist (gadopentetate dimeglumine, Gd-DTPA), Multihance (gadobenate dimeglumine, Gd-BOPTA), Omniscan (gadodiamide, Gd-DTPABMA), Optimark (gadoversetamide, Gd-DTPA-BMEA), and Prohance (gadoteridol, Gd-DO3A-HP).37 GBCAs can be divided into two groups: ionic and nonionic agents. Ionic contrast agents typically, but not always, have higher osmolality and more side effects. Nonionic contrast agents have lower osmolality and tend to have fewer side effects. The properties of GBCAs are summarized in Table 5-1.

Although GBCAs are considered to be relatively safe in clinical studies, they are not completely without risks. Recently, associations between nephrogenic systemic fibrosis (NSF) and GBCAs have been reported in a small subset of patients.55,105,116 NSF may result in fatal or debilitating fibrosis. Patients with acute, chronic, and severe kidney disease (glomerular filtration rate [GFR] < 30 mL/min/1.73 m2) have very poor elimination of GBCAs and could have an increased risk of developing NSF due to GBCAs. Because GBCAs increase the risk of NSF, patients with poor elimination should not be administered such contrast agents. Patients in MRI examination with GBCAs should be properly screened and be closely observed during the procedure. Upon any reaction to the drugs, the appropriate therapy must be used, and any case of NSF should be reported to the FDA (1-800-FDA-1088 or www.fda.gov/medwatch). Because GBCAs are paramagnetic, they generate a magnetic moment when placed within a static magnetic field, influencing water protons within their local magnetic field and shortening T1, T2, and T2* relaxation times to enhance image contrast. T1 is the spin-lattice or longitudinal relaxation time, which measures how fast the proton magnetization recovers to its equilibrium position. Shortening T1 increases the recovery speed and further increases MR signal in the tissue. T2 is the spin-spin relaxation time, which represents how fast MR transverse magnetization disappears. T2 shortening increases the rate at which the transverse magnetization disappears and further decreases MR signal in the tissue. T2* is the transverse relaxation time including T2 and magnetic field inhomogeneity effects. Although T2* has an effect on MR signal that is similar to that of T2, T2*’s effect is usually much smaller than that of T2. In MR imaging, spin echo sequences generate T2-weighted images and gradient echo sequences usually generate T2*-weighted images. GBCAs do not cross an intact blood-brain barrier (BBB) and therefore do not accumulate in normal brain. However, disruption of the BBB allows accumulation of GBCAs in lesions such as neoplasms. Within the body, there is no barrier to hinder extravasation of the smaller GBCAs into the surrounding tissue. The rate and degree to which this extravasation occurs are highly elevated in many neoplasms and can be used to accentuate the conspicuity of the lesions. In this chapter, we will describe the measurement of perfusion and permeability in tissues performed by using dynamic contrast-enhanced MRI (DCE-MRI) and dynamic susceptibility contrast MRI (DSC-MRI).

DCE-MRI AND DSC-MRI Dynamic MRI with GBCA can be used to measure two basic physiologic properties: perfusion and capillary permeability.23,93,104,126 Perfusion is a faster process and requires a faster acquisition with high

81

82

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

TABLE 5-1  FDA-Approved Gadolinium-Based Contrast Agents (GBCAs)

Indication

Children/ Adults

Elimination Half-Life (hours)

Elimination Percentage (%)

0.12

MRA

A

0.48/16.3a

78.7c

2008

Ionic

0.2

CNS

C/A

1.4 in females, 2.0 in malesb

72.9 in females, 85.4 in malesd

2013

Ionic

0.1

Liver

A

0.9b

None

2008

Nonionic

0.1

CNS

C/A

1.81b

>90e

2011

Ionic

0.2

CNS/ body

C/A

0.2/1.6a

91e

1988

Ionic

0.2

CNS/ MRA

C/A

2.02 b

>80e

2004

Nonionic

0.2

CNS/ body

C/A

0.062/1.30a

95e

1993

Nonionic

0.2

CNS/ Liver

C/A

0.23/1.73a

95e

1999

Nonionic

0.2

CNS

C/A

0.2/1.57a

94e

1992

Drug Name

Active Ingredients

Nonionic/ Ionic

Ablavar

Gadofosveset trisodium Gadoterate meglumine (Gd-DTOA) Gadoxetate disodium (Gd-EOB-DTPA) Gadobutrol (Gd-DO3A-butriol) Gadopentetate dimeglumine (Gd-DTPA) Gadobenate dimeglumine (Gd-BOPTA) Gadodiamide (Gd-DTPA-BMA) Gadoversetamide (Gd-DTPA-BMEA) Gadoteridol (Gd-DO3A-HP)

Ionic

Dotarem

Eovist

Gadavist Magnevist

MultiHance

Omniscan Optimark Prohance

Recommended Dose (mL/kg)

Year Approved

CNS, central nervous system; MRA, magnetic resonance angiography. a Using the two-compartment model; busing the one-compartment model; celimination percentage within 72 hours; delimination percentage within 48 hours; eelimination percentage within 24 hours.

temporal resolution to capture the first pass of the bolus of contrast agent while it is still intravascular. DSC-MRI is mostly used to measure perfusion with high temporal and low spatial resolution. Permeability is usually a relatively slower process and requires a longer acquisition time to characterize the slow transfer of GBCA from capillaries to the interstitial space. DCE-MRI is generally used to measure permeability using higher spatial resolution and a longer acquisition time. DSC-MRI acquires a series of T2*-weighted or T2-weighted dynamic MR images before, during, and immediately after a bolus injection of a GBCA. DSC-MRI is widely used for perfusion measurements in brain imaging, which generates a larger signal change with the same concentration of GBCA than does DCE-MRI; this is because of the higher T2* relaxivity of GBCA.59 DSC-MRI can be routinely acquired as a relatively low spatial resolution volume data set with a temporal resolution of approximately 1 second68 and an acquisition time of approximately 60 seconds, making DSC-MRI ideal for the study of first-pass kinetics of the GBCA bolus. However, it is difficult for DSC-MRI to measure permeability because the decrease of signal due to T2* effects counteracts the increase in signal caused by T1 effects during the leakage and thus confounds the signal change.8,112 DCE-MRI acquires a series of T1-weighted dynamic MR images before, during, and after a bolus injection of a GBCA. T2* effects can be minimized by using a short echo time. The T1 relaxation rate has a linear relationship with drug concentration, and the T1 relaxivity is largely independent of tissue type.29,94,103 When first introduced, DCE-MRI had a relatively low temporal resolution (>1 minute per image) because of acquisition sequence design, hardware limitations, and the requirement for higher spatial resolution.65,120 However, because of improvements in the acquisition sequences and hardware improvements, DCE-MRI can now be acquired as a volume data set

with a temporal resolution of approximately 3 seconds.68 DCE-MRI can be used to measure perfusion and permeability simultaneously but has to have both a high temporal resolution and a sufficiently long acquisition time of approximately 8 to 10 minutes.68,113 DCE-MRI measures of perfusion have a lower contrast-to-noise ratio (CNR) than DSC-MRI does because of its lower T1 relaxivity in comparison with the high T2* relaxivity in DSC-MRI. This drawback can be largely mitigated when imaging highly vascularized cerebral tumors, where high drug concentrations can be reached.10,48

DCE-MRI A complete DCE-MRI study usually includes baseline T1 mapping, dynamic data acquisition, arterial input function measurement, and dynamic data analysis. In the following sections, we will discuss each part of the whole process. DCE-MRI images can be analyzed by using a heuristic method or a suitable tracer-kinetic model to get the proper physiologic metrics related to perfusion and permeability. For convenience, all physiologic parameters in this chapter are summarized in Table 5-2 for later reference.

Baseline T10 Measurement The baseline longitudinal relaxation time T10, before drug injection, is required for pharmacokinetic modeling analysis. Several methods can be used to measure T1. The variable flip angle (VFA) method is one of most widely used methods, calculating a T1 map by using multiple spoiled gradient echo acquisitions with different flip angles.9,22,42 The VFA method allows rapid high-resolution threedimensional (3D) acquisitions because of the modern implementation with very short repetition time.22 Radio frequency (RF) field (B1)

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging

83

TABLE 5-2  Summary of Physiologic Parameters Modeled in DCE- and DSC-MRI Category

Parameters

Normal Range

Unit

Interpretation

Perfusion

Fp vp

20-400 1-30

mL/min/mL %

Mixed

Ktrans kep PS

E × Fp Ktrans/ve 0.01-0.1

1/min 1/min mL/min/100 mL

ve

20-30

%

E τ

∝PS < 10-20 <5

% Second

Plasma flow Plasma volume <5% for brain Volume transfer constant Interstitium-to-plasma rate constant Permeability–surface area product for normal muscle Interstitial volume fraction (extracellular extravascular space [EES] fraction) Extraction fraction Mean intracellular water molecule lifetime

Permeability

DCE, dynamic contrast-enhanced; DSC, dynamic susceptibility contrast.

TABLE 5-3  Summary of Major Parameters for T1 Mapping Protocols Parameters

Variable Flip Angle

Inversion Recovery

B0 Sequence Vendors Sequence name Time of echo (TE) Time of repetition (TR) Flip angles Time of inversion (TI) Slice number Slice thickness Acquisition matrix Average number Partial Fourier Time/measures Total time

1.5 T 3D spoiled gradient echo GE Phillips SPGR T1-FFE 1.0 ms 5.0 ms 2°, 5°, 10°, 15°, 20°, 25°, 30° None 16 5 mm 256 × 160 × 16 4 None 0.85 minutes 5.97 minutes

1.5 T 3D single-shot fast spin echo GE Phillips Single-shot FSE Single-shot TSE 78 ms 4000 ms 90°-180° 100, 300, 900, 1500, 2200, 3300 ms 16 5 mm 256 × 128 × 16 1 6/8 in slice; 5/8 in phase 0.83 minutes 5 minutes

Siemens FLASH

inhomogeneity leads to variable actual flip angles in different locations even when the same nominal flip angle is prescribed. To ensure more accurate T1 mapping, B1 mapping has to be assessed for correction of errors caused by B1 inhomogeneity.72,142,143 B1 inhomogeneity correction increases the complexity of the data processing and requires additional acquisition time, which diminishes the advantage of the time-efficient VFA method. A potential alternative method for T1 mapping, T1 inversion recovery, uses a single-shot fast-spin echo sequence with multiple inversion of times (TI).22,83 These sequences are less sensitive to B1 inhomogeneity and require similar acquisition times. An example of T1 protocols for the VFA method and the inversion recovery method are given in Table 5-3 for readers’ reference.

DCE-MRI Data Acquisition DCE-MRI data should be acquired by using a 3D T1-weighted spoiled gradient echo sequence, which is widely available on 1.5 and 3 T scanners from major vendors. For better B1 homogeneity, a body coil should be used as the transmit coil; the receiver coil can be a body or phased array coil, depending on the anatomy being studied. The acquisition volume should cover the whole lesion (or the major lesions) and a feeding vessel with in-plane flow for measuring arterial input function (AIF) if feasible. The ideal echo time (TE) should be less than 1.5 milliseconds to improve SNR and to diminish the T2* effects on signal change. The repetition time (TR) should be approximately 3 to 5 milliseconds to achieve higher temporal resolution and optimal T1

Siemens HASTE

quantification. The temporal resolution should be less than 10 seconds for quantifying perfusion. A small flip angle could cause saturation effects when used with higher tracer concentrations. Therefore 25- to 30-degree flip angles are recommended to minimize saturation effects and balance specific absorption rate (SAR), which measures the amount of energy deposited into the human body by radiofrequency pulses. Faster acquisition techniques, including partial Fourier and parallel imaging, can be used with additional attention to ensure consistency on all scanners. The total acquisition time should be no less than 5 minutes; the optimal is about 8 to 10 minutes to ensure the accuracy of estimating kinetic parameters related to the relatively slow process of permeability.68,113 An example of a DCE-MRI protocol is shown in Table 5-4 and can be modified for specific applications. Additional details and sample protocols can be found in a workshop report by the Imaging Committee of the Experimental Cancer Medicine Centres (ECMC)68 and in the Quantitative Imaging Biomarkers Alliance (QIBA) protocols and profiles.26 Contrast injection should start after at least five measurements have been acquired to allow for a robust nonenhanced baseline reference. The preferred method for contrast agent delivery is a bolus-like injection consisting of a fast constant rate injection of generally 2 to 4 mL/s to ensure that the most meaningful perfusion information is obtained in the early enhancement phase. Because of the relatively small volumes of GBCA being administered, the injection should be immediately followed by a saline flush of 20 to 40 mL administered at the same rate

84

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

as the bolus to ensure that all contrast agent is cleared from the line. Although a manual bolus injection can be used, a power injector synchronized with the MRI scanner should be used for the most consistent results.26,68

MRI Signal vs. Tracer Concentration MRI signal is measured in arbitrary units and varies substantially with many factors, including the scanners, coils, field inhomogeneity, and imaging location. Therefore MRI signal is not a suitable measure with which to directly quantify physiologic parameters in large clinical studies. However, MRI signal can be converted to tracer concentration for further modeling analyses. The first step in this conversion is to convert the gradient echo signal into T1 relaxation rate (R1 = 1/T1). The second step is to calculate the tracer concentration by using an assumed linear relationship: R1 = R10 + r1 ⋅ C, where the baseline R10 = 1/T10 is measured in a precontrast scan as described previously, r1 is the T1 relaxivity of the specific GBCA administered, and C is the GBCA concentration.46 It has been demonstrated that the difference in T1 relaxation rates (ΔR1 = R1 − R10) is approximately linearly proportional to the GBCA concentration. Most in vivo studies use an r1 of approximately 4.0 kg/s/mmol, which is a value originally measured in cartilage.29 However, r1 has also been reported to vary in different

TABLE 5-4  An Example DCE-MRI Protocol Parameters

DCE-MRI

B0 Sequence Vendors Sequence name Time of echo (TE) Time of repetition (TR) Flip angles Parallel imaging Slice number Slice thickness Acquisition matrix Bandwidth Partial Fourier Time/measures Measurement number Total time

1.5 T 3D spoiled gradient echo GE Phillips Siemens SPGR T1-FFE FLASH 1.2 ms 3.5 ms 30° GRAPPA R=2, Nref=24 16 5 mm 256 × 192 × 16 360 Hz/pixel 6/8 in slice and phase 7 seconds 50 (minimum) 80 (optimal) 5:50 min 9:20 min

anatomic locations in animal models for different agents. For example, r1 for Gd-EOB-DTPA changes from 2.0 in kidney to 10.7 in liver, and r1 for Gd-DTPA changes from 1.2 in kidney to 4.8 in liver.108 Additional care should be taken when using a constant r1 value in studies that may involve various types of tissue.

Heuristic Analysis Heuristic analyses attempt to extract the important features or attributes of the contrast-enhancing signal curve without any assumptions or inferences about the underlying physiology of the tissue. The main heuristic parameters include initial slope, peak enhancement, time to peak, terminal slope, and initial area under the Gd concentration-time curve (IAUGC) as shown in Figure 5-1.34,43,57,77,99,100 The IAUGC refers to the time period defined by the first 30, 60, 90, 120, or 180 seconds after bolus arrival, with the IAUGC_60 being the most commonly used in DCE-MRI studies.67 These parameters are attractive clinically because they provide a simple and straightforward measure of the contrast-enhancing signal curve without requiring further measurements, such as an AIF, but have less physiologic significance and could vary greatly with different protocols, contrast agent injections, and scanners.67 The IAUGC parameter has a higher reproducibility because it is calculated from the tracer concentration curve by acquiring the additional baseline T10.34,44 Therefore measurements of IAUGC have become one of the primary end points in many DCE-MRI studies. Pattern analysis methods have been proposed to classify DCE-MRI signal curves into several groups on the basis of their curve shapes to differentiate features of benign and malignant tumors.31,66 These methods quantify areas with different curve patterns from DCE-MRI data. Pattern analysis method can also be combined with additional quantification methods to automatically generate several quantitative parameters, which are less sensitive to acquisition protocols but still do not require additional measures of an AIF and baseline T10 measurement.46 One parameter in this hybrid method is highly correlated with the kinetic parameter Kep in the modeling analyses described later and could provide some physiologic significance. The application of this method in DCE-MRI studies is straightforward and can be applied to existing DCE-MRI studies that may not have contained the additional data needed for modeling analyses.

Arterial Input Function For more elaborate kinetic models, an AIF is necessary. An AIF is one of the dominant sources of variability in evaluating kinetic parameters. Three types of AIFs can be used: an assumed AIF, a measured AIF, and a reference AIF. An assumed AIF can be any kind of predefined function. In the early 1990s, assumed simple AIFs, such as a biexponential

∆S

∆S kt ki

IAUGC

∆Smax 0

0

A

∆t

Time

B

Time 60 sec

FIG 5-1  Diagrams demonstrating dynamic contrast-enhanced (DCE)-MRI heuristic parameters. A, Initial slope ki, peak enhancement ΔSmax, time to peak Δt, and terminal slope kt. B, Initial area under the Gd concentration-time curve (IAUGC) in the first 60 seconds after the bolus point. ΔS = (S(t) − S(0))/S(0).

85

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging 7

Cb (t) (mMol)

Model 1

High temporal resolution AIF

6

Weinmann AIF

5 4 Arteriole

3 2

Venule

A

1

Model 2

0 0

1

2

3

4

5

6

Time (min)

FIG 5-2  Comparison of an assumed arterial input function (AIF) (Weinmann biexponential AIF) and a measured averaged AIF. (From Parker GJ, et al: Experimentally-derived functional form for a populationaveraged high-temporal-resolution arterial input function for dynamic contrast-enhanced MRI. Magn Reson Med 56:993–1000, 2006.)

Arteriole

Venule

B function, were used in model analysis.120 The advantage of using an assumed AIF in the analysis is the simplicity of implementation, but the disadvantage is that it could lead to large systematic errors.86 An assumed AIF cannot reflect any change in shape of the bolus for different organs or different injections. A measured AIF can be extracted from the vessel in the dynamic data or measured separately in a previous examination. A visual comparison between an assumed biexponential AIF and a measured AIF from a large artery is shown in Figure 5-2.87 A reference AIF can be explicitly evaluated or be implicitly assessed by using information from a reference tissue such as muscle. Because an AIF is crucial for the accuracy of estimating kinetic parameters, researchers began to measure individual AIFs within  each subject. To measure it accurately, high temporal resolution (2-3 seconds) is required. A low concentration of contrast agent must be administered to avoid signal saturation in the blood vessel. Both the shape and amplitude of the AIF can still change dramatically depending on which artery is used and where the region of interest is located within the artery to extract the AIF. Therefore vessel orientation, inflow effects, and tumor position must be considered to measure an accurate AIF.12,87,102,113 However, even with great care, the amplitude of the AIF may still vary substantially (≈50% change) when measured in the same location from a single subject over five visits because of partial volume corruption and timing effects.87 In certain tumors (e.g., breast tumors, osteosarcoma), no large arteries in which to extract an AIF may be visible within the imaging sections acquired in many subjects. An averaged AIF extracted from a smaller subset of subjects with large arteries within the imaging plane can serve as a surrogate AIF for the whole cohort of subjects. Another option is to use a population-averaged AIF from a cohort of subjects imaged multiple times, which can have an experimentally derived functional form to model the AIF. These AIFs are similar to the assumed AIF, but their amplitudes and shapes are closer to those of measured AIFs and can be used with proper adjustment.47,87 This approach of using a population-averaged AIF may substantially improve measurement repeatability.87 The reference AIF approach is derived explicitly or implicitly  from one or more reference regions of normal tissue such as muscle.36,40,64,106,137-139 These methods assume normal tissue pharmacokinetic properties to allow reverse calculation of the AIF or direct calculation of kinetic parameters in tumors. However, there are two concerns with these methods. First, muscle usually has a very limited contrast enhancement compared to that of tumor and especially artery.

Model 3

Arteriole

Venule

C FIG 5-3  Three models demonstrating different numbers of tissue compartments. A, Model 1 represents a one-compartment model. B, Model 2 represents a two-compartment model, including plasma and interstitial space. C, Model 3 represents a three-compartment model, including plasma, interstitial space, and cellular space (light blue).

Robustness and accuracy become serious problems when trying to estimate large changes in tumors from small signal changes in muscle. Second, pharmacokinetic properties in muscle may be altered by therapy or atrophy during treatment. These variations from the assumed values may cause substantial errors in the derived tumor properties. Although researchers have expended much time and effort to improve its accuracy and reliability, AIF estimation still remains a challenging and dominant factor influencing DCE-MRI analysis.

Modeling Analysis Within an MR image, each voxel usually includes different tissue compartments such as a vascular space, a cellular space, and an interstitial space (extravascular extracellular space [EES]). The kinetic models describing the dynamic contrast-enhancing process can be divided into three categories: one-, two-, and three-compartment pharmacokinetic models, as shown in Figure 5-3. In the one-compartment model shown in Figure 5-3A, a single compartment is assumed, where the dominant effect is the relationship between the signal and drug concentration within the vascular space while contributions from the other two tissue compartments can be neglected owing to relatively small size and/or extremely rapid water exchange between compartments. In the two-compartment model shown in Figure 5-3B, effects from both vascular space and EES are considered for fitting the data more accurately. This model assumes that contrast does not cross the

http://radiologyebook.com

86

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging 250

250

200

200

150

150

100

100

50

50

0

0

A D

C A

0

B

200

300

400

C

250

250

200

200

150

150

100

100

50

50

0

0 0

B

100

100

200

300

400

Time (sec)

0

100

0

100

D

200

300

400

200

300

400

Time (sec)

FIG 5-4  MRI signal curves from four regions in a large heterogeneous bone tumor (osteosarcoma) in the right distal femur. A postcontrast T1-weighted image is shown on the left. Four signal curves corresponding to four different regions (A-D) are shown on the right. (From Reddick WE, et al: Dynamic MR imaging (DEMRI) of microcirculation in bone sarcoma. J Magn Reson Imaging 10:277–285, 1999.)

cellular membrane and that the water exchange across the membrane is slow enough that the cellular space can be ignored. The threecompartment model shown in Figure 5-3C includes all three tissue spaces and becomes much more complicated. Although the GBCA still does not cross the cellular membrane, the water exchange across the cellular membrane is fast enough that the water within the cell also experiences the shortened T1 effects of the contrast. The number of fitted parameters usually increases as the number of compartments increases, which should provide a better fit to the data. However, overfitting, which occurs when a model is excessively complex with too many parameters, could become an issue. The effects from random error or noise are emphasized instead of the underlying physiologic relationships in case of overfitting. If this occurs, then the model generally has poor predictive performance in clinical trials because it can exaggerate minor fluctuations in the data. Therefore the optimal model is the simplest model that fits the data with the least number of parameters. DCE-MRI analysis was first applied to the averaged concentration curve for a whole tumor because it is easy to perform and the averaged curve has higher SNR than that of a single voxel. Researchers found that signal curves could be dramatically different in different tumor regions, as shown in Figure 5-4.101 Data have to be analyzed pixel-bypixel to generate parametric maps for capturing physiologic variation inside heterogeneous tumors. DCE-MRI data can vary dramatically for different types of tumors, with a model performing well for one kind of tumor and failing in another. This characteristic resulted in many different kinetic models being proposed in the past, but under certain conditions and assumptions, these kinetic models could all be derived from one another. Three of the most popular kinetic models will be described in detail below: Tofts (one-compartment), extended Tofts (two-compartment), and shutter-speed models (three-compartment). For further reading to understand the details and complicated relationships among different kinds of kinetic models, we recommend the excellent review on DCE-MRI modeling by Sourbron et al.111

One-Compartment Model A one-compartment model describes a voxel or a region of interest as a single compartment to simplify the impulse response function to a monoexponential decay. The single compartment could refer to vascular space, EES, or the combination of both. Therefore there are four regimes: (1) fast exchange regime, (2) weakly vascularized regime, (3) highly vascularized regime, and (4) no exchange regime.111 The fast exchange regime assumes the GBCA exchange between vascular space and EES is very fast, such that these two spaces can be treated as a single compartment. The weakly vascularized regime applies to tissues such as connective tissues or partially necrotic tumor tissues, which have a less abundant supply of capillaries. The highly vascularized regime applies to tissues such as large vessels or tumor edge, which have a dominant supply of capillaries. The no exchange regime refers to the brain, in which there is no drug exchange between the vascular space and the EES because of the BBB. In the early 1990s, Tofts proposed a simple one-compartment model, which is also referred to as the Tofts/Kety model and belongs to the weakly vascularized regime.117,120 The Tofts model can be written as: Ct (t) = Cp (t) ⊗ I(t) = Cp (t) ⊗ K

trans

⋅e

−

K trans ⋅t ve



(1)

Ct(t) represents the contrast concentration in tissue at time t, and Cp(t) represents the contrast concentration in plasma (AIF). I(t) is the impulse response function and ⊗ represents the convolution operator. The impulse response function is a monoexponential decay characterized by ve, the fractional volume of the EES, and the volume transfer constant Ktrans. The volume transfer constant is a combination of the arterial plasma flow Fp and the permeability–surface area product PS. The ratio of Ktrans/ve is often referred to as kep and represents the efflux rate from the EES to the plasma.144 This model was originally

http://radiologyebook.com

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging developed for applications in multiple sclerosis lesions, which are weakly vascularized.120 However, researchers started to use this simple model in other body parts such as tumors, where the assumption of a weakly vascularized regime was likely not valid for many regions.

Two-Compartment Model Because one-compartment models often fail to adequately describe the observed signal behavior, two-compartment models were proposed to take into account both the vascular space and the EES. Tofts et al. proposed an extended Tofts model that considered the tracer in the plasma.118 The equation of the dependence of drug concentration in tissue is as follows:

Ct (t) = Cp (t) ⋅ vp + Cp (t) ⊗ K trans ⋅ e

−

K trans ⋅t ve



(2)

Although most notations within this equation are the same as those in Equation (1), there is now a contribution from vp that represents the plasma fractional volume. This plasma fraction is a function of the blood fractional volume vb and the hematocrit of the microvasculature Hct as given by the relationship vp = vb ⋅ (1 − Hct). The average hematocrit for a human adult is approximately 0.45 but could vary between individuals, which would cause an error in estimates of Ktrans and vp.35 The extended Tofts model is valid in highly and weakly vascularized tissues with a well-mixed interstitium. This model has become the most widely used model in DCE-MRI and has proven useful in many clinical applications because it provides useful estimates of Ktrans, ve, and vp at relatively low temporal resolution.74 However, the extended Tofts model still provides a poor fit for certain types of tumors, including those of breast, neck, and cervical tissue.30,58,69,140 Therefore new models were proposed to improve the fitting of DCE-MRI data in these tumors by considering effects from compartments that are neglected in the extended Tofts model.70,88,140

Three-Compartment Model A three-compartment model was proposed to account for contributions to the signal from water within all three tissue spaces. These models require estimating at least five parameters: two water exchange rates across the vascular endothelium and the cellular cytolemma, and the three kinetic parameters from the extended Tofts model.70,88 The SNR in DCE-MRI data is limited for voxel-by-voxel analysis and can usually support the fitting of only three or four unknown model parameters.2,54 Therefore full three-compartment models with five unknown parameters are not recommended for use in regular DCE-MRI analyses. To make three-compartment models more practical, the number of unknown parameters can be reduced by either neglecting water exchange effects or fixing two parameters.35 The shutter speed model (SSM, or shutter speed approach [SSA]) is a simplified three-parameter water exchange model that eliminates vp and the water exchange rate across the vascular endothelium.52,141 The shutter speed model estimates three parameters: Ktrans, ve, and the mean intracellular water lifetime τ1, which is related to the water exchange rate across the cellular cytolemma.52 Several studies have shown that the three-parameter SSM provided a better fit in com­ parison with Tofts and extended Tofts models for some applications.52,58,69,140 SSM assumes a nonlinear relationship between the contrast concentration and the change of relaxation rate, which is assumed to be linear in both Tofts models.118,140 This nonlinearity is caused by the limited water exchange rate across the cellular cytolemma, which is assumed to be extremely fast in the extended Tofts model.141 Comparisons of fitting results obtained by using the Tofts model and SSM in breast tumor are shown in Figure 5-5, which clearly demonstrate that SSM provides a much better fit for the

87

measured data in two different types of breast lesions.53 Although SSM has provided better fits in some example tumors than the extended Tofts model has, very few systematic studies have compared these models in different tumors and regions.35,58 One study in internal obturator muscle found no advantage in using SSM in comparison with the extended Tofts model. Furthermore, clinical DCE-MRI data are often unsuitable for assessing cellular-interstitial water exchange because this exchange rate is too fast to estimate by using the relatively slower acquisitions.11

Discussion and Limitations DCE-MRI has become widely accepted in clinical practice and heavily used in clinical trials research. It is also heavily used in tumor grading of individual subjects to facilitate the design and optimization of individualized treatment strategies to improve clinical outcomes. Histologic tumor grading has been the historical gold standard but is limited by its invasive nature, sampling errors, and subjective variability.56 DCE-MRI has been used in brain tumors to differentiate grades of glioma (grade II and IV) as shown in Figure 5-6. There is a clear indication of increased blood volume and leakage in the grade IV glioma relative to both the surrounding brain and the grade II glioma.56 These noninvasive grading results in glioma could be further used in predicting prognosis.20,76 DCE-MRI also has utility in differentiating treatment-induced effects such as pseudoprogression and necrosis from recurrent progressive tumor3,97 and in monitoring and evaluating tumor response.46,82,145 Even with these examples of clinical utility, DCE-MRI results can still vary dramatically depending on the model used to process the data, which emphasizes the need to understand the tissue being examined and choose the most appropriate model for the particular application. Even when using the same Tofts model and the same data, kinetic parameters from four commercial software packages can still vary by an order of magnitude.50 Even the heuristic parameter IAUGC is significantly different among four commercial software solutions. These results emphasize the need to understand the approach used in each of the commercial software solutions to better understand the limitations and appropriateness of using the software in a particular application. Given the complexity of the analysis models and the variability between commercial software solutions, many of the DCE-MRI applications remain within research institutions. Even within an experienced research institution, there are still some limitations for routine DCE-MRI imaging. A series of reproducibility studies were conducted to determine the variation that could be expected in DCE-MRI parameters to establish the magnitude of change needed for reliable detection in multiinstitutional clinical trials. Using the same analysis implementation, coefficients of variation for Ktrans, ve, kep, vp, and IAUGC60 were up to 20.3%, 23.6%, 21%, 30%, and 18.2%, respectively.26 These variations are primarily attributed to the different DCE-MRI acquisitions and demonstrate the need to ensure that the highest quality DCE-MRI data are acquired. Many DCE-MRI kinetic models were developed on the basis of different assumptions and purposes.35,111 Model selection is complicated and confusing for less experienced researchers. Because each kinetic model has its own limits and could fit some data well and other data poorly, automatic model selection has become a promising approach.2,54,73,111 These tools can provide objective model selection and produce robust and reproducible estimates of model parameters, which will enhance the utility of DCE-MRI in clinical applications. Because DCE-MRI requires a long acquisition time, voluntary motion during the scan can be a large problem, which is further  complicated in imaging targets such as lung and liver tumors that undergo involuntary motion as well. Most DCE-MRI acquisitions are

http://radiologyebook.com

88

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging 5.0 4.5 4.0

Lesion R1 (S–1)

3.5 3.0 2.5 2.0 1.5

SA SSA

1.0

C

0.5 [CRp] (mM)

A

R1 (S–1)

3.5

Lesion

3.0 2.5 2.0

4 2

2 4 6 time (min)

1.5 1.0 0.5

B

1

D

2

3

4

5

6

time (min)

FIG 5-5  Comparison of results fitted by using standard approach (SA: Tofts model) and shutter speed approach (SSA). A and B show two different breast lesions. C and D show corresponding results of fitting. The inset in D shows the arterial input function used for all analyses. (From Huang W, et al: Discrimination of benign and malignant breast lesions by using shutter-speed dynamic contrast-enhanced MR imaging. Radiology 261:394–403, 2011.)

now volumetric, with multiple sections acquired at each time point; voluntary motion can be compensated for to some degree by using simple affine registration methods to adjust for shifts and rotations. For body parts such as liver and lung, involuntary motion during image acquisition causes ghosting and blurring in addition to the  rigid shifts and rotations. Although the cooperation of patients can mitigate motion effects to some extent, fast imaging techniques with motion compensation are fundamental to overcoming the motion problems. The golden-angle radial sequence with self-gating has shown promise for DCE-MRI in lung and liver lesions.39,71 A comparison between Cartesian and self-gated golden-angle radial images in free breathing is shown in Figure 5-7. The golden-angle radial acquisition with self-gating achieved a much better image quality even with free breathing.

DSC-MRI Dynamic susceptibility contrast (DSC) MRI is a well-established  perfusion quantification method in brain imaging. In DSC-MRI, T2*or T2-weighted MR images are acquired to extract kinetic parameters in brain perfusion instead of using the T1-weighted images used in DCE-MRI. DSC-MRI is primarily restricted to use in brain imaging because most GBCAs leak from the vasculature throughout the body except in the brain, where the agents are blocked by the BBB. Several useful kinetic parameters can be calculated by using DSC-MRI, including cerebral blood flow (CBF), cerebral blood volume (CBV), mean transit time (MTT), and time to maximum (Tmax). DSC-MRI parametric maps are shown in Figure 5-8. Although DSC-MRI acquisition and data processing involve many technical aspects, we have focused on

some of the more important aspects to introduce in the following sections. Several good review articles describe the more detailed aspects of DSC-MRI.13,62,130

Administration and Acquisition GBCAs are usually administered intravenously, and the bolus of agent circulates through a lengthy pathway from heart, lungs, back to heart, and finally to the brain. This long circulation leads to apparent dispersion of the bolus shape. To reduce this bolus broadening effect, a higher rate of injection of contrast agent should be used.7,121 Lower injection rates (<3 mL/s) lead to underestimation of the CBF according to simulations.125 Therefore in DSC-MRI studies, the contrast should be injected at a constant rate of 5 mL/s and should be immediately followed by a saline flush of 20 to 40 mL administered at the same rate to ensure all contrast agent is cleared from the line. Any further increase in rate only causes a minor change in the shape of the bolus for brain imaging.62,125 Echo planar imaging (EPI) has an intrinsic advantage in DSCEMRI studies because of its fast acquisition. T2*-weighted single-shot gradient echo (GE) EPI with two-dimensional (2D) interleaved slices has been widely used in clinical DSC-MRI studies because it can achieve very high temporal resolutions (≤1.5 seconds).45,49,109,135 Any slower temporal sampling could decrease estimates of CBF and increase the standard deviation, as demonstrated in simulations.125 Although segmented EPI and some other alternative sequences have also been tested in DSC-MRI studies, those sequences are not widely used because of either T1-contaminating effects or their lower temporal resolution.90,122,127 T1 contamination can be corrected by using multiecho EPI sequences.80 Recently the multiband or simultaneous

http://radiologyebook.com

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging

A

B

Vp = 8.45 %

C

D

E

Vp = 1.22 %

F

89

Ktrans = 0.21 min–1

Ktrans = 0.02 min–1

FIG 5-6  Dynamic contrast-enhanced (DCE)-MRI differentiation of grade IV and grade II glioma. A postcontrast T1-weighted image, vp, and Ktrans parametric maps for a grade IV glioma are shown in the first row (A-C). The corresponding results for a grade II glioma are shown in the second row (D-F). (From Jain R: Measurements of tumor vascular leakiness using DCE in brain tumors: Clinical applications. NMR Biomed 26:1042– 1049, 2013.)

multislice technique based on EPI sequences was developed to achieve even faster acquisitions and are employed in different applications, such as diffusion imaging.38 Multiband EPI in DSC-MRI can either improve the temporal resolution to approximately 500 milliseconds or increase the volumetric coverage to 36 slices with a more traditional temporal resolution of 1.4 seconds.19,128 In DSC-MRI, the change in signal caused by the GBCA concentration mainly depends on the TE used in the acquisition. GBCA concentration in brain tissue is usually quite low owing to the very small vascularization percentage (<5%) of the tissue. A longer TE increases the magnitude of the signal drop due to the contrast agent, but the TE should not be too long because this would result in a loss of almost all signal within the brain at the highest contrast concentration. Therefore an optimal TE (about 30-45 milliseconds at 1.5 T) can be found by balancing the larger signal drop and a better SNR in the baseline signal.110,135 However, AIF measurements require a shorter TE for arteries, and an acquisition with two TEs (a short one for AIF and long one for tissue) has been suggested.92,127 In addition, T1 contamination

effects on the signal caused by a shorter TR and flip angle and fresh inflow effects are potential complications. These issues can be addressed by using a smaller flip angle (60-90 degrees) and a longer TR.18 Because TR also determines the temporal resolution, an optimal TR (≈1.5 seconds) must be selected to balance the temporal resolution and T1 contamination. A sample clinical DSC-MRI protocol is shown in Table 5-5, which can be modified according to the specific applications.135

MRI Signal vs. Tracer Concentration Although the conversion of MRI signal to contrast agent concentration in DCE-MRI is rather complicated, the conversion for DSC-MRI is simple and straightforward. MR signal curves in arteries and  tumor tissues can be converted to GBCA concentration curves as shown in Figure 5-9. The change of T2* relaxation rate due  to GBCA injection at time t can be calculated directly by using  the signal S(t) and the averaged baseline signal S0 according to the 1  S0  ln  equation: ∆R *2 =  . The relationship between ∆R *2 and TE  S(t)

http://radiologyebook.com

90

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

B

A

FIG 5-7  Comparison of image quality between Cartesian and golden-angle self-gated radial DCE-MRI with free breathing. A, Cartesian image. B, Self-gated radial image. Lesions are indicated by arrows. (From Lin W, et al: Respiratory motion-compensated radial dynamic contrast-enhanced [DCE]-MRI of chest and abdominal lesions. Magn Reson Med 60:1135–1146, 2008.) FLAIR

A

CBF

B

CBV

MTT

112

25

6.0

90

20

4.8

67

15

3.6

45

10

2.4

22

5

1.2

0 (mL/100 g/min)

0 (mL/100 g)

C

D

(s)

0

Parameters

DSC-MRI

B0 Sequence name Time of echo (TE) Time of repetition (TR) Flip angles Slice number Slice thickness Acquisition matrix Time/measures Measurement number Total time

1.5 T 3.0 T Single shot gradient echo EPI 35-45 ms 25-30 ms 1500 ms 60-90° 60° 12 5 mm 128 × 128 × 12 1.5 second 60-80 90-120 seconds

Artery Tissue

Concentration

TABLE 5-5  An Example DSC-MRI Protocol

MR signal (A.U.)

FIG 5-8  Dynamic susceptibility contrast (DSC)-MRI parametric maps of a pediatric patient with an ependymoma. A, Fluid-attenuated inversion recovery (FLAIR) image with lesion. B, Cerebral blood flow (CBF) map. C, Cerebral blood volume (CBV) map. D, Mean transit time (MTT) map. Tumor region has high CBF and CBV, and the surrounding edematous regions have very low CBF and CBV values.

Time

Time

FIG 5-9  Examples of dynamic susceptibility contrast (DSC)-MRI signal curves and corresponding Gd contrast agent concentration in artery and tissue.

http://radiologyebook.com

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging Input

System

×

Arterial input function

91

Output

=

Unknown

Residue function

Tissue concentration curve

FIG 5-10  Diagram of dynamic susceptibility contrast (DSC)-MRI analysis. Arterial input function is an input; residue function represents the properties of the system, which is brain tissue; the output is the tissue concentration curve. DSC-MRI analysis determines the unknown residue function. ⊗, convolution sign.

tracer concentration is assumed to be almost linear in tissue, and has been further verified by using simulations.61 Therefore the GBCA concentration C can be calculated by using ∆R *2 = r2* ⋅ C , where r2* is the T2* relaxivity, which is approximately 0.044(ms ⋅ mM)−1 at 1.5 T and 0.087(ms ⋅ mM)−1 at 3 T for Gd-DTPA in brain tissue.61 There have been no differences demonstrated for r2* between gray and white matter in simulations using gradient echo acquisition.61 However, according to another study comparing CT perfusion with DSC-MRI in healthy volunteers, there may be some degree of tissue-dependent r2* difference.81 Currently one single value of r2* is assumed in most studies.

Residue function

CBF

CBV

Arterial Input Function.  An accurate AIF is very important in the quantification of DSC-MRI. Measurement of a global AIF, a single AIF for a whole brain, has several technical challenges, including bolus dispersion, nonlinearity of the relationship between the tracer concentration and MR signal, T1 contamination, partial volume effects due to low spatial resolution, and geometric distortions.14-16,18,21,51,123 There is a clear nonlinear relationship between ∆R *2 and the tracer concentration observed in arteries.124 However, the linear relationship is still used in practice to calculate the arterial concentration in most studies. This linear assumption leads to substantial CBF errors when the magnitude of MR signal is used to calculate the tracer concentration.14 A quadratic relationship can be used to extract more accurate estimations of tracer concentration from the MR signal, which makes the concentration less dependent on the hematocrit and the assumed relaxivity.14,124 Another way to estimate the tracer concentration is by using the MR signal phase changes instead of magnitude changes caused by the contrast. A good linear relationship exists between the phase and tracer concentration, and this relationship is independent of hematocrit levels.24,124 Using phase can also substantially increase SNR of the AIF in comparison with those magnitude-based AIFs.1,25,63 Despite these advantages, a phase-based AIF is still not widely used in clinical DSC-MRI protocols, because the phase data are usually not saved in these protocols. Another approach using local AIFs, individualized AIFs for local brain regions, has been proposed to reduce the CBF quantification errors related to bolus dispersion.17 Because of some controversy about the validity of the local AIF method and its technical implementation, the local AIF method is not widely used in clinical studies.5,131 A good comprehensive review about these issues in DSC-MRI has been published for those who are interested in more details about methods for estimating AIF.13

Data Analysis Data analysis in DSC-MRI is relatively simple compared to postprocessing in DCE-MRI. The tracer concentrations in artery and tissue

Time

FIG 5-11  The relationship between residue function and dynamic susceptibility contrast (DSC)-MRI parameters such as cerebral blood flow (CBF) and cerebral blood volume (CBV). CBF is the maximum amplitude of the residue function; CBV is the total gray area of the residue function. MTT is equal to CBV/CBF.

can be calculated as described in the previous sections. The relationship between the tissue and arterial concentrations shown in Figure 5-10 can be connected by a residue function, R(t), which represents the fraction of tracer remaining in the voxel at a given time, t. The tissue tracer concentration, Ct(t), in a voxel can be written as a convolution of the AIF, Ca(t), and the tissue residue function R(t)85:

Ct (t) = I(t) ⊗ Ca (t) = CBF ⋅ R(t) ⊗ Ca (t)

(3)

where ⊗ represents the convolution. I(t) = CBF ⋅ R(t) is the unknown function to be solved. In Equation (3), I(t) is usually solved first and CBF is commonly estimated to be equal to the maximum of this function, max(I(t)), because the initial value of I(t) is very sensitive to noise errors and bolus dispersion errors.114 The MTT is a measure of how fast the bolus goes through the voxel, which can be calculated as the ratio of residual ∞ area to its maximum height, MTT = ∫ 0 I(t)dt max(I(t)). CBV is a measure of how much blood is in the voxel and reflects the capillary density or total blood volume in the voxel. CBV can be calculated as ∞ the integral of the I(t) function, CBV = ∫ 0 I(t)dt , which is the area under the I(t) function. Alternatively, CBV can be calculated as the product of CBF and MTT, CBV = CBF ⋅ MTT, according to the central volume theorem.85,91 The relationships of CBF, CBV, and the residue function are shown in Figure 5-11. The reproducibility of all these

http://radiologyebook.com

92

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

parameters are functions of both the accuracy of the AIF and the approach used to solve Equation (3). Solving Equation (3) is a typical inverse problem known as deconvolution. In DSC-MRI, it is usually an ill-posed problem, which means that the solution is highly sensitive to noise and a small error can cause a large variation in the estimates of CBF.85 Regularization has to be applied to ensure that the estimates of the parameters are within a plausible range. Singular value decomposition (SVD) is an approach widely used to solve the matrix form of Equation (3) as an inverse matrix problem.85 However, perfusion parameters in Equation (3) highly depend on the bolus delay between the AIF and the tissue concentration curve.15 Oscillation-limited SVD (oSVD) was proposed as a method to overcome this problem by minimizing the variation of the derived residue function and further stabilized DSC-MRI parameters.136 Bolus delay effects are mitigated by using oSVD instead of standard SVD, as shown in Figure 5-12.136 Many other approaches have been proposed to solve Equation (3) and have various strengths and weaknesses. For example, an analytical function can be assumed to ensure a smoothly decaying estimate of the residue function but is very sensitive to noise.84 Statistical regression approaches such as maximumlikelihood expectation maximization algorithms132,133 and bayesian algorithms79 are also proposed to extract more accurate characterizations of the residue function, but their implementations are still very limited because of their intensive computation.

Discussion and Limitations Both gradient echo and spin echo EPI sequences have been extensively investigated in DSC-MRI.49,109,115 Gradient echo EPI is T2*-weighted and more sensitive to larger vessels (>4 µm), whereas spin echo EPI is

T2-weighted and more sensitive to smaller vessels, providing more information about capillary perfusion.6,115,129 Gradient echo EPI sequences have a greater response to a given concentration of contrast because of susceptibility effects, meaning that the signal drop is larger for gradient echo–based than spin echo–based EPI measures. This difference leads to larger CBF and CBV estimates when using gradient echo sequences than when using spin echo, as shown in Figure 5-13.107 For DSC-MRI, gradient echo EPI has been shown to be preferred over or equivalently to spin echo EPI.75,109 Gradient and spin echo information can be used together to assess the vessel size, which is sensitive to the tumor type.60 The combination of gradient and spin echo measurements has been reported to reveal information on vessel type and function besides vessel size, using a temporal shift between gradient and spin echo signals.33 BBB leakage is one of the primary issues in DSC-MRI, because the BBB is assumed to be intact in all the data analyses described in this chapter. BBB leakage causes strong local T1 contamination such that the MR signal could even become T1 dominated instead of T2*/T2 dominated. The model must be adapted according to different situations.117 However, when the BBB is disrupted, the GBCA goes into the tissue and causes both T1 and T2* shortening, making deciphering the underlying MR signal more complicated. The tracer concentration curve could be substantially distorted depending on the sequence type, protocol parameters, predose usage, and degree of leakage. CBV measures could be substantially over- or underestimated according to simulations.95 An in vivo example has demonstrated typical leakage effects on tumor curves, which could be T1 or T2* dominated, as shown in Figure 5-14. These leakage effects can be corrected in data analysis by modifying the model with a leakage factor as shown in AIF 1

AIF 1

ROI 1

AIF 2

ROI 2

AIF 2

12 sSVD

10

∆R2 (t)

8

B

C

6 4 2 oSVD 0

–2

A

10

20

30 Time (s)

40

D

E CBF maps

FIG 5-12  An arterial input function (AIF) delay causes issues for standard singular value decomposition (sSVD), which is mediated by using minimal oscillation SVD (oSVD). A shows two AIFs and ∆R* 2 curves from

two tissue ROIs. An inset image shows the lesions with an arrow. B and C show CBF maps produced by using sSVD, which demonstrates dramatic changes caused by the shift of the two AIFs. D and E are CBF maps produced by using oSVD, which show similar results for the two AIFs. (From Wu O, et al: Tracer arrival timing-insensitive technique for estimating flow in MR perfusion-weighted imaging using singular value decomposition with a block-circulant deconvolution matrix. Magn Reson Med 50:164–174, 2003.)

http://radiologyebook.com

93

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging CBF

CBV

MTT

Tmax

T1w GE

SE

0 (mL/100 g/min)

30

0

(mL/100 g)

6 0

(s)

20

0

(s)

10

FIG 5-13  Dynamic susceptibility contrast (DSC)-MRI comparison using gradient echo (GE) and spin echo (SE) sequences in a patient with a subacute infarct in the left posterior middle cerebral artery, marked by the arrow on the T1-weighted postcontrast image. Each column uses the same color scale. (From Schmiedeskamp H, et al: Combined spin- and gradient-echo perfusion-weighted imaging. Magn Reson Med 68:30– 40, 2012.)

50 40

AIF

30

∆R2* (1 / sec)

∆R2* (1 / sec)

50

Tumor GM

20 10 0 –10

∆R2* (1 / sec)

A

Tumor Corrected

0 20

0

80

60

40 Time (sec)

–20

80 Uncorrected

–20

C

35 30

Tumor AIF

25

GM

12

20 15 10 5 0

B

–5

0

20

40 60 Time (sec)

80

100

D

Uncorrected CBV

Corrected CBV

FIG 5-14  Leakage effects on dynamic susceptibility contrast (DSC)-MRI tumor results. A, T1-dominated

extravasation effects on tumor ∆R*2 curve from one subject, which drop below zero after the peak marked by the red arrow. B, T2*-dominated extravasation effects on tumor ∆R*2 curve from another subject; the curve tail stays high and does not return to zero even after the arterial input function (AIF) drops close to zero. C, A diagram of leakage correction for the tumor curve shown in A, with the corrected tumor curve. D, Uncorrected and corrected cerebral blood volume (CBV) maps show different contrast for tumor regions marked by yellow arrows. (A, B, and D from Bjornerud A, et al: T1- and T2*-dominant extravasation correction in DSC-MRI: Part I–Theoretical considerations and implications for assessment of tumor hemodynamic properties. J Cereb Blood Flow Metab 31:2041–2053, 2011.)

http://radiologyebook.com

0 (mL/100 g)

94

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

Figure 5-14. An alternative approach would be to use a dual echo sequence or a predose of GBCA to minimize the T1 effects, thus resulting in more accurate and robust CBV measures.89 Several postprocessing correction methods have also been proposed to account for leakage effects.4,8,32,96 Although these methods can improve the measures to some extent, there is no consensus on the optimal approach to correcting this BBB leakage effect. Another way to avoid this problem is to use DCE-MRI instead of DSC-MRI, because DCE-MRI is intrinsically less affected by BBB leakage. Other issues in DSC-MRI include partial volume effects, arterial delay and dispersion, and geometric distortion, which have been summarized in review articles.62,130

REFERENCES 1. Akbudak E, Norberg RE, Conturo TE: Contrast-agent phase effects: An experimental system for analysis of susceptibility, concentration, and bolus input function kinetics. Magn Reson Med 38(6):990–1002, 1997. 2. Bagher-Ebadian H, Jain R, Nejad-Davarani SP, et al: Model selection for DCE-T1 studies in glioblastoma. Magn Reson Med 68(1):241–251, 2012. 3. Bisdas S, Naegele T, Ritz R, et al: Distinguishing recurrent high-grade gliomas from radiation injury: A pilot study using dynamic contrastenhanced MR imaging. Acad Radiol 18(5):575–583, 2011. 4. Bjornerud A, Sorensen AG, Mouridsen K, et al: T1- and T2*-dominant extravasation correction in DSC-MRI: Part I–Theoretical considerations and implications for assessment of tumor hemodynamic properties.   J Cereb Blood Flow Metab 31(10):2041–2053, 2011. 5. Bleeker EJ, Webb AG, van Walderveen MA, et al: Evaluation of signal formation in local arterial input function measurements of dynamic susceptibility contrast MRI. Magn Reson Med 67(5):1324–1331, 2012. 6. Boxerman JL, Bandettini PA, Kwong KK, et al: The intravascular contribution to fMRI signal change: Monte Carlo modeling and diffusion-weighted studies in vivo. Magn Reson Med 34(1):4–10, 1995. 7. Boxerman JL, Rosen BR, Weisskoff RM: Signal-to-noise analysis of cerebral blood volume maps from dynamic NMR imaging studies.   J Magn Reson Imaging 7(3):528–537, 1997. 8. Boxerman JL, Schmainda KM, Weisskoff RM: Relative cerebral blood volume maps corrected for contrast agent extravasation significantly correlate with glioma tumor grade, whereas uncorrected maps do not. AJNR Am J Neuroradiol 27(4):859–867, 2006. 9. Brookes JA, Redpath TW, Gilbert FJ, et al: Accuracy of T1 measurement in dynamic contrast-enhanced breast MRI using two- and threedimensional variable flip angle fast low-angle shot. J Magn Reson Imaging 9(2):163–171, 1999. 10. Bruening R, Kwong KK, Vevea MJ, et al: Echo-planar MR determination of relative cerebral blood volume in human brain tumors: T1 versus T2 weighting. AJNR Am J Neuroradiol 17(5):831–840, 1996. 11. Buckley DL, Kershaw LE, Stanisz GJ: Cellular-interstitial water exchange and its effect on the determination of contrast agent concentration in vivo: Dynamic contrast-enhanced MRI of human internal obturator muscle. Magn Reson Med 60(5):1011–1019, 2008. 12. Buckley DL, Shurrab AE, Cheung CM, et al: Measurement of single kidney function using dynamic contrast-enhanced MRI: Comparison of two models in human subjects. J Magn Reson Imaging 24(5):1117– 1123, 2006. 13. Calamante F: Arterial input function in perfusion MRI: A comprehensive review. Prog Nucl Magn Reson Spectrosc. 74:1–32, 2013. 14. Calamante F, Connelly A, van Osch MJ: Nonlinear DeltaR*2 effects in perfusion quantification using bolus-tracking MRI. Magn Reson Med 61(2):486–492, 2009. 15. Calamante F, Gadian DG, Connelly A: Delay and dispersion effects in dynamic susceptibility contrast MRI: Simulations using singular value decomposition. Magn Reson Med 44(3):466–473, 2000. 16. Calamante F, Gadian DG, Connelly A: Quantification of perfusion using bolus tracking magnetic resonance imaging in stroke: Assumptions, limitations, and potential implications for clinical use. Stroke 33(4):1146–1151, 2002.

17. Calamante F, Morup M, Hansen LK: Defining a local arterial input function for perfusion MRI using independent component analysis. Magn Reson Med 52(4):789–797, 2004. 18. Calamante F, Vonken EJ, van Osch MJ: Contrast agent concentration measurements affecting quantification of bolus-tracking perfusion MRI. Magn Reson Med 58(3):544–553, 2007. 19. Cantrell DR, Gallagher TA, Spottiswoode B, et al: Clinical implementation of slice accelerated EPI-DSC MR perfusion weighted imaging. Paper presented at: the 22nd Annual Meeting of ISMRM2014; Milan, Italy. 20. Cao Y, Nagesh V, Hamstra D, et al: The extent and severity of vascular leakage as evidence of tumor aggressiveness in high-grade gliomas. Cancer Res 66(17):8912–8917, 2006. 21. Chen JJ, Smith MR, Frayne R: The impact of partial-volume effects in dynamic susceptibility contrast magnetic resonance perfusion imaging. J Magn Reson Imaging 22(3):390–399, 2005. 22. Cheng HL, Wright GA: Rapid high-resolution T(1) mapping by variable flip angles: Accurate and precise measurements in the presence of radiofrequency field inhomogeneity. Magn Reson Med 55(3):566– 574, 2006. 23. Choyke PL, Frank JA, Girton ME, et al: Dynamic Gd-DTPA-enhanced MR imaging of the kidney: Experimental results. Radiology 170(3 Pt 1): 713–720, 1989. 24. Conturo TE, Akbudak E, Kotys MS, et al: Arterial input functions for dynamic susceptibility contrast MRI: Requirements and signal options. J Magn Reson Imaging 22(6):697–703, 2005. 25. Conturo TE, Smith GD: Signal-to-noise in phase angle reconstruction: Dynamic range extension using phase reference offsets. Magn Reson Med 15(3):420–437, 1990. 26. DCE-MRI Technical Committee: DCE-MRI Quantification Profile, Quantitative Imaging Biomarkers Alliance. Version 1.0. Publicly Reviewed Version. QIBA, July 1, 2012. Available from: RSNA.ORG/ QIBA. 27. Detre JA, Leigh JS, Williams DS, et al: Perfusion imaging. Magn Reson Med 23(1):37–45, 1992. 28. Detre JA, Subramanian VH, Mitchell MD, et al: Measurement of regional cerebral blood flow in cat brain using intracarotid 2H2O and 2H NMR imaging. Magn Reson Med 14(2):389–395, 1990. 29. Donahue KM, Burstein D, Manning WJ, et al: Studies of Gd-DTPA relaxivity and proton exchange rates in tissue. Magn Reson Med 32(1):66–76, 1994. 30. Donaldson SB, Buckley DL, O’Connor JP, et al: Enhancing fraction measured using dynamic contrast-enhanced MRI predicts disease-free survival in patients with carcinoma of the cervix. Br J Cancer 102(1):23–26, 2010. 31. Eida S, Ohki M, Sumi M, et al: MR factor analysis: Improved technology for the assessment of 2D dynamic structures of benign and malignant salivary gland tumors. J Magn Reson Imaging 27(6):1256– 1262, 2008. 32. Emblem KE, Bjornerud A, Mouridsen K, et al: T(1)- and T(2) (*)-dominant extravasation correction in DSC-MRI: Part II–Predicting patient outcome after a single dose of cediranib in recurrent glioblastoma patients. J Cereb Blood Flow Metab 31(10):2054–2064, 2011. 33. Emblem KE, Mouridsen K, Bjornerud A, et al: Vessel architectural imaging identifies cancer patient responders to anti-angiogenic therapy. Nat Med 19(9):1178–1183, 2013. 34. Evelhoch JL: Key factors in the acquisition of contrast kinetic data for oncology. J Magn Reson Imaging 10(3):254–259, 1999. 35. Ewing JR: Bagher-Ebadian H. Model selection in measures of   vascular parameters using dynamic contrast-enhanced MRI: Experimental and clinical applications. NMR Biomed 26(8): 1028–1041, 2013. 36. Fan X, Haney CR, Mustafi D, et al: Use of a reference tissue and blood vessel to measure the arterial input function in DCEMRI. Magn Reson Med 64(6):1821–1826, 2010. 37. FDA: Information on Gadolinium-Based Contrast Agents.  
http://radiologyebook.com

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55. 56. 57.

58.

postmarketdrugsafetyinformationforpatientsandproviders/ ucm142882.htm>. Accessed Sept. 19, 2014. Feinberg DA, Moeller S, Smith SM, et al: Multiplexed echo planar imaging for sub-second whole brain FMRI and fast diffusion imaging. PLoS ONE 5(12):e15710, 2010. Feng L, Grimm R, Block KT, et al: Golden-angle radial sparse parallel MRI: Combination of compressed sensing, parallel imaging, and golden-angle radial sampling for fast and flexible dynamic volumetric MRI. Magn Reson Med 72(3):707–717, 2014. Fluckiger JU, Schabel MC, Dibella EV: Model-based blind estimation of kinetic parameters in dynamic contrast enhanced (DCE)-MRI. Magn Reson Med 62(6):1477–1486, 2009. Fork FT, Lindstrom C, Ekelund GR: Reliability of routine doublecontrast examination (DCE) of the large bowel in polyp detection: A prospective clinical study. Gastrointest Radiol 8(2):163–172, 1983. Fram EK, Herfkens RJ, Johnson GA, et al: Rapid calculation of T1 using variable flip angle gradient refocused imaging. Magn Reson Imaging 5(3):201–208, 1987. Fujii K, Fujita N, Hirabuki N, et al: Neuromas and meningiomas: Evaluation of early enhancement with dynamic MR imaging. AJNR Am J Neuroradiol 13(4):1215–1220, 1992. Galbraith SM, Lodge MA, Taylor NJ, et al: Reproducibility of dynamic contrast-enhanced MRI in human muscle and tumours: Comparison of quantitative and semi-quantitative analysis. NMR Biomed 15(2):132– 142, 2002. Griffiths PD, Pandya H, Wilkinson ID, et al: Sequential dynamic gadolinium magnetic resonance perfusion-weighted imaging: Effects on transit time and cerebral blood volume measurements. Acta Radiol 47(10):1079–1084, 2006. Guo JY, Reddick WE: DCE-MRI pixel-by-pixel quantitative curve pattern analysis and its application to osteosarcoma. J Magn Reson Imaging 30(1):177–184, 2009. Guo J, Reddick WE, Glass JO, et al: Dynamic contrast-enhanced magnetic resonance imaging as a prognostic factor in predicting event-free and overall survival in pediatric patients with osteosarcoma. Cancer 118(15):3776–3785, 2012. Haroon HA, Patankar TF, Zhu XP, et al: Comparison of cerebral blood volume maps generated from T2* and T1 weighted MRI data in intra-axial cerebral tumours. Br J Radiol 80(951):161–168, 2007. Heiland S, Kreibich W, Reith W, et al: Comparison of echo-planar sequences for perfusion-weighted MRI: Which is best? Neuroradiology 40(4):216–221, 1998. Heye T, Davenport MS, Horvath JJ, et al: Reproducibility of dynamic contrast-enhanced MR imaging. Part I. Perfusion characteristics in the female pelvis by using multiple computer-aided diagnosis perfusion analysis solutions. Radiology 266(3):801–811, 2013. Hou L, Yang Y, Mattay VS, et al: Optimization of fast acquisition methods for whole-brain relative cerebral blood volume (rCBV) mapping with susceptibility contrast agents. J Magn Reson Imaging 9(2):233–239, 1999. Huang W, Li X, Morris EA, et al: The magnetic resonance shutter speed discriminates vascular properties of malignant and benign breast tumors in vivo. Proc Natl Acad Sci U S A 105(46):17943–17948, 2008. Huang W, Tudorica LA, Li X, et al: Discrimination of benign and malignant breast lesions by using shutter-speed dynamic contrastenhanced MR imaging. Radiology 261(2):394–403, 2011. Ingrisch M, Sourbron S, Morhard D, et al: Quantification of perfusion and permeability in multiple sclerosis: Dynamic contrast-enhanced MRI in 3D at 3T. Invest Radiol 47(4):252–258, 2012. ISMRM: MR contrast agents—Public comment on NSF. 2009; . Accessed Sept. 18, 2014. Jain R: Measurements of tumor vascular leakiness using DCE in brain tumors: Clinical applications. NMR Biomed 26(8):1042–1049, 2013. Kaiser WA, Zeitler E: MR imaging of the breast: Fast imaging sequences with and without Gd-DTPA. Preliminary observations. Radiology 170(3 Pt 1):681–686, 1989. Kim S, Quon H, Loevner LA, et al: Transcytolemmal water exchange in pharmacokinetic analysis of dynamic contrast-enhanced MRI data in

59.

60. 61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

95

squamous cell carcinoma of the head and neck. J Magn Reson Imaging 26(6):1607–1617, 2007. Kiselev VG: Transverse relaxation effect of MRI contrast agents: A crucial issue for quantitative measurements of cerebral perfusion.   J Magn Reson Imaging 22(6):693–696, 2005. Kiselev VG, Strecker R, Ziyeh S, et al: Vessel size imaging in humans. Magn Reson Med 53(3):553–563, 2005. Kjolby BF, Ostergaard L, Kiselev VG: Theoretical model of intravascular paramagnetic tracers effect on tissue relaxation. Magn Reson Med 56(1):187–197, 2006. Knutsson L, Stahlberg F: Wirestam R. Absolute quantification of perfusion using dynamic susceptibility contrast MRI: Pitfalls and possibilities. MAGMA 23(1):1–21, 2010. Kotys MS, Akbudak E, Markham J, et al: Precision, signal-to-noise ratio, and dose optimization of magnitude and phase arterial input functions in dynamic susceptibility contrast MRI. J Magn Reson Imaging 25(3):598–611, 2007. Kovar DA, Lewis M, Karczmar GS: A new method for imaging perfusion and contrast extraction fraction: Input functions derived from reference tissues. J Magn Reson Imaging 8(5):1126–1134, 1998. Larsson HB, Stubgaard M, Frederiksen JL, et al: Quantitation of blood-brain barrier defect by magnetic resonance imaging and gadolinium-DTPA in patients with multiple sclerosis and brain tumors. Magn Reson Med 16(1):117–131, 1990. Lavini C, de Jonge MC, van de Sande MG, et al: Pixel-by-pixel analysis of DCE MRI curve patterns and an illustration of its application to the imaging of the musculoskeletal system. Magn Reson Imaging 25(5):604– 612, 2007. Leach MO, Brindle KM, Evelhoch JL, et al: The assessment of antiangiogenic and antivascular therapies in early-stage clinical trials using magnetic resonance imaging: issues and recommendations. Br J Cancer 92(9):1599–1610, 2005. Leach MO, Morgan B, Tofts PS, et al: Imaging vascular function for early stage clinical trials using dynamic contrast-enhanced magnetic resonance imaging. Eur Radiol 22(7):1451–1464, 2012. Li X, Huang W, Yankeelov TE, et al: Shutter-speed analysis of   contrast reagent bolus-tracking data: Preliminary observations in benign and malignant breast disease. Magn Reson Med 53(3):724–729, 2005. Li X, Rooney WD, Springer CS, Jr: A unified magnetic resonance imaging pharmacokinetic theory: Intravascular and extracellular contrast reagents. Magn Reson Med 54(6):1351–1359, 2005. Lin W, Guo J, Rosen MA, et al: Respiratory motion-compensated radial dynamic contrast-enhanced (DCE)-MRI of chest and abdominal lesions. Magn Reson Med 60(5):1135–1146, 2008. Lin W, Song HK: Improved signal spoiling in fast radial gradient-echo imaging: Applied to accurate T(1) mapping and flip angle correction. Magn Reson Med 62(5):1185–1194, 2009. Luypaert R, Ingrisch M, Sourbron S, et al: The Akaike information criterion in DCE-MRI: Does it improve the haemodynamic parameter estimates? Phys Med Biol 57(11):3609–3628, 2012. Luypaert R, Sourbron S, de Mey J: Validity of perfusion parameters obtained using the modified Tofts model: A simulation study. Magn Reson Med 65(5):1491–1497, 2011. Marstrand JR, Rostrup E, Rosenbaum S, et al: Cerebral hemodynamic changes measured by gradient-echo or spin-echo bolus tracking and its correlation to changes in ICA blood flow measured by phase-mapping MRI. J Magn Reson Imaging 14(4):391–400, 2001. Mills SJ, Patankar TA, Haroon HA, et al: Do cerebral blood volume and contrast transfer coefficient predict prognosis in human glioma? AJNR Am J Neuroradiol 27(4):853–858, 2006. Moate PJ, Dougherty L, Schnall MD, et al: A modified logistic model to describe gadolinium kinetics in breast tumors. Magn Reson Imaging 22(4):467–473, 2004. Mouraviev V, Venkatraman TN, Tovmasyan A, et al: Mn-porphyrins   as novel molecular magnetic resonance imaging contrast agents.   J Endourol 26(11):1420–1424, 2011.

http://radiologyebook.com

96

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

79. Mouridsen K, Friston K, Hjort N, et al: Bayesian estimation of cerebral perfusion using a physiological model of microvasculature. Neuroimage 33(2):570–579, 2006. 80. Newbould RD, Skare ST, Jochimsen TH, et al: Perfusion mapping with multiecho multishot parallel imaging EPI. Magn Reson Med 58(1):70– 81, 2007. 81. Newman GC, Hospod FE, Patlak CS, et al: Experimental estimates of the constants relating signal change to contrast concentration for cerebral blood volume by T2* MRI. J Cereb Blood Flow Metab 26(6):760–770, 2006. 82. O’Connor JP, Jackson A, Parker GJ, et al: DCE-MRI biomarkers in the clinical evaluation of antiangiogenic and vascular disrupting agents.   Br J Cancer 96(2):189–195, 2007. 83. Ogg RJ, Kingsley PB: Optimized precision of inversion-recovery T1 measurements for constrained scan time. Magn Reson Med 51(3):625– 630, 2004. 84. Ostergaard L, Chesler DA, Weisskoff RM, et al: Modeling cerebral blood flow and flow heterogeneity from magnetic resonance residue data.   J Cereb Blood Flow Metab 19(6):690–699, 1999. 85. Ostergaard L, Weisskoff RM, Chesler DA, et al: High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. Part I: Mathematical approach and statistical analysis. Magn Reson Med 36(5):715–725, 1996. 86. Parker GJ: Monitoring contrast agent kinetics using dynamic MRI: Quantitative and qualitative analysis. [Ph.D. Dissertation], University of London, London, UK, 1997. 87. Parker GJ, Roberts C, Macdonald A, et al: Experimentally-derived functional form for a population-averaged high-temporal-resolution arterial input function for dynamic contrast-enhanced MRI. Magn Reson Med 56(5):993–1000, 2006. 88. Paudyal R, Bagher-Ebadian H, Nagaraja TN, et al: Modeling of Look-Locker estimates of the magnetic resonance imaging estimate of longitudinal relaxation rate in tissue after contrast administration. Magn Reson Med 66(5):1432–1444, 2011. 89. Paulson ES, Schmainda KM: Comparison of dynamic susceptibilityweighted contrast-enhanced MR methods: Recommendations for measuring relative cerebral blood volume in brain tumors. Radiology 249(2):601–613, 2008. 90. Pedersen M, Klarhofer M, Christensen S, et al: Quantitative cerebral perfusion using the PRESTO acquisition scheme. J Magn Reson Imaging 20(6):930–940, 2004. 91. Perl W, Lassen NA, Effros RM: Matrix proof of flow, volume and mean transit time theorems for regional and compartmental systems. Bull Math Biol 37(6):573–588, 1975. 92. Perman WH, Gado MH, Larson KB, et al: Simultaneous MR acquisition of arterial and brain signal-time curves. Magn Reson Med 28(1):74–83, 1992. 93. Pettigrew RI, Avruch L, Dannels W, et al: Fast-field-echo MR imaging with Gd-DTPA: Physiologic evaluation of the kidney and liver. Radiology 160(2):561–563, 1986. 94. Pintaske J, Martirosian P, Graf H, et al: Relaxivity of gadopentetate dimeglumine (Magnevist), gadobutrol (Gadovist), and gadobenate dimeglumine (MultiHance) in human blood plasma at 0.2, 1.5, and 3 Tesla. Invest Radiol 41(3):213–221, 2006. 95. Quarles CC, Gochberg DF, Gore JC, et al: A theoretical framework to model DSC-MRI data acquired in the presence of contrast agent extravasation. Phys Med Biol 54(19):5749–5766, 2009. 96. Quarles CC, Ward BD, Schmainda KM: Improving the reliability of obtaining tumor hemodynamic parameters in the presence of contrast agent extravasation. Magn Reson Med 53(6):1307–1316, 2005. 97. Rajan Jain, Jayant Narang, Ali Syed Arbab, et al: Role of non-modelbased semi-quantitative indices obtained from DCE T1 MR perfusion in differentiating pseudo-progression from true-progression. NeuroOncol 13(Suppl 3):1, 2011. 98. Ramaswamy S, Schornack PA, Smelko AG, et al: Superparamagnetic iron oxide (SPIO) labeling efficiency and subsequent MRI tracking of native cell populations pertinent to pulmonary heart valve tissue engineering studies. NMR Biomed 25(3):410–417, 2012.

99. Reddick WE, Bhargava R, Taylor JS, et al: Dynamic contrast-enhanced MR imaging evaluation of osteosarcoma response to neoadjuvant chemotherapy. J Magn Reson Imagin 5(6):689–694, 1995. 100. Reddick WE, Langston JW, Meyer WH, et al: Discrete signal processing of dynamic contrast-enhanced MR imaging: Statistical validation and preliminary clinical application. J Magn Reson Imaging 4(3):397–404, 1994. 101. Reddick WE, Taylor JS, Fletcher BD: Dynamic MR imaging (DEMRI) of microcirculation in bone sarcoma. J Magn Reson Imaging 10(3):277– 285, 1999. 102. Roberts C, Little R, Watson Y, et al: The effect of blood inflow and B(1)-field inhomogeneity on measurement of the arterial input function in axial 3D spoiled gradient echo dynamic contrast-enhanced MRI. Magn Reson Med 65(1):108–119, 2011. 103. Rohrer M, Bauer H, Mintorovitch J, et al: Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol 40(11):715–724, 2005. 104. Runge VM, Clanton JA, Herzer WA, et al: Intravascular contrast agents suitable for magnetic resonance imaging. Radiology 153(1):171–176, 1984. 105. Sam AD 2nd, Morasch MD, Collins J, et al: Safety of gadolinium contrast angiography in patients with chronic renal insufficiency. J Vasc Surg 38(2):313–318, 2003. 106. Schabel MC, DiBella EV, Jensen RL, et al: A model-constrained Monte Carlo method for blind arterial input function estimation in dynamic contrast-enhanced MRI: II. In vivo results. Phys Med Biol 55(16):4807– 4823, 2010. 107. Schmiedeskamp H, Straka M, Newbould RD, et al: Combined spin-   and gradient-echo perfusion-weighted imaging. Magn Reson Med 68(1):30–40, 2012. 108. Shuter B, Tofts PS, Wang SC, et al: The relaxivity of Gd-EOB-DTPA and Gd-DTPA in liver and kidney of the Wistar rat. Magn Reson Imaging 14(3):243–253, 1996. 109. Simonsen CZ, Ostergaard L, Smith DF, et al: Comparison of gradientand spin-echo imaging: CBF, CBV, and MTT measurements by bolus tracking. J Magn Reson Imaging 12(3):411–416, 2000. 110. Smith MR, Lu H, Frayne R: Signal-to-noise ratio effects in quantitative cerebral perfusion using dynamic susceptibility contrast agents. Magn Reson Med 49(1):122–128, 2003. 111. Sourbron SP, Buckley DL: Classic models for dynamic contrastenhanced MRI. NMR Biomed 26(8):1004–1027, 2013. 112. Sourbron S, Heilmann M, Biffar A, et al: Bolus-tracking MRI with a simultaneous T1- and T2*-measurement. Magn Reson Med 62(3):672– 681, 2009. 113. Sourbron S, Ingrisch M, Siefert A, et al: Quantification of cerebral blood flow, cerebral blood volume, and blood-brain-barrier leakage with DCE-MRI. Magn Reson Med 62(1):205–217, 2009. 114. Sourbron S, Luypaert R, Van Schuerbeek P, et al: Choice of the regularization parameter for perfusion quantification with MRI. Phys Med Biol 49(14):3307–3324, 2004. 115. Speck O, Chang L, DeSilva NM, et al: Perfusion MRI of the human brain with dynamic susceptibility contrast: gradient-echo versus spin-echo techniques. J Magn Reson Imaging 12(3):381–387, 2000. 116. Thomsen HS: How to avoid CIN: Guidelines from the European Society of Urogenital Radiology. Nephrol Dial Transplant 20(Suppl 1): i18–i22, 2005. 117. Tofts PS: Modeling tracer kinetics in dynamic Gd-DTPA MR imaging.   J Magn Reson Imaging 7(1):91–101, 1997. 118. Tofts PS, Brix G, Buckley DL, et al: Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: Standardized quantities and symbols. J Magn Reson Imaging 10(3):223– 232, 1999. 119. Tofts PS, Kermode AG: Blood brain barrier permeability in multiple sclerosis using labelled DTPA with PET, CT and MRI. J Neurol Neurosurg Psychiatry 52(8):1019–1020, 1989. 120. Tofts PS, Kermode AG: Measurement of the blood-brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med 17(2):357–367, 1991.

http://radiologyebook.com

CHAPTER 5  Contrast-Enhanced Magnetic Resonance Imaging 121. Tombach B, Benner T, Reimer P, et al: Do highly concentrated gadolinium chelates improve MR brain perfusion imaging? Intraindividually controlled randomized crossover concentration comparison study of 0.5 versus 1.0 mol/L gadobutrol. Radiology 226(3):880–888, 2003. 122. van Gelderen P, Grandin C, Petrella JR, et al: Rapid three-dimensional MR imaging method for tracking a bolus of contrast agent through the brain. Radiology 216(2):603–608, 2000. 123. van Osch MJ, van der Grond J, Bakker CJ: Partial volume effects on arterial input functions: Shape and amplitude distortions and their correction. J Magn Reson Imaging 22(6):704–709, 2005. 124. van Osch MJ, Vonken EJ, Viergever MA, et al: Measuring the arterial input function with gradient echo sequences. Magn Reson Med 49(6):1067–1076, 2003. 125. van Osch MJ, Vonken EJ, Wu O, et al: Model of the human vasculature for studying the influence of contrast injection speed on cerebral perfusion MRI. Magn Reson Med 50(3):614–622, 2003. 126. Villringer A, Rosen BR, Belliveau JW, et al: Dynamic imaging with lanthanide chelates in normal brain: Contrast due to magnetic susceptibility effects. Magn Reson Med 6(2):164–174, 1988. 127. Vonken EJ, van Osch MJ, Bakker CJ, et al: Measurement of cerebral perfusion with dual-echo multi-slice quantitative dynamic susceptibility contrast MRI. J Magn Reson Imaging 10(2):109–117, 1999. 128. Wang D, Cantrell C, Spottiswoode B, et al: Slice accelerated EPI for dynamic-susceptibility contrast enhanced (DSC) MRI. Paper presented at: the 21st Annual Meeting of ISMRM2013; Salt Lake City, Utah. 129. Weisskoff RM, Zuo CS, Boxerman JL, et al: Microscopic susceptibility variation and transverse relaxation: Theory and experiment. Magn Reson Med 31(6):601–610, 1994. 130. Willats L, Calamante F: The 39 steps: Evading error and deciphering the secrets for accurate dynamic susceptibility contrast MRI. NMR Biomed 26(8):913–931, 2013. 131. Willats L, Christensen S, Ma HK, et al: Validating a local arterial input function method for improved perfusion quantification in stroke.   J Cereb Blood Flow Metab 31(11):2189–2198, 2011. 132. Willats L, Connelly A, Calamante F: Improved deconvolution of perfusion MRI data in the presence of bolus delay and dispersion. Magn Reson Med 56(1):146–156, 2006. 133. Willats L, Connelly A, Calamante F: Minimising the effects of bolus dispersion in bolus-tracking MRI. NMR Biomed 21(10):1126–1137, 2008.

97

134. Williams DS, Detre JA, Leigh JS, et al: Magnetic resonance imaging of perfusion using spin inversion of arterial water. Proc Natl Acad Sci U S A 89(1):212–216, 1992. 135. Wintermark M, Albers GW, Alexandrov AV, et al: Acute stroke imaging research roadmap. Stroke 39(5):1621–1628, 2008. 136. Wu O, Ostergaard L, Weisskoff RM, et al: Tracer arrival timinginsensitive technique for estimating flow in MR perfusion-weighted imaging using singular value decomposition with a block-circulant deconvolution matrix. Magn Reson Med 50(1):164–174, 2003. 137. Yang C, Karczmar GS, Medved M, et al: Estimating the arterial input function using two reference tissues in dynamic contrast-enhanced MRI studies: Fundamental concepts and simulations. Magn Reson Med 52(5):1110–1117, 2004. 138. Yang C, Karczmar GS, Medved M, et al: Multiple reference tissue method for contrast agent arterial input function estimation. Magn Reson Med 58(6):1266–1275, 2007. 139. Yankeelov TE, Luci JJ, Lepage M, et al: Quantitative pharmacokinetic analysis of DCE-MRI data without an arterial input function: A reference region model. Magn Reson Imaging 23(4):519–529, 2005. 140. Yankeelov TE, Rooney WD, Huang W, et al: Evidence for shutter-speed variation in CR bolus-tracking studies of human pathology. NMR Biomed 18(3):173–185, 2005. 141. Yankeelov TE, Rooney WD, Li X, et al: Variation of the relaxographic “shutter-speed” for transcytolemmal water exchange affects the CR bolus-tracking curve shape. Magn Reson Med 50(6):1151–1169, 2003. 142. Yarnykh VL: Actual flip-angle imaging in the pulsed steady state: A method for rapid three-dimensional mapping of the transmitted radiofrequency field. Magn Reson Med 57(1):192–200, 2007. 143. Yarnykh VL: Optimal radiofrequency and gradient spoiling for improved accuracy of T1 and B1 measurements using fast steady-state techniques. Magn Reson Med 63(6):1610–1626, 2010. 144. Yeung WT, Lee TY, Del Maestro RF, et al: In vivo CT measurement of blood-brain transfer constant of iopamidol in human brain tumors.   J Neurooncol 14(2):177–187, 1992. 145. Zahra MA, Hollingsworth KG, Sala E, et al: Dynamic contrast-enhanced MRI as a predictor of tumour response to radiotherapy. Lancet Oncol 8(1):63–74, 2007.

http://radiologyebook.com

6  Magnetic Resonance Imaging in the Pediatric Patient Aaryani Tipirneni-Sajja, Sue C. Kaste, and Claudia M. Hillenbrand

trast, techniques to minimize motion artifacts, standard imaging protocols, and emerging techniques in pediatric MRI.

INTRODUCTION The demand for pediatric magnetic resonance imaging (MRI) is steadily increasing because MRI is radiation free, has exquisite soft tissue contrast, and provides quantitative structural, functional, and vascular information for diagnosing a wide spectrum of diseases. Furthermore, the advent of genomics has led to improvements in understanding the genetic basis of a variety of heritable disorders associated with childhood neoplasms, so serial laboratory and radiologic screening protocols in patients with cancer predispositions (e.g., BeckwithWiedemann syndrome, von Hippel-Lindau disease, Li-Fraumeni syndrome) are established.100,137 Computed tomography (CT) and whole-body 18F-fluorodeoxyglucose positron emission tomography/ CT (FDG-PET/CT) are common surveillance modalities. However, these modalities expose children to ionizing radiation, and cumulative radiation doses from these relatively low-dose individual imaging studies may increase their lifetime cancer risk.26,121 A retrospective cohort analysis focusing on radiation exposure from CT scans in 178,604 pediatric patients found children with cumulative doses of over 50 mGy to the red bone marrow and 60 mGy to the brain (≈2-3 head CT scans) to be about three times more susceptible to developing radiation-induced leukemia and brain cancer, respectively.111 In addition, given the possible radiation sensitivity of patients with cancer syndromes such as Li-Fraumeni syndrome or ataxia-telangiectasia, localized and whole-body MRI may be a promising alternative for cancer prescreening and surveillance in children,12,125,137 thereby making every reasonable effort to avoid or limit ionizing radiation as far below dose limits as practical (ALARA [“as low as reasonably achievable”]).103 Although MRI has been extensively used in adult imaging for decades, MR knowledge and methodology are not directly transferable to pediatric imaging, because “children are not just small adults.”44,55 There are both practical and technical challenges when imaging infants and children. These include anatomic differences, developmental changes, differences in physiologic parameters, different spectrum of diseases, challenges in holding still for long periods of time, patient cooperation, availability of pediatric equipment (e.g., smaller-sized radiofrequency [RF] coils, MR-compatible incubators), and specific absorption rate (SAR) issues. Also, MRI requests may be nonspecific because children are poor reporters and often describe their symptoms very vaguely. All these factors could render pediatric MRI exams potentially complicated, costly, and lengthy. Nonetheless these problems can be overcome by proper patient preparation and optimization of techniques to obtain diagnostic image quality. This chapter gives an overview of pediatric MRI focused on techniques and parameters that differ from adult imaging. Topics include preparing and coaching a pediatric patient, special equipment, addressing safety concerns, optimizing signal-to-noise ratio (SNR) and con-

98

PATIENT PREPARATION Adequate patient preparation before the procedure is very important in reducing a child’s anxiety and promoting his/her cooperation during the MR examination. Developing a trusting relationship with the parent or guardian is key to successful pediatric imaging. Preprocedural preparation for the pediatric patient begins with a brief medical history describing the problems that warrant the MR examination. The patient or parent of a minor patient fills out a screening form51,133 indicating whether the patient has had any surgical interventions, including implanted prostheses, or if the child has experienced any accidents that would have resulted in embedded metal. If the patient has an implant, its MR compatibility must be determined before  the MR examination; “the list” (www.mrisafety.com) is a convenient resource for this purpose. Furthermore, the staff member conducting the screening should actively gather information from both parent and child, because the child could have items stuck in the ear or nose, unknown to the parent. For patients who may need sedation, the parent is also asked whether the child has ever experienced any symptoms that relate to certain conditions that may increase the risks of sedation. A history of upper airway obstruction, chronic liver disease, mental retardation, and congenital heart disease are some clinical conditions that may indicate an increased risk of sedation or risk of sedation failure. If a patient has an acute illness (e.g., unexplained fever, upper respiratory tract infection), an elective or nonurgent MR examination should be temporarily postponed. To minimize the risks of aspiration, pediatric patients should abstain from eating and drinking for a designated time period prior to sedation. According to NPO (nothing by mouth) guidelines for intravenous (IV) sedation, patients can have clear liquids up until 2 hours, breast milk up until 4 hours, infant formula up until 6 hours, and solid food and any other liquids up until 8 hours before the procedure.5 For oral sedation, patients typically can have solid food up until 4 hours and clear liquids up until 2 hours before the procedure.

Sedation Children referred for MRI often feel anxious or distressed about the unknown, unfamiliar surroundings and long examination times (on average, ≈ 45 min), and they may also be biased by their parents’ anxiety. Children who have had an MRI before may still be afraid of the large and noisy machine, confined space, the need to lie still, discomfort from surface coils, and the possibility of IV contrast injection. Given these factors, many children are unable to cooperate and require sedation. To successfully complete the study, sedation is generally

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient

A

99

B FIG 6-1  Atelectasis and sedation. Chest MRI in a sedated 2-year-old girl with neuroblastoma. A, Precontrast axial T1 BLADE image at the beginning of the scan showing no atelectasis. B, Postcontrast axial T1 BLADE FS image acquired 20 minutes later showing significant atelectasis in the left lung.

recommended for children younger than age 6 to 8 years and for older children with behavioral problems.9,84 However, establishing definitive age limits for sedation is not recommended; the need for sedation should be decided on an individual basis.

Sedation Agents.  The American Academy of Pediatrics (AAP) has published guidelines for sedation and monitoring of children undergoing imaging studies.5 The most common drugs used for sedation in pediatric MRI are propofol, dexmedetomidine, chloral hydrate, and pentobarbital.20,131 Propofol is often preferred for pediatric sedation because it is effective, has short induction (2 minutes) and recovery times (8 minutes), can easily be titrated to the required level of sedation, and has low incidence of complications.36,89 Typical propofol doses for sedation are 2 to 5 mg/kg/h IV. Dexmedetomidine is an alternative sedative for MRI with less respiratory side effects compared to propofol.90 It is administered at a loading dose of 2 to 3 µg/kg over 10 minutes, followed by a 1 to 2 µg/kg/h maintenance infusion for sedation. However, dexmedetomidine may have side effects such as bradycardia and hypotension and is therefore unsuitable for patients with compromised cardiac function.70,131 Chloral hydrate is another widely used sedative drug with barbiturate-like features. The dosing is 25 to 100 mg/kg, the onset time is 15 to 30 minutes, and duration of effect is 60 to 120 minutes. However, this drug has side effects such as nausea, vomiting, long recovery times, and postoperative agitation that lead to high failure rates for its use in MRI.131 Pentobarbital is another short-acting barbiturate and is given either orally or rectally at 3 to 6 mg/kg, with an onset time of 15 to 60 minutes; the duration of effect is 60 to 240 minutes.131 However, its use is restricted because it may be associated with respiratory and cardiovascular depression.52 Although a variety of sedative drugs are available, one should have a clear understanding of the pharmocokinetic effects of each agent when choosing the most appropriate drug for a particular patient and scenario. Institutional guidelines are usually in place for the selection and administration of effective and safe sedation in children during radiologic procedures.92,132 In pediatric MRI, establishing high sedation success rates largely depends on (1) filtering out children in whom sedation is unsafe or will likely be unsuccessful, (2) administering appropriate drugs and dosages, (3) continuously monitoring vital parameters, and (4) having experienced staff on site for immediate resuscitation.160

Monitoring During Sedation.  Every child undergoing sedation during an MR examination should be continuously monitored with MR-safe or MR-conditional equipment. Blood pressure, pulse rate, oxygen saturation, respiratory rate, electrocardiogram (ECG), and temperature are the parameters commonly monitored. Of these, the most reliable and immediate indicators of patient distress are increase in pulse rate and decrease in transcutaneous oxygen saturation,  because the most frequent complications of sedation are related to respiratory depression or airway obstruction.91 All patients should be monitored not only during the procedure but also following the procedure until postsedation recovery is complete.91

Imaging Without Sedation Sedation is frequently needed in pediatrics because some children are unable to lie immobile for the period of time needed to complete an MR study. Though often necessary, sedation is not without risk. Longer sedation is needed for MRI, owing to longer scan times that can put the child at risk for respiratory depression with oversedation.43 Patients on a ventilator or those who hypoventilate are at risk for atelectasis.88 Figure 6-1 shows images of a sedated patient who developed partial atelectasis in the left lung within 20 minutes. A potential issue with opacities caused by atelectasis is that they may obscure underlying pathology such as metastatic disease or sites of infection. Sedation services are also costly, requiring a team of pediatric anesthesiologists, intensivists, emergency department pediatricians, and pediatric nurses who work to ensure a safe procedure. Hence imaging without sedation could improve patient safety and reduce associated costs. Some techniques that have been shown to be effective in improving patient compliance and hence image quality without the need for sedation follow.

Feed and Sleep.  Feed and sleep is a technique in which sleep is induced in very young children through proper feeding and by providing warmth and comfort. This technique is routinely used in infants younger than 3 months of age undergoing MR examination (Fig. 6-2). The infant is made to fast for 4 hours and then fed until satisfied at least 30 minutes before the scan time.94 After the child goes to sleep or settles down, he/she is swaddled using infant sheets and secured in an immobilization device, sound protection is applied, and then the child is taken into the scan room escorted by a parent or a guardian.4 In one

http://radiologyebook.com

100

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging be able to cooperate during the examination. During the procedure the CLS can serve as liaison between technologists, patient, and parents, engage in nonstudy talking, distract the patient, provide emotional support, and so forth.3,95,144

Play Therapy and Mock MRI.  Play therapy is used to prepare chil-

FIG 6-2  Preparation of an infant for MRI. The infant is swaddled with blankets and covered with warm hot packs to encourage sleep. The ears are protected with neonatal earmuffs, and an MR-compatible pulse oximeter with fiberoptic probe is attached to the infant’s right foot to monitor the vital parameters. (Reprinted from van Wezel-Meijler G, et al: Magnetic resonance imaging of the brain in newborn infants: Practical aspects. Early Hum Dev 85:85–92, 2009, Fig. 2; ©Elsevier.)

study the feed-and-sleep approach with immobilization devices enabled 32 of 36 infants to successfully complete MRI scans with good image quality and without any sedation.65 Another study by Nordahl et al. showed good success rates in acquiring diagnostic-quality MRIs without the use of sedation in autistic children aged 2.5 to 4.5 years by scanning them in the evening during their normal bedtime.106

Communication and Positive Reinforcement.  Professionals who work with children can give reassurance and raise confidence in young patients to undergo the procedure without the need for sedation through verbal communication and body language. Positive reinforcement is also an effective way to keep children motivated to complete the scan successfully; reinforcement can be through verbal praise, certificates, trophies, and other prizes appropriate to the child’s age. Rewarding or praising pediatric patients for their efforts at each step is very helpful in gaining and maintaining cooperation and keeping them immobilized during the procedure.19

Child Life Specialist.  In recent years, child life specialists (CLSs) have become an integral part of the imaging team. Their knowledge and ability to relate to parents and to children of different ages and levels of maturity can mean the difference between successful and unsuccessful imaging, and the difference between the need for sedation and completing a study without sedation.144 CLSs are positioned to educate patients and parents about an imaging study prior to the child presenting for the examination, thereby allaying anxiety and enhancing patient cooperation.95,144 The CLS may intervene prior to the study by discussing the procedure with the patient and parent, introducing the family to the imaging team, building rapport, presenting coping skills, and using strategies such as play therapy, mock MRI, and distraction techniques in an attempt to increase the likelihood the child will

dren older than about age 4 years for MRI scans by explaining the procedure in a play-based manner with the help of medical toys, tape recordings, mockup scanners, or using real equipment.7 Showing pictures of children undergoing or having their favorite soft toys undergo an MR scan to explain what the examination entails makes a positive impact on how children will perceive the whole experience. Playing with a toy scanner model with sliding table and listening to an audio recording of the MR scanner noise (Fig. 6-3) help children become familiar with the procedure and develop coping strategies, allowing them to lie still and complete the scan.117 Recent studies have shown that sedation rates were considerably decreased in children exposed to a mock MRI scanner.29,40 The mock scanner room is well decorated and made appealing to children; it is usually in close proximity to the clinical MRI room (Fig. 6-4). In the mock scanner the child experiences the same sounds and visuals he/ she will experience during the real examination, such as lying on the table within the bore, having coils placed, hearing the sounds when sequences are played, and sometimes even rehearsing breath holds and developing coping strategies for IV contrast administration. This helps children cope with stressful conditions and makes the real experience less intimidating and frightening. A recent study showed that adequate play therapy in combination with a mock MRI session greatly improved patient cooperation and facilitated clinical MRI scans without sedation in 110 (84%) of 131 patients.38

Distraction.  Stress during MR procedures can be effectively reduced in pediatric patients by using distraction strategies. Simple physical contact between a parent and child (e.g., parent placing a hand on the child’s hand or foot) is very useful in allaying fear. Parents can bring pacifiers, MR-safe toys, books, or other comfort items that will distract the child.19 Making the physical environment more appealing by having a bright, colorful, and lively room with pictures of animals or characters can also distract older children and help them handle the procedure.122 However, young infants may prefer a warm and darkened environment that facilitates sleep.45 Auditory distraction techniques like playing music and storytelling are effective ways to calm and distract pediatric patients.104 Watching a video during the scan is an extremely valuable distraction tool. Older video system designs projected a video onto a screen in the scanner room, which was reflected by a mirror attached to the head coil (Fig. 6-5A). This setup worked well for brain scans but not for body scans, because as the table moved, the screen went out of focus. Also, this setup did not allay claustrophobic feelings some children experience within small-bore magnets. Modern commercial entertainment systems use MR-compatible video goggles that do not interfere with table movement and completely enclose the patient’s eyes, thereby minimizing claustrophobia (see Fig. 6-5B). It has been shown that the use of MR-compatible video goggles greatly reduces the need for sedation and provides a smoother workflow while obtaining good-quality images.66,75

Open MRI.  Traditional MRI scanners have long, narrow horizontal tubes into which the patient slides until the targeted body part is positioned in the center of the magnet. Such a closed environment can be daunting for pediatric patients and add to their anxieties and fears. Open MRI scanners have been designed to mitigate claustrophobic

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient

A

101

B FIG 6-3  Play therapy and model MRI scanner. A child life specialist explains the MRI procedure to the child by (A) using a small-scale model MRI with a movable table and (B) employing a play-based approach using toys, showing pictures of the MRI and the body, and playing MRI sounds and videos on an electronic device.

FIG 6-4  Mock MRI scanner. The mock MRI machine looks like a real MRI machine, but it does not have a magnet and does not take MRIs. (Courtesy Children’s MRI Centre, Royal Children’s Hospital, Melbourne.)

feelings. Several open magnet types are on the market. In the “pancake” design, the poles of the magnet are arranged horizontally, making the magnet three times larger and open on all sides. This allows parents to lie next to their child in the magnet, hold their child’s hand, and/or maintain eye contact during the scan, which can reduce anxiety for both child and parent and ultimately eliminate the need for sedation. Other designs include traditional magnets with shorter bore length and larger diameter, a so-called double-doughnut design with an open space in the middle, and a standup MRI configuration that allows patients to stand or sit. Historically, image quality was a problem with open MRIs because of their low magnetic field strength (0.2-0.35 tesla [T]). New open MRI technology has both open design and field strength of 1 T or higher, with images of the same quality as traditional scanners.14,41

SAFETY MR Screening Compared to adult scans, where typically only a technologist and the patient enter the magnet room, a relatively large number of people

(e.g., parents/guardians, child life personnel, sedation/anesthesia team) may enter for pediatric scans. If someone accidentally carries any ferrous object into the magnet room, it may become a projectile owing to the strong forces from the static magnetic field. Although nonferrous metals are not attracted by the magnetic field, they may experience Lenz’s forces either due to rapid switching of gradients  or when moved in the static magnetic field.72 Moreover the strong static field or the high RF power may cause electronic equipment to fail. RF fields inside the magnet bore can induce currents in conducting structures such as wires, electrode leads, or even sweaty skin, which in turn can induce eddy currents in the body that may cause burns in the patient. Hence everybody needs to be screened for medical conditions or for metal items inside or on the body, especially for pacemakers and implants, before entering the magnet room.72 Generally each individual entering the MR room must complete a screening form as part of the screening process, as previously mentioned.114 The American College of Radiology (ACR) has published extensive guidelines  on MR safety,51 and all MR personnel should be trained and follow MR safety guidelines to avoid any MR-specific risks. These riskmanagement and training requirements are now also demanded by The Joint Commission.139

Equipment All equipment used in the MR suite needs to be labeled according to the scheme proposed by the U.S. Food and Drug Administration (FDA)11,72,159: 1. MR safe: A square green label indicates that the item is “MR safe” and can be used in the MR environment and poses no known hazard. Items include nonmetallic, nonmagnetic, and nonconducting objects such as plastic or wood. 2. MR conditional: A triangular yellow label represents that the item is “MR conditional” and has been tested and demonstrated to pose no known hazards in a specific MR environment (e.g., maximum static magnetic field, gradients, SAR) under specified conditions of use. The conditions of use need to be known by  the person introducing the item into the scanner room and handling it. 3. MR unsafe: A round red label indicates that the items (i.e., ferromagnetic objects) are “MR unsafe” and known to pose a direct threat to persons and equipment in the scanner room and must not be brought into the scanner room.

http://radiologyebook.com

102

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B FIG 6-5  Distraction using MR-compatible entertainment solutions. A, A movie is projected onto an LCD screen in the scanner room, and the picture is reflected by a mirror attached to the head support or head coil. B, Modern commercial entertainment system with MR-compatible video goggles and advanced audio technology. Such entertainment systems help reduce anxiety and claustrophobia and may lead to a qualitatively better exam in nonsedated children. (A courtesy Children’s MRI Centre, Royal Children’s Hospital, Melbourne; B courtesy Resonance Technology Inc., Northridge, CA [www.mrivideo.com].)

Required non–MR-safe equipment like ventilators and monitoring devices can be used by placing them outside the magnet room and running nonmetallic connectors to the patient. Nowadays many MRI suites are provided with wave guides (wall openings) between the scan room and the control room, where fiber optic leads, plastic tubes, and other nonconducting leads or hoses can be fed through. Electrically conductive leads cannot be run through because they will act as antennas, compromise the RF shielding of the magnet room, lead to MRI artifacts, and also may be dangerous to the patient because of RF heating. Typical equipment necessary for neonates or sick children follows.10,69,146

User interactive display MRI incubator MR-safe skin temperature probes (×2)

MRI incubator trolley

Incubator.  MRI of critically ill preterm and term neonates is made feasible with the availability of MR-compatible incubators with integrated RF coils.21,49 In the neonatal intensive care unit (NICU), infants are transferred from their incubator to the MRI-compatible incubator (Fig. 6-6) with all necessary equipment (IV pumps, ventilators, monitors) connected; earmuffs and RF coils are also attached. They are  then transported to the MR room and scanned within the incubator. The MR-compatible incubator not only offers a safe humidity- and temperature-controlled environment to the infant at all times, but also optimizes the workflow because there is no additional transfer of the infant necessary when entering and leaving the magnet room. If an MR-compatible incubator is not available, a standard incubator can be used to transport the infant from NICU to the room adjacent to the scanner room. However, the infant will need to be taken out of the incubator for scanning. Most recently, Cincinnati Children’s Hospital pioneered an infant-sized MRI scanner (Fig. 6-7) installed in the NICU141,140; by avoiding transportation to the radiology department, a safer way of performing MRI in critically ill and premature newborns is provided.

Ventilator.  Patients being ventilated should be transferred to an MR-compatible ventilator with MR-conditional gas cylinders before undergoing the examination. In situations where MR-compatible ventilators are not available, standard ventilators can be used. However, they need to remain outside the magnet room and the extended hoses

Battery backup

FIG 6-6  MR-compatible incubator. Neonate Imaging Sub-System consists of a modular MRI incubator, nonmagnetic trolley with aluminum air/oxygen tanks, backup battery power supply, interactive touchscreen graphics user interface, and custom RF coils (not shown) to enable safe transport and effective MRI of preterm and term newborns and small babies. (Courtesy SREE Medical Systems, Cleveland, OH [www.sreemedical.com].)

are then guided into the scanner room; personnel must ensure that the ventilator is not accidentally transferred into the room.146

Intravenous Pumps.  Only absolutely necessary IV solutions should be administered during MRI. The IV pumps should be replaced with MR-compatible pumps if available. Otherwise, regular IV pumps can

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient

103

include using a removable sound-insulating foam insert and electrodynamic headphones with embedded MiniMuff ear pads.39 Ear protection is not only important for patients but also for parents or staff members accompanying the patient in the magnet room. In recent years the availability of MR-compatible active noise-cancelling headphones that introduce an out-of-phase secondary acoustic wave form to cancel the sound produced by the gradients has become an effective approach for improving the safety of both patient and accompanying person.123

RF Exposure

FIG 6-7  Neonatal MRI scanner. A new infant MRI machine was developed for the neonatal intensive care unit at Cincinnati Children’s Hospital Medical Center. (Technical details on this dedicated scanner and applied neonatal MRI protocols are described in Tkach JA, et al: An MRI system for imaging neonates in the NICU: Initial feasibility study. Pediatr Radiol 42:1347–1356, 2012. Image courtesy Dr. Charles L. Dumoulin, Cincinnati Children’s Hospital.)

be used by placing them outside the magnet room and running them through extended lines, as with non–MR-compatible ventilators, following the same precautions.

Resuscitation Equipment.  An emergency cart should always be available and readily accessible when pediatric patients, especially sedated ones, are undergoing MR examination. The cart must contain necessary age- and size-appropriate drugs and equipment that can resuscitate an unconscious or a nonbreathing child.112 This includes catheters, laryngoscopes, oral airways, positive pressure oxygen delivery system, wall suction and oxygen, portable oxygen cylinders, and suction equipment. Because most of these items are not MR-compatible, the resuscitation cart should be kept outside the magnet room but in a room adjacent to the scanner for easy access. In an emergency, the patient should be taken into the adjacent room for resuscitation in accordance with ACR recommendations for MR safety.51

Ear Protection During MRI examination, switching of the gradient fields creates loud acoustic noise, typically above the safe level of 85 dB. Prolonged exposure to such loud noise can potentially cause hearing loss, especially in children, owing to their sensitive ears. Hence noise reduction is crucial when imaging pediatric patients. There are some active and passive measures that can be implemented to reduce the acoustic noise during the MRI scan.39 Active measures include modification of the imaging pulse sequences to reduce imaging gradient rise times and slew rates, and to soften sharp changes in the gradient pulses. Passive measures

During MRI scanning, the patient is exposed to RF fields. The amount of RF energy deposited into the patient tissue during imaging is known as the specific absorption rate (SAR), expressed in watts per kilogram (W/kg). SAR depends on several factors, such as the strength of the static magnetic field, type of RF coil used, exact pulse sequence (number and type of pulses, including flip angle and waveform), patient size, and body part being imaged.48,120 According to the FDA, the maximum allowable SAR is 3 W/kg (averaged over 10 minutes) for head imaging and 4 W/kg (averaged over 15 minutes) for whole-body imaging.58 Higher SAR levels increase the body temperature, causing patient heating. However, the actual increase in temperature due to RF exposure is dependent on the patient’s thermoregulatory system. Children are particularly sensitive to RF-induced heating because they have relatively immature systems for controlling body temperature, especially infants and premature neonates. Hence monitoring SAR is very important for pediatric MRI. MR manufacturers use built-in SAR models to estimate the predicted SAR for a given patient size and pulse sequence. However, because MRI scanners are primarily built for adults, adult SAR models may not be reliable for use in pediatric patients, especially infants and young children with their relatively small size and different morphology compared to adults.34 Kussman et al.77 reported that an infant developed fever (38°C) during cardiac MRI even though the predicted SAR value did not reach the SAR limit; this suggested that SAR limits may need to be revised for infants and young children. SAR increases with the square of the magnetic field strength secondary to an increase in the RF frequency applied. For instance, there is a fourfold increase in SAR at 3 T compared to 1.5 T using the same scanning parameters, which consequently may lead to further rise in body temperatures at higher field strengths. Hence implementing measures to reduce SAR is of utmost importance in pediatric MRI. Ways to reduce SAR10,32 are to increase repetition time (TR), decrease flip angle, reduce the number of slices, use parallel imaging techniques to decrease the number of RF pulses applied, intersperse gradient echo and spin echo sequences, minimize the use of saturation bands, and use special transmit coil designs that are restricted to the body part of interest, because SAR increases with coil size. Furthermore the room should be cool and the patient should be dressed in a loose hospital gown for better heat dissipation.32 Infants are usually wrapped or swaddled to keep them warm and comfortable in the scanner; however, this may further impair heat dissipation. Thus body temperature should be constantly monitored in infants younger than age 1 year94,146 (e.g., via MR-compatible fluoro-optic temperature probes attached to the abdomen).

TECHNIQUES AND APPLICATIONS Strategies to overcome practical difficulties in performing pediatric MRI have been addressed in this chapter, but there are also technical challenges when imaging pediatric patients: poor signal-to-noise ratio (SNR), contrast optimization, and motion artifacts. Described in the

http://radiologyebook.com

104

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C

FIG 6-8  Selected commercially available infant array coils. A, A neonate head birdcage coil designed to cover 95th percentile of newborns weighing up to 4.5 kg and 55 cm whole-body length. The coil is designed to accommodate intubated patients as well (a notch is present on the anterior section of head coil for tube clearance). B, A 16-channel integrated infant head/spine array (IHSA) coil suited for high-resolution head/spine imaging in infants. C, A 20-channel infant cocoon suited for brain, spine, cardiac, and torso imaging of 95th percentile of 6-month-old infants. (Courtesy SREE Medical Systems, Cleveland, OH [www.sreemedical.com].)

following sections are some techniques used to overcome these technical problems.

Optimizing SNR The four major factors that significantly contribute to the SNR in MRI are field strength, voxel size, acquisition time, and coil selection (see Chapter 3 for a detailed discussion). 1. Field strength: The higher the field strength, the more surplus spins align with the main magnetic field, participate in image formation, and contribute to the SNR. 3-T magnets have an approximately twofold higher SNR compared to 1.5-T magnets. However, scanning at higher field strength imposes challenges such as altered T1 contrast due to longer T1 times, shorter T2* decay due to magnetic field inhomogeneities, increased chemical shift artifacts, and higher SAR.127 2. Voxel size: SNR is directly proportional to voxel volume (i.e., SNR decreases with voxel size). In children, anatomic structures are significantly smaller than in adults; the size of the brain of an average-weight term newborn is approximately 25% that of an average adult.35,76 Therefore scans with higher spatial resolution (i.e., smaller voxel size) are needed to depict anatomic features in the same detail. Increasing the resolution decreases the voxel volume and hence the SNR. For example, if the resolution is doubled in all three spatial dimensions, an eightfold reduction in the voxel volume and SNR results. 3. Acquisition time: SNR is proportional to the square root of image acquisition time for a given pulse sequence, coil selection, and voxel size; if the voxel volume is halved, the acquisition time must be quadrupled to maintain the same SNR. Higher SNR can be achieved by increasing the number of averages, decreasing receiver bandwidth, reducing the parallel imaging factor,

and oversampling the k-space; all these measures lead to longer acquisition times. In pediatric imaging, longer acquisition times may, however, not always be feasible, owing to limited patient cooperation or time constraints due to sedation. 4. RF coils: Appropriate coil selection is another means to optimize the SNR in pediatric MRI. There are two important factors for boosting the SNR: first, selecting a coil that is appropriate in size for the anatomy being scanned (Fig. 6-8), and second, using multichannel coils with smaller elements. The size of the RF coil should be approximately 1.5 times larger than the area of interest. At many sites, infants are imaged using adult head or knee coils because of the lack of dedicated neonatal head or body coils.8,94,146 Often the size of adult volume coils is too big for infants and young children, which results in a suboptimal filling factor (i.e., the ratio of tissue volume to the volume of the coil). A coil that tightly fits the anatomy of interest has a higher filling factor and produces better SNR than one with a lower filling factor when scanning under identical conditions (Fig. 6-9). State-of-the-art array coil systems consist of up to 128 coils with separate receiver chains. Higher SNR can be achieved with such arrays because each coil element detects magnetization and noise from its immediate vicinity only (i.e., coil sensitivity area; see Chapter 3). Coils that are not contributing signal to the imaging volume should be turned off to reduce image noise. Overall, phased array coils allow acquisition of fewer signal averages, which shortens the scan time while maintaining high SNR and resolution.152

Optimization of Contrast Optimal image contrast can partly compensate for low SNR in pediatric MRI by using a combination of different image weightings and

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient

Neonate

6 mo

1 yro

4 yro

105

7 yro

A 1 yro

4 yro

7 yro

1000

32 Ch Adult

600

Birdcage

800

200

400

Signal-to-noise ratio

6 mo

32 Ch Pedi

Neonate

0

B FIG 6-9  Size-optimized 32-channel receiver array coils for pediatric brain imaging. A, Posterior coil segments of five constructed pediatric array coils. The coil formers are based on the 95th percentile MRI contours of corresponding-aged children and have been dilated to accommodate foam padding. The coil formers were 3D printed and enclosed in a plastic box. Mounted mirrors are used to project visual stimulus for research studies. B, SNR comparisons between sagittal images obtained from the sized-matched head phantoms using the pediatric brain array coils (first row), the 32-channel adult brain array coil (second row), and a CP birdcage coil (third row). Images show that the highest SNR gain occurs closest to the surface of the constructed array. In the “brain” center of the phantoms the SNR is only slightly improved. The superimposed ROI on the birdcage coil SNR maps correspond to the regions used in the average brain SNR measurement. The 7-year-old phantom was too big to fit into the pediatric birdcage coil. (From Keil B, et al: Size-optimized 32-channel brain arrays for 3T pediatric imaging. Magn Reson Med 66:1777–1787, 2011, Figs. 2 and 5. ©Wiley-Liss, Inc. Reprinted with permission of publisher.)

by using contrast agents.109 The spectrum of disease, the T1s, and the T2s for tissues and lesions are very different in children compared to adults, and thus parameters used for T1 and T2 weighting in adult imaging cannot be automatically replicated for pediatric imaging. In addition, changing one sequence parameter may necessitate changes in multiple other parameters. Hence familiarity with the basic sequence properties and their tradeoffs is necessary for the technologist performing MRI in the pediatric population.

with maturation there is conversion of the epiphyseal and apophyseal cartilage to bone, as well as conversion of hematopoietic marrow to fatty marrow.82 Knowledge of these developmental changes is imperative for accurate interpretation of studies. These developmental changes also cause different tissue contrast in children compared to adults when the same imaging parameters (TR, echo time [TE]) are used, and hence create challenges in optimization of the parameters for pediatric imaging. In general, age-adapted protocols should be used.71,124

Changing Contrast Due to Maturation.  In children the anatomy is

Native Contrast.  One of the main advantages of MRI over other

not only smaller than in adults, but there are normal changes in the appearance of many structures as children grow, develop, and mature. For instance, the water content in the body is 75% at birth and decreases to only 55% to 65% for an average adult.105 The developmental changes in children are more apparent in the brain, bone, and cartilage.4,16,50,82 In the brain, sulcation and myelination of white matter tracts occur rapidly during the first 2 to 3 years of life15 (Fig. 6-10). In the skeleton,

imaging modalities is the ability to produce different tissue contrasts (principally T1-, T2-, proton density [PD], and diffusion-weighted) by exploiting the inherent properties of different tissue types. Optimal weighting in MRI is based on the combination of sequence parameters (TR, TE, inversion time [TI], flip angle, b value) chosen to provide good contrast that can differentiate a lesion from its surrounding tissue. Selection of the parameters would be easy if the exact values of

http://radiologyebook.com

106

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

T1-W

T2-W

PD-W

FLAIR

1.1 yrs

1.4 yrs

2.3 yrs

3.7 yrs

FIG 6-10  White matter maturation during the first years of life. MRI images of a 1.1-year-old acute lymphoblastic leukemia patient at baseline and at multiple time points during treatment. As the brain develops, the images show temporal maturation of the white matter and a better differentiation between white matter and gray matter (apparent on terminal zones of myelination and in frontal lobe on T2-W, FLAIR, and PD-W images). Volumetric gray and white matter growth is clearly appreciated on T1-weighted imaging over the course of 2.5 years. (Courtesy Dr. Wilburn E. Reddick, St. Jude Children’s Research Hospital, Memphis.)

http://radiologyebook.com

107

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient T1 (for TR and flip angle), T2 (for TE), and apparent diffusion coefficient (for b value) for the tissue of interest are known. Unfortunately these values change because of the pathology, interindividual differences, and even intraindividual change over time (e.g., change in tumor characteristics over the course of treatment). One solution is to use a wider spectrum of weightings to obtain weakly and heavily weighted images,108 and another solution is to fuse images with

different weightings (e.g., a combination of T1, T2, and diffusion) using postprocessing software.153

Intravenous Contrast Agents.  Gadolinium (Gd)-based contrast agents greatly improve diagnostic image quality and provide clinically relevant information in addition to native MRI, as for adults. Table 6-1 gives an overview of the main applications for which contrast 

TABLE 6-1  Applications of Pediatric MRI by Body Region Body Region

Applications of MRI

Advantages of MRI

MRI Protocol

Brain and spine

– Tumors (e.g., ependymoma, medulloblastoma, cerebellar low-grade astrocytoma) – Congenital malformations – Demyelinating diseases – Neurodegenerative disease – Inflammatory diseases – Epilepsy – Pulmonary diseases of alveolar infiltration or exudation patterns (e.g., segmental pneumonia or bronchopneumonia, pulmonary edema) – Tumors – Interstitial pulmonary diseases, fibrotic processes – Lung malformations – Congenital malformations (e.g., shunts, fistulae, regurgitant valves)

Extensive experience of anatomic and functional characterization of CNS pathologies

Brain: axial T2-weighted, coronal FLAIR, and coronal and sagittal T1-weighted images Spine: sagittal, fast-spin echo T1- and T2-weighted sequences Gadolinium (Gd) enhancement used in suspected inflammation, tumors/ metastases, white matter disorders, neurocutaneous disorders T2-weighted (turbo-spin echo) T1-weighted spin echo or 3D gradient echo sequence after applying Gd contrast (if suspicion of abscess, assessment of fibrosis activity) Combination of cardiac and ventilator gating often required (use fast imaging technique)

Chest

Cardiovascular system

Abdomen

– – – – – –

Musculoskeletal system

– – – – –

– – – –

Genitourinary tract (urography)

– – – –

Acute abdomen Unexplained abdominal pain Appendicitis Inflammatory bowel disease (Crohn’s disease, ulcerative colitis) Motility disorders Congenital GI malformations (e.g., biliary atresia, cloacal malformations) GI tumors Pancreatitis Ovarian pathology Trauma Skeletal, congenital, and developmental disorders (e.g., hip dysplasia, Meyer’s dysplasia) Rheumatic diseases (e.g., juvenile spondyloarthropathies) Trauma (bone fracture, tendon, and muscle) Bone tumors (benign, malignant) Soft tissue masses (e.g., vascular malformations, cysts, fibromatous tumors, neurofibromas, soft tissue malignancies) Congenital anatomic abnormalities Vesicoureteric reflux Hydronephrosis Obstructive uropathy

Superior visualization of interstitial processes, inflammatory disease

Provision of 3D anatomic and hemodynamic information beyond echocardiography and catheterization Anatomic depiction of complete abdominal organ systems Enterography (oral contrast distention of bowel combined with intravenous Gd) provides increased sensitivity for bowel wall abnormalities

Breath-held, ECG-gated, balanced steadystate free precession (bSSFP) cine image Gd enhancement used in bSSFP and MRA

Versatile depiction of bone marrow, cartilage, joints, and soft tissues to identify and localize pathology

T1-weighted, T2-weighted, and proton density sequences (at least one combined with fat saturation), STIR sequence T1-weighted images performed with Gd contrast

Evolving technique for generating high-quality anatomic scans (kidneys, ureters, bladder) and renal function assessments (e.g., split renal function and drainage)

T2-weighted imaging (static-fluid MR urography) T1-weighted fat-suppressed postcontrast imaging (excretory urography)

Coronal T2-weighted or short tau inversion recovery (STIR) images in combination with axial T2-weighted and/or fat-suppressed (STIR) T2-weighted images Further sequences obtained according to underlying pathology Intravenous hyoscine or glucagon to reduce peristalsis Gd enhancement used in suspected inflammatory bowel disease

Continued

http://radiologyebook.com

108

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

TABLE 6-1  Applications of Pediatric MRI by Body Region—cont’d Body Region

Applications of MRI

Advantages of MRI

MRI Protocol

Infections

CNS – Bacterial intracranial infection (e.g., epidural and subdural empyema, meningitis, pyogenic abscess) – Spinal infection (e.g., spondylodiskitis, epidural abscess) – Viral meningoencephalitis (e.g., herpes simplex virus) – HIV Non-CNS – Musculoskeletal (e.g., osteomyelitis) – Gastrointestinal (e.g., cholangitis) – Vascular (e.g., vasculitis) – Stroke (arterial, venous, hemorrhagic) – Hypoxic-ischemic brain injury – Hereditary metabolic diseases (e.g., peroxisomal disorders, lysosomal storage disorders, disorders of amino acid and organic acid metabolism) – Brain malformations – Vascular malformations – Tumors – Multifocal lesions (e.g., metastases, storage disorders, soft tissue disorders, multifocal osteomyelitis) – Fever of unknown origin – Nonaccidental trauma

Sensitive and specific imaging, providing early diagnosis

T1- (pre- and post-Gd) and T2-weighted images Gd enhancement provides additional information for differential diagnosis

Depiction of small/subtle pathology

Axial T2-weighted turbo-spin echo, axial FLAIR, T2*-weighted gradient echo sequences, DWI, and sagittal T1-weighted acquisition Gd-enhanced T1-weighted images for inflammatory diseases or tumors

3D anatomic visualization for determining location and extent of lesions; functional/ quantitative capabilities

STIR or fat-suppressed T2 spin echo, DWI, fat-suppressed T1 spin echo, 3D spoiled gradient echo sequences in arterial or portovenous phase following Gd contrast, fat-suppressed T1 spin echo (post-Gd)

Metabolic disorders and malformations

Whole body

From Bhargava R, et al: Contrast-enhanced magnetic resonance imaging in pediatric patients: Review and recommendations for current practice. Magn Reson Insights 6:95–111, 2013.

enhancement in pediatric MRI is indicated. Dynamic contrastenhanced (DCE)-MRI assists in measuring important physiologic parameters such as perfusion, capillary permeability, or glomerular filtration rate (GFR) (see Chapter 5 for further reading). DCE-MRI is more challenging in children because the smaller vessels limit both the pressure with which a contrast bolus may be administered (i.e., slower injection rates, smaller-gauge catheters) and the volume of contrast used.6 Furthermore, the faster circulatory dynamics and smaller anatomy typically found in young children may pose an additional challenge in acquiring high spatiotemporal images for first-pass studies. However, recent studies demonstrated the use of combined parallel imaging and compressed sensing techniques to resolve the challenges of high spatiotemporal resolution and respiratory motion in pediatric body DCE-MRI.164,166 Details on FDA-approved Gd contrast agents and their recommended doses for use in pediatric MRI can be found in Table 5-1 in Chapter 5. In children aged 2 years and older, the pharmacokinetics of the currently available Gd agents are similar to adults.13,61,155 Thus the dosage can be given based on body weight as in adults, and no agedependent dose adjustment is required. More information on practical recommendations for the use of contrast media in pediatric MRI  (e.g., injection technique, injection rate, needle size to be used) is summarized by Bhargava et al.20 Risks such as nephrogenic systemic fibrosis (NSF) associated with using Gd chelates in children with reduced renal function are similar to those in adults.96,113 Stable chelates like

macrocyclic Gd agents are generally preferred for pediatric use, because they are considered to have low risk for NSF.20

Motion Artifacts Patient motion is a major concern in MRI because motion manifests as artifacts in MRIs, making the images hard to interpret and increasing the possibility of misdiagnosis (see Chapter 3). Especially in pediatric MRI, even minor motion artifacts in small patients can result in a nondiagnostic exam. Two types of motion cause artifacts on MRI images if not addressed. Gross motion (i.e., bulk patient movement) can be difficult to control in children. They might lack understanding of the scan’s importance and make voluntary movements. Sometimes their ability to control gross motion (e.g., limb movement) is limited. However, using good patient preparation strategies as explained earlier can help reduce gross motion artifacts. Involuntary physiologic motion due to cardiac pulsations, blood flow, respiratory motion, and gastrointestinal peristalsis can also lead to image artifacts. Compensating for this motion is very challenging in children because of their high blood flow and heart and respiratory rates; normal heart and respiratory rates in infants are 140/ min and 40/min, respectively, compared to only 60-100/min and 15-20/min in adults. Such high rates cause significant motion artifacts in chest and abdominal imaging that can compromise the quality of the study. Described below are some techniques to minimize motion artifacts.

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient Motion Compensation Techniques Breath-hold scans.  Breathing causes severe motion artifacts in thoracic and abdominal imaging. Breath-holding techniques during image acquisition can suppress respiratory motion artifacts. However, children younger than 8 years typically have difficulty holding their breath.63 In children who can do short breath holds, acquiring fewer slices in each breath hold and doing more breath holds for multislice sequences, or using faster sequences, are probably the best options for eliminating breathing artifacts. Faster sequences like half-Fourier acquisition single-shot turbo-spin echo (HASTE), balanced steadystate free precession (bSSFP), and T1-weighted three-dimensional (3D) gradient echo (GRE) can be performed within short breath-hold times. Gated sequences.  Gated sequences acquire or use images that are obtained during a well-defined window where the motion  is supposed to be minimal. For cardiac gating, this window is usually in end-diastole and is detected by an ECG monitoring device. For respiratory gating, the images are usually acquired during endexpiration as determined by external bellows or by using navigator methods that monitor the position of the diaphragm.46 Because the signal is only acquired during short windows, the imaging time  tends to be very long, usually at least twice the nongated scan time. Children’s rapid heartbeat and respiratory rates add to the gating challenge. Phase-encoding direction.  If gating is not feasible, one solution is to swap the frequency and phase-encoding directions, depending on the anatomy of interest or lesion location. This will shift the direction of artifacts but will not reduce them.108 Signal averaging.  Signal averaging refers to adding signal from multiple image acquisitions. This reduces motion artifacts by decreasing the relative contribution of motion-corrupted signal compared to the overall signal. Increasing the number of signal averages (NSA) also increases the SNR by a factor of the square root of NSA but at the expense of increased imaging time. Signal averaging is an alternative to gating in sedated and in young children who cannot hold their breath for chest and abdominal scans. Fat suppression.  Respiratory motion can induce ghosting artifacts that arise from high-signal-intensity subcutaneous abdominal fat and may obscure underlying pathologies. To avoid these motioninduced artifacts, fat suppression techniques such as short tau inversion recovery (STIR)28 and chemical shift selective suppression (CHESS)60 can be applied. However, incorporating fat suppression requires additional RF and gradient pulses and thus can increase imaging time and SAR. Radial k-space sampling.  In radial sampling, every readout samples the center of k-space with varying phase-encoding direction and fills the k-space in a nonlinear fashion. Multishot spin echo sequences with radial k-space filling have been given trade names like BLADE (Siemens Healthcare, Erlangen, Germany), PROPELLER (GE Medical Systems, Milwaukee, Wis.), and MultiVane (Philips, Best, The Netherlands).116 The main advantage of this technique is that motion artifacts are not propagated in the phase-encoding direction because it changes with each radial section. Instead the artifacts are dispersed throughout the radial dimensions and become less prominent.101 Downsides of radial k-space sampling include longer acquisition time, higher SAR, and streaking artifacts, which may not be acceptable for certain clinical applications. Parallel imaging.  Shortening the image acquisition time can reduce the risk of motion artifacts.101 Parallel imaging techniques reduce acquisition time and are now used routinely in clinical imaging. Parallel imaging exploits the unique spatial information from each

109

receiver coil element (channel) within a multichannel array to reduce the number of k-space measurements needed for an image. This is achieved by knowing the coil sensitivity maps, depicting the spatial profile of signal sensitivity across the imaging FOV for each coil element (see Chapter 3). The imaging speed can be further increased by combining parallel imaging with other undersampling strategies such as radial undersampling and compressed sensing.147 These techniques have a high potential in pediatric imaging because they reduce the sedation need, depth, and duration. The main disadvantage of these methods is an SNR reduction that at least corresponds to the reduction in acquisition time. Attenuation of peristalsis.  Peristalsis of the stomach and bowel causes noticeable artifact in the phase-encoding direction on T2-  weighted imaging. Attenuation of peristalsis is therefore important in abdominal and pelvic imaging. One way to reduce gastrointestinal tract motility is to ensure that the patient has been NPO for a long period. Another way is to use drugs such as glucagon or hyoscine-Nbutylbromide that cause an adequate antiperistaltic effect, thereby minimizing these artifacts.109

Imaging Protocols Conventional MRI sequences include T1-weighted, T2-weighted, fluid-attenuated inversion recovery (FLAIR), and diffusion-weighted imaging (DWI). Typically T1-weighted images are used for depiction of anatomy, and T2-weighted and FLAIR images are used for depicting tissue pathology. T2*-weighted GRE images are sensitive to susceptibility-induced phase changes and can therefore be used to differentiate between hemorrhage and calcification.162 Diffusion images reflect changes in the mobility of water molecules within tissues that help characterize pathologies. Table 6-1 lists a comprehensive number of applications for pediatric MRI and the standard imaging protocols used for various body regions.20

EMERGING TECHNIQUES IN PEDIATRIC MRI Whole-Body MRI State-of-the-art MRI equipment with high-performance gradients, parallel imaging, and sliding table capabilities allows whole-body (WB) MRI from head to toe to investigate multifocal diseases.83,128,129 Because children are generally smaller than adults, pediatric WB  MRI requires fewer stations and is usually more time efficient than in adults. The primary application of WB MRI is in oncology  (Fig. 6-11) for the evaluation of metastases and staging of a known primary tumor.125,130 Accurate staging is very crucial in children and adolescents with cancer, because the presence and location of tumors throughout the body determines efficient treatment regimens. Other nononcologic applications of pediatric WB MRI37 include osteonecrosis,18,31 multifocal osteomyelitis,74,149 myopathies,30 fever of unknown origin,86 body fat mapping,25,85 muscle volume quantification,85 vascular malformations, infectious or inflammatory diseases, nonaccidental trauma, virtual autopsy,1,138 and child abuse.47,115 Standard protocols used for WB MRI for various applications are listed in Table 6-1.20 Conventional WB MRI sequences detect primary tumor and metastases mainly based on morphologic changes but do not provide functional information as seen in PET imaging. However, recent studies showed that the introduction of rapid DWI MRI sequences improved tumor delineation by providing metabolic information  on tumor cellularity and necrosis.79 Initially it was thought that respiration motion limited the successful application of DWI to WB examinations, owing to breath holding or respiratory triggering, which lengthens acquisition time, a major concern in pediatric imaging.79

http://radiologyebook.com

110

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C

FIG 6-11  Whole-body MR and PET imaging of an 11-year-old girl with newly diagnosed metastatic Ewing’s sarcoma of the pelvis. Coronal STIR (A), T1-weighted (B), and grayscale inverted DWI (b 800 sec/mm2) (C) show a very large left pelvic and proximal thigh tumor with osseous involvement (solid arrows) as well as bilateral pulmonary metastases (dotted arrows).

However, Takahara et al.136 introduced a new technique of performing free breathing, strongly DWI (b = 1000 s/mm2) with fat-suppression pulses and thinner slices, which has increased the interest in WB DWI for the pediatric population.81 Recent studies comparing WB MRI including DWI with PET/CT showed promising results.87,135,143 However, the main limitations of DWI are differentiating between benign and certain malignant pathologies and in evaluating the structures close to the heart.78,81 In a recent WB DWI study, 48% of children showed asymmetrically restricted diffusion in the pelvic skeleton and lumbar spine as a normal finding, indicating caution has to be taken in its interpretation for pediatric cancer staging.110 Nevertheless the sensitivity and specificity of PET/ CT examinations also depend on tumor type, applied tracer, and tissue of interest. Therefore when selecting the appropriate staging modality, one should know the pitfalls. A study by Kwee et al.80 demonstrates that WB DWI and 18F-FDG PET play complementary roles in the characterization of lesions and in predicting response to therapy. Other investigators suggest that integrated PET/MRI would be beneficial for accurate diagnosis.154 Furthermore, in the context of pediatric imaging and ALARA, PET/MRI scanners are promising126 because patient radiation dose is reduced by more than 50% compared to PET/CT when

replacing CT imaging with MRI.22 In addition, radiotracer doses can be reduced because PET sampling can now be extended over the entire longer MR examination time.107

Arterial Spin Labeling Arterial spin labeling (ASL) is an emerging technology that enables perfusion imaging without contrast administration and is potentially a very useful technique in children.150 Until recently, measurement of perfusion involved serial imaging after injection of a paramagnetic contrast agent to dynamically track the passage of a labeled bolus through the vasculature. However, the use of an exogenous contrast agent limits its applicability in longitudinal studies and in patients with renal insufficiency because of the associated risk of nephrogenic systemic fibrosis.27 Especially in pediatrics, its application has been limited owing to reduced patient comfort and technical difficulties in administering the IV contrast agent. ASL is a noninvasive MRI technique that uses arterial blood water as an endogenous tracer to measure perfusion and hence is considered ideal for pediatric perfusion imaging. In addition, the faster blood flow and the shorter perfusion distance in children because of their relatively smaller size greatly improves the image quality of the ASL

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient

111

calcified from hemorrhagic lesions.161 However, familiarity with pediatric SWI findings and knowledge of potential pitfalls is essential to prevent misdiagnosis.24

800

600

400

RBF (mL/100 g/min)

Diffusion Tensor Imaging

200

0

FIG 6-12  Coronal renal blood flow (RBF) map in a 1.5-year-old girl undergoing treatment for neuroblastoma. Perfusion-weighted images of the kidneys were acquired with a FAIR-TrueFISP ASL sequence under free-breathing. The RBF map is overlaid on the first control image (M0).

perfusion maps compared to adults.150,151 The important applications of ASL in the pediatric population include distinguishing high-grade brain tumors from low-grade,163 investigating regions of hypo- and hyperperfusion in pediatric ischemic stroke,33,158 sickle cell disease,68 and moyamoya disease.56 The most recent study by Hales et al.62 shows the potential role of ASL in characterizing cerebral perfusion during normal maturation from late childhood into adulthood as the brain changes structurally and functionally during adolescence. Although most of the applications of ASL are in brain, recent studies show its applications outside of the brain, with kidney (Fig. 6-12) being one of the first and most studied sites.67,134,157 In recent years, ASL has become important for measurement of liver perfusion,73,119 but this application is still in its initial stages.

Susceptibility-Weighted Imaging Susceptibility-weighted imaging (SWI) is emerging as an important technique for diagnosing a wide variety of intracranial pathologies.59 SWI is a 3D gradient echo sequence with high in-plane spatial resolution that uses both phase and magnitude data to accentuate the magnetic properties of various substances such as blood, blood products, non-heme iron, and calcification.99 SWI is sensitive to the subtle differences in the paramagnetic susceptibility properties of oxygenated versus deoxygenated hemoglobin. Depending on the degree of deoxygenation, SWI shows veins with varying degrees of hypointensity, thereby enhancing the venous contrast. As in adults, SWI appears promising in children for the evaluation of hemorrhage, vascular malformations, traumatic brain injury, brain tumors, congenital infections, and neurodegenerative disorders.98,142,148 Recent studies reported that SWI offers higher detectability of microbleeds compared with T2*-weighted and conventional MR sequences.2,102 SWI has also been shown as a useful noncontrast imaging technique in evaluation and follow-up of pediatric stroke.23,97 SWI may also play a role in evaluating and monitoring hemodynamic alterations in patients with sickle cell disease156 (Fig. 6-13). Furthermore, the phase information from SWI can be used to differentiate

Diffusion tensor imaging (DTI) is an advanced application of DWI that describes the 3D distribution of water motion at each voxel in the MRI. Compared to conventional MRI, DTI provides better anatomic information and quantitative assessment of the white matter tracts. The main pediatric applications of DTI are to assess brain maturation118,165 and several pathologic conditions such as developmental delay,53 white-matter diseases, and brain injuries.93

Magnetic Resonance Spectroscopy MR spectroscopy (MRS) is a noninvasive MR technique used to identify and measure cellular metabolites. In pediatrics, MRS is used in the assessment of encephalopathic neonates,17,64 detection and diagnosis of some inborn errors of metabolism,42,57 developmental delay,54 and posttreatment assessment of intracranial tumors.145 Spectra obtained with short echo time (TR = 2 milliseconds, TE = 20-25 milliseconds) are used to identify the peaks of metabolites in various disease conditions, especially in patients with inborn errors of metabolism. Spectra obtained with long echo time (TR = 2 milliseconds, TE = 272-288 milliseconds) are used for accurate quantification of metabolites such as lactate and N-acetylaspartate (NAA) in injury states, and choline, creatinine, and NAA in tumors.

SUMMARY This chapter describes the practical problems, patient preparation procedures, safety issues, technical challenges and optimization, applications, and emerging techniques in pediatric MRI. The role of MRI in pediatrics is increasing tremendously because it avoids ionizing radiation, is noninvasive in nature, and has superior soft tissue contrast and resolution. However, MRI can be challenging in pediatrics with respect to sedation, MR-compatible equipment, dedicated pediatric RF coils, SAR, smaller anatomy, physiologic differences, and changing contrast. Adequate patient preparation before the procedure using strategies such as feed and sleep, play therapy, distraction, and reinforcement, with child life specialists on board, can greatly reduce the child’s anxiety and make the imaging successful even without sedation. Recent advances in hardware have increased the availability of MR-compatible equipment and dedicated pediatric coils, thereby making MRI more efficient in pediatrics. In contrast to adult MRI where standard imaging protocols are used, the sequences need to be optimized for pediatric MRI, taking into account the age, size, contrast differences, and the different spectrum of childhood diseases. Modern MRI techniques, while once the domain of research, are now available in clinical practice, and clinicians should be familiar with their current applications in pediatrics.

ACKNOWLEDGMENTS The authors thank Drs. Charles Dumoulin and Jean Tkach (Cincinnati Children’s Hospital), Michael Kean (Royal Children’s Hospital Melbourne), Dr. Gene Reddick (St. Jude Children’s Research Hospital), Olivia Shayesteh (Resonance Technology, Inc.), and Ravi Srinivasan (SREE Medical Systems/Advanced Imaging Research, Inc.) for providing information and figures. The authors thank Dr. Ralf Loeffler, Dr. Ruitian Song, Dr. Beth McCarville, Dr. Nathan Artz, Chris Goode, and Adam Winchell from St. Jude Children’s Research Hospital for helpful discussions.

http://radiologyebook.com

112

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging MIP

Vein

SCD

Control

SWI

FIG 6-13  SWI, mIP SWI, and segmented vein maps for a 15-year-old control subject (top row, normalized visible venous volume [NVVV] = 0.054) and a 12-year-old patient with sickle cell disease (SCD) (bottom row, NVVV = 0.0071). The mIP SWI of the SCD patient shows overall less venous contrast compared to the control subject; the black arrows point to the apparent difference in venous conspicuity in temporal and deep medullary areas. The loss in visible venous volume of the SCD patient can also be appreciated in the segmented vein map.

REFERENCES 1. Addison S, Arthurs OJ, Thayyil S: Post-mortem MRI as an alternative to non-forensic autopsy in foetuses and children: From research into clinical practice. Br J Radiol 87(1036):20130621, 2014. 2. Akter M, Hirai T, Hiai Y, et al: Detection of hemorrhagic hypointense foci in the brain on susceptibility-weighted imaging clinical and phantom studies. Acad Radiol 14(9):1011–1019, 2007. 3. Alexander M: Managing patient stress in pediatric radiology. Radiol Technol 83(6):549–560, 2012. 4. Almli CR, Rivkin MJ, McKinstry RC, et al: The NIH MRI study of normal brain development (Objective-2): Newborns, infants, toddlers, and preschoolers. Neuroimage 35(1):308–325, 2007. 5. American Academy of Pediatrics, American Academy of Pediatric Dentistry, Cote CJ, Wilson S, Work Group on S: Guidelines for

6.

7. 8. 9. 10.

monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures: An update. Pediatrics 118(6):2587–2602, 2006. American College of Radiology: ACR Manual on Contrast Media, ed 9, 2013. . Accessed June 22, 2015. Armstrong TS, Aitken HL: The developing role of play preparation in paediatric anaesthesia. Paediatr Anaesth 10(1):1–4, 2000. Arthur R: Magnetic resonance imaging in preterm infants. Pediatr Radiol 36(7):593–607, 2006. Arthurs OJ, Bjorkum AA: Safety in pediatric imaging: An update. Acta Radiol 54(9):983–990, 2013. Arthurs OJ, Edwards A, Austin T, et al: The challenges of neonatal magnetic resonance imaging. Pediatr Radiol 42(10):1183–1194, 2012.

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient 11. ASTM: Standard practice for marking medical devices and other items for safety in the magnetic resonance environment, vol F2503-05, West Conshohocken, PA, 2005, ASTM International. 12. Atkin KL, Ditchfield MR: The role of whole-body MRI in pediatric oncology. J Pediatr Hematol Oncol 36(5):342–352, 2014. 13. Baker JF, Kratz LC, Stevens GR, et al: Pharmacokinetics and safety of the MRI contrast agent gadoversetamide injection (OptiMARK) in healthy pediatric subjects. Invest Radiol 39(6):334–339, 2004. 14. Bangard C, Paszek J, Berg F, et al: MR imaging of claustrophobic patients in an open 1.0T scanner: Motion artifacts and patient acceptability compared with closed bore magnets. Eur J Radiol 64(1):152–157, 2007. 15. Barkovich AJ: Concepts of myelin and myelination in neuroradiology. AJNR Am J Neuroradiol 21(6):1099–1109, 2000. 16. Barkovich AJ: Pediatric neuroimaging, ed 4, Philadelphia, 2005, Lippincott Williams & Wilkins. 17. Barkovich AJ, Baranski K, Vigneron D, et al: Proton MR spectroscopy for the evaluation of brain injury in asphyxiated, term neonates. AJNR Am J Neuroradiol 20(8):1399–1405, 1999. 18. Beer M, Stenzel M, Girschick H, et al: Whole-body MR imaging in children with suspected osteonecrosis after intensive chemotherapy: Preliminary results. Rofo 180(3):238–245, 2008. 19. Beth Linder JM, Schiska AD: Imaging children: Tips and tricks. J Radiol Nurs 26(1):23–25, 2007. 20. Bhargava R, Hahn G, Hirsch W, et al: Contrast-enhanced magnetic resonance imaging in pediatric patients: Review and recommendations for current practice. Magn Reson Insights 6:95–111, 2013. 21. Bluml S, Friedlich P, Erberich S, et al: MR imaging of newborns by using an MR-compatible incubator with integrated radiofrequency coils: Initial experience. Radiology 231(2):594–601, 2004. 22. Boellaard R, O’Doherty MJ, Weber WA, et al: FDG PET and PET/CT: EANM procedure guidelines for tumour PET imaging: Version 1.0. Eur J Nucl Med Mol Imaging 37(1):181–200, 2010. 23. Bosemani T, Poretti A, Orman G, et al: Pediatric cerebral stroke: Susceptibility-weighted imaging may predict post-ischemic malignant edema. Neuroradiol J 26(5):579–583, 2013. 24. Bosemani T, Verschuuren SI, Poretti A, et al: Pitfalls in susceptibilityweighted imaging of the pediatric brain. J Neuroimaging 24(3):221–225, 2014. 25. Brennan DD, Whelan PF, Robinson K, et al: Rapid automated measurement of body fat distribution from whole-body MRI. AJR Am J Roentgenol 185(2):418–423, 2005. 26. Brenner DJ, Doll R, Goodhead DT, et al: Cancer risks attributable to low doses of ionizing radiation: Assessing what we really know. Proc Natl Acad Sci U S A 100(24):13761–13766, 2003. 27. Broome DR: Nephrogenic systemic fibrosis associated with gadoliniumbased contrast agents: A summary of the medical literature reporting. Eur J Radiol 66(2):230–234, 2008. 28. Bydder GM, Young IR: MR imaging: Clinical use of the inversion recovery sequence. J Comput Assist Tomogr 9(4):659–675, 1985. 29. Carter AJ, Greer ML, Gray SE, et al: Mock MRI: Reducing the need for anaesthesia in children. Pediatr Radiol 40(8):1368–1374, 2010. 30. Castro TC, Lederman H, Terreri MT, et al: Whole-body magnetic resonance imaging in the assessment of muscular involvement in juvenile dermatomyositis/polymyositis patients. Scand J Rheumatol 43(4):329–333, 2014. 31. Castro TC, Lederman H, Terreri MT, et al: Detection of multifocal osteonecrosis in an adolescent with dermatomyositis using whole-body MRI. Pediatr Radiol 40(9):1566–1568, 2010. 32. Chavhan GB, Babyn PS, Singh M, et al: MR imaging at 3.0 T in children: Technical differences, safety issues, and initial experience. Radiographics 29(5):1451–1466, 2009. 33. Chen J, Licht DJ, Smith SE, et al: Arterial spin labeling perfusion MRI in pediatric arterial ischemic stroke: Initial experiences. J Magn Reson Imaging 29(2):282–290, 2009. 34. Conil E, Hadjem A, Lacroux F, et al: Variability analysis of SAR from 20 MHz to 2.4 GHz for different adult and child models using finitedifference time-domain. Phys Med Biol 53(6):1511–1525, 2008.

113

35. Dagia C, Ditchfield M: 3T MRI in paediatrics: Challenges and clinical applications. Eur J Radiol 68(2):309–319, 2008. 36. Dalal PG, Murray D, Cox T, et al: Sedation and anesthesia protocols used for magnetic resonance imaging studies in infants: Provider   and pharmacologic considerations. Anesth Analg 103(4):863–868, 2006. 37. Darge K, Jaramillo D, Siegel MJ: Whole-body MRI in children: Current status and future applications. Eur J Radiol 68(2):289–298, 2008. 38. de Amorim e Silva CJ, Mackenzie A, Hallowell LM, et al: Practice MRI: Reducing the need for sedation and general anaesthesia in children undergoing MRI. Australas Radiol 50(4):319–323, 2006. 39. Dean DC, 3rd, Dirks H, O’Muircheartaigh J, et al: Pediatric neuroimaging using magnetic resonance imaging during non-sedated sleep. Pediatr Radiol 44(1):64–72, 2014. 40. de Bie HM, Boersma M, Wattjes MP, et al: Preparing children with a mock scanner training protocol results in high quality structural and functional MRI scans. Eur J Pediatr 169(9):1079–1085, 2010. 41. de Bucourt M, Streitparth F, Wonneberger U, et al: Obese patients in an open MRI at 1.0 Tesla: Image quality, diagnostic impact and feasibility. Eur Radiol 21(5):1004–1015, 2011. 42. deGrauw TJ, Salomons GS, Cecil KM, et al: Congenital creatine transporter deficiency. Neuropediatrics 33(5):232–238, 2002. 43. Dial S, Silver P, Bock K, et al: Pediatric sedation for procedures titrated to a desired degree of immobility results in unpredictable depth of sedation. Pediatr Emerg Care 17(6):414–420, 2001. 44. Editorial: Time to be serious about children’s health care. Lancet 358(9280):431, 2001. 45. Edwards AD, Arthurs OJ: Paediatric MRI under sedation: Is it necessary? What is the evidence for the alternatives? Pediatr Radiol 41(11):1353–1364, 2011. 46. Ehman RL, Felmlee JP: Adaptive technique for high-definition MR imaging of moving structures. Radiology 173:1989. 47. Eltermann T, Beer M, Girschick HJ: Magnetic resonance imaging in child abuse. J Child Neurol 22(2):170–175, 2007. 48. EN/IEC 60601-2-33—Medical Electrical Equipment—Part 2: Particular requirements for the safety of magnetic resonance equipment for medical diagnosis. IEC 60601: International Standard, 1995. 49. Erberich SG, Friedlich P, Seri I, et al: Functional MRI in neonates using neonatal head coil and MR compatible incubator. Neuroimage 20(2):683–692, 2003. 50. Evans AC, Brain Development Cooperative G: The NIH MRI study of normal brain development. Neuroimage 30(1):184–202, 2006. 51. Expert Panel on MRS, Kanal E, Barkovich AJ, et al: ACR guidance document on MR safe practices: 2013. J Magn Reson Imaging 37(3):501–530, 2013. 52. Field KJ, White WJ, Lang CM: Anaesthetic effects of chloral hydrate, pentobarbitone and urethane in adult male rats. Lab Anim 27(3):258– 269, 1993. 53. Filippi CG, Lin DD, Tsiouris AJ, et al: Diffusion-tensor MR imaging in children with developmental delay: Preliminary findings. Radiology 229(1):44–50, 2003. 54. Filippi CG, Ulug AM, Deck MD, et al: Developmental delay in children: Assessment with proton MR spectroscopy. AJNR Am J Neuroradiol 23(5):882–888, 2002. 55. Gillis J, Loughlan P: Not just small adults: The metaphors of paediatrics. Arch Dis Child 92(11):946–947, 2007. 56. Goetti R, O’Gorman R, Khan N, et al: Arterial spin labelling MRI for assessment of cerebral perfusion in children with moyamoya disease: Comparison with dynamic susceptibility contrast MRI. Neuroradiology 55(5):639–647, 2013. 57. Grodd W, Krageloh-Mann I, Klose U, et al: Metabolic and destructive brain disorders in children: Findings with localized proton MR spectroscopy. Radiology 181(1):173–181, 1991. 58. Guidance for industry and FDA staff. Criteria for significant risk investigations of magnetic 58 resonance diagnostic devices. Rockville, MD, 2014. 59. Haacke EM, Xu Y, Cheng YC, et al: Susceptibility weighted imaging (SWI). Magn Reson Med 52(3):612–618, 2004.

http://radiologyebook.com

114

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

60. Haase A, Frahm J, Hanicke W, et al: 1H NMR chemical shift selective (CHESS) imaging. Phys Med Biol 30(4):341–344, 1985. 61. Hahn G, Sorge I, Gruhn B, et al: Pharmacokinetics and safety of gadobutrol-enhanced magnetic resonance imaging in pediatric patients. Invest Radiol 44(12):776–783, 2009. 62. Hales PW, Kawadler JM, Aylett SE, et al: Arterial spin labeling characterization of cerebral perfusion during normal maturation from late childhood into adulthood: Normal ‘reference range’ values and their use in clinical studies. J Cereb Blood Flow Metab 34(5):776–784, 2014. 63. Hallowell LM, Stewart SE, de Amorim ESCT, et al: Reviewing the process of preparing children for MRI. Pediatr Radiol 38(3):271–279, 2008. 64. Hanrahan JD, Cox IJ, Azzopardi D, et al: Relation between proton magnetic resonance spectroscopy within 18 hours of birth asphyxia and neurodevelopment at 1 year of age. Dev Med Child Neurol 41(2):76–82, 1999. 65. Hansen SS: Feed-and-sleep: A non-invasive and safe alternative to general anaesthesia when imaging very young children. Radiographer 56:5–8, 2009. 66. Harned RK, 2nd, Strain JD: MRI-compatible audio/visual system: Impact on pediatric sedation. Pediatr Radiol 31(4):247–250, 2001. 67. He X, Aghayev A, Gumus S, et al: Estimation of single-kidney glomerular filtration rate without exogenous contrast agent. Magn Reson Med 71(1):257–266, 2014. 68. Helton KJ, Paydar A, Glass J, et al: Arterial spin-labeled perfusion combined with segmentation techniques to evaluate cerebral blood flow in white and gray matter of children with sickle cell anemia. Pediatr Blood Cancer 52(1):85–91, 2009. 69. Hillenbrand CM, Reykowski A: MR imaging of the newborn: A technical perspective. Magn Reson Imaging Clin N Am 20(1):63–79, 2012. 70. Ingersoll-Weng E, Manecke GR, Jr, Thistlethwaite PA: Dexmedetomidine and cardiac arrest. Anesthesiology 100(3):738–739, 2004. 71. Jaramillo D, Laor T: Pediatric musculoskeletal MRI: Basic principles to optimize success. Pediatr Radiol 38(4):379–391, 2008. 72. Kanal E, Barkovich AJ, Bell C, et al: ACR guidance document for safe MR practices: 2007. AJR Am J Roentgenol 188(6):1447–1474, 2007. 73. Katada Y, Shukuya T, Kawashima M, et al: A comparative study between arterial spin labeling and CT perfusion methods on hepatic portal venous flow. Jpn J Radiol 30(10):863–869, 2012. 74. Kennedy MT, Murphy T, Murphy M, et al: Whole body MRI in the diagnosis of chronic recurrent multifocal osteomyelitis. Orthop Traumatol Surg Res 98(4):461–464, 2012. 75. Khan JJ, Donnelly LF, Koch BL: A program to decrease the need for pediatric sedation for CT and MRI. Appl Radiol 36(4):30–33, 2007. 76. Kliegman MR, Behrman RE, Jenson HB, et al: Nelson textbook of pediatrics edition, 18th edition & Atlas of pediatric physical diagnosis, 5th edition package, ed 18, Philadelphia, PA, 2007, Saunders. 77. Kussman BD, Mulkern RV, Holzman RS: Iatrogenic hyperthermia during cardiac magnetic resonance imaging. Anesth Analg 99(4):1053– 1055, 2004. table of contents. 78. Kwee TC, Takahara T, Niwa T, et al: Influence of cardiac motion on diffusion-weighted magnetic resonance imaging of the liver. MAGMA 22(5):319–325, 2009. 79. Kwee TC, Takahara T, Ochiai R, et al: Diffusion-weighted whole-body imaging with background body signal suppression (DWIBS): Features and potential applications in oncology. Eur Radiol 18(9):1937–1952, 2008. 80. Kwee TC, Takahara T, Ochiai R, et al: Complementary roles of whole-body diffusion-weighted MRI and 18F-FDG PET: The state of the art and potential applications. J Nucl Med 51(10):1549–1558, 2010. 81. Kwee TC, Takahara T, Vermoolen MA, et al: Whole-body diffusionweighted imaging for staging malignant lymphoma in children. Pediatr Radiol 40(10):1592–1602, quiz 1720–1591, 2010. 82. Laor T, Jaramillo D: MR imaging insights into skeletal maturation: What is normal? Radiology 250(1):28–38, 2009.

83. Lauenstein TC, Semelka RC: Emerging techniques: Whole-body screening and staging with MRI. J Magn Reson Imaging 24(3):489–498, 2006. 84. Lawson G, Bray R: Sedation of children for magnetic resonance imaging. Arch Dis Child 82(2):150–153, 2000. 85. Lee S, Kuk JL: Changes in fat and skeletal muscle with exercise training in obese adolescents: Comparison of whole-body MRI and dual energy x-ray absorptiometry. Obesity 21(10):2063–2071, 2013. 86. Ley S, Ley-Zaporozhan J, Schenk JP: Whole-body MRI in the pediatric patient. Eur J Radiol 70(3):442–451, 2009. 87. Lichy MP, Aschoff P, Plathow C, et al: Tumor detection by diffusionweighted MRI and ADC-mapping—Initial clinical experiences in comparison to PET-CT. Invest Radiol 42(9):605–613, 2007. 88. Lutterbey G, Wattjes MP, Doerr D, et al: Atelectasis in children undergoing either propofol infusion or positive pressure ventilation anesthesia for magnetic resonance imaging. Paediatr Anaesth 17(2):121– 125, 2007. 89. Machata AM, Willschke H, Kabon B, et al: Propofol-based sedation regimen for infants and children undergoing ambulatory magnetic resonance imaging. Br J Anaesth 101(2):239–243, 2008. 90. Mahmoud M, Gunter J, Donnelly LF, et al: A comparison of dexmedetomidine with propofol for magnetic resonance imaging sleep studies in children. Anesth Analg 109(3):745–753, 2009. 91. Malviya S, Voepel-Lewis T, Eldevik OP, et al: Sedation and general anaesthesia in children undergoing MRI and CT: Adverse events and outcomes. Br J Anaesth 84(6):743–748, 2000. 92. Mason KP: The pediatric sedation service: Who is appropriate to sedate, which medications should I use, who should prescribe the drugs, how do I bill? Pediatr Radiol 38(Suppl 2):S218–S224, 2008. 93. Mathur AM, Neil JJ, Inder TE: Understanding brain injury and neurodevelopmental disabilities in the preterm infant: The evolving role of advanced magnetic resonance imaging. Semin Perinatol 34(1):57–66, 2010. 94. Mathur AM, Neil JJ, McKinstry RC, et al: Transport, monitoring, and successful brain MR imaging in unsedated neonates. Pediatr Radiol 38(3):260–264, 2008. 95. McGee K: The role of a child life specialist in a pediatric radiology department. Pediatr Radiol 33(7):467–474, 2003. 96. Mendichovszky IA, Marks SD, Simcock CM, et al: Gadolinium and nephrogenic systemic fibrosis: Time to tighten practice. Pediatr Radiol 38(5):489–496, quiz 602–483, 2008. 97. Meoded A, Poretti A, Benson JE, et al: Evaluation of the ischemic penumbra focusing on the venous drainage: The role of susceptibility weighted imaging (SWI) in pediatric ischemic cerebral stroke.   J Neuroradiol 41(2):108–116, 2014. 98. Meoded A, Poretti A, Northington FJ, et al: Susceptibility weighted imaging of the neonatal brain. Clin Radiol 67(8):793–801, 2012. 99. Mittal S, Wu Z, Neelavalli J, et al: Susceptibility-weighted imaging: Technical aspects and clinical applications, part 2. AJNR Am J Neuroradiol 30(2):232–252, 2009. 100. Monsalve J, Kapur J, Malkin D, et al: Imaging of cancer predisposition syndromes in children. Radiographics 31(1):263–280, 2011. 101. Morelli JN, Runge VM, Ai F, et al: An image-based approach to understanding the physics of MR artifacts. Radiographics 31(3):849– 866, 2011. 102. Mori N, Miki Y, Kikuta K, et al: Microbleeds in moyamoya disease: Susceptibility-weighted imaging versus T2*-weighted imaging at 3 tesla. Invest Radiol 43(8):574–579, 2008. 103. Munro L: Basics of radiation protection for everyday use: How to achieve ALARA: Working tips and guidelines, 2004. 104. Nguyen TN, Nilsson S, Hellstrom AL, et al: Music therapy to   reduce pain and anxiety in children with cancer undergoing lumbar puncture: A randomized clinical trial. J Pediatr Oncol Nurs 27(3): 146–155, 2010. 105. Nicolaidis S: Physiology of thirst. In Arnaud MJ, editor: Hydration throughout life, Montrouge, 1998, John Libbey Eurotext. 106. Nordahl CW, Simon TJ, Zierhut C, et al: Brief report: Methods for acquiring structural MRI data in very young children with autism

http://radiologyebook.com

CHAPTER 6  Magnetic Resonance Imaging in the Pediatric Patient

107.

108. 109. 110.

111.

112.

113. 114.

115. 116.

117.

118.

119.

120. 121. 122. 123.

124.

125. 126.

127. 128.

129.

130.

131.

without the use of sedation. J Autism Dev Disord 38(8):1581–1590, 2008. Oehmigen M, Ziegler S, Jakoby BW, et al: Radiotracer Dose Reduction in Integrated PET/MR: Implications from National Electrical Manufacturers Association Phantom Studies. J Nucl Med 55(8):1361–1367, 2014. Olsen OE: Practical body MRI—A paediatric perspective. Eur J Radiol 68(2):299–308, 2008. Olsen OE: MRI: How to perform a pediatric scan. Acta Radiol 54(9):991–997, 2013. Ording Muller LS, Avenarius D, Olsen OE: High signal in bone marrow at diffusion-weighted imaging with body background suppression (DWIBS) in healthy children. Pediatr Radiol 41(2):221–226, 2011. Pearce MS, Salotti JA, Little MP, et al: Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: A retrospective cohort study. Lancet 380(9840):499–505, 2012. Pediatrics AAO: Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures: An update. Pediatrics 118(6):2587–2602, 2006. Penfield JG: Nephrogenic systemic fibrosis and the use of gadoliniumbased contrast agents. Pediatr Nephrol 23(12):2121–2129, 2008. Pennock JM: Patient preparation, safety and hazards in imaging infants and children. Chapter 01. In Rutherford MA, editor: ed2002-2011, . Accessed June 22, 2015. Perez-Rossello JM, Connolly SA, Newton AW, et al: Whole-body MRI in suspected infant abuse. AJR Am J Roentgenol 195(3):744–750, 2010. Pipe JG: Motion correction with PROPELLER MRI: application to head motion and free-breathing cardiac imaging. Magn Reson Med 42(5):963–969, 1999. Pressdee D, May L, Eastman E, et al: The use of play therapy in the preparation of children undergoing MR imaging. Clin Radiol 52(12):945–947, 1997. Provenzale JM, Liang L, DeLong D, et al: Diffusion tensor imaging assessment of brain white matter maturation during the first postnatal year. AJR Am J Roentgenol 189(2):476–486, 2007. Ramasawmy R, Campbell-Washburn AE, Wells JA, et al: Hepatic arterial spin labelling MRI: An initial evaluation in mice. NMR Biomed 28(2):272–280, 2015. Reykowski A: RF Coil Safety. Joint Annual Meeting ISMRM-ESMRMB; May 19-25, 2007; Berlin (Germany). Robbins E: Radiation risks from imaging studies in children with cancer. Pediatr Blood Cancer 51(4):453–457, 2008. Roberton DM, South MJ: Practical paediatrics, London, 2007, Longman. Rudd BW, Lim TC, Lee J-H, et al: Evaluation of MRI compatible headphones for active noise control. Noise Control Engr J 61(1):41–49, 2013. Saunders DE, Thompson C, Gunny R, et al: Magnetic resonance imaging protocols for paediatric neuroradiology. Pediatr Radiol 37(8):789–797, 2007. Schaefer JF, Schlemmer HP: Total-body MR-imaging in oncology. Eur Radiol 16(9):2000–2015, 2006. Schafer JF, Gatidis S, Schmidt H, et al: Simultaneous whole-body PET/ MR imaging in comparison to PET/CT in pediatric oncology: Initial results. Radiology 273(1):220–231, 2014. Schick F: Whole-body MRI at high field: Technical limits and clinical potential. Eur Radiol 15(5):946–959, 2005. Schlemmer HP, Schafer J, Pfannenberg C, et al: Fast whole-body assessment of metastatic disease using a novel magnetic resonance imaging system: Initial experiences. Invest Radiol 40(2):64–71, 2005. Schmidt GP, Baur-Melnyk A, Herzog P, et al: High-resolution whole-body magnetic resonance image tumor staging with the use of parallel imaging versus dual-modality positron emission tomographycomputed tomography: Experience on a 32-channel system. Invest Radiol 40(12):743–753, 2005. Schmidt GP, Reiser MF, Baur-Melnyk A: Whole-body MRI for the staging and follow-up of patients with metastasis. Eur J Radiol 70(3):393–400, 2009. Schulte-Uentrop L, Goepfert MS: Anaesthesia or sedation for MRI in children. Curr Opin Anaesthesiol 23(4):513–517, 2010.

115

132. Shankar VR: Sedating children for radiological procedures: An intensivist’s perspective. Pediatr Radiol 38(Suppl 2):S213–S217, 2008. 133. Shellock FG: Reference manual for magnetic resonance safety, implants, and devices: 2015 edition, ed 2015, Los Angeles, CA, 2015, Biomedical Research Publishing Company. 134. Song R, Loeffler RB, Hillenbrand CM: Improved renal perfusion measurement with a dual navigator-gated Q2TIPS fair technique. Magn Reson Med 64(5):1352–1359, 2010. 135. Stecco A, Romano G, Negru M, et al: Whole-body diffusion-weighted magnetic resonance imaging in the staging of oncological patients: Comparison with positron emission tomography computed tomography (PET-CT) in a pilot study. Radiol Med (Torino) 114(1):1– 17, 2009. 136. Takahara T, Imai Y, Yamashita T, et al: Diffusion weighted whole body imaging with background body signal suppression (DWIBS): Technical improvement using free breathing, STIR and high resolution 3D display. Radiat Med 22(4):275–282, 2004. 137. Teplick A, Kowalski M, Biegel JA, et al: Educational paper: Screening in cancer predisposition syndromes: Guidelines for the general pediatrician. Eur J Pediatr 170(3):285–294, 2011. 138. Thayyil S, Sebire NJ, Chitty LS, et al: Post-mortem MRI versus conventional autopsy in fetuses and children: A prospective validation study. Lancet 382(9888):223–233, 2013. 139. The Joint Commission: The Joint Commission prepublication requirements, Oak Brook, IL, 2015, Joint Commission Resources. 140. Tkach JA, Hillman NH, Jobe AH, et al: An MRI system for imaging neonates in the NICU: Initial feasibility study. Pediatr Radiol 42(11):1347–1356, 2012. 141. Tkach JA, Merhar SL, Kline-Fath BM, et al: MRI in the neonatal ICU: Initial experience using a small-footprint 1.5-T system. AJR Am J Roentgenol 202(1):W95–W105, 2014. 142. Tong KA, Ashwal S, Obenaus A, et al: Susceptibility-weighted MR imaging: A review of clinical applications in children. AJNR Am J Neuroradiol 29(1):9–17, 2008. 143. Tsuji K, Kishi S, Tsuchida T, et al: Evaluation of staging and early response to chemotherapy with whole-body diffusion-weighted MRI   in malignant lymphoma patients: A comparison with FDG-PET/CT.   J Magn Reson Imaging 41(6):1601–1607, 2014. 144. Tyson ME, Bohl DD, Blickman JG: A randomized controlled trial: Child life services in pediatric imaging. Pediatr Radiol 44(11):1426–1432, 2014. 145. Tzika AA, Vajapeyam S, Barnes PD: Multivoxel proton MR   spectroscopy and hemodynamic MR imaging of childhood brain tumors: Preliminary observations. AJNR Am J Neuroradiol 18(2):203– 218, 1997. 146. van Wezel-Meijler G, Leijser LM, de Bruine FT, et al: Magnetic resonance imaging of the brain in newborn infants: Practical aspects. Early Hum Dev 85(2):85–92, 2009. 147. Vasanawala SS, Alley MT, Hargreaves BA, et al: Improved pediatric MR imaging with compressed sensing. Radiology 256(2):607–616, 2010. 148. Verschuuren S, Poretti A, Buerki S, et al: Susceptibility-weighted imaging of the pediatric brain. AJR Am J Roentgenol 198(5):W440– W449, 2012. 149. von Kalle T, Heim N, Hospach T, et al: Typical patterns of bone involvement in whole-body MRI of patients with chronic recurrent multifocal osteomyelitis (CRMO). Rofo 185(7):655–661, 2013. 150. Wang J, Licht DJ: Pediatric perfusion MR imaging using arterial spin labeling. Neuroimaging Clin N Am 16(1):149–167, ix, 2006. 151. Wang J, Licht DJ, Jahng GH, et al: Pediatric perfusion imaging using pulsed arterial spin labeling. J Magn Reson Imaging 18(4):404–413, 2003. 152. Wang J, Reykowski A, Dickas J: Calculation of the signal-to-noise ratio for simple surface coils and arrays of coils. IEEE Trans Biomed Eng 42(9):908–917, 1995. 153. Watson AT, Olsen OE: Fusion and subtraction post-processing in body MRI. Pediatr Radiol 45(2):273–282, 2015. 154. Wehrl HF, Wiehr S, Divine MR, et al: Preclinical and translational PET/ MR imaging. J Nucl Med 55(Suppl 2):11S–18S, 2014.

http://radiologyebook.com

116

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

155. Wible JH, Jr, Tata PN, Napoli AM, et al: Pharmacokinetics of gadoversetamide injection, a gadolinium-based contrast agent, in pediatric patients. Magn Reson Imaging 27(4):512–518, 2009. 156. Winchell AM, Taylor BA, Song R, et al: Evaluation of SWI in children with sickle cell disease. AJNR Am J Neuroradiol 35(5):1016–1021, 2014. 157. Winter JD, St Lawrence KS, Cheng HL: Quantification of renal perfusion: Comparison of arterial spin labeling and dynamic contrastenhanced MRI. J Magn Reson Imaging 34(3):608–615, 2011. 158. Wintermark P, Warfield SK: New insights in perinatal arterial ischemic stroke by assessing brain perfusion. Transl Stroke Res 3(2):255–262, 2012. 159. Woods TO: Guidance for Industry and FDA Staff: Establishing Safety and Compatibility of Passive Implants in the Magnetic Resonance (MR) Environment. In: FDA, ed2008. 160. Woodthorpe C, Trigg A, Alison G, et al: Nurse led sedation for paediatric MRI: Progress and issues. Paediatr Nurs 19(2):14–18, 2007.

161. Wu Z, Mittal S, Kish K, et al: Identification of calcification with MRI using susceptibility-weighted imaging: A case study. J Magn Reson Imaging 29(1):177–182, 2009. 162. Yamada N, Imakita S, Sakuma T, et al: Intracranial calcification on gradient-echo phase image: Depiction of diamagnetic susceptibility. Radiology 198(1):171–178, 1996. 163. Yeom KW, Mitchell LA, Lober RM, et al: Arterial spin-labeled perfusion of pediatric brain tumors. AJNR Am J Neuroradiol 35(2):395–401, 2014. 164. Yoruk U, Saranathan M, Loening AM, et al: High temporal resolution dynamic MRI and arterial input function for assessment of GFR in pediatric subjects. Magn Reson Med 2015 [Epub ahead of print]. 165. Yoshida S, Oishi K, Faria AV, et al: Diffusion tensor imaging of normal brain development. Pediatr Radiol 43(1):15–27, 2013. 166. Zhang T, Cheng JY, Potnick AG, et al: Fast pediatric 3D free-breathing abdominal dynamic contrast enhanced MRI with high spatiotemporal resolution. J Magn Reson Imaging 41(2):460–473, 2015.

http://radiologyebook.com

7  Tissue Characterization in Liver Imaging Using Advanced Magnetic Resonance Techniques Daniel T. Boll, Tobias Heye, Mustafa R. Bashir, and Elmar M. Merkle

INTRODUCTION In the evaluation of diffuse and focal liver disorders, magnetic resonance imaging (MRI) has been proven to provide superior characterization of disease processes and therapy outcomes compared to other modalities in the imaging armamentarium.60,80 The particular strength of hepatic MRI is based on its unique ability to synergistically extract information on various soft tissue components in combination with assessment of functional liver status.98 To achieve and maintain this performance, continuous optimization of existing MR sequences as well as recognition of current limitations with the goal to establish novel acquisition sequences and processing algorithms are of paramount importance.1,16,60 Accurate tissue characterization through detection and quantification of imaging biomarkers specific to liver disorders represent one possible approach to redefine hepatic MRI workflows and will be the main focus of this chapter.84 Traditionally, hepatic MRI sequences have been designed to either highlight locoregional tissue-inherent differences in relaxivity or, alternatively, emphasize focal contrast uptake characteristics compared to surrounding tissue; both phenomena are usually detected in a subjective fashion and reported qualitatively based on grayscale-encoded intensity levels of tissues under investigation.77 Novel hepatic MRI sequences that focus on detection and quantification of liver imaging biomarkers aim to assess specific pathophysiologic processes of the hepatobiliary systems and yield functional quantitative results.77 Although histopathologic analyses of hepatic parenchyma may yield a wide spectrum of answers regarding underlying hepatic diseases, complications from invasive sampling techniques, sampling errors, and inability to sample entire organ systems in association with associated costs limit the use of histologic sampling in many clinical scenarios, in particular if the underlying abnormality itself has a low complication or exacerbation rate.58,59,72 Of the existing functional MRI techniques seeking to extract reproducible and specific hepatic imaging biomarkers, this chapter focuses on MR diffusion-weighted imaging (DWI), dynamic perfusion imaging using various contrast agents, and MR relaxometry and elastography, all employed in specific disease scenarios that potentially benefit from biomarker quantitation unique to hepatic disease status or, alternatively, assessment of their longitudinal evolution over time.

DIFFUSION-WEIGHTED IMAGING Over more than 2 decades, diffusion-weighted MRI has evolved into a reliable and proven technique in the neuroradiology field, mainly for stroke detection and visualization of white matter tracts, employing diffusion tensor imaging (DTI). Despite being one of the earliest sequence techniques supplying anatomic and functional information

of an organ system, hepatic diffusion-weighted MRI still faces many challenges, and its clinical usefulness remains under investigation. In this section, basic concepts of DWI will be highlighted. Several signal models have been investigated for tissue characterization, including the apparent diffusion coefficient (ADC), intravoxel incoherent motion (IVIM), diffusion kurtosis, and stretched exponential models. The discussion will mainly focus on ADC as the model that has been explored to the greatest extent. Lastly, potential applications for hepatic tissue characterization using ADC will be discussed.

Technical Aspects After spins experience a radiofrequency (RF) excitation pulse, they undergo dephasing because of spin-spin interactions, T2 relaxation, and effects originating from external magnetic field inhomogeneities, T2* relaxation. Furthermore, a small component of this dephasing is related to diffusion of water molecules as a direct function of random thermally induced brownian movement. This effect in particular is enhanced in DWI, which usually employs a T2-weighted sequence with an additional symmetric pair of diffusion-sensitive gradients enveloping the refocusing pulse. Water molecules in a cellular environment leading to restricted diffusion undergo an initial phase shift as a result of the first diffusion gradient, which will be completely reversed by the second diffusion gradient without residual phase shift at time of signal readout. Water molecules in a cellular environment without substantial diffusion restriction, in contrast, acquire phase information from the first gradient, but because of their inherent motion, their signal will not be completely rephased by the second gradient, leading to a signal loss at the time of the signal readout. The degree of diffusion restriction is apparent as reduced grayscale intensity of the measured signal intensity at DWI. The sensitivity for detecting diffusion restriction using DWI sequences can be varied by changing the amplitude of the diffusionweighted gradient, the duration of the applied gradient, and the time interval between the paired gradients. The combination of gradient characteristics is expressed as a single numeric parameter referred to as the b value. The sensitivity for detecting restrictions in diffusion is varied by changing the b value, which is measured in s/mm2. Most frequently, the b value is varied by altering the gradient amplitude.12 In general, DWI is performed as a spin echo pulse sequence with echo-planar imaging (EPI) readout, with the additional set of diffusion gradients mentioned before. The main advantage of EPI sequences remains the combination of relatively short acquisition times and high signal-to-noise ratios (SNRs).79 EPI is, however, limited by spatial resolution and its high sensitivity to susceptibility variations. In particular, stronger magnetic fields may lead to pronounced susceptibility artifacts, inhomogeneities in the base magnetic field (B0), that result in shorter T2* relaxation times, as well as amplified inhomogeneities in

http://radiologyebook.com

117

118

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

the B1 transmit fields, with a higher characteristic frequency spectrum compared to less intense magnetic field strength.12,83 Mainly to improve image quality by increasing spatial resolution and decreasing the impact of susceptibility variations, improvements in DWI rely on implementation of parallel imaging techniques, half-Fourier acquisitions of single-shot turbo-spin echo (HASTE) sequence designs, and radial k-space acquisitions through periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER).77 Hepatic DWI can be performed as a free-breathing or breathhold technique, but the latter is usually limited by a maximum possible b value of less than 500 s/mm2 if a breathheld acquisition is desired. Additional cardiac gating, in particular, improves left hepatic lobe lesion characterization.25,53,86 For liver imaging, respiratory gating techniques and short echo times may be used because the liver undergoes considerable translation, rotational motion, and deformation during the respirational cycle.57 Owing to the T2 weighting of DWI sequences, certain focal hepatic lesions such as hemangiomas or cysts, which appear intrinsically hyperintense on standard T2-weighted imaging sequences, may also appear hyperintense on DWI. The hyperintense signal on DWI may,

however, not be a sign of restricted diffusion but rather represent the strong contribution of the T2 hyperintensity bleeding into the DWI sequence, because all signal intensities on DWI images are composed of mixed contributions from the intrinsic T2 properties of the tissue being examined, as well as spin density detected after defocusing and refocusing of the diffusion gradients. This phenomenon is known as the T2 shine-through effect.27 To differentiate T2 shine-through effects from true areas of restricted diffusion, apparent diffusion coefficient (ADC) maps can be calculated to remove signal dependence on underlying tissue T2. To be able to calculate ADC maps for each pixel in a DWI image series, at least two DWI datasets with differing b values need to be obtained. The ADC map represents grayscale-encoded ratios of logarithmized differences in signal intensities over the differences of b values of at least two sequential DWI sequences measured in mm2/s. Areas of restricted diffusion demonstrate hyperintense voxels on DWI, with corresponding hypointense voxels on the ADC map (Fig. 7-1). In particular for low b values, the effects of microscopic perfusion and diffusion are contributing to the grayscale-encoded intensity of individual voxels. Intravoxel incoherent motion (IVIM) analyses seek

A

B

C

D FIG 7-1  A 49-year-old man with metastatic renal cell cancer to the liver, as well as a hepatic cyst, seen on diffusion-weighted imaging at 3 T, showing a low b value (b 50) in A, a high b value (b 400) in B, and the apparent diffusion coefficient (ADC) map (C), as well as T2-weighted image (D) series. Note the T2 shinethrough effect on A, absent on B for the cyst (white) and the confirmation of true restricted diffusion on C for the metastasis (red).

http://radiologyebook.com

CHAPTER 7  Tissue Characterization in Liver Imaging Using Advanced MR Techniques to separate capillary microcirculation from molecular diffusion.44,45 IVIM uses multiple b values, and by employing theoretic models/ curve fitting, allows separation of perfusion from diffusion effects throughout the entire b-value spectrum. DWI-derived parameters such as free molecular-based (D) and perfusion-related (D*, f) diffusion parameters can be subsequently calculated. IVIM-DWI imaging is especially beneficial for patients with focal lesions for whom tissue vascularity and cellular treatment response need to be assessed individually.70 IVIM-DWI imaging allows differentiation among viable tumor tissue and tumor necrosis, as well as developing fibrosis, and can potentially be used as a surrogate marker of tumor response to treatment.15 The pathophysiologic features of many tumor tissues, such as distinct cytologic patterns and unique forms of clustering, are linked to tumor grading and resultant aggressiveness. These patterns and clusters affect molecular diffusion properties.95 The basic hypothesis of DWI is free and unrestricted gaussian diffusion of water protons; complex biologic tissues, however, often show a non-gaussian distribution of the water displacement profile when compartmentalization and/or tissue barriers are present. Therefore the measurement of a non-gaussian diffusion distribution (i.e., kurtosis) may allow quantitation of diffusional heterogeneity and improve characterization of water diffusion properties in the tumoral microenvironment.95

Clinical Applications Potential clinical use of DWI focuses on a wide range of hepatic disease scenarios, owing to the fact that by now DWI sequence sets have been routinely incorporated into standard MR liver imaging protocols. Diffuse liver disease generally is associated with lower ADC values than normal liver parenchyma. Mechanisms leading to diffusion restriction differ, however, based on the underlying cellular pathodynamic environment.11 In the disease evolution ranging from early hepatic fibrosis to end-stage cirrhosis, progressive deposition of collagen with lower unbound water content compared to normal hepatic parenchyma into the extracellular space occurs in combination with a decrease in hepatic perfusion.4,63,71 The resultant reduction in ADC values allows differentiation of normal liver parenchyma from end-stage fibrosis. However, ADC values seem to have a limited capability to distinguish early stages of fibrosis from advanced disease.78 In hepatic steatosis, on the other hand, the reduced ADC values result from reduced diffusivity of water protons associated with intra- and extracellular fat. Significant ADC reduction has been observed in patients with fatty infiltration that show hepatic fat fractions of greater than 5%.21,63 Lastly, in patients with diffuse iron deposition, as seen in hemochromatosis, the local susceptibility artifacts induced by high parenchymal iron concentration results directly in decreased ADC values. For all types of diffuse hepatic disease, the longitudinal evolution of ADC parameters over time appears more meaningful than defining disease-specific thresholds. In the detection of focal liver lesions, increased lesional conspicuity can be achieved through enhanced lesion-to-background tissue contrast by employing smaller diffusion-weighting values (b < 50-150 s/mm2), essentially nulling the intrahepatic vascular signal and resulting in black blood images. Lesion characterization requires higher b values (b > 500 s/mm2) to differentiate between underlying cystic or solid nature.50,87 Whereas a differentiation between cystic and solid focal lesions usually does not represent a diagnostic dilemma, differentiation between differing solid hepatic lesions is a greater diagnostic challenge owing to substantial overlap in morphologic appearances. Generally it has been shown that benign lesions show a higher ADC value than malignant ones; it seems that a threshold value of 1.7 × 10−3 mm2/s shows a fairly high sensitivity and specificity for differ-

119

entiating benign from malignant disease.13,32,69 Metastatic disease from mucinous primaries (e.g., ovarian or colorectal cancers) can, however, mimic benign lesions. Therefore quantitative ADC measurements performed for lesion characterization should always be used in concert with other quantitative markers as well as clinical background and temporal context. It has to be emphasized that none of the described biomarkers in this review article appear to be robust enough to allow deducing diagnoses without placing the extracted biomarker values into the appropriate clinical settings. Additionally, acquisition of serial b values can highlight changes in grayscale-encoded diffusivity, depending on diffusion weighting. Focal liver lesions generally appear hyperintense on b0 images, but benign lesions transition to iso- or hypointensity on b800 DWI, in contrast to malignant lesions, which remain hyperintense.10 For hepatocellular carcinomas (HCCs), histopathologic grading is one of the most important prognostic indicators and of paramount importance to determine treatment. It has been shown that threshold ADC values allowed differentiation of poorly differentiated tumors from moderate to poorly differentiated HCCs when a cutoff of 0.972 × 10−3 mm2/s was used.64,67 In general, lower mean ADC values indicate less differentiated and thereby worse histopathologic grades compared to well-differentiated HCCs. In addition, it is conceivable that cardiac motion artifacts may affect extraction of reproducible and reliable ADC values, in particular in the left hepatic lobe. Finally, DWI can, as part of the quantitative imaging arsenal, also be used in assessment of treatment response and disease prognostication. Although anatomic descriptive methods such as tumor size and enhancement patterns are not able to document evidence of treatment response or lack thereof for several weeks or months following treatment—in particular if novel cytostatic therapeutic agents are used—DWI may potentially provide an earlier assessment. Within hours of systemic chemotherapy, malignant tumors respond with an increase in cellular size, resulting in a reduction in ADC value as the intracellular/extracellular volume ratio per voxel increases. Eventually, cellular necrosis and a resultant increase in ADC are observed.20,24 This pattern of ADC change over time has been particularly documented in patients showing at least partial response to treatment.19 In contrast, focal hepatic metastases, which either show high ADC values before treatment or lack of ADC value increase early following treatment, may be characterized by a more adverse response to chemotherapy.52 For patients receiving focal therapy such as transcatheter arterial chemoembolization or radiofrequency ablation, analogously lower ADC values on posttreatment DWI correlated with tumor recurrence and thereby allowed differentiation of viable from necrotic portions of the treated lesion.47 Because pretreatment ADC values are composed of variable combinations of diffusion and perfusion effects, quantitated IVIM might be helpful in predicting the effectiveness of transcatheter arterial chemoembolization prior to treatment70 (Fig. 7-2).

DYNAMIC CONTRAST-ENHANCED MRI Similar to DWI, which evaluates microenvironments by emphasizing the degree of restriction in diffusion processes, perfusion and capillary permeability imaging focuses on microenvironmental changes in small blood vessels that are beyond the resolution of MRI for direct visualization.22,26,38,39,40,51 The basic principle of imaging these microcirculatory parameters is based on identification and quantification of signal changes as a tracer transits through tissue under investigation over time to determine blood flow, capillary permeability, sizes of intravascular and interstitial spaces, and potentially the differentiation between predominantly arterial or portal venous supply.89 This type of quantifiable microcirculatory information may potentially be helpful in the

http://radiologyebook.com

120

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C

D

E FIG 7-2  A 52-year-old man with hepatocellular carcinoma treated with external beam radiation the previous day on precontrast (A) and arterial phase postcontrast (B) fat-saturated T1-weighted imaging. Peripheral components of the hepatocellular carcinoma show necrotic transformation (white arrows); due to the high intrinsic T1 signal on precontrast imaging (SI 358), the assessment of the more solid-appearing components to differentiate between treatment change and residual tumor (red arrow) remains challenging (SI 218). Diffusion-weighted imaging at 3 T was performed with low b value (b 50) (cystic 93, solid SI 150) (C), high b value (b 800) (cystic SI 33 and solid SI 57) (D), and subsequent calculation of an apparent diffusion coefficient (ADC) map (cystic ADC 1333 and solid ADC 650) (E). Diffusion restriction in the solid component confirms histopathologically proven residual viable tumor tissue.

http://radiologyebook.com

CHAPTER 7  Tissue Characterization in Liver Imaging Using Advanced MR Techniques process of identifying pathodynamic features of underlying tissue, as well as in assessing early response to therapy. MR-based dynamic contrast-enhanced (DCE) imaging techniques incorporate T1-weighted DCE-MRI, dynamic susceptibility contrast (DSC) imaging, and arterial spin labeling (ASL) methodologies. DSC uses signal intensity changes in T2* as a direct result of first-pass gadolinium-based contrast material transit through tissue and allows identification of vascular density, mean transit time, and relative blood volume. ASL, in contrast, uses magnetically labeled blood as the intravenous tracer. Although these latter two techniques are well established in neuroradiologic imaging, their application in the liver is still experimental and in its early stages.28,77 In this section, a brief overview of qualitative tissue characterization relevant for dynamic MRI will be supplied, basic concepts of T1-weighted DCE-MRI will be outlined, and possible clinical applications will be described.

Technical Aspects In a voxel-based DCE analysis, most voxels contain elements of both the intravascular space and interstitial space, with both the arteriole of the intravascular space and the interstitial space itself being too small to be directly visualized by MRI. In general, the signal contribution related to the intravascular space is characterized by a maximum signal in the arterial phase, with subsequent gradual signal decrease during portal venous and equilibrium imaging phases; changes in intravascular signal intensity are predominantly based on mixing phenomena. In histopathologic settings of enlarged intravascular spaces, as can be seen in hypervascular lesions such as HCCs, neuroendocrine hepatic metastases, nodular hyperplasia, and adenomas, the voxels’ signal intensity is predominantly based on flow dynamics occurring in the intravascular space. Therefore hypervascular lesions show a characteristic rapid wash-in and wash-out perfusion pattern. Signal changes in the interstitial space occur through differences in hydrostatic pressure as well as gradients of contrast agent concentration between the arteriole and the interstitium, convection and diffusion, respectively. In tumors with relatively increased interstitial spaces due to histopathologic desmoplasia and resultant smaller intravascular spaces, signal dynamics in each voxel are predominantly determined by the contrast uptake of the interstitial space. Lesions such as colorectal hepatic metastasis and cholangiocarcinoma show progressive enhancement over time.89 In normal hepatic parenchyma, the hepatocytes form microscopic plates that are impenetrable by extracellular contrast material. Microcirculatory hepatic intravascular spaces are predominantly composed of sinusoids that have large fenestrations; the hepatic interstitial space, the space of Disse, is located between the sinusoid and liver plate. The large fenestrations allow free transfer of contrast material between sinusoids and spaces of Disse, making the interstitial space in normal hepatic parenchyma effectively zero because the contrast agent concentration in the interstitial and intravascular spaces are the same.54 It has to be emphasized that normal liver parenchyma is usually supplied by the hepatic artery and portal vein, and although the majority of liver tumors are predominantly supplied by the hepatic artery, there also may be a portal venous perfusion component.88 The main goal of DCE imaging is to identify and quantitate tracer kinetics, depending on the microcirculation between these intravascular and interstitial spaces. The most important quantitative parameter extracted by DCE imaging is, on the one hand, Ktrans, which allows the quantitation of volume transfer between the intravascular and extravascular/extracellular space and, on the other hand, the volume of the interstitial space, ve, per time volume, usually expressed as a fraction between 0 and 1. The underlying curve modeling and fitting

121

procedures rely highly on a defined arterial input function (AIF), which, uniquely to the liver, may be affected by hepatic arterial and portal venous components and differentiated into a single-input or dual-input AIF (Fig. 7-3).68 Whereas in computed tomography (CT) imaging, attenuation coefficients resulting from underlying tissue as well as contrast material and the subsequently calculated Hounsfield scale are linearly and proportionally related to each other, gadolinium concentration and MR signal intensity changes do not show this proportional linearity. There is, however, a linear relationship to the difference in the inverse of the T1 relaxation time between the contrast-enhanced and unenhanced MR examination. To exploit this difference and translate into gadolinium concentrations, T1 relaxation maps of the underlying tissue prior to contrast administration need to be calculated. The estimation of precontrast T1 relaxation time is usually performed by acquiring two identical T1-weighted image series with slightly different excitation angles. Subsequent perfusion imaging is then performed with either twodimensional or three-dimensional T1-weighted parallel-accelerated gradient echo sequences, potentially employing cyclic k-space filling techniques with selective undersampling.34 It has to be highlighted, however, that temporal and spatial resolution remain conflicting requirements in hepatic function imaging, in particular when this technique is used for assessment of diffuse liver disease, where not only high spatial but also high temporal resolution is required, in particular during the early arterial perfusion phases. DCE-MRI is traditionally performed with extracellular gadoliniumbased contrast agents that reside in the intravascular and intercellular spaces and are eventually excreted via the kidneys, but hepatobiliary contrast agents such as gadobenate dimeglumine and gadoxetate  disodium have the additional property of being actively taken up  into hepatocytes through the adenosine triphosphate–driven organic anion-transporting polypeptides and eventually excreted into the biliary system by multidrug resistance–associated protein-2.6,65 These hepatocyte uptake and biliary excretion properties of hepatobiliary contrast agents are similar to iminodiacetic acid, a compound used in hepatobiliary scintigraphy to determine hepatic function.66

Clinical Applications Clinically, because of the growing prevalence of steatohepatitis and chronic hepatitis in the general population, quantification of liver fibrosis and cirrhosis is essential but remains challenging, in particular when distinguishing intermediate grades of fibrosis (F2 and F3) from both minimum fibrosis (F1) and cirrhosis.17 Histopathologically, the hepatic fibrosis/cirrhosis transformation process is characterized by a decrease in sinusoids and increase in interstitial space.89 Using CT and MR contrast imaging, it has been shown that the mean transit time was increased in clinical scenarios of intermediate fibrosis compared to minimum fibrosis. Hepatic DCE-MRI imaging, in particular, was able to highlight that functional physiologic parameters such as hepatic distribution volume and arterial flow were highly sensitive and specific for the diagnosis of advanced fibrosis in histopathologic correlation.33,71,73 Use of hepatocyte-specific contrast agents has allowed for quantitative assessment of agent uptake by  MR as an evaluation of hepatic function on a segmental level, of particular importance when underlying disease such as hepatic fibrosis and cirrhosis results in a heterogeneous distribution of liver function.66 In general, as fibrosis increases and liver function worsens, uptake of hepatocyte-specific contrast agent becomes measurably lower.7,29,49 In the workup of metastatic hepatic lesions it has been shown that an enlarged background liver distribution volume, expressed as

http://radiologyebook.com

122

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

B

A

C

D

MC 1 Min/Max: 225.00/343.00 Mean/SD: 286.36/31.13 Area: 0.24 cm2

Enhancement as enhancement ratio from baseline

AORTIC ENHANCEMENT OVER TIME 900 800 700 600 500 400 300 200 100 0

MC 1

–100 0

E

25

50

75 100 125 150 175 200 225 250 Time in sec

FIG 7-3  A 36-year-old woman with a focal nodular hyperplasia in the right hepatic lobe, as seen on compressed-sensing golden-angle radial sparse parallel (GRASP) T1-weighted, fat-suppressed imaging during the early arterial (A), late arterial (B), and portal venous (C) contrast phases. GRASP imaging allowed acquisition of 20 free-breathing contrast phases, with a temporal resolution of 8 seconds for the noncontrast phase, 4 seconds for the arterial phases, and 14 seconds for the portal venous and equilibrium contrast phases. For differentiation between arterial (D-E) and portal venous

http://radiologyebook.com

123

CHAPTER 7  Tissue Characterization in Liver Imaging Using Advanced MR Techniques

Enhancement as enhancement ratio from baseline

PORTAL VENOUS ENHANCEMENT OVER TIME

MC 2 Min/Max: 1372.00/2381.00 Mean/SD: 2005.05/270.61 Area: 0.52 cm2

F

150 125 100 75 50 25 0 MC 2

–25 0

25

50

G

75 100 125 150 175 200 225 250 Time in sec

H

Enhancement as enhancement ratio from baseline

CAVAL ENHANCEMENT OVER TIME

MC 3 Min/Max: 1593.00/1776.00 Mean/SD: 1656.00/50.27 Area: 0.46 cm2

900 800 700 600 500 400 300 200 100 0

MC 3

–100 0

25

I

50

75 100 125 150 175 200 225 250 Time in sec

FIG 7-3, cont’d  (F-G) perfusion components, input functions need to be defined; peak caval (H-I) enhancement occurs between arterial and portal venous maximum enhancement. Continued

http://radiologyebook.com

124

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

MC 2 Min/Max: 1260.00/1381.00 Mean/SD: 1326.31/35.17 Area: 0.73 cm2 MC 1 Min/Max: 1165.00/1455.00 Mean/SD: 1308.53/62.41 Area: 0.69 cm2

J

K PARENCHYMAL ENHANCEMENT OVER TIME Enhancement as enhancement ratio from baseline

300 250 200 150 100 50 0

MC 1 Liver MC 2 Lesion

–50 0

L

25

50

M

75 100 125 150 175 200 225 250 Time in sec

FIG 7-3, cont’d  Combining input function with parenchymal enhancement assessment of normal hepatic

parenchyma (J) and hepatic lesion (K-L) allows calculation of color-encoded parametric maps targeting either Ktrans, ve, or, in this case, area under the gadolinium curve (M).

increased intravascular space, could potentially indicate the presence of occult micrometastases.62 The predominant area of hepatic MR microcirculation imaging remains, however, the monitoring of therapy and prediction of response, in particular to antiangiogenic chemotherapeutic agents. Posttherapeutic changes in the permeability constant Ktrans revealed that a decrease of less than 40% was associated with a progression-free survival of 10 months, compared to a progression-free survival of 16 months when Ktrans was decreased by more than 40% in a patient population with colorectal hepatic metastases.46 In particular for DCE imaging and signal modeling, many  factors can influence the calculation of physiologic parameters. Standardization of choice of AIF, type of kinetic model (e.g., number of compartments), and the model time intervals (e.g., contrast uptake phase) is essential to assess the response to antiangiogenic therapy longitudinally.

MR RELAXOMETRY A broad spectrum of hepatopathies (e.g., viral, cryptogenic, autoimmune, alcohol- and steatosis-induced hepatitis) can initiate a pathodynamic pathway ranging from mild hepatic fibrosis to cirrhosis. Evolutionary stages of this pathway can be differentiated using either the Model for End-Stage Liver Disease (MELD) or Child-Pugh scoring system. Whereas the MELD score incorporates purely objective parameters (creatinine, bilirubin level, prothrombin time, and international normalized ratio [INR]), the Child-Pugh classification adds additional subjective clinical parameters to help characterize differentiation (hepatic encephalopathy and ascites in addition to albumin and bilirubin levels and INR).46,48 In clinical MRI, image contrasts are usually based on relative signal contributions from transversal T2 and longitudinal T1 relaxation of various tissues. Owing to the variable contributions of both relaxation

http://radiologyebook.com

CHAPTER 7  Tissue Characterization in Liver Imaging Using Advanced MR Techniques mechanisms for individual tissues and tissue components, as well as differing imaging parameters to visualize differences in relaxivity— namely echo time (TE), relaxation time (TR) and inversion time (TI)—a direct correlation between imaging appearance and a standardized chronic liver disease classification system has not been achieved. MR relaxometry is based on a different concept; here, either T1 TR or T2/T2* TRs are extracted to generate grayscale/color-encoded maps of various organs. To date, it has been used to evaluate brain, breast, muscle, and hepatic tissues.14 In this section, a qualitative tissue characterization relevant for MR relaxometry will be given, the basic concept of extracting T1, T2, and T2* TRs will be outlined, and lastly, possible clinical applications for quantification of hepatic storage disease, monitoring of chronic liver disease progression, and possible combination with hepatocyte-specific contrast agents to assess liver function will be described.

Technical Aspects Fibrotic remodeling of the liver parenchyma, as seen particularly in early stages of chronic liver disease, is characterized by subtle pathophysiologic processes occurring in the hepatic microenvironment, such as focal fibrogenesis with associated inflammation, edema, and neosynthesis of extracellular matrix.96 T1, T2, and T2* TRs are all tissue-specific parameters that are dependent variables based on the physical, chemical, and biologic conditions of the surrounding microenvironments. Changes in relaxivity may be associated with the augmented hepatic water content as a result of the edema and the synthesis of collagen fibers, in which protons are less abundant and tightly bound. In addition, the structural change of the intracellular space and the increased ratio of free to bound water may contribute in particular to changes in longitudinal relaxation.35,61 The large majority of total body iron, approaching 80%, is functional and can be found in hemoglobin, myoglobin, and iron-containing enzymes. The remainder of the body’s iron is found in either storage form, using the protein ferritin, or participates in iron metabolism, being bound to the protein transferrin. Iron elimination occurs through sloughing of intestinal lining cells and skin cells, as well as menstruation, and is a nonregulated process in the human body; therefore an excessive iron delivery will result in a systemic iron overload situation. Most frequently, excess iron delivery either occurs through increased intestinal absorption or administration of blood transfusions. Excess iron will initially be stored in ferritin as well as hemosiderin; however, as the storage capacity is maximally utilized, free iron will accumulate in organs such as the liver.31,81,82 Of particular importance for hepatic iron storage is hereditary hemochromatosis through increased intestinal iron absorption. Initially iron is deposited in hepatocytes and the biliary epithelium; Kupffer cells are typically spared until later stages of disease. Conditions such as thalassemia and sickle cell disease, which require frequent blood transfusions, will lead initially to iron deposition in the reticuloendothelial cells (in particular in liver, spleen, and bone marrow) and eventually to accumulation of iron in the hepatocytes during later stages of disease. Lastly, different conditions associated with chronic liver disease, such as a viral hepatitis, alcohol-induced liver disease, and nonalcoholic fatty liver disease, may lead to increased iron deposition in both ferritin and hemosiderin forms; the resulting hepatic iron concentration is, however, lower than in hemochromatosis or transfusion-based disease.5,81,85 The intrahepatic ferritin clusters and hemosiderin aggregates lead to local field inhomogeneities and show paramagnetic properties. Greater iron concentration may induce concentration-dependent local magnetic field inhomogeneities and result in shorter T2 relaxation and prolonged R2 effects (as R2 = 1/T2) compared to less affected hepatic segments (Fig. 7-4).14,37

125

Impaired hepatic function can be further quantitated by determining exact concentrations of gadolinium-based hepatocyte-specific contrast agents during delayed imaging phases, heterogeneously distributed throughout abnormal hepatic parenchyma; a downregulated membrane transporter expression has been described in cirrhotic livers.91 A correlation of hepatocellular uptake of hepatocyte-specific contrast agents and their paramagnetic property to shorten T1 longitudinal TRs in functioning and impaired areas allows successful correlation between gadolinium concentration and end-stage liver disease scoring methodologies such as MELD.35 MR sequences and postprocessing solutions necessary to perform quantitation of voxel-to-voxel changes in relaxivity are readily available on 1.5 tesla (T) and 3 T scanners, but require capable gradient systems. The measurement of relaxation rates is based on the physical aspects of nuclei returning to their ground state after having been excited by radiofrequency pulses. T1 relaxation is produced by tissue-dependent variations in local magnetic field strength and subsequent realignment longitudinally; T2 relaxation results from transverse dephasing of individual magnetic moments of water protons as they experience magnetic field inhomogeneities as present in the main magnetic fields, as well as local microenvironments.14,61 T1 estimation can be achieved using either a variable flip angle methodology as described in the DCE-MRI section, or, more frequently employed, by using a saturation recovery gradient echo approach. The latter technique employs a nonselective saturation pulse train to prepare the longitudinal magnetization prior to readout. Practically, varying saturation recovering times prior to a segmented fast low-angle short (FLASH) readout scheme with centric reordering are used.41 For T2 and R2 estimation, a half-Fourier acquisition single-shot turbo-spin echo (HASTE) sequence with varying TEs is frequently used. To estimate T2* relaxation rates and R2* values, a similar approach using a gradient echo sequence (e.g., FLASH) is employed.41 Following acquisition, image series with varying saturation recovering times for T1 estimation and differing TEs using spin echo  and gradient echo sequences for T2 and T2* estimation, respectively, have to be postprocessed through dedicated curve fitting algorithms.41 Importantly, relatively sophisticated techniques correcting for confounders such as hepatic steatosis must be used in patient populations in whom abnormal hepatic fat and iron deposition may coexist.8,9,99

Clinical Applications Clinically, MR relaxometry has been shown to successfully differentiate normal hepatic parenchyma from chronic liver disease of various extents according to the Child-Pugh classification by assessing T1 TRs. With worsening chronic liver disease, a sequential increase in T1 TR has been observed; a T1 threshold of 748 milliseconds was found to discriminate between normal hepatic parenchyma and patients with any degree of cirrhosis.41 Normal hepatic parenchyma has been characterized by a T1 TR of around 675 ± 153 milliseconds.23,56,90 Similarly, a shortening in T2* TRs was observed with worsening disease, and  T2* TRs were able to differentiate normal hepatic parenchyma from cirrhotic hepatic parenchyma with various degrees of a fibrotic remodeling. T2 TRs were not able to differentiate between normal and  abnormal hepatic parenchyma.41 For T2 and T2* TRs, however, no mathematic models yet exist to statistically predict threshold levels to differentiate various extents of chronic liver disease. For patients with progressive iron deposition, clinical studies have shown significant curvilinear relationships between R2 and liver iron concentration in a concentration range from 0.3 to 42.7 mg Fe/g liver as well as significant linear relationships between R2* and liver iron

http://radiologyebook.com

126

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

A

B

C FIG 7-4  Estimation of hepatic iron concentration using R2* estimation based on multiple (12) echo time (TE) acquisition approach. A, A 26-year-old man with mildly elevated hepatic iron concentration (5.3 mg/g) and a corresponding R2* value of 206/sec. B, A 21-year-old man with moderately elevated hepatic iron concentration (8.1 mg/g) and a corresponding R2* value of 304/sec. C, A 24-year-old woman with severely elevated hepatic iron concentration (14.5 mg/g) and a corresponding R2* value of 535/sec.

concentration in a concentration range from 1.3 to 32.9 mg Fe/g liver.36,85,94 Assessment of hepatic function through correlation of gadolinium concentration with the MELD methodology showed a significant increase in the mean T1 TR with increasing liver damage. To differentiate normal control group (MELD ≤ 10) from patients with mild hepatic dysfunction (MELD 11-18), a T1 TR postcontrast administration cutoff of 341.0 milliseconds has been proposed, yielding a sensitivity and specificity of 78% and 73%, respectively. Differentiating mild hepatic dysfunction (MELD 11-18) from severe hepatic dysfunction (MELD > 18), a T1 TR postcontrast administration cutoff of 488.4 milliseconds has been identified, with a sensitivity and specificity of 82% and 83%, respectively.35

MR ELASTOGRAPHY While quantifying the viscoelastic properties of hepatic tissue and its change under pathologic conditions, it has been shown that the elasticity of hepatic parenchyma shows a strong correlation with the degree of hepatic fibrosis and association with increased vascular resistance, as evident in elevated portal venous pressure.42,92,93 Originally described

as ultrasound-based methodology, external compressional pulse waves are visualized as shear waves propagate through liver parenchyma and are directly dependent on and impacted by elasticity of liver parenchyma.3,97 Substituting limited-field-of-view Doppler ultrasound with whole-organ MRI, the combination of mechanical stimulation and high-contrast resolution imaging allows quantification of the liver’s viscoelastic properties, in particular for the assessment of hepatic fibrosis.18 Assessing the viscoelastic properties of various focal liver lesions is another area in which the technique holds promise.30

Technical Aspects MR elastography combines an external sinusoidal vibration pattern through propagating mechanical shear waves (ranging in frequency from 20 to 200 Hz) to probe the mechanical properties of hepatic parenchyma with an EPI or gradient echo elastography MR sequence that applies motion-encoding gradients along the slice-selection encoding direction to detect cyclic motion in the through-plane direction; external acoustic driver and motion-encoding gradients have to be synchronized in frequency and phase.93 As tissue stiffens, the waves propagate more rapidly compared to softer tissue; when waves are applied continuously, the speed of propagation is directly linked to the

http://radiologyebook.com

CHAPTER 7  Tissue Characterization in Liver Imaging Using Advanced MR Techniques wavelength within the tissue under investigation.3 These motionsensitive MRI sequences allow quantification of detected displacement fields. It is from those displacement fields that viscoelastic shear properties of hepatic parenchyma are deduced. In general, shear properties can be quantitated as complex numbers, with the real part being the storage modulus G′, and its imaginary counterpart being the loss modulus G″, describing shear elasticity and shear viscosity, respectively. In a simplified assumption that the wave traverses linearly within an incompressible and viscoelastic medium, a wave-vector conversion allows one to extract propagation in mm−1, and attenuation in mm−1, which in turn enables us to derive and quantitate hepatic elasticity and viscosity.93 Ultimately, quantitative stiffness maps are calculated, which express voxel-by-voxel stiffness in kilopascals (kPa) or velocity of shear waves in meters per second. Regional assessment of

127

elasticity correlated well with histologic staging of fibrosis for various types of chronic liver diseases; disease-dependent cutoff values have been observed (Fig. 7-5).18,74,75

Clinical Applications In assessing diffuse liver disease, it has been shown that MR elastography is a robust, reproducible, and accurate method to detect and stage liver fibrosis.43,76 Various diffuse parenchymal conditions may increase the viscoelastic properties of the liver, such as inflammation, cholestasis, congestion, and portal hypertension.76 MR elastography may prove beneficial in disease characterization and differentiation for these conditions, because liver fibrosis severity mainly correlates with tissue elasticity, whereas inflammatory conditions mainly correlate with wave scattering coefficient. Lastly, severity of portal hypertension appears to

A

B

C

D FIG 7-5  Evaluation of hepatic parenchyma with MR elastography, outlining differences on stiffness maps as well as shear wave maps in normal parenchyma and various degrees of hepatic fibrosis and cirrhosis, with histopathologic correlation of disease severity. A, A 46-year-old man without liver disease, showing normal parenchyma on stiffness map (1.7 ± 0.2 kPa); note relatively short wavelength of the visualized shear waves in the liver (B). C, A 51-year-old man with moderate fibrosis on stiffness maps (4.7 ± 0.9 kPa); note the increasing wavelength of the visualized shear waves in the liver (D). Continued

http://radiologyebook.com

128

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

E

F FIG 7-5, cont’d E, A 62-year-old woman with extensive hepatic cirrhosis on stiffness maps (9.5 ± 1.3 kPa); note the further increasing wavelength of the visualized shear waves in the liver (F).

correlate with liver and spleen viscosity. In addition it has been shown that primary sclerosing cholangitis shows greater stiffness values on MR elastography than viral hepatitis or alcoholic-induced liver disease of the same histologic stage.76 Normal liver parenchyma has shear stiffness values of less than 3 kPa.2,76 In the assessment of focal liver lesions, a general trend was observed that malignant lesions showed substantially higher stiffness than benign hepatic lesions, and a cutoff value of 5 kPa has been identified to differentiate those entities.30 Finally, the assessment of liver transplants showed that MR elastography can identify early parenchymal disease but may not be able to differentiate between acute rejection and onset of recurrent viral hepatitis.55

CONCLUSION AND OUTLOOK This chapter sought to discuss the various hepatic MRI techniques that allow quantification of liver imaging biomarkers that may be helpful in understanding and staging disease processes of the hepatobiliary system. Reproducible and specific hepatic imaging biomarkers based on MR DWI, dynamic perfusion imaging using extracellular and hepatocyte-specific contrast agents, MR relaxometry, and MR elastography were discussed. Standardization of MR sequences for tissue characterization remains one of the greatest obstacles before these techniques can be synergistically performed throughout larger patient populations. Furthermore, many of the described techniques are still based on sequence designs that do not fully integrate all possible forms of image acquisition acceleration and therefore may prolong the examination unnecessarily. Many of the techniques described in this chapter have not yet left the realm of pure research-related applications. It may be difficult to predict how many of the described techniques will affect patient care. With the current reference standard being liver biopsy—with its own spectrum of drawbacks such as invasiveness, cost, poor patient acceptance, proneness to interobserver variability and sampling errors, poor repeatability, and carrying a risk of complications (estimated at 3%, with a mortality rate of 0.03%)—a justification to bring these described biomarkers into clinical practice is clearly evident and is driving further developments.58,72 The multiparametric capabilities of MRI may hold the opportunity of improved diagnostic performance through a combination of biomarkers described in this review. As radiologic imaging techniques evolve over the next decade to further emphasize the patient-centered and precision medicine

approach, the utilization of imaging modalities that allow extracting reproducible and reliable biomarkers will very likely become part of routine clinical practice.

REFERENCES 1. Arif-Tiwari H, Kalb B, Chundru S, et al: MRI of hepatocellular carcinoma: An update of current practices. Diagn Interv Radiol 20:209–221, 2014. 2. Asbach P, Klatt D, Hamhaber U, et al: Assessment of liver viscoelasticity using multifrequency MR elastography. Magn Reson Med 60:373–379, 2008. 3. Asbach P, Klatt D, Schlosser B, et al: Viscoelasticity-based staging of hepatic fibrosis with multifrequency MR elastography. Radiology 257:80–86, 2010. 4. Aube C, Racineux PX, Lebigot J, et al: Diagnosis and quantification of hepatic fibrosis with diffusion weighted MR imaging: Preliminary results. J Radiol 85:301–306, 2004. 5. Bacon BR, Adams PC, Kowdley KV, et al: Diagnosis and management of hemochromatosis: 2011 practice guideline by the American Association for the Study of Liver Diseases. Hepatology 54:328–343, 2011. 6. Bashir MR: Magnetic resonance contrast agents for liver imaging. Magn Reson Imaging Clin N Am 22:283–293, 2014. 7. Bashir MR, Gupta RT, Davenport MS, et al: Hepatocellular carcinoma in a North American population: Does hepatobiliary MR imaging with Gd-EOB-DTPA improve sensitivity and confidence for diagnosis?   J Magn Reson Imaging 37:398–406, 2013. 8. Bashir MR, Merkle EM, Smith AD, et al: Imaging for in vivo differentiation of steatosis, iron deposition and combined storage disorder: Single-ratio in/opposed phase analysis vs. dual-ratio Dixon discrimination. Eur J Radiol 81:e101–e109, 2012. 9. Bashir MR, Zhong X, Dale BM, et al: Automated patient-tailored screening of the liver for diffuse steatosis and iron overload using MRI. AJR Am J Roentgenol 201:583–588, 2013. 10. Battal B, Kocaoglu M, Akgun V, et al: Diffusion-weighted imaging in   the characterization of focal liver lesions: Efficacy of visual assessment.   J Comput Assist Tomogr 35:326–331, 2011. 11. Bharwani N, Koh DM: Diffusion-weighted imaging of the liver: An update. Cancer Imaging 13:171–185, 2013. 12. Boll DT, Merkle EM: Imaging at higher magnetic fields: 3 T versus 1.5 T. Magn Reson Imaging Clin N Am 18:549, 2010. 13. Bruegel M, Holzapfel K, Gaa J, et al: Characterization of focal liver lesions by ADC measurements using a respiratory triggered diffusionweighted single-shot echo-planar MR imaging technique. Eur Radiol 18:477–485, 2008.

http://radiologyebook.com

CHAPTER 7  Tissue Characterization in Liver Imaging Using Advanced MR Techniques 14. Carneiro A, Vilela G, Araujo D, et al: MRI relaxometry: Methods and Applications. Braz J Phys 36:9–15, 2006; 16 A.D. 15. Chiaradia M, Baranes L, Van Nhieu JT, et al: Intravoxel incoherent motion IVIM MR imaging of colorectal liver metastases: Are we   only looking at tumor necrosis? J Magn Reson Imaging 39:317–325, 2014. 16. Chundru S, Kalb B, Arif-Tiwari H, et al: MRI of diffuse liver disease: The common and uncommon etiologies. Diagn Interv Radiol 19:479–487, 2013. 17. Cobbold JF, Patel D, Taylor-Robinson SD: Assessment of inflammation and fibrosis in non-alcoholic fatty liver disease by imaging-based techniques. J Gastroenterol Hepatol 27:1281–1292, 2012. 18. Coco B, Oliveri F, Maina AM, et al: Transient elastography: A new surrogate marker of liver fibrosis influenced by major changes of transaminases. J Viral Hepat 14:360–369, 2007. 19. Cui Y, Zhang XP, Sun YS, et al: Apparent diffusion coefficient: potential imaging biomarker for prediction and early detection of response to chemotherapy in hepatic metastases. Radiology 248:894–900, 2008. 20. Culverwell AD, Sheridan MB, Guthrie JA, et al: Diffusion-weighted MRI of the liver—Interpretative pearls and pitfalls. Clin Radiol 68:406–414, 2013. 21. d’Assignies G, Ruel M, Khiat A, et al: Noninvasive quantitation of human liver steatosis using magnetic resonance and bioassay methods. Eur Radiol 19:2033–2040, 2009. 22. Davenport MS, Heye T, Dale BM, et al: Inter- and intra-rater reproducibility of quantitative dynamic contrast enhanced MRI using TWIST perfusion data in a uterine fibroid model. J Magn Reson Imaging 38:329–335, 2013. 23. de Certaines JD, Henriksen O, Spisni A, et al: In vivo measurements of proton relaxation times in human brain, liver, and skeletal muscle: A multicenter MRI study. Magn Reson Imaging 11:841–850, 1993. 24. Deckers F, De FB, Van MF, et al: Apparent diffusion coefficient measurements as very early predictive markers of response to chemotherapy in hepatic metastasis: A preliminary investigation of reproducibility and diagnostic value. J Magn Reson Imaging 40:448–456, 2014. 25. de Souza DA, Parente DB, de Araujo AL, et al: Modern imaging evaluation of the liver: Emerging MR imaging techniques and indications. Magn Reson Imaging Clin N Am 21:337–363, 2013. 26. Do RK, Rusinek H, Taouli B: Dynamic contrast-enhanced MR imaging of the liver: Current status and future directions. Magn Reson Imaging Clin N Am 17:339–349, 2009. 27. Duran R, Ronot M, Kerbaol A, et al: Hepatic hemangiomas: Factors associated with T2 shine-through effect on diffusion-weighted MR sequences. Eur J Radiol 83:468–478, 2014. 28. Essig M, Nguyen TB, Shiroishi MS, et al: Perfusion MRI: The five most frequently asked clinical questions. AJR Am J Roentgenol 201:W495– W510, 2013. 29. Feier D, Balassy C, Bastati N, et al: Liver fibrosis: Histopathologic and biochemical influences on diagnostic efficacy of hepatobiliary contrastenhanced MR imaging in staging. Radiology 269:460–468, 2013. 30. Garteiser P, Doblas S, Daire JL, et al: MR elastography of liver tumours: Value of viscoelastic properties for tumour characterisation. Eur Radiol 22:2169–2177, 2012. 31. Gossuin Y, Muller RN, Gillis P: Relaxation induced by ferritin: A better understanding for an improved MRI iron quantification. NMR Biomed 17:427–432, 2004. 32. Gourtsoyianni S, Papanikolaou N, Yarmenitis S, et al: Respiratory gated diffusion-weighted imaging of the liver: Value of apparent diffusion coefficient measurements in the differentiation between most commonly encountered benign and malignant focal liver lesions. Eur Radiol 18:486–492, 2008. 33. Hagiwara M, Rusinek H, Lee VS, et al: Advanced liver fibrosis: Diagnosis with 3D whole-liver perfusion MR imaging—Initial experience. Radiology 246:926–934, 2008. 34. Haider MA, Farhadi FA, Milot L: Hepatic perfusion imaging:   Concepts and application. Magn Reson Imaging Clin N Am 18:465–475, 2010.

129

35. Haimerl M, Verloh N, Fellner C, et al: MRI-based estimation of liver function: Gd-EOB-DTPA-enhanced T1 relaxometry of 3T vs. the MELD score. Sci Rep 4:5621, 2014. 36. Hankins JS, McCarville MB, Loeffler RB, et al: R2* magnetic resonance imaging of the liver in patients with iron overload. Blood 113:4853–4855, 2009. 37. Hernando D, Levin YS, Sirlin CB, et al: Quantification of liver iron with MRI: State of the art and remaining challenges. J Magn Reson Imaging 2014. 38. Heye T, Boll DT, Reiner CS, et al: Impact of precontrast T10 relaxation times on dynamic contrast-enhanced MRI pharmacokinetic parameters: T10 mapping versus a fixed T10 reference value. J Magn Reson Imaging 39:1136–1145, 2014. 39. Heye T, Davenport MS, Horvath JJ, et al: Reproducibility of   dynamic contrast-enhanced MR imaging. Part I. Perfusion   characteristics in the female pelvis by using multiple computer-  aided diagnosis perfusion analysis solutions. Radiology 266:801–811, 2013. 40. Heye T, Merkle EM, Reiner CS, et al: Reproducibility of dynamic contrast-enhanced MR imaging. Part II. Comparison of intra- and interobserver variability with manual region of interest placement versus semiautomatic lesion segmentation and histogram analysis. Radiology 266:812–821, 2013. 41. Heye T, Yang SR, Bock M, et al: MR relaxometry of the liver: Significant elevation of T1 relaxation time in patients with liver cirrhosis. Eur Radiol 22:1224–1232, 2012. 42. Huwart L, Sempoux C, Salameh N, et al: Liver fibrosis: Noninvasive assessment with MR elastography versus aspartate aminotransferase-toplatelet ratio index. Radiology 245:458–466, 2007. 43. Huwart L, Sempoux C, Vicaut E, et al: Magnetic resonance elastography for the noninvasive staging of liver fibrosis. Gastroenterology 135:32–40, 2008. 44. Jerome NP, Orton MR, d’Arcy JA, et al: Comparison of free-breathing with navigator-controlled acquisition regimes in abdominal diffusionweighted magnetic resonance images: Effect on ADC and IVIM statistics. J Magn Reson Imaging 39:235–240, 2014. 45. Kakite S, Dyvorne H, Besa C, et al: Hepatocellular carcinoma: Shortterm reproducibility of apparent diffusion coefficient and intravoxel incoherent motion parameters at 3.0T. J Magn Reson Imaging 41:149– 156, 2015. 46. Kamath PS, Kim WR: The model for end-stage liver disease MELD. Hepatology 45:797–805, 2007. 47. Kamel IR, Liapi E, Reyes DK, et al: Unresectable hepatocellular carcinoma: Serial early vascular and cellular changes after transarterial chemoembolization as detected with MR imaging. Radiology 250:466– 473, 2009. 48. Kim WR, Biggins SW, Kremers WK, et al: Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med 359:1018–1026, 2008. 49. Kim JY, Lee SS, Byun JH, et al: Biologic factors affecting HCC conspicuity in hepatobiliary phase imaging with liver-specific contrast agents. AJR Am J Roentgenol 201:322–331, 2013. 50. Kim T, Murakami T, Takahashi S, et al: Diffusion-weighted single-shot echoplanar MR imaging for liver disease. AJR Am J Roentgenol 173:393–398, 1999. 51. Koh DM, Padhani AR: Functional magnetic resonance imaging of the liver: Parametric assessments beyond morphology. Magn Reson Imaging Clin N Am 18:565–585, 2010. 52. Koh DM, Scurr E, Collins D, et al: Predicting response of colorectal hepatic metastasis: Value of pretreatment apparent diffusion coefficients. AJR Am J Roentgenol 188:1001–1008, 2007. 53. Koh DM, Takahara T, Imai Y, et al: Practical aspects of assessing tumors using clinical diffusion-weighted imaging in the body. Magn Reson Med Sci 6:211–224, 2007. 54. Koh TS, Thng CH, Lee PS, et al: Hepatic metastases: In vivo assessment of perfusion parameters at dynamic contrast-enhanced MR imaging with dual-input two-compartment tracer kinetics model. Radiology 249:307–320, 2008.

http://radiologyebook.com

130

PART I  Principles of Computed Tomography and Magnetic Resonance Imaging

55. Lee VS, Miller FH, Omary RA, et al: Magnetic resonance elastography and biomarkers to assess fibrosis from recurrent hepatitis C in liver transplant recipients. Transplantation 92:581–586, 2011. 56. Lerski RA, Straughan K, Schad LR, et al: MR image texture analysis—An approach to tissue characterization. Magn Reson Imaging 11:873–887, 1993. 57. Luciani A, Vignaud A, Cavet M, et al: Liver cirrhosis: Intravoxel incoherent motion MR imaging—Pilot study. Radiology 249:891–899, 2008. 58. Maharaj B, Maharaj RJ, Leary WP, et al: Sampling variability and its influence on the diagnostic yield of percutaneous needle biopsy of the liver. Lancet 1:523–525, 1986. 59. Maharaj B, Maharaj RJ, Leary WP, et al: Sampling variability and its influence on the diagnostic yield of percutaneous needle biopsy of the liver. Lancet 1:523–525, 1986. 60. Marti-Bonmati L, Alberich-Bayarri A, Sanchez-Gonzalez J: Overload hepatitides: Quanti-qualitative analysis. Abdom Imaging 37:180–187, 2012. 61. Mathur-De VR: Biomedical implications of the relaxation behaviour of water related to NMR imaging. Br J Radiol 57:955–976, 1984. 62. Miyazaki K, Orton MR, Davidson RL, et al: Neuroendocrine tumor liver metastases: Use of dynamic contrast-enhanced MR imaging to monitor and predict radiolabeled octreotide therapy response. Radiology 263:139–148, 2012. 63. Naganawa S, Kawai H, Fukatsu H, et al: Diffusion-weighted imaging of the liver: Technical challenges and prospects for the future. Magn Reson Med Sci 4:175–186, 2005. 64. Nakanishi M, Chuma M, Hige S, et al: Relationship between diffusionweighted magnetic resonance imaging and histological tumor grading of hepatocellular carcinoma. Ann Surg Oncol 19:1302–1309, 2012. 65. Nassif A, Jia J, Keiser M, et al: Visualization of hepatic uptake transporter function in healthy subjects by using gadoxetic acid-enhanced MR imaging. Radiology 264:741–750, 2012. 66. Nilsson H, Blomqvist L, Douglas L, et al: Gd-EOB-DTPA-enhanced MRI for the assessment of liver function and volume in liver cirrhosis. Br J Radiol 86:20120653, 2013. 67. Nishie A, Tajima T, Asayama Y, et al: Diagnostic performance of apparent diffusion coefficient for predicting histological grade of hepatocellular carcinoma. Eur J Radiol 80:e29–e33, 2011. 68. Pandharipande PV, Krinsky GA, Rusinek H, et al: Perfusion imaging of the liver: Current challenges and future goals. Radiology 234:661–673, 2005. 69. Parikh T, Drew SJ, Lee VS, et al: Focal liver lesion detection and characterization with diffusion-weighted MR imaging: Comparison with standard breath-hold T2-weighted imaging. Radiology 246:812–822, 2008. 70. Park YS, Lee CH, Kim JH, et al: Using intravoxel incoherent motion IVIM MR imaging to predict lipiodol uptake in patients with hepatocellular carcinoma following transcatheter arterial chemoembolization: A preliminary result. Magn Reson Imaging 32:638–646, 2014. 71. Patel J, Sigmund EE, Rusinek H, et al: Diagnosis of cirrhosis with intravoxel incoherent motion diffusion MRI and dynamic contrastenhanced MRI alone and in combination: Preliminary experience.   J Magn Reson Imaging 31:589–600, 2010. 72. Piccinino F, Sagnelli E, Pasquale G, et al: Complications following percutaneous liver biopsy. A multicentre retrospective study on 68,276 biopsies. J Hepatol 2:165–173, 1986. 73. Ronot M, Asselah T, Paradis V, et al: Liver fibrosis in chronic hepatitis C virus infection: Differentiating minimal from intermediate fibrosis with perfusion CT. Radiology 256:135–142, 2010. 74. Ronot M, Lambert S, Elkrief L, et al: Assessment of portal hypertension and high-risk oesophageal varices with liver and spleen threedimensional multifrequency MR elastography in liver cirrhosis. Eur Radiol 24:1394–1402, 2014. 75. Ronot M, Lambert SA, Wagner M, et al: Viscoelastic parameters for quantifying liver fibrosis: Three-dimensional multifrequency MR elastography study on thin liver rat slices. PLoS ONE 9:e94679, 2014.

76. Rouviere O, Yin M, Dresner MA, et al: MR elastography of the liver: Preliminary results. Radiology 240:440–448, 2006. 77. Sandrasegaran K: Functional MR imaging of the abdomen. Radiol Clin North Am 52:883–903, 2014. 78. Sandrasegaran K, Akisik FM, Lin C, et al: Value of diffusion-weighted MRI for assessing liver fibrosis and cirrhosis. AJR Am J Roentgenol 193:1556–1560, 2009. 79. Schmid-Tannwald C, Oto A, Reiser MF, et al: Diffusion-weighted MRI of the abdomen: Current value in clinical routine. J Magn Reson Imaging 37:35–47, 2013. 80. Semelka RC, Martin DR, Balci C, et al: Focal liver lesions: Comparison of dual-phase CT and multisequence multiplanar MR imaging including dynamic gadolinium enhancement. J Magn Reson Imaging 13:397–401, 2001. 81. Siegelman ES, Mitchell DG, Semelka RC: Abdominal iron deposition: Metabolism, MR findings, and clinical importance. Radiology 199:13–22, 1996. 82. Sirlin CB, Reeder SB: Magnetic resonance imaging quantification of liver iron. Magn Reson Imaging Clin N Am 18:359–381, 2010. 83. Soher BJ, Dale BM, Merkle EM: A review of MR physics: 3T versus 1.5T. Magn Reson Imaging Clin N Am 15:277–290, 2007. 84. Springer F, Machann J, Claussen CD, et al: Liver fat content determined by magnetic resonance imaging and spectroscopy. World J Gastroenterol 16:1560–1566, 2010. 85. St Pierre TG, Clark PR, Chua-anusorn W, et al: Noninvasive measurement and imaging of liver iron concentrations using proton magnetic resonance. Blood 105:855–861, 2005. 86. Taouli B, Koh DM: Diffusion-weighted MR imaging of the liver. Radiology 254:47–66, 2010. 87. Taouli B, Vilgrain V, Dumont E, et al: Evaluation of liver diffusion isotropy and characterization of focal hepatic lesions with two single-shot echo-planar MR imaging sequences: Prospective study in 66 patients. Radiology 226:71–78, 2003. 88. Terayama N, Terada T, Nakanuma Y: An immunohistochemical study of tumour vessels in metastatic liver cancers and the surrounding liver tissue. Histopathology 29:37–43, 1996. 89. Thng CH, Koh TS, Collins D, et al: Perfusion imaging in liver MRI. Magn Reson Imaging Clin N Am 22:417–432, 2014. 90. Thomsen C, Christoffersen P, Henriksen O, et al: Prolonged T1 in patients with liver cirrhosis: An in vivo MRI study. Magn Reson Imaging 8:599–604, 1990. 91. Tsuda N, Harada K, Matsui O: Effect of change in transporter expression on gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acidenhanced magnetic resonance imaging during hepatocarcinogenesis in rats. J Gastroenterol Hepatol 26:568–576, 2011. 92. Van Beers BE, Daire JL, Garteiser P: New imaging techniques for liver diseases. J Hepatol 62:690–700, 2015. 93. Venkatesh SK, Yin M, Ehman RL: Magnetic resonance elastography of liver: Technique, analysis, and clinical applications. J Magn Reson Imaging 37:544–555, 2013. 94. Wood JC, Enriquez C, Ghugre N, et al: MRI R2 and R2* mapping accurately estimates hepatic iron concentration in transfusion-  dependent thalassemia and sickle cell disease patients. Blood 106:1460–1465, 2005. 95. Wu D, Li G, Zhang J, et al: Characterization of breast tumors using diffusion kurtosis imaging DKI. PLoS ONE 9:e113240, 2014. 96. Wynn TA: Cellular and molecular mechanisms of fibrosis. J Pathol 214:199–210, 2008. 97. Yin M, Talwalkar JA, Glaser KJ, et al: Assessment of hepatic fibrosis with magnetic resonance elastography. Clin Gastroenterol Hepatol 5:1207– 1213, 2007. 98. Yu H, Shimakawa A, McKenzie CA, et al: Multiecho water-fat separation and simultaneous R2* estimation with multifrequency fat spectrum modeling. Magn Reson Med 60:1122–1134, 2008. 99. Zhong X, Nickel MD, Kannengiesser SA, et al: Liver fat quantification using a multi-step adaptive fitting approach with multi-echo GRE imaging. Magn Reson Med 72:1353–1365, 2014.

http://radiologyebook.com

8  Normal Anatomy Haibo Xu, Haojun Shi, and Hanping Wu

ANATOMY OF THE BRAIN The advent of high-resolution computed tomography (CT) and magnetic resonance imaging (MRI) scanners has allowed the fine anatomic structure to be seen in detail. In this chapter the gross surface anatomy of the brain is briefly reviewed, followed by discussion of CT and MRI techniques, followed by the illustration of sectional anatomy in multiplanar axes in these two modalities.

OVERVIEW OF BRAIN ANATOMY

1,11,17,18,24,28,33,38,40

The brain is semisolid and conforms to the shape of the skull. Its hemispheric surface is convoluted and has gyri and sulci. The brain consists of the cerebrum, cerebellum, and brainstem.

Dura and Dural Structures The brain is protected (from the outer to the inner layer) by the skull, meninges, and cerebrospinal fluid (CSF). The meninges are organized into dura mater, arachnoid membrane, and pia mater, in order of their proximity to the skull. The pia mater follows all the gyri and is separated from the arachnoid membrane by CSF. The potential subdural space in turn separates the arachnoid membrane from the dura. The outer parietal layer of the dura adheres to the skull. The dural venous sinuses are made between the dural reflections and their opposing edges, thereby forming the superior and interior sagittal, transverse, sigmoid, cavernous, and straight sinuses. These sinuses drain to the jugular venous system.

Ventricles There are four ventricles within the brain. They are lined with ependyma and contain the CSF, produced by the choroid plexus.

Lateral (Left and Right) Ventricles.  Lateral ventricles are formed by the two ependyma-lined cavities of the cerebral hemisphere and communicate with the third ventricle via the midline foramen of Monro (Fig. 8-1). They can be divided into five parts: the anterior (frontal) horn, the ventricular body, the collateral (atrium) trigone, the inferior (temporal) horn, and the posterior (occipital) horn.

Third Ventricle.  The third ventricle is a slitlike ventricle midline cavity of the diencephalons. It communicates with the lateral ventricles via the interventricular foramina of Monro and with the fourth ventricle via the cerebral aqueduct. It contains choroid plexus in its roof.

Fourth Ventricle.  The fourth ventricle is a rhomboid-shaped cavity overlying the pons and medulla, extending from the central canal of the upper cervical spinal cord to the cerebral aqueduct of the midbrain. There is a small lateral recess on each side of the fourth ventricle that

contains choroid plexus that protrudes through the foramina of Luschka into the subarachnoid space. A small median aperture in the caudal part of the ventricle is known as the foramen of Magendie. Via the two lateral foramina of Luschka and the single medial foramen of Magendie, CSF flows into the ventricular system into the subarachnoid spaces.

Cerebral Hemispheres The two cerebral hemispheres are separated by interhemispheric fissures and falx cerebri. On the lateral surface of the brain, the sylvian fissure (lateral fissure) and the rolandic fissure (central fissure) separate the cerebral hemisphere into the frontal lobe, temporal lobe, parietal lobe, and a line drawn from the parietooccipital sulcus onto the preoccipital notch, delineating the boundaries of the parietal and temporal lobes from that of the occipital lobe. The frontal lobe, the largest of all the brain, has four principal gyri: the precentral gyrus and the superior frontal, middle frontal, and inferior frontal gyri. The precentral gyrus, parallel to the central sulcus, together with the anterior bank of the central sulcus comprises the primary motor area, which is one of the most important cortical areas for movement. Rostral to the precentral sulcus is the premotor area, another important area for movement. The middle frontal gyrus contains Brodmann’s area 8, known as the frontal eye field, which is important for conjugate eye movements. Another important motor area for speech, Broca’s area, is located at the triangular and opercular parts of the inferior frontal gyrus in the dominant hemisphere. In the parietal lobe there are a postcentral gyrus, a superior parietal lobule, and an inferior parietal lobule. The postcentral gyrus is a primary somesthetic area involved in general body sensation. The superior parietal lobule is involved with the behavioral interaction of an individual with the surrounding space, and the inferior parietal lobule with the integration of diverse sensory information for speech and perception. In the temporal lobe the superior, middle, and inferior temporal gyri are separated by the two transverse sulci. The superior temporal gyrus contains two important functional structures. One is the transverse temporal gyri of Heschl, which constitutes the primary auditory area, and another is Wernicke’s area caudal to the transverse gyri of Heschl, which is involved in the comprehension of spoken language. The inferior temporal gyrus is involved with perception of visual form and color. In the occipital lobe the occipital pole contains a portion of the primary visual area, which is more extensive on the medial aspect of the occipital lobe. On the medial surface of the hemisphere, most convolutions are somewhat flat, but the corpus callosum is the most prominent structure; it connects the two hemispheres and plays an important role in the transfer of information between the two hemispheres. Ventral to the corpus callosum is the septum pellucidum, which separates the two

http://radiologyebook.com

133

134

PART II  CT and MR Imaging of the Whole Body Foramen of Monro

Cingulate gyrus

Third ventricle and massa intermedia

Cingulate sulcus Parieto-occipital fissure Occipital horn of lateral ventricle

Rostrum of corpus callosum

Frontal horn of lateral ventricle

Inferior quadrigeminal plate Aqueduct of cerebrum

Fourth ventricle

Nodulus of vermis cerebelli

FIG 8-1  Medial surface of the brain.

lateral ventricles. Medially the paracentral lobule represents medial extensions of the precentral and postcentral gyri. Caudal to the paracentral lobule is the precuneus, which lies immediately rostral to the parietooccipital sulcus. The parietooccipital sulcus separates the parietal and temporal lobes from the occipital lobe. The occipital lobe is divided into the lingual and the cuneus gyri by the calcarine sulcus. On each bank of the calcarine sulcus is the primary visual area. Between the medial aspect of the temporal lobe and the lateral aspect of the midbrain is the hippocampal gyrus, which mediates verbal learning and memory. The hippocampal sulcus separates the hippocampal gyrus from the parahippocampal gyrus. The parahippocampal gyrus in turn is separated from the fusiform gyrus by the collateral sulcus and terminates in the uncus. The medial surface of the cerebral hemisphere is superior to the delineating components of the limbic lobe, such as the subcallosal gyrus, cingulated gyrus, parahippocampal gyrus, and uncus.

Hemispheric White Matter The massive white matter contains basically three types of fibers: projection fibers, association fibers, and commissural fibers. The projection fibers convey impulses from the cortex to the distant loci or vice versa. The association fibers connect the adjacent cortex within the same hemisphere. The commissural fibers connect corresponding cortex of the two hemispheres. The projection fibers are the following: • Corona radiata—convey impulses to and from the cortex, enter the projection fibers arranged as radiating bundles. Near the upper part of the brainstem these fibers form the internal capsule. • Optic radiation (also known as the geniculocalcarine tract or the geniculostriate pathway)—the fibers arise from the most posterior component of the posterior limb of the internal capsule and radiate toward the calcarine sulcus. • Thalamocortical radiation—arises mainly from the thalamus and projects to nearly all regions of the cortex The intrahemispheric commissures (association fibers) are the following: • Uncinate fasciculus—connect the orbital frontal gyri and parts of the inferior and middle frontal gyri to anterior portions of the temporal lobe

• Arcuate fasciculus—connect the superior and middle frontal gyri to parts of the temporal lobe • Cingulate fibers—connect the frontal and parietal lobes to the parahippocampal gyrus and adjacent temporal gyrus • Superior longitudinal fasciculus—connect the occipital lobes to the parietal and frontal lobes • Visual radiations—connect the lateral geniculate bodies to the occipital lobes • Fornix—connects the hippocampus to the ipsilateral mammillary body The interhemispheric commissures (commissural fibers) are the following: • Anterior commissure—connect the temporal lobes and olfactory structures on the two sides • Posterior commissure—connect the rostral midbrain nuclei, located behind the third ventricle • Corpus callosum—connect all corresponding cortex of the opposite hemisphere

Posterior Fossa The posterior fossa contains the cerebellum and brainstem. The posterior fossa is outlined by the clivus and petrous bones anteroinferiorly, the tentorium cerebelli superiorly, and the occipital bone posteroinferiorly. The cerebellum is located posteriorly in the two thirds of the posterior fossa, separated from the brainstem by the fourth ventricle. The brainstem occupies the anterior third of the posterior fossa, including the midbrain, pons, and medulla oblongata. The midbrain consists of cerebral peduncles anteriorly and tectum posteriorly. The prominent cerebral peduncles leave a triangular depressed area called the interpeduncular fossa. The basilar artery lies in this fossa and the third cranial nerve emerges from it. The dorsal surface of the midbrain is composed of the tectum and tegmentum. The tectum represents the superior and inferior colliculi. Caudal to the inferior colliculus, the fourth cranial nerve, the only cranial nerve, exits from the dorsal aspect of the brainstem. Dorsal to the tegmentum is the cerebral aqueduct, connecting the third and fourth ventricles. The pons on the ventral surface is separated from the crus cerebri by the superior pontine sulcus and from the medulla by the inferior pontine sulcus. The massive transverse fibers in the ventral portion of

http://radiologyebook.com

CHAPTER 8  Normal Anatomy the pons form the middle cerebellar peduncle. On the dorsal surface the pontine tegmentum is continuous with the reticular formation  of the medulla and midbrain. Laterally the cerebellopontine angle (CPA) is formed by the junction of pons, medulla, and cerebellum. From the CPA emerge the seventh and eighth cranial nerves. Caudally the ninth and tenth cranial nerves cross the CPA; superiorly the fifth nerve crosses it. The medulla is the most caudal portion of the brainstem and continuous with the cervical cord at the level of the foramen magnum. Anteriorly the pyramid has a protrusive appearance formed largely by the corticospinal tracts. Posterior and lateral to the pyramid is a very shallow groove called the ventrolateral sulcus in which the vertebral arteries usually lie. Lateral to the pyramid is the olive, corresponding to the inferior olivary nucleus. Ventrolaterally the ninth, tenth, and eleventh cranial nerves emerge from the postolivary sulcus, and the twelfth cranial nerve emerges from the preolivary sulcus. Laterally the inferior cerebellar peduncles, connecting the spinal cord and medulla with the cerebellum, make the superior protrusion.

The cerebellum is attached to the brainstem by three cerebellar peduncles and forms the roof of the fourth ventricle. It consists of the vermis and hemispheres. The two cerebellar hemispheres are united by the vermis as a midline structure. On the inferior surface, two deep sulci clearly separate the vermis from the hemispheres. The cerebellum is divided by the primary fissure into anterior and posterior lobes and is divided by the horizontal fissure into superior and inferior halves.

Vascular Supply2,10,11,16,24,28 The brain derives its vascular supply (Fig. 8-2) from two carotid and two vertebral arteries. The internal carotid artery bifurcates terminally into the anterior and middle cerebral arteries. The two vertebral arteries unite at the caudal border of the pons to form the basilar artery. Anterior to the interpeduncular fossa, the basilar artery divides into the two posterior cerebral arteries. Anterior and posterior communicating arteries and proximal portions of the anterior, middle, and posterior cerebral arteries form the cerebral arterial circle called the circle of Willis.

Frontal

Frontal

Frontal

Par Parietal

Parietal

Temporal Par

Occp Occp

Frontal

Frontal

Frontal

Temporal Temporal

Temporal Occipital

135

Occp

Anterior cerebral artery

Posterior cerebral artery

Anterior cerebral artery, deep

Posterior cerebral artery, deep

Middle cerebral artery

Anterior choroidal artery

Middle cerebral artery, deep

FIG 8-2  Demonstration of vascular territories on axial images. (Modified from Gonzalez C, et al: Head and spine imaging, New York, 1985, Churchill Livingstone, p 145.)

http://radiologyebook.com

136

PART II  CT and MR Imaging of the Whole Body

The anterior cerebral artery supplies blood mainly to the frontal lobe medially and mediolaterally, with some supply to the parasagittal parietal lobe. The medial lenticulostriate arteries, arising from the horizontal portion of the anterior cerebral artery, supply the anterior aspects of the putamen and caudate nucleus, the anteroinferior part of the internal capsule, and the dorsal aspect of the chiasm and the hypothalamus. The pericallosal branches supply the upper surface of the corpus callosum and septum pellucidum. The medial surface of the hemisphere is supplied by the hemispheric branches of the anterior cerebral artery. The anterior choroidal branch of the internal carotid artery supplies the remaining portion of the internal capsule as well as the choroid plexus. The middle cerebral artery, the largest of the cerebral arteries, gives off five to eight branches into the insular region called the sylvian triangle and branches from the lateral sulcus over the lateral convexity of the hemisphere. These branches supply blood mainly to the frontal lobe inferiorly and laterally, most of the temporal lobe, the temporooccipital and angular areas, and the motor and premotor areas of the parietal lobe. The lenticulostriate arteries, the first branches to arise from the horizontal portion of the middle cerebral artery, supply the lateral basal ganglia. The posterior cerebral artery divides mainly into the posterior temporal artery and the internal occipital arteries. The former supplies the posterior temporal lobe and the occipitotemporal and lingual gyri; the latter supplies the occipital lobe, portions of the splenium of the corpus callosum, and part of the temporal lobe. The anterior cerebral territory is located between the anterior part of the interhemispheric fissure and a line along the anterior lateral wall of the lateral ventricle, and the posterior cerebral territory is located between the posterior portion of the interhemispheric fissure and a line along the posterior lateral wall of the lateral ventricle. The area between the lines along the lateral ventricle wall is the middle cerebral territory. The ill-defined anterior area between the anterior cerebral artery and the middle cerebral artery is called the anterior watershed region. The area between the middle cerebral and posterior cerebral arteries is the posterior watershed area. These two areas are more susceptible to hypotensive infarctions than other areas of the brain are. In the posterior fossa, each half of the cerebellum is supplied by the posterior inferior cerebellar artery, the anterior inferior cerebellar artery, and the superior cerebellar artery. The posterior inferior cerebellar artery, derived from the vertebral artery, supplies the inferior vermis, inferior cerebellar hemispheres, and choroid plexus of the fourth ventricle. The anterior inferior cerebellar arteries, arising from the caudal part of the basilar artery, supply mainly the area of the CPA and the anterior inferior cerebellar hemispheres. The perforator branches of the basilar artery supply the pons. The superior cerebellar arteries, arising from the rostral part of the basilar artery, supply the superior portion of the cerebellar hemispheres.

CT TECHNIQUE

11,25,39,42

Axial head CT scanning is performed traditionally with the angulation of the gantry parallel to the orbitomeatal line, defined as passing through the lateral canthus and middle of the external ear canal. To avoid radiation through the eyes when performing axial head CT, a 15- to 20-degree angulation of the gantry to the canthomeatal line is advisable. To match the MRI slices, axial head CT scanning is advocated to parallel the Talairach anterior commissure–posterior commissure (AC-PC) line with the angle 12 degrees from the hard palate line, as visualized on the lateral CT scout view.39 Eight- to 10-mm-thick sections are routinely obtained, but thinner slices with 1.5- to 3-mm thickness are necessary in evaluation of the orbits, pituitary gland, suprasellar and parasellar regions, and the CPA. In these situations,

reformatted images in coronal and sagittal planes are useful. With the advent of the MR scanner, however, head CT is limited mainly to detect bleeding in acute stroke, acute trauma, calcification, and bony diseases.

Contrast Studies Intravenous (IV) contrast medium is helpful to improve the detection and characterization of intracranial lesions. Enhanced CT scanning is not used in patients with acute trauma, patients with acute stroke to exclude a bleed, follow-up of trauma patients, patients with renal disease, or patients with a history of allergy to iodinated contrast medium. About 100 to 150 mL of a 350 mg I/mL iodinated contrast agent (adult dosage) should be administered for optimal study results.

MRI TECHNIQUE7,34 MRI parameters include repetition time (TR), echo time (TE), matrix size, field of view (FOV), slice thickness, and number of excitations (NEX). TR and TE are the only parameters that affect T1 and T2 weighting of the image. T1 images have a short TR of less than 1000 milliseconds and a short TE of less than 30 milliseconds. T2-weighted images have a long TR of more than 1500 milliseconds with dual echoes. The first echo of T2 images with shorter TE is called proton density, or a mixture of T1 and T2. The second echo of T2 images with a longer TR of more than 60 milliseconds represents real T2 images. T2-weighted images are most sensitive for detecting brain pathology, owing to the higher water content in most of the lesions. Gadolinium compounds, which are paramagnetic IV contrast media for brain imaging, are especially helpful for identification of extraaxial tumors and intraaxial lesions associated with disrupted blood-brain barrier. The regular dose is about 0.1 mmol/kg of body weight, but in a case with equivocal findings or solitary metastasis, a dose of 0.2 to 0.3 mmol/kg of body weight is advisable. The same structures that enhance with IV use of contrast medium on CT scans enhance similarly on MRI scans.

SECTIONAL ANATOMY Normal Axial CT and MRI Anatomy11,12,24,25 On CT and MR scans the brain has been briefly viewed in infratentorial and supratentorial sections, as described below. CT scans are performed with a 15- to 20-degree angulation to the canthomeatal line at 8-mm increments. MRI scans are generally obtained parallel to the AC-PC line in the axial plane, with 6-mm slice thickness. Using the sagittal view the coronal sections are acquired parallel to the brainstem and the sagittal sections are obtained perpendicular to the axial section.

Infratentorial Sections Below the fourth ventricle (Fig. 8-3).  In this section, most of the structures in the anterior and middle fossa are components of the base of the skull and the orbits. In the middle fossa the foramen ovale and foramen spinosum can be visualized with a wide window setting on CT scans. Through these foramina pass the third branch of the fifth cranial nerve and the middle meningeal artery, respectively. The cerebellar hemispheres can be seen lateral to the medulla. On MRI studies, cranial nerves IX and X can be demonstrated at this level because they emerge from the postolivary sulcus. The posterior aspect of the cerebellar hemispheres is outlined by the inferior portion of the cisterna magna. Fourth ventricular level (Fig. 8-4).  The lower pons is seen in front of the fourth ventricle, connecting to the cerebellar hemispheres by the middle cerebellar peduncles. The pons is outlined by CSF, with the

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Eye globe Ethmoid sinus

Sphenoid sinus

Auricular canal Medulla Mastoid cells Cerebellar hemisphere

A

Ethmoid sinus Eye globe Orbital fat

Sphenoid sinus Temporal lobe Internal jugular vein

Petrous bone Medullary cistern

Medulla oblongata Cerebellar tonsil

Cerebellar hemisphere

B FIG 8-3  A, Plain axial CT scan above the foramen magnum. B, Axial T1 MRI scan above the foramen magnum.

Continued

http://radiologyebook.com

137

138

PART II  CT and MR Imaging of the Whole Body

Lens Eye globe

Ethmoid sinus

Optic nerve

Temporal lobe Sphenoid sinus Basilar artery Cranial nerves IX and X

Medullary cistern

Medulla oblongata

Cerebellar hemisphere

C FIG 8-3, cont’d  C, Axial T2 MRI scan above the foramen magnum.

anterior and lateral pontine cisterns. Anterior to the pons appears the basilar artery, with signal void on regular MRIs. Posteriorly the fourth ventricle is outlined by the cerebellar nodulus in the midline and by the cerebellar hemispheres laterally. On MRI examinations the seventh and eighth cranial nerves can be seen at the CPA. Above the fourth ventricle (Fig. 8-5).  At the center, the midbrain appears outlined by the suprasellar cistern, ambient cistern, and quadrigeminal cistern. The suprasellar cistern contains the internal carotid artery, the anterior and middle cerebral arteries, the optic chiasm, the infundibulum, the mammillary bodies, and the top of the basilar artery. Lateral to the midbrain are the temporal lobes, where the uncus, amygdaloid nucleus, and hippocampal formation can be demonstrated. The temporal horn can be seen in the middle of the temporal lobe at this level. In the anterior fossa, the frontal lobes are separated from the temporal lobes by the sylvian fissure. Most inferior portions of the frontal lobe can be seen at this level. In the posterior fossa, the aqueduct of Sylvius is present behind the midbrain. The superior cerebellar surface is seen with separation of the two hemispheres by the superior vermis.

Supratentorial Sections Third ventricular level (Fig. 8-6).  The frontal lobes can be seen separated by the interhemispheric fissure medial to them. Superficially the sylvian fissures extend medially to separate the frontal lobe from the temporal lobe. Medial to the medial surface of the sylvian fissure, the insular cortex, external capsule, putamen, and globus pallidus can be demonstrated. The third ventricle in the midline is seen as a slitlike cavity surrounded by the thalamus laterally. Behind the third ventricle, the internal cerebral vein, the great vein of Galen, and the cerebellar vermis can be seen. Sometimes the pineal gland behind the

third ventricle is seen at this level. Occasionally some calcification of the habenula occurs anterior to the pineal gland calcification. Within the lateral ventricle the most commonly calcified choroid plexus can be identified. Posteriorly the occipital lobes are visualized. Low ventricular level (Fig. 8-7).  The most superior aspects of the frontal lobes are visualized. The superior portion of the frontal horns appears outlined by the head of the caudate nuclei laterally. Anteriorly the frontal horns are shaped by indentation of the genu of the corpus callosum. The corpus callosum is seen separated from the cingulate gyrus by the cingulate sulcus. Posteriorly the occipital horns are indented by the splenium of the corpus callosum. Within the lateral ventricle the choroid plexus can be visualized on MRIs. Behind the splenium of the corpus callosum, the straight sinus is presented as signal void on regular MRIs. Midventricular level (Fig. 8-8).  The superior extension of the sylvian fissure and the superior temporal gyrus are seen. The central sulcus separates the temporal lobe from the parietal lobe. The most superior aspect of the frontal horns is indented laterally by the most superior aspect of the caudate nuclei. The posterior medial aspect of the occipital horns is seen bound by the white matter fibers of the splenium of the corpus callosum. The posterior portion of the cingulate sulcus, separating the cingulate gyrus from the occipital lobe, can be seen connecting to the posterior interhemispheric fissure. Above the ventricular level (Fig. 8-9).  This section mainly includes the frontal lobe, parietal lobe, and a small portion of the occipital lobe. The superior frontal sulcus, separating the superior frontal gyrus from the middle frontal gyrus, is well delineated parallel to the interhemispheric fissure. Three parallel sulci including the precentral, central, and postcentral sulci are well visualized at this level.

http://radiologyebook.com

Text continued on p. 148

CHAPTER 8  Normal Anatomy

Lens Eye globe

Sphenoid sinus Temporal lobe

Pons Mastoid cell

Fourth ventricle

Cerebellar hemisphere

Occipital bone

A

Eye globe Orbital fat

Sphenoid sinus

Temporal lobe Basilar artery Pons Middle cerebellar peduncle Cerebellar vermis

Cerebellar hemisphere

Confluens sinuum

B FIG 8-4  A, Plain axial CT scan at the level of the fourth ventricle. B, Axial T1 MRI scan at the level of the fourth ventricle.

Continued

http://radiologyebook.com

139

140

PART II  CT and MR Imaging of the Whole Body

Orbital fat Ethmoid sinus Optic nerve Temporal lobe

Sphenoid sinus

Pontine cistern Basilar artery Cranial nerves VII and VIII

Pons

Cerebellar hemisphere

C FIG 8-4, cont’d  C, Axial T2 MRI scan at the level of the fourth ventricle.

Frontal sinus

Gyrus rectus Hypophysial stalk

Middle cerebral artery

Dorsum sellae

Suprasellar cistern

Mesencephalon Quadrigeminal cistern

Midbrain tegmentum Vermis

Cerebellar hemisphere

A FIG 8-5  A, Plain axial CT scan at the level above the fourth ventricle.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Superior frontal gyrus Middle frontal gyrus Frontal lobe Inferior frontal gyrus

Optic tract

Infundibulum

Amygdala

Uncus Ambient cistern

Temporal horn of lateral ventricle

Midbrain tegmentum

Hippocampus

Quadrigeminal plate cistern

Suprasellar cistern

Cerebellum

B

Frontal lobe

Anterior cerebral artery Optic tract Middle cerebral artery

Substantia nigra

Infundibulum Quadrigeminal cistern

Suprasellar cistern

Midbrain tegmentum

Vermis

Confluens sinuum

C FIG 8-5, cont’d  B, Axial T1 MRI scan at the level above the fourth ventricle. C, Axial T2 MRI scan at the level above the fourth ventricle.

http://radiologyebook.com

141

142

PART II  CT and MR Imaging of the Whole Body

Frontal lobe

Frontal sinus

Anterior horn of lateral ventricle

Head of caudate nucleus Lenticular nucleus

Cistern of lateral sulcus

External capsule

Insula

Thalamus

Third ventricle Cornu occipitale ventriculi lateralis

Great cerebral vein

Internal cerebral veins

Falx cerebri

A

Falx cerebri Superior frontal gyrus

Cingulate gyrus Frontal horn Caudate nucleus

Septum pellucidum

Third ventricle

Lenticular nucleus

Temporal horn

Thalamus

Vermis

B FIG 8-6  A, Plain axial CT scan at the third ventricular level. B, Axial T1 MRI scan at the third ventricular level.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy Superior sagittal sinus

Superior frontal gyrus

Cingulate gyrus Genu of corpus callosum

Pericallosal artery

Frontal horn of lateral ventricle

Anterior limb of internal capsule

Head of caudate nucleus

Fornix Superior temporal gyrus

Lentiform nucleus Posterior limb of internal capsule

Third ventricle

Thalamus

Trigone of lateral ventricle

Internal cerebral vein Cerebellar vermis

C Great cerebral vein

Occipital lobe

FIG 8-6, cont’d  C, Axial T2 MRI scan at the third ventricular level.

Cingulate sulcus Genu of corpus callosum

Falx cerebri

Frontal horn

Anterior limb of internal capsule Septum pellucidum

Head of caudate nucleus

Lenticular nucleus

Genu of internal capsule

Posterior limb of internal capsule

Occipital horn Fornix

Thalamus

Internal cerebral veins

Straight sinus

Falx cerebri

A FIG 8-7  A, Plain axial CT scan at the low ventricular level. Continued

http://radiologyebook.com

143

144

PART II  CT and MR Imaging of the Whole Body

Genu of corpus callosum

Cingulate gyrus

Frontal lobe

Head of caudate nucleus

Septum pellucidum

Lenticular nucleus

Interventricular foramen

Fornix Superior temporal gyrus

Internal capsule

Choroid plexus

Thalamus

Straight sinus

Splenium of corpus callosum

Occipital lobe

Calcarine sulcus

B

Superior frontal gyrus

Cingulate sulcus Cingulate gyrus

Middle frontal gyrus

Head of caudate nucleus

Inferior frontal gyrus Central sulcus

Lateral ventricle

Superior temporal gyrus

Lenticular nucleus Splenium of corpus callosum

Choroid plexus

Straight sinus

Occipital lobe

Calcarine sinus

C FIG 8-7, cont’d  B, Axial T1 MRI scan at the low ventricular level. C, Axial T2 MRI scan at the low ventricular level.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Superior frontal gyrus

Middle frontal gyrus

Cingulate gyrus Anterior horn of lateral ventricle

Septum pellucidum

Head of caudate nucleus

Choroid plexus Trigone of lateral ventricle

Splenium of corpus callosum

Angular gyrus

Falx cerebri

Parieto-occipital sulcus

A

Cingulate sulcus

Superior frontal gyrus

Corpus callosum Septum pellucidum Choroid plexus

Body of fornix

Lateral ventricle Optic radiation

Calcarine fissure Superior sagittal sinus

B FIG 8-8  A, Plain axial CT scan at the midventricular level. B, Axial T1 MRI scan at the midventricular level. Continued

http://radiologyebook.com

145

146

PART II  CT and MR Imaging of the Whole Body

Superior sagittal sinus Cingulate gyrus

Inferior frontal sulcus Precentral gyrus

Centrum semiovale

Central sulcus

Lateral ventricle Supramarginal gyrus

Splenium of corpus callosum

Calcarine sulcus Occipital lobe

C FIG 8-8, cont’d  C, Axial T2 MRI scan at the midventricular level.

Superior frontal gyrus Middle frontal gyrus Precentral gyrus Centrum semiovale

Central sulcus Postcentral gyrus

Cingulate gyrus Falx cerebri

A FIG 8-9  A, Plain axial CT scan above the ventricular level.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Superior frontal gyrus

Superior sagittal sinus

Middle frontal gyrus Precentral gyrus Central sulcus Marginal sulcus

Postcentral sulcus

Parieto-occipital fissure

Angular gyrus

B

Superior frontal gyrus Superior frontal sulcus

Falx cerebri

Cingulate gyrus

Middle frontal gyrus

Centrum semiovale Central sulcus Intraparietal sulcus Superior sagittal sinus

C FIG 8-9, cont’d  B, Axial T1 MRI scan above the ventricular level. C, Axial T2 MRI scan above the ventricular level.

http://radiologyebook.com

147

148

PART II  CT and MR Imaging of the Whole Body

The central sulcus can be identified in the midportion of the scan. Precentral and postcentral sulci outline motor and sensory cortices. The interhemispheric fissure can be seen in its entire length, with the falx between. Lateral to the interhemispheric fissure the centrum semiovale, being massive white mater, is well documented.

Review of Coronal Scans.  Coronal sections of the brain are briefly discussed from the frontal to the occipital horn level of the lateral ventricles on MRIs. Frontal horn level.  See Figure 8-10. Third ventricular level.  See Figure 8-11. Brainstem level.  See Figure 8-12. Midventricular level.  See Figure 8-13. Occipital horn level.  See Figure 8-14.

Review of Sagittal Scans.  Sagittal images are essential in the evaluation of sellar and parasellar lesions, posterior fossa lesions, and intraventricular lesions, as well as for evaluation of the vascular anatomy. Moreover, sagittal images are necessary in three-dimensional visualization and accurate localization of intracranial lesions. Midsagittal level.  See Figure 8-15. Parasagittal level through the lateral ventricular body.  See Figure 8-16. Lateral orbital level (cortical level).  See Figure 8-17.

Physiologic Calcifications5,13,23,27,43 The following calcifications seen on CT and MRI of the brain are considered normal.

Choroid Plexus.  Calcification occurs mainly in the glomus, located within the atrium of the lateral ventricles. Calcification of the choroid plexus is uncommon in patients younger than 10 years. With age the incidence increases, reaching 86% by the eighth decade.

Basal Ganglia.  Calcification within the basal ganglia is common after the age of 40 years but should be considered pathologic in persons younger than age 40 unless proved otherwise. Furthermore, calcification elsewhere in the basal ganglia, dentate nuclei, or cortex should be considered pathologic.

Pineal Gland and Habenula.  Pineal gland calcification is common after the age of 30 years, but it is rare in persons younger than age 10. However, age is not the only important factor; displacement and size of the calcification are significant. If calcification of the pineal gland is larger than 12 × 12 × 12 mm, the possibility of a pineal gland–region tumor should be considered.

Falx Calcification.  Calcification of the falx is commonly seen in the posterior interhemispheric fissure in adults. If increased attenuation of the falx extends to the medial sulci, the possibility of subarachnoid bleeding should be excluded.

Dentate Nuclei.  Calcification of the dentate nucleus is uncommon compared to calcification within the basal ganglia. If massive calcification of the dentate nucleus occurs with basal ganglia calcification, the possibility of Fahr’s disease should be ruled out.

ANATOMY OF THE SPINE The techniques of CT and MRI scanning are introduced briefly, followed by discussion of the normal anatomy of the spine using those modalities.

CT TECHNIQUE22,31,41 CT is still frequently performed in evaluation of the spine, especially of the osseous structures, despite the rapid development and wide acceptance of MRI techniques. Routine 5- or 3-mm-thick scans are obtained parallel to the disk space, but for evaluation of the craniocervical junction, axial CT scanning can be performed at 0-degree angulation to the orbitomeatal line, with a contiguous cut of 1.5- or 3-mm thickness. The reformatted coronal or sagittal scans are useful for evaluating the odontoid process, the articulation of C1 with C2, and the articulation of the occipital condyles with the superior articular facets of C1 (Fig. 8-18). In evaluation of the cervical osseous articulations and disk disease, 1.5-mm slice thickness is advisable. There are two methods of choice to perform scanning. One is a stacking method without gantry angulation. This method has the advantage of visualizing the entire scanned area in one column without interruption to reconstruct the images in a different orientation for further evaluation. The other method is performed parallel to the disk space with gantry angulation. Scans require different angulation at each level, and thus not all of the slices can be reconstructed together. Scans are imaged by standard and bone algorithms. Generally the use of IV contrast may facilitate the diagnostic accuracy of CT in evaluation of both cervical disk herniation and postoperative spine in an attempt to visualize the epidural venous plexus.31

MRI TECHNIQUE3,6,21,31 MRI scanning of the spine is usually carried out in the orthogonal sections—that is, coronal, sagittal, and transverse. Our routine MRI technique includes T1 and T2 fast-spin echo images in the sagittal plane and T2 axial images. Sometimes oblique imaging is added to demonstrate the neuronal foramina if necessary, because the exit zone of the spinal nerve is oriented at a 45-degree angle to the neural canal. IV administration of gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA) is useful in evaluation of postoperative spine studies and better characterization of other lesions of the entire spine.

NORMAL ANATOMY* There are 32 or 33 vertebrae in the vertebral column, including 7 segments in the cervical, 12 in the thoracic, 5 in the lumbar, 5 in the sacral, and 3 or 4 in the coccygeal area. The osseous and soft tissue structures of the spinal anatomy are discussed separately for CT scans and MRIs as follows.

Osseous Structures

Vertebral Bodies.  In the cervical region (Fig. 8-19), the transverse diameter is greater than the anteroposterior diameter. The vertebral bodies are oval in the axial plane. The vertebral bodies become wider from C2 to C7 as weight bearing increases. In the thoracic area, the transverse and anteroposterior diameters are the same (Fig. 8-20). The vertebral bodies appear cone shaped in cross section. In the lumbar region (Fig. 8-21), the transverse diameter is wider than the anteroposterior diameter and generally larger than that at the thoracic level. As viewed from above the lumbar bodies are kidney shaped in the axial plane, and two venous channels for the basivertebral veins emerge close to the midline of the posterior cortex and form a Y-shaped defect on the vertebral body.

*References 3, 4, 6, 8, 9, 14, 15, 19-21, 26, 29, 30, 32, 35-37.

http://radiologyebook.com

Text continued on p. 159

CHAPTER 8  Normal Anatomy Superior frontal gyrus

Superior sagittal sinus

Body of corpus callosum

Middle frontal gyrus

Temporalis muscle

Interhemispheric fissure Cingulate gyrus

Inferior frontal gyrus

Centrum semiovale

Anterior horn of lateral ventricle

Sylvian fissure

Septum pellucidum

Temporal lobe Caudate nucleus

Internal carotid artery

Fornices

Optic chiasm

Internal capsule

Sphenoid sinus Pituitary gland Pharyngeal tonsil

A Mesopharynx Superior frontal gyrus

Nasopharynx Superior sagittal sinus

Body of corpus callosum

Middle sagittal sinus

Frontal horn Septum pellucidum

Interhemispheric fissure

Lentiform nucleus

Centrum semiovale Temporalis muscle

Sylvian fissure Caudate nucleus

Anterior cerebral artery

Fornices Optic chiasm

Internal capsule

Internal carotid artery

Middle cerebral artery

Sphenoid sinus Pituitary gland

Mesopharynx

B Nasopharynx

FIG 8-10  A, Coronal T1 MRI scan at the level of the frontal horns. B, Coronal T2 MRI scan at the level of the frontal horns.

http://radiologyebook.com

149

150

PART II  CT and MR Imaging of the Whole Body

Superior sagittal sinus

Cingulate sulcus

Falx Central sulcus Postcentral gyrus

Head of caudate nucleus

Precentral gyrus

Septum pellucidum

Thalamus Third ventricle

Occipital horn Sphenoid sinus

Hippocampus Parahippocampal gyrus

Pituitary gland

A Superior sagittal sinus

Cingulate sulcus

Cerebral falx

Precentral gyrus Central sulcus

Body of corpus callosum

Postcentral gyrus

External capsule

Sylvian fissure

Head of caudate nucleus

Superior temporal gyrus

Occipital horn

Middle temporal gyrus

Hippocampus

Sphenoid sinus Parahippocampal gyrus

Inferior temporal gyrus

Third ventricle

Putamen

B Internal capsule

FIG 8-11  A, Coronal T1 MRI scan at the level of the third ventricle. B, Coronal T2 MRI scan at the level of the third ventricle.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy Superior sagittal sinus

Superior frontal gyrus

Precentral gyrus

Cingulate gyrus

Postcentral gyrus

Fornix

Central sulcus

Pineal body Lateral ventricle

Red nucleus

Hippocampus Middle cerebellar peduncle

Cerebral peduncle

Medullary pyramid

A Superior sagittal sinus Superior frontal gyrus

Precentral gyrus

Cingulate gyrus

Postcentral gyrus

Fornix

Central sulcus Lateral ventricle

Red nucleus

Pineal body

Hippocampus Middle cerebellar peduncle

Cerebral peduncle

Medullary pyramid

B FIG 8-12  A, Coronal T1 MRI scan through the midbrain. B, Coronal T2 MRI scan through the midbrain.

http://radiologyebook.com

151

152

PART II  CT and MR Imaging of the Whole Body

Falx Superior sagittal sinus

Interhemispheric fissure

Lateral ventricle Superior temporal gyrus

Splenium of corpus callosum

Middle temporal gyrus

Quadrigeminal cistern

Cerebellar vermis Horizontal fissure

Nodules of cerebellar vermis

Fourth ventricle Cisterna magna Cerebellar tonsil

A Superior sagittal sinus

Falx Interhemispheric fissure

Postcentral sulcus

Lateral ventricle

Central sulcus

Superior temporal gyrus

Supramarginal gyrus

Middle temporal gyrus

Sylvian fissure

Inferior temporal gyrus

Cerebellar vermis Cerebellar hemisphere

Quadrigeminal cistern

Nodules Cerebellar tonsil

Cisterna magna

B FIG 8-13  A, Coronal T1 MRI scan at the mid-to-posterior ventricular level. B, Coronal T2 MRI scan at the mid-to-posterior ventricular level.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy Superior sagittal sinus

Falx

Superior parietal lobe

Interhemispheric fissure

Middle temporal gyrus Superior cerebellar cistern

Calcarine sulcus

Lingual gyrus Occipital horn

Transverse sinus Cerebellar vermis Horizontal fissure

Tentorium of cerebellum

A Superior sagittal sinus

Precuneus

Cingulate sulcus

Superior parietal lobe

Calcarine sulcus Middle temporal gyrus Superior cerebellar cistern Occipital horn Tentorium of cerebellum

Internal hemispheric fissure

Horizontal fissure

Cerebellar vermis

Lingual gyrus

B FIG 8-14  A, Coronal T1 MRI scan slightly posterior to the occipital horns. B, Coronal T2 MRI scan slightly posterior to the occipital horns.

http://radiologyebook.com

153

154

PART II  CT and MR Imaging of the Whole Body Fornix

Superior frontal gyrus

Central sulcus Paracentral lobule Cingulate gyrus

Body of corpus callosum

Superior sagittal sinus

Cingulate sulcus

Splenium of corpus callosum

Lateral ventricle Parieto-occipital sulcus

Genu of corpus callosum

Internal cerebral vein

Anterior commissure

Tectal plate

Optic chiasm

Tegmentum of midbrain Pituitary gland Cerebellum Fourth ventricle

Pons

Tonsil cerebellum

Medulla oblongata

A

Foramen magnum Interthalamic adhesion Fornix

Superior frontal gyrus

Central sulcus Paracentral lobule

Body of corpus callosum

Superior sagittal sinus

Cingulate sulcus Lateral ventricle

Parieto-occipital sulcus

Genu of corpus callosum

Cingulate gyrus

Splenium of corpus callosum

Anterior commissure

Internal cerebral vein

Optic chiasm

Cerebellum Tectal plate Pituitary gland

Fourth ventricle Tonsil cerebellum

B

Tegmentum of midbrain Interthalamic adhesion

Pons

Medulla Foramen magnum oblongata

FIG 8-15  A, Sagittal T1 MRI scan at the midsagittal level. B, Sagittal T2 MRI scan at the midsagittal level.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy Central sulcus

Precentral gyrus

Postcentral gyrus Superior frontal gyrus

Thalamus Lateral ventricle

Putamen Tail of hippocampus

Amygdala

Body of hippocampus Internal carotid artery

Temporal horn of lateral ventricle

Maxillary sinus Cerebellum Head of hippocampus Flocculus

A

Central sulcus Precentral gyrus Postcentral gyrus Superior frontal gyrus

Lateral ventricle Thalamus

Putamen

Tail of hippocampus

Internal carotid artery

Body of hippocampus Cerebellum

Maxillary sinus

Temporal horn of lateral ventricle

Head of hippocampus

B

FIG 8-16  A, Sagittal T1 MRI scan through the body of the lateral ventricle. B, Sagittal T2 MRI scan through the body of the lateral ventricle.

http://radiologyebook.com

155

156

PART II  CT and MR Imaging of the Whole Body Precentral gyrus

Middle frontal gyrus

Central sulcus Posterior short insular gyri

Postcentral gyrus Anterior long insular gyri

Middle short insular gyri

Posterior long insular gyri

Anterior short insular gyri

Heschl gyri

Middle cerebral artery

Lateral ventricle Horizontal fissure

Middle temporal gyrus

Cerebellum Hippocampus

Maxillary sinus

Internal acoustic meatus

Inferior temporal gyrus

Apex of insula

A Internal carotid artery Middle frontal gyrus

Temporal horn of lateral ventricle Precentral gyrus Central sulcus

Posterior short insular gyri

Postcentral gyrus Anterior long insular gyri

Middle short insular gyri

Posterior long insular gyri

Anterior short insular gyri

Heschl gyri

Apex of insula Middle cerebral artery

Lateral ventricle Superior semilunar lobule

Middle temporal gyrus

Horizontal fissure Maxillary sinus Cerebellum Inferior temporal gyrus

Inferior semilunar lobule

B

Hippocampus Internal carotid Temporal horn artery of lateral ventricle

Internal acoustic meatus

FIG 8-17  A, Sagittal T1 MRI scan at the lateral orbital level. B, Sagittal T2 MRI scan at the lateral orbital level.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Anterior arch of atlas Odontoid process

Lateral mass of atlas

Foramen magnum

Occiput

A

Occiput

Lateral mass of atlas

Odontoid process

Intervertebral space

Transverse process

Odontoid vertebra

B FIG 8-18  A, Axial CT scan through the articulation of C1 with C2. The spatial relation of the odontoid process with the anterior arch and lateral mass of atlas is clearly identified. B, Coronal reconstruction well delineates the articulation of the convex occipital condyles with the concave superior articular facets of C1, and the articulation of C1 with C2. Continued

http://radiologyebook.com

157

158

PART II  CT and MR Imaging of the Whole Body

Anterior arch of atlas Odontoid process Posterior arch of atlas Odontoid vertebra Spinous process Intervertebral space

Spinal canal

C7 vertebral body

C FIG 8-18, cont’d  C, Sagittal reconstruction clearly demonstrates the relation of the odontoid process and anterior arch of atlas.

Vertebral body Lamina Hyoid bone

Sternocleidomastoid muscle

Transverse foramen Transverse process

Anterior subarachnoid space Spinous process

Cervical cord Semispinalis cervicis muscle Multifidus muscle Trapezius muscle

A

FIG 8-19  A, Axial CT scan through a cervical vertebral body showing the transverse foramina for passage of the vertebral arteries.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

159

Intervertebral foramen External jugular vein Vertebral body Carotid artery Anterior subarachnoid space

Cervical cord

Sternocleidomastoid muscle

B FIG 8-19, cont’d  B, Axial CT scan through the intervertebral foramen in the midcervical region. The large amount of cerebrospinal fluid in front and back of the cervical cord help delineate the cervical cord and roots. (Courtesy M. Hossain Naheedy.)

CT well presents cortical bone and the trabecular structure of  the vertebral bodies with higher attenuation value, whereas MRI demonstrates these structures by their signal void (see Fig. 8-20B). On axial CT scans, it should not be mistaken for a destructive process that the central part of the vertebral body shows decreased density due to partial volume averaging of the end plate and disk space (see Fig. 8-20A). With MRI studies, vertebral bodies containing yellow marrow appear isointense or high signal on T1 images and low signal on T2 images.

Pedicles (Fig. 8-22).  Pedicles extend from the vertebral bodies posteriorly to the articular pillars and form the neural arch enclosing the spinal canal posteriorly. They are short in the cervical area (see Fig. 8-19A). In the thoracic area (see Fig. 8-20A), the pedicles arise from the upper half of the vertebral bodies, resulting in a higher relative position of the intervertebral foramina in relation to disk spaces than those in the cervical area. In the lumbar region (see Fig. 8-21A), the superior portion of the pedicle is larger than the inferior portion. On axial scans above or below the pedicles, the osseous canal appears incomplete.

Laminae (see Fig. 8-20).  The laminae consist of two bony plates extending posteriorly and medially to unite in the midline at each level and form the spinous process. In the cervical area, the laminae become thicker from C3 to C7. In the thoracic area, the laminae are broader and shorter, and they overlap in the adjacent vertebrae. In the lumbar region, the laminae are relatively flat and do not overlap each other. A larger space is found between adjacent laminae in the upper lumbar spine than in the lower lumbar spine.

Spinous Processes (see Figs. 8-20 and 8-21).  The spinous processes protrude posteriorly and downward from the fusion of two laminae. In midcervical areas, they frequently appear bifid. In the thoracic area, they are slender and longer, with a triangular shape in the coronal plane. In the lumbar area, they appear rectangular in the sagittal plane.

Transverse Processes (see Figs. 8-20A and 8-21A).  In the cervical area the transverse processes consist of two components,  one extending anterolaterally and one posterolaterally from the vertebral bodies to form the transverse foramina. In these foramina run  the vertebral arteries and small veins from the C7 to the C2 level (see Fig. 8-19A). In the thoracic region (see Fig. 8-20A) the transverse processes project outward, upward, and posteriorly. They are close to the heads, necks, and tubercles of the corresponding ribs. In the lumbar area the transverse processes extend laterally and posteriorly. The transverse processes of L1 and L5 are thicker and shorter than those of the remaining lumbar vertebrae.

Articulating Pillars.  The articulating pillar is formed by the superior and inferior facets and projects from the junction of the pedicle and lamina.

Intervertebral Foramina (Fig. 8-23; see Figs. 8-21 and 8-22).  The intervertebral foramina, also called neural foramina, are outlined by the inferior pedicle cortex superiorly, the superior pedicle cortex inferiorly, the posterior aspect of the vertebral bodies anteriorly,  and the posterior elements posteriorly. Within the foramina are the corresponding spinal nerves, veins, fat, arteries, and connective tissue. In the thoracic area the foramina are at the level of the inferior half of the vertebral bodies. In the lumbar region the foramina are larger and longer. The neural foraminal ligaments of the lumbar spine start from or near the intervertebral disk and cross the foramen to attach on the superior articular process, pedicle, transverse process, or ligamentum flavum. The ligaments are most conspicuous in the inferior portion of the foramen. Identification of these ligaments on CT scans and MRIs is useful for the diagnosis of lateral spinal stenosis. Within each neural foramen the isointense nerve root is outlined by the low signal intensity of CSF and the high signal intensity of the fat in the epidural space on T1 images (see Fig. 8-23). In contradistinction the higher-signal-intensity CSF outlines the spinal nerve roots and cord on proton-density and T2 images (see Fig. 8-23).

http://radiologyebook.com

Text continued on p. 165

160

PART II  CT and MR Imaging of the Whole Body

Vertebral body Intervertebral disk Rib

Pedicle

Dural sac Transverse process

Epidural fat

Lamina

Spinous process

A

Vertebral body

Spinal cord Superior articulating facet

Rib Subarachnoid space

Facet joint

Epidural fat

Inferior articulating facet

Spinous process

Lamina

B FIG 8-20  A, Axial CT scan through the thoracic vertebra demonstrating the full circle of bony canal made by the vertebral body, pedicles, laminae, and spinous process. The costovertebral articulation is also clearly seen. B, Axial T1 MRI scan at the midthoracic level. The thoracic cord appears rounded. Note the larger subarachnoid space posteriorly with low-signal cerebrospinal fluid and large amount of high-signal fat in the epidural space posterior to the thecal sac.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Vertebral body Psoas muscle

Dural sac Pedicle

Superior articulating facet

Transverse process

Facet joint

Ligamentum flavum

Inferior articulating facet Multifidus muscle

Spinous process

A

Intervertebral disk

Intervertebral foramen

Dural sac Superior articulating facet

Facet joint Inferior articulating facet

Lamina Epidural fat

Ligamentum flavum

Spinous process

B FIG 8-21  A, Axial CT scan through the upper half of the vertebral body in the lumbar region demonstrating the lumbar vertebra larger than the thoracic, and slightly high density of ligamentum flavum. B, Axial CT scan through the intervertebral foramen in the midlumbar region. Minimal averaging of the cortex causes increased attenuation value at the periphery. With a narrow window setting, the facet joints cannot be clearly seen, but slightly high density of ligamentum flavum is easily identified. Low attenuation of fat and the slightly increased attenuation of ligamentum flavum help delineate the border of the thecal sac. Continued

http://radiologyebook.com

161

162

PART II  CT and MR Imaging of the Whole Body

Intervertebral disk

Dural sac

Intervertebral foramen

Superior articulating facet

Facet joint Lamina

Inferior articulating facet

Spinous process

C FIG 8-21, cont’d  C, Axial CT scan through intervertebral foramen. The wide window setting allows clear visualization of low attenuation of the facet joint between the facets and its constructive components.

Pedicle

Vertebral body

Intervertebral space

Intervertebral foramen Superior articulating facet

Facet joint Inferior articulating facet

A FIG 8-22  A, Parasagittal reconstructionhttp://radiologyebook.com of CT scan in the thoracic region demonstrating the pedicle, intervertebral foramen, and facet joint.

CHAPTER 8  Normal Anatomy

Vertebral body

Pedicle

Intervertebral disk

Nerve root

Epidural fat Superior articulating facet

Facet joint

Inferior articulating facet

B

Vertebral body

Pedicle

Intervertebral disk

Nerve root

Epidural fat Superior articulating facet Facet joint Inferior articulating facet

C FIG 8-22, cont’d  B, Parasagittal T1 MRI in the lumbar region. C, Parasagittal gradient echo study of the lumbar spine.

http://radiologyebook.com

163

164

PART II  CT and MR Imaging of the Whole Body

Vertebral body

Epidural fat Nerve root ganglion

Anterior subarachnoid space Epidural fat

Posterior nerve root

Cervical cord

Subarachnoid space Spinous process

A

Vertebral body Nerve root ganglion Anterior subarachnoid space Anterior median fissure

Subarachnoid space Posterior nerve root

Cervical cord

Posterior subarachnoid space

Spinous process

B FIG 8-23  A, Axial T1 MRI scan through the intervertebral foramen in the cervical region. The nerve root ganglion is outlined by the high-signal fat and lower-signal cerebrospinal fluid (CSF) within the intervertebral foramen. B, Axial T2 MRI scan through the intervertebral foramen in the cervical region. The high signal intensity of CSF outlines the low signal intensity of the spinal nerve roots and cord (similar to the appearance on intrathecal contrast-enhanced CT scan).

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

165

Thoracic vertebral body

Anterior nerve root

Thoracic cord

Nerve root ganglion

Ligamenta flava

Epidural fat Spinous process

Posterior nerve root

C FIG 8-23, cont’d  C, Axial T1 MRI scan through the intervertebral foramen in the thoracic region. There is abundant epidural fat posteriorly and laterally at the neural foramen but little anteriorly in the epidural space. The epidural fat outlines the anterior and posterior nerve roots as well as the nerve root ganglion.

On CT scans the nerve root and ganglion have the water attenuation value within the intervertebral foramina, and some low attenuation of the fat appears in the epidural space (see Fig. 8-21).

Spinal Canal.8,37,41  In the cervical area, the diameter of the spinal canal is almost uniform from C2 to C7 except for being largest at the C1 level. The upper limit of the canal is about 27 mm at the C1 level and 21 mm in the lower cervical area in the sagittal midline plane. The lower limit is 12 mm in the lower cervical area and 15 to 16 mm at C1-C2.37 In the thoracic area the canal has fairly constant size and contour. The transverse and anteroposterior diameters of the canal are the same. The canal appears round except in the lower region, where it is triangular. In the lumbar area the canal is round or oval in the upper lumbar region and triangular in the middle and lower lumber regions. The anteroposterior diameters of the canal are about 15 to 25 mm at the L1-L5 levels. CT scans well demonstrate the osseous canal, outlined by the bony margins of the spinous processes, laminae, pedicles, and posterior aspects of the vertebral body (see Figs. 8-19A and 8-20A). Also the anteroposterior diameter of the canal can easily be measured on MR sagittal images (Figs. 8-24, 8-25, and 8-26).

Soft Tissues, Joints, and Ligaments

Intervertebral Disk.  The intervertebral disk consists of two components—the nucleus pulposus, composed of a hydrophilic mucopolysaccharide/collagen gel, and the anulus fibrosus, composed of concentric layers of collagen fibers. The nucleus pulposus, softer with greater water content, is surrounded by the anulus fibrosus with hard fibrous tissue.

In the cervical region, intervertebral disks separate the two adjacent vertebral bodies, but there is no disk space between C1 and C2. The disk is smaller laterally than is the vertebral body, and it is thicker anteriorly than posteriorly. Also the disk is smaller and intermediate in thickness compared to those of the thoracic and lumbar regions. The thoracic disk has a uniform thickness and is thinner but  larger in the cross-sectional area. The lumbar disks are thicker and vary in height from 8 to 12 mm. They are kidney shaped with a concavity in the posterior margin. Unlike lumbar disks the height of the lumbosacral disk is less than 10 mm and frequently less than 5 mm. It appears convex in the posterior margin, bulging into the canal centrally. On CT scan, a disk has a water attenuation value close to that of muscle (see Figs. 8-21B and C). The disk cannot be distinguished from ligaments because of their similar attenuation values. On T1 images a disk shows low intensity (see Figs. 8-24A, 8-25A, and 8-26A), with increased intensity on gradient echo images (see Figs. 8-24B, 8-25B, and 8-26B) and T2 images. Spin echo T2-weighted imaging demonstrates the nuclei of the disks with variable intensity and shows the low signal of the anulus fibrosus, posterior longitudinal ligament, and dura with indistinguishability. Gradient refocused T2-weighted imaging is useful in differentiating the high signal of disk from the low signal of bone and yields an exquisite bone-CSF interface. On sagittal images the relationship of the anterior and posterior margins of the disk with the anterior and posterior longitudinal ligaments is well documented (see Fig. 8-26). Interruption of the ligament may suggest a tear resulting from a herniated disk.

Uncovertebral Joints.4  The uncovertebral joint is formed between the uncinate process and the depression in the inferior aspect of the

http://radiologyebook.com

166

PART II  CT and MR Imaging of the Whole Body

Posterior arch of atlas

Odontoid process Anterior arch of atlas

Posterior subarachnoid space

Odontoid vertebra Anterior subarachnoid space

Cervical cord

Intervertebral disk

Nuchal ligament

Spinous process Vertebral body Subcutaneous fat

Posterior longitudinal ligament

Posterior epidural fat

A

Posterior arch of atlas

Odontoid process Anterior arch of atlas

Posterior subarachnoid space

Odontoid vertebra Anterior subarachnoid space

Cervical cord

Nuchal ligament

Intervertebral disk

Spinous process Vertebral body Subcutaneous fat Posterior longitudinal ligament

Posterior epidural fat

B

FIG 8-24  A, Sagittal T1 MRI in cervical area. The intervertebral disk has slightly higher intensity relative to outlining cortical bone and the ligaments. The diameter of spinal canal is almost uniform from C2 to C7. B, Sagittal gradient echo study of cervical spine showing increased signal intensity of cerebrospinal fluid similar to that in a T2 MRI study. The healthy intervertebral disk has higher signal intensity due to its water content. The anterior and posterior longitudinal ligaments can readily be seen outlining the spinal canal.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy

Vertebral body Spinous process

Intervertebral disk

Posterior epidural fat

Thoracic cord

Posterior subarachnoid space

Anterior subarachnoid space

A

Vertebral body Spinous process

Intervertebral disk

Posterior epidural fat Thoracic cord Anterior subarachnoid space

Posterior subarachnoid space

B FIG 8-25  A, Sagittal T1 MRI study of the thoracic spine. There is more cerebrospinal fluid (CSF) posteriorly in the thoracic area. The spinal cord can be measured more accurately with this technique than with gradient echo or T2 studies. B, Sagittal gradient echo study of the thoracic spine. The cord is enlarged from T9 to T12, outlined by high-signal CSF. The high signal intensity indicates a healthy disk. The very low signal intensity of the cortical bone helps visualize the upper and lower border of the intervertebral disk.

http://radiologyebook.com

167

168

PART II  CT and MR Imaging of the Whole Body Spinal cord

Conus medullaris

Subarachnoid space Posterior epidural fat

Vertebral body

Cauda equina

Intervertebral disk

Interspinous ligament

Anterior longitudinal ligament

Spinous process Posterior longitudinal ligament

Anterior epidural fat

A Spinal cord

Conus medullaris

Subarachnoid space Posterior epidural fat

Vertebral body

Cauda equina

Intervertebral disk

Interspinous ligament

Anterior longitudinal ligament

Spinous process Posterior longitudinal ligament

Anterior epidural fat

B FIG 8-26  A, Sagittal T1 MRI study of the lumbar spine in midline. The epidural fat behind the thecal sac and in front of the distal lumbar area shows high signal intensity. The intervertebral disk space has slightly lower intensity relative to the vertebral body. B, Sagittal gradient echo study of the lumbar area in midline. The cerebrospinal fluid has slightly increased signal intensity. The anterior and posterior longitudinal ligaments are seen outlining the vertebral bodies and intervertebral disk spaces. The interspinous ligaments connect the two adjacent spinous processes.

http://radiologyebook.com

CHAPTER 8  Normal Anatomy vertebral body from C2 to C7 level. The space between the uncinate process and adjacent vertebral body is narrow and filled with loose connective tissue.

169

area there are two sets of the articulations—that is, the joint spaces between the heads of the ribs and the vertebral bodies, as well as between the tubercles of the ribs and the transverse processes. The costotransverse articulations lie posterior and lateral to the costovertebral articulations, the joints between the ribs and the vertebral bodies, in a plane behind the articular pillar. Moreover, these ribs—like ribs 11, 12, and occasionally 10—do not articulate with the transverse processes.

shows high signal intensity on T2 images. Gradient refocused T2-weighted images are superior to spin echo T2-weighted images in delineation of the cartilage of facet joints. With CT scans the joint space appears outlined by the uniformly thick articular margins of the adjacent articular processes. The dense cortex of the articulating facet is easily separated by the low attenuation value of the facet joint space (see Fig. 8-21). Sagittal reconstruction of CT scan and sagittal MRIs are useful for evaluation of the facet joints and their orientation. Each facet joint receives innervation from the posterior primary ramus of the posterior root ganglion at its level and the level above.29 Thus the facet joint diseases may produce back pain, which should be differentiated from the cause of disk herniation with nerve root compression.

Facet Joints (Fig. 8-27; see Figs. 8-21 and 8-22).  Facet joints are

Ligaments

Costovertebral Articulations (see Fig. 8-20A).  In the thoracic

formed between the superior articular facet of the lower vertebra and the inferior facet of the upper vertebra. In the cervical area they are oriented in the oblique plane halfway between the axial and coronal planes. The superior articular facet of the lower vertebra is always anterior to the inferior facet of the upper vertebra. In the thoracic region the facet joints are oriented approximately in the coronal plane. The facet joints are oriented in a parasagittal plane in the upper lumbar area but are more oblique in the lower lumbar area. The joint spaces are narrow, with a range of 2 to 4 mm.4 The joint surfaces are lined with articular cartilage. The synovial space extends medially just beneath the anterior margin of the ligamentum flavum. The facet capsule and the anterior portion of the ligamentum flavum cannot be separately delimited on CT scans. The articular cartilage is clearly visible on T1 and proton-density images, which produce a moderately bright signal (see Fig. 8-27). The cartilage

Anterior Longitudinal Ligament (see Figs. 8-22 to 8-26).  This ligament extends from the axis as the anterior atlantoaxial ligament to the sacral segments, connecting the anterior aspects of the vertebral bodies and disk spaces (see Fig. 8-26). On CT scans the ligament cannot be distinguished from the bony structures and the disk space, whereas T1 MRIs present it as a line of low signal, characteristic of fibrous structures outlining the anterior aspect of the vertebral bodies and disk spaces in the sagittal plane.

Posterior Longitudinal Ligament (see Figs. 8-24 to 8-26).  This ligament begins on the posterior surface of the axis and extends to the sacrum. It widens and becomes adherent at the disk space and at the posterior aspects of the vertebral body. In the cervical spine the posterior longitudinal ligament extends cephalad to merge into the tectorial membrane and dura mater. Anterior to this, just behind the dens, are

Intervertebral disk Subarachnoid space Intervertebral foramen Superior articulating facet

Nerve plexus Facet joint

Lamina

Inferior articulating facet

Spinous process

FIG 8-27  Axial proton-density MRI scan through the facet joint in the lumbar area. The articular cartilage lining on the superior and inferior facets is clearly visible, appearing as a moderately bright signal in this image.

http://radiologyebook.com

170

PART II  CT and MR Imaging of the Whole Body

Pons

Superior cerebellar cistern

Prepontine cistern Cerebellar vermis

Clivus

Cisterna magna Anterior arch of atlas

Cerebellar tonsil

Odontoid process

Medulla

Anterior subarachnoid space

Posterior subarachnoid space Cervical cord

Spinous process

FIG 8-28  Spinal reconstruction of a CT scan with administration of intrathecal contrast agent at the craniocervical junction. The upper cervical cord, medulla, pons, cerebellar tonsils, and vermis are outlined clearly by contrast material.

the inferior and superior cruciate ligaments, which merge to form the transverse ligament. The posterior longitudinal ligament may serve as a barrier to posterior disk herniation. It appears as a soft tissue density on CT and as low signal intensity on T1 MRIs (see Fig. 8-26A).

Interspinous Ligament (see Fig. 8-26).  This ligament connects the spinous processes. It is easily seen on CT scans and T1 MRIs.

Nuchal Ligament.  This ligament connects the base of the occipital bone to the spinous processes of C1 to C7 (see Fig. 8-24).

Ligamentum Flavum.  This ligament extends from the axis to the sacrum and attaches at the ventral surface of the lamina of the upper vertebra and at the dorsal superior surface of the lamina below. It has both stretchability and retractability functions. It measures 3 to 5 mm in thickness and appears as a higher attenuation value on CT scans (see Fig. 8-21) and as a low signal intensity on T1 MRIs (see Fig. 8-23C). Hypertrophy of the ligament results in narrowing of the canal posteriorly and laterally.

Epidural Space (see Figs. 8-20, 8-21, and 8-23).  The epidural space is formed between the dura and the spinal canal. It contains fat, vessels, and neural elements. In the cervical area the spinal nerve can be demonstrated by CT, owing to plenty of fat in the upper portion of the neural foramen and in the perivertebral region. Within the space is the internal vertebral venous plexus with no valve, thereby facilitating the spread of infectious processes and metastasis. The basivertebral vein penetrates the vertebral body to join the internal venous plexus. These veins cannot be identified as separate structures by nonenhanced CT scans. Therefore dural and epidural venous enhancement by IV injection of contrast medium may provide useful anatomic landmarks and help evaluate extradural diseases. In the thoracic region there is abundant epidural fat posteriorly between the neural arch and the dura and laterally at the neural foramen15 but little anteriorly in the epidural space (see Figs. 8-20 and 8-23). The pattern of the thoracic epidural vein is similar to the pattern in the cervical and lumbar areas, except that there is a possible link between the anterior and posterior vertebral plexus with the intercostal veins and the azygos system in the thoracic region.35 In the lumbar area, epidural fat can normally be seen anteriorly and posteriorly at all lumbar disk levels. Obliteration of fat in the epidural space is suggestive

of an epidural process. Therefore the fat within the epidural space and the intervertebral foramina is helpful for evaluating disease processes affecting the margins of the spinal canal. In addition, conjoined nerves, two nerves exiting from the thecal sac at the same level, is an anomaly seen in up to 1% of the population and most frequently involves the L5 and S1 nerves.

Dura Mater, Intradural Space, and Spinal Cord.  The dura mater forms a sleeve around the subarachnoid space, covering the cord and intracanal component of the nerve roots. It extends beyond the canal to fuse with the perineurium of the spinal nerves (see Fig. 8-23B). The cervical cord is slightly flat on the ventral surface, indented by  the anterior median fissure (see Fig. 8-23B). Its posterior surface is indented by the shallow posterior median fissure.36 Because of the large subarachnoid space in the cervical area, the cord can be seen on plain CT scans (see Fig. 8-19). With intrathecal contrast, visualization of the cervical cord and roots can be enhanced significantly (Fig. 8-28). Eight pairs of dorsal and ventral roots arising from the cervical cord unite lateral to the dural sheath to form the spinal nerves. The first and second cervical spine nerves pass posterior to the atlantooccipital and atlantoaxial joints. The remaining nerves pass through the intervertebral foramina. In the thoracic area the cord appears round30,35 and is enlarged from T9 to T12,37 terminating in the conus with rapid tapering in its diameter. Its ventral surface is indented by an anterior median fissure. On the dorsal cord surface the posterointermediate and posterolateral sulci appear shallow and indistinct. In the upper thoracic region the cord segment is about two levels lower in number than the corresponding vertebra. The difference increases to three vertebrae in the lower thorax. In the lumbar area the dura terminates at the level of S2 and fuses with the filum terminale to end at the coccyx. The lumbar canal contains the conus, cauda equina, and filum terminale. The conus medullaris, continuous with the spinal cord, ends at the L1-L2 level with a taper (see Fig. 8-26) and then becomes the filum terminale. Normally the conus is above the mid-L2 level and the filum terminale is 2 mm or less in thickness at the level of the L5-S1 interspace. Generally a conus level below the mid-L2 in a person older than 12 years is referred to as an anomaly.29 On T1 images the spinal cord is clearly seen because of its isointensity surrounded by the low signal intensity of CSF (Figs. 8-25A and 8-26A). On T2 images these signals reverse, and the

http://radiologyebook.com

CHAPTER 8  Normal Anatomy decreased signal cord and roots of the cauda equina are outlined by high-signal CSF (Figs. 8-25B and 8-26B). Unlike the thoracic spinal nerves, the lumbar spinal nerves can be seen on plain CT scans, exiting medial to the corresponding pedicles.

REFERENCES 1. Afifi AK, Bergman RA: Functional neuroanatomy, New York, 1998, McGraw-Hill. 2. Berman SA, Hayman LA, Hinck VC: Correlation of CT cerebral vascular territories with function: 1. Anterior cerebral artery. AJNR Am J Neuroradiol 1:259–263, 1980. 3. Berry I, Sigal R, Lebas J, et al: Magnetic resonance imaging: Principles, technique and imaging protocols. In Manelfe C, editor: Imaging of the spine and spinal cord, New York, 1992, Raven Press, pp 157–194. 4. Carrera GF, Haughton VM, Syvertsen A: Computed tomography of the lumbar facet joints. Radiology 134:145–148, 1980. 5. Cohen CR, Duchesneau PM, Weinstein MA: Calcification of the basal ganglia as visualized by computed tomography. Radiology 134:97–99, 1980. 6. Donovan Post JM, Sze G, Quencer RM: Gadolinium-enhanced MR in spinal infection. J Comput Assist Tomogr 14:721–729, 1990. 7. Drayer BP, Burger P, Dawrin R, et al: MRI of brain iron. AJNR Am J Neuroradiol 7:373–380, 1986. 8. Dryer PB, Rosenbaum AE, Higman HB: Cerebrospinal fluid imaging using serial metrizamide CT cisternography. Neuroradiology 13:7–17, 1977. 9. Epstein BS: The spine: A radiological text and atlas, Philadelphia, 1976, Lea & Febiger. 10. Fix JD: Neuroanatomy, ed 2, Baltimore, 1995, Williams & Wilkins. 11. Goss C, editor: Gray’s anatomy, ed 29, Philadelphia, 1973, Lea & Febiger. 12. Hanaway J, Scott W, Strother C: Atlas of the human brain and the orbit for computed tomography, St. Louis, 1977, Warren Green. 13. Harwood-Nash DC, Fitz CR: Neuroradiology in infants and children, vol 2, St. Louis, 1976, CV Mosby, pp 484–486. 14. Haughton VM, Syvertsen A, Williams AL: Soft tissue anatomy within the spinal canal as seen on computed tomography. Radiology 134:649–655, 1980. 15. Haughton VM, Williams AL: CT anatomy of the spine. CRC Diagn Imaging 15:173–192, 1981. 16. Hayman LA, Berman SA, Hinck VC: Correlation of CT cerebral vascular territories with function: II. Posterior cerebral artery. AJR Am J Roentgenol 137:13–19, 1981. 17. Holodny AI, George AE, Golomb J, et al: The perihippocampal fissures: Normal anatomy and disease states. Radiographics 18:653–665, 1998. 18. Kido DK, LeMay M, Levinson AW, et al: Computed tomographic localization of the precentral gyrus. Radiology 135:373–377, 1980. 19. Kido DK, O’Reilly GVA, Naheedy MH: Radiological perspective on idiopathic low back pain. In White AA, Gordon SL, editors: Symposium on idiopathic low back pain, St. Louis, 1982, CV Mosby, pp 178–194. 20. Kjos BO, Norman D: Strategies for efficient imaging of the lumbar spine. In Brant-Zawadzki M, Norman D, editors: Magnetic resonance imaging of the central nervous system, New York, 1987, Raven Press, pp 279–287. 21. Labischong P, et al: Normal anatomy of the spine, spinal cord, and nerve roots. In Manelfe E, editor: Imaging of the spine and spinal cord, New York, 1992, Raven Press, pp 1–91.

171

22. LaMasters DL, et al: Computed tomography of the spine and spinal cord. In Newton TH, Potts DG, editors: Modern neuroradiology, vol 1, San Anselmo, CA, 1983, Clavadel Press, pp 53–113. 23. Macpherson P, Matheson MS: Comparison of calcification of pineal, habenular commissure, and choroid plexus on plain films and computed tomography. Neuroradiology 18:67–72, 1979. 24. Masdeu JC, Grossman BC, editors: Head and spine imaging, New York, 1985, John Wiley & Sons. 25. Matsui T, Hirano A: An atlas of the human brain for computed tomography, New York, 1978, Igaku-Shoin. 26. Naidich TP, Moran CJ, Pudlowski RM, et al: Advances in diagnosis: Cranial and spinal computed tomography. Med Clin North Am 63:849–895, 1979. 27. Norman D, Diamond C, Boyd D: Relative detectability of intracranial calcifications on computed tomography and skull radiography. J Comput Assist Tomogr 2:61–64, 1978. 28. Parent A: Carpenter’s human neuroanatomy, ed 9, Baltimore, 1996, Williams & Wilkins. 29. Raghavan N, Barkovich AJ, Edwards M, et al: MR imaging in the tethered spinal cord syndrome. AJR Am J Roentgenol 152:843–852, 1989. 30. Resjo IM, Harwood-Nash DC, Fitz CR: Normal cord in infants examined with computed tomographic metrizamide myelography. Radiology 130:691–696, 1979. 31. Ross JS, Modic MT, Masaryk TJ, et al: Assessment of extradural degenerative disease with Gd-DTPA-enhanced MR imaging: Correlation with surgical and pathologic findings. AJNR Am J Neuroradiol 10:1243– 1249, 1989. 32. Russell EJ, D’Angelo CM, Zimmerman RD, et al: Cervical disc herniation: CT demonstration after contrast enhancement. Radiology 152:703–712, 1984. 33. Shapiro R, Galloway SJ, Shapiro MD: Minimal asymmetry of the brain: A normal variant. AJR Am J Roentgenol 147:753–756, 1986. 34. Sherman JL, Citrin CM: MR demonstration of normal CSF flow. AJNR Am J Neuroradiol 7:3–6, 1986. 35. Taylor AJ, Haughton VM, Doust BD: CT imaging of the thoracic spinal cord without intrathecal contrast medium. J Comput Assist Tomogr 4:223–224, 1980. 36. Theron J, Moret J: Spinal phlebography: Lumbar and cervical techniques, New York, 1978, Springer-Verlag. 37. Thijssen HO, Rombouts JJ, Walder HA: Morphology of the cervical spinal cord on computed myelography. Neuroradiology 18:57–62, 1979. 38. Truex RC, Carpenter MB: Human neuroanatomy, ed 6, Baltimore, 1969, Williams & Wilkins. 39. Weiss KL, Storrs J, Weiss JL, et al: CT brain prescriptions in Talairach space: A new clinical standard. AJNR Am J Neuroradiol 25:233–237, 2004. 40. Williams PL, Warwick R: Functional neuroanatomy of man, Philadelphia, 1975, WB Saunders. 41. Wolpert SM: Appropriate window settings for CT anatomic measurements. Radiology 132:775, 1979. 42. Zatz LM: Basic principle of computed tomography scanning. In Newton TH, Potts DG, editors: Radiology of the skull and brain: technical aspects of computed tomography, vol 5, St. Louis, 1981, CV Mosby, pp 3853–3876. 43. Zimmerman RA, Bilaniuk LT: Age-related incidence of pineal calcification detected by computed tomography. Radiology 142:659–662, 1982.

http://radiologyebook.com

9  Intracranial Neoplasms Kristine A. Blackham, Jesse Montagnese, Stephen A. Kieffer, and Jeffrey R. Brace

INTRODUCTION Based on data collected from several central cancer registries during the years 2007-2011, the Central Brain Tumor Registry of the United States (CBTRUS) estimates that 67,900 new cases of primary brain and central nervous system (CNS) tumors will be reported in 2014—one third of which will be malignant.282 The brain and CNS are also common sites of secondary tumor implantation; metastases may occur up to 10 times more frequently than primary brain tumors.264 The 2014 CBTRUS statistical report found that brain and CNS tumors are the most common neoplasm among those aged 0 to 19 years, with an average age-adjusted incidence rate slightly higher than leukemia. Children and young adults generally have better survival outcomes for most histologies; survival estimates have a large variation depending on tumor histology, ranging from a 5-year survival rate of 94% for pilocytic astrocytoma to 5% for glioblastoma. The fourth edition of the World Health Organization (WHO) classification of tumors of the CNS was published in 2007 and continues to improve and expand histologic and clinical diagnostic criteria that enable the possibility of worldwide communication, clinical trials, and epidemiologic studies. In 2007 several new entities and histologic variants were added and genetic profiles were updated and discussed in a review article.230 The WHO-edited book publication includes concise commentary on the clinicopathologic characteristics of each tumor type.229 Box 9-1 is a list of tumors discussed in this chapter, which have been organized into clinically and imaging-relevant histology groups and reflect the 2007 WHO classification. Although many brain and CNS tumors are not histologically confirmed, particularly those with benign behavior, the most frequently reported histology overall in the United States from 2014 CBTRUS data was meningioma (36%) followed by glioblastoma (15%). Tumors of the pituitary gland and nerve sheath tumors combined account  for about one fourth of all tumors, the vast majority of which are nonmalignant.

CLINICAL PRESENTATION The presenting symptoms in most patients with intracranial neoplasms are often nonspecific. Headache has been shown to occur in a majority of patients with primary brain tumors; indeed, headache is a common initial presentation for patients with high-grade gliomas. The headache is frequently described as diffuse and more severe in the morning and often dissipates in a few hours even without treatment. When unilateral, it can accurately indicate the hemisphere involved by tumor. Episodes of confusion or change in behavior also commonly bring the patient to the physician but can be confused with depression and result in delay of diagnosis. It is the persistent and progressive

172

nature of the symptoms, often with development of visual disturbances or protracted nausea and vomiting, that eventually leads to consideration of brain tumor as a diagnostic possibility. Less common but more declarative presentations are focal neurologic symptoms (e.g., unilateral weakness, aphasia) and onset of seizures, the type and associated symptoms of which vary by tumor location. Seizures occur more frequently in low-grade gliomas and meningiomas than in higher-grade gliomas and lymphomas.81,112,312 Intracranial neoplasms produce symptoms by three mechanisms: (1) infiltration and destruction of normal tissues (e.g., by a glioblastoma), (2) localized cortical irritation or depression (e.g., by a convexity meningioma), or (3) expansion within the rigid and unyielding cranial vault. Edema of the white matter adjacent to the tumor is often extensive and may account for much of the patient’s symptomatology and neurologic deficit, an impression that is verified by the striking improvement, both clinically and on imaging, exhibited by many patients after systemic administration of corticosteroids. An increase in intracranial pressure above the normal level of 180 mm H2O commonly accompanies an expanding intracranial mass. Headache, nausea, vomiting, and sixth cranial nerve palsy may reflect increased intracranial pressure. Prolonged or significantly elevated intracranial pressure manifests clinically as loss of attentiveness, apathy, and drowsiness. These signs of depressed cerebral function are most likely caused by diminished cerebral blood flow secondary to distention of the intracranial contents. This distention also produces traction on the overlying meninges, probably accounting for the headache, nausea, and vomiting that occur with chronic intracranial hypertension.76,81,108

DIAGNOSTIC IMAGING Brain imaging is routinely performed using computed tomography (CT) or magnetic resonance imaging (MRI), typically after intravenous (IV) administration of iodinated contrast media (CT) or gadolinium contrast (MRI). The goals of diagnostic imaging in the patient with suspected intracranial tumor include detection of the presence of a neoplasm, localization of the extent of the tumor (including definition of involvement of key structures and assessment of the presence and severity of secondary changes [e.g., edema, herniation, hemorrhage]), and characterization of the nature of the process. Imaging (usually MRI) is also extensively used in treatment planning before surgery or radiation to define the gross tumor margins and permit selection of the safest approach to the lesion. Increasingly, various functional/physiologic applications of MRI are employed preoperatively to determine the location of functionally eloquent cortex and critical white matter tracts, which can then be avoided during surgery. Interactive computerized neuronavigational devices use these data (1) to allow administration of very high doses of precisely focused

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

173

BOX 9-1  Classification of Tumors of the Nervous System Tumors of the Cranial Nerves Schwannoma Neurofibroma Malignant peripheral nerve sheath tumors

Gliomas Astrocytic Tumors Circumscribed Pilocytic astrocytoma Pleomorphic xanthoastrocytoma Subependymal giant cell astrocytoma Diffuse Low-grade astrocytoma Optic pathway glioma Brainstem glioma Anaplastic astrocytoma Glioblastoma Gliomatosis cerebri Gliosarcoma

Tumors of the Meninges Meningioma Melanocytic tumors (melanocytoma, malignant melanoma, meningeal melanomatosis, diffuse melanocytosis) Hemangioblastoma Mesenchymal tumors Lipoma Chondrosarcoma Rhabdomyosarcoma Hemangiopericytoma

Oligodendrogliomas Low-grade oligodendroglioma Anaplastic oligodendroglioma

Lymphomas and Hematopoietic Neoplasms Primary CNS lymphoma Secondary CNS lymphoma Granulocytic sarcoma (chloroma)

Ependymomas Ependymoma Anaplastic ependymoma Subependymoma

Germ Cell Tumors Germinoma Teratoma Other germ cell tumors (embryonal carcinoma, yolk sac tumor, and choriocarcinoma)

Choroid Plexus Tumors Choroid plexus papilloma Choroid plexus carcinoma

Tumors of the Sellar Region Pituitary adenoma Microadenoma Macroadenoma Craniopharyngioma Rathke’s cleft cyst Tumors of the neurohypophysis (granular cell tumor, Langerhans cell histiocytosis) Chordoma

Nonglial Tumors Neuronal and Mixed Neuronal/Glial Tumors Ganglioglioma Gangliocytoma Dysplastic gangliocytoma of the cerebellum (Lhermitte-Duclos) Desmoplastic infantile ganglioglioma Dysembryoplastic neuroepithelial tumor Central neurocytoma

Metastatic Tumors Metastases to the brain Metastases to the skull and intracranial dura Leptomeningeal metastasis (leptomeningeal carcinomatosis)

Pineal Parenchymal Tumors Pineoblastoma Pineocytoma Embryonal Tumors Medulloblastoma CNS primitive neuroectodermal tumor

Adapted and modified from Louis DN, et al: The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 114:97– 109, 2007.

radiation to small (up to 3 cm) intracranial tumors while sparing the adjacent normal brain and (2) to accurately register the previously acquired image data set onto intraoperative space, thus enabling precise intraoperative localization of surgical instrumentation relative to tumor margins, key vascular structures, and other critical anatomy. A continually evolving application is image-guided surgery, in which the entire surgical resection is guided by direct real-time MRI in the operating room environment.136,143,223,416

Contrast Enhancement IV administration of iodinated contrast medium causes a transient increase in the attenuation coefficient of normal gray matter and

accentuates the difference in radiographic density on CT images between white matter and gray matter in the normal brain. Similarly, IV administration of paramagnetic contrast agent (e.g., gadolinium chelates) transiently induces shortening of both the T1 and T2 relaxation times of normal gray matter and accentuates the difference in MR signal intensity between normal cerebral gray matter and white matter. Most WHO grade IV and, to a slightly lesser extent, WHO grade III primary intracranial neoplasms and certain lower-grade primary brain tumors exhibit “contrast enhancement” in all or part of the tumor. The presence, extent, and pattern of tumor contrast enhancement have proved to be of significant value in improving detection, localization, and characterization of intracranial neoplasms

http://radiologyebook.com

174

PART II  CT and MR Imaging of the Whole Body

by CT and MRI. Although the passage of intravascular contrast agent through brain parenchyma causes brief transient enhancement of the normal cerebral cortex, leakage of contrast material into the extravascular extracellular space accounts for the more pronounced and somewhat longer lasting contrast enhancement seen in many tumors and other structural abnormalities of the brain on both CT and MRI. The mechanisms involved in contrast enhancement of intracranial neoplasms on CT and MRI and in the uptake of radionuclide by such tumors appear to be identical.1,17,43,99,104 In the normal brain, tight intercellular junctions between capillary endothelial cells create a blood-brain barrier that prevents escape of radionuclide or contrast agent from the intravascular space. In gliomas, electron microscopy studies have demonstrated that the endothelial junctions are frequently patent and the level of junctional patency is proportional to the degree of tumor malignancy. It therefore seems reasonable to postulate that the intensity of contrast enhancement in gliomas and other intracranial tumors reflects the degree of bloodbrain barrier breakdown. Experimental data support this hypothesis by demonstrating that the peak intensity of tumor contrast enhancement in malignant gliomas occurs 20 to 60 minutes later than the peak serum concentration of contrast agent. Furthermore, glucocorticoids, which are known to diminish defects in the blood-brain barrier, also significantly reduce the extent of contrast enhancement in primary and secondary brain tumors.142,227,277,344,352 Factors involved in maximizing contrast enhancement of intracranial neoplasms on CT or MRI include the quantity of injected contrast agent and the timing of the images. The standard dose of paramagnetic contrast agent (typically a chelate of gadolinium) administered IV for MRI is 0.1 millimoles per kilogram of body weight and is usually assessed on T1-weighted images. Application of a strong radiofrequency pulse that is slightly offset from the resonance frequency of water protons before initiation of the standard T1-weighted spin echo pulsing sequence produces saturation of protons in macromolecules, which is then transferred to the adjacent free water protons, thus diminishing their signal intensity. Because this maneuver, known as magnetization transfer (MT), does not affect the T1 shortening caused by IV administration of gadolinium, the intensity and conspicuity of resultant contrast enhancement are increased. However, because of greater suppression of white matter signal, MT alters tissue contrast relationships such that certain structures (basal ganglia, pulvinar, substantia nigra) become more conspicuous on noncontrast MT images, whereas other structures that normally enhance (choroid plexus, cerebral veins and dural venous sinuses, body of caudate nucleus, pituitary gland) do so more conspicuously on MT images acquired after contrast administration. The reader must therefore exercise caution in interpreting areas of apparent contrast enhancement as possibly signifying disease.107

COMPLICATIONS OF BRAIN TUMORS Brain Edema Edema or swelling of the brain is a common accompaniment of many brain tumors and other structural abnormalities of the brain. When sufficiently severe, edema may be responsible for both focal and generalized signs of brain dysfunction.108 Edema is not a single pathologic response to a variety of insults but rather occurs in at least three different forms: vasogenic (secondary to tumor, inflammation, hemorrhage, extensive infarction, or contusion), cytotoxic (in response to hypoxia, early infarction, or water intoxication), and interstitial (resulting from acute obstruction to the flow or absorption of cerebrospinal fluid [CSF]).186 Vasogenic edema is the form of brain swelling most typically associated with intracranial neoplasms. It is caused by a breakdown in the

blood-brain barrier with seepage of a plasma filtrate containing plasma proteins through patent junctions between capillary endothelial cells. Gray matter is relatively resistant to the development of edema, and the extracellular plasma filtrate mainly accumulates in the white matter, extending along the major white matter fiber tracts. The subcortical arcuate (U) fibers between adjacent gyri offer greater resistance to the spread of edema than the long white matter tracts and are therefore involved relatively later and in more severe cases. The severity of edema associated with various brain tumors varies widely even between lesions of identical histology. In general, white matter swelling is greatest in association with carcinomatous metastases, and it is not unusual for a small metastasis to provoke a disproportionately large amount of edema. In order of declining severity, edema is associated with metastases, glioblastomas, meningiomas, and low-grade gliomas, but exceptions to this order are common. Rarely a low-grade glioma may be surrounded by extensive edema, whereas occasional metastases or meningiomas may have little associated white matter swelling.254,423 On CT, vasogenic edema appears as widening and diminished density of the major white matter tracts with fingerlike extensions into the arcuate fibers in each gyrus. The overlying cortical gray matter is compressed and thinned by the expanded pseudopods of edematous white matter (see Figs. 9-14 and 9-17).95,254 On noncontrast MRI, the high water content in the edematous peritumoral white matter causes prolonged T1 and T2 relaxation in the involved white matter; these findings manifest as an increase in signal intensity on T2-weighted images and as a less conspicuous decrease in signal intensity on T1-weighted images (see Figs. 9-1A, 9-15, and 9-20B). It is often difficult to delineate the boundary between tumor and edema on these noncontrast images; IV administration of paramagnetic contrast agent permits gross demarcation on T1-weighted (see Fig. 9-20C) and fluid-attenuated inversion recovery (FLAIR) images in many tumors (e.g., metastases) but not in malignant gliomas.41,96 Systemic administration of glucocorticoids minimizes the bloodbrain barrier defect inherent in most higher-grade brain tumors, with resultant reduction in fluid and protein extravasation. Diminution in peritumoral white matter swelling as well as in the volume and intensity of tumor contrast enhancement is often evident on follow-up imaging studies within a few days after institution of steroid therapy.76,142 Interstitial edema secondary to obstruction of CSF pathways appears on CT as poorly circumscribed periventricular hypodensity and on T2-weighted MRI as bandlike periventricular hyperintensity of varying thickness and margination. These findings are often symmetric and are most evident surrounding the anterosuperior margins of the dilated frontal horns of the lateral ventricles (see Fig. 9-2D) and the posterior margins of the occipital horns. Fluid accumulates in the periventricular white matter as a result of transependymal seepage of ventricular fluid across microscopic breaks in the ependymal lining of the ventricles. Systemic glucocorticoids do not affect this type of edema, but surgical insertion of a ventriculostomy shunt catheter above the level of obstruction usually results in prompt decompression of the ventricles and disappearance of the periventricular hypodensity/hyperintensity.95,426

Herniations An expanding mass within the rigid cranial vault, whether due to tumor, edema, or a combination of both processes, compresses and distorts the adjacent normal brain, producing internal herniation  of the brain under the relatively rigid falx or through the tentorial incisura. Laterally placed masses in the cerebral hemispheres, particularly those located in the superior temporal, midfrontal, or frontoparietal

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

175

B FIG 9-1  Right high frontal convexity meningioma causing subfalcine herniation. A, Axial T1-weighted noncontrast MRI at the level of the cingulate gyri. The right cingulate gyrus is displaced to the left of the midline under the free edge of the falx cerebri by a large, lobulated, rounded, heterogeneous, hypointense and isointense parasagittal mass (arrow). Note that the mass is surrounded by focal signal voids representing enlarged vessels on the surface of this extraaxial tumor. A wide band of slight hypointensity surrounding the mass represents vasogenic edema of the peritumoral white and gray matter. B, Coronal T1-weighted MRI at the level of the mid–third ventricle following gadolinium administration demonstrates marked downward and medial displacement of the right side of the corpus callosum and the frontal horn of the right lateral ventricle. The third ventricle (arrow) is bent and displaced across the midline by this large homogeneously contrast-enhancing dural-based tumor.

regions, displace the deep central structures (basal ganglia, thalamus, third ventricle, lateral ventricles, septum pellucidum) medially. The ipsilateral cingulate gyrus is compressed and displaced across the midline under the free edge of the falx (subfalcine herniation). Medially located high frontoparietal (parasagittal) masses depress and displace the cingulate gyrus contralaterally but without significantly affecting the deep central structures. Subfalcine herniation is most clearly depicted on coronal images, which also demonstrate depression and contralateral displacement of the corpus callosum and the underlying body of the ipsilateral ventricle (Fig. 9-1; see Fig. 9-11). The septum pellucidum and third ventricle are bowed away from the side of the mass. On axial CTs and MRIs, the ipsilateral ventricle appears compressed and displaced contralaterally.334 Tumors of the temporal lobe and middle cranial fossa displace the uncus and parahippocampal gyrus on the medial aspect of the temporal lobe toward the midline, with resultant encroachment on the lateral aspect of the suprasellar and ambient (circummesencephalic) cisterns and the tentorial incisura (descending transtentorial herniation) (see Figs. 9-7 and 9-18A). The ipsilateral margin of the midbrain is compressed, displaced contralaterally, and rotated by the impinging temporal lobe. Noncontrast MRI, postcontrast CT, or both may demonstrate medial displacement of the posterior communicating and posterior cerebral arteries, narrowing of the contralateral crural and circummesencephalic cisterns, and widening of their ipsilateral counterparts behind the impinging temporal lobe. Late secondary signs include hemorrhages in the compressed midbrain and unilateral or bilateral medial occipital infarctions (due to occlusion of one or both posterior cerebral arteries). Large centrally located cerebral masses for which the

primary vector of force is directly downward may cause bilateral transtentorial herniation.373 Ascending transtentorial herniation with upward displacement of the superior cerebellar vermis and hemisphere through the posterior aspect of the tentorial incisura is caused by expanding masses in the superior portion of the posterior compartment of the posterior cranial fossa, including the upper half of the cerebellum.341 As the superior cerebellar vermis is forced anteriorly and superiorly into the incisura, it protrudes into and compresses the superior cerebellar cistern from below and flattens the normal posterior convexity of the quadrigeminal cistern from behind. Increasing severity of herniation leads to reversal of that convexity and eventually to obliteration of the quadrigeminal and superior cerebellar cisterns (Fig. 9-2A and B) with flattening of the posterior margin of the third ventricle.280 Tumors of the lower half of the cerebellum and other masses in the inferior portion of the posterior compartment of the posterior fossa may produce downward displacement of the cerebellar tonsils through the foramen magnum (tonsillar herniation). This situation can be visualized directly on T1-weighted MRI in the sagittal and coronal planes (see Fig. 9-2A and C) and can be suggested on axial CT or MRI if the cisterna magna and upper cervical posterior subarachnoid spaces are encroached on and partially or completely obliterated by soft tissue density.

Hemorrhage Hemorrhage is not a common accompaniment of most intracranial neoplasms at initial presentation. In a series of 973 intracranial tumors, CT demonstrated acute intratumoral bleeding in 28 patients (3%) at

http://radiologyebook.com

176

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 9-2  Juvenile pilocytic astrocytoma of the cerebellum causing interstitial edema and upward transtentorial herniation and tonsillar herniation. A, Left parasagittal T1-weighted MRI without IV contrast. A large sharply marginated cyst in the left cerebellar hemisphere causes downward herniation of the cerebellar tonsils into the upper cervical spinal canal (arrow). The cystic mass compresses the fourth ventricle and pons from behind and displaces the superior vermis upward through the tentorial notch, obliterating the superior cerebellar and quadrigeminal cisterns and elevating and displacing forward the floor and posterior wall of the third ventricle. B, Axial T1-weighted noncontrast image at the level of the quadrigeminal cistern demonstrates forward displacement and compression of the quadrigeminal plate and cistern on the left (arrow) by the upwardly herniating superior cerebellar vermis and hemisphere. Note enlargement of the aqueduct, third ventricle, and temporal horns of the lateral ventricles secondary to obstruction of the fourth ventricle. C, Coronal T1-weighted image at the level of the posterior aspect of the foramen magnum following IV gadolinium. Both cerebellar tonsils have herniated downward through the foramen magnum (arrow). Note homogeneous contrast enhancement of a solid tumor nodule in the floor of the left cerebellar hemisphere cystic mass. D, Axial T2-weighted image at the level of the superior aspect of the lateral ventricles. The ventricles are markedly distended, and CSF has dissected across the ependymal margins of the ventricles into the adjacent periventricular white matter anteriorly (arrow) and in the corpus callosum.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms the time of initial diagnosis and in 7 additional patients (0.7%) who experienced clinical deterioration during their subsequent course.425 In Cushing and Eisenhardt’s87 meticulously reported experience, hemorrhage was found in association with intracranial gliomas in 31 of 832 cases (3.7%) (see Fig. 9-14). Acute hemorrhage was demonstrated on CT in 6 of a series of 131 meningiomas (4.6%).342 Intratumoral bleeding is more common in certain tumors, notably metastases from choriocarcinoma, melanoma (see Fig. 9-63), carcinomas of the lung and thyroid, and renal cell carcinoma, as well as in neuroblastoma, lymphoma, and medulloblastoma. In metastatic melanoma and choriocarcinoma, the typically hyperdense appearance of the metastatic nodules on CT has been attributed to the presence of acute hemorrhage or hemosiderin within the tumor.16,119,236,420 Because many of the breakdown products of blood have paramagnetic properties, MRI provides a unique and highly sensitive method for detection of subacute or chronic intracranial bleeding. Hemorrhage associated with intracranial tumors may occur centrally within a necrotic tumor cavity (in glioblastoma [see Fig. 9-14] and some metastases) or peripherally around the tumor (seen in other metastases and meningiomas). Signal intensity (MRI) or density (CT) is typically more heterogeneous in intratumoral hemorrhage than in benign intracranial hemorrhage. On occasion, hemorrhage may completely mask the presence of the underlying neoplasm, but images obtained after injection of contrast medium may demonstrate contrast enhancement of tumor along a margin of the hematoma or in other locations within the brain.16,425 Hemorrhage into a pituitary adenoma is a common cause of “pituitary apoplexy” and may obliterate not only the entire tumor but also the normal pituitary (see Figs. 9-56 and 9-57).380

TYPES OF BRAIN TUMORS Gliomas Gliomas include all primary brain tumors of astrocytic, oligoden­ droglial, or ependymal origin—astrocytomas, oligodendrogliomas, and ependymomas—as well as choroid plexus papillomas and carcinomas.

Astrocytomas.  Astrocytomas (tumors derived from astrocytes) are the most common of all primary intracranial neoplasms. In the 1993 revision of the WHO classification of brain neoplasms, astrocytomas were subdivided into four histologic grades. This histologic grading system has demonstrated a high positive correlation with the biological potential and behavior of tumors. Tumors of lower histologic grades (I and II) demonstrate few mitoses, little cellular structural variation (pleomorphism), and no vascular proliferation or necrosis. Tumors in grades III (anaplastic astrocytoma) and IV (glioblastoma multiforme) are characterized by more frequent mitoses, higher degrees of cellular dedifferentiation, and increasing angioneogenesis at the tumor periphery and by necrosis within the more central portion of the tumor.188 Astrocytomas are characterized on both gross dissection and diagnostic imaging into two groups: circumscribed and diffuse. In general, the diffuse astrocytomas are more common, tend to occur more in adulthood, and are more infiltrative or aggressive with spread along the white matter tracts. Circumscribed astrocytomas.  Three histologic types of astrocytoma are assigned to the circumscribed group. They are pilocytic astrocytoma (PA), pleomorphic xanthoastrocytoma (PXA), and subependymal giant cell astrocytoma (SCGA). All three types occur mainly in younger individuals, are more localized, and have a less aggressive biological potential than the diffuse astrocytomas. Pilocytic astrocytoma.  The most common type of circumscribed astrocytoma, PA, occurs in the cerebellum (60% of cases) in

177

children (Fig. 9-3; see Fig. 9-2), with a peak age distribution between 5 and 15 years. PA also frequently (30% of cases) originates in the optic pathways and the adjacent hypothalami (Fig. 9-4), usually in slightly younger children (2-3 years), often in association with neurofibromatosis type 1 (NF1).10 PA is the classic well-circumscribed brain neoplasm, but it lacks  a true tumor capsule. It grows mainly by expansion rather than  by infiltration. It is widely considered to be a benign (biologically stable) neoplasm and is classified histologically as grade I. It may  be densely cellular or more loosely arranged with intervening microcysts, and both patterns may coexist in different portions of the same tumor. Histologic characteristics include the frequent presence of eosinophilic Rosenthal fibers in tumor cell cytoplasm, although these can also be seen in reactive gliosis. Mitoses are rare in this lesion,  and necrosis is not seen. Most often the tumor represents a mural nodule in the wall of a well-circumscribed cyst. The origin and nature of the cyst are not well understood, but the cyst fluid is proteinaceous and likely secreted from coiled capillaries found in the midst of the nodular tumor. With the exception of the mural tumor nodule, the wall of the cyst consists of compacted normal brain or nonneoplastic gliotic tissue. Intratumoral calcification is occasionally identified (5%-25% of cases), but hemorrhage into or adjacent to the tumor  is rare. The clinical presentation of children with PA depends on the tumor site. The cerebellar tumors typically manifest with headache, nausea, and vomiting because of hydrocephalus secondary to obstruction of the fourth ventricle. Weakness and loss of equilibrium are not uncommon. The optic tumors typically cause difficulties with vision, but signs of hydrocephalus may appear if the tumor has extended into the hypothalamus and is obstructing the third ventricle.10,46 On CT the typical cerebellar PA appears as a smoothly marginated hypodense cystlike mass with a well-defined less hypodense tumor nodule on one wall (see Fig. 9-3A). The nodule may contain one or more areas of dense calcification. The cyst fluid is less hypodense than the CSF (15-25 Hounsfield units [HU]) because of its high protein concentration.267 Typically after IV administration of a contrast agent, dense homogeneous contrast enhancement of the mural nodule but not of the other walls of the cyst is seen (see Fig. 9-3B and C). More extensive enhancement of the cyst walls suggests a more aggressive (higher histologic grade) tumor.214,232,267 On MRI the mural nodule appears homogeneously hyperintense to gray matter on T2-weighted images and hypointense to isointense on T1-weighted images. Intratumoral calcification may cause a  more heterogeneous appearance. The associated cyst is even more hyperintense on T2-weighted images and even more hypointense  on T1-weighted images (see Fig. 9-2A). Edema of the adjacent  white matter is usually minimal. Homogeneous contrast enhancement of the tumor nodule is characteristic (see Figs. 9-2C and 9-3B and C), although a calcific focus (if present) does not demonstrate enhancement. The clinical prognosis of cerebellar PA is guardedly optimistic. The cerebellar tumor nodule can often be totally excised, and long-term survival rates exceeding 75% are commonly reported. The location and extent of the optic-chiasmatic-hypothalamic tumors typically preclude total excision, but irradiation and chemotherapy often achieve longterm tumor control.424 Optic pathway PA manifests on both CT and MRI with enlargement of the optic nerve as an intraconal mass, with characteristic kinking or buckling of the nerve secondary to the neoplasm itself and vascular congestion. Isointensity on T1-weighted MRI and heterogeneous hyperintensity on T2-weighted MRI are typical, with variable enhancement following IV contrast material administration.158,194

http://radiologyebook.com

178

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-3  Juvenile pilocytic astrocytoma of the cerebellum. A, Axial CT image without IV contrast. A large well-circumscribed hypodense mass is noted in the left cerebellar hemisphere. A flat nodular density is seen on the posteromedial wall of this cystic hypodensity (arrow). Axial (B) and parasagittal (C) T1-weighted MRIs following IV gadolinium demonstrate intense homogeneous contrast enhancement of the mural nodule (arrows) within the cystic hypointense mass.

Pleomorphic xanthoastrocytoma.  PXA was newly designated in the 1993 WHO revised classification.188 The number of cases documented in the literature is still small. PXA is presumed to arise from subpial astrocytes near the surface of the cerebral hemispheres.69 Rare tumors have also been described in the posterior cranial fossa and in the spinal cord. The gross morphology resembles the mural nodule and cystic appearance of cerebellar PA. The nodule is often based on the pial surface of the brain.

On CT and on MRI the appearance is similar to that of PA but in a different location: well circumscribed, hypodense to isodense on CT, hypointense on T1- and hyperintense on T2-weighted MRI, often with a cystic component that is isointense to CSF (Fig. 9-5A). However, the location is supratentorial (in order of descending frequency: temporal, frontal, parietal) and superficial, with little or no mass effect. Although the tumor presents a circumscribed appearance, it usually grows slowly and may infiltrate the adjacent brain and the overlying pia, even

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

179

B

C FIG 9-4  Juvenile pilocytic astrocytoma of the optic chiasm and hypothalamus with spread along visual pathways in a patient with NF1. Axial T1-weighted MRIs obtained with fat saturation following IV gadolinium. A, At the level of the mid-orbits, arrows indicate contrast enhancement in and marked widening of the intraorbital optic nerves bilaterally due to tumor spread. B, At the level of the optic chiasm, note widening of the chiasm and extension of the contrast-enhancing tumor into the left optic tract (arrow). C, At the level of the mid–third ventricle, intense homogeneous contrast enhancement indicates tumor extension into the basal ganglia bilaterally (arrows).

causing focal thickening of the meninges (dural tail). Intense homogeneous contrast enhancement of the solid portion of the tumor and the overlying meninges is the rule193,194,224 (see Fig. 9-5B and C). Histologically, as the name implies, the tumor nodule is pleomorphic, but cells with a foamy myxoid cytoplasm are common. PXA is usually classified as grade II, but later dedifferentiation into highergrade neoplasm may occur. The clinical presentation most commonly consists of seizures that may be intractable. Most reported cases have been in adolescents or young adults; the median age at diagnosis is 26 years.193 Treatment is usually surgical. Malignant transformation has

been described in 10% to 25% of cases, but the 10-year survival is reported as 70%.117,389 Subependymal giant cell astrocytoma.  This low-grade (WHO grade I) tumor occurs almost exclusively in patients with tuberous sclerosis. It is estimated that 10% to 15% of patients with tuberous sclerosis develop this neoplasm, usually in the late first or second decade of life. These tumors arise from subependymal nodules located on the lateral wall of the lateral ventricle overlying the head of the caudate nucleus. By convention, subependymal nodules that are more than 1 cm in diameter or become symptomatic are considered giant

http://radiologyebook.com

180

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-5  Pleomorphic xanthoastrocytoma of the left frontal convexity. A, Coronal turbo FLAIR MRI displays a semilunar hyperintense bandlike lesion indenting the left frontal convexity (arrow). More superficially the mass appears hypointense and mushrooms outward, eroding the calvarium and bulging into the overlying scalp. T1-weighted coronal (B) and parasagittal (C) images following IV gadolinium demonstrate contrast enhancement of the mural nodule (arrows). The findings suggest a superficial cystic tumor with mural nodule of long standing.

cell astrocytomas. The tumor does not invade into the underlying caudate head but rather grows exophytically into the ventricular lumen near the foramen of Monro, which may be obstructed either unilaterally or bilaterally. Despite this tendency for intraventricular growth, the overlying ependyma remains intact and dissemination via CSF is rare. Intratumoral calcification is common and may be heavy.267,355,357,358 Histologically the tumor is seen to contain large multinucleated giant cells of uncertain origin. On cytochemical analysis, some of these cells display GFAP (glial fibrillary acidic protein) reactivity, a feature consistent with an astrocytic origin, whereas others contain neuronspecific enolase, a finding consistent with neuronal origin.24 Subependymal giant cell astrocytomas present a characteristic appearance on CT.232,358 They appear as large, hypodense, polypoid,

sharply marginated intraventricular masses at the level of the foramen of Monro (Fig. 9-6A). The tumor may be heavily calcified and may obstruct the foramen unilaterally or bilaterally, causing gross enlargement of one or both lateral ventricles. Other manifestations of   tuberous sclerosis are usually evident, including cortical tubers and subependymal hamartomatous nodules. On MRI, subependymal giant cell astrocytoma appears as a heterogeneous sharply demarcated intraventricular mass that is mildly hyperintense on T2-weighted images and hypointense or isointense on T1-weighted images. It is located within the frontal horn of the left ventricle, without evidence for deep invasion or spread within the basal ganglia. On both CT and MRI, intense but heterogeneous contrast enhancement is often demonstrated (see Fig. 9-6B). The heterogeneous signal on MRI both without

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

181

B FIG 9-6  Subependymal giant cell astrocytoma in a patient with tuberous sclerosis. Axial CT images prior to (A) and following (B) IV administration of iodinated contrast medium. A partially calcified (arrow in A) multilobulated mass arising on the medial surface of the head of the right caudate nucleus fills and obscures the frontal horn of the right lateral ventricle. On the left, a calcified tuber (not neoplastic) obstructs the foramen of Monro; the dilated left frontal horn is compressed and displaced by a cyst, which is likely secondary to previous ventriculostomy complication. The noncalcified portion of the right frontal intraventricular tumor demonstrates homogeneous contrast enhancement (arrow in B).

and with contrast enhancement reflects the presence of dense calcification in these tumors. Differentiation of this tumor from the subependymal hamartomatous nodules that are so ubiquitous in patients with tuberous sclerosis is sometimes problematic, but the factors usually considered are the characteristic tumor location, its larger size, and the presence of contrast enhancement; contrast enhancement is not seen in the nontumorous nodules. Clinically these patients present with signs of increased intracranial pressure (headache, nausea, vomiting, and obtundation) due to obstruction of the foramen of Monro. Treatment is surgical, and the recurrence rate is low if the resection is complete.195 Diffuse astrocytomas.  Poorly marginated diffuse astrocytomas demonstrate much more aggressive infiltrative behavior than circumscribed astrocytomas. Progression from lower histologic grade (astrocytoma, WHO grade II) to higher grades (anaplastic astrocytoma, WHO grade III; glioblastoma, WHO grade IV) is common; it is estimated that 50% of low-grade astrocytomas undergo subsequent dedifferentiation into higher-grade tumors (anaplastic astrocytoma or glioblastoma).358 Diffuse astrocytomas are characterized genetically by inactivation of the TP53 gene, which is located on the short arm of chromosome 17. The TP53 gene encodes a protein that is involved in regulation of the cell cycle and DNA formation. Inactivation of this gene by mutation may lead to formation of abnormal DNA as well as other changes that can contribute to malignant transformation of the cells. It is interesting to note that primary glioblastomas that arise de novo and not from preexisting lower-grade astrocytoma do not exhibit loss of the TP53 gene.

Low-grade astrocytoma.  Low-grade (WHO grade II) supratentorial astrocytomas are hypercellular tumors that consist of welldifferentiated fibrillary or gemistocytic astrocytes and demonstrate few mitoses and moderate pleomorphism but no vascular proliferation or necrosis. On immunohistochemistry studies, they demonstrate strong affinity for GFAP.188 They arise in white matter and grow by infiltration predominantly along white matter tracts, extending between and separating anatomic landmarks. The involved area of the brain eventually appears larger than its contralateral counterpart. Cerebral astrocytomas are usually solid masses that, because of their infiltrative characteristics, are poorly circumscribed and blend imperceptibly with the adjacent cerebral parenchyma. Small and large cysts are occasionally found within these tumors, but hemorrhage is rare.216,408 Approximately two thirds of low-grade astrocytomas are supratentorial; these arise mainly in the frontal, parietal, and temporal lobes (Figs. 9-7 and 9-8). They typically provoke little or no edema in the surrounding tissues. Because such a tumor begins in the white matter without affecting the more eloquent cortical centers, symptoms are usually vague and nonlocalizing until the lesion gains sufficient size, extends into eloquent cortex, or undergoes dedifferentiation into a grade III or IV tumor. The peak age range for occurrence of low-grade astrocytoma is the third through fifth decades of life. Some diffuse (initially low-grade) astrocytomas occur in childhood, notably involving the optic nerves and chiasm, optic tracts,  and visual pathways as well as the hypothalamus and thalamus. More than half of the optic pathway gliomas of early childhood involve the optic chiasm and hypothalamus, and 40% of these are diffuse fibrillary astrocytomas.7 In contradistinction to the young child with NF1

http://radiologyebook.com

182

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-7  Low-grade astrocytoma of the temporal lobe. Axial T1-weighted MRIs prior to (A) and following (B) IV gadolinium. A mostly ill-defined homogeneously hypointense mass expands the left anterior temporal lobe (arrow). There is no contrast enhancement and no surrounding edema. Note medial displacement of the medial margin (uncus and parahippocampal gyrus) of the left temporal lobe indicating early transtentorial herniation with compression of the left ambient cistern and the anterolateral aspect of the left cerebral peduncle.

in whom more circumscribed PAs develop along the visual pathways,10 the more diffuse fibrillary astrocytomas arise without any identifiable genetic predilection, although in a similar age group (peak age, 2-3 years); these are termed sporadic optic pathway astrocytomas. However, the diffuse fibrillary tumors share with their adult counterparts a poorer prognosis and thus merit a more aggressive course of management.7 Another relatively common location of diffuse low-grade astrocytoma in childhood is in the brainstem. Brainstem gliomas constitute 10% to 20% of all pediatric brain tumors.232 The peak age of occurrence is 5 to 10 years. These tumors arise mainly in the pons but commonly involve the midbrain and medulla and may extend into the upper cervical spinal cord.424 Despite their proximity to the aqueduct and fourth ventricle, considerable tumor growth may take place without causing obstruction of the ventricular system. Multiple cranial nerve deficits, long tract signs, and ataxia are the most common presenting findings.232,238,424 Although these tumors are often regarded as low-grade lesions, histologic heterogeneity with intermixed areas of higher-grade neoplastic change is common. Approximately 60% of brainstem gliomas demonstrate foci of glioblastomatous change with central necrosis. Five-year survival rates range from 15% to 30%.6,232 On CT, low-grade diffuse cerebral astrocytomas appear most commonly as poorly marginated areas of decreased attenuation involving the white matter of one or more lobes of the brain with mild to moderate mass effect.43,216,242,326,358 A more heterogeneous appearance with patchy areas of normal or slightly higher tissue density intermixed with hypodensity is seen in a small number of cases. Contrast enhancement is usually absent, reflecting a lack of tumor vascularity, blood-brain barrier breakdown, or both.358

Because low-grade astrocytomas are predominantly hypodense (15-30 HU) and usually do not exhibit contrast enhancement, it is often not possible with CT to differentiate the margin of the tumor from the adjacent white matter. However, because the blood-brain barrier is intact, extensive peritumoral edema is unusual in low-grade astrocytomas.216 Grossly identifiable areas of intratumoral hemorrhage or necrosis are also not typically seen on CT or MRI studies in these lesions. Calcification is identified in 20% of low-grade diffuse astrocytomas, and cysts are occasionally demonstrated.326 The presence of hemorrhage, necrosis, edema, and/or contrast enhancement suggests progression to a higher tumor grade. Optic pathway diffuse astrocytomas appear on CT and MR as rounded or lobulated thickening of the more posterior optic pathway, typically involving the optic chiasm and retrochiasmal segment of the optic pathway. This typically appears homogeneously hypodense on CT, T1 isointense to brain, and variably hyperintense on T2-weighted images. The tumor may encase the arteries of the circle of Willis  and invade the floor of the third ventricle, causing obstructive hydrocephalus.7,358 Necrosis and hemorrhage are uncommon. In contrast to optic pathway PAs, cystic foci may be seen. Contrast enhancement is highly variable and typically does not signify malignant progression, a rare occurrence in sporadic diffuse optic pathway astrocytoma in childhood. Brainstem glioma typically appears on CT as ill-defined diffuse heterogeneously hypodense and isodense asymmetric expansion of the pons and medulla, with dorsal displacement of the floor of the fourth ventricle and aqueduct and ventral encroachment on the pontine cistern (Fig. 9-9A and B).6,28,424 Despite the impression of the tumor on the aqueduct and the floor of the fourth ventricle, hydrocephalus is a

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

183

B

C FIG 9-8  Low-grade astrocytoma of the frontal lobe. Axial (A) and coronal (B) turbo FLAIR images without contrast and coronal T1-weighted postgadolinium (C) image in a 30-year-old man complaining of headache following a questionable seizure. A large well-demarcated homogeneously T2 hyperintense mass is centered in the right frontal opercular subcortical white matter on the FLAIR images (A and B). Following IV gadolinium, the T1-weighted image (C) demonstrates no intrinsic contrast enhancement of this large homogeneously hypointense tumor. The margins of this mass appear less well defined on the T1-weighted images. Although the overlying sulci are effaced, there is no gross surrounding edema or midline shift.

relatively late finding. Approximately 50% of brainstem gliomas exhibit varying degrees of contrast enhancement on CT and MRI after IV administration of contrast medium (see Fig. 9-9C). Contrast enhancement at presentation suggests higher grade and is usually patchy and heterogeneous. However, new enhancement after treatment does not

necessarily suggest disease progression but may actually represent treatment response.28,226 On MRI, diffuse low-grade astrocytomas appear relatively homogeneous and demonstrate varying degrees of prolonged T1 and T2 relaxation. They therefore appear homogeneously hypointense on

http://radiologyebook.com

184

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-9  Brainstem glioma. Axial T2-weighted (A) and T1-weighted (B) noncontrast MRIs through the midpons demonstrate a poorly circumscribed mass with heterogeneous signal intensity involving the pons, which is larger on the left than on the right (arrows). This expansile and invasive tumor causes thinning of the prepontine cistern and flattening and posterior displacement of the floor of the fourth ventricle. C, On this midsagittal T1-weighted image following IV gadolinium, note extension of the tumor (arrow) superiorly into the midbrain, with posterior bowing of the aqueduct and floor of the fourth ventricle and patchy intratumoral contrast enhancement.

T1-weighted images and hyperintense on T2-weighted images in comparison with normal brain (see Figs. 9-7A and 9-8A and B).15,96,326 Because of the inherently higher contrast resolution of MRI, astrocytomas are often better demarcated from adjacent normal white matter on MRI than on CT, reflecting their higher water content. Nevertheless, as on CT, the apparent margins of the lesion on MRI (without or  with contrast enhancement) do not closely correlate with histopathologic evidence of tumor extent.96,200 As on CT, there is no grossly evident contrast enhancement on MRI in diffuse low-grade cerebral

astrocytomas (see Figs. 9-7B and 9-8C); the demonstration of contrast enhancement should raise the strong suspicion of a higher tumor grade.43,98,128,326,358 The differential imaging diagnosis of low-grade diffuse astrocytoma on CT and MRI is ubiquitous. Principal considerations are cerebral infarction, cerebritis, and higher-grade glioma (e.g., anaplastic astrocytoma). Infarctions tend to involve overlying cortical gray matter more extensively and are often more sharply delineated from the adjacent cerebral parenchyma.254 Inflammatory processes and higher-grade

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms gliomas are often associated with blood-brain barrier breakdown and exhibit contrast enhancement and surrounding edema.326 The prognosis in low-grade astrocytoma is guarded. Many of these tumors in young adults progress to a higher grade. Median survival has been reported as 5 to 8 years.247,326 Anaplastic astrocytoma.  As the name implies, anaplastic astrocytomas (WHO grade III) exhibit considerable variation in cellular morphology, with high cellularity, frequent mitoses, and foci of vascular proliferation, findings that are not present in low-grade (grade II) astrocytomas.358 However, no necrosis is present in anaplastic astrocytomas on either gross inspection or microscopic evaluation.326 These tumors tend to occur in a slightly older age group (peak age, mid-40s) than the grade II lesions and often represent a progressive dedifferentiation of a previously low-grade tumor. Reflecting the increased histologic variation, the imaging findings are also less homogeneous than in the lower-grade tumors (Figs. 9-10A and 9-11A). On T2-weighted MRI, anaplastic astrocytomas are most often heterogeneously hyperintense, frequently involve the overlying cortex, and are associated with a greater degree of vasogenic edema and overall mass effect (see Fig. 9-11A). Restricted diffusion is often identified within these tumors, likely reflecting foci of increased cellularity. Contrast enhancement is common on both CT and T1-weighted MRI (see Figs. 9-10B and 9-11B and C); compared with glioblastoma (grade IV) the enhancement pattern is often not ringlike but rather patchier. Although the region of contrast enhancement may appear

A

185

discrete, infiltration of tumor cells beyond the margin of enhancement into the edematous surrounding white matter is nearly always present, often referred to as infiltrative edema.96 Current estimates indicate that anaplastic astrocytomas represent about 25% of all primary gliomas of the brain.275 Median survival for patients with anaplastic astrocytoma is 2 to 3 years. Most tumors progress relatively rapidly to frank glioblastoma with intratumoral necrosis. Glioblastoma.  Glioblastoma (WHO grade IV) is the most common primary malignant tumor of the CNS, accounting for more than half of all intracranial gliomas. It is biologically the most aggressive of the gliomas, with rapid progression of clinical signs and symptoms and a mean survival time in the range of 12 months. Like all other astrocytomas, glioblastoma characteristically involves the cerebral white matter, but this highly malignant process readily infiltrates and destroys the adjacent gray matter with loss of gray-white differentiation. As their name implies, these highly malignant tumors have a variegated histologic appearance with interspersed areas of hypercellularity, cellular pleomorphism, endothelial proliferation, and intratumoral necrosis.42 It is probable that most glioblastomas arise by anaplastic change within preexisting low-grade cerebral astrocytomas, although some likely occur de novo.341 Although most commonly focal and singular in their presentation, glioblastomas may be multifocal or multicentric in origin. The peak age of incidence of glioblastoma is during the fifth

B FIG 9-10  Anaplastic astrocytoma, left frontotemporal. Axial CT images prior to (A) and following (B) IV administration of iodinated contrast medium. On the noncontrast image (A), patchy areas of hyperdensity (arrows) are noted on both sides of the left sylvian fissure, involving the medial (insular) and lateral (temporal) aspects. There is mild mass effect on the lateral margin of the atrium of the left lateral ventricle by the posterior extension of this tumor (posterior arrow). Note the ill-defined diffuse hypodensity extending between and surrounding the hyperdense regions; this likely represents a mixture of infiltrating tumor cells and reactive edema. B, On the postcontrast image, there is diffuse contrast enhancement of the previously hyperdense regions (arrows). Note that the enhancement reaches the ependymal surface of the left atrium. There is no evidence of intratumoral necrosis.

http://radiologyebook.com

186

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-11  Anaplastic astrocytoma in the deep frontal lobe. Axial noncontrast turbo FLAIR (A) and T1-weighted axial (B) and coronal (C) postgadolinium images in a 31-year-old man who presented following a seizure. A, A large poorly demarcated heterogeneously T2 hyperintense mass is identified in the right deep frontal region involving the putamen, the head of the caudate nucleus, and the overlying insula and subinsular white matter. This mass displaces the sylvian fissure laterally, and the third ventricle and interhemispheric fissure are displaced to the left of the midline. B and C, Postgadolinium images demonstrate irregular contrast enhancement in the region of the anterior putamen.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

187

B FIG 9-12  Glioblastoma multiforme involving the septum pellucidum. A, Axial T2-weighted MRI through the inferior aspect of the frontal horns of the lateral ventricles. An ovoid heterogeneously hyperintense mass (arrow) arising from the inferior aspect of the septum pellucidum indents and partially occludes the frontal horns bilaterally. Note the irregularly marginated intratumoral marked hyperintensity suggesting central necrosis. B, Following IV gadolinium, a coronal T1-weighted image demonstrates intense contrast enhancement (arrow) of the thick peripheral rind with nonenhancement of the central cavity.

and sixth decades of life, some 10 to 20 years later than the peak age for astrocytoma. However, glioblastoma is not rare in younger adults or even in children; of all glioblastomas, approximately 3% occur in childhood, mainly in the brainstem and cerebellum. On gross inspection, glioblastomas appear more or less spherical when confined to one lobe of a cerebral hemisphere. However, these rapidly growing neoplasms invariably infiltrate into adjacent lobes, the corpus callosum (“butterfly gliomas”), and the contralateral hemisphere. Tumors that have extended widely tend to have more irregular configurations with lobulated margins of varying thickness. Infiltration of the ependyma of the lateral and third ventricles and penetration of the cerebral cortex with invasion of the overlying meninges occur commonly and may lead to seeding through the CSF pathways.288,341 Vascular endothelial proliferation with formation of large sinusoidal channels (angioneogenesis) is a typical finding in glioblastomas, and intratumoral hemorrhage is common.424 The central portions of these tumors frequently outgrow their blood supply and undergo ischemic coagulative necrosis, leaving only a peripheral rim of viable tumor of variable thickness.341 Cysts of varying size representing the residua of central necrosis may be found within the more solid periphery of many glioblastomas. The gross pathologic findings described are accurately portrayed on CT and MRI. Although smaller tumors are generally rounded, more extensive lesions have irregular configurations indicating their spread along white matter pathways and penetration into and through cerebral cortex. These tumors typically appear on noncontrast CT as heterogeneous masses with irregular borders of normal or slightly increased density with central cavitation of diminished attenuation

(see Fig. 9-14).216 The irregular hyperdense borders demonstrate avid contrast enhancement. Findings on MRI parallel those on CT. Because of the greater inherent sensitivity of MRI for detecting subacute and chronic hemorrhage and for demonstrating tumor hypervascularity (curvilinear and racemose signal voids), the overall appearance of the lesion is even more heterogeneous on MRI than on CT. Areas of intratumoral central necrosis, which tend to dominate the appearance of the most aggressive of these tumors, appear notably hyperintense on T2-weighted images except where interrupted by more solid masses of cellular debris (Fig. 9-12A). Compared with this central necrotic T2 hyperintensity, the surrounding rim of viable and growing tumor appears significantly less hyperintense (owing to its hypercellularity) with unsharp margins and may blend with or be indistinguishable from the peritumoral infiltrative edema. Virtually all glioblastomas are associated with edema in the surrounding peritumoral white matter, usually graded as moderate or severe (Figs. 9-13A; see Fig. 9-15A).216 The origin of this edema is uncertain; it may reflect production of a vascular permeability factor by the tumor cells.15 On CT and on T1-weighted MRI, grossly evident mildly hyperdense or slightly hyperintense tumor margins are typically unsharp, but after IV administration of a contrast agent, they stand out in contrast to the surrounding hypodense or hypointense edematous white matter (see Figs. 9-13B and 9-15B).43,96 Hemorrhage within the tumor is common in these aggressive tumors with extensive angioneogenesis. The appearance of acute intratumoral hemorrhage in glioblastoma on both CT and MRI differs from acute nonneoplastic intracerebral hemorrhage in that the

http://radiologyebook.com

188

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-13  Left deep frontoparietal glioblastoma with extension through corpus callosum. A, Axial T2-weighted MRI through the level of the upper third ventricle. A large area of diffuse hyperintensity involves the left posterior thalamus and basal ganglia and extends posteriorly to involve the subcortical white matter in the left parietooccipital convexity. It is not possible on this image to either define the tumor margins or separate gross tumor from surrounding edema. The third ventricle and the flow voids of the two parallel internal cerebral veins are bowed to the right of the midline secondary to the left-sided deep mass effect. Note that this hyperintensity also extends into and widens the splenium of the corpus callosum (arrow), crossing the midline into the contralateral hemisphere and creating a “butterfly glioma” pattern. B, Axial T1-weighted image following IV gadolinium demonstrates patchy heterogeneous contrast enhancement within the splenium (arrow) and surrounding the dilated and posteriorly displaced atrium of the left lateral ventricle; the tumor has blocked the posterior portion of the ventricle, and the marginal contrast enhancement of the atrium suggests subependymal tumor spread.

former is more sharply circumscribed and more heterogeneous in density/signal because of the surrounding and intermixed tumor and necrotic debris (Fig. 9-14), whereas the latter is more irregularly marginated and homogeneous in density/signal because it dissects along otherwise normal white matter tracts. Also, evolution of intratumoral hemorrhage usually proceeds more slowly compared with hyper­ tensive, posttraumatic, or other nonneoplastic types of parenchymal hematomas.16 Contrast enhancement of viable tumor, including the peripheral rim and the intratumoral solid portions, is nearly universal in glioblastomas (see Figs. 9-12B, 9-13B, and 9-15B).216,242 The opacifying tumor ring is typically irregular and of varying thickness. The intensity of the enhancement correlates with the level of cellular anaplasia,43 but it does not accurately reflect the extent of microscopic tumor infiltration.41,96 Careful imaging-histopathologic correlation studies in patients with malignant gliomas have conclusively established the presence of viable tumor cells well beyond the margin of contrast enhancement, and the zone of edema in the surrounding white matter must be understood as including foci of microscopic tumor infiltration42,96,182; thus the term infiltrative edema.51 Differentiation by CT or MRI between glioblastoma (or other grade IV glioma), carcinomatous metastasis, and lymphoma is often difficult. All three processes may appear as irregularly rounded, contrast-

enhancing, ringlike masses with walls of varying thickness. In all, the viable tumor rim is typically irregular and nonuniform in thickness, and histologic examination is usually required for definitive diagnosis. Primary CNS lymphoma is typically more central in location and homogeneously solid (see Figs. 9-46 and 9-74), but central necrosis, infiltration into the corpus callosum and other major white matter pathways, and tumor extension into the meninges and along the ependyma (all common late manifestations in glioblastoma) are not rare occurrences with lymphoma. Additionally, a large area of active demyelination with inflam­ mation and edema secondary to previously undiagnosed multiple  sclerosis may also simulate a glioblastoma on CT or MRI. Such tumefactive demyelinating lesions (e.g., in multiple sclerosis or in acute disseminated encephalomyelitis [ADEM]) often demonstrate arclike (open toward the adjacent cortex) contrast enhancement and increased or facilitated diffusion,4 findings that are much less common in glioblastomas. When images demonstrate multiple nodular or ringlike masses, multicentric glioblastoma must be included in the differential diagnosis (Fig. 9-15), along with metastases and pyogenic cerebral abscesses. The peripheral rim of an abscess is typically (but not always) thinner and more uniform than that of a glioblastoma or a metastasis. Diffusion-weighted imaging (DWI) demonstrates restriction of

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

FIG 9-14  Intratumoral hemorrhage into large right posterior frontoparietal glioblastoma. Axial CT image without IV contrast. A large welldemarcated hypodense cystic cavity in the deep right frontoparietal white matter is surrounded by poorly defined hypodensity. On the posterior margin of this cavity there is a sharply defined scythelike hyperdense collection (arrow) representing intratumoral hemorrhage. Note the considerable mass effect secondary to diffuse edema in the right frontal lobe, with shift of the lateral and third ventricles and the septum pellucidum to the left of midline. The atrium of the left lateral ventricle is compressed and displaced anteromedially by this cavitated tumor.

diffusion in the central cavity of an abscess but not in the coagulative necrosis of a primary or metastatic neoplasm.86 Besides DWI, other advanced neuroimaging techniques can be used to differentiate glioblastoma, solitary metastases, and abscesses, namely magnetic resonance spectroscopy (MRS), perfusion MRI, and positron emission tomography (PET) imaging, as discussed later in this chapter (see “Advanced Neuroimaging Techniques”). A commonly encountered clinical problem is the need to differentiate radiation necrosis from recurrent glioblastoma in a patient who has undergone surgical resection and a complete course of radiation therapy and several months after the radiation therapy demonstrates a change in neurologic status. If the follow-up imaging study demonstrates interval increase in the size of the apparent tumor mass and the extent of contrast enhancement, differentiation of recurrent tumor from radiation necrosis in this lesion on the basis of CT or MRI is not possible. In such circumstances, advanced neuroimaging techniques are often employed with perfusion MRI, MRS, and PET imaging, using both 18F FDG and non-FDG radiotracers, greatly improving sensitivity and specificity, as discussed later in this chapter (see “Advanced Neuroimaging Techniques”). Gliomatosis cerebri.  A comparatively uncommon pattern of glial neoplasia, gliomatosis cerebri is characterized by widespread diffuse neoplastic infiltration in multiple lobes and both cerebral hemispheres without formation of an obvious tumor mass. The cell of origin is uncertain; the tumor is classified as WHO grade III. The extent of infiltration is out of proportion to the other histopathologic

189

features, such as the level of anaplasia and the relative lack of destruction of normal tissues. Despite the extent of tumor infiltration, neural connections are preserved, and the clinical signs and symptoms are relatively mild. Nevertheless the clinical course is relentlessly progressive over periods varying from months to a few years.399 The process tends to extend along the white matter tracts but also involves the adjacent gray matter with loss of clear gray-white distinction both on gross pathology and on imaging studies. The involved parenchyma is both expanded and distorted by the tumor infiltration. Microscopically the tumor cells are elongated and infiltrate between myelinated fibers without destroying them, much like in a low-grade astrocytoma, but there is greater cellular atypia and mitotic frequency, and foci of necrosis are occasionally seen. Diagnostic imaging studies demonstrate extensive and scattered areas of parenchymal involvement with ill-defined areas of hyperintensity on T2-weighted MRI and subtle areas of hypodensity on CT scans (Fig. 9-16A). No well-demarcated or circumscribed tumor masses are identified on imaging in this disorder.297 Focal expansion and distortion are seen as areas of diffuse effacement of sulci and compression and displacement of the adjacent ventricles and cisterns. The incidence and level of contrast enhancement are often less than the extent of apparent tumor infiltration (see Fig. 9-16B) because enhancement relates closely to the degree of cellular dedifferentiation.15 Clinically the principal differential diagnosis is with multiple sclerosis, but the imaging pattern of multiple widespread areas of expansion and distortion as well as the relatively rapid and consistent progressive clinical course favor a diagnosis of neoplasm. Gliosarcoma.  A rare tumor, gliosarcoma consists of both glial and mesenchymal elements. Most reported patients have been between 50 and 70 years old. The origin of the mesenchymal element remains uncertain and controversial; it presumably arises from spontaneous neoplastic transformation of vascular elements within the glial neoplasm, but prior radiation therapy has been implicated in a few case reports. The mesenchymal element is usually a fibrosarcoma, but other sarcomas have been reported. Immunohistochemistry studies demonstrate both GFAP reactivity, confirming the astrocytic nature of the glial component, and positive staining for collagen and reticulin, confirming the mesenchymal component.69 The glial component of these mixed tumors is typically glioblastoma, and gross pathology as well as both CT and MRI demonstrate the variegated pattern of that tumor with areas of central necrosis and peripheral infiltration both along white matter tracts and into adjacent gray matter (Fig. 9-17). Invasion into and involvement of the overlying meninges are often noted (see Fig. 9-17B); invasion into bone and extracranial soft tissue likely reflects the sarcomatous element within the tumor, and systemic metastasis (a rare occurrence in glioblastoma) is well described for gliosarcoma.69

Oligodendroglioma.  Oligodendrogliomas (tumors derived from oligodendrocytes) represent 15% to 20% of intracranial gliomas  and 5% to 10% of all intracranial neoplasms.70 The diagnosis is made much more commonly in recent years than in the past, reflecting a reevaluation of the histologic and immunochemical criteria (oligodendrogliomas are GFAP negative) that define the distinction between oligodendroglioma and astrocytoma.69,70 These tumors occur principally in the supratentorial white matter. Like low-grade astrocytomas, they tend to infiltrate along white matter tracts in a nondestructive manner. However, more commonly than astrocytomas, oligodendrogliomas also involve the subcortical white matter and the overlying cortex, and involvement of deep structures is also common.341 Eightyfive percent occur in the cerebral hemispheres, more than half in the frontal lobes.79,216,354

http://radiologyebook.com

190

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-15  Multicentric glioblastoma. Coronal MRIs through the posterior atrial region. A, T2-weighted image demonstrates a hyperintense rounded mass in and deep to the high right frontoparietal convexity (white arrow) and a second similar mass in the inferomedial right temporooccipital cortex and white matter (black arrow). Between these masses is a large area of greater hyperintensity, likely representing edema with probable microscopic tumor infiltration, which effaces the overlying convexity cortex and severely compresses and obscures the atrium of the right lateral ventricle. B, On a T1-weighted image following IV gadolinium administration, both lesions demonstrate inhomogeneous contrast enhancement (arrows). The inferomedial tumor has invaded the splenium of the corpus callosum, extending across the midline into the left hemisphere. Note also extension of this tumor inferiorly, protruding into the tentorial notch and impinging on the right side of the superior cerebellar vermis (black arrow).

A

B FIG 9-16  Gliomatosis cerebri in the left cerebral hemisphere. A, On this axial T2-weighted MRI, there is evidence of diffuse enlargement of the left frontal, temporal, and parietal lobes, with extensive but ill-defined hyperintensity involving the temporoparietal convexity cortex and adjacent white matter (arrows). B, Axial T1-weighted MRI following IV gadolinium demonstrates widely separated patchy areas of faint contrast enhancement (arrows) that represent scattered foci of malignant gliomatous infiltration.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

191

B FIG 9-17  Gliosarcoma. A, Axial CT image following IV contrast demonstrates a lobulated mass with variegated contrast enhancement in the high right frontal convexity, with multiple intratumoral calcifications (arrow) surrounding an area of central hypodensity that likely represents necrosis. The contrast enhancement crosses the midline into the high left frontal convexity. Extensive white matter hypodensity, likely representing edema, involves the surrounding frontal lobe white matter bilaterally. B, Coronal T1-weighted MRI after IV gadolinium confirms the presence of a lobulated high right frontal convexity contrast-enhancing tumor (arrow) and also clearly demonstrates the extensive spread of this tumor along the inner table of the skull bilaterally, with bandlike meningeal thickening over the convexities of both cerebral hemispheres.

The peak age of incidence of oligodendrogliomas is in the fourth and fifth decades of life; these tumors are much less common in childhood or adolescence.69,325 The initial clinical presentation is most often a seizure, reflecting the predilection for cortical involvement. Tumor growth is generally slower than in astrocytomas of similar grade (WHO grade II) and the prognosis is better, but a more biologically aggressive category is now recognized in the revised WHO classification as anaplastic oligodendroglioma (WHO grade III).69,70 Cytogenetic studies have noted loss of the 1p and 19q chromosomes in approximately half of patients with oligodendroglioma; on imaging studies, patients with loss of 1p and 19q have demonstrated a significant association with positive response to multiagent nitrosoureabased chemotherapy.45,298 Oligodendrogliomas are typically densely cellular and solid, contain multiple irregular masses of dystrophic calcification, and grow very slowly. Median survival is approximately 10 years. They can attain considerable size before they produce symptoms.69 They may contain a dense network of branching capillaries. Areas of cystic degeneration are found in 20% of cases, mainly in the larger tumors. Intratumoral hemorrhage is also a frequent finding.212 Treatment is primarily surgical resection with adjuvant chemotherapy and radiotherapy, but local recurrence is common. Anaplastic oligodendrogliomas on histology demonstrate greater cellular atypia and mitotic activity, more extensive tumor vascularity, and necrosis. The prognosis is less favorable, with median survival of 4 years and 10-year survival of approximately 15%. Foci of astrocytoma are frequently found in association with oligodendroglioma, and when these are prominent, the tumor is termed  an oligoastrocytoma. Oligoastrocytomas are typically WHO grade II

tumors and anaplastic oligoastrocytomas WHO grade III.269 When there is a clear preponderance of neoplastic oligodendrocytes (75%90%), the tumor is classified as an oligodendroglioma.358 On CT, oligodendrogliomas most often appear as masses of heterogeneous density (predominantly hypodense) in the frontal lobe, with irregular and poorly defined margins involving the more superficial white matter and frequently infiltrating into the overlying cortex and obscuring the gray-white interface. The characteristic feature in the great majority of cases is the presence of large nodular or clumplike calcifications within the tumor (Fig. 9-18A). The frequency of intratumoral calcification in oligodendrogliomas (70%-90%) is the highest of any cerebral neoplasm, but overall, intratumoral calcification is most common in astrocytomas (frequency, 25%) because of the much greater prevalence of that lesion. Intratumoral cyst formation is common in oligodendrogliomas (20%),212 but frank intratumoral necrosis is unusual and indicates a poor prognosis. Contrast enhancement is noted in approximately 50% of oligodendrogliomas (see Fig. 9-18C), but the degree of enhancement is variable and appears to correlate with higher histologic grade and lower survival.358 On MRI, oligodendrogliomas appear heterogeneous on both T1and T2-weighted images because of the presence of intratumoral cysts and calcifications. On T1-weighted images, the tumors appear predominantly hypointense to gray matter, and on T2-weighted images they are most often hyperintense, with small intratumoral cysts and focal calcifications and hemorrhages contributing to the overall heterogeneity (see Fig. 9-18B). Peritumoral edema is noted in about one third of cases.212 Enhancement is variable and typically heterogeneous, with anaplastic oligodendrogliomas (grade III) enhancing more commonly than low-grade oligodendroglioma (grade II). Also, as with

http://radiologyebook.com

192

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-18  Oligodendroglioma. A, Axial noncontrast CT image demonstrates a large mass with poorly defined margins in the left midtemporal lobe (arrow). The mass is predominantly hypodense with a large amorphous central calcification and involves the convexity cortex as well as the underlying white matter. The expanded left temporal lobe herniates across the tentorial margin and compresses the lateral aspect of the left cerebral peduncle. B, On this axial T2-weighted image, the large intratemporal tumor appears heterogeneously hyperintense, with the calcified portion (arrow) exhibiting lesser intensity. C, Axial T1-weighted MRI following IV gadolinium. The tumor appears predominantly hypointense, has ill-defined margins, and involves mainly white matter but also extends into the convexity cortex and demonstrates faint heterogeneous contrast enhancement (arrow).

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms other gliomas, new enhancement within a previously nonenhancing oligodendroglioma suggests anaplastic transformation.100

Ependymoma.  The ependyma lining the floor of the fourth ventricle is the most common site of origin of intracranial ependymomas (65%). Nevertheless, approximately 30% of ependymomas occur above the tentorium, arising mainly from ependymal cell rests in the cerebral white matter adjacent to the lateral ventricles.267,341,358,379 Infratentorial ependymomas occur predominantly in children between 1 and 6 years of age; supratentorial ependymomas occur mainly in the second through the fourth decades of life.11,114,379 Posterior fossa ependymomas make up approximately 10% to 15% of all primary CNS neoplasms of childhood and approximately 3% to 5% of all intracranial neoplasms.341 They are the third or fourth most common posterior fossa tumor of childhood, exceeded only by PA, medulloblastoma, and possibly brainstem glioma.271 Ependymomas are usually slow-growing tumors of moderate malignancy (WHO grade II), but the prognosis is guarded because of their notable tendency to recur locally and disseminate via the subarachnoid spaces. The incidence of seed metastases via the CSF is approximately 10% overall but is considerably higher in the infratentorial tumors of early childhood; tumor seeding into the subarachnoid spaces is found initially or on follow-up studies in approximately one third of fourth ventricle ependymomas.267,341 Supratentorial ependymomas are commonly paraventricular, located most frequently within cerebral parenchyma near the atrium of the lateral ventricle or the foramen of Monro and presumably arising from ependymal cell rests.267 Infratentorial ependymomas are lobulated, well-circumscribed, soft exophytic masses that expand within the cavity of the fourth ventricle rather than infiltrate into the surrounding parenchyma. They often extend in a tonguelike fashion through the lateral recesses into the lateral medullary and cerebellopontine angle cisterns (15%) or through the foramen of Magendie into the vallecula (10%).267,341,358 Tumor growth within the fourth ventricle eventually leads to occlusion of the ventricle, with resultant obstructive hydrocephalus. Ependymomas are highly cellular with small uniform cells that tend to form characteristic perivascular pseudorosettes. Although they do not invade through the ventricular walls, they are firmly attached to the ventricular floor, and complete resection of the tumor base is frequently not possible. The rate of local tumor recurrence after surgical resection is 90% to 95% within 3 years, and the 5-year survival is currently 45% to 70%.147,271,359 Furthermore, about 25% of ependymomas exhibit features of anaplasia, with a high mitotic rate, cellular pleomorphism, and intratumoral necrosis; these tumors are known as anaplastic ependymomas (WHO grade III).24,69 Intratumoral calcification and cyst formation are common in ependymomas (20%-50% of reported cases).11,267,379,418 Peritumoral edema is a prominent and frequent finding in association with both cerebellar and cerebral ependymomas. On noncontrast CT, fourth ventricle ependymomas appear as hypodense or isodense heterogeneous, well-circumscribed, rounded masses within the fourth ventricle that partially or completely obliterate the cavity of the ventricle.232,267,358,379 The tumor is typically well defined by a prominent surrounding hypodense halo of peritumoral edema. Aggregates of calcification, usually small and round but sometimes quite large, are found in up to half of these lesions, and focal lucencies resulting from cyst formation are nearly as common.232 In supratentorial tumors, the cysts are both larger and more frequent (up to 80%) than in fourth ventricle ependymomas (see Fig. 9-20A).11 Intratumoral hemorrhage is relatively less common, with reported frequency of 20% or less.232 After IV administration of contrast agent,

193

the solid portions and cyst walls of the tumor mass exhibit moderate but variable heterogeneous contrast enhancement.11,114,232,379,418 The MRI appearance of ependymomas is heterogeneous, reflecting mixed signals from intratumoral calcification, cysts, and hemorrhage. Usually these tumors are isointense to hypointense relative to white matter on T1-weighted images and isointense to hyperintense to white matter on T2-weighted images (Fig. 9-19A and B; Fig. 9-20B).114,232,369 Contrast enhancement is the rule, but it is variable in degree and  typically not uniform, seen in both ependymoma and anaplastic ependymoma. (see Figs. 9-19C and 9-20C). Although the MRI features  of ependymoma are nonspecific,133,369 the multiplanar capabilities of this modality offer unique visualization of infratentorial tumor extent within the cranial vault and the region of the cervicocranial junction.15 On both CT and MRI as well as in the clinical setting, the presentation of ependymoma of the fourth ventricle may closely resemble that of medulloblastoma and PA. The incidence and extent of intratumoral calcification, heterogeneity of density or signal intensity, and propensity for tumor extension into the fourth ventricular recesses are greater in ependymomas than in medulloblastomas or astrocytomas. In general, medulloblastomas appear more dense on CT with a lower frequency of intratumoral calcification, and they exhibit more homogeneous contrast enhancement than ependymomas. Differentiation can be difficult but is aided by DWI, as discussed later in this chapter. Astrocytomas tend to occur in slightly older patients and are much less frequently centered in the midline. Subependymoma.  A rare variant of ependymoma that actually consists of both highly differentiated ependymal cells and astrocytes, subependymoma is classified as a benign (WHO grade I) glial neoplasm. It is considered to arise from beneath the ventricular lining and projects into the ventricular cavity as a sharply outlined intraventricular mass but does not extend deeply from its base into the adjacent brain parenchyma (Fig. 9-21). Subependymomas are typically solid and homogeneous, although larger tumors may contain calcifications, microcystic changes, and evidence of intratumoral hemorrhage.16 Subependymomas occur most commonly (50%-60%) in the fourth ventricle but also arise in the lateral ventricles. Most present clinically in adults, but many of these lesions are asymptomatic and found  incidentally at autopsy.15 Those that become symptomatic (causing obstructive hydrocephalus) occur mostly in the lateral ventricles (Fig. 9-22) in relation to the septum pellucidum and foramen of Monro and much less commonly in the region of the aqueduct. On CT and MRI, symptomatic lesions in the fourth or lateral ventricles usually appear homogeneous in density and signal intensity. They appear homogeneously isodense or slightly hypodense on CT. On T2-weighted MRI the tumor appears homogeneously hyperintense to adjacent brain. There is variable contrast enhancement after IV administration of contrast medium.195 However, larger tumors may exhibit internal heterogeneity of density and signal owing to intratumoral calcification and hemorrhage.114,369 The differential diagnosis includes central neurocytoma, ependymoma, and subependymal giant cell astrocytoma. Patient age, nature of the symptoms, presence of associated disease, tumor size, presence and extent of calcification, cystic change, evidence of hemorrhage, and contrast enhancement are the key factors in making the differentiation.

Choroid Plexus Papilloma and Carcinoma.  Papillomas arising in the choroid plexus are benign intraventricular neoplasms that locally expand the involved ventricle and cause hydrocephalus, which may  be secondary to ventricular obstruction, overproduction of CSF, or

http://radiologyebook.com

Text continued on p. 198

194

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-19  Ependymoma of the fourth ventricle. A, Axial T1-weighted MRI demonstrates a large predominantly hypointense mass filling and distending the fourth ventricle and compressing the pons from behind. Only a small more hypointense crevice at the right lateral margin of the tumor (arrow) indicates the presence of CSF in the markedly compressed lumen of the fourth ventricle. Small foci of hypointensity within the tumor likely represent areas of calcification. Note the marked dilatation of the temporal horns secondary to obstruction of the fourth ventricle. B, On this axial T2-weighted image, the intraventricular tumor appears heterogeneously mildly hyperintense, while the signal within the residual compressed fourth ventricle (arrow) is strongly hyperintense. Intratumoral calcification is strongly hypointense. C, Following IV gadolinium, an axial T1-weighted MRI demonstrates irregular heterogeneous intratumoral contrast enhancement (arrow).

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

C FIG 9-20  Supratentorial ependymoma. A, Axial CT image after IV contrast demonstrates a cystic-appearing hypodense mass with irregular rimlike contrast enhancement (arrow) in the medial aspect of the left temporal lobe. B, On this axial T2-weighted MRI, the tumor (arrow) exhibits mixed signal intensity; it is predominantly isointense to gray matter and hypointense to CSF. A tiny ovoid hypointensity in its posteromedial surface likely represents calcification. There is extensive surrounding edema in the temporal lobe and in the lateral aspect of the left cerebral peduncle. C, Axial T1-weighted postgadolinium image demonstrates irregular rimlike and central contrast enhancement (arrow) with medial extension into the lateral margin of the midbrain.

http://radiologyebook.com

195

196

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-21  Subependymoma of the fourth ventricle and vestibular schwannoma in the right cerebellopontine angle. Axial T1-weighted (A) and T2-weighted (B) MRIs demonstrate a small, sharply marginated, isointense (T1-weighted)/hyperintense (T2-weighted) nodular intraventricular mass in the central and left lateral aspects of the fourth ventricle (white arrows). The mass appears continuous with the left lateral wall of the ventricle. Also evident is a rounded heterogeneously hypointense (T1-weighted)/hyperintense (T2-weighted) mass in the right cerebellopontine angle and internal auditory canal that indents the lateral aspect of the pons and middle cerebellar peduncle and the anterior margin of the right cerebellar hemisphere (black arrows). C, Axial T1-weighted image after IV gadolinium administration demonstrates nearly homogeneous contrast enhancement of both lesions (arrows); the uptake of contrast in the intraventricular nodule is more subtle than in the cerebellopontine angle lesion.

http://radiologyebook.com

A

B

C

D FIG 9-22  Large supratentorial subependymoma. A, Axial T1-weighted MRI demonstrates an intraventricular mass with slight signal heterogeneity in the body of the right lateral ventricle. This mass abuts both the septum pellucidum medially and the corona radiata laterally. Both lateral ventricles appear enlarged. B, On the axial T2-weighted image, the mass also appears mildly heterogeneous. There are small foci within the mass that follow CSF signal intensity on both the T1- and T2-weighted images. C, On the axial turbo FLAIR image, the mass demonstrates heterogeneous signal hyperintensity. The focal areas that followed CSF signal intensity on the T1- and T2-weighted images have their signal suppressed on this image and are consistent with cystic areas containing simple CSF-like fluid. There is no evidence for edema in the right corona radiata adjacent to the mass. There is abnormal signal hyperintensity in the subependymal regions of the brain adjacent to the anterior and posterior aspects of lateral ventricles bilaterally that, along with the lateral ventricular enlargement, suggests hydrocephalus with transependymal flow of CSF; this is secondary to obstruction of the foramen of Monro bilaterally. D, Axial T1-weighted MRI after administration of gadolinium demonstrates only a few scattered linear areas of enhancement within the mass that may represent small blood vessels. There is no significant enhancement of the mass itself.

http://radiologyebook.com

198

PART II  CT and MR Imaging of the Whole Body

impairment of CSF absorption. Choroid plexus papillomas represent less than 1% of all brain tumors but are more common in young children (3% of all pediatric brain tumors), particularly in the neonatal period; 85% occur in the first 5 years of life, and these tumors represent 40% of all brain tumors in the first 60 days of life.232,323,358 The atria of the lateral ventricles and fourth ventricle are the most common sites of involvement. An age disparity exists with regard to tumor location, with lateral ventricle tumors occurring mainly in young children, and fourth ventricle papillomas appearing more commonly in adults.341 It is not uncommon for fourth ventricle tumors to extend through the lateral recesses and protrude into the cerebellopontine angle cisterns.294 Histologically, choroid plexus papillomas consist of welldifferentiated proliferation of both the surface epithelium of the choroid plexus and the underlying vascular connective tissue core and are usually classified as WHO grade I.341 They grow slowly and are noninvasive, tending to expand within the confines of the ventricle or its outlet foramina.294,358 Occasionally, however, the epithelium may exhibit malignant change, and in such cases the tumors are classified as carcinomas; they may extend into the adjacent brain parenchyma.257 The prognosis in these malignant choroid plexus tumors depends on the completeness of surgical resection, but overall the 5-year survival ranges from 26% to 40%. Because of the friability and high vascularity of choroid plexus papillomas, intratumoral and intraventricular hemorrhage are common.69,359 On CT, choroid plexus papillomas appear as large, rounded or lobulated, isodense or hyperdense intraventricular masses that engulf and separate the normal choroid plexus calcifications and exhibit intense homogeneous contrast enhancement. Small hypodense nonenhancing foci are occasionally identified within these tumors.189,257 On MRI, choroid plexus papillomas are usually isointense or hypointense with respect to brain on T1-weighted images and heterogeneously hyperintense relative to brain on T2-weighted images (Fig. 9-23A and B)232; the heterogeneity of signal intensity within the tumor mass may be due to areas of vascularity, calcification, or previous hemorrhage (see Fig. 9-23A). As noted previously the tumor typically expands the ventricle locally at the site of origin and may obstruct ventricular drainage, causing dilatation (entrapment) of the more proximal portion of the affected ventricle. On both CT and MRI, these tumors usually demonstrate intense contrast enhancement (see Fig. 9-23C), which may be homogeneous or heterogeneous depending on the tumor contents. The differential diagnosis of choroid plexus papilloma on imaging studies usually includes ependymoma in older children and adults and meningioma in adults. The lobulated tumor margins, lack of invasion of brain parenchyma despite the large tumor size, and heterogeneous signal pattern within the tumor all aid in distinguishing this lesion.

Nonglial Tumors

Neuronal and Mixed Neuronal and Glial Tumors.  In this group of tumors of neuroepithelial origin, neuronal and glial differentiation is present in varying degrees. This chapter will discuss the following recognized entities: ganglioglioma, gangliocytoma, dysplastic gangliocytoma of the cerebellum (Lhermitte-Duclos disease), desmoplastic infantile ganglioglioma, dysembryoplastic neuroepithelial tumor, and central neurocytoma.188,358 Ganglioglioma and gangliocytoma.  Gangliogliomas are un­ common low-grade tumors consisting of a mixture of neoplastic  astrocytes and dysplastic neurons. Slow-growing neoplasms with  well-differentiated ganglion cells and a low-grade astrocytomatous stroma,341 these tumors are considered WHO grade I or II.69 Ganglio-

cytomas (also known as ganglioneuromas) are relatively less common and contain only neuronal elements; they are classified as WHO  grade I.232 Together these tumors account for 1.3% of all primary brain tumors. Gangliogliomas and gangliocytomas occur in young people (peak incidence in the second decade of life) and constitute approximately 3% to 4% of primary brain tumors in children.267 They are found most commonly in the temporal lobes but may occur anywhere in the cerebral hemispheres.318 The clinical presentation is usually with seizures of long duration, and gangliogliomas are considered the most common cause of chronic temporal lobe epilepsy. These tumors are usually wellcircumscribed solid masses but may include one or more cysts.48 Gross total resection of a tumor is the treatment of choice. Overall, gangliogliomas and gangliocytomas have a 92% 3-year survival.173 In approximately 10% of cases, however, the tumor is more aggressive; in these cases, the malignant element is always glial.69 Diagnostic imaging findings are relatively nonspecific and similar to those in other low-grade intracerebral neoplasms.232 Noncontrast CT scans demonstrate relatively well-circumscribed hypodense (occasionally isodense with adjacent brain) lesions typically located in the periphery of the hemisphere, with little associated mass effect or surrounding vasogenic edema. Foci of calcification are identified in 33% to 40% of cases, and cysts in 50% of cases.422 Calcification is less common in noncystic tumors. Rarely a peripherally located indolent tumor may erode the inner table of the overlying skull. Hemorrhage is rare in these tumors. Contrast enhancement is observed in about 50% of cases and may be homogeneous or heterogeneous depending on the presence and size of cystic and calcific changes.48,173,267,422 On MRI, these tumors (like many other intracerebral neoplasms) are usually heterogeneous in appearance, predominantly hypointense to gray matter on T1-weighted images and hyperintense on T2 (Fig. 9-24A).48,385 Mild to moderate contrast enhancement of the solid noncystic portion of the tumor is found in the majority of cases on both CT and MRI (see Fig. 9-24B). Findings on CT and MRI in these ganglion cell tumors are not characteristic; oligodendroglioma is a common differential consideration. However, in a young patient with a protracted history of seizures and a relatively well-circumscribed cystic tumor containing calcification in the periphery of the temporal lobe with little or no associated mass effect, ganglioglioma should be strongly considered.48,173,326,422 Dysplastic gangliocytoma of the cerebellum.  A rare lesion, dysplastic gangliocytoma of the cerebellum (Lhermitte-Duclos disease) is considered a complex hamartoma or malformation and is classified as WHO grade I. Dysplastic gangliocytoma may exist as an isolated entity or as part of Cowden’s disease, a rare syndrome with multiple mucosal hamartomas. The clinical presentation is usually ataxia and signs of increased intracranial pressure in a young adult. A large dysplastic mass occupies most or all of the cerebellar hemisphere, with thickening of the granular and molecular layers of the overlying cerebellar cortex (Fig. 9-25A). The enlarged mass causes displacement and compression of the fourth ventricle. On MRI the mass is poorly demarcated, hypointense on T1-weighted images and hyperintense on T2-weighted images, and the thickened folia present a distorted striate pattern (see Fig. 9-25B). Contrast enhancement  of either the tumor or the thickened cerebellar folia is unusual (see Fig. 9-25C).201 Desmoplastic infantile ganglioglioma.  Desmoplastic infantile ganglioglioma (DIG) is an unusual lesion that occurs in young infants. It consists of immature neurons and astrocytes together with extensive fibrocollagenous (desmoplastic) thickening and cyst formation. The solid desmoplastic portion is often located adjacent to the meninges. Despite its variegated appearance, this tumor is classified as WHO

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

C FIG 9-23  Choroid plexus papilloma of the left lateral ventricle. A, Axial T1-weighted MRI demonstrates an intraventricular mass with heterogeneous signal intensity in the body of the left lateral ventricle (arrow). The lateral portion of this mass is hyperintense (likely indicating the presence of subacute hemorrhage), while the medial portion is mainly hypointense relative to the white matter. B, On this axial T2-weighted image, the tumor (arrow) appears slightly hyperintense to adjacent brain but hypointense to the fluid in the lateral ventricles. C, Axial T1-weighted postgadolinium image demonstrates heterogeneous contrast enhancement of the tumor mass (arrow).

http://radiologyebook.com

199

200

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-24  Ganglioglioma of the right temporal lobe. A, Axial T2-weighted MRI demonstrates a wellcircumscribed cystic mass (arrow) with a hypointense rim (secondary to calcification) in the region of the right parahippocampal gyrus and surrounding hyperintensity (due to edema). There is associated mass effect with effacement of the sulci of the right temporooccipital convexity, but there is no evidence of tentorial herniation. B, On this postgadolinium T1-weighted coronal image, peripheral rimlike contrast enhancement (arrow) outlines the margins of this cystic tumor, which is located immediately above the right leaf of the tentorium. The contents of the cyst are slightly hyperintense relative to CSF, reflecting a higher protein content.

grade I and has a favorable long-term prognosis after complete excision.173 DIGs are large and superficial hemispheric tumors that occur mainly in the frontal and parietal lobes and present a markedly heterogeneous appearance on both gross inspection and MRI (Fig. 9-26). The solid portions of this tumor are hypointense relative to normal brain on T2-weighted images and frequently demonstrate intense contrast enhancement. DIG differs from the usual ganglioglioma in that it presents in infancy, is predominantly frontoparietal in location, and has dense desmoplasia.232 On MRI, the intermixed collagenous and multicystic pattern of this lesion most resembles that of a primitive neuroectodermal tumor (PNET), which carries a much less favorable prognosis; however, PNET is also characterized by the common presence of intratumoral calcification, hemorrhage, and necrosis, findings that are uncommon in DIG.173,232 Dysembryoplastic neuroepithelial tumor.  Dysembryoplastic neuroepithelial tumors (DNETs) are uncommon benign intracortical tumors that characteristically occur in young patients (<20 years) above the tentorium, mainly in the temporal lobe (60%-80%) or frontal lobe.55,370,192,358 They can be multifocal and rarely arise in the septum pellucidum. Patients typically present with intractable partial seizures. Partial or complete excision is the treatment of choice, and no recurrences have been reported. The tumor is termed dysembryoplastic because of its multinodularity, with areas of cortical dysplasia at the margins between the tumor nodules and adjacent brain.55,370

DNETs are lesions of long standing that most frequently involve the convexity cortex and often protrude beyond the adjacent cortical margin, eroding the overlying inner table of the calvarium.192,358 The imaging findings, except for the superficial cortical location, are similar to those of other low-grade glial tumors (astrocytoma, oligodendroglioma). On CT, DNET typically appears as a well-circumscribed hypodense superficial mass, occasionally with intratumoral calcification or cystic change or both.358 MRI demonstrates a mass centered in the convexity cortex and bulging externally that is hypointense to adjacent brain on T1-weighted images (Fig. 9-27B) and hyperintense on T2-weighted images (see Fig. 9-27A) with no surrounding edema.192,358 The protruding external margin may present a “soap bubble” appearance, reflecting internal cystic change.203 Contrast enhancement is seen in only a minority of these lesions (see Fig. 9-27C).192,281 Central neurocytoma.  A relatively rare benign slow-growing (WHO grade II) intraventricular tumor, central neurocytoma has now been recognized as a separate entity.69,358 It occurs almost exclusively in the anterior portion of the lateral ventricle, arising from the superolateral ventricular wall and extending medially adjacent to the septum pellucidum and the foramen of Monro; extension through the foramen into the third ventricle is unusual. No such tumors have been reported arising in the third or fourth ventricles, and only a few hemispheric lesions have been found. Central neurocytomas are tumors of young adults and are rare in children and older adults. Patients typically present clinically with headache and signs of increased intracranial

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

C FIG 9-25  Dysplastic gangliocytoma of the cerebellum (Lhermitte-Duclos). A, Axial CT image without IV contrast demonstrates a poorly marginated heterogeneous predominantly hypodense mass (arrow) occupying most of the right cerebellar hemisphere and displacing the fourth ventricle to the right of the midline. A striate pattern of calcification is noted within this process. B, On this T1-weighted axial noncontrast MRI the poorly marginated hypointense mass presents a distorted internal architecture (arrow) and compresses and displaces the pons anteriorly and to the left. C, Axial T1-weighted image post IV gadolinium; there is no evident contrast enhancement of this hamartomatous process (arrow).

http://radiologyebook.com

201

202

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-26  Desmoplastic infantile ganglioglioma (DIG). A, Axial T2-weighted MRI demonstrates a wellcircumscribed heterogeneous left high frontal convexity tumor that is predominantly hypointense to adjacent brain (arrow). On its posteromedial aspect, there are two small cysts and one large ovoid cyst on the left and a single large cyst on the right, all with hyperintense contents. There is extensive white matter edema bilaterally surrounding the large cysts and extending into the left frontoparietal convexity gyri. Following IV gadolinium, axial (B) and left parasagittal (C) T1-weighted images demonstrate intense contrast enhancement of the superficial solid area of this tumor (arrows), which consists of thick dense fibrocollagenous (desmoplastic) tissue. Contrast enhancement is also seen to a variable degree in the rims of the cystic elements of this lesion. On C, note the posteroinferior displacement and compression of the atrium and temporal horn of the left lateral ventricle by the large left cystic mass.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

C FIG 9-27  Dysembryoplastic neuroepithelial tumor (DNET). A, Axial T2-weighted MRI demonstrates a wellcircumscribed superficial cystic mass with heterogeneous signal hyperintensity in the right midfrontal convexity cortex with an outwardly bulging external surface. B, Axial turbo FLAIR image through the lesion demonstrates suppression of signal in the anterior aspect consistent with relatively simple fluid and persistence of mild signal hyperintensity in the posterior aspect consistent with either proteinaceous fluid or a soft tissue component. C, Axial T1-weighted image after IV gadolinium demonstrates no contrast enhancement within the cyst. The lesion is eroding the inner table of the overlying calvarium.

http://radiologyebook.com

203

204

PART II  CT and MR Imaging of the Whole Body

pressure secondary to ventricular obstruction.69 Partial or complete surgical excision is the treatment of choice, and no deaths from tumor growth or recurrence have been reported.413 Histologically, central neurocytoma is identical to oligodendroglioma. It is characterized by nests of small well-differentiated cells separated by thin fibrovascular septa.15 Intratumoral calcification and multiple small cysts are common findings, but hemorrhage (either intratumoral or intraventricular) is unusual. It is only with electron microscopy and immunohistochemistry that its neuronal origin can be determined.69 CT demonstrates a sharply marginated, inhomogeneous, isodense to slightly hyperdense intraventricular mass with a broad-based attachment to the wall of the frontal horn or anterior body of the lateral ventricle and abutting the septum pellucidum. Calcifications, either coarse or globular, and multiple small intratumoral hypodense cysts are found in 50% or more of central neurocytomas. Mild to moderate contrast enhancement of the isodense to hyperdense regions is common.123,413 The findings on MRI are similar. The tumor demonstrates a heterogeneous pattern with intermixed foci of isointensity (compared with adjacent gray matter) and hypointensity (due to calcifications and cysts) on T1-weighted images and isointensity to hyperintensity on T2-weighted images (Fig. 9-28A). Inhomogeneous mild contrast enhancement is common (see Fig. 9-28B).123,413 The differential diagnosis is that of a well-circumscribed intraventricular mass at or near the foramen of Monro and includes subepen-

A

dymal giant cell astrocytoma and subependymoma. The absence of deep extension into adjacent brain parenchyma, associated nodular tubers and intratumoral hemorrhage, and the relatively young age of the patient favor the diagnosis of central neurocytoma.15

Pineal Parenchymal Tumors.  Tumors arising within the parenchyma of the pineal gland constitute less than 1% of all primary brain tumors in adults.69 However, pineal tumors represent nearly 10% of all pediatric brain tumors.232 The most common tumors to involve the pineal gland are the germ cell tumors, which constitute 50% to 70% of pineal region tumors and are discussed elsewhere in this chapter. Pineal parenchymal tumors represent only 15% to 30% of   pineal region neoplasms.360,419 They include a range from primitive undifferentiated (pineoblastoma) through transitional forms (primary parenchymal tumor of intermediate differentiation) to welldifferentiated (pineocytoma).232,350 Pineoblastoma.  Pineoblastomas are pediatric tumors with a peak incidence in the first 2 decades of life and rare after age 30. These malignant highly cellular undifferentiated small cell tumors are histologically and histochemically identical to other primitive neuroectodermal tumors and exhibit similar biological behavior. They are locally invasive and may spread into the third ventricle, thalami, and quadrigeminal plate. Pineoblastomas frequently contain areas of hemorrhage and necrosis, and intratumoral calcification is common.15,350 Presenting symptoms include precocious puberty (possibly related to destruction of normal pineal tissue with loss of melatonin secretion),26

B FIG 9-28  Central neurocytoma in the anterior body and frontal horn of the left lateral ventricle. A, Axial T2-weighted MRI demonstrates a large multilobulated heterogeneous but predominantly isointense (to gray matter) intraventricular mass arising from the lateral wall of the left lateral ventricle (arrow) with multiple small hyperintense foci and one large intratumoral hyperintense cyst. The tumor expands the left lateral ventricle. Enlargement of both lateral ventricles is likely secondary to obstruction of the foramen of Monro by this mass. B, Coronal T1-weighted spoiled gradient echo image following IV gadolinium administration; the plane of section passes through the anterior portion of the tumor. The more solid portion of this tumor demonstrates faint contrast enhancement (arrow). Note extension of the tumor inferiorly and medially at a level just anterior to the foramen of Monro.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms Parinaud’s syndrome and other cranial nerve deficits, and obstructive hydrocephalus. Dissemination via the CSF with leptomeningeal and ependymal metastases occurs in 10% of cases and may be found at the time of initial diagnosis. The 5-year survival rate is 58%.350 On CT, pineoblastomas appear hyperdense with irregular intratumoral calcification, infiltration into neighboring structures, and mass effect on the posterior aspect of the third ventricle and the tectum of the midbrain. They demonstrate dense contrast enhancement. On MRI, pineoblastomas appear heterogeneous in signal, predominantly hypointense to gray matter on T1-weighted images and isointense to gray matter on T2-weighted images, with interspersed foci of hypointensity secondary to intratumoral necrosis and calcification15,360,388,419 (Fig. 9-29A and B). As on CT, contrast enhancement of the solid portion of the tumor is the rule (see Fig. 9-29C). The chief differential consideration is germinoma, which may look identical on CT and MRI. A rare occurrence that is nevertheless important for recognition and management is the association of bilateral retinoblastomas with pineoblastoma, often termed trilateral retinoblastoma. This condition, an inherited disorder occurring in very young children (<2 years), is found in 3% of patients with bilateral retinoblastomas.318 Pineocytoma.  Pineocytomas are tumors of adults, with a mean age of presentation of 36 years. Compared with pineoblastomas, these lesions are less densely cellular with well-differentiated cells that have more cytoplasm arranged in lobules in an architecture similar to that of the normal pineal gland.24,350 They tend to follow a slower and more benign clinical course, but transitional tumors intermediate in cellularity and mitotic activity also exist.69,350 Intratumoral calcification is present in more than 50%. Subarachnoid seeding is not seen in the more benign neoplasms. The 5-year survival for pineocytomas is 67%.350 On CT, the typical appearance is of a rounded, well-circumscribed, homogeneously hyperdense lobulated expansile mass with intratumoral calcification situated at and often obstructing the posterior margin of the third ventricle232,318 (Fig. 9-30A). There is no evidence of invasion of neighboring structures. Multiplanar MRI demonstrates  a rounded or ovoid mass compressing the posterior margin of the  third ventricle and the superior aspect of the quadrigeminal plate. The mass is homogeneously hypointense to isointense on T1-weighted images and hyperintense relative to gray matter on T2-weighted images.232,360,388,419 The signal characteristics reflect the higher water content of these more mature tumors compared with pineoblastomas.15 Contrast enhancement of all or a major portion of a pineocytoma is the rule on both CT (see Fig. 9-30B) and MRI.

Embryonal Tumors.  The terminology regarding primitive or embryonal tumors of neuroepithelial origin has been a source of controversy for more than 2 decades. Gradually, however, a consensus appears to be emerging concomitant with advances in applied molecular biology.24 PNETs of the CNS appear to be one group of a larger category of embryonal tumors that occur throughout the body (e.g., neuroblastomas of the adrenal and sympathetic ganglia).24,335 These are the second most common intracranial tumors of childhood, representing approximately 15% to 20% of all childhood brain tumors and exceeded in frequency only by astrocytomas. They have a common histology consisting of densely packed undifferentiated round cells with a high nucleus-to-cytoplasm ratio and high mitotic activity but follow divergent patterns of differentiation.335 The classic PNET is the medulloblastoma of the cerebellar vermis, which constitutes 85% of this group of tumors. The remaining 15%, almost all supratentorial, include pineoblastomas (see earlier), ependymoblastomas, and medulloepitheliomas. The histogenesis of PNETs

205

is undetermined, but one hypothesis is that they arise from immature pluripotential precursor cells in the subependymal layers of the developing ventricular system and in the external granular layer of the cerebellum.24,335 Medulloblastoma.  An invasive and highly malignant tumor (WHO grade IV), medulloblastoma is the second most common primary brain tumor of childhood and the most common pediatric posterior fossa neoplasm.126 Although medulloblastomas may occur at any age, 75% present clinically before age 15. There are two distinct age peaks, at 4 to 8 years and 15 to 35 years. In children, medulloblastomas are primarily tumors of the cerebellar vermis (Fig. 9-31), but in adults they tend to be located more laterally in the cerebellar hemispheres (see Fig. 9-31).7,191,335,341 The typical clinical presentation consists of signs of obstructive hydrocephalus, cerebellar dysfunction, or both—headache, nausea, vomiting, and ataxia. These aggressive tumors grow rapidly with extension from the vermis into the fourth ventricle and spread inferiorly into its recesses and outlet foramina and superiorly into the aqueduct.358,392 More than 90% present clinically with hydrocephalus.232 They have a strong tendency to seed via the CSF, and approximately 40% demonstrate subarachnoid or retrograde ependymal metastases at the time of initial diagnosis.107 They are highly radiosensitive, and the standard treatment approach is gross total resection plus chemotherapy and total craniospinal axis irradiation. Ten-year survival rates are in the range of 50% to 70%, but the prognosis is poorer in children younger than 3 years.92,232 Recurrence after the initial treatment regimen is about equally distributed between local regrowth at the primary tumor site and distant metastasis. Systemic metastasis, a highly unusual occurrence with most CNS primary neoplasms, is identified in approximately 5% of patients, mainly to bone. Compared with other highly aggressive brain tumors, medulloblastomas appear relatively discrete and solid with convex margins on gross inspection but may contain foci of necrosis and hemorrhage. Microscopically the tumor cells tend to be arranged in pseudorosettes, and areas of vascular proliferation, cellular pleomorphism, and calcification are occasionally identified. Electron microscopy and immunostaining usually confirm neuronal differentiation. Approximately 15% of medulloblastomas contain a significant amount of collagen and reticulin and may demonstrate a nodular pattern; designated desmoplastic medulloblastomas, these tumors are relatively more common in the cerebellar hemispheres in adolescents and young adults.24,220 On noncontrast CT, medulloblastomas appear as rounded or oblong, mainly homogeneous, isodense to slightly hyperdense masses centrally located in the inferior vermis and cavity of the fourth ventricle (see Fig. 9-31A).267 Differentiation of medulloblastoma from diffuse solid cerebellar astrocytoma is most reliably made on noncontrast CT. The solid hyperdense appearance of medulloblastoma reflects the dense cellular nature of this tumor, whereas astrocytomas are typically rather uniformly hypodense.113,115 Obstruction of the fourth ventricle, inferior aqueduct, or both with resulting enlargement of the third and lateral ventricles is commonly seen on CT at the time of clinical presentation. Punctate or nodular intratumoral calcifications are identified in 10% to 20% of cases.191,392,420 Hypodense intratumoral foci of necrosis and cysts are found in up to 50% of medulloblastomas (see Fig. 9-31A).267,420 Many medulloblastomas are surrounded by a thick poorly marginated hypodense halo representing extensive peritumoral edema (see Fig. 9-31A and B). The imaging findings in young adults are similar except for the more lateral (hemispheric) location of the tumor. After IV administration of contrast medium, moderate to intense homogeneous enhancement of the solid tumor mass is the rule  in medulloblastomas,191,392 although approximately 10% do not

http://radiologyebook.com

206

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-29  Pineoblastoma. A, Axial T2-weighted MRI demonstrates a small ovoid tumor mass (arrow) arising in the region of the pineal gland and protruding posteriorly into the quadrigeminal plate cistern. The tumor is heterogeneous in signal but predominantly hyperintense to gray matter. Axial T1-weighted images before (B) and after (C) IV gadolinium administration demonstrate intense homogeneous contrast enhancement of the entire tumor mass (arrows). Note the blunting of the posterior margin of the third ventricle by this pineal tumor.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

207

B FIG 9-30  Pineocytoma. Axial noncontrast (A) and postcontrast (B) CT images demonstrate a wellcircumscribed markedly hyperdense mass in the midline (arrow in A) in the pineal region, with bilateral lobulated intense contrast enhancement impinging on the medial aspects of both thalami (arrow in B). The tumor is partially obstructing the posterior aspect of the third ventricle, accounting for the enlarged lateral and anterior third ventricles.

demonstrate contrast enhancement on CT.267 Areas of tumor seeding in the posterior fossa cisterns or in the third or lateral ventricles exhibit similar homogeneous contrast enhancement in a nodular or sheetlike configuration. The appearance of medulloblastomas is less specific on MRI than on CT. The tumors are mildly hypointense to isointense relative to the adjacent cerebellar vermis on T1 and isointense to slightly hyperintense on T2-weighted images (Fig. 9-32A; see Fig. 9-31B and C). Isointensity (or even hypointensity) of the tumor matrix on T2-weighted images is likely due to the hypercellularity and high nucleus-to-cytoplasm ratio of this tumor.80,133,226 However, heterogeneity of signal intensity is often seen on T2-weighted images (see Figs. 9-31C and 9-32B) secondary to intratumoral cysts, necrosis, vascular flow voids, and hemorrhage.232 Contrast enhancement of the solid portions of the tumor, seen in more than 90% of patients, is typically intense and homogeneous (see Figs. 9-32B and 9-75A) but may be irregular and patchy (see Fig. 9-31D). MRI with contrast enhancement is very sensitive for detecting tumor spread and metastatic seeding in the cranial and spinal subarachnoid spaces.332,358,392 CNS primitive neuroectodermal tumor.  Some 15% of all PNETs of the CNS occur above the tentorium.24 However, they represent less than 1% of all pediatric brain tumors. Like their infratentorial counterparts, they occur mainly in the very young (65% of patients < 5 years; 85% < 10 years). Patients may present clinically with seizures or with symptoms and signs of increased intracranial pressure. The distribution of these aggressive tumors within the cerebral hemispheres is, in diminishing order of frequency, frontal, parietal, temporal, and occipital.232 Approximately one third are pineoblastomas; the remain-

der are central neuroblastomas, ependymoblastomas, ganglioneuroblastomas, medulloepitheliomas, and other rare immature round cell tumors.52,80,92,335 On CT, CNS PNETs appear heterogeneous and well circumscribed. The heterogeneity reflects cystic or necrotic changes and intratumoral hemorrhage. The solid elements are hyperdense because of their high cellularity and demonstrate contrast enhancement. Coarse calcifications are described in 50% to 70%, a much higher proportion than  in cerebellar medulloblastomas.232,335 CNS PNETs share with their infratentorial counterparts a strong propensity for spread via CSF.80 MRI also demonstrates heterogeneity of signal intensity on both T1- and T2-weighted images (Fig. 9-33; see Fig. 9-73). These are often large tumors with relatively lesser degrees of surrounding edema.80,92 The differential diagnosis includes other large well-circumscribed supratentorial tumors such as ependymomas and oligodendrogliomas.232 As on CT, the solid portions of the tumor demonstrate strong contrast enhancement, a feature that may aid differentiation.

Tumors of the Cranial Nerves Approximately one third of all intracranial masses are nonglial primary neoplasms of the CNS, almost all of which are extraaxial in origin and location. The most common of these are schwannomas of the cranial nerves and meningiomas. Within the brain and spinal cord, the nerve fiber tracts are surrounded by oligodendrocytes. However, as the cranial nerves emerge from the brain, there is a transition zone beyond which the oligodendrocytes do not extend but are instead replaced by Schwann cells (also called neurilemmal cells), which surround and ensheath the extraaxial

http://radiologyebook.com

208

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 9-31  Medulloblastoma of the cerebellar vermis in an 11-year-old child. A, Axial noncontrast CT image demonstrates a large lobulated hyperdense round tumor (arrow) with an internal hypodense cavity to the right of the midline that compresses the fourth ventricle from behind. An ill-defined faintly hypodense band surrounding the hyperdense mass represents white matter edema. The intratumoral area of marked hypodensity to the right of the midline likely represents necrosis. A smaller focus of lesser hypodensity on the left posterolateral tumor margin suggests an intratumoral cyst. These findings are confirmed on noncontrast axial T1-weighted (B) and T2-weighted (C) MRIs; the solid portion of the tumor appears mildly hypointense on T1-weighting and mildly hyperintense on T2-weighting (arrow). Following IV gadolinium, an axial T1-weighted image (D) demonstrates irregular patchy contrast enhancement of the solid areas of the tumor (arrow).

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

209

B FIG 9-32  Medulloblastoma of the anteroinferior right cerebellar hemisphere in a 26-year-old patient. A, Axial T2-weighted MRI demonstrates a poorly circumscribed mass with a heterogeneous signal pattern in the anteroinferior portion of the right cerebellar hemisphere. Irregular areas of pronounced hyperintensity (arrow) are interspersed in a mildly hyperintense background that likely represents surrounding edema. The tumor extends medially into the inferior vermis. B, After IV gadolinium the tumor demonstrates homogeneous contrast enhancement with well-circumscribed margins (arrow). Punctate intratumoral hypointensities represent enlarged vascular channels and/or focal calcifications.

portions of the cranial nerves, spinal nerve roots and nerves, and peripheral nerves. Within the cranial vault, nearly all tumors arising from these nerves represent schwannomas. The other major primary tumor of the cranial nerves is neurofibroma. Neurofibromas differ from schwannomas not only in site of origin but also in histology and natural history. Malignant degeneration is essentially unknown in schwannomas, but both malignant degeneration of neurofibromas and primary malignant peripheral nerve sheath tumors are recognized but uncommon occurrences.

Schwannoma.  Schwannoma is a benign (WHO grade I) encapsulated tumor that arises from Schwann cells of nerve sheaths of cranial and spinal nerves. It is estimated to account for between 5% and 10% of primary intracranial neoplasms and for nearly 30% of intraspinal tumors.47 Intracranial schwannomas occur in all age groups, but the peak incidence is in the fourth through seventh decades of life. There is a distinct sex predilection, with a female-to-male ratio of 2 : 1, although no such predilection exists in regard to spinal schwannomas.47 By far the most common site of intracranial involvement is the superior vestibular division of the eighth cranial nerve.241 Trigeminal nerve schwannomas are much less common, with facial, trochlear, and abducens nerve tumors only rarely reported. Only the first (olfactory) and second (optic) cranial nerves lack Schwann cell sheaths and are therefore not potential sites of origin. The clinical presentation varies according to the site of origin. Vestibular schwannoma (also called acoustic neuroma or neurinoma) typically presents with symptoms of a mass in the cerebellopontine angle, including tinnitus, sensorineural

hearing loss, and facial paresthesias. Despite the tumor’s origin from the vestibular nerve, symptoms and signs of vestibular dysfunction do not become manifest until relatively late. These tumors have a distinct propensity to involve the sensory nerve roots; motor symptoms are uncommon. Multiplicity of intracranial schwannomas strongly suggests the presence of neurofibromatosis type 2 (NF2), a congenital inherited disorder associated with a mutation on the long arm of chromosome 22.262 The occurrence of bilateral vestibular schwannomas (see Fig. 9-35) is pathognomonic of this disorder, which has also been termed MISME (multiple inherited schwannomas, meningiomas, and ependymomas).359 In fact, labeling this disorder “neurofibromatosis” is a misnomer because neurofibromas are not a part of its constellation of abnormalities. Patients with this disorder typically present clinically in the second and third decades of life, much earlier than those with sporadic intracranial schwannoma. On gross inspection, schwannomas appear as focal wellcircumscribed globular or ovoid masses that do not infiltrate the nerve but rather arise and grow eccentrically, displacing the uninvolved portion of the cranial nerve of origin to the side. As the tumor grows and matures, it may undergo cystic degeneration or develop patchy areas of lipid accumulation. Vestibular schwannomas typically enlarge within the cerebellopontine angle cistern and may displace and compress the adjacent pons, medulla, and fourth ventricle (see Fig. 9-21). They can attain considerable size before becoming clinically evident. Obstruction of the ipsilateral cerebellopontine angle cistern may occur with formation of one or more extratumoral arachnoid cysts.361 Trigeminal schwannomas tend to occur more commonly in the parasellar

http://radiologyebook.com

210

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-33  Primitive neuroectodermal tumor in the left parietooccipital region (postcraniotomy). A, Axial T2-weighted MRI demonstrates an ovoid heterogeneously hyperintense mass in the left parietooccipital white matter with extension into the overlying convexity cortex and a single large intratumoral sharply marginated central collection of marked hyperintensity, likely representing a combination of necrosis, surgical resection cavity, and/or cyst formation (arrow). The mass compresses and displaces the atrium of the left lateral ventricle from behind. B, Following IV gadolinium, an axial T1-weighted image demonstrates irregular contrast enhancement within the tumor interstices without enhancement of the central cavity (arrow).

region of the middle cranial fossa than in the posterior fossa, although approximately 25% extend through the tentorial notch, involving both fossae. Microscopic evaluation of a schwannoma typically demonstrates a well-encapsulated tumor containing collections of spindle-shaped Schwann cells in both compact cellular (Antoni A) and less cellular and more loosely textured (Antoni B) arrangements. Histologic variations are commonly encountered in larger tumors, including intratumoral clusters of lipid-laden cells, confluent microcysts and larger cysts, and dilated vascular sinusoids surrounded by foci of hemorrhage. The diagnosis of vestibular schwannoma on CT rests on demonstration of a well-circumscribed globular or ovoid, hypodense to isodense (with respect to the adjacent pons and cerebellum) extraaxial mass in the cerebellopontine angle cistern, with its base on the posterior aspect of the petrous temporal bone in the region of the internal auditory meatus. Images displayed in a bone window may demonstrate fusiform widening of the internal auditory canal with erosion of its bony margins. However, soft tissue detail within the cisternal portion of the tumor is often obscured by beam-hardening artifacts caused by the adjacent dense cortical bone and pneumatized air cells of the petrous pyramid. Tumors up to about 1 cm in diameter typically demonstrate homogeneous density and homogeneously strong contrast enhancement, but larger tumors often demonstrate heterogeneity of both density and contrast enhancement because of the presence of intratumoral cystic degeneration, xanthomatous change, or intermixed areas of hypercellularity and hypocellularity.66 On MRI, smaller tumors can be more reliably detected, especially within the internal auditory canal, because of the lack of bone-induced

artifact and the multidimensional capability of this modality. The transition zone between the central (oligodendroglial) and peripheral (Schwann cell) portions of the eighth (vestibulocochlear) nerve, the presumed site of origin of schwannomas of this nerve, occurs within the canal. Tumors less than 5 mm in diameter can be reliably identified on thin-section T1-weighted MRI, appearing as homogeneously mildly hypointense or isointense (to adjacent brain) ovoid or tubular intracanalicular masses with intense homogeneous contrast enhancement2,326,361 (Fig. 9-34). On T2-weighted pulsing sequences, small vestibular schwannomas appear mildly to markedly hyperintense and may be obscured by the similarity in signal intensity to that of the surrounding CSF. As these tumors enlarge, the intratumoral degenerative changes previously described cause increasing heterogeneity of signal within the main tumor mass in the cerebellopontine angle cistern. On axial images the tumor often has a comma-like shape with a globular cisternal mass medially and a short tapered fusiform extension laterally into the internal auditory canal. Contrast enhancement is seen in nearly all schwannomas and may be homogeneous (two thirds of cases) or heterogeneous (Fig. 9-35; see Figs. 9-21 and 9-34). Larger tumors compress and displace the adjacent anterolateral margin of the cerebellar hemisphere posteromedially; displace, compress, and rotate the pons and medulla; and narrow the fourth ventricle, elongating it in the anteroposterior direction (see Figs. 9-21 and 9-35). Peritumoral vasogenic edema is frequently observed, and (as noted previously) associated arachnoid cysts are visualized in 5% to 10% of cases.260,361 Vestibular schwannomas represent 80% to 85% of cerebellopontine angle masses. The major differential diagnostic considerations include

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

FIG 9-34  Vestibular schwannoma. Axial T1-weighted image of the posterior fossa and skull base following IV gadolinium. A small intensely enhancing intracanalicular ovoid mass protrudes slightly medially into the cerebellopontine angle cistern on the patient’s right (arrow).

211

meningioma and epidermoid tumor. Meningioma involving the dura of the posterior margin of the petrous temporal bone can project posteriorly into the cerebellopontine angle and simulate a vestibular schwannoma (Fig. 9-36). Such lesions represent about 10% of cerebellopontine angle masses. Although meningiomas are typically more dense on CT, these two neoplasms may have identical signal intensity characteristics on MRI. Intense homogeneous or slightly heterogeneous contrast enhancement is another shared characteristic (see Fig. 9-36B).326,361 However, meningiomas are usually situated eccentric to the internal auditory canal and form a more obtuse angle with the petrous ridge than schwannomas.208 Extension of meningioma into the internal auditory canal is uncommon but not unknown, and rarely schwannoma may simulate the appearance of a dural tail on postcontrast images.293 Epidermoid may resemble vestibular schwannoma on noncontrast CT but typically appears more lobulated and more hypodense and tends to insinuate more extensively into the subarachnoid spaces adjacent to the pons without causing major compression of the pons or fourth ventricle. On both CT and MRI, epidermoid does not demonstrate contrast enhancement. On MRI, larger epidermoids appear more homogeneous than schwannomas of similar size and may mimic CSF in intensity with high T2 signal and low T1 signal326,361,384; differentiation from arachnoid cyst is made by the demonstration on DWI of restricted diffusion in epidermoid.4

Neurofibroma.  Neurofibromas are well-demarcated infiltrative

FIG 9-35  Bilateral vestibular schwannomas and right trigeminal schwannoma in a patient with NF2. Coronal postgadolinium T1-weighted MRI demonstrates three intensely contrast-enhancing lesions in the posterior cranial fossa. The large left-sided lobulated vestibular schwannoma fills the left cerebellopontine angle cistern (white arrow); compresses, indents, and displaces the left lateral margin of the pons inward; and extends laterally into the widened left internal auditory canal. A tiny punctate focus of contrast enhancement in the right petrous temporal region (arrowhead) represents a small intracanalicular vestibular schwannoma. A third focus of contrast enhancement abuts and indents the right lateral margin of the pons (black arrow); this is a trigeminal (fifth cranial nerve) schwannoma.

intraneural or diffusely infiltrative extraneural benign tumors that consist of a mixture of neoplastic Schwann cells, perineurial cells, and fibroblasts in a collagenous and mucoid matrix. They are histologically benign and considered WHO grade I. With only extremely rare exceptions, neurofibromas do not arise within the intracranial portions of the cranial nerves. However, they can arise peripherally and extend retrogradely into the cranial vault, notably in the facial nerve (CN VII) and the ophthalmic division of the trigeminal nerve (CN V1). The tumors exhibiting this retrograde intracranial extension are usually plexiform neurofibromas that involve multiple nerve fascicles or trunks with fusiform multinodular enlargement often described as ropelike. On noncontrast CT, they appear isodense to adjacent muscle, and on MRI they are isointense to adjacent muscle on T1-weighted images and hyperintense on T2-weighted images. Contrast enhancement is variable, with areas of moderate to strong enhancement after IV administration of gadolinium. Multiple neurofibromas are typically associated with NF1 (von Recklinghausen’s disease), an autosomal dominant congenital disorder associated with a gene defect on the long arm of chromosome 17. Approximately 50% of patients report no family history and are presumed to have new spontaneous germline mutations. In addition to multiple neurofibromas of the spinal and peripheral nerves, NF1 also includes optic nerve and chiasm gliomas (see earlier); foci of “hamartomatous” change of uncertain etiology in the region of the basal ganglia and in the optic tracts and radiations, cerebral peduncles, and brainstem; astrocytomas of the spinal cord; multiple café au lait spots; axillary and inguinal freckling; malignant peripheral nerve sheath tumors; iris hamartomas (Lisch nodules); and dysplastic osseous lesions (sphenoid wing dysplasia, pencil-like hypoplasia of the metaphyses of long bones).17,262 Patients with NF1 usually present clinically very early in life, whereas those with sporadic solitary neurofibroma involving a cranial nerve may present in later youth and adult life. On noncontrast CT, plexiform neurofibroma appears as a poorly circumscribed irregular widening of the involved nerve root, which is isodense with adjacent muscles. Erosion of adjacent bone with widening of bony foramina may be evident. MRI demonstrates the widened

http://radiologyebook.com

212

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-36  Meningioma arising from the posterior aspect of the left temporal bone. A, Axial T2-weighted MRI demonstrates a large rounded extraaxial tumor mass in the right side of the posterior fossa with its base on the posterior dura of the left temporal bone. The mass is slightly hyperintense relative to gray matter and compresses and displaces the right cerebellar hemisphere posteriorly and medially. A narrow cleft of CSF (arrow) separates the medial margin of the tumor from the adjacent compressed cerebellum and pons. Note that the internal auditory canals bilaterally are filled with CSF. B, Following IV gadolinium, an axial T1-weighted image demonstrates intense homogeneous contrast enhancement of this large dural-based tumor (arrow).

nerve, which is isointense to adjacent muscle on T1-weighted images and hyperintense on T2-weighted images. Contrast enhancement after IV administration of gadolinium is the rule, varying from moderate to intense.24

Malignant Peripheral Nerve Sheath Tumors.  Malignant peripheral nerve sheath tumors are rare tumors that appear to arise de novo in the cranial nerves, the fifth cranial nerve being more frequently involved than any other. Elsewhere in the body, almost two thirds of reported cases of malignant peripheral nerve sheath tumors have arisen from preexisting neurofibromas, often in the setting of NF1. Approximately 50% occur in association with NF1. The incidence of malignant progression in patients with plexiform neurofibroma is approximately 5%. Imaging findings include irregularity and loss of clear tumor margins together with inhomogeneous contrast enhancement. Evidence of diffuse invasion of the skull base may also be present.

Tumors of the Meninges A wide variety of primary tumors can arise from and develop within the meninges. By far the most common is meningioma. Additionally, nonmeningothelial mesenchymal tumors—both benign (lipoma, chondroma, benign fibrous histiocytoma) and malignant (chondrosarcoma, osteosarcoma)—can originate in the meninges. Also, tumors of melanocytic origin can in rare cases arise within the intracranial meninges. Finally, although the histogenesis is uncertain, neuropathologists usually assign hemangioblastomas to this category.

Meningioma.  Meningiomas are solid, well-circumscribed, highly cellular, slow-growing tumors that are usually benign histologically (WHO grade I). These sometimes spherical, sometimes lobulated, and sometimes flat and broad-based tumors are composed of neoplastic meningothelial cells originating from the arachnoid layer of the meninges, with a broad attachment to the adjacent dura. Most commonly they project inward from the dura, indenting and compressing

the underlying brain, causing neurologic symptoms and signs through compression of the adjacent cortex. Meningiomas are the most common primary extracerebral tumors of the CNS, accounting for approximately 20% of primary brain tumors. They occur mainly in middle and old age, with a peak incidence in the fifth through seventh decades of life; these tumors can, however, be found in all age groups. Meningiomas exhibit a strong sex predilection, occurring preponderantly in females, with a female-tomale ratio of at least 2 : 1. However, meningiomas associated with hereditary tumor syndromes such as NF2 generally occur in younger patients and do not demonstrate a gender predilection.210,331 Common sites of origin are in the frontal and parietal convexities and the parasagittal region (in close association with the falx cerebri); tumors in these locations constitute about 50% of all intracranial meningiomas. Meningiomas arising in relation to the sphenoid wings, olfactory grooves, sylvian fissures, and parasellar regions represent about 35%.124 Less than 10% arise below the level of the tentorium, mainly from the meninges overlying the clivus and petrous pyramid and the leaves and free edge of the tentorium. Multiple meningiomas are reported in 6% to 9% of cases,341 occurring as a component of NF2 and less commonly sporadically.262,336 Although most meningiomas are slow-growing and wellencapsulated globular masses, variations in shape and histology are not unusual. Invasion of the dura and encasement or invasion of the nearby dural venous sinuses are common. Tumors arising in relation to the sphenoid wing are frequently flat (en plaque) and tend to invade through both layers of the cranial dura into the adjacent bone, provoking a notable bony reaction with thickening and sclerosis (meningioma bone).124 Such bony changes are a source of considerable controversy; many clinical neuroscientists regard bony sclerosis and thickening as highly indicative of tumor infiltration into the haversian canals of the calvaria, whereas others state that such hyperostosis can also occur without histologic evidence of bone infiltration by tumor. Meningiomas arising adjacent to the cribriform plate and planum sphenoidale

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

213

B FIG 9-37  Sphenoid wing meningioma with underlying hyperostosis (“meningioma bone”). A, Axial postcontrast CT image demonstrates intense homogeneous contrast enhancement of a large round extracerebral tumor with its base on the right greater sphenoid wing (arrow). On its inner aspect the tumor is surrounded by a thick band of hypodensity representing edema of the adjacent cerebral cortex and white matter. There is a marked shift of the midline vessels to the left. B, On an axial CT image displayed in a bone window setting, the underlying greater sphenoid wing and squamous portion of the temporal bone demonstrate heterogeneous bony thickening (arrow).

and those occurring over the high cerebral convexities are more typically rounded and invaginate the underlying brain; however, they can also show a propensity for stimulating adjacent bony thickening and sclerosis (Fig. 9-37). Meningiomas are typically benign, and several different histologic types of benign meningioma have been described.299,341 A small but significant minority (up to 10%) of meningiomas are identified as atypical (WHO grade II) on the basis of increased mitotic activity and high nucleus-to-cytoplasm ratio. Anaplastic (malignant) meningiomas exhibit histologic features of atypical meningiomas as well as frank malignancy and often invade the adjacent cerebral parenchyma. These are classified as WHO grade III. Atypical and malignant meningiomas occur more frequently in the high cerebral convexities and the falx. Neuroimaging features such as heterogeneous appearance and enhancement, marked perilesional edema, irregular cerebral surface, mushrooming on the outer edge of the lesion, and bone destruction are more suggestive of atypical/malignant meningiomas. However, these conventional CT and MRI features have been found to be unreliable in differentiating between histologic meningioma grades.265 DWI, perfusion MR (permeability), and PET imaging have proven useful in meningioma grading, as discussed later in this chapter under “Advanced Neuroimaging Techniques.” Most meningiomas (up to 90%) are typical (benign) meningiomas and demonstrate characteristic features including homogeneous solid enhancement, but foci of necrosis and scarring (probably secondary  to ischemia), microcystic change, or areas of heavy lipid storage (lipomatous/lipoblastic meningiomas) (see Fig. 9-49) can be identified

in approximately 5% to 15% of excised tumors regardless of grade.98,234,342 Approximately three fourths of meningiomas appear on noncontrast CT as sharply circumscribed, rounded, and sometimes lobulated homogeneous masses of slightly increased density (40-50 HU) compared with adjacent brain. The hyperdensity is related to the dense cellularity of these tumors and does not reflect psammomatous change within the tumor. Meningioma typically has a broad base on a dural surface. The invaginating mass displaces and compresses the underlying brain and causes flattening of the adjacent cerebral cortex (see Fig. 9-37A). Inward buckling of the white matter can be identified in about 40% of meningiomas.124 A hypodense cleft (representing trapped CSF and occasionally blood vessels [see Fig. 9-1A]) may be insinuated between the invaginating mass and the invaginated cerebral parenchyma; cystic changes in the arachnoid may also be identified adjacent to the tumor margins. About 10% of meningiomas appear isodense with the adjacent brain on noncontrast CT, and an occasional tumor contains areas of hypodensity that correlate pathologically with foci of ischemic scar, microcysts, or lipoblastic change.37,98,234,273,342 Vasogenic edema of the adjacent cerebral white matter is identified in 50% to 75% of meningiomas.8 The mechanism responsible for the occurrence of peritumoral intracerebral edema is uncertain, and there is no correlation between tumor size and extent of edema. Significant edema is found more commonly in patients with lateral  convexity tumors than in those with tumors in other sites. Edema may be secondary to compressive ischemia, venous stasis, or aggressive growth and parasitization of pial vessels. The occurrence of

http://radiologyebook.com

214

PART II  CT and MR Imaging of the Whole Body

intracerebral edema appears to correlate with a poorer prognosis, a correlation that may be related to reduced resectability. Hyperostosis and thickening of adjacent bone in sphenoid wing, skull base, and high convexity meningiomas, as described previously, can be clearly delineated on CT if high center and wide window settings are used (see Fig. 9-37B).273 After IV administration of contrast medium, meningiomas typically display striking homogeneous contrast enhancement (increase of ≥40-50 HU) (Fig. 9-38; see Fig. 9-37A). Even the occasional meningioma that is isodense with adjacent cerebral cortex exhibits this intense opacification. However, meningiomas that are predominantly microcystic may not enhance strongly or uniformly. In 5%  to 15% of cases, focal areas of nonenhancement can be identified within the tumor mass. These areas have been correlated pathologically with regions of necrosis, old hemorrhage, scarring, cystic degeneration, and lipoblastic change; they usually do not interfere with the accuracy of diagnosis. Occasionally, however, the heterogeneity of contrast enhancement may be sufficient to raise the differential of glioblastoma.273,342 In a multiinstitutional prospective study, the sensitivity of postcontrast CT for detection of an intracranial mass in patients with meningioma was 96%; 10% of the masses were not detected on noncontrast scans. The differential diagnosis of a hyperdense or isodense peripherally situated extracerebral mass with homogeneous contrast enhancement includes metastatic malignancy (particularly carcinoma of the lung and melanoma), acute leukemia, systemic lymphoma, and (at the skull base) schwannoma of a cranial nerve.119,314,341 Criteria for differentiation include lesion location and configuration, definition of margins, and presence or absence of associated bony changes (erosion or hyperostosis). On noncontrast MRI, the majority of meningiomas present a homogeneous appearance similar to that seen on CT. Tumor signal intensity on T1-weighted images tends to approximate that of the adjacent cerebral cortex in about 50% of cases and to be slightly hypointense to cortex in 50%. On T2-weighted images, about 50% are mildly hyperintense relative to adjacent gray matter (Fig. 9-39A) and 50% are isointense to the cortex (see Fig. 9-36A).98 On T2-weighted images, comparison of signal intensity of tumor with that of cortex may have histologic correlation; in one reported series, hyperintensity correlated strongly with either syncytial or angioblastic types of meningioma, whereas fibroblastic and transitional meningiomas failed to demonstrate hyperintensity. A minority of meningiomas appear heterogeneous in signal pattern on both T1- and T2-weighted images because of the presence of intratumoral lipoblastic or cystic changes (see Fig. 9-39A and B), calcifications, or prominent vessels.98 MRI enables more accurate localization and evaluation of these extracerebral tumors than CT. This superiority is due to visualization of the mass in all three dimensions and the lack of beam-hardening artifact at the base of the brain with MRI. In approximately two thirds of cases, the interface between the inner margin of the extracerebral tumor mass and the invaginated or displaced cortex can be recognized on MRI from the presence of a cleft of CSF (see Fig. 9-36A) or interposition of vascular flow voids of the displaced arteries and veins on the pial surface of the brain (see Fig. 9-1A). Also, invasion of adjacent brain by meningiomas arising from the tentorium, falx, and dura of the lateral wall of the cavernous sinus can be recognized as a breech in the hypointense dural rim at the tumor margin.15 Finally, arterial encasement and partial dural venous sinus invasion are more readily and accurately depicted by the contrast between the flow void and the tumor tissue. Specifically, parasagittal meningiomas of sufficient size may compress (see Fig. 9-38) or invade the adjacent superior sagittal sinus; interruption of the sinus can be identified on coronal postcon-

trast CT or noncontrast MRI and is an important preoperative finding for which the radiologist must search. As on CT, meningiomas usually demonstrate rapid and pronounced contrast enhancement after IV administration of paramagnetic contrast agent, and the strong (often striking) homogeneous contrast enhancement seen in most meningiomas enables their accurate detection and localization (Fig. 9-40B; see Figs. 9-1B, 9-36B, 9-38B and C, and 9-39C).326 A thickened tapered extension of contrast-enhancing dura is commonly identified at the margins of the tumor (see Fig. 9-36C). This dural tail may indicate the presence of tumor infiltration or may be due to a proliferative dural reaction.266 Because a major prognostic factor in the recurrence of meningiomas after surgery is the extent of tumor resection, careful assessment must be made of preoperative postcontrast MRI for the presence of a dural tail.126 The arachnoid contributes embryologically to formation of the choroid plexuses, so intraventricular meningiomas may arise within the choroid plexus of the lateral ventricle (80%) or the third ventricle (15%) and locally expand the ventricle to conform to the size and shape of the tumor (see Fig. 9-40).257,356 These tumors share the imaging characteristics of extracerebral meningiomas as previously described. The normally compact choroid glomus calcifications may be fragmented and spread apart by the expanding tumor mass. Differentiation from choroid plexus papilloma is based on (1) patient age (choroid plexus papillomas of the lateral ventricles occur mainly in infants and very young children), (2) tumor surface characteristics (papillomas tend to have a nodular irregular margin, whereas meningiomas tend to be smoothly rounded), and (3) size and shape of the cerebral ventricles (papillomas are often associated with diffuse enlargement of all ventricles, whereas meningiomas cause only focal enlargement).

Melanocytic Tumors.  A wide spectrum of tumors of melanocytic origin, varying from benign and diffuse (melanocytosis) to benign and well circumscribed (melanocytoma) to malignant (primary malignant melanoma), can originate within the meninges and CNS. They are thought to arise from leptomeningeal melanocytes derived from the neural crest. All these tumors are uncommon lesions, but the benign forms are less rare than the primary malignancies. Melanocytomas of the meninges, which represent less than 0.1% of all brain tumors, present as solitary solid, round, or flat extraaxial masses occurring mainly near the foramen magnum or in the region of Meckel’s cave. Histologically they demonstrate monomorphic cells with variable melanin content in their cytoplasm and a low mitotic rate. Immunohistochemical studies show most tumors reacting to antimelanosomal antibody. Presenting symptoms are those related to local compression of the underlying brain or obstruction of CSF flow. They can be found in patients of all age groups, with a peak incidence in the fifth decade of life. Females are affected more commonly than males, with a reported 2 : 1 ratio. On MRI, melanocytomas of the meninges appear as well-circumscribed extraaxial masses that are isointense to hyperintense (depending on the quantity of melanin in the tumor cells) on T1-weighted images and hypointense to gray matter on T2-weighted images. Like the majority of meningeal tumors, they exhibit homogeneous contrast enhancement.225,270,397

Hemangioblastoma.  Hemangioblastomas are benign (WHO grade I) vascular neoplasms of uncertain origin that consist of abundant endothelial pericytes forming a rich capillary network intermixed with vacuolated (lipid-containing) stromal cells. Although relatively uncommon overall (1%-2.5% of all primary CNS tumors), hemangioblastoma is the most common primary intraaxial posterior fossa tumor in adults.337,341 The great majority (80%-85%) of CNS hemangioblastomas occur in the cerebellum, 2% to 3% in the medulla, and only 1%

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

C FIG 9-38  Multiple meningiomas. Transcalvarial tumor extension with subgaleal mass and dural tail. Axial T2-weighted (A) and postgadolinium T1-weighted (B) images demonstrate a large right frontoparietal convexity extracerebral tumor that is homogeneously mildly hyperintense to gray matter on T2-weighting and intensely and homogeneously enhances following IV gadolinium. This aggressive tumor extends through the calvaria (black arrow in B) and presents as a broad-based mass under the scalp (white arrows in A and B). Note also the presence of a second similar tumor arising in the high posterior falx cerebri on its left lateral aspect and displacing the superior sagittal sinus posterosuperiorly and to the right of the midline. C, Coronal T1-weighted postgadolinium image demonstrates intense homogeneous contrast enhancement of both the convexity and the parasagittal tumors and also of a third large similarly enhancing tumor in the left cerebellopontine angle. Note the “dural tail” of tumor extension over the right frontal convexity (arrow).

http://radiologyebook.com

215

216

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-39  Lipoblastic meningioma. Axial T1-weighted (A) and T2-weighted (B) MRIs demonstrate a curvilinear focus of T1 hyperintensity in the far right lateral temporooccipital convexity (black and white arrows). The remainder of this dural-based extracerebral tumor cannot be clearly distinguished from the adjacent compressed brain parenchyma on the T1-weighted noncontrast image. The lesion (arrow) is also barely discernible on the fast T2-weighted image (B). Following IV gadolinium (C), the tumor, which is based on the most lateral aspect of the right leaf of the tentorium, enhances intensely on T1-weighted imaging, except for the focus of lipid storage (arrow).

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

217

B FIG 9-40  Intraventricular meningioma. A, A large round mildly hyperintense (to gray matter) tumor fills and distends the atrium of the right lateral ventricle (arrow) on this axial T2-weighted MRI. Note dilatation of the anterior right temporal horn proximal to the obstruction caused by this mass. Edema of the surrounding white matter contributes to the right-sided mass effect, which causes compression of the right frontal horn and marked shift of the septum pellucidum to the left of midline. B, An axial T1-weighted postgadolinium image demonstrates intense homogeneous contrast enhancement of this large multilobulated intraventricular meningioma (arrow).

to 1.5% above the tentorium.154,341 The remainder (10%-15%) arise within the spinal cord.154 Between 75% and 85% of hemangioblastomas are sporadic in occurrence and solitary. They typically present clinically in young adults, with peak incidence in the third to fifth decades of life, and are rare in children. The natural history is that of a slow-growing mass in the cerebellum that is frequently associated with a cyst. Approximately 60% of these tumors are cystic at the time of clinical presentation. Symptoms are usually related to obstruction of CSF flow in the fourth ventricle by the solid tumor or the associated cyst. Hemorrhage is uncommon in hemangioblastoma despite the prominent vascularity, and necrosis is rare. The advent of microsurgical excision techniques has significantly improved the prognosis in patients with sporadically occurring hemangioblastomas; mortality is low, and permanent neurologic deficits are rare.154 On gross inspection, hemangioblastomas are well-demarcated, highly vascularized, small (5 to 10 mm), rounded solid lesions located peripherally within the cerebellum, brainstem, or spinal cord, usually abutting a pial surface106 (Fig. 9-41). The most striking lesions are those associated with well-circumscribed cysts in which the solid tumor is a mural nodule.341 Like the cyst associated with juvenile PA of the cerebellum with a mural nodule, the cyst associated with a hemangioblastoma is essentially extratumoral, lacking any tumor tissue in its margin except for the eccentrically located solid nodule. The cyst wall is composed of compressed brain parenchyma or reactive gliosis.154 Approximately 25% of hemangioblastomas are associated with von Hippel-Lindau disease, a phakomatosis that is inherited as an autosomal dominant trait caused by a germline mutation of a suppressor gene

on the 3p chromosome. Lindau originally described this disease complex in 1926.222 It is characterized by development of capillary hemangioblastomas in the CNS in association with identical tumors in the retina (originally described by von Hippel in 1906); clear cell renal carcinomas, pheochromocytomas, and pancreatic, renal, and hepatic cysts are also frequently found in patients with the disease.387 Multiplicity of hemangioblastomas of the cerebellum, brainstem, and spinal cord is common (10%-40% of cases) in patients with von Hippel-Lindau disease,106 and multiple tumors are found almost exclusively in patients with the disease (Fig. 9-42). Patients with von HippelLindau disease become symptomatic at an earlier age (mean age at diagnosis, 29 years) than those with solitary sporadically occurring tumors. Despite improved surgical techniques, the most common cause of death in patients with von Hippel-Lindau disease is hemangioblastoma.179,272 Noncontrast CT usually demonstrates a large, sharply marginated, hypodense, cystic cerebellar hemispheric mass with a slightly more dense rounded mural nodule. There is typically little or no peritumoral edema, and the solid lesion or mural nodule may not be distinguishable from surrounding normal cerebellar parenchyma, particularly if the lesion is located inferiorly within the posterior fossa where it may be masked by dense adjacent bone and beam-hardening artifact. Calcifications have not been detected in these tumors. After IV administration of a nonionic iodinated contrast agent, the solid tumor or mural nodule enhances homogeneously and intensely adjacent to a pial surface. Because of the intense vascularity of even the smallest solid tumor nodules, angiography may be more sensitive than CT or MRI in the detection of hemangioblastoma.179

http://radiologyebook.com

218

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-41  Hemangioblastoma of the cerebellum. A, Coronal T2-weighted MRI demonstrates a sharply marginated ovoid hyperintense mass in the superior aspect of the left cerebellar hemisphere. On the superior margin of this cystic mass, a flat nodule that is isointense to adjacent gray matter is identified (arrow). B, Following IV gadolinium, a coronal T1-weighted image demonstrates intense homogeneous contrast enhancement of the mural nodule (arrow), which is located on the superior surface of the cerebellum. The adjacent cyst is hypointense and its margins do not demonstrate contrast enhancement.

On MRI the cysts typically are sharply and smoothly marginated and homogeneously hypointense relative to adjacent brain parenchyma on T1-weighted images and hyperintense on T2-weighted images (see Fig. 9-41A).154 The solid tumors and mural nodules are usually inhomogeneous in signal pattern but predominantly  isointense to normal gray matter on T1-weighted images and slightly hyperintense on T2-weighted images. Areas of increased signal within the nodule or the cyst on T1-weighted images are occasionally noted secondary to hemorrhage. Serpentine signal voids due to enlarged vessels supplying and draining the tumor may be identified at the periphery of the mass or nodule, a finding that should strongly suggest the diagnosis in the appropriate clinical setting. As with CT, contrast enhancement of the solid tumor or mural nodule on T1-weighted MRI is characteristically homogeneous and intense15,154,211 (see Figs. 9-41B and 9-42).

Mesenchymal Tumors.  A diverse group of relatively rare tumors, mesenchymal nonmeningothelial tumors may arise within the cranial vault in the skull base, the calvaria and sutures, the meninges, or (most rarely) the cerebral parenchyma. They may be of primitive mesenchymal origin but have not differentiated along the meningothelial line. Rather, these unusual lesions exhibit fibrous, fibrohistiocytic, adipose, myoid, endothelial, chondroid, or osseous differentiation and cannot be distinguished histologically from their more common counterparts in the somatic soft tissues. These tumors vary widely in location, histology, pathologic grade, and clinical behavior. Some (e.g., lipomas and hemangiomas) are benign (WHO grade I) and mainly asymptomatic, and some (e.g., osteosarcomas and rhabdomyosarcomas) are highly malignant (WHO grade IV) and highly invasive. Tumors arising in the meninges are more common than those originating within cerebral parenchyma.

Although most of these lesions occur supratentorially in adults, many rhabdomyosarcomas occur below the tentorium in young children. Lipomas typically occur in or near the midline and are associated with malformations of the corpus callosum, whereas chondrosarcomas arise most commonly in the skull base off the midline (e.g., petroclival suture). Lipomas are not true neoplasms but instead likely represent congenital malformations. They enlarge only with somatic growth. Intracranial lipomas are usually subarachnoid in location and occur most commonly in the pericallosal sulcus. Other sites are the chiasmatic, circummesencephalic, interpeduncular, and quadrigeminal cisterns and, in the posterior fossa, the cerebellopontine angle cistern. More than 50% of pericallosal lipomas are associated with partial agenesis or dysgenesis of the adjacent corpus callosum, with the rostrum, genu, and anterior portion of the body more commonly affected than the posterior body or splenium.395 Two types of pericallosal lipomas are recognized. The tubonodular lipomas are large and rounded, are located anteriorly, and have a high incidence of associated corpus callosal dysgenesis as well as anomalies of the frontal lobes. The curvilinear type tend to be more posteriorly situated, appear long and thin, curve around the splenium, and are associated with a normal or only slightly dysgenetic corpus callosum (Fig. 9-43). These lesions are tightly adherent to the adjacent brain, and lipoma in the cerebellopontine angle cistern may envelop and compress the seventh and eighth cranial nerves where they pass through the cistern.65 Lipomas are typically asymptomatic and discovered only incidentally when the head is imaged for other reasons. Even on plain skull radiographs, the tubonodular masses are recognized by their welldemarcated homogeneous lucency in the midline anteriorly, often associated with marginal curvilinear calcification bilaterally (see 

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms Fig. 9-43). On CT the smoothly demarcated mass appears more hypodense than CSF (50-100 HU) and is often marginated laterally by nodular or curvilinear calcification (see Fig. 9-43). On MRI, lipomas appear mainly homogeneously hyperintense on both T1-weighted and fast-spin echo T2-weighted images, but areas of low signal secondary to the marginal calcification are commonly noted peripherally. Central flow voids, representing pericallosal arteries coursing through the substance of the tumor, are prominent within the tumor mass.116 Osteocartilaginous tumors occasionally arise within dural structures such as the falx. The CNS is the most common site of occurrence of extraosseous chondrosarcoma. This mesenchymally derived malignancy occurs more frequently in the skull base, notably in  the region of the intrasphenoid synchondrosis or the adjacent petroclival suture, and it may invade the overlying brain (Fig. 9-44). Whether arising from dura or from skull base structures, chondrosarcomas manifest as soft tissue masses with focal bone destruction. Intratumoral calcifications, identified in about half of tumors, contribute to a heterogeneous appearance on MRI, with intratumoral nodules that appear hypointense to isointense to adjacent brain on T1-weighted images and hyperintense to brain on T2-weighted images. Intense contrast enhancement of the soft tissue mass is the rule248 (see Fig. 9-44B). Other mesenchymal malignancies may also rarely originate within the cranial vault but more commonly arise in extracranial structures and secondarily invade the skull and brain, notably embryonal rhabdomyosarcoma. This is the most common soft tissue sarcoma in children, and the head and neck are the most common sites of occurrence, especially the orbit and nasopharynx as well as the middle ear. The tumor is almost uniformly fatal within 2 years of presentation, despite intensive chemotherapy and radiotherapy. An extremely aggressive neoplasm, embryonal rhabdomyosarcoma commonly invades through the skull base into the adjacent brain. Evidence of invasion of  the anterior skull base, cavernous sinuses, or both is identified on  MRI in 35% of embryonal rhabdomyosarcomas originating in the nasopharynx.205 Perineural spread and meningeal spread of tumor are also common in this tumor. Imaging studies demonstrate large soft tissue masses with evidence of aggressive bone destruction. On MRI, these lesions appear isointense to muscle on T1-weighted images and hyperintense to brain and muscle on T2-weighted images, findings that are nonspecific and similar to those seen in nasopharyngeal carcinoma, myeloma, and neuroblastoma. Embryonal rhabdomyosarcomas exhibit contrast enhancement, which varies in both extent and intensity.205

Hemangiopericytoma.  Hemangiopericytoma is now recognized in the WHO classification of tumors of the CNS as a distinct entity separate from meningioma. This dural-based tumor was formerly considered an “angioblastic” meningioma because of its hypervascularity. However, despite many gross pathologic, histopathologic, and imaging similarities to meningioma, the aggressive clinical behavior of hemangiopericytomas differs considerably from the majority of menin­ giomas, and it is now possible to distinguish the two entities immunohistochemically. The cell of origin is the pericyte, a perivascular cell of mesenchymal origin, and the intracranial tumors are identical histologically to their counterparts in the somatic soft tissues.146 Like meningiomas, hemangiopericytomas are dural-based extraaxial rounded masses, well circumscribed and highly cellular. They are often more lobulated in contour than the typical meningioma and are more commonly located in the occipital region near the confluence of the dural venous sinuses, to which they are often attached. They are comparatively rare, with a ratio of 1 hemangiopericytoma to 40 or 50 meningiomas. They tend to occur in younger patients, typically in their

219

40s, and the gender distribution (male-to-female ratio 1.4 : 1) differs markedly from that of meningiomas.134 Histologically, hemangiopericytomas appear highly cellular with moderate nuclear atypia and prominent mitotic activity comparable to that of anaplastic meningiomas. They are richly vascular tumors with arterial supply from both meningeal and cerebral arteries, and numerous large and branching vascular channels (“staghorn sinusoids”) with perivascular fibrosis are a prominent microscopic finding. Invasion and destruction of overlying bone are not unusual, but these tumors do not provoke a hyperostotic (“meningioma bone”) response. Infiltration of underlying brain parenchyma can also occur. Hemangiopericytomas exhibit a marked tendency to recur locally despite apparently complete surgical excision and postoperative irradiation of the former tumor bed; in one large series, the 15-year recurrence rate was 85%. They also tend to metastasize beyond the cranial cavity; in the same large series, 60% of tumors had metastasized after 15 years, mainly to bone, lung, and liver.134 The findings on diagnostic imaging tend to mimic those of meningiomas. These sharply circumscribed extraaxial masses appear hyperdense on noncontrast CT and are generally homogeneous and hypointense to white matter and isointense to gray matter on T1-weighted noncontrast MRI (Fig. 9-45A). On T2-weighted images, hemangiopericytomas typically are slightly hyperintense and more heterogeneous in appearance (see Fig. 9-45B). As noted previously, hemangiopericytomas often appear more lobulated and exhibit a higher incidence and degree of heterogeneity of attenuation (CT) and signal intensity (MRI) than meningiomas owing to the presence of multiple flow voids. Contrast enhancement of hemangiopericytomas is usually intense and homogeneous (see Fig. 9-45C). Points of differentiation from meningioma include a greater frequency of internal heterogeneity and external lobulation, a tendency toward a more narrow base of dural attachment, and absence of tumor calcification and hyperostosis of adjacent bone.61

Lymphomas and Hematopoietic Neoplasms During the last 2 decades of the 20th century, tumors and tumorlike conditions of the hematopoietic system involving the CNS increased in both frequency and clinical and societal significance. Most notable in this regard was a greater than 10-fold increase in incidence of primary CNS non-Hodgkin’s lymphoma in the United States. The vast majority of these tumors were malignant B-cell lymphomas, which carry a poor prognosis. Their site of origin is unknown because the normal brain lacks lymphatics and lymphoid tissue. The rise in incidence of these tumors was associated with the human immunodeficiency virus (HIV) epidemic; up to 10% of patients in the terminal stages of acquired immunodeficiency syndrome (AIDS) experience an Epstein-Barr virus–associated B-cell malignant cerebral lymphoma. The concomitantly growing number of non–HIV-positive immunocompromised patients who have primary cerebral lymphoma, notably those undergoing long-term immunosuppressive therapy after allograft organ transplantation, has also contributed to the higher incidence of primary CNS lymphoma.15,155,341 Lymphomas arising systemically may involve the brain and meninges secondarily. These include T-cell lymphoma, plasmacytoma, angiotropic lymphoma, and Hodgkin’s disease. In current practice, secondary involvement of the CNS by systemic lymphoma is much less common than the primary intracranial malignant B-cell variety.249 A heterogeneous group of tumors and tumorlike processes of histiocytic origin rarely involve the brain, notably Langerhans cell histiocytosis. Although leukemias in their advanced stages may involve the meninges, leading to detection of malignant cells on CSF cytology, radiologically detectable dural-based or intracerebral masses are rare.286

http://radiologyebook.com

220

PART II  CT and MR Imaging of the Whole Body

C

A

B

D

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

221

FIG 9-42  Multiple hemangioblastomas in a patient with von Hippel-Lindau disease. A, Axial T1-weighted MRI following IV gadolinium demonstrates a sharply defined, round, homogeneously contrast-enhancing nodule on the dorsal aspect of the lower medulla (arrow). Note the absence of a surrounding cyst. B, Axial T1-weighted postgadolinium image at the level of the suprasellar cistern. A large lobulated homogeneously intensely contrasting hypothalamic tumor protrudes inferiorly, filling and distending the suprasellar cistern (arrow) and insinuating into the interpeduncular cistern, with spreading apart of the cerebral peduncles. Intratumoral hypointense linear and round foci likely represent flow voids within the rich tumor vasculature. This is a most unusual location for a hemangioblastoma. Also, the comparatively large size of this mass without an associated cyst is unusual. C, Axial T1-weighted postgadolinium image of the orbits and anterior brain obtained with fat saturation. A relatively large contrastenhancing nodule is identified within the lateral aspect of the left globe (arrow); a partially intranodular cyst is visualized anteriorly. D, Coronal T1-weighted postgadolinium image of the thoracic spine demonstrates an elongated nodular homogeneously enhancing tumor located eccentrically within the lower thoracic spinal canal on the right and protruding into and widening the right lateral subarachnoid space (arrow); the differential diagnosis of this lesion would include meningioma, schwannoma, and leptomeningeal metastasis.

FIG 9-43  Pericallosal lipoma and partial agenesis of the corpus callosum. Axial noncontrast CT image demonstrates widely separated and parallel bodies of the lateral ventricles with an intervening elevated and dilated third ventricle (black arrow); these findings are consistent with partial agenesis of the anterior portion of the corpus callosum. In the midline more posteriorly, a markedly hypodense rhomboid-shaped curvilinear lipoma (white arrow) with two internal foci of calcification separates the lateral ventricles.

Primary CNS Lymphoma.  Primary malignant lymphomas of the CNS are extranodal tumors arising in the CNS in patients with no obvious lymphoma outside the nervous system at the time of initial diagnosis. Historically this entity constituted about 1% of all primary tumors of the CNS; by the early 1990s, however, this proportion had increased to 6.6%,251 mainly as a consequence of the AIDS epidemic. The incidence of this neoplasm in the AIDS population (4.7 per 1000 person-years) is approximately 3600 times that in the general population.89 Approximately 10% of patients with AIDS experience primary CNS lymphoma, mainly during the late stages of their illness.58 In

organ transplant recipients who are immunosuppressed, the incidence of primary CNS lymphoma exceeds 20%.249,295 Primary CNS lymphoma affects all ages, but the peak incidence in immunocompetent individuals is in the sixth and seventh decades. In patients who have AIDS or have received organ transplants and are receiving immunosuppressive therapy, the median age of onset is in the late 30s. This tumor shows a definite gender predilection, with a 3 : 2 male-to-female ratio in immunocompetent patients and a 9 : 1 male-to-female ratio in patients with AIDS. The presenting symptoms and signs are nonspecific; focal neurologic deficits are most common, but seizures, signs of increased intracranial pressure, neuropsychological symptoms, and visual disturbances are also frequent. Primary CNS lymphoma carries a guarded prognosis in immunocompetent patients and a poor prognosis in immunocompromised and immunosuppressed patients. In immunocompetent patients, the initial response rate to radiotherapy, chemotherapy, or both is 85%, but the survival rate is 40% to 70% at 2 years and 25% to 45% at 5 years; in patients with AIDS, the median survival after multimodal therapy is 13.5 months.249 On gross inspection, primary CNS lymphoma presents as one or more (usually multiple) firm or friable centrally located deep-seated masses with variable demarcation from adjacent cerebral parenchyma. The tumors are often located near the ventricles and may infiltrate through and along the ependymal ventricular walls69 (Fig. 9-46). The corpus callosum and basal ganglia are common sites of involvement. Intratumoral hemorrhage and necrosis are uncommon, and peritumoral edema is usually mild in immunocompetent patients, but in immunocompromised patients lesion growth is more rapid, hemorrhage and necrosis are much more common, and edema of the  surrounding white matter is often extensive.69 Involvement of the leptomeninges is a common but late finding. Histologically, 98% of primary CNS lymphomas are high-grade B-cell tumors that characteristically demonstrate an angiocentric infiltration pattern in which collars of small neoplastic lymphocytes  surround and infiltrate the walls of the small penetrating vessels of   the brain and the perivascular (Virchow-Robin) extensions of the subarachnoid space. From these perivascular foci, tumor cells invade the adjacent neural parenchyma and coalesce, forming diffuse masses. In AIDS-associated cases, the diffuse masses tend to be multifocal with intervening areas of necrosis.69,249 CT demonstrates the tumor masses involving the deep gray matter structures, periventricular white matter, and corpus callosum; they appear homogeneously isodense to moderately hyperdense (reflecting dense packing of small tumor cells with a high nucleus-to-cytoplasm ratio), have relatively ill-defined margins, and demonstrate diffuse strong homogeneous contrast enhancement (see Fig. 9-46).15,164,368

http://radiologyebook.com

222

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-44  Chondrosarcomas of the skull base. A, Axial noncontrast CT image displayed in a bone window demonstrates bilateral irregularly marginated areas of bone destruction in the region of the intrasphenoid synchondrosis with multiple intratumoral hyperdensities representing foci of calcification and/or bony sequestra. The bony destruction involves the right lateral margin of the clivus (arrow) and the right petrous apex. B, In another patient, an axial CT image following IV contrast displayed in a soft tissue window demonstrates more extensive bone destruction involving the entire body of the sphenoid bone, right ethmoid sinus, and right orbit, with foci of intratumoral calcification and/or bony sequestra. The soft tissue components of this tumor demonstrate heterogeneous contrast enhancement (arrow) with multiple ill-defined areas of nonenhancement.

Peritumoral edema is less extensive than that seen in association with primary gliomas and metastases of similar size.11 Central necrosis with peripheral ringlike contrast enhancement is uncommon in immunocompetent patients but is noted more frequently and involves wider areas with larger lesions in immunocompromised patients.164 Extension along the ventricular walls with ependymal contrast enhancement (see Fig. 9-46C) and leptomeningeal tumor spread with hyperdensity in and contrast enhancement of the cisterns and sulci are common late findings.155 On MRI the deeply seated tumor nodules exhibit a variable signal intensity pattern. Most often the tumors are homogeneously isointense to gray matter on both T1- and T2-weighted images, findings similar to those for other small cell hypercellular tumors.333 However, some lesions may exhibit marked T2 hyperintensity. Almost all primary CNS lymphomas demonstrate intense contrast enhancement after IV administration of gadolinium. In immunocompromised patients, this enhancement is often ringlike (reflecting central necrosis) and associated with extensive peritumoral edema, and the clinical and radiologic presentation can simulate that of toxoplasmosis, a common occurrence in patients with AIDS.91 The differential diagnosis favors toxoplasmosis if there is associated hemorrhage, whereas the presence of subependymal extension with contrast enhancement of the ventricular walls favors lymphoma.15 In such circumstances a short trial of anti-Toxoplasma therapy followed by evaluation of the clinical and imaging response may be sufficient to avoid invasive biopsy. Leptomeningeal seeding occurs in up to 60% of patients with primary CNS lymphoma, generally late in the course of the disease.

Although malignant cells may be readily detected on CSF cytology, detection of subarachnoid tumor spread on imaging studies has been problematic. Imaging verification of leptomeningeal neoplasm, which is visible as parallel linear streaklike contrast enhancement, may be more easily and reliably detected on FLAIR and T1-weighted postgadolinium MRI than on postcontrast CT.155,199,287,381 An unusual form of presentation of primary CNS lymphoma simulates diffuse white matter disease with patchy and diffuse infiltration of deep white matter as well as deep gray matter. Poorly marginated areas of T2 hyperintensity are seen scattered through the deep cerebral white and gray matter and in the pons. The pattern is similar to that of gliomatosis cerebri. This entity has been termed lymphomatosis cerebri. In an immunocompetent patient, the differential diagnosis of solitary or multiple deep-seated tumors that are relatively homogeneous, appear isointense to gray matter on MRI and hyperdense on CT, and demonstrate dense contrast enhancement with mild surrounding edema would include, in addition to primary CNS lymphoma, highgrade glioma (e.g., anaplastic astrocytoma), sarcoid, and intraventricular meningioma.

Secondary CNS Lymphoma.  Approximately one third of patients with systemic lymphoma demonstrate secondary brain or spinal involvement.20 The majority of secondary CNS lymphomas are T-cell lymphomas with a propensity to first involve the leptomeninges surrounding the brain and then extend into cerebral parenchyma along

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

C FIG 9-45  Hemangiopericytoma in the floor of the posterior cranial fossa on the right. Axial T1-weighted (A) and T2-weighted (B) MRIs. A large, well-circumscribed, lobulated, homogeneously hypointense (T1weighted)/hyperintense (T2-weighted) extraaxial mass (arrow) deeply invaginates and displaces the inferior aspect of the right cerebellar hemisphere and vermis medially across the midline and indents the right posterolateral aspect of the medulla. C, Following IV gadolinium, intense homogeneous contrast enhancement is noted in this lobulated tumor (arrow).

http://radiologyebook.com

223

224

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-46  Primary CNS lymphoma in a patient with AIDS. A, Axial CT image demonstrates bandlike hyperdensity in the subependymal white matter at the margins of the frontal horns of both lateral ventricles (arrows) with surrounding peritumoral edema. B, Following IV administration of contrast medium, diffuse enhancement of the periventricular tumor is noted (arrows). C, Axial postcontrast CT image at a slightly higher level demonstrates symmetric subependymal tumor enhancement with surrounding white matter edema along the margins of the frontal horns (arrows), the atria and occipital horns of both lateral ventricles, and the lateral margins of the third ventricle.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms the Virchow-Robin spaces in the periphery of the cerebral hemispheres. Both focal and diffuse thickening of the arachnoid occurs, and infiltration also extends into the overlying dura. As in the primary CNS lymphomas, multicentricity is common, but the masses are predominantly peripherally rather than deeply situated. Thick plaques of tumor on the brain surface that are isointense to the adjacent brain on T1and T2-weighted images and demonstrate homogeneous contrast enhancement (Fig. 9-47) may simulate the appearance of meningioma; in general the tumor masses of lymphoma are more numerous and more extensive than those of meningioma and often extend into the subarachnoid space and underlying brain parenchyma.15

Granulocytic Sarcoma (Chloroma).  Granulocytic sarcoma (formerly known as chloroma) is a focal collection of leukemic cells that form a solid mass projecting inward from the meninges or within the brain parenchyma contiguous with the ventricular wall ependyma in patients with systemic leukemia, typically acute or chronic myelogenous leukemia. In approximately one third of cases, the tumor may actually precede development of the systemic disease. On CT the lesion appears hyperdense or isodense relative to brain parenchyma, with unsharp margins. On MRI it demonstrates heterogeneous isointensity or hypointensity relative to gray matter on T1-weighted images and isointensity to hyperintensity on T2-weighted images. Homogeneous contrast enhancement of the tumor is the rule121,130,286 (Fig. 9-48).

Germ Cell Tumors Tumors arising from primordial germ cells represent only 0.3% to 0.5% of all primary intracranial neoplasms, although in Asia the figure is 2%. Germ cell tumors represent about 3% of primary pediatric intracranial neoplasms, except in Asia, where the proportion has been reported as 10% to 15%.245 These tumors are the intracranial morphologic homologues of germinal neoplasms arising in the gonads and in other sites. They present mainly in children and adolescents, with an age peak of 10 to 12 years and a strong (2 : 1) male predilection. Approximately 90% arise in patients younger than 20 years. Intracranial germ cell tumors occur mainly in or near the midline in the vicinity of the third ventricle. The most common sites of occurrence are the pineal (65%) and suprasellar (30%) regions. Nearly two thirds (65%) of intracranial germ cell neoplasms are germinomas, the intracranial morphologic equivalent of the testicular seminoma, whereas 16% represent teratomas, 6% are endodermal sinus (yolk sac) tumors or embryonal cell carcinomas, 4% are choriocarcinomas, and the remaining 9% are mixed tumors incorporating elements of two or more germ cell types. These lesions may be multifocal with involvement of both the pineal and suprasellar regions.172 The clinical presentation depends on the location of the tumor mass. Patients with pineal region tumors typically present with headache because the tumor compresses and obstructs the aqueduct, causing a noncommunicating hydrocephalus. Compression on or invasion of the quadrigeminal plate causes a Parinaud’s syndrome with paralysis of upward gaze and convergence. Tumors in the floor of the third ventricle or the suprasellar region typically impinge on the optic chiasm (causing visual field defects) or the hypothalamic-pituitary pathways (causing diabetes insipidus, growth retardation, or precocious puberty). Examination of the serum and CSF for the presence of certain oncoproteins elaborated by germ cell tumors often provides valuable differential diagnostic information and is also useful in monitoring  the course of treatment. Elevation of placental alkaline phosphatase (PLAP) in these fluids favors the diagnosis of germinoma, whereas a rise in concentration of α-fetoprotein suggests the presence of yolk sac endoderm, as in an endodermal sinus (yolk sac) tumor or an immature

225

teratoma. Human chorionic gonadotropin is elaborated by syncytial trophoblasts, and elevation of this marker points to the presence of choriocarcinoma. There is a higher risk of intracranial and mediastinal germ cell tumors in patients with Klinefelter’s syndrome, a complex chromosomal disorder with a 47XXY karyotype. The most common cause of hypogonadism in young males, Klinefelter’s syndrome is characterized by testicular atrophy, gynecomastia, and lack of secondary sex characteristics.172 Affected individuals are also at greater risk for development of breast carcinoma. Histopathologic examination of tumor tissue is essential for treatment planning and prognosis.103 Although germinomas and teratomas often occur as pure tumor types, the nongerminomatous neoplasms are often of mixed cellular composition, and immunohistochemical studies performed on excised tissue to detect and localize the oncoproteins expressed by these tumors (mentioned in the preceding paragraph) are valuable for their characterization.

Germinoma.  Germinomas are usually solid tumors that may contain cystic foci. A pure germinoma is the most common germ cell tumor, composed of sheets or lobules of uniform cells with large nuclei and clear cytoplasm that exhibit immunohistochemical labeling for PLAP on their cell membrane surfaces. Mitoses are common, but intratumoral hemorrhage and necrosis are uncommon. These tumors are not encapsulated and tend to grow by invasion of neighboring structures. In the pineal region they surround and engulf the pineal calcifications and may invade the adjacent thalami and quadrigeminal plates.113,360 In the suprasellar region these tumors appear as rounded or lobulated masses in the floor of the third ventricle, compressing and invading the optic chiasm and surrounding the pituitary stalk; they may also extend superiorly, infiltrating the walls of the third ventricle. Seeding of tumor cells may occur into the CSF with both ependymal and subarachnoid spread of tumor. More than 90% of pineal region germinomas occur in males, whereas for tumors in the suprasellar region, the sex distribution is approximately equal.419 Pure germinomas are highly radiosensitive, and 5-year survival rates of 65% to 95% are reported. Adjuvant chemotherapy may permit reduction in radiation dosage with comparable survival rates. Extent of disease at the time of diagnosis is an important prognostic factor.172 On both CT and MRI, germinomas in the pineal and suprasellar regions appear as well-circumscribed and mostly homogeneous rounded or lobulated masses.113,360,419 Their hypercellularity explains their notable hyperdensity on CT (Fig. 9-49A) and their isointensity to gray matter and hypointensity relative to CSF on T2-weighted MRI (Fig. 9-50A and B). Intratumoral cysts are occasionally found and are easily distinguished from the surrounding solid tumor. Intratumoral hemorrhage and necrosis are rare occurrences. CT visualization of intratumoral calcification in a cluster likely signifies engulfment or displacement of the normal calcified pineal gland; widely dispersed calcifications favor a diagnosis of a primary pineal cell neoplasm. The solid parts of these tumors exhibit uniform contrast enhancement, and spread of tumor from the pineal region into the quadrigeminal plate or thalami or from the floor of the third ventricle into the optic chiasm and pituitary stalk can be clearly delineated on either CT or MRI after IV administration of contrast medium (see Figs. 9-49B and 9-50C). Differentiation of suprasellar germinoma from PA on either CT or MRI may be difficult, but clinical differences in age and presenting symptoms usually permit accurate characterization.

Teratoma.  Teratomas are the second most common germ cell tumors. They originate from multipotential germ cells that recapitulate normal somatic development by developing elements of the three embryonic

http://radiologyebook.com

226

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-47  Systemic lymphoma with secondary CNS involvement. A, An ill-defined isointense (to gray matter) mass involves the right posterior temporooccipital convexity cortex (arrow) and the adjacent white matter on this axial T1-weighted MRI. B, An axial T2-weighted image at the same level demonstrates a multilobulated right temporooccipital convexity tumor with extensive surrounding edema. Like a meningioma, this heterogeneous tumor is based on the convexity dura, but the deep extension (arrow) suggests spread into the cerebral parenchyma via the perivascular (Virchow-Robin) spaces. There is extensive edema of the adjacent forceps major white matter resulting in considerable mass effect with compression of the right sylvian fissure and atrium of the right lateral ventricle, bowing of the third ventricle across the midline, and effacement of the sulci of the right temporal and occipital lobes. C, Following IV gadolinium, an axial T1-weighted image demonstrates heterogeneous contrast enhancement in this peripheral multilobulated tumor (arrow).

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B FIG 9-48  Granulocytic sarcoma (formerly known as chloroma) involving the left orbit and cavernous sinus. A and B, Axial CT images following IV contrast administration. In A, a sharply outlined irregular contrastenhancing mass is seen in the lateral aspect of the left orbit (arrow). The mass slightly displaces the superior aspect of the left globe and the composite density of the levator palpebrae superioris and superior rectus muscles medially. At a slightly lower level (B), the tumor extends into and diffusely widens the left cavernous sinus (arrow).

A

B FIG 9-49  Germinoma of the pineal region. A, Axial CT image demonstrates a well-circumscribed round hyperdense mass in the pineal region extending slightly to the left and indenting the posteromedial aspect of the left thalamus (arrow) with a thin hypodense rim (likely representing edema) at the interface with the left thalamus. The lateral and third ventricles are enlarged owing to obstruction of the aqueduct by the tumor mass. B, Following IV contrast administration the mass demonstrates diffuse contrast enhancement (arrow).

http://radiologyebook.com

227

228

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-50  Germinoma of the suprasellar region. A, Sagittal T1-weighted MRI demonstrates a bulky lobulated ill-defined isointense (to gray matter) tumor that nearly completely fills the third ventricle and extends downward into the sella (arrow) and posteriorly into the prepontine cistern. The belly of the pons is flattened and displaced posteriorly by this extensive tumor. B, Axial T2-weighted image through the level of the sella. The homogeneously isointense intrasellar tumor (arrow) extends bilaterally to encase the adjacent internal carotid arteries and posteriorly into the prepontine cistern, abutting the ventral surface of the pons on the right. C, Following IV gadolinium, a sagittal T1-weighted image demonstrates patchy contrast enhancement of this extensive tumor (arrow).

germ layers—ectoderm, endoderm, and mesoderm. They are subdivided pathologically into mature and immature teratomas. Mature teratomas are composed entirely of fully differentiated tissue elements that may be arranged in patterns simulating those of normal organogenesis and originating from ectoderm (skin, brain, choroid), mesoderm (bone, tooth, cartilage, fat, muscle), and endoderm (intestine, respiratory). Immature teratomas occur more commonly,232 demonstrate greater hypercellularity and mitotic activity, and are composed of incompletely differentiated tissue elements resembling fetal tissues (e.g., the developing neural tube). An occasional teratoma contains a malignant component, such as rhabdomyosarcoma, undifferentiated sarcoma, squamous cell carcinoma, or adenocarcinoma.

Teratomas occur almost exclusively in males in the pineal, parapineal, and suprasellar regions. The clinical presentation for these lesions is earlier than for germinomas, usually during the first decade of life.15,157,232 On gross inspection, teratomas are typically well-defined lobulated masses containing multiple structures, including mucusfilled cysts, nodules of cartilage and bone, and rarely teeth and hair. In the pineal region these tumors are not usually invasive. For mature teratomas that can be completely excised, the 5-year survival rate is 80% to 90%.232 On both CT and MRI, teratomas are characterized by their striking heterogeneity in density and signal.15,113,360,419 Intratumoral cysts with well-defined margins and variable density or signal of contents (watery,

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms proteinaceous, lipid) are intermixed with calcifications or ossifications and areas of variegated soft tissue appearance, including hemorrhage. Dilatation of the lateral and third ventricles may indicate compression of the cerebral aqueduct by the tumor mass (see Fig. 9-49). Contrast enhancement is variable and when patchy or ringlike raises the possibility of malignant degeneration (Fig. 9-51). Invasion of neighboring structures can best be assessed on postcontrast images.

Other Germ Cell Tumors.  Endodermal sinus (yolk sac) tumor is characterized by accumulation of a myxoid matrix within a meshwork of primitive epithelial cells. Embryonal carcinoma is composed of large primitive cells that form abortive papillae and may attempt to replicate the structure of the early embryo. Choriocarcinoma consists of very large multinucleated trophoblastic cells that attempt to simulate placental tissue and include ectatic vascular channels that are prone to hemorrhage. Elements of these tumors may be found in germinomas or teratomas, and such tumors are then classified as mixed germ cell tumors. Approximately 10% of germ cell tumors contain more than one cell type.103 All of these neoplasms, including mixed germ cell tumors, have a high mitotic rate and a worse prognosis than pure germinomas or teratomas.172,245 These lesions exhibit a strong propensity for seeding via the CSF, for hematogenous metastasis to lung and bone, and even for spread of tumor via a ventriculoperitoneal shunt catheter into the peritoneal cavity. Differentiation of these lesions on CT or MRI studies from pure germinoma or teratoma depends on demonstration of a pineal region or suprasellar region mass with areas of intratumoral necrosis and hemorrhage. Hemorrhage is particularly common in choriocarcinoma, reflecting the vascular stroma of that tumor. However, these findings are nonspecific, and individual tumor type is best established with immunohistochemical evaluation of excised tumor tissue.360

Tumors of the Sellar Region Tumors of the pituitary gland and adjacent structures reflect the variety of tissues involved in the formation of these structures. Intrasellar, suprasellar, and parasellar tumors affect the CNS by expansion, with resultant compression of cranial nerves and the base of the brain, or by interference with the hypothalamic-neurohypophysial pathway. Alternatively, certain neoplasms, mainly intrasellar, may become clinically manifest because of endocrine hyperfunction. The most common neoplasms in this region are pituitary adenomas and craniopharyngiomas, both of which are considered to arise from derivatives of Rathke’s pouch, the embryonic infolding of endoderm that extends superiorly from the stomodeum and gives rise in early fetal development to the adenohypophysis (anterior lobe of the pituitary gland). The anterior lobe constitutes 75% of the volume of the pituitary gland. The posterior lobe of the pituitary gland is of neuroectodermal origin, extending inferiorly from the hypothalamus through the pituitary stalk (infundibulum). Neoplasms arising in the posterior pituitary are extremely rare. Less common masses in the sellar region, all of nonhypophysial origin, include chordomas, meningiomas, gliomas of the optic pathways, germinomas, teratomas, and metastases.

Pituitary Adenoma.  Adenomas of the anterior pituitary are relatively common benign neoplasms, representing 10% to 15% of all intracranial tumors. They are mainly tumors of adulthood, with less than 3% occurring in individuals younger than 18 years. Pituitary adenomas vary greatly in size, but the vast majority are less than 10 mm in diameter and are designated by convention as microadenomas. In a series of 1000 unselected autopsy specimens, 3.1% of

229

pituitary glands contained microadenomas, all in individuals older than 40 years. Microadenoma.  Microadenomas are typically well-demarcated lesions. Although they lack a definite capsule, they can be distinguished on gross inspection from the normal pituitary gland because they have a pseudocapsule of compressed pituitary tissue. Approximately 75% of pituitary adenomas are hormonally active; compared with nonfunctional (endocrinologically inactive) adenomas, hormonally active adenomas usually present earlier in their course of evolution when they are much smaller with symptoms and signs of endocrine hyperfunction. They are generally classified on the basis of the hormone produced—for example, lactotroph (prolactin), somatotroph (growth hormone), corticotroph (corticotropin [ACTH]), thyrotroph (thyroidstimulating hormone [TSH]), and gonadotroph (follicle-stimulating hormone [FSH]).196 Cells that secrete prolactin and growth hormone and the tumors arising from them tend to be located in the lateral portions of the gland, whereas cells that secrete ACTH, TSH, and FSH and the tumors arising from them tend to be located centrally. Approximately 50% of hormonally active adenomas are prolactinomas. They arise mainly in women (female-to-male ratio 4 to 5 : 1), who typically present in young adulthood with amenorrhea, galactorrhea, and infertility. Loss of libido and impotence are the most common presenting symptoms in males. The symptoms and signs are less pronounced in postmenopausal women and in men, in whom the tumors may manifest clinically only when they have become large enough to compress the visual pathways or cause pituitary hypofunction. The great majority of prolactin-secreting tumors lie in the lateral aspect of the anterior pituitary. Elevation of the serum prolactin concentration above the normal upper limit of 20 ng/mL raises the level of suspicion for the presence of prolactinoma, but elevated serum prolactin values can also be  associated with ingestion of certain drugs (notably phenothiazines  and tricyclic antidepressants), the presence of a suprasellar mass  compressing the hypothalamus or pituitary stalk, and primary hypothyroidism. Although markedly elevated serum prolactin values (>150 ng/mL) are almost always due to the presence of a prolactinoma, many patients with this tumor present with prolactin values between 20 and 150 ng/mL; in this group, diagnostic imaging often provides evidence allowing differentiation of prolactinoma from nonneoplastic causes of hyperprolactinemia.97 Somatotropic and corticotropic tumors each account for approximately 15% of microadenomas. The former cause gigantism in children and acromegaly in adults and are usually located laterally within the pituitary gland. The latter are located centrally, cause Cushing’s disease, and are typically very small. The great majority (75%) of patients presenting with Cushing’s disease are female. Thyrotropic and gonadotropic tumors are relatively rare. About 10% of microadenomas secrete more than one hormone, most commonly prolactin and growth hormone. Approximately 25% of pituitary adenomas are nonfunctional— that is, they do not cause nor are they directly associated with endocrine hyperfunction. These tumors are detected accidentally as microadenomas when patients undergo diagnostic imaging for other clinical indications, are discovered incidentally at autopsy, or present clinically at a later stage as macroadenomas when they are large enough to compress the pituitary stalk, optic chiasm, cranial nerves in the cavernous sinus or suprasellar cistern, or base of the brain.97 On CT and MRI, the pituitary gland is best visualized in the coronal plane. Because of the lack of bone artifact and the greater intrinsic soft tissue contrast on MRI, this modality has largely supplanted CT for initial detection and localization of pituitary microadenoma. Approximately 80% to 85% of these tumors demonstrate focal subtle

http://radiologyebook.com

230

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-51  Teratoma in the suprasellar region. A, On this axial noncontrast CT image an ovoid wellcircumscribed heterogeneous mass containing several cystic areas (arrow) fills the suprasellar cistern. B, Axial T1-weighted MRI displays a heterogeneously hypointense (to gray matter) suprasellar tumor with several small and large intratumoral cysts (arrow). C, Sagittal T1-weighted image following IV gadolinium demonstrates a multicystic suprasellar mass (arrow) that elevates and posteriorly displaces the floor of the third ventricle. The solid elements of this tumor demonstrate diffuse contrast enhancement, giving it a honeycomb-like appearance.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms hypointensity on T1-weighted images and focal mild hyperintensity on T2-weighted images. However, the differential signal intensity between the tumor and surrounding normal pituitary tissue is often very slight (Fig. 9-52A), thus accounting for the 15% to 20% falsenegative rate for MRI evaluation.200 Intratumoral calcification occurs only rarely in pituitary adenomas and even more rarely in microadenomas. Intratumoral necrosis, cyst formation, and hemorrhage are also infrequent occurrences that are much more likely in larger tumors. These changes typically cause  heterogeneity of the density or signal pattern within the lesion. The use of thin-section T1-weighted images in the coronal plane obtained after IV bolus administration of gadolinium improves sensitivity and specificity of MRI for detection and localization of pituitary microadenomas. The normal pituitary gland lacks a blood-brain barrier and thus opacifies rapidly and homogeneously after IV administration of contrast medium. Uptake of contrast agent by microadenomas is slower than that of the normal gland, and this difference provides greater conspicuity of the relatively hypointense tumor over the first 1 to 2 minutes after injection (see Fig. 9-52B). Rapid-sequence coronal single-slice T1-weighted imaging of the sella immediately after IV bolus administration of gadolinium, with images obtained consecutively at 10- to 15-second intervals, provides a “dynamic” demonstration of the difference in contrast uptake between tumor and normal pituitary (Fig. 9-53). A similar result can be obtained on CT with the use of thin-section coronal scans obtained after IV bolus administration of iodinated contrast medium; on early rapid-sequence serial images, microadenoma appears hypodense relative to the homogeneously contrast-enhancing normal pituitary tissue. This difference in contrast-enhancement effect fades after several minutes, and later or delayed images may mask the presence of the tumor or rarely may demonstrate reversal of the image contrast pattern, with the adenoma enhancing more strongly than the surrounding tissue. Laterally placed microadenomas may cause unilateral upwardly convex bulging of the overlying superior margin of the pituitary gland into the suprasellar cistern, downward protrusion of the inferior margin into the floor of the sella, or both. Also, the pituitary stalk may be displaced contralaterally by the expanding microadenoma. However, these findings are inconstant, depending on the size of the tumor and the location of its epicenter; when present they may raise the level of confidence in the diagnosis of microadenoma, but their absence does not detract from the significance of an intrasellar hypodense or hypointense mass. Also, pituitary stalk deviation can be seen in individuals without symptoms, signs, or imaging evidence of intrasellar or suprasellar tumor.2 The differential diagnosis of pituitary microadenoma on MRI or CT is limited. Pituitary adenomas may contain cells that are undergoing mitosis or may demonstrate dural invasion without being classified as malignant. The diagnosis of primary pituitary carcinoma is made only rarely and then usually in the circumstance of distant metastasis. Most malignancies occurring within the gland are metastases from primary carcinomas of the breast, lung, kidney, and gastrointestinal tract,176 although metastases account for only about 1% of intrasellar and parasellar tumors. Often multiple and tiny, metastases are demonstrated only on postcontrast high-resolution images. Nonneoplastic inflammatory entities that may involve the anterior lobe of the pituitary gland include lymphocytic hypophysitis and sarcoidosis. Lymphocytic hypophysitis is a rare disorder, likely autoimmune, that occurs mainly in young women during late pregnancy or postpartum. It is characterized by diffuse lymphocytic infiltration of the anterior pituitary. On imaging studies the pituitary gland appears diffusely enlarged and demonstrates contrast enhancement.219 Sarcoidosis, another autoimmune disorder, affects the CNS in about 5% of

231

cases, notably involving the meninges and parenchymal structures at the base of the brain. On imaging studies, multiple foci of ill-defined contrast enhancement are seen within and enveloping the surfaces of the hypothalamus, optic chiasm, and pituitary stalk71 (Fig. 9-54). Intrapituitary focal abnormalities are commonly identified on MRI of asymptomatic volunteers and as incidental findings in large autopsy series. These include not only microadenomas but also Rathke’s cleft cysts (see Fig. 9-59) (see later) and other benign cysts of uncertain origin. The incidence of such cysts on autopsy series varies from 5.8% to 8.3% of individuals older than 30 years. In one double-blind series of asymptomatic adult volunteers, MRI demonstrated intrapituitary focal hypointensities on T1-weighted imaging that were consistent with microadenoma or benign pituitary cyst in 15% of subjects. Another possible explanation proposed for these “incidentalomas” is magnetic susceptibility induction of signal distortion in the floor of the sella near the junction with the sphenoid sinus. Most of these lesions, apparent or real, are very small (<3 mm) and in the absence of correlating clinical symptoms or signs can be followed with periodic monitoring on an annual or biannual basis.137 Macroadenoma.  When an adenoma of the pituitary gland reaches or exceeds 10 mm in diameter, it is designated by convention a macroadenoma. The great majority of such lesions are endocrinologically nonfunctional, although immunohistochemical and electron microscopic studies now suggest that these “null cell” tumors may be elaborating proteins that are involved in formation or control of hormones. The tumors often become clinically evident in middle age when they reach sufficient size to compress the optic chiasm, causing a disturbance in vision, typically an asymmetric bitemporal hemianopia. Detailed clinical and laboratory evaluation may reveal evidence of hypopituitarism due to compression of the normal pituitary gland or stalk, but this is rarely a presenting symptom. Macroadenomas originate as intrasellar masses that slowly expand in all directions, notably upward out of the sella, protruding into the suprasellar cistern and eventually elevating and compressing the optic chiasm, hypothalamus, and the floor of the third ventricle (Fig. 9-55B and C). They also spread laterally and can invade across dura into the cavernous sinus and partially or completely encase the intracavernous internal carotid artery. Slow tumor growth in the inferior direction leads to thinning and erosion of the floor of the sella, with protrusion into the underlying sphenoid sinus. On plain films and CT images, this chronic slow enlargement causes expansion of the sellar cavity with thinning of the bony cortex of the dorsum, floor, and anterior wall of the sella. Histologically, pituitary adenomas consist of small oval or polyhedral cells in a sheetlike monomorphous array that contrasts markedly with the pleomorphic acinar pattern of the normal anterior pituitary gland. Cellular pleomorphism and even occasional mitoses may be seen in pituitary adenomas but do not indicate aggressive biological behavior. Calcification in these tumors is rare (in the range of 1%) but may be more common in prolactin-secreting tumors.93 Intratumoral necrosis, cystic change, and hemorrhage all occur more frequently in macroadenomas than in microadenomas. Hemorrhage is present in 20% to 30% of macroadenomas; it may be associated with pituitary apoplexy, an acute clinical event characterized by sudden onset of severe headache, visual loss, and hypotension. More often, however, intratumoral hemorrhage is subclinical and discovered incidentally on MRI283 (Fig. 9-56A; see Fig. 9-55B). Hemorrhage occurs even more commonly in patients with prolactinomas who are being treated with bromocriptine.414 On CT, macroadenomas often appear as large, homogeneously isodense, rounded midline masses arising out of the sella and insinuating upward to elevate and compress the optic chiasm and overlying third ventricle (see Fig. 9-55A). On occasion a

http://radiologyebook.com

232

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-52  Prolactin-secreting pituitary microadenoma. A, Coronal T1-weighted thin-section MRI through the mid-sella demonstrates a round isointense mass arising from the right side of the pituitary gland and projecting upward into the suprasellar cistern (arrow). B, Coronal T1-weighted image obtained shortly following IV gadolinium. Normal pituitary tissue demonstrates homogeneous contrast enhancement, while the round mass (arrow) shows little or no enhancement.

FIG 9-53  Pituitary microadenoma. Coronal T1-weighted MRI obtained during a dynamic scanning sequence following bolus IV administration of gadolinium. A 5-mm-diameter ovoid hypointense lesion in the left inferior aspect of the pituitary gland (arrow) fails to take up the contrast medium as rapidly as the surrounding normal pituitary tissue.

FIG 9-54  Sarcoidosis involving the base of the brain. Axial postgadolinium T1-weighted MRI demonstrates patchy diffuse contrast enhancement of the leptomeninges enveloping the base of the brain. A large focus (arrow) involves the pituitary stalk.

http://radiologyebook.com

B

A

D

C

E FIG 9-55  Pituitary macroadenoma with intratumoral hemorrhage and cystic changes. A, Axial noncontrast CT image demonstrates a soft tissue mass filling the suprasellar cistern. The tumor is mainly isodense to adjacent brain, but a faintly hyperdense focus at the right lateral margin (arrow) suggests recent hemorrhage. On a sagittal T1-weighted MRI (B) the tumor is seen to enlarge the sella and extend into the suprasellar cistern, elevating and slightly posteriorly displacing the optic chiasm and the anterior floor of the third ventricle. It is mainly homogeneously isointense to adjacent brain, but it contains several hyperintense foci (the largest is located in the posterior portion of the tumor and is indicated by an arrow) that likely represent subacute hemorrhage. C, Coronal T2-weighted image at a level through the anterior third ventricle. The dome of the tumor contains a large cyst. The mass displaces the supraclinoid internal carotid arteries laterally (arrow) and elevates and flattens the floor of the third ventricle. Sagittal (D) and coronal (E) T1-weighted images following IV gadolinium demonstrate homogeneous contrast enhancement of the solid portions of this large tumor. The arrow in D delineates the large focus of subacute hemorrhage in the posterior aspect of the tumor. The arrow in E demonstrates elevation and stretching of the proximal (A1) segment of the left anterior cerebral artery by this large intrasellar and suprasellar mass.

http://radiologyebook.com

234

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-56  Pituitary macroadenoma with intratumoral hemorrhage. A, Coronal T1-weighted MRI demonstrates a very large multilobulated well-circumscribed homogeneously hypointense tumor widening the sella and extending superiorly and bilaterally, more to the right than to the left. The most superolateral extension of this mass contains a large focal hyperintensity representing late subacute hemorrhage (arrow). The pituitary stalk is deviated to the left of the midline by the predominantly right-sided tumor. Note that the supraclinoid internal carotid arteries on both sides are compressed and displaced far laterally, as are the medial aspects of both temporal lobes. B, Following IV gadolinium, an axial T1-weighted MRI demonstrates diffuse homogeneous contrast enhancement of this multilobulated tumor (arrow).

tumor may grow so large as to invaginate the floor and walls of the third ventricle and obstruct the foramen of Monro. Areas of intratumoral necrosis or cystic degeneration appear as hypodensities within the tumor mass, whereas acute or subacute hemorrhage causes focal intratumoral hyperdensity. Because MRI offers multiplanar imaging capabilities and is not degraded by beam-hardening artifacts due to adjacent bone, it is the preferred modality for visualizing the pituitary gland. Pituitary macroadenomas appear hypointense relative to gray matter on T1-weighted images and hyperintense on T2-weighted images (see Figs. 9-55B and C and 9-56A). The T2 hyperintensity appears to correlate with softening of the tumor observed at the time of surgical removal and may reflect intratumoral necrosis. Normal pituitary tissue is usually severely compressed downward into the sellar floor by these bulky tumors and cannot be recognized.364 After IV administration of a contrast agent, delayed moderate contrast enhancement of the solid portions of the tumor mass appears on both CT and MRI200,274 (see Figs. 9-55D and E and 9-56B). A major advantage of MRI is the ability to demonstrate the location and status of the internal carotid arteries relative to the tumor. Lateral tumor extension into the cavernous sinus may displace the carotid flow void laterally (see Figs. 9-55C and 9-56). Further tumor extension may surround the flow void, indicating the presence of vessel encasement.73 In contradistinction to meningioma, such encasement by macroadenoma rarely causes arterial narrowing or occlusion.

Craniopharyngioma.  Craniopharyngiomas are benign partly cystic epithelial tumors that occur almost exclusively in the sellar region and are presumed to arise from remnants of Rathke’s pouch epithelium. They represent 1.2% to 4.6% of all intracranial tumors but are more common in children, in whom they represent 5% to 10% of intracranial neoplasms.40 They are most frequently suprasellar in location, with an intrasellar component; less commonly (≈20% of cases) they may be entirely suprasellar, and approximately 5% are purely intrasellar.141

These tumors have a bimodal age distribution that largely corresponds to two clinicopathologic forms, adamantinomatous and papillary. Approximately two thirds of cases of craniopharyngioma occur in the first 2 decades of life, mainly between ages 5 and 14 years; the great majority of these are of the adamantinomatous variety.233,425 The second age peak is in middle age (40-50 years); almost all papillary craniopharyngiomas occur in this group.85 Craniopharyngioma has no gender predilection. Craniopharyngiomas are slow-growing tumors that extend superiorly and often insinuate into the third ventricle. They may also expand anteriorly under the frontal lobes (30%), posteriorly into the interpeduncular cistern with impingement on and posterior displacement of the ventral margins of the midbrain and upper pons (20%), and laterally into the medial aspect of the middle cranial fossa (20%).141 Adamantinomatous tumors provoke an intense gliosis in the adjacent brain and become firmly adherent to adjacent brain and vessels. On gross inspection they have a multilobulated inner aspect with nodules of solid tissue separated by cysts of varying size. The cystic component is usually predominant in the adamantinomatous tumors and much less prominent in the papillary variety. Papillary tumors often involve the third ventricle. Calcification is present in the solid portions of nearly 90% of adamantinomatous craniopharyngiomas but is much less common in papillary tumors.348 The cyst contents have a “machine oil” color and consistency and frequently contain suspended debris with a high cholesterol content. Microscopically, adamantinomatous tumors consist  of broad strands and cords of stratified squamous epithelium with nodules of keratin and dystrophic calcification. In papillary tumors, sheets of squamous epithelium form papillae with a fibrovascular stroma, but keratin, cholesterol deposits, and calcification are typically lacking.77 Like pituitary macroadenomas, these slow-growing tumors present clinically most commonly with visual disturbances secondary to compression of the optic chiasm, nerves, or tracts. Endocrine deficiencies

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms are slightly less frequent. Compression of the neurohypophysis leads to diabetes insipidus in up to 15% of children and 30% of adults. Symptoms and signs of increased intracranial pressure (headache, nausea, vomiting, papilledema) secondary to third ventricular obstruction that may lead to cognitive impairment and personality changes are also common presenting findings.

A

235

On CT these tumors typically appear as heterogeneously hypodense suprasellar masses with nodular solid areas intermixed with cystic areas of varying size. The cyst contents are more dense than CSF. More than 80% of craniopharyngiomas in children (usually of the adamantinomatous variety) demonstrate nodular and curvilinear calcifications in the solid nodules and in the cyst walls141,425 (Fig. 9-57A).

B

C FIG 9-57  Craniopharyngioma in a child. A, Axial postcontrast CT image of a large suprasellar cyst with several nodular calcifications of varying size (arrow) in the wall of the cyst. B, Axial T2-weighted MRI demonstrates strongly hyperintense homogeneous cyst contents. The well-circumscribed cyst (arrow) displaces the anterior cerebral arteries anteriorly and the middle cerebral arteries bilaterally. C, Coronal T1-weighted postgadolinium image demonstrates a cystic intrasellar/suprasellar mass with strong contrast enhancement of the cyst wall (arrow). The cyst contents are isointense to gray matter, reflecting their high protein content.

http://radiologyebook.com

236

PART II  CT and MR Imaging of the Whole Body

Papillary craniopharyngiomas in adults appear more solid and rarely demonstrate calcification.77 Cysts are found in approximately 50% of papillary tumors versus about 90% of adamantinomatous neoplasms.348 In contrast to pituitary adenomas, slow-growing craniopharyngiomas do not usually enlarge the sella but rather gradually erode the posterior clinoid processes and the upper portion of the dorsum sellae. Adamantinomatous suprasellar tumors appear as heterogeneous masses of variable intensity on T1-weighted MRI. Areas containing calcification usually appear more hypointense. Cystic components are well marginated and vary in signal but are commonly hyperintense relative to gray matter, reflecting high concentration of protein in the cyst contents (Fig. 9-58B and C; see Fig. 9-57C). On T2-weighted images, the solid areas vary from hypointense to mildly hyperintense as compared to gray matter, whereas the cysts present a more internally uniform hyperintense appearance232 (see Figs. 9-57B and 9-58A). Papillary tumors have a more uniform and homogeneous appearance.348 On both CT and MRI, after IV administration of contrast agents, enhancement of the noncalcified solid components of adamantinomatous tumors and of the solid papillary tumors is typically intense and uniform77,141 (see Figs. 9-57C and 9-58B and C). Both adamantinomatous and papillary craniopharyngiomas are managed surgically. Ten-year recurrence-free survival rates vary from 64% to 96%.77 The extent of surgical resection is the most important factor in predicting recurrence; because of the intense gliotic reaction provoked by the tumor as it impinges on adjacent brain and the adhesion of large tumors to vascular structures in the suprasellar region, incomplete surgical resection remains a significant problem, especially in tumors larger than 5 cm in diameter.

Rathke’s Cleft Cyst.  Both Rathke’s cleft cysts and craniopharyngiomas are considered to arise from embryonic rests of Rathke’s pouch. In normal embryonic development, Rathke’s pouch regresses by the sixth week of gestation into a narrow cleft. Persistence and enlargement of that cleft is believed to be the cause of Rathke’s cleft cysts, which are lined by cuboidal or columnar epithelium and are usually located in the pars intermedia between the anterior lobe (adenohypophysis) and posterior lobe (neurohypophysis) of the pituitary gland. Rathke’s cleft cysts are typically small (2-10 mm) thin-walled smoothly marginated intrapituitary cysts with mucinous contents. Less commonly the contents may be serous.75 They are usually asymptomatic and found incidentally on MRI or at autopsy. In an unselected series of 1000 cases, incidental Rathke’s cleft cysts were found in 11.3%.387 Nearly 90% occur in the center of the pituitary gland. Larger cysts may be both intrasellar and suprasellar and may become symptomatic if they compress the pituitary gland, pituitary stalk, or optic chiasm, but such compression is unusual. They are found in patients in all age groups but are usually identified in adults. A 2 : 1 to 3 : 1 female predominance has been reported, but this figure may reflect selection bias because more women than men undergo pituitary imaging, mainly for menstrual problems, hyperprolactinemia, and infertility. On CT, 75% of Rathke’s cleft cysts appear as round, well-marginated, hypodense intrapituitary lesions. The density of the cyst likely reflects the concentration of mucoprotein in the cyst contents. Calcification is not found in these cysts. On MRI also, they appear discrete and sharply marginated. On T1-weighted images, approximately two thirds of Rathke’s cleft cysts appear hyperintense and one third hypointense relative to gray matter (Fig. 9-59B); this variance is also likely related to the protein content of the cyst fluid. On T2-weighted MRI, approximately 50% exhibit moderate to strong hyperintensity relative to gray matter (see Fig. 9-59A), another 25% appear isointense, and a like

number are hypointense. After IV administration of contrast agent,  a Rathke’s cleft cyst does not enhance on either CT or MRI, although the surrounding compressed pituitary tissue may exhibit rimlike enhancement.338

Tumors of the Neurohypophysis.  Neoplasms arising in the posterior lobe of the pituitary gland or the pituitary stalk are very uncommon. Granular cell tumor of the neurohypophysis is a rarely occurring benign slow-growing intrasellar or suprasellar mass that originates within the posterior pituitary or the pituitary stalk and presents as a well-defined rounded mass composed of densely packed nests of large cells with a granular cytoplasm. These tumors, considered to be of neural crest or glial origin, exhibit intense immunoreactivity for the S-100 tumor marker. Like many intrasellar and suprasellar neoplasms, granular cell tumors usually develop insidiously and present with hormonal and visual disturbances secondary to compression of normal pituitary and optic pathway structures. On imaging studies they appear as well-circumscribed round masses within the posterior pituitary or on the posterior aspect of the pituitary stalk. The tumors are isodense with brain on CT and isointense with gray matter on T1- and T2-weighted MRI. The posterior pituitary “bright spot” on T1-weighted images is effaced or absent. Granular cell tumors are richly vascular and demonstrate strong contrast enhancement.67 Langerhans cell histiocytosis involves the sellar region as a component of the Hand-Schüller-Christian syndrome; granulomas may be found in the pituitary stalk, hypothalamus, or both in association with destruction of the bony margins of the sella (see earlier). Imaging studies may demonstrate thickening and contrast enhancement of these structures390 (Fig. 9-60).

Chordoma.  Chordomas are histologically benign neoplasms derived from remnants of the embryonic notochord. Despite their benign histology, they are locally invasive and destructive, behaving clinically like low-grade malignancies. These tumors arise in or near the midline and occur most commonly at the proximal (clivus-basisphenoid) and distal (sacrum) ends of the notochord. Approximately 40% of chordomas arise in the basisphenoid region. They are uncommon tumors, representing less than 0.2% of all intracranial neoplasms. Chordomas can present clinically in any age group, but their peak incidence is in patients between 20 and 50 years of age; less than 5% are identified in children.69 Chordomas arising in the upper clivus expand slowly and inexorably in all directions, infiltrating and eroding the surrounding bone. Posterior extension leads to the presence of a bulky lobulated mass in the prepontine cistern, compressing and displacing (but not invading) the pons and midbrain posteriorly. Anterior extension may manifest as a nasopharyngeal mass. Upward growth causes destruction of the walls of the sella with extension of tumor into and above the sella. Tumor infiltration into the cavernous sinuses with encasement of arteries and nerves is common. Sequestra of intratumoral necrotic bone and calcification and foci of intratumoral hemorrhage are common findings. Histologically, chordoma consists of clusters of large vacuolated physaliferous (Greek for “bubble-bearing”) cells interspersed with abundant extracellular mucinous material and separated into lobules by fibrous septa.69 The clinical presentation of this slowly expanding neoplasm is insidious and depends on its precise location. The majority of patients present with head or neck pain and cranial nerve deficits. The most common presenting complaint is diplopia, noted in 60% to 90% of cases. Upward extension of chordoma causes visual disturbances and hypopituitarism, whereas posterior extension leads to extraocular

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

237

B

C FIG 9-58  Craniopharyngioma in an adult. A, Axial T2-weighted MRI demonstrates a multilobulated multicystic hyperintense midline tumor (arrow) with irregular margins that has insinuated itself between the basal ganglia of the two cerebral hemispheres. The tumor extends more to the right than to the left of the midline, and the right globus pallidus is compressed and displaced laterally. The cyst contents are more hyperintense than the solid tumor rim in this papillary tumor. Axial (B) and right parasagittal (C) T1-weighted MRIs following IV gadolinium demonstrate contrast enhancement of the solid tumor rim (arrow). The cyst contents are hyperintense relative to CSF, reflecting their high protein content. On the sagittal image, note that this cystic tumor extends into and nearly obliterates the third ventricle with its superior margin projecting into the floor of the right lateral ventricle; posteroinferiorly, the tumor projects into the interpeduncular cistern and displaces the midbrain posteriorly.

nerve palsies. Hearing loss and dysphagia are relatively common presenting findings that correlate with downward and anterior spread  of tumor.111 CT of clival chordoma demonstrates irregularly marginated erosion and destruction of the bony clivus, with heterogeneous attenuation of

the associated intraosseous and extraosseous lobulated soft tissue mass. The heterogeneity is due to the presence of hyperdense intratumoral bony sequestra and calcifications intermixed with areas of hypodensity and hyperdensity secondary to cystlike mucinous collections as well as intratumoral hemorrhage.

http://radiologyebook.com

238

PART II  CT and MR Imaging of the Whole Body

A

B FIG 9-59  Rathke’s cleft cyst. Axial T2-weighted (A) and coronal postgadolinium T1-weighted (B) images demonstrate a small smoothly marginated cystic mass (arrow) within and projecting above the pituitary gland. The cyst appears slightly hyperintense relative to gray matter on both T1 weighting (B) and T2 weighting (A). There is no contrast enhancement of its contents or margins.

A

B FIG 9-60  Langerhans cell histiocytosis of the hypothalamus and pituitary stalk. A, Axial T2-weighted MRI displays a large ill-defined area of hyperintensity (likely representing edema) in the region of the hypothalamus (arrow). Within this hyperintensity can be seen two smaller isointense (to gray matter) foci. B, Coronal T1-weighted postgadolinium image demonstrates homogeneous contrast enhancement of the two foci; the larger lesion (arrow) extends into and widens the pituitary stalk.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

239

B

A

C FIG 9-61  Chordoma of the lower clivus and foramen magnum. Axial T1-weighted (A) and T2-weighted (B) MRIs through the level of the lower clivus demonstrate a sharply marginated lobulated mass in and to the right of the midline that is hypointense on T1 weighting and hyperintense on T2 weighting and has destroyed the lower clivus, infiltrating anteriorly into the longus colli muscles (arrow in B) and extending posteriorly into the premedullary cistern (arrow in A), displacing the medulla posteriorly and indenting its right ventrolateral margin. In B, note that the tumor has encased the vertebral arteries bilaterally. C, Sagittal T1-weighted postgadolinium image demonstrates faint homogeneous contrast enhancement of this extensive tumor that presents anteriorly as a submucosal nasopharyngeal mass (arrow) and projects posteriorly to compress the medulla and upper cervical spinal cord against the posterior margin of the foramen magnum.

On T1-weighted MRI, the infiltrating tumor appears isointense to gray matter in 75% of patients and mildly hypointense in the remaining 25% (Fig. 9-61A and C). In either case the tumor presents a striking contrast to the diffusely hyperintense T1 signal of the adjacent normal marrow fat of the clivus. On T2-weighted images, chordoma demonstrates moderate heterogeneity of signal with areas of hyperintensity intermixed with linear strands of hypointensity (see Fig. 9-61B).248,382 After IV administration of contrast material, varying and heterogeneous contrast enhancement is seen within the tumor mass on both CT and MRI (see Fig. 9-61C).93 Differentiation of chordoma from chondrosarcoma—by both imaging and gross and microscopic examination—may be very difficult. There is considerable overlap of imaging and histologic appear-

ances between these two tumors. Tumor location may be of limited help in this differentiation; chordoma arises in or near the midline, whereas the majority of chondrosarcomas of the skull base tend to arise laterally in or near the petroclival junction. However, chondrosarcomas often spread medially into the clivus and posteriorly into the cerebellopontine angle, extending into the ipsilateral cavernous sinus and encasing the arteries of the circle of Willis in a manner similar to chordoma.93 Immunohistochemical studies also aid in this differentiation; chordoma is of neural crest origin and stains positively for epithelial membrane antigen and cytokeratin, but chondrosarcoma, being of mesodermal origin, does not take up these markers.337 Other neoplasms occurring in or metastasizing to the sellar region (e.g., meningiomas, gliomas of the optic pathways, epidermoid tumors,

http://radiologyebook.com

240

PART II  CT and MR Imaging of the Whole Body

germ cell tumors, carcinomatous metastases) are considered elsewhere in this chapter.

Metastatic Tumors Hematogenous metastases from extracranial primary malignant neoplasms may involve the skull, dura, leptomeninges, or brain parenchyma. In adults, metastasis to the brain is most common.

Metastases to the Brain.  More than 30% of intracranial tumors detected in a multiinstitutional prospective CT study were metastatic malignancies.17 Autopsy studies have demonstrated the presence of intracranial metastasis in approximately 25% of patients with cancer.311,312 The incidence of intracranial metastasis is rising because of (1) the wider availability and greater refinement of cross-sectional diagnostic imaging and (2) continuing improvements in other diagnostic and treatment modalities, with resultant prolonged survival of patients with systemic tumors. The most common sources of intracranial metastasis, in order of decreasing frequency, are carcinomas of the lung and breast, malignant melanoma, and carcinomas of the kidney and gastrointestinal tract.15,314,341 Certain primary malignancies, particularly bronchogenic carcinoma,383,386 carcinoma of the breast,384 choriocarcinoma, and melanoma,119 have a notable propensity to metastasize to the brain. Between 30% and 65% of patients with carcinoma of the lung exhibit cerebral metastases at autopsy, many of which were not clinically evident during life.175,341 The most common cause of death in patients with melanoma is metastasis to the brain; autopsy studies reveal cerebral metastases in up to 90% of patients with melanoma.39 Cerebral metastases occur most frequently at the junction of cortex and underlying white matter,314 a finding likely related to obstruction by circulating tumor cell emboli of the penetrating arteries, which abruptly narrow as they enter the subcortical white matter.341 Eighty percent of brain metastases are located in the arterial distribution zones of the cerebral hemispheres, with 3% in the basal ganglia and 15% in the cerebellum. A notable departure from this pattern occurs in the case of mucinous adenocarcinomas of gastrointestinal origin; although these tumors represent only 5% of intracranial metastases, approximately 50% occur in the posterior fossa, mainly in the cerebellum. Metastasis to the cerebellum occurs mainly at the junction of the superior and inferior cerebellar artery distributions.82 Differentiation between solitary and multiple cerebral metastases has a major impact on the further management of the patient. The relative ratio of solitary to multifocal cerebral metastases in autopsy series has been reported as approximately 50 : 50. However, experience with advances in diagnostic imaging, particularly with postcontrast MRI, suggests that about 70% of patients with cerebral metastasis demonstrate more than one lesion.15 On gross inspection, small metastases are usually discrete spherical well-circumscribed tumors (Fig. 9-62). However, microscopic examination often reveals tumor cell infiltration into the adjacent surrounding normal tissue with angioneogenesis, perivascular invasion, and reactive gliosis.341 Histologically, metastases typically recapitulate the pattern of the tumor of origin. These are markedly malignant neoplasms with high rates of mitotic activity. Edema of the adjacent white matter is often disproportionately extensive in comparison with the small size of the tumor and may spread along white matter fiber tracts for a considerable distance.341,383 As the tumor implant grows, it penetrates and distorts the surrounding gray and white matter. Further growth is typically associated with central necrosis like that seen in glioblastoma, leaving recognizable tumor tissue only at the periphery and around the blood vessels. Intratumoral hemorrhage is found in about 20% of cerebral metastases, notably in melanomas (Fig. 9-63),

FIG 9-62  Hematogenous metastases of small cell carcinoma of the lung to the brain. Axial T1-weighted postgadolinium image demonstrates multiple contrast-enhancing ring and ovoid lesions of varying sizes (arrows) at the junction of gray and white matter in both cerebral hemispheres. Although numerous, these tumors are relatively small and have not yet provoked grossly evident surrounding white matter edema.

choriocarcinomas, and carcinomas of the lung, kidney, and thyroid.15,236 Although metastases to the CNS can occur in all age groups, more than 75% are found in older patients. The age-specific incidence rises steeply after age 45.25,296 Approximately 30% of patients with cerebral metastases are asymptomatic at the time of initial diagnosis—that is, the tumors are discovered incidentally during evaluation for other causes (e.g., head trauma) or are found on screening imaging surveys of patients who have primary tumors with a known high incidence of hematogenous seeding to the brain. When symptoms occur, they are usually of insidious onset and slow progression and relate either to increased intracranial pressure or to localized mass effect; they include headache (in ≈ 50%), nausea and vomiting, mental status or behavioral change, progressive neurologic deficit, and seizures. These symptoms do not differ from those caused by primary gliomas in the same area. Mean survival time after pathologic confirmation of the diagnosis of cerebral metastasis is 1 to 3 months.29 Despite advances in surgery, irradiation, and chemotherapy, the prognosis for patients with cerebral metastases from extracranial primary malignant neoplasms has only minimally improved. When only a solitary parenchymal lesion of the brain can be identified and the extent of the disease in other organs is not progressing, surgical excision of the cerebral metastasis correlates with improvement in both quality and length of life.377 The appearance of hematogenous cerebral metastasis on noncontrast CT scans corresponds closely to the gross pathologic changes already described. Small tumors appear as rounded homogeneously isodense or, less commonly, slightly hyperdense nodules relative  to adjacent normal brain.15,314 The isodense lesions merge imperceptibly with the adjacent cortex and may be detected only when they distort the inner margin of the cortex. The hyperdense nodules

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

241

B FIG 9-63  Hematogenous metastases of melanoma to the brain with intratumoral hemorrhage. A, An axial noncontrast CT image demonstrates at least four foci of varying sizes located at gray-white junctions in both cerebral hemispheres. All four lesions contain blood-fluid levels (arrows) indicating the presence of intratumoral hemorrhage. The lesions in the frontoparietal regions bilaterally and the left deep frontal area have provoked considerable surrounding edema. B, An axial CT image at an adjacent level following IV contrast injection demonstrates faint ringlike contrast enhancement of the margins of several of these tumors (arrows).

typically represent small round-cell tumors or other tumors with high nucleus-to-cytoplasm ratios. However, most metastases from melanoma and choriocarcinoma as well as some metastases from carcinomas of the thyroid and kidney appear hyperdense because of hemorrhage within the tumor (see Fig. 9-63A). Some small lesions and virtually all large metastases have undergone central necrosis by the time they are detected and appear as ringlike masses with central lucency and a peripheral isodense or slightly hyperdense rim of irregular thickness.314 As already noted, edema (hypodensity) of the surrounding white matter is usually extensive and produces considerable mass effect (see Fig. 9-63). Calcification is rarely found in association with intracerebral metastases because of their rapid growth; however, it may occasionally be found in metastases originating from carcinoma of the breast or gastrointestinal tract.82 Contrast enhancement is virtually universal in these tumors. Less than 4% of the metastases in a multiinstitutional prospective CT study did not demonstrate contrast enhancement of the viable peripheral tumor rim (in lesions with central necrosis) or of the entire tumor (in small solid lesions of homogeneous density).314 Enhancement is typically intense and fairly uniform within viable portions of the neoplasm, enabling clear separation of small tumor foci that are isodense on precontrast CT images from surrounding edema. Definitive preoperative differentiation of metastasis from glioblastoma with CT is often not possible in an individual case, even though larger metastases generally retain a spherical or ovoid outline and extensive glioblastomas most commonly have much more irregular lobulated configurations and exhibit greater heterogeneity of density on contrast-enhanced images. Definitive differentiation of multiple

metastases from multiple pyogenic cerebral abscesses also is not usually possible with CT; although virtually all abscesses demonstrate uniformity and smoothness of contrast-enhancing wall thickness, so may many small metastases. Extensive surrounding edema is also characteristic of both processes. As noted previously, CT with contrast injection has been used for the preoperative screening evaluation of patients with carcinoma of the lung, malignant melanoma, and choriocarcinoma who have no cerebral symptoms or neurologic deficit. In one study of individuals without clinical evidence of brain involvement, cerebral metastases were found on CT in 40% of patients with adenocarcinoma of the lung and 11% of patients with oat cell carcinoma of the lung386; in another study, 11% of patients with melanoma were found on CT  to have metastatic intracranial lesions.119 Demonstration of unsuspected cerebral metastases in such patients may eliminate unnecessary surgery and permit earlier institution of more appropriate forms of treatment. On evaluation with MRI, most intracerebral metastatic lesions have prolonged relaxation times (diminished signal on T1-weighted images and increased signal on T2-weighted images compared with normal brain).15 As with CT, differentiation of the lesion or lesions from surrounding white matter edema is often problematic on noncontrast images. Additionally, signal intensity of the tumor can be highly variable depending on the tumor constituents; the signal characteristics in a given metastasis depend on the cellularity of the lesion, extent of intratumoral necrosis, presence and age of hemorrhage, presence and extent of calcification, and severity of surrounding edema.14,16 Melanoma has a somewhat characteristic appearance if there has not been

http://radiologyebook.com

242

PART II  CT and MR Imaging of the Whole Body

previous hemorrhage; the lesion is hyperintense on T1-weighted images and isointense to brain on T2-weighted images, most likely because of the free radical content of the melanin.16 Metastases from mucinous adenocarcinomas of gastrointestinal origin often appear hypointense on T2-weighted images because of the high protein content of mucin.15 If intratumoral hemorrhage has occurred, the signal pattern is determined by the nature of the blood degradation products (i.e., age of the hemorrhage).16,127 Contrast-enhanced MRI has proven even more sensitive than contrast-enhanced CT for detection of intracranial metastatic disease.148,343 Multiplanar T1-weighted MRI obtained after IV administration of gadolinium, usually in a dose of 0.1 mmol/kg of body weight, are currently the most sensitive method for evaluation of intracranial metastatic disease.15 As on CT, contrast enhancement on MRI allows the reader to distinguish the gross margins of the metastasis from the surrounding edema (see Fig. 9-62), a differentiation not often permitted on precontrast T1- or T2-weighted MRI. Differentiation of metastasis from primary glioma can be significantly aided with the use of MRS, DWI, and perfusion MRI or by  a combination of these methodologies, as discussed later in this chapter. Differentiation of metastasis from abscess is aided with the use of DWI.86 Several investigators have suggested that higher doses of gadolinium (0.2-0.3 mmol/kg) further improve the sensitivity of this examination for detecting additional lesions, although unlike CT, delayed MRI does not increase the contrast between the tumor and surrounding brain.145,417 In a patient with known extracranial primary malignancy in whom standard-dose postcontrast MRI of the brain appears to demonstrate a solitary parenchymal contrast-enhancing metastasis and who does not have compromised renal function, it would seem appropriate and cost-effective to repeat the postcontrast images with a higher dose of gadolinium to more conclusively exclude the possibility of more metastases before surgical resection of the apparent solitary lesion is undertaken.

Metastases to the Skull and Intracranial Dura.  Metastasis to the cranial vault or the base of the skull is associated with intracranial metastasis in 5% to 10% of cases.149,341 Hematogenous metastasis to the calvarial diploë often leads to destruction of the internal and external tables and extension of tumor into the adjacent epidural space. In one autopsy series of patients with primary extracranial malignancy and intracranial metastasis, 18% demonstrated dural involvement (including epidural or subdural neoplasm) as the only site of intracranial metastasis. Occasionally dural metastasis may occur without an associated calvarial lesion, either as a nodular plaquelike focus or as a more extensive diffuse thickening of the dura. Epidural or dural metastasis is most often secondary to carcinoma of the breast, lymphoma, carcinoma of the prostate, and neuroblastoma. Other common sources are carcinomas of the lung and kidney. Although the dura is usually an effective barrier to further deep invasion, spread of tumor in the epidural space with nodular or bandlike soft tissue thickening is common. Symptoms usually relate to compression of the underlying brain parenchyma. Irregularly marginated focal osseous destruction of the bony calvaria or the skull base caused by metastasis to the diploë or marrow can be best defined on noncontrast CT scans obtained through the use of a bone-targeting algorithm with wide window and high center settings. Destruction of the inner table of the skull allows extension of tumor into the intracranial epidural space, which can be demonstrated on CT scans obtained after IV administration of iodinated contrast medium with the use of a soft tissue algorithm and appropriate (“subdural window”) soft tissue window and center settings. These tumors

FIG 9-64  Diffuse pachymeningeal and calvarial metastasis from carcinoma of the breast. Axial T1-weighted postgadolinium MRI demonstrates diffuse nodular and bandlike contrast-enhanced thickening of the dura over the high right frontoparietal convexity. The process extends into the falx and across the midline into the high left frontal convexity dura. Several ovoid and plaquelike masses in the right frontoparietal convexity represent outcroppings of dural metastasis (arrow) and compress the underlying frontal and parietal lobes, with resultant effacement of the sulci. Note the loss of normal fatty marrow signal in the diploë of the overlying right parietal bone, indicating the presence of calvarial tumor involvement. Immediately posterior and lateral to the arrow is a focal plaque of tumor extension into the subgaleal space, with elevation of the overlying subcutaneous fat of the scalp.

typically demonstrate intense contrast enhancement, but recognition of the contrast-enhancing biconvex epidural tumor mass can be masked by the adjacent bone unless relatively wide window and high center settings are used. In current practice, early detection of small metastases in the diploë or dura can best be accomplished by using T1-weighted MRI both before and after IV administration of gadolinium. Small foci of intradiploic metastasis appear hypointense on noncontrast T1-weighted images and differ sharply from the adjacent hyperintense fatty marrow in the diploic space (Fig. 9-64). Postgadolinium T1-weighted images obtained with fat suppression typically demonstrate intense contrast enhancement of tumor in the epidural space or dura, which appears as nodular or curvilinear bandlike thickening (see Fig. 9-64); the underlying cerebral cortex is displaced inward, and (unlike leptomeningeal metastasis) the contrast-enhancing tumor does not extend into the cerebral sulci. The findings may simulate those of meningioma, but epidural or dural metastasis typically appears hyperintense to adjacent cerebral cortex on T2-weighted images, whereas meningioma is nearly isointense to cortex on this pulsing sequence.313

Leptomeningeal Metastasis (Leptomeningeal Carcinomatosis).  Hematogenous metastasis from primary extracranial malignant

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms tumors may occur directly to the meninges, but leptomeningeal carcinomatosis may also result from parenchymal neoplasm that breaches either the ependymal lining of the ventricles or the pia overlying the cortex with seeding of tumor cells into the CSF and subsequent spread by implantation on meningeal surfaces.196,341 These metastases tend to occur most commonly in the basal cisterns and in areas of relative CSF stasis. The malignant infiltrate induces a reactive inflammatory response of the leptomeninges, and communicating hydrocephalus may develop as a result of obstruction of the basal cisterns or the pacchionian granulations. Sources include most primary parenchymal neoplasms of the brain, even relatively low-grade gliomas,64 as well as carcinomas of the breast, lung, and gastrointestinal tract; melanoma; lymphoma; and leukemias, notably acute lymphatic leukemia. Of the secondary causes, carcinomas of the breast and lung are most common; in as many as 50% of patients with small cell carcinoma of the lung, leptomeningeal metastasis may be found at autopsy. Leptomeningeal metastasis is nearly always symptomatic, and simultaneous presentation of signs and symptoms in widely scattered areas of the CNS is typical. Headache is the most common initial complaint, occurring in 50% of patients, and weakness is found on clinical examination in 80%. Cranial nerve palsies are presenting complaints in 40% of patients, and lethargy and confusion are found in 20%.405 Although on occasion the diagnosis of leptomeningeal metastasis may be established before the primary tumor becomes symptomatic, the interval between primary diagnosis of neoplasm and diagnosis of leptomeningeal involvement is typically between 6 months and 3 years; at this stage, most patients also have evidence of metastatic disease outside the CNS. The diagnosis is usually established by lumbar puncture and microscopic examination of the CSF; abnormal cytologic results are obtained in approximately 50% of patients on the initial analysis and in 95% after six lumbar punctures. Supporting findings include elevation of CSF pressure above 160 mm Hg, increase in protein concentration above 20 g/L, and elevated white blood cell count (usually lymphocytes), all of which are found in at least half of patients.263 Although CT has proved highly sensitive for detection of cerebral parenchymal metastases, leptomeningeal metastasis has been demonstrated on contrast-enhanced CT images in less than a third of patients with cytologic evidence of CSF seeding.17,119,211 The sensitivity of MRI without and with paramagnetic contrast enhancement for detection of leptomeningeal metastasis appears to be approximately twice that of CT.120,381,415 MRI findings strongly suggestive of leptomeningeal metastasis include (1) nodular or plaquelike thick curvilinear hyperintensity of the involved subarachnoid cisterns and spaces on FLAIR images and (2) homogeneous contrast enhancement on postcontrast T1-weighted images that diffusely involves one or more subarachnoid cisterns, coats the surfaces of the adjacent brain parenchyma, and extends into the sulci (Fig. 9-65). The leptomeningeal tumor may also invade the underlying brain parenchyma and infiltrate the perivascular spaces surrounding the penetrating vessels. The greater sensitivity of MRI for detecting leptomeningeal tumor spread has established this technique as the diagnostic modality of choice for evaluation of possible leptomeningeal carcinomatosis. The differential diagnosis of leptomeningeal contrast enhancement includes bacterial and fungal meningitis, sarcoid (see Fig. 9-54), and postoperative inflammatory changes in the meninges. On the basis of imaging alone, definitive differentiation of tumor from inflammation or infection is not possible. Local or diffuse meningeal contrast enhancement at the site of previous craniotomy or at the site of insertion of a ventriculostomy shunt catheter reflects an inflammatory response of the traumatized dura and leptomeninges, with eventual fibrosis. Nodularity of the contrast-enhancing tissue may favor

243

FIG 9-65  Leptomeningeal metastasis/carcinomatosis from carcinoma of the breast. Axial T1-weighted postgadolinium MRI. Numerous and diffuse nodular foci of leptomeningeal thickening with intense contrast enhancement are seen in the sylvian fissures, over the convexity surfaces, and in the convexity and parasagittal sulci bilaterally. A large nodule fills the pericallosal sulcus anteriorly (arrow). There is associated mass effect in the right frontotemporal region with resultant compression of the frontal horn of the right lateral ventricle and slight displacement of the third ventricle to the left of the midline.

neoplastic over inflammatory processes, but by no means is this true in all cases.162,301

CHEMOTHERAPY AND BRAIN RADIOTHERAPY In addition to surgery, many patients with brain tumors will undergo treatment with chemotherapy, radiation therapy, or a combination of both. Chemotherapy and brain radiotherapy may result in changes in the appearance of the brain parenchyma. In some cases these changes may mimic the appearance of brain tumor. Treatment effect is categorized radiologically as white matter injury (WMI) (leukoencephalopathy) or radiation necrosis. The nomenclature can be confusing: chemotherapy can result only in WMI, but radiation injury, depending on the stage, can result in WMI or radiation necrosis.62 Radiation therapy results in neurotoxicity via vascular injury, glial and white matter damage, effects on the fibrinolytic enzyme system, and immune mechanisms.202 Chemotherapy induces neurotoxicity through plurifactorial mechanisms including induction of blood-brain barrier permeability and stimulation of aging mechanisms (e.g., cell apoptosis, DNA damage, free radical production).365 Chemotherapy-induced WMI results in diffuse homogeneous T2 hyperintensity throughout the cerebral white matter without or with cerebral volume loss (Fig. 9-66). In the absence of concurrent radiation therapy, additional findings such as contrast enhancement, necrosis, hemorrhage, and infarct are rare. In the case of young children

http://radiologyebook.com

244

PART II  CT and MR Imaging of the Whole Body Similar to radiation necrosis, another radiologic phenomenon  that mimics early tumor recurrence in treated high-grade gliomas  on conventional imaging is pseudoprogression. Pseudoprogression typically occurs in the first 3 months (up to 60%) after radiation therapy but has been described up to 6 months post irradation.32 Pseudoprogression occurs with radiation therapy alone but is widely believed to be more common with concomitant chemoradiotherapy (RT-temozolamide [TMZ]).33 Pathologically, pseudoprogression has been found to correspond to gliosis and radiation-induced changes without evidence of viable tumor.396 Imaging findings of pseudoprogression are increased enhancing lesion size/peritumoral enhancement in the early posttreatment period. Follow-up imaging will typically demonstrate spontaneous resolution, with some cases progressing to radiation necrosis. Like radiation necrosis, pseudoprogression is difficult to distinguish from progressive neoplasm on conventional imaging, but occurrence in the early posttreatment period and resolution on follow-up imaging distinguishes pseudoprogression from recurrent/progressive neoplasm.163

ADVANCED NEUROIMAGING TECHNIQUES FIG 9-66  Chemotherapy-induced white matter injury. Axial turbo FLAIR MRI of the brain in a 56-year-old woman who received fludarabine prior to bone marrow transplant for multiple myeloma. The image demonstrates diffuse abnormal signal hyperintensity throughout the cerebral white matter without mass effect.

undergoing intrathecal methotrexate therapy either for prevention of CNS relapse or for treatment of CNS relapse, most of these white matter abnormalities are reversible and without long-term neurologic sequelae.62 Radiation neurotoxicity can be divided into acute (1-6 weeks postirradiation), early-delayed (3 weeks to several months postirradiation) and late-delayed (months to several years postirradiation) injury. Acute radiation injury leads to increased capillary permeability and vasodilatation, manifested radiologically as mild transient vasogenic edema. Early-delayed radiation injury leads to vasogenic edema and demyelination that is often irreversible. Typical radiologic findings include focal or multiple T2 hyperintense white matter lesions with edema. Also, unlike acute radiation injury, patchy enhancement is often present.400 Symptoms related to acute and early-delayed radiation injury often respond to corticosteroids.402 Radiation necrosis is seen in the setting of radiation therapy or more commonly with combination chemoradiation therapy and occurs in 5% to 24% of patients undergoing brain radiation therapy. Radiation necrosis typically presents as a solid or ring-shaped contrastenhancing mass with central necrosis occurring at the site of the original treated neoplasm, without or with associated surrounding vasogenic edema.202 Other less common patterns include: (1) multiple enhancing lesions, (2) lesions in the contralateral hemisphere, (3) lesions arising remotely from a primary cerebral site, and (4) subependymal lesions. Prototypical patterns of enhancement include “soap bubble” and “Swiss cheese” patterns, but linear and nodular enhancement can also be seen. Radiation necrosis may therefore present with the same imaging characteristics as recurrent brain tumor. Unfortunately, distinguishing radiation necrosis from recurrent brain tumor based only on imaging morphology is often not possible, and advanced neuroimaging techniques are often needed, as discussed in the next section.261

Brain tumor imaging continues to evolve, particularly MRI techniques and PET imaging. Conventional MRI sequences provide anatomic and morphologic information, whereas advanced techniques can provide physiologic, metabolic, cellular, functional, and cytoarchitectural data. This additional biological information has greater clinical impact with regard to tumor diagnosis and grading, prognostication and treatment planning, response monitoring, and posttreatment surveillance— specifically, differentiation of recurrent or progressive disease (PD) from radiation change (RC). The most commonly employed advanced MRI techniques are perfusion imaging (including permeability imaging), DWI (including diffusion tensor imaging [DTI]), MRS, and blood oxygen level–dependent (BOLD) functional MRI (fMRI). Many of these techniques can be performed in a single MR examination, providing a wealth of multiparametric data.

MR Perfusion and Permeability Imaging In brain tumors, perfusion and permeability MRI proposes to measure tumor angiogenesis and capillary permeability, both of which are important biological markers of malignancy grading and prognosis, particularly in gliomas.51 T2*-weighted first-pass dynamic susceptibility contrast (DSC) MRI is the most widely used technique to measure tumor perfusion, typically quantified by the relative cerebral blood volume (rCBV). rCBV measurements assume the absence of recirculation and contrast material leakage, which is violated in high-grade leaky tumors. To avoid under- or overestimation of rCBV, correction techniques are typically employed, specifically use of a preloading  dose of contrast, dual-echo pulse sequences, baseline subtraction,  and model-based permeability compensation methods. Significant  differences in rCBV values have been found depending on the technical parameters and correction algorithms used, which limits a generalized application of rCBV threshold values.292 In clinical practice, institution-specific rCBV thresholds or normalized rCBV ratios are often employed. Another important limitation of DSC perfusion imaging is susceptibility artifact, particularly problematic with cortical or skull base lesions or postoperative lesions with blood products or overlying craniotomy.204 Several studies have shown a statistically strong correlation between tumor rCBV and astrocytoma grading.13,190,217,374 In general the maximal tumor rCBV tends to increase with increasing astrocytoma grade209 (Fig. 9-67). In clinical studies, 95% to 100% sensitivity has been

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

245

B PERMEABILITY

D

C FIG 9-67  Anaplastic astrocytoma. A, Contrast-enhanced T1-weighted MRI demonstrates nodular enhancing left frontal lesion (short arrow). B, DSC MR perfusion rCBV map shows increased rCBV (short arrow), suggestive of high-grade neoplasm. C, DCE T1-weighted permeability Ktrans map demonstrates increased permeability (short arrow). D, Time-concentration permeability curve demonstrates typical curve morphology for high-grade neoplasm with steep upslope in the initial vascular phase (long arrow) followed by plateau (red line, lesion; green line, contralateral normal white matter). Biopsy confirmed anaplastic astrocytoma.

reported for differentiating high-grade glioma from low-grade glioma when rCBV thresholds of 1.5 or 1.75 have been applied.209,217 How­ ever, in the same studies, specificity was relatively low, ranging from 57.5% to 69%, although this may be partly explained by spatial heterogeneity of neoplasm with potential histologic sampling bias.204 This theory is supported by study results suggesting an independent prognostic value of perfusion MR, with rCBV greater than 1.75 being an independent predictor of progression in both low-grade and highgrade gliomas.209,218 Furthermore, significant increase in rCBV has been identified in transforming low-grade gliomas up to 12 months before new enhancement is seen, suggesting that rCBV may predict malignant transformation.78 However, a few notable tumors demonstrate elevated rCBV independent of grade, namely PA, oligodendroglioma, and choroid plexus

tumors.53 Primary CNS lymphoma (PCL) tends to show elevated rCBV but not to the degree of glioblastoma.306 Metastatic lesions demonstrate variable, typically elevated, rCBV dependent on primary cell of origin. rCBV alone is, however, not a reliable method to distinguish between lymphoma, metastatic lesions, or high-grade glioma. Peritumoral rCBV can be useful in differentiating metastatic tumors from primary gliomas, with more elevated rCBV levels in glioma owing to infiltrative neoplasm.208 In addition, differentiation of neoplasm from nonneoplastic enhancing lesions can be improved. Specifically, neoplasm can be distinguished from a tumefactive demyelinating lesion (TDL), an occasional neoplasm mimic, by relative diminished rCBV within a TDL compared to high-grade neoplasm317 (Fig. 9-68). DSC perfusion imaging has also proven to be a sensitive technique in differentiating recurrent or progressive neoplasm from radiation

http://radiologyebook.com

246

PART II  CT and MR Imaging of the Whole Body

A

B

P PERFUSION

10 0 –10 –20 –30 –40 –50 –60 –70 –80 –90 –100 –110

D

25

75

100

125

150

175

200

C FIG 9-68  Tumefactive demyelinating lesion. Axial contrast-enhanced T1-weighted MRI (A) demonstrates a discontinuous peripherally enhancing lesion in the left middle cerebellar peduncle, with corresponding hyperintense T2 FLAIR signal (B). C, DSC MR perfusion rCBV map demonstrates rCBV similar to contralateral cerebellar peduncle. D, Perfusion time-signal intensity curve shows overlap of the lesional (yellow) and contralateral normal rCBV (pink), confirming qualitative assessment. Biopsy confirmed tumefactive demyelinating lesion (TDL).

change.18,19,184,244,250,284,376 High rCBV typical of recurrent high-grade neoplasm (see Fig. 9-78) can be differentiated from the typically low rCBV values of radiation necrosis (Fig. 9-69). Most studies apply either absolute rCBV or rCBV ratio (typically to contralateral normal gray or white matter) thresholds with varying sensitivity and specificity depending on chosen threshold. Studies have also suggested DSC

perfusion imaging using rCBV to be superior to FDG PET for this purpose.184,315 Without the application of correction methods, other quantitative perfusion parameters can be obtained, namely percentage signal recovery (PSR) and peak height (PH), which have been shown to be useful independent parameters in differentiating glioma from metastasis and

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

Initial study

247

C

B 220 200

PERMEABILITY

180 160 140 120 100 80 60 40 20 0 -20

E

F

D

10 month F/U

FIG 9-69  Radiation necrosis following resection and chemoradiation therapy in a patient with a history of right frontotemporal anaplastic oligodendroglioma. A, Contrast-enhanced T1-weighted MRI (initial study) demonstrates new ill-defined enhancement in right centrum semiovale (short arrow). Axial 18F FDG PET/MRI (B) and PET (C) images show no significant lesional FDG uptake. D, DSC MR perfusion rCBV map demonstrates rCBV similar to contralateral white matter. E, Time-concentration permeability curve demonstrates gradual upslope in the initial vascular phase, with continued progressive gradual upslope typical of radiation change (red line, lesion; green line, contralateral normal white matter). F, Ten-month follow-up contrastenhanced T1-weighted MRI shows near-complete resolution of enhancement typical of radiation necrosis.

recurrent metastatic disease from radiation necrosis.18 Additionally, PSR has recently been shown to better distinguish between glioblastoma, solitary metastasis, and PCL than rCBV.237 Permeability MRI or dynamic contrast-enhanced (DCE) MRI uses dynamic steady-state T1-weighted imaging after a gadolinium contrast bolus. High-grade intraaxial malignancy, namely glioma and metastases, are dependent on neovascularity for growth, with formation of immature and “leaky” blood vessels resulting in increased capillary permeability.167 The most commonly measured and studied quantitative permeability parameter is Ktrans, a measure of the volume transfer of contrast between the blood plasma and extracellular space.402 Ktrans has been shown to correlate strongly with tumor grade,231,289,328 although threshold values are difficult to apply owing to variability in technique and the requirement of a complex model-based analysis. Therefore, various nonmodel-based semiquantitative indices have been suggested, which are derived from the time-signal intensity curve. The maximum slope of the initial vascular phase (MSIVP) has shown promise in differentiating radiation change (RC) and recurrent/ progressive neoplasm (PD) with RC, demonstrating lower MSIVP

values (see Fig. 9-69) than PD, which typically demonstrates a steep upslope in the initial vascular phase owing to permeable blood vessels268 (see Fig. 9-67). Using similar diagnostic criteria, high-grade neoplasm can often be differentiated from nonneoplastic lesions based on permeability imaging.167 Additionally, preliminary studies have demonstrated correlation between Ktrans and histologic grade for atypical and typical meningiomas, a distinction that has not been demonstrated with rCBV.412 In summary, perfusion and permeability MRI are robust techniques that measure tumor vascularity, typically using DSC and DCE T1 MRI techniques, respectively. Perfusion imaging using absolute rCBV or rCBV ratios, and permeability imaging, typically using Ktrans, aids primarily in tumor grading and prognosis, differentiation of neoplasm from nonneoplastic lesions, and distinguishing between recurrent progressive neoplasm and radiation change. Semiquantitative indices through DSC (peak height, percentage signal recovery) and DCE T1 (MSIVP) MR techniques are additional parameters that can be helpful in differentiating tumor types and distinguishing between recurrent primary or secondary neoplasm and radiation change.

http://radiologyebook.com

248

PART II  CT and MR Imaging of the Whole Body

Normal

Cr I:49.5

NAA I:58.1

Tumor

Cho I:76.8

Cho I:36.8 4

6 Cr2 I:73.6 4

2

Cr2 I:33.9

Cr I:35.9

NAA I:21.2

2

0

0

4

3

2

1

4

3

2

1

FIG 9-70  MRS. Spectrum from normal white matter on the left (Normal). Spectrum from patient with anaplastic astrocytoma (Tumor), which demonstrates typical findings of increased choline (Cho) and decreased creatinine (Cr) and N-acetylaspartic acid (NAA). Cho/NAA ratio was 3.6, indicative of high-grade neoplasm.

Magnetic Resonance Spectroscopy MRS is a well-established methodology routinely using the proton (1H) nucleus, with utility in brain tumor imaging. In MRS, the resonance signal is used to generate a frequency spectrum (expressed in parts per million [ppm]) of a sampled voxel, reflecting the intravoxel cellular metabolism. MRS can be performed using either single-voxel or multivoxel techniques, with multivoxel techniques allowing for sampling of larger three-dimensional (3D) volumes, optimizing evaluation of metabolically heterogeneous tumors.160 N-acetylaspartic acid (NAA), creatinine (Cr), and choline (Cho) are the most prevalent brain metabolites. NAA is a marker of neuronal and glial cell viability and density, demonstrating the highest peak amplitude in normal brain. Cr is involved in energy-dependent processes of cellular metabolism in several cell types, typically with the second highest peak amplitude. Cho reflects the metabolism of the cellular membrane, usually with the third highest peak amplitude. Brain tumors generally have decreased NAA and Cr levels and increased Cho levels, leading to increased Cho/NAA and Cho/Cr ratios36,206 (Figs. 9-70 and 9-71C). Other important metabolic peaks occur in the range of 0.9 to 1.3 ppm and reflect mobile lipids and lactate. Lipids and lactate are not normal constituents of brain metabolism. The regions of lactate most likely represent areas of hypoxic but viable tumor, whereas the regions of mobile lipids represent necrotic nonviable tissue.51 Elevated lactate levels can be seen in both low- and high-grade tumors,151 but elevated lipid levels are typically only seen with high-grade (necrotic) neoplasm, both primary and secondary48,51 (Fig. 9-72C; see Fig. 9-71C). Using short echo time (TE) spectra, elevated myoinositol levels can also be seen, reflecting increased glial cells. These have been reported particularly with grade II gliomas49,144 and in patients with glioblastoma cerebri.346 Also, elevated alanine levels in a tumor are specific for meningiomas.27 In astrocytomas, several studies (but not all) have suggested an association between tumor grade and Cho levels, with higher-grade tumors having greater Cho concentrations.12,118,160 Statistically signifi-

cantly higher Cho/Cr and Cho/NAA ratios can differentiate low- and high-grade astrocytoma.209 Cho/Cr is the most frequently used ratio for glioma grading, with varying cutoff values of 2.0 to 2.5.27 MRS also better distinguishes high- from low-grade oligodendroglioma as compared to DSC perfusion MR using rCBV.327,366,411 However, whereas elevated Cho levels are typically seen in nonnecrotic high-grade brain tumors (anaplastic astrocytoma, glioblastoma, primary CNS lymphoma), necrotic glioblastoma and metastases can demonstrate low levels of all metabolites and increased lipids.159 Lesional MRS features do not reliably differentiate high-grade glioma from metastases, but evaluation of peritumoral regions can reveal elevated Cho in glioma versus metastases owing to the infiltrative nature of high-grade glioma.60,102 MRS can also be useful in differentiating tumors from nonneo­ plastic lesions, namely abscesses, encephalitis, cortical dysplasia, and infarct.4,5,44,87,235,253,309,310,324,401,403 Published MRS findings of nonneoplastic lesions typically demonstrate normal or low Cho levels, distinguishing them from elevated Cho levels characteristic of high-grade neoplasm. MRS studies show a sensitivity of 79% and specificity of 77% for a Cho/NAA ratio greater than 1 as an indicator of neoplasm.60 Intracranial abscesses usually have decreased NAA, Cho, and Cr levels as well as increased levels of compounds not typically seen in neoplasia (i.e., amino acids, lactate, alanine, acetate, pyruvate, succinate).4 However, MRS does not confidently differentiate demyelinating lesions from neoplasm, insofar as demyelinating lesions often present with elevated Cho and decreased NAA levels as well as increased lactate levels.345 The ability of MRS to predict survival has been evaluated in both adult and pediatric brain tumor populations. High Cho/NAA ratio and high levels of lipid and lactate have been negatively correlated with survival time in both adult and pediatric tumors.54,74,239,347,404 MRS can also be helpful in identifying infiltrate tumor in peritumoral white matter, which can differentiate high-grade glioma from metastases208 as well as potentially aid in optimizing tumor coverage for radiation therapy. MRS has also proven useful in biopsy guidance, both

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

249

B

A I: Integral P: Position 0.5

Cho I:5.01 P:3.29

0.4

Cr I:3.46 P:3.10

0.3

Ins Cr2 0.2 I:1.84 I:2.56 P:3.92 P:3.60

NAA I:1.50 Lac P:2.04 Lac I:0.433 I:0.0550 P:1.38 P:1.35

0.1 0.0

ppm 4

3

2

1

C FIG 9-71  High-grade glioma most likely representing an anaplastic astrocytoma in a 10-year-old child with MRS correlation. A, Axial turbo FLAIR image demonstrates a rounded and slightly lobulated well-circumscribed mildly T2-hyperintense mass in the right temporal lobe, with surrounding edema. B, Following IV gadolinium, axial T1-weighted imaging demonstrates avid contrast enhancement within the tumor, with only a few small ill-defined areas of nonenhancement. C, Single-voxel MRS interrogation of the right temporal lobe tumor demonstrates depression of the NAA peak, with marked elevation of the choline peak, consistent with the presence of active tumor growth. Note also the slight elevation of the lactate peak, indicating a small amount of central necrosis within this tumor. This suggests that the tumor is transitioning toward glioblastoma.

independently and in combination with other advanced neuroimaging techniques.94,138,135,153,243 Lastly, MRS has been reported to accurately differentiate recurrent or progressive tumor from radiation change, specifically when using Cho/NAA and Cho/Cr ratios as well as relative increase in Cho peak (compared to contralateral normal brain).94,305,362,409,421 In irradiated

gliomas, abnormal increase in Cho peak greater than 50% of the contralateral brain or Cho/NAA ratio greater than 2 has moderate to high sensitivity (64%-95%) and high specificity (82%-100%) for identifying active tumor.9,305,393 Radiation change typically shows decreased NAA with variable changes in Cho and Cr intensities over time,321 with Cho levels variably elevated early and both Cho and Cr diminished in

http://radiologyebook.com

250

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 9-72  Radiation necrosis and recurrent tumor following irradiation of malignant tumor of thalami and brainstem. A, Axial T2-weighted MRI demonstrates diffuse hyperintensity of both thalami, likely representing edema (arrow). There is an ovoid heterogeneously hyperintense mass behind and continuous with the left thalamus, with displacement of the internal cerebral veins to the right of the midline. B, Axial T2-weighted image at a slightly higher level than A. The voxel for MRS interrogation is placed in the left posterior thalamic mass. C, MRS interrogation of the left posterior thalamic region demonstrates depression of the NAA peak (central arrow) with elevation of the choline peak (left arrow), indicating the presence of active tumor growth. The other markedly elevated and abnormal peak (right arrow) is lactate, likely representing necrosis of tumor secondary to radiation therapy.

delayed radiation necrosis. However, mixed radiation necrosis and neoplasm is not reliably differentiated with MRS, especially without multivoxel technique.101,177,330,329,351,378 In summary, MRS can serve as a useful adjunctive MR technique, particularly in tumor grading, biopsy guidance, and differentiation of radiation change from progressive or recurrent neoplasm.

Diffusion-Weighted Imaging DWI uses the microscopic mobility of water, classic brownian motion, to obtain biological MR data. DWI is typically acquired through echo planar gradient echo T2*-weighted pulse sequences or less commonly using spin echo techniques. The primary clinical applications of DWI

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms in brain tumor imaging are assessing tumor grade and cellularity, differentiating tumor type, evaluating peritumoral edema, assessing treatment response, and characterizing postoperative injury.51 Also, unlike other advanced MR techniques, DWI is a rapidly acquired and routinely performed MR sequence. DWI enables calculation of apparent diffusion coefficient (ADC), which represents the magnitude of water diffusion inside voxels.402 Generally speaking, diffusivity is inversely correlated with cellularity, with reduced diffusivity (diminished ADC signal) often correlating with higher-grade (hypercellular) tumor.275 Several reports have shown that glioma grade correlates inversely with minimum ADC, likely on the basis of increasing cellularity and tumor grade.38,50,132,185,197,375 However, preoperative assessment of glioma grade using only ADC is often unreliable owing to spatial histologic tumor heterogeneity. ADC values have also been correlated with cellularity and grade of nonglial brain tumors,51 as well as pediatric brain tumors of varying types50 (Fig. 9-73). When compared with gliomas, lymphomas are shown to

251

have lower ADC, likely owing to the characteristic marked hypercellularity of lymphoma132 (Fig. 9-74). Additionally, atypical (WHO grade II) and malignant (grade III) meningiomas often demonstrate lower ADC values compared with typical benign meningiomas (grade I).105 A decrease in ADC and normalized ADC (normalized to normal white matter) on follow-up imaging may suggest dedifferentiation to higher tumor grade.265 ADC maps have also recently been shown to differentiate high- versus low-grade neoplasms for supratentorial tumors presenting in the first year of life, with significant negative correlation between ADC values and tumor grade.50 DWI has also proven helpful in distinguishing between the three most common pediatric posterior fossa neoplasms: ependymoma, juvenile pilocytic astrocytoma (JPA), and medulloblastoma, with medulloblastoma demonstrating diminished ADC compared to ependymoma and JPA (Fig. 9-75). However, more recent studies have demonstrated overlap in ADC between ependymoma/JPA and ependymoma/medulloblastoma. Atypical teratoid/rhabdoid tumor (AT/RT) demonstrates diminished

A

B

C

D

E

F

FIG 9-73  Primitive neuroectodermal tumor (PNET) in a 4-year-old patient with heterogeneous left hemispheric mass on T2-weighted (A), T2 FLAIR (B), T1-weighted (E), and contrast-enhanced T1-weighted MRI (F). Hyperintense signal corresponding to the enhancing portion of tumor on DWI images (C) with corresponding diminished ADC (D). Findings are suggestive of high-grade neoplasm. Surgical resection revealed PNET (WHO grade IV).

http://radiologyebook.com

252

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 9-74  Primary CNS lymphoma. A, DWI; B, ADC map; and contrast-enhanced axial (C) and parasagittal (D) T1-weighted MRIs. Left periatrial enhancing lesion (blue arrows in C and D) demonstrates hyperintense signal on DWI (red arrow in A) with corresponding diminished signal on the ADC map (orange arrow in B). Decreased diffusivity is typical of hypercellular neoplasm such as primary CNS lymphoma.

ADC values overlapping with medulloblastoma—both WHO grade IV tumors.170,198,340 Diffusion imaging has also been used in monitoring treatment  and predicting response. Numerous preclinical studies support the hypothesis that serial change in ADC may be a biomarker of treatment response.58,59,139,150,246,371 Increasing ADC values early in the posttreatment period of primary gliomas, metastases, and medulloblastoma have been correlated with positive clinical response and reduction in lesion size.161,240,353,391 Additionally, more recent methods, namely “functional diffusion mapping,”252 can provide voxel-by-voxel data, which is advantageous given the typical spatial heterogeneity of highgrade tumors such as glioma.140 DWI is useful in evaluating postoperative injury and distinguishing subsequent enhancing infarcts from recurrent neoplasm. After brain

tumor resection, acute cellular damage is likely to occur for a variety of reasons, such as direct surgical trauma, retraction and vascular injury, and devascularization of tumor, manifesting as regions of  diffusion restriction. Areas of postoperative diffusion restriction invariably evolve to heterogeneously enhancing lesions, which can be misinterpreted as tumor recurrence.363 If interpreted in the context of the immediate postoperative DWI, new areas of enhancement after surgery can be more accurately characterized, avoiding unnecessary treatment and associated toxicity. DTI measures the effects of tissue microstructure on the random motion of water molecules in the brain. Fractional anisotropy (FA), the most widely reported measure in DTI, represents the eccentricity of the displacement of water molecules. Because the most relevant factor affecting fractional anisotropy is the intravoxel orientation

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

A

B

253

C

FIG 9-75  Medulloblastoma. A, Pediatric patient with enhancing fourth ventricular mass on contrast-enhanced T1-weighted MRI. On DWI (B) increased signal within the tumor is seen, with corresponding diminished ADC (C). Findings are most suggestive of medulloblastoma (WHO grade IV).

coherence of white matter tracts, DTI allows performance of white matter tract mapping. DTI is typically evaluated through quantitative parametric maps displayed in grayscale (FA maps), color maps showing principal orientation of diffusion for each voxel, and 3D white matter tract maps (tractography). DTI is becoming more commonly used for preoperative planning as well as more recently in the evaluation of peritumoral white matter.21,68,255,303,372 Several studies have demonstrated the utility of DTI in evaluating peritumoral white matter, specifically differentiating infiltrated white matter from pure vasogenic edema. This can be useful in differentiating primary and metastatic brain tumors, based on the premise that glioblastoma typically has an infiltrative growth pattern invading surrounding tissue, whereas metastatic disease is typically encapsulated, surrounded by vasogenic edema.42 FA has been shown to be more often decreased in infiltrated peritumoral white matter (surrounding glioma) as compared to vasogenic noninfiltrative edema (surrounding metastases).83,174,316,349,394 DTI is, however, primarily useful in preoperative planning. Directional DTI color maps can indicate whether a mass is displacing, infiltrating, or destroying the main white matter tracts171 (Fig. 9-76). MR tractography can be used to virtually dissect functionally critical white matter tracts, such as the corticospinal tract and arcuate fasciculus, enabling the neurosurgeon to identify and preserve the tract during resection.207

Functional MRI fMRI has tremendous potential in preoperative evaluation, combining anatomic detail with physiologic information to demonstrate spatial relationships between functionally eloquent cortex and intracranial pathology. T2* BOLD MR sequences are used to image local hemodynamic changes supporting cortical neuronal activity during performance of tasks.30,72,215,300 Presurgical fMRI mapping of eloquent cortex involved in language, sensory, motor, and visual function and processing is acquired through appropriate tasks. Mapping of motor and sensory functions is relatively straightforward, primarily because of reliable surface cortical landmarks. However, localization of language eloquent cortex is far more complex and the most critical factor in planning the cortical trajectory and resection of a hemispheric cerebral mass.122,165,256,319

Language localization is typically acquired through varying tasks, the most common and reliable of which are verbal fluency with word generation, verb generation, sentence comprehension, and passive listening with the combined use of multiple tasks improving sensitivity and specificity of the exam. fMRI has been shown to have a sensitivity of 81% to 92% for mapping language areas in comparison with intraoperative electrocortical stimulation (ECS), the neurosurgical gold standard.109,339 However, fMRI has limitations with high-grade neoplasms that alter local hemodynamics owing to associated neoangiogenesis, as well as with high- and low-grade gliomas through tumor infiltration and associated distortion.122 fMRI is often combined with DTI in preoperative mapping and has changed the way neurosurgeons look at patients with a mass in the dominant hemisphere (see Fig. 9-76). The aims of presurgical mapping are fourfold: (1) to determine hemispheric dominance, (2) to select candidates for awake craniotomy, (3) to guide the intraoperative surgical procedure, and (4) to assess the risk of postoperative neurologic deficits by measuring the distance between the margin of the lesion resection and critical functional structures.30

Positron Emission Tomography PET imaging has demonstrated a role in brain tumor imaging, providing valuable metabolic information. PET imaging is typically performed with fluorine-18 (18F) fluorodeoxyglucose (FDG), with higher-grade tumor typically showing increased uptake.90 However, FDG PET has some well-demonstrated limitations, primarily because of the high level of normal brain uptake. Several non-FDG tracers have been investigated, particularly amino acid tracers, which have proven advantageous owing to excellent tumor uptake, even in low-grade tumors (WHO grade II), in combination with a comparatively low level of normal brain uptake.131,181 Similar to advanced MR techniques, PET imaging contributes important information to overall assessment of tumor diagnosis and grade, prognostication/treatment planning, response monitoring, and posttreatment surveillance, specifically differentiation of recurrent or progressive disease from radiation change. 18

F FDG PET

18

F FDG has many well-established oncologic applications. 18F FDG is actively transported across the blood-brain barrier into tumor cells,

http://radiologyebook.com

254

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F

FIG 9-76  Preoperative task-based (tb)-fMRI and DTI tractography. A, Contrast-enhanced T1-weighted MRI shows heterogeneous peripherally enhancing left frontotemporoparietal lesion compatible with glioblastoma. tb-fMRI with expressive language and motor paradigms demonstrates no involvement of Broca’s area (B) or motor activity clusters (C). D, Receptive language activity paradigm tb-fMRI image shows anterior displacement of the left superior temporal gyrus (primary auditory cortex) by the lesion. E, Directional DTI map demonstrates relationship of lesion to corona radiata (corticospinal tract) and adjacent association fibers, which are minimally displaced but not invaded. F, Intraoperative contrast-enhanced T1-weighted MRI following resection demonstrates intact superior temporal gyrus. Patient experienced mild transient receptive aphasia postoperatively that subsequently resolved.

and 18F FDG uptake in high-grade tumors, particularly gliomas, has been well demonstrated.90 Although limited by heterogeneous data sampling, several studies have also shown correlation between the level of FDG uptake and tumor grade.3,31,90,305 In general, low-grade and high-grade gliomas demonstrate different degrees of FDG uptake, with low-grade and high-grade glioma demonstrating FDG uptake similar to white matter and gray matter, respectively. FDG PET imaging can also provide important prognostic information. Although there have been mixed results, several studies have correlated degree of FDG avidity in malignant glioma3,90,305 and primary pediatric brain tumors31 inversely with median and 1-year survival rates. Also, high FDG uptake in a previously known low-grade tumor implies anaplastic transformation22,89,285,291 (Fig. 9-77). Several studies have evaluated the utility of FDG PET in differentiating recurrent or progressive disease from radiation change (Fig. 9-78) with great variability in performance, particularly regarding malignant glioma. Earlier reports had shown sensitivity of 81% to 86%

and specificity of 40% to 90%. A recent meta-analysis demonstrated a summary sensitivity and specificity of 77% and 78% when high- and low-grade gliomas were included and 79% and 70% when only highgrade gliomas were evaluated.276,290 High normal brain FDG uptake, a well established FDG PET limitation, provides diagnostic challenges when evaluating small tumors, tumors near cortex, or low-grade tumors. To avoid diagnostic error, two interpretation techniques have been suggested: (1) coregistration with MRI and (2) delayed PET imaging. Coregistration of FDG PET with MRI greatly improves the performance of FDG PET. In general, any area of FDG uptake corresponding to suspicious MR findings that is greater than the expected background level of the adjacent brain should be considered viable neoplasm, even if less than or equal to cortical uptake. Delayed PET imaging 3 to 8 hours after administration has been shown to increase tumor background uptake ratio, improving conspicuity of tumor detection. Standard uptake value (SUV), a semiquantitative parameter of metabolic activity, is often difficult to

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms

Fibrillary Astrocytoma

A

D

P

C

B

255

12 Month F/U

F

E

FIG 9-77  Astrocytoma, anaplastic transformation. Frontoparietal infiltrative lesion on T2 FLAIR (arrow in A) without associated enhancement (B), suggestive of low-grade glioma. Biopsy yielded fibrillary astrocytoma (WHO grade II). A 12-month follow-up T2 FLAIR (C) demonstrates progressive left hemispheric hyperintense white matter signal with new area of nodular enhancement on contrast-enhanced T1 MRI (arrow in D). Axial 18 F FDG PET/CT PET (E) and PET (F) images demonstrate focal FDG uptake corresponding to frontal nodular enhancing lesion (arrows). Findings are suggestive of anaplastic transformation. Biopsy yielded anaplastic astrocytoma.

reliably use as an independent parameter, so SUV ratios are commonly used, including ratios of tumor to contralateral gray matter, white matter, or ipsilateral adjacent brain.56,275,279,367,410 In summary, although there are certain advantages of newer radiotracers, 18F FDG remains the most commonly used radiotracer in brain tumor imaging because of its overall acceptable diagnostic accuracy, low cost, widespread availability, and favorable half-life. FDG PET imaging of brain tumors, particularly if coregistered with MRI, can provide useful metabolic information, especially in high-grade tumors.

Non-FDG PET To circumvent the shortcomings of FDG PET, several additional radiotracers have been studied for brain tumor PET imaging. Amino acid analogue radiotracers are advantageous owing to diminished background brain uptake and avidity for both low- and high-grade tumors.

The most widely studied amino acid tracer is carbon-11 (11C) methionine (MET), which has demonstrated improved performance compared to 18F FDG.152 The overall sensitivity of MET PET for malignant gliomas (both low- and high-grade) ranges from 76% to 95% and for low-grade gliomas, 65% to 85%.152,278 Multiple studies have demonstrated improved tumor delineation with MET PET imaging compared to anatomic imaging such as CT or MRI.26,125,251,259,304 MET uptake has also been shown to extend beyond the margins of enhancement in high-grade glioma, with correlation between resected tumor size and MET uptake suggesting a role for MET PET in operative planning.178 MET PET has also proven valuable in grading of brain neoplasms  as well as differentiation of neoplasm from nonneoplastic processes.84,85,166,181,221,258 In a large series of 196 patients with suspected low-grade glioma, MET PET was able to accurately distinguish between high-grade gliomas, low-grade gliomas, and chronic or subacute

http://radiologyebook.com

256

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

FIG 9-78  Recurrent glioblastoma in a patient with a history of left temporal glioblastoma. A, On contrastenhanced T1-weighted MRI following resection and chemoradiation therapy, a recurrent heterogeneously enhancing lesion within and adjacent to the resection cavity is seen. B to E, Multiparametric 18F FDG PET/MRI performed to differentiate between progressive/recurrent neoplasm and radiation change. Axial PET (B) and PET/MRI (C) images demonstrate FDG uptake above background along the peripheral (enhancing) portion of the lesion. DSC MR perfusion rCBV map coregistered with contrast-enhanced T1-weighted MRI (D) demonstrates peripheral increased lesional rCBV in a pattern similar to FDG uptake (B and C). E, DCE T1-weighted MRI Ktrans map demonstrates corresponding increased permeability. Findings are suggestive of recurrent/progressive glioblastoma.

nonneoplastic lesions.258 A more recent study comparing MET PET, FDG PET, and 11C-choline PET showed that when using a tumor/ normal brain uptake ratio, diffuse astrocytoma (grade II), anaplastic astrocytoma (grade III), and glioblastoma (grade IV) could be differentiated with all three radiotracers, but only MET PET could distinguish same-grade gliomas of different types (oligodendroglioma vs. astrocytoma).181 Similar to FDG, MET uptake has prognostic implications correlating with length of patient survival. MET PET is also an accurate tool in differentiating radiation change from progressive or recurrent disease, with several studies demonstrating an advantage over FDG PET.63,88,180,183,228,302,398 A recent meta-analysis demonstrated similar summary sensitivity for FDG PET (79%) and MET PET (70%) but with increased specificity for MET PET (93%) compared to FDG PET (70%) in primary glioma detection (grade II-IV).181 Improved specificity with amino acid tracers compared to 18F FDG relates to the nonspecificity of glucose uptake (i.e., inflammation) compared to the more specific uptake of amino acids in tumors. 11 C MET is, however, limited by short tracer half-life (20 minutes), requiring an on-site cyclotron, which precludes generalized use.275

Therefore other novel fluorine-based amino acid tracers have been studied, and those shown to have tumor uptake similar to 11C MET are 18F-fluoroethyltyrosine (FET) and 3,4-dihydroxy-6-18F-fluoro-lphenyalanine (18F FDOPA).23,406 FET PET has similar diagnostic potential to MET PET, with tumor delineation appearing to be identical.22 FET PET has also been shown to reliably differentiate tumor recurrence from radiation change with high sensitivity and specificity.307,322,407 FET PET has demonstrated great promise in prognostication, par­ ticularly with low-grade astrocytoma.110,168,308,320 Recent research has focused on dynamic FET PET imaging, which may provide an advantage over static images in determining prognosis as well as tumor grading. FDOPA PET imaging has also proven advantageous compared to FDG PET in tumor detection but with comparable relatively low specificity.57 FDOPA demonstrates excellent visualization of highand low-grade tumors and is postulated to be beneficial in the posttreatment period, specifically in evaluating for recurrent low-grade tumor.56 18 F fluorothymidine (FLT) is an additional amino acid tracer that serves as a biomarker for cell proliferation. Hypoxic PET agents

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms such as 18F fluoromisonidazole (18F FMISO) and 18F EF5 [2-(2-nitro1 H-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl)-acetamide] enable quantification of hypoxia in tumors. Hypoxia in non-CNS tumors has been correlated with aggressive tumor phenotype, increased metastatic potential, poor response to radiation and chemotherapy, and overall poor clinical outcome.35,34,129,156,169 FLT and hypoxic PET agents show promise, particularly in evaluation of prognosis and treatment response monitoring. Decreased FLT uptake early in treatment has been suggested to represent a treatment response, which correlated with survival in preliminary studies.56

Hybrid PET/MRI Brain tumor evaluation with PET imaging is greatly improved by fusion to anatomic images, typically CT in current clinical practice. As discussed earlier, it is well demonstrated that interpretation of brain PET images benefits from coregistration with conventional MRI, which improves sensitivity and specificity.410 Hybrid PET/MRI is an emerging technique that allows for improved lesion and anatomic coregistration, which improves the accuracy of PET or MRI alone. Hybrid PET/MRI also allows for the performance of adjunctive advanced MRI sequences, serving as a “one-stop-shop” imaging evaluation and providing a wealth of multiparametric information (see Fig. 9-78). It is easy to envision a role for multiparametric hybrid PET/MRI in neurooncologic imaging, with clinical experience growing in the last decade.275

REFERENCES 1. Abdullah ND, Mathews VP: Contrast issues in brain tumor imaging. Neuroimaging Clin N Am 9(4):733–749, 1999. 2. Ahmadi H, Larsson EM, Jinkins JR: Normal pituitary gland: Coronal MR imaging of infundibular tilt. Radiology 177(2):389–392, 1990. 3. Alavi JB, Alavi A, Chawluk J, et al: Positron emission tomography in patients with glioma. A predictor of prognosis. Cancer 62(6):1074– 1078, 1988. 4. Al-Okaili RN, Krejza J, Wang S, et al: Advanced MR imaging techniques in the diagnosis of intraaxial brain tumors in adults. Radiographics 26(Suppl 1):173–189, 2006. 5. Al-Okaili RN, Krejza J, Woo JH, et al: Intraaxial brain masses: MR imaging-based diagnostic strategy–Initial experience. Radiology 243(2):539–550, 2007. 6. Albright AL, Guthkelch AN, Packer RJ, et al: Prognostic factors in pediatric brain-stem gliomas. J Neurosurg 65(6):751–755, 1986. 7. Alshail E, Rutka JT, Becker LE, et al: Optic chiasmatic-hypothalamic glioma. Brain Pathol 7(2):799–806, 1997. 8. Amundsen P, Dugstad G, Syvertsen AH: The reliability of computer tomography for the diagnosis and differential diagnosis of meningiomas, gliomas, and brain metastases. Acta Neurochir (Wien) 41(1–3):177–190, 1978. 9. Ando K, Ishikura R, Nagami Y, et al: Usefulness of Cho/Cr ratio in proton MR spectroscopy for differentiating residual/recurrent glioma from non-neoplastic lesions. Nihon Igaku Hoshasen Gakkai Zasshi 64(3):121–126, 2004. 10. Aoki S, Okada Y, Nishimura K, et al: Normal deposition of brain iron in childhood and adolescence: MR imaging at 1.5 T. Radiology 172(2):381–385, 1989. 11. Armington WG, Osborn AG, Cubberley DA, et al: Supratentorial ependymoma: CT appearance. Radiology 157(2):367–372, 1985. 12. Arnold DL, Shoubridge EA, Villemure JG, et al: Proton and phosphorus magnetic resonance spectroscopy of human astrocytomas in vivo. Preliminary observations on tumor grading. NMR Biomed 3(4):184– 189, 1990. 13. Aronen HJ, Gazit IE, Louis DN, et al: Cerebral blood volume maps of gliomas: Comparison with tumor grade and histologic findings. Radiology 191(1):41–51, 1994.

257

14. Atlas SW: Adult supratentorial tumors. Semin Roentgenol 25(2):130– 154, 1990. 15. Atlas SW, editor: Magnetic resonance imaging of the brain and spine, Philadelphia, 2002, Lippincott, Williams and Wilkins. 16. Atlas SW, Grossman RI, Gomori JM, et al: Hemorrhagic intracranial malignant neoplasms: Spin echo MR imaging. Radiology 164(1):71–77, 1987. 17. Baker HL, Jr, Houser OW, Campbell JK: National Cancer Institute study: Evaluation of computed tomography in the diagnosis of intracranial neoplasms. I. Overall results. Radiology 136(1):91–96, 1980. 18. Barajas RF, Chang JS, Segal MR, et al: Differentiation of recurrent glioblastoma multiforme from radiation necrosis after external beam radiation therapy with dynamic susceptibility-weighted contrastenhanced perfusion MR imaging. Radiology 253(2):486–496, 2009. 19. Barajas RF, Chang JS, Sneed PK, et al: Distinguishing recurrent intra-axial metastatic tumor from radiation necrosis following gamma knife radiosurgery using dynamic susceptibility-weighted contrastenhanced perfusion MR imaging. AJNR Am J Neuroradiol 30(2):367– 372, 2009. 20. Barnard RO, Scott T: Patterns of proliferation in cerebral lymphoreticular tumours. Acta Neuropathol Suppl (Suppl 6):125–130, 1975. 21. Basser PJ, Mattiello J, LeBihan D: MR diffusion tensor spectroscopy and imaging. Biophys J 66(1):259–267, 1994. 22. Basu S, Alavi A: Molecular imaging (PET) of brain tumors. Neuroimaging Clin N Am 19(4):625–646, 2009. 23. Becherer A, Karanikas G, Szabo M, et al: Brain tumour imaging with PET: A comparison between [18F]fluorodopa and [11C]methionine. Eur J Nucl Med Mol Imaging 30(11):1561–1567, 2003. 24. Becker LE: Pathology of pediatric brain tumors. Neuroimaging Clin N Am 9(4):671–690, 1999. 25. Bentson JR, Steckel RJ, Kagan AR: Diagnostic imaging in clinical cancer management: Brain metastases. Invest Radiol 23(5):335–341, 1988. 26. Bergstrom M, Collins VP, Ehrin E, et al: Discrepancies in brain tumor extent as shown by computed tomography and positron emission tomography using [68Ga]EDTA, [11C]glucose, and [11C]methionine.   J Comput Assist Tomogr 7(6):1062–1066, 1983. 27. Bertholdo D, Watcharakorn A, Castillo M: Brain proton magnetic resonance spectroscopy: Introduction and overview. Neuroimaging Clin N Am 23(3):359–380, 2013. 28. Bilaniuk LT, Zimmerman RA, Littman P, et al: Computed tomography of brainstem gliomas in children. Radiology 134(1):89–95, 1980. 29. Bindal RK, Sawaya R, Leavens ME, et al: Surgical treatment of multiple brain metastases. J Neurosurg 79(2):210–216, 1993. 30. Bizzi A: Presurgical mapping of verbal language in brain tumors with functional MR imaging and MR tractography. Neuroimaging Clin N Am 19(4):573–596, 2009. 31. Borgwardt L, Hojgaard L, Carstensen H, et al: Increased fluorine-18 2-fluoro-2-deoxy-D-glucose (FDG) uptake in childhood CNS tumors is correlated with malignancy grade: A study with FDG positron emission tomography/magnetic resonance imaging coregistration and image fusion. J Clin Oncol 23(13):3030–3037, 2005. 32. Brandes AA, Franceschi E, Tosoni A, et al: MGMT promoter methylation status can predict the incidence and outcome of pseudoprogression after concomitant radiochemotherapy in newly diagnosed glioblastoma patients. J Clin Oncol 26(13):2192–2197, 2008. 33. Brandsma D, Stalpers L, Taal W, et al: Clinical features, mechanisms, and management of pseudoprogression in malignant gliomas. Lancet Oncol 9(5):453–461, 2008. 34. Brizel DM, Scully SP, Harrelson JM, et al: Radiation therapy and hyperthermia improve the oxygenation of human soft tissue sarcomas. Cancer Res 56(23):5347–5350, 1996. 35. Brizel DM, Sibley GS, Prosnitz LR, et al: Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 38(2):285–289, 1997. 36. Bruhn H, Frahm J, Gyngell ML, et al: Noninvasive differentiation of tumors with use of localized H-1 MR spectroscopy in vivo: Initial

http://radiologyebook.com

258

37.

38.

39. 40. 41.

42.

43.

44.

45.

46. 47.

48. 49.

50.

51. 52.

53.

54.

55. 56.

57.

58.

PART II  CT and MR Imaging of the Whole Body experience in patients with cerebral tumors. Radiology 172(2):541–548, 1989. Buetow MP, Buetow PC, Smirniotopoulos JG: Typical, atypical, and misleading features in meningioma. Radiographics 11(6):1087–1106, 1991. Bulakbasi N, Guvenc I, Onguru O, et al: The added value of the apparent diffusion coefficient calculation to magnetic resonance imaging in the differentiation and grading of malignant brain tumors.   J Comput Assist Tomogr 28(6):735–746, 2004. Bullard DE, Cox EB, Seigler HF: Central nervous system metastases in malignant melanoma. Neurosurgery 8(1):26–30, 1981. Bunin GR, Surawicz TS, Witman PA, et al: The descriptive epidemiology of craniopharyngioma. J Neurosurg 89(4):547–551, 1998. Burger PC, Heinz ER, Shibata T, et al: Topographic anatomy and CT correlations in the untreated glioblastoma multiforme. J Neurosurg 68(5):698–704, 1988. Burger PC, Vogel FS, Green SB, et al: Glioblastoma multiforme and anaplastic astrocytoma. Pathologic criteria and prognostic implications. Cancer 56(5):1106–1111, 1985. Butler AR, Horii SC, Kricheff II, et al: Computed tomography in astrocytomas. A statistical analysis of the parameters of malignancy and the positive contrast-enhanced CT scan. Radiology 129(2):433–439, 1978. Butzen J, Prost R, Chetty V, et al: Discrimination between neoplastic and nonneoplastic brain lesions by use of proton MR spectroscopy: The limits of accuracy with a logistic regression model. AJNR Am J Neuroradiol 21(7):1213–1219, 2000. Cairncross JG, Ueki K, Zlatescu MC, et al: Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst 90(19):1473–1479, 1998. Campbell JW, Pollack IF: Cerebellar astrocytomas in children.   J Neurooncol 28(2–3):223–231, 1996. Casadei GP, Komori T, Scheithauer BW, et al: Intracranial parenchymal schwannoma. A clinicopathological and neuroimaging study of nine cases. J Neurosurg 79(2):217–222, 1993. Castillo M, Kwock L: Proton MR spectroscopy of common brain tumors. Neuroimaging Clin N Am 8(4):733–752, 1998. Castillo M, Smith JK, Kwock L: Correlation of myo-inositol levels and grading of cerebral astrocytomas. AJNR Am J Neuroradiol 21(9):1645– 1649, 2000. Castillo M, Smith JK, Kwock L: Proton MR spectroscopy in patients with acute temporal lobe seizures. AJNR Am J Neuroradiol 22(1):152– 157, 2001. Cha S: Update on brain tumor imaging: From anatomy to physiology. AJNR Am J Neuroradiol 27(3):475–487, 2006. Cha S, Knopp EA, Johnson G, et al: Intracranial mass lesions: Dynamic contrast-enhanced susceptibility-weighted echo-planar perfusion MR imaging. Radiology 223(1):11–29, 2002. Cha S, Tihan T, Crawford F, et al: Differentiation of low-grade oligodendrogliomas from low-grade astrocytomas by using quantitative blood-volume measurements derived from dynamic susceptibility contrast-enhanced MR imaging. AJNR Am J Neuroradiol 26(2):266–273, 2005. Chan AA, Lau A, Pirzkall A, et al: Proton magnetic resonance spectroscopy imaging in the evaluation of patients undergoing gamma knife surgery for grade IV glioma. J Neurosurg 101(3):467–475, 2004. Chassoux F, Daumas-Duport C: Dysembryoplastic neuroepithelial tumors: Where are we now? Epilepsia 54(Suppl 9):129–134, 2013. Chen W, Delaloye S, Silverman DHS, et al: Predicting treatment response of malignant gliomas to bevacizumab and irinotecan by imaging proliferation with [18F] fluorothymidine positron emission tomography: A pilot study. J Clin Oncol 25(30):4714–4721, 2007. Chen W, Silverman DHS, Delaloye S, et al: 18F-FDOPA PET imaging of brain tumors: Comparison study with 18F-FDG PET and evaluation of diagnostic accuracy. J Nucl Med 47(6):904–911, 2006. Chenevert TL, McKeever PE, Ross BD: Monitoring early response of experimental brain tumors to therapy using diffusion magnetic resonance imaging. Clin Cancer Res 3(9):1457–1466, 1997.

59. Chenevert TL, Stegman LD, Taylor JM, et al: Diffusion magnetic resonance imaging: An early surrogate marker of therapeutic efficacy in brain tumors. J Natl Cancer Inst 92(24):2029–2036, 2000. 60. Chiang IC, Kuo Y-T, Lu C-Y, et al: Distinction between high-grade gliomas and solitary metastases using peritumoral 3-T magnetic resonance spectroscopy, diffusion, and perfusion imagings. Neuroradiology 46(8):619–627, 2004. 61. Chiechi MV, Smirniotopoulos JG, Mena H: Intracranial hemangiopericytomas: MR and CT features. AJNR Am J Neuroradiol 17(7):1365–1371, 1996. 62. Chu WCW, Chik K-W, Chan Y-L, et al: White matter and cerebral metabolite changes in children undergoing treatment for acute lymphoblastic leukemia: Longitudinal study with MR imaging and 1H MR spectroscopy. Radiology 229(3):659–669, 2003. 63. Chung J-K, Kim YK, Kim S-K, et al: Usefulness of 11C-methionine PET in the evaluation of brain lesions that are hypo- or isometabolic on 18F-FDG PET. Eur J Nucl Med Mol Imaging 29(2):176–182, 2002. 64. Civitello LA, Packer RJ, Rorke LB, et al: Leptomeningeal dissemination of low-grade gliomas in childhood. Neurology 38(4):562–566, 1988. 65. Cohen TI, Powers SK, Williams DW: MR appearance of intracanalicular eighth nerve lipoma. AJNR Am J Neuroradiol 13(4):1188–1190, 1992. 66. Cohen LM, Schwartz AM, Rockoff SD: Benign schwannomas: Pathologic basis for CT inhomogeneities. AJR Am J Roentgenol 147(1):141–143, 1986. 67. Cone L, Srinivasan M, Romanul FC: Granular cell tumor (choristoma) of the neurohypophysis: Two cases and a review of the literature. AJNR Am J Neuroradiol 11(2):403–406, 1990. 68. Conturo TE, Lori NF, Cull TS, et al: Tracking neuronal fiber pathways in the living human brain. Proc Natl Acad Sci U S A 96(18):10422– 10427, 1999. 69. Coons SW, Ashby LS: Pathology of intracranial neoplasms. Neuroimaging Clin N Am 9(4):615–649, 1999. 70. Coons SW, Johnson PC, Scheithauer BW, et al: Improving diagnostic accuracy and interobserver concordance in the classification and grading of primary gliomas. Cancer 79(7):1381–1393, 1997. 71. Cooper SD, Brady MB, Williams JP, et al: Neurosarcoidosis: Evaluation using computed tomography and magnetic resonance imaging.   J Comput Tomogr 12(2):96–99, 1988. 72. Cosgrove GR, Buchbinder BR, Jiang H: Functional magnetic resonance imaging for intracranial navigation. Neurosurg Clin N Am 7(2):313– 322, 1996. 73. Cottier JP, Destrieux C, Brunereau L, et al: Cavernous sinus invasion by pituitary adenoma: MR imaging. Radiology 215(2):463–469, 2000. 74. Crawford FW, Khayal IS, McGue C, et al: Relationship of pre-surgery metabolic and physiological MR imaging parameters to survival for patients with untreated GBM. J Neurooncol 91(3):337–351, 2009. 75. Crenshaw WB, Chew FS: Rathke’s cleft cyst. AJR Am J Roentgenol 158(6):1312, 1992. 76. Crocker EF, Zimmerman RA, Phelps ME, et al: The effect of steroids on the extravascular distribution of radiographic contrast material and technetium pertechnetate in brain tumors as determined by computed tomography. Radiology 119(2):471–474, 1976. 77. Crotty TB, Scheithauer BW, Young WF, et al: Papillary craniopharyngioma: A clinicopathological study of 48 cases.   J Neurosurg 83(2):206–214, 1995. 78. Danchaivijitr N, Waldman AD, Tozer DJ, et al: Low-grade gliomas: Do changes in rCBV measurements at longitudinal perfusion-weighted MR imaging predict malignant transformation? Radiology 247(1):170–178, 2008. 79. Daumas-Duport C, Tucker ML, Kolles H, et al: Oligodendrogliomas. Part II: A new grading system based on morphological and imaging criteria. J Neurooncol 34(1):61–78, 1997. 80. Davis PC, Wichman RD, Takei Y, et al: Primary cerebral neuroblastoma: CT and MR findings in 12 cases. AJNR Am J Neuroradiol 11(1):115– 120, 1990. 81. DeAngelis LM: Brain tumors. N Engl J Med 344(2):114–123, 2001. 82. Delattre JY, Krol G, Thaler HT, et al: Distribution of brain metastases. Arch Neurol 45(7):741–744, 1988.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms 83. Deng Z, Yan Y, Zhong D, et al: Quantitative analysis of glioma cell invasion by diffusion tensor imaging. J Clin Neurosci 17(12):1530–1536, 2010. 84. Derlon JM, Bourdet C, Bustany P, et al: [11C]L-methionine uptake in gliomas. Neurosurgery 25(5):720–728, 1989. 85. Derlon JM, Petit-Taboue MC, Chapon F, et al: The in vivo metabolic pattern of low-grade brain gliomas: A positron emission tomographic study using 18F-fluorodeoxyglucose and 11C-L-methylmethionine. Neurosurgery 40(2):276–287, 1997. 86. Desprechins B, Stadnik T, Koerts G, et al: Use of diffusion-weighted MR imaging in differential diagnosis between intracerebral necrotic tumors and cerebral abscesses. AJNR Am J Neuroradiol 20(7):1252– 1257, 1999. 87. De Stefano N, Caramanos Z, Preul MC, et al: In vivo differentiation of astrocytic brain tumors and isolated demyelinating lesions of the type seen in multiple sclerosis using 1H magnetic resonance spectroscopic imaging. Ann Neurol 44(2):273–278, 1998. 88. De Witte O, Goldberg I, Wikler D, et al: Positron emission tomography with injection of methionine as a prognostic factor in glioma. J Neurosurg 95(5):746–750, 2001. 89. De Witte O, Levivier M, Violon P, et al: Prognostic value positron emission tomography with [18F]fluoro-2-deoxy-D-glucose in the low-grade glioma. Neurosurgery 39(3):470–476, discussion 476–477, 1996. 90. Di Chiro G, DeLaPaz RL, Brooks RA, et al: Glucose utilization of cerebral gliomas measured by [18F] fluorodeoxyglucose and positron emission tomography. Neurology 32(12):1323–1329, 1982. 91. Dina TS: Primary central nervous system lymphoma versus toxoplasmosis in AIDS. Radiology 179(3):823–828, 1991. 92. Dorsay TA, Rovira MJ, Ho VB, et al: Ependymoblastoma: MR presentation. A case report and review of the literature. Pediatr Radiol 25(6):433–435, 1995. 93. Doucet V, Peretti-Viton P, Figarella-Branger D, et al: MRI of intracranial chordomas. Extent of tumour and contrast enhancement: Criteria for differential diagnosis. Neuroradiology 39(8):571–576, 1997. 94. Dowling C, Bollen AW, Noworolski SM, et al: Preoperative proton MR spectroscopic imaging of brain tumors: Correlation with histopathologic analysis of resection specimens. AJNR Am J Neuroradiol 22(4):604–612, 2001. 95. Drayer BP, Rosenbaum AE: Brain edema defined by cranial computed tomography. J Comput Assist Tomogr 3(3):317–323, 1979. 96. Earnest FT, Kelly PJ, Scheithauer BW, et al: Cerebral astrocytomas: Histopathologic correlation of MR and CT contrast enhancement with stereotactic biopsy. Radiology 166(3):823–827, 1988. 97. Elster AD: Modern imaging of the pituitary. Radiology 187(1):1–14, 1993. 98. Elster AD, Challa VR, Gilbert TH, et al: Meningiomas: MR and histopathologic features. Radiology 170(3 Pt 1):857–862, 1989. 99. Elster AD, Moody DM, Ball MR, et al: Is Gd-DTPA required for routine cranial MR imaging? Radiology 173(1):231–238, 1989. 100. Engelhard HH, Stelea A, Mundt A: Oligodendroglioma and anaplastic oligodendroglioma: Clinical features, treatment, and prognosis. Surg Neurol 60(5):443–456, 2003. 101. Esteve F, Rubin C, Grand S, et al: Transient metabolic changes observed with proton MR spectroscopy in normal human brain after radiation therapy. Int J Radiat Oncol Biol Phys 40(2):279–286, 1998. 102. Fan G, Sun B, Wu Z, et al: In vivo single-voxel proton MR spectroscopy in the differentiation of high-grade gliomas and solitary metastases. Clin Radiol 59(1):77–85, 2004. 103. Felix I, Becker LE: Intracranial germ cell tumors in children: An immunohistochemical and electron microscopic study. Pediatr Neurosurg 16(3):156–162, 1990. 104. Felix R, Schorner W, Laniado M, et al: Brain tumors: MR imaging with gadolinium-DTPA. Radiology 156(3):681–688, 1985. 105. Filippi CG, Edgar MA, Ulug AM, et al: Appearance of meningiomas on diffusion-weighted images: Correlating diffusion constants with histopathologic findings. AJNR Am J Neuroradiol 22(1):65–72, 2001.

259

106. Filling-Katz MR, Choyke PL, Oldfield E, et al: Central nervous system involvement in Von Hippel-Lindau disease. Neurology 41(1):41–46, 1991. 107. Finelli DA, Hurst GC, Gullapali RP, et al: Improved contrast of enhancing brain lesions on postgadolinium, T1-weighted spin-echo images with use of magnetization transfer. Radiology 190(2):553–559, 1994. 108. Fishman RA: Brain edema. N Engl J Med 293(14):706–711, 1975. 109. FitzGerald DB, Cosgrove GR, Ronner S, et al: Location of language in the cortex: A comparison between functional MR imaging and electrocortical stimulation. AJNR Am J Neuroradiol 18(8):1529–1539, 1997. 110. Floeth FW, Pauleit D, Sabel M, et al: Prognostic value of O-(2-18Ffluoroethyl)-L-tyrosine PET and MRI in low-grade glioma. J Nucl Med 48(4):519–527, 2007. 111. Forsyth PA, Cascino TL, Shaw EG, et al: Intracranial chordomas: A clinicopathological and prognostic study of 51 cases. J Neurosurg 78(5):741–747, 1993. 112. Forsyth PA, Posner JB: Headaches in patients with brain tumors: A study of 111 patients. Neurology 43(9):1678–1683, 1993. 113. Fujimaki T, Matsutani M, Funada N, et al: CT and MRI features of intracranial germ cell tumors. J Neurooncol 19(3):217–226, 1994. 114. Furie DM, Provenzale JM: Supratentorial ependymomas and subependymomas: CT and MR appearance. J Comput Assist Tomogr 19(4):518–526, 1995. 115. Gado MH, Phelps ME, Coleman RE: An extravascular component of contrast enhancement in cranial computed tomography. Part II. Contrast enhancement and the blood-tissue barrier. Radiology 117(3 Pt 1):595–597, 1975. 116. Georgy BA, Hesselink JR, Jernigan TL: MR imaging of the corpus callosum. AJR Am J Roentgenol 160(5):949–955, 1993. 117. Giannini C, Scheithauer BW, Burger PC, et al: Pleomorphic xanthoastrocytoma: What do we really know about it? Cancer 85(9):2033–2045, 1999. 118. Gill SS, Thomas DG, Van Bruggen N, et al: Proton MR spectroscopy of intracranial tumours: In vivo and in vitro studies. J Comput Assist Tomogr 14(4):497–504, 1990. 119. Ginaldi S, Wallace S, Shalen P, et al: Cranial computed tomography of malignant melanoma. AJR Am J Roentgenol 136(1):145–149, 1981. 120. Ginsberg LE: Contrast enhancement in meningeal and extra-axial disease. Neuroimaging Clin N Am 4(1):133–152, 1994. 121. Ginsberg LE, Leeds NE: Neuroradiology of leukemia. AJR Am J Roentgenol 165(3):525–534, 1995. 122. Giussani C, Roux F-E, Ojemann J, et al: Is preoperative functional magnetic resonance imaging reliable for language areas mapping in brain tumor surgery? Review of language functional magnetic resonance imaging and direct cortical stimulation correlation studies. Neurosurgery 66(1):113–120, 2010. 123. Goergen SK, Gonzales MF, McLean CA: Interventricular neurocytoma: Radiologic features and review of the literature. Radiology 182(3):787– 792, 1992. 124. Gold LH, Kieffer SA, Peterson HO: Intracranial meningiomas. A retrospective analysis of the diagnostic value of plain skull films. Neurology 19(9):873–878, 1969. 125. Goldman S, Levivier M, Pirotte B, et al: Regional methionine and glucose uptake in high-grade gliomas: A comparative study on PET-guided stereotactic biopsy. J Nucl Med 38(9):1459–1462, 1997. 126. Goldsher D, Litt AW, Pinto RS, et al: Dural “tail” associated with meningiomas on Gd-DTPA-enhanced MR images: Characteristics, differential diagnostic value, and possible implications for treatment. Radiology 176(2):447–450, 1990. 127. Gomori JM, Grossman RI, Goldberg HI, et al: Intracranial hematomas: Imaging by high-field MR. Radiology 157(1):87–93, 1985. 128. Graif M, Bydder GM, Steiner RE, et al: Contrast-enhanced MR imaging of malignant brain tumors. AJNR Am J Neuroradiol 6(6):855–862, 1985. 129. Gray LH, Conger AD, Ebert M, et al: The concentration of oxygen dissolved in tissues at the time of irradiation as a factor in radiotherapy. Br J Radiol 26(312):638–648, 1953.

http://radiologyebook.com

260

PART II  CT and MR Imaging of the Whole Body

130. Guermazi A, Feger C, Rousselot P, et al: Granulocytic sarcoma (chloroma): Imaging findings in adults and children. AJR Am J Roentgenol 178(2):319–325, 2002. 131. Gulyas B, Halldin C: New PET radiopharmaceuticals beyond FDG   for brain tumor imaging. Q J Nucl Med Mol Imaging 56(2):173–190, 2012. 132. Guo AC, Cummings TJ, Dash RC, et al: Lymphomas and high-grade astrocytomas: Comparison of water diffusibility and histologic characteristics. Radiology 224(1):177–183, 2002. 133. Gusnard DA: Cerebellar neoplasms in children. Semin Roentgenol 25(3):263–278, 1990. 134. Guthrie BL, Ebersold MJ, Scheithauer BW, et al: Meningeal hemangiopericytoma: Histopathological features, treatment, and long-term follow-up of 44 cases. Neurosurgery 25(4):514–522, 1989. 135. Hall WA, Galicich W, Bergman T, et al: 3-Tesla intraoperative MR imaging for neurosurgery. J Neurooncol 77(3):297–303, 2006. 136. Hall WA, Kim P, Truwit CL: Functional magnetic resonance imagingguided brain tumor resection. Top Magn Reson Imaging 19(4):205–212, 2009. 137. Hall WA, Luciano MG, Doppman JL, et al: Pituitary magnetic resonance imaging in normal human volunteers: Occult adenomas in the general population. Ann Intern Med 120(10):817–820, 1994. 138. Hall WA, Martin A, Liu H, et al: Improving diagnostic yield in brain biopsy: Coupling spectroscopic targeting with real-time needle placement. J Magn Reson Imaging 13(1):12–15, 2001. 139. Hall DE, Moffat BA, Stojanovska J, et al: Therapeutic efficacy of DTI-015 using diffusion magnetic resonance imaging as an early surrogate marker. Clin Cancer Res 10(23):7852–7859, 2004. 140. Hamstra DA, Galban CJ, Meyer CR, et al: Functional diffusion map as an early imaging biomarker for high-grade glioma: Correlation with conventional radiologic response and overall survival. J Clin Oncol 26(20):3387–3394, 2008. 141. Harwood-Nash DC: Neuroimaging of childhood craniopharyngioma. Pediatr Neurosurg 21(Suppl 1):2–10, 1994. 142. Hatam A, Bergstrom M, Yu ZY, et al: Effect of dexamethasone treatment on volume and contrast enhancement of intracranial neoplasms. J Comput Assist Tomogr 7(2):295–300, 1983. 143. Hatiboglu MA, Weinberg JS, Suki D, et al: Impact of intraoperative high-field magnetic resonance imaging guidance on glioma surgery: A prospective volumetric analysis. Neurosurgery 64(6):1073–1081, discussion 1081, 2009. 144. Hattingen E, Raab P, Franz K, et al: Myo-inositol: A marker of reactive astrogliosis in glial tumors? NMR Biomed 21(3):233–241, 2008. 145. Haustein J, Laniado M, Niendorf HP, et al: Triple-dose versus standard-dose gadopentetate dimeglumine: A randomized study in 199 patients. Radiology 186(3):855–860, 1993. 146. Hayman LA, Evans RA, Hinck VC: Delayed high iodine dose contrast computed tomography: Cranial neoplasms. Radiology 136(3):677–684, 1980. 147. Healey EA, Barnes PD, Kupsky WJ, et al: The prognostic significance of postoperative residual tumor in ependymoma. Neurosurgery 28(5):666– 671, 1991. 148. Healy ME, Hesselink JR, Press GA, et al: Increased detection of intracranial metastases with intravenous Gd-DTPA. Radiology 165(3):619–624, 1987. 149. Healy JF, Marshall WH, Brahme FJ, et al: CT of intracranial metastases with skull and scalp involvement. AJNR Am J Neuroradiol 2(4):335–338, 1981. 150. Henning EC, Azuma C, Sotak CH, et al: Multispectral tissue characterization in a RIF-1 tumor model: Monitoring the ADC and T2 responses to single-dose radiotherapy. Part II. Magn Reson Med 57(3):513–519, 2007. 151. Herholz K, Heindel W, Luyten PR, et al: In vivo imaging of glucose consumption and lactate concentration in human gliomas. Ann Neurol 31(3):319–327, 1992. 152. Herholz K, Holzer T, Bauer B, et al: 11C-methionine PET for differential diagnosis of low-grade gliomas. Neurology 50(5):1316–1322, 1998.

153. Hermann EJ, Hattingen E, Krauss JK, et al: Stereotactic biopsy in gliomas guided by 3-tesla 1H-chemical-shift imaging of choline. Stereotact Funct Neurosurg 86(5):300–307, 2008. 154. Ho VB, Smirniotopoulos JG, Murphy FM, et al: Radiologic-pathologic correlation: Hemangioblastoma. AJNR Am J Neuroradiol 13(5):1343– 1352, 1992. 155. Hochberg FH, Miller DC: Primary central nervous system lymphoma.   J Neurosurg 68(6):835–853, 1988. 156. Hockel M, Schlenger K, Aral B, et al: Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res 56(19):4509–4515, 1996. 157. Hoffman HJ, Otsubo H, Hendrick EB, et al: Intracranial germ-cell tumors in children. J Neurosurg 74(4):545–551, 1991. 158. Hollander MD, FitzPatrick M, O’Connor SG, et al: Optic gliomas. Radiol Clin North Am 37(1):59–71, ix, 1999. 159. Horska A, Barker PB: Imaging of brain tumors: MR spectroscopy   and metabolic imaging. Neuroimaging Clin N Am 20(3):293–310, 2010. 160. Hourani R, Horska A, Albayram S, et al: Proton magnetic resonance spectroscopic imaging to differentiate between nonneoplastic lesions and brain tumors in children. J Magn Reson Imaging 23(2):99–9107, 2006. 161. Huang C-F, Chou H-H, Tu H-T, et al: Diffusion magnetic resonance imaging as an evaluation of the response of brain metastases treated by stereotactic radiosurgery. Surg Neurol 69(1):62–68, 2008. 162. Hudgins PA, Davis PC, Hoffman JC: Gadopentetate dimeglumineenhanced MR imaging in children following surgery for brain tumor: Spectrum of meningeal findings. AJNR Am J Neuroradiol 12(2):301– 307, 1991. 163. Hygino da Cruz LC, Jr, Rodriguez I, Domingues RC, et al: Pseudoprogression and pseudoresponse: Imaging challenges in the assessment of posttreatment glioma. AJNR Am J Neuroradiol 32(11):1978–1985, 2011. 164. Jack CR, O’Neill BP, Banks PM, et al: Central nervous system lymphoma: Histologic types and CT appearance. Radiology 167(1):211– 215, 1988. 165. Jack CR, Thompson RM, Butts RK, et al: Sensory motor cortex: Correlation of presurgical mapping with functional MR imaging and invasive cortical mapping. Radiology 190(1):85–92, 1994. 166. Jager PL, Vaalburg W, Pruim J, et al: Radiolabeled amino acids: Basic aspects and clinical applications in oncology. J Nucl Med 42(3):432–445, 2001. 167. Jain R, Griffith B, Narang J, et al: Blood-brain barrier imaging in brain tumors: Concepts and methods. Neurographics 2(2):48–59, 2012. 168. Jansen NL, Graute V, Armbruster L, et al: MRI-suspected low-grade glioma: Is there a need to perform dynamic FET PET? Eur J Nucl Med Mol Imaging 39(6):1021–1029, 2012. 169. Janssen HL, Haustermans KM, Balm AJ, et al: Hypoxia in head and neck cancer: How much, how important? Head Neck 27(7):622–638, 2005. 170. Jaremko JL, Jans LBO, Coleman LT, et al: Value and limitations of diffusion-weighted imaging in grading and diagnosis of pediatric posterior fossa tumors. AJNR Am J Neuroradiol 31(9):1613–1616, 2010. 171. Jellison BJ, Field AS, Medow J, et al: Diffusion tensor imaging of cerebral white matter: A pictorial review of physics, fiber tract anatomy, and tumor imaging patterns. AJNR Am J Neuroradiol 25(3):356–369, 2004. 172. Jennings MT, Gelman R, Hochberg F: Intracranial germ-cell tumors: Natural history and pathogenesis. J Neurosurg 63(2):155–167, 1985. 173. Johnson JH, Hariharan S, Berman J, et al: Clinical outcome of pediatric gangliogliomas: Ninety-nine cases over 20 years. Pediatr Neurosurg 27(4):203–207, 1997. 174. Johnson PC, Hunt SJ, Drayer BP: Human cerebral gliomas: correlation of postmortem MR imaging and neuropathologic findings. Radiology 170(1 Pt 1):211–217, 1989. 175. Johnson DH, Windham WW, Allen JH, et al: Limited value of CT brain scans in the staging of small cell lung cancer. AJR Am J Roentgenol 140(1):37–40, 1983.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms 176. Juneau P, Schoene WC, Black P: Malignant tumors in the pituitary gland. Arch Neurol 49(5):555–558, 1992. 177. Kaminaga T, Shirai K: Radiation-induced brain metabolic changes in the acute and early delayed phase detected with quantitative proton magnetic resonance spectroscopy. J Comput Assist Tomogr 29(3):293– 297, 2005. 178. Kapouleas I, Alavi A, Alves WM, et al: Registration of threedimensional MR and PET images of the human brain without markers. Radiology 181(3):731–739, 1991. 179. Karsdorp N, Elderson A, Wittebol-Post D, et al: Von Hippel-Lindau disease: New strategies in early detection and treatment. Am J Med 97(2):158–168, 1994. 180. Kaschten B, Stevenaert A, Sadzot B, et al: Preoperative evaluation of 54 gliomas by PET with fluorine-18-fluorodeoxyglucose and/or carbon11-methionine. J Nucl Med 39(5):778–785, 1998. 181. Kato T, Shinoda J, Oka N, et al: Analysis of 11C-methionine uptake in low-grade gliomas and correlation with proliferative activity. AJNR Am J Neuroradiol 29(10):1867–1871, 2008. 182. Kelly PJ, Daumas-Duport C, Scheithauer BW, et al: Stereotactic histologic correlations of computed tomography- and magnetic resonance imaging-defined abnormalities in patients with glial neoplasms. Mayo Clin Proc 62(6):450–459, 1987. 183. Kim S, Chung JK, Im SH, et al: 11C-methionine PET as a prognostic marker in patients with glioma: Comparison with 18F-FDG PET. Eur J Nucl Med Mol Imaging 32(1):52–59, 2005. 184. Kim YH, Oh SW, Lim YJ, et al: Differentiating radiation necrosis from tumor recurrence in high-grade gliomas: Assessing the efficacy of 18F-FDG PET, 11C-methionine PET and perfusion MRI. Clin Neurol Neurosurg 112(9):758–765, 2010. 185. Kitis O, Altay H, Calli C, et al: Minimum apparent diffusion   coefficients in the evaluation of brain tumors. Eur J Radiol 55(3):393–400, 2005. 186. Klatzo I: Pathophysiological aspects of brain edema. Acta Neuropathol 72(3):236–239, 1987. 187. Reference deleted in proofs. 188. Kleihues P, Burger PC, Scheithauer BW: The new WHO classification of brain tumours. Brain Pathol 3(3):255–268, 1993. 189. Klibanski A, Zervas NT: Diagnosis and management of hormonesecreting pituitary adenomas. N Engl J Med 324(12):822–831, 1991. 190. Knopp EA, Cha S, Johnson G, et al: Glial neoplasms: dynamic contrast-enhanced T2*-weighted MR imaging. Radiology 211(3):791– 798, 1999. 191. Koci TM, Chiang F, Mehringer CM, et al: Adult cerebellar medulloblastoma: Imaging features with emphasis on MR findings. AJNR Am J Neuroradiol 14(4):929–939, 1993. 192. Koeller KK, Dillon WP: Dysembryoplastic neuroepithelial tumors: MR appearance. AJNR Am J Neuroradiol 13(5):1319–1325, 1992. 193. Koeller KK, Henry JM: From the archives of the AFIP: Superficial gliomas: Radiologic-pathologic correlation. Armed Forces Institute of Pathology. Radiographics 21(6):1533–1556, 2001. 194. Koeller KK, Rushing EJ: From the archives of the AFIP: Pilocytic astrocytoma: Radiologic-pathologic correlation. Radiographics 24(6):1693–1708, 2004. 195. Koeller KK, Sandberg GD: From the archives of the AFIP. Cerebral intraventricular neoplasms: Radiologic-pathologic correlation. Radiographics 22(6):1473–1505, 2002. 196. Kokkoris CP: Leptomeningeal carcinomatosis. How does cancer reach the pia-arachnoid? Cancer 51(1):154–160, 1983. 197. Kono K, Inoue Y, Nakayama K, et al: The role of diffusion-weighted imaging in patients with brain tumors. AJNR Am J Neuroradiol 22(6):1081–1088, 2001. 198. Koral K, Zhang S, Gargan L, et al: Diffusion MRI improves the accuracy of preoperative diagnosis of common pediatric cerebellar tumors among reviewers with different experience levels. AJNR Am J Neuroradiol 34(12):2360–2365, 2013. 199. Kremer S, Abu Eid M, Bierry G, et al: Accuracy of delayed post-contrast FLAIR MR imaging for the diagnosis of leptomeningeal infectious or tumoral diseases. J Neuroradiol 33(5):285–291, 2006.

261

200. Kucharczyk W, Davis DO, Kelly WM, et al: Pituitary adenomas: High-resolution MR imaging at 1.5 T. Radiology 161(3):761–765, 1986. 201. Kulkantrakorn K, Awwad EE, Levy B, et al: MRI in Lhermitte-Duclos disease. Neurology 48(3):725–731, 1997. 202. Kumar AJ, Leeds NE, Fuller GN, et al: Malignant gliomas: MR imaging spectrum of radiation therapy- and chemotherapy-induced necrosis of the brain after treatment. Radiology 217(2):377–384, 2000. 203. Kuroiwa T, Bergey GK, Rothman MI, et al: Radiologic appearance of the dysembryoplastic neuroepithelial tumor. Radiology 197(1):233–238, 1995. 204. Lacerda S, Law M: Magnetic resonance perfusion and permeability imaging in brain tumors. Neuroimaging Clin N Am 19(4):527–557, 2009. 205. Laine FJ, Nadel L, Braun IF: CT and MR imaging of the central skull base. Part 2. Pathologic spectrum. Radiographics 10(5):797–821, 1990. 206. Langkowski JH, Wieland J, Bomsdorf H, et al: Pre-operative localized in vivo proton spectroscopy in cerebral tumors at 4.0 Tesla–First results. Magn Reson Imaging 7(5):547–555, 1989. 207. Laundre BJ, Jellison BJ, Badie B, et al: Diffusion tensor imaging of the corticospinal tract before and after mass resection as correlated with clinical motor findings: Preliminary data. AJNR Am J Neuroradiol 26(4):791–796, 2005. 208. Law M, Cha S, Knopp EA, et al: High-grade gliomas and solitary metastases: Differentiation by using perfusion and proton spectroscopic MR imaging. Radiology 222(3):715–721, 2002. 209. Law M, Yang S, Wang H, et al: Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 24(10):1989–1998, 2003. 210. Laws ER, Thapar K: Brain tumors. CA Cancer J Clin 43(5):263–271, 1993. 211. Lee YY, Glass JP, Geoffray A, et al: Cranial computed tomographic abnormalities in leptomeningeal metastasis. AJR Am J Roentgenol 143(5):1035–1039, 1984. 212. Lee SR, Sanches J, Mark AS, et al: Posterior fossa hemangioblastomas: MR imaging. Radiology 171(2):463–468, 1989. 213. Reference deleted in proofs. 214. Lee YY, Van Tassel P, Bruner JM, et al: Juvenile pilocytic astrocytomas: CT and MR characteristics. AJR Am J Roentgenol 152(6):1263–1270, 1989. 215. Lee CC, Ward HA, Sharbrough FW, et al: Assessment of functional MR imaging in neurosurgical planning. AJNR Am J Neuroradiol 20(8):1511– 1519, 1999. 216. Leeds NE, Elkin CM, Zimmerman RD: Gliomas of the brain. Semin Roentgenol 19(1):27–43, 1984. 217. Lev MH, Ozsunar Y, Henson JW, et al: Glial tumor grading and outcome prediction using dynamic spin-echo MR susceptibility mapping compared with conventional contrast-enhanced MR: Confounding effect of elevated rCBV of oligodendrogliomas [corrected]. AJNR Am J Neuroradiol 25(2):214–221, 2004. 218. Lev MH, Rosen BR: Clinical applications of intracranial perfusion MR imaging. Neuroimaging Clin N Am 9(2):309–331, 1999. 219. Levine SN, Benzel EC, Fowler MR, et al: Lymphocytic adenohypophysitis: Clinical, radiological, and magnetic resonance imaging characterization. Neurosurgery 22(5):937–941, 1988. 220. Levy RA, Blaivas M, Muraszko K, et al: Desmoplastic medulloblastoma: MR findings. AJNR Am J Neuroradiol 18(7):1364–1366, 1997. 221. Lilja A, Bergstrom K, Hartvig P, et al: Dynamic study of supratentorial gliomas with L-methyl-11C-methionine and positron emission tomography. AJNR Am J Neuroradiol 6(4):505–514, 1985. 222. Lindau R: Studien ueber Kleinhirncysten: Bau, Pathogenese und Bezeihungen zur Angiomatosis Retinae. Arch Pathol Microbiol Scand (Suppl 1):1926. 223. Linskey ME, Andrews DW, Asher AL, et al: The role of stereotactic radiosurgery in the management of patients with newly diagnosed brain metastases: A systematic review and evidence-based clinical practice guideline. J Neurooncol 96(1):45–68, 2010.

http://radiologyebook.com

262

PART II  CT and MR Imaging of the Whole Body

224. Lipper MH, Eberhard DA, Phillips CD, et al: Pleomorphic xanthoastrocytoma, a distinctive astroglial tumor: Neuroradiologic   and pathologic features. AJNR Am J Neuroradiol 14(6):1397–1404, 1993. 225. Litofsky NS, Zee CS, Breeze RE, et al: Meningeal melanocytoma: Diagnostic criteria for a rare lesion. Neurosurgery 31(5):945–948, 1992. 226. Lizak PF, Woodruff WW: Posterior fossa neoplasms: Multiplanar imaging. Semin Ultrasound CT MR 13(3):182–206, 1992. 227. Long DM: Capillary ultrastructure and the blood-brain barrier in human malignant brain tumors. J Neurosurg 32(2):127–144, 1970. 228. Lote K, Egeland T, Hager B, et al: Survival, prognostic factors, and therapeutic efficacy in low-grade glioma: A retrospective study in 379 patients. J Clin Oncol 15(9):3129–3140, 1997. 229. Louis DN, Ohgaki H, Wiestler OD, et al, editors: WHO classification of tumours of the central nervous system, Lyon, 2007, IARC. 230. Louis DN, Ohgaki H, Wiestler OD, et al: The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 114(2):97– 109, 2007. 231. Ludemann L, Grieger W, Wurm R, et al: Comparison of dynamic contrast-enhanced MRI with WHO tumor grading for gliomas. Eur Radiol 11(7):1231–1241, 2001. 232. Luh GY, Bird CR: Imaging of brain tumors in the pediatric population. Neuroimaging Clin N Am 9(4):691–716, 1999. 233. Maciunas RJ, Berger MS, Copeland B, et al: A technique for interactive image-guided neurosurgical intervention in primary brain tumors. Neurosurg Clin N Am 7(2):245–266, 1996. 234. Maiuri F, Iaconetta G, de Divitiis O, et al: Intracranial meningiomas: Correlations between MR imaging and histology. Eur J Radiol 31(1):69–75, 1999. 235. Majos C, Aguilera C, Alonso J, et al: Proton MR spectroscopy improves discrimination between tumor and pseudotumoral lesion in solid brain masses. AJNR Am J Neuroradiol 30(3):544–551, 2009. 236. Mandybur TI: Intracranial hemorrhage caused by metastatic tumors. Neurology 27(7):650–655, 1977. 237. Mangla R, Kolar B, Zhu T, et al: Percentage signal recovery derived from MR dynamic susceptibility contrast imaging is useful to differentiate common enhancing malignant lesions of the brain. AJNR Am J Neuroradiol 32(6):1004–1010, 2011. 238. Mantravadi RV, Phatak R, Bellur S, et al: Brainstem gliomas: An autopsy study of 25 cases. Cancer 49(6):1294–1296, 1982. 239. Marcus KJ, Astrakas LG, Zurakowski D, et al: Predicting survival of children with CNS tumors using proton magnetic resonance spectroscopic imaging biomarkers. Int J Oncol 30(3):651–657, 2007. 240. Mardor Y, Pfeffer R, Spiegelmann R, et al: Early detection of response to radiation therapy in patients with brain malignancies using conventional and high b-value diffusion-weighted magnetic resonance imaging. J Clin Oncol 21(6):1094–1100, 2003. 241. Mark AS: Contrast-enhanced magnetic resonance imaging of the temporal bone. Neuroimaging Clin N Am 4(1):117–131, 1994. 242. Marks JE, Gado M: Serial computed tomography of primary brain tumors following surgery, irradiation, and chemotherapy. Radiology 125(1):119–125, 1977. 243. Martin AJ, Liu H, Hall WA, et al: Preliminary assessment of turbo spectroscopic imaging for targeting in brain biopsy. AJNR Am J Neuroradiol 22(5):959–968, 2001. 244. Matsusue E, Fink JR, Rockhill JK, et al: Distinction between glioma progression and post-radiation change by combined physiologic MR imaging. Neuroradiology 52(4):297–306, 2010. 245. Matsutani M, Sano K, Takakura K, et al: Primary intracranial germ   cell tumors: A clinical analysis of 153 histologically verified cases.   J Neurosurg 86(3):446–455, 1997. 246. McConville P, Hambardzumyan D, Moody JB, et al: Magnetic resonance imaging determination of tumor grade and early response to temozolamide in a genetically engineered mouse model of glioma. Clin Cancer Res 13(10):2897–2904, 2007. 247. McCormack BM, Miller DC, Budzilovich GN, et al: Treatment and survival of low-grade astrocytoma in adults–1977-1988. Neurosurgery 31(4):636–642, 1992.

248. Meyers SP, Hirsch WL, Curtin HD, et al: Chondrosarcomas of the skull base: MR imaging features. Radiology 184(1):103–108, 1992. 249. Miller DC, Hochberg FH, Harris NL, et al: Pathology with clinical correlations of primary central nervous system non-Hodgkin’s lymphoma. The Massachusetts General Hospital experience 1958-1989. Cancer 74(4):1383–1397, 1994. 250. Mitsuya K, Nakasu Y, Horiguchi S, et al: Perfusion weighted magnetic resonance imaging to distinguish the recurrence of metastatic brain tumors from radiation necrosis after stereotactic radiosurgery.   J Neurooncol 99(1):81–88, 2010. 251. Miwa K, Shinoda J, Yano H, et al: Discrepancy between lesion distributions on methionine PET and MR images in patients with glioblastoma multiforme: Insight from a PET and MR fusion image study. J Neurol Neurosurg Psychiatry 75(10):1457–1462, 2004. 252. Moffat BA, Chenevert TL, Lawrence TS, et al: Functional diffusion map: A noninvasive MRI biomarker for early stratification of clinical brain tumor response. Proc Natl Acad Sci U S A 102(15):5524–5529, 2005. 253. Moller-Hartmann W, Herminghaus S, Krings T, et al: Clinical application of proton magnetic resonance spectroscopy in the diagnosis of intracranial mass lesions. Neuroradiology 44(5):371–381, 2002. 254. Monajati A, Heggeness L: Patterns of edema in tumors vs. infarcts: Visualization of white matter pathways. AJNR Am J Neuroradiol 3(3):251–255, 1982. 255. Mori S, Crain BJ, Chacko VP, et al: Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 45(2):265–269, 1999. 256. Morioka T, Yamamoto T, Mizushima A, et al: Comparison of magnetoencephalography, functional MRI, and motor evoked potentials in the localization of the sensory-motor cortex. Neurol Res 17(5):361–367, 1995. 257. Morrison G, Sobel DF, Kelley WM, et al: Intraventricular mass lesions. Radiology 153(2):435–442, 1984. 258. Mosskin M, Bergstrom M, Collins VP, et al: Positron emission tomography with 11C-methionine of intracranial tumours compared with histology of multiple biopsies. Acta Radiol Suppl 369:157–160, 1986. 259. Mosskin M, Ericson K, Hindmarsh T, et al: Positron emission tomography compared with magnetic resonance imaging and computed tomography in supratentorial gliomas using multiple stereotactic biopsies as reference. Acta Radiol 30(3):225–232, 1989. 260. Mulkens TH, Parizel PM, Martin JJ, et al: Acoustic schwannoma: MR findings in 84 tumors. AJR Am J Roentgenol 160(2):395–398, 1993. 261. Mullins ME, Barest GD, Schaefer PW, et al: Radiation necrosis versus glioma recurrence: Conventional MR imaging clues to diagnosis. AJNR Am J Neuroradiol 26(8):1967–1972, 2005. 262. Mulvihill JJ, Parry DM, Sherman JL, et al: NIH conference. Neurofibromatosis 1 (Recklinghausen disease) and neurofibromatosis 2 (bilateral acoustic neurofibromatosis). An update. Ann Intern Med 113(1):39–52, 1990. 263. Murray JJ, Greco FA, Wolff SN, et al: Neoplastic meningitis. Marked variations of cerebrospinal fluid composition in the absence of extradural block. Am J Med 75(2):289–294, 1983. 264. Nabors LB, Portnow J, Ammirati M, et al: Central nervous system cancers, version 2.2014: Featured updates to the NCCN guidelines.   J Natl Compr Canc Netw 12(11):1517–1523, 2014. 265. Nagar VA, Ye JR, Ng WH, et al: Diffusion-weighted MR imaging: Diagnosing atypical or malignant meningiomas and detecting tumor dedifferentiation. AJNR Am J Neuroradiol 29(6):1147–1152, 2008. 266. Nagele T, Petersen D, Klose U, et al: The “dural tail” adjacent to meningiomas studied by dynamic contrast-enhanced MRI: A comparison with histopathology. Neuroradiology 36(4):303–307, 1994. 267. Naidich TP, Zimmerman RA: Primary brain tumors in children. Semin Roentgenol 19(2):100–114, 1984. 268. Narang J, Jain R, Arbab AS, et al: Differentiating treatment-induced necrosis from recurrent/progressive brain tumor using nonmodel-based semiquantitative indices derived from dynamic contrast-enhanced T1-weighted MR perfusion. Neuro Oncol 13(9):1037–1046, 2011.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms 269. Naugle DK, Duncan TD, Grice GP: Oligoastrocytoma. Radiographics 24(2):598–600, 2004. 270. Naul LG, Hise JH, Bauserman SC, et al: CT and MR of meningeal melanocytoma. AJNR Am J Neuroradiol 12(2):315–316, 1991. 271. Nazar GB, Hoffman HJ, Becker LE, et al: Infratentorial ependymomas in childhood: Prognostic factors and treatment. J Neurosurg 72(3):408– 417, 1990. 272. Neumann HP, Eggert HR, Scheremet R, et al: Central nervous system lesions in von Hippel-Lindau syndrome. J Neurol Neurosurg Psychiatry 55(10):898–901, 1992. 273. New PF, Hesselink JR, O’Carroll CP, et al: Malignant meningiomas: CT and histologic criteria, including a new CT sign. AJNR Am J Neuroradiol 3(3):267–276, 1982. 274. Newton DR, Dillon WP, Norman D, et al: Gd-DTPA-enhanced MR imaging of pituitary adenomas. AJNR Am J Neuroradiol 10(5):949–954, 1989. 275. Nguyen N, Montagnese J, Rogers LR, et al: Positron emission tomography-magnetic resonance imaging in the evaluation of brain tumors: Current status and future prospects. Semin Roentgenol 49(3):275–289, 2014. 276. Nihashi T, Dahabreh IJ, Terasawa T: Diagnostic accuracy of PET for recurrent glioma diagnosis: A meta-analysis. AJNR Am J Neuroradiol 34(5):944–950, S941–911, 2013. 277. Norman D, Stevens EA, Wing SD, et al: Quantitative aspects of contrast enhancement in cranial computed tomography. Radiology 129(3):683– 688, 1978. 278. Ogawa T, Shishido F, Kanno I, et al: Cerebral glioma: Evaluation with methionine PET. Radiology 186(1):45–53, 1993. 279. Olivero WC, Dulebohn SC, Lister JR: The use of PET in evaluating patients with primary brain tumours: Is it useful? J Neurol Neurosurg Psychiatry 58(2):250–252, 1995. 280. Osborn AG, Heaston DK, Wing SD: Diagnosis of ascending transtentorial herniation by cranial computed tomography. AJR Am J Roentgenol 130(4):755–760, 1978. 281. Ostertun B, Wolf HK, Campos MG, et al: Dysembryoplastic neuroepithelial tumors: MR and CT evaluation. AJNR Am J Neuroradiol 17(3):419–430, 1996. 282. Ostrom QT, Gittleman H, Liao P, et al: CBTRUS statistical report: Primary brain and central nervous system tumors diagnosed in the United States in 2007-2011. Neuro Oncol 16(Suppl 4):iv1–iv63, 2014. 283. Ostrov SG, Quencer RM, Hoffman JC, et al: Hemorrhage within pituitary adenomas: How often associated with pituitary apoplexy syndrome? AJR Am J Roentgenol 153(1):153–160, 1989. 284. Ozsunar Y, Mullins ME, Kwong K, et al: Glioma recurrence versus radiation necrosis? A pilot comparison of arterial spin-labeled, dynamic susceptibility contrast enhanced MRI, and FDG-PET imaging. Acad Radiol 17(3):282–290, 2010. 285. Padma MV, Said S, Jacobs M, et al: Prediction of pathology and survival by FDG PET in gliomas. J Neurooncol 64(3):227–237, 2003. 286. Parker K, Hardjasudarma M, McClellan RL, et al: MR features of an intracerebellar chloroma. AJNR Am J Neuroradiol 17(8):1592–1594, 1996. 287. Parmar H, Sitoh Y-Y, Anand P, et al: Contrast-enhanced flair imaging in the evaluation of infectious leptomeningeal diseases. Eur J Radiol 58(1):89–95, 2006. 288. Parsa AT, Wachhorst S, Lamborn KR, et al: Prognostic significance of intracranial dissemination of glioblastoma multiforme in adults.   J Neurosurg 102(4):622–628, 2005. 289. Patankar TF, Haroon HA, Mills SJ, et al: Is volume transfer coefficient (K(trans)) related to histologic grade in human gliomas? AJNR Am J Neuroradiol 26(10):2455–2465, 2005. 290. Patronas NJ, Di Chiro G, Brooks RA, et al: Work in progress: [18F] fluorodeoxyglucose and positron emission tomography in the evaluation of radiation necrosis of the brain. Radiology 144(4):885–889, 1982. 291. Patronas NJ, Di Chiro G, Kufta C, et al: Prediction of survival in glioma patients by means of positron emission tomography. J Neurosurg 62(6):816–822, 1985.

263

292. Paulson ES, Schmainda KM: Comparison of dynamic susceptibilityweighted contrast-enhanced MR methods: Recommendations for measuring relative cerebral blood volume in brain tumors. Radiology 249(2):601–613, 2008. 293. Paz-Fumagalli R, Daniels DL, Millen SJ, et al: Dural “tail” associated with an acoustic schwannoma in MR imaging with gadopentetate dimeglumine. AJNR Am J Neuroradiol 12(6):1206, 1991. 294. Pencalet P, Sainte-Rose C, Lellouch-Tubiana A, et al: Papillomas and carcinomas of the choroid plexus in children. J Neurosurg 88(3):521– 528, 1998. 295. Penn I, Porat G: Central nervous system lymphomas in organ allograft recipients. Transplantation 59(2):240–244, 1995. 296. Percy AK, Elveback LR, Okazaki H, et al: Neoplasms of the central nervous system. Epidemiologic considerations. Neurology 22(1):40–48, 1972. 297. Peretti-Viton P, Brunel H, Chinot O, et al: Histological and MR correlations in Gliomatosis cerebri. J Neurooncol 59(3):249–259, 2002. 298. Perry JR: Oligodendrogliomas: Clinical and genetic correlations. Curr Opin Neurol 14(6):705–710, 2001. 299. Perry A, Stafford SL, Scheithauer BW, et al: Meningioma grading: An analysis of histologic parameters. Am J Surg Pathol 21(12):1455–1465, 1997. 300. Petrella JR, Shah LM, Harris KM, et al: Preoperative functional MR imaging localization of language and motor areas: Effect on therapeutic decision making in patients with potentially resectable brain tumors. Radiology 240(3):793–802, 2006. 301. Phillips ME, Ryals TJ, Kambhu SA, et al: Neoplastic vs inflammatory meningeal enhancement with Gd-DTPA. J Comput Assist Tomogr 14(4):536–541, 1990. 302. Piepmeier J, Christopher S, Spencer D, et al: Variations in the natural history and survival of patients with supratentorial low-grade astrocytomas. Neurosurgery 38(5):872–878, discussion 878–879, 1996. 303. Pierpaoli C, Jezzard P, Basser PJ, et al: Diffusion tensor MR imaging of the human brain. Radiology 201(3):637–648, 1996. 304. Pirotte B, Goldman S, David P, et al: Stereotactic brain biopsy guided by positron emission tomography (PET) with [F-18]fluorodeoxyglucose and [C-11]methionine. Acta Neurochir Suppl 68:133–138, 1997. 305. Plotkin M, Eisenacher J, Bruhn H, et al: 123I-IMT SPECT and 1H MR-spectroscopy at 3.0 T in the differential diagnosis of recurrent or residual gliomas: A comparative study. J Neurooncol 70(1):49–58, 2004. 306. Pollack IF, Lunsford LD, Flickinger JC, et al: Prognostic factors in the diagnosis and treatment of primary central nervous system lymphoma. Cancer 63(5):939–947, 1989. 307. Pöpperl G, Götz C, Rachinger W, et al: Value of O-(2-[18F] fluoroethyl)-l-tyrosine PET for the diagnosis of recurrent glioma. Eur J Nucl Med Mol Imaging 31(11):1464–1470, 2004. 308. Popperl G, Kreth FW, Mehrkens JH, et al: FET PET for the evaluation of untreated gliomas: Correlation of FET uptake and uptake kinetics with tumour grading. Eur J Nucl Med Mol Imaging 34(12):1933–1942, 2007. 309. Poptani H, Gupta RK, Roy R, et al: Characterization of intracranial mass lesions with in vivo proton MR spectroscopy. AJNR Am J Neuroradiol 16(8):1593–1603, 1995. 310. Poptani H, Kaartinen J, Gupta RK, et al: Diagnostic assessment of brain tumours and non-neoplastic brain disorders in vivo using proton nuclear magnetic resonance spectroscopy and artificial neural networks. J Cancer Res Clin Oncol 125(6):343–349, 1999. 311. Posner JB: Neurological complications of systemic cancer. Med Clin North Am 63(4):783–800, 1979. 312. Posner JB: Brain tumors. CA Cancer J Clin 43(5):261–262, 1993. 313. Posner JB, Chernik NL: Intracranial metastases from systemic cancer. Adv Neurol 19:579–592, 1978. 314. Potts DG, Abbott GF, von Sneidern JV: National Cancer Institute study: Evaluation of computed tomography in the diagnosis of intracranial neoplasms. III. Metastatic tumors. Radiology 136(3):657–664, 1980. 315. Prat R, Galeano I, Lucas A, et al: Relative value of magnetic resonance spectroscopy, magnetic resonance perfusion, and 2-(18F) fluoro-2deoxy-D-glucose positron emission tomography for detection of

http://radiologyebook.com

264

316.

317.

318.

319.

320.

321.

322.

323. 324.

325. 326. 327.

328.

329.

330.

331.

332.

333.

334.

335. 336. 337.

PART II  CT and MR Imaging of the Whole Body recurrence or grade increase in gliomas. J Clin Neurosci 17(1):50–53, 2010. Provenzale JM, McGraw P, Mhatre P, et al: Peritumoral brain regions in gliomas and meningiomas: Investigation with isotropic diffusionweighted MR imaging and diffusion-tensor MR imaging. Radiology 232(2):451–460, 2004. Provenzale JM, Wang GR, Brenner T, et al: Comparison of permeability in high-grade and low-grade brain tumors using dynamic susceptibility contrast MR imaging. AJR Am J Roentgenol 178(3):711–716, 2002. Provenzale JM, Weber AL, Klintworth GK, et al: Radiologic-pathologic correlation. Bilateral retinoblastoma with coexistent pineoblastoma (trilateral retinoblastoma). AJNR Am J Neuroradiol 16(1):157–165, 1995. Puce A, Constable RT, Luby ML, et al: Functional magnetic resonance imaging of sensory and motor cortex: Comparison with electrophysiological localization. J Neurosurg 83(2):262–270, 1995. Pyka T, Gempt J, Ringel F, et al: Prediction of glioma recurrence using dynamic (1)(8)F-fluoroethyltyrosine PET. AJNR Am J Neuroradiol 35(10):1924–1929, 2014. Rabinov JD, Lee PL, Barker FG, et al: In vivo 3-T MR spectroscopy in the distinction of recurrent glioma versus radiation effects: Initial experience. Radiology 225(3):871–879, 2002. Rachinger W, Goetz C, Popperl G, et al: Positron emission tomography with O-(2-[18F]fluoroethyl)-l-tyrosine versus magnetic resonance imaging in the diagnosis of recurrent gliomas. Neurosurgery 57(3):505– 511, 2005. Radkowski MA, Naidich TP, Tomita T, et al: Neonatal brain tumors: CT and MR findings. J Comput Assist Tomogr 12(1):10–20, 1988. Rand SD, Prost R, Haughton V, et al: Accuracy of single-voxel proton MR spectroscopy in distinguishing neoplastic from nonneoplastic brain lesions. AJNR Am J Neuroradiol 18(9):1695–1704, 1997. Razack N, Baumgartner J, Bruner J: Pediatric oligodendrogliomas. Pediatr Neurosurg 28(3):121–129, 1998. Ricci PE: Imaging of adult brain tumors. Neuroimaging Clin N Am 9(4):651–669, 1999. Rijpkema M, Schuuring J, van der Meulen Y, et al: Characterization of oligodendrogliomas using short echo time 1H MR spectroscopic imaging. NMR Biomed 16(1):12–18, 2003. Roberts HC, Roberts TP, Brasch RC, et al: Quantitative measurement   of microvascular permeability in human brain tumors achieved using dynamic contrast-enhanced MR imaging: Correlation with histologic grade. AJNR Am J Neuroradiol 21(5):891–899, 2000. Rock JP, Hearshen D, Scarpace L, et al: Correlations between magnetic resonance spectroscopy and image-guided histopathology, with special attention to radiation necrosis. Neurosurgery 51(4):912–919, 2002. Rock JP, Scarpace L, Hearshen D, et al: Associations among magnetic resonance spectroscopy, apparent diffusion coefficients, and imageguided histopathology with special attention to radiation necrosis. Neurosurgery 54(5):1111–1117, 2004. Rohringer M, Sutherland GR, Louw DF, et al: Incidence and clinicopathological features of meningioma. J Neurosurg 71(5 Pt 1):665–672, 1989. Rollins N, Mendelsohn D, Mulne A, et al: Recurrent medulloblastoma: Frequency of tumor enhancement on Gd-DTPA MR imaging. AJNR Am J Neuroradiol 11(3):583–587, 1990. Roman-Goldstein SM, Goldman DL, Howieson J, et al: MR of primary CNS lymphoma in immunologically normal patients. AJNR Am J Neuroradiol 13(4):1207–1213, 1992. Ropper AH: Lateral displacement of the brain and level of consciousness in patients with an acute hemispheral mass. N Engl J Med 314(15):953–958, 1986. Rorke LB, Trojanowski JQ, Lee VM, et al: Primitive neuroectodermal tumors of the central nervous system. Brain Pathol 7(2):765–784, 1997. Rosenbaum AE, Rosenbloom SB: Meningiomas revisited. Semin Roentgenol 19(1):8–26, 1984. Rosenberg AE, Nielsen GP, Keel SB, et al: Chondrosarcoma of the base of the skull: A clinicopathologic study of 200 cases with emphasis on its distinction from chordoma. Am J Surg Pathol 23(11):1370–1378, 1999.

338. Ross DA, Norman D, Wilson CB: Radiologic characteristics and results of surgical management of Rathke’s cysts in 43 patients. Neurosurgery 30(2):173–178, 1992. 339. Roux FE, Boulanouar K, Ibarrola D, et al: Functional MRI and intraoperative brain mapping to evaluate brain plasticity in patients with brain tumours and hemiparesis. J Neurol Neurosurg Psychiatry 69(4):453–463, 2000. 340. Rumboldt Z, Camacho DLA, Lake D, et al: Apparent diffusion coefficients for differentiation of cerebellar tumors in children. AJNR Am J Neuroradiol 27(6):1362–1369, 2006. 341. Russell DS: Pathology of tumours of the nervous system, Baltimore, 1989, Williams and Wilkins. 342. Russell EJ, George AE, Kricheff II, et al: Atypical computed tomography features of intracranial meningioma: Radiological-pathological correlation in a series of 131 consecutive cases. Radiology 135(3):673– 682, 1980. 343. Russell EJ, Geremia GK, Johnson CE, et al: Multiple cerebral metastases: Detectability with Gd-DTPA-enhanced MR imaging. Radiology 165(3):609–617, 1987. 344. Sage MR: Blood-brain barrier: Phenomenon of increasing importance to the imaging clinician. AJR Am J Roentgenol 138(5):887–898, 1982. 345. Saindane AM, Cha S, Law M, et al: Proton MR spectroscopy of tumefactive demyelinating lesions. AJNR Am J Neuroradiol 23(8):1378– 1386, 2002. 346. Saraf-Lavi E, Bowen BC, Pattany PM, et al: Proton MR spectroscopy   of gliomatosis cerebri: Case report of elevated myoinositol with   normal choline levels. AJNR Am J Neuroradiol 24(5):946–951, 2003. 347. Saraswathy S, Crawford FW, Lamborn KR, et al: Evaluation of MR markers that predict survival in patients with newly diagnosed GBM prior to adjuvant therapy. J Neurooncol 91(1):69–81, 2009. 348. Sartoretti-Schefer S, Wichmann W, Aguzzi A, et al: MR differentiation of adamantinous and squamous-papillary craniopharyngiomas. AJNR Am J Neuroradiol 18(1):77–87, 1997. 349. Scherer HJ: A critical review: The pathology of cerebral gliomas.   J Neurol Psychiatry 3(2):147–177, 1940. 350. Schild SE, Scheithauer BW, Schomberg PJ, et al: Pineal parenchymal tumors. Clinical, pathologic, and therapeutic aspects. Cancer 72(3):870– 880, 1993. 351. Schlemmer HP, Bachert P, Henze M, et al: Differentiation of radiation necrosis from tumor progression using proton magnetic resonance spectroscopy. Neuroradiology 44(3):216–222, 2002. 352. Schorner W, Laniado M, Niendorf HP, et al: Time-dependent changes in image contrast in brain tumors after gadolinium-DTPA. AJNR Am J Neuroradiol 7(6):1013–1020, 1986. 353. Schubert MI, Wilke M, Muller-Weihrich S, et al: Diffusion-weighted magnetic resonance imaging of treatment-associated changes in recurrent and residual medulloblastoma: Preliminary observations in three children. Acta Radiol 47(10):1100–1104, 2006. 354. Shaw EG, Scheithauer BW, O’Fallon JR, et al: Oligodendrogliomas: The Mayo Clinic experience. J Neurosurg 76(3):428–434, 1992. 355. Shepherd CW, Scheithauer BW, Gomez MR, et al: Subependymal giant cell astrocytoma: A clinical, pathological, and flow cytometric study. Neurosurgery 28(6):864–868, 1991. 356. Silver AJ, Ganti SR, Hilal SK: Computed tomography of tumors involving the atria of the lateral ventricles. Radiology 145(1):71–78, 1982. 357. Sinson G, Sutton LN, Yachnis AT, et al: Subependymal giant cell astrocytomas in children. Pediatr Neurosurg 20(4):233–239, 1994. 358. Smirniotopoulos JG: The new WHO classification of brain tumors. Neuroimaging Clin N Am 9(4):595–613, 1999. 359. Smirniotopoulos JG, Murphy FM: The phakomatoses. AJNR Am J Neuroradiol 13(2):725–746, 1992. 360. Smirniotopoulos JG, Rushing EJ, Mena H: Pineal region masses: Differential diagnosis. Radiographics 12(3):577–596, 1992. 361. Smirniotopoulos JG, Yue NC, Rushing EJ: Cerebellopontine angle masses: Radiologic-pathologic correlation. Radiographics 13(5):1131– 1147, 1993.

http://radiologyebook.com

CHAPTER 9  Intracranial Neoplasms 362. Smith EA, Carlos RC, Junck LR, et al: Developing a clinical decision model: MR spectroscopy to differentiate between recurrent tumor and radiation change in patients with new contrast-enhancing lesions. AJR Am J Roentgenol 192(2):45–52, 2009. 363. Smith JS, Cha S, Mayo MC, et al: Serial diffusion-weighted magnetic resonance imaging in cases of glioma: Distinguishing tumor recurrence from postresection injury. J Neurosurg 103(3):428–438, 2005. 364. Snow RB, Johnson CE, Morgello S, et al: Is magnetic resonance imaging useful in guiding the operative approach to large pituitary tumors? Neurosurgery 26(5):801–803, 1990. 365. Soussain C, Ricard D, Fike JR, et al: CNS complications of radiotherapy and chemotherapy. Lancet 374(9701):1639–1651, 2009. 366. Spampinato MV, Smith JK, Kwock L, et al: Cerebral blood volume measurements and proton MR spectroscopy in grading of oligodendroglial tumors. AJR Am J Roentgenol 188(1):204–212, 2007. 367. Spence AM, Muzi M, Mankoff DA, et al: 18F-FDG PET of gliomas at delayed intervals: Improved distinction between tumor and normal gray matter. J Nucl Med 45(10):1653–1659, 2004. 368. Spillane JA, Kendall BE, Moseley IF: Cerebral lymphoma: Clinical radiological correlation. J Neurol Neurosurg Psychiatry 45(3):199–208, 1982. 369. Spoto GP, Press GA, Hesselink JR, et al: Intracranial ependymoma and subependymoma: MR manifestations. AJNR Am J Neuroradiol 11(1):83–91, 1990. 370. Stanescu Cosson R, Varlet P, Beuvon F, et al: Dysembryoplastic neuroepithelial tumors: CT, MR findings and imaging follow-up: A study of 53 cases. J Neuroradiol 28(4):230–240, 2001. 371. Stegman LD, Rehemtulla A, Hamstra DA, et al: Diffusion MRI detects early events in the response of a glioma model to the yeast cytosine deaminase gene therapy strategy. Gene Ther 7(12):1005–1010, 2000. 372. Sternberg EJ, Lipton ML, Burns J: Utility of diffusion tensor imaging   in evaluation of the peritumoral region in patients with primary   and metastatic brain tumors. AJNR Am J Neuroradiol 35(3):439–444, 2014. 373. Stovring J: Descending tentorial herniation: Findings on computed tomography. Neuroradiology 14(3):101–105, 1977. 374. Sugahara T, Korogi Y, Kochi M, et al: Correlation of MR imagingdetermined cerebral blood volume maps with histologic and angiographic determination of vascularity of gliomas. AJR Am J Roentgenol 171(6):1479–1486, 1998. 375. Sugahara T, Korogi Y, Kochi M, et al: Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 9(1):53–60, 1999. 376. Sugahara T, Korogi Y, Tomiguchi S, et al: Posttherapeutic intraaxial brain tumor: The value of perfusion-sensitive contrast-enhanced MR imaging for differentiating tumor recurrence from nonneoplastic contrast-enhancing tissue. AJNR Am J Neuroradiol 21(5):901–909, 2000. 377. Sundaresan N, Galicich JH: Surgical treatment of brain metastases. Clinical and computerized tomography evaluation of the results of treatment. Cancer 55(6):1382–1388, 1985. 378. Sundgren PC, Fan X, Weybright P, et al: Differentiation of recurrent brain tumor versus radiation injury using diffusion tensor imaging in patients with new contrast-enhancing lesions. Magn Reson Imaging 24(9):1131–1142, 2006. 379. Swartz JD, Zimmerman RA, Bilaniuk LT: Computed tomography of intracranial ependymomas. Radiology 143(1):97–9101, 1982. 380. Symon L, Mohanty S: Haemorrhage in pituitary tumours. Acta Neurochir (Wien) 65(1–2):41–49, 1982. 381. Sze G, Soletsky S, Bronen R, et al: MR imaging of the cranial meninges with emphasis on contrast enhancement and meningeal carcinomatosis. AJNR Am J Neuroradiol 10(5):965–975, 1989. 382. Sze G, Uichanco LS, Brant-Zawadzki MN, et al: Chordomas: MR imaging. Radiology 166(1 Pt 1):187–191, 1988. 383. Tally PW, Laws ER, Scheithauer BW: Metastases of central nervous system neoplasms. Case report. J Neurosurg 68(5):811–816, 1988. 384. Tampieri D, Melanson D, Ethier R: MR imaging of epidermoid cysts. AJNR Am J Neuroradiol 10(2):351–356, 1989.

265

385. Tampieri D, Moumdjian R, Melanson D, et al: Intracerebral gangliogliomas in patients with partial complex seizures: CT and MR imaging findings. AJNR Am J Neuroradiol 12(4):749–755, 1991. 386. Tarver RD, Richmond BD, Klatte EC: Cerebral metastases from lung carcinoma: Neurological and CT correlation. Work in progress. Radiology 153(3):689–692, 1984. 387. Teramoto A, Hirakawa K, Sanno N, et al: Incidental pituitary lesions in 1,000 unselected autopsy specimens. Radiology 193(1):161–164, 1994. 388. Tien RD, Barkovich AJ, Edwards MS: MR imaging of pineal tumors. AJR Am J Roentgenol 155(1):143–151, 1990. 389. Tien RD, Cardenas CA, Rajagopalan S: Pleomorphic xanthoastrocytoma of the brain: MR findings in six patients. AJR Am J Roentgenol 159(6):1287–1290, 1992. 390. Tien RD, Newton TH, McDermott MW, et al: Thickened pituitary stalk on MR images in patients with diabetes insipidus and Langerhans cell histiocytosis. AJNR Am J Neuroradiol 11(4):703–708, 1990. 391. Tomura N, Narita K, Izumi J-I, et al: Diffusion changes in a tumor and peritumoral tissue after stereotactic irradiation for brain tumors: Possible prediction of treatment response. J Comput Assist Tomogr 30(3):496–500, 2006. 392. Tortori-Donati P, Fondelli MP, Rossi A, et al: Medulloblastoma in children: CT and MRI findings. Neuroradiology 38(4):352–359, 1996. 393. Traber F, Block W, Flacke S, et al: 1H-MR Spectroscopy of brain tumors in the course of radiation therapy: Use of fast spectroscopic imaging and single-voxel spectroscopy for diagnosing recurrence. Rofo 174(1):33–42, 2002. 394. Tropine A, Vucurevic G, Delani P, et al: Contribution of diffusion tensor imaging to delineation of gliomas and glioblastomas. J Magn Reson Imaging 20(6):905–912, 2004. 395. Truwit CL, Barkovich AJ: Pathogenesis of intracranial lipoma:   An MR study in 42 patients. AJNR Am J Neuroradiol 11(4):665–674, 1990. 396. Tsien C, Galban CJ, Chenevert TL, et al: Parametric response map as an imaging biomarker to distinguish progression from pseudoprogression in high-grade glioma. J Clin Oncol 28(13):2293–2299, 2010. 397. Uematsu Y, Yukawa S, Yokote H, et al: Meningeal melanocytoma: Magnetic resonance imaging characteristics and pathological features. Case report. J Neurosurg 76(4):705–709, 1992. 398. Van Laere K, Ceyssens S, Van Calenbergh F, et al: Direct comparison of 18F-FDG and 11C-methionine PET in suspected recurrence of glioma: Sensitivity, inter-observer variability and prognostic value. Eur J Nucl Med Mol Imaging 32(1):39–51, 2005. 399. Vates GE, Chang S, Lamborn KR, et al: Gliomatosis cerebri: A review of 22 cases. Neurosurgery 53(2):261–271, 2003. 400. Vazquez E, Lucaya J, Castellote A, et al: Neuroimaging in pediatric leukemia and lymphoma: Differential diagnosis. Radiographics 22(6):1411–1428, 2002. 401. Venkatesh SK, Gupta RK, Pal L, et al: Spectroscopic increase in choline signal is a nonspecific marker for differentiation of infective/ inflammatory from neoplastic lesions of the brain. J Magn Reson Imaging 14(1):8–15, 2001. 402. Verma N, Cowperthwaite MC, Burnett MG, et al: Differentiating tumor recurrence from treatment necrosis: A review of neuro-oncologic imaging strategies. Neuro Oncol 15(5):515–534, 2013. 403. Vuori K, Kankaanranta L, Hakkinen A-M, et al: Low-grade gliomas and focal cortical developmental malformations: Differentiation with proton MR spectroscopy. Radiology 230(3):703–708, 2004. 404. Warren KE, Frank JA, Black JL, et al: Proton magnetic resonance spectroscopic imaging in children with recurrent primary brain tumors. J Clin Oncol 18(5):1020–1026, 2000. 405. Wasserstrom WR, Glass JP, Posner JB: Diagnosis and treatment of leptomeningeal metastases from solid tumors: Experience with 90 patients. Cancer 49(4):759–772, 1982. 406. Weber WA, Wester H-J, Grosu AL, et al: O-(2-[18F]fluoroethyl)-Ltyrosine and l-[methyl-11C]methionine uptake in brain tumours: Initial results of a comparative study. Eur J Nucl Med 27(5):542–549, 2000.

http://radiologyebook.com

266

PART II  CT and MR Imaging of the Whole Body

407. Weckesser M, Langen KJ, Rickert CH, et al: O-(2-[18F]fluorethyl)-Ltyrosine PET in the clinical evaluation of primary brain tumours. Eur J Nucl Med Mol Imaging 32(4):422–429, 2005. 408. Weir B, Grace M: The relative significance of factors affecting postoperative survival in astrocytomas, grades one and two. Can J Neurol Sci 3(1):47–50, 1976. 409. Weybright P, Sundgren PC, Maly P, et al: Differentiation between brain tumor recurrence and radiation injury using MR spectroscopy. AJR Am J Roentgenol 185(6):1471–1476, 2005. 410. Wong TZ, Turkington TG, Hawk TC, et al: PET and brain tumor image fusion. Cancer J 10(4):234–242, 2004. 411. Xu M, See SJ, Ng WH, et al: Comparison of magnetic resonance spectroscopy and perfusion-weighted imaging in presurgical grading of oligodendroglial tumors. Neurosurgery 56(5):919–926, 2005. 412. Yang S, Law M, Zagzag D, et al: Dynamic contrast-enhanced perfusion MR imaging measurements of endothelial permeability: Differentiation between atypical and typical meningiomas. AJNR Am J Neuroradiol 24(8):1554–1559, 2003. 413. Yasargil MG, von Ammon K, von Deimling A, et al: Central neurocytoma: Histopathological variants and therapeutic approaches.   J Neurosurg 76(1):32–37, 1992. 414. Yousem DM, Arrington JA, Zinreich SJ, et al: Pituitary adenomas: Possible role of bromocriptine in intratumoral hemorrhage. Radiology 170(1 Pt 1):239–243, 1989. 415. Yousem DM, Patrone PM, Grossman RI: Leptomeningeal metastases: MR evaluation. J Comput Assist Tomogr 14(2):255–261, 1990. 416. Yu CS, Li KC, Xuan Y, et al: Diffusion tensor tractography in patients with cerebral tumors: A helpful technique for neurosurgical planning and postoperative assessment. Eur J Radiol 56(2):197–204, 2005.

417. Yuh WT, Fisher DJ, Runge VM, et al: Phase III multicenter trial of high-dose gadoteridol in MR evaluation of brain metastases. AJNR Am J Neuroradiol 15(6):1037–1051, 1994. 418. Zee CS, Segall HD, Ahmadi J, et al: Computed tomography of posterior fossa ependymomas in childhood. Surg Neurol 20(3):221–226, 1983. 419. Zee CS, Segall H, Apuzzo M, et al: MR imaging of pineal region neoplasms. J Comput Assist Tomogr 15(1):56–63, 1991. 420. Zee CS, Segall HD, Miller C, et al: Less common CT features of medulloblastoma. Radiology 144(1):97–9102, 1982. 421. Zeng QS, Li CF, Liu H, et al: Distinction between recurrent glioma and radiation injury using magnetic resonance spectroscopy in combination with diffusion-weighted imaging. Int J Radiat Oncol Biol Phys 68(1):151–158, 2007. 422. Zentner J, Wolf HK, Ostertun B, et al: Gangliogliomas: Clinical, radiological, and histopathological findings in 51 patients. J Neurol Neurosurg Psychiatry 57(12):1497–1502, 1994. 423. Zhang M, Olsson Y: Hematogenous metastases of the human brain–Characteristics of peritumoral brain changes: A review.   J Neurooncol 35(1):81–89, 1997. 424. Zimmerman RA: Neuroimaging of primary brainstem gliomas: Diagnosis and course. Pediatr Neurosurg 25(1):45–53, 1996. 425. Zimmerman RA, Bilaniuk LT: Computed tomography of acute intratumoral hemorrhage. Radiology 135(2):355–359, 1980. 426. Zimmerman RD, Fleming CA, Lee BC, et al: Periventricular hyperintensity as seen by magnetic resonance: Prevalence and significance. AJR Am J Roentgenol 146(3):443–450, 1986.

http://radiologyebook.com

10  Cerebral Infections and Inflammation Alexander Lerner, Anandh Rajamohan, Mark S. Shiroishi, John L. Go, Paul E. Kim, Chia-Shang J. Liu, and Chi-Shing Zee

Computed tomography (CT) and magnetic resonance imaging (MRI) have become indispensable for diagnosis, evaluation, and treatment of intracranial infectious disease, contributing to significant decline in morbidity and mortality of these patients.110 CT is useful as a rapid screening tool to detect complications of intracranial infections such as hydrocephalus, extraaxial collections, and mass effect. MRI is superior for assessment of intracranial infections and allows for a more sensitive evaluation of parenchymal lesions and leptomeningeal disease. CT angiography (CTA) is frequently used for detection of vascular complications such as vasculitis and mycotic aneurysms, with conventional cerebral angiography reserved for selected difficult cases and endovascular interventions. Implementation of advanced MRI techniques such as MR perfusion imaging, diffusion imaging, and proton MR spectroscopy (MRS) has improved differentiation of neoplastic processes from intracranial infections and allowed for more accurate characterization of these disorders. Functional MRI (fMRI) and diffusion tensor imaging (DTI) have allowed for improved preoperative planning when surgical intervention is necessary. There are four common routes of spread of infectious disease to the intracranial compartment. 1. Hematogenous spread is the most common route, with pathogens from a distant site in the body disseminating to the choroid plexus, meninges, and brain parenchyma. In cerebral parenchyma the pathogens usually lodge within small vessels, resulting in a subcortical distribution of such lesions. Anomalous hemodynamics, such as those seen in children with cyanotic heart disease, have a high incidence of intracranial infection secondary to a right-to-left shunt.112 Retrograde venous spread can also occur via anastomotic connections between superficial veins of the face, scalp, and cortical veins. 2. Direct extension from an adjacent site of infection, such as mastoiditis, otitis media, sinusitis, or orbital infection is another possibility. 3. Direct intracranial introduction of a pathogen may occur during trauma or as a complication of surgery or lumbar puncture.109 4. Retrograde spread to the central nervous system (CNS) via peripheral nerves can occur in herpes simplex and rabies virus infections.89 Multiple coexisting conditions may increase the risk of intra­ cranial infections (e.g., diabetes mellitus, alcoholism, malignancy,  agammaglobulinemia, radiation therapy, chemotherapy, and steroid therapy). Patients with acquired immunodeficiency syndrome (AIDS) are particularly susceptible to intracranial infection; however, the disease spectrum and imaging manifestations in these patients are different.141

A variety of infectious disorders involve the CNS. Pyogenic infection is discussed in detail in this chapter, with categorization according to the anatomic site of involvement. Other infectious processes are also discussed, including viral infections, granulomatous disease, spirochete infections, fungal diseases, parasitic diseases, and inflammatory disease.

LEPTOMENINGITIS Three distinct layers of connective tissue known as the meninges surround the brain within the intracranial compartment. The dura mater layer is also referred to as the pachymeninges. The thinner inner arachnoid and pial meningeal layers are referred to as leptomeninges. There are two principal patterns of meningeal enhancement on MRI: leptomeningeal and pachymeningeal enhancement. Because the dura mater lacks the blood-brain barrier, normal pachymeninges usually demonstrate only minimal discontinuous linear enhancement except for dural venous sinuses. The leptomeninges usually demonstrate no enhancement except for adjacent vascular structures. Meningitis primarily involves the leptomeninges (pia mater and arachnoid membrane) and the adjacent subarachnoid space, resulting in a leptomeningeal pattern of enhancement.64,89 The diagnosis of meningitis is frequently made by clinical history and cerebrospinal fluid (CSF) analysis. Clinical presentation in adults may include the classic triad of fever, neck stiffness, and altered mental status, as well as headache and photophobia. CSF analysis in acute pyogenic meningitis classically demonstrates high protein, low glucose, and neutrophilic pleocytosis. CT of the head should be obtained prior to lumbar puncture in patients with suspected increased intracranial pressure, focal neurologic deficits, or moderate to severe impairment of consciousness.57 Imaging may also be used to confirm meningitis and evaluate for complications such as abscess, ventriculitis, and subdural empyema. Additionally, cross-sectional imaging may elucidate the origin of intracranial infection, because CT and MRI are sensitive for sinusitis and mastoiditis. The infectious agents responsible for acute pyogenic meningitis vary by age group and immune status. In neonates, causes include group B Streptococcus and Escherichia coli. In infants and young children, typical pathogens may include Haemophilus influenzae and Neisseria meningitides. In immunocompetent adults, Streptococcus pneumoniae and Listeria monocytogenes are the common causative agents. In immunocompromised adults, causative agents for acute meningitis include Klebsiella, Pseudomonas, E. coli, and fungal organisms.134

http://radiologyebook.com

267

268

PART II  CT and MR Imaging of the Whole Body

Initial CT and MRI of the brain may be normal in uncomplicated cases of meningitis. Occasionally, in more advanced meningitis, noncontrast CT may demonstrate basal cistern obliteration. This may represent increased attenuation of exudate in the subarachnoid space  and thickened, inflamed leptomeninges. Ventricular enlargement and diffuse cerebral swelling may also be seen in more advanced cases of meningitis. Postcontrast CT may show enhancement of leptomeninges in the basal cisterns and sylvian fissures.135 MRI is more sensitive for detection of meningitis, especially when using fluid-attenuated inversion recovery (FLAIR) sequences. T2-  FLAIR is sensitive to subtle changes in CSF and may demonstrate hyperintensity in the subarachnoid space (presumably due to exudate and elevated protein in this region) even when T1-weighted images may appear normal. However, T2-FLAIR sulcal hyperintensity may also be seen in other subarachnoid space processes such as leptomeningeal carcinomatosis and subarachnoid hemorrhage.65,124 Postcontrast T2-FLAIR has been reported to have high sensitivity and specificity for leptomeningeal enhancement.100 T1-weighted images may show obliteration of the basal cisterns, and T2-weighted images may show cortical hyperintensity. Postcontrast T1-weighted images may show linear continuous sulcal or cisternal enhancement, frequently involving the basal cisterns or sylvian fissures and extending deep into the base of the sulci. Sulcal or cisternal enhancement on three contiguous slices has been reported to correlate with significant meningeal pathology107 (Fig. 10-1). On postcontrast MRI and CT imaging, differential diagnosis of abnormal subarachnoid enhancement includes meningitis, lepto­ meningeal carcinomatosis, and subacute subarachnoid hemorrhage, although the clinical presentation is usually very different in these entities. Contrast-enhanced MRI has been shown to be more sensitive than contrast-enhanced CT in detection of meningitis and its complications.135,141 Leptomeningeal enhancement on postcontrast MRI is seen in only 50% of meningitis cases.89

A

Thin linear enhancement most evident along cerebral convexities is typically seen in acute pyogenic or viral meningitis. Nodular thick enhancement most evident in the basal cisterns is typically seen in chronic meningitides such as tuberculous meningitis, fungal meningitis, and sarcoidosis. Granulomatous disease may also result in pachymeningeal involvement. Pachymeningeal thickening and enhancement may be seen in a variety of disorders including intracranial hypotension, meningiomas, metastatic disease, lymphoma, and granulomatous disease.89 Postoperative meningeal enhancement may be pachymeningeal or leptomeningeal in appearance. Occasionally, pachymeningeal enhancement can be seen in patients who had uncomplicated lumbar puncture.87 Complications of meningitis include hydrocephalus, ventriculitis, subdural effusions, subdural empyema, epidural empyema, cranial nerve inflammation, cerebritis, abscess, and infarction due to arterial or venous involvement.18,61,64,135 Hydrocephalus is the most common complication and usually presents as transient communicating hydrocephalus due to subarachnoid space inflammation and secondary decreased CSF resorption in pacchionian granulations. Noncommunicating hydrocephalus may also occur, especially when meningitis is complicated by ventriculitis.89 Magnetization transfer (MT) sequences have been reported to improve visualization of abnormal meninges by showing more prominent enhancement on postcontrast images. These sequences may also demonstrate basal cistern hyperintensity on noncontrast images in tuberculous meningitis. MT ratio (MTR) may help determine the cause of meningitis, with tuberculous meningitis reported to show lower MTR than fungal and pyogenic meningitis and higher MTR than viral meningitis.51,63 Recent investigations of DTI have demonstrated high sensitivity  in detecting bacterial meningitis, showing increased fractional anisotropy (FA) within enhancing and nonenhancing cortex. High FA is thought to be due to increased orientation of inflammatory cells in the

B FIG 10-1  Meningitis. A, Contrast-enhanced MRI study shows enhancement along the leptomeninges in cerebral sulci bilaterally. B, Sulcal hyperintensities are seen bilaterally along the convexities of temporal and occipital lobes on this T2-FLAIR image.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

269

pia-arachnoid meninges closely related to the cortical ribbon.93 FA of the cortex has also shown a strong correlation with neuroinflammatory markers in neonatal meningitis.138

restriction also supports the diagnosis of empyema.18 Hyperintense signal on T2-FLAIR and T2-weighted images compared to CSF is not specific, because it may be seen in subdural effusions and empyemas.

SUBDURAL EFFUSION

EPIDURAL EMPYEMA

Subdural effusion is a sterile fluid collection between the dura and the arachnoid. Reactive subdural effusions are common complications of acute bacterial meningitis, especially in infants. S. pneumoniae and H. influenzae meningitis are more frequently associated with this complication.61 These collections are similar to posttraumatic and postsurgical subdural effusions and may result from tears in the arachnoid membrane and subsequent leakage of CSF into the subdural space. Alternatively, subdural effusions may be caused by inflammation of subdural veins or dural irritation. CT and MRI demonstrate extraaxial collections with attenuation and signal intensity similar to CSF, which are most frequently seen  as large bilateral fluid collections over the frontal and temporal convexities. Subdural effusions may be isointense or hyperintense to CSF on FLAIR images, depending on protein concentration of the fluid. Lack of transiting venous structures within these collections on T2-weighted images allows for differentiation from prominent subarachnoid spaces. There is no diffusion restriction seen within these collections, and typically they show no significant enhancement.61 Occasionally, fibrin strands and membranes develop; they may enhance and appear similar to subdural empyema. Subdural effusions regress spontaneously and do not require intervention except when they are large and result in significant mass effect.18,89

Epidural empyema, or epidural abscess, is a purulent collection between the dura and the calvarium and usually occurs as a complication of otitis media, mastoiditis, sinusitis, or osteomyelitis of the skull.54 Ten percent of epidural empyemas may be associated with subdural empyema.141 Epidural empyema is separated from the subarachnoid space and brain parenchyma by the dura, which may act as a barrier to infection. As a result this collection usually has a more insidious and benign clinical course.89 Owing to its extradural location, this collection usually has biconvex morphology and may cross and displace the dural venous sinuses but does not cross the calvarial sutures. Typically an avidly enhancing T2 hypointense rim is seen. Adjacent parenchymal involvement is rare (Fig. 10-3). Imaging characteristics are otherwise similar to subdural empyema.

SUBDURAL EMPYEMA Subdural empyema is a purulent collection between the dura and the arachnoid; it may arise as a complication of acute pyogenic meningitis, mastoiditis, otitis media, paranasal sinusitis, trauma, or postsurgical infection.39,136 The adjacent infection is thought to spread to the subdural space via retrograde thrombophlebitis of the calvarial emissary veins, bridging veins, or dural venous sinuses. More rarely, an infection of a subdural effusion may occur, resulting in a subdural empyema.39,59 These collections may cross the calvarial sutures but do not cross the midline. Subdural empyemas represent up to 15% of all extraaxial collections18 and are seen in 13% to 20% of all cases of intracranial infection.141 Subdural empyema may progress and result in multiple compli­ cations such as thrombophlebitis, dural venous sinus thrombosis, infarction, cerebritis, and abscess formation. For this reason it usually requires aggressive intervention, and early accurate diagnosis is critical.64,136 CT imaging usually demonstrates a hypodense to isodense crescentic or lenticular extraaxial collection along the cerebral convexity or falx, which frequently appears similar to a subdural effusion. MRI  is extremely useful when empyema is suspected, because it allows  for reliable diagnosis of this entity and evaluation of associated complications. Diffusion restriction is the most useful characteristic for diagnosis of empyema, which is seen as very high signal on diffusion-weighted images (DWI) and low values on apparent diffusion coefficient (ADC) maps. In contradistinction, subdural effusions show signal intensity similar to CSF on DWI. Empyemas are also typically slightly hyperintense to CSF on T1-weighted images and usually demonstrate rim enhancement (Fig. 10-2). The presence of adjacent cerebritis, abscess, or infarction with associated parenchymal enhancement or diffusion

EPENDYMITIS Ependymitis or ventriculitis is inflammation of the ependyma along the inner surface of the cerebral ventricles. Intraventricular infection may result from intraventricular rupture of a cerebral abscess, hematogenous spread to subependyma or choroid plexus, retrograde spread from extraventricular CSF spaces, trauma, or surgery. Staphylococcus and Enterobacter are the most common causative agents. Imaging plays an important role in diagnosis; signs and symptoms may be subtle.43 Noncontrast CT frequently shows no significant abnormality, although occasionally dependent debris or slightly hyperdense ependyma may suggest the diagnosis. MRI findings include enhancing thickened ependyma, T2 hyperintense ventricular walls, dependent intraventricular debris, and ventricular enlargement. Dependent debris is a characteristic finding in ventriculitis and usually demonstrates diffusion restriction and hyperintensity on T2-FLAIR images43 (Fig. 10-4). Intraventricular adhesions and septations may cause noncommunicating hydrocephalus or result in a trapped ventricle or segmental dilatation.18,59,61,89

CEREBRITIS AND ABSCESS Cerebritis is a poorly defined regional acute inflammation of cerebral parenchyma with increased local vascular permeability, which is usually caused by pyogenic bacteria. Untreated cerebritis evolves into a cerebral abscess, which is characterized as a focal parenchymal collection of pus with a surrounding vascular collagenous capsule. Britt and Enzmann divided the process of abscess formation into four stages: early cerebritis, late cerebritis, early capsule formation, and late capsule formation.15 1. Early or acute cerebritis (first 3-5 days). Bacteria invade the parenchyma and trigger an immune response, with infiltration by inflammatory cells, increased vascular permeability of the blood-brain barrier, perivascular exudates, petechial hemorrhages, microthrombosis, early necrosis, and associated edema. Imaging is rarely performed at this stage, because the patient may still be asymptomatic. CT shows a poorly defined area  of parenchymal hypodensity, with mass effect and variable enhancement, including no enhancement, mild nodular, or ring enhancement. MRI is more sensitive for detection of cerebritis and abscess. A poorly defined parenchymal area of edema with

http://radiologyebook.com

270

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 10-2  Subdural empyema. A, T2-FLAIR image demonstrates a right frontoparietal hyperintense subdural fluid collection. In addition, hyperintense fluid is seen in the adjacent cerebral sulci. The adjacent cortex is hyperintense. B, Diffusion-weighted image shows hyperintensity of the subdural collection. ADC (apparent diffusion coefficient) map confirmed restricted diffusion by demonstrating low ADC values in the collection. C and D, Axial and coronal postcontrast T1-weighted images show linear enhancement along the lateral and medial walls of the collection, with adjacent sulcal leptomeningeal enhancement.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

271

B FIG 10-3  Epidural empyema. A, T2-FLAIR image demonstrates a left frontal lentiform hyperintense epidural fluid collection. A low signal rim is seen posteriorly, representing the dura. B, Postcontrast T1-weighted image shows rim enhancement of this collection. Notice the lack of leptomeningeal or parenchymal enhancement. There was restricted diffusion within the collection (not shown). This collection extended across the midline above the superior sagittal sinus (not shown).

T2 hyperintensity and T1 iso- to hypointensity is seen with variable ill-defined enhancement39,109 (Fig. 10-5). 2. Late cerebritis (second week). There is progression of necrosis with formation of a large necrotic zone. Peripheral fibroblast deposition occurs, with formation of reticulin matrix. MRI and CT demonstrate thick, irregular, ring enhancement around a central necrotic zone with a surrounding area of edema. Central necrosis shows decreased attenuation on CT, with T1 hypointensity, T2 hyperintensity, and diffusion restriction on MRI. A rim may be seen around the necrotic zone on MRI, which is T1 iso- to hyperintense and T2 iso- to hypointense39 (Fig. 10-6). 3. Early abscess—early capsule formation (end of second week). Granulation tissue forms at the periphery of the abscess. A thin capsule is formed by fibroblasts, with reticulin matrix and scant collagen. Liquefaction occurs in the central necrotic zone, with reduction in surrounding edema. CT frequently demonstrates iso- to slightly hyperdense capsule interposed between the hypodense central area and surrounding hypodense edema (Fig. 10-7). MRI demonstrates a well-demarcated T2 hypointense capsule surrounding a central T2 hyperintense area, demonstrating diffusion restriction. Typically there is well-defined avid enhancement of the capsule, which may be thinner on the side closer to the ventricle (Fig. 10-8). Daughter abscesses may appear at this stage.39 4. Mature abscess—late capsule formation (after second week). A thick collagen capsule is formed, with surrounding gliosis. The wall of the mature abscess consists of three layers: (1) an inner inflammatory layer of granulation tissue containing macro-

phages, (2) a middle collagenous layer, and (3) an outer gliotic layer. The capsule may become thinner and better defined as surrounding edema decreases. Over time and especially with treatment, the capsule may become less T2 hypointense, enhancement may decrease, and the abscess may gradually decrease in size.39 Cerebral abscesses are caused by multiple mechanisms, including direct spread from a local infection, hematogenous dissemination from a remote infection, trauma, or postsurgical infection. Up to 30% of cerebral abscesses are reported to be cryptogenic. Sources of direct spread include sinusitis, otitis media, mastoiditis (Fig. 10-9), and odontogenic infections, with pathogens invading the brain parenchyma via adjacent osteomyelitis or via emissary veins.109 Sources of hematogenous seeding include endocarditis, pulmonary abscesses, and abdominal infections. Congenital cyanotic heart disease and pulmonary arteriovenous malformation patients are at high risk for hematogenous dissemination of infections to the brain. Presenting symptoms depend on size and location of the abscess and may include headache, fever, nausea and vomiting, neurologic deficits, and seizure. Absence of leukocytosis and normal CSF analysis do not exclude cerebral abscess, and CSF cultures are often sterile. Infectious agents include microaerophilic streptococci, anaerobic bacteria, Staphylococcus aureus, facultative gram-negative bacteria, and polymicrobial infections. Nocardial abscesses have been associated with high mortality. There are three characteristic findings of a cerebral abscess on MRI: (1) central liquefied cavity with diffusion restriction; (2) smooth, thin, continuous capsular enhancement; (3) T2 hypointense signal in the capsule. Frequently there is thinning along the medial wall of the

http://radiologyebook.com

272

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 10-4  Ventriculitis. A, T2-FLAIR image demonstrates marked ventricular enlargement and periventricular parenchymal hyperintensity, most prominently on the right. B, Postcontrast T1-weighted image shows enhancement of the ependymal lining of the lateral ventricles. Notice the lack of leptomeningeal or parenchymal enhancement. C, Diffusion-weighted image shows an irregular region of restricted diffusion in the dependent portion of the right lateral ventricle, presumably representing interventricular debris.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

B

A

FIG 10-5  Early cerebritis. A, T2-weighted image demonstrates two ill-defined areas of high signal intensity involving the cortex and subcortical white matter of the left frontal lobe. B, Gadolinium-enhanced MRI study shows mild ill-defined peripheral enhancement. Subsequent imaging (not shown) demonstrated formation of two mature abscesses in these areas.

A

B FIG 10-6  Late cerebritis. A, Gadolinium-enhanced MRI study shows thick, smooth, ringlike enhancement with surrounding edema. B, Gadolinium-enhanced MRI study (coronal view) shows a second small adjacent ringlike enhancement.

http://radiologyebook.com

273

274

PART II  CT and MR Imaging of the Whole Body

FIG 10-7  Mature abscess. Noncontrast CT demonstrates a central hypodense cavity with an iso- to slightly hyperdense rim and surrounding edema.

capsule. The capsule is usually T1 iso- to hyperintense compared to gray matter (see Fig. 10-8). Usually only a moderate amount of edema is seen surrounding a mature abscess. Gas-forming organisms may result in gas accumulation within the central liquefied zone. Typical T2 hypointensity of the capsule is thought to be due to the presence of free radicals, which are generated by macrophages within the wall of the abscess (see Fig. 10-8A). Satellite lesions associated with an abscess also support the diagnosis (Fig. 10-10). Clinical improvement is correlated to decrease in enhancement and surrounding edema on CT and MRI. Complications of cerebral abscess include hydrocephalus, meningitis, subdural empyema, and thrombophlebitis. Ependymitis resulting from rupture of an abscess into the ventricular system is a severe  complication with high mortality. The medial wall of the abscess is often thinner owing to less extensive blood supply in periventricular white matter, which predisposes an abscess to rupture medially (Fig. 10-11). The classic ring enhancing appearance of cerebral abscess is not specific. CT and MRI imaging differential diagnosis includes highgrade primary CNS neoplasm, metastasis, nonpyogenic infection, radiation necrosis, demyelinating lesion, thrombosed aneurysm, subacute infarct, and resolving hematoma. Hematoma demonstrates typical MRI signal patterns of blood products, and infarcts demonstrate gyriform enhancement or cortical laminar necrosis, leading to the correct diagnosis. Neoplasms and granulomatous diseases typically have thicker and more nodular enhancement. Generally the presence of characteristic imaging features of cerebral abscess offers high specificity and sensitivity for this diagnosis. The presence of central diffusion restriction (hyperintensity on DWI and low values on ADC map) is particularly important in differentiation of abscess from other pathology and has been reported to have sensitivity of 95.2% and specificity of 95.7%68 (Fig. 10-12; see also Fig. 10-8). In contradistinction, necrotic tumors usually demonstrate central diffusion facilitation.

Diffusion restriction within a liquefied center of an abscess is now thought to be primarily due to high density of viable inflammatory cells in pus, with high viscosity, macromolecules, and necrotic debris as possible secondary contributors.86 Recent investigations suggest that diffusion imaging may be helpful to follow up evaluation of treatment response, with high ADC reported in treated abscesses and low ADC after therapy reported in treatment failure.17,38 The features of abscess tend to differ in immunocompromised patients receiving steroids. In these patients, edema may be minimal and the abscess capsule may be thick and irregular. Hypointensity  of the capsule seen on T2-weighted images may not be obvious.  Extraparenchymal spread into the ventricles and meningitis are more common.141 Thallium-201 has been investigated for differentiation of abscess from necrotic tumor. However, a report of a false-positive result with uptake in a brain abscess suggests that this technique may not be reliable.82 However, this technique has been shown to be useful in differentiation of CNS lymphoma from Toxoplasmosis gondii infection in HIV patients. Advanced MRI imaging techniques such as 1H MRS, DTI, and MR perfusion have been recently employed to assist in differentiation of cerebral abscess from other lesions such as neoplasms with central necrosis. The most characteristic MRS findings within the liquefied nonenhancing center of a mature cerebral abscess include presence of cytosolic amino acids (valine, leucine, and isoleucine) at 0.9 ppm, lactate at 1.3 ppm, and absence of normal metabolites such as N-acetylaspartate (NAA), choline, and creatine. Acetate at 1.9 ppm and succinate at 2.4 ppm may also be present. Lipid and lactate are not specific and may be seen in tumors and abscesses. Detection of amino acids is  the most useful finding for diagnosis of abscess; these metabolites are not detected in vivo in necrotic or cystic neoplasms.23,50 Amino acids are generated during protein breakdown in pus owing to release  of proteolytic enzymes by neutrophils. Amino acids, acetate, and  succinate are generated during glycolysis and may be detected in abscesses caused by anaerobes or facultative anaerobes.69,97 Amino acids, acetate, and succinate have been reported to become undetectable after treatment.16 Investigation of MR perfusion by Chiang et al. demonstrated lower mean relative cerebral blood volume (rCBV) in the enhancing wall of an abscess when compared to the enhancing rim of a necrotic tumor; this is due to the relatively poor vascularity of the abscess wall.23 Recent studies using DTI in brain abscess showed increased fractional anisotropy (FA) within the abscess’s liquefied center. High FA is thought to be due to organized orientation of inflammatory cells, caused by neuroinflammatory molecules in pus, which may assist in differentiation from a necrotic tumor.92 Additionally DTI may be useful for evaluation of treatment response, because FA has been reported to decline during treatment, while mean diffusivity (MD) was relatively unchanged.94,109

ENCEPHALITIS Encephalitis is a diffuse inflammation of the brain with clinical evidence of brain dysfunction, typically caused by a viral infection. Infectious encephalitis may present with fever, headache, and altered level of consciousness. Viral causes include herpes simplex virus type 1 (HSV-1), herpes zoster, arbovirus, and enteroviruses in adults; and herpes simplex virus type 2 (HSV-2), cytomegalovirus (CMV), rubella, and human immunodeficiency virus (HIV) in neonates. In immunocompromised patients, common viral causes include HIV, CMV, John Cunningham virus (JC virus), and human herpesvirus 6 (HHV-6).

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

B

C

D FIG 10-8  Mature abscess. A, T2-weighted image shows a mature abscess with a hypointense rim, central cavity, and adjacent surrounding edema. B, Diffusion-weighted image demonstrates internal high signal intensity. Corresponding ADC images (not shown) demonstrated low values representing diffusion restriction. C, T1-weighted image demonstrates hypointense central cavity with iso- to hyperintense rim. D, T1-weighted postcontrast image demonstrates smooth ringlike enhancement corresponding to the hypointense rim seen on T2-weighted image.

http://radiologyebook.com

275

276

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 10-9  Mastoiditis causing meningitis, subdural empyema, cerebritis, thrombophlebitis, and abscess. A and B, T1-weighted postcontrast coronal and axial images demonstrate enhancement in the right mastoid bone and central filling defect with peripheral enhancement in the adjacent sigmoid sinus (arrow). Coronal image demonstrates rim enhancing abscess in the adjacent inferior temporal lobe (arrowhead). C, T2-FLAIR image demonstrates right temporal subdural, sulcal, cortical and subcortical hyperintensity due to meningitis, subdural empyema, cerebritis, and abscess formation. D, Diffusion-weighted image demonstrates ill-defined temporal lobe hyperintensity representing cerebritis. There is also focal diffusion restriction in the posterior right temporal lobe corresponding to a rim enhancing abscess on postcontrast images. Anterior temporal convexity subdural empyema also shows diffusion restriction.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

277

Nonviral infectious encephalitis may also be seen in immunocom­ promised patients, with causative agents such as Toxoplasma, Asper­ gillus, and Listeria monocytogenes. Noninfectious encephalitides of autoimmune etiology include numerous entities such as acute disseminated encephalomyelitis (ADEM), Rasmussen encephalitis, and anti-NMDAR-mediated encephalitis.62,109,133

Herpes Simplex Encephalitis

Herpes Simplex Type 1 (HSV-1) Encephalitis.  HSV-1 is estimated

FIG 10-10  Mature abscess with daughter abscesses. Coronal postcontrast T1-weighted image demonstrates a large ring enhancing lesion with surrounding hypointense edema and multiple adjacent small ring enhancing lesions.

A

to cause 10% of the encephalitis cases in the United States and is the most common cause of sporadic encephalitis worldwide. HSV-1 encephalitis occurs in children, adults, and immunocompromised patients. Infection occurs early in life, with virus remaining in latent form in the vagus nerve, sacral nerves, and trigeminal ganglia. Reactivation later in life may result in rapidly progressive encephalitis with headache, fever, seizures, and altered mental status. Untreated HSV-1 encephalitis may result in death within 2 weeks, with a mortality rate of approximately 70%. For this reason early diagnosis is critical, with immediate treatment with acyclovir in clinically suspected cases.19,133 Characteristic findings on MRI include bilateral asymmetric T2-  FLAIR hyperintensities in the cortex and subcortical white matter, in addition to sulcal effacement involving the hippocampi and mesial temporal lobes. The inferior frontal lobes, insula, and cingulate gyri may also be involved. The basal ganglia are spared, helping in differentiation from other encephalitides that frequently involve the basal ganglia. Patchy diffusion restriction may be seen early and may precede

B FIG 10-11  Abscess causing ventriculitis in a child. A, Gadolinium-enhanced MRI study shows a rim enhancing right peritrigonal abscess with rim enhancement rupturing into the right atrium. Ependymal enhancement is seen in the posterior right lateral ventricle, consistent with ventriculitis. Enlarged enhancing choroid plexus represents choroid plexitis. B, Diffusion-weighted images (DWI) demonstrate marked hyperintensity within the abscess cavity extending into the right lateral ventricle. Note the dependent DWI hyperintense debris in the left atrium. ADC images (not shown) confirmed diffusion restriction in these areas.

http://radiologyebook.com

278

PART II  CT and MR Imaging of the Whole Body T2-FLAIR hyperintensities. Petechial hemorrhages may be demonstrated in areas of involvement. Later in the course of disease, patchy enhancement may also be seen (Fig. 10-13; see also Fig. 10-12). CT scan may show only minimal changes, with loss of gray-white matter junction, sulcal effacement, and hypodensity in the areas of involvement.19,109,133 MRS frequently demonstrates decreased NAA, increased choline, and increased lactate.85,114 Definitive diagnosis is established by CSF polymerase chain reaction (PCR) or brain biopsy. Chronic sequelae of HSV-1 encephalitis may include volume loss, encephalomalacia, and calcifications within involved regions of the brain.

Herpes Simplex Type 2 (HSV-2) Encephalitis.  HSV-2 causes 80%

FIG 10-12  Herpes simplex virus (HSV)-1 encephalitis. Axial T2-FLAIR image shows hyperintense lesions in both temporal lobes, more on the right side.

A

of neonatal HSV infections and is transmitted from mother to neonate, most frequently during delivery as the baby passes through the vaginal canal. More rarely, infection may occur in utero or in the postpartum period. Infection may result in skin, eye, or mouth involvement, encephalitis, or disseminated disease. Typical imaging presentation includes T2-FLAIR hyperintensity involving basal ganglia as well as the cortex and subcortical white matter in the temporal and frontal lobes. T2 hyperintensity may be demonstrated in white matter diffusely. Contrast enhancement is  often seen in these areas. In early encephalitis only diffusion restriction may be evident. Hemorrhages may be seen later in disease. Chronic sequelae include atrophy with cystic encephalomalacia and ventricular enlargement. When acquired in utero, the patient may present with cataracts, microcephaly, cerebral calcifications, and ventricular enlargement as  a part of TORCH syndrome.19 Diagnosis is made by CSF PCR or

B FIG 10-13  Herpes simplex virus (HSV)-1 encephalitis. Axial postcontrast T1-weighted image demonstrates linear streaky enhancement along the sylvian fissure, insular cortex, and bilateral temporal frontal gyri (A, B).

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation detection of virus in mucocutaneous lesions. Acyclovir therapy is recommended in suspected cases.133

Human Herpesvirus 6 (HHV-6) Encephalitis HHV-6 infection also occurs early in life, with virus remaining in latent form in immunocompetent adults. HHV-6 reactivation and subsequent encephalitis, pneumonitis, and bone marrow suppression may occur in transplant recipients. Imaging findings of HHV-6 encephalitis are similar to HSV-1 encephalitis, except that in HHV-6 encephalitis, contrast enhancement is not demonstrated.133

Varicella Zoster Virus (VZV) VZV is a herpesvirus usually causing infection in childhood and presenting as chickenpox (varicella) and subsequently remaining in latent phase in ganglionic neurons, especially in dorsal root ganglia. In children, VZV encephalitis may occur as a complication of chickenpox or later reactivation. VZV encephalitis resulting from vaccine has been reported.96 In adults, VZV reactivation presents as shingles (zoster) as the virus travels along a sensory nerve and results in a painful rash. Ramsay Hunt syndrome occurs with reactivation of VZV in the geniculate ganglion of cranial nerve VII, resulting in facial weakness and  a rash known as zoster oticus.77 MRI may demonstrate facial nerve enhancement. VZV reactivation may occur in older immunocompetent patients or in immunosuppressed patients and may also result in vasculopathy, retinal necrosis, cerebellitis, leukoencephalopathy, and myelopathy. CNS involvement may occur in patients without skin lesions. Acyclovir is used for treatment of VZV. Vasculopathy is thought to be the dominant component of cerebral involvement by VZV, although meningoencephalitis may also be seen. Large- and small-vessel arterial involvement may be present. Clinical manifestations include infarcts and transient ischemic attacks, headaches, cranial neuropathies, and neurologic deficits.109 Vasculitis may be demonstrated on MR angiography (MRA), CT angiography (CTA), or digital subtraction angiography (DSA). Characteristic findings include multiple vessels with segmental narrowing (Fig. 10-14). Brain imaging features of VZV involvement are variable and may show T2-FLAIR hyperintensities at the gray-white matter

279

junction and less typically localizing to white matter or gray matter. Enhancing lesions and infarcts may also be seen. Rarer sequelae include aneurysm formation, dissection, and hemorrhage (parenchymal or subarachnoid).46 In children, VZV encephalitis may occur 10 days after appearance of the rash. MRI shows T2 hyperintense lesions at the gray-white matter junction, cortex, basal ganglia, and cerebellum. Vasculitis  and associated infarction may occur 1 to 4 months later.99 VZV may also cause acute cerebellitis and present with ataxia after infection  or vaccination. Numerous other viruses such as Epstein-Barr virus (EBV), HSV-1, HHV-6, enteroviruses, measles, and mumps may cause  cerebritis with similar clinical and radiologic presentations. Usually symptoms resolve over weeks to months. MRI demonstrates diffuse bilateral cerebellar hemisphere T2-FLAIR hyperintensity and diffusion restriction.30,133

Arbovirus Encephalitis Arboviruses are viruses that are transmitted by arthropod vectors such as ticks and mosquitoes to humans. Arboviral encephalitides include West Nile virus (WNV) encephalitis, eastern and western equine encephalitis, St. Louis encephalitis, and Japanese encephalitis (JE), with varying epidemiology, morbidity, and mortality in these entities. Arboviral encephalitides have no specific treatment, and imaging may help differentiate these diseases from treatable encephalitides caused by HSV and VZV. WNV is transmitted by a mosquito vector, with 1 out of 150 infections resulting in meningitis or encephalitis. MRI findings are nonspecific and include T2 hyperintensities in the basal ganglia, thalami, and brainstem, with frequent associated diffusion restriction. Mesial temporal lobe and long segment of spinal cord involvement may also occur.102 JE virus is a leading cause of viral encephalitis in Asia but is less common in the United States. Imaging findings are similar to WNV encephalitis, with bilateral T2 hyperintensities in thalami and brainstem. Bilateral thalamic hemorrhages are characteristic for this disease. Although the mesial temporal lobe may be involved, JE spares the anterior temporal lobe and insula.55,109,133

Acute Disseminated Encephalomyelitis (ADEM) ADEM is an autoimmune demyelinating disease that usually occurs 2 to 31 days following a viral infection or vaccination. This is usually a monophasic disease, but a less common multiphasic form may be seen with relapses occurring shortly after presentation, known as multipha­ sic disseminated encephalomyelitis (MDEM). Autoimmune response to myelin is thought to be responsible for ADEM. MRI is superior to CT for detection of ADEM lesions. Typically multiple bilateral, asymmetric, poorly defined, T2-FLAIR hyperintense, nonenhancing lesions are seen in the subcortical white matter. The corpus callosum and periventricular white matter are usually spared in contradistinction to multiple sclerosis (MS). Gray matter involvement is frequently seen in the thalami, basal ganglia, and cortex. Brainstem and spinal cord lesions may also be present. Occasionally contrast enhancement may be seen associated with the lesions. All lesions usually have similar imaging features due to the typical monophasic course of ADEM and characteristically show improvement after treatment with steroids (Fig. 10-15). Imaging differential diagnosis includes MS and vasculitis, as well as infectious and metabolic disorders77,81,111

P57

FIG 10-14  Herpes zoster vasculitis. MR angiogram demonstrates narrowing of the left middle cerebral artery branches as well as the proximal left posterior cerebral artery. There is a small infarct in the left middle cerebral artery territory.

Other Encephalitides

Creutzfeldt-Jakob Disease (CJD).  CJD is a transmissible spongiform encephalopathy presenting as progressive dementia. CJD is

http://radiologyebook.com

280

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 10-15  Acute disseminated encephalomyelitis (ADEM). A and B, T2-FLAIR images show numerous asymmetric rounded hyperintense, predominantly subcortical, white matter lesions. Some lesions involve the cortex. A right pulvinar lesion is also seen. C and D, Postcontrast T1-weighted image demonstrates incomplete ring enhancement associated with these lesions. All the lesions show similar imaging features. Marked improvement was seen after steroid therapy.

caused by prions, which are abnormal protein particles that induce abnormal folding of normal proteins in the brain, resulting in neuronal loss and spongiform changes typical of CJD. Three forms of the disease are recognized: sporadic, familial, and acquired. Acquired forms include kuru, iatrogenic CJD, and new variant CJD. The new variant of CJD is due to bovine spongiform encephalopathy (so-called mad cow disease). Iatrogenic CJD may result from use of contaminated human growth hormone, corneal transplantation, and cadaveric dural matter transplantation. Sporadic CJD with no known source of infection occurs in adults in their late 50s and accounts for most cases (≈90%). Rapidly progressive dementia leading to death within several

weeks to months and myoclonus are usually seen. Synchronous discharges on EEG are seen in the sporadic form.109 In sporadic CJD, MRI typically demonstrates diffusion restriction of the basal ganglia, thalami, cortex of the insula, cingulate gyri, superior frontal gyri, and other areas near the midline129 (Fig. 10-16). Associated T2 hyperintensity may be also seen in these regions in some cases. MRS is nonspecific, demonstrating decreased NAA. The new variant or bovine spongiform encephalopathy may present with bilateral pulvinar T2 hyperintensities known as pulvinar sign or with a combination of dorsomedial thalamic nucleus and pulvinar hyperintensity known as hockey stick sign.29,130,133

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

281

[AR]

A

B

[PF]

FIG 10-16  Creutzfeldt-Jakob disease. A, T2-FLAIR image demonstrates hyperintensity in both thalami, in the periatrial white matter, and in the anterior cingulate gyri. B, Diffusion-weighted image shows restricted diffusion in both thalami and cingulate gyri.

Subacute Sclerosing Panencephalitis (SSPE).  SSPE affects older children and young adults 6 to 8 years after a primary measles infection. This subacute encephalitis progresses through four clinical stages and usually results in death within 2 to 4 years. Early stages present with personality and behavioral changes, incoordination, tremor, and myoclonus and may show no imaging abnormalities. Late-stage disease results in coma, and MRI may demonstrate hyperintensities on FLAIR and T2-weighted images in periventricular white matter, splenium of the corpus callosum, thalami, basal ganglia, brainstem, and cerebellum.88 Cerebral and cerebellar atrophy may be seen on CT and MRI.77

Reye’s Syndrome.  Reye’s syndrome is a noninflammatory encephalopathy and liver dysfunction that typically presents after an upper respiratory tract infection and is associated with use of salicylates. This entity is now rarely seen, in part because of a decline in pediatric aspirin use. MRI may reveal cerebral edema and T2 hyperintensities with diffusion restriction in the midbrain, thalamus, cortical, and subcortical regions. Pontine T2 hyperintensity may also be demonstrated. CT may show diffuse cerebral edema.118

Encephalitis in Immunocompromised Patients

HIV Encephalitis.  HIV causes a subacute encephalitis through direct involvement of the brain and may result in a spectrum of clinical presentations, including neurocognitive abnormalities and AIDS dementia complex (ADC). ADC, a progressive dementia with behavioral and motor abnormalities, has become less common with the advent of highly active antiretroviral therapy (HAART). However, mild neurocognitive disorders remain common. Typical MRI manifestations include cerebral atrophy and symmetric patchy or confluent

periventricular white matter lesions, sparing the subcortical U-fibers, which are T2-FLAIR hyperintense and T1 isointense. CT is less sensitive, and areas of abnormal white matter may be seen as hypodensities on CT in advanced disease. Severity of atrophy has been shown to correlate with severity of neuropsychiatric abnormalities.48 MRS shows decreased NAA, elevated choline, and elevated myoinositol. Investigations of this entity using DTI have reported increased MD and decreased FA in normal-appearing white matter of HIVinfected individuals with detectable viral loads.40,103 However, a more recent study found DTI not to be useful for detection of early HIV infection in the brain.109,127 HIV may also cause vacuolar myelopathy of the cord, presenting with bilateral T2 hyperintense dorsal columns extending over multiple segments. These findings are similar to subacute combined degeneration in B12 deficiency.25,133

Progressive Multifocal Leukoencephalopathy (PML).  PML is caused by reactivation of a latent JC virus, a human polyomavirus that infects oligodendroglial cells of the brain. JC virus infects a majority of the population early in life, persisting in latent form and occasionally reactivating with immunosuppression. PML occurs in AIDS and less frequently in hematologic malignancies or with immunosuppressive medications, including therapy for transplant patients and MS patients on natalizumab.28 Patients present with altered mental status, progressive neurologic deficits, and lethargy. Without treatment, death may result within 6 to 12 months.133 Imaging findings include multiple bilateral, asymmetric, and often confluent subcortical and periventricular white matter lesions that  are T2-FLAIR hyperintense, T1 hypointense, and hypodense on CT.

http://radiologyebook.com

282

PART II  CT and MR Imaging of the Whole Body

FIG 10-17  Progressive multifocal leukoencephalopathy (PML). T2weighted image shows high signal intensity in the parietooccipital white matter bilaterally extending to the subcortical U-fibers.

peripheral neuropathy. CMV encephalitis may be classified in two forms that may be concurrent: microglial nodular encephalitis and ventriculoencephalitis. Microglial nodular encephalitis is similar to HIV encephalitis, causing progressive dementia. Unlike HIV encephalitis, progression is more rapid and both cranial nerve palsies and disorientation may occur. Imaging may be unremarkable or may show periventricular white matter T2-FLAIR hyperintensities and atrophy similar to HIV encephalopathy. Ventriculoencephalitis is a subacute necrotizing infection of the ependyma and subependymal tissue presenting as ataxia, cranial nerve palsies, progressive confusion, and lethargy, resulting in death within weeks. MRI demonstrates ependymal and subependymal T2-FLAIR hyperintensity and enhancement. However, imaging findings may be subtle or absent in some cases.109,121 Rarely meningeal enhancement or a ring enhancing mass lesion may be seen. CT is less sensitive, showing hypodensity in the periventricular white matter  and atrophy. Quantitative CSF viral PCR is used to establish the diagnosis.48 CMV is the most common cause of congenital infection in the United States and may be acquired by transplacental transmission, during birth, or through breast milk ingestion. CNS involvement may result in microcephaly, chorioretinitis, hearing impairment, cerebral palsy, epilepsy, and mental retardation. Most severe damage occurs in first and second trimester infections, resulting in lissencephaly, polymicrogyria, small cerebellum, and ventriculomegaly. Later infections result in white matter injury, with T2-FLAIR hyperintensity, periventricular calcifications, ventriculomegaly, and periventricular cysts. Subependymal distribution of calcifications differs from the randomly scattered distribution of calcifications seen in congenital toxoplasmosis133 (Fig. 10-18).

Lesions may show central diffusion facilitation with peripheral diffusion restriction and usually show no significant enhancement (Fig. 10-17). Occasionally faint peripheral enhancement is seen, which is associated with favorable outcomes.11 The lesions may involve any part of the brain, including posterior fossa structures, but they more frequently involve the supratentorial brain, basal ganglia, and thalami.105 PML lesions usually involve subcortical U-fibers, which are spared in HIV encephalopathy. Additionally, PML lesions are frequently T1 hypointense, whereas HIV lesions are isointense to white matter. MTR is markedly reduced in PML lesions compared to HIV lesions; presumably this is due to myelin macromolecule destruction in PML.37 MRS shows decreased NAA, increased choline, and possibly the presence of lactate and lipid.21 Diagnosis may be confirmed with CSF PCR. Immune reconstitution inflammatory syndrome (IRIS) represents worsening of neurologic deficits following initiation of therapy and is due to recovery of the immune system and enhanced inflammatory response to infection. PML-IRIS may occur in 23% of HIV-associated PML27 and may also be seen in MS patients on monoclonal antibody therapy. Steroid therapy may be useful in this setting.125 MRI reveals progression of findings a few weeks following initiation of therapy, with contrast enhancement in lesions, increased lesion size, increased edema, and mass effect. These findings later improve and resolve.48,109,133

SEPTIC EMBOLUS AND MYCOTIC ANEURYSM

Cytomegalovirus (CMV) Encephalitis.  CMV is a ubiquitous DNA

GRANULOMATOUS INFECTION

virus in the Herpesviridae family and is present in latent form in a majority of the general population. In immunocompetent patients  it may cause a self-limited febrile illness. In immunocompromised patients CMV may cause disseminated disease, including respiratory, gastrointestinal (GI), and CNS infections. CMV may cause ventriculitis, meningoencephalitis, retinitis, radiculomyelopathy, and

Infective endocarditis is most frequently caused by Staphylococcus aureus, and its major complication is septic emboli and stroke. Other causes of septic emboli to the brain include pulmonary infection, pulmonary arteriovenous malformation with sepsis, intravenous drug abuse, and cyanotic heart disease. The majority of septic emboli cases are seen in patients with a history of drug abuse. Intracranial septic emboli may lead to formation of mycotic aneurysms, abscess formation, or obliterative vasculitis. Mycotic aneurysms probably result from infarction of the arterial wall secondary to infected emboli. Mycotic aneurysms may be solitary or multiple, and they usually involve middle cerebral artery branches. The relative frequency of middle cerebral artery branch involvement is probably explained by the likelihood of emboli being carried into this territory.141 Rupture of a mycotic aneurysm may produce an intracerebral or subdural hematoma117 (Fig. 10-19). An unruptured mycotic aneurysm may occasionally be detected by contrast-enhanced CT or MRI studies.4 Associated infarction or small areas of cerebritis or abscess formation may be seen on imaging. Fungal infectious aneurysms frequently result from craniofacial infections. Typically they present as large proximal vessel fusiform aneurysms and are associated with a worse prognosis12 (see Fig. 10-29).

Tuberculosis (TB) Mycobacterium tuberculosis is the most common cause of CNS tuberculous infection. Hematogenous spread to the CNS may occur from pulmonary infection or other distant sites of infection. Initially an inflammatory focus known as the Rich focus is established in the

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

283

B

A

FIG 10-18  Cytomegalic inclusion disease. A and B, Nonenhanced CT scans in a newborn show periventricular calcification and ventricular dilatation.

meninges, subpial, or subependymal tissue of the brain or spinal; this subsequently ruptures into the ventricular system or subarachnoid space and causes causes ventriculitis, meningitis, or both. Occasionally direct extension may occur from the skull, paranasal sinuses, mastoids, or orbits. CNS tuberculosis remains a worldwide problem owing to the high incidence of this infection in immunocompromised patients  and emergence of drug-resistant strains of mycobacteria. CNS TB includes leptomeningeal and pachymeningeal meningitis as well as parenchymal disease including tuberculomas, tuberculous abscesses, and cerebritis.48 TB meningitis is a subacute inflammation of the leptomeninges that typically involves the basal cisterns and sylvian fissures, producing a thick gelatinous exudate. This exudate may obstruct CSF flow and result in communicating hydrocephalus. Less frequently, noncommunicating or complex hydrocephalus may occur secondary to ventricular outlet obstruction, ventriculitis, or mass effect from a parenchymal TB lesion. TB meningitis may cause a vasculitis of small to medium-sized vessels of the circle of Willis, resulting in focal infarcts in the basal ganglia or distal MCA territory. Cranial nerve involvement due to ischemia or entrapment of cranial nerves in thick exudate or chronic fibrosis frequently occurs. Occasionally, parenchymal lesions may result in cranial neuropathy due to involvement of a cranial nerve nucleus. Patients with TB meningitis may present with fever, weight loss, behavioral changes, decreased mental status, and meningeal signs. Advanced disease may present with stupor, infarcts, and cranial neuropathy. CSF analysis maybe normal in 10% to 20% of cases but usually shows increased protein, decreased glucose, and lymphocytic pleocytosis. The most characteristic imaging finding on postcontrast CT and MRI is nodular leptomeningeal enhancement in the basal cisterns and sylvian fissures, which may extend to the cerebral convexities in advanced cases. Presence of associated hydrocephalus and focal infarcts in the basal ganglia or MCA territory is specific for TB meningitis 

(Figs. 10-20 and 10-21). Cranial nerve involvement may be seen as thickening and enhancement of the cisternal segment of the nerve. However, meningeal enhancement may be absent or minimal in early disease and in patients with AIDS. Occasionally, CT may show basal meningeal hyperdensity. Noncontrast magnetization transfer (MT) T1-weighted imaging may demonstrate basal meningeal hyperintensity. MRA or CTA may show arterial narrowing, beading, or occlusion of middle cerebral and lenticulostriate arteries. The chronic sequelae of TB meningitis include meningeal calcifications, volume loss, hydrocephalus, syringobulbia, and syringomyelia.101 TB pachymeningitis may occur as the only finding or in combination with leptomeningeal and parenchymal involvement. MRI may show diffuse or localized dural thickening, which is isointense on T1-weighted images, hypo- to isointense on T2-weighted images, and demonstrates moderate to avid enhancement on postcontrast studies. On noncontrast CT, the dural thickening may appear hyperdense. Tuberculomas are the most common form of parenchymal TB and present as solitary or multiple lesions in any region of the brain. Infratentorial lesions predominate in children, whereas supratentorial lesions are more common in adults. Frequently tuberculomas are seen at the corticomedullary junction owing to hematogenous dissemination to narrow vessels in this region. Occasionally tuberculomas may complicate meningitis, presumably the result of extension of infection via perivascular spaces or via cortical veins. Patients may present with headache, seizures, high intracranial pressure, or focal neurologic deficits. CSF analysis may be normal. Noncaseating tuberculomas may progress to caseating tuberculomas that may later undergo central liquefaction. Different imaging characteristics are seen at each stage. 1. Noncaseating tuberculomas are iso- to slightly hyperdense on noncontrast CT and demonstrate T1 and T2 iso- to hypointensity on MRI. Postcontrast CT and MRI demonstrates homogeneous enhancement. MT T1-weighted images demonstrate internal hyperintensity (Fig. 10-22).

http://radiologyebook.com

284

PART II  CT and MR Imaging of the Whole Body

A

B FIG 10-19  Mycotic aneurysm. A, Proton-density-weighted image shows high-signal-intensity hemorrhage in the left occipital lobe as well as in the ventricle. A small ringlike low-signal-intensity area is seen within the hemorrhage. B, A mycotic aneurysm was found on cerebral angiogram, corresponding to the small ringlike low-signal-intensity area seen on the MRI study.

A

B FIG 10-20  Tuberculous meningitis in a child. A and B, Postcontrast CT images demonstrate intense enhancement in the suprasellar cistern, sylvian cistern, and prepontine cistern. Dilatation of the ventricular system is seen, consistent with associated hydrocephalus.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

285

B

C FIG 10-21  Tuberculous meningitis and caseating tuberculoma. A, T1-weighted image demonstrates iso- to slightly hypointense signal and effacement of the suprasellar and right sylvian cisterns. B, T2-weighted image shows corresponding predominantly hypointense heterogeneous signal within the basal cisterns, with focal areas of hyperintensity. A hypointense rounded lesion is noted in the right uncus. Parenchymal hyperintensity representing edema and possible cerebritis is seen in the adjacent right temporal lobe and midbrain. C, Postcontrast T1-weighted image demonstrates corresponding intense nodular enhancement in the basal cisterns surrounding arterial structures. Right uncus lesion shows rim enhancement surrounding a central nonenhancing region, corresponding to a hypointense area on T2-weighted image, which is consistent with caseating tuberculoma.

2. Caseating tuberculomas are T1 and T2 iso- to hypointense with peripheral rim demonstrating T2 iso- to hyperintensity and MT T1 hyperintensity. The central region usually shows no diffusion restriction. Postcontrast imaging shows peripheral rim enhancement. MRS typically shows a prominent lipid peak (1.33 ppm)52 (Fig. 10-23; see also Fig. 10-21). 3. Caseating tuberculoma with liquid center demonstrates central hypodensity on CT and T2 hyperintensity on MRI. Central diffusion restriction is seen. Peripheral rim is T2 hypointense, enhances, and may mimic an abscess on MRI. Absence of amino

acids on MRS and thick rim enhancement help in differentiation from abscess.128 A central nidus of calcification with surrounding rim of enhancement on CT has been described as “target sign” and is pathognomonic for tuberculoma.10 Occasionally tuberculoma may demonstrate heterogeneous enhancement and may demonstrate a choline peak as well as lipid, presumably due to greater cellularity.52 Miliary tuberculomas are caused by hematogenous dissemination, presenting with numerous small enhancing foci that may be difficult to visualize on conventional sequences. These small lesions may be

http://radiologyebook.com

286

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 10-22  Noncaseating tuberculoma. A, T1-weighted image demonstrates a focal rounded lesion in the right midbrain, with an isointense central region and iso- to hyperintense rim. B, T2-weighted image shows corresponding heterogeneous iso- to hypointense signal in this lesion, with surrounding hyperintense edema. C, Coronal postcontrast T1-weighted image demonstrates avid homogeneous enhancement of the lesion. D, Diffusion-weighted image demonstrates no diffusion restriction.

better seen on MT T1-weighted images.51 Tuberculous cerebritis is a rare presentation of parenchymal TB in patients without AIDS. MRI shows regional gyral enhancement and adjacent edema. Tuberculous abscess represents a collection of pus with a large number of tuberculous bacilli surrounded by a capsule. This is a rare form of parenchymal

TB and may arise from a tuberculoma or as a complication of TB meningitis. MRI usually demonstrates a large multiloculated ringenhancing lesion with surrounding edema and central diffusion restriction.128 The enhancing rim is thicker than a typical pyogenic abscess. MRS usually demonstrates lipid, lactate, phosphoserine peaks,

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

287

B

A

FIG 10-23  Caseating tuberculoma. A, Postcontrast T1-weighted image demonstrating a ring enhancing lesion in the left cerebellar hemisphere. B, Diffusion-weighted image demonstrates no diffusion restriction.

and absence of amino acid peaks5 (Fig. 10-24). Low MT ratio is demonstrated in the rim and may also assist in differentiation from a pyogenic abscess.53 MRI and CT are frequently used to follow response to therapy. Occasionally paradoxical increase in volume may occur despite adequate therapy. A recent investigation using dynamic contrast-enhanced (DCE) imaging demonstrated that decrease in permeability (Ktrans) and edema volume are associated with therapeutic response even in cases of paradoxical volume increase.56

SPIROCHETE AND NOCARDIA INFECTIONS Lyme Disease Lyme borreliosis is an infectious disease transmitted by Ixodes deer ticks harboring the spirochete Borrelia burgdorferi in the United States and B. garinii and B. afzelii in Europe. Similar to syphilis, Lyme disease presents in three stages: (1) early localized disease, (2) early disseminated disease, and (3) persisting late disease. The early stage may present with a characteristic enlarging erythematous skin rash at the site of the tick bite termed erythema chronicum migrans, flulike symptoms, and lymphadenopathy. The early disseminated stage may present with arthritis, meningitis, cranial neuropathy (frequently Bell’s palsy), painful radiculopathy, and cardiac disease. The late disease may occur in untreated patients 2 to 3 years after infection, presenting with chronic arthritis, polyneuropathy, and encephalitis. CNS disease may occur without history of preceding rash or other symptoms.6 MRI is often unremarkable even in patients with CNS symptoms. MRI findings include multiple subcortical and periventricular T2-  FLAIR hyperintense white matter lesions that sometimes involve the callososeptal interface and may mimic MS. Investigation using MT MRI and DTI suggests that in contradistinction to MS, Lyme disease is unlikely to result in occult cerebral tissue damage or cervical cord involvement.3 Occasionally multiple enhancing lesions may be seen. MRI may also demonstrate leptomeningeal and cranial nerve enhancement.2 Prominent perivascular spaces may be seen in pediatric Lyme disease. Vasculitis may also occur, resulting in infarcts or intracranial hemorrhage. Rarely orbital involvement may occur, which may mimic

orbital pseudotumor. Transverse myelitis and nerve root enhancement have also been reported.6

Syphilis Syphilis is a chronic disease caused by Treponema pallidum and is usually transmitted by sexual contact. The primary stage with local ulcer (chancre) and regional lymphadenopathy is followed by the secondary stage with hematogenous dissemination and rash. After a prolonged latent period, some patients develop tertiary syphilis, with approximately one third of these patients developing neurosyphilis. AIDS patients may have more rapid progression of disease and may show CNS involvement within 3 to 18 months. The most common forms of CNS syphilis include meningovascular syphilis and syphilitic gummas.48 Meningovascular syphilis causes meningeal thickening and peri­ vascular lymphocytic infiltrates and may present with headaches, meningeal signs, and cranial nerve palsies. MRI may show meningeal enhancement, cranial nerve enhancement (most frequently cranial nerves VII and VIII), cochlear enhancement, and hydrocephalus. Vasculitis may result in arterial occlusion involving the middle cerebral artery or basilar artery branches. Medium to large infarcts may be seen.6,119 A gumma is a mass of granulation tissue caused by a focal  leptomeningeal inflammatory reaction during syphilitic meningitis. Gummas form in the leptomeninges and associated blood vessels and may extend into the adjacent brain parenchyma. MRI demonstrates solitary or multiple masses in the brain parenchyma, which are adherent to the dura and may demonstrate a dural tail. These lesions are iso- to hypodense on noncontrast CT and demonstrate T1 isointensity and T2 hyperintensity on MRI. Gummas demonstrate nodular or ring enhancement and may mimic intracranial neoplasms. 18F fluorodeoxyglucose positron emission tomography (FDG PET) may show high uptake similar to a hypermetabolic tumor,76 making correct diagnosis more challenging. Parenchymal gyriform or heterogeneous enhancement and nonspecific T2 white matter hyperintensities may also be seen. Diffuse parenchymal disease and ependymitis may result in cerebral atrophy

http://radiologyebook.com

288

PART II  CT and MR Imaging of the Whole Body

A

B

TR: 1265.0 TE: 144.0 NEX: 4.0

C

4

3

2

1

0

D

FIG 10-24  Tubercular abscesses. A, T2-weighted image shows large left frontal and right parietal lesions, demonstrating a hypointense rim and a heterogeneously hyperintense central cavity. Extensive surrounding area of edema is noted. Focal hypointense lesion in the left parietal lobe represents a noncaseating tuberculoma. B, Postcontrast T1-weighted image demonstrates corresponding thick rim enhancement of the two cavitary lesions. Homogeneous enhancement is seen in the left parietal tuberculoma. C, Singel-voxel MR spectroscopy of the central cavity demonstrates a large lipid peak. D, Diffusion-weighted image demonstrates internal high signal intensity representing central diffusion restriction.

and hydrocephalus.48 Spinal cord involvement may occur as tabes dorsalis with cord myelopathy involving the dorsal columns. Sensory nerve damage may also be present.6

Nocardiosis Nocardiosis is usually caused by an aerobic gram-positive, variably acid-fast bacteria, Nocardia asteroides. The bacteria form filamentous branching hyphae and may be found in soil and decaying vegetables. Nocardial infections usually occur in immunocompromised individu-

als, including transplant recipients and patients with AIDS. Inhalation of the organism causes pulmonary infection, which may subsequently disseminate hematogenously to the CNS in 15% to 44% of patients.80,137 Solitary or multiple brain abscesses represent the most common manifestation of nocardiosis in the CNS.116 Hydrocephalus and clinical evidence of meningitis may be present. Preceding pulmonary infection is frequently identified on chest imaging.72 Noncontrast CT may show a hypodense lesion. Postcontrast MRI and CT typically demonstrate solitary or multiple ring enhancing

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

B

C

D

289

FIG 10-25  Nocardial abscesses in a renal transplant patient. A and B, Postcontrast T1-weighted image shows two large ring enhancing lesions. Note the bilobed appearance of the left subinsular abscess and the lobulated appearance of the right cerebellar abscess. C, T2-weighted image shows an abscess with a hypointense rim, central cavity, and adjacent surrounding edema corresponding to ring enhancing lesion on T1-weighted image. D, Diffusion-weighted image demonstrates diffusion restriction. Diffusion restriction is seen in the dependent portion of the ventricles, likely representing intraventricular debris due to ependymitis.

lesions. Multiple conglomerate or budding ring enhancing lesions may be seen.34,116 Other imaging characteristics are similar to pyogenic abscesses. The abscess wall usually demonstrates hypointensity on T2 weighted images.116 The central cavity is usually T2 hyperintense and shows diffusion restriction (Fig. 10-25). MRS is reported to demonstrate lactate and amino acids.122 Subependymal nodules may also be seen.72

FUNGAL DISEASE Fungal organisms are ubiquitous but rarely result in clinically significant disease in immunocompetent individuals. The recent increase in invasive fungal infections is attributed to the rising prevalence of immunosuppression due to AIDS, higher numbers of transplantations, and increased resistance to antifungal medications. Fungal spores may

http://radiologyebook.com

290

PART II  CT and MR Imaging of the Whole Body

be inhaled, infecting the lungs and subsequently disseminating hematogenously to the CNS. Direct extension may occur from fungal disease or postsurgical infection. Coccidioides, Cryptococcus, Histoplasma, and Blastomyces may cause disease in immunocompetent individuals, whereas Candida and Aspergillus infections occur in the immunocompromised state. CNS fungal disease may be classified as meningitis, parenchymal lesions, and vasculitis. Morphologically, fungi may be classified as small unicellular yeasts such as Candida and Cryptococcus, large branching colonies of hyphae such as Aspergillus and Mucor, and dimorphic organisms such as Coccidioides that can exist in both forms. The morphology of a fungal organism predicts the pattern of CNS disease, with small yeast organisms entering small meningeal vessels and frequently causing meningitis, whereas large hyphal organisms typically localize to larger cerebral arterioles, resulting in parenchymal lesions. A large mass of branching hyphae may form, which is known as a mycelium. The imaging presentation of fungal disease is similar to CNS tuberculosis. Fungal meningitis is usually due to Cryptococcus or Candida and may demonstrate nodular enhancement of the leptomeninges in the basal cisterns and subarachnoid hyperintensity on FLAIR in these areas. Lower MTR values may help distinguish fungal meningitis from TB. Parenchymal lesions include cerebritis, granuloma, and fungal abscess. Fungal granulomas are most frequently caused by Aspergillus and members of the order Mucorales, with nonspecific MR and CT appearance including faint enhancement and surrounding edema. Fungal cerebritis usually demonstrates T2-FLAIR hyperintensity, variable enhancement, and heterogeneous diffusion restriction. Fungal abscess typically demonstrates ring enhancement and a T2 hypointense rim that may show diffusion restriction. Fungal abscesses are frequently multiple and typically localize to the corticomedullary junction and basal ganglia. Pyogenic abscess are more often solitary and spare the basal ganglia. Crenated borders and nonenhancing diffusionrestricting intracavitary projections are also suggestive for a fungal abscess.78 MRS may show amino acids, lipid, and lactate, similar to a pyogenic abscess. However, the presence of trehalose, seen as multiple peaks at 3.6 to 3.8 ppm, is specific for a fungal etiology. Vascular involvement is common in aspergillosis and mucormycosis, which produce elastase. Vasculitis usually involves medium to large vessels in basal cisterns and may result in infarction. Hemorrhage occurs more frequently in Aspergillus infection. Rarely mycotic aneurysms may form, causing focal aneurysm or fusiform dilatation of the proximal cerebral vessels.83

Aspergillosis Aspergillus is a common environmental mold and usually causes infection in immunocompromised individuals. Inhalation of Aspergillus fumigatus spores may result in chronic pulmonary infection in an immunocompromised patient and may subsequently cause hematogenous dissemination. Aspergillus may also invade the CNS by direct extension secondary to trauma or sinus infection. This hyphal organism usually causes parenchymal lesions including aspergillomas (granulomas), cerebritis, and abscess. Aspergillomas are typically T2 iso- to hypointense with surrounding T2 hyperintense regions of edema (Fig. 10-26). Aspergillus is the most common pathogen in fungal brain abscesses. Aspergillus abscess typically demonstrates a T2 and T2* hypointense rim, presumably due to small amounts of hemorrhage. Contrast enhancement may or may not be seen following intravenous injection of contrast material, owing to the variable ability to mount an inflammatory response in an immunosuppressed state. This organism produces elastase and has a propensity to invade vessels, resulting in thrombosis, vasculitis, and infarction with frequent associated

A

B

L

FIG 10-26  Aspergillosis. A, T2-weighted image demonstrates curvilinear hypointense areas with surrounding hyperintense edema in the left basal ganglia. B, Postcontrast T1-weighted image shows nodular and incomplete ring enhancement in the left basal ganglia, as well as small nodular enhancement in the left occipital and right temporal parietal region.

hemorrhage.5 Meningitis and meningoencephalitis are seen less frequently. There is a lack of correlation between the clinical presentations and imaging findings in patients with aspergillosis. This is probably due to the fulminant aggressive nature of the disease process and slow development of the imaging findings in immunocompromised patients who may be on steroid therapy.141

Mucormycosis Mucormycosis, also known as zygomycosis and phycomycosis, represents infection with fungi in the Mucorales order. These ubiquitous hyphal organisms cause infection in immunocompromised patients, especially those with diabetes, chronic renal failure, and cirrhosis. Spores in nasal mucosa may germinate and form angioinvasive hyphae, causing infarction and gangrenous changes in the involved mucosa. Infection may rapidly spread to the sinuses and orbits, with frequent subsequent intracranial invasion via the cribriform plate, ophthalmic arteries, or cavernous sinus (described as rhinocerebral mucormycosis). A high level of suspicion for mucormycosis is needed in diabetic patients with sinusitis, eye pain, or facial pain. Clinical symptoms may

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

291

tis, skin, and CNS involvement. Meningitis is the most common form of CNS disease and has been reported in up to 50% of cases with disseminated infection. Meningitis is typically seen on MRI as enhancement of the basal cisterns, sylvian fissures, and pericallosal leptomeninges (Fig. 10-30). Associated hydrocephalus is frequently seen. Vascular complications with associated deep infarcts may also occur.35 Parenchymal lesions such as granulomas and abscesses are seen infrequently and demonstrate ring enhancement on postcontrast imaging. A lytic lesion involving the calvarium may be seen without parenchymal disease. A high prevalence of concurrent brain and intraspinal disease has been reported, including arachnoiditis and cord signal abnormalities, supporting a low threshold for imaging evaluation of the spine in patients with intracranial coccidioidomycosis.70

Cryptococcosis

FIG 10-27  Mucormycosis. Noncontrast CT demonstrates air-fluid level in the right maxillary sinus. Note increased density and inflammatory stranding of the right periantral fat and masticator space, consistent with invasive fungal sinusitis.

also include bloody nasal discharge, chemosis, exophthalmos, and cranial neuropathy. Hematogenous spread is less common, and isolated cerebral mucormycosis is seen more often in patients with a history of intravenous drug abuse, where it often presents with basal ganglia involvement.123,126 Imaging presentation is similar to aspergillosis, with a propensity for vascular invasion that may result in thrombosis and infarction and frequent involvement of the orbits and paranasal sinuses. Early diagnosis of rhinocerebral mucormycosis is critical because rapid progression of this disease results in high morbidity and mortality. Early CT findings include invasion of periantral fat and orbital extension from the ethmoid sinuses, with later development of aggressive bone destruction9 (Fig. 10-27). Characteristic MRI findings of early sinonasal infection include regional mucosal T2 hypointensity, diffusion restriction, and lack of mucosal enhancement, which has been described as the “black turbinate sign.” These findings reflect infarcted devitalized tissue and presence of hyphae60,113 (Fig. 10-28). Sinus disease may exhibit near signal void (similar to air) on T2-weighted sequences, indistinguishable from normal aerated sinuses, which has been suggested to be the result of a high concentration of iron and manganese within the fungal mass.126 Intracranial extension may present with leptomeningeal enhancement, cerebritis, cranial nerve involvement, and formation of granuloma or abscess. Vascular complications such as cavernous sinus thrombosis, arterial occlusion, aneurism or pseudoaneurysm formation, and infarction may be seen83 (Fig. 10-29; see also Fig. 10-28).

Coccidioidomycosis Coccidioidomycosis is caused by a dimorphic fungus Coccidioides immitis. It is endemic in the southwestern United States and parts of Central and South America. Spore inhalation may result in a selflimited respiratory infection. Hematogenous dissemination is reported in less than 1% of immunocompetent individuals and is more common in patients with HIV (up to 30%).42 Patients may develop osteomyeli-

Cryptococcus neoformans is a ubiquitous yeast organism with a polysaccharide capsule, usually causing infection in immunocompromised individuals. It is most frequently seen in AIDS patients; however,  up to 30% cases of cryptococcosis are seen in immunocompetent individuals. Cryptococcosis is the most common fungal disease involving the CNS; a majority of infected individuals develop CNS involvement. Cryptococcal IRIS may be seen after initiation of HAART, presenting with worsening of clinical symptoms and radiologic findings (Fig. 10-31). Meningitis with basilar predominance is the most common form of CNS disease; however, imaging typically demonstrates absent or minimal meningeal enhancement, especially in immunocompromised patients. When avid leptomeningeal enhancement is seen in AIDS and IRIS is not suspected, other causes of meningitis should be considered, including bacterial, TB, CMV, or leptomeningeal involvement by lymphoma.22 Increased intracranial pressure and hydrocephalus are common. Infection may extend from the subarachnoid space into the Virchow-Robin (VR) perivascular spaces, which may become enlarged owing to growth of yeast organisms and later coalesce into larger gelatinous pseudocysts. MRI usually demonstrates nonenhancing lesions that are isointense to CSF on T1- and T2-weighted images in basal ganglia, thalamus, and midbrain, representing dilated perivascular spaces or larger gelatinous pseudocysts. The lesions are usually bilateral and symmetric (Fig. 10-32). Although enlarged perivascular spaces may be a normal finding—especially in older individuals— concern for cryptococcosis should be raised in the appropriate clinical setting.48 Occasionally pseudocysts may show diffusion restriction. Meningitis may cause vasculitis with subsequent infarcts (Fig. 10-33). Meningitis may also result in parenchymal involvement with formation of granulomas (cryptococcomas). These lesions are usually seen in the basal ganglia and demonstrate T1 iso- to hypointensity, T2 hyperintensity, and rarely show enhancement (Fig. 10-34). Intraparenchymal cystic lesions may also be seen and may demonstrate surrounding edema.83 Rarely spinal cord involvement may occur.

Candidiasis Most cases of candidiasis are caused by Candida albicans, a yeast organism that is part of the normal flora in human mucocutaneous regions and the GI tract. Disseminated disease usually occurs in immunocompromised patients and is thought to result from an imbalance in the normal flora due to medications such as antibiotics, steroids, and chemotherapy. Renal infection is the most common manifestation of disseminated disease, with CNS involvement seen in 18% to 52% of cases. CNS involvement includes cerebral microabscesses, meningitis, cerebral macroabscesses, and vascular complications.

http://radiologyebook.com

Text continued on p. 296

292

PART II  CT and MR Imaging of the Whole Body

A

C

B

D FIG 10-28  Rhinocerebral mucormycosis. A and B, Postcontrast coronal and axial T1-weighted image demonstrating lack of enhancement in the right nasal cavity mucosa, maxillary sinus, and orbit, consistent with infarcted tissue due to invasive sinusitis. Axial image shows absence of cavernous internal carotid artery (ICA) flow void due to occlusion. Additionally the right cavernous sinus demonstrates no enhancement, representing thrombosis. C, Diffusion-weighted image demonstrates diffusion restriction in the right nasal and maxillary sinus mucosa corresponding to nonenhancing areas on postcontrast MRI. Right temporal lobe hyperintensities represent cerebritis. D, T2-weighted image demonstrates corresponding low signal in the right nasal mucosa, lack of right ICA flow void, and right temporal lobe hyperintensity.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

B FIG 10-29  Mycotic aneurysms secondary to mucormycosis. A, Time-of-flight MR angiogram maximum intensity projection (MIP) image demonstrates large irregular fusiform aneurysms of the bilateral cavernous segments of internal carotid arteries. B, Reformatted CT angiogram image in the sagittal plane shows an irregular fusiform aneurysm of the right internal carotid artery and adjacent erosions of the sphenoid sinus wall and sinus opacification.

A

B FIG 10-30  Coccidioidomycosis meningitis. A, T2-FLAIR image shows hyperintensity along the leptomeninges of the suprasellar, ambient, quadrigeminal plate, and bilateral sylvian cisterns. B, Postcontrast T1-weighted image demonstrates corresponding intense leptomeningeal enhancement in the basal cisterns and sylvian fissures. Smaller foci of leptomeningeal enhancement are seen in multiple sulci.

http://radiologyebook.com

293

294

PART II  CT and MR Imaging of the Whole Body

A

B FIG 10-31  Immune reconstitution inflammatory syndrome in cryptococcal meningitis. A, Axial postcontrast T1-weighted image shows unusually thick avid enhancement in the cerebral sulci bilaterally. B, T2-FLAIR image demonstrates bilateral frontal lobe hyperintensity and sulcal effacement consistent with edema. These findings became evident after initiation of HAART.

A

B FIG 10-32  Cryptococcal meningitis. A, T2-weighted image shows multiple hyperintense lesions in the basal ganglia. B, T2-FLAIR image demonstrates high signal in the basal ganglia lesions, consistent with dilated perivascular spaces and gelatinous pseudocysts. High signal is seen in cortical sulci. Postcontrast images demonstrated no enhancement. Diffusion-weighted image showed no diffusion restriction (not shown).

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

B FIG 10-33  Cryptococcal meningitis with vasculitis and infarcts. A, Coronal postcontrast T1-weighted image shows enhancement along the leptomeninges in the cisterns and cerebral sulci bilaterally. B, Diffusionweighted image demonstrates multiple bilateral cortical areas of diffusion restriction, consistent with multiple infarcts caused by vasculitis.

A

B FIG 10-34  Cryptococcomas. A, T2-weighted image shows a large hyperintense rounded lesion in the left lentiform nucleus. B, Postcontrast T1-weighted image demonstrates corresponding thick irregular peripheral enhancement. Two smaller homogeneously enhancing rounded lesions are seen in right frontoparietal and left temporal subcortical white matter.

http://radiologyebook.com

295

296

PART II  CT and MR Imaging of the Whole Body

L 9 5

A

B

C FIG 10-35  Candidiasis. A, T1-weighted image reveals a hyperintense area in the left thalamus, consistent with focal hemorrhage. B, FLAIR image demonstrates abnormal hyperintense areas in the corpus striati bilaterally and in the left thalamus. C, Postcontrast T1-weighted image demonstrates patchy abnormal enhancement in the corpus striati bilaterally as well as in the left thalamus.

Microabscesses are smaller than 3 mm and may be seen on MRI as multiple hypointense parenchymal T2 hyperintense or hemorrhagic T2 hypointense lesions that may show enhancement and diffusion restriction. CT may show multiple enhancing iso- to hypodense lesions. Macroabscesses are unusual and may be solitary or multiple. Vascular involvement may result in infarctions or hemorrhages, most frequently in the basal ganglia, and may rarely cause a mycotic aneurism (Fig. 10-35).

PARASITIC DISEASE Cysticercosis The life cycle of the pork tapeworm Taenia solium includes humans as the definitive hosts and pigs as the intermediate hosts. Ingestion of poorly cooked pork containing larvae (cysticerci) of the organism results in intestinal infection by mature tapeworms. This infection  is known as taeniasis and may be asymptomatic or result in mild GI symptoms. Pigs acting as intermediate hosts may ingest feed  contaminated with human feces containing eggs of the organism, resulting in hematogenous dissemination of larvae to the muscle, forming numerous cysticerci. Cysticercosis occurs when humans ingest food contaminated with human fecal material containing T. solium eggs, with subsequent hematogenous dissemination of the larvae to muscle, subcutaneous tissues, and CNS, with formation of numerous larval cysticercal cysts. This disease is endemic in parts of Oceania, Latin America, Eastern Europe, Asia, and Africa and is seen more frequently in the southern and southwestern United States because of immigration from endemic regions. CNS involvement,

known as neurocysticercosis, is common and may present with parenchymal, intraventricular, or meningeal involvement. Seizures represent the most common clinical presentation of neurocysticercosis. Other complications include hydrocephalus, intracranial hypertension, mass effect, and infarction.44 Parenchymal neurocysticercosis is typically seen as multiple lesions in the cerebral corticomedullary junction and is due to hematogenous deposition of these lesions in small vessels in these regions. Cerebellar and basal ganglia lesions are occasionally seen. Initially viable cysticerci are able to suppress the immune response. However, in months to years the immune response is triggered and the organism dies, resulting in edema and enhancement on imaging as the cysts undergo degeneration. Occasionally cyst rupture may result in a more severe inflammatory response. These lesions are classified in four stages as they evolve and degenerate.44,74 1. The vesicular stage represents a viable cysticercal cyst containing an invaginated scolex. CT and MRI demonstrate a cyst with attenuation and signal similar to CSF. MRI demonstrates a thin linear rim of enhancement or lack of enhancement on postcontrast images. A scolex may be seen in some lesions, which is frequently best visualized on FLAIR images as a hyperintense focus occasionally demonstrating enhancement or diffusion restriction95 (Fig. 10-36). 2. The colloidal vesicular stage represents an early degenerating cyst with destruction of the scolex, proteinaceous internal fluid, and associated significant inflammatory response. CT typically demonstrates a hyperdense cyst and ring enhancement. MRI usually demonstrates a cyst with T1 hyperintense fluid, no visible

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

297

B FIG 10-36  Parenchymal cysticercosis: vesicular stage. A and B, T2-weighted image shows two cortical and subcortical rounded high-signal-intensity lesions. Internal eccentric hypointense focus represents the scolex. Note lack of adjacent edema. Postcontrast images showed no significant enhancement associated with these lesions (not shown). Ill-defined linear hyperintensities in the suprasellar cistern represent coexisting subarachnoid or “racemose” form of cysticercosis with associated ventricular enlargement due to hydrocephalus.

A

B FIG 10-37  Parenchymal cysticercosis. A, T2-weighted image shows no definite abnormality. B, Postcontrast T1-weighted image shows two ringlike enhancing lesions in the right insular cortex.

scolex, and a peripheral enhancing rim that may be T2 hypointense. Surrounding edema is usually seen. Unlike a pyogenic abscess, no diffusion restriction is seen95 (Fig. 10-37). 3. The granular nodular stage represents a more advanced stage of cyst degeneration forming a smaller nodule. A nodular or small ring enhancing lesion with mild surrounding edema is usually demonstrated on MRI and CT.95 4. The nodular calcified stage represents the inactive end stage of cyst degeneration, resulting in a small calcified granuloma. Typically

no enhancement or edema is seen. CT is the most sensitive modality for these lesions, demonstrating hyperdense foci. MRI is less sensitive, showing small T1 and T2 hypointense foci.95,141 Occasionally reactivation occurs at this stage, resulting in a calcified lesion with enhancement and surrounding edema that may present with seizure. Presumably this is due to an inflammatory response to residual antigens within the lesions91,90 (Fig. 10-38). Multiple cysts may coexist at different stages within the brain.74

http://radiologyebook.com

298

PART II  CT and MR Imaging of the Whole Body

FIG 10-38  Parenchymal cysticercosis: nodular calcified stage. Non­

FIG 10-39  Intraventricular cysticercosis. Coronal postcontrast T1-

contrast CT scan reveals multiple intracranial calcifications at the corticomedullary junction, consistent with small calcified granulomas representing the nonviable inactive stage of neurocysticercosis. A coexisting vesicular stage lesion in subcortical white matter of posterior left cingulate gyrus represents a viable larva.

weighted image shows a rim enhancing lesion in the third ventricle. There is marked enlargement of the lateral ventricles due to associated noncommunicating hydrocephalus.

Intraventricular neurocysticercosis may be seen as the only form of CNS disease or in combination with parenchymal disease. Typically a solitary cyst is present in the fourth ventricle or in the third ventricle. Lateral ventricle or cerebral aqueduct cysts are less common.8 Migration of the intraventricular cyst may occur.142 Complications include ependymitis with development of ependymal adhesions and obstructive hydrocephalus (Fig. 10-39). Imaging diagnosis is challenging because fluid within the cyst is often the same density and signal intensity as CSF on CT and MRI. Occasionally the cyst wall may be visible as a T2 hypointense linear structure, or cyst fluid may show hyperintensity on FLAIR or T1-weighted sequences. Rarely the internal scolex may be visualized or the cyst wall may show enhancement. Usually the cyst is evident on imaging only because of ventricular deformity, distention, or hydrocephalus74 (Fig. 10-40). High-resolution MR cisternography sequences improve cyst visualization and localization, which may be useful for surgical planning.49 The subarachnoid form of neurocysticercosis may involve the basal cisterns and sylvian fissure. Basal meninges are frequently involved, resulting in the meningobasal or “racemose” form. The suprasellar, prepontine, or ambient cisterns are typically filled with complex cystic lesions with no visible scolex, usually resulting in meningitis. These lesions are difficult to visualize on CT and MRI owing to isointensity and isodensity to CSF, respectively. Enhancement is frequently absent in early disease. Meningobasilar neurocysticercosis is usually identified as a result of distortion of basal cisterns or presence of linear T2 hypointensities within the cisterns. Avid leptomeningeal enhancement may be seen; this is due to an inflammatory response causing meningitis, with frequent formation of meningeal adhesions and associated hydrocephalus (Fig. 10-41; see also Fig 10-36). Associated vasculitis may also occur, occasionally resulting in infarctions. Meningeal calcifications may form in late disease. Occasionally large cysticercosis cysts

may form in the sylvian fissures. Rarely neurocysticercosis may involve leptomeninges of the spine.74 MR cisternography sequences allow for improved visualization of intraventricular and leptomeningeal neurocysticercosis.49 T2-FLAIR sequences with supplemental 100% O2 increase CSF signal intensity and may also be helpful for characterization of extraaxial neurocysticercosis.14 MRS of the cyst fluid typically shows elevated choline, lactate, lipid, succinate, acetate, and alanine and decrease of other normal metabolites.98 MRS may be useful in differentiation of neurocysticercosis from a necrotic neoplasm, which demonstrates only lipid and lactate peaks in the nonenhancing central region. Racemous neurocysticercosis may show a pyruvate peak. Pyogenic abscess may demonstrate MRS findings similar to neurocysticercosis; however, the lack of diffusion restriction in neurocysticercosis allows for accurate diagnosis.31,108 MR perfusion may also help differentiate a parenchymal lesion from a neoplasm. Enhancing areas of inflammatory lesions such as neurocysticercosis typically demonstrate no significant rCBV elevation, in contradistinction to neoplasms, which demonstrate rCBV elevation.74,75

Echinococcosis Echinococcosis or hydatid disease is caused by the tapeworm larvae  of two Echinococcus species, E. granulosus and E. multilocularis. This disease is endemic in Middle East, Mediterranean region, South America, and Australia. Dogs or other carnivores are the definitive hosts for Echinococcus, and sheep or other ruminants are the intermediate hosts. Humans are infected by accidental ingestion of food or water contaminated with dog fecal material containing ova of the tapeworm. Liver and lungs are involved most frequently, with CNS involvement reported in 3% of cases.131 E. granulosus infection is more common, typically presenting on CT and MRI as a solitary supratentorial, MCA territory, round cyst. CT usually demonstrates internal fluid similar in attenuation to CSF and an iso- to hyperdense thin wall (Fig. 10-42). MRI usually

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

299

B

C FIG 10-40  Intraventricular cysticercosis. A, Sagittal T1-weighted image shows expansion of the fourth ventricle due to a large thin-walled cyst. Note abnormally rounded morphology of the fourth ventricle and blunting of the fastigial point. Isointense eccentric internal focus represents the scolex. B, FLAIR image demonstrates thin wall of the cyst in the posterior fourth ventricle. C, Diffusion-weighted image shows restricted diffusion in the scolex.

demonstrates internal fluid isointense to CSF on all sequences and a well-defined smooth, thin, T1 and T2 hypointense wall. Although most cysts show no enhancement, partial or complete ring enhancement is occasionally seen. Most cases demonstrate no surrounding edema unless the cyst is superinfected, helping in differentiation from an abscess. Rarely calcifications or daughter cysts may be seen. Multiple cysts may result from rupture of a solitary cystic lesion.6 Recurrent disease following surgery may present with intense enhancement of the cyst wall and surrounding edema.33,141

E. multilocularis infection is rare and may mimic malignancy. The lesions demonstrate a semisolid structure with multiple vesicles in a dense connective tissue stroma. Calcifications and surrounding edema are frequently seen.6

Paragonimiasis Paragonimiasis is caused by the trematode Paragonimus westermani, which is endemic in East Asia and Southeast Asia. Humans are infected by ingestion of undercooked freshwater crabs and crayfish. The lung

http://radiologyebook.com

300

PART II  CT and MR Imaging of the Whole Body

A

B FIG 10-41  Subarachnoid “racemose” form of neurocysticercosis A, T2-weighted image demonstrates prepontine cistern distortion and poorly defined linear hypointense septations within the cistern. B, Postcontrast T1-weighted image shows linear enhancement of the leptomeninges and linear septations in the prepontine cistern. Note poor definition of individual cysts and no identifiable scolex. This patient had associated hydrocephalus and coexisting parenchymal lesions (same patient as Fig. 3-39).

A

B FIG 10-42  Echinococcosis. A and B, Contrast-enhanced CT scans show a large lobulated cyst with marginal calcification (arrow in B) and minimal cyst wall enhancement.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation is the most common site of involvement, with brain involvement seen in 2% to 27% of cases. Granulomatous reaction occurs in the brain in response to the organism. MRI may reveal multiple ring enhancing, conglomerate lesions with surrounding edema and frequent associated hemorrhage. A tunnel sign is a T1 hypointense and T2 hyperintense columnar or fusiform lesion that may be seen in some cases.1 Most of the chronic granulomas calcify. Plain films show characteristic “soap bubble” intracranial calcifications. CT scans show multiple densely calcified areas with various nodular shapes around and within the lateral ventricles and high over the cerebral convexities.139,141

Toxoplasmosis Toxoplasmosis is caused by an intracellular protozoan, Toxoplasma gondii. Humans are infected by ingestion of cysts in undercooked pork or food contaminated with cat feces. Initial infection with rapidly multiplying tachyzoites is subclinical or mildly asymptomatic in immunocompetent patients. Subsequently, latent infection develops, with formation of cysts containing nonreplicating bradyzoites in different tissues including the brain. Reactivation of latent disease occurs in immunocompromised individuals, which is seen most frequently in AIDS. Toxoplasmosis is the most common opportunistic infection of CNS in AIDS despite recent decline due to HAART.6 Toxoplasmosis usually presents as multiple parenchymal abscesses, typically seen in the basal ganglia, thalamus, and corticomedullary junction. A solitary lesion is demonstrated on MRI in 14% of cases.104 CT typically demonstrates hypodense lesions with surrounding edema. Rarely, hyperdense toxoplasmosis lesions may be seen as a result of intralesional hemorrhage. Primary CNS lymphoma (PCNSL), in contrast, typically presents with hyperdense focal lesions. MRI may show T1 iso- to hypointense ring or nodular enhancing lesions with variable signal intensity on T2-weighted images and extensive surrounding edema (Fig. 10-43). The eccentric target sign, representing a ring enhancing lesion with internal eccentric enhancing nodule, is seen in 30% of cases and is 95% specific.67 The concentric target sign has been recently described in deep parenchymal lesions on T2-weighted images as a central hyperintensity surrounded by at least three zones of alternating bands of hypointensity and hyperintensity, not including the surrounding edema.79 Toxoplasmosis and PCNSL typically present as solitary or multiple enhancing parenchymal masses in AIDS patients. Differentiation of toxoplasmosis lesions from PCNSL masses remains a diagnostic challenge. Presence of subependymal spread and ventricular encasement are characteristic for PCNSL. Antitoxoplasmal therapy response evaluation is important for evaluation of such lesions in AIDS patients. Toxoplasmosis lesions typically decrease in size and number within 2 to 3 weeks after initiation of therapy. Treated lesions may calcify or hemorrhage.48 Thallium-201 single-photon emission CT (SPECT) and 18F FDG PET may be used for further evaluation of equivocal lesions and demonstrate avid uptake in PCNSL and poor uptake in toxoplasmosis. MRS has been described to exhibit high lipid and lactate with decreased or absent normal metabolites, in contrast with PCNSL, which typically exhibits marked elevation of choline with mild increase in lipid and lactate. However, recent investigations have shown that MRS is of limited value in this setting, in part owing to difficulties in voxel placement.24 Diffusion imaging is of limited value in this setting because significant overlap in ADC values exists between toxoplasmosis and lymphoma.115 MR perfusion imaging may be helpful, with significantly higher regional cerebral blood volume values seen in lymphoma in comparison to toxoplasmosis lesions.36 In newborns with toxoplasmosis, noncontrast CT scans may show hydrocephalus and scattered calcifications in the parenchyma as well

301

as in the periventricular region (Fig. 10-44). These findings may be seen in cytomegalic inclusion disease, herpes encephalitis, and congenital rubella involving newborns. With cytomegalic inclusion disease, however, calcifications are seen predominantly in the periventricular region.141

INFLAMMATORY DISEASES Sarcoidosis Sarcoidosis is a systemic granulomatous inflammatory disease that most frequently involves the lungs and lymph nodes. It usually presents in the third and fourth decades of life and is more common in females; 10% of cases demonstrate CNS involvement on imaging. Rarely CNS involvement may be the only manifestation of this disease, presenting as parenchymal lesions, leptomeningitis, hypothalamic and pituitary involvement, cranial nerve involvement, and dural involvement. Orbital involvement is occasionally seen and may mimic orbital pseudotumor. Occasionally, well-circumscribed “punched-out” skull lesions may be seen.58,120 Leptomeningeal involvement is the most common form of CNS sarcoidosis and is typically seen in the basal cisterns. Diffuse or nodular meningeal enhancement and thickening is seen on MRI (Fig. 10-45). Spread along perivascular spaces and associated formation of parenchymal enhancing lesions may be seen. Some cases demonstrate involvement of the hypothalamus, pituitary stalk, and rarely the pituitary gland. Secondary communicating hydrocephalus may develop. Findings are similar to leptomeningeal involvement by lymphoma and tuberculosis. Cranial nerve involvement may be an isolated finding or may be seen in association with meningeal disease. Involvement of the optic nerves and facial nerves is seen more frequently.140 Parenchymal lesions are typically seen as nonenhancing periventricular T2 hyperintense lesions similar in appearance to vascular disease or MS. This may represent changes secondary to vascular involvement by this disease.32 Solitary or multiple enhancing parenchymal lesions may also occur.132 Usually these granulomatous lesions demonstrate T2 hypointensity and homogeneous enhancement. CT typically demonstrates iso- to hyperdense lesions. Differential diagnosis may include tumefactive demyelinating lesions and neoplasms. Some lesions may result from parenchymal extension of meningitis via perivascular spaces.26 Dural involvement may manifest as diffuse dural thickening or as focal dural mass lesions. These lesions are typically T2 hypointense and show homogeneous enhancement. Differential diagnosis includes meningioma, lymphoma, and dural metastases71,120 (Fig. 10-46).

Granulomatosis with Polyangiitis (Wegener’s Granulomatosis) Granulomatosis with polyangiitis (GPA), formerly known as Wegener’s granulomatosis, is a systemic granulomatous disease that may involve the lungs, kidneys, heart, orbits, paranasal sinuses, skin, joints, or peripheral nerves. Occasionally, intracranial involvement may be seen. MRI may demonstrate diffuse or focal pachymeningeal thickening and enhancement. Nonspecific T2-FLAIR hyperintense lesions may be seen in the cerebral white matter and brainstem. Parenchymal granulomas have also been reported. Granulomatous vasculitis may result in cerebral infarcts or hemorrhages.106

Behçet’s Disease Behçet’s disease is a chronic inflammatory systemic disorder that manifests as vasculitis involving the GI, cardiovascular, pulmonary, and nervous systems. This disease is most common in the eastern

http://radiologyebook.com

302

A

PART II  CT and MR Imaging of the Whole Body

B

C FIG 10-43  Toxoplasmosis. A, T2-weighted image shows a low-signal-intensity lesion with thin hyperintense and hypointense rims. Extensive surrounding edema is seen. B, Postcontrast T1-weighted image shows two ring enhancing lesions in the left frontal lobe, with the superior lesion corresponding to the lesion seen on T2-weighted image. C, Diffusion-weighted image demonstrates low signal within the lesions, allowing differentiation from a pyogenic abscess.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

303

Mediterranean and eastern Asia. Neurologic involvement is seen in 10% to 50% of cases. Parenchymal involvement is characteristically seen as T2-FLAIR hyperintense and CT hypodense small focal lesions in the brainstem, especially in cerebral peduncles and pons. Similar lesions may be seen in the thalamus and brainstem. Nodular or ringlike enhancement may be seen (Fig. 10-47). Less frequently, cerebral and spinal cord lesions may be seen. Meningitis and vascular manifestations may also occur, including dural venous thrombosis, aneurysm formation, and arterial thrombosis.20,66,73 The differential diagnosis on imaging studies includes MS, primary arteritis of the CNS, and sarcoidosis.141

Whipple’s Disease Whipple’s disease is a systemic disease caused by Tropheryma whippelii and may involve the GI tract, CNS, and cardiac tissues. MRI may show T2-FLAIR hyperintense symmetric lesions in the midbrain, hypothalamus, and mesial temporal lobes and occasionally within the corticospinal tracts.84 No diffusion restriction is seen, and only minimal enhancement may be demonstrated.13 Early diagnosis and antibiotic treatment may result in improved symptoms.

UNUSUAL PEDIATRIC ENCEPHALITIDES Rasmussen’s Encephalitis FIG 10-44  Neonatal toxoplasmosis. Noncontrast CT scan shows hydrocephalus and scattered intracranial calcifications, with a few seen in the periventricular region.

A

Rasmussen’s encephalitis is a chronic progressive inflammatory disease, usually seen in childhood. Patients usually present with intractable focal motor seizures and may develop hemiplegia and cognitive deterioration. Imaging may be normal early in this disease. MRI in more

B FIG 10-45  Sarcoidosis. A, FLAIR image shows diffuse irregular, abnormal high signal intensity extending from the subarachnoid space into the brain parenchyma of frontal lobes and basal ganglia. B, Gadoliniumenhanced T1-weighted image demonstrates patchy contrast enhancement involving the frontal lobes and basal ganglia.

http://radiologyebook.com

304

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 10-46  Sarcoid. A, T1-weighted image shows an isointense extraaxial lesion near the vertex. B, T2weighted image shows the lesion to be of low signal intensity. C, Postcontrast T1-weighted image demonstrates intense enhancement of the lesion, which is better visualized compared to noncontrast images.

advanced disease may demonstrate atrophy and T2-FLAIR hyperintensities in the gray and white matter, predominantly involving one cerebral hemisphere (Fig. 10-48). 18F FDG PET demonstrates decreased activity within the affected hemisphere.41,45

Acute Necrotizing Encephalitis (ANE) ANE occurs in infants and young children following a mycoplasma or viral disease (HSV-6, influenza A and B) and is seen predominantly in

the East Asia region. Patients present with acute encephalopathy, seizures, and impaired consciousness. Serum aminotransferase may be elevated, and CSF analysis may show increased protein with lack of pleocytosis. Prognosis is variable, ranging from full recovery to severe morbidity and death. MRI characteristically shows FLAIR-T2 hyperintense bilateral, symmetric, hemorrhagic thalamic lesions (Fig. 10-49A and B). Diffusion restriction and rim enhancement may be seen in thalamic lesions

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation

A

B FIG 10-47  Behçet’s disease. A and B, Axial and coronal postcontrast T1-weighted image showing multiple enhancing lesions in the pons and in bilateral cerebral peduncles. (Courtesy Dr. Bruno Telles.)

A

B FIG 10-48  Rasmussen’s encephalitis. A, T1-weighted image demonstrates volume loss predominantly involving the left cerebral hemisphere and left basal ganglia. B, Coronal T2-weighted image demonstrates associated asymmetric widening of left sylvian fissure and lateral ventricle.

http://radiologyebook.com

305

306

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 10-49  Acute necrotizing encephalitis. A, T1-weighted image reveals bilateral thalamic hemorrhage. B, FLAIR image demonstrates hyperintensity in both thalami, with adjacent edema. C, MR spectroscopy demonstrates elevated lipids, glutamine/glutamate complex, and lactate.

in the acute stage. Frequently other symmetric lesions are demonstrated in brainstem tegmentum, periventricular white matter, internal capsule, putamen, and cerebellum. CT may show hypodensities in these regions. MRS may show elevation in glutamate/glutamine complex, lipids, and lactate7,47,77 (see Fig. 10-49C).

REFERENCES 1. Abdel Razek AA, Watcharakorn A, Castillo M: Parasitic diseases of the central nervous system. Neuroimaging Clin N Am 21(4):815–841, viii, 2011. 2. Agarwal R, Sze G: Neuro-lyme disease: MR imaging findings. Radiology 253(1):167–173, 2009.

3. Agosta F, Rocca MA, Benedetti B, et al: MR imaging assessment of brain and cervical cord damage in patients with neuroborreliosis. AJNR Am J Neuroradiol 27(4):892–894, 2006. 4. Ahmadi J, Segall HD, Zee CS, et al: Computed tomography   in childhood intracranial infections. Semin Neurol 2:166–188, 1982. 5. Aiken AH: Central nervous system infection. Neuroimaging Clin N Am 20(4):557–580, 2010. 6. Akgoz A, Mukundan S, Lee TC: Imaging of rickettsial, spirochetal,   and parasitic infections. Neuroimaging Clin N Am 22(4):633–657, 2012. 7. Albayram S, Bilgi Z, Selcuk H, et al: Diffusion-weighted MR imaging findings of acute necrotizing encephalopathy. AJNR Am J Neuroradiol 25(5):792–797, 2004.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation 8. Amaral L, Maschietto M, Maschietto R, et al: Unusual manifestations   of neurocysticercosis in MR imaging: Analysis of 172 cases. Arq Neuropsiquiatr 61(3A):533–541, 2003. 9. Aribandi M, McCoy VA, Bazan C, 3rd.: Imaging features of invasive and noninvasive fungal sinusitis: A review. Radiographics 27(5):1283–1296, 2007. 10. Bargallo J, Berenguer J, Garcia-Barrionuevo J, et al: The “target sign”: Is it a specific sign of CNS tuberculoma? Neuroradiology 38(6):547–550, 1996. 11. Berger JR, Levy RM, Flomenhoft D, et al: Predictive factors for prolonged survival in acquired immunodeficiency syndrome-associated progressive multifocal leukoencephalopathy. Ann Neurol 44(3):341–349, 1998. 12. Biondi A: Intracranial aneurysms associated with other lesions, disorders or anatomic variations. Neuroimaging Clin N Am 16(3):467– 482, viii, 2006. 13. Black DF, Aksamit AJ, Morris JM: MR imaging of central nervous system Whipple disease: A 15-year review. AJNR Am J Neuroradiol 31(8):1493–1497, 2010. 14. Braga F, Rocha AJ, Gomes HR, et al: Noninvasive MR cisternography with fluid-attenuated inversion recovery and 100% supplemental O(2) in the evaluation of neurocysticercosis. AJNR Am J Neuroradiol 25(2):295–297, 2004. 15. Britt RH, Enzmann DR: Clinical stages of human brain abscesses on serial CT scans after contrast infusion. Computerized tomographic, neuropathological, and clinical correlations. J Neurosurg 59(6):972–989, 1983. 16. Burtscher IM, Holtas S: In vivo proton MR spectroscopy of untreated and treated brain abscesses. AJNR Am J Neuroradiol 20(6):1049–1053, 1999. 17. Cartes-Zumelzu FW, Stavrou I, Castillo M, et al: Diffusion-weighted imaging in the assessment of brain abscesses therapy. AJNR Am J Neuroradiol 25(8):1310–1317, 2004. 18. Castillo M: Imaging of meningitis. Semin Roentgenol 39(4):458–464, 2004. 19. Castillo M, Thurnher M: Imaging viral and prion infections. Semin Roentgenol 39(4):482–494, 2004. 20. Chae EJ, Do KH, Seo JB, et al: Radiologic and clinical findings of Behcet disease: Comprehensive review of multisystemic involvement. Radiographics 28(5):e31, 2008. 21. Chang L, Ernst T, Tornatore C, et al: Metabolite abnormalities in progressive multifocal leukoencephalopathy by proton magnetic resonance spectroscopy. Neurology 48(4):836–845, 1997. 22. Charlier C, Dromer F, Leveque C, et al: Cryptococcal neuroradiological lesions correlate with severity during cryptococcal meningoencephalitis in HIV-positive patients in the HAART era. PLoS ONE 3(4):e1950, 2008. 23. Chiang IC, Hsieh TJ, Chiu ML, et al: Distinction between pyogenic brain abscess and necrotic brain tumour using 3-tesla MR spectroscopy, diffusion and perfusion imaging. Br J Radiol 82(982):813–820, 2009. 24. Chinn RJ, Wilkinson ID, Hall-Craggs MA, et al: Toxoplasmosis and primary central nervous system lymphoma in HIV infection: Diagnosis with MR spectroscopy. Radiology 197(3):649–654, 1995. 25. Chong J, Di Rocco A, Tagliati M, et al: MR findings in AIDS-associated myelopathy. AJNR Am J Neuroradiol 20(8):1412–1416, 1999. 26. Christoforidis GA, Spickler EM, Recio MV, et al: MR of CNS sarcoidosis: Correlation of imaging features to clinical symptoms and response to treatment. AJNR Am J Neuroradiol 20(4):655–669, 1999. 27. Cinque P, Koralnik IJ, Gerevini S, et al: Progressive multifocal leukoencephalopathy in HIV-1 infection. Lancet Infect Dis 9(10):625– 636, 2009. 28. Clifford DB, De Luca A, Simpson DM, et al: Natalizumab-associated progressive multifocal leukoencephalopathy in patients with multiple sclerosis: Lessons from 28 cases. Lancet Neurol 9(4):438–446, 2010. 29. Collie DA, Summers DM, Sellar RJ, et al: Diagnosing variant Creutzfeldt-Jakob disease with the pulvinar sign: MR imaging findings in 86 neuropathologically confirmed cases. AJNR Am J Neuroradiol 24(8):1560–1569, 2003.

307

30. De Bruecker Y, Claus F, Demaerel P, et al: MRI findings in acute cerebellitis. Eur Radiol 14(8):1478–1483, 2004. 31. do Amaral LL, Ferreira RM, da Rocha AJ, et al: Neurocysticercosis: Evaluation with advanced magnetic resonance techniques and atypical forms. Top Magn Reson Imaging 16(2):127–144, 2005. 32. Dumas JL, Valeyre D, Chapelon-Abric C, et al: Central nervous system sarcoidosis: Follow-up at MR imaging during steroid therapy. Radiology 214(2):411–420, 2000. 33. El-Shamam O, Amer T, El-Atta MA: Magnetic resonance imaging of simple and infected hydatid cysts of the brain. Magn Reson Imaging 19(7):965–974, 2001. 34. Enzmann DR, Brant-Zawadzki M, Britt RH: CT of central nervous system infections in immunocompromised patients. AJR Am J Roentgenol 135(2):263–267, 1980. 35. Erly WK, Bellon RJ, Seeger JF, et al: MR imaging of acute coccidioidal meningitis. AJNR Am J Neuroradiol 20(3):509–514, 1999. 36. Ernst TM, Chang L, Witt MD, et al: Cerebral toxoplasmosis and lymphoma in AIDS: Perfusion MR imaging experience in 13 patients. Radiology 208(3):663–669, 1998. 37. Ernst T, Chang L, Witt M, et al: Progressive multifocal leukoencephalopathy and human immunodeficiency virus-associated white matter lesions in AIDS: Magnetization transfer MR imaging. Radiology 210(2):539–543, 1999. 38. Fanning NF, Laffan EE, Shroff MM: Serial diffusion-weighted MRI correlates with clinical course and treatment response in children with intracranial pus collections. Pediatr Radiol 36(1):26–37, 2006. 39. Ferreira NP, Otta GM, do Amaral LL, et al: Imaging aspects of pyogenic infections of the central nervous system. Top Magn Reson Imaging 16(2):145–154, 2005. 40. Filippi CG, Ulug AM, Ryan E, et al: Diffusion tensor imaging of patients with HIV and normal-appearing white matter on MR images of the brain. AJNR Am J Neuroradiol 22(2):277–283, 2001. 41. Fiorella DJ, Provenzale JM, Coleman RE, et al: (18) F-fluorodeoxyglucose positron emission tomography and MR imaging findings in Rasmussen encephalitis. AJNR Am J Neuroradiol 22(7):1291–1299, 2001. 42. Fish DG, Ampel NM, Galgiani JN, et al: Coccidioidomycosis during human immunodeficiency virus infection. A review of 77 patients. Medicine 69(6):384–391, 1990. 43. Fukui MB, Williams RL, Mudigonda S: CT and MR imaging features   of pyogenic ventriculitis. AJNR Am J Neuroradiol 22(8):1510–1516, 2001. 44. Garcia HH, Del Brutto OH: Imaging findings in neurocysticercosis. Acta Trop 87(1):71–78, 2003. 45. Geller E, Faerber EN, Legido A, et al: Rasmussen encephalitis: Complementary role of multitechnique neuroimaging. AJNR Am J Neuroradiol 19(3):445–449, 1998. 46. Gilden D, Cohrs RJ, Mahalingam R, et al: Varicella zoster virus vasculopathies: Diverse clinical manifestations, laboratory features, pathogenesis, and treatment. Lancet Neurol 8(8):731–740, 2009. 47. Goo HW, Choi CG, Yoon CH, et al: Acute necrotizing encephalopathy: Diffusion MR imaging and localized proton MR spectroscopic findings in two infants. Korean J Radiol 4(1):61–65, 2003. 48. Gottumukkala RV, Romero JM, Riascos RF, et al: Imaging of the brain in patients with human immunodeficiency virus infection. Top Magn Reson Imaging 23(5):275–291, 2014. 49. Govindappa SS, Narayanan JP, Krishnamoorthy VM, et al: Improved detection of intraventricular cysticercal cysts with the use of threedimensional constructive interference in steady state MR sequences. AJNR Am J Neuroradiol 21(4):679–684, 2000. 50. Grand S, Passaro G, Ziegler A, et al: Necrotic tumor versus brain abscess: Importance of amino acids detected at 1H MR spectroscopy– Initial results. Radiology 213(3):785–793, 1999. 51. Gupta RK, Kathuria MK, Pradhan S: Magnetization transfer MR imaging in CNS tuberculosis. AJNR Am J Neuroradiol 20(5):867–875, 1999. 52. Gupta RK, Roy R, Dev R, et al: Finger printing of Mycobacterium tuberculosis in patients with intracranial tuberculomas by using in vivo,

http://radiologyebook.com

308

53.

54. 55.

56.

57.

58.

59.

60. 61.

62. 63.

64. 65. 66.

67.

68.

69.

70.

71.

72.

73. 74. 75. 76.

PART II  CT and MR Imaging of the Whole Body ex vivo, and in vitro magnetic resonance spectroscopy. Magn Reson Med 36(6):829–833, 1996. Gupta RK, Vatsal DK, Husain N, et al: Differentiation of tuberculous from pyogenic brain abscesses with in vivo proton MR spectroscopy and magnetization transfer MR imaging. AJNR Am J Neuroradiol 22(8):1503–1509, 2001. Handel SF, Klein WC, Kim YW: Intracranial epidural abscess. Radiology 111(1):117–120, 1974. Handique SK, Das RR, Barman K, et al: Temporal lobe involvement in Japanese encephalitis: Problems in differential diagnosis. AJNR Am J Neuroradiol 27(5):1027–1031, 2006. Haris M, Gupta RK, Husain M, et al: Assessment of therapeutic response in brain tuberculomas using serial dynamic contrast-enhanced MRI. Clin Radiol 63(5):562–574, 2008. Hasbun R, Abrahams J, Jekel J, et al: Computed tomography of the head before lumbar puncture in adults with suspected meningitis.   N Engl J Med 345(24):1727–1733, 2001. Hayes WS, Sherman JL, Stern BJ, et al: MR and CT evaluation of intracranial sarcoidosis. AJR Am J Roentgenol 149(5):1043–1049, 1987. Hazany S, Go JL, Law M: Magnetic resonance imaging of infectious meningitis and ventriculitis in adults. Top Magn Reson Imaging 23(5):315–325, 2014. Hoxworth JM, Glastonbury CM: Orbital and intracranial complications of acute sinusitis. Neuroimaging Clin N Am 20(4):511–526, 2010. Hughes DC, Raghavan A, Mordekar SR, et al: Role of imaging in the diagnosis of acute bacterial meningitis and its complications. Postgrad Med J 86(1018):478–485, 2010. Jones KC, Benseler SM, Moharir M: Anti-NMDA receptor encephalitis. Neuroimaging Clin N Am 23(2):309–320, 2013. Kamra P, Azad R, Prasad KN, et al: Infectious meningitis: Prospective evaluation with magnetization transfer MRI. Br J Radiol 77(917):387– 394, 2004. Kanamalla US, Ibarra RA, Jinkins JR: Imaging of cranial meningitis and ventriculitis. Neuroimaging Clin N Am 10(2):309–331, 2000. Kastrup O, Wanke I, Maschke M: Neuroimaging of infections of the central nervous system. Semin Neurol 28(4):511–522, 2008. Kocer N, Islak C, Siva A, et al: CNS involvement in neuro-Behcet syndrome: An MR study. AJNR Am J Neuroradiol 20(6):1015–1024, 1999. Kumar GG, Mahadevan A, Guruprasad AS, et al: Eccentric target sign in cerebral toxoplasmosis: Neuropathological correlate to the imaging feature. J Magn Reson Imaging 31(6):1469–1472, 2010. Lai PH, Hsu SS, Ding SW, et al: Proton magnetic resonance spectroscopy and diffusion-weighted imaging in intracranial cystic mass lesions. Surg Neurol 68(Suppl 1):S25–S36, 2007. Lai PH, Li KT, Hsu SS, et al: Pyogenic brain abscess: Findings from in vivo 1.5-T and 11.7-T in vitro proton MR spectroscopy. AJNR Am J Neuroradiol 26(2):279–288, 2005. Lammering JC, Iv M, Gupta N, et al: Imaging spectrum of CNS coccidioidomycosis: Prevalence and significance of concurrent brain and spinal disease. AJR Am J Roentgenol 200(6):1334–1346, 2013. Larner AJ, Ball JA, Howard RS: Sarcoid tumour: Continuing diagnostic problems in the MRI era. J Neurol Neurosurg Psychiatry 66(4):510–512, 1999. LeBlang SD, Whiteman ML, Post MJ, et al: CNS nocardia in AIDS patients: CT and MRI with pathologic correlation. J Comput Assist Tomogr 19(1):15–22, 1995. Lee SH, Yoon PH, Park SJ, et al: MRI findings in neuro-Behcet’s disease. Clin Radiol 56(6):485–494, 2001. Lerner A, Shiroishi MS, Zee CS, et al: Imaging of neurocysticercosis. Neuroimaging Clin N Am 22(4):659–676, 2012. Lev MH, Rosen BR: Clinical applications of intracranial perfusion MR imaging. Neuroimaging Clin N Am 9(2):309–331, 1999. Lin M, Darwish BS, Chu J: Neurosyphilitic gumma on F18-2-fluoro-2deoxy-D-glucose (FDG) positron emission tomography: An old disease investigated with a new technology. J Clin Neurosci 16(3):410–412, 2009.

77. Lo CP, Chen CY: Neuroimaging of viral infections in infants and young children. Neuroimaging Clin N Am 18(1):119–132, viii, 2008. 78. Luthra G, Parihar A, Nath K, et al: Comparative evaluation of fungal, tubercular, and pyogenic brain abscesses with conventional and diffusion MR imaging and proton MR spectroscopy. AJNR Am J Neuroradiol 28(7):1332–1338, 2007. 79. Mahadevan A, Ramalingaiah AH, Parthasarathy S, et al: Neuropathological correlate of the “concentric target sign” in MRI of HIV-associated cerebral toxoplasmosis. J Magn Reson Imaging 38(2):488–495, 2013. 80. Mamelak AN, Obana WG, Flaherty JF, et al: Nocardial brain abscess: Treatment strategies and factors influencing outcome. Neurosurgery 35(4):622–631, 1994. 81. Marin SE, Callen DJ: The magnetic resonance imaging appearance of monophasic acute disseminated encephalomyelitis: An update post application of the 2007 consensus criteria. Neuroimaging Clin N Am 23(2):245–266, 2013. 82. Martinez del Valle MD, Gomez-Rio M, Horcajadas A, et al: False positive thallium-201 SPECT imaging in brain abscess. Br J Radiol 73(866):160–164, 2000. 83. Mathur M, Johnson CE, Sze G: Fungal infections of the central nervous system. Neuroimaging Clin N Am 22(4):609–632, 2012. 84. Mendel E, Khoo LT, Go JL, et al: Intracerebral Whipple’s disease diagnosed by stereotactic biopsy: A case report and review of the literature. Neurosurgery 44(1):203–209, 1999. 85. Menon DK, Sargentoni J, Peden CJ, et al: Proton MR spectroscopy in herpes simplex encephalitis: Assessment of neuronal loss. J Comput Assist Tomogr 14(3):449–452, 1990. 86. Mishra AM, Gupta RK, Saksena S, et al: Biological correlates of diffusivity in brain abscess. Magn Reson Med 54(4):878–885, 2005. 87. Mittl RL, Jr, Yousem DM: Frequency of unexplained meningeal enhancement in the brain after lumbar puncture. AJNR Am J Neuroradiol 15(4):633–638, 1994. 88. Modi G, Campbell H, Bill P: Subacute sclerosing panencephalitis. Changes on CT scan during acute relapse. Neuroradiology 31(5):433– 434, 1989. 89. Mohan S, Jain KK, Arabi M, et al: Imaging of meningitis and ventriculitis. Neuroimaging Clin N Am 22(4):557–583, 2012. 90. Nash TE, Pretell J, Garcia HH: Calcified cysticerci provoke perilesional edema and seizures. Clin Infect Dis 33(10):1649–1653, 2001. 91. Nash TE, Pretell EJ, Lescano AG, et al: Perilesional brain oedema and seizure activity in patients with calcified neurocysticercosis: A prospective cohort and nested case-control study. Lancet Neurol 7(12):1099–1105, 2008. 92. Nath K, Agarwal M, Ramola M, et al: Role of diffusion tensor imaging metrics and in vivo proton magnetic resonance spectroscopy in the differential diagnosis of cystic intracranial mass lesions. Magn Reson Imaging 27(2):198–206, 2009. 93. Nath K, Husain M, Trivedi R, et al: Clinical implications of increased fractional anisotropy in meningitis associated with brain abscess.   J Comput Assist Tomogr 31(6):888–893, 2007. 94. Nath K, Ramola M, Husain M, et al: Assessment of therapeutic response in patients with brain abscess using diffusion tensor imaging. World Neurosurg 73(1):63–68, discussion e66, 2010. 95. Noujaim SE, Rossi MD, Rao SK, et al: CT and MR imaging of neurocysticercosis. AJR Am J Roentgenol 173(6):1485–1490, 1999. 96. Pahud BA, Glaser CA, Dekker CL, et al: Varicella zoster disease of the central nervous system: Epidemiological, clinical, and laboratory features 10 years after the introduction of the varicella vaccine. J Infect Dis 203(3):316–323, 2011. 97. Pal D, Bhattacharyya A, Husain M, et al: In vivo proton MR spectroscopy evaluation of pyogenic brain abscesses: A report of 194 cases. AJNR Am J Neuroradiol 31(2):360–366, 2010. 98. Pandit S, Lin A, Gahbauer H, et al: MR spectroscopy in neurocysticercosis. J Comput Assist Tomogr 25(6):950–952, 2001. 99. Parmar H, Ibrahim M: Pediatric intracranial infections. Neuroimaging Clin N Am 22(4):707–725, 2012.

http://radiologyebook.com

CHAPTER 10  Cerebral Infections and Inflammation 100. Parmar H, Sitoh YY, Anand P, et al: Contrast-enhanced FLAIR imaging in the evaluation of infectious leptomeningeal diseases. Eur J Radiol 58(1):89–95, 2006. 101. Patkar D, Narang J, Yanamandala R, et al: Central nervous system tuberculosis: Pathophysiology and imaging findings. Neuroimaging Clin N Am 22(4):677–705, 2012. 102. Petropoulou KA, Gordon SM, Prayson RA, et al: West Nile virus meningoencephalitis: MR imaging findings. AJNR Am J Neuroradiol 26(8):1986–1995, 2005. 103. Pomara N, Crandall DT, Choi SJ, et al: White matter abnormalities in HIV-1 infection: A diffusion tensor imaging study. Psychiatry Res 106(1):15–24, 2001. 104. Porter SB, Sande MA: Toxoplasmosis of the central nervous system in the acquired immunodeficiency syndrome. N Engl J Med 327(23):1643– 1648, 1992. 105. Post MJ, Yiannoutsos C, Simpson D, et al: Progressive multifocal leukoencephalopathy in AIDS: Are there any MR findings useful to patient management and predictive of patient survival? AIDS Clinical Trials Group, 243 Team. AJNR Am J Neuroradiol 20(10):1896–1906, 1999. 106. Provenzale JM, Allen NB: Wegener granulomatosis: CT and MR findings. AJNR Am J Neuroradiol 17(4):785–792, 1996. 107. Quint DJ, Eldevik OP, Cohen JK: Magnetic resonance imaging of normal meningeal enhancement at 1.5 T. Acad Radiol 3(6):463–468, 1996. 108. Raffin LS, Bacheschi LA, Machado LR, et al: Diffusion-weighted MR imaging of cystic lesions of neurocysticercosis: A preliminary study. Arq Neuropsiquiatr 59(4):839–842, 2001. 109. Rath TJ, Hughes M, Arabi M, et al: Imaging of cerebritis, encephalitis, and brain abscess. Neuroimaging Clin N Am 22(4):585–607, 2012. 110. Rosenbaum ML, Hoff JT, Norman D: Decreased mortality from brain abscesses since the advent of computerized tomography. J Neurosurg 49:658–668, 1978. 111. Rossi A: Imaging of acute disseminated encephalomyelitis. Neuroimaging Clin N Am 18(1):149–161, ix, 2008. 112. Saez-Llorens XJ, Umana MA, Odio CM, et al: Brain abscess in infants and children. Pediatr Infect Dis J 8(7):449–458, 1989. 113. Safder S, Carpenter JS, Roberts TD, et al: The “black turbinate” sign: An early MR imaging finding of nasal mucormycosis. AJNR Am J Neuroradiol 31(4):771–774, 2010. 114. Samann PG, Schlegel J, Muller G, et al: Serial proton MR spectroscopy and diffusion imaging findings in HIV-related herpes simplex encephalitis. AJNR Am J Neuroradiol 24(10):2015–2019, 2003. 115. Schroeder PC, Post MJ, Oschatz E, et al: Analysis of the utility of diffusion-weighted MRI and apparent diffusion coefficient values in distinguishing central nervous system toxoplasmosis from lymphoma. Neuroradiology 48(10):715–720, 2006. 116. Shin JH, Lee HK: Nocardial brain abscess in a renal transplant recipient. Clin Imaging 27(5):321–324, 2003. 117. Simmons KC, Sage MR, Reilly PL: CT of intracerebral haemorrhage due to mycotic aneurysms–Case report. Neuroradiology 19(4):215–217, 1980. 118. Singh P, Goraya JS, Gupta K, et al: Magnetic resonance imaging findings in Reye syndrome: Case report and review of the literature. J Child Neurol 26(8):1009–1014, 2011. 119. Smith MM, Anderson JC: Neurosyphilis as a cause of facial and vestibulocochlear nerve dysfunction: MR imaging features. AJNR Am J Neuroradiol 21(9):1673–1675, 2000. 120. Smith JK, Matheus MG, Castillo M: Imaging manifestations of neurosarcoidosis. AJR Am J Roentgenol 182(2):289–295, 2004. 121. Smith AB, Smirniotopoulos JG, Rushing EJ: From the archives of the AFIP: Central nervous system infections associated with human

122.

123. 124.

125.

126.

127.

128. 129.

130. 131. 132.

133.

134.

135. 136.

137. 138.

139.

140. 141.

142.

309

immunodeficiency virus infection: Radiologic-pathologic correlation. Radiographics 28(7):2033–2058, 2008. Soto-Hernandez JL, Moreno-Andrade T, Gongora-Rivera F, et al: Nocardia abscess during treatment of brain toxoplasmosis in a patient with AIDS, utility of proton MR spectroscopy and diffusionweighted imaging in diagnosis. Clin Neurol Neurosurg 108(5):493–498, 2006. Stave GM, Heimberger T, Kerkering TM: Zygomycosis of the basal ganglia in intravenous drug users. Am J Med 86(1):115–117, 1989. Stuckey SL, Goh TD, Heffernan T, et al: Hyperintensity in the subarachnoid space on FLAIR MRI. AJR Am J Roentgenol 189(4):913– 921, 2007. Tan IL, McArthur JC, Clifford DB, et al: Immune reconstitution inflammatory syndrome in natalizumab-associated PML. Neurology 77(11):1061–1067, 2011. Terk MR, Underwood DJ, Zee CS, et al: MR imaging in rhinocerebral and intracranial mucormycosis with CT and pathologic correlation. Magn Reson Imaging 10(1):81–87, 1992. Thurnher MM, Castillo M, Stadler A, et al: Diffusion-tensor MR imaging of the brain in human immunodeficiency virus-positive patients. AJNR Am J Neuroradiol 26(9):2275–2281, 2005. Torres C, Riascos R, Figueroa R, et al: Central nervous system tuberculosis. Top Magn Reson Imaging 23(3):173–189, 2014. Tschampa HJ, Kallenberg K, Kretzschmar HA, et al: Pattern of cortical changes in sporadic Creutzfeldt-Jakob disease. AJNR Am J Neuroradiol 28(6):1114–1118, 2007. Tschampa HJ, Zerr I, Urbach H: Radiological assessment of CreutzfeldtJakob disease. Eur Radiol 17(5):1200–1211, 2007. Tuzun M, Hekimoglu B: Hydatid disease of the CNS: Imaging features. AJR Am J Roentgenol 171(6):1497–1500, 1998. Urbach H, Kristof R, Zentner J, et al: Sarcoidosis presenting as an intra- or extra-axial cranial mass: Report of two cases. Neuroradiology 39(7):516–519, 1997. Vachha B, Rojas R, Prabhu SP, et al: Magnetic resonance imaging in viral and prion diseases of the central nervous system. Top Magn Reson Imaging 23(5):293–302, 2014. van de Beek D, de Gans J, Spanjaard L, et al: Clinical features and prognostic factors in adults with bacterial meningitis. N Engl J Med 351(18):1849–1859, 2004. Wong J, Quint DJ: Imaging of central nervous system infections. Semin Roentgenol 34(2):123–143, 1999. Wong AM, Zimmerman RA, Simon EM, et al: Diffusion-weighted MR imaging of subdural empyemas in children. AJNR Am J Neuroradiol 25(6):1016–1021, 2004. Wood M, Anderson M: Neurological infections, Philadelphia, 1988, WB Saunders. Yadav A, Malik GK, Trivedi R, et al: Correlation of CSF neuroinflammatory molecules with leptomeningeal cortical subcortical white matter fractional anisotropy in neonatal meningitis. Magn Reson Imaging 27(2):214–221, 2009. Yoshida M, Moritaka K, Kuga S, et al: CT findings of cerebral paragonimiasis in the chronic state. J Comput Assist Tomogr 6(1):195– 196, 1982. Zajicek JP, Scolding NJ, Foster O, et al: Central nervous system sarcoidosis—Diagnosis and management. QJM 92(2):103–117, 1999. Zee CS, Geng D, Go JL, et al: Cerebral infections and inflammation. In Haaga JR, Dogra VS, Forsting M, et al, editors: CT and MR imaging of the whole body, vol 1, ed 5, Philadelphia, 2008, Mosby Elsevier, pp 145–186. Zee CS, Segall HD, Apuzzo ML, et al: Intraventricular cysticercal cysts: Further neuroradiologic observations and neurosurgical implications. AJNR Am J Neuroradiol 5(6):727–730, 1984.

http://radiologyebook.com

11  Stroke Simon Rupe Khangure, Cathy Soufan, Henry Ma, Thanh Phan, Stephen L. Stuckey, and Ronil V. Chandra

INTRODUCTION Stroke is the second leading cause of death worldwide.53 Globally an estimated 16.9 million cases of new stroke occurred in 2010, resulting in 5.9 million stroke deaths.53 In the United States, a stroke occurs every 40 seconds, with a stroke death every 4 minutes.64 Although stroke mortality rates have reduced over the last 2 decades, the number of stroke survivors and disability-adjusted life-years lost from stroke is increasing.53 Work capacity is reduced in 70% of stroke survivors, with 30% requiring assistance with self-care.140 By 2030, there will be an estimated 70 million stroke survivors globally.53 Stroke was defined by the World Health Organization in 1970 as “rapidly developing clinical signs of focal (or global) disturbance of cerebral function lasting more than 24 hours with no apparent cause other than of vascular origin.”2 Over the last 45 years since this definition was proposed, there have been advances in clinical assessment, imaging, and treatment of stroke. Thus while the term stroke can be used broadly to include ischemic stroke, cerebral hemorrhage, subarachnoid hemorrhage, and cerebral venous thrombosis, this chapter will focus on the imaging of ischemic cerebral stroke, which can be defined as neurologic dysfunction caused by focal cerebral arterial ischemia.64 The diagnosis and imaging of acute ischemic cerebral stroke can be challenging, particularly when time-critical decisions need to be made for stroke treatment. Furthermore, approximately 30% of patients evaluated by an acute stroke team have a stroke mimic; 15% of those receiving intravenous tissue plasminogen activator (ivtPA) are diagnosed with a stroke mimic.134 Ultimately the early diagnosis of acute ischemic cerebral stroke is based on clinical assessment, but modern computed tomography (CT) and magnetic resonance imaging (MRI) are important to exclude intracranial hemorrhage, aid in the diagnosis of ischemic stroke, and aid in triage for treatment of acute major ischemic stroke. There is also emerging literature to support the greater use of imaging in minor stroke and transient ischemic attack (TIA), and imaging is important in the diagnosis of alternate mechanisms of stroke. Moreover, with increasing complexity of care for ischemic stroke patients, radiologists are integral members of the modern multidisciplinary acute stroke team. Thus the focus of this chapter will be to familiarize the radiologist with the evolving role of CT and MRI for acute ischemic stroke.

PATHOPHYSIOLOGY OF STROKE Cerebral infarction occurs when there is insufficient blood flow to perfuse cerebral tissue, most commonly due to arterial occlusion. The vascular compromise that ultimately leads to ischemic injury is

310

dynamic. Factors that influence tissue injury include the rate of onset and duration of ischemia, size of the downstream territory of the occluded blood vessel, the presence and adequacy of collateral circulatory mechanisms, and systemic factors such as body temperature and hyperglycemia. Patients experiencing very short durations of minor ischemia may have no detectable ischemic injury on MRI.19 However, if a central artery is occluded, the large size of the downstream parenchymal territory often results in severe and sustained neurologic deficit. A robust collateral circulatory mechanism may prevent or reduce the extent of ischemic injury. Conversely, an occlusion in a small perforating end-arterial vessel will often result in an infarct even with a short duration of ischemia, owing to the lack of collateral circulatory mechanisms. In addition, elevated body temperature at stroke admission is associated with greater infarct size and higher mortality,164 and persistent poststroke hyperglycemia is independently associated with greater tissue injury and poorer functional outcome.11 Much of the CT and MRI work in stroke pathophysiology has focused on identifying and distinguishing nonviable and potentially viable tissue. The term infarct core refers to nonviable tissue that is irreversibly injured (infarcted) or with such critically reduced tissue perfusion that cellular death is inevitable. Ischemic penumbra refers to tissue that is potentially viable (i.e., with reduced perfusion that results in cellular dysfunction but not inevitable cell death). Thus penumbral tissue is dysfunctional and at risk of infarction but may recover with adequate and timely reperfusion. This is an operationalized concept that may be used to guide ischemic stroke treatment. Specifically, reperfusion treatments could be potentially provided or withheld if there are imaging tests that can accurately distinguish between core and penumbra. This is the “holy grail” of CT and MRI in acute ischemic stroke, and current imaging methods and applications used in major stroke centers will be explored below.

CLINICAL ASSESSMENT OF ACUTE ISCHEMIC STROKE It is important for the interpreting radiologist to be aware of the degree of suspicion for ischemic stroke, common clinical syndromes and localizing features, and the severity and time of onset of neurologic symptoms; these variables may prompt different CT and MRI protocols and aid in interpretation of imaging studies. One of the first questions faced by the stroke team is whether the patient is experiencing an ischemic stroke. Initial assessment requires consideration of all causes of acute neurologic symptoms, including nonischemic stroke, intracranial tumor or infection, or metabolic causes. The potential symptoms of ischemic stroke are varied and include facial weakness or asymmetry, limb weakness or sensory loss

http://radiologyebook.com

CHAPTER 11  Stroke or incoordination, aphasia, ataxia, visual loss, and altered conscious state. Features that increase the likelihood of a stroke include the presence of facial palsy, abnormal visual fields, exact time of onset, ability to lateralize symptoms to the right or left side of the brain, history of hypertension or angina, and age older than 50 years. Conversely, patients with convulsions or history of epilepsy or migraine may have a stroke mimic.58,75,120,199 Often it is useful to consider the pattern of neurologic symptoms and signs to identify whether they may fit a recognized and stereotyped pattern for focal arterial occlusion. This increases the likelihood of a stroke over a stroke mimic.75 Broadly, ischemic stroke can be classified into the anterior circulation or posterior circulation territories. The more common anterior circulation stroke refers to the anterior and middle cerebral artery territories supplied by the internal carotid arteries, whereas posterior circulation refers to territory supplied by the vertebrobasilar system, including the posterior cerebral territory. Differentiation between anterior and posterior circulation stroke may be difficult owing to the potential for considerable overlap in clinical features; however, the presence of ataxia, nausea and vomiting, vertigo, diplopia, nystagmus, oculomotor palsy, visual field defects, or crossed motor or sensory deficits should point to a posterior circulation stroke.187 Similarly, ischemic stroke can be divided into cortical or subcortical stroke. A common pattern for left (dominant) anterior circulation cortical stroke includes aphasia, which may be expressive due to injury to Broca’s area, and/or receptive aphasia due to injury to Wernicke’s area, right face, arm, and leg weakness, and/or sensory loss due to injury to the motor and sensory cortex. The left middle cerebral artery (MCA) territory is more commonly affected than the anterior cerebral artery territory, and thus arm weakness is typically more pronounced than leg weakness. A common pattern for right anterior circulation cortical stroke includes hemineglect due to damage to cortical association areas, and left face, arm, and leg weakness and/or sensory loss due to injury to the motor and sensory cortex. Neurologic symptoms and signs are variable and depend on the size and eloquence of the territory affected; however, in general the presence of aphasia and/or neglect suggests cortical involvement. Subcortical ischemic stroke often produces pure motor or sensory symptoms due to injury to the hemispheric white matter, corona radiata, internal capsule, or basal ganglia from occlusion of lenticulostriate or thalamoperforate vessels. The use of the terms lacune or lacunar infarct for all subcortical infarcts can be problematic. Historically, lacunar infarction was thought to be almost exclusively associated with small vessel lipohyalinosis and thrombosis related to cardiovascular risk factors. However, modern imaging has proved this to be incorrect—subcortical or “lacunar” infarcts may be caused by an embolic mechanism.9 In fact, complete or partial occlusion of the MCA is not uncommon in patients with large subcortical infarcts,110 concordant with an embolic mechanism. Additional important factors are the severity of neurologic deficit and the time of symptom onset. The severity of neurologic deficit is usually communicated in terms of a standardized acute stroke severity scoring system, such as the National Institutes of Health Stroke Scale (NIHSS).21,65 The NIHSS is one of the most frequently used tools to rapidly and objectively quantify the neurologic impairment caused by acute stroke. A higher NIHSS score indicates greater severity of neurologic deficit and provides information that affects treatment decisions and patient outcome. However, practitioners must be aware that the NIHSS is weighted to dominant hemisphere stroke and may be less sensitive in posterior circulation stroke.135 Time of onset is a critical variable in the clinical assessment of ischemic stroke. Patients presenting to hospital early after the onset of

311

symptoms may be eligible for reperfusion therapies that can improve long-term functional outcome. In the United States, approximately 30% of patients arrive within 4.5 hours of symptom onset, within the time window for use of ivtPA reperfusion therapy.192 However, a significant number of patients have an unclear symptom onset time. In these situations, it is important to differentiate between the time when the patient was “last seen well” and when the patient was found with symptoms. The time difference could render patients eligible or ineligible for reperfusion therapies. For patients with unclear time of onset, and in particular in patients who wake up with stroke symptoms, there is increasing interest in whether advanced CT and MRI can provide a “tissue-clock” that could potentially increase the number of patients eligible for reperfusion therapies.123,188

TREATMENT OF ACUTE ISCHEMIC STROKE Historically the treatment of ischemic stroke has primarily been supportive therapy, rehabilitation, and secondary prevention. This includes consideration of platelet inhibition, anticoagulation, and modification of cardiovascular risk factors such as smoking, diabetes, hypertension, and cholesterol. Acute reperfusion therapy with intravenous thrombolysis has been extensively investigated over the past 2 decades and is now widely used and accepted. The benefit of intravenous thrombolysis for acute stroke was first adequately established in the 1995 National Institute of Neurological Disorders and Stroke (NINDS) tPA trial. As compared with patients receiving a placebo, those treated with tPA were more likely to have no symptoms or no significant disability (modified Rankin Scale [mRS] score 0-1) by 3 months, despite an increase in the rate of symptomatic intracerebral hemorrhage.144 However, patients were excluded if they had had serious head trauma or stroke within the preceding 3 months; major surgery within 14 days; history of previous intracranial hemorrhage; systolic blood pressure above 185 mm Hg; rapidly improving/minor symptoms; gastrointestinal or urinary tract hemorrhage within the previous 3 weeks; arterial puncture at a noncompressible site within the previous week; seizure at the onset of stroke; taking anticoagulants, received heparin, and had an elevated partial thromboplastin time, abnormally low platelet count, or severe hypoglycemia or hyperglycemia.144 Subsequent trials have confirmed the benefit of the NINDS finding and have also investigated extending the time window of treatment. Pooled results of 5 major trials suggested benefit of intravenous thrombolysis in the 3- to 4.5-hour time window after symptom onset.71 The third European Cooperative Acute Stroke Study (ECASS III) was specifically designed to investigate this suggested benefit. The trial confirmed benefit of therapy in this time frame but with additional exclusions for patient selection: patients older than 80 years, use of oral anticoagulation therapy regardless of INR, baseline NIHSS score greater than 25, and history of a combination of previous stroke and diabetes mellitus.72 Other trials are investigating the benefit of further extension to the time window of intravenous thrombolysis based on the use of advanced MRI.123 Although the benefit of thrombolysis has been demonstrated up to 4.5 hours after symptom onset, greater benefit is observed with earlier treatment. Patients are less likely to die or experience symptomatic intracerebral hemorrhage, and more likely to be ambulant at discharge and to be discharged home by reducing treatment delay170; each 15 minutes gained translates to an extra month of disability-free life.133 Since the positive Prolyse in Acute Cerebral Thromboembolism II trial in 1999, there has been increasing interest in the role of endovascular therapy for ischemic stroke.61 For patients with MCA occlusion and less than 6 hours since symptom onset, patients treated with local intraarterial fibrinolysis with prourokinase and heparin infusion were more likely to have no or slight disability (mRS 0-2) as compared with

http://radiologyebook.com

312

PART II  CT and MR Imaging of the Whole Body

patients receiving the heparin infusion alone, in spite of an increase in the rate of symptomatic intracerebral hemorrhage.61 However, exclusions were similar to those applied in the NINDS trial. Since this trial, multiple intraarterial therapies have been considered. These include local intraarterial fibrinolysis or thrombolysis, mechanical thromboaspiration, mechanical thrombectomy, and intracranial stenting.131 The evidence for endovascular therapy is complex to assess because of the multiplicity of treatment options, ongoing technologic advances, and ability to combine intraarterial and intravenous therapies. In 2013, three randomized controlled trials of endovascular treatment had neutral results, contributing to uncertainty regarding the efficacy of endovascular therapy.20,36,99 However, concerns were raised regarding the relatively long delays to endovascular treatment, the limited use of pretreatment CT or MR angiography to confirm proximal vessel occlusions, and the limited use of more recent-generation mechanical thrombectomy devices such as retrievable stents. More recent publications in 2015 have found benefit to endovascular therapy compared to intravenous thrombolysis alone.16,28,27,66 For patients with a proximal anterior circulation occlusion treated rapidly (within 6 hours of symptom onset), endovascular therapy achieves higher rates of early neurologic recovery and functional independence without significant differences in symptomatic hemorrhage.16,27,66 The respective roles of various intravenous and endovascular reperfusion therapies continues to evolve, with continuing investigation of patient selection for treatment, novel intravenous thrombolytics, novel methods of endovascular reperfusion, and refining measures of treatment effectiveness.

IMAGING GOALS OF ACUTE ISCHEMIC STROKE With the advent of both intravenous and endovascular reperfusion therapies for stroke, the radiologist has become an integral multidisciplinary member of the team evaluating patients with ischemic stroke arriving within the treatment window, and has a clear role in the stroke “chain of survival.”87 The overall goal is to provide rapid access to brain imaging and expert interpretation, which is critical for the diagnosis and triage of patients with acute severe ischemic stroke. Quality metrics proposed for stroke teams include a door-to-CT time of less than 25 minutes and door-to-CT interpretation time of less than 45 minutes.87 Time metrics have been proposed for endovascular treatment, and radiology departments will need to expedite multidisciplinary workflow to meet the demands of modern stroke treatment.130 Clinical assessment and noncontrast CT (NCCT) remains sufficient to satisfy the inclusion criteria for the NINDS trial, provided there is no intracranial hemorrhage.144 The impact of ischemic changes on patient outcome after ivtPA treatment was investigated in a post hoc analysis of ECASS I trial data. This revealed no treatment benefit in patients with over one third MCA territory infarction, and is the source of the commonly used exclusion criteria for ivtPA treatment used in the subsequent ECASS trials.197 However, since the original ECASS and NINDS trials were published in 1995, there has been increasing interest in further refining acute stroke imaging and treatment triage. Specifically, apart from exclusion of intracranial hemorrhage and ensuring there is not a large established infarct, there is interest in confirming the diagnosis of ischemic stroke over a stroke mimic to potentially avoid treatment risk from reperfusion therapy, identifying the presence/absence of a proximal vessel occlusion that may affect choice of reperfusion therapy and prognosis, identifying the brain parenchymal territory involved, identifying how much infarct core and penumbra are present, and whether an etiology (e.g., embolus from carotid stenosis/dissection) can be identified, and determining how these factors affect prognosis, particularly in the acute setting. It must be considered that acquiring this extra information may incur

treatment delay, and that institutional efforts should be directed to a rapid, consistent, and reliable approach to provide timely and rapid access to reperfusion therapies. Both CT and MRI provide the platform for rapid multimodal imaging; often an institution will choose to adopt either CT or MRI as its primary strategy (Box 11-1). The utility of both CT and MRI to answer the imaging goals of ischemic stroke is explored below.

Exclude Intracranial Hemorrhage Any acute intracranial hemorrhage (ICH) is a contraindication for acute reperfusion therapy (Fig. 11-1). This may be intracerebral hemorrhage, subarachnoid hemorrhage, and subdural or extradural hematoma. Although NCCT is used as the standard-of-care test to exclude

BOX 11-1  Monash Health Acute Stroke

Imaging Protocol

Noncontrast CT, brain (skull base to vertex) • 5-mm axial, coronal, and sagittal reformats • 2.5-mm axial, coronal, and sagittal reformats CT angiogram (aortic arch to vertex) • 0.7-mm axial acquisition • 15/2-mm maximum intensity projection (MIP) axial, coronal, and sagittal reformats of brain (skull base to vertex) • 10/2-mm MIP coronal and sagittal reformats of neck CT perfusion (8-cm field of view) • MTT, TTP, CBV, and CBF parametric color maps Postcontrast CT, brain • 5-mm axial, coronal, and sagittal reformats

FIG 11-1  Intracranial hemorrhage on CT in a 50-year-old woman who presented with headache and left hemiplegia; CT was performed at 90 minutes post onset. Axial NCCT reveals large right intracerebral hemorrhage (circled area).

http://radiologyebook.com

CHAPTER 11  Stroke intracranial hemorrhage, apart from a small number of patients with autopsy correlation,86 there are no level I data to definitively evaluate the sensitivity of NCCT for ICH detection.112 Regardless, CT has been used in multiple stroke trials for detection of ICH. The appearance of ICH on CT depends on the time elapsed since the hemorrhage. In general there is a linear relationship between CT attenuation and hemorrhage hemoglobin protein content.182 In the hyperacute phase, the ICH consists of red and white blood cells, platelets, and serum that confer an attenuation of approximately 30 to 60 Hounsfield units (HU).182 Over the next few hours, clot retraction occurs with increasing attenuation, generally to 60 to 80 HU, but potentially up to 100 HU.182 The overall CT appearance is dependent on the plasma hemoglobin content,128 patient coagulation mechanisms,156 and slice thickness of the CT imaging technique. As further time elapses, the attenuation of the ICH on NCCT reduces, and by 3 weeks is generally less than the surrounding brain tissue.15 MRI is considered equivalent to CT for detection of intracranial hemorrhage.98 The appearance of ICH on MRI is more complex, and the signal characteristics depend on the state of the iron atoms within the hemoglobin of the ICH and the red cell membranes. Intraparenchymal hemorrhage evolves through five distinct stages depending on time from onset. This can be divided into the hyperacute (first few hours), acute (hours to days), early subacute (days to weeks), late subacute (weeks to months), and chronic phases (months). Typical findings are detailed in Box 11-2 and seen in Figure 11-2. Acute subarachnoid hemorrhage and intraventricular hemorrhage are best detected with a combination of fluid-attenuated inversion recovery (FLAIR) or double inversion recovery (DIR) MR images, and susceptibility weighted imaging (SWI).196,79

Is It an Ischemic Stroke? The presence of radiologic changes on imaging tests that are concordant with the clinical symptoms and signs is helpful in the diagnosis of a true ischemic stroke and excluding a stroke mimic. Ultimately, however, the diagnosis of ischemic stroke is clinical and not based on any sole radiologic test. Moreover, even with sensitive MRI techniques, imaging can be negative in cases of true ischemic stroke, particularly if symptoms are transient or clinically localized to the brainstem or subcortical territory.186 NCCT findings of parenchymal hypoattenuation or swelling in a topographic distribution concordant with the clinical symptoms can

BOX 11-2  MRI Signal of

Parenchymal Blood

Hyperacute (oxyhemoglobin) • T1 isointense • T2 isointense to hyperintense Acute (deoxyhemoglobin) • T1 isointense • T2 hypointense Early subacute (intracellular methemoglobin) • T1 hyperintense at periphery • T2 hypointense Late subacute (extracellular methemoglobin) • T1 hyperintense • T2 hyperintense Chronic (hemosiderin) • T1 hypointense • T2 hypointense

313

confirm the clinical diagnosis of stroke (Fig. 11-3). However, NCCT findings are time dependent and may be normal or near normal even in the setting of established or large infarct, particularly within the first 3 hours of stroke onset121 (Fig. 11-4). Additional NCCT detection of a hyperdense thrombosed intracranial vessel can also confirm the stroke diagnosis.12 However, even using modern CT scanner technology for patients who may be eligible for reperfusion therapies, the diagnostic sensitivity for NCCT is approximately 50%.30 The addition of CT angiography (CTA) for detection of vessel occlusion increases sensitivity by another 10% to 15%. In addition, if the typical topographic distribution of infarction is discernable on the source images of the CTA, this too confirms the diagnosis of ischemic stroke rather than stroke mimic.174 Overall, a multimodal imaging incorporating CT perfusion (CTP) provides the highest CT-based imaging sensitivity for confirming diagnosis of ischemic stroke30,51 (Fig. 11-5). Postcontrast CT may be helpful to aid rapid recognition of enhancing tumors that may mimic stroke by presenting with postictal neurologic deficit (Fig. 11-6). Nonetheless, CT-based imaging strategies remain limited in their sensitivity for detection of subcortical and brainstem stroke. MRI findings of parenchymal FLAIR and T2 hyperintensity and swelling in a topographic distribution concordant with the clinical symptoms also confirm the clinical diagnosis of stroke (see Fig. 11-3). However, similar to the parenchymal findings on CT, FLAIR may be negative, particularly within the first 3 hours of stroke onset, and the infarct only detectible on diffusion-weighted imaging (DWI).189 Thus it has been proposed that a rapid stroke imaging protocol using DWI scans alone may be of benefit in confirming the diagnosis of ischemic stroke and excluding stroke mimics.115 However, there remain issues with detection of small subcortical and brainstem strokes, particularly when DWI is performed within the first few hours of stroke onset.149 The use of MR angiography (MRA) and MR perfusion techniques may also increase diagnostic confidence for ischemic stroke if vessel occlusions or typical ischemic perfusion patterns are discernible. Moreover, by performing further MR sequences, an alternate cause for symptoms may be discernible (Fig. 11-7).

Is There an Intracranial Vessel Occlusion? Detection of a hyperdense thrombosed intracranial vessel on NCCT can provide confirmation of ischemic stroke diagnosis before parenchymal changes are discernible. Common locations include the horizontal or M1 segment of the MCA, the sylvian or M2 segment of   the MCA, the terminal internal carotid artery, and the basilar artery. Identification of the hyperdense vessel on NCCT provides information about thrombus burden and has prognostic and therapeutic information. In particular, patients with a hyperdense vessel have greater thrombus burden,101 greater likelihood of early ischemic parenchymal changes, more severe neurologic presentation, and less likelihood of early neurologic recovery even after intravenous thrombolysis.119,125 Importantly, use of thin sections (≤2.5 mm) allows greater detection of hyperdense vessels165 (Fig. 11-8). CTA is highly accurate for detection of central large vessel occlusions, with a  “flow-gap” or vessel “cutoff ” being diagnostic for a vessel occlusion.118 However, CTA only provides a snapshot in time and thus may not allow sufficient time for collateral flow to reach the distal end of the thrombus. This can result in overestimation of thrombus length on CTA,139 and the use of NCCT may be more accurate206 (Fig. 11-9). On MRI, vessel occlusion may result in altered arterial T1, T2, or FLAIR signal, or T2* imaging, but greater sensitivity is achieved with SWI80 (Fig. 11-10). Similar to CTA, MRA can be used for detection of vessel occlusion; however, MRA is prone to overestimation of vessel stenosis78 and typically limited to the more central vasculature compared to CTA.

http://radiologyebook.com

314

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 11-2  Evolution of intracranial hemorrhage on MRI. A, MRI day 1 post intracerebral hemorrhage; B, 4 weeks post hemorrhage; and C, 4 months post hemorrhage. Axial T1 (top row), axial T2 (middle row), and axial SWI (bottom row) reveal typical signal and morphologic changes with evolution to a hemosiderin-lined slitlike cavity by 4 months.

http://radiologyebook.com

CHAPTER 11  Stroke

A

B

C

D FIG 11-3  Features of acute infarction on CT and MRI in an 82-year-old woman who presented with speech disturbance with unclear time of onset. A, Axial NCCT reveals parenchymal hypoattenuation and loss of gray-white differentiation in a vascular topographic distribution, concordant with the clinical symptoms of established stroke. MRI FLAIR (B), DWI (C), and ADC maps (D) are concordant with CT findings demonstrating parenchymal swelling, FLAIR hyperintensity, and diffusion restriction.

http://radiologyebook.com

315

316

A

PART II  CT and MR Imaging of the Whole Body

C

B

FIG 11-4  Large infarcts may be difficult to discern on NCCT. This 39-year-old woman presented with a right MCA syndrome with hemiplegia and neglect. A, Axial NCCT at 1 hour 40 minutes displayed at standard window width and level; the infarct is just appreciable. B, Axial MRI DWI reveals faint DWI changes in the right MCA territory. C, Axial MRI ADC map confirms diffusion restriction and large established right MCA infarct. CT acquisition to MRI acquisition time was 40 minutes.

A

C

B

FIG 11-5  Utility of CTP for detection of small distal territory strokes. This 59-year-old man presented with mild dysarthria, NIHSS 1. A, No definitive stroke is evident on NCCT at 2 hours post onset. B, Concurrent CTP reveals perfusion abnormality in the left frontal lobe, confirming the diagnosis of stroke. C, Subsequent MRI DWI confirms infarct at corresponding location as CTP abnormality.

http://radiologyebook.com

CHAPTER 11  Stroke

A

317

B FIG 11-6  Utility of contrast-enhanced CT in a 55-year-old woman who presented with speech disturbance, NIHSS 8. CT was performed 3.5 hours after onset. A, Axial NCCT reveals faint hyperattenuating area (oval) without intracerebral hemorrhage. B, Axial contrast-enhanced CT clearly shows abnormal parenchymal enhancement. Final diagnosis was seizure activity from glioblastoma.

How Big Is the Infarct Core? The term infarct core refers to nonviable tissue that is irreversibly injured or with such critically reduced tissue perfusion that cellular death is inevitable. DWI is considered the gold standard for established and irreversible infarct,172 because DWI reversal is considered uncommon.29 The DWI sequence is sensitive to disruption of the normal random movement (diffusion) of water molecules. Proposed mechanisms for reduced water diffusivity in cerebral infarction include cell wall dysfunction with associated intracellular/extracellular water shift, increased intracellular viscosity, and increase in extracellular space tortuosity.60 DWI is obtained by using gradient pulses incorporated into a T2-weighted sequence. Both T2 and diffusion contribute to image contrast, with the magnitude of the diffusion weighting known as the b value. The b value is dependent on gradient duration, amplitude, and interval, with higher b values reflecting greater diffusion weighting. At high b values, diffusion properties are the main determinant of image contrast; however, the T2 component may remain visible, potentially leading to an effect known as T2 shine-through. The DWI image should be interpreted with the apparent diffusion coefficient (ADC) map, a postprocessed image where tissue contrast is dependent only on diffusion. Restricted diffusion, confirming infarct core, is characterized by increased DWI signal with typical corresponding decrease in ADC signal. In contrast, T2 shine-through is characterized by increased DWI signal with corresponding increase in signal on the ADC map. Changes of restricted diffusion manifest within minutes after acute infarction and peak between 24 and 72 hours. However, occasionally, early DWI may result in only faint or minimal changes on DWI,114 and thus both DWI and ADC maps should be carefully scrutinized in cases of hyperacute stroke (see Fig. 11-4). Although some institutions use MRI as their primary strokeimaging strategy because of the increased sensitivity of DWI over CT

for established infarct, other institutions have been searching for a reliable and valid alternative using CT, given the rapidity and greater availability of CT imaging techniques. NCCT signs of established infarct are parenchymal hypoattenuation and swelling. Specifically in the MCA territory, loss of gray-white attenuation of the insular ribbon and hypoattenuation in the lentiform nuclei are considered early signs of MCA infarction.141 Adjusting the window width and level of the CT image increases detection of early ischemic changes117 (Fig. 11-11). More standardized methods of assessment for early ischemic change is achieved by using standardized reporting templates, such as the Alberta Stroke Program Early CT Score (ASPECTS).155 ASPECTS was proposed as a more accurate and practical alternate method for clinicians and radiologists to assess the extent of parenchymal injury. Moreover, since an increase in the rates of death and dependency occurred with patients treated with intravenous thrombolysis if pretreatment ASPECTS was 7 or less, it has been suggested as an alternate to the “1/3 MCA rule.”155 However, there is only moderate interrater reliability for using ASPECTS in such a dichotomized fashion, which may have an impact on utility.69 As an alternate to NCCT, the source image from CTA provides additional information for the detection of ischemic change. Although initially considered to be as reliable as DWI in assessing the size of the infarct core,174 CTA source images on faster modern CT scanners commonly overestimate infarct core size.161 CTP techniques show increasing promise in the accurate detection of infarct core. CTP data are obtained by monitoring the passage of a bolus of iodinated contrast media through the cerebral parenchyma via serial CT scans. Complex mathematical equations are used to calculate the average time for the contrast bolus to pass through each part of the imaged brain parenchyma (mean transit time [MTT]), the time to peak enhancement (time to peak [TTP]), the volume of blood within each part of the imaged brain parenchyma (cerebral blood volume [CBV]), and the volume of blood passing through each part

http://radiologyebook.com

318

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 11-7  Hemiplegic migraine mimicked acute stroke in a 5-year-old boy who presented with headache, right-sided hemiparesis, and altered conscious state. Axial T2 MRI (A) and DWI (B) do not reveal any stroke, nor intracranial vessel occlusion on (C) MR angiography. D, There is asymmetric cortical venous signal on SWI, a phenomenon reported in pediatric hemiplegic migraine. The patient made a full recovery.

of the brain parenchyma over time (cerebral blood flow [CBF]). Each CT image is usually obtained in the axial plane, typically every 1 to 1.5 seconds for approximately 60 seconds. Interpretation is typically by visual inspection of color-coded parametric maps displaying MTT, TTP, CBV, and CBF data. Modern multislice scanners are able to assess the whole brain in a single examination, with the potential to display perfusion data in multiple planes (Fig. 11-12). There is debate on which is the best CTP parameter to delineate infarct core. Early data suggested that CBV provided the most accurate measure of core volume.202 More recent data suggest that relative CBF may be the most robust parameter.25 Importantly, there is significant

variation in perfusion maps depending on the CT scanner and postprocessing software used, and thus each local institution should gain their own experience using their own postprocessing software.107 Current research trials are investigating the feasibility of automated software to improve reproducibility and provide standardization,31 and these may be available for routine clinical use in the near future.

How Big Is the Penumbra? The term penumbra refers to tissue that is potentially viable (i.e., with reduced perfusion that results in cellular dysfunction but not inevitable cell death). Thus penumbral tissue is at risk of infarction but may

http://radiologyebook.com

CHAPTER 11  Stroke

B

A

FIG 11-8  Use of thin-section NCCT increases intravascular thrombus detection. This 85-year-old woman presented with right MCA syndrome; image shows NCCT at 90 minutes post onset. A, Axial NCCT at 5-mm-thick reconstruction fails to convincingly demonstrate right MCA intravascular thrombus (arrow). B, Axial NCCT at 2.5-mm-thick reconstruction shows the hyperdense right MCA (arrow).

5.8mm

8.2mm

20.1mm

A

C

B

FIG 11-9  Clot length measurement with CT demonstrating overestimation of clot length that can occur with CTA. This 47-year-old woman presented with rapidly improving right MCA symptoms; CT was performed 2 hours post onset. A, NCCT reconstructed in oblique plane measures clot in the right MCA branch at 8 mm. B, CTA measures segment of nonopacification at 20 mm, overestimating true clot length. C, Delayed postcontrast CT provides additional time for the collateral circulation to reach the distal end of the clot and measures segment of nonopacification at 6 mm.

http://radiologyebook.com

319

320

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 11-10  Utility of SWI for clot imaging in a 52-year-old man who presented with ataxia. Axial T2 MRI (A) and DWI (B) reveal an acute left lateral medullary infarct and a tiny cerebellar infarct (arrows). The abnormal signal of the thrombus (circled) on the axial SWI (C) is easier to discern than the lack of flow signal on the MR angiogram (D).

recover with adequate and timely reperfusion. Importantly the distinction of infarct core and penumbra were not used in the NINDS or ECASS trials, and thus intravenous thrombolysis is generally administered after initial NCCT imaging at most centers. Within 3 hours of stroke onset, penumbral information may be evident on NCCT. It has been postulated that areas of brain parenchyma that are isoattenuating and slightly swollen may represent penumbral tissue rather than established infarct.142,153 However, the majority of ischemic penumbra appears normal on NCCT. CT and MR perfusion MTT or TTP maps are most commonly used in comparison to the CBF or CBV maps to estimate the size of the penumbra. In addition TMax maps may also be obtained based on further mathematical equations. MTT and TTP abnormalities may be discernible immediately after symptom onset, making CTP a valuable tool to confirm the suspicion of ischemic stroke. Large cortical perfusion abnormalities, such as those that occur with proximal vessel occlusion, are easier to detect than smaller or subcortical perfusion abnormalities. MTT and TTP abnormalities are difficult to detect in the posterior fossa, because this area is prone to artifact on CT and may not be adequately included during CTP. Importantly, for clinical use it must

be noted that there is variation in the appearance of the perfusion maps depending on the CT scanner and postprocessing software used.57,90 For ischemic stroke trial use, more reproducible information is required, and thus various CTP parameters have been validated against MRI, with CT and MR TMax greater than 6 seconds considered reliable in the assessment of penumbra.26 The great interest in core and penumbral assessment is in the advancement of the mismatch hypothesis. The mismatch hypothesis is based on identification of the infarct core and the penumbra, with the volumetric difference representing the mismatch (i.e., tissue at risk of infarction without timely reperfusion) (Fig. 11-13). Restoration of perfusion to the penumbral tissue results in clinical recovery,183 and thus patients with mismatch could be selected for reperfusion therapies. This promising concept is being investigated in a number of trials, particularly those outside traditional treatment windows (wake-up strokes, >4.5 hours since presentation),123 because mismatch may persist for at least 48 hours in selected cases.124 The mismatch concept could facilitate treatment of a larger number of stroke patients, as the majority of stroke patients present to hospital outside the traditional intravenous thrombolysis treatment window of 4.5 hours.192

http://radiologyebook.com

CHAPTER 11  Stroke

B

A

FIG 11-11  Utility of variable CT window width and level. This 39-year-old woman presented with a right MCA syndrome with hemiplegia and neglect. Axial NCCT at 1 hour 40 minutes displayed at (A) standard window width and level (W:75 L:35) and (B) varied window width and level (W:35 L:35). The large right MCA infarct (oval) is easier to discern using the varied “Stroke” window width and level.

A

B FIG 11-12  Modern CT scanners may be able to assess perfusion of the whole brain and display CTP data as a brain volume reconstruction. This 80-year-old man presented with left hemisensory changes, NIHSS 2. A, Superior view of whole brain volume CTP demonstrating elevated MTT in the right MCA territory. B, Oblique whole brain volume reconstruction.

http://radiologyebook.com

321

322

PART II  CT and MR Imaging of the Whole Body

A

B

MTT (sec)

CBV (ml/100g)

FIG 11-13  Mismatch on CTP in a 24-year-old man who presented with left MCA syndrome, NIHSS 11. CT perfusion at 4.5 hours post initial TIA and 3 hours post recurrence of hemiparesis. A, Axial MTT map reveals large area of MTT abnormality. B, Near-normal CBV in the corresponding territory, indicating perfusion mismatch.

Is There an Etiology Evident? There are a number of risk factors for ischemic stroke, such as age, male sex, hypertension, diabetes, hypercholesterolemia, smoking,  and family history of cerebrovascular disease.76,147 However, the exact stroke trigger that initiated the ischemic event often remains  unclear. Over 20 years ago, the Trial of Org 10172 in Acute Stroke Treatment (TOAST) investigators created a clinical classification system of inferred stroke etiology that has been widely adopted in stroke registries and trials.1 The TOAST classification recognizes 5 etiologies: (1) large-artery atherosclerosis, (2) cardioembolism, (3) small-artery occlusion, (4) stroke of other determined etiology,  and (5) stroke of undetermined etiology. Diagnosis is based on  clinical features, laboratory tests, and imaging of the brain, heart,  and cervicocranial vessels. For a large-artery atherosclerosis mechanism, more than 50% stenosis of the relevant intra- or extracranial vessel is required, as is exclusion of a cardiac embolic cause. However, over the last 20 years, there have been advances in imaging, and  in particular, use of DWI has become routine in the evaluation of ischemic stroke. This has shed further light on the vascular mechanisms of stroke and facilitates more accurate, consistent, and evidencebased classification of the cause of ischemic stroke in the individual patient.7,104 The use of modern CT and CTA or MR and MRA techniques can help determine ischemic stroke etiology. A large-artery mechanism is most likely in the setting of high-grade stenoocclusive disease with an infarct in the relevant vascular territory and the absence of acute infarction in other vascular territories. However, while the TOAST classification would consider greater than 50% stenosis of the relevant vessel for a large-artery atherosclerosis mechanism, complex plaque with lesser degrees of stenosis may also provide an embolic source. Thus the degree of stenosis is best used to determine the need for carotid endarterectomy rather than necessarily determine the stroke etiology.

MRI with DWI provides the greatest sensitivity for detection of small acute infarcts. Similarly, detection of multiple acute infarcts closely related in time in both the right or left anterior and/or posterior circulations makes a central embolic cause most likely. In certain cases, the intramural hemorrhage from arterial dissection and secondary occlusion or embolic infarction may be evident on CT or MRI (Fig. 11-14). In particular, cervicocranial arterial dissection should be considered in patients who do not have typical stroke risk factors, when there is a history of neck trauma or manipulation, connective tissue disorders, or where there are supporting clinical features such as neck pain.

Prognostication Although multiple factors such as age and severity of neurologic deficit affect the clinical outcome of a patient with ischemic stroke, there are imaging factors that also provide early prognostic information. The size of the infarct core prior to acute reperfusion therapy is a predictor of clinical outcome and response to therapy and can be assessed using CT and MRI techniques as described previously. In particular, patients with NCCT ASPECTS less than 7 and DWI volumes of more than 70 to 100 cc4,67,77,207 are more likely to have poor functional outcome regardless of reperfusion and may incur harm with acute reperfusion therapies.176 In addition to the core volume, infarct location in more eloquent areas—in particular the internal capsule or corona radiata—adversely affects outcome.33 The presence of proximal intracranial vessel occlusion is another strong predictor of clinical outcome,181 with greater thrombus burden leading to poorer clinical outcome.160 Moreover, greater thrombus burden may be less responsive to intravenous reperfusion therapies and require alternate endovascular approaches.16,27,66,166

Complications of Acute Stroke Treatment The major concern after acute reperfusion therapy for ischemic stroke is intracranial hemorrhage. In the NINDS trial, as assessed by CT scan

http://radiologyebook.com

CHAPTER 11  Stroke

A

B

323

C

FIG 11-14  Utility of fat-suppressed T1 MRI for diagnosis of arterial dissection. This 46-year-old man presented with headache, neck pain, and Horner’s syndrome. A, Contrast-enhanced MR angiogram reveals severe narrowing (arrow) of the distal cervical ICA. Sagittal (B) and coronal (C) reconstructions of 3D volumetric T1 fat-suppressed MRI clearly show the intramural hematoma (arrows).

performed at 24 hours or with clinical suspicion for hemorrhage before 36 hours, asymptomatic hemorrhage occurred in 11% of patients treated with intravenous thrombolysis; 6% overall were symptomatic.144 In the ECASS III trial, as assessed by CT or MRI performed before 36 hours, 27% of treated patients had any intracranial hemorrhage; overall 2% were symptomatic.72 The rates of any intracranial hemorrhage after endovascular reperfusion approximates 20% to 30%, with rates of symptomatic hemorrhage ranging between 2% and 7%.16,27,66,145,154 However, multiple variables have an impact on the likelihood of symptomatic hemorrhage, including age, severity of neurologic deficit, infarct size, and serum glucose levels.129,176 To accurately characterize complications, it is important to distinguish between hemorrhagic infarction and parenchymal hematoma formation. For accuracy of interpretation it is useful to consider the definitions used in stroke trials, and in particular the ECASS trials. This defined hemorrhagic events on CT as hemorrhagic infarction (HI) types I and II and parenchymal hematoma (PH) types I and II. HI I was defined as small petechiae along the infarct margins of the infarct; HI II, confluent petechiae within the infarct but without mass effect; PH I as focal blood clot in less than 30% of the infarct area, with mild mass effect; and PH II as focal blood clot greater than 30% of the infarct volume, with significant mass effect. PH II was a significant predictor of poor clinical outcome; other types may have less impact on clinical outcome.56 Of note, these hemorrhagic events were typically assessed with CT, and the advent of new MRI techniques such as SWI can increase the detection of intracranial hemorrhage (Fig. 11-15). This makes it even more important to distinguish between hemorrhagic infarct and parenchymal hematoma, because HI is common after ischemic stroke and particularly after reperfusion therapies. In addition, little is known about the normal temporal evolution of ischemic stroke on SWI, which may include development of minor HI that is not discernible on CT or conventional MRI.32 This may more closely reflect the higher incidence of HI evident on autopsy studies as compared to CT studies.211 Hemorrhagic events should also be distinguished from the common parenchymal or subarachnoid contrast staining that can mimic hemorrhage after endovascular stroke therapy.122,208 Although this may be distinguishable with delayed interval CT or MRI, the dual energy CT techniques provide a useful alternative for immediate distinction on

posttreatment CT scans.157 Endovascular stroke therapy may also lead to procedural embolic complications and infarcts in remote territories. Although this occurs in fewer than 15% of cases,16,27,154 it may adversely affect clinical outcomes.191 Thus interpretation of posttreatment imaging should assess for the presence of hemorrhagic complications, contrast staining, and new embolic infarcts.

IMAGING OF ACUTE SUBCORTICAL/PERFORATOR ARTERY TERRITORY STROKE Subcortical ischemic stroke occurs in the territories of the deep or superficial perforating vessels. Deep perforating arteries include medial and lateral lenticulostriate arteries arising from the proximal anterior and middle cerebral arteries, branches of the anterior choroidal artery, and thalamoperforator arteries arising from the posterior communicating and posterior cerebral arteries. These supply the deep gray nuclei, internal capsule, and a variable volume of surrounding white matter. Superficial perforating arteries are the terminal branches of cortical arteries, extending from the cortex toward the lateral ventricles, supplying subcortical and deep white matter. The clinical diagnosis of a subcortical/perforator territory stroke may be suspected by the absence of cortical symptoms and pure sensory or motor symptoms. However, this is not specific; small cortical infarcts can mimic the clinical symptoms of a subcortical infarct.163 Small hemorrhages in an appropriate territory can produce similar findings.85,138 Moreover, the etiology may be single perforator involvement or multiple perforator involvement, such as that evident with a striatocapsular stroke involving the caudate, putamen, and anterior limb of the internal capsule, which is most commonly caused by embolic occlusion of the MCA.43 These patients do derive benefit from the use of intravenous thrombolysis,150 and the use of CT and MRI for assessment of the brain parenchyma and vasculature helps confirm the diagnosis and infarct territory and assess for the etiology. NCCT may be normal early in acute subcortical/perforator territory stroke and may become evident as a focal area of hypoattenuation. However, this may be indistinguishable from more established infarcts or dilated perivascular (Virchow-Robin) spaces, and thus MRI with DWI is the examination of choice for radiologic confirmation148 (Fig. 11-16). Established perforator territory infarcts may have central 

http://radiologyebook.com

324

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 11-15  Hemorrhagic infarction on MRI in a 69-year-old man who presented with right hemiparesis outside the time frame for treatment; MRI was performed at 9 days post onset. Axial T1 (A), T2 (B), FLAIR (C), and SWI (D) reveal large left subacute MCA territory infarct, with cortical blood products easiest to identify on the SWI sequences.

http://radiologyebook.com

CHAPTER 11  Stroke

A

325

B FIG 11-16  Acute anterior choroidal territory infarct in a 64-year-old woman who presented with right-sided ataxic hemiparesis; MRI was performed 2 days after onset. A, Axial DWI MRI reveals focal hyperintensity in the posterior limb of the internal capsule, typical for anterior choroidal territory infarction. B, Axial FLAIR MRI reveals only faint hyperintensity.

cavitation and a typical rim of elevated FLAIR signal, whereas perivascular spaces typically follow cerebrospinal fluid (CSF) signal on all sequences, without a rim of elevated FLAIR signal. Vascular imaging with CTA or MRA should also be considered because occasionally patients with a clinical subcortical/perforator territory stroke and normal CT scan may harbor a large vessel occlusion.163 The spatial resolution of current clinical CTA and MRA is not sufficient to accurately resolve the perforating vessels and reliably assess for perforator vessel occlusion. There is increasing interest in examining the role of CTP techniques in the diagnosis of subcortical/perforator territory stroke, but there remain issues with sensitivity owing to small lesion size.59,127

IMAGING OF ACUTE BRAINSTEM AND CEREBELLAR STROKE Brainstem and cerebellar ischemic strokes occur in the vertebrobasilar territory; they account for 20% of ischemic strokes.18,194 The paired vertebral arteries fuse to form the basilar artery; the basilar artery terminates by branching into the paired posterior cerebral arteries. Along the course of the vertebral and basilar arteries, small perforating vessels supply the brainstem, and larger cerebellar vessels arise to supply the superior (superior cerebellar artery [SCA]), petrosal (anterior inferior cerebellar artery [AICA]), and inferior (posterior inferior cerebellar artery [PICA]) surfaces of the cerebellum. The clinical diagnosis of a brainstem or cerebellar infarct may be suspected by the presence of ataxia, nausea and vomiting, vertigo, diplopia, nystagmus, oculomotor palsy, or crossed motor or sensory deficits.187 About half of vertebrobasilar strokes occur in the brainstem, with the majority in the pons.18 Ventral pontine infarcts may involve cranial nerve nuclei, the reticular activating system, and the corticospinal, corticobulbar, and corticopontine tracts and thus can produce

severe clinical deficits10,23; these are typically associated with basilar stenoocclusive disease. Medullary infarcts typically involve the lateral medullary territory and classically produce Wallenberg’s syndrome, which includes ipsilateral Horner’s syndrome, ipsilateral ataxia, and contralateral hemisensory findings. These are typically associated with intracranial vertebral artery stenoocclusive disease or posterior inferior cerebellar artery occlusion.91,100 Isolated midbrain infarction is rare; the majority involve additional arterial territories.126 These typically are associated with oculomotor palsy and may have additional symptoms of ataxia or hemiparesis if the superior cerebellar peduncle or corticospinal tracts are also involved.126 Cerebellar infarcts typically present with more nonspecific symptoms such as gait instability, nausea and vomiting, and dizziness, which lead to high rates of initial misdiagnosis.48 There is also significant overlap between symptoms in the vascular territories of the SCA, AICA, and PICA190 and limited sensitivity for detection by NCCT, which also limits the rate of early diagnosis. Multiplanar CT reconstructions may be helpful to increase sensitivity (Fig. 11-17). Moreover, about 15% of patients with cerebellar stroke have additional brainstem infarction,190 because the SCA, AICA, and PICA vessels also provide local perforating vessels to the brainstem during their course. Clinical symptoms from basilar arterial occlusion are variable, depending on which segment of the basilar artery is occluded and the extent of collateral flow. Occlusion of the basilar terminus alone results in loss of consciousness and oculomotor palsy, with less motor and sensory symptoms.106,171 Occlusion of the proximal and middle basilar segments may produce the locked-in syndrome by ischemic damage to the pontine tracts. Typically basilar artery occlusion is caused by atherothrombotic disease or embolism.173 NCCT findings of parenchymal hypoattenuation or swelling in a topographic distribution concordant with the clinical symptoms can confirm the clinical diagnosis of brainstem or cerebellar stroke.

http://radiologyebook.com

326

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 11-17  Utility of multiplanar NCCT for stroke assessment. This 68-year-old woman presented with vertigo and unsteady gait. The right inferior cerebellar infarct is difficult to detect on (A) axial NCCT but easier to detect on (B) coronal and (C) sagittal reconstructions (ovals).

A

B FIG 11-18  Pontine perforator stroke on CTP in a 62-year-old woman who presented with ataxia and dysarthria, NIHSS 3. A, CTP 3.5 hours post onset reveals altered perfusion in a topographic distribution, compatible with pontine perforator infarction. B, MRI DWI confirms infarct at the same location as the CTP abnormality.

However, the sensitivity of NCCT is poor, particularly in the first 24 hours.82 NCCT, however, may detect a hyperdense thrombosed basilar or posterior cerebral vessel that can help confirm the stroke diagnosis.49 The sensitivity for NCCT diagnosis of basilar artery occlusion is poor,37 and thus CTA should be performed if there is any clinicoradiologic suspicion. Data on the use of CTP in the assessment of brainstem and cerebellar stroke are sparse; however, larger areas of ischemia may be discernible (Fig. 11-18). If using perfusion techniques that include  the posterior fossa, the radiologist should be aware of the phenomena of crossed cerebellar diaschisis. Crossed cerebellar diaschisis can result in perfusion abnormalities in the cerebellar hemisphere contralateral to the side of supratentorial infarct, which can mimic cerebellar

ischemia.89 MRI with DWI remains the radiologic examination of choice for assessment of brainstem and cerebellar infarcts and can accurately localize the infarct vascular territory. However, DWI may also be negative within the first 24 hours, particularly if lesion volume is small or deficits are mild/transient.149,193

IMAGING OF MINOR ISCHEMIC STROKE AND TRANSIENT ISCHEMIC ATTACKS Over the last 20 years, improvements in preventive medical therapy have reduced the proportion of patients presenting with major ischemic stroke. Thus TIA and minor stroke (NIHSS < 4) account for the

http://radiologyebook.com

CHAPTER 11  Stroke

A

327

C

B

FIG 11-19  MRI imaging baseline maximizes sensitivity for detection of small strokes. This 41-year-old man presented with ophthalmoplegia and transient hemiparesis. A, Axial DWI at 4.5 hours post symptom onset reveals bilateral acute thalamic infarcts. B, Axial FLAIR MRI at 3 weeks post onset clearly demonstrates the bilateral thalamic infarcts. C, Axial FLAIR at 3 months post onset; the thalamic infarcts are no longer discernible.

majority of acute ischemic events.198 The original definition of TIA was that of “a sudden, focal neurological deficit of presumed vascular origin lasting < 24 hours.”47 However, the 24-hour time frame was chosen arbitrarily in the 1960s.3 This was chosen to help distinguish between TIA and ischemic stroke prior to the advent of neuroimaging techniques that could assess for injury to the brain parenchyma. Moreover, this was prior to the advent of reperfusion therapies, so early treatment was not considered a priority. These minor stroke and TIA patients are generally not eligible for acute reperfusion therapies, as the NINDS and ECASS III trials excluded patients with mild or rapidly improving symptoms. They provide an excellent opportunity for secondary stroke prevention, and CT and MRI play an important role in investigation and management. MRI in particular has led to a new definition of TIA. A third of patients defined as having a TIA by the original definition have evidence of tissue injury on MRI with DWI.47 This has led to movement toward a tissue-based definition of TIA: “a transient episode of neurological dysfunction caused by focal ischemia but without acute infarction.”3 Although this definition relies heavily on imaging, it is not constrained by the sensitivity of the imaging technique. Small perforator infarcts in the brainstem may not be discernible on early MRI with DWI, particularly in the first 24 hours. However, the patient may have an acute onset of a prolonged neurologic deficit and a clinical syndrome consistent with a small perforator territory infarct.149,193 In some cases, repeat interval MRI may later reveal the infarct. Traditionally in TIA and minor stroke patients, NCCT was used to exclude hemorrhage as the cause of the neurologic symptoms, with no further emergent imaging typically performed. Carotid ultrasonography to screen for cervical carotid stenoocclusive disease accompanied NCCT to identify patients who may benefit from carotid endarterectomy. However, increasingly both CT and MRI techniques are being used to distinguish TIA from minor stroke and to help in the prognostication of future stroke risk.205 Although various clinical scoring systems are also available to identify patients at risk of recurrent events, they may not be accurate for rapid decision making for individual patients.168

For the confirmation of minor stroke diagnosis rather than TIA, MRI with DWI is the imaging modality of choice. Greatest sensitivity for minor stroke diagnosis is at initial baseline review; up to a third of infarcts may be missed if only late MRI is performed137 (Fig. 11-19). DWI also allows risk stratification—patients with “TIA” who have DWI-positive scans have a significantly higher risk of recurrent in-hospital stroke and TIA than patients with DWI-negative scans.8 They are also at higher risk of asymptomatic recurrent events confirmed on DWI.38 Moreover, 10% to 15% of patients with a traditional definition of TIA have a central vascular occlusion, which is a predictor of poor outcome.44,158 This may explain some of the poor outcomes in patients with mild or rapidly improving stroke who are not treated with intravenous thrombolysis.180 Thus the use of CTA or MRA aids in identification of central vascular occlusion and can replace carotid ultrasound if the cervical vessels are also included. There is increasing interest in the use of both CT and MR perfusion techniques in the evaluation of TIA and minor stroke patients. For patients with a traditional definition of TIA, the addition of perfusion imaging may increase the yield of detection of ischemia by a further 15%.103 This may not only confirm TIA or minor stroke diagnosis, but in the future may also identify patients at risk for poorer outcome and thus appropriate for more intensive treatment. Imaging of acute TIA and minor stroke continues to evolve, with increasing uptake of the new tissue-based definition of TIA and the use of advanced CT and MRI to not only confirm stroke diagnosis but aid in prognostication.

IMAGING OF SUBACUTE AND CHRONIC STROKE There is no consistent clinical or radiologic definition for subacute and chronic stroke. In general, subacute infarction can be considered between 24 hours and 4 weeks, reflecting corresponding histologic changes of acute neuronal injury and acute inflammation.132 Similarly chronic infarction can be considered as occurring beyond 4 weeks, reflecting the time frame for histologic resolution of inflammation and features of resorptive change.132

http://radiologyebook.com

328

A

PART II  CT and MR Imaging of the Whole Body

C

B

FIG 11-20  CT fogging. A, Axial NCCT reveals subacute left MCA and ACA territory infarcts. B, NCCT 10 days later demonstrates CT fogging, and the large infarct is not clearly discernible. C, NCCT 3 weeks after initial CT reveals that the infarct is now discernible again.

After 24 hours, infarcts are more readily discerned on NCCT compared to earlier timepoints.143 The histologic changes of acute neurologic injury and cytotoxic edema manifest as focal areas of reduced parenchymal CT attenuation with loss of differentiation between gray and white matter. The infarct margins are more readily discernible and local swelling may be present. Changes of hemorrhagic transformation peak with the first 48 hours.203 Whereas some influx of intact red cells into the infarct territory may be a normal phenomenon,97 particularly with reperfusion, hemorrhagic transformation associated with mass effect can be symptomatic and lead to secondary neurologic deterioration. Parenchymal swelling typically peaks within 2 to 5 days, but swelling leading to herniation can occur as early as 20 hours after complete MCA territory infarction.73 Mortality rates are extremely high in this cohort of patients, and the term malignant MCA infarct is used to describe the complete MCA infarct with symptomatic edema.73 Similar symptomatic swelling may occur after large PICA and SCA infarcts. Decompressive craniectomy has emerged as a life-saving option for patients with symptomatic mass effect and neurologic deterioration despite maximal medical therapy, and thus serial NCCT for the first 48 hours is useful in identifying patients at high risk of neurologic deterioration.201 DWI is the neuroimaging technique of choice for identification of acute infarction within the first 14 days.6 Diffusion restriction typically peaks between 1 and 4 days (i.e., increased DWI signal and corresponding decreased ADC signal).60 FLAIR and T2 MRI can also identify the area of infarction in the subacute period. However, DWI is often more valuable because it enables rapid identification of small recent infarcts that may be overlooked or potentially dismissed as areas of background white matter T2 hyperintense lesions. After the first week, there may be a progressive increase in the NCCT attenuation of the infarct. This is related to the influx of polymorphonuclear leukocytes, mononuclear inflammatory cells, and macrophages that are involved in the inflammatory process. By 2 to 3 weeks after onset, the infarct may be isoattenuating to the adjacent brain parenchyma, a phenomenon known as CT fogging.13 This may not occur in all infarcts and may occur in only part of rather than the entire infarct.178 Recognition of this possible phenomenon is important because fogging can result in large infarcts being missed on NCCT (Fig. 11-20). On DWI, after the first week, the DWI signal reduces and the ADC values increase until they both return to baseline. They generally

continue the trend, with subsequent reduction in DWI signal and elevation of ADC signal in the transition from the subacute to the chronic phase. The time course of DWI signal elevation is variable, with many returning to baseline by 2 weeks.22 However, this is variable, with DWI in some patients returning to baseline within the first 5 days and others remaining elevated for months22,111 (Fig. 11-21). The ADC signal evolution is also variable, generally transitioning through baseline after 5 to 14 days; however, this may occur as early as 24 hours.200 Similarly, infarcts may transition through a period from early elevation of FLAIR and T2 signal to normal signal compared to the surrounding parenchyma, a process known as MR fogging. The mechanisms are considered to be similar to CT fogging. This may affect part of a large infarct or an entire small cortical or subcortical infarct.34,146 This is important to recognize because infarcts may be occult on MRI. This typically occurs between 2 and 6 weeks after onset.5 Contrast-enhanced CT or MRI can improve infarct detection in the subacute period, and in particular when infarcts are affected by the fogging phenomenon179 (Fig. 11-22). CT contrast enhancement is generally found between 1 and 4 weeks after onset but may occur as early as 3 days and persist into the second month.24,40 Parenchymal MRI contrast enhancement, however, can occur earlier, including within  the first 24 hours of onset.93,209 Parenchymal contrast enhancement becomes more common later in the first week.39,93 Typically, MRI for ischemic stroke is performed without contrast; however, in the subacute time frame after a week, contrast may increase the sensitivity  for detection compared to conventional MRI sequences and DWI sequences.6,204 Enhancement may persist for up to 3 to 6 months after symptom onset; mild enhancement after a year has been reported.50,105,177 Not uncommonly, T1 hyperintensity is detected in the area of infarct, more commonly in the cortex after large vessel ischemic stroke. The mechanism is unclear but is postulated to relate to increased protein concentration, infiltration with lipid-laden macrophages, and necrosis within the cortex.102,105,177 These are not considered to reflect cortical hemorrhage, and in general SWI sequences are not abnormal95,102 (Fig. 11-23). This phenomenon is considered part of normal stroke evolution and is generally observed by 2 weeks post onset.105 Moreover, there may be an increase in T1 signal intensity by 1 to 3 months and then gradual reduction; in some cases, elevated T1 signal may remain for years.102,105 As time progresses, the histologic changes of inflammation resolve and transition to a resorptive phase with necrosis and cavitation.132

http://radiologyebook.com

CHAPTER 11  Stroke

A

B FIG 11-21  Elevated DWI signal may persist for over a year. A, Axial MRI DWI 4 months after large right MCA infarct reveals a small persisting focus of elevated DWI signal in the top image, while majority of the signal in the chronic infarct approaches CSF. B, Axial MRI DWI 14 months later reveals persisting elevated DWI signal at the same location.

http://radiologyebook.com

329

330

PART II  CT and MR Imaging of the Whole Body

A

B FIG 11-22  MR DWI fogging and T1 MRI contrast enhancement in subacute stroke. This 24-year-old man presented with right hemiparesis. A, Axial DWI (top) 3 days post onset confirms subcortical left MCA stroke; no contrast enhancement is discernible (bottom). B, Axial DWI 9 days later reveals near normalization of DWI signal (top) and prominent contrast enhancement (bottom). Small infarcts may not be discernible in the subacute period without the use of contrast.

http://radiologyebook.com

CHAPTER 11  Stroke

A

B

C

D FIG 11-23  Cortical T1 hyperintensity in chronic stroke. This 83-year-old woman was imaged 9 months after left PCA stroke. Sagittal (A) and axial (B) T1 MRI reveals prominent T1 cortical hyperintensity in the left PCA territory. C, Axial FLAIR reveals that the center of the infarct approaches CSF signal, while the periphery has elevated FLAIR signal from gliosis. D, Axial SWI confirms that the elevated T1 signal is not due to cortical hemorrhage.

http://radiologyebook.com

331

332

A

PART II  CT and MR Imaging of the Whole Body

C

B

FIG 11-24  Changes of chronic infarction, 14 months after right MCA infarction. A, Axial FLAIR above, T2 below demonstrates that the center of the infarct approaches CSF signal, while the periphery has elevated T2 signal from gliosis. Note the enlargement of the ipsilateral lateral ventricle from negative mass effect.  B, DWI above, ADC below shows that there is reduced diffusion signal and elevated ADC signal, respectively. C, MR angiogram shows diminution in the right middle cerebral artery vasculature from reduced metabolic demand.

Tissue resorption leads to parenchymal volume loss, evident on CT and MRI as widening of sulci and ex vacuo dilation of the adjacent ventricle. The center of the infarct may approach CSF attenuation and signal on CT and MRI, respectively, and there may be surrounding gliosis with reduced attenuation on CT and increased signal on FLAIR and T2 MRI (Fig. 11-24). Longer-term imaging findings include wallerian degeneration and calcification. Wallerian degeneration refers to antegrade neuronal degeneration that is most frequently observed in the corticospinal tracts. This is most commonly detected in the late stage with increased signal in the corticospinal tracts and resultant localized brainstem atrophy84,108 (Fig. 11-25). However, the process commences shortly after onset and may be detectable earlier on DWI and ADC.41 Calcification of an infarct is not common, generally developing months to years after onset.92,109 Although microscopic mineralization in infarcts is not uncommon, the pathologic mechanism leading to calcification in only few infarcts remains unclear.109 Calcium deposition may be accelerated in cases of abnormal calcium metabolism35 and may explain why in some cases calcification in infarcts has been reported to occur after several weeks.151,175

IMAGING OF ALTERNATE MECHANISMS OF STROKE Hypoxic-Ischemic Injury Hypoxic-ischemic injury is typically caused by hypoxemia and hypoxia. In children, perinatal asphyxia may lead to initial brain and cardiac hypoxemia, which in turn leads to reduced cardiac output and secondary brain ischemia. However, in adults, cardiac arrest with direct reduction of blood flow to the brain leads to primary ischemic injury. Importantly, the brain tissues have different sensitivities to hypoxicischemic injury, and the variable CT and MRI manifestations are related to multiple variables including age, duration and severity of hypoxic-ischemic injury, and timing of imaging. In general, the pattern of hypoxic-ischemic injury regardless of age is of bilateral involvement of gray matter, in particular the cortex and deep nuclei. This relates  to the increased sensitivity of neurons to hypoxic-ischemic injury  compared to oligodendrocytes and astrocytes in the white matter.70 Even within the deep nuclei, there is often preferential sparing of the thalami and brainstem, and more severe involvement of the caudate and putamen.52

http://radiologyebook.com

CHAPTER 11  Stroke

A

B

C

D

333

FIG 11-25  Wallerian degeneration after right MCA infarction. Sequential axial T2 MRI 14 months after symptom onset, demonstrating large chronic right MCA territory infarct (ovals) (also see Fig. 11-24), with secondary increased signal in the corticospinal tracts from wallerian degeneration (arrows in A-D). Note the ipsilateral pontine and midbrain atrophy.

Similar to ischemic stroke from vessel occlusion, CT may be normal early after onset. In cases of severe or prolonged hypoxic-ischemic injury there may be generalized loss of gray-white matter differentiation, hypoattenuation in the basal ganglia, or diffuse cerebral swelling and effacement of sulci detected on CT (Fig. 11-26). An important pitfall is the relative hyperattenuation of the subarachnoid cisterns, the tentorium cerebelli, and cortical sulci in patients with severe brain edema, termed pseudosubarachnoid hemorrhage.63 This CT imaging phenomenon is considered to occur from the differential attenuation between the edematous and hypoattenuating brain and the congested and hyperattenuating superficial veins in the effaced subarachnoid cisterns. Measurement of the attenuation values can help distinguish between true subarachnoid hemorrhage and pseudosubarachnoid hemorrhage, because true subarachnoid hemorrhage typically is above 45 HU.210 A similar CT imaging phenomenon is the relatively hyperattenuating appearance of the cerebellum compared to the diffusely edematous and hypoattenuating cerebrum.74 MRI is the imaging modality of choice for hypoxic-ischemic injury and is highly sensitive in the acute phase (Fig. 11-27). DWI changes often precede the onset of cortical and gray nuclei FLAIR and T2

hyperintensity and swelling. Extensive cortical and deep gray involvement within the first 12 hours of onset is predictive of a poor clinical outcome.152 In some patients, there may be postanoxic leukoencephalopathy that manifests as confluent areas of white matter diffusion restriction and T2 hyperintensity, typically days to weeks after onset.68 The exact mechanism remains unclear but is considered to be similar to wallerian degeneration.81

Venous Ischemia and Stroke Venous sinus or cortical vein thrombosis leading to stroke is rare compared to other types of stroke.169,184 However, it is more common in children and young females and can have devastating neurologic consequences if not recognized and treated early. The mechanism of brain injury is twofold. Thrombosis leads to obstruction of venous outflow that causes local venous hypertension and edema. Both cytotoxic and vasogenic edema may develop. There is further ischemic neuronal damage, with petechial hemorrhage that can merge to become large hematomas.184 The second mechanism of injury is via disruption of CSF absorption and secondary intracranial hypertension. Both mechanisms of injury are common after venous thrombosis.184

http://radiologyebook.com

334

PART II  CT and MR Imaging of the Whole Body recanalization occurs in only half of all cases, with 20% of affected sinuses remaining occluded.162,185 Nonetheless, overall prognosis is generally favorable. Only 1 in 10 patients are dead or dependent by 6 months.54 The presence of coma, deep venous system involvement, coexistent malignancy, or central nervous system (CNS) infection are predictors of poor outcome, and this patient subgroup may benefit from earlier invasive treatments.54 These include endovascular strategies to recanalize the affected venous sinus or decompressive craniectomy for focal mass effect. These are generally considered in the setting of clinical deterioration despite anticoagulation, or increased local mass effect from increasing edema or hemorrhage.55,169

Nonatherosclerotic Vasculopathy and Stroke

FIG 11-26  CT findings of global hypoxic injury in a 48-year-old man imaged with CT 3 days after cardiac arrest. There is generalized loss of gray-white matter differentiation, hypoattenuation in the basal ganglia, and diffuse cerebral swelling with effacement of sulci.

Severe headache is the most common symptom but also not specific.54 There may be associated nausea and vomiting. The rate of initial misdiagnosis and delayed diagnosis is high,14 and thus radiologists should have a high index of suspicion for young female patients presenting with unexplained severe headache. NCCT may detect the thrombus or signs of secondary injury (Fig. 11-28). The thrombosed vein or sinus may be hyperattenuating; similar to thrombus detection in arterial ischemia, sensitivity may be increased with use of thinner NCCT slices. Secondary areas of parenchymal edema that do not conform to an arterial territory or unexplained parenchymal hemorrhage, particularly in the parasagittal or posterior temporal locations (superior sagittal sinus thrombosis and transverse sinus thrombosis patterns, respectively), should prompt further investigation. Contrastenhanced CT venography can provide rapid diagnosis at the time of detection of NCCT abnormalities96 (Fig. 11-29). However, MRI provides superior assessment of the extent of brain parenchymal injury, in particular areas of parenchymal edema, which may have areas of diffusion restriction. Importantly, the mechanisms of venous ischemia and arterial ischemia are different, and diffusion restriction may not indicate irreversible injury and ultimately reverse.45 MRI may detect the thrombus from the loss of normal expected flow voids or altered T1, T2, DWI, and SWI signals within the sinus or cortical vein113 (Fig. 11-30). MR venography should also be performed; contrast-enhanced MR venography is superior to other MR venography techniques.116 Notably, SWI imaging may have higher sensitivity for cortical vein thrombosis than MR venography.83 The mainstay of treatment for venous thrombosis is anticoagulation, with the rationale of preventing thrombus propagation and facilitating endogenous recanalization.169 If there is a small burden of thrombus, recanalization of the affected sinus after anticoagulation may occur as early as 24 hours.195 However, in the long-term, complete

The nonatherosclerotic vasculopathies comprise a heterogeneous list of diseases of various etiologies including the reversible cerebral vasoconstrictive syndromes, primary angiitis of the CNS (PACNS), systemic diseases and vasculitides that can affect the CNS, and the infectious/postinfectious vasculopathies. Although the list of potential nonatherosclerotic vasculopathy that can cause ischemic stroke is extensive, a rational approach that involves consideration of the demographic features, the duration and nature of symptoms, the pattern of imaging findings (in particular the vessel branch size and location affected), the presence/absence of systemic features of infection/ inflammation, CSF, and other serologic findings can help narrow down the potential etiology. These are discussed briefly below.

Reversible Cerebral Vasoconstrictive Syndromes.  Reversible cerebral vasoconstrictive syndrome (RCVS) is not a disease entity but a common clinicoradiologic syndrome that is caused by a heterogeneous group of disorders. The primary uniting feature is segmental vasoconstriction involving the cerebral vessels that resolves (typically by 3 months). This syndrome includes previous disorders including Call-Fleming syndrome, drug-induced vasculopathy, migrainous vasculopathy, postpartum angiopathy, and benign cerebral angiopathy. RCVS typically affects patients aged 20 to 50, with a female predominance, and presentation usually involves acute or severe headache, focal neurologic deficit, or seizure.136 Brain imaging may reveal intracranial hemorrhage or ischemic infarcts. Vascular imaging is typically normal in the first few days, and maximal involvement is discernible on CTA or MRA by the third week.136 Importantly, although there is no clear cause-effect relationship, there have been cases of clinical worsening after digital subtraction angiography (DSA),94 and thus DSA should be avoided if possible. The clinical course is uniphasic, but there may be initial clinical worsening by 7 to 10 days, commonly due to ischemia.94 CSF studies and serologic tests of inflammation are typically normal.136 Confirmation of the diagnosis requires resolution of the vascular changes by 3 months, so patients are often managed as having “probable” RCVS (Fig. 11-31). Treatment may involve oral calcium channel antagonists; 95% of patients achieve good neurologic outcomes.46

Primary Angiitis of the CNS.  Biopsy-proven PACNS is a rare disease, and most patients suspected with PACNS have an alternate final diagnosis.17 The diagnosis is ideally made with histologic evidence, as this typically affects vessels smaller than 200 µm in diameter.159 Importantly, DSA is frequently normal even in histologically proven cases, and CTA- and MRA-detected large vessel abnormalities are more likely to result from an alternate diagnosis.159 Distinguishing between the more common RCVS is particularly important. In PACNS there is a male predominance, presence of a more insidious onset of symptoms, chronic headache and encephalopathy reflecting diffuse

http://radiologyebook.com

CHAPTER 11  Stroke

A

B FIG 11-27  MRI findings of global hypoxic injury in a 34-year-old man imaged with MRI 4 days after cardiac arrest. A, DWI above and FLAIR below reveal abnormal cortical hyperintensity, particularly in the medial occipital lobes, and abnormal hyperintensity in the basal ganglia. B, The same patient had an MRI 2 years earlier; normal DWI and FLAIR imaging provided for comparison.

http://radiologyebook.com

335

336

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 11-28  Superior sagittal sinus thrombosis in a 66-year-old woman who presented with altered conscious state and right-sided hemiparesis. A, Axial NCCT reveals left frontal parasagittal hemorrhage. This pattern should prompt further assessment of the venous sinuses. Coronal (B) and sagittal (C) NCCT more clearly demonstrate abnormal hyperattenuation in the superior sagittal sinus from extensive venous sinus thrombosis (arrows).

A

C

B

FIG 11-29  Superior sagittal sinus thrombosis in the patient shown in Figure 11-28, who presented with altered conscious state and right-sided hemiparesis. Axial NCCT revealed left frontal parasagittal hemorrhage. A, Sagittal CTA. There is often enough venous contrast on the CTA to assess for venous sinus thrombosis. Note the absence of opacification of the superior sagittal sinus but normal enhancement of the straight sinus. Confirmatory CTV (B) sagittal reconstruction and (C) volumetric reconstruction reveal similar findings, confirming superior superior sagittal sinus thrombosis (arrows).

cortical dysfunction, and common CSF and MRI abnormalities.17 Treatment usually involves high-dose immunosuppression.

Systemic Diseases and Vasculitides.  There are myriad of systemic diseases and vasculitides that can cause cerebral vasculopathy and ischemic stroke. One of the more common conditions implicated is systemic lupus erythematosus (SLE). Although the pathogenesis of some of the brain parenchymal abnormalities in SLE likely relate to a small vessel vasculopathy,88 a detectible large vessel vasculopathy is rare. Patients presenting with ischemic stroke are more likely to have the more typical embolic mechanisms rather than SLE cerebral vasculitis.42 The appropriate clinical context with inflammatory CSF and serologic markers may suggest the diagnosis of SLE cerebral vasculitis, but brain biopsy remains the definitive test.167 Treatment often involves high-dose immunosuppression.

Infectious/Postinfectious Vasculopathy.  There are multiple potential infectious causes for vasculopathy. One of the more common viruses implicated, particularly in pediatric stroke, is varicella zoster.

Although early reports emphasized young patients presenting with acute severe ischemic stroke with unifocal middle artery stenosis or occlusion, there is increasing recognition of a more chronic and protracted course of multifocal and small vessel involvement.62 Brain and vascular imaging typically reveal ischemic parenchymal changes with areas of segmental constriction in the vessels. A modest mononuclear pleocytosis is common; virologic confirmation is required. Acute treatment typically involves intravenous acyclovir.

CONCLUSION The diagnosis and imaging of acute ischemic cerebral stroke is challenging, particularly in the modern neuroimaging era when timecritical treatments are available. Each institution will decide on their own acute CT and/or MR stroke imaging strategy, but the emphasis will have to be on neuroimaging techniques that improve treatment effectiveness and enhance patient safety. Radiologists have a great opportunity to remain relevant and important members of the multidisciplinary acute stroke team. Thus the focus of this chapter has been

http://radiologyebook.com

CHAPTER 11  Stroke

A

B

C

D FIG 11-30  Venous sinus thrombosis on MRI. Use all sequences available to detect and confirm venous thrombus. This 66-year-old woman presented with extensive venous sinus thrombosis. A, Axial T1 reveals abnormal elevated T1 signal in the right transverse sinus (oval). B, Axial T2 reveals corresponding loss of the normal flow void. C, Axial SWI reveals altered signal in the right transverse sinus. D, DWI offers superior detection of the small volume of thrombus in the contralateral transverse venous sinus (arrow).

http://radiologyebook.com

337

338

PART II  CT and MR Imaging of the Whole Body

A

B FIG 11-31  Typical MRA changes of reversible cerebral vasoconstriction syndrome (RCVS) in a 59-year-old woman who presented with initial thunderclap headache. A, Axial MRA 3 weeks after symptom onset reveals diffuse and multifocal areas of cerebral vasoconstriction. B, Axial MRA 6 weeks after symptom onset reveals resolution of vascular changes.

to familiarize the radiologist with the assessment, imaging, and treatment of acute ischemic stroke. As modern stroke treatment along with CT and MRI techniques continue to evolve, so must the knowledge and role of the radiologist.

REFERENCES 1. Adams HP, Jr, Bendixen BH, Kappelle LJ, et al: Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 24(1):35–41, 1993. 2. Aho K, Harmsen P, Hatano S, et al: Cerebrovascular disease in the community: Results of a WHO collaborative study. Bull World Health Organ 58(1):113–130, 1980. 3. Albers GW, Caplan LR, Easton JD, et al: Transient ischemic attack— Proposal for a new definition. N Engl J Med 347(21):1713–1716, 2002. 4. Albers GW, Thijs VN, Wechsler L, et al: Magnetic resonance imaging profiles predict clinical response to early reperfusion: The diffusion and perfusion imaging evaluation for understanding stroke evolution (DEFUSE) study. Ann Neurol 60(5):508–517, 2006. 5. Asato R, Okumura R, Konishi J: “Fogging effect” in MR of cerebral infarct. J Comput Assist Tomogr 15(1):160–162, 1991. 6. Augustin M, Bammer R, Simbrunner J, et al: Diffusion-weighted imaging of patients with subacute cerebral ischemia: Comparison   with conventional and contrast-enhanced MR imaging. AJNR Am J Neuroradiol 21(9):1596–1602, 2000. 7. Ay H, Furie KL, Singhal A, et al: An evidence-based causative classification system for acute ischemic stroke. Ann Neurol 58(5):688– 697, 2005. 8. Ay H, Koroshetz WJ, Benner T, et al: Transient ischemic attack with infarction: A unique syndrome? Ann Neurol 57(5):679–686, 2005. 9. Ay H, Oliveira-Filho J, Buonanno FS, et al: Diffusion-weighted imaging identifies a subset of lacunar infarction associated with embolic source. Stroke 30(12):2644–2650, 1999.

10. Baird TA, Muir KW, Bone I: Basilar artery occlusion. Neurocrit Care 1(3):319–329, 2004. 11. Baird TA, Parsons MW, Phan T, et al: Persistent poststroke hyperglycemia is independently associated with infarct expansion and worse clinical outcome. Stroke 34(9):2208–2214, 2003. 12. Barber PA, Demchuk AM, Hudon ME, et al: Hyperdense sylvian fissure MCA “dot” sign: A CT marker of acute ischemia. Stroke 32(1):84–88, 2001. 13. Becker H, Desch H, Hacker H, et al: CT fogging effect with ischemic cerebral infarcts. Neuroradiology 18(4):185–192, 1979. 14. Bentley JN, Figueroa RE, Vender JR: From presentation to follow-up: Diagnosis and treatment of cerebral venous thrombosis. Neurosurg Focus 27(5):E4, 2009. 15. Bergstrom M, Ericson K, Levander B, et al: Variation with time of the attenuation values of intracranial hematomas. J Comput Assist Tomogr 1(1):57–63, 1977. 16. Berkhemer OA, Fransen PS, Beumer D, et al: A randomized trial of intraarterial treatment for acute ischemic stroke. N Engl J Med 372(1):11–20, 2015. 17. Birnbaum J, Hellmann DB: Primary angiitis of the central nervous system. Arch Neurol 66(6):704–709, 2009. 18. Bogousslavsky J, Van Melle G, Regli F: The Lausanne Stroke Registry: Analysis of 1,000 consecutive patients with first stroke. Stroke 19(9):1083–1092, 1988. 19. Brazzelli M, Chappell FM, Miranda H, et al: Diffusion-weighted imaging and diagnosis of transient ischemic attack. Ann Neurol 75(1):67–76, 2014. 20. Broderick JP, Palesch YY, Demchuk AM, et al: Endovascular therapy after intravenous t-PA versus t-PA alone for stroke. NEJM 368(10):893– 903, 2013. 21. Brott T, Adams HP, Jr, Olinger CP, et al: Measurements of acute cerebral infarction: A clinical examination scale. Stroke 20(7):864–870, 1989. 22. Burdette JH, Ricci PE, Petitti N, et al: Cerebral infarction: Time course of signal intensity changes on diffusion-weighted MR images. AJR Am J Roentgenol 171(3):791–795, 1998.

http://radiologyebook.com

CHAPTER 11  Stroke 23. Burger KM, Tuhrim S, Naidich TP: Brainstem vascular stroke anatomy. Neuroimaging Clin N Am 15(2):297–324, x, 2005. 24. Caille JM, Guibert F, Bidabe AM, et al: Enhancement of cerebral infarcts with CT. Comput Tomogr 4(2):73–77, 1980. 25. Campbell BC, Christensen S, Levi CR, et al: Cerebral blood flow is the optimal CT perfusion parameter for assessing infarct core. Stroke 42(12):3435–3440, 2011. 26. Campbell BC, Christensen S, Levi CR, et al: Comparison of computed tomography perfusion and magnetic resonance imaging perfusiondiffusion mismatch in ischemic stroke. Stroke 43(10):2648–2653, 2012. 27. Campbell BCV, Mitchell PJ, Kleinig TJ, et al: Endovascular therapy for ischemic stroke with perfusion-imaging selection. NEJM 372(11):1009– 1018, 2015. 28. Campbell BC, Mitchell PJ, Yan B, et al: A multicenter, randomized, controlled study to investigate EXtending the time for Thrombolysis in Emergency Neurological Deficits with Intra-Arterial therapy (EXTEND-IA). Int J Stroke 9(1):126–132, 2014. 29. Campbell BC, Purushotham A, Christensen S, et al: The infarct core is well represented by the acute diffusion lesion: Sustained reversal is infrequent. J Cereb Blood Flow Metab 32(1):50–56, 2012. 30. Campbell BC, Weir L, Desmond PM, et al: CT perfusion improves diagnostic accuracy and confidence in acute ischaemic stroke. J Neurol Neurosurg Psychiatry 84(6):613–618, 2013. 31. Campbell BC, Yassi N, Ma H, et al: Imaging selection in ischemic stroke: Feasibility of automated CT-perfusion analysis. Int J Stroke 10(1):51–54, 2015. 32. Castillo M, Scatliff JH, Kwock L, et al: Postmortem MR imaging of lobar cerebral infarction with pathologic and in vivo correlation. Radiographics 16(2):241–250, 1996. 33. Cheng B, Forkert ND, Zavaglia M, et al: Influence of stroke infarct location on functional outcome measured by the modified rankin scale. Stroke 45(6):1695–1702, 2014. 34. Choi P, Srikanth V, Phan T: “Fogging” resulting in normal MRI 3 weeks after ischaemic stroke. BMJ Case Rep 2011, 2011. 35. Chuang TJ, Tang WH, Hung YJ, et al: Aggressive gyriform calcifications and seizures after ischemia stroke in a patient with primary hyperparathyroidism. QJM 107(7):567–569, 2014. 36. Ciccone A, Valvassori L, Nichelatti M, et al: Endovascular treatment for acute ischemic stroke. NEJM 368(10):904–913, 2013. 37. Connell L, Koerte IK, Laubender RP, et al: Hyperdense basilar artery sign—A reliable sign of basilar artery occlusion. Neuroradiology 54(4):321–327, 2012. 38. Coutts SB, Hill MD, Simon JE, et al: Silent ischemia in minor stroke and TIA patients identified on MR imaging. Neurology 65(4):513–517, 2005. 39. Crain MR, Yuh WT, Greene GM, et al: Cerebral ischemia: Evaluation with contrast-enhanced MR imaging. AJNR Am J Neuroradiol 12(4):631–639, 1991. 40. Davis KR, Ackerman RH, Kistler JP, et al: Computed tomography of cerebral infarction: Hemorrhagic, contrast enhancement, and time of appearance. Comput Tomogr 1(1):71–86, 1977. 41. DeVetten G, Coutts SB, Hill MD, et al: Acute corticospinal tract wallerian degeneration is associated with stroke outcome. Stroke 41(4):751–756, 2010. 42. Devinsky O, Petito CK, Alonso DR: Clinical and neuropathological findings in systemic lupus erythematosus: The role of vasculitis, heart emboli, and thrombotic thrombocytopenic purpura. Ann Neurol 23(4):380–384, 1988. 43. Donnan GA, Bladin PF, Berkovic SF, et al: The stroke syndrome of striatocapsular infarction. Brain 114(Pt 1A):51–70, 1991. 44. Dubuc V, Singh D, Modi J, et al: TIA and minor stroke patients with intracranial occlusions in both proximal and distal vessels are most at risk for symptom progression. Cerebrovasc Dis 38(5):389–390, 2014. 45. Ducreux D, Oppenheim C, Vandamme X, et al: Diffusion-weighted imaging patterns of brain damage associated with cerebral venous thrombosis. AJNR Am J Neuroradiol 22(2):261–268, 2001. 46. Ducros A, Hajj-Ali RA, Singhal AB, et al: Reversible cerebral vasoconstriction syndrome. JAMA Neurol 71(3):368, 2014.

339

47. Easton JD, Saver JL, Albers GW, et al: Definition and evaluation of transient ischemic attack: A scientific statement for healthcare professionals from the American Heart Association/American Stroke Association Stroke Council; Council on Cardiovascular Surgery and Anesthesia; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular Nursing; and the Interdisciplinary Council on Peripheral Vascular Disease: The American Academy of Neurology affirms the value of this statement as an educational tool for neurologists. Stroke 40(6):2276–2293, 2009. 48. Edlow JA, Newman-Toker DE, Savitz SI: Diagnosis and initial management of cerebellar infarction. Lancet Neurol 7(10):951–964, 2008. 49. Ehsan T, Hayat G, Malkoff MD, et al: Hyperdense basilar artery. An early computed tomography sign of thrombosis. J Neuroimaging 4(4):200–205, 1994. 50. Elster AD: MR contrast enhancement in brainstem and deep cerebral infarction. AJNR Am J Neuroradiol 12(6):1127–1132, 1991. 51. Ezzeddine MA, Lev MH, McDonald CT, et al: CT angiography with whole brain perfused blood volume imaging: Added clinical value in the assessment of acute stroke. Stroke 33(4):959–966, 2002. 52. Falini A, Barkovich AJ, Calabrese G, et al: Progressive brain failure after diffuse hypoxic ischemic brain injury: A serial MR and proton MR spectroscopic study. AJNR Am J Neuroradiol 19(4):648–652, 1998. 53. Feigin VL, Forouzanfar MH, Krishnamurthi R, et al: Global and regional burden of stroke during 1990-2010: Findings from the Global Burden of Disease Study 2010. Lancet 383(9913):245–254, 2014. 54. Ferro JM, Canhao P, Stam J, et al: Prognosis of cerebral vein and dural sinus thrombosis: Results of the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT). Stroke 35(3):664–670, 2004. 55. Ferro JM, Crassard I, Coutinho JM, et al: Decompressive surgery in cerebrovenous thrombosis: A multicenter registry and a systematic review of individual patient data. Stroke 42(10):2825–2831, 2011. 56. Fiorelli M, Bastianello S, von Kummer R, et al: Hemorrhagic transformation within 36 hours of a cerebral infarct: Relationships with early clinical deterioration and 3-month outcome in the European Cooperative Acute Stroke Study I (ECASS I) cohort. Stroke 30(11):2280–2284, 1999. 57. Forkert ND, Kaesemann P, Treszl A, et al: Comparison of 10 TTP and Tmax estimation techniques for MR perfusion-diffusion mismatch quantification in acute stroke. AJNR Am J Neuroradiol 34(9):1697–1703, 2013. 58. Forster A, Griebe M, Wolf ME, et al: How to identify stroke mimics in patients eligible for intravenous thrombolysis? J Neurol 259(7):1347– 1353, 2012. 59. Forster A, Kerl HU, Wenz H, et al: Diffusion- and perfusion-weighted imaging in acute lacunar infarction: Is there a mismatch? PLoS ONE 8(10):e77428, 2013. 60. Fung SH, Roccatagliata L, Gonzalez RG, et al: MR diffusion imaging in ischemic stroke. Neuroimaging Clin N Am 21(2):345–377, xi, 2011. 61. Furlan A: Intra-arterial Prourokinase for Acute Ischemic Stroke: The PROACT II Study: A randomized controlled trial. JAMA 282(21):2003– 2011, 1999. 62. Gilden D, Cohrs RJ, Mahalingam R, et al: Varicella zoster virus vasculopathies: Diverse clinical manifestations, laboratory features, pathogenesis, and treatment. Lancet Neurol 8(8):731–740, 2009. 63. Given CA, 2nd, Burdette JH, Elster AD, et al: Pseudo-subarachnoid hemorrhage: A potential imaging pitfall associated with diffuse cerebral edema. AJNR Am J Neuroradiol 24(2):254–256, 2003. 64. Go AS, Mozaffarian D, Roger VL, et al: Heart disease and stroke statistics—2014 update: A report from the American Heart Association. Circulation 129(3):e28–e292, 2014. 65. Goldstein LB, Bertels C, Davis JN: Interrater reliability of the NIH stroke scale. Arch Neurol 46(6):660–662, 1989. 66. Goyal M, Demchuk AM, Menon BK, et al: Randomized assessment   of rapid endovascular treatment of ischemic stroke. NEJM 372(11):1019–1030, 2015. 67. Goyal M, Menon BK, Coutts SB, et al, Penumbra Pivotal Stroke Trial Investigators CSP: Effect of baseline CT scan appearance and time to

http://radiologyebook.com

340

68. 69.

70.

71.

72.

73.

74.

75.

76. 77.

78.

79.

80.

81. 82.

83.

84.

85.

86.

87.

88.

89.

PART II  CT and MR Imaging of the Whole Body recanalization on clinical outcomes in endovascular thrombectomy of acute ischemic strokes. Stroke 42(1):93–97, 2011. Greer D, Scripko P, Bartscher J, et al: Serial MRI changes in comatose cardiac arrest patients. Neurocrit Care 14(1):61–67, 2011. Gupta AC, Schaefer PW, Chaudhry ZA, et al: Interobserver reliability of baseline noncontrast CT Alberta Stroke Program Early CT Score for intra-arterial stroke treatment selection. AJNR Am J Neuroradiol 33(6):1046–1049, 2012. Gutierrez LG, Rovira A, Portela LA, et al: CT and MR in non-neonatal hypoxic-ischemic encephalopathy: Radiological findings with pathophysiological correlations. Neuroradiology 52(11):949–976, 2010. Hacke W, Donnan G, Fieschi C, et al: Association of outcome with early stroke treatment: Pooled analysis of ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet 363(9411):768–774, 2004. Hacke W, Kaste M, Bluhmki E, et al: Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med 359(13):1317–1329, 2008. Hacke W, Schwab S, Horn M, et al: “Malignant” middle cerebral artery territory infarction: Clinical course and prognostic signs. Arch Neurol 53(4):309–315, 1996. Han BK, Towbin RB, De Courten-Myers G, et al: Reversal sign on CT: Effect of anoxic/ischemic cerebral injury in children. AJNR Am J Neuroradiol 10(6):1191–1198, 1989. Hand PJ, Kwan J, Lindley RI, et al: Distinguishing between stroke and mimic at the bedside: The brain attack study. Stroke 37(3):769–775, 2006. Hankey GJ: Potential new risk factors for ischemic stroke: What is their potential? Stroke 37(8):2181–2188, 2006. Hill MD, Rowley HA, Adler F, et al: Selection of acute ischemic stroke patients for intra-arterial thrombolysis with pro-urokinase by using ASPECTS. Stroke 34(8):1925–1931, 2003. Hirai T, Korogi Y, Ono K, et al: Prospective evaluation of suspected stenoocclusive disease of the intracranial artery: Combined MR angiography and CT angiography compared with digital subtraction angiography. AJNR Am J Neuroradiol 23(1):93–101, 2002. Hodel J, Aboukais R, Dutouquet B, et al: Double inversion recovery MR sequence for the detection of subacute subarachnoid hemorrhage. AJNR Am J Neuroradiol 36(2):251–258, 2015. Hodel J, Leclerc X, Khaled W, et al: Comparison of 3D multi-echo gradient-echo and 2D T2* MR sequences for the detection of arterial thrombus in patients with acute stroke. Eur Radiol 24(3):762–769, 2014. Huang BY, Castillo M: Hypoxic-ischemic brain injury: Imaging findings from birth to adulthood. Radiographics 28(2):417–439, quiz 617, 2008. Hwang DY, Silva GS, Furie KL, et al: Comparative sensitivity of computed tomography vs. magnetic resonance imaging for detecting acute posterior fossa infarct. J Emerg Med 42(5):559–565, 2012. Idbaih A, Boukobza M, Crassard I, et al: MRI of clot in cerebral venous thrombosis: High diagnostic value of susceptibility-weighted images. Stroke 37(4):991–995, 2006. Inoue Y, Matsumura Y, Fukuda T, et al: MR imaging of wallerian degeneration in the brainstem: Temporal relationships. AJNR Am J Neuroradiol 11(5):897–902, 1990. Iwasaki Y, Kinoshita M: Lacunar syndrome and intracerebral hemorrhage: Clinico-computed tomographic correlations. Comput Med Imaging Graph 12(6):359–363, 1988. Jacobs L, Kinkel WR, Heffner RR, Jr: Autopsy correlations of computerized tomography: Experience with 6,000 CT scans. Neurology 26(12):1111–1118, 1976. Jauch EC, Saver JL, Adams HP, Jr, et al: Guidelines for the early management of patients with acute ischemic stroke: A guideline for healthcare professionals from the American Heart Association/ American Stroke Association. Stroke 44(3):870–947, 2013. Jennekens FG, Kater L: The central nervous system in systemic lupus erythematosus. Part 2. Pathogenetic mechanisms of clinical syndromes: A literature investigation. Rheumatology 41(6):619–630, 2002. Jeon YW, Kim SH, Lee JY, et al: Dynamic CT perfusion imaging for the detection of crossed cerebellar diaschisis in acute ischemic stroke. Korean J Radiol 13(1):12–19, 2012.

90. Kamalian S, Kamalian S, Konstas AA, et al: CT perfusion mean transit time maps optimally distinguish benign oligemia from true “at-risk” ischemic penumbra, but thresholds vary by postprocessing technique. AJNR Am J Neuroradiol 33(3):545–549, 2012. 91. Kameda W, Kawanami T, Kurita K, et al: Lateral and medial medullary infarction: A comparative analysis of 214 patients. Stroke 35(3):694– 699, 2004. 92. Kapila A: Calcification in cerebral infarction. Radiology 153(3):685–687, 1984. 93. Karonen JO, Partanen PL, Vanninen RL, et al: Evolution of MR contrast enhancement patterns during the first week after acute ischemic stroke. AJNR Am J Neuroradiol 22(1):103–111, 2001. 94. Katz BS, Fugate JE, Ameriso SF, et al: Clinical worsening in reversible cerebral vasoconstriction syndrome. JAMA Neurol 71(1):68–73, 2014. 95. Kesavadas C, Santhosh K, Thomas B, et al: Signal changes in cortical laminar necrosis-evidence from susceptibility-weighted magnetic resonance imaging. Neuroradiology 51(5):293–298, 2009. 96. Khandelwal N, Agarwal A, Kochhar R, et al: Comparison of CT venography with MR venography in cerebral sinovenous thrombosis. AJR Am J Roentgenol 187(6):1637–1643, 2006. 97. Khatri R, McKinney AM, Swenson B, et al: Blood-brain barrier, reperfusion injury, and hemorrhagic transformation in acute ischemic stroke. Neurology 79(13 Suppl 1):S52–S57, 2012. 98. Kidwell CS, Chalela JA, Saver JL, et al: Comparison of MRI and CT for detection of acute intracerebral hemorrhage. JAMA 292(15):1823–1830, 2004. 99. Kidwell CS, Jahan R, Gornbein J, et al: A trial of imaging selection and endovascular treatment for ischemic stroke. NEJM 368(10):914–923, 2013. 100. Kim JS: Pure lateral medullary infarction: Clinical-radiological correlation of 130 acute, consecutive patients. Brain 126(Pt 8):1864– 1872, 2003. 101. Kim EY, Yoo E, Choi HY, et al: Thrombus volume comparison between patients with and without hyperattenuated artery sign on CT. AJNR Am J Neuroradiol 29(2):359–362, 2008. 102. Kinoshita T, Ogawa T, Yoshida Y, et al: Curvilinear T1 hyperintense lesions representing cortical necrosis after cerebral infarction. Neuroradiology 47(9):647–651, 2005. 103. Kleinman JT, Zaharchuk G, Mlynash M, et al: Automated perfusion imaging for the evaluation of transient ischemic attack. Stroke 43(6):1556–1560, 2012. 104. Ko Y, Lee S, Chung JW, et al: MRI-based algorithm for acute ischemic stroke subtype classification. J Stroke 16(3):161–172, 2014. 105. Komiyama M, Nakajima H, Nishikawa M, et al: Serial MR observation of cortical laminar necrosis caused by brain infarction. Neuroradiology 40(12):771–777, 1998. 106. Kubik CS, Adams RD: Occlusion of the basilar artery; a clinical and pathological study. Brain 69(2):73–121, 1946. 107. Kudo K, Sasaki M, Yamada K, et al: Differences in CT perfusion maps generated by different commercial software: Quantitative analysis by using identical source data of acute stroke patients. Radiology 254(1):200–209, 2010. 108. Kuhn MJ, Mikulis DJ, Ayoub DM, et al: Wallerian degeneration after cerebral infarction: Evaluation with sequential MR imaging. Radiology 172(1):179–182, 1989. 109. Kuzuhara S, Naito Y, Namura Y, et al: CT demonstration of calcification within old cerebral infarcts. J Comput Assist Tomogr 9(2):268–271, 1985. 110. Kvistad CE, Oygarden H, Logallo N, et al: A dark side of subcortical diffusion-weighted lesions? Characteristics, cause, and outcome in large subcortical infarction: The Bergen Norwegian Stroke Cooperation Study. Stroke 45(9):2710–2716, 2014. 111. Lansberg MG, Thijs VN, O’Brien MW, et al: Evolution of apparent diffusion coefficient, diffusion-weighted, and T2-weighted signal intensity of acute stroke. AJNR Am J Neuroradiol 22(4):637–644, 2001. 112. Latchaw RE, Alberts MJ, Lev MH, et al: Recommendations for imaging of acute ischemic stroke: A scientific statement from the American Heart Association. Stroke 40(11):3646–3678, 2009.

http://radiologyebook.com

CHAPTER 11  Stroke 113. Leach JL, Strub WM, Gaskill-Shipley MF: Cerebral venous thrombus signal intensity and susceptibility effects on gradient recalled-echo MR imaging. AJNR Am J Neuroradiol 28(5):940–945, 2007. 114. Lefkowitz D, LaBenz M, Nudo SR, et al: Hyperacute ischemic stroke missed by diffusion-weighted imaging. AJNR Am J Neuroradiol 20(10):1871–1875, 1999. 115. Leker RR, Keigler G, Eichel R, et al: Should DWI MRI be the primary screening test for stroke? Int J Stroke 9(6):696–697, 2014. 116. Lettau M, Laible M, Barrows RJ, et al: 3-T contrast-enhanced MR angiography with parallel imaging in cerebral venous and sinus thrombosis. J Neuroradiol 38(5):275–282, 2011. 117. Lev MH, Farkas J, Gemmete JJ, et al: Acute stroke: Improved nonenhanced CT detection—Benefits of soft-copy interpretation by using variable window width and center level settings. Radiology 213(1):150–155, 1999. 118. Lev MH, Farkas J, Rodriguez VR, et al: CT angiography in the rapid triage of patients with hyperacute stroke to intraarterial thrombolysis: Accuracy in the detection of large vessel thrombus. J Comput Assist Tomogr 25(4):520–528, 2001. 119. Li Q, Davis S, Mitchell P, et al: Proximal hyperdense middle cerebral artery sign predicts poor response to thrombolysis. PLoS ONE 9(5):e96123, 2014. 120. Libman RB, Wirkowski E, Alvir J, et al: Conditions that mimic stroke in the emergency department. Implications for acute stroke trials. Arch Neurol 52(11):1119–1122, 1995. 121. Lin K, Rapalino O, Law M, et al: Accuracy of the Alberta Stroke Program Early CT Score during the first 3 hours of middle cerebral artery stroke: Comparison of noncontrast CT, CT angiography source images, and CT perfusion. AJNR Am J Neuroradiol 29(5):931–936, 2008. 122. Lummel N, Schulte-Altedorneburg G, Bernau C, et al: Hyperattenuated intracerebral lesions after mechanical recanalization in acute stroke. AJNR Am J Neuroradiol 35(2):345–351, 2014. 123. Ma H, Parsons MW, Christensen S, et al: A multicentre, randomized, double-blinded, placebo-controlled phase III study to investigate EXtending the time for Thrombolysis in Emergency Neurological Deficits (EXTEND). Int J Stroke 7(1):74–80, 2012. 124. Ma H, Wright P, Allport L, et al: Salvage of the PWI/DWI mismatch up to 48 h from stroke onset leads to favorable clinical outcome. Int J Stroke 10(4):565–570, 2015. 125. Manelfe C, Larrue V, von Kummer R, et al: Association of hyperdense middle cerebral artery sign with clinical outcome in patients treated with tissue plasminogen activator. Stroke 30(4):769–772, 1999. 126. Martin PJ, Chang HM, Wityk R, et al: Midbrain infarction: Associations and aetiologies in the New England Medical Center Posterior Circulation Registry. J Neurol Neurosurg Psychiatry 64(3):392–395, 1998. 127. Maruya J, Yamamoto K, Ozawa T, et al: Simultaneous multi-section perfusion CT and CT angiography for the assessment of acute ischemic stroke. Acta Neurochir (Wien) 147(4):383–391, discussion 91-2, 2005. 128. Matano F, Murai Y, Nakagawa S, et al: Atypical radiological and intraoperative findings of acute cerebral hemorrhage caused by ruptured cerebral aneurysm in a patient with severe chronic anemia.   J Nippon Med Sch 81(4):264–268, 2014. 129. Mazya M, Egido JA, Ford GA, et al: Predicting the risk of symptomatic intracerebral hemorrhage in ischemic stroke treated with intravenous alteplase: Safe Implementation of Treatments in Stroke (SITS) symptomatic intracerebral hemorrhage risk score. Stroke 43(6):1524– 1531, 2012. 130. Mehta BP, Leslie-Mazwi TM, Chandra RV, et al: Reducing door-topuncture times for intra-arterial stroke therapy: A pilot quality improvement project. J Am Heart Assoc 3(6):e000963, 2014. 131. Mehta B, Leslie-Mazwi TM, Chandra RV, et al: Assessing variability   in neurointerventional practice patterns for acute ischemic stroke.   J Neurointerv Surg 5(Suppl 1):i52–i57, 2013. 132. Mena H, Cadavid D, Rushing EJ: Human cerebral infarct: A proposed histopathologic classification based on 137 cases. Acta Neuropathol 108(6):524–530, 2004.

341

133. Meretoja A, Keshtkaran M, Saver JL, et al: Stroke thrombolysis: Save a minute, save a day. Stroke 45(4):1053–1058, 2014. 134. Merino JG, Luby M, Benson RT, et al: Predictors of acute stroke mimics in 8187 patients referred to a stroke service. J Stroke Cerebrovasc Dis 22(8):e397–e403, 2013. 135. Meyer BC, Lyden PD: The modified National Institutes of Health Stroke Scale: Its time has come. Int J Stroke 4(4):267–273, 2009. 136. Miller TR, Shivashankar R, Mossa-Basha M, et al: Reversible cerebral vasoconstriction syndrome, part 1: Epidemiology, pathogenesis, and clinical course. AJNR Am J Neuroradiol 36(8):1392–1399, 2015. 137. Moreau F, Modi J, Almekhlafi M, et al: Early magnetic resonance imaging in transient ischemic attack and minor stroke: Do it or lose it. Stroke 44(3):671–674, 2013. 138. Mori E, Tabuchi M, Yamadori A: Lacunar syndrome due to intracerebral hemorrhage. Stroke 16(3):454–459, 1985. 139. Mortimer AM, Little DH, Minhas KS, et al: Thrombus length estimation in acute ischemic stroke: A potential role for delayed contrast enhanced CT. J Neurointerv Surg 6(3):244–248, 2014. 140. Moskowitz MA, Lo EH, Iadecola C: The science of stroke: Mechanisms in search of treatments. Neuron 67(2):181–198, 2010. 141. Moulin T, Cattin F, Crepin-Leblond T, et al: Early CT signs in acute middle cerebral artery infarction: Predictive value for subsequent infarct locations and outcome. Neurology 47(2):366–375, 1996. 142. Muir KW, Baird-Gunning J, Walker L, et al: Can the ischemic penumbra be identified on noncontrast CT of acute stroke? Stroke 38(9):2485–2490, 2007. 143. Mullins ME, Schaefer PW, Sorensen AG, et al: CT and conventional and diffusion-weighted MR imaging in acute stroke: Study in 691 patients at presentation to the emergency department. Radiology 224(2):353– 360, 2002. 144. National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group: Tissue plasminogen activator for acute ischemic stroke.   N Engl J Med 333(24):1581–1587, 1995. 145. Nogueira RG, Gupta R, Jovin TG, et al: Predictors and clinical relevance of hemorrhagic transformation after endovascular therapy for anterior circulation large vessel occlusion strokes: A multicenter retrospective analysis of 1122 patients. J Neurointerv Surg 7(1):16–21, 2015. 146. O’Brien P, Sellar RJ, Wardlaw JM: Fogging on T2-weighted MR after acute ischaemic stroke: How often might this occur and what are the implications? Neuroradiology 46(8):635–641, 2004. 147. O’Donnell MJ, Xavier D, Liu L, et al: Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the INTERSTROKE study): A case-control study. Lancet 376(9735):112–123, 2010. 148. Oliveira-Filho J, Ay H, Schaefer PW, et al: Diffusion-weighted magnetic resonance imaging identifies the “clinically relevant” small-penetrator infarcts. Arch Neurol 57(7):1009–1014, 2000. 149. Oppenheim C, Stanescu R, Dormont D, et al: False-negative diffusionweighted MR findings in acute ischemic stroke. AJNR Am J Neuroradiol 21(8):1434–1440, 2000. 150. Pantoni L, Fierini F, Poggesi A: Thrombolysis in acute stroke   patients with cerebral small vessel disease. Cerebrovasc Dis 37(1):5–13, 2014. 151. Parisi J, Place C, Nag S: Calcification in a recent cerebral infarct— Radiologic and pathologic correlation. Can J Neurol Sci 15(2):152–155, 1988. 152. Park JS, Lee SW, Kim H, et al: Efficacy of diffusion-weighted magnetic resonance imaging performed before therapeutic hypothermia in predicting clinical outcome in comatose cardiopulmonary arrest survivors. Resuscitation 88:132–137, 2014. 153. Parsons MW, Pepper EM, Bateman GA, et al: Identification of the penumbra and infarct core on hyperacute noncontrast and perfusion CT. Neurology 68(10):730–736, 2007. 154. Pereira VM, Gralla J, Davalos A, et al: Prospective, multicenter, single-arm study of mechanical thrombectomy using solitaire flow restoration in acute ischemic stroke. Stroke 44(10):2802–2807, 2013. 155. Pexman JH, Barber PA, Hill MD, et al: Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol 22(8):1534–1542, 2001.

http://radiologyebook.com

342

PART II  CT and MR Imaging of the Whole Body

156. Pfleger MJ, Hardee EP, Contant CF, Jr, et al: Sensitivity and specificity of fluid-blood levels for coagulopathy in acute intracerebral hematomas. AJNR Am J Neuroradiol 15(2):217–223, 1994. 157. Phan CM, Yoo AJ, Hirsch JA, et al: Differentiation of hemorrhage from iodinated contrast in different intracranial compartments using dual-energy head CT. AJNR Am J Neuroradiol 33(6):1088–1094, 2012. 158. Poisson SN, Nguyen-Huynh MN, Johnston SC, et al: Intracranial large vessel occlusion as a predictor of decline in functional status after transient ischemic attack. Stroke 42(1):44–47, 2011. 159. Powers WJ: Primary angiitis of the central nervous system: Diagnostic criteria. Neurol Clin 33(2):515–526, 2015. 160. Puetz V, Dzialowski I, Hill MD, et al: Intracranial thrombus extent predicts clinical outcome, final infarct size and hemorrhagic transformation in ischemic stroke: The clot burden score. Int J Stroke 3(4):230–236, 2008. 161. Pulli B, Yoo AJ: CT Angiography source images with modern multisection CT scanners: Delay time from contrast injection to imaging determines correlation with infarct core. AJNR Am J Neuroradiol 33(4):E61, 2012. 162. Putaala J, Hiltunen S, Salonen O, et al: Recanalization and its correlation to outcome after cerebral venous thrombosis. J Neurol Sci 292(1–2):11–15, 2010. 163. Rajajee V, Kidwell C, Starkman S, et al: Diagnosis of lacunar infarcts within 6 hours of onset by clinical and CT criteria versus MRI.   J Neuroimaging 18(1):66–72, 2008. 164. Reith J, Jorgensen HS, Pedersen PM, et al: Body temperature in acute stroke: Relation to stroke severity, infarct size, mortality, and outcome. Lancet 347(8999):422–425, 1996. 165. Riedel CH, Jensen U, Rohr A, et al: Assessment of thrombus in acute middle cerebral artery occlusion using thin-slice nonenhanced computed tomography reconstructions. Stroke 41(8):1659–1664, 2010. 166. Riedel CH, Zimmermann P, Jensen-Kondering U, et al: The importance of size: Successful recanalization by intravenous thrombolysis in acute anterior stroke depends on thrombus length. Stroke 42(6):1775–1777, 2011. 167. Rowshani AT, Remans P, Rozemuller A, et al: Cerebral vasculitis as a primary manifestation of systemic lupus erythematosus. Ann Rheum Dis 64(5):784–786, 2005. 168. Sanders LM, Srikanth VK, Blacker DJ, et al: Performance of the ABCD2 score for stroke risk post TIA: Meta-analysis and probability modeling. Neurology 79(10):971–980, 2012. 169. Saposnik G, Barinagarrementeria F, Brown RD, Jr, et al: Diagnosis and management of cerebral venous thrombosis: A statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 42(4):1158–1192, 2011. 170. Saver JL, Fonarow GC, Smith EE, et al: Time to treatment with intravenous tissue plasminogen activator and outcome from acute ischemic stroke. JAMA 309(23):2480–2488, 2013. 171. Savitz SI, Caplan LR: Vertebrobasilar disease. N Engl J Med 352(25):2618–2626, 2005. 172. Schellinger PD, Bryan RN, Caplan LR, et al: Evidence-based guideline: The role of diffusion and perfusion MRI for the diagnosis of acute ischemic stroke: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 75(2):177–185, 2010. 173. Schonewille WJ, Wijman CA, Michel P, et al: Treatment and outcomes of acute basilar artery occlusion in the Basilar Artery International Cooperation Study (BASICS): A prospective registry study. Lancet Neurol 8(8):724–730, 2009. 174. Schramm P, Schellinger PD, Fiebach JB, et al: Comparison of CT and CT angiography source images with diffusion-weighted imaging in patients with acute stroke within 6 hours after onset. Stroke 33(10):2426–2432, 2002. 175. Sener RN: Gyral calcifications detected on the 45th day after cerebral infarction. Pediatr Radiol 23(7):570–571, 1993. 176. Singer OC, Kurre W, Humpich MC, et al: Risk assessment of symptomatic intracerebral hemorrhage after thrombolysis using DWI-ASPECTS. Stroke 40(8):2743–2748, 2009.

177. Siskas N, Lefkopoulos A, Ioannidis I, et al: Cortical laminar necrosis in brain infarcts: Serial MRI. Neuroradiology 45(5):283–288, 2003. 178. Skriver EB, Olsen TS: Transient disappearance of cerebral infarcts on CT scan, the so-called fogging effect. Neuroradiology 22(2):61–65, 1981. 179. Skriver EB, Olsen TS: Contrast enhancement of cerebral infarcts. Incidence and clinical value in different states of cerebral infarction. Neuroradiology 23(5):259–265, 1982. 180. Smith EE, Fonarow GC, Reeves MJ, et al: Outcomes in mild or rapidly improving stroke not treated with intravenous recombinant tissue-type plasminogen activator: Findings from Get With the Guidelines—Stroke. Stroke 42(11):3110–3115, 2011. 181. Smith WS, Lev MH, English JD, et al: Significance of large vessel intracranial occlusion causing acute ischemic stroke and TIA. Stroke 40(12):3834–3840, 2009. 182. Smith EE, Rosand J, Greenberg SM: Imaging of hemorrhagic stroke. Magn Reson Imaging Clin N Am 14(2):127–140, v, 2006. 183. Sobesky J: Refining the mismatch concept in acute stroke: Lessons learned from PET and MRI. J Cereb Blood Flow Metab 32(7):1416– 1425, 2012. 184. Stam J: Thrombosis of the cerebral veins and sinuses. N Engl J Med 352(17):1791–1798, 2005. 185. Strupp M, Covi M, Seelos K, et al: Cerebral venous thrombosis: Correlation between recanalization and clinical outcome—A long-term follow-up of 40 patients. J Neurol 249(8):1123–1124, 2002. 186. Sylaja PN, Coutts SB, Krol A, et al: When to expect negative diffusionweighted images in stroke and transient ischemic attack. Stroke 39(6):1898–1900, 2008. 187. Tao WD, Liu M, Fisher M, et al: Posterior versus anterior circulation infarction: How different are the neurological deficits? Stroke 43(8):2060–2065, 2012. 188. Thomalla G, Fiebach JB, Ostergaard L, et al: A multicenter, randomized, double-blind, placebo-controlled trial to test efficacy and safety of magnetic resonance imaging-based thrombolysis in wake-up stroke (WAKE-UP). Int J Stroke 9(6):829–836, 2014. 189. Thomalla G, Rossbach P, Rosenkranz M, et al: Negative fluid-attenuated inversion recovery imaging identifies acute ischemic stroke at 3 hours or less. Ann Neurol 65(6):724–732, 2009. 190. Tohgi H, Takahashi S, Chiba K, et al: Cerebellar infarction. Clinical   and neuroimaging analysis in 293 patients. The Tohoku Cerebellar Infarction Study Group. Stroke 24(11):1697–1701, 1993. 191. Tomsick TA, Yeatts SD, Liebeskind DS, et al: Endovascular revascularization results in IMS III: Intracranial ICA and M1 occlusions. J Neurointerv Surg 2014. 192. Tong D, Reeves MJ, Hernandez AF, et al: Times from symptom onset to hospital arrival in the Get With the Guidelines—Stroke Program 2002 to 2009: Temporal trends and implications. Stroke 43(7):1912–1917, 2012. 193. Tsuyusaki Y, Sakakibara R, Kishi M, et al: “Invisible” brainstem infarction at the first day. J Stroke Cerebrovasc Dis 23(7):1903–1907, 2014. 194. Turney TM, Garraway WM, Whisnant JP: The natural history of hemispheric and brainstem infarction in Rochester, Minnesota. Stroke 15(5):790–794, 1984. 195. Vanden Abeele B, Saliou G, Lehmann P, et al: Recanalization of cerebral venous thrombosis within 24 hours: A case report. Neurology 69(2):221, 2007. 196. Verma RK, Kottke R, Andereggen L, et al: Detecting subarachnoid hemorrhage: Comparison of combined FLAIR/SWI versus CT. Eur J Radiol 82(9):1539–1545, 2013. 197. von Kummer R, Allen KL, Holle R, et al: Acute stroke: Usefulness of early CT findings before thrombolytic therapy. Radiology 205(2):327– 333, 1997. 198. von Weitzel-Mudersbach P, Andersen G, Hundborg HH, et al:   Transient ischemic attack and minor stroke are the most common manifestations of acute cerebrovascular disease: A prospective, population-based study—the Aarhus TIA study. Neuroepidemiology 40(1):50–55, 2013.

http://radiologyebook.com

CHAPTER 11  Stroke 199. Vroomen PC, Buddingh MK, Luijckx GJ, et al: The incidence of stroke mimics among stroke department admissions in relation to age group.   J Stroke Cerebrovasc Dis 17(6):418–422, 2008. 200. Welch KM, Windham J, Knight RA, et al: A model to predict the histopathology of human stroke using diffusion and T2-weighted magnetic resonance imaging. Stroke 26(11):1983–1989, 1995. 201. Wijdicks EF, Sheth KN, Carter BS, et al: Recommendations for the management of cerebral and cerebellar infarction with swelling: A statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 45(4):1222–1238, 2014. 202. Wintermark M, Flanders AE, Velthuis B, et al: Perfusion-CT assessment of infarct core and penumbra: Receiver operating characteristic curve analysis in 130 patients suspected of acute hemispheric stroke. Stroke 37(4):979–985, 2006. 203. Xavier AR, Qureshi AI, Kirmani JF, et al: Neuroimaging of stroke: A review. South Med J 96(4):367–379, 2003. 204. Yamada N, Imakita S, Sakuma T: Value of diffusion-weighted imaging and apparent diffusion coefficient in recent cerebral infarctions: A correlative study with contrast-enhanced T1-weighted imaging. AJNR Am J Neuroradiol 20(2):193–198, 1999. 205. Yoo AJ, Chandra RV, Leslie-Mazwi TM: Catching strokes before they happen: The importance of early neuroimaging findings in TIA and

206.

207.

208.

209.

210.

211.

343

minor ischemic stroke. Expert Rev Cardiovasc Ther 10(7):847–850, 2012. Yoo AJ, Leslie-Mazwi TM, Riedel CH, et al: Pathologic validation of clot length determined using thin section non-contrast CT. J Neurointerv Surg 5(4):e16, 2013. Yoo AJ, Verduzco LA, Schaefer PW, et al: MRI-based selection for intra-arterial stroke therapy: Value of pretreatment diffusion-weighted imaging lesion volume in selecting patients with acute stroke who will benefit from early recanalization. Stroke 40(6):2046–2054, 2009. Yoon W, Jung MY, Jung SH, et al: Subarachnoid hemorrhage in a multimodal approach heavily weighted toward mechanical thrombectomy with solitaire stent in acute stroke. Stroke 44(2):414–419, 2013. Yuh WT, Crain MR, Loes DJ, et al: MR imaging of cerebral ischemia: Findings in the first 24 hours. AJNR Am J Neuroradiol 12(4):621–629, 1991. Yuzawa H, Higano S, Mugikura S, et al: Pseudo-subarachnoid hemorrhage found in patients with postresuscitation encephalopathy: Characteristics of CT findings and clinical importance. AJNR Am J Neuroradiol 29(8):1544–1549, 2008. Zhang J, Yang Y, Sun H, et al: Hemorrhagic transformation after cerebral infarction: Current concepts and challenges. Ann Transl Med 2(8):81, 2014.

http://radiologyebook.com

12  Cerebral Aneurysms and Cerebrovascular Malformations Theodore C. Larson III

CEREBRAL ANEURYSMS INCIDENCE AND NATURAL HISTORY The incidence of a cerebral aneurysm in the general population is 0.5% to 5%.46,60,169,185,312,328,338,343,351,354,443,446,454 Women more commonly harbor a cerebral aneurysm than men, and aneurysm development increases with advancing age.185,312,328,338,443 Some families demonstrate a genetic predisposition to develop cerebral aneurysms,46,294 with a prevalence of 4% when a first-degree relative has a confirmed aneurysm, increasing to 8% if two first-degree relatives have aneurysms.60 Twenty percent of patients with unruptured cerebral aneurysms have a family member with an aneurysm or a history of subarachnoid hemorrhage.446 A genetic propensity without confirmed genetic locus mutation to develop cerebral aneurysms can be found in some families without known syndromic disorders.60,445 The risk of new aneurysm development in patients with previously treated or untreated cerebral aneurysms is approximately 0.7% to 2% per year.185,425 Subarachnoid hemorrhage is the most common presenting sign of a cerebral aneurysm, and a cerebral aneurysm is the second most common cause of subarachnoid hemorrhage, trauma being the most common etiology.351 From 80% to 90% of nontraumatic subarachnoid hemorrhages are due to a ruptured cerebral aneurysm, a far more common cause than an arteriovenous malformation (AVM), which accounts for 5% of subarachnoid hemorrhages.195,199,397,399 Other less common presenting signs and symptoms include cranial nerve palsy, headache, seizure, and transient or permanent neurologic dysfunction that may be due to mass effect, neighboring inflammation, aneurysm thrombosis, or aneurysm-generated emboli.328 These nonhemorrhagic symptoms can be seen with aneurysms of any size but occur more commonly with larger aneurysms and those in the posterior circulation or located along the internal carotid artery.328

Ruptured Cerebral Aneurysm The annual incidence of rupture of an asymptomatic cerebral aneurysm has been estimated at 0.5% to 2%, or roughly 7 to 10 cases per 100,000 population.60,140,171,177,185,186,338,343,353,366,397,417 The rate of aneurysm rupture is cumulative over decades, so there is a 10% risk after 10 years and a 30% risk after 30 years.186 Symptomatic intact aneurysms have a higher rate of bleeding—4% or more per year.46,338 Some studies have suggested that for aneurysms less than 7 to 10 mm in diameter, the incidence of annual rupture is less than 0.05% to 0.7%.264,332,353,421,446 Other reports, however, have documented that 70% to 85% of ruptured aneurysms are less than 9 to 10 mm in diameter,169,196,343,347 and that even small (2-3 mm) aneurysms may cause subarachnoid hemorrhage.366,385,417,457 The larger the aneurysm, the greater the risk of rupture,60 with large aneurysms (10-24 mm) having

344

a hazard ratio of 9, and giant aneurysms (≥25 mm) having a hazard ratio of 76 compared with a 3- to 4-mm aneurysm.417 Several arbitrary size classification systems designate aneurysms as small, large, or giant, the medium category having been neglected.140 The presence of a daughter sac, multilobulation, irregular surface, a tubular or bottleneck shape (increased aspect ratio or aneurysm height-to-neck ratio), increased size or volume orifice area ratios (ratio of maximum aneurysm diameter to parent vessel diameter; ratio of aneurysm volume  to neck area), and interval growth probably raise the likelihood of   rupture.60,261,347,353,366,417,446 Smoking and larger initial aneurysm size are associated with interval aneurysm growth.60 Anterior and posterior communicating aneurysms are almost twice as likely to rupture  as aneurysms in other locations.60,171,338,417,446 In addition, vertebral basilar and posterior cerebral aneurysms carry a significant risk of rupture.46,80,171,338 A second unruptured aneurysm discovered in a patient with subarachnoid hemorrhage has an increased risk of causing a future second subarachnoid hemorrhage.171,338,446 Patients with a treated ruptured cerebral aneurysm have an increased incidence of new aneurysm formation and recurrent subarachnoid hemorrhage.140 Younger patients and women have a greater propensity for aneurysmal hemorrhage, especially if they have multiple aneurysms, although statistically a patient with a ruptured aneurysm is more likely to be of increased age.46,60,80,185,186,338,446 Some of the factors implicated in aneurysm rupture include a positive family history of aneurysmal subarachnoid hemorrhage, smoking and exposure to smoke, hypertension including variable episodes of blood pressure elevation, excessive alcohol intake, use of stimulants and vasoconstricting agents, excessive caffeine ingestion, specific genetic syndromes (see following), and extreme straining or Valsalva maneuvers including sexual intercourse.60,80,140,186,332,446 Smoking and positive family history may be additive.60,172 The majority of cerebral aneurysms, however, do not bleed, and there is no convincing evidence that pregnancy and childbirth are associated with aneurysm rupture.46,80 A ruptured cerebral aneurysm carries significant morbidity and mortality, with the 30-day mortality traditionally being stated as 40% to 65%46,47,160,171,185,196,343,354; 12% to 15% expire before reaching medical care.80 Older age, severe comorbidities, worse presentation grade (Tables 12-1 and 12-2), large amount of subarachnoid hemorrhage, intraventricular or intraparenchymal hemorrhage, and global cerebral edema predict worse outcome.80,171,397 Untreated patients with ruptured cerebral aneurysms have a rebleeding rate of 2% to 4% in the first 24 hours and then 15% to 20% in the first 2 weeks.46 Thereafter the bleeding rate is approximately 3% per year.177,185 The risk of rebleeding is 10 times higher in the 10% to 40% of patients with a sentinel aneurysmal bleed, a symptom most often deduced as a preceding headache that in some instances can quickly abate.80,397 Rebleeding carries a 50% to 67% mortality.48,310,397,455 Initial aneurysm size is not a prognostic factor

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations TABLE 12-1  Hunt and Hess Grading Scale

for Subarachnoid Hemorrhage Grade

Characteristics

I II

Asymptomatic or minimal headache Moderate to severe headache, nuchal rigidity, oculomotor palsy, normal level of consciousness Confusion, drowsiness, mild focal neurologic deficit Stupor or hemiparesis Comatose or moribund with decerebrate posturing

III IV V

TABLE 12-2  World Federation of

Neurological Surgeons Grading of Subarachnoid Hemorrhage Grade 1 2 3 4 5

Glasgow Coma Scale Score 15 13-14 13-14 7-12 3-6

Clinical Evaluation No motor deficit No motor deficit Motor deficit ± Motor deficit ± Motor deficit

influencing rebleeding in the acute period. Ruptured aneurysms typically increase in size if left untreated.455 From 21% to 25% of patients who enter the hospital for treatment of subarachnoid hemorrhage eventually die as a result of the condition.361,397 Approximately 14% to 23% of survivors of a ruptured cerebral aneurysm are left with significant morbidity361; in 13% to 32% the morbidity is the direct result of delayed vasospasm-caused ischemia.259,313,317,360 More than half to two thirds will be able to resume independent living and return to work, one third requiring lifetime care.140,172 Some 20% to 45% are left with long-lasting cognitive deficits, and more than 50% have more subtle residual memory, emotional, or neuropsychological impairments.80,397 Surgical clipping of acutely ruptured aneurysms results in improved functional outcome compared with delayed intervention.299 These data also argue for elective treatment of an unruptured aneurysm.455

Unruptured Cerebral Aneurysm Unruptured cerebral aneurysms are most often asymptomatic, identified incidentally on a noninvasive imaging study. Focal mass effect can manifest as compression of adjacent cranial nerves or brain parenchyma, a typical example being third-nerve palsy caused by a posterior communicating aneurysm. Other diseases, however, can cause similar neurologic deficits. For example, third-nerve palsy could be caused by diabetes or microvascular disease instead of a posterior communicating aneurysm. When magnetic resonance imaging (MRI), MR angiography (MRA), or computed tomography angiography (CTA) fail to diagnose the cause of the patient’s third-nerve palsy, catheter cerebral angiography may be required to exclude a cerebral aneurysm, because the morbidity and mortality associated with cerebral aneurysms are so significant. When catheter cerebral angiography does not demonstrate an aneurysm in such a patient with no history of subarachnoid hemorrhage, the patient can be presumed to have small-vessel cerebrovascular disease. Sometimes the angiogram supports this diagnosis by showing other areas of atherosclerotic disease.

345

Growth and (potentially) rupture of a cerebral aneurysm has been linked to untreated or variably elevated systolic blood pressure, smoking, female sex, age older than 74, aneurysm location in the internal carotid artery or middle cerebral artery, aneurysm diameter of 7 mm or greater, and multiple cerebral aneurysms.167,213

TYPES OF ANEURYSMS, ETIOLOGY, AND PATHOLOGY Aneurysms may be classified by etiology and characterized according to morphology, size, or location. Most cerebral aneurysms are acquired lesions and are saccular or fusiform. Older medical reports proposed that saccular or berry cerebral aneurysms were due to an embryologic abnormality of the walls of cerebral vessels.303 Stehbens392 and others have refuted this theory. Cerebral aneurysms can develop from  hemodynamic stresses,168,354,392 which typically occur at vessel bifur­ cations or terminations or along the outer curvature of looping vessels.168,267,333,366,392,446 The aneurysm’s location is a continuation of the primary flow direction in the parent vessel, with greater prevalence in tortuous arteries and increased arterial bifurcation angles.333,347 Rupture and thrombosis of cerebral aneurysms may also be related to hemodynamic forces.129,392,446 Supporting this theory is the higher incidence of cerebral aneurysms in patients with intracranial vascular anomalies, including fenestrations93,428 and congenital or iatrogenic vascular variants of the circle of Willis (e.g., persistent trigeminal artery, unilateral absence or hypoplasia of the A1 segment of an anterior cerebral artery [Fig. 12-1], azygos anterior cerebral artery A2 segment [Fig. 12-2; also see Fig. 12-5], and occlusion or absence of an internal carotid artery).66,318,347,357,365 Disturbed or unstable blood flow patterns, resonant frequency–generated vibrations, oscillatory wall shear stress, transmural pressure, and entrance and internal aneurysmal flow rates and patterns—many characteristics investigated by computational fluid dynamics—are proposed mechanisms of aneurysm development and rupture446 (see Fig. 12-30). Supporting this reasoning is that flowgenerated aneurysms associated with cerebral AVMs have an increased risk of rupture.446 There is no significant pressure difference within aneurysms compared with parent arteries, but intraaneurysmal flow velocities are lower; wall shear stress analysis is complex.347 Regions of relative flow stagnation have been linked to aneurysm growth and rupture.347 Fusiform aneurysms are vascular ectasias caused by advanced focal atherosclerotic disease that has degraded the vessel’s media.392 The vertebrobasilar system is more commonly affected than the anterior circulation (Fig. 12-3). The involved vessel is frequently elongated and tortuous, so in effect the native vessel is replaced by aneurysmal dilatation. Although fusiform aneurysms may rupture, thrombosis due to blood stagnation may be the mode of presentation. Local mass effect, neighboring tissue edema, wall calcification, ischemia due to penetrating vessel occlusion or emboli generated by mural thrombus (see Fig. 12-3C), and cranial nerve palsies are other pathologic sequelae.140,399 Rarely, fusiform aneurysms can form giant serpentine aneurysms due to repeated thrombosis and recanalization.140 An infundibulum represents the residua of a developmental vessel that has undergone incomplete regression. It most commonly involves the junction of the internal carotid artery and posterior communi­ cating artery and less commonly involves the origin of the anterior choroidal artery, ophthalmic artery, middle cerebral branch, a lenticulostriate or thalamoperforating artery, or rarely other intracranial artery origins (Fig. 12-4). Infundibula are usually 3 mm or less in diameter. They can be difficult to differentiate from cerebral aneurysms unless the characteristic origination of the emerging vessel from the summit of the dome of the infundibulum is identified.

http://radiologyebook.com

346

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-1  Reconstructed 3D CTA in a patient with polycystic kidney disease illustrating (A, oblique view) a ruptured small, right anterior communicating aneurysm (arrow), and (B, basal view) a hypoplastic left anterior cerebral artery A1 segment (arrow).

A

B

FIG 12-2  Reconstructed 3D CTA in (A) lateral and (B) frontal projections demonstrating an azygous A3 segment aneurysm (arrows).

Infundibula have rarely been reported to cause subarachnoid hemorrhage.83

Etiology Leading causes of cerebral aneurysms are: • Atherosclerosis338 • Hypertension46,90,167 • Smoking46 • Familial genetic predisposition,46,338 sometimes transmitted as autosomal dominant transmission,140 with a reported incidence of 17.9%278

• Abuse of cocaine, methamphetamine, ephedrine, heroin, and other drugs that produce arteritis or hypertension174,185,281,310 (Fig. 12-5) • Vascular malformations (≈6%-10% incidence of aneurysm, both proximal flow–related and nidal)12,50,88,116,209,254,275 (Fig. 12-6), and moyamoya disease446 • Specific vascular diseases and syndromes, such as spontaneous cervical carotid or vertebral artery dissection364 (more commonly seen in women),365 spontaneous or posttraumatic intracranial dissection, polycystic kidney disease181 (10%-22.5% incidence of aneurysm, increasing with age60,286,338; see Fig. 12-1),

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

B A

C FIG 12-3  Sagittal (A) and coronal (B) individual slice reconstructions of a CTA demonstrating an atherosclerotic fusiform basilar aneurysm (arrow). Note the calcification within the wall of the aneurysm. C, Axial noncontrast CT scan illustrating an acute right occipital infarction, a small chronic left occipital infarction, and an unrelated chronic right frontal infarction (arrows). The occipital infarctions may be secondary to emboli from the atherosclerotic fusiform basilar aneurysm.

http://radiologyebook.com

347

348

PART II  CT and MR Imaging of the Whole Body

B

A FIG 12-4  A, Axial source image of a CTA, with the arrow indicating the site of a left posterior communicating infundibulum that cannot confidently be diagnosed on CTA alone. B, The arrow highlights the easily verified left posterior communicating infundibulum on this lateral view of a catheter left carotid cerebral angiogram, with the posterior communicating artery emanating from it.

FIG 12-5  Sagittal CTA demonstrating a pericallosal aneurysm (arrow) in a cocaine user with an azygous A2 segment of the anterior cerebral arteries.

FIG 12-6  Catheter cerebral angiogram, AP view, of a right middle cerebral territory AVM with three flow-related aneurysms (arrows).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

349

C

FIG 12-7  Woman with carotid fibromuscular dysplasia seen on the CTA sagittal reconstruction of the left carotid (arrow in A), better illustrated on the cervical lateral view of the catheter left carotid angiogram (arrow in B), and also indicated in the right carotid artery on the lateral right carotid cerebral angiogram (white arrow in C) that also demonstrates a small ruptured right posterior communicating aneurysm (black arrow in C).

α1-antitrypsin deficiency,140 coarctation of the aorta (10% incidence),79,81,90 bicuspid aortic valve (10% incidence),60 neurofibromatosis type 1 (von Recklinghausen’s disease),158 hereditary hemorrhagic telangiectasia (HHT, Osler-Weber-Rendu disease; 2.5% incidence),140,458 and Takayasu’s arteritis145 • Connective tissue disorders, such as fibromuscular dysplasia (20%-50% incidence of aneurysm, more commonly seen in women)152,365 (Fig. 12-7), Marfan’s syndrome,60 pseudoxanthoma elasticum,397 and Ehlers-Danlos syndrome type IV (in which aneurysms of the carotid siphon to supraclinoid segment are more common)60,365,446 (Fig. 12-8) • Inflammation or mycosis due to septic emboli causing destruction of the endothelium, spread of infection to the vasa vasorum with subsequent vessel wall destruction and focal dilatation, or local infectious/inflammatory vessel wall invasion (mycotic or infectious aneurysm) (Fig. 12-9) • Neoplasm, especially pituitary adenoma338 • Penetrating and nonpenetrating trauma • Radiation281 (Fig. 12-10) Hypertension and smoking may be additive causal factors.325 Estrogen deficiency as a hormonal causative factor has been proposed.60 Rarely cerebral aneurysms may be congenital.185 Except for congenital aneurysms, aneurysms with uncommon causes often develop in unusual locations that are distal to the circle of Willis.

Dissection.  Dissection of the internal carotid or vertebral arteries may be intracranial or extracranial and can be spontaneous, posttraumatic, or secondary to disease of the vessel wall, such as with fibromuscular dysplasia.273 A pseudoaneurysm accompanies up to one third of internal carotid dissections and 7% of vertebral artery dissections386 (Fig. 12-11). Dissections and dissecting pseudoaneurysms may manifest as ischemic symptoms due to emboli or vessel stenosis or

occlusion. Intracranial dissections and dissecting pseudoaneurysms most often present in young to middle-aged men and often present as subarachnoid hemorrhage, especially when occurring in typically involved vertebral or posterior inferior cerebellar arteries.110,140 Rebleed rates are particularly high, with accompanying significant mortality reaching 80%.140,273 Dissections with or without pseudoaneurysms can less commonly involve the basilar, posterior cerebral, or middle cerebral arteries.140 A blister-type aneurysm may represent a small focal dissection, most commonly involving the proximal intracranial internal carotid artery.140 Connective tissue disorders, fibromuscular dysplasia, and α1-antitrypsin deficiency have an increased incidence of dissections and cerebral aneurysms.60

Trauma.  Trauma typically results in a pseudoaneurysm rather than a true aneurysm, because no normal vascular components constitute the sac within a hematoma surrounding the injured vessel. Posttraumatic pseudoaneurysms due to blunt trauma often occur in unusual locations, including the proximal skull base (internal carotid artery), posterior or anterior cerebral arteries (caused by tentorial or falcine impaction, respectively),280,292 and meningeal or scalp arteries.374,411,415 Penetrating head trauma near the circle of Willis predicts arterial injury and pseudoaneurysm formation.38,39 Traumatic pseudoaneurysms may not be visualized initially but then can rapidly enlarge (Fig. 12-12)280; 50% of untreated posttraumatic aneurysms bleed, with an almost uniformly fatal outcome.11,280 Traumatic pseudoaneurysms may also be iatrogenic.

Neoplasm.  Uncommon neoplastic aneurysms (e.g., cardiac myxomatous aneurysms) are pathologically pseudoaneurysms resulting from primary or metastatic tumor invasion of a vessel, or metastatic emboli to the vessel lumen or its vasa vasorum.157,293 Pituitary adenomas are associated with a higher incidence of cerebral aneurysms, with

http://radiologyebook.com

350

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 12-8  Patient with genetically confirmed Ehlers-Danlos syndrome and new-onset headache. A, Axial CTA section documenting a new pseudoaneurysm (arrow) involving the subarachnoid intracranial right vertebral artery, not seen on a prior CTA (B) (arrow). Note the interval apparent vertebral artery lumen enlargement. C, Lateral oblique view of the right vertebral cerebral catheter angiogram confirming the pseudoaneurysm (arrow) located proximal to posterior inferior cerebellar artery (PICA). The widened vertebral artery segment proximal to the pseudoaneurysm indicates this is a dissection.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations acromegaly-producing adenomas theorized to sometimes produce diffuse cerebrovascular ectasia or unilateral or bilateral cavernous carotid aneurysms due to growth hormone secretion.12,444 (Fig. 12-13).

Pathology The walls of saccular cerebral aneurysms contain intima and adventitia, but the media and internal elastic membrane are thinned or absent, supporting the proposition that degeneration of elastica and muscle cells are the foundations of aneurysm development.46,60,446 Tunica

351

media pathology has been confirmed in cerebral aneurysms found in patients with genetic predispositions.140 Aneurysm walls may contain variable cellularity of disordered and dedifferentiated smooth muscle cells, fibroblasts, macrophages, leukocytes, mast cells, deficient collagen, and a hyaline-like proteinaceous substance, with the composition progressively abnormal from neck to dome.60,347,365 Alterations in levels or activity of numerous substances including elastin, fibronectin, tumor necrosis factor (TNF)-α, matrix metalloproteinases, oxidative stress reactants, and other proinflammatory isolates have been reported.60,347

Location

100 mm

FIG 12-9  CTA demonstrating a distal posterior cerebral artery territory mycotic pseudoaneurysm (arrow) with adjacent parenchymal hematoma in a patient with mitral valve vegetations.

A

Most cerebral aneurysms arise from the circle of Willis and middle cerebral artery bifurcations; 90% involve the anterior circulation and 10% the posterior circulation.12,304 The most common sites are the anterior communicating artery at its junction with the adjacent anterior cerebral artery (30%-35%), the posterior communicating artery at its union with the internal carotid artery (30%-35%), the middle cerebral artery bifurcation (20%-35%), the basilar terminus (5%), the carotid terminus (5%), and the origin of the ophthalmic artery (5%).140,307,312,417 The posterior communicating location is typically categorized with the posterior circulation. Less commonly, aneurysms occur along the course of the anterior cerebral arteries (particularly at the origin of the callosomarginal or pericallosal arteries), middle cerebral artery M1 segment and its branches, posterior cerebral arteries, superior cerebellar arteries, anterior inferior cerebellar arteries, and posterior inferior cerebellar arteries.17,153,169,333,348 Cavernous carotid aneurysms may grow to large size, possibly because of rupture prevention by the cavernous sinus dura.446 Multiple aneurysms are found in 15% to 30% of patients, more often in women than in men by a 5 : 1 ratio,46,185,196,395,422,446 and most

B FIG 12-10  Axial image of a CTA (A) revealing a small left supraclinoid carotid segment aneurysm (black arrow) that had produced subarachnoid hemorrhage in a patient with previous radiation therapy for a suprasellar mass as a child. Note the right craniotomy and right suprasellar subarachnoid catheter (white arrow). B, 3D reconstruction image from a rotational left carotid catheter cerebral angiogram confirms the aneurysm (arrow).

http://radiologyebook.com

352

PART II  CT and MR Imaging of the Whole Body

B

A

FIG. 12-11  Coronal reconstructed CTA image (A) of an intracranial dissecting right vertebral pseudoaneurysm (arrow). The right vertebral catheter cerebral angiogram, lateral oblique view (B), further evaluates the dissecting right vertebral pseudoaneurysm (arrow) with adjacent proximal and distal arterial stenosis.

frequently involve the middle cerebral artery.457 Of the patients with multiple cerebral aneurysms, 75% have two, 15% have three, and 10% have four or more.395 Bilateral symmetric aneurysms are called mirror aneurysms, and they most often involve the internal carotid arteries or middle cerebral artery bifurcations395 (Fig. 12-14). When one of multiple aneurysms ruptures and causes subarachnoid hemorrhage, it is most often the largest,457 although this is not always the case. For instance, Nehls and colleagues283 have reported the propensity for anterior communicating aneurysms to rupture when several are present simultaneously. Vasculopathies such as fibromuscular dysplasia and connective tissue disorders, as well as syndromes such as polycystic kidney disease, are associated with a higher incidence of multiple cerebral aneurysms.122 Familial aneurysms tend to be larger, multiple, rupture at a younger age and smaller size, and more frequently occur at the middle cerebral artery bifurcation.140,345

IMAGING AND CLINICAL EVALUATION Imaging objectives for patients with a cerebral aneurysm are (1) visualization of subarachnoid hemorrhage if present; (2) confirmation of the size, location, and morphology of the aneurysm; (3) evaluation of the adjacent cerebral vasculature, including evidence of spasm, atherosclerosis, vascular displacement, and incorporation of arterial branches into the aneurysm’s wall; and (4) demonstration of any accompanying adjacent brain pathology. The typical patient with a ruptured cerebral aneurysm is a woman between age 40 and 60 years who presents with “the worst headache of my life,” neck pain (meningismus), nausea/vomiting, and photophobia, with or without a neurologic deficit. Clinical assessment  classifies the patient as Hunt and Hess grades I through V (see Table 12-1)164 or World Federation of Neurological Surgeons grades 1 through 5 (see Table 12-2).331 CT scan of the head identifies subarachnoid hemorrhage, and the patient next typically undergoes CTA, which is used to triage the patient to surgery for clipping or endovascular coiling plus dictating other management.199,200 In patients in whom the CT scan shows no hemorrhage, lumbar puncture should typically be

performed. The presence of xanthochromia is the sine qua non to confirm subarachnoid hemorrhage; without it, any amount of red blood cells in the cerebrospinal fluid (CSF) may be unreliable because of the possibility that the lumbar puncture itself is the source of the hemorrhage.46 Xanthochromia may take 12 hours to manifest.397 In some situations, MRI and MRA may be performed to evaluate the subarachnoid hemorrhage and to identify a cerebral aneurysm. These examinations can be especially important in the patient with impaired renal function, thereby avoiding iodinated or gadolinium (Gd)-containing intravenous (IV) contrast. MRI without and with contrast may be beneficial to determine if there is an unusual cause for subarachnoid hemorrhage such as hemorrhagic primary or metastatic malignancies. These noninvasive MRI tests may be used routinely when no aneurysm is evident on CTA in the patient with subarachnoid hemorrhage; reserved for cases in which the clinical suspicion for a cerebral aneurysm is low; obtained to better evaluate brain parenchyma status, cranial nerve compression, or hemorrhage age and location; or performed in the patient who is a poor candidate for cerebral angiography or aneurysm intervention (i.e., elderly, extremely poor grade, or poor operative or endovascular candidate because of severe medical comorbidities). MRI and MRA may be contraindicated in patients with implanted ferromagnetic lines or devices such as pacemakers, or in patients who would require sedation because of claustrophobia or uncooperative clinical status when concurrent clinical evaluation is considered paramount.323 Patient motion may significantly degrade MRI quality. MRA or CTA may also be employed as screening tests in populations known to have a higher incidence of cerebral aneurysms46,286,446 (see preceding discussion), for monitoring known untreated aneurysms,181 and to reveal intracranial vascular lesions in patients with headaches.7 MRA may be preferred as a screening noninvasive imaging study for patients with a low suspicion of a cerebral aneurysm, in young patients, and in patients who require frequent periodic aneurysm evaluation.326 Better spatial resolution, inherent in CTA, may be reserved for patients with aneurysms less than 3 mm in diameter and when planning intervention at the expense of x-ray radiation.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C

D FIG 12-12  A, Axial T2-weighted image demonstrates a mass with peripheral hemosiderin in the right temporal lobe uncus (arrow). B, Reconstructed collapsed MRA view does not depict the abnormality in the medial right temporal lobe well (arrow), although the source MRA image (C) does (arrow). D, Lateral view of the catheter right carotid cerebral angiogram documents the traumatic right anterior choroidal pseudoaneurysm (arrow). The pseudoaneurysm was absent on previous catheter cerebral angiography. Slow flow within the pseudoaneurysm accounted for its appearance on MRA.

http://radiologyebook.com

353

354

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-13  A, Axial CTA image of a ruptured right supraclinoid carotid blister aneurysm (arrow) in a patient with a previously surgically treated pituitary adenoma. B, AP oblique view illustrating the blister aneurysm (arrow) during Pipeline (Covidien Corp.) endovascular treatment.

FIG 12-14  Rotated collapsed basal view of a noncontrast MRA illustrating small mirror middle cerebral aneurysms (arrows).

Computed Tomography

Acute Subarachnoid Hemorrhage.  The most important role for CT in the patient with a cerebral aneurysm is in identification of acute subarachnoid hemorrhage, which is demonstrated as increased density within a cisternal space. The location of the subarachnoid hemorrhage may help determine the location of the underlying cerebral aneurysm1,348,370,395 (Fig. 12-15). For example, blood in the sylvian fissure

most likely has arisen from a middle cerebral aneurysm, and blood in the fourth ventricle is a common sign of rupture of a posterior inferior cerebellar artery aneurysm.348 Some subarachnoid hemorrhage patterns are not diagnostic of aneurysm location, and none is foolproof. The CT demonstration of subarachnoid hemorrhage progressively decreases until it disappears by approximately 3 weeks,141,348 depending on the initial amount of subarachnoid hemorrhage. Subarachnoid hemorrhage is not identified with CT acutely in only a small percentage of cases but may be missed in up to half of patients by 7 days owing to clearance.23,80,140,199,397 A small amount of aneurysmal subarachnoid hemorrhage anterior to the brainstem may mimic perimesencephalic (prepontine and interpeduncular) hemorrhage, a finding believed to be due to the rupture of small perimesencephalic veins189,340 or a perforating artery, or secondary to basilar artery dissection (Fig. 12-16).367 Perimesencephalic hemorrhage accounts for as much as 10% of spontaneous subarachnoid hemorrhage and usually results in a good outcome, unlike an aneurysm rupture.199,339 Conversely, the chance of a ruptured aneurysm producing a perimesencephalic pattern of subarachnoid hemorrhage is 2.5% to 8.9%.199 CT is also excellent for identifying hydrocephalus (which develops in 15%-20% of patients after subarachnoid hemorrhage)46,397 (Fig. 12-17; see Fig. 12-19), intraventricular hemorrhage (13%-28% of cases) (Fig. 12-18),348 parenchymal hematoma (20%-30%) (see Fig. 12-9), and the occasional subdural hematoma. Brain parenchymal hemorrhage develops from rupture of the cerebral aneurysm into adjacent brain substance after the blood passes through overlying arachnoid and pia. Both parenchymal and subarachnoid hemorrhages may obscure an underlying ruptured cerebral aneurysm on CT or MRI.

Cerebral Aneurysm.  CT examination is secondarily helpful for identifying cerebral aneurysms typically 5 mm or larger in diameter. The rate of identification of cerebral aneurysms by thin-section, highresolution, contrast-enhanced CT scanning has been reported to be at

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

355

B

A

C FIG 12-15  A, CT scan illustrating acute subarachnoid hemorrhage (arrow) centered in the expected location of a right posterior communicating aneurysm, confirmed on two contiguous sagittal reconstructed CTA images (B and C, arrows).

least 67% for aneurysms 3 to 5 mm in diameter and to approach 100% for larger aneurysms.363,368 Focal, slightly dense areas of luminal blood in typical locations where aneurysms arise463 (Fig. 12-19), globular or elongated areas of contrast enhancement284 or a similarly shaped filling defect within a collection of blood (Fig. 12-20), variable calcification involving the aneurysm’s walls, and clot within larger aneurysms characterize cerebral aneurysms.143 Larger cerebral aneurysms are typically more easily identifiable because of their conspicuous size, but they are also often identified by the presence of internal clot and peripheral wall calcifications, especially in giant aneurysms (>25 mm in diameter) (Fig. 12-21). Giant aneurysms are most commonly identified in middle-aged women; are typically located in the extradural carotid artery, the middle cerebral artery bifurcation, or the basilar summit; and usually manifest with mass effect but can present as subarachnoid hemorrhage.52,196 Any large aneurysm can produce mass effect on adjacent brain or cranial nerves (see Fig. 12-19). CT is excellent for identifying skull base erosion due to large or giant aneurysms, especially if thin sections (≤3 mm) are obtained.

Computed Tomography Angiography CTA using multirow CT detectors is typically the first study performed to identify a cerebral aneurysm. CTA is excellent for evaluating the morphology, location, and orientation of a cerebral aneurysm, its relationship to adjacent osseous and brain parenchymal structures, and revealing perforating or branch arteries arising next to or from the dome of the aneurysm (Fig. 12-22). After a variable delay based on bolus tracking, thin-section spiral CT slices of 1 mm or less with less than 10% slice gap are acquired during rapid IV infusion of 40 to 70 mL of a high iodine–containing contrast agent for 4 to 5 mL/sec, with kVp (peak kilovoltage) and mAs (milliamperes per second)  settings of 80 to 120 and 150 to 400, respectively.201,243,436 Although typical 16-, 32-, or 64-row-detector CT scanners are commonplace, more recent commercial offerings are 256- and 320-row-detector machines.151,440 The acquired axial images and the reconstructed two- and three-dimensional (2D and 3D) images created from any  of a variety of reconstruction methodologies (e.g., multiplanar

http://radiologyebook.com

356

PART II  CT and MR Imaging of the Whole Body

FIG 12-16  CT scan of a patient with benign perimesencephalic hemorrhage (arrow) and no cerebral aneurysm documented on catheter cerebral angiography.

FIG 12-17  CT scan illustrating acute subarachnoid hemorrhage (black arrow) from a ruptured anterior communicating aneurysm presenting a star appearance in the basilar cisterns (same patient as in Fig. 12-1) and temporal horn enlargement indicating early communicating hydrocephalus (white arrow).

LCCA

A

B FIG 12-18  A, CT image demonstrating intraventricular hemorrhage as well as subarachnoid hemorrhage in a patient with a ruptured left anterior communicating aneurysm. B, The aneurysm is demonstrated on the oblique AP view of the catheter left carotid cerebral angiogram (arrow).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B FIG 12-19  A, Axial noncontrast CT scan illustrates a large, mildly hyperdense mass (black arrow) at the level of the third ventricle, causing obstruction of the bilateral foramina of Monro and resulting in obstructive hydrocephalus with bilateral frontal horn and bilateral temporal horn enlargement. Note the periventricular transependymal CSF flow adjacent to both frontal horns (white arrows). B, Coronal (turbo-spin echo) T2-weighted image from a 3D data set demonstrates that this is a giant left superior hypophyseal aneurysm causing the obstructive hydrocephalus. Note the flow-generated phase-encoding artifact in the aneurysm and adjacent brain (short arrows) as well as the bilateral frontal periventricular CSF flow (long arrows).

A

B FIG 12-20  A, CT image depicting a left posterior communicating aneurysm (arrow) set in negative relief by surrounding subarachnoid hemorrhage. B, Lateral view of the accompanying left carotid cerebral angiogram illustrating the left posterior communicating aneurysm adjacent to the left posterior cerebral artery having a fetal origin from the left internal carotid artery (arrow).

http://radiologyebook.com

357

358

PART II  CT and MR Imaging of the Whole Body

0

A

0

12

12

B

C FIG 12-21  CT examination demonstrating calcification and old thrombus within a giant anterior communicating aneurysm. A, CT scan without contrast documenting calcification within the wall of the aneurysm wall (arrow). B, CT study with contrast demonstrating the contrast-enhancing patent lumen (short arrow) and the majority of the aneurysm filled with thrombus (long arrow). C, AP view of the right carotid angiogram illustrating the patent portion of the anterior communicating aneurysm (arrow) that is much smaller than its total volume.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

359

B

A

FIG 12-22  CTA of a basilar summit aneurysm (arrow). A, Sagittal reconstruction. B, Coronal reconstruction.

reformation, maximum intensity projection [MIP], shaded surface display, ray-sum projection, direct volume rendering) reproduce the cerebral vasculature for scutiny.140,282,297,430 Both the source images and the reconstructed images should be reviewed because sometimes one will be superior to the other.469 The exclusion of adjacent skull base bone and the contrast enhancement of nonarterial structures can be problematic297,355,368; however, bone removal algorithms and improved contrast bolus timing have been used to mitigate these issues.3,355 Electrocardiographic gating of 3D CTA (4D CTA) permits visualization of aneurysm pulsation and calculation of volume change, which may predict unruptured aneurysm growth.108,150 3D CTA has a reported sensitivity of 77% to 100% for identifying cerebral aneurysms.46,58,140,159,178,199,233,236,264,297,311,325,327,427,429,432,447,469 Aneurysms measuring 5 mm or less in diameter may be missed.296,325,465 The sensitivity of CTA for identifying aneurysms 3 to 5 mm or less in diameter is 61% to 93%, significantly better than MRA for this size of aneurysm.285,448 CTA has some benefits over MRA, such as for evaluation of calcified atherosclerotic disease of adjacent vessels or in the dome of the aneurysm, although calcification can be suppressed on reconstructed images. CTA source images can be superior to MRA in some instances for visualization of the aneurysm neck or dome and their relationship to osseous skull base structures such as the anterior clinoid process and optic strut, helping to determine whether the aneurysm neck is intradural or within the cavernous sinus.130,147,410 CTA has been recommended as sufficient for aneurysm treatment planning by several centers.58,94,151,159,311,432,440 Its advantages include identification of aneurysm dome and parent vessel calcification that influence surgical clipping, rapid acquisition and reconstruction speed, a multitude of processing and reconstruction algorithms including bone subtraction, multiple viewing angles, image manipulations, and relative noninvasiveness.68,410,440 Software reconstruction enhancements for CTA (as well as MRA) can provide endoscopic views of the dome and neck of an aneurysm.256 Disadvantages include radiation exposure (although efforts have been made to decrease patient dose),436 difficulty differentiating small aneurysms from infundibula or permitting their identification in regions of complex vascular anatomy (see Fig. 12-4), obscuration by adjacent bone, the “kissing vessel” artifact (creating a false-positive aneurysm), variably visualizing fenestrations and small arteries, and potential contrast complications such as allergic reaction and contrast nephropathy.199,325,410 CTA using MIP reconstructions can have difficulty separating the contrast-enhancing patent portion of the aneurysm lumen from intraluminal and perianeurysmal

clot, and it can miss some small cerebral aneurysms that are detectable only with catheter cerebral angiography.194,410 CTA has been compared to both catheter cerebral angiography and MRA for aneurysm identification and characterization, including mural thrombus detection with mixed conclusions.96,236,245,327,432 It has been suggested that CTA may be sufficient (with a 100% negative predictive value) to exclude a cerebral aneurysm in patients with a CT scan negative for subarachnoid hemorrhage, those with perimesencephalic subarachnoid hemorrhage pattern, or those with CSF xanthochromia or bloody CSF only, especially considering lumbar puncture complications.96,190,199,236,245,327,385

Cerebral Vasospasm and Delayed Cerebral Ischemia Due to Subarachnoid Hemorrhage.  Catheter cerebral angiographic vasospasm is identified in as many as 70% of aneurysmal subarachnoid hemorrhage patients.397 The development of vasospasm is related to the amount of subarachnoid blood, with predictive grades constructed by several authors.75,109,156,397 Approximately 25% to 45% of patients with aneurysmal subarachnoid hemorrhage develop symptomatic cerebral vasospasm, with as many as 30% developing delayed cerebral ischemia.100,397 Approximately 50% of patients with large vessel vasospasm are symptomatic.80 The risk is greater in younger patients.188 Roughly 35% to 55% recover and 25% to 30% die.100 Transcranial Doppler evaluation has limited sensitivity and specificity for cerebral vasospasm.431 CTA is beneficial in the noninvasive determination of large proximal-vessel cerebral vasospasm, with reported sensitivity of 77% to 87% and specificity of 93% to 99% compared with catheter cerebral angiography32,134; as a screening examination it can be used to avoid the more invasive examination.63 Although CTA’s sensitivity and specificity for large proximal arterial vasospasm that is clinically symptomatic is over 90% (Fig. 12-23), catheter cerebral angiography remains necessary when CTA is negative, and also for vasospasm treatment with angioplasty or intraarterial infusion of vasodilation agents.373,123 CT perfusion performed by rapid CT imaging during IV bolus contrast administration can determine when cerebral vasospasm compromises tissue perfusion with a sensitivity of 74%, a specificity of 93%, and high positive and negative predictive values.134,202 Scan parameters are typically 5 mm thick with minimal slice gap acquired during an IV infusion of a high iodine–containing contrast agent at a rate of 7 mL/sec using an mAs of 100 and kVp of 80. Diminished cerebral blood flow and increased mean transit time or time to peak

http://radiologyebook.com

360

PART II  CT and MR Imaging of the Whole Body parameters are particularly helpful91,155,358,359 and can be predictive of delayed cerebral ischemia, infarction, and permanent neurologic deficits91,155,193,202,270,358 (Fig. 12-24). CT perfusion is superior to transcranial Doppler to predict the need for endovascular treatment of cerebral vasospasm, and its combination with CTA can accurately diagnose cerebral vasospasm.456 Neither severe vasospasm nor decreased CT perfusion always lead to delayed cerebral ischemia, and patients without spastic cerebral arteries or abnormal cerebral perfusion can go on to infarction.80,86,92 MRA does not appear to be reliable for imaging cerebral vasospasm.149,405

Magnetic Resonance Imaging

Acute Subarachnoid Hemorrhage.  The identification of acute

FIG 12-23  CTA demonstrating bilateral middle cerebral artery vasospasm (arrows) in a patient with recent subarachnoid hemorrhage.

subarachnoid hemorrhage by routine MRI has been notoriously poor,13,295 although there are reports that claim otherwise.180 The following two reasons have been postulated for the lack of visualization of acute subarachnoid hemorrhage on MRI137: 1. Normal Po2 (partial pressure of oxygen) in CSF inhibits the development of deoxyhemoglobin. 2. Lysis of red blood cells in subarachnoid CSF prevents heterogeneous magnetic susceptibility. However, with the employment of pulse sequences including intermediate-weighted spin echo or fast-spin echo with TE (echo time) values of 22 to 45 milliseconds, T1-weighted spin echo, FLAIR

MTT

A

(sec)

CBF

B

(ml/100g/min)

FIG 12-24  CT perfusion scan in a patient with subarachnoid hemorrhage and symptomatic vasospasm demonstrating (A) elevated mean transit time (MTT) and (B) diminished cerebral blood flow (CBF) in the watershed regions between the bilateral anterior cerebral, middle cerebral, and posterior cerebral territories creating a cerebral “X”.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations AH

361

AH

A

B AH

C FIG 12-25  Acute subarachnoid hemorrhage within the prepontine cistern (anterior arrow) and acute intraventricular hemorrhage in the fourth ventricle (oblique arrow) on T1-weighted (A), T2-weighted (B), and T2* gradient echo (C) axial MRIs.

(fluid-attenuated inversion recovery) imaging, gradient echo imaging, and susceptibility-weighted imaging (SWI), acute subarachnoid hemorrhage can frequently be identified (Figs. 12-25 and 12-26). Other imaging problems can obscure precise identification of subarachnoid hemorrhage on MRI, including CSF and vascular pulsation artifact

and magnetic susceptibility artifact along the skull base. With experience, acute subarachnoid hemorrhage often can be correctly identified on MRI. Subarachnoid hemorrhage older than 12 to 24 hours can routinely be identified on MRIs, often helping pinpoint the underlying ruptured cerebral aneurysm by a collection of blood.

http://radiologyebook.com

362

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 12-26  T2-weighted FLAIR (A), SWI (B), and CT scan (C) demonstrating acute subarachnoid hemorrhage (black arrows) in a patient with a ruptured right posterior communicating aneurysm. The white arrow in B illustrates the sensitivity of SWI to intraventricular hemorrhage.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

363

B

A

0

S

C

D FIG 12-27  MRI and MRA examination demonstrating flow effects associated with a giant left cavernous carotid aneurysm. A, Axial short-TE, long-TR MRI demonstrating the aneurysm (long arrow), with internal and external phase-encoding artifact (short arrows) resulting from mismapped flow. B, Sagittal T1-weighted image falsely gives the appearance of thrombus within the aneurysm (arrow) owing to slow flow. C, 3D time-of-flight (TOF) MRA poorly outlines the aneurysm (arrow) because of internal slow flow. D, AP view of left carotid angiogram confirms giant left cavernous carotid aneurysm (arrow). (Case courtesy Daryl Harp, MD.)

Cerebral Aneurysm.  MRI is superior to CT for the localization of cerebral aneurysms, their relationship with adjacent structures, and associated changes in neighboring brain tissue (see Fig. 12-19). Thin-section MRI examination can often identify cerebral aneurysms over 3 mm in diameter267; identification may be easier if the aneurysm is unruptured. It may be difficult, however, to differentiate normal vascular tortuous anatomy or bone such as the anterior clinoid process from a cerebral aneurysm. When large enough, cerebral aneurysms can be positively identified on MRI because of flow artifacts or characteristic findings with the use of different pulse sequences. Phaseencoding artifact due to flow-producing signal mismapped in the phase-encoding direction can be visualized on a variety of pulse sequences (Fig. 12-27; see Fig. 12-19). Other flow effects are also often

helpful to identify cerebral aneurysms. These include a swirling appearance inside the aneurysm due to spin dephasing, rephasing, and turbulence; even-echo or gradient rephasing; diastolic pseudogating; slow flow; variable intraluminal flow rates; entry slice enhancement; and signal void due to rapidly flowing blood visualized on most pulse sequences (Figs. 12-28 and 12-29).15,342,435 Contrast enhancement of the aneurysm highlights its central lumen and can exaggerate the aforementioned artifacts, particularly the phase mismapping effects (see Fig. 12-77).408 For instance, these clues can often reliably diagnose cavernous-carotid segment aneurysms and differentiate them from pituitary masses. A partially thrombosed aneurysm contains peripheral layers of clot of different ages with an onion-like appearance, usually with the oldest

http://radiologyebook.com

364

PART II  CT and MR Imaging of the Whole Body

FIG 12-28  Axial T2-weighted MRI demonstrating chronic left middle cerebral artery territory infarctions (black arrows) in a patient with bilateral middle cerebral (mirror aneurysms) and basilar summit aneurysms (white arrows). Same case as illustrated in Figure 12-29.

clot located most peripherally.16,143,408 MRI may identify intramural thrombus in dissecting pseudoaneurysms and fusiform aneurysms (Figs. 12-30 and 12-31). The signal intensities of intraluminal clot correspond to the expected age of hemorrhage on MRI according to the magnetic field strength employed.16,143 Fibrosis and cellular infiltration can accompany the clot, typically reducing signal intensity, and perianeurysmal edema may be present to a variable extent. Completely thrombosed aneurysms or aneurysms with extremely slow flow can rarely mimic tumors, including sellar and parasellar masses (see Fig. 12-27). Catheter cerebral angiography may be necessary in these problematic cases to prevent “surprises” during craniotomy and potentially catastrophic hemorrhage during transsphenoidal or other operations. Brain parenchymal changes adjacent to cerebral aneurysms may be absent unless the aneurysm is large enough to cause edema and localized mass effect (see Fig. 12-31). MRI is superior for evaluating the relationship of the aneurysm to adjacent brain structures because of inherent soft tissue contrast limitations and beam-hardening artifact seen with CT. If the aneurysm has ruptured, an adjacent intraparenchymal, subarachnoid, or subdural clot may be identified. This finding is useful for determining which aneurysm has ruptured in a patient with multiple aneurysms, especially when clinical localizing signs are not helpful.141,283,395 The imaging location of the clot can subsequently guide treatment.141 Chronic changes include reactive astrocytosis (gliosis) due to infarction from prior vasospasm or aneurysm clot emboli (see Fig. 12-28) and deposition of hemosiderin or ferritin (see Fig. 12-29); the latter two substances represent hemorrhagic breakdown products, which are best demonstrated on long-TE, long-TR, T2-weighted spin echo or fast-spin echo, T2* gradient echo sequences, or SWI. Superficial siderosis represents intracellular (most often macrophage) and extracellular hemosiderin or ferritin coating the leptomeninges, cranial nerves, and superficial central nervous system parenchyma.126 The siderosis is secondary to any cause of subarachnoid hemorrhage, usually repetitive, that can also occur on the ventricular ependymal surfaces.126 Focal or generalized neurologic deficits correspond to the location and extent of central nervous system involvement by siderosis.126

Magnetic Resonance Angiography

FIG 12-29  Flow artifact (long arrow) within a left middle cerebral aneurysm on an axial T2* gradient echo MRI. Note the left temporal lobe cortical hemosiderin (short arrow) in a chronic infarction. Same case as illustrated in Figure 12-28.

Identification of cerebral aneurysms using MRA has a reported  sensitivity of 69% to 97% and a specificity of 89% to  93%46,67,269,326,344,352,360,369,448 (Fig. 12-32). MRA can be used alone for presurgical planning,67,360,441 although some authors have claimed its inferiority to catheter cerebral angiography, especially for depiction of aneurysms less than 3 mm in diameter4 (Fig. 12-33). Sensitivity of MRA for detection of aneurysms can be increased when MRA is combined with MRI,344,369 thereby overcoming the inherent difficulties of imaging slow flow, turbulent flow, other flow artifacts, proteinaceous fluid, or intraluminal clot with MRA140 (see Fig. 12-27). Because almost all aneurysms larger than 3 mm can be visualized,211,269 MRA can function as a screening examination.326 The sensitivity of MRA for detecting aneurysms 3 mm or smaller decreases to 38% to 74%.269,448 The most commonly employed sequence is time-of-flight (TOF) 3D MRA, which typically provides submillimeter in-plane resolution. A variant of this is MOTSA (multiple overlapping thin-slab acquisition),36,37,314 which attempts to combine the advantages of 3D with 2D MRA. MOTSA is especially helpful when the slow flow in an aneurysm located distally within the imaging slab is not imaged on a 3D TOF study because of flow saturation (Fig. 12-34). The T1 shortening effect of subacute to chronic subarachnoid blood or intraluminal or perianeurysmal clot can obscure an aneurysm or mimic flowing blood on a TOF imaging sequence,166 although MRA source images can be

http://radiologyebook.com

Text continued on p. 369

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C

D FIG 12-30  Acute thrombus within a dissecting right posterior communicating pseudoaneurysm shown on (A) axial T2-weighted MRI (arrow), axial (B) (arrow), and sagittal (C) (arrow) T1-weighted MRI, with a patent lumen located posteriorly and inferiorly on the sagittal study and axial CTA scan (D) (arrow). Continued

http://radiologyebook.com

365

366

PART II  CT and MR Imaging of the Whole Body

E

F

G FIG 12-30, cont’d  More cephalad CTA image (E) confirms the thrombosed component of the pseudoaneurysm (arrow). The patient presented with headache without subarachnoid hemorrhage and was treated with aspirin. F and G, Two sequential lateral views of the catheter right carotid cerebral angiogram acquired 1 month later illustrate the thrombus has resorbed and the pseudoaneurysm is now completely patent (arrows). Note the swirling pattern of flow within the pseudoaneurysm in F.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

C

B

D FIG 12-31  A, Reconstructed noncontrast MRA of a stented right posterior inferior cerebellar artery (PICA) dissecting pseudoaneurysm. Subacute thrombosis of the majority of the pseudoaneurysm results in increased signal intensity on the TOF MRA (long arrow); however, the proximal and distal portions of the right PICA are patent with good flow (short arrows). B, Axial source image of the MRA documents patency of the right PICA, with peripheral diminished signal intensity caused by the Neuroform microstent (Boston Scientific Corp.) (arrow), and no flow in the pseudoaneurysm. Axial noncontrast T1-weighted (C) and axial T2-weighted (D) MRIs confirming patency of the right PICA (short arrow) and both subacute and acute thrombus in the right PICA dissecting pseudoaneurysm (long arrow). Note the edema in the medulla on the T2-weighted image. Continued

http://radiologyebook.com

367

368

PART II  CT and MR Imaging of the Whole Body

E

F

FIG 12-31, cont’d  E and F, Early and late arterial phase oblique lateral images of the right vertebral 3D rotational angiogram immediately after endovascular treatment, illustrating the proximal and distal markers of the Neuroform microstent (arrows in E) and the patent portion of the pseudoaneurysm that went on to thrombose (arrow in F).

FIG 12-32  MRA of a basilar summit aneurysm (arrow).

FIG 12-33  MRA of a small left middle cerebral aneurysm (arrow). (Case courtesy David Watts, MD.)

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations helpful in differentiating the patent lumen from mural thrombus4 (see Fig. 12-31). An alternative pulse sequence is phase-contrast (PC) MRA, which has excellent background, susceptibility artifact, and stagnant blood suppression and may help image in-plane flow and slow flow, two causes of flow saturation degradation of TOF MRA.101,165 It can be combined with cardiac gating using electrocardiogram leads or peripheral photoplethysmography to create cine-PC MRA, a technique that has been used to document a greater pulsatility change in aneurysm size with ruptured than with unruptured aneurysms. PC MRA has poorer spatial resolution, depends on a prescribed velocity-encoding gradient with the potential for aliasing, can be altered by complex flow, and requires longer imaging and processing times than TOF MRA.165 Although PC MRA may identify fewer aneurysms than TOF MRA,182,266 the slow flow within fusiform aneurysms may be better imaged with PC MRA. A patent double lumen in a dissecting aneurysm might also be best imaged with PC MRA. MRA combined with MRI can often demonstrate internal flow patterns within large and giant cerebral aneurysms; faster flow is located more peripherally and stagnant flow more centrally.129,396 TOF MRA often inadequately visualizes large and giant aneurysms (see Figs. 12-27 and 12-34).175 Gd-based IV contrast agents—although not routinely employed in the identification of cerebral aneurysms—can be helpful to highlight aneurysms that are small, are present in atypical locations, or have slow or stagnant flow, or to document the residual patent lumen in an incompletely thrombosed or giant aneurysm,175,342,401,408 consequently making contrast-enhanced MRA in some instances superior to noncontrast TOF MRA401 (see Fig. 12-34). The use of contrast enhancement with MRI or MRA increases flowdependent artifact and can obscure cerebral aneurysms because of enhancement of dural sinuses, intracranial veins, and extracranial structures85 unless a PC MRA technique is used.165 Artifacts can also be reduced using bolus timing and centric k-space acquisition to improve aneurysm detection74,175,266 or by employing a precontrast subtraction technique, a process that has been reported to improve detection of ruptured cerebral aneurysms compared with nonenhanced MRA.420 In large or giant partially or completely thrombosed aneurysms, the aneurysmal wall may occasionally show enhancement. Special MRA techniques may help in the positive identification of cerebral aneurysms; they include magnetization transfer pulses,105,266 phase cycling,266 progressively increasing radiofrequency pulse flip angles,266 and multiple postprocessing algorithms including volume rendering.4,18,21,67,163,236,250,413 It is important that both the 3D reconstructed MRA images and the source 2D planar images are scrutinized so that small aneurysms and aneurysms with slow flow are not missed.4 Standard MIP reconstructions have some artifact-derived limitations and may be unreliable4,5,250 (see Figs. 12-27 and 12-34). Targeted MIPs select a region of interest and help exclude confounding extraneous vessels and other structures including bone.165 Vascular loops, fenestrations, infundibula, and artifact may be confused with cerebral aneurysms on MRA.352,404 Review of the source images also may best demonstrate the neck of the aneurysm and its relationship to the parent vessel. Vessel loops and magnetic susceptibility artifact from air-filled sinuses or bone at the skull base may obscure aneurysms352; however, MRA source images may be invaluable in determining whether any part of an aneurysm is in the subarachnoid space or is extradural, such as in the cavernous sinus (Fig. 12-35). The availability of 3-tesla (T) MRI units has produced an improved signal-to-noise ratio that yields better quality and superior detection of cerebral aneurysms using high-field MRA, with or without the use of IV contrast.67,120,121,323 Acquisition time can be decreased, spin inflow signal is more robust, and some flow-based artifacts may be reduced,

369

but metallic and diamagnetic artifacts can be accentuated. MRA at 3 T is generally superior to 1.5-T MRA for aneurysm detection and evaluation, including analyzing large and giant aneurysms, aneurysm morphology, and adjacent or aneurysmal arterial branches.67,277,352 The spatial resolution for 3-T MRA can be 0.6 mm vs. approximately 1.0 mm at 1.5 T.140

Catheter Cerebral Angiography Catheter cerebral angiography remains the standard for evaluation of cerebral aneurysms. 3-T MRA is inferior to catheter cerebral angiography for aneurysm detection and when evaluating aneurysm dimensions, bleb formation, and arterial branches arising from the aneurysm dome.269,323 Catheter-based angiography has better spatial resolution (typically 0.2 mm) than MRA or CTA, but it evaluates only the intraluminal portion of the aneurysm that contains flowing blood. The inability of catheter cerebral angiography to demonstrate aneurysmal clot can lead to underestimation of the size of the aneurysm. This deficiency of catheter cerebral angiography is significant in comparison with MRI and MRA source images, which not only can demonstrate clot but also can show its age as well as the relationship of the aneurysm to adjacent brain structures.4 Cerebral angiography has the benefit, however, of better demonstrating cerebral vasospasm,  the aspects of the aneurysm’s morphology (e.g., dome irregularity or  lobulation, which are possible predictors of hemorrhage), and a focal tit or dimple indicating the site of recent aneurysmal hemorrhage.457 Aneurysms evaluated soon after rupture often are smaller than when intact because of clot compression and spasm of either the aneurysm or adjacent artery, but they subsequently enlarge if surgery is delayed.196,283,366,443 Catheter cerebral angiography fails to identify a source of subarachnoid hemorrhage in as many as 20% of patients.46 Repeat catheter cerebral angiography in 7 to 14 days, however, may identify a cerebral aneurysm in approximately 14% of initially negative studies in patients with aneurysmal subarachnoid hemorrhage.80 The use of 3D rotational catheter cerebral angiography is considered the best imaging evaluation of cerebral aneurysms, superior to 3D CTA, 3D MRA, and 2D catheter angiography140,285 (Fig. 12-36). 3D digital subtraction angiography (DSA) yields a spatial resolution slightly less than the 0.2 mm of 2D DSA, and better than the typical 0.4 mm to 1 mm of CTA and MRA, respectively.140 Not only does it provide the ability to obtain a multiplicity of views of the aneurysm, but it also permits manipulations of the reconstructed images on an independent work station. Catheter cerebral angiography is an invasive procedure whose risk must always be kept in mind. The incidence of neurologic events during cerebral digital subtraction angiography ranges from 0.1% to 2.6%.140,198,199 Other risks include femoral or retroperitoneal hematoma, vascular injury, contrast-induced renal injury, allergic reaction, and radiation exposure.46

EVALUATION AFTER OPEN AND ENDOVASCULAR SURGERY Most cerebral aneurysm clips are currently MRI compatible378 and safe in a magnetic field, but they create local ferromagnetic artifact on MRI/ MRA that obscures precise evaluation of the neck of the clipped aneurysm135 (Figs. 12-37 and 12-38). Issues have been raised regarding clip-to-clip variability, distortion of the aneurysm clip during surgical application, manufacturer liability, warnings from the U.S. Food and Drug Administration (FDA) regarding rotational and translational aneurysm clip motion in a magnetic field, and the necessity of institutional assurance that all aneurysm clips be MRI compatible.192 Many facilities have denied MRI to patients with cerebral aneurysm clips,

http://radiologyebook.com

370

PART II  CT and MR Imaging of the Whole Body

100 mm

A

B

100 mm

C

100 mm

D FIG 12-34  A, Reconstructed noncontrast TOF MRA using MOTSA (multiple overlapping thin-slab acquisition) demonstrating the interfaces of the three acquisition slabs (arrows). This technology results in greater signal intensity at the top of the aneurysm owing to maximal flow-related enhancement in the superior slab. Inferior (B), middle (C), and superior axial source (D) images illustrate that the majority of the aneurysm either is thrombosed or contains extremely slow flow, with rapid blood flow seen only along its medial wall and superiorly (arrow).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

371

E FIG 12-34, cont’d E, Contrast-enhanced MRA without MOTSA demonstrates that the entire aneurysm is patent and does not contain clot. Note the venous contamination with the arrow pointing out the right sigmoid sinus. Same patient as in Figure 12-10.

FIG 12-36  Reconstructed 3D rotational right carotid cerebral catheter angiogram illustrating a large cavernous carotid aneurysm (red arrow) and a small posterior communicating aneurysm (green arrow).

A

B FIG 12-35  A, Reconstructed MRA of a right superior hypophyseal aneurysm (arrow). B, The source image indicates that the aneurysm is exposed to the subarachnoid space (arrow).

http://radiologyebook.com

372

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-37  Ferromagnetic artifact produced by a cerebral aneurysm clip on MRI. Axial short-TE, long-TR (A) and coronal postcontrast T1-weighted (B) spin echo images demonstrate an ophthalmic aneurysm clip (arrow).

even though an MRI exam permits better evaluation of brain parenchyma than CT. Fatal cerebral hemorrhage attributed to the magnetic field during MRI has been reported in a patient in whom a stainless steel MRI-incompatible aneurysm clip had been placed on a middle cerebral artery aneurysm.206 Beam-hardening artifact from surgical clips significantly degrades CTA images, prohibiting accurate identification of a potential remnant aneurysm.410 Guglielmi and other detachable platinum coils used to treat cerebral aneurysms endovascularly are MRI compatible.376,377 Localized artifact caused by the platinum coils on MRI, MRA, or CTA often do not permit precise evaluation of the aneurysm, missing remnants or recurrences of 1.8 to 5 mm in diameter43,82,99,107,128,173,301,320,414,442 (Fig. 12-39). 3D TOF MRA employing a short TE of 1.54 to 1.6 milliseconds, a high spatial resolution of 0.3 mm3, volume acquisition positioning to maximize arterial inflow, and the use of 3 T can improve the detection of residual blood flow within coiled cerebral aneurysms.97,251,370 Contrast-enhanced MRA has been reported non-uniformly to be better at imaging aneurysm remnants or recurrences after endovas­ cular treatment than standard TOF MRA, in some instances superior to catheter cerebral angiography owing to the ability of cross-  sectional imaging to demonstrate internal patent aneurysm com­part ments5,10,19,40,231,251,62,80,88,140,279,370 (Fig. 12-40). Contrast-enhanced MRA lessens turbulent- or slow-flow intravascular signal degradation, in-flow saturation, and metallic artifact compromise but remains confounded by luminal thrombus and enhancement of thrombus and occasionally the wall of the aneurysm117,279 (Fig. 12-41; see Fig. 12-44). Whether MRA is sufficient to identify residual aneurysm patency after endovascular coiling that requires additional treatment has been debated in several published analyses.9,117,118,125,225,247,265,291,371,437 Although less metallic artifact is present in the endovascularly treated aneurysm patient imaged with MRA compared with CTA, CTA

is typically superior to MRA to evaluate the surgically clipped aneurysm and neighboring arteries438 (Fig. 12-42). Slice thickness less than 1 mm, 0.6-mm pitch, dual energy, dual volume, advanced software, and fusion packages may improve CTA evaluation of aneurysm interface conspicuity against clips, coils, and stents,125,170,379,438 although overcoming a multiplicity of these metallic devices and those containing cobalt can still be problematic.438 Clip orientation to the CT detector array can worsen or lessen metallic artifact.438 The use of intracranial stents or vascular reconstruction devices, including flow diverters, to assist in endovascular treatment of cerebral aneurysms can also be evaluated with CTA, MRA, or MRI5,19,40,70 (Fig. 12-43; see Fig. 12-31). MRA suffers from metallic artifact that produces or exaggerates in-stent stenosis, and furthermore the use of contrast can obscure coil localization, overestimate patent lumen dimensions, and create artifacts5,225,231,279,321 (Fig. 12-44). In most cases, however, 3D reconstruction digital subtraction catheter angiography remains the best examination (including superior to 2D DSA) to determine whether the aneurysm has been completely excluded from the parent circulation (see  Fig. 12-41).19,225,322,372,438

CEREBROVASCULAR MALFORMATIONS Cerebrovascular malformations can be divided into four primary types45,242,346 (Fig. 12-45): • Arteriovenous malformation (AVM) • Cavernous malformation or cavernous angioma • Capillary telangiectasia • Developmental venous anomaly (venous angioma) These types of vascular malformations can also be mixed, incorporating elements of two or more types.22,103,131,305,336,337,403 The four classic types of cerebrovascular malformations are based on their congenital

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C FIG 12-38  Subtracted (A) and unsubtracted (B) oblique AP right carotid cerebral catheter angiogram demonstrating a well-clipped right superior hypophyseal aneurysm (arrows). MRA axial source image (C) of the same patient is degraded by clip metallic artifact (red arrow), and no information regarding relationship of the clip to the right carotid artery is available. Note the patent left internal carotid artery (blue arrow) and the right frontal mucocele (green arrow).

http://radiologyebook.com

373

374

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-39  A, Reconstructed coronal CTA of a large, coiled, right middle cerebral aneurysm (arrow) degraded by metallic artifact is unable to evaluate adjacent arteries or brain parenchyma. B, CT lateral localizer image illustrating the coils within the right middle cerebral aneurysm (arrow).

origin and do not include arteriovenous cerebral fistulas, which are typically acquired, rarely congenital, and represent a direct communication between artery and vein or are dural based. The term occult cerebrovascular malformations has been used to describe cavernous angiomas, capillary telangiectasias, and thrombosed AVMs because they could not be visualized with catheter cerebral angiography. Because all of these lesions can be demonstrated on MRI, they are no longer truly “occult,” and the term therefore has little current relevance. Sinus pericranii represents anomalous venous drainage through the skull. The overall incidence of vascular malformations of all types in the general population is estimated to be roughly 3%.

ARTERIOVENOUS MALFORMATIONS Incidence Approximately 0.1% of the general population harbor a cerebral AVM, the most common symptomatic cerebrovascular malformation.308,316 AVMs are most commonly identified in women in the second to fourth decades.450 There may be a mildly increased genetic familial occurrence even without a defined associated syndrome known to predispose to the incidence of cerebral AVMs, such as HHT.424,458 Although the concept that cerebral AVMs are congenital is entrenched, there are rare reports of documented de novo AVM development, raising the possibility that these are acquired vascular lesions.53,276 More than 80% of cerebral AVMs are located supratentorially and the remainder within the posterior fossa. AVMs may be multiple in up to 2% of patients,356 often as part of HHT or Osler-Weber-Rendu disease,2,258 an autosomal dominant inherited condition in which mucosal, skin, gastrointestinal, and cerebral HHT, brain cavernous malformations, and pulmonary AVMs are also found, and in the Wyburn-Mason syndrome (also termed mesencephalo-oculofacial angiomatosis), which is signified by the presence of cutaneous nevi/ angiomas and retinal AVMs or angiomas together with brain AVMs.274,329,451 Approximately 3% to 7.5% of HHT patients have cerebral AVMs that are typically small, cortical, and supratentorial.458 This

particular patient population more frequently experiences a transient ischemic attack (TIA) or stroke from a pulmonary AVM.246 AVMs may also be located extracranially, or intracranial AVMs may be associated with additional extracranial AVMs.

Pathology and Histology Pathologically the cerebral AVM represents a direct communication between arteries and veins without intervening capillaries.133,195,304 The nidus represents the central compact tangle of low-resistance vessels constituting the junction of the feeding arteries and draining veins. The nidus does not contain normal brain parenchyma, although the enlarged feeding arteries and draining veins can traverse normal tissue.54 Fistulas often reside within the AVM nidus. Large diffuse AVMs may contain normal neurons and white matter.69 Brain AVMs may be located in the subarachnoid space, intraparenchymally, or intraventricularly or may traverse several spaces, including the cranial vault. Larger AVMs are often located in a vascular watershed territory and assume a pyramidal shape extending from cortex to the lateral ventricle, with its apex directed deeply (see Figs. 12-48 and 12-50). The AVM nidus volume together with characteristics such as the presence of fistulas, chronicity of the lesion (i.e., the age of the patient), and rate of flow through the AVM determine the size and tortuosity of feeding arteries and draining veins. AVMs may be supplied by any of the cerebral vessels, including meningeal contributors in 15% to 50% of cases.271 Approximately 10% of AVMs have an arterial feeder, nidus, or remote arterial aneurysm,254,314,453 so-called flow-related aneurysms most often located on feeding vessels that can involute after AVM cure88,209,272,298,314 (see Fig. 12-6). Some of these flow-related aneurysms are pseudoaneurysms.315 The aneurysms contribute to the incidence of hemorrhage from AVMs.453 Varices can develop in draining veins and can become quite large. Areas of stenosis, thrombosis, or occlusion can occur in supplying arteries or exiting veins. The findings of vascular stenoses and aneurysm/varix formation are attributed to local hemodynamic stresses.315

http://radiologyebook.com

Text continued on p. 380

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

D

C FIG 12-40  Unenhanced (A) and contrast-enhanced (B) MRA of a recurrent coiled left posterior communicating aneurysm (arrows). AP (C) and lateral (D) views from the left carotid catheter cerebral angiogram reveal no aneurysm patency (arrows) located adjacent to a fetal posterior cerebral artery. This case supports tomographic imaging vs. planar imaging for demonstrating aneurysm recurrence.

http://radiologyebook.com

375

376

PART II  CT and MR Imaging of the Whole Body

A

C

B

D

E

FIG 12-41  Axial T2-weighted (A) and coronal postcontrast T1-weighted (B) MRIs demonstrate a coiled right superior hypophyseal aneurysm (arrow) causing localized metallic signal loss. Peripheral foci of increased signal intensity are seen as well on the postcontrast T1-weighted image. C, The axial source image of the TOF MRA illustrates a similar finding (arrow) with peripherally located arteries or aneurysm flow. D, Reconstructed MRA, oblique AP view, indicates that the supraclinoid internal carotid artery is narrowed (arrow) without aneurysm patency. E, The oblique AP right carotid cerebral angiogram documents that the aneurysm is completely occluded by the endovascular coils (arrow) without internal carotid artery narrowing.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C FIG 12-42  CTA of a residual/recurrent surgically clipped left anterior communicating aneurysm. A, Axial source image of the CTA examination demonstrating the residual/recurrent aneurysm (red arrow), the aneurysm clip (green arrow), and the craniotomy defect (white arrow). B, 3D reconstructed CTA oblique image profiling the residual/recurrent aneurysm (red arrow) and the aneurysm clip beneath it (green arrow). C, Oblique lateral view of the left carotid cerebral catheter angiogram confirming the residual/recurrent left anterior communicating aneurysm (red arrow) and the adjacent aneurysm clip (green arrow).

http://radiologyebook.com

377

378

PART II  CT and MR Imaging of the Whole Body

A

C

B

D

E FIG 12-43  A, Oblique AP view of a left carotid catheter cerebral angiogram illustrating a large left cavernous carotid aneurysm (arrow). B, Magnified oblique AP view during Pipeline (Covidien eV3 Neurovascular) (arrow) treatment producing contrast stasis within the aneurysm. C, Six-month lateral view of the diagnostic catheter angiogram reveals resolution of the left cavernous carotid aneurysm (arrow). Non–contrast-enhanced (D) and contrast-enhanced (E) axial source MRA images obtained at 3 T demonstrate patency of the flow diverter and no flow within the aneurysm (arrows).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

D C FIG 12-44  A, Unenhanced axial MRA source image demonstrating a right vertebral dissecting pseudoaneurysm (arrow) in a patient with Ehlers-Danlos syndrome (same patient as Fig. 12-8). After endovascular treatment with two coaxial Enterprise vascular reconstruction devices (Codman Neurovascular) an axial T2-weighted fast-spin echo image (B) cannot determine whether the pseudoaneurysm is thrombosed or patent (arrow). Noncontrast (C) and contrast-enhanced (D) MRA axial source images demonstrate contrast enhancement of the pseudoaneurysm and thrombus (arrows), but it appears the vascular reconstruction devices are occluded. Continued

http://radiologyebook.com

379

380

PART II  CT and MR Imaging of the Whole Body as a result of previous hemorrhage, which is also denoted by hemosiderin or ferritin deposition.55

Presentation

E FIG 12-44, cont’d  Reconstructed sagittal CTA image (E) reveals the vertebral artery and the vascular reconstruction devices are patent, and the pseudoaneurysm is occluded. Arrow in E points out the distal vascular reconstruction device markers.

FIG 12-45  Schematic illustrating the four basic types of cerebrovascular malformations. A cavernous malformation is seen in the frontal lobe, a true AVM is present in the cingulate gyrus with a pericallosal aneurysm, a capillary telangiectasia is located in the brainstem at the pontomidbrain junction, and a DVA is identified in the cerebellum. (Image courtesy Judith Gault, PhD.)

Histologic examination of cerebral AVMs demonstrates thickened walls of both variably enlarged feeding arteries and draining veins, the walls demonstrating smooth muscle hyperplasia, fibroblast infiltration, and increased connective tissue.14 Arteries are differentiated from veins on the basis of the presence of elastic lamina and muscular layers within arteries. Adjacent brain parenchyma may exhibit atrophy, reactive astrocytosis (gliosis), and demyelinization54,302,303 due to ischemic steal by the AVM,62 compressive mass effect, venous hypertension, and prior hemorrhage. If previous hemorrhage has occurred, ferritin and hemosiderin staining of brain will be demonstrated. Calcification may be identified within component vessel walls as well as in adjacent brain parenchyma because of chronic venous hypertension or hemorrhage. Overlying leptomeninges may be thickened by venous engorgement or

Roughly 30% to 55% of patients present with intracranial hemorrhage,2,49,84,298 a heralding event that is also most common in children with cerebral AVMs. Seventy percent of patients with intracranial hemorrhage due to AVMs present before age 40 years and are more likely to be women.20,49,124,203 Intracranial hemorrhage is most commonly intraparenchymal but may be intraventricular or subarachnoid. Nontraumatic subarachnoid and intraventricular hemorrhages are more commonly due to rupture of cerebral aneurysms. Spontaneous thrombosis of a draining vein and rupture of a nidal aneurysm are believed to be the most common causes of AVM hemorrhage. Exclusive deep venous drainage, high-grade venous stenosis, deep brain location, a periventricular or intraventricular location, and associated nidal or feeding artery aneurysms have been reported as other characteristics indicating a propensity for AVM hemorrhage.203,253,350,391 Peripheral venous drainage or mixed peripheral and central venous drainage, along with angiomatous change, a condition characterized by dilated arterial collaterals that pass through brain to supply the AVM, have been correlated with a decreased incidence of AVM hemorrhage.253,254 AVMs 3 cm or less in diameter have been reported to bleed more frequently and to produce larger hematomas than larger AVMs.177,319 The rate of hemorrhage of a cerebral AVM is estimated to be 1% to 3.5% per year, increasing with advancing age, and may be as high as 34% depending on the AVM characteristics and presentation.30,203,298,391 Lifetime percentage risk of AVM bleeding is estimated by the equation of 105 minus the patient’s age.298 The potential yearly bleeding rate is higher than the yearly rupture rate for cerebral aneurysms.20,49,84,132,154,306,394 Pregnancy and the puerperium do not increase the risk of AVM hemorrhage, although the rate of recurrent hemorrhage may be increased in a pregnant woman who presents with hemorrhage from an AVM.239,298 Mortality of 10% to 30% is associated with each instance of hemorrhage,49,84,113,298,450 for an overall annual as well as per-event mortality rate of less than 1%.145 The risk of permanent neurologic deficit is 20% to 50% with each AVM hemorrhage,20,49,84,113,124,146,306,320 again resulting in an annual incidence of approximately 1%; long-term disability with each hemorrhage is estimated to be 10% to 20%.298 The risk of rebleeding after a first hemorrhage is approximately 6% in the first year132 and subsequently decreases to an annual rate of 3%.49,316,409 An AVM that is initially diagnosed after causing an intracranial hemorrhage is more likely to bleed again than an unruptured AVM otherwise identified.203,391,458 Clinical deterioration due to vasospasm from a ruptured AVM is unusual.298 The second most common clinical presentation of cerebral AVMs is seizure, reported in 22% to 60% of cases.132,262,298,316,439 Seizures manifest more commonly in patients who are younger than 30 years and with AVMs that are large or diffuse. Large AVMs are also more commonly identified with progressive neurologic deficit. The overall risk of a seizure disorder in a patient with an AVM is 18%, and the risk of a neurologic deficit is 25% to 30%,5,49,84 although neurologic deficits are reportedly uncommon secondary to regional AVM steal.298 There is an approximately 1 in 4 chance that a patient will become intellectually disabled because of a brain AVM, without or with previous hemorrhage. Brain AVMs cause nonhemorrhagic symptoms due to enlarged AVM feeding arteries and nidus vessels, regional venous hypertension due to the engorged high-flow draining veins of the AVM, and possibly blood flow steal from adjacent normal brain territories. Clinical sequelae are seizures, neurologic deficit, headache, and pulsatile

http://radiologyebook.com

381

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations tinnitus.298 Acute neurologic deficits, however, are most commonly associated with intracranial hemorrhage due to an AVM. Occasionally brain AVMs can result in localized mass effect. Obstructive hydrocephalus may be caused by AVM compression of CSF pathways; communicating hydrocephalus may be secondary to previous hemorrhage and arachnoid granulation CSF resorption dysfunction. Headache is identified in more than 50% of patients with brain AVMs.459 Thirty percent of patients report a bruit, which is due to transmission of blood pulsations from the AVM.49 Calvarial auscultation of a cranial bruit is unusual, however. Occasionally an AVM may cause either hemifacial spasm or cranial nerve neuralgia or palsy if the AVM in whole or in part is located in the basilar cisterns459 or if subarachnoid hemorrhage has occurred previously, producing cisternal siderosis and cranial nerve irritation.

without the administration of a contrast agent (Fig. 12-46). Acute hemorrhage is well documented when present22 (Fig. 12-47). Occasionally, mass effect, surrounding edema, calcification (25%-30% of cases), or adjacent brain atrophy may be identified.215 IV contrast enhancement is essential for identifying the typical “bag of worms” appearance of a brain AVM (see Fig. 12-46) as well as its enlarged supplying arteries and draining veins.

Magnetic Resonance Imaging.  MRI examination is for the most part superior to CT for identifying and evaluating cerebral AVMs. MRI not only outlines the nidus and permits evaluation of its size but also identifies supplying arteries and draining veins. A cluster of vessels of variable signal intensities, depending on the imaging sequences used, characterizes the AVM nidus. Flow-dependent phenomena such as

Grading TABLE 12-3  Cerebral Arteriovenous

The most commonly used AVM grading system was proposed by Spetzler and Martin.387 This system assigns a numerical grade to the AVM on the basis of the size of the nidus, the location of the nidus in relationship to eloquent functional brain, and the venous drainage (superficial or deep) (Table 12-3). The intent of such a grading system is to predict therapeutic risk and outcome, and its application is fairly simple; the details of many AVMs are not completely explained by this classification system, however. A newer condensed classification system combines grades I and II and grades IV and V, leaving grade III unaltered.388

Malformation Spetzler Martin Grade Parameter

Imaging Evaluation

Computed Tomography.  CT imaging of the brain should be performed without and with contrast to evaluate for intracranial hemorrhage and to visualize the abnormal intravascular contrast within the AVM. AVMs demonstrate a focal collection of isointense to slightly hyperintense vessels on nonenhanced scans but may be unremarkable

A

Assigned Points

Nidus Size Small (<3 cm) Medium (3-6 cm) Large (>6 cm)

1 2 3

Adjacent Brain Eloquence Noneloquent Eloquent

0 1

Venous Drainage Superficial only Deep

0 1

Total number of points = Grade I-V.

B FIG 12-46  A, Noncontrast CT scan demonstrating an isodense to slightly hyperdense AVM (arrow). B, Contrast-enhanced CT scan illustrating the typical appearance of a cerebral AVM (arrow) as a “bag of worms.” (Case courtesy Joe Allen, MD.)

http://radiologyebook.com

382

PART II  CT and MR Imaging of the Whole Body

B A

C FIG 12-47  A-C, Acute right cerebellar hematoma (arrow in A) seen on noncontrast CT scan is secondary to an AVM (white arrow in B and C), better demonstrated on T1-weighted (B) and T2-weighted (C) axial MRIs. The cerebellar hematoma is visible on the T1-weighted MRI study (black arrow in B) but is not well demonstrated with the T2-weighted MRI sequence (C).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C

D

383

FIG 12-48  A, Short-TE, long-TR axial MRI illustrating a cerebral AVM extending from cortex to ventricle. The AVM nidus is of decreased signal intensity owing to rapid flow, which produces phase-encoding artifact (arrow). B, Axial T2-weighted MRI demonstrates some of the draining veins of the AVM (short arrows) and less uniform decreased signal intensity in smaller nidus vessels, some of which have increased signal intensity (long arrow). C, Coronal T1-weighted MRI illustrates the triangular shape of the AVM. D, Coronal T1-weighted contrast-enhanced image illustrates variable contrast enhancement of the nidus of the AVM owing to differential flow rates, with faster flow producing less enhancement. Note the phase-encoding artifact (arrow).

turbulence, dephasing, rephasing, slow flow, entry slice enhancement, and phase-encoding mismapping may all be demonstrated (Figs. 12-48 and 12-49). Most often the nidus vessels demonstrate signal void on spin echo or fast/turbo-spin echo pulse sequences and bright or dark signal intensity on gradient echo sequences, depending on pulse sequence parameters (see Fig. 12-49). The appearance of enlarged feeding arteries and draining veins is the product of blood flow patterns and velocity characteristics. Adjacent brain parenchyma can be well evaluated for edema or reactive astrocytosis (gliosis), which have bright signal intensity on T2-weighted spin echo or fast/turbo-spin echo and FLAIR studies55,383 (Fig. 12-50). MRI is crucial for localizing the AVM in relationship to eloquent cortex or deep gray matter structures and significant white matter fiber tracts, as well as demonstrating

any mass effect (Fig. 12-51). MRI well demonstrates the expected  evolutionary appearance of acute, subacute, and chronic hemorrhage (see Fig. 12-47). Chronic hemorrhage with resultant ferritin- or hemosiderin-stained brain, including superficial or ependymal hemosiderosis,55,126,383 is displayed as areas of parenchymal decreased signal intensity on T2-weighted spin echo or fast/turbo-spin echo, T2*weighted gradient echo sequences, and SWI.62 Gd-based contrast agents have a limited role in the MRI evaluation of cerebral AVMs because of inconsistent enhancement and the production of excessive flow artifact. Contrast-enhanced MRI studies can demonstrate slow arterial or venous flow, however. Functional MRI uses cortical blood flow changes based on deoxyhemoglobin levels to create activation maps, which may be used to

http://radiologyebook.com

384

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 12-49  A, Short-TE, long-TR spin echo axial MRI demonstrating a right lentiform nuclei/internal capsule/ thalamic AVM (arrow). B, Intermediate-weighted gradient echo axial MRI of the same AVM, in which much of the nidus is bright (long arrow) and the draining vein, an enlarged vein of Rosenthal, remains black owing to rapid, turbulent flow (short arrow) but demonstrates spatial phase-encoding mismapping.

evaluate brain function within and adjacent to an AVM before therapeutic interventions are undertaken57,248,324,390 (see Fig. 12-51). In some instances the tested neurologic function is expressed in an unexpected cerebral region secondary to neural plasticity, a consequence of the congenital and long-standing nature of the AVM and the development of alternative neural circuits.248 Increased regional cerebral blood volume and regional cerebral blood flow with a decrease in regional mean transit time displayed using MR perfusion imaging illustrate relative AVM steal phenomena.138 Positron emission tomography (PET) and single-photon emission computed tomography (SPECT) scintigraphy can document AVM nidal and perinidal diminished metabolism and perfusion, validating local tissue steal (Fig. 12-52). Diffusion tensor tractography is an MRI technique that has been used to confirm AVM involvement of nearby white matter tracts173,300,464 (see Fig. 12-51). These demonstrable spatial relationships have been correlated with the patient’s neurologic deficits.169

MRA and CTA.  MRA and CTA can help in the evaluation of cerebral AVMs, but these studies may be inconclusive for AVMs that are small or have slow flow.102 CTA and MRA of the circle of Willis as well as the territory containing the nidus must be performed so that all critical features of the AVM are imaged (Fig. 12-53). Evaluation of the source planar images acquired during MRA may reveal a small AVM as bright vessels differentiated from intravascular or extravascular clot, which may obscure an AVM on conventional MRI and the reconstructed 3D MRA. Both TOF and PC methods can be used in MRA examinations.255 MOTSA and PC MRA are better suited to diagnose AVMs with slow flow.36,255 PC MRA is problematic because of the requirement to select a velocity-encoding parameter that is unable to accommodate different speeds of blood flow within different components of the AVM, therefore incompletely evaluating the anatomy of the AVM

unless multiple interrogations with several different velocity-encoding selections are carried out. PC MRA with radial readout has been reported to evaluate AVM hemodynamic parameters.61 MRA is notorious for incompletely visualizing AVM details such as nidus aneurysms because of variable flow effects and related artifact that can be accentuated with IV contrast102,400 (Fig. 12-54). Its sensitivity for identifying cerebral AVMs has been reported to be less than MRI.136 Both MRA and CTA have difficulty discriminating the supplying arteries of the AVM from draining veins, and venous stenosis can be missed. Unenhanced and contrast-enhanced 4D MRA, especially when performed at 3-T field strength and using advanced acquisition techniques such as parallel imaging and specialized k space readout, address some of these issues and may be valuable in both pre- and posttreatment evaluation of cerebral AVMs.104,142,214,330,385,461 Selective arterial spin labeling MRA at 3 T has also been reported to contribute additional anatomic and physiologic information.214 Supplemental surgical planning information can be gained from the fusion of contrast-enhanced MRA and MRI.28 CTA has been reported to be superior to MRA for evaluating both the AVM nidus and especially intranidal aneurysms.136 The precise angioarchitecture of cerebral AVMs is best demonstrated with traditional catheter angiography, which can identify flow-related and nidal aneurysms, the rate of AVM flow, characteristic early filling veins, all of the contributing arteries and draining veins, and arterial and venous stenoses.

Differential Diagnosis The differential diagnosis of cerebral AVMs is limited. Incompletely  or totally thrombosed AVMs can appear similar to other masses, including vascular neoplasms such as hemangioblastoma. When a thrombosed AVM has an accompanying intraparenchymal hematoma,

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C

D FIG 12-50  Brain AVM with bordering reactive astrocytosis (gliosis). A, T2-weighted axial MR study demonstrating a high-flow right frontoparietal AVM and draining vein (arrow). B, More inferior T2-weighted axial section reveals enlarged right middle cerebral artery (lateral arrows) and anterior cerebral artery (medial arrow) branches supplying the AVM. C, Coronal T2-weighted image illustrating the extent of the AVM from cortex to ventricle. D, FLAIR image highlighting reactive astrocytosis (gliosis) bordering the AVM, seen as bright signal intensity (arrow).

http://radiologyebook.com

385

386

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 12-51  A, Axial T2-weighted MRI illustrating a right frontoparietal insular AVM with regional mass effect. Note sulcal effacement and shift of midline structures to the left. Arrow indicates a draining varix. B, Coronal functional MRI study demonstrates left finger-tapping cortical activity (red focus) is located superior to the AVM nidus (arrow). C, MR tractography reveals disruption of the right corona radiata by the AVM. The blue arrow points out the normal left corona radiata.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

387

B FIG 12-52  A, Axial T2-weighted MRI demonstrates an AVM with venous drainage to an enlarged left thalamostriate vein (arrow). B, Axial Ceretec SPECT brain scan demonstrating diminished cerebral perfusion within the AVM (arrow).

rhage,398,467 transient or permanent ischemic stroke, or seizures and is seen in both children and adults. The most typical age at presentation in adults is in the fourth decade. Moyamoya is characterized by progressive obliteration of large proximal anterior circulation vessels, with the development of small “puff-of-smoke” (translation of Japanese moyamoya) arterial collaterals. The posterior circulation is less frequently involved, or its proximal arterial narrowing may be present to a lesser degree. Causes of findings similar to those of moyamoya include neurofibromatosis, sickle cell disease, prior radiation therapy, and Down’s syndrome. The MRI/MRA and contrast-enhanced CT/ CTA appearance of moyamoya is notable for the numerous small vessels seen within the inferior brain substance, which represent the collaterals for severely narrowed or occluded distal internal carotid or proximal anterior or middle cerebral arteries112,148 (Fig. 12-55). Hemorrhage or infarction may also be imaged.

Treatment FIG 12-53  Reconstructed collapsed MRA projection illustrates a left temporal lobe AVM (short arrow) that is supplied by an enlarged left anterior choroidal artery (long arrow).

differentiating it from a hemorrhagic metastasis or high-grade glioma, or a nonmalignant cause of hemorrhage such as hypertension or amyloid angiopathy, can be difficult. Occasionally, moyamoya can mimic basilar cistern or basal ganglia AVMs. An unusual occurrence in North America, moyamoya is more commonly identified in Japan. It may manifest as intraparenchymal (60%), intraventricular (40%) or (rarely) subarachnoid hemor-

The therapy of brain AVMs is based on one of the following four options: conservative observation, surgical resection, radiosurgery, and endovascular embolization. The various treatment options may be combined. Complete surgical resection has been reported in approximately 80% of brain AVMs, with a combined morbidity and mortality of 2% for small AVMs, 15% for medium-sized AVMs, and 40% to 60% for large complex AVMs.154,30 Cerebral edema, ischemia, and hemorrhage can result after surgical resection of a cerebral AVM.298 Radiosurgery is typically employed for brain AVMs that are 3 cm or less in diameter, but it can be used for an AVM whose nidus is larger. Radiosurgery has obliterated more than 80% of AVMs at 3-year follow-up in multiple reported series, with an even greater cure rate for small AVMs.56,298,466 The treatment has a complication rate of less than 3%,

http://radiologyebook.com

A

B FIG 12-54  A, Oblique AP MRA reconstruction of a cingulate AVM. Note the absence of nidal aneurysms. B, Nidal aneurysms are seen on the catheter left carotid cerebral angiogram (arrows).

B

A

C

D

E

FIG 12-55  A, Coronal T2-weighted MRI with abnormal vascularity in the expanded right chiasmatic cistern (arrow). B, Reconstructed oblique coronal noncontrast MRA illustrates right moyamoya vessels (arrow). C, On the axial CTA image, similar findings are seen that mimic an AVM (arrow). Confirmation of unilateral moyamoya disease (arrows) is obtained on the AP and oblique lateral (D and E) right carotid cerebral cerebral angiogram. Note the moyamoya arteries supplying the right middle cerebral and right posterior cerebral http://radiologyebook.com arteries but not the right anterior cerebral artery.

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

389

FIG 12-56  Sagittal postcontrast T1-weighted MRI demonstrating radiation necrosis (arrow) after radiosurgery for a cerebral AVM.

adverse events being most often due to radiation necrosis.56,106,241,242,294,393 Embolization is often performed to reduce the flow and size of brain AVMs before surgical resection or radiosurgery, but in some instances, complete cure can be achieved with embolization alone.29 Conservative management has been advocated in some trials and observational studies.8 MRI examination is useful to evaluate cerebral AVMs after therapy, particularly radiosurgery or embolization,205,383 to demonstrate decreased size, reduced flow,241,251,252 and bordering edema. The use of a contrast agent is helpful for identifying any residual AVM nidus; however, contrast enhancement may also be secondary to postoperative reactive astrocytosis (gliosis), radiation necrosis, infarction, or resolving hemorrhage (Fig. 12-56). TOF MRA without contrast has been used to document AVM nidus reduction after radiosurgery.460 4D MRA with or without contrast may be helpful in identifying residual nidus after surgical resection or radiosurgery, with a negative predictive value of 78% to 90%.142,235,385 Perinidal edema after radiosurgery, seen as increased signal intensity on T2-weighted spin echo MRIs, is expected and does not necessarily indicate impending radiation necrosis. A 2% risk of hemorrhage may manifest as a component of radiation necrosis, occurring as soon as 3 months after radiation therapy, or result from AVM rupture, with or without thrombosis.466 CTA may be superior to MRA after treatment, especially when compared to using 1.5-T field strength119 (Fig. 12-57). 3D CTA employing postprocessing segmentation and fusion with MRI for anatomic localization can be useful for presurgical AVM planning.268 Catheter cerebral angiography using a reconstructed dynamic CT data set and fusion with stereotactic CT or MR can also be beneficial in assisting AVM resection.407 Catheter cerebral angiography remains the standard for documenting obliteration of treated cerebral AVMs.

CAVERNOUS MALFORMATION OR CAVERNOUS ANGIOMA

FIG 12-57  Sagittal reconstruction CTA image demonstrating residual frontal AVM nidus (red arrows) after gamma knife radiation and Onyx (Covidien eV3 Neurovascular) embolization, located anterior and superior, inferior and posterior to the AVM (blue arrows).

0.5% of the general population.315,470 They are the second most common type of symptomatic vascular malformation. Cavernous malformations typically present as focal parenchymal intracranial hemorrhages between the third and fifth decades of life. Those found in the brain occur equally in males and females. The most common location is in the cerebral hemispheres, with only 10% identified in deep gray matter or white matter tracts. Approximately 25% are found infratentorially in equal numbers within the brainstem and cerebellum. The pons is the most common location in the brainstem. Unusual cases may arise from the leptomeninges,382 dura,33 ependyma, choroid plexus, or cranial nerves, or extradurally, such as in the cavernous sinus.384 In 30% to 50% of cases, cavernous malformations are multiple (Fig. 12-58). Spinal cord cavernous malformations may be isolated or may be associated with cerebral cavernous malformations.44,77 A genetic predisposition manifests when cavernous malformations are found in 10% to 15% of family members in addition to the index case.31,81,335 When a familial tendency exists, the propensity for multiple cavernous malformations rises to 80%,308,470 and the new development of cavernous malformations is also significantly increased in relatives.219,230 Recent investigations have mapped cavernous malformation loci to KRIT1/CCM1, MGC4607/CCM2, and PDCD10/CCM3 genetic locations73 that code for proteins shown to be frequently mutated,59,220,240 although pathologic penetrance maybe be incomplete even though inheritance is in autosomal dominant fashion.98 Multiple cavernous malformations can be seen in HHT or Osler-Weber-Rendu disease, with an incidence of about 1%.458

Pathology

Incidence Cavernous malformations account for approximately 10% to 15% of all cerebral vascular malformations309 and are found in approximately

On pathologic examination, cavernous malformations are well circumscribed and typically have a mulberry configuration. The lesions are not encapsulated but contain a compact or sometimes racemose network of multiple endothelium-lined, sinusoidal, nonuniform

http://radiologyebook.com

390

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-58  Axial T2-weighted MRI (A) shows no cavernous malformations in a patient with hereditary hemorrhagic telangiectasias; these are identified only on the axial T2* gradient echo scan (B).

vascular channels containing thrombosed blood of varying ages.54,302 The endothelium demonstrates fenestrations, large intercellular gap junctions, and internal vesicles.416 The extent of tight junction deficiency appears to correlate with a propensity to grow in size and bleed.176 Compartment walls are irregularly thickened with collagen, hyalin, and scattered calcium deposition and contain no basement membrane or muscularis or elastic lamina layers.416 No interposed brain tissue is identified within the cavernous malformation. Adjacent brain may be calcified, laden with hemosiderin or ferritin, atrophic, and gliotic. Cavernous malformations are typically parenchymal lesions, sometimes marginating a pial or ependymal surface. Hemorrhage from a cavernous malformation uncommonly extends into the ventricular system or the subarachnoid space. Capillary telangiectasias may coexist with or may be mixed with a cavernous malformation.22,284 Developmental venous anomalies (DVAs) are found in approximately 20% of patients in whom a cavernous malformation is identified,2 the DVA usually neighboring the cavernous angioma.131,337,403 When a cavernous malformation and a DVA coexist, they are more frequently found in the posterior fossa and the cavernous malformations are often multiple.2 A cerebral AVM and a cavernous malformation are infrequently present in the same patient, although this is more likely in HHT or Osler-Weber-Rendu disease.

Presentation The clinical presentation of cavernous malformations typically consists of seizure activity or acute or progressive neurologic deficit due to cerebral hemorrhage,337 but in one study, only 61% of patients were symptomatic.470 Seizures, the most common symptom, occur in approximately 50% of patients.98 Seizures are believed to be secondary to mass effect from the cavernous malformation and adjacent brain parenchymal hemosiderin or ferritin and reactive astrocytosis (gliosis). Symptomatic hemorrhage manifesting as headache, seizures, or acute

neurologic deficit most often occurs in the second and third decades of life and has an incidence of approximately 0.5% up to 4.5% per year.26,89,179,210,341 Cavernous malformations having once bled, and lesions containing subacute hemorrhage have a significantly increased risk of recurrent hemorrhage.179,210 Subclinical hemorrhage is not uncommon, and clinically significant hemorrhage occurs in only 10% to 13% of cases.54 When a cavernous angioma is large enough, mass effect can produce cranial nerve palsies and focal neurologic deficits.133 Cavernous malformations may initially be tiny and grossly unrecog­ nizable and then suddenly enlarge because of new hemorrhage and thrombosis. They often change over time in size, shape, and character because of intermittent and recurrent hemorrhage and thrombosis, recanalization, vascular ingrowth, and granulation tissue.44,380 Elevated serum levels of vascular endothelial growth factor (VEGF) have been reported during cavernous malformation enlargement and evolution.184 Cranial radiation therapy may cause thrombosis and hemorrhage to develop in preexistent unrealized cavernous malformations, making them appear as new findings on serial MRI studies; or it is possible that radiation causes new cavernous malformations.27 Cavernous malformations in the brainstem have been reported to exhibit a bleeding rate of 2.46% per year and a rebleeding rate of 5.1% per year and present as new hemorrhage more commonly when larger than 10 mm in diameter and if hemorrhage beyond the geographic confines of the malformation is preexisting.216 The majority of brainstem cavernous malformations remain stable in size, with approximately 15% enlarging and 15% shrinking in size over time.216 They are much less likely to bleed in patients older than age 35, and fewer than 10% cause significant morbidity.216

Imaging Evaluation

Computed Tomography.  CT scans of cavernous malformations typically demonstrate an isodense to hyperdense focal area within the

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

391

B

A

C FIG 12-59  A, CT scan of a cingulate gyrus cavernous malformation with calcification in its margin (arrow). Sagittal T1-weighted image (B) and axial T2-weighted spin echo image (C) of the same cavernous malformation (arrows).

brain parenchyma that may or may not contain calcification (Fig. 12-59). They may or may not enhance after IV administration of an iodinated contrast agent362 (Fig. 12-60). Mass effect and edema associated with the cavernous malformation may be present, depending on its size and hemorrhage status.213

Magnetic Resonance Imaging.  MRI is superior to CT for identification of cavernous malformations.335 It classically demonstrates a mass characterized as “popcorn” in appearance owing to the varying ages of internal hemorrhages127 (Fig. 12-61), the typical Zabramski type 2,470 with a margin of decreased signal intensity on T2*-weighted gradient echo, SWI, and T2-weighted spin echo or fast/turbo-spin echo

images.2,127 This latter finding is due to magnetic susceptibility–caused phase incoherence created by the hemosiderin and ferritin chronic residua from previous hemorrhage and thrombosis.44 Hemosiderin and ferritin may cause a dark, flaring appearance extending into the adjacent brain parenchyma, because macrophages and astrocytes remove these and other blood breakdown products along white matter fibers54 (Fig. 12-62). The signal intensity of the rim of the cavernous angioma is also decreased because of fibrin and collagen deposition. If calcification is present, it too can contribute to the signal loss. T2*weighted gradient echo scans or SWI scans, which are best for identification of hemosiderin and ferritin, should always be performed in the evaluation for intracranial hemorrhage and will increase the

http://radiologyebook.com

392

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 12-60  A, Noncontrast CT scan reveals a pontine cavernous malformation (arrow) with hemorrhage of varying ages and possible calcification within its margin. B, Contrast-enhanced CT scan demonstrates no contrast enhancement of a brainstem cavernous malformation (arrow).

conspicuity of cavernous malformations234 (Fig. 12-63). In some cases the T2* gradient echo or SWI study reveals multiple cavernous malformations when the T2-weighted spin echo or fast/turbo-spin echo study illustrates none, fewer, or only a solitary lesion.98,230 A T2* gradient echo or SWI examination may be the only imaging study that identifies a cavernous malformation (see Fig. 12-58). Cavernous malformations presenting as circumscribed subacute hemorrhage with a hemosiderin or ferritin margin would be classified as Zabramski type 1 (see Fig. 12-59), and those that have involuted and fibrosed with only a collection of hemosiderin or ferritin remaining would be classified type 3470 (see Fig. 12-62). Some cavernous malformations are identified as black spots on T2* gradient echo or SWI that are often multiple and although small, still present a low risk of future hemorrhage.287 These would be classified as Zabramski type 4470 and can be mimicked by tiny capillary telangiectasias336 (see Fig. 12-58). Increasing the magnetic field strength to 3 T or even 7 T further increases lesion detection sensitivity.111 Edema adjacent to a cavernous malformation is best illustrated on MRI and may correlate with the patient’s having recently become symptomatic due to new hemorrhage (Fig. 12-64; see Fig. 12-61). Owing to the exquisite sensitivity of MRI for blood degradation products, cavernous malformations may be incidentally discovered as asymptomatic lesions.335 Gd-based contrast agents may produce some enhancement of the cavernous malformation. Unusual extraaxial cavernous malformations may have a typical or atypical MRI appearance. Extraaxial cavernous malformations may resemble meningiomas or pituitary adenomas, with homogeneous intense contrast enhancement on T1-weighted images and bright signal intensity on T2-weighted spin echo or fast/turbo-spin echo images.33,191,197,212,238,384 Catheter cerebral angiography is indicated only to exclude a true AVM as the cause of intracranial hemorrhage when the diagnosis of

cavernous malformation using MRI is uncertain. Catheter angiography displays cavernous malformations as avascular masses, occasionally demonstrating very late capillary staining without arteriovenous shunting.362 When present, an associated DVA may be demonstrated.

Treatment Therapy for cavernous malformations consists of medical treatment for seizure control, surgical excision when the lesion is accessible and clinically indicated, and radiosurgery.

Differential Diagnosis Differential diagnosis of cavernous malformations includes previous hypertensive hemorrhage, treated toxoplasmosis, previous radiation therapy, amyloid angiopathy, and disseminated intravascular coagulopathy. Demonstration of a lesion containing hemorrhages of multiple ages, a complete peripheral hemosiderin/ferritin ring with flaring into neighboring brain, and lack of surrounding edema helps differentiate a cavernous malformation from hemorrhagic metastasis. A positive PET examination can identify a malignancy when the mass is of sufficient size, depending on the spatial resolution of the scanner employed. A cavernous malformation will not take up a significant amount of 19-fluorodeoxyglucose. In some instances, particularly  with melanoma, only follow-up MRI examinations documenting typical hemorrhage evolution can confirm the diagnosis of cavernous malformation.

CAPILLARY TELANGIECTASIA Incidence and Histology Capillary telangiectasias are small, usually clinically silent lesions that most commonly occur in the pons or cerebellum but can occur in other locations in the brain parenchyma. They are the second most

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C FIG 12-61  MRI examination of a cavernous malformation demonstrating typical variegated appearance containing hemorrhage of varying ages. A, T2-weighted image of a right subinsular/lenticular nucleus/ thalamic/internal capsule cavernous malformation. Note the peripheral hemosiderin/ferritin (long arrow) and edema located at the site of recent hemorrhage (short arrow). B, T1-weighted image without contrast. Note the subacute hemorrhage (arrow). C, T1-weighted image with contrast. Note the lack of appreciable contrast enhancement. (Case courtesy David Watts, MD.)

http://radiologyebook.com

393

394

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-62  A, Axial short-TE, long-TR spin echo MRI demonstrating cavernous malformations in the right middle cerebellar peduncle and adjacent cerebellar hemisphere and another in the left cerebellar hemisphere. B, Axial T2-weighted MRI revealing a typical cavernous malformation in the right frontal lobe. Both images illustrate hemosiderin- and ferritin-laden macrophages and astrocytes removing blood breakdown products migrating in the adjacent white matter, visualized as bordering decreased signal intensity.

common cerebral vascular malformation after DVAs and are usually multiple.303,308 The histology of capillary telangiectasias is characterized by dilated thin-walled capillaries. Unlike cavernous malformations, normal brain parenchyma is identified within the capillary telangiectasia.54,302 Adjacent brain is most often normal but may contain reactive astrocytosis (gliosis) or hemosiderin/ferritin from previous hemorrhage.54 Capillary telangiectasias are believed to rarely bleed. Other types of cerebral vascular malformations can be found with capillary telangiectasias, especially cavernous malformations337; some have proposed that these two cerebral vascular lesions represent a continuous pathologic spectrum.336 Cerebral capillary telangiectasias can be found in HHT or Osler-Weber-Rendu disease (incidence ≈ 2%),458 along with telangiectasias in other visceral, mucosal, and cutaneous sites in addition to pulmonary and cerebral true AVMs.3,235,246 Cerebellar capillary telangiectasias may also be a component of the ataxia-telangiectasia syndrome.

Imaging Evaluation Capillary telangiectasias are most often not identified on nonenhanced MRI studies using all available standard pulse sequences, but sometimes they are visualized as a poorly circumscribed focus of slightly increased signal intensity on short-TE, long-TR (repetition time), or T2-weighted spin echo or fast/turbo-spin echo imaging.227 Rare capillary telangiectasias that have hemorrhaged demonstrate expected MR findings of a collection of blood, but most often the hemorrhage is minor and remote, best demonstrated on T2*-weighted gradient echo or SWI studies. Capillary telangiectasias typically lack evidence of hemosiderin or ferritin deposition. SWI and T2*-weighted gradient echo sequences have also been reported to identify brainstem capillary telangiectasias producing focal areas of marked decreased signal intensity, presumably owing to deoxyhemoglobin within the malformation.227,468 Enhancement with Gd-based contrast agents on a

T1-weighted spin echo pulse sequence is the best imaging study and demonstrates a lacy network of vessels of increased signal intensity or a focal patch of variable hyperintensity constituting the capillary telangiectasia25,227 (Fig. 12-65). A coexisting cavernous malformation may obscure identification of the capillary telangiectasia.22 CT usually reveals no abnormality.

DEVELOPMENTAL VENOUS ANOMALIES (VENOUS ANGIOMAS) Incidence and Histology DVAs, or venous angiomas, are the most common cerebrovascular malformation, accounting for almost two thirds of the total and having an incidence in the general population of approximately 2%. They are twice as common in patients with HHT or Osler-Weber-Rendu disease.228,303,309,412,458 This vascular malformation represents a developmental variant of normal venous drainage222,353 composed entirely of venous elements, and most authors prefer to call them developmental venous anomalies, dropping the term angioma.222 The association of DVAs with neuronal migration anomalies and the absence of normal venous drainage from regional brain containing the DVA emphasizes their embryologic origin222,334,412 (Fig. 12-66). DVAs have an increased incidence in patients with periorbital lymphatic and mixed lymphaticvenous malformations.34 DVAs are typically solitary and located in any part of the brain parenchyma, although they are most often located in the frontal lobe, less commonly in the cerebellum.309,423 They range in size from small to large enough to drain an entire hemisphere.412 Multiple DVAs are not rare, particularly when they occur bilaterally in the cerebellum,412 and their multiplicity has been described in the blue rubber bleb nevus syndrome.114 DVAs are characterized by a medusa head of enlarged medullary veins draining to a collector vein, which continues to either the

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B FIG 12-63  A, T2-weighted axial MRI scan at the level of the lateral ventricles and immediately supraventricularly a cavernous malformation in the anterior left cingulate gyrus/corpus callosum genu and adjacent white matter (arrows) that becomes much more conspicuous on the axial T2* gradient echo MR scan (B).

http://radiologyebook.com

395

396

PART II  CT and MR Imaging of the Whole Body

superficial or subependymal cerebral venous drainage system or a combination of these.131,412,423,446 DVAs may occasionally have several collector veins, or there may be several DVAs located in the same brain region. DVAs are occasionally found intraventricularly. Unusual cases of a DVA draining to the foramen of Monro or causing aqueductal stenosis that results in obstructive hydrocephalus have been reported.412,423,462 The brain parenchyma through which the DVA passes is typically normal and usually shows no evidence of previous bleeding.

Presentation The vast majority of DVAs are silent and incidentally discovered. Some authors believe that DVAs may rarely cause venous infarction with or without hemorrhage or may cause parenchymal or subarachnoid

hemorrhage144,208,224,418,423,452,453 (Fig. 12-67). Hemorrhage from a DVA is possibly due to acquired venous outflow restriction by stenosis or thrombosis100,222 secondary to hemodynamic stresses. A venous varix or ectatic component may be rarely seen and is explained by similar obstructive mechanisms or hemodynamic forces.222 The incidence of associated contiguous cavernous malformations is roughly 10% to 30%,131,263,308 however, and is often used to explain any clinically significant hemorrhage being derived from the cavernous malformation, not from the DVA22,237,260,337,412 (Fig. 12-68). Patients with multiple DVAs, DVAs located infratentorially, and DVAs with more than one draining collector vein are prone to have associated cavernous malformations.263 Rarely seizures, focal neurologic deficit, cerebellar signs, tinnitus, and headache have been ascribed to DVAs,35,222,309,418,423,452,453 depending on their location, but these clinical presentations are controversial.

Imaging Evaluation

FIG 12-64  Edema (arrow) borders an anterior right parietal cavernous malformation seen on this T2-weighted axial MRI scan.

A

Contrast-enhanced CT examination of a DVA demonstrates a dense vascular lesion with a thistle-shaped collection of veins coursing to a single dominant draining vein412,423 (Fig. 12-69). Calcification is unusual, and without calcification they are typically not identified on unenhanced CT scans. A large DVA, however, may be visualized on CT as a slightly dense solitary vessel or collection of vessels.423 Parenchymal calcification of the caudate and putamen sparing the internal capsule has been reported to be associated with DVAs, with a proposed mechanism of chronic venous ischemia being the pathophysiologic etiology.95 CTA, 4D CTA, or CT perfusion can identify and evaluate DVAs. CT perfusion is used to document alterations in perfusion parameters dependent on the rapidity of venous drainage, which may also explain how these lesions produce symptoms.71 MRI has revealed the true incidence of DVAs in the general population. On unenhanced MRI a DVA appears as a cluster of small vessels with signal void if flow within them is rapid, turbulent, or dephased; or the vessels may produce increased signal intensity due

B

FIG 12-65  Contrast-enhanced T1-weighted MRI examination (A) demonstrating a pontine capillary telangiectasia (arrow) that is not well visualized on the T2-weighted study (B).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C FIG 12-66  DVA associated with a neuronal migration anomaly. A, T2-weighted MRI demonstrates a right parietal schizencephalic cleft (short arrow) containing a DVA (long arrow) and bordering cortical dysplasia (curved arrow). B, Coronal postcontrast T1-weighted MRI illustrating the DVA (arrow) within the schizencephaly. C, AP view of right carotid cerebral angiogram confirming the DVA (arrow). (Case courtesy David Watts, MD.)

http://radiologyebook.com

397

398

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 12-67  A, T2-weighted axial fast-spin echo image of a thrombosed collector vein of a large DVA (red arrow). Peripheral to the collector vein is a small varix (blue arrow), and posteriorly within the surrounding edema is parenchymal hemorrhage (green arrow). B, SWI illustrates the caput medusa appearance of the DVA, with punctate hemorrhages (red arrow) and the parenchymal hematoma (green arrow). C, The contrastenhanced MRA demonstrates the enhancing territory of the congested DVA (red arrow) due to venous thrombosis.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C FIG 12-68  MRI examination demonstrating a DVA in association with a cavernous angioma. A, T2-weighted image demonstrates a cavernous angioma in the anterior medial right occipital lobe (long arrow), with a neighboring DVA (short arrow). B and C, T1-weighted images after contrast demonstrating the periphery of the DVA (arrow in B) and its collector vein (arrow in C). (Case Courtesy Robert Weil, MD.)

http://radiologyebook.com

399

400

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 12-69  Contrast-enhanced CT examination demonstrating a frontal lobe DVA (arrow in A). Note the collector vein (arrow) illustrated in B.

to slow flow, entry slice phenomena, even-echo rephasing, or when gradient-moment nulling techniques are used. Spatial misregistration artifact occurs in DVAs when gradient-moment nulling techniques are used on T2-weighted spin echo or fast/turbo-spin echo images.221 The rate of blood flow within the DVA usually depends on its overall size and determines its signal intensity on MRI452 together with its orientation to the plane of acquisition. Some T2*-weighted gradient echo sequences, TOF MRA, or PC MRA can demonstrate DVAs as a bright grouping of veins if the flow is fast enough. MR examinations crafted to take advantage of slow flow (e.g., thin-slice 2D TOF oriented perpendicular to venous inflow), including those typically employed in MR venography (MRV), can also identify DVAs309 (Fig. 12-70). SWI and T2*-weighted gradient echo scans designed to visualize flowing blood containing deoxygenated blood as black linear structures can exquisitely demonstrate DVAs. Routine MRA often fails to visualize DVAs; however, neither MRA nor MRV are typically necessary to image them.228 Contrast-enhanced T1-weighted spin echo or gradient echo MRI sequences often best demonstrate DVAs and their relationship to adjacent brain parenchyma.228,229,309,449,452,453 Contrast-enhanced CT or MRI may reveal a focal parenchymal stain around the venous medusa of the DVA due to regional slow venous flow and vascular congestion. MR perfusion imaging can document elevated cerebral blood volume, cerebral blood flow, and mean transit time in normal-appearing brain neighboring a DVA, with or without a concomitant cavernous malformation.375 Frequently DVAs are imaged on MRI as incidental findings when the MRI study is performed for other indications. They also can be overlooked on MRI scans if small and similar to normal venous drainage. DVAs should be differentiated from aberrant venous drainage on the basis of the classic findings of a cluster of enlarged medullary veins continuing into a larger collector vein, aberrant venous drainage

lacking the tuft of upstream smaller veins.229 The MRI appearance of DVAs is so characteristic that catheter cerebral angiography rarely should be performed in their evaluation. Small DVAs can be overlooked on catheter cerebral angiography, frequently considered another example of variant venous drainage, although characteristically they appear later and persist longer than normal vein architecture.309

SINUS PERICRANII Sinus pericranii is a soft, compressible venous malformation that lies within and superficial to the skull and communicates with the intracranial dural sinuses, most often by enlarged emissary veins.41,349 Its etiology is most likely congenital, although it has been reported to occur spontaneously or after trauma.41,349 Sinus pericranii may be associated with other vascular malformations such as cavernous malformations or DVAs, or may be seen as a component of the blue rubber bleb nevus syndrome.41,114,349 Sinus pericranii most commonly occurs in the frontal scalp near the midline.41,349 An asymptomatic cosmetic scalp lump is found typically in a patient younger than 30 years. The lesion is twice as common in men as in women and is remarkable for its enlargement with Valsalva maneuvers.349 It may disappear when erect then form a soft tissue swelling when recumbent.6 On imaging, sinus pericranii has been described as a vascular soft tissue scalp mass that is slightly hyperdense on nonenhanced CT, demonstrates flow on MRI, and enhances strongly with use of a contrast agent on both CT and MRI.41,349 CT best identifies associated osseous defects in the calvarium associated with the vascular mass and its associated transdiploic veins.41,349 CT venography (less so MRV owing to slow flow) and catheter cerebral angiography demonstrate the enlarged scalp veins and their communication with the intracranial dural sinus and cortical veins (Fig. 12-71).

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

401

B

C FIG 12-70  MR venography examination illustrating DVA. Slightly oblique lateral (A) and posterior coronal view (B) of a left frontal DVA (arrow) demonstrated on MR venography. C, Lateral view of venous phase during carotid angiography confirms the DVA (arrow).

ARTERIOVENOUS FISTULAS Etiology, Pathology, Location, Incidence, Presentation, and Classification Cerebral arteriovenous fistulas (AVFs) are usually acquired64 but may be congenital in some instances. Congenital fistulas most often represent a rare anomaly of cerebrovascular development and are part of a spectrum of cerebrovascular disorders, such as an infrequent component of HHT or Osler-Weber-Rendu disease.449,458 Single or multiple arteriovenous connections or a true AVM nidus with one or more associated AVFs illustrates the variety of pathologic presentations (Fig. 12-72). AVFs may be located intracranially or extracranially within the

scalp, neck, or facial structures. The arteriovenous shunting with increased venous pressure and flow within an AVF can result in variably sized varices that when large enough can produce mass effect such as brain parenchymal or cranial nerve compression or obstructive hydrocephalus.97,161,162,223,249,433,434 Intracranial and extracranial posttraumatic AVFs represent another category of fistulas best illustrated by the direct carotid-cavernous fistula, although this can similarly be produced by a ruptured intracavernous aneurysm. Most dural AVFs are spontaneous and are believed to be acquired after thrombosis of a dural sinus or cortical vein, with the subsequent development of multiple fistulous shunts in the involved sinus or vein.162 This has been characterized to occur in three stages: thrombosis, fistula inception, and recanalization.139 Trauma, surgery, infection,

http://radiologyebook.com

402

PART II  CT and MR Imaging of the Whole Body

R

A

F

B

C FIG 12-71  A, Coronal slice of a CT venogram demonstrating sinus pericranii with a congenital skull defect. B, Reconstruction of the CT venogram. C, Carotid cerebral angiogram, venous phase, lateral view of the same patient. Arrows in each illustration indicate the scalp venous malformation component of the sinus pericranii. (Images courtesy Allan Brook, MD.)

inflammation, systemic hypertension, metabolic aberrations, and hypercoagulable states—including those caused by pregnancy, hormonal contraception medications, and dehydration—may be etiologic events that produce dural sinus thrombosis.24,161,162,249 D-dimer levels are elevated in patients with dural AVFs, indicating thrombophilia; these return to normal with AVF treatment.139 Chronic venous hypertension without dural sinus thrombosis has also been postulated to produce dural AVFs.139 Angiogenesis has been linked to the pathogenesis of dural AVFs because venous hypertension causes relative hypoxia,  a pro-angiogenic state, and elevated levels of angiogenic and inflammatory substances such as VEGF, hypoxia inducible factor 1 (HIF-1), macrophage stromal cell–derived factor (SDF)-1α, neutrophil matrix metalloproteinase (MMP)-9, nitric oxide, and cytokines including interleukin (IL)-6.139,249 Dural AVFs are characterized pathologically by abnormal direct communication between normal sinus wall arteries and the sinus

interior or adjacent veins.288 Hemodynamic stresses lead to pathologic disorganization and thickening (less commonly, thinning) of   the walls of supplying arteries, draining cortical and medullary veins, and the dural sinus at the site of the fistula; these stresses also lead  to thrombosis, stenosis, and occlusion of the involved vascular elements, most commonly the dural sinus itself and least commonly the feeding arteries.162,249,288 Sometimes there is a parallel abnormal venous channel next to the involved dural sinus that participates in the  dural AVF.139 Dural AVFs represent approximately 10% to 15% of intracranial vascular malformations51,162,257,434 although not counted in the traditional four categories. They are most often diagnosed in the fifth to seven decades and are more common in women than men,434 but men may present with more significant neurologic manifestations.72,139 Most often the dural AVF involves a single sinus, sometimes with involvement of contiguous sinuses. Multiple (two to four) disparate

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

403

B

C FIG 12-72  Congenital single-feeder AVF between a right middle cerebral artery branch and an adjacent cortical vein documented on MRI. A demonstrates an enlarged right middle cerebral artery (arrow), whereas B illustrates the enlarged angular artery (short arrows) and proximal draining vein varix (long arrow) on these T2-weighted images. C, AP modified Townes view of the right carotid injection confirming the fistula.

dural AVFs are unusual (found in ≈ 8%)51,139 and are typically near each other,249 even the bilateral dural carotid-cavernous fistulas, possibly related by the interconnecting circular sinus and clival veins. In one third to two thirds of cases, dural AVFs involve the sigmoid and transverse sinuses.51,54 The cavernous sinus is the second most common location, accounting for 12% to 20% of dural AVFs; however, this is the most common variety in Asia.139 Dural carotid cavernous fistulas are typically discovered in middle-aged women.51,97 The symptoms of cavernous sinus dural AVFs include palsies of cranial nerves III, IV, and VI, which produce diplopia, proptosis, conjunctival

engorgement, chemosis, retroorbital or cranial nerve V pain (trigeminal neuralgia), increased intraocular pressure, and secondary glaucoma; all of these are due to the venous hypertension within the cavernous sinus and retrograde venous drainage into the involved orbit, most evident as an enlarged superior ophthalmic vein.161,207,406 More ominous consequences are papilledema, choroidal or retinal detachment, central retinal vein thrombosis, diminished visual acuity,249 and stroke or intracranial hemorrhage due to intracranial venous drainage from the cavernous sinus. An orbital bruit can be auscultated in roughly 50%.

http://radiologyebook.com

404

PART II  CT and MR Imaging of the Whole Body

Other sites of dural AVF are the inferior petrosal sinus, occipital and marginal sinuses, hypoglossal canal involving the anterior condylar vein or confluence (which may have an intraosseous component), sometimes included with the marginal sinus variety),389 inferior and superior sagittal sinus, straight sinus, vein of Galen, and the anterior, middle, and posterior cranial fossae venous structures, including cortical veins.97,249,257 In less common sites, dural AVFs may more commonly present with significant neurologic symptoms and overall have a worse prognosis.72 Dural AVFs most often become symptomatic in patients between 40 and 60 years of age. Pulsatile tinnitus and secondly headache are the most common complaints,51,389,434 but these may be mild enough and of such chronicity that the patient does not seek medical attention until other symptoms develop (e.g., cranial nerve palsies or irritation).223 Objective tinnitus is when a bruit is auscultated over the fistulous site.139 In a patient with a normal otoscopic examination and objective tinnitus, the most common diagnosis is a transverse-sigmoid dural AVF.139 Spinal venous drainage can produce myelopathy.389 Neonates may present with congestive heart failure.139 Communicating hydrocephalus may be the result of the dural AVFs producing dural sinus hypertension, dural sinus thrombosis, or prior subarachnoid hemorrhage, thereby affecting arachnoid granulation CSF resorption.223 Increased intracranial pressure without communicating hydrocephalus may be due to cerebral venous congestion, increased cerebral blood volume, elevated intrasinus pressure, and impaired cerebral venous drainage.162,434 Dural AVFs may cause seizures, parkinsonism, or neurologic  deficits secondary to venous hypertension–caused hemorrhage  or ischemia/infarction secondary to cortical, brainstem, or spinal venous drainage, with ischemia less commonly caused by arterial steal.51,162,223,402,419,434 Cortical venous drainage can be significant enough to cause venous hypertensive encephalopathy or dementia in up to 12%, which is often reversible with fistula cure.139 Superior sagittal sinus dural AVFs can produce psychiatric manifestations.139 Dural AVFs may spontaneously obliterate,17,249,389 a course most frequently observed in the cavernous sinus variety139; however, closure of the fistula may be associated with intracranial hemorrhage.223 Approximately 15% of patients present with intraparenchymal, subarachnoid, intraventricular, or subdural hemorrhage, listed in decreasing order of frequency.51,162,223,381 The most common form of hemorrhage—intraparenchymal—is often lobar.139 Significant in the production of intracranial hemorrhage is reversal of flow within the involved dural sinus into variably dilated cortical veins, which results in venous rupture or venous hypertension in the involved cerebral territory162,223,319,389 (Fig. 12-73). The overall risk of hemorrhage from a dural AVF is 1.8% per year, with a 20% mortality rate associated with each hemorrhage.51 Risk factors for producing hemorrhage include (1) AVF location in the anterior cranial fossa (especially ethmoidal, which often present as a higher grade and with a neurologic deficit or hemorrhage, including the unusual subdural hematoma), superior petrosal sinus, tentorium (often categorized as higher grade; hemorrhage risk is 60%-74% per year with an accompanying high mortality; 10%  per year neurologic presentation), or deep veins; (2) leptomeningeal or pial venous drainage; (3) draining vein stenosis; and (4) a venous varix.51,139,249,257,434,471 Dural AVFs have been categorized on the basis of their venous drainage pattern by Cognard and Borden42,76 (Tables 12-4 and 12-5) (see Figs. 12-73, 12-77, and 12-78). An older classification was proposed by Merland and Djindjan, and Lalwani has specified characteristics of transverse-sigmoid sinus fistulas.139 Over time, dural AVFs may change and so be assigned to a different type or grade249 because of new arteriovenous shunts, increased arterial supply, or venous stenosis

TABLE 12-4  Cognard Classification of

Dural Arteriovenous Malformations76 Type

Description

I II   a   b

Antegrade drainage into sinus

  a and b III IV V

Antegrade and retrograde drainage into sinus Antegrade drainage into sinus and retrograde drainage into cortical veins Antegrade drainage into sinus and retrograde drainage into sinus and cortical veins Retrograde drainage into cortical veins Retrograde drainage into cortical veins with venous ectasia Drainage into spinal veins

TABLE 12-5  Borden Classification of Dural

Arteriovenous Fistulas42 Type

Description

I II

Drainage into dural sinuses and/or meningeal veins Type I drainage plus retrograde drainage into subarachnoid veins Drainage into subarachnoid veins only

III

or thrombosis.139 The risk of advancing type or grade is approximately 1% per year.139 Symptomatology and risk of hemorrhage roughly correspond to more abnormal venous drainage,249,389 so a dural AVF with no cortical venous drainage may have an annual risk of cerebral hemorrhage close to 0%, whereas exclusive cerebral or spinal venous drainage may have an annual risk of 10%.139 Furthermore, patients who present with clinical symptomatology have an enhanced chance of incurring a cerebral hemorrhage that ranges from 10% to 89% depending on the degree of abnormal venous drainage of the fistula.139 Arterial supply to dural AVFs is predominantly from meningeal branches of the external carotid artery (e.g., middle meningeal, ascending pharyngeal, and occipital arteries), the internal carotid artery (e.g., tentorial and clival arteries of the meningohypophyseal trunk), and the vertebral artery (e.g., posterior meningeal artery).51 Pial arteries derived from the carotid or vertebrobasilar system can also provide input to dural AVFs.

Imaging Evaluation Nonenhanced CT scans of dural AVFs may be normal unless hydrocephalus, intracranial hemorrhage, edema or infarction from venous hypertension, or thrombosis of a dural sinus or draining vein is identified (see Fig. 12-73). Dystrophic parenchymal calcification may be present if the dural AVF has caused chronic venous ischemia465 (Fig. 12-74). Contrast-enhanced scans may demonstrate enlarged cortical or medullary veins or stenotic or enlarged dural sinuses (see Fig. 12-74). These findings are noted in up to half of patients with dural carotid-cavernous fistulas, with expansion of the involved cavernous sinus and increased size of the draining superior ophthalmic vein (see Fig. 12-75 for similar findings in a posttraumatic carotid-cavernous fistula). CTA can demonstrate enlarged arteries and draining veins or varices from an AVF (see Figs. 12-73 and 12-78), and its sensitivity may be improved with time-resolved or 4D techniques.139 CT venography (CTV) is useful to define dural sinus thrombosis and stenosis.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

R

C

R

D FIG 12-73  A, Axial image from a CTA demonstrating an inferior right occipital intraparenchymal hematoma with surrounding edema (long arrow) caused by a right transverse sinus dural AVF, indicated by the enlarged right occipital cortical veins (short arrow). The curved arrow points out the medial aspect of the right transverse sinus. B, The same CTA examination at a more superior level again demonstrates an enlarged medial right occipital cortical vein (straight black arrow) projecting into the intraparenchymal hematoma (white arrow). Note the small subdural hematoma lateral to the right temporal and occipital lobes (curved arrow) and the mass effect on the brainstem. C, Lateral early arterial image from a right external carotid angiogram demonstrates right occipital (short arrow) and right middle meningeal (long arrow) arterial supply to the lateral right transverse sinus dural AVF (curved arrow). D, The late arterial phase angiographic image demonstrates the direct retrograde cortical venous drainage from the dural AVF, as well as faint opacification of the transverse sinus (short arrow), indicating that this example is a Borden type II or Cognard type IV with venous ectasia (long arrow).

http://radiologyebook.com

405

406

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 12-74  Contrast-enhanced CT examination of a congenital superior sagittal sinus dural arterial venous fistula demonstrating (A) an enlarged superior sagittal sinus (long arrow) and subcortical calcifications (small arrows) and (B) an enlarged torcula (long arrow), right transverse sinus and tentorial veins (open arrow), and deep perimesencephalic veins (small arrows). Note also the subcortical and lenticular nuclei calcifications (small curved arrows). C, Lateral left external carotid angiogram illustrates left middle meningeal branches (short arrows) and left occipital branches (curved arrow) supplying the congenital superior sagittal sinus dural AVF (long arrows). Note the duplication of the superior sagittal sinus (open arrow).

MRI demonstration of cerebral AVFs depends on the type of fistula encountered. With congenital direct AVFs, MRI demonstrates one or more enlarged supplying arteries that communicate directly with solitary or multiple enlarged draining veins (see Fig. 12-72). In a posttraumatic direct carotid-cavernous fistula, MRI shows enlargement of the cavernous sinus containing diminished signal intensity owing to excessive blood flow within the sinus, and dilated superior and inferior

ophthalmic veins as well as other smaller orbital veins (Fig. 12-75). Accompanying these findings are proptosis, globe elongation, and extraocular muscle enlargement with enhancement. Dural carotidcavernous fistulas often show similar but less dramatic findings207 (Fig. 12-76). Either type of carotid-cavernous fistula may produce bilateral cavernous sinus and orbital findings on MRI because of intersinus midline venous communications. The MRI findings with other dural

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

B

C

D

407

FIG 12-75  A, Non–contrast-enhanced axial CT image illustrating an enlarged superior ophthalmic vein (arrow) in a patient with a traumatic carotid cavernous fistula. B, Noncontrast axial T1-weighted MRI of an enlarged superior ophthalmic vein (arrow) in a different patient with a traumatic carotid cavernous fistula. C, Same patient as in B, illustrating dilated compartments with increased flow in the left cavernous sinus (arrow) on the same axial T1-weighted scan owing to the carotid cavernous fistula. D, Catheter left carotid cerebral angiogram, lateral view, of the patient in B and C, documenting the rapid flow from the left internal carotid artery to the adjacent cavernous sinus (long arrow) and the enlarged superior ophthalmic vein (short arrow).

AVFs may be subtle if the more easily appreciated findings of intracranial hemorrhage, edema, or infarction are not present. MRI and MRA are notorious for missing uncomplicated dural AVFs unless enlarged supplying arteries or the venous drainage into dilated cortical or medullary veins are demonstrated.78,97,426 Venous varices may also be identified (Fig. 12-77). MRI can well image hemorrhage, edema, ischemia, infarction, and hydrocephalus due to a dural AVF.78,426 Although MRI can be used to document an occluded or stenosed dural sinus or vein,78,204,218 the evaluation can be problematic because of variable venous signal intensity caused by flowing blood within dural sinuses or draining veins on various pulse sequences204,218 and the

changing appearance of clot depending on its age (see Fig. 12-79). Contrast enhancement may make visualization of leptomeningeal and medullary fistula vascular pathology more readily apparent than on noncontrast images.204,218 Diminished ADC values with diffusionweighted imaging (DWI) may be absent unless venous ischemia or infarction are present139 (Fig. 12-78). SWI is best for visualizing abnormal draining veins139 (see Fig. 12-78). MRV can be useful to define patent, occluded, or stenotic dural sinuses; however, if a TOF instead of a PC sequence is used, clot can mimic flowing blood. MRA can illustrate dilated supplying arteries and high flow within the dural sinus or draining cortical or medullary veins78,290; however, the typical small

http://radiologyebook.com

A

B

C

D

E

F FIG 12-76  Dural carotid cavernous fistula. A, Coronal T1-weighted MRI illustrates multiple enlarged left orbital veins (arrow). B, Axial postcontrast T1-weighted MRI demonstrates enlarged medial and lateral rectus muscles (red arrows) and non-enhancing left cavernous sinus compartments due to rapid fistula flow (green arrow). C and D, Two early arterial phase lateral views from the left external carotid catheter angiogram show external carotid branches causing early opacification of the left cavernous sinus (arrow in C) and anteriorly directed flow in an enlarged left superior ophthalmic vein (arrow in D). The lateral (E) and AP (F) views of the catheter left carotid cerebral angiogram reveal the arterial supply from the internal carotid artery branches to the left cavernous sinus dura (red arrows). Venous drainage to the contralateral cavernous sinus and inferior petrosal sinus are indicatedhttp://radiologyebook.com by the black arrow and purple arrow, respectively.

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations caliber of the feeding arteries in dural AVFs may make only the more proximal arterial vascular supply appear abnormal, and slow or absent venous flow may be missed.218,290 MR digital subtraction angiography has been reported to demonstrate early venous sinus filling.78,290 Pseudocontinuous arterial spin labeling MRA may improve conspicuity of dural AVFs, as can time-resolved or 4D MRA.226,139 Both MRA and MRI may be normal when the dural shunts are located next to flowing blood and near cortical bone, making their identification difficult. Patients with retrograde cortical venous flow from a dural AVF have been documented to have increased regional cerebral blood volume, decreased regional cerebral blood flow, elevated regional oxygen extraction fraction, and impaired regional oxygen metabolism on PET217 or MR perfusion imaging.289 SPECT nuclear medicine imaging displays hypoperfused brain corresponding to edema or ischemia in areas of retrograde cortical or subarachnoid venous drainage from dural AVFs.187,471 SPECT evaluation of vasoreactivity in these same regions using acetazolamide challenge can possibly differentiate infarcted from recoverable tissue.187 Catheter cerebral angiography is always necessary for full evaluation of the angioarchitecture of cerebral AVFs139; 3D rotational catheter cerebral angiography may be contributory in some instances.139

Treatment Direct AVFs are usually treated endovascularly. Dural AVFs can be managed with transvenous or transarterial embolization and sometimes radiosurgery or conventional surgery, with radiosurgery being least effective, especially if cortical venous drainage is present.65,381,389 Treatment can reverse associated brain or spinal pathology due to venous hypertension.471

A

409

VEIN OF GALEN MALFORMATIONS Incidence and Presentation The vein of Galen malformation is a congenital anomaly in which AVFs supply the embryologic precursor to the vein of Galen. Some authors have divided vein of Galen malformations into two types: choroidal and mural. The more common choroidal type (type I) consists of one or more anterior and posterior choroidal, distal pericallosal, thalamoperforating, and sometimes superior cerebellar arteries directed to the dorsal vein of the prosencephalon, which continues into the median prosencephalic vein. With this type the posterior choroidal vessels are typically the dominant supply. The mural type of malformation (type II) consists of collicular and posterior choroidal arteries supplying the wall of the median prosencephalic vein, although additional arteries (e.g., thalamoperforators) may also be contributors. These features distinguish the vein of Galen malformation from congenital parenchymal AVMs, acquired dural AVFs, and vein of Galen or straight sinus stenoses, which also produce a dilated vein of Galen. The primitive median prosencephalic vein, the precursor of the vein of Galen, is usually abnormally dilated in vein of Galen malformations. Normal deep veins drain not into the median prosencephalic vein but through alternative routes because the true vein of Galen is absent, although this pattern is not inviolate.232 The persistent median prosencephalic vein typically drains to a falcine sinus rather than the straight sinus, and the straight sinus is often absent. Venous outflow restriction is common in the vein of Galen malformation and contributes to further dilatation of the median prosencephalic vein, also  commonly referred to as a dilated vein of Galen. Other venous anomalies are often associated with the vein of Galen malformation,

B FIG 12-77  A, Axial noncontrast CT scan demonstrating a dense left occipital mass (arrow) representing a varix and adjacent anterior edema in the left temporal occipital region. B, Axial T2-weighted MRI better illustrates the varix with central and medial high flow (medial arrow) and peripheral clot (lateral arrow). Adjacent edema is also well demonstrated. Continued

http://radiologyebook.com

410

PART II  CT and MR Imaging of the Whole Body

C

S

120 mm

D

E

FIG 12-77, cont’d  C, Sagittal noncontrast T1-weighted MRI best confirms that the inferior aspect of the varix is thrombosed (arrow) by subacute clot and that the superior aspect is patent, containing rapid flow. D, Contrast-enhanced axial fat-suppressed T1-weighted MRI best documents flow within the varix (long arrow) and flow-produced phase-encoding artifact (short arrow). E, AP view of the left external carotid angiogram, arterial phase, elucidates the anatomic features of the medial left transverse sinus dural AVF. The horizontal arrows indicate supplying branches of the left occipital artery, the thick arrow shows the supply from the left middle meningeal artery, the white arrow shows the site of the fistula, and the curved arrow shows the large draining vein varix. Because the AVF drains only to cortical veins with accompanying ectasia, it would be classified as a Borden type III or Cognard type IV.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations

A

C

B

D FIG 12-78  Three axial CTA images illustrate (A) minimal focal patency of the right transverse and sigmoid sinuses due to chronic thrombosis (red arrows), compared with the normal left transverse sinus (green arrow); (B) cortical draining vein (red arrow) indicating the presence of a dural AVF; and (C) numerous enlarged right hemisphere draining veins from the dural AVF. D, SWI highlights the small right hemisphere parenchymal hemorrhages (arrow). Continued

http://radiologyebook.com

411

412

PART II  CT and MR Imaging of the Whole Body

E

F

G FIG 12-78, cont’d  Although the DWI scan (E) demonstrates abnormal parenchymal increased signal intensity in the territory of the abnormal cortical veins (arrow), this represents venous edematous congestion and not infarction because there is no corresponding decreased signal intensity in the same region on the ADC map (F). The lateral right external carotid angiogram (G) demonstrates an isolated right sigmoid sinus dural AVF (blue arrow) supplied by branches of the right posterior auricular, occipital, and middle meningeal arteries (green arrows), with cortical venous drainage (red arrows). Classified as Cognard type III, Borden type III.

including absence of hypoplastic sigmoid sinuses and jugular veins,  duplicated sinuses, and persistent marginal or occipital sinuses.115 Venous-to-arterial shunting through a patent foramen ovale or persistent ductus arteriosus may coexist.87,244 The venous drainage of the enlarged median prosencephalic vein or vein of Galen typically demonstrates a dilated venous varix continuing into an accessory or inferior falcine sinus. If venous outflow stenosis is present and severe enough, some vein of Galen malformations have been documented to proceed to thrombosis that may lead to

spontaneous cure but also can result in acute hydrocephalus, intracranial hemorrhage, or central venous hypertension. The extent of high-flow shunting and the amount of venous outflow restriction in a vein of Galen malformation is associated with the age of the patient at presentation. Patients with high-flow multiple fistulas and no venous outflow stenosis (usually type I) are often identified at birth with high-output congestive heart failure, intracranial bruit, macrocephaly, and hydrocephalus. These patients have a grave prognosis.183 If the vein of Galen shunt is severe enough, infants can

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations demonstrate metabolic lactic acidosis due to multiorgan hypoperfusion and quickly progress to death. Patients who have decreased flow through a vein of Galen malformation, typically those with high-grade venous outflow restriction (usually type II), may present later in childhood, even as adults. In such patients, cardiomegaly may be asymptomatic, and developmental delay and ocular symptoms due to vascular-caused collicular dysfunction are more common. Seizures, enlarging head circumference, intracranial hemorrhage (that may be parenchymal, subarachnoid, or intraventricular), focal neurologic deficit, pulsatile tinnitus, and cranial bruit can be seen in any patient with a vein of Galen malformation, although hemorrhage is rare and usually seen only with venous drainage thrombosis.

Imaging Evaluation CT scans clearly demonstrate the enlarged primitive median prosencephalic vein, which appears as a mass in the tentorial incisura of slightly increased density on nonenhanced scans and markedly enhances with use of an iodinated contrast agent (Fig. 12-79). IV contrast administration also may demonstrate the feeding arteries and

A

the disproportionate outflow restriction, which is often characterized by diffuse or focal narrowing of the falcine sinus or (less commonly) the straight sinus. Nonenhanced CT scanning is valuable for identifying brain parenchymal calcifications located in the hemispheric venous watershed zones, owing to chronic deep venous hypertension. MRI provides a 3D display of the vein of Galen malformation, documenting the dilated primitive median prosencephalic vein, falcine sinus, and abnormal arterial supply (Fig. 12-80). Flow-related artifacts are related to the size of and rate of flow through the vein of Galen malformation (see Fig. 12-80). MRA and CTA can also display the enlarged vascular structures of the malformation and any venous stenoses. MRA evaluation will be compromised by the small caliber of supplying arteries and disorganized rapid flow in dilated arteries, veins, or sinuses. MRI, CT, and ultrasonography can demonstrate brain parenchymal encephalomalacia, atrophy, hemorrhage, vascular thrombosis, and ischemic changes. Each modality’s efficacy depends on the patient’s age at presentation and the corresponding degree of brain myelination or, in the case of ultrasonography, the size of the skull fontanels. Each

B

C

413

D FIG 12-79  CT examination of a vein of Galen malformation before (A) and after (B-D) contrast. In A and B, note the enlarged primitive median prosencephalic vein (short arrow) continuing to the falcine sinus (long arrow). Both structures are dense without contrast in an infant. C and D illustrate numerous enlarged arteries (arrows) supplying this choroidal type of vein of Galen malformation.

http://radiologyebook.com

414

PART II  CT and MR Imaging of the Whole Body

A

B FIG 12-80  A, Short-TE, long-TR axial MRI of a vein of Galen malformation. Note the enlarged vein of Galen (primitive median prosencephalic vein) (long arrow) producing flow artifact (short arrows). B, Sagittal T1-weighted MRI illustrates the vein of Galen malformation including the enlarged vein of Galen (primitive median prosencephalic vein) (straight arrow) and the draining falcine sinus (curved arrow). (Case courtesy Michelle Michel, MD.)

imaging technique also accurately depicts hydrocephalus. Ventricular system enlargement is due to venous hypertension and impedance of CSF resorption (communicating hydrocephalus) or less commonly to obstruction of the aqueduct of Sylvius by the enlarged primitive median prosencephalic vein or vein of Galen (obstructive hydrocephalus). In utero hydrocephalus may produce cerebral developmental dysplasia. Transuterine or transcranial ultrasound examination in the fetus or neonate, respectively, is often used to provide the initial diagnosis of vein of Galen malformations. Catheter cerebral angiography is necessary to define the angioarchitecture of the malformation, correctly categorize it, and plan therapy.

Treatment Vein of Galen malformations are typically treated using transarterial with or without transvenous embolization or combined operative or ultrasound-guided transtorcular endovascular methods.

REFERENCES 1. Aalmaani WS, Richardson AE: Multiple intracranial aneurysms: Identifying the ruptured lesion. Surg Neurol 9:303–305, 1978. 2. Abe T, Singer R, Marks M, et al: Coexistence of occult vascular malformations and developmental venous anomalies in the central nervous system: MR evaluation. AJNR Am J Neuroradiol 19:51–57, 1998. 3. Abrahams JM, Saha PK, Hurst RW, et al: Three-dimensional bone-free rendering of the cerebral circulation by use of computed tomographic angiography and fuzzy connectedness. Neurosurgery 51:264–268, Discussion 268–269, 2002.

4. Adams WM, Laitt RD, Jackson A: The role of MR angiography in the pretreatment assessment of intracranial aneurysms: A comparative study. AJNR Am J Neuroradiol 21:1618–1628, 2000. 5. Agid R, Schaaf M, Farb R: CE-MRA for follow-up of aneurysms post stent-assisted coiling. Interv Neuroradiol 18:275–283, 2012. 6. Akram H, Prezerakos G, Haliasos N, et al: Sinus pericranii: An overview and literature review of a rare cranial venous anomaly (a review of the existing literature with case examples). Neurosurg Rev 35:15–26, 2012. 7. Alons IM, van den Wijngaard IR, Verheaul RJ, et al: The value of CT angiography in patients with acute severe headache. Acta Neurol Scand 131:164–168, 2015. 8. Al-Shahi Salman R, White PM, Counsell CE, et al: Outcome after conservative management or intervention for unruptured brain arteriovenous malformations. JAMA 311:1661–1669, 2014. 9. Anzalone N, Righi C, Simionato F, et al: Three-dimensional time-offlight MR angiography in the evaluation of intracranial aneurysms treated with Guglielmi detachable coils. AJNR Am J Neuroradiol 21:746–752, 2000. 10. Anzalone N, Scomazzoni F, Cirillo M, et al: Follow-up of coiled cerebral aneurysms at 3T: Comparison of 3D time-of-flight MR angiography and contrast-enhanced MR angiography. AJNR Am J Neuroradiol 29:1530–1536, 2008. 11. Arabi B: Traumatic aneurysms of brain due to high velocity missile head wounds. Neurosurgery 22:1056–1063, 1988. 12. Atkinson JLD, Sundt TM, Jr, Houser OW, et al: Angiographic frequency of anterior circulation intracranial aneurysms. J Neurosurg 70:551–555, 1989. 13. Atlas SW: MR imaging is highly sensitive to acute subarachnoid hemorrhage. Not! Radiology 186:319–322, 1993. 14. Atlas S, editor: Intracranial vascular malformations and aneurysms: Magnetic resonance imaging of the brain and spine, ed 2, Philadelphia, 1996, Lippincott-Raven.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations 15. Atlas SW, Fram EK, Mark AS, et al: Vascular intracranial lesions: Applications of fast scanning. Radiology 169:455–461, 1988. 16. Atlas SW, Grossman RI, Goldberg HI, et al: Partially thrombosed giant intracranial aneurysms: Correlation of MR and pathologic findings. Radiology 162:111–114, 1987. 17. Atlas SW, Listerud J, Chung W, et al: Intracranial aneurysms: Depiction on MR angiograms with a multi-feature extraction post-processing algorithm. Radiology 192:129–139, 1994. 18. Atlas SW, Sheppard L, Goldberg HI, et al: Intracranial aneurysms: Detection and characterization with MR angiography with use of an advanced postprocessing technique in a blinded-reader study. Radiology 203:807–814, 1997. 19. Attali J, Benaissa A, Soize S, et al: Follow-up of intracranial aneurysms treated by flow diverter: Comparison of three-dimensional time-offlight MR angiography (3D-TOF-MRA) and contrast-enhanced MR angiography (CE-MRA) sequences with digital subtraction angiography as the gold standard. J Neurointerv Surg 2014. 20. Auger RG, Wiebers DO: Management of unruptured intracranial arteriovenous malformations: A decision analysis. Neurosurgery 30:561–569, 1992. 21. Augustyn G, Scott J, Olson E, et al: Cerebral venous angiomas: MR imaging. Radiology 156:391–395, 1985. 22. Awad I, Robinson J, Mohanty S, et al: Mixed vascular malformations of the brain: Clinical and pathogenetic considerations. Neurosurgery 33:179–188, 1993. 23. Bambakidis NC, Selman WR: Subarachnoid hemorrhage. In Suarez JI, editor: Critical care neurology and neurosurgery, Totowa, NJ, 2004, Humana Press, pp 365–378. 24. Barnwell SL, Halbach VV, Dowd CF, et al: Multiple dural arteriovenous fistulas of the cranium and spine. AJNR Am J Neuroradiol 12:441–445, 1991. 25. Barr R, Dillon W, Wilson C: Slow-flow vascular malformations in the pons: Capillary telangiectasias? AJNR Am J Neuroradiol 17:71–78, 1996. 26. Barrow D, Krisht A: Cavernous malformations and hemorrhage. In Awad I, Barrow D, editors: Cavernous malformations, Park Ridge, IL, 1993, American Association of Neurological Surgeons, pp 65–80. 27. Baumgartner J, Ater J, Ha C, et al: Pathologically proven cavernous angiomas of the brain following radiation therapy for pediatric brain tumors. Pediatr Neurosurg 39:201–207, 2003. 28. Bekelis K, Missios S, Desai A, et al: Magnetic resonance imaging/ magnetic resonance angiography fusion technique for intraoperative navigation during microsurgical resection of cerebral arteriovenous malformations. Neurosurg Focus 32:E7, 2012. 29. Berenstein A, Lasjaunias P: Surgical neuroangiography, New York, 1992, Springer-Verlag. 30. Bervini D, Morgan MK, Ritson EA, et al: Surgery for unruptured arteriovenous malformations of the brain is better than conservative management for selected cases: A prospective cohort study. J Neurosurg 121:878–890, 2014. 31. Bicknell J, Carlow T, Kornfield M, et al: Familial cavernous angiomas. Arch Neurol 35:746–749, 1978. 32. Binaghi S, Colleoni ML, Maeder P, et al: CT angiography and perfusion CT in cerebral vasospasm after subarachnoid hemorrhage. AJNR Am J Neuroradiol 28:750–758, 2007. 33. Biondi A, Clemenceau S, Dormont D, et al: Intracranial extra-axial cavernous (HEM) angiomas: Tumors or vascular malformations?   J Neuroradiol 29:91–104, 2002. 34. Bisdorff A, Mulliken JB, Carrico J, et al: Intracranial vascular anomalies in patients with periorbital lymphatic and lymphaticovenous malformations. AJNR Am J Neuroradiol 28:335–341, 2007. 35. Blackmore CC, Mamourian AC: Compression from venous angioma: MR findings. AJNR Am J Neuroradiol 17:458–460, 1996. 36. Blatter DD, Parker DL, Ahn SS, et al: Cerebral MR angiography with multiple overlapping thin slab acquisition. Part II: Early clinical experience. Radiology 183:379–389, 1992. 37. Blatter DD, Parker DL, Robison RO: Cerebral MR angiography with multiple overlapping thin slab acquisition. Part I: Quantitative analysis of vessel visibility. Radiology 179:805–811, 1991.

415

38. Bodanapally UK, Saksobhavivat N, Shanmuganathan K, et al:   Arterial injuries after penetrating brain injury in civilians: Risk factors on admission head computed tomography. J Neurosurg 122:219–226, 2015. 39. Bodanapally UK, Shanmuganathan K, Boscak AR, et al: Vascular complications of penetrating brain injury: Comparison of helical CT angiography and conventional angiography. J Neurosurg 121:1275–1283, 2014. 40. Boddu SR, Tong FC, Dehkharghani S, et al: Contrast-enhanced time-resolved MRA for follow-up of intracranial aneurysms treated with the pipeline embolization device. AJNR Am J Neuroradiol 35:2112–2118, 2014. 41. Bollar A, Allut AG, Prieto A, et al: Sinus pericranii: Radiological and etiopathological considerations. J Neurosurg 77:469–472, 1992. 42. Borden JA, Wu JK, Shucart WA: A proposed classification for spinal and cranial dural arteriovenous fistulous malformations and implications for treatment. J Neurosurg 82:166–179, 1995. 43. Boulin A, Pierot L: Follow-up of intracranial aneurysms treated with detachable coils: Comparison of gadolinium-enhanced 3D time-offlight MR angiography and digital subtraction angiography. Radiology 219:108–113, 2001. 44. Bourgouin PM, Tampieri D, Johnston W, et al: Multiple occult vascular malformations of the brain and spinal cord: MRI diagnosis. Neuroradiology 34:110–111, 1992. 45. Bradley WG, Waluch V: Blood flow: Magnetic resonance imaging. Radiology 154:443–450, 1985. 46. Brisman JL, Song JK, Newell DW: Cerebral aneurysms. N Engl J Med 355:928–939, 2006. 47. Broderick JP, Brott TG, Duldner JE, et al: Initial and recurrent bleeding are the major causes of death following subarachnoid hemorrhage. Stroke 25:1342–1347, 1994. 48. Brott T, Mandybur TI: Case-control study of clinical outcomes after aneurysmal subarachnoid hemorrhage. Neurosurgery 19:891–895, 1986. 49. Brown RD, Jr, Wiebers DO, Forbes G, et al: The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 68:352–357, 1988. 50. Brown RD, Wiebers DO, Forbes GS: Unruptured intracranial aneurysms and arteriovenous malformations: Frequency of   intracranial hemorrhage and relationship of lesions. J Neurosurg 73:859–863, 1990. 51. Brown RD, Jr, Wiebers DO, Nichols DA: Intracranial dural arteriovenous fistulae: Angiographic predictors of intracranial hemorrhage and clinical outcome in nonsurgical patients. J Neurosurg 81:531–538, 1994. 52. Bull J: Massive aneurysms at the base of the brain. Brain 92:535–570, 1969. 53. Bulsara KR, Alexander MJ, Villavicencio AT, et al: De novo cerebral arteriovenous malformation: Case report. Neurosurgery 50:1137–1140, 2002. 54. Burger PC, Schuthauer BW: Vascular tumors and tumor-like lesions. In Tumors of the central nervous system, Washington, DC, 1994, Armed Forces Institute of Pathology, pp 287–299. 55. Burger PC, Schuthauer BW, Vogel FS: Surgical pathology of the nervous system and its coverings, ed 3, New York, 1991, Churchill Livingstone. 56. Burrow AM, Link MJ, Pollock BE: Is stereotactic radiosurgery   the best treatment option for patients with a radiosurgery-based arteriovenous malformation score < 1? World Neurosurg 82: 1144–1147, 2014. 57. Cannestra AF, Pouratian N, Forage J, et al: Functional magnetic resonance imaging and optical imaging for dominant-hemisphere perisylvian arteriovenous malformations. Neurosurgery 55:804–812, 2004. 58. Caruso R, Colonnese C, Elefante A, et al: Use of spiral computerized tomography angiography in patients with cerebral aneurysm. Our experience. J Neurosurg Sci 46:4–9, 2002. 59. Cave-Riant F, Denier C, Labauge P, et al: Spectrum and expression analysis of KRIT1 mutations in 121 consecutive and unrelated patients

http://radiologyebook.com

416

60. 61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71. 72. 73.

74.

75.

76.

77.

78.

79.

80.

PART II  CT and MR Imaging of the Whole Body with cerebral cavernous malformations. Eur J Hum Genet 10:733–740, 2002. Chalouhi N, Hoh BL, Hasan D: Review of cerebral aneurysm formation, growth, and rupture. Stroke 44:3613–3622, 2013. Chang W, Loecher MW, Wu Y, et al: Hemodynamic changes in patients with arteriovenous malformations assessed using high-resolution 3D radial phase-contrast MR angiography. AJNR Am J Neuroradiol 33:1565–1572, 2012. Chappell PM, Steinberg GK, Marks MP: Clinically documented hemorrhage in cerebral arteriovenous malformations: MR characteristics. Radiology 183:719–724, 1992. Chaudhary SR, Ko N, Dillon WP, et al: Prospective evaluation of multidetector-row CT angiography for the diagnosis of vasospasm following subarachnoid hemorrhage: A comparison with digital subtraction angiography. Cerebrovasc Dis 25:144–150, 2008. Chaudhary MY, Sachdev VP, Cho SH, et al: Dural arteriovenous malformations of the major venous sinuses: An acquired lesion. AJNR Am J Neuroradiol 3:13–19, 1982. Chen CJ, Lee CC, Ding D, et al: Stereotactic radiosurgery for intracranial dural arteriovenous fistulas: A systematic review. J Neurosurg 122:353–362, 2015. Chen H, Li MH: A1 segment hypoplasia accompanied by AcomA aneurysms assessed with magnetic resonance angiography. Surg Radiol Anat 36:353–357, 2014. Chen YC, Sun ZK, Li MH, et al: The clinical value of MRA at 3.0 T for the diagnosis and therapeutic planning of patients with subarachnoid haemorrhage. Eur Radiol 22:1404–1412, 2012. Cheng B, Cai W, Sun C, et al: 3D bone subtraction CT angiography for the evaluation of intracranial aneurysms: A comparison study with 2D bone subtraction CT angiography and conventional non-subtracted CT angiography. Acta Radiol 56:1127–1134, 2014. Chin LS, Raffel C, Gonzales-Gomez I, et al: Diffuse arteriovenous malformations: A clinical, radiological, and pathological description. Neurosurgery 31:863–868, 1992. Cho WS, Kim SS, Lee SJ, et al: The effectiveness of 3T time-of-flight magnetic resonance angiography for follow-up evaluations after the stent-assisted coil embolization of cerebral aneurysms. Acta Radiol 55:604–613, 2014. Chong W, Patel H, Holt M: Developmental venous anomalies (DVA): What are they really? Neuroradiol J 24:59–70, 2011. Chung SJ, Kim JS, Kim JC, et al: Intracranial dural arteriovenous fistulas: Analysis of 60 patients. Cerebrovasc Dis 13:79–88, 2002. Cigoli MS, Avemaria F, De Benedetti S, et al: PDCD10 gene mutations in multiple cerebral cavernous malformations. PLoS ONE 9:e110438, 2014. Cirillo M, Scomazzoni F, Cirillo L, et al: Comparison of 3D TOF-MRA and 3D CE-MRA at 3T for imaging of intracranial aneurysms. Eur J Radiol 82:e853–e859, 2013. Claassen J, Bernardini G, Kreiter K, et al: Effect of cisternal   and ventricular blood on risk of delayed cerebral ischemia   after subarachnoid hemorrhage. Stroke 32:2012–2020, 2011. Cognard C, Gobin YP, Picrot L, et al: Cerebral dural arteriovenous fistulas: Clinical and angiographic correlation with a revised classification of venous drainage. Radiology 194:671–680, 1995. Cohen-Gadol AA, Jacob JT, Edwards DA, et al: Coexistence of intracranial and spinal cavernous malformations: A study of prevalence and natural history. J Neurosurg 104:376–384, 2006. Coley SC, Romanowski CA, Hodgson TJ, et al: Dural arteriovenous fistulae: Noninvasive diagnosis with dynamic MR digital subtraction angiography. AJNR Am J Neuroradiol 23:404–407, 2002. Connolly HM, Huston J, III, Brown RD, Jr, et al: Intracranial aneurysms in patients with coarctation of the aorta: A prospective magnetic resonance angiographic study of 100 patients. Mayo Clin Proc 78:1491–1499, 2003. Connolly ES, Jr, Rabinstein AA, et al: Guidelines for the management   of aneurysmal subarachnoid hemorrhage. Stroke 43:1711–1737, 2012.

81. Cook SC, Hickey J, Maul TM, et al: Assessment of the cerebral circulation in adults with coarctation of the aorta. Congenit Heart Dis 8:289–295, 2013. 82. Cottier JP, Bleuzen-Couthon A, Gallas S, et al: Intracranial aneurysms treated with Guglielmi detachable coils: Is contrast material necessary in the follow-up with 3D time-of-flight MR angiography. AJNR Am J Neuroradiol 24:1797–1803, 2003. 83. Coupe NJ, Athwal RK, Marshman LA, et al: Subarachnoid hemorrhage emanating from a ruptured infundibulum: Case report and literature review. Surg Neurol 67:204–206, 2007. 84. Crawford PM, West CR, Chadwick DW, et al: Arteriovenous malformations of the brain: Natural history in unoperated patients. J Neurol Neurosurg Psychiatry 49:1–10, 1986. 85. Creasy J, Price RR, Presbrey T, et al: Gadolinium-enhanced MR angiography. Radiology 175:280–283, 1990. 86. Cremers CH, van der Schaaf IC, Wensink E, et al: CT perfusion and delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis. J Cereb Blood Flow Metab 34:200–207, 2014. 87. Cumming GR: Circulation in neonates with intracranial arteriovenous fistula and cardiac failure. Am J Cardiol 45:1019–1024, 1980. 88. Cunha e Sa MJ, Stein BM, Solomon RA, et al: The treatment of associated intracranial aneurysms and arteriovenous malformation. J Neurosurg 77:853–859, 1992. 89. Curling O, Kelly D: The natural history of intracranial cavernous and venous malformations. Persp Neurol Surg 1:19–39, 1990. 90. Curtis SL, Bradley M, Wilde P, et al: Results of screening for intracranial aneurysms in patients with coarctation of the aorta. AJNR Am J Neuroradiol 33:1182–1186, 2012. 91. Dankbaar JW, de Rooij NK, Rijsdijk M, et al: Diagnostic threshold values of cerebral perfusion measured with computed tomography for delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage. Stroke 41:1927–1932, 2010. 92. Dankbaar JW, Rijsdijk M, van der Schaar IC, et al: Relationship between vasospasm, cerebral perfusion, and delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage. Neuroradiology 51:813– 819, 2009. 93. de Gast AN, van Rooij WJ, Sluzewski M: Fenestrations of the anterior communicating artery: Incidence on 3D angiography and relationship to aneurysms. AJNR Am J Neuroradiol 29:296–298, 2008. 94. Dehdashti AR, Rufenacht DA, Delavelle J, et al: Therapeutic decision and management of aneurysmal subarachnoid haemorrhage based   on computed tomographic angiography. Br J Neurosurg 17:46–53, 2003. 95. Dehkharghani S, Dillon WP, Bryant SO, et al: Unilateral calcification of the caudate and putamen: Association with underlying developmental venous anomaly. AJNR Am J Neuroradiol 31:1848–1852, 2010. 96. Delgado AJE, Jagadeesan BD, Refai D, et al: Diagnostic yield of computed tomography angiography and magnetic resonance angiography in patients with catheter angiography-negative subarachnoid hemorrhage. J Neurosurg 117:309–315, 2012. 97. De Marco JK, Dillon WP, Halback VV, et al: Dural arteriovenous fistulas: Evaluation with MR imaging. Radiology 175:193–199, 1990. 98. Denier C, Labauge P, Brunereau L, et al: Clinical features of cerebral cavernous malformations in patients with KRIT1 mutations. Ann Neurol 55:213–220, 2004. 99. Derdeyn CP, Graves VB, Turski PA, et al: MR angiography of saccular aneurysms after treatment with Guglielmi detachable coils: Preliminary experience. AJNR Am J Neuroradiol 18:279–286, 1997. 100. Dorsch N: A clinical review of cerebral vasospasm and delayed ischaemia following aneurysm rupture. Acta Neurochir Suppl 110(Pt 1):5–6, 2011. 101. Dumoulin CL, Souza SP, Walker MF, et al: Three-dimensional phase contrast angiography. Magn Reson Med 9:139–149, 1989. 102. Duran M, Schoenberg SO, Yuh WT, et al: Cerebral arteriovenous malformations: Morphologic evaluation by ultrashort 3D   gadolinium-enhanced MR angiography. Eur Radiol 12:2957–2964, 2002.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations 103. Ebeling JD, Tranmer BI, Davis KA, et al: Thrombosed arteriovenous malformations: A type of occult vascular malformation. Neurosurgery 23:605–610, 1988. 104. Eddleman CS, Jeong HJ, Hurley MC, et al: 4D radial acquisition contrast-enhanced MR angiography and intracranial arteriovenous malformations: Quickly approaching digital subtraction angiography. Stroke 40:2749–2753, 2009. 105. Edelman RR, Ahn SS, Chien D, et al: Improved time-of-flight MR angiography of the brain with magnetization transfer contrast. Radiology 184:395–399, 1992. 106. Fabrikant J, Levy R, Steinberg G, et al: Charged particle radiosurgery for intracranial vascular malformations. Neurosurg Clin North Am 3:99–139, 1992. 107. Farb RI, Nag S, Scott JN, et al: Surveillance of intracranial aneurysms treated with detachable coils: A comparison of MRA techniques. Neuroradiology 47:507–515, 2005. 108. Firouzian A, Manniesing R, Metz CT, et al: Quantification of intracranial aneurysm morphodynamics from ECG-gated CT angiography. Acad Radiol 20:52–58, 2013. 109. Fisher C, Kistler J, Davis J: Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery 6:1–9, 1980. 110. Friedman AH, Drake CG: Subarachnoid hemorrhage from intracranial dissecting aneurysm. J Neurosurg 60:325–334, 1984. 111. Frischer JM, God S, Gruber A, et al: Susceptibility-weighted imaging   at 7T: Improved diagnosis of cerebral cavernous malformations and associated development venous anomalies. Neuroimage Clin 1:116–120, 2012. 112. Fujisawa I, Asato R, Nishimura K, et al: Moyamoya disease: MR imaging. Radiology 164:103–105, 1987. 113. Fults D, Kelly DL: Natural history of arteriovenous malformations of the brain: A clinical study. Neurosurgery 15:658–662, 1984. 114. Gabikian P, Clatterbuck RE, Gailloud P, et al: Developmental venous anomalies and sinus pericranii in the blue rubber-bleb nevus syndrome. Case report. J Neurosurg 99:409–411, 2003. 115. Garcia-Monaco R, Lasjaunias P, Berenstein A: Therapeutic management of vein of Galen aneurysmal malformations in interventional neuroradiology. In Vinvela F, Halback VV, Dion JE, editors: Endovascular therapy of the central nervous system, New York, 1992, Raven Press. 116. Garcia-Monaco R, Rodesch G, Alvarez H, et al: Pseudoaneurysms within ruptured intracranial arteriovenous malformations: Diagnosis and early endovascular management. AJNR Am J Neuroradiol 14:315–321, 1993. 117. Gauvrit JY, Leclerc X, Caron S, et al: Intracranial aneurysms treated with Guglielmi detachable coils: Imaging follow-up with contrastenhanced MR angiography. Stroke 37:1033–1037, 2006. 118. Gauvrit JY, Leclerc X, Pernodet M, et al: Intracranial aneurysms treated with Guglielmi detachable coils: Usefulness of 6-month imaging follow-up with contrast-enhanced MR angiography. AJNR Am J Neuroradiol 26:515–521, 2005. 119. Giesel FL, Essig M, Zabel-Du-Bois A, et al: High-contrast computed tomographic angiography better detects residual intracranial arteriovenous malformations in long-term follow-up after radiotherapy than 1.5-tesla time-of-flight magnetic resonance angiography. Acta Radiol 51:64–70, 2010. 120. Gibbs GF, Huston J, III, Bernstein MA, et al: Improved image quality of intracranial aneurysms: 3.0-T versus 1.5-T time-of-flight MR angiography. AJNR Am J Neuroradiol 25:84–87, 2004. 121. Gibbs GF, Huston J, 3rd, Bernstein MA, et al: 3.0-tesla MR   angiography of intracranial aneurysms: Comparison of time-of-flight and contrast-enhanced techniques. J Magn Reson Imaging 21:97–102, 2005. 122. Gibbs GF, Huston J, 3rd, Qian Q, et al: Follow-up of intracranial aneurysms in autosomal-dominant polycystic kidney disease. Kidney Int 65:1621–1627, 2004. 123. Goldsher D, Shreiber R, Shik V, et al: Role of multisection CT angiography in the evaluation of vertebrobasilar vasospasm in patients

124. 125.

126.

127. 128.

129.

130.

131.

132.

133. 134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

417

with subarachnoid hemorrhage. AJNR Am J Neuroradiol 25:1493–1498, 2004. Golfinos JG, Wascher TM, Zabramski JM, et al: The management of unruptured intracranial vascular malformations. BNI Q 8:2–11, 1992. Golitz P, Struffert T, Kaschka I, et al: Optimized angiographic CT using intravenous contrast injection: A noninvasive imaging option for the follow-up of coiled aneurysms? AJNR Am J Neuroradiol 35:2341–2347, 2014. Gomori JM, Grossman RI, Bilaniuk LT, et al: High-field MR imaging of superficial siderosis of the central nervous system. J Comput Assist Tomogr 9:972–975, 1985. Gomori JM, Grossman RI, Goldberg HI, et al: Occult cerebral vascular malformations: High-field MR imaging. Radiology 158:707–713, 1986. Gönner F, Heid O, Remonda L, et al: MR angiography with ultrashort echo time in cerebral aneurysms treated with Guglielmi detachable coils. AJNR Am J Neuroradiol 19:1324–1328, 1998. Gonzalez CF, Cho YI, Ortega HV, et al: Intracranial aneurysms: Flow analysis of their origin and progression. AJNR Am J Neuroradiol 13:181–188, 1992. Gonzalez LF, Walker MT, Zabramski JM, et al: Distinction between paraclinoid and cavernous sinus aneurysms with computed tomographic angiography. Neurosurgery 52:1131–1137, 2003. Goulao A, Alvarez H, Monaco RG, et al: Venous anomalies and abnormalities of the posterior fossa. Neuroradiology 31:476–482, 1990. Graf CJ, Perret GE, Torner JC: Bleeding from cerebral arteriovenous malformations as part of their natural history. J Neurosurg 58:331–337, 1990. Graves VB, Duff TA: Intracranial arteriovenous malformations: Current imaging and treatment. Invest Radiol 25:952–960, 1990. Greenberg ED, Gold R, Reichman M, et al: Diagnostic accuracy of CT angiography and CT perfusion for cerebral vasospasm: a meta-analysis. AJNR Am J Neuroradiol 31:1853–1860, 2010. Grieve JP, Stacey R, Moore E, et al: Artefact of MRA following aneurysm clipping: An in vitro study and prospective comparison with conventional angiography. Neuroradiology 41:680–686, 1999. Gross BA, Frerichs KU, Du R: Sensitivity of CT angiography, T2-weighted MRI, and magnetic resonance angiography in detecting cerebral arteriovenous malformations and associated aneurysms. J Clin Neurosci 19:1093–1095, 2012. Grossman RI, Kemp SS, Yu IC, et al: The importance of oxygenation in the appearance of acute subarachnoid hemorrhage on high-field magnetic resonance imaging. Acta Radiol Suppl 369:56–58, 1986. Guo WY, Wu YT, Wu HM, et al: Toward normal perfusion after radiosurgery: Perfusion MR imaging with independent component analysis of brain arteriovenous malformations. AJNR Am J Neuroradiol 25:1636–1644, 2004. Hacein-Bey L, Konstas AA, Pile-Spellman J: Natural history, current concepts, classification, factors impacting endovascular therapy, and pathophysiology of cerebral and spinal dural arteriovenous fistulas. Clin Neurol Neurosurg 121:64–75, 2014. Hacein-Bey L, Provenzale JM: Current imaging assessment and treatment of intracranial aneurysms. AJR Am J Roentgenol 196:32–44, 2011. Hackney DB, Lesnick JE, Zimmerman RA, et al: MR identification of bleeding site in subarachnoid hemorrhage with multiple intracranial aneurysms. J Comput Assist Tomogr 10:878–880, 1986. Hadizadeh DR, Kukuk GM, Steck DT, et al: Noninvasive evaluation of cerebral arteriovenous malformations by 4D-MRA for preoperative planning and postoperative follow-up in 56 patients: Comparison with DSA and intraoperative findings. AJNR Am J Neuroradiol 33:1095– 1101, 2012. Hahn FJ, Chg E, McComb R, et al: Peripheral signal void ring in giant vertebral aneurysm: MR and pathology findings. J Comp Assist Tomogr 10:1036–1038, 1986. Hammond D, Beauchamp N, Wityk R, et al: Ischemic complication of a cerebral developmental venous anomaly: Case report and review of the literature. J Comput Assist Tomogr 26:633–636, 2002.

http://radiologyebook.com

418

PART II  CT and MR Imaging of the Whole Body

145. Hargraves RW, Spetzler RF: Takayasu’s arteritis: Case report. BNI Q 7:20–23, 1991. 146. Hartmann A, Mast H, Mohr JP, et al: Morbidity of intracranial hemorrhage in patients with cerebral arteriovenous malformations. Stroke 29:931–934, 1998. 147. Hashimoto K, Nozaki K, Hashimoto N: Optic strut as a radiographic landmark in evaluating neck location of a paraclinoid aneurysm. Neurosurgery 59:880–895, 2006. 148. Hasuo K, Tamura S, Kudo S, et al: Moyamoya disease: The use of   digital subtraction angiography in its diagnosis. Radiology 157:107–111, 1985. 149. Hattingen E, Blasel S, Dumesnil R, et al: MR angiography in patients with subarachnoid hemorrhage: Adequate to evaluate vasospasminduced vascular narrowing? Neurosurg Rev 33:431–439, 2010. 150. Hayakawa M, Tanaka T, Sadato A, et al: Detection of pulsation in unruptured cerebral aneurysms by ECG-gated 3D-CT angiography (4D-CTA) with 320-row area detector CT (ADCT) and follow-up evaluation results: Assessment based on heart rate at the time of scanning. Clin Neuroradiol 24:145–150, 2014. 151. Hayashida E, Sasao A, Hirai T, et al: Can sufficient preoperative information of intracranial aneurysms be obtained by using 320-row detector CT angiography alone? Jpn J Radiol 31:600–607, 2013. 152. Healton EB: Fibromuscular dysplasia. In Barnett HJM, editor: Stroke (vol 2), New York, 1986, Churchill Livingstone, pp 831–843. 153. Herneshiemi J, Tapaninaho A, Vapalahti M, et al: Saccular aneurysms of the distal anterior cerebral artery and its branches. Neurosurgery 31:994–999, 1992. 154. Heros RC, Korosue K: Arteriovenous malformations of the brain. Curr Opin Neurol Neurosurg 3:63–67, 1990. 155. Hickmann AK, Langner S, Kirsch M, et al: The value of perfusion computed tomography in predicting clinically relevant vasospasm in patients with aneurysmal subarachnoid hemorrhage. Neurosurg Rev 36:267–278, 2013. 156. Hijdra A, Brouwers P, Vermeulen M, et al: Grading the amount of blood on computed tomograms after subarachnoid hemorrhage. Stroke 21:1156–1161, 1990. 157. Hofmann E, Becker T, Romberg-Hahnloser R, et al: Cranial MRI and CT in patients with left atrial myxoma. Neuroradiology 34:57–61, 1992. 158. Hoffmann KR, Hosten N, Liebig T, et al: Giant aneurysm of the vertebral artery in neurofibromatosis type 1: Report of a case and review of the literature. Neuroradiology 40:245–248, 1998. 159. Hoh BL, Cheung AC, Rabinov JD, et al: Results of a prospective protocol of computed tomographic angiography in place of catheter angiography as the only diagnostic and pretreatment planning study for cerebral aneurysms by a combined neurovascular team. Neurosurgery 54:1329–1340, 2004. 160. Hop JW, Rinkel GJ, Algra A, et al: Case-fatality rates and functional outcome after subarachnoid hemorrhage: A systematic review. Stroke 28:660–664, 1997. 161. Houser O, Baker HJ, Rhoton AJ, et al: Intracranial dural arteriovenous malformations. Radiology 163:55–64, 1972. 162. Houser O, Campbell J, Campbell R, et al: Arteriovenous malformation affecting the transverse dural venous sinus: An acquired lesion. Mayo Clin Proc 54:651–661, 1979. 163. Hu X, Alperin N, Levin DN, et al: Visualization of MR angiographic data with segmentation and volume-rendering techniques. J Magn Reson Imaging 1:539–546, 1991. 164. Hunt WE, Hess RM: Surgical risks as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg 28:14–20, 1968. 165. Huston J, III, Rufenacht DA, Ehman RL, et al: Intracranial aneurysms and vascular malformations: Comparison of time-of-flight and phase-contrast MR angiography. Radiology 181:721–730, 1991. 166. Ida M, Kurisu Y, Yamashita M: MR angiography of ruptured aneurysms in acute subarachnoid hemorrhage. AJNR Am J Neuroradiol 18:1025– 1032, 1997. 167. Igase M, Igase K, Kohara K, et al: Visit-to-visit variability in systolic blood pressure is a novel risk factor for the growth of intracranial aneurysms. Cerebrovasc Dis 36:401–406, 2013.

168. Imbesi SG, Kerber CW: Analysis of slipstream flow in two ruptured intracranial cerebral aneurysms. AJNR Am J Neuroradiol 20:1703–1705, 1999. 169. Inagawa T, Hirano A: Autopsy study of unruptured incidental intracranial aneurysms. Surg Neurol 34:361–365, 1990. 170. Inoue S, Hosoda K, Fujita A, et al: Utility of 320-detector row CT for diagnosis and therapeutic strategy for paraclinoid and intracavernous aneurysm. Acta Neurochir (Wien) 156:505–514, 2014. 171. International Study of Unruptured Intracranial Aneurysms Investigators: Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 362:103–110, 2003. 172. Isaksen J, Egge A, Waterloo K, et al: Risk factors for aneurysmal subarachnoid haemorrhage: The Tromso study. J Neurol Neurosurg Psychiatry 73:185–187, 2002. 173. Itoh D, Aoki S, Maruyama K, et al: Corticospinal tracts by diffusion tensor tractography in patients with arteriovenous malformations. J Comput Assist Tomogr 30:618–623, 2006. 174. Jacobs IG, Roszler MH, Kelly JK, et al: Cocaine abuse: Neurovascular complications. Radiology 170:223–227, 1989. 175. Jager HR, Ellamushi H, Moore EA, et al: Contrast-enhanced MR angiography of intracranial giant aneurysms. AJNR Am J Neuroradiol 21:1900–1907, 2000. 176. Jakimovski D, Schneider H, Frei K, et al: Bleeding propensity of cavernous malformations: Impact of tight junction alterations   on the occurrence of overt hematoma. J Neurosurg 121:613–620, 2014. 177. Jane JA, Kassell NF, Torner JC, et al: The natural history of aneurysms and arteriovenous malformations. J Neurosurg 62:321–323, 1985. 178. Jayaraman MV, Mayo-Smith WW, Tung GA, et al: Detection of intracranial aneurysms: Multi-detector row CT angiography compared with DSA. Radiology 230:510–518, 2004. 179. Jeon JS, Kim JE, Chung YS, et al: A risk factor analysis of prospective symptomatic haemorrhage in adult patients with cerebral cavernous malformation. J Neurol Neurosurg Psychiatry 85:1366–1370, 2014. 180. Jenkins A, Hadley DM, Teasdale GM, et al: Magnetic resonance   imaging of acute subarachnoid hemorrhage. J Neurosurg 68:731–736, 1988. 181. Jiang T, Wang P, Qian Y, et al: A follow-up study of autosomal dominant polycystic kidney disease with intracranial aneurysms using 3.0 T three-dimensional time-of-flight magnetic resonance angiography. Eur J Radiol 82:1840–1845, 2013. 182. Jinkins JR, Dadsetan MR, Sener RN, et al: Value of acute-phase angiography in the detection of vascular injuries caused by gunshot wounds to the head: Analysis of 12 cases. AJNR Am J Neuroradiol 169:365–368, 1992. 183. Jones BV, Ball WS, Tomsick TA, et al: Vein of Galen aneurismal malformation: Diagnosis and treatment of 13 children with extended clinical follow-up. AJNR Am J Neuroradiol 23:1717–1724, 2002. 184. Jung KH, Chu K, Jeong SW, et al: Cerebral cavernous malformations with dynamic and progressive course: Correlation study with vascular endothelial growth factor. Arch Neurol 60:1613–1618, 2003. 185. Juvela S, Porras M, Heiskanen O: Natural history of unruptured intracranial aneurysms: A long-term follow-up study. J Neurosurg 79:174–182, 1993. 186. Juvela S, Poussa K, Lehto H, et al: Natural history of unruptured intracranial aneurysms. A long-term follow-up study. Stroke 44:2414– 2421, 2013. 187. Kai Y, Hamada J, Morioka M, et al: Pre- and post-treatment MR imaging and single photon emission CT in patients with dural arteriovenous fistulas and retrograde leptomeningeal venous drainage. AJNR Am J Neuroradiol 24:619–625, 2003. 188. Kale SP, Edgell RC, Alshekhlee A, et al: Age-associated vasospasm in aneurysmal subarachnoid hemorrhage. J Stroke Cerebrovasc Dis 22:22–27, 2013. 189. Kallmes DF, Clark HP, Dix JE, et al: Ruptured vertebrobasilar aneurysms: Frequency of the nonaneurysmal perimesencephalic pattern of hemorrhage on CT scans. Radiology 201:657–660, 1996.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations 190. Kalra VB, Wu X, Matouk CC, et al: Use of follow-up imaging in isolated perimesencephalic subarachnoid hemorrhage: A meta-analysis. Stroke 46:401–406, 2015. 191. Kanaan I, Jallu A, Alwatban J, et al: Extra-axial cavernous hemangioma: Two case reports. Skull Base 11:287–295, 2001. 192. Kanal E, Shellock FG, Lewin JS: Aneurysm clip testing for ferromagnetic properties: Clip variability issues. Radiology 200:576–578, 1996. 193. Kanazawa R, Kato M, Ishikawa K, et al: Convenience of the computed tomography perfusion method for cerebral vasospasm detection after subarachnoid hemorrhage. Surg Neurol 67:604–611, 2007. 194. Kangasniemi M, Makela T, Koskinen S, et al: Detection of intracranial aneurysms with two-dimensional and three-dimensional multislice helical computed tomographic angiography. Neurosurgery 54:336–340, 2004. 195. Kaplan HA, Aronson SM, Browder EJ: Vascular malformations of the brain: An anatomical study. J Neurosurg 27:630–635, 1961. 196. Kassel NJ, Torner JC: Size of intracranial aneurysms. Neurosurgery 12:291–297, 1983. 197. Katayama Y, Tsubokawa T, Miyasaki S, et al: Magnetic resonance imaging of cavernous sinus cavernous hemangioma. Neuroradiology 33:118–122, 1991. 198. Kaufmann TJ, Huston J, 3rd, Mandrekar JN, et al: Complications of diagnostic cerebral angiography: Evaluation of 19,826 consecutive patients. Radiology 243:812–819, 2007. 199. Kelliny M, Maeder P, Binaghi S, et al: Cerebral aneurysm exclusion by CT angiography based on subarachnoid hemorrhage pattern: A retrospective study. BMC Neurol 11:8, 2011. 200. Khosravani H, Mayer SA, Demchuk A, et al: Emergency noninvasive angiography for acute intracerebral hemorrhage. AJNR Am J Neuroradiol 34:1481–1487, 2013. 201. Kidoh M, Nakaura T, Ogata T, et al: Subtracted 3D CT angiography for the evaluation of intracranial aneurysms in 256-slice multidetector CT: Usefulness of the 80-kVp plus compact contrast medium bolus protocol. Eur Radiol 23:3012–3019, 2013. 202. Kileen RP, Gupta A, Delaney H, et al: Appropriate use of CT perfusion following aneurysmal subarachnoid hemorrhage: A bayesian analysis approach. AJNR Am J Neuroradiol 35:459–465, 2014. 203. Kim H, Al-Shahi Salman R, McCulloch CE, et al: Untreated brain arteriovenous malformation: Patient-level meta-analysis of hemorrhage predictors. Neurology 83:590–597, 2014. 204. Kitajima M, Hirai T, Korogi Y, et al: Retrograde cortical and deep venous drainage in patients with intracranial dural arteriovenous fistulas: Comparison of MR imaging and angiographic findings. AJNR Am J Neuroradiol 26:1532–1538, 2005. 205. Kjellberg RN, Hanamura T, Davis KR, et al: Bragg-peak proton beam therapy for arteriovenous malformations of the brain. N Engl J Med 309:269–274, 1983. 206. Klucznik R, Carrier D, Pyka R, et al: Placement of a ferromagnetic intracerebral aneurysm clip in a magnetic field with a fatal outcome. Radiology 187:612–614, 1993. 207. Komiyama M, Fu Y, Yagura H, et al: MR imaging of dural AV fistulas at the cavernous sinus. J Comp Assist Tomogr 14:397–401, 1990. 208. Konan AV, Raymond J, Bourgouin P, et al: Cerebellar infarct caused by spontaneous thrombosis of a developmental venous anomaly of the posterior fossa. AJNR Am J Neuroradiol 20:256–258, 1999. 209. Kondiziolka D, Nixon BJ, Lasjaunas P, et al: Cerebral arteriovenous malformations with associated arterial aneurysms: Hemodynamic and therapeutic considerations. Can J Neurol Sci 15:130–134, 1988. 210. Kondziolka D, Monaco EA, 3rd, Lunsford LD: Cavernous malformations and hemorrhage risk. Prog Neurol Surg 27:141–146, 2013. 211. Korogi Y, Takahashi M, Mabuchi N, et al: Intracranial aneurysms: Diagnostic accuracy of the three dimensional, Fourier transform, time-of-flight MR angiography. Radiology 193:181–186, 1994. 212. Krief O, Sichez JP, Chedid G, et al: Extraaxial cavernous hemangioma with hemorrhage. AJNR Am J Neuroradiol 12:988–990, 1990. 213. Kubo Y, Koji T, Kashimura H, et al: Female sex as a risk factor for the growth of asymptomatic unruptured cerebral saccular aneurysms in elderly patients. J Neurosurg 121:599–604, 2014.

419

214. Kukuk GM, Hadizadeh DR, Bostrom A, et al: Cerebral arteriovenous malformations at 3.0T: Intraindividual comparative study of 4D-MRA in combination with selective arterial spin labeling and digital subtraction angiography. Invest Radiol 45:126–132, 2010. 215. Kumar AJ, Vinuela F, Fox AJ, et al: Unruptured intracranial arteriovenous malformations do cause mass effect. AJNR Am J Neuroradiol 6:29–32, 1985. 216. Kupersmith MJ, Kalish H, Epstein F, et al: Natural history of brainstem cavernous malformations. Neurosurgery 48:47–53, 2001. 217. Kuroda S, Furukawa K, Shiga T, et al: Pretreatment and posttreatment evaluation of hemodynamic and metabolic parameters in intracranial dural arteriovenous fistulae with cortical venous reflux. Neurosurgery 54:585–591, 2004. 218. Kwon BJ, Han MH, Kang HS, et al: MR imaging findings of intracranial dural arteriovenous fistulas: Relations with venous drainage patterns. AJNR Am J Neuroradiol 26:2500–2507, 2005. 219. Labauge P, Brunereau L, Laberge S, et al: Prospective follow-up of 33 asymptomatic patients with familial cerebral cavernous malformations. Neurology 57:1825–1828, 2001. 220. Labauge P, Denier C, Bergametti F, et al: Genetics of cavernous angiomas. Lancet Neurol 6:237–244, 2007. 221. Larson TC, III, Kelly WM, Ehman RL, et al: Spatial misregistration of vascular flow during MR imaging of the CNS: Cause and clinical significance. AJNR Am J Neuroradiol 11:1041–1048, 1990. 222. Lasjaunias P, Burrows P, Planet C: Developmental venous anomalies (DVA): The so-called venous angioma. Neurosurg Rev 9:233–244, 1986. 223. Lasjaunias P, Chue M, TerBrugge K: Neurological manifestations of intracranial dural arteriovenous malformations. J Neurosurg 64:724– 730, 1986. 224. Latchaw RE, Truwit CL, Heros RC: Venous angioma, cavernous angioma, and hemorrhage. AJNR Am J Neuroradiol 15:1255–1257, 1994. 225. Lavoie P, Gariepy JL, Milot G, et al: Residual flow after cerebral aneurysm coil occlusion: Diagnostic accuracy of MR angiography. Stroke 43:740–746, 2012. 226. Le TT, Fischbein NJ, Andre JB, et al: Identification of venous signal on arterial spin labeling improves diagnosis of dural arteriovenous fistulas and small arteriovenous malformations. AJNR Am J Neuroradiol 33:61–68, 2012. 227. Lee R, Becher M, Benson M, et al: Brain capillary telangiectasia: MR imaging appearance and clinicohistopathologic findings. Radiology 205:797–805, 1997. 228. Lee M, Kim MS: Image findings in brain developmental venous anomalies. J Cerebrovasc Endovasc Neurosurg 14:37–43, 2012. 229. Lee C, Pennington M, Kenney C, III: MR evaluation of developmental venous anomalies: Medullary venous anatomy of venous angiomas. AJNR Am J Neuroradiol 17:61–70, 1996. 230. Lehnhardt FG, von Smekal U, Ruckriem B, et al: Value of gradient echo magnetic resonance imaging in the diagnosis of familial cerebral cavernous malformation. Arch Neurol 62:653–658, 2005. 231. Levent A, Yuce I, Eren S, et al: Contrast-enhanced and time-of-flight MR angiographic assessment of endovascular coiled intracranial aneurysms at 1.5T. Interv Neuroradiol 20:686–692, 2014. 232. Levrier O, Gailloud PH, Souei M, et al: Normal galenic drainage of the deep cerebral venous system in two cases of vein of Galen aneurismal malformation. Childs Nerv Syst 20:91–97, 2004. 233. Li Q, Lv F, Yao G, et al: 64-section multidetector CT angiography for evaluation of intracranial aneurysms: Comparison with 3D rotational angiography. Acta Radiol 55:840–846, 2014. 234. Liang L, Korogi Y, Sugahara T, et al: Detection of intracranial hemorrhage with susceptibility weighted MR sequences. AJNR Am J Neuroradiol 20:1527–1534, 1999. 235. Lim HK, Choi CG, Kim SM, et al: Detection of residual brain arteriovenous malformations after radiosurgery: Diagnostic accuracy of contrast-enhanced four-dimensional MR angiography at 3.0 T. Br J Radiol 85:1064–1069, 2012. 236. Lim LK, Dowling RJ, Yan B, et al: Can CT angiography rule out aneurysmal subarachnoid haemorrhage in CT scan-negative

http://radiologyebook.com

420

237. 238. 239.

240.

241.

242.

243.

244.

245.

246.

247.

248.

249.

250.

251.

252.

253.

254.

255.

256.

257.

PART II  CT and MR Imaging of the Whole Body subarachnoid haemorrhage patients? J Clin Neurosci 21:191–193, 2014. Lindquist C, Guo W-Y, Karlsson B, et al: Radiosurgery for venous angiomas. J Neurosurg 78:531–536, 1993. Linsky ME, Sekhar LN: Cavernous sinus hemangioma: A series, a review, and an hypothesis. Neurosurgery 3:101–108, 1992. Liu XJ, Wang S, Zhao YL, et al: Risk of cerebral arteriovenous malformation rupture during pregnancy and puerperium. Neurology 82:1798–1803, 2014. Liquori CL, Berg MJ, Siegel AM, et al: Mutations in a gene encoding a novel protein containing a phosphotyrosine-binding domain cause type 2 cerebral cavernous malformations. Am J Hum Genet 73:1459–1464, 2003. Lunsford L, Flickinger J, Coffey R: Stereotactic gamma knife radiosurgery: Initial North American experience in 207 patients. Arch Neurol 7:169–175, 1990. Lunsford L, Kondziolka D, Bissonette D, et al: Stereotactic radiosurgery of brain vascular malformations. Neurosurg Clin North Am 3:79–98, 1992. Luo S, Zhang LJ, Meinel FG, et al: Low tube voltage and low contrast material volume cerebral CT angiography. Eur Radiol 24:1677–1685, 2014. Lylyk P, March AD, Kohan GA, et al: Alternative therapeutic approaches in intravascular embolization of vein of Galen vascular malformations. In Vinvela F, Halback VV, Dion JE, editors: Endovascular therapy of the central nervous system, New York, 1992, Raven Press. MacKinnon AD, Clifton AG, Rich PM: Acute subarachnoid haemorrhage: Is a negative CT angiogram enough? Clin Radiol 68:232–238, 2013. Maher CO, Piepgras DG, Brown RD, Jr, et al: Cerebrovascular manifestations in 321 cases of hereditary hemorrhagic telangiectasia. Stroke 32:877–882, 2001. Majoie CB, Sprengers ME, van Rooij WJ, et al: MR angiography at 3T versus digital subtraction angiography in the follow-up of intracranial aneurysms treated with detachable coils. AJNR Am J Neuroradiol 26:1349–1356, 2005. Maldjian J, Howard R, Alsop DA, et al: Functional MRI in AVMs prior to surgical or endovascular therapy (abstract). In Book of Abstracts, Annual Meeting of the Radiological Society of North America, 1994. Malek AM, Halbach VV, Dowd CF, et al: Diagnosis and treatment of dural arteriovenous fistulas. Neuroimaging Clin North Am 8:445–468, 1998. Mallouhi A, Felber S, Chemelli A, et al: Detection and characterization of intracranial aneurysms with MR angiography: Comparison of volume-rendering and maximum-intensity-projection algorithms. AJR Am J Roentgenol 180:55–64, 2003. Marks MP, DeLapaz RL, Fabrikant JI, et al: Intracranial vascular malformations: Imaging of charged-particle radiosurgery. Part I: Results of therapy. Radiology 168:447–455, 1988. Marks MP, DeLapaz RL, Fabrikant JI, et al: Intracranial vascular malformations: Imaging of charged-particle radiosurgery. Part II: Complications. Radiology 168:457–462, 1988. Marks MP, Lane B, Steinberg GK, et al: Hemorrhage in intracerebral arteriovenous malformations: Angiographic determinants. Radiology 176:807–813, 1990. Marks MP, Lane B, Steinberg GK, et al: Intracranial aneurysms in cerebral arteriovenous malformations: Evaluation and endovascular treatment. Radiology 183:355–360, 1992. Marks MP, Pelc MJ, Ross MR, et al: Determination of cerebral blood flow with a phase-contrast cine MR imaging technique: Evaluation   of normal subjects and patients with arteriovenous malformation. Radiology 182:467–476, 1992. Marro B, Galanaud D, Valery CA, et al: Intracranial aneurysm: Inner view and neck identification with CT angiography virtual endoscopy.   J Comput Assist Tomogr 21:587–589, 1997. Martin N, King W, Wilson C, et al: Management of dural arteriovenous malformations of the anterior cranial fossa. J Neurosurg 72:692–697, 1990.

258. Matsubara S, Manzia J, ter Brugge K, et al: Angiographic and clinical characteristics of patients with cerebral arteriovenous malformations associated with hereditary hemorrhagic telangiectasia. AJNR Am J Neuroradiol 21:1016–1020, 2000. 259. McCormick PW, McCormick J, Zimmerman R, et al: The pathophysiology of acute subarachnoid hemorrhage. BNI Q 7:18–26, 1991. 260. McCormick P, Michelson W: Management of cavernous and venous malformations. In Barrow D, editor: Intracranial vascular malformations, Park Ridge, IL, 1990, AANS, pp 197–217. 261. Mehan WA, Jr, Romero JM, Hirsch JA, et al: Unruptured intracranial aneurysms conservatively followed with serial CT angiography: Could morphology and growth predict rupture? J Neurointerv Surg 6:761–766, 2014. 262. Mendalow A, Erfurth A, Grossary K, et al: Do cerebral arteriovenous malformations increase in size? J Neurol Neurosurg Psychiatry 50:980–987, 1987. 263. Meng G, Bai C, Yu T, et al: The association between cerebral developmental venous anomaly and concomitant cavernous malformation: An observational study using magnetic resonance imaging. BMC Neurol 14:50, 2014. 264. Menke J, Larsen J, Kallenberg K: Diagnosing cerebral aneurysms by computed tomographic angiography: Meta-analysis. Ann Neurol 69:646–654, 2011. 265. Menke J, Schramm P, Sohns JM, et al: Diagnosing flow residuals in coiled cerebral aneurysms by MR angiography: Meta-analysis. J Neurol 261:655–662, 2014. 266. Metens T, Rio F, Baleriaux D, et al: Intracranial aneurysms: Detection with gadolinium-enhanced dynamic three-dimensional MR angiography-initial results. Radiology 216:39–46, 2000. 267. Meyer FB, Huston J, III, Reiderer SS: Pulsatile increases in aneurysm size determined by cine phase-contrast MR angiography. Neurosurgery 78:879–883, 1993. 268. Mikami T, Hirano T, Sugino T, et al: Presurgical planning for arteriovenous malformations using multidetector row CT. Neurosurg Rev 35:393–399, 2012. 269. Mine B, Pezzulio M, Roque G, et al: Detection and characterization of unruptured intracranial aneurysms: Comparison of 3T MRA and DSA. J Neuroradiol 14:250–258, 2014. 270. Mir DI, Gupta A, Dunning A, et al: CT perfusion for detection of delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis. AJNR Am J Neuroradiol 35:866– 871, 2014. 271. Miyachi S, Negoro M, Harda T, et al: Contribution of meningeal arteries to cerebral arteriovenous malformations. Neuroradiology 35:205–209, 1993. 272. Miyasaka K, Wolpert SM, Prager RJ: The association of cerebral aneurysms, infundibula and intracranial arteriovenous malformations. Stroke 13:196–203, 1982. 273. Mizutani T, Aruga T, Kirino T, et al: Recurrent subarachnoid hemorrhage from untreated ruptured vertebrobasilar dissecting aneurysms. J Neurosurg 36:905–913, 1995. 274. Mizutani T, Tanaka H, Aruza T: Multiple arteriovenous   malformations located in the cerebellum, posterior fossa, spinal   cord, dura, and scalp with associated port wine stain and   supratentorial venous anomaly. Neurosurgery 31:137–141, 1992. 275. Monaco RG, Alvarez H, Goulao A, et al: Posterior fossa arteriovenous malformations: Angioarchitecture in relation to their hemorrhagic episodes. Neuroradiology 31:471–475, 1990. 276. Morales-Valero SF, Bortolotti C, Sturiale CL, et al: Are parenchymal AVMs congenital lesions? Neurosurg Focus 37:E2, 2014. 277. Nael K, Villablanca JP, Saleh R, et al: Contrast-enhanced MR angiography at 3T in the evaluation of intracranial aneurysms: A comparison with time-of-flight MR angiography. AJNR Am J Neuroradiol 27:2118–2121, 2006. 278. Nakagawa T, Hashi K: The incidence and treatment of asymptomatic, unruptured cerebral aneurysms. J Neurosurg 80:217–223, 1994.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations 279. Nakiri GS, Santos AC, Abud TG, et al: A comparison between magnetic resonance angiography at 3 teslas (time-of-flight and contrastenhanced) and flat-panel digital subtraction angiography in the assessment of embolized brain aneurysms. Clinics 66:641–648, 2011. 280. Nakstad P, Nornes H, Hauge HN: Traumatic aneurysms of the pericallosal arteries. Neuroradiology 28:335–338, 1986. 281. Nanney AD, 3rd, El Tecle NE, El Ahmadieh TY, et al: Intracranial aneurysms in previously irradiated fields: Literature review and case report. World Neurosurg 81:511–519, 2014. 282. Napel S, Marks MP, Rubin GP, et al: CT angiography with spiral CT and maximum intensity projection. Radiology 185:607–610, 1992. 283. Nehls DG, Flom RA, Carter LP, et al: Multiple intracranial aneurysms: Determining the site of rupture. J Neurosurg 63:342–348, 1985. 284. Newell DW, LeRoux PD, Dacey RG, Jr, et al: CT infusion scanning   for the detection of cerebral aneurysms. J Neurosurg 71:175–179, 1989. 285. Ni W, Tian Y, Jiang H, et al: Preliminary experience of 256-row multidetector computed tomographic angiography for detecting cerebral aneurysms. J Comput Assist Tomogr 37:233–241, 2013. 286. Niemczyk M, Gradzik M, Niemczyk S, et al: Intracranial aneurysms in autosomal dominant polycystic kidney disease. AJNR Am J Neuroradiol 34:1556–1559, 2013. 287. Nikoubashman O, Wiesmann M, Tournier-Lasserve E, et al: Natural history of cerebral dot-like cavernomas. Clin Radiol 68:e453–e459, 2013. 288. Nishijima M, Takaku A, Endo S, et al: Etiological evaluation of dural arteriovenous malformations of the lateral and sigmoid sinuses based on histopathological examinations. J Neurosurg 76:600–606, 1992. 289. Noguchi K, Kubo M, Kuwayama N, et al: Intracranial dural arteriovenous fistulas with retrograde cortical venous drainage: Assessment with cerebral blood volume by dynamic susceptibility contrast magnetic resonance imaging. AJNR Am J Neuroradiol 27:1252–1256, 2006. 290. Noguchi K, Melhem ER, Kanazawa T, et al: Intracranial dural arteriovenous fistulas: Evaluation with combined 3D time-of-flight MR angiography and MR digital subtraction angiography. AJR Am J Roentgenol 182:183–190, 2004. 291. Nome T, Bakke SJ, Nakstad PH: MR angiography in the follow-up of coiled cerebral aneurysms after treatment with Guglielmi detachable coils. Acta Radiol 43:10–14, 2002. 292. Nov AA, Cromwell LD: Traumatic pericallosal artery aneurysm. J Neuroradiol 11:3–8, 1984. 293. Nucifora PG, Dillon WP: MR diagnosis of myxomatous aneurysms: Report of two cases. AJNR Am J Neuroradiol 22:1349–1352, 2001. 294. Obuchowski NA, Modic MT, Magdinec M: Current implications for the efficacy of noninvasive screening for occult intracranial aneurysms in patients with a family history of aneurysms. J Neurosurg 83:42–49, 1995. 295. Ogawa T, Inugami A, Shimosegawa E, et al: Subarachnoid hemorrhage: Evaluation with MR imaging. Radiology 186:345–351, 1993. 296. Ogawa S, Lee TM, Kay AR, et al: Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A 87:9868–9872, 1990. 297. Ogawa T, Okudera T, Noguchi K, et al: Cerebral aneurysms: Evaluation with three-dimensional CT angiography. AJNR Am J Neuroradiol 18:790–792, 1997. 298. Ogilvy CS, Stieg PE, Awad I, et al: Recommendations for the management of intracranial arteriovenous malformations. Stroke 32:1458–1471, 2001. 299. Ohman J, Heiskanen O: Timing of operation for ruptured supratentorial aneurysms: A prospective randomized study. J Neurosurg 70:55–60, 1989. 300. Okada T, Miki Y, Kikuta K, et al: Diffusion tensor fiber tractography for arteriovenous malformations: Quantitative analyses to evaluate the corticospinal tract and optic radiation. AJNR Am J Neuroradiol 28:1107–1113, 2007. 301. Okahara M, Kiyosue H, Hori Y, et al: Three-dimensional time-of-flight MR angiography for evaluation of intracranial aneurysms after

302. 303. 304. 305. 306.

307. 308. 309.

310.

311.

312.

313.

314. 315.

316.

317.

318.

319.

320.

321.

322.

323.

324.

421

endosaccular packing with Guglielmi detachable coils: comparison   with 3D digital subtraction angiography. Eur Radiol 14:1162–1168, 2004. Okazaki H: Fundamentals of neuropathology, New York, 1983, Igaku-Shoin. Okazaki H: Cerebrovascular disease. In Fundamentals of neuropathology, ed 2, New York, 1989, Igaku-Shoin, pp 27–94. Okazaki H: Malformative vascular lesions. In Fundamentals of neuropathology, ed 2, New York, 1989, Igaku-Schoin, pp 70–74. Okazaki H, Scheithauer B: Atlas of neuropathology, New York, 1988, Gower Medical. Ondra S, Troupp H, George E, et al: The natural history of symptomatic arteriovenous malformations of the brain. J Neurosurg 73:387–391, 1990. Osborn A: Intracranial aneurysms. In Diagnostic neuroradiology, St. Louis, 1994, Mosby-Year Book. Osborn A: Intracranial vascular malformations. In Diagnostic neuroradiology, St. Louis, 1994, Mosby-Year Book. Osterton B, Solymosi L: Magnetic resonance angiography of cerebral developmental anomalies: Its role in differential diagnosis. Neuroradiology 35:97–104, 1993. Oyesiku N, Colohan ART, Barow DL, et al: Cocaine-induced aneurysmal rupture: An emergent negative factor in the natural history of intracranial aneurysms? Neurosurgery 32:518–526, 1993. Papke K, Kuhl C, Fruth M, et al: Intracranial aneurysms: Role of multidetector CT angiography in diagnosis and endovascular therapy planning. Radiology 244:532–540, 2007. Park S, Lee DH, Ryu CW, et al: Incidental saccular aneurysms on head MR angiography: 5 years’ experience at a single large-volume center. J Stroke 16:189–194, 2014. Parkarinen S: Incidence, aetiology, and prognosis of primary subarachnoid hemorrhage. A study based on 589 cases diagnosed in a defined urban population during a defined period. Acta Neurol Scand 43:1–57, 1967. Parker DL, Yuan C, Blatter DD: MR angiography by multiple thin slab 3D acquisition. Magn Reson Med 17:434–451, 1991. Perata HJ, Tomsick TA, Tew JM, Jr: Feeding artery pedicle aneurysms: Association with parenchymal hemorrhage and arteriovenous malformation in the brain. J Neurosurg 80:631–634, 1994. Perret G, Nishioka H: Arteriovenous malformations: An analysis of 545 cases for craniocerebral arteriovenous malformations and fistulae reported to the cooperative study. J Neurosurg 25:467–490, 1966. Phillips LH, II, Whisnant JP, O’Fallon WM, et al: The unchanging pattern of subarachnoid hemorrhage in a community. Neurology 30:1034–1040, 1980. Piepgras DG, Sundt TM, Jr, Ragoowansi AT, et al: Seizure outcome in patients with surgically treated cerebral arteriovenous malformations. J Neurosurg 78:5–11, 1993. Pierot L, Chiras J, Meder J-F, et al: Dural arteriovenous fistulas of the posterior fossa draining into subarachnoid veins. AJNR Am J Neuroradiol 13:315–323, 1992. Pierot L, Delcourt C, Bouquigny F, et al: Follow-up of intracranial aneurysms selectively treated with coils: Prospective evaluation of contrast-enhanced MR angiography. AJNR Am J Neuroradiol 27:744– 749, 2006. Pierot L, Portefaix C, Boulin A, et al: Follow-up of coiled intracranial aneurysms: Comparison of 3D time-of-flight and contrast-enhanced magnetic resonance angiography at 3T in a large, prospective series. Eur Radiol 22:2255–2263, 2012. Pierot L, Portefaix C, Gauvrit JY, et al: Follow-up of coiled intracranial aneurysms: Comparison of 3D time-of-flight MR angiography at 3T and 1.5T in a large prospective series. AJNR Am J Neuroradiol 33:2162–2166, 2012. Pierot L, Portefaix C, Rodriguez-Regent C, et al: Role of MRA in the detection of intracranial aneurysm in the acute phase of subarachnoid hemorrhage. J Neuroradiol 40:204–210, 2013. Pouratian N, Bookheimer SY, Rex DE, et al: Utility of preoperative functional magnetic resonance imaging for identifying language

http://radiologyebook.com

422

325.

326.

327.

328. 329.

330.

331.

332.

333. 334.

335.

336.

337.

338.

339.

340.

341. 342. 343.

344.

345.

346.

PART II  CT and MR Imaging of the Whole Body cortices in patients with vascular malformations. J Neurosurg 97:21–32, 2002. Pradilla G, Wicks RT, Hadelsberg U, et al: Accuracy of computed tomography angiography in the diagnosis of intracranial aneurysms. World Neurosurg 80:845–852, 2013. Raaymakers TW, Buys PC, Verbeeten B, Jr, et al: MR angiography as a screening tool for intracranial aneurysms: Feasibility, test characteristics and interobserver agreement. AJR Am J Roentgenol 173:1469–1475, 1999. Ramgren B, Siemund R, Nilsson OG, et al: CT angiography in non-traumatic subarachnoid hemorrhage: The importance of arterial attenuation for the detection of intracranial aneurysms. Acta Radiol 2014. Raps EC, Rogers JD, Galeta SL, et al: The clinical spectrum of unruptured intracranial aneurysms. Arch Neurol 50:265–268, 1993. Reck SD, Zacks DN, Eibschitz-Tsimhoni M: Retinal and intracranial arteriovenous malformations: Wyburn-Mason syndrome. J Neuroophthalmol 25:205–208, 2005. Reinacher P, Reinges MH, Simon VA, et al: Dynamic 3-D contrastenhanced angiography of cerebral tumours and vascular malformations. Eur Radiol 17(Suppl 6):F52–F62, 2007. Report of World Federation of Neurological Surgeons Committee on a Universal Subarachnoid Hemorrhage Grading Scale. J Neurosurg 68:985–986, 1988. Reynolds MR, Willie JT, Zipfel GJ, et al: Sexual intercourse and cerebral aneurysmal rupture: Potential mechanisms and precipitants. J Neurosurg 114:969–977, 2011. Rhoton A: Anatomy of saccular aneurysms. Surg Neurol 14:59–66, 1980. Riel-Romero RM, Mattingly M: Developmental venous anomaly in association with neuromigrational anomalies. Pediatr Neurol 32:53–55, 2005. Rigamonti D, Hadley MN, Drayer BP, et al: Cerebral cavernous malformations: Incidence and familial occurrence. N Engl J Med 319:343–347, 1988. Rigamonti D, Johnson PC, Spetzler RF, et al: Cavernous malformations and capillary telangiectasia: A spectrum within a single pathological entity. Neurosurgery 28:60–64, 1991. Rigamonti D, Spetzler D: The association of venous and cavernous malformations: Report of form cases and discussions of the pathophysiological, diagnostic and therapeutic implications. Acta Neurochir (Wien) 92:100–105, 1988. Rinkel GJE, Djibuti M, Algra A, et al: Prevalence and risk of rupture   of intracranial aneurysms. A systematic review. Stroke 29:251–256, 1998. Rinkel GJ, Wijdicks EF, Hasan D, et al: Outcome in patients with subarachnoid haemorrhage and negative angiography according to pattern of haemorrhage on computed tomography. Lancet 338(8773):964–968, 1991. Rinkel GJE, Wijdicks EFM, Vermeulen M, et al: Nonaneurysmal perimesencephalic subarachnoid hemorrhage: CT and MR patterns that differ from aneurysmal rupture. AJNR Am J Neuroradiol 12:829–834, 1991. Robinson J, Awad I, Little J: Natural history of the cavernous angioma. J Neurosurg 75:709–714, 1991. Rolen PB, Sze G: Small, patent aneurysms: Atypical appearances at 1.5-T MR imaging. Radiology 208:129–136, 1998. Rosenorn J, Eskesen V, Schmidt K: Unruptured intracranial aneurysms: An assessment of the annual risk of rupture based on epidemiological and clinical data. Br J Neurosurg 2:369–378, 1988. Ross JS, Masaryk TJ, Modic MT, et al: Intracranial aneurysms: Evaluation by MR angiography. AJNR Am J Neuroradiol 11:449–456, 1990. Ruigrok YM, Rinkel GJ, Algra A, et al: Characteristics of intracranial aneurysms in patients with familial subarachnoid hemorrhage. Neurology 62:891–894, 2004. Russell DS, Rubinstein LJ: Pathology of tumors of the nervous system, ed 5, Baltimore, 1989, Williams & Wilkins.

347. Sadasivan C, Fiorella DJ, Woo HH, et al: Physical factors effecting cerebral aneurysm pathophysiology. Ann Biomed Eng 41:1347–1365, 2013. 348. Sadato N, Numaguchi T, Rigamonti D, et al: Bleeding patterns in ruptured posterior fossa aneurysms: A CT study. J Comp Assist Tomogr 15:612–617, 1991. 349. Sadler LR, Tarr RW, Jungreis CA, et al: Sinus pericranii: CT and MR findings. J Comp Assist Tomogr 14:124–127, 1990. 350. Sahlein DH, Mora P, Becske T, et al: Features predictive of brain arteriovenous malformation hemorrhage: Extrapolation to a physiologic model. Stroke 45:1964–1970, 2014. 351. Sahs AL, Perret GE, Locksley HB, et al: Intracranial aneurysms and subarachnoid hemorrhage: A cooperative study, Philadelphia, 1969, JB Lippincott. 352. Sailer AM, Wagemans BA, Nelemans PJ, et al: Diagnosing intracranial aneurysms with MR angiography: Systematic review and meta-analysis. Stroke 45:119–126, 2014. 353. Saito Y, Kobayashi N: Cerebral venous angiomas: Clinical evaluation and possible etiology. Radiology 139:87–94, 1981. 354. Sakaki T, Tominaga M, Miyamoto K, et al: Clinical studies of de novo aneurysms. Neurosurgery 32:512–517, 1993. 355. Sakamoto S, Kiura Y, Shibukawa M, et al: Subtracted 3D CT angiography for evaluation of internal carotid artery aneurysms: Comparison with conventional digital subtraction angiography. AJNR Am J Neuroradiol 27:1332–1337, 2006. 356. Salcman M, Scholtz H, Numaguchi Y: Multiple intracerebral arteriovenous malformations: Report of three cases and review of the literature. Surg Neurol 38:121–128, 1992. 357. Sanders WP, Sorek PA, Mehta BA: Fenestration of intracranial arteries with special attention to associated aneurysms and other anomalies. AJNR Am J Neuroradiol 14:675–680, 1993. 358. Sanelli PC, Anumula N, Johnson CE, et al: Evaluating CT perfusion using outcome measures of delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage. AJNR Am J Neuroradiol 34:292–298, 2013. 359. Sanelli PC, Ugorec I, Johnson CE, et al: Using quantitative CT perfusion for evaluation of delayed cerebral ischemia following aneurysmal subarachnoid hemorrhage. AJNR Am J Neuroradiol 32:2047–2053, 2011. 360. Sato M, Nakano M, Sasanuma J, et al: Preoperative cerebral aneurysm assessment by three-dimensional magnetic resonance angiography: Feasibility of surgery without conventional catheter angiography. Neurosurgery 56:903–912, 2005. 361. Saveland H, Hillman J, Brandt L, et al: Overall outcome in aneurysmal subarachnoid hemorrhage. J Neurosurg 76:729–734, 1992. 362. Savoiardo M, Strada L, Passerini A: Intracranial cavernous hemangioma: Neuroradiologic review of 36 operated cases. AJNR Am J Neuroradiol 4:945–950, 1983. 363. Schievink WI, Limburg M, Dreissen J, Jr, et al: Screening for unruptured familial aneurysms: Subarachnoid hemorrhage 2 years after angiography negative for aneurysms. Neurosurg 29:434–438, 1991. 364. Schievink WI, Mokri B, Michels VV, et al: Familial association of intracranial aneurysms and cervical artery dissections. Stroke 22:1426– 1430, 1991. 365. Schievink WI, Mokri B, Piepgras DG: Angiographic frequency of saccular intracranial aneurysms in patients with spontaneous cervical arterial dissection. J Neurosurg 76:62–66, 1992. 366. Schievink WI, Piepgras DG, Wirth FP: Rupture of previously documented small asymptomatic saccular intracranial aneurysms. J Neurosurg 76:1019–1024, 1992. 367. Schievink WI, Wijdicks EF: Origin of pretruncal nonaneurysmal subarachnoid hemorrhage: Ruptured vein, perforating artery, or intramural hematoma? Mayo Clin Proc 75:1169–1173, 2000. 368. Schmid UD, Steiger HJ, Huber P: Accuracy of high resolution computed tomography in direct diagnosis of cerebral aneurysms. Neuroradiology 29:152–159, 1987. 369. Schuierer G, Huk WJ, Laub G: Magnetic resonance angiography of intracranial aneurysms. Neuroradiology 35:50–54, 1992.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations 370. Scotti G, Ethier R, Melancon D, et al: Computed tomography in the evaluation of intracranial aneurysms and subarachnoid hemorrhage. Radiology 123:85–90, 1977. 371. Serafin Z, Strzesniewski P, Lasek W, et al: Comparison of remnant size in embolized intracranial aneurysms measured at follow-up with DSA and MRA. Neuroradiology 54:1381–1388, 2012. 372. Serafin Z, Strzesniewski P, Lasek W, et al: Follow-up after embolization of ruptured intracranial aneurysms: A prospective comparison of two-dimensional digital subtraction angiography, three-dimensional digital subtraction angiography, and time-of-flight magnetic resonance angiography. Neuroradiology 54:1253–1260, 2012. 373. Shankar JJ, Tan IY, Krings T, et al: CT angiography for evaluation of cerebral vasospasm following acute subarachnoid haemorrhage. Neuroradiology 54:197–203, 2012. 374. Sharma A, Tyagi G, Sahai A, et al: Traumatic aneurysm of superficial temporal artery: CT demonstration. Neuroradiology 33:510–512, 1991. 375. Sharma A, Zipfel GJ, Hildebolt C, et al: Hemodynamic effects of developmental venous anomalies with and without cavernous malformations. AJNR Am J Neuroradiol 34:1746–1751, 2013. 376. Shellock FG, Detrick MS, Brant-Zawadski MN: MR compatibility of Guglielmi detachable coils. Radiology 203:568–570, 1997. 377. Shellock FG, Gounis M, Wakhloo A: Detachable coil for cerebral aneurysms: In vitro evaluation of magnetic field interactions, heating, and artifacts at 3 T. AJNR Am J Neuroradiol 26:363–366, 2005. 378. Shellock FG, Kanal E: Yasargil aneurysm clips: Evaluation of interactions with a 1.5-T MR system. Radiology 207:587–591, 1998. 379. Shinohara Y, Sakamoto M, Iwata N, et al: Usefulness of monochromatic imaging with metal artifact reduction software for computed tomography angiography after intracranial aneurysm coil embolization. Acta Radiol 55:1015–1023, 2014. 380. Sigal R, Krief O, Houtteville JP, et al: Occult cerebrovascular malformations: Follow-up with MR imaging. Radiology 176:815–819, 1990. 381. Signorelli F, Della Pepa GM, Sabatino G, et al: Diagnosis and management of dural arteriovenous fistulas: A 10 years single-center experience. Clin Neurol Neurosurg 128:123–129, 2015. 382. Simard JM, Garcia-Bengochea F, Ballinger WEJ, et al: Cavernous angioma: A review of 126 collected and 12 new clinical cases. Neurosurgery 18:162–172, 1986. 383. Smith HJ, Strother CM, Kikuchi Y, et al: MR imaging in the management of supratentorial intracranial AVMs. AJNR Am J Neuroradiol 9:225–235, 1988. 384. Sohn CH, Kim SP, Kim IM, et al: Characteristic MR imaging findings of cavernous hemangiomas in the cavernous sinus. AJNR Am J Neuroradiol 24:1148–1151, 2003. 385. Soize S, Bouquigny F, Kadziolda K, et al: Value of 4D MR angiography at 3T compared with DSA for the follow-up of treated brain arteriovenous malformation. AJNR Am J Neuroradiol 35:1903–1909, 2014. 386. Sorek PA, Silbergleit R: Multiple asymptomatic cervical cephalic aneurysms. AJNR Am J Neuroradiol 14:31–33, 1993. 387. Spetzler RF, Martin NA: A proposed scheme for grading intracranial arteriovenous malformations. J Neurosurg 65:476–483, 1986. 388. Spetzler RF, Ponce FA: A 3-tier classification of cerebral arteriovenous malformations. Clinical article. J Neurosurg 114:842–849, 2011. 389. Spittau B, Millan DS, El-Sherifi S, et al: Dural arteriovenous fistulas of the hypoglossal canal: systematic review on imaging anatomy, clinical findings, and endovascular management. J Neurosurg 21:2–21, 2014. 390. Stancanello J, Cavedon C, Francescon P, et al: Bold fMRI integration into radiosurgery treatment planning of cerebral vascular malformations. Med Phys 34:1176–1184, 2007. 391. Stapf C, Mast H, Sciacca RR, et al: Predictors of hemorrhage in patients with untreated brain arteriovenous malformation. Neurology 66:1350– 1355, 2006. 392. Stehbens WE: Etiology of intracranial berry aneurysms. J Neurosurg 70:823–831, 1989.

423

393. Steiner L: Radiosurgery in cerebral arteriovenous malformations. In Flamm E, editor: Cerebrovascular surgery, New York, 1986, Springer-Verlag. 394. Steiner L, Lindquist C, Adler JR, et al: Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 77:1–8, 1992. 395. Stone JL, Crowell RM, Gandhi YN, et al: Multiple intracranial aneurysms: Magnetic resonance imaging for determination of the site of rupture. Neurosurgery 23:97–100, 1988. 396. Strother CM, Graves VB, Rappe A: Aneurysm hemodynamics: An experimental study. AJNR Am J Neuroradiol 13:1089–1095, 1992. 397. Suarez JI, Tarr RW, Selman WR: Aneurysmal subarachnoid hemorrhage. N Engl J Med 354:387–396, 2006. 398. Suzuki J, Kodama N: Moyamoya disease: A review. Stroke 14:104–109, 1983. 399. Suzuki S, Kayama T, Sakurai Y, et al: Subarachnoid hemorrhage of unknown origin. Neurosurg 21:310–313, 1987. 400. Suzuki M, Matsui O, Kobayashi K, et al: Contrast-enhanced MRA for investigation of cerebral arteriovenous malformations. Neuroradiology 45:231–235, 2003. 401. Suzuki IM, Matsui, Ueda F, et al: Contrast-enhanced MR angiography (enhanced 3-D fast gradient echo) for diagnosis of cerebral aneurysms. Neuroradiology 44:17–20, 2002. 402. Takahashi S, Tomura N, Watarai J, et al: Dural arteriovenous fistula of the cavernous sinus with venous congestion of the brain stem: Report of two cases. AJNR Am J Neuroradiol 20:886–888, 1999. 403. Takamiya Y, Takayama H, Kobayashi K, et al: Familial occurrence of multiple vascular malformations of the brain. Neurol Med Chir (Tokyo) 24:271–277, 1984. 404. Takhtani D, Dundamadappa S, Puri AS, et al: Flow artifact in the anterior communicating artery resembling aneurysm on the time of flight MR angiogram. Acta Radiol 55:1253–1257, 2014. 405. Tamatani S, Sasaki O, Takeuchi S, et al: Detection of delayed cerebral vasospasm, after rupture of intracranial aneurysm, by magnetic resonance angiography. Neurosurgery 40:748–753, 1997. 406. Taniguchi RM, Goree JA, Odom GL: Spontaneous carotid cavernous shunts presenting diagnostic problems. J Neurosurg 35:384–391, 1971. 407. Tee JW, Dally M, Madan A, et al: Surgical treatment of poorly visualized and complex cerebrovascular lesions using pre-operative angiographic data as angiographic DynaCT datasets for frameless stereotactic navigation. Acta Neurochir (Wien) 154:1159–1167, 2012. 408. Teng MM, Nasir Qadri SM, et al: MR imaging of giant intracranial aneurysm. J Clin Neurosci 10:460–464, 2003. 409. Thomas LE, Czuczman AD, Boulanger AB, et al: Low risk for subsequent subarachnoid hemorrhage for emergency department patients with headache, bloody cerebrospinal fluid, and negative findings on cerebrovascular imaging. J Neurosurg 121:24–31, 2014. 410. Tomandl BF, Kostner NC, Schempershofe M, et al: CT angiography of intracranial aneurysms: A focus on postprocessing. Radiographics 24:637–655, 2004. 411. Toro VE, Fravel JF, Weidman TA: Posttraumatic pseudoaneurysm of the posterior meningeal artery associated with intraventricular hemorrhage. AJNR Am J Neuroradiol 14:264–266, 1993. 412. Truwit CL: Venous angioma of the brain: History, significance and imaging findings. AJR Am J Roentgenol 159:1299–1307, 1992. 413. Tsuchiya K, Katase S, Yoshino A, et al: Preliminary evaluation of volume-rendered three-dimensional display of time-of-flight MR angiography in the diagnosis of intracranial aneurysms. Neuroradiology 43:633–636, 2001. 414. Tsuruda J, Saloner D, Norman D: Artifacts associated with MR neuroangiography. AJNR Am J Neuroradiol 13:1411–1422, 1992. 415. Tsutsumi M, Kazekawa K, Tanaka A, et al: Traumatic middle   meningeal artery pseudoaneurysm and subsequent fistula   formation with the cavernous sinus: Case report. Surg Neurol 58:325–328, 2002. 416. Tu J, Stoodley MA, Morgan MK, et al: Ultrastructural characteristics of hemorrhagic, nonhemorrhagic, and recurrent cavernous malformations. Neurosurgery 103:903–909, 2005.

http://radiologyebook.com

424

PART II  CT and MR Imaging of the Whole Body

417. UCAS Japan Investigators: The natural course of unruptured cerebral aneurysms in a Japanese cohort. N Engl J Med 366:2474–2482, 2012. 418. Uchino A, Imador H, Ohno M: Magnetic resonance imaging of intracranial venous angioma. Clin Imaging 14:309–314, 1990. 419. Uchino A, Kato A, Kuroda S, et al: Pontine venous congestion caused by dural carotid-cavernous fistula: Report of two cases. Eur Radiol 7:405–408, 1997. 420. Unlu E, Cakir B, Gocer B, et al: The role of contrast-enhanced MR angiography in the assessment of recently ruptured intracranial aneurysms: A comparative study. Neuroradiology 47:780–791, 2005. 421. Unruptured intracranial aneurysms—Risk of rupture and risks of surgical intervention. N Engl J Med 339:1725–1733, 1998. 422. Vajda J: Multiple intracranial aneurysms: A high-risk condition. Acta Neurochir (Wien) 118:59–75, 1992. 423. Valavanis A, Wellauer J, Yasargil MG: The radiological diagnosis of cerebral venous angioma: Cerebral angiography and computed tomography. Neuroradiology 24:193–199, 1983. 424. van Beijnum J, vander Worp HB, Algra A, et al: Prevalence of brain arteriovenous malformations in first-degree relatives of patients with a brain arteriovenous malformation. Stroke 45:3231–3235, 2014. 425. van der Schaaf IC, Velthuis BK, Wermer MJ, et al: New detected aneurysms on follow-up screening in patients with previously clipped intracranial aneurysms: Comparison with DSA or CTA at the time of SAH. Stroke 36:1753–1758, 2005. 426. van Dijk JM, Willinsky RA: Venous congestive encephalopathy related to cranial dural arteriovenous fistulas. Neuroimaging Clin N Am 13:55–72, 2003. 427. van Gelder JM: Computed tomographic angiography for detecting cerebral aneurysms: Implications of aneurysm size distribution for the sensitivity, specificity, and likelihood ratios. Neurosurgery 3:597–605, 2003. 428. Van Rooij SB, Bechan RS, Peluso JP, et al: Fenestrations of intracranial arteries. AJNR Am J Neuroradiol 36:1167–1170, 2015. 429. Velthuis BK, Rinkel GJE, Ramos LMP, et al: Subarachnoid hemorrhage: Aneurysm detection and preoperative evaluation with CT angiography. Radiology 208:423–430, 1998. 430. Velthuis BK, van Leeuwen MS, Witkamp TD, et al: CT angiography: Source images and postprocessing techniques in the detection of cerebral aneurysms. AJR Am J Roentgenol 169:1411–1417, 1997. 431. Vergouwen MD, Vermeulen M, van Gijn J, et al: Definition of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage as an outcome event in clinical trials and observational studies: Proposal of a multidisciplinary research group. Stroke 41:2391–2395, 2010. 432. Villablanca JP, Martin N, Jahan R, et al: Volume-rendered helical computerized tomography angiography in the detection and characterization of intracranial aneurysms. J Neurosurg 93:254–264, 2000. 433. Vinuela F, Drake CG, Fox AJ, et al: Giant intracranial varices secondary to high-flow arteriovenous fistulae. J Neurosurg 66:198–203, 1987. 434. Vinuela F, Fox AJ, Pelz DM, et al: Unusual clinical manifestations of dural arteriovenous malformations. J Neurosurg 64:554–558, 1986. 435. von Schulthess GK, Higgins CB: Blood flow imaging with MR: Spin-phase phenomena. Radiology 157:687–695, 1985. 436. Waaijer A, Prokop M, Velthuis BK, et al: Circle of Willis at CT angiography: Dose reduction and image quality—Reducing tube voltage and increasing tube current settings. Radiology 242:832–839, 2007. 437. Wallace R, Karis J, Partovi S, et al: Noninvasive imaging of treated cerebral aneurysms, part I: MR angiographic follow-up of coiled aneurysms. AJNR Am J Neuroradiol 28:1001–1008, 2007. 438. Wallace RC, Karis JP, Partovi S, et al: Noninvasive imaging of treated cerebral aneurysms, part II: CT angiographic follow-up of surgically clipped aneurysms. AJNR Am J Neuroradiol 28:1207–1212, 2007. 439. Waltimo O: The relationship of size, density and localization of intracranial arteriovenous malformations to the type of the initial symptom. J Neurol Sci 19:13–19, 1973. 440. Wang H, Li W, He H, et al: 320-detector row CT angiography for detection and evaluation of intracranial aneurysms: Comparison with

441.

442.

443. 444. 445.

446.

447.

448.

449. 450. 451.

452.

453. 454.

455.

456.

457.

458.

459.

460.

461.

462.

conventional digital subtraction angiography. Clin Radiol 68:e15–e20, 2013. Watanabe Z, Kikuchi Y, Izaki K, et al: The usefulness of 3D MR angiography in surgery for ruptured cerebral aneurysms. Surg Neurol 55:359–364, 2001. Weber W, Yousry TA, Felber SR, et al: Noninvasive follow-up of GDC-treated saccular aneurysms by MR angiography. Eur Radiol 11:1792–1797, 2001. Weibers DO, Whisnant JP, Sundt T, et al: The significance of unruptured intracranial aneurysms. J Neurosurg 66:23–29, 1987. Weir B: Pituitary tumors and aneurysms: Case report and review of the literature. Neurosurgery 30:585–591, 1992. Wermer MJ, Rinkel GJ, van Gijn J: Repeated screening for intracranial aneurysms in familial subarachnoid hemorrhage. Stroke 34:2788–2791, 2003. Wiebers DO, Piepgras DG, Meyer FB, et al: Pathogenesis, natural history, and treatment of unruptured intracranial aneurysms. Mayo Clin Proc 79:1572–1583, 2004. White PM, Teasdale EM, Wardlaw JM, et al: Intracranial aneurysms:   CT angiography and MR angiography for detection: Prospective blinded comparison in a large patient cohort. Radiology 219:739–749, 2001. White PM, Wardlaw JM, Easton V: Can noninvasive imaging accurately depict intracranial aneurysms? A systematic review. Radiology 217:361–370, 2000. Wilkins RH: Natural history of intracranial vascular malformation:   A review. Neurosurgery 16:421–430, 1985. Wilkins RH, Rengachary SS: Vascular malformations. In Neurosurgery, New York, 1985, McGraw-Hill, pp 1448–1473. Willinsky RA, Lasjaunias P, TerBrugge K, et al: Multiple cerebral arteriovenous malformations (AVMs): Review of our experience from 203 patients with cerebral vascular lesions. Neuroradiology 32:207–210, 1990. Wilms G, Demaerel P, Marchi G, et al: Gadolinium-enhanced MR imaging of cerebral venous angiomas with emphasis on their drainage. J Comput Assist Tomogr 15:199–206, 1991. Wilms G, Marchal G, Vas Hecke P, et al: Cerebral venous angioma: MR imaging at 1.5 tesla. Neuroradiology 32:81–85, 1990. Winn HR, Jane JA, Sr, Taylor J, et al: Prevalence of asymptomatic incidental aneurysms: Review of 4568 arteriograms. J Neurosurg 96:43–49, 2002. Winn HR, Richardson AE, Jane JA: The long-term prognosis in untreated cerebral aneurysms: I. The incidence of late hemorrhage in cerebral aneurysm: A 10-year evaluation of 364 patients. Ann Neurol 1:358–370, 1977. Wintermark M, Ko NU, Smith WS, et al: Vasospasm after subarachnoid hemorrhage: Utility of perfusion CT and CT angiography on diagnosis and management. AJNR Am J Neuroradiol 27:26–34, 2006. Wood EH: Angiographic identification of the ruptured lesion in patients with multiple cerebral aneurysms. J Neurosurg 21:182–198, 1964. Woodall MN, McGettigan M, Figueroa R, et al: Cerebral vascular malformations in hereditary hemorrhagic telangiectasia. J Neurosurg 120:87–92, 2014. Woodard E, Barrow D: Clinical presentation of intracranial arteriovenous malformations. In Barrow D, editor: Intracranial vascular malformations, Park Ridge, IL, 1990, AANS, pp 53–61. Wowra B, Muacevic A, Tonn JC, et al: Obliteration dynamics in cerebral arteriovenous malformations after cyberknife radiosurgery: Quantification with sequential nidus volumetry and 3-tesla 3-dimensional time-of-flight magnetic resonance angiography. Neurosurgery 64(2 Suppl):A102–A109, 2009. Xu J, Shi D, Chen C, et al: Noncontrast-enhanced four-dimensional MR angiography for the evaluation of cerebral arteriovenous malformation: A preliminary trial. J Magn Reson Imaging 34:1199–1205, 2011. Yagmulu B, Fitoz S, Atasoy C, et al: An usual cause of hydrocephalus: Aqueductal developmental venous anomaly. Eur Radiol 15:1159–1162, 2005.

http://radiologyebook.com

CHAPTER 12  Cerebral Aneurysms and Cerebrovascular Malformations 463. Yamada N, Hayashi K, Murao K, et al: Time-of-flight MR angiography targeted to coiled intracranial aneurysms is more sensitive to residual flow than is digital subtraction angiography. AJNR Am J Neuroradiol 25:1154–1157, 2004. 464. Yamada K, Kizu O, Ito H, et al: Tractography for arteriovenous malformations near the sensorimotor cortices. AJNR Am J Neuroradiol 26:598–602, 2005. 465. Yang MS, Chen CC, Cheng YY, et al: Unilateral subcortical calcification: A manifestation of dural arteriovenous fistula. AJNR Am J Neuroradiol 26:1149–1151, 2005. 466. Yen CP, Ding D, Cheng CH, et al: Gamma knife surgery for incidental cerebral arteriovenous malformations. J Neurosurg 121:1015–1021, 2014. 467. Yonakawa Y, Handa J, Okuno T: Moyamoya disease: Diagnosis, treatment, and recent achievement. In Barnett HJ, Stein BM, Mohr JP,

468.

469.

470.

471.

425

et al, editors: Stroke: Pathophysiology, diagnosis, and management, ed 2, New York, 1985, Churchill Livingstone, pp 805–831. Yoshida Y, Terae S, Kudo K, et al: Capillary telangiectasia of the brainstem diagnosed by susceptibility-weighted imaging. J Comput Assist Tomogr 30:980–982, 2006. Young N, Dorsch NW, Kingston RJ: Pitfalls in the use of spiral CT   for identification of intracranial aneurysms. Neuroradiology 41:93–99, 1999. Zambramski JM, Wascher TM, Spetzler RF, et al: The natural history of familial cavernous malformations: Results of an ongoing study. J Neurosurg 80:422–432, 1994. Zhao LB, Suh DC, Lee DG, et al: Association of pial venous reflux with hemorrhage or edema in dural arteriovenous fistula. Neurology 82:1897–1904, 2014.

http://radiologyebook.com

13  Traumatic Brain Injury David Dreizin, Uttam K. Bodanapally, Daniel Mascarenhas, and Thorsten R. Fleiter

INTRODUCTION Nearly 2 million cases of traumatic brain injury (TBI) occur each year.129,145 Worldwide, trauma is the leading cause of death and disability in those younger than 45 years, with nearly half of fatalities resulting from TBI.65,148 In younger patients, traffic accidents are the leading cause of TBI, and in the elderly, falls are the principal cause.10,32,65,81,102 TBI is a heterogeneous disorder with different forms of presentation.81 Approximately 15% to 25% of TBIs are serious injuries requiring specialist care, but the majority are considered mild.81,145 Mortality from severe TBI has decreased substantially with protocols that depend in part on computed tomography (CT) to monitor and recognize findings that may result in further injury.40,65 Magnetic resonance imaging (MRI) is an important problem-solving tool in moderate to severe TBI and has shown increasing utility for risk-stratifying mild TBI (mTBI), which can be associated with insidious long-term cognitive impairment.8,55

CLASSIFICATION OF TRAUMATIC BRAIN INJURY TBI is most broadly classified based on blunt or penetrating mechanism.10 This chapter emphasizes issues pertaining to the imaging of TBI from blunt head trauma. TBI may be intra- or extraparenchymal10; intraaxial injuries are further subdivided as focal or diffuse,148 the former occurring more commonly from falls in the elderly and the latter in high-velocity traffic accidents.65 Injury may be primary, resulting directly from external forces and structural strain to cerebral tissue at the time of the traumatic event,70 or secondary, representing evolution of the primary injury as well as consequences and complications of increased intracranial pressure (ICP).

MECHANISM OF INJURY—PRIMARY TBI Intraaxial Lesions Diffuse Axonal Injury, Contusions, and Intracerebral Hematomas.  TBIs usually results from either deceleration impact or impulsive acceleration and deceleration forces (e.g., “whiplash”), which cause variable degrees of compression, tension, and shear.70,102 Shear stress is caused by differential acceleration between peripheral and central portions of the brain with respect to the axis of rotation (often about the fulcrum of the cervical spine102,145 or from wave propagation through the head after direct impact.10,73 Diffuse axonal injury (DAI) is seen in approximately half of severe TBI cases73,76 and refers to the microstructural damage to white matter tracts resulting from shear injury.36,90,105,110 Injury occurs primarily at interfaces of different density such as the gray-white matter junction and the corpus callosum, and also tends  to cluster where the brain can collide and splay out against dural  reflections such as the falx and tentorium.10 This may account for the

426

predilection of DAI to the midline of the splenium105 and the dorsolateral midbrain (Fig. 13-1). Generally the more central the injury, the greater the overall severity,102 and severity is greatest when the midbrain or brainstem is involved. Contusions occur in approximately 45% of patients with TBI.76 Coup contusions are relatively common in falls that result in lateral and occipital blows to the calvarium.10,82 These are sometimes seen deep to the margins of depressed skull fractures73 (Fig. 13-2). Because coup contusions tend to involve the cortex immediately subjacent to the point of impact, as in DAI, contusions extending more centrally also tend to be more severe.102 Stress waves resulting from motion of the buoyant brain against the fixed inelastic calvarium and skull base account for the frequent presence of both coup and contrecoup contusions.65,68,102 Contrecoup contusions are usually more severe and are also more common than coup contusions, possibly owing to cavitation effects as the brain reflects off the inner table of the skull68 and the gliding movement of the undersurface of the brain over the rough surface of the anterior and middle cranial fossae. This also explains the susceptibility of the orbitofrontal regions (especially the gyrus rectus) and anterior inferior temporal lobes to contrecoup contusions10,68,73 (Fig. 13-3). Intraparenchymal hematomas involving deeper structures such as the thalamus, basal ganglia, hippocampus, and parahippocampal regions are distinct from cortical contusions, and like DAI can result from shear strain or may be caused by compression against the tentorium after blows to the vertex.10

Extraaxial Lesions Tensile failure of bridging veins that extend from the subarachnoid perivascular spaces to the dura accounts for the high frequency of subdural hematomas (SDH), particularly in the elderly and in children, where the vessels are maximally preloaded as a result of increased extraaxial space.73 The subdural space is normally occupied by fluid similar to cerebrospinal fluid (CSF).70 Because of this, the greatest amount of motion occurs in the subdural space, predisposing bridging veins to shearing.70 SDHs tend to spread out along the entire hemisphere73 because there is very little attachment between the inner leaflet of the dura and the arachnoid73 (Fig. 13-4). SDHs sometimes occur as a result of isolated impulsive forces without impact for this reason.8,102 Tearing of cortical veins resulting in SDH often occurs in the same distribution as coup and contrecoup contusions,70 and like contusions, a large proportion of SDHs, epidural hematomas (EDHs), and subarachnoid hemorrhage (SAH) result from lateral blows.10 Knowledge of the anatomy of fluid spaces and the relationship of the meningeal layers to the brain surface and inner table of the skull is important to understand common appearances of different extraaxial hemorrhagic lesions. The pia lines the cortical gyri as well as perforating vessels (bridging veins), forming a perivascular space (Virchow-Robin space)71 that communicates with the subarachnoid space, lying between the arachnoid and pia. SAH can result

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

B

A

FIG 13-1  Axial FLAIR images show high signal in the splenium (A) and dorsolateral midbrain (B) in a 21-yearold woman following a motor vehicle accident.

A

B FIG 13-2  Axial noncontrast CT of the head shows a depressed comminuted “eggshell”-type right parietal skull fracture (A) in a 28-year-old unhelmeted man after a motorcycle collision. B, A right parietal coup contusion is seen deep to the depressed skull fracture (arrow).

http://radiologyebook.com

427

428

PART II  CT and MR Imaging of the Whole Body

A

B FIG 13-3  This 37-year-old male pedestrian was struck by a vehicle. On CT (A), a left posterior temporal coup contusion is seen, with edema and hemorrhage resulting in a typical “salt-and-pepper” appearance (white arrow). Hemorrhagic contrecoup contusions involve the gyri recti (black arrows). B, On the axial FLAIR image, nonhemorrhagic cortical contrecoup contusions are also demonstrated, involving the bilateral anterior inferior temporal lobes (arrows). Previously described coup-contrecoup contusions are also seen to better advantage on this image.

from laceration of small cortical, pial, or subarachnoid vessels in the subarachnoid space, or the pial/subarachnoid portions of bridging veins. SAH may also result from transependymal diffusion of intraventricular hemorrhage (IVH) if the ependyma lining the ventricles is compromised,73 from direct migration through the ventricular outflow foramina,10,68 from direct extension of hemorrhage within a cortical contusion, from SDH with arachnoid tear, and from EDHs with dural and arachnoid tears. EDHs are more closely correlate with skull fractures than SDH and often occur in the temporoparietal region because the middle meningeal artery, which runs within a groove in the inner table of the squamous temporal bone, or its branches are the usual source of bleeding.68 The dura is a two-layer membrane with an inner meningeal layer and outer periosteal layer that is tightly bound to the skull. The two layers encase the venous sinuses.70 The epidural space is a potential space between the periosteal layer of the dura and inner table of the skull.70 The periosteal layer of the dura is tightly adherent to the inner table, particularly at the sutures, preventing most EDHs from crossing sutural margins10,70,73 (Fig. 13-5). Small EDHs without clinical deterioration can be treated conservatively but need to be followed closely because approximately one quarter enlarge within 36 hours after injury, usually within the first 5 to 8 hours70,140 (Fig. 13-6). IVH may result from shearing of subependymal veins during rotational acceleration, direct extension from parenchymal hematoma, or from retrograde migration through ventricular outflow foramina.73 FIG 13-4  The patient in Figure 13-3 had a hyperdense holohemispheric left subdural hematoma (arrow) on the same side as the coup contusion. There is associated bulk hemispheric swelling and subfalcine herniation (rightward shift) measuring more than 1 cm.

MECHANISM OF INJURY—SECONDARY TBI TBI is not a one-time event but rather a continuously progressive injury.68 Secondary brain injury may refer to the evolution and

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

B FIG 13-5  A, Left parietal epidural hematoma confined by the lambdoid suture posteriorly and coronal suture anteriorly in a 21-year-old following a motor vehicle accident. B, Oblique off-axial nondepressed fracture is shown on thin-section coronal CT image (arrow), traversing the left parietal bone, overlying the epidural hematoma.

A

C

B

FIG 13-6  A, A 16-year-old with right frontal epidural hematoma measuring 1.3 cm in thickness after being hit in the head by a falling tree limb. B, Because the hematoma contained an area of low density representing active bleeding, a follow-up CT was performed 2 hours later. On this study, the hematoma had increased in size, now measuring 1.7 cm, and new areas of low density indicating continued active bleeding were seen. C, Postoperative changes of craniotomy for evacuation of the hematoma. Two burr holes are seen at the anterior and posterior margins of the craniotomy on this axial image. To create the craniotomy flap, burr holes are first made and connected with a craniotome. These were sealed with burr hole covers (solid arrow). There is a subgaleal drain (open arrow), subgaleal gas, swelling, and overlying skin staples in keeping with immediate postoperative changes.

http://radiologyebook.com

429

430

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 13-7  A, CT of a 51-year-old with TBI after a fall. SDH, SAH, and bifrontal hemorrhagic contusions (right greater than left) are displayed. B, On follow-up CT the bifrontal contusions have both increased in size. Worsening mass effect from enlargement of contusions and SDH necessitated a wide decompressive craniectomy to allow brain parenchyma to herniate through the defect. Notice that there is essentially no midline shift after the wide craniectomy. C, On the T2-weighted MR image the corresponding mixed–signal intensity appearance is typical of hemorrhagic bifrontal contusions. Low signal intensity corresponds with acute blood product.

enlargement of the primary injury around a “traumatic penumbra” (analogous to the ischemic penumbra in stroke) (Fig. 13-7), or it may refer to the potentially preventable consequences of resultant mass effect and elevated ICP, including herniation, ischemia and infarction, and global hypoxic injury.10 Secondary brain injury is responsible for over 50% of deaths related to trauma within the first 48 hours  after injury.10,108,148 Brain swelling after TBI can be the result of both cytotoxicity and vasogenic edema, which have a synergistic effect.  The initial mechanical disruption results in cellular damage and  cytotoxicity85 that progresses throughout the first week after injury.70 Ischemic damage after primary brain injury is most commonly perilesional, resulting from impaired cerebral perfusion and oxygenation, excitotoxic injury, focal arterial or venous microthrombi, and vasospasm.8,27,81,106,133 On a cellular level, traumatic injury induces a vicious cycle beginning with the release of excitatory neurotransmitters such as glutamate, which causes a massive influx of calcium and sodium into neurons, leading to cellular swelling, mitochondrial dysfunction, free radical production, and calcium-mediated proteolysis. This in turn leads to ischemic and apoptotic cell death and more neurotransmitter release.58,68,70,129,148 Although cytotoxic edema appears to be the primary cause of cerebral swelling after TBI (Fig. 13-8), the bloodbrain barrier typically becomes compromised, leading to some degree of vasogenic edema.8,68 Mass lesions and swelling cause increased ICP, with resulting herniation that can occlude vessels and cause territorial infarction (Fig. 13-9). During the subacute period after the initial injury, microglial cells enter areas of focal injury and phagocytize blood products and damaged tissue.70 A cavitary area of encephalomalacia lined by glia results (Fig. 13-10). Small lesions may heal only with a glial scar.70 The overall effect is widespread volume loss beyond any focal area of TBI.

INDICATIONS FOR IMAGING The Glasgow Coma Scale (GCS) is a ubiquitously used scoring system for clinical grading of TBI severity, with eye, verbal, and motor components added to give a sum score ranging from 3 to 15.81 The lowest score in the first 48 hours is used to classify TBI. Scores of 13 to 15

correspond with mild TBI, scores of 9 to 12 with moderate TBI, and scores of 3 to 8 with severe TBI. In practice, the GCS can falsely suggest brain injury in the setting of intoxication, sedation, or paralysis. Another limitation is that most patients with TBI have mild injuries and may exhibit residual neurocognitive sequelae despite having normal or near-normal GCS scores.81 Nevertheless, the GCS serves as a useful triage tool for determining the need for CT. CT is generally indicated for severe and moderate TBI (GCS < 13).73,76 Some argue for imaging any patient with a GCS below 15,31,60,76,123,134,135 even though less than 20% of patients with minor head trauma have positive findings on CT and only up to 5% require intervention.62,76,135 Additional factors, such as high-velocity mechanism, contribute to the likelihood of positive CT findings138 even when GCS scores are normal.76 Other predictors used to determine the need for head CT in otherwise clinically mild head injury include risk factors such as advanced age (>60-65 years),53 loss of consciousness greater than 5 minutes,73 worsening level of consciousness, failure to improve over time,76 periods of transient amnesia or confusion,76,138 and headache with episodes of vomiting.8,53,73,76,136,137 Scanning should be performed more liberally in those on anticoagulation therapy or with bleeding diatheses73 and should be routinely performed for depressed skull fractures.73,76,121 CT is also important in allowing the neurosurgeon to clear patients prior to surgical treatment of non–life-threatening extracranial injuries.65 In unconscious patients, fixed dilated pupils are a sign of herniation requiring urgent imaging and decompression.65 Deteriorating mental status and delayed development of coma may reflect evolving secondary injury with worsening ICP and herniation or brainstem compression. DAI, which may not be apparent on CT, can also be the sole source of posttraumatic coma.36,129 Because of the time, resources, and logistics involved in MR scanning, particularly in polytraumatized patients with anesthetic and monitoring equipment, CT has long been used as the workhorse  to screen patients for prognostic indicators for surgical intervention and for following the dynamics of lesion development,14,16,17,73,76,81,101,111 although MRI remains an important problem-solving tool. Any focal lesions detected on CT, regardless of size, require follow-up CT scanning. New lesions will arise in around 15% of patients, and 25% to

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

C

B

E

D

FIG 13-8  A, Nonhemorrhagic contusions in the frontal and temporal regions are difficult to distinguish from a background of diffuse cerebral edema and extensive SAH at CT in this 38-year-old following a motor vehicle accident. B, T2-weighted image better shows the extensive contrecoup cortical contusions. The suprasellar and quadrigeminal plate cisterns are completely effaced, consistent with uncal and descendant transtentorial herniation, respectively. C, Uncal and transtentorial herniation is seen on sagittal T1-weighted image. Basal cisterns are completely obliterated. D, DWI shows lesional and perilesional cytotoxic edema in and around the regions of contusion. E, Corresponding low ADC in the contused areas confirms the high signal on DWI is from cytotoxicity rather than T2 shine-through, which would be seen in vasogenic edema. Restricted diffusion due to concurrent DAI is also seen in the corpus callosum.

A

C

B

FIG 13-9  This 36-year-old bicyclist was struck by an automobile. A, At CT, extensive bilateral contusions were shown to result in complete effacement of the suprasellar cistern from uncal herniation (white arrows) and near-total obliteration of the quadrigeminal plate cistern (black arrow) from descendant transtentorial herniation. This resulted in infarction of the superior right cerebellar hemisphere from compression of the superior cerebellar artery, which arises near the tip of the basilar artery and courses around the midbrain. B, DWI shows infarct of the superior right cerebellar hemisphere, resulting from compression of the superior cerebellar artery from uncal and descendant transtentorial herniation. The infarct may have also been caused by vasospasm from large central subarachnoid hemorrhage. C, At ADC, an area of low signal intensity corresponding with the territory of high signal on DWI is seen, confirming infarct.

http://radiologyebook.com

431

432

PART II  CT and MR Imaging of the Whole Body 45% of cerebral contusions will significantly enlarge81 (see Fig. 13-7). CT progression generally occurs within 6 to 9 hours after injury.81,96 CT facilitates a preemptive approach in which surgery is used to prevent deterioration rather than treating lesions once neurologic deterioration has already occurred.81 This strategy has been shown to improve outcome.20,119 CT is also used to determine the need for ICP monitoring.81,95 In general, ICP should be monitored in all patients with clinical evidence of severe TBI (GCS of 3-8) and any abnormal findings on CT.68,95

STRATIFYING INJURY SEVERITY WITH CT CT signs that have been correlated with increased ICP, coma, and death include midline shift greater than 3 to 5 mm (see Figs. 13-4 and 13-5), obliterated basal cisterns87,88 (see Figs. 13-3, 13-8, and 13-9), diffuse hemispheric swelling, SDHs greater than 10 mm in thickness27,68,70,153 (Fig. 13-11), EDHs greater than 15 mm in thickness65,68,75,77,113 (see Figs 13-5A and 13-6B), SAH, and IVH.27,68,80,106,122 Any intracranial mass lesion may warrant surgical evacuation, whether intracerebral hematoma (ICH), hemorrhagic contusion, SDH, or EDH once the size is large enough to cause brain distortion and elevated ICP.148 SDHs have a higher mortality than acute EDHs,68 which may seem counterintuitive given that EDHs often have an arterial source. However, the force required to separate the dura from its periosteal attachment to the inner table of the calvarium is considerable, whereas SDHs can continue to increase with little restriction as additional bridging veins tear when stretched beyond their capacity. Significant improvement in mortality is seen when SDHs are evacuated less than 4 hours after

FIG 13-10  Evolving encephalomalacia replacing contrecoup contusions in a 41-year-old woman after motor vehicle collision (arrow). There is associated ex vacuo dilation of the right temporal horn.

A

B FIG 13-11  A, CT in an 82-year-old after a fall from ladder shows a hemorrhagic contusion in the right frontal region (arrow), with subdural blood (open arrow) that likely migrated from the contusion through a tear in the arachnoid into the subdural space. The holohemispheric SDH is greater than 1 cm in thickness and is associated with bulk right hemispheric swelling and marked leftward transfalcine herniation. B, Marked transfalcine herniation has resulted in entrapment of the left temporal horn (solid arrow), and there is associated locoregional transependymal spread of CSF (open arrow).

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury injury.119,148 Intraparenchymal mass lesions typically result in less mass effect than extraaxial hemorrhagic mass lesions and are less ominous predictors of poor outcome27; contusions or ICHs, however, particularly in elderly patients with low GCS scores, are likely to benefit from prompt intervention.44,76 Classification systems relying on mass effect and mass lesions on CT are limited in patients with DAI.81

CT Findings in Primary Injury

Cortical Contusions.  Contusions are common, occurring in 43% of closed head injuries, frequently with coexisting coup and contrecoup injuries,76 and can be multiple or solitary.10 Cortical contusions represent bruising and laceration of the brain and primarily involve gray matter. Contusions are often based at the cortical surface and extend toward the center of the brain.10 On noncontrast CT, contusions appear as peripheral low-attenuation areas76 with loss of gray-white differentiation, reflecting the combination of cytotoxic edema and vasogenic edema involved. In most cases, some hemorrhage is present at MRI, but this may not be detectable at first with CT. When contusions are nonhemorrhagic on MRI, they may also be difficult to appreciate  on CT,38,54 with subtle areas of low-density edema68 (Fig. 13-12). Hemorrhagic contusions may take the form of multiple areas of microhemorrhage interspersed with low density, giving a heterogeneous “salt-and-pepper” appearance73 (see Fig. 13-3A), or they may appear as more coalescent areas of hemorrhage, giving a more uniformly hyperdense appearance76 (see Fig. 13-7B). As the injury progresses, small hemorrhages may coalesce into larger hematomas.70

Intracerebral

Hematoma.  Whereas ICHs can result from

coalescent hemorrhage in contusions (Fig. 13-13), deeper lesions often

A

433

result from shear-induced hemorrhage and rupture of small intraparenchymal vessels. Unlike contusions, these often occur in areas of otherwise normal-appearing brain and initially have less surrounding edema.73 These often occur within frontotemporal white matter, sparing the cortex. Deeper lesions may occur in the basal ganglia, but bleeding in this area may also suggest a possible hypertensive hemorrhage.73 ICHs can be a marker of concurrent DAI even when typical hemorrhagic foci at the gray-white matter junction are absent on CT73 (Fig. 13-14).

Diffuse Axonal Injury.  DAI is thought to be present in over 50% of severe head injuries and in more than 85% of severe head injuries resulting from motor vehicle accidents.58 The key point to remember is that DAI is more the rule than the exception in severe TBI patients. DAI presents with the classic CT findings of petechial hemorrhage (an indirect sign of axonal injury) in only 10% to 50% of cases50,68,76,91,94,107 (Fig. 13-15). DAI can be appreciated on CT when small hemorrhagic foci are evident, typically occurring at the gray-white junction  (Fig. 13-16) and, in more severe cases, in the corpus callosum complex, which includes the septum pellucidum, fornix, and choroid.  Hemorrhagic DAI can also occur in the periventricular white matter at the corner of the frontal horns and in the hippocampal and parahippocampal areas, internal capsule, and cerebellum37,105 (Fig. 13-17). Common areas of hemorrhage should be examined very thoroughly for minimal lesions.10

Subdural Hematoma.  In general, the incidence of SDH and EDH is higher when skull fractures are present.10,82 SDHs occur along the convexities, tentorium, and falx in descending order of frequency.68

B FIG 13-12  Images of a 37-year-old male pedestrian struck by a car. A, Hypodensity is seen in the bifrontal regions, right (open arrow) greater than left. There is sulcal SAH (solid arrow), but no intraparenchymal hemorrhage is appreciated. B, On FLAIR, nonhemorrhagic cortical contusions corresponding to the areas of low density on CT are visible (open arrow). The acute subarachnoid blood appears as low signal intensity on T2/FLAIR (solid arrow). The round area of low signal intensity in the right frontal region corresponds with an extraventricular drain seen in A. Continued

http://radiologyebook.com

434

PART II  CT and MR Imaging of the Whole Body

C

D FIG 13-12, cont’d  C, On SWI, which is highly sensitive to para- and ferromagnetic substances in blood, the SAH is redemonstrated. Tiny foci of subarachnoid blood are also seen on the right. D, On DWI, cortical contusions are apparent in a wider distribution than what is appreciable on FLAIR.

H

A

B FIG 13-13  Images of a 70-year-old man after a rollover motor vehicle collision. A, Right frontal intracerebral hematoma (open arrow) resulting from coalescent hemorrhage in a contusion deep to a complex skull fracture (shown with volumetric imaging in B). There is an overlying SDH, likely resulting from active bleeding arising from the intraparenchymal hematoma, with a relatively low-density area representing liquid blood (solid arrow). There is also scattered SAH. There is a depressed “eggshell” fracture involving the vertex (B) and vertically oriented displaced fractures propagating down the frontal and parietal bones on the right.

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

435

C

B

D

E

FIG 13-14  Images of a 68-year-old following a motor vehicle collision. A, CT shows intraparenchymal hematoma with thin halo of edema in the left temporal region, sparing the cortex, and surrounded by otherwise normal-appearing brain parenchyma. B, Low T2 signal within the intraparenchymal hematoma on FLAIR, with surrounding high signal, corresponding with the acute hemorrhage and rim of edema seen at CT. C, On SWI, blooming artifact accentuates the deoxyhemoglobin within the intraparenchymal hematoma. Foci of hemorrhage are also seen at SWI throughout the gray-white junction of the bilateral convexities (D), along the septum pellucidum, splenium of the corpus callosum, and within the lateral ventricles (E).

SDHs follow dural reflections, spreading out along the anterior or posterior falx cerebri and typically the supratentorial aspect of the tentorium cerebelli (Fig. 13-18), which may be symmetric or appear as asymmetric high density68 (Fig. 13-19). Unlike EDHs, SDHs are not restricted by suture lines.68 SDHs are usually described as being crescent shaped,10 in contradistinction to EDHs, which dissect between the dura and inner table, resulting in a biconvex (“lentiform”) appearance. The density of acute SDH is high because of clot formation and retraction (see Fig. 13-7A). Density is highest in the first week after injury and then gradually decreases with protein degradation and liquefaction.73 In acute SDHs, heterogeneity can result from active bleeding9,68,73 (see Fig. 13-13A). SDHs may appear isodense to brain parenchyma in the acute setting if there is severe anemia or traumainduced coagulopathy or when there is admixture with subarachnoid fluid crossing through an arachnoid tear68 (Fig. 13-20). Diffuse underlying cortical swelling is common with SDH10; 85% of cases of diffuse swelling and bulk enlargement of a cerebral hemisphere are associated with SDHs.79

Epidural Hematoma.  EDHs are relatively uncommon, occurring in approximately 4% of head trauma patients.76 EDHs mostly occur

in young adults. In the elderly, the dura is more strongly affixed to  the skull, and in children, plasticity of the calvarium mitigates  against EDH. In 91% of EDHs, a skull fracture is seen on CT70,152 (see Fig. 13-5). EDH typically occurs at the site of impact (coup site) in contradistinction to SDH.148 As EDHs dissect along the tightly adherent inner table and dura, the shape formed by the leading edges is well defined.10 As with SDH, mixed density may indicate active bleeding. Areas of often central and irregular low density within hyperdense hematoma, representing admixture of liquid blood and fresh clot, has been termed swirl sign.10,152 Low-density liquid blood can also occur eccentrically or peripherally (see Figs. 13-5 and 13-6). Low-density liquid blood with an EDH is associated with massive hemorrhage and poor outcome.41 Unlike SDHs, EDHs cross dural reflections. EDHs usually do not cross suture lines unless the sutures themselves are disrupted or diastatic (Fig. 13-21). The sagittal sutural line is an exception because the dura is not as tightly attached to the periosteum as it is elsewhere73 (Fig. 13-22). Occipital EDHs, often venous in origin, can extend both across the midline and above and below the tentorium (Fig. 13-23), whereas SDHs are restricted by dural reflections  (Fig. 13-24). Venous EDHs may cause the dural sinuses to be displaced away from the inner table of the calvarium68 (see Fig. 13-21C). Venous

http://radiologyebook.com

436

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F

G

H

I

FIG 13-15  Images of a 20-year-old woman following a motor vehicle collision. CT at the level of the corpus callosum (A) and at the level of the basal ganglia (B) show no visible foci of hemorrhage, but in A, there are small foci of intraventricular hemorrhage in the occipital horns. C, At SWI, numerous petechial foci of hemorrhage are seen in classic locations, including the septum pellucidum, choroid, and splenium of the corpus callosum. D, More caudally, foci of hemorrhage are also seen in the basal ganglia. E, On conventional T2*GRE, foci of hemorrhage are less conspicuous and fewer in number than at SWI. Several petechial foci are seen in the right basal ganglia and right parahippocampal region. F, On DWI, several nonhemorrhagic foci of DAI are identified in the deep white matter along the bodies of the lateral ventricles (arrows) that were not appreciated on the other sequences. These foci show restricted diffusion on the ADC map (G). H, On the FLAIR sequence, CSF is nulled out. The bright signal within the parietal sulci represents subacute subdural blood (arrows). I, The SAH is obscured on conventional T2-weighted imaging because the CSF is also bright.

http://radiologyebook.com

A

C

B

E

D

FIG 13-16  This 22-year-old suffered DAI after a motor vehicle collision. A, At CT, a small hemorrhagic lesion is seen at the parasagittal gray-white matter junction near the vertex (arrow). B, FLAIR shows the more extensive shear injury in this region, most of which was not appreciable on CT. Additional areas of DAI are revealed with FLAIR in the left basal ganglia (C), right parahippocampal region (D [open arrow]), and dorsolateral brainstem (D [solid arrow]). E, Hemorrhage is more conspicuous in the same areas at SWI.

A

B FIG 13-17  A, Several hemorrhagic lesions are seen at the corner of the right frontal horn in a 31-year-old with DAI following motor vehicle collision. B, At SWI, much more extensive hemorrhagic lesions are apparent in the same region at the corner of the right frontal horn, also involving the genu of the corpus callosum. Lesions not apparent on CT are revealed in the left basal ganglia (solid arrow) and fornix (open arrow).

http://radiologyebook.com

438

A

PART II  CT and MR Imaging of the Whole Body

B FIG 13-18  A 19-year-old with TBI following a motor vehicle accident. On CT, acute SDH is seen propagating along the posterior falx (A [arrow]), and tentorial leaflets (B [arrows]).

A

B FIG 13-19  A 40-year-old woman with TBI following a motor vehicle accident. A, A high-density SDH is seen spreading out along the right tentorial leaflet. Unlike EDHs, SDHs may be indistinct and can blend with SAH, seen in B in the quadrigeminal plate cistern (solid arrow). A small focus of SAH is also seen in the interpeduncular cistern (B [open arrow]).

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

439

C

B

FIG 13-20  Images of a 25-year-old driver of a sedan that was “run over” by a tractor trailer. The patient sustained injury to multiple body regions and large-volume resuscitation. A, A thin isodense right SDH is barely perceptible on head CT, even with the use of low-level wide window settings, which accentuate subdural blood. There is resultant bulk hemispheric swelling on the right, resulting in a 3-mm leftward midline shift. B, An image from the arterial phase neck CT is shown. There is displacement of the cortical vessels away from the inner table of the calvarium (arrows). C, On the contralateral side, vessels extend all the way to the inner table (arrow).

A

B

C

FIG 13-21  A 21-year-old after a fall while skateboarding. A, A biconvex epidural hematoma is seen, which crosses the lambdoid suture. An indentation of the dura (arrow) is seen immediately over the suture, likely representing the prior point of sutural attachment. B, Diastatic lambdoid suture is shown on bone window (arrow). C, CT venogram shows a distorted extrinsically narrowed left transverse sinus displaced off the inner table of the calvarium (arrow), with overlying (likely venous) EDH and sutural diastasis.

EDHs (representing only 15% of EDHs overall) typically occur in three locations: the posterior fossa from torcular injury, middle cranial fossa from sphenoparietal sinus injury, and vertex from injury to the superior sagittal sinus.73 Venous EDH also occurs in the regions of the transverse and sigmoid sinuses.68,70 These EDHs tend to enlarge less quickly than arterial EDHs.

Intraventricular Hemorrhage.  IVH is uncommon, occurring in approximately 3% of newly diagnosed cases of TBI, but it contributes significantly to morbidity and mortality, with half of patients with IVH developing increased ICP, and 10% requiring ventricular drainage.76 IVH usually appears as a hyperdense blood-fluid level layering dependently in the ventricular system.73 In small hemorrhages this can be

subtle, with punctate foci appearing in the tips of the occipital horns (see Fig. 13-15A). Occasionally IVH may appear tumefactive as a cast within the ventricle73 (Fig. 13-25).

Subarachnoid Hemorrhage.  SAH occurs in 40% of cases of moderate to severe head injury and is usually associated with other types of intracranial hemorrhage.68 The volume of extravasated blood is rarely as large or extensive as in aneurysmal SAH and more often has a peripheral distribution, sometimes adjacent to areas of cortical contusion.70,111 Acute SAH is readily identified on noncontrast CT as linear and serpentine areas of high attenuation extending into and con­ forming to the cerebral sulci and sylvian fissures without mass effect68,73 (Fig. 13-26; see also Fig. 13-12A). SAH along the convexity, tentorium,

http://radiologyebook.com

440

PART II  CT and MR Imaging of the Whole Body

A

B FIG 13-22  A 21-year-old after a fall from height. A, EDH is seen over the vertex on coronal CT. B, This crosses the midline and results in sagittal suture and coronal suture diastasis, seen on volumetric 3D CT image.

A

B FIG 13-23  This 17-year-old sustained an occipital bone fracture after a fall from height and was found to have an EDH (A), which enlarged on short-interval follow-up CT (B). The EDH extended above (A) and below (B) the tentorium. Areas of low density correspond with active bleeding.

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

441

B FIG 13-24  A, Occipital SDH in a 71-year-old woman after a fall from standing. Unlike occipital EDHs, SDHs do not cross dural reflections. B, Sagittal reformatted image shows the SDH coursing along the inferior aspect of the tentorium.

or falx can be difficult to differentiate from SDH.73 Small foci of SAH at the interpeduncular cistern may be easily missed without conscious attention to this area68 (see Fig. 13-19B). Although SAH is usually associated with other findings, it may be the sole finding in a minority of cases.68,122 In the subacute setting, SAH becomes isodense and may be very subtle; effacement of sulci may be the only imaging clue.73 The reason for increased mortality is uncertain but is hypothesized to be related to vasospasm.27

CT OF SKULL FRACTURES There are three main types of skull fractures: linear and depressed calvarial fractures and basilar fractures. Linear fractures are the most common, seen frequently with falls from standing,102 and are least likely to be associated with intracranial injury. Linear fractures  of the calvarium are easily appreciated on axial CT images if the plane of the fracture is vertically oriented. Axially oriented fractures can be easily missed but are better seen on multiplanar reformats (see Fig. 13-5B) and three-dimensional (3D) volume rendered images70; 3D images are also useful for depicting comminuted and depressed fractures. Basilar fractures are best evaluated with dedicated thin-section scanning (see Fig. 13-25D). Depressed or markedly displaced fractures are more commonly associated with intracranial abnormalities such as dural tears, which can result in CSF leak. Thin-section CT has been shown to be both more sensitive than radionuclide or CT cisternography for identifying the site of  leak (commonly near paranasal sinuses and mastoid air cells) and correlates more closely with operative findings78,139 (Fig. 13-27). Depressed fractures can also cause arachnoid tears, which can result in low-density hygromas. Depressed fractures are explored and repaired, and bone fragments elevated both for cosmesis70 and to prevent additional cortical laceration, dural tear, or vascular injury by sharp bony fragments. In general, open skull fractures depressed more than the full thickness of the skull should be surgically 

elevated.76 Débridement may be necessary if there is significant fragmentation of the skull.65

CT of Vascular Injuries Vascular injuries are commonly related to skull base fractures. CT venogram may be warranted for fractures that propagate in the region of the transverse and sigmoid sinuses. CT venography is at least as accurate as MR venography for dural venous sinus thrombosis (DVST) and is better at visualizing small cerebral veins104 (Fig. 13-28). In the absence of a fracture line extending to a dural venous sinus or jugular bulb, the likelihood of traumatic DVST is very low.23 On the other hand, when skull fractures extend to a dural sinus or jugular bulb, approximately 40% will have DVST, which is occlusive in over half of cases and is associated with hemorrhagic venous infarct at MRI in 7%.23 Arterial injuries may occur with skull base fractures extending through the petrous temporal bone into the carotid canal or propagating through the sphenoid in the vicinity of the cavernous carotid.70 These fractures warrant further evaluation with CT angiography (CTA) because vascular injuries can result in infarcts and cerebral ischemia.20,73 Carotid cavernous fistulae are characterized by asymmetric enhancement of the cavernous sinus and venous engorgement, especially of the superior ophthalmic vein and inferior petrosal sinus,73 as well as enhancement of draining veins73 (Fig. 13-29). CTA offers slightly improved spatial resolution compared to current MR angiography (MRA) protocols, with fewer flow-related artifacts and better delineation of adjacent bony anatomy45,76,146; however, MRA allows follow-up of injuries without additional exposure to ionizing radiation. In addition to skull base fracture, neck trauma is another major risk factor for cerebral infarction, and neck CTA is often performed concurrently. Bilateral segmental wedge-shaped areas of low density and loss of gray-white differentiation on noncontrast head CT can suggest thromboembolic phenomena from blunt cerebrovascular injury (Fig. 13-30).

http://radiologyebook.com

442

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 13-25  TBI and skull base fracture in a 28-year-old following a motorcycle collision. Patient had previously undergone decompressive frontotemporoparietal craniectomy. Two weeks after the initial injury, he developed delayed intraparenchymal hematoma (A) and substantial IVH, which formed casts within the lateral, third, and fourth ventricles. B, Tumefactive IVH in the third ventricle is shown (arrow). C, A repeat CT angiogram revealed the source of hemorrhage to be a newly formed pseudoaneurysm of the right anterior cerebral artery (arrow). D, Skull base fracture extending through the sphenoid is shown.

CT PERFUSION Perfusion CT is frequently used in acute stroke but may have a role in TBI patients,76,149 in whom abnormalities on perfusion CT reflect abnormalities in autoregulation. Patients with minor head trauma  and intact autoregulation will maintain or have slightly increased  cerebral perfusion, whereas patients with severe injury have impaired autoregulation and pressure-passive flow, resulting in oligemia and decreased perfusion68 (Fig. 13-31). Multiple sequential dynamic images are required for perfusion CT to track contrast material as it passes through the brain,76 and high radiation dose remains a concern,68 although radiation doses are lower with newer MDCT scanners,

improved detectors, and low mAs protocols. Normal perfusion  or hyperemia (high cerebral blood volume [CBV] and cerebral blood flow [CBF]) are associated with favorable clinical outcomes, whereas oligemia (low CBV and CBF) are associated with unfavorable  outcomes68,76,149 (Fig. 13-32).

LIMITATIONS OF CT CT can miss small amounts of blood, linear fractures, and foci of pneumocephalus, owing to volume averaging.76 Beam hardening limits evaluation of the posterior fossa, frontal, and temporal regions.20,64,76 CT findings may lag behind clinical deterioration, and exams

http://radiologyebook.com

FIG 13-26  In the 21-year-old woman shown in Figure 13-1, DAI was not apparent on CT, but focal SAH was seen within right frontal lobe peripheral sulci (arrow). There is a large overlying right frontal scalp contusion and hematoma.

FIG 13-27  Coronal CT of the face of a 17-year-old with CSF rhinorrhea shows fracture of the left planum sphenoidale (arrow), which also extended to the cribriform plate (not shown). The CSF leak was successfully treated with shunting.

A

B

C

D

E

F

FIG 13-28  A, CT of a 20-year-old after motor vehicle collision. High-density material is seen in the expected location of the left sigmoid sinus. B, CT venogram showed segmental nonfilling in keeping with occlusive thrombus in the left sigmoid sinus (arrow). C, MR venogram without gadolinium shows loss of flow-related enhancement from occlusive thrombus in the left sigmoid sinus (arrow). D, Full-volume maximum intensity projection (MIP) image from MRV confirms lack of flow in the distal transverse sinus, sigmoid sinus, and jugular bulb. E, Sagittal T1-weighted image shows high signal thrombus within the transverse and sigmoid sinus (arrows). F, Temporal bone CT shows fracture involving the petrous and squamous temporal bone extending through the sigmoid notch.

http://radiologyebook.com

444

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 13-29  In this 37-year-old pedestrian struck by a car, asymmetric opacification of the cavernous sinus (A) and engorged early draining right ophthalmic vein (B [arrow]) were highly suggestive of carotid cavernous fistula, later confirmed angiographically and treated with coil embolization.

performed within several hours of trauma may underestimate injury.29 There is still debate over whether CT scans should be repeated after normal admission CT.76,126

MRI IN PRIMARY INJURY MRI is superior to CT for many forms of posttraumatic lesions,3,37,67,77,103,111 although skull fractures are a major exception.10,103 MRI is often reserved for screening for focal or diffuse parenchymal lesions in the setting of a CT scan that does not adequately explain depressed consciousness.10,70,111 This includes DAI (see Figs. 13-1 and 13-14 through 13-17) and nonhemorrhagic contusions77,111 (see Figs. 13-8 and 13-12). Owing to its multiplanar capabilities, MRI is far superior to CT for detecting herniation, especially through the tentorial incisura and foramen magnum30,37,76 (see Figs. 13-8B and C). MRI is also better for characterizing injury evolution and recognizing complications, and for characterizing degenerating blood products, which vary in signal intensity with time.

Intraparenchymal Injury A variety of MR sequences can be used concurrently to identify, characterize, and determine the extent of focal brain injuries. Because fluid-attenuated inversion recovery (FLAIR) nulls the high signal from CSF, this sequence is sensitive for detecting periventricular lesions such as foci of DAI (see Figs. 13-1, 13-14B, and 13-16B-D) as well as superficial cortical contusions20,73 (see Figs. 13-3B and 13-12B) and vasogenic edema.3,73,76 Gradient echo images (GRE) are T2* weighted and sensitive to magnetic susceptibility effects of para- and ferromagnetic substances112,144 (see Fig. 13-15E). In recent times, GRE has been all but supplanted by susceptibility-weighted imaging (SWI). Unlike GRE,

SWI enhances contrast of magnetic susceptibility by combining magnitude and filtered phase images from a high-resolution 3D gradient. SWI has been shown to be up to 5 or 6 times more sensitive to hemorrhage than GRE8,112,118,120,142 and is less prone to artifacts at air-tissue interfaces.39,46,132 SWI is highly sensitive to field inhomogeneities caused by paramagnetic substances, including deoxyhemoglobin and intracellular methemoglobin, and is also highly sensitive for detecting ferritin and hemosiderin, which are both ferromagnetic.8,20 SWI is therefore useful for detecting hemorrhage in acute, subacute, and chronic phases. One major disadvantage of SWI is that images take longer to perform than GRE and can easily be degraded by motion in agitated patients.39 Diffusion-weighted imaging (DWI) detects decreases in random brownian motion in injured swollen cells, resulting in bright signal on DWI as water moves from the extracellular to intracellular space.20,70 DWI is sensitive to acute cytotoxic edema and by extension is  highly sensitive for detecting nonhemorrhagic traumatic lesions (see Figs. 13-8D, 13-12D, and 13-15F).

Contusions and Intraparenchymal Hematomas On MRI, hemorrhagic contusions and intraparenchymal hematomas appear as ill-defined areas of mixed signal intensity on T2-weighted images73 (see Fig. 13-7C). Because cerebral contusions primarily involve the periphery of the brain, contusions often have a gyral morphology73 (see Figs. 13-8B and 13-12). Nonhemorrhagic contusions may be difficult to detect initially on conventional sequences, including FLAIR, but the resultant cellular dysfunction causes early cytotoxicity. Restriction on DWI may occur as early as 1 hour after injury51 and gives a more accurate estimate of the true extent of injury than FLAIR10,70,73 (compare Fig. 13-12B and D). High signal on DWI can also be caused by T2 shine-through from vasogenic edema, and images

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

B

C

D

445

FIG 13-30  A, Volume-rendered image from neck CTA shows left carotid artery pseudoaneurysm (arrow) in a 27-year-old pedestrian struck by a car. B, Bilateral wedge-shaped hypodensities with loss of gray-white matter differentiation are seen at the convexities, in keeping with embolic infarcts. C, Filling defect from embolized thrombus is seen in the proximal left MCA. D, DWI shows large area of restricted diffusion consistent with left MCA territory infarct.

must be interpreted together with apparent diffusion coefficient (ADC).8,26,57 DWI does not routinely distinguish between cytotoxic edema from trauma and cytotoxicity from ischemia, but the latter is suggested by a vascular distribution rather than perilesional involvement (compare Fig. 13-31D and E).

Brainstem Injuries A very small percentage of brainstem lesions are seen at CT, owing to beam-hardening.37,52,10,83 Brainstem lesions are encountered in up to 64% of comatose patients after severe head injury.10,33 Contusion of the brainstem can occur in isolation, but this is rare, and concomitant

lesions are seen in most cases.33 Bilateral pontine lesions are associated with poor outcome and are often fatal.18 Decompressive craniectomy or other invasive measures may be medically futile in such cases.33 Primary and secondary brainstem lesions may be distinguishable based on their MR characteristics.33 Direct traumatic brain laceration typically occurs at the dorsolateral midbrain and pons, where these structures can collide posteriorly with the dural reflections of the tentorial incisura. Because of the severe forces required to produce shear and direct destruction of tissue in this region, most are hemorrhagic,33 whereas secondary lesions caused by herniation are less likely to bleed.33 Although Duret hemorrhages can occur in brainstem

http://radiologyebook.com

446

PART II  CT and MR Imaging of the Whole Body

B

A

C

D

E

FIG 13-31  CTP study and MRI from a 36-year-old bicyclist struck by a car (same patient as Fig. 13-9). The patient has had a decompressive craniotomy, and brain contents are seen herniating through the right-sided craniotomy defect. There is decreased blood volume (A) and blood flow (B) within areas of bilateral frontal lobe hemorrhagic contusion, herniating frontotemporal brain contents, and right cerebellar hemisphere infarct (C). D and E, On DWI, corresponding restricted diffusion is seen in all these areas.

herniation, these are usually differentiated from primary traumatic injury by their location in the ventral and paramedian midbrain68,105 (Fig. 13-33). Duret hemorrhages are thought to occur as a result  of venous thrombosis or laceration of the pontine perforators of   the basilar artery from tension related to caudal herniation of the brainstem.68,70

Diffuse Axonal Injury Even with the best imaging techniques, only a minority of DAI lesions, which are primarily microscopic, can be identified.58,105 Brainstem injuries are very rarely present in the absence of DAI involving more peripheral structures. Mild DAI (grade I) involves only peripheral gray-white matter junctions; moderate DAI (grade II) involves  the corpus callosum (usually the splenium and posterior body); and severe DAI (grade III) involves the dorsolateral midbrain or brainstem.1,68,70,73,105 In the midbrain, the posterolateral area adjacent to the superior cerebellar peduncle is most frequently affected105 (see Fig. 13-1B). DWI is the most sensitive sequence for detecting nonhemorrhagic DAI in the acute setting. Lesions are hyperintense on DWI and hypointense on ADC.58,68 SWI is extremely sensitive for the detection of concomitant microhemorrhages in DAI68,120 (see Figs. 13-14 through 13-17). Although FLAIR is less sensitive for acute DAI lesions, sagittal and coronal FLAIR images can be helpful for detecting DAI involving the corpus callosum and fornix.68,73

Extraaxial Injuries MRI is more sensitive for thin extraaxial “smear collections,” SAH, and IVH when FLAIR is used4,73,76,99 (Fig. 13-34). Hypointensity on T2/ FLAIR is in keeping with acute blood (see Fig. 13-12B), and hyperintensity on T2 suggests subacute bleed or hygroma147 (Fig. 13-35). With SWI, MRI has excellent sensitivity for detecting acute extraaxial hematomas but little added value for this purpose with respect to CT. MRI has greater utility for characterizing extraaxial collections in the subacute and chronic phases, discussed in the subsequent section.

ROLE OF IMAGING IN SECONDARY BRAIN INJURY AND COMPLICATIONS Common findings in secondary brain injury include worsening or  new hemorrhage and worsening vasogenic or cytotoxic edema, all of which contribute to increased ICP and can result in herniation, ischemia, and infarction. CT is used to determine the need for extraventricular drain placement, hematoma evacuation, and wide craniectomy,70 and also for monitoring improvement and determining when a bone flap can be replaced (Fig. 13-36). When ICP increases in patients with ICP monitors, CT is performed to exclude delayed traumatic ICH, infarct, or posttraumatic hydrocephalus.148 CT can also screen for

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

A

B

C

D FIG 13-32  A, CT demonstrates diffuse cerebral hypodensity with relatively increased density of the cerebellum (“white cerebellum sign”) in a 19-year-old pedestrian struck by a car. B, On follow-up CT, the marked diffuse cerebral hypodensity is more apparent (“big black brain sign”). Cerebral hypodensity brings the meninges into relief, which appear relatively hyperdense (pseudosubarachnoid hemorrhage sign). Findings are consistent with anoxic brain injury. Areas of low CBV (C) and low CBF (D) are seen throughout most of the cerebral hemispheres bilaterally, with an area of relative sparing in the posterior left temporal region. High CBV and CBF are seen in the cerebellum, indicating compensatory increased perfusion through the vertebrobasilar circulation. Shortly thereafter, nuclear scintigraphy showed no perfusion of the the cerebral hemispheres, confirming a clinical diagnosis of brain death.

http://radiologyebook.com

447

448

PART II  CT and MR Imaging of the Whole Body

A

B FIG 13-33  A, In this 51-year-old with TBI after a fall, SWI shows Duret hemorrhage (arrow) in the midline of the brainstem, emanating from the ventral pons and appearing to arise from the basilar artery (arrow). This likely resulted from injury to a pontine perforating branch. B, More inferiorly, the Duret hemorrhage is seen to continue to extend from the ventral to dorsal pons. Additionally there are two dorsolateral hemorrhagic foci (arrows), consistent with primary traumatic injury resulting from collision of the midbrain with the tentorial leaflets.

postoperative complications such as hematomas, which occur in 15% of TBI patients.148 MRI is generally considered diagnostically superior to CT 48 to 72 hours after injury; this is because the ability of MRI to depict the full extent of hematomas improves over time as blood composition changes.76 In the subacute stage, SAH may be isodense to brain parenchyma on CT but is high signal intensity on FLAIR (see Fig. 13-15G). Chronic SAH is also better seen owing to superficial hemosiderosis.73 MRI is better at visualizing and characterizing all herniations because of its multiplanar capabilities. DWI and ADC are exquisitely sensitive for ischemia and infarction resulting from herniation (see Figs. 13-31E and 13-35C) and for characterizing  the full extent of both hemorrhagic and nonhemorrhagic contusions (see Fig. 13-12D).

Contusions, Intracerebral Hematomas, and Hypoxic Encephalopathy

FIG 13-34  A thin subacute subdural hematoma is seen along the left convexity (arrow) in a 52-year-old woman after a fall from standing. This is well visualized as CSF is nulled out on FLAIR.

Within the first week after injury, it is common for hemorrhage within contusions to progress and for large contusions to develop new delayed hemorrhage and worsening edema (see Fig. 13-7). Surprisingly, perilesional edema does not appear to be a strong predictor of subsequent size increases in hemorrhagic contusions.7 Perilesional edema also seems to be uncommon in contusions that do evolve.7 ICHs may develop in a delayed fashion in a part of the brain previously appearing normal on CT; this is known as delayed traumatic ICH.68 In one study, all traumatic ICHs developed within 24 hours of onset, but only  56% of ICHs developed within 6 hours of injury.68,151 With diffuse cytotoxic edema, such as may result from global hypoxic episodes, the

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

C

B

A

449

FIG 13-35  Extraaxial collection is high signal on T2/FLAIR (A) and high signal in sagittal T1-weighted image (B) of a 43-year-old after a fall. The high T1 signal differentiates this subacute hematoma from a subdural hygroma, which would have similar signal intensity to CSF on T1. C, On ADC image, right uncal herniation from SDH and associated cerebral swelling resulted in extrinsic compression of the right PCA and right PCA-territory infarct.

cerebellum and brainstem are usually spared,73 which can result in a “big black brain” appearance, “white cerebellum sign,” or “pseudoSAH”10 on CT, where the cerebellum, dura, and vasculature are brought into relief by the low attenuation of the brain (see Fig. 13-32).

SDH, EDH, and SAH In the course of evolving TBI, SDHs may rebleed. It may be difficult to appreciate this in a heterogeneous subacute SDH, but when new heterogeneity is seen within a previously homogeneous SDH, this is relatively specific for small amounts of rebleeding. New sedimentation effect is also suggestive of rebleed66,102 (Fig. 13-37). Sometimes a clear interface may be seen along the long axis of a SDH,102 resulting from formed neomembranes in the subacute or chronic setting (see Fig. 13-37). Layered-type SDH has the highest tendency for subsequent rebleed due to hyperfibrinolytic activity.100 At 1 to 4 weeks after bleeding, extraaxial hematomas are generally isodense to brain parenchyma, making them difficult to detect on CT.9,10,69,74,76 It is important to use low-level, wide window settings (sometimes referred to as a subdural window) to maximize the likelihood of detecting subacute SDHs.76 Subacute SDHs are suggested by indirect findings such as sulcal effacement, displacement of the gray matter and gray-white junction away from the inner table, regional sulcal effacement, white matter buckling, and midline shift.73 Capsules and neomembranes that have formed around or within a subacute or chronic SDH may enhance on contrastenhanced CT secondary to neovascularity and disrupted blood-brain barrier10,70 (Fig. 13-38). Subacute SDHs are well appreciated at MRI. As collections liquefy, red cells break down, deoxyhemoglobin oxidizes to methemoglobin, and extraaxial collections become hyperintense on T1-weighted images11,76 (see Fig. 13-35B). In the subacute phase, methemoglobin is primarily responsible for variation in T1 and T2 signal intensity.25 Chronic SDH has attenuation similar to CSF both on CT and conventional MR sequences. A new low-density collection not seen on a recent prior exam is highly suggestive of hygroma (Fig. 13-39). MR signal intensity within SDH depends on the evolution of hemorrhage.10 Vasospasm from SAH is especially common in the subacute phase.70

Traumatic Hydrocephalus In the acute setting, traumatic hydrocephalus can result either from impaired resorption of CSF by the arachnoid villi secondary to

arachnoiditis in communicating hydrocephalus68 or obstruction of the third ventricle, cerebral aqueduct, fourth ventricle, and ventricular outflow tracts by blood or extrinsic compression in noncommunicating hydrocephalus73 (see Fig. 13-25). Hydrocephalus appears acutely as dilated ventricles, with sulcal effacement and periventricular low density from transependymal spread of CSF in severe cases.73 Delayed traumatic hydrocephalus is a common complication of traumatic SAH as arachnoid villi become engorged by phagocytes and fibrose68 (see Fig. 13-39). CT helps in determining the need for an extraventricular drain for therapeutic drainage of CSF.148 Although asymmetric ventricles do not have a high predictive value for elevated ICP, dilation/ entrapment of the contralateral temporal horn can result from compression of ventricular outflow at the foramen of Monro, which correlates highly with increased ICP27 (see Fig. 13-11B). The ipsilateral ventricle is usually compressed.73

Traumatic Brain Herniation, Ischemia, and Infarction Traumatic brain herniation occurs secondary to mass effect produced by primary or secondary injuries, resulting in unmitigated increases in ICP.68,73 Common patterns of brain herniation correlate with patterns of compressive and ischemic injury.70 Multiple patterns often occur simultaneously and have overlapping appearances.63 The most common form is subfalcine herniation, where the cingulate gyrus  is displaced across the midline under the falx cerebri, resulting in midline shift at CT and MRI68,73 (see Figs. 13-4 and 13-5A). This causes the anterior cerebral artery (ACA) and callosomarginal branches to become trapped, resulting in ischemia or infarction involving the ipsilateral paracentral region and superior frontal gyrus.10,73,114 Uncal herniation, where the uncus of the hippocampus displaces through the tentorial incisura, can be unilateral or bilateral and results in effacement of the suprasellar cistern68,73 (see Figs. 13-8B and C and 13-9A). This can compress the anterior choroidal artery, leading to infarction of the posterior limb of the internal capsule, as well as peduncular and posterior cerebral artery (PCA) territory infarction68,93 (see Fig. 13-35C). Uncal herniation also results in compression of the ipsilateral oculomotor nerve, causing blown pupil.73 In descendent transtentorial herniation, the medial temporal lobes herniate through the tentorial incisura10,48 (see Fig. 13-8B and C). Upward herniation can result from lesions in the posterior fossa, with herniation of portions of the cerebellar vermis or hemispheres.73 Transtentorial herniation causes effacement of the quadrigeminal and

http://radiologyebook.com

450

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 13-36  After a motor vehicle accident, CT showing basal cisternal effacement in this 19-year-old with SDH and SAH (A), together with elevated pressure detected using an extraventricular drain (B), was used to determine the need for wide frontotemporoparietal craniotomy (C). There is mild herniation of brain contents through the right-sided defect in C. Some of the high-density material seen overlying the herniated brain contents is from duraplasty (i.e., dural repair [arrows]). D, Once swelling subsided, a cranioplasty was performed. There has been interval bone absorption of the cranioplasty flap, which can occur with either cryo- or subcutaneous preservation. In this case the flap was cryopreserved.

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

451

CHRONIC SEQUELAE OF TBI Moderate to Severe TBI Approximately 40% of patients with TBI discharged from acute hospitalizations develop TBI-related long-term disability.21,70 Encephalomalacia is a common chronic sequela of parenchymal injury. Small areas of encephalomalacia are often clinically asymptomatic but can serve as seizure foci.73 Encephalomalacia is detected at CT as well-defined areas of simple fluid density with volume loss.10,73 On MRI, encephalomalacic areas follow signal intensity of CSF on T1- and T2-weighted imaging. Areas of gliosis at the parenchymal margins are high signal on FLAIR. Encephalomalacia becomes  completely cystic 4 to 6 months after trauma10 (see Fig. 13-39B). Other late sequelae of trauma include hydrocephalus, CSF leak with itinerant pneumocephalus, and generalized atrophy.10 In cases of DAI, involved portions of the corpus callosum become shrunken and cystic.2,105

Mild TBI

FIG 13-37  In this 82-year-old man who hit the back of his head after falling out of bed, there is dependent sedimentation effect as well as a clear interface spanning the long axis of the left SDH, separating acute from chronic blood. Both areas of high density represent rebleed. Layered-type SDHs have a high propensity toward rebleeding.

other basal cisterns.68 Infarct can occur in the bilateral occipital lobes or brainstem as the PCAs or distal basilar artery become trapped.70,93 In tonsilar herniation, the cerebellar tonsils herniate into the foramen magnum.61,68,73 Bulk hemispheric mass effect can cause MCA territory infarct. Vasospasm alone can also result in territorial ischemia and infarction.93 Peripheral venous infarcts can be caused by extraaxial hematomas that exert mass effect on the adjacent cortex, compressing cortical veins.68

Postoperative Appearances and Complications Wide craniectomies are performed to decrease ICP by allowing  the brain to herniate through the defect, effectively increasing intracranial volume148 (see Fig. 13-31). This often results in rapid radiologic improvement in terms of midline shift and cisternal effacement.68 In diffuse swelling, bilateral decompressive craniectomies are sometimes performed.27,109 Cranioplasty can be performed immediately after evacuation of hematomas (see Fig. 13-6), or the bone flap can be temporarily removed when hemispheric swelling prevents replacement27 (see Fig. 13-36C and D). Duraplasty can be performed after durotomy, and thin extraaxial high density subjacent to the cranioplasty or craniectomy site may represent synthetic dural graft.148 CT and MRI can detect postoperative complications including  subdural empyema, abscess, brainstem hemorrhage, edema, tension pneumocephalus, and intracerebral hemorrhage76 (Fig. 13-40). Complications of ventriculostomy can include hematoma, meningitis, and ventriculitis.148

mTBI is especially difficult to diagnose but can result in persistent cognitive deficits in a minority of patients.28,72 Advanced MRI techniques have the potential to help distinguish between organic  and nonorganic causes of insidious neuropsychiatric symptoms  after mTBI.8 Interest in prognosticating outcomes in cases of mTBI have increased with the recent attention to the long-term consequences in athletes following repeat concussions or subconcussive episodes.8,12,20,22,49,70,111,130,141

SWI and DWI Cerebral microhemorrhages may be an important marker of longterm complications in mTBI49,111 resulting from low grade DAI. This phenomenon has been termed DAI syndrome.129 DWI and SWI can improve the ability to detect axonal shear injury in mTBI cases, but the clinical relevance of this increased sensitivity in the setting of mTBI is still controversial.127,118,117,143,34,111,39,132 Mean ADC values in the whole brain have been shown to be among the best predictors of outcome among all degrees of TBI.34,111

Magnetic Resonance Spectroscopy Although MRS is not widely used after TBI, particularly in the early period after injury,124 it has been shown to be clinically useful and can depict injury in a brain that otherwise appears normal at imaging.13,15,20,35,56 MRS is potentially widely available, as it can be readily implemented on existing MRI machines.20 Within a selected tissue volume, a spectrum of metabolites, including N-acetylaspartate (NAA), creatine, and choline are examined. Heights of each compound correspond with their relative abundance. The NAA peak occurring  at 2.0 parts per million (ppm) is a measure of neuronal integrity  and mitochondrial function or impairment.70,84 The creatine peak at 3.0 ppm consists of protons involved in the creatine kinase reaction, basic to cellular energy and metabolism.70 At 3.2 ppm, the choline peak consists of protons found in the phospholipid bilayer of cell membranes. Choline is a marker of myelin and cell membrane breakdown.70,111 The creatine peak is relatively stable and acts as a convenient benchmark to gauge the relative abundance of other compounds. Decrease in NAA, indicating axonal or neuronal damage, is seen in the early acute setting after head injury15,20,76,111,115,124, (Fig. 13-41). Lower NAA-to-creatine ratios have been correlated with poorer clinical  outcomes.19,35,59,76,124,128 In repeat concussive episodes, the severity of biochemical abnormality depends on the interval between traumatic events, and NAA measurement by MRS may be valid in assessing metabolic recovery.70

http://radiologyebook.com

452

PART II  CT and MR Imaging of the Whole Body

A

B FIG 13-38  A 63-year-old with acute-on-chronic right SDH after a fall. A, The chronic portion of the hematoma is low density on the noncontrast CT image. Several high-density strands, in keeping with neomembranes, are seen spanning the SDH. An area of high density represents rebleed, which is contained by the formed membranes (arrow). On the left there is a subacute, nearly isodense subdural hematoma. Notice the subtle inward buckling of the cortex. B, Contrast-enhanced CT shows that the formed fibrous membranes traversing the right chronic SDH enhance (arrow). Both the right chronic and left subacute subdural have enhancing capsules (arrows).

A

C

B

FIG 13-39  A, In this 36-year-old bicyclist struck by a car (same patient as Figs. 13-9 and 13-31) a left hygroma had developed, which was not seen on a recent prior. Hygromas result from arachnoid tears, which allow CSF to enter the subdural space. A single burr hole evacuation was subsequently performed, and the hygroma resolved with drainage. B and C, Brain swelling gradually resolved, and hemorrhagic contusions and right cerebellar infarct were replaced by evolving encephalomalacic changes. Third and fourth ventricles are dilated (B [arrows]) as a result of impaired resorption of CSF at the arachnoid villi. Communicating hydrocephalus is redemonstrated, and there is further evolution of encephalomalacia on follow-up (C).

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

B

A

C

FIG 13-40  A 44-year-old with abscess demonstrated on CT (A [arrow]) and sagittal T1 postgadolinium MRI (B [arrow]) after decompressive bifrontal craniotomy for TBI. Because of concurrent osteomyelitis, the flap was replaced with a synthetic cranioplasty implant (C).

NAA

NAA

Choline Choline

A

Cre

Cre

B FIG 13-41  A, MRS in a healthy control shows normal NAA/Cr ratio in the splenium of the corpus callosum. B, MRS in a 40-year-old male mTBI patient with impaired neurocognitive function following a motor vehicle collision shows a relative decrease in the NAA peak compared with the creatine peak, and decreased NAA/ Cr ratio. (From Bodanapally UK et al: Imaging of traumatic brain injury. Radiol Clin North Am 53:695–715, 2015.)

http://radiologyebook.com

453

454

PART II  CT and MR Imaging of the Whole Body

C

B

A

FIG 13-42  DTI fiber tracking at the corpus callosum from a normal volunteer (A) and a 40-year-old woman with mTBI after a fall and GCS of 15 at presentation (B and C). Fiber tracking at the corpus is shown 10 days (B) and 1 month (C) after injury. B and C show temporary loss of visualization of fiber tracks in the genu/ anterior body junction of the corpus (B [arrow]), with return of normal low radial and high axial diffusivity, resulting in visualization of tracks in this region at 1 month (C [arrow]). The patient had a normal-appearing brain on conventional MRI sequences, yet acutely impaired cognitive scores, with normal neurocognitive function by 1 month. (From Bodanapally UK et al: Imaging of traumatic brain injury. Radiol Clin North Am 53:695–715, 2015.)

C

B

A

FIG 13-43  Abused infant with shaken baby syndrome. Right subdural collection, which contained locules that followed CSF on T1-weighted images and T2/FLAIR, consistent with hygroma or chronic subdural hematoma (A and B [solid arrows]), and other areas that were high signal intensity on both T1 (C [solid arrow]) and T2, consistent with subacute SDH. Areas of gliosis (high FLAIR signal) and dystrophic calcification (high T1 signal) are seen in the occipital lobes, more prominently on the right (A and C [open arrows]). The child was found to have retinal hemorrhages on funduscopic exam.

Diffusion Tensor Imaging Unlike scalar measures such as ADC, a tensor provides information about directional diffusivity by derivation of “eigenvectors,” which correspond with the three orthotopic axes of a white matter tract’s direction.5 Unlike conventional DWI, which encodes diffusion in three axes, an effective diffusion tensor is determined with DTI by assessing diffusion in up to 25 to 30 directions.68 The directionality of a diffusion tensor is expressed quantitatively as fractional anisotropy (FA), which reflects the propensity of water molecules within neurons to travel in an axial rather than radial direction.5,6,145,150 Nondamaged regions of the brain are highly organized into coherent fibers (Fig. 13-42A), which yield a high FA due to low radial diffusivity,8,131 whereas FA is reduced when radial diffusivity is increased, such as with traumatic axonal disruption or swollen and leaky axons.8,70,98,145 As such, decreased FA is a biomarker for DAI and other white matter lesions in the acute setting (see Fig. 13-42B). Swollen axons may recover (see Fig. 13-42C), but persistent decreased FA is predictive of long-term impairment.86,89,92,97,145

Over long periods of time ranging from months to years, as axons continue to degrade through wallerian degeneration, FA values can continue to gradually decrease.8 Specific tracts that are variably involved include the genu (see Fig. 13-42) and splenium of the corpus callosum, with the latter involved in more severe injuries.116,145 White matter in the frontal and temporal lobes is also affected.145 FA values can show compensatory gradual increases in the internal capsule and centrum semiovale, which correlates with more favorable outcomes and likely reflects neuroplasticity.125,145 Lower FA values correspond with greater degrees of axonal injury and greater disability.8,68,70,81

CONSIDERATIONS FOR PEDIATRIC TBI The pediatric population is much more susceptible to brain swelling following TBI than adults.68 This has been attributed to disproportionate increases in hyperemia, vasodilation, and increased blood volume. Leptomeningeal cysts are more common in pediatric TBI.73 These

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury result from dural tears that act as one-way valves, causing herniation of arachnoid and pressure remodeling of bone, lytic lesions, or sutural widening.73 The bony manifestations of leptomeningeal cysts are sometimes referred to as growing fractures. Findings specific to child abuse can be missed without an appropriate index of suspicion. The hallmark of abuse is the presence of multiple injuries without adequate explanation. Intracranial injuries among abused infants occur with a high incidence even in the absence of clinical signs. The infant skull is highly elastic,102 and intracranial injury can result without skull fracture. The threshold for head CT in suspected abuse should be very low. Children are prone to SDH from bridging vein disruption43,102 with repetitive shaking (Fig. 13-43). The impulsive forces also cause retinoschisis and retinal hemorrhage by indirect tugging on the retina from to-and-fro motions of the relatively dense lens and resultant ocular fluid pressure waves.42,102 Hadley et al. introduced the term infant whiplash shake injury syndrome47 in 1989 to describe this constellation of findings, later renamed shaken baby syndrome.24 Shaking is also associated with a high frequency of injuries to the spinal cord,47,102 and routine MRI of the cervical spine should be performed in abused children with head injuries.102

REFERENCES 1. Adams J, Doyle D, Ford I, et al: Diffuse axonal injury in head injury: Definition, diagnosis and grading. Histopathology 15(1):49–59, 1989. 2. Adams J, Mitchell D, Graham D, et al: Diffuse brain damage of immediate impact type: Its relationship to ‘primary brain-stem damage’ in head injury. Brain 100(3):489–502, 1977. 3. Ashikaga R, Araki Y, Ishida O: MRI of head injury using FLAIR. Neuroradiology 39(4):239–242, 1997. 4. Bakshi R, Kamran S, Kinkel PR, et al: Fluid-attenuated inversionrecovery MR imaging in acute and subacute cerebral intraventricular hemorrhage. AJNR Am J Neuroradiol 20(4):629–636, 1999. 5. Basser PJ, Mattiello J, LeBihan D: MR diffusion tensor spectroscopy and imaging. Biophys J 66(1):259–267, 1994. 6. Beaulieu C: The basis of anisotropic water diffusion in the nervous system–A technical review. NMR Biomed 15(7–8):435–455, 2002. 7. Beaumont A, Gennarelli T: CT prediction of contusion evolution after closed head injury: The role of pericontusional edema. Acta Neurochir Suppl 96:30–32, 2006. 8. Benson RR, Gattu R, Sewick B, et al: Detection of hemorrhagic and axonal pathology in mild traumatic brain injury using advanced MRI: Implications for neurorehabilitation. Neurorehabilitation 31(3):261–279, 2012. 9. Bergström M, Ericson K, Levander B, et al: Computed tomography of cranial subdural and epidural hematomas: Variation of attenuation related to time and clinical events such as rebleeding. J Comput Assist Tomogr 1(4):449–455, 1977. 10. Bešenski N: Traumatic injuries: Imaging of head injuries. Eur Radiol 12(6):1237–1252, 2002. 11. Bradley WG: MR appearance of hemorrhage in the brain. Radiology 189(1):15–26, 1993. 12. Breedlove EL, Robinson M, Talavage TM, et al: Biomechanical correlates of symptomatic and asymptomatic neurophysiological impairment in high school football. J Biomech 45(7):1265–1272, 2012. 13. Brooks WM, Friedman SD, Gasparovic C: Magnetic resonance spectroscopy in traumatic brain injury. J Head Trauma Rehabil 16(2):149–164, 2001. 14. Bullock R, Chesnut R, Clifton G, et al: Guidelines for the management of severe head injury. Brain Trauma Foundation. Eur J Emerg Med 3(2):109–127, 1996. 15. Carpentier A, Galanaud D, Puybasset L, et al: Early morphologic and spectroscopic magnetic resonance in severe traumatic brain injuries can detect “invisible brain stem damage” and predict “vegetative states”.   J Neurotrauma 23(5):674–685, 2006.

455

16. Chesnut RM: Implications of the guidelines for the management of severe head injury for the practicing neurosurgeon. Surg Neurol 50(3):187–193, 1998. 17. Chesnut R, Marshall L, Klauber M, et al: The role of secondary brain injury in determining outcome from severe head injury. J Trauma 34(2):216–222, 1993. 18. Chew BG, Spearman CM, Quigley MR, et al: The prognostic significance of traumatic brainstem injury detected on T2-weighted MRI. J Neurosurg 117(4):722–728, 2012. 19. Choe B, Suh T, Choi K, et al: Neuronal dysfunction in patients with closed head injury evaluated by in vivo 1H magnetic resonance spectroscopy. Invest Radiol 30(8):502–506, 1995. 20. Coles JP: Imaging after brain injury. Br J Anaesth 99(1):49–60, 2007. 21. Corrigan JD, Selassie AW, Orman JA: The epidemiology of traumatic brain injury. J Head Trauma Rehabil 25(2):72–80, 2010. 22. De Beaumont L, Théoret H, Mongeon D, et al: Brain function decline in healthy retired athletes who sustained their last sports concussion in early adulthood. Brain 132(3):695–708, 2009. 23. Delgado Almandoz JE, Kelly HR, Schaefer PW, et al: Prevalence of traumatic dural venous sinus thrombosis in high-risk acute blunt head trauma patients evaluated with multidetector CT venography. Radiology 255(2):570–577, 2010. 24. Duhaime A-C, Christian CW, Rorke LB, et al: Nonaccidental head injury in infants—the “shaken-baby syndrome”. NEJM 338(25):1822– 1829, 1998. 25. Ebisu T, Naruse S, Horikawa Y, et al: Nonacute subdural hematoma: Fundamental interpretation of MR images based on biochemical and in vitro MR analysis. Radiology 171(2):449–453, 1989. 26. Ebisu T, Naruse S, Horikawa Y, et al: Discrimination between different types of white matter edema with diffusion-weighted MR imaging.   J Magn Reson Imaging 3(6):863–868, 1993. 27. Eisenberg HM, Gary HE, Aldrich EF, et al: Initial CT findings in   753 patients with severe head injury. J Neurosurg 73(5):688–698, 1990. 28. Englander J, Hall K, Stimpson T, et al: Mild traumatic brain injury in an insured population: Subjective complaints and return to employment. Brain Inj 6(2):161–166, 1992. 29. Fainardi E, Chieregato A, Antonelli V, et al: Time course of CT evolution in traumatic subarachnoid haemorrhage: A study of 141 patients. Acta Neurochir (Wien) 146(3):257–263, 2004. 30. Feldmann E, Gandy S, Becker R, et al: MRI demonstrates descending transtentorial herniation. Neurology 38(5):697–701, 1988. 31. Feuerman T, Wackym P, Gade G, et al: Value of skull radiography, head computed tomographic scanning, and admission for observation in cases of minor head injury. Neurosurgery 22(3):449–453, 1988. 32. Finfer SR, Cohen J: Severe traumatic brain injury. Resuscitation 48(1):77–90, 2001. 33. Firsching R, Woischneck D, Diedrich M, et al: Early magnetic resonance imaging of brainstem lesions after severe head injury. J Neurosurg 89(5):707–712, 1998. 34. Galloway NR, Tong KA, Ashwal S, et al: Diffusion-weighted imaging improves outcome prediction in pediatric traumatic brain injury.   J Neurotrauma 25(10):1153–1162, 2008. 35. Garnett MR, Cadoux-Hudson TAD, Styles P: How useful is magnetic resonance imaging in predicting severity and outcome in traumatic brain injury? Curr Opin Neurol 14(6):753–757, 2001. 36. Gennarelli T: Mechanisms of brain injury. J Emerg Med 11(Suppl 1):5–11, 1993. 37. Gentry LR, Godersky JC, Thompson B: MR imaging of head trauma: Review of the distribution and radiopathologic features of traumatic lesions. AJR Am J Roentgenol 150(3):663–672, 1988. 38. Gentry LR, Godersky JC, Thompson B, et al: Prospective comparative study of intermediate-field MR and CT in the evaluation of closed head trauma. AJR Am J Roentgenol 150(3):673–682, 1988. 39. Geurts BHJ, Andriessen TMJC, Goraj BM, et al: The reliability of magnetic resonance imaging in traumatic brain injury lesion detection. Brain Injury. 26(12):1439–1450, 2012.

http://radiologyebook.com

456

PART II  CT and MR Imaging of the Whole Body

40. Ghajar J, Hariri RJ, Narayan RK, et al: Survey of critical care management of comatose, head-injured patients in the United States. Crit Care Med 23(3):560–567, 1995. 41. Greenberg J, Cohen WA, Cooper PR: The “hyperacute” extraaxial intracranial hematoma: Computed tomographic findings and clinical significance. Neurosurgery 17(1):48–56, 1985. 42. Greenwald M, Weiss A, Oesterle C, et al: Traumatic retinoschisis in battered babies. Ophthalmology 93(5):618–625, 1986. 43. Guthkelch A: Infantile subdural haematoma and its relationship to whiplash injuries. Br Med J 2(5759):430–431, 1971. 44. Gutman MB, Moulton RJ, Sullivan I, et al: Risk factors predicting operable intracranial hematomas in head injury. J Neurosurg 77(1):9– 14, 1992. 45. Guyot L, Kazmierczak C, Diaz F: Vascular injury in neurotrauma. Neurol Res 23(2–3):291–296, 2001. 46. Haacke EM, Mittal S, Wu Z, et al: Susceptibility-weighted imaging: Technical aspects and clinical applications, Part 1. AJNR Am J Neuroradiol 30(1):19–30, 2009. 47. Hadley M, Sonntag V, Rekate H, et al: The infant whiplash-shake injury syndrome: A clinical and pathological study. Neurosurgery 24(4),1989. 48. Hahn F, Gurney J: CT signs of central descending transtentorial herniation. AJNR Am J Neuroradiol 6(5):844–845, 1985. 49. Hähnel S, Stippich C, Weber I, et al: Prevalence of cerebral microhemorrhages in amateur boxers as detected by 3T MR imaging. AJNR Am J Neuroradiol 29(2):388–391, 2008. 50. Hammoud D, Wasserman B: Diffuse axonal injuries: Pathophysiology and imaging. Neuroimaging Clin N Am 12(2):205–216, 2002. 51. Hanstock C, Faden A, Bendall M, et al: Diffusion-weighted imaging differentiates ischemic tissue from traumatized tissue. Stroke 25(4):843– 848, 1994. 52. Hashimoto T, Nakamura N, Richard K, et al: Primary brain stem lesions caused by closed head injuries. Neurosurg Rev 16(4):291–298, 1993. 53. Haydel MJ, Preston CA, Mills TJ, et al: Indications for computed tomography in patients with minor head injury. NEJM 343(2):100–105, 2000. 54. Hesselink JR, Dowd CF, Healy ME, et al: MR imaging of brain contusions: A comparative study with CT. AJR Am J Roentgenol 150(5):1133–1142, 1988. 55. Holm L, Cassidy J, Carroll L, et al: Summary of the WHO Collaborating Centre for Neurotrauma Task Force on Mild Traumatic Brain Injury. J Rehabil Med 37(3):137–141, 2005. 56. Holshouser BA, Tong KA, Ashwal S, et al: Spectroscopic imaging depicts diffuse axonal injury in children with traumatic brain injury. AJNR Am J Neuroradiol 26(5):1276–1285, 2005. 57. Hossmann K-A, Fischer M, Bockhorst K, et al: NMR imaging of the apparent diffusion coefficient (ADC) for the evaluation of metabolic suppression and recovery after prolonged cerebral ischemia. J Cereb Blood Flow Metab 14(5):723–731, 1994. 58. Huisman T, Gregory S, Hergan K, et al: Diffusion-weighted imaging for the evaluation of diffuse axonal injury in closed head injury. J Comput Assist Tomogr 27(1):5–11, 2003. 59. Hunter JV, Thornton RJ, Wang ZJ, et al: Late proton MR spectroscopy in children after traumatic brain injury: Correlation with cognitive outcomes. AJNR Am J Neuroradiol 26(3):482–488, 2005. 60. Ingebrigtsen T, Romner B, Kock-Jensen C: Scandinavian guidelines for initial management of minimal, mild, and moderate head injuries.   J Trauma 48(4):760–766, 2000. 61. Ishikawa M, Kikuchi H, Fujisawa I, et al: Tonsillar herniation on magnetic resonance imaging. Neurosurgery 22(1 Pt 1):77–81, 1988. 62. Jeret JS, Mandell M, Anziska B, et al: Clinical predictors of abnormality disclosed by computed tomography after mild head trauma. Neurosurgery 32(1):9–16, 1993. 63. Johnson P, Eckard D, Chason D, et al: Imaging of acquired cerebral herniations. Neuroimaging Clin N Am 12(2):217–228, 2002. 64. Jones TR, Kaplan RT, Lane B, et al: Single– versus multi–detector row CT of the brain: Quality assessment. Radiology 219(3):750–755, 2001.

65. Juul N, Duch B, Rasmussen M: Clinical management of patients with head injury. Curr Anaesth Crit Care 20(3):132–137, 2009. 66. Kao M-C: Sedimentation level in chronic subdural hematoma visible on computerized tomography. J Neurosurg 58(2):246–251, 1983. 67. Kelly A, Zimmerman R, Snow R, et al: Head trauma: Comparison of MR and CT—Experience in 100 patients. AJNR Am J Neuroradiol 9(4):699–708, 1988. 68. Kim J, Gean A: Imaging for the diagnosis and management of traumatic brain injury. Neurother 8(1):39–53, 2011. 69. Kostanian V, Choi J, Liker M, et al: Computed tomographic characteristics of chronic subdural hematomas. Neurosurg Clin N Am 11(3):479–489, 2000. 70. Kubal WS: Updated imaging of traumatic brain injury. Radiol Clin North Am 50(1):15–41, 2012. 71. Kwee RM, Kwee TC: Virchow-Robin spaces at MR imaging. Radiographics 27(4):1071–1086, 2007. 72. Langlois JA, Rutland-Brown W, Wald MM: The epidemiology and impact of traumatic brain injury: A brief overview. J Head Trauma Rehabil 21(5):375–378, 2006. 73. Le TH, Gean AD: Neuroimaging of traumatic brain injury. Mt Sinai J Med 76(2):145–162, 2009. 74. Lee KS, Bae WK, Bae HG, et al: The computed tomographic attenuation and the age of subdural hematomas. J Korean Med Sci 12(4):353–359, 1997. 75. Lee E-J, Hung Y-C, Wang L-C, et al: Factors influencing the functional outcome of patients with acute epidural hematomas: Analysis of 200 patients undergoing surgery. J Trauma 45(5):946–952, 1998. 76. Lee B, Newberg A: Neuroimaging in traumatic brain imaging. NeuroRx 2(2):372–383, 2005. 77. Lee H, Wintermark M, Gean AD, et al: Focal lesions in acute mild traumatic brain injury and neurocognitive outcome: CT versus 3T MRI. J Neurotrauma 25(9):1049–1056, 2008. 78. Lloyd M, Kimber P, Burrows E: Post-traumatic cerebrospinal fluid rhinorrhoea: Modern high-definition computed tomography is all   that is required for the effective demonstration of the site of leakage. Clin Radiol 49(2):100–103, 1994. 79. Lobato RD, Sarabia R, Cordobes F, et al: Posttraumatic cerebral hemispheric swelling. J Neurosurg 68(3):417–423, 1988. 80. Maas A, Hukkelhoven C, Marshall L, et al: Prediction of outcome in traumatic brain injury with computed tomographic characteristics:   A comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery 57(6):1173–1182, 2005. 81. Maas AIR, Stocchetti N, Bullock R: Moderate and severe traumatic brain injury in adults. Lancet Neurol 7(8):728–741, 2008. 82. Macpherson B, MacPherson P, Jennett B: CT evidence of intracranial contusion and haematoma in relation to the presence, site and type of skull fracture. Clin Radiol 42(5):321–326, 1990. 83. Mannion RJ, Cross J, Bradley P, et al: Mechanism-based MRI classification of traumatic brainstem injury and its relationship to outcome. J Neurotrauma 24(1):128–135, 2007. 84. Marmarou A, Signoretti S, Fatouros P, et al: Mitochondrial injury measured by proton magnetic resonance spectroscopy in severe head trauma patients. Acta Neurochir Suppl 95:149–151, 2005. 85. Marmarou A, Signoretti S, Fatouros PP, et al: Predominance of cellular edema in traumatic brain swelling in patients with severe head injuries.   J Neurosurg 104(5):720–730, 2006. 86. Marquez de la Plata CD, Yang FG, Wang JY, et al: Diffusion tensor imaging biomarkers for traumatic axonal injury: Analysis of three analytic methods. J Int Neuropsychol Soc 17(01):24–35, 2011. 87. Marshall LF, Marshall SB, Klauber MR, et al: A new classification of head injury based on computerized tomography. Spec Suppl 75(1s):S14–S20, 1991. 88. Marshall L, Marshall S, Klauber M, et al: The diagnosis of head injury requires a classification based on computed axial tomography.   J Neurotrauma 9(Suppl 1):S287–S292, 1992. 89. Matsushita M, Hosoda K, Naitoh Y, et al: Utility of diffusion tensor imaging in the acute stage of mild to moderate traumatic brain injury

http://radiologyebook.com

CHAPTER 13  Traumatic Brain Injury

90. 91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106. 107.

108.

109.

110.

111. 112.

for detecting white matter lesions and predicting long-term cognitive function in adults. J Neurosurg 115(1):130–139, 2011. Maxwell W: Histopathological changes at central nodes of Ranvier after stretch-injury. Microsc Res Tech 34(6):522–535, 1996. Meythaler JM, Peduzzi JD, Eleftheriou E, et al: Current concepts: Diffuse axonal injury–associated traumatic brain injury. Arch Phys Med Rehabil 82(10):1461–1471, 2001. Miles L, Grossman RI, Johnson G, et al: Short-term DTI predictors of cognitive dysfunction in mild traumatic brain injury. Brain Inj 22(2):115–122, 2008. Mirvis S, Wolf A, Numaguchi Y, et al: Posttraumatic cerebral infarction diagnosed by CT: Prevalence, origin, and outcome. AJR Am J Roentgenol 154(6):1293–1298, 1990. Mittl R, Grossman R, Hiehle J, et al: Prevalence of MR evidence   of diffuse axonal injury in patients with mild head injury and normal head CT findings. AJNR Am J Neuroradiol 15(8):1583–1589, 1994. Narayan RK, Greenberg RP, Miller JD, et al: Improved confidence of outcome prediction in severe head injury. J Neurosurg 54(6):751–762, 1981. Narayan RK, Maas AIR, Servadei F, et al: Progression of traumatic intracerebral hemorrhage: A prospective observational study.   J Neurotrauma 25(6):629–639, 2008. Newcombe V, Chatfield D, Outtrim J, et al: Mapping traumatic axonal injury using diffusion tensor imaging: Correlations with functional outcome. PLoS ONE 6(5):e19214, 2011. Newcombe VFJ, Williams GB, Nortje J, et al: Analysis of acute traumatic axonal injury using diffusion tensor imaging. Br J Neurosurg 21(4):340–348, 2007. Noguchi K, Ogawa T, Seto H, et al: Subacute and chronic subarachnoid hemorrhage: Diagnosis with fluid-attenuated inversion-recovery MR imaging. Radiology 203(1):257–262, 1997. Nomura S, Kashiwagi S, Fujisawa H, et al: Characterization of local hyperfibrinolysis in chronic subdural hematomas by SDS-PAGE and immunoblot. J Neurosurg 81(6):910–913, 1994. Ommaya A: Computerized axial tomography of the head: The EMI-scanner, a new device for direct examination of the brain “in vivo”. Special article. Surg Neurol 1:217–222, 1973. Ommaya AK, Goldsmith W, Thibault L: Biomechanics and neuropathology of adult and paediatric head injury. Br J Neurosurg 16(3):220–242, 2002. Orrison W, Gentry L, Stimac G, et al: Blinded comparison of cranial CT and MR in closed head injury evaluation. AJNR Am J Neuroradiol 15(2):351–356, 1994. Ozsvath RR, Casey SO, Lustrin ES, et al: Cerebral venography: Comparison of CT and MR projection venography. AJR Am J Roentgenol 169(6):1699–1707, 1997. Parizel PM, Özsarlak Ö, Van Goethem JW, et al: Imaging findings in diffuse axonal injury after closed head trauma. Eur Radiol 8(6):960– 965, 1998. Pasqualin A, Vivenza C, Rosta L, et al: Cerebral vasospasm after head injury. Neurosurgery 15(6):855–858, 1984. Paterakis K, Karantanas AH, Komnos A, et al: Outcome of patients with diffuse axonal injury: The significance and prognostic value of MRI in the acute phase. J Trauma 49(6):1071–1075, 2000. Peng R, Chang C, Gilmore D, et al: Epidemiology of immediate and early trauma deaths at an urban level I trauma center. Am Surg 64(10):950–954, 1998. Polin RS, Shaffrey ME, Bogaev CA, et al: Decompressive bifrontal craniectomy in the treatment of severe refractory posttraumatic cerebral edema. Neurosurgery 41(1):84–94, 1997. Povlishock J, Jenkins L: Are the pathobiological changes evoked by traumatic brain injury immediate and irreversible? Brain Pathol 5(4):415–426, 1995. Provenzale JM: Imaging of traumatic brain injury: A review of the recent medical literature. AJR Am J Roentgenol 194(1):16–19, 2010. Reichenbach J, Venkatesan R, Yablonskiy D, et al: Theory and application of static field inhomogeneity effects in gradient-echo imaging. J Magn Reson Imaging 7(2):266–279, 1997.

457

113. Rivas J, Lobato R, Sarabia R, et al: Extradural hematoma: Analysis of factors influencing the courses of 161 patients. Neurosurgery 23(1):44– 51, 1988. 114. Rothfus W, Goldberg A, Tabas J, et al: Callosomarginal infarction secondary to transfalcial herniation. AJNR Am J Neuroradiol 8(6):1073– 1076, 1987. 115. Rubin Y, Cecil K, Wehrli S, et al: High-resolution 1H NMR spectroscopy following experimental brain trauma. J Neurotrauma 14(7):441–449, 1997. 116. Rutgers DR, Fillard P, Paradot G, et al: Diffusion tensor imaging characteristics of the corpus callosum in mild, moderate, and severe traumatic brain injury. AJNR Am J Neuroradiol 29(9):1730–1735, 2008. 117. Schaefer PW, Huisman TAGM, Sorensen AG, et al: Diffusion-weighted MR imaging in closed head injury: High correlation with initial Glasgow Coma Scale score and score on Modified Rankin Scale at discharge. Radiology 233(1):58–66, 2004. 118. Scheid R, Preul C, Gruber O, et al: Diffuse axonal injury associated with chronic traumatic brain injury: Evidence from T2*-weighted gradientecho imaging at 3 T. AJNR Am J Neuroradiol 24(6):1049–1056, 2003. 119. Seelig JM, Becker DP, Miller JD, et al: Traumatic acute subdural hematoma: Major mortality reduction in comatose patients treated within four hours. NEJM 304(25):1511–1518, 1981. 120. Sehgal V, Delproposto Z, Haacke EM, et al: Clinical applications of neuroimaging with susceptibility-weighted imaging. J Magn Reson Imaging 22(4):439–450, 2005. 121. Servadei F, Faccani G, Roccella P, et al: Asymptomatic extradural haematomas. Results of a multicenter study of 158 cases in minor head injury. Acta Neurochir (Wien) 96(1–2):39–45, 1989. 122. Servadei F, Murray GD, Teasdale GM, et al: Traumatic subarachnoid hemorrhage: Demographic and clinical study of 750 patients from the European Brain Injury Consortium survey of head injuries. Neurosurgery 50(2):261–269, 2002. 123. Shackford S, Wald S, Ross S, et al: The clinical utility of computed tomographic scanning and neurologic examination in the management of patients with minor head injuries. J Trauma 33(3):385–394, 1992. 124. Shutter L, Tong KA, Lee A, et al: Prognostic role of proton magnetic resonance spectroscopy in acute traumatic brain injury. J Head Trauma Rehabil 21(4):334–349, 2006. 125. Sidaros A, Engberg AW, Sidaros K, et al: Diffusion tensor imaging during recovery from severe traumatic brain injury and relation to clinical outcome: A longitudinal study, 2008. 126. Sifri ZC, Livingston DH, Lavery RF, et al: Value of repeat cranial computed axial tomography scanning in patients with minimal head injury. Am J Surg 187(3):338–342, 2004. 127. Sigmund GA, Tong KA, Nickerson JP, et al: Multimodality comparison of neuroimaging in pediatric traumatic brain injury. Pediatr Neurol 36(4):217–226, 2007. 128. Sinson G, Bagley LJ, Cecil KM, et al: Magnetization transfer imaging and proton MR spectroscopy in the evaluation of axonal injury: Correlation with clinical outcome after traumatic brain injury. AJNR Am J Neuroradiol 22(1):143–151, 2001. 129. Smith DH, Meaney DF, Shull WH: Diffuse axonal injury in head trauma. J Head Trauma Rehabil 18(4):307–316, 2003. 130. Solomon GS, Haase RF: Biopsychosocial characteristics and neurocognitive test performance in National Football League   players: An initial assessment. Arch Clin Neuropsychol 23(5):563–577, 2008. 131. Song S-K, Sun S-W, Ramsbottom MJ, et al: Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage 17(3):1429–1436, 2002. 132. Spitz G, Maller JJ, Ng A, et al: Detecting lesions after traumatic brain injury using susceptibility weighted imaging: A comparison with fluid-attenuated inversion recovery and correlation with clinical outcome. J Neurotrauma 30(24):2038–2050, 2013. 133. Stein SC, Chen X-H, Sinson GP, et al: Intravascular coagulation: A major secondary insult in nonfatal traumatic brain injury. J Neurosurg 97(6):1373–1377, 2002.

http://radiologyebook.com

458

PART II  CT and MR Imaging of the Whole Body

134. Stein SC, O’Malley KF, Ross SE: Is routine computed tomography scanning too expensive for mild head injury? Ann Emerg Med 20(12):1286–1289, 1991. 135. Stein S, Ross S: The value of computed tomographic scans in patients with low-risk head injuries. Neurosurgery 26(4):638–640, 1990. 136. Stiell IG, Lesiuk H, Wells GA, et al: The Canadian CT Head Rule Study for patients with minor head injury: Rationale, objectives, and methodology for phase I (derivation). Ann Emerg Med 38(2):160–169, 2001. 137. Stiell IG, Lesiuk H, Wells GA, et al: Canadian CT Head Rule Study for patients with minor head injury: Methodology for phase II (validation and economic analysis). Ann Emerg Med 38(3):317–322, 2001. 138. Stiell IG, Wells GA, Vandemheen KL, et al: The Canadian C-Spine Rule for radiography in alert and stable trauma patients. JAMA 286(15):1841–1848, 2001. 139. Stone JA, Castillo M, Neelon B, et al: Evaluation of CSF leaks: High-resolution CT compared with contrast-enhanced CT and radionuclide cisternography. AJNR Am J Neuroradiol 20(4):706–712, 1999. 140. Sullivan TP, Jarvik JG, Cohen WA: Follow-up of conservatively managed epidural hematomas: Implications for timing of repeat CT. AJNR Am J Neuroradiol 20(1):107–113, 1999. 141. Talavage TM, Nauman EA, Breedlove EL, et al: Functionally-detected cognitive impairment in high school football players without clinically-diagnosed concussion. J Neurotrauma 31(4):327–338, 2010. 142. Tong KA, Ashwal S, Holshouser BA, et al: Hemorrhagic shearing lesions in children and adolescents with posttraumatic diffuse axonal injury: Improved detection and initial results. Radiology 227(2):332–339, 2003.

143. Tong KA, Ashwal S, Holshouser BA, et al: Diffuse axonal injury in children: Clinical correlation with hemorrhagic lesions. Ann Neurol 56(1):36–50, 2004. 144. Tsushima Y, Endo K: Hypointensities in the Brain on T2*-weighted gradient echo magnetic resonance imaging. Curr Probl Diagn Radiol 35(4):140–150, 2006. 145. Voelbel GT, Genova HM, Chiaravalotti ND, et al: Diffusion tensor imaging of traumatic brain injury review: Implications for neurorehabilitation. Neurorehabilitation 31(3):281–293, 2012. 146. Wellwood J, Alcantara A, Michael D: Neurotrauma: The role of CT angiogram. Neurol Res 24(Suppl 1):S13–S16, 2002. 147. Wilms G, Marchal G, Geusens E, et al: Isodense subdural haematomas on CT: MRI findings. Neuroradiology 34(6):497–499, 1992. 148. Winter CD, Adamides AA, Lewis PM, et al: A review of the current management of severe traumatic brain injury. Surgeon 3(5):329–337, 2005. 149. Wintermark M, van Melle G, Schnyder P, et al: Admission perfusion CT: Prognostic value in patients with severe head trauma. Radiology 232(1):211–220, 2004. 150. Xu J, Rasmussen I-A, Lagopoulos J, et al: Diffuse axonal injury in severe traumatic brain injury visualized using high-resolution diffusion tensor imaging. J Neurotrauma 24(5):753–765, 2007. 151. Yamaki T, Hirakawa K, Ueguchi T, et al: Chronological evaluation of acute traumatic intracerebral haematoma. Acta Neurochir (Wien) 103(3–4):112–115, 1990. 152. Zimmerman RA, Bilaniuk LT: Computed tomographic staging of traumatic epidural bleeding. Radiology 144(4):809–812, 1982. 153. Zumkeller M, Behrmann R, Heissler HE, et al: Computed tomographic criteria and survival rate for patients with acute subdural hematoma. Neurosurgery 39(4):708–713, 1996.

http://radiologyebook.com

14  Spinal Cord Injury David Dreizin and Daniel Mascarenhas

INTRODUCTION More than 10,000 spinal cord injuries are diagnosed each year in the United States, with over half involving the cervical spinal cord.2,7,69 Multidetector computed tomography (CT) is recommended by the American College of Radiology as the first-line screening exam for any patient for whom spine injury cannot be excluded on clinical grounds alone.24 Coronal and sagittal spine reformats are also commonly incorporated into whole-body CT protocols for severe blunt trauma.15,14 CT is more cost-effective for initial cervical spine screening and appears to have a greater risk-benefit profile than plain radiography despite an order of magnitude greater radiation dose.14,63 This is largely due to the rate of missed injury on plain films, which can be higher than 50%.45 The use of magnetic resonance imaging (MRI) for spine clearance is controversial because CT is exquisitely sensitive for detection of unstable fracture patterns, and isolated signal abnormalities at MRI when bony relationships are normal are of questionable clinical significance.30 On the other hand, CT has little utility for directly evaluating the spinal cord and identifying extraaxial soft tissue pathology, owing to beam hardening from adjacent bony structures and poor contrast resolution.50 This chapter focuses primarily on issues pertaining to MRI of spinal cord injury (SCI). MRI is frequently used after trauma to evaluate patients with persistent neck tenderness (not explainable by CT), myelopathy, absence of clinical improvement, and for suspected epidural hematomas or traumatic disk herniations that compress the spinal cord. MRI is  also useful for prognosticating severity of SCI, determining surgical approach, confirming sufficient decompression of the spinal cord after surgery, evaluating progressive syrinx formation or other complications if recovery reverses in the chronic stage after injury, and guiding the duration of bracing in conservatively managed cases.6,16,35,50 Despite its widespread use, evidence supporting the use of MRI in the early acute setting for management purposes is not robust.6 The prognostic value of MRI for short- and long-term outcomes after SCI is relatively well established.21,22,51,55,59

SPINAL CORD INJURY MECHANISMS Most fractures in both adult and pediatric patients with spinal cord injuries occur in the highly mobile cervical spine. These usually result from impulsive loading (i.e., whiplash).14 In adults the fulcrum of the cervical spine is at C5-C6, and unstable hyperflexion injuries tend  to cluster around this motion segment (Fig. 14-1). Hyperextension injuries have a higher center of rotation, often resulting from deceleration motor vehicle collisions with facial impact or in the elderly with face-planting after trivial falls. In up to one third of patients,11,71 SCI occurs without skeletal or ligamentous disruption appreciable on CT. This has come to be called

by the terms spinal cord injury without radiographic abnormality (SCIWORA) in pediatric patients47 and spinal cord injury without radiographic evidence of trauma (SCIWORET) in adults—in the latter to account for nontraumatic abnormalities usually related to spondylosis.4,61 In adults, SCI in this setting commonly manifests as central cord syndrome, in which the upper extremities are disproportionately affected (Fig. 14-2). Young children have highly elastic spinal columns that can recoil back to normal alignment and exhibit no evidence of bone or discoligamentous injury despite severe SCIs, including transection. Some athletes involved in contact sports may be predisposed to “stinger syndrome” and “spinal cord concussion,” in which myelopathic symptoms are typically minor and resolve after short periods.32,53,62,64,72 Patients with congenitally narrowed spinal canals appear to be predisposed to SCI, but the positive predictive value of various measurement parameters of spinal stenosis appears to be limited.29,49,65 Extraneural findings that can benefit from decompression (e.g., traumatic disk herniations, epidural hematomas) are more common in older patients, especially in the setting of fused spine from ankylosing spondylitis, diffuse idiopathic skeletal hyperostosis, chronic spondylotic changes, or prior spinal surgery16,71 (Fig. 14-3). In those older than 60 years, profound residual deficits, morbidity, and high mortality are common after SCI.13,61

MRI PROTOCOL CONSIDERATIONS 1.5-tesla (T) and 3-T protocols vary considerably from institution to institution. Sagittal T2-weighted imaging or short tau inversion recovery (STIR) images are acquired with near ubiquity and are by far the most important from a prognostic standpoint because of their ability to visualize both the presence of spinal cord hemorrhage and the  craniocaudal extent of edema.6 Sagittal proton density images can very clearly depict disruptions in the continuity of discoligamentous structures, but the added value of this sequence to T2-weighted imaging and STIR, which are also used for this purpose, remains unproven. Findings of ligamentous injury in the thoracolumbar  and cervical spine at STIR have been shown to correlate well with findings at surgery,26,37 and STIR has the added benefit of identifying marrow edema and bone bruising due to fat suppression50,68 (Fig. 14-4). Signal patterns useful for prognostication have not been established with axial images.6,16 Axial gradient echo (GRE) images are highly sensitive to susceptibility artifact from blood product and can depict areas of spinal cord hemorrhage with greater conspicuity than T2 (see Fig. 14-1B). Nevertheless, the added prognostic value of GRE remains unknown, and GRE is incorporated in only about one fourth of protocols in published studies.6 Axial T2-weighted images have not been shown to have prognostic value but can identify clinically relevant lesions such as disk herniation and spinal cord compression that may

http://radiologyebook.com

459

460

PART II  CT and MR Imaging of the Whole Body

*

A

C

B

D FIG 14-1  This 30-year-old man sustained a teardrop fracture at C5-C6 after a fall and presented with complete motor and sensory loss. A, Sagittal T2-weighted imaging shows high signal, in keeping with cord edema from C3-C7 (cranial and caudal extent demarcated by long arrows). Foci of low signal intensity at C5 and C6 are consistent with intramedullary hemorrhage (open arrow). The anterior longitudinal ligament is stripped off the vertebral body of C6 (short arrow), which is retropulsed into the canal, compressing the spinal cord. There is extensive perivertebral hyperintensity, consistent with hemorrhage and edema (*). B, On axial GRE there is low signal intensity and blooming artifact at the level of hemorrhage (open arrows). C, On axial SWI there is low signal involving a greater proportion of the cord surface area, but the sequence is limited by motion. D, The full craniocaudal extent of cord hemorrhage (arrows) is well visualized on sagittal SWI.

http://radiologyebook.com

CHAPTER 14  Spinal Cord Injury

FIG 14-2  This 77-year-old man presented with central cord syndrome after a fall. Sagittal STIR image shows injured disk and ruptured anterior longitudinal ligament, resulting in anteriorly divergent widening at C4-C5 (short arrow). This abnormality was not appreciated at CT because of the extensive underlying degenerative changes. There is multilevel cord edema spanning C3-C6 (demarcated by long arrows). The patient underwent decompression and instrumented fusion.

alter management6 (see Fig. 14-3D). Rapid MRI sequences such as fast imaging employing steady-state acquisition (FIESTA) have gradient echo properties but do not accentuate blood breakdown products, owing to the use of balanced steady-state pulse sequences. These may decrease motion artifact but add little to conventional T2 sequences. The use of advanced sequences such as diffusion-weighted imaging (DWI), diffusion tensor imaging (DTI), and susceptibility-weighted imaging (SWI) is not widespread, and data supporting their utility remains more controversial and less robust than in traumatic brain injury. There are diminishing returns from the acquisition of multiple sequences, and this must be weighed with considerations of safety, logistics, and resource allocation for patients who often require monitoring and anesthetic equipment and close oversight by healthcare personnel.50 Reducing the length of the total exam also increases the availability of the MR scanner.6

PROGNOSTIC SIGNIFICANCE OF MR FINDINGS AFTER SCI SCI severity occurs along a spectrum that is encapsulated in a simplified four-pattern classification system introduced in the late 1980s and modified in the early 1990s.5,36 This has been shown in a handful of studies to correlate both with initial neurologic impairment (determined mostly using the American Spinal Cord Injury Association [ASIA] motor score) and degree of long-term improvement.21,22,51,55,59 Injuries are classified as (1) nonedematous cord, (2) edema spanning a single vertebral level, (3) edema spanning multiple levels, and (4) mixed hemorrhage and edema. After “normal” nonedematous spinal cord, which is seen in 10% to 20% of patients with SCI, single-level

461

edema has the best prognosis.6 Approximately one fourth of patients with single-level edema recover normal motor and sensory function (ASIA E). Patients with diffuse edema (Fig. 14-5) will almost invariably have permanent, at least partial, deficits with lower degrees of recovery. Those with diffuse edema who are initially ASIA A (complete motor and sensory deficit) will rarely show any meaningful improvement. Appreciable spinal cord hemorrhage is usually associated with complete motor and sensory loss and will rarely show meaningful improvement.6 Hemorrhage is rarely seen in the absence of edema. The term cord contusion has been variably used in the literature to refer to edema alone or to a specific pattern of edema with central isointensity51 or to denote frank spinal cord hemorrhage (i.e., synonymous with hemorrhagic contusion).16 Prognosis from spinal cord hemorrhage appears to be slightly better when less than 50% of the spinal cord cross-sectional area is involved.27,52,62 The length of hemorrhage does not correlate with injury severity, likely because of the uniformly poor outcome in these patients regardless of whether single or multiple levels are involved (see Fig. 14-1D). For the same reason, complete cord transections are not included in prognostic grading scales6 (Fig. 14-6). Cord expansion is seen with more severe injury resulting from accumulation of both intracellular and interstitial fluid above and below the injury.16 The utility of cord expansion as a predictor of outcome is controversial because the degree of enlargement can be limited by posttraumatic, degenerative, or congenital stenosis. Greater degrees of cord compression and canal compromise, best seen using axial T2 images, may correlate with worse prognosis.6,12,22,42,56 Spinal cord compression specifically by extraaxial hematoma may be another independent predictor of poor outcome,56 as is severity of prevertebral soft tissue injury, although this remains controversial.61

EVOLUTION OF SPINAL CORD INJURY—FINDINGS AT MRI Because of the small caliber of the cord and narrow space within the spinal canal, even small foci of hemorrhage within the spinal cord are associated with devastating neurologic outcomes and poor improvement over time. As such, SCI is not generally thought of in the discrete terms of primary and secondary injury that are used to characterize traumatic brain injury. The relationship between evolution of signal abnormality and time after injury has been studied, but major discrepancies between exam and electrophysiologic testing and MRI appearance are common.35 The ideal time interval for performing MRI after initial injury remains controversial. Surgical decompression within 24 hours for SCI patients and “very early” decompression within 12 hours for those with incomplete cervical SCIs has been recommended based on a worldwide survey of 971 spine surgeons.19 Although these recommendations would seem to support MRI in the early acute phase, it is not uniformly practiced at many centers. Bondurant et al. previously recommended that MRI be performed within 24 to 72 hours after trauma.5 Generally MRI plays an ancillary role to neurophysiologic testing, with additional prognostic information provided by injury morphology, as described in the previous section. As SCI transitions from acute to chronic phases, it is important to understand expected secondary morphologic changes resulting from injury evolution in the posttraumatic lesion area.35 Initially, edema with or without hemorrhage spreads cranially and caudally from the injury epicenter (Fig. 14-7). Shortly after injury, spinal cord hemorrhage is usually appreciated as low signal on T2 and edema as high signal.16,59 Extraneural blood may take only hours to develop high signal from methemoglobin (see Fig. 14-4B), whereas

http://radiologyebook.com

462

PART II  CT and MR Imaging of the Whole Body

A

C

B

D FIG 14-3  Images of an 82-year-old man after a fall down stairs. Bony fusion of the vertebral bodies and posterior elements were present throughout the cervical spine. A, A coronally oriented fracture plane is seen traversing the posterior vertebral body of C4. An epidural hematoma is faintly visualized and suggested by loss of the normal dorsal epidural fat plane (solid arrow), which is preserved below C7 (open arrow). The epidural hematoma is well visualized on the sagittal PD (B) and STIR (C) images (solid white arrows). On the STIR image (C), long arrows point to injured disks at C5-C6 and C6-C7, and there is discontinuity of the ligamentum flavum at C6-C7, consistent with ligamentum flavum tear (open arrow). D, Axial T2-weighted imaging shows the epidural hematoma (open arrow) severely compressing the cord (solid arrow). This necessitated urgent surgical decompression.

http://radiologyebook.com

CHAPTER 14  Spinal Cord Injury

463

*

B

A

FIG 14-4  This 65-year-old man sustained a T12 burst fracture after a fall off a ladder. A, Sagittal STIR image shows marrow edema (*) and low-signal-intensity transverse fracture lines spanning the vertebral body. Edema is seen within the conus (open arrow). There is traumatic disk herniation at T12-L1, with an overlying epidural hematoma (arrows), which appears slightly hyperintense on the sagittal T1 image (B [arrows]).

Posttraumatic cysts are synonymous with syringomyelia or syrinx. These usually follow the signal intensity of cerebrospinal fluid (CSF) and are well defined (Fig. 14-9). They result from rents in the ependymal lining of the central canal but are not covered by ependyma. Syringomyelia can sometimes be difficult to differentiate from hydromyelia (dilation of the central canal), and in such cases the term syringohydromyelia is used. Once a posttraumatic cyst has fully formed—usually within 1 year after injury—the cyst and spinal cord area are usually stable. In some instances posttraumatic cyst enlargement is observed (Fig. 14-10), but this does not necessarily correlate with symptomatology. Posttraumatic syrinx develops in 5% to 8% of SCI.17,35 Tethering and syrinx progression are suspected if the myotomes affected become more rostral. Surgical intervention may be warranted purely on the basis of MR findings if the rostral extent threatens to encroach on the brainstem.35

Concurrent Soft Tissue and Bony Injury in SCI

FIG 14-5  Hyperextension injury from trivial fall in a 50-year-old man, resulting in multilevel spinal cord edema spanning C3 and C4 (arrows). Clinically the patient had incomplete spinal cord injury, and he improved one ASIA score over time.

intraneurally this may take days to weeks, limiting the utility of T1-weighted imaging in the early acute phase after injury (Fig. 14-8). Sagittal T2-weighted and STIR images offer a global view of both the cord and discoligamentous structures along the full extent of the imaged spine.16,34 Acute injury is followed by a second subacute stage marked by regression of edema and hemorrhage,35 during which the transverse diameter of the cord may subside to normal or shrink. In the final chronic stage, a posttraumatic cyst may form.35

Important soft tissue structures to evaluate at MRI in patients with SCI include the prevertebral musculature, anterior longitudinal ligament (ALL), intervertebral disk, and posterior longitudinal ligament (PLL) ventrally; and facet capsules, ligamentum flavum, interspinous ligament, supraspinous ligament, and paraspinal musculature dorsally.50 The craniocervical junction should be evaluated for integrity of the transverse and alar ligaments, which may be disrupted in craniocervical distraction injury. Normal appearances of these structures are shown in Figure 14-11. The ALL and facet capsules are the most important determinants of mechanical stability.14,69 The sine qua non of clinically important ligamentous injury is discontinuity of normal low ligamentous signal with intervening gap and high T2 signal indicative of hemorrhage and edema. These findings are best appreciated on sagittal STIR or T2-weighted imaging (see Figs. 14-1A, 14-2, 14-3C, 14-4A, 14-6A, 14-7A and B, and 14-8A). Despite its importance for determining stability, the ALL is among the most difficult structures to visualize because the inner layer is tightly adherent to the anterior

http://radiologyebook.com

464

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 14-6  A, Sagittal STIR image of a T4-T5 fracture-dislocation in a 64-year-old woman involved in a motor vehicle collision. The cord is transected at this level, and there is an intervening gap (open arrow). B, On sagittal T1-weighted image, dorsal epidural fat outlines the severed and displaced rostral and caudal ends of the spinal cord (open arrows).

cortex of the vertebral body and disk anulus. The sensitivity of MRI for cervical ALL injuries in patients with SCI ranges from 46% to 71%; for disk injury, approximately 93%; and for ligamentous injuries  posterior to the vertebral bodies, from 67% to 95%.6,26,40,50 Posteriorly, sensitivity is lowest for the ligamentum flavum, the thin ligament bridging the laminae; however, this structure has a relatively minor role in determining spinal stability. Posteriorly, sensitivity is highest for supraspinous ligament tears.26 Muscle injury is managed conservatively but may help give a general picture of the extent of injury. SCI is more likely with severe injuries that cause retropulsion of bone  into the canal or narrowing as the result of rotation and translation, which may crush the cord. MRI most precisely details degree of canal encroachment and spinal cord compression in these cases. MRI misses a substantial proportion of fractures, and as mentioned, CT is generally superior for this purpose. This is particularly the case for posterior elements where there is a paucity of cancellous bone.16 Evidence of concomitant facet capsular injury at MRI may call attention to a fracture. MRI provides superior definition of extraaxial entities that may be causative of SCI and require urgent decompression, such as epidural hematoma and traumatic disk herniation. In the fused spine, MRI may reveal nondisplaced fractures or disk disruptions not readily apparent on CT (see Fig. 14-3). These injuries nevertheless have a high propensity for severe instability because the cranial and caudal portions of the fractured spine act as long lever arms. The threshold for obtaining MRI in these patients should be low. The bulky flowing osteophytes seen in diffuse idiopathic skeletal hyperostosis result in a propensity for injury planes traversing the vertebral  body, whereas disk-level disruption is more common in ankylosing spondylitis.16,28,46 Dural tears are found in 7% to 28% of patients operated on for spinal injuries and are most common in lumbar burst fractures.8,33,44 One quarter of patients with lumbar burst fracture and neurologic deficit have dural tears requiring repair, and the great majority of

lumbar burst fractures with dural tear have a neurologic deficit.33 MRI can be used to identify dural tears acutely, although diagnosis with any conventional imaging modality is challenging.44 Dural tears are very difficult to visualize when rents are less than 1  cm in length.38 On T2 or STIR images, accumulations of CSF may be ill defined and blend into high interfacial and intermuscular signal within adjacent soft tissue. In more severe trauma, larger meningeal lacerations result in more spread-out collections. Indirect findings on MRI, including laminar fracture with gap, wide interpedicular distance, and marked canal narrowing, all increase the likelihood of dural tear.38 Isotropic sequences with heavy T2 weighting can have added benefit in making the diagnosis and identifying the origin and course of CSF fistulae.38 CT myelography with intrathecal contrast may be an option in patients who cannot undergo MRI, but the procedure is often not tolerated in the trauma setting. The use of MR myelography has been described with intrathecal gadolinium administration, but the use of gadolinium for this purpose remains off label in both the United States and abroad.44 Dural tears are clinically important because they can act as foci for herniation of neural elements and nerve root entrapment,8,38 which may require decompression with a posterior surgical approach. Meningeal infection or inflammation may cause failure to heal and chronic posttraumatic meningocele.38 Chronically, neural structures may become entrapped in scar tissue.44 In brachial and (less commonly) in lumbosacral plexus traction injuries, nerve root avulsions are usually associated with meningeal tears and CSF collections. Reepithelialization of these collections results in formation of pseudomeningoceles54 (Fig. 14-12). In the cervical spine these are virtually pathognomonic for nerve root avulsion.54

Canal Stenosis as a Predictor of SCI A number of anatomic characteristics may predispose individuals to SCI after trauma.1,18,20,25,60,64-66 The most widely evaluated radiographic measurement parameter of canal stenosis is the Torg-Pavlov ratio

http://radiologyebook.com

CHAPTER 14  Spinal Cord Injury

*

A

B

C FIG 14-7  A, Sagittal STIR image of a 53-year-old woman with hyperflexion injury after fall, showing spinal cord edema spanning C3-C5 (arrows). There is interspinous widening at C5-C6 (*), with mild focal kyphosis at this level. B, Parasagittal STIR image shows C5-C6 facet subluxation. The patient underwent decompressive laminectomy and anterior corpectomy, diskectomy, and fusion (ACDF). C, On follow-up MRI several days later there is evolution of the SCI, with increased craniocaudal extent of spinal cord edema and expansion of the spinal cord.

http://radiologyebook.com

465

466

PART II  CT and MR Imaging of the Whole Body

A

C

B

D FIG 14-8  This 50-year-old man had an unstable burst fracture at C4 and quadriplegia. A, Sagittal T2-weighted image shows disk disruption and stripping of the anterior and posterior longitudinal ligaments (open arrows) at C4-C5. There is spinal cord edema from C2-C6, and a focus of low signal intensity at C4-C5 consistent with hemorrhage (solid arrow). B, On sagittal SWI the hemorrhage appears more extensive (arrow). C, On the corresponding sagittal T1-weighted image the hemorrhage is not appreciated. D, On follow-up scan 1 week later, there is high signal within the cord on T1 (arrow), corresponding with the area of low signal previously identified on sagittal T2. This reflects evolution of blood products and accumulation of methemoglobin in the subacute phase of SCI.

http://radiologyebook.com

CHAPTER 14  Spinal Cord Injury

FIG 14-9  This 37-year-old woman presented with central cord syndrome after a fall 2 years prior to this study. On this sagittal STIR image, there is atrophic thinning of the cord (solid arrows) and a focus of syrinx at C4-5 (open arrow).

FIG 14-10  A 56-year-old man with remote craniocervical injury after a motor vehicle collision. The patient developed progressive lower extremity weakness and was found to have a long segment of syringohydromyelia spanning C5-T9, with only a thin residual rim of spinal cord along the periphery (arrow). This was surgically treated with laminectomy and drainage by marsupialization.

D

O

O

* A

467

B FIG 14-11  Normal spinal column in a 20-year-old man after a motor vehicle collision. A, Sagittal T2-weighted image of the cervical spine shows intact anterior longitudinal ligament (black open arrow), posterior longitudinal ligament (white open arrow), ligamentum flavum (thin white arrow), interspinous ligaments (*), and supraspinous ligament (thin black arrow). B, Axial T2 image through the craniocervical junction shows intact alar ligaments (arrow), which fan out from the dens (D) to the occipital condyles (O). Continued

http://radiologyebook.com

468

PART II  CT and MR Imaging of the Whole Body

D

A

A

C

D FIG 14-11, cont’d  C, Axial T2 also shows an intact transverse ligament, which extends from the medial aspect of one lateral mass of the atlas (A) to the other, completing the osseoligamentous ring around the dens (D), formed anteriorly by the anterior arch of the atlas. D, In the parasagittal T2-weighted image, normal alignment of the facets and normal signal within the facet capsules are seen.

risk of SCI after minor trauma.1 The canal diameter at MRI is measured from the outer dorsal and ventral margins of the subarachnoid space (Fig. 14-13). The clinical implications of a narrow spinal canal in asymptomatic patients remain unclear.

ADVANCED IMAGING TECHNIQUES

FIG 14-12  A remote motor vehicle collision caused right brachial plexus injury with nerve root avulsion in this 53-year-old patient. A chronic posttraumatic pseudomeningocele is seen at C6-C7, extending through the right neural foramen at this level (arrow).

(TPR),66,65 derived by dividing the narrowest midvertebral sagittal canal diameter in the cervical spine by the sagittal diameter of the corresponding vertebral body at the same level. The TPR is a plain radiographic parameter with limited positive predictive value, partly attributable to variability in vertebral body dimensions, a factor not directly dependent on canal diameter.3,29,31,49,65 The TPR also does not take into account spondylotic changes occurring at the disk level. Recently, measurement parameters have been reevaluated with MRI. A disk-level canal diameter of less than 8 mm is associated with a high

Advanced techniques will be necessary to explain variations in improvement between patients with similar imaging findings on conventional MRIs or for monitoring regeneration.6,10,35,39,48 Reproducibility and re-test reliability is critical for following injury evolution longitudinally, especially for DTI.41,43 At present, advanced sequences that are gaining widespread use for traumatic brain injury, including DWI, SWI, and DTI, have limited clinical applicability in the spinal cord. These sequences are limited by motion artifacts, the narrow caliber of the spinal cord, and susceptibility and field inhomogeneity resulting from adjacent bony structures and lungs.39,48,70 Conventional MR sequences detect gross morphologic abnormalities but do not have the ability to resolve the integrity of long white matter tracts. DWI has the potential to detect early pathologic changes not apparent on conventional sequences; however, the small body of literature evaluating the utility of DWI at 1.5 T in the early acute setting shows similar sensitivity to T2, with no added benefit.48,58,67 SWI may identify smaller amounts of hemorrhage than GRE, but it is even more prone to motion disturbances (see Fig. 14-1B and C) and can result in artifactual microhemorrhage. SWI has been shown to be more sensitive for hemorrhage than T2, although the study lacked a reference standard.70 DTI quantifies the diffusion of water molecules along tensors parallel and perpendicular to axonal tracts and may eventually be used in the clinical setting to track axons and distinguish pathology involving white and gray matter.43,57 To date, only a few small studies have shown a correlation between decreased fractional anisotropy with DTI and clinical SCI severity,9,43 but fiber tracking suitable for clinical use will require substantial technical improvements.39 The cervical spine becomes most stenotic with extension and  is narrowest at the C4 level.25 Dynamic imaging may be useful for evaluation of critical stenosis during hyperextension or

http://radiologyebook.com

CHAPTER 14  Spinal Cord Injury

469

B

A

FIG 14-13  This 60-year-old man was hit in the head by falling scaffolding and presented clinically with an incomplete spinal cord injury. A, Sagittal CT shows ossification of the posterior longitudinal ligament from C2-C3 to C4 (arrow), with resultant severe narrowing of the spinal canal. B, Sagittal T2-weighted image showed no abnormal signal in the cord despite a minimum disk-level canal diameter of only 5 mm (doubleheaded arrow).

hyperflexion after SCI, as well as the effect on the cord and nerve roots; however, motorized or stepwise positioning devices are still largely experimental.23

REFERENCES 1. Aebli N, Rüegg TB, Wicki AG, et al: Predicting the risk and severity of acute spinal cord injury after a minor trauma to the cervical spine. Spine J 13(6):597–604, 2013. 2. Berkowitz M: Assessing the socioeconomic impact of improved treatment of head and spinal cord injuries. J Emerg Med 11(Suppl 1):63–67, 1993. 3. Blackley HR, Plank LD, Robertson PA: Determining the sagittal dimensions of the canal of the cervical spine: The reliability of ratios   of anatomical measurements. J Bone Joint Surg Br 81-B(1):110–112, 1999. 4. Boese CK, Lechler P: Spinal cord injury without radiologic abnormalities in adults: A systematic review. J Trauma Acute Care Surg 2013;75(2):320330 10.1097/TA.0b013e31829243c9. 5. Bondurant F, Cotler H, Kulkarni M, et al: Acute spinal cord injury. A study using physical examination and magnetic resonance imaging. Spine 15(3):161–168, 1990. 6. Bozzo A, Marcoux J, Radhakrishna M, et al: The role of magnetic resonance imaging in the management of acute spinal cord injury.   J Neurotrauma 28(8):1401–1411, 2010. 7. Bracken M, Freeman D, Jr, Hellenbrand K: Incidence of acute traumatic hospitalized spinal cord injury in the United States, 1970-1977. Am J Epidemiol 113(6):615–622, 1981. 8. Cammisa F, Eismont F, Green B: Dural laceration occurring with burst fractures and associated laminar fractures. J Bone Joint Surg 71(7):1044– 1052, 1989. 9. Cheran S, Shanmuganathan K, Zhuo J, et al: Correlation of MR diffusion tensor imaging parameters with ASIA motor scores in hemorrhagic and nonhemorrhagic acute spinal cord injury. J Neurotrauma 28(9):1881– 1892, 2011.

10. Clark CA, Werring DJ: Diffusion tensor imaging in spinal cord: Methods and applications—A review. NMR Biomed 15(7–8):578–586, 2002. 11. Como JJ, Samia H, Nemunaitis GA, et al: The misapplication of the term spinal cord injury without radiographic abnormality (SCIWORA) in adults. J Trauma Acute Care Surg 73(5):1261–1266, 2012. 10.097/ TA.0b013e318265cd8c. 12. Dai L, Jia L: Central cord injury complicating acute cervical disc herniation in trauma. Spine 25(3):331–336, 2000. 13. Daneshvar P, Roffey DM, Brikeet YA, et al: Spinal cord injuries related to cervical spine fractures in elderly patients: Factors affecting mortality. Spine J 13(8):862–866, 2013. 14. Dreizin D, Letzing M, Sliker CW, et al: Multidetector CT of blunt cervical spine trauma in adults. Radiographics 34(7):1842–1865, 2014. 15. Dreizin D, Munera F: Blunt polytrauma: evaluation with 64-section whole-body CT angiography. Radiographics 32(3):609–631, 2012. 16. Dundamadappa S, Cauley K: MR imaging of acute cervical spinal ligamentous and soft tissue trauma. Emerg Radiol 19(4):277–286, 2012. 17. Edgar R, Quail P: Progressive post-traumatic cystic and non-cystic myelopathy. Br J Neurosurg 8(1):7–22, 1994. 18. Eismont F, Clifford S, Goldberg M, et al: Cervical sagittal spinal canal size in spine injury. Spine (Phila Pa 1976) 9(7):663–666, 1984. 19. Fehlings M, Rabin D, Sears W, et al: Current practice in the timing   of surgical intervention in spinal cord injury. Spine (Phila Pa 1976) 35(21 Suppl):S166–S173, 2010. 20. Firooznia H, Ahn J, Rafii M, et al: Sudden quadriplegia after a minor trauma. The role of preexisting spinal stenosis. Surg Neurol 23(2):165– 168, 1985. 21. Flanders AE, Spettell CM, Friedman DP, et al: The relationship between the functional abilities of patients with cervical spinal cord injury and the severity of damage revealed by MR imaging. Am J Neuroradiol 20(5):926–934, 1999. 22. Flanders A, Spettell C, Tartaglino L, et al: Forecasting motor recovery after cervical spinal cord injury: Value of MR imaging. Radiology 201(3):649–655, 1996.

http://radiologyebook.com

470

PART II  CT and MR Imaging of the Whole Body

23. Gerigk L, Bostel T, Hegewald A, et al: Dynamic magnetic resonance imaging of the cervical spine with high-resolution 3-dimensional T2-imaging. Clin Neuroradiol 22(1):93–99, 2012. 24. Geurts BHJ, Andriessen TMJC, Goraj BM, et al: The reliability of magnetic resonance imaging in traumatic brain injury lesion detection. Brain Injury 26(12):1439–1450, 2012. 25. Ghogawala Z, Whitmore RG: Asymptomatic cervical canal stenosis:   Is there a risk of spinal cord injury? Spine J 13(6):613–614, 2013. 26. Goradia D, Linnau KF, Cohen WA, et al: Correlation of MR imaging findings with intraoperative findings after cervical spine trauma. Am J Neuroradiol 28(2):209–215, 2007. 27. Grabb PA, Pang D: Magnetic resonance imaging in the evaluation of spinal cord injury without radiographic abnormality in children. Neurosurgery 35(3):406–414, 1994. 28. Hanson JA, Mirza S: Predisposition for spinal fracture in ankylosing spondylitis. Am J Roentgenol 174(1):150, 2000. 29. Herzog R, Wiens J, Dillingham M, et al: Normal cervical spine morphometry and cervical spinal stenosis in asymptomatic professional football players. Plain film radiography, multiplanar computed tomography, and magnetic resonance imaging. Spine 16(6 [Suppl]):S178–S186, 1991. 30. Hogan GJ, Mirvis SE, Shanmuganathan K, et al: Exclusion of unstable cervical spine injury in obtunded patients with blunt trauma: Is MR imaging needed when multi–detector row CT findings are normal? Radiology 237(1):106–113, 2005. 31. Hukuda S, Xiang L, Imai S, et al: Large vertebral body, in addition to narrow spinal canal, are risk factors for cervical myelopathy. J Spinal Disord 9(3):177–186, 1996. 32. Kasimatis GB, Panagiotopoulos E, Megas P, et al: The adult spinal cord injury without radiographic abnormalities syndrome: Magnetic resonance imaging and clinical findings in adults with spinal cord injuries having normal radiographs and computed tomography studies.   J Trauma Acute Care Sur 2008;65(1):86-93 10.1097/ TA.0b013e318157495a. 33. Keenen T, Antony J, Benson D: Dural tears associated with lumbar burst fractures. J Orthop Trauma 4(3):243–245, 1990. 34. Krakenes J, Kaale BR: Magnetic resonance imaging assessment of craniovertebral ligaments and membranes after whiplash trauma. Spine 31(24):2820–2826, 2006. 10.1097/01.brs.0000245871.15696.1f. 35. Kramer J, Freund P, Curt A: Traumatic spinal cord injury: Chronic spinal cord injury and recovery. In Cohen-Adad J, editor: Quantitative MRI of the Spinal Cord, Amsterdam, 2014, Elsevier Inc, pp 49–55. 36. Kulkarni MV, Bondurant FJ, Rose SL, et al: 1.5 Tesla magnetic resonance imaging of acute spinal trauma. Radiographics 8(6):1059–1082, 1988. 37. Lee H-M, Kim H-S, Kim D-J, et al: Reliability of magnetic resonance imaging in detecting posterior ligament complex injury in thoracolumbar spinal fractures. Spine 25(16):2079–2084, 2000. 38. Lee IS, Kim HJ, Lee JS, et al: Dural tears in spinal burst fractures: Predictable MR imaging findings. Am J Neuroradiol 30(1):142–146, 2009. 39. Lundell H, Barthelemy D, Biering-Sørensen F, et al: Fast diffusion tensor imaging and tractography of the whole cervical spinal cord using point spread function corrected echo planar imaging. Magn Reson Med 69(1):144–149, 2013. 40. Martínez-Pérez R, Paredes I, Cepeda S, et al: Spinal cord injury after blunt cervical spine trauma: Correlation of soft-tissue damage and extension of lesion. Am J Neuroradiol 35(5):1029–1034, 2014. 41. Middleton DM, Mohamed FB, Barakat N, et al: An investigation of motion correction algorithms for pediatric spinal cord DTI in healthy subjects and patients with spinal cord injury. Magn Reson Imaging 32(5):433–439, 2014. 42. Miyanji F, Furlan JC, Aarabi B, et al: Acute cervical traumatic spinal cord injury: MR imaging findings correlated with neurologic outcome— Prospective study with 100 consecutive patients. Radiology 243(3):820– 827, 2007. 43. Mulcahey MJ, Samdani A, Gaughan J, et al: Diffusion tensor imaging in pediatric spinal cord injury: Preliminary examination of reliability and clinical correlation. Spine 37(13):E797–E803, 2012. 10.1097/ BRS.0b013e3182470a08.

44. Muñoz A, Mateo I, Lorenzo V, et al: MR cisternography/myelography of post-traumatic spinal CSF fistulae and meningeal lesions in small animals. Acta Radiol 54(5):569–575, 2013. 45. Nuñez D, Jr, Ahmad A, Coin C, et al: Clearing the cervical spine in multiple trauma victims: A time-effective protocol using helical computed tomography. Emerg Radiol 1(6):273–278, 1994. 46. Paley D, Schwartz M, Cooper P, et al: Fractures of the spine in diffuse idiopathic skeletal hyperostosis. Clin Orthop Relat Res 267:22–32, 1991. 47. Pang D, Wilberger JJ: Spinal cord injury without radiographic abnormalities in children. J Neurosurg 57(1):114–129, 1982. 48. Pouw MH, van der Vliet AM, van Kampen A, et al: Diffusion-weighted MR imaging within 24 hours post-injury after traumatic spinal cord injury: A qualitative meta-analysis between T2-weighted imaging and diffusion-weighted MR imaging in 18 patients. Spinal Cord 50(6):426– 431, 2012. 49. Presciutti SM, DeLuca P, Marchetto P, et al: Mean subaxial space available for the cord index as a novel method of measuring cervical spine geometry to predict the chronic stinger syndrome in American football players. J Neurosurg Spine 11(3):264–271, 2009. 50. Provenzale J: MR imaging of spinal trauma. Emerg Radiol 13(6):289–297, 2007. 51. Ramón S, Domínguez R, Ramírez L, et al: Clinical and magnetic resonance imaging correlation in acute spinal cord injury. Spinal Cord 35(10):664–673, 1997. 52. Regenbogen V, Rogers L, Atlas S, et al: Cervical spinal cord injuries in patients with cervical spondylosis. AJR Am J Roentgenol 146(2):277–284, 1986. 53. Rozzelle CJ, Aarabi B, Dhall SS, et al: Spinal cord injury without radiographic abnormality (SCIWORA). Neurosurgery 72:227–233, 2013. 10.1227/NEU.0b013e3182770ebc. 54. Sasaka KK, Phisitkul P, Boyd JL, et al: Lumbosacral nerve root avulsions: MR imaging demonstration of acute abnormalities. Am J Neuroradiol 27(9):1944–1946, 2006. 55. Schaefer D, Flanders A, Osterholm J, et al: Prognostic significance of magnetic resonance imaging in the acute phase of cervical spine injury.   J Neurosurg 76(2):218–223, 1992. 56. Selden NR, Quint DJ, Patel N, et al: Emergency magnetic resonance imaging of cervical spinal cord injuries: Clinical correlation and prognosis. Neurosurgery 44(4):785–792, 1999. 57. Shanmuganathan K, Gullapalli RP, Zhuo J, et al: Diffusion tensor MR imaging in cervical spine trauma. Am J Neuroradiol 29(4):655–659, 2008. 58. Shen H, Tang Y, Huang L, et al: Applications of diffusion-weighted MRI in thoracic spinal cord injury without radiographic abnormality. Int Orthop 31(3):375–383, 2007. 59. Shimada K, Tokioka T: Sequential MR studies of cervical cord injury: Correlation with neurological damage and clinical outcome. Spinal Cord 37(6):410–415, 1999. 60. Song K-J, Choi B-W, Kim S-J, et al: The relationship between spinal stenosis and neurological outcome in traumatic cervical spine injury: An analysis using Pavlov’s ratio, spinal cord area, and spinal canal area. Clin Orthop Surg 1(1):11–18, 2009. 61. Sun L, Shen Y, Li Y: Quantitative magnetic resonance imaging   analysis correlates with surgical outcome of cervical spinal cord injury without radiologic evidence of trauma. Spinal Cord 52(7):541–546, 2014. 62. Tewari MK, Gifti DS, Singh P, et al: Diagnosis and prognostication of adult spinal cord injury without radiographic abnormality using magnetic resonance imaging: Analysis of 40 patients. Surg Neurol 63(3):204–209, 2005. 63. Theocharopoulos N, Chatzakis G, Damilakis J: Is radiography justified for the evaluation of patients presenting with cervical spine trauma? Med Phys 36(10):4461–4470, 2009. 64. Torg JS, Corcoran TA, Thibault LE, et al: Cervical cord neurapraxia: Classification, pathomechanics, morbidity, and management guidelines.   J Neurosurg 87(6):843–850, 1997.

http://radiologyebook.com

CHAPTER 14  Spinal Cord Injury 65. Torg J, Naranja R, Jr, Pavlov H, et al: The relationship of developmental narrowing of the cervical spinal canal to reversible and irreversible injury of the cervical spinal cord in football players. J Bone Joint Surg Am 78(9):1308–1314, 1996. 66. Torg J, Pavlov H, Genuario S, et al: Neurapraxia of the cervical spinal cord with transient quadriplegia. J Bone Joint Surg Am 68(9):1354–1370, 1986. 67. Tsuchiya K, Fujikawa A, Honya K, et al: Value of diffusion-weighted MR imaging in acute cervical cord injury as a predictor of outcome. Neuroradiology 48(11):803–808, 2006. 68. Utz M, Khan S, O’Connor D, et al: MDCT and MRI evaluation of cervical spine trauma. Insights Imaging 5(1):67–75, 2014.

471

69. Vaccaro AR, Hulbert RJ, Patel AA, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the discoligamentous complex. Spine 32(21):2365–2374, 2007. 10.1097/ BRS.0b013e3181557b92. 70. Wang M, Dai Y, Han Y, et al: Susceptibility weighted imaging in detecting hemorrhage in acute cervical spinal cord injury. Magn Reson Imaging 29(3):365–373, 2011. 71. Yucesoy K, Yuksel KZ: SCIWORA in the MRI era. Clin Neurol Neurosurg 110(5):429–433, 2008. 72. Zwimpfer T, Bernstein M: Spinal cord concussion. J Neurosurg 72(6):894–900, 1990.

http://radiologyebook.com

15  Neurodegenerative Disorders Federica Agosta, Francesca Caso, and Massimo Filippi

It is often clinically difficult to differentiate among the various neurodegenerative disorders such as dementias, hydrocephalus, and movement disorders, and even between pathologic and normal aging. In recent years, more specific and sensitive neuroimaging criteria have evolved to more accurately establish the correct diagnosis in these disorders. Magnetic resonance imaging (MRI), with its ability to image both the structure and function of the living brain, is an invaluable tool. In general, the tendency is to move away from simply excluding other (brain) diseases toward finding specific pointers to a diagnosis in the field of neurodegenerative disorders. The contribution of MRI to the evaluation of neurodegenerative disorders will no doubt increase as the population ages and as new pharmacologic and surgical therapies for these disorders develop.

NORMAL AGING Normal aging of the brain can be defined as a set of structural, metabolic and functional abnormalities of both the gray (GM) and white (WM) matter. Incidental WM hyperintensities (WMHs) can be seen on computed tomography (CT) scans, but they are better visualized on proton density–weighted, T2-weighted, and fluidattenuated inversion recovery (FLAIR) MRIs of the brain in about 30% of healthy subjects older than 60 years.128 Their prevalence rises steadily with increasing age. Age-related WMHs are usually located in the deep and subcortical WM and around the ventricles (Fig. 15-1).45,63 Brainstem lesions are less common. They have different features on MRI scans, the most typical of which are punctate hyperintensities, periventricular caps and halos, and large lesions (which can be confluent).45,63 The severity of normally occurring brain volume loss is variable, ranging from minimal to moderately severe. Age-related changes in brain volume are apparent in both postmortem and in vivo MRI studies (see Fig. 15-1).8,117 Pathologic studies have found that brain weight peaks by the middle to late teens and declines slowly (0.1%0.2% a year) until the age of 60 to 70, after which losses accelerate.117 Longitudinal MRI studies report that rates of global atrophy in healthy people increase gradually with age from an annual rate of 0.2% a year at age 30 to 50 to 0.3% to 0.5% at age 70 to 80.8 GM volume starts to decrease early in life (at the end of the first decade), whereas WM volume starts to decrease at the fourth decade. On average, GM volume loss is greater in the cortex than in subcortical structures. Age-related cortical density decrease seems to follow a gradient, with greatest and earliest changes occurring in association areas, especially those of the prefrontal cortex.116 However, recent MRI studies have described, in addition to an age-related atrophy of associative cortices, a widespread age-related thinning of a large portion of the cortical ribbon, including several primary areas, which were previously considered to be spared by aging.116 Cerebral WM also exhibits various types of degenerative

472

changes. Although WM loss starts later in life than that of GM, the rate of such a tissue loss seems then to be faster.103 WM age-related loss is more pronounced in subcortical regions (with a preferential involvement of the frontal lobe). Diffusion tensor (DT) MRI studies confirmed reduced WM integrity in elderly people, especially in the frontal lobes.140

DEMENTIA Dementia can be caused by myriad pathologic processes, including anatomic (e.g., abscess, tumor, subdural hematoma, posttraumatic encephalomalacia, diffuse axonal injury), metabolic (e.g., electrolyte imbalance, nutritional deficiency, endocrinopathy, toxic exposure, medications), psychiatric (depression), degenerative (e.g., Alzheimer’s disease [AD], Parkinson’s disease [PD], frontotemporal dementia [FTD], dementia with Lewy bodies [DLB]), vascular (e.g., cerebral infarction, Binswanger’s disease, CADASIL), infectious/inflammatory (e.g., chronic meningitis, vasculitis, prion disease, Hashimoto’s encephalopathy), demyelinating disease (e.g., multiple sclerosis), and paraneoplastic phenomena (e.g., limbic encephalitis). It is often not possible to establish the cause by clinical examination alone. Neuroimaging assessment, traditionally and currently, is used to rule out the presence of some treatable causes of dementia. Current imaging techniques allow us to determine whether findings are consistent with or even typical of a certain diagnosis.47

Alzheimer’s Disease AD, the most common dementing disorder in older adults,90 has been recognized as one of the most significant health problems of the 20th and 21st centuries. AD is estimated to affect 10% of people older than 65 years and 50% of individuals older than 85 years.107 Two abnormal protein aggregates characterize AD pathology: neuritic plaques and neurofibrillary tangles (NFTs).107 Neuritic plaques are extracellular deposits and consist of a dense central core of amyloid β fibrils with inflammatory cells and dystrophic neurites in its periphery. The second major proteinopathy in AD is aggregated tau, which consists of intraneuronal polymers primarily composed of hyperphosphorylated tau in the form of NFTs.107 In typical late-onset (arbitrarily defined as age at onset > 65 years) AD, the medial temporal lobes (MTL), especially the hippocampus and entorhinal cortex, are among the earliest sites of pathologic involvement11,124 (Fig. 15-2). MTL atrophy in the presence of memory loss is now one of the supportive biomarkers to make a diagnosis of AD proposed by the new diagnostic criteria.38,39,90 Other severely affected regions include the posterior portion of the cingulate gyrus and the precuneus on the medial surface (see Fig. 15-2), and the parietal, posterior superior temporal, and frontal regions on the lateral cerebral surfaces.48,124

http://radiologyebook.com

473

CHAPTER 15  Neurodegenerative Disorders

A

B

D

E

C

FIG 15-1  Axial proton density–weighted (A), T2-weighted (B), and FLAIR (C) MRIs of the brain from a 73-year-old healthy subject. Multiple hyperintense lesions suggestive of multifocal white matter pathology are visible. In C, CSF signal suppression allows better identification of the lesions located in the periventricular and juxtacortical regions. D and E, Axial T1-weighted MRIs of the brain of an 82-year-old healthy subject. Enlargement of the cortical sulci and ventricular size is evident.

R

L

FIG 15-2  Axial and coronal T1-weighted images from a typical case of Alzheimer’s disease (AD) showing prominent bilateral hippocampal and medial parietal atrophy (arrows). R, right; L, left.

http://radiologyebook.com

474

PART II  CT and MR Imaging of the Whole Body

Structural MRI studies in mild cognitive impairment (MCI)  have produced mixed results, both in terms of hippocampal and posterior cingulate and parietal involvement (absent, unilateral, or bilateral).5,14,26,137 The reasons for this variability may consist of different subject selection (i.e., diverse diagnostic inclusion criteria), small sample size (i.e., studies are not adequately powered to pick up differences even at the group level), and methodological differences. It is also worth noting that the largest source of variance in MCI studies is likely to be the intrinsic heterogeneity of the MCI population, because a relevant proportion of these subjects will not progress to dementia. MCI patients with predominant memory impairment (amnestic MCI), who are at increased risk of developing AD, have atrophy in a consistent set of cortical regions (the “cortical signature of AD”), including the MTL and temporoparietal cortex. Conversely, nonamnestic MCI shows a different pattern of atrophy characterized by relative sparing of the MTL and a regional involvement that is typically highly consistent with the observed clinical deficits. It should be emphasized that MTL atrophy may occur in other diseases as well; thus MTL atrophy alone lacks the specificity to confidently exclude other dementias, in particular in patients at the MCI stage. Early-onset AD patients (i.e., onset of symptoms before age 65) showed less prominent MTL atrophy and greater involvement of the parietal, lateral temporal, and frontal regions compared to late-onset AD cases.22,50 A specific visual rating scale has been designed, evaluating the posterior cingulate, precuneus, and superior parietal regions.77

The utility of such a scale has been assessed in pathologically proven (mostly early-onset) AD and frontotemporal lobar degeneration (FTLD) patients.80 Thirty percent of AD patients had posterior atrophy in the absence of abnormal MTL atrophy, whereas only 7% of the FTLD group had abnormal posterior atrophy scores and normal MTL.80 Adding the posterior atrophy to the MTL visual rating score improved discrimination of early-onset AD from normal controls and all AD from FTLD cases.80 Cerebral blood flow (CBF) single-photon emission computed tomography (SPECT) and 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) scans of typical AD patients demonstrate predominant hypoperfusion or reduced glucose metabolism in the temporoparietal regions, including the precuneus and the posterior cingulate cortex13 (Fig. 15-3). Functional frontal lobe involvement is also often reported in AD but usually in conjunction with and characteristically less severe than temporoparietal involvement.95 Overall hypoperfusion or hypometabolism in early-onset AD is much greater in magnitude and extent than that of late-onset AD, with similar dementia severity.108 The primary visual and sensorimotor cortices, cerebellum, thalamus, and basal ganglia are relatively spared in AD.95 FDG PET differentiates patients with MCI from healthy controls. Amnestic MCI typically shows regional hypometabolism consistent with AD, although the magnitude of reduction is milder than that in clinically probable AD cases.96,100 Longitudinal studies of patients with MCI found that if the baseline FDG PET scan suggests an AD-like

NC > FTD

AD > FTD

NC > AD

FTD > AD

FIG 15-3  FDG-PET glucose metabolism. Normal controls (NC) > frontotemporal dementia (FTD) (upper left): hypometabolic regions in FTD patients compared with NC subjects. Alzheimer’s disease (AD) > FTD (upper right): hypometabolic regions in FTD patients compared with AD patients. NC > AD (lower left): hypometabolic regions in AD patients compared with NC subjects. FTD > AD (lower right): hypometabolic regions in AD patients compared with FTD patients. (From Kanda T, et al: Comparison of grey matter and metabolic reductions in frontotemporal dementia using FDG-PET and voxel-based morphometric MR studies. Eur J Nucl Med Mol Imaging 35:2227–2234, 2008.)

http://radiologyebook.com

475

CHAPTER 15  Neurodegenerative Disorders pattern, the probability of clinical progression within several years is extremely high.25,37

Frontotemporal Dementia FTD is the umbrella term encompassing a group of progressive proteinopathies that are heterogeneous with regard to etiology and neuropathology but share (1) atrophy of the frontal and/or temporal cortex as a morphologic feature and (2) deposition of abnormal ubiquitinated protein inclusions in the cytoplasm and nucleus of neuronal and glial cells as major pathologic constituents.110 FTD includes three clinical syndromes and three major underlying neuropathologic subtypes. The clinical syndromes, which are distinguished by the early  and predominant symptoms, are: a behavioral dysexecutive disorder (behavioral variant [bv]FTD); a language disorder (primary progressive aphasia [PPA] variants); and a motor disorder such as amyotrophic lateral sclerosis (ALS), corticobasal syndrome (CBS), and progressive supranuclear palsy (PSP) syndrome. The neuropathologic

subtypes are characterized by an abnormal accumulation of proteins82: microtubule-associated protein tau (MAPT), TAR DNA-binding protein (TDP) 43, and fused in sarcoma protein (FUS). FTLD-tau, FTLD-TDP, and FTLD-FUS represent 45%, 50%, and 5% of all FTLD cases, respectively, at postmortem examination. The designation of probable bvFTD by the revised diagnostic criteria111 restricts diagnosis to patients with demonstrable functional decline and typical neuroimaging findings, including frontal and/or temporal atrophy, and hypoperfusion or hypometabolism on PET  or SPECT. Structural MRI studies showed that classic bvFTD presents  with a combination of medial frontal, orbital-insular and anterior temporal cortical atrophy12,114,119 (Fig. 15-4). Such an atrophy pattern can be readily appreciated on coronal T1-weighted MRI scans (knifeedge atrophy) (see Fig. 15-4). The MTL is more affected anteriorly (i.e., the amygdala is more affected than the hippocampus and posterior hippocampus often appears normal). Nevertheless, the typical

bvFTD

Nonfluent PPA

Semantic PPA

R

L

FIG 15-4  MRIs of patients with frontotemporal lobar degeneration syndromes. Top row, Axial and coronal T1-weighted images show a pattern of knife-edge frontal atrophy (arrows) in a patient with the behavioral variant of frontotemporal dementia (bvFTD). Middle row, Axial and coronal T1-weighted images show marked left frontal and perisylvian atrophy (arrows) in a patient with the nonfluent variant of primary progressive aphasia (PPA). Bottom row, Major involvement of the anterior temporal lobes, with left predominance (arrows), is shown in a patient with the semantic variant of PPA. R, right; L, left.

http://radiologyebook.com

476

PART II  CT and MR Imaging of the Whole Body

pattern is not necessarily present in all cases,75,92,105 particularly in patients with FTD and motor neuron disease, and the pattern of atrophy in bvFTD varies significantly across different cohorts.83,138 In some cases, bvFTD presents with remarkable atrophy of the  right anterior temporal lobe and lesser involvement of the frontal regions.136 BvFTD is identified on SPECT or PET scans by patterns of hypoperfusion or hypometabolism in frontal, insular, and anterior temporal regions that are typically quite asymmetrically centered into the frontolateral cortex69,71,94 (see Fig. 15-3). The regions mostly impaired are the medial frontal cortex, followed by the frontolateral and anterior temporal cortices. The regional pattern of predominantly frontal functional impairment in bvFTD, with relative sparing of posterior brain regions, usually allows a clear distinction between these patients and those with AD (see Fig. 15-3). In patients clinically diagnosed with PPA, who are then divided into clinical variants based on specific speech and language features characteristic for each subtype, an “imaging-supported” diagnosis can be made if the expected pattern of focal atrophy on structural MRI scans or functional involvement on SPECT and FDG is found.59 Semantic variant PPA is associated with left anterior temporal atrophy (temporal pole) affecting the lateral and ventral temporal surfaces (see Fig. 15-4) as well as particularly the anterior hippocampus, amygdala, and fusiform gyrus.51,58,97,114 Semantic patients may have left hippocampal atrophy that is at least as severe as that seen  in AD patients. In these patients, the hippocampal atrophy is  predominantly located anteriorly, with relative preservation of the posterior hippocampal regions. As the disease progresses, the right temporal lobe becomes more involved.15 The nonfluent PPA variant is associated with a characteristic pattern of left anterior perisylvian atrophy involving inferior, opercular, and insular portions of the frontal lobe58 (see Fig. 15-4). Motor and premotor regions and Broca’s area are also involved.58 Compared with controls, nonfluent patients also have atrophy of the left hippocampus; however, it is less  severe than that in AD patients.127 In the logopenic PPA variant, the pattern of atrophy primarily affects the left temporoparietal junction, including the left posterior superior and middle temporal gyri, as well as the inferior parietal lobule.57,58,93 Involvement of the left MTL is reported less consistently.58 Such a posterior temporoparietal pattern of atrophy chiefly discriminates this syndrome from the other  subtypes of PPA. In patients with semantic variant PPA, FDG PET studies showed asymmetric hypometabolism of the temporal lobes, more marked on the left side.36,99,109 A functional deficit of the left frontal opercular regions of the brain has been reported in nonfluent variant PPA patients.100,101,109 In these cases, functional involvement of bilateral caudate nuclei and thalami was also described. Logopenic PPA patients usually show a pattern of left posterior temporoparietal hypometabolism on FDG PET scans.109

Dementia with Lewy Bodies DLB is the second most common type of degenerative dementia, accounting for 10% to 15% of cases.89 The core clinical features of DLB include fluctuating attention, recurrent visual hallucinations, and spontaneous parkinsonism, as well as cognitive impairment characterized by deficits of attention, executive functions (e.g., planning deficits), and complex visual abilities.89 Memory deficits are not inevitably present, particularly in the early stages. Fluctuations of cognitive function over minutes, hours, or days are a core feature and affect mainly the level of arousal. DLB is characterized by nigrostriatal dopaminergic neurodegeneration, making dopaminergic imaging a potentially useful diagnostic tool

in the differential diagnosis with AD33 (Fig. 15-5). Low dopamine transporter uptake in the basal ganglia demonstrated by SPECT or PET has been included as a suggestive feature in the diagnostic criteria for DLB.89 On the contrary, its negativity does not exclude a clinical diagnosis of probable DLB; about 20% of probable DLB cases have a normal or inconclusive scan.102 Numerous studies reported predominant medial occipital cortex hypoperfusion or hypometabolism in DLB patients compared with AD, with a parietotemporal reduction common to both the diseases.3,67,95,96 Occipital lobe hypometabolism differentiated patients with DLB from AD in both clinically diagnosed96 and autopsy confirmed3,67,95 cohorts. No clear signature pattern of cerebral atrophy associated with DLB has been established so far. Similar to AD, a diffuse pattern of global GM atrophy including temporal, parietal, frontal, and insular cortices may occur in DLB,10,20,19 but at the same time a pattern of cortical GM loss restricted to frontal and parietal lobes has also been reported.7,139 On the whole, several volumetric studies have not found significant or disproportionate occipital atrophy in DLB.20,139 A relatively robust MR finding in DLB is that of a relative preservation of the MTL when compared with AD of similar clinical severity.19,21,139 This finding is supported by a prospective MRI study with pathologic verification that found that MTL atrophy on MRI has a robust discriminatory power for distinguishing AD from DLB (sensitivity of 91% and specificity of 94%).19 Thus a relative preservation of MTL structures on CT or MRI supports a diagnosis of DLB in the consensus diagnostic criteria.89 Subcortical structural alterations in terms of putamen atrophy have been described in some cases of DLB relative to AD,32 whereas no significant atrophy was detected in the caudate nucleus.4 A pattern of relatively focused atrophy of the midbrain, hypothalamus, and substantia innominata, with relative sparing of the hippocampus and temporoparietal cortex, has been found in DLB compared to AD cases.139 Whether these findings can contribute to an early diagnosis remains unknown. Furthermore a substantial overlap between DLB and AD with regard to atrophy in these regions detracts from the usefulness of these markers in individual cases.47

Creutzfeldt-Jakob Disease Creutzfeldt-Jakob disease (CJD) is a rare cause of rapidly progressive dementia and is one of several associated neurodegenerative illnesses whose pathogenesis is related to a small, nonviral, 30- to 35-kD proteinaceous infectious particle known as a prion.72 The loci of pathology involve the cerebral and cerebellar cortices as well as the basal ganglia, where neuronal loss, reactive astrocytosis, and formation of cytoplasmic vacuoles within the glia and neurons give the tissue a characteristic spongiform appearance on light microscopy.72 The clinical syndrome is typically one of rapid cognitive decline, often with psychosis  and delirium.72 Motor abnormalities of cerebellar dysfunction can appear, and almost all patients show pronounced myoclonus before the final phase of deepening unresponsiveness and coma. The time course of the disease from presentation until death is usually less than 1 year. Brain MRI, particularly diffusion-weighted imaging (DWI), has become increasingly important in the clinical diagnosis of prion diseases, both excluding nonprion forms of rapidly progressive dementias and demonstrating features considered typical of prion diseases. In sporadic CJD (sCJD), FLAIR and especially DWI sequences show hyperintensity in subcortical GM nuclei (striatum and thalamus) and cortical GM (so-called cortical ribboning)143 (Fig. 15-6). In addition, increased signal on T2 images is accompanied by restricted diffusion with signal hypointensity on apparent diffusion coefficient (ADC) maps (see Fig. 15-6), probably owing to specific histopathologic

http://radiologyebook.com

CHAPTER 15  Neurodegenerative Disorders

477

B

A

R

C

L

D FIG 15-5  Dopaminergic imaging (123I ioflupane dopamine transporter SPECT) in (A) a healthy control, (B) a patient with early Parkinson’s disease (PD; Hoehn and Yahr score I), (C) a patient with dementia with Lewy bodies (DLB), and (D) a patient with Alzheimer’s disease (AD). Normal control and AD patient show normal symmetric intense tracer uptake in striatum (both caudate nucleus and putamen). PD and DLB patients show bilaterally reduced uptake in the striatum, predominantly reduced in the putamen. Note the asymmetry (left reduced more than right) in PD patient (arrow in B). (From Kostic VS, et al: Parkinsonism. In Filippi M, Simon JH, editors: Imaging acute neurologic disease: A symptom-based approach. Cambridge, UK, Cambridge University Press, 2014, pp 270–290.)

findings, in particular vacuolation, gliosis, and prion deposition.53,84 Because typical vacuoles in sCJD are filled with fluid, it makes sense they would restrict water movement. In sCJD, MRI hyperintensities usually present with a typical neuroanatomic pattern of involvement.91,132 Three main patterns of GM involvement are recognized on DWI and/or FLAIR images91,143: concomitant cortical and subcortical involvement (≈two thirds of patients); only cortical involvement (
both hemispheres, often asymmetrically, especially the cingulate cortex.132,143 Other cortical regions frequently involved are the superior and middle frontal gyri, insula, precuneus, angular gyrus, and parahippocampal gyrus, with relative sparing of the precentral gyrus.132 At a subcortical level, striatal involvement is supportive for sCJD diagnosis, frequently showing an anterior-posterior gradient, with the caudate head and anterior putamen being brighter than  the posterior putamen.132,143 Thalamic DWI involvement in sCJD is often coupled with bilateral striatal increased signal involving the

http://radiologyebook.com

478

PART II  CT and MR Imaging of the Whole Body

FLAIR

DWI

ADC

A

B

C FIG 15-6  Three common variations of sporadic Creutzfeldt-Jakob disease presentation on MRI. A, Neocortical (solid arrow), limbic (dashed arrow), and subcortical gray matter hyperintensities (dotted arrow) on DWI and FLAIR scans. B, Neocortical and limbic cortex involvement. C, Limbic and subcortical involvement. Note that DWI shows the hyperintensities much more than the corresponding FLAIR sequences, and that DWI hyperintensities often have corresponding ADC hypointensity. ADC abnormalities are most easily identified in the basal ganglia. (From Vitali P, et al: Diffusion-weighted MRI hyperintensity patterns differentiate CJD from other rapid dementias. Neurology 76:1711–1719, 2011.)

dorsomedian and posterior (pulvinar) regions. A study evaluating the relationship between DWI alterations and the severity of the clinical course in sCJD suggested that patients showing hyperintensities in both the cortex and basal ganglia experience a significantly shorter interval from disease onset to akinetic mutism than those with only cortical ribbon hyperintensity.142 DWI signal changes usually are more evident and earlier in the disease course than T2 or FLAIR hyperintensities,132 showing focal changes not yet apparent on FLAIR images in up to 20% of cases.132,143 In sCJD, DWI has approximate sensitivity and specificity greater than 90%,132,143 which are higher than those of current clinically available cerebrospinal fluid (CSF) biomarkers such as 14-3-3, total tau, and neuron-specific enolase.30,146 The 2009 European criteria for probable sCJD consider increased signal intensity on FLAIR or DWI as an equivalent diagnostic finding to either elevated CSF 14-3-3 protein or periodic sharp and slow wave complexes on the electroencephalogram.146 In clinical practice, sCJD is frequently misdiagnosed as other conditions. According to UCSF (University of California, San Francisco) MRI criteria, four MRI features may help differentiate sCJD from nonprion diagnoses132: (1) greater hyperintensity on DWI than on FLAIR, (2) subcortical GM hyperintensities on DWI associated with

ADC hypointensity (i.e., restricted diffusion), (3) cortical GM involvement usually not isolated to only limbic structures, and (4) a characteristic neuroanatomic pattern of involvement. The imaging hallmark of variant CJD is a bilateral FLAIR/T2 hyperintensity of the posterior portion of the thalamus (pulvinar), which is brighter than the anterior putamen (pulvinar sign).29,145 FLAIR hyperintensity of the dorsomedial thalamic nuclei (“double hockey stick” sign) is also common in vCJD, although this finding is found in other forms of CJD.145

Normal-Pressure Hydrocephalus The term normal-pressure hydrocephalus (NPH) was introduced in 1965 by Adams and Hakim and their colleagues1,64 to describe a condition of hydrocephalus presenting with a clinical triad of symptoms (gait instability, mental deterioration, urinary incontinence) associated with ventricular enlargement and normal CSF pressure. Primary (idiopathic) NPH (iNPH) is distinguished from secondary NPH (sNPH), whose common causes are subarachnoid hemorrhage (23%), meningitis (4.5%), and traumatic brain injury (12.5%).66 A common feature of iNPH and sNPH is that neither involves any obstruction to the flow of CSF within the ventricular system of the brain.

http://radiologyebook.com

CHAPTER 15  Neurodegenerative Disorders

A

B

C

D

E

F

479

FIG 15-7  A to C, Axial T2-weighted MRIs obtained in a 71-year-old patient with normal-pressure hydrocephalus (NPH), showing triventricular hydrocephalus with ballooning of the frontal horns and cortical effacement. There is little or no sign of cortical atrophy (A and B). There is no evidence of dilation of the parahippocampal fissures or hippocampal atrophy (C). D, Coronal T1-weighted MRI showing the sharp angle of the frontal horns characteristic of NPH. E and F, Sagittal T1-weighted MRIs showing upward bowing of the corpus callosum (E) and distal widening of the aqueduct of Sylvius (F). (From Tarnaris A, et  al: Noninvasive biomarkers in normal pressure hydrocephalus: Evidence for the role of neuroimaging. J Neurosurg 110:837–851, 2009.)

CT or, preferably, MRI is used to evaluate the presence of the cardinal findings in iNPH, which are disproportionate widening of the ventricles in comparison to the cerebral sulci, upward bowing of the corpus callosum or small corpus callosum angle, and dilatation of   the temporal horns65 (Fig. 15-7). Specifically, ventricular widening is defined as an Evan’s index (frontal horns to inner table skull) greater than 0.3.112 A coronal section at the level of the posterior commissure reveals a narrow subarachnoid space surrounding the outer surface of the brain (a “tight convexity”) and narrow medial cisterns. The third ventricle is usually enlarged as well, while the fourth ventricle may or may not be enlarged. As a result of transependymal CSF diapedesis, iNPH can be associated with periventricular lucencies141 (i.e., relatively thin, slightly blurry zone of increased signal best seen on T2-weighted or FLAIR images) extending homogeneously around the whole ventricular system, which have to be differentiated from periventricular diffuse, smooth hyperintensities extended to the corona radiata and centrum semiovale due to microvascular disease. Symptoms can be relieved by ventricular shunting, with the greatest improvement usually seen in gait.66 The typical patient with iNPH who is likely to respond to a shunt exhibits a significant motor deficit and gait disturbance in the absence of cognitive deficits. Radiologically,

absence of WM changes,18 Evan’s index greater than 0.4, and absence of cortical atrophy have all been associated with good response. Enlargement of temporal horns has been reported as a poor prognostic sign.66,130

WHITE MATTER DISEASES MRI is exquisitely sensitive to the detection of WM disease. As in the atrophic changes described in the first part of this chapter, understanding the MRI characteristics and underlying pathophysiology of WM abnormalities can allow one to arrive at the accurate diagnosis. Because this chapter focuses on neurodegenerative diseases, it covers only the limited number of WM diseases that apply to this category.

Vascular Dementia Vascular dementia is the second most common cause of dementia in older adults, representing 10% to 20% of cases in patients older than 65 years.8 Clinically these patients may present with sudden cognitive change and a clinical course marked by periods of acute deterioration and spontaneous improvement, often with motor impairment. These patients present radiologically with focally dilated fissures and multiple

http://radiologyebook.com

480

PART II  CT and MR Imaging of the Whole Body

areas of focal cortical or subcortical volume loss. Multiple subcortical infarcts or single strategic infarcts (e.g., involving parts of the thalamus, caudate, cingulate gyrus, MTL, or internal capsule) can also result in dementia. Cerebrovascular disease can be detected by CT and structural MRI.8 Although both modalities perform relatively well in depicting large vessel infarcts, MRI is more sensitive to subtle small vessel vascular changes than CT. T2-weighted and FLAIR sequences are highly sensitive for detecting major strokes as well as small strategic infarcts and small vessel ischemic WM damage. FLAIR performs less well than T2-weighted sequences in detecting thalamic infarcts.9 Extensive WM changes visible as diffuse hyperintense abnormalities on T2 and FLAIR, predominantly involving the periventricular and deep WM but relatively sparing the U-fibers, are the imaging correlate of Binswanger’s disease.42 Marked hypointensity on T1-weighted images usually represents tissue destruction in the presence of a complete large vessel infarct, whereas WM changes are usually not prominently hypointense on T1-weighted images. Finally, lacunar infarcts are focal complete infarcts of deep small vessels, which are hyperintense on T2-weighted images and markedly hypointense on T1 and FLAIR images.42 On FLAIR, lacunae are often surrounded by a hyperintense rim. According to the NINDS-AIREN (National Institute of Neurological Disorders and Stroke and Association Internationale pour la Recherché et l’Enseignement en Neurosciences) international criteria,113 structural brain imaging is an essential element for the diagnosis of vascular dementia, and without it vascular dementia will be “possible” at best. In addition the operational radiologic definitions for the NINDS-AIREN criteria provided indications on the topography and severity of vascular lesions.129 Bilateral infarcts in the area of the anterior cerebral artery, posterior cerebral artery, association areas, or in the watershed regions are thought to be causative of large vessel vascular dementia.129 Extensive WM lesions involving at least 25% of the WM, or multiple basal ganglia, thalamic, and frontal WM lacunar infarcts, or bilateral thalamic lesions are considered relevant radiologic lesions associated with small vessel vascular dementia.129 Research criteria to specifically diagnose subcortical ischemic vascular dementia have also been proposed42: the presence of extensive periventricular and deep WM lesions and lacunar infarcts in the deep GM, or multiple lacunae in the deep GM and at least moderate WM lesions in the absence of cortical and/or cortical-subcortical (nonlacunar) territorial infarcts, watershed infarcts, hemorrhages, and other specific causes of WM lesions. Different methods can be used to measure the extent of WM changes to diagnose subcortical ischemic vascular dementia. Visual rating of WM hyperintensities is relatively easy, and several scales43,118,134 are available with good reproducibility. Widely used visual rating scales include those introduced by Fazekas and Scheltens. The Fazekas scale rates WMHs in the periventricular and subcortical region, combined, on a 0- to 3-point scale.43 The Scheltens scale rates WMHs separately in the periventricular region on a 0- to 6-point scale and in the subcortical region on a 0- to 24-point scale on the basis of the size and number of lesions. This scale also includes ratings for basal ganglia and infratentorial areas.118 Figure 15-8 shows examples of brain scans graded according to the Fazekas scale. In addition to these scales, Wahlund and coworkers introduced a scale applicable to both CT and MRI, which is easy to use and compare between modalities.134 Cerebral microbleeds (or microhemorrhages) are small, rounded, dotlike lesions of low signal intensity in the brain that can be observed on T2*-weighted images such as gradient echo.44 Susceptibilityweighted imaging has considerably increased microbleed detection rates compared with gradient echo sequences.56,62 The sensitivity to

detect microbleeds is also dependent on slice thickness and magnetic field strength.98 Microbleeds in deep brain regions are most likely to be associated with vasculopathy due to hypertension, whereas their distribution is mostly lobar in specific disorders such as sporadic cerebral amyloid angiopathy (CAA).61 Sporadic CAA is the most common cause of lobar intracerebral hemorrhages in the elderly and results from cerebrovascular deposition of amyloid β protein in the media and adventitia of small and medium-sized vessels of the superficial layers of the cerebral cortex and leptomeninges, with sparing of the deep GM nuclei.131 According to a set of validated criteria (Boston criteria),76 a diagnosis of probable CAA can be reached in elderly patients with  at least two acute or chronic lobar hemorrhagic lesions (including microbleeds) in the absence of another definite cause of intracerebral hemorrhage. Although an increased WM lesion load suggests vascular disease, particularly in combination with lesions in the basal ganglia, a critical clinical challenge in subjects with vascular lesions is determining  the relationship of cerebrovascular disease and cognitive symptoms. To appropriately diagnose vascular cognitive impairment or dementia, there should be a clear relationship in the severity and pattern of cognitive impairment and the presence of vascular lesions with neuroimaging.113 However, this association can be confounded by the frequent co-occurrence of depression on a cerebrovascular basis, as well as by the frequent coexistence with neurodegenerative diseases, especially AD.121 Small vessel disease is frequently observed in MRI scans of AD patients in the form of WMHs, lacunae, and microbleeds. Several studies have indicated that the prevalence of microbleeds in AD ranges from 15% to 32%.31,104 Microbleeds are also considerably more frequent in AD compared with other neurodegenerative dementia.31

Cerebral Autosomal-Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy (CADASIL) Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is a hereditary disease of   small cerebral arteries that affects middle-aged adults, associated with a mutation of the Notch3 gene.23 CADASIL typically presents with cerebral ischemic events, cognitive impairment, and (usually migraine-type) headache. The two most distinctive MRI features suggestive of CADASIL are the presence of T2 hyperintensity of the temporal pole, the U-fibers at the vertex and external capsule or insular region, and multiple microhemorrhagic foci (basal ganglia, internal capsule, thalamus, and pons), which can be seen on gradient echo scans.23

HUNTINGTON’S DISEASE Huntington’s disease (HD) is a degenerative neurologic disease that is inherited in an autosomal dominant fashion with complete penetrance.115 The causative gene is located on chromosome 4 and contains an excessive number of trinucleotide CAG repeats. Patients usually become symptomatic before age 50 and tend to present with abnormalities of affect and personality. With time, dementia becomes evident, accompanied by gradual disintegration of motor control and emergence of choreoathetoid movements. Occasionally, rigidity rather than chorea is the most salient motor abnormality. Neuropathologically, the most conspicuous finding in patients with HD is atrophy of the basal ganglia, generally proceeding medial to lateral and dorsal to ventral.133 In early cases, neuronal loss is generally first appreciated in the head of the caudate,133 appearing as focal enlargement of frontal horns. On MRI, atrophy of the caudate is

http://radiologyebook.com

CHAPTER 15  Neurodegenerative Disorders Fazekas 1

Fazekas 2

481

Fazekas 3

Nature Reviews Neurology

FIG 15-8  Axial FLAIR images illustrating Fazekas scores. Fazekas 0 (not shown): no white matter hyperintensities. Fazekas 1: focal or punctate lesions (single lesions ≤ 9 mm; grouped lesions < 20 mm); Fazekas 2: beginning confluent lesions (single lesions 10-20 mm; grouped lesions > 20 mm in any diameter; no more than connecting bridges between individual lesions); Fazekas 3: confluent lesion (single lesions or confluent areas of hyperintensity ≥ 20 mm in any diameter). (From Prins ND, Scheltens P: White matter hyperintensities, cognitive impairment and dementia: An update. Nat Rev Neurol 11:157–165, 2015.)

detectable up to 10 years prior to onset of the motor clinical manifestation and worsens as the disease progresses122 (Fig. 15-9). Degenerative changes can also affect the frontal and temporal cortices.122 In addition, GM reductions can be seen in the hypothalamus and opercular cortex.115 A number of investigators have reported that in addition to atrophic changes, patients with HD present with areas of abnormal T2 signal hyperintensity in the basal ganglia.115

PARKINSON’S DISEASE Idiopathic PD is by far the most common cause of parkinsonism and should always be the diagnosis if a definite secondary cause cannot be identified. Definitive diagnosis of idiopathic PD requires histologic demonstration of intraneuronal Lewy body inclusions in the

substantia nigra (SN) pars compacta.54 In most cases the diagnosis of probable PD can be made on clinical grounds, and no ancillary investigations are needed. However, in early PD the full triad of clinical symptoms and signs (bradykinesia, tremor at rest and rigidity) may not yet be manifested.88 CT scans can show nonspecific atrophy with enlarged ventricles and sulci. Conventional MRI at 1.5 T with visual assessment of T2- and T1-weighted imaging does not reveal disease-specific abnormalities in PD and, particularly in the early phases, the MRI appears normal.120 Its main clinical role is to exclude subcortical vascular pathology, rare secondary causes of parkinsonism (e.g., Wilson’s disease, NPH, or tumors, granulomas, or calcification of basal ganglia), and in discriminating atypical parkinsonian syndromes. With disease progression, findings remain nonspecific, although a signal decrease in the SN on

http://radiologyebook.com

482

PART II  CT and MR Imaging of the Whole Body Control

A

Premanifest stage

40

40

Clinically manifest stage

45

50

Age (years) Gray matter

White matter

Pre-HD

Early HD

B FIG 15-9  MRI scans of control individuals, mutation carriers, and patients with Huntington’s disease (HD). A, Coronal images from controls and mutation carriers during the transition from premanifest stage to symptomatic disease over 10 years. Over this period, progressive atrophy of the striatum, whole brain, and white matter is evident and accompanied by increased lateral ventricle sizes. B, Voxel-based morphometric images from the TRACK-HD study showing regions for which the gray matter and white matter atrophy rates over 12 months were significantly different from individuals in the control group. (From Weir DW, et al: Development of biomarkers for Huntington’s disease. Lancet Neurol 10:573–590, 2011.)

proton density and T2-weighted images, as well as mild frontal and hippocampal atrophy, has been described.144 Nuclear medicine techniques including PET and SPECT allow confirmation of the clinical suspicion of PD by demonstrating presynaptic loss of the dopamine transporter. In PD, tracers of presynaptic dopaminergic function show a characteristic pattern of asymmetric reduction of striatal uptake, more pronounced contralateral to the clinically more affected side of the body33 (see Fig. 15-5). Typically the nigrostriatal dopaminergic projections to the posterior (dorsal-caudal) putamen are affected earlier and more severely than those to the caudate nucleus. The caudal-rostral gradient is attributed to (1) predominant degeneration of the ventrolateral portion of the SN that projects to the putamen and (2) relative preservation of the dorsomedial SN that projects to the caudate nucleus.60 Reductions in striatal tracer uptake correlate with pathologic measures (i.e., cell loss) and degree of motor disability, particularly bradykinesia and to a lesser extent rigidity; progression of the disease can be monitored using these tracers in longitudinal studies.16 In PD, FDG PET reveals a characteristic profile characterized by increased metabolism in the putamen/globus pallidus, thalamus, cerebellum, pons, and sensorimotor cortex and reduced metabolism in the lateral frontal, paracentral, and parietooccipital areas.41 The PD-related covariance pattern linearly correlates with clinical motor ratings (mainly with bradykinesia and rigidity rather than tremor) and increases with disease progression.

Atypical Parkinsonisms Atypical parkinsonisms can be caused by different neurodegenerative syndromes, such as the parkinsonian variant of multiple system atrophy (MSA-P), PSP, and CBS, which are frequently accompanied by denervation of both pre- and postsynaptic dopaminergic pathways and are usually levodopa resistant. MSA is a progressive neurodegenerative disorder characterized clinically by autonomic dysfunction, parkinsonism, cerebellar ataxia, and pyramidal signs.55 MSA can be classified into two subgroups: a cerebellar (MSA-C) and a parkinsonian (MSA-P) variant.55 The histologic hallmarks of MSA are α-synuclein-positive glial cytoplasmic inclusions in the oligodendroglia, which are required for the diagnosis of definite MSA.55 PSP and CB degeneration are pathologically distinct causes of progressive atypical parkinsonism, characterized by hyperphosphorylated tau aggregates in the brain accompanied by neuronal loss and gliosis in a characteristic distribution.35 The large majority of PSP patients presents with Richardson’s syndrome, also known as PSP syndrome,81 characterized by postural instability leading to backward falls within the first year of symptom onset, axial rigidity, progressive vertical ophthalmoplegia, dementia, and personality changes. CBD pathology can cause multiple different neurologic syndromes, including CBS, PSP syndrome, bvFTD, or PPA. CBS includes a mixed movement disorder (e.g., levodopa-unresponsive rigidity associated with apraxia, dystonia, myoclonus, and alien limb) and impaired cognition.6 In the early stage of the diseases, differential

http://radiologyebook.com

CHAPTER 15  Neurodegenerative Disorders diagnosis among different atypical parkinsonian syndromes, and each of them with PD, is challenging because of similarity of symptoms and lack of preclinical markers of the disease. MRI can aid in the classification of these patients, and several neuroimaging signs have been described for differentiating atypical parkinsonism syndromes from PD. A number of conventional MRI findings has been described as suggestive of MSA-P. These include a posterolateral putaminal decreased signal intensity on T2-weighted images, mainly due to iron deposition, covered by a rim of increased signal, mainly due to gliosis (Fig. 15-10). If pontocerebellar degeneration is present, lateral and longitudinal pontine fibers become evident as high signal manifesting as the “hot cross bun” sign (i.e., a cruciform linear area of high signal on T2-weighted images in the pons, with tiny round darker areas within the checkerboard of the cross; see Fig. 15-10). Cerebellar and pontine atrophy can be visually detected, along with hyperintensity of

A

C

483

the MCP (see Fig. 15-10). Although putaminal atrophy appears to discriminate MSA-P from PD, T2 putaminal hyperintensity and a putaminal hyperintense rim may also occur in the latter condition.120 Specificity of those abnormalities in differentiating MSA-P from PD and healthy controls is considered to be relatively high; in contrast, sensitivity seems to be insufficient, especially in the early stages of disease.120 PSP is characterized by a significant dorsal midbrain atrophy, especially of the anteroposterior diameter. Indirect signs of midbrain atrophy are the “penguin silhouette” or “hummingbird” signs on sagittal views120 (Fig. 15-11), where the shapes of the midbrain tegmentum (i.e., the bird’s head) and pons (i.e., the bird’s body) resemble a lateral view of a standing king penguin or hummingbird, and an abnormal superior midbrain profile (flat or concave vs. convex aspect in healthy subjects) is present. Other MRI findings in PSP patients are enlargement of the third ventricle, signal increase on T2-weighted images of

B

E

D

FIG 15-10  Axial T2* gradient echo (A) and T2-weighted turbo-spin echo (B) MRIs show symmetric, abnormally low signal intensity of the putamen in two different patients with multiple system atrophy–parkinsonian type (MSA-P), which is surrounded by a regular hyperintense rim in B. C, “Hot cross bun” sign. D, Axial T2-weighted turbo-spin echo shows subtle diffuse hyperintensity of the white matter of the pons, middle cerebellar peduncles, and cerebellar white matter, with sparing of the corticospinal tracts in the basis pontis, which creates the “hot cross bun” sign (box). Note the marked thinning of the basis pontis and middle cerebellar peduncles. E, Sagittal T1-weighted gradient echo demonstrates diffuse atrophy of the brainstem (more pronounced in the basis pontis) and thinning of the folia of the cerebellar vermis. (From Mascalchi M, et al: Movement disorders: Role of imaging in diagnosis. J Magn Reson Imaging 35:239–256, 2012.)

http://radiologyebook.com

484

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 15-11  A, Hummingbird. B, Sagittal T1-weighted MRI from a patient with progressive supranuclear palsy shows the characteristic thinning of the anteroposterior diameter and the abnormal shape of the rostral midbrain tegmentum, exhibiting a concave upper profile and the posterior and central portions of floor of the third ventricle resembling the hummingbird’s long, thin, sharp beak. C, Axial T2-weighted image shows thinning of the midbrain and hypointensity of the substantia nigra. (From Mascalchi M, et al: Movement disorders: Role of imaging in diagnosis. J Magn Reson Imaging 35:239–256, 2012.)

the midbrain and inferior olives, atrophy and increased signal of the superior cerebellar peduncles on T2-weighted and FLAIR images, hypointense putamen on T2-weighted images (due to increased iron content), and frontal and parietal lobe atrophy. Only a few conventional MRI studies have assessed CBS patients. Despite this, several suggestive findings have been described, including a characteristically asymmetric frontal and/or parietal atrophy (including the pre- and postcentral gyri and central sulcus), with less frequent involvement of the temporal lobe (sometimes even global and symmetric), and a subcortical gliosis in the atrophic cortical gyri seen as high intensity on T2-weighted images. Unfortunately, none of these MRI abnormalities is considered to be of clearly diagnostic or even pathognomonic relevance for CBS, mainly because of the pathologic heterogeneity of this syndrome. SPECT and PET tracers investigating presynaptic dopaminergic function cannot differentiate PD from atypical parkinsonisms, given a similar nigrostriatal involvement in these conditions, although asymmetry of binding loss tends to be more pronounced in PD. FDG PET studies can contribute in the diagnostic workup between atypical parkinsonisms and PD. In contrast to PD, atypical parkinsonian syndromes show reduced lentiform nucleus glucose metabolism, which continues to decline as disease progresses.40 In addition, FDG PET shows bilateral hypometabolism of the putamen, brainstem, and cerebellum in MSA-P and hypometabolism of the brainstem, thalamus, and frontal cortex in PSP.40 The FDG PET pattern of CBS is characterized by a unilateral hypometabolism contralateral to the clinically affected side, with a predominance in the parietotemporal, prefrontal, cingular, and motor cortices as well as in the caudate and thalamus.52,68

AMYOTROPHIC LATERAL SCLEROSIS Motor neuron diseases (MND) are progressive neurodegenerative conditions that lead to severe disability and death.74 There is considerable variability in presentation, clinical course, and prognosis. These conditions are divided into several different clinical subtypes.74 ALS is the

most common (90% of all MND cases) and is characterized by upper (UMN) and lower motor neuron involvement of the bulbar, upper, and lower limb territories. Although a detailed clinical assessment remains the basis of the evaluation of patients suspected of having ALS, current international diagnostic criteria17 and European guidelines46 recommend that “all patients suspected of having ALS, where a plausible alternative unifying neuroanatomical explanation exists, should undergo an MRI of either or both the brain and whole cord depending on the clinical presentation.” Indeed, routine brain and spinal cord MRI can be useful in excluding several ALS mimic syndromes, including cerebral lesions (e.g., multiple sclerosis and cerebrovascular disease), skull base lesions, cervical spondylotic myelopathy, other myelopathy (e.g., foramen magnum lesions, intrinsic and extrinsic tumors, syringomyelia), conus lesions, and thoracolumbar-sacral radiculopathy. Corticospinal tract (CST) hyperintensities on T2-weighted, proton density-weighted, and FLAIR images are frequently found in ALS patients.24,125 CST hyperintensities, which are best followed on coronal scans, have been reported mostly bilaterally and are most frequently seen in the caudal portion of the posterior limb of the internal capsule. They typically extend downward to the ventral portion of the brainstem and less consistently upward through the corona radiata. Such lesions may occur more often in younger patients with greater disability.70 However, the reported frequency of conventional MRI abnormalities in ALS patients is very heterogeneous, ranging from 15% to 76%. More importantly, increased CST signal intensity has also been described in healthy individuals and, strikingly, in patients with other conditions such as Krabbe disease, X-linked Charcot-Marie-Tooth neuropathies, adrenomyeloneuropathy, and after hepatic transplantation.46 In ALS patients, the precentral cortex can present a low signal intensity (hypointense rim) on T2-weighted images.27,125 This so-called ribbon-like hypointensity is sharply contrasted by the hyperintense signal of CSF in the adjacent sulci. Over the past 15 years there have been significant advances in the identification of advanced neuroimaging patterns in MND.28 International consensus was reached about essential and desirable protocols

http://radiologyebook.com

CHAPTER 15  Neurodegenerative Disorders

485

B

A

FIG 15-12  Sagittal (A) and axial (B) T1-weighted images show atrophy of the medulla and cervical spinal cord (arrow) with normal bulk of the pons and cerebellar vermis in a 13-year-old girl with Friedreich’s ataxia. (From Mascalchi M: Spinocerebellar ataxias. Neurol Sci 29[Suppl 3]:311–313, 2008.)

for MRI for studies in ALS,126 which are of the utmost importance in multicenter and longitudinal settings. In the absence of definitive biomarkers of UMN involvement, a significant cortical thinning of the precentral gyrus, damage to the CST and corpus callosum assessed using DT MRI, and altered N-acetylaspartate levels in the primary motor cortex and CST should be considered suggestive of MND and hold promise for detecting UMN involvement before clinical symptoms become apparent.2

FRIEDREICH’S ATAXIA Friedreich’s ataxia is the most common inherited ataxia affecting  children and young adults. It is commonly due to an expanded GAA trinucleotide repeat in both the alleles of a gene encoding the mitochondrial protein frataxin.78 Pathologic hallmarks of Friedreich’s ataxia are neuronal loss and shrinkage of the spinal ganglion cells and Clarke’s column of the spinal cord, with degeneration of peripheral nerves and spinocerebellar gracilis and cuneatus tracts.78 The cerebellar cortex is spared, whereas there is cell loss in the dentate nuclei and gliosis of the cerebellar WM.78 Cerebral cortical GM abnormalities are usually limited to the primary motor cortex, although damage to the visual cortex can occur. Hypertrophic cardiomyopathy can be the cause of death in advanced phases. MRI reveals a pattern of spinal atrophy85,86 in which the neurons of the sensory ganglion of the spinal nerves and the Clarke’s columns in the spinal GM are predominantly affected (Fig. 15-12). MRI shows atrophy of the spinal cord extending to the medulla, with a slight widening of the fourth ventricle. Symmetric signal changes in the posterior and lateral columns of the cervical spinal cord are found. Atrophy of the cerebellar or cerebral cortical GM is not remarkable in Friedreich’s ataxia, whereas volume loss of the deep cerebellar WM, which correlates with severity of the clinical deficits, was reported.34 Atrophy of the posterior cingulate gyrus, paracentral lobule, and middle frontal gyrus was also found.49 Acknowledgment: We would like to thank Drs. A.I. Holodny, A.E. George, M.J. de Leon, S. Karimi, and J. Golomb for the earlier version of the chapter.

REFERENCES 1. Adams RD, Fisher CM, Hakim S, et al: Symptomatic occult hydrocephalus with “normal” cerebrospinal-fluid pressure. A treatable syndrome. N Engl J Med 273:117–126, 1965. 2. Agosta F, Al-Chalabi A, Filippi M, et al: The El Escorial criteria: Strengths and weaknesses. Amyotroph Lateral Scler Frontotemporal Degener 16:1–7, 2015. 3. Albin RL, Minoshima S, D’Amato CJ, et al: Fluoro-deoxyglucose positron emission tomography in diffuse Lewy body disease. Neurology 47:462–466, 1996. 4. Almeida OP, Burton EJ, McKeith I, et al: MRI study of caudate nucleus volume in Parkinson’s disease with and without dementia with Lewy bodies and Alzheimer’s disease. Dement Geriatr Cogn Disord 16:57–63, 2003. 5. Apostolova LG, Thompson PM, Green AE, et al: 3D comparison of low, intermediate, and advanced hippocampal atrophy in MCI. Hum Brain Mapp 31:786–797, 2010. 6. Armstrong MJ, Litvan I, Lang AE, et al: Criteria for the diagnosis of corticobasal degeneration. Neurology 80:496–503, 2013. 7. Ballmaier M, O’Brien JT, Burton EJ, et al: Comparing gray matter loss profiles between dementia with Lewy bodies and Alzheimer’s disease using cortical pattern matching: diagnosis and gender effects. Neuroimage 23:325–335, 2004. 8. Barkhof F, Fox NC, Bastos Leite AJ, et al: Neuroimaging in dementia, ed 1, Berlin-Heidelberg, 2011, Springer. 9. Bastos Leite AJ, van Straaten EC, Scheltens P, et al: Thalamic lesions in vascular dementia: Low sensitivity of fluid-attenuated inversion recovery (FLAIR) imaging. Stroke 35:415–419, 2004. 10. Beyer MK, Larsen JP, Aarsland D: Gray matter atrophy in Parkinson disease with dementia and dementia with Lewy bodies. Neurology 69:747–754, 2007. 11. Bloudek LM, Spackman DE, Blankenburg M, et al: Review and meta-analysis of biomarkers and diagnostic imaging in Alzheimer’s disease. J Alzheimers Dis 26:627–645, 2011. 12. Boccardi M, Sabattoli F, Laakso MP, et al: Frontotemporal   dementia as a neural system disease. Neurobiol Aging 26:37–44, 2005. 13. Bohnen NI, Djang DS, Herholz K, et al: Effectiveness and safety of 18F-FDG PET in the evaluation of dementia: A review of the recent literature. J Nucl Med 53:59–71, 2012.

http://radiologyebook.com

486

PART II  CT and MR Imaging of the Whole Body

14. Bozzali M, Filippi M, Magnani G, et al: The contribution of voxel-based morphometry in staging patients with mild cognitive impairment. Neurology 67:453–460, 2006. 15. Brambati SM, Rankin KP, Narvid J, et al: Atrophy progression in semantic dementia with asymmetric temporal involvement: A tensor-based morphometry study. Neurobiol Aging 30:103–111, 2009. 16. Brooks DJ: Imaging approaches to Parkinson disease. J Nucl Med 51:596–609, 2010. 17. Brooks BR, Miller RG, Swash M, et al: El Escorial revisited: Revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord 1:293–299, 2000. 18. Bugalho P, Alves L: Normal-pressure hydrocephalus: White matter lesions correlate negatively with gait improvement after lumbar puncture. Clin Neurol Neurosurg 109:774–778, 2007. 19. Burton EJ, Barber R, Mukaetova-Ladinska EB, et al: Medial temporal lobe atrophy on MRI differentiates Alzheimer’s disease from dementia with Lewy bodies and vascular cognitive impairment: A prospective study with pathological verification of diagnosis. Brain 132:195–203, 2009. 20. Burton EJ, Karas G, Paling SM, et al: Patterns of cerebral atrophy in dementia with Lewy bodies using voxel-based morphometry. Neuroimage 17:618–630, 2002. 21. Burton EJ, McKeith IG, Burn DJ, et al: Cerebral atrophy in Parkinson’s disease with and without dementia: A comparison with Alzheimer’s disease, dementia with Lewy bodies and controls. Brain 127:791–800, 2004. 22. Canu E, Frisoni GB, Agosta F, et al: Early and late onset Alzheimer’s disease patients have distinct patterns of white matter damage. Neurobiol Aging 33:1023–1033, 2013. 23. Chabriat H, Joutel A, Dichgans M, et al: Cadasil. Lancet Neurol 8:643–653, 2009. 24. Charil A, Corbo M, Filippi M, et al: Structural and metabolic changes   in the brain of patients with upper motor neuron disorders: A multiparametric MRI study. Amyotroph Lateral Scler 10:269–279, 2009. 25. Chetelat G, Desgranges B, de la Sayette V, et al: Mild cognitive impairment: Can FDG-PET predict who is to rapidly convert to Alzheimer’s disease? Neurology 60:1374–1377, 2003. 26. Chetelat G, Landeau B, Eustache F, et al: Using voxel-based morphometry to map the structural changes associated with rapid conversion in MCI: A longitudinal MRI study. Neuroimage 27:934–946, 2005. 27. Cheung G, Gawel MJ, Cooper PW, et al: Amyotrophic lateral sclerosis: Correlation of clinical and MR imaging findings. Radiology 194:263– 270, 1995. 28. Chiò A, Pagani M, Agosta F, et al: Neuroimaging in amyotrophic lateral sclerosis: Systematic insight into structural and functional changes. Lancet Neurol 13:1228–1240, 2014. 29. Collie DA, Summers DM, Sellar RJ, et al: Diagnosing variant Creutzfeldt-Jakob disease with the pulvinar sign: MR imaging findings in 86 neuropathologically confirmed cases. AJNR Am J Neuroradiol 24:1560–1569, 2003. 30. Collins SJ, Sanchez-Juan P, Masters CL, et al: Determinants of diagnostic investigation sensitivities across the clinical spectrum of sporadic Creutzfeldt-Jakob disease. Brain 129:2278–2287, 2006. 31. Cordonnier C, van der Flier WM, Sluimer JD, et al: Prevalence and severity of microbleeds in a memory clinic setting. Neurology 66:1356–1360, 2006. 32. Cousins DA, Burton EJ, Burn D, et al: Atrophy of the putamen in dementia with Lewy bodies but not Alzheimer’s disease: An MRI study. Neurology 61:1191–1195, 2003. 33. Cummings JL, Henchcliffe C, Schaier S, et al: The role of dopaminergic imaging in patients with symptoms of dopaminergic system neurodegeneration. Brain 134:3146–3166, 2011. 34. Della Nave R, Ginestroni A, Tessa C, et al: Brain white matter tracts degeneration in Friedreich ataxia. An in vivo MRI study using tract-based spatial statistics and voxel-based morphometry. Neuroimage 40:19–25, 2008.

35. Dickson DW, Kouri N, Murray ME, et al: Neuropathology of frontotemporal lobar degeneration-tau (FTLD-tau). J Mol Neurosci 45:384–389, 2011. 36. Diehl J, Grimmer T, Drzezga A, et al: Cerebral metabolic patterns at early stages of frontotemporal dementia and semantic dementia. A PET study. Neurobiol Aging 25:1051–1056, 2004. 37. Drzezga A, Grimmer T, Riemenschneider M, et al: Prediction of individual clinical outcome in MCI by means of genetic assessment and (18)F-FDG PET. J Nucl Med 46:1625–1632, 2005. 38. Dubois B, Feldman HH, Jacova C, et al: Revising the definition of Alzheimer’s disease: A new lexicon. Lancet Neurol 9:1118–1127, 2010. 39. Dubois B, Feldman HH, Jacova C, et al: Research criteria for the diagnosis of Alzheimer’s disease: revising the NINCDS-ADRDA criteria. Lancet Neurol 6:734–746, 2007. 40. Eckert T, Barnes A, Dhawan V, et al: FDG PET in the differential diagnosis of parkinsonian disorders. Neuroimage 26:912–921, 2005. 41. Eidelberg D, Moeller JR, Dhawan V, et al: The metabolic anatomy of Parkinson’s disease: Complementary [18F]fluorodeoxyglucose and [18F]fluorodopa positron emission tomographic studies. Mov Disord 5:203–213, 1990. 42. Erkinjuntti T, Inzitari D, Pantoni L, et al: Research criteria for subcortical vascular dementia in clinical trials. J Neural Transm Suppl 59:23–30, 2000. 43. Fazekas F, Chawluk JB, Alavi A, et al: MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol 149:351–356, 1987. 44. Fazekas F, Kleinert R, Roob G, et al: Histopathologic analysis of foci of signal loss on gradient-echo T2*-weighted MR images in patients with spontaneous intracerebral hemorrhage: evidence of microangiopathyrelated microbleeds. AJNR Am J Neuroradiol 20:637–642, 1999. 45. Fazekas F, Ropele S, Enzinger C, et al: MTI of white matter hyperintensities. Brain 128:2926–2932, 2005. 46. Filippi M, Agosta F, Abrahams S, et al: EFNS guidelines on the use of neuroimaging in the management of motor neuron diseases. Eur J Neurol 17:526–e20, 2010. 47. Filippi M, Agosta F, Barkhof F, et al: EFNS task force: The use of neuroimaging in the diagnosis of dementia. Eur J Neurol 19:e131–e140, 1487–1501, 2012. 48. Fox NC, Crum WR, Scahill RI, et al: Imaging of onset and progression of Alzheimer’s disease with voxel-compression mapping of serial magnetic resonance images. Lancet 358:201–205, 2001. 49. Franca MC, Jr, D’Abreu A, Yasuda CL, et al: A combined voxel-based morphometry and 1H-MRS study in patients with Friedreich’s ataxia.   J Neurol 256:1114–1120, 2009. 50. Frisoni GB, Pievani M, Testa C, et al: The topography of grey matter involvement in early and late onset Alzheimer’s disease. Brain 130:720–730, 2007. 51. Galton CJ, Patterson K, Graham K, et al: Differing patterns of temporal atrophy in Alzheimer’s disease and semantic dementia. Neurology 57:216–225, 2001. 52. Garraux G, Salmon E, Peigneux P, et al: Voxel-based distribution of metabolic impairment in corticobasal degeneration. Mov Disord 15:894–904, 2000. 53. Geschwind MD, Potter CA, Sattavat M, et al: Correlating DWI MRI with pathologic and other features of Jakob-Creutzfeldt disease. Alzheimer Dis Assoc Disord 23:82–87, 2009. 54. Gibb WR, Lees AJ: The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 51:745–752, 1988. 55. Gilman S, Wenning GK, Low PA, et al: Second consensus statement on the diagnosis of multiple system atrophy. Neurology 71:670–676, 2008. 56. Goos JD, van der Flier WM, Knol DL, et al: Clinical relevance of improved microbleed detection by susceptibility-weighted magnetic resonance imaging. Stroke 42:1894–1900, 2011. 57. Gorno-Tempini ML, Brambati SM, Ginex V, et al: The logopenic/ phonological variant of primary progressive aphasia. Neurology 71:1227–1234, 2008.

http://radiologyebook.com

CHAPTER 15  Neurodegenerative Disorders 58. Gorno-Tempini ML, Dronkers NF, Rankin KP, et al: Cognition and anatomy in three variants of primary progressive aphasia. Ann Neurol 55:335–346, 2004. 59. Gorno-Tempini ML, Hillis AE, Weintraub S, et al: Classification of primary progressive aphasia and its variants. Neurology 76:1006–1014, 2011. 60. Goto S, Hirano A, Matsumoto S: Subdivisional involvement of nigrostriatal loop in idiopathic Parkinson’s disease and striatonigral degeneration. Ann Neurol 26:766–770, 1989. 61. Greenberg SM, Vernooij MW, Cordonnier C, et al: Cerebral microbleeds: A guide to detection and interpretation. Lancet Neurol 8:165–174, 2009. 62. Haacke EM, DelProposto ZS, Chaturvedi S, et al: Imaging cerebral amyloid angiopathy with susceptibility-weighted imaging. AJNR Am J Neuroradiol 28:316–317, 2007. 63. Hachinski VC, Potter P, Merskey H: Leuko-araiosis. Arch Neurol 44:21–23, 1987. 64. Hakim S, Adams RD: The special clinical problem of symptomatic hydrocephalus with normal cerebrospinal fluid pressure. Observations on cerebrospinal fluid hydrodynamics. J Neurol Sci 2:307–327, 1965. 65. Hashimoto M, Ishikawa M, Mori E, et al: Study of Ioni. Diagnosis of idiopathic normal pressure hydrocephalus is supported by MRI-based scheme: A prospective cohort study. Cerebrospinal Fluid Res 7:18, 2010. 66. Hebb AO, Cusimano MD: Idiopathic normal pressure hydrocephalus: A systematic review of diagnosis and outcome. Neurosurgery 49:1166– 1184, discussion 84–86, 2001. 67. Higuchi M, Tashiro M, Arai H, et al: Glucose hypometabolism and neuropathological correlates in brains of dementia with Lewy bodies. Exp Neurol 162:247–256, 2000. 68. Hosaka K, Ishii K, Sakamoto S, et al: Voxel-based comparison of regional cerebral glucose metabolism between PSP and corticobasal degeneration. J Neurol Sci 199:67–71, 2002. 69. Ishii K, Sakamoto S, Sasaki M, et al: Cerebral glucose metabolism in patients with frontotemporal dementia. J Nucl Med 39:1875–1878, 1998. 70. Ishikawa K, Nagura H, Yokota T, et al: Signal loss in the motor cortex on magnetic resonance images in amyotrophic lateral sclerosis. Ann Neurol 33:218–222, 1993. 71. Jeong Y, Cho SS, Park JM, et al: 18F-FDG PET findings in frontotemporal dementia: An SPM analysis of 29 patients. J Nucl Med 46:233–239, 2005. 72. Johnson RT: Prion diseases. Lancet Neurol 4:635–642, 2005. 73. Kanda T, Ishii K, Uemura T, et al: Comparison of grey matter and metabolic reductions in frontotemporal dementia using FDG-PET and voxel-based morphometric MR studies. Eur J Nucl Med Mol Imaging 35:2227–2234, 2008. 74. Kiernan MC, Vucic S, Cheah BC, et al: Amyotrophic lateral sclerosis. Lancet 377:942–955, 2011. 75. Knopman DS, Boeve BF, Parisi JE, et al: Antemortem diagnosis of frontotemporal lobar degeneration. Ann Neurol 57:480–488, 2005. 76. Knudsen KA, Rosand J, Karluk D, et al: Clinical diagnosis of cerebral amyloid angiopathy: Validation of the Boston criteria. Neurology 56:537–539, 2001. 77. Koedam EL, Lehmann M, van der Flier WM, et al: Visual assessment of posterior atrophy development of a MRI rating scale. Eur Radiol 21:2618–2625, 2011. 78. Koeppen AH: Friedreich’s ataxia: Pathology, pathogenesis, and molecular genetics. J Neurol Sci 303:1–12, 2011. 79. Kostic VS, Agosta F, Filippi M: Parkinsonism. In Filippi M, Simon JH, editors: Imaging acute neurologic disease: A symptom-based approach, Cambridge, UK, 2014, Cambridge University Press, pp 270–290. 80. Lehmann M, Koedam EL, Barnes J, et al: Posterior cerebral atrophy in the absence of medial temporal lobe atrophy in pathologically confirmed Alzheimer’s disease. Neurobiol Aging 33(627):e1–e12, 2012. 81. Litvan I, Agid Y, Calne D, et al: Clinical research criteria for the diagnosis of progressive supranuclear palsy (Steele-RichardsonOlszewski syndrome): Report of the NINDS-SPSP international workshop. Neurology 47:1–9, 1996.

487

82. Mackenzie IR, Neumann M, Bigio EH, et al: Nomenclature and nosology for neuropathologic subtypes of frontotemporal lobar degeneration: An update. Acta Neuropathol 119:1–4, 2010. 83. Mahoney CJ, Beck J, Rohrer JD, et al: Frontotemporal dementia   with the C9ORF72 hexanucleotide repeat expansion: Clinical, neuroanatomical and neuropathological features. Brain 135:736–750, 2012. 84. Manners DN, Parchi P, Tonon C, et al: Pathologic correlates of diffusion MRI changes in Creutzfeldt-Jakob disease. Neurology 72:1425–1431, 2009. 85. Mascalchi M: Spinocerebellar ataxias. Neurol Sci 29(Suppl 3):311–313, 2008. 86. Mascalchi M, Vella A: Magnetic resonance and nuclear medicine imaging in ataxias. Handb Clin Neurol 103:85–110, 2012. 87. Mascalchi M, Vella A, Ceravolo R: Movement disorders: Role of imaging in diagnosis. J Magn Reson Imaging 35:239–256, 2012. 88. Massano J, Bhatia KP: Clinical approach to Parkinson’s disease: Features, diagnosis, and principles of management. Cold Spring Harb Perspect Med 2:a008870, 2012. 89. McKeith IG, Dickson DW, Lowe J, et al: Diagnosis and management of dementia with Lewy bodies: Third report of the DLB Consortium. Neurology 65:1863–1872, 2005. 90. McKhann GM, Knopman DS, Chertkow H, et al: The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7:263–269, 2011. 91. Meissner B, Kallenberg K, Sanchez-Juan P, et al: MRI lesion profiles in sporadic Creutzfeldt-Jakob disease. Neurology 72:1994–2001, 2009. 92. Mendez MF, Shapira JS, McMurtray A, et al: Accuracy of the clinical evaluation for frontotemporal dementia. Arch Neurol 64:830–835, 2007. 93. Migliaccio R, Agosta F, Rascovsky K, et al: Clinical syndromes associated with posterior atrophy: Early age at onset AD spectrum. Neurology 73:1571–1578, 2009. 94. Miller BL, Cummings JL, Villanueva-Meyer J, et al: Frontal lobe degeneration: Clinical, neuropsychological, and SPECT characteristics. Neurology 41:1374–1382, 1991. 95. Minoshima S, Foster NL, Sima AA, et al: Alzheimer’s disease versus dementia with Lewy bodies: Cerebral metabolic distinction with autopsy confirmation. Ann Neurol 50:358–365, 2001. 96. Mosconi L, Tsui WH, Herholz K, et al: Multicenter standardized 18F-FDG PET diagnosis of mild cognitive impairment, Alzheimer’s disease, and other dementias. J Nucl Med 49:390–398, 2008. 97. Mummery CJ, Patterson K, Price CJ, et al: A voxel-based morphometry study of semantic dementia: Relationship between temporal lobe atrophy and semantic memory. Ann Neurol 47:36–45, 2000. 98. Nandigam RN, Viswanathan A, Delgado P, et al: MR imaging detection of cerebral microbleeds: Effect of susceptibility-weighted imaging, section thickness, and field strength. AJNR Am J Neuroradiol 30:338– 343, 2009. 99. Nestor PJ, Fryer TD, Hodges JR: Declarative memory impairments in Alzheimer’s disease and semantic dementia. Neuroimage 30:1010–1020, 2006. 100. Nestor PJ, Fryer TD, Smielewski P, et al: Limbic hypometabolism in Alzheimer’s disease and mild cognitive impairment. Ann Neurol 54:343–351, 2003a. 101. Nestor PJ, Graham NL, Fryer TD, et al: Progressive non-fluent aphasia is associated with hypometabolism centered on the left anterior insula. Brain 126:2406–2418, 2003b. 102. O’Brien JT, McKeith IG, Walker Z, et al: Diagnostic accuracy of 123I-FP-CIT SPECT in possible dementia with Lewy bodies. Br J Psychiatry 194:34–39, 2009. 103. Pagani E, Agosta F, Rocca MA, et al: Voxel-based analysis derived from fractional anisotropy images of white matter volume changes with aging. Neuroimage 41:657–667, 2008. 104. Pettersen JA, Sathiyamoorthy G, Gao FQ, et al: Microbleed topography, leukoaraiosis, and cognition in probable Alzheimer disease from the Sunnybrook dementia study. Arch Neurol 65:790–795, 2008.

http://radiologyebook.com

488

PART II  CT and MR Imaging of the Whole Body

105. Pijnenburg YA, Mulder JL, van Swieten JC, et al: Diagnostic accuracy of consensus diagnostic criteria for frontotemporal dementia in a memory clinic population. Dement Geriatr Cogn Disord 25:157–164, 2008. 106. Prins ND, Scheltens P: White matter hyperintensities, cognitive impairment and dementia: An update. Nat Rev Neurol 11:157–165, 2015. 107. Querfurth HW, LaFerla FM: Alzheimer’s disease. N Engl J Med 362:329–344, 2010. 108. Rabinovici GD, Furst AJ, Alkalay A, et al: Increased metabolic vulnerability in early-onset Alzheimer’s disease is not related to amyloid burden. Brain 133:512–528, 2010. 109. Rabinovici GD, Jagust WJ, Furst AJ, et al: Abeta amyloid and glucose metabolism in three variants of primary progressive aphasia. Ann Neurol 64:388–401, 2008. 110. Rademakers R, Neumann M, Mackenzie IR: Advances in understanding the molecular basis of frontotemporal dementia. Nat Rev Neurol 8:423–434, 2012. 111. Rascovsky K, Hodges JR, Knopman D, et al: Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain 134:2456–2477, 2011. 112. Relkin N, Marmarou A, Klinge P, et al: Diagnosing idiopathic normal-pressure hydrocephalus. Neurosurgery 57:S4–S16, discussion ii–v, 2005. 113. Roman GC, Tatemichi TK, Erkinjuntti T, et al: Vascular dementia: Diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 43:250–260, 1993. 114. Rosen HJ, Gorno-Tempini ML, Goldman WP, et al: Patterns of brain atrophy in frontotemporal dementia and semantic dementia. Neurology 58:198–208, 2002. 115. Ross CA, Aylward EH, Wild EJ, et al: Huntington disease: Natural history, biomarkers and prospects for therapeutics. Nat Rev Neurol 10:204–216, 2014. 116. Salat DH, Buckner RL, Snyder AZ, et al: Thinning of the cerebral cortex in aging. Cereb Cortex 14:721–730, 2004. 117. Savva GM, Wharton SB, Ince PG, et al: Age, neuropathology, and dementia. N Engl J Med 360:2302–2309, 2009. 118. Scheltens P, Barkhof F, Leys D, et al: A semiquantitative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging. J Neurol Sci 114:7–12, 1993. 119. Seeley WW, Crawford R, Rascovsky K, et al: Frontal paralimbic network atrophy in very mild behavioral variant frontotemporal dementia. Arch Neurol 65:249–255, 2008. 120. Seppi K, Poewe W: Brain magnetic resonance imaging techniques in the diagnosis of parkinsonian syndromes. Neuroimaging Clin N Am 20:29–55, 2010. 121. Snyder HM, Corriveau RA, Craft S, et al: Vascular contributions to cognitive impairment and dementia including Alzheimer’s disease. Alzheimers Dement 2014. [Epub ahead of print]. 122. Tabrizi SJ, Scahill RI, Owen G, et al: Predictors of phenotypic progression and disease onset in premanifest and early-stage Huntington’s disease in the TRACK-HD study: Analysis of 36-month observational data. Lancet Neurol 12:637–649, 2013. 123. Tarnaris A, Kitchen ND, Watkins LD: Noninvasive biomarkers in normal pressure hydrocephalus: Evidence for the role of neuroimaging. J Neurosurg 110:837–851, 2009. 124. Thompson PM, Hayashi KM, de Zubicaray G, et al: Dynamics of gray matter loss in Alzheimer’s disease. J Neurosci 23:994–1005, 2003. 125. Thorpe JW, Moseley IF, Hawkes CH, et al: Brain and spinal cord MRI in motor neuron disease. J Neurol Neurosurg Psychiatry 61:314–317, 1996. 126. Turner MR, Grosskreutz J, Kassubek J, et al: Towards a neuroimaging biomarker for amyotrophic lateral sclerosis. Lancet Neurol 10:400–403, 2011.

127. van de Pol LA, Hensel A, van der Flier WM, et al: Hippocampal  atrophy on MRI in frontotemporal lobar degeneration and   Alzheimer’s disease. J Neurol Neurosurg Psychiatry 77:439–442, 2006. 128. van Dijk EJ, Prins ND, Vermeer SE, et al: Frequency of white   matter lesions and silent lacunar infarcts. J Neural Transm Suppl 25-39:2002. 129. van Straaten EC, Scheltens P, Knol DL, et al: Operational definitions for the NINDS-AIREN criteria for vascular dementia: An interobserver study. Stroke 34:1907–1912, 2003. 130. Virhammar J, Laurell K, Cesarini KG, et al: Preoperative prognostic value of MRI findings in 108 patients with idiopathic normal   pressure hydrocephalus. AJNR Am J Neuroradiol 35:2311–2318, 2014. 131. Viswanathan A, Greenberg SM: Cerebral amyloid angiopathy in the elderly. Ann Neurol 70:871–880, 2011. 132. Vitali P, Maccagnano E, Caverzasi E, et al: Diffusion-weighted MRI hyperintensity patterns differentiate CJD from other rapid dementias. Neurology 76:1711–1719, 2011. 133. Vonsattel JP, Myers RH, Stevens TJ, et al: Neuropathological classification of Huntington’s disease. J Neuropathol Exp Neurol 44:559–577, 1985. 134. Wahlund LO, Barkhof F, Fazekas F, et al: A new rating scale for age-related white matter changes applicable to MRI and CT. Stroke 32:1318–1322, 2001. 135. Weir DW, Sturrock A, Leavitt BR: Development of biomarkers for Huntington’s disease. Lancet Neurol 10:573–590, 2011. 136. Whitwell JL, Przybelski SA, Weigand SD, et al: Distinct anatomical subtypes of the behavioural variant of frontotemporal dementia: A cluster analysis study. Brain 132:2932–2946, 2009. 137. Whitwell JL, Przybelski SA, Weigand SD, et al: 3D maps from multiple MRI illustrate changing atrophy patterns as subjects progress from mild cognitive impairment to Alzheimer’s disease. Brain 130:1777–1786, 2007a. 138. Whitwell JL, Weigand SD, Boeve BF, et al: Neuroimaging signatures of frontotemporal dementia genetics: C9ORF72, tau, progranulin and sporadics. Brain 135:794–806, 2012. 139. Whitwell JL, Weigand SD, Shiung MM, et al: Focal atrophy in dementia with Lewy bodies on MRI: A distinct pattern from Alzheimer’s disease. Brain 130:708–719, 2007b. 140. Wozniak JR, Lim KO: Advances in white matter imaging: A review of in vivo magnetic resonance methodologies and their applicability to the study of development and aging. Neurosci Biobehav Rev 30:762–774, 2006. 141. Yamada F, Fukuda S, Samejima H, et al: Significance of pathognomonic features of normal-pressure hydrocephalus on computerized tomography. Neuroradiology 16:212–213, 1978. 142. Yi SH, Park KC, Yoon SS, et al: Relationship between clinical course and diffusion-weighted MRI findings in sporadic Creutzfeldt-Jakob disease. Neurol Sci 29:251–255, 2008. 143. Young GS, Geschwind MD, Fischbein NJ, et al: Diffusion-weighted and fluid-attenuated inversion recovery imaging in Creutzfeldt-Jakob disease: High sensitivity and specificity for diagnosis. AJNR Am J Neuroradiol 26:1551–1562, 2005. 144. Zarei M, Ibarretxe-Bilbao N, Compta Y, et al: Cortical thinning is associated with disease stages and dementia in Parkinson’s disease.   J Neurol Neurosurg Psychiatry 2013. 145. Zeidler M, Sellar RJ, Collie DA, et al: The pulvinar sign on magnetic resonance imaging in variant Creutzfeldt-Jakob disease. Lancet 355:1412–1418, 2000. 146. Zerr I, Kallenberg K, Summers DM, et al: Updated clinical diagnostic criteria for sporadic Creutzfeldt-Jakob disease. Brain 132:2659–2668, 2009.

http://radiologyebook.com

16  Functional Magnetic Resonance Imaging Sachin K. Gujar, Haris I. Sair, and Jay J. Pillai

INTRODUCTION Magnetic resonance imaging (MRI) is very adept at displaying structural anatomy of the brain in great detail. However, the correlation between anatomy and function is not always clear or apparent, and this becomes particularly difficult in situations where the anatomy deviates from the expected owing to developmental anatomic variation or if there is structural distortion by disease. A few short years after the clinical use of MRI for imaging the human brain, Ogawa et al.37,38 described the imaging technique they called blood oxygenation level–dependent (BOLD) contrast for MRI. They demonstrated that the sensitivity to paramagnetic effects of deoxyhemoglobin could be increased with the use of gradient echo (GRE) techniques and that this BOLD contrast followed physiologic changes and could provide in vivo real-time maps of blood oxygenation. The initial studies described drug-induced changes,38 but this was soon followed by studies describing the BOLD response to physiologic visual and primary sensory activation by two groups almost simultaneously in 1992.4,28,33 This BOLD technique is the most commonly used method for functional MRI (fMRI).

HEMODYNAMIC RESPONSE FUNCTION BOLD fMRI measures or demonstrates the effects of neuronal activity via the changes in the hemodynamic correlate.1 Deoxyhemoglobin is paramagnetic, whereas oxyhemoglobin is diamagnetic. The presence of paramagnetic deoxyhemoglobin in a blood vessel results in local susceptibility differences that produce signal decreases (hypointensity) on GRE T2* images. Because oxyhemoglobin is diamagnetic and does  not cause similar susceptibility changes, the changes in oxygenation of blood can be observed as signal changes on the T2* images.59 On the onset of neural activity, because oxygen consumption is expected to increase, it is also expected that local deoxyhemoglobin concentration would increase and result in decreased MRI signal. However, following a small transient decrease in BOLD signal at the onset of neural activity, it has been observed that there is a much more pronounced increase in signal, implying a net decrease in deoxyhemoglobin concentration. This has been thought to result from a much larger increase in cerebral blood flow in response to the initial increased oxygen extraction, bringing in more oxygenated blood (oxyhemoglobin), with consequent reduction in deoxyhemoglobin levels in the local voxels. This variation of the BOLD fMRI signal as a function of time in response to neuronal activity is known as the hemodynamic response function (HRF).21 The relationship between the underlying neural activity and the resultant hemodynamic response is also known as neurovascular coupling.1 BOLD signal changes occur in tissue (extravascular BOLD effect) as well as within the vasculature (intravascular), with additional signal changes also occurring within the larger vessels.25

Although the physics behind the BOLD response is fairly well understood, there have been numerous theories proposed to explain the neural basis of the BOLD response.13 The favored explanation is a relatively direct correlation between fMRI signal and synaptic activity via a neurotransmitter-driven hyperemia response.1,2

CLINICAL fMRI Most fMRI studies performed for clinical use are task-based studies that require the patient to perform a motor, language, or visual task in the MRI scanner while (typically) GRE T2*-weighted echo planar images (EPI) are rapidly acquired. Another less clinically established but promising BOLD technique that will be discussed later in this chapter is resting state functional MRI (rs-fMRI).

Task-Based fMRI Most fMRI performed clinically is task based, with ultrafast images (e.g., single-shot T2* GRE echo planar technique) obtained during performance of defined motor or sensorimotor, language, or other cognitive or visual tasks. The magnitude of the BOLD signal changes is relatively small and can be more pronounced on high-field-strength magnets (at our institution, we perform our fMRI studies on a 3-tesla [T] scanner). Use of a block design for the fMRI examination (described later) and multiple repetitions also improve signal-to-noise ratio and BOLD contrast. The rapid EPI technique allows imaging of the whole brain with a temporal resolution of approximately 2000 milliseconds and a spatial resolution of approximately 3 to 5 mm. A typical fMRI study paradigm would acquire generally between 90 and 120 whole brain volumes in 3 to 4 minutes. Disadvantages of the EPI technique include heightened susceptibility artifacts, especially at the skull base. Spin echo methods with acquisition of T2-weighted images can also be employed to minimize these artifacts; however, they provide lower BOLD effect and lower contrast between images acquired during rest and activity states.25 The stimulus or the activation paradigms are essentially of two types: block design paradigms or event-related paradigms. Block design paradigms consist of alternating blocks/epochs of a previously defined task alternating with equal-duration blocks/epochs of rest or a control task (Fig. 16-1). In most sensorimotor or language paradigms, each epoch duration in a block design is approximately 20 to 30 seconds. Event-related paradigms or single event paradigms are designed with short periods of activation alternating with longer periods of rest,  with MR acquisition methods designed to measure hemodynamic changes over several seconds.25 Advantages of event-related design include the ability to present very brief stimuli at irregular intervals. Disadvantages of such design include greater task difficulty, particularly for neurologically impaired patients, and much lower statistical power for equivalent-duration paradigms. In clinical fMRI, block

http://radiologyebook.com

489

490

PART II  CT and MR Imaging of the Whole Body BOLD signal intensity Rest

Task

Rest

Task

Rest

Task

Activity

FIG 16-1  Typical block design used for task-based fMRI examinations is depicted in this diagram. Equivalentduration epochs of rest alternate with epochs of sustained repetitive activity (motor, language, or visual task performance). The upper half of the image displays the summation of individual stimulus-related hemodynamic responses as effectively sustained increase in BOLD signal during the activity blocks compared to the resting blocks.

design paradigms are generally used because they are generally easier to perform, are more efficient, and produce more statistically robust results.65

Motor Tasks.  All the motor tasks we use are designed as a block design paradigm, with 30 seconds of rest alternating with 30 seconds of activity within each cycle, with the visual (or aural) instructions to “stop” or “go.” The most common sensorimotor tasks include a foot movement task (with continuous bilateral gentle plantar flexion and dorsiflexion of the foot at the ankle at a steady pace, or alternatively toe flexion and extension), a finger movement task (which includes bilateral simultaneous sequential finger tapping), and a vertical tonguemovement task (with continuous gentle up-and-down movement of the tongue at a steady pace, with special care to avoid movement of the jaw) that elicits activation related to the face motor area. Alternatively, lip-puckering may be used for eliciting activation in the lower face sensorimotor cortex. A variety of other hand movement tasks (e.g., hand opening/closing, squeezing) can be used to activate the hand representation area of the primary motor cortex. If one thinks of the diagrammatic representations of the motor homunculus from physiology textbooks, the hand representation is always displayed to be overrepresented and takes up a large portion of the motor strip. Because the hand motor tasks are relatively simple to perform and show good activation of the motor cortex, hand motor paradigms should be performed if the primary motor cortex is to be localized. The bilateral finger-tapping task is successful in eliciting activation of the primary motor as well as the somatosensory cortex of the precentral and postcentral gyrus, respectively, in addition to the supplementary motor area (SMA) along the medial aspect of the cerebral hemisphere5 (see Fig. 16-3). We often employ a more complex three-phase hand opening/ closing task (right hand followed by left hand and then by a rest phase) using a dynamic visual display as the cue to obtain motor cortical activation maps in the precentral gyrus while minimizing sensory cortex activation (Fig. 16-2). The foot/ankle movement task results in robust activity within the precentral gyrus near the superior termination of the central sulcus and more prominently in the paracentral lobule, with activation of the supplementary motor area. Cerebellar activation is commonly observed during motor tasks in the hemisphere contralateral to the activated primary motor cortex.

Language Tasks.  Functional MRI has been shown to be reliable for the prediction of hemispheric language dominance.17,50,51 Numerous paradigms are typically used to elicit the activation in eloquent cortex related to language function. Certain paradigms have been specially designed to elicit activation primarily in the speech-production areas (expressive paradigms) in the receptive language regions (receptive paradigms), whereas others activate both regions. Language-activation paradigms may be further divided into semantic, phonologic, or verbal fluency paradigms. In general, for clinical language fMRI, only covert

(silently thought) rather than overt (physically spoken) language tasks are used to minimize head/mandibular motion. The expressive paradigms have been designed to elicit activation in the inferior frontal gyrus (IFG) corresponding to the region of Broca’s area, the dorsolateral prefrontal cortex (DLPFC) over the convexity of the frontal lobe, usually in the hemisphere dominant for language, and in the premotor or language supplementary motor area (pre-SMA) along the medial aspect of the frontal lobes (Fig. 16-3). The receptive paradigms have been designed to elicit activation in the receptive language areas predominantly in the temporal and parietal lobes, most prominently in Wernicke’s area, which technically involves Brodmann area 22 in the posterior-most aspect of the superior temporal gyrus (STG), although activation is typically seen along the posterior aspect of the superior temporal sulcus (STS), middle temporal gyrus (MTG), and temporal occipital junction, as well as in the inferior parietal lobule, including the angular and supramarginal gyrus (SMG) (Fig. 16-4). Expressive paradigms.  Some of the expressive language tasks commonly used at our institution include covert word-generation tasks such as the silent word–generation task and the object-naming task. During the silent word–generation task, the patient is asked to covertly generate multiple words starting with the presented letter of the alphabet. In the object-naming task, the patient is asked to silently name a series of presented objects. The control task uses a nonsense drawing for visual fixation in order to minimize the effects of visual cortex activation. Silent word generation has been compared favorably with the Wada test in the lateralization of language function.62,63 With this task, activation is typically identified within the DLPFC and IFG, and variably within the cingulate language regions, as well as in the pre-SMA regions. This task produces robust activation and effectively lateralizes speech functions. A similar task is a verb-generation task where instead of a letter of the alphabet, clues are visually or aurally presented. The patient is then asked to generate related verbs. Object naming is a simple covert word-generation task intended to activate language cortical functions of the dominant hemisphere, including in the IFG in addition to other areas. An example of a phonologic task is the rhyming task. The more complex rhyming task alternates epochs of a series of word pairs with a series of images showing two rows of stick figures, where the subjects respond electronically if the word pairs rhyme or if the rows  of stick figures match. Typical activation includes DLPFC, IFG, STG, and STS. Receptive paradigms.  Examples of commonly used receptive (i.e., comprehension or semantic) paradigms include the sentence listening comprehension and sentence reading comprehension tasks, as well as the passive story-listening task. The sentence listening comprehension task alternates blocks of real sentences with garbled speech and requires the subject to respond  via button presses on a keypad only if the content of the presented sentence is true. The reading comprehension task is a similar paradigm

http://radiologyebook.com

CHAPTER 16  Functional Magnetic Resonance Imaging

491

FIG 16-2  This 63-year-old right-handed woman with a right frontal lobe anaplastic astrocytoma (WHO grade III) underwent BOLD fMRI motor cortical mapping with three different tasks: right hand opening/closing is depicted in yellow, left hand opening/closing in red, and vertical tongue movement in light blue (cyan). Note that the tumor is expanding the right precentral gyrus and resulting in effacement of the right central sulcus. The tumor involves the portion of the precentral gyrus just inferior to the hand representation area and extends into the face representation area, which demonstrates asymmetrically decreased activation compared to the normal contralateral precentral gyrus. The asymmetric decrease in ipsilesional activation may be a reflection of tumor-induced neurovascular uncoupling.

that uses written sentences instead of aural stimuli. These are useful for clinical questions regarding language localization in Broca’s area and Wernicke’s area. Another example of a simple receptive task is a passive listening task that alternates segments of a short story with garbled speech sounds. Activation is most typically seen within the STG and in the cortex lining the STS. Additionally, activation is seen within the STG and MTG, more anteriorly. Semantic paradigms.  Semantic paradigms are more general purpose for language mapping than the expressive and receptive

paradigms described. These are designed to be useful for language lateralization as well as localization of key expressive and receptive speech areas in the dominant hemisphere. The sentence completion task requires the patient to covertly complete written sentences that have been displayed with the last word missing. The control block consists of simple designs or a nonsense series of scrambled letters arranged in the form of a sentence. For this task, the dominant IFG activation is seen, consistent with the dominant expressive speech area, together with left MTG activation, consistent with the dominant receptive speech area.

http://radiologyebook.com

492

PART II  CT and MR Imaging of the Whole Body

FIG 16-3  This 35-year-old right-handed (with moderate ambidexterity) woman with a WHO grade III anaplastic oligodendroglioma involving the medial left frontal lobe underwent presurgical mapping with BOLD fMRI using both language and motor paradigms. The T2/FLAIR hyperintense mass is seen infiltrating and expanding the left superior frontal gyrus as well as a portion of the left middle frontal gyrus. This figure depicts the composite activation map from five different tasks: sentence reading comprehension in magenta, silent word generation in red, sentence completion in yellow, and rhyming in light blue (cyan), as well as a bilateral simultaneous sequential finger-tapping task in green. The second row arrow displays the pre-supplementary motor area, and the top row arrow designates the slightly more posterior motor SMA. The large cluster of language-related activation just lateral to the mass in the middle row represents dorsolateral prefrontal cortex expressive language activation, whereas Broca’s activation is seen more inferiorly in the left inferior frontal gyrus in the bottom row. The hand representation area of the primary motor cortex is seen depicted in green just posterior to the posterior margin of the mass.

http://radiologyebook.com

CHAPTER 16  Functional Magnetic Resonance Imaging

493

FIG 16-4  This 37-year-old right-handed woman with a WHO grade II oligoastrocytoma involving the left parietal operculum underwent language mapping with three language tasks, as well as mapping of the face representation area of the primary motor cortex (PMC) with a vertical tongue-movement task (color-coded in magenta). The three language tasks include a passive story-listening task in dark blue, sentence-listening comprehension task in light blue (cyan), and a silent word–generation task in red. Note that the face representation area of the PMC is located just above the superior margin of the tumor, which involves the postcentral gyrus. The top arrow indicates Broca’s area involving the left inferior frontal gyrus, and the lower arrow indicates functional Wernicke’s area along the left superior temporal sulcus. This patient was lefthemispheric dominant for both expressive and receptive components of language.

There are other semantic decision paradigms such as noun-verb association tasks and word category association tasks that are not described in detail here.

Visual Paradigms.  Activation of the visual cortex involves a large portion of the occipital cortex and has been investigated since the early days of fMRI research. Visual stimulation results in robust activation in the visual cortex, and visual stimuli typically involve different varieties of flashing checkerboard patterns. Such stimuli have been effective for both retinotopic mapping and assessing foveal vision.

Memory Paradigms.  Memory paradigms have not gained general acceptance for clinical fMRI, although much work has been performed at a research level using such paradigms.17 These can be difficult to perform and interpret, and susceptibility artifacts near the skull base can be a major limitation. Reliable memory lateralization with such

paradigms has been difficult to accomplish. These are beyond the scope of this chapter and are not discussed in detail here.

PERFORMING CLINICAL fMRI The Setup Because the magnitude of the BOLD signal is small, use of a high-fieldstrength magnet is essential; although 1.5 T is acceptable, the current standard is 3 T. At our institution, we perform fMRI studies on a 3-T scanner using a multichannel head coil. The stimulus paradigm is administered visually by displaying text or instructions via projectormirror system or specialized goggles, or with auditory stimulus using specialized head phones especially for language/semantic tasks.

Pre-Scan Interview and Training Reviewing recent prior imaging studies allows planning of the fMRI study and selection of paradigms tailored to the individual patient.

http://radiologyebook.com

494

PART II  CT and MR Imaging of the Whole Body

Furthermore, pre-scan patient training is essential to familiarize patients with the tasks, assess their ability to adequately perform them, and alleviate anxiety.44

allows greater flexibility in discovering task-related and non–taskrelated signals without necessitating assumptions about their time courses.8,9

Monitoring

Role of Streamlined Postprocessing Tools

Performance of foot, finger, or hand motor tasks can usually be monitored visually. Performance of some of the language tasks is monitored via patient responses/button presses, but performance of other covert language tasks cannot be directly monitored and must be assumed on the basis of the performance during the pre-scan training. Real-time maps allow monitoring of task performance and also permit assessment of bulk head motion. These enable the neuroradiologist to make real-time informed decisions to repeat poor runs that are due to inadequate performance or excessive head motion that may have otherwise resulted in a nondiagnostic exam.

Research-level BOLD fMRI postprocessing software is very accurate but relatively cumbersome, requiring significant experience in image processing, as well as computer programming, for generation of custom-made scripts for postprocessing. The main drawbacks in the use of such software include lack of U.S. Food and Drug Administration (FDA) approval and inability to efficiently export postprocessed images to picture archiving and communication systems (PACS) servers and neuronavigation systems. However, there are commercially available FDA-approved software packages now available that are more user friendly, allow for better image overlays on highresolution anatomic images, and are compatible with both PACS servers and neuronavigation software.18,43 This has enabled import of functional images into the operating room, such that neurosurgeons can use the data in surgical planning. These packages also provide multiple options for statistical thresholding (e.g., cross-correlation,  P value, T-statistic) and quality control analysis. At the same time,  the research software allows for additional processing in situations where processing with the commercially available packages is inadequate.

Postprocessing and Quality Control After acquisition of data, several preprocessing steps are needed before the final statistical analysis and creation of activation maps. As a part of postprocessing at our institution, we perform an extensive quality control analysis that includes assessment of patient head motion, functional-anatomic alignment, susceptibility artifacts, and data outlier volumes.

Analysis of fMRI Images A variety of methods are available for analysis of fMRI. The most widely used is a regression analysis using a general linear model (GLM).16 The GLM can be written as Y = XB + U, where Y represents the fMRI data, X represents the design of the experiment, B represents the estimation parameters, and U represents the error term. For each voxel, a B is estimated that minimizes the error between the observed (dependent) data and a reference, or predictor. In clinical fMRI, a hemodynamic response function (HRF) convolved with the boxcar function of the task serves as the reference in a block-design paradigm. For cases with multiple tasks, each may be designated separately in the design matrix X, and corresponding beta coefficients calculated.15 Following GLM fitting, an F statistic is calculated to determine how well the model explains a particular voxel’s time course. A high F statistic or low p value indicates that there is a high correlation between the model and the modulation of the voxel time course. A T-statistic is then calculated by contrasting the signal in one condition to a second (usually “control” or “rest”) condition. A typical fMRI study at 3 T may have approximately 100,000 voxels. As a separate GLM analysis is performed for each voxel, this is also referred to as a massive univariate analysis. Given the large number of calculations performed, various methods have been employed to ensure that type I (false-positive) errors in activation are minimized.30 Ensuring specificity using multiple correction methods such as familywise error (FEW), Bonferroni correction, false discovery rate (FDR), or cluster-based thresholding must be balanced with the need to avoid type II (false-negative) errors, which are more serious in clinical fMRI because surgical resection of areas labeled as no activation may result in significant morbidity. There are many assumptions that must be met for fMRI analysis using GLM. Although most of these assumptions are beyond the scope of this book, one warrants discussion: fMRI GLM analysis relies on a canonical HRF to function as the basis for the reference. The HRF, however, may be variable across brain regions, across different age ranges, and even vary with specific stimulus conditions.26,36,41 Thus alternate methods have been investigated for model-free analysis  of fMRI, including independent component analysis (ICA), which

Challenges and Limitations BOLD fMRI is very sensitive to susceptibility and motion artifacts. Susceptibility artifacts can be especially problematic in patients who have had significant hemorrhage in the vicinity of the areas of interest or in those who have had prior surgery, with susceptibility artifacts from calvarial metallic implants/fixation devices. There may be situations where patients are not able to perform the tasks owing to lack of understanding, disability, or inability to cooperate for the exam. BOLD fMRI is also affected by so-called venous artifacts because of intravascular BOLD effects. These changes in blood flow and oxygenation may propagate downstream in the brain vasculature and appear in larger draining veins distant from the active cortical areas. The venous location of the activation can, however, be confirmed  by comparison of functional activation maps with high-resolution (contrast-enhanced) structural images. Another problem associated with fMRI in the clinical setting is neurovascular uncoupling. As mentioned earlier, BOLD fMRI is an indirect measure of neuronal activity based upon the related hemodynamic changes. The absence of activation on BOLD fMRI does not imply the lack of neuronal activity. There are conditions under which the BOLD response may be disturbed, as with large vascular malformations or brain tumors with associated neovascularity, some low-grade tumors, or in situations with severe proximal arterial stenosis or strokes, which may give rise to false-negative activation.17,44,64,65 Maps of cerebrovascular reactivity obtained using hypercapnia stimuli56 can be tools that provide an overall whole-brain view of the BOLD response. At our institution, we routinely use a breath-hold (BH) technique44,64 for this purpose. This BH technique involves shortduration BHs alternating with longer periods of normal self-paced relaxed breathing (we use a 16-second BH block with a normal breathing block of 40 seconds), repeated multiple times. Monitoring of task performance is performed using a standard respiratory belt. Postprocessing, however, needs to be performed with specific modeling to account for the differing durations of the BH and normal breathing blocks. The information obtained can be used to validate the ability of a perfusion bed to respond to a physiologic stimulus and the reliability of the fMRI mapping data.44

http://radiologyebook.com

CHAPTER 16  Functional Magnetic Resonance Imaging

CLINICAL APPLICATIONS OF fMRI Clinical use of MRI is a relatively recent phenomenon, with over 15 years of collective experience. The main clinical applications are in presurgical evaluation and mapping of patients with structural lesions (e.g., brain tumors) and in patients with epilepsy.42 Presurgical mapping has been used extensively in patients with brain tumors. Preoperative fMRI studies have been demonstrated  to be helpful to surgeons for mapping of eloquent cortex and proximity to the target lesion, allowing informed decisions to be made regarding potential risks or safety of surgical resection and the  surgical trajectory; fMRI may influence the need for intraoperative cortical mapping.39 When fMRI is used for presurgical planning, the choice of paradigms depends mainly upon the location of the lesion.6 For example, if the tumor is located in the posterior frontal lobe, motor paradigms can be used to map the motor strip/precentral gyrus. Presurgical mapping has also been extensively used in epilepsy patients, primarily for hemispheric language lateralization. In epilepsy patients, a higher percentage of patients display atypical language lateralization than in normal individuals, and sometimes there may be discordance between expressive and receptive language lateralization. It has also been used for motor and visual cortical mapping in cases where malformations of cortical development or other resectable  epileptogenic lesions are thought to be in close spatial proximity to eloquent cortical regions.

RESTING STATE fMRI Principles Resting state fMRI (rs-fMRI) has emerged as a complementary tool to task-based fMRI (tb-fMRI) for analysis of brain function.20 In tb-fMRI an “active” state corresponding to the task is followed by a “rest” state in which a control task is presented to nullify or diminish taskassociated signals of no interest (e.g., signals related to attention in a task not specifically interrogating attention mechanisms). In rs-fMRI there is no explicit modulatory task given, and spontaneous changes in brain activity are recorded. As in tb-fMRI, currently the majority of MR-based imaging of resting state brain function uses the BOLD technique described previously. The spontaneous changes in brain activity are not random, with spatially distinct brain areas demonstrating patterns of variable temporal synchrony. This temporal synchrony occurs generally in the frequency range of approximately 0.1 to 0.01 Hz. When linking highly synchronous brain regions together, a set of intrinsic networks of the brain emerge.14

Processing Processing of rs-fMRI partially mirrors that of tb-fMRI to include slice timing correction when appropriate, motion correction/realignment, detrending, and spatial smoothing if necessary. However, additional processing steps are necessary depending on the analysis method used to diminish false correlations related to physiologic noise and motion. In fact, motion can significantly alter functional connectivity measurements, enhancing short-range connectivity and diminishing longrange connectivity.45,46 In studies where motion characteristics may differ on a global scale between groups, caution must be taken to account for these motion-related effects.23 In addition to motionrelated artifacts, physiologic noise of cardiopulmonary origin can  also contribute to rs-fMRI artifacts.34 Here, ideal nuisance reduction involves prospective recording of cardiac and pulmonary signals to  be regressed from the fMRI volumes obtained. However, insofar as

495

real-time recording is often impractical, solutions have been derived that allow retrospective removal of noise, typically involving regression of signal from white matter and cerebrospinal fluid from the rs-fMRI volumes.3,35 Need for removal of average global signal is debated.10,61 High-pass or bandpass filtering is commonly performed in seed-based analysis to limit signal detection in frequency bands of interest. Highmotion volumes of rs-fMRI may be removed to improve overall quality of the remaining signal.

Analysis There are multiple methods to analyze rs-fMRI data; however, the most common can be divided into seed-based correlation analysis and blind source separation.24 In seed-based analysis, seed regions may be placed in certain brain areas based on a priori hypotheses of connectivity in the study group. At its most basic, two seed regions can be placed to determine the connectivity between those regions. A voxel-wise correlation analysis of the whole brain may also be performed using a time course derived from a single seed. Multiple seeds or regions of interest can also be used to determine pairwise correlations across each, and graph theoretical methods can then be employed to quantitatively describe network measures. Graph theory offers an ideal method of describing whole-brain correlations; the basis of graph theory is the description of multiple pairwise correlations.55 In graph terminology, a node or vertex refers to each unit that forms a graph (e.g., each seed or ROI). The edge represents the connection between the nodes/vertices (i.e., correlation). Graphs may be directed or undirected, the latter indicating that information regarding edge directionality is absent. They may be weighted, where edges may have varying values depending on the strength of correlations, or unweighted, where magnitude information of the edges is absent. The total number of edges associated with a particular node represents the degree of the node. For a specific graph, an average of the degrees of each node can be calculated, and any node with a higher degree compared to this average is considered a hub. Various clustering methods can be used to group together nodes with high degrees of interconnected edges to form communities or modules. These modules often correspond to different groups of intrinsic brain networks described below. Independent component analysis (ICA) has been the most widely used method of blind source separation in rs-fMRI. Here, multivariate rs-fMRI signal is decomposed into a set of maximally independent signals. ICA assumes that the source signals are independent of each other, the signals are linearly mixed, and the distribution of signals is non-gaussian. Multiple algorithms for ICA exist. In rs-fMRI, spatial ICA has consistently demonstrated a set of intrinsic brain networks (IBNs) that demonstrate unique spatial distributions.12 Although there is no consensus yet as to the exact constituents or the exact number of these IBNs, there are a limited group of IBNs that have been reliably demonstrated in the literature11,60 (Fig. 16-5).

Intrinsic Brain Networks.  The most widely studied IBN is the default mode network (DMN). Regions of the DMN include the precuneus, the posterior cingulate (specifically the retrosplenial cortex), the medial prefrontal cortex, and the inferior parietal lobules.19 Although the exact function of the DMN remains unknown to date, it is thought to be involved in self-referential tasks or internal tasks.47 The DMN may be subdivided into several subcomponents depending on the data processing and analysis method used. The attention network may be separated into the dorsal attention network (DAN) and the ventral attention network (VAN). The  DAN includes bilateral parietal lobes at the banks of the intraparietal

http://radiologyebook.com

496

PART II  CT and MR Imaging of the Whole Body

DMN

ATT

ECN

SMN

VIS

AUD

FIG 16-5  Canonical intrinsic networks of the brain extracted from rs-fMRI using group independent component analysis (ICA). For each network, only the right hemisphere is shown in lateral, dorsal, and medial views. Some of the networks are split across different ICA components for this specific dataset; each subnetwork is shown in the same row, separated by dashed lines. Yellow indicates regions with high correlations for each map. DMN, default mode network; ATT, attention network; ECN, executive control network; SMN, sensorimotor network; VIS, visual network; AUD, auditory network.

sulci, as well as the frontal eye fields. The VAN includes the rostral aspect of the inferior frontal gyri bilaterally and the inferior parietal lobules. The executive control network (ECN), also referred to as the frontoparietal network, is related to tasks of executive functioning, typically lateralized, and consists of dorsal medial prefrontal cortex and the ipsilateral superior parietal lobule. The salience network (SAL) includes bilateral insula and the anterior cingulate gyrus. This network is thought to mediate the assigning of salience, or relevance, of stimuli and may engage other appropriate higher-order cognitive networks to subsequently process the data.32 A language network (LAN) has been variably described, and when shown includes portions of the inferior frontal gyrus, posterior banks of the superior temporal sulcus, and the supramarginal gyrus.57,58 The primary networks include the sensorimotor network (SMN; also known as the somatomotor network), involving the primary sensory and motor regions of the pre- and postcentral gyri as well as the supplementary motor area; the visual network (VIS), which as expected

includes the occipital lobe subserving the primary visual cortex; and the auditory network (AUD), which includes bilateral primary auditory cortex (Heschl’s gyri) and adjacent superior temporal gyrus. Additional networks are less commonly described in the literature, such as the basal ganglia network or the cerebellar network.

Applications of rs-fMRI Although significant progress has been made in attempting to use rs-fMRI as a viable clinical tool, at the current time the clinical applications are limited. At the group level, network dysfunction has been well described across a plethora of disorders. The most widely studied network, the DMN, appears to be universally affected.7 DMN aberrations in Alzheimer’s disease is well documented.53 Connectivity of the DMN has also been demonstrated to be different in patients with mild cognitive impairment who progress to Alzheimer’s, compared to those who do not.40 Interestingly the spatial distribution of atrophy in different types of neurodegenerative disorders appears to correlate with the spatial distribution of IBNs.52

http://radiologyebook.com

CHAPTER 16  Functional Magnetic Resonance Imaging The greatest potential of rs-fMRI in clinical practice is in the setting of presurgical mapping, where tb-fMRI is routinely used.54,67 There are several advantages of rs-fMRI compared to tb-fMRI in this setting. Because there is no need to perform a specific task in rs-fMRI (other than to try to keep as motionless as possible), this technique is helpful in situations where patients are not able to perform tb-fMRI owing to lack of understanding (e.g., related to cognitive deficits or language barriers) or physical disability (e.g., inability to move an extremity). In addition, rs-fMRI potentially can be obtained in less time than is necessary for a typical tb-fMRI and gives a wealth of information pertaining to multiple networks in the brain, as opposed to a narrow functional domain specifically related to the task in tb-fMRI. rs-fMRI is also easier to implement because no stimulus presentation hardware or software is needed. tb-fMRI does have an advantage at the current time in having multiple options for easy-to-use software packages for processing and analysis. These packages, however, will undoubtedly offer rs-fMRI–based network visualization in the future as processing and analysis methods become standardized.

Future of rs-fMRI Significant advances have been made in assessing functional connectivity of the brain since the discovery of synchronous BOLD signals in the brain. Many studies have assumed temporally stationary correlations across rs-fMRI acquisition; however, recently dynamic functional correlations have been investigated and offer insight into time-varying connectivity changes.22,29,66 Although network dysfunction has been demonstrated at the group level across many disorders, single subject– level diagnosis has been challenging. Characterizing normal versus abnormal dynamic connectivity may potentially improve subject-level applications.49 rs-fMRI also has shown some promise in prediction, including cognitive function and recovery from disease. This may be helpful in clinical settings where outcome following injury is uncertain, or it may be helpful in selection of patients who would maximally benefit from intervention.27,31,48

REFERENCES 1. Arthurs OJ, Boniface S: How well do we understand the neural origins of the fMRI BOLD signal? Trends Neurosci 25(1):27–31, 2002. 2. Attwell D, Buchan AM, Charpak S, et al: Glial and neuronal control of brain blood flow. Nature 468(7321):232–243, 2010. 3. Behzadi Y, Restom K, Liau J, et al: A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. Neuroimage 37(1):90–101, 2007. 4. Belliveau JW, Kennedy DN, Jr, McKinstry RC, et al: Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254(5032):716–719, 1991. 5. Boecker H, Kleinschmidt A, Requardt M, et al: Functional cooperativity of human cortical motor areas during self-paced simple finger movements. A high-resolution MRI study. Brain 117(Pt 6):1231–1239, 1994. 6. Bogomolny DL, Petrovich NM, Hou BL, et al: Functional MRI in   the brain tumor patient. Top Magn Reson Imaging 15(5):325–335, 2004. 7. Broyd SJ, Demanuele C, Debener S, et al: Default-mode brain dysfunction in mental disorders: A systematic review. Neurosci Biobehav Rev 33(3):279–296, 2009. 8. Calhoun VD, Adali T, McGinty VB, et al: fMRI activation in a visualperception task: Network of areas detected using the general linear model and independent components analysis. Neuroimage 14(5):1080– 1088, 2001. 9. Calhoun VD, Adali T, Pearlson GD, et al: Spatial and temporal independent component analysis of functional MRI data containing a pair of task-related waveforms. Hum Brain Mapp 13(1):43–53, 2001.

497

10. Chen G, Chen G, Xie C, et al: A method to determine the necessity for global signal regression in resting-state fMRI studies. Magn Reson Med 68(6):1828–1835, 2012. 11. Chen S, Ross TJ, Zhan W, et al: Group independent component analysis reveals consistent resting-state networks across multiple sessions. Brain Res 1239:141–151, 2008. 12. Damoiseaux JS, Rombouts SA, Barkhof F, et al: Consistent resting-state networks across healthy subjects. Proc Natl Acad Sci U S A 103(37):13848–13853, 2006. 13. Fox PT: The coupling controversy. Neuroimage 62(2):594–601, 2012. 14. Fox MD, Snyder AZ, Vincent JL, et al: The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci U S A 102(27):9673–9678, 2005. 15. Friston KJ: Analysing brain images: Principles and overview. In Frackowiak RSJ, Friston KJ, Frith CD, et al, editors: Human brain function, San Diego, CA, 1997, Academic Press, p 25. 16. Friston KJ, Holmes AP, Worsley KJ, et al: Statistical parametric maps in functional imaging: A general linear approach. Hum Brain Mapp 2(4):189–210, 1994. 17. Gaillard WD: Functional MR imaging of language, memory, and sensorimotor cortex. Neuroimaging Clin N Am 14(3):471–485, 2004. 18. Gonzalez-Ortiz S, Oleaga L, Pujol T, et al: Simple fMRI postprocessing suffices for normal clinical practice. AJNR Am J Neuroradiol 34(6):1188– 1193, 2013. 19. Greicius MD, Supekar K, Menon V, et al: Resting-state functional connectivity reflects structural connectivity in the default mode network. Cereb Cortex 19(1):72–78, 2009. 20. Gusnard DA, Raichle ME, Raichle ME: Searching for a baseline: Functional imaging and the resting human brain. Nat Rev Neurosci 2(10):685–694, 2001. 21. Heeger DJ, Ress D: What does fMRI tell us about neuronal activity? Nat Rev Neurosci 3(2):142–151, 2002. 22. Hutchison RM, Womelsdorf T, Allen EA, et al: Dynamic functional connectivity: Promise, issues, and interpretations. Neuroimage 80:360– 378, 2013. 23. Jo HJ, Gotts SJ, Reynolds RC, et al: Effective preprocessing procedures virtually eliminate distance-dependent motion artifacts in resting state FMRI. J Appl Math 2013:2013. doi: 10.1155/2013/935154. 24. Joel SE, Caffo BS, van Zijl PC, et al: On the relationship between seed-based and ICA-based measures of functional connectivity. Magn Reson Med 66(3):644–657, 2011. 25. Jones RA, Brooks JA, Moonen CTW: Ultra-fast fMRI. In Jezzard P, Matthews PM, Smith SM, editors: Functional MRI, an introduction to methods, New York, 2001, Oxford University Press, p 93. 26. Kannurpatti SS, Motes MA, Rypma B, et al: Neural and vascular variability and the fMRI-BOLD response in normal aging. Magn Reson Imaging 28(4):466–476, 2010. 27. Klumpp H, Keutmann MK, Fitzgerald DA, et al: Resting state amygdalaprefrontal connectivity predicts symptom change after cognitive behavioral therapy in generalized social anxiety disorder. Biol Mood Anxiety Disord 4(1):14-014-0014-5, 2014. eCollection 2014. 28. Kwong KK, Belliveau JW, Chesler DA, et al: Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proc Natl Acad Sci U S A 89(12):5675–5679, 1992. 29. Leonardi N, Van De Ville D: On spurious and real fluctuations of dynamic functional connectivity during rest. Neuroimage 104:430–436, 2015. 30. Loring DW, Meador KJ, Allison JD, et al: Now you see it, now you don’t: Statistical and methodological considerations in fMRI. Epilepsy Behav 3(6):539–547, 2002. 31. Mennes M, Kelly C, Zuo XN, et al: Inter-individual differences in resting-state functional connectivity predict task-induced BOLD activity. Neuroimage 50(4):1690–1701, 2010. 32. Menon V: Large-scale brain networks and psychopathology: A unifying triple network model. Trends Cogn Sci 15(10):483–506, 2011. 33. Menon RS, Ogawa S, Kim SG, et al: Functional brain mapping using magnetic resonance imaging. signal changes accompanying visual stimulation. Invest Radiol 27(Suppl 2):S47–S53, 1992.

http://radiologyebook.com

498

PART II  CT and MR Imaging of the Whole Body

34. Murphy K, Birn RM, Bandettini PA: Resting-state fMRI confounds and cleanup. Neuroimage 80:349–359, 2013. 35. Muschelli J, Nebel MB, Caffo BS, et al: Reduction of motion-related artifacts in resting state fMRI using aCompCor. Neuroimage 96:22–35, 2014. 36. Neumann J, Lohmann G, Zysset S, et al: Within-subject variability of BOLD response dynamics. Neuroimage 19(3):784–796, 2003. 37. Ogawa S: Finding the BOLD effect in brain images. Neuroimage 62(2):608–609, 2012. 38. Ogawa S, Lee TM, Kay AR, et al: Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A 87(24):9868–9872, 1990. 39. Petrella JR, Shah LM, Harris KM, et al: Preoperative functional MR imaging localization of language and motor areas: Effect on therapeutic decision making in patients with potentially resectable brain tumors. Radiology 240(3):793–802, 2006. 40. Petrella JR, Sheldon FC, Prince SE, et al: Default mode network connectivity in stable vs progressive mild cognitive impairment. Neurology 76(6):511–517, 2011. 41. Pfeuffer J, McCullough JC, Van de Moortele PF, et al: Spatial dependence of the nonlinear BOLD response at short stimulus duration. Neuroimage 18(4):990–1000, 2003. 42. Pillai JJ: The evolution of clinical functional imaging during the past 2 decades and its current impact on neurosurgical planning. AJNR Am J Neuroradiol 31(2):219–225, 2010. 43. Pillai JJ: The significance of streamlined postprocessing approaches for clinical FMRI. AJNR Am J Neuroradiol 34(6):1194–1196, 2013. 44. Pillai JJ, Mikulis DJ: Cerebrovascular reactivity mapping: An evolving standard for clinical functional imaging. AJNR Am J Neuroradiol 2014. 45. Power JD, Barnes KA, Snyder AZ, et al: Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage 59(3):2142–2154, 2012. 46. Power JD, Schlaggar BL, Petersen SE: Recent progress and outstanding issues in motion correction in resting state fMRI. Neuroimage 105C:536– 551, 2015. 47. Raichle ME, Snyder AZ: A default mode of brain function: A brief history of an evolving idea. Neuroimage 37(4):1083–1090, discussion 1097–1099, 2007. 48. Ranasinghe KG, Hinkley LB, Beagle AJ, et al: Regional functional connectivity predicts distinct cognitive impairments in Alzheimer’s disease spectrum. Neuroimage Clin 5:385–395, 2014. 49. Rashid B, Damaraju E, Pearlson GD, et al: Dynamic connectivity states estimated from resting fMRI identify differences among schizophrenia, bipolar disorder, and healthy control subjects. Front Hum Neurosci 8:897, 2014. 50. Rutten GJ, Ramsey NF, van Rijen PC, et al: fMRI-determined language lateralization in patients with unilateral or mixed language dominance according to the Wada test. Neuroimage 17(1):447–460, 2002.

51. Sabbah P, Chassoux F, Leveque C, et al: Functional MR imaging in assessment of language dominance in epileptic patients. Neuroimage 18(2):460–467, 2003. 52. Seeley WW, Crawford RK, Zhou J, et al: Neurodegenerative diseases target large-scale human brain networks. Neuron 62(1):42–52, 2009. 53. Sheline YI, Raichle ME: Resting state functional connectivity in preclinical Alzheimer’s disease. Biol Psychiatry 74(5):340–347, 2013. 54. Shimony JS, Zhang D, Johnston JM, et al: Resting-state spontaneous fluctuations in brain activity: A new paradigm for presurgical planning using fMRI. Acad Radiol 16(5):578–583, 2009. 55. Sporns O: Structure and function of complex brain networks. Dialogues Clin Neurosci 15(3):247–262, 2013. 56. Thomason ME, Foland LC, Glover GH: Calibration of BOLD fMRI using breath holding reduces group variance during a cognitive task. Hum Brain Mapp 28(1):59–68, 2007. 57. Tie Y, Rigolo L, Norton IH, et al: Defining language networks from resting-state fMRI for surgical planning—A feasibility study. Hum Brain Mapp 35(3):1018–1030, 2014. 58. Tomasi D, Volkow ND: Resting functional connectivity of language networks: Characterization and reproducibility. Mol Psychiatry 17(8):841–854, 2012. 59. Turner R, Le Bihan D, Moonen CT, et al: Echo-planar time course MRI of cat brain oxygenation changes. Magn Reson Med 22(1):159–166, 1991. 60. van den Heuvel MP, Mandl RC, Kahn RS, et al: Functionally linked resting-state networks reflect the underlying structural connectivity architecture of the human brain. Hum Brain Mapp 30(10):3127–3141, 2009. 61. Yeh CJ, Tseng YS, Lin YR, et al: Resting-state functional magnetic resonance imaging: The impact of regression analysis. J Neuroimaging 25(1):117–123, 2015. 62. Yetkin FZ, Hammeke TA, Swanson SJ, et al: A comparison of functional MR activation patterns during silent and audible language tasks. AJNR Am J Neuroradiol 16(5):1087–1092, 1995. 63. Yetkin FZ, Swanson S, Fischer M, et al: Functional MR of frontal lobe activation: Comparison with Wada language results. AJNR Am J Neuroradiol 19(6):1095–1098, 1998. 64. Zaca D, Hua J, Pillai JJ: Cerebrovascular reactivity mapping for brain tumor presurgical planning. World J Clin Oncol 2(7):289–298, 2011. 65. Zaca D, Pillai JJ: BOLD fMRI for presurgical planning: Part 1. In Pillai JJ, editor: Functional brain tumor imaging, New York, 2014, Springer Science + Business Media. 66. Zalesky A, Fornito A, Cocchi L, et al: Time-resolved resting-state brain networks. Proc Natl Acad Sci U S A 111(28):10341–10346, 2014. 67. Zhang D, Johnston JM, Fox MD, et al: Preoperative sensorimotor mapping in brain tumor patients using spontaneous fluctuations in neuronal activity imaged with functional magnetic resonance imaging: Initial experience. Neurosurgery 65(6 Suppl):226–236, 2009.

http://radiologyebook.com

17  Brain Proton Magnetic Resonance Spectroscopy Lester Kwock, Mauricio Castillo, Jay J. Pillai, and Alena Horská

INTRODUCTION The first reported in vivo localized proton (1H) magnetic resonance spectroscopy (H1-MRS) studies of the human brain were first reported over 20 years ago.37,109,132 Since these early studies, in vivo H1-MRS brain studies have been shown to be a powerful technique to noninvasively investigate the biochemistry of the human brain and to assess in situ the neurochemical profile of brain tissue. Thus they can provide biomarkers of neurologic disorders even in cases where lesions are not seen in conventional anatomic MRIs.21,96 The noninvasive quality of H1-MRS and the spatially localized MR techniques that have been developed to examine in vivo the neurochemical profile of brain tissue makes them suitable not only for diagnostic purposes but also longitudinal follow-up studies. They have proven to be important tools at many institutions for the noninvasive clinical assessment of numerous neurologic disorders such as brain tumors, epilepsy, white matter disease processes, metabolic disorders, and brain trauma.40,188,243,259,347 However, even though proton MRS has been incorporated into clinical protocols and is accepted at many institutions worldwide,253 it is still considered an investigational technique by some healthcare organizations, who indicate that at the present time there has been “no large multicenter trial published demonstrating any added benefit of MRS over MRI in diagnosing or monitoring pathological processes such as brain tumors, and no clinical trials demonstrating improved outcomes evaluated with MRS alone compared to patients evaluated with conventional imaging modalities have been reported.”3 What is not understood by these healthcare organizations is that MRS and MRI are not competitive techniques but rather complementary techniques that can “add value” to the MR diagnostic procedure and lead to better patient management decisions. MRS provides information about the metabolic profiles of lesions, whereas conventional MRI provides anatomic profiles, and in most cases it should not be used alone in making clinical diagnostic decisions but used in conjunction with the clinical history of the patient and information from conventional anatomic and advanced MRI techniques. As an example, Moeller-Hartmann and coworkers237 examined the diagnostic benefits of adding MRS to a conventional anatomic MRI study of intracranial mass lesions (i.e., infarctions, primary brain tumors, metastatic lesions, primitive neuroectodermal tumors, abscesses). In this study they found that with conventional MRI alone, 96 out of 176 correct diagnoses were made (55.1%). Adding H1-MRS information to the MRI findings led to an increase in the number of correct diagnoses from 96 to 124 out of the 176 cases (a 15.4% increase in correct diagnoses). By adding MRS to the MRI study, the number of incorrect diagnoses decreased from 27 (15.3%) for MRI alone to 16 (9.1%). Similarly, in a 2012 study165 it was shown that H1-MRS added value to conventional MRI information in the preoperative characterization of the type and grade of brain

tumors. This study found that information obtained from “H1-MRS significantly improved the radiologists’ MRI based characterization of grade IV tumors (glioblastomas (GBs), metastases, medulloblastomas, and lymphomas) in the cohort” (area under the receiver operating curve [AUC] from 0.85 in the MRI [alone] to 0.93 in the MRI plus MRS reevaluation [MRI plus MRS] and also in the less malignant glial tumors [AUC in the MRI reevaluation was 0.93 versus 0.81 in the MRI alone]).165 This chapter will not review the basic principles of MRS and in vivo H1-MRS/MRSI of the human brain. The fundamental principles of this technique have been presented in detail elsewhere.187,271 The main body of this chapter will concentrate on the complementary use of H1-MRS with advanced MRI techniques (i.e., diffusion-weighted imaging [DWI], perfusion-weighted imaging [PWI], and permeability), and conventional MRI to improve the accuracy of diagnosis and delineation of brain tumors and in therapeutic decision making. The major reason for this is because of the important findings and recommendations on the use of MRS recently published by the MR Spectroscopy Consensus Group.253 This group was formed in 2011 under the auspices of the International Society of Magnetic Resonance in Medicine and is composed of leading imaging scientists, neuroradiologists, neurologists, oncologists, and clinical neuroscientists from universities, as well as vendors from the United States, Europe, and Asia. This international group was charged with the task of documenting “the impact of H-1 MR spectroscopy in the clinical evaluation of the central nervous system.” One of the major conclusions and recommendations of this group was that “MR spectroscopy adds diagnostic and prognostic benefits to MR imaging and aids in treatment planning and monitoring of brain cancers,” and that “clinical imaging centers specializing in combined use of MR imaging and spectroscopy should be established in all major clinical neurologic centers that offer standardized MR spectroscopy procedures for improved patient management.”253 This conclusion was based on the substantial body of both basic science and clinical MRS research that has been performed over the past 2 decades worldwide, with consistent results found across laboratories. This chapter will demonstrate that MRS, especially if combined with advanced MRI techniques, (1) can improve the diagnostic accuracy identifying neoplastic from nonneoplastic brain processes, (2) can identify the presence or absence of specific metabolites in the 1H MR spectrum that can be used as biomarkers to identify various tumor processes, and (3) that the combination of conventional and advanced MRI and MRS can reduce the need for more invasive diagnostic procedures to assess the cellular characteristics of untreated and treated brain tumor processes. The MR paradigms described in the brain tumor studies can  serve as models for the potential development of combined MRS and MRI techniques to increase diagnostic accuracy and modify patient

http://radiologyebook.com

499

500

PART II  CT and MR Imaging of the Whole Body

treatment planning to improve the therapeutic outcomes in other central system disorders presented in this chapter and in the MR Spectroscopy Consensus Group report.253

H2O

H

H C

MRI VS. MRS There is no difference between the physical principals of MRI and MRS; both techniques are governed by the same principles of magnetism. MRI and MRS differ only in the manner in which the data obtained are analyzed and the type of information provided. In the case of MRI, data collected are analyzed in the time domain (namely, free induction decay signal; signal intensity vs. time) to obtain relaxation time (TR) information (namely, T1 [spin-lattice] and T2 [spinspin]) of the nuclei. The data from the time domain information is then used to generate an anatomic image. In MRS, the time domain information is converted to the frequency domain (signal intensity vs. frequency) via Fourier transformation of the free induction decay time domain signal. The frequency information generated via Fourier transformation of the time domain signal is used to form a distribution of the intensities of chemical groups associated with various metabolites versus their Larmor resonance frequencies (Fig. 17-1) to give a spectral profile of the metabolites within this region of the anatomic image. Thus the two MR techniques (i.e., conventional MRI vs. MRS) give different but complementary information about the region being examined. In the case of conventional MRI (namely, T1- and T2-weighted MRI), the information is mainly anatomic in nature, generated via water proton interactions with the tissues, whereas MRS gives information about the biochemical/metabolic profile of the anatomic region being examined. Advanced MRI techniques such as PWI312 and DWI255 give physiologic information about the vascularity and cellularity of the anatomic region being evaluated. This information is also complementary to the information obtained from conventional anatomic MRI studies and biochemical/metabolic MRS studies. The major advantage of employing MR techniques to assess a lesion is that these complementary diagnostic techniques are noninvasive and can be performed during the same examination session (normally < 1.5 hours) to obtain multiparametric diagnostic information.

HCOH

C

CH2

C

CH3

NH

8

6

4

2

0

Chemical shift, δ (ppm)

FIG 17-1  Resonance frequency positions of various proton groups. Chemical shifts are in parts per million of main magnetic field (i.e., at 1.5 T the central resonance frequency of protons is 64 MHz, each part per million [ppm] = 64 Hz).

NAA 0.08 Cr Cho

0.06

0.04 Cr2 0.02

0.00

4

3

A

2

1

ppm NAA

EVALUATION OF BRAIN METABOLITES

0.15 Cr

MRS Observable In Vivo Brain Metabolites Several important metabolites are evaluated in long echo time (TE) (135-288 milliseconds) proton MR spectra (Fig. 17-2): • N-acetylaspartate (NAA) • Choline (Cho) • Creatine/phosphocreatine (Cr/PCr) • Lactate (Lac) When short TEs (20-30 milliseconds) are used, a greater  number of metabolites can be identified in the MR spectra; in  addition to NAA, Cho, Cr, and Lac, the following may be identified34,51,92,149,240,274,290,296,354: • Glutamate (Glu) • Glutamine (Gln) • γ-Aminobutyric acid (GABA) • Myoinositol (MI) • Alanine (Ala) • Glucose (Gc) • Lipids and proteins • Scylloinositol/taurine Although it may appear advantageous to obtain spectra routinely at only short TEs to distinguish among different clinical entities, some

MI

0.10

0.05

Cho

Cr2 Glx

0.00 4

B

3

2

1

ppm

FIG 17-2  Normal brain spectra. A, Long echo TE = 135 ms. B, Short echo TE = 30 ms.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy disadvantages of short TE studies exist. For example, short TE spectra display greater baseline distortion, and estimating signal areas calls for more sophisticated processing software algorithms.248 However, to maximize metabolite information, both a short and long TE MRS study should be performed.92,132 N-Acetylaspartate.  NAA accounts for the majority of the NAA resonance at 2.01 ppm; this signal is the most prominent one in normal adult brain proton MRS and is used as a reference for determination of chemical shift.24,180,181 The NAA signal also contains contributions from other N-acetyl groups, such as N-acetylaspartyl glutamate (NAAG), N-acetylated glycoproteins, and amino acid residues in peptides.24,180,181 NAA is second only to Glu as the most abundant free amino acid in the normal adult brain.211 The function of this amino acid is not fully understood despite its early discovery in 1956 by Tallan.330,331 From animal studies, NAA is believed to be involved in coenzyme A (CoA) interactions and in lipogenesis within the brain.16,31,44,86 Specifically, such studies suggest that NAA is synthesized in the mitochondria from aspartate and acetyl CoA and transported into the cytosol where it is converted by aspartoacylase into aspartate and acetate.31,57,86 Although NAA is widely regarded as a nonspecific neuronal marker, it has also been detected in immature oligodendrocytes and astrocyte progenitor cells.296,346 Normal absolute concentrations of NAA in the adult brain are generally in the range of 8 to 9 mmol/kg, although regional and agerelated variations in NAA concentration have been noted by Kreis and others.178,229,274 In normal adults, NAA concentrations in cortical gray matter are higher than those in white matter; in infants the concentrations in gray and white matter are similar (highly active lipid synthesis in immature white matter accounts for this difference from the adult pattern).44,234 NAA concentrations are decreased in many brain disorders, resulting in neuronal and/or axonal loss, such as in neurodegenerative diseases, stroke, brain tumors, epilepsy, and multiple sclerosis, but are increased in Canavan’s disease.36

Creatine.  The main Cr signal is present at 3.03 ppm and demonstrates major contributions from methyl protons of creatine and phosphocreatine as well as minor contributions from GABA, lysine, and glutathione.180,181 A second, usually smaller Cr signal is seen at 3.94 ppm. Cr is probably involved in maintenance of energy-dependent systems in brain cells by serving as a reserve for high-energy phosphates in neurons and as a buffer in cellular adenosine triphosphate/diphosphate (ATP-ADP) reservoirs.38,53,57,145,178,231 Thus the Cr signal is an indirect indicator of brain intracellular energy stores. The Cr signal is often used as an internal reference standard for characterizing other metabolite signal intensities because it tends to be relatively constant in each tissue type in normal brain; however, this is not always true in abnormal brain tissue, particularly in areas of necrosis.205 Cr concentrations in the brain are relatively high, with progressive increases noted from white matter to gray matter to cerebellum.274,296 Kreis and coworkers noted a mean absolute Cr concentration in normal adult brains of 7.49 ± 0.12 mmol/kg on the basis of a sample of 10 normal subjects,178 whereas Michaelis and colleagues reported a similar value of 5.3 mmol/kg.167,229 Total Cr values tend to be abnormally reduced in brain tumors, particularly malignant ones.52

Choline.  The Cho resonance is present at 3.2 ppm and is attributable to trimethyl ammonium residues of free Cho as well as phosphocholine, phosphatidylcholine, and glycerophosphocholine. This signal reflects cell membrane synthesis and degradation.36 Thus all processes

501

resulting in hypercellularity (e.g., primary brain neoplasms or gliosis) or myelin breakdown (demyelinating diseases) lead to locally increased Cho concentration, whereas hypomyelinating diseases result in decreased Cho levels.36,52 Kreis and coworkers have reported a mean absolute Cho concentration in normal adult brain tissue of 1.32 ± 0.07 mmol/kg.178 Michaelis and others have reported a similar value of 1.6 mmol/kg.167,229

Myoinositol.  MI produces two signals noted at 3.56 ppm and 4.06 ppm. MI is the major component of the signal at 3.56 ppm, although contributions from MI-monophosphate and glycine are also present.290 MI is believed to be a glial marker because it is present primarily in glial cells and is absent in neurons.38 A role in osmotic regulation of the brain has been attributed to MI.296 In addition, MI may represent both a storage pool for membrane phosphoinositides involved in synaptic transmission and a precursor of glucuronic acid, which is involved in cellular detoxification.51,53,151,177 A derivative, MI-1,4,5-triphosphate, may act as a second messenger of intracellular calcium-mobilizing hormones.53,60 The mean absolute concentration of MI in normal brain tissues obtained in the Kreis series was 6.56 ± 0.43 mmol/kg.178 MI concentrations are abnormally increased in patients with demyelinating diseases, Alzheimer’s disease, and low-grade brain tumors.53,54,229

Lactate.  Lac resonance is identified as a doublet (splitting into two distinct resonant signals separated by 7 Hz), produced by magnetic field interactions among adjacent protons (referred to as J-coupling) centered at 1.32 ppm. A second Lac signal is present at 4.1 ppm but tends to be inconspicuous on spectra obtained with water suppression owing to its proximity to the water signal.53 Because Lac levels in normal brain tissue are absent or extremely low (<0.5 mmol/L), they are essentially undetectable on normal spectra.296 The presence of a visible Lac signal constitutes a nonspecific indicator of cellular anaerobic glycolysis, which may be seen with brain neoplasms, infarcts, hypoxia, metabolic disorders, or seizures.296 Lac may also accumulate within cysts or foci of necrosis.53,296 Changing the TE using a point-resolved spectroscopy (PRESS) sequence enables confirmation of the presence of an abnormal Lac doublet and differentiation from lipid signals; at TE = 20 and 272 milliseconds the Lac doublet projects above baseline, and at TE = 136 milliseconds it is inverted below the baseline.53 In encephalopathic neonates, however, a doublet very similar to Lac may be seen at 1.15 ppm, which corresponds to propan-1,2-diol and may easily be mistaken for Lac; propan-1,2-diol is a solvent commonly used for administration of anticonvulsant medications to neonates.25

Glutamate, Glutamine, and GABA.  The glutamate and glutamine (“Glx”) and GABA signals are a complex set of resonances noted between 2.1 and 2.5 ppm and consist of Glu, Gln, and GABA components. Glu is an excitatory neurotransmitter involved in neuronal mitochondrial metabolism and is the most abundant amino acid in the human brain.53,57,349 Gln is involved in both cellular detoxification and regulation of neurotransmitter activities.53,229 GABA is a product of decarboxylation of Glu via glutamic acid decarboxylase and serves as an inhibitory neurotransmitter.296 Abnormalities in this signal complex have been noted in schizophrenia and epilepsy.57

Lipids and Proteins.  Lipids produce multiple resonances, the most important of which are noted at 0.8 to 0.9 and 1.2 to 1.3 ppm because of methyl and methylene protons of fatty acids, respectively.53,157,272,296 Membrane lipids are not usually identified unless very short TEs are employed, because they have very short TRs and are not normally

http://radiologyebook.com

502

PART II  CT and MR Imaging of the Whole Body

observed.53 Although artifactual accentuation of these signals may be seen with voxel contamination by adjacent subcutaneous fat, they can be eliminated by using outer volume suppression saturation pulses. High-grade gliomas (HGGs), meningiomas, demyelinating processes, necrotic foci, and inborn errors of metabolism can show the presence of lipids owing to the presence of lipid droplets within necrotic tumors, loosening of the lipid-rich myelin sheath, and breakdown of plasma membrane structures of cells.296

Quantitation of Metabolite Levels Currently, metabolite concentrations are most often presented as peak area ratios of the metabolite resonance relative to the peak area resonance of Cr at 3.03 ppm rather than absolute concentrations. Despite the success of this qualitative approach in defining changes in metabolite levels in disease processes, it has recently been suggested that relative quantitation may introduce substantial errors and lead to erroneous metabolite values.201,212 Alterations in the peak area ratios do not necessarily reflect a change in the metabolite; these can be due to a change in the metabolite of interest or Cr levels or both, which will affect the ratio or possibly even change the TR behavior of the metabolites, which will also affect the ratio. This needs to be understood when interpreting changes in relative peak area ratios. If changes in the absolute metabolite concentrations are essential in the interpretation of the spectral data, then relevant methods to obtain absolute metabolite concentrations in either single-voxel or multivoxel spectroscopy studies have been described and are available and readily applied.159

CLINICAL APPLICATIONS OF PROTON MRS Evaluation of Normal Brain Development and Regional Brain Differences Although most neurons are formed during intrauterine life, proliferation of dendrites, astroglia, and oligodendroglia continues after birth.178,350 The processes of myelination and neuronal and dendritic development are associated with changes in various brain metabolites during infancy and early childhood, which are demonstrated by MRS.153,155,172,350 MI and Cho form the most prominent signals in the newborn MR spectrum.178 The reason for the prominence of these two components in the infant spectrum is presumed to be the presence of active myelination.178,350 MRS may enable earlier detection of myelination abnormalities in infants than is possible with conventional MRI sequences, because changes in the Cho and NAA signals occur before white matter signal abnormalities on standard imaging sequences.178,350 Age-related differences in metabolite levels were reported in both gray and white matter during brain development.70,178,273,340,353 NAA concentration, which is very low in the newborn brain, rapidly increases during the first few months of life and continues to increase until 2 to 3 years. Preterm infants were noted to have lower brain NAA levels than full-term infants.353 Cho levels in newborns are twice as high as in adults.153,178 After birth, Cho concentration exponentially decreases and reaches adult levels by approximately 2 to 3 years of age.178 The high MI levels decline rapidly in the first 3 to 4 months and decrease to adult levels by approximately 1 year of age.178,296 Cr levels rise to adolescent levels by 4 months of age, presumably as a result of increasing energy demands of the developing brain.281,340 Glu levels do not exhibit any development-related differences.273 The scylloinositol signal is high in newborns and decreases over time.138,151,228 Agerelated differences in metabolite levels are less pronounced during childhood and adolescence, and the spectral pattern stabilizes 

during early adulthood.143,273 A meta-analysis evaluating literature data on metabolite levels in normal aging (in subjects > 60 years) reported decreased frontal NAA and increased parietal Cho and Cr with aging.131 Lac signal is seen in preterm and small-for-gestational-age infants.19,250 However, in appropriate-for-gestational-age term infants, it is abnormal to find higher than trace amounts of Lac, particularly following the first few hours of life, and such abnormal amounts of Lac signify brain injury.25 In preterm infants, the levels of Lac decrease progressively until the age of 40 postconceptional weeks.25,197,201,264 Significant regional heterogeneity in spectral patterns has  been noted both in the developing brain and in mature  brains.7,136,166,178,273,353,359 In the developing brain, proton MR spectra demonstrate that different parts of the brain mature at different rates and at different times and that the more metabolically mature areas demonstrate lower MI and higher NAA levels than the less mature regions of brain; specifically, the basal ganglia, perirolandic cortex, and visual cortex mature before areas such as the frontal white matter, prefrontal cortex, and temporal cortex.25,155,207,353 NAA/Cho ratios in the frontal region were reported to be lower than those in the parietal lobe at birth, with subsequent increase during the first 6 months of life.7,154 Lower NAA and higher Cho levels were detected in the allocortex compared to the isocortex in healthy children and adults.8,71 Tedeschi and colleagues have noted significant regional variations in metabolite ratios in adults335; NAA/Cho and NAA/Cr ratios varied by more than a factor of 2 for different brain regions, with high levels of Cho and Cr and low levels of NAA found in the cerebellum.158,359 High Cho levels were also reported in insular cortex, thalamus, and centrum semiovale compared to parietal or occipital gray and white matter; high NAA levels were detected in the centrum semiovale.22 In the mesial temporal lobe, the anterior mesial temporal lobe was found to have higher Cho concentration than the middle and posterior mesial temporal lobe.8

Evaluation of Primary Brain Tumors (Gliomas) Many different types of brain tumors have been studied with MRS, and fairly consistent patterns of metabolic abnormalities have been described in both glial and nonglial tumors.18,278 Proton MRS allows reliable differentiation of tumor margins from adjacent brain edema, which is often not possible with gadolinium (Gd)-enhanced MRI,52 which may under- or overestimate tumor size in approximately 40% of cases.252 A multitude of applications of proton MRS exist in the field of brain tumor diagnosis and management, and MRS is clearly playing an increasingly valuable clinical role that complements the roles of conventional structural MRI and other advanced MRI techniques such as PWI and diffusion tensor imaging (DTI).

Noninvasive Histologic Grading of Gliomas.  Although it is possible to clearly distinguish glial neoplasms from normal brain tissue by MRS, controversy exists regarding the reliability of MRS in distinguishing among different histologic grades of astrocytomas and other brain tumors.53,183,247,311 For example, Shimizu and associates demonstrated that through semiquantitation of MRS signal intensities as a ratio to that of an external reference, it was possible to predict tumor malignancy15,311,310; higher-grade brain tumors were associated with higher Cho/reference values and lower NAA/reference values in their series.318 This, in conjunction with findings from Tamiya’s group regarding high correlation between Cho/NAA and Cho/Cr ratios and cellular proliferative activity,332 provides substantial evidence that proton MRS may be useful in the prediction of tumor grade.199

http://radiologyebook.com

503

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy In a series involving children, linear discriminant analysis  and proton MRS demonstrated an 83% success rate in establishing the correct diagnosis of histologic grade of brain tumors.358 In a series of 27 patients with biopsy-confirmed brain tumors,  Meyerand and colleagues showed that the combination of Lac/water and Cho/water ratios obtained from regions of contrast-enhancing brain tumor (with normalization of each metabolite signal area  to the area of the unsuppressed water signal) permitted differentiation of low-grade astrocytomas from anaplastic astrocytomas  and GBs.226 However, not all studies have confirmed the ability of proton  MRS to differentiate between different tumor grades. For example, Burtscher and colleagues have described a series of 26 intracranial tumors in which MRS allowed differentiation of infiltrative processes from circumscribed lesions but did not allow differentiation of different types of lesions within each category.45 In a multicenter study involving 86 cases of glial tumors, Negendank and coauthors showed that each type and grade of tumor was a metabolically heterogeneous group with significant overlap in spectral NAA/Cr and Cho/Cr ratios with tumors in other categories; all tumors demonstrated abnormally decreased NAA/Cr and increased Cho/Cr ratios with respect to normal brain parenchyma.247 Some of the reasons for these findings in early studies (i.e., before the year 2000) regarding the inability of proton MRS to reliably differentiate various tumor grades have included inconsistencies in quality of spectroscopic data among different institutions, use of only long TE spectra, and use of only singlevoxel technique; many of these problematic issues have been addressed in more recent studies.249 Studies have suggested that advanced physiologic MRI methods in conjunction with MRS may be more accurate for determination of tumor grade than MRS or anatomic MRI alone. For example, Law and colleagues demonstrated that relative cerebral blood volume (rCBV) perfusion MRI combined with proton MRS metabolite ratios data improved both sensitivity and positive predictive value (PPV) for the determination of glioma grade over structural MRI alone.192 The combination of rCBV, Cho/Cr, and Cho/NAA resulted in sensitivity, specificity, PPV, and negative predictive value of 93.3%, 60%, 87.5%, and 75%, respectively, compared to 72.5%, 65%, 86.1%, and 44.1%, respectively, with conventional MRI alone. The low specificity was due in part to the fact that high Cho levels were found in some low-grade lesions. El-Sherbeny and coworkers100 recently showed that inclusion of MI/Cr ratios in the MRS analysis in conjunction with the MRI information obtained from conventional and advanced MRI techniques added additional diagnostic value to the differentiation of the grade of the brain tumor. A number of studies have shown that MI/Cr peak area ratios over 0.6 are highly characteristic of low-grade gliomas.29,54,148 With the addition of MI/Cr ratios to MRS data, the sensitivity and specificity of the MRI/MRS paradigm employed to determine the grade of the lesion was 91% and 90.5%, respectively, compared to  the study of Law et al.,192 which found a 93.3% sensitivity and only a 60% specificity. The study of El-Sherbeny examined 93 patients  with histologically confirmed brain tumors (31% low grade and 69% high grade). More recently, Caulo and colleagues18 reported a quantitative multiparametric MRI evaluation of 110 patients with either low- or highgrade gliomas (rCBV, apparent diffusion coefficient [ADC]), MRS (Cho/Cr and Cho/NAA), and conventional MRI that incorporated the analysis of four heterogeneous MRI tumor regions. They found that taking into account the heterogeneous nature of the gliomas and combining the studies with both advanced and conventional MR techniques significantly improved the discrimination between low- and

high-grade brain gliomas (sensitivity of 84.4%, specificity of 100%, and an AUC of 0.95). A meta-analysis by Hollingworth and colleagues published in 2006 explored the overall clinical utility of MRS in the characterization of brain tumors and shed some important light on the potential role of MRS in the differentiation of various tumor types.140 It concluded that MRS was accurate in differentiating high- and low-grade tumors.140 Specifically, based on receiver operating characteristic (ROC) curve analysis of multiple studies examining sensitivity and specificity of proton MRS for differentiating high- from low-grade tumors, the study found excellent values for this purpose in the AUC in multiple studies.140 In studies by Devos and Lukas and colleagues, AUC values of 94% and 96% in studies using long and short TE, respectively, were obtained.91,207 In Hollingworth’s meta-analysis, two studies also showed excellent ability of proton MRS to differentiate between high- and low-grade gliomas; specifically, Herminghaus and coworkers showed in their study of 90 patients that the sensitivity and specificity of proton MRS for the differentiation of high-grade from low-grade tumors were 95% and 93%, respectively.135,140 In addition, Astrakas and associates demonstrated in their cohort of 66 patients that by using a combination of MRS parameters (namely Cho/Cr and Lac/Cr; threshold value = 1.8), an AUC value of 96% was obtained for the sensitivity and a specificity of 88% for differentiation of high-grade from lowgrade tumors.9,141 Astrocytomas typically demonstrate reduced NAA levels, moderately reduced Cr levels, and elevated Cho levels compared to normal brain parenchyma (Fig. 17-3 and Table 17-1).43 These absolute reductions result in abnormally low NAA/Cr ratios and elevated Cho/Cr ratios (see Table 17-1).296 NAA levels in astrocytomas are reduced to 40% to 70% of the levels seen in normal brain parenchyma.220 Furthermore, Cho has been reported to be substantially elevated in more malignant astrocytomas, and this may be secondary to increased cellularity and cell membrane synthesis.220,350 There is evidence to suggest that the elevation of Cho is proportional to the degree of tumor malignancy.52 However, highly malignant primary brain tumors with extensive necrosis may actually demonstrate decreased levels of Cho. Elevated Cho levels are seen more consistently in anaplastic astrocytomas (which do not demonstrate histologic evidence of necrosis) than in glioblastoma multiforme (GBM), and ependymomas display higher Cho/Cr ratios than those noted for astrocytomas in general.52,358 Lac levels may be elevated in some astrocytomas as well. However, a higher incidence of Lac in more aggressive or higher-grade astrocytomas is controversial because Lac may also be present in benign pilocytic astrocytomas.156,247,252,344 It is thought that Lac may be present

TABLE 17-1  Choline/Creatine and

Myoinositol/Creatine Proton Metabolite Peak Area Ratios in Normal Brain and Gliomas TE = 135 msec Normal (n = 20) Grade 1 (n = 5)* Grade 2 (n = 5)† Grade 3 (n = 5) Grade 4 (n = 10)

TE = 30 msec

(Cho/Cr)

(Cho/Cr)

(MI/Cr)

1.08 ± 0.19 2.25 ± 0.44 1.66 ± 0.28 2.72 ± 0.49 3.23 ± 1.06

0.91 ± 0.15 1.57 ± 0.26 1.28 ± 0.63 2.06 ± 0.36 2.15 ± 0.66

0.42 ± 0.16 1.95 ± 0.74 1.02 ± 0.13 0.49 ± 0.21 0.33 ± 0.17

*WHO grade 1 lesions were biopsy-confirmed pilocytic astrocytomas. † WHO grade 2 lesions were confirmed oligodendrogliomas.

http://radiologyebook.com

Myoinositol

Cho

Cho

1.5

Ins dd1 Ins dd2

1.0

0.5

0.10

Cr

Cr

Gtx m3

Gtx dd Cr2

NAA NAA NAA dd2 dd1 Gtx m1 Gtx m2

0.05 Cr2 NAA

0.0

0.00

4

3

2

1

4

ppm TE = 20

A

3

2 ppm TE = 135

Cho 0.4

0.3 Cr

0.2 Cr2 0.1

NAA 0.0 4

3

B

2

1

ppm

Cho

0.6 Cr NAA

0.4 Cr2

0.2

Ins dd1

0.0 4

C

3

2

1

ppm

FIG 17-3  A, Grade II astrocytoma. A representative single-volume proton MR spectrum of a histologically proven low-grade glioma using a STEAM sequence for the TE = 20 msec spectrum and a PRESS sequence for the TE = 135 msec spectrum. The TR used was 1500 msec in both the TE = 20 and TE = 135 msec studies. Note the intensity of the myoinositol (3.56 ppm) and choline (3.2 ppm) resonances in the low-grade lesion compared to the higher-grade gliomas. B, Glioblastoma multiforme (GBM). Representative singlevolume proton MR spectrum of a histologically proven GBM using a PRESS sequence at TE = 135 msec and TR = 1500 msec. C, TE = 30 msec proton spectrum of GBM. Note the lower myoinositol resonance compared with the grade II lesion and the elevated choline signal at both TE = 30 and 135 msec.

http://radiologyebook.com

1

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy across the entire spectrum of grades of astrocytomas because Lac may arise not only from hypoxia developing within the tumor itself as a result of disruption of normal vascular networks but also from necrosis or cysts within the tumor.52,53,311 Lac levels may be elevated in all cysts regardless of etiology.52 The presence of Lac has been attributed to (1) the extent of anaerobic glycolysis, (2) the efficiency of electron transport, and (3) the rate of washout from tumor tissue.156,202,296 In highly vascular tumors, Lac may be rapidly removed from the tumor as a result of increased blood flow; therefore, even in high-grade vascular tumors, Lac may not be present in the MR spectra.52 The value of Lac identification following radiation therapy or surgery is controversial, but the Lac concentration may be proportional to the original tumor size.52,338 Lac has been reported following radiation therapy and surgery, including stereotactic biopsy, although postsurgical porencephaly may lead to artifactual increases in Lac levels because cerebrospinal fluid (CSF) is known to be rich in Lac.52,208 The presence of elevated lipid levels has also been used to differentiate low-grade from high-grade neoplasms.202,246,251,272,344 Some studies suggest that mobile lipids tend to be present in higher-grade neoplasms, with highest levels noted in GB; however, although high levels of mobile lipids may be specific for anaplastic astrocytoma or GB, their absence in MR spectra does not exclude the possibility of such high histologic grades.247 Lipids may originate from tumor cells within highgrade astrocytomas, macrophages along the tumor periphery, or areas of necrosis.52 Lipids may be present as a result of microscopic cellular necrosis, which may not be apparent on conventional MRIs.52 There are inherent limitations of histopathologic grading of gliomas that make it a somewhat imperfect gold standard for tumor assessment. For example, there is the intrinsic sampling error associated with current stereotactic biopsy, which is frequently based on enhancing components of tumor, when HGGs are known to infiltrate nonenhancing parenchyma following the vascular channels of white matter tracts.170,192 Furthermore, in adults the biological behavior of low-grade astrocytomas with identical histologic grades is heterogeneous, with more than half of low-grade tumors eventually recurring as or evolving into more aggressive tumors.241 Some investigators believe that differences in MRS ratios of various metabolites, like differences in glucose metabolism in positron emission tomography (PET), may be of prognostic significance even if they are not of diagnostic value in such cases of low-grade neoplasms.142,170,171,247,261 Physiologic MRI, such as proton MRS, not only allows whole lesion evaluation but also allows for evaluation of metabolic abnormalities beyond the tumor margins.192

Defining and Delineating Metabolically Different Tumor Regions A major potential role for proton MRSI is in treatment planning and monitoring of tumor response. Clinical oncologists face a challenge when attempting to implement a treatment protocol that not only will effectively treat the patient’s central nervous system (CNS) tumor to improve survivability but also maintains the patient’s neurologic functions and avoids major treatment-associated morbidities. The capability of MRI to obtain anatomic images in any orientation and to vary between T1- and T2-weighted contrasts has greatly aided in identifying regions containing tumor and in the planning and monitoring of surgical and radiotherapeutic treatments. Furthermore, the use of intravenous Gd-containing contrast agents along with perfusion and water diffusion MRI techniques allow more accurate differentiation of tumor from edematous regions. However, confirming the presence and extent of malignant disease with conventional MR techniques is still problematic. This is especially true for highly diffuse primary brain tumors. For instance, not all

505

visible MR Gd contrast enhancement may correspond to the presence of tumor. Other biological processes besides the presence of a tumor can cause breakdown of the blood-brain barrier, which leads to  contrast enhancement. Tumefactive inflammatory processes and treatment-induced necrosis are known to cause breakdown in the blood-brain barrier, leading to an MR appearance suggestive of tumor process.107 In addition, absence of Gd contrast enhancement does not always indicate that tumor is not present. WHO grade 2 gliomas commonly do not exhibit Gd contrast enhancement,89 and furthermore the heterogeneity of HGGs suggests that some regions of the tumor may enhance while other regions of the same tumor do not.200,279 Therefore treatments based solely on Gd contrast–enhanced T1-weighted MRIs to delineate the extent of the tumor may not target the full extent of the tumor and/or the most active regions of the tumor.277 Similar ambiguities with T2-weighted and FLAIR MR tumor imaging have also been found.324 Accurate diagnosis and delineation of tumor extent is critical to the clinical management of patients with intracranial tumors. Both surgical and radiotherapy tumor treatment plans are based on the conventional MRI techniques described previously and have been shown to be inadequate by themselves to accurately diagnose and delineate tumor extent. Depending on the tumor type, these MR techniques can only diagnose intracranial lesions with a 30% to 90% success rate.370 Biopsy is still considered the gold standard for determining cancer type and degree of malignancy. Once the lesion is diagnosed as a malignant brain tumor, planning of the surgical and radiotherapy treatments is based solely on the anatomic abnormalities observed on the conventional MRI studies. These conventional MRI studies, as indicated earlier, do not always provide an accurate picture of the extent or location of the tumor.195,268,277 This is especially true in determining the zone(s) of tumor infiltration with low to moderate rates of proliferation, regions of nonmigrating actively proliferating tumor, and regions of quiescent pseudopalisading migrating hypoxic tumor cells adjacent to necrotic regions.39,72 Usually, large uniform surgical and radiotherapy margins between 1 and 4 cm are treated to account for these regions.195,324,370 These margins may underestimate the extent of tumor infiltration, treat areas devoid of tumor (which increases the risk for normal tissue toxicity events), and/or undertreat tumor cells that are not actively proliferating in the planned treatment volumes. To obtain the most benefit from the newer neurosurgical and radiotherapy treatment approaches, such as computer-guided stereotactic neurosurgical techniques and the use of intensity-modulated and conformal radiotherapy treatment techniques, it is important that regions identified for special attention be defined accurately. Areas of active tumor proliferation need to be identified from areas suspicious for tumor extension, which are less proliferative but more infiltrative from areas that contain quiescent pseudopalisading migratory tumor cells around regions of necrosis. Noninvasive techniques that can provide information to improve our ability to characterize the genetic and molecular differences in the tumor populations present in a lesion and delineate the spatial extent of each population need to be utilized. This information, especially in HGG patients, may improve our ability to treat these lesions more accurately with the techniques now available and aid in the development of more targeted therapies to obtain better patient outcomes. In 2001, Dowling et al.95 took an important step in establishing MRS as a potential technique to aid in the mapping of tumor regions. Using a three-dimensional (3D) MRS technique, they evaluated the 3D MR spectra obtained from 28 brain tumor patients and correlated the histologic findings from specific sites to the corresponding MR spectral voxels. Their results showed the following:

http://radiologyebook.com

506

PART II  CT and MR Imaging of the Whole Body

(1) If the level of NAA is normal, the tissue within that voxel is normal. Conversely, abnormal levels of NAA correlate with abnormal tissue regardless of the underlying histologic findings (Figs. 17-4 and 17-5). (2) When a lesion contains a Cho peak area intensity greater than the peak area intensity of NAA and larger than the normal peak area intensity of Cho, tumor was always present (see Figs. 17-5 and 17-6). (3) The peak area intensity of Cho correlated with percentage of tumor in the sampled voxel. (4) Undetectable to near-normal levels of NAA and Cho are seen in areas of gliosis and necrosis (Fig. 17-6). (5) Image areas with similar appearance and enhancement may show different spectral patterns and histologic characteristics (see Fig. 17-6). These observations suggest that 3D MRSI could be a potentially important method that can be used to define more accurately tumor margins, identify areas with the highest proliferative fractions, and identify posttherapy sites of tumor recurrence. For example, Pirzkall et al.268 studied whether the metabolic information provided by MRSI-defined volumes could improve the accuracy of target volumes for radiotherapy treatment planning compared to target volumes defined by MRI. Using elevated Cho and decreased NAA as criteria, Pirzkall et al.268 compared the MRSI-defined volumes to that defined by the Gd-enhanced T1- and T2-weighted MRI studies. They found that the T2 studies estimated, as judged by the volume of the hyperintense T2 region, the region at risk of containing microscopic disease as being as much as 50% greater than by MRSI. On the other hand, the Gd-enhanced T1 MRI studies suggested a smaller tumor volume and a different location for highly active disease compared to the MRSI study (Fig. 17-7). They concluded that “the use of MRSI to define target volumes for RT treatment planning would increase, and change the location of the volume receiving a boost dose as well as reduce the volume receiving a standard dose.” This study suggests that inclusion of MRSI data in the treatment planning process could

potentially improve tumor control while reducing normal tissue radiation complications. This was clearly shown to have occurred in an external beam radiation therapy study of 26 patients with grade 4 gliomas in a study conducted by Chan et al.61 Patients with MRSI abnormalities (i.e., elevated Cho/NAA levels) outside of the MRIdefined target volume had decreased median survival times relative to those with MRSI abnormalities inside the MRI-defined target volumes (10.4 vs. 15.7 months). Applying a multiparametric MR scheme using information from H1-MRSI, PWI, DWI, and conventional MRI, Di Constanzo et  al.93 found that the combination of these techniques provided useful information that could be used to further define and improve the delineation and extent of tumor processes. In particular, this multiparametric study characterized more fully some of the more extreme spatial heterogeneities of tumor and could identify patterns for just tumor and edema alone, tumor and edema patterns in perienhancing regions of interest (ROIs) with abnormal signals on conventional MRI, and a tumor infiltrated pattern in apparently normal-appearing perienhancing ROIs. More importantly, this study suggested that changes in the MR parameters NAA, Cho, ADC, and rCBV could accurately classify the different perienhancing ROI patterns and be used reliably for guiding stereotactic biopsies, surgical resections, and radiation treatments of these regions.

MRS Spectral Profile of Proliferative, Hypoxic/ Pseudopalisading, and Infiltrating Glioma Populations One of the major factors that contributes to the poor therapy outcome in HGG patients is the subjective grouping of all HGGs by only  their histologic characteristics, without considering the differences in molecular and genetic characters of different tumor populations within the observed HGG—namely, the main proliferating and stationary tumor body, the infiltrative population of HGG (which has a lower proliferation rate), and the quiescent pseudopalisading migratory tumor population adjacent to necrotic regions of the lesion. The

42Y NAA 0.3 Cho 0.2

0.1

Cr

Cr2

0.0 ppm 4

3

2

1

FIG 17-4  Proton MRSI of uninvolved brain region adjacent to low-grade lesion. Note level of NAA in abnormal FLAIR tumor region vs. uninvolved tumor region shown in square.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy

507

Cho

0.4

0.3 Cr

NAA

0.2

0.1

Cr2

0.0 ppm 4

A

3

2

1

Cho 0.3

0.2 Cr

0.1 NAA

Cr2 0.0

ppm

B

4

3

2

1

FIG 17-5  Infiltrative grade 4 glioma (A,B). Note level of Cho relative to Cr and NAA in enhancing and nonenhancing brain regions. Both spectra are indicative of tumor being present.

molecular and/or genetic characteristics of these populations can affect different proliferative and/or infiltrative signaling pathways that are not normally considered when designing treatment plans. For example, with radiotherapy treatment of HGG, concomitant temozolamide chemotherapy is delivered; the reason for this is the mechanism of action of this drug. Temozolamide is a DNA alkylating agent that induces nicks in DNA, and if not repaired, daughter cells can inhibit replication, leading to increased apoptosis and increased radiation sensitivity. Inhibition of cellular replication occurs in the G2-M phase of the cell cycle,267 the most radiosensitive phase of the reproductive cycle. Thus the major effect of temozolamide is on tumor cells

in the most radioresponsive portion of the cell cycle. However, the level of response to temozolamide has been shown to be dependent on the genetic character of the HGG.221 Deletion on chromosome 10 (PTEN) of the phosphatase and tensin homolog leads to increased sensitivity to temozolamide. PTEN is a lipid phosphatase that plays a basic role in attenuating the phosphatidylinositol 3-kinase (P13K)/ Akt-1 signaling pathway; hence loss of PTEN can affect major oncogenic events during glioma genesis. Loss of PTEN occurs in about 36% of patients diagnosed with grade 4 gliomas; thus only about a third of the patients diagnosed with HGG will benefit from temozolamide treatment.193

http://radiologyebook.com

508

PART II  CT and MR Imaging of the Whole Body

Choline

A

3.5

3.0

2.5

2.0 ppm

1.5

1.0

3.5

B

3.0

2.5

2.0 ppm

1.5

Cho/tNAA

FIG 17-6  SPGR (spoiled gradient recalled acquisition in steady state) image and proton MRS spectral location of a patient with an oligoastrocytoma. A, MRI location and the MRS spectrum for a biopsy that yielded 75% tumor. Note the elevated Cho resonance and absence of NAA in this voxel. B, MRI location and MRS spectrum for a biopsy that correlated with the pathologic finding of necrosis, astrogliosis, and white matter. Note the low levels of Cho, Cr, and NAA resonances in this voxel region. Although the MRI intensity is similar to the region in A, significant differences in both the spectra and pathologic findings were found in these regions. (From Dowling C, et al: Preoperative proton MR spectroscopic imaging of brain tumors: Correlation with histopathologic analysis of resection specimens. AJNR Am J Neuroradiol 22:604–612, 2001.)

2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 Glioma grade II

A

B

Glioma grade III

C

FIG 17-7  Transverse T2-weighted (6490/98) MRIs in (A) patient 6 (astrocytoma grade II) and (B) patient 7 (oligodendroglioma grade III) superimposed with color-coded segmented Cho/NAA ratio image. Voxels in predominantly red areas (enclosed by black line) were determined as voxels in the tumor center; all other voxel positions were determined as voxels in the tumor border (enclosed by white line). Green lines show volume of interest of the MRS examination. Violet and blue, minimum value; red, maximum value. C, Box plot shows a significant and definite difference (P = .001) for quantified Cho/tNAA ratio averaged over the tumor center between patients with glioma grades II (n = 9) and III (n = 17). The central box represents values from lower to upper quartile (25th-75th percentile), the middle line represents the median, and vertical bars extend from minimum to maximum value. Small black squares show individual data points. (From Stadlbauer A, et al: Preoperative grading of gliomas by using metabolite quantification with high-spatialresolution proton MR spectroscopic imaging. Radiology 238:958–969, 2006.)

http://radiologyebook.com

1.0

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy However, what about the two other tumor populations present in the HGGs in patients who respond to temozolamide—namely, the pseudopalisading migratory cells adjacent to necrotic regions and diffusively infiltrating tumor cells? Both of these tumor cell types should be refractory to temozolamide, because neither proliferate rapidly.72 Around areas of necrosis in HGG, the tumor cells show psuedopalisading.72 Data suggest that the pseudopalisading zone is the result of tumor cells migrating away from the necrotic region, and that necrosis selects for tumor cells that are more aggressive and more resistant to apoptosis-inducing therapeutic regimens.284 Similarly, data indicate that glioma cells that migrate out of the main tumor body have a decreased proliferative capacity and decreased susceptibility to apoptosis.164 These two populations may be the tumor cells responsible for regrowth and recurrence of HGG. Thus more information should be obtained that characterizes molecular signaling pathways activated in these populations, which will allow for more effective therapies to be developed. In the case of the migratory HGG cells, evidence indicates that treatment with inhibitors of migration does not affect apoptosis of migration-restricted/stationary glioma cells but can sensitize migrating glioma cells to cytotoxic agents.164

Proliferative Glioma Population: Cho/NAA Ratio, ADC, and rCBV Parameters.  At present, the major role of MRSI in the surgical treatment planning for HGG is in guiding the neurosurgeon to regions that show high metabolic activity on biopsy (regions with elevated Cho levels and low NAA levels relative to normal brain tissue).95,223 In a study conducted by Gupta and colleagues,130 18 glioma patients were preoperatively examined with H1-MRSI and DWI along with conventional MR studies. This study was conducted to correlate quantitative presurgical H1-MRSI and DWI results with quantitative histopathologic characteristics of resected tumor tissue. The primary hypotheses to be tested were: “(1) glioma choline correlates with cell density, (2) glioma apparent water diffusion coefficient (ADC) correlates inversely with cell density, and (3) glioma choline signal correlates with cell proliferative index” as measured by MIB-1 immunostaining. The tumor-to–contralateral normalized Cho signal ratio (nCho = Chotumor voxel/Chocontralateral normal volume) and the ADC from resected tumor regions were determined from the preoperative MRI studies. The results showed that cell density correlated linearly with nCho but inversely with ADC values. No correlation was found between nCho and cell proliferative capacity. However, in a similar study conducted by McKnight et al.,223 using a normalized Cho/NAA (CNI) and Cho/Cr (CCrI) index instead of nCho index, they found that the normalized Cho/NAA but not the Cho/Cr ratios correlated significantly with the proliferation index of the tumors as determined by MIB-1 immunostaining (P < 0.001) and cell density measurements (P < 0.001) of the biopsy-obtained voxel sample. This finding suggests that voxels containing elevated normalized Cho/NAA ratios have the highest probability of being a region that has “high cell density, high proliferative fraction, and/or high ratio of cell proliferation to cell death.”223 As discussed previously, for surgical, chemotherapeutic, and radiation therapy clinical patient treatment decisions it is important to be able to delineate and identify, as confidently as possible, the various heterogeneous regions of the tumor so that the correct treatment plans can be applied to these regions. Thus the more noninvasive parameters or changes in the values of the parameters one can use to identify the tumor ROIs, the more accurate and more reliable the tumor map will be, which can be used to devise the most effective treatment paradigm for the patient. In the case of proliferative tumor regions, addition of the rCBV maps obtained from PWI studies can be

509

used to further identify and delineate the major proliferating regions of the tumor. Advanced MRI techniques have been developed that allow noninvasive study of tumor vascularity. Dynamic susceptibilityweighted contrast-enhanced perfusion MRI, used widely in clinical practice, relies on the T2* signal intensity change that occurs when the contrast agent passes through the tissues. This change in signal intensity allows calculation of the rCBV, and this parameter has been correlated with tumor vascularity.13,326 Studies have shown that gliomas that have higher rCBV also have higher mitotic activity,13,298 as judged from the histopathologic determination of the number of mitotic cells found in tissue sections from stereotactic biopsies obtained from the perfusion MRI study. More recently, Price et al.276 has shown that the rCBV in HGGs linearly correlated with cellular proliferation, as judged by MIB-1 immunostaining and histopathologic analysis of resected tumor section analyzed from preoperative PWI; r = 0.66, P < 001). Thus, using the correct thresholds for the Cho/NAA ratios, ADC, and rCBV for proliferating tumor regions, and coregistering this information with the anatomic MRIs can be used to improve the delineation and more accurately assess tumor regions having high proliferative capacity.

Hypoxic/Pseudopalisading Tumor Population: MRS Metabolic Profile and the Role of PET Imaging.  Two other cell populations within the tumor that can lead to recurrence and/or resistance to therapy also need to be evaluated: (1) the quiescent pseudopalisading migratory hypoxic tumor population adjacent to necrotic regions and (2) the active infiltrating but low mitotically active tumor population migrating from the main tumor body.39,72 If neither of these populations are either resected or targeted for further treatment, they will lead to repopulation/recurrence and spread into other areas of the brain. In the case of the hypoxic/pseudopalisading population, this population is normally quiescent and if not surgically resected will be more resistant to standard radiotherapy treatments normally applied following surgery; this will lead to repopulation and spread of the tumor. In addition to their intrinsic resistance to radiation damage, this hypoxic tumor population is often refractory to the cytotoxic actions of conventional chemotherapeutic agents.316 Thus it is important to try to identify the location of this population for the design  of augmented surgical, radiotherapy, and chemotherapy treatment plans. MRSI and conventional MRI, in conjunction with positron emission tomography (PET) of [F-18] fluoromisonidazole uptake (F18FMISO) may be of use to map out the regions containing the hypoxic/ pseudopalisading tumor cells. F18-FMISO is a fluoronitroimadazole compound that was one of the first PET agents used for the pretherapy detection of hypoxic tumor populations in patients with tumors outside the CNS.283 With the advent of PET-MR clinical scanners, PET studies can be readily registered into the MRI and MRSI studies to produce anatomic maps combined with functional maps showing the distribution of proliferating and quiescent tumor populations in hypoxic regions. F18-FMISO PET can be used to image tumor hypoxia, because tumor cells show increased F18-FMISO uptake, and once taken up, it is exclusively trapped in hypoxic tumor cells because of the reduced cellular oxygen environment of these cells. F18-FMISO is cleared in normoxic tumor cells.364 The presence of hypoxia appears to be a common feature of HGGs that accumulate F18-FMISO.39,339 The hypoxic regions found in HGG have been shown to be associated with the regions adjacent to areas of necrosis where the hypoxic pseudopalisading cells reside.39,371 Recently Bruehlmeier et al.41 found that uptake and retention of F18-FMISO was consistently found in the periphery of the lesion adjacent to necrotic regions of the HGG.

http://radiologyebook.com

510

PART II  CT and MR Imaging of the Whole Body LIP/LAC

0.25

0.20

0.15

0.10 Cho 0.05

NAA

Ins dd1 Cr

0.00 ppm

A

3

2

B

1

FIG 17-8  A and B, MRSI volume showing necrosis (Lip/Lac) and tumor spectrum (elevated Cho/Cr levels).

MRSI can add support to the F18-FMISO PET findings because it has been shown that regions showing low Cho and NAA levels relative to normal brain normally indicate necrosis, astrogliosis, or regions with macrophage infiltration.95,323 However, these regions that appear to be necrotic may also contain quiescent pseudopalisading migratory tumor cells adjacent to necrotic centers,39,72 especially if they show spectra (obtained at TE = 30 milliseconds) that have elevated Cho/ NAA and Cho/Cr ratios, presence of lipid/Lac resonances, but overall lower levels of metabolites relative to normal tissue (Fig. 17-8). This spectral profile in fact may represent the most aggressive tumor population within HGGs and the major population responsible for the resistance and repopulation of tumors.39,284 In a study conducted by Martin et al.217 to determine the efficacy of MRSI in guiding biopsy, they found that in 17 out of 21 patients who had histologically confirmed biopsy tumor samples, the Cho levels were elevated relative to normal levels. However, in four patients with low Cho levels similar to those observed in necrotic regions, histologic evaluation of the biopsy sampled showed tumor. The representative MRSI spectra of one of these patients (Fig. 17-9) shows relatively elevated Cho/Cr and Cho/NAA ratios and an elevated lipid/ Lac resonance, suggestive of a necrotic region with actively proliferating tumor cells.

Infiltrative Glioma Population: Cho/Cr and Cho/NAA Ratio Profiles, rCBV, and Fractional Anisotropy.  When targeting tumor biopsies or defining tumor treatment volumes, the infiltrating tumor population must be taken into account. To determine the usefulness of MRSI in assessing the degree of tumor infiltration, Ganslandt et al.112 performed MRSI studies on 7 patients with untreated grade 2 and 3 gliomas. The MRSI data sets were fused with the 3D MRI data sets and integrated into a frameless stereotactic system for imageguided interactive neuronavigation surgery. Tissue samples were obtained from three regions based on the Cho/NAA ratios found in the MRSI study: (1) normal-appearing brain region, (2) zone between spectroscopically pathologic and suspicious for infiltrating tumor volumes, and (3) maximum spectroscopically pathologic tumor volumes. The implementation of the MRSI data set into the frameless

stereotactic system was successful in all cases, and stereotactic biopsies were obtained from the MRSI-defined regions. In this study, a relationship between the tumor cell density and Cho/NAA metabolic changes were found (60%-100% in the maximum pathologic areas to 5%-15% in the border zones suspicious for infiltrating tumor). Another important finding of this study was that the tumor areas defined by the metabolic maps, which were confirmed histopathologically, exceeded the T2-weighted abnormal signal change in all cases by 6% to 32%. However, in four patients, biopsies obtained with normal Cho/NAA ratios showed the presence of tumor infiltration. It was suggested that the false-negative finding in these patients was due to the low resolution of MRSI with respect to the glioma borders. Another explanation could be that the Cho/NAA ratio reflects primarily the proliferative capacity of the tumor cells and not necessarily the migratory capacity. Thus in infiltrative tumor zones, Cho/NAA levels may be normal or only slightly elevated because tumor cells are not dividing and/or displacing or destroying normal neural cells.368 Highly infiltrative cells normally only proliferate at vascular branch points,72 therefore in infiltrative white matter tumor regions the Cho/ NAA ratios may not be elevated. However, the Cho/Cr ratios may be elevated.368 This is shown by Wright to represent an infiltrative spectral tumor pattern.368 The infiltrating tumor pattern of gliomas is closely associated with the expression of matrix metalloproteinase 2 (MMP-2).69 Invasion of glioma cells involves the attachment of invading cells to the extracellular matrix (ECM), disruption of the ECM components, and subsequent penetration of the invading cell into adjacent brain structures. This is accomplished in part by the secretion of MMPs by the invading glioma cell to break down the ECM barrier, which impedes the infiltrating tumor cells from extending.62 Zhang et al.373 found a significant correlation between the Cho/NAA and Cho/Cr ratios and MMP-2 expression; it appears that the higher the ratios, the more invasive/ infiltrative the tumor. This is in agreement with the image-guided MRSI brain tumor biopsy studies of Croteau et al.84 Histologic analyses of tissue samples obtained from the MRSI-guided biopsy sites found that using contralateral Cr and Cho levels for normalization or ipsilateral NAA appeared to correlate with the degree of tumor

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy

Lipid

SVS

Cho Cr

511

C

NAA

A TSI

D

B 3.0

2.0

1.0

0.0

E

Chemical shift (ppm)

FIG 17-9  This patient had prominent lipid resonance and otherwise reduced metabolic levels on both SVS and MRSI. Although a modest degree of Cho elevation is evident on both spectra, it is difficult to definitively infer the presence of tumor. However, on biopsy this area was confirmed to be recurrent glioblastoma. (From Martin AJ, et al: Preliminary assessment of turbo spectroscopic imaging for targeting in brain biopsy. AJNR Am J Neuroradiol 22:959–968, 2001.)

infiltration. They found that the degree of tumor infiltration increased with increasing Cho/NAA ratio and Cho/Cr ratios (minimal level of tumor infiltration had Cho/NAA = 1.09 ± 0.12 and Cho/nCr = 1.62 ± 0.09; maximum level of infiltration had Cho/NAA = 2.01 ± 0.18 and Cho/nCr = 1.98 ± 0.18). Using the Cho/Cr and Cho/NAA threshold ratios defined in  the Croteau study for minimal level of tumor infiltration may be  useful in delineating the more metabolic and proliferative tumor populations (Fig. 17-10) from the less proliferative infiltrating tumor ones (Figs. 17-11 and 17-12). Figure 17-11 shows a tumor that appears to be infiltrating into areas beyond the Gd T1-weighted enhancing regions of tumor. The Cho/Cr peak area ratios in these areas are 1.5 or greater. The region outside the enhancing area also shows Cho/NAA peak area ratios of 1.0 or less. This suggests that in areas outside the Gd-enhancing or T2-weighted hyperintense areas, in which Cho/Cr ratios greater than 1.5 and Cho/NAA levels of less than 1.0 occur (see

Fig. 17-12), there may be regions containing a mixture of normal cells and infiltrating tumor cells with lower rates of proliferation. In the area of enhancement or extending just beyond this area, where the MRSI Cho/Cr ratios are greater than 1.5 and the Cho/NAA ratios are greater than 1.0, are areas of active tumor proliferation and infiltration (see Fig. 17-11). The spectral pattern observed for the infiltrative tumor process is consistent with the infiltrative tumor spectral pattern found by Wright et al.368 (Fig. 17-13). Cho/NAA ratios may be low (<1.0) in areas of tumor infiltration, because the normal neuronal cells are not being eliminated in these regions by actively proliferating tumor cells, as would be expected in regions in and peripheral to the Gd-enhancing area where Cho/NAA ratios are greater than 1.0. Figure 17-10 shows an HGG that appears to be well circumscribed. The Cho/Cr ratio is greater than 1.5 only in or adjacent to the enhancing region where the Cho/NAA peak area ratio is greater than 1.0. It does not appear to extend beyond this region

http://radiologyebook.com

512

PART II  CT and MR Imaging of the Whole Body

Cho/Cr

Cho/NAA

FIG 17-10  MRSI of proliferative glioma pattern. Note both Cho/Cr and Cho/NAA patterns are similar. (Threshold ratios: Cho/Cr > 1.5 and Cho/NAA > 1.0.)

FIG 17-11  Infiltrative tumor pattern. Note the proliferative tumor population (with Cho/Cr ratios > 1.5 and Cho/NAA > 1.0) is located primarily in enhancing region of tumor, whereas the predominantly migratory population (with elevated Cho/Cr ratio > 1.5 and Cho/NAA < 1.0) is outside the enhancing region.

as defined by the Cho/Cr peak area ratio of 1.5 or greater for tumor threshold (namely, where the Cho/Cr is > 1.5, it is highly probable tumor is present in this voxel). This observed Cho/Cr and Cho/NAA ratio pattern may suggest that the tumor is expanding via a more proliferative rather than an infiltrative pathway; this may be due to the more cortical location of this lesion, which predominantly affects white matter.72 Di Constanza et al.93 used an MR multiparametric approach to assess the extent and malignancy of cerebral gliomas. Thirty-one patients (21 high grade and 10 low grade) with gliomas underwent conventional MRI, H1-MRSI, DWI, and PWI studies prior to surgery and histologic examination of resected tissue. ROIs with high Cho (>1.3; Chovolume on tumor side/Chocontralateral normal volume) and or abnormal

Cho/NAA greater than 1.0 (normalized to ratio found in contralateral volume) were considered tumor or infiltrated tumor on the basis of previous neuropathologic correlation studies. From this multiparametric study of high-grade lesions, three tumor regions could be delineated: (1) Gd-enhancing tumor region (margin and mass), (2) perienhancing abnormal region containing tumor and edema, and (3) normal-appearing perienhancing region containing infiltrated tumor and normal brain parenchyma. The high-grade lesions containing the bulk tumor mass and enhancing margins (1) showed very high mean Cho/NAA ratios greater than 4.0 and Cho/Cr greater than 2.0 (normalized relative to contralateral normal region), high mean ADC values (1.24 ± 0.31 × 10−3 mm2/sec in bulk tumor vs. 0.77 ± 0.05 × 10−3 mm2/sec in normal brain region), and mean elevated

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy NORMAL UNINVOLVED REGION SLICE 2

SUSPECTED INFILTRATIVE TUMOR SPECTRUM

I : Integral

I : Integral

NAA I : 0.393

0.03

513

NAA I : 0.175 Cho I : 0.166

0.010

Cho I : 0.245 Cr I : 0.230

0.02

0.005 Cr2 0.01

Cr I : 0.0512

I : 0.0232

Cr2 I : 0.0868

0.000 0.00 ppm 4

3

2

ppm

1

4

3

2

1

FIG 17-12  Spectral pattern of (left) normal and (right) infiltrative tumor process (non–Gd-enhancing region).

rCBV of 3.70 in tumor vs. 0.98 in normal tissue. In the perienhancing abnormal-appearing region containing tumor and edema (2), the normalized mean Cho/NAA was found to be 3.3 and Cho/Cr was 1.76, the mean tumor ADC value was 0.92 ± 0.17 × 10−3 mm2/sec, and mean tumor rCBV was 1.68 ± 0.41. In the normal-appearing perienhancing region (3), the normalized mean infiltrated tumor Cho/NAA ratio was 1.55 and Cho/Cr ratio was 1.28, the mean ADC of this region was 0.86 ± 0.08 × 10−3 mm2/sec, and the mean rCBV of this region was 1.28 ± 0.16. The primary change in the ratios of Cho/NAA in the normalappearing infiltrated tumor region of a high-grade lesion is due to a change in the Cho levels and not to changes in the level of NAA. The mean normalized NAA levels in infiltrated tumor regions were 0.94 ± 0.23 vs. 0.96 ± 0.14 for the levels found in the normal contralateral side. In the bulk tumor mass, the mean NAA level was 0.31 ± 0.11, and in the perienhancing abnormal region (2) the mean NAA level was 0.58 ± 0.22. Thus in both regions 1 and 2 (as per above), the increased changes observed in the metabolite ratios appear to be due to decreased levels of NAA and increased Cho. This suggests higher tumor proliferative capacities in these regions as measured by increased Cho levels and loss of NAA levels. NAA is a neuronal marker, so the loss of NAA suggests either loss of neuronal activity or destruction due to tumor proliferation and invasion into these regions. However, in region 3 the almost normal level of NAA suggests that changes observed are due primarily to infiltrating tumor cells that are not proliferating rapidly. This interpretation is consistent with the changes in rCBV observed in tumor regions (2-3) of this study. Price et al.276 showed that high-grade brain lesions that have increased mitotic activity also have higher rCBV values. These investigators showed that rCBV correlated significantly with MIB-1 labeling index and not with tumor cell packing, which suggested that increased tumor vascularity as measured by rCBV was more dependent on cellular proliferation than just the number of   cells present. Low-grade lesions demonstrated the same trend with respect to Cho/NAA, Cho/Cr, ADC, and rCBV values in bulk tumor versus

infiltrated tumor regions. The bulk of the tumor showed much higher values for all parameters compared to the normal-appearing adjacent infiltrated tumor regions, but in both the bulk tumor and the infiltrated tumor regions, values were found to be higher than in the contralateral normal brain regions. Tumor infiltration into white matter can induce widening of white matter fiber bundles, which affects the directionality of the motion of water molecules within the white matter fiber bundles. Changes in the directionality of water molecules within white matter tracts can be measured with DTI. By employing at least six diffusion gradient directions, compared to three gradient directions, DWI studies enable not only the macroscopic detection and imaging of the highly anisotropic water diffusion found in white matter bundles but can also detect changes in the anisotropic motion of water (i.e., directionality) within white matter due to tumor invasion, which cause disruption of white matter bundles.275 A DTI study generates a diffusion tensor D for each image voxel, and from each D voxel value, the magnitude and directionality of water diffusion can be calculated. The magnitude is known as the mean diffusivity (MD) and is mathematically equivalent to the ADC obtained from a  conventional DWI study. The directionality is known as fractional anisotropy (FA). In a 2006 publication, Stadlbauer et al.323 used DTI to retrospectively correlate changes in FA and MD with the degree of tumor cell infiltration determined histologically. Histopathologic findings of 77 MRI-guided stereotactic biopsies in 20 glioma patients were correlated with the FA and MD values found at the biopsy sites. FA was found to be inversely correlated to percent tumor infiltration and was found  to be better than MD for assessment and delineation of tumor  infiltration. For low tumor cell numbers as shown in Figure 17-14,  the logarithmic regression lines showed a strong decrease in FA (see Fig. 17-14A) and an increase in MD (see Fig. 17-14B) followed by an asymptotic plateauing trend for high tumor cell number. From these findings, the investigators hypothesized that their findings “reflect  the infiltrative invasion mechanism of gliomas”—namely, initial

http://radiologyebook.com

514

PART II  CT and MR Imaging of the Whole Body

IC1

Lip and lac

g1 Cho

40

35

30

25

A

20

15

10

05

40

35

30

30

25

B

20

15

10

05

35

30

05

10

05

10

05

NAA

40

35

30

25

20

15

ppm

NAA g3 Cr

Cho

40

10

Cr

E

ppm

IC3

C

15

g2 Cho

35

20 ppm

IC2

40

25

D

ppm

25

20 ppm

15

10

40

05

35

F

30

25

20

15

ppm

FIG 17-13  A-C, The three independent components (ICs) generated from the 1517 voxels, normalized as described by Wright and coworkers.368 D, The median of all absorption spectra, with IC1 as the component with the highest coefficient (g1, necrotic tumor core), illustrated with the 25% and 75% quartiles above and below (dashed lines). E and F, The median and quartiles of spectra for the highest-coefficient components of IC2 and IC3, respectively (g2 [infiltrative tumor growth] and g3 [normal brain]). D-F, Key metabolite resonances, indicated as Lip and lac (lipids and lactate), Cho (cholines), Cr (creatine), and NAA (N-acetylaspartate); g2 (E), the median and quartiles of spectra for the highest-coefficient IC2 (mixture of normal and infiltrative tumor pattern); g3 (F), the median and quartiles of spectra for the highest coefficient IC3 (normal brain pattern). (From Wright AJ, et al: Pattern recognition of MRSI data shows region of glioma growth that agrees with DTI markers of brain tumor infiltration. Magn Reson Med 62:1646–1651, 2009.)

http://radiologyebook.com

515

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy 2.25

0.35

2.00 MD value [x 10–3 mm2/s]

FA value

0.30

0.25

0.20

0.15

1.75 1.50 1.25 1.00 0.75

R = –0.796

0.10 0

A

20

R = 0.521 40

60

80

100

Tumor infiltration [%]

0

20

B

40

60

80

100

Tumor infiltration [%]

FIG 17-14  Scatterplot of fractional anisotropy (FA) and mean diffusivity (MD) vs. percentage TI. A, Linear regression (solid line) of FA versus percentage TI for all biopsies sampled. Regression analysis resulted in R = 0.796 and P < 0.001. Linear regression fit is represented by FA = 0.28-1.5 × 10−3 percentage TI. B, Linear regression (solid line) of mean diffusivity versus percentage TI for all biopsies sampled. Regression analysis resulted in R = 0.521 and P < 0.001. Linear regression fit is represented by mean diffusivity = 1.17-5.9 × 10−3 percentage TI. Dotted lines = 95% CIs. (From Stadlbauer A, et al: Preoperative grading of gliomas by using metabolite quantification with high-spatial-resolution proton MR spectroscopic imaging. Radiology 238:958– 969, 2006.)

TABLE 17-2  Median Values of FA and MD in the Tissue-Segmented Groups g1 (Tumor Core), g2 (Infiltrative Tumor Growth), and g3 (Normal Brain) FA

MD

g1

g2

g3

g1

g2

g3

Median Significantly different from g1 (Mann-Whitney U test)

0.128 — —

0.240 Yes P = 3 × 10−49 No P = 0.11

1.23 (×10−3) —

Significantly different from g2 (Mann-Whitney U test)

0.233 Yes P = 7 × 1048−48 —

1.13 (×10−3) Yes P = 1 × 10−4 —

0.909 (×10−3) Yes P = 2 × 10−30 Yes P = 8 × 10−27

—

From Wright AJ, et al: Pattern recognition of MRSI data shows region of glioma growth that agrees with DTI markers of brain tumor infiltration. Magn Reson Med 62:1646–1651, 2009.

enlargement of the extracellular space as the tumor invades, leading to a logarithmic decrease in FA and a logarithmic increase in MD but preserving the normal microstructure of the white matter fiber bundles in the border zone. This hypothesis is consistent with the data of  Di Constanza et al.93 in which the NAA levels remained relatively unchanged while the Cho levels increased by almost 50% in suspected tumor-infiltrated normal-appearing ipsilateral white matter. This lack of change in NAA reflects only the initial phase of tumor infiltration into these regions, which induces only mild disruption of white matter bundles. As more tumor cells infiltrate and proliferate, there is a decrease in the extracellular space (e.g., edema) and a decrease in directionality of diffusion due to tumor-induced derangement and destruction of white matter microstructures, which leads to decreases in NAA along with increased Cho levels. A number of investigators have proposed the combined use of proton MRSI and DTI to map out the infiltrative regions of gliomas.33,117 Wright et al.368 performed independent component analysis of MRSI

data from human gliomas to segment tissues into tumor core, tumor infiltration, and normal brain. DTI analysis was then used to confirm the correctness of the MRSI segmentation pattern (see Fig. 17-13). They found that the MD and FA data found for these regions were consistent with previous studies that compared anomalies in isotropic and anisotropic diffusion images to determine regions of potential glioma infiltration (Table 17-2). They showed that coefficients of independent components from their analysis could then be used to create colored images overlaid onto conventional anatomic images to visualize regions of infiltrative tumor growth. MRSI and DTI both provide markers that represent infiltrative growth, and the two techniques are complementary. DTI can detect abnormalities with high spatial resolution, and MRSI can provide validation of the presence of proliferating and/or infiltrating tumor cells from the metabolic profile. The combination of markers will provide greater certainty in correctly identifying regions of tumor infiltration, which can then be treated appropriately.

http://radiologyebook.com

516

PART II  CT and MR Imaging of the Whole Body

Monitoring Tumor Response to Therapy MRS may be able not only to facilitate diagnosis and accurate classification of de novo brain tumors but also to allow differentiation of recurrent tumor and tumor progression from radiation necrosis, posttreatment effects, and edema.36,52,53,296 Thus MRS may be useful in evaluating the response to therapy in patients with brain tumors. Radiation necrosis occurs from approximately 6 to 24 months after completion of conventional therapy, is seen more commonly with high-grade astrocytomas, and may be indistinguishable from recurrent tumor by conventional Gd-enhanced MRI.53 Elevated Lac levels are seen in proton MR spectra of patients who have received 40 Gy or more of radiation to the brain, even when the conventional MRI study does not yet demonstrate any structural abnormality within the resection bed.53 One study involving 25 patients with cerebral astrocytomas who received a combination of radiation and chemotherapy demonstrated increased Cho/NAA and Cho/Cr ratios as well as the presence of Lac in recurrent tumors, compared to markedly decreased levels of NAA, Cho, and Cr and the presence of a broad intense signal between 0 and 2.0 ppm in cases of radiation necrosis.53,102 This broad signal, between 0 and 2.0 ppm, consists of free fatty acids, Lac, and amino acids.53,102 However, because most therapy-induced tissue damage occurs in combination with areas of viable tumor, single-voxel techniques are not optimal for evaluation of these patients, and 3D MRSI is preferable for distinguishing among areas of residual or recurrent tumor, radiation necrosis, and viable normal brain tissue.52,372 Sensitivity and specificity of proton MRS for the detection of residual/recurrent tumor in radiated patients in one series were 71% and 100%, respectively, and serial MRS in the same series allowed differentiation of necrosis  and tumor progression; progressive decreases in Cho levels and mild increases in NAA levels correlated well with therapy success.52,333 Elmogy et al.99 in a 23-patient study with 2D MRSI found that Cho/ NAA ratios were the best discriminators of recurrent/residual tumor from radiation injury. In this study using a cutoff of 1.8 for the Cho/ NAA ratios, an accuracy of 96%, a sensitivity of 93%, and specificity of 100% were found. In a 3D MRSI of 28 patients, Zeng et al.,372 using a cutoff of 1.71 for either Cho/Cr or Cho/NAA, resulted in a ROC analysis of 96.2% for accuracy, 94.1% for sensitivity, and 100% for specificity of MRSI in discriminating recurrent/residual tumor from radiation injury. Tedeschi and coauthors demonstrated that interval changes in Cho levels during long-term (3.5 years) follow-up with MRS allowed differentiation of stable from progressive glioma.36,336 A study by Henry et al. suggested that MRS and MR perfusion imaging (namely, rCBV mapping) may be complementary modalities that when combined may be better able to noninvasively differentiate tumor from necrosis, posttreatment effects, or edema in patients with treated gliomas than either modality alone.134 Graves et al.126 examined 18 patients with recurrent malignant gliomas with proton MRSI to characterize the temporal tissue responses in these tumors after receiving high-dose gamma knife radiosurgery treatment. Only six patients with recurring high-grade gliomas responded to the treatment and showed reduction of Cho and an increase in Lac/lipid resonances, typically within 6 months post treatment. Of the 18 patients treated with radiosurgery, 12 recurred after  treatment (recurrence was defined as > 2 cc increase in Gd contrastenhancement areas). In each case, an elevated Cho/NAA ratio was associated with at least one voxel within the area of recurrence. In 9 of the 12 recurring patients, voxels coinciding with the recurring regions showed increased Cho/NAA ratios relative to normal-appearing voxels

prior to increases in contrast enhancement. In the remaining 3 patients, the new areas of contrast enhancement appeared in regions and were characterized by pronounced increases in Cho/NAA ratios. Other groups have reported similar success in distinguishing radiation necrosis from recurrent tumor with proton MRS, although a few have reported contradictory results; the reasons for differences in results may be due to the fact that Cho levels may be elevated in early radiation-induced lesions because of demyelination and reactive astrocytosis.62,68,79,161 In addition, Bizzi and coworkers reported that MRS is useful in the surveillance of lymphoma following treatment.35,52

Role of Proton MRSI in Predicting Tumor Prognosis Long-term prognosis for patients with HGG remains dismal even with advances in both imaging and therapy. In a 2003 survival analysis of 766 GBM patients, only 2% of these patients were found to be still alive after 5 years.224 Given such a bleak prognosis, there is a need to identify noninvasive biomarkers that can be used to characterize individual lesions and predict outcome. This information can then be used to stratify patients into high- or low-risk categories to determine which patients may benefit from new and/or experimental therapies. Several studies have evaluated the role of proton MRSI in the prediction of survival of GBM patients.61,80,303 In a multiparametric MR study of 56 GBM patients, proton MRSI, conventional MRI, DWI,  and PWI were combined to examine whether pretreatment characteristics of the tumors obtained from these studies could be used to predict survival of these patients after treatment.80 This study showed that shorter survival was associated with the volume of tissue  with abnormally high Cho/NAA ratio [CNI > 2; (Cho/NAA)tumor vol/ (Cho/NAA)normal contralateral vol], contrast-enhancing tissue and high rCBV, and tissue with abnormally low ADC values. Higher Lac and level of lipids were also associated with a worse prognostic outcome. Several studies have also examined the prognostic value of proton MRS in evaluating pediatric brain tumors. In a study of 76 pediatric brain tumors, a low Cho/Cr ratio—less than 1.8 index value (normalized to contralateral Cr peak area)—was found to be a strong predictor of survival.216 In a different report, a high Cho/NAA ratio in children with recurrent gliomas was associated with decreased survival.361 In a multiparametric longitudinal MRI study of 34 patients conducted at the National Cancer Institute (Bethesda, Md.), Hipp et al. used proton MRS, conventional MRI, and perfusion MRI data to predict outcome in children with diffuse intrinsic pontine gliomas.139 They found that at baseline MRI only, increased ratio of Cho/NAA and increased perfusion on dynamic susceptibility contrast (DSC)-MRI predicted shorter survival. When examined at later time points, any volumes with an observed maximum Cho/NAA ratio on MRSI (i.e., >4.5) or increased Cho/NAA on either single- or multivolume MRSI compared to baseline increased perfusion, and the presence of enhancement predicted shorter survival. They concluded that changes in Cho/NAA over time, as determined by either single- or multivolume MRS were “prognostic and can potentially be used as indicators of response or lack of response to treatment.” They also recommended that routine baseline and subsequent imaging for children with diffuse intrinsic pontine gliomas should, at a minimum, incorporate both DSC-MRI and single-voxel spectroscopy.

Differentiation of Neoplastic from Pseudotumoral Process The differential diagnosis of a brain lesion may vary depending on its solid and necrotic characteristics. For example, if the brain mass encountered on a conventional MR examination appears more necrotic in nature, the differential diagnosis would favor that of either an

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy aggressive malignant brain tumor (e.g., metastatic brain lesion or GB) or nonneoplastic pseudotumoral brain lesion that mimics a brain tumor (e.g., abscess, tuberculous granuloma, parasitic infection, or radiation necrosis if the patient had received radiation treatment for  a brain tumor). Conversely if a brain lesion appears solid on the  MRI the main diagnosis would be either an HGG or a pseudotumoral demyelinating disease process such as multiple sclerosis. MRI examination suggesting a pseudotumoral nonneoplastic process would lead to further clinical tests to determine the type of process so that the appropriate therapy could be applied. However, if a neoplastic process is suggested from the MRI exam, then a stereotactic biopsy or surgical resection would be considered to determine whether the lesion is  an HGG or metastatic lesion and used to determine whether further chemotherapeutic and/or radiation therapy treatments are needed. However, in many cases reliable differentiation of neoplastic from nonneoplastic brain lesions cannot be unequivocally made from conventional MRI.184,256 Proton MRS can help differentiate between brain tumors and nonneoplastic brain lesions.46,237,269,282 In two retrospective studies, discriminant function analysis using proton MRSI data correctly classified 78% of pediatric cases with primary brain lesions based on Cho/Cr ratio (the study included 32 subjects).147 In 69 adult patients with untreated primary brain lesions, proton MRS imaging correctly classified 84% of originally grouped cases based on NAA/Cho, NAA/ Cr, and relative NAA and Cho levels (compared to the contralateral

hemisphere). Combining proton MRSI and perfusion MRI, sensitivity of 72% and specificity of 92% were achieved using cutoff points corresponding to the diagnosis of tumor of 0.61 for the ratio of NAA/Cr and 1.5 for rCBV.146 Recently the role of routinely used advanced MRI in differentiation of intracranial masses in adults was evaluated5 and the accuracy of an MRI-based strategy to differentiate among histologically confirmed neoplastic and nonneoplastic lesions was assessed. This review and assessment led to the development and evaluation of a MR-based decision paradigm that could be used to differentiate neoplastic from nonneoplastic processes.6 This “practical MRI-based” paradigm proposed in Figure 17-15 used results from postcontrast Gd-enhanced, DWI, PWI, and multivoxel H1-MRSI with well-defined thresholds to determine how well this multiparametric approach could not only accurately differentiate neoplastic from nonneoplastic lesions but also classify the type of neoplastic lesion (e.g., glioma from lymphoma and metastatic lesions) and nonneoplastic processes (e.g., abscess from tumefactive process).55 The MR-based diagnostic paradigm was evaluated in 40 patients who had complete data from all MR techniques to differentiate between tumors and nonneoplastic pseudotumoral processes. The accuracy, sensitivity, and specificity of the paradigm using the threshold values used in the ROC curves for each technique was found to be 90%, 97%, and 67%, respectively. For discrimination  of high-grade from low-grade gliomas, the accuracy, sensitivity, and

Intraaxial brain mass Conventional CE MRI Does the lesion enhance? Yes

No

Diffusion MRI Is diffusion facilitated over 1.1/100mm2/ADC?

Magnetic Resonance Spectroscopy Is there elevation of Cho/NAA over 2.2? Yes

No

Yes

No

Low-grade neoplasm or Encephalitis

Is there necrosis on CE MRI? Perfusion MRI Is perfusion increased over 1.75 rCBV? No

Low-grade neoplasm

No

Yes No

Yes

Lymphoma

Yes

High-grade neoplasm

TDL or abscess Abscess

Magnetic Resonance Spectroscopy Is there perienhancement infiltration over 1 Cho/NAA? No Metastasis

517

Yes High-grade glioma

FIG 17-15  Flow chart for determining brain lesion type based on conventional contrast-enhanced MRI, DWI, MRS, and PWI. 1.1/100mm2/ADC, 1.1 × 10−3  mm2/sec; ADC, apparent diffusion coefficient; CE, contrastenhanced; rCBV, relative cerebral blood volume; TDL, tumefactive demyelinating lesion. (Modified from AlOkaili RN, et al: Intraaxial brain masses: MR imaging–based diagnostic strategy–Initial experience. Radiology 243:539–550, 2007.)

http://radiologyebook.com

518

PART II  CT and MR Imaging of the Whole Body

specificity was 90%, 88%, and 100%, respectively, and 85%, 84%, and 87% for discrimination of high-grade tumors and lymphomas from low-grade tumors and nonneoplastic pseudotumoral processes. The results of this paper clearly indicate that integration of advanced complementary MRI techniques with conventional anatomic MRIs can improve the diagnosis and classification of brain lesions, which in turn will add value to the clinical decisions being made for patients with these types of lesions (see Fig. 17-15). Typically, tumors exhibit elevated Cho and decreased levels of NAA. The greatest benefit of adding MRS to a clinical examination may be to exclude processes in the diagnosis that have markedly different spectral characteristics but similar MRI characteristics to a low-grade tumor such as focal cortical dysplasias or strokes. Neither of these latter two lesions show elevated Cho levels but show decreased levels of NAA.196,360 Similarly, spectra from brain abscesses are quite different from HGGs, but the anatomic MRI appearance is similar to necrotic malignant tumors, especially GBs.189 However, the MR spectral profiles of abscesses compared to high-grade tumors normally show decreased Cho, NAA, and Cr resonances and increased resonances from compounds not normally observed in brain tumors, such as alanine, acetate, acetoacetate, and succinate. Another application of proton MRS is the differentiation of gliomas from hamartomas, particularly in the setting of phakomatoses such as neurofibromatosis type 1.296 NAA/Cho, NAA/Cr, and Cr/Cho ratios in hamartomas are closer to those of normal brain tissue than to those of gliomas; specifically, studies have demonstrated Cho/Cr ratios greater than 2.0 in gliomas, between 1.3 and 2.0 in hamartomas, and less than 1.3 in normal brain tissue.48,119,296

Metastases Brain metastases are the most common intracranial tumors in adults in 20% to 40% of cancer patients,12,270 although they usually do not pose much of a diagnostic challenge when multiple brain lesions are observed on conventional MRI. However, metastases can be problematic when they are solitary, because it may be difficult to distinguish them from primary brain neoplasms.42 Unfortunately, proton MR spectra of intracranial metastases are often nonspecific and indistinguishable from those of primary brain tumors; at long TE, they display low NAA levels, low Cr levels, and elevated Cho levels as shown in Figures 17-16 and 17-17.51-53,150,183,252 Although in theory NAA should not be present in metastases because of their lack of neural components, NAA is frequently present in proton MR spectra, presumably secondary to voxel contamination with adjacent brain parenchyma or due to the presence of N-acetylated metabolites on their cell membranes.52 One study indicated that although no statistically significant difference existed between signal areas of Cho, Cr, and NAA in the spectra of metastases and those of high-grade astrocytomas, a trend toward higher Cho/Cr and Cho/NAA ratios in astrocytomas did exist, and lipid and Lac signals were more common in intracranial metastatic lesions.52 Lipid signals are often present in metastases, particularly those from breast carcinoma.53,318 Detection of an enhancing brain lesion on conventional crosssectional imaging invokes a much broader list of malignant and benign differential considerations. In fact, as many as 11% of solitary brain lesions identified on MRI may represent processes other than metastases.260 In these cases, adjunctive tests such as proton MRS and PWI can be helpful both in distinguishing malignant from benign lesions and in differentiating primary brain tumors from metastases.150,191 Malignant brain tumors tend to demonstrate both high Cho and increased vascularity compared to benign processes. Huang et al.150 found a very strong positive association between Cho/Cr ratios and rCBV in brain metastases. Law et al.191 showed that both Cho/Cr levels

and rCBV remain relatively higher in the peritumoral region (1 cm from enhancing margin) in GBMs compared to the uninvolved contralateral region, whereas in metastatic peritumoral regions, both Cho/ Cr levels and rCBV are near normal compared to the contralateral uninvolved regions. Metastatic lesions compared to primary malignant brain tumors show a more circumscribed lesion with almost normal peritumoral Cho/Cr ratios compared to high-grade primary brain tumors (Fig. 17-18; compare with Fig. 17-11).

Other Tumors MRS findings in gliomatosis cerebri include elevated Cho/Cr and Cho/ NAA ratios as well as varying degrees of decreased NAA/Cr ratios.30 A maximum Cho/NAA ratio of 1 : 3 was detected in low-grade lesions, compared with 2.5 for anaplastic tumors; in addition, Lac signals were noted in grade III and IV lesions.30 MR spectra of acoustic schwannomas show absence of Cr, marked reduction in NAA, and increased lipids.43,52 Other MRS studies have attempted to distinguish astrocytomas from ependymomas and primitive neuroectodermal tumors (PNETs); these studies have demonstrated reduced NAA/Cho and elevated Lac/Cho ratios in astrocytomas and ependymomas in comparison with PNETs.296,327,328,345,357 Meningiomas have also been extensively studied with proton MRS, and they display certain characteristic features on MR spectra. Marked elevation of Cho levels up to three times that of normal brain parenchyma have been reported, particularly in recurrent menigiomas.52,53,183 The Cho/Cr ratio has been reported to be higher in malignant meningiomas than in benign menigiomas.52,252

Traumatic Brain Injury Recent studies of concussion have revealed how complex this often misdiagnosed condition can be. A single episode of traumatic brain injury (TBI), induced by the mechanical trauma of acceleration and deceleration forces on the brain, can set off a cascade of complex neurochemical and neurometabolic brain processes23 that, depending on the severity of the TBI, can be either reversible or irreversible. Rotational forces from TBI can lead to axonal stress and strain, causing diffuse axonal injury (DAI).219 After the initial brain insult, a destructive series of biochemical processes occur, such as activation of inflammatory responses, imbalances in Na+, K+, and Ca+2 ion levels, increases in the presence of excitatory amino acids, dysregulation of neurotransmitter synthesis, and increased production of free radicals.363 As a result of these responses to the brain insult, individuals present with myriad of clinical symptoms.47 What is more disturbing is that in 15% of individuals diagnosed with mild TBI (mTBI) report physical, cognitive, and emotional symptoms that persists for more than 1 year post injury.175,366 These persistent symptoms can lead to postconcussive syndrome and long-term disabilities. Despite the cascade of destructive pathophysiologic and biochemical processes that occur after MTI and especially mild brain injuries, conventional neuroimaging techniques (i.e., CT and MRI) and neuropsychological tests (i.e., Glasgow Coma Score and Glasgow Outcome Score) fail to be sensitive or specific enough to detect changes in the early subacute phase of injury and also in individuals who have had previous mTBI.8 Recently, advanced MRI techniques (i.e., SWI, DTI) and multivolume H1-MRS imaging have emerged as promising techniques that can more accurately assess the severity and regional distribution of injury. Complimentary data obtained from these techniques can then be used to obtain reliable metrics to predict the severity and outcome after a TBI event. Accurate early prediction of a patient’s long-term prognosis enables the physician to determine appropriate treatment paradigms.9

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy NAA I : 0.136 0.020

Cho I : 0.134 Met. NSCL Ca: Cho/Cr > 2.0

0.015

0.010 Cr Cr2 I : 0.0170 I : 0.103 0.005

0.000

ppm

A

3

2

1

Cho I : 3.77 0.5

CSI VOXEL MET BREAST CA Cho/cr > 2.0

0.4

0.3

0.2

Cr I : 0.845 NAA I : 0.608

0.1

0.0 ppm

B

3

2

1

FIG 17-16  A, Metastatic non–small cell lung cancer spectrum. B, Metastatic breast cancer. Continued

http://radiologyebook.com

519

520

PART II  CT and MR Imaging of the Whole Body

0.20

MET MELANOMA

0.15

0.10

NAA I : 1.05

Cr I : 0.763 0.05

0.00 ppm

C

3.5

3.0

2.5

2.0

1.5

1.0

FIG 17-16, cont’d  C, Metastatic melanoma.

Ratios of Cho/Cr, NAA/Cr, and Cho/NAA for different cancer types Breast

5

Lung

Melanoma

Mixed

Control

4

3

2 2.814

2.821

3.054

3.063

2.815

1

2.851

1.959 1.44

1.109

1.09

1.408

1.191 1.241

1.514 0.576

0 NAA/Cr

Cho/Cr

NAA/Cr

FIG 17-17  Metabolic ratios for different metastatic tumors.

In the case of MRS, using either single- or multivolume MRSI technique, studies have found consistently large increases in Cho/Cr ratios and diffuse decreases in NAA/Cr and NAA/Cho ratios in the brains of TBI patients.9,120,218 In a pediatric TBI patient study, Aswal et al.9 compared the differences between the acutely determined MRS variables (metabolite ratios and presence of Lac) and 6- to 12-month

outcomes. They found that the metabolite levels of NAA/Cr and NAA/ Cho were significantly lower in patients with poor outcome. Lac was evident in 91% of infants and 80% of children with poor outcomes. At best, using the standard clinical criteria alone predicted the outcome in 77% of the infant cohort and 86% of the children cohort. The presence of observable Lac alone predicted an outcome in 96% of both the

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy

521

Cho/Cr

A

B FIG 17-18  Metastatic melanoma. A, Spectral map of metastatic melanoma. Note that tumor spectral pattern does not extend more than 1 voxel from Gd-enhancing rim. B, Metabolic map of metastatic melanoma. Note that Cho/Cr ratios greater than 1.5 do not extend more than 1 voxel beyond Gd-enhancing rim.

infant and children cohorts. In an MRS study conducted in adults, Marino et al.218 found similar results. Both the decrease in NAA/ metabolite ratio and the presence of Lac correlated closely with the Glasgow Coma Scale and the Glasgow Outcome Scale at 3 months. Again as in the pediatric study, the presence of detectable MRS levels of Lac is associated with poorer clinical outcome. In an H1-MRS study of clinically asymptomatic athletes recovering from single and multiple mTBIs, Johnson and colleagues163 evaluated the NAA/Cho, NAA/Cr, and Cho/Cr ratios in two regions within the corpus callosum, the genu and splenium. These two regions were chosen because the corpus callosum is a major site of TBI.321 The major finding in this study was that a single concussive episode, whether it was an initial or subsequent injury, produced significant reductions in NAA/Cho and NAA/Cr ratios in the genu of the corpus callosum but not in the splenium for mTBI subjects. In addition, there does not appear to be any further reduction in NAA ratios in the genu or splenium as a result of an increase in the number of mTBI episodes. However, they observed that as the number of mTBIs increased, so did the length of time for symptom resolution. The changes observed in a DTI study of the corpus callosum of mTBI patients supports the MRS findings that mTBI induces changes primarily in the genu of the corpus callosum, with little alterations in the splenium.297 Patients with mTBI examined less than 3 months post trauma, compared to controls, had reduced FA and increased MD. However, patients with moderate to severe TBI examined at less than 3 months post trauma had reduced FA and increased MD in the genu and reduced FA in the splenium, without significant change in the MD. This suggests a larger contribution of vasogenic edema in the genu than in the splenium in TBI.

Epilepsy Proton MRS has a major clinical application in the localization of subtle epileptogenic foci that are not evident on conventional structural MRI sequences at 1.5 T or 3 T, as well as in the planning 

of epilepsy surgery. Single-voxel MRS may confirm a structurally abnormal epileptogenic focus, whereas MRSI may detect foci that appear normal on conventional structural MRI but demonstrate  definite metabolic abnormalities, such as those associated with very subtle malformations of cortical development.77,248 One study suggests that the sensitivity of MRS for detecting subtle neuronal dysfunction is greater than that of PET.2,296 Development of different surgical approaches to epilepsy, as well as the potential for future less invasive treatment modalities such as gamma knife radiosurgery, has made precise localization of seizure foci critical.14,248 Most of the applications of MRS in epilepsy have involved temporal lobe epilepsy (TLE), the most common type of partial epilepsy. Many cases of TLE are refractory to medical therapy and thus MRS can play an important role in presurgical planning in these patients.36 The most common lesion associated with TLE is mesial temporal sclerosis (also called hippocampal sclerosis), noted in approximately 65% of cases of TLE.17,36,77 Although in classic cases of mesial temporal sclerosis the involved hippocampus is easily identified by a  combination of volume loss and signal hyperintensity on T2-weighted and fluid-attenuated inversion recovery (FLAIR) sequences, 20% of patients with TLE have normal structural MRI scans, and the findings in children generally tend to be more subtle than those in adults.248,296 NAA, NAA/Cho, NAA/Cr, and NAA/(Cho + Cr) all are decreased in atrophic hippocampi as well as in nonatrophic hippocampi with abnormal electroencephalographic (EEG) findings, according to the results from the series by Ende and coworkers.36,101,187 Even when seizure onset and structural MRI abnormalities are clearly unilateral, MRS has shown that bilateral temporal lobe abnormalities are present; bilateral metabolic abnormalities are found in approximately 40% to 50% of TLE patients, and in such cases these abnormalities appear to be more diffuse than the corresponding structural MRI abnormalities would suggest.76,77,83,101,137,250 Kuzniecky and colleagues argued that the lack of correlation between structural hippocampal volume loss and proton MRSI 

http://radiologyebook.com

522

PART II  CT and MR Imaging of the Whole Body

metabolic abnormalities reflects the presence of distinct pathophysiologic processes that are coexistent in cases of mesial temporal sclerosis.186 Gadian and colleagues, examining a mixed group of patients with partial seizures, mostly with TLE, reported a 9% decrease in the NAA content of the epileptogenic temporal lobe, with an increase of 14% and 17% in the levels of Cr and Cho, respectively, suggestive of gliosis.111,174,245 Multivoxel MRS studies have shown not only reduced NAA levels in the diseased hippocampi but also the presence of Lac signals in the spectra of epileptogenic foci when studied within 6 hours of seizure activity.58,73,250,296 The postictal Lac signal has been described as having potential lateralizing value, because only one side demonstrates a signal in Lac, even in patients with bilateral disease.73 Cendes and associates noted that correct lateralization of the seizure focus in patients with TLE was possible in 83% of cases in their large series with MR volumetry and 86% with MRSI.59,174 Ende and coauthors found that the NAA/(Cho + Cr) ratio was the most sensitive measure for lateralization compared with ictal EEG.101,174 They also noted that in unilateral TLE, the reduced levels of NAA in the hippocampus contralateral to the atrophic one predict poor clinical outcome following surgical resection of the epileptic focus.36,101 Thus MRS may provide important prognostic as well as diagnostic information. MRS has also shown promise in evaluating patients with extratemporal epilepsy, including those with neocortical epilepsy. This class includes the second largest group of patients with complex partial seizures refractory to medical therapy—those with frontal lobe epilepsy.1 Garcia and coworkers113 reported that in a series of eight patients with frontal lobe epilepsy, the mean NAA/Cr ratio in the epileptogenic frontal lobe was decreased by 27% compared with that of the contralateral frontal lobe, with decreases of at least 5% in each individual.1 In addition, in patients with structural MR evidence of malformations of cortical development or neuronal migration disorders (NMD), MRS provides insight into both the pathophysiology and the true spatial extent of the disease processes. Li and coworkers showed that the maximal NAA/Cr ratio decrease, indicative of metabolic dysfunction, localized to the structural malformation noted on conventional MRI in 23 cases of malformations of cortical development (including focal cortical dysplasia, heterotopia, polymicrogyria, and tuberous sclerosis)198; however, less impressive decreases also extended to normal-appearing areas of brain tissue adjacent to the structural lesion. Simone and colleagues, in their series of 15 patients with NMD, noted abnormally decreased NAA/Cr and Cho/Cr ratios in these lesions when compared with gray and white matter of neurologically normal controls319; they also noted abnormally decreased Cho/Cr ratios in the normal-appearing brain contralateral to the identified NMD compared with gray and white matter of controls. The absence of correlation between the NAA/Cr decrease, EEG abnormalities, and NMD lateralization suggested that the metabolic abnormality may be related more to the underlying structural and functional alterations within the focus of NMD than to actual epileptic activity in the lesions, and that the Cho/Cr ratio decreases may reflect more extensive diffuse hypomyelination than what the structural MRI suggested.319 H1-MRS can be combined with phosphorus 31 MRS (31P) to confirm the metabolic changes observed in the H1-MRS studies in epilepsy patients to identify the location(s) of the seizure foci. These studies have shown ictal increases in inorganic phosphorus (Pi) and Lac and decreases in phosphocreatine (PCr) levels and pH as measured by P31-MRS in TLE.174,313,369 Postictally, pH was noted to increase with development of alkalosis before Lac normalization.174,317 In addition, ipsilateral

increased Pi and decreased phosphomonoester (PME; phosphocholine and/or phosphoethanolamine) levels without significant differences in other metabolites have been noted in TLE.152 Kuzniecky et al. have found in their study of patients with TLE that the PCr/Pi ratio was reduced by 50% in the epileptogenic temporal lobe compared with controls and by 35% when compared with the unaffected contralateral temporal lobe185; in their study, however, they did not observe any significant differences in intracellular pH between controls and patients.185

Alzheimer’s Disease Alzheimer’s disease is the most common form of dementia in older adults and represents approximately 40% to 60% of all dementias.36,53 The disease is characterized by decreased cortical acetylcholine, neuronal loss, amyloid deposits, and neurofibrillary tangles. Diagnosis, which may be difficult, is based primarily on clinical criteria, and imaging has been used mostly for excluding other possible causes of dementia such as intracranial mass lesions and subcortical vascular disease. Nevertheless structural MRI can confirm clinical findings in Alzheimer’s disease by demonstrating preferential atrophy of the hippocampi and temporoparietal cortex. Proton MRS, however, may enable earlier diagnosis of this disease.253,287,288 The proton MR spectrum in Alzheimer’s disease demonstrates an abnormally elevated MI/Cr ratio and an abnormally decreased NAA/Cr ratio.168,232,235,236,291,314 This combination of spectroscopic findings distinguishes Alzheimer’s disease from normal conditions in the elderly,168,314 whereas the elevated MI/Cr ratio distinguishes Alzheimer’s disease from other dementias, with the possible exception of Pick’s disease.291 Reduced levels of NAA in Alzheimer’s disease are noted in the frontoparietal regions, temporal lobes (including hippocampi), and posterior cingulate region.36,168,305,334 In Pick’s disease, however, decreased NAA levels and elevated MI levels  are noted in the frontal lobes. Because no such similar metabolic abnormalities are present in these brain regions in those with Alzheimer’s disease, the differentiation between Alzheimer’s and Pick’s disease can be made on the basis of proton MRS findings. More than 90% of cases may be correctly classified based on these differences in MRS findings.36,103 However, many confounding metabolic disorders can also result in abnormal MI/Cr ratios; renal  failure, diabetes mellitus, chronic hypoxic encephalopathy, and hypernatremia can all result in elevated MI concentrations, whereas hepatic encephalopathy and hyponatremia can result in decreased MI concentrations.179-181,194,230,291,352

Degenerative and Metabolic Brain Disorders Various metabolic disorders, including primary leukodystrophies and mitochondrial disorders, have been studied with proton MRS. Although the MRS findings, conventional MRI findings, and clinical symptoms of most of these degenerative brain disorders (many of which present in childhood) are nonspecific, MRS can sometimes be useful in distinguishing the various clinical entities. Wang and Zimmerman have divided brain metabolic disorders into five general categories359: 1. Disorders of lipid metabolism 2. Disorders of carbohydrate metabolism and respiratory chain 3. Disorders of amino acid metabolism and the urea cycle 4. Primary white matter disorders 5. Miscellaneous metabolic disorders not fitting into the previously described categories Peroxisomal and mitochondrial disorders may be considered to be subcategories of these five categories.359 Disorders of lipid metabolism include the following359:

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy • Abnormalities of long-chain fatty acid metabolism (peroxisomal diseases), such as Zellweger’s syndrome (in which peroxisomes are deficient or absent) • Rhizomelic chondrodysplasia punctata (in which several peroxisomal enzymes are absent) • Adrenoleukodystrophy (ALD) (in which a single peroxisomal enzyme is deficient) ALD is unique in that among hereditary leukodystrophies it is the only one associated with an increased Cho signal at times other than during the early stage of acute demyelination.36 In childhood ALD, decreased NAA/Cr ratio and increased Cho/Cr ratio are present in addition to elevated Lac, Glu, Gln, and MI signals in the white matter lesions.53,343 A recent quantitative study performed at 4 T identified seven metabolites as markers of lesion development. Of these metabolites, NAA, Gln, and [lipids + Lac] were the strongest markers.254 The decline in the NAA/Cr ratio in X-linked ALD both parallels disease progression and can predate the emergence of white matter hyperintensities on conventional T2-weighted MRI scans.74,309 Elevated Cho and reduced NAA have been reported in normal-appearing white matter in ALD patients.182,300 Reduction in NAA/Cho in normalappearing white matter on conventional MRI may be suggestive of   the onset of white matter degeneration and may predict lesion progression.98 Disorders of carbohydrate metabolism include diabetes mellitus and galactosemia. In diabetes mellitus, MRS reveals elevated glucose signal and an acetone signal at 2.2 ppm, owing to hyperglycemia and ketogenesis.180,294 It is suggested that the extent of brain injury in severe diabetic ketoacidosis may be evaluated with MRS, because elevated Lac levels have been noted in individuals who eventually die of cerebral herniation and cardiac arrest.359 Increased glucose levels are observed in patients with type 1 diabetes, recurrent hypoglycemia, and hypoglycemia unawareness.81 No difference in brain glucose concentration was detected between patients with poorly controlled diabetes and healthy volunteers evaluated at the same concentrations of plasma glucose.307 In galactosemia, a condition in which one of various inborn errors of metabolism prevents normal conversion of galactose to glucose, neonates demonstrate elevated brain galactitol signals in proton MR spectra at 3.67 and 3.74 ppm at the time of symptom onset; this corresponds to elevated levels of galactitol in brain autopsy specimens in fatal cases.355,356,362 Disorders of the respiratory chain include mitochondrial disorders such as mitochondrial myopathy, encephalopathy, lactic acidosis,  and strokes (MELAS) and myoclonic epilepsy with ragged red fibers (MERRF), Kearns-Sayre syndrome, and pyruvate dehydrogenase deficiency (PDD). MRS may provide useful information regarding the extent of metabolic derangement, severity of disease, and response to therapy.91,133,173,359 All these diseases are characterized by brain Lac accumulation secondary to impaired mitochondrial oxidative phosphorylation.359 However, MRS may not detect elevated CNS Lac in all cases.88 Presence of elevated Lac in the spectrum may be affected by differences in types and severity of mitochondrial disorders, timing of the scan with respect to the course of the disease, and selection of examined brain region with respect to the affected brain tissue.203 MRSI is the preferred approach for evaluating distributions of cerebral metabolites in these disorders because the distribution of Lac in the brain may vary significantly across different anatomic regions and because metabolic abnormalities may be present even when no corresponding conventional brain MRI abnormalities are detected.50,213 Typically the most striking Lac elevations are noted in regions of brain displaying the most marked structural abnormalities on conventional MRI. Elevated Lac has been noted in the parietooccipital gray and white matter in MELAS; occipital cortex in MERRF; brainstem, basal ganglia, and

523

occipital cortex in Leigh’s syndrome; and cortical gray matter in Kearns-Sayre syndrome.296 All these syndromes are also associated with decreased NAA/Cr ratios, presumably related to neuronal loss or dysfunction.127,213 However, PDD has been associated with more marked reductions in NAA/Cr ratios in addition to elevated Lac levels  and decreased Cho/Cr ratios309,359,374; Zimmerman and Wang have noted one case in which NAA was completely absent from the spectrum.309,359,374 Disorders of amino acid metabolism include entities such as phenylketonuria (PKU), the most prevalent of this class of disorders, and maple syrup urine disease (MSUD). PKU results from a deficiency  of the hepatic enzyme phenylalanine hydroxylase and demonstrates an autosomal recessive mode of inheritance.306 In patients with PKU, the proton MR spectra are remarkable only for an abnormal elevation of the phenylalanine signal at 7.3 ppm, which is specific for this disease.359 In patients with MSUD, long TE proton MR spectra reveal a signal at 0.9 to 1.0 ppm in addition to a reversible decrease in  the NAA/Cr ratio during acute metabolic decompensation.106,358,359 Abnormal accumulation of valine, leucine, and isoleucine occurs in the CSF, blood, and tissues as a consequence of defective oxidative decarboxylation.94,359 Ornithine transcarbamylase deficiency is the most common  disorder of the urea cycle and is inherited in an X-linked dominant fashion. In patients with this condition, hyperammonemia and intracerebral Gln accumulation lead to cerebral edema, neuronal loss,  and white matter gliosis.176,359 Reported MRS abnormalities include elevated Gln levels in a case of hyperammonemia, and decreased  MI with a normal Gln level in one patient with a normal serum ammonia level.75,294 The primary white matter disorders include Canavan’s disease, Alexander’s disease, and Pelizaeus-Merzbacher disease, among others.20 In patients with Canavan’s disease, which is the only primary white matter disorder with truly specific findings on MRS, markedly increased NAA levels have been noted, probably secondary to impaired NAA breakdown related to a deficiency in the enzyme aspartoacyclase.36,53,127,210 Canavan’s disease is the only metabolic disorder associated with an elevated NAA level.358,359,367 Other findings in Canavan’s disease include elevated MI levels and decreased Cho and Cr signals; the decreased Cho is probably due to failure of myelination, whereas a decreased Cr level may reflect spongy degeneration.359 In patients with Alexander’s disease, decreased NAA levels and elevated Lac levels are observed, most prominently in the frontal lobe; the structural white matter abnormality extends from the frontal lobes posteriorly.127,359 Pelizaeus-Merzbacher disease has been characterized by normal spectra or slight decreases in NAA in the basal ganglia in the early stages and by more prominent NAA decreases and Cho increases in advanced disease.127,190 In contrast, a recent single-voxel proton MRS study of five patients with genetically proven PelizaeusMerzbacher disease showed increased concentrations of NAA, Gln, MI, and Cr and decreased Cho in affected white matter.239 It was concluded that the observed spectroscopic pattern, suggesting increased neuroaxonal density, astrogliosis, and reduction of oligodendroglia, was consistent with the histopathologic characteristics of dysmyelination and hypomyelination.239 Several as-yet unclassified primary white matter diseases have also been reported. For example, van der Knaap and coauthors348 have reported a new leukoencephalopathy in nine children, manifested by severe white matter degeneration with elevated Lac and Glu and virtual disappearance of most other metabolites in MR spectra.359 A leukoencephalopathy associated with an inborn error of metabolism of   the polyols arabinitol and ribitol (resonance signals in the range 3.5-4.0 ppm) was described recently.239 Tedeschi and associates have

http://radiologyebook.com

524

PART II  CT and MR Imaging of the Whole Body

described a new white matter syndrome resulting from a metabolic defect producing hypomyelination and secondary axonal dysfunction characterized by a prominent decrease in NAA, Cho, and Cr levels and an increase in Lac.36,337 In general, childhood demyelinating disorders and neuronal degenerative disorders typically demonstrate decreased NAA/Cr ratios, with the degree of decrease corresponding to the extent of white matter disease and degree of cortical atrophy, respectively, depicted on conventional MRIs.53,351 Decreased NAA levels and elevated Lac levels are noted in Schilder’s and Cockayne’s diseases in addition to the disorders described earlier.53,127 Miscellaneous metabolic disorders not fitting into the previously described categories are characterized by specific pathologic changes in proton MR spectra. For example, abnormal lipid signals are seen in Niemann-Pick disease type C (a lysosomal disorder), low Cr levels are present in guanidinoacetate methyltransferase deficiency, and elevated Gly levels are present in nonketotic hyperglycinemia.295,325 In hepatic encephalopathy, increased levels of Glx (due to increased concentrations of Gln) and decreased levels of MI and Cho in proton MRS  have been described; the Gln elevation is thought to be secondary to hyperammonemia.292,359 MRS may even be able to detect subclinical encephalopathy, in that the reduction in MI level may occur even in patients with hepatic disease without neurologic symptoms.53,181,293 The specificity of these abnormalities may be useful in both diagnosis and monitoring of therapeutic efficacy in these conditions.

Stroke Stroke is another clinical condition in which MRS has been used. Its utility in the acute stroke setting is less than that of DWI, PWI, or even noncontrast CT, because rapid imaging is critical in the hyperacute and acute settings for decision making regarding need for thrombolysis. However, proton MRS may be useful in the chronic phase248 for assessing prognosis and monitoring recovery from stroke.248,258,259,263 MRS has also been an important research tool for evaluating the mechanisms of ischemic damage to brain tissue.251,320 The first detectable change in the MR spectrum that occurs following the onset of cerebral ischemia is the appearance of the Lac doublet.286 The Lac signal has been noted within minutes of induction of ischemia in animal models, and its level tends to rise over the first few hours after onset.3,32,238,242,286 A blood flow threshold of 20 mL/100 g of brain tissue per minute has been described in animal models for the appearance of Lac, but whether a similar threshold exists for humans is not certain.82 Studies in humans have also noted the presence of Lac during the first few days following symptom onset, correlated with ischemia and resultant edema.19,248,360 Not only does MRS demonstrate spectral changes before comparable changes in signal intensity of ischemic tissue on conventional T2-weighted MRIs occur, but also the extent of cerebral tissue displaying metabolic abnormalities exceeds the extent of signal hyperintensity on the T2-weighted images.19,120,242,286 In addition, the concentration of Lac in ischemic tissue is not homogeneous; rather, the center of the infarct tends to demonstrate higher concentrations than the infarct periphery.114 Concentrations of Lac during the evolution of an infarct also demonstrate temporal as well as spatial variability.121,286 Although the presence of Lac signifies ischemia, it does not necessarily signify actual irreversible infarction.19,32,116,286 A later and more specific spectral change signifying actual cerebral infarction is reduced NAA concentration.286 Several studies have demonstrated both a reduction in NAA and appearance of Lac in human stroke.19,123,304,360 The reduced NAA level is thought to be related to actual neuronal damage (neuronal cell death or replacement by non–NAA-containing cells such as glial cells, which occurs in a delayed

fashion relative to the changes seen in Lac levels) and is correlated highly with poor clinical outcome.242,244,248,286 Pereira and coworkers used proton MRS and T2-weighted structural MRI to study a series of 31 patients with new middle cerebral artery distribution infarcts within the first 72 hours following clinical onset of stroke symptoms. The authors noted that the combination  of infarct volume and NAA concentration could accurately predict clinical outcome.266 In the acute stage, decreases in NAA levels can often be identified in initial MR spectra. An animal model demonstrated NAA reduction within 30 to 60 minutes following induction of cerebral infarction.138,262,286 Regional variability in NAA concentrations is noted, just as with Lac, and the areas demonstrating the greatest elevation of   Lac within an infarct also demonstrate the most marked NAA depression.19,105,121,286 Changes in Cho and Cr concentrations during acute infarction are more variable than changes involving Lac and NAA. Many studies have reported decreases in Cr levels.42,105,108,114,122,304 Nevertheless, a few studies documented no consistent change in Cr levels,108,114,121,258 and two studies have even noted increased levels of Cr.19,120 Variability in Cho levels has been equally problematic. Elevated Cho levels have been reported in acute infarction, and this elevation may be related to ischemic demyelination.19,120,286 However, this finding is not consistent; several studies have noted decreased42,121 or unchanged108,114,304 Cho levels. In the subacute stage, Lac levels decrease progressively from those present in the acute stage, with an approximately 36% reduction in Lac levels noted per week.122 Lac may disappear or persist at low levels in the chronic stage; it may also disappear in the subacute stage and recur in the chronic stage.19,115,121,122,144,302 Reasons for the persistence of Lac signals into the subacute and chronic phases are not entirely clear.286 In the chronic stage, NAA, Cr, and Cho levels decrease.286,304 Specifically the NAA level decreases in an irreversible fashion at an average rate of approximately 29% per week.19,105,114,122,144 The rate of decrease of Cr and Cho is less than that of NAA, although these levels also decrease in the subacute and chronic phases.97,105,114,286 The smaller relative decrease of Cho and Cr is probably related to both the relatively decreased vulnerability of glial cells to ischemia compared to neurons, and the reactive gliosis that occurs as a result of ischemic brain injury.122,286

Hypoxic-Ischemic Injury/Encephalopathy MRS has been used to evaluate brain damage due to perinatal asphyxia, or hypoxic-ischemic injury (HIE).4,28,124,128 HIE is the main cause of developmental injury leading to cerebral palsy. Standard images in neonates (cranial sonography, CT, and conventional MRI) obtained in the first few days after injury may appear near normal. Proton MRS and DWI can detect brain injury on the first day of life. However, to reveal the overall pattern and extent of HIE, DWI should be performed 2 to 4 days after insult.27,222 MRS can detect acute brain injury even in cases with negative conventional MRI and DWI.257 However, similar to DWI, the pattern of MRS evolves with time after the injury. Both diffusivity and metabolite ratios were reported to be abnormal on the first or second day of life.27,222 The detected abnormal findings continued to worsen in most patients until the fourth or fifth day, after which the diffusivity and metabolite ratios started to normalize.27 In many cases the most severe damage (selective neuronal necrosis) involves the occipital gray matter and is manifested as a reduction in NAA and Cr levels and a proportional increase in Cho levels.4 This suggests that significant astrocyte survival may be present in these cases of necrosis secondary to HIE.4 In particularly severe cases, a Lac

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy signal may also be present, with the Lac signal appearing during the first 24 hours and persisting for at least 48 hours.265 This finding often precedes the reduction in the NAA/Cr ratio and the increase in the Cho/Cr ratio, both of which have been noted on MRS studies performed 1 to 2 weeks after the initial insult.309 The severity of the MRS changes appears to correlate with the severity of the initial insult and clinical outcome; persistent NAA and Cho signal abnormalities and the presence of Lac are associated with poor prognosis.26,128,141,315 Specifically, the presence of Lac in the acute phase has been associated with poor neurocognitive status at 1 year of age.26 Barkovich and coauthors have demonstrated that Lac/Cho ratios in the basal nuclei had particularly highly statistically significant associations with clinical outcome.26 The distribution of Lac elevation in the brain also correlates well with the severity of the injury; in severe asphyxia, the basal ganglia demonstrate greater Lac elevation than the arterial watershed regions; in mild to moderate asphyxia, the watershed regions demonstrate greater Lac elevation.25 Penrice and colleagues have shown that Lac/ NAA ratios in the thalami in particular correlate highly with degree of severity of asphyxia; the ratios in normal infants were less than 0.3, the ratios in infants with asphyxia were greater than 0.4, and the ratios in severely affected infants were greater than 0.5.25,264 Pu and others have shown that cases of moderate to severe  HIE display detectable brain Glu/Gln (Glx) peaks on proton MR spectra.280 Elevated Glx/(Cr + PCr) ratios in these cases correlated positively with the Sarnat stage of HIE on both initial and follow-up MRS studies.280

Cerebral Infections Most applications of proton MRS in evaluating intracranial infections have involved the study of acquired immunodeficiency syndrome (AIDS) and AIDS-related conditions. Nevertheless, evaluation of cerebral abscesses with MRS has also been documented. Cerebral abscesses can be differentiated from necrotic brain neoplasms by MRS. Grand and coworkers, in their series of 34 cystic intracranial lesions, showed that at a TE of 136 milliseconds, detection of an amino acid resonance at 0.9 ppm in bacterial abscesses allows differentiation from necrotic neoplasms, which do not demonstrate this spectral signal.125 MR spectra of brain abscesses demonstrate absent or low Cho and Cr levels and a relatively decreased level of NAA, as well as the possible presence of substantial amounts of Lac.53,285 MR spectra of abscesses may show other signals at 0.97, 1.24, 1.36, 1.50, 1.89, 2.02, and 2.14 ppm, and the exact significance of these additional signals is not known.53 Mishra and colleagues have reported results from their study of 52 patients with cystic intraparenchymal ringenhancing lesions with surrounding edema, in which they compared the relative sensitivity of DWI (with ADC mapping) and proton MRS.233 Their chosen criteria for abscess diagnosis were ADC values less than 0.9 ± 1.3 × 10-3 mm2/sec and presence of Lac cytosolic amino acids either with or without presence of succinate, acetate, alanine, or glycine signals on MRS.194 Using these criteria, they found sensitivity for differentiation of abscess from nonabscess to be 0.72 for DWI/ADC and 0.96 for MRS.233 Lac and amino acid signals with or without other metabolites were observed in 25 of 29 cases of abscesses.233 In three patients with neurocysticercosis in their cohort, nonspecific amino acid signals (n = 2), Lac (n = 3), acetate (n =1), succinate (n = 3), Cho (n = 2), and alanine (n = 3) were seen.233 Whereas diffusion restriction was suggestive of brain abscess, in the absence of diffusion restriction, proton MRS was necessary to accurately distinguish brain abscesses from cystic tumors.233 Various studies have shown resonances from acetate, succinate, and some amino acids (signals not seen in brain neoplasms) in addition to

525

Lac, Ala, and lipid signals (which may also be seen with certain brain tumors).56 Similar findings have been reported by Castillo’s group and others in cases of toxoplasmosis and cysticercosis.53,56 The few reports in the literature regarding MR spectra of other infectious processes include studies of intracranial tuberculomas, Creutzfeldt-Jakob disease, and herpes simplex encephalitis.56 Prominent lipid resonances have been noted in intracranial tuberculomas, with particularly important signals at 1.3 and 0.9 ppm corresponding to the methylene and terminal methyl groups of fatty acids, respectively.56 Abnormally decreased NAA levels in both white and gray matter, as well as abnormally increased levels of MI in the white matter, have been noted in Creutzfeldt-Jakob disease, whereas reductions in NAA/Cho and NAA/Cr ratios and the presence of Lac signals have been noted in herpes encephalitis.56

Acquired Immunodeficiency Syndrome Various studies of applications of MRS to the evaluation of   metabolic disturbances related to human immunodeficiency virus (HIV) infection have been recently performed; these studies  have demonstrated reduced NAA concentrations in HIV-positive  patients,63-65,67,160,231,301,329,342 including those who are asymptomatic227 or who display normal structural MRI findings.342 Increases in Cho and progressive decreases in NAA levels have been noted with disease progression.63-65,67,160,231,301,329,342 Increased levels of Gln/Glu (Glx) and MI have been noted in  the late stages of HIV dementia.64 One study reported a decrease of both NAA and Cr along with increases in Cho and MI, both in structural brain lesions and normal-appearing areas of brain (on conventional MRI) in patients with AIDS.36,225 It is not clear whether the metabolic abnormalities in AIDS are due directly to the virus or represent secondary effects related to opportunistic infection or AIDS-related neoplasms.36 Nevertheless it appears that MRS may be able to detect subclinical metabolic abnormalities prior to clinical and brain structural manifestations of the disease.295 The diffuse cortical atrophic changes and white matter hyperintensities noted  on T2-weighted images are late findings in the AIDS dementia complex. MRS may be particularly useful in the evaluation of infants of HIVinfected mothers; it may be able to demonstrate abnormalities in the infant brain as early as 10 days after birth, whereas seroconversion is difficult to determine during the first 6 months and conventional brain MRIs may be normal up to about 1 year of age.53,78,206 Toxoplasmosis, lymphoma, cryptococcal infection, and progressive multifocal leukoencephalopathy (PML) are all frequently  seen in HIV-positive patients and may be difficult to distinguish  based on structural MRI alone.204 The distinction among these conditions is critical for determination of optimal therapy, because major differences in treatment approaches exist among these conditions.204 MRS can allow differentiation because each entity presents with a relatively unique metabolic signature.66,204 Toxoplasmosis demonstrates decreased MI, Cho, Cr, and NAA, with a 15- to 20-fold increase in Lac/lipids.66,204 CNS lymphomas demonstrate increased Cho, Lac, and lipids (similar magnitude) as well as decreased NAA and MI.66,204 PML tends to demonstrate increased MI and Cho and decreased NAA and Cr.66,204 Cryptococcal infection is associated with decreased MI, Cho, Cr, NAA, and increased lipids but  not Lac.66,204 MRS may also prove to be a useful means of monitoring both disease progression and treatment effects in patients with AIDS.65,248,301 Specifically it has been suggested that monitoring of the efficacy of antiretroviral therapy—and even prediction of a patient’s response to therapy—may be performed with MRS.295,365

http://radiologyebook.com

526

PART II  CT and MR Imaging of the Whole Body

Multiple Sclerosis Multiple sclerosis (MS) is a disease that is manifested primarily as multifocal demyelination, although significant axonal injury and wallerian degeneration are usually present as well.11,71,155,215,248,295 In MS, the axonal loss rather than the demyelination may be responsible for the neurologic impairment seen in these patients.214,295 Active demyelinating plaques generally demonstrate enhancement on MR scans following Gd administration, in addition to displaying hyperintensity on T2-weighted and FLAIR conventional MR sequences.53 Conventional MRI is the most sensitive imaging method to diagnose MS; however, MRI’s ability to detect lesion heterogeneity and involvement of surrounding brain tissue is limited. Proton MRS can evaluate the degree of metabolic impairment in lesions, normalappearing white matter on conventional MRI, and gray matter.90 Substantial reductions in NAA concentration are noted in  acute (active) MS lesions, and these reductions correlate highly  with clinical neurologic impairment. Partial recovery of NAA levels is not uncommon, and these changes are also reflected in clinical improvement; whether the partial recovery of NAA levels is due  to reduction of edema or to actual recovery of neuronal function is not clear.48,295 In the acute phase, marked increases in Cho concentration, moderate increases in Lac, and increases in mobile lipid signals and “marker signals” in the range of 2.0 to 2.6  ppm may be present.11,36,129 The increase in Cho levels is due to release of phosphocholine and glycerophosphocholine during active demyelination.36 The mobile lipid signals are products of myelin breakdown, and the marker signals are of uncertain etiology but appear to be typical of demyelinating conditions in general.36,53 It has been suggested that lipid concentration increases may precede signal hyperintensity development on T2-weighted conventional MR sequences.246,295 The presence of Lac may be related to local inflammatory infiltrates associated with the demyelinating plaques and their effects on local cerebral vasculature.11,295 A transient but significant decrease in Cr levels in the hyperacute stage may revert to normal in the subacute and chronic stages.295 Increases in MI concentrations may be seen on short TE spectra; MI is usually detected when the demyelinating process is severe.36,88,295 Elevated Glu levels reported recently in acute MS lesions (and normal-appearing white matter but not chronic lesions) suggest an association between axonal injury and impaired metabolism  of Glu.322 Long TE proton MR spectra from acute MS lesions showing decreased NAA and elevated Cho and Lac levels may resemble spectra of brain tumors.299 However, elevated Glu/Gln levels in short TE MR spectra in tumefactive demyelinating lesions may help differentiate demyelination from neoplasms, which do not typically show elevated Glu/Gln signals.69,322 Reduction of NAA/Cr ratios has been noted not only in structural lesions (visible demyelinating plaques) but even in white matter, which appears normal by conventional T2-weighted MRI; thus, these MRS findings suggest that conventional MR sequences may underestimate the true extent of brain tissue involvement by MS.36,87,110,292 Although much of the reduction has been attributed to decreases in absolute concentrations of NAA, controversy exists regarding possible increases in Cr levels as well.87,248,289 The metabolite changes noted in MRS may be better correlated with actual clinical status than the number or volume of hyperintense structural brain lesions on conventional T2-weighted sequences.10,341 The reduction in NAA in normalappearing white matter correlates best with the degree of clinical disability.295 Gonen and colleagues118 have suggested that measurement of whole-brain NAA concentration may serve as an effective and

reproducible measure of disease load in MS and may be used to measure disease progression. In the chronic phase of MS, axonal loss and mild cortical atrophy may occur. Although acute plaques demonstrate edema and demyelination, chronic plaques demonstrate gliosis and mild associated neuronal loss.36 Incomplete recovery of axonal damage leads to irreversible clinical disabilities. NAA levels are decreased both in the structural lesions identified on conventional MR sequences and in surrounding normalappearing white matter in the chronic phase, and this NAA decrease corresponds to the axonal and mild neuronal loss.295 Normal NAA levels in the acute phase of MS may suggest that the axons in these cerebral regions have not yet sustained permanent damage, whereas in the chronic phase, when axonal loss has occurred, NAA levels are decreased.53 Free mobile lipids and marker signals may also be seen in MR spectra in the chronic phase in MS.53 Because MRS provides valuable insight into the various stages of the evolution of MS plaques and the physiologic and metabolic changes associated with this evolution, in the near future it may also provide a reliable means of effectively monitoring responses to new therapeutic interventions.104 A recent study has already explored serial spectroscopic changes in active MS plaques during pharmacologic intervention.209 These interventions may be designed to target particular stages in the development and maturation of these plaques and thus may prevent progression of the disease.

OTHER APPLICATIONS OF PROTON MRS IN THE CENTRAL NERVOUS SYSTEM Other CNS applications of proton MRS include the study of psychiatric diseases, upper motor neuron disease (e.g., amyotrophic lateral sclerosis [ALS]), Huntington’s disease, and parkinsonian syndromes, among other clinical entities). In patients with schizophrenia, the most consistent findings  include decreased NAA levels in the temporal lobes.169 In bipolar affective disorder, elevated Cho, NAA, and MI levels in the basal  ganglia and elevated levels of Glu in the parietal lobes have been described.49,308 Children with attention deficit hyperactivity disorder (ADHD) demonstrate abnormally elevated NAA/Cr, Glu/Cr, and Cho/Cr  ratios in the frontal lobes when compared with age-matched controls.49 In patients with ALS, the NAA/Cr ratio is significantly decreased in the motor and premotor cortices, and similar decreases are present  in the brainstem.36,85 In Huntington’s disease, a reduced NAA/Cho ratio is present in both the basal ganglia and the cerebral cortex, and Lac elevations  in these areas may also be present.162,295 MRS evaluation of the parkinsonian syndromes has revealed typical absence of metabolic abnormalities in the basal ganglia in idiopathic Parkinson’s disease, as opposed to reductions in the basal ganglia NAA/Cr and NAA/Cho ratios in other parkinsonian syndromes such as progressive supranuclear palsy and corticobasal degeneration.295 Another entity for which proton MRS has been clinically useful is hepatic encephalopathy (HE). Ross et al.293 used proton MRS to study 20 patients with cirrhosis and found that compared with control  subjects, patients with liver disease and no HE showed a reduction in Cho/Cr without altered MI/Cr or Glx/Cr. Patients with subclinical HE  had, in addition to lower Cho/Cr, a substantial reduction in MI/Cr  and increases in the beta-, gamma-, and alpha-Glx regions of the spectrum.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy

REFERENCES 1. Achten E: Aspects of proton MR spectroscopy in the seizure patient. Neuroimaging Clin N Am 8:849–862, 1998. 2. Achten E, Santens P, Boon P, et al: Single-voxel proton MR spectroscopy and positron emission tomography for lateralization of refractory temporal lobe epilepsy. AJNR Am J Neuroradiol 19:1–8, 1998. 3. Aetna Clinical Policy Bulletin: Magnetic Resonance Spectroscopy (MRS), April 29, 2014. 4. Allen K, Busza AL, Crockard HA, et al: Acute cerebral ischaemia: Concurrent changes in cerebral blood flow, energy metabolites, pH, and lactate measured with hydrogen clearance and 31P and 1H nuclear magnetic resonance spectroscopy: III. Changes following ischaemia.   J Cereb Blood Flow Metab 8:816–821, 1988. 5. Al-Okaili RN, Krejza J, Wang S, et al: Advanced MR imaging techniques in the diagnosis of intraaxial brain tumors in adults. Radiographics 26:S173–S189, 2006. 6. Al-Okaili RN, Krejza J, Woo JH, et al: Intraaxial brain masses: MR imaging-based diagnostic strategy-initial experience. Radiology 243(2):539–550, 2007. 7. Ando K, Ishikura R, Morikawa T, et al: Regional differences of in vivo proton MR spectroscopy in developing human brain. Nippon Igaku Hoshasen Gakkai Zasshi 60:199–204, 2000. 8. Ashwal S, Babikian T, Gardener-Nichols J, et al: Susceptibility-weighted imaging and proton magnetic resonance spectroscopy in assessment of outcome after pediatric traumatic brain injury. Arch Phys Med Rehabil 87(12 Suppl 2):S50–S58, 2006. 9. Ashwal S, Holshouser B, Simmons PL, et al: Predictive value of proton magnetic resonance spectroscopy in pediatric closed head injury. Pediatr Neurol 23:730–737, 2000. 10. Arnold DL, Matthews PM, Francis GS, et al: Proton magnetic resonance spectroscopy of human brain in vivo in the evaluation of multiple sclerosis: Assessment of the load of the disease. Magn Reson Med 14:154–159, 1990. 11. Arnold DL, Matthews PM, Francis GS, et al: Proton magnetic resonance spectroscopic imaging for metabolic characterization of demyelinating plaques. Ann Neurol 31:235–241, 1992. 12. Arnold SM, Patchell RA: Diagnosis and management of brain metastases. Hematol Oncol Clin North Am 15:1085–1107, 2001. 13. Aronen HJ, Granzit E, Louis DN, et al: CBV maps of gliomas comparison with tumor grade and histologic findings. Radiology 191:41–51, 1994. 14. Arslanoglu A, Bonekamp D, Barker PB, et al: Quantitative proton MR spectroscopic imaging of the mesial temporal lobe. J Magn Reson Imaging 20:772–778, 2004. 15. Astrakas LG, Zurakowski D, Tzika AA, et al: Noninvasive magnetic resonance spectroscopic imaging biomarkers to predict the clinical   grade of pediatric brain tumors. Clin Cancer Res 10:8220–8228, 2004. 16. Austin SJ, Connelly A, Gadian DG, et al: Localized 1H NMR spectroscopy on Canavan’s disease: A report of two cases. Magn Reson Med 19:439–445, 1991. 17. Babb TL, Brown WJ: Pathological findings in epilepsy. In Engel J, Jr, editor: Surgical treatment of the epilepsies, New York, 1987, Raven Press, pp 511–540. 18. Barker P: Diagnosis and characterization of brain tumors: MR spectroscopic imaging. In Pillai JJ, editor: Functional brain tumor imaging, New York, 2014, Springer Publishing, pp 39–55. 19. Barker PB, Gillard JH, van Ziji PCM, et al: Acute stroke: Evaluation   with serial proton MR spectroscopic imaging. Radiology 192:723–732, 1994. 20. Barker PB, Horska A: Neuroimaging in leukodystrophies. J Child Neurol 19:559–570, 2004. 21. Barker PB, Lin DD: In vivo proton MR spectroscopy of the human brain. Prog Nucl Magn Reson Spectrosc 49:99–128, 2006. 22. Barker PB, Szopinski K, Horska A: Metabolic heterogeneity at the level of the anterior and posterior commissures. Magn Reson Med 43:348– 354, 2000.

527

23. Barkhoudarian G, Hovda DA, Giza CC: The molecular pathophysiology of concussive brain injury. Clin Sports Med 30:33–48, 2011. 24. Barkovich AJ: Pediatric neuroimaging, ed 2, Philadelphia, 1996, Lippincott-Raven, pp 47–52. 25. Barkovich AJ: Pediatric neuroimaging, ed 3, Philadelphia, 2000, Lippincott Williams & Wilkins, pp 55–59, 168, 199–202. 26. Barkovich AJ, Baranski K, Vigneron D, et al: Proton MR spectroscopy for the evaluation of brain injury in asphyxiated term neonates. AJNR Am J Neuroradiol 20:1399–1405, 1999. 27. Barkovich AJ, Miller SP, Bartha A, et al: MR imaging, MR spectroscopy, and diffusion tensor imaging of sequential studies in neonates with encephalopathy. AJNR Am J Neuroradiol 27:533–547, 2006. 28. Barkovich AJ, Westmark K, et al: Perinatal asphyxia: MR findings in the first 10 days. AJNR Am J Neuroradiol 16:427–438, 1995. 29. Bartha R, Megyesi JF, Watling CJ: Low grade glioma: Correlation of short echo time 1H-MR spectroscopy with 23Na MR imaging. AJNR Am J Neuroradiol 29:464–470, 2008. 30. Bendszus M, Warmuth-Metz M, Klein R, et al: MR spectroscopy in gliomatosis cerebri. AJNR Am J Neuroradiol 21:375–380, 2000. 31. Benuck M, D’Adamo JAF: Acetyl transport mechanisms: Metabolism on N-acetyl-l-aspartic acid in the non-nervous tissues of the rat. Biochem Biophys Acta 152:611–618, 1968. 32. Berkelbach van der Sprenkel JW, Luyten PR, et al: Cerebral lactate detected by regional proton magnetic resonance spectroscopy in a patient with cerebral infarction. Stroke 19:1556–1560, 1988. 33. Bieza A, Krumina G: The value of magnetic resonance spectroscopy and diffusion tensor imaging in characterization of glioma growth patterns and treatment efficiency. J Biomed Sci Eng 6:518–526, 2013. 34. Birken DL, Oldendorf WH: N-acetyl-l-aspartic acid: A literature review of a compound prominent in 1H-NMR spectroscopic studies of brain. Neurosci Biobehav Rev 13:23–31, 1989. 35. Bizzi A, Movsas B, Tedeschi G, et al: Response of non-Hodgkin lymphoma to radiation therapy: Early and long-term assessment with H-1 spectroscopic imaging. Radiology 194:271–276, 1995. 36. Bonavita S, Di Salle F, Tedeschi G: Proton MRS in neurological disorders. Eur J Radiol 30:125–131, 1999. 37. Bottomley PA, Edelstein WA, Foster TH, et al: In vivo solventsuppressed localized hydrogen nuclear magnetic resonance (NMR): A new window to metabolism? Proc Natl Acad Sci U S A 82:2148–2152, 1985. 38. Brand A, Richter-Landsberg C, Leibfritz D: Multinuclear NMR studies on the energy metabolism of glial and neuronal cells. Dev Neurosci 15:289–298, 1993. 39. Brat DJ, Castellano-Sanchez AA, Hunter SB: Pseudopallisades in glioma are hypoxic, express extracellular matrix proteases, and are formed   by and actively migrating cell population. Cancer Res 64:920–927, 2004. 40. Bray MD, Mullins ME: Metabolic white matter diseases and the utility of MR spectroscopy. Radiol Clin North Am 52:403–411, 2014. 41. Bruehlmeier M, Roelcke U, Schbiger PA, et al: Assessment of hypoxia and perfusion in human brain tumors using PET with 18 F-fluoromisonidazole and 15O-H2O. J Nucl Med 45:1851–1859, 2004. 42. Bruhn H, Frahm J, Gyngell ML, et al: Cerebral metabolism in man after acute stroke: New observations using localized proton NMR spectroscopy. Magn Reson Med 9:126–131, 1989. 43. Bruhn H, Frahm J, Gyngell ML, et al: Noninvasive differentiation of tumors with use of localized H-1 MR spectroscopy in vivo: Initial experience in patients with cerebral tumors. Radiology 172:541–548, 1989. 44. Burri R, Steffen C, Herschkowitz N: N-acetyl-l-aspartate is a major source of acetyl groups for lipid synthesis during rat brain development. Dev Neurosci 13:403–411, 1991. 45. Burtscher IM, Skagerberg G, Geijer B, et al: Proton MR spectroscopy and preoperative diagnostic accuracy: An evaluation of intracranial mass lesions characterized by stereotactic biopsy findings. AJNR Am J Neuroradiol 21:84–93, 2000. 46. Butzen J, Prost R, Chetty V, et al: Discrimination between neoplastic and nonneoplastic brain lesions by use of proton MR spectroscopy:

http://radiologyebook.com

528

47.

48.

49.

50. 51. 52. 53.

54.

55.

56. 57.

58.

59.

60. 61.

62.

63. 64.

65.

66. 67.

68.

69.

PART II  CT and MR Imaging of the Whole Body The limits of accuracy with a logistic regression model. AJNR Am J Neuroradiol 21:1213–1219, 2000. Byrant RA, Harvey AG: Postconcussive symptoms and posttraumatic stress disorder after mild traumatic brain injury. J Nerv Ment Dis 187:302–305, 1999. Castillo M, Green C, Kwock L, et al: Proton MR spectroscopy in patients with neurofibromatosis type I: Evaluation of hamartomas and clinical correlation. AJNR Am J Neuroradiol 16:141–147, 1995. Castillo M, Kwock L, Courvoisie HE, et al: Proton MR spectroscopy in psychiatric and neurodevelopmental childhood disorders: Early experience. Neuroimaging Clin N Am 8:901–912, 1998. Castillo M, Kwock L, Green C: MELAS syndrome: Imaging and proton MR spectroscopic findings. AJNR Am J Neuroradiol 16:233–239, 1995. Castillo M, Kwock L, Mukherji SK: Clinical applications of proton MR spectroscopy. AJNR Am J Neuroradiol 17:1–15, 1996. Castillo M, Kwock L: Proton MR spectroscopy of common brain tumors. Neuroimaging Clin N Am 8:733–752, 1998. Castillo M, Kwock L, Scatliff J, et al: Proton MR spectroscopy in neoplastic and non-neoplastic brain disorders. Magn Reson Imaging Clin N Am 6:1–20, 1998. Castillo M, Smith JK, Kwock L: Correlation of myo-inositol levels and grading of cerebral astrocytomas. AJNR Am J Neuroradiol 21:1645– 1649, 2000. Caulo M, Panara V, Tortora D, et al: Data-driven grading of brain gliomas: A multiparametric MR imaging study. Radiology 2014. doi: 10.1148/radiol.14132040. Cecil KM, Lenkinski RE: Proton MR spectroscopy in inflammatory and infectious brain disorders. Neuroimaging Clin N Am 8:863–880, 1998. Cecil KM: Technical aspects of performing and interpreting proton MR spectroscopy (syllabus, advanced imaging symposium: Preparing the neuroradiologist for the new millennium, pp 29-36). Paper presented at Annual Meeting of the American Society of Neuroradiology, Atlanta, 2000. Cendes F, Andermann F, Preul MC, et al: Lateralization of temporal lobe epilepsy based on regional metabolic abnormalities in proton magnetic resonance spectroscopic images. Ann Neurol 35:211–216, 1994. Cendes F, Caramano Z, Andermann F, et al: Proton magnetic resonance spectroscopic imaging and magnetic resonance imaging volumetry in the lateralization of temporal lobe epilepsy: A series of 100 patients. Ann Neurol 42:737–746, 1997. Ceodan S, Parrilla R, Santoro J, et al: H-1 NMR detection of cerebral myoinositol. FEBS Lett 187:167–172, 1985. Chan A, Lau A, Pirzkall A: Proton magnetic resonance spectroscopy imaging in the evaluation of patients undergoing gamma knife surgery for grade IV glioma. J Neurosurg 101:460–475, 2004. Chan Y-l, Yeung DKW, Leung S-F, et al: Proton magnetic resonance spectroscopy of late delayed radiation-induced injury of the brain.   J Magn Reson Imaging 10:130–137, 1999. Chang L: In vivo magnetic resonance spectroscopy in HIV and HIV-related brain diseases. Rev Neurosci 6:365–378, 1995. Chang L, Ernst T: MR spectroscopy and diffusion-weighted MR imaging in focal brain lesions in AIDS. Neuroimaging Clin N Am 7:409–426, 1997. Chang L, Ernst T, Leonido-Yee M, et al: Cerebral metabolite abnormalities correlate with clinical severity of HIV-1 cognitive motor complex. Neurology 52:100–108, 1999. Chang L, Miller BL, McBride D, et al: Brain lesions in patients with AIDS: H-1 MR spectroscopy. Radiology 197:525–531, 1995. Chang L, Ernst T, Tornatore C, et al: Metabolite abnormalities in progressive multifocal leukoencephalopathy by proton magnetic resonance spectroscopy. Neurology 48:836–845, 1997. Chiang CS, McBride WH, Withers HR: Radiation-induced astrocytic   and microglial responses in mouse brain. Radiother Oncol 29:60–68, 1998. Chintala SK, Tonn JC, Rao JS: Matrix metalloproteases and their biological function in human gliomas. Int J Dev Neurosci 17:495–502, 1999.

70. Choi CG, Ko TS, Lee HK, et al: Localized proton MR spectroscopy of the allocortex and isocortex in healthy children. AJNR Am J Neuroradiol 21:1354–1358, 2000. 71. Cianfoni A, Niku S, Imbesi SG: Metabolite findings in tumefactive demyelinating lesions utilizing short echo time proton magnetic resonance spectroscopy. AJNR Am J Neuroradiol 28:272–277, 2007. 72. Claes A, Idema AJ, Wessling P: Diffuse glioma growth: A guerilla war. Acta Neuropathol 114:443–458, 2007. 73. Comair YG, Ng T, Xue M, et al: Early post-ictal lactate detection in temporal lobe epilepsy for localization of seizure focus: A chemical shift imaging study (abstract). In Book of Abstracts. San Francisco, Society of Magnetic Resonance in Medicine, 1994, p 401. 74. Confort-Gouny S, Vion-Dury J, Chabrol B, et al: Localized proton magnetic resonance spectroscopy in X-linked adrenoleukodystrophy. Neuroradiology 37:568–575, 1995. 75. Connelly A, Cross J, Gadian D, et al: Magnetic resonance spectroscopy shows increased brain glutamine in ornithine carbamoyl transferase deficiency. Pediatr Res 33:77–81, 1993. 76. Connelly A, Jackson GD, Duncan JS, et al: Proton MRS in temporal lobe epilepsy. Neurology 44:1411–1417, 1994. 77. Connelly A, Paesschen WV, Porter DA, et al: Proton magnetic resonance spectroscopy in MRI-negative temporal lobe epilepsy. Neurology 51:61–66, 1998. 78. Cortey A, Jarvik JG, Lenkinski RE, et al: Proton MR spectroscopy of brain abnormalities in neonates born to HIV-positive mothers. AJNR Am J Neuroradiol 15:1853–1859, 1994. 79. Cousins JP, Seymour PA, Weaver S, et al: Clinical differentials of CNS tumors and radiation necrosis using single voxel H1-MRS correlated with pathology. Proceedings of the 5th Annual Scientific Meeting of the International Society for Magnetic Resonance in Medicine, Vancouver, BC, 1997, p 1126. 80. Crawford FW, Khayal IS, McGue C, et al: Relationship of pre-surgery metabolic and physiological MR imaging parameters to survival for patients with untreated GBM. J Neurooncol 91:337–351, 2009. 81. Criego AB, Tkac I, Kumar A, et al: Brain glucose concentrations in patients with type 1 diabetes and hypoglycemia unawareness. J Neurosci Res 79:42–47, 2005. 82. Crockard HL, Gadian DG, Frackowiak SJ, et al: Acute cerebral ischaemia: Concurrent changes in cerebral blood flow, energy metabolites, pH, and lactate measured with hydrogen clearance and 31P and 1H nuclear magnetic resonance spectroscopy: II. Changes during ischaemia. J Cereb Blood Flow Metab 7:394–402, 1987. 83. Cross JH, Connelly A, Jackson GD, et al: Proton magnetic resonance spectroscopy in children with temporal lobe epilepsy. Ann Neurol 39:107–113, 1996. 84. Croteau D, Scarpace L, Hearshen D, et al: Correlation between magnetic resonance spectroscopy imaging and image-guided biopsies: Semiquantitative and qualitative histopathological analyses of patients with untreated glioma. Neurosurgery 49(4):823–829, 2001. 85. Cwik VA, Hanstock CC, Allen PS, et al: Estimation of brainstem neuronal loss in amyotrophic lateral sclerosis with in vivo proton magnetic resonance spectroscopy. Neurology 50:72–77, 1998. 86. D’Adamo AF, Jr, Smith JC, Woiler C: The occurrence of N-acetylaspartate amidohydrolase (aminoacylase II) in the developing rat. J Neurochem 20:1275–1278, 1973. 87. Davie CA, Barker GJ, Webb S, et al: Persistent functional deficit in multiple sclerosis and autosomal dominant cerebellar ataxia is associated with axon loss: Part 6. Brain 118:1583–1592, 1995. 88. Davie CA, Hawkins CP, Barker GJ, et al: Serial proton magnetic resonance spectroscopy in acute multiple sclerosis lesions: Part 1. Brain 117:49–58, 1994. 89. DeAngelis LM: Brain tumors. N Engl J Med 344:114–123, 2001. 90. De Stefano N, Filippi M: MR spectroscopy in multiple sclerosis.   J Neuroimaging 17(Suppl 1):31–35, 2007. 91. De Stefano N, Matthews P, Ford B, et al: Short-term dichloroacetate treatment improves indices of cerebral metabolism in patients with mitochondrial disorders. Neurology 45:1193–1198, 1995.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy 92. Devos A, Lukas L, Suykens JA, et al: Classification of brain tumours using short echo time 1H MR spectra. J Magn Reson 170:164–175, 2004. 93. Di Constanzo A, Scarabino T, Trojsi F, et al: Multiparametric 3T MR approach to the assessment of cerebral gliomas: Tumor extent and malignancy. Neuroradiology 48:622–631, 2006. 94. Diezel P, Martin K: Die Ahornsirupkrankheit mit familarem Befall. Virchows Arch Pathol Anat 337:425–445, 1964. 95. Dowling C, Bollen AW, Noworolski SM, et al: Preoperative proton   MR spectroscopic imaging of brain tumors: Correlation with histopathologic analysis of resection specimens. AJNR Am J Neuroradiol 22:604–612, 2001. 96. Duarte JM, Len H, Miynarik V, et al: The neurochemical profile quantified by in vivo 1H NMR spectroscopy. Neuroimage 61:342–362, 2012. 97. Duijn JH, Matson GB, Maudsley AA, et al: Human brain infarction: Proton MR spectroscopy. Radiology 183:711–718, 1992. 98. Eichler FS, Barker PB, Cox C, et al: Proton MR spectroscopic imaging predicts lesion progression on MRI in X-linked adrenoleukodystrophy. Neurology 58:901–907, 2002. 99. Elmogy SA, Mousa AE, Elashry MS, et al: MR spectroscopy in post-treatment follow-up of brain tumors. Egypt J Radiol Nuc Med 42:413–424, 2011. 100. El Sherbeny A, El-Shafey H, Biomy SL, et al: Diagnostic yield of combined magnetic resonance spectroscopy and diffusion weighted imaging in intracranial neoplasms. Egypt J Radiol Nuc Med. 20(4):849– 858, 2014. 101. Ende G, Laxer K, Knowlton R, et al: Temporal lobe epilepsy: Bilateral hippocampal metabolite changes revealed at proton MR spectroscopic imaging. Radiology 202:809–817, 1997. 102. Ende J, Scatliff JH, Powers S, et al: Spectral proton and P-31 MR spectroscopy patterns of treated human brain tumors. Paper presented at the Annual Meeting of the Society of Magnetic Resonance in Medicine, 1992. 103. Ernst T, Chang L, Melchor R, et al: Frontotemporal dementia and early Alzheimer disease: Differentiation with frontal lobe H-1 MR spectroscopy. Radiology 203:829–836, 1997. 104. Fazekas F, Soelberg-Sorensen P, Comi G, et al: MRI to monitor treatment efficacy in multiple sclerosis. J Neuroimaging 17(Suppl 1):50–55, 2007. 105. Felber SR, Aichner FT, Sauter R, et al: Combined magnetic resonance imaging and proton magnetic resonance spectroscopy of patients with acute stroke. Stroke 23:1106–1110, 1992. 106. Felber SR, Sperl W, Chemelli A, et al: Maple syrup urine disease: Metabolic decompensation monitored by proton magnetic resonance imaging and spectroscopy. Ann Neurol 33:396–401, 1993. 107. Felix R, Schorner W, Laniado M, et al: MR imaging and gadoliniumDPTA. Radiology 156:681–690, 1885. 108. Fenstermacher MJ, Narayana PD: Serial proton magnetic resonance spectroscopy of ischemic brain injury in humans. Radiology 25:1034– 1039, 1990. 109. Frahm J, Bruhn H, Gyngell ML, et al: Localized high-resolution   proton NMR spectroscopy using stimulated echoes: Initial   applications to human brain in vivo. Magn Reson Med 9(1):79–93, 1989. 110. Fu L, Matthews PM, De Stefano N, et al: Imaging axonal damage of normal appearing white matter in multiple sclerosis. Brain 121:103– 113, 1998. 111. Gadian DF, Connelly A, Duncan JS, et al: 1H magnetic resonance spectroscopy in the investigation of intractable epilepsy. Acta Neurol Scand Suppl 152:116–121, 1994. 112. Ganslandt O, Stadlbauer A, Fahlbusch R, et al: Proton magnetic resonance spectroscopic imaging integrated into image-guided surgery: Correlation to standard magnetic resonance imaging and tumor cell density. Neurosurgery 56(Suppl 2):291–298, 2005. 113. Garcia PA, Laxer KD, van der Grond J, et al: Proton magnetic resonance spectroscopic imaging in patients with frontal lobe epilepsy. Ann Neurol 37:279–281, 1995.

529

114. Gideon P, Henriksen O, Sperling B, et al: Early time course of N-acetylaspartate, creatine, and phosphocreatine, and compounds containing choline in the brain after acute stroke: A proton magnetic resonance spectroscopy study. Stroke 23:1566–1572, 1992. 115. Gideon P, Sperling B, Arlien-Soborg P, et al: Long-term follow-up of cerebral infarction patients with proton magnetic resonance spectroscopy. Stroke 25:967–973, 1994. 116. Gillard JH, Barker PB, van Zijl PCM, et al: Proton MR spectroscopy in acute middle cerebral artery stroke. AJNR Am J Neuroradiol 17:873–886, 1996. 117. Goebell E, Fiehler J, Ding X-O, et al: Disarrangement of fiber tracts and decline of neuronal density correlate in glioma patients: Combined diffusion tensor imaging and proton magnetic resonance spectroscopy study. AJNR Am J Neuroradiol 27:1426–1431, 2006. 118. Gonen O, Catalaa I, Babb JS, et al: Total brain N-acetylaspartate: A new measure of disease load in MS. Neurology 54:15–19, 2000. 119. Gonen O, Wang ZJ, Viswanathan AK, et al: Three-dimensional multivoxel proton MR spectroscopy of the brain in children with neurofibromatosis type I. AJNR Am J Neuroradiol 20:1333–1341, 1999. 120. Govindaraju V, Gauger GE, Manley GT, et al: Volumetric proton spectroscopic imaging of mild traumatic brain injury. AJNR Am J Neuroradiol 25:730–737, 2004. 121. Graham GD, Blamire AM, Howseman AM, et al: Proton magnetic resonance spectroscopy of cerebral lactate and other metabolites in stroke patients. Stroke 23:333–340, 1992. 122. Graham GD, Blamire AM, Rothman DL, et al: Early temporal variation of cerebral metabolites after human stroke. Stroke 24:1891–1896, 1993. 123. Graham GD, Kalvach P, Blamire AM, et al: Clinical correlates of proton magnetic resonance spectroscopy findings after acute cerebral infarction. Stroke 26:225–229, 1995. 124. Graham SH, Meyerhoff DJ, Bayne L, et al: Magnetic resonance spectroscopy of N-acetylaspartate in hypoxic-ischemic encephalopathy. Ann Neurol 35:490–494, 1994. 125. Grand S, Passaro G, Ziegler A, et al: Necrotic tumor versus brain abscess: Importance of amino acids detected at 1H MR spectroscopy— initial results. Radiology 213:785–793, 1999. 126. Graves EE, Nelson SJ, Vigneron DB, et al: Serial proton MR spectroscopic imaging of recurrent malignant gliomas after gamma knife radiosurgery. AJNR Am J Neuroradiol 22:613–624, 2001. 127. Grodd W, Krageloh-Mann W, Klose U, et al: Metabolic and destructive brain disorders in children: Findings with localized proton MR spectroscopy. Radiology 181:173–181, 1991. 128. Groenendaal F, Veenhoven RH, et al: Cerebral lactate and N-acetylaspartate/choline ratios in asphyxiated full-term neonates demonstrated in vivo using proton magnetic resonance spectroscopy. Pediatr Res 35:148–151, 1994. 129. Grossmann RI, Lenkinski RE, Ramer KN, et al: MR proton spectroscopy in multiple sclerosis. AJNR Am J Neuroradiol 13:1535– 1543, 1992. 130. Gupta RK, Cloughesy TF, Sinha U, et al: Relationship between choline magnetic resonance spectroscopy, apparent diffusion coefficient, and quantitative histopathology in human glioma. J Neurooncol 50:215–226, 2000. 131. Haga KK, Khor YP, Farrall A, et al: A systematic review of brain metabolite changes, measured with (1)H magnetic resonance spectroscopy, in healthy aging. Neurobiol Aging [Epub August 2007]. 132. Hanstock CC, Rothman DL, Pritchad JW, et al: Spatially localized 1H NMR spectra of metabolites in the human brain. Proc Natl Acad Sci U S A 85:1821–1825, 1988. 133. Harada M, Tanouchi M, Arai K, et al: Therapeutic efficacy of a case of pyruvate dehydrogenase complex deficiency monitored by localized proton magnetic resonance spectroscopy. Magn Reson Imaging 14:129–133, 1996. 134. Henry RG, Vigneron DB, Fischbein NJ, et al: Comparison of relative cerebral blood volume and proton spectroscopy in patients with treated gliomas. AJNR Am J Neuroradiol 21:357–366, 2000.

http://radiologyebook.com

530

PART II  CT and MR Imaging of the Whole Body

135. Herminghaus S, Dierks T, Pilatus U, et al: Determination of histopathological tumor grade in neuroepithelial brain tumors by using spectral pattern analysis of in vivo spectroscopic data. J Neurosurg 98:74–81, 2003. 136. Hetherington HP, Mason GF, Pan JW, et al: Evaluation of cerebral gray and white matter metabolite differences by spectroscopic imaging at 4.1 T. Magn Reson Med 32:565–571, 1994. 137. Hetherington H, Kuzniecky R, Pan J, et al: Proton nuclear magnetic resonance spectroscopic imaging of human temporal lobe epilepsy at 4.1 T. Ann Neurol 38:396–404, 1995. 138. Hida K: In vivo 1H and 31P NMR spectroscopy of the developing rat brain. Hokkaido J Med Sci 67:272–281, 1992. 139. Hipp SJ, Steffen-Smith E, Hammpud D, et al: Predicting outcome of children with diffuse intrinsic pontine gliomas using multiparametric imaging. Neuro-Oncol 13:904–909, 2011. doi: 10.1093/neuonc/nor076. 140. Hollingworth W, Medina LS, Lenkinski RE, et al: A systematic literature review of magnetic resonance spectroscopy for the characterization of brain tumors. AJNR Am J Neuroradiol 27:1404–1411, 2006. 141. Holshouser BA, Ashwal S, Luh GY, et al: Proton MR spectroscopy after acute nervous system injury: Outcome prediction in neonates, infants, and children. Radiology 202:487–496, 1997. 142. Holzer T, Herholz K, Jeske J, et al: FDG-PET as a prognostic indicator in radiochemotherapy of glioblastoma. J Comput Assist Tomogr 17:681–687, 1993. 143. Horska A, Kaufmann WE, Brant LJ, et al: In vivo quantitative proton MRSI study of brain development from childhood to adolescence.   J Magn Reson Imaging 15:137–143, 2002. 144. Houkin K, Kamada K, Kamiyama H, et al: Longitudinal changes in proton magnetic resonance spectroscopy in cerebral infarction. Stroke 24:1316–1321, 1993. 145. Hoult DI, Busby SJW, Gadian DG, et al: Observation of tissue metabolites using phosphorus nuclear magnetic resonance. Nature 252:285–287, 1974. 146. Hourani R, Brant LJ, Rizk T, et al: Can proton MR spectroscopic and perfusion imaging differentiate between neoplastic and nonneoplastic brain lesions in adults? AJNR Am J Neuroradiol 29:366–372, 2008. 147. Hourani R, Horska A, Albayram S, et al: Proton magnetic resonance spectroscopic imaging to differentiate between nonneoplastic lesions and brain tumors in children. J Magn Reson Imaging 23:99–107, 2006. 148. Howe FA, Barton SJ, Cudlip SA, et al: Metabolic profiles of human brain tumors using quantitative in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 49:223–232, 2003. 149. Howe FA, Maxwell RJ, Saunders DE, et al: Proton spectroscopy in vivo. Magn Reson Q 9:31–59, 1993. 150. Huang BY, Kwock L, Castillo M, et al: Association of choline levels and tumor perfusion in brain metastases assessed with proton MR spectroscopy and dynamic susceptibility contrast-enhanced perfusion weighted MRI. Technol Cancer Res Treat 9:327–338, 2010. 151. Hueppi PS, Posse S, Lazeyras F, et al: Developmental changes in 1H spectroscopy in human brain. In Lafeber HN, editor: Fetal and neonatal physiological measurements, New York, 1991, Elsevier, pp 33–41. 152. Hugg JW, Laxer KD, Matson GB, et al: Lateralization of human focal epilepsy by 31P magnetic resonance spectroscopic imaging. Neurology 42:2011–2018, 1992. 153. Huppi P, Posse S, et al: Magnetic resonance in preterm and term newborns: H spectroscopy in developing human brain. Pediatr Res 30:574–578, 1991. 154. Huppi PS, Fusch C, Boesch C, et al: Regional metabolic assessment of human brain during development by proton magnetic resonance spectroscopy in vivo and by high-performance liquid chromatography/ gas chromatography in autopsy tissue. Pediatr Res 37:145–150, 1995. 155. Husted CA, Goodin DS, Hugg JW, et al: Biochemical alterations in multiple sclerosis lesions and normal-appearing white matter detected by in vivo 31P and 1H spectroscopic imaging. Ann Neurol 36:157–165, 1994. 156. Hwang J-H, Egnaczyk GF, Ballard E, et al: Proton MR spectroscopic characteristics of pediatric pilocytic astrocytomas. AJNR Am J Neuroradiol 19:535–540, 1998.

157. Hwang J-H, Graham GD, Behar KL, et al: Short echo time proton magnetic resonance spectroscopic imaging of macromolecule and metabolite signal intensities in the human. Magn Reson Med 35:633– 639, 1996. 158. Jacobs MA, Horska A, van Zijl PC, et al: Quantitative proton MR spectroscopic imaging of normal human cerebellum and brain stem. Magn Reson Med 46:699–705, 2001. 159. Jansen JFA, Backes WH, Nicolay K, et al: 1H MR spectroscopy of the brain: Quantification of metabolites. Radiology 240(2):318–332, 2006. 160. Jarvik JG, Lenkinski RE, Saykin AJ, et al: Proton spectroscopy in asymptomatic HIV-infected adults: Initial results in a prospective cohort study. J Acquir Immune Defic Syndr Hum Retrovirol 13:247–253, 1996. 161. Jaysundar R, Bakshi A, Singh VP, et al: Diagnostic potential of proton MRS in differentiating between tumor recurrence and radiation necrosis? Paper presented at the Fifth Annual Scientific Meeting of the International Society for Magnetic Resonance in Medicine, 1997, Vancouver, BC, p 1142. 162. Jenkins BG, Koroshetz WJ, Beal MF, et al: Evidence for impairment of energy metabolism in vivo in Huntington’s disease using localized 1H NMR spectroscopy. Neurology 43:2689–2695, 1993. 163. Johnson B, Gay M, Zhang K, et al: The use of magnetic resonance spectroscopy in the subacute evaluation of athletes recovering from single and multiple traumatic brain injury. J Neurotrauma 29:2297– 2304, 2012. 164. Joy AM, Beaudry CE, Tran NL, et al: Migrating glioma cells activate the P13K pathway and display decreased susceptibility to apoptosis. J Cell Sci 116(21):4409–4417, 2003. 165. Julia-Sape M, Coronel I, Majos C, et al: Prospective diagnostic performance evaluation of single voxel 1H MRS for typing and grading of brain tumors. NMR Biomed 25(4):661–673, 2012. 166. Kadota T, Horinouchi T, Kuroda C: Development and aging of the cerebrum: Assessment with proton MR spectroscopy. AJNR Am J Neuroradiol 22:128–135, 2001. 167. Kaibara T, Tyson RL, Sutherland GR: Human cerebral neoplasms studied using MR spectroscopy: A review. Biochem Cell Biol 76:477– 486, 1998. 168. Kantarci K, Jack CR, Jr, Xu YC, et al: Regional metabolic patterns in mild cognitive impairment and Alzheimer’s disease: A 1H MRS study. Neurology 55:210–217, 2000. 169. Kegeles LS, Humaran TJ, Mann JJ: In vivo neurochemistry of the brain in schizophrenia as revealed by magnetic resonance spectroscopy. Biol Psychiatry 44:382–398, 1998. 170. Kelly PJ, Daumas-Duport C, Scheithauer BW, et al: Stereotactic histologic correlations of computed tomography- and magnetic resonance imaging-defined abnormalities in patients with glial neoplasms. Mayo Clin Proc 62:450–459, 1987. 171. Kim CK, Alavi JB, Alavi A, et al: New grading system of cerebral gliomas using positron emission tomography with F-18 fluorodeoxyglucose. J Neurooncol 10:85–91, 1991. 172. Kimura H, Fujii Y, Matsuda T, et al: Metabolic alterations in the neonate and infant brain during development: Evaluation with proton MR spectroscopy. Radiology 194:483–489, 1995. 173. Kimura S, Ohtuki N, Nezu A, et al: Clinical and radiologic improvements in mitochondrial encephalomyelopathy following sodium dichloroacetate therapy. Brain Dev 19:535–540, 1997. 174. King D, Baltuch GH: Magnetic resonance imaging and temporal lobe epilepsy. Acta Neurol Scand 98:217–223, 1998. 175. Kiraly MA, Kiraly SJ: Traumatic brain injury and delayed sequelae:   A review—Traumatic brain injury and mild traumatic brain injury (concussion) are precursors to later onset brain disorders, including early onset dementia. Scientificworldjournal 7:1768–1776, 2007. 176. Kornfeld M, Woodfin B, Papile L, et al: Neuropathology of ornithine carbamyl transferase deficiency. Acta Neuropathol 65:261–264, 1985. 177. Kreis R, Ernst T, Ross BD: Absolute quantitation of water and metabolites in the human brain: II. Metabolite concentrations. J Magn Reson 102:9–15, 1993.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy 178. Kreis R, Ernst T, Ross BD: Development of the human brain: In vivo quantification of metabolite and water content with proton magnetic spectroscopy. Magn Reson Med 30:424–437, 1993. 179. Kreis R, Farrow NA, Ross BD: Localized 1H NMR spectroscopy in patients with chronic hepatic encephalopathy: Analysis of changes in cerebral glutamine, choline and inositols. NMR Biomed 4:109–116, 1991. 180. Kreis R, Ross BD: Cerebral metabolic disturbances in patients with subacute and chronic diabetes mellitus: Detection with proton MR spectroscopy. Radiology 184:123–130, 1992. 181. Kreis R, Ross BD, Farrow NA, et al: Metabolic disorders of the brain in chronic hepatic encephalopathy detected with H-1 MR spectroscopy. Radiology 182:19–27, 1992. 182. Kruse B, Barker PB, van Zijl PC, et al: Multislice proton magnetic resonance spectroscopic imaging in X-linked adrenoleukodystrophy. Ann Neurol 36:595–608, 1994. 183. Kugel H, Heindel W, Ernestus RI, et al: Human brain tumors: Spectral patterns detected with localized H-1 MR spectroscopy. Radiology 183:701–709, 1992. 184. Kurihara N, Takahashi S, Furuta A, et al: MR imaging of multiple sclerosis simulating brain tumor. Clin Imaging 20:171–177, 1996. 185. Kuzniecky R, Elgavish GA, Hetherington HP, et al: In vivo 31P nuclear magnetic resonance spectroscopy of human temporal lobe epilepsy. Neurology 42:1586–1590, 1992. 186. Kuzniecky R, Hugg JW, Hetherington H, et al: Relative utility of 1H spectroscopic imaging and hippocampal volumetry in the lateralization of mesial temporal lobe epilepsy. Neurology 51:66–71, 1998. 187. Kwock L: Localized MR spectroscopy. Neuroimaging Clin N Am 8:713–731, 1998. 188. Kwock L: Proton magnetic resonance spectroscopy and spectroscopic imaging of primary brain tumors. In Pillai JJ, editor: Functional brain tumor imaging, New York, 2014, Springer Publishing, pp 143–167. 189. Lai PH, Weng HH, Chen CY, et al: In vivo differentiation of aerobic brain abscesses and necrotic GBM using proton MRSI. AJNR Am J Neuroradiol 29:1511–1518, 2008. 190. Lam WWM, Wang ZJ, Zhao H, et al: 1H MR spectroscopy of the basal ganglia in childhood: A semiquantitative analysis. Neuroradiology 40:315–323, 1998. 191. Law M, Cha S, Knopp E, et al: High grade gliomas and solitary metastases: Differentiation by using perfusion and proton spectroscopic MR imaging. Radiology 222:715–721, 2002. 192. Law M, Yang S, Wang H, et al: Glioma grading: sensitivity, specificity and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 24:1989–1998, 2003. 193. Laws ER, Parney IF, Huang W, et al: Survival following surgery and prognostic factors for recently diagnosed malignant glioma: Data from the Glioma Outcomes Project. J Neurosurg 99:467–473, 2003. 194. Lee JH, Arcinue E, Ross BD: Organic osmolytes in the brain of an infant with hypernatremia. N Engl J Med 331:439–442, 1994. 195. Lee SW, Benedick BA, Marsch LH, et al: Patterns of failure following high dose 3-D conformal radiotherapy for high grade astrocytomas: a quantitative dosimetric study. Int J Radiat Oncol Biol Phys 43(1):79–88, 1999. 196. Leite CC, Lucato LT, Sato JR, et al: Multivoxel proton magnetic resonance spectroscopy in malformations of cortical development. AJNR Am J Neuroradiol 28:1071–1075, 2007. 197. Leth H, Toft PB, Pryds O, et al: Brain lactate in preterm and growthretarded neonates. Acta Paediatr 84:495–499, 1995. 198. Li LM, Cendes F, Bastos AC, et al: Neuronal metabolic dysfunction in patients with cortical developmental malformations: A proton magnetic resonance spectroscopic imaging study. Neurology 50:755–759, 1998. 199. Li X, Jin H, Chang S, et al: Identification of MRI and 1H MRSI parameters that may predict survival for patients with malignant gliomas. NMR Biomed 17:10–20, 2004. 200. Li X, Lu Y, Pirzkall A, et al: Analysis of the spatial characteristics of metabolic abnormalities in newly diagnosed glioma patients. J Magn Reson Imaging 16:229–237, 2002.

531

201. Li BS, Wang H, Gonen O: Metabolite ratios to assumed creatine level may confound the quantification of proton brain MR spectroscopy. Magn Reson Imaging 21:923–928, 2003. 202. Lichter T, Dohrman GJ, Gets GS: Respiratory deficiency and increased glycolysis in benign human tumors. Surg Forum 35:486–488, 1994. 203. Lin DD, Crawford TO, Barker PB: Proton MR spectroscopy in the diagnostic evaluation of suspected mitochondrial disease. AJNR Am J Neuroradiol 24:33–41, 2003. 204. Lin A, Ross BD, Harris K, et al: Efficacy of proton magnetic resonance spectroscopy in neurological diagnosis and neurotherapeutic decision making. NeuroRx 2:197–214, 2005. 205. Lowry OH, Berger SJ, Chi M-Y, et al: Diversity of metabolic patterns in human brain tumors: I. High energy phosphate compounds and basic composition. J Neurochem 29:959–977, 1977. 206. Lu D, Pavlakis SG, Frank Y, et al: Proton MR spectroscopy of the basal ganglia in healthy children and children with AIDS. Radiology 199:423–428, 1996. 207. Lukas L, Devos A, Suykens JA, et al: Brain tumor classification based on long echo proton MRS signals. Artif Intell Med 31:73–89, 2004. 208. Luyten PR, Marien AJH, Heindel W, et al: Metabolic imaging of patients with intracranial tumors: H-1 MR spectroscopic imaging and PET. Radiology 176:791–799, 1990. 209. Mader I, Roser W, Kappos L, et al: Serial proton MR spectroscopy of contrast-enhancing multiple sclerosis plaques: Absolute metabolic values over 2 years during a clinical pharmacological study. AJNR Am J Neuroradiol 21:1220–1227, 2000. 210. Marks HG, Caro PA, Wang Z, et al: Use of computed tomography, magnetic resonance imaging, and localized 1H magnetic resonance spectroscopy in Canavan’s disease: A case report. Ann Neurol 30:106– 110, 1991. 211. Matalon R, Michals K, Sevesta D, et al: Asparto-acylase deficiency and N-acetylaspartic aciduria in patients with Canavan’s disease. Am J Med Genet 20:463–471, 1988. 212. Matsuma A, Isobe T, Anno I, et al: Correlation between choline and MIB-1 index in human gliomas. J Clin Neurosci 12(4):416–420, 2005. 213. Matthews PM, Andermann F, Silver K, et al: Proton MR spectroscopic characterization of differences in regional brain metabolic abnormalities in mitochondrial encephalomyopathies. Neurology 43:2484–2490, 1993. 214. Matthews PM, De Stefano N, Narayanan S, et al: Putting magnetic resonance spectroscopy studies in context: Axonal damage and disability in multiple sclerosis. Semin Neurol 18:327–336, 1998. 215. Matthews PM, Francis G, Antel J, et al: Proton magnetic resonance spectroscopy for metabolic characterization of plaques in multiple sclerosis. Neurology 41:1251–1256, 1991. correction in Neurology 41:1828, 1991. 216. Marcus KJ, Astakas IG, Zurakowski D, et al: Predicting survival of children with CNS tumors using proton magnetic resonance spectroscopy. Int J Oncol 30:651–670, 2007. 217. Martin AJ, Liu H, Hall WA, et al: Preliminary assessment of turbo spectroscopic imaging for targeting in brain biopsy. AJNR Am J Neuroradiol 22:959–968, 2001. 218. Marino S, Zei E, Battaglini M, et al: Acute metabolic brain changes following traumatic brain injury and their relevance to clinical severity and outcome. J Neurol Neurosurg Psychiatry 78:501–507, 2007. 219. Maruta J, Suh M, Niogi SN, et al: Visual tracking synchronization as a metric for concussion screening. J Head Trauma Rehabil 25:293–305, 2010. 220. Maudsley AA, Twieg DB, Sappey-Marinier D, et al: Spin echo 31P spectroscopic imaging in the human brain. Magn Reson Med 14:415– 422, 1990. 221. McEllin B, Camacho CV, Mukherjee B, et al: PTEN loss compromises homologous recombination repair in astrocytes: Implications for glioblastoma therapy with temozolamide or poly(ADP-ribose) polymerase inhibitors. Cancer Res 70(13):5457–5464, 2010. 222. McKinstry RC, Miller JH, Snyder AZ, et al: A prospective, longitudinal diffusion tensor imaging study of brain injury in newborns. Neurology 59:824–833, 2002.

http://radiologyebook.com

532

PART II  CT and MR Imaging of the Whole Body

223. McKnight TR, von dem Bussche MH, Vigneron DB, et al: Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence.   J Neurosurg 97:794–802, 2002. 224. McLendon RE, Halpern EC: Is the long-term survival of patients with GBM overstated? Cancer 98:1745–1748, 2003. 225. Menon DK, Baudouin CJ, Tomlinson D, et al: Proton MR spectroscopy and imaging of the brain in AIDS: Evidence of neuronal loss in regions that appear normal with imaging. J Comput Assist Tomogr 14:882–885, 1990. 226. Meyerand ME, Pipas JM, Mamourian A, et al: Classification of biopsy-confirmed brain tumors using single-voxel MR spectroscopy. AJNR Am J Neuroradiol 20:117–123, 1999. 227. Meyerhoff DJ, Bloomer C, Cardenas V, et al: Elevated subcortical choline metabolites in cognitively and clinically asymptomatic HIV+ patients. Neurology 52:995–1003, 1999. 228. Michaelis T, Helms KD, Merboldt W, et al: First observation of scyllo-inositol in proton NMR spectra of human brain in vitro and in vivo. Paper presented at Annual Meeting of the Society of Magnetic Resonance in Medicine, Berlin, 1992. 229. Michaelis T, Merboldt KD, Bruhn H, et al: Absolute concentrations of metabolites in the adult human brain in vivo: Quantification of localized proton MR spectra. Radiology 187:219–227, 1993. 230. Michaelis T, Videen J, Linsey MS, et al: Cerebral abnormalities in end stage renal disease. Paper presented at the Second Annual Meeting of the Society of Magnetic Resonance in Medicine, San Francisco, 1994,   p 306. 231. Miller B: A review of chemical issues in 1H NMR spectroscopy: N-acetyl-aspartate, creatine and choline. NMR Biomed 4:47–52, 1991. 232. Miller BL, Moats R, Shonk T, et al: Alzheimer disease: Depiction of increased cerebral myoinositol with proton MR spectroscopy. Radiology 187:433–437, 1993. 233. Mishra AM, Gupta RK, Jaggi RS, et al: Role of diffusion-weighted imaging and in vivo proton magnetic resonance spectroscopy in the differential diagnosis of ring-enhancing intracranial cystic mass lesions. J Comput Assist Tomogr 28:540–547, 2004. 234. Miyake M, Kakimoto Y: Developmental changes of N-acetyl-l-aspartic acid, N-acetyl-l-aspartylglutamic acid, and beta-citryl-l-glutamic acid in different brain regions and spinal cords of rat and guinea pig.   J Neurochem 37:1064–1067, 1981. 235. Moats RA, Ernst T, Kreis R, et al: Well-localized standardized, quantitized short TE 1H MRS permits new diagnostic accuracy in patients with Alzheimer disease. TAMU Newsletter 413:13–14, 1993. 236. Moats RA, Ernst T, Shonk TK, et al: Abnormal cerebral metabolite concentrations in patients with probable Alzheimer disease. Magn Reson Med 32:110–115, 1994. 237. Moller-Hartmann W, Herminghaus S, Krings T, et al: Clinical application of proton magnetic resonance spectroscopy in the diagnosis of intracranial mass lesions. Neuroradiology 44:371–381, 2002. 238. Monsein LH, Mathews VP, Barker PB, et al: Irreversible regional ischemia: Serial MR imaging and proton MR spectroscopy in a non-human primate model. AJNR Am J Neuroradiol 14:963–970, 1993. 239. Moolenaar SH, van der Knaap MS, Engelke UF, et al: In vivo and in vitro NMR spectroscopy reveal a putative novel inborn error involving polyol metabolism. NMR Biomed 14:167–176, 2001. 240. Moore GJ: Proton magnetic resonance spectroscopy in pediatric neuroradiology. Pediatr Radiol 28:805–814, 1998. 241. Morantz RA: The management of the patient with low-grade cerebral astrocytomas. In Morantz RA, Walsh JW, editors: Brain tumors: A comprehensive text, New York, 1994, Marcel Dekker, pp 387–415. 242. Moseley ME, Cohen Y, Mintorovitch J, et al: Early detection of regional ischemia in cats: Comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med 14:330–346, 1990. 243. Mukherji SK, editor: Brain proton magnetic resonance spectroscopy. Neuroimaging Clin N Am 23(3):359–555, 2013. 244. Nagatomo Y, Wick M, Prielmeier F, et al: Dynamic monitoring of cerebral metabolites during and after transient global ischemia in rats

245. 246.

247.

248.

249. 250.

251.

252. 253. 254.

255.

256. 257.

258.

259.

260. 261.

262.

263.

264.

265.

266.

by quantitative proton NMR spectroscopy in vivo. NMR Biomed 8:265–270, 1995. Najm IM, Wang Y, Shedid D, et al: MRS metabolic markers of seizures and seizure-induced neuronal damage. Epilepsia 39:244–250, 1998. Narayana PA, Doyle TJ, Lai D, et al: Serial proton magnetic resonance spectroscopic imaging, contrast-enhanced magnetic resonance imaging, and quantitative lesion volumetry in multiple sclerosis. Ann Neurol 43:56–71, 1998. Negendank WG, Sauter R, Brown TR, et al: Proton magnetic resonance spectroscopy in patients with glial tumors: A multicenter study.   J Neurosurg 84:449–458, 1996. Nelson SJ: MR Spectroscopy in adults: Current and future applications (syllabus, Advanced Imaging Symposium: Preparing the neuroradiologist for the new millennium, pp 37-43). Presented at the American Society of Neuroradiology Annual Meeting, Atlanta, April 2000. Nelson SJ: Minireview: Multivoxel magnetic resonance spectroscopy of brain tumors. Mol Cancer Ther 2:497–507, 2003. Ng TC, Comair YG, Xue M, et al: Temporal lobe epilepsy: Presurgical localisation with proton chemical shift imaging. Radiology 193:465–472, 1994. Nicoli F, Lefur Y, Denis B, et al: Metabolic counterpart of decreased apparent diffusion coefficient during hyperacute ischemic stroke: A brain proton magnetic resonance spectroscopic imaging study. Stroke 34:e82–e87, 2003. Ott D, Hennig J, Ernst T: Human brain tumors: Assessment with in vivo proton MR spectroscopy. Radiology 186:745–752, 1993. Oz G, Alger JR, Barker PB, et al: Clinical proton MR spectroscopy in central nervous system disorders. Radiology 270(3):658–679, 2014. Oz G, Tkac I, Charnas LR, et al: Assessment of adrenoleukodystrophy lesions by high field MRS in non-sedated pediatric patients. Neurology 64:434–441, 2005. Padhani AR, Liu G, Mu-Koh D, et al: Diffusion-weighted magnetic resonance imaging as a cancer biomarker: Consensus and recommendations. Neoplasia 11(2):102–125, 2009. Paley RJ, Persing JA, Doctor A, et al: Multiple sclerosis and brain tumor. J Emerg Med 7:241–244, 1989. Panigrahy A, Bluml S: Advances in magnetic resonance neuroimaging techniques in the evaluation of neonatal encephalopathy. Top Magn Reson Imaging 18:3–29, 2007. Parsons MW, Barber PA, Desmond PM, et al: Acute hyperglycemia adversely affects stroke outcome: A magnetic resonance imaging and spectroscopy study. Ann Neurol 52:20–28, 2002. Parsons MW, Li T, Barber PA, et al: Combined (1)H MR spectroscopy and diffusion-weighted MRI improves the prediction of stroke outcome. Neurology 55:498–505, 2000. Patchell RA: Brain metastases. Handb Clin Neurol 25:135–149, 1997. Patronas NJ, Di Chiro G, Kufta C, et al: Prediction of survival in glioma patients by means of positron emission tomography. J Neurosurg 62:816–822, 1985. Peeling J, Wong D, Sutherland GR: Nuclear magnetic resonance study of regional metabolism after forebrain ischemia in rats. Stroke 20:633–640, 1989. Pendlebury ST, Blamire AM, Lee MA, et al: Axonal injury in the internal capsule correlates with motor impairment after stroke. Stroke 30:956–962, 1999. Penrice J, Cady EB, Lorek A, et al: Proton magnetic resonance spectroscopy of the brain in normal preterm and term infants and early changes after perinatal hypoxia-ischemia. Pediatr Res 40:6–14, 1996. Penrice J, Lorek A, Cady EB, et al: Proton magnetic resonance spectroscopy of the brain during acute hypoxia-ischemia and delayed cerebral energy failure in the newborn piglet. Pediatr Res 41:795–802, 1997. Pereira AC, Saunders DE, Doyle VL, et al: Measurement of initial N-acetyl aspartate concentration by magnetic resonance spectroscopy and initial infarct volume by MRI predicts outcome of patients with middle cerebral artery territory infarction. Stroke 30:1577–1582, 1999.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy 267. Phillips HS, Kharbanda S, Chen R, et al: Molecular subclasses of high grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell 9(3):157–173, 2006. 268. Pirzlkall A, McKnight TR, Graves EE, et al: MR-spectroscopy guided target delineation for high grade gliomas. Int J Radiat Oncol Biol Phys 50(4):915–928, 2001. 269. Poptani H, Kaartinen J, Gupta RK, et al: Diagnostic assessment of brain tumours and non-neoplastic brain disorders in vivo using proton nuclear magnetic resonance spectroscopy and artificial neural networks. J Cancer Res Clin Oncol 125:343–349, 1999. 270. Posner JB: Brain metastases: 1995. A brief review. J Neurooncol 27:287–293, 1996. 271. Posse S, Otazo R, Dager SR: MR spectroscopic imaging: Principles and recent advances. J Magn Reson Imaging 37:1301–1325, 2013. 272. Posse S, Schuknecht B, Smith ME, et al: Short echo time proton MR spectroscopic imaging. J Comput Assist Tomogr 17:1–14, 1994. 273. Pouwels PJ, Brockmann K, Kruse B, et al: Regional age dependence of human brain metabolites from infancy to adulthood as detected by quantitative localized proton MRS. Pediatr Res 46:474–485, 1999. 274. Pouwels PJW, Frahm J: Regional metabolite concentrations in human brain determined by quantitative localized proton MRS. Magn Reson Med 39:53–60, 1998. 275. Price SJ, Burnet NG, Donovan T, et al: Diffusion tensor imaging of brain tumors at 3T: A potential tool for assessing white matter tract invasion. Clin Radiol 58:455–462, 2003. 276. Price SJ, Green HAL, Dean AF, et al: Correlation of MR relative cerebral blood volume measurements with cellular density and proliferation in high grade gliomas: An image guided biopsy study. AJNR Am J Neuroradiol 32:501–506, 2011. 277. Price SJ, Jena R, Burnet NG, et al: Improved delineation of glioma margins and regions of infiltration with the use of diffusion tensor imaging: An image guided biopsy study. AJNR Am J Neuroradiol 27:1969–1974, 2006. 278. Pruel MC, Caramanos Z, Collins DL, et al: Linear discriminate analysis based on proton MR spectroscopic imaging of human brain tumors improves pre-operative diagnosis. Nat Med 2(3):323–325, 1996. 279. Pronin IN, Holodny AI, Petrakin AV: MRI of high grade glial tumors: Correlation between the degree of contrast enhancement and volume of surrounding edema. Neuroradiology 39:348–350, 1997. 280. Pu Y, Li QF, Zeng CM, et al: Increased detectability of alpha brain glutamate/glutamine in neonatal hypoxic-ischemic encephalopathy. AJNR Am J Neuroradiol 21:203–212, 2000. 281. Rakic P, Bourgeois JP, Eckenhoff MF, et al: Concurrent overproduction of synapses in diverse regions of the primate cerebral cortex. Science 232:232–235, 1986. 282. Rand SD, Prost R, Haughton V, et al: Accuracy of single-voxel proton MR spectroscopy in distinguishing neoplastic from nonneoplastic brain lesions. AJNR Am J Neuroradiol 18:1695–1704, 1997. 283. Rasey JS, Koh W-J, Evans ML, et al: quantifying regional hypoxia in human tumors with positron emission tomography of [18F] Fluoromisonidazole: A pretherapy study of 37 patients. Int J Radiat Oncol Biol Phys 36(2):417–428, 1996. 284. Raza SM, Lang FF, Aggarwal BB: Necrosis and glioblastoma: A friend or a foe? A review and hypothesis. Neurosurgery 51(1):2–13, 2002. 285. Remy C, Grand S, Lai ES, et al: 1H MRS of human brain abscesses in vivo and in vitro. Magn Reson Med 34:508–514, 1995. 286. Ricci PE: Proton MR spectroscopy in ischemic stroke and other vascular disorders. Neuroimaging Clin N Am 8:881–900, 1998. 287. Richards TL, Dager SR, Panagiotides HS, et al: Functional MR spectroscopy during language activation: A preliminary study using proton echo-planar spectroscopic imaging (PEPSI). Int J Neuroradiol 3:490, 1997. 288. Richards TL, Dager SR, Posse S: Functional MR spectroscopy of the brain. Neuroimaging Clin N Am 8:823–834, 1998. 289. Rooney WD, Goodkin DE, Schuff N, et al: H MRS of normal   appearing white matter in multiple sclerosis. Mult Scler 3:231–237, 1997.

533

290. Ross BD: Biochemical considerations in 1H spectroscopy. Glutamate and glutamine: Myoinositol and related metabolites. NMR Biomed 4:59–63, 1991. 291. Ross BD, Bluml S, Cowan R, et al: In vivo MR spectroscopy of human dementia. Neuroimaging Clin N Am 8:809–822, 1998. 292. Ross B, Danielsen E, Bluml S: Proton magnetic resonance spectroscopy: The new gold standard for diagnosis of clinical and subclinical hepatic encephalopathy? Dig Dis 14(Suppl 1):30–39, 1996. 293. Ross BD, Jacobson S, Villamil F, et al: Subclinical hepatic encephalopathy: Proton MR spectroscopic abnormalities. Radiology 193:457–463, 1994. 294. Ross B, Kreis R, Ernst T: Clinical tool for the 90s: Magnetic resonance spectroscopy and metabolite imaging. Eur J Radiol 14:128–140, 1992. 295. Rudkin TM, Arnold DL: Proton magnetic resonance spectroscopy for the diagnosis and management of cerebral disorders. Arch Neurol 56:919–926, 1999. 296. Ruggieri PM: Practical MR spectroscopy in pediatric neuroradiology (syllabus, Advanced Imaging Symposium: Preparing the neuroradiologist for the new millennium, pp 45-53). Paper presented at the Annual Meeting of the American Society of Neuroradiology, Atlanta, April 2000. 297. Rutgers DR, Fillard P, Paradot G, et al: Diffusion tensor imaging characteristics of the corpus callosum in mild, moderate, and severe traumatic brain injury. AJNR Am J Neuroradiol 29:1730–1735, 2008. 298. Sadeghi N, Salmon I, Decaestecker C, et al: Stereotactic comparison among cerebral blood volume, methionine uptake, and histopathology in brain glioma. AJNR Am J Neuroradiol 28:455–461, 2007. 299. Saindane AM, Cha S, Law M, et al: Proton MR spectroscopy of tumefactive demyelinating lesions. AJNR Am J Neuroradiol 23:1378– 1386, 2002. 300. Salvan AM, Confort-Gouny S, Chabrol B, et al: Brain metabolic impairment in non-cerebral and cerebral forms of X-linked adrenoleukodystrophy by proton MRS: Identification of metabolic patterns by discriminant analysis. Magn Reson Med 41:1119–1126, 1999. 301. Salvan AM, Vion-Dury J, Confort-Gouny S, et al: Brain proton magnetic resonance spectroscopy in HIV-related encephalopathy: Identification of evolving metabolic patterns in relation to dementia and therapy. AIDS Res Hum Retroviruses 13:1055–1066, 1997. 302. Sappey-Marinier D, Calabrese G, Hetherington HP, et al: Proton magnetic resonance spectroscopy of the human brain: Applications to normal white matter, chronic infarction, and MRI white matter signal hyperintensities. Magn Reson Med 26:313–327, 1992. 303. Saraswathy SS, Crawford FW, Lamborn KR: Evaluation of MR markers that predict survival in patients with newly diagnosed GBM prior to adjuvant therapy. J Neurooncol 91:69–81, 2009. 304. Saunders DE, Howe FA, Van Den Boogaart A: Continuing ischemic damage after acute middle cerebral artery infarction in humans demonstrated by short-echo proton spectroscopy. Stroke 26:1007–1013, 1995. 305. Schuff N, Amend D, Ezekiel BA, et al: Changes of hippocampal N-acetyl aspartate and volume in Alzheimer’s disease: A proton MR spectroscopic imaging and MRI study. Neurology 49:1513–1521, 1997. 306. Scriver C, Kaufman S, Woo S: The hyperphenylalaninemias. In Scriver CR, Beaudet AL, Sly WS, et al, editors: The metabolic basis of inherited disease, New York, 1989, McGraw-Hill, pp 495–546. 307. Seaquist ER, Tkac I, Damberg G, et al: Brain glucose concentrations in poorly controlled diabetes mellitus as measured by high-field magnetic resonance spectroscopy. Metabolism 54:1008–1013, 2005. 308. Sharma R, Venkatasubramanian PN, Barany M, et al: Proton MRS of the brain in schizophrenia and affective disorders. Schizophr Res 8:43–49, 1992. 309. Shevell MI, Ashwal S, Novotny E: Proton magnetic resonance spectroscopy: Clinical applications in children with nervous system diseases. Semin Pediatr Neurol 6:68–77, 1999. 310. Shimizu H, Kumabe T, Shirane R, et al: Correlation between choline level measured by proton MR spectroscopy and Ki-67 labeling index in gliomas. AJNR Am J Neuroradiol 21:659–665, 2000.

http://radiologyebook.com

534

PART II  CT and MR Imaging of the Whole Body

311. Shimizu H, Kumabe T, Tominaga T, et al: Noninvasive evaluation of malignancy of brain tumors with proton MR spectroscopy. AJNR Am J Neuroradiol 17:737–747, 1996. 312. Shiroishi MS, Jones JGA, Muradyan N, et al: MR perfusion imaging: ASL, T2*-weighted DSC, and T1-weighted DCE methods. In Pillai JJ, editor: Functional brain tumor imaging, New York, 2014, Springer Publishing, pp 3–26. 313. Shnall MD, Yoshizaki K, Chance B, et al: Triple nuclear NMR studies of cerebral metabolism during generalized seizure. Magn Reson Med 6:15–23, 1988. 314. Shonk TK, Moats RA, Gifford P, et al: Probable Alzheimer disease: Diagnosis with proton MR spectroscopy. Radiology 195:65–72, 1995. 315. Shu SK, Ashwal S, Holshouser BA, et al: Prognostic value of 1H MRS in perinatal CNS insults. Pediatr Neurol 17:309–318, 1997. 316. Siemann DW: Modification of chemotherapy by nitroimidazoles. Int J Radiat Oncol Biol Phys 10:1585–1594, 1984. 317. Siesjo BK, von Hanwehr R, Nergelius G, et al: Extra- and intracellular pH in the brain during seizures and in the recovery period following arrest of seizure activity. J Cereb Blood Flow Metab 5:47–57, 1985. 318. Sijens PE, van Dijk P, Oudkerk M: Correlation between choline level and Gd-DTPA enhancement in patients with brain metastases of mammary carcinomas. Magn Reson Med 32:549–555, 1994. 319. Simone IL, Federico F, Tortorella C, et al: Metabolic changes in neuronal migration disorders: Evaluation by combined MRI and proton MR spectroscopy. Epilepsia 40:872–879, 1999. 320. Singhal AB, Ratai E, Benner T, et al: Magnetic resonance spectroscopy study of oxygen therapy in ischemic stroke. Stroke 38:2851–2854, 2007. 321. Smits M, Houston GC, Dippel DWJ, et al: Microstructural brain injury in post-concussion syndrome after minor head injury. Neuroradiology 53:553–563, 2010. 322. Srinivasan R, Sailasuta N, Hurd R, et al: Evidence of elevated glutamate in multiple sclerosis using magnetic resonance spectroscopy at 3 T. Brain 128:1016–1025, 2005. 323. Stadlbauer A, Gruber S, Nimsky C, et al: Preoperative grading of gliomas by using metabolite quantification with high-spatial-resolution proton MR spectroscopic imaging. Radiology 238(3):958–969, 2006. 324. Stall B, Zach L, Ning H, et al: comparison of T2 and FLAIR imaging for target delineation in high grade gliomas. Radiat Oncol 5:5, 2010. . 325. Stockler S, Holzbach U, Hanefeld F, et al: Creatine deficiency in the brain: A new, treatable inborn error of metabolism. Pediatr Res 36:409–413, 1994. 326. Sugahara T, Korogi Y, Kochi M, et al: Correlation of MR imaging determined cerebral blood volume maps with histologic and angiographic determination of vascularity of gliomas. AJR Am J Roentgenol 171:1479–1486, 1998. 327. Sutton LN, Wang Z, Gusnard D, et al: Proton magnetic resonance spectroscopy of pediatric brain tumors. Neurosurgery 31:195–202, 1992. 328. Sutton LN, Wehrli SL, Gennareli L, et al: High-resolution 1H-magnetic resonance spectroscopy of pediatric posterior fossa tumors in vitro.   J Neurosurg 81:443–448, 1994. 329. Swindells S, McConnell JR, McComb RD, et al: Utility of cerebral proton magnetic resonance spectroscopy in differential diagnosis of HIV-related dementia. J Neurovirol 1:268–274, 1995. 330. Tallan HH: Studies on the distribution of N-acetyl-aspartic acid in brain. J Biol Chem 224:41–45, 1956. 331. Tallan HH, Moore S, Stein WH: N-acetyl-l-aspartic acid in brain. J Biol Chem 219:257–264, 1956. 332. Tamiya T, Kinoshita K, Ono Y, et al: Proton magnetic resonance spectroscopy reflects cellular proliferative activity in astrocytomas. Neuroradiology 42:333–338, 2000. 333. Taylor JS, Langston JW, Reddick WE, et al: Clinical value of proton magnetic resonance spectroscopy for differentiating recurrent or residual brain tumor from delayed cerebral necrosis. Int J Radiat Oncol Biol Phys 36:1251–1261, 1996. 334. Tedeschi G, Bertolino A, Lundbom N, et al: Cortical and subcortical chemical pathology in Alzheimer’s disease as assessed by multislice

335.

336.

337.

338.

339.

340.

341.

342.

343.

344.

345.

346.

347.

348.

349.

350.

351.

352.

353.

354.

proton magnetic resonance spectroscopic imaging. Neurology 47:696–704, 1996. Tedeschi G, Bertolino A, Righini A, et al: Brain regional distribution pattern of metabolite signal intensities in young adults by proton magnetic resonance spectroscopic imaging. Neurology 45:1384–1391, 1995. Tedeschi G, Lundbom N, Raman R, et al: Increase of choline signal coincides with malignant degeneration of cerebral gliomas: A serial proton magnetic resonance spectroscopic imaging study. J Neurosurg 87:516–524, 1997. Tedeschi G, Schiffmann R, Barton NW, et al: Proton magnetic resonance spectroscopic imaging in childhood ataxia with diffuse white matter hypomyelination. Neurology 45:1526–1532, 1995. Tien RD, Lai PH, Smith JS, et al: Single-voxel proton brain spectroscopy exam (PROBE/SV) in patients with primary brain tumors. AJR Am J Roentgenol 167:201–209, 1996. Tochon-Danguy HJ, Sachinidis JL, Chan F, et al: Imaging and quantitation of the hypoxic cell fraction in an animal model of intracerebral high grade glioma using [18F]fluoromisonidazole (FMISO). Nucl Med Biol 29:191–197, 2002. Toft PB, Leth H, Lou HC, et al: Metabolite concentrations in the developing brain estimated with proton MR spectroscopy. J Magn Reson Imaging 4:674–680, 1994. Tourbah A, Stievenart JL, Gout O, et al: Localized proton magnetic resonance spectroscopy in relapsing remitting versus secondary progressive multiple sclerosis. Neurology 53:1091–1097, 1999. Tracey I, Lane J, Chang I, et al: 1H magnetic resonance spectroscopy reveals neuronal injury in a simian immune deficiency virus macaque model. J Acquir Immune Defic Syndr Hum Retrovirol 15:21–27, 1997. Tzika AA, Ball WS, Vigneron DB, et al: Childhood adrenoleukodystrophy: Assessment with proton MR spectroscopy. Radiology 189:467–480, 1993. Tzika AA, Vajapeyam S, Barnes PD: Multivoxel proton MR spectroscopy and hemodynamic MR imaging of childhood brain tumors: Preliminary observations. AJNR Am J Neuroradiol 18:203–218, 1997. Tzika AA, Vigneron DB, Dunn RS, et al: Intracranial tumors in children: Small single-voxel proton MR spectroscopy using short- and long-echo sequences. Neuroradiology 38:254–263, 1996. Urenjak J, Williams SR, Gadian DG, et al: Specific expression of N-acetylaspartate in neurons, oligodendrocyte-type 2 astrocyte progenitors, and immature oligodendrocytes in vitro. J Neurochem 59:55–61, 1992. Vagnozzi R, Signoretti S, Cristofori L, et al: Assessment of metabolic brain damage and recovery following mild traumatic brain injury: A multicenter proton magnetic resonance spectroscopic study in concussed patients. Brain 133:3232–3242, 2010. van der Knaap MS, Barth P, Gabreels F, et al: A new leukoencephalopathy with vanishing white matter. Neurology 48:845– 855, 1997. van der Knaap MS, Ross B, Valk J: Uses of MR in inborn errors of metabolism. In Kucharczyk J, Mosely M, Barkovich AJ, editors: Magnetic resonance neuroimaging, Boca Raton, FL, 1994, CRC Press, pp 245–318. van der Knaap MS, van der Grond J, van Rijen PC, et al: Agedependent changes in localized proton and phosphorus MR spectroscopy of the brain. Radiology 176:509–515, 1990. van der Knaap MS, van der Grond J, Luyten PR, et al: H-1 and P-31 magnetic resonance spectroscopy of the brain in degenerative cerebral disorders. Ann Neurol 31:202–211, 1993. Videen JS, Michaelis T, Pinto P, et al: Human cerebral osmolytes during chronic hyponatremia: A proton magnetic resonance spectroscopy study. J Clin Invest 95:788–793, 1995. Vigneron DB, Barkovich AJ, Noworolski SM, et al: Three-dimensional proton MR spectroscopic imaging of premature and term neonates. AJNR Am J Neuroradiol 22:1424–1433, 2001. Vion-Dury J, Meyerhoff DJ, Cozzone PJ, et al: What might be the impact on neurology of the analysis of brain metabolism by in vivo magnetic resonance spectroscopy? J Neurol 241:354–371, 1994.

http://radiologyebook.com

CHAPTER 17  Brain Proton Magnetic Resonance Spectroscopy 355. Wang ZJ, Berry GT, Dreha SF, et al: In vivo proton brain MRS of galactosemia. Paper presented at the Sixth Annual Meeting of the International Society of Magnetic Resonance in Medicine, 1998, Sydney, Australia, p 538. 356. Wang ZJ, Berry GT, Dreha SF, et al: Proton magnetic resonance spectroscopy of brain metabolites in galactosemia. Ann Neurol 50:266–269, 2001. 357. Wang Z, Sutton LN, Cnaan A, et al: Proton MR spectroscopy of pediatric cerebellar tumors. AJNR Am J Neuroradiol 16:1821–1833, 1995. 358. Wang Z, Zimmerman RA, Sauter R: Proton MR spectroscopy of the brain: Clinically useful information obtained in assessing CNS diseases in children. AJR Am J Roentgenol 167:191–199, 1996. 359. Wang ZJ, Zimmerman RA: Proton MR spectroscopy of pediatric   brain metabolic disorders. Neuroimaging Clin N Am 8:781–807, 1998. 360. Wardlaw JM, Marshall I, Wild J, et al: Studies of acute ischemic stroke with proton magnetic resonance spectroscopy: Relation between time from onset, neurological deficit, metabolite abnormalities in the infarct, blood flow, and clinical outcome. Stroke 29:1618–1624, 1998. 361. Warren KE, Frank JA, Black JL, et al: Proton MRSI in children with recurrent brain tumors. J Clin Oncol 18:1020–1026, 2007. 362. Wells W, Pittman T, Wells H, et al: The isolation and identification of galactitol from the brains of galactosemia patients. J Biol Chem 240:1002–1004, 1965. 363. Wheaton P, Mathias JL, Vink R: Impact of pharmacological treatments on cognitive and behavioral outcome in the postacute stages of adult traumatic brain injury: A meta-analysis. J Clin Psychopharmacol 31:745–757, 2011. 364. Whitmore GF, Varghese AJ: The biological properties of reduced nitroheterocyclics and possible biological mechanisms. Biochem Pharmacol 35(1):97–103, 1986.

535

365. Wilkinson ID, Lunn S, Miszkiel KA, et al: Proton MRS and quantitative MRI assessment of the short term neurological response to antiretroviral therapy in AIDS. J Neurol Neurosurg Psychiatry 63:477–482, 1997. 366. Witt ST, Lovejoy DW, Pearlson GD, et al: Decreased prefrontal cortex activity in mild traumatic brain injury during performance of an auditory oddball task. Brain Imaging Behav 4:232–247, 2010. 367. Wittsack H, Kugel H, Roth B, et al: Quantitative measurements with localized 1H MR spectroscopy in children with Canavan’s disease. J Magn Reson Imaging 6:889–893, 1996. 368. Wright AJ, Fellows G, Byrnes TJ, et al: Pattern recognition of MRSI data shows region of glioma growth that agrees with DTI markers of brain tumor infiltration. Magn Reson Med 62:1646–1651, 2009. 369. Yaksh TL, Anderson RE: In vivo studies on intracellular pH, focal   flow and vessel diameter in the cat cerebral cortex: Effects of altered CO2 and electrical stimulation. J Cereb Blood Flow Metab 7:332–341, 1987. 370. Yu X, Liu Z, Yian Z, et al: Stereotactic biopsy for space occupying lesions: Clinical analysis of 550 cases. Stereotact Funct Neurosurg 75:103–108, 2000. 371. Zagzag D, Zhong H, Scalzitti JM, et al: Expression of hypoxia-inducible factor 1-alpha in brain tumors: association with angiogenesis, invasion and progression. Cancer 88:2606–2618, 2000. 372. Zeng Q-S, Li C-F, Zhang K, et al: Multivoxel 3D MR spectroscopy in the distinction of recurrent glioma from radiation injury. J Neurooncol 84:63–69, 2007. 373. Zhang K, Li C, Liu Y: Evaluation of invasiveness of astrocytoma using 1 H-magnetic resonance spectroscopy: correlation with expression of matrix metalloproteinase-2. Neuroradiology 49(11):913–919, 2007. 374. Zimmerman RA, Wang Z: Proton spectroscopy of the pediatric brain. Riv Neuroradiol 5(Suppl 1):5–8, 1992.

http://radiologyebook.com

18  Meningeal Processes Melanie B. Fukui, Sheilah Curran-Melendez, and Carolyn Cidis Meltzer

ANATOMY AND EMBRYOLOGY Meninges The meninges, which form the coverings of the brain and spinal cord, develop from the meninx primitiva.71 The neural tube is surrounded by this dense cellular layer shortly after the neural tube closes.126 As early as 32 days and as late as 44 days of gestation, the meninx primitiva begins to cavitate to form the cerebral cisterns by gradually decreasing its cellular component and enlarging its intercellular space.126 The periphery of the meninx primitiva, however, develops more dense cellularity to become the primitive dura mater.126 The earliest subarachnoid space (SAS) is ventral to the brainstem. As this space expands, the prepontomedullary cisterns and anterior spinal SAS are formed.126,158 At approximately 41 days, the space is extended to create perimesencephalic and dorsal mesencephalic cisterns.125,126,158 The primitive meninx is composed of totipotential mesenchymal cells of neural crest origin.71 The remnants of incomplete differentiation of these pluripotential cells may be seen as deposits of fat, or lipomas, in and around the basal cisterns, corpus callosum, and cavernous sinuses.158 The order of regression of the primitive meninx is reflected in the distribution of lipomas.158 Thus intracranial lipomas are thought most appropriately to represent developmental rather than neoplastic pathology of the meninges.158 This network of concentric membranes consists of the pachymeninx (dura mater) and the leptomeninges (arachnoid and pia mater) (Fig. 18-1). The dura mater is the most superficial membrane, a thick tough structure composed of dense connective tissue.24 The dura is composed of two layers: (1) an outer periosteal layer, which is highly vascularized, serves as the true periosteum of the inner table of the calvaria, and is not of meningeal origin,145 and (2) an inner meningeal layer, which is derived from the meninx. The cranial dural layers split to form the venous sinuses. The term dura (from Latin durus, meaning “hard”) is an apt descriptor of the structure that maintains the position of the cerebral hemispheres and posterior fossa structure by its reflections, such as the falx cerebri and tentorium cerebelli. The arachnoid and pia mater constitute the leptomeninges (from Greek lepto and meninx, meaning “slender membrane”). The delicate arachnoid is adjacent to the inner surface of the dura and is thinner over the convexities than at the base of the brain. The pia mater is a fine membrane that extends into the depths of the sulci.

Extraaxial Spaces The meninges delimit the extraaxial compartments of the central nervous system (CNS). The epidural space is created when the dura is separated from the calvaria. Although the subdural space (between the

536

dura and arachnoid membranes) has conventionally been characterized as a potential space containing minimal fluid, cells of the arachnoid actually form an intimate network with those of the meningeal dural layer.49 Electron microscopy has provided evidence that cells belonging to the inner dural layer may be found on both sides of this space when collections form in the subdural space.49,67 The subdural space, therefore, is formed by cleavage through the inner layer of the dura rather than by a true separation of dura and arachnoid and as such probably exists only in pathologic states.67 The SAS (between the arachnoid and pia mater) contains cerebrospinal fluid (CSF), which flows throughout the CNS and drains into the venous sinuses through the valves of the arachnoid granulations.145 The pia and the arachnoid are joined by fine connective tissue and cellular septa that traverse the SAS.21 Near the base of the brain, though, the pia and arachnoid widely separate to accommodate the basal cisterns. Perivascular, or Virchow-Robin, spaces were originally thought to be potential pathways connecting the CSF and deep brain structures by virtue of continuity with the SAS.44 More recent studies, however, have concluded that perivascular spaces are within the subpial space, separated from the SAS by pia mater4 (Fig. 18-2).

Extraaxial Collections Fluid collections in the epidural space assume a localized biconvex configuration as a result of the strong force needed to detach the firmly adherent dura from the inner table of the calvaria. The outer dural layer is most tightly adherent at sutures; classic teaching therefore holds that epidural collections do not cross suture lines.63 Uncommon exceptions to this rule do occur, however. After administration of paramagnetic contrast medium, the dura adjacent to an epidural hematoma (EDH) has a curvilinear enhancing appearance. An inflammatory reaction with formation of granulation tissue on the outer surface of the dura may produce increased thickness and intensity of enhancement in the dura immediately subjacent to an EDH, as demonstrated in Figure 18-3.70 Greater tissue contrast and multiplanar imaging capability contribute to the superior sensitivity of magnetic resonance imaging (MRI) compared with computed tomography (CT) in detecting small subdural fluid collections.87 MRI is especially useful in cases of subacute subdural hematoma (SDH), which may appear isodense to cortex on CT. Subdural hygromas result from leakage of CSF into the subdural space, probably after a tear in the arachnoid. MRI can distinguish SDH from subdural hygroma by improved detection of blood products, which are absent in a hygroma, but it cannot distinguish normal from infected subdural fluid, because peripheral dural enhancement may be seen in both infected and noninfected subdural collections (Fig. 18-4). A diffusely enhancing infected subdural collection may mimic an  en plaque meningioma.105

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

537

Calvaria Dura Barrier cells Subarachnoid space Pia mater

Arachnoid

FIG 18-1  Schematic representation of the dura and leptomeninges. (From Meltzer CC, et al: MR imaging of the meninges: Part 1. Normal anatomic features and nonneoplastic disease. Radiology 201:297–308, 1996.)

Arachnoid Subarachnoid space

Arachnoid vessel Pia mater

FIG 18-3  Epidural abscess. Enhanced axial T1-weighted image shows a lentiform epidural collection of pus with dural enhancement (arrows) that could mimic an epidural hematoma.

Parenchymal vessel

VirchowRobin space

FIG 18-2  Diagram illustrating the anatomy of the perivascular (VirchowRobin) space with respect to the pia mater and subarachnoid space. (Modified from Fukui MB, et al: MR imaging of the meninges: Part 2. Neoplastic disease. Radiology 201:605–612, 1996.)

MAGNETIC RESONANCE IMAGING MRI is substantially more sensitive than CT for visualizing both normal and abnormal meninges40,58,73,108,155,157 (Fig. 18-5). Beamhardening and other artifacts adjacent to the calvaria may be partly responsible for the diminished detection of meningeal enhancement with CT. Experimental evidence suggests that more intense enhancement results in areas of blood-brain barrier disruption with gadolinium (Gd)-DTPA–enhanced MRI than with CT performed after iodinated contrast.94,139 It has been suggested that MRI may be as sensitive as or even superior to CT in detecting subarachnoid hemorrhage (SAH) in the subacute and chronic phases and when a fluid-attenuated inversion recovery (FLAIR) pulse sequence is used.119,121 Other evidence suggests that FLAIR sequences may result in false-negative interpretations in the setting of SAH.171

Normal Meninges on MRI The normal meninges may demonstrate short segments of thin, low signal intensity on standard spin echo sequences.46 Intravenous (IV) administration of Gd-DTPA results in enhancement of the normal cranial dura, which lacks a blood-brain barrier, in an interrupted pattern of short linear segments that is typically most prominent

parasagittally30,155 (Figs. 18-6 and 18-7). Field strength may influence the conspicuity of “normal” meningeal enhancement as well as detection of pathologic enhancement. Earlier literature reported a distinct lack of enhancement of normal meningeal structures with relatively low-field-strength MRI.23 The greater signal-to-noise levels achieved at relatively higher field strengths potentiate increased detection of meningeal enhancement (see Fig. 18-6). Cohen and colleagues reported that meningeal enhancement present on more than three contiguous 1.5 tesla (T) spin echo MRI was highly correlated with significant intracranial abnormality.30 Thick, long, or intensely enhancing segments, as well as nodular meningeal enhancement, are particularly suspect. The normal falx and dura may occasionally enhance in a thin uniform pattern.88 The typical lack of normal meningeal enhancement observed in MRI of the spine may reflect the absence of the vascular outer dural layer that is found in the cranial pachymeninx.

MRI for Detection of Meningeal Disease The optimal imaging protocol for detection of meningeal disease is based on (1) technique, (2) sensitivity, (3) role, (4) cardinal signs, and (5) principal patterns (Box 18-1).

Imaging Technique The selection of MRI has a significant impact on the conspicuity of normal meningeal enhancement and sensitivity in detecting meningeal disease (see Figs. 18-6 and 18-7). Numerous factors influence the appearance of the meninges on MRI, including (1) the presence and amount of MR contrast agent administered, (2) the type of pulse sequence performed, and (3) the exact pulse sequence parameters. For partial saturation (spin echo or gradient echo), short echo time (TE) sequences, crucial imaging parameters include the repetition time (TR) and excitation flip angle. The shorter the TR, the greater the degree of saturation of magnetization (i.e., decreased signal intensity) from all imaged tissues. As a result, contrast-enhancing tissue will be more conspicuous against the more saturated (i.e., hypointense)

http://radiologyebook.com

538

PART II  CT and MR Imaging of the Whole Body

A

B FIG 18-4  Subdural empyema. A and B, Enhanced axial CT images at the level of the falx cerebri show dural enhancement surrounding a subdural empyema.

A

B FIG 18-5  Superior sensitivity of MRI over CT in detection of meningeal disease: Pott’s puffy tumor. A, Enhanced axial CT image at the level of an extracranial subperiosteal collection of pus. B, Enhanced axial MRI obtained after administration of gadolinium at the same level shows enhancement of the subjacent dura-arachnoid (arrows). Detection of this meningeal enhancement prompted a change of antibiotics to improve CSF drug penetration.

background tissues on a typical short TR, large flip-angle gradient echo study than on a spin echo study. Therefore the normal meninges usually exhibit diffuse enhancement on such ultrashort TR, large flipangle gradient echo imaging sequences (see Fig. 18-7).46 Imaging plane also affects visualization of meningeal enhancement; the coronal plane is preferred to axial MRI data for evaluating meningeal enhancement over the cerebral convexity. Similarly, factors that either increase signal from the meninges or decrease signal from background tissues enhance the contrast-to-noise ratio (CNR) and therefore the conspicuity of the enhancing meninges. Double-dosing and triple-dosing of paramagnetic contrast agents have demonstrated improved detection of parenchymal lesions with MRI,72

and there is evidence to support a similar dose effect for enhanced MRI of meningeal disease.86,140 More recently, cases of leptomeningeal metastases that had been diagnosed by MRI and high-dose (0.3 mmol/ kg) gadolinium that were not visualized with standard dosing (0.1 mmol/kg) technique have been reported62,86 (Fig. 18-8). Also, the use of fat saturation or magnetization transfer options increases the CNR by decreasing the signal from the background tissues.41,140

Sensitivity of MRI In the past, nonenhanced spin echo MRI was insensitive in detecting meningeal disease.108 The advent of the FLAIR sequence has greatly improved the sensitivity of MRI performed in the absence of

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

A

B FIG 18-6  Effects of field strength on conspicuity of normal meningeal enhancement. A, Axial T1-weighted image acquired at 0.5 T immediately after injection of 0.1 mmol/kg gadolinium. B, One day later in the same patient, axial T1-weighted image acquired at 1.5 T immediately after injection of the same dose of gadolinium shows more prominent enhancement of the falx (arrows) than in A. (From Meltzer CC, et al: MR imaging of the meninges: Part 1. Normal anatomic features and nonneoplastic disease. Radiology 201:297–308, 1996.)

A

B FIG 18-7  Effects of pulse sequence on conspicuity of normal meningeal enhancement. A, Axial T1-weighted spin echo image in a patient shows faint short-segment meningeal enhancement (arrow). B, Axial spoiled gradient recalled echo image in the same patient shows thin continuous enhancement in a dura-arachnoid pattern. (From Meltzer CC, et al: MR imaging of the meninges: Part 1. Normal anatomic features and nonhttp://radiologyebook.com neoplastic disease. Radiology 201:297–308, 1996.)

539

540

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 18-8  Effect of contrast dose on meningeal enhancement. Axial T1-weighted images obtained after administration of 0.1 mmol/kg (A) and 0.2 mmol/kg (B) doses of gadolinium show increased enhancement in a pia–subarachnoid space pattern at the higher dose in this patient with carcinomatosis of the meninges from gastric carcinoma.

BOX 18-1  Protocol for Magnetic

Resonance Imaging Performed at 1.5 Tesla • Axial FLAIR image • Axial T1-weighted image • Axial and coronal T1-weighted image after administration of a double or a triple dose of gadolinium with a fat-saturation pulse or magnetization transfer

result in false-negative diagnoses of meningeal or SAS disease in cases of infectious meningitis, carcinomatosis, and SAH.171 In this series a false-positive diagnosis of meningeal or SAS pathology was made in the presence of normal and hyperintense cortex, susceptibility artifact, prominent pial vessels, concatenated saturation pulse, or flow artifact.171 Neoplastic or inflammatory processes involving the SAS are best evaluated with paramagnetic contrast agents for visualization of enhancement of the accompanying meningeal involvement (Fig. 18-9).

FLAIR, fluid-attenuated inversion recovery.

Role of Imaging

gadolinium to detect meningeal and SAS abnormalities.148,171 Singer and coworkers evaluated the use of FLAIR MRI in 62 patients (21 with proven SAS or meningeal disease) and 41 control patients.148 The sensitivity, specificity, and accuracy of FLAIR for SAS disease were 85%, 93%, and 90%, respectively.148 Although all six cases of acute SAH were interpreted as abnormal on FLAIR images in the Singer series, this was a source of false-negative interpretation in the series by Williams and associates.148,171 Poor detection of SAH has been attributed to the relatively high oxygen tension in the SAS, which does not permit evolution to deoxyhemoglobin,65 and to the diluting effect of CSF. In 24 patients who underwent both FLAIR and Gd-enhanced T1-weighted MRI, FLAIR imaging (sensitivity, specificity, and accuracy of 86%, 91%, and 89%, respectively) was superior to Gd-enhanced T1-weighted imaging (43%, 88%, and 74%).148 Williams and coauthors prospectively evaluated 376 consecutive cases performed with FLAIR imaging and showed that FLAIR may

Before and after gadolinium administration, MRI plays an important role in the diagnosis of meningeal neoplasm in the asymptomatic patient when CSF cytologic examination results are equivocal or when lumbar puncture is contraindicated.26,50,55,155 Imaging, however, does not replace CSF examination in diagnosis of meningeal neoplasm. The limited sensitivity of CSF cytologic examination continues to be a significant obstacle and reinforces the complementary role of imaging.56 The percentage of positive spinal CSF cytology in cases of primary CNS tumors with histologically confirmed meningeal involvement varies from 12% to 63% and is increased in symptomatic patients.8,59,91,115 With meningeal metastases from non-CNS neoplasms, CSF cytology was positive in 45% to 80% of cases, with higher yields after multiple spinal taps.122,165 CSF flow cytometry increases the detection of CSF spread of hematologic malignancies over cytologic examination alone.22 The modest sensitivity of CSF cytology has fueled the search for CSF tumor markers to detect dissemination of neoplasm. Recent literature suggests that markers such as CA 15-3 may have value in detecting breast carcinoma in CSF, and nuclear magnetic spectroscopy may have value in detecting metabolites in CSF indicating disseminated lung adenocarcinoma.6,99 Additional tumor markers that

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

A

541

B FIG 18-9  Axial FLAIR image in a patient with leptomeningeal spread of breast carcinoma (A) shows signal abnormality in a pia–subarachnoid space pattern (arrowheads) that enhances on the T1-weighted image after administration of gadolinium (B).

may be detected in CSF to indicate leptomeningeal neoplasm include transforming growth factor (TGF)-β1, vascular endothelial growth factor (VEGF), urokinase-type plasminogen activator (uPA), and tissue-type plasminogen activator (tPA).161 Like CSF cytology, tumor detection with CSF markers may also be hampered by suboptimal sensitivity and specificity.117,160 The sensitivity of MRI in detecting meningeal neoplasm was originally reported to be lower than that of CSF cytology, with high falsenegative rates: 30% to 33% for MRI and 58% for CT.27,177 The early studies, however, may have unwittingly introduced a selection bias by using positive cytology as an inclusion criterion.27,177 Later series comparing CT myelography with MRI and cytology showed a higher detection rate of CSF metastases with MRI (range, 65%-72%) than with CT myelography (range, 45%-47%) and cytology (29%).73,93 A Memorial Sloan-Kettering Cancer Center study showed that when the patient cohort was not restricted to patients with positive cytologic findings, the rate of detection of meningeal carcinomatosis was increased.50 This study examined 137 patients with signs and symptoms of meningeal disease.50 CT and MRI were assessed for signs of meningeal or SAS neoplasm, including hydrocephalus, enhancement of the dura, leptomeninges, and cranial nerves.50 Leptomeningeal metastases were identified in 77 of 137 patients.50 The diagnosis of leptomeningeal metastases was based on the clinical and imaging findings alone in 31% of those cases.50 Abnormal imaging findings were reported much more frequently in cases of solid tumor primaries (90%) than in hematologic neoplasms (55%).50 More recent studies confirm these findings.178 A caveat in using MRI for the diagnosis of leptomeningeal tumor involvement, therefore, is that MRI is less sensi-

tive in detecting involvement of the meninges resulting from hematologic malignancies than in detecting detecting solid tumors.27,50,177,178 When the distinction between meningeal neoplasm and inflammatory meningeal disease cannot be established by CSF and clinical data, MRI can support the diagnosis of neoplasm and guide meningeal biopsy.84,98 Although infectious meningitides are often uncovered by CSF analysis, MRI may be used to target meningeal biopsies when needed. Cheng and coworkers reported an improved yield from meningeal biopsy specimens in cases of chronic meningitis when tissue specimens were obtained from enhancing regions identified on MRI study29 (Fig. 18-10). MRI has an advantage over CSF examination in its ability to characterize bulky neoplastic disease that may be more responsive to radiation therapy.26 Because CSF cytologic examination produces a higher false-negative rate in focal than in diffuse disease,15,56 imaging may be of particular value in detecting focal spread of neoplasm to the meninges or SAS. MRI also has the potential to allow noninvasive monitoring of treatment response, although it is conceivable that the imaging abnormalities may persist beyond the eradication of neoplastic cells in the CSF.173 In a series of patients with coccidioidal meningitis, diminution of meningeal enhancement was seen in treated patients with improving CSF profiles.175 Therefore MRI also may be useful as a means of monitoring response to therapy in fungal meningeal disease.175 Imaging is a useful adjunct for the diagnosis of neoplastic meningeal disease in the appropriate clinical setting, along with CSF examination to exclude infectious or noninfectious processes of the meninges.50

http://radiologyebook.com

542

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 18-10  Chronic meningitis. A, Thick slightly irregular dura-arachnoid enhancement on axial T1-weighted imaging proved to be idiopathic hypertrophic pachymeningitis on biopsy. B and C, In another patient, thin slightly irregular dura-arachnoid enhancement on axial (B) and coronal (C) T1-weighted imaging also proved to be idiopathic hypertrophic pachymeningitis on biopsy.

Finally, imaging may allow the clinician to assess outcome, in that diffuse leptomeningeal involvement confers a poor prognosis.16

Cardinal Imaging Signs of Meningeal Pathology The main imaging findings that have been associated with meningeal pathology are as follows33,43,50,82,93,136,154,155,177: • Hydrocephalus (Fig. 18-11) • Dura and arachnoid enhancement or signal abnormality (see Fig. 18-10)

• Pia and SAS enhancement or signal abnormality (see Fig. 18-11) • Subependymal enhancement or signal abnormality (Fig. 18-12) Hydrocephalus may or may not be associated with enhancement  of the meninges or ependyma. In this setting, hydrocephalus alone implies a resorptive block to CSF flow. Hydrocephalus may occur in the setting of SAH, infectious or noninfectious meningitis, and neoplasm. In the patient with neoplastic meningeal disease, hydrocephalus is more likely when leptomeningeal invasion or SAS has occurred rather than when neoplasm is limited to the dura.166

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

B

A

FIG 18-11  Hydrocephalus with ependymal enhancement and pia–subarachnoid space abnormality. A, Axial T1-weighted image obtained after gadolinium administration shows enlargement of the lateral ventricles and thin linear ependymal enhancement. B, There is a pia-subarachnoid enhancement pattern between the cerebellar folia and along the third cranial nerves, indicating meningeal spread of small cell lung carcinoma in a patient with multiple cranial neuropathies.

A

B FIG 18-12  Focal nodular ependymal enhancement. A, Nodular enhancement of the ependyma of the frontal horns was the probable entry point for subarachnoid spread of melanoma. B, Sagittal T1-weighted image of the lumbar spine shows nodular enhancing drop metastases to the cauda equina (arrowheads).

http://radiologyebook.com

543

544

PART II  CT and MR Imaging of the Whole Body

A

B FIG 18-13  Diffuse nodular pia–subarachnoid space enhancement. A and B, Multiple pial-based enhancing nodules are present in this patient with disseminated lung carcinoma.

Enhancement of the meninges may occur in the spine or brain. Meningeal or subependymal enhancement may be focal (Fig. 18-13) or diffuse (see Fig. 18-12) and may have either a smooth or nodular contour (Fig. 18-14; see Fig. 18-13). Diffuse leptomeningeal involvement is a harbinger of a worse prognosis than focal disease.16 Although a nodular pattern of enhancement may suggest neoplasm, it is not specific; nonneoplastic entities (e.g., sarcoidosis) may also produce nodular thickening of the meninges (see Fig. 18-14A and B). Infiltration of the leptomeninges overlying the convexities or in the basal cisterns may result in sulcal or cisternal obliteration on a noncontrast T1-weighted image (Fig. 18-15) or a FLAIR sequence (see Fig. 18-9). Minimal shortening of T1 and T2 relaxation times in the cisternal CSF (“dirty CSF sign”) may be seen on nonenhanced MRI (see Fig. 18-15).18 In rare instances, subarachnoid tumor or inflammatory exudates may result in distention of the SAS or perivascular (subpial) spaces  (Fig. 18-16).

Imaging Patterns Two distinct patterns of meningeal enhancement or signal abnormality may be observed with MRI. The dura-arachnoid pattern follows the inner contour of the calvaria (Fig. 18-17), whereas the pia mater–SAS pattern extends into the depths of the sulci (Fig. 18-18). Enhancement or signal abnormality surrounding the brainstem is always of the pia mater–SAS type, in that the arachnoid is clearly separated from the pia mater by the intervening basal cisterns in this region. Although pia mater–SAS enhancement does occur more commonly in the setting of meningitis than with tumor involvement, both inflammatory and neoplastic processes may demonstrate either pattern.89,132 A diffuse appearance favors an inflammatory etiologic mechanism, whereas nodular meningeal enhancement suggests a neoplasm (Boxes 18-2 to 18-6). Dura-arachnoid pattern.  Because the dura and arachnoid are closely approximated, the distribution of meningeal enhancement  does not reliably distinguish purely pachymeningeal (dural) from  leptomeningeal (arachnoidal and pial) involvement. As a result the dura-arachnoid pattern of enhancement does not imply that the leptomeninges or the SAS is spared. This fact has practical importance because some authors report a lower incidence of positive CSF

cytologic findings with the dural pattern, in contrast to the pial pattern, implying a lack of SAS neoplasm.127 A surprisingly moderate rate (55%) of positive CSF cytologic results was reported in a subset of 11 patients with a dura-arachnoid enhancement pattern on MRI.50 The presence of malignant cells in the CSF of patients in this subgroup implies neoplastic involvement of the arachnoid and SAS. This confirms the limitation of MRI in distinguishing abnormal enhancement of the dura from that of the arachnoid.50 When restricted to the dura, however, meningeal carcinomatosis often results in negative CSF cytology; in this setting, MRI can play an important role in disease detection. The dural tail sign, once considered specific for meningioma, may be observed in a wide variety of other extraaxial lesions, including schwannoma (Fig. 18-19), dural metastases (Fig. 18-20), lymphoma, tuberculoma, and sarcoidosis.109 Occasionally a dural tail may be seen in association with intraaxial mass lesions such as gliomas or non-CNS metastases.109 Although this imaging feature has been useful in suggesting the diagnosis of meningioma, its lack of specificity may occasionally cause misleading interpretation of MRI.109 Similar signal, shape, and enhancement characteristics found with several of these lesions may further confound the distinction between meningiomas and other entities.109 Pia mater–subarachnoid space pattern.  Similarly, the pia mater–SAS pattern of enhancement may reflect neoplasm within the SAS, tumor infiltration of the pia mater, or both.177 The pia mater–SAS pattern is more common in patients with infectious meningitis than in those with a neoplasm,89,132 although the pia mater–SAS distribution is not at all specific for inflammation. A higher rate of positive CSF cytologic findings has been reported in conjunction with the pia mater–SAS pattern than with the duraarachnoid pattern of enhancement in cases of neoplasm.127 This may reflect the anatomy of the blood-brain barrier with respect to the meninges.89 The dura lacks a blood-brain barrier because its capillary endothelium has a discontinuous cell layer, whereas the outer layer of the arachnoid has capillaries with a continuous cell layer and tight junctions.142,143 Bloodborne neoplastic cells, therefore, may gain access to the dura more easily than to the SAS.89 A pia mater–SAS pattern

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

A

B

C FIG 18-14  Nonneoplastic focal nodular and linear meningeal enhancement. A and B, Nodular enhancement in a pia-subarachnoid pattern (arrows) mimics neoplasm in a case of sarcoidosis. C, Focal linear enhancement in a linear dura-arachnoid pattern (arrows) in a different case of sarcoidosis.

BOX 18-2  Differential Diagnosis: Focal

BOX 18-3  Differential Diagnosis: Diffuse

Infectious Adjacent petrous apicitis/sinusitis/mastoiditis Adjacent abscess or cerebritis Fungal infection (Aspergillus)

Infectious Bacterial (unusual)

Dura-Arachnoid Pattern

Noninfectious Inflammatory Reactive Calvarial metastasis Subdural or epidural hematoma Acute infarction Iatrogenic: catheter or craniotomy Sarcoidosis Neoplastic Meningioma Breast carcinoma Prostate carcinoma Lymphoma Posttransplant lymphoproliferative disorder (PTLD)

Dura-Arachnoid Pattern

Noninfectious Inflammatory Reactive Diffuse calvarial metastases Extensive subdural hematoma Iatrogenic: response to catheter or craniotomy Low intracranial pressure states Cerebrospinal fluid leak After lumbar puncture (unusual) Spontaneous intracranial hypotension (rare) Hypertrophic cranial pachymeningitis (rare) Wegener’s granulomatosis (rare) Multiple sclerosis (rare) Neoplastic Breast carcinoma http://radiologyebook.com Prostate carcinoma

545

546

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 18-15  “Dirty CSF sign” and ventricular debris. T1-weighted image (A) demonstrating obliteration of bilateral sulci, and FLAIR image (B) showing material that is hyperintense to CSF within the subarachnoid space (arrows) in a patient with gram-positive meningitis. C, After gadolinium administration there is extensive pia–subarachnoid space enhancement.

may reflect SAS enhancement, pial enhancement, or both.48 Gadolinium may leak through capillary tight junctions that have been disrupted by neoplastic invasion of the meninges48 and directly enter the CSF.

NONNEOPLASTIC MENINGEAL DISEASE Infectious Meningitis Bacterial Meningitis Bacterial meningitis results from hematogenous spread of infection or occasionally from direct extension from a paranasal sinus or mastoid

source.90 The mechanism of entry of bacteria from the intravascular space into the CSF is not well understood. Animal studies have suggested that bacterial cell wall elements provoke an inflammatory response that brings about opening of tight intercellular junctions in the arachnoidal capillary bed.89 This breakdown in the leptomeningeal blood-brain barrier allows bacteria to gain access to the SAS. Early congestion and hyperemia of the leptomeninges are succeeded by mixed inflammatory cell infiltration with exudate in the SAS, especially over the cerebral convexity (see Fig. 18-15). A link between meningeal enhancement and the degree of inflammatory cell infiltration of the leptomeninges has been proposed by

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

A

547

B FIG 18-16  Perivascular space pattern. Axial T2-weighted image (A) and enhanced T1-weighted image (B) show distention of the perivascular spaces at the base of the brain by the gelatinous nonenhancing pseudocysts of cryptococcosis.

FIG 18-17  Diagram illustrating the dura-arachnoid pattern of meningeal abnormality that follows the inner table of the skull and the dural reflections (arrows). (From Meltzer CC, et al: MR imaging of the meninges: Part 1. Normal anatomic features and nonneoplastic disease. Radiology 201:297–308, 1996.)

FIG 18-18  Diagram illustrating the pia–subarachnoid space pattern of meningeal abnormality that invaginates into sulci (arrows). (From Meltzer CC, et al: MR imaging of the meninges: Part 1. Normal anatomic features and nonneoplastic disease. Radiology 201:297–308, 1996.)

http://radiologyebook.com

548

PART II  CT and MR Imaging of the Whole Body

BOX 18-4  Differential Diagnosis: Focal Pia

BOX 18-6  Differential Diagnosis: Diffuse

Infectious Bacterial meningitis Tuberculous meningitis (basal) Fungal (Aspergillus, Cryptococcus, Coccidioides) Viral (herpes) Adjacent abscess or cerebritis Spirochetal (neurosyphilis)

Infectious Bacterial Spirochetal Neurosyphilis Lyme disease Viral (“aseptic”) meningitis CMV Varicella-zoster (spinal in AIDS)

Mater–Subarachnoid Space Pattern

Noninfectious Inflammatory Reactive Sarcoid Subarachnoid hemorrhage Multiple sclerosis (rare)

Pia Mater–Subarachnoid Space Pattern

Noninfectious Inflammatory Reactive Irritative (“aseptic”) meningitis Response to foreign material: Contrast agents Chemical (e.g., ruptured dermoid) Subarachnoid hemorrhage

Vascular Pial vascular malformation (e.g., Sturge-Weber syndrome) Subarachnoid hemorrhage Superficial siderosis Granulomatous angiitis

Vascular Subarachnoid hemorrhage

Hamartomatous Meningioangiomatosis

Neoplastic Primary CNS AIDS lymphoma Glioblastoma multiforme Astrocytoma Primitive neuroectodermal tumor Primary leptomeningeal gliomatosis (very rare) Secondary neoplasm Melanoma Breast carcinoma Lung carcinoma False-positive result on FLAIR image Concatenated saturation pulse

Neoplastic Primary CNS Glioblastoma multiforme Astrocytoma Primitive neuroectodermal tumor Medulloblastoma Germinoma Ependymoma Secondary neoplasm Melanoma Breast carcinoma Lung carcinoma False-positive on FLAIR image Normal, hyperintense cortex (especially at convexities) Susceptibility artifact Prominent pial vessels Flow artifact (especially in basal cisterns)

AIDS, acquired immunodeficiency syndrome; CMV, cytomegalovirus; CNS, central nervous system; FLAIR, fluid-attenuated inversion recovery.

CNS, central nervous system; FLAIR, fluid-attenuated inversion recovery.

BOX 18-5  Differential Diagnosis:

Perivascular Space Pattern Infectious Enhancing Tuberculosis Nonenhancing Cryptococcosis Neoplastic Carcinomatosis Breast carcinoma Lung carcinoma

FIG 18-19  Focal dura-arachnoid enhancement. A dural tail (arrowheads) adjacent to a left acoustic schwannoma illustrates the nonspecificity of this sign as an indicator of meningioma on an enhanced axial T1-weighted image.

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

A

B

C

D

549

FIG 18-20.  Focal dura-arachnoid mass. CT scan (A, B) shows colon adenocarcinoma metastases to the calvaria (arrow) extending to the dura (arrows) on enhanced axial and coronal T1-weighted images (C, D).

Mathews and coworkers.108 The lack of gadolinium enhancement in some areas of minimal inflammation suggests a threshold effect whereby a critical degree of inflammatory reaction is required before meningeal enhancement may be observed.108 MRI is superior to CT in the evaluation of complications of meningitis, including subdural empyemas, dural venous thrombosis, secondary ischemia, and parenchymal lesions.28 Septic thrombosis of the superior sagittal sinus, although uncommon since the advent of antibiotics, may complicate bacterial meningitis150 (Fig. 18-21). When thrombosis is suspected, MRI is the imaging modality of choice for confirming the presence of thrombus. Absent flow may be detected as lack of a normal flow void in the sagittal sinus on short TR images and may be optimally demonstrated by magnetic resonance angiography (MRA) using two-dimensional phase-contrast tech-

niques.85 Conversely, localized meningeal enhancement may also be seen overlying an adjacent parenchymal infectious process, such as a brain abscess or encephalitis. Bacterial meningitis may result in ventriculitis, especially in the setting of gram-negative and iatrogenic infections.51 An early finding in ventriculitis may be that of irregular debris in the dependent portions of the ventricles.51 Infectious meningeal processes may occasionally spread across the pial barrier to cause cerebritis.

Mycobacterial Meningitis CNS tuberculosis is increasing in frequency, partly because of its  proclivity to occur with human immunodeficiency virus (HIV) infection. Tuberculous meningitis commonly occurs as a result of disseminated pulmonary infection, usually along with parenchymal brain

http://radiologyebook.com

550

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 18-21  Diffuse dura-arachnoid enhancement. A, Coronal T1-weighted image shows dura-arachnoid enhancement (arrows) in a patient with meningitis and subsequent superior sagittal sinus thrombosis. B, There is an absence of flow in the superior sagittal sinus on a sagittal phase-contrast MR venogram.

involvement, although it rarely may be seen as the sole manifestation of the disease.110 In contrast to bacterial meningitis, tuberculous infection is associated with a more insidious onset, fewer changes in the CSF profile, and higher rates of mortality and complications such as infarction. Although tuberculosis tends to involve the basal meninges, tuberculomas may occur within the brain parenchyma or subarachnoid, subdural, or epidural spaces18,133 (Fig. 18-22). In one series, meningeal enhancement was observed in 36% of HIV-infected patients with CNS tuberculosis evaluated with MRI.169 MRI of tuberculous meningitis demonstrates meningeal enhancement, most often in the basal cisterns, reflecting the known predilection of tuberculosis for the base of the brain. Calcification of the basal meninges may also be seen and is more easily appreciated with CT than with MRI. Focal or diffuse dura-arachnoid enhancement may be observed on contrast-enhanced MRI scans of the spine in cases of suspected active tuberculous meningitis.95 Thickening, inflammation, and fibrosis of the leptomeninges are commonly seen at histopathologic examination. Meningitis may also be caused by a virus or fungus, a fact that is of particular importance in immunocompromised patients. In patients with acquired immunodeficiency syndrome (AIDS), a polyradiculomyelitis may develop in rare circumstances as a result of cytomegalovirus (CMV) or varicella-zoster virus infection.54 Diffuse enhancement of the cauda equina on MRI should alert the clinician to exclude these causes in AIDS patients.168

FIG 18-22  Multiple ring-enhancing tuberculomas are distributed throughout the perivascular spaces and basal pial surfaces on an axial T1-weighted image after administration of gadolinium.

Fungal Meningitis Fungal meningitis usually results from hematogenous dissemination of disease and is also often seen in immunosuppressed hosts. Fungal meningitides, such as those caused by Aspergillus, Cryptococcus, and Coccidioides species, may be demonstrated by contrast-enhanced MRI.107,113,175 Aspergillosis is an important cause of morbidity and mortality in organ transplant recipients and may be associated with focal, thick dura-arachnoid enhancement that resolves with treatment.113 Cryptococcus frequently infects patients with AIDS, but meningeal enhancement is seen only occasionally in cryptococcal meningitis, probably because of the blunted host response.108,144 Meningeal

enhancement was observed in only one in five autopsy-proven cases of AIDS-related cryptococcosis studied with Gd-enhanced MRI.107 Cryptococcal infection tends to spread from the basal cisterns via perivascular spaces to the basal ganglia, brainstem, internal capsule, and thalamus.167 This produces a characteristic MRI appearance of nonenhancing dilated perivascular spaces, caused by infestation with the gelatinous pseudocysts of cryptococcal material (see Fig. 18-16). In contrast to the pattern seen with Cryptococcus, Wrobel and colleagues reported MRI enhancement of the meninges in 7 of 11 patients with coccidioidal meningitis.175 Enhancement was most prominent in

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

FIG 18-23  Coccidioidomycosis. Axial T1-weighted image demonstrates diffuse meningeal enhancement within the suprasellar cistern in a patient with coccidioidomycosis diagnosed with lumbar puncture.

A

551

B

FIG 18-24  Lyme disease. Sagittal T1-weighted images of the lumbar

the basilar cisterns, the sylvian and interhemispheric fissures (Fig. 18-23), and the upper cervical SAS. Spirochetal infections may exhibit both localized and diffuse meningeal enhancement.20 Similarly, smooth diffuse pia-SAS enhancement of the brainstem and the entire spinal cord has been reported in Lyme disease (Fig. 18-24), even in the absence of MRI evidence of parenchymal disease.36

Aseptic Meningitis Meningitis is designated “aseptic” when CSF cultures are sterile. Aseptic meningitis may have a variety of causes, including viral infection and irritation after introduction of foreign substances into the CSF (e.g., blood, chemical agents, contrast materials), or it may be seen against the backdrop of connective tissue disorders.110 MRI may demonstrate meningeal enhancement or subarachnoid signal abnormality (Fig. 18-25). In cases of suspected aseptic meningitis, performing MRI after administration of paramagnetic contrast medium may help prevent delays in diagnosis.45

Noninfectious Inflammatory Meningitis Noninfectious granulomatous disease can also involve the meninges. Approximately 5% of patients with sarcoidosis have neurologic manifestations; of those, the most common sign is that of cranial nerve palsies.153 Gd-enhanced MRI is the study of choice to document meningeal involvement and may show either the pia mater–SAS or the dura-arachnoid pattern (see Fig. 18-14).146,147,179 Striking thickening of the meninges may also be demonstrated. Sarcoid was present in 31% of 37 patients with chronic meningitis of unknown etiology examined with enhanced MRI and meningeal biopsy and as such was the most frequent cause of chronic meningitis in that series.29 Meningeal abnormality may also occur in Wegener’s granulomatosis135 (Fig. 18-26). Thickened nodular meninges on both enhanced CT and MRI in a patient with Wegener’s granulomatosis were reported by Tishler and colleagues.156 The imaging features reflect the histopathologic findings of dural and brain biopsies that revealed granulomatous

spine before (A) and after (B) administration of intravenous gadolinium demonstrate thick linear enhancement of the spinal conus and all nerve roots of the cauda equina in a patient with Lyme disease.

infiltration of the dura and development of fibrosis and granulomas of the pia mater and arachnoid. Hypertrophic cranial pachymeningitis is a rare type of diffuse granulomatous inflammation that produces thickening, chronic inflammation, and fibrosis of the dura.104,106 Common presenting symptoms include headache, cranial nerve palsies, and ataxia. Diffuse dural enhancement (on enhanced T1-weighted images) or hypointense and thickened dura (on T2-weighted images) may be demonstrated on MRI (see Fig. 18-10). Diagnosis requires confirmation with dural biopsy. The disease may progress despite corticosteroid therapy. In rare instances, meningeal enhancement has been reported to accompany demyelinating disease. We are aware of few reports of meningeal enhancement on MRI in cases of multiple sclerosis (MS). One case displayed a diffuse dura-arachnoid pattern; another showed focal pial enhancement.11,35 Serial MRI of MS plaques suggests that parenchymal lesion enhancement is related to the inflammatory process associated with active demyelination.11 Analogously, meningeal enhancement in MS may result from extension of perivascular lymphoplasmacytic infiltration into the leptomeninges, found in 41% of autopsy cases.66 A similar inflammatory etiology was proposed by Fulbright and coauthors52 in describing enhancement of multiple cranial nerves in Guillain-Barré syndrome, a disorder characterized by peripheral demyelination (Fig. 18-27).

Iatrogenic Meningeal Enhancement Benign meningeal enhancement on MRI may result from mechanical disruption of the meninges from a variety of causes. Localized or diffuse dural enhancement may occur after craniotomy and may persist indefinitely23,40 (Fig. 18-28). It may be difficult to distinguish normal reactive postoperative enhancement from meningeal tumor involvement following craniotomy for tumor resection. Postoperative meningeal enhancement usually regresses over time, often resolving

http://radiologyebook.com

552

PART II  CT and MR Imaging of the Whole Body

A

B FIG 18-25  Aseptic meningitis. Thin linear pia–subarachnoid space enhancement (arrowheads) is seen in a patient with headache and no organism growth on CSF culture on axial (A) and coronal (B) enhanced T1-weighted images.

A

B FIG 18-26  Wegener’s granulomatosis. A and B, Diffuse thickening and enhancement of the meninges in a dura-arachnoid pattern (arrows) are present in a patient with biopsy-proven Wegener’s granulomatosis on a contrast-enhanced CT scan.

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

553

catheter may also result in diffuse or localized dural enhancement. In rare instances, dura-arachnoid enhancement may result from uncomplicated lumbar puncture.111 Dural enhancement after lumbar puncture occurs more commonly, however, when intracranial hypotension results, suggesting a CSF leak.17

Vascular Disease

FIG 18-27  Guillain-Barré syndrome. Sagittal T1-weighted image of the lumbar spine demonstrates thick linear enhancement of multiple nerve roots of the cauda equina in a patient with Guillain-Barré syndrome.

FIG 18-28  Postoperative meningeal enhancement. Coronal T1weighted image shows diffuse linear enhancement in a patient who had undergone previous craniotomy for meningioma.

Occasionally an acute cerebral infarction may demonstrate adjacent meningeal enhancement (Fig. 18-29). This “meningeal enhancement sign” occurs between day 2 and day 6 after infarction. It is seen  with large supratentorial infarcts and usually has a dura-arachnoid configuration.39,42 Although the cause of meningeal enhancement in cerebral infarction is not established, possible contributing factors include reactive hyperemia and local inflammation following meningeal irritation.39 Hemorrhage into the SAS also irritates the leptomeninges, causing inflammation of the pia mater and arachnoid.69 In severe cases, fibrosis may ensue and eventually may obliterate the SAS. Superficial siderosis is a rare condition resulting from deposition of hemosiderin in the leptomeninges and surface of the brain following recurrent SAH and is a cause of hearing loss.25,80,83 The leptomeninges are thickened and pigmented on pathologic examination. A characteristic hypointense rim is observed along the brain surface on T2-weighted images19,25 (Fig. 18-30). This hypointense border of hemosiderin deposits in superficial siderosis becomes exaggerated, and thus more readily detected, on gradient echo MRI as a result of augmented magnetic susceptibility effects.83 The thickened meninges may also enhance with gadolinium. Vasculitis rarely involves the meninges. Preferential involvement of small leptomeningeal vessels in granulomatous angiitis may result in meningeal enhancement.116 MRI is the modality of choice for evaluation of Sturge-Weber syndrome, a phakomatosis characterized by deranged vasculature of the face, brain, and meninges.149 Poor venous drainage and chronic ischemic damage to the underlying cortex develop as a result of the leptomeningeal vascular malformation that originates in the SAS. MRI of the brain shows superficial gyriform enhancement. This enhancement likely results from slow flow within the leptomeningeal vascular malformation with or without enhancement of ischemia13,164 (Fig. 18-31). Meningioangiomatosis is a rare hamartomatous disorder of the leptomeninges distinguished by meningovascular proliferation and leptomeningeal calcification.3,68,130 A cortical mass with meningeal extension and heterogeneous signal intensity on T2-weighted images and demonstrating contrast enhancement on MRI has been reported130,170 (Fig. 18-32). Halper and coworkers have theorized that meningioangiomatosis is caused by a proliferation of both meningothelial cells and leptomeningeal vessels within perivascular spaces as they penetrate the cortex.68 A combined pia mater–SAS and dura-arachnoid pattern of enhancement has been rarely reported in association with migraine headaches.32,100 This unusual finding is attributed to hyperperfusion, documented on single-photon emission computed tomography (SPECT) or transcranial Doppler imaging.32,100

Toxic Meningeal Enhancement between 1 and 2 years after craniotomy.40 In addition, the distribution of enhancement may support a benign etiologic mechanism. In contrast to the typically basilar pattern of neoplastic meningitis, postoperative enhancement occurs more often over the convexities.23 The morphology of enhancement may also provide clues as to the cause. A nodular dural enhancement pattern or thick pia mater–SAS enhancement suggests recurrent tumor.79 Placement of a ventricular

Chemical meningitis may arise following rupture of dermoid or epidermoid cyst. This inflammation is likely provoked by the irritating effect of cholesterin crystals and keratin material discharged into the CSF.34,101 In such cases the diagnosis of ruptured dermoid is usually verified by demonstration of droplets of high-signal lipid material dispersed throughout the SAS on nonenhanced T1-weighted MRI. Meningeal irritation resulting in contrast enhancement on MRI has also been reported in the setting of adverse drug reactions. Eustace and

http://radiologyebook.com

C

A

FIG 18-29  Meningeal enhancement after infarction. A, Axial proton-density–weighted image shows the left frontal infarct. B, Sagittal T1-weighted image shows bright signal in the cortex of the left temporal infarct (arrows). C, There is gyriform enhancement on the axial T1-weighted image that is probably a combination of cortical and pial enhancement (arrows).

B

A

B FIG 18-30  Siderosis. The dark signal from hemosiderin outlines the pial surface of frontal and parietal lobes

http://radiologyebook.com (A) and cerebellum (B) in this patient with superficial siderosis on axial T2*-weighted images.

CHAPTER 18  Meningeal Processes

A

555

B

C

D

FIG 18-31  Sturge-Weber syndrome. A, Gyriform enhancement defines the pial vascular malformation and adjacent ischemic occipital cortex on an axial T1-weighted image. B, In another patient, axial CT image demonstrates gyriform calcification. C, Axial T2*-weighted image shows susceptibility artifact within left frontoparietal cortex and pia (arrow), consistent with calcification seen on CT. D, Coronal T1-weighted image shows gyriform enhancement due to pial vascular malformation.

Buff reported a case of ibuprofen-induced meningitis depicted by meningeal enhancement on MRI.45

Spontaneous Intracranial Hypotension Like the low-intracranial-pressure headache after complicated lumbar puncture, spontaneous intracranial hypotension is identified by a postural headache that resolves after placement of an epidural blood patch.138 A CSF pressure of less than 60 mm H2O in the absence of previous lumbar puncture establishes the diagnosis. MRI has shown striking thick, diffuse dural-arachnoid enhancement.47,78 Subdural fluid collections may occur in conjunction with this low-pressure state.138 An unsuspected spontaneous CSF leak may be the inciting factor (Fig. 18-33), and thus recognition of this combination of imaging and clinical findings should prompt a search 

for the source of the leak. Identification of this treatable cause of meningeal abnormality in association with the germane clinical  features may prevent diagnostic delay and unnecessary meningeal biopsy.

NEOPLASTIC MENINGEAL DISEASE Primary Involvement Primary dural neoplasms are rare. Neoplasms arising from the dura are mesenchymal in origin because the dura is composed of fibrous tissue.141 Primary dural tumors bridge the entire range of benign (fibromas) to intermediate (fibromatosis) to malignant (sarcomas).137 An assortment of dural sarcomas has been reported, including chondrosarcoma, fibrosarcoma, and malignant fibrous histiocytoma.7,97,103

http://radiologyebook.com

556

PART II  CT and MR Imaging of the Whole Body

A

B FIG 18-32  Meningioangiomatosis. A, Axial T2-weighted image reveals left temporal gyriform hypointensity with subcortical hyperintensity (arrows) that may represent the meningovascular proliferation. B, This axial spoiled gradient recalled acquisition in a steady state (GRASS) image does not reveal enhancement.

A

B FIG 18-33  Intracranial hypotension. A, Coronal T1-weighted image demonstrates diffuse dura-arachnoid enhancement in this patient who presented with severe headaches and whose CSF pressure was low on lumbar puncture. B, The axial T2-weighted image shows fluid presumed to be CSF in the soft tissues of the upper neck (arrows). The patient recovered after surgery despite the fact that although myelography revealed a leak, the exact site could not be found at operation.

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

A

557

B

FIG 18-34  A, Focal linear dura-arachnoid enhancement characterizes this right tentorial en plaque meningioma. B, In another patient, sagittal T1-weighted image of the thoracic spine demonstrates an enhancing intradural extramedullary mass with dural thickening (arrows) demonstrating the meningeal origin of the mass.

Previous radiation therapy for another disease has been connected with some sarcomas.7 The most common primary tumor arising from the leptomeninges is meningioma141 (Figs. 18-34 and 18-35). Meningiomas originate from meningothelial cells of the arachnoid.141 A hormonal influence has been implicated in the pathogenesis of meningiomas, because they are more common in women than in men, occur more often in patients with breast cancer, and may become symptomatic during pregnancy.14,81,141 Radiation therapy has also been implicated as a causative factor in meningioma,53,152 especially in patients treated for tinea capitis or vascular nevi in childhood. Fifty percent of meningiomas are isointense with gray matter on T1-weighted and T2-weighted imaging; the remaining half vary in signal intensity.151 Heterogeneous signal intensity on MRI may result from calcified, vascular, or cystic components.151 Additional primary leptomeningeal neoplasms such as glioma37,77,98,129 (Fig. 18-36), melanoma,5 melanocytoma114 (Fig. 18-37), sarcoma,131,141 and lymphoma96 are very rare. Primary leptomeningeal gliomas probably originate from heterotopic rests of glial tissue within the meninges and SAS.31,174

Secondary Involvement Breast carcinoma, lymphoma, leukemia, lung carcinoma, malignant melanoma, gastrointestinal carcinoma, and genitourinary carcinoma are the most common neoplasms to progress to meningeal dissemination.74,122,165 Primary CNS neoplasms such as medulloblastoma and primitive neuroectodermal tumor (PNET), pineoblastoma (Fig. 18-38), ependymoma, germ cell tumor (Fig. 18-39), astrocytoma (Fig. 18-40), and glioblastoma (Fig. 18-41) also have a propensity for meningeal spread.8,9,123,124,128 Most tumors tend to produce focal involvement in a dura-arachnoid or pia-SAS distribution. Exceptions to this are aggressive tumors such as lung and breast carcinoma, which may result in diffuse meningeal neoplasm.

Risk Factors for Leptomeningeal Dissemination of Neoplasm The risk factors for leptomeningeal dissemination differ somewhat for primary CNS and non-CNS neoplasms but are largely related to tumor dedifferentiation and proximity to CSF spaces. Characteristics of primary CNS neoplasms that correlate highly with a tendency to CSF spread are (1) poor glial fibrillary acidic protein (GFAP) staining, indicating poor differentiation of neoplasm,123 and (2) close proximity to CSF spaces.60,162 Patient risk factors for leptomeningeal dissemination of primary CNS neoplasm include extended survival9 and previous operation (see Fig. 18-41).8,10 In cases of non-CNS neoplasm, prolonged survival120,122,176 and the presence of other metastatic foci16,120,159 increase the probability of meningeal dissemination. Dispersion of non-CNS neoplastic cells is likely a multistep process according to the analysis of the pathways of metastases performed by Viadana and colleagues.163 This “cascade” phenomenon postulates that spread of the primary tumor occurs to at least one intermediate organ, from which it may disseminate widely.163 The observation that vertebral body (especially in breast carcinoma),16,159 bone marrow, and liver metastases (especially in lung carcinoma)120 are commonly present in cases of meningeal carcinomatosis lends credence to this predisposition of tumors to spread first to an intervening organ before disseminating widely. The breast, lung, and stomach, in addition to melanoma, are primary sites of solid tumors that tend to metastasize to the leptomeninges.74,75,122,165 More recently, specific biological subtypes of neoplasms have been identified that indicate increased risk for developing meningeal carcinomatosis over other subtypes. This phenomenon has been particularly well established in breast carcinoma: the “triple negative” (estrogen receptor, progesterone receptor, HER2 receptor negative) and lobular subtypes demonstrate a propensity to leptomeningeal dissemination.118 As a result of increasingly targeted treatment options, breast carcinoma

http://radiologyebook.com

558

PART II  CT and MR Imaging of the Whole Body

A

B FIG 18-35  Meningiomatosis in neurofibromatosis type 2 (NF2). A, Axial CT image demonstrates bilateral soft tissue plaques with calcification (arrow). B, Axial T1-weighted image reveals homogeneously enhancing meningeal plaques extending along the bilateral convexities and falx, which represent multiple meningiomas.

A

B FIG 18-36  Primary leptomeningeal gliomatosis. This rare primary malignant meningeal neoplasm was diagnosed only at autopsy. There is both linear and nodular enhancement in a pia–subarachnoid space pattern (arrows) on a coronal image (A) and a sagittal enhanced T1-weighted image (B). (Courtesy Bernhard C. Sander, MD, and Tibor Mitrovics, MD, Freie Universitatsklinikum Rudolf Virchow, Berlin, Germany.)

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

B

A

FIG 18-37  Primary leptomeningeal melanocytosis. Sagittal T1-weighted images obtained before (A) and after (B) administration of gadolinium show multiple tiny nodular foci of hyperintensity (arrowheads) along the pial surface of the cervical spinal cord, with enhancement in this case of primary leptomeningeal melanocytosis. There is also an intramedullary lesion (curved arrow).

A

B FIG 18-38  Subarachnoid spread of pineoblastoma. A and B, Coronal enhanced T1-weighted images show the pineal region mass with pia–subarachnoid space enhancement (arrows) in this patient with CSF dissemination of pineoblastoma.

http://radiologyebook.com

559

560

PART II  CT and MR Imaging of the Whole Body

tumor subtypes also influence overall survival rates; one study demonstrated improved prognosis for patients with HER2+ tumor types compared with other subtypes, which may have been related to the tumor subtype itself, differences in overall treatment, or be multifactorial.1 Additional tumor characteristics, such as expression of VEGF, also indicate increased risk for leptomeningeal metastases.76 Hematologic neoplasms such as leukemia and lymphoma are also inclined to CSF spread.75

Mechanisms of Meningeal Neoplastic Dissemination Dura

The two major mechanisms of spread of neoplasm to the dura are hematogenous dissemination and direct extension. Hematogenous spread.  One pathway for neoplasm to reach the dura is as a sequela of bone metastases; this route is common in patients with breast carcinoma. Vertebral bone metastases disseminate hematogenously through Batson’s plexus and then to the intracranial dural venous sinuses159 (Fig. 18-42). For example, dural metastases are as likely as parenchymal brain metastases in cases of disseminated breast carcinoma, reflecting the high incidence of vertebral involvement.159 Direct extension.  A second route is dural infiltration, with neoplasm extending directly from vertebral or calvarial lesions2 (Fig. 18-43).

Leptomeninges

FIG 18-39  Disseminated germ cell tumor. This pineal region neoplasm demonstrates ependymal and subarachnoid spread involving the optic chiasm (arrowheads) on an enhanced coronal T1-weighted image.

A

The major mechanisms of seeding into the CSF and leptomeninges are also hematogenous spread and direct extension, but these may take a variety of forms. Hematogenous spread.  There are three main modes by which neoplasm is spread hematogenously to the leptomeninges and CSF. First, the choroid plexus may become seeded hematogenously, resulting in tumor deposits that are then sloughed into the CSF.38,61,74,75 This pathway is presumably more common than is clinically evident. The frequency of choroid metastases is likely to be underestimated on MRI, because the normal choroid plexus is irregular in contour. Metastases occur frequently when the choroid plexus is sectioned at pathologic examination; however, the choroid plexus may not be examined routinely.112,122 The second hematogenous route is via parenchymal blood vessels in the Virchow-Robin (perivascular) spaces to the pia mater and then into the SAS (see Fig. 18-2).57,75 This mechanism of seeding of the

B FIG 18-40  CSF spread of astrocytoma in a 30-year-old patient. Sagittal T1-weighted images obtained before (A) and after (B) gadolinium administration show such extensive enhancement of the subarachnoid space that the enhanced T1-weighted image mimics a T2-weighted sequence. The patient had been treated for astrocytoma 3 years earlier.

http://radiologyebook.com

CHAPTER 18  Meningeal Processes

561

B

A

FIG 18-41  CSF spread of glioblastoma multiforme after biopsy in a 40-year-old patient. Six weeks after stereotactic biopsy, an enhanced coronal T1-weighted image shows the primary right thalamic glioblastoma and extensive thick dissemination of neoplasm in a pia–subarachnoid space pattern.

FIG 18-43  Focal direct extension of breast carcinoma calvarial metastasis to dura. Coronal enhanced T1-weighted image shows focal dural thickening (arrows) and enhancement deep to a calvarial focus of breast carcinoma (see also Fig. 18-20).

FIG 18-42  Diagram of the route of hematogenous spread to dura via Batson’s plexus. (From Fukui MB, et al: MR imaging of the meninges: Part 2. Neoplastic disease. Radiology 201:605–612, 1996.)

http://radiologyebook.com

562

PART II  CT and MR Imaging of the Whole Body

leptomeninges and SAS may result in tumor spread either to or from the SAS.57,75,112 Third, neoplasm may interrupt the thin walls of microscopic vessels in the arachnoid to enter the SAS (see Fig. 18-2).134 This pathway of spread occurs more commonly in leukemia and lymphoma than in solid tumors.134 Direct extension.  Direct extension, the other major means by which neoplasm gains access to the leptomeninges, can also take one of three forms. First, parenchymal neoplasm close to CSF borders, such as cortical or subependymal tumor, may shed cells into the CSF (see Fig. 18-12).38,172 Non-CNS neoplasms may induce a fibrous reaction that tends to prevent CSF spread; thus, this pathway may be more common among tumors of CNS origin.74,122 Second, an overlying dural neoplasm may invade the leptomeninges directly.64,74,75 Third, spinal epidural neoplasm can spread contiguously along nerve roots or their lymphatics.12,16,38,92,122,159 This route of spread is somewhat controversial; some authors assert that the frequency of leptomeningeal carcinomatosis would be greater if this were an important pathway, considering the common occurrence of vertebral metastases.74

Complications of Meningeal Neoplasms Once the CSF has been seeded, neoplasm has the potential to disseminate widely from a focus in the leptomeninges, throughout the arachnoid and pia mater, through the perivascular (Virchow-Robin) spaces, or along the sleeves of cranial nerves and spinal nerve roots.61 Extension into the perivascular spaces or partial obstruction of the vessel lumen by tumor cells can result in ischemic complications of leptomeningeal carcinomatosis122,165 (Fig. 18-44). The term carcinomatous

A

encephalitis refers to the rare occurrence of diffuse perivascular (Virchow-Robin space) infiltration resulting in ischemia.74,102 Neoplasm adjacent to the meninges or a small site of meningeal tumor involvement may also provoke a diffuse fibrous response of the meninges.122 Because both the meningeal reaction to neoplasm and the neoplasm itself usually enhance on MRI, this finding is not specific.

SUMMARY A wide spectrum of nonneoplastic and neoplastic conditions may affect the meninges. MRI with FLAIR sequences or enhanced T1-weighted images is an effective modality for characterizing meningeal pathology. The cardinal MRI findings in meningeal disease are hydrocephalus and smooth or nodular enhancement or signal abnormality of the dura-arachnoid, the pia mater–SAS, and the ependyma of the brain or spine. Two major patterns of disease are identified: a dura-arachnoid pattern that follows along the inner table of the calvarium, and a pia mater–SAS pattern that extends into the sulcal spaces. Nodular disease in particular may suggest neoplasm. Meningeal neoplasm may be clinically silent, and false-negative results on CSF cytologic study are common.8,10,15,122 The addition of MRI to the clinical data and CSF examination may increase detection of meningeal involvement and may allow tumor staging. Understanding the routes of neoplastic spread to the meninges permits identification of imaging features of the primary tumor that predispose to CSF dissemination, such as proximity to the SAS or ventricles. Although there is considerable overlap in the MR appearance of meningeal involvement in various diseases, the pattern and distribution of enhancement may provide guidance for diagnosis, management, and treatment monitoring of disorders affecting the

B FIG 18-44  Carcinomatous encephalitis. A, Axial CT of the head obtained before contrast medium shows hypoattenuation in the posterior parietal lobe (arrowheads) in a 42-year-old patient who presented with a 4-month history of seizures. B, After contrast administration, pial enhancement is minimal (arrowheads). The patient was initially thought to have subarachnoid hemorrhage, but autopsy showed extensive adenocarcinoma in the perivascular spaces and meninges, with multiple ischemic foci in the brain and a primary focus in the lung.

http://radiologyebook.com

CHAPTER 18  Meningeal Processes meninges. Correlation of imaging findings with clinical and CSF data is essential to ensure that treatable disease is not overlooked. Recognition of meningeal neoplasm is increasingly important as treatment options targeted to specific tumor subtypes allow the potential for prolonged survival.

REFERENCES 1. Abouharb S, Ensor J, Loghin ME, et al: Leptomeningeal disease and breast cancer: The importance of tumor subtype. Breast Cancer Res Treat 146:477–486, 2014. 2. Ahmadi J, Hinton D: Dural invasion by craniofacial and calvarial neoplasms: MR imaging and histopathologic evaluation. Radiology 188(3):747–749, 1993. 3. Aizpuru R, Quencer R, Norenberg M, et al: Meningioangiomatosis: Clinical, radiologic, and histopathologic correlation. Radiology 179:819–821, 1991. 4. Alcolado R, Weller R, Parrish E, et al: The cranial arachnoid and   pia mater in man: Anatomical and ultrastructural observations. Neuropathol Appl Neurobiol 14:1–17, 1988. 5. Allcutt D, Michowitz S, Weitzman S, et al: Primary leptomeningeal melanoma: An unusually aggressive tumor in childhood. Neurosurgery 32:721–729, 1993. 6. An YJ, Cho HR, Kim TM, et al: An NMR metabolomics approach   for the diagnosis of leptomeningeal carcinomatosis in lung adenocarcinoma cancer patients. Int J Cancer 136:162–171, 2015. 7. Anonymous: Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 26-1988. A 40-year-old man with a persistent dural abnormality after treatment for an astrocytoma. N Engl J Med 318:1742–1750, 1988. 8. Arita N, Taneda M, Hayakawa T: Leptomeningeal dissemination of malignant gliomas, incidence, diagnosis and outcome. Acta Neurochir 126:84–92, 1994. 9. Awad I, Bay J, Rogers L: Leptomeningeal metastasis from supratentorial malignant gliomas. Neurosurgery 19(2):247–251, 1986. 10. Balhuizen J, Bots G, Schaberg A, et al: Value of cerebrospinal fluid cytology for the diagnosis of malignancies in the central nervous system. J Neurosurg 48:747–753, 1978. 11. Barkhof F, Valk J, Hommes O, et al: Meningeal Gd-DTPA enhancement in multiple sclerosis. AJNR Am J Neuroradiol 13:397–400, 1992. 12. Batson O: The function of the vertebral veins and their role in the spread of metastases. Ann Surg 112(1):138–149, 1940. 13. Benedikt R, Brown D, Walker R, et al: Sturge-Weber syndrome: Cranial MR imaging with Gd-DTPA. AJNR Am J Neuroradiol 14:409–415, 1993. 14. Bickerstaff E, Small J, Guest I: The relapsing course of certain meningiomas in relation to pregnancy and menstruation. J Neurol Neurosurg Psychiatry 21:89–91, 1958. 15. Bigner S, Johnston W: The cytopathology of cerebrospinal fluid. II. Metastatic cancer, meningeal carcinomatosis and primary central nervous system neoplasms. Acta Cytol 25(5):461–479, 1981. 16. Boogerd W, Hart A, van der Sande J, et al: Meningeal carcinomatosis in breast cancer. Prognostic factors and influence of treatment. Cancer 67(6):1685–1695, 1991. 17. Bourekas E, Wildenhain P, Lewin J, et al: The dural tail sign revisited. AJNR Am J Neuroradiol 16:1514–1516, 1995. 18. Bowen B, Donovan Post M: Intracranial infection. In Atlas S, editor: Magnetic resonance imaging of the brain and spine, New York, 1991, Raven Press, pp 501–538. 19. Bracchi M, Savoiardo M, Triulzi F, et al: Superficial siderosis of the CNS: MR diagnosis and clinical findings. AJNR Am J Neuroradiol 14:227–236, 1993. 20. Brightbill T, Ihmeidan I, Donovan Post M, et al: Neurosyphilis is HIV-positive and HIV-negative patients: Neuroimaging findings.   AJNR Am J Neuroradiol 16:703–711, 1995. 21. Brodal P: The central nervous system, New York., 1992, Oxford University Press. Vol.

563

22. Bromberg JE, Breems DA, Kraan J, et al: CSF flow cytometry greatly improves diagnostic accuracy in CNS hematologic malignancies. Neurology 68:1674–1679, 2007. 23. Burke J, Podrasky A, Bradley W: Meninges: Benign postoperative enhancement on MR images. Radiology 174:99–102, 1990. 24. Carpenter AB: Meninges and cerebrospinal fluid. In Wilkins W, editor: Core text of neuroanatomy, ed 2, Baltimore, 1978, Williams & Wilkins, pp 1–14. 25. Castelli M, Husband A: Superficial siderosis of the central nervous system: An underestimated cause of hearing loss. J Laryngol Otol 111:60–62, 1997. 26. Chamberlain M: Current concepts in leptomeningeal metastasis. Curr Opin Oncol 4:533–539, 1992. 27. Chamberlain M, Sandy A, Press G: Leptomeningeal metastasis:   A comparison of gadolinium-enhanced MR and contrast-enhanced CT of the brain. Neurology 40:435–438, 1990. 28. Chang K, Han M, Roh J, et al: Gd-DTPA-enhanced MR imaging of the brain in patients with meningitis: Comparison with CT. AJR Am J Roentgenol 11:69–76, 1990. 29. Cheng T, O’Neill B, Scheithauer B, et al: Chronic meningitis: The role of meningeal or cortical biopsy. Neurosurgery 34:590–596, 1994. 30. Cohen J, Quint D, Eldevik O: Patterns of normal meningeal enhancement on 1.5-T MR imaging. American Roentgen Ray Society Proceedings Book:127, 1995. 31. Cooper I, Craig M, Kernohan J: Tumors of the spinal cord. Primary extramedullary gliomas. Surg Gynecol Obstet 92:183–190, 1951. 32. Crawford J, Konkol R: Familial hemiplegic migraine with crossed cerebellar diaschisis and unilateral meningeal enhancement. Headache 37:590–593, 1997. 33. Davis P, Friedman N, Fry S, et al: Leptomeningeal metastasis: MR imaging. Radiology 163:449–454, 1987. 34. De Klerk D, Spence J: Chemical meningitis with intracranial tumours.   S Afr Med J 48:131–135, 1974. 35. Demaerel P, Robberecht W, Casteels I, et al: Focal leptomeningeal MR enhancement along the chiasm as a presenting sign of multiple sclerosis. J Comput Assist Tomogr 19(2):297–298, 1995. 36. Demaerel P, Wilms G, Van Lierde S, et al: Lyme disease in childhood presenting as primary leptomeningeal enhancement without parenchymal findings of MR. AJNR Am J Neuroradiol 15:302–304, 1994. 37. Dietrich P-Y, Aapro M, Pizzolato G: Primary diffuse leptomeningeal gliomatosis (PDLG): A neoplastic cause of chronic meningitis.   J Neurooncol 15:275–283, 1993. 38. Dinsdale H, Taghavy A: Carcinomatosis of the meninges. Can Med Assoc J 90:505–512, 1964. 39. Elster A: Magnetic resonance contrast enhancement in cerebral infarction. Neuroimaging Clin N Am 4:89–100, 1994. 40. Elster A, DiPersio D: Cranial postoperative site: Assessment with contrast-enhanced MR imaging. Radiology 174:93–98, 1990. 41. Elster A, Mathews V, King J, et al: Improved detection of gadolinium enhancement using magnetization transfer imaging. Neuroimaging Clin N Am 4:185–399, 1994. 42. Elster A, Moody D: Early cerebral infarction: Gadopentetate dimeglumine enhancement. Radiology 177:627–632, 1990. 43. Enzmann D, Krikorian J, Yorke C, et al: Computed tomography in leptomeningeal spread of tumor. J Comput Assist Tomogr 2:448–455, 1978. 44. Esiri M, Gay D: Immunological and neuropathological significance of the Virchow-Robin space. J Neurol Sci 100:3–8, 1990. 45. Eustace S, Buff B: Magnetic resonance imaging in drug-induced meningitis. Can Assoc Radiol J 45:463–465, 1994. 46. Farn J, Mirowitz S: MR imaging of the normal meninges: Comparison of contrast-enhancement patterns on 3D gradient-echo and spin-echo images. AJR Am J Roentgenol 162:131–135, 1994. 47. Fishman R, Dillon W: Dural enhancement and cerebral displacement secondary to intracranial hypotension. Neurology 43:609–611, 1993. 48. Frank J, Girton M, Dwyer A, et al: Meningeal carcinomatosis in the VX2 rabbit tumor model: Detection with Gd-DPTA-enhanced MR imaging. Radiology 167(3):825–829, 1988.

http://radiologyebook.com

564

PART II  CT and MR Imaging of the Whole Body

49. Frederickson R: The subdural space interpreted as a cellular layer of meninges. Anat Rec 230:38–51, 1991. 50. Freilich R, Krol G, DeAngelis L: Neuroimaging and cerebrospinal fluid cytology in the diagnosis of leptomeningeal metastasis. Ann Neurol 38:51–57, 1995. 51. Fukui M, Williams R, Mudigonda S: CT and MR imaging signs of pyogenic ventriculitis. Proceedings of the American Society of Neuroradiology, 1999. 52. Fulbright R, Erdum E, Sze G, et al: Cranial nerve enhancement in   the Guillain-Barre syndrome. AJNR Am J Neuroradiol 16:923–925, 1995. 53. Giaquinto S, Massi G, Ricolfi A, et al: On six cases of radiation meningiomas from the same community. Ital J Neurol Sci 5:173–175, 1984. 54. Gilden D, Beinlich B, Rubenstien E, et al: Varicella-zoster virus myelitis: An expanding spectrum. Neurology 44:1818–1823, 1994. 55. Ginsberg L, Leeds N: Neuroradiology of leukemia. AJR Am J Roentgenol 165:525–534, 1995. 56. Glass J, Melamed M, Chernik N, et al: Malignant cells in cerebrospinal fluid (CSF): The meaning of a positive CSF cytology. Neurology 29:1369–1375, 1979. 57. Globus J, Meltzer T: Metastatic tumors of the brain. Arch Neurol Psychiatry 48:163–226, 1942. 58. Goldsher D, Litt A, Pinto R, et al: Dural “tail” associated with meningiomas on Gd-DTPA-enhanced MR images: Characteristics, differential diagnostic value, and possible implications for treatment. Radiology 176:447–450, 1990. 59. Gondos B, King E: Cerebrospinal fluid cytology: Diagnostic accuracy and comparison of different techniques. Acta Cytol 20:542–547, 1976. 60. Grabb P, Albright A, Pang D: Dissemination of supratentorial malignant gliomas via the cerebrospinal fluid in children. Neurosurgery 30:64–71, 1992. 61. Grain G, Karr J: Diffuse leptomeningeal carcinomatosis. Clinical and pathologic characteristics. Neurology 5:706–722, 1955. 62. Gray L, MacFall J, Provenzale J, et al: High dose MR contrast agent   for the diagnosis of leptomeningeal disease. Radiology 197(P):410, 1995. 63. Greenberg R, Lane E, Cinnamon J, et al: The cranial meninges: Anatomic considerations. Semin Ultrasound CT MR 15(6):454–465, 1994. 64. Griffin J, Thompson R, Mitchinson M, et al: Lymphomatous leptomeningitis. Am J Med 51:200–208, 1971. 65. Grossman R, Kemp S, Yu I, et al: The importance of oxygenation in the appearance of acute subarachnoid hemorrhage on high-field magnetic resonance imaging. Acta Radiol Suppl 369:56–58, 1986. 66. Guseo A, Jellinger K: The significance of perivascular infiltrations in multiple sclerosis. J Neurol 211:51–60, 1975. 67. Haines D: On the question of a subdural space. Anat Rec 230:3–21, 1991. 68. Halper J, Scheithauer B, Okazaki H, et al: Meningio-angiomatosis:   A report of six cases with special reference to the occurrence of neurofibrillary tangles. J Neuropathol Exp Neurol 45:426–446, 1986. 69. Hammes E: Reaction of the meninges to blood. Arch Neurol Psychiatry 52:505–514, 1944. 70. Hardman J: Cerebrospinal trauma. In Davis R, Robertson D, editors: Textbook of neuropathology, ed 2, Baltimore, 1991, Williams & Wilkins, pp 962–1002. 71. Harvey S, Burr H: The development of the meninges. Arch Neurol Psychiatry 15:545–567, 1926. 72. Haustein J, Laniado M, Niendorf H-P, et al: Triple-dose versus standard-dose gadopentetate dimeglumine: A randomized study in 199 patients. Radiology 186:855–860, 1993. 73. Heinz R, Wiener D, Friedman H, et al: Detection of cerebrospinal fluid metastasis: CT myelography or MR? AJNR Am J Neuroradiol 16:1147– 1151, 1995. 74. Henson R, Urich H: Carcinomatous meningitis. In Henson R, Urich H, editors: Cancer and the nervous system, Boston, 1982, Blackwell Scientific Publications, pp 101–119.

75. Henson R, Urich H: Diffuse infiltration by lymphoma and leukemia. In Henson R, Urich H, editors: Cancer and the central nervous system, Boston, 1982, Blackwell Scientific Publications, pp 227–267. 76. Herrlinger U, Wiendl H, Renninger M, et al: Vascular endothelial growth factor (VEGF) in leptomeningeal metastasis: Diagnostic and prognostic value. Br J Cancer 91:219–224, 2004. 77. Ho K-L, Hoschner J, Wolfe D: Primary leptomeningeal gliomatosis. Symptoms suggestive of meningitis. Arch Neurol 38:662–666, 1981. 78. Hochman M, Naidich T, Kobetz S, et al: Spontaneous intracranial hypotension with pachymeningeal enhancement on MRI. Neurology 42:1628–1630, 1992. 79. Hudgins P, Davis P, Hoffman J, Jr: Gadopentetate dimeglumineenhanced MR imaging in children following surgery for brain tumor: Spectrum of meningeal findings. AJNR Am J Neuroradiol 12:301–307, 1991. 80. Hughes J, Oppenheimer D: Superficial siderosis of the central nervous system. Acta Neuropathol (Berl) 13:56–74, 1969. 81. Jacobs D, McFarlane M, Holmes F: Female patients with meningioma of the sphenoid ridge and additional primary neoplasms of the breast and genital tract. Cancer 60:3080–3082, 1987. 82. Jaeckle K, Krol G, Posner J: Evolution of computed tomographic abnormalities in leptomeningeal metastases. Ann Neurol 17(1):85–89, 1985. 83. Janss A, Galetta S, Freese A, et al: Superficial siderosis of the central nervous system: Magnetic resonance imaging and pathological correlation. J Neurosurg 79:756–760, 1993. 84. Jennings M, Slatkin N, D’Angelo M, et al: Neoplastic meningitis as the presentation of occult primitive neuroectodermal tumors. J Child Neurol 8:306–312, 1993. 85. Johnson B, Fram E: Cerebral venous occlusive disease. Pathophysiology, clinical manifestations, and imaging. Neuroimaging Clin N Am 2:769–783, 1992. 86. Kallmes D, Gray L, Glass J: High-dose gadolinium-enhanced MRI for diagnosis of meningeal metastasis. Neuroradiology 40:23–26, 1998. 87. Kelly A, Zimmerman R: Head trauma: comparison of MR and CT–Experience in 100 patients. AJNR Am J Neuroradiol 9:699–708, 1988. 88. Kilgore D, Breger R, Daniels D, et al: Cranial tissues: Normal MR appearance after intravenous injection of Gd-DTPA. Radiology 160:757–761, 1986. 89. Kioumehr F, Dadsetan M, Feldman N, et al: Postcontrast MRI of the cranial meninges: Leptomeningitis versus pachymeningitis. J Comput Assist Tomogr 19:713–720, 1995. 90. Kirkpatrick J: Neurologic infections due to bacteria, fungi, and parasites. In Davis R, Robertson D, editors: Textbook of neuropathology, ed 2, Baltimore, 1991, Williams & Wilkins, pp 719–803. 91. Kline T: Cytological examination of the cerebrospinal fluid. Cancer 15:591–597, 1962. 92. Kokkoris C: Leptomeningeal carcinomatosis. How does cancer reach the pia-arachnoid? Cancer 51:154–160, 1983. 93. Kramer E, Rafto S, Packer R, et al: Comparison of myelography with CT follow-up versus gadolinium MRI for subarachnoid metastatic disease in children. Neurology 41:46–50, 1991. 94. Krol G, Sze G, Malkin M, et al: MR of cranial and spinal meningeal carcinomatosis: Comparison with CT and myelography. AJR Am J Roentgenol 151:583–588, 1988. 95. Kumar A, Montanera W, Willinsky R, et al: Case report. MR features of tuberculous arachnoiditis. J Comput Assist Tomogr 17(1):127–130, 1993. 96. Lachance D, O’Neill B, Macdonald D, et al: Primary leptomeningeal lymphoma: Report of 9 cases, diagnosis with immunocytochemical analysis, and review of literature. Neurology 41:95–100, 1991. 97. Lee Y-Y, Van Tassel P, Raymond A: Intracranial dural chondrosarcoma. AJNR Am J Neuroradiol 9:1189–1193, 1988. 98. Leproux F, Melanson D, Mercier C, et al: Leptomeningeal gliomatosis: MR findings. J Comput Assist Tomogr 17(2):317–320, 1993. 99. Le Rhun E, Kramar A, Salingue S, et al: CSF CA 15-3 in breast cancer-related leptomeningeal metastases. J Neurooncol 117:117–124, 2014.

http://radiologyebook.com

CHAPTER 18  Meningeal Processes 100. Lindner A, Reiners K, Toyka K: Meningeal hyperperfusion visualized by MRI in a patient with visual hallucinations and migraine. Headache 36:53–57, 1996. 101. Lunardi P, Missori P, Rizzo A, et al: Chemical meningitis in ruptured intracranial dermoid. Case report and review of the literature. Surg Neurol 32:449–452, 1989. 102. Madow L, Alpers B: Encephalitic form of metastatic carcinoma. Arch Neurol Psychiatry 65:161–173, 1951. 103. Malat J, Virapongse C, Palestro C, et al: Primary intraspinal fibrosarcoma. Neurosurgery 19:434–436, 1986. 104. Mamelak A, Kelly W, Davis R, et al: Idiopathic hypertrophic cranial pachymeningitis. Report of three cases. J Neurosurg 79:270–276, 1993. 105. Mark A: Nondegenerative, non-neoplastic diseases of the spine and spinal cord. In Atlas S, editor: Magnetic resonance imaging of the brain and spine, New York, 1991, Raven Press, pp 967–1011. 106. Martin N, Masson C, Henin D, et al: Hypertrophic cranial pachymeningitis: Assessment with CT and MR imaging. AJNR Am J Neuroradiol 10:477–484, 1989. 107. Mathews VP, Alo PL, Glass JD, et al: AIDS-related CNS Cryptococcus: Radiologic-pathologic correlation. AJNR Am J Neuroradiol 13:1477– 1486, 1992. 108. Mathews V, Kuharik M, Edwards M, et al: Gd-DTPA-enhanced MR imaging of experimental bacterial meningitis: Evaluation and comparison with CT. AJR Am J Roentgenol 152:131–136, 1989. 109. Meltzer C, Smirniotopoulos J, Fukui M: The dural tail. Int J Neuroradiol 4:33–40, 1998. 110. Miller J, Jubelt B: Bacterial infections. In Rowland L, editor: Merritt’s textbook of neurology, ed 8, Philadelphia, 1989, Lea & Febiger, pp 63–96. 111. Mittl R, Yousem D: Frequency of unexplained meningeal enhancement in the brain after lumbar puncture. AJNR Am J Neuroradiol 15:633–638, 1994. 112. Moberg A, Reis G: Carcinosis meningum. Acta Med Scand 170:747–755, 1961. 113. Murai H, Kira J, Kobayashi T, et al: Hypertrophic cranial pachymeningitis due to Aspergillus flavus. Clin Neurol Neurosurg 94:247–250, 1992. 114. Naul L, Hise J, Bauserman S, et al: CT and MR of meningeal melanocytoma. AJNR Am J Neuroradiol 12:315–316, 1991. 115. Naylor B: The cytologic diagnosis of cerebrospinal fluid. Acta Cytol 8:141–148, 1964. 116. Negishi C, Sze G: Vasculitis presenting as primary leptomeningeal enhancement with minimal parenchymal findings. AJNR Am J Neuroradiol 14:26–28, 1993. 117. Newton H, Fleisher M, Schwartz M, et al: Glucose phosphate isomerase as a CSF marker for leptomeningeal metastasis. Neurology 41:395–398, 1991. 118. Niwinska A, Rudnicka H, Murawska M: Breast cancer leptomeningeal metastasis: Propensity of breast cancer subtypes for leptomeninges and the analysis of factors influencing survival. Med Oncol 30:408, 2013. 119. Noguchi K, Ogawa T, Inugami A, et al: MR of acute subarachnoid hemorrhage: A preliminary report of fluid-attenuated inversionrecovery pulse sequences. AJNR Am J Neuroradiol 15:1940–1943, 1994. 120. Nugent J, Bunn P, Matthews M, et al: CNS metastases in small cell bronchogenic carcinoma. Cancer 44:1885–1893, 1979. 121. Ogawa T, Inugami A, Fujita H, et al: MR diagnosis of subacute and chronic subarachnoid hemorrhage: Comparison with CT. AJR Am J Roentgenol 165:1257–1262, 1995. 122. Olson M, Chernik N, Posner J: Infiltration of the leptomeninges by systemic cancer. Arch Neurol 30:122–137, 1974. 123. Onda K, Tanaka R, Takahashi H, et al: Cerebral glioblastoma with cerebrospinal fluid dissemination: A clinicopathological study of 14 cases examined by complete autopsy. Neurosurgery 25(4):533–540, 1989. 124. Onda K, Tanaka R, Takahashi H, et al: Symptomatic cerebrospinal fluid dissemination of cerebral glioblastoma. Neuroradiology 32:146–150, 1990. 125. O’Rahilly R, Muller F: The meninges in human development.   J Neuropathol Exp Neurol 45:588–608, 1986.

565

126. Osaka K, Handa H, Matsumoto S, et al: Development of the cerebrospinal fluid pathway in the normal and abnormal human embryos. Childs Brain 6:26–38, 1980. 127. Paako E, Patronas N, Schellinger D: Meningeal Gd-DTPA enhancement in patients with malignancies. J Comput Assist Tomogr 14:542–546, 1990. 128. Packer R, Siegel K, Sutton L, et al: Leptomeningeal dissemination of primary central nervous system tumors of childhood. Ann Neurol 18:217–221, 1985. 129. Park J, Van den Noort S, Kim R, et al: Primary diffuse leptomeningeal gliomatosis with signs of increased intracranial pressure and progressive meningeal enhancement on MRI. J Neuroimaging 6:250–254, 1996. 130. Partington C, Graves V, Hegstrand L: Meningioangiomatosis. AJNR Am J Neuroradiol 12:549–552, 1991. 131. Pfluger T, Weil S, Weis S, et al: MRI of primary meningeal sarcomas in two children: Differential diagnostic considerations. Neuroradiology 39:225–228, 1997. 132. Phillips M, Ryals T, Kambhu S, et al: Neoplastic vs inflammatory meningeal enhancement with Gd-DTPA. J Comput Assist Tomogr 14(4):536–541, 1990. 133. Praharaj S, Sharma M, Prasad K, et al: Unilateral meningeal thickening: A rare presentation of tuberculous meningitis. Clin Neurol Neurosurg 99:60–62, 1997. 134. Price R, Johnson W: The central nervous system in childhood leukemia: I. The arachnoid. Cancer 31:520–533, 1973. 135. Provenzale J, Allen N: MR findings in Wegener’s granulomatosis. ASNR Proceedings Book:130, 1995. 136. Pui M, Langston J, Arai Y: Gd-DTPA Enhancement of CSF in meningeal carcinomatosis. J Comput Assist Tomogr 17:940–944, 1993. 137. Quest D, Salcman M: Fibromatosis presenting as a cranial mass lesion.   J Neurosurg 44:237–240, 1976. 138. Rando T, Fishman R: Spontaneous intracranial hypotension: Report of two cases and review of the literature. Neurology 42:481–487, 1992. 139. Runge V, Clanton J, Price A, et al: Dyke Award. Evaluation of contrastenhanced MR imaging in a brain-abscess model. AJNR Am J Neuroradiol 6(2):139–147, 1985. 140. Runge V, Wells J, Williams N, et al: Detectability of early brain meningitis on MR images with pathologic correlation. Radiology 197P:480, 1995. 141. Russell D, Rubinstein L: Pathology of tumours of the nervous system, ed 5, London, 1989, Edward Arnold. 142. Sage M: Blood-brain barrier: Phenomenon of increasing importance to the imaging clinician. Am J Neuroradiol 3:127–138, 1982. 143. Sage M, Wilson A: The blood-brain barrier: An important concept in neuroimaging. AJNR Am J Neuroradiol 15:601–622, 1994. 144. Sakamoto S, Kitagaki H, Ishii K, et al: Gadolinium enhancement   of the cerebrospinal fluid in a patient with meningeal fibrosis and cryptococcal infection. Neuroradiology 39:504–505, 1997. 145. Sarnat H, Netsky M: Evolution of the nervous system, ed 2, New York., 1981, Oxford University Press. 146. Seltzer S, Mark A, Atlas S: CNS sarcoidosis: Evaluation with contrastenhanced MR imaging. AJNR Am J Neuroradiol 12:1227–1233, 1991. 147. Sherman J, Stern B: Sarcoidosis of the CNS: Comparison of unenhanced and enhanced MR images. AJNR Am J Neuroradiol 11:915–923, 1990. 148. Singer M, Atlas S, Drayer B: Subarachnoid space disease: Diagnosis with fluid-attenuated inversion-recovery MR imaging and comparison with gadolinium-enhanced spin-echo MR imaging—Blinded reader study. Radiology 208:417–422, 1998. 149. Smirniotopoulos J, Murphy F: The phakomatoses. AJNR Am J Neuroradiol 13:725–746, 1992. 150. Southwick F, Richardson E, Swartz M: Septic thrombosis of the dural venous sinuses. Medicine 65:82–106, 1986. 151. Spagnoli M, Goldberg H, Grossman R, et al: Intracranial meningiomas: High-field MR imaging. Radiology 161:369–375, 1986. 152. Spallone A, Gagliardi F, Vagnozzi R: Intracranial meningiomas related to external cranial radiation. Surg Neurol 12:153–159, 1979.

http://radiologyebook.com

566

PART II  CT and MR Imaging of the Whole Body

153. Stern B, Krunholz A, Johns C, et al: Sarcoidosis and its neurological manifestations. Arch Neurol 42:909–917, 1985. 154. Sze G: Diseases of the intracranial meninges: MR imaging features.   AJR Am J Roentgenol 160:727–733, 1993. 155. Sze G, Soletsky S, Bronen R, et al: MR imaging of the cranial meninges with emphasis on contrast enhancement and meningeal carcinomatosis. AJNR Am J Neuroradiol 10:965–975, 1989. 156. Tishler S, Williamson T, Mirra S, et al: Wegener granulomatosis with meningeal involvement. AJNR Am J Neuroradiol 14:1248–1252, 1993. 157. Tokumaru A, O’uchi T, Tsuneyoshi E, et al: Prominent meningeal enhancement adjacent to meningioma on Gd-DTPA-enhanced MR images: Histopathologic correlation. Radiology 175:431–433, 1990. 158. Truwit C, Barkovich A: Pathogenesis of intracranial lipoma: An MR study in 42 patients. AJR Am J Roentgenol 155:855–864, 1990. 159. Tsukada Y, Fouad A, Pickren J, et al: Central nervous system metastasis from breast carcinoma. Autopsy study. Cancer 52(12):2349–2354, 1993. 160. Twijnstra A, Van Zanten A, Nooyen W, et al: Cerebrospinal fluid beta2-microglobulin: A study in controls and patients with metastatic and non-metastatic neurological disease. Eur J Cancer Clin Oncol 22:387–391, 1986. 161. van de Langerijt B, Gijtenbeek JM, de Reus HP, et al: CSF levels of growth factors and plasminogen activators in leptomeningeal metastases. Neurology 67:114–119, 2006. 162. Vertosick F, Selker R: Brainstem and spinal metastases of supratentorial glioblastoma muliforme: A clinical series. Neurosurgery 27:516–522, 1990. 163. Viadana E, Bross I, Pickren JW: Cascade spread of blood-borne metastases in solid and nonsolid cancers of humans. In Weiss L,   Gilbert HA, editors: Pulmonary metastasis, Boston, 1978, GK Hall, pp 142–167. 164. Vogl T, Stemmler J, Bergman C, et al: MR and MR angiography of Sturge-Weber syndrome. AJNR Am J Neuroradiol 14:417–425, 1993. 165. Wasserstrom W, Glass J, Posner J: Diagnosis and treatment of leptomeningeal metastases from solid tumors: Experience with 90 patients. Cancer 49:759–772, 1982. 166. Watanabe M, Tanaka R, Takeda N: Correlation of MRI and clinical features in meningeal carcinomatosis. Neuroradiology 35:512–515, 1993.

167. Wehn S, Heinz E, Burger P, et al: Dilated Virchow-Robin spaces in cryptococcal meningitis associated with AIDS: CT and MR findings.   J Comput Assist Tomogr 13:756–762, 1989. 168. Whiteman M, Dandapani B, Shebert R, et al: MRI of AIDS-related polyradiculomyelitis. J Comput Assist Tomogr 18:7–11, 1994. 169. Whiteman M, Espinoza L, Post M, et al: Central nervous system tuberculosis in HIV-infected patients: Clinical and radiographic findings. AJNR Am J Neuroradiol 16:1319–1327, 1995. 170. Wiebe S, Munoz D, Smith S, et al: Meningioangiomatosis. A comprehensive analysis of clinical and laboratory features. Brain 122:709–726, 1999. 171. Williams R, Fukui M, Tishkoff N, et al: MR imaging of the subarachnoid space utilizing FLAIR: Pathology and pitfalls. Radiology 213(P):552, 1999. 172. Willis R: Pathology of tumours, ed 4, New York, 1967, AppletonCentury-Crofts. 173. Witham T, Fukui M, Meltzer C, et al: Survival using intrathecal thiotriethylenephosphoramide (thio-TEPA) for the treatment of ependymal or leptomeningeal gliomatosis in patients with high grade gliomas. Cancer 86:1347–1353, 1999. 174. Wolbach S: Congenital rhabdomyoma of the heart: Report of a case associated with multiple nests of neuroglial tissue in the meninges of the spinal cord. J Med Res 16:495–519, 1907. 175. Wrobel C, Meyer S, Johnson R, et al: MR findings in acute and chronic coccidioidomycosis meningitis. AJNR Am J Neuroradiol 13:1241–1245, 1992. 176. Yap H-Y, Yap B-S, Tashima C, et al: Meningeal carcinomatosis in breast cancer. Cancer 42:283–286, 1978. 177. Yousem D, Patrone P, Grossman R: Leptomeningeal metastases:   MR evaluation. J Comput Assist Tomogr 14:255–261, 1990. 178. Zeiser R, Burger JA, Bley TA, et al: Clinical follow-up indicates differential accuracy of magnetic resonance imaging and immunocytology of the cerebral spinal fluid for the diagnosis of neoplastic meningitis—A single centre experience. Br J Haematol 124:762–768, 2004. 179. Zouaoui A, Maillard J-C, Dormont D, et al: MRI in neurosarcoidosis.   J Neuroradiol 19:271–284, 1992.

http://radiologyebook.com

19  Demyelinating Disease and Leukoencephalopathies John P. Nazarian, Leo Wolansky, Amit Gupta, and Michael Coffey

INTRODUCTION The white matter of the central nervous system (CNS) is made up of myelinated neuronal axons. Myelin is a dielectric material composed in large part of lipids; it allows for rapid propagation of nerve impulses and is critical for normal development and functioning of the nervous system. Myelination begins early in the second trimester of fetal development and is largely complete by the end of the second year of postnatal life, although some myelination continues into adolescence. Destruction of myelin, or demyelination, is the pathologic hallmark of several neurologic diseases. The prototypical demyelinating disease is multiple sclerosis, one of the most common causes of neurologic disability in young adults. Advances in imaging, particularly magnetic resonance imaging (MRI), have allowed radiologists to play an increasing role in both the initial diagnosis and treatment monitoring of patients with this disease. Other demyelinating diseases, including acute disseminated encephalomyelitis and progressive multifocal leukoencephalopathy, are seen in association with infections and immunologic phenomena. MRI plays an important role in diagnosing these diseases and distinguishing them from other conditions. The disease entities in this chapter are broadly grouped by etiology. In addition to the demyelinating diseases mentioned above, this chapter also covers several other white matter diseases, or leukoencephalopathies. These diverse pathologic processes have different causes but share in common their propensity to cause white matter abnormalities often visible on diagnostic imaging studies. A basic understanding of their imaging characteristics, when combined with  a thorough clinical workup, will facilitate accurate diagnosis and appropriate management. Prior to any discussion of imaging of white matter diseases, an important caveat must be given: not all computed tomography (CT) hypodensities and T2/FLAIR (fluid-attenuated inversion recovery) hyperintense foci in the white matter indicate demyelinating disease. White matter signal changes on MRI are very common in the elderly and increase in prevalence with age. These so-called unspecific bright objects (UBOs) are not without clinical relevance in elderly patients; they have been associated with an increased risk of cognitive decline, dementia, and stroke.72 Other causes of T2-hyperintense white matter lesions include migraines and sequelae of cerebrovascular disease.29

MULTIPLE SCLEROSIS Multiple sclerosis (MS) is the most common demyelinating disease of the CNS. MS usually presents between the ages of 20 and 50 and is three times more common in women than in men.27,48 The hallmark of the disease is multiplicity in time and space.59 In most cases, patients present with a subacute clinically isolated syndrome;  common presentations include optic neuritis, motor and sensory 

deficits, transverse myelitis, and brainstem syndromes.47,52 The McDonald criteria, revised in 2010, allow for diagnosis of MS based on various clinical and paraclinical assessments, including MRI.54 Several MS subtypes (clinical phenotypes) have been defined on the basis of their clinical course; these definitions were most recently revised in 2013 and play an important role in disease prognostication and design of clinical trials.40 The pathophysiology of MS is complex and not fully understood but is likely autoimmune in nature. The main pathologic characteristic of the disease is the demyelinated plaque, a lesion induced by a variety of immune mechanisms.34 Plaques are responsible for the characteristic hyperintense white matter abnormalities visible on conventional fluid-sensitive MR sequences.18 In addition, more advanced MR techniques have permitted detection of other abnormalities in the so-called normal-appearing white matter (NAWM)62 and gray matter that also occur in MS.15,23 MRI is the imaging method of choice in the diagnosis and monitoring of MS, owing to its high sensitivity and specificity44 as well as its ability to detect or exclude possible alternative conditions. CT is limited in use where MRI is available. The MRI criteria integrated into the McDonald criteria have demonstrated a high specificity for subsequent development of MS in patients who present with a clinically isolated syndrome suggestive of MS.12 In such patients the MRI protocol should always include T2-weighted and FLAIR sequences, preferably with three-dimensional (3D)-Fourier transform (volumetric) to minimize partial volume averaging in the commonly small lesions. MS lesions are seen in characteristic locations: where the white matter contacts the cerebrospinal fluid (CSF). Periventricular, optic nerve, spinal cord, and superficial brainstem lesions are typical, including the cerebellar peduncles, near the trigeminal root entry zone, the tectum, and fourth ventricular floor. Corpus callosum lesions are also characteristic and are much more common in MS than in other white matter diseases.29,63 MS lesions have a characteristic size and shape. On T2-weighted and FLAIR images, demyelinating plaques in the brain appear as highintensity lesions that are often 10 to 15 mm in size and ovoid, although this shape disappears in more advanced cases when lesions become confluent (Fig. 19-1). Periventricular lesions tend to be oriented with their major axis perpendicular to the ependymal surface, referred to as Dawson’s fingers (Fig. 19-2). Periventricular and juxtacortical lesions are better delineated on FLAIR images.19 T1-weighted images are generally less sensitive for detection of plaques, but a subset of T2-hyperintense lesions appear dark on T1 images. These “black holes” have signal characteristics ranging from mildly to severely hypointense; the degree of hypointensity correlates with the degree of pathologic severity.19,71 When persistent, these lesions are associated with demyelination and axonal loss and have been used in clinical trials as markers of accumulated disease burden.18 Degree of T1-hypointensity is also influenced by scanning technique (Fig. 19-3).

http://radiologyebook.com

567

568

PART II  CT and MR Imaging of the Whole Body

A

FIG 19-1  FLAIR transaxial image cropped for the right cerebral hemisphere. Characteristic ovoid lesions are commonly 10 to 15 mm in size.

B FIG 19-2  Sagittal MRI of two different patients with MS. T2 (A) and FLAIR images (B). Multiple ovoid lesions with the long axis perpendicular to the ventricular surface, “Dawson’s fingers.” (B from Bakshi R, et al: The use of magnetic resonance imaging in the diagnosis and long-term management of multiple sclerosis. Neurology 63[Suppl 5]: S3–S11, 2004.)

A

B

C

FIG 19-3  Transaxial T1-weighted images of the right cerebral hemisphere. T1-weighted (A), T1 with fat saturation and Gd (B), and T1 with fat saturation without Gd (C). Noncontrast T1 reveals the presence of a white matter hypointense lesion (arrowhead in A), which on contrast-enhanced image appears to enhance (B). Noncontrast fat-suppressed T1-weighted image reveals that the hyperintensity (arrow in C) is due to fat suppression and not contrast administration due to magnetization transfer effect.

http://radiologyebook.com

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies

A

569

B

FIG 19-4  Contrast-enhanced images with standard 0.1 mmol/kg Gd dose (A) and cumulative 0.3 mmol/kg dose (B). Triple-dose image (arrow in B) reveals an enhancing lesion that was not seen with single dose (arrow in A).

A

B

FIG 19-5  Axial T1 with Gd (A) and FLAIR (B). The size of the ring-enhancing lesion of left middle cerebellar peduncle is seen to match the size of FLAIR hyperintensity—unlike most ring lesions, for which the area of perifocal edema on FLAIR is significantly larger than the enhancing focus.

In addition, lesion enhancement is considered an excellent surrogate for disease activity after a study revealed contrast enhancement was associated with active inflammation in 40 of 40 pathologically proven cases.51 Triple-dose (0.3 mmol/kg) delayed imaging (20-40 minutes) has shown an increase of more than two times the enhancing lesion detection rate (Fig. 19-4) but is not standard, given concerns about nephrogenic systemic fibrosis.76,77 Enhancement of lesions may be homogeneous or ringlike.56 A differential feature from ring lesions of many other etiologies is  the margin of enhancement coinciding with the margin of FLAIR signal rather than extending significantly beyond the enhancing border (Fig. 19-5).

Most lesions are up to 1.5 cm in diameter, but in rare cases larger ring-enhancing lesions may be seen with obvious mass effect (tumefactive demyelination) (Fig. 19-6). A characteristic feature is an openring appearance thought to represent a front of demyelination. In tumefactive cases, perfusion imaging can be of value because ringenhancing malignant neoplasms typically display elevation of relative cerebral blood volume (rCBV) and rCBV ratio, whereas demyelination does not (Fig. 19-7). The “life cycle” of an MS lesion typically includes a short enhancing phase, the majority enhancing for no more than 8 weeks. Oftentimes the enhancement of each lesion appears independent of the others (Fig. 19-8).

http://radiologyebook.com

A

C

B

FIG 19-6  Axial T1 images; baseline without Gd (A) and with Gd (B). C, Follow-up several months later (with Gd). Noncontrast image (A) displays obvious edema with sulcal effacement. Postcontrast image (B) reveals nonspecific-appearing tumefactive ring-enhancing lesion. Follow-up (C) demonstrates complete regression.

A

D

C

B

E

F

FIG 19-7  Three lesions in same patient with perfusion images (A, C, E), with corresponding Gd-T1 images (B, D, F). Each lesion is larger than typical MS plaque but demonstrates open ring of enhancement and low relative cerebral blood volume (rCBV) (light blue open-ring in A, C, E). Ratio of lesion to control brain was http://radiologyebook.com approximately equal to 1. On follow-up, the lesions resolved.

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies

571

13 10

5

9 11

A

B

C

D

E

F

FIG 19-8  Gd-T1 on 6 consecutive months in same patient. Several enhancing lesions present on the initial scan (A) resolve by 1 month later (B), while new lesions appear. Those lesions, in turn, disappear within 1 month (C), by which time new lesions appear (D). Two months later another new lesion appears (E), which is gone within 1 month (F).

With the development of immunomodulatory treatment, MRI is periodically used (e.g., annually) even in clinically stable patients to monitor therapeutic efficacy. Contrast enhancement or interim development of new nonenhancing FLAIR lesions are considered by many to indicate failure to arrest disease. Although MS is the epitome of white matter disease, more  recently developed sequences (including volumetric T1, FLAIR, and double-inversion recovery) have revealed significant gray matter involvement.15,23

In addition to the characteristic relapsing-remitting clinical pattern, MS can also be a progressive diffuse degenerative  disease. This is manifest as cerebral atrophy and abnormalities  in NAWM. Magnetization transfer contrast62 and diffusion tensor imaging (DTI)16 have shown measurable differences in the white matter of MS vs. healthy controls that were imperceptible  to visual inspection. Functional (f)MRI has also been used  to study connectivity abnormalities associated with cognitive dysfunction.24

http://radiologyebook.com

572

PART II  CT and MR Imaging of the Whole Body

VIRAL AND POSTVIRAL DEMYELINATING SYNDROMES Acute Disseminated Encephalomyelitis Acute disseminated encephalomyelitis (ADEM) is a demyelinating disease of the CNS that commonly follows a febrile illness such  as an upper respiratory infection. Numerous infectious agents— predominantly viruses but also several bacterial species—have been reported in association with ADEM. Less commonly the disease  may occur following a vaccination. Children with ADEM commonly present with nonspecific symptoms such as fever, headache, and lethargy; adults are more likely to have motor and sensory deficits, as might be seen with an acute episode of MS.14,46,68 Though they can present in similar fashion, distinction between ADEM and MS is critical because of different prognostic and thera-

peutic implications for the patient. ADEM is more common in  children and, unlike MS, is typically monophasic rather than relapsing. There are also underlying pathologic differences: though the lesions of both MS and ADEM are due to immune-mediated inflammation and demyelination, in ADEM the demyelination remains restricted to the perivascular area, unlike the confluent demyelination observed in MS.46 MRI is generally considered the most valuable diagnostic  tool in the workup of suspected ADEM.46 T2-weighted and FLAIR images demonstrate multiple asymmetric hyperintense white  matter lesions, particularly in the subcortical regions (Fig. 19-9). Periventricular and corpus callosum lesions are less common than in MS. Enhancement of lesions on postcontrast images is variable; larger lesions may show peripheral or focal irregular central enhancement.29 Spinal cord lesions can also be seen but are uncommon.10

A

B

C

D

E

F

FIG 19-9  Acute disseminated encephalomyelitis. A and B, Initial axial FLAIR images in this 32-year-old man who presented with numbness in his lower extremities, fever, and headache demonstrate numerous hyperintense foci in the subcortical white matter bilaterally. C, Post-Gd T1 images demonstrate enhancement of several of these lesions. Corresponding FLAIR (D and E) and post-Gd T1 (F) images obtained 2 months later (after steroid therapy and plasma exchange) showed resolution of most lesions.

http://radiologyebook.com

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies Given the overlap in MRI appearance between the two conditions, it is not always possible to distinguish ADEM from MS on a single MRI examination. For this reason, follow-up scans with an interval of at least 6 months are very helpful; lesions that resolve after corticosteroid therapy or remain unchanged suggest ADEM, but new lesions raise concern for MS.29,30

Progressive Multifocal Leukoencephalopathy Progressive multifocal leukoencephalopathy (PML) is a demyelinating disease that results from an opportunistic infection of oligodendrocytes by a type of human polyomavirus called the JC virus. The disease occurs in the setting of immunosuppression; although first described

573

in patients with Hodgkin’s disease and chronic lymphocytic leukemia, it has been most frequently associated with human immunodeficiency virus (HIV) infection. PML can also be seen in the context of biologic and immunomodulating therapies, such as natalizumab for MS.26,29 Patients usually present with insidious onset of neurologic deficits  that may affect motor function, speech, vision, personality, and cognition.39 MRI findings in PML include multiple hyperintense foci in the subcortical white matter on T2-weighted and FLAIR images. These lesions are frequently bilateral and characteristically involve the subcortical U fibers (Fig. 19-10). The parietal, occipital, and frontal lobes are commonly affected; other common sites include the posterior fossa and the deep gray matter nuclei.29,69,75

A

B

C

D

FIG 19-10  Progressive multifocal leukoencephalopathy. FLAIR images (A) obtained in this HIV patient showed asymmetric subcortical white matter signal changes in the left and right parietal white matter without significant enhancement visible on post-Gd T1 images (B). One month later, corresponding FLAIR and contrast-enhanced (C and D) images demonstrate progression of white matter changes (including involvement of the splenium of the corpus callosum) but still no significant lesion enhancement.

http://radiologyebook.com

574

A

D

PART II  CT and MR Imaging of the Whole Body

B

C

E

F

FIG 19-11  Progressive multifocal leukoencephalopathy—immune reconstitution inflammatory syndrome. A, Axial FLAIR images in this 31-year-old woman with HIV (viral load > 80,000 copies/mL) at her hospital presentation for weakness show asymmetric subcortical white matter hyperintensity in the left frontal lobe, which involves the U fibers. Follow-up images obtained 2 (B) and 14 months (C) later (after initiation of combined antiretroviral therapy) show progression, with eventual involvement of the contralateral hemisphere. D-F, Corresponding post-Gd T1 images show progressive enhancement over time, suggesting an evolving inflammatory process.

Other MRI features of PML allow distinction from other neurologic complications of the immunosuppressed state, including cerebral abscess, toxoplasmosis, and primary CNS lymphoma. PML lesions  do not demonstrate mass effect or associated edema, differentiating them from these other conditions. Additionally, contrast enhancement is usually absent in PML lesions, unlike the ringlike enhancement  commonly seen in cerebral abscesses or toxoplasmosis.55 Special note should be made of an entity known as immune reconstitution inflammatory syndrome (IRIS or PML-IRIS), commonly observed in HIV-positive patients with PML undergoing combination antiretroviral therapy. This condition is due to a reconstituted immunity, allowing infiltration of lymphocytes into PML lesions, with corresponding CNS inflammation. In cases of PML-IRIS, PML lesions may demonstrate mass effect and gadolinium (Gd) enhancement (Fig. 19-11).74

HIV Leukoencephalitis CNS manifestations of HIV infection include those related to primary HIV infection, those related to opportunistic infections, and those that result from neoplastic or vascular complications.41 We will focus here on the effects of primary HIV infection. Neuroimaging findings in HIV patients may be completely normal. In patients with findings of HIV encephalitis, early MRI findings include symmetric multifocal T2-hyperintense lesions in the periventricular white matter and centrum semiovale, which  generally spare the subcortical U fibers.6 Chronic HIV infection results in progression of these changes, with the initially patchy T2-hyperintense lesions becoming more confluent and extensive, sometimes extending to involve the basal ganglia, cortex, cerebellum, brainstem, and spine. The white matter lesions are typically slightly

http://radiologyebook.com

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies

A

575

B

FIG 19-12  HIV leukoencephalitis. A and B, Axial T2-weighted images acquired 12 months apart in this 57-year-old man with HIV show extensive symmetric and confluent white matter signal abnormalities. Even when accounting for slight differences in slice position, progression of cerebral atrophy in the interval between the scans was perceptible.

hypointense on T1 images and usually do not enhance. Another prominent feature of chronic HIV infection is progressive cerebral atrophy41 (Fig. 19-12).

TOXIC AND TRAUMATIC LEUKOENCEPHALOPATHIES Pathologic alteration of the cerebral white matter can be caused by exposure to a variety of agents, such as cranial irradiation, chemotherapeutic agents (including antineoplastic and immunosuppressive drugs), drugs of abuse, environmental toxins, and physical trauma. Clinical findings are highly variable; however, patients often present with neurobehavioral deficits ranging from inattention, forgetfulness, and personality changes to severe neurologic dysfunction, which can be easily confused with other white matter diseases.20,49

Radiation-Induced Demyelination Radiation injury to the brain is commonly seen as a complication of therapeutic radiation therapy for intracranial disease (most commonly primary and metastatic brain tumors, but also arteriovenous malformations) and head and neck neoplasms. It takes two forms: focal injury/radiation necrosis and diffuse white matter injury, also known as diffuse radiation-induced leukoencephalopathy.70

Radiation Necrosis Radiation necrosis (RN) is severe local tissue reaction to high-dose focal radiation therapy and is usually seen at the site of maximum radiation dose—in and around the tumor bed. The reported incidence is 3% to 24%, which increases steeply if doses exceed 65 Gy. RN is most commonly seen 3 to 12 months following radiotherapy, with 70%  of cases occurring during first 2 years, but can rarely develop even decades later.17,70 CT findings are nonspecific: RN typically appears as a focal hypodensity with mass effect and variable contrast enhancement. MRI

is more sensitive and specific than CT in evaluating RN, demonstrating T2-weighted and FLAIR hyperintensity with or without irregular contrast enhancement. Differentiation between recurrent neoplasm and RN is a frequent diagnostic dilemma because they share imaging characteristics, including origin at or close to the original tumor site, contrast enhancement, interval growth, and surrounding mass effect. Certain diagnostic clues that favor RN have been described, such as a “soap bubble” or “Swiss cheese–like” pattern of enhancement and  periventricular location of the lesion. Conversely, corpus callosum involvement, in conjunction with multiple enhancing lesions with  or without extension across the midline and subependymal spread, suggests tumor progression.33,50 In many cases, however, the distinction can still not be made with conventional imaging, owing to the coexistence of tumor and RN. In such cases, advanced imaging techniques like MR perfusion imaging (rCBV, vascular transfer constant [Ktrans]), MR spectroscopy (choline [Cho], creatine [Cr], and N-acetylaspartate [NAA]), and positron emission tomography (PET) may be helpful (Fig. 19-13). Restricted diffusion, elevated rCBV and Ktrans values, high Cho/Cr and Cho/NAA ratios, and high PET uptake are more frequently seen in recurrent tumor than in radiation necrosis.60

Diffuse Radiation-Induced Leukoencephalopathy Diffuse radiation-induced leukoencephalopathy refers to diffuse white matter injury secondary to large-volume or whole-brain radiation. RN and diffuse radiation-induced leukoencephalopathy can be seen concurrently in a patient or may occur sequentially. Typical CT findings are low-attenuation lesions involving the bilateral cerebral hemispheres, without any contrast enhancement or mass effect. The MRI is more sensitive and shows corresponding white matter hyperintensities on T2-weighted and FLAIR sequences. The severity of involvement ranges from small foci at the angles of the frontal and occipital horns to confluent and symmetric involvement of the entire cerebral white matter. More severe abnormalities are

http://radiologyebook.com

576

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 19-13  Radiation necrosis. This 35-year-old woman with a biopsy-proven WHO grade II right temporal oligodendroglioma was treated with resection and subsequent radiation therapy. A, Post-Gd T1-weighted image demonstrates multiple irregular foci of enhancement in the right frontal centrum semiovale. B, The corresponding corrected rCBV map fails to show elevated perfusion related to the right frontal abnormal enhancement. C, The corresponding FDG-PET image demonstrates asymmetrically diminished activity in the right frontal lobe, favoring radiation-induced changes over viable neoplasm. D, Six-month follow-up post-Gd T1-weighted image following treatment with hyperbaric oxygen therapy shows almost complete resolution of the right frontal enhancement, confirming the diagnosis of radiation necrosis.

noted with increase in patient age, volume of brain irradiated, radiation dose, and interval between radiotherapy and imaging. In most cases, the findings are indistinguishable from white matter changes seen with other leukoencephalopathies.42,70

Disseminated Necrotizing Leukoencephalopathy Disseminated necrotizing leukoencephalopathy (DNL) refers to diffuse white matter injury following treatment with intrathecal or systemic

chemotherapeutic agents. This condition was first reported in children with leukemia who were treated with intrathecal methotrexate,  but several other agents, including BCNU, melphalan, fludarabine,  cytarabine, 5-fluorouracil, levamisole, and cisplatin, have also been implicated.13,35 The prevalence of injury increases substantially when chemotherapy and radiation therapy are used together. However, the latent period between treatment and onset of symptoms is shorter after chemotherapy than after radiation.36,70

http://radiologyebook.com

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies

A

577

C

B

FIG 19-14  Acute methotrexate neurotoxicity. MRI was performed after this 23-year-old man undergoing methotrexate therapy for acute lymphoblastic leukemia began to experience facial hemiparesis and speech difficulties. DWI (A and B) obtained 16 hours apart demonstrate asymmetric ovoid regions of diffusion restriction involving the left and right centrum semiovale; interestingly the laterality of the imaging pattern reversed on the second exam. The axial FLAIR image (C) corresponding to image A reveals no appreciable abnormality. (Images courtesy Vasant Garg, MD and Stephanie Soriano, MD.)

The pathologic findings in DNL resemble those of radiationinduced white matter change and include axonal swelling, multifocal demyelination, coagulation necrosis, and gliosis. The clinical presentation and imaging appearance of DNL is also similar to diffuse radiationinduced leukoencephalopathy. In mild cases, both MRI and CT scan show diffuse symmetric bihemispheric white matter changes that are usually subclinical and transient and indistinguishable from diffuse radiation injury. However, in severe cases there is rapid neurologic deterioration, and imaging may demonstrate lesions with intense tumorlike solid enhancement and associated mass effect.35,70 Methotrexate has specifically been implicated as a causative agent when acute neurotoxicity is seen in patients who have been treated  for leukemia and lymphoma. Diffusion-weighted imaging (DWI) allows early detection of white matter injury in these patients. Typical findings include asymmetric bilateral diffusion restriction in the  deep hemispheric white matter, commonly in the centrum semiovale (Fig. 19-14). Contemporaneously acquired T2-weighted and FLAIR images often appear completely normal; corresponding regions of increased signal intensity may be seen on follow-up exams several days after presentation.21

Mineralizing Microangiopathy Mineralizing microangiopathy is a neuroradiologic abnormality characterized by intracerebral calcifications usually seen after combined radiation and chemotherapy for the treatment of CNS neoplasms in childhood. The calcifications are typically seen in the basal ganglia and subcortical white matter (Fig. 19-15) but rarely can also occur in cerebral cortex and the dentate nuclei of the cerebellum. The findings are usually superimposed on diffuse cortical atrophy and white matter changes.37 CT is more sensitive than MRI in detecting the calcifications. Areas of calcification generally demonstrate a decreased signal on conventional MRI sequences; this is due to a paucity of mobile protons. Rarely calcification may reveal paradoxically increased signal on T1-weighted images, which has been attributed to a surface relaxation mechanism associated with particulate calcium.61,67

Marchiafava-Bignami Disease Marchiafava-Bignami Disease (MBD) is a progressive disease characterized by demyelination and necrosis of the corpus callosum associated with chronic alcoholism. The condition was first described in Italian red wine drinkers but is occasionally seen in nonalcoholic patients as well. The disease may have an acute, subacute, or  chronic clinical presentation and may lead to death within weeks to months.5 Although CT may reveal hypoattenuation in the corpus callosum, MRI is currently the most sensitive diagnostic tool. The lesions are typically hyperintense on T2-weighted and FLAIR sequences and hypointense on T1-weighted sequences. Lesions most commonly involve the corpus callosum (Fig. 19-16) but may also occur in the adjacent hemispheric white matter and middle cerebral peduncles; the subcortical U fibers are usually spared. The lesions usually do not cause mass effect but may show peripheral enhancement in the acute phase. The chronic lesions have a more cystic appearance. On DWI, diffusion restriction can be seen, which may resolve completely without any apparent permanent damage.25,45,57

HYPOXIC, ISCHEMIC, AND METABOLIC LEUKOENCEPHALOPATHIES The cerebral white matter is susceptible to injury from a variety of processes that deprive it of adequate oxygenated blood supply or cause metabolic derangements. Several of these entities are described in the following sections.

Hypoxic-Ischemic Encephalopathy Inadequate oxygen delivery to the brain can occur in a number of clinical situations, including cardiopulmonary arrest, drowning, carbon monoxide poisoning, and severe anemia.66 Hypoxic brain injury can also be an unfortunate complication of childbirth. In these critically ill patients, imaging can help guide short-term clinical management and in some cases may provide prognostic information.

http://radiologyebook.com

578

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 19-15  Cerebral calcifications in a similar pattern to mineralizing microangiopathy. These noncontrast CT images obtained in a 79-year-old woman with multiple medical problems demonstrate coarse basal ganglia calcifications (A) as well as posterior subcortical and parenchymal calcifications (B). Although there was no documented history of chemotherapy or radiation treatment in this patient, the CT appearance is certainly similar to findings seen in mineralizing microangiopathy.

A

C

B

FIG 19-16  Marchiafava-Bignami disease. MRI (images cropped for splenium) was performed in this 40-yearold woman with alcohol dependence who presented with left-sided sensory deficits. Axial FLAIR image (A) reveals focal increased signal in the splenium of the corpus callosum. Associated abnormal diffusion restriction is seen on DWI (B) and ADC map (C).

CT is frequently the first imaging study obtained in cases of anoxic or hypoxic brain injury. Typical findings include diffuse brain swelling, relative loss of gray-white matter differentiation, and hypodensity of the basal ganglia. It is vital to remember that these changes can be very subtle and progress over time. The appearance of an initial head CT may be completely unremarkable.66 MRI, particularly DWI,4 is more sensitive for detecting subtle abnormalities in the early phases of brain injury. Findings are described as occurring in four phases: an acute phase (within 24 hours after the hypoxic insult), an early subacute phase (days 1-13), a late subacute phase (days 14-20), and a chronic phase (after 21 days). During the acute and early subacute phases, T2-weighted and DWI sequences show increased signal in the cortex, thalamus, and basal ganglia  (Fig. 19-17). In the late subacute phase, the hyperintensity on DWI

fades and diffuse white matter abnormalities may develop. During the chronic phase, the DWI signal has returned to normal and diffuse atrophy and ventricular dilatation is visible.73

Posterior Reversible Encephalopathy Syndrome Posterior reversible encephalopathy syndrome (PRES) has been described as a neurotoxic state accompanied by a unique CT or MRI appearance.7 The clinical manifestations of the syndrome can include a broad range of signs and symptoms, including headache, visual changes, paresis, altered mental status, and generalized seizures.22 Probably the best-known feature of PRES, however, is the symmetric pattern of posterior-predominant bihemispheric cerebral edema visible on CT and MRI studies (Fig. 19-18). PRES has been described in association with a number of conditions, including preeclampsia/

http://radiologyebook.com

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies

A

B

C

D

579

FIG 19-17  Hypoxic-ischemic encephalopathy. This 2-year-old boy experienced cardiac arrest during surgical removal of a Wilms’ tumor. DWI (A) and ADC map (B) show extensive abnormal diffusion restriction in the bihemispheric white matter and deep gray structures. T2-weighted (C) and FLAIR (D) images show corresponding increased signal.

eclampsia, allogeneic bone marrow and solid organ transplantation, multiple autoimmune diseases, and in cancer patients receiving highdose chemotherapy.7 The pathophysiologic mechanism underlying PRES is still not clearly established and has been a source of controversy. A current popular theory holds that development of vasogenic edema in  PRES is due to severe hypertension overwhelming the limits of cerebrovascular autoregulation, resulting in hyperperfusion. This theory is challenged by the fact that in some cases, PRES occurs with only mild or absent elevation of blood pressure. Another, older theory that has received renewed support argues that systemic toxicity leads to endothelial dysfunction, resulting in vasoconstriction and hypoperfusion that gives rise to the characteristic imaging findings.8

Classic imaging findings on CT and MR are of symmetric vasogenic edema, most commonly affecting the parietal and occipital lobes. This edema usually completely resolves and is manifested as confluent hypodensity in the posterior subcortical and deep white matter on CT, with corresponding T2 and FLAIR hyperintensity on MRI. The typical pattern of lesion location follows the distribution of brain watershed zones. A variety of other patterns, however, have been reported in the literature, including edema involving the basal ganglia, cerebellum, and brainstem. Associated hemorrhage and foci of diffusion restriction have also been reported in a minority of cases.7,65

Binswanger’s Disease Binswanger’s disease is a form of subcortical vascular dementia caused by arteriosclerosis affecting small vessels supplying the cerebral white

http://radiologyebook.com

580

PART II  CT and MR Imaging of the Whole Body

C

B

A

FIG 19-18  Posterior reversible encephalopathy syndrome (PRES). A, Emergent head CT was performed for acute hypertension in this 2-year-old girl with a history of severe aplastic anemia undergoing bone marrow transplantation. A symmetric pattern of posterior-predominant vasogenic edema was evident. Also noteworthy is the moderate parenchymal volume loss, likely related to the patient’s other underlying medical conditions. Subsequently obtained T2-weighted (B) and FLAIR (C) MRIs showed symmetric abnormally high signal in the posterior cerebral white matter.

matter. The disease is strongly associated with hypertension and lacunar infarction (>90% of cases). The clinical presentation is characterized primarily by loss of memory and worsening cognitive function; emotional changes and evidence of subcortical dysfunction (e.g., gait abnormalities, incontinence) may also be present.2,53 Diffuse asymmetric hypodensity of the central cerebral white matter is visible on CT studies. MRI findings include T2-hyperintense lesions in the subcortical and periventricular lesions, especially in  the frontal horns and centrum semiovale. As mentioned earlier, lacunar infarcts (in the basal ganglia and thalami) are extremely common. Diffuse cerebral atrophy is another consistent finding of this entity.

Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy CADASIL is an inherited form of vascular dementia that usually presents between ages 30 and 50. Common presentations include migraines and ischemic events, with eventual progression to subcortical dementia.72 The disease has been linked to mutations of the Notch3 gene on chromosome 19, which result in severe changes to vascular smooth muscle cells.28 MRI demonstrates confluent T2-hyperintense areas in the  periventricular white matter that become more diffuse over time, with  eventual involvement of the external capsule and temporal poles  (Fig. 19-19). The basal ganglia and thalamus are also affected, allowing differentiation from MS. Other findings include lacunar infarcts and microhemorrhages.11

Osmotic Demyelination Syndrome Osmotic demyelination syndrome (ODS) refers to central pontine myelinolysis (CPM) and extrapontine myelinolysis (EPM). CPM is a syndrome of acute pontine demyelination, originally described in alcoholics1 but most commonly seen in association with electrolyte abnormalities and rapid correction of hyponatremia.31,32,64 Other predisposing conditions include hepatic dysfunction, liver transplantation, and hyperosmolar states.3 The typical clinical course of CPM is marked by neurologic deterioration 2 to 8 days following correction

of hyponatremia; dysarthria and dysphagia occur first, followed by progression from flaccid quadriparesis (corticospinal tract) to spastic quadriparesis (basis pontis). Involvement of the tegmentum may result in pupillary and oculomotor dysfunction. Severe cases may result in “locked in” syndrome.3,64 Although the pontine white matter is most susceptible to this process, extrapontine involvement (EPM) is common and sometimes even occurs independent of CPM. Common presenting symptoms of EPM include behavioral disorders, movement disorders, seizures, polyradiculopathy, and neuropathy.3 Pathologic features of ODS include a symmetric noninflammatory loss of myelin, with preservation of neuronal cell bodies and axons of the central pons and sparing of the peripheral fibers and axons of the corticospinal tract. In cases of EPM, similar findings are seen in other areas, including the cerebellum, lateral geniculate body, external capsule, basal ganglia, thalamus, gray-white junction of the cerebral cortex, and hippocampi.32 MRI has permitted recognition of ODS as it develops and is the most sensitive imaging modality for this entity. DWI is particularly useful for early diagnosis.58 Increased DWI signal with corresponding decreased ADC values in the central pons is characteristically seen in the central pons in a symmetric trident pattern. Increased T2 and FLAIR signal in an identical pattern occurs 7 to 10 days after the appearance of diffusion restriction (Fig. 19-20). There is no mass effect and typically no enhancement of lesions, allowing differentiation from neoplasms and acute MS plaques. In cases of EPM, similar findings occur bilaterally and symmetrically along a similar time course in the aforementioned commonly affected extrapontine locations.3

MISCELLANEOUS Amyloid Leukoencephalopathy Cerebral amyloid angiopathy (CAA) is a cerebrovascular disease caused by deposition of β-amyloid protein in the walls of blood vessels. CAA is usually seen in elderly patients and is well known as a cause 

http://radiologyebook.com

FIG 19-19  CADASIL. FLAIR images in

A

B

C

D

this patient with CADASIL show confluent areas of signal hyperintensity in the periventricular white matter (A) and right temporal pole (B). Several lacunar infarcts are also apparent. Foci of acute ischemia are apparent on DWI (C and D).

FIG 19-20  Central pontine myelinolysis. MRI (images cropped for pons) was performed in this 19-year-old liver transplant recipient. T2-weighted (A) and FLAIR (B) images show abnormal signal in the pons with sparing of the corticospinal tracts, resulting in the characteristic “trident” pattern.

A

B http://radiologyebook.com

582

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F

FIG 19-21  Amyloid leukoencephalopathy. This 75-year-old woman was transferred from an outside hospital with altered mental status; the working diagnosis was “possible brain tumor.” A large region of abnormal high signal in the right frontal white matter is evident on T2-weighted and FLAIR images (A and B). Susceptibility-weighted images (C) revealed punctate foci of low signal in the right frontal and left parietal lobes, suggesting microhemorrhages. Pre- and postcontrast T1-weighted images (D and E) failed to show enhancement in the region of T2 signal abnormality. Fused PET/CT images (F) demonstrated decreased FDG uptake corresponding to the region of signal abnormality, adding further evidence that this was not a malignant neoplasm.

of intracerebral hemorrhage. In uncommon cases, an inflammatory manifestation of this disease can result in cognitive decline, headaches, and seizures.9,43 This inflammatory entity, which we call amyloid leukoencephalopathy (also referred to as CAA-related inflammation or cerebral amyloid inflammatory vasculopathy) presents on MRI with areas of high signal intensity in the cerebral white matter, which may be solitary or multifocal and bilateral.38 No contrast enhancement is present.43 The diagnosis may be suggested on gradient echo images, which will allow visualization of associated cerebral microhemorrhages or hemosiderin foci (Fig. 19-21), common findings in CAA.

REFERENCES 1. Adams RD, Victor M, Mancall EL: Central pontine myelinolysis:   A hitherto undescribed disease occurring in alcoholic and   malnourished patients. AMA Arch Neurol Psychiatry 81(2):154–172, 1959. 2. Akiguchi I, Budka H, Shirakashi Y, et al: MRI features of Binswanger’s disease predict prognosis and associated pathology. Ann Clin Transl Neurol 1(10):813–821, 2014. 3. Alleman AM: Osmotic demyelination syndrome: Central pontine myelinolysis and extrapontine myelinolysis. Semin Ultrasound CT MR 35(2):153–159, 2014.

http://radiologyebook.com

CHAPTER 19  Demyelinating Disease and Leukoencephalopathies 4. Arbelaez A, Castillo M, Mukherji SK: Diffusion-weighted MR imaging of global cerebral anoxia. AJNR Am J Neuroradiol 20(6):999–1007, 1999. 5. Arbelaez A, Pajon A, Castillo M: Acute Marchiafava-Bignami disease: MR findings in two patients. AJNR Am J Neuroradiol 24(10):1955–1957, 2003. 6. Bakshi R: Neuroimaging of HIV and AIDS related illnesses: A review. Front Biosci 9:632–646, 2004. 7. Bartynski WS: Posterior reversible encephalopathy syndrome, part 1: Fundamental imaging and clinical features. AJNR Am J Neuroradiol 29(6):1036–1042, 2008. 8. Bartynski WS: Posterior reversible encephalopathy syndrome, part 2: Controversies surrounding pathophysiology of vasogenic edema. AJNR Am J Neuroradiol 29(6):1043–1049, 2008. 9. Bernstein RA, Gibbs M, Hunt Batjer H: Clinical diagnosis and successful treatment of inflammatory cerebral amyloid angiopathy. Neurocrit Care 14(3):453–455, 2011. 10. Bester M, Petracca M, Inglese M: Neuroimaging of multiple sclerosis, acute disseminated encephalomyelitis, and other demyelinating diseases. Semin Roentgenol 49(1):76–85, 2014. 11. Chabriat H, Joutel A, Dichgans M, et al: CADASIL. Lancet Neurology 8(7):643–653, 2009. 12. Charil A, Yousry TA, Rovaris M, et al: MRI and the diagnosis of multiple sclerosis: Expanding the concept of “no better explanation.” Lancet Neurology 5(10):841–852, 2006. 13. Cossaart N, SantaCruz KS, Preston D, et al: Fatal chemotherapy-induced encephalopathy following high-dose therapy for metastatic breast cancer: A case report and review of the literature. Bone Marrow Transplant 31(1):57–60, 2003. 14. Dale RC, de Sousa C, Chong WK, et al: Acute disseminated encephalomyelitis, multiphasic disseminated encephalomyelitis and multiple sclerosis in children. Brain 123(Pt 12):2407–2422, 2000. 15. de Graaf WL, Zwanenburg JJ, Visser F, et al: Lesion detection at seven Tesla in multiple sclerosis using magnetisation prepared 3D-FLAIR and 3D-DIR. Eur Radiol 22(1):221–231, 2012. 16. Droogan AG, Clark CA, Werring DJ, et al: Comparison of multiple sclerosis clinical subgroups using navigated spin echo diffusion-weighted imaging. Magn Reson Imaging 17(5):653–661, 1999. 17. Fatterpekar GM, Galheigo D, Narayana A, et al: Treatment-related change versus tumor recurrence in high-grade gliomas: A diagnostic conundrum—Use of dynamic susceptibility contrast-enhanced (DSC) perfusion MRI. AJR Am J Roentgenol 198(1):19–26, 2012. 18. Filippi M, Rocca MA: MR imaging of multiple sclerosis. Radiology 259(3):659–681, 2011. 19. Filippi M, Rocca MA, De Stefano N, et al: Magnetic resonance techniques in multiple sclerosis: The present and the future. Arch Neurol 68(12):1514–1520, 2011. 20. Filley CM, Kleinschmidt-DeMasters BK: Toxic leukoencephalopathy.   N Engl J Med 345(6):425–432, 2001. 21. Fisher MJ, Khademian ZP, Simon EM, et al: Diffusion-weighted MR imaging of early methotrexate-related neurotoxicity in children. AJNR Am J Neuroradiol 26(7):1686–1689, 2005. 22. Garg RK: Posterior leukoencephalopathy syndrome. Postgrad Med J 77(903):24–28, 2001. 23. Geurts JJ, Pouwels PJ, Uitdehaag BM, et al: Intracortical lesions in multiple sclerosis: Improved detection with 3D double inversionrecovery MR imaging. Radiology 236(1):254–260, 2005. 24. Hawellek DJ, Hipp JF, Lewis CM, et al: Increased functional connectivity indicates the severity of cognitive impairment in multiple sclerosis. Proc Natl Acad Sci U S A 108(47):19066–19071, 2011. 25. Hlaihel C, Gonnaud PM, Champin S, et al: Diffusion-weighted magnetic resonance imaging in Marchiafava-Bignami disease: Follow-up studies. Neuroradiology 47(7):520–524, 2005. 26. Hoepner R, Faissner S, Salmen A, et al: Efficacy and side effects of natalizumab therapy in patients with multiple sclerosis. J Cent Nerv Syst Dis 6:41–49, 2014. 27. Jacobs LD, Wende KE, Brownscheidle CM, et al: A profile of multiple sclerosis: The New York State Multiple Sclerosis Consortium. Mult Scler 5(5):369–376, 1999.

583

28. Joutel A, Corpechot C, Ducros A, et al: Notch3 mutations in CADASIL,   a hereditary adult-onset condition causing stroke and dementia. Nature 383(6602):707–710, 1996. 29. Kanekar S, Devgun P: A pattern approach to focal white matter hyperintensities on magnetic resonance imaging. Radiol Clin North Am 52(2):241–261, 2014. 30. Kesselring J, Miller DH, Robb SA, et al: Acute disseminated encephalomyelitis. MRI findings and the distinction from multiple sclerosis. Brain 113(Pt 2):291–302, 1990. 31. Kleinschmidt-DeMasters BK, Norenberg MD: Rapid correction of hyponatremia causes demyelination: Relation to central pontine myelinolysis. Science 211(4486):1068–1070, 1981. 32. Kleinschmidt-Demasters BK, Rojiani AM, Filley CM: Central and extrapontine myelinolysis: Then … and now. J Neuropathol Exp Neurol 65(1):1–11, 2006. 33. Kumar AJ, Leeds NE, Fuller GN, et al: Malignant gliomas: MR imaging spectrum of radiation therapy- and chemotherapy-induced necrosis of the brain after treatment. Radiology 217(2):377–384, 2000. 34. Lassmann H, Bruck W, Lucchinetti CF: The immunopathology of multiple sclerosis: An overview. Brain Pathol 17(2):210–218, 2007. 35. Laxmi SN, Takahashi S, Matsumoto K, et al: Treatment-related disseminated necrotizing leukoencephalopathy with characteristic contrast enhancement of the white matter. Radiat Med 14(6):303–307, 1996. 36. Lee YY, Nauert C, Glass JP: Treatment-related white matter changes in cancer patients. Cancer 57(8):1473–1482, 1986. 37. Lewis E, Lee YY: Computed tomography findings of severe mineralizing microangiopathy in the brain. J Comput Tomogr 10(4):357–364, 1986. 38. Loes DJ, Biller J, Yuh WT, et al: Leukoencephalopathy in cerebral amyloid angiopathy: MR imaging in four cases. AJNR Am J Neuroradiol 11(3):485–488, 1990. 39. Love S: Demyelinating diseases. J Clin Pathol 59(11):1151–1159, 2006. 40. Lublin FD, Reingold SC, Cohen JA, et al: Defining the clinical course of multiple sclerosis: The 2013 revisions. Neurology 83(3):278–286, 2014. 41. Mahan M, Karl M, Gordon S: Neuroimaging of viral infections of the central nervous system. Handb Clin Neurol 123:149–173, 2014. 42. Martins AN, Johnston JS, Henry JM, et al: Delayed radiation necrosis of the brain. J Neurosurg 47(3):336–345, 1977. 43. Martucci M, Sarria S, Toledo M, et al: Cerebral amyloid angiopathyrelated inflammation: Imaging findings and clinical outcome. Neuroradiology 56(4):283–289, 2014. 44. McDonald WI, Compston A, Edan G, et al: Recommended diagnostic criteria for multiple sclerosis: Guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 50(1):121–127, 2001. 45. Menegon P, Sibon I, Pachai C, et al: Marchiafava-Bignami disease: Diffusion-weighted MRI in corpus callosum and cortical lesions. Neurology 65(3):475–477, 2005. 46. Menge T, Hemmer B, Nessler S, et al: Acute disseminated encephalomyelitis: An update. Arch Neurol 62(11):1673–1680, 2005. 47. Miller D, Barkhof F, Montalban X, et al: Clinically isolated syndromes suggestive of multiple sclerosis, part I: Natural history, pathogenesis, diagnosis, and prognosis. Lancet Neurol 4(5):281–288, 2005. 48. Milo R, Kahana E: Multiple sclerosis: Geoepidemiology, genetics and the environment. Autoimmun Rev 9(5):A387–A394, 2010. 49. Moser FG, Hilal SD, Bello JA, et al: Magnetic resonance imaging of white matter changes in traumatic leukoencephalopathy. J Comput Tomogr 12(3):171–173, 1988. 50. Mullins ME, Barest GD, Schaefer PW, et al: Radiation necrosis versus glioma recurrence: Conventional MR imaging clues to diagnosis. AJNR Am J Neuroradiol 26(8):1967–1972, 2005. 51. Nesbit GM, Forbes GS, Scheithauer BW, et al: Multiple sclerosis: Histopathologic and MR and/or CT correlation in 37 cases at biopsy and three cases at autopsy. Radiology 180(2):467–474, 1991. 52. Noseworthy JH, Lucchinetti C, Rodriguez M, et al: Multiple sclerosis.   N Engl J Med 343(13):938–952, 2000. 53. Olsen CG, Clasen ME: Senile dementia of the Binswanger’s type. Am Fam Physician 58(9):2068–2074, 1998.

http://radiologyebook.com

584

PART II  CT and MR Imaging of the Whole Body

54. Polman CH, Reingold SC, Edan G, et al: Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria.” Ann Neurol 58(6):840–846, 2005. 55. Price RW: Neurological complications of HIV infection. Lancet 348(9025):445–452, 1996. 56. Qian P, Cadavid D, Wolansky LJ, et al: Heterogeneity in longitudinal evolution of ring-enhancing multiple sclerosis lesions. Ann Neurol 70(4):668–669, author reply 669–670, 2011. 57. Raina S, Mahesh DM, Mahajan J, et al: Marchiafava-Bignami disease.   J Assoc Physicians India 56:633–635, 2008. 58. Ruzek KA, Campeau NG, Miller GM: Early diagnosis of central pontine myelinolysis with diffusion-weighted imaging. AJNR Am J Neuroradiol 25(2):210–213, 2004. 59. Schumacher GA, Beebe G, Kibler RF, et al: Problems of experimental trials of therapy in multiple sclerosis: Report by the Panel on the Evaluation of Experimental Trials of Therapy in Multiple Sclerosis.   Ann N Y Acad Sci 122:552–568, 1965. 60. Shah R, Vattoth S, Jacob R, et al: Radiation necrosis in the brain: Imaging features and differentiation from tumor recurrence. Radiographics 32(5):1343–1359, 2012. 61. Shanley DJ: Mineralizing microangiopathy: CT and MRI. Neuroradiology 37(4):331–333, 1995. 62. Silver NC, Lai M, Symms MR, et al: Serial magnetization transfer imaging to characterize the early evolution of new MS lesions. Neurology 51(3):758–764, 1998. 63. Simon JH, Holtas SL, Schiffer RB, et al: Corpus callosum and subcallosal-periventricular lesions in multiple sclerosis: Detection with MR. Radiology 160(2):363–367, 1986. 64. Singh TD, Fugate JE, Rabinstein AA: Central pontine and extrapontine myelinolysis: A systematic review. Eur J Neurol 21(12):1443–1450, 2014. 65. Stevens CJ, Heran MK: The many faces of posterior reversible encephalopathy syndrome. Br J Radiol 85(1020):1566–1575, 2012.

66. Stevens RD, Pustavoitau A, Chalela JA: Brain imaging in intensive care medicine. Semin Neurol 28(5):631–644, 2008. 67. Suzuki S, Nishio S, Takata K, et al: Radiation-induced brain calcification: Paradoxical high signal intensity in T1-weighted MR images. Acta Neurochir (Wien) 142(7):801–804, 2000. 68. Tenembaum S, Chamoles N, Fejerman N: Acute disseminated encephalomyelitis: A long-term follow-up study of 84 pediatric patients. Neurology 59(8):1224–1231, 2002. 69. Thurnher MM, Thurnher SA, Muhlbauer B, et al: Progressive multifocal leukoencephalopathy in AIDS: Initial and follow-up CT and MRI. Neuroradiology 39(9):611–618, 1997. 70. Valk PE, Dillon WP: Radiation injury of the brain. AJNR Am J Neuroradiol 12(1):45–62, 1991. 71. van Walderveen MA, Kamphorst W, Scheltens P, et al: Histopathologic correlate of hypointense lesions on T1-weighted spin-echo MRI in multiple sclerosis. Neurology 50(5):1282–1288, 1998. 72. Vitali P, Migliaccio R, Agosta F, et al: Neuroimaging in dementia. Semin Neurol 28(4):467–483, 2008. 73. Weiss N, Galanaud D, Carpentier A, et al: Clinical review: Prognostic value of magnetic resonance imaging in acute brain injury and coma. Critical Care 11(5):230, 2007. 74. Weissert R: Progressive multifocal leukoencephalopathy. J Neuroimmunol 231(1-2):73–77, 2011. 75. Whiteman ML, Post MJ, Berger JR, et al: Progressive multifocal leukoencephalopathy in 47 HIV-seropositive patients: Neuroimaging with clinical and pathologic correlation. Radiology 187(1):233–240, 1993. 76. Wolansky LJ, Bardini JA, Cook SD, et al: Triple-dose versus single-dose gadoteridol in multiple sclerosis patients. J Neuroimaging 4(3):141–145, 1994. 77. Wolansky LJ, Finden SG, Chang R, et al: Gadoteridol in multiple sclerosis patients. A comparison of single and triple dose with immediate vs. delayed imaging. Clin Imaging 22(6):385–392, 1998.

http://radiologyebook.com

20  Orbit Colin S. Poon, Michael Abrahams, and James J. Abrahams

ANATOMY The soft tissue structures of the orbit are contained within a bony cavity and include the globe, extraocular muscles, optic nerve–sheath complex, lacrimal apparatus, and various vascular and nerve structures (Figs. 20-1 to 20-4).

Bony Anatomy The bony orbit is a conical structure with the apex pointing posteriorly. The orbital roof is composed of the frontal bone and is thinner anteriorly. The medial wall is composed of the frontal process of the maxillary bone anteriorly, the lamina papyracea of the ethmoid air cells at the midportion, and the sphenoid bone posteriorly. The lamina papyracea is very thin, and not surprisingly it is a common site of orbital blowout fracture and spontaneous dehiscence of orbital fat. The lateral orbital wall is formed by the orbital surface of the zygomatic bone. The orbital floor is formed by the orbital plate of the maxilla, the orbital process of the palatine bone, and the orbital surface of the zygomatic bone. The orbital plate of the maxilla is thin and a common site of inferior blowout fracture. Multiple foramina and canals go through the bony orbits (Box 20-1) (see Figs. 20-1 and 20-2). The optic canal (also called optic foramen) is located at the orbital apex. It is bordered by two bony spikes of the lesser wing of the sphenoid bone, commonly referred to as the optic struts. The canal contains the optic nerve and ophthalmic artery, both of which are contained within a dural sheath. The superior orbital fissure is located at the margin between the lateral wall and the orbital roof. The greater wing of the sphenoid bone forms its lateral boundary, and the lesser wing forms its medial boundary. The superior orbital fissure contains the superior ophthalmic vein; the oculomotor (III), trochlear (IV), and abducens (VI) nerves; and the ophthalmic division of the trigeminal nerve (V1). The superior orbital fissure forms the largest communication between the orbit and intracranial structures and therefore forms a conduit for infectious or neoplastic processes between the orbital apex and the cavernous sinus. The inferior orbital fissure is located at the margin between the lateral wall and the orbital floor. It contains the infraorbital (branch of V2) and zygomatic nerves, the nerve branches from the pterygopalatine ganglion, and venous connection between the inferior ophthalmic vein and the pterygoid plexus. The inferior orbital fissure connects with the pterygopalatine fossa and the masticator space/infratemporal fossa, allowing the spread of deep facial infection and neoplasm to the orbital apex. The globe is essentially a spherical structure, with the wall consisting of three layers: retina (innermost), choroids (middle), and sclera (outermost) (Fig. 20-3). These layers cannot be resolved with current

586

clinical computed tomography (CT) or magnetic resonance imaging (MRI) technology unless they are separated by pathologic processes (e.g., retinal detachment). The globe is divided into three fluid-filled cavities: anterior chamber, posterior chamber, and vitreous cavity.4,43 The anterior chamber and posterior chamber constitute the anterior segment, and the vitreous cavity constitutes the posterior segment. The anterior chamber extends from the cornea to the iris. The posterior chamber extends from the posterior surface of the iris to the anterior surface of the vitreous. The vitreous cavity is posterior to the posterior chamber. The anterior border of the orbit is formed by the orbital septum (see Fig. 20-2D), a fibrous structure adherent to the inner margin of the orbital rim, with central portions that extend into the tarsus of the eyelids. Although there are a few orifices for passage of vessels, nerves, and ducts, the septum forms an effective barrier to prevent superficial processes from extending into the orbit proper. A pathologic process such as cellulitis may be designated as preseptal versus postseptal. A postseptal process signals involvement of more critical structures of the orbit and the possibility of extension into the cavernous sinus and intracranial structures.

Soft Tissue Anatomy There are seven extraocular muscles: the superior, inferior, medial and lateral rectus; the superior and inferior oblique; and the levator palpebrae superioris muscles. The levator palpebrae muscle can be seen immediately above the superior rectus muscle. With the exception of the inferior oblique muscle, all extraocular muscles originate from the annulus of Zinn, a tendinous ring in the orbital apex. They pass anteriorly and insert on the globe just behind the corneoscleral border. The four rectus muscles and the fibrous septa connecting between them form the muscle cone of the orbit. The intraconal space is filled with orbital fat. Orbital vessels, sensory and motor nerves to the extraconal muscles, and the optic nerve–sheath complex also traverse the intraconal space. The optic nerve may appear straight or slightly tortuous depending on the eye position. It consists of three segments: orbital, canalicular, and intracranial. The orbital segment is covered by the same meningeal sheaths as the brain. The normal diameter of the optic nerve is up to 4 mm. A layer of cerebrospinal fluid can be seen between the meningeal sheath and optic nerve. The extraconal space represents the area between the muscle cone and bony orbit. This space contains orbital fat and the lacrimal gland. The lacrimal gland is located superolateral to the globe. The upper margin of the gland is convex. The lower margin is concave and lies on the levator palpebrae and lateral rectus muscles. The lacrimal system drains through the lacrimal ductal system near the medial canthus. It consists of the superior and inferior puncta, their associated

http://radiologyebook.com

587

CHAPTER 20  Orbit c

d

c

d

e

f

g

b e b

f

a

h

j

a

A

i

B

g Frontal bone

Frontozygomatic suture

Lamina papyracea Zygomatic bone

C

Maxillary bone

Infraorbital foramen Superior orbital fissure

D

Optic canal

Inferior orbital fissure

Anterior clinoid process

Optic strut

FIG 20-1  A, Coronal CT scan: normal anatomy. Lateral rectus (a), superior rectus (b), medial rectus (c), superior oblique (d), levator palpebrae superioris (e), lacrimal gland (f), inferior oblique (g). B, Medial rectus (a), superior oblique (b), ophthalmic artery (c), superior rectus/levator palpebrae superioris complex (d), dural sheath (e), superior ophthalmic vein (f), subarachnoid space (g), optic nerve (h), inferior rectus (i), lateral rectus (j). C, Bony orbits at midanterior level. D, Bony orbits at orbital apex.

ducts, the lacrimal sac, lacrimal duct, and the valve of Hasner, which is a draining orifice inferolateral to the inferior nasal turbinate. The vascular anatomy of the orbits can be well demonstrated on high-resolution MRI16 and CT angiography (CTA). The primary arterial supply to the orbit is the ophthalmic artery. It is superior to the optic nerve and can be seen crossing the optic nerve almost perpendicularly (Fig. 20-4). The ophthalmic artery most often originates from the internal carotid artery. The origin is usually at the anteromedial aspect of the internal carotid artery as it exits the cavernous sinus. Variants of its origin include the cavernous segment of the internal carotid artery and the middle meningeal artery (i.e., external carotid artery branch). Secondary arterial supply to the orbits comes from the external carotid artery. Because the orbits receive blood supply from both the internal and external carotid arteries, orbital arteries may serve as anastomoses between the two arterial systems. The largest orbital vein visualized on CT or MRI is the superior ophthalmic vein. It can be seen arising near the base of the nose, coursing anteromedially to posterolaterally, and draining into the cavernous

sinus. It crosses over the optic nerve in its midcourse at approximately 20 degrees (see Fig. 20-4). The midportion of the superior ophthalmic vein is an intraconal structure that lies between the superior rectus muscle and the ophthalmic artery. The inferior ophthalmic vein is much smaller than the superior ophthalmic vein. It is usually not well visualized on CT or MRI studies. Both the superior ophthalmic vein and inferior ophthalmic vein receive tributaries from the veins of the face and nose.

IMAGING TECHNIQUES The major modalities for imaging the orbits include CT and MRI. The abundance of intraorbital fat provides good intrinsic soft tissue contrast on CT for most clinical applications. The advances of multidetector CT technology now make high-resolution CT imaging possible. The source images can be reformatted in different planes, providing high-resolution isotropic imaging. This renders the previous advantage of multiplanar capability of MRI obsolete. CT is superior to MRI

http://radiologyebook.com

588

PART II  CT and MR Imaging of the Whole Body Lacrimal sac and duct

A

Inferior rectus m.

Ciliary bodies

Inferior orbital fissure

Medial rectus m.

Anterior chamber Lens Vitreous cavity Lateral rectus m. Optic nerve and sheath

B

Ophthalmic artery Lacrimal gland

C

Lacrimal vein

Superior orbital fissure Superior orbital fissure

Superior ophthalmic vein

FIG 20-2  Normal orbital anatomy. Direct axial CT scanning from inferior to superior. A to D, Soft tissue window.

http://radiologyebook.com

CHAPTER 20  Orbit Superior oblique m.

D

Trochlear m.

589

Orbital septum

Superior rectus m. Lacrimal gland Nasomaxillary Lamina suture papyracea

Zygomatic bone

Sphenozygomatic suture Greater wing of sphenoid Sphenotemporal suture

E

Superior orbital fissure

Optic canal

Optic strut

FIG 20-2, cont’d E, Bone window.

BOX 20-1  Major Foramina of the Orbit

and Their Neurovascular Contents Optic Canal Optic nerve Ophthalmic artery Superior Orbital Fissure Cranial nerves: III, IV, VI, V1 Lacrimal and frontal nerves Superior and inferior ophthalmic veins Inferior Orbital Fissure Cranial nerve: V2 Zygomatic nerve Infraorbital vessels

for delineation of osseous structures and calcifications. It requires short imaging time and is therefore less sensitive to motion of the globe and eyelid. CT imaging can be completed quickly and requires less patient cooperation, making it ideal for imaging orbital trauma. Compared to CT, MRI provides superior soft tissue contrast. It also provides better imaging details of the intracranial structures. When it is important to assess intracranial abnormalities, either as direct

extension of orbital lesions or as associated lesions in certain diseases (e.g., in multiple sclerosis), MRI is superior to CT. In the past, evaluation of suspected vascular lesions of the orbits required conventional angiography. The advances in CTA and MRA now allow many vascular lesions to be evaluated noninvasively. In some cases, conventional angiography can be foregone. CT and MRI often provide complementary roles in orbital imaging. The choice of CT versus MRI for initial imaging of the orbits depends on the clinical problem. CT is usually preferred for trauma, for evaluation of the bony orbits or calcified lesions, and when MRI is contraindicated. For other applications, MRI is generally preferred because of the absence of radiation risks and its high soft tissue contrast. MRI is the initial imaging of choice for evaluation of the optic nerve, other cranial nerves, and intracranial lesions. Exceptions can be found in a small number of optic nerve meningiomas, which are very small and mostly calcified. These lesions may be missed by MRI and are better detected by CT.

Computed Tomography The orbits are often included in routine CT head or maxillofacial CT examinations. These screening examinations are usually performed according to the standard head or maxillofacial CT protocols. When dedicated orbital CT is performed, thin sections (usually < 3 mm and preferably < 1.5 mm) are acquired. Coronal images are especially important in that cross-sectional evaluation of all of the

http://radiologyebook.com

590

PART II  CT and MR Imaging of the Whole Body

Horizontal section

Cornea

Zonular fibers (suspensory ligament of lens) Scleral venous sinus (canal of Schlemm)

Capsule of lens

Anterior chamber

Lens Iris

Scleral spur

Posterior chamber Iridocorneal angle Ciliary processes

Ciliary body and ciliary muscle

Bulbar conjunctiva

Ciliary part of retina

Ora serrata

Tendon of lateral rectus muscle

Tendon of medial rectus muscle

Optic (visual) part of retina Vitreous body

Choroid

Hyaloid canal

Perichoroidal space Sclera

Lamina cribrosa of sclera

Fascial sheath of eyeball (Tenon’s capsule)

Optic nerve (II)

Episcleral space Fovea centralis in macula (lutea)

Central retinal artery and vein

Outer sheath of optic nerve Subarachnoid space

FIG 20-3  Illustration of the globe showing the structures of the anterior and posterior segments. The anterior segment is divided into anterior and posterior chambers. (From Netter FH: Atlas of human anatomy, ed 4. St. Louis, 2006, Elsevier, p 87. Netterimages.com Image ID 4627. ©2006 Elsevier Inc. All rights reserved. www.netterimages.com.)

http://radiologyebook.com

CHAPTER 20  Orbit

591

b a

a

A

a

B

b

a

a

C FIG 20-4  CTA of the orbits, superior to inferior (A to C). The ophthalmic artery (a) arises from the internal carotid artery as it exits the cavernous sinus. It enters the orbit through the optic canal and crosses the optic nerve underneath the superior rectus muscle. The superior ophthalmic veins (b) cross the optic nerve more distally and at a more obtuse angle.

intraorbital structures is optimal (e.g., extraocular muscles, optic nerve–sheath–nasal complex, vessels, and globe) (see Fig. 20-1). This plane is also imperative for assessing spread of processes from surrounding structures (e.g., paranasal sinuses, trauma, tumor). A typical orbital CT protocol can be performed with scanning in the axial plane. This plane is usually chosen to be parallel to the orbital long axis. In practice, imaging is performed in the plane parallel to the infraorbitalmeatal line. Coronal images should be included in the routine protocol and can be obtained by multiplanar reformation. This can be performed in the plane perpendicular to the axial plane. Parasagittal reformation in a plane parallel to the long axis of the optic nerve may also be added. Intravenous contrast is often used in the evaluation of inflammatory, infectious, neoplastic, and vascular orbital diseases. For evaluation of vascular lesions, a bolus injection may be used for better depiction of its arterial blood supply. When orbital varix is suspected, the CT study should be repeated without and with the Valsalva maneuver. Enlargement of a lesion with the Valsalva maneuver is indicative of an orbital varix. Less commonly, cavernous hemangiomas may enlarge with the Valsalva maneuver.18 In

patients unable to cooperate, similar effects can be obtained by positioning the patient prone during scanning. CTA can provide good depiction of the major vascular anatomy in the orbits. In addition to the ophthalmic artery and superior ophthalmic vein, their branches, tributaries, and many other smaller vessels can often be seen and traced. The study can be performed as part of a CT angiographic study of the head and neck (see Fig. 20-4).70 Bolus injection of iodinated contrast is required. It is important to use a field of view sufficiently wide to include extraocular pathology that may be associated with the vascular orbital lesions, such as carotid-cavernous fistula. CT dacryocystograms can be performed by administration of contrast material into the nasolacrimal duct to evaluate for patency1 (Fig. 20-5). This requires cannulation of the lacrimal duct, usually by an ophthalmologist.

Magnetic Resonance Imaging MRI of the orbits can be performed with the head coil. For highspatial-resolution imaging of the anterior orbital structures, special orbital surface coils may be advantageous. However, the sensitivity of

http://radiologyebook.com

592

PART II  CT and MR Imaging of the Whole Body

A

B FIG 20-5  A, Coronal reformatted image from a CT dacryocystogram shows dilation of the lacrimal duct (arrow). The nasal septum and inferior turbinate are deviated leftward and cause obstruction at the valve of Hasner. B, The obstruction is only partial, as evidenced by the presence of contrast material in the posterior nasopharynx (arrow). a f e

d

b

c

A d e

f g

c b

h

a

B

C FIG 20-6  A, Coronal T1-weighted MRI study: normal anatomy. Superior ophthalmic vein (a), lateral rectus (b), inferior rectus (c), medial rectus (d), superior oblique (e), superior rectus-levator palpebrae superioris complex (f). B, Axial T1-weighted MRI study: normal anatomy. Optic nerve in the optic canal (a), optic nerve sheath complex (b), medial rectus (c), anterior chamber (d), lens (e), lid (f), medial and lateral aspects of the orbital septum (arrow; g), lateral rectus (h). C, Axial T2-weighted MRI study: posterior visual apparatus. Position of lateral geniculate body (arrows), path of optic radiations (arrowheads).

surface coils decreases rapidly with distance from the coils, leading to rapid signal falloff and inadequate coverage of deeper structures. For routine imaging, the field of view should include the cavernous sinus, optic chiasm, optic tracts and radiations, and the nuclei of the oculomotor, abducens, and trochlear nerves in the midbrain and pons. The protocol should include T1- and T2-weighted imaging in axial and coronal planes (Fig. 20-6). Intravenous gadolinium (Gd) contrast

is routinely used. For dedicated orbital imaging, fat suppression is usually performed for T2-weighted imaging and post-Gd imaging to prevent obscuration of enhancing lesions by the high intraorbital fat signal (Figs. 20-7 and 20-8). Fat suppression is often performed with standard frequency selective presaturation radiofrequency (RF) pulses. Alternative advanced fat-suppression techniques can provide more robust fat suppression in

http://radiologyebook.com

CHAPTER 20  Orbit adverse conditions when magnetic field inhomogeneity is encountered. These techniques may be based on specialized RF pulses (e.g., adiabatic pulses) or modifications of the Dixon technique (e.g., iterative decomposition of water and fat with echo asymmetry and the least-squares estimation [IDEAL]).14 The fat suppression for fluidsensitive imaging (i.e., T2-weighting) can also be performed effectively using inversion recovery32 (Fig. 20-9). Orbital MRI is susceptible to image artifacts because of several factors.27 First, chemical shift artifacts may be seen at the interface of the orbital fat and the globe. Similar artifacts may also be present if silicone oil is used to fill the globe in treatment of retinal detachment. These chemical shift artifacts can be reduced by using fat or silicone saturation, using a higher gradient strength, or narrowing the bandwidth. Second, the proximity of orbital structures to the air cavities of

a h

b

c

e f

FIG 20-7  Coronal post-Gd T1-weighted MRI study with fat saturation: normal anatomy. a, superior rectus/levator palpebrae superioris complex; b, superior ophthalmic vein; c, optic nerve; d, lateral rectus; e, dural sheath; f, inferior rectus; g, medial rectus; h, superior oblique.

A

paranasal sinuses makes orbital imaging susceptible to image artifacts. Exogenous metallic materials (e.g., cosmetics) can also lead to susceptibility artifacts. Third, motion artifacts may be present. To minimize motion of the globe, a patient can be asked to fixate his or her vision at a certain object when the eyes are open. Temporal averaging can also be performed. MR dacrocystography can be performed in a similar fashion as CT dacrocystography, by cannulation and instillation of Gd contrast material into the nasolacrimal duct.46 It has comparable sensitivity to CT dacrocystography.46

APPROACH TO DIFFERENTIAL DIAGNOSIS A large number of disease processes can involve the orbits, and orbital complaints such as proptosis, orbital pain, visual loss, and ophthalmoplegia are nonspecific. Proptosis is abnormal protrusion of the globe; exophthalmos is abnormal prominence of the globe. On imaging, proptosis is best evaluated at a level of the lens on axial images. A line connecting the most distal tips of the lateral orbital walls is drawn. The distance from the anterior margin of the globe to this line should not exceed 21 mm.30

General Principle d

g

593

Various characteristics of an orbital lesion can be used to help construct a differential diagnosis. These include its location, anatomic structure, and imaging features and the clinical presentation of the patient.23,36,38,62 Using a compartmental approach a lesion is first localized to one of the four compartments: globe, optic nerve–sheath complex, intraconal space, or extraconal space. Differential diagnosis of an extraconal lesion can be further refined if it can be determined to be associated with the lacrimal gland and apparatus. Once the primary location of a lesion is determined, other parameters including imaging features (e.g., characteristics of margin, associated bony changes, enhancement patterns), pathophysiologic basis, age of presentation, and chronicity can be considered to further reduce the differential diagnosis. The presence of calcification may also be helpful in refining the differential diagnosis, especially for globe lesions.

B FIG 20-8  A, Precontrast axial T1-weighted image performed without fat suppression demonstrates a mass at the left orbital apex in a patient with known cutaneous lymphoma. B, Postcontrast T1-weighted image of the same patient in which the fat-saturation pulse failed to suppress the orbital fat (this may be due to dental artifact). The lesion demonstrates marked contrast enhancement and is now indistinguishable from the high signal of the orbital fat. This case illustrates the importance of performing nonsuppressed precontrast images and fat-suppressed postcontrast studies.

http://radiologyebook.com

594

PART II  CT and MR Imaging of the Whole Body BOX 20-2  Differential Diagnosis of

Globe Lesions23,36

Congenital Persistent hyperplastic primary vitreous Coats’ disease Coloboma Globe hypoplasia/aplasia

b

a

Degenerative Optic nerve drusen Phthisis bulbi Staphyloma

c

Trauma Vitreous hemorrhage Choroidal hematoma Choroidal effusion Foreign body Inflammatory Orbital pseudotumor (uveal/scleral thickening) Sclerosing endophthalmitis (Toxocara canis) FIG 20-9  Coronal inversion recovery fast-spin echo MRI study: normal anatomy. a, optic nerve; b, subarachnoid space with cerebrospinal fluid; c, optic nerve sheath.

Obviously some lesions may extend over more than one compartment. Nevertheless this compartmental approach helps simplify the diagnostic thought process. The optic nerve–sheath complex, strictly speaking, is also an intraconal structure. However, because of its unique significance, it can be considered a separate compartment to improve the specificity of the differential diagnosis. Differential diagnosis of orbital lesions is summarized in Boxes 20-2 to 20-7. Apart from aiding differential diagnosis, localization of a lesion in the extraconal space versus the intraconal space may also have management implications. In general, intraconal lesions may require surgical attention, whereas extraconal lesions may be amenable to medical management.

Specific Clinical Scenarios It may be helpful to give a few clinical scenarios their own differential consideration. The first is lesions of the lacrimal gland and apparatus.1 Lacrimal lesions are most often benign inflammatory processes, with tumors being less common.69 Viral adenitis is the most common acute process. More chronic inflammatory processes include sarcoidosis, Wegener’s granulomatosis, and Sjögren’s syndrome. Histologically the lacrimal gland is analogous to the minor salivary gland in other regions of the head and neck. They therefore share many common pathologic processes. Most lacrimal gland tumors are epithelial cell tumors, with half of these being benign mixed tumors and half carcinomas. Lymphoma also occurs commonly at the lacrimal gland fossa (see further discussion under “Lacrimal Gland and Apparatus”). In a young patient presenting with leukokoria, one will need to exclude retinoblastoma. Other differential considerations include developmental and congenital conditions such as retinopathy of prematurity, Coats’ disease, persistent hyperplastic primary vitreous, toxocariasis, retina dysplasia, and congenital retinal fold.

Neoplasm Uveal melanoma (adults) Retinoblastoma (children) Metastasis Choroidal hemangioma Medulloepithelioma

BOX 20-3  Differential Diagnosis of Globe

Lesions Associated with Calcification23,36 Congenital Degenerative Cataracts Optic nerve drusen Phthisis bulbi Retinal detachment (chronic) Retrolental fibroplasia Calcification of ciliary muscle insertion Iatrogenic (e.g., scleral banding) Trauma Foreign body

Inflammatory Infection (cytomegalovirus, herpes simplex, rubella, syphilis, toxoplasmosis, tuberculosis) Neoplasm Astrocytic hamartoma (neurofibromatosis, tuberous sclerosis, von HippelLindau syndrome) Retinoblastoma (children) Choroidal osteoma Metabolic Hypercalcemia Sarcoidosis

http://radiologyebook.com

CHAPTER 20  Orbit

595

BOX 20-4  Differential Diagnosis of Optic

BOX 20-5  Differential Diagnosis of Conal/

Trauma Contusion Hematoma Optic nerve avulsion

Trauma Hematoma Foreign body

Nerve Sheath Lesions23,36

Intraconal Lesions23,36

Infection Cellulitis Abscess

Infection Toxoplasmosis Tuberculosis Syphilis

Noninfectious Inflammatory Thyroid ophthalmopathy Pseudotumor Sarcoidosis Wegener’s granulomatosis

Noninfectious Inflammatory Thyroid ophthalmopathy Optic neuritis Pseudotumor Sarcoidosis

Vascular Carotid-cavernous fistula Venous varix Superior ophthalmic vein thrombosis Venous angioma Arteriovenous malformation Cavernous hemangioma (adults) Capillary hemangioma (children) Lymphangioma

Vascular Central retinal vein occlusion Neoplasm Optic nerve glioma Meningioma Neurofibroma Schwannoma Lymphoma/leukemia Metastasis Hemangioblastoma Hemangiopericytoma

Neoplasm Lymphoma Metastasis Rhabdomyosarcoma (children) Hemangiopericytoma Neurofibroma/schwannoma (cranial nerve III, IV, VI) Ectopic meningioma

Miscellaneous Increased intracranial pressure Optic hydrops

Orbital diseases can be categorized based on their pathophysiology: trauma, infection, noninfectious inflammation, neoplasm, vascular lesions, congenital and developmental abnormalities, and degenerative conditions. A more popular approach for differential diagnosis is based on a compartmental approach (see earlier).

Evaluation for foreign bodies is best performed with thin-section CT (Fig. 20-11). Wood fragments pose a challenge to CT evaluation because they may have variable densities owing to differences in hydration. Wood may appear hypodense, isodense, or hyperdense. Air may be present within a wood fragment.59 Therefore unusual air pockets should be evaluated carefully. In certain situations, wooden and organic foreign bodies may be better evaluated with MRI.39

Trauma

Infection

CT is the imaging method of choice for evaluation of orbital trauma. The most common traumatic injury is fracture of the orbital walls. Less commonly, hemorrhage in the globe, globe rupture, perforation and penetrating injury, and contusion or avulsion of the optic nerve sheath may occur. A common type of orbital fracture is “blowout” fracture, which results from increased intraorbital pressure transmitted to the orbital walls secondary to blunt trauma (Fig. 20-10). Blowout fractures most often involve the inferior and medial walls because they are the thinnest. Inferior blowout fracture commonly involves the infraorbital foramen, which is the weakest point of the orbital floor. Intraorbital soft tissue contents may herniate through the fracture. Muscle entrapment is a potential complication of orbital fractures. Because the extraocular muscles are tethered to the orbital walls by tiny fibrous strands that are too small to image on CT or MRI, muscle entrapment may occur even without herniation of the muscle itself.37

Orbital infection is most often caused by direct extension from adjacent structures; hematogenous infection is less common. It is important to localize orbital infection to the following compartments: (1) preseptal versus postseptal and (2) extraconal versus intraconal. Preseptal or extraconal infection can usually be treated by standard antimicrobial therapy. Postseptal or intraconal infection requires more aggressive management because of the risk of neurovascular injury and further intracranial spread.22 Identification of orbital abscesses is also crucial because they may require surgical intervention. Orbital infection is most commonly caused by contiguous spread of sinusitis or a superficial periorbital cellulitis of the face. In children, infection is most commonly secondary to extension from ethmoid air cells (Fig. 20-12), whereas in adults, extension from the frontal sinus is most common67 (Fig. 20-13). Common organisms include Streptococcus pneumoniae and β-hemolytic streptococci. Haemophilus influenzae, staphylococci, and anaerobes are less common.

PATHOPHYSIOLOGY

http://radiologyebook.com

596

PART II  CT and MR Imaging of the Whole Body

BOX 20-6  Differential Diagnosis of

BOX 20-7  Differential Diagnosis of

Trauma Fracture Hematoma Infection Cellulitis Abscess

Trauma Hematoma

Extraconal Lesions23,36

Lacrimal Gland and Apparatus Lesions1,23,36

Infection Dacryoadenitis Noninfectious Inflammatory Pseudotumor Postviral syndrome Sarcoidosis Sjögren’s syndrome Mikulicz’s syndrome Wegener’s granulomatosis

Noninfectious Inflammatory Pseudotumor Postviral syndrome (lacrimal gland) Sjögren’s syndrome (lacrimal gland) Mikulicz’s syndrome (lacrimal gland) Neoplasm Metastasis Primary malignancy from adjacent structures Benign mixed tumor (lacrimal gland) Adenoid cystic carcinoma (lacrimal gland) Non-Hodgkin’s lymphoma Rhabdomyosarcoma (children)

Neoplasm Papilloma Benign mixed tumor (pleomorphic adenoma) Adenoid cystic carcinoma Mucoepidermoid carcinoma Adenocarcinoma Malignant mixed tumor Undifferentiated carcinoma Squamous cell carcinoma Sebaceous carcinoma Primary malignancy from adjacent structures Non-Hodgkin’s lymphoma Metastasis Dermoid/epidermoid

Congenital Cephalocele Dermoid/epidermoid

Congenital Dacryocele Dacryocystocele

m

m

FIG 20-11  CT scan shows intraconal metallic foreign bodies just FIG 20-10  CT scan shows blowout fracture of the left orbital floor with herniation of extraconal fat and inferior rectus muscle (arrow). m, maxillary sinus.

medial to the medial rectus (large arrow) and intraocular (small arrow). Scleral band in place in right globe (arrowheads).

http://radiologyebook.com

CHAPTER 20  Orbit CT is the imaging modality of choice because it can demonstrate inflammatory soft tissue changes, fluid collections/abscesses, and bone changes (e.g., osteomyelitis). Imaging findings vary from mild mucoperiosteal thickening or elevation to frank intraorbital abscesses.

Inflammation

Thyroid-Associated Ophthalmopathy.  Thyroid-associated ophthalmopathy is an autoimmune-mediated inflammation of the extra-

FIG 20-12  CT scan shows an extraconal subperiosteal abscess (small arrowheads). This is a complication of ethmoid sinusitis. A thickened displaced medial rectus (large arrowhead) and preseptal soft tissue swelling (arrows) are seen.

597

ocular muscles and periorbital connective tissues. It is most often associated with Graves’ disease, although association with other thyroid diseases such as Hashimoto’s thyroiditis, thyroid carcinoma, and neck irradiation has also been reported. In approximately 10% to 20% of patients, thyroid-associated ophthalmopathy may present before any other clinical symptoms or signs. Clinical presentation may include eyelid retraction, proptosis, chemosis, periorbital edema, and impaired ocular motility. The classic imaging findings are fusiform enlargement of the extraocular muscles with sparing of the tendinous attachments (which may be difficult to appreciate when muscle enlargement is severe), and inflammatory changes of the periorbital fat (Fig. 20-14). There may be increase of intraorbital fat. The lacrimal glands may also be affected. The extraocular muscles most often affected, in descending order of frequency, are the inferior rectus, medial rectus, and superior rectus– levator palpebrae muscle complex.52 Periorbital soft tissue swelling and proptosis may be seen. Involvement is usually bilateral but may be asymmetric. Direct involvement of the globe and optic nerve sheath is uncommon. However, secondary compression of the optic nerve sheath may occur and can lead to irreversible visual loss. It is important to assess for this possibility on imaging.

Orbital Pseudotumor.  Orbital pseudotumor is also known as idiopathic orbital inflammatory disease. It is an idiopathic nongranulomatous inflammatory process that often involves the extraocular muscles and orbital fat. Less frequently, other intraorbital structures including the uveal tract, sclera, optic nerve, and lacrimal glands may be involved.13,17,47,48 Involvement of these structures can occur in isolation without extraocular muscle involvement. Tolosa-Hunt represents a variant form of pseudotumor that involves the cavernous sinus. When muscular involvement is present, it is usually diffuse. As opposed to thyroid-associated ophthalmopathy, involvement is often unilateral and there is usually extension to the muscular tendon attachments (Figs. 20-15 and 20-16). Orbital pseudotumor is usually painful, which helps distinguish it from thyroid-associated ophthalmopathy. Orbital pseudotumor may be difficult to differentiate from other tumefactive inflammatory processes and neoplasms. A quick response to a trial steroid therapy may help establish the diagnosis.49 Alternative diagnosis needs to be excluded when there is poor response to therapy or recurrence.

Sarcoidosis.  Sarcoidosis is a noninfectious granulomatous disease FIG 20-13  CT scan shows subperiosteal abscess of the superior orbit (arrowheads) from ethmoid or frontal sinus disease.

A

that may affect any part of the optic pathway from the globe to the optic radiations.9 The lacrimal gland, anterior layer of the globe, and

B FIG 20-14  A, Axial contrast-enhanced CT scan shows Graves’ ophthalmopathy characterized by enlarged superior, medial, and inferior recti with compromise of the orbital apex. Note the sparing of the muscle tendon insertions. B, Coronal contrast-enhanced CT scan in the same patient.

http://radiologyebook.com

598

PART II  CT and MR Imaging of the Whole Body

A

B FIG 20-15  A, Axial CT scan shows pseudotumor of the orbit with swollen bilateral medial rectus (arrows), which includes tendinous insertion on the globe. Thickening and enhancement of the globe (arrowheads) are also shown. B, Coronal CT scan shows pseudotumor of the orbit in the same patient.

FIG 20-16  CT scan of orbital pseudotumor with bilateral medial rectus and left lacrimal involvement.

FIG 20-18  Axial T1-weighted MRI study with Gd shows sarcoid of the chiasm (long arrow), left cerebral peduncle (large arrowhead), lateral geniculate body (short arrow), and superior colliculus (small arrowhead).

On MRI, retrobulbar masses may demonstrate marked hypointensity. Although this finding is not unique, it strongly suggests Wegener’s granulomatosis.12,56 FIG 20-17  Axial T1-weighted MRI study with Gd shows sarcoid of the anterior left globe (arrowheads).

eyelids are commonly involved. The imaging findings can simulate pseudotumor (Figs. 20-17 and 20-18).

Wegener’s Granulomatosis.  This is a form of necrotizing granulomatous vasculitis. Orbital involvement is common and seen in slightly more than 50% of patients.25,31,51 Any orbital structures can be involved. Findings may include conjunctivitis, episcleritis, scleritis, uveitis, optic nerve vasculitis, retinal artery occlusion, nasolacrimal duct obstruction, and retrobulbar diseases. CT examination may demonstrate nonspecific nodules or infiltrates in the retrobulbar space.56,66 Enhancement is generally present.

Optic Neuritis.  Optic neuritis represents nonspecific inflammation of the optic nerve that can be associated with infection, granulomatous diseases, pseudotumor, postradiation, or demyelinating diseases. A large proportion of cases are idiopathic. Association with multiple sclerosis is established in approximately 50% of patients.3 Imaging findings are best demonstrated on MRI, which may include enhancement and T2 prolongation (Fig. 20-19). These findings can be subtle. It is important to include the whole brain in image evaluation to exclude intracranial lesions, particularly the presence of demyelinating lesions.

Neoplasms

Lymphoma.  Lymphoma is the most common neoplasm in the orbit, accounting for just over half of all cases.69 B-cell lymphomas of the non-Hodgkin’s type are by far the most common, although T-cell lineages have also been described.11 Usually orbital lymphomas are

http://radiologyebook.com

CHAPTER 20  Orbit primary to the orbit, but occasionally orbital manifestation of a systemic lymphoproliferative process is seen. The usual appearance is a well-defined mass within the muscle cone (Figs. 20-20 and 20-21; see Fig. 20-8). Less frequently, extraconal masses or diffuse infiltration of the orbital fat can be seen. The differential diagnosis includes pseudotumor and metastasis. Lacrimal gland involvement can be seen either in isolation or in

599

combination with the other manifestations of lymphoma within the orbit.

Optic Glioma.  Optic gliomas most often occur in children, especially between the ages of 2 and 6 years. They are usually benign, but a small number of lesions may develop aggressive behavior.73 This lesion usually involves the anterior optic apparatus (e.g., optic nerves, chiasm, and optic tracts) and causes enlargement and often tortuosity of these structures. About half of all optic gliomas occur in patients with neurofibromatosis type 1, and 10% to 15% of neurofibromatosis type 1 patients develop optic gliomas15 (Fig. 20-22). These lesions do not calcify.8 MRI has become the modality of choice, given the necessity of evaluating the intracranial extent of the tumor (Fig. 20-23). Optic gliomas are typically either nonenhancing or weakly enhancing. The lesions are generally isointense to slightly hypointense on T1-weighted images and hyperintense on T2-weighted images.21 CT can help in assessing bony changes and is especially valuable in detecting expansion of the optic canal. CT thus complements MRI in evaluation of these lesions.

Optic Nerve Sheath Meningioma.  Optic nerve sheath meningioFIG 20-19  Coronal MRI study, inversion recovery fast-spin echo, shows left optic neuritis. Contrast the increased signal at the left optic nerve (long arrow) with the low signal of the normal right optic nerve (short arrow).

A

mas (ONSMs) are meningiomas that arise from the meninges surrounding the optic nerve. It is not an uncommon tumor, making up between 5% and 7% of primary orbital tumors.33 The onset occurs at a median age of 38 years and is seen four times more frequently in

B

FIG 20-20  A, T1-weighted axial MRI study shows a large mass

C

replacing the intraconal fat in the right orbit (arrow). B, T2-weighted axial image of the same patient. C, Gd-enhanced axial T1-weighted image of the same patient demonstrates marked enhancement (black arrow). Fat suppression allows the smaller lesion to be visible at the apex of the left orbit (white arrow).

http://radiologyebook.com

600

PART II  CT and MR Imaging of the Whole Body

B

A

M M

M M

C FIG 20-21  Orbital lymphoma. A, Contrast-enhanced CT scan demonstrates a homogeneously enhancing intraconal mass (black arrow) adjacent to the left optic nerve, causing medial deviation of the nerve (white arrow). B, Axial post-Gd fat-suppressed T1-weighted image confirms the CT findings (long arrow shows the enhancing mass; short arrows show the optic nerve). C, Coronal post-Gd fat-suppressed images more clearly demonstrate the enhancing mass (long arrow) separate from the nonenhancing left optic nerve (short arrow). M, extraocular muscles.

FIG 20-22  Axial T1-weighted MRI study shows optic glioma of bilateral optic nerves, with involvement of the chiasm (arrows) in a patient with neurofibromatosis.

females than males.33 Because meningiomas in general occur more frequently in patients with neurofibromatosis type 2, ONSM also occurs more frequently in these patients. The presenting symptom with ONSM is usually diminished visual acuity from optic nerve compression or proptosis. If there is no evidence of visual loss or intracra-

nial extension, these lesions are often treated by close observation. In the setting of visual loss, radiation treatment is frequently used. Surgery is usually reserved for intracranial extension and larger tumors. On axial imaging, the most common presentation is the wellknown tram track appearance caused by the enhancing tumor wrapping around the sheath (Fig. 20-24). Inflammation of the dura from other causes may occasionally have a similar appearance. ONSM can also present a fusiform enlargement of the sheath on one side (Fig. 20-25). As with all meningiomas, they enhance vividly with contrast and often demonstrate calcification. Hyperostosis may occasionally be seen when the lesion is at the orbital apex or in the optic canal. Optic nerve glioma may initially have the appearance of a meningioma on the axial images, but on the coronal fat-suppressed enhanced MRI,  the nerve should be seen separate from the surrounding enhancing meningioma. Although MRI is the imaging modality of choice, thin-section CT is often helpful because it demonstrates the calcifications or hyperostosis that may be present, thus aiding in the differential diagnosis. A noncontrast CT scan should be performed first so that the enhancing tumor does not hide the calcifications. ONSMs, particularly when in the optic canal, can be quite small and yet cause significant symptoms. As a result they can be easily missed unless there is a high degree of suspicion and careful inspection. MRI can therefore be extremely useful for finding these lesions. Coronal and axial contrast-enhanced MRI with fat suppression allows the enhancing lesion to be seen against the fat and bone, which turn dark.

http://radiologyebook.com

CHAPTER 20  Orbit

A

B FIG 20-23  A, T1-weighted axial MRI study with Gd enhancement shows optic glioma of the optic chiasm (arrows). B, T1-weighted coronal MRI study with Gd.

A

B FIG 20-24  A, Coronal contrast-enhanced CT scan shows optic nerve sheath meningioma (arrow). B, Axial contrast-enhanced CT scan shows same patient with “tram track” appearance (arrow).

g

m

FIG 20-25  Axial contrast-enhanced CT scan shows optic nerve sheath meningioma (m). g, globe; arrow, displaced optic nerve sheath complex emerging from mass.

http://radiologyebook.com

601

602

PART II  CT and MR Imaging of the Whole Body

FIG 20-26  Coronal contrast-enhanced CT scan shows melanoma of the left ciliary body.

Melanoma.  Primary orbital melanoma usually presents as an ocular lesion. It originates in the uveal tract (iris, choroid, and ciliary bodies) and may extend posteriorly to the rest of the orbit. On CT imaging, melanomas appear as focal soft tissue masses with mild to moderate enhancement45 (Fig. 20-26). MRI studies may help differentiate melanomas from other ocular lesions, evaluate their intraorbital extent, and search for metastatic disease.55 On MRI studies the amount of melanin contained in melanoma determines the signal characteristics. Melanin shortens T1 and T2, thereby causing increased signal on T1-weighted images and mildly decreased signal on T2-weighted images (Fig. 20-27). MR signal is also affected by hemorrhage, which is not uncommon in patients with melanotic lesions. With hemorrhage the differential diagnosis includes retinal and choroidal detachment from other causes (Fig. 20-28). The presence of Gd enhancement favors melanomas and helps differentiate the tumor from retinal detachment.6,44 Although melanotic lesions have characteristic appearances, nonpigmented melanomas cannot be reliably differentiated from other masses.58

Metastatic Disease.  In adults the most common tumor to metastasize to the orbit is breast carcinoma. Other primary sites include lung, colon, and prostate (Fig. 20-29). In children the most common primary lesions include neuroblastoma, leukemia, and Ewing’s sarcoma. Metastatic lesions may affect any of the intraorbital structures as well as the bony orbit itself29,64 (Fig. 20-30). None of the available imaging techniques offers specificity to differentiate metastases from the many other orbital lesions. The findings may be subtle, with small areas of focal thickening of the globe, or large destructive lesions.23 In addition, extension of tumor from an adjacent structure (e.g., paranasal sinuses) may occur (Fig. 20-31). Enophthalmos may be present in primary disease that is often associated with extensive fibrous response, such as scirrhous carcinoma of the breast.

Retinoblastoma.  Retinoblastoma is seen primarily in infants and has an occurrence of 1 in every 18,000 to 30,000 live births.54 It is responsible for 1% of all childhood cancer-related deaths in the United States.63 Early diagnosis extends the 5-year survival rate to more than 90%; however, if the tumor extends beyond the globe, the mortality rate approaches 100%.35 Retinoblastoma has been strongly linked to mutations on the RB1 allele of chromosome 13. Whereas about 10% of cases are said to be inherited, most retinoblastoma cases are not inherited. Hence there is both a familial hereditary form of retinoblastoma and a nonfamilial sporadic form. Aside from the hereditary  differences, the tumors are the same. Patients with nonfamilial retino-

FIG 20-27  Ocular melanoma of the inferior aspect of the globe. Top, T1-weighted coronal MRI study. Bottom, T2-weighted coronal MRI study.

blastoma have unilateral solitary tumors, whereas patients with the familial form have a much higher rate of bilateral than unilateral disease. Patients with the familial form of retinoblastoma have a high incidence of nonocular cancers as well. The term “trilateral retinoblastoma” refers to a patient who has bilateral retinoblastomas and a third midline tumor. The midline tumor is histologically the same as the intraocular tumor and may occur in the pineal region, suprasellar region, or fourth ventricle. It is important to remember that the midline tumor may not be seen at the same time as the ocular tumors but may be discovered several years later. Because most of these lesions that arise from the retina are calcified, CT is extremely important in their diagnosis (Fig. 20-32). The lesions will also enhance with intravenous contrast. The tumor may spread in the lymphatics or along the optic nerve to gain intracranial access. If a tumor is discovered in one globe, very close inspection of the other globe is necessary to exclude bilateral disease. On initial evaluation and on follow-up examinations, close inspection of the pineal region, suprasellar region, and fourth ventricle is important to seek out trilateral disease.

http://radiologyebook.com

CHAPTER 20  Orbit

A

R

603

L

FIG 20-29  Axial contrast-enhanced CT scan shows prostate metastasis to the left orbit roof.

B

FIG 20-30  Axial contrast-enhanced CT scan shows bilateral neuroblastoma metastasis.

C FIG 20-28  A, T1-weighted sagittal MRI study shows retinal detachment with hemorrhage. B, T2-weighted axial MRI study of the same patient as in A. C, Coronal CT scan (different patient) shows retinal detachment. (A and B courtesy Guy Wilms, MD, Universitaire Ziekenhuizer, Leuven, Belgium.)

A

B FIG 20-31  A, Axial contrast-enhancedhttp://radiologyebook.com CT scan shows extension of squamous cell carcinoma of the maxillary sinus to the orbit. B, Coronal contrast-enhanced CT scan in the same patient.

604

PART II  CT and MR Imaging of the Whole Body

R

FIG 20-32  Axial contrast-enhanced CT scan shows calcified retinoblastoma of the left eye.

FIG 20-33  Rhabdomyosarcoma of the orbit in a young child. CT demonstrates enhancing soft tissue mass (R). There is aggressive bone destruction, with the tumor extending into the ethmoid sinus (arrow).

Because these tumors enhance, MRI with contrast and fat suppression is excellent for identifying the lesion; however, CT is better at identifying the calcification. With bilateral disease the diagnosis is easy, but with a unilateral tumor it may be more difficult. Entities such as Coats’ disease and Toxocara canis infection can be confused with retinoblastoma, but these typically lack contrast enhancement.

Rhabdomyosarcoma.  Even though the most common malignant ocular tumors in children are retinoblastomas, the most common malignant orbital tumors in children are rhabdomyosarcomas.10 They may arise primarily or secondarily in the orbits. They are very aggressive tumors and may grow rapidly. On CT imaging, they are seen as enhancing soft tissue masses with associated permeative or lytic bone destruction (Fig. 20-33). On MRI they are hypointense to isointense on T1-weighted images and isointense to hyperintense on T2-weighted images. Enhancement is variable.2

Langerhans Cell Histiocytosis.  Langerhans cell histiocytosis is not a true neoplasm but a reticuloendothelial disorder of unknown origin. Like rhabdomyosarcoma, it occurs most often in children. Because its clinical presentation and imaging features are often similar to neoplastic processes, it is often included in the differential consideration of a soft tissue mass. The imaging appearance can simulate rhabdomyosarcoma.10 On CT imaging an isodense to hyperdense soft tissue mass is seen. Enhancement of the lesion and associated lytic bone changes are usually present.2 On MRI study the mass is isointense to hypointense on T1-weighted images and isointense to hyperintense on T2-weighted images.

Teratoma.  The teratoma is a rare benign lesion that contains mixed endodermal, mesodermal, and ectodermal elements. It usually calcifies. Because the teratoma is usually seen in neonates, knowing a patient’s age can help one decide whether to include this entity in a differential diagnosis26,74 (Fig. 20-34).

Vascular Abnormalities

Carotid Cavernous Fistula.  Carotid cavernous fistula is an abnormal high-flow communication between the arterial and venous circulations. This results in transmission of arterial flow into the cavernous sinuses, consequently leading to reversal of flow in venous structures draining into the cavernous sinus. Two types of carotid cavernous fistulas have been described: direct and indirect. The more common type, direct fistula, is formed by an abnormal communication between the internal carotid artery and the cavernous sinus. The less common

FIG 20-34  Axial contrast-enhanced CT scan shows calcified teratoma with areas of enhancement.

indirect fistula is an abnormal communication between the dural external carotid artery and the cavernous sinus. Common causes of direct carotid cavernous fistula include rupture of the cavernous internal carotid artery secondary to either trauma or spontaneous rupture of an aneurysm. CT or MRI findings include proptosis and engorgement of the superior ophthalmic vein and cavernous sinus (Fig. 20-35). The abnormally high flow in the cavernous sinus may lead to the appearance of flow void on MRI. There is usually enlargement of extraocular muscles due to impaired venous drainage. Imaging abnormalities may be seen bilaterally if intercavernous venous connections are present. On CTA, engorgement of the cavernous sinuses and ophthalmic veins can be seen. The venous engorgement often extends to the branches of the facial veins through their anastomoses with the ophthalmic veins (Fig. 20-36). Indirect carotid cavernous fistula can be caused by trauma. The imaging findings are more subtle than those for direct carotid cavernous fistula.

http://radiologyebook.com

CHAPTER 20  Orbit

605

F

c

m

A

B FIG 20-35  Carotid cavernous fistula. A, Coronal T1-weighted MRI study shows marked enlargement of the left superior ophthalmic vein (arrow). B, Lateral digital subtraction angiogram shows fistulous communication between branches of the external carotid and the cavernous sinus. m, middle meningeal artery; F, fistula; c, cavernous sinus.

FIG 20-36  CTA of posttraumatic carotid cavernous fistula. Note the enlarged superior ophthalmic veins (long arrows) and branches of the facial veins in the periorbital and maxillofacial regions (short arrows).

Cavernous hemangiomas are the most common adult orbital mass lesions (Fig. 20-39). They are actually venous malformations rather than true neoplasms. They tend to be well encapsulated with distinct margins and are usually intraconal. On CT imaging, they are often hyperdense. Phleboliths may be present. On MRIs, signal on T1-weighted images are variable, but hyperintensity may be associated with thrombosis. Cavernous hemangiomas are usually hyperintense on T2-weighted images. Intense enhancement is usually seen with both CT and MRI. Capillary hemangiomas are also known as infantile hemangioma, juvenile hemangioma, hypertrophic hemangioma, and hemangioblastoma (Fig. 20-40). Because they represent true vascular neoplasm, previous authors have advocated simply using the term “hemangioma” to refer to these lesions.50 Capillary hemangiomas often are found in infants during the first year of life. They tend to regress with age. Prominent arterial supply from the external or internal carotid arterial systems usually is seen. On CT and MRI, they present as an intensely enhancing lobulated mass with low intensity to isointensity on T1-weighted images and hyperintensity on T2-weighted images.18

Venous Varix.  Venous varix represents abnormal dilatation of

Lymphangioma.  Lymphangiomas usually present at an early age,

orbital veins. The superior ophthalmic vein is most often affected. Etiologies include congenital venous malformation or venous wall weakness; it can also occur in association with intraorbital or intracranial arteriovenous malformation. For evaluation of venous varix, CT is best performed with contrast enhancement, repeated without and with the Valsalva maneuver65,72 (Fig. 20-37). The varix is seen as a large tortuous vein or lobulated masslike confluence of small veins. There is strong contrast enhancement, although heterogeneity can be present if thrombosis is present. Venous varix enlarges significantly with the Valsalva maneuver, helping differentiate it from other enhancing lesions. CTA can also be used to diagnose and map the anatomic course of an orbital varix71 (Fig. 20-38).

from infancy to the first decade of life. They are congenital vascular malformations, and the term venous lymphatic vascular malformation has also been used.34,61 The lesions may involve the extraconal compartment, intraconal compartment, or both. Histologically they may contain a variable amount of lymphatic channels, lymphocytic aggregates, dysplastic vascular channels, loose connective tissue, and smooth muscle fibers. Hemorrhage of different stages may be present. The imaging appearance is variable and reflects the histologic composition of individual lesions (Figs. 20-41 and 20-42). Most lesions appear poorly defined, multilobulated, and inhomogeneous.20,24 Septations may be present. Enhancement is variable. On CT imaging, phleboliths or dystrophic calcifications may be seen occasionally.18 MRI can provide better details of these lesions. On T1- and T2-weighted images, both hyperintensity and hypointensity may be seen.5

Hemangiomas.  Hemangiomas can be classified into two distinct entities: cavernous and capillary.5 They are seen in different age groups; cavernous hemangiomas are seen in young adults, most commonly between the second and fourth decades of life, whereas capillary hemangiomas are seen in young pediatric patients.

Arteriovenous Malformation.  Most orbital arteriovenous malformations are extensions of intracranial lesions with drainage into the orbital veins via the cavernous sinus.19 On CT or MRI, multiple

http://radiologyebook.com

606

PART II  CT and MR Imaging of the Whole Body

A

B FIG 20-37  Axial contrast-enhanced CT scan. Orbital varix both before (A) and during (B) Valsalva maneuver demonstrates increase in size of the lesion. (Courtesy Robert E. Peyster, MD, Hahnemann Hospital, Philadelphia.)

A

B

C

D

E

F

G

H

I

J

K

L

M

N

FIG 20-38  Contiguous axial CTA images of right orbital varix, from superior orbit (A) to inferior orbit (N). A, Superior ophthalmic vein identified by white arrow. B to F, Small serpentine vein (arrows) arises from inferior margin of superior ophthalmic vein. G to J, This courses medially and inferiorly into larger upper lobe of lesion (white arrows). I to K, Inferior ophthalmic vein (black arrow) courses into dilated venous confluence. L to N, Venous confluence demonstrates blind-ended anterior extending saccular dilation (arrow), representing lower lobe of lesion. M, Contrast level in lower lobe of lesion (arrow) indicates venous stasis. N, Venous confluence drains through superior orbital fissure into cavernous sinus (arrow). (From White JH, et al: Diagnosis and anatomic mapping of an orbital varix by computed tomographic angiography. Am J Ophthalmol 140:945–947, 2005.)

tortuous vessels can be seen. The superior ophthalmic vein may also be dilated.

Congenital and Developmental Abnormalities and Anomalies

Coloboma.  Coloboma is a congenital defect in the globe at the insertion of the optic nerve. It results from incomplete or inadequate fusion of the fetal optic fissure and occurs along the inferomedial aspect of the globe and optic nerve.47 The defect can be bilateral. On CT or MRI the affected globe is usually small. The defect can be seen as a cystic outpouching of vitreous at the site of optic nerve attachment to the globe (Fig. 20-43). A retroocular cyst may be associated. It may present as a small cone-shaped defect at the optic disc or as a large retinal cyst.

Retinopathy of Prematurity.  Retinopathy of prematurity is also known as retrolental fibroplasia. It is seen in premature infants requiring long-term ventilation support. Prolonged exposure to high oxygen concentration leads to proliferation of the retinal vascular buds. Involvement is usually bilateral but may be asymmetric. On CT imaging, there is increased density in the vitreous bilaterally. Microphthalmia is often seen, but calcification is uncommon except in advanced cases; these two features distinguish retinopathy of prematurity from retinoblastoma.

Persistent Hyperplastic Primary Vitreous.  Persistent hyperplastic primary vitreous is caused by hyperplasia of the residual primary vascular vitreous, which usually involutes by the second trimester. This

http://radiologyebook.com

CHAPTER 20  Orbit

A

C

FIG 20-39  A, Axial T1-weighted MRI study shows cavernous

B

A

hemangioma of the medial right orbit. B and C, Axial T2-weighted MRI study in the same patient.

B

FIG 20-40  Capillary hemangioma. A, Coronal inversion recovery

C

fast-spin echo image demonstrates lobulated hyperintense mass occupying both intraconal and extraconal space. Note the dark flow voids in the lesion (white arrow) indicating the presence of distinct blood vessels. There is extraorbital extension (black arrow). B, Coronal fat-suppressed post-Gd image demonstrates intense enhancement of the orbital (white arrows) and extraorbital lesions (black arrow). C, Reformatted coronal CTA confirms the presence of small vessels in the enhancing orbital mass (arrow).

http://radiologyebook.com

607

608

PART II  CT and MR Imaging of the Whole Body

L

L

A FIG 20-42  Recurrent lymphangioma in a 34-year-old woman. Reformatted axial image demonstrates lobulated extraconal lesion with an anterior and a posterior component (L). The lesion is heterogeneous with mild wall enhancement. The posterior component is lateral and abutting the medial rectus muscle (arrow). Note the orbital wall is intact, consistent with a nonaggressive slow-growing process.

B

FIG 20-43  Axial CT scan shows left orbit coloboma. Note the cystic outpouching is at the site of optic nerve attachment (arrow), distinguishing this from staphyloma.

C FIG 20-41  Lymphangioma. A lobulated lesion is seen in the medial aspect of extraconal space. A, Coronal T1-weighted image. B, Coronal T2-weighted image. C, Coronal CT image.

condition may present in male infants as unilateral leukokoria. Therefore retinoblastoma is a strong differential consideration. On CT imaging, the globe is small and demonstrates hyperdensity in the vitreous humor. Enhancement is present. A tissue density band may extend from the back of the lens to the posterior inner surface of the globe.

Coats’ Disease.  This is a congenital vascular malformation of the retina characterized by multiple telangiectatic vessels. Leakage of serous and lipoproteinaceous exudates from these vessels leads to retinal detachment and central vision loss. On CT imaging, hyperdensity is seen in part or all of the vitreous. Calcification is usually absent, and the lesion is hyperintense on T2-weighted images, distinguishing this entity from retinoblastoma.42 Enhancement is usually seen in the detached leaves of the retina, which may be displaced to the midline by the massive subretinal effusion.41

Microphthalmia.  Microphthalmia represents congenital underdevelopment or acquired diminution in the size of the globe. Congenital

associations include coloboma and craniofacial anomalies such as hemifacial microsomia. Acquired etiologies include intrauterine infections such as rubella.

Dermoids/Epidermoids.  Dermoid cysts are the most common benign congenital lesion of the orbit. They arise from epithelial rests sequestered in sutures and most often present during the first decade of life.53 Although they are most commonly located near the frontozygomatic suture and the lacrimal fossa, dermoid cysts can be found anywhere in the orbits. When they rupture they can elicit a severe inflammatory response. Imaging features include cysts with welldefined walls that may or may not enhance (Fig. 20-44). Fat is often seen within the lesions, giving rise to pathognomonic fat-fluid levels. Other tissues such as hair or sweat glands may also be present. The lesion may demonstrate hyperintensity on T1-weighted images owing to the presence of fat. Marginal calcification may be present. Epidermoid cysts do not contain fat, sweat glands, or hair. On MRIs they are usually hypointense on T1-weighted images, but they may be hyperintense when hemorrhage or high proteinaceous material is present (Fig. 20-45).

Degenerative Conditions

Calcified Ciliary or Extraocular Muscle Insertions.  Calcification of the insertions of the ciliary muscles (ciliary bodies) or extraocular

http://radiologyebook.com

CHAPTER 20  Orbit muscles is commonly seen in older adults. This is usually an incidental finding on CT studies (Fig. 20-46).

Phthisis Bulbi.  Phthisis bulbi represents the end stage of injury to the globe. This may be secondary to trauma, infection, or inflammatory disease. On imaging, the globe is usually shrunken, atrophic, and misshapen. Heavy calcification is usually present (Fig. 20-47).

Optic Drusen.  Optic drusen is usually seen in adults. The lesion may present with prominence of the optic disc that can be misinterpreted as papilledema during funduscopic examination (pseudopapilledema).60 Drusen represents a collection of hyaline-like material on the surface of the optic disc. Discrete calcification is usually present,7 making CT an ideal modality for its evaluation57,68 (Fig. 20-48). In 75% of cases there is bilateral involvement.

Retinal and Choroidal Detachment.  The three layers of the globe include the posterior hyaloid membrane, sensory retina, and retinal pigment epithelium layers. Posterior hyaloid detachment represents detachment of the posterior hyaloid membrane from the sensory retina, and retinal detachment represents separation of the sensory retina from the retinal pigment epithelium. Separation of the choroid from the sclera is called choroidal detachment. Choroidal detachment may demonstrate lenticular or moundlike abnormalities. Retinal

attachment may have any shape or size but is more commonly V-shaped with a small attachment at the optic disc (Fig. 20-49; see Fig. 20-28). Etiologies of retinal and choroidal detachment include trauma, infection, inflammation, and neoplasm. It is important to consider the possibility of these underlying etiologies (e.g., malignancies), which may have dire effects.

Staphyloma.  Staphyloma represents focal outpouching of the scleraluveal coats of the globe (Fig. 20-50). As opposed to coloboma, staphyloma defect is located off-center from the optic disc, typically temporal to the disc. It is an acquired defect secondary to weakness and thinning of the scleral-uveal coats. Patients often present with severe axial myopia. Associations include glaucoma, previous inflammatory process, and trauma.

LACRIMAL GLAND AND APPARATUS The unique histology of the lacrimal gland and apparatus gives rise to its own differential consideration. The most common lesions affecting the lacrimal gland are benign inflammatory processes, with tumors being less common.40 Inflammatory, neoplastic, and traumatic processes can affect the lacrimal gland.69 The patient can present with deviation of the globe, proptosis, or conjunctival injection, depending on the extent of the lesion. With the possible exception of traumatic disruption, glandular

FIG 20-44  Fat-suppressed T2-weighted image of dermoid cyst. A small recurrent dermoid cyst presents as a small well-circumscribed heterogeneous lesion at medial aspect of the right orbit (arrow), the second most common location of orbital dermoids.

A

609

FIG 20-46  Calcified ciliary muscle insertion (arrows).

B FIG 20-45  Epidermoid cysts. A, T2-weighted image demonstrates an epidermoid cyst (arrow). Note the homogeneous signal intensity and typical location at frontozygomatic suture. B, Post-Gd image of the cyst (arrow) demonstrates hyperintensity owing to highly proteinaceous material, but no enhancement.

http://radiologyebook.com

610

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 20-47  A, Axial CT scan shows bilateral phthisis bulbi. B, Axial T1-weighted MRI study of the same patient. C, Axial T2-weighted MRI study of the same patient.

C

FIG 20-48  Axial CT scan shows optic disc drusen, right optic nerve head calcification, and optic nerve atrophy.

FIG 20-50  T1-weighted MRI shows staphyloma. Note the thinning of the scleral-uveal outpouching (white arrow), which is temporal to the optic disc (black arrow), thus distinguishing it from coloboma (see Fig. 11-43).

FIG 20-49  CT image showing chronic retinal detachment in the left globe (arrow).

enlargement is common to most lesions. Contrast enhancement does not distinguish between inflammation and neoplasia. In the acute setting, viral adenitis accounts for most of the lesions of the lacrimal fossa and tends to affect a younger population. Chronic disease is usually secondary to granulomatous disease (e.g., sarcoid or

Wegener’s) or autoimmune disease (e.g., Sjögren’s syndrome). Orbital pseudotumor may affect the lacrimal gland. Unilateral or bilateral glandular enlargement as well as variability of enhancement may be seen. The chronic entities tend to be more well marginated than the acute inflammatory processes.40 Tumors of the lacrimal gland may be either benign or malignant. Approximately half of lacrimal neoplasms are tumors of epithelial origin with histopathologic types similar to that of the salivary gland. Benign tumors are most commonly pleomorphic adenoma (also known as benign mixed tumor). They may contain small cystic components. As in the salivary glands, pleomorphic adenoma carries a low accumulative risk of malignant transformation into carcinoma ex pleomorphic (5% at 10 years and 20% at 30 years). This should be considered when more aggressive or invasive features are present.18 Adenoid cystic carcinoma is the most common malignant tumor of the lacrimal gland (Fig. 20-51). Mucoepidermoid carcinoma and adenocarcinoma are also seen. Lymphocytic hyperplasia and lymphoma (Fig. 20-52) make up the bulk of additional tumors.28

http://radiologyebook.com

CHAPTER 20  Orbit

611

F M

M

A

B

FIG 20-51  A, T1-weighted axial MRI study shows adenoid cystic carcinoma (M) in the left lacrimal gland. The lesion demonstrates T1 hypointensity. Note the normal right lacrimal gland (arrow). B, Coronal fat-saturated T1-weighted post-Gd image demonstrates marked homogeneous enhancement of the mass (M). In the absence of invasive features, it is difficult to differentiate this malignant tumor from benign pleomorphic adenoma. C, Invasive features, such as invasion of preseptal soft tissues (arrows), in another patient with adenoid cystic carcinoma suggest a malignant tumor.

C

A

B FIG 20-52  A, Coronal contrast-enhanced CT scan shows left lymphoma with extension into the sinuses and anterior cranial fossa. B, Axial contrast-enhanced CT scan in the same patient.

CT scanning may demonstrate bone involvement in lacrimal tumors, including pleomorphic adenoma, adenoid cystic carcinoma, adenocarcinoma, and mucoepidermoid carcinoma.28 When bone involvement is present, pleomorphic adenoma demonstrates smooth bone scalloping and remodeling. Irregular bone destruction raises the suspicion of malignant tumors. A study by Hesselink and coworkers28 revealed variable enhancement in both inflammatory (Fig. 20-53) and neoplastic lesions, and therefore enhancement is not a reliable predictor of aggressive pathology. Both pleomorphic adenoma and adenoid cystic carcinoma can demonstrate intense enhancement (see Fig. 20-51). The malignant

mucoepidermoid can demonstrate marked enhancement, but the adenocarcinoma may show no enhancement.

INDIRECT INVOLVEMENT OF THE ORBIT AND OPTIC PATHWAYS The orbit or any part of the optic pathways may be affected by adjacent intracranial lesions. The most common lesions arise from the sella turcica and parasellar regions. The differential diagnosis includes pituitary macroadenoma (Fig. 20-54), meningioma, craniopharyngioma, metastasis, aneurysm, chordoma, and other less common entities. The

http://radiologyebook.com

612

PART II  CT and MR Imaging of the Whole Body

REFERENCES

FIG 20-53  Axial contrast-enhanced CT study shows lacrimal gland pseudotumor with extension to preseptal soft tissues on the right (arrowheads).

M

FIG 20-54  Pituitary macroadenoma (M) compressing on the optic chiasm, causing gradual visual loss in this patient. Arrows, optic nerves.

FIG 20-55  Axial T1-weighted MRI study with Gd shows left sphenoid wing meningioma (arrows), with compromise of the optic canal (curved arrow).

posterior optic apparatus—superior colliculi, lateral geniculate bodies (see Fig. 20-18), optic radiation, and occipital lobes—may be affected by any infiltration or mass lesion (e.g., tumor, infarction, or inflammatory process). One of the more common locations of an intracranial meningioma is the sphenoid wing. Visual symptoms can result from proximity to the optic canal or by mass effect on the anterior optic apparatus. Direct extension into the orbit is also possible (Fig. 20-55).

1. Ansari SA, Pak J, Shields M: Pathology and imaging of the lacrimal drainage system. Neuroimaging Clin N Am 15:221–237, 2005. 2. Barnes P, Robson C, Robertson R, et al: Pediatric orbital and visual pathway lesions. Neuroimaging Clin N Am 6:179–198, 1996. 3. Beck RW, Arrington J, Murtagh F: Brain magnetic resonance imaging in optic neuritis: Experience of the optic neuritis study group. Arch Neurol 50:841–846, 1993. 4. Belden CJ: MR imaging of the globe and optic nerve. Magn Reson Imaging Clin N Am 10:663–678, 2002. 5. Bilaniuk LT: Vascular lesions of the orbit in children. Neuroimaging Clin N Am 15:107–120, 2005. 6. Bond JB, Haik BG, Mihara F: Magnetic resonance imaging of choroidal melanoma with and without gadolinium contrast enhancement. Ophthalmology 98:459–466, 1991. 7. Boyce SW, Platia EV, Green WR: Drusen of the optic nerve head. Ann Ophthalmol 10:645–704, 1978. 8. Chutorian AM, Schwartz JF, Evans RA, et al: Optic gliomas in children. Neurology 14:83–95, 1964. 9. Collison JMT, Miller NR, Green WR: Involvement of orbital tissues by sarcoid. Am J Ophthalmol 102:302–307, 1986. 10. Conneely MF, Mafee MF: Orbital rhabdomyosarcoma and simulating lesions. Neuroimaging Clin N Am 15:121–136, 2005. 11. Coupland SE: T-cell and T/natural killer cell lymphomas involving ocular and ocular adnexal tissues: A clinicopathologic, immunohistochemical, and molecular study of seven cases. Ophthalmology 106:2109–2120, 1999. 12. Courcoutsakis NA, Langford CA, Sneller MC, et al: Orbital involvement in Wegener granulomatosis: MR findings in 12 patients. J Comput Assist Tomogr 21:452–458, 1997. 13. Curtin HD: Pseudotumor. Radiol Clin North Am 25:583–599, 1987. 14. Del Grande F, Santini F, Herzka DA, et al: Fat suppression techniques for 3-T MR imaging of the musculoskeletal system. Radiographics 34(1):217–233, 2014. 15. DiMario FJ, Ramsby G, Grennastein R, et al: Neurofibromatosis type I: Magnetic resonance imaging findings. J Child Neurol 8:32–39, 1993. 16. Ettl A, Kramer J, Daxer A, et al: High resolution magnetic resonance imaging of neurovascular orbital anatomy. Ophthalmology 104:869–877, 1997. 17. Flanders AE, Mafee MF, Rao VM, et al: CT characteristics of orbital pseudotumors and other orbital inflammatory processes. J Comput Assist Tomogr 13:40–47, 1989. 18. Font RL, Patipa M, Rosenbaum PS, et al: Correlation of computed tomographic and histopathologic features in malignant transformation of benign mixed tumor of lacrimal gland. Surv Ophthalmol 34(6):449– 452, 1990. 19. Forbes GS: Vascular lesions in the orbit. Neuroimaging Clin N Am 6:113–122, 1996. 20. Graeb DA, Rootman J, Robertson WD, et al: Orbital lymphangiomas: Clinical, radiologic, and pathologic characteristics. Radiology 175:417– 421, 1990. 21. Haik BG, Saint Louis L, Bierly J, et al: Magnetic resonance imaging in evaluation of optic nerve gliomas. Ophthalmology 94:709–717, 1987. 22. Handler LC, Davey IC, Hill JC, et al: The acute orbit: Differentiation of orbital cellulites from subperiosteal abscess by computed tomography. Neuroradiology 33:15–18, 1991. 23. Harnsberger HR: Handbook of head and neck imaging, ed 2, St. Louis, 1995, Mosby. 24. Harris GJ, Sakol PJ, Bonavolonta G, et al: An analysis of thirty cases of orbital lymphangioma: Pathophysiologic considerations and management recommendations. Ophthalmology 97:1583–1592, 1990. 25. Haynes BF, Fishman ML, Fauci AS, et al: The ocular manifestations of Wegener’s granulomatosis. Am J Med 63:131–141, 1977. 26. Henderson JW: Orbital cysts. In Orbital tumors, ed 2, New York, 1980, BC Decker, pp 75–115. 27. Herrick RC, Hayman LA, Taber KH, et al: Artifacts and pitfalls in MR imaging of the orbit: A clinical review. Radiographics 17:707–724, 1997.

http://radiologyebook.com

CHAPTER 20  Orbit 28. Hesselink JR, Davis KR, Dallow RL, et al: Computed tomography of masses in the lacrimal gland region. Radiology 131:143–147, 1979. 29. Hesselink JR, Davis KR, Weber AL, et al: Radiologic evaluation of orbital metastasis, with emphasis on computed tomography. Radiology 137:363–366, 1980. 30. Hilal SK, Trokel SL: Computerized tomography of the orbit using thin sections. Semin Roentgenol 12:137–147, 1977. 31. Hoffman GS, Kerr GS, Leavitt RY, et al: Wegener’s granulomatosis: An analysis of 158 patients. Ann Intern Med 116:488–498, 1992. 32. Johnson G, Miller DM, MacManus D, et al: STIR sequences in NMR imaging of the optic nerve. Neuroradiology 29:238–245, 1987. 33. Jones IS, Jacobiec FA, editors: Diseases of the orbit, New York, 1979, Harper & Row. 34. Katz SE, Rootman J, Vangveeravong S, et al: Combined venous lymphatic malformations of the orbit (so-called lymphangiomas): Association with noncontiguous intracranial vascular anomalies. Ophthalmology 105:176–184, 1998. 35. Kodilyne HC: Retinoblastoma in Nigeria: Problems in treatment. Am J Ophthalmol 63:467–481, 1967. 36. Koeller KK, Smirniotopoulos G: Orbital masses. Semin Ultrasound CT MR 19:272–291, 1998. 37. Koornneef L, Zonneveld F: The role of direct multiplanar high resolution CT in the assessment and management of orbital trauma. Radiol Clin North Am 25:753–766, 1987. 38. Kubal WS: The pathological globe: Clinical and imaging analysis. Semin Ultrasound CT MR 18:423–436, 1997. 39. Kubal WS: Imaging of orbital trauma. Radiographics 28:1729–1739, 2008. 40. Mafee MF: The orbit proper. In Som PM, Curtin HD, editors: Head and neck imaging, ed 4, St. Louis, 2003, Mosby, pp 193–259. 41. Mafee MF, Ainbinder D, Afshani E, et al: The eye. Neuroimaging Clin N Am 6:29–59, 1996. 42. Mafee MF, Goldberg MF, Greenwald MJ, et al: Retinoblastoma and simulating lesions: Role of CT and MR imaging. Radiol Clin North Am 25:667–682, 1987. 43. Mafee MF, Karimi A, Shah J, et al: Anatomy and pathology of the eye: Role of MR imaging and CT. Neuroimaging Clin N Am 15:23–47, 2005. 44. Mafee MF, Peyman GA: Retinal and choroidal detachments: Role of magnetic resonance imaging and computed tomography. Radiol Clin North Am 25:487–507, 1987. 45. Mafee MF, Peyman GA, McKusick MA: Malignant uveal melanoma and similar lesions studied by computed tomography. Radiology 156:404– 408, 1985. 46. Manfre L, de Maria M, Mangiameli A, et al: MR dacrocystography: Comparison with dacrocystography and CT dacrocystography. AJNR Am J Neuroradiol 21(6):1145–1150, 2000. 47. Mann I: Developmental abnormalities of the eye, ed 2, Philadelphia, 1957, JB Lippincott, pp 68–94. 48. McNichols MMJ, Power WJ, Griffen JF: Idiopathic inflammatory pseudotumor of the orbit: CT features correlated with clinical outcome. Clin Radiol 44:3–7, 1991. 49. Motton-Lippa L, Jakobiec FA, Smith M: Idiopathic inflammatory orbital pseudotumor in childhood: II. Results of diagnostic tests and biopsies. Ophthalmology 88:565–574, 1981. 50. Mulliken JB, Glowacki J: Hemangiomas and vascular malformations in infants and children: A classification based on endothelial characteristics. Plast Reconstr Surg 69:412–420, 1982. 51. Nishino H, Rubino FA, DeRemee RA, et al: Neurological involvement in Wegener’s granulomatosis: An analysis of 324 consecutive patients at the Mayo Clinic. Ann Neurol 33:4–9, 1993.

613

52. Nugent RA, Belkin RI, Meigel JM, et al: Graves orbitopathy: Correlation of CT and clinical findings. Radiology 177:675–682, 1990. 53. Nugent RA, Lapointe JS, Rootman J, et al: Orbital dermoids: Features on CT. Radiology 165:475–478, 1987. 54. Pendergrass TW, Davis S: Incidence of retinoblastoma in the United States. Arch Ophthalmol 98:1204–1210, 1980. 55. Peyman GA, Mafee MF: Uveal melanoma and similar lesions: The role of magnetic resonance imaging and computed tomography. Radiol Clin North Am 25:471–486, 1987. 56. Provenzale JM, Mukherji S, Allen NB, et al: Orbital involvement by Wegener’s granulomatosis: Imaging findings. AJR Am J Roentgenol 166:929–934, 1996. 57. Ramirez H, Blatt ES, Hibri HS: Computed tomography identification of calcified optic nerve drusen. Radiology 148:137–139, 1983. 58. Raymond WR, Char DH, Norman D, et al: Magnetic resonance imaging evaluation of uveal tumors. Am J Ophthalmol 111:633–641, 1991. 59. Roberts CF, Leehey PJ, III: Intraorbital wood foreign body mimicking air at CT. Radiology 185:507–508, 1992. 60. Rosenberg MA, Savino PJ, Glaser JS: A clinical analysis of pseudopapilledema: I. Population, laterality, refractive error, ophthalmoscopic characteristics and coincident disease. Arch Ophthalmol 97:65–70, 1979. 61. Rootman J: Vascular malformations of the orbit: Hemodynamic concepts. Orbit 22:103–120, 2003. 62. Rothfus WE, Kapoor V: Orbital masses, schemata for differential diagnosis. In Latchaw RE, Kucharczyk J, Moseley ME, editors: Imaging of the nervous system—Diagnostic and therapeutic applications, St. Louis, 2005, Mosby, pp 1007–1031. 63. Schulman JA, Peyman GA, Mafee MF, et al: The use of magnetic resonance imaging in the evaluation of retinoblastoma. J Pediatr Ophthalmol Strabismus 23:144–147, 1986. 64. Shields JA, Bakewell B, Augsburger JJ, et al: Classification and incidence of space-occupying lesions of the orbit: A survey of 645 biopsies. Arch Ophthalmol 102:1606–1611, 1984. 65. Shnier R, Parker GD, Hallinan JM, et al: Orbital varices: A new   technique for noninvasive diagnosis. AJNR Am J Neuroradiol 12:717–718, 1991. 66. Simmons JT, Leavitt R, Kornblat AD, et al: CT of the paranasal sinuses and orbits in patients with Wegener’s granulomatosis. Ear Nose Throat J 66:134–140, 1987. 67. Som PM, Brandwein MS: Inflammatory diseases. In Som PM, Curtin HD, editors: Head and neck imaging, ed 4, St. Louis, 2003, Mosby, pp 193–259. 68. Turner RM, Gutman I, Hilal SK, et al: CT of drusen bodies and other calcified lesions of the optic nerve: Case report and differential diagnosis. AJNR Am J Neuroradiol 4:175–178, 1983. 69. Vaidhyanath R, Kirke R, Brown L, et al: Lacrimal fossa lesions: Pictorial review of CT and MRI features. Orbit 27(6):410–418, 2008. 70. Valvassori GE: Imaging of orbital lymphoproliferative disorders. Radiol Clin North Am 37:135–150, 1999. 71. White JH, Fox AJ, Symons SP: Diagnosis and anatomic mapping of an orbital varix by computed tomographic angiography. Am J Ophthalmol 140:945–947, 2005. 72. Winter J, Centeno RS, Bentson JR: Maneuver to aid diagnosis of   orbital varix by computed tomography. AJNR Am J Neuroradiol 3:39–40, 1982. 73. Wright JE, McNab AA, McDonald WI: Optic nerve glioma and the management of optic nerve tumours in the young. Br J Ophthalmol 73:967–974, 1989. 74. Youssefi B: Orbital tumors in children: A clinical study of 62 cases. J Pediatr Ophthalmol 6:177–181, 1969.

http://radiologyebook.com

21  Temporal Bone Ellen Hoeffner, Rickin Shah, Dilan Samarawickrama, and Suyash Mohan

EMBRYOLOGY AND DEVELOPMENT OF TEMPORAL BONE The ear consists of three major components, the external, middle,  and internal ear. Embryologically the ear has a dual development  with development of the inner ear structures (internal auditory canal [IAC], cochlea, vestibule, and semicircular canals [SCC]) independent of the middle and external ear structures (ossicles, middle ear air spaces, mastoid antrum, tympanic membrane [TM], external auditory canal [EAC], and temporomandibular joint [TMJ]). The neural soundperceiving apparatus of the inner ear develops from a thickened  ectodermal placode. The mechanical sound-collecting and soundtransmitting structures of the external and middle ear, respectively, develop from the first and second pharyngeal (branchial) arches and the intervening first pharyngeal cleft and pharyngeal pouch. Adjacent mesenchyme contributes to the development of the inner, middle, and external ear.19 This explains why developmental abnormalities of the inner ear are usually not seen with deformities of the external and middle ear, and vice versa.10,81 The endolymphatic (membranous), perilymphatic (periotic), and bony labyrinths of the inner ear all develop from the ectodermal otic placode that invaginates to form the otocyst or otic vesicle. The otic vesicle elongates, forming a dorsal vestibular region and a ventral cochlear region. The endolymphatic duct also develops as a short projection from the dorsomedial surface of the otic vesicle.19 The perilymphatic labyrinth develops as mesenchyme surrounding the membranous labyrinth resorbs.23 The bony labyrinth (otic capsule) also develops from mesenchyme around the membranous labyrinth. This begins between 4 and 8 weeks’ gestation, continues to grow between 8 and 16 weeks’ gestation, and ossifies by week 24 of gestation.95 The external and middle ear develop from the pharyngeal apparatus. The first pharyngeal groove deepens to become the primitive EAC and over time contributes to the development of the mature EAC, the outer layer of the TM, and the tympanic ring. The auricle or pinna develops around the first pharyngeal groove from six nodular mesenchymal masses or hillocks arising from the first and second pharyngeal arches.23 The first pharyngeal pouch expands to form a tubotympanic recess.19 This tubotympanic recess gives rise to the eustachian tube, the tympanic cavity, the mastoid antrum, and their epithelial lining.23 Just dorsal to the end of the tubotympanic recess, a condensation of mesenchyme from the first and second pharyngeal arches appears, which develops into the ossicles.19 The head and neck of the malleus and body and short process of the incus arise from the first arch. The second arch forms Reichert’s cartilage, from which the remainder of the malleus and incus and the stapes superstructure develop. The stapes footplate is a bilaminar structure, with an outer portion developing

614

from Reichert’s cartilage and an inner portion developing from the ectodermal otocyst.23 The middle ear cavity and mastoid antrum are fluid filled until birth. The neonate’s initial crying and breathing fill the eustachian tube system and the middle ear with air. The mastoid air cells develop as saclike extensions from the mastoid antrum, commencing around the time of birth and continuing for several years.

NORMAL TEMPORAL BONE ANATOMY External Auditory Canal The EAC consists of a funnel-shaped, undulating, cartilaginous lateral portion and an osseous medial segment (Figs. 21-1 through 21-5;  see Figs. 21-1B, 21-2C, 21-3A, 21-5A). The cartilaginous portion is continuous with the auricle, flexible and surrounded by fat. The osseous portion, surrounded by the tympanic portion of the temporal bone, constitutes two thirds of the length of the EAC and is covered by skin and periosteum only. The EAC extends medially to the TM. The TM attaches to an annular ridge of bone termed the tympanic anulus.50

Middle Ear Cavity and Mastoid The tympanic cavity, or middle ear, is the air space within the petrous portion of the temporal bone between the eustachian canal anteriorly and the mastoid air cells posteriorly that contains the ossicles (see Figs. 21-1D-G, 21-2C-F, 21-3B, 21-4D). Its other borders consist of the TM laterally, otic capsule and cochlear promontory medially, tegmen tympani superiorly, and jugular floor inferiorly. The middle ear cavity can be subdivided into three spaces—hypotympanum, mesotympanum, and epitympanum—based on the relationship to the TM (see Fig. 21-2C). The hypotympanum is the most inferior portion of the tympanic cavity, inferior to the level of the TM, and connects anteromedially to the opening of the eustachian tube. The semicanal for the tensor tympani muscle is parallel and superior to the eustachian tube (see Fig. 21-2A).50 This muscle arises from the superior surface of the cartilaginous eustachian tube and courses posteriorly in the medial portion of the middle ear, turning sharply laterally at the level of the neck of the malleus, to which it attaches. The thin plate of bone separating these two semicanals is termed the cochleariform process (see Figs. 21-1F and 21-2B-C).23 The mesotympanum, the central portion of the tympanic cavity, is medial to the TM. It is bordered laterally by the scutum (or spur) and TM and medially by the otic capsule. The scutum is a bony crest separating the EAC from the epitympanic space (see Fig. 21-2C-D). The TM attaches to the scutum superiorly and to the limbus inferiorly. The mesotympanum contains the majority of the ossicles.

http://radiologyebook.com

Text continued on p. 623

CHAPTER 21  Temporal Bone

Petrooccipital fissure

condyle

* *

Condyle and TMJ

*

*

EAC

Carotid canal

Carotid canal

Cochlear aqueduct Descending VII

Descending VII Jugular foramen

B

A

Carotid canal Tensor tympani muscle Manubrium Long process

Tympanic membrane

1 Descending VII

Round window niche

***

Basal turn

Pyramidal eminence

Vestibular aqueduct

C

2

D FIG 21-1  Axial CT images from caudal to cranial. A, Jugular foramen level. The caroticojugular spine separates the carotid canal and jugular foramen. The descending facial nerve canal (descending VII) is lateral to the jugular foramen. The mandibular condyle and temporomandibular joint (*) are seen. B, Inferior tympanic level. The eustachian canal (black arrows) and its opening into the hypotympanum (*) are visualized just lateral to the carotid canal. A portion of the external auditory canal (EAC) is seen just anterior to the descending facial nerve canal (descending VII). The mandibular condyle and TMJ are seen anterior to the EAC. The cochlear aqueduct and petrooccipital fissure are seen. C-E, Midtympanic level. The descending portion of the facial nerve canal (descending VII) is seen posterior to the pyramidal eminence (C, D). The facial nerve recess (1) can be seen lateral to and the sinus tympani (2) medial to the pyramidal eminence (D). Continued

http://radiologyebook.com

615

616

PART II  CT and MR Imaging of the Whole Body

Tensor tympani t. Ant. crus 3 4

Manubrium

5

Post. crus

Long process Stapes

E

F Interscalar septae

Tympanic VII i

m

mo v

IAC

Singular canal

G

H FIG 21-1, cont’d  The manubrium of the malleus (manubrium) is anterior to the long process of the incus (long process) (D, E). The stapes suprastructure is medial to the incus (E). The apical (3), middle (4), and basal (5) turns of cochlea and round window niche are seen with overlying cochlear promontory (*) (C-E). The tensor tympani muscle is lateral and parallel to the carotid canal (C). F-J, Epitympanic–internal auditory canal level. The anterior (ant. crus) and posterior (post. crus) crura of stapes are seen extending to oval window; the tensor tympani tendon (tensor tympani t.) attaches to neck of malleus (F). At the incudomalleolar articulation the “ice-cream-cone” configuration of the head of malleus (m) and body and short process of incus (i) is seen (G). The modiolus (mo) and interscalar septae of the cochlea are seen along with the cochlear nerve canal (black arrowheads) (H).

http://radiologyebook.com

CHAPTER 21  Temporal Bone

Tympanic VII

L S S C

*

Labyrinthine VII

Sup. vest. nerve c.

V

J

I

SSCC Aditus

Common crus

Antrum PSCC

K

L FIG 21-1, cont’d  The vestibule (v) and lateral semicircular canal (LSSC) are seen as are the canals for the superior (sup. vest. nerve c.) and inferior (singular canal) vestibular nerves as the exit the internal auditory canal (IAC) (G-I). The labyrinthine segment (labyrinthine VII), anterior genu (*) and tympanic segment (tympanic VII) of the facial nerve canal are seen in their characteristic course (G, J). K and L, Mastoid antrum level. The aditus ad antrum (aditus) is seen connecting the mastoid antrum and epitympanum (K). Portions of the posterior (PSSC) and superior (SSSC) semicircular canals and their shared common crus are visualized (K, L).

http://radiologyebook.com

617

618

PART II  CT and MR Imaging of the Whole Body

Geniculate ganglion Tensor tympani Cochlea

* *

Tensor tympani

Carotid canal

*

Cochleariform process

Condyle

Carotid canal

Eustachian tube

A

B

Tympanic VII TT tendon

*

T

L

Cochlea

N

** Incus LP

EAC

C

Cochlea

D FIG 21-2  Coronal CT images from anterior to posterior. A, Temporomandibular joint (TMJ) level. The tensor tympani muscle is seen in its semicanal just above the eustachian tube. The TMJ (*), mandibular condyle (condyle) and carotid canal are seen. B, Geniculate ganglion level. The tensor tympani muscle and cochleariform process are seen along medial wall of middle ear. The geniculate ganglion is seen as a small lucency just superior to anteriormost portion of cochlea. The carotid canal is just inferior to the cochlea. C, Anterior tympanic level. The external auditory canal (EAC) and tympanic membrane (TM) (white arrowheads) are seen, with the TM attaching to the scutum (*) superiorly. The malleus is seen, with the tensor tympani tendon (TT tendon) attaching to the neck (N) of the malleus. The labyrinthine (L) and tympanic (T) segments of facial nerve canal are seen as two lucencies superior to cochlea. D-E, Midtympanic level. The long process of the incus is seen (incus LP), forming the medial border of Prussak’s space (white *); the scutum (black *) forms the lateral border of this space (D).

http://radiologyebook.com

CHAPTER 21  Temporal Bone

Superior SCC Tympanic VII

Oval window

Lateral SCC

V

Crista falciformis

IAC

Tympanic VII Stapes I-S joint

E

F

Antrum

Superior SCC

Pyramidal eminence Lateral SCC 1

2

Round w. niche

G

Jugular foramen

H FIG 21-2, cont’d  The lenticular process of the incus and head of the stapes form the incudostapedial joint (I-S joint) (E). The tympanic segment of the facial nerve canal (tympanic VII) is seen superolateral to the cochlea and cochlear promontory (black arrowheads) (D, E). F, Oval window level. The internal auditory canal (IAC) is visualized divided by the crista falciformis. The oval window is seen along the lateral margin of the vestibule (v), with the stapes extending toward it. The tympanic segment of the facial nerve canal (tympanic VII) is seen just below the lateral SCC. The anterior limb of the superior SCC is also seen. G, Posterior tympanic level. The facial nerve recess (1) is lateral to and the sinus tympani (2) medial to the pyramidal eminence. The round window niche (round w. niche) is seen along the basal turn of the cochlea. H, Jugular foramen level. The mastoid segment of the facial nerve canal (black arrowheads) is seen in its vertical course. Parts of the lateral SCC and superior SCC are seen. The mastoid antrum and jugular foramen are seen.

http://radiologyebook.com

619

620

PART II  CT and MR Imaging of the Whole Body

Malleus

Incus

Lateral SCC

E A C

Condyle

B

A

Common crus

V

Cochlea

Superior SCC

I A C

V. aqueduct

*

Carotid canal Jugular foramen

C

D FIG 21-3  Sagittal CT images from lateral to medial. A, External auditory canal (EAC) level. The EAC is posterior to the TMJ composed of the glenoid fossa (arrowheads) and the mandibular condyle (condyle). B, Descending facial nerve canal level. The posterior genu and descending facial nerve canal (arrowheads) have an upside-down hockey-stick shape, and the malleus and incus together have a molar tooth configuration. Part of the lateral SCC is seen. C, Vestibule level. The vestibule (V) is central, with the common crus and a portion of the superior SCC posterior to the vestibule. The round window (arrowhead) is just posterior to the basal turn of the cochlea (*). The vestibular aqueduct (V. aqueduct) is seen along the posterior margin of the petrous bone. D, Internal auditory canal (IAC) level. The IAC is seen as an oval lucency posterior to a portion of the cochlea. The carotid canal and jugular foramen are seen.

http://radiologyebook.com

Superior SCC

Superior SCC Posterior SCC

Common crus

A

B Incudomallear joint

Cochlea

I M

C

D Superior SCC

Posterior SCC

m

Tympanic VII

E

F FIG 21-4  Stenvers’ and Pöschl’s views. A, Stenvers’ view midsuperior SCC level. The superior SCC is seen in cross section. B, Stenvers’ view common crus level. In addition to common crus, portions of the posterior SCC and superior SCC are seen. C, Stenvers’ view cochlea level. The turns of the cochlea are seen, as is the round window (arrowhead) and descending facial nerve canal (arrows). D, Stenvers’ view incudomalleolar joint level. The incudomalleolar joint is well seen between body of incus (I) and head of malleus (M). E, Pöschl’s view midsuperior SCC level. The superior SCC is seen as a ring, and the posterior SCC is seen in cross section. The tympanic segment of the facial nerve canal is seen in cross section (tympanic VII). http://radiologyebook.com F, Pöschl’s view modiolus level. The modiolus (m) is seen along its axis.

622

PART II  CT and MR Imaging of the Whole Body

Cochlea

* EAC

** Posterior SCC Endolymphatic duct

A

B

Cochlea

* V

Lateral SCC

IAC

Posterior SCC

C

Cochlear n.

Inferior vestibular n.

D FIG 21-5  Axial MRIs from caudal to cranial. A, Jugular foramen level, postcontrast T1-weighted image. The descending portion of the facial nerve is seen in cross section (arrow) as a focus of intermediate signal lateral to the jugular bulb (**), which has mixed signal. Mixed signal also seen in internal carotid artery (*). The EAC can be seen. B, Midtympanic level, T2-weighted image. Inner ear structures (cochlea and posterior SCC) are of high signal, as is endolymphatic duct. C, Epitympanic level, T2-weighted image. Inner ear structures (cochlea, vestibule [V], lateral SCC, and posterior SCC) are of high signal; modiolus (*) is seen as a low-signal structure within cochlea. The cochlear nerve (cochlear n.) and inferior vestibular nerve (inferior vestibular n.) are seen as linear low-signal structures in fluid-filled IAC. D, Epitympanic level, T1-weighted postcontrast image. The geniculate ganglion (long arrow) and tympanic segment of the facial nerve (short arrows) are of intermediate signal. The IAC can be seen medially.

http://radiologyebook.com

CHAPTER 21  Temporal Bone The triangular epitympanum forms the superior portion of the middle ear, superior to the level of the TM. It contains the head of the malleus and the body and short process of the incus (see Fig. 21-1G). The posterior wall of the cavity is irregular and contains the facial nerve recess laterally and the sinus tympani medially, separated by the pyramidal eminence. The pyramidal eminence overlies the stapedius muscle, which inserts onto the head of the stapes (see Figs. 21-1D   and 21-2G). Prussak’s space is another important area of the middle ear cavity bounded by the scutum and pars flaccid of the TM laterally, the neck of the malleus medially, and the lateral malleal ligament superiorly (see Fig. 21-2D). The cone-shaped antrum is located in the anterosuperior part of the mastoid portion of the temporal bone and communicates with the epitympanum via a narrow channel termed the aditus ad antrum. The antrum is surrounded by smaller variable-sized mastoid air cells (see Fig. 21-1K).23 The size of the antrum varies considerably as a result of the alterations in pneumatization of the air cells that drain into it. The antrum is bordered superiorly by the tegmen tympani. The medial margin of the antrum is the promontory formed by the otic capsule of the lateral SCC. The mastoid air cells are divided by Koerner’s septum, a thin bony structure formed by the petrosquamous suture that extends posteriorly from the epitympanum, into medial and lateral components.

623

Superior SCC

Lateral SCC Cochlea

A

B

Ossicles The ossicles are suspended by the TM, the ossicular ligaments to the epitympanic walls, and the oval window. The manubrium (or handle) and lateral process of the malleus attach to the TM. The neck of the malleus is connected to the tensor tympani tendon (see Fig. 21-2C). The anterior malleal ligament attaches the head of the malleus to the anterior epitympanic wall; the superior malleal ligament attaches the head to the roof of the middle ear cavity. The lateral malleal ligament attaches the head or neck of the malleus to the tympanic notch, a defect in the superior portion of the tympanic anulus.50,56 The circular head of the malleus articulates with the triangular body of the incus (see Figs. 21-1G and 21-3B). The short process of the incus projects posteriorly and is suspended within the epitympanic space, pointing toward the aditus ad antrum. The posterior incudal ligament attaches the short process to the fossa incudis, a small depression in the posteroinferior part of the epitympanic recess. The long process of the incus projects inferiorly, parallel to and behind the manubrium of the malleus (see Fig. 21-2D). The tip of the long process bends medially to end in the lenticular process, which articulates with the head of the stapes (see Fig. 21-2E). The stapes consists of the head, two crura, and a footplate (see Figs. 21-1E and F). The footplate of the stapes attaches to the oval window of the vestibule and is secured by the annular ligament (see Fig. 21-2F).

Inner Ear The inner ear is located within the petrous portion of the temporal bone and comprises the osseous labyrinth, which consists of the cochlea, vestibule, and SCCs. The osseous labyrinth encapsulates the membranous labyrinth, which contains endolymph and is surrounded by perilymph. The vestibular cavity is the central portion of the membranous labyrinth that communicates with the SCCs and cochlea (Fig. 21-6; see Figs. 21-1G, 21-1I, 21-2F, 21-3C, 21-5C, 21-6A). The vestibule lies immediately lateral to the fundus of the IAC, and the oval window of the middle ear forms its lateral margin. The vestibular aqueduct is a narrow bony canal that contains the small endolymphatic duct and sac (see Figs. 21-1C, 21-3C, 21-5B). The endolymphatic duct originates

FIG 21-6  Coronal MRIs from anterior to posterior. A, Oval window level, T2-weighted image. The IAC (arrowheads) is of high signal, with low-signal nerves traversing it. The inner ear structures (vestibule [V], superior SCC, lateral SCC, and cochlea) are of high signal. B, Oval widow level, T1-weighted postcontrast image. The IAC (arrowheads) is of intermediate signal.

from the posteromedial aspect of the vestibule, courses past the common crus, and assumes a hockey-stick shape. The endolymphatic sac, an extension of the endolymphatic duct, is a dead-end fluid space arising from the vestibule that lies beneath the dura of the posterior temporal bone. The bony vestibular aqueduct surrounds the endolymphatic duct.50 The SCCs are three orthogonally arranged circular structures of the membranous labyrinth arising from the vestibule. The lateral SCC approximates the horizontal plane (see Fig. 21-1I). The posterior and superior SCCs share a crus posteriorly, formed by a fusion of the posterior crus of the superior SCC and the anterior crus of the posterior SCC (see Figs. 21-1K, 21-2F, 21-3C, 21-4B). The vestibule and SCCs are responsible for balance and equilibrium. The oval window of the vestibule is the interface between the mechanical and neural elements of the ear (see Fig. 21-2F). It lies just inferior to the horizontal facial nerve canal and the lateral SCC. It is covered by the annular ligament, which surrounds the footplate of the stapes. The spiral cochlear apparatus contains the membranous labyrinth components for hearing. The cochlea lies anterior to the vestibule and has 2.5 to 2.75 turns (basal, middle, and apical) around its modiolus, the central bony spiral axis. The turns of the cochlea are separated by interscalar septae (see Figs. 21-1E and H, 21-2C, 21-4C, 21-5B and C). The most lateral portion of the basal turn of the cochlea projects into the tympanic cavity, forming a promontory of bone (see Figs. 21-1D, 21-2E). The round window is the bony opening of the basal turn of the cochlea into the middle ear, lying inferior and slightly posterior to the oval window and covered by a fibrous membrane (see

http://radiologyebook.com

624

PART II  CT and MR Imaging of the Whole Body

1 2

3 4

FIG 21-7  IAC level, sagittal T2-weighted image. The high-signal IAC is seen in cross section, with the facial (1), cochlear (2), superior vestibular (3) and inferior vestibular (4) nerves seen within.

Figs. 21-1C, 21-2G, 21-3C). The cochlear nerve passes from the IAC through the bony canal for the cochlear nerve into the modiolus (see Figs. 21-1H, 21-5C). The cochlear aqueduct, a channel that connects the cochlea near the round window with the subarachnoid space just superior to the jugular tubercle, lies inferior to the IAC (see Fig. 21-1B).

Internal Auditory Canal The IAC is a bony conduit within the petrous portion of the temporal bone that transmits cranial nerves VII (facial) and VIII (vestibulocochlear) from the pontomedullary junction of the brainstem across the cerebellopontine angle (CPA) cistern to the inner ear (Fig. 21-7; see Figs. 21-1G, 21-2F, 21-5C, 21-6A and B). Its medial orifice is the porus acusticus. The IAC is partially divided by a transverse bony crest called the crista falciformis (see Fig. 21-2F), which runs parallel to the long axis of the canal and divides it into a superior and inferior portion. A vertical crest termed Bill’s bar divides the superior component into anterior and posterior parts. The facial nerve is located in anterosuperior compartment, the cochlear nerve in the anteroinferior compartment, and the superior and inferior vestibular nerves in the posterosuperior and posteroinferior compartments, respectively (see Fig. 21-7). The lateral portion of the canal or fundus is perforated by the cranial nerves where they enter the cochlea, vestibule, and facial nerve canal (see Fig. 21-1H-J). The IACs should be nearly symmetric. Although there is wide variation in the exact shape and size of the canals, asymmetry of greater than 2 mm suggests pathology.

Facial Nerve and Facial Nerve Canal The facial nerve has a highly tortuous course and comes in close contact with almost every other important component of the temporal

bone. It may be described as having six segments: cisternal, intracanalicular, labyrinthine, tympanic, mastoid, and extracranial. The facial nerve departs the brainstem at the lower border of the pons at the pontomedullary junction. Within the CPA cistern, the cisternal facial nerve is the most anterior and the vestibulocochlear nerve is the most posterior, with the nervus intermedius between the two. The facial nerve then enters the porus acusticus of the IAC and traverses the  IAC in its intracanalicular segment (see Fig. 21-7). The labyrinthine portion of the facial nerve exits the superior anterior portion of the IAC fundus and follows an anterolaterally curving course in the fallopian canal to the geniculate ganglion. The greater superficial petrosal nerve also arises from the geniculate ganglion (see Figs. 21-1J, 21-2B and C, 21-5D). At the geniculate ganglion, an acute reverse (inverted V) angle of the nerve and canal is seen. The distal limb of the facial nerve doubles back posteriorly to become the tympanic segment in the medial wall of the tympanic cavity (see Figs. 21-1C, 21-2D-F, 21-5D). This portion of the facial nerve canal has a thin inferior bony covering that may be dehiscent. The tympanic portion of the facial nerve runs horizontally directly inferior to the lateral SCC. The facial nerve then abruptly turns inferiorly (at an angle of 95 to 125 degrees) to form the posterior genu near the sinus tympani and the pyramidal eminence (see Fig. 21-3B). The vertical, or descending, facial nerve in the mastoid portion of the facial nerve canal runs vertically just posterior to the EAC. In this vertical portion, the nerve to the stapedius muscle and the chorda tympani (the first and second branches, respectively, of the facial nerve distal to the geniculate ganglion) arise (see Figs. 21-1A-C, 21-2H, 21-5A). The facial nerve then exits the mastoid portion of the temporal bone via the funnel-shaped fat-filled stylomastoid foramen. The extracranial portion of the facial nerve delivers branches to the muscles and pierces the parotid gland to form the parotid plexus.

Carotid Canal The carotid canal enters the base of the skull, ascends vertically, then turns horizontally and medially toward the petrous apex (see Figs. 21-1A-C, 21-2A and B, 21-3D, 21-5A). This canal lies anterior and inferior to the cochlea and is separated from the middle ear cavity by a thin bony plate.

Jugular Foramen and Fossa The jugular foramen is divided into a smaller anteromedial neural compartment (pars nervosa) containing cranial nerves IX, X, and XI, and a larger posterolateral vascular compartment (the pars vascularis) containing the jugular vein. The floor of the tympanic cavity normally forms the roof of the jugular fossa. The jugular bulbs are frequently asymmetric, and there may be a small diverticulum from their apices (see Figs. 21-1A, 21-2H, 21-3D, 21-5A).

TEMPORAL BONE IMAGING TECHNIQUES Computed tomography (CT) and magnetic resonance imaging (MRI) are currently the most widely used techniques for imaging the temporal bone, having largely replaced the former modalities. CT and MRI studies are complementary. Each technique has advantages and disadvantages, and often more than one examination is necessary for a complete temporal bone evaluation. CT is excellent for assessing the osseous structures of the temporal bone but not ideal for evaluating the soft tissue contents of the otic capsule, brain, or vessels. In contrast, MRI is superior to CT in characterizing the cerebrospinal fluid (CSF), brain, and cranial nerves. CT and MR angiography can be used to assess vascular structures.

http://radiologyebook.com

CHAPTER 21  Temporal Bone Computed Tomography The temporal bone has a high inherent contrast resolution, having both the densest bone in the body and air-filled spaces. High-spatialresolution CT scanning is the best method for evaluating bone and air space anatomy and disorders. A major advance in CT technology was achieved with the introduction of multidetector-row helical CT. Compared to earlier CT techniques, multidetector CT provides improved quality of the axial source images as well as two- and three-dimensional (2D and 3D) reconstructions; this is due to a reduction in partial volume and motion artifacts and higher resolution along the z-axis with nearly isotropic voxel size.48,113 Multiplanar CT reformations can display the complex anatomy of the temporal bone in all planes while exposing the patient to radiation only once.54,62 High-resolution CT images are usually acquired with thin sections (0.5-1 mm) and special bone algorithms for high detail. At the authors’ institution, scans are acquired in the axial plane in the helical mode with a slice thickness of 0.625 mm. The image data set is reconstructed into magnified axial, sagittal, and coronal images with 0.625-mm thickness at an interval of 0.312 mm. Contrast enhancement is not essential for evaluation of pathology isolated to bone or air spaces. For some clinical indications, additional reconstructed images are obtained in the planes of Stenvers and Pöschl (see Fig. 21-4A-F). Stenvers’ plane is perpendicular to the superior SCC. Pöschl’s plane is parallel to the superior SCC.23 Oblique reformatted images can be obtained in virtually any other plane to highlight specific structures of the temporal bone.54 The high contrast between the bony structures and the air spaces allows for virtual endoscopy of the middle ear by using surfacerendering technique. Axial images and 2D reformatted images remain the main modality for evaluating temporal bone structures and pathology. 3D reformatted virtual endoscopy (VE) images may serve as an adjunct in assessing the temporal bone, particularly in assessing the ossicular chain and possibly the round window; however, the overall sensitivity and specificity of this technique for detecting middle ear abnormalities is 59% and 25%, respectively.41 VE images may be of additional benefit to surgeons because they are more similar to their visual view in the operating room. VE images may also be helpful in teaching and in ear surgery simulators.62

Magnetic Resonance Imaging MRI also has revolutionized the temporal bone examination and expanded the range of pathology that can be accurately evaluated, because it can image many soft tissue entities. It is much better than CT in characterizing the CSF, brain, and cranial nerves, and it can be useful in the evaluation of blood vessel–related disorders of the temporal bone. Many different techniques can be used to acquire MRIs, depending on the suspected abnormality and the equipment available. T2-weighted (repetition time [TR] = 3000 milliseconds or longer; echo time [TE] = 80 milliseconds or longer) axial spin echo images display the CSF and endolymphatic spaces and many disease processes as highsignal-intensity regions. Submillimeter images can be obtained using a 3D turbo-spin echo (TSE)/fast-spin echo (FSE) sequences.20 These sequences display the otic capsule structures as high-signal-intensity regions surrounded by signal void. The facial nerve and the three branches of the vestibulocochlear nerve can be distinguished inside the IAC. This type of sequence is often used as a screening technique for vestibular schwannomas.1 T1-weighted (TR = 600 milliseconds, TE = 20 milliseconds) spin echo coronal or axial images are often acquired before and after administration of gadolinium-based contrast agent. The fat spaces have high

625

signal intensity, and the brain has intermediate signal intensity. CSF demonstrates low signal intensity, and air and bone spaces appear as voids. Postcontrast T1-weighted images are particularly sensitive for evaluating abnormalities that alter the blood-brain barrier or are vascular. The sensitivity of contrast-enhanced T1-weighted images to detect these lesions can be exaggerated by using fat-saturation techniques.64

Normal Temporal Bone Images Axial Images: Caudal to Cranial

Axial jugular foramen level (see Figs. 21-1A, 21-5A).  CT: The carotid canal lies just anterior to the jugular fossa, forming a “snowman”-like configuration. Both demonstrate sharp cortical margins. Inferiorly only a small spine, the caroticojugular spine, separates the two as they converge to enter the carotid sheath. The descending facial nerve canal is lateral to the jugular foramen, seen as a rounded well-corticated lucency. The mandibular condyle and TMJ can also  be seen.23,39 MRI: The descending portion of the facial nerve can be seen just lateral to the jugular bulb on T1- and T2-weighted images as an intermediate- to high-signal rounded structure. The blood vessels generally demonstrate signal void due to flow but may be seen as areas of intermediate to high signal if slow or turbulent flow is present.23,39 Axial inferior tympanic level (see Fig. 21-1B).  CT: The opening of the eustachian canal is triangular, with the apex extending parallel to the carotid canal and communicating with the anterior hypotympanum. The anterior and posterior walls of the bony EAC demonstrate dense, sharp cortical margins without soft tissue covering. The anterior margin of the EAC forms the posterior lip of the TMJ. The medial funnel-shaped opening of the cochlear aqueduct can be seen as a triangular lucency facing the CPA, progressively enlarging from lateral to medial. The opening of the aqueduct may be large and mimic the IAC. The descending facial nerve canal is easily identified posterior to the EAC. The petrooccipital fissure separates the temporal bone from the occiput.23,39 MRI: The facial nerve is seen as an intermediate- to high-signalintensity circular structure surrounded by a large area of signal void of the temporal bone and mastoid air spaces. The cochlear aqueduct is not well seen on MRI studies, possibly because of the combination of its small diameter and CSF motion artifacts.23,39 Axial midtympanic level (see Figs. 21-1C-E, 21-5B).  CT: The carotid canal can be seen early in its anteromedial course through the skull base, parallel and medial to the semicanal of the tensor tympani muscle at this level. The normally thin TM may be visible. The manubrium of the malleus parallels the TM. The manubrium of the malleus lies parallel and anterior to the long process of the incus. The stapedial superstructure (head, crura, and tympanic portion of the footplate) is often seen, forming an arch over the oval window. The cochlear promontory can be seen along the complex medial wall of the middle ear. Along the posterior wall of the middle ear is the pyramidal eminence, with the facial nerve recess lateral and the sinus tympani medial to the eminence. The descending facial nerve canal can be seen just posterior to the pyramidal eminence. The apical, middle, and basal cochlear turns are seen at this level along with the round window niche at the basal turn. The vestibular aqueduct is seen as a thin bony lucency along the posterior margin of the temporal bone, near its opening to the posterior cranial fossa.23,39 MRI: The cochlea is seen as a high-signal-intensity structure on T2-weighted images. The endolymphatic duct and sac may be seen as a tubular, thin, high-signal-intensity structure along the posterior margin of the temporal bone. The middle ear and its contents are not well seen unless filled with fluid.23,39

http://radiologyebook.com

626

PART II  CT and MR Imaging of the Whole Body

Axial epitympanic–internal auditory canal level (see Figs. 21-1F-J, 21-5C and D).  CT: The crura of the stapes can be seen extending toward the oval window, and the tensor tympani tendon can be seen making a 90-degree turn and attaching to the neck of the malleus. At the level of the incudomalleolar articulation, the round head of the malleus and the triangular body and short process of the incus are seen in their characteristic “ice cream cone” configuration. The lateral SCC is seen protruding into the middle ear. The modiolus and interscalar septae of the cochlea are visualized, as is the vestibule, posterolateral to the cochlea. The IAC is slightly funnel-shaped and ends in an ovoid fundus. The canals for the facial nerve, cochlear nerve, superior vestibular nerve, and inferior vestibular nerve (singular canal) can be seen leaving the fundus of the IAC. The labyrinthine segment of the facial nerve canal (fallopian canal) can be seen extending to the anterior genu, and the tympanic segment is seen coursing along the medial wall of the tympanic cavity.23,39 MRI: MRI studies demonstrate CSF in the IAC, and with highresolution images the individual nerves are seen as low-signal filling defects within the CSF. The fluid contents of the cochlea, vestibule, and SCCs are seen as high signal on T2-weighted images. The modiolus and interscalar septae are seen as areas of low signal within the otherwise bright cochlea. The geniculate ganglion and horizontal segment of the facial nerve are seen as intermediate-signal-intensity structures surrounded by the signal void resulting from their bony walls on T1-weighted images.23,39 Axial mastoid antrum level (see Fig. 21-1K and L).  CT: The aditus ad antrum is situated between the epitympanum and mastoid antrum. The mastoid antrum lies posterior and lateral to the aditus ad antrum and opens into many mastoid air cells. On CT scans, the posterior SCC and common crus of the posterior and superior SCCs can be seen. More superiorly the superior SCC is present.23,39 MR: Portions of the SCCs are seen as areas of high signal on T2-weighted images.23,39

Coronal Images: Anterior to Posterior Coronal temporomandibular joint level (see Fig. 21-2A).  CT: The horizontal carotid canal is seen as an oval structure just lateral to petrooccipital suture. The semicanal for the tensor tympani muscle is seen as a small lucency lateral to the carotid canal. The air-filled eustachian tube is inferior to the tensor tympani, and the TMJ is seen laterally.23,39 Coronal geniculate ganglion level (see Fig. 21-2B).  CT: The tensor tympani muscle is seen along the medial wall of the middle ear, with the cochleariform process separating it from the eustachian tube below. The anteriormost part of the cochlea is medial to the tensor tympani. The geniculate ganglion is seen as a small lucency superior to the cochlea.23,39 MRI: The geniculate ganglion is seen as a punctate focus of intermediate signal on T1-weighted images.23,39 Coronal anterior tympanic level (see Fig. 21-2C).  CT: The superior and inferior walls of the EAC are seen well. The TM may be identified as a thin filamentous structure extending from the scutum superiorly and coursing parallel to the plane of the long process of the malleus to attach to the limbus inferiorly. The head and neck of the malleus can be seen in the epitympanic space, with the tendon of the tensor tympani muscle attaching to neck. The basal and second turn of the cochlea are visualized. The labyrinthine and tympanic segments of the facial nerve are seen as two lucencies superior to the cochlea. The floor and roof (tegmen) of the middle ear are well visualized.23,39 MRI: The cochlea is of high signal on T2-weighted images. Coronal midtympanic level (see Fig. 21-2D and E).  CT: This level shows the long process and lenticular process of the incus and the

incudostapedial articulation as an L-shaped configuration. The stapes projects medially and superiorly from the lenticular process of the incus toward the oval window, located above the cochlear promontory. Prussak’s space is seen between the incus and scutum. The tympanic segment of the facial nerve canal is seen along the medial wall of the middle ear just superior and lateral to the cochlea.23,39 Coronal oval window level (see Figs. 21-2F, 21-6A and B).  CT: The full extent of the IAC is well visualized at this level, with the central crista falciformis dividing the canal into two portions. On CT scans, the oval window is seen as a bony defect in the lateral portion of the vestibule. The stapes is oriented slightly superiorly toward the oval window, just inferior to the lateral SCC. Also, beneath the lateral SCC, the horizontal portion of the facial nerve canal appears as a small circular structure. The anterior limb of the superior SCC can be seen. The epitympanic space lies just lateral to the lateral SCC. The floor and roof (tegmen) of the middle ear are well visualized on CT scans.23,39 MRI: On T2-weighted images the facial and vestibulocochlear nerves can be seen as linear low-signal-intensity structures surrounded by the high-signal fluid in the IAC. The inner ear structures are of high signal on T2-weighted images. On T1-weighted images the IAC and its contents are of intermediate signal.23,39 Coronal posterior tympanic level (see Fig. 21-2G).  CT: The vestibule and round window niche are seen at this level. The sinus tympani—which extends between the labyrinthine wall and the pyramidal eminence—is seen, as is the facial nerve recess, which is lateral to the pyramidal eminence.23,39 Coronal jugular foramen level (see Fig. 21-2H).  CT: The jugular foramen frequently has a dome-shaped outline. The mastoid segment of the facial nerve canal can be identified lateral to the jugular foramen, running nearly vertical and extending toward the stylomastoid foramen. The mastoid antrum is seen superiorly and laterally. Portions of the lateral and superior SCCs can be seen.23,39 MRI: Variable signal can be seen in the jugular bulb within the jugular foramen, depending on the sequence, flow, and presence of contrast. The mastoid segment of the facial nerve is seen as a linear intermediate signal structure lateral to the jugular bulb.23,39

Sagittal Images: Lateral to Medial Sagittal external auditory canal level (see Fig. 21-3A).  CT: The EAC is seen in cross section just posterior to the TMJ. The relationship of the mandibular condyle to the glenoid fossa and EAC is well visualized.23

Sagittal descending facial nerve canal level (see Fig. 21-3B). 

CT: The descending facial nerve and posterior genu form an upsidedown hockey-stick shape. The head and manubrium of the malleus and the body and long process of the incus can be seen in a “molar tooth” configuration. The lateralmost part of the lateral SCC is visualized.23 Sagittal vestibule level (see Fig. 21-3C).  CT: The vestibule is seen as a central lucency. The common crus of the superior and posterior SCCs can be seen just posterior and superior to the vestibule. The round window is visualized, with the basal turn of the cochlea just anterior to the round window. The vestibular aqueduct appears as a thin linear structure underlying a flange of the posterior petrous bone.23 MRI: The vestibule and portions of the cochlea and SCCs are seen as areas of high signal on T2-weighted images. The contained endolymphatic duct and sac are seen as areas of high signal underlying a flange of the petrous bone.23

Sagittal internal auditory canal level (see Figs. 21-3D, 21-7). 

CT: The IAC is seen in cross section as an oval lucency. A portion of

http://radiologyebook.com

CHAPTER 21  Temporal Bone the cochlea can be seen anterior to the IAC. The carotid canal is anterior and the jugular fossa inferior to the otic capsule structures.23 MRI: MRI studies can demonstrate the individual nerves in the IAC, with the facial nerve superior and anterior, the cochlear nerve inferior and anterior, and the vestibular nerves posterior.23

627

ABNORMAL TEMPORAL BONE

and must be assessed along with possible presence of a cholesteatoma. EAC atresia can be complete or stenotic.1 The EAC is considered narrowed if its diameter is less than 4 mm. Canal stenosis may be from a web or band of soft tissue, a bony plate at the level of the TM, or both.67 Many ossicular anomalies can occur with EAC maldevelopment. The ossicles may be fused, rotated, hypoplastic, or absent.1 The most common anomaly is fusion of the malleus and incus. The fused ossicles may also be fused to the atretic plate, usually at the region of the neck of malleus.112 The facial nerve canal must be carefully evaluated in all cases of suspected external and middle ear anomalies. Anomalous development of the middle ear and temporal bone may alter the course of the facial nerve, with the tympanic and mastoid segments most commonly affected. The tympanic segment is usually displaced inferiorly as low as the round window and may be medially displaced and cross the oval window. The mastoid segment is displaced anterolaterally and may exit the temporal bone at the level of the round window or into the TMJ1,67,83 (Fig. 21-8). Various grading systems and classifications have been developed for EAC atresia.47,86 The most common isolated congenital ossicular anomaly is incudostapedial discontinuity. In this anomaly the long process of the incus may be absent or it may be elongated and assume a more posterior trajectory than normal with respect to the oval window. Anomalies of the stapes suprastructure are often associated with abnormalities of the long process of the incus.112

Middle Ear and External Ear Malformations

Congenital Inner Ear Malformations

Congenital aural atresia is failure of development of the EAC. It is often associated with abnormalities of the auricle (microtia), ossicles, and middle ear. The inner ear is usually normal.27 The external ear, EAC, size of middle ear, ossicles, mastoid air cells, oval window, round window, and facial nerve course can potentially be dysplastic or absent

The different congenital inner ear malformations result from arrest during various stages of embryogenesis21,42,46,87 (Table 21-1). Michel’s deformity, or complete labyrinthine dysplasia, is absence of the inner ear structures. Cranial nerve VIII and its ganglion will be absent because the otic vesicle does not develop. The absence of cranial

Stenvers’ and Pöschl’s Views Stenvers’ view (see Fig. 21-4A-D).  This plane is perpendicular to that of the superior SCC, which is seen in cross section, and parallel to that of the posterior SCC, which is well seen, as is the common crus. The turns of the cochlea are well seen on this view, as are the round window, descending facial nerve canal, and incudomalleolar joint. This view is perpendicular to the labyrinthine segment and parallels the tympanic segment of the facial nerve canal. It is also parallel to the tensor tympani muscle.23,54 Pöschl’s view (see Fig. 21-4E and F).  This plane is parallel to the plane of the superior SCC, which can be seen throughout its course appearing as a ring, and the bone separating it from the middle cranial fossa is clearly defined. This plane is perpendicular to the posterior SCC. This plane is also perpendicular to the tympanic segment of the facial nerve canal, which is seen in cross section on this view. This plane is parallel to the plane of the axis (modiolus) of the cochlea and the labyrinthine segment of the facial nerve canal.23,54

A

B

FIG 21-8  EAC atresia. Newborn boy with unilateral ear deformity. Axial high-resolution multidetector CT image of the left temporal bone (A, superior; B, inferior) shows membranous and bony EAC atresia (long white arrow) and atretic middle ear cavity (short white arrow). Malleus and incus are rotated and dysplastic, with fusion at the incudomalleolar articulation and middle ear wall (long black arrow). Note aberrant inferolateral course of the facial nerve (short black arrow).

http://radiologyebook.com

628

PART II  CT and MR Imaging of the Whole Body

nerve VIII results in decreased caliber of the IAC. Other associated anomalies include hypoplasia of the petrous bone, diminished volume of the mastoid and middle ear, flattening of the medial wall of the middle ear, and stapes aplasia or malformation.42,75,87 In cochlear aplasia, there is complete absence of the cochlea. The vestibule is present but often dysplastic. In cochlear hypoplasia, the cochlea and vestibule are separated but their dimensions are smaller than usual. The hypoplastic cochlea is seen as a small bud off of  the IAC.42,87 Common cavity deformity is characterized by a single featureless labyrinthine cavity with no distinct vestibule, cochlea, or SCCs.42,87 Mondini’s malformation, more recently classified as incomplete partition (IP) type II by Sennaroglu, has a cochlea with 1.5 turns. The middle and apical turns of the cochlea fuse to form one cavity owing to an interscalar defect. Dilatation of the vestibular aqueduct and vestibule are frequent associated anomalies42,87,107 (Fig. 21-9). The IP type 1 deformity (pseudo-Mondini’s or cystic vestibulocochlear malformation) consists of a cystic-appearing cochlea without a modiolus and a dilated vestibule without any internal structure. This is not associated with an enlarged vestibular aqueduct. IP type I is thought to result from an insult at an earlier stage of development than IP type II.87

TABLE 21-1  Congenital Malformations of

the Inner Ear

Week of Gestation

Malformation

3 Late 3 4 5 6 7

Michel deformity Cochlear aplasia Common cavity Incomplete partition type I Cochlear hypoplasia Incomplete partition type II

A

Partition defects more subtle than those seen in Mondini’s malformation have also been described, consisting of modiolar and interscalar defects, often in association with an enlarged vestibular aqueduct.55 Enlarged vestibular aqueduct is a common cause of progressive hearing loss in children, with sudden exacerbations often occurring after minor trauma or illness. In 90% of cases, this malformation is bilateral. The aqueduct is considered dilated if it is larger than the diameter of an SCC or more than 1.5 mm wide.67 The endolymphatic sac within the aqueduct may or may not be proportionately enlarged.112 Many of these patients have cochlear abnormalities.55 The superior SCC is the first of the SCCs to form, and the lateral SCC is the last to form. In most cases, only the lateral SCC can have isolated anomalous development; the two important exceptions to this rule occur in Waardenburg’s and Alagille’s syndromes, where there is isolated absence of the posterior SCC.42 CT evaluation of congenital inner ear malformations should include evaluation of the cochlear and vestibular aqueducts. In patients with large cochlear aqueducts, there is a high incidence of CSF leaks near the oval window (stapes gusher). Gushing occurs because the absence of bone at the base of cochlea allows for communication between the CSF in the IAC and perilymph. Preoperative recognition of this entity plays an important role in identifying patients at risk for perilymphatic stapes gushing and avoids worsening of sensorineural hearing loss.53 Cochlear implants have become a promising treatment option for children with sensorineural hearing loss. Before placement of cochlear implants, patients are studied with CT and MRI to characterize the inner ear anomalies. CT is superior to study the bony structures for surgical planning, whereas MRI can confirm that the space within the bony otic capsule is filled with fluid and allows evaluation of the vestibulocochlear nerve.107

Semicircular Canal Dehiscence Syndrome.  SCC dehiscence syndrome results from absence of the osseous covering of an SCC, usually

B FIG 21-9  Incomplete partition type II or Mondini’s malformation. A and B, Axial high-resolution multidetector CT images of temporal bones show an incomplete partition of the cochlea, with normal basal turn and cystic apices, with 1.5 turns (black arrows). Dilatation of the vestibule (long white arrows) and the vestibular aqueduct is also seen (short white arrows).

http://radiologyebook.com

CHAPTER 21  Temporal Bone

A

629

B FIG 21-10  Superior semicircular canal dehiscence. Coronal oblique high-resolution multidetector CT image of the right (A) and left (B) temporal bones shows an absence of osseous covering overlying superior SCC on the left (arrow). Note normal osseous covering on the right (A).

the superior SCC. The dehiscence creates a “third window” in the labyrinth in addition to the oval and round windows.63 It is unclear whether this dehiscence is congenital or acquired.69 The abnormal connection between the middle cranial fossa and SCC results in abnormal vestibular function that may be provoked by sound (Tullilo’s syndrome) or changes in intracranial pressure. SCC dehiscence is seen as absence of bone overlying an SCC on CT or absence of overlying low signal from cortical bone on thin-section heavily T2-weighted MRIs16 (Fig. 21-10). Browaeys et al. have suggested using MRI with heavily T2-weighted images as a screening test in patients with vestibulocochlear symptoms, because in their study MRI had a sensitivity and negative predictive value of 100% for SCC dehiscence. Only patients with positive findings on MRI should undergo subsequent evaluation with CT; false positives can occur with MRI.16

Vascular Anomalies It is important to be aware of vascular anomalies to prevent vascular injury during surgery or mistakenly characterizing them as some other pathology. Agenesis of the petrous internal carotid artery can be mistaken for internal carotid artery occlusion. The carotid canal will not develop in cases of agenesis of the carotid artery. On CT the absence of the carotid canal and carotid sulcus is the clue that distinguishes agenesis of the petrous carotid artery from carotid artery occlusion. The internal carotid artery can have an aberrant course within the middle ear cavity.61 An intratympanic internal carotid artery is thought to occur from involution of a segment of the internal carotid artery during development. The actual intratympanic segment may be a collateral branch that occurs from the anastomoses of the inferior tympanic branch of the ascending pharyngeal artery (from the external carotid artery) with the caroticotympanic artery of the internal carotid artery. This aberrant course of the artery may be asymptomatic, but it can cause pulsatile tinnitus or conductive hearing loss if the artery is close enough to the ossicles. This anomaly may be confused with a middle ear mass such as a glomus tympanicum or hemangioma. Awareness of this entity and its imaging findings is important because it can prevent it from being biopsied or resected. On CT the carotid artery will bulge into the hypotympanum owing to dehiscence of the bony wall of the carotid canal. Other CT findings include absence of the vertical carotid canal and reduced caliber of the aberrant internal carotid artery as it passes through an enlarged inferior tympanic canal. On CT angiogram or MR angiogram, the intratympanic carotid artery will be located more posterior and lateral than normal and appear hypoplastic.13,61

Another type of intratympanic vascular anomaly is a persistent stapedial artery. The stapedial artery normally exists only during embryogenesis. If the artery does not regress, the middle meningeal artery does not develop. The persistent stapedial artery will then supply the structures normally supplied by the middle meningeal artery.88 Because the persistent stapedial artery originates from the petrous internal carotid artery, this implies an internal carotid artery route to the usual middle meningeal artery territory instead of an external carotid artery route. After branching from the petrous internal carotid artery, the persistent stapedial artery runs with the facial nerve through the tympanic cavity and then exits into the middle cranial fossa to supply the structures normally supplied by the middle meningeal artery. A persistent stapedial artery may be asymptomatic but may also cause tinnitus. It is important to be aware of persistent stapedial artery in patients undergoing stapes surgery, because the artery courses through the obturator foramen of the stapes and may be injured. Clues to an intratympanic course of the stapedial artery are a small canaliculus leaving the carotid canal, a linear structure crossing the middle ear over the promontory, enlargement of the tympanic portion of the facial nerve canal, or a separate canal parallel to the facial nerve and absence of the ipsilateral foramen spinosum. The foramen spinosum will be absent owing to absence of the middle meningeal artery that normally courses within it97 (Fig. 21-11). Anomalies of the jugular vein and bulb are common. A high-riding jugular bulb is a bulb present at the level of the IAC or cochlea. By itself it is of no clinical significance. However, if there is also dehiscence of the bony plate separating the jugular bulb from the hypotympanum, the high jugular bulb can then protrude into the middle ear and cause pulsatile tinnitus (Fig. 21-12). A jugular bulb diverticulum appears as an outpouching of the jugular bulb posterior to the IAC (Fig. 21-13).11 It is important to be aware of this benign entity to prevent mistaking it for a tumor. On MRI, slow flow in a diverticulum may mimic a mass.

Inflammatory Lesions Both CT and MRI are important in identifying the primary changes and complications of temporal bone inflammatory disease. CT is useful for detecting bony erosions and soft tissue masses associated with middle ear inflammatory disease. Contrast-enhanced MRI studies can be used to show enhancing granulation tissue.

Acute Otomastoiditis.  Acute otomastoiditis appears as middle ear and mastoid air cell opacification with fluid and nonspecific debris. The fluid in serous otitis media cannot be distinguished from the pus seen with purulent otitis. Subacute otitis media shows similar findings

http://radiologyebook.com

630

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 21-11  Persistent stapedial artery. Axial high-resolution multidetector CT image of the right (A, B) and left (C) temporal bones shows persistent stapedial artery contained in an osseous covering originating from the petrous internal carotid artery, crossing the middle ear over the promontory (white arrows). Note absence of the right foramen spinosum and normal left foramen spinosum (black arrow).

in the mastoid air cells and middle ear, with the additional finding of focal or diffuse mucosal thickening. Spread of infection to the bone with erosion of the mastoid septae, lateral mastoid cortex, or cortex over the sigmoid plate is termed coalescent mastoiditis. High-resolution CT is the optimal imaging modality to assess for evidence of bone erosion94 (Fig. 21-14). MRI is mostly helpful for detecting suspected complications such as epidural abscess, subdural empyema, petrous apicitis, or thrombosis of the sigmoid sinus.103

Chronic Otomastoiditis.  Chronic otomastoiditis follows repeated bouts of otitis media. It results in nonspecific opacification of the middle ear and mastoid air cells by granulation tissue and effusion.50 Reactive sclerosis results in thickening of the bony septa of the mastoid. If chronic otomastoiditis occurs during maturation of the mastoid during childhood, there will be a gradual reduction in the number of mastoid air cells, resulting in poorly developed mastoids. Chronic suppurative otitis media may occur in isolation or have an associated

http://radiologyebook.com

631

CHAPTER 21  Temporal Bone A

*

FIG 21-14  Otomastoiditis and subperiosteal abscess. Axial highFIG 21-12  High-riding jugular bulb. Axial high-resolution multidetector CT image of the right temporal bone shows a high-riding jugular bulb with thinning of the sigmoid plate (arrow).

resolution multidetector CT image of the left temporal bone shows fluid in the middle ear cavity and mastoid air cells, with osseous erosion (arrows) and associated left postauricular soft tissue mass (*).

A

A

B FIG 21-13  Jugular bulb diverticulum. Axial (A) and coronal (B) high-resolution multidetector CT image of the right temporal bone shows an outpouching of the jugular bulb posterior to the right internal auditory canal (arrows).

http://radiologyebook.com

632

PART II  CT and MR Imaging of the Whole Body

FIG 21-15  Tympanosclerosis in a 45-year-old man with bilateral chronic otitis media and conductive hearing loss. Axial high-resolution multidetector CT image of the left temporal bone shows lack of visualization of normal ossicular articulations, with an osseous mass (black arrow) and thickened, retracted tympanic membrane (white arrow).

cholesteatoma or middle ear cholesterol granuloma. Ossicular erosions can occur with chronic otomastoiditis even in the absence of a cholesteatoma, most likely related to the action of osteoclasts and histiocytes.94 Conductive hearing loss can occur from chronic middle ear infection secondary to postinflammatory ossicular fixation. The fixation may be due to fibrous tissue, hyalinization of cartilage, or new bone formation. Fibrous tissue will appear as soft tissue density around  the ossicles. Hyalinization of cartilage (tympanosclerosis) will appear as punctate, nodular, or weblike calcifications in the middle ear,  often involving ossicular tendons or ligaments (Fig. 21-15). If the calcifications occur in the oval window, they can mimic fenestral otosclerosis. New bone formation will be apparent as dense bone in the epitympanum.94

Petrous Apicitis.  Petrous apicitis is an infectious osteitis resulting most often from adjacent spread of otomastoiditis, usually in the setting of a pneumatized petrous apex. Cranial nerves V and VI course close to the petrous apex, separated from the bone by dura. The osteitis may cause Gradenigo’s syndrome, which manifests as facial pain or diplopia due to secondary inflammation of the adjacent cranial nerves. CT findings of petrous apicitis include opacification of a pneumatized petrous apex and bone destruction (Fig. 21-16A and B). On MRI the petrous apex is opacified and enhances (see Fig. 21-16C-E). MRI may also show enhancement of cranial nerves V and VI, Meckel’s cave, and dura adjacent to the petrous apex and cavernous sinus.38,80

Cholesteatoma.  A cholesteatoma is a conglomerate mass of reactive cells involved in a localized chronic inflammatory process. Cholestea-

tomas are lined by stratified keratinizing squamous epithelium with subepithelial fibroconnective or granulation tissue and keratin debris.106 Cholesteatomas are often clinically problematic because of their growth over time, the osseous erosions they cause, and their tendency to recur after resection. Most cholesteatomas are intratympanic.50 Very rarely they may occur in the EAC, mastoid, or petrous apex. The vast majority (98%) of cholesteatomas in the middle ear are acquired. A variety of theories have been proposed to explain the development of acquired cholesteatomas, which are beyond the scope of this chapter. Acquired cholesteatomas can be classified as primary acquired and secondary acquired lesions. Primary acquired cholesteatomas (80% of all cholesteatomas) result from chronic infection with development of granulation tissue behind an apparently intact TM, usually in the region of the pars flaccida. Secondary acquired cholesteatomas (18% of all cholesteatomas) develop through a perforation in the TM, usually involving the pars tensa and less often the pars flaccida.9 On CT, primary acquired cholesteatomas are usually seen as a rounded, expansile, soft tissue mass in Prussak’s space, with erosion of the scutum and medial displacement of the ossicles. Secondary acquired cholesteatomas are often located in the facial recess and sinus tympani. Osseous erosion by a cholesteatoma may additionally lead to facial nerve canal dehiscence, labyrinthine fistulas (usually affecting the lateral SCC), and dehiscence of the tegmen tympani with or without meningoencephalocele formation50,78,104 (Fig. 21-17). Congenital cholesteatomas (2% of all cholesteatomas) develop from embryonic epithelial rests that can occur anywhere in the temporal bone, including the middle ear. In the middle ear, congenital cholesteatomas most frequently originate in the anterior superior quadrant of the middle ear at the eustachian tube orifice near the anterior tympanic ring.90 A small nodular mass is usually seen on CT (Fig. 21-18). Ossicular erosions are uncommon with anteriorly located lesions, but erosion of the facial nerve canal may occur.9 It is important to note on imaging whether a cholesteatoma occupies the sinus tympani or facial nerve recess of the middle ear cavity. The sinus tympani and facial nerve recess are blind spots during surgical resection of cholesteatomas and are the most common and second most common sites of recurrent cholesteatoma, respectively. With a traditional operating microscope, the surgeon can only visualize structures directly ahead. This straight-line view is what creates these blind pockets during surgery. If the surgeon is aware of a cholesteatoma in the sinus tympanic and facial nerve recess, the surgical approach may be altered to remove the cholesteatoma from these sites and prevent a source of recurrence.5,43 Differentiation of cholesteatoma from a chronic otitis media may be impossible on CT owing to similar appearance of the entities  and lack of bony destruction with small cholesteatomas. Further  complicating differentiation is the fact that cholesteatomas may or  may not be accompanied by chronic otitis media. Standard MRI sequences cannot differentiate cholesteatomas from other inflammatory lesions. However, half-Fourier acquisition single-shot turbo-spin echo (HASTE) diffusion-weighted sequences have been shown to be  helpful in detecting cholesteatomas larger than 5 mm. Cholesteatomas  will appear hyperintense on diffusion-weighted imaging (DWI)4,44 (Fig. 21-19).

Malignant External Otitis.  Malignant (necrotizing) external otitis is most often associated with a Pseudomonas infection in older diabetic patients or immunosuppressed patients. Clinically it will be apparent because of otalgia and pus in the EAC.84 The infection may affect just the soft tissues and cartilage. However, it can also lead to osteomyelitis of the adjacent temporal bone

http://radiologyebook.com

CHAPTER 21  Temporal Bone 22

A

B

C FIG 21-16  Gradenigo’s syndrome in a 49-year-old woman with history of left otitis media. Axial highresolution multidetector CT image of the right (A) and left (B) temporal bones shows opacified left mastoids, left middle ear cavity, and left petrous apex (white arrow in B). Axial T2-weighted MRI (C) demonstrates fluid opacification of the left petrous apex (white arrow in C), with enhancement (white arrow in E) on postcontrast T1-weighted image. Continued

http://radiologyebook.com

633

634

PART II  CT and MR Imaging of the Whole Body

E

D

FIG 21-16, cont’d  Note normal marrow signal in the right petrous apex (black arrow) and fluid signal in left petrous apex (white arrow) on non–fat-saturated precontrast T1-weighted image (D).

P

A

P

B FIG 21-17  Cholesteatoma in a 43-year-old woman with left facial paralysis. Coronal high-resolution multidetector CT images (A and B) of the left temporal bone show marked thickening of the tympanic membrane, with a perforation in the superior half and calcifications involving the lower half (white arrow). Abnormal soft tissue is seen within the middle ear, with eroded incus and stapes, and is contiguous with the dehiscent tympanic segment of the facial nerve (black arrow).

http://radiologyebook.com

CHAPTER 21  Temporal Bone

A

635

B FIG 21-18  Congenital cholesteatoma in a 12-year-old boy with left conductive hearing loss and a white mass on otoscopy. Axial (A) and coronal (B) high-resolution multidetector CT images of the left temporal bone show a small, well-circumscribed lobular mass (white arrows) within the posterior mesotympanum, with demineralization of the long process of the incus (black arrow).

structures or skull base, with resultant cranial nerve palsies. Cranial nerve VII is the most common nerve to be involved, owing to dehiscence of the facial nerve canal or infection spreading through the stylomastoid foramen or mastoid segment of the facial nerve.70 Cranial nerves IX, X, and XI will be affected if infection spreads to the jugular foramen, and cranial nerves V and VI will be affected if infection spreads to the petrous apex. The infection can often smolder in the skull base if inadequately treated. Patients who have resolution of canal symptoms may present with cranial nerve palsies of seemingly unknown cause if osteomyelitis is not suspected.18,76 CT findings in external otitis include soft tissue thickening of the EAC, enhancement of canal granulation tissue, and osseous erosions (Fig. 21-20). MRI is more useful to assess soft tissue spread of disease, skull base osteomyelitis, and cranial nerve palsies (Fig. 21-21).50,76

Labyrinthitis.  Labyrinthitis is an inflammatory process of the perilymphatic space that results in changes in the endolymphatic space. It can result from infection, trauma, or autoimmune disease. MRI is sensitive to subtle changes of the membranous labyrinth in patients who present with sudden symptoms such as vertigo or hearing loss. The basal turn of the cochlea is most commonly affected. Acutely, inner ear enhancement is seen on MRI. Increased precontrast T1 signal can represent hemorrhage within the labyrinth.105 In the intermediate stage, fibrosis results in loss of fluid signal in the affected inner ear structures on heavily T2-weighted MRI images. Labyrinthitis ossificans is a late stage of labyrinthitis in which a cascade of inflammatory reactions results in ossification with new bone formation.29 It is only in the late stage when the CT becomes abnormal with bony attenuation in the cochlea, vestibule, and/or SCCs29,50 (Fig. 21-22).

Bell’s Palsy.  Bell’s palsy is the most common cause of acute facial paralysis. It is thought to be due to reactivation of a viral infection in the geniculate ganglion. On MRI there may be variable pathologic contrast enhancement of cranial nerve VII, but pathologic enhancement of the nerve is not specific for Bell’s palsy; it can occur with other diseases. Normal areas of contrast enhancement along cranial nerve VII occur at the geniculate ganglion and the tympanic and mastoid segments, which is actually enhancement of the normal vascular plexus that surrounds these portions of the nerve. Enhancement is pathologic

when it occurs in the cisternal, meatal, labyrinthine, and extracranial (just beyond the stylomastoid foramen) segments of the nerve.32

Neoplastic Lesions Benign Neoplasms

Osteomas and exostoses.  Osteomas are rare benign tumors that are often incidentally found on CT imaging. They are commonly located at the tympanosquamous suture within the EAC but can also arise in the mastoid process and IAC.6,17,26 They are considered true neoplasms as opposed to exostoses, which are considered reactive periosteal bony growths, most commonly in response to prolonged cold water exposure. Radiographically, osteomas are unilateral and solitary pedunculated hyperdense masses on CT, whereas exostoses present as multiple sessile, bilateral, hyperdense masses (Fig. 21-23). Often these lesions are clinically asymptomatic and do not require surgery unless there is conductive hearing loss, recurrent EAC infections, or significant cerumen retention due to the significant surgical complications associated with their removal, including TM perforation, infection, ossicular disruption, and injuries to the TMJ and facial nerve.102 Even with surgical removal, significant exostoses often recur despite cessation of cold water exposure.99 Schwannomas.  Schwannomas are benign tumors of the temporal bone typically arising along cranial nerves VII and VIII. Vestibular schwannomas (VS) are the most common IAC/CPA cistern tumor and the second most common extraaxial neoplasm in adults. Facial nerve schwannomas are rare.15 VS most commonly involve the inferior vestibular nerve and classically arise at the porus acusticus. VS should be considered in patients with symptoms of sensorineural hearing, tinnitus, disequilibrium, or facial nerve palsy, although they are often asymptomatic. MRI has better sensitivity for detecting these tumors than CT. Classic MR features of vestibular schwannomas include a cylindrically shaped (smaller size) or “ice cream cone”–shaped (larger size) mass with T2 hypointense signal and avid enhancement (Fig. 21-24). As they enlarge they can extend into the posterior fossa, causing mass effect on the brainstem or hydrocephalus. Typically the border of a VS extending into the CPA forms an acute angle with the temporal bone. The rate of growth can be slow; therefore long-term MR surveillance is appropriate.7

http://radiologyebook.com

636

PART II  CT and MR Imaging of the Whole Body

*

FD

L

L

5 cm 5 cm

A

B

C

D FIG 21-19  Recurrent cholesteatoma in a 27-year-old woman with skull base fibrous dysplasia (FD), status post right partial mastoidectomy. Axial high-resolution multidetector CT images of the right temporal bone (A, superior; B, inferior) show a soft tissue mass in the right mastoidectomy bed (* on A) with dehiscence of the lateral wall of the descending mastoid segment of the facial nerve (black arrow on B). Note that the external auditory canal is closed off owing to prior surgery (long white arrow), with a dysmorphic and eroded head of the malleus and portions of the incus (short white arrow). On DWI this mass was hyperintense on b-1000 imaging sequences (C, white arrow), with low ADC values (D, white arrows).

http://radiologyebook.com

637

CHAPTER 21  Temporal Bone

*

L

* * A A

B FIG 21-20  External otitis. Axial (A) and coronal (B) high-resolution multidetector CT images of the left temporal bone show debris in the external auditory canal, with osseous erosion along the posterior external auditory canal (white arrow). Note the extensive swelling of the left ear and left periauricular soft tissues (*). Minimal tissue in Prussak’s space (black arrow), with partial opacification of the mastoid air cells, without scutum erosion or confluent mastoiditis.

There are three distinct surgical approaches for VS every radiologist should know: the middle cranial fossa (MCF), suboccipital, and translabyrinthine approaches. The MCF approach is used for patients with good hearing and mainly intracanalicular lesions with less than 1 cm CPA extension, because it is the only technique that does not violate the inner ear structures. However, it has a slightly increased risk of facial nerve palsy. For lesions with greater CPA extension, the suboccipital approach is used and can preserve hearing function; however, it has limited access into the lateral IAC and is prone to CSF leaks. Finally, the translabyrinthine approach is reserved for patients who have lost hearing function, but it offers lowest recurrence rates.89 Facial nerve schwannomas can occur along the entire course of the facial nerve but most commonly involve the geniculate ganglion.109 The presenting symptoms can provide insight on the location of the schwannoma. For example, sensorineural hearing loss is caused by compression of cranial nerve VIII at the CPA, whereas conductive hearing loss is caused by ossicular chain dysfunction from tympanic cavity schwannomas. Classic MR features include a fusiform mass with T1 hypo- to isointense signal, T2 hyperintense signal, and avid enhancement along the facial nerve. Modern MR and CT techniques with thinner sections have recently shown a variety of appearances, including “dumbbell shape” in the CPA-IAC, multilobular morphology in the tympanic segment, and an “invasive” appearance in the mastoid segment.109 CT can evaluate for bony scalloping/remodeling involving the IAC and geniculate ganglion, which can distinguish facial from vestibular schwannomas (Fig. 21-25). If the tumor is proximal to the geniculate ganglion with preserved hearing and minimal CPA extension, the MCF approach is used. The suboccipital approach is used when there is more than 1-cm CPA extension. If hearing function is lost, the translabyrinthine approach is favored, given the direct access. Tympanic-segment schwannomas are reached via the transmastoid approach; mastoid-segment schwannomas may also require following the nerve into the parotid gland.109

Meningiomas.  Meningiomas are a common intracranial neoplasm but rarely involve the temporal bone.37,98 Although histologically they arise from arachnoid caps cells like their intracranial counterparts, there is controversy regarding the exact origin.98 When they occur the common locations include the tegmen tympani, jugular foramen, and IAC. Identifying not only the location of the meningioma but spread of the tumor is important in the temporal bone. Jugular foramen and tegmen tympani meningiomas characteristically spread into the middle ear cavity, whereas IAC meningiomas spread into the intralabyrinthine structures. When in the middle ear cavity, they can encase the ossicular chain, resulting in conductive hearing loss. Middle ear cavity meningiomas can be distinguished from cholesteatomas by lack of bony erosions and enhancement and can be distinguished from glomus jugulare paragangliomas by permeative-sclerotic bony changes rather than permeative-destructive bony changes.37 Temporal bone meningiomas can present with a similar clinical presentation to other commonly found lesions; therefore differentiating imaging features can aid significantly in surgical planning. On CT, meningiomas can be iso- to hyperdense, with some containing calcification, but a majority exhibit avid uniform enhancement (Fig. 21-26A and B). On MRI, they can be iso- to hypointense on T1 and intermediate signal on T2. Marked T2 hyperintense signal and flow voids are not typical. The classic “dural tail” enhancement is typically present, which distinguishes it from schwannoma (see Fig. 21-26C and D). The border between a meningioma in the CPA cistern and the temporal bone is usually an obtuse angle.37 Paragangliomas.  Paragangliomas, also known as glomus tumors, are derived from neural crest paraganglion chief cells and develop throughout the body. They are classified based on their location. Specifically, paragangliomas that develop along the jugular foramen wall are called glomus jugulare, and ones that develop along the tympanic plexus of Jacobsen’s nerve, a glossopharyngeal nerve branch, are called glomus tympanicum. They are the most common middle ear tumor

http://radiologyebook.com

638

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 21-21  Skull base osteomyelitis. A, Axial high-resolution multidetector CT image of the left temporal bone shows areas of cortical deficiency, with osseous erosion and destruction along the inner cortical table of the left temporal bone at the level of the mastoid and junction of the left sigmoid and distal transverse sinuses (long white arrows). B, Precontrast axial T1-weighted MRI shows marrow signal abnormality of the left skull base involving the left occipital bone, occipital condyle, and lateral clivus (white arrows). No gadolinium-based contrast agent was administered, because the patient was in renal failure. Opacification of the left mastoid air cells is seen on axial CT image (short white arrow in A), as well as on axial T2-weighted MRI (C), along with osseous narrowing of the left external auditory canal (black arrow in A).

and second most common temporal bone tumor after vestibular schwannomas. MRI features include a mass with low to intermediate signal on T1-weighted and hyperintense signal on T2-weighted images. A mixed pattern with hyperintense and hypointense foci (“salt-andpepper appearance”) on T1- and T2-weighted images is often seen with paragangliomas, particularly larger lesions.73 Paragangliomas have avid enhancement. On CT there can be permeative-destructive bony changes, particularly at the jugular foramen, which distinguishes them from meningiomas. Glomus jugulare tumors characteristically cause dehiscence of the hypotympanum of the middle ear, whereas glomus tympanicum tumors usually arise on the cochlear promontory without bony destruction (Fig. 21-27). Most occur as a solitary lesion and are benign, but there are no reliable imaging features to identify malignant tumors unless metastatic lesions are present. MRI and digital subtraction angiography is used preoperatively to assess the collateral cerebral blood flow and contralateral sigmoid sinus and internal jugular vein if the ipsilateral vessels will need to be sacrificed.72,100,101

Treatment options are controversial, with considerable debate in the literature; however, a consensus paper by Gjuric and Gleeson states patients with multifocal tumors must be assessed very carefully before surgical intervention owing to increased risk for profound neurologic deficits including laryngeal function and ability to swallow and hear.34 Epidermoids (epidermoid cysts).  Epidermoid tumors, also known as cholesteatomas, are rare intracranial tumors, with about half occurring in the CPA cistern.85 They arise from ectodermal rests and have a thin, keratinized, stratified squamous epithelium capsule. They are slow growing and nonneoplastic, but because the capsule can become densely adherent on adjacent neurovascular structures, surgical removal can be difficult. Aside from the CPA, they can occur in the EAC, middle ear cavity, and petrous apex. On MRI, they are hypointense on T1-weighted and hyperintense on T2-weighted images, making identification and delineation of their extent difficult. On fluid-attenuated inversion recovery (FLAIR) images they are usually heterogeneously hyperintense, allowing easier detection than on conventional MR sequences. However, the heterogeneous signal and

http://radiologyebook.com

CHAPTER 21  Temporal Bone

A

B FIG 21-22  Labyrinthitis ossificans. Axial (A) and coronal (B) high-resolution multidetector CT images of the right temporal bone shows bony attenuation in the cochlea (long white arrow in A; white arrow in B), vestibule (short white arrow in A) and semicircular canals (black arrow).

A

B FIG 21-23  External auditory canal exostoses in a 52-year-old man with chronic ear infections. Axial highresolution multidetector CT images of the right (A) and left (B) temporal bone show extensive external auditory canal exostoses (surfer’s ear), causing moderate to severe narrowing of the external auditory canals, right worse than left (arrows).

http://radiologyebook.com

639

640

PART II  CT and MR Imaging of the Whole Body

FIG 21-24  Vestibular schwannoma. Axial postcontrast T1-weighted image shows bilateral vestibular schwannomas (arrows) in a patient with neurofibromatosis type 2.

A

adjacent CSF artifacts associated with intracranial epidermoids limit the ability of FLAIR images to define the border of the mass. DWI is very sensitive in detecting epidermoids, which appear hyperintense on this sequence and of low signal on apparent diffusion coefficient (ADC) images.36 Constructive interference in the steady state (CISS) sequences may help delineate tumor extent and differentiate the mass form adjacent nerves and vessels. On CISS sequences, epidermoids  are typically slightly hypointense and heterogeneous compared to CSF59 (Fig. 21-28). Typical CT features include a well-defined, lowattenuating, nonenhancing mass, but atypical features include high attenuation and partly calcified components.91 Miscellaneous benign lesions.  Hemangiomas arise from the capillary bed of the epineurium and can cause compression or invasion of the nerve. In the temporal bone, they most commonly affect the geniculate ganglion but can also occur in the IAC.60 On CT they have been described as having a “honeycomb” high-density matrix due to small spicules of calcium, with enlargement of the facial nerve canal. Partial calcification can be helpful to distinguish facial nerve hemangiomas from facial nerve schwannomas.24 On MRI they are usually isointense on T1-weighted and hyperintense on T2-weighted images relative to gray matter, with small foci of low signal on both sequences, likely related to calcification. There is significant enhancement after contrast administration.111 There is often extension into the temporal bone beyond the facial nerve canal24 (Fig. 21-29). As with the vestibular nerve, schwannomas can also occur in  the vicinity of the temporal bone from the other cranial nerves, including cranial nerves V, IX, X, XI, and XII. Lower cranial nerve lesions within the jugular foramen can mimic paragangliomas but are less vascular.31 Lipomas can also occur, usually within the IAC. Typical features include T1 hyperintense signal without enhancement (Fig. 21-30). Fat saturation techniques are helpful to confirm the diagnosis.25 Arachnoid cysts, particularly in the CPA cistern, can mimic epidermoid cysts. These are nonneoplastic intraarachnoid lesions containing normal CSF. Their attenuation and signal characteristics on CT and MRI, respectively, follow CSF characteristics. They are distinguished from epidermoids, given lack of high signal on DWI. Rarely dermoid cysts or neuroenteric cysts can occur in the CPA.12 Cholesterol granulomas can occur in the temporal bone, usually in the middle ear and mastoid air cells and less commonly in the petrous apex. They characteristically are expansile with hyperintense signal on T1- and T2-weighted images (Fig. 21-31). Although there are two competing theories regarding their pathogenesis, they are slowgrowing collections and are nonneoplastic. When they occur in the petrous apex, they must be distinguished from other diseases processes, including effusion, cholesteatoma, trapped fluid (“leave-me-alone lesion”), petrous apicitis, mucocele, cephalocele, and malignant tumors.2,52,66,77,92

Malignant Neoplasms.  Primary malignant neoplasms of the tem-

B FIG 21-25  Presumed facial nerve schwannoma in a patient with longstanding abnormal left facial movements. A, Axial CT through temporal bones shows fusiform expansion of the left anterior genu (arrow). B, Axial postcontrast T1-weighted image shows an avidly enhancing mass in the left internal auditory canal, extending to the left fallopian canal and geniculate ganglion of the left facial nerve (arrow). Note the normal right side for comparison.

poral bone are relatively uncommon; the most common are squamous or basal cell carcinoma, which most commonly involve the EAC or mastoid region. They are aggressive when they occur, often extending to the EAC, middle ear cavity, or mastoid air cells, with involvement of the facial nerve and TMJ.57 Distant metastases are rare for these tumors.22 Adenocarcinoma can present locally with similar features as squamous cell carcinoma, but lymph node involvement is more common. Chondrosarcomas and less commonly chordomas involve the petrous apex, generally arising along the petrosphenoidal and petrooccipital synchondroses.74 Endolymphatic sac tumors are low-grade slow-growing papillary cystadenomatous tumors that usually present along the retrolabyrinthine petrous bone around the

http://radiologyebook.com

CHAPTER 21  Temporal Bone

* B

A

*

* C

D FIG 21-26  Meningioma. A, Axial CT image at the level of the temporal bones shows a large, partially calcified extraaxial mass in the left CPA cistern (*). B, Note normal appearance of the left IAC (arrow) on axial bone window image. On MRI there is avid postcontrast enhancement (*) on axial (C) and coronal (D) postcontrast T1-weighted images, with presence of “dural tail” (arrow).

GT

*

FIG 21-27  Glomus tympanicum. Axial high-resolution multidetector CT image of the left temporal bone shows a soft tissue mass (GT) at the cochlear promontory (*).

http://radiologyebook.com

641

642

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 21-28  CPA epidermoid cyst. A, Axial CT shows a CSF attenuation lesion in the left CPA cistern (arrow). On MRI it is bright on heavily T2-weighted sequence (B), with no postcontrast enhancement (C). D, Note characteristic diffusion restriction on b-1000 images.

http://radiologyebook.com

CHAPTER 21  Temporal Bone

A

643

B FIG 21-29  Presumed left facial nerve hemangioma in a patient with progressive left facial paralysis. A, Axial high-resolution multidetector CT image of the left temporal bone shows a mixed lytic and sclerotic lesion involving the left fallopian canal and left anterior genu of the facial nerve (long white arrow). Soft tissue from this lesion extends into the middle ear cavity, with dehiscence of the tegmen tympani (short white arrow). B, Intense enhancement of mass (white arrow) is seen on postcontrast axial T1-weighted image.

vestibular aqueduct.8,68 In children the most common primary malignant tumor in the middle ear cavity is embryonal rhabdomyosarcoma. This is a highly aggressive tumor that spreads with bone destruction, early involvement of the facial nerve, and intracranial spread.51 CT and MRI are considered complementary imaging techniques for evaluation of temporal bone malignancies, with CT being highly sensitive in detecting bone erosion, while MRI can delineate the tumor extent and differentiate tumor from nonneoplastic soft tissue.33,65 Malignant tumors can also locally spread into the temporal bone. The most common pathways of extension include nasopharyngeal carcinomas and salivary gland tumors via the eustachian tube and EAC or squamous cell carcinoma and adenoid cystic carcinoma via perineural spread along the skull base cranial nerves. The facial nerve is the most commonly involved nerve for perineural spread71 (Fig. 21-32). Metastases of the temporal bone are not uncommon and can spread hematogenously from the breast, lung, stomach, prostate gland, or kidney. Malignant tumors can mimic meningiomas or paragangliomas, given the permeative bony changes and enhancement.

Traumatic Injuries (Fractures) The classic way to categorize a temporal bone fracture is by the direction it courses relative to the petrous portion of the temporal bone. Longitudinal temporal bone fractures course parallel to the long axis of the petrous pyramid, whereas transverse fractures course perpendicular to the long axis of the petrous pyramid. Fractures of the temporal bone can be complex, with both longitudinal and transverse components96 (Fig. 21-33). Alternative classifications have been

proposed because the prognostic value of the traditional classification has been questioned. These newer classification systems include otic capsule sparing versus otic capsule violating and petrous versus nonpetrous involvement.45,58,79 Longitudinal fractures typically extend from the temporal squamosa across the roof or posterior wall of the EAC into the tympanic cavity and tegmen tympani toward the petrous apex. These often involve the ossicular chain and facial nerve near the anterior genu. Conductive hearing loss is a possible complication of longitudinal fracture if it results in ossicular dislocation or hemotympanum. Facial nerve paralysis occurs in about 20% of patients. There may be extension to the carotid canal with vascular injury. These comprise 70% to 90% of temporal bone fractures.96 Transverse fractures typically extend from the jugular foramen and foramen magnum to the middle cranial fossa. These fractures often extend through the vestibular aqueduct, otic capsule, and IAC near the fundus. Sensorineural hearing loss can occur if there is disruption of the otic capsule around the membranous labyrinth.96 Facial nerve paralysis occurs more often with transverse fractures than longitudinal fractures. It can occur either due to disruption of the facial nerve canal or facial nerve edema from contusion. If there is immediate facial nerve paralysis after a fracture, it is usually due to disruption of the nerve. If there is delayed facial nerve paralysis, it is often due to nerve edema or subtotal injury.110 Pneumolabyrinth is the presence of air in the normally fluid-filled labyrinth. In the presence of trauma, pneumolabyrinth implies an otic capsule disruption with communication between the fluid-filled inner ear and the air-filled middle ear cavity35 (Fig. 21-34). However, air in

http://radiologyebook.com

644

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 21-30  Facial nerve lipoma. Incidental detection of focal fat in the distal mastoid segment of the left facial nerve (arrows), bright on T1- (A) and T2-weighted (B) images, with signal dropout on DWI (C) in a patient being evaluated for dizziness. Note that all commercial diffusion-weighted sequences use some sort of fat suppression to suppress chemical shift artifacts.

http://radiologyebook.com

CHAPTER 21  Temporal Bone

A

B

C FIG 21-31  Cholesterol granuloma in a 65-year-old woman with left-sided aural fullness and vertigo. A, Axial high-resolution multidetector CT image of the left temporal bone shows well-corticated expansile lesion of the left petrous bone (white arrow). On MRI, this lesion is mildly hyperintense on T1-weighted imaging (B), without pathologic enhancement (C) (white arrow). Note the tip of the petrous apex is marrow containing (black arrow on A).

http://radiologyebook.com

645

646

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 21-32  An 80-year-old man with remote history of parotid adenocarcinoma presenting with new left facial paralysis. A, Axial precontrast T1-weighted image shows a spiculated mass in the parotidectomy bed (arrow). Axial (B) and coronal (C) postcontrast fat-suppressed T1-weighted images demonstrate thickening and enhancement of the entire course of the left facial nerve involving the mastoid portion, proximal and distal tympanic segments, geniculate ganglion, and labyrinthine portion, filling the internal auditory canal (arrows) and consistent with perineural spread of neoplasm.

cc JF

A

* *

B FIG 21-33  Complex temporal bone fracture in a 26-year-old male trauma patient. A, Axial high-resolution multidetector CT image of the left temporal bone shows comminuted oblique fracture involving the left mastoid and petrous segments, left jugular foramen (JF), and left carotid canal (CC). Partial opacification of the left mastoid air cells and middle ear cavity (*) from posttraumatic hemorrhage. Gas locules and hypodensity seen along the course of the left transverse and sigmoid sinuses (arrows). B, Axial source images from a CT angiogram demonstrate venous thrombosis of the left transverse and sigmoid sinuses (arrows), extending to the proximal left internal jugular http://radiologyebook.com vein (not shown).

CHAPTER 21  Temporal Bone

647

the labyrinth or cochlea is not specific for trauma, because it can also result from a spontaneous inner ear–middle ear fistula.

Metabolic and Dysplastic Lesions The petrous temporal bone may be affected by a variety of primary bone diseases that can involve other portions of the skeleton. When they involve the temporal bone, the significance is related to disturbances of hearing, balance, or facial nerve function.

Otosclerosis.  Otosclerosis is a hereditary osteodystrophy of the otic capsule that results in hearing loss. In the majority of cases the bilateral otic capsules are affected. Pathologically, otosclerosis consists of an active phase of resorption (otospongiosis) and an inactive phase of immature bone deposition (sclerosis). Otosclerosis can affect any part of the otic capsule. The most common types of otosclerosis are fenestral and cochlear. Fenestral otosclerosis, which involves the oval window and stapes footplate, causes conductive hearing loss. Cochlear otosclerosis, which involves the labyrinthine capsule or cochlea, will cause progressive sensorineural hearing loss.82 Fenestral otosclerosis most commonly begins in the area of the fissula ante fenestram of the oval window82 (Fig. 21-35). It can spread across the stapedial annular ligament to fix the stapes footplate.14 On CT, early fenestral otosclerosis appears as small lucencies within the otic capsule at the anterior margin of the oval window. Later, otosclerotic foci may cover the oval window, stapes, and round window.93

FIG 21-34  Pneumolabyrinth in a trauma patient. Axial high-resolution multidetector CT image of the right temporal bone shows small foci of air within the vestibule (black arrow) and cochlea (white arrow). There was a nondisplaced occipital bone fracture and bifrontal parenchymal contusions (not shown), but no temporal bone fracture or ossicular disruption was identified.

71

5

A

B FIG 21-35  Fenestral otosclerosis in a 42-year-old man with moderately severe mixed hearing loss in the left ear. Axial high-resolution multidetector CT images of the right (A) and left (B) temporal bones show focal area of demineralization anterior to the left oval window (arrow in B). The right side is normal.

http://radiologyebook.com

648

PART II  CT and MR Imaging of the Whole Body

On CT, cochlear otosclerosis appears as a crescentic hypodense rim paralleling the margins of the basal turn of the cochlea, the entire cochlea, or the whole otic capsule. Other patients may have areas of sclerosis. T1-weighted MRI reveals fuzzy soft tissue signal intensities in the cochlear capsule. Contrast enhancement may be seen, which is presumably due to contrast medium pooling in the numerous blood vessels of the otospongiotic foci.28,93 Cavitary otosclerosis, a rare form of otosclerosis, appears as hypodense foci in the anterior wall of the IAC.14 Most patients with otosclerosis are treated with stapedectomy and prosthetic implants. Recurrent hearing loss after prosthetic implants may be due to migration or dislocation of the implant or necrosis/ erosion of the ossicular attachment of the implant.108

Other.  A variety of other metabolic and dysplastic disorders can involve the temporal bone, including osteogenesis imperfecta, Paget’s disease, fibrous dysplasia, and osteopetrosis.3,30,40,49

REFERENCES 1. Abele TA, Wiggins RH: Imaging of the temporal bone. Radiol Clin North Am 53:15–36, 2015. 2. Achilli V, Danesi G, Caverni L, et al: Petrous apex arachnoid cyst: A case report and review of the literature. Acta Otorhinolaryngol Ital 25:296–300, 2005. 3. Alkadhi H, Rissmann D, Kollias SS: Osteogenesis imperfecta of the temporal bone: CT and MR imaging in Van der Hoeve-de Kleyn syndrome. AJNR Am J Neuroradiol 25:1106–1109, 2004. 4. Alvo A, Garrido C, Salas A, et al: Use of non-echo-planar diffusionweighted MR imaging for the detection of cholesteatomas in high-risk tympanic retraction pockets. AJNR Am J Neuroradiol 35:1820–1824, 2006. 5. Badr-el-Dine M: Value of ear endoscopy in cholesteatoma surgery. Otol Neurotol 23:631–635, 2002. 6. Baik FM, Nguyen L, Doherty JK, et al: Comparative case series of exostoses and osteomas of the internal auditory canal. Ann Otol Rhinol Laryngol 120:255–260, 2011. 7. Bakkouri W, Kania RE, Guichard JP, et al: Conservative management of 386 cases of unilateral vestibular schwannoma: Tumor growth and consequences for treatment. J Neurosurg 110:662–669, 2009. 8. Baltacioglu F, Ekinci G, Ture U, et al: MR imaging, CT, and angiography features of endolymphatic sac tumors: Report of two cases. Neuroradiology 44:91–96, 2002. 9. Baráth K, Huber AM, Stämpfli P, et al: Neuroradiology of cholesteatomas. AJNR Am J Neuroradiol 32:221–229, 2011. 10. Benton C, Bellet PS: Imaging of congenital anomalies of the temporal bone. Neuroimaging Clin N Am 10:35–53, 2000. 11. Bilgen C, Kirazli T, Ogut F, et al: Jugular bulb diverticula: Clinical and radiologic aspects. Otolaryngol Head Neck Surg 128:382–386, 2003. 12. Bonneville F, Savatovsky J, Chiras J: Imaging of cerebellopontine angle lesions: An update. Part 1: Enhancing extra-axial lesions. Eur Radiol 17:2472–2482, 2007. 13. Botma M, Kell RA, Bhattacharya J, et al: Aberrant internal carotid artery in the middle-ear space. J Laryngol Otol 114:784–787, 2000. 14. Bou-Assaly W, Mukherji S, Srinivasan A: Bilateral cavitary otosclerosis: A rare presentation of otosclerosis and cause of hearing loss. Clin Imaging 37:1116–1118, 2013. 15. Brackmann DE, Bartels LJ: Rare tumors of the cerebellopontine angle. Otolaryngol Head Neck Surg 88:555–559, 1980. 16. Browaeys P, Larson TL, Wong ML, et al: Can MRI replace CT in evaluating semicircular canal dehiscence. AJNR Am J Neuroradiol 34:1421–1427, 2013. 17. Carbone P, Nelson B: External auditory osteomas. Head Neck Pathol 6:244–246, 2012. 18. Carfrae MJ, Kesser BW: Malignant otitis externa. Otolaryngol Clin North Am 41:537–549, viii–ix, 2008.

19. Carlson BA: Human embryology and developmental biology, Philadelphia, 2014, Elsevier Saunders, pp 269–293. 20. Casselman JW, Kuhweide R, Deimling M, et al: Constructive interference in steady state 3DFT MRI of the inner ear and CP angle. AJNR Am J Neuroradiol 14:47–57, 1993. 21. Casselman JW, Offeciers EF, De Foer B, et al: CT and MR imaging of congenital abnormalities of the inner ear and internal auditory canal. Eur J Radiol 40:94–104, 2001. 22. Chee G, Mok P, Sim R: Squamous cell carcinoma of the temporal bone: Diagnosis, treatment and prognosis. Singapore Med J 41:441–446, 451, 2000. 23. Curtin HD, Gupta R, Bergeron RT: Embryology, anatomy and imaging of the temporal bone. In Som PM, Curtin HD, editors: Head and neck imaging, 5th ed, St. Louis, 2011, Elsevier Mosby, pp 1053–1096. 24. Curtin HD, Jensen JE, Barnes L, et al: “Ossifying” hemangiomas   of the temporal bone: Evaluation with CT. Radiology 164:831–835, 1987. 25. Dahlen RT, Johnson CE, Harnsberger HR, et al: CT and MR imaging characteristics of intravestibular lipoma. AJNR Am J Neuroradiol 23:1413–1417, 2002. 26. Davis TC, Thedinger BA, Greene GM: Osteomas of the internal auditory canal: A report of two cases. Am J Otol 21:852–856, 2000. 27. De la Cruz A, Teufert KB: Reconstruction of the auditory canal and tympanum. In Flint PW, Haughey BH, Lund VJ, et al, editors: Cummings otolaryngology head and neck surgery, 5th ed, Philadelphia, 2010, Elsevier Mosby, pp 2752–2760. 28. de Oliveira Vicente A, Yamashita H, Manguabeira Albernaz P, et al: Computed tomography in the diagnosis of otosclerosis. Otolaryngol Head Neck Surg 134:685–692, 2006. 29. deSouza C, Paparella MM, Schachern P, et al: Pathology of labyrinthine ossification. J Laryngol Otol 105:621–624, 1991. 30. Dozier TS, Duncan IM, Klein AJ, et al: Otologic manifestations of malignant osteopetrosis. Otol Neurotol 26:762–766, 2005. 31. Eldevik OP, Gabrielsen TO, Jacobsen EA: Imaging findings in schwannomas of the jugular foramen. AJNR Am J Neuroradiol 21:1139–1144, 2000. 32. Gebarski SS, Telian SA, Niparko JK: Enhancement along the normal facial nerve in the facial canal: MR imaging and anatomic correlation. Radiology 183:391–394, 1992. 33. Gillespie MB, Francis HW, Chee N, et al: Squamous cell carcinoma of the temporal bone: A radiographic. pathologic correlation. Arch Otolaryngol Head Neck Surg 127:803–807, 2001. 34. Gjuric M, Gleeson M: Consensus statement and guidelines on the management of paragangliomas of the head and neck. Skull Base 19(1):109–116, 2009. 35. Gross M, Ben-Yaakov A, Goldfarb A, et al: Pneumolabyrinth: An unusual finding in a temporal bone fracture. Int J Pediatr Otorhinolaryngol 67:553–555, 2004. 36. Hakyemez B, Aksoy U, Yildiz H, et al: Intracranial epidermoid cysts: Diffusion-weighted, FLAIR and conventional MR findings. Eur J Radiol 54:214–220, 2005. 37. Hamilton BE, Salzman KL, Patel N, et al: Imaging and clinical characteristics of temporal bone meningioma. AJNR Am J Neuroradiol 27:2204–2209, 2006. 38. Hardjasudarma M, Edwards RL, Ganley P, et al: Magnetic resonance imaging of Gradenigo’s syndrome. Am J Otolaryngol 16:247–250, 1995. 39. Harnsberger HR: Temporal bone and skull base. In Harnsberger HR, Osborn AG, Macdonald AJ, et al, editors: Diagnostic and surgical imaging anatomy: Brain, head and neck, spine, 1st ed, Salt Lake City, 2006, Amirsys, Inc, pp II-2–II-87. 40. Heimert TL, Lin DD, Yousem DM: Case 48: Osteogenesis imperfecta of the temporal bone. Radiology 224:166–170, 2002. 41. Ho ACW, Lee R, Co MTH, et al: Is virtual endoscopy of the middle ear useful? Ear Nose Throat J 90:256–260, 2011. 42. Huang BY, Zdanski C, Castillo M: Pediatric sensorineural hearing loss, Part 1: Practical aspects for neuroradiologists. AJNR Am J Neuroradiol 33:211–217, 2012.

http://radiologyebook.com

CHAPTER 21  Temporal Bone 43. Hughes EK, Hughes JP, Madani G: Interpretation of computed tomography of the petrous temporal bone. Otorhinolaryngologist 6:91–98, 2013. 44. Huins CT, Singh A, Lingam RK, et al: Detecting cholesteatoma with non-echo planar (HASTE) diffusion-weighted magnetic resonance imaging. Otolaryngol Head Neck Surg 143:141–146, 2010. 45. Ishman ST, Friedland DR: Temporal bone fractures: Traditional classification and clinical relevance. Laryngoscope 114:1734–1741, 2004. 46. Jackler RK, Luxford WM, House WF: Congenital malformations of the inner ear: A classification based on embryogenesis. Laryngoscope 97(Suppl 40):2–14, 1987. 47. Jahrsdoerfer RA, Yeakley JW, Aguilar EA, et al: Grading system for the selection of patients with congenital aural atresia. Am J Otol 13:6–12, 1992. 48. Jager L, Bonell H, Liebl M, et al: CT of the normal temporal bone: Comparison of multi- and single-detector row CT. Radiology 235:133–141, 2005. 49. Jee WH, Choi KH, Choe BY, et al: Fibrous dysplasia: MR imaging characteristics with radiopathologic correlation. AJR Am J Roentgenol 167:1523–1537, 1996. 50. Juliano AF, Ginat DT, Moonis G: Imaging review of the temporal bone: Part 1. Anatomy and inflammatory and neoplastic processes. Radiology 269:17–33, 2013. 51. Komorowska A, Makowska-Piontek A: Rhabdomyosarcoma of the middle ear in childhood. Otolaryngol Pol 53:417–421, 1999. 52. Koral K, Dowling M: Petrous apicitis in a child: Computed tomography and magnetic resonance imaging findings. Clin Imaging 30:137–139, 2006. 53. Kumar G, Castillo M, Buchman CA: X-linked stapes gusher: CT findings in one patient. AJNR Am J Neuroradiol 24:1130–1132, 2003. 54. Lane JI, Lindell EP, Witte RJ, et al: Middle and inner ear: Improved depiction with multiplanar reconstruction of volumetric CT data. Radiographics 26:112–124, 2006. 55. Lemmerling MM, Mancuso AA, Antonelli PJ, et al: Normal modiolus: CT appearance in patients with a large vestibular aqueduct. Radiology 204:213–219, 1997. 56. Lemmerling MM, Stambuk HE, Mancuso AA, et al: CT of the normal suspensory ligaments of the ossicles of the middle ear. AJNR Am J Neuroradiol 18:471–477, 1997. 57. Leonetti JP, Smith PG, Kletzker GR, et al: Invasion patterns of advanced temporal bone malignancies. Am J Otol 17:438–442, 1996. 58. Little SC, Kesser BW: Radiographic classification of temporal bone fractures: Clinical predictability using a new system. Arch Otolaryngol Head Neck Surg 132:1300–1304, 2006. 59. Liu P, Saida Y, Yoshioka H, et al: MR imaging of epidermoids at the cerebellopontine angle. Magn Reson Med Sci 2:109–115, 2003. 60. Lo W, Shelton C, Waluch V, et al: Intratemporal vascular tumors: Detection with CT and MR imaging. Radiology 171:445–448, 1989. 61. Lo W, Solti-Bohman LG, McElveen JT: Aberrant carotid artery: Radiologic diagnosis with emphasis on high-resolution computed tomography. Radiographics 5:985–993, 1985. 62. Mehanna AM, Baki FA, Eid M, et al: Comparison of different computed tomograph post-processing modalities in assessment of various middle ear disorders. Eur Arch Otorhinolaryngol 2014. doi: 10.1007/ s00405-014-2920-y. 63. Minor LB, Solomon D, Zinreich JS, et al: Sound- and/or pressureinduced vertigo due to bone dehiscence of the superior semicircular canal. Arch Otolaryngol Head Neck Surg 124:249–258, 1998. 64. Mohan S, Hoeffner E, Bigelow DC, et al: Applications of magnetic resonance imaging in adult temporal bone disorders. Magn Reson Imaging Clin N Am 20:545–572, 2012. 65. Moody SA, Hirsch BE, Myers EN: Squamous cell carcinoma of the external auditory canal: An evaluation of a staging system. Am J Otol 21:582–588, 2000. 66. Moore KR, Harnsberger HR, Shelton C, et al: “Leave me alone” lesions of the petrous apex. AJNR Am J Neuroradiol 19:733–738, 1998. 67. Mukerji SS, Parmar HA, Ibrahim M, et al: Congenital malformations of the temporal bone. Neuroimaging Clin N Am 21:603–619, 2011.

649

68. Mukherji SK, Albernaz VS, Lo WW, et al: Papillary endolymphatic sac tumors: CT, MR imaging, and angiographic findings in 20 patients. Radiology 202:801–808, 1997. 69. Nadgir RN, Ozonoff A, Devaiah AK, et al: Superior semicircular canal dehiscence: Congenital or acquired condition? AJNR Am J Neuoradiol 32:947–949, 2011. 70. Nawas MT, Daruwalla VJ, Spirer D, et al: Complicated necrotizing otitis externa. Am J Otolaryngol 34:706–709, 2013. 71. Nemzek WR, Hecht S, Gandour-Edwards R, et al: Perineural spread of head and neck tumors: How accurate is MR imaging? AJNR Am J Neuroradiol 19:701–706, 1998. 72. Noujaim SE, Pattekar MA, Cacciarelli A, et al: Paraganglioma of the temporal bone: Role of magnetic resonance imaging versus computed tomography. Top Magn Reson Imaging 11:108–122, 2000. 73. Olsen WL, Dillon WP, Kelly WM, et al: MR imaging of paragangliomas. AJR Am J Roentgenol 148:201–204, 1987. 74. Oot RF, Melville GE, New PF, et al: The role of MR and CT in evaluating clival chordomas and chondrosarcomas. AJR Am J Roentgenol 151:567–575, 1988. 75. Ozgen B, Oguz KK, Atas A, et al: Complete labyrinthine aplasia: Clinical and radiologic findings with review of the literature. AJNR Am J Neuroradiol 30:774–780, 2009. 76. Patmore H, Jebreel A, Uppal S, et al: Skull base infection presenting with multiple lower cranial nerve palsies. Am J Otolaryngol 31:376–380, 2010. 77. Profant M, Steno J: Petrous apex cholesteatoma. Acta Otolaryngol 120:164–167, 2000. 78. Quaranta N, Liuzzi C, Zizzi S, et al: Surgical treatment of labyrinthine fistula in cholesteatoma surgery. Otolaryngol Head Neck Surg 140:406– 411, 2009. 79. Rafferty MA, Mc Conn Walsh R, Walsh MA: A comparison of temporal bone fracture classification systems. Clin Otolaryngol 31:287–291, 2006. 80. Razek AA, Huang BY: Lesions of the petrous apex: Classification and findings at CT and MR imaging. Radiographics 32:151–173, 2012. 81. Rodriguez K, Shah RK, Kenna M: Anomalies of the middle and inner ear. Otolaryngol Clin North Am 40:81–96, 2007. 82. Roland PS, Meyerhoff WL: Otosclerosis. In Bailey BJ, editor: Head and neck surgery—Otolaryngology, 2nd ed, Philadelphia, 1998, LippincottRaven Publishers, pp 2083–2096. 83. Romo LV, Casselman JW, Robson CD: Congenital anomalies of the temporal bone. In Som PM, Curtin HD, editors: Head and neck imaging, 5th ed, St. Louis, 2011, Elsevier Mosby, pp 1097–1165. 84. Rubin Grandis J, Branstetter BF, 4th, Yu VL: The changing face of malignant (necrotising) external otitis: Clinical, radiological, and anatomic correlations. Lancet Infect Dis 4:34–39, 2004. 85. Schiefer T, Link M: Epidermoids of the cerebellopontine angle: A 20 year experience. Surg Neurol 70:584–590, 2008. 86. Schuknecht HF: Congenital aural atresia. Laryngoscope 99:908–917, 1989. 87. Sennaroglu L, Saatci I: A new classification for cochleovestibular malformations. Laryngoscope 112:2230–2241, 2002. 88. Silbergleit R, Quint DJ, Mehta BA, et al: The persistent stapedial artery. AJNR Am J Neuroradiol 21:572–577, 2000. 89. Silk P, Lane J, Driscoll C: Surgical approaches to vestibular schwannomas: What the radiologist needs to know. Radiographics 29:1955–1970, 2009. 90. Sina-Khadiv M, Alemi AS, Ziai K, et al: Congenital cholesteatoma originating within the facial nerve canal. Otolaryngol Head Neck Surg 143:708–709, 2010. 91. Sirin S, Gonul E, Kahraman S, et al: Imaging of posterior fossa epidermoid tumors. Clin Neurol Neurosurg 107:461–467, 2005. 92. Srinivasan A, Ibrahim M, Mukherji SK: Radiology quiz case 2. Bilateral petrous apex cephaloceles (PACs). Arch Otolaryngol Head Neck Surg 132:1265–1267, 2006. 93. Swartz JD: The otodystrophies: Diagnosis and differential diagnosis. Semin Ultrasound CT MR 25:305–318, 2004. 94. Swartz JD, Hagiwara M: Inflammatory diseases of the temporal bone. In Som PM, Curtin HD, editors: Head and neck imaging, 5th ed, St. Louis, 2011, Elsevier Mosby, pp 1183–1229.

http://radiologyebook.com

650

PART II  CT and MR Imaging of the Whole Body

95. Swartz JD, Harnsberger HR: Imaging of the temporal bone, 3rd ed, New York, 1998, Thieme, pp 240–317. 96. Swartz JD, Kang MD: Trauma to the temporal bone. In Som PM, Curtin HD, editors: Head and neck imaging, 5th ed, St. Louis, 2011, Elsevier Mosby, pp 1167–1181. 97. Thiers FA, Sakai O, Poe DS, et al: Persistent stapedial artery: CT findings. AJNR Am J Neuroradiol 21:1551–1554, 2000. 98. Thompson L, Bouffard J, Sandberg G, et al: Primary ear and temporal bone meningiomas: A clinicopathologic study of 36 cases with a review of the literature. Mod Pathol 16:236–245, 2003. 99. Timofeev I: Exostoses of the external auditory canal: A long term follow-up study of surgical treatment. Clin Otolaryngol Allied Sci 29:588–594, 2004. 100. van den Berg R: Imaging and management of head and neck paragangliomas. Eur Radiol 15:1310–1318, 2005. 101. van den Berg R, Schepers A, de Bruine FT, et al: The value of MR angiography techniques in the detection of head and neck paragangliomas. Eur J Radiol 52:240–245, 2004. 102. Vasama J: Surgery for external auditory canal exostoses: A report of 182 operations. J Otol Rhinol Laryngol 65:189–192, 2003. 103. Vazquez E, Castellote A, Piqueras J, et al: Imaging of complications of acute mastoiditis in children. Radiographics 23:359–372, 2003. 104. Vikram B, Udayashankar S, Naseeruddin K, et al: Complications in primary and secondary acquired cholesteatoma: A prospective comparative study of 62 ears. Am J Otolaryngol 29:1–6, 2008. 105. Weissman JL, Curtain HD, Hirsch BE, et al: High signal from the otic labyrinth on unenhanced magnetic resonance imaging. AJNR Am J Neuroradiol 13:1183–1187, 1992.

106. Wenig BW: Non-neoplastic lesions/disease of the ear and temporal bone. In Atlas of head and neck pathology, Philadelphia, 2008, Saunders Elsevier, pp 719–759. 107. Westerhof JP, Rademaker J, Weber BP, et al: Congenital malformations of the inner ear and the vestibulocochlear nerve in children with sensorineural hearing loss: Evaluation with CT and MRI. J Comput Assist Tomogr 25:719–726, 2001. 108. Whetstone J, Nguyen A, Nguyen-Huynh A, et al: Surgical and clinical confirmation of temporal bone CT findings in patients with otosclerosis with failed stapes surgery. AJNR Am J Neuroradiol 35:1195–2011, 2014. 109. Wiggins RH, 3rd, Harnsberger HR, Salzman KL, et al: The many faces of facial nerve schwannoma. AJNR Am J Neuroradiol 27:694–699, 2006. 110. Woodson EA, Smith RJH: Bilateral pneumolabyrinth diagnostic for otic capsule fractures without high-resolution imaging. Otolaryngol Head Neck Surg 137:969–971, 2007. 111. Yue Y, Jin Y, Yang B, et al: Retrospective case series of the imaging findings of facial nerve haemangioma. Eur Arch Otorhinolaryngol 2014 272:2497–2503, 2015. 112. Yuen HY, Ahuja AT, Wong KT, et al: Computed tomography of common congenital lesions of the temporal bone. Clin Radiol 58:687–693, 2003. 113. Zhen J, Liu C, Wang S, et al: The thin section anatomy of the temporal bone correlated with multislice spiral CT. Surg Radiol Anat 29:409–418, 2007.

http://radiologyebook.com

22  Pharynx Suyash Mohan, Supreeta Arya, Tushar Chandra, and Ellen Hoeffner

INTRODUCTION The pharynx is a fibromuscular tube situated directly anterior to the vertebral column. It extends from the skull base to the lower border of the cricoid cartilage. It comprises a group of six muscles that  are predominantly responsible for the voluntary act of swallowing: three pharyngeal constrictor muscles (superior, middle, and inferior constrictor muscles) and three vertically oriented muscles (stylo­ pharyngeus, salpingopharyngeus, and palatopharyngeus). They are innervated by the pharyngeal branch of the vagus nerve (cranial nerve [CN] X), with the exception of the stylopharyngeus muscle, which is innervated by the glossopharyngeal nerve (CN IX). Based on the location, the pharynx is often subdivided into the following three sections: • Nasopharynx: top of the pharynx, located behind the nasal cavity and extends from the skull base to the soft palate • Oropharynx: located behind the oral cavity and extends from the uvula to the hyoid bone • Hypopharynx: extends from the hyoid bone to the cricopharyn­ geus muscle, located at C5-C6 level, at the lower end of the cricoid cartilage

IMAGING RATIONALE AND TECHNIQUES Computed tomography (CT) and magnetic resonance imaging (MRI) are currently the primary modalities for investigating pathologies of the pharynx. Although the superficial extent of mucosal lesions can be readily identified by the examining physician, their deep submucosal extent is almost totally inaccessible to clinical or endoscopic evalua­ tion. CT offers rapid image acquisition and is helpful in the acute setting for rapidity of acute diagnosis. In particular the high spatial resolution of CT can provide precise anatomic detail. MRI on the other hand offers benefits over CT because of its higher soft tissue contrast resolution, multiplanar capability, and superiority in detecting peri­ neural tumor spread and intracranial invasion. The role of advanced biological imaging techniques such as MR spectroscopy, MRI diffusion and perfusion, CT perfusion, and fluorodeoxyglucose positron emis­ sion tomography (FDG-PET) is discussed toward the end of this chapter.

NASOPHARYNX Anatomy The nasopharynx is the superiormost portion of the pharynx, located immediately below the skull base. It is attached to the undersurface of the clivus by the pharyngobasilar fascia of the superior constrictor muscle. The posterior and lateral walls of the nasopharynx are in

continuity with the posterior and lateral oropharyngeal walls, respec­ tively, and it is inferiorly limited by the soft palate. Along the lateral nasopharyngeal wall there are two indentations and one protrusion into the lumen. The indentations (recesses) are the eustachian tube orifice and the lateral pharyngeal recess (often known by its eponym, the fossa of Rosenmüller), whereas the protru­ sion is the torus tubarius (the cartilaginous portion of the eustachian tube). On axial images, the lateral pharyngeal recess is located pos­ terior to the torus tubarius, and it is superior to the torus tubarius on coronal images (Fig. 22-1A-B). The posterior pharyngeal wall  has an undulating contour from the prevertebral longus capitis muscles, which demarcates the posteromedial aspect of the lateral pharyngeal recess. There is a small lateral hiatus or opening in the skull base attach­ ment known as the foramen of Morgagni through which the eustachian (pharyngotympanic) tube and levator veli palatini muscle pass (see Fig. 22-1C-D). The levator veli palatini and the tensor veli palatini muscles are identifiable on T1-weighted MRI scans in the lateral wall of the nasopharynx and are responsible for opening the eustachian tube and for tensing and elevating the palate to prevent oronasal reflux. The eustachian tube communicates with the middle ear, allowing both middle ear aeration and drainage. The foramen of Morgagni is an essential anatomic structure but is also a potential weak spot through which nasopharyngeal neoplasms or infections may spread directly to the skull base or laterally to the parapharyngeal fat.

Congenital Lesions

Thornwaldt’s Cyst.  Thornwaldt’s cyst, named after German physi­ cian Gustav Ludwig Thornwaldt, is a common developmental benign midline nasopharyngeal mucosal cyst. It is usually asymptomatic and is detected incidentally on imaging. Peak incidence has been variably reported between the ages of 15 and 60 years. It has an autopsy preva­ lence of approximately 4%, with no gender predilection. It results from focal adhesion of pharyngeal mucosa to the notochord (a midline col­ lection of cells that aids embryonic long-axis and neural plate develop­ ment), which is then carried up as the notochord ascends to the developing skull base. This creates a potential space known as the pharyngeal bursa, whose orifice becomes obliterated after an attack of pharyngitis, thereby forming the cyst. A Thornwaldt’s cyst lacks osseous involvement and corresponds to a superior level in the clivus. These are almost always asymptomatic; however, if they become infected the patient presents with persistent or periodic nasopharyngeal drainage, halitosis, and foul taste. Some may present with otitis media due to obstruction of the eustachian tube. Dull occipital headache has also been described. A symptomatic cyst is also called Thornwaldt’s disease.68 Thornwaldt’s cyst appears as a well-circumscribed, thin-walled, midline (but may be slightly paramidline) posterior nasopharyngeal

http://radiologyebook.com

651

652

PART II  CT and MR Imaging of the Whole Body

*

*

LC

B

A

*+

*

*+ LC

C

D FIG 22-1  Normal nasopharynx on CT and MRI. A, Axial postcontrast CT image shows eustachian tube orifice (arrow in front), torus tubarius (*), and lateral pharyngeal recess (arrow behind) along the lateral wall of nasopharynx. Deep to posterior wall of nasopharynx are longus capitus (LC) muscles. B, Reformatted coronal CT image in same patient shows lateral pharyngeal recess (arrow) and torus tubarius (*). C, Axial postcontrast T1-weighted MRI shows eustachian tube orifice (arrow in front), torus tubarius (*), and lateral pharyngeal recess (arrow behind) along lateral wall of nasopharynx. Deep to posterior wall of enhancing nasopharyngeal mucosa are longus capitus (LG) muscles. D, On axial postcontrast T1-weighted MRI slightly inferior to C, the tensor levator palatini (*) and levator veli palatini (+) muscles are seen.

mucosal space cyst lodged between the prevertebral muscles. It varies in diameter from a few millimeters to several centimeters. On CT scans, a small cyst may be missed, but a larger one usually is seen as a hypodense fluid-attenuation lesion (Fig. 22-2A). CT attenuation of the cyst increases, as does its protein content, and occasionally it may mimic a soft tissue mass. On MRI the T1-weighted signal intensity increases from low to high with increasing protein content. The cyst is

bright on T2-weighted (see Fig. 22-2B) and fluid-attenuated inversion recovery (FLAIR) images, with thin peripheral enhancement on post­ contrast images.

Persistent Canalis Basilaris Medianus.  Persistent canalis basilaris medianus (CBM) is a rare and typically asymptomatic con­ genital anatomic variant of the basiocciput, believed to arise from

http://radiologyebook.com

CHAPTER 22  Pharynx

A

653

B FIG 22-2  Thornwaldt cyst. A, Low-attenuation well-circumscribed mass (arrows) in midline of nasopharynx on postcontrast axial CT image. B, The mass (arrows) is of fluid signal on axial T2-weighted image.

FIG 22-3  Persistent canalis basilaris medianus is seen as a well-corticated lucency in the basiocciput (arrow).

notochordal remnants. It occurs in approximately 2% to 3% of adults as a well-defined corticated tubular structure, usually over 2 mm in diameter, originating on the intracranial surface of the basiocciput  in close proximity to the anterior rim of the foramen magnum (Fig. 22-3). It is thought to represent an embryologic remnant demarcating the cephalic end of the notochord and corresponds to the course of the notochordal canal in the caudal basiocciput. A complete CBM traverses entirely through the basioccipital bone, whereas an incom­ plete CBM can extend partially through either the cranial or pharyn­ geal portions of the basiocciput but does not extend all the way through to the other surface. Clival defects from this vestige can be associated with several anomalies. MRI may detect accompanying

FIG 22-4  Fossa navicularis is seen as a notch-like defect in the basiocciput (arrow).

cysts and help differentiate between other possible etiologies, such as a cephalocele, and ecchordosis physaliphora, which is usually intradu­ ral within the prepontine cistern and is attached to the dorsal wall of the clivus via pedicles.51

Fossa Navicularis.  Fossa navicularis is a notchlike bone defect in the basiocciput (Fig. 22-4) and is also an incidental imaging finding. Its incidence is around 3%, with many being less than 2 mm in  size. Rarely, lymphoid tissue can be found within a prominent fossa

http://radiologyebook.com

654

PART II  CT and MR Imaging of the Whole Body

navicularis, what some authors have termed fossa navicularis magna. Transmission of infection from oropharyngeal soft tissues via the fossa navicularis has been rarely reported as a cause of clival osteomyelitis, raising the possibility of affected patients being more susceptible to other infections such as meningitis. It is important to be aware of this rare congenital variant to prevent mischaracterization with the many more worrisome possibilities and possibly avoid biopsy or unnecessary treatment attempts.72

Craniopharyngeal Canal.  The craniopharyngeal canal (CPC) is a rare well-corticated defect through the midline of the sphenoid bone, extending from the floor of the sella turcica to the roof of the naso­ pharynx (Fig. 22-5). The CPC is postulated to arise from an error in the normal development of the pituitary gland. It is a rare but impor­ tant entity to recognize in the evaluation of nasopharyngeal or midline skull base lesions, because correct diagnosis may indicate pituitary dysfunction and obviate the need for surgery, thus preventing iatro­ genic hypopituitarism or cerebrospinal fluid (CSF) leak. CPCs have previously been described as either small (<1.5-mm diameter) or large, with or without mass effect, but more recently have been classified into three types that more accurately describe size and associated patho­ logic appearances. Type 1, small incidental canals, may be found in patients with other craniofacial or neural congenital anomalies. Type 2 CPCs are medium-sized canals with ectopic pituitary tissue. Type 3 CPCs are large canals containing cephalocele (type 3A), tumors (type 3B), or both (type 3C). Type 3 large CPCs containing tumors and/or cephaloceles often contain ectopic pituitary tissue within the naso­ pharynx; thus, care should be taken during surgery to avoid iatrogenic hypopituitarism and/or CSF leak.1

Transsphenoidal Encephalocele.  Transsphenoidal encephalocele (Fig. 22-6) is thought to be due to the persistence of a craniopharyn­ geal canal. The clinical features include respiratory difficulties,  feeding difficulties, episodes of recurrent meningitis, and endocrine

abnormalities. Respiratory and feeding difficulties are related to a mass in the oral or nasal cavity. Associated congenital anomalies are seen in about a third of the patients and include hypertelorism, median nasal fissure, broad nasal root, and cleft lip or palate. The optic malforma­ tions include anophthalmia or microphthalmia, colobomas, retinal abnormalities, morning glory syndrome, and optic nerve or chiasm hypoplasia. The cerebral malformations include agenesis of the corpus callosum in up to 50% of cases, hydrocephalus, and pituitary dystopia or hypoplasia. Imaging is necessary to confirm the diagnosis as well as to define any neural or vascular elements in the sac. Noncontrast CT with threedimensional (3D) reconstruction is invaluable in reconstruction of   the skull base defect. MRI is helpful to define the contents of the  sac. Endocrine assessment is critical in every patient; hypothalamicpituitary dysfunction is often present, with deficiency of antidiuretic hormone and growth hormones being the most common finding. Various surgical approaches have been described to repair these lesions, including transcranial, transbasal, and transpalatal approaches as well as endoscopic techniques.67

Teratomas.  Teratomas are true neoplasms originating from pluripo­ tent cells and are composed of tissues from all three germ layers. Tera­ tomas occur in 1 of 4000 live births and show a female predominance, with head and neck teratomas accounting for less than 5% of the total. Head and neck teratomas are most commonly cervical, with the naso­ pharynx being the second commonest location. It is theorized that the teratoma interferes with fusion of embryonic tissues in the early devel­ opmental period and may have associated cranial anomalies like palatal fissures, hemicrania, and anencephaly. It may be sessile or peduncu­ lated and often protrudes from the mouth. The most common present­ ing symptoms are upper airway obstruction, dysphagia, and failure to gain weight. Accurate prenatal diagnosis is definitive to plan proper peripartum management. On prenatal ultrasound, polyhydramnios is often associ­ ated due to impaired swallowing. Maternal serum α-fetoprotein levels may be elevated. On CT and MRI, a teratoma can be suspected when a multiloculated lesion with focal areas of low (fat) attenuation and high signal intensity is seen on CT (Fig. 22-7) and T1-weighted MRI, respectively. Areas of bone and tooth formation, if present,  are characteristic, and these are in the form of mature tissue, unlike other bone-forming tumors. The primary differential diagnostic con­ siderations include an encephalocele or meningoencephalocele (all noncalcified and mainly cystic) and hemangioma (mainly solid with phleboliths).42

Inflammatory Lesions

Adenoidal Hypertrophy.  Adenoid is the condensation of lymphoid

FIG 22-5  Craniopharyngeal canal in a 2-year-old with nasal congestion. Sagittal reformatted CT image shows a well-corticated defect extending from sellar floor to nasopharynx (arrows). Associated soft tissue mass in nasopharynx and nasal cavity was a nasal glioma.

tissue on the posterosuperior wall of nasopharynx and part of   Waldeyer’s ring (palatine tonsils, adenoids/nasopharyngeal tonsils, and lingual tonsils). Hypertrophy of these lymphoid tissues represents immunologic activity and is not a disease process per se. Adenoidal hypertrophy occurs physiologically in children between the ages of 6 and 10 years and atrophies by age 16. Males are more commonly involved (70%) than females, with the most commonly involved age group being 16 to 25 years (60%). Failure to visualize adenoidal tissue in a young child should raise the possibility of an immune deficiency state. The common causes of adenoidal hypertrophy in adults are chronic infection and allergy. Pollution and smoking are also impor­ tant predisposing factors. Sometimes it is also associated with sinonasal malignancy, lymphoma, and human immunodeficiency virus (HIV) infection. Therefore all cases of adult adenoidal hypertrophy should be carefully evaluated to exclude underlying ominous causes.

http://radiologyebook.com

CHAPTER 22  Pharynx

A

655

B

C FIG 22-6  Transsphenoidal encephalocele in a 6-month-old who had difficulty breathing at birth. A, Sagittal FLAIR image shows a CSF-containing structure extending through skull base into nasal cavity and nasopharynx (arrows). Intracranially there is agenesis of corpus callosum. B, Sagittal postcontrast T1-weighted image shows similar findings (arrows). C, Coronal reformatted postcontrast CT image shows osseous defect in skull base (arrows) through which encephalocele extends.

Enlarged adenoids can become nearly the size of Ping-Pong balls and completely block airflow through the nasal passages. Infants present with difficulty in feeding, whereas older children present with symptoms of nasal obstruction, such as mouth breathing and snoring. Otitis media may result from encroachment on the orifice of the eusta­ chian tube. Hypertrophic adenoids appear as a homogeneous fullness of the nasopharyngeal mucosal space, which obliterates the fossae of Rosenmüller and encroaches on the nasopharyngeal airway. On CT scans, the adenoid is isodense to the underlying prevertebral muscles and may contain small cysts and calcifications (Fig. 22-8). On T1-weighted MRI scans, it is isointense to the muscles as well but can be identified by its bright T2-weighted signal. On postcontrast scans, the pharyngobasilar fascia normally enhances as a thin continuous line outlining the deep adenoid surface. Identification of this line is a reli­ able sign of the noninvasive nature of the mass, although it does not guarantee benignity. In fact it may be impossible to differentiate between hypertrophic adenoids and nasopharyngeal lymphoma on imaging alone, and a biopsy is necessary for making a definitive diag­ nosis. Although adenoid hypertrophy is a benign lesion, it shows low

apparent diffusion coefficient (ADC) values in both children and adults. This should be kept in mind to avoid possible misinterpreta­ tions as a malignant lesion.80 Adenoidectomy is indicated when hyper­ trophy results in nasopharyngeal airway obstruction, chronic otitis media, or serous middle ear effusion.

Pharyngitis.  Pharyngitis is defined as an infection or irritation of the pharynx. The etiology is usually infectious, with most cases being of viral origin and most bacterial cases attributable to group A streptococci (GAS). Other causes include allergy, trauma, toxins, and neoplasia. Imaging is usually not performed in acute and uncompli­ cated cases. Contrast-enhanced CT may demonstrate mild promi­ nence of the posterior pharyngeal soft tissues, with linear striated enhancement.

Mucous Retention Cysts.  Postpharyngitis sequelae include forma­ tion of mucous retention cysts in the inflamed obstructed mucous glands. These retention cysts are usually asymptomatic but may be large enough to cause a superficial pharyngeal mass or otitis media

http://radiologyebook.com

656

PART II  CT and MR Imaging of the Whole Body

* B

A

FIG 22-7  Teratoma (residual) in a 3-year-old who was found to have a neck mass prenatally. A, Axial postcontrast CT image shows a heterogeneous mass in the nasopharynx (short white arrows) extending into the parapharyngeal space (black arrows), with soft tissue attenuation and calcification (long white arrow). B, Axial postcontrast CT image slightly inferior to A shows the mass (arrows) also contains fat attenuation (*).

commonly results from rupture of a suppurated retropharyngeal node. On imaging, abscesses appear as low-attenuation rim-enhancing col­ lections with mass effect and flattening of the prevertebral muscles (Fig. 22-9). They tend to displace the carotid sheath laterally and the parapharyngeal space anterolaterally. Life-threatening complications such as airway obstruction and aspiration may occur following rupture of an abscess into the airway. Treatment frequently requires surgical intervention for abscess drainage and intravenous antibiotics. Aggres­ sive airway management is necessary in these patients to avoid airway obstruction and aspiration of pus. MRI may help delineate retropharyngeal abscess better and dif­ ferentiate it from cellulitis. Symmetric and smooth expansion of the retropharyngeal space by fluid density without an enhancing rim (termed retropharyngeal edema or nonabscess fluid) may be seen in early infection, internal jugular vein thrombus, postradiation states, and prevertebral calcific tendonitis. The differentiation of retropharyngeal edema from abscess on imaging is important because the former does not require surgical drainage. Retropharyngeal edema lacks an enhancing rim and is almost always confined to the level of oropharynx. Retropharyngeal abscesses may extend posteriorly to involve the danger and prevertebral spaces as well as cause osteomyelitis of the spine.73 FIG 22-8  Adenoid hypertrophy found on physical exam in 3-year-old with fever. Axial postcontrast CT image confirms prominence of adenoid tissue (arrows).

secondary to eustachian tube orifice compromise. The cyst appears as a well-defined mass usually a few centimeters in diameter but can grow to a large size. Both CT and MRI demonstrate the appearance of a typical cyst (i.e., hypodense on CT scans, dark on T1-weighted images, and bright on T2-weighted images). However, high protein content increases CT density and T1-weighted signal intensity.

Retropharyngeal Abscess/Adenitis.  The most common source of retropharyngeal abscess in children is suppurative adenitis, which is difficult to differentiate from frank abscess. A retropharyngeal abscess

Benign Neoplasms The commonest benign neoplasm of the nasopharynx is juvenile naso­ pharyngeal angiofibroma (JNA). Less frequently the nasopharynx can be a location for entities such as benign mixed tumors of salivary glands and posttransplant lymphoproliferative disorder.

Juvenile Nasopharyngeal Angiofibroma Etiopathogenesis and symptoms.  JNA is a highly vascular benign but locally aggressive tumor arising from mesenchymal tissue and is typically seen in adolescent males, with a mean age of 15 years.88 The most frequent presenting symptoms are epistaxis and nasal speech and less commonly facial deformity.23 Imaging features and spread patterns.  The exact site of origin is contentious because these masses usually present when they have

http://radiologyebook.com

CHAPTER 22  Pharynx

657

A

C

B

FIG 22-9  Suppurative retropharyngeal node vs. retropharyngeal abscess. A, An 18-month-old with fever, odynophagia, and neck swelling. Axial postcontrast CT image shows an oval mass (arrows) with a lowattenuation center with peripheral enhancement in the right retropharyngeal space, consistent with a suppurative node. At surgery, pus was aspirated from a necrotic-appearing retropharyngeal node. B, A 7-month-old on amoxicillin for otitis media with persistent fever, cough, neck swelling, and respiratory distress. Axial postcontrast CT shows an irregular rim-enhancing mass in the retropharyngeal space (black arrows) crossing the midline, compatible with a retropharyngeal abscess likely resulting from rupture of suppurative  node. There is also an abscess in the paraspinal space (white arrow). C, Sagittal reformatted postcontrast CT image of the patient in B shows extension of retropharyngeal abscess (arrows) from nasopharynx to mediastinum.

reached considerable size. However, most authors agree that it origi­ nates from the posterior choanal tissues in the region of the spheno­ palatine foramen (Fig. 22-10A). Anteriorly the tumor can extend into the nasal cavity and posteriorly through the choana into the contralat­ eral nasal cavity. The ipsilateral nasal cavity may be expanded, with displacement of the nasal septum without frank erosion obliterating the contralateral nasal cavity (see Fig. 22-10B). Superiorly it can extend into the ethmoid air cells and sphenoid sinus.15,23 Posteriorly JNA extends to the nasopharynx, with potential airway compromise. The lesion can also erode the floor of sphenoid sinus into the sinus cavity. Most commonly, JNA spreads laterally through the

pterygopalatine fossa, causing widening of this fossa with typical bowing of the posterior wall of maxillary antrum without osseous erosion. Focal erosion of the pterygoid lamina has been reported and is considered a more characteristic feature (see Fig. 22-10C). Further lateral extension can occur into the pterygomaxillary fissure, reaching the infratemporal fossa inferiorly and temporal fossa superiorly. Exten­ sion into the orbit may occur from the pterygopalatine fossa through a widened inferior orbital fissure15,23,88 (see Fig. 22-10D). Superiorly, intracranial extension occurs through two pathways: either from the orbit through the superior orbital fissure into the middle cranial fossa or by directly eroding the roof of the sphenoid

http://radiologyebook.com

658

PART II  CT and MR Imaging of the Whole Body

*

A

B

* * *

D

C

FIG 22-10  Juvenile nasopharyngeal angiofibroma. A, Axial postcontrast T1-weighted MRI shows an enhancing lesion centered in the right sphenopalatine foramen (arrow). B, Coronal reformatted postcontrast CT image shows a large enhancing mass (*) occupying entire nasal cavity, displacing the nasal septum (arrow) to the opposite side. C, Axial postcontrast CT image shows an enhancing mass anteriorly, bowing of the posterior maxillary sinus wall (black arrow), and focal erosion of the pterygoid plate (white arrow). Mass extends into the infratemporal fossa (***). D, Mass also extends into orbital apex (arrow).

sinus into the central skull base into the cavernous sinus.15,23 Inferiorly the tumor extends into the buccal space laterally and the oropharynx and central oral cavity medially. Staging.  Based on extent of spread on imaging, a grading system has been proposed (1996); grade 1 tumors are confined to the naso­ pharynx, grade 2 tumors extend into pterygopalatine fossa and masti­ cator space/infratemporal fossa, and grade 3 tumors exhibit orbital or intracranial extension. The grading system can predict prognosis and possibility of recurrence and help plan the surgical approach. Preop­ erative embolization is usually performed prior to surgery.66 Recently a new staging has been proposed for JNA (University of Pittsburgh Medical Center, 2010), which relies not only on tumor size and extent but also on the route of intracranial extension and residual vascularity after angioembolization.76 The authors report improved prediction of

intraoperative blood loss, the need for staged resections, and tumor recurrence.

Minor Salivary Gland Tumor/Benign Mixed Tumor.  Pleomor­ phic adenomas or benign mixed tumors most commonly arise from the soft palate, followed by tongue base or submucosa of the pharyn­ geal wall. Imaging features are nonspecific, and biopsy is required for diagnosis. The lesion is seen as a pedunculated homogenous mass with bright signal on T2-weighted sequences88 (Fig. 22-11).

Posttransplant Lymphoproliferative Disorder.  Posttransplant lymphoproliferative disorder (PTLD) is a serious complication of organ transplantation. Although the precise etiology is unknown, the Epstein-Barr virus and immunosuppressive agents appear to be risk

http://radiologyebook.com

CHAPTER 22  Pharynx

659

B

A

C FIG 22-11  Benign mixed tumor in an 83-year-old presenting with neck pain. A, Axial T1-weighted precontrast image shows a mass (arrows) nearly isointense to muscle centered in the left nasopharynx, with extension into the parapharyngeal space. B, The mass is markedly hyperintense on axial T2-weighted image. C, There is prominent heterogeneous enhancement on axial T1-weighted postcontrast image.

factors. The first manifestations of PTLD are frequently observed in the head and neck area, with adenoidal and/or tonsillar enlargement (Fig. 22-12). PTLD remains an elusive disease, with diversity in its clinical presentation, course, and histology all contributing to the dif­ ficulty in diagnosing and treating these patients. The incidence after bone marrow or organ transplantation has been reported from 1% to 10% but varies based on the type of transplantation. Adenotonsillar PTLD is also more prevalent in younger patients, females, and liver and lung transplant recipients. A lower rate is reported in other immu­ nocompromised patients, such as renal transplant and bone marrow recipients. Many of these cases are diagnosed when the child undergoes excision for adenotonsillar hypertrophy. It has been found that Waldeyer’s ring involvement appears to occur in both adults and chil­ dren. Treatment continues to be controversial but is generally designed to accommodate the specific patient. Typically, although aggressive non-Hodgkin’s B-cell lymphoma is treated with chemotherapy, PTLD often can be treated with a major reduction of immunosuppressive medications. Surgical resection generally is attempted for limited disease.43,82

Malignant Neoplasms The commonest malignant tumor of the nasopharynx is nasopharyn­ geal carcinoma, which is a type of squamous cell carcinoma that com­ prises 70% of all neoplasms, followed by lymphomas that account for 20%.5 Minor salivary gland tumors, adenocarcinomas, rhabdomyosar­ comas, extramedullary plasmacytoma, and melanomas are other less common neoplasms.

Nasopharyngeal Carcinoma Etiopathogenesis and symptoms.  The highest incidence of naso­ pharyngeal carcinoma (NPC) is in southwest China, followed by  Singapore and Hong Kong; Eskimoan and Mediterranean populations are also disproportionately affected.15 NPC has a seven times higher incidence in Americans of Chinese ancestry than in non–Chinese Americans. A higher incidence is also found in African Americans. The etiology appears to be the result of genetic susceptibility combined with environmental factors, some of which include consumption of nitrosamines in dry salted fish and chronic nasal infections with Epstein-Barr virus.57,88 The male-to-female incidence is 3 : 1, the peak

http://radiologyebook.com

660

PART II  CT and MR Imaging of the Whole Body

FIG 22-12  Posttransplant lymphoproliferative disorder (PTLD) in 9-year-old with history of a renal transplant and 6 weeks of coughing and stridor. There is diffuse thickening of hypopharyngeal and subglottic soft tissues on axial postcontrast CT images. Biopsy revealed PTLD. Involvement of the larynx is much less common than adenotonsillar involvement.

age being 50 to 60 years, with a smaller peak in late childhood. NPC frequently presents with cervical metastatic nodes (in >70%) despite early small primary lesions, the other symptoms being otitis media (due to eustachian tube obstruction), sore throat, epistaxis, headache, and cranial nerve neuropathies in advanced cases.15,23 Staging and treatment principles.  The criteria given in the seventh edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual for tumor-node-metastasis (TNM) staging of NPC are provided in Box 22-1. Surgery is unable to achieve clear resec­ tion margins, and radiotherapy (RT) remains the mainstay of treat­ ment. Early-stage disease (T1 tumors without nodal metastases) treated with RT have a 5-year survival of 75% to 90%.28 In more advanced NPC, concurrent chemoradiation is the standard of care.47 Recently, transoral robotic surgery (TORS) has been used for local recurrence of NPC. Spread patterns.  NPC usually originates in the lateral pharyngeal recess (Fig. 22-13A) and spreads mucosally, submucosally, and beyond the confines of the nasopharynx as well. Spread to the torus tubarius results in obstruction of the eustachian tube with resultant serous otitis media. Spread to the nasal cavity anteriorly or inferiorly into orophar­ ynx is classified as T1 disease. More commonly, early disease spread is seen posterolaterally into the parapharyngeal space (T2 disease) through the sinus of Morgagni. Further posterolateral spread into the carotid space may occur, with invasion of CNs IX, X, XI, and XII, with resultant palsies (T4 disease).16 On axial images, hypoglossal nerve palsy results in fatty replacement of the affected side of the tongue along with prolapse of the hemi­ tongue into the oropharynx, indicating hypotonia.13 Posterior spread to the retropharyngeal space, prevertebral muscles, and vertebrae may occur and is associated with increased incidence of cervical nodal metastases.17 Superior extension into the sphenoid sinus and skull base upstages to T3 disease (see Fig. 22-13B). Cranial spread can also occur directly

through the foramen ovale (see Fig. 22-13C) or foramen lacerum (located immediately superior to the lateral pharyngeal recess) into the middle cranial fossa or through the jugular foramen into the posterior cranial fossa, upstaging the tumor to T4 disease.17 Perineural spread along the mandibular nerve through the foramen ovale (see Fig. 22-13D) can cause denervation atrophy of the muscles of mastica­ tion.64 Intracranial spread may extend to involve the dura mater, cav­ ernous sinus, Meckel’s cave (see Fig. 22-13D) and prepontine cistern. Lateral extension to the masticator space or infratemporal fossa is also T4 disease (see Fig. 22-13D) and can cause trismus.17 Anteriorly, lateral spread through the sphenopalatine foramen extends into the pterygopalatine fossa, from where tumor can spread (a) along the maxillary nerve (Fig. 22-14A) through the foramen rotundum to the intracranial cavity, (b) through the inferior orbital fissure into the orbital apex (T4 disease), (c) through the vidian canal into the carotid canal, (d) through the superior orbital fissure into the cranial cavity, and (e) through the pterygomaxillary fissure into the infratemporal fossa.12 The primary echelon nodes in NPCs are level IIb and lateral retro­ pharyngeal group (see Fig. 22-14B), the other nodal levels being III, IV, and V.17 Spread to supraclavicular nodes is regarded as N3b (see Box 22-1). Pretreatment imaging features.  On CT and MRI, NPC shows nonspecific imaging features and cannot be differentiated with cer­ tainty from lymphoma or minor salivary gland tumors. Hence the primary role of pretreatment imaging is complete disease staging  to help plan optimal treatment. CT and MRI are complementary  in evaluating the locoregional extent of NPC, but MRI is the tech­ nique of choice because it is more sensitive for depicting skull base invasion, perineural spread, invasion of dura and cavernous sinuses, and retropharyngeal adenopathy.36 Cortical erosion is best studied on CT.59 Asymmetry of the nasopharynx is an early feature, but this may also be seen in asymmetric lymphoid hyperplasia; obliteration of the fat stripe between the tensor and levator veli palatini is a more reliable early imaging feature of NPC. The above features should prompt atten­ tion to serous mastoiditis, which suggests eustachian tube dysfunction, and lymphadenopathy in the ipsilateral neck, particularly in the retro­ pharyngeal group.88 In more advanced cases, imaging reveals invasion of adjacent spaces. Noncontrast and non–fat-suppressed T1-weighted sequences are most useful to demonstrate obliteration of the fat intensity between the adjacent space and tumor (which is isointense). NPC displays intermediate signal on T2-weighted sequences, with intense enhancement on postgadolinium images. Fat-suppressed contrastenhanced images, particularly in the coronal plane, are valuable for confirming skull base marrow invasion (see Fig. 22-13B), and for demonstrating orbital extension, perineural spread (see Figs. 22-13D and 14A), and dural invasion. Thickening and enhancement of dura mater alone, however, could either represent intracranial extension of tumor or reactive inflammation (see Fig. 22-13C).16 Foraminal invasion due to perineural spread is seen as widening and increased enhancement (see Fig. 22-13D). Denervation atrophy due to perineural spread is seen as hyperintense signal in the mastica­ tor muscles on T2-weighted images, with increased enhancement in the acute-subacute stage, and as diffuse muscle hyperintensity on both T1- and T2-weighted images, with reduced bulk in the chronic phase. Direct tumor invasion increases the muscle bulk, is focal rather than diffuse, and has signal intensity similar to tumor.64 NPC and nodal metastases are usually FDG avid. Currently FDG-PET/CT is considered useful for mapping precise nodal burden, for evaluating distant metastases, and in the workup of unknown

http://radiologyebook.com

CHAPTER 22  Pharynx

661

BOX 22-1  AJCC 7Th Edition T & N Staging of Primary Tumors of Nasopharynx Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of primary tumor Tis Carcinoma in situ Nasopharynx T1 Tumor confined to the nasopharynx, or tumor extends to oropharynx and/or nasal cavity without parapharyngeal extension* T2 Tumor with parapharyngeal extension* T3 Tumor involves bony structures of skull base and/or paranasal sinuses T4 Tumor with intracranial extension and/or involvement of cranial nerves, hypopharynx, orbit or with extension to infratemporal fossa/masticator space Oropharynx T1 Tumor 2 cm or less in greatest dimension T2 Tumor more than 2 cm but not more than 4 cm in greatest dimension T3 Tumor more than 4 cm in greatest dimension or extension to lingual surface of epiglottis T4a Moderately advanced local disease Tumor invades the larynx, extrinsic muscle of tongue, medial pterygoid, hard palate, or mandible^ T4b Very advanced local disease Tumor invades lateral pterygoid muscle, pterygoid plates, lateral nasopharynx, or skull base or encases carotid artery Regional Lymph Nodes (N) Nasopharynx The distribution and the prognostic impact of regional lymph node spread from nasopharyngeal cancer, particularly of the undifferentiated type, are different from those of other head and neck cancers and justify the use of a different N classification scheme.

NX—Regional lymph nodes cannot be assessed N0—No regional lymph node metastasis N1—Unilateral metastasis in lymph node(s), 6 cm or less in greatest dimension, above the supraclavicular fossa, and/or unilateral or bilateral, retropharyngeal lymph nodes, 6 cm or less, in greatest dimension.+ N2—Bilateral metastasis in cervical lymph node(s), 6 cm or less in greatest dimension, above the supraclavicular fossa+ N3—Metastasis in a lymph node(s)+ > 6 cm and/or to supraclavicular fossa+ N3a—Greater than 6 cm in dimension N3b—Extension to the supraclavicular fossa++ Oropharynx^^ NX—Regional lymph nodes cannot be assessed N0—No regional lymph node metastasis N1—Metastasis in a single ipsilateral lymph node, 3 cm or less in greatest dimension N2—Metastasis in a single ipsilateral lymph node, more than 3 cm but not more than 6 cm in greatest dimension, or in multiple ipsilateral lymph nodes, none more than 6 cm in greatest dimension, or in bilateral or contralateral lymph nodes, none more than 6 cm in greatest dimension N2a—Metastasis in a single ipsilateral lymph node, more than 3 cm but not more than 6 cm in greatest dimension N2b—Metastasis in multiple ipsilateral lymph nodes, none more than 6 cm in greatest dimension N2c—Metastasis in bilateral or contralateral lymph nodes, none more than 6 cm in greatest dimension N3—Metastasis in a lymph node more than 6 cm in greatest dimension

From Edge S, Byrd DR, Compton CC, et al (eds.): AJCC Cancer Staging Manual, ed 7, New York, 2010, Springer-Verlag. *Note: Parapharyngeal extension denotes posterolateral extension of tumor. ^ Note: Mucosal extension to lingual surface of epiglottis from primary tumors of base tongue or vallecula does not constitute invasion of larynx. + Midline nodes are considered ipsilateral nodes. ++ Supraclavicular zone or fossa is relevant to the staging of nasopharyngeal carcinoma and is the triangular region originally described by Ho. It is defined by three points: (1) the superior margin of the sternal end of the clavicle, (2) the superior margin of the lateral end of the clavicle, (3) the point where the neck meets the shoulder. Note that this would include caudal portions of levels IV and VB. All cases with lymph nodes (whole or part) in the fossa are considered N3b. ^^ Note: Metastases at level VII are considered regional lymph node metastases.

primary cancers presenting with cervical adenopathy.61,90 National Comprehensive Cancer Network (NCCN) practice guidelines recom­ mend PET-CT to evaluate for distant metastases when stage III or IV disease is present. Imaging is also valuable for treatment planning. Accurate delineation of target volume with MRI as well as PET/CT results in more precise delivery of RT and indirectly improves local control and disease-specific survival.36 Imaging can also provide prog­ nostic indicators. The most important prognostic criterion predictive of outcome and survival in NPC was CT-derived primary tumor volume greater than 50 mL during RT planning.8 Response assessment and surveillance.  The NCCN guidelines recommend a baseline scan to be obtained within 6 months of comple­ tion of chemoradiation in patients with T3-4 and/or N2-3 disease. Further imaging is only recommended “as indicated based on signs/ symptoms.”65 Response assessment can be performed either with MRI or PET/ CT, with equivalent reported accuracy for detection and restaging.19 CT has inferior soft tissue resolution to differentiate between posttreat­ ment changes and residual or recurrent disease. A baseline posttreat­ ment MRI scan is usually performed at 2 to 3 months after therapy because it is perceived to be valuable for detecting subclinical residual

adenopathy and also serves as a baseline to detect future, more subtle recurrences. If the pretreatment study showed infiltration of the para­ pharyngeal space, pterygopalatine fossa, orbits, or the skull base, abnormal signal intensity may persist. The initial scan may also show significant posttreatment edema of the pharyngeal mucosa. There should be no residual enlarged lymph nodes, however; if these are found on the new baseline scan, a neck dissection is typically per­ formed. The timing of PET/CT is critical, with a systematic review and meta-analysis reporting a high negative predictive value of 95% in scans done after 12 or more weeks.32 Earlier response assessment to chemoradiotherapy at 20 days and 50 days has been evaluated with diffusion-weighted MRI with promising results.9 High pretreatment ADC values and significant increase in ADC following initiation of induction chemotherapy showed better response to definitive treat­ ment with concurrent chemoradiation. Use of plasma Epstein-Barr virus (EBV) DNA for initial assessment, followed by PET/CT or MRI in positive cases, has been recommended. Wang et al. found plasma EBV DNA to have 100% accuracy in detect­ ing recurrence of NPC.83 A recent study also found that patients with a negative 3-month PET/CT had limited benefit from subsequent PET surveillance.38

http://radiologyebook.com

662

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 22-13  Nasopharyngeal carcinoma. A, Axial T1-weighted postcontrast MRI shows a small enhancing lesion centered in the left lateral pharyngeal recess (arrow). B, Coronal T1-weighted postcontrast MRI shows an enhancing mass invading skull base marrow (arrow). C, Coronal T1-weighted postcontrast MRI shows an enhancing mass invading the foramen ovale (arrow), with intracranial extension (T4 tumor) causing dural thickening. D, Coronal T1-weighted postcontrast MRI shows perineural spread through a widened foramen ovale into Meckel’s cave (long arrow). Short arrow on the contralateral side shows the normal Meckel’s cave. Dashed arrow shows extension to masticator space (T4 tumor).

On MRI, nonenhancing tissue with dark signal on T2-weignted sequences represents mature fibrosis. However, immature fibrosis (being cellular) may show intermediate to high T2-weighted signal intensity and may enhance, mimicking residual/recurrent disease.60 Following RT the nasopharynx may appear asymmetric and tends  to lose its normal contours and develops a fibrosed appearance with effacement of the lateral pharyngeal recess primary site (Fig. 22-15AB). Recurrent/residual tumor has bulging contours with enhancement (see Fig. 22-15C) that can be detected with certainty on PET/CT as an FDG-avid lesion (see Fig. 22-15D). Regression on serial scans may indicate post-RT changes, but unchanged lesions remain suspicious for occult or recurrent disease. Progressing masses or the appearance of a new node suggest recurrence.11 Another uncommon manifestation of

recurrence is meningeal infiltration in the posterior fossa through the jugular foramen or foramen magnum without any visible nasopharyn­ geal component, which should prompt meticulous scrutiny of these regions.14

Nasopharyngeal Lymphoma.  The head and neck is the second most common site of non-Hodgkin’s lymphoma (NHL). NHL is most frequently found in Waldeyer’s ring. Within Waldeyer’s ring the naso­ pharynx is the second most common site of disease after the tonsil. It is a homogeneous tumor that tends to diffusely involve all walls of the nasopharynx and spread in an exophytic fashion to fill the airway, rather than infiltrating into the deep tissues (Fig. 22-16). Primary NHL more commonly spreads superficially to involve the nasal cavity or

http://radiologyebook.com

CHAPTER 22  Pharynx

663

B

A

FIG 22-14  Nasopharyngeal carcinoma. A, Axial T1-weighted postcontrast MRI shows enhancement along pterygopalatine fossa, suggestive of perineural spread along V2 (maxillary nerve). B, Axial T2-weighted MRI shows a rounded retropharyngeal node (arrow) located medial to the carotid vessels.

oropharynx; lymphadenopathy is frequent and extensive, and the nodes may show necrosis and matting. A large tumor that fills the nasopharynx, with no or minimal inva­ sion into deep structures and a propensity to extend down into the tonsil rather than up into the skull base, may suggest the diagnosis of NHL over nasopharyngeal carcinoma.44

Nasopharyngeal Rhabdomyosarcoma.  Rhabdomyosarcoma (RMS) is the most common soft tissue tumor in children, with the head and neck region accounting for 35% to 40% of cases. It is a highly malignant tumor thought to arise from striated muscle progenitor cells, with four major histologic subtypes. Embryonal RMS is the most common subtype in the head and neck region, has an inter­ mediate prognosis, and rarely involves regional lymph nodes. Alveo­ lar RMS and undifferentiated sarcoma are both associated with a relatively poorer prognosis. Nasopharyngeal RMS presents with indolent minimal and nonspecific symptoms of nasal, aural, or  sinus obstruction, with or without a mucopurulent or sanguineous discharge. Imaging is critical in defining the exact size and location of the lesion and offering a differential diagnosis. It makes an essential con­ tribution to staging and detecting extension into the intracranial  compartment, involvement of cranial nerves, osseous erosion of the  skull base, orbital involvement, encasement of the carotid or vertebral arteries, perineural spread, lymphadenopathy, and distant metastases (Fig. 22-17). Imaging also helps guide a biopsy procedure by showing the best route to obtain representative material. Treatment is usually multimodal, with chemotherapy, surgery, and radiation therapy.26,34

Miscellaneous

Inflammatory Pseudotumor of the Nasopharynx.  Inflammatory pseudotumor of the nasopharynx is a rare benign idiopathic disease often mistaken for a neoplasm or infection, owing to its aggressive behavior and clinical presentation. It can present as a progressively

destructive skull base mass and should be considered when repeated tissue biopsies reveal acute or chronic inflammation without evidence of malignant disease or infection. Recognition of inflammatory pseu­ dotumor is critical because it usually mimics squamous cell carcinoma or an aggressive angioinvasive fungal infection on imaging. Much like other inflammatory or neoplastic processes of the nasopharynx, it can present with unilateral hearing loss or otitis media owing to eustachian tube obstruction, multiple cranial nerve palsies resulting from skull base or cavernous sinus invasion, and stroke resulting from compro­ mise of intracranial vasculature (primarily internal carotid arteries at or below the skull base). On imaging it is seen as an ill-defined infiltrative soft tissue lesion with inflammatory fat stranding on CT or MRI. Inflammatory pseu­ dotumor is a diagnosis of exclusion, with differential considerations including inflammatory myofibroblastic tumor, immunoglobulin  (Ig)G4-related sclerosing lesion, Rosai-Dorfman disease, EBV-related inflammatory pseudotumor, calcifying fibrous pseudotumor, lym­ phoma, and NPC. Because of their aggressive presentation, tissue sam­ pling is essential for making the diagnosis, which usually shows only inflammatory changes. An important diagnostic clue is the absence of cervical lymphadenopathy despite the presence of an aggressive destructive mass. Despite their aggressive appearance, these lesions respond well to steroid treatment.18

OROPHARYNX Anatomy The oropharynx is the part of the pharynx that lies behind the oral cavity (can be seen through the open mouth). It is separated from the oral cavity by the circumvallate papillae of the tongue, the anterior tonsillar pillars, and the soft palate. The base of tongue is the posterior third of the tongue, bounded anteriorly by the circumvallate papillae, laterally by the glossotonsillar sulci, and posteriorly by the epiglottis. The vallecula is a strip of mucosa that is the transition from the base

http://radiologyebook.com

664

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 22-15  Posttreatment MRI in nasopharyngeal carcinoma. Axial (A) and coronal (B) T1-weighted contrastenhanced MRI in two different patients shows mild asymmetry and enhancement in the lateral pharyngeal recess (arrows). These are posttreatment changes, and follow-up showed complete response. C, Contrastenhanced axial CT shows residual heterogeneously enhancing lesion along the left lateral nasopharyngeal wall (arrow) and parapharyngeal space, suspicious for recurrence. D, PET/CT shows FDG-avid mass (arrow) suggestive of residual disease.

http://radiologyebook.com

CHAPTER 22  Pharynx

*

A

B

*

C

D FIG 22-16  Nasopharyngeal lymphoma in a 63-year-old male with 5-month history of nasal congestion progressing to nasal obstruction. A, Axial postcontrast CT image shows mass filling the nasopharynx, with extension into the parapharyngeal spaces (long black arrows), carotid spaces (long white arrows), and retropharyngeal and prevertebral spaces (short black arrows). B, Axial CT image more inferiorly shows extension of the mass into the oropharynx (arrows) and numerous enlarged lymph nodes (*). C, The mass (*) is associated with diffuse infiltration into clivus, basisphenoid, and sphenoid sinus (arrows), as seen on sagittal precontrast T1-weighted image. D, The mass is only mildly hyperintense on axial T2-weighted image (long arrows) with deep extension into carotid spaces visualized (short arrows).

http://radiologyebook.com

*

665

666

PART II  CT and MR Imaging of the Whole Body

A

B

*

* C

*

D

E FIG 22-17  Nasopharyngeal rhabdomyosarcoma in an 8-year-old with 1-week history of headache, left eye pain, ptosis, and anisocoria. A, On axial postcontrast CT there is a hypodense rim-enhancing mass (arrows) centered in the left nasopharynx, with extension into parapharyngeal space. B, The mass (arrows) is isointense to muscle on axial T1-weighted precontrast image. Infiltration into the left carotid space is present (arrowhead). C, The mass is hyperintense on axial T2-weighted image (arrows), with carotid space extension (arrowhead) of similar signal. D, There is enhancement following contrast on fat-saturated axial T1-weighted image. E, Coronal postcontrast T1-weighted image shows extension of the mass (long black arrows) through the skull base in region of the foramen ovale into the cavernous sinus (short black arrows). There is dural enhancement along the floor of the left middle cranial fossa (arrowhead). Normal foramen ovale is seen on the right (white arrow).

http://radiologyebook.com

CHAPTER 22  Pharynx

667

* V

A

P

* V

*

V

H

C

B

FIG 22-18  Normal oropharynx on CT and MRI. A, Axial postcontrast CT image shows the vallecula (V) between the base of tongue (*) and epiglottis. B, The same structures are seen on axial postcontrast T1-weighted image. C, Sagittal T1-weighted MRI without contrast shows the palate (P), hyoid (H), tongue base (*), and vallecula (V), the latter being just anterior to the epiglottis.

of the tongue to the epiglottis and is considered part of the oropharynx (Fig. 22-18). Thus the oropharynx includes the base of the tongue (posterior third of the tongue and the lingual tonsil), the vallecula, the soft palate and uvula, the lateral pharyngeal walls including the palatine tonsils and tonsillar pillars, and the posterior pharyngeal wall, extending from the plane of the soft palate/hard palate junction to the level of the pharyngoepiglottic folds at the hyoid bone. Posteriorly the superior and middle constrictor muscles separate the oropharynx from the prevertebral muscles overlying the  second and third cervical vertebrae. Superiorly the oropharynx  communicates with the nasopharynx, separated by the soft palate. Inferiorly the oropharynx is separated from the hypopharynx by the pharyngoepiglottic fold and from the larynx by the epiglottis and glossoepiglottic fold.33,56,62

The features of the lateral oropharyngeal walls include two tonsillar pillars (faucial arches): • The anterior tonsillar pillar (palatoglossus muscle) • The posterior tonsillar pillar (palatopharyngeus muscle) Between them lies the tonsillar fossa, which contains the palatine tonsil on each side.

Congenital Lesions

Lingual Thyroid.  Ectopic thyroid tissue at the base of the tongue/ floor of the mouth is referred to as lingual thyroid. Embryologic descent of thyroid anlage occurs between the third and seventh week of gesta­ tion from the foramen caecum, situated at the base of the tongue, to its usual position in the neck. Complete failure of descent results in a lingual thyroid. This may occur in association with thyroglossal duct cyst and is seen more commonly in females.58 This abnormality is

http://radiologyebook.com

668

PART II  CT and MR Imaging of the Whole Body

A

B FIG 22-19  Lingual thyroid found incidentally. A, Axial postcontrast CT image shows round hyperdense mass (arrow) in midline of tongue base in expected location of foramen cecum. B, The lingual thyroid (arrow) is well seen on sagittal reformatted CT image.

usually incidentally detected, but some patients may develop dyspha­ gia, dyspnea, or obstructive sleep apnea. On imaging, these lesions appear as well-circumscribed round or oval lesions at the floor of the mouth/base of the tongue in the midline, with features similar to thyroid tissue. Characteristically the lesions are hyperdense in attenu­ ation on noncontrast CT and demonstrate avid enhancement (Fig. 22-19). On MRI, the lesion is usually isointense to skeletal muscle on T1-weighted and hyperintense on T2-weighted images. The lesion demonstrates high uptake on Tc-99m pertechnetate or radioiodine nuclear scan.70 Surgery may be performed if the patient has obstructive symptoms.

Thyroglossal Duct Cyst.  A thyroglossal duct cyst (TGDC) is the most common congenital neck mass, accounting for 70% of all con­ genital neck anomalies.31,45,46 Embryologically the thyroglossal duct (which means pertaining to the thyroid gland and the tongue) extends from the foramen cecum in the base of tongue to the thyroid bed through the hyoid bone. Any portion of this duct that fails to involute results in a TGDC. It is located in the midline (75%) or slightly paramidline (25%), embedded in the infrahyoid strap muscles. The most common location of a TGDC is infrahyoid (65%), whereas only 15% are at the level of the hyoid bone, and 20% are suprahyoid in the location in the anterior neck. The more inferior the cyst, the more likely it is to be off the midline. About 50% of patients present before 20 years of age, though there is a small per­ centage (15%) who present at older than 50 years, with no reported gender predilection. The most common clinical presentation is that of a painless palpable cystic neck mass near the hyoid bone in the midline. Approximately one third present with a concurrent or prior infection. As a result of their intimate relationship with the foramen cecum and hyoid bone, the TGDC usually moves with swallowing or protrusion of the tongue on physical exam.2,46,84 The uncomplicated TGDC can be anechoic, homogeneously hypoechoic, or heteroge­ neous in echogenicity on ultrasound, without internal vascularity on

color Doppler ultrasound. The uncomplicated TGDC is well defined with low CT attenuation, has fluid signal intensity on MRI, and lacks a solid enhancing component. The thin cyst wall may demonstrate peripheral contrast enhancement (Fig. 22-20). A complicated TGDC with superimposed infection may have peripheral contrast enhance­ ment, internal septations, and may be associated with edema and stranding of the surrounding soft tissues. Ectopic thyroid tissue may be present in the TGDC and can be a site of thyroid carcinoma, 80% of which are of papillary cell origin. The surgical resection of a symp­ tomatic TGDC is Sistrunk’s procedure, which involves resection of the thyroglossal remnant together with a central portion of hyoid bone and cuff of tissue around the thyroglossal tract from the hyoid bone to the foramen cecum.2

Congenital Cysts.  Congenital cysts of the oropharynx have been described in the medical literature by various names such as tongue base cysts, epiglottic cysts, vallecular cysts, or lingual cysts.85 Histologi­ cally the cyst contains mucous glands with an external lining of squa­ mous epithelium.39 These cysts are usually detected incidentally on imaging studies. Infrequently these can be seen in children who present with airway obstruction. Diagnosis is made on flexible laryngoscopy or bronchoscopy. CT and MRI are useful to narrow the differential diagnosis. Typical imaging findings are a sharply demarcated cystic lesion that does not enhance. Surgical treatment involves marsupializa­ tion under anesthesia.

Lymphatic/Venolymphatic Malformation.  A lymphatic malfor­ mation is a developmental lesion composed of dilated lymphatic chan­ nels. These are caused by rests of lymph sacs left during embryogenesis. A venolymphatic malformation has combined elements of venous and lymphatic malformations. Most lesions are found in the head and neck region. Approximately 50% of these lesions are found at birth and  80% to 90% are detected within the first 2 decades.89 These are benign lesions and do not have malignant potential. Depending on the size of

http://radiologyebook.com

CHAPTER 22  Pharynx

A

669

B FIG 22-20  Thyroglossal duct cyst in 27-year-old with neck pain. A, Axial postcontrast CT image shows lowattenuation mass with thin enhancing rim (arrows) in midline of infrahyoid neck, embedded within strap muscles. B, The cyst (arrows) is seen just below hyoid on sagittal reformatted CT image.

A

B FIG 22-21  Lymphatic malformation in 4-year-old with neck mass since birth. A, Axial postcontrast CT image shows multispatial heterogeneous, predominantly cystic neck mass (long arrows) with extension to oropharynx (short arrows). B, Coronal reformatted CT image shows extension to oropharynx (arrows).

the cysts, these lesions can be classified as microcystic or macrocystic. These can be located in a single neck space or can be transspatial (Fig. 22-21A). On imaging these lesions appear as uniloculated or multi­ loculated cystic neck masses with imperceptible walls. Fluid-fluid levels may be seen. Typically these lesions do not demonstrate enhancement or may have minimal peripheral enhancement (see Fig. 22-21B). Enhancement within a lesion that otherwise resembles a lymphatic malformation suggests it is a mixed lesion with a component of venous malformation. These lesions can rapidly enlarge in size if there is an infection or intralesional hemorrhage and cause symptoms, predomi­ nantly due to locoregional mass effect on surrounding structures. 

Macrocystic lymphatic malformations have a low tendency to recur, whereas microcystic malformations have a high recurrence rate. Treat­ ment options include percutaneous sclerotherapy or surgery. Percuta­ neous sclerotherapy has shown good results both as primary treatment and also treatment of lesions that recur after surgery.75

Macroglossia.  Macroglossia refers to enlargement of the tongue. It is usually seen in children and is associated with conditions like Down syndrome, Beckwith-Wiedemann syndrome, congenital hypothyroid­ ism, and mucopolysaccharidoses. The tongue can also appear large  in children with micrognathia, such as in cases of Pierre Robin 

http://radiologyebook.com

670

PART II  CT and MR Imaging of the Whole Body

syndrome.25 Rarely macroglossia can be seen in adults in cases of lym­ phoma and amyloidosis. In some cases, macroglossia may result in dysphagia or airway compromise and require surgical correction.53

Oropharyngeal Airway Compromise (Obstructive Sleep Apnea and Pickwickian Syndrome).  Obstructive sleep apnea (OSA) is the most common type of sleep apnea. It is caused by upper airway obstruction during sleep. Common signs and symptoms include loud snoring, excessive daytime sleepiness and fatigue, morning headaches, and insomnia. OSA has been implicated to cause sleep fragmentation and hypoxemia, which in turn contribute to various adverse effects,4 which include increased risk of motor vehicle accidents and adverse cardiovascular risk. These patients are also at an increased risk for stroke and hypertension.86 Most cases of OSA result from narrowing of the posterior nasopharynx or retroglossal airways. Enlarged lingual or palatine tonsils and adenoids contribute to airway narrowing.  Glossoptosis, defined as posterior motion of the tongue during sleep, is also implicated because it causes narrowing of the airway. Glossoptosis can occur in cases of macroglossia or decreased muscular tone. Hypo­ pharyngeal collapse can also occur during sleep in patients with decreased muscular tone. Cine-MR studies have been used to docu­ ment the dynamic change in tongue position and the configuration of the hypopharynx during sleep. Pickwickian syndrome (also known as obesity hypoventilation syndrome) is named after a character Charles Dickens originally described in his novel The Pickwick Papers. It refers to a condition of hypoventilation, hypoxemia, somnolence, and hyper­ capnia in obese people.

Inflammatory Lesions

FIG 22-22  Peritonsillar abscess in 25-year-old with 3 days of left throat pain and odynophagia. There is enlargement of left palatine tonsil containing a low-attenuation rim-enhancing collection.

Tonsillitis and Peritonsillar Abscess.  Tonsillitis and peritonsillar abscess are the most commonly encountered acute neck infections among adolescents and young adults, with peritonsillar abscess accounting for one third of all soft tissue abscesses of the head and neck.7,71 Symptoms include severe unilateral sore throat, fever, tender cervical lymphadenopathy, dysphagia, pharyngotonsillar exudates, otalgia. and trismus. The most common pathogens are β-hemolytic Streptococcus, Staphylococcus aureus, Pneumococcus, and Haemophilus influenzae. Acute tonsillitis may suppurate and internally cavitate to create an intratonsillar abscess; however, a true tonsillar abscess is rare. Infection that penetrates the fibrous tonsillar capsule and the periton­ sillar space—a potential space between the tonsillar capsule and the superior constrictor muscle—is more common. The infection may then extend into the adjacent parapharyngeal, masticator, or subman­ dibular spaces. The resulting peritonsillar cellulitis resolves over several days when antibiotics are administered; if it goes untreated, a periton­ sillar abscess develops, typically along the superior tonsillar pole. Imaging is not routinely performed if the diagnosis is clinically apparent.35 However, contrast-enhanced CT is used if the diagnosis is uncertain, a full clinical examination is difficult because of severe trismus, a deep neck space infection or complication is suspected, or the patient does not respond to therapy. CT findings of peritonsillar cellulitis include tonsillar enlargement and linear striated enhance­ ment of the palatine tonsils and posterior pharyngeal soft tissues. Medial apposition of the enlarged tonsils results in a “kissing tonsils” appearance. Central liquefaction with surrounding rimlike enhance­ ment is diagnostic of peritonsillar abscess (Fig. 22-22). Peritonsillar cellulitis is treated with antibiotics, whereas abscess requires needle aspiration or surgical drainage.54,71 When a peritonsillar abscess is clinically suspected, the radiologist’s responsibilities are the following: • To confirm the diagnosis • To determine the degree of airway compromise

• To determine whether the abscess has spread into the deep neck spaces • To assess the status of the internal carotid artery and internal jugular vein in the case of a deeply extending abscess

Postinflammatory Sequelae.  Occasionally a second branchial cleft cyst drains into the tonsillar fossa, and this should be considered when a fistula there is identified. Alternatively, actinomycosis or dermal sinus tracts could cause peritonsillar or tonsillar fistulae. After tonsillitis, dystrophic calcification or clumps of tonsillar calcification called tonsilloliths may be detected incidentally on CT scans; these are indicative of previous or chronic tonsillitis. Less commonly they may be seen in the adenoids or the lingual tonsil.

Benign Neoplasms Benign tumors are rare and include benign mixed tumor, schwannoma, hemangioma, myofibroma, and amyloidoma. These have nonspecific imaging features, and biopsy is required for diagnosis.

Malignant Neoplasms The commonest neoplasm of the oropharynx (90%) is squamous cell carcinoma (SCC).74 Lymphomas, minor salivary gland tumors, and sarcomas comprise the remainder of malignant neoplasms. Imaging cannot reliably differentiate these neoplasms from SCC.

Oropharyngeal Squamous Cell Carcinoma Etiopathogenesis and symptoms.  Oropharyngeal SCC (OPSCC) has two etiologic subtypes—one related to alcohol and tobacco use and another less aggressive variety related to human papillomavirus (HPV) infection, which has a better prognosis.20 The presenting symptoms are nonspecific, such as sore throat, dysphagia, and otalgia in the vast majority of patients, with pain and trismus in more invasive tumors.74

http://radiologyebook.com

CHAPTER 22  Pharynx Staging and principles of treatment.  The seventh edition AJCC Cancer Staging Manual criteria for TNM staging of OPSCC are pro­ vided in (see Box 22-1. OPSCC is treated with RT/surgery or a com­ bination of both. Although TORS may have a role for early-stage oropharyngeal cancers, stage III and IV disease are usually treated with concurrent chemoradiation.24 In view of better prognosis of HPVpositive cancers, less intensive treatment regimens (deescalation treat­ ment protocols) are currently being investigated.52

671

Spread patterns.  OPSCC can arise from the tonsil, tongue base, soft palate, or posterior pharyngeal wall. Tonsillar SCC mostly arises from the anterior tonsillar pillar (ATP) and tonsil, lesions from the posterior tonsillar pillar being rare. Anterior spread is along the pala­ toglossus muscle to the base of the tongue inferiorly (Fig. 22-23A-C) and soft palate superiorly. Lateral spread is common, causing oblitera­ tion of the parapharyngeal fat. Superior and inferior spread occurs along the nasopharyngeal and hypopharyngeal wall. Spread along the

A

B

C

D FIG 22-23  Oropharyngeal carcinoma. A, Axial T2-weighted MRI shows bilateral prominent palatine tonsils that appear hyperintense (arrows). B, Axial T2-weighted MRI shows a bulky mass involving left lateral pharyngeal wall and palatine tonsil (solid arrow). Dashed arrow shows contralateral normal palatine tonsil. C, Axial contrast-enhanced MRI shows an enhancing mass involving left tonsil extending into tongue base (arrows). Long arrow on the right shows contralateral normal lingual tonsil. Dashed arrow shows a necrotic node. D, Sagittal T2-weighted MRI shows a squamous cancer of the soft palate as a lobulated mass (arrows).

http://radiologyebook.com

672

PART II  CT and MR Imaging of the Whole Body

pterygomandibular raphe to the retromolar trigone is often seen. Advanced cases invade the mandible, masticator space, or infratempo­ ral fossa to reach the skull base.49 Lymphatic spread is seen in 76% of patients at presentation, with level II being the first-echelon nodes; level IB, II-V, and retropharyngeal nodes may also be involved.27 Soft palate SCC usually arises from the oral surface of the palate (see Fig. 22-23D). It spreads anteriorly to the hard palate and inferiorly along the ATP to the tongue base. Lateral extension to the parapharyn­ geal space may extend to the skull base. Superior spread to the

nasopharynx and occasionally perineural spread along the palatine nerves may reach the pterygopalatine fossa.27,84 Lymphatic spread occurs to level I, II, and retropharyngeal nodes, and bilateral nodal metastases are known.49,84 Tongue base SCC may spread anteriorly into the oral tongue and further below to the sublingual space to invade the lingual neurovas­ cular bundle as well as the hyoglossus, genioglossus, and mylohyoid muscles (Fig. 22-24A-B). Perineural spread may occur along the lingual nerve to reach the mandibular nerve. Perineural spread,

* *

A

B

C

D FIG 22-24  Oropharyngeal carcinoma. A, Axial T2-weighted MRI shows normal extrinsic muscles, paired genioglossi (*), and hyoglossi (arrows). B, Axial T2-weighted MRI shows large intermediate signal–intensity tumor (outlined by arrows) invading both-sided extrinsic muscles (compare with A). C, Axial postcontrast T1-weighted image shows normal enhancement in bilateral lingual tonsils (arrows); hence careful attention to asymmetry in enhancement is needed to detect tumor. D, Coronal postcontrast T1-weighted image shows brightly enhancing tumor crossing midline (arrow).

http://radiologyebook.com

CHAPTER 22  Pharynx

A

673

B FIG 22-25  Posterior pharyngeal wall squamous carcinoma. A, Sagittal reformatted postcontrast CT image shows a posterior pharyngeal wall mass (arrows). B, Axial postcontrast CT image shows a posterior pharyngeal wall mass (arrow). Dashed arrow shows intact fat plane with prevertebral muscles. This plane is best evaluated on T1-weighted MRI sequences and when well seen on MRI rules out prevertebral fascial invasion with a high negative predictive value.

vascular invasion, and high T stage are adverse prognostic factors. Spread along the palatoglossus can reach the ATP and further into the soft palate. Posterior spread to the valleculae and preepiglottic space may reach the pyriform sinus. However, early tongue base OPSCC arising from the lingual tonsil may be difficult to differentiate from normal lymphoid tissue in this region (see Fig. 22-24C). Loss of sym­ metry due to spread to adjacent structures can help identify disease.49 Lymphatic metastases occur in 60% and are seen to level II-IV nodes and to submandibular nodes when the floor of the mouth is invaded. Disease may cross the midline (see Fig. 22-24D) and cause bilateral nodal metastases.27,84 Posterior pharyngeal SCCs are least common (Fig. 22-25) but may be involved by spread from adjacent lateral pharyngeal walls. Inferior spread to the hypopharynx may occur. It is important to rule out fixa­ tion of the prevertebral fascia (see Fig. 22-25B), which precludes surgi­ cal resection, although the majority of these lesions are treated with RT or concurrent chemoradiation.49 Isolated retropharyngeal adenop­ athy may be seen. Pretreatment imaging features.  MRI is superior to CT for evalu­ ating the locoregional extent of OPSCC (better soft tissue resolution), but CT best depicts cortical erosion.84 Normal tonsils are well seen on MRI, with hyperintense signal on T2-weighted sequences (see Fig. 22-23A). Although they can appear asymmetric, any obliteration of the parapharyngeal fat with an enlarged tonsil or associated adenopathy (see Fig. 22-23B-C) should be viewed with suspicion.27 Soft palate SCCs are best imaged with coronal and sagittal images (see Fig. 22-23D). In tongue base SCC, coronal and axial postcontrast T1-weighted sequences depict lingual vascular bundles. T2-weighted sequences best depict the extrinsic tongue muscles (see Fig. 22-24A-B), and the preepiglottic space is best seen on sagittal noncontrast T1-weighted sequences. Invasion of the preepiglottic space implies radical surgery, including a supraglottic laryngectomy, or alternatively

could be a relative surgical contraindication. Spread across the midline with invasion of both lingual vascular bundles also needs mention because this may require total glossectomy or could be an indication for nonsurgical therapy (chemoradiation). MRI is also useful to evaluate invasion of prevertebral fascia  and musculature,40,50 to detect retropharyngeal adenopathy (using T2-weighted, short tau inversion recovery [STIR], and fat-suppressed contrast-enhanced sequences), and to depict skull base marrow involvement and intracranial dural invasion (using noncontrast T1 and fat-suppressed contrast-enhanced T1-weighted sequences). It can also detect perineural tumor spread along the greater and lesser pala­ tine nerves into the pterygopalatine fossa (best seen with contrastenhanced fat-suppressed coronal T1-weighted sequences).30 Thickened and enhancing nerves and Meckel’s cave and bulging of the cavernous sinus may be direct signs of perineural spread. Indirect signs include foraminal enlargement, loss of normal fat around foramina or in the pterygopalatine fossa (seen with maxillary nerve invasion12), and denervation atrophy of the masticator muscles, visualized with man­ dibular nerve invasion. Imaging features suggestive of prevertebral fascial invasion (e.g., absent retropharyngeal fat plane) have an accu­ racy of only 60%. Imaging features suggestive of prevertebral muscle invasion, seen as altered T2 signal in the muscle, with irregular contour and enhancement also have a low positive predictive value but need to be recorded.50 However, preservation of the retropharyngeal fat plane on T1-weighted noncontrast sequences has high negative predictive value (97.5%) for absence of prevertebral fascial invasion.40 Marrow invasion is seen as replacement of the bright fat signal on coronal T1-weighted sequences with dark tumor signal intensity and with enhancement on postcontrast sequences. Imaging cannot reliably differentiate between HPV-positive and HPV-negative OPSCC. However, studies have reported that HPVpositive OPSCCs have cystic nodal metastases with well-defined

http://radiologyebook.com

674

PART II  CT and MR Imaging of the Whole Body

A

B FIG 22-26  PET-CT for detecting unknown primary. A, Axial postcontrast CT image shows mild thickness of the median glossoepiglottic fold (arrow). B, PET-CT detects the unknown primary as an FDG-avid lesion.

borders, whereas HPV-negative OPSCCs have metastatic nodes with ill-defined borders invading muscle.6 PET-CT is useful in high T stage cancers to evaluate for distant metastases and in cases presenting with cervical adenopathy to detect an unknown primary that may be seen in the palatine or lingual tonsils87,90 (Fig. 22-26). The main imaging-dependent prognostic indi­ cators in OPSCC are high T and N stage, which are more significant than tumor volume.3 Studies evaluating MRI diffusion and perfusion and PET-CT to identify prognostic markers are ongoing. Response assessment and surveillance.  MRI and PET-CT are both useful for response assessment. Diffusion-weighted imaging is also useful to detect residual disease. A recent study evaluating PET-CT at 3 months in HPV-positive OPSCC demonstrated high negative pre­ dictive value for locoregional failure in complete responders but poor sensitivity and positive predictive value.81 In patients with N2/N3 nodal disease treated with concurrent chemoradiation, absence of nodal disease on clinical examination with lack of nodal uptake on PET-CT is a safe indication to observe the neck without a planned neck dissection.21 The surveillance policy in OPSCC treated with RT or concurrent chemoradiation is not clearly defined. Imaging is known to detect local failure earlier than clinical examination. Asymmetry of the oropharynx with enhancement and bulging contours may suggest recurrent tumor and can be confirmed by PET-CT and/or biopsy. Baseline MRI and CT studies are useful for comparison with future studies, particularly in those with persisting asymmetry and negative biopsies. Routine sub­ sequent PET surveillance in those with negative 3-month PET-CT is of limited benefit.38

HYPOPHARYNX Anatomy The hypopharynx lies behind the larynx in front of the C3-C6 verte­ brae and extends from the tip of the epiglottis (or level of the hyoid bone) superiorly to the inferior border of the cricoid cartilage. It com­

municates anteriorly with the larynx, superiorly with the oropharynx, and inferiorly with the esophagus. It can be divided into three ana­ tomic subsites: the pyriform sinus, the postcricoid area, and the pos­ terior pharyngeal wall.

Pyriform Sinus.  The pyriform sinus is a funnel-shaped or an inverted pyramid–shaped structure that begins superiorly at the glossoepiglot­ tic fold and extends inferiorly, with its apex at the level of the crico­ pharyngeus muscle. It is bounded laterally by the thyroid lamina and posteriorly by the lateral wall of the hypopharynx. Its medial boundary is the lateral surface of the arytenoid cartilage and the aryepiglottic fold (Fig. 22-27A-B). Posteriorly it is open and continuous with the posterior pharyngeal wall. Its apex is where the anterior, lateral, and medial walls meet inferiorly.

Postcricoid Region.  The postcricoid region forms the anterior wall of the hypopharynx and spans along the mucosa lining the posterior wall of the cricoid cartilage (see Fig. 22-27C-D). This area is the most difficult to examine by both mirror exam and direct laryngoscopy.

Posterior Pharyngeal Wall.  The posterior pharyngeal wall extends from a plane drawn at the level of the tip of the epiglottis (some say the level of the vallecula or hyoid) to a plane at the inferior border of the cricoid. The superior and inferior margins of the hypopharynx blend with the posterior wall of the oropharynx and esophagus, respec­ tively. It covers middle and inferior constrictor muscles and is sepa­ rated from prevertebral fascia by the retropharyngeal space. The lining of the hypopharynx is stratified squamous epithelium and has a rich submucosal network of lymphatics that exit superiorly through the thyrohyoid membrane into the superior and middle jugular nodes. Inferiorly the lymphatics drain to the paratracheal and low jugular nodes. The hypopharynx functions as a dynamic conduit for food and helps prevent aspiration. As the food bolus is propelled past the epiglottis, contraction of the constrictor muscles propels the food toward the cricopharyngeus. The cricopharyngeus relaxes as the

http://radiologyebook.com

CHAPTER 22  Pharynx

PS

** *

PS

A

B

C

D

*

675

*

FIG 22-27  Normal hypopharynx. A, Axial postcontrast CT image shows piriform sinus (PS) lateral to aryepiglottic fold (*). B, Similar structures are seen on axial postcontrast fat-saturated T1-weighted image. C, The postcricoid region is delineated by arrows on this axial postcontrast CT image. D, Axial postcontrast T1-weighted image shows postcricoid region (white arrows) and posterior pharyngeal wall (black arrows).

food enters the esophagus, where peristaltic action propels the food to the stomach. Congenital and inflammatory lesions of the hypopharynx will be discussed here, whereas hypopharyngeal tumors are covered along with the larynx.

Congenital and Inflammatory Lesions

Branchial Apparatus Anomalies.  During embryologic develop­ ment, the tissues of the neck are derived from branchial arches that are separated externally by grooves and internally by pharyngeal pouches. Incomplete or aberrant fusion of two adjacent arches can result in the formation of branchial cleft anomalies that include cysts, fistulas, and

sinuses. The majority of branchial apparatus anomalies are cysts, which can originate from the branchial cleft, branchial arch, or pha­ ryngeal pouch remnant. Therefore branchial apparatus cyst is the pre­ ferred terminology and is synonymous with branchial cleft cyst (BCC).55 BCC typically presents in children from 10 years of age to adults aged 30-40 years.46 BCCs are categorized into three types by their embryo­ logic development. First branchial apparatus anomaly.  First BCC develops from the first branchial arch, which extends from the external auditory canal (EAC) through the parotid gland to the submandibular space. The most common location of first BCC is inferior/posterior to the EAC or in the parotid gland.

http://radiologyebook.com

676

PART II  CT and MR Imaging of the Whole Body

Second branchial cleft cyst.  Second BCC is the most common branchial cleft anomaly.46 The most common presentation is presence of a painless fluctuant mass or recurrent inflammation-infection in the submandibular space at angle of the jaw. The classic radiologic appear­ ance of second BCC is that of a well-defined, usually unilocular cyst that shows a predilection for the angle of the jaw, along the anterior margin of sternocleidomastoid muscle, behind the submandibular gland, and lateral to the carotid sheath structures. Varying embryologic development results in other locations for BCC-II: (1) superficial to sternocleidomastoid muscle, (2) medial to the sternocleidomastoid muscle and lateral to the carotid, with a characteristic beak between the common carotid artery bifurcation toward the pharynx, and (3) in the pharyngeal mucosa.46 Third and fourth branchial cleft cysts.  Third BCCs are located in the posterior cervical space posterior to the common or internal carotid artery and the sternocleidomastoid muscle. They may present as a fistula along the anterior border of sternocleidomastoid. The tract passes posterior to the common carotid artery and enters the thyroid membrane to reach the pharynx at the pyriform sinus. Despite its rarity, third BCC is the second most common congenital cystic lesion in the posterior triangle, after cystic hygromas, which are the most common lesions of this space.46,55 Fourth branchial cleft anomalies are generally sinus tracts or fis­ tulae and arise from the pyriform sinus, pierce the thyrohyoid mem­ brane, and descend along the tracheoesophageal groove along  the course of the recurrent laryngeal nerve.46,55 Both anomalies may have fistulous communication from the pyriform sinus through the thyrohyoid membrane into the neck. These anomalies are much less common than second BCC and lymphatic malformation. However, when recurrent neck infection occurs lateral to the pyriform sinus and/or along the tracheoesophageal groove without risk factors or clearly defined etiology, the presence of third or fourth branchial cleft fistula should be suspected. A fluoroscopic barium swallow exam is valuable to demonstrate the fistula for complete surgical excision in these cases.

Miscellaneous

Tortuous ICA.  The internal carotid artery may have a medial course in the retropharyngeal region. This may be unilateral or bilateral (Fig. 22-28). When bilateral, these are called kissing carotids. When seen in children, this represents persistent fetal angulation of carotids. In adults this is usually a manifestation of atherosclerotic changes causing ectasia and tortuosity. Patients are usually asymptomatic, and this is an incidental finding on imaging studies. However, some patients may present with a pulsatile retropharyngeal mass. Diagnosis is important because this condition is a potential risk for pharyngeal surgery and may result in catastrophic bleeding. Therefore, diagnosis prior to imaging prevents iatrogenic bleeding.

Pharyngocele.  Pharyngocele or lateral pharyngeal diverticulum refers to lateral pharyngeal bulging through one of the weak areas of lateral pharyngeal wall. A true pharyngocele is herniation of pharyngeal mucosa beyond the thyrohyoid membrane. Congenital pharyngoceles are branchial cleft sinus tracts that connect internally with the pharynx. Acquired pharyngoceles are pulsion-type diverticula. These arise from the posterior faucial pillar and pyriform fossa and are caused by chron­ ically raised intrapharyngeal pressure accompanied by loss of muscular tone. The diagnosis is generally made by barium swallow studies. Dif­ ferential diagnosis includes Zenker’s diverticulum, laryngocele, esoph­ ageal diverticulum, cricopharyngeal spasm, foreign body, neurologic disease, myopathies, and globus hystericus. Surgery is indicated if the patient is symptomatic.

FIG 22-28  Tortuous internal carotid arteries (ICAs). Axial postcontrast CT image shows right ICA extending into retropharyngeal space (white arrow) while the left ICA is also medial in location (black arrow).

Diverticula/Fistulae.  Zenker’s diverticulum is a pulsion-type diver­ ticulum, a mucosa-lined outpouching from the hypopharynx just proximal to the upper esophageal sphincter. It is a false diverticulum because it does not contain all layers of the pharyngeal wall. It arises from the dehiscence of Killian, a site of focal weakness in the cleavage plane between the circular and oblique fibers of the cricopharyngeus. Most patients are elderly, in the seventh or eighth decade of life. A barium swallow study can depict an outpouching during swallowing, best seen on the lateral view in the posterior midline at the C5-C6 level. Killian Jamieson diverticulum is a true diverticulum that is located just below the cricopharyngeus muscle and is situated anteriorly and later­ ally. These are usually smaller than Zenker’s diverticulum and less often symptomatic.

Dermoid.  Pharyngeal dermoids are choristiomatous developmental anomalies arising from the first branchial cleft area. These are epithe­ lial inclusion cysts resulting from faulty separation of ectoderm from neuroectoderm. Dermoids usually present in the second and third decades of life. On CT these appear as low-density cystic nonenhanc­ ing structures in the midline. Fat, mixed fluid density, and calcifica­ tion may also be seen. On MRI, these lesions are usually hyperintense on both T1- and T2-weighted images and do not demonstrate restricted diffusion. The imaging features are usually pathognomonic. In the pharynx, these lesions can present with airway obstruction and breathing difficulties.

Amyloid.  Amyloidosis is a diverse group of disorders of uncertain etiology. The disease is characterized pathologically by abnormal protein deposition in the extracellular tissue. Within the aerodigestive tract, amyloid preferentially involves the tracheobronchial airway. Involvement of the pharynx by amyloidosis is exceedingly rare. The

http://radiologyebook.com

CHAPTER 22  Pharynx disease can be classified as systemic or localized. Localized disease involves only one organ and has a better prognosis, whereas systemic disease involves multiple organs and has a poorer prognosis. MRI is the technique of choice to demonstrate features of amyloidosis. Amyloid deposits appear isointense on T1-weighted and hypointense on T2-weighted sequences.29

Drugs/Cocaine.  In patients with heavy nasal cocaine abuse, there are reports describing an aggressive intranasal and intrapharyngeal destructive process simulating Wegener’s granulomatosis and midline reticulosis.22 Prominent nasopharyngeal and oropharyngeal ulcers have been reported in these cases, which improve on abstinence from cocaine abuse.

Trauma Traumatic injury to the pharynx is usually seen with direct trauma, instrumentation such as endoscopy, or following ingestion of fish or chicken bones. This may result in disruption of pharyngeal walls and formation of a retropharyngeal abscess. Plain radiographs may dem­ onstrate widening of the prevertebral space and air in the mediastinum or neck. CT can more accurately depict the extent of injury and assess any complications of the injury, such as retropharyngeal abscess, medi­ astinitis, and septic thrombosis of the internal jugular vein. CT is also useful for localization of foreign bodies and aids in management of these cases.

POSTTHERAPEUTIC NECK The management of head and neck cancer involves multidisciplinary evaluation and treatment, including surgery, radiation therapy, and chemotherapy. A multitude of surgical approaches with or without tissue reconstruction, different types of neck dissection, a variety of

radiation therapy techniques, and concurrent and neoadjuvant che­ motherapy regimens may complicate imaging findings in the post­ therapeutic setting. Knowledge of common surgical procedures  and their expected postoperative appearances is vital to avoid falsepositive interpretations. An array of advanced biological imaging  techniques such as MR spectroscopy, CT perfusion, MRI diffusion  and perfusion, and FDG-PET may prove useful in this regard, insofar as they have the ability to look beyond size, morphology, and enhance­ ment and can focus on the molecular-level differences between active tumor and benign tissue.69 This section will review the common postsurgical appearances after various types of neck dissection, recon­ structive techniques, postradiation changes, complications, and a brief discussion on the role of advanced imaging in posttreatment neck evaluation.

Neck Dissection The three major types of neck dissection are radical neck dissection (RND), modified radical neck dissection (MRND), and selective neck dissection (SND). RND is the gold standard by which all other neck dissections are compared. It involves en bloc removal of all the ipsilateral lymph nodes (LN) (levels I-V) and includes the sternocleidomastoid muscle (SCM), internal jugular vein (IJV), submandibular gland (SMG), and spinal accessory nerve (SAN)48,77 (Fig. 22-29A). Indications for RND are extensive cervical involvement or lymph nodes with gross extracapsu­ lar spread and invasion into the adjacent tissues. MRND is the same as RND but with preservation of one or more of the nonlymphatic structures (SCM, SAN, SMG, and IJV) (see Fig. 22-29B). MRND is indicated in patients with less spread and invasion and has some advantages; for example, preservation of the SAN pre­ vents development of frozen shoulder (adhesive capsulitis) and causes less cosmetic deformity than RND.63

*

* SCM

IJV

*

SCM

+ A

677

B

C

FIG 22-29  Neck dissections. A, The submandibular gland (arrow), internal jugular vein (IJV) and sternocleidomastoid muscle (SCM) are present on the right but absent in the left neck after radical neck dissection (RND). There has also been resection of much of the fat in the left neck. B, Modified RND. There is absence of the left submandibular gland (white *), with normal gland seen on the right (arrow). The SCM and IJV (arrow on left) are preserved. There is diminished fat volume deep to (white *) and posterior (+) to the SCM muscle, which is removed along with nodes at levels I-V. C, Posterolateral neck dissection. There is diminished fat volume deep to (arrow in front) and posterior to (arrow behind) to the SCM muscle. The submandibular gland (*) is preserved.

http://radiologyebook.com

678

PART II  CT and MR Imaging of the Whole Body

TABLE 22-1  Typical Postoperative

BOX 22-2  Surgical Reconstruction

Type of Neck Dissection

Pedicle • Rotated regional skin and muscle • Preserved vascular pedicle • Preserved nutrient feeders • Preferred for closing defects in irradiated necks

Appearance After Neck Dissections

RND

Extended RND

Supraomohyoid SND

Lateral SND

Techniques

Postoperative Appearance Absence of: • Level I-V LN • SMG • IJV • SCM RND plus: • Level VI and/or VII lymph node dissections • Retropharyngeal LN dissection, and/or • Internal carotid artery, hypoglossal nerve, and vagus nerve • Level I-III dissection • Absence of right SMG • Preserved right SCM and IJV • Level II-IV dissection • Preserved SCM and IJV

IJV, internal jugular vein; LN, lymph nodes; RND, radical neck dissection; SCM, sternocleidomastoid; SMG, submandibular gland; SND, selective neck dissection.

SND has four common subtypes: the supraomohyoid type (levels I-III and ipsilateral SMG), the lateral type (levels II-IV), the postero­ lateral type (levels II-V as well as posterior auricular and suboccipital LN) (see Fig. 22-29C), and the anterior type (levels VI and VII, typi­ cally bilateral). SND has the advantage that it preserves the functional and cosmetically relevant structures. Extended radical neck dissection is the same as RND but includes removal of additional nodes (levels VI and VII), retropharyngeal LN dissection, and/or nonlymphatic structures such as the internal carotid artery, hypoglossal nerve, and vagus nerve. The typical CT and MRI appearance after neck dissections is sum­ marized in Table 22-1. Note that it may be very difficult to determine that surgery has been performed after SND, and careful correlation with clinical history and surgical notes is mandatory.

Surgery with Reconstruction Curative surgical resection requires a wide local excision to obtain negative margins; however, given the anatomic complexity of the head and neck, complex reconstructive techniques are often required to close the defect and maintain cosmesis and function (Fig. 22-30). Various reconstructive techniques have been introduced to close the surgical defect, and these techniques are mainly classified into three types as outlined in Box 22-2.

Postradiation Changes Radiation therapy for head and neck cancer is classified into two types: external beam radiation therapy and brachytherapy. External beam radiation therapy includes 3D conformal radiation therapy, intensitymodulated radiation therapy, and stereotactic radiosurgery.5 External beam radiation therapy uses photon, electron beam, or proton beam radiation delivered from a source external to the patient. Brachyther­ apy uses radioactive sources such as iodine seeds, iridium, or cesium, which are implanted in the patient. Definitive doses of radiation with external beam radiation therapy for head and neck cancer consist of 66 to 70 Gy delivered daily during a period of 7 weeks. In the past 10 years, intensity-modulated radiation therapy has become the favored

Free Flap • Transferred from distant sites to the surgical field • Skin, soft tissue, muscle, bone, and bowel • No vascular pedicle • Microvascular techniques are employed Prosthesis • Wide spectrum of material and techniques • Maxillofacial and craniofacial • Improved cosmesis and function

technique for administering photon external beam radiation therapy to patients with head and neck cancer, because it allows sparing of multiple organs at risk, including the bilateral parotid glands, pharyn­ geal constrictor muscles, and optic structures, and also allows better coverage of the tumor. Changes after radiation therapy are commonly divided into early and late reactions. CT and MRI findings of early postradiation changes include thickening of the skin and platysma, reticulation of the sub­ cutaneous fat, edema and fluid in the retropharyngeal space, increased enhancement of the major salivary glands, thickening and increased enhancement of the pharyngeal walls, and thickening of the laryngeal structures (Fig. 22-31). Late reactions to radiation therapy include atrophy of the salivary glands and thickening of the pharyngeal con­ strictor muscles, platysma, and skin.79

Complications of Therapy These can be divided into complications after surgery, osseous com­ plications, vascular complications, and radiation-induced complica­ tions, as summarized in Table 22-2.

Imaging Appearances of Tumor Recurrence The most common locations for tumor recurrence are in the operative bed and at the margins of the surgical site. Tumors typically recur within the first 2 years after treatment. Early recurrence within weeks after surgery—before adjuvant radiation therapy—is sometimes seen because of accelerated repopulation. Tumor recurrence is identified as an expansile lesion in the operative bed or as progressive thickening of soft tissues deep to the flap.10,37,41,63 CT demonstrates recurrence as an infiltrating slightly hyperattenu­ ating enhancing mass, with or without bone destruction10,37,63 (Fig. 22-32). Tumor recurrence has attenuation similar to that of muscle. Therefore if a suspected mass has lower attenuation than that of muscle, it is unlikely to be a malignancy and often is related to edema.77 MRI demonstrates tumor recurrence as an infiltrative mass with intermediate T1-weighted signal intensity, intermediate to high T2-  weighted signal intensity, and postcontrast enhancement. The differ­ ential diagnosis for tumor recurrence includes a vascularized scar, which represents early fibrosis.48 Fundamental tenets for the evaluation of tumor recurrence include careful utilization of comparison scans, with a high index of suspicion

http://radiologyebook.com

CHAPTER 22  Pharynx

* A

C

D

B FIG 22-30  Surgical reconstruction. A and B, Pectoralis major myocutaneous pedicle flap (arrows) is seen along anterior neck on postcontrast axial CT image (A) and reformatted sagittal image (B) and was used to fill defect of floor of mouth and neck following multiple surgeries for recurrent ameloblastoma. C, Postcontrast coronal reformatted CT image shows rectus abdominis myocutaneous free flap (*) filling the defect in oral cavity and masticator space following anterior two-thirds glossectomy and composite resection for recurrent tongue cancer invading the mandible. D, Fibular free tissue transfer (arrows) was used to reconstruct mandible.

http://radiologyebook.com

679

680

PART II  CT and MR Imaging of the Whole Body

* *

A

B FIG 22-31  Postradiation changes. A, Axial postcontrast CT image shows thickening of platysma muscles (arrows) and reticulation of subcutaneous fat. B, Axial postcontrast CT image shows diffuse thickening of the epiglottis (*).

A

B FIG 22-32  Recurrent disease in a patient with piriform sinus cancer status post total laryngopharyngectomy with anterolateral thigh free flap. A, Axial postcontrast CT image shows a mass (arrows) near the lower end of flap. B, Sagittal reformatted postcontrast CT image shows the recurrent mass (arrows) just superior to the tracheostomy and tracheoesophageal puncture.

http://radiologyebook.com

CHAPTER 22  Pharynx

A

681

B FIG 22-33  Physiologic imaging. A, Axial postcontrast CT image shows mass along right upper margin of anterolateral thigh free flap (arrows), of concern for recurrence in a patient status post total laryngopharyngectomy for recurrent hypopharyngeal cancer. B, PET-CT shows hypermetabolism in mass of concern on CT (arrow on right). There is also hypermetabolism along left side of flap, compatible with recurrence (arrow) not suspected on CT.

TABLE 22-2  Complications of Therapy Infection

Vascular

Radionecrosis

Chylous fistula

• • • • • • • •

Neuroma

• • •

Dermal seeding

•

Other

• • • • •

Cellulitis Phlegmon Abscess IJV thrombosis Arterial injury Radiation arteritis Can occur in native and grafted bone Imaging findings of osteoradionecrosis: 1. Cortical disruption 2. Rarefaction 3. Sequestration 4. Pathologic fracture 5. Diffuse soft tissue thickening Usually occurs after a level IV LN dissection Appears as a persistent fluid collection Rare proliferative reparative response to transected peripheral nerves; not a true neoplasm Most commonly occurs in patients with extracapsular spread of disease or in immunocompromised patients Tissue fibrosis Delayed central nervous system reaction Radiation myelopathy Cranial nerve palsy Secondary tumors

IJV, internal jugular vein; LN, lymph nodes.

for any enlarging mass or new lymphadenopathy, which should be considered a recurrence unless proven otherwise.

ADVANCED BIOLOGICAL IMAGING TECHNIQUES The traditional role of imaging has remained in the anatomic domain, but recent advanced imaging techniques such as MR spectroscopy, MRI diffusion and perfusion, CT perfusion, and FDG-PET provide information about cellularity, perfusion, neoangiogenesis, and metab­ olism and hold promise in assessment of pharyngeal cancers. These techniques are useful in monitoring tumor response during chemora­ diation, assessing response at the end of therapy, and detecting residual or recurrent disease and metastasis (Fig. 22-33). Information that can be obtained with these techniques will help propel us toward custom­ ization of treatments and better prognostication for patients in this era of personalized medicine.78

REFERENCES 1. Abele TA, Salzman KL, Harnsberger HR, et al: Craniopharyngeal canal and its spectrum of pathology. AJNR Am J Neuroradiol 35(4):772–777, 2014. 2. Acierno SP, Waldhausen JH: Congenital cervical cysts, sinuses   and fistulae. Otolaryngol Clin North Am 40:161–176, vii–viii, 2007. 3. Agarwal JP, Mallick I, Bhutani R, et al: Prognostic factors in oropharyngeal cancer—Analysis of 627 cases receiving definitive radiotherapy. Acta Oncol 48(7):1026–1033, 2009.

http://radiologyebook.com

682

PART II  CT and MR Imaging of the Whole Body

4. Aloia MS, Arnedt JT, Davis JD, et al: Neuropsychological sequelae of obstructive sleep apnea-hypopnea syndrome: A critical review. J Int Neuropsychol Soc 10:772–785, 2004. 5. Burri RJ, Kao J, Navada S, et al: Nonsurgical treatment of head and neck cancer. In Som PM, Curtin HD, editors: Head and neck imaging, ed 5, St. Louis, 2011, Mosby, pp 2893–2914. 6. Cantrell SC, Peck BW, Li G, et al: Differences in imaging characteristics of HPV-positive and HPV-negative oropharyngeal cancers: A blinded matched-pair analysis. AJNR Am J Neuroradiol 34(10):2005–2009, 2013. 7. Capps EF, Kinsella JJ, Gupta M, et al: Emergency imaging assessment of acute, nontraumatic conditions of the head and neck. Radiographics 30(5):1335–1352, 2010. 8. Chen C, Fei Z, Pan J, et al: Significance of primary tumor volume and T-stage on prognosis in nasopharyngeal carcinoma treated with intensity-modulated radiation therapy. Jpn J Clin Oncol 41(4):537–542, 2011. 9. Chen Y, Liu X, Zheng D, et al: Diffusion-weighted magnetic resonance imaging for early response assessment of chemoradiotherapy in patients with nasopharyngeal carcinoma. Magn Reson Imaging 32(6):630–637, 2014. 10. Chong VF: Post treatment imaging in head and neck tumours. Cancer Imaging 5(1):8–10, 2005. 11. Chong VF, Fan YF: Detection of recurrent nasopharyngeal carcinoma: MR imaging versus CT. Radiology 202:463–470, 1997. 12. Chong VF, Fan YF: Pterygopalatine fossa and maxillary nerve infiltration in nasopharyngeal carcinoma. Head Neck 19(2):121–125, 1997. 13. Chong VF, Fan YF: Hypoglossal nerve palsy in nasopharyngeal carcinoma. Eur Radiol 8(6):939–945, 1998. 14. Chong VF, Fan YF: Meningeal infiltration in recurrent nasopharyngeal carcinoma. Australas Radiol 44:23–27, 2000. 15. Chong VF, Fan YF: Radiology of the nasopharynx: Pictorial essay. Australas Radiol 44(1):5–13, 2000. Review. 16. Chong VF, Fan YF, Khoo JB: Nasopharyngeal carcinoma with intracranial spread: CT and MR characteristics. J Comput Assist Tomogr 20(4):563–569, 1996. 17. Chong VF, Mukherji SK, Ng SH, et al: Nasopharyngeal carcinoma: Review of how imaging affects staging. J Comput Assist Tomogr 23(6):984–993, 1999. 18. Chwang WB, Jain R, Narayan A, et al: Inflammatory pseudotumor of the nasopharynx and skull base: Mimicking an aggressive neoplasm or infection. Arch Otolaryngol Head Neck Surg 138(8):765–769, 2012. 19. Comoretto M, Balestreri L, Borsatti E, et al: Detection and restaging of residual and/or recurrent nasopharyngeal carcinoma after chemotherapy and radiation therapy: Comparison of MR imaging and FDG PET/CT. Radiology 249(1):203–211, 2008. 20. Corey AS, Hudgins PA: Radiographic imaging of human papillomavirus related carcinomas of the oropharynx. Head Neck Pathol 6(Suppl 1): S25–S40, 2012. 21. Corry J, Peters L, Fisher R, et al: N2-N3 neck nodal control without planned neck dissection for clinical/radiologic complete respondersresults of Trans Tasman Radiation Oncology Group Study 98.02. Head Neck 30(6):737–742, 2008. 22. Daggett RB, Haghighi P, Terkeltaub RA: Nasal cocaine abuse causing an aggressive midline intranasal and pharyngeal destructive process mimicking midline reticulosis and limited Wegener’s granulomatosis.   J Rheumatol 17(6):838–840, 1990. 23. Dähnert W: Ear, nose, and throat disorders. In Radiology review manual, ed 4, Baltimore, 1999, Williams & Wilkins, pp 314–335. 24. de Almeida JR, Moskowitz AJ, Miles BA, et al: Transoral robotic surgery is cost-effective compared to (chemo)radiotherapy for early T-classification oropharyngeal carcinoma: A cost-utility analysis. Head Neck 2014. [Epub ahead of print]. 25. Donnelly LF, Jones BV, Strife JL: Imaging of pediatric tongue abnormalities. AJR Am J Roentgenol 175(2):489–493, 2000. 26. Freling NJ, Merks JH, Saeed P, et al: Imaging findings in craniofacial childhood rhabdomyosarcoma. Pediatr Radiol 40(11):1723–1738, quiz 1855, 2010.

27. Gale DR: CT and MRI of the oral cavity and oropharynx. In Valvassori GE, Mafee MF, Carter BL, editors: Imaging of the head and neck, Stuttgart, 1995, Thieme, pp 445–474. 28. Geara FB, Sanguineti G, Tucker SL, et al: Carcinoma of the nasopharynx treated by radiotherapy alone: Determinants of distant metastasis and survival. Radiother Oncol 43(1):53–61, 1997. 29. Gilad R, Milillo P, Som PM: Severe diffuse systemic amyloidosis with involvement of the pharynx, larynx, and trachea: CT and MR findings. AJNR Am J Neuroradiol 28(8):1557–1558, 2007. Review. PubMed PMID: 17846211. 30. Ginsberg LE, DeMonte F: Imaging of perineural tumor spread from palatal carcinoma. AJNR Am J Neuroradiol 19(8):1417–1422, 1998. 31. Goins MR, Beasley MS: Pediatric neck masses. Oral Maxillofac Surg Clin North Am 24:457–468, 2012. 32. Gupta T, Master Z, Kannan S, et al: Diagnostic performance of post-treatment FDG PET or FDG PET/CT imaging in head and neck cancer: A systematic review and meta-analysis. Eur J Nucl Med Mol Imaging 38(11):2083–2095, 2011. 33. Harnsberger HR: Head and neck imaging. In Osborne AG, Bragg DC, editors: Handbooks in radiology (series), ed 2, St. Louis, 1995, Mosby-Year Book. 34. Healy JN, Borg MF: Paediatric nasopharyngeal rhabdomyosarcoma: A case series and literature review. J Med Imaging Radiat Oncol 54(4):388– 394, 2010. 35. Hegde AN, Mohan S, Pandya A, et al: Imaging in infections of the head and neck. Neuroimaging Clin N Am 22(4):727–754, 2012. 36. Hermans R: Head and neck cancer: How imaging predicts treatment outcome. Cancer Imaging 6:S145–S153, 2006. 37. Hermans R: Posttreatment imaging in head and neck cancer. Eur J Radiol 66(3):501–511, 2008. 38. Ho AS, Tsao GJ, Chen FW, et al: Impact of positron emission tomography/computed tomography surveillance at 12 and 24 months for detecting head and neck cancer recurrence. Cancer 119(7):1349– 1356, 2013. 39. Hsieh WS, Yang PH, Wong KS, et al: Vallecular cyst: An uncommon cause of stridor in newborn infants. Eur J Pediatr 159(1–2):79–81, 2000. 40. Hsu WC, Loevner LA, Karpati R, et al: Accuracy of magnetic resonance imaging in predicting absence of fixation of head and neck cancer to the prevertebral space. Head Neck 27:95–100, 2005. 41. Hudgins PA: Flap reconstruction in the head and neck: Expected appearance, complications, and recurrent disease. Eur J Radiol 44(2):130–138, 2002. 42. Kadlub N, Touma J, Leboulanger N, et al: Head and neck teratoma: From diagnosis to treatment. J Craniomaxillofac Surg 42(8):1598–1603, 2014. PubMed PMID: 24954765. 43. Khedmat H, Taheri S: Post-transplantation lymphoproliferative disorders localizing in the adenotonsillar region: Report from the PTLD.Int survey. Ann Transplant 16(1):109–116, 2011. 44. King AD, Lei KI, Richards PS, et al: Non-Hodgkin’s lymphoma of the nasopharynx: CT and MR imaging. Clin Radiol 58(8):621–625, 2003. 45. Koch BL: Cystic malformations of the neck in children. Pediatr Radiol 35:463–477, 2005. 46. Koeller KK, Alamo L, Adair CF, et al: Congenital cystic masses of the neck: Radiologic-pathologic correlation. Radiographics 19:121–146, quiz 152–123, 1999. 47. Langendijk JA, Leemans CR, Buter J, et al: The additional value of chemotherapy to radiotherapy in locally advanced nasopharyngeal carcinoma: A meta-analysis of the published literature. J Clin Oncol 22(22):4604–4612, 2004. 48. Lell M, Baum U, Greess H, et al: Head and neck tumors: Imaging recurrent tumor and post-therapeutic changes with CT and MRI. Eur J Radiol 33(3):239–247, 2000. 49. Lenz M, Hermans R: Imaging of the oropharynx and oral cavity. Part II: Pathology. Eur Radiol 6(4):536–549, 1996. 50. Loevner LA, Ott IL, Yousem DM, et al: Neoplastic fixation to the prevertebral compartment by squamous cell carcinoma of the head and neck. AJR Am J Roentgenol 170:1389–1394, 1998.

http://radiologyebook.com

CHAPTER 22  Pharynx 51. Lohman BD, Sarikaya B, McKinney AM, et al: Not the typical Tornwaldt’s cyst this time? A nasopharyngeal cyst associated with canalis basilaris medianus. Br J Radiol 84(1005):e169–e171, 2011. 52. Masterson L, Moualed D, Liu ZW, et al: De-escalation treatment protocols for human papillomavirus-associated oropharyngeal squamous cell carcinoma: A systematic review and meta-analysis of current clinical trials. Eur J Cancer 50(15):2636–2648, 2014. 53. Morgan WE, Friedman EM, Duncan NO, et al: Surgical management of macroglossia in children. Arch Otolaryngol Head Neck Surg 122:326–339, 1996. 54. Mukherji SK: Pharynx. In Som PM, Curtin HD, editors: Head and neck imaging, ed 4, St. Louis, 2003, Mosby, pp 1465–1520. 55. Mukherji SK, Fatterpekar G, Castillo M, et al: Imaging of congenital anomalies of the branchial apparatus. Neuroimaging Clin N Am 10:75–93, 2000. 56. Mukherji SK, Holliday RA: Pharynx. In Head and neck imaging, ed 3, St. Louis, 1996, Mosby-Year Book, pp 437–487. 57. Neel HB, Slavitt DH: Nasopharyngeal cancer. In Baily BJ, editor: Head and neck surgery: Otolaryngology, Philadelphia, 1993, JB Lippincott, pp 1257–1260. 58. Neville BW, Damm DD, Allen CM, et al: Oral & maxillofacial pathology, ed 2, Philadelphia, 2002, W.B. Saunders, p 2. 59. Ng SH, Chang TC, Ko SF, et al: Nasopharyngeal carcinoma: MRI and CT assessment. Neuroradiology 39:741–746, 1997. 60. Ng SH, Chang JT, Ko SF, et al: MRI in recurrent nasopharyngeal carcinoma. Neuroradiology 41:855–862, 1999. 61. Ng SH, Chan SC, Yen TC, et al: Staging of untreated nasopharyngeal carcinoma with PET/CT: Comparison with conventional imaging work-up. Eur J Nucl Med Mol Imaging 36(1):12–22, 2009. 62. Norbash AM: Nasopharynx and deep facial spaces. In Eldman RR, Hesselink JR, Zlatkin MB, editors: Clinical magnetic resonance imaging, ed 2, Philadelphia, 1996, WB Saunders, pp 1079–1109. 63. Offiah C, Hall E: Post-treatment imaging appearances in head and neck cancer patients. Clin Radiol 66(1):13–24, 2011. 64. Ong CK, Chong VF: Imaging of perineural spread in head and neck tumours. Cancer Imaging 10(SpecA):S92–S98, 2010. 65. Pfister DG, et al: Cancers of the nasopharynx. In Head and Neck Cancers. NCCN clinical practice guidelines in oncology, ed 2, 2011, National Comprehensive Cancer Network (NCCN). Available at: . 66. Radkowski D, McGill T, Healy GB, et al: Angiofibroma. Changes in staging and treatment. Arch Otolaryngol Head Neck Surg 122(2):122–129, 1996. 67. Rathore YS, Sinha S, Mahapatra AK: Transsellar transsphenoidal encephalocele: A series of four cases. Neurol India 59(2):289–292, 2011. 68. Righi S, Boffano P, Pateras D, et al: Thornwaldt cysts. J Craniofac Surg 25(5):e456–e457, 2014. PubMed PMID: 25148637. 69. Saito N, Nadgir RN, Nakahira M, et al: Posttreatment CT and MR imaging in head and neck cancer: What the radiologist needs to   know. Radiographics 32(5):1261–1282, 2012. PubMed PMID: 22977017. 70. Schoen EJ, Clapp W, To TT, et al: The key role of newborn thyroid scintigraphy with isotopic iodide (123I) in defining and managing congenital hypothyroidism. Pediatrics 114(6):e683–e688, 2004. 71. Schraff S, McGinn JD, Derkay CS: Peritonsillar abscess in children:   A 10-year review of diagnosis and management. Int J Pediatr Otorhinolaryngol 57(3):213–218, 2001.

683

72. Segal N, Atamne E, Shelef I, et al: Intracranial infection caused by spreading through the fossa naviclaris magna—A case report and review of the literature. Int J Pediatr Otorhinolaryngol 77(12):1919–1921, 2013. PubMed PMID: 24148862. 73. Shefelbine SE, Mancuso AA, Gajewski BJ, et al: Pediatric retropharyngeal lymphadenitis: Differentiation from retropharyngeal abscess and treatment implications. Otolaryngol Head Neck Surg 136(2):182–188, 2007. 74. Sheman LJ: Diseases of the oropharynx. In Lee KJ, editor: Textbook of otolaryngology and head and neck surgery, New York, 1989, Elsevier, pp 407–414. 75. Shiels WE, 2nd, Kang DR, Murakami JW, et al: Percutaneous treatment of lymphatic malformations. Otolaryngol Head Neck Surg 141(2):219– 224, 2009. 76. Snyderman CH, Pant H, Carrau RL, et al: A new endoscopic staging system for angiofibromas. Arch Otolaryngol Head Neck Surg 136(6):588– 594, 2010. 77. Som PM, Lawson W, Genden EM: The posttreatment neck: Clinical   and imaging considerations. In Som PM, Curtin HD, editors:   Head and neck imaging, ed 5, St. Louis, 2011, Mosby, pp 2771–2822. 78. Srinivasan A, Mohan S, Mukherji SK: Biologic imaging of head and neck cancer: The present and the future. AJNR Am J Neuroradiol 33(4):586– 594, 2012. PubMed PMID: 22194364. 79. Stone HB, Coleman CN, Anscher MS, et al: Effects of radiation on normal tissue: Consequences and mechanisms. Lancet Oncol 4(9):529– 536, 2003. 80. Surov A, Ryl I, Bartel-Friedrich S, et al: Diffusion weighted imaging of nasopharyngeal adenoid hypertrophy. Acta Radiol 56(5):587–591, 2015. PubMed PMID: 24855289. 81. Vainshtein JM, Spector ME, Stenmark MH, et al: Reliability of post-chemoradiotherapy F-18-FDG PET/CT for prediction of locoregional failure in human papillomavirus-associated oropharyngeal cancer. Oral Oncol 50(3):234–239, 2014. 82. Vargas H, Nazeer T, Conti D, et al: Posttransplant lymphoproliferative disorder of the nasopharynx. Am J Rhinol 16(1):37–42, 2002. 83. Wang WY, Twu CW, Lin WY, et al: Plasma Epstein-Barr virus DNA screening followed by (1)(8)F-fluoro-2-deoxy-D-glucose positron emission tomography in detecting posttreatment failures of nasopharyngeal carcinoma. Cancer 117:4452–4459, 2011. 84. Wesolowski JR, Mukherji SK: Pathology of the pharynx. In Som PM, Curtin HD, editors: Head & neck imaging (vol 2), ed 5, St. Louis, 2011, Elsevier Mosby. 85. Wong KS, Huang YH, Wu CT: A vanishing tongue-base cyst. Turk J Pediatr 49(4):451–452, 2007. 86. Yaggi H, Cancato J, Kernan W, et al: Obstructive sleep apnea as a risk factor for stroke and death. N Engl J Med 353:2034–2041, 2005. 87. Yongkui L, Jian L, Wanghan, et al: 18FDG-PET/CT for the detection of regional nodal metastasis in patients with primary head and neck cancer before treatment: A meta-analysis. Surg Oncol 22(2):e11–e16, 2013. 88. Yousem DM, Chalian AA: Oral cavity and pharynx. Radiol Clin North Am 36:967–981, 1998. 89. Zadvinskis DP, Benson MT, Kerr HH, et al: Congenital malformations of the cervicothoracic lymphatic system: Embryology and pathogenesis. Radiographics 12(6):1175–1189, 1992. 90. Zhu L, Wang N: 18F-fluorodeoxyglucose positron emission tomographycomputed tomography as a diagnostic tool in patients with cervical nodal metastases of unknown primary site: A meta-analysis. Surg Oncol 22(3):190–194, 2013.

http://radiologyebook.com

23  Paranasal Sinuses Varsha Joshi, Rima Sansi, Saugata Sen, and Piyush Saxena

IMAGING OPTIONS AND PROTOCOLS Computed tomography (CT) and magnetic resonance imaging (MRI) are the primary imaging modalities for assessment of the paranasal sinuses. Both these imaging techniques are far superior to plain radiographs in the evaluation of the sinonasal cavities. Plain radiographs  of the paranasal sinuses are limited in their ability to delineate the complex bony anatomy or to characterize sinonasal disease and hence are almost obsolete today. They also suffer from high false-negative results.

CT CT scan is the primary imaging modality for evaluation of sinonasal disease. The concept of screening sinus CT (SSCT) performed in  the coronal plane with single window settings, 3-mm sections, and without administration of contrast was promoted in the early 1990s when functional endoscopic sinus surgery (FESS) became popular among endoscopic surgeons. With the evolution of CT scanners to multidetector (MD) CT technology, direct coronal acquisition is no longer commonly performed. Today sinus CT is performed in the axial plane with the patient supine on the scanning table and neutral position of the gantry using 0.625-mm collimation. The volumetric dataset is then reconstructed in coronal, axial, and sagittal planes to provide up to 0.9-mm-thick images in bone (width 3500, center 500) and soft tissue (width 270, center 70) algorithms. When a contrast CT is required, 3- to 4-mm-thick slices are generated in soft tissue algorithm.

MRI MRI of the paranasal sinuses is performed using a head and neck  coil. The protocol includes nonenhanced high-resolution 3- to 4-mm T1- and T2-weighted images, diffusion-weighted imaging (DWI), and contrast-enhanced T1-weighted images with fat suppression. Axial and coronal planes are preferred, and sagittal sequences may be performed when required. Care must be taken to include the adjacent orbits and intracranial cavity along with the sinuses.

ANATOMY Nose, Nasal Cavity, and Lateral Nasal Wall The nose implies the external nose that projects anteriorly from the face. It consists of a bony portion superiorly and a cartilaginous portion inferiorly. The anatomic subunits of the nose include the nasal root, dorsum, side walls, and nasal septum. The nasal septum is the midline partition that separates the right and left side of the nose and is made up of the septal (quadrangular) cartilage anteriorly, the perpendicular plate of the ethmoid bone posterosuperiorly, and the vomer posteroinferiorly. The external nose leads posteriorly into the nasal cavity, which in turn leads posteriorly into the nasopharynx.

684

The lateral wall of the nasal cavity shows three (sometimes four) constant projections—the superior, middle, and inferior turbinates— that divide the lateral nasal wall into three meati lateral to the turbinates—the superior, middle, and inferior meati, respectively.  A fourth turbinate (supreme turbinate) may be present, and the supreme meatus is the space above the superior turbinate and lateral to the supreme turbinate. The turbinates and meati are best seen on coronal CT sections (Fig. 23-1). The middle turbinate and meatus form the most important anatomic area in the lateral wall of the nose. The attachments of the middle turbinate have anatomic and surgical relevance and provide fairly good stability to the turbinate. Anteriorly the turbinate runs in the sagittal plane and attaches superiorly to the cribriform plate (see Fig. 23-1). As it continues behind, it turns laterally into the coronal plane and attaches to the lamina papyracea to form the basal lamella that divides the ethmoid cells into the anterior and posterior ethmoid cells. The middle meatus contains the drainage pathways of the anterior ethmoid air cells, the maxillary and frontal sinuses (osteomeatal unit [OMU]). This region is most commonly involved in the pathophysiology of chronic rhinosinusitis and will be described in greater detail in the following sections.

Paranasal Sinuses There are four paired paranasal sinuses, one on either side of the midline. They develop from ridges in the lateral nasal wall by the  eighth week of embryogenesis and continue pneumatization until  early adulthood. Each sinus is named after the bone in which it is located. The relevant imaging anatomy is described in the following sections.4,5,57,62,75

Maxillary Sinus.  Each maxillary sinus is located within the maxillary bone and is the largest and most constant paranasal sinus. It is present at birth and shows a bimodal growth pattern between ages 1 and 3 and 7 and 18 years. The roof of the sinus is formed by the floor of the orbit and has a canal for the second branch of fifth cranial nerve (see Fig. 23-1). The medial wall is formed by portions of the ethmoid, palatine, and lacrimal bones. A fairly large part of the medial wall, however, does not contain any bone and is formed by connective tissue and sinus mucosa. This membranous area is called the fontanelle and can break down secondary to sinus infection, forming an accessory ostium of the sinus. The accessory ostium is nonfunctional and drains the sinus only when the main ostium is blocked. The primary maxillary sinus ostium is located on the highest part of the medial wall of the maxillary sinus (see Fig. 23-1). The floor of the sinus is formed by the alveolar process of the maxilla and palatine bones. The posterolateral wall separates it from the pterygomaxillary fissure medially and infratemporal fossa laterally (see Fig. 23-23A, B, and D). The maxillary sinus drains through the primary ostium into the ethmoid infundibulum (see Fig. 23-1).

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

685

thick anterior table, and the posterior wall is the thinner posterior table that separates the sinus from the cranial cavity. The floor is  the anterior roof of the orbit. The frontal sinus drains through the frontal sinus ostium and the frontal recess into the middle meatus (see Fig. 23-2).

Sphenoid Sinus.  The sphenoid sinus is located in the body of the

B

sphenoid. It begins to grow at age 3 and is fully pneumatized by about age 18. The sphenoid sinuses on the two sides are separated by the intersphenoid septum. The roof of the sinus is formed by the sellar floor, and the lateral wall demarcates the sinus from the cranial  cavity. The carotid artery is situated in the lateral wall, the optic nerve superolateral to the sinus, and the vidian nerve canal in its floor (see Fig. 23-5A). The sphenoid sinus ostium is located medially in the anterosuperior portion of the sinus and opens into the sphenoethmoid recess (Fig. 23-3).

MT MS IT

FIG 23-1  Coronal CT image through the osteomeatal unit. Normal anatomy is depicted on the left side. Note the uncinate process (elbow arrow), ethmoid infundibulum (dotted line), hiatus semilunaris (curved line), ethmoid bulla (B), frontal recess (dashed line), maxillary sinus (MS) ostium (white circle), middle meatus (double line), Kero’s type I ethmoid roof on left and type II ethmoid roof on right (vertical white arrows), and fovea ethmoidales forming the roof of the ethmoid cells (vertical white arrowheads). Also seen are the infraorbital foramen in the roof of the maxillary sinus, inferior turbinate (IT), middle turbinate (MT), lamina papyracea (horizontal white arrowheads), and the opening of the anterior ethmoid artery (horizontal black arrow). The middle turbinate is seen attaching superiorly to the cribriform plate. Note the infundibular pattern of sinonasal disease on the right side, with opacified maxillary sinus and infundibulum (curved white arrow).

Ethmoid Sinus (Ethmoid Air Cells).  Each ethmoid sinus is located within the ethmoid labyrinth and has a honeycomb appearance. It is present at birth and continues to grow until about 12 years of age. The ethmoid labyrinth is situated inferolateral to the olfactory fossa, which is bounded by the horizontal and vertical lamellae of the cribriform plate. The roof of the sinus is formed by the fovea ethmoidales of the frontal bone, which joins the vertical lamella of the cribriform plate (see Fig. 23-1). The medial wall is formed by the lateral surface of the middle and superior turbinate, the lateral wall by the lamina papyracea (see Fig. 23-1), and the posterior wall by the anterior wall of the sphenoid sinus (Fig. 23-2). The ethmoid sinus is divided into multiple discrete cells by bony lamellae that extend laterally to the lamina papyracea and superiorly to the fovea ethmoidales. Of these, the basal lamella divides the ethmoid air cells into anterior and posterior groups of ethmoid cells. The anterior ethmoid air cells drain into the ethmoid infundibulum, and the posterior ethmoid air cells drain into the superior meatus (see Figs. 23-1 and 23-2). The pneumatization of the ethmoid air cells shows the highest variation among all the sinuses. There are two primary types of ethmoid air cells: intramural and extramural. Intramural cells remain within the ethmoid bone and include the ethmoid bulla, suprabullar cell, and frontal bullar cell. The extramural cells develop external to the ethmoid labyrinth and include the agger nasi cell, frontal cell, supraorbital ethmoid cell, Haller cell, and Onodi cell. These are considered separately in the section on anatomic variants.

Frontal Sinus.  Each frontal sinus lies within the frontal bone. The frontal sinus is rudimentary at birth and begins to grow by age 6, continuing until the late teens. The frontal sinuses on two sides are separated by the interfrontal sinus septum. The anterior wall is the

Drainage Pathways of the Paranasal Sinuses The paranasal sinuses are broadly divided into two major groups: the anterior sinuses and the posterior sinuses. The anterior sinuses comprise the frontal sinus, anterior ethmoid air cells, and maxillary sinus. These drain into a common area centered in the middle meatus, the osteomeatal unit (OMU). The posterior sinuses are the posterior ethmoid air cells and the sphenoid sinuses. These drain into the  sphenoethmoid recess.

Osteomeatal Unit.  The OMU is the common drainage pathway of the anterior sinuses, the key area in the pathophysiology of chronic sinusitis and the center of interest since the advent of FESS. It is a narrow anatomic region bounded by the middle turbinate medially, the lamina papyracea laterally, and the basal lamella posteriorly. This region and all its components are best visualized on coronal CT sections (see Fig. 23-1). The OMU comprises the uncinate process, hiatus semilunaris, ethmoid infundibulum, and the ethmoid bulla (bulla ethmoidales). The uncinate process is a key component of the osteomeatal unit and a very important surgical landmark for endoscopic sinus surgery. It is a hook-shaped bone that arises from the posteromedial wall of the nasolacrimal duct and is best seen on coronal CT sections as the superior extension of the medial wall of the maxillary sinus (see Fig. 23-1). The maxillary sinus ostium is located behind the uncinate process. The hiatus semilunaris is a crescent-shaped cleft located just posterior and superior to the uncinate process and anteroinferior to the ethmoid infundibulum (see Fig. 23-1). It serves as the communication of the ethmoid infundibulum with the middle meatus. The ethmoid infundibulum receives drainage from the ethmoid bulla in its anterior portion and the maxillary sinus ostium in  its floor. It opens into the hiatus semilunaris and is located inferomedial to the ethmoid bulla and superolateral to the uncinate process (see Fig. 23-1). The ethmoid bulla is the largest and most constant of the anterior ethmoid air cells. It receives drainage from the anterior ethmoid air cells and drains into the ethmoid infundibulum. There may be variation in the amount of pneumatization of the ethmoid bulla. If it is not pneumatized, it is termed torus ethmoidales; excessive pneumatization may form a giant bulla that may fill the entire middle meatus. The ethmoid bulla is located just posterior and superior to the hiatus semilunaris and posterior to the frontal recess. The lateral wall is formed by the lamina papyracea, the roof by the skull base, and the posterior wall by the basal lamella (see Figs. 23-1 and 23-2). Sometimes an air recess may form between the bulla and the basal lamella; this is called a retrobullar recess or sinus lateralis.

http://radiologyebook.com

686

PART II  CT and MR Imaging of the Whole Body Frontal Recess and Frontal Sinus Drainage Pathway.  The

FS SB B A

PE SS

FIG 23-2  Sagittal CT image through the frontal sinus drainage pathway. Frontal sinus (FS) narrows down to the frontal sinus ostium at the level of the nasofrontal spine (black arrow) and then the frontal recess (dashed white line). Agger nasi cell (A) and type I frontal cell (elbow black arrow) are seen anterior to the frontal recess, and suprabullar cell (SB) and bulla ethmoidales (B) are seen posteriorly draining into the infundibulum (dash-dot line). Note the posterior ethmoid air cells (PE) draining into the superior meatus (white line). The posterior wall of the ethmoid air cells is formed by the anterior wall of the sphenoid sinus.

frontal sinus drainage pathway (FSDP) comprises the frontal sinus infundibulum, frontal sinus ostium (FSO), and frontal recess (FR).29,36,44 Only the FR is a part of the OMU. The entire anatomy of the FSDP is best seen on coronal and sagittal images. The frontal sinus tapers down to a funnel-shaped inferior portion called the frontal sinus infundibulum. This frontal infundibulum leads to the FSO, which is the narrowest part of the FSDP and located in its most dependent portion in the floor of the sinus at the level of the nasofrontal spine (see Fig. 23-2). The frontal infundibulum and the FSO are best visualized on sagittal images. The portion of the FSDP below the ostium is the FR, and this drains into the middle meatus (see Fig. 23-2). It measures about 13  mm in anterior-to-posterior dimension and has the shape of an inverted funnel with its tip at the frontal ostium. The FR is surrounded by the ethmoid air cells—the agger nasi cell anteriorly, the ethmoid bulla posteriorly and laterally, and the most anterior and superior portion of the middle turbinate medially (see Figs. 23-1 and 23-2).

Sphenoethmoid Recess.  The sphenoid sinus ostium opens into the sphenoethmoid recess (SER). The SER lies just lateral to the posterosuperior nasal septum and behind the tail of the superior turbinate and is best appreciated on sagittal and axial CT images (see Fig. 23-3). The posterior ethmoid air cells drain into the superior meatus and from there into the SER (see Fig. 23-2).

IMPORTANT ANATOMIC VARIATIONS AND THEIR RELEVANCE Numerous variations in the sinonasal anatomy have been described. Many of them have important clinical and surgical relevance, and 

FS FB

SS

A

B FIG 23-3  A, Axial CT image shows the sphenoid sinus ostia (arrows). B, Sagittal image shows the sphenoid sinus ostium (arrow) leading into the sphenoethmoid recess (dashed line). Note the frontal bullar cell (FB) protruding into the frontal sinus (FS) posteriorly.

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses it is very important to identify them on the presurgical scans obtained in such patients.4,5,57,65 MDCT with its multiplanar reformations provides the best details of the complex bony anatomy and anatomic variants of the sinonasal cavities. It is extremely important to review the CT images in all three planes for better understanding of the anatomy and identification of these anatomic variations. If available, review of images on a workstation is extremely beneficial for reporting the scans.

Frontal Recess Cells

687

Frontal Cells (Kuhn’s Cells).  Four types of frontal cells are described.29,44,45 Types 1 to 3 frontal cells are commoner and located directly above the ANC. The type I cell is a single anterior ethmoid air cell seen above the ANC (see Fig. 23-2). Type II cells are two or more anterior ethmoid air cells above the ANC (see Fig. 23-3). The type III cell is a single large cell above the ANC that bulges into the frontal sinus. The type IV cell is an isolated air cell located completely within the frontal sinus, simulating a “cell within a cell” appearance; it is visualized infrequently (see Fig. 23-3).

These cells are variations in the pneumatization of the anterior ethmoid air cells. They are broadly divided into the anterior and posterior groups of frontal recess cells. The anterior group is located anterior to the frontal recess and comprises the agger nasi cells and the frontal cells. The posterior frontal cells are located posterior to the frontal recess and include the supraorbital cells, frontal bullar cells, and suprabullar cells; they are bordered posteriorly or superiorly by the base of the skull.29,36,44 These cells may obstruct frontal sinus drainage at the level of the frontal recess and contribute to the development of frontal sinusitis. Also, if these cells are not addressed during endoscopic surgery on the frontal recess, they often lead to surgical failure and recurrence of the disease process. Therefore it is very important for the radiologist to identify them and comment on their presence using the standard nomenclature as described below.

Supraorbital Ethmoid Cell.  This cell represents pneumatization of

Agger Nasi Cell.  The agger nasi cell (ANC) is a fairly constant ante-

Onodi Cell (Sphenoethmoid Cell)

riormost ethmoid cell seen in up to 98% patients at CT. It is located anterior to the vertical attachment of the middle turbinate and is best visualized on sagittal and coronal CT sections (Fig. 23-4; also see Fig. 23-2). The degree of pneumatization of the ANC varies and has a bearing on the size of the frontal sinus ostium and frontal recess.6

The Onodi cell is an anatomic variation of the most posterior ethmoid air cell when it extends into the sphenoid bone, superior or superolateral to the sphenoid sinus, pushing the sphenoid sinus inferiorly. The cell is intimately related to the optic nerve canal and can be mistaken for sphenoid sinus at endoscopy, increasing the risk of orbital

the orbital plate of the frontal bone posterior to the frontal recess and the frontal sinus.27,34,54 It simulates the appearance of septate or hyperpneumatized frontal sinuses on coronal images, and axial images confirm their location posterior and lateral to the frontal sinus (see Fig. 23-14).

Frontal Bullar Cell and Suprabullar Cell.  The ethmoid bulla forms the posterior boundary of the frontal recess. The frontal bullar cell, if present, lies above the ethmoid bulla and posterior to the frontal sinus and frontal recess.24,34,54 It extends into the frontal sinus and is best seen on sagittal images (see Fig. 23-3B). When the cell lies below the level of the frontal sinus ostium and does not extend into the frontal sinus, it is called a suprabullar cell (see Fig. 23-2).

FS FS

FS

A

A

A

FIG 23-4  Sagittal CT images through the frontal sinuses (FS) shows the types of frontal cells located atop the agger nasi cell (A). Type II frontal cells (arrowhead), type III frontal cell (curved arrow), type IV frontal cell (straight arrow).

http://radiologyebook.com

688

PART II  CT and MR Imaging of the Whole Body

A

A

A

A ON

SS SS

ON

SS

B

A

FIG 23-5  Coronal CT images at the level of the sphenoid sinus (SS). A, The hyperpneumatized SS, pneumatization of the anterior clinoid processes (A), type III optic canals (black arrows), foramen rotundum (white arrows), and vidian nerve canal (elbow arrows). Note the intersphenoid septum attaching to the right optic canal (white arrowhead). B, Bilateral Onodi cells (ON) with type IV optic canals (black arrows). Note the mucosal disease in the right sphenoid sinus (dashed black arrow).

complications. Hence it is of paramount importance to report their presence at CT examinations. Identification of this cell requires careful review of coronal, axial, and parasagittal CT images.27 Presence of a horizontal septum dividing the sphenoid sinuses into “two floors” or giving a cruciform appearance on coronal CT images suggests the presence of Onodi cells4,24,34 (Fig. 23-5).

Haller Cell (Orbitomaxillary Or Infraorbital Ethmoid Cell)

C

This is an extramural anterior ethmoid cell that pneumatizes inferior to the orbital floor, extending from the ethmoid labyrinth below the ethmoid bulla toward the interior of the maxillary sinus (Fig. 23-6). Although debate exists on whether or not their presence contributes to the occurrence of sinusitis, most agree that it is their number and size that influence the development of inflammatory sinus disease.7,67

Uncinate Process Variations

C

MT IT

FIG 23-6  Coronal CT image shows Haller cells on left side (arrow-

Three types of uncinate processes have been described, depending  on the superior attachment that impacts the drainage of the frontal sinus.34,54,57,62 Type I uncinate process is the commonest type and attaches to the lamina papyracea, separating the ethmoidal infundibulum and the frontal recess (Fig. 23-7). Type II uncinate attaches to the skull base, and type III turns medially and attaches to the middle turbinate (see Fig. 23-7). In type II and III uncinate processes, the frontal recess opens into the ethmoid infundibulum. The uncinate process may also be medialized or lateralized. An uncinate bulla may be seen.

Nasal Septal Variations The nasal septum may be deviated and make endoscopic access difficult. The deviation may be focal or more of a smooth curvature and is often associated with septal spurs and a concha bullosa or a lamellar concha (see Fig. 23-6). A nasal septal air cell may be seen in the posterosuperior portion of the nasal septum and may communicate with the sphenoid sinus.

heads), paradoxical curvature of the middle turbinates (MT), nasal septal deviation to the right side with a bony septal spur (curved arrow), and lamellar concha on either side (C). Hypertrophy of the left inferior turbinate (IT) is noted.

Middle Turbinate Variations The middle turbinate may be pneumatized. When pneumatization involves the bulbous portion of the middle turbinate, it is termed concha bullosa. If only the attachment portion of the middle turbinate is pneumatized, it is termed lamellar concha (see Fig. 23-6). A small concha may not be clinically significant, but a larger one—commonly associated with a nasal septal deviation—may obstruct the ethmoid infundibulum and lead to sinusitis.7 The convexity of the middle turbinate may be directed laterally and is called paradoxical curvature of the turbinate (see Fig. 23-6). A small paradoxical turbinate may not be significant, but a larger one may impair drainage at the OMU.7

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

689

Other Variants The maxillary sinus may show septations, resulting in inadequate drainage of sinus secretions. An accessory ostium may be present. The importance of this lies in the fact that an antrochoanal polyp may sometimes exit through the accessory ostium rather than the natural ostium. An aerated crista galli may communicate with the frontal recess, causing obstruction of the ostium and leading to chronic sinusitis.

INFLAMMATORY SINONASAL DISEASE—SINUSITIS

FIG 23-7  Coronal CT image shows the attachments of the uncinate processes (arrowheads). Type I uncinate on the left side attaches to lamina papyracea and the frontal recess (thin dotted line) opens into the middle meatus. Type III uncinate on the right side attaches to the middle turbinate, and the frontal recess (thick dotted line) opens into the ethmoid infundibulum.

Variations of the Ethmoid Roof Kero’s classification describes three types of ethmoid roofs, depending on the depth of the olfactory fossa, determined by measuring the height of the lateral lamella of the cribriform plate. Type I ethmoid roof has a depth of 1 to 3 mm, type II a depth of 4 to 7 mm, and type III a depth of 8 to 16 mm (see Fig. 23-1). Type III is associated with more potential damage from iatrogenic injury.50

Variations Related to Sphenoid Sinus Four types of optic nerve canals (OCs) have been described, depending upon the relationship of the OC with the posterior sinuses.12 Type I OC is the commonest type and runs immediately adjacent to the sphenoid sinus, without indentation of the wall or contact with the posterior ethmoid air cell. Type II OC courses like the type I but indents the sphenoid sinus wall. Type III OC runs through the sphenoid sinus, with at least 50% of the nerve being surrounded by air, and type IV OC lies immediately adjacent to the sphenoid sinus and the posterior ethmoid air cells (see Fig. 23-5). Anterior clinoid process pneumatization is associated with type II and type III OCs and predisposes the optic nerve to injury during FESS.12,35,60 The carotid artery canal may form a bulge into the sphenoid sinus wall. Sometimes the sinus wall separating the optic nerve or the carotid artery from the sinus may be dehiscent, or the intersphenoid sinus septum may be attached to the sinus wall covering them, and thus arterial or nerve injury may occur when the septum is avulsed during surgery (see Fig. 23-5A).

Sinusitis means inflammation of the sinus mucosa. Several forms are described based on the duration of symptoms. Acute sinusitis is almost always sudden in onset and may last up to 4 weeks. Subacute sinusitis is a continuation of the acute process and lasts anywhere between 4 and 12 weeks. Chronic sinusitis is defined as sinonasal inflammation that lasts for 12 weeks and beyond. The disease is termed recurrent acute sinusitis when there are more than four episodes of acute sinusitis. Acute sinusitis is most commonly an inflammatory response to an upper respiratory viral infection that causes sinonasal mucosal congestion and edema that may lead to sinus obstruction and bacterial proliferation. Chronic sinusitis develops secondary to acute disease that is refractory to treatment and is generally a sequelae to the interference of the normal mucociliary drainage of the paranasal sinuses. Imaging is neither routinely recommended nor performed for every patient who presents with sinusitis. The American Academy of Otolaryngology, in its 2007 clinical practice guidelines, recommended against diagnostic imaging for patients with acute or subacute sinusitis unless an intraorbital or intracranial complication is suspected. Imaging is recommended in patients who fail to respond to medical treatment, present with recurrent sinusitis, or have unilateral recurrent symptoms when an alternate diagnosis of neoplasia or fungal infection is suspected.41,47,53 CT is clearly superior to MRI for delineation of the bony anatomy and anatomic variants. CT shows the presence and extent of the  sinonasal disease, nature of the sinonasal secretions, and presence of any intrasinus calcifications. Hence CT is the preferred technique in preoperative evaluation of the paranasal sinuses and the accepted gold standard for delineation of inflammatory sinus disease due to obstruction. Coronal CT images almost simulate the appearance of the sinonasal cavities at endoscopy and are most preferred by endoscopic surgeons. Contrast CT or MRI have a role to play when there is suspicion of an intraorbital or intracranial extension of the disease.62,74

Imaging Findings

Acute Sinusitis.  CT and MRI may show nonspecific mucosal thickening, submucosal edema, air-fluid levels, or sinus secretions interspersed with air bubbles (Fig. 23-8). Acute sinonasal secretions are of a mucoid nature (−10 to 25 HU) and are typically hypointense on T1 and hyperintense on T2 sequences. An isolated air-fluid level as the only finding in the sinus is fairly characteristic for acute sinusitis but may not be seen in all patients.62 Complications of acute sinusitis include orbital cellulitis, orbital abscess, empyema, meningitis, brain abscess, and superior ophthalmic vein and/or cavernous sinus thrombosis. Although they can be visualized on contrast CT, intracranial complications and assessment of the orbital apex are best done on contrast MRI studies.

http://radiologyebook.com

690

PART II  CT and MR Imaging of the Whole Body

FIG 23-8  Axial CT image shows a fluid level in the left maxillary sinus (arrow) with presence of air bubbles. Acute left maxillary sinusitis.

FIG 23-10  Axial CT section in soft tissue algorithm shows complete opacification of the right maxillary sinus with sclerosis of the walls of the sinus. The sinus disease shows a thin rim of hypodensity bordering the mildly hyperdense inner contents. Chronic right maxillary sinusitis.

both the T1 and T2 signal intensity, mimicking a signal void.62,63,76 In such cases, sinus disease may be completely missed on MRI, while CT may show the sinus filled with chronic desiccated secretions. Other features of chronic sinusitis include retention cysts, polyps, and mucoceles.

Retention Cyst.  Two forms of retention cysts have been described:

FIG 23-9  Axial CT image shows sclerosis of the walls of the right maxillary sinus (white arrows) with presence of mucosal thickening (arrowheads). Chronic right maxillary sinusitis.

serous and mucous retention cysts. They are very common incidental findings. Serous cysts are due to accumulation of serous fluid in the submucosal layer of the sinus mucosa, whereas mucous forms result from obstruction of a submucous mucinous gland. They are most commonly found in the maxillary sinuses and cannot be differentiated from one another. They are both seen as mucoid low-density welldefined lesions on CT and low signal on T1-weighted and high signal on T2-weighted MRI sequences.

Polyp.  Polyps are the commonest expansile masses in the sinuses

Chronic Sinusitis Imaging features include nonspecific mucosal thickening, sinus opacification, intrasinus calcifications, and sclerosis (reactive osteitis) of the bony walls of the sinus (Figs. 23-9 and 23-10). Mucosal thickening is common to both acute and chronic forms of sinusitis. Presence of a sclerotic thickened bone is a fairly characteristic feature of chronic sinusitis but may not be seen in all patients.62 Calcifications are generally more peripheral and scattered.73 Air-fluid levels may be seen. Whereas secretions in acute sinusitis are watery, chronic secretions vary in their fluid and protein content, thereby affecting their CT and MR appearances. With chronicity, secretions lose water and become more proteinaceous and hence appear denser on CT. A thin hypodense line is usually seen separating the dense sinus secretions from the bony wall and represents thickened mucosa and submucosal edema (see Fig. 23-10).62,75 Progressive increase in the protein content of the secretions from 5% to about 25% results in an increase in the T1-weighted signal from low to high intensity. With increase in protein content beyond 25% to about 35%, the T2 signal begins to drop (Fig. 23-11). A very high protein content beyond 35% causes loss of almost all water from the secretions and results in a fall of

and may be solitary or multiple (Fig. 23-12). They are generally small but may grow and become very deforming. They result from accumulation of fluid in the deeper lamina propria of the sinus mucosa. Allergic polyps are generally multiple and occur most commonly in the ethmoid air cells. An antrochoanal polyp is usually unilateral, solitary, and seen in young adults. It is a large polyp in the maxillary sinus that expands and fills the antrum and prolapses through the primary or accessory ostium into the nasal cavity. The ostium and the infundibulum are widened. The polyp fills the nasal cavity and extends behind into the nasopharynx (see Fig. 23-12). Polyps and cysts cannot be differentiated on CT or MRI; however, it is of little consequence because they are both treated in the same manner. When multiple polyps are present, they may show varying densities on CT and have variable T1 and T2 signals due to the differing nature of secretions; this helps distinguish them from masses that have more homogeneous signals.

Mucocele.  Mucocele is the commonest expansile lesion of the paranasal sinuses, most commonly seen in the frontoethmoid sinuses. It develops as a result of obstruction of a sinus ostium or a compartment of a septate sinus and hence is lined by the sinus mucosa. CT shows

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

IS

A

691

IS

B FIG 23-11  A, Axial T1-weighted image with mild hyperintensity of the contents of the left maxillary sinus (IS). B, Corresponding T2-weighted image shows intrasinus hypointensity due to inspissated secretions (IS). Note the hyperintense mucosal thickening surrounding the inspissated secretions. Thin rim of mucosal disease is also seen in the right maxillary sinus.

B

A

FIG 23-12  Coronal CT (A) and axial CT (B) images show an antrochoanal polyp on the right side. Note the polyp prolapsing out of the sinus through the accessory ostium (arrow). A maxillary sinus polyp/retention cyst is seen in the left maxillary sinus (arrowhead).

an expanded airless sinus cavity with bony remodeling of the sinus walls, filled with fairly homogenous mucoid attenuation62,63 (Fig. 23-13). If the secretions are thick and proteinaceous, attenuation can increase to 25 to 40 HU. MRI typically shows low T1 and high T2 signals. If the secretions become proteinaceous, the T2 signal drops, with mild increase in T1 signals.

Patterns of Inflammatory Sinonasal Disease Sinonasal inflammatory disease has been categorized into five patterns for better understanding and interpretation and importantly to group the patients into nonsurgical (normal CT), routine surgical (I and II), and complex surgical groups (III and IV). Groups I to III are obstructive patterns and based on the major routes of mucociliary drainage that have been previously described.66 I. Infundibular pattern: obstruction is at the maxillary ostium and infundibulum and causes disease limited to the maxillary sinus (see Fig. 23-1). The remainder of the anterior OMU is patent, and therefore the anterior ethmoid and frontal sinuses are normal. Because the frontal sinus and anterior ethmoids may also drain

FIG 23-13  Axial CT image shows a mucocele arising from the left anterior ethmoid air cells (curved arrow).

http://radiologyebook.com

692

PART II  CT and MR Imaging of the Whole Body

SO

FS

SO SO SO C

A

B FIG 23-14  A, Coronal CT image depicting left-sided osteomeatal pattern of sinonasal inflammatory disease. Opacification of the middle meatus, frontal recess, ethmoid infundibulum, anterior ethmoid air cells, and supraorbital ethmoid air cell (SO) on the left side is noted. Lamellar concha (C) is seen on the right side. The supraorbital ethmoid air cell (SO) is seen posterior and lateral to the frontal sinus, better appreciated in axial CT image (B) at the level of the frontal sinuses (FS). Note mucosal disease in left frontal sinus and the supraorbital cell on this image.

into the infundibulum, the obstruction is presumed to be in the posteroinferior aspect of the infundibulum. II. OMU pattern: obstruction is at the level of the middle meatus, leading to inflammatory sinonasal disease within the ipsilateral maxillary, frontal, and anterior and middle ethmoid sinuses (Fig. 23-14). There may be variable amounts of disease in the affected sinuses owing to slight variations in the placement of the draining ostia. This has been specifically noted for the frontal recess that drains the frontal sinus, and hence the frontal sinus may be spared in the OMU pattern of disease. Similarly, isolated frontal sinus disease may also occur when the obstruction is limited to the anterior aspect of the middle meatus. III. Sphenoethmoid recess pattern: obstruction is at sphenoethmoid recess, causing inflammatory disease limited to the sphenoid sinus and posterior ethmoid air cells. Isolated sphenoid disease without posterior ethmoid sinus disease can be seen in the SER pattern because the sphenoid sinus drains directly into the SER and the posterior ethmoid initially drains more anteriorly into the superior meatus. IV. Sinonasal polyposis: characterized by polyps diffusely present within the nasal cavity and paranasal sinuses, filling the nasal vault and sinuses, causing bilateral infundibular enlargement, convex (bulging) ethmoid sinus walls, and attenuation of the bony nasal septum and ethmoid trabeculae (Fig. 23-15). Truncation of the bulbous portion of the middle turbinates may be seen.39 Isolated polyps are not included in this disease pattern. It is often associated with allergic fungal sinusitis. V. Sporadic pattern: diagnosed when inflammatory sinonasal disease is not attributable to obstruction of known mucous drainage routes or polyposis and there is randomly placed disease noted anywhere within the sinuses. This group includes individual inflammatory lesions such as retention cysts and mucoceles.

FUNGAL SINUSITIS Fungal infections of the paranasal sinuses are not uncommon, and there has been an increase in their incidence across the world over 

FIG 23-15  Coronal CT image shows diffuse opacification of the sinonasal cavities with widening of the osteomeatal units and deossification of the ethmoid trabeculae. Sinonasal polyposis.

the last decade. They are associated with high morbidity, and certain forms of these infections can be fatal. Although CT and MRI offer important diagnostic clues and depict certain characteristic features in most forms of these infections, the right diagnosis may still be elusive in some cases. A sound knowledge of the radiologic features, however, enables the radiologist to alert the surgeon about the diagnosis, prompting adequate sampling for fungal stains, cultures, and histopathologic examination. Etiologic fungi in most cases belong to the Aspergillus genus or the Phycomycetes group.9,20 Fungal sinusitis is broadly divided into invasive and noninvasive forms. The diagnosis of invasive fungal sinusitis is made when there is (a) invasion of the mucosa, submucosa, or blood vessels by the fungal hyphae or (b) tissue necrosis with minimal host inflammatory cell infiltration.20 Although the noninvasive forms occur almost exclusively in immunocompetent patients, they may progress to invasive forms if the immunologic status

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

A

693

B FIG 23-16  A, Axial CT section in a patient following renal transplant shows left maxillary sinus disease. Mild edema in the left retroantral fat (arrows) and a thin rim of retroantral soft tissue (arrowhead) are seen. B, Bone window shows subtle focal irregularity of the posterolateral wall of the left maxillary sinus. Fungal culture showed Aspergillus. Early acute invasive fungal sinusitis.

of the patient changes. These forms are further divided into five major subtypes based on their natural history and clinical and radiologic features.9

Invasive Fungal Sinusitis

Acute Invasive Fungal Sinusitis.  Acute invasive fungal sinusitis (AIFS) is the most dangerous and rapidly progressive form of fungal sinusitis, seen almost exclusively in individuals with poorly controlled diabetes mellitus or in immunosuppressed patients. The primary sites of infection are the nasal cavity and the middle turbinate, and a distinct predilection for unilateral involvement of the sphenoethmoid sinuses is noted.21,22 The fungi have a propensity to spread through the perivascular channels and hence cross the bony sinus walls through the penetrating vessels to reach the periantral soft tissues of the maxillofacial region, the orbit, and the cranial cavity. Early features of the disease on noncontrast CT include hypodense mucosal thickening and intranasal or intrasinus soft tissue, with bony erosion. The mucosal thickening is hypointense on T1-weighted and hyperintense on T2-weighted images. These findings may be subtle, and it can be very difficult to differentiate early AIFS from other nonspecific sinusitis at imaging. Presence of unilateral nasal cavity soft tissue thickening is the most consistent, though nonspecific, early CT finding.2,22 Early extrasinus spread is seen as obliteration or edema of the normal fat density in the periantral fat and within the orbit across normal-appearing sinus walls (Fig. 23-16). Deep spread into the periantral soft tissues, with destruction of sinus walls and intraorbital and intracranial extension are late, though fairly specific, findings in AIFS.13,37,43 Bone changes are best appreciated on CT, but intraorbital and intracranial extension are best delineated on contrast MRI. As the intraorbital disease spreads, there may be thickened extraocular muscles, thrombosis of the  superior ophthalmic vein, and even cavernous sinus thrombosis. Features of intracranial extension include leptomeningeal enhancement, granulomas, cerebritis, cerebral abscess formation, carotid artery invasion, pseudoaneurysm formation, and intracerebral infarct and hemorrhage.2,19,37,62

Chronic Invasive Fungal Sinusitis.  CIFS is a slowly progressive form of fungal infection that develops over months to years and 

is also associated with high morbidity. It is commonly seen in immunocompetent patients, but immunocompromised patients and  those with diabetes mellitus may also be affected. Patients generally present with history of chronic sinusitis. Noncontrast CT shows  a mildly hyperdense soft tissue mass in the paranasal sinus, sinus  wall erosion, and extension of the mass beyond the walls of the sinus.2,13,19,37,43,59,68 On MRI, the soft tissue is iso- to hypointense on T1 images and markedly hypointense on T2 sequences (Fig. 23-17). Invasion of the adjacent periantral region, the orbit, and anterior cranial fossa that are seen with AIFS may also be seen with the chronic form and are best appreciated on MRI. The differential diagnoses for intrasinus hyperdensity at CT include chronic inspissated secretions, fungal infection, intrasinus hemorrhage, and calcifications.62,63 Although the diagnosis of intrasinus hemorrhage is most often facilitated by the history and possible associated fractures, it may be almost impossible to differentiate chronic inspissated inflammatory sinus disease from fungal disease in the absence of bone erosion at imaging. Presence of mottled irregular bone erosion is perhaps the only sign that favors the diagnosis of an aggressive fungal infection over chronic inspissated secretions.18 Although intrasinus hyperdensity at CT almost completely excludes the diagnosis of a sinonasal malignancy, at times the differentiation of CIFS from sinonasal malignancy may be difficult.63 At such times, the markedly hypointense signal of CIFS on T2-weighted images is a differentiating feature from a sinonasal malignant neoplastic mass, which shows an intermediate signal on T2 images63 (Fig. 23-18).

Chronic Invasive Granulomatous Sinusitis.  This is a rare condition, very similar clinically and radiologically to CIFS and occurring almost exclusively in immunocompetent patients from North Africa and South America.9,20

Noninvasive Fungal Sinusitis

Allergic Fungal Sinusitis.  Allergic fungal sinusitis (AFS) is the commonest form of fungal sinusitis. Unlike the other forms, it is not an infection but postulated to be a hypersensitivity reaction to inhaled fungal organisms, resulting in a chronic inflammatory  disease process. It is commonly seen in warm humid climates  and affects younger immunocompetent individuals.9,20,58 Patients

http://radiologyebook.com

694

PART II  CT and MR Imaging of the Whole Body

A

B FIG 23-17  A, Axial CT section in a patient with chronic rhinosinusitis with symptoms of right-sided orbital apex syndrome shows patchy opacification of the right sphenoethmoid sinuses with erosion of the right lamina papyracea and lateral wall of the right sphenoid sinus. Abnormal soft tissue is seen in the region of the right orbit, orbital apex, and cavernous sinus. B, T2-weighted MRI section at the same level. Mucosal disease is seen in the right sphenoethmoid sinus. Note the hypointense soft tissue in the right orbit, orbital apex, and right cavernous sinus (arrowheads). Note also hypointense soft tissue in right anterior ethmoid air cells (curved arrow). Chronic invasive fungal sinusitis.

E

IS

IS

A

B FIG 23-18  A, Coronal CT reformatted section shows hyperdense soft tissue in the nasoethmoid region, with erosion of the anterior skull base and fairly large intracranial extension (arrowheads) and edema in the bifrontal lobes (E). Intrasinus hyperdensity is also seen in the left maxillary sinus. B, Coronal T2-weighted MRI shows the soft tissue with profoundly hypointense signal (arrowheads) and mild edema in the bifrontal lobes. Hypointense signal in left maxillary sinus is due to chronic inspissated secretions (IS); this shows a hyperintense signal on T1-weighted images not shown here.

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

A

695

B FIG 23-19  A, Axial CT image shows bilateral polypoid expansile disease with marked hyperdensity in the ethmoid sinuses bilaterally and left sphenoid sinus. Note erosion of the lamina papyracea with intraorbital extension of the disease. B, Disease in the maxillary sinuses. Note erosion of the posterolateral wall of left maxillary sinus. Allergic fungal sinusitis.

generally have a history of atopy, chronic sinusitis, and polyposis.14,49 The diagnosis of AFS may often be first suggested at CT, and the radiologist plays a crucial role in alerting the physician to the  possibility of this entity. Presence of allergic mucin at endoscopy  is characteristic of the disease. The disease may be unilateral or  bilateral but asymmetric and shows near-complete opacification of multiple sinuses with polypoid masses (Fig. 23-19). Presence of allergic mucin produces the characteristic intrasinus hyperdensity in a background of hypodense polypoid mucosal disease, giving rise to the “double-density sign.”37,43,49 This hyperdensity is best appreciated on soft tissue algorithm images and is due to the combination of heavy metals (iron and manganese), calcium, high protein, and low water content of the allergic mucin.56 There may be sinus expansion encroaching onto the adjacent orbits and erosion of the walls, with associated extrasinus extension into the orbit and the cranial cavity (see Fig. 23-19). Presence of sinus expansion with thinning of the walls instead of reactive sclerosis differentiates AFS from chronic sinusitis with desiccated secretions.49,56 Whereas T1 intensity of allergic fungal sinusitis is variable, the T2 signal is very low owing to metals concentrated by the fungal organisms, high protein, and  low water content of the mucin and may lead to underdiagnosis; hence CT is best suited for the diagnosis. Contrast study shows enhancement of the peripheral mucosa with central areas of nonenhancement.

Fungus Ball (Mycetoma).  This relatively uncommon type of fungal sinusitis is a dense extramucosal conglomeration of fungal hyphae without any allergic mucin, found in immunocompetent patients in older age groups. It generally occurs in a single sinus cavity, the commonest sinus being the maxillary sinus, followed by the sphenoid sinus.2,19,37 Noncontrast CT shows a hyperdense mass due to the dense matted fungal hyphae within the involved sinus, with occasional intrasinus calcifications (Fig. 23-20). Most of these calcifications occur centrally within the sinus.73 Surrounding hypodensity due to mucosal thickening is usually present, and there may be reactive osteitis due to chronic sinusitis. No bony erosion is seen. The fungus ball is hypointense on T1- and T2-weighted images owing to the absence of free water, calcifications, and paramagnetic metals such as iron, magnesium, and manganese.

FIG 23-20  Axial CT image shows thickened walls of the right maxillary sinus with intrasinus central hyperdensity and calcifications. Mycetoma was found at histopathology.

GRANULOMATOUS DISEASES OF THE SINONASAL CAVITIES Granulomatous diseases can be of an infective or noninfective nature. Infectious causes of granulomatoses include actinomycosis, tuberculosis, leprosy, and syphilis. Wegener’s granulomatosis and sarcoidosis are noninfectious granulomatous processes. Wegener’s granulomatosis is a systemic vasculitis that can affect any organ system but primarily involves the upper and lower respiratory tracts and kidneys. Imaging findings in Wegener’s granulomatosis can be broadly divided into mucosal and bony changes. Mucosal lesions range from nonspecific mucosal thickening and sinus opacification to granuloma formation that are low signal on T1 and intermediate to high signal on T2-weighted sequences.48 Bony changes include erosion, mainly involving the septum, turbinates, and the medial wall of the maxillary sinuses, followed by sclerosing osteitis of the sinus wall. The combination of bone erosion, remodeling, and sclerosis gives rise to

http://radiologyebook.com

696

PART II  CT and MR Imaging of the Whole Body BOX 23-1  Classification of Paranasal

Sinus Neoplasms25 Epithelial Tumors • Papilloma • Squamous cell carcinoma • Adenocarcinoma • Adenoid cystic carcinoma • Metastases

FIG 23-21  Coronal CT image shows reduced caliber of the sinus, mildly thickened sinus walls, opacification of the sinus and the ethmoid infundibulum (arrow), with inward retraction of the walls of the right maxillary sinus (arrowheads). Note the increase in right orbital volume. Silent sinus syndrome.

the tramline effect in the antrum and may eventually lead to complete bony obliteration of these sinuses.40,72 Although these findings are suggestive of Wegener’s granulomatosis, they are not pathognomonic of the disease and should be combined with clinical and laboratory studies to confirm the diagnosis.

SILENT SINUS SYNDROME This is a unilateral atelectasis of the maxillary sinus, with partial or complete opacification of the sinus; it is seen in adults. Patients generally present with progressive enophthalmos and facial asymmetry. Although most cases are idiopathic, trauma to the lateral nasal wall may be the cause in some patients. The primary imaging finding is reduced caliber of a fully formed maxillary sinus due to inward retraction of the sinus walls, occlusion of the maxillary ostium and infundibulum, and sinus opacification and reactive osteitis of the sinus walls due to chronic inflammation (Fig. 23-21).28,31 The orbital floor is almost always retracted and a resultant increase in the orbital volume is noted. Chronic occlusion of the maxillary sinus ostium results in gradual resorption of the air, with subsequent generation of negative pressure that causes gradual inward bowing of the sinus wall.

TUMORS OF THE PARANASAL SINUSES Sinonasal tumors are rare neoplasms and account for less than  3% of head and neck cancers. A variety of sinonasal tumors have been described and have an overall poor prognosis. There are differences in the age at presentation of the affected patients; sinonasal carcinomas are generally seen in the elderly population, but some tumors like lymphomas, olfactory neuroblastomas, and minor salivary gland tumors also present in younger patients.3 The sinonasal tumors are broadly classified into epithelial and nonepithelial types (Box 23-1). Patients with sinonasal masses often present late and with nonspecific symptoms that may be attributed to those of chronic sinusitis. Hence a plain CT scan is generally the first test performed  on these patients. Features at CT that should alert the radiologist to

Nonepithelial Tumors Neuroectodermal and Nervous System Tumors • Malignant melanoma • Nasal glioma • Olfactory neuroblastoma • Neuroendocrine tumor • Peripheral nerve sheath tumor • Ewing’s sarcoma and PNET • Meningioma Mesenchymal Tumors • Juvenile nasopharyngeal angiofibroma • Angiomatous polyp • Hemangioma • Hemangiopericytoma Osseous and Cartilaginous Tumors • Osteosarcoma • Chondrosarcoma • Soft tissue sarcoma • Rhabdomyosarcoma Lymphoreticular Tumors • Lymphoma • Plasmacytoma Fibroosseous Tumors • Fibrous dysplasia • Ossifying fibroma • Cementifying fibroma • Cemento-ossifying fibroma • Osteoma PNET, primitive neuroectodermal tumor.

the presence of a possible neoplastic mass lesion include persistent unilateral sinus opacification, extrasinus disease, bony remodeling, and bone destruction. Whereas benign masses expand the sinus cavity and may cause bone remodeling and pressure erosion, local bone destruction is characteristic of malignant tumors (see Figs. 23-23D and Fig. 23-24).61 Benign processes generally have a smooth polypoid margin, and even when they extend intracranially, they usually retain this contoured appearance. Malignant masses, however, frequently show irregular and nodular margins.10 Erosion of the nasoethmoid roof, posterolateral wall of the maxillary sinus, or the lamina papyracea are all very well appreciated on CT.16,25 Often contrast MDCT optimally shows the tumor extension, with changes of bony remodeling or destruction and presence of intratumoral calcifications; it also delineates the tumor from the nonenhancing sinonasal secretions. MRI delineates the soft tissue extent and early extrasinus spread into the orbit, skull base, nasopharynx, and intracranial cavity with

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

697

of the tumor alongs the second division of the fifth cranial nerve,  better depicted at MRI than at CT. Spread into the dura and brain parenchyma is better seen on a contrast MRI. Linear dural enhancement is not specific for dural invasion and is generally due to reactive changes. Signs of dural involvement by malignant tumors include focal nodularity, dural thickening of more than 5 mm, and pial enhancement.17 There may also be edema in the adjacent brain parenchyma (see Fig. 23-25). Brain invasion is seen as enhancement of the brain parenchyma.

Papilloma

FIG 23-22  Coronal T2-weighted MRI section shows a mass lesion centered in the middle meatus, lateral nasal wall, and maxillary antrum, with a convoluted cerebriform pattern. Note superb delineation of the mass from the surrounding sinonasal secretions. Inverting papilloma.

greater precision than contrast CT. MRI is also preferred over CT to detect perineural spread. Another major advantage of MRI is better differentiation of tumor from the coexistent inflammatory sinonasal disease. Malignant masses are highly cellular and have very little free water, and hence they show an intermediate signal on T2-weighted MR scans (see Fig. 23-23B).10,61 The more benign and glandular-type tumors such as papillomas and minor salivary gland tumors usually have sufficient water to produce hyperintensity on T2-weighted MRIs (Fig. 23-22).42,63 It is important to note that although the T2 signal of the mass is intermediate, it is still higher than the T1 signal, known as tumor brightening on T2-weighted scans (Fig. 23-23A and B).10,63,61 A T2 signal in a mass that is lower than the T1 signal almost excludes a sinonasal tumor.63 The intermediate T2 signal intensity also allows separation of the tumor mass from high-intensity inflammatory changes (see Fig. 23-23B). Specific histologic diagnosis is not possible using MRI, because the observed signal intensities in different tumors show substantial overlap.75 The neck nodes are not routinely assessed on MRI of the sinuses, but when spread to the pharynx or oral cavity is seen, the deep cervical nodes should be evaluated. The presence of metastatic lymph nodes signifies poor prognosis.70 The primary role of imaging in sinonasal neoplasia is to delineate the exact extent of the tumor, and CT and MRI often complement each other in most aspects. Sinonasal malignancies often show spread into the adjoining orbit, skull base, cranial cavity, pterygopalatine fossa, and the masticator space. Spread into the anterior skull base and intracranial cavity is common in patients with nasoethmoid cancers. Invasion of the infratemporal fossa, pterygopalatine fossa, and the masticator space are commonly seen with maxillary sinus cancers (see Fig. 23-23). Orbital invasion may occur with both maxillary and ethmoid tumors. The pterygopalatine fossa may get involved also by perineural spread

Papillomas are uncommon sinonasal masses. Three histologic varieties of papillomas are described: fungiform, inverted, and oncocytic, collectively called schneiderian papillomas.69 Fungiform (or exophytic) papillomas arise from the nasal septum and are solitary and unilateral. These have a verrucous appearance and are not considered premalignant. Inverted (or endophytic) papillomas arise from the lateral nasal wall in the region of the middle turbinate and extend into the sinuses. The tumor shows a fairly strong male predilection. Although this is  a benign tumor, it may behave aggressively, may commonly show recurrence after resection, and in 7% to 15% of patients squamous  cell carcinoma (SCC) may coexist with the tumor.30,51 It derives its name from the characteristic invagination of the epithelium seen on histopathology. The anatomic location and its extent may be seen fairly well on CT. Sclerosis of the sinus wall may be seen at times. MRI  features may be nonspecific in smaller masses, but larger masses show a convoluted “cerebriform” pattern on T2-weighted and contrastenhanced T1-weighted images (see Fig. 23-22). This pattern consists of low-signal-intensity and relatively high-signal-intensity striations on T2-weighted images that probably represent metaplastic squamous epithelium and edematous stroma, respectively.32,55,74 Contrastenhanced T1-weighted images show well-enhancing stroma and less-enhancing epithelium that create a convoluted cerebriform pattern. Oncocytic (or cylindrical) papillomas have imaging features similar to the inverted papilloma variety.

Carcinoma SCC is the commonest malignant tumor of the sinonasal cavity, constituting about 60% of sinonasal tumors, followed by adenocarcinoma and adenoid cystic carcinoma. An increased risk of developing adenocarcinoma has been seen in woodworkers.3 Although the maxillary sinus is the commonest site for occurrence of SCC, adenocarcinomas are seen more commonly in the ethmoid sinuses and the superior part of the nasal cavity. There are no specific features, and MRI shows an intermediate-signal-intensity mass on T1 and T2 sequences, with bone destruction. Adenoid cystic carcinoma (ACC) is the commonest minor salivary gland tumor seen in the sinonasal cavities, most commonly found in the maxillary sinus, followed by the nasal cavity. However, more commonly ACC arises in the palate and secondarily involves the sinonasal cavities. Unlike SCC, ACC is a slow-growing neoplasm. It is a locally aggressive tumor with a high propensity for local recurrence, perineural spread, and distant metastases.23

Malignant Melanoma These rare tumors arise from the melanocyte precursors present in the sinonasal mucosa. They have a poor prognosis, a high rate of recurrence, and also a high incidence of distal metastases. Most arise in the nasal cavity, followed by the maxillary sinus.11 Imaging shows a fairly well-defined polypoid nasal or intrasinus mass. Bone remodeling is generally more common than bone destruction. A characteristic T1 hyperintensity due to melanin or hemorrhage is seen with an intermediate signal on T2-weighted MR scans.

http://radiologyebook.com

698

PART II  CT and MR Imaging of the Whole Body

B

A

D

C FIG 23-23  Axial T1 (A) and T2 (B) sections show a large left maxillary sinus mass with extension into the periantral soft tissues (arrowheads). Posteriorly the mass extends into the retroantral fat and left masticator space (curved white arrow). Note the intermediate signal of the mass on T2-weighted images. Note the normal retroantral fat (double-headed white arrow) and masticator space (dotted white lines) posterior to the right maxillary sinus; the pterygoid plates are better seen medially on axial CT. C, Contrast-enhanced T1-weighted MRI with fat suppression shows mild extension of the mass into the floor of the left orbit. D, Axial CT shows significant destruction of the walls of the left maxillary sinus (black arrows). Squamous cell carcinoma of the left maxillary sinus. Chronic right-sided maxillary sinusitis is seen.

Schwannoma

Olfactory Neuroblastoma

This is an infrequent tumor of the sinonasal cavity that arises from the first or second division of the fifth cranial nerve. On contrast CT, the lesion is seen as a moderately enhancing fairly well-defined mass that frequently remodels the bone as it grows (Fig. 23-24). On MRI the mass is hyperintense on T2 images owing to a fairly good amount of free water within it. Necrosis and cystic changes may lead to an inhomogeneous appearance.15 Presence of aggressive bone destruction may be considered an indication of malignant transformation.

This is a rare tumor of neuroectodermal origin that arises from the olfactory epithelium in the roof of the ethmoidal sinuses, in the  cribriform region, upper part of the nasal septum, and superior turbinates, generally seen in adults between 20 and 40 years of age. Imaging shows an enhancing soft tissue mass, with or without bone destruction, centered in the nasal vault, with or without extension into the ethmoid sinuses, orbit, and cranial cavity, depending on the stage of the disease. Calcification may be seen on CT. Intracranial extension is

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses

FIG 23-24  Contrast-enhanced axial CT image shows a mildly enhancing expansile mass in the right nasal cavity causing remodeling of the walls of the nasal cavity. Intranasal schwannoma. Chronic right maxillary sinusitis with inspissated secretions is noted.

699

FIG 23-26  Axial contrast CT image in a 12-year-old boy with epistaxis shows moderately enhancing mass centered in the left sphenopalatine foramen, extending into the nasal cavity, nasopharynx, pterygopalatine fossa, retromaxillary region, and masticator space, eroding the pterygoid plates and posterolateral wall of the left maxillary sinus with intrasinus extension. Juvenile nasopharyngeal angiofibroma. Note mucosal disease in right maxillary sinus.

E paranasal sinuses. Stage C includes tumors with orbital, skull base, or intracranial spread and distant metastases.

Juvenile Nasopharyngeal Angiofibroma Juvenile nasopharyngeal angiofibroma is a benign but locally aggressive tumor with a very rich vascular supply from the internal maxillary artery and the ascending pharyngeal artery. It occurs almost exclusively in male adolescents, and patients generally present with epistaxis. The site of origin is the sphenopalatine foramen and pterygopalatine fossa, and the sinonasal space is involved secondarily. They frequently grow to very large sizes and spread into adjacent regions along the skull base and superiorly into the cranial cavity. In the early stages, there may be anterior bowing of the posterior wall of the maxillary sinus and widening of the pterygopalatine fossa; frank bone destruction may be seen as the tumor grows. Contrast CT shows an intensely enhancing mass with bone destruction in the characteristic location and the characteristic age group (Fig. 23-26). The arterial phase of contrast injection shows the hypervascular nature of the mass. On MRI, heterogeneous signals are seen on T1- and T2-weighted scans, with presence of flow voids. Intense enhancement is seen following administration of contrast. FIG 23-25  Coronal T1 contrast MRI shows an enhancing mass centered in the superior nasal cavity, eroding the anterior skull base and extending into the cranial cavity. Note the cysts at the growing margin of the mass lesion (arrowhead). Mild edema of the bifrontal lobes is seen (E). Note mucosal disease in both the maxillary sinuses.

often dumbbell shaped, the narrow part being in the region of the cribriform plate. Cysts are often present at the interface between the brain and tumor64 (Fig. 23-25). These tumors are clinically staged using the Kadish staging system.33 Stage A includes tumors limited to the nasal cavity, stage B includes tumors limited to the nasal cavity and

Lymphoma Paranasal sinuses are an uncommon extranodal site for lymphomas. There are usually two types of lymphomas that involve the paranasal sinuses, the B-cell type and the T-cell type. The T-cell type is more prevalent in Asian and South American populations. The T-cell type usually occurs in the nasal cavity and can be very aggressive, with destructive changes.15 Many of the sinonasal lymphomas are EpsteinBarr virus associated. The destructive “lethal midline granuloma” described in the literature has been proven to be the T-cell variety of lymphoma involving the sinonasal region. B-cell lymphomas occur in the sinuses and are indolent in their behavior. They are more common in Western populations. The lesions appear as bulky soft tissue masses. There is bone remodeling, with the maxillary sinus being commonly

http://radiologyebook.com

700

PART II  CT and MR Imaging of the Whole Body

FIG 23-27  Coronal CT image shows smooth expansion with ground-

FIG 23-28  Axial CT shows fracture of the pterygoid plates bilaterally,

glass appearance of the walls of the ethmoid labyrinth on the right side and the right middle turbinate. Fibrous dysplasia.

posterolateral wall of both maxillary sinuses, anterior wall of right maxillary sinus, and the right zygomatic arch.

involved by B-cell disease.26 On MRI, the lesions are of intermediate signal and show mild enhancement. Unlike carcinomas, these lesions do not show invasion toward the skull base.

keratocyst is seen as a multiloculated cystic mass with scalloped borders that may bulge into the maxillary sinus. Most odontogenic tumors are common in the mandible but may occur in the maxilla. Ameloblastoma is a benign epithelial odontogenic tumor that may be seen in the maxilla. It is a well-defined unilocular or multilocular expansile cystic lesion with interspersed isodense solid areas that causes moderate expansion, thinning, and often destruction of the maxilla. It may be difficult to differentiate an ameloblastoma from an odontogenic keratocyst.

Fibroosseous Lesions Fibroosseous lesions are probably the most common noninflammatory benign masses of the sinonasal region. This group comprises masses wherein the normal bone is replaced by benign fibrous tissue containing various amounts of mineralization/calcification.1,8,46 All of them are very well appreciated on CT. Fibrous dysplasia commonly affects the maxilla and mandible in the craniofacial region and is  generally seen in children and adolescents. The medullary bone is expanded, with a hazy ground-glass density compared with the normal compact bone; the cortex is intact and the margins may be poorly defined1,8 (Fig. 23-27). Ossifying fibroma is an expansile mass lesion that is more commonly seen in the mandible, followed by the maxilla, in close proximity to the roots of the teeth. A juvenile form of the tumor is seen in adolescents, but the tumor is otherwise seen in young adults. CT shows a well-circumscribed expansile osteolytic lesion with homogenous fibrous matrix that may have a sclerotic border and a few discrete foci of osseous matrix. Osteoma is generally an incidental finding and is the commonest osseous tumor in the maxillofacial region. It is frequently seen in the frontoethmoid sinuses. It appears as a characteristic sharp, well-delineated sclerotic lesion attached by a broad base or pedicle to the bone. It produces a complete signal void on all MR sequences and may often be overlooked owing to the background of sinus air.

Odontogenic Cysts and Tumors Odontogenic cysts and tumors that arise from the maxillary alveolar ridge may extend into the maxillary sinus and present as sinus masses.71 All of them are very well delineated at CT. The odontogenic cysts of significance to the sinonasal region include the radicular cyst, dentigerous cyst, and odontogenic keratocyst. The radicular cyst is seen in relation to a carious tooth and appears as a well-defined cystic lesion bordered by a thin rim of cortical bone. Dentigerous cysts and odontogenic keratocysts are more common in the mandible. A dentigerous cyst is seen as a well-defined unilocular or multilocular cyst in relation to the crown of an unerupted tooth; the crown of the tooth projects into the cystic cavity. An odontogenic

Metastases The commonest cancer to metastasize to the paranasal sinuses is renal cell carcinoma, followed by lung and breast.38 Most metastases are to the bony walls.

MAXILLOFACIAL TRAUMA MDCT is the modality of choice for evaluation of maxillofacial trauma.52 For ease of understanding and interpretation, the facial anatomy is divided into upper, middle, and lower thirds. The upper third of the face consists of the frontal bone (including the frontal sinuses). The middle third of the face extends from the superior orbital rims inferiorly to the maxilla and thus includes the orbits, nasal cavity, and the maxillary, ethmoid, and sphenoid sinuses. The middle third of the face is bounded posterolaterally by the zygomaticotemporal sutures, which connect the midface to the calvaria, and posteromedially by the pterygoid plates, which connect it to the skull base. The lower third of the face consists of the mandible. Describing fractures by their location with regard to these facial thirds may be helpful for planning surgical access. Bony injury may involve the nasal bones or the bony walls of the paranasal sinuses, and various combinations may be noted (Fig. 23-28). In addition to the bony injury, it is important to comment on the presence of intrasinus hemorrhage and intrasinus displacement  of fracture fragments. Assessment of the orbital walls, orbital apex,  and optic nerve canal and the ethmoid and sphenoid roof for risk of cerebrospinal fluid rhinorrhea and the visualized cranial cavity are of immense importance. Also of importance is the identification of Le Fort fractures that are associated with disruption of the pterygoid plate (see Fig. 23-28). Because of the close proximity and associated

http://radiologyebook.com

CHAPTER 23  Paranasal Sinuses disruption of the pterygoid venous plexus, it is important to recognize these fractures because of fairly high risk of large nasopharyngeal hematomas.

REFERENCES 1. Araghi HM, Haery C: Fibro-osseus lesions of cranio-facial bones: Role of imaging. Radiol Clin North Am 31:121–134, 1993. 2. Aribandi M, McCoy VA, Bazan C, III: Imaging features of invasive and noninvasive fungal sinusitis: A review. Radiographics 27:1283–1296, 2007. 3. Barnes L, Brandwein M, Som PM: Diseases of the nasal cavity, paranasal sinuses and nasopharynx. In Barnes L, editor: Surgical Pathology of the Head and Neck, ed 2, New York, 2003, Marcel Dekker, pp 509–517. 4. Beale TJ, Madani G, Morley SJ: Imaging of the paranasal sinuses and nasal cavity: Normal anatomy and clinically relevant anatomical variants. Semin Ultrasound CT MR 30:2–16, 2009. 5. Bolger WE: Anatomy of the paranasal sinuses. In Kennedy DW, Bolger WE, Zinreich J, editors: Diseases of the Sinuses, Diagnosis and Management, New York, 2001, Marcel Dekker. 6. Bruner E, Jacobs JB, Lebowitz RA: Role of the agger nasi cell in chronic frontal sinusitis. Ann Otol Rhinol Laryngol 105:694–700, 1996. 7. Caughey RJ, Jameson MJ, Gross CW, et al: Anatomic risk factors for sinus disease: Fact or fiction? Am J Rhinol 19(4):334–339, 2005. 8. Celenk P, Zengin Z, Muglali M, et al: Computed tomography of cranio-facial fibrous dysplasia. Eur J Radiol Extra 69:e85–e87, 2009. 9. Chakrabarti A, Denning DW, Ferguson BJ, et al: Fungal rhinosinusitis: A categorisation and definitional schema addressing current controversies. Laryngoscope 119:1809–1818, 2009. 10. Chow JM, Leonetti JP, Mafee MF: Epithelial tumors of the paranasal sinuses and nasal cavity. Radiol Clin North Am 31:61–73, 1993. 11. Conley JJ: Melanomas of the mucous membrane of the head and neck. Laryngoscope 99:1248–1254, 1989. 12. DeLano MC, Fun FY, Zinreich SJ: Optic nerve relationship to the posterior paranasal sinuses. CT Anatomic study. AJNR Am J Neuroradiol 17:669–675, 1996. 13. Delgaudio JM, Swain RE, Jr, Kingdom TT: CT findings in patients with invasive fungal sinusitis. Arch Otolaryngol Head Neck Surg 129:236–240, 2003. 14. DeShazo RD, Swain RE: Diagnostic criteria for allergic fungal sinusitis.   J Allergy Clin Immunol 96(1):24–35, 1995. 15. Dublin AB, Dedo HH, Bridger WH: Intranasal schwannoma: Magnetic resonance and computed tomography appearance. Am J Otolaryngol 16:251–254, 1995. 16. Eisen MD, Yousem DM, Loevner LA, et al: Preoperative imaging to predict orbital invasion by tumor. Head Neck 22:456–462, 2000. 17. Eisen MD, Yousem DM, Montone KT, et al: Use of preoperative MR to predict dural, perineural, and venous sinus invasion of skull base tumors. AJNR Am J Neuroradiol 17:1937–1945, 1996. 18. Fatterpekar G, Delman BM, Som PM: Imaging the paranasal sinuses– Where are we and where we are going. Anat Rec (Hoboken) 291:1564– 1572, 2008. 19. Fatterpekar G, Mukherji S, Arbealez A, et al: Fungal diseases of the paranasal sinuses. Semin Ultrasound CT MR 20:391–401, 1999. 20. Ferguson BJ: Definition of fungal rhinosinusitis. Otolaryngol Clin North Am 33(2):389–398, 2000. 21. Gillespie MB, Huchton DM, O’Malley BW: Role of middle turbinate biopsy in the diagnosis of fulminant invasive fungal rhinosinusitis. Laryngoscope 110:1832–1836, 2000. 22. Gillespie MB, O’Malley BW, Jr, Francis HW: An approach to fulminant invasive fungal rhinosinusitis in the immunocompromised host. Arch Otolaryngol Head Neck Surg 124(5):520–526, 1998. 23. Ginsberg LE: Imaging of perineural tumor spread in head and neck cancer. Semin Ultrasound CT MR 20(3):175–186, 1999. 24. Goncalves G, Jovem CL, Moura LO: Computed tomography of intra and extramural ethmoid cells: Iconographic essay. Radiol Bras 44(5):321–326, 2011.

701

25. Hermans R: Neoplasms of the sinonasal cavities. Head and Neck Cancer Imaging. Robert Hermans, 2006, Springer, pp 191–218. 26. Hermans R, Horvath M, De Schrijver T, et al: Extranodal non-Hodgkin lymphoma of the head and neck. J Belge Radiol 77:72–77, 1994. 27. Hoang JK, Eastwood JD, Tebbit CL, et al: Multiplanar sinus CT: A systematic approach to imaging before functional sinus surgery. AJR Am J Roentgenol 194(6):W527–W536, 2010. 28. Hourany R, Aygun N, Santina CCD, et al: Silent sinus syndrome: An acquired condition. AJNR Am J Neuroradiol 26:2390–2392, 2005. 29. Huang BY, Lloyd KM, Delgaudio JM, et al: Failed endoscopic sinus surgery: Spectrum of CT findings in the frontal recess. Radiographics 29:177–195, 2009. 30. Hyams V: Papillomas of the nasal cavity and paranasal sinuses: A clinicopathologic study of 315 cases. Ann Otol Rhinol Laryngol 80:192–206, 1971. 31. Illner A, Davidson HC, Harnsberger HR, et al: The silent sinus syndrome: Clinical and radiographic findings. AJR Am J Roentgenol 178:503–506, 2002. 32. Jeon TY, Kim HJ, Chung SK, et al: Sinonasal inverted papilloma: Value of convoluted cerebriform pattern on MR imaging. AJNR Am J Neuroradiol 29:1556–1560, 2008. 33. Kadish S, Goodman M, Wang CC: Olfactory neuroblastoma: A clinical analysis of 17 cases. Cancer 37:1571–1576, 1976. 34. Kantarci M, Karasen RM, Alper F, et al: Remarkable anatomic variations in paranasal sinus region and their clinical importance. Eur J Radiol 50(3):296–302, 2004. 35. Kazkayasi M, Karadeniz Y, Aarikan OK: Anatomic variations of sphenoid sinus on CT. Rhinology 43(2):109–114, 2005. 36. Kew J, Rees GL, Close D, et al: Multiplanar reconstructed computed tomography images improve depiction and understanding of the anatomy of frontal sinus and recess. Am J Rhinol 16(2):119–123, 2002. 37. Khattar VS, Hathiram BT: Radiologic appearances of fungal sinusitis. Otorhinolaryngology clinics: An international journal. 1(1):15–23, 2009. 38. Lee HM, Kang HJ, Lee SH: Metastatic renal cell carcinoma presenting as epistaxis. Eur Arch Otorhinolaryngol 262:69–71, 2005. 39. Liang EY, Lam WW, Woo JK, et al: Another CT sign of sinonasal polyposis: Truncation of the bony middle turbinate. Eur Radiol 6(4):553–556, 1996. 40. Lohrmann C, Uhl M, Warnatz K, et al: Sinonasal computed tomography in patients with Wegener’s granulomatosis. J Comput Assist Tomogr 30:122–125, 2006. 41. Madani G, Beale TJ: Sinonasal inflammatory disease. Semin Ultrasound CT MR 30:17–24, 2009. 42. Mafee MF: Nonepithelial tumors of paranasal sinuses and nasal cavity. Radiol Clin North Am 31:61-75-90, 1993. 43. Mahmud MB, Ilica TA, Maluf F, et al: The many faces of fungal disease of the paranasal sinuses : CT and MR findings. Diagn Interv Radiol 19:195–200, 2013. 44. McLaughlin RB, Jr, Rehl RM, Lanza DC: Clinically relevant frontal   sinus anatomy and physiology. Otolaryngol Clin North Am 34(1):1–22, 2001. 45. Meyer Tk, Kocak M, Smith MM, et al: Coronal CT analysis of frontal cells. Am J Rhinol 17:163–168, 2003. 46. Mohammadi HA, Haery C: Fibro-osseus lesions of cranio-facial bones—Role of imaging. Radiol Clin North Am 31:121–134, 1993. 47. Momeni AK, Roberts CC, Chew FS: Imaging of chronic and exotic sinonasal disease : A review. AJR Am J Roentgenol 189:S35–S45, 2007. 48. Muhle C, Reinhold-Keller E, Richter C, et al: MRI of the nasal cavity, the paranasal sinuses and orbits in Wegener’s granulomatosis. Eur Radiol 7(4):566–570, 1997. 49. Mukherji SK, Figueroa RE, Ginsberg LE, et al: Allergic fungal sinusitis : CT findings. Radiology 207:417–422, 1998. 50. Nouraei SA, Elisay AR, Dimarco A, et al: Variations in paranasal sinus anatomy: Implications for the pathophysiology of chronic rhinosinusitis and safety of endoscopic sinus surgery. J Otolaryngol Head Neck Surg 38(1):32–37, 2009. 51. Pasquini E, Sciaretta V, Farneti G, et al: Inverted papilloma: Report of 89 cases. Am J Otolaryngol 25:178–185, 2004.

http://radiologyebook.com

702

PART II  CT and MR Imaging of the Whole Body

52. Pathria MN, Blaser SI: Diagnostic imaging of craniofacial fractures. Radiol Clin North Am 27:839–853, 1989. 53. Phillips CD: Current status and new developments in techniques for imaging the nose and sinuses. Otolaryngol Clin North Am 30(3):371–387, 1997. 54. Reddy UDMA, Dev Bhawna: Pictorial essay: Anatomical variations of paranasal sinuses on multidetector CT–How does it help FESS surgeons. Indian J Radiol Imaging 22(4):317–324, 2012. 55. Roobottom CA, Jewell FM, Kabala J: Primary and recurrent   inverting papilloma: Appearances with MRI. Clin Radiol 50:472–475, 1995. 56. Saravanan K, Panda NK, Chakrabarti A, et al: Allergic fungal rhinosinusitis: An attempt to resolve the diagnostic dilemma. Arch Otolaryngol Head Neck Surg 132(2):173–178, 2006. 57. Sargi ZB, Casiano RR: Surgical anatomy of the paranasal sinuses. In Kountakis SE, Önerci M, eds: In Rhinologic and Sleep Apnea Surgical Techniques, Berlin Heidelberg, 2007, Springer, pp 17–25. 58. Schubert MS: Allergic fungal sinusitis. Otolaryngol Clin North Am 37(2):301–326, 2004. 59. Silverman CS, Mancuso AA: Periantral soft tissue infiltration and its relevance to the early detection of invasive fungal sinusitis: CT and MR findings. AJNR Am J Neuroradiol 19(2):321–325, 1998. 60. Sirikci A, Bazavit YA, Bayram M, et al: Variations of sphenoid and related structures. Eur Radiol 10(5):844–848, 2000. 61. Som P: Tumors and tumor-like conditions of sino-nasal cavity. In Som P, Bergeron RT, editors: Head and Neck Imaging, ed 4, St. Louis, 2003, Mosby, pp 261–373. 62. Som PM, Brandwein MS: Inflammatory diseases. In Som PM, Curtin DC, editors: Head and Neck Imaging, ed 4, St. Louis, 2003, Mosby, pp 193–259. 63. Som PM, Curtin HD: Inflammatory lesions and tumors of the nasal cavities and paranasal sinuses with skull base involvement. Neuroimaging Clin N Am 4:499–513, 1994.

64. Som PM, Lidov M, Brandwein M, et al: Sinonasal esthesioneuroblastoma with intracranial extension: Marginal tumor cysts as a diagnostic MR finding. AJNR Am J Neuroradiol 15(7):1259–1262, 1994. 65. Som PM, Shugar JMA: Anatomy and physiology. In Som PM, Curtin DC, editors: Head and Neck Imaging, ed 4, St Louis, 2003, Mosby, pp 87–148. 66. Sonkens JW, Harnsberger HR, Blanch GM, et al: The impact of screening sinus CT on the planning of functional endoscopic sinus surgery. Otolaryngol Head Neck Surg 105(6):802–813, 1991. 67. Stackpole SA, Edelstein DR: The anatomic relevance of the Haller cell in sinusitis. Am J Rhinol 11:219–223, 1997. 68. Stringer SP, Ryan MW: Chronic invasive fungal rhinosinusitis. Otolaryngol Clin North Am 33(2):375–387, 2000. 69. Syrjanen K, Syrjanen S: Detection of human papillomavirus in sinonasal papillomas: Systematic review and meta-analysis. Laryngoscope 123:181–192, 2013. 70. Tart RP, Mukherji SK, Avino AJ, et al: Facial lymph nodes: Normal and abnormal CT appearance. Radiology 188:695–700, 1993. 71. Weber AL: Imaging of cysts and odontogenic tumors of the jaw: Definition and classification. Radiol Clin North Am 31:101–120, 1993. 72. Yang C, Talbot JM, Hwang PH: Bony abnormalities of the paranasal sinuses in Wegener’s granulomatosis. Am J Rhinol 15(2):121–125, 2001. 73. Yoon JH, Na DG, Byun HS, et al: Calcification in chronic maxillary sinusitis: Comparison of CT findings with histopathologic results. AJNR Am J Neuroradiol 20:571–574, 1999. 74. Younis RT, Anand VK, Davidson B: The role of computed tomography and magnetic resonance imaging in patients with sinusitis with complications. Laryngoscope 112(2):224–229, 2002. 75. Yousem DM, Fellows DW, Kennedy DW, et al: Inverted papilloma: Evaluation with MR imaging. Radiology 185:501–505, 1992. 76. Zinreich SJ, Abayram S, Benson ML, et al: The osteomeatal complex and functional endoscopic surgery. In Som PM, Curtin HD, editors: Head and Neck Imaging, ed 4, St. Louis, 2003, Mosby, pp 149–174.

http://radiologyebook.com

24  Cervical Adenopathy and Neck Masses Sotirios Bisdas, Katarina Surlan-Popovic, and Thomas J. Vogl

CLASSIFICATION Cancer of the head and neck, which includes cancers of the larynx, nasal passages and nose, oral cavity, pharynx, salivary glands, buccal regions, and thyroid, is the sixth most frequent cancer worldwide.78 Neck masses can be grouped into two major types: nodal masses  and nonnodal masses. Both types can appear as benign or malignant lesions. Most malignant nonnodal masses are epithelial malignancies of the mucous membranes of the upper aerodigestive tract, known as head and neck squamous cell carcinoma (SCC). A second major group of head and neck nonnodal masses arises from the glandular tissue, including thyroid and salivary glands. Less frequent head and neck nonnodal neoplasms include soft tissue and bone tumors (e.g., sarco­ mas), neuroectodermal tissue tumors (e.g., paragangliomas, malignant melanomas), and skin cancer (e.g., SCC, basal cell carcinoma).

EPIDEMIOLOGY Chronic consumption of alcoholic beverages and smoking are accepted risk factors for head and neck cancer. In the upper aerodigestive  tract, local morphologic, metabolic, and functional alterations lead to increased susceptibility to carcinogens and cell proliferation in the mucosa, resulting in genetic changes with the development of dys­ plasia, leukoplasia, and carcinoma. Also, radiation exposure is an  established environmental factor for the development of glandular neoplasia. Nodal masses in the head and neck result from regional metastases of malignant lesions of the head and neck or represent reactive enlarge­ ment due to inflammatory lesions or benign neoplasms. Lymphoma of the head and neck also presents with enlarged nodes. Cervical node metastases have variable incidence rates, and their presence is associ­ ated with a decrease in global survival to roughly half of affected patients; in addition, they are associated with higher recurrence rates.62

NORMAL ANATOMY Gross Anatomy The neck is composed of the posteriorly located nucha and the ante­ riorly located cervix. The nucha consists primarily of the vertebral column and its associated musculature, and the cervix (which also means “neck”) can be thought of as a cylinder of soft tissue whose superior extent is a line connecting the occiput and the tip of the chin and whose inferior extent parallels the course of the first rib at the thoracic inlet.

Muscles and Bones.  The most important landmarks to be grossly identified when one is studying the neck in any plane are the sterno­

cleidomastoid muscle and the hyoid bone. The sternocleidomastoid muscle takes its origin from the mastoid tip and digastric notch at the skull base and extends anteriorly, inferiorly, and medially on each side to insert as two heads on the medial third of the clavicle and the manubrium. The course of the sternocleidomastoid muscle divides the soft tissues of the neck into two paired spaces, the anterior triangles and the posterior triangles (one on each side) (Fig. 24-1). All structures located anterior to the sternocleidomastoid muscle lie within the ante­ rior triangles, and structures deep and posterior to it are in the poste­ rior triangles. The anterior triangles have a common side abutting each other in the midline, and their superior border is the mandible. The hyoid bone divides these triangles into a suprahyoid portion and an infrahyoid portion. The suprahyoid division contains the laterally located submandibular triangles and the medially located submental triangle. Superiorly these triangles are separated from the oral cavity by the mylohyoid muscle. By definition, structures above the mylohy­ oid muscle are in the oral cavity; those below are in the neck. The base of each submandibular triangle is formed by the body of the mandible. The other two sides of each submandibular triangle are formed by the anterior and posterior bellies of the digastric muscle. The digastric muscle takes its origin from a depression on the skull base between the mastoid tip and the styloid process known as the digastric notch. The posterior belly of the digastric muscle extends anteriorly, inferiorly, and medially from the digastric notch to the greater cornu of the hyoid bone. The anterior belly of the digastric muscle extends from the greater cornu anteriorly, superiorly, and medially to insert on the internal surface of the mandible inferior to the attachment of the mylohyoid muscle. The base of the submental triangle is formed by the hyoid bone. The other two sides of the sub­ mental triangle are formed by the anterior bellies of the digastric muscles. The infrahyoid division of each anterior triangle is divided by the superior belly of the omohyoid muscle into two parts, the carotid triangle superolaterally and the muscular triangle inferomedially. The infrahyoid portion of the anterior triangle contains the larynx, hypo­ pharynx, trachea, esophagus, lymph nodes, and thyroid and parathy­ roid glands. The two sides of each posterior triangle are the sternocleidomastoid muscle anteriorly and the trapezius muscle posteriorly. The base of each triangle is formed by the clavicle. The inferior belly of the omo­ hyoid muscle crosses the inferior aspect of the posterior triangle and divides it into two unequal parts, the occipital triangle superiorly and the subclavian triangle inferiorly. The major structures located within the occipital triangle are the spinal accessory nerve (cranial nerve XI) and its associated chain of lymph nodes. The major structures located within the subclavian triangle are the transverse cervical vessels and their associated chain of lymph nodes.

http://radiologyebook.com

703

704

PART II  CT and MR Imaging of the Whole Body

2

1 B

4a

A

2 C 4b

3 5 E 6

D 5

4

M

F

THY

FIG 24-1  Triangles of the neck. A, Submental triangle; B, submandibular triangle; C, carotid triangle; D, muscular triangle; E, occipital triangle; F, subclavian triangle. 1, Anterior belly of the digastric muscle; 2, posterior belly of the digastric muscle; 3, superior belly of the omohyoid muscle; 4, inferior belly of the omohyoid muscle; 5, sternocleidomastoid muscle; 6, trapezius muscle. (Modified from Reede DL, et al: CT of the soft tissue structures of the neck. Radiol Clin North Am 22:239, 1984.)

Fasciae and Spaces of the Neck.  Interest in the fasciae of the neck stems from early anatomic investigations performed by surgeons who were seeking ways to predict the spread of infection. The importance of the fasciae of the neck lies in their ability to limit the spread of infections and some tumors. There is no consistent definition of what constitutes fascia, so there are varied descriptions of the anatomic boundaries of the fasciae (Figs. 24-2 and 24-3).18,47,51 Traditionally there are two major cervical fasciae, the superficial cervical fascia (SCF) and the deep cervical fascia (DCF). Superficial cervical fascia.  The SCF is a layer of fatty loose con­ nective tissue that covers the head, face, and neck and contains the thin platysma muscles, the muscles of facial expression, the subcutaneous nerves and lymphatics, and portions of the anterior and external jugular veins. Its primary function is to allow the skin to glide easily over the deeper structures of the neck. Infections that track along the SCF are often secondary to skin infections and rarely track deeper into the neck. Deep cervical fascia.  The DCF is made up of thinner but denser, better-defined layers and extends in the neck below the skull base and encloses the muscles of the neck, the mandible, and the muscles of mastication and deglutition. It consists of three layers: (1) the super­ ficial (or investing) layer (SLDCF) surrounding all the important structures of the neck; (2) the middle layer, with the pretracheal and visceral layers (MLDCF) surrounding the aerodigestive tract; and (3) the deep layer (DLDCF) surrounding the vertebral column and para­ vertebral muscles. The DLDCF has two divisions separated by a poten­ tial space: the alar layer anteriorly and the prevertebral layer posteriorly. All three layers are closely associated in the anterolateral neck, where they contribute to the formation of the carotid sheath, which sur­ rounds the carotid artery, internal jugular vein, and vagus nerve. Superficial layer of the deep cervical fascia.  The SLDCF is attached posteriorly to the ligamentum nuchae and the spinous pro­ cesses of the cervical vertebrae. As it extends anteriorly the fascia splits

1

FIG 24-2  The layers of deep cervical fascia and the cervical spaces, sagittal view. Solid line, investing fascia; dotted line, visceral fascia; dashed line, prevertebral fascia. 1, Suprasternal space of Burns; 2, visceral space; 4a, retrovisceral component of the retropharyngeal space; 4b, danger space; 5, prevertebral space; M, mylohyoid muscle; THY, thyroid gland. (Adapted from Smoker WRK, Harnsberger HR: Normal anatomy of the neck. In Som PM, Bergeron RT, editors: Head and neck imaging, ed 2, St. Louis, 1990, Mosby Year Book, pp 498–518.) STR THY TR E 2 4

CCA IJV SCM

3

5

TRP

FIG 24-3  The layers of deep cervical fascia and the cervical spaces of the infrahyoid neck, axial view. Solid line, investing fascia; dotted line, visceral fascia; dashed line, prevertebral fascia. 2, Visceral space; 3, carotid space; 4, retropharyngeal space; 5, prevertebral space; CCA, common carotid artery; E, esophagus; IJV, internal jugular vein; SCM, sternocleidomastoid muscle; STR, strap muscles; THY, thyroid gland; TR, trachea; TRP, trapezius muscle. (Adapted from Smoker WRK, Harnsberger HR: Normal anatomy of the neck. In Som PM, Bergeron RT, editors: Head and neck imaging, ed 2, St. Louis, 1990, Mosby Year Book, pp 498–518.)

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses to envelop the trapezius and sternocleidomastoid muscles, crossing the posterior triangle of the neck. In the suprahyoid neck, the SLDCF passes over the muscles below the floor of the mouth, splits to sur­ round the submandibular glands, and when it reaches the mandible, divides into superficial and deep leaflets that enclose the muscles of mastication and form the masticator space. The SLDCF is attached superiorly to the external occipital protuberance, mastoid process, and skull base. In the infrahyoid neck, the SLDCF encircles the strap muscles (ster­ nothyroid, sternohyoid, and thyrohyoid muscles) anterior to the larynx and trachea (although some authors believe this fascia is part of the MLDCF). The fascia also invests the omohyoid muscle, holding the muscle close to the clavicle so that contraction of the muscle results in a downward rather than a lateral pull on the hyoid bone. Caudally in the midline the SLDCF splits into two layers enclosing the manubrium. This creates the variably sized suprasternal space of Burns (or Gruber), which contains fat and a communicating vein between the left and right anterior jugular veins. If this space is entered during a tracheos­ tomy, inadvertent transection of the communicating vein may result in considerable blood loss. Caudally the SLDCF is attached into the pectoral and axillary regions. Middle layer of the deep cervical fascia.  The MLDCF is the most arbitrarily described of the fascial layers. The MLDCF in the infrahyoid neck is most commonly considered the fascia between the sternocleidomastoid muscles, passing behind the strap muscles and lying in front of the thyroid gland and trachea (many authors consider this loose connective tissue about the trachea and thyroid gland not  to be part of the MLDCF). The deepest portion of the MLDCF is closely adherent to the muscular walls of the pharynx and esophagus and extends ventrally via the pterygomandibular raphe over the buc­ cinator muscles (buccopharyngeal fascia). Inferiorly the MLDCF extends behind the sternum into the origin of the strap muscles and fuses into the fibrous pericardium (at about the level of the fourth thoracic vertebra), forming the anterior border of the retropharyngeal space. Cranially the MLDCF fuses with the thyroid cartilage and the hyoid bone, and laterally it contributes to the carotid sheath on either side of the neck. The MLDCF in the suprahyoid neck extends from the skull base and follows the outer surface of the pharyngobasilar fascia (cranial continuation of the superior pharyngeal constrictor muscle) and the pharyngeal constrictors. Here the fascia is firmly adherent to the mus­ cular wall of pharynx and esophagus, although some authors argue for the existence of a potential visceral space between the fascia and the viscera. Deep layer of the deep cervical fascia.  The DLDCF, like the SLDCF, begins in the posterior midline and extends to the side, cover­ ing and investing the muscles that form the floor of the posterior tri­ angle. The deep layer is attached laterally to the posterior tubercles of the transverse processes of the vertebrae, where it splits into the ante­ rior alar and posterior prevertebral layers as they extend anterior to the vertebral body (between them lies a layer of loose connective tissue). The alar and prevertebral layers have different craniocaudal extensions. Although both attach superiorly at the skull base, the alar layer blends with the visceral layer along the posterior margin of the esophagus at the level between the sixth cervical and fourth thoracic vertebrae; the prevertebral layer extends from the skull base to the coccyx. The proximal portion of each phrenic nerve lies deep to the prevertebral fascia on the anterior face of each anterior scalene muscle. On each side of the neck, the DLDCF extends laterally from the trans­ verse process of the seventh cervical vertebra, covers the dome of the pleura, and attaches to the rib medially. This fascia separates the lower neck from the thorax and is called Sibson’s fascia.

705

Neck spaces.  The three layers of the DCF delineate spaces in and through which bacterial infections can spread in the infrahyoid neck. Visceral space.  The visceral space includes all structures within the confines of the MLDCF (pharynx, cervical esophagus, trachea, thyroid gland, parathyroid glands, larynx, recurrent laryngeal nerves, and portions of the sympathetic trunk) and is continuous with the anterior mediastinum.4 It is considered to have two subdivisions: an anterior pretracheal space and a more posterior retrovisceral space (see later), which communicate freely around the sides of the larynx, the lowermost pharynx, and the upper cervical esophagus between the levels of the thyroid cartilage and the inferior thyroid artery. Caudal to this level, the pretracheal space is separated from the retrovisceral space by dense connective tissue. The posterior margin of the retrovis­ ceral space is the alar layer of the DLDCF. Retropharyngeal space.  The retropharyngeal space is situated in the midline directly posterior to the pharynx.18 The anterior bound­ ary of the retropharyngeal space is formed by the MLDCF fascia, and the posterior boundary is formed by the DLDCF. The retropharyngeal space can be further divided by the alar layer of the DLDCF into the retrovisceral space anteriorly and the danger space posteriorly. Depend­ ing on its location, infection in the retropharyngeal space can extend inferiorly in the retrovisceral and pretracheal space (it is common that a retropharyngeal abscess can affect the thyroid gland and anterior mediastinum) to the level of approximately C7 (where the alar layer and the visceral layer fuse) or in the danger space just above the level of the diaphragm, where the fused alar and visceral fasciae fuse with the prevertebral fascia. Because the danger space is a closed one, the infection must penetrate its walls to enter the space. Because it is impossible to separate these two smaller spaces radiographically, the retropharyngeal space may be considered a single radiographic space, with the recognition that all infections in the retropharyngeal space must be evaluated in full, and scans of either the superior mediastinum or entire chest may be necessary (Fig. 24-4). Prevertebral space.  The vertebral space is a potential space that includes those structures (vertebral bodies, paravertebral and scalene muscles, vertebral arteries) that are surrounded by the DLDCF. Infec­ tions of the prevertebral space are usually secondary to vertebral infec­ tion (osteomyelitis) or posterior extension arising in the retropharyngeal space. The vast majority of the pathology that affects this space arises from the adjacent vertebral bodies, disks, and nerves. Carotid space.  The carotid space is a potential space within the carotid sheath.71 Investigators doubt whether this cavity can act as a space that allows spread of infections. Furthermore, because little areolar tissue is present within the sheath, actual infection of the carotid space is rare. The most accepted theory is that the carotid sheath may be a true space only below the carotid bifurcation and above the root of the neck. Similarly there is controversy over whether the carotid sheath in the suprahyoidal neck should be considered a separate “carotid space” or part of the parapharyngeal space. We support the latter opinion, which follows the anatomic and surgical literature. Remarkably the carotid sheath, with its contributions from all three layers of the deep cervical fascia, can act as a conduit of infec­ tion from one space to another. Septic or reactive thrombosis of the internal jugular vein may produce swelling within the carotid space. The internal jugular vein and its associated chain of lymph nodes closely parallel the deep surface of the anterior margin of the sterno­ cleidomastoid muscle, thereby bridging the anterior and posterior  triangles. The arterial and neural components of the carotid sheath, common carotid artery, internal carotid artery, and vagus nerve also traverse the neck within the anterior triangle. Body of the mandible.  Another space in the suprahyoid neck is the space of the body of the mandible, which is confined by the deep

http://radiologyebook.com

706

PART II  CT and MR Imaging of the Whole Body Sectional Anatomy

*

* * *

T3

FIG 24-4  Retropharyngeal space abscess. Sagittal T2-weighted MRI demonstrates a hyperintense mass in the retropharyngeal space (asterisks) with anterior displacement of the airway. The abscess extends inferiorly to the level of the third thoracic vertebra (T3), implying involvement of the retrovisceral space rather than the danger space. (From Holliday RA, Prendergast NC: Imaging inflammatory processes of the oral cavity and suprahyoid neck. Oral Max Surg Clin North Am 4:215, 1992.)

leaflet of the SLDCF on the lingual surface of the mandible along the line of origin of the mylohyoid muscle and the lingual cortex of the mandible. This potential space is limited anteriorly by the attachment of the anterior belly of the digastric muscle and posteriorly by the attachment of the internal pterygoid muscle. Submandibular space.  The submandibular space is the upper­ most boundary of the neck as it extends from the hyoid bone to the mandible. The mucosa in the floor of the mouth is its cranial border. The space is divided into an upper (sublingual space) and lower portion by the mylohyoid muscle. The portions communicate freely around the dorsal margin of the muscle. The submandibular gland is folded around the back of this muscle, lying partially above and par­ tially below its dorsal edge. The radiologist should also identify whether a tumor is in the sublingual space, because that will eventually lead to an intraoperative sacrifice of the lingual nerve (trigeminal nerve branch that carries the taste fibers of the anterior two thirds of the tongue) and hypoglossal nerve, which runs from the floor of the mouth to the tongue base. Other radiographically important anatomic landmarks of the above-described spaces are the deep portion of the submandibular gland and the submandibular duct (Wharton’s duct) in the sublingual space and the anterior belly of the digastric muscle, the superficial portion of the submandibular gland, and the lymph nodes in the submandibular space. The above-mentioned submental and submandibular triangles of the neck are the superficial landmarks that correspond to the lower portion of the submandibular space.

The sternocleidomastoid muscle is a constant landmark that can always be identified on axial computed tomography (CT) scans or magnetic resonance imaging (MRI) through the neck. As one pro­ gresses from superior to inferior, the location of the sternocleidomas­ toid muscle moves from a far lateral position to a paramedian position. With this change in position of the sternocleidomastoid muscle, a change in relative sizes of the anterior and posterior triangles can be identified. Superiorly the anterior triangle occupies the majority of the cross-sectional area of the neck because there is little space between the sternocleidomastoid and trapezius muscles. Inferiorly the posterior triangle occupies the majority of the cross-sectional area of the neck because there is little space between the anterior surface of the sterno­ cleidomastoid muscle and the midline. The mylohyoid muscle, the major landmark separating the oral cavity from the suprahyoid neck, is readily identified on both coronal and axial images. On axial images (Fig. 24-5A and B), the muscle appears in cross section as two separate muscle bundles medial to the mandible. As axial images are obtained from superior to inferior, the distance between the separate bundles decreases. The superior aspect of the submandibular gland is identified on axial images at the poste­ rior margin of the mylohyoid muscle. On coronal images, the muscle has the configuration of a hammock suspended from the medial (lingual) surface of the mandible (see Fig. 24-5C and D). Immediately inferior to the mylohyoid muscle, the anterior bellies of the digastric muscle can be identified (Fig. 24-6). Depending on the degree of angu­ lation used on axial imaging, varying amounts of mylohyoid muscle can be seen projecting between the anterior bellies of the digastric muscle. The posterior belly of the digastric muscle can usually be identified medial and posterior to the parotid gland, separating the gland from the contents of the carotid sheath. The hyoid bone is the major bony landmark of the anterior  neck. The hyoid is best identified on axial CT images, where its central body and greater horns (cornua) appear in the midline as an inverted U approximately at the level of the C3-C4 disk space (Fig. 24-7).  The carotid bifurcation is typically located at or near the level of the hyoid bone. Separation of the larynx into the supraglottis, glottis, and subglottis seems fundamental for the interpretation of scans of patients with laryngeal cancer. The supraglottis includes the false vocal cords, aryte­ noids, epiglottis, and aryepiglottic folds. The glottis includes the true vocal cords, the anterior and posterior commissures, and the vocal ligament extending from the arytenoid cartilage to the thyroid carti­ lage. The laryngeal ventricle, which is supposed to separate the supra­ glottis and glottis, is itself a part of the supraglottis. Finally, the subglottis begins 1 cm below the ventricle and extends to the first tracheal ring. The thyroid cartilage has an inverted V appearance on axial images (Fig. 24-8). The superior third of the cartilage is not fused in the midline at the level of the thyroid notch. Immediately superficial to the two halves of the thyroid cartilage are the strap muscles. The degree and pattern of ossification of the cartilages vary with the age and sex of the patient. Gross cartilage invasion in laryngeal cancer can be detected with CT, but nonossified hyaline cartilage shows about the same density values as tumor on CT images. Separation of nonossified cartilage from the overlying strap muscles and detection of intracartilaginous alterations is accomplished more easily with MRI than with CT. The superior cornua of the thyroid cartilage are easily identified as paired calcified structures, and the inferior cornua of the thyroid cartilage articulate with the posterior aspect of the cricoid cartilage.

http://radiologyebook.com

707

CHAPTER 24  Cervical Adenopathy and Neck Masses

2 2 3 3 9 7

9

A

7

5 4

1

1

5 4

1

B

2 8 2

C

8

D

FIG 24-5  Normal sectional anatomy at the junction of the oral cavity and neck. A, Contrast-enhanced CT image at the level of the mandible. The sternocleidomastoid muscle is posterior to the parotid gland. B, T1-weighted MRI at a similar level in a different patient. Note the greater tissue contrast between the mylohyoid muscle and the submandibular gland on MRI. Coronal CT scan (C) and coronal T1-weighted MRI (D) demonstrate the anterior belly of the digastric inferior to the sling-shaped mylohyoid muscle. 1, Sternocleidomastoid muscle; 2, mylohyoid muscle; 3, submandibular gland; 4, internal jugular vein; 5, internal carotid artery; 6, posterior belly of digastric muscle; 7, parotid gland; 8, anterior belly of digastric muscle; 9, longus colli muscle.

The cricoid cartilage is the only intact ring of the larynx with a complete posterior component and has a variable degree of ossification on axial imaging. In adult patients, the cricoid cartilage appears on CT scans as a thin rim of calcification surrounding a lucent medullary center. On T1-weighted MRI, the cricoid appears as a ring of tissue isointense to fat. The level of the cricoid cartilage also marks the point at which the omohyoid muscle crosses over the internal jugular vein. The cricoarytenoid joint is a readily recognizable structure that demar­ cates glottic from supraglottic and subglottic structures.

The paraglottic tissues can be separated into the preepiglottic and paraglottic fat (dark on CT scanning) and the thyroarytenoid muscle (soft tissue density on CT and intermediate T1-weighted on MRI) of the true vocal cords. The true vocal cords meet in the midline at the anterior commissure, which should be no more than 1 to 2 mm thick. The posterior commissure refers to the mucosa between the two vocal processes on the anterior surface of the arytenoid cartilage. At approximately the level of the cricoid cartilage (Fig. 24-9), the superior pole of the thyroid gland appears on axial images as a

http://radiologyebook.com

708

PART II  CT and MR Imaging of the Whole Body

8

8

2

3

3 9

6

5 4 1

FIG 24-6  Normal axial anatomy of the suprahyoid neck. T1-weighted MRI in the same patient as in Figure 24-5B. The submental triangle is located between the two anterior bellies of the digastric muscle. 1, Sternocleidomastoid muscle; 2, mylohyoid muscle; 3, submandibular gland; 4, internal jugular vein; 5, internal carotid artery; 6, posterior belly of digastric muscle; 8, anterior belly of digastric muscle; 9, longus colli muscle.

triangular soft tissue structure situated between the cricoid cartilage medially and the carotid sheath structures laterally. The normal gland (owing to its iodide content) has a density of 80 to 100 Hounsfield units (HU) on CT. A homogeneous gland on noncontrast CT corre­ lates well with thyroid function. The gland enhances diffusely after application of iodinated intravenous contrast material. The latter can alter the radioactive iodine uptake measurements (i.e., in nuclear imaging) for up to 6 weeks following the application. This altera­ tion does not happen after gadolinium application. On noncontrast T1-weighted images, the gland is homogeneous in appearance and slightly hyperintense to skeletal muscle. On T2-weighted imaging, the gland is hyperintense relative to the neck musculature. Axial and sagit­ tal T1- and T2-weighted images (with fat saturation) as well as post­ contrast axial fat-saturated T1-weighted images are essential for MRI of the thyroid and parathyroid glands. The posterior triangle of the normal supracricoid neck appears as a fat-filled cleft containing a few small soft tissue structures represent­ ing nutrient vessels, lymph nodes, and nerves. Beginning at the level of the cricoid cartilage, the anterior scalene muscle can be identified in the medial aspect of the posterior triangle, posterior to the carotid sheath structures. The anterior scalene muscle, which arises from the transverse processes of C4-C6, is the key reference point needed to locate the major neural and vascular structures in the region. The subclavian vein is located anterior to the muscle that divides the sub­ clavian artery in three anatomic portions. The phrenic and vagus nerves cross the root of the neck anterior to the first portion of the subclavian artery laterally and medially of it, respectively. The anterior scalene muscle lies directly anterior to the exiting trunks of the brachial plexus. At the level of the first tracheal ring, the brachial plexus is

3

3

3 1

5 4

7 6

7 6

9 1

A

4

9

5

B FIG 24-7  Normal axial anatomy at the level of the hyoid bone. A, Contrast-enhanced CT scan. Note that the normal retrofacial vein or proximal external carotid artery might be mistaken for a lymph node or other mass on a noncontrast examination. B, T1-weighted MRI demonstrates the same soft tissue and vascular anatomy without intravenous contrast. 1, Sternocleidomastoid muscle; 3, submandibular gland; 4, internal jugular vein; 5, internal carotid artery; 6, external carotid artery; 7, retrofacial vein; 9, longus colli muscle.

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

709

2

2

3

4

3

1

9

4

1

9

B

A

FIG 24-8  Normal axial anatomy at the level of the thyroid cartilages. A, CT scan. B, T1-weighted MRI. Note how the anterior margin of the sternocleidomastoid muscle is located anterior to the vertebral body. Compare with Figure 24-5B. 1, Sternocleidomastoid muscle; 2, strap muscle; 3, internal carotid artery; 4, internal jugular vein; 9, longus colli muscle.

identified as a heterogeneous low-density or low-signal-intensity focus immediately posterior to the anterior scalene muscle.

Nodal Anatomy Forty percent of the body’s lymph nodes are located in the head and neck region, which represents about 20% of the body’s volume. Each lymph node is enclosed by a fibrous capsule, and lymph moves into a node via several lymphatic vessels and emerges by one or two efferent vessels. Fibrous septa (trabeculae) extend from the covering capsule toward the center of the node. Cortical nodules found within the sinuses along the outer region of the node are separated from each other by these trabeculae. The center, or medulla, of a lymph node is composed of sinuses and cords. Both the cortical and medullary sinuses are lined with specialized reticuloendothelial cells (fixed mac­ rophages) capable of phagocytosis. When the lymph nodes elicit the immune response, the nodes enlarge (reactive hyperplasia). However, the morphologic analysis of the reactivity has a controversial prognos­ tic value, probably owing to a variety of response patterns in an indi­ vidual patient as well as in an individual node.50,70 Nodes that drain areas of frequent infection (e.g., jugulodigastric or submental nodes) tend to enlarge over many years relative to other cervical nodes. Normal lymph nodes have an elliptical shape whose long axis is directed parallel to the anatomic structures along which the nodes are distributed and thus cannot be shown in axial CT slices. The ratio of the longitudinal (maximum) diameter to the transverse (minimum) diameter of a lymph node is a major morphologic criterion for assess­ ing metastatic disease. In addition the minimum diameter more closely correlates with the volume of the lymph node and is less dependent on the scan plane, and therefore it is useful for follow-up imaging. The CT attenuation of lymph nodes on noncontrast CT scans is equivalent to that of other soft tissue structures.42 The node hilum can sometimes be seen to contain a small amount of fat. Normal lymph nodes display moderate homogeneous enhancement following intravenous injection of contrast material. On MRI, lymph nodes appear homogeneous with low to interme­ diate precontrast T1-weighted intensity and appear at fairly high

intensity in high-resolution T2-weighted images. Moreover, lymph nodes have smooth noninfiltrating margins and a slight homogeneous enhancement after gadolinium administration. High-resolution MR microimaging offers a detailed structural analysis of the nodes using microscopy coils and demonstrating medullary sinus as regions of low signal intensity and follicles as high-intensity structures; the method can effectively characterize the morphologic details of benign and metastatic nodes without using gadolinium enhancement.70 Neverthe­ less, microimaging is limited to the superficial nodes and requires additional time for examination and evaluation of the imaging data.

Classification.  As the pathways of node metastasis were recognized in the last century, the necessity of a nodal classification that identifies the precise location of the nodes emerged. This is important because neoplastic lymph nodes are mainly regionally localized according to the sentinel lymph node concept. The most commonly used classifica­ tion of Rouvìere in 1938, mainly based on the superficial triangles of the neck (Fig. 24-10), was followed by that of Shah and colleagues in 1981 (Fig. 24-11), which employed a simpler level-based system to help the surgeon select the most appropriate type of nodal resection.61 The latter system can be easily transferred to the axial imaging (Fig. 24-12). Since then, further classifications such as that of the American Joint Committee on Cancer (AJCC) in 1997 and the American Academy of Otolaryngology—Head and Neck Surgery (AAO-HNS) have modified and improved the initial level-based system.56 To address specific issues raised in both the AJCC and AAO-HNS classifications and to bring to nodal classification the anatomic detail and reproducibility of imaging the vast majority of patients currently receive, an imaging-based (CT and MRI) classification was suggested in 1999.66 A brief presentation of this classification is given in Box 24-1.

IMAGING TECHNIQUES Various imaging techniques are currently used in the evaluation of patients with neck masses before, during, and after treatment.21,25 Each of these imaging modalities has its own advantages and disadvantages.

http://radiologyebook.com

710

PART II  CT and MR Imaging of the Whole Body

1 1 2

4

9

4

3 4

3

3 9

8

A

8

B

FIG 24-9  Normal axial anatomy at the level of the cricoid cartilage and inferior cornua of the thyroid cartilage. A, CT scan. The asymmetry in the diameters of the internal jugular veins is a normal variant. B, T1-weighted MRI. 1, Sternocleidomastoid muscle; 2, thyroid gland; 3, internal carotid artery; 4, internal jugular vein; 8, trapezius muscle; 9, longus colli muscle.

Parotid nodes Occipital nodes Facial nodes

Mastoid nodes Jugulodigastric node

Submandibular nodes Submental nodes

Spinal accessory nerve

External carotid artery

Greater auricular nerve

Internal jugular chain (common carotid artery, internal jugular vein, vagus nerve)

Spinal accessory chain External jugular vein Transverse cervical chain

Anterior jugular chain

FIG 24-10  Nodal chains of the neck. (Modified from Reede DL, Bergeron RT: CT of cervical lymph nodes. J Otolaryngol 11:411, 1982.)

In most patients, especially in developed countries, CT and MRI are performed for diagnosing and staging of neck masses. In the pediatric population, ultrasonography has a special role in follow-up of nodal masses.73 Whether CT or MRI is the primary imaging modality depends on the institution, although head and neck radiologists tend to use CT for imaging of laryngeal cancer, hypopharyngeal cancer,

oropharyngeal cancer, and inflammatory lesions and also use it to complement the diagnostic evaluation of tumors with presumed bone destruction. The relative low cost of CT examinations, its wide avail­ ability, the short examination time, and the high-quality multiplanar imaging provided, as well as the higher sensitivity and specificity regarding nodal involvement in malignant neoplasms, are indisputable

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

711

G

G

II

I

A III VII VI

V

IV

A d

advantages of CT over MRI. On the other hand, the relatively low soft tissue contrast resolution of CT, the need to administer iodinated contrast agents, and the severe image quality degradation caused by dental fillings and other foreign objects are disadvantages of CT imaging. MRI in turn may be jeopardized by motion artifacts and orthodontic material. Whatever the admission imaging modality, the patient’s follow-up has to be performed with the same modality and the same examination protocol.

s

*

FIG 24-11  Simplified nodal classification. Diagram of the head and neck in left anterior oblique projection. Palpable nodes are indicated with use of a simplified nomenclature of Roman numerals I through VII. See Box 24-1 for explanation. A, Common carotid artery. (From Som PM: Lymph nodes of the neck. Radiology 165:596, 1987.)

d a v

B FIG 24-12  Sectional imaging using simplified nodal classification. A, Axial contrast-enhanced CT scan demonstrates bilateral lymph nodes (curved white arrows) anterior to each submandibular gland (G). They are level I (submandibular triangle) lymph nodes. Note the lucent zone of fat eccentrically located in each node (straight arrows). These zones represent prominent nodal hila. B, Axial contrast-enhanced CT scan demonstrates a homogeneous 2.5-cm diameter lymph node (asterisk) in the left posterior triangle deep to the sternocleidomastoid muscle (s) and separate from the internal jugular vein (v), internal carotid artery (a), and posterior belly of the digastric muscle (d). This is a level V (spinal accessory chain) lymph node. Asymmetry in the appearance of right and left digastric muscles is secondary to slight skewing of the patient’s head within the scanner gantry.

Computed Tomography CT may be regarded as the workhorse of head and neck imaging. There are many types of CT scanners, most of which are multislice units. Thus it is not possible to define the optimal protocol. The ideal neck CT scan provides the best possible contrast of soft tissue (with the choice of an appropriate delay, contrast agent volume, flow rate, and scanning time), allows visualization of the arterial and venous vascular structures, and provides thin multiplanar reconstructions in soft tissue and bone algorithms in an angle appropriate for evaluation of the suspected lesion.27,28,33 CT is performed with the patient supine on the scanner gantry and during quiet respiration. The patient’s head may be cushioned to avoid motion artifacts during the hot flush of contrast agent injection. The patient’s shoulders must be dropped as low as possible. In case of incremental or single spiral CT, malposition of the patient can cause artifacts that simulate disease. One common solution is to acquire images from the superior orbital rim to the hard palate with the use of gantry angulation parallel to the central skull base. Images from the alveolar ridge of the mandible to the lung apex are then obtained with a gantry angulation parallel to the body of the

mandible. If the patient’s chin is fully extended, this latter series may be obtained with a 0-degree gantry angulation. Intravenous contrast enhancement is essential for CT examinations to facilitate both tissue characterization of neck masses and separation of neck masses from normal vascular structures. Most investigators recommend a bolus of approximately 80 to 100 mL of high-density iodinated contrast material administered at a rate of 1 to 2 mL/sec and a delay of 80 to 100 sec before image acquisition, followed by a steady rapid-drip saline infusion at the same rate. Power injectors, commonly used for abdominal or thoracic CT scanning, can be readily adapted for neck CT examinations. When evaluating the neck, contiguous 5-mm-thick axial images are usually obtained from the level of the superior orbital rim to the lung apex (scanning from cranial to caudal) to reduce artifacts at the level of the thoracic inlet caused by the beam-hardening effects of the con­ trast agent. The field of view must be as small as possible to enhance the spatial resolution (reconstruction of the raw data in a greater field of view can be subsequently performed to encounter possible

http://radiologyebook.com

712

PART II  CT and MR Imaging of the Whole Body

BOX 24-1  Imaging-Based Classification for

Cervical Lymph Nodes

I Submental and submandibular nodes, located above the hyoid bone, below the mylohyoid muscle, and anterior to the back of the submandibular gland Ia Submental nodes Ib Submandibular nodes II Upper internal jugular nodes, lying from the skull base to the level of the bottom of the body of the hyoid bone, posterior to the back of the submandibular gland, and anterior to the back of the sternocleidomastoid muscle IIa Nodes anterior, medial, lateral, and posterior to the internal jugular vein IIb Nodes posterior to the jugular vein, with a fat plane separating them from the vein III Midjugular nodes, located from the level of the bottom of the hyoid bone to the level of the bottom of the cricoid arch, anterior to the back of the sternocleidomastoid muscle IV Low jugular nodes, lying from the level of the bottom of the cricoid arch to the level of the clavicle, lateral to the carotid arteries V Nodes in the posterior triangle, lying posterior to the back of the sternocleidomastoid muscle from the skull base to the level of the bottom of the body of the cricoid arch and posterior to a line connecting the back of the sternocleidomastoid muscle and the posterolateral margin of the anterior scalene muscle from the level of the bottom of the cricoid arch to the level of the clavicle. They also lie anterior to the anterior edge of the trapezius muscle. Va Upper level V nodes, located from the skull base to the level of the bottom of the body of the cricoid arch Vb Lower level V nodes, located from the level of the bottom of the cricoid arch to the level of the clavicle. VI Upper visceral nodes, lying between the carotid arteries from the level of the bottom of the hyoid bone to the level of the top of the manubrium VII Superior mediastinal nodes, lying between the carotid arteries from the level of the top of the manubrium to the level of the innominate vein Supraclavicular nodes lie at or caudal to the clavicles and lateral to the carotid artery. Retropharyngeal nodes lie within 2 cm of the skull base, medial to the internal carotid arteries. From Som PM, et al: An imaging-based classification for the cervical nodes designed as an adjunct to recent clinically based nodal classifications. Arch Otolaryngol Head Neck Surg 125:388–396, 1999.

pathologic conditions in these regions). The images are reconstructed in 3-mm slices for evaluation of the oropharynx (parallel to the hard palate on the axial section and coronal to the mouth floor) and 2-mm slices for evaluation of the hypopharyngeal-laryngeal region (parallel to the vocal cords as well as vertical to them). In cases of extended osteolytic lesions, three-dimensional (3D) imaging provides valuable information to surgeons. A significant advantage of the fast image acquisition of multislice CT scanners is the easier performance of a dynamic maneuver during the examination. Such maneuvers are the [i]-phonation (for better visualization of the laryngeal ventricle) and the modified Valsalva (for better visualization of the pyriform sinuses, postcricoid region, and gingivobuccal tumors). CT imaging techniques that can complement the standard CT study include dynamic acquisition of neck mass sections during administration of the bolus of contrast agent. The morphologic infor­ mation obtained by cross-sectional imaging of SCC of the upper

aerodigestive tract is often insufficient because of the highly infiltrative character of the tumor, posttherapy edema or fibrosis, and the diffi­ culty of reliably recognizing lymph node metastases in apparently “normal” lymph nodes. Perfusion CT, based on various tracer kinetic models, aims at measurement of tissue blood supply, tissue vascularity, and contrast clearance. Because every neoplasm is characterized by neovascularity and increased angiogenic activity, perfusion CT is a potential imaging tool of tumor activity, theoretically before the tumor proceeds in a gross anatomic distortion.7

Magnetic Resonance Imaging MRI of the neck is routinely performed at magnetic fields of 1.5 tesla (T) and recently at 3.0 T. Possibly more important than field strength is the choice of the right surface coils and their optimal positioning, along with comfortable positioning of the patient and instructions to avoid any movement, coughing, and swallowing during the examina­ tion. Usually the choice of the appropriate coil depends on the specific organ or lesion being imaged; however, in our opinion the combined head-neck coil should be used every time a neck MRI study is per­ formed. This ensures a comprehensive evaluation of the entire extra­ cranial head and neck—evaluating all nodal masses, a second primary neoplasm, and perineural tumor spread. The possible drawback of inhomogeneous recipient field characteristics at the crossover between the two coils is of minor importance and can be partially overcome by shimming—improved coils with identical receiver coil elements—and sequences with short echo trains. A conventional MR survey is performed using a dedicated headand-neck circular polarization surface coil and (at the beginning) per­ forming a localizer in three planes. The axial plane is aligned along the patient’s hard palate which in turn has to be perpendicular to the tabletop. The slices should be 3 to 4 mm with either no interslice gap or a 10% interslice gap. The entire neck should be studied to evaluate lymphadenopathy, especially in patients with esophageal or thyroid cancer who present with superior mediastinal node metastases. The field of view may be approximately 20 to 22 cm and the matrix, if possible, 512 × 512 (256 × 256) pixels. Inversion recovery fast-spin echo (FSE) sequences with fat saturation (spectral presaturation by inver­ sion recovery [SPIR], spectral selection attenuated inversion recovery [SPAIR], and turbo inversion recovery magnitude [TIRM]) have proven useful in imaging lymph nodes by improving the conspicuity of small nodes adjacent to or surrounded by fat.58 Alternatively, con­ trast can be improved by performing the sequence with a chemical shift selective suppression (CHESS) technique. The FSE inversion recovery sequence can be complementary, followed by an FSE T2-weighted sequence (preferably on the axial plane). Subsequently, axial and coronal spin echo T1-weighted images are performed before and after application of gadolinium-containing contrast agent with a fat satura­ tion in the postcontrast axial sections. Another fat saturation in the coronal plane can be renounced because the saturated images are prone to susceptibility artifacts and have a long acquisition time. Finally, the radiologist ought to avoid very fast imaging using singleshot techniques and instead take advantage of parallel imaging in high-field scanners. In addition to conventional contrast-enhanced MRI, axial diffusionweighted imaging (DWI) may be also performed.1,39,69 DWI allows visualization as well as separation of molecular diffusion from micro­ circulation of the blood in the capillary network (perfusion) of bio­ logical tissues. Neoplasms are associated with alterations in water diffusivity and microcirculation of the node. DWI can be obtained with a single-shot echo planar imaging (EPI) sequence by using the same coil. The sequence can be repeated for at least three values of the motion-probing gradients (usually b = 0, 500, 700, or 1000 sec/mm2).

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses Suggested section thickness is 4 to 5 mm with an intersection gap of 1 mm. Scanning time does not usually exceed 2 minutes. Quantifica­ tion of diffusion abnormalities requires calculation of the apparent diffusion coefficient (ADC). Another promising MRI technique is proton magnetic resonance spectroscopy (MRS), wherein data can be acquired on the same scanner and with the same coils that were used for the imaging.6,38 The MRI examination is used for localizing the MRS. Single-voxel spin echo sequence may be used. The echo time (TE) allows in-phase detection of the choline signal. Weak water suppression can be achieved by a CHESS technique. The volume of interest (VOI) fits the size of the suspected lesion. Automatic shimming of the system is usually applied and controlled by interactive shimming by the radiologist. Postpro­ cessing of the data includes identification of choline and creatine metabolites. Similar to perfusion CT, dynamic MRI offers the potential of visual­ izing dynamically the tumor vascularity, including permeability disor­ ders with indirect implications for tissue oxygenation status.23 EPI or a VIBE (volumetric interpolated breath-hold imaging) sequence may be used. For postprocessing of the acquired data, a semiquantitative analysis using mean signal intensity versus time curves may be per­ formed. Parameters such as peak time (in seconds), peak enhancement (% baseline), and washout slope estimate can be calculated from the late point versus the peak point. Another approach uses commercially available deconvolution-based analysis of the dynamic data, which were first used in perfusion studies of the brain. MRI also enables contrast material–enhanced intravenous lym­ phography. The first contrast medium used for intravenous lymphog­ raphy was dextran-coated ultrasmall superparamagnetic iron oxide (USPIO) particles, which accumulate slowly in the phagocytes of the lymph nodes.17,41 Based on the T2* effects of the iron oxide particles, after a single intravenous administration of USPIO, T2-weighted images show areas of focal signal intensity loss in medullary sinuses corresponding to the distribution of uptake by macrophages. Lymph follicles appear unchanged in signal intensity because they are largely devoid of macrophages. Infiltration of lymph nodes with tumor alters the architecture, with tumor replacing the macrophages; hence the nodes retain their high signal intensity after USPIO administration. Several studies have demonstrated enhanced sensitivity and specificity for lymph node evaluation after administration of USPIO particles for head and neck malignancies. However, very small lesions may go unde­ tected because of the “blooming effect,” which reflects the extreme sensitivity of T2*-weighted imaging to susceptibility artifacts. Other pitfalls of the method include the poor contrast of T2- and T2*weighted imaging between a subcapsular node metastasis and the  surrounding extralymphatic tissue. Another promising lymph node–specific contrast agent is gado­ fluorine M (T1 contrast agent).15 Gadofluorine M is a water-soluble paramagnetic gadolinium-based agent whose mechanism of lymph node uptake is presumed to be direct transcapillary passage through interendothelial junctions into medullary sinuses.

NODAL NECK MASSES Staging Nodal staging is concerned with the presence or absence of nodal disease, nodal size, and assessment of the involved nodes. In 1987 the AJCC and the International Union Against Cancer (UICC) developed a common staging system intended to overcome the pitfalls of subjec­ tive palpation of the nodes. The latest refinement of this staging (Box 24-2) refers to all head and neck cancers except nasopharyngeal carci­ noma and thyroid carcinoma, which need separate nodal staging.

713

BOX 24-2  AJCC Nodal (N) Staging System

for Cervical Lymph Nodes Nx N0 N1

The lymph nodes cannot be assessed clinically. There are no metastatic lymph nodes. Single ipsilateral lymph node metastasis 3 cm or less in greatest dimension N2a Single ipsilateral lymph node metastasis between 3 and 6 cm in greatest dimension N2b Multiple ipsilateral lymph node metastases, none of them greater than 6 cm in greatest dimension N2c Bilateral or contralateral lymph node metastases, none of them greater than 6 cm in greatest dimension N3 Metastasis in lymph nodes that are more than 6 cm in greatest dimension From Edge S, et al, editors: AJCC cancer staging manual, ed 7, American Joint Committee on Cancer, New York, 2010, Springer-Verlag.

Role of Imaging in Staging.  Because invasive clinical and labora­ tory examinations are expensive and result in complications, radiology plays an important role in lymph node staging.75 Lymph node staging and localization of pathologic lymph nodes is mandatory for choosing therapy, either (neo) adjuvant or surgical, and is a major factor in establishing the prognosis in head and neck cancer patients.52 Nodal metastases can be categorized in two groups: overt (clinical) or nonovert (occult). Nonovert metastases may be further categorized as metastases detectable by traditional methods (e.g., staining) and as “submicroscopic” metastases, which are evident only with immuno­ histochemical or molecular analysis. Overt node metastases are imaged by cross-sectional imaging. The detection of nonovert metastasis is under development in that many patients initially classified as having cN0 in fact have occult metastatic disease (pN1).40 CT and MRI evaluation must avoid false-positive results that could lead to a selective or radical neck dissection, which is associated with increased morbidity and mortality, thus overshadowing the improve­ ment in survival. Even in the case of N0, CT and MRI play a role as close observation tools.40 Neither CT nor MRI for node staging in the neck reaches 100% sensitivity or specificity, and the accuracy of the exact number of metastases or levels involved has not been studied.59 Thus neck dissec­ tion with subsequent pathologic examination remains the gold stan­ dard for node staging. CT and MRI aim to evaluate nodal size and shape, nodal grouping, and signs of nodal metastasis such as necrosis and extranodal tumor spread.64 The latter two criteria can be better assessed with CT imaging, although MRI does have advantages  owing to the increased soft tissue contrast and the ability to obtain tissue characteristics in different sequences, including diffusion- and perfusion-weighted sequences as well as proton spectroscopy imaging. The lack of radiation burden makes MRI suitable for close follow-up of the patient, and imaging with the use of new intravenous contrast material (e.g., USPIO) seems superior to conventional imaging.

Conventional CT and MRI Findings on Neck Nodal Masses A number of diseases in addition to metastatic SCC affect the cervical lymph nodes and require imaging evaluation. These include viral,  bacterial, and protozoal diseases (Toxoplasma), fungal lymphad­ enitides, reactive lymphadenopathies, lymphoproliferative disorders, vascular lymphadenopathies, proliferative histiocytic disorders, and

http://radiologyebook.com

714

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 24-13  Nonmetastatic small nodes. MRI in axial planes of nodes in level II bilaterally (arrows) in a patient with laryngeal cancer. A, Axial T2-weighted imaging shows the small nodes in level II on both sides and level V on the right side, which according to the size criterion (see text) are not metastatic. T1-weighted imaging (B) and axial TIRM-weighted images (C) also reveal no signs of nodal metastatic disease.

lymphadenopathies related to clinical syndromes such as Kimura’s disease, sarcoidosis, and Castleman’s disease. However, because imaging characteristics in all these diseases are nonspecific, the diagnosis is based on the combination of history, clinical findings, imaging, and laboratory data. The role of imaging is to locate the node disease and evaluate any associated findings such as necrosis or abscess formation as well as soft tissue infiltration. Especially for the cervical lymph nodes, which have a major effect on the prognosis and treatment of head and neck cancer, criteria for metastatic disease include size and shape of the node, abnormality of the internal architecture (including nodal necrosis), and extracapsular tumor spread.43

Size and Shape of the Node.  Size is the most frequently used cri­ terion for diagnosis; however, sensitivity and specificity vary widely depending on the threshold value used and the level at which the node is located (Fig. 24-13). The size criterion has long been recognized by surgeons, who initially believed that level I and level II nodes larger than 1.5 cm at their maximum diameter and that nodes elsewhere in the neck larger than 1 cm contained metastases. The inaccuracy of these criteria in about one third of cases led to reaffirmation of these measurements and introduction of the minimum axial diameter of the node. These criteria suggest that the minimum diameter should not exceed 11 mm in level II nodes and should not exceed 10 mm at nodes elsewhere in the neck.22 Based on the observation that metastatic nodes tend to be spherical, whereas nonmetastatic nodes tend to be oblong or lima bean–shaped, the ratio of the maximum longitudinal length to the maximum axial nodal length was introduced.64 The ratio exceeds the value of 2 in normal hyperplastic nodes and is less than 2 in meta­ static nodes. Finally, node groups with maximal diameters of 8 to 15 mm in level II or minimal diameters of 9 to 10 mm in level IV are suggestive of metastatic disease. Similarly, for three or more nodes elsewhere in the neck, the cutoff value is suggested as 8 to 9 mm. Overall it is remarkable that even if the higher value of the normal node is considered 10 mm, 16% of patients with negative imaging have metastasis. This implicates the insufficiency of nodal size alone as a criterion for diagnosis of metastases; for instance, an acceptable

negative predictive value of 90% can be achieved with a cutoff value of 6 mm. Notably the aforementioned studies are performed in homo­ geneous sharply differentiated nodes without taking into account any other criteria. Finally, van den Brekel and colleagues suggested a minimal axial diameter of 8 to 9 mm in level II and 7 to 8 mm for the rest of neck.76

Nodal Necrosis.  Detection of nodal necrosis in untreated patients seems to be the most reliable sign of metastatic disease. Nodal necrosis appears to increase with nodal size (56%-63% of nodes >1.5 cm in diameter show necrosis), but more importantly, 10% to 33% of malig­ nant nodes smaller than 1 cm in diameter show necrosis.19 The classic appearance of nodal necrosis on CT imaging is a low-attenuated area, whereas on MRI (fat-suppressed, T2-weighted images) it is a hyperin­ tense area inside the node (Fig. 24-14). This MRI appearance seems to correspond to liquefaction necrosis, whereas coagulation necrosis gives a hypointense appearance relative to the residual nodal parenchyma signal. The advent of new MR scanners with surface coils and matrices with smaller pixels improved the quality of the MRI and thus revived its role in the detection of nodal necrosis,36 with diagnostic accuracy and sen­ sitivity of 91% to 99% and 93%, respectively, which are comparable to CT rates. Overall, a distorted internal architecture seems to be a pathog­ nomonic feature for metastatic nodes from head and neck carcinomas, whereas other investigators report distorted internal architecture to occur in 67% of metastatic nodes, 14% of lymphomas, and 9% of benign nodes.30 Considering that central nodal necrosis has a low inci­ dence in metastatic nodes,30 it seems more reliable for the radiologist to look for distorted nodal architecture. Nevertheless, one should keep in mind the increased prevalence of heterogeneous nodal architecture (including necrosis) in nodal lymphomas (including non-Hodgkin’s disease), unlike what it was believed until recently (Fig. 24-15).35,70

Extracapsular Neoplastic Spread.  Another imaging characteristic of lymph node involvement is extracapsular neoplastic spread (com­ monly seen in SCC as well as in lymphoma metastases), which can be

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

715

FIG 24-16  Extranodal spread of tumor. Axial T1-weighted contrastFIG 24-14  Nodal necrosis. Axial T2-weighted MRI in the epipharynx demonstrates three metastatic lymph nodes (arrowheads) in the retropharyngeal space and in the parapharyngeal space on the left side. The inhomogeneous hyperintense signal inside the nodes corresponds with liquefaction necrosis. The patient was referred to our institution with a squamous cell carcinoma on both tonsils.

enhanced fat-saturated MR scan demonstrates an irregular interface and diffuse infiltration between the large left necrotic nodal mass and the cervical soft tissues (arrowheads). Anteriorly the overlying platysma is thickened and irregular; posteriorly the node metastasis extends in the erector spinae muscle and splenius cervicis; medially the tumor reaches the paravertebral space; laterally there is extension into the subcutaneous fat. This nodal mass was “fixed” on physical examination and appeared with an intense swelling. Note how the mass engulfs the left internal carotid artery (arrow) and the internal jugular vein cannot be recognized owing to the nodal mass effect.

well appreciated with MRI and CT. It is characterized by the presence of indistinct nodal margins, irregular nodal capsular enhancement, or infiltration in the adjacent fat or muscle (Fig. 24-16). As the node enlarges, the incidence of extracapsular spread rises and is macroscopi­ cally easy to recognize.37,76 Here, special attention must be given to the patient’s history. Recent nodal infection, surgical manipulation, and irradiation may obliterate tissue planes around vessels and lymph nodes, mimicking extracapsular spread and leading to false-positive results. The significance of false-positive results lies in the fact that extranodal spread is the best indicator for treatment failure and decreases survival by 50%.46 Further grave prognostic findings of metastatic nodal disease include arterial invasion to the adjacent carotid artery. Imaging of nodal microinvasion in the arterial adventitia before a gross invasion into the muscularis and intima (recognized by narrowing of the lumen) is beyond the scope and capabilities of available CT and MRI techniques. However, findings such as the degree of nodal circumfer­ ential extension and the degree of obliteration of the adjacent fat plane make arterial invasion more plausible (see Fig. 24-16). Although the survival of such patients is remarkably small, carotid artery graft replacement during the surgery may be more beneficial than tumor dissection and adjuvant chemotherapy.77 FIG 24-15  Coronal T1-weighted MRI shows a clinically palpable bilateral manifestation of nodal disease (arrows) in level III in a patient with non-Hodgkin’s lymphoma. The nodes appear to be enlarged and to have a heterogeneous architecture, which indicates metastatic disease.

Other Imaging Techniques With DWI of nodal disease, metastatic nodes show generally high signal intensity in diffusion maps (irrelevant to the b value) 

http://radiologyebook.com

716

PART II  CT and MR Imaging of the Whole Body

A

B FIG 24-17  DWI of cervical nodes in a patient with non-Hodgkin’s lymphoma. A, DWI reveals a group of enlarged lymph nodes on the left side (arrow) with vivid hyperintense signal. B, The corresponding apparent diffusion coefficient map shows a uniform hypointense signal (more pronounced in the posterior larger node) indicating the restricted water diffusion in the nodes (arrow).

corresponding to low ADC values and reflecting the altered architec­ tural changes in the proportion of extracellular to intracellular water protons (Figs. 24-17 and 24-18). Necrotic areas, however, will show mixed signal intensity at different b values and high signal intensity  in ADC maps (compared with the high signal intensity in T2-  weighted images). Therefore ADC maps provide helpful information about necrotic areas and help differentiate malignant from benign nodes.1 However, DWI fails to image readily the contours and content of small nodes (usually < 0.9 cm) because of associated susceptibility artifacts. Dynamic contrast-enhanced (perfusion) modalities may be easily integrated into any CT or MRI protocol.67 Dynamic CT acquisition presents the first pass of contrast material through the vascular system and is used to assess perfusion and blood volume. For permeability measurements, a second phase is obtained, prolonging the acquisition up to 2 minutes. The main quantitative parameters are: BV (blood volume), expressing the volume of functional vascular network; BF (blood flow) and MTT (mean transit time), presenting the rate of delivery of oxygen and nutrients; and PS (permeability surface area product), which describes the rate of contrast leakage into the extracel­ lular space45,53 (Fig. 24-19). Furthermore, several recent studies have proven that noninvasively acquired perfusion CT parameters reflect angiogenesis and can be predictive of molecular biomarkers.3,26,32 Per­ fusion CT technique has been demonstrated to be both precise and reproducible when applied to the study of head and neck tumor vas­ culature.11 Head and neck SCC and metastatic lymph nodes larger than 1 cm demonstrate increased PS, BF, BV, and decreased MTT compared with normal tissue or benign lesions.9,11,57,72 However, perfusion param­ eters could not differentiate between benign and malignant lymph nodes smaller than 1 cm.7 Furthermore, perfusion CT may be used for the prognosis of short- and long-term response to therapy, a parameter of major importance for patient-tailored therapeutic regimens.10,24,79 Another clinically meaningful application of dynamic cross-sectional imaging in neck masses is differentiation between posttherapeutic changes and tumor recurrence, which remains a challenging issue on

conventional CT or MRI. In a study comparing CTP parameters between recurrent tumors and posttherapy changes after chemoradia­ tion treatment, recurrent tumor demonstrated significantly higher BF compared to tissue changes after chemoradiation.7 Dynamic contrastenhanced (DCE) MRI, also widely referred to as permeability MRI, is based on the acquisition of dynamic T1-weighted images during a bolus of gadolinium-based contrast agent. DCE-MRI offers superior spatial resolution and less susceptibility artifacts in the neck than dynamic susceptibility contrast-enhanced (DSC) MRI, which, however, profits from superior temporal resolution. The most frequently used metric in DCE MR perfusion is the transfer constant of contrast agent between intra- and extravascular space (Ktrans), whose value depends on blood flow and permeability and therefore can be used for tumor neoangiogenesis measurements (Fig. 24-20). Studies using DCE-MRI focused on predicting values of baseline perfusion parameters in meta­ static lymph nodes or primary tumor, suggesting that higher pretreat­ ment Ktrans is predictive for short-term tumor response and local control in head and neck SCC treated with chemoradiation.31,34,49 Another functional MRI diagnostic technique that can be imple­ mented in every MR survey for evaluation of nodal disease (with an additional acquisition time of ≈ 5 minutes) is proton MRS. Metastatic nodal disease seems to have characteristic spectra that could differenti­ ate it from adjacent healthy tissue and immunostimulated lymph nodes, because metastatic nodes present higher choline concentrations due to the neoplastic tissue6 (Fig. 24-21). The technique seems feasible for nodal volumes up to 9 mm3; because of MRS-inherent problems (e.g., inadequate signal reception, motion artifacts, strong lipid signal, shimming), there have been a restricted number of in vivo series in the neck region. Special reference should be made to specific imaging findings of metastatic thyroid carcinoma, which is usually of papillary or follicular type and rarely of anaplastic type. The nodal metastases may have a great variety of imaging findings, including enhancement, scattered calcifications (the differential diagnosis includes lymphomas before or after irradiation or chemotherapy, tuberculosis, metastatic mucinous

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

717

B

*

A

C FIG 24-18  DWI in a patient with bilateral lymphadenopathy, on the right side appearing as neck mass, due to tonsillar carcinoma. A, Fat-suppressed T1-weighted MRI shows the necrotic neck mass (asterisk) with the surrounding inflammatory reaction and extracapsular spread as well as three suspected lymph nodes on the contralateral side (arrows). B, DWI (b value 1000 s/mm2) shows residual hyperintensity on the left nodes (arrows) as well as in the right-sided neck mass, more pronounced in the necrotic (abscess-like) part. C, Calculated ADC values are markedly low, indicating metastatic nodal disease, which was histopathologically confirmed.

adenocarcinomas, prostate metastasis, the result of prior granuloma­ tous disease, and seminomas), or even findings of a benign cyst with or without hemorrhagic areas.44 On MRI, most metastatic nodes have low to intermediate T1-weighted and high T2-weighted signal inten­ sity. High T1-weighted signal may be seen in papillary thyroid carci­ noma nodal metastases because of the high intranodal concentrations of macroglobulin. Hemorrhage will also demonstrate high intensity  in both T1- and T2-weighted imaging. Miscellaneous diseases with adenopathy include Castleman’s disease (angiofollicular lymph node

hyperplasia), sarcoidosis, cat-scratch disease (caused by Bartonella henselae), and Kimura and Kikuchi diseases. Finally, the identification of certain patterns of nodal disease facili­ tates clinical evaluation as follows: 1. Large homogeneous lymph nodes or “foamy”-appearing nodes, sometimes with a thin capsule, are most commonly encoun­ tered in lymphoma, sarcoidosis, and infectious mononucleosis. Cross-sectional imaging cannot reliably differentiate Hodgkin’s from non-Hodgkin’s lymphoma.

http://radiologyebook.com

718

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 24-19  A, Contrast-enhanced CT imaging (early arterial phase) of a patient with an enlarged lymph node (indicated by the free-hand drawn ROI) in level 2 on the left side. The patient underwent perfusion CT, and blood volume (B) as well as permeability (C) were estimated in the lymph node of interest. Blood volume seems not significantly elevated, whereas permeability value is pathologically high. The histopathologic examination showed metastatic disease from ipsilateral oropharyngeal carcinoma.

2. Total or relatively uniform enhancement of an enlarged node is usually encountered in inflammatory processes. Exceptions  to this rule include nodal involvement by Kaposi’s sarcoma  (Fig. 24-22), thyroid carcinoma, and renal cell carcinoma  (Fig. 24-23). 3. A cluster of nodes with a variety of appearances (homogeneous, necrotic, enhancing, and calcified) is most compatible with granulomatous disease of cervical lymph nodes or lymphomas (Fig. 24-24). In urban centers, tuberculosis is the most common cause54 (Fig. 24-25).

NONNODAL NECK MASSES The most common etiologies of nonnodal neck masses include con­ genital lesions and their complications, inflammatory lesions, vascular lesions, neural lesions, and malignant lesions. The role of the radiolo­ gist is to differentiate among these conditions using imaging modali­ ties such as ultrasound (US) with color Doppler, CT, and MRI. CT and MRI can then be used to determine the extent of the lesion and whether local invasion is present. Both CT and MRI can provide mul­ tiplanar images, but MRI will more effectively demonstrate the soft

http://radiologyebook.com

719

CHAPTER 24  Cervical Adenopathy and Neck Masses

Roi1

0.61

0.71

A

0.02

B

0

FIG 24-20  DCE-MRI–derived K

trans

measurements overlaid onto the anatomic fat-suppressed postcontrast T1-weighted images in a patient with lateral lymphadenopathy on the left side. The elevated Ktrans values in the small submental (level 1) node (A) as well as in the necrotic node (level 2) (B) indicate metastatic disease. Note the pronounced heterogeneity in the tumor microenvironment, as depicted in the different degrees of neoangiogenesis (elevated Ktrans) values on tumor rim.

Cho I: 4.21 W: 6.30 0.6

I: Integral W: Width

0.4 Cr I: 1.63 W: 7.00 0.2

0.0

A

B

4.0

3.5

3.0

2.5

ppm

FIG 24-21  Proton MRS of a nodal metastasis. A, Coronal T1-weighted MR scan, used as a localizer for placement of the voxel of interest, shows the extended nodal tumor on the left side with infiltration of the adjacent soft tissue. B, The spectrum of the subsequently performed MRS reveals high choline signal indicating a cell-proliferating disease. The choline-to-creatine integrals ratio is 2.6, which is in essential agreement with the published data.

http://radiologyebook.com

2.0

720

PART II  CT and MR Imaging of the Whole Body

K K K

*

FIG 24-23  Renal cell carcinoma nodal metastasis. Coronal T1-weighted FIG 24-22  Kaposi’s sarcoma. Axial contrast-enhanced CT scan at the

MR scan demonstrates a circumscribed node on the left side (arrowheads) with heterogeneous architecture and a hypointense capsule.

level of the anterior bellies of the digastric muscles demonstrates multiple uniformly enhancing masses in the anterior and posterior triangles of the neck bilaterally, the largest identified in the left spinal accessory chain (asterisk). Multiple oropharyngeal lesions are also present (K).

FIG 24-25  Tuberculous adenitis in an 18-year-old female patient. FIG 24-24  Sarcoidosis of cervical lymph nodes (axial contrast-enhanced fat-saturated T1-weighted MRI). A cluster of nodes (arrows) with a slightly hypointense center and an enhancing rim on the left side.

Coronal contrast-enhanced T1-weighted MR scan demonstrates a circumscribed lymph node (arrow) in the supraclavicular fossa that has a variably homogeneous and partially necrotic parenchyma without infiltrating margins.

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses tissue characteristics of a lesion. The CT and MR appearances of nodal and nonnodal neck masses often overlap. Careful attention to imaging findings and correlation with clinical history are essential to establish the correct diagnosis.

Masses of Developmental Origin Congenital lesions of the neck are usually not difficult to diagnose and involve lymphangiomas (cystic hygromas), branchial cleft cysts, thyro­ glossal duct cysts, and (epi)dermoids.

Lymphangioma.  Cystic hygroma is the most common form of lymphangiomatous malformation occurring in the neck (other types include cavernous lymphangioma, capillary lymphangioma, and vas­ culolymphatic malformation) (Figs. 24-26 and 24-27).2 Cystic hygroma is a nonencapsulated lymphatic rest that is left behind during embryo­ genesis. It typically presents at birth or within first 2 years of life and may occasionally be associated with a more generalized failure of

S

FIG 24-26  Cystic hygroma. Axial contrast-enhanced CT scan at the level of the first tracheal ring demonstrates a multilocular cystic mass without rim enhancement occupying the entire right posterior triangle extending toward the right axilla. S, Sternocleidomastoid muscle.

lymphatic system development, as is seen in cases of the 45,X chromo­ some karyotype (Turner’s syndrome). Its imaging appearance is that of a uniloculated or multiloculated nonenhancing mass, without significant mass effect, that insinuates between normal structures. Its CT appearance is a hypodense nonen­ hancing poorly circumscribed cystic mass. On MRI, however, it appears as a low T1-weighted and high T2-weighted signal mass with no or subtle rim enhancement, with the latter implicating venous or heman­ giomatous components. Finally, hyperintensity in T1-weighted images may suggest hemorrhage. Lingual thyroid tissue can also be observed as a lobular mass in the midline sublingual space and presents with the aforementioned features. Because up to 10% of cervical cystic hygro­ mas may extend into the mediastinum, imaging is performed to assess the extent of the lesion and its relationship to adjacent structures before surgical resection.

Branchial Cleft Cyst.  Special attention has to be given to the diag­ nosis of the second branchial cyst (the Bailey type II cyst, which is secondary to persistence of the embryologic cervical sinus) as the most common developmental mass identified on imaging and the most common branchial cleft anomaly.5 It appears as an intermittent soft, painless mass along the course of the anterior margin of the sterno­ cleidomastoid muscle and affects children, teenagers, and young adults. The classic location of the second branchial cyst, which is actually a branchial apparatus remnant, is in the posterior submandibular space at the angle of mandible (Figs. 24-28 and 24-29). In contrast, the type III second branchial cleft cyst courses between the external and internal carotid arteries. The most specific imaging characteristic of the lesion is displacement of the submandibular gland anteromedially, the carotid space medially, and the sternocleidomastoid muscle posterolaterally.73 The lesions appear cystic on CT imaging and show a bright signal on T2-weighted images. T1-weighted signal is variable depending on the protein content of the cyst. Enhancement is usually present in infected cysts, in which case cellulitis may appear in the surrounding soft tissues. Septations are uncommon and are usually found only in cases of previous infection or needle aspiration. Infected cysts may mimic

S PT

C

* A

721

V

*

B

FIG 24-27  Cystic hygroma. A, Sagittal T1-weighted MRI demonstrates a lobulated posterior triangle mass (asterisk) that is moderately hyperintense to muscle (C). B, The mass (asterisk) is located deep to the posterior margin of the sternocleidomastoid muscle (S). Note how the hygroma tracks posterior to the internal jugular vein (V). The hygroma appears virtually isointense to fat in the right posterior triangle (PT).

http://radiologyebook.com

722

PART II  CT and MR Imaging of the Whole Body

necrotic jugular chain lymph nodes or cervical abscesses on both CT and MRI.

Thyroglossal Duct Cyst.  The persistence of the thyroglossal duct (normally it involutes during the eighth to tenth week) gives rise to midline (75%) or paramedian (<2 cm) cysts suprahyoidally (20%), at the level of the hyoid (15%), or anterior to the strap muscles infrahy­ oidally (65%) (Fig. 24-30). The more common infrahyoid thyroglossal duct cysts often have both midline and off-midline components, the latter being embedded in the strap muscles. It is the location of an infrahyoid thyroglossal duct cyst in the strap muscles—not its signal or attenuation characteristics—that allows for a definitive radiographic

diagnosis. Thyroglossal duct cyst is one of the most common congeni­ tal masses; it usually presents before the age of 20 and occurs four to five times more frequently in females than males. Thyroglossal duct anomalies can be associated with thyroid carcinoma, usually of the papillary type. Nodularity or a soft tissue mass within the cyst is sug­ gestive for malignancy. Routine sectional imaging of thyroglossal duct cysts is not needed in children because the clinical presentation (a painless midline neck mass that moves on swallowing) is typical.2,14 Ultrasonography or radioisotope scanning may be performed to confirm a normal position of the thyroid before surgery. CT or MRI is usually performed preop­ eratively in cases of suspected thyroglossal duct cyst in adults to confirm the diagnosis and exclude other nodal masses. Contrastenhanced CT imaging reveals a well-circumscribed low-density mass. Peripheral rim enhancement or internal septations are occasionally seen. The MR signal characteristics of thyroglossal duct cysts are the same as those of branchial cleft cysts or small cystic hygromas.2

Dermoid Cyst.  Dermoid cysts are the rarest of the congenital neck G

S

lesions. Only 7% of all dermoid tumors occur in the neck. The majority of head and neck dermoids occur in the orbit, nasal cavity, or oral cavity. Dermoid cyst is a general term that includes epidermoid, dermoid, and teratoid cysts. Most cervical dermoid cysts occur in the midline or just off midline and are usually present at birth (Fig. 24-31). In contrast to thyroglossal duct cysts, dermoid cysts typically do not move with swallowing. A distinct imaging characteristic of a dermoid cyst is the fat or fluid components that result in very bright T2-weighted signal and a low to high T1-weighted signal with an absence of contrast enhancement. In CT imaging, the fatty parts of the cyst are hypodense.29

Masses of Inflammatory Origin FIG 24-28  Type II second branchial cleft cyst. Axial contrast-enhanced CT scan demonstrates a unilocular cystic mass with minimal rim enhancement along the anterior margin of the left sternocleidomastoid muscle (S), displacing the left submandibular gland (G) anteriorly.

It is often difficult to determine clinically whether a patient with an erythematous, painful, and swollen neck has a simple cellulitis, an infected retention cyst, a cervical manifestation of a systemic inflam­ matory disease (e.g., sarcoidosis), or a potentially life-threatening deep neck abscess. With sectional imaging, in many cases it is possible not only to diagnose an abscess but also to differentiate it.

*

S

S

A

B

FIG 24-29  Type II second branchial cyst. A, Sagittal T1-weighted MRI demonstrates a well-circumscribed anterior triangle mass (asterisk) mildly hyperintense to muscle along the anterior margin of the sternocleidomastoid muscle (S). B, Axial T2-weighted MRI demonstrates a hyperintense mass along the anterior margin of the left sternocleidomastoid muscle (S). Compare with Figure 24-26.

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

FIG 24-30  Thyroglossal duct cyst. Sagittal TIRM-weighted MRI at a level immediately inferior to the hyoid bone demonstrates a welldefined cystic midline anterior triangle mass (arrow) located under the subcutaneous fat tissue, extending from the base of the tongue to the suprahyoidal level.

723

FIG 24-32  Calculous disease in the submandibular space. Axial contrast-enhanced CT scan demonstrates obstructing calculi (arrow) in the anterior submandibular duct associated with enlarged submandibular gland (arrowheads) on the left side.

and drainable pus, for which the term abscess is more appropriate. In cases of CT imaging, a precontrast scan is not essential, the images should also be reviewed in bone algorithm, and thin multiplanar reconstructions are necessary. MRI demonstrates an enhancing sub­ mandibular gland (do not forget the fat saturation) and offers the advantage of MR sialography.60 Fat infiltration or gland size reduction, compared to the healthy side, in MRI may also indicate chronic disease with sialadenitis.68

Ranula.  The second inflammatory lesion of the sublingual space is

FIG 24-31  Dermoid cyst. Axial contrast-enhanced CT scan demonstrates a midline fluid-containing mass without peripheral rim enhancement in the floor of the mouth (arrows).

the ranula (Fig. 24-33). The lesion is actually a mucous retention cyst that may result from an obstructed sublingual or minor salivary gland or from trauma or retained secretions that may lead to rupture of acini. Notably a ranula may herniate through the mylohyoid muscle into the submandibular or lower parapharyngeal space and evolve in a plung­ ing (diving) ranula. Ranulas have homogeneous high T2- and low T1-weighted signals. The multiplanar capacity of MRI may facilitate documentation of a tail of tissue within the floor of the mouth. On CT, ranulas have fluid attenuation, and the cyst wall may or may not enhance, depending on the history of infection. Notably a unilocular cystic hygroma or an epidermoid cyst cannot be differentiated radio­ logically from a simple ranula. A diving ranula, however, has a distinct “tail sign” within the sublingual space, with the bulk of the cyst seen in the submandibular space.

Calculous Disease.  Calculous disease is the most common cause of

Abscess in Inflammatory Diseases.  Inflammatory lesions in the

salivary inflammatory lesions, especially in the submandibular space, owing to the upward secretion of the more mucinous, viscous, and alkaline saliva. Patients with calculous disease have painful glands, and calculous disease is often associated with fever. Cross-sectional imaging should be performed (Fig. 24-32). CT imaging is preferred to MRI for identifying calculi, but MRI can also be applied to image mandibular infection (osteomyelitis) and to differentiate between cellulitis/myositis

parapharyngeal space, including the carotid sheath, are cervical  adenitis coming from adjacent adenitis or mucosal space infections. Tuberculous adenitis and suppurative adenopathy are associated with pharyngitis and with enlarged lymph nodes with a low-attenuation center (suggestive of liquefaction necrosis) or perimucosal abscesses. On CT imaging, abscesses appear as single or multiloculated lowdensity masses that conform to fascial spaces; however, a necrotic

http://radiologyebook.com

724

PART II  CT and MR Imaging of the Whole Body

* *

R

A

D

D

*

S

M R

FIG 24-34  Metastatic lymph node mimicking abscess. Axial contrastG

R

B FIG 24-33  Ranula (R). A, Axial contrast-enhanced CT scan at the level of the thyroid cartilage demonstrates a unilocular cystic anterior triangle mass deforming the anterior margin of the right sternocleidomastoid muscle (S). On the basis of this single image, differential diagnosis would include second branchial cleft cyst, lymphangioma, and necrotic lymph node. Thyroglossal duct cyst should be excluded from the differential diagnosis because the mass is extrinsic to the strap muscles (asterisks). B, Axial contrast-enhanced CT image at the level of the mandible demonstrates the origin of the ranula in the posterior oral cavity, medial to the enhancing mylohyoid muscle (arrows). M, Right mylohyoid muscle; G, right submandibular gland.

tumor, nodal disease with extension into the adjacent soft tissues, and a thrombosed vessel may mimic an abscess on sectional images (Fig. 24-34). Abscesses demonstrate peripheral rim enhancement. Cutane­ ous and subcutaneous manifestations of infection include enlargement of adjacent muscles (myositis), thickening of the overlying skin, oblit­ eration of the fat planes, and enhancement with thickening of the fascial planes. The presence of these cutaneous and subcutaneous manifestations without a low-density fluid collection is consistent  with cellulitis (although inflammatory reactions after radiotherapy or surgery can mimic cervical infections). Cellulitis can easily spread to adjacent fascial spaces, upward or downward. It is possible that a patient may be studied before a cellulitis has “ripened” to a frank col­ lection of pus. If a diagnosis of cellulitis is made and the patient shows no response to appropriate antibiotic therapy, a second examination should be performed to ensure that an abscess cavity has not formed. A rare complication of rapidly progressive neck infection that can hardly be differentiated from cellulitis is necrotizing fasciitis (if gas is present, it aids diagnosis). On the other hand, abscesses conform to fascial spaces and can occur anywhere in the neck. In MRI, abscesses appear as single or multiloculated masses that are hypointense to skeletal muscle on T1-weighted images and markedly hyperintense on T2-weighted images (Fig. 24-35). The use of MRI contrast agents should also demonstrate an enhancing rim in a mature

enhanced CT scan demonstrates a low-density right submandibular triangular mass (asterisk) surrounded by an irregular rim of contrast enhancement. The overlying platysma muscle is thickened (arrows). The patient, who was afebrile, had a history of treatment for squamous cell carcinoma of the right oropharynx 2 years before this examination. Needle aspiration of the mass yielded atypical squamous cells. D, digastric muscle.

FIG 24-35  Abscess. Axial TIRM-weighted MRI demonstrates a mass in the left submandibular triangle that is hyperintense to the muscle tissue and has a mass affect in the floor of the mouth (arrow). Note that inflammatory changes in the fat lateral to the abscess (arrowhead) are easily appreciated on MR scans.

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses abscess. There are several considerations when interpreting contrastenhanced MRI, however. First, the cutaneous manifestations of infec­ tion are not easily obvious on MRI; second, subcutaneous manifestations of infection may be obscured on MRI if imaging parameters are  not carefully monitored; third, after intravenous administration of contrast agent, the enhancing rim may become isointense with sur­ rounding fat, limiting detection or delineation of the abscess unless fat-suppression techniques are used; fourth, both areas of abscess for­ mation and cellulitis tend to be hyperintense on T2-weighted images, hampering differentiation between the two processes. Thus we recom­ mend the use of contrast-enhanced CT to assess suspected inflamma­ tory disease in adults. In children and in adults whose clinical status precludes use of iodinated contrast material, MRI is the study of initial choice, although a limited noncontrast CT study may be required for evaluation of osseous integrity. If MRI is not available, ultrasonogra­ phy is an alternative modality for localizing suspected collections for surgical drainage.

Thrombophlebitis.  Other inflammatory lesions include thrombo­ phlebitis of the internal jugular vein associated either with mucosal or peritonsillar inflammation (Ludwig’s angina) or with iatrogenic causes. A halo of edema around the thrombosed vein and enhancement of the vessel and the perivenous soft tissue with fat-suppressed MRI are typical findings (Fig. 24-36).

Osteomyelitis.  Osteomyelitis of the jaw is another important dif­ ferential diagnosis in inflammatory masses of the neck. Osteomyelitis may develop as a result of odontogenic infection, focal injury after tooth extraction, or complicated fracture or may occur in an immu­ nocompromised patient (e.g., in a patient with multiple myeloma). It may present radiographically as suppurative, sclerosing, or tuberculous and may present with periostitis. The MR appearance of the osteomy­ elitis may depend on the patient’s age, type of infecting organism, and

FIG 24-36  Internal jugular vein thrombosis. Axial contrast-enhanced CT scan at the level of the thyroid gland demonstrates hypodensity (straight arrow) lateral to the right common carotid artery compared with the hyperattenuated jugular vein on the left side (curved arrow). Note the effacement of the adjacent soft tissue on the right side (arrowhead) indicating inflammatory changes in this region.

725

underlying medical conditions. The bone marrow may have a normal appearance in the first days, but later, low signal intensity on T1-weighted images and high signal intensity on T2-weighted images are characteristic of the disease. Soft tissue involvement with edema and abnormal enhancement is also common in osteomyelitis. However, this pattern is also found in osteoradionecrosis in irradiated patients or in bisphosphonate-induced osteonecrosis.8,16

Masses of Vascular Origin Neck abnormalities of vascular origin are associated with partially or totally occluded vessels and may present a diagnostic problem on sec­ tional MRI; care has to be taken not to confuse flow-related enhance­ ment with thrombosis. The MR diagnosis of venous thrombosis often requires gradient echo and short flip angle imaging or phase imaging to confirm findings identified on spin echo sequences.12 In this case, the clotted or slowly flowing blood will give high signal intensity. Whereas internal jugular vein thrombosis can be iatrogenic or result from prior surgery and irradiation, occlusion of the internal carotid artery (ICA) is secondary to atherosclerotic disease or surgical trauma. Aneurysm of the ICA can give the impression of a cystic mass. Helpful characteristics for differentiation of aneurysm from schwanno­ mas or other rare tumors are peripheral calcification (e.g., calcified lymph node vs. papillary thyroid carcinoma) and obliteration of the ICA lumen. CT and MR angiography (MRA) can confirm the diagno­ sis of aneurysm. Dissection of the extracranial ICA presents commonly with head­ ache and focal neurologic deficits (ocular or lower cranial nerve palsies)46 and may have numerous causes, including minor trauma, fibromuscular dysplasia, and Marfan’s syndrome. Catheter angiogra­ phy is the state-of-the-art method of imaging an ICA dissection, but CT and MRI provide a noninvasive means of disease evaluation. The intramural hematoma demonstrates a CT signal that is hypodense to muscle and a hyperintense T1- and T2-weighted signal. Enhanced CT or fat-suppressed T1-weighted MRI demonstrates peripheral enhance­ ment of the vasa vasorum. Another common cause of nonnodal neck masses are paraganglio­ mas, which can occur extracranially in two different varieties. The most common form is the glomus vagale, which arises from the paragangli­ onic cells about the nodose ganglion, located just below the skull base. The second type of paraganglioma is the glomus caroticum, which originates from the carotid body cells near the carotid bifurcation (Figs. 24-37 and 24-38). These hypervascular tumors can occur spo­ radically or as autosomal dominant familial tumors. They present a significant enhancement (on CT and MR scans), which may be inho­ mogeneous in large lesions with necrosis and hemorrhage. Dynamic scanning reveals a sharp filling peak and a rapid washout of the tumor with characteristic “dropout” effect in the dynamic MRI. A “salt-andpepper” MR appearance has been traditionally attributed to paragan­ gliomas, owing to scattered foci of signal voids (mostly in lesions >2 cm), which represent larger-caliber vessels. Rapid-flow dominant tumor vessels in all sequences are thought to be specific (in conjunc­ tion with the T2-weighted signal profile of the lesion) for the tumor, so that contrast enhancement may be unnecessary. However, other neck masses, such as nodal metastases from renal cell or thyroid car­ cinoma, may display areas of signal void on MRI. Some venous hemangiomas may have a similar hypervascular imaging appearance with flow voids (Fig. 24-39). A key feature for the differential diagnosis is lack of displacement of the adjacent vessels. Carotid body tumors arise in the juxtahyoid neck and separate  the internal and external carotid arteries on sectional imaging.  Glomus vagale tumors arise in the suprahyoid neck and displace the internal carotid artery anteromedially and the internal jugular vein

http://radiologyebook.com

726

PART II  CT and MR Imaging of the Whole Body

A

B FIG 24-37  Paraganglioma of the neck. A, Axial TIRM-weighted MRI demonstrates a heterogeneous mass (arrows) at the bifurcation of the common carotid artery splaying the external carotid artery anteriorly and the internal carotid artery posteriorly. B, Axial T1-weighted MRI at the same level demonstrates a left anterior triangle mass (arrows) with an internal anterior area of signal void.

*

FIG 24-38  Paraganglioma of the neck. Sagittal MIP (maximum intensity projection) reconstruction of contrast-enhanced CT imaging demonstrates an enhancing homogeneous mass (asterisk) at the bifurcation of the common carotid artery splaying the external and internal carotid (with calcified plaques) arteries.

FIG 24-39  Venous hemangioma. Axial contrast-enhanced fat-saturated MR scan demonstrates an avidly enhancing lesion in the left submental triangle (arrowheads) without infiltration of the adjacent muscle tissue and the submandibular gland.

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses posterolaterally. When there is gross asymmetry in the splaying of the internal and external carotid arteries, the possibility of a vascular-type nodal metastasis or even a schwannoma of the cervical sympathetic chain (usually lateral to the carotid vessels) should be considered. Finally, MR findings after radiation therapy of the paragangliomas include variable alteration in T2 signal, decreased heterogeneous enhancement, and a reduction in flow voids.48 For imaging paragangliomas and their feeders, 3D time-of-flight (TOF) MRA is superior to 3D phase-contrast angiography and 2D TOF MRA, but like CT angiography it cannot replace conventional digital subtraction angiography, especially for carotid body tumors.74

Masses of Neural Origin Schwannomas and neurofibromas are the most common types of neu­ rogenic tumors found in the head and neck. Common sites for schwannomas and neurofibromas in the neck are the vagus nerve, less commonly the glossopharyngeal nerve, the ventral and dorsal cervical nerve roots, the cervical sympathetic chain, and the brachial plexus (Fig. 24-40). When small, schwannomas rarely cause symptoms. When they are large, associated motor dysfunction and pain in the distribu­ tion of a sensory nerve are varying clinical findings because, for example, salivary gland tumors tend to invade nerves. Because the cervical sympathetic chain runs in a relatively loose fascial compart­ ment, compression, such as is seen in other schwannomas, is exceed­ ingly rare. Without a clinical history of neurofibromatosis, it is impossible to distinguish between schwannomas and neurofibromas on sectional imaging. The most demanding aspect of treatment of these lesions is distinguishing the benign cervical sympathetic schwannoma from other pathology in the parapharyngeal space. The differential diagnosis of a parapharyngeal space mass is based on the division of the space into pre- and poststyloid compartments. The prestyloid compartment contains the parotid gland, fatty tissue, and lymph nodes. It has the highest incidence of tumors. The poststyloid compartment contains

727

the carotid sheath with the sympathetic chain and cranial nerves IX through XII. Thus masses arising in the poststyloid compartment include carotid artery lesions, paragangliomas arising from the vagus nerve or the carotid body, neurogenic tumors involving cranial nerves IX to XII, or sympathetic chain neurogenic lesions. Most schwannomas are fairly homogeneous soft tissue masses and appear hypodense or isodense to skeletal muscle on noncontrast CT and tend to be hypoin­ tense or isointense to skeletal muscle on T1-weighted and variably hyperintense on T2-weighted MRI. Despite their hypovascularity, they enhance significantly on both CT and MRI and can mimic a paragan­ glioma. Enhancement of schwannomas is seen at least 2 minutes after contrast injection and depicts the equilibrium phase of the contrast agent and the poor venous drainage of the tumor. Dynamic scans can reveal the true nature of the lesion and differentiate it from hypervas­ cular lesions. The enhancement pattern of neural tumors can vary; it may be an inhomogeneous enhancement (owing to necrosis and hem­ orrhage) or even lack of enhancement.20 There are some useful hints to establish the diagnosis of a neural tumor in the neck.65 For example, a tumor arising from the vagus nerve manifests as a mass in the anterior triangle, displacing the internal or common carotid artery anteromedially and the internal jugular vein posterolaterally (Fig. 24-41), whereas a neoplasm of the cervical nerve roots manifest as a mass in the posterior triangle, which may extend into one of the neural foramina of the cervical spine (Fig. 24-42). Sympathetic chain tumors demonstrate a constant relationship with the longus colli muscles, and lesions of the cervical sympathetic chain, as they grow and expand, tend to push the common carotid or internal and external carotids anteriorly. Brachial plexus schwannomas and neurofibromas in the infrahyoid neck often displace the anterior scalene muscle anteriorly.

Masses of Mesenchymal Origin Lipomatous tumors such as lipoma, lipoblastoma, and liposarcoma are the most common cervical neoplasms of mesenchymal origin (Fig. 24-43).63 They typically manifest as painless masses, occurring most commonly in the midline of posterior neck or the posterior triangle. Cervical lipoblastoma typically presents as a rapidly enlarging mass.13 Although they are generally asymptomatic, there are cases of lipoma­ tous tumors with respiratory compromise, two cases of Horner’s syn­ drome, a case of hemiparesis, and a case of upper extremity weakness caused by compression of the spinal cord.13

AL

5cm

FIG 24-40  Schwannoma. Coronal contrast-enhanced MR scan demonstrates an enhancing heterogeneous mass (arrows) medial to the right sternocleidomastoid muscle in the parapharyngeal space near the skull base.

FIG 24-41  Vagal schwannoma. Sagittal reconstruction of a contrastenhanced CT scan demonstrates a predominantly middle-density mass (arrows) medial to the left sternocleidomastoid muscle, without signs of soft tissue infiltration. Note the site of the schwannoma in relation to the jugular foramen at the level of the skull base.

http://radiologyebook.com

728

PART II  CT and MR Imaging of the Whole Body

S

*

A

*

B FIG 24-42  Dorsal cervical nerve root schwannoma. A, Axial contrastenhanced CT scan at the level of the thyroid cartilages demonstrates a rim-enhancing solid posterior triangle mass (asterisk) deep to the left sternocleidomastoid muscle (S). On this single image, differentiation of neural tumor from enhancing lymph node would be impossible. B, Axial contrast-enhanced CT scan at the level of the hyoid bone demonstrates that the mass (asterisk) involves the neural foramen (arrows), allowing the correct radiologic diagnosis of neural tumor.

CT can be used to demonstrate the presence of a mass (homoge­ neous nonenhancing mass isodense to subcutaneous fat), but MRI is particularly helpful in its ability to suggest the histologic components of the tumor.55 Lipomas may or may not have a demonstrable capsule on sectional images. Lipoblastoma typically appears hyperintense on both T1- and T2-weighted images, although it is consistently less intense than mature fat on T1-weighted images, likely owing to its fibrous septa and the variable degree of lipomatous differentiation. Lipocytes, or mature fat cells, exhibit relatively high signal intensity  on T1-weighted images, whereas lipoblasts show lower intensity.  The degree of intensity on T2-weighted images also varies, primarily depending on the amount of myxoid and fibrous components. Myxoid stroma and lipoblasts exhibit high intensity on T2-weighted images. The technique of fat suppression sequencing is particularly valuable in the assessment of lipoblastoma in its ability to demonstrate the pres­ ence of fatty tissue even when standard T1 and T2 images fail to demonstrate sufficient hyperintensity of the tumor to indicate such content. Liposarcoma is one of the most frequent malignant soft tissue tumors in adults. It should be suspected whenever a large (>5 cm) fat-containing soft tissue tumor is found, especially when it is clinically

associated with pain and rapid growth. Well-differentiated liposar­ coma displays a hypointense fat intensity (≤20 HU) mass on CT; the tumor shows signal intensities similar to those of subcutaneous normal fat on MRI. Poorly differentiated liposarcoma displays tissue density on CT and low signal intensity on spin echo T1-weighted and high signal intensity on T2-weighted MRI with intense early enhancement on dynamic acquisition contrast-enhanced images. Ill-defined borders, and especially nonfatty heterogeneous content, indicate the high  probability of malignant tumor. Other sarcomas that may be present are rhabdomyosarcoma, fibro­ sarcoma, osteosarcoma, chondrosarcoma, malignant fibrous histiocy­ toma, leiomyosarcoma, alveolar soft part sarcoma, Ewing’s sarcoma, and synovial sarcoma. The imaging findings of these entities frequently are nonspecific. However, imaging, particularly MRI, has a major role in defining the extent of these tumors. This is important because com­ plete surgical excision is the preferred method of treatment. Imaging also is useful in planning radiation therapy and determining prognosis. Sarcomas tend to destroy bone and distort soft tissue planes. The ramus of the mandible is the most common site of Ewing’s sarcoma of the head and neck. The typical manifestations are rapidly growing mass with associated pain and local paresthesias. Mandibular involve­ ment may cause additional associated symptoms, such as loosening  of teeth and otitis media. Images of Ewing’s sarcoma and osteogenic sarcoma show large soft tissue masses that originate in the marrow, permeate and destroy the cortical bone, and extend into the masticator space. The extent of bone marrow involvement is best evaluated with T1-weighted imaging, whereas T2-weighted imaging is of value for defining the margins of the tumor and its relation to neurovascular bundles and adjacent muscle. Desmoids and other benign mesenchymal tumors tend to have sharply circumscribed borders and typically do not infiltrate adjacent soft tissue or destroy bone. They may exhibit local aggressiveness in a small subset of patients. The superiority of CT in depicting cortical bone involvement is evident; however, the true extent of the infiltrative process within the trabecular bone network is better assessed with MRI showing displacement of the normal fatty marrow.

Masses Arising from the Aerodigestive Tract Masses detected on sectional images of the neck may represent cervical extensions of diseases arising within the oral cavity, larynx, or hypo­ pharynx. A mucus-filled cyst may be present in the glossoepiglottic fold. More often a laryngocele, which presents an abnormal dilatation of the saccule (appendix) of the laryngeal ventricle, is found in crosssectional imaging (Fig. 24-44). Although a small proportion of laryn­ goceles are congenital, the majority are the result of chronic increases in intralaryngeal pressure. Neoplasms or localized inflammation and edema may occasionally obstruct the ventricle, and imaging is required to exclude this situation. The resulting ball-valve mechanism traps air within the saccule. Mucus produced by the respiratory epithelium of the ventricle may result in a fluid-filled laryngocele (laryngeal muco­ cele). Internal laryngoceles are limited to the paralaryngeal space. External or combined internal and external laryngoceles extend from the paralaryngeal space into the anterior triangle of the neck via fen­ estrations in the thyrohyoid membrane. On CT or MRI, the external component of a laryngocele appears as a thin-rimmed fluid-filled or air-filled mass directly lateral to the thyrohyoid membrane. Continuity with the internal component can be demonstrated on axial or direct coronal images. A pharyngocele is an abnormal dilatation of the pyriform sinus, which may also herniate through the thyrohyoid membrane to mani­ fest in the anterior triangle. The less common pharyngocele is differ­ entiated from a laryngocele by the demonstration of continuity with

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

A

729

B FIG 24-43  Lipoblastoma. A, Coronal T1-weighted MRI at the level of the supraglottis in an adolescent patient demonstrates a fat-intensity mass with variable degree of fat content in its center (arrow) originating from the right pyriform sinus. B, Axial contrast-enhanced fat-saturated MR scan shows a hypointense mass (arrow) after fat saturation, with enhanced wall.

the pyriform sinus rather than the laryngeal ventricle. Neoplasms or abscesses involving the apex of the pyriform sinus may extend into the anterior triangle of the infrahyoid neck through the cricothyroid membrane.

REFERENCES

FIG 24-44  Mixed (external) laryngocele. Axial contrast-enhanced fatsaturated MR scan shows an air-filled, thin-walled, space-occupying lesion in the right lower submandibular space mass (arrow).

1. Abdel Razek AA, Soliman NY, Elkhamary S, et al: Role of diffusionweighted MR imaging in cervical lymphadenopathy. Eur Radiol 16:1468–1477, 2006. 2. Ahuja AT, Wong KT, King AD, et al: Imaging for thyroglossal duct cyst: The bare essentials. Clin Radiol 60:141–148, 2005. 3. Ash L, Teknos TN, Gandhi D, et al: Head and neck squamous cell carcinoma: CT perfusion can help noninvasively predict intratumoral microvessel density. Radiology 251:422–428, 2009. 4. Babbel RW, Smoker WR, Harnsberger HR: The visceral space: The unique infrahyoid space. Semin Ultrasound CT MR 12:204–223, 1991. 5. Benson MT, Dalen K, Mancuso AA, et al: Congenital anomalies of the branchial apparatus: Embryology and pathologic anatomy. Radiographics 12:943–960, 1992. 6. Bisdas S, Baghi M, Huebner F, et al: In vivo proton MR spectroscopy of primary tumours, nodal and recurrent disease of the extracranial head and neck. Eur Radiol 17:251–257, 2007. 7. Bisdas S, Baghi M, Smolarz A, et al: Quantitative measurements of perfusion and permeability of oropharyngeal and oral cavity cancer, recurrent disease, and associated lymph nodes using first-pass contrastenhanced computed tomography studies. Invest Radiol 42:172–179, 2007. 8. Bisdas S, Chambron-Pinho N, Smolarz A, et al: Biphosphonate-induced osteonecrosis of the jaws: CT and MRI spectrum of findings in 32 patients. Clin Radiol 63:71–77, 2008.

http://radiologyebook.com

730

PART II  CT and MR Imaging of the Whole Body

9. Bisdas S, Medov L, Baghi M, et al: A comparison of tumor perfusion assessed by deconvolution-based analysis of dynamic contrast-enhanced CT and MR imaging in patients with squamous cell carcinoma of the upper aerodigestive tract. Eur Radiol 18:843–850, 2008. 10. Bisdas S, Rumboldt Z, Surlan-Popovic K, et al: Perfusion CT in squamous cell carcinoma of the upper aerodigestive tract: Long-term predictive value of baseline perfusion CT measurements. AJNR Am J Neuroradiol 31:576–581, 2010. 11. Bisdas S, Surlan-Popovic K, Didanovic V, et al: Functional CT of squamous cell carcinoma in the head and neck: Repeatability of tumor and muscle quantitative measurements, inter- and intra-observer agreement. Eur Radiol 18:2241–2250, 2008. 12. Braun IF, Hoffman JC, Jr, Malko JA, et al: Jugular venous thrombosis: MR imaging. Radiology 157:357–360, 1985. 13. Brodsky JR, Kim DY, Jiang Z: Cervical lipoblastoma: Case report, review of literature, and genetic analysis. Head Neck 29:1055–1060, 2007. 14. Brousseau VJ, Solares CA, Xu M, et al: Thyroglossal duct cysts: Presentation and management in children versus adults. Int J Pediatr Otorhinolaryngol 67:1285–1290, 2003. 15. Choi SH, Han MH, Moon WK, et al: Cervical lymph node metastases: MR imaging of gadofluorine M and monocrystalline iron oxide nanoparticle-47 in a rabbit model of head and neck cancer. Radiology 241:753–762, 2006. 16. Chong J, Hinckley LK, Ginsberg LE: Masticator space abnormalities associated with mandibular osteoradionecrosis: MR and CT findings in five patients. AJNR Am J Neuroradiol 21:175–178, 2000. 17. Curvo-Semedo L, Diniz M, Migueis J, et al: USPIO-enhanced magnetic resonance imaging for nodal staging in patients with head and neck cancer. J Magn Reson Imaging 24:123–131, 2006. 18. Davis WL, Smoker WR, Harnsberger HR: The normal and diseased infrahyoid retropharyngeal, danger, and prevertebral spaces. Semin Ultrasound CT MR 12:241–256, 1991. 19. Don DM, Anzai Y, Lufkin RB, et al: Evaluation of cervical lymph node metastases in squamous cell carcinoma of the head and neck. Laryngoscope 105:669–674, 1995. 20. Eldevik OP, Gabrielsen TO, Jacobsen EA: Imaging findings in schwannomas of the jugular foramen. AJNR Am J Neuroradiol 21:1139–1144, 2000. 21. Eskey CJ, Robson CD, Weber AL: Imaging of benign and malignant soft tissue tumors of the neck. Radiol Clin North Am 38:1091–1104, xi, 2000. 22. Feinmesser R, Freeman JL, Noyek AM, et al: Metastatic neck disease. A clinical/radiographic/pathologic correlative study. Arch Otolaryngol Head Neck Surg 113:1307–1310, 1987. 23. Fischbein NJ, Noworolski SM, Henry RG, et al: Assessment of metastatic cervical adenopathy using dynamic contrast-enhanced MR imaging. AJNR Am J Neuroradiol 24:301–311, 2003. 24. Gandhi D, Chepeha DB, Miller T, et al: Correlation between initial and early follow-up CT perfusion parameters with endoscopic tumor response in patients with advanced squamous cell carcinomas of the oropharynx treated with organ-preservation therapy. AJNR Am J Neuroradiol 27:101–106, 2006. 25. Gor DM, Langer JE, Loevner LA: Imaging of cervical lymph nodes in head and neck cancer: The basics. Radiol Clin North Am 44:101–110, viii, 2006. 26. Hoefling N, McHugh J, Light E, et al: Human papillomavirus, p16, and epidermal growth factor receptor biomarkers and CT perfusion values in head and neck squamous cell carcinoma. AJNR Am J Neuroradiol 34:1062–1066, 2013. 27. Huda W, Lieberman KA, Chang J, et al: Patient size and x-ray technique factors in head computed tomography examinations. II. Image quality. Med Phys 31:595–601, 2004. 28. Hudgins PA: Contrast enhancement in head and neck imaging. Neuroimaging Clin N Am 4:101–115, 1994. 29. Imhof H, Czerny C, Hormann M, et al: Tumors and tumor-like lesions of the neck: From childhood to adult. Eur Radiol 14(Suppl 4):L155– L165, 2004. 30. Ishikawa M, Anzai Y: MR imaging of lymph nodes in the head and neck. Magn Reson Imaging Clin N Am 10:527–542, 2002.

31. Jansen JF, Schöder H, Lee NY, et al: Noninvasive assessment of tumor microenvironment using dynamic contrast-enhanced magnetic resonance imaging and 18F-fluoromisonidazole positron emission tomography imaging in neck nodal metastases. Int J Radiat Oncol Biol Phys 77:1403–1410, 2010. 32. Jo SY, Wang PI, Nör JE, et al: CT perfusion can predict overexpression of CXCL8 (interleukin-8) in head and neck squamous cell carcinoma. AJNR Am J Neuroradiol 34:2338–2342, 2013. 33. Keberle M, Tschammler A, Hahn D: Single-bolus technique for spiral CT of laryngopharyngeal squamous cell carcinoma: Comparison of different contrast material volumes, flow rates, and start delays. Radiology 224:171–176, 2002. 34. Kim S, Loevner LA, Quon H, et al: Prediction of response to chemoradiation therapy in squamous cell carcinomas of the head and neck using dynamic contrast-enhanced MR imaging. AJNR Am J Neuroradiol 31:262–268, 2010. 35. King AD, Lei KI, Ahuja AT: MRI of neck nodes in non-Hodgkin’s lymphoma of the head and neck. Br J Radiol 77:111–115, 2004. 36. King AD, Tse GM, Ahuja AT, et al: Necrosis in metastatic neck nodes: Diagnostic accuracy of CT, MR imaging, and US. Radiology 230:720– 726, 2004. 37. King AD, Tse GM, Yuen EH, et al: Comparison of CT and MR imaging for the detection of extranodal neoplastic spread in metastatic neck nodes. Eur J Radiol 52:264–270, 2004. 38. King AD, Yeung DK, Ahuja AT, et al: Human cervical lymphadenopathy: Evaluation with in vivo 1H-MRS at 1.5 T. Clin Radiol 60:592–598, 2005. 39. Koc O, Paksoy Y, Erayman I, et al: Role of diffusion weighted MR in the discrimination diagnosis of the cystic and/or necrotic head and neck lesions. Eur J Radiol 62:205–213, 2006. 40. Layland MK, Sessions DG, Lenox J: The influence of lymph node metastasis in the treatment of squamous cell carcinoma of the oral cavity, oropharynx, larynx, and hypopharynx: N0 versus N. Laryngoscope 115:629–639, 2005. 41. Mack MG, Balzer JO, Straub R, et al: Superparamagnetic iron oxideenhanced MR imaging of head and neck lymph nodes. Radiology 222:239–244, 2002. 42. Mancuso AA, Harnsberger HR, Muraki AS, et al: Computed tomography of cervical and retropharyngeal lymph nodes: Normal anatomy, variants of normal, and applications in staging head and neck cancer. Part I: Normal anatomy. Radiology 148:709–714, 1983. 43. Mancuso AA, Harnsberger HR, Muraki AS, et al: Computed tomography of cervical and retropharyngeal lymph nodes: Normal anatomy, variants of normal, and applications in staging head and neck cancer. Part II: Pathology. Radiology 148:715–723, 1983. 44. Maremonti P, Califano L, Longo F, et al: Detection of laterocervical metastases from oral cancer. J Craniomaxillofac Surg 25:149–152, 1997. 45. Miles KA, Lee TY, Goh V, et al: Current status and guidelines for the assessment of tumor vascular support with dynamic contrast-enhanced computed tomography. Eur Radiol 22:1430–1441, 2012. 46. Mokri B, Silbert PL, Schievink WI, et al: Cranial nerve palsy in spontaneous dissection of the extracranial internal carotid artery. Neurology 46:356–359, 1996. 47. Mukherji SK, Castillo M: A simplified approach to the spaces of the suprahyoid neck. Radiol Clin North Am 36:761–780, v, 1998. 48. Mukherji SK, Kasper ME, Tart RP, et al: Irradiated paragangliomas of the head and neck: CT and MR appearance. AJNR Am J Neuroradiol 15:357–363, 1994. 49. Ng SH, Lin CY, Chan SC, et al: Dynamic contrast-enhanced MR imaging predicts local control in oropharyngeal or hypopharyngeal squamous cell carcinoma treated with chemoradiotherapy. PLoS ONE 8:e72330, 2013. 50. Okura M, Kagamiuchi H, Tominaga G, et al: Morphological changes of regional lymph node in squamous cell carcinoma of the oral cavity. J Oral Pathol Med 34:214–219, 2005. 51. Parker GD, Harnsberger HR, Smoker WR: The anterior and posterior cervical spaces. Semin Ultrasound CT MR 12:257–273, 1991. 52. Pentenero M, Gandolfo S, Carrozzo M: Importance of tumor thickness and depth of invasion in nodal involvement and prognosis of oral

http://radiologyebook.com

CHAPTER 24  Cervical Adenopathy and Neck Masses

53. 54. 55. 56. 57. 58.

59.

60. 61.

62.

63. 64.

65. 66.

67.

squamous cell carcinoma: A review of the literature. Head Neck 27:1080–1091, 2005. Razek AA, Tawfik AM, Elsorogy LG, et al: Perfusion CT of head and neck cancer. Eur J Radiol 83:537–544, 2014. Reede DL, Bergeron RT: Cervical tuberculous adenitis: CT manifestations. Radiology 154:701–704, 1985. Reiseter T, Nordshus T, Borthne A, et al: Lipoblastoma: MRI appearances of a rare paediatric soft tissue tumour. Pediatr Radiol 29:542–545, 1999. Robbins KT: Classification of neck dissection: Current concepts and future considerations. Otolaryngol Clin North Am 31:639–655, 1998. Rumboldt Z, Al-Okaili R, Deveikis JP: Perfusion CT for head and neck tumors: Pilot study. AJNR Am J Neuroradiol 26:1178–1185, 2005. Sadick M, Sadick H, Hormann K, et al: Diagnostic evaluation of magnetic resonance imaging with turbo inversion recovery sequence in head and neck tumors. Eur Arch Otorhinolaryngol 262:634–639, 2005. Sarvanan K, Bapuraj JR, Sharma SC, et al: Computed tomography and ultrasonographic evaluation of metastatic cervical lymph nodes with surgicoclinicopathologic correlation. J Laryngol Otol 116:194–199, 2002. Shah GV: MR imaging of salivary glands. Neuroimaging Clin N Am 14:777–808, 2004. Shah JP, Strong E, Spiro RH, et al: Surgical grand rounds. Neck dissection: Current status and future possibilities. Clin Bull 11:25–33, 1981. Shingaki S, Takada M, Sasai K, et al: Impact of lymph node metastasis on the pattern of failure and survival in oral carcinomas. Am J Surg 185:278–284, 2003. Smith MM: Nonsquamous cell neoplasms of the adult head and neck. Top Magn Reson Imaging 10:304–324, 1999. Som PM: Detection of metastasis in cervical lymph nodes: CT and MR criteria and differential diagnosis. AJR Am J Roentgenol 158:961–969, 1992. Som PM, Curtin HD: Lesions of the parapharyngeal space. Role of MR imaging. Otolaryngol Clin North Am 28:515–542, 1995. Som PM, Curtin HD, Mancuso AA: An imaging-based classification for the cervical nodes designed as an adjunct to recent clinically based nodal classifications. Arch Otolaryngol Head Neck Surg 125:388–396, 1999. Srinivasan A, Mohan S, Mukherji SK: Biologic imaging of head and neck cancer: the present and the future. AJNR Am J Neuroradiol 33:586–594, 2012.

731

68. Sumi M, Izumi M, Yonetsu K, et al: The MR imaging assessment of submandibular gland sialoadenitis secondary to sialolithiasis: Correlation with CT and histopathologic findings. AJNR Am J Neuroradiol 20:1737–1743, 1999. 69. Sumi M, Sakihama N, Sumi T, et al: Discrimination of metastatic cervical lymph nodes with diffusion-weighted MR imaging in patients with head and neck cancer. AJNR Am J Neuroradiol 24:1627–1634, 2003. 70. Sumi M, Van Cauteren M, Nakamura T: MR microimaging of benign and malignant nodes in the neck. AJR Am J Roentgenol 186:749–757, 2006. 71. Tom BM, Rao VM, Guglielmo F: Imaging of the parapharyngeal   space: Anatomy and pathology. Crit Rev Diagn Imaging 31:315–356, 1991. 72. Trojanowska A, Trojanowski P, Bisdas S, et al: Squamous cell cancer of hypopharynx and larynx—Evaluation of metastatic nodal disease based on computed tomography perfusion studies. Eur J Radiol 81:1034–1039, 2012. 73. Turkington JR, Paterson A, Sweeney LE, et al: Neck masses in children. Br J Radiol 78:75–85, 2005. 74. van den Berg R, Wasser MN, van Gils AP, et al: Vascularization of head and neck paragangliomas: Comparison of three MR angiographic techniques with digital subtraction angiography. AJNR Am J Neuroradiol 21:162–170, 2000. 75. van den Brekel MW, Castelijns JA: Imaging of lymph nodes in the neck. Semin Roentgenol 35:42–53, 2000. 76. van den Brekel MW, Castelijns JA: What the clinician wants to know: Surgical perspective and ultrasound for lymph node imaging of the neck. Cancer Imaging 5 Spec No A:S41–S49, 2005. 77. Vergeer MR, Doornaert P, Leemans CR, et al: Control of nodal metastases in squamous cell head and neck cancer treated by radiation therapy or chemoradiation. Radiother Oncol 79:39–44, 2006. 78. White JS, Weissfeld JL, Ragin CC, et al: The influence of clinical and demographic risk factors on the establishment of head and neck squamous cell carcinoma cell lines. Oral Oncol 43:701–712, 2007. 79. Zima A, Carlos R, Gandhi D, et al: Can pretreatment CT perfusion predict response of advanced squamous cell carcinoma of the upper aerodigestive tract treated with induction chemotherapy? AJNR Am J Neuroradiol 28:328–334, 2007.

http://radiologyebook.com

25  Larynx Hugh D. Curtin

Imaging of the larynx is a challenge.4,12,17,19,27 Motion is a constant problem because breathing and swallowing are very difficult to control, particularly in patients with lesions that impinge on the airway. The patient’s larynx moves slightly with each breath. Despite the difficulties, imaging can still provide important information regarding potential involvement of deeper soft tissues and cartilage. Imaging must be considered in view of the capabilities of modern endoscopy.1,25,31,33 As the technology of imaging has progressed, so too has the instrumentation available for direct visualization. Almost all of the mucosa can be examined very effectively. To be relevant, imaging must provide information that cannot be had by direct visualization. Thus the intent of the radiologist in most cases is to evaluate the deeper tissues. In some cases a bulky lesion of the upper larynx can block the view of the lower larynx, and imaging can assist in defining the caudal extent of disease. Tumors of the lowermost part of the larynx may be difficult to visualize, obscured by the vocal folds. The radiologist must be familiar with the anatomy. The behavior of various disease processes and particularly growth patterns of cancer are extremely important. The radiologist must also be familiar with the available therapeutic options to emphasize the information that will be important in making clinical decisions. There is variability of opinion about some of the more recent surgical strategies, and it is important to work closely with the individual surgeons to ensure that the precise regions of concern are addressed. This chapter begins with a section on anatomy, including the normal appearance in the various imaging planes. A brief discussion of the technical considerations of applying computed tomography (CT) and magnetic resonance imaging (MRI) to the task of examining the larynx is followed by sections on imaging of various pathologic conditions. The important clinical considerations in each disease will be stressed.

ANATOMY The larynx is a system of mucosal folds supported by a cartilaginous framework.12,18 Tension and movement of the mucosal folds is effected by the actions of small muscles pulling against this cartilaginous framework.

Mucosa Clinical descriptions of lesions of the larynx emphasize the mucosal anatomy, so the mucosa is a good starting point for the present discussion. When viewed from superiorly (the view of the endoscopist), the first landmark seen is the epiglottis (Fig. 25-1). The upper edge of the epiglottis represents the most superior limit of the larynx. The epiglottis is the anterior boundary of the entrance into the larynx. Anteriorly

732

two small sulci, the valleculae, separate the free portion of the epiglottis from the base of the tongue (Fig. 25-2). From the lateral margin of the epiglottis, the aryepiglottic (AE) folds curve around to reach the small interarytenoid notch. Together these structures complete the boundaries of the airway at the entrance into the larynx. The inner mucosal surface of the larynx, or endolarynx, can be thought of as the working part of the organ. Two prominent parallel folds stretch from front to back along the lateral aspect of each side of the airway (see Figs. 25-1 and 25-2). These are the true and false vocal cords or folds. They are in the horizontal plane. The true vocal fold (the glottis) is the key functional component in the generation of voice and has a fine edge at the medial margin. The more superiorly placed false vocal fold has a more blunted medial surface. Separating these two folds is one of the most important landmarks in the larynx, the ventricle. The ventricle is a thin cleft between the true and false folds, stretching from the most anterior limit of the fold to a point close to the posterior limit of the larynx. Above the false vocal fold the mucosa sweeps upward and outward to the AE folds. Inferiorly the mucosa covering of the true fold sweeps downward into the subglottic area, eventually merging smoothly with the mucosa of the trachea. A final important mucosal landmark is the anterior commissure. This is the point where the true vocal folds converge anteriorly to attach to the inner (posterior) surface of the anterior angle of the thyroid cartilage. The three parallel structures—true vocal fold, false vocal fold, and ventricle—organize the larynx into three regions: the supraglottis, the glottis, and the subglottis. The true fold or cord is the glottis. The glottic region of the larynx extends from the upper surface of the true fold to a line somewhat arbitrarily chosen as being 1 cm below the level of the ventricle. The subglottic region is between this arbitrary line and the inferior edge of the cricoid cartilage, the lower margin of the larynx. There is no defined mucosal structure representing the exact boundary or separation of the glottic and subglottic regions. The supraglottic region is the part of the larynx above the ventricle. This region includes the false folds, epiglottis, and AE folds.

Laryngeal Framework The laryngeal cartilages represent the supporting framework of the larynx. The major cartilages include the cricoid, thyroid, arytenoid, and epiglottic cartilages (Fig. 25-3). The smaller cartilages, the corniculate and cuneiform, reside in the AE fold and are not of particular importance to the radiologist. The cricoid cartilage is the foundation of the larynx and is the only complete ring. The posterior part is larger than the anterior, giving the cartilage the appearance of a signet ring facing posteriorly. The upper

http://radiologyebook.com

CHAPTER 25  Larynx

733

E H

1 2

P

C

FIG 25-1  Endoscopic view of the larynx using an office endoscope. The image is rotated 180 degrees from the usual perspective of the endoscopist. The true vocal fold (1) extends from anterior to posterior and is separated from the false vocal fold (2) by the ventricle (arrowheads). The true folds meet anteriorly at the anterior commissure (small arrow). Larger arrows, aryepiglottic fold. E, epiglottis; P, pyriform sinus. The epiglottis appears fatter than normal because of the perspective.

V H

E

F

T 2

1

FIG 25-2  Larynx split sagittally. View of the lateral wall of the airway. The mucosa over the true vocal fold (1) has been removed to show the thyroarytenoid muscle, which extends from the arytenoid to the thyroid cartilage. The upper edge of the muscle (arrow) indicates the level of the ventricle. The false vocal fold (2) is just above the ventricle. E, epiglottis; F, preepiglottic fat; H, hyoid; T, thyroid cartilage; V, vallecula. (From Curtin HD: The larynx. In Som PM, Bergeron RT, editors: Head and neck imaging, ed 2, St. Louis, 1991, Mosby.)

A

B

FIG 25-3  A, Lateral view of the cartilaginous skeleton. The thyroid cartilage is the largest of the cartilages and articulates with the cricoid cartilage (C) posteriorly and inferiorly. The epiglottis is partially hidden as it disappears behind the thyroid cartilage. H, hyoid bone. B, The thyroid cartilage is now partially transparent, showing the lower  extension of the epiglottis (petiole). The small arytenoid cartilage is seen perched on the cricoid cartilage posteriorly. (From Curtin HD: The larynx. In Som PM, Curtin HD, editors: Head and neck imaging, ed 4, St. Louis, 2003, Mosby.)

margin of the cricoid slopes inferiorly toward the anterior midline. The upper margin of the larger posterior part of the cricoid cartilage reaches the level of the cricoarytenoid joint and the true vocal fold. The lower margin of the cricoid cartilage represents the lower margin of the larynx. The thyroid and arytenoid cartilages articulate with the cricoid cartilage. The thyroid cartilage is large and can be thought of as forming a shield for most of the inner larynx. The arytenoid cartilages perch on the superior edge of the posterior cricoid cartilage. The arytenoid cartilage is situated in the posterior larynx and vertically crosses the level of the ventricle. The vocal process at the level of the true vocal fold projects anteriorly from the lowermost part of the arytenoid cartilage. Because of its characteristic shape and position, the arytenoid cartilage can help localize the ventricle on axial scanning. This is most helpful with CT, because the vocal process may be directly visualized. The upper margin of the arytenoid is at the level of the lower false fold just above the ventricle. The vocal process is at the level of the true fold just below the ventricle. Indeed, the vocal ligament, which represents the medial margin of the true vocal fold, attaches to the vocal process. The epiglottic cartilage, with the exception of its superior tip, is contained within the external framework of the larynx. This cartilage is made of elastic fibrocartilage and does not ossify. Grossly this cartilage has multiple perforations resembling more a mesh than a solid plate, so the epiglottic cartilage is not a major barrier to tumor spread. Inferiorly the epiglottic cartilage is connected to the inner surface  of the anterior part of the thyroid cartilage by the thyroepiglottic ligament.

Muscles and Ligaments The cartilages are connected by a system of muscles and ligaments. Several muscles are mentioned in the remainder of the chapter; one

http://radiologyebook.com

734

PART II  CT and MR Imaging of the Whole Body tions of the paraglottic space vary, this discussion will use Tucker’s original descriptions, with the medial boundary represented by the conus elasticus and quadrangular membrane.32 The lateral boundary is the external skeleton of the larynx, predominantly formed by the inner cortex of the thyroid cartilage. At the level of the supraglottic larynx, the paraglottic space is predominantly filled with fat. Below the ventricle the TAM fills the paraglottic region. A small recess of the ventricle extends upward into the paraglottic space of the supraglottic larynx. This is called the laryngeal saccule or appendix. This structure is within the paraglottic fat lateral to the quadrangular membrane. The preepiglottic space is between the epiglottis posteriorly and the thyroid cartilage and thyrohyoid membrane anteriorly. The hyoepiglottic ligament forms the roof of the preepiglottic and paraglottic spaces.

M

A

IMAGING CONSIDERATIONS C

FIG 25-4  View from above, looking down at the skeletonized larynx. The thyroarytenoid muscle (M) is seen stretching from the arytenoid (A) to the thyroid cartilage. C, cricoid. The arrow indicates the point where paraglottic spread of tumor is expected to intersect the lateral vocal fold. The opposite side shows the vocal ligament stretching from the vocal process of arytenoid to the thyroid cartilage. (From Curtin HD: MR and CT of the larynx. In Thrall JH, editor: Current practice of radiology, ed 3, St. Louis, 1994, Mosby.)

deserves special attention in a discussion of larynx imaging. The thyroarytenoid muscle (TAM) stretches from the arytenoid cartilage to the inner surface of the anterior thyroid cartilage (Fig. 25-4; see Fig. 25-2). This muscle parallels the true fold and makes up the bulk of the true vocal fold. The characteristic shape of this muscle allows identification of the true vocal fold level at imaging. The cricothyroid and thyrohyoid membranes span the greater part of the intervals between the cricoid cartilage, thyroid cartilage, and hyoid bone and represent, along with the cartilages, the outer limits of the larynx. Notably the cricothyroid membrane is considered to be incomplete lateral to the anterior midline. Two membranes are found just deep to the mucosa. The conus elasticus (also called the cricovocal ligament or lateral cricothyroid ligament) stretches from the vocal ligament to the upper margin of the cricoid cartilage. Some descriptions include the membrane attaching to the inner surface of the ring as well. This membrane or fascial layer merges with the anterior cricothyroid ligament in the anterior midline. A similar structure is seen in the supraglottic larynx. A thin ligament called the ventricular ligament is found in the lower margin of the false fold, and the quadrangular membrane sweeps superiorly from the ventricular ligament and terminates in the AE fold. The fan-shaped hyoepiglottic ligament stretches from the epiglottis to the hyoid and divides the supraglottic larynx into a superior (suprahyoepiglottic) area and an inferior (infrahyoepiglottic) area.36

The radiologist can use either CT or MRI to image the larynx. MRI has a theoretical advantage in differentiating soft tissues, but CT has a definite advantage in speed. Radiologists at this institution begin with CT and occasionally do limited MRI to clarify a specific issue. Multidetector CT can cover the entire larynx in just a few seconds. The thin slice thickness allows multiplanar reformatted images comparable in quality to direct imaging (see Fig. 25-6). Because the scan can be done so quickly, the contrast can be ideally planned for best visualization of the anatomy and the pathology. MRI can separate or distinguish various soft tissues slightly better than CT. This may allow better analysis of potential cartilage invasion and has been considered to have an advantage in defining the tumor-muscle interface. Motion artifact is a significant problem when imaging the larynx. CT scanning avoids the problem by using fast imaging. The multidetector acquisition covers the larynx in a few seconds. The examination can be performed during slow shallow respiration, or the scan can be done with the breath held. More elaborate schemes are required for MRI studies. At this institution the patient practices breathing with the abdominal muscles rather than the chest muscles. This is done before the examination begins, usually before the patient is put on the examination table. Specialized neck coils or surface coils are positioned so that any chest movement does not bump the coil. Although speed of imaging is always an advantage, some have recommended using multiple excitations on the T1-weighted sequence. This gives longer imaging times but averages out some of the motion. At this institution, contrast material is usually used when imaging tumor cases. Gadolinium combined with fat suppression can emphasize tumor margins. Fat suppression with the T2-weighted images can help in the evaluation of cartilage involvement. Imaging emphasizes the importance of the ventricle. Axial images are taken along the plane of the ventricle. The position of the ventricle is estimated in axial images by the appearance of the cartilages or by the transition from fat to muscle in the paraglottic region. The coronal images are perpendicular to the ventricle. Even if the ventricle is not actually seen, its position can be predicted by the transition from fat to muscle in the paraglottic space.

Spaces

PATHOLOGY

The paraglottic space and the preepiglottic space are found between the mucosal surface and the cartilaginous outer limit of the larynx. The paraglottic space is found laterally and represents much of the soft tissue wall of the larynx (Figs. 25-5 and 25-6). Although descrip-

Most laryngeal imaging is done for evaluation of tumors or for assessment of patients with trauma to the larynx.4,11,12,17,19 The tumors can be subdivided into mucosal and submucosal groups. In most instances the larynx is directly visualized—if not by endoscopy, at least by

http://radiologyebook.com

CHAPTER 25  Larynx

T

PES

S E C J

A

B

TAM

A

C

D

C

E FIG 25-5  Axial contrast-enhanced CT scan through the larynx from superior to inferior. A, Supraglottic larynx (level of upper thyroid cartilage); most of the wall of the larynx is fat density. C, carotid artery; E, epiglottis; J, jugular vein; PES, preepiglottic space. B, Supraglottic larynx inferior to A. The paraglottic space (arrow) is filled with fat density. Arrowhead represents the aryepiglottic fold. SCM, sternocleidomastoid muscle; T, thyroid cartilage. C, Just above the ventricle (level of the false vocal fold) the paraglottic space is filled with fat (arrow). The upper arytenoid cartilage (A) is visible. D, At the true vocal fold level, the thyroarytenoid muscle (TAM) fills the paraglottic space. Arrowhead represents the cricoarytenoid joint. The upper margin of the cricoid cartilage is visible. E, At the subglottis, the mucosa is tightly applied to the inner surface of the cricoid ring (C) without intervening soft tissue.

http://radiologyebook.com

C M

735

736

PART II  CT and MR Imaging of the Whole Body moderate-sized tumors, in whom some type of voice-sparing therapy is considered. Such treatment options include partial laryngectomies, radiotherapy, or organ-sparing strategies that include radiation and chemotherapy as an initial approach. The feasibility of these therapies depends on the origin and extent of the tumor. In very extensive tumors where total laryngectomy is likely, imaging gives a baseline evaluation and helps evaluate the lymph nodes. Ultrasound has  been useful in evaluating lymph nodes, and many institutions perform positron emission tomography (PET)/CT to evaluate nodes and possible distant metastases. PET/CT will also find occasional second primaries. The precise information required by the various clinicians varies. Perhaps the most precise anatomic analysis is targeted at the feasibility of doing a surgical partial laryngectomy. Originally these voice-sparing partial laryngectomies were performed through open neck approaches. Now more partial resections are done via endoscope using lasers. Many patients who would have been candidates for open partial laryngectomies are now treated with various chemoradiation (chemoRT) organpreserving protocols. Even though open partial laryngectomies are done much less common than previously, we still use the landmarks described for these techniques in our evaluation of larynx cancers. This approach with very slight modifications provides the information required by any of the therapeutic strategies considered. The discussion of imaging of carcinoma of the larynx is organized based on the site of origin of the tumor.

T TAM C

A

H

Supraglottic Tumors.  The classic voice-sparing surgical option for T

B FIG 25-6  Coronal reformat through the larynx (multidetector CT). A, Quiet respiration. The thyroarytenoid muscle (TAM) is seen in cross section with a triangular shape. This muscle represents the bulk of the true vocal fold. The level of the ventricle (arrowhead) is predicted as the transition between the muscle density of the true fold and fat density of the paraglottic space (arrow) at the supraglottic level. C, cricoid cartilage; T, thyroid cartilage. The lower half of the thyroid cartilage is heavily ossified; the upper half is nonossified, giving the difference of density. B, The patient is holding his breath. The position of the closed airway is represented by the dashed line. Note the density of fat in the paraglottic space at the supraglottic level (arrow). The TAM (arrowhead) forms the bulk of the true vocal fold (glottic level). H, hyoid bone; T, thyroid gland. The thyroid cartilage and cricoid cartilage are also visible.

mirror—and the radiologist should pursue information that is not obtainable by direct visualization.

Mucosal Tumors Most tumors evaluated by imaging are squamous cell carcinomas arising from the mucosa of the larynx.3,28 With the possible exception of extremely rare lesions arising deep within the ventricle, these tumors are almost always detected before imaging. Indeed, imaging cannot currently rule out a small malignancy and so is not a substitute for direct visualization. The otolaryngologist can completely assess smaller lesions. The margins of a small tumor may be readily visible and no further  information is required. Imaging is important for patients with

supraglottic squamous cell carcinoma is the supraglottic laryngectomy22,31 (Fig. 25-7). The resection is made along the ventricle, and the entire supraglottic larynx is removed, leaving at least one and usually both arytenoid cartilages. The patient retains the true folds and so generates voice in the usual manner with the normal neuromuscular mechanisms. The protective function of the supraglottic larynx, particularly the epiglottis, is lost, and the patient must learn to swallow without aspirating. The classic open surgical partial laryngectomy is done via open skin incisions, but currently larynx surgeons perform these partial resections endoscopically using lasers.31 The anatomic landmarks emphasizing the ventricle are the same for both approaches. Tumor crossing the ventricle is the primary contraindication for the standard supraglottic laryngectomy (Box 25-1). The incision passing along the ventricle would cut through tumor. Tumor growing along the mucosa to reach the true vocal fold may be obvious at direct visualization. Tumors may invade into the deeper tissues laterally and follow the paraglottic space around the ventricle into the lateral edge of the true vocal fold. This type of spread is unusual as an isolated phenomenon but should be checked when imaging a supraglottic tumor. Axial imaging parallels the ventricle and therefore is somewhat limited in precisely identifying the level of this key structure. The paraglottic space at the false vocal fold level is predominantly fat, but at the true vocal fold level is predominantly muscle. The transition between the two represents the approximate level of the ventricle. If a normal slice can be found below the tumor but still within the supraglottic larynx, the supraglottic laryngectomy is technically possible (Fig. 25-8). If the tumor can be identified at the level of the vocal fold as well as in the supraglottis, the patient cannot have the standard supraglottic resection (Fig. 25-9). If surgery is chosen as the therapy, the usual procedure would be a total laryngectomy although some more extensive procedures have been described, taking larger parts of the larynx (see later). As endoscopic resections advance, new strategies can extend the ability to take tumors at lower levels. Describing the most inferior extent of the tumor relative to the ventricle is very

http://radiologyebook.com

CHAPTER 25  Larynx

E

A E

2

T

C

FIG 25-7  Diagram of a supraglottic laryngectomy. Dotted line shows the incision of a supraglottic laryngectomy, passing along the ventricle between the true vocal fold and the false vocal fold (2). AE, aryepiglottic fold; C, cricoid cartilage; E, epiglottis; T, thyroid cartilage. (From Curtin HD: Imaging of the larynx: Current concepts. Radiology 173: 1–11, 1989.)

BOX 25-1  Contraindications to Classic

Supraglottic Laryngectomy • • • • • • •

Tumor extension onto the cricoid cartilage Bilateral arytenoid involvement Arytenoid fixation Extension onto the glottis or impaired vocal cord mobility Thyroid cartilage invasion Involvement of the apex of the piriform sinus or postcricoid region Involvement of the base of the tongue more than 1 cm posteriorly to the circumvallate papillae

From Lawson W, et al: Cancer of the larynx. In Suen JY, Myers E, editors: Cancer of the head and neck, New York, 1989, Churchill Livingstone, pp 533–591.

important in such cases. Alternatively these patients may be considered for organ-sparing chemoRT protocols. On axial images, the lateral edge of the TAM should be carefully examined for the earliest evidence of transglottic (from false fold to true fold) spread. A narrow extension of the paraglottic fat is often seen along the outer (lateral) margin of the muscle (see Fig. 25-9). Tumor spreading around the ventricle will grow into this fat and eventually appear to “pry” the muscle away from the thyroid cartilage. Coronal images are perpendicular to the ventricle and may also be helpful in

737

showing the relationship between tumor and the lateral aspect of the true vocal fold. Tumor can be distinguished readily from the fat in the supraglottic paraglottic space. The interface between tumor and muscle at the true fold level is more difficult to visualize. MRI studies have the advantage of better tissue differentiation using various pulse sequences, but this advantage can be realized only if the patient is able to control motion (Fig. 25-10). In our experience the ability to optimize the contrast injection with CT scanning may allow good differentiation of the tumor from muscle, diminishing the soft tissue advantage of MRI. Midline tumor can cross the level of the ventricle to involve the anterior commissure. This area can be difficult to evaluate either by direct visualization or by imaging. If there is deep growth from the anterior commissure into the attachment of the cord or through the cricothyroid membrane, imaging may detect tumor that is unappreciated clinically. Anterior growth from a supraglottic carcinoma brings the tumor into the preepiglottic space (see Figs. 25-9 and 25-10). The preepiglottic space and paraglottic space have a rich lymphatic drainage, and tumor invasion heightens the concern regarding nodal metastases. Invasion is easily appreciated on axial imaging because the tumor is contrasted against the fat. Such involvement is not a contraindication to a supraglottic laryngectomy. Thyroid cartilage invasion is usually considered a contraindication to the supraglottic laryngectomy. Such invasion is extremely uncommon unless the tumor is large and has crossed the ventricle.20 In such a case the patient would be excluded as a candidate on the basis of ventricular involvement. Involvement of the epiglottic cartilage is not a contraindication. Cartilage assessment is discussed under the heading “True Vocal Fold Tumors.” The supraglottic partial laryngectomy is the classic surgical therapy for tumors limited to the area above the ventricle. More extensive resections have been used for lesions extending beyond the boundaries of supraglottic laryngectomy.31 Pearson’s near-total laryngectomy can be done for supraglottic cancer extending to one vocal fold. The resection is continued to include the vocal fold and take the upper margin of the cricoid. The supracricoid partial laryngectomy with cricohyoidopexy takes the anterior vocal folds to the level of cricoid. Minor involvement of the thyroid cartilage is not a contraindication. In either of these approaches the inferior extent remains the most important assessment at imaging. As stated previously, the modern alternative to partial laryngectomy is endoscopic laser surgery.31 Lesions of the suprahyoid epiglottis, AE fold, and false vocal folds are more appropriate for this type of surgery than are lesions of the laryngeal surface of the epiglottis. The latter tumors are partially hidden by the epiglottis and thus less accessible to endoscopic resection. Radiation therapy and chemoRT protocols are also performed  for supraglottic cancer. The same landmarks and considerations for supraglottic laryngectomy pertain to assessment for these therapeutic options. The bulk of the tumor is also a prognostic factor. Larger tumors do worse than smaller. One study used a 6-mL volume as a separation, with larger tumors showing significantly worse prognosis than smaller ones.26 Nodal metastases are very frequent in supraglottic tumors and can be bilateral.

True Vocal Fold Tumors.  There are several options for therapy of relatively small glottic carcinomas.31 These include simple resection, endoscopic laser resection, and radiotherapy. The classic vertical hemilaryngectomy may be considered for a more extensive lesion at the level

http://radiologyebook.com

738

PART II  CT and MR Imaging of the Whole Body

N

T

A

B

C

D FIG 25-8  Supraglottic carcinoma not reaching the level of the ventricle or true fold. A, Axial plane. Tumor (T) of supraglottis/aryepiglottic fold. N, metastatic node. B, Axial plane, inferior to A. The paraglottic fat (arrow) is normal at this level, situated between tumor and the level of the ventricle. C, True vocal fold level is normal. D, Coronal. The inferior margin (arrow) of the tumor separable from the level of the ventricle and upper margin of the true vocal fold (arrowhead).

of the true vocal fold. This procedure removes the true and false vocal folds on one side of the larynx using a vertical incision through the thyroid cartilage. The lesion can cross the anterior commissure and involve the anterior one third of the opposite vocal fold and still be treatable by vertical hemilaryngectomy. The incision through the thyroid cartilage is moved laterally toward the opposite side. At our institution, most resections of glottic-level tumors are performed endoscopically using a standard cutting laser or the KTP (potassium titanylphosphate) angiolytic laser to interrupt the blood supply to a part of the tumor.2,35 No matter the preferred approach, evaluating the lesion as though the patient is a candidate for open classic vertical hemilaryngectomy will emphasize the important surgical landmarks. Various strategies to extend some of the limits have been developing rapidly in the laryngeal surgical community. The contraindications for the classic vertical hemilaryngectomy are listed in Box 25-2.22 The most important assessments done at imaging involve the inferior extension of the tumor, deep invasion at the anterior commissure, and cartilage invasion. Submucosal superior extension crossing the ventricle into the supraglottic larynx is only an occasional problem. The proximity of tumor to the cricoarytenoid joint should be defined. Deep posterior paraglottic extension along the lateral aspect of the vocal process has been described as being particularly important in endoscopic resections.27 Inferior extension is quantified relative to the upper margin of the cricoid cartilage. The cricoid cartilage is the foundation of the larynx,

BOX 25-2  Contraindications to Vertical

Frontolateral Hemilaryngectomy*

• Tumor extension from the ipsilateral vocal cord across the anterior commissure to involve more than one third of the contralateral vocal cord • Extension subglottically >10 mm anteriorly and >5 mm posterolaterally • Extension across the ventricle to the false cord • Thyroid cartilage invasion • Impaired vocal cord mobility a relative contraindication From Lawson W, et al: Cancer of the larynx. In Suen JY, Myers E, editors: Cancer of the head and neck, New York, 1989, Churchill Livingstone, pp 533–591. *This technique can still be used if the vocal process and anterior surface of the arytenoid are involved, but involvement of the cricoarytenoid joint, interarytenoid area, opposite arytenoid, or rostrum of the cricoid is a contraindication.

and classically only very limited partial resections or cortical shavings have been possible. Recently some surgeons have devised strategies to resect segments of the cricoid and reconstruct a scaffold to support the airway. For instance, an aortic graft has been strong enough to replace a section of this important cartilage.37

http://radiologyebook.com

CHAPTER 25  Larynx

T T

SCM

A

B

T

D

C

C

T

E

C

F FIG 25-9  Postcontrast CT scan shows transglottic tumor. The lesion appeared to arise from the supraglottic larynx. A, Axial image through the level of the supraglottis. The tumor (T) fills much of the paraglottic space at the level of the supraglottis. Note the normal fat (arrow) in the opposite paraglottic space. Arrowhead indicates thyroid cartilage. B, Axial image slightly inferior to A. Tumor (T) continues inferiorly. Note the airfilled laryngeal appendix (arrow) in the paraglottic fat on the normal side. SCM, sternocleidomastoid muscle. C, Axial image level of the ventricle/upper true vocal fold. Tumor (T) fills the right side. On the left there is a small amount of normal fat (arrow) just lateral to the thyroarytenoid muscle (TAM). The cricoarytenoid joint (arrowhead) is also an indicator that the scan is entering the true vocal fold level. D, Axial image of true vocal fold level. The tumor distorts the true fold on the right. The normal TAM (arrowheads) indicates the level of the true vocal fold. A small amount of residual normal paraglottic fat (arrow) reaches this level. E, Axial image at the level of the upper subglottis. The tumor (arrowhead) is seen along the inner cortex of the cricoid cartilage (C). F, Coronal reformatted image. The tumor (T) fills the supraglottis on the right and infiltrates the true vocal fold (arrow), massively increasing its size. The tumor touches the cricoid (C) in the subglottis. The approximate level of the ventricle on the left is indicated by the dashed line. A small invagination of paraglottic fat (arrowhead) is demonstrated at the upper outer margin of the TAM at the true vocal fold level on the normal side.

http://radiologyebook.com

739

740

PART II  CT and MR Imaging of the Whole Body

T

A

B

T

V TAM

C D FIG 25-10  Supraglottic tumor. A, Axial T1-weighted image. The tumor (T) fills most of the right supraglottic larynx and crosses the midline. Paraglottic and preepiglottic spaces are invaded. Note that the tumor  and the prelaryngeal strap muscles (arrow) have the same signal intensity. The strap muscle would have the same signal as the thyroarytenoid muscle (TAM). B, Axial T2-weighed image. The tumor has significantly more signal than the prelaryngeal strap muscle. Arrowhead indicates fat in the medullary cavity of  the ossified thyroid cartilage. C, Axial T2-weighted image at the level of the true vocal fold. The TAM (arrow) is normal with approximately the same signal as the strap muscles. The tumor has not reached this level. D, Coronal postcontrast T1-weighted image. The tumor (T) fills the supraglottis but is separated from the TAM and thus the true fold by a small amount of fat (arrow) in the paraglottic space at the level of the ventricle (V).

At imaging, if the tumor can be identified within the ring of the cricoid cartilage, the classic open vertical laryngectomy cannot be done (Fig. 25-11). Deep invasion at the anterior commissure and anterior midline can involve the thyroid cartilage or the cricothyroid membrane (Figs. 25-12 and 25-13). The subtle fat plane just external to the membrane should be examined closely on the axial image. Cartilage invasion is considered a contraindication for vertical hemilaryngectomy and is very important in radiation or chemoRT protocols. Minor thyroid cartilage invasion has been treated with supracricoid partial laryngectomy with cricohyoepiglottopexy. This procedure removes more of the supraglottic larynx, thus suspending the cricoid cartilage from the remaining supraglottic tissues and hyoid. If the tumor reaches the cricoid, a segment of the cricoid can still be

removed if there is an appropriate strategy to reconstruct (e.g., placement of a segment of cadaver aortic graft).37

Cartilage Invasion.  Cartilage invasion is considered a contraindication to both the standard supraglottic and vertical hemilaryngectomies. Assessment of the cartilage is important for radiation planning or for inclusion in various chemoRT strategies. The subject is discussed here because such invasion is of more concern in lesions involving the true vocal fold. With both CT and MRI studies, the most reliable sign of cartilage involvement is identification of tumor on the extralaryngeal or outer surface of the cartilage. Minor degrees of cartilage invasion can be difficult to detect. The variability of ossification of the major cartilages can lead to problems detecting cartilage invasion.

http://radiologyebook.com

CHAPTER 25  Larynx The nonossified part of the cartilage can have approximately the same appearance as tumor on CT scans. MRI studies have an advantage of better tissue discrimination, and although there are no large series available, limited results suggest that tumor can be differentiated from nonossified cartilage based on signal.9,10 If there is bright signal on the T1-weighted image, that part of the cartilage can be reliably considered normal (Fig. 25-14). The high signal represents fat in

741

ossified cartilage; carcinoma is never that bright. On T1-weighted images, tumor and nonossified cartilage are usually intermediate or dark. The nonossified cartilage may be slightly darker than tumor. Regions that are dark on T1-weighted images are examined on the long-TR (repetition time) sequence. Nonossified cartilage remains relatively dark on long-TR sequence images. Tumor tends to be significantly brighter than nonossified cartilage (see Fig. 25-14). We prefer fat suppression on T2-weighted sequences so that all normal cartilage is dark and abnormal cartilage is of higher signal. Nonossified cartilage does not enhance with gadolinium, whereas tumor invading cartilage does enhance to some extent. Again, we prefer to use fat suppression on the T1 sequences taken after gadolinium. Tumor, then, is dark or intermediate on T1-weighted images and is usually relatively brighter on T2-weighted images and enhances after gadolinium. This signal pattern does not definitely mean tumor. Edema, fibrosis, and even red marrow have been described with this pattern.6,8 Edema is of particular concern because squamous cell carcinoma often has a peritumoral inflammatory response. Edema may be brighter than tumor on T2-weighted imaging, but more experience

A

N

FIG 25-12  Tumor of the anterior commissure extending anteriorly. A

B FIG 25-11  Tumor of vocal fold with subglottic extension. Axial CT with contrast. A, Tumor (arrow) at level of true vocal fold. B, Subglottic extension shows tumor (arrow) along the inner cortex of cricoid inferior to the level of the upper margin.

small tumor was seen at the anterior commissure (white arrow). Anterior extension carried the tumor through the thyroid cartilage (arrowheads), with a prominent nodule in the anterior soft tissues (black arrow). N, metastatic node. (From Curtin HD: The larynx. In Som PM, Bergeron RT, editors: Head and neck imaging, ed 2, St. Louis, 1991, Mosby.)

T c

A

B

FIG 25-13  Cancer of true vocal fold/ventricle extending via paraglottic space through cricothyroid interval. A, Coronal postcontrast CT image shows tumor (T) centered on the ventricle, filling the paraglottic space. The tumor extends inferiorly to reach the cricoid cartilage (C). B, Axial image shows tumor (arrow) extending into fat anterior to the cricothyroid membrane (interval). Compare to normal fat on opposite side.

http://radiologyebook.com

742

PART II  CT and MR Imaging of the Whole Body T2 FS

T1 Pre

NOC

NOC

T OC

OC

A

B

Post Gad FS

Pre T1

NOC

T

E

T OC

C

D Post Gad FS

T2 FS T

T

E

E

E

F FIG 25-14  Squamous cell cancer invading lower thyroid cartilage. A-C, Normal noninvaded cartilage. A, Axial T1-weighted MRI shows high signal of fat in ossified cartilage (OC). Nonossified cartilage (NOC) shows low signal similar to the tumor (T). B, Both OC and NOC have low signal on fat-suppressed T2-weighted MRI. C, With postcontrast fat suppression, neither OC nor NOC enhance. D-F, Direct invasion anteriorly (T) and inflammatory edema (E) posteriorly. D, Axial T1-weighted MRI shows the tumor (T) invading the cartilage; invading tumor has approximately the same signal as the main bulk of the tumor. More posteriorly, inflammatory edema (E) has similar signal pattern as the tumor. E, Fat-suppressed T2-weighted MRI shows intermediate signal of the invading tumor (T) similar to the bulk of the tumor, but the more posterior cartilage shows inflammatory edema (E) with significantly higher signal than that of the tumor. F, Postcontrast fatsuppressed T1-weighted axial image shows enhancement of the invading tumor (T) similar to the bulk of the tumor. More intense enhancement is present in the segment with inflammatory edema (E).

http://radiologyebook.com

CHAPTER 25  Larynx is needed before a definite statement can be made regarding making this differentiation. Edema of the cartilage presumably means that tumor is at least very close. In our experience, at least the perichondrium has been invaded if there is “edema” (high T2 signal) in the cartilage. A CT irregularity of the inner cortex may suggest that tumor is eroding, but this sign is questionable owing to the variability of ossification.13 The cortex may be interrupted normally. Asymmetric sclerosis, particularly of the thyroid and cricoid cartilages, may indicate tumor.7 The normal arytenoid cartilage can ossify asymmetrically, so sclerosis is not as good an indicator of tumor.34 Tumor may obliterate the medullary fat within the cartilage. Very recently dual-energy CT has been suggested as an approach to evaluating possible invasion of cartilage.15,21 Tumor tends to invade ossified cartilage rather than nonossified cartilage.20 Invasion of ossified cartilage shows loss of bone density and loss of medullary fat density. An intact bone density cortex of ossified cartilage is considered evidence that tumor has not invaded. However, nonossified cartilage can have the same density as tumor; both are relatively dense compared to muscle. The problem becomes distinguishing the density of enhancing tumor from the relatively high density of nonossified cartilage. Nonossified cartilage can mimic tumor invading and destroying ossified cartilage. By scanning with two different energies, one may be able to make the differentiation. Because of the presence of the iodine peak, density due to iodine can be separated from density due to protein  or other substances that, although dense, do not contain significant iodine. Nonossified cartilage does not have a significant blood supply and thus does not show a prominent iodine-based enhancement. Further study regarding the significance of various signal patterns and CT findings is needed. For instance, the effect of various imaging findings on the prognosis of various therapies, particularly radiotherapy, could potentially change management.10,23,24

Subglottic Tumors.  Subglottic tumors are rare. When they do occur the only surgical option usually is a total laryngectomy because of the proximity to the cricoid cartilage. Because the cricoid is such an important landmark for both areas, glottic and subglottic tumors are often considered together in imaging strategies. This allows separation of lesions into two groups: supraglottic and glottic/subglottic.

Hypopharyngeal Tumors.  The hypopharynx is not actually part of the larynx but is so intimately associated that discussion is warranted in this chapter. Decisions about potential surgical resection usually relate to the proximity of the tumor to landmarks within the larynx, so the assessment is similar to that of lesions within the larynx itself. Two regions of the hypopharynx have important relationships to the larynx: the pyriform sinuses and the postcricoid region. The pyriform sinuses indent the posterior wall of the larynx. The anterior wall of the pyriform sinuses represents the posterior wall of the paraglottic space. The pyriform sinus makes up the lateral aspect of the AE fold. If a tumor is localized to the upper pyriform, a supraglottic resection may be considered along with resection of the tumor. A tumor that invades the paraglottic region of the larynx can follow the paraglottic pathway to reach the level of the true vocal fold (Fig. 25-15). The postcricoid region is that area of the lower hypopharynx that covers the posterior aspect of the cricoid cartilage. To achieve an appropriate margin the cricoid must usually be removed, and thus a total laryngectomy is most commonly performed or organ preservation strategies considered. In these patients the inferior extension within the pharynx should be estimated to help the surgeon decide the appropriate method of reconstructing the food passage. The key landmark

C

743

T

FIG 25-15  Tumor (T) of the pyriform sinus. The lesion protrudes through the thyroarytenoid gap between thyroid cartilage and arytenoid (arrow). The tumor thus invades the paraglottic space (arrowhead) of the supraglottic larynx. Compare with the fat in the paraglottic space on the normal side. C, carotid artery.

is the transition from the oval appearance of the pharyngeal muscu­ lature to the round appearance of the esophageal musculature as appreciated in the axial plane (Fig. 25-16). This transition is closely approximated by the lower margin of the cricoid cartilage.

Lymph Node Metastases.  Metastasis to the lymph nodes is a major consideration in carcinoma of the larynx. The subject of lymph nodes is discussed in Chapter 24, but brief comments about major pathways of spread are appropriate. Nodal metastases are a major concern with supraglottic tumors. Spread to the jugular nodes is very frequent because of the rich lymphatic drainage. The margin of the true vocal fold does not have any lymphatics. Therefore tumors of the true vocal fold do not spread to the lymph nodes unless there is deep invasion or, alternatively, the tumor has grown along the mucosa to involve the subglottic region. The subglottic mucosa does have lymphatic drainage, so a primary tumor arising in this location or a glottic tumor extending into the area secondarily can metastasize to the nodes. In these patients the nodes along the esophagus and in the upper mediastinum become a significant concern.

Radiation Therapy and Chemoradiation for Mucosal Tumors.  The evaluation of patients from a surgical perspective uses landmarks that are slightly more precise, but many of the same findings are appropriate for those patients for whom radiotherapy or chemoRT is chosen as the treatment. The dimensions of the tumor are relevant in that this reflects the bulk of the tumor and relates to the depth of invasion into the soft tissues. The volume of the tumor has been used to suggest prognosis after treatment with radiation therapy.26 The vertical extent should be estimated, although the precision relating to the ventricle is less crucial. Extension into the subglottic area is very important because the treatment portals may be changed to include the paratracheal and upper mediastinum regions. Formerly, cartilage involvement was considered a contraindication to radiation therapy. Cartilage invasion makes tumor recurrence more

http://radiologyebook.com

744

PART II  CT and MR Imaging of the Whole Body

A

T J C

B FIG 25-16  Axial CT scan showing transition from pharynx to esophagus. A, Axial image through the inferior pharynx. The pharynx has an oval appearance (arrowheads) stretching from side to side. This is approximately at the level of the cricopharyngeus where the muscles attach to the more lateral cricoid cartilage. Arrow indicates cricoid cartilage. B, Level of the esophagus. The esophagus is seen as a round structure (arrow). C, carotid artery; J, jugular vein; T, thyroid gland.

frequent, as is radiation chondritis and chondronecrosis. Recently, however, some have expressed the opinion that relatively subtle involvement is not an absolute contraindication.

Other Mucosal Tumors.  Many benign lesions such as polyps and papillomas arise from the mucosa of the larynx. These are seldom evaluated by CT or MRI studies. Submucosal tumors can ulcerate through the mucosa (see “Submucosal Tumors”). Although almost all malignancies arising from the mucosa of the larynx are squamous cell carcinoma, occasional malignancies arise from the minor salivary glands, with histologies such as adenocarcinoma, adenoid cystic carcinoma, and mucoepidermoid carcinoma. Verrucous carcinoma is an exophytic slow-growing variant of squamous cell carcinoma. Usually it does not metastasize but can be locally aggressive. There are no particular imaging features, but the lesion has a typical exophytic “warty” appearance at direct inspection.

Submucosal Tumors Lesions presenting beneath an intact mucosa usually arise from the nonepithelial elements of the larynx.3,5,12,14,28 Such tumors include chondroid lesions, hemangiomas, vascular malformations, neurogenic tumors, and rare lesions such as leiomyoma, rhabdomyoma, and lipomas. Sarcomas arising from the same various mesenchymal

elements are often submucosal but with increasing size can lead to mucosal ulceration. Rarely a squamous cell carcinoma can appear to be entirely submucosal. Such a lesion arises in the ventricle or ventricular saccule (appendix). The tumor expands upward into the paraglottic space and downward into the true vocal fold rather than medially toward the lumen of the larynx. A submucosal tumor represents a different problem than the mucosal lesion. Unlike the squamous cell carcinoma, where the actual tumor can be seen and easily biopsied, the submucosal tumor presents as a bulge beneath an intact mucosa. In these patients, imaging is used not only to show the extent of the lesion but also to help identify the type of tumor. In certain tumors, the findings are characteristic enough that the diagnosis can be made accurately. In patients in whom the diagnosis is not definite after imaging, certain findings can still be helpful to the otolaryngologist. The first question is whether the tumor is a chondroid lesion. Chondrosarcomas and chondromas arise from the cartilaginous framework of the larynx16,30 (Figs. 25-17 and 25-18). Most arise from the cricoid. The thyroid cartilage is the second most common site of origin. Differentiation between benign and malignant can be impossible even at histopathologic examination. Most are considered to  be low-grade chondrosarcomas. These tumors can compromise the airway, and treatment is usually surgical. A partial laryngectomy can be curative if the entire lesion is resected. If the cricoid cartilage is significantly involved, a total laryngectomy may be considered because of the key role of this cartilage in maintaining a patent airway. Many surgeons will consider partial resection with debulking the lesion, leaving the smallest amount of tumor possible. Close imaging surveillance is mandatory with this approach. If partial resection is planned, the relationship of the tumor to the cricoarytenoid joint is extremely important. On either CT scan or MRI study, the cartilage origin may be obvious and give a clue to the identity of the lesion (see Fig. 25-17). The cartilage may be expanded, indicating a lesion arising within rather than eroding into the cartilage from a more superficial mucosal origin. The cartilaginous matrix is the most characteristic finding at imaging (see Fig. 25-18). CT scans are better than MRI studies at demonstrating the calcifications within the matrix of the tumor. Such calcifications are extremely rare in other lesions. Relapsing polychondritis can give fairly extensive calcification but is extremely rare and should not involve the larynx exclusively. If the lesion does not have the imaging characteristics of a cartilage lesion, the radiologist tries to determine whether the lesion is vascular. Hemangiomas, venous vascular malformations, and paragangliomas enhance intensely on CT scans if a bolus or rapid drip of intravenous contrast material is used. Knowledge that a tumor is vascular has obvious implications if biopsy or resection is planned. Other than chondrosarcomas and vascular lesions, soft tissue tumors do not have specific identifying characteristics. They do not have a cartilaginous matrix and do not enhance brightly. The absence of these two imaging characteristics is useful information in limiting the diagnostic possibilities.

Cysts Cysts also present as submucosal masses. Three types of cysts typically arise within or in intimate association with the larynx. The saccular cyst (laryngocele) and the mucosal cyst arise within the larynx. The thyroglossal duct cyst arises just outside the larynx in close association with the strap muscles. The saccular cyst (laryngocele) represents a dilatation of the ventricular saccule or appendix (Fig. 25-19). Strict terminology refers to

http://radiologyebook.com

CHAPTER 25  Larynx

A

C

C

B

A

V

C

C

C

D FIG 25-17  Chondroid lesion arising in the cricoid cartilage. A, Axial CT. The chondrosarcoma (C) expands the cricoid cartilage. Small bits of cartilage matrix (arrow) suggest the diagnosis. B, Axial CT bone algorithm. The chondrosarcoma (C) extends up to the superior margin of the cricoid, reaching the cricoarytenoid joint. A, arytenoid. Normal cricoarytenoid joint on opposite side (arrowhead). C, Sagittal CT. The chondrosarcoma (C) expands the cricoid cartilage. V, ventricle. D, The chondrosarcoma (C) reaches the cricoarytenoid and displaces the arytenoid (arrow).

A

B FIG 25-18  Axial CT scan showing chondrosarcoma. A, The tumor expands (arrowhead) the thyroid lamina. B, Note the stippled density (arrow) within the cartilage, representing cartilage formation.

http://radiologyebook.com

745

746

PART II  CT and MR Imaging of the Whole Body The thyroglossal duct cyst arises from the remnant left as the thyroid descends to the appropriate position in the lower neck. Such a cyst can arise above or below the hyoid. Those that arise below the hyoid are usually found just off midline in the region of the strap muscles anterior to the larynx (Fig. 25-22). The appearance is characteristic and should not be confused with a saccular cyst arising within the larynx itself. A thyroglossal duct cyst may rarely bulge over the anterior notch of the thyroid cartilage but should not pass through the lateral aspect of the thyrohyoid membrane. The paraglottic space should not be involved. Any enhancing tissue or nodule within a thyroglossal duct cyst must be considered concerning for malignancy such as papillary thyroid carcinoma.

Posttherapy Imaging

FIG 25-19  Diagram of a laryngocele. Coronal section through the larynx shows the ventricle (arrow) and the ventricular appendix (arrowhead) on the normal side. When this appendix is obstructed, it dilates (fills with either fluid or air) to form a saccular cyst or laryngocele. In the diagram the lateral extent of the saccular cyst protrudes through the lateral cricothyroid membrane. (From Curtin HD: The larynx. In Som PM, Bergeron RT, editors: Head and neck imaging, ed 2, St. Louis, 2003, Mosby.)

the abnormality as a saccular cyst if fluid filled and as a laryngocele if air filled. Many simply refer to the lesions as air-filled or fluid-filled laryngoceles. The laryngocele is a supraglottic abnormality, and the true vocal fold is normal. Both types present as a submucosal mass or bulge in the supraglottic larynx (Figs. 25-20 and 25-21). The air or fluid within is readily identifiable, and the extent can be determined. A small saccular cyst or laryngocele may be confined to the supraglottic paraglottic space and is referred to as internal. A larger lesion may protrude laterally through the thyrohyoid membrane just anterior to the upper horn of the thyroid cartilage. Any component outside the membrane is referred to as external. A pure external laryngocele is rare. Occasionally a small air-filled external laryngocele is identified with no obvious dilatation of the part of the saccule within the paraglottic space. This is usually an incidental finding. The external component is more commonly seen in combination with an internal component. This is called a mixed laryngocele. When the external component is large, the laryngocele can present as a neck mass. Laryngoceles and saccular cysts are benign. However, these abnormalities can occur when tumor in the ventricle causes obstruction of the laryngeal saccule; therefore both the radiologist and the endoscopist must carefully examine this key area. A mucosal cyst is considered to be an obstructed mucous gland and can present anywhere there is mucosa. The benign nature can be suggested by the smooth appearance, and this diagnosis is considered when a cystic-appearing mass is found on a mucosal surface. An oncocytic cyst is considered to be oncocytic metaplasia or degeneration of minor salivary glands of the larynx. These may have a high intrinsic density. This may be mistaken to represent enhancement on contrastenhanced CT scan.

Both surgery and radiotherapy significantly distort the normal appearance of the larynx. After a total laryngectomy, the food passage may dilate to mimic an airway, but the laryngeal landmarks are no longer seen. Lower slices show the tracheostomy site. The appearance of the glottic and subglottic levels is unchanged by a supraglottic laryngectomy. The supraglottic larynx has been removed, so the characteristic landmarks of the paraglottic and preepiglottic spaces are not identified. The hyoid bone may be removed. The major portion of the thyroid cartilage above the level of the vocal fold is also taken. The vertical hemilaryngectomy can give a variable appearance. One vocal fold is removed, leading to significant asymmetry at the glottic level. Part of the upper larynx is removed on the ipsilateral side. The position of the original lesion dictates the position of the cuts through the thyroid cartilage, resulting in significant variability in appearance among patients with vertical hemilaryngectomies. At times an attempt is made to reconstruct the true vocal fold using a small slip taken from the strap muscles. This can lead to significant soft tissue density that cannot be readily distinguished from tumor. With the advent of endoscopic resections, the postoperative appearance of the larynx has become more variable. Close coordination with the surgeon is crucial in understanding these imaging findings. No two resections have the same appearance. Radiotherapy can cause characteristic apparent swelling of the soft tissues over the arytenoids and supraglottic region (Fig. 25-23). Subtle reticulation or stranding of the supraglottic fat is seen. Initially this represents local inflammation and edema, but small vessel changes make the appearance more permanent. Recurrent tumor usually is seen as a small enhancing area that cannot be explained as a postoperative or postradiation change. With further technologic progress and higher resolution, PET may play an expanding role in postoperative assessment of laryngeal tumors.

TRAUMA Direct trauma can result in fractures or dislocations of the cartilages in the larynx.17 Soft tissue injuries such as mucosal tears and hematomas can occur in isolation but are often seen in conjunction with fractures. CT is usually used to evaluate trauma cases because of   the ability to visualize the calcified cartilage and because of the speed of imaging. The fragile airway is more easily managed and monitored in the CT scanner than in the more confined area of the MRI machine. In young patients the cartilage may not be calcified but can often be seen as slightly increased density against the contiguous soft tissues. Fractures usually affect the thyroid cartilage or the cricoid cartilage. Fractures of the thyroid cartilage can be vertical or horizontal

http://radiologyebook.com

CHAPTER 25  Larynx

L

747

L

A

B

L

D

C

L

FIG 25-20  Saccular cyst (fluid-filled laryngocele). Postcontrast CT. A, The fluid-filled laryngocele (L) causes a submucosal bulge in the supraglottic larynx. B, Slightly inferiorly the cyst (L) fills much of the paraglottic space but has a smooth margin. C, Just inferior to B the cyst (L) is visible at the false vocal fold level. D, The true vocal fold level is normal. Note the fat (arrow) in the lower paraglottic space along the lateral margin of the thyroarytenoid muscle. E, Coronal reformat shows the cyst (L) in the paraglottic space of the supraglottis. The cyst stops at the level of the ventricle.

E

(Fig. 25-24). The vertical fractures are easily detected in the axial scan plane, but the horizontal fracture may be more difficult to diagnose unless there is some displacement of the fracture fragments. Reformatted images may help define horizontal fractures. These fractures are less common. The cricoid fracture is characteristically a double break (Fig. 25-25). The force of the blow pushes the anterior arch posteriorly, fracturing the ring in two or more places. The ring collapses inward. There is swelling in the subglottic larynx, with apparent soft tissue against the inner surface of the cricoid cartilage, whereas normally the mucosa is very thin or almost invisible at imaging. The concern is compromise of the airway. Any inward displacement reduces the lumen of the airway. The possibility of a tear of the mucosa is also a concern. A fracture fragment protruding through the mucosa often results in chondritis and chondronecrosis with further airway

restriction. If there is any suspicion of a fragment pushed toward or into the airway, direct visualization via endoscopy or open exploration is indicated to thoroughly assess the mucosa. If there is sufficient force to fracture the thyroid cartilage, the lower epiglottis can be torn or avulsed. This can be seen particularly with horizontal fractures where the fragments are pushed posteriorly, creating a shearing force across the lower epiglottis. Because the epiglottis does not calcify significantly, this abnormality may not be detectable by CT scans. The only finding may be hematoma obscuring the fat in the preepiglottic space. The curve of the epiglottis can often be outlined against the preepiglottic fat in the sagittal plane, so sagittal reformatted images can be helpful in evaluating trauma cases. The arytenoid cartilage is not often fractured, but the cricoarytenoid joint can be dislocated. This can occur with fairly minimal trauma to the lateral larynx or result from a more significant insult associated

http://radiologyebook.com

748

PART II  CT and MR Imaging of the Whole Body

C

A

B FIG 25-21  Saccular cyst (laryngocele). A, The cystic structure has both an internal and an external component. The saccular cyst (C) protrudes through the thyrohyoid membrane (approximately at arrowheads). Note the sharp margin of the cyst. B, Slightly lower slice shows the dilated appendix (arrow) just above the ventricle. Again note the sharp margin. The fat in the paraglottic space indicates that this is still above the ventricle. Normal appendix is on the opposite side (arrowhead).

H E

B

A

FIG 25-22  Thyroglossal duct cyst. Postcontrast CT. A, Axial image shows cyst (arrows) along the outer lamina of the thyroid cartilage embedded in the prelaryngeal strap muscles. B, Sagittal image shows the cyst bulging between the thyroid cartilage and hyoid bone (H). The cyst attaches to the posterior margin of the hyoid (arrow). E, epiglottis.

with fractures of the major cartilages. The arytenoid is normally perched just off midline. The radiologist tries to verify that the cartilage is in the normal position or grossly displaced. Currently a subtle displacement cannot be confidently excluded. The lower horns of the thyroid cartilage articulate with the lateral cricoid cartilage. This cricothyroid joint can be dislocated, usually in association with fractures of either the thyroid or cricoid cartilages. Because the attachment of the horn (cornua) to the cricoid cartilage is fairly strong, a fracture of the lower horn is considered to be more common than dislocation. One must be careful in making this diagnosis; slight obliquity of the slice can cause apparent asymmetry of the space between the two cartilages. Each of the inferior horns must be followed down to the point of articulation.

MISCELLANEOUS PATHOLOGY Although tumors and trauma account for most imaging of the larynx, there are several other problems for which imaging can be useful.

Vocal Cord Paralysis Vocal cord paralysis is most commonly a problem affecting the recurrent laryngeal nerve. This nerve innervates all the muscles in the larynx except the cricothyroid muscle. The most common characteristic imaging findings are the result of atrophy of the TAM. Normally this muscle makes up the bulk of the true cord. With atrophy the characteristic CT density or MRI intensity shape is either smaller or no longer seen (Fig. 25-26). The ventricle enlarges into the volume vacated by

http://radiologyebook.com

CHAPTER 25  Larynx

749

E A

A

B

C

D FIG 25-23  Postradiation change. Postcontrast CT. A, Scout view for CT shows the thickened epiglottis (E) and swollen arytenoid (A) prominence. B, Axial image through the level of the thickened epiglottis. C, Axial image at the level of the aryepiglottic folds shows pronounced swelling (arrow). D, Swollen supraglottis (arrows).

FIG 25-25  Fracture of the cricoid cartilage. The ring has been fractured, FIG 25-24  Axial CT scan of vertical fracture of the larynx. Arrow indicates fracture.

and fragments from the anterior ring (arrowheads) have been pushed back into the airway. These fragments would be in danger of perforating the mucosa. The lamina of the cricoid (arrow) is not fractured. Note the air in the soft tissues and the soft tissue within the ring of the cricoid cartilage. (From Curtin HD: MR and CT of the larynx. In Thrall JH, editor: Current practice of radiology, ed 3, St. Louis, 1994, Mosby.)

http://radiologyebook.com

750

PART II  CT and MR Imaging of the Whole Body

A

FIG 25-27  Vocal fold injection. Radiodense injection (Gore-Tex) (arrow) displaces the true vocal fold medially. The arytenoid (A) is also slightly rotated.

A

the atrophied muscle, so air can be seen where there should be muscle. The vocal ligament remains, so the arytenoid may be in a fairly normal position or it may be tipped slightly. The pyriform sinus on the affected side often enlarges as well. These findings can be visualized in both the axial or coronal planes. The posterior cricoarytenoid muscle is very thin, but because the axial plane provides a perfect cross section, atrophy of this muscle may also be appreciated.29 The muscle is located along the posterior aspect of the cricoid cartilage. In this case, fat is seen closely applied to the posterior surface of the cricoid cartilage instead of the typical muscle density (see Fig. 25-26). Surgical treatment of a vocal fold paralysis may include injection of various materials or even placement of fat to add bulk to the paralyzed vocal fold (Fig. 25-27). Prostheses have been designed to add bulk and reposition the arytenoid (Fig. 25-28). In most instances, atrophied muscle is seen in a patient with known paralysis. The diagnosis is usually made by direct visualization by mirror or office endoscopy. If the cause of the paralysis is not known, the course of the vagus nerve and the recurrent laryngeal nerve is examined. CT is usually performed because of the distance that must be covered and because bone detail is helpful at the level of the jugular foramen in the temporal bone. The vagus nerve is just posterior to the carotid artery and jugular vein on transit through the neck. Lower in the neck, the nerve moves forward with respect to the carotid. The left recurrent laryngeal nerve loops around the aortic arch and the right around the subclavian artery. The nerve on both sides ascends along the tracheoesophageal groove to reach the larynx. A lesion at any level can involve the nerve and cause paralysis. Superior laryngeal nerve paralysis is much less common, and characteristic radiographic findings have not been described.

B

C FIG 25-26  Vocal fold paralysis, recurrent nerve. A, Axial T1-weighted image. The ventricle is enlarged (arrow) as the thyroarytenoid muscle (TAM) atrophies. Arrowhead indicates arytenoid. B, Axial T1 image, slightly lower than A. The posterior cricoarytenoid muscle has atrophied and is replaced by fat (arrow). Note the normal posterior cricoarytenoid muscle (arrowhead) on the opposite side. C, Coronal reformatted CT image. The ventricle (arrowhead) has enlarged as the TAM (arrow) atrophies. The vocal cord (fold) is more pointed on the abnormal side.

Stenosis Subglottic stenosis and tracheal stenosis can be evaluated by plain films, but sectional imaging can give information about the crosssectional area of the remaining airway. CT can show narrowing or stenosis of anterior and posterior commissures. The position of the cartilage relative to the airway can also be visualized. CT scans can determine whether the cartilage has collapsed into the airway or whether the cartilage remains in the normal position, with the stenosis the result of soft tissue within the cartilage ring (Fig. 25-29). Such narrowing can be congenital or the result of trauma. Stenosis can follow prolonged intubation. Granulation tissue can develop after

http://radiologyebook.com

CHAPTER 25  Larynx

A

B

A

C

D FIG 25-28  Prosthesis for repositioning paralyzed vocal fold. Axial CT from inferior to superior. A, The prosthesis (arrow) is “popped” into a defect cut through the thyroid cartilage. A small ridge is present at each margin to hold the prosthesis in place. B, Slightly superior to A. C, Slightly superior to B. The arytenoid (A) is pushed medially. D, Superior to C. The tip of the prosthesis (arrowhead) projects posteriorly.

E

A

B

FIG 25-29  Subglottic stenosis. History of prolonged intubation. A, Axial postcontrast CT at the level of the lower cricoid cartilage. There is soft tissue (arrowheads) along the inner cortex of the cricoid cartilage (arrow) that is not collapsed into the airway. This is a “soft” stenosis. B, Coronal reformat airway “cast” showing the level and extent of the narrowing. Note the level of the anterior commissure and ventricle (arrowhead). E, epiglottis.

http://radiologyebook.com

751

752

PART II  CT and MR Imaging of the Whole Body

tracheostomy. The length of the narrowing and the relationship to the vocal folds should be estimated. Plain films can be very helpful in estimating the length of the abnormality, but this information can easily be derived from reformatting of images from CT.

Granulomatous Disease and Inflammation Granulomatous disease, either of infectious or noninfectious etiology, can involve the larynx.17 Though rare, it can involve any of the mucosal surfaces and can be mistaken for tumor. The findings at imaging are nonspecific. Tuberculosis usually occurs with concurrent pulmonary disease. Radiation causes an apparent inflammatory response with swelling of the supraglottis (see Fig. 25-23). Rarely adult supraglottitis can cause severe swelling, predominantly of the supraglottis. More subtle changes can be seen with severe gastroesophageal reflux extending to the pharynx and larynx.

Rheumatoid Arthritis Rheumatoid arthritis can involve the cricoarytenoid joint. This process is not usually evaluated radiologically.

Polychondritis Relapsing polychondritis is a rare disease of unknown etiology. The cartilages of the larynx and trachea (among others) become inflamed, with edema of surrounding tissues. The cartilages may actually appear to enlarge. Calcifications of the abnormal tissue are characteristic and can be very prominent. If particularly extensive, the increased soft tissues and calcifications can mimic chondroid lesions.

SUMMARY CT or MRI can be used for tumor imaging. One approach is to use CT as the initial imaging evaluation, reserving MRI for answering specific questions remaining after CT. MRI is then done in a small fraction of cases directed at a limited specific part of the tumor. Contrast-enhanced CT scans are preferred for imaging of submucosal masses. CT scans are done without enhancement in trauma cases.

REFERENCES 1. Agrawal N, Ha PK: Management of early-stage laryngeal   cancer. Otolaryngol Clin North Am 41:757–769, vi–vii, 2008. 2. Barbu AM, Burns JA, Lopez-Guerra G, et al: Salvage endoscopic angiolytic KTP laser treatment of early glottic cancer after   failed radiotherapy. Ann Otol Rhinol Laryngol 122:235–239, 2013. 3. Barnes L: Diseases of the larynx, hypopharynx, and esophagus. In Barnes L, editor: Surgical pathology of the head and neck, ed 2, New York, 2001, Marcel Dekker, pp 127–237. 4. Becker M, Burkhardt K, Dulguerov P, et al: Imaging of the larynx and hypopharynx. Eur J Radiol 66:460–479, 2008. 5. Becker M, Moulin G, Kurt AM, et al: Non-squamous cell neoplasms of the larynx: Radiologic-pathologic correlation. Radiographics 18:1189– 1209, 1998. 6. Becker M, Zbaren P, Casselman JW, et al: Neoplastic invasion of laryngeal cartilage: Reassessment of criteria for diagnosis at MR imaging. Radiology 249:551–559, 2008. 7. Becker M, Zbaren P, Delavelle J, et al: Neoplastic invasion of the laryngeal cartilage: Reassessment of criteria for diagnosis at CT. Radiology 203:521–532, 1997. 8. Becker M, Zbaren P, Laeng H, et al: Neoplastic invasion of the laryngeal cartilage: Comparison of MR imaging and CT with histopathologic correlation. Radiology 194:661–669, 1995.

9. Castelijns JA, Gerritsen GJ, Kaiser MC, et al: MRI of normal or cancerous laryngeal cartilages: Histopathologic correlation. Laryngoscope 97:1085–1093, 1987. 10. Castelijns JA, van den Brekel MW, Smit EM, et al: Predictive value of MR imaging-dependent and non-MR imaging-dependent parameters for recurrence of laryngeal cancer after radiation therapy. Radiology 196:735–739, 1995. 11. Curtin HD: Imaging of the larynx: Current concepts. Radiology 173:1–11, 1989. 12. Curtin HD: Anatomy, imaging, and pathology of the larynx. In Som PM, Curtin HD, editors: Head and neck imaging, ed 5, Philadelphia, 2011, Mosby, pp 1905–2039. 13. Dadfar N, Seyyedi M, Forghani R, et al: Computed tomography appearance of normal nonossified thyroid cartilage: Implication for tumor invasion diagnosis. J Comput Assist Tomogr 39:240–243, 2015. 14. Dubal PM, Svider PF, Kanumuri VV, et al: Laryngeal chondrosarcoma: A population-based analysis. Laryngoscope 124:1877–1881, 2014. 15. Forghani R, Levental M, Gupta R, et al: Different spectral Hounsfield unit curve and high energy virtual monochromatic image characteristics of squamous cell carcinoma compared to non-ossified thyroid cartilage. AJNR Am J Neuroradiol 2015. In press. 16. Franco RA, Jr, Singh B, Har-El G: Laryngeal chondroma. J Voice 16:92–95, 2002. 17. Glastonbury CM: Non-oncologic imaging of the larynx. Otolaryngol Clin North Am 41:139–156, vi, 2008. 18. Gray H, Strandring S, Ellis H, et al: Gray’s anatomy: The anatomical basis of clinical practice, Edinburgh, 2005, Elsevier Churchill Livingstone. 19. Huang BY, Solle M, Weissler MC: Larynx: Anatomic imaging for diagnosis and management. Otolaryngol Clin North Am 45:1325–1361, 2012. 20. Kirchner JA: Two hundred laryngeal cancers: Patterns of growth and spread as seen in serial section. Laryngoscope 87:474–482, 1977. 21. Kuno H, Onaya H, Iwata R, et al: Evaluation of cartilage invasion by laryngeal and hypopharyngeal squamous cell carcinoma with dualenergy CT. Radiology 265:488–496, 2012. 22. Lawson W, Biller H, Suen J: Cancer of the larynx. In Suen J, Myers E, editors: Cancer of the head and neck, New York, 1989, Churchill Livingstone, pp 533–591. 23. Ljumanovic R, Langendijk JA, Hoekstra OS, et al: Pre- and postradiotherapy MRI results as a predictive model for response in laryngeal carcinoma. Eur Radiol 18:2231–2240, 2008. 24. Ljumanovic R, Langendijk JA, van Wattingen M, et al: MR imaging predictors of local control of glottic squamous cell carcinoma treated with radiation alone. Radiology 244:205–212, 2007. 25. Loehn BC, Kunduk M, McWhorter AJ: Advanced laryngeal cancer. In Johnson JT, Rosen CA, Bailey BJ, editors: Bailey’s Head and Neck Surgery—Otolaryngology, ed 5, Philadelphia, 2013, Wolters Kluwer Health/Lippincott Williams & Wilkins, pp 1961–1977. 26. Mancuso AA, Mukherji SK, Schmalfuss I, et al: Preradiotherapy computed tomography as a predictor of local control in supraglottic carcinoma. J Clin Oncol 17:631–637, 1999. 27. Maroldi R, Ravanelli M, Farina D: Magnetic resonance for laryngeal cancer. Curr Opin Otolaryngol Head Neck Surg 22:131–139, 2014. 28. Pilch BZ: Head and neck surgical pathology, Philadelphia, 2001, Lippincott Williams & Wilkins. 29. Romo LV, Curtin HD: Atrophy of the posterior cricoarytenoid muscle as an indicator of recurrent laryngeal nerve palsy. AJNR Am J Neuroradiol 20:467–471, 1999. 30. Sakai O, Curtin HD, Faquin WC, et al: Dedifferentiated chondrosarcoma of the larynx. AJNR Am J Neuroradiol 21:584–586, 2000. 31. Sinha P, Okuyemi O, Haughey BH: Early laryngeal cancer. In Johnson JT, Rosen CA, Bailey BJ, editors: Head and neck surgery—Otolaryngology, ed 5, Philadelphia, 2013, Wolters Kluwer Health /Lippincott Williams & Wilkins, pp 1940–1960. 32. Tucker GF, Jr, Smith HR, Jr: A histological demonstration of the development of laryngeal connective tissue compartments. Trans Am Acad Ophthalmol Otolaryngol 66:308–318, 1962.

http://radiologyebook.com

CHAPTER 25  Larynx 33. Tufano RP, Stafford EM: Organ preservation surgery for laryngeal cancer. Otolaryngol Clin North Am 41:741–755, vi, 2008. 34. Zan E, Yousem DM, Aygun N: Asymmetric mineralization of the arytenoid cartilages in patients without laryngeal cancer. AJNR Am J Neuroradiol 32:1113–1118, 2011. 35. Zeitels SM, Burns JA: Oncologic efficacy of angiolytic KTP laser treatment of early glottic cancer. Ann Otol Rhinol Laryngol 123:840–846, 2014.

753

36. Zeitels SM, Kirchner JA: Hyoepiglottic ligament in supraglottic cancer. Ann Otol Rhinol Laryngol 104:770–775, 1995. 37. Zeitels SM, Wain JC, Barbu AM, et al: Aortic homograft reconstruction of partial laryngectomy defects: A new technique. Ann Otol Rhinol Laryngol 121:301–306, 2012.

http://radiologyebook.com

26  Imaging of the Head and Neck in the Pediatric Patient Indu Rekha Meesa and Suresh Mukherji

INTRODUCTION The old adage that “children are not young adults” is certainly true when evaluating pediatric head and neck lesions. The differential diag­ nosis of neck masses differs compared to masses that arise in adults. The most common pediatric head and neck lesions are congenital or developmental in origin, followed by inflammatory processes and tumors. This chapter will focus on the congenital/developmental  and neoplastic processes; inflammatory lesions are covered in other chapters.

BENIGN CONGENITAL MASSES OF THE NECK Thyroglossal duct cysts (TDCs) and anomalies of the branchial appa­ ratus are the most common benign masses of the neck in children. Less common congenital benign masses of the head and neck in the pedi­ atric age group include dermoid/epidermoid cysts and teratomas.54

Thyroglossal Duct Cyst TDC is the most common congenital neck mass and accounts for nearly 90% of nonodontogenic congenital cysts.217 TDC is the second most common benign cervical mass in children after reactive ade­ nopathy and almost three times more common than branchial cleft cysts.54,148,204,206

Embryogenesis of Thyroglossal Duct.  At about 3 weeks’ gestation, the thyroid gland begins to develop as a midline endodermal diverticu­ lum from the floor of the pharynx, and over the following 4 weeks of development, the primitive thyroid gland enlarges to form a bilobed diverticulum.54 For a short period of time, the developing thyroid gland is connected to the tongue by a narrow tube, the thyroglossal duct.195 This duct courses from the region of the junction of the ante­ rior two thirds and posterior two thirds of the tongue (foramen cecum) to the hyoid bone and farther caudally to the region of the future loca­ tion of the pyramidal lobe of the thyroid gland in the inferior neck.195 A TDC or fistula forms when a portion of this duct fails to involute and leaves behind rests of secretory epithelial cells that when stimu­ lated by an inflammatory process cause a cyst to develop.195

Presentation.  The TDC is usually a midline or near-midline lesion and can occur anywhere along the path of the duct. About 20% to 25% occur in the suprahyoid neck, 15% to 50% at the level of the hyoid bone, and 25% to 65% in the infrahyoid neck.159 They commonly occur near the hyoid bone and can be superior, anterior, inferior, or posterior to this bone.195 TDCs usually present in the first 5 years of life, with about 66% noted before age 7 and 90% before age 10.54 A small percentage are diagnosed

754

in patients older than 50 years. Patients typically present with a history of a gradually enlarging mass in the midline or paramedian neck and demonstrate a nontender, mobile, 2- to 4-cm subcutaneous midline or paramedian neck mass of variable firmness. Other presentations include complications of cyst infection, fistula formation, or rupture.54 The lining of the cyst may be stratified squamous, pseudostratified ciliated, simple cuboidal, or columnar epithelium, and there may be residual thyroid tissue in the enlarging mass in the midline of the neck.195 Coexisting carcinoma in the wall of the cyst is reported in less than 1% of patients with TDC and typically occurs in adults,  with papillary carcinoma being most common followed by papillaryfollicular carcinoma.3,11,23,35

Imaging Findings.  Suprahyoid neck cysts are almost always in the midline. When the cysts occur just caudal to the hyoid bone at the level of the thyrohyoid membrane of the larynx, they can stretch this mem­ brane and bow it posteriorly, appearing to lie in the preepiglottic space of the larynx. In the infrahyoid neck, the cyst lies just off the midline, adjacent to the outer surface of the thyroid cartilage and deep to the infrahyoid strap muscles.195 In the pediatric population, initial imaging evaluation is frequently performed with ultrasound, which demonstrates a well-defined, thinwalled, midline or paramedian mass with increased through transmis­ sion and variable internal echogenicity and no internal blood flow. Increased internal echogenicity can be due to multiple factors includ­ ing inflammatory debris, hemorrhage, or proteinaceous content.120,216 On computed tomography (CT), the TDC appears as a mass of relatively decreased attenuation to muscle. When the cyst is not infected, it will also demonstrate a thin, smooth rim, but when infected the cyst wall thickens and enhances (Fig. 26-1). The attenuation of the cyst will vary based on its contents; usually they will have mucoid attenuation (10-25 Hounsfield units [HU]), but if there has been pre­ vious infection or hemorrhage, attenuation of the cyst contents can approach that of muscle.195 On magnetic resonance imaging (MRI) the T1 signal intensity can vary from low to high, while T2 signal intensity remains high. The variations in signal intensity are based on the vari­ able protein content of the cyst. With infection or hemorrhage, septa­ tions can also arise in the cyst. A TDC with an internal eccentric solid mass containing calcifica­ tions or infiltrative soft tissue characteristics should raise concern for an associated carcinoma.54 The primary differential considerations of TDC by imaging include dermoid/epidermoid cyst, branchial cleft cyst (if paramedian in loca­ tion), or rarely a sebaceous cyst.148

Treatment.  Treatment is surgical removal of all tissue along the course of the duct as far as the foramen cecum (Sistrunk procedure).

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

755

B

A

FIG 26-1  Thyroglossal duct cyst (TDC). A, Axial contrast-enhanced CT image at the level of the hyoid bone demonstrates a large midline nonenhancing low-attenuation mass (arrows), most consistent with a TDC. B, Axial contrast-enhanced CT image at the level of the hyoid bone demonstrates a large midline predominantly low-attenuation mass with some enhancing components centrally (arrows), concerning for a TDC complicated by papillary thyroid carcinoma.

The hyoid bone rotates during maturation before assuming its final adult position. During this rotation, the thyroglossal duct is adherent to the hyoid along its anterior inferior edge and may be drawn poste­ riorly and cranially to lie behind the body of the hyoid bone; rarely the duct can be incorporated into the hyoid bone, trapped between the second and third arch components of the hyoid’s body.15,58 Owing to this close relationship between the body and the TDC, Sistrunk pro­ posed that the body of the hyoid bone should be removed during surgi­ cal resection of a thyroglossal duct, leading to a significant decrease in recurrences, reduced from nearly 50% to less than 4%.186

Anomalies of the Branchial Apparatus

Embryogenesis of the Branchial Apparatus.  The anatomic struc­ tures of the face and neck predominantly derive from the branchial apparatus, which is a complex structure derived from neural crest cells that develops between the second and seventh weeks of gestation.27,54 The branchial apparatus consists of six paired arches separated on their outer surface by five paired ectodermal clefts and on their inner surface by five paired endodermally derived pharyngeal pouches.27 By the end of the fourth week of life, four well-defined pairs of arches are visible and the fifth and sixth arches are rudimentary. Each arch is composed of a central core of mesoderm and is lined externally by ectoderm and internally by endoderm. The cervical sinus of His is formed after the branchial arches appear by accelerated mesodermal growth cranially of the first arch and a portion of the second arch, and caudally by growth of the epipericar­ dial ridge, which develops from the mesoderm lateral to the fifth-sixth arch. The branchial apparatus typically disappears between the fourth and sixth weeks of life.27

Etiology of Branchial Anomalies.  The etiology of branchial anom­ alies is controversial, and the most accepted theories to explain the

development of branchial abnormalities propose that they are either vestigial remnants from incomplete obliteration of the branchial appa­ ratus or the result of buried epithelial cell rests.126,29

Classification.  Anomalies of the branchial apparatus are best under­ stood by being classified into a spectrum of developmental anomalies that includes fistulas, sinuses, and cysts.195

First Branchial Anomalies.  First branchial cleft anomalies account for up to 5% to 8% of all branchial anomalies: 68% are cysts, 16% are sinuses, and 16% are fistulas.51,63,139 Although Arnot and Work tried to subclassify first branchial anomalies into two distinct subtypes in 1971 and 1972, Olsen—after reviewing multiple cases—concluded  that it is difficult to subclassify all these anomalies into one of the subtypes and proposed a simplified classification (cysts, sinuses, and fistulas).70,152,183,209,224 Embryology.  The first branchial apparatus gives rise to portions of the middle ear, external auditory canal, eustachian tube, mandible, and maxilla, and the formation of these structures is completed by the sixth and seventh weeks of gestation.54 The parotid gland and facial nerve form and migrate between the sixth and eighth weeks of gestation. Because the facial nerve and parotid gland have a somewhat later embryologic development, vestigial branchial anomaly can be located in variable relationship to the parotid gland and the facial nerve.195 First branchial apparatus anomalies arise from incomplete closure of the ectodermal portion of the first branchial cleft. They can start at the junction of the bony and cartilaginous external auditory canal or in the cartilaginous portion of the external canal, the plane between the mandibular and hyoid arches, and end in the subman­ dibular area. As a result, first branchial apparatus anomalies can arise in the middle ear cavity, external auditory canal, superficial or deep lobes of the parotid gland, superficial to the parotid gland, at the

http://radiologyebook.com

756

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 26-2  First branchial cleft anomaly. A, Coronal contrastenhanced CT image demonstrates a low-attenuation lesion (arrow) within the right parotid gland associated with asymmetric enlargement and high attenuation. B, Axial T2-weighted MRI at the level of the parotid gland demonstrates a rounded T2 bright lesion (arrow) within the right parotid gland. C, Coronal T1 postcontrast MRI demonstrates a nonenhancing cystic lesion in the right parotid gland, associated with asymmetric enlargement and enhancement of the parotid tissue secondary to an inflamed branchial cleft cyst (arrows).

C

angle of the mandible, anterior or posterior to the pinna, or along the nasopharynx.14,209 Presentation.  Although the majority of first branchial apparatus anomalies occur in the first decade of life, they can also be encountered in middle-aged patients, especially women. Patients typically present with recurrent abscesses or inflammatory processes with tenderness and swelling at the angle of the mandible or in the ear. If the process is related to the parotid gland, patients tend to present with recurrent parotid abscesses not responding to therapy. If there is drainage of the cyst into the external auditory canal, the initial presentation is typically otorrhea.54 Simple cysts can also present as a painless soft tissue mass. Sinus tracts are rare with first branchial anomalies. When present, they are usually found in the first decade of life.204 Imaging findings.  First branchial cysts are usually related to the parotid gland, external canal of the ear, and/or the lower margin of the pinna or at the angle of the mandible.54 If a tract can be identi­

fied, directed toward the external auditory canal, then the diagnosis can be made with certainty. If a tract is not seen and the cyst lies within the parotid gland, possibilities would include a first branchial cyst, parotid suppurative lymph node and abscess, inflamed ectatic salivary gland, a lymphoepithelial cyst (human immunodeficiency virus [HIV]), Sjögren’s syndrome, or a rare localized obstructive mucocele.54,105,195 Uncomplicated cysts demonstrate mild rim enhancement or no enhancement. Infected cysts have variable wall thickness and enhance­ ment characteristics (Fig. 26-2). If the anomaly is located within the parotid gland, the imaging findings will be nonspecific. For example, the cyst contents of a first branchial cyst, a lymphoepithelial cyst, or an obstructive mucocele or sialocele are usually of mucoid attenuation on CT, and on MRI usually have low to intermediate T1-weighted and high T2-weighted signal intensity.195 T2-weighted or short tau inver­ sion recovery (STIR) MRI sequences may be helpful in identification

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient of small fluid-filled tracts leading to the external auditory canal.54 If the cyst is infected and has a tract, coronal T2-weighted and postcon­ trast T1-weighted fat-suppressed MRIs would best demonstrate the tract.195 Treatment.  Treatment is complete surgical resection and carries an excellent prognosis. The major complication of surgery is related to facial nerve injury. Residual wall remnants can lead to recurrence.54

Second Branchial Anomalies.  About 95% of all branchial anoma­ lies are related to the second branchial apparatus, and these anomalies can occur anywhere from the tonsillar fossa to the supraclavicular region of the neck.54,80 The majority of these anomalies, about three fourths, are cysts.205 Cysts are more common between 10 and 40 years of age, and fistulas or sinuses usually present before age 10.133,194 Classification.  Second branchial cysts are classified into four sub­ types based on location per Bailey’s classification. Type 1 cyst lies deep

to the platysma muscle and anterior to the sternocleidomastoid (SCM) muscle (Fig. 26-3). This type of cyst is felt to represent a remnant of the tract between the sinus of His and the skin. Type II cyst is the most common type and is thought to be the result of the persistence of the sinus of His. This cyst is located posterior and lateral to the subman­ dibular gland, anterior and medial to the SCM muscle, and anterior and lateral to the carotid space (Fig. 26-4). Type III cyst/fistula is thought to arise from a dilated pharyngeal pouch and courses medially between the internal and external carotid arteries; it can extend up to the lateral wall of the pharynx or skull base (Fig. 26-5). Type IV cyst lies in the mucosal space of the pharynx adjacent to the pharyngeal wall and is thought to arise from a remnant of the pharyngeal pouch5,54,135,147,179 (Fig. 26-6). Presentation.  The most common presentation of a second bran­ chial cleft cyst is as a fluctuant nontender mass at the lateral aspect of the mandibular angle.135,191,194 When infected the patient will present

B

A

FIG 26-3  Second branchial cleft anomaly, type 1. A, Axial contrast-enhanced CT image at the level of the thyroid cartilage demonstrates a nonenhancing rounded lesion (arrow) adjacent to the strap muscle. B, Axial contrast-enhanced CT image at the level of the thyroid gland demonstrates inflammatory changes along the platysma and adjacent to the left sternocleidomastoid muscle (arrows).

A

757

C

B

FIG 26-4  Second branchial cleft anomaly, type 2. Axial (A), coronal (B), and sagittal (C) contrast-enhanced CT images at the level of the submandibular gland demonstrate a rounded fluid-attenuation lesion located anterior to the right sternocleidomastoid muscle, anterolateral to the carotid space and posterior to the submandibular gland, resulting in mild mass effect and anterior displacement of the right submandibular gland (arrows).

http://radiologyebook.com

758

PART II  CT and MR Imaging of the Whole Body

A

C

B

D FIG 26-5  Second branchial cleft anomaly, type 3. Axial contrast-enhanced CT images demonstrate an enhancing fistulous tract extending from the left tonsillar fossa (A) to the skin surface at the left of the sternocleidomastoid (SCM) muscle (B) in the lower neck (arrows). Coronal CT image of the fistula injection demonstrates a tract extending from the right tonsillar region (C) to the right SCM and carotid jugular sheath (D) (arrows).

with a history of a slowly enlarging painful mass. Because the cysts contain lymphoid tissue, stimulus such as an upper respiratory tract infection can lead to an increase in size of the cyst.54 With a fistula, an ostium is often noted at birth at the anterior border of the junction of the middle and inferior thirds of the SCM muscle. The tract then courses deep to the platysma muscle, ascends laterally along the carotid sheath, lateral to the hypoglossal and glossopharyngeal nerves, and then passes between the internal and external carotid arteries before it terminates in the region of the palatine tonsillar fossa.5,14 Imaging findings.  Ultrasound is usually the initial imaging study because the cysts often present as a neck mass in a young child and can range in size from 1 to 10 cm.5,14,46,54,205 If the anomaly is uncomplicated

it will demonstrate sonographic characteristics of a simple cyst (thin wall, increased through transmission, anechoic, and compressible). On CT the anomaly will demonstrate a thin wall with decreased attenua­ tion. On MRI it will appear as hypointense to slightly hyperintense  to muscle on T1-weighted images and hyperintense to muscle on T2-weighted images. When it is infected it can demonstrate a thicker wall with increased internal echogenicity on ultrasound, increased attenuation on CT, and hyperintense T1-weighted signal on MRI, with variable enhancement. If the cysts are chronically complicated, they can demonstrate septations.14,80,135,139,204 Sometimes there may be a “beak sign” present, which is pathogno­ monic of a Bailey type III second branchial cyst, where the medial

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

759

B

A

FIG 26-6  Second branchial cleft anomaly, type 4. Axial (A) and coronal (B) contrast-enhanced CT images demonstrate a rounded nonenhancing cystic lesion in the right tonsillar fossa (arrows).

aspect of the cyst is compressed and forms a beak as it extends between the internal and external carotid arteries on axial CT or MRI.78 In children, differential considerations include suppurative adenopathy/deep neck infection, lymphatic malformation, dermoid/ epidermoid cyst, and thymic cyst. Other rare considerations in this age group include cystic schwannoma of the vagal nerve, minor salivary gland tumor, and necrotic malignant adenopathy.80,143,181 Because the jugulodigastric node is seen in the same location as the most common form of second branchial cleft cyst, clinicians may mistake it for an enlarged, suppurative, reactive, or tumor-infiltrated jugulodigastric node.78 Treatment.  Treatment is complete surgical resection, which carries an excellent prognosis.54

Third and Fourth Branchial Anomalies.  These are rare anomalies, and third branchial anomalies account for 3% and fourth branchial anomalies account for about 1% to 2%.40 Third branchial anomalies are centered in the posterior cervical space, and despite their overall rarity they are the second most common congenital lesions of the posterior cervical space after lymphatic malformation. These must sometimes be distinguished from the large second branchial cleft cyst, which can protrude posteriorly in the posterior cervical space.54 Embryology.  Third branchial anomalies most likely arise second­ ary to failure of involution of the third branchial apparatus, and fourth branchial anomalies are likely due to failure of regression of the fourth branchial pouch or distal sinus of His.40,155 Third branchial apparatus cysts are located posterior to the common carotid or internal carotid artery, between the hypoglossal nerve (below) and glossopharyngeal nerve (above). If it is in the form of a fistula, there will be a cutaneous opening similar to a second branchial fistula that is anterior to the lower SCM muscle. It then courses pos­ terior to the common or internal carotid artery, anterior to the vagus nerve and between the hypoglossal and glossopharyngeal nerves, and then pierces the thyrohyoid membrane and enters the pyriform sinus anterior to the superior laryngeal nerve.54 Presentation and imaging findings.  Third branchial cleft cysts usually present as painless, fluctuant, 2- to 5-cm masses in the pos­ terior triangle of the neck, often after a viral upper respiratory infec­

FIG 26-7  Third branchial cleft anomaly. Postcontrast axial CT image at the level of the mandible demonstrates a low-attenuation lesion in the posterior cervical space posterior to the carotid vessels and the sternocleidomastoid muscle (arrow).

tion (Fig. 26-7). The presentation is at an earlier age with a sinus or a fistula. On CT and MRI, uncomplicated cysts usually present as a unilocular cystic mass on imaging, and complicated cysts demonstrate increased wall thickness with variable enhancement characteristics and internal proteinaceous fluid.54 Anomalies related to the fourth branchial pouch are usually in the form of a sinus tract that arises from the pyriform sinus, pierces the

http://radiologyebook.com

760

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 26-8  Fourth branchial cleft anomaly. A, Contrast-enhanced axial CT image at the level of the thyroid gland demonstrates abnormal low attenuation, enlargement, and enhancement of the left thyroid lobe (arrow). Axial T2-weighted (B) and contrast-enhanced T1-weighted (C) MRIs demonstrate abnormal signal and enhancement with inflammatory changes in the left thyroid lobe (arrows). D, Esophagram demonstrates a sinus tract from the left pyriform sinus (arrow).

thyrohyoid membrane, descends along the tracheoesophageal groove, and continues into the mediastinum. These anomalies have variant presentations such as a nontender fluctuant mass anterior to the infe­ rior SCM muscle for uncomplicated cases, and suppurative thyroiditis or recurrent neck abscesses for complicated cases54 (Fig. 26-8). For diagnosing a fistulous tract by imaging, a contrast-enhanced CT fol­ lowing ingestion of barium or water-soluble contrast or direct injec­ tion of a fistulous tract is the imaging study of choice.54 Although it is difficult to differentiate third from fourth branchial anomalies by imaging alone, the key anatomic markers for correctly diagnosing them include the carotid artery and the superior laryngeal nerve. Third branchial anomalies lie posterior to the internal carotid

artery and above the superior laryngeal nerve, whereas fourth bran­ chial anomalies are anterior to the internal carotid artery and lie below the superior laryngeal nerve.8,54 Third and fourth branchial anomalies can be related to the pyriform sinus and can appear similar to external laryngoceles on imaging.54 Other differential considerations for third and fourth branchial anomalies include other branchial anomalies, TDC, lymphatic malfor­ mation, thyroid abscess, thyroid, parathyroid, or thymic cyst, and sup­ purative adenopathy.14,139 Treatment.  For both third and fourth branchial anomalies, com­ plete surgical resection is the treatment of choice. Persistence of   the tract to the pyriform sinus can result in recurrent episodes of

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient thyroiditis and can necessitate a partial thyroidectomy. If fourth bran­ chial anomalies are complicated by infection and secondary thyroiditis, medical treatment should precede surgical treatment.54

Cervical Thymic Remnant Embryology and etiology.  Cervical thymic remnants are very rare lesions, and two main theories, congenital theory and acquired theory, have been reported to explain their etiology.54 The congenital theory states that the cysts are due to persistence of thymopharyngeal duct remnants and can be located anywhere from the level of the pyriform sinus to the superior mediastinum and located either beneath or medial to the SCM muscle. The acquired theory for thymic cysts pro­ poses progressive cystic degeneration of thymic (Hassall’s) corpuscles, primitive endodermal cells, lymphocytes, and epithelial reticulum of the thymus as sources for thymic remnants. Presentation.  The thymus reaches its greater relative size at age 2 to 4 years and its greatest absolute size at puberty. Cervical thymic remnants are very rare lesions and can be found anywhere along the path of the thymopharyngeal duct. Most are detected in childhood, and two thirds are found in the first decade of life, often between 3 and 8 years of age, and the remain­ der in the second and third decades of life.22,78,194,219 Most patients are asymptomatic, and about 80% to 90% present with a painless slowly enlarging neck mass near the thoracic inlet, anterior or deep to the SCM muscle.41,52,57,59 In 50% of cases, a connec­ tion between the cervical thymic anomaly and the mediastinal thymic gland occurs in the form of direct extension or a remnant of the thy­ mopharyngeal duct.54 Mediastinal thymic cysts can occur without a cervical component. There is a frequent association between cervical thymic cysts and thyroid and parathyroid inclusion cysts, and a fibrous strand is sometimes present between the cervical thymic cyst and the thyroid gland.8,57 Imaging findings.  Because most ectopically located thymic masses are often cystic on ultrasound, CT and MRI will demonstrate imaging characteristics consistent with an uncomplicated cyst. A cystic lesion will course parallel to the SCM muscle and along the carotid sheath or lateral aspect of the visceral space on CT (Fig. 26-9). Imaging charac­ teristics will vary if the cysts are complicated by hemorrhage, choles­ terol, proteinaceous fluid, or infection.22,194 Aberrant thymic tissue,

A

761

parathyroid tissue, or lymphoid tissue can be present as soft tissue components related to the wall of the cyst.54 Differential diagnosis.  A second branchial apparatus fistula is the main differential consideration of thymic remnants. A second bran­ chial fistula will pass between the internal and external carotid arteries and end in the superior tonsillar pillar, whereas a thymic cyst will pass posterior to the carotid bifurcation and terminate in the pyriform sinus.54 Additionally, about 50% of cervical thymic cysts extend into the superior mediastinum, and some can be seen in the inferior  pole of the thyroid gland.8,104,146 Other differential considerations for a thymic cyst include a fourth branchial anomaly, TDC, thyroid/ parathyroid cyst, lymphatic malformation, cyst neuroblastoma, lymph­ adenopathy, external laryngocele, and vallecular cyst.54 Treatment.  Treatment is surgical, and the prognosis is excellent if the resection is complete. There is a high recurrence rate with incom­ plete resection.22,98

Dermoid/Epidermoid Cyst Because of overlapping features, the terminology and classification of dermoid, epidermoid, and teratoid cysts in the head and neck can lead to confusion.

Embryology.  These masses are thought to arise from trapped epithe­ lial cell rests at the site of midline closure. Dermoids and epidermoids are ectodermal-lined inclusion cysts, and teratoids can contain tissues of all three germ-cell layers. Dermoids contain epithelial and dermal elements, whereas epidermoids contain only epithelial elements.54

Presentation.  There is no gender predilection with dermoid and epidermoid cysts. Dermoid cysts of the neck commonly occur near the midline, with the most common location being the floor of the mouth, (11.5% of all dermoids), followed by other locations including the anterior neck, tongue, palate, and orbit.101,144,145 Epidermoid cysts are rarely found in the head and neck, and when they occur they usually present in infancy.191 Dermoid/epidermoid cysts become clinically apparent in the second or third decades of life, when they present as a painless, soft, slow-growing mass in the suprahyoid midline neck. The cyst can be variable in size, and rapid increase in size can be seen secondary to

B FIG 26-9  Thymic cyst. Axial (A) and coronal (B) contrast-enhanced CT images demonstrate a cystic mass (arrows) in the lateral infrahyoid neck at the level of the thyroid gland.

http://radiologyebook.com

762

PART II  CT and MR Imaging of the Whole Body

association with a sinus tract, pregnancy, or increasing desquamation of skin appendage products.89,191,194 Masses lying in the sublingual space are often clinically inapparent and demonstrate minimal mass effect, but masses in the submandibular space present with a more obvious swelling or mass effect.89,215 Additionally, a distinguishing factor between TDCs and dermoid/epidermoid cysts are that whereas TDCs move with protrusion of the tongue, owing to their association with the hyoid bone, dermoid/epidermoid cysts are nonmobile with this maneuver.154 Dermoid cysts are well-defined encapsulated masses lined with squamous epithelium and may contain skin appendages in the form of sweat or sebaceous glands or hair follicles, which can lead to the formation of keratin and sebaceous material, including occasional hair. Epidermoid cysts lack skin appendages but have a histologic appear­ ance similar to dermoids.191 The incidence of malignant degeneration into squamous cell carcinoma is about 5% for dermoids.

The signal characteristics of this lesion can be confused with other cystic lesions such as an uncomplicated lymphatic malformation or ranula. It is important to accurately localize the lesions when they are located in the floor of the mouth. If the lesion is located above the mylohyoid muscle, it lies in the sublingual space and is amenable to an intraoral surgical approach. If it lies in the submandibular space, a submandibular approach would be used. The optimal imaging plane for localizing these masses into the appropriate space would be coronal CT or MRI.191

Teratoma Approximately 9% of all pediatric head and neck neoplasms are tera­ tomas, and less than 5% of all teratomas occur in the head and neck.38,99 In the pediatric population, most teratomas are extragonadal in loca­ tion, with 82% in the sacrococcygeal area; locations in the head and neck include neck, nasal cavity, paranasal sinus, oral cavity, orophar­ ynx, nasopharynx, temporal bone, ear, and orbit.7,112,201,202 Teratomas are thought to arise from misplaced pluripotential pri­ mordial germ cells and are classified as true neoplasms because of their biological behavior with progressive and invasive growth patterns.170

Imaging Findings.  On CT a dermoid cyst can appear as a thinwalled unilocular mass with decreased attenuation and at times with CT attenuation values approximating that of fat (Fig. 26-10). Some­ times a virtually pathognomonic “sack-of-marbles” appearance, repre­ senting multiple small fatty nodules within the fluid, is seen within the cyst. These cysts can also present with a more heterogeneous appear­ ance owing to differences in composition of the various skin append­ age components, including calcification. Thin peripheral enhancement can be seen following contrast administration.191 On MRI, signal changes vary depending on the composition of the various cyst components. T1 signal can be hyperintense due to lipid components or isointense to muscle, and T2 signal is usually hyperin­ tense to muscle. If calcification is present, this will cause low T2 signal. If a dermoid contains minimal complex elements, it can appear similar to an epidermoid. Postcontrast images may or may not demonstrate rim enhancement. T1 postcontrast fat-saturation images can be helpful to diagnose dermoids by assessing for signal dropout. Epidermoid cysts usually follow water signal, demonstrating homogenous low attenuation on CT and hypointense T1-weighted and hyperintense T2-weighted signal to muscle on MRI (Fig. 26-11).

A

Cervical Teratomas.  Cervical teratomas are large and can present with extensive unilateral or more diffuse cervical swelling that can lead to obstetric complications such as difficult labor with malpresentation, premature labor, stillbirth, acute neonatal respiratory distress, and maternal polyhydramnios. The tumors that present later in life are smaller but have a higher incidence of malignancy.170 Embryology.  Cervical teratomas can contain variable degrees of maturation, with mature elements derived from ectoderm, mesoderm, and endoderm and immature tissue mostly from neuroectoderm. Both mature and immature neuroectodermal elements tend to predominate in cervical teratomas.54 Imaging findings.  On antenatal ultrasound, teratomas can appear well defined or infiltrative and demonstrate heterogeneous echo­ genicity with multiloculated areas of cystic and solid regions and acoustic shadowing from calcifications on sonography.54,170

B FIG 26-10  Dermoid. A and B, Axial CT images demonstrate large fat-containing lesions (arrows) in the sublingual space.

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

763

B

A

FIG 26-11  Epidermoid. A, Axial contrast-enhanced CT image demonstrates a nonenhancing midline lowdensity lesion (arrow) in the oral cavity, most consistent with an epidermoid. B, Sagittal T2-weighted fetal MRI demonstrates a T2 bright lesion (arrow) at the floor of the mouth.

A

B

C

FIG 26-12  Teratoma. A, Lateral radiograph demonstrates a large soft tissue mass with calcifications (arrow). Noncontrast coronal CT images in soft tissue (B) and bone windows (C) demonstrate a large soft tissue mass with calcifications and fat (arrows).

On CT, the imaging characteristics will reflect the pathologic makeup of the tumor. Most cervical teratomas are large, measuring up to 12 cm, and can appear primarily cystic, solid, or multicystic; about 50% contain calcification170,221 (Fig. 26-12). On MRI, the mass shows heterogeneous signal on all sequences and contains areas of  T1 hyperintense signal in tissues containing fatty components and hypointense or hyperintense signal in areas of calcification.170,221 Vari­ able enhancement is seen after contrast administration. Because these tumors are large and can lead to airway compromise, intubation is often required.

Differential diagnosis.  The primary differential diagnosis is lym­ phatic malformation, which more closely follows water signal on all pulse sequences. Because cervical teratoma can be difficult to distin­ guish from a lymphatic malformation with prenatal ultrasound, fetal MRI has been found to be helpful in such cases. Other differential considerations include cervical neuroblastoma (which can contain cal­ cification), complicated TDC or branchial cleft cyst, goiter, hemangi­ oma, and external laryngocele.45 Treatment.  For congenital cervical teratomas, surgery is manda­ tory and the mortality rate is 80% to 100% secondary to airway 

http://radiologyebook.com

764

PART II  CT and MR Imaging of the Whole Body

complication. Therefore differentiating this tumor from other con­ genital space-occupying masses of the neck is important. The inci­ dence of true malignancy is less than 5%, and associated anomalies are uncommon. Prognosis with surgery is excellent.75,157

CONGENITAL MASSES OF THE NOSE

derm with them as they retreat, creating a dermal/ectodermal tract that extends from the glabella through a canal at the frontonasal suture to the crista galli or beyond the crista to the interdural space between the two leaves of the falx.34,73,175 Depending on the contents and patency of the diverticulum, the resultant lesion can be a dermal sinus, dermoid cyst, nasal glioma, or encephalocele.83,156,229

Nasal Dermal Sinus Cyst

Presentation.  NDSC commonly presents as a painless cystic enlarge­

Nasal dermal sinuses are thin epithelium-lined tubes that arise at exter­ nal ostia situated along the midline of the nose and extend deeply for a variable distance, extending sometimes to the intracranial compart­ ment. Nasal dermoid and epidermoid cysts are midline epitheliumlined cysts and may coexist with dermal sinuses or present as isolated masses. Nasal dermal sinus cyst (NDSC) is the most common midline congenital lesion and accounts for 1% to 3% of all dermoids and 3.7% to 12.6% of all dermoids of the head and neck.151,161,162,175 Although most are sporadic, several familial cases have been reported.18,138 They can present at any age but are most commonly found in younger chil­ dren, with a mean age of presentation of 3 years. There is a male predilection, and these lesions are found anywhere between the gla­ bella and the base of the columella.161,162 About 56% present as midline cysts and 44% as midline sinus ostia.156

ment of any part of the nose, associated with one or more dimples.156 Externally, midline pits, fenestra, or discrete masses can be present, and they may be associated with hair protruding from the defect.18,21,88,175 With intermittent episodes of inflammation or discharge of sebaceous or inflammatory material there can be a variation in the size of the mass. Although there is no consistent associated pattern of malfor­ mation of the face and brain, an increased incidence of intracranial extension of the NDSC has been found in patients with multiple malformations.175,218 The extension of nasal dermoid cysts and dermal sinuses is variable, which may end blindly in the superficial tissues or extend both intra­ cranially and extracranially.175 The reported intracranial extension varies between 0% and 57%.131,151 Intracranial extension of the der­ moids is typically extraaxial, with extension usually occurring through the foramen cecum into the epidural space in the region of the crista galli (Fig. 26-13). When intracranial extension is present, signs and symptoms of meningitis or behavioral changes related to life-threatening frontal lobe abscess can be seen. Other complications include nasal abscess, periorbital-nasal cellulitis, osteomyelitis, cavernous sinus thrombosis, and seizures.151,175,218

Embryology.  Early in the embryo, the developing frontal bones are separated from the developing nasal bones by the fonticulus frontona­ salis.73,131,175 The nasal bones are separated from the cartilaginous nasal capsule by the prenasal space, which extends from the base of the brain to the nasal tip.131 The most accepted theory for the development of NDSC is the failure of involution of the anterior neuropore, with lack of regression of the diverticulum of dura that extends through the foramen cecum. The midline diverticula of dura normally project anteriorly into the fonticulus frontonasalis and anteroinferiorly into the prenasal space and touch the ectoderm. These diverticula usually regress before closure of the bone plates of the anterior skull base, but if the embryonic diverticula of dura become adherent to the superficial ectoderm, they may not regress normally and instead may pull ecto­

A

Imaging Findings.  Both CT and MRI are useful in evaluating the course of the sinus tract, cysts, and sequelae of infection. On CT, the course of a dermal sinus is seen as a well-defined gap between the nasal bones in the midline or in the frontal bone. An enlarged foramen cecum and bifid-appearing or distorted crista galli are suggestive of but not pathognomonic for intracranial extension. These findings can present with or without intracranial extension or with intracranial

B FIG 26-13  Nasal dermoid. Axial (A) and sagittal (B) T1-weighted MRIs demonstrate a fat-containing lesion at the nasofrontal region, with intracranial extension (arrows).

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

765

extension of a fibrous cord without dermoid elements. The intracranial ends of dermoid cysts usually lie in a hollowed-out space along the anterior surface of a thickened enlarged crista galli. The ostium and tract usually appear as isodense fibrous channels or as lucent dermoid channels that can extend inward for variable distances. The bony canals indicate the course of the tract through the nasal bones, ossified nasal septum, and skull base. On CT, noninfected dermoid cysts appear as nonenhancing masses of fat attenuation with an isodense rim to soft tissue. Stranding of the surrounding soft tissues and change in attenuation value of the cyst suggest secondary infection. An infected dermoid cyst has an appear­ ance similar to an abscess, with a peripheral rim of enhancement. A foramen cecum can be incompletely ossified, a normal finding in chil­ dren younger than 1 year of age, appearing as a bony defect.54 In such cases, contrast administration can aid in the differentiation of enhanc­ ing cartilage in a young child from a nonenhancing true skull base defect, similar to differentiation of enhancing nasal mucosa from non­ enhancing dermoids.97,161,162 MRI is extremely valuable in assessing the intracranial and nasal components of these lesions. Thin-slice sagittal, coronal, and axial T1-weighted, T2-weighted with fat saturation, and fast-spin echo inversion recovery sequences should be obtained. Sagittal T1-weighted and inversion recovery sequences are best for demonstrating enlarge­ ment of the foramen cecum with intracranial extension, and postcon­ trast imaging with fat saturation has similar appearances as CT. Dermoid cysts commonly have close to fat signal as compared to brain on all sequences, with associated chemical shift artifact and suppres­ sion with fat saturation. Fibrous tracts demonstrate an isointense signal to brain on T1-weighted sequences.1

and communicate via a defect in the nasal bones or around the lateral edges.137,225

Treatment.  The nasal dermal sinuses are resected for three major

Imaging Findings.  On imaging, nasal gliomas and encephaloceles

reasons: for cosmetic purposes, to avoid/treat complications of local infection, and to avoid/treat secondary complications such as second­ ary meningitis and cerebral abscess.167 To avoid unnecessary cranioto­ mies, surgical studies suggest that the best approach is to dissect the extracranial portion of the tract along its entire length from the super­ ficial ostium to the extracranial surface of the enlarged foramen cecum and sever the tract at the extracranial end of the foramen cecum. Then send the severed end for pathologic examination, and if the specimen shows (epi)dermal elements at the foramen cecum, the dissection is extended intracranially. If no (epi)dermal elements are found at the foramen cecum, the procedure is concluded without intracranial exploration.102,175

appear as masses of soft tissue signal isointense to gray matter.  The differentiation between a nasal glioma and an encephalocele  is made based upon identification of intracranial communication. If an intracranial communication is identified, the mass is classified as an encephalocele. This is best evaluated with MRI in the sagittal plane.54

Nasal Gliomas (Heterotopias) Nasal gliomas are congenital masses of glial tissue that may occur intranasally and/or extranasally at or near the root of the nose and may or may not be connected to the brain by a pedicle of glial tissue.160 They do not contain cerebrospinal fluid (CSF)-filled space that is con­ nected with either of the CSF-containing spaces in the brain (ventricles or subarachnoid space). They are uncommon and account for only 4.5% of congenital nasal masses.175,180 They occur sporadically with no familial tendency and usually affect both genders equally.73 About 15% of patients with nasal heterotopias also manifest multiple cerebral heterotopias.225 Nasal gliomas are subclassified into three types: extranasal (60%), intranasal (14%-30%), and mixed (10%-14%) forms. Extranasal gliomas lie external to the nasal bones and nasal cavities and typically occur at the bridge of the nose, to the left or right of the midline. Intranasal gliomas lie within the nasal or nasopharyngeal cavities. Mixed nasal gliomas consist of extranasal and intranasal components

Presentation.  Extranasal gliomas present in early infancy or child­ hood as firm, slightly elastic, reddish to bluish, skin-covered masses, and capillary telangiectasias may cover the lesion.180 They do not pulsate, do not increase in size with the Valsalva maneuver, and do not pulsate or swell following compression of the ipsilateral jugular vein.9,19,73 Intranasal gliomas lie within the nasal or nasopharyngeal cavities and present as large, firm, polypoid, submucosal masses that may extend inferiorly toward the nostril.73,187 They may protrude through the nostril and usually attach to the turbinates and come to lie medial to the middle turbinate between the middle turbinate and nasal septum.73 These can lead to obstruction of the nasal passage or blockage of the nasolacrimal duct, causing epiphora on the affected side. CSF rhinorrhea, meningitis, and epistaxis may be additional com­ plications and presenting complaints. Intranasal gliomas are commonly confused with inflammatory polyps, which usually lie medial to the middle turbinate, whereas inflammatory polyps typically lie inferolateral to the middle turbinate. Nasal gliomas usually present in infancy, whereas ordinary nasal polyps are rare in children younger than 5 years.199 Histologically, nasal gliomas resemble reactive gliosis rather than neoplasia and consist of large aggregates or small islands of glial tissue.33,199,225 Fibrous connective tissue enwraps the blood vessels and extends outward to form septa that divide the mass.187 Extranasal gliomas are surrounded by dermis, and intranasal gliomas are sur­ rounded by minor salivary glands, fibrovascular tissue, and nasal mucosa.73

Cephaloceles Cephaloceles are defined as herniations of intracranial contents through a cranial defect.141 When the herniation contains only menin­ ges, it is defined as a cranial meningocele, and if the herniation contains brain, then it is called a meningoencephalocele. They are classified by the site of the cranial defect through which the brain and meninges herniate.141,185 Cephaloceles are common lesions, occurring in 1 per 4000 live births.17,103 The types of cephaloceles vary based on different populations. Sincipital cephaloceles are found in 1 in 35,000 live births in North America and Europe but are more common in Southeast Asia, where they are found in 1 per 5000 to 6000 live births.30,88 In Cauca­ sians, occipital cephaloceles are most common (67%-80%) whereas sincipital cephaloceles (2%-15%) and basal cephaloceles (10%) are uncommon.62,141 Sincipital encephaloceles commonly present in the neonatal or infantile period with a nasal mass or nasal stuffiness, and nasopharyngeal basal cephaloceles present toward the end of the first decade with mouth breathing or nasal stuffiness.54

Anterior (Sincipital Cephaloceles).  Anterior cephaloceles, also known as sincipital cephaloceles, are situated in the anterior part of the skull and include both interfrontal and frontoethmoidal cephalo­ celes.199 They always present as external masses along the nose, orbital margin, or forehead. With the interfrontal cephalocele, the cranial

http://radiologyebook.com

766

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 26-14  Anterior basal encephalocele. Axial T1-weighted (A) and coronal postcontrast (B) T1-weighted MRIs at the anterior skull base demonstrate a nonenhancing CSF-filled structure extending through a defect in the dura into the posterior ethmoid air cells and sphenoid sinus from the right middle cranial fossa (arrows). Sagittal T1-weighted (C) and axial T2-weighted (D) MRIs demonstrate a CSF signal lesion extending through a defect in the dura into the posterior ethmoid and sphenoid sinuses (arrows).

defect lies between two frontal bones and the cephalocele presents anteriorly as a midline mass between the defect. Frontoethmoid cephaloceles emerge outward from the skull  defect at the junction of the frontal and ethmoid bones, immediately anterior to the crista galli. They are subclassified into nasofrontal, nasoethmoidal, and nasoorbital subtypes based on the point at which the skull defect and hernia emerge externally. Various complications can occur depending on the size of the mass. For example, large masses stretch and thin the skin, obscure vision, block the airway, and may also cause pressure deformities of adjacent soft tissue and bone at the forehead, nose, and orbits, causing telecanthus or true bony hypertelorism.163,172

Basal Cephaloceles.  Basal cephaloceles are subdivided into trans­ ethmoid, transsphenoidal, sphenoethmoid, and frontospenoid.54 The most common type is the sphenoethmoid type, and it usually presents as a pharyngeal mass with signs and symptoms of airway obstruction or CSF leak (Fig. 26-14). Other complications that can be seen with basal cephaloceles (transsphenoidal or sphenoethmoidal) include pituitary and hypothalamic dysfunction and decreased visual acuity secondary to herniation of the pituitary gland, hypothalamus, and optic apparatus into the sac.54 Embryogenesis.  There is no universal agreement on the etiology of anterior cephalocele, but it is thought to be secondary to imperfect closure of the anterior neuropore of the neural tube at about the 25th

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient day of gestation. The basal cephaloceles, those related to the sphenoid bone, are thought to be secondary to persistence of the craniopharyn­ geal canal or failure of normal union of basilar ossification centers.54,158 Histology.  In a meningoencephalocele, the herniated neural tissue contains neurons with prominent astrogliosis, and the cyst wall is composed of glial and fibrous connective tissue without neurons.159 Imaging.  CT is useful for imaging the bony margins of the skull base defect along with the soft tissue component of the lesion. MRI is the best imaging modality for evaluating the contents of the cephalo­ cele and in defining the cephalocele in relation to the pituitary stalk and gland as well as the position and state of the optic nerves, chiasm, and optic tracts.54

Nasopharyngeal Teratomas Head and neck nasopharyngeal teratomas are presumed to arise from primitive germ cells that get trapped in the normal migratory channel during development. They contain all three germ layers and arise from embryonic tissue that breaks off or fails to migrate to its normal des­ tination. More than 50% of nasopharyngeal teratomas are seen in the first year of life and often arise from the superior or lateral wall of the nasopharynx. There is a female predilection for teratomas in the nasal cavity, and the typical clinical presentation is a neonate with severe respiratory distress, stridor, and dysphagia. CT and MRI are the best imaging modalities for teratomas, which can be identified by the pres­ ence of fat and and calcification.200 Differential considerations for nasopharyngeal teratomas include dermoids, encephalocele, nasal glioma, rhabdomyosarcoma, hemangioma, lipoma, neurofibroma, and lymphoma.189

MALIGNANT NEOPLASMS

767

and eosinophils. Classic Hodgkin’s lymphoma represents 95% of cases and has been subclassified into nodular sclerosing (60%-80%), mixed cellularity (15%-30%), lymphocyte predominant (5%), and lympho­ cyte depleted (<1%) variants.54 Treatment consists of chemotherapy and radiation for advanced disease, with an overall 15-year survival rate of 68%. The majority of relapses occur within 3 years of treatment.123,211

Non-Hodgkin’s Lymphoma.  About 15% of non-Hodgkin’s lym­ phoma arises in the head and neck, tends to occur in children between the ages of 7 and 11 years, and may involve any structure in the head and neck. There is a 3 : 1 male predominance, and patients may have widespread disease with bone involvement at diagnosis.54 Clinically it presents as a neck mass or masses and can have various clinical presentations depending on the location. For example, extra­ nodal lymphatic disease may occur in the Waldeyer’s ring, leading to middle ear effusion or nasal obstruction, or in cases of Burkitt’s lym­ phoma (extranodal extralymphatic) can present as a large mass involv­ ing the jaw, thyroid, or parotid gland. The most common histologic types of non-Hodgkin’s lymphoma are malignant lymphoma, lymphoblastic (30%); malignant lymphoma, undifferentiated (Burkitt’s and non-Burkitt’s) (40%-50%); and large cell (15%-20%).24,42,104 Leukemia has primarily bone marrow involve­ ment, and non-Hodgkin’s lymphoma has primarily lymph node involvement. If greater than 25% of bone marrow cells are lympho­ blasts, the diagnosis is leukemia. Compared to Hodgkin’s lymphoma, non-Hodgkin’s lymphoma follows a more unpredictable course with chemotherapy and radiation therapy treatments. Prognosis is related to initial tumor stage and response to therapy.54

Imaging Findings.  Imaging is often not helpful in differentiating

Neoplasms of the head and neck comprise approximately 5% of   all malignant neoplasms in the pediatric age group. Although lym­ phoma and rhabdomyosarcoma are the most common tumors, other tumors in this category include primary neuroblastoma, nasopharyn­ geal carcinoma (NPC), thyroid carcinoma, synovial sarcoma, and esthesioneuroblastoma.54,94

Lymphoma Lymphoreticular malignancies account for about 40% of all malignant disorders in children, with lymphoma representing 11% of pediatric malignancies and 50% of pediatric head and neck malignancies.54,213 It is the second most common malignancy in the head and neck in this age group and is equally divided between non-Hodgkin’s and Hodg­ kin’s lymphoma.54,82

Hodgkin’s Lymphoma.  Hodgkin’s lymphoma represents 14% of all cases of lymphoma and has a bimodal age distribution: the first peak in teenagers and young adults and the second peak in adults older than 40 years. The disease is uncommon in children younger than age 5, and there is a male predominance of 3 : 1 in the preteen years. Although  the etiology remains unknown, there is a known association with Epstein-Barr virus in up to 50% of cases. There is also increased inci­ dence of the disease in patients with HIV, which carries a poorer prognosis. The disease is less common among Asian and African American populations.54 The majority of patients present with painless adenopathy, with 80% to 90% involving cervical lymph nodes. Some of the common constitutional symptoms include fever, night sweats, weight loss, anemia, fatigue, and pruritus.210 The tumor consists of malignant Reed-Sternberg cells and their mononuclear variants and normal reactive cells including lymphocytes

Hodgkin’s from non-Hodgkin’s disease or metastatic lymphadenopa­ thy. Tumor deposition sites are subdivided into nodal (most common), extranodal lymphatic (Waldeyer’s ring), and extranodal extralym­ phatic (orbit, paranasal sinuses, deep fascial spaces, mandible, salivary glands, skin, and larynx).79 Hodgkin’s disease typically spreads from one nodal group to the next contiguous nodal group, with extranodal disease being rare (Fig. 26-15). Usual presentation of this disease in children involves cervical and mediastinal adenopathy in a young patient.26,79,118 In non-Hodgkin’s disease, Waldeyer’s ring, paranasal sinuses, and nasal cavity are the most common extranodal sites of disease.118 Both can present as pain­ less neck masses with unilateral or bilateral cervical adenopathy, with involvement of middle and lower jugular nodal chains and relative sparing of upper jugular chains.79,118 The nodal and extranodal masses are of intermediate attenuation on CT, but on MRI demonstrate homogeneous intermediate-signalintensity masses on all pulse sequences.76 On sonography, cervical adenopathy demonstrates multiple areas of decreased echogenicity, simulating cystic masses. Evaluation of patients with head and neck lymphoma should include CT of the chest and abdomen in addition to imaging the neck. For both Hodgkin’s and non-Hodgkin’s lymphoma, gallium-67 scan­ ning has been used for initial staging of tumor, evaluation of response to treatment, and detection of recurrent and metastatic disease. Fluo­ rodeoxyglucose positron emission tomography (FDG PET) is increas­ ingly becoming the imaging modality of choice.67,69,223

Rhabdomyosarcoma Soft tissue sarcomas account for 7% of all malignancies in children younger than age 15, and greater than 50% of these tumors are rhab­ domyosarcomas (RMS). It is primarily a tumor of children and young

http://radiologyebook.com

768

PART II  CT and MR Imaging of the Whole Body

B

A

C

D FIG 26-15  Lymphoma. Coronal postcontrast CT images demonstrate extensive cervical (A) and mediastinal (B) lymphadenopathy (arrows). Sonographic images (C) demonstrate two enlarged lymph nodes with abnormal internal morphology (arrow). PET scan (D) demonstrates metabolically active conglomeration of rightsided cervical lymph nodes (arrow).

adults, with most cases arising in patients younger than 12 years and 43% in patients younger than age 5. Approximately 38% are found in the head and neck.127 RMS is a malignant tumor that arises from primi­ tive pluripotential mesenchymal cells capable of differentiating into skeletal muscle, and as a result these tumors may develop anywhere  in the body. The most common locations for primary head and neck RMS in children are the orbit and nasopharynx, followed by paranasal sinuses and middle ear.28,107,115 There is a bimodal age distribution occurring at 2 to 5 years of age and the late teenage years.54 Different types include embryonal, alveolar, pleomorphic, botryoid, and mixed types. More than half of RMS are embryonal variety, 70% to 90% arise in the head and neck or genitourinary tract, and most are seen in patients younger than 10 years.10 The Intergroup Rhabdomyo­ sarcoma Study divided head and neck RMS into three categories based on site of origin: orbital, parameningeal, (nasopharynx, sinonasal,

middle ear, mastoid, pterygoid fossa), and all other head and neck sites. The paramedian tumors extend through the skull base and invade the meninges. Tumors spread to lymph nodes of the neck in 12% to 50% of cases, and hematogenous spread also occurs to lung, bone, brain, bone marrow, and breast.28,220,228

Presentation.  RMS is a solitary rapidly enlarging firm mass that involves striated muscle. Clinical presentation depends on the ana­ tomic site of the tumor. Tumors in the neck present with a painless enlarging mass, and parameningeal RMS presents with a bloody dis­ charge from the nose or ear, chronic otitis media with or without facial nerve palsy, or symptoms attributed to an upper respiratory infec­ tion.54 Tumor location also makes a difference in prognosis. For example, tumor in the orbit tends to be more localized and has a lower incidence of lymph node metastasis than other supraclavicular tumors.

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

A

B

769

C

FIG 26-16  Rhabdomyosarcoma. A, Axial contrast-enhanced CT image at the level of the nasal cavity demonstrates a large enhancing aggressive soft tissue mass involving the right maxillary sinus and the infratemporal fossa (arrows). Axial (B) and coronal (C) contrast-enhanced T1-weighted MRIs demonstrate an infiltrative mass involving the pharyngeal mucosal space, left infratemporal masticator space, skull base, and sphenoid sinus (arrows).

Imaging Findings.  RMS is seen as a densely cellular-infiltrating neo­ plasm on imaging, and skeletal muscle is always seen as the site of origin. On both CT and MRI, the masses are usually homogeneous, with destruction of adjacent bone (Fig. 26-16). On MRI, the tumor is isointense to muscle on T1-weighted images and hyperintense on T2-weighted images.117 The degree of enhancement is similar to that of normal skeletal muscle, and just over half of cases enhance homogeneously. Meningeal involvement is the worst prognostic indicator in head and neck RMS, so sagittal and coronal postcontrast MRI with fat satu­ ration is crucial for evaluating for dural/meningeal disease and intra­ cranial extension. Coronal CT is recommended for evaluation of bone destruction.16,28

Treatment and Prognosis.  Tumors that are not amenable to good surgical outcome are treated with chemotherapy and radiotherapy. If intracranial extension is detected, then the patient undergoes addi­ tional radiotherapy to the base of the brain and whole brain, along with intrathecal chemotherapy. With the remainder of these tumors, the assumption is that patients have systemic disease at presentation, so they are treated locally and regionally with surgery and radiotherapy as well as chemotherapy.128 Prognosis is related to initial extent of disease.153 Patients with parameningeal disease commonly present with advanced disease and carry a 35% to 64% 2-year disease-free survival rate. This poor survival rate is due to a tendency for invasion of the skull base. Orbital RMS has the best prognosis, with an 80% to 90% 2-year disease-free survival rate.72,84,115

of small cells, sometimes arranged in whirls or rosette-like structures intermingled in larger primitive ganglion-like cells and neuronal tissue. A spectrum of maturational changes from the most immature malignant neuroblastoma to benign ganglioneuroma and bone marrow involvement is common.180

Imaging Findings.  Because these patients are often young infants presenting with a neck mass, ultrasound is often the initial imaging modality of choice. With ultrasound, the findings are often nonspecific and show a hypoechoic or complex mass along with cervical adenopa­ thy. Small echogenic foci consistent with calcifications are sometimes present2 (Fig. 26-17). Approximately 50% of neuroblastomas demon­ strate calcifications on routine images. A sharply circumscribed mass in the poststyloid parapharyngeal space is seen on CT, with calcifica­ tion in up to 90% of cases on noncontrast images. When the mass originates posterior to the vascular sheath, the carotid and jugular vein are displaced anteriorly and demonstrate enhancement to the same degree or slightly greater than muscle.54 MRI is done to demonstrate the full extent of the mass, chest wall invasion, and extradural intra­ spinal involvement. Increased signal intensity to muscle is seen on T1- and T2-weighted MRI.50,64 Because it is more commonly a meta­ static lesion, whole-body imaging should be done in the initial evalu­ ation to rule out other sites of involvement.

Differential Diagnosis and Treatment.  Differential diagnoses

Neuroblastoma is a neural crest tumor arising from the sympathetic nervous system. It is the most common malignant tumor in children younger than age 1. The most frequent sites are the adrenal gland and retroperitoneum.109 Cervical neuroblastoma in infants is rare and accounts for less than 5% of all cases of neuroblastoma. Most that are located in the neck likely represent metastatic disease from an abdomi­ nal or thoracic primary.25,188

are limited to processes that involve the poststyloid parapharyngeal and carotid spaces and primarily include nodal tumors (metastases, lymphoma) and neurogenic tumors (schwannoma, neurofibroma, paraganglioma).54 Treatment is based on multiple factors, including stage of the disease, age of the patient at diagnosis, and other biologically deter­ mined factors.20 Treatment can include surgery, radiotherapy, chemo­ therapy, and bone marrow transplantation. Cervical neuroblastoma is usually localized, and infants younger than 1 year of age have the best overall survival rate (74%-82%). This may be due to early presentation, spontaneous regression, and maturation of the tumor to a benign ganglioneuroma.

Presentation.  Patients often present with a painless, smooth, firm

Nasopharyngeal Carcinoma

mass and symptoms of dysphagia, hoarseness, stridor, and respiratory difficulty.2,25,188 Neuroblastomas are densely cellular tumors composed

NPC is rare in the pediatric population and accounts for less than 1% of pediatric malignancies, with children 10 to 19 years being the most

Neuroblastoma

http://radiologyebook.com

770

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 26-17  Neuroblastoma. A and B, Axial contrast-enhanced CT images at the level of the mandible demonstrate a large enhancing soft tissue mass (B) with calcifications (A) in the left poststyloid carotid space (arrows). Abnormally enlarged right level two cervical node is seen at this level. C, Coronal contrast-enhanced CT image demonstrates a mildly enhancing soft tissue mass in the right lower neck (arrows) resulting in mass effect on the adjacent structures.

A

B

C

FIG 26-18  Nasopharyngeal carcinoma. A, Axial noncontrast CT image at the level of the nasopharynx demonstrates a soft tissue mass in the right posterior nasopharynx extending to the right fossa of Rosenmüller (white arrow). Opacification of the right mastoid air cells is seen (black arrow). Axial pre- (B) and postcontrast (C) T1-weighted MRIs demonstrate a large enhancing right nasopharyngeal mass (arrows).

commonly affected.196 There is increased incidence of this tumor in certain geographic locations and populations (e.g., southern China and Alaska).54 NPC is the most common cancer among Chinese males.90 In the United States, there is a higher incidence in African American and Chinese American populations.53,227

type III, undifferentiated carcinoma. Types 1 and II are usually associ­ ated with EBV, whereas type III is associated with childhood NPC. In the pediatric age group, NPC is usually an undifferentiated epidermoid carcinoma or lymphoepithelioma.61,96,122

Presentation and Classification.  The most common presentation

shows a large inhomogeneous mass filling the nasopharyngeal airway, with skull base erosion and intracranial extension (Fig. 26-18). The tumor usually arises in the posterolateral wall of the nasopharynx in the region of the fossa of Rosenmüller and can obstruct the lateral recess and cause asymmetry of the airway secondary to mass effect. It can also arise from the posterosuperior wall, posterior wall, or anterior wall of the nasopharynx.13 Necrotic cervical adenopathy involving the retropharyngeal and high jugular chain nodes is common. CT is ideal for evaluation of bony erosion of the skull base. The tumors tend to be noncalcified with a variable enhancement pattern.

in childhood is with tender cervical adenopathy. Serous otitis media can be seen secondary to eustachian tube obstruction by the tumor or by dysfunction of muscular action of the tensor veli palatini.54 Nasal obstruction, epistaxis, pain referred to the ear, neck, or throat or cranial nerve paresis are some of the clinical symptoms that may be seen at presentation. Most children have metastatic cervical adenopathy at the time of presentation.43 Multiple cell patterns can be seen with tumors. The World Health Organization proposed three types: type 1, keratinizing squamous cell carcinoma; type II, nonkeratinizing squamous cell carcinoma; and

Imaging Findings.  On CT and MRI, nasopharyngeal carcinoma

http://radiologyebook.com

771

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

3.5

A

B

C

FIG 26-19  Papillary thyroid cancer. A, Axial contrast-enhanced CT image demonstrates a low-attenuation mass in the right lobe of the thyroid gland (arrow). B, Further workup revealed metastatic pulmonary nodules (arrow) on chest CT. Thyroid ultrasound (C) in a different patient demonstrated a heterogeneous mass with calcifications in the left lobe of the thyroid gland (arrows).

Widening of the petroclival fissure with extension to the parapharyn­ geal space has been reported in children with NPC, as has bilateral  and unilateral cervical adenopathy and lateral retropharyngeal nodes greater than 1 cm in maximal transverse dimension. MRI has been found to be superior to CT in the evaluation of primary lesions of the nasopharynx in both adults and children. Undifferentiated carcinomas are invasive tumors with poorly defined margins and associated bony destruction. These masses are hypointense on T1-weighted sequences, hyperintense on T2-weighted signal intensity with inhomogeneous enhancement after contrast administration. Necrotic cervical adenopathy is common, especially in the retropharyngeal and high jugular chains. Primary tumor and meta­ static disease is strongly FDG avid.54 Differential considerations by imaging would include lymphoma, RMS, neuroblastoma, and juvenile angiofibroma (JAF), or inflamma­ tory processes that destroy bone. The most common sites of metastatic disease are the cervical lymph nodes, lungs, and bone. Survey CT of the chest and nuclear bone scanning would be included in the workup. Treatment is focused on local control of disease with radiotherapy, with or without chemotherapy, before metastatic spread occurs. The 5-year relapse-free survival rate is 36%, with an overall 5-year survival rate of 51%.94

Thyroid Carcinoma Thyroid carcinoma is uncommon in the pediatric age group and rep­ resents 1% to 1.5% of all malignancies in children. It more commonly occurs in older children and adolescents. Although thyroid carcinoma tends to be more aggressive in children than adults at the time of diagnosis, the prognosis is generally good.71,230 In the past, about 20% to 50% of children with thyroid carcinoma had a history of radiation exposure.174,232 This was due to frequent use of radiotherapy for benign conditions such as thymic enlargement, tonsillar and adenoidal hypertrophy, hemangioma, and lymphatic malformation, peaking in the 1940s and resulting in an increase in thyroid carcinoma in the 1960s and 1970s.60,214 Even though the use of radiotherapy for such benign conditions has significantly decreased, there are more children surviving cancer with adjuvant radiotherapy. As a result, delayed thyroid carcinoma may become more of a concern in the future because the period between exposure to radiation and occurrence of thyroid carcinoma may be as long as 25 years.232

Presentation.  Thyroid carcinoma is subdivided into differentiated (papillary or follicular), medullary, and anaplastic or nondifferentiated

carcinoma. Papillary carcinoma is the most common type in the pedi­ atric age group, accounting for about 90% of cases (Fig. 26-19). It is a well-differentiated tumor, and squamous metaplasia, psammoma bodies, oxyphilic change, and lymphatic invasion are frequent. Pure follicular adenocarcinoma is the second most common type occurring in children and accounts for 5% to 10% of cases.66,77,222 Medullary carcinoma accounts for about 5% of thyroid neoplasms in children and originates from parafollicular cells, which are derived from neural crest tissue that migrates to the thyroid during embryogenesis. Ana­ plastic carcinoma is very uncommon in children.1 Differentiated carcinomas (papillary and follicular) most com­ monly present with cervical lymphadenopathy or a thyroid nodule. There is a high incidence of cervical metastases (43%-87%) and pul­ monary metastases (up to 15%) at the time of presentation in children with this type of thyroid carcinoma.71,171,174

Imaging Findings.  The imaging workup of a thyroid mass involves ultrasound and radionuclide scanning. A thyroid carcinoma is usually “cold” nonfunctioning but rarely may be “hot.” Sonography can be used to characterize the mass to determine the cystic and solid com­ ponents. Cystic masses are usually benign, representing an adenoma or colloid cyst, but they may be complicated by hemorrhage or debris simulating a solid or mixed tumor. On occasion, malignant thyroid lesions are purely cystic. Most thyroid carcinomas have ill-defined margins and variable echogenicity compared to normal thyoid.54 Microcalcifications, irregular margins, and detectable flow in the mass by Doppler increase the suspicion of malignancy. Echogenic foci within the mass or lymph nodes, which represent calcifications, are characteristic of primary or metastatic medullary carcinoma.71,74 In patients with known differentiated thyroid carci­ noma, MRI should be performed to evaluate for metastatic lymphade­ nopathy, because iodine use may postpone I-131 therapy. Because iodine therapy is not used for medullary thyroid carcinoma, CT is useful for evaluation of metastatic disease.71,74,190 MRI can be helpful in differentiating recurrent thyroid carcinoma from fibrosis. Recurrent carcinoma appears on MRI as hypointense to intermediate T1-T2 signal intensity, and hypointense signal on both T1 and T2 sequences with fibrosis.54

Treatment.  Surgery is the main treatment for primary thyroid carci­ noma. The best survival rates are found with papillary carcinoma (86.3%), follicular (85.3%), medullary (44%), and anaplastic (19%) types; the overall mortality rate for children with thyroid cancer is 17%.66,71,75,174,222

http://radiologyebook.com

772

PART II  CT and MR Imaging of the Whole Body Imaging Findings.  Esthesioneuroblastomas are usually superior

Synovial Sarcoma Synovial sarcoma accounts for about 8% to 10% of all soft tissue malignancies; 85% occur in the extremities. Synovial sarcomas are not usually found in the head and neck and are thought to arise from pluripotential mesenchymal tissue that can differentiate into synovi­ oblastic cells.54 They usually arise in or near tendon or tendon sheaths rather than joints or bursal surfaces.11,52,95 The most common locations for synovial sarcoma in the head and neck are cervical soft tissues and parapharyngeal sites.

Presentation.  Patients usually present between the ages of 15 and 35, and the most common clinical presentation is a painless mass with symptoms of dysphagia, hoarseness, and headache. Histologically they are divided into biphasic and monophasic types. The biphasic type contains epithelial and spindle cell components, and the monophasic contains only spindle cells.85

Imaging Findings.  CT is useful for demonstrating calcification, present in the form of round concretions, and for evaluating erosive changes of bone associated with synovial sarcomas.54 MRI is con­ sidered the imaging modality of choice for detecting and staging  this tumor, because tumor extent, vascular invasion, and intratu­ moral hemorrhage are better assessed with MRI.184 They have vari­ able heterogeneous MRI signal on T1- and T2-weighted sequences and demonstrate variable heterogeneous enhancement after contrast administration.184 The most common imaging appearance is a well-defined mass with an irregular wall, necrotic center, and minimal wall enhancement on CT. It may be solid, cystic, or mixed in composition. Because 12.5% of patients have metastases to cervical lymph nodes, CT or MRI of the neck should be included in the initial evaluation, along with CT of the chest because of pulmonary metastases.169

Treatment.  Treatment is wide local excision, and radiation and che­ motherapy may be used as adjunctive treatment if the excision is not complete.136 The rate of local recurrence is high, and metastasis to the lungs usually occurs within 2 years, with low survival rates of 40% at 5 years and 25% at 10 years.95,169,182

Esthesioneuroblastoma Esthesioneuroblastoma is a neurogenic tumor related to the olfactory apparatus. It is a neural crest tumor originating from olfactory epithe­ lium of the superior nasal cavity or cribriform plate and represents 2% to 3% of all malignant nasal tumors.32,142,178 Olfactory epithelium can also be found in the middle turbinates and adjacent paranasal sinuses. It is common to see spread of tumor to paranasal sinuses, anterior cranial fossa, olfactory bulbs, orbits, and cavernous sinus.56

Presentation.  Although it occurs in all age groups, a bimodal age peak has been noted in the second and sixth decades.93 The most common clinical presentations are epistaxis related to the vasculature of the tumor and unilateral nasal obstruction.47,173 Other presenting symptoms, although less common, include anosmia, sinus pain, and diploplia.47 The diagnosis of this tumor depends on demonstration of neural elements by histochemical or biochemical techniques using electron microscopy. The presence of intercellular fibrillary background or Homer-Wright rosettes is diagnostic for esthesioneuroblastoma.31,134,203 This tumor can be confused with other small cell malignancies such as undifferentiated squamous cell carcinoma or lymphoma.212

nasal cavity lesions demonstrating both remodeling and destructive bony growth properties.173 On CT they appear as homogenous mod­ erately enhancing soft tissue masses centered in the superior nasal cavity, frequently with extension into the ethmoid and maxillary sinuses, orbit, and anterior cranial fossa.92,212 On MRI, the imaging characteristics are nonspecific. The mass can be isointense to hypoin­ tense on T1-weighted and isointense to hyperintense on T2-weighted images, with variable enhancement.47,173 Cysts are sometimes seen along the intracranial margin of the mass, and these findings have been reported to be highly suggestive of esthesioneuroblastoma. With exten­ sion to the anterior nasal cavity, the mass has a dumbbell appearance, with the waist at the level of the cribriform plate. They can also be completely intranasal in location with superior location in the nasal cavity.39 Destructive bony changes are usually best evaluated with coronal CT, but findings can also be detected with MRI.173 Calcification has also been reported as highly suggestive of esthesioneuroblastoma, but it is often difficult to distinguish true calcification from secondary bony destruction from the tumor.32,47,39 Coronal postcontrast MRIs are best for evaluation of intracranial extension of disease and extension to orbits or paranasal sinuses (Fig. 26-20). MRI is very useful in dif­ ferentiation of nonenhancing obstructive sinus disease from enhanc­ ing tumor or recurrent disease. Sinus disease is usually hyperintense on T2-weighted images and demonstrates peripheral enhancement. Tumor or recurrent disease is usually intermediate on T2-weighted sequences, with solid enhancement.39,47,212 Primary differential consid­ erations for this tumor include non-Hodgkin’s lymphoma, nasal meningioma, invasive aspergilloma and polyps, JAF, and squamous cell and undifferentiated carcinoma.28,32,47,173

Prognosis and Treatment.  There are staging and grading systems by Kadish and Hyams that are important in determining the progno­ sis.39,49 Hematogenous metastasis to cervical lymph nodes occurs in 17% of patients at presentation, so if cervical lymphadenopathy is present, then CT or MRI of the neck should be performed. Widespread disease to lungs, bone, and liver occurs in 14% to 20% of patients.54 Treatment is determined by stage of disease at presentation; the majority of patients undergo surgical resection and adjuvant radio­ therapy.1,48 Disease-free survival and overall survival rates are 77% and 61% at 5 years and 53% and 42% at 10 years.124

BENIGN MASSES Juvenile Angiofibroma JAF is the most commonly encountered vascular mass of the nasal cavity and comprises 0.5% of head and neck tumors in children.12 The tumor originates in the nasal cavity, with secondary spread to the nasopharynx. It is a benign but locally invasive tumor; these lesions are almost exclusively found in adolescent males and usually present between 7 and 21 years of age, with a peak incidence between 14 and 18 years.12,44 The majority of JAFs arise from the fibrovascular stroma at the margin of the sphenopalatine foramen in the posterior nasal cavity and extend into the nasopharynx, pterygopalatine fossa, and more anteriorly into the nasal cavity.12 Rare extranasopharyngeal masses have also been reported, which tend to occur more commonly in older age and in women.81

Presentation.  Patients usually present with complaints of unilateral nasal obstruction or severe spontaneous epistaxis. Other signs and symptoms include purulent nasal rhinorrhea from associated sinus infection, nasal speech, anosmia, facial deformity, serous otitis media,

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

A

B

773

C

FIG 26-20  Esthesioneuroblastoma. Sagittal (A) and axial (B) precontrast T1-weighted and coronal (C) postcontrast T1-weighted MRIs demonstrate an avidly enhancing mass in the superior nasal cavity, with intracranial extension through the cribriform plate (arrows).

hearing loss, exophthalmos, and sometimes signs of intracranial mass effect from extension of the tumor. Episodes of recurrent epistaxis have been related to vascular fragility of the tumor, thought to be secondary to increased levels of circulating estrogens.12,114 JAFs are unencapsulated masses composed of spindle or stellate fibrocytes in a connective tissue stroma with many wide thin-walled vessels. The myofibroblast represents the principal cell of tumor and can interfere with synthesis of collagen and elastin and is capable of transforming into smooth muscle.114 Fine vessels can invade the fibro­ vascular stroma and produce pseudolymphatic endothelialized spaces. The final appearance represents a spectrum from a cellular or purely fibrous mass to a cavernous-like mass with a fibrovascular stroma. The incidence of local invasion and frequency of epistaxis are decreased if the fibrous component of the tumor predominates.44

Imaging Findings.  JAFs are slow-growing masses tending to extend along and through fissures and foramina, remodeling rather than destroying bone.87 Imaging gives important information for both diag­ nosis and management of these tumors. Imaging findings include nasopharyngeal soft tissue mass, widen­ ing of the pterygopalatine fossa, anterior displacement of the posterior maxillary sinus wall with widening of the pterygomaxillary fissure, opacification of one or more paranasal sinuses, erosion of the sphenoid bone with a mass in the sphenoid sinus, erosion of the hard palate or medial wall of the maxillary sinus, deviation of the nasal septum, and intracranial extension (Fig. 26-21). JAFs can slowly enlarge over time and may involve the paranasal sinuses (ethmoid, maxillary, and sphe­ noid) and orbits; 5% to 20% of these tumors can demonstrate skull base erosion.176 The tumor can be evaluated with CT, MRI, and angiography. CT is useful for evaluation of bony involvement, and MRI is better for eval­ uation of determining intracranial extension of tumor and for differ­ entiating tumor extension into sinuses from postobstructive sinus disease. Coronal and axial planes should be performed with and without contrast. Involvement and extent of invasion of the sphenoid sinus is best evaluated on CT, which is the main determinant for recurrence. On MRI, JAF is isointense to hypointense to muscle on T1-weighted images and isointense to hyperintense to muscle on T2-weighted images.74,121 Fat saturation should be used with postcontrast

MRIs. Flow voids are frequently seen, and homogeneous intense enhancement is usually seen on both CT and MRI after contrast administration.44,176,177 MR angiography can show enlargement of external carotid and internal maxillary arteries, and angiography is useful for defining the vascular supply of the tumor for preoperative embolization. JAF is supplied by branches of the external carotid artery, commonly the internal maxillary or ascending pharyngeal artery, and numerous hypertrophied and tortuous vessels are seen, without segmental nar­ rowing or aneurysm formation.54 A persistent dense blush is present throughout the capillary and venous phase, and early-draining veins and AV shunting are usually absent.113,168 To fully define the vascular supply of the tumor, bilateral internal and external carotid artery injec­ tions along with selective injections of the ipsilateral internal maxillary, ascending pharyngeal, and ascending palatine arteries should be performed.113,114 The primary differential consideration from an imaging standpoint is an angiomatous polyp. Patients can present with similar symptoms, although these polyps usually occur at a slightly older age and do not invade the pterygopalatine fossa or intracranial cavity. There is only rare extension of the mass into the sphenoid sinus, and the enhance­ ment pattern is less intense than JAF and flow voids are not present on MRI.192 Other differential considerations include lymphoma, RMS, NPC, and esthesioneuroblastoma.11,192

Treatment and Prognosis.  A staging classification was developed by Sessions and colleagues, based on the extent of the lesion, to assist in selecting the best surgical approach. Total resection within 24 hours of embolization is the most common approach to therapy, which is per­ formed to decrease intraoperative blood loss. Different surgical approaches that can be used include open, endoscopic, or laser-assisted surgery.54 Radiotherapy can be used for inoperative disease or residual disease and has been used at some institutions for patients with advanced JAF.164 Intracranial extension does not preclude surgery, but parasellar disease can make en bloc resection very difficult, so these patients usually undergo debulking with extensive beam radiotherapy. Hormonal therapy and cryotherapy are reported as adjunctive treat­ ments to surgery. The long-term prognosis for this tumor is good, with a 25% to  60% recurrence rate with surgery and 25% recurrence rate with

http://radiologyebook.com

774

PART II  CT and MR Imaging of the Whole Body

A

B

D

E

C

FIG 26-21  Juvenile angiofibroma. Axial noncontrast (A) and coronal contrast-enhanced (B) CT images demonstrate a large mass filling the right nasal cavity and nasopharynx, with medial extension into the maxillary sinus and posterolaterally into the pterygomaxillary fissure (arrows). More superiorly the mass also demonstrates intracranial extension to the right middle cranial fossa (arrows) (C). Axial T2-weighted (D) MRI demonstrates flow voids (arrow), and preembolization external carotid injection (E) shows a hypervascular mass (arrows) being supplied by branches of the external carotid artery.

radiotherapy alone. It is common to have recurrence within first 12 months of therapy and unusual after 24 months. There is increased predisposition to recurrence owing to the multilobulated nature of the tumor, leading to invasion of adjacent bones, sinuses, and fascial planes.54

Neurogenic Tumors Tumors of neurogenic origin comprise a small but significant group of head and neck masses, with neurofibromas and schwannomas being the most common nerve sheath tumors of the peripheral nerves in the head and neck.13,37

Presentation.  Solitary neurofibromas and schwannomas can be seen as sporadic lesions, and plexiform and multiple neurofibromas occur in association with neurofibromatosis type 111,65,102 (Fig. 26-22). About 25% of solitary neurogenic tumors originate along the course of a cervical or cranial nerve, with the exception of olfactory and optic nerves.65,165 From 25% to 45% of schwannomas occur in the head and neck, and the most common locations include the face, scalp, orbit, oral cavity, parapharyngeal space, middle ear, larynx, and neck.54 The normal nerve fiber is ensheathed by Schwann cells and sur­ rounded by endoneurial fibroblasts; the Schwann cell is thought of as the precursor of the neurofibroma, schwannoma, and likely neuro­ genic sarcoma.208 Neurofibromas are encapsulated nerve sheath tumors composed of Schwann cells with collagen in a mucous matrix. Solitary neurofibromas are subcutaneous and well defined, and plexiform

neurofibromas are infiltrative with ill-defined margins. From 5% to 13% of plexiform neurofibromas undergo malignant degeneration.11 Schwannomas are encapsulated tumors attached to or surrounded by a nerve. They are composed of two types of tissues: cellular Antoni type A and less cellular Antoni type B tissue. They are almost never associated with neurofibromatosis type 1 and rarely undergo malig­ nant change.11 Neurofibromas can be seen at any age, but schwannomas most commonly present between 20 and 50 years of age, with occasional occurrences in childhood. Neurofibromas are usually asymptomatic, whereas schwannomas are often painful and tender. The pain may be localized or radicular in nature and associated with paresthesias.11

Imaging Findings.  The imaging appearance of peripheral nerve sheath tumors depends on their histologic composition, and as a result neurofibromas and schwannomas may be indistinguishable by imaging. Solitary neurofibromas and schwannomas present as well-defined lobulated masses and tend to cross fascial planes. They can appear as isointense masses to muscle, with rare calcification and variable enhancement on CT. Some 73% of solitary or plexiform neurofibro­ mas demonstrate relatively decreased attenuation before and after  contrast administration on CT. Schwannomas enhance more homoge­ neously with either iodinated contrast or gadolinium. On MRI, neu­ rofibromas are hypointense to isointense to muscle on T1-weighted and hyperintense to muscle on T2-weighted images. Variable enhance­ ment is seen after contrast administration.

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

A

775

B

FIG 26-22  Nerve sheath tumors. A, Coronal postcontrast T1-weighted MRI demonstrates an enhancing infiltrative plexiform neurofibroma in the left lower neck (arrows). B, Sagittal postcontrast T1-weighted MRI demonstrates an enhancing well-circumscribed schwannoma in the lower neck (arrow).

Schwannomas have intermediate T1-weighted signal and hyperin­ tense T2-weighted signal to muscle, with uniform enhancement after contrast administration. Heterogeneous signal intensity can be seen centrally in the tumors owing to cystic degeneration and hemorrhage, and is seen more commonly in schwannomas. The characteristic finding of neurofibroma is the target sign. On MRI, an area of hypointense central signal with peripheral hyperin­ tense signal is seen on T2-weighted sequences, with central enhance­ ment of the lesion present on CT. Loss of the central hypointense signal, large size, infiltrating margins, rapid growth, central necrosis, and prominent enhancement pattern are imaging findings concerning for malignant degeneration.65,193 Plexiform neurofibromas are poorly defined infiltrative masses that can erode bone and expand neural foramina of the spine. They have a complex appearance on MRI, with wavy cords of hypointense signal surrounded by areas of mixed isointensity and hyperintensity on T2 images. STIR sequences may be helpful in evaluation of patients with multiple or plexiform neurofibromas where the masses demonstrate diffuse hyperintense signal.65,193

VASCULAR AND LYMPHATIC MALFORMATIONS Hemangioma of Infancy and Vascular Malformation There has been widespread use of different terminology regarding vascular and lymphatic malformations of the head and neck in chil­ dren. Mulliken and Glowacki proposed a classification system of   hemangiomas and vascular malformations based on cellular turnover, histology, natural history, and physical findings. The two major catego­ ries are hemangioma of infancy and vascular malformation.140 Hemangioma of infancy (Fig. 26-23) has a predictable biological behavior. The masses usually appear in the first month of life (1%-2%

of neonates), increase in size secondary to endothelial cell proliferation during the first year of life (proliferating phase), slowly involute from age 1 to age 7 (involuting phase), and then completely regress by age 8 to 12 (involuted phase).6,15 Vascular malformations are due to abnormal development of blood or lymphatic vessels. The lesions grow, have normal endothelial mitotic activity, and fail to involute.140 Hemangiomas of infancy are obvious in the first month of life, whereas vascular malformations do not become apparent until late infancy or childhood even though lesions are present at birth. They can be more complex in composition, with multiple vascular elements. They can rapidly increase in size, as seen with infection, trauma, or endocrine changes (pregnancy, puberty).

Classification.  Vascular malformations are categorized based on the dominant component of the malformation: capillary (Fig. 26-24), venous, lymphatic, arterial (with or without fistulas), combined (capil­ lary venous, lymphaticovenous, arteriovenous [AVM]).140 Hemangioma of infancy and vascular malformation can also  be characterized according to flow characteristics as follows: AVMs  and hemangiomas—high-flow lesions; venous and lymphatic malformations—low-flow lesions; combined malformation—low- or high-flow lesions.6,100,132

Imaging Findings.  Hemangioma of infancy is a well-defined mass of isodense attenuation to muscle on CT and is isointense on T1-weighted and hyperintense on T2-weighted signal to muscle on MRI.6 The flow voids within the lesion on MRI are a secondary finding that is helpful in distinguishing this mass from other soft tissue lesions in infancy (see Fig. 26-23).132 Postcontrast images on CT and MRI show a diffuse intense homo­ geneous enhancement pattern, and these masses progressively enlarge

http://radiologyebook.com

776

PART II  CT and MR Imaging of the Whole Body

A

C

B

D FIG 26-23  Hemangioma. A, Axial T2-weighted MRI demonstrates a T2 bright lesion with flow voids in the right parotid gland (arrows). B, Ultrasound of the left parotid gland demonstrates asymmetric enlargement with extensive vascularity (arrows). Axial T2-weighted (C) and postcontrast T1-weighted (D) MRI in another patient demonstrate an avidly enhancing lesion with flow voids in the left masseter muscle (arrows).

FIG 26-24  Capillary malformation. Axial contrast-enhanced T1-weighted MRI demonstrates a low-flow, partially enhancing soft tissue mass (arrows) in the soft tissues of the left side of the face at the level of the submandibular gland.

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

777

FIG 26-25  Venous malformation. Axial T2-weighted MRI demonstrates a hyperintense to muscle mass in the left masseter muscle with a small hypointense area, most consistent with a phlebolith (arrows).

over a period of several months before spontaneously involuting. Areas of hyperintense T1-weighted signal on MRI can sometimes be seen owing to secondary fatty replacement.6,132 Venous malformations are well-defined predominantly solid soft tissue masses associated with bony remodeling or deformity. The masses tend to cross fascial planes and do not spontaneously involute. They have characteristics of low-flow lesions and lack intralesional flow voids (Fig. 26-25). On MRI, intermediate T1-weighted signal and heterogeneous hyperintense T2-weighted signal to muscle with homo­ geneous or mildly heterogeneous enhancement after contrast admin­ istration is seen.132 Heterogeneous signal due to intralesional phleboliths (hypointense on T1- and T2-weighted) and venous lakes (hyperintense on T2-weighted signal) may be found within the lesion, and heteroge­ neous contrast enhancement may also be present if the lesion is a mixed venolymphatic malformation.6,132 AVMs do not usually cause mass effect, owing to lack of an underly­ ing stroma within a lesion. On MRI, serpiginous flow voids on T1-   and T2-weighted sequences and decreased T1-weighted signal due  to intraosseous extension with marrow involvement can be seen.  The signal voids correlate with AV shunting present on angiography. Invasive combined vascular malformations are seen as infiltrative  solid masses with serpiginous flow voids, and intermediate T1-weighted and hyperintense T2-weighted signal is seen with mixed lesions  (Fig. 26-26). Lymphatic malformations are vascular malformations composed of primitive embryonic lymph sacs of different sizes separated by connec­ tive tissue. Although lymphatic malformations can be diagnosed pre­ natally, most of them do not present until after birth and typically grow along with growth of the child. They do not spontaneously involute and may show rapid increase in association with upper respiratory tract infection or intralesional hemorrhage. They can be characterized as microcystic, macrocystic, or mixed. Large cystic spaces of a lym­ phatic malformation are hypoechoic on ultrasound and low attenua­ tion on CT and on MRI; they follow water signal, with T1 hypointense and T2 hyperintense signal intensity relative to muscle125 (Fig. 26-27). They usually lack significant contrast enhancement, and internal septa­ tions may enhance if there is an associated venous malformation. They may be focal and well defined, unilocular, multilocular, or infiltrative. They can be confined to one space in the head and neck or cross mul­

FIG 26-26  Arteriovenous malformation. Axial T1-weighted MRI demonstrates a large soft tissue mass with serpiginous flow voids (arrows) in the floor of the mouth.

tiple spaces of the neck (transspatial). Infection or hemorrhage into the mass would alter the imaging characteristics, and changes are dependent on the stage of breakdown of blood products. On angiography, hemangiomas of infancy are distinguished from vascular malformations by the presence of a well-defined mass showing intense tissue staining in an organized lobular pattern. The findings are variable for vascular malformation and depend on the predomi­ nant vascular channel type. They are typically diffuse lesions composed of vessels without intervening tissue stain. Although hemangiomas of infancy are thought to be high-flow lesions, they usually spontaneously involute; however, lesions occasion­ ally require systemic or intralesional steroids, alpha-interferon, laser therapy, or surgery.54 Venous and lymphatic malformations can be treated with sclerotherapy. For complicated AVM cases, selective arte­ riography is indicated because they may be candidates for embolization along with surgical intervention. Combined vascular malformations may also be highly invasive and resistant to all forms of therapy.100

Rapidly Involuting Congenital Hemangioma, Noninvoluting Congenital Hemangioma, and Infantile Hemangioma Rapidly involuting congenital hemangioma (RICH) and noninvolut­ ing congenital hemangioma (NICH) are uncommon entities. The infantile hemangioma and these two conditions have overlapping clinical and pathologic features. With RICH, the full size of the tumor is present at birth and regresses by 12 to 14 months of age. NICH is present at birth and grows in proportion to the patient and does not involute.54 They have similarities in appearance, location, and size and similar imaging characteristics on MRI. RICH demonstrates areas of inhomogeneous signal and larger flow voids. Arterial aneurysms, along with large irregular feeding vessels, direct arteriovenous shunting, and intravascular thrombi, can be seen on angiography.51,106

PHACE PHACE is an acronym given to a rare neurocutaneous syndrome  of unknown etiology. Findings include posterior fossa malformation

http://radiologyebook.com

778

PART II  CT and MR Imaging of the Whole Body

C

B

A

FIG 26-27  Venolymphatic malformation. A, Axial T2-weighted MRI demonstrates a large T2 hyperintense lesion on the left, with extension into multiple facial and neck spaces (arrows). B, Sagittal T1-weighted MRI in another patient demonstrates a large lymphatic malformation, with hyperintense material likely representing blood or protein contents (arrows). C, Coronal contrast-enhanced image demonstrates a partially enhancing transspatial lesion (arrows) along the right lower face.

A

B FIG 26-28  Fibromatosis. Pre- (A) and postcontrast (B) sagittal T1-weighted MRIs demonstrate abnormal enlargement and enhancement (arrows) of muscle in the floor of the mouth.

(Dandy-Walker malformation), hemangiomas (large facial hemangio­ mas), arterial malformations, coarctation of the aorta and other cardiac anomalies, and eye abnormalities.36,68

categorized into superficial and deep lesions. The two deep lesions that can occur in the neck in children are aggressive infantile fibromatosis (Fig. 26-28) and fibromatosis colli166 (Fig. 26-29).

MASSLIKE CONDITIONS

Fibromatosis Colli.  Also referred to as muscular or congenital tor-

Musculoskeletal Fibromatoses These are a diverse group of lesions that occur at different ages and locations but share a common histopathologic appearance. They are

ticollis and SCM tumor or pseudotumor of infancy, fibromatosis colli is a rare benign form of infantile fibromatosis and is related to enlargement of the SCM muscle. The two main theories regarding the etiology of this condition include one related to birth trauma and the

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

779

B

A

FIG 26-29  Fibromatosis coli. Sonographic image of the right sternocleidomastoid muscle (A) demonstrates fusiform enlargement (arrows) of the muscle compared to the contralateral side (B).

other one related to fetal intrauterine malpositioning with torticollis. The first theory attributes birth trauma with intramuscular hema­ toma and subsequent fibrosis.91,111,166 The second theory postulates the condition of fetal intrauterine malpositioning with torticollis causing an SCM compartment syndrome. With the second theory, on histo­ logic examination there is fibrocollagenous infiltration without hem­ orrhage.130,207 Greater than 90% have a history of difficult or forceps delivery.166 Presentation.  Infants present at 2 to 4 weeks of age with torticollis (14%-20% of patients) and a firm, unilateral, palpable neck mass that conforms to the shape of the SCM.54 The most commonly affected portions are the middle to lower third, but masses can occur anywhere along the muscle. Though the mass may grow for days to a few weeks, the natural history is regression of the process by 4 to 8 months of age in 80% of infants. The condition is slightly more common in males, with a right-sided predominance and rarely bilateral cases. Some of the reported secondary findings include delayed torticollis, plagiocephaly, and facial asymmetry.32,142 Imaging findings.  The primary role of imaging is to exclude other causes of a mass in the muscle. Ultrasound is the imaging modality of choice, and the imaging patterns in this condition depend on the his­ tologic composition, especially the fibrous content in the SCM muscle. On ultrasound, the findings include focal or diffuse enlargement of   the SCM, with variable internal echogenicity compared to normal muscle.1,57,108,226 A peripheral rim of decreased echogenicity has been reported in up to 90% of cases and is thought to represent subjacent normal muscle.1 CT and MRI are used for cases where the diagnosis is not confirmed by ultrasound. On CT the findings include a focally or diffusely enlarged SCM, with the enlarged area appearing isodense to muscle. With contrast administration, the mass demonstrates a variable enhancement pattern. On MRI the enlarged mass is isointense to slightly hyperintense to muscle on T2-weighted images and isointense to hypointense on T1-weighted images. The lesions with a greater fibrous content have hypointense T2 signal, and diffuse contrast

enhancement can be seen after contrast administration.54 Secondary inflammatory changes are usually absent. Radiographs are usually normal, but there have been reported cases of lytic lesions at the  clavicular attachment site of the muscle.166 Treatment.  Primary and initial treatment is with physical therapy, with a 70% to 85% success rate. Despite therapy, contraction of the fibrous tissue in the SCM leads to muscular torticollis and ipsilateral head tilt in 20% of cases. When conservative treatment is not success­ ful, a muscle release, muscle lengthening, or tenotomy procedure may be performed.32

Aggressive Infantile Fibromatosis.  Aggressive fibromatosis is a rare condition representing 0.03% to 0.1% of all tumors and an inci­ dence of 2 to 4 cases per million population per year.54 There is a wide age range, and most cases are found in the third and fourth decades. Head and neck lesions represent about 11% of extraabdominal cases, and the remainder occur in the abdomen, abdominal wall, or other extraabdominal locations.54 Presentation.  Aggressive infantile fibromatosis represents the childhood counterpart of extraabdominal desmoid tumor, also known as musculoaponeurotic fibromatosis. Their biological behavior lies between benign fibrous lesions and fibrosarcoma, but they are benign histologically and do not metastasize. They can be large at presenta­ tion, grow rapidly and be locally aggressive, and invade or encase neurovascular or bony structures. In children the head and neck are the most common locations for this tumor, with a male predominance in children.119,166,232 Patients usually present in the first 2 years of life with a firm, nodular, painless soft tissue mass that is usually within the skeletal muscle, fascia, or periosteum.54 The condition is characterized by pro­ liferation of fibrous tissue surrounded and separated by collagen and demonstrates a locally aggressive behavior and a tendency toward recurrence after surgery.166 Imaging findings.  MRI is the best imaging modality for evaluation of the condition, and the masses are usually intermuscular in location

http://radiologyebook.com

780

PART II  CT and MR Imaging of the Whole Body

and follow fascial planes. The lesion demonstrates well-defined or infiltrative margins, and invasion into the muscle is common.54 Signal intensity depends on the histologic composition of the cellular tissue, myxoid tissue, and collagen, and the pattern is highly variable. For example, some lesions have a hypointense signal on all pulse sequences, reflecting a high collagen composition. Usually the pattern is more heterogeneous in more cellular masses with less collagen, with T1-weighted signal isointense to slightly hyperintense to muscle and T2-weighted signal similar to or greater than fat. Most cases demon­ strate moderate to marked enhancement on MRI, but approximately 10% of lesions lack significant enhancement.54 CT can be useful for evaluation of secondary bony involvement. Extraabdominal desmoid tumors can have a similar imaging appearance and are indistinguishable but tend to present in an older age group. Treatment and prognosis.  Local recurrence is common after sur­ gical resection, owing to the aggressive lesion growth and locally inva­ sive nature of the mass.54 Radiotherapy has been useful both as a postoperative adjuvant to surgery and a primary treatment; despite the high rate of recurrence, the overall prognosis is good.166,231,232

Adenopathy/Adenitis of the Neck Cervical adenopathy is a very common finding in children, and in the vast majority of pediatric patients, imaging is not required. The  adenopathy is usually related to a developing immune system respond­ ing to exposure to various multiple new antigens, and the adenopathy is reactive rather than metastatic.54

Presentation.  Cervical adenopathy can be divided by location into deep (retropharyngeal or parapharyngeal) and lateral or superficial (anterior/posterior cervical and submandibular). With inflammatory processes in the neck, the child usually presents with a painful and tender neck mass with cervical lymph nodal masses (Fig. 26-30). Potential etiologies include upper respiratory infection, underlying immune disorder, foreign body ingestion, upper aerodigestive or odontogenic source, and osteomyelitis of the spine.54

A

Imaging Findings.  For atypical clinical presentation or cases that are refractory to medical treatment or if abscess is suspected, imaging may be requested.54 It is important to distinguish between cellulitis and drainable fluid collection, evaluate for evidence of complications in the form of vascular compromise or airway compression, and determine the presence of associated osteomyelitis.54 It is important to distinguish between cellulitis and drainable abscess, because the former is usually treated medially and the latter may need surgical intervention.55 Although ultrasound is a noninvasive method to evaluate for the presence of an abscess and has a reported sensitivity of 90% to 95%, it lacks specificity in distinguishing suppurative masses from infected congenital cystic masses.110,116,129 CT is most commonly used in attempting to make this differentia­ tion and also helps differentiate cellulitis from an abscess. Cellulitis appears as an ill-defined area of decreased attenuation with mild diffuse enhancement and infiltration of the surrounding soft tissue. A drainable abscess is seen as a single or multiloculated area of decreased attenuation, with peripheral rim enhancement of the wall conforming to the fascial spaces.54 Secondary findings that can be seen include myositis, infiltration with stranding of surrounding fat, skin thicken­ ing, and engorgement of veins and lymphatics.150,225 Distinguishing an early abscess from cellulitis is often difficult on imaging if there is lack of a complete circumferential ring of enhancement, which is the  most sensitive imaging finding of an abscess. Because similar imaging appearance can also be seen with necrotic nodal disease, infected bran­ chial cleft cyst, necrotic neoplasm, and thrombosed vessels, pertinent clinical history, physical findings, and laboratory data are important factors.54 Cat-scratch disease and nontuberculous mycobacteria (espe­ cially nodes with areas of central necrosis) should be considered in children with significant cervical adenopathy without associated sur­ rounding or adjacent inflammatory changes, because bacterial adenitis is usually associated with inflammatory changes4,55,197 (Fig. 26-31). In cases of recurrent deep neck infection, other etiologies must be con­ sidered, including complicated underlying congenital lesion such as a branchial apparatus anomaly, lymphatic malformation, cervical thymic cyst, or a TDC.86,149,153

B FIG 26-30  Cervical adenopathy. A, Contrast-enhanced CT image demonstrates enlarged left cervical lymph nodes (arrows). B, Contrast-enhanced CT image demonstrates a large enlarged left cervical node with low attenuation (arrows).

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient

781

B

A

C FIG 26-31  Unilateral cervical adenopathy: cat-scratch disease and non-TB Mycobacterium. A, Patient with history of exposure to cats presented with left-sided unilateral cervical adenopathy (arrows). B, Another patient with unilateral cervical adenopathy demonstrated calcified left cervical lymph nodes (arrows) on bone (B) and soft tissue windows (C).

REFERENCES 1. Ablin DS, Jain K, Howell L, et al: Ultrasound and MR imaging of fibromatosis colli (sternomastoid tumor of infancy). Pediatr Radiol 28:230–233, 1998. 2. Abramson SJ, Berfon WE, Ruzal-Shapiro C, et al: Cervical neuroblastoma in eleven infants—A tumor with favorable prognosis. Clinical and radiologic (US, CT, MRI) findings. Pediatr Radiol 23:253–257, 1993. 3. Allard RH: The thyroglossal cyst. Head Neck Surg 5:134–146, 1982. 4. Bagla S, Tunkel D, Kraut MA: Nontuberculous mycobacterial lymphadenitis of the head and neck: Radiologic observations and clinical context. Pediatr Radiol 33:402–406, 2003. 5. Bailey H: Branchial cysts and other essays on surgical subjects in the faciocervical region, London, 1929, Lewis. 6. Baker LL, Dillon WP, Hieshima GB, et al: Hemangiomas and vascular malformations of the head and neck. MR characterization. AJNR Am J Neuroradiol 14:307–314, 1993. 7. Bale PM, Painter DM, Cohen D: Teratomas in childhood. Pathology 7:209–218, 1975. 8. Barat M, Sciubba JJ, Abramson AL: Cervical thymic cyst: Case report and review of literature. Laryngoscope 95:89–91, 1985. 9. Barkovich AJ, Vandermarck P, Edwards MSB, et al: Congenital nasal masses: CT and MR imaging features in 16 cases. AJNR Am J Neuroradiol 12:105–116, 1991. 10. Barnes L, Verbin R, Gnepp D: Diseases of the nose, paranasal sinuses and nasopharynx. In Barnes L, editor: Surgical pathology of the head and neck (vol 1). New York, 1985, Marcel Dekker, pp 725–880.

11. Batsakis JG: Teratomas of the head and neck. In Batsakis JG, editor: Tumors of the head and neck: Clinical and pathological considerations, Baltimore, 1979, Williams and Wilkins. 12. Batsakis JG: Vasoformative tumors. In Batsakis JG, editor: Tumors of the head and neck: Clinical and pathological considerations, Baltimore, 1979, Williams & Wilkins. 13. Batsakis JG, Solomon AR, Rice DH: The pathology of head and neck tumors: Carcinoma of the nasopharynx. Head Neck Surg 3:511–524, 1981. 14. Benson MT, Dalen K, Mancuso AA, et al: Congenital anomalies of the branchial apparatus: Embryology and pathologic anatomy. Radiographics 12:943–960, 1992. 15. Berenguer B, Mulliken JB, Enjolras O, et al: Rapidly involuting congenital hemangioma: Clinical and histopathologic features. Pediatr Dev Pathol 6:495–510, 2003. 16. Berry MP, Jenkin RD: Parameningeal rhabdomyosarcoma in the young. Cancer 48:281–288, 1981. 17. Blumenfeld R, Skolnik EM: Intranasal encephaloceles. Arch Otolaryngol 82:527–531, 1965. 18. Bradley PJ: Nasal dermoids in children. Int J Pediatr Otorhinolaryngol 3:63–70, 1981. 19. Braun M, Bman F, Hascoet JM, et al: Brain tissue heterotopia in the nasopharynx: Contribution of MRI to assessment of extension. J Neuroradiol 19:68–74, 1992. 20. Breslow N, McCann B: Statistical estimation of prognosis for children with neuroblastoma. Cancer Res 31:2098–2103, 1971. 21. Brownstein MH, Shapiro L, Slevin R: Fistula of the dorsum of the nose. Arch Dermatol 109:227–229, 1974.

http://radiologyebook.com

782

PART II  CT and MR Imaging of the Whole Body

22. Burton EM, Mercado-Deane MG, Howell CG, et al: Cervical   thymic cysts: CT appearance of two cases including a persistent thymopharyngeal duct cyst. Pediatr Radiol 25:363–365, 1995. 23. Butler EC, Dickey JR, Shill OS, Jr, et al: Carcinoma of the thyroglossal duct remnant. Laryngoscope 79:264–271, 1969. 24. Callihan TR, Berard CW: Childhood non-Hodgkin’s lymphomas in current histologic perspective. In Rosenberg HS, Berstein J, editors: Perspectives in pediatric pathology, New York, 1982, Raven Press. 25. Casselman JW, Smet MH, Van Damme B, et al: Primary cervical neuroblastoma: CT and MR findings. J Comput Assist Tomogr 12:684–686, 1988. 26. Castellino RA: Hodgkin disease: Practical concepts for the diagnostic radiologist. Radiology 159:305–310, 1986. 27. Castillo M, Mukherji S: The Neck. Imaging of the pediatric head, neck, and spine, Philadelphia, 1996, Lippincott-Raven, 15: pp 563–577. 28. Castillo M, Pillsbury HC, 3rd: Rhabdomyosarcoma of the middle ear: Imaging features in two children. AJNR Am J Neuroradiol 14:730–733, 1993. 29. Chandler J, Mithchell B: Branchial cleft cysts, sinuses, and fistulae. Otolaryngol Clin North Am 14:175–186, 1981. 30. Charoonsmith T, Suwanwela C: Frontoethmoidal encephalomeningocele with special reference to plastic reconstruction. Clin Plast Surg 1:27–47, 1974. 31. Chaudhry AP, Haar JG, Koul A, et al: Olfactory neuroblastoma (esthesioneuroblastoma): A light and ultrastructural study of two cases. Cancer 44:564–579, 1979. 32. Cheng JC, Tang SP, Chen TM: Sternocleidomastoid pseudotumor and congenital muscular torticollis in infants: A prospective study of 510 cases. J Pediatr 134:712–716, 1999. 33. Choudhury AR, Bandey SA, Haleem A, et al: Glial heterotopias of the nose. A report of two cases. Childs Nerv Syst 12:43–47, 1996. 34. Choudhury AR, Taylor JC: Primary intranasal encephalocele: report of four cases. J Neurosurg 57:552–555, 1982. 35. Choy FJ, Ward R, Richardson R: Carcinoma of the thyroglossal duct. Am J Surg 108:361–369, 1964. 36. Coats DK, Paysse EA, Levy ML: PHACE: A neurocutaneous syndrome with important ophthalmologic implications: Case report and literature review. Ophthalmology 106:1739–1741, 1999. 37. Coffin CM, Dehner LP: Peripheral neurogenic tumors of the soft tissues in children and adolescents: A clinicopathologic study of 139 cases. Pediatr Pathol 9:387–407, 1989. 38. Cohen AF, Mistudo S, Ruben RJ: Nasopharyngeal teratoma in the neonate. Int J Pediatr Otorhinolaryngol 14(2–3):187–195, 1987. 39. Constantinidis J, Steinhart H, Koch M, et al: Olfactory neuroblastoma: The University of Erlangen-Nuremberg experience 1975-2000. Otolaryngol Head Neck Surg 130:567–574, 2004. 40. Cote DN, Gianoli GJ: Fourth branchial cleft cyst. Otolaryngol Head Neck Surg 114:95–97, 1996. 41. Cressman WR, Myer CM, 3rd: Pathologic quiz case 1. Cervical thymic cyst. Arch Otolaryngol Head Neck Surg 118:772–774, 1992. 42. Crist WM, Kelly DR, Ragab AH, et al: Predictive ability of LukesCollins classification for immunologic phenotypes of childhood non-Hodgkin lymphoma: An institutional series and literature review. Cancer 48:2070–2075, 1981. 43. Cunningham MJ, Mcguirt WF, Myers EN: Malignant tumors of the head and neck. In Bluestone CD, Stool SE, Kenna MA, editors: Pediatric otolarngology, ed 3, (vol 2). Philadelphia, 1996, WB Saunders, pp 1557–1583. 44. Davis KR: Embolization of epistaxis and juvenile nasopharyngeal angiofibroma. AJR Am J Roentgenol 148:209–218, 1987. 45. Day LH, Arnold GE: Rare tumors of the ear, nose and throat: Second series: Uncommon benign tumors of the head and neck. Laryngoscope 81:1138–1174, 1971. 46. Deane SA, Telander RL: Surgery for thyroglossal duct and branchial cleft anomalies. Am J Surg 136:348–353, 1978. 47. Dedeyn CP, Moran CJ, Wippold FJ, 2nd, et al: MRI of esthesioneuroblastoma. J Comput Assist Tomogr 18:16–21, 1994.

48. Devaiah AK, Larsen C, Tawfik O, et al: Esthesioneuroblastoma: Endoscopic nasal and anterior craniotomy resection. Laryngoscope 113:2086–2090, 2003. 49. Dias FL, Sa GM, Lima RA, et al: Patterns of failure and outcome in esthesioneuroblastoma. Arch Otolaryngol Head Neck Surg 129:1186– 1192, 2003. 50. Dietrich RB, Kangarloo H: Retroperitoneal mass with intradural extension: Value of magnetic resonance imaging in neuroblastoma. AJR Am J Roentgenol 146:251–254, 1986. 51. Donegan J: Congenital neck masses. In Cummings C, Frederick J, Harker L, et al, editors: Otolaryngology head and neck surgery, St. Louis, 1993, Mosby-Year Book, pp 1554–1565. 52. Drugman ME, Rosin HD, Toker C: Synovial sarcoma of the head and neck. Arch Otolaryngol 98:53–54, 1973. 53. Easton JM, Levine PH, Hyams VJ: Nasopharyngeal carcinoma in the United States. A pathologic study of 177 US and 30 foreign cases. Arch Otolaryngol 106:88–91, 1981. 54. Egelhoff JC, Kock BL: Pediatric Head and Neck Imaging. In Haaga J, editor: CT and MRI of the whole body, Philadelphia, 2008, Mosby. 55. Elden LM, Grundfast KM, Vezina G: Accuracy and usefulness of radiographic assessment of cervical neck infections in children. J Otolaryngol 30:82–89, 2001. 56. Elkon D, Hightower SI, Lim ML, et al: Esthesioneuroblastoma. Cancer 44:1087–1094, 1979. 57. Ellis HA: Cervical thymic cysts. Br J Surg 54:17–20, 1967. 58. Ellis P, Van Nostranel A: The applied anatomy of the thyroglossal tract. Laryngoscope 8:765–770, 1977. 59. Faerber EN, Swartz JD: Imaging of neck masses in infants and children. Crit Rev Diagn Imaging 31(3–4):283–314, 1991. 60. Favus MJ, Schneider AB, Stachura ME, et al: Thyroid cancer occurring as a late consequence of head and neck irradiation. Evaluation of 1056 patients. N Engl J Med 294:1019–1025, 1976. 61. Fedder M, Gonzalez MF: Nasopharyngeal carcinoma. Brief review.   Am J Med 79:365–369, 1985. 62. Finerman WB, Pick EI: Intranasal encephalomeningocele. Ann Otol Rhinol Laryngol 62:114–120, 1953. 63. Finn D, Buchalter I, Srti E, et al: First branchial cleft cysts: Clinical update. Laryngoscope 97:136–140, 1987. 64. Fletcher BD, Kopiwoda SY, Strandjord SE, et al: Abdominal neuroblastoma: Magnetic resonance imaging and tissue characterization. Radiology 155:699–703, 1985. 65. Fortman BJ, Kuszyk BS, Urban BA, et al: Neurofibromatosis type 1: A diagnostic mimicker at CT. Radiographics 21:601–612, 2001. 66. Frankenthaler RA, Sellin RV, Cangir A, et al: Lymph node metastasis from papillary-follicular thyroid carcinoma in young patients. Am J Surg 160:341–343, 1990. 67. Friedberg JW, Chengazi V: PET scans in the staging of lymphoma: Current status. Oncologist 8:438–447, 2003. 68. Frieden IJ, Reese V, Cohen D: PHACE syndrome. The association of posterior fossa brain malformations, hemangiomas, arterial anomalies, coarctation of the aorta and cardiac defects, and eye abnormalities. Arch Dermatol 132:307–311, 1996. 69. Front D, Ben-Haim S, Israel O, et al: Lymphoma: Predictive value   of Ga-67 scintigraphy after treatment. Radiology 182:359–363, 1992. 70. Gainsford JC, Anderson US: First branchial cleft cysts and sinuses. Plast Reconstr Surg 55:299–304, 1975. 71. Garcia CJ, Daneman A, McHugh K, et al: Sonography in thyroid carcinoma in children. Br J Radiol 65(779):977–982, 1992. 72. Ginsberg LE: Neoplastic diseases affecting the central skull base: CT an MR imaging. AJR Am J Roentgenol 159:581–589, 1992. 73. Gorenstein A, Kern EB, Facer GW, et al: Nasal gliomas. Arch Otolaryngol 106:536–540, 1980. 74. Gorman B, Charboneau JW, James EM, et al: Medullary thyroid carcinoma: Role of high-resolution US. Radiology 162(1 Pt 1):147–150, 1987. 75. Gundry SR, Wesley JR, Klein MD, et al: Cervical teratomas in the newborn. J Pediatr Surg 18:382–386, 1983.

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient 76. Han MH, Chang KH, Kim IO, et al: Non-Hodgkin lymphoma of the central skull base: MR manifestations. J Comput Assist Tomogr 17:567–571, 1993. 77. Harnes JK, Thompson NW, McLeod MK, et al: Differentiated thyroid carcinoma in children and adolescents. World J Surg 16:547–553, 1992. 78. Harnsberger HR: Handbook of head and neck imaging, St. Louis, 1995, Mosby. 79. Harnsberger HR, Bragg DG, Osborn AG, et al: Non-Hodgkin’s lymphoma of the head and neck: CT evaluation of nodal and extranodal sites. AJR Am J Roentgenol 149:785–791, 1987. 80. Harnsberger HR, Mancuso AA, Muraki AS, et al: Branchial cleft anomalies and their mimics: Computed tomographic evaluation. Radiology 152:739–748, 1984. 81. Haung RY, Damrose EJ, Blackwell KE, et al: Extranasopharyngeal angiofibroma. Int J Pediatr Otorhinolaryngol 56:59–64, 2000. 82. Healy GB: Malignant tumors of the head and neck in children: Diagnosis and treatment. Otolaryngol Clin North Am 13:483–488, 1980. 83. Hedlund G: Congenital frontonasal masses: Developmental anatomy, malformations, and MR imaging. Pediatr Radiol 367:647–662, 2006. 84. Herrmann BW, Sotelo-Avila C, Eisenbeis JF: Pediatric sinonasal rhabdomyosarcoma: Three cases and a review of the literature. Am J Otolaryngol 24:174–180, 2003. 85. Hirsch RJ, Yousem DM, Loevner LA, et al: Synovial sarcomas of the head and neck: MR findings. AJR Am J Roentgenol 169:1185–1188, 1997. 86. Hollinshead WH: Anatomy for surgeons. The head and neck (vol 1). New York, 1968, Harper and Row. 87. Holman CB, Miller WE: Juvenile nasopharyngeal fibroma. AJR Am J Roentgenol 94:292, 1965. 88. Hoving EW: Nasal encephaloceles. Childs Nerv Syst 16:702–706, 2000. 89. Howell CJ: The sublingual dermoid cyst. Report of five cases and review of the literature. Oral Surg Oral Med Oral Pathol 59:578–580, 1985. 90. Hsu MM, Haung SC, Lynn TC, et al: The survival of patients with nasopharyngeal carcinoma. Otolaryngol Head Neck Surg 90(3 Pt 1):289–295, 1982. 91. Hulbert KF: Congenital torticollis. J Bone Joint Surg Br 32:50, 1950. 92. Hurst RW, Erickson S, Cail WS, et al: Computed tomographic features of esthesioneuroblastoma. Neuroradiology 31:253–257, 1989. 93. Hyams VJ, Batsakis JG, Michaels L: Tumors of the upper respiratory tract and ear. Fascicle 25. In Atlas of Tumor Pathology, Washington, DC, Armed Forces Institute of Pathology, 1988. 94. Jaffe BF, Jaffe N: Head and neck tumors in children. Pediatrics 51:731–740, 1973. 95. Jajdu SI, Shiu MH, Fortner JG: Tendosynovial sarcoma: A clinicopathological study of 136 cases. Cancer 39:1201–1217, 1977. 96. Jenkin RD, Anderson JR, Jereb B, et al: Nasopharyngeal carcinoma—A retrospective review of patients less than thirty years of age: A report of Children’s Cancer Study Group. Cancer 47:360–366, 1981. 97. Johnson GF, Weisman PA: Radiological features of dermoid cysts of the nose. Radiology 82:1016–1021, 1964. 98. Jones JE, Hession B: Cervical thymic cysts. Ear Nose Throat J 75:678– 680, 1996. 99. Jordan RB, Gauderer MW: Cervical teratomas: An analysis. Literature review and proposed classification. J Pediatr Surg 23:583–591, 1988. 100. Kaban LB, Mulliken JB: Vascular anomalies of the maxillofacial region. J Oral Maxillofac Surg 44:203–213, 1986. 101. Kapadia SB, Popek EJ, Barnes L: Pediatric otorhinolaryngic pathology: Diagnosis of selected lesions. Pathol Annu 29(Pt1):159–209, 1994. 102. Katz AD, Passy V, Kaplan L: Neurogenous neoplasms of major nerves of face and neck. Arch Surg 103:51–56, 1971. 103. Kennedy EM, Gruber DP, Billmire DA, et al: Transpalatal approach for the extracranial surgical repair of transsphenoidal cephaloceles in children. J Neurosurg 87:677–681, 1997. 104. Kiyosue H, Miyake H, Komatsu E, et al: MRI of cervical masses of thymic origin. J Comput Assist Tomogr 18:206–208, 1994. 105. Koeller KK, Alamo L, Adair CF, et al: Congenital cystic masses of the neck: Radiologic-pathologic correlation. Radiographics 19:121–146, 1999.

783

106. Konez O, Burrows PE, Mulliken JB, et al: Angiographic features of rapidly involuting congenital hemangioma (RICH). Pediatr Radiol 33:15–19, 2003. 107. Koscielniak E, Morgan M, Treuner J: Soft tissue sarcoma in children. Prognosis and management. Paediatr Drugs 4:21–28, 2002. 108. Kraus R, Han BK, Babcock DS, et al: Sonography of neck masses in children. AJR Am J Roentgenol 146:609–613, 1986. 109. Kretschmar CS, Colbach C, Bhan I, et al: Desmoplastic small cell tumor: A report of three cases and a review of the literature. J Pediatr Hematol Oncol 18:293–298, 1996. 110. Kreutzer EW, Jafek BW, Johnson ML, et al: Ultrasonography in the preoperative evaluation of neck abscesses. Head Neck Surg 4:290–295, 1982. 111. Kurtycz DF, Logrono R, Hoerl HD, et al: Diagnosis of fibromatosis colli by fine-needle aspiration. Diagn Cytopathol 23:338–342, 2000. 112. Lack EE: Extragonadal germ cell tumors of the head and neck region: Review of 16 cases. Hum Pathol 16:56–64, 1985. 113. Lasjaunias P, Marsot-Dupuch K, Doyon D: The radio-anatomical   basis of arterial embolization for epistaxis. J Neuroradiol 6:45–53, 1973. 114. Lasjaunias P, Picard L, Manelfe C, et al: Angiofibroma of the nasopharynx. A review of 53 cases treated by embolization. The role of pretherapeutic angiography. Pathophysiological hypotheses. J Neuroradiol 7:73–95, 1980. 115. Latack JT, Hutchinson RJ, Heyn RM: Imaging of rhabdomyosarcomas of the head and neck. AJNR Am J Neuroradiol 8(2):353–359, 1987. 116. Latifi HR, Siegel MJ: Color Doppler flow imaging of pediatric soft tissue masses. J Ultrasound Med 13:165–169, 1994. 117. Lee J, Lee M, Lee B, et al: Rhabdomyosarcoma of the head and neck in adults: MR and CT findings. AJNR Am J Neuroradiol 17:1923–1928, 1996. 118. Lee YY, Van Tassel P, Nauert C, et al: Lymphomas of the head and neck: CT findings at initial presentation. AJR Am J Roentgenol 149:575–581, 1987. 119. Lewin JS, Lavertu P: Aggressive fibromatosis of the prevertebral and retropharyngeal spaces: MR and CT characteristics. AJNR Am J Neuroradiol 16(4 Suppl):897–900, 1995. 120. Lim-Dunham JE, Feinstein KA, Yousefzadeh DK, et al: Sonographic demonstration of a normal thyroid gland excludes ectopic thyroid in patients with thyroglossal duct cyst. AJR Am J Roentgenol 164:1489– 1491, 1995. 121. Lloyd GA, Phelps PD: Juvenile angiofibroma: imaging by magnetic resonance, CT and conventional techniques. Clin Otolaryngol Allied Sci 11:247–259, 1986. 122. Lombardi F, Gasparini M, Gianni C, et al: Nasopharyngeal carcinoma in childhood. Med Pediatr Oncol 10:243–250, 1982. 123. Lukes RJ, Butler J, Hicks EB: The prognosis of Hodgkin’s disease according to the histologic type and the clinical stage. Role of the reactions of the host. Nouv Rev Fr Hematol 6:15–22, 1966. 124. Lund VJ, Howard D, Wei W, et al: Olfactory neuroblastoma: Past, present, and future? Laryngoscope 113:502–507, 2003. 125. Mancuso AA, Dillon WP: The neck. Radiol Clin North Am 27:407–434, 1989. 126. Maran AGD, Buchanan DR: Branchial cysts, sinuses, and fistulae. Clin Otolaryngol 3:77–92, 1978. 127. Maurer HM: The Intergroup Rhabdomyosarcoma Study: Update, November 1978. Natl Cancer Inst Monogr 56:61–68, 1981. 128. Maurer HM, Moon T, Donaldson M, et al: The intergroup rhabdomyosarcoma study: A preliminary report. Cancer 40:2015–2026, 1977. 129. McCurdy JA, Jr, Nadalo LA, Yim DW: Evaluation of extrathyroid masses of the head and neck with gray scale ultrasound. Arch Otolaryngol 106:83–87, 1980. 130. McDonald D: Sternomastoid tumour and muscular torticollis. J Bone Joint Surg 51:432–443, 1969. 131. McQuown SA, Smith JD, Gallo AE, Jr: Intracranial extension of nasal dermoids. Neurosurgery 12:531–535, 1983.

http://radiologyebook.com

784

PART II  CT and MR Imaging of the Whole Body

132. Meyer JS, Hoffer FA, Barnes PD, et al: Biological classification of soft-tissue vascular anomalies: MR correlation. AJR Am J Roentgenol 157:559–564, 1991. 133. Michael AS, Mafee MF, Valvassori EG, et al: Dynamic computed tomography of the head and neck: Differential diagnostic value. Radiology 154:413–419, 1985. 134. Micheau C: A new histochemical approach to olfactory esthesioneuroma. A nasal tumor of neural crest origin. Cancer 40:314–318, 1977. 135. Miller MD, Rao VM, Tom BM: Cystic masses of the head and neck: Pitfalls in CT and MR interpretation. AJR Am J Roentgenol 159:601– 607, 1992. 136. Miser AW, Miser JS: The treatment of cancer pain in children. Pediatr Clin North Am 36:979–999, 1989. 137. Morgan DW, Evans JNG: Developmental nasal anomalies. J Laryngol Otol 104:394–403, 1990. 138. Muhlbauer WD, Dittmar W: Hereditary median dermoid cysts of the nose. Br J Plast Surg 29:334–340, 1976. 139. Mukherji S, Tart R, Slattery W, et al: Evaluation of first branchial anomalies by CT and MR. J Comput Assist Tomogr 17:576–581, 1993. 140. Mulliken JB, Glowacki J: Hemangiomas and vascular malformation in infants and children: A classification based on endothelial characteristics. Plast Reconstr Surg 69:412–422, 1982. 141. Naidich TP, Altman NR, Braffman BH, et al: Cephaloceles and related malformations. AJNR Am J Neuroradiol 13:655–690, 1992. 142. Nakagawa T, Takashima CP, Snow JB, Jr: Structure of human fetal   and adult olfactory neuroepithelium. Arch Otolaryngol 110:641–646, 1984. 143. Nakagawa T, Takashima T, Tomiyama K: Differential diagnosis of a lateral cervical cyst and solitary cystic lymph node metastasis of occult thyroid papillary carcinoma. J Laryngol Otol 115:240–242, 2001. 144. New G: Congenital cysts of the tongue, the floor of the mouth, the pharynx, and the larynx. Arch Otolaryngol 45:145–158, 1947. 145. New G, Eric J: Dermoid cysts of the head and neck. Surg Gynecol Obstet 65:48–55, 1937. 146. Nguyen Q, deTar M, Wells W, et al: Cervical thymic cyst: Case reports and review of the literature. Laryngoscope 106(3 Pt 1):247–252, 1996. 147. Nicolai P, Luzzago F, Maroldi R, et al: Nasopharyngeal cysts. Report of seven cases with review of the literature. Arch Otolaryngol Head Neck Surg 115:860–864, 1989. 148. Noyek AM, Friedberg J: Thyroglossal duct and ectopic thyroid disorders. Otolaryngol Clin North Am 14:187–201, 1981. 149. Nusbaum AO, Som PM, Rothschild MA, et al: Recurrence of a deep neck infection: A clinical indication of an underlying congenital lesion. Arch Otolaryngol Head Neck Surg 125:1379–1382, 1999. 150. Nyberg DA, Jeffrey RB, Brant-Zawadzki M, et al: Computed tomography of cervical infections. J Comput Assist Tomogr 9:288–296, 1985. 151. Okuda Y, Oi S: Nasal dermal sinus and dermoid cyst with intrafalcial extension. Case report and review of literature. Childs Nerv Syst 3:40–43, 1987. 152. Olsen KD, Maragos NE, Weiland LH: first branchial cleft anomalies. Laryngoscope 90:423–436, 1980. 153. Paonessa DF, Goldstein JC: Anatomy and physiology of head and neck infections (with emphasis on the fascia of the face and neck). Otolaryngol Clin North Am 9:561–580, 1976. 154. Park YW: Evaluation of neck masses in children. Am Fam Physician 51:1904–1912, 1995. 155. Parker GD, Harnsberger HR, Smoker WR: The anterior and posterior cervical spaces. Semin Ultrasound CT MR 12:257–273, 1991. 156. Pensler JM, Bauer BS, Naidich TP: Craniofacial dermoids. Plast Reconstr Surg 82:953–958, 1988. 157. Pinelli V, Pierro V, Bottero S, et al: Dysontogenetic neoplasms of the thyroid gland in infancy: Two case reports. Int J Pediatr Otorhinolaryngol 10:101–110, 1985. 158. Pollock JA, Newton TH, Hoyt WF: Transsphenoidal and transethmoidal encephaloceles. A review of clinical and roentgen features in 8 cases. Radiology 90:442–453, 1968.

159. Pounds LA: Neck masses of congenital origin. Pediatr Clin North Am 28:841–844, 1981. 160. Puppala B, Mangurten HH, McFadden J, et al: Nasal glioma   presenting as neonatal respiratory distress. Clin Pediatr (Phila) 29:49–52, 1990. 161. Rahbar R, Resto VA, Robson CD, et al: Nasal glioma and   encephalocele: Diagnosis and management. Laryngoscope 113:2069–2077, 2003. 162. Rahbar R, Shah P, Mulliken JB, et al: The presentation and management of nasal dermoid: A 30-year experience. Arch Otolaryngol Head Neck Surg 129:464–471, 2003. 163. Rappoport RL, II, Dunn RC, Alhady F: Anterior encephalocele.   J Neurosurg 54:213, 1981. 164. Reddy KA, Mendenhall WM, Amdur RJ, et al: Long-term results of radiation therapy for juvenile nasopharyngeal angiofibroma. Am J Otolaryngol 22:172–175, 2001. 165. Rice DH, Coulthard SW: Neurogenic tumors of the head and neck in children. Ann Plast Surg 2:441–444, 1979. 166. Robbin MR, Murphey MD, Temple HT, et al: Imaging of musculoskeletal fibromatosis. Radiographics 21:585–600, 2001. 167. Rohrich RJ, Lowe JB, Schwartz MR: The role of open rhinoplasty in the management of nasal dermoid cysts. Plast Reconstr Surg 104:1459–1471, 1999. 168. Rosen L, Hanafee W, Nahum A: Nasopharyngeal angiofibroma, an angiographic evaluation. Radiology 86:103–107, 1966. 169. Roth JA, Enzinger FM, Tannenbaum M: Synovial sarcoma of the neck: A followup study of 24 cases. Cancer 35:1243–1253, 1975. 170. Rothschild MA, Catalano P, Urken M, et al: Evaluation and management of congenital cervical teratoma. Case report and review. Arch Otolaryngol Head Neck Surg 120:444–448, 1994. 171. Samuel AM, Sharma SM: Differentiated thyroid carcinomas in children and adolescents. Cancer 67:2186–2190, 1991. 172. Sargent LA, Seyfer AE, Gunby EN: Nasal encephaloceles: Definitive one-stage reconstruction. J Neurosurg 68:571–575, 1988. 173. Schuster JJ, Phillips CD, Levine PA: MR of esthesioneuroblastoma (olfactory neuroblastoma) and appearance after craniofacial resection. AJNR Am J Neuroradiol 15:1169–1177, 1994. 174. Segal K, Sidi J, Levy R, et al: Thyroid carcinoma in children and adolescents. Ann Otol Rhinol Laryngol 94(4 Pt 1):346–349, 1985. 175. Sessions RB: Nasal dermal sinuses-new concepts and explanations. Laryngoscope 92(8 Pt 2, Suppl 29):1–28, 1982. 176. Sessions RB, Bryan RN, Naclerio RM, et al: Radiographic staging of juvenile angiofibroma. Head Neck Surg 3:279–283, 1981. 177. Sessions RB, Wills PI, Alford BR, et al: Juvenile nasopharyngeal angiofibroma. Radiographic aspects. Laryngoscope 86:2–18, 1976. 178. Shah JP, Feghali J: Esthesioneuroblastoma. Am J Surg 142:456–458, 1981. 179. Shidara K, Uruma T, Yasuoka Y, et al: Two cases of nasopharyngeal branchial cyst. J Laryngol Otol 107:453–455, 1993. 180. Shimada H, Chatten J, Newton WA, Jr, et al: Histopathologic prognostic factors in neuroblastic tumors: Definition of subtypes of ganglioneuroblastoma and an age-linked classification of neuroblastomas. J Natl Cancer Inst 73:405–416, 1984. 181. Shin JH, Lee HK, Kim SY, et al: Parapharyngeal second branchial cyst manifesting as cranial nerve palsies: MR findings. AJNR Am J Neuroradiol 22:510–512, 2001. 182. Shmookler BM, Enzinger FM, Brannon RB: Orofacial synovial sarcoma: A clinicopathologic study of 11 new cases and review of the literature. Cancer 50:269–276, 1982. 183. Sichel JY, Halperin D, Dano I, et al: Clinical update on type II first branchial cleft cysts. Laryngoscope 108:1524–1527, 1998. 184. Sigal R, Chancelier MD, Luboinski B, et al: Synovial sarcomas of the head and neck: CT and MR findings. AJNR Am J Neuroradiol 13:1459–1462, 1992. 185. Simpson DA, David DJ, White J: Cephaloceles: Treatment, outcome and antenatal diagnosis. Neurosurgery 15:14–26, 1984. 186. Sistrunk W: The surgical treatment of cysts of the thyroglossal tract. Ann Surg 71:121–122, 1920.

http://radiologyebook.com

CHAPTER 26  Imaging of the Head and Neck in the Pediatric Patient 187. Smith KR, Jr, Schwartz HG, Luse SA, et al: Nasal gliomas: A report of five cases with electron microscopy of one. J Neurosurg 20:968–982, 1963. 188. Smith MC, Smith RJ, Bailey CM: Primary cervical neuroblastoma in infants. J Laryngol Otol 99:209–214, 1985. 189. Snow JB, Jr: Neoplasms of the nasopharynx in children. Otolaryngol Clin North Am 10:11–24, 1977. 190. Solbiati L, Cioffi V, Ballarati E: Ultrasonography of the neck. Radiol Clin North Am 30:941–954, 1992. 191. Som PM: Cystic lesions of the neck. Postgrad Radiol 7:211–236, 1987. 192. Som PM, Cohen BA, Sacher M, et al: The angiomatous polyp and the angiofibroma: Two different lesions. Radiology 144:329–334, 1982. 193. Som PM, Curtin HD: Head and neck imaging, Chicago, 1996, Mosby-Year Book. 194. Som PM, Sacher M, Lanzieri CF, et al: Parenchymal cysts of the lower neck. Radiology 157:399–406, 1985. 195. Som PM, Smoker WRK, Curtin HD, et al: Congenital lesions. Head and neck imaging, ed 4, 35: 1828–1864, 2003, Mosby, St. Louis. 196. Stambuk HE, Patel SG, Mosier KM, et al: Nasopharyngeal carcinoma: Recognizing the radiographic features in children. AJNR Am J Neuroradiol 26:1575–1579, 2005. 197. Stone ME, Link DT, Egelhoff JC, et al: A new role for computed tomography in the diagnosis and treatment of pyriform sinus fistula. Am J Otolaryngol 21:323–325, 2000. 198. Sutow WW: Cancer of the head and neck in children. JAMA 190:414– 416, 1964. 199. Suwanwela C, Hongsaprabhas C: Frontoethmoidal encephalomeningocele. J Neurosurg 25:172–182, 1966. 200. Takhtani D, Melhem E, Carson B: A Heterotopic cerebellum presenting as a suprasellar mass with associated nasopharyngeal teratoma. AJNR Am J Neuroradiol 21:1119–1121, 2000. 201. Tanimura F, Tachibana M, Dejima K, et al: Teratoma of the soft palate with multifocal cyst formation in the tongue and the brain. Arch Otorhinolaryngol 243:320–323, 1986. 202. Tapper D, Lack EE: Teratomas in infancy and childhood. Ann Surg 94:1489–1492, 1984. 203. Taxy JB, Hidvegi DF: Olfactory neuroblastoma: An ultrastructural study. Cancer 39:131–138, 1977. 204. Telander RL, Deane SA: Thyroglossal and branchial cleft cysts and sinuses. Surg Clin North Am 57:779–791, 1977. 205. Telander RL, Filston HC: Review of head and neck lesions in infancy and childhood. Surg Clin North Am 72:1429–1447, 1992. 206. Thomas JR: Thyroglossal-duct cysts. Ear Nose Throat J 58:510–514, 1979. 207. Tom LW, Rossiter JL, Sutton LN, et al: Torticollis in children. Otolaryngol Head Neck Surg 105:1–5, 1991. 208. Toriumi DM, Atiyah RA, Murad T, et al: Extracranial neurogenic tumors of the head and neck. Otolaryngol Clin North Am 19:609–617, 1986. 209. Trigilia JM, Nicollas R, Ducroz V, et al: First branchial cleft anomalies: A study of 39 cases and a review of the literature. Arch Otolaryngol Head Neck Surg 124:291–295, 1998. 210. Ultmann JE, Cunningham JK, Gellhorn A: The clinical picture of Hodgkin’s disease. Cancer Res 26:1047–1062, 1966. 211. Urquhart A, Berg R: Hodgkin’s and non-Hodgkin’s lymphoma of the head and neck. Laryngoscope 111:1565–1569, 2001.

785

212. Vanhoenacker P, Hermans R, Sneyers W, et al: Atypical aesthesioneuroblastoma: CT and MRI findings. Neuroradiology 35:466–467, 1993. 213. Vazquez E, Lucaya J, Castellote A, et al: Neuroimaging in pediatric leukemia and lymphoma: Differential diagnosis. Radiographics 22:6, 2002. 214. Viswanathan K, Gierlowski TC, Schnider AB: Childhood thyroid cancer. Characteristics and long-term outcome in children irradiat3ed for benign conditions of the head and neck. Arch Pediatr Adolesc Med 148:260–265, 1994. 215. Vogl TJ, Steger W, Ihrler S, et al: Cystic masses in the floor of the mouth: Value of MR imaging in planning surgery. AJR Am J Roentgenol 161:183–186, 1993. 216. Wadsworth DT, Siegel MJ: Thyroglossal duct cysts: Variability of sonographic findings. AJR Am J Roentgenol 163:1475–1477, 1994. 217. Ward G, Hendrick J, Chamber R: Thyroglossal tract abnormalities, cysts, and fistula. Surg Gynecol Obstet 89:727–734, 1949. 218. Wardinsky TD, Pagon RA, Kropp RJ, et al: Nasal dermoid sinus cysts: Association with intracranial extension and multiple malformations. Cleft Palate Craniofac J 28:87–95, 1991. 219. Weber PC, Rueger RG, Pickeral J: Thymic cysts. Am J Otolaryngol 17:64–66, 1996. 220. Wexler LH, Helman LJ: Pediatric soft tissue sarcomas. CA Cancer J Clin 44:211–247, 1994. 221. Wiatrak BJ, Myer CM, 3rd, Bratcher GO: Report of a nasopharyngeal teratoma evaluated with magnetic resonance imaging. Otolaryngol Head Neck Surg 102:186–190, 1990. 222. Winship T, Rosvoll RF: Thyroid carcinoma in childhood: Final report on a 20 year study. Clin Proc Children’s Hosp Wash 26:327, 1970. 223. Wirth A, Seymour JF, Hicks RJ, et al: Fluorine-18 fluorodeoxy-glucose positron emission tomography, gallium-67 scintigraphy, and conventional staging for Hodgkin’s disease and non-Hodgkin’s lymphoma. Am J Med 112:262–268, 2002. 224. Work WP: Newer concepts of first branchial cleft defects. Laryngoscope 82:1581–1593, 1972. 225. Yeoh GPS, Bale PM, de Silva M: Nasal cerebral heterotopia: The so-called nasal glioma or sequestered encephalocele and its variants. Pediatr Pathol 9:531–549, 1989. 226. Youkilis RA, Koch B, Myer CM, 3rd: Ultrasonographic imaging of sternocleidomastoid tumor of infancy. Ann Otol Rhinol Laryngol 104(4 Pt 1):323–325, 1995. 227. Young JL, Jr, Miller RW: Incidence of malignant tumors in US children. J Pediatr 86:254–258, 1975. 228. Yousem DM, Lexa FJ, Bilaniuk LT, et al: Rhabdomyosarcomas in the head and neck: MR imaging evaluation. Radiology 177:683–686, 1990. 229. Zerris VA, Annino D, Heilman CB: Nasofrontal dermoid sinus cyst: Report of two cases. Neurosurgery 51:811–814, 2002. 230. Zimmerman D, Hay ID, Gough IR, et al: Papillary thyroid carcinoma in children and adults: Long-term followup of 1039 patients conservatively treated at one institution during three decades. Surgery 104:1157–1166, 1988. 231. Zlotecki RA, Scarborough MT, Morris CG, et al: External beam radiotherapy for primary and adjuvant management of aggressive fibromatosis. Int J Radiat Oncol Biol Phys 54:177–181, 2002. 232. Zohar Y, Strauss M, Laurian N: Adolescent versus adult thyroid carcinoma. Laryngoscope 96:555–559, 1986.

http://radiologyebook.com

27  Noninfectious Inflammatory Diseases Affecting the Spinal Cord Kenneth Lury

INTRODUCTION Most noninfectious inflammatory diseases of the spinal cord are autoimmune in nature.6,10 Although the number of autoimmune diseases affecting the spinal cord is large, their clinical imaging manifestations are limited, and with few exceptions the most common presentation is transverse myelitis (TM). The typical magnetic resonance imaging (MRI) features of TM (Fig. 27-1) are abnormal high signal on T2-weighted images, often accompanied by contrast enhancement (35%-75%) on postgadolinium T1-weighted images and swelling/ expansion (50%) of the spinal cord. Contrast enhancement may be diffuse, poorly defined, and heterogeneous, nodular, or peripheral. TM usually extends over one to four vertebral segments and more often affects the thoracic region. Diffusion tensor imaging (DTI) is more sensitive in outlining the true extent of TM than T2-weighted images and has better clinical correlation with the affected levels and severity of symptoms. DTI yields information about axonal fiber tract integrity, and loss of that integrity shows a low fractional anisotropy (FA). DTI shows decreased FA within and distal to the area of TM.9,15 In autoimmune disorders such as rheumatoid arthritis and ankylosing spondylitis, systemic lupus erythematosus (SLE), mixed connective tissue disease, scleroderma, Sjögren’s syndrome, multiple sclerosis (MS), acute disseminated encephalomyelitis (ADEM), neuromyelitis optica (NMO), and antiphospholipid syndrome, TM is the most common clinical spinal cord manifestation.15 In this chapter we review the most salient clinical and imaging aspects of the noninfectious inflammatory disorders in which the spinal cord may be involved.

SYSTEMIC LUPUS ERYTHEMATOSUS SLE most commonly affects African American and Asian females between the ages of 10 and 50 years. Although in most patients SLE is considered to be idiopathic, it may be caused by medications, the most common being isoniazid, hydralazine, and procainamide.16 TM is a rare complication in SLE, affecting mostly young patients and occurring in the early stage of the disease.16 MRI findings are variable, and 16% to 30% of patients with clinical myelitis have a normal MRI study. Of the remainder, most have a long region of central cord hyperintensity on T2-weighted images, with expansion and contrast enhancement of the regions. A small percentage of patients have small multifocal bright lesions on T2-weighted imaging.9 Longitudinal myelitis is a variant of acute TM in which the cord abnormality on MRI is continuous over more than four vertebral segments. There is abnormal high T2 intramedullary signal intensity accompanied by swelling of the entire transverse section of the spinal cord and often with contrast enhancement. Longitudinal myelitis is a rare but severe complication of SLE and tends to occur in the early

phase of the disease in young SLE patients. Early diagnosis and treatment with prednisone and cyclophosphamide may ameliorate its poor prognosis.5,19

SARCOIDOSIS Sarcoidosis is a systemic granulomatous disease of unknown etiology that involves multiple organs and is characterized microscopically by the presence of noncaseating granulomas. It is estimated that up to 4 to 20 out of 10,000 people in the United States have sarcoidosis. Overall, central nervous system involvement is diagnosed in 10% to 15% of patients with systemic sarcoidosis.6,20 Involvement of the spinal cord in sarcoidosis is extremely rare, occurring in only 0.34% of patients with systemic disease.20 Sarcoidosis most often occurs between age 20 and 40, with women affected more than men. The disease is 10 to 17 times more common in African Americans than whites. People of Scandinavian, German, Irish, or Puerto Rican origin are also more prone to the disease.18,21 Although isolated central nervous system involvement occurs in fewer than 1% of patients, it is often the cause of presenting symptoms. Spinal disease may involve the nerve roots, intracanalicular extradural space, dura, leptomeninges, parenchyma, and surrounding bony spine (vertebral bodies and disks). Imaging findings in spinal cord sarcoidosis are nonspecific, and their differential diagnosis includes tumor, MS, and fungal infections. Fusiform enlargement of the cord, with high T2 signal and low T1 signal intensity and patchy contrast enhancement is common (Fig. 27-2). MRI shows abnormal linear and/or nodular contrast enhancement along the surface of the cord. The disease starts at the surface of the cord and then moves into the parenchyma centripetally, presumably via the perivascular spaces. There may also be involvement of the spinal nerve roots that enhance with gadolinium. Spinal cord sarcoidosis occurs at any level as follows: 56% in the cervical region, 37% in the thoracic level, and 7% at the lumbar and sacral levels.20 With resolution of the inflammatory process, the cord may return to normal size or become atrophic (Fig. 27-3) and show no further contrast enhancement. Spinal cord sarcoidosis is difficult to diagnose because it  commonly mimics other diseases, but lesions at more than three distinct levels help distinguish it from TM, which is generally a single lesion.6,18,20,21,26 Rapid onset of profound weakness correlates with extensive cord involvement and poor outcome. Rapid resolution of MRI findings following therapy suggests a favorable outcome.9 Cerebrospinal fluid (CSF) evaluation usually shows elevated protein and/ or pleocytosis. Serum and CSF angiotensin-converting enzyme levels are neither sensitive nor specific for neurosarcoidosis.6 A confident diagnosis can be made by a positive Kviem test, in which spleen or liver tissues from a patient with known sarcoidosis are intradermally injected, followed by a skin biopsy 4 to 6 weeks later. A positive result

http://radiologyebook.com

787

788

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 27-1  Transverse myelitis. A, Sagittal T2-weighted MRI through the thoracic spine demonstrates diffuse abnormal high signal involving the spinal cord (arrows). B, Axial T2-weighted MRI at the level of thoracic spine demonstrates abnormal high signal involving most of the transverse diameter of the cord.

FIG 27-2  Sarcoidosis. Sagittal postgadolinium T1-weighted MRI through cervical spine demonstrates focal dorsal enhancement of the spinal cord (arrow).

FIG 27-3  Sarcoidosis. Sagittal T2-weighted MRI through the thoracic spine demonstrates diffuse abnormal high signal (long arrows) and a long segment of cord atrophy (short arrow).

http://radiologyebook.com

CHAPTER 27  Noninfectious Inflammatory Diseases Affecting the Spinal Cord

A

789

B

FIG 27-4  Multiple sclerosis. A, Sagittal T2-weighted MRI through cervical spine demonstrates multiple oblong (cigar shaped) foci of abnormal high signal. B, Sagittal postgadolinium T1-weighted MRI through cervical spine again demonstrates multiple oblong areas of abnormal enhancement (short arrows) as well as a ring of enhancement (long arrow).

will show typical sarcoid granulomata.21 Biopsy of affected organs may be helpful and can most easily be taken from the skin or the outer membrane of the eyes.12 Fluorine-18 fluorodeoxyglucose positron emission tomography (18FDG-PET) may be useful in evaluating the systemic extent of sarcoidosis, with a sensitivity of 80%, specificity of 67%, and accuracy of 76%.25 Patients with neurologic sarcoidosis are generally treated with corticosteroids and other immunomodulatory agents.6

MULTIPLE SCLEROSIS MS is an autoimmune disease in which myelin around axons is destroyed by inflammation. The disorder is most commonly diagnosed between ages 20 and 40 years but may be seen at any age.13 In the United States approximately 350,000 individuals are affected by MS, and the chance of developing it is 1 : 750. Females are two to three times more likely to develop MS than males, and whites are at a higher risk for it than blacks or Asians.3 MS may clinically mimic the symptoms of many other disorders; an important component of the diagnosis is oligoclonal bands in CSF.13 Clinical features of spinal cord involvement include ataxia, weakness (mono/hemi/para/quadriparesis), and paresthesia.1 On MRI, TM may be a rare initial presentation of MS. Enlargement of the cord is usually seen during the active phase of MS, but this enlargement is subtle and less than expected for the size of the lesions.7 Spinal cord lesions are cigar shaped and extend less than two vertebral bodies in length. Larger active lesions may have extensive edema with associated cord expansion and may appear as enhancing nodules, rings, or arcs (Fig. 27-4). The period of contrast enhancement may last 2 to 8 weeks. In contrast to TM, these lesions involve less than half of the spinal cord diameter, lie eccentrically, rarely show mass effect, and

preferentially affect the lateral cord and posterior columns (Fig. 27-5). The association between spinal cord atrophy and disability progression is strong. Chronic lesions show no contrast enhancement and often result in focal atrophy.7,9 DTI is useful in evaluating patients with MS, and cord lesions show low FA and high values of mean diffusivity, implying myelin sheath breakdown and disorganization of the normal directionality of extracellular water movement.8 Commonly, treatment of cord lesions consists of immunomodulatory therapy with interferon and the monoclonal antibody natalizumab.11

ACUTE DISSEMINATED ENCEPHALOMYELITIS ADEM is a severe acute demyelinating disease usually triggered by an inflammatory response to viral or bacterial infections and/or vaccinations. Patients commonly present with nonspecific symptoms including headache, vomiting, drowsiness, fever, and lethargy, all of which are relatively uncommon in MS.1 The course of ADEM is usually monophasic, and it affects children more commonly than adults, with no predilection for gender. In general, ADEM is self-limited and prognosis is favorable. Although 35% of patients have relapses, complete recovery from ADEM is reported for 57% to 89% of them. One episode of ADEM can be the first manifestation of the classical relapsing form of MS; 21% of patients with ADEM develop MS after 2.36 years and 27% after 5.64 years.4 Several clinical features help to distinguish between MS and ADEM. In one study the mean age at presentation was 14 years in MS compared with 7 years in ADEM. Encephalopathy and seizures are more common in ADEM, whereas other symptoms such as headache, tremor/dysmetria, ataxia, weakness, and paresthesias are present in both diseases. In ADEM, large active lesions have extensive edema with associated spinal cord expansion and may show

http://radiologyebook.com

790

PART II  CT and MR Imaging of the Whole Body enhancing nodules, rings, or arcs1,9 (Fig. 27-6). Similar to MS, spinal cord lesions are cigar shaped and extend over less than two vertebral bodies. In contrast to TM, ADEM lesions involve less than 50% of the spinal cord diameter, lie eccentrically, rarely show mass effect, and preferentially affect the lateral cord and posterior columns.8,9 First-line treatment for ADEM includes intravenous high-dose corticosteroids, which in nonresponsive patients are followed by plasma exchange or immunoglobulin. Immunosuppressive agents such as mitoxantrone or cyclophosphamide are alternative therapies if antiinflammatory treatment fails.4

NEUROMYELITIS OPTICA NMO is a severe inflammatory necrotizing disorder distinct from MS. NMO was considered to be a monophasic illness, defined by the concomitant occurrence of acute bilateral visual loss and longitudinally extensive myelitis without brain abnormalities, but it is now recognized as a predominantly relapsing condition with acute visual loss and typical spinal cord lesions.2 From 10% to 40% of patients have a coexistent autoimmune disease such as SLE and or Sjögren’s syndrome.9 Early and accurate diagnosis is important because NMO carries a poorer prognosis than MS and treatment differs. Pathology in NMO shows fragmentation and loss of astrocytes that precede demyelination, and granulocyte infiltration with perivascular immunoglobulin and complement deposition found in later lesions. The autoantibody NMO immunoglobulin (Ig)G targets aquaporin 4 (AQP4), a cell membrane water channel protein found in the periependymal regions surrounding the third ventricle, cerebral aqueduct, fourth ventricle, and central spinal canal. NMO-IgG is found in the

FIG 27-5  Multiple sclerosis. Axial T2-weighted MRI at level of cervical spine demonstrates focal abnormal high signal in the lateral aspect of the cord (arrow).

A

B

FIG 27-6  ADEM. A, Sagittal T2-weighted MRI through the cervical and proximal thoracic spine demonstrates diffuse swelling as well as diffuse abnormal high signal. B, Sagittal postgadolinium T1-weighted MRI through the cervical and proximal thoracic spine demonstrates areas of abnormal enhancement (short arrows) as well as nonenhancing areas of edema (long arrow).

http://radiologyebook.com

CHAPTER 27  Noninfectious Inflammatory Diseases Affecting the Spinal Cord

A

791

B

FIG 27-7  Neuromyelitis optica. A, Sagittal T2-weighted MRI through the cervical and proximal thoracic spine demonstrates diffuse swelling and confluent abnormal high signal (long arrow) along with more focal cystlike areas of abnormal high signal (short arrow). B, Sagittal postgadolinium T1-weighted MRI of the cervical and proximal thoracic spine demonstrates multiple areas of patchy abnormal enhancement.

serum of 70% to 90% of NMO patients, and its presence is 90% specific in patients presenting with visual disturbances and symptoms related to the spinal cord. NMO-IgG is not detected in MS patients. MRI of the spinal cord is the single most useful diagnostic test for NMO, showing longitudinally extensive spinal cord lesions spanning more than three vertebral segments in length, which in the acute phase demonstrate patchy contrast enhancement and swelling2 (Fig. 27-7). The lesions are hyperintense on T2-weighted imaging, frequently hypointense on T1-weighted imaging, and may acutely expand the cord and enhance throughout their length after contrast administration.9 However, these findings are subject to study timing issues because a long T2-weighted image lesion may not have developed fully in the first few days after clinical symptom onset or may have diminished or resolved over time. NMO lesions preferentially involve spinal central gray matter.2 T1-weighted image hypointensity and cord atrophy are prominent findings in established NMO, reflecting severe tissue injury, and there is a propensity for recurrent bouts of inflammatory activity to occur in older lesions that will again enhance during acute relapse. Additionally, short-segment T2 hyperintense lesions have been described in NMO. Although a confluence of multiple short-segment MS lesions may occasionally mimic the MRI features of spinal NMO, contiguous T2 hyperintensity of central gray matter over many segments is rarely observed in MS. Current diagnostic criteria require the occurrence of both optic neuritis and myelitis and at least two of three supportive criteria, including long cord lesions, seropositivity for NMO-IgG, and a brain MRI study that is nondiagnostic for MS.2 Early and aggressive immunosuppression prevents or reduces severe disabilities. Routine imaging surveillance is necessary to monitor for complications of immunosuppressive therapy. Excellent clinical

and imaging recovery may be seen in some patients. NMO does not respond to traditional MS immunomodulatory therapy with interferon, which in some NMO patients may even be harmful. Severe exacerbations of NMO have been reported in patients treated with natalizumab. One case series suggests that NMO responds to immunosuppressive therapy aimed at reducing serum levels of NMO-IgG. The B-cell–depleting anti-CD20 monoclonal antibody rituximab has been shown to be effective but has been associated with progressive multifocal leukoencephalopathy in patients with systemic autoimmune disease.2

SJÖGREN’S SYNDROME Sjögren’s syndrome is a common autoimmune disease mostly affecting postmenopausal women.24 Between 400,000 and 3.1 million adults in the United States have Sjögren’s syndrome. This condition affects people of any age, but symptoms usually appear between age 45 and 55. It affects 10 times more women than men. About 50% of patients have rheumatoid arthritis or other connective tissue diseases such as SLE.17 Its diagnosis depends on a combination of symptoms, physical examination, and blood tests. Typical antibodies include antinuclear antibodies, anti-SSA and SSB antibodies, or rheumatoid factor. Biopsies of major salivary glands or those under the surface of the inner lips may be used to make a diagnosis. Neurologic symptoms occur in 20% to 27% of patients and may mimic those of MS and NMO; many patients have a coexistent autoimmune disease such as SLE.9,24 In most patients the neurologic symptoms precede the diagnosis of Sjögren’s syndrome, with symptoms developing acutely and usually including severe neck and interscapular pain followed by sensory and motor deficits below the level of the cord lesion.24

http://radiologyebook.com

792

PART II  CT and MR Imaging of the Whole Body

There is some disagreement about the MRI spinal cord findings. According to some reports, MRI more often shows a long-segment central cord lesion, high signal on T2-weighted images with expansion and variable contrast enhancement, but some patients have multifocal lesions.9 Other reports state that TM is the most frequent form of spinal cord involvement in Sjögren’s syndrome, with abnormal high signal on T2-weighted images, often with contrast enhancement on postgadolinium T1-weighted images, accompanied by swelling of the entire transverse section of the spinal cord (70%). The lesions extend longitudinally over one to four vertebral segments15,24 Sjögren’s myelitis carries a high mortality.24 Hydroxychloroquine (Plaquenil), an antimalarial drug used in SLE and rheumatoid arthritis, may be helpful in some patients with Sjögren’s syndrome.17

ANTIPHOSPHOLIPID SYNDROME Antiphospholipid syndrome (APS) is an autoimmune condition characterized by arterial and venous thromboses and persistently positive antiphospholipid antibodies. In addition to peripheral arterial and venous thromboses affecting vessels of any size, secondary APS occurs with other connective tissue diseases, most often SLE. TM is the most common spinal cord manifestation of secondary APS associated with SLE but is rare in primary APS. Some authors suggest that vascular thrombosis affecting the blood-cord barrier promotes an inflammatory response leading to myelitis. Treatment with steroids and anticoagulants may be effective in patients with myelopathy associated  with APS.22,23,27

POSTRADIATION MYELOPATHY

FIG 27-8  Postradiation myelopathy. Sagittal T1-weighted MRI with fat saturation through thoracic spine demonstrates abnormal areas of enhancement adjacent to vertebral bodies, with areas of decreased signal (arrows) indicative of fatty marrow replacement.

Postradiation myelopathy has been reported in patients up to 10 years following radiotherapy. MRI shows focal cord swelling and diffuse hyperintensity on T2-weighted images corresponding to the radiation portal. Often there is patchy or ringlike contrast enhancement following gadolinium administration (Fig. 27-8), and the lesion may simulate a tumor. Long-term follow-up studies show normalization of cord signal with development of severe atrophy. Increased vertebral body fatty marrow signal on T1-weighted images (decreased signal if fat saturation used) is often seen in the same geographic region radiation portal and helps one suggest the diagnosis of radiation-induced myelitis.9

GUILLAIN-BARRÉ SYNDROME Guillain-Barré syndrome (GBS) is a rare disease affecting 1 person in 100,000. It occurs a few days or weeks after symptoms of a respiratory or gastrointestinal viral infection. Occasionally surgery will trigger GBS. In rare instances vaccinations may increase its risk. Pathologically the immune system starts to destroy the myelin sheaths that surround the axons of peripheral nerves or the axons themselves.14 Clinical presentation of GBS is characterized by progressive and ascending weakness of the extremities and areflexia. Progression of symptoms is rapid and may include sensory disturbances, autonomic dysfunction, pain, numbness, paresthesia, and respiratory failure. Diagnosis is usually established on the basis of the clinical presentation but may be aided by CSF characteristics, electrophysiologic examination, and neuroimaging findings. An elevated CSF protein level without pleocytosis is typical of GBS. Thickening and contrast enhancement in the ventral (and sometimes dorsal) cauda equina nerve roots on MRI is characteristic (Fig. 27-9) and has been shown to be present in 95% of patients with GBS. Because GBS is an ascending myelitis, the more cephalad ventral gray matter and nerve roots

A

B

FIG 27-9  Guillain-Barré syndrome. A, Axial postgadolinium T1-weighted MRI at the level of the conus medullaris demonstrates abnormal enhancement of ventral nerve roots. B, Sagittal postgadolinium T1weighted MRI through lumbar and lower thoracic spine demonstrates abnormal enhancement of ventral roots of the cauda equina (short arrow) as well as ventral horns of the distal spinal cord (long arrow).

http://radiologyebook.com

CHAPTER 27  Noninfectious Inflammatory Diseases Affecting the Spinal Cord

FIG 27-10  Guillain-Barré syndrome. Axial postgadolinium T1-weighted MRI at the level of the midthoracic spine demonstrates abnormal enhancement of the ventral nerve roots.

may show enhancement (Fig. 27-10). Occasionally the ventral gray matter horns in the spinal cord may enhance, and rarely this abnormality is the only one found.28 There is no known cure for GBS, but therapies such as plasma exchange (plasmapheresis) and high-dose immunoglobulin lessen the severity of the illness and accelerate recovery in most patients.14

REFERENCES 1. Alper G, Heyman R, Wang L: Multiple sclerosis and acute disseminated encephalomyelitis diagnosed in children after long-term follow-up: Comparison of presenting features. Dev Med Child Neurol 51(6):480– 486, 2009. Published online October 24, 2008. Accessed 12/15/13. 2. Barnett Y, Sutton IJ, Ghadiri M, et al: Conventional and Advanced Imaging in Neuromyelitis Optica. Published online June 13, 2013. 3. Burks J, Bigley GK, Hill H: Accessed 11/4/13. 4. Cañellas A, Gols A, Izquierdo J, et al: Idiopathic inflammatorydemyelinating diseases of the central nervous system. Neuroradiology 49(5):393–409, 2007. 5. Dai YJ, Liu Q, He J, et al: Myelopathy associated with systemic lupus erythematosus: A clinical analysis of 10 cases and review of literature. Zhonghua Nei Ke Za Zhi Mar 52(3):213–217, 2013. [Article in Chinese] Abstract. 6. Eisen A: Disorders affecting the spinal cord. . Accessed 10/22/13. 7. Fertikh D, Smirniotopoulos JG: Imaging of the spine in multiple sclerosis. Updated: Feb 3, 2012. Accessed 10/30/13.

793

8. Filippi M, Rocca M, De Stefano N, et al: Magnetic resonance techniques in multiple sclerosis. The present and the future. Arch Neurol 68(12):1514–1520, 2011. 9. Goh C, Phal PM, Desmond PM: Neuroimaging in acute transverse myelitis. Neuroimag Clin North Am 21(4):951–973, 2011. 10. Grossman R, Yousem D: The requisites: Neuroradiology, ed 2, Philadelphia, 2003, Mosby, pp 800–804. 11. . Accessed 12/15/13. 12. . Accessed 12/24/13. 13. Multiple sclerosis MS; Demyelinating disease Last reviewed: September 26, 2011. Accessed 10/30/13. 14. Accessed 10/31/13. 15. Accessed 10/7/13. 16. Accessed 10/22/13. 17. American College of Rheumatology. 18. Accessed 10/7/13. 19. Huang LK, Chung CC, Chen BZ, et al: Systemic lupus erythematosus presented as extensive longitudinal myelitis. Acta Neurol Taiwan 22(2):67–71, 2013. Abstract. 20. Kobayashi S, Nakata W, Sugimoto H: Spinal magnetic resonance imaging manifestations at neurological onset in Japanese patients with spinal cord sarcoidosis. Intern Med 52:2041–2050, 2013. 21. Lury KM, Smith JK, Matheus MG, et al: Neurosarcoidosis: Review of imaging findings. Semin Roentgenol 39(4):495–504, 2004. 22. Miroslav M, Mislav C: Antiphospholipid syndrome and central nervous system. Clin Neurol Neurosurg 112(7):602–608, 2010. 23. Rodrigues CE, de Carvalho JF: Clinical, radiologic, and therapeutic analysis of 14 patients with transverse myelitis associated with antiphospholipid syndrome: Report of 4 cases and review of the literature. Semin Arthritis Rheumatol 40(4):349–357, 2011. [Epub August 10, 2010]. 24. Roman G, Ruiz P: Sjogren syndrome: Neurologic complications. . Originally released June 28, 2006; last updated June 3, 2012. 25. Rubini G, Cappabianca S, Altini C, et al: Current clinical use of 18FDG-PET/CT in patients with thoracic and systemic sarcoidosis. Radiol Med 2013. [Epub ahead of print]. 26. Sohn M, Culver DA, Judson MA, et al: Spinal cord neurosarcoidosis. Am J Med Sci 2013. . [Epub ahead of print]; Accessed 10/31/13. 27. Tsutsumi Y, Mochizuki A, Maruyama K, et al: Myelopathy in patients with antiphospholipid antibodies: Clinical features, pathogenesis, and review of literature. Rinsho Shinkeigaku 44(10):655–660, 2004. Abstract [Article in Japanese]. 28. Zuccoli A, Panigrahy A, Fitz C: Redefining the Guillain-Barré spectrum in children: Neuroimaging findings of cranial nerve involvement. AJNR Am J Neuroradiol 32:639–642, 2011.

http://radiologyebook.com

28  Spinal Trauma Roy Riascos and Eliana Bonfante

Imaging studies play a central role in the diagnosis of spinal trauma. The biomechanics, natural history, and treatment methods differ significantly in cervical and thoracolumbar injuries and are therefore addressed separately in this chapter. Multiple classification systems have been developed and continue to evolve. We will focus on the analyses of injury characteristics and will discuss the classification systems that have been evaluated for validity and reliability. Vascular injuries of the cervical vessels and aorta, although common in the setting of spinal trauma, are beyond the scope of this chapter. The appropriateness criteria for imaging studies will be reviewed because they are essential for the radiology practice.

CERVICAL SPINE TRAUMA INDICATIONS The imaging modalities that are available for the evaluation of spinal trauma include plain radiographs of the spine, computed tomography (CT) of the cervical spine with and without contrast, magnetic resonance imaging (MRI) of the cervical spine with and without contrast, CT post myelography, CT of the head and neck with contrast, magnetic resonance angiography (MRA) of the neck with and without contrast, and cervicovertebral angiography. The indications for imaging the cervical spine after trauma remain a controversial topic. The 2012 revision of the appropriateness criteria for imaging of suspected cervical spine trauma published by the American College of Radiology (ACR) estimates that over 1 million blunt trauma patients present to U.S. emergency rooms with a potential cervical injury each year.22 Cervical spinal cord injury occurs in 5 of every 10,000 cases per year15 and is currently the most commonly missed severe injury with serious implications.24 To improve utilization of radiographic studies in such instances, an extensive review of the literature that included findings from the National Emergency X-radiography Utilization Study (NEXUS) and the Canadian C-spine Rule (CCR) group established criteria that safely excludes cervical fracture without the use of imaging in low-risk patient groups. To do this, patients were divided into groups and the relevance of each of the imaging modalities was analyzed. In the high-risk patient group, CT remains the preferred initial imaging modality.11 NEXUS and CCR criteria provide methods for the referring physician to evaluate the need for imaging studies in suspected cervical lesions. Studies show sensitivities ranging from 0.83 to 1.0 and specificities ranging from 0.01 to 0.77 for the criteria; the CCR has slightly better diagnostic accuracy.68,99 The NEXUS criteria, published in 2000, were intended to decrease unnecessary radiographic evaluation of the cervical spine in trauma patients. The criteria recommend cervical spine imaging in patients with one or more of these findings: evidence

794

of intoxication, posterior midline tenderness, focal neurologic deficit, clinically apparent distraction injury, or a decreased level of alertness.39 There are no data available to establish the reliability of NEXUS criteria in young children, owing to radiation exposure concerns in this population. The CCR criteria are used to evaluate patients who are alert and in stable condition where spinal injury is suspected; patients are categorized into high- or low-risk groups. High-risk patients are older than 65 years, present with the presence of paresthesia, or have a history of a dangerous mechanism of injury, including falls from an elevation greater than 3 feet or 5 stairs, axial load to the head, motor vehicle collision at speeds greater than 100 km/hour (62 miles/h [mph]) or rollover or ejection, a collision involving a motorized recreational vehicle, or a bicycle collision. Indicators of low risk are simple rear-end motor vehicle collision, a patient in the sitting position in the emergency department, ambulation at any time, delayed onset of neck pain, and absence of midcervical tenderness. No imaging is necessary for patients fulfilling low risk-factor criteria who are able to actively rotate the neck 45 degrees from left to right.99,100 Similarly, meta-analysis of 14 studies including NEXUS and CCR criteria indicates that asymptomatic patients without a distracting injury who maintain a full range of motion of the cervical spine can be cleared without radiographic evaluation, with a sensitivity of 98% and negative predictive value of 99.8%.7

IMAGING MODALITIES Plain Radiographs Three-view plain radiographs—anteroposterior (AP), cross-table lateral, and odontoid views—offer a fast and low-cost imaging option but should only be used when the suspicion of injury is low. Plain radiographs may fail to demonstrate lesions visible in other crosssectional imaging modalities; missed injury rates of 25% to 60% have been demonstrated in plain film radiography.67 In a study of 10,496 cervical spine injuries conducted by Mower et al.,70 plain radiographs detected 932 injuries in 498 patients and missed 564 injuries in 320 patients. Evidence does not support the use of additional oblique views,32 and although dynamic views (flexion and extension) have been shown to demonstrate unstable cervical lesions, successful evaluation requires an adequate range of motion, which can be limited in the acute setting.16,42,56 Dynamic supervised views are best used for follow-up of symptomatic patients. Radiography is recommended in children younger than 14 years.22,62

Computed Tomography CT is the current study of choice for suspected cervical spine injury.2 CT outperforms plain radiography for evaluation of blunt trauma,5,11,18,19

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

A

795

B

FIG 28-1  Extruded disk in a 34-year-old with normal CT of the cervical spine after a motor vehicle accident, who presented with neurologic symptoms. MRI axial T2-weighted (A) and sagittal STIR (B) images show an extruded disk at C5-C6 (arrows) causing spinal cord compression.

with measured sensitivities of 98% compared to 52% for plain films in a meta-analysis of 7 studies of high-risk patients.2 Newer multidetector CT imaging of the spine with submillimeter acquisitions allows the generation of high-quality multiplanar reformatted (MPR) sagittal and coronal reconstructions without the stairstep artifacts seen with single-detector studies.8 Although this improves interpretations of CT scans in trauma patients, particularly the ability to differentiate stable from unstable fractures, it also increases the volume of images radiologists are expected to read.9,88 Therefore the ACR criteria recommend a slice thickness of 1.25 mm for sagittal and coronal reconstructions. Because CT scans expose patients to higher mean effective doses of radiation than plain films,24,25 these studies should be limited to patients at high risk for cervical injury. The use of multislice CT protocols in modern scanners has contributed to decreases in radiation exposure following CT in recent years.25,26

Magnetic Resonance Imaging MRI can be used for evaluation of cervical trauma but is typically indicated for suspected myelopathy, treatment planning for the mechanically unstable spine, patients who cannot be clinically evaluated for more than 48 hours, patients with neurologic deficits, or suggested ligamentous injury. The main role of MRI is detection of the type and extent of spinal cord injury, which affects patient management.81 MRI is better suited for exclusion of spinal cord lesions than detection of ligamentous or other soft tissue injuries. In a 2001 study correlating MRI findings with cadaveric cryosections, more than half of soft tissue injuries were not detected97; however, MRI performs slightly better in detection of clinically significant lesions. No studies compare more current MRI technology with cadaveric studies, but MRI has only a slight advantage in the detection of occult instability injuries not detected on CT. In fact, Pourtaheri et al. found that MRI does not provide any additional clinically relevant information in patients who are alert, neurologically intact, and younger than age 60.81

INCIDENCE The major causes of cervical spine injury in the young population are motor vehicle accidents (MVAs), acts of violence, and sports injuries19,46; falls are more common in the geriatric population.105 The cervical spine is the most likely region to fracture following motor

vehicle collision (Fig. 28-1). A recent study reviewed the data from 33,015 front-seat occupants sustaining MVAs and found 0.4% had cervical spine fractures, 0.24% had thoracic spine fractures, and 0.3% had lumbar spine fractures.72 The mean relative collision speed for cervical fractures was 48 km/h (30 mph) compared to 40 km/h (25 mph) for the lower segments. In 2012, the Association for the Advancement of Automotive Medicine reported an increase in incidence of injuries of the thoracolumbar spine in frontal impact crashes, finding burst-type fractures at T12, L1, or L5 and wedge compression deformities commonly at L1.77 Additionally the elderly are more prone to spinal injury due to MVAs, possibly owing to changes in biomechanical tolerance.10

FRACTURES The study of cervical spine injuries can be approached based on the mechanism of injury: flexion, extension, compression/axial load, and rotational.12 Lesions may involve the bony elements, soft tissues (ligaments, joint capsule or peripheral soft tissues), spinal canal, or neural elements including the spinal cord. Injuries of the cervical spine are of great importance owing to the proximity of the spinal cord and other neural structures.12 Cervical spine fractures are divided into upper and lower cervical lesions.

Craniocervical Injuries Upper cervical fractures include structures from the occiput to the second cervical vertebra and may also involve the occipital condyles, the atlas, the odontoid process of C2, the pars interarticularis of C2 (hangman’s facture), and the atlantoaxial joint.

Anatomy of the Craniocervical Junction.  The craniocervical junction (CCJ) has a unique anatomy that makes evaluation of this region challenging. It is crucial to understand the complex morphology to appropriately evaluate potential lesions of this region. Three bony structures constitute the CCJ: the occiput, the first cervical vertebra (atlas [C1]), and the second cervical vertebra (axis [C2]). These bones articulate in a way that maximizes motion of  the spine region while avoiding injury to the neural and vascular  components. The structures approximate at two major joints: The

http://radiologyebook.com

796

PART II  CT and MR Imaging of the Whole Body

atlantooccipital and the atlantoaxial.107 The biomechanical properties of the atlantooccipital joint depend upon bony structure relationships, whereas the atlantoaxial joint relies on ligamentous relationships.118 Bony anatomy.  The occiput has two paired inferior projections called condyles that articulate with the superior articular surfaces of the lateral masses of C1. At the anterior aspect of the base of these condyles, foramina allow the hypoglossal nerve to exit the skull. The next structure, the atlas, is a ring-shaped vertebra with bilateral paired lateral masses fused by an anterior and a posterior arch. It is the only vertebra that does not have a body or a relationship with an intervertebral disk. The posterior aspect of the anterior arch has an articular surface for the odontoid process of C2 (dens). The lateral masses articulate superiorly with the occipital condyles and inferiorly with C2. The superior aspect of the lateral masses has a groove allowing passage of the vertebral artery before it enters the foramen magnum.118 The final structure is the axis, a unique vertebra that has a superior projection called the dens. The axis has paired superior articular processes to articulate with the lateral masses of the atlas and inferior articular processes to articulate with C3. Ligamentous anatomy.  The broad complex of ligaments involved in the CCJ can be divided into intrinsic and extrinsic ligaments  (Fig. 28-2). The intrinsic ligaments form three layers anterior to the dura matter and include the tectorial membrane, the cruciate ligament, and the odontoid ligaments.43 The tectorial membrane is a continuation of the posterior longitudinal ligament that runs posterior to the cruciform ligament. It attaches to the occipital bone and is in intimate contact with the dura matter.107 The cruciate ligament is composed of the transverse ligament and superior and inferior limbs. The transverse ligament is located in the atlas and traverses the posterior aspect of the dens, fixating it to the posterior aspect of the anterior arch of the atlas. This ligament divides the ring of the atlas into two compartments: an anterior compartment including the dens and a posterior compartment that houses the thecal sac and its contents. The odontoid ligaments comprise the alar and apical ligaments. The alar ligaments are paired structures that attach the base of the skull and the axis at the posterolateral aspect of the odontoid process; they act as stabilizers of

Anterior atlanto-occipital membrane

Apical ligament Cruciate ligament

A

Tectorial membrane

the CCJ and limit axial rotation. The apical ligament, also known as the suspensory ligament, joins the tip of the dens with the basion. The extrinsic ligaments are fibroelastic membranes that replace the anterior longitudinal ligament and the ligamentum flavum.43 The nuchal ligament is a continuation of the interspinous ligament that attaches to the occipital bone, restricting hyperflexion. Ligamentous function.  The atlantooccipital and atlantoaxial joints are both necessary for full range of motion. The atlantooccipital permits flexion and extension while limiting rotation, and the atlantoaxial allows rotation while prohibiting other movements. The anatomic morphology of the ligaments in these joints explains the range of motion. Flexion at the atlantooccipital joint is limited by the relationship of the tip of the dens with the basion; extension is limited by the tectal membrane. The alar ligaments at this level minimize axial rotation.98 Therefore CCJ hyperextension injuries often result in tectal membrane injuries and also can encompass alar ligament injury when rotational force factors into the trauma. The rotational range of motion at the atlantoaxial joint is permitted by the functional relationship of the transverse ligament, alar ligaments, and capsular ligaments.118 Special attention has been paid to evaluation of the ligamentous structures of the CCJ using MRI.47 Thin-slice MRI in three orthogonal planes provides good visualization of clinically relevant ligamentous structures.49,59,95,115 However, controversy still exists over the value of the imaging findings. Recent studies revealed that MRI signal changes in the alar and transverse ligaments did not correlate with whiplash injuries.59,113 Craniometry.  The normal landmarks of the CCJ are best evaluated in sagittal and coronal cross sections. For accurate craniometrics, the basion, hard palate, and tuberculum sella, which are often excluded from exams, must be visible.95 However, in the setting of trauma, the usefulness of these measurements is questionable. Interobserver agreement demonstrates that the best correlation between plain film and cross-sectional imaging measurements is the McRae line, extending from the basion to the opisthion. Changes in craniometry secondary to flexion and extension, as well as irregularity of the clivus as seen in cross-sectional imaging, might explain this controversy.51

Tectorial membrane

Superior crus of the cruciate ligament Apical ligament Alar ligament

Posterior atlanto-occipital ligament

Atlanto-occipital capsule

Anterior longitudinal ligament

Dens

Posterior longitudinal ligament

Axio-atlantal capsule

Transverse ligament

B FIG 28-2  Ligamentous anatomy. A, Diagram of a cross-sectional view of the craniocervical junction (CCJ) demonstrating the ligamentous anatomy. B, Diagram of a posterior view of the CCJ; the posterior elements have been removed and ligaments are demonstrated. In the center of the image lies the cruciate ligament.

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

Normal

Type I

Type II

Type III

FIG 28-3  Classification of atlantooccipital dislocation (AOD). In type I dislocations, the occiput is anteriorly displaced. In type II there is widening of the atlantooccipital space. In type III dislocations, the occipital bone is posteriorly displaced. Close attention to the craniocervical junction relation is crucial for this diagnosis. Occipitocervical fixation is the preferred treatment. (Adapted from Anderson PA: Injuries to the occipital cervical articulation. In Clark CR, et al, editors: The cervical spine, ed 3, Philadelphia, 1998, Lippincott-Raven, p 393.)

Types of Craniocervical Injury Atlantooccipital dissociation.  Atlantooccipital dissociation (AOD) can be life threatening, typically resulting from high-energy trauma. Although this is an uncommon injury, in fatal traffic accidents the incidence is as high as 8% to 31%.18,52,119 The high fatality rate results from concomitant brainstem or vascular injury.71 Three types of AOD have been described; the most common is ventral dislocation of the occiput—type I; the most unstable is axial distraction—type II; and the third is dorsal dislocation of the occiput—type III.106 The most frequent mechanism of injury that produces AOD includes extreme hyperextension of the neck along with some degree of rotation (Figs. 28-3 and 28-4). There is also an association with AOD and the presence of condylar fractures. Clinically these patients present with neurologic symptoms suggesting brainstem injury, and urgent reduction of the dissociation has been shown to improve long-term survival rates.2 A higher incidence has been reported in the pediatric population.18 As discussed in previous sections, plain radiographs have limited usefulness in this situation, and MPR CT images are ideally needed for proper visualization of the craniocervical articulation. MRI can be used to identify major ligamentous disruptions and evaluate lesions to the spinal cord.35 A displacement greater than 10 mm between the basion and the tip of the dens (basion-dens interval) is a reliable method to measure AOD.117 In a comparison of basion-dens interval, condylar gap, Harris method (basion-atlas index), and the BC : OA Powers’ ratio, Dziurzynki et al. found the basion-dens interval to have the highest sensitivity and specificity for determination of AOD.27 Occipital condyle fractures.  Occipital condyle fractures are uncommon lesions in blunt trauma, generally resulting from com­ pression mechanisms. They can be classified into three types:

797

comminuted—type I; extension of a linear skull fracture—type II; and avulsion injury—type III.104 Condylar fractures may be bilateral in up to 50% of cases. Special attention should be paid to type III condylar fractures because they can be associated with injury of the alar ligament, causing instability of the CCJ5 (Figs. 28-5 and 28-6). Plain films have a very low sensitivity to detect these lesions; CT is the best diagnostic tool to identify them. MRI can be helpful in identifying ligamentous involvement. With this injury, cranial nerve damage, particularly of the hypoglossal nerve, is often a concern. Close attention should be paid to the other cervical segments when encountering this fracture, because other associated lesions such as type II odontoid, axis, and subaxial fractures may also be encountered.104 Symptoms tend to improve with cervical traction, the initial treatment of choice. Fractures of the atlas and transverse ligament rupture.  Fractures of the atlas account for 2% to 13% of acute cervical fractures.18,93 Also referred to as Jefferson’s fracture, they typically occur after application of an axial load to the skull, resulting in a burst fracture of the atlas at its weakest points, the anterior and posterior arches. These fractures may be isolated or associated with other lesions such as transverse ligament injury or odontoid fractures28 (Figs. 28-7 and 28-8). Evaluation of the anterior transverse ligament is essential for the prognosis. Rupture of this structure destabilizes the atlantodental relationship and can result in posterior displacement of the dens in relation to the atlas.60,65 If the atlas is considered as a ring, the presence of single fractures is unusual; typically at least two fracture sites will be present. The most common fractures will be (1) posterior arch— associated with hyperextension (2) burst—associated with axial load, and (3) lateral mass fracture—associated with lateral flexion43 (see Fig. 28-7). Isolated C1 fractures tend to not have neurologic compromise, but when displacement of the lateral masses occurs, concomitant rupture of the anterior transverse ligament is possible. Disruption of the anterior transverse ligament without a fracture of the C1 ring can also occur; these lesions usually are accompanied by disruption of the alar ligaments. They have been classified as type I (lesions of the ligament alone) and type II (presence of avulsed bone fragments). Evaluation of the atlas typically requires an open-mouth view for a frontal radiograph. A discontinuity of the lateral line that extends from the lateral masses of C1 to C2 would be present. Displacement of more than 7 mm indicates anterior transverse ligamentous injuries.96 Dynamic (flexion and extension) radiographs can be used as a screening tool. Widening of the atlantoodontoid space would suggest anterior transverse ligament disruption in the trauma setting, but if instability is suspected, dynamic views should be avoided.25 CT is helpful for delineating the fracture lines in the atlas. Coronal multiplanar images can demonstrate the discontinuity of C1 and C2 in a similar way to open-mouth radiographs (Fig. 28-9). MRI may be beneficial to study the ligaments,28 but its usefulness is limited because typical cervical spine MRIs do not cover the atlas in axial views, which hinders evaluation of the anterior transverse ligament. In a normal study the ligaments should be low intensity and homogeneous on gradient recalled echo sequences. Lesions appear as discontinuity of the ligament or significant signal change.25 MRI has limited use in the initial evaluation of ligamentous involvement of high cervical injuries as well.114 Atlantoaxial rotatory deformity.  Atlantoaxial subluxations are rare events that are usually nontraumatic and can be caused by an incomplete odontoid process or an incomplete transverse ligament; they are highly associated with rheumatoid arthritis.41 The atlantoaxial articulation is the most mobile portion of the spine, with articular facets that are mostly flat, allowing ample movement in multiple

http://radiologyebook.com

798

PART II  CT and MR Imaging of the Whole Body

A

D

B

C

E

F

FIG 28-4  Atlantooccipital dislocation (AOD). Digital scout view (A), sagittal CT reconstruction (B and C), and coronal reconstruction (D) show increase in the space between the basion and tip of the dens (arrowheads). Widening of the atlantooccipital space (black arrows) and atlantoaxial space (white arrows) is seen. Sagittal T2-weighted MRIs (E and F) show increased signal and widening of the atlantooccipital space (arrows) and thinning with partial interruption of fibers of the tectorial membrane (arrowhead).

directions. Dislocation of the joint can result in compression of the thecal sac and neural structures. Traumatic rotatory subluxation occurs secondary to flexion and rotation force with lesions of the alar ligament; it is more frequent in children than in adults.44 Odontoid fractures.  Odontoid fractures are a common injury in the cervical spine, accounting for 5% to 15% of cervical fractures.4 They can occur in hyperextension or hyperflexion injuries. The classification of odontoid fractures by Anderson and D’Alonzo4 links the location of the fracture line with prognosis. Type I fractures compromise the upper part of the dens; type II fractures involve the base of the odontoid process; type III fractures affect the body of the axis (Figs. 28-10 and 28-11). However, a limitation of the AndersonD’Alonzo classification is the lack of descriptive terms for fracture lines.73 Roy-Camille et al.87 proposed an alternate classification that defines the angle of the fracture and the potential of dens displacement: type 1—oblique linear fracture in which its line slopes forward, with dens displacement in an anterior direction; type 2—oblique linear fracture in which its line slopes backward, with dens displacement in a posterior direction; and type 3—horizontal fracture line, with either anterior or posterior dens displacement. Treatment varies with type of injury. Type II fractures have the lowest potential for healing, especially in the elderly,65,73 and there is controversy regarding their management. These fractures may be life threatening if secondary spinal cord lesions occur, but this event is extremely rare. Type II fractures have a nonunion rate ranging from

10% to 77%; risk factors for nonunion include displacement greater than 5 mm, inability to maintain reduction, early initial treatment, and advanced patient age.91 Evaluation of nondisplaced odontoid fractures with x-rays is limited. High-resolution CT with multiplanar reformats allows proper visualization of the fracture lines and eases classification, which will be critical for treatment decision making. Most odontoid fractures are treated with halo vest immobilization. Traumatic spondylolisthesis of the axis.  Traumatic spondylolisthesis of the axis, representing around 15% of cervical spine fractures, is caused by hyperflexion of the cervical spine, producing bilateral fractures of the pars articularis of the axis. This fracture typically occurs in the setting of an MVA and tends to decompress the neural axis; patients are typically stable and have a good prognosis.31,36 Traumatic spondylolisthesis is similar in appearance to judicial hanging, which occurs after hyperextension of the neck, fracturing the posterior elements of C2 and causing severe spinal cord compression.21 The term hangman’s fracture is used loosely in the literature to describe traumatic pars articularis defects of the axis. The classification of C2 spondylolisthesis takes into account the angulation and translation between C2 and C3. It describes three types: type I—fracture through the neural arch with no deviation; type II— fracture with significant displacement over angulation; type IIA— oblique fractures with angulation, with minimal translation; and type III—fracture of the posterior arch of C2, with unilateral or bilateral

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

799

facet dislocation31 (Figs. 28-12 and 28-13). These fractures are rare in the pediatric population and if present, child abuse must be suspected.84 Difficulty evaluating these children arises because of anatomic variation seen as a line connecting the spinous processes of C1 and C3.102 CT and MRI are recommended if the diagnosis is not clear.84 Pain radiographs have demonstrated great reproducibility in characterizing translation and angulation of hangman’s fractures.13 MRI is helpful to evaluate spinal cord compromise if the fracture is suspected to be secondary to hyperextension. Vertebral artery lesions tend to be more prominent in comminuted C2 fractures that affect the foramen transversarum.26

Subaxial Cervical Spine Trauma The subaxial cervical vertebrae articulate with one another via intervertebral disks between each vertebral body and bilateral zygapophyseal joint facets, which are both flat and obliquely angulated, allowing subluxation. The spinous processes are typically bifid. Each transverse process is pierced by the foramen transversarium. The vertebral arteries enter the foramen transversarium between C7 and C6 and ascend the remaining cervical vertebrae protected within this structure. Ligaments provide structural support in the space between vertebrae, maintaining proper alignment; the ligamentum flavum connects adjacent lamina, and the thick interspinous ligament connects adjacent spinous processes. The neural foramen, at the center of all of these structures, is oriented obliquely and anteriorly. Greater than 60% of all cervical spine fractures and more than three fourths of the dislocations are subaxial.108 Different classification systems have been developed, and there is controversy regarding their accuracy and impact on clinical outcome. Most imagers have resorted to using descriptive terminology to characterize patterns of injury.108 The Allen and Fergusson classification has prevailed as the system most often used for description of fractures.3,65 This system groups cervical fractures based on mechanism of injury: compressive flexion, vertical compression, distractive flexion, compressive extension, and lateral flexion. Different stages exist within each group for a total of 21 different possible types. This system does not quantify the severity of a lesion or guide treatment, which is considered a limitation.37,101 This classification was expanded in the mid-1980s to include a combination of description and mechanism of injury. Today this expansion is rarely used in clinical practice.50 Recent efforts to further classify subaxial lesions have been made, giving rise to the subaxial ligamentous injury classification (SLIC; Box 28-1) and the cervical spinal injury severity score (CSISS).

Type I

Type II

Subaxial Ligamentous Injury Classification.  The SLIC was pro-

Type III

FIG 28-5  Occipital condyle fracture classification. Type I and II fractures are considered stable and are treated with a collar for 6 to 8 weeks. Type III injuries, if unstable, are treated with a halo vest. If incongruity is present, cervical fusion is considered. (Adapted from Anderson PA: Injuries to the occipital cervical articulation. In Clark CR, et al, editors: The cervical spine, ed 3, Philadelphia, 1998, LippincottRaven, p 391.)

posed by Vaccaro et al. in 2007108 to provide a comprehensive scoring system that went beyond mechanism of injury and is based on injury morphology, addressing both discoligamentous complex (DLC) integrity and the neurologic status of the patient.7 The advantage of using this classification is that it guides management strategies and has demonstrated better interrater reliability than the Allen and Fergusson classification.108 The SLIC algorithm has three components: injury morphology, integrity of the DLC, and neurologic status of the patient (see Box 28-1). Morphology.  The first component, morphology, evaluates the anterior support structures and is determined by the pattern of spinal column disruption seen on imaging studies. It is subdivided into three categories that describe the relation of the vertebral bodies to each other: compression, distraction, and translation/rotation. Compression.  Compression injuries demonstrate a visible loss of height of the vertebral body or disruptions of the end plates. The Text continued on p. 806

http://radiologyebook.com

800

PART II  CT and MR Imaging of the Whole Body

A

B FIG 28-6  Occipital condyle fracture. Coronal reconstruction (A), axial CT without contrast (B). Type III fracture; a small bony fragment is avulsed from the right occipital condyle (arrow).

Posterior arch

Burst

Lateral mass

Transverse ligament

FIG 28-7  Common types of fractures of the atlas. Arch fractures are stable. In Jefferson or burst fractures the lateral masses are typically laterally displaced. The lateral mass fractures can cause unilateral lateral mass displacement. Atlas fractures can be associated with transverse ligament lesions. (Adapted from Jackson RS, et al: Upper cervical spine injuries. J Am Acad Orthop Surg 10:271–280, 2002.)

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

801

A

B

C FIG 28-8  Burst fracture of the atlas. A and B, Axial and coronal reconstruction of a CT without contrast show widening of the space between the lateral masses (arrows) in relation to C2 (arrowheads). Sagittal STIR MRI (C) shows prevertebral fluid and fluid signal around the posterior elements of the atlas.

FIG 28-9  Assessment for odontoid alignment. Interruption of the continuity of a vertical line traced from the lateral aspect of the lateral mass to the lateral aspect of the articular process of the axis confirms misalignments of the dens with the lateral masses of the atlas. (Adapted from Jackson RS, et al: Upper cervical spine injuries. J Am Acad Orthop Surg 10:271–280, 2002.)

http://radiologyebook.com

802

PART II  CT and MR Imaging of the Whole Body

Type I

Type II

Type III

FIG 28-10  Odontoid fractures. Type I fractures affect the distal dens and are considered stable. Type II fractures course through the neck of the dens and are the most unstable and unlikely to heal. Type III fractures affect the body of the axis, tend to be stable, and have a high likelihood of healing. (Adapted from Anderson LD, D’Alonzo RT: Fractures of the odontoid process of the axis. J Bone Joint Surg Am 56:1663– 1674, 1974.)

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

A

B

C

D

E

G

F

H FIG 28-11  Odontoid fracture. Instability of a type II fracture in a 76-year-old patient after trauma. A and B, The fracture is identified at the base of the dens (arrows). Three months after the initial episode the patient returns with neck pain after a new trauma. CT of the cervical spine (C and D) shows widening of the previously described fracture (arrows). Sagittal T2- (E) and STIR-weighted (F) images show the fracture line (arrows). G and H, Another patient with an acute type II odontoid fracture with displacement. Axial and http://radiologyebook.com coronal reconstruction of a CT show an anteriorly displaced fracture of the base of the dens (arrows).

803

804

PART II  CT and MR Imaging of the Whole Body

Type I

Type IA

Type II

Type IIA

Type III

FIG 28-12  Traumatic spondylolisthesis of C2 is classified as type I when a fracture crosses the neural arch with no angulation or displacement or with minimal angulation and elongation (type IA). Type II fractures have significant angulation and displacement, while type IIA have important angulation with minimal translation. If the fracture shows translation it is considered a type III. These are very unstable fractures and typically need surgical correction to prevent neurologic deficits. (Redrawn from Torretti JA, Sengupta DK: Cervical spine trauma. Indian J Orthop 41:255–267, 2007.)

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

C

A

B

D FIG 28-13  Traumatic spondylolisthesis of C2. Axial CT of the cervical spine (A) and sagittal reconstruction (B) show a traumatic pars interarticularis defect of C2 on the right (arrows). Sagittal T1- (C) and T2-weighted (D) MRIs demonstrate the pars defect (arrows). The T2-weighted images show increased signal in the posterior elements due to edema.

http://radiologyebook.com

805

806

PART II  CT and MR Imaging of the Whole Body

BOX 28-1  Subaxial Ligamentous Injury

Classification (SLIC)

Subaxial Cervical Spine Injury Morphology 0 No abnormality 1 Compression 2+ Burst fracture 3 Distraction 4 Rotation/translation Anterior column

Discoligamentous Complex (DLC) 0 Intact 1 Indeterminate 2 Disrupted

DLC disruption

Neurologic Status 0 Intact 1 Root injury 2 Complete cord injury 3 Incomplete cord injury +1 Continuous cord compression (in setting of neurologic deficit) Treatment • SLICS ≤ 3: non-operative • SLICS = 4: consider for operative or non-operative intervention • SLICS ≥ 5: operative

Laminar fractures

Background SLICS = Subaxial Cervical Spine Injury Classification and Severity Scale—a composite score of morphology, DLC, and neurologic status. From Vaccaro AR, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374, 2007.

fractures included in this category include traditional compression fractures, burst fractures, sagittal and coronal fractures of the vertebral body, and “teardrop” fractures. Concomitant posterior element injuries may be present, especially when unevenly distributed trauma happens, involving the lateral masses or facet joints, as long as no distraction is present. Plain radiographs, CT, and MRI studies are useful to detect theses injuries108 (Figs. 28-14 through 28-17). Distraction.  Distraction is mainly seen as anatomic dissociation of the vertebral axis. The mechanisms needed to disrupt the anterior and posterior elements that keep the integrity of the spine have a larger mechanical force than the ones involved in the compression category. In hyperflexion injuries, distraction occurs when joint facet capsules and interspinous ligaments fail. In hyperextension, distraction will happen when the anterior longitudinal ligament, intervertebral disk, or vertebral body is disrupted by the mechanical forces, resulting in widening of the anterior intervertebral space. Bony fractures of the posterior elements (e.g., spinous processes) can be associated with these lesions.108 These disruptions typically spread through the intervertebral disk and facet joints. Evaluation of the facet joints is crucial in this category, because these joints are the primary posterior stabilization force78 (Fig. 28-18). MRI may guide in identifying ligamentous injuries, but its limitations have been described previously in this chapter. Translation/rotation.  When horizontal displacement between two segments of the cervical spine is evidenced through imaging, the

Non-displaced lateral mass fracture

FIG 28-14  Diagrams demonstrating simple compression morphology. The lesions show loss of anterior column height. They can affect the anterior column and be accompanied by DLC disruption or laminar fractures. Nondisplaced lateral mass or facet fractures are also considered simple compression injuries. (Adapted from Vaccaro AR, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374, 2007.)

injury morphology is classified as a translation/rotation. Dynamic images may be used to evidence the lesions. These lesions imply disruption of both the anterior and posterior elements. The normal  rage of rotation suggested for the infraaxial spine is less than 11 degrees,116 and the pathologic degree of translation is considered to be 3.5 mm in non-bony traumatic injuries.116 Unilateral or bilateral joint facet dislocations or fractures, bilateral pedicle fractures, and separation of the lateral mass are typically seen in this category108 (Figs. 28-19 and 28-20). If no morphologic abnormality is seen, a score of zero is given. Simple compression injuries are scored as 1, burst fractures receive a 2, distraction injuries receive 3 points, and translation/rotation lesions 4 points.108 Discoligamentous complex integrity.  The second SLIC component is integrity of the DLC (see Box 28-1). The DLC is composed of

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

807

B

A

FIG 28-15  Simple compression fracture of the spine. A, Axial CT shows a nondisplaced facet joint fracture line on the right side (arrow). B, Sagittal STIR image shows increased signal at the fracture site (arrow) and soft tissue edema.

FIG 28-16  Burst fracture. Severe compression lesion that involves the whole vertebral body and can be associated with retropulsed fragments. Lateral (left) and sagittal (right) diagrams of the cervical spine. (Adapted from Vaccaro AR, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374, 2007.)

disruption such as increased T2 signal, seen horizontally in the intervertebral disk, or increased signal in the intervertebral ligaments.108 No DLC injury receives zero points in the classification. Obvious DLC injury is given 2 points. When imaging studies are inconclusive about the presence of a DLC lesion, a 1-point score is given.108 Neurologic status.  The last component of the SLIC may be the single most important predictor for management decision making, focusing on decompressing the spinal cord to avoid further neurologic inury.108 The greater the disruption of the osseous and ligamentous structure of the spine, the more likely a nerve root or spinal cord injury may exist. If no neurologic injury is evidenced, a zero score is given. Nerve root injury is given 1 point, complete spinal cord injury receives 2 points, and 3 points are assigned to lesions that present as incomplete spinal cord injury. An additional point can be added if continuous cord compression in the setting of neurologic deficit is evidenced by MRI. How to use the SLIC.  The three categories of the SLIC system are identified and reported as seen in Box 28-1. The report should include the spinal level, injury level morphology, bony injury descriptor, DLC status and descriptors, neurologic status, and confounders (e.g.,  presence of ankylosing spondylitis [AS], diffuse idiopathic skeletal hyperostosis [DISH], osteoporosis, previous surgery, spondylosis).108 A numeric value is assigned to each category and the numeric sum is the SLIC score. The higher the score number, the worse the injury, increasing the likelihood for surgical intervention. If injury is noted at different levels, each level is assigned a separate score.

Cervical Spine Injury Severity Score.  The CSISS, published by the intervertebral disk, the anterior and posterior longitudinal ligaments, the ligamentum flavum, the interspinous ligaments, and the joint facet capsules. In imaging studies, indirect signs of DLC injury would include “fanning” of the spinous processes, joint facet dislocation, subluxation of the vertebral bodies, or a widened intervertebral disk space.108 DLC disruption often accompanies translation/rotation injuries. Widening of the joint facet greater than 2 mm or apposition of the facet joints of less than 50% and widening of the anterior disk space in neutral position or extension are considered absolute indication of DLC disruption. MRI may provide other clues to DLC

Moore et al. in 2006,6 is another method proposed to grade the severity of subaxial injuries to try to determine the prognosis of injuries. One of the main goals of this classification is to be able to assess cervical stability and translate that into the likelihood of considering surgery as a treatment option. This system considers the bony and ligamentous components. The spine is divided into four columns: anterior, posterior, right pillar, and left pillar. The anterior column includes the vertebral body, intervertebral disk, and the anterior and posterior longitudinal ligaments; the posterior column includes the spinous process, interspinous ligament, lamina, and ligamenta flava; the lateral pillars include the pedicle, pars articularis, joint facet with the capsules, and transverse process. For each column a score of 0 to 5 is given, with

http://radiologyebook.com

808

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 28-17  Burst fracture. A, Sagittal T2-weighted image demonstrates a burst fracture of C5 with a retropulsed fragment (arrows) that narrows the spinal canal. B, Axial gradient echo recalled image shows retropulsed fragments narrowing the spinal canal (arrow).

Circumferential

Facet dislocation

Anterior distraction with extension

Anterior distraction with flexion

FIG 28-18  Distraction morphology is identified by anatomic dissociation in the vertical axis involving the DLC. Lesions may be circumferential and can have associated facet dislocations. Anterior distraction in hyperextension may be associated with spinous process fractures. Distraction in hyperflexion can be associated with posterior ligamentous injury. (Adapted from Vaccaro AR, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374, 2007.)

Translation in sagittal plane

Translation with pedicle fracture

Translation with facet fracture

Rotational injury

FIG 28-19  Diagrams demonstrating translation and rotation injuries. These injuries are characterized with complete DLC disruption. Translation can happen in the sagittal plane, associated with pedicle fractures or joint facet fractures. Rotational injuries can also be present in addition to the axial translation. (Adapted from Vaccaro AR, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374, 2007.)

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

A

809

C

B

FIG 28-20  Translation injury. Axial CT (A) and sagittal reconstruction (B) show translation in the sagittal plane injury associated with a joint facet fracture. C, Sagittal T2-weighted MRI shows complete disruption of the DLC with sagittal translation. The spinal cord is compressed at the site of the injury (arrow).

0 representing no injury and 5 being the most severe fracture or dislocation possible for that column (Fig. 28-21). The numeric values are then summed in all four columns, and a final score of 0 to 20 will result for each level. The ability of this classification to predict prognosis is yet to be validated. CSISS is an image only–based classification. The authors found that patients with scores greater than 7 were likely to be treated surgically; 3 out of 20 patients with scores lower than 7 required surgical reduction.6

THORACOLUMBAR SPINAL TRAUMA INDICATIONS The indications for imaging studies in thoracolumbar (TL) trauma have been reviewed extensively. After a comprehensive review of the literature, Hsu et al. recommended imaging the TL spine if any of the following are present: (1) back pain or midline tenderness, (2) local signs of TL injury, (3) abnormal neurologic signs, (4) cervical spine fracture, (5) Glasgow Coma Score less than 15, (6) major distracting injury, (7) alcohol or drug intoxication. Fractures found in one level of the spine indicate an increased risk of spinal fractures elsewhere. Thus identification of a spinal fracture may imply a need to survey the remainder of the spine.40 The ACR recommends TL spine imaging in patients with blunt trauma with mechanisms of injury such as motor vehicle crash at greater than 35 mph, falls of greater than 15 feet, automobile hitting pedestrian with pedestrian thrown more than 10 feet, assaulted with a depressed level of consciousness. It is also recommended in patients with known cervical injury and rigid spine disease (e.g., AS, DISH).

IMAGING MODALITIES According to the ACR appropriateness criteria, blunt trauma of the TL spine warrants evaluation with CT whether or not there are localizing signs. In patients with neurologic abnormalities, CT and MRI are complementary examinations and both should be performed. Radiography is the imaging procedure of choice in children aged 14 years or younger. In the cervical region recommended views are AP, lateral, and open-mouth; in the thoracic and lumbar regions, AP and lateral only. Radiography has limited use in adults and should be used primarily for resolving nondiagnostic CT studies due to motion artifacts.

Flexion-extension radiography is not useful in the acute injury period because of muscle spasm. Multidetector-row CT is the procedure of choice for evaluating for possible thoracic and/or lumbar injuries. In patients who undergo CT of the chest, abdomen, and pelvis, the images will be adequate to evaluate the spine. Sagittal and coronal reconstructions should be performed. Because the incidence of multiple noncontiguous fractures is as high as 25% of such patients, the panel recommends imaging of the entire spine when there are known fractures in any segment.45,103 MRI is recommended, especially in patients with neurologic deficits, in the evaluation of the disk, ligaments, and neural elements. In other injuries, principally TL burst fractures, MRI has been used to obtain information about the status of the posterior ligamentous complex (PLC), a critical determinant of surgical decision making. Disks and ligaments play important roles in fracture stability, because failure of the posterior tension band may lead to progressive kyphosis. MRI can now be considered key to accurate fracture classification, detection of occult and distant injuries, and as the basis for therapeutic decision making. It is therefore strongly recommended in the early assessment of spine trauma patients.79

INCIDENCE Thoracic and lumbar spinal fractures are commonly encountered in blunt trauma patients. Approximately 50% of all vertebral fractures occur in the TL spine, and the incidence of TL fractures in trauma patients presenting to level 1 trauma centers is 4% to 5%. Neurologic injury to the spinal cord occurs in 19% to 50% of these patients, and a delay in diagnosis of TL spine fractures can result in up to an eightfold increase in neurologic deficits.94

ANATOMY A spinal motion segment is the principal functional unit of the spine and consists of two vertebrae and the interconnecting soft tissues.  The anterior portion of the functional unit contains two aligned vertebral bodies, the intervertebral disk, and the anterior and posterior longitudinal ligaments. The posterior portion consists of the vertebral arches, facet joints, and posterior elements. The coronal orientation of the thoracic facet joints minimizes extension but allows rotation, whereas the sagittal oblique orientation of the lumbar articular facets

http://radiologyebook.com

810

PART II  CT and MR Imaging of the Whole Body

Anterior

Left

Right

Posterior

A

0

Non displaced fracture

Displaced 1–3 mm

1

2

1

2

Displaced 3–5 mm 3

4

3

4

Displaced more than 5 mm 5

Bony Ligamentous 0

Mild 1–3 mm

Moderate 3–5 mm

5 Severe more than 5 mm

B FIG 28-21  Cervical Spine Injury Severity Score (CSISS). A, Diagram of cervical spine. The spine is divided into four columns: anterior, posterior, right pillar, and left pillar. B, CSISS classification. (Adapted from Anderson PA, et al: Cervical spine injury severity score. Assessment of reliability. J Bone Joint Surg Am 89:1057– 1065, 2007.)

minimizes rotation. The posterior portion of the motion segment contains the PLC, which is composed of the supraspinous ligament, interspinous ligaments, articular facet capsules, and ligamentum flavum.48 The ribcage protects the thoracic spine and limits its mobility in comparison with the cervical and lumbar spine. The TL junction  is a point of high kinetic energy transfer and often results in TL fractures.33

CLASSIFICATION Numerous TL classification systems have been developed to guide clinical and surgical treatment. Early classification systems were based on anatomic structures or inferred mechanism of injuries.33,55,57,82,90 The AOSpine (Arbeitsgemeinschaft für Osteosynthesefragen) classification, introduced by Magerl in 1994, is one of the most systematic and detailed TL spine classification systems; however, it has low interobserver reliability and is less useful in therapeutic decision making and prognostic purposes.1,63 The AOSpine Trauma Knowledge Forum, an international group of academic spine surgeons, has recently modified the AO classification to include fracture morphology and

clinical factors relevant to surgical decision making, such as the presence of neurologic deficit85,110 (Fig. 28-22).

AOSpine Thoracolumbar Spine Injury Classification System Morphology Classification

Type A. Compression injuries A0. No fracture of the vertebra or insignificant fractures of the spinous or transverse processes A1. Wedge compression of impaction fractures with fracture of a single end plate without involvement of the posterior wall of the vertebral body (Fig. 28-23) A2. Split- or pincer-type fractures in which the fracture line involves both end plates but does not involve the posterior vertebral wall (Fig. 28-24) A3. Vertebral fractures affecting a single end plate with any involvement of the posterior vertebral wall and spinal canal A4. Vertebral body fractures involving both end plates as well as the posterior wall. Split fractures that also involve the posterior vertebral body are included in this group (Fig. 28-25).

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

A

B

811

C

FIG 28-22  AOSpine Thoracolumbar Spine Injury Classification. Morphologic classification of the fractures. Type A: compression injuries. Failure of anterior structures under compression with intact tension band. Type B: failure of the posterior or anterior tension band. Type C: failure of all elements, leading to dislocation or translation. (Adapted from Vaccaro AR, et al: AOSpine thoracolumbar spine injury classification system: Fracture description, neurological status, and key modifiers. Spine [Phila Pa 1976] 38:2028–2037, 2013.)

B

A

FIG 28-23  Compression injuries. A, CT sagittal reformatted image demonstrates superior end plate fracture in a thoracic vertebral body without loss of height (arrow). B, Sagittal fat-suppressed T2-weighted MRI of same patient demonstrates superior end plate fractures and bone marrow edema in three consecutive vertebral bodies. This corresponds to type A1 compression fracture in the AOSpine Thoracolumbar Spine Injury Classification, and Compression morphology in the TLICSS.

Type B. Injury affecting either anterior or posterior tension band; may be seen in combination with type A fractures of the vertebral body B1. Monosegmental osseous failure of the posterior tension band extending into the vertebral body, known as Chance fractures (Fig. 28-26) B2. Disruption of the posterior tension band with or without osseous involvement (Fig. 28-27) B3. Disruption of the anterior longitudinal ligament, which serves as the anterior tension band of the spine, preventing hyperextension. The injury may pass through either the intervertebral

disk or through the vertebral body itself (particularly in the ankylosed spine), but there is an intact posterior element hinge preventing gross displacement (Fig. 28-28). Type C. Injuries characterized by displacement beyond the physiologic range of the cranial and caudal parts of the spinal column in any plane (Fig. 28-29). Any associated vertebral body fracture and/or tension band injury should be specified separately (e.g., A0, A1, A2, A3, A4, B1, B2, B3). In the AOSpine TL classification system, type A injuries may affect a single vertebral body in isolation or occur in combination with type

http://radiologyebook.com

812

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 28-24  Compression injuries. Sagittal CT reformatted image (A) and axial CT image (B) of the thoracic spine. There is a comminuted fracture of the T7 vertebral body, with coronal and sagittal fracture lines not involving the posterior wall. This corresponds to an A2 split fracture in the AOSpine Thoracolumbar Spine Injury Classification, and Compression morphology in the TLICSS.

C

A

B

D

FIG 28-25  Compression injuries. Coronal CT reformatted image (A), sagittal CT reformatted image (B), and axial CT images (C and D) of the thoracic and lumbar spine demonstrate fractures of the T11 and T12 vertebral bodies. At T11 the fracture affects the inferior end plate and the posterior wall but not the superior end plate and corresponds to type A3 and Compression morphology in the AOSpine Thoracolumbar Spine Injury Classification and TLICSS, respectively. The T12 fracture affects both end plates and the posterior wall and therefore corresponds to type A4 in the AOSpine Thoracolumbar Spine Injury Classification, and Burst morphology in the TLICSS.

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

A

B

C

FIG 28-26  Tension band/distraction injuries. A-C, Sagittal CT reformatted images demonstrate a compression fracture of the L2 vertebral body as well as fractures with distraction through the pars interarticularis on one side and through the pedicle on the other side. This is the classic Chance fracture, which corresponds to type B1 tension band injury in the AOSpine Thoracolumbar Spine Injury Classification, and Distraction morphology in the TLICSS.

A

B

C

FIG 28-27  Tension band/distraction injuries. A-C, Sagittal CT reformatted images of the lumbar spine in a pediatric patient. There is a Chance-type injury through the L2-L3 level, with avulsion of the L3 superior end plate (arrow in C), widening of the intervertebral disk, and subluxation of the facet joints bilaterally. There is disruption of the posterior tension band, resulting in two-segment distraction, conforming the B2 type of posterior band injury in the AOSpine Thoracolumbar Spine Injury Classification, and Distraction morphology in the TLICSS.

http://radiologyebook.com

813

814

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 28-28  Tension band/distraction injuries. Sagittal CT reformatted image (A) and sagittal T2-weighted MRI (B). There is mild widening of the T8-T9 intervertebral disk space anteriorly as a result of disruption of the anterior longitudinal ligament and intervertebral disk. This injury corresponds to type B3 anterior tension band injury in the AOSpine Thoracolumbar Spine Injury Classification, and Distraction morphology in the TLICSS.

A

B

C

FIG 28-29  Displacement/translation injuries. Sagittal CT reformatted image (A), axial CT image (B), and sagittal T2-weighted MRI (C). Translation injury at the T7-T8 level is demonstrated. A fragment of the T8 vertebral body is rotated and posteriorly displaced into the spinal canal. There is anterolisthesis of T7 on T8. The anterior and posterior tension bands are disrupted. This injury corresponds to type C translational injury in the AOSpine Thoracolumbar Spine Injury Classification, and Translation/Rotation morphology in the TLICSS.

http://radiologyebook.com

815

CHAPTER 28  Spinal Trauma B or C injuries. Type B2, B3, and C injuries affect a motion segment and are coded accordingly by motion segment (e.g., T12-L1), whereas type A and B1 injuries are coded by the single vertebral level they affect (e.g., L2). Multilevel injuries should be classified separately and listed according to declining severity.

Grading of Neurologic Deficit N0. Neurologically intact N1. Transient neurologic deficit, no longer present N2. Symptoms or signs of radiculopathy N3. Incomplete spinal cord injury or cauda equina injury N4. Complete spinal cord injury (ASIA grade A)

TABLE 28-1  Thoracolumbar Injury

Case-Specific Modifiers M1. Fractures with an indeterminate injury to the tension band based on spinal imaging such as MRI or clinical examination M2. Patient-specific comorbidity, which might argue either for or against surgery for those patients with relative indications for surgery. Examples of an M2 modifier include disorders such as (but not limited to) AS, rheumatologic conditions, DISH, osteopenia/ osteoporosis, or burns affecting the skin overlying the injured spine.

Thoracolumbar Injury Classification Scale The Spine Trauma Study Group, led by Alexander Vaccaro, redefined stability of the TL spine to take into account the immediate mechanical stability (suggested by the morphology of injury), long-term stability (indicated by integrity of the PLC), and neurologic stability (indicated by the presence or absence of a deficit). Based on this definition, they developed the thoracolumbar injury severity scale (TLISS).14,45,53,76,109 This system was modified in 2005 based on observations from the validation study and further analysis. The modification is referred  to as the thoracolumbar injury classification severity scale (TLICSS; Table 28-1).38

Thoracolumbar Spine Injury Classification and Severity Scale Morphology of Injury (Fig. 28-30)

Compression (value point 1). Compression injuries are caused by axial loading or lateral flexion and are defined by imaging as variable loss

A

of vertebral body height or disruption of the vertebral end plate. Less severe injuries affect only the anterior portion of the vertebral body. The revised TLICSS removed the lateral compression subcategory of compression morphology. Compression and burst fractures may be deemed unstable in the event of PLC injury54 (see Figs. 28-23 and 28-24). Burst (value point 2). Axial loading or lateral flexion injuries with increased force result in burst morphology with involvement of the posterior body and varying degrees of retropulsion (see Fig. 28-25).

Classification Severity Scale (TLICSS) Injury Category

Value Point

Morphology of Injury Compression Burst Translation or rotation Distraction

1 2 3 4

Integrity of the Posterior Ligamentous Complex Intact Injury suspected or indeterminate Injured

0 2 3

Neurologic Status Intact Nerve root involvement Spinal cord or conus medullaris injury incomplete Spinal cord or conus medullaris injury complete Cauda equina syndrome

0 2 3 2 3

From Harrop JS, et al: Intrarater and interrater reliability and validity in the assessment of the mechanism of injury and integrity of the posterior ligamentous complex: A novel injury severity scoring system for thoracolumbar injuries. Invited submission from the Joint Section Meeting on Disorders of the Spine and Peripheral Nerves, March 2005. J Neurosurg Spine 4:118–122, 2006.

B

C

FIG 28-30  Thoracolumbar Injury Classification and Severity Score (TLICSS). A, Compression morphology. The vertebral body buckles under load to produce a compression or burst fracture. B, Translation/rotation morphology. The vertebral column is subjected to shear or torsional forces that cause the rostral part of the spinal column to translate or rotate with respect to the caudal part. C, Distraction morphology. The rostral spinal column becomes separated from the caudal segment due to distractive forces. Combinations of these morphologic patterns may occur.

http://radiologyebook.com

816

PART II  CT and MR Imaging of the Whole Body

Translation or rotation (value point 3). Torsional and shear forces cause horizontal displacement or rotation of one vertebral body with respect to another. Anatomically the TL spine is configured to move in flexion and extension but to resist significant rotation and translation. Hence failure from torsion and/or shear requires considerably more destruction of normal anatomy and imparts considerably more instability than failure from compression.109 Imaging in the AP plane demonstrates horizontal separation of the spinous processes or altered alignment of the pedicles above and below. The sagittal plane better depicts the facet fractures and dislocations. In the axial plane, shifts across the sagittal plane can be observed across the injury site (see Fig. 28-28). Distraction (value point 4). In distraction injuries one part of the spinal column is separated from the other, leaving a space in between. This can occur through disruption of anterior or posterior ligaments, through anterior or posterior bony elements, or a combination of both. These are often very unstable injuries because by definition, the spinal column is disrupted circumferentially. Angulation is frequently seen in the sagittal and/or coronal planes across the fracture site. The distraction morphology is seen in flexion distraction injuries (as in a Chance fracture) and in distraction extension injuries (as in patients with DISH or AS). Flexion distraction mechanism and PLC injury should be suspected if an inferior or superior posterior end plate fracture is seen, because it reflects avulsion fracture from the comparatively strong anulus fibrosus of the vertebral disk (see Figs. 28-26 through 28-28).

Integrity of the Posterior Ligamentous Complex.  In the TLICSS the PLC includes the supraspinous ligament, interspinous ligament, ligamentum flavum, and facet joint capsules. Intact (value point 0) Injury suspected or indeterminate (value point 2) Injured (value point 3). Injuries to the disk, anterior longitudinal ligament, and posterior longitudinal ligament are not scored as a part of the PLC but taken into account in the analysis of the morphology of the injury.

Neurologic Status Intact (value point 0) Nerve root involvement (value point 2) Spinal cord or conus medullaris injury incomplete (value point 3) Spinal cord or conus medullaris injury complete (value point 2) Cauda equina syndrome (value point 3) Patients with incomplete spinal cord injury or cauda equina syndrome may receive greater potential benefit from surgical decompression than patients with complete spinal cord injury or no initial neurologic injury. The composite injury severity score derived from this classification system assigns between 1 and 4 points to three critical components of an injury. Fractures with 3 points or less are considered nonoperative cases. Fractures with scores of 4 points can be considered for nonoperative or operative intervention. Fractures with 5 or greater points are considered surgical cases. In operative candidates, features of this classification system such as posterior ligamentous integrity and the neurologic status of the patient serve to direct the optimal surgical approach.

EVALUATION OF THE LIGAMENTOUS STRUCTURES The PLC protects the spine against excessive flexion, extension, rotation, translation, and distraction. Owing to its poor healing ability,

once disrupted it generally requires surgical intervention. PLC injury is a risk factor for late kyphosis after burst fractures.83 In plain radiographs and CT, injuries to the PLC are inferred by indirect signs such as widening of the interspinous space, avulsion fracture of the superior or inferior aspects of contiguous spinous processes, facet joint widening, subluxation, or dislocation (i.e., perched or locked facets), and vertebral body translation or rotation.48 In a study by Radcliff et al., translation greater than 3.5 mm is associated with PLC injury, whereas loss of vertebral body height and kyphotic angulation greater than 20 degrees are not predictive of PLC injury in TL burst fractures.83 The reported diagnostic accuracy of MRI for the evaluation of ligamentous injury is variable. In general it tends to demonstrate relatively high negative predictive values and relatively low positive predictive values for PLC injuries.86,111,112 According to Dai et al., MRI is reliable for detecting ligamentous injury, especially PLC injury in TL burst fractures, but the ligamentous injury as shown by MRI is not correlated with neurologic function or fracture severity.23 MRI shows higher sensitivity, specificity, and accuracy in distinguishing ligamentous lesions versus CT and may reduce the risk of failure to diagnose a PLC injury and associated late deformity.54 MRI signs of PLC injury include disruption of the low-signalintensity black stripe on sagittal T1- or T2-weighted images (indicative of supraspinatus ligament or ligamentum flavum tear), fluid in the facet joint capsules, and edema in the interspinous region. In a study by Lee et al. in a setting of normal-appearing plain radiographs, PLC injury as displayed on T1-weighted MRI, posterior edema (high signal intensity) in the region of PLC elements on T2 short tau inversion recovery (STIR) or FAT SAT (fat saturation) sagittal MRI, and diastasis of the facet joints on CT scan were considered the most popular  determinants of probable PLC injury on imaging studies54 (Figs. 28-31 and 28-32).

EVALUATION OF CORD INJURY Acute traumatic spinal cord injury (SCI) affects 12,000 to 15,000 people in North America each year. The functional consequences of an acute SCI are variable. Patients with incomplete SCI have some chance of neurologic recovery; the prognosis is far less optimistic for patients with complete injuries. The 2015 American Spinal Injury Association (ASIA) Standards is used for the neurologic assessment of adult patients with acute SCI (Fig. 28-33). MRI is often helpful in determining the location and severity of neurologic compromise and identifying injury to non-bony structures.110 SCI findings can be subdivided into those that relate to extrinsic compression and those that traumatize the cord parenchyma itself. Fractures with bony retropulsion, disk extrusions, and epidural hematomas can result in cord compression. Preexisting spondylosis and spinal stenosis will accentuate the vulnerability of the cord to extrinsic compression (Fig. 28-34). The extent of maximum spinal canal compression is more reliable than presence of canal stenosis for predicting the neurologic outcome after SCI.30,69 The spectrum of cord parenchymal injuries include edema (low T1 and high T2 signal), swelling (defined as a smooth enlargement of the cord contour), and hemorrhage (Fig. 28-35). Acute intramedullary hemorrhage is seen as a focus of T2 shortening (hypointensity).  Cervical cord hematomas have a strong correlation with a complete neurologic deficit: ASIA A, with the highest lesions carrying an increased risk of fatality (especially at the cervicomedullary junction in the setting of occipitoatlantal dislocation).69 Spinal cord edema increases significantly during the early time period after injury, whereas intramedullary hemorrhage is comparatively static. At any level the

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

A

B

C

FIG 28-31  PLC injury. A, Sagittal T2-weighted MRI of the thoracic spine demonstrates PLC injury. Note the disruption of the black line that corresponds to the ligamentum flavum (arrow). B, Widening and increased T2 signal in the interspinous ligament (arrow). C, Disruption of the black line that corresponds to the supraspinous ligament (arrow).

A

B FIG 28-32  PLC injury. A, Axial CT demonstrates bilateral dislocation of the facet joints. B, Axial T2-weighted MRI on a different patient demonstrates increased T2 signal consistent with disruption of the ligamentum flavum on the right (arrow).

http://radiologyebook.com

817

http://radiologyebook.com

Spinal Injury Association, Atlanta, GA.)

FIG 28-33  American Spinal Injury Association International Standards for Neurological Classification of Spinal Cord Injury, revised 2013, updated 2015. (Courtesy American

818 PART II  CT and MR Imaging of the Whole Body

819

FIG 28-33, cont’d

CHAPTER 28  Spinal Trauma

http://radiologyebook.com

820

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 28-34  Extrinsic findings contributing to cord injury. Sagittal T2-weighted MRIs in two different patients. A, Disk osteophyte complex at C4-C5 causes narrowing of the spinal canal and predisposes this patient to cord contusion even with minor trauma. B, Acute disk protrusion at C5-C6 associated to distraction of the posterior elements, with cord swelling from C4 to C6 and cord edema above and below those levels.

A

B

C

FIG 28-35  Spectrum of cord parenchymal injuries. Sagittal T2-weighted MRIs in three different patients. A, Swelling of the spinal cord at the level of C6. B, Extensive swelling of the cord from C2 to C6, with multiple punctate areas of low signal intensity corresponding to parenchymal hemorrhage. There is posterior translation of C4 with respect to C5. C, Distraction injury in the thoracic spine. There is complete transection of the spinal cord.

cord is vulnerable to transection if the applied forces are sufficient58 (see Fig. 28-35C). In the subacute stage, cord injuries evolve to myelomalacia, with areas of T2 hyperintensity and T1 signal intermediate between the cord and cerebrospinal fluid (CSF). Progressive posttraumatic myelomalacic myelopathy (PTMM) is a poorly understood entity described in

the late subacute period (generally > 2 months post injury). It likely represents a presyrinx state following disruption of normal transparenchymal CSF transit. A contributory role for cord tethering by adhesions has also been proposed, which may be amenable to intraoperative lysis.29 Chronic posttraumatic cord cysts or syringes (Fig. 28-36) are identified as expansile structures that are isointense with CSF on all

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

FIG 28-36  Posttraumatic syrinx. Sagittal T2-weighted MRI. Kyphotic deformity in the thoracic spine due to old compression fracture with posterior distraction. An extensive complex syrinx replaces the cord parenchyma proximal to the level of the injury. There is atrophy of the cord distally.

821

and that are more sharply marginated compared with myelomalacia. Cord atrophy is a common finding in patients imaged more than 20 years after the initial trauma. Atrophy has been defined as an AP dimension of 7 mm or less in the cervical cord and 6 mm or less in the thoracic region.34,80 In 1982 Pang and Wilberger introduced the concept of spinal cord injury without radiographic abnormality (SCIWORA), originally described in the pediatric population as the presence of clinical findings (neurologic deficit) in the setting of normal x-rays and CT. The advent of MRI permitted characterization of the underlying spinal cord injury even when radiographs and CT scan were normal. In view of these advances, Pang reviewed his work in 2004 and advocated that only cases with neural injuries seen on MRI or those with normal MRI findings should be counted as SCIWORA, and purely extraneural compressive lesions should be excluded from the definition of SCIWORA.74 Children diagnosed with SCIWORA but with normal MRI findings have shown more favorable clinical outcomes than those with cervical cord abnormalities on MRI.64 In adults with SCIWORA, parenchymal spinal cord injury is the single most important determinant in the long-term outcome. Cord hemorrhage has the worst prognosis, cord edema has the best. Longitudinal signal extension and associated extraneural injuries are also associated with poorer outcomes. Cases with purely neural injuries can be managed conservatively, but associated extraneural injuries, especially disk prolapse and ligamentous instability, warrant surgical management.92

THORACOLUMBAR FRACTURES IN PATIENTS WITH ANKYLOSING SPONDYLITIS TL fractures are significantly less common than cervical fractures in patients with AS (Fig 28-37). A majority of these fractures occur at the

A

B

C

FIG 28-37  Atypical injuries in patients with ankylosing spondylitis (AS). A, CT sagittal reformatted image of the thoracic spine. Retrolisthesis of the T8 vertebral body is due to disruption of the intervertebral disk and anterior and posterior longitudinal ligaments. There is also translation of the posterior elements. Note calcification of the anterior longitudinal ligament in this patient with AS. B and C, CT sagittal reformatted images of the cervical spine in a different patient. Widening of the C6-C7 disk with pseudoarthrosis of the facet joint (arrow) in a patient with AS.

http://radiologyebook.com

822

PART II  CT and MR Imaging of the Whole Body

TL junction. Trent et al. originally classified these fractures into  three types: shearing injury (distractive flexion injury or distractive extension injury), wedge compression, or pseudarthrosis from chronic nonunion. Typically the shearing injuries are seen acutely; compression fractures generally have a chronic presentation, whereas pseudarthrosis can be seen subacutely after a missed fracture or in patients with microfractures leading to fibrosis. Most of these injuries are highly unstable and often require surgical treatment. The forces acting across the fracture site are drastically increased by the long lever  arms of the fused thoracic and lumbar spine segments. This is further potentiated at the TL junction owing to the sheer weight of the  thorax above.20

SACRAL INSUFFICIENCY FRACTURES Insufficiency fractures are a subtype of stress fracture that results from normal stress applied to abnormal bone that has lost its elastic resistance. Bone insufficiency is often the result of osteoporosis, metabolic bone disease, osseous metastatic disease, and marrow replacement. Sacral insufficiency fractures (SIFs) are a common cause of debilitating back pain in the elderly. Antecedent trauma is not identified in two thirds of patients and when present is usually minor. The standard of care for the treatment of SIFs has been conservative management; more recently, sacroplasty has been advocated as an alternative to conservative therapy. SIFs most commonly involve the sacral ala lateral to the neural foramina and medial to the sacroiliac joints (zone 1). Fractures may be unilateral or bilateral and are reported with relatively equal frequency in the literature. There may also be a horizontal component to the fracture through the sacral bodies. This unique fracture pattern may be related to axial loading and weight bearing transmitted through the spine, resulting in sacral alar strain. Additionally, osteoporosis causes asymmetric loss of bony trabeculae in the sacral ala compared with the vertebral bodies, placing the lateral aspect of the sacrum at increased risk of insufficiency. On plain films SIFs appear as vertical bands of sclerosis oriented parallel to the sacroiliac joints. Cortical disruption and/or fracture lines may be evident. Occasionally fractures can have an aggressive appearance, simulating malignancy, with areas of sclerosis and periosteal reaction. Bone scintigraphy with technetium 99m–labeled methylenediphosphonate (99mTc MDP) is one of the most sensitive examinations for the detection of SIFs. Bone scintigraphy has a reported sensitivity of 96% for detection of SIFs, with a positive predictive value of 92%. Various patterns of radiotracer uptake have been described, with the so-called Honda sign, or H pattern, considered diagnostic of SIFs in the correct clinical setting. However, this pattern of radiopharmaceutical uptake is seen only in 20% to 40% of patients. Variations in the pattern of radiopharmaceutical activity in SIFs include uptake oriented unilaterally in the sacral ala, unilaterally with a horizontal strut, bilaterally without a horizontal strut, and as multiple foci of activity. Posterior planar images are the most sensitive (i.e., when the sacrum is closest to the detector). Follow-up scintigraphic results are quite variable, with changes in the radiopharmaceutical uptake pattern during a 10- to 33-month interval ranging from resolution of abnormal activity to no change or even worsening activity. CT may demonstrate sclerosis in the sacral ala and lateral sacrum paralleling the sacroiliac joints (Fig. 28-38) or fracture lines with or without bony callous. Coronal reformatted images are useful to detect the horizontal component. CT may also be useful to determine whether the fracture lines extend into the neural foramina, creating a potential pathway for cement if sacroplasty is being considered as a treatment option. CT is not as sensitive for detecting SIFs when compared with

FIG 28-38  Sacral insufficiency fracture. Axial CT image. Note the subtle lucencies oriented in the sagittal plane in the sacral ala (arrows) secondary to insufficiency fracture of the sacrum in this patient with unexplained back pain.

bone scintigraphy or MRI, with a reported sensitivity between 60% and 75%. MRI can detect early changes of sacral insufficiency and, similar to bone scintigraphy, has a reported sensitivity at or near 100%. Marrow edema is demonstrated as areas of increased signal intensity on T2-weighted and STIR images and low signal intensity on T1-weighted images. A hypointense fracture line is usually evident within the area of edema, although it is not seen in 7% of cases. Coronal oblique images in the plane of the sacrum better demonstrate the vertically oriented fractures and should be included in the imaging protocol if there is clinical suspicion. MRI can usually differentiate marrow edema secondary to SIFs from malignancy, with fat-saturation and postgadolinium imaging being particularly useful in this regard. Occasionally MRI can be confusing, especially if a fracture line is not evident, and correlative CT or follow-up imaging may be useful.61

CHRONIC CONDITIONS ATTRIBUTED TO TRAUMA OF THE THORACOLUMBAR SPINE Cord Herniation Spinal cord herniation is an unusual condition and can be defined as protrusion of the spinal cord beyond its dural sleeve. Patients present with variable symptoms or cord dysfunction, including Brown-Séquard syndrome and spastic paralysis. Although progression of neurologic deficits can sometimes be very slow, reduction of the spinal cord and repair of the defect are crucial in stopping or reversing neurologic deterioration.89 Congenital malformation of the meninges, trauma, and more recently trauma to the anterior surface of the thoracic dura by disk protrusions or osteophytes have been proposed as possible etiologies of spinal cord herniation.17 MRI findings of spinal cord herniation on axial images include protrusion through the dura in an anterolateral or anterior position. On sagittal images an anterior C-shaped kink of cord and secondary expansion of the dorsal subarachnoid space can be seen (Fig. 28-39). Cord deviation is generally limited to one or two thoracic spine

http://radiologyebook.com

CHAPTER 28  Spinal Trauma

823

B

A

C

FIG 28-39  Cord herniation. A, Sagittal T2-weighted MRI of thoracic spine demonstrated anterior C-shaped kink of the cord and expansion of the posterior subarachnoid space. B and C, Axial images demonstrate anterior displacement and flattening of the cord (arrows).

segments. An extradural masslike area of signal intensity similar to that of the spinal cord may be observed occasionally and represents the herniated spinal cord. The dural defect occurs in the thoracic spine, most commonly between the levels of the T4 and T7 vertebrae. Associated cord atrophy and high T2 signal intensity may be observed within the thoracic cord. Occasionally scalloping of the vertebral body also may be seen.75

REFERENCES 1. Aebi M: Classification of thoracolumbar fractures and dislocations.   Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 19(Suppl 1):S2–S7, 2010. 2. Ahuja A, Glasauer FE, Alker GJ, Jr, et al: Radiology in survivors of traumatic atlanto-occipital dislocation. Surg Neurol 41:112–118, 1994. 3. Allen BL, Jr, Ferguson RL, Lehmann TR, et al: A mechanistic classification of closed, indirect fractures and dislocations of the lower cervical spine. Spine 7:1–27, 1982. 4. Anderson LD, D’Alonzo RT: Fractures of the odontoid process of the axis. J Bone Joint Surg Am 56:1663–1674, 1974. 5. Anderson PA, Montesano PX: Morphology and treatment of occipital condyle fractures. Spine 13:731–736, 1988. 6. Anderson PA, Moore TA, Davis KW, et al: Cervical spine injury severity score. Assessment of reliability. J Bone Joint Surg Am 89:1057–1065, 2007. 7. Anderson PA, Muchow RD, Munoz A, et al: Clearance of the asymptomatic cervical spine: A meta-analysis. J Orthop Trauma Febr 24:100–106, 2010.

8. Begemann PGC, Kemper J, Gatzka C, et al: Value of multiplanar reformations (MPR) in multidetector CT (MDCT) of acute vertebral fractures: Do we still have to read the transverse images? J Comput Assist Tomogr 28:572–580, 2004. 9. Reference deleted in proofs. Duplicate of Reference 8. 10. Bilston LE, Clarke EC, Brown J: Spinal injury in car crashes: Crash factors and the effects of occupant age. Inj Prev 17:228–232, 2011. 11. Blackmore CC: Evidence-based imaging evaluation of the cervical spine in trauma. Neuroimaging Clin N Am 13:283–291, 2003. 12. Bogner EA: Imaging of Cervical Spine Injuries in Athletes. Sports Health 1:384–391, 2009. 13. Bono CM, Schoenfeld AJ, Anderson PA, et al: Observer variability of radiographic measurements of C2 (axis) fractures. Spine 35:1206–1210, 2010. 14. Bono CM, Vaccaro AR, Hurlbert RJ, et al: Validating a newly proposed classification system for thoracolumbar spine trauma: Looking to the future of the thoracolumbar injury classification and severity score.   J Orthop Trauma 20:567–572, 2006. 15. Bracken MB, Freeman DH, Jr, Hellenbrand K: Incidence of acute traumatic hospitalized spinal cord injury in the United States, 1970-1977. Am J Epidemiol 113:615–622, 1981. 16. Brady WJ, Moghtader J, Cutcher D, et al: ED use of flexion-extension cervical spine radiography in the evaluation of blunt trauma. Am J Emerg Med 17:504–508, 1999. 17. Brus-Ramer M, Dillon WP: Idiopathic Thoracic Spinal Cord Herniation: Retrospective Analysis Supporting a Mechanism of Diskogenic Dural Injury and Subsequent Tamponade. Am J Neuroradiol 33:52–56, 2012.

http://radiologyebook.com

824

PART II  CT and MR Imaging of the Whole Body

18. Bucholz RW, Burkhead WZ: The pathological anatomy of fatal atlanto-occipital dislocations. J Bone Joint Surg Am 61:248–250, 1979. 19. Burney RE, Maio RF, Maynard F, et al: Incidence, characteristics, and outcome of spinal cord injury at trauma centers in North America. Arch Surg Chic Ill 1960 128:596–599, 1993. 20. Chaudhary SB, Hullinger H, Vives MJ: Management of Acute Spinal Fractures in Ankylosing Spondylitis. ISRN Rheumatol 2011, 2011. . Accessed 2014 Mar 13. 21. Coric D, Wilson JA, Kelly DL, Jr: Treatment of traumatic spondylolisthesis of the axis with nonrigid immobilization: A review of 64 cases. J Neurosurg 85:550–554, 1996. 22. Daffner RH, Hackney DB: ACR Appropriateness criteria on suspected spine trauma. J Am Coll Radiol JACR 4:762–775, 2007. 23. Dai L-Y, Ding W-G, Wang X-Y, et al: Assessment of ligamentous injury in patients with thoracolumbar burst fractures using MRI. J Trauma 66:1610–1615, 2009. 24. Demetriades D, Charalambides K, Chahwan S, et al: Nonskeletal cervical spine injuries: Epidemiology and diagnostic pitfalls. J Trauma 48:724–727, 2000. 25. Dickman CA, Greene KA, Sonntag VK: Injuries involving the transverse atlantal ligament: Classification and treatment guidelines based upon experience with 39 injuries. Neurosurgery 38:44–50, 1996. 26. Ding T, Maltenfort M, Yang H, et al: Correlation of C2 fractures and vertebral artery injury. Spine 35:E520–E524, 2010. 27. Dziurzynski K, Anderson PA, Bean DB, et al: A blinded assessment of radiographic criteria for atlanto-occipital dislocation. Spine 30:1427– 1432, 2005. 28. Esses S, Langer F, Gross A: Fracture of the atlas associated with fracture of the odontoid process. Injury 12:310–312, 1981. 29. Falcone S, Quencer RM, Green BA, et al: Progressive posttraumatic myelomalacic myelopathy: Imaging and clinical features. Am J Neuroradiol 15:747–754, 1994. 30. Fehlings MG, Furlan JC, Massicotte EM, et al: Interobserver and intraobserver reliability of maximum canal compromise and spinal cord compression for evaluation of acute traumatic cervical spinal cord injury. Spine 31:1719–1725, 2006. 31. Ferro FP, Borgo GD, Letaif OB, et al: Traumatic spondylolisthesis of the axis: Epidemiology, management and outcome. Acta Ortop Bras 20:84–87, 2012. 32. Freemyer B, Knopp R, Piche J, et al: Comparison of five-view and three-view cervical spine series in the evaluation of patients with cervical trauma. Ann Emerg Med 18:818–821, 1989. 33. Ghobrial GM, Jallo J: Thoracolumbar spine trauma: Review of the evidence. J Neurosurg Sci 57:115–122, 2013. 34. Goldberg AL, Kershah SM: Advances in Imaging of Vertebral and Spinal Cord Injury. J Spinal Cord Med 33:105–116, 2010. 35. Grabb BC, Frye TA, Hedlund GL, et al: MRI diagnosis of suspected atlanto-occipital dissociation in childhood. Pediatr Radiol 29:275–281, 1999. 36. Hadley MN, Browner C, Sonntag VK: Axis fractures: A comprehensive review of management and treatment in 107 cases. Neurosurgery 17:281–290, 1985. 37. Harris JH, Jr, Edeiken-Monroe B, Kopaniky DR: A practical classification of acute cervical spine injuries. Orthop Clin North Am 17:15–30, 1986. 38. Harrop JS, Vaccaro AR, Hurlbert RJ, et al: Intrarater and interrater reliability and validity in the assessment of the mechanism of injury and integrity of the posterior ligamentous complex: A novel injury severity scoring system for thoracolumbar injuries. Invited submission from the Joint Section Meeting On Disorders of the Spine and Peripheral Nerves, March 2005. J Neurosurg Spine 4:118–122, 2006. 39. Hoffman JR, Mower WR, Wolfson AB, et al: Validity of a Set of Clinical Criteria to Rule Out Injury to the Cervical Spine in Patients with Blunt Trauma. N Engl J Med 343:94–99, 2000. 40. Hsu JM, Joseph T, Ellis AM: Thoracolumbar fracture in blunt trauma patients: Guidelines for diagnosis and imaging. Injury 34:426–433, 2003.

41. Iizuka H, Iizuka Y, Kobayashi R, et al: Characteristics of idiopathic atlanto-axial subluxation: A comparative radiographic study in patients with an idiopathic etiology and those with rheumatoid arthritis. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 22:54–59, 2013. 42. Insko EK, Gracias VH, Gupta R, et al: Utility of flexion and extension radiographs of the cervical spine in the acute evaluation of blunt trauma. J Trauma 53:426–429, 2002. 43. Jackson RS, Banit DM, Rhyne AL, 3rd, et al: Upper cervical spine injuries. J Am Acad Orthop Surg 10:271–280, 2002. 44. Jain VK: Atlantoaxial dislocation. Neurol India 60:9–17, 2012. 45. Joaquim AF, Patel AA: Thoracolumbar spine trauma: Evaluation and surgical decision-making. J Craniovertebral Junction Spine 4:3–9, 2013. 46. Jones C, Jazayeri F: Evolving standards of practice for cervical spine imaging in trauma: A retrospective review. Australas Radiol 51:420–425, 2007. 47. el-Khoury GY, Kathol MH, Daniel WW: Imaging of acute injuries of the cervical spine: Value of plain radiography, CT, and MR imaging. AJR Am J Roentgenol 164:43–50, 1995. 48. Khurana B, Sheehan SE, Sodickson A, et al: Traumatic thoracolumbar spine injuries: What the spine surgeon wants to know. Radiogr Rev Publ Radiol Soc N Am Inc 33:2031–2046, 2013. 49. Krakenes J, Kaale BR, Rorvik J, et al: MRI assessment of normal ligamentous structures in the craniovertebral junction. Neuroradiology 43:1089–1097, 2001. 50. Kwon BK, Vaccaro AR, Grauer JN, et al: Subaxial cervical spine trauma. J Am Acad Orthop Surg 14:78–89, 2006. 51. Kwong Y, Rao N, Latief K: Craniometric Measurements in the Assessment of Craniovertebral Settling: Are They Still Relevant in the Age of Cross-Sectional Imaging? Am J Roentgenol 196:W421–W425, 2011. 52. Labler L, Eid K, Platz A, et al: Atlanto-occipital dislocation: Four case reports of survival in adults and review of the literature. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 13:172–180, 2004. 53. Lee JY, Vaccaro AR, Lim MR, et al: Thoracolumbar injury classification and severity score: A new paradigm for the treatment of thoracolumbar spine trauma. J Orthop Sci Off J Jpn Orthop Assoc 10:671–675, 2005. 54. Lee JY, Vaccaro AR, Schweitzer KM, Jr, et al: Assessment of injury to the thoracolumbar posterior ligamentous complex in the setting of normal-appearing plain radiography. Spine J Off J North Am Spine Soc 7:422–427, 2007. 55. Lenarz CJ, Place HM, Lenke LG, et al: Comparative reliability of 3 thoracolumbar fracture classification systems. J Spinal Disord Tech 22:422–427, 2009. 56. Lewis LM, Docherty M, Ruoff BE, et al: Flexion-extension views in   the evaluation of cervical-spine injuries. Ann Emerg Med 20:117–121, 1991. 57. Lewkonia P, Paolucci EO, Thomas K: Reliability of the thoracolumbar injury classification and severity score and comparison with the Denis classification for injury to the thoracic and lumbar spine. Spine 37:2161–2167, 2012. 58. Leypold BG, Flanders AE, Burns AS: The early evolution of spinal cord lesions on MR imaging following traumatic spinal cord injury. AJNR Am J Neuroradiol 29:1012–1016, 2008. 59. Li Q, Shen H, Li M: Magnetic resonance imaging signal changes of alar and transverse ligaments not correlated with whiplash-associated disorders: A meta-analysis of case-control studies. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 22:14–20, 2013. 60. Lipson SJ: Fractures of the atlas associated with fractures of the odontoid process and transverse ligament ruptures. J Bone Joint Surg Am 59:940–943, 1977. 61. Lyders EM, Whitlow CT, Baker MD, et al: Imaging and Treatment of Sacral Insufficiency Fractures. Am J Neuroradiol 31:201–210, 2010. 62. MacDonald RL, Schwartz ML, Mirich D, et al: Diagnosis of cervical spine injury in motor vehicle crash victims: How many X-rays are enough? J Trauma 30:392–397, 1990.

http://radiologyebook.com

CHAPTER 28  Spinal Trauma 63. Magerl F, Aebi M, Gertzbein SD, et al: A comprehensive classification of thoracic and lumbar injuries. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 3:184–201, 1994. 64. Mahajan P, Jaffe DM, Olsen CS, et al: Spinal cord injury without radiologic abnormality in children imaged with magnetic resonance imaging. J Trauma Acute Care Surg 75:843–847, 2013. 65. Marcon RM, Cristante AF, Teixeira WJ, et al: Fractures of the cervical spine. Clinics 68:1455–1461, 2013. 66. Reference deleted in proofs. Duplicate of Reference 65. 67. Mathen R, Inaba K, Munera F, et al: Prospective Evaluation of Multislice Computed Tomography Versus Plain Radiographic Cervical Spine Clearance in Trauma Patients. J Trauma-Inj Infect 62:1427–1431, 2007. 68. Michaleff ZA, Maher CG, Verhagen AP, et al: Accuracy of the Canadian C-spine rule and NEXUS to screen for clinically important cervical spine injury in patients following blunt trauma: A systematic review. CMAJ Can Med Assoc J J Assoc Medicale Can 184:E867–E876, 2012. 69. Miyanji F, Furlan JC, Aarabi B, et al: Acute Cervical Traumatic Spinal Cord Injury: MR Imaging Findings Correlated with Neurologic Outcome—Prospective Study with 100 Consecutive Patients. Radiology 243:820–827, 2007. 70. Mower WR, Hoffman JR, Pollack CV, et al For the NEXUS Group: Use of plain radiography to screen for cervical spine injuries. Ann Emerg Med 38:1–7, 2001. 71. Mueller FJ, Kinner B, Rosskopf M, et al: Incidence and outcome of atlanto-occipital dissociation at a level 1 trauma centre: A prospective study of five cases within 5 years. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 22:65–71, 2013. 72. Müller CW, Otte D, Decker S, et al: Vertebral fractures in motor vehicle accidents—A medical and technical analysis of 33,015 injured front-seat occupants. Accid Anal Prev 66C:15–19, 2014. 73. Pal D, Sell P, Grevitt M: Type II odontoid fractures in the elderly: An evidence-based narrative review of management. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 20:195–204, 2011. 74. Pang D: Spinal cord injury without radiographic abnormality in children, 2 decades later. Neurosurgery 55:1325–1342, discussion 1342–1343, 2004. 75. Parmar H, Park P, Brahma B, et al: Imaging of Idiopathic Spinal Cord Herniation. Radiographics 28:511–518, 2008. 76. Patel AA, Dailey A, Brodke DS, et al: Thoracolumbar spine trauma classification: The Thoracolumbar Injury Classification and Severity Score system and case examples. J Neurosurg Spine 10:201–206, 2009. 77. Pintar FA, Yoganandan N, Maiman DJ, et al: Thoracolumbar Spine Fractures in Frontal Impact Crashes. Ann Adv Automot Med Annu Sci Conf 56:277–283, 2012. 78. Pitzen T, Lane C, Goertzen D, et al: Anterior cervical plate fixation: Biomechanical effectiveness as a function of posterior element injury.   J Neurosurg 99:84–90, 2003. 79. Pizones J, Castillo E: Assessment of acute thoracolumbar fractures: Challenges in multidetector computed tomography and added value of emergency MRI. Semin Musculoskelet Radiol 17:389–395, 2013. 80. Potter K, Saifuddin A: Pictorial review: MRI of chronic spinal cord injury. Br J Radiol 76:347–352, 2003. 81. Pourtaheri S, Emami A, Sinha K, et al: The role of magnetic resonance imaging in acute cervical spine fractures. Spine J 2014.
825

85. Reinhold M, Audigé L, Schnake KJ, et al: AO spine injury classification system: A revision proposal for the thoracic and lumbar spine. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 22:2184–2201, 2013. 86. Rihn JA, Yang N, Fisher C, et al: Using magnetic resonance imaging to accurately assess injury to the posterior ligamentous complex of the spine: A prospective comparison of the surgeon and radiologist.   J Neurosurg Spine 12:391–396, 2010. 87. Roy-Camille R, Saillant G, Judet T, et al: Factors of severity in the fractures of the odontoid process (author’s transl). Rev Chir Orthopédique Réparatrice Appar Mot 66:183–186, 1980. 88. Rubin GD: Data explosion: The challenge of multidetector-row CT.   Eur J Radiol 36:74–80, 2000. 89. Sasani M, Ozer AF, Vural M, et al: Idiopathic Spinal Cord Herniation: Case Report and Review of the Literature. J Spinal Cord Med 32:86–94, 2009. 90. Sethi MK, Schoenfeld AJ, Bono CM, et al: The evolution of thoracolumbar injury classification systems. Spine J Off J North Am Spine Soc 9:780–788, 2009. 91. Seybold EA, Bayley JC: Functional outcome of surgically and conservatively managed dens fractures. Spine 23:1837–1845, discussion 1845–1846, 1998. 92. Sharma S, Singh M, Wani IH, et al: Adult Spinal Cord Injury without Radiographic Abnormalities (SCIWORA): Clinical and radiological correlations. J Clin Med Res 0:165–172, 2009. 93. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone Joint Surg Am 52:1017–1024, 1970. 94. Sixta S, Moore FO, Ditillo MF, et al: Screening for thoracolumbar spinal injuries in blunt trauma: An Eastern Association for the Surgery of Trauma practice management guideline. J Trauma Acute Care Surg 73:S326–S332, 2012. 95. Smoker WRK, Khanna G: Imaging the craniocervical junction.   Childs Nerv Syst ChNS Off J Int Soc Pediatr Neurosurg 24:1123–1145, 2008. 96. Spence KF, Jr, Decker S, Sell KW: Bursting atlantal fracture associated with rupture of the transverse ligament. J Bone Joint Surg Am 52:543–549, 1970. 97. Stäbler A, Eck J, Penning R, et al: Cervical spine: Postmortem assessment of accident injuries—Comparison of radiographic, MR imaging, anatomic, and pathologic findings. Radiology 221:340–346, 2001. 98. Steinmetz MP, Mroz TE, Benzel EC: Craniovertebral junction: Biomechanical considerations. Neurosurgery 66:7–12, 2010. 99. Stiell IG, Clement CM, McKnight RD, et al: The Canadian C-Spine Rule versus the NEXUS Low-Risk Criteria in Patients with Trauma.   N Engl J Med 349:2510–2518, 2003. 100. Stiell IG, Wells GA, Vandemheen KL, et al: The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA J Am Med Assoc 286:1841–1848, 2001. 101. Stone AT, Bransford RJ, Lee MJ, et al: Reliability of classification systems for subaxial cervical injuries. Evid-Based Spine-Care J 1:19–26, 2010. 102. Swischuk LE: Anterior displacement of C2 in children: Physiologic or pathologic. Radiology 122:759–763, 1977. 103. Takami M, Nohda K, Sakanaka J, et al: Usefulness of full spine computed tomography in cases of high-energy trauma: A prospective study. Eur J Orthop Surg Traumatol Orthop Traumatol 2013. 104. Theodore N, Aarabi B, Dhall SS, et al: Occipital condyle fractures. Neurosurgery 72(Suppl 2):106–113, 2013. 105. Touger M, Gennis P, Nathanson N, et al: Validity of a decision rule to reduce cervical spine radiography in elderly patients with blunt trauma. Ann Emerg Med 40:287–293, 2002. 106. Traynelis VC, Marano GD, Dunker RO, et al: Traumatic atlantooccipital dislocation. Case report. J Neurosurg 65:863–870, 1986. 107. Tubbs RS, Hallock JD, Radcliff V, et al: Ligaments of the craniocervical junction. J Neurosurg Spine 14:697–709, 2011. 108. Vaccaro AR, Hulbert RJ, Patel AA, et al: The subaxial cervical spine injury classification system: A novel approach to recognize the

http://radiologyebook.com

826

109.

110.

111.

112.

113.

PART II  CT and MR Imaging of the Whole Body importance of morphology, neurology, and integrity of the discoligamentous complex. Spine 32:2365–2374, 2007. Vaccaro AR, Lehman RA, Jr, Hurlbert RJ, et al: A new classification of thoracolumbar injuries: The importance of injury morphology, the integrity of the posterior ligamentous complex, and neurologic status. Spine 30:2325–2333, 2005. Vaccaro AR, Oner C, Kepler CK, et al: AOSpine thoracolumbar spine injury classification system: Fracture description, neurological status, and key modifiers. Spine 38:2028–2037, 2013. Vaccaro AR, Rihn JA, Saravanja D, et al: Injury of the posterior ligamentous complex of the thoracolumbar spine: A prospective evaluation of the diagnostic accuracy of magnetic resonance imaging. Spine 34:E841–E847, 2009. Van Middendorp JJ, Patel AA, Schuetz M, et al: The precision, accuracy and validity of detecting posterior ligamentous complex injuries of the thoracic and lumbar spine: A critical appraisal of the literature. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 22:461–474, 2013. Vetti N, Kråkenes J, Eide GE, et al: Are MRI high-signal changes of alar and transverse ligaments in acute whiplash injury related to outcome? BMC Musculoskelet Disord 11:260, 2010a.

114. Vetti N, Kråkenes J, Eide GE, et al: Are MRI high-signal changes of alar and transverse ligaments in acute whiplash injury related to outcome? BMC Musculoskelet Disord 11:260, 2010b. 115. Volle E, Montazem A: MRI video diagnosis and surgical therapy of soft tissue trauma to the craniocervical junction. Ear Nose Throat J 80:41–44, 46–48, 2001. 116. White AA, 3rd, Panjabi MM: Update on the evaluation of instability of the lower cervical spine. Instr Course Lect 36:513–520, 1987. 117. Wholey MH, Bruwer AJ, Baker HL, Jr: The lateral roentgenogram of the neck; with comments on the atlanto-odontoid-basion relationship. Radiology 71:350–356, 1958. 118. Wolfla CE: Anatomical, biomechanical, and practical considerations in posterior occipitocervical instrumentation. Spine J Off J North Am Spine Soc 6:225S–232S, 2006. 119. Zivot U, Di Maio VJ: Motor vehicle-pedestrian accidents in adults. Relationship between impact speed, injuries, and distance thrown. Am J Forensic Med Pathol 14:185–186, 1993.

http://radiologyebook.com

29  Degenerative Disease Michael K. Brooks, Joseph P. Mazzie, and A. Orlando Ortiz

INTRODUCTION Acute and chronic low back pain is a ubiquitous complaint encountered by primary care physicians throughout the United States, accounting for the second most common symptomatic reason for office visits. Low back pain is the leading cause of disability for patients younger than age 45, resulting in lost time from work and decreased work productivity, with increasing cost expenditures for evaluation and treatment of back pain estimated at $90 billion annually ($4695 per year in 1997 vs. $6096 in 2005); it is thus considered a major public health issue in the United States. The most common reason why patients undergo imaging of the spine is pain. Degenerative/acquired and congenital spinal stenosis, neoplasm, infection, trauma, inflammatory or arthritic processes are all known causes of back pain. However, a majority of patients presenting with neurogenic intermittent claudication and radiculopathy is the result of acquired “degenerative” spondylotic changes as the cause of central canal, subarticular recess, and/or neuroforaminal compromise.63 Acquired degenerative disease is due to a combination of bone, ligament, joint, and disk disease, most commonly affecting the lumbar spine, followed by the cervical spine. Thoracic disk disease is now more commonly recognized with the increased use of magnetic resonance imaging (MRI).39

IMAGING GUIDELINES, TECHNIQUES, AND METHODS Advanced imaging has led to a more accurate understanding of radiographic changes throughout the spine that are now known to occur as part of the normal aging process.9-11,19,30,35,44,50,55,111 However, imaging does allow for accurate assessment of mechanical causes of nerve root compression, offering objective evidence in support of a clinical diagnosis of back pain or radicular symptoms.33 Conversely there is a high prevalence of abnormal computed tomography (CT) and MRI findings in asymptomatic patients.9,41 Degenerative changes involving the intervertebral disks or facet joints have not been proven to correlate with back pain, because many patients without back pain demonstrate zygapophyseal joint osteoarthrosis,11,35,99 disk desiccation, reduced disk height, annular tears, and disk protrusions.7,11,19,49,96 Thus degenerative changes may be coincidental, given the poor association between low back pain and anatomic findings,17,29,41 and when found at imaging are not correlative of a patient’s pain4 or predictive for the development of back pain.12,17 The use of advanced spine imaging has continued to rise rapidly, with MRI often being performed on patients not having received a  trial of conservative therapy as currently recommended. The natural history of disk herniation is spontaneous regression, with extrusions and sequestrations demonstrating a higher trend toward resolu-

tion.4,15,23,25,28,61,67,69 Although multiple sources have confirmed that uncomplicated acute low back pain and/or radiculopathy is a benign self-limited condition that does not warrant imaging, many patients continue to undergo radiographic or advanced imaging evaluations without a clear indication. Improvement in pain and functional status usually occurs in most patients presenting with acute back pain (with or without radiculopathy) within 4 weeks regardless of treatment.22,86,112 It has been shown that routine imaging does not improve clinical outcomes, may expose patients to unnecessary radiation and invasive procedures, and contributes to rising healthcare costs. Owing to the ever-increasing trend in spine imaging and the associated expenditures, multiple medical professional organizations have issued recommendations and practice guidelines to address the overutilization of spine imaging. Updated in 2015, the American College of Radiology (ACR)27 published appropriateness criteria for diagnostic imaging in patients with low back pain presenting with various scenarios, which were in parallel with previous guidelines from the American College of Physicians and the American Pain Society. In summary, the ACR appropriateness criteria recommend lumbar radiography as the initial evaluation for low back pain in patients with recent trauma, osteoporosis, on chronic steroids, patients suspected of vertebral compression fracture, and in the evaluation of young patients for ankylosing spondylitis. However, low back pain complicated by “red flags” may justify the use of CT or MRI. In the setting of major trauma, patients with neurologic compromise, history of malignancy, and patients at risk for vertebral infection for which clinical decision making may be directly influenced, advanced imaging should be performed. Furthermore, for low back pain complicated by radicular symptoms lasting longer than 6 weeks following a trial of conservative therapy, MRI should be performed, especially in patients who may be candidates for surgical intervention or percutaneous epidural steroid injection. As can be expected, the challenge for the referring physician is to distinguish this smaller group of patients who may require further workup and advanced imaging because of suspicion of a more serious underlying process.27 Multiple imaging modalities have been used in the evaluation of degenerative disk disease, including radiography, diskography, CT, myelography with postmyelographic CT, MRI, and scintigraphy, which contributes to the dilemma of their application.

Radiography As mentioned, radiographs may provide sufficient data in the initial evaluation of low back pain in patients with recent trauma, osteoporosis, or age older than 70. They provide an inexpensive screening tool of the spine, including osseous detail, alignment, stability, and postoperative evaluation of hardware and fusion. Most patients experience symptoms that occur in axially loaded positions such as standing or sitting; however, advanced imaging studies performed in the supine

http://radiologyebook.com

827

828

PART II  CT and MR Imaging of the Whole Body

position do not reproduce the physiologic changes of weight bearing.48 Standing or upright frontal and lateral lumbar spine radiographs are an appropriate initial imaging study in patients with back pain, with dynamic imaging reserved for the presurgical setting.36,48

CT/CT Myelography The advancement of multidetector (MD)CT has allowed for rapid acquisition, with datasets reconstructed into multiple planes without the loss of spatial resolution or additional radiation exposure. CT imaging provides superior detail of cortical and trabecular bone  compared with MRI. Similarly, MDCT can demonstrate ligamentous ossification, spondylosis, spondylolysis, facet joint arthropathy with encroaching osteophytes, and fracture. Although CT can accurately distinguish disk herniations, CT has been shown to be less reliable than MRI in detecting nerve compressive lesions.110 Although minimally invasive, CT myelography is recommended in patients with contraindications to MRI or in the postoperative spine with preexisting hardware. In patients with cervical radicular symptoms, CT myelography may be critical in determining osseous versus diskogenic impingement, as well as show lateral recess lesions that are less conspicuous on MRI.6 Furthermore an ACR expert panel on chronic neck pain recommends CT myelography as a viable alternative to MRI for patients with suspected cord involvement when MRI cannot be performed.75

MRI The superior soft tissue contrast resolution of MRI has allowed for noninvasive detailed anatomic evaluation of the spine without radiation exposure and is considered the imaging modality of choice for detection of an objective cause of radicular symptoms versus normal age-related changes. Application of different pulse sequences allows  for image acquisition to exploit tissue composition.17 T1-weighted sequences provide optimal contrast between bright epidural fat and the internal components of the spinal canal. Osseous anatomy can also be evaluated using T1 sequences, owing to the abundance of fat within the vertebral body marrow. Accordingly an infiltrative or marrowreplacing process, as well as focal bone lesions, can be detected as hypointense signal on T1-weighted images. The intervertebral disks are slightly lower signal than the adjacent vertebral bodies. Cerebrospinal fluid (CSF) is lower signal on T1-weighted images than the relatively higher-signal spinal cord, nerve roots, and peripheral epidural fat. The exiting nerve roots and dorsal root ganglia are also surrounded by epidural fat. In general, T2-weighted images have greater contrast differentiation compared to T1-weighted images. For example, the cortical bone of the vertebral bodies demonstrates hypointense T2 signal, whereas the vertebral bone marrow remains higher signal owing to its rich fat content. The nucleus pulposus is made of water-binding glycosaminoglycans, thus demonstrating intermediate T1 and hyperintense T2 signal. The anulus fibrosus, which is the less-hydrated fibrous capsule surrounding the nucleus pulposus, demonstrates hypointense signal on all pulse sequences. CSF demonstrates high T2 signal, allowing for more sensitive identification of the intermediate-signal spinal cord and nerve roots. Fat-suppression techniques or the application of short  tau inversion recovery (STIR) sequences allows for more conspicuous detection of marrow edema. Fluid-weighted sequences are thus useful for visualizing disk desiccation, high-intensity zones, and Modic endplate changes. In the cervical spine, conventional spin echo imaging poorly differentiates cortical osteophytes from disk material. Gradient echo sequences allow for delineation of bone and disk margins, providing excellent contrast between the spinal cord and surrounding subarachnoid space, as well as clear visualization of the exiting cervical nerve roots.

LUMBAR SPINE Anatomy The changes that compromise the central canal and neuroforamen are directly related to the three-joint functional unit of the spine allowing motion: the end plate–intervertebral disk–end plate complex and the paired zygapophyseal (facet) joints. A review of the lumbar diskovertebral complex as they relate to acquired degenerative disease is useful. The anterior column consists of the anterior portion of the vertebral body, intervertebral disk, and anterior longitudinal ligament. The middle column consists of the posterior half of the vertebral body, intervertebral disk, and posterior longitudinal ligament. The posterior column includes the pedicles, facet joints, laminae, spinous processes, ligamentum flavum, and interspinous ligaments50 (Fig. 29-1). The spinal canal is formed by the posterior cortex of the vertebral body anteriorly, pedicles laterally, laminae posterolaterally, and the base of the spinous processes posteriorly, forming a protective ring surrounding the neural elements. Each spinal nerve exists within intervertebral neural foramen, bordered superiorly and inferiorly by adjacent pedicles, anteriorly by the intervertebral disk and vertebral body, and posteriorly by the facet joints. When viewed sagittally, the foramina should have a vertical oval appearance, with the exiting nerve roots passing through the upper portions just posterior to the diskovertebral complex. All lumbosacral nerve roots originate at approximately the T12-L1 level where the spinal cord terminates as the conus medullaris. The nerve roots course inferiorly through the spinal canal, forming the cauda equina, with the most laterally positioned roots leaving the bundle until they exit through their respective neural foramen. The exiting nerve roots carry the same number designation as the adjacent pedicle they travel beneath (e.g., the L5 nerve root exits under the L5 pedicle). The lateral recesses, which are located on the anteromedial aspects of the pedicles, contain the traversing nerve roots. The nerve roots travel within the lateral recesses after leaving the thecal sac but before entering the next exiting neural foramina. Epidural fat concentrically surrounds the thecal sac, which encases the exiting spinal nerves and fills the neural foramina (Figs. 29-2 and 29-3). The superior articular process, which is directed posteromedially, and the inferior articular process, which is directed anterolaterally, extend cranially and caudally and are joined by the pars interarticularis. The lumbar facet articulations are true synovial joints lined by hyaline cartilage, innervated by medial branches of the primary dorsal rami that arise from the spinal nerve just peripheral to the dorsal root ganglion. The ligamentum flavum, the interspinous and supraspinous ligaments, and the facet joint capsules constitute the posterior ligamentous complex. The ligamentum flavum bridges the interlaminar interval, attached to the facet joint capsules laterally and to the midline interspinous ligament medially, forming the posterior wall of the  vertebral canal.

Imaging of Lumbar Degenerative Disk Disease

Pathophysiology.  The diskovertebral complex is a cartilaginous joint formed by the intervertebral disk components and the hyaline cartilage–lined vertebral end plates. Surrounding the inner nucleus pulposus is a fibrous capsule, the anulus fibrosus, which is attached to the vertebral apophyseal ring by Sharpey fibers and continuous with the anterior and posterior longitudinal ligaments. Water bound to proteoglycans comprises a majority of the inner nucleus pulposus, acting to disperse axial loads in a radial distribution. With aging, disk desiccation ensues as the proteoglycan concentration and waterbinding capacity decreases, causing the nucleus to become more fibrous and less elastic, thus transferring axial loads to the posterior

http://radiologyebook.com

CHAPTER 29  Degenerative Disease

829

V D P

*I

A

S

B

FIG 29-1  A-B, Parasagittal and axial T1 images demonstrating the normal neural foramina bordered superiorly and inferiorly by adjacent pedicles (P), anteriorly by the intervertebral disk (D) and vertebral body (V), and posteriorly by the facet joints. The exiting nerve roots pass through the upper portions just posterior to the diskovertebral complex. Arrow indicates normal epidural fat surrounding nerve roots. (*), ligamentum flavum; S, indication superior facet of lower vertebra; I, indicating inferior facet of upper vertebra.

V

V

L

L

S

A

B FIG 29-2  A, Axial T2 image of the lumbar spine demonstrating normal appearance of nerve roots surrounded by cerebrospinal fluid (arrow). B, Corresponding CT in same patient demonstrating low-density epidural fat (arrow) and normal appearance of the ligamentum flavum (L). S, spinous process; V, vertebral body.

aspect of the anulus. As the anulus continues to stretch, fissures can develop, resulting in internal disk disruption, axial pain, and disk herniation. Excessive physical loading and stretching of the posterior anulus allows for mechanical stimulation of the pain receptors within the annular fibers.11,83,85 However, radiculopathy and radicular pain are the products of both mechanical nerve compression and an inflammatory response, likely mediated by tumor necrosis factor (TNF)-α, although other substances may play a role in the pain cascade. Evidence to support an inflammatory response as a component of radicular pain

includes: (1) pain persists despite adequate surgical decompression, (2) large disk herniations may be asymptomatic, (3) size or shape of herniation has no significant correlation with severity of symptoms, and (3) conservative therapy is usually effective.28,72,83 Furthermore, the nucleus pulposus has been shown to induce an inflammatory response and nerve dysfunction,28,53,72,73,80 which is relieved by anti-inflammatory agents,24,28 but concomitant compression is necessary to cause pain.72 Degenerative changes or degenerative disk disease applies to the process of intervertebral disk desiccation, manifesting as loss of the

http://radiologyebook.com

830

PART II  CT and MR Imaging of the Whole Body

EF D

A

C

D

D

LF

B

E FIG 29-3  Lumbar spine anatomy. A, Sagittal T1-weighted image shows vertebral bone marrow to be brighter than the water-filled disk (D). Nerve roots (long arrow) are slightly hyperintense to CSF (short arrow) on T1. Notice the bright epidural fat (EF) on the T1 images. B, Corresponding sagittal T2-weighted fat-suppression image shows normal hyperintense T2 signal of the intervertebral disk, the hyperintense T2 signal of CSF (long arrow, and lower signal nerve roots (short arrow). C, Axial T2-weighted image at the level of the L4-L5 disco-vertebral junction demonstrates the traversing L5 nerve root (black circle) identified within the lateral recess. D, Corresponding axial T1-weighted image at the level of the L4-L5 disco-vertebral junction demonstrates the exiting L4 nerve root (white circle) within the neural foramen. E, Corresponding CT myelogram shows the nerve roots (white arrow) within the thecal sac. D, disc; LF, ligamentum flavum.

normal T2 signal within the nucleus, which may be an early sign of degeneration or reflect normal age-related changes.83,85 Spondylosis is often synonymous with disk disease but should be reserved as a descriptor for marginal anterior and lateral osteophyte formation without substantial disk height loss.20 Osseous remodeling can occur in a relatively normal joint that is exposed to excessive external loads, or as a response to the nucleus pulposus matrix degradation.28,39 In a study by Nathan, anterior and lateral osteophytes were identified in all cadavers older than age 40, and posterior osteophytes were present in a majority of specimens older than age 80. However, failure of the nucleus sets in motion a cascade of events termed intervertebral osteochondrosis, manifesting as decreased disk space height, posterior osteophytes, radial annular fissures, end-plate erosion and marrow changes, and nitrogen gas within the nucleus, which are pathologic processes that are not features of normal aging.28,31,64 Osteophytes are formed as attempts to increase the surface area of the diskovertebral complex by creating additional articular surfaces to reduce or redis-

tribute the axial loading compressive forces throughout the joint.11,17 With loss of disk height and compressive stress redistribution across the joint, load-bearing capacity is shifted toward the posterior column. The facet joints endure increased stresses, leading to thinning and erosion of the articular cartilage, joint space narrowing, subchondral sclerosis, osteophyte formation, synovitis, and synovial cyst formation. Facet hypertrophy and osteophyte formation primarily arising from the superior articular facet can result in central canal stenosis as well as lateral recess and neural foraminal narrowing with subsequent compromise of traversing or exiting nerve roots. Disk degeneration with height loss and facet joint osteoarthrosis causes increased stress on the posterior ligamentous structures, resulting in ligamentum flavum hypertrophy, laxity, and anterior buckling50 and thus spinal stenosis21,42,50,63,98,100 (Fig. 29-4A1-A3). Increased stresses also result in degeneration of the interspinous process ligaments, leading to close apposition and hypertrophy of the spinous processes (“kissing”), causing flattening or a grooved appearance, subchondral sclerosis, and

http://radiologyebook.com

CHAPTER 29  Degenerative Disease

**

A2

A1

A3

*

B1

B2

B3

B4

FIG 29-4  A, Disk extrusion with acquired spinal stenosis. Disk desiccation with a focal central disk extrusion at L3-L4. There is associated ligamentum flavum hypertrophy, the combination of which results in moderate central canal stenosis. B, Spinal stenosis with degenerative spondylolisthesis. Sagittal T1-weighted image (B1) shows grade I anterolisthesis of L4, resulting in uncovering of the L4-L5 intervertebral disk (arrow). Sagittal (B2) and axial (B3) T2-weighted images demonstrate L4-L5 disk desiccation with inward buckling of the hypertrophied ligamentum flavum (*) as well as severe facet arthropathy (arrows), the combination of which results in severe central canal stenosis. Corresponding axial CT (B4) clearly demonstrates the advanced hypertrophic osseous facet joint changes, with subchondral sclerosis and osteophytes encroaching upon the thecal sac, contributing to spinal stenosis.

http://radiologyebook.com

831

832

PART II  CT and MR Imaging of the Whole Body

cystic degeneration and interspinous bursa formation (see Fig. 29-6B1-B4). Moreover, spondylolisthesis or subluxation of the facet joints allows for vertebral body motion or slippage, most commonly at L4-L5, causing progressive central canal or neural foraminal narrowing (see Fig. 29-4B1-B4). The dynamic changes that occur in the anterior and posterior columns of the spine in combination contribute to narrowing of the central canal, lateral recesses, and neural foramen.

Modic Changes.  Signal changes that occur within the subchondral bone along the vertebral end plates in response to continued nuclear matrix degradation and end-plate articular cartilage disruption have correlated with histologic findings. Originally described by Modic  and colleagues,17,68 the spectrum of subchondral vertebral marrow changes in association with degenerative disk disease has been classified into three types. Type I change is thought to represent an active inflammatory state, type II a more dormant state, and type III a postinflammatory state. However, it is not uncommon to visualize mixed 

I/II changes, suggesting that these changes can convert between types.87 Modic II changes are the most common, followed by type I and mixed type I/II.51,52,113 Type I changes show subchondral bone marrow edema, as demonstrated by decreased signal intensity on T1-weighted images and increased signal intensity on T2-weighted images (Fig. 29-5A1-A3). Histologically the end plates are disrupted and fissured, with the subchondral marrow infiltrated by fibrovascular granulation tissue, and thus exhibit enhancement on postcontrast images. Patients with Modic I changes have been documented to have high levels of C-reactive protein, which may support the presence of a local inflammatory response within the vertebral end plate.90 Furthermore, elevated levels of immunoreactive nerve fibers and TNF-immunoreactive cells were significantly higher in end plates demonstrating Modic I changes in comparison with patients with normal end plates on MRI or in end plates exhibiting Modic type II changes.64,78 Luoma et al.57 documented accelerated diskovertebral degenerative change in patients with Modic I changes, which supports a distinct pathologic process at the level of

A3

A2

A1

B1

B2

FIG 29-5  A, Disk protrusion and Modic I changes. Sagittal (A1) and axial (A2) T2-weighted images show a focal central disk protrusion at L5-S1, characterized by a broader base than the distance of the disk material extension in the anteroposterior plane. Sagittal T1- (A3) and T2-weighted (A1) images demonstrate hypointense T1 and hyperintense T2 signal along the subchondral end plates characteristic of Modic I changes. B, Modic II changes. Sagittal T1-weighted image (B1) demonstrates increased T1 signal paralleling the inferior end plate of L5 and superior end plate of S1, with corresponding loss of signal on the sagittal T2-weighted fat-saturation image (B2), compatible with Modic II marrow signal changes. Incidental note is made of an intraosseous disk herniation, or Schmorl’s node, along the superior end plate of L4, with surrounding edema on the T2-weighted sagittal image.

http://radiologyebook.com

CHAPTER 29  Degenerative Disease the diskovertebral unit. Overall, Modic I change has a stronger association with low back pain1 than type II changes. Type II changes are characterized by increased signal intensity on T1- and T2-weighted images, correlating to yellow or fatty marrow replacement within the subjacent vertebral body end plates (see  Fig. 29-5B1-B2). Type III changes are characterized by decreased signal intensity on both T1- and T2-weighted sequences, correlating with subchondral end-plate sclerosis.

High-Intensity Zones.  Initially described by April and Bogduk,3 a high-intensity zone (or annular fissure/tear) on MRI indicates internal disk disruption and the presence of trapped nuclear material within the periphery of the anulus fibrosus, correlating with grade IV annular fissuring at postdiskography CT and reproducible pain during provocative diskography.3,17,45,54,84,97 The trapped nuclear material demonstrates hyperintense T2 signal greater than that of the parent nucleus, as well as enhancement on postgadolinium images, secondary to the localized inflammatory process97 (Fig. 29-6A1-A4). Although there is

A1

A3

833

A2

A4

FIG 29-6  A1-A4, Disk extrusion. Axial T2-weighted image (A1) through the L4-L5 disk level shows a central/ left paracentral disk fragment displacing the thecal sac and lateral recess impinging upon the traversing left L5 nerve root. Extruded disk herniation is characterized by a base that is narrower than the distance of disk material extension in the anteroposterior plane. Sagittal T1- (A2) and T2-weighted (A3) images demonstrate the disk extrusion displacing the posterior longitudinal ligament, as well as a high-intensity zone of trapped nuclear material. Postcontrast sagittal T1-weighted image (A4) demonstrates expected enhancement of the high-intensity zone and adjacent epidural tissues secondary to the localized inflammatory response. Continued

http://radiologyebook.com

834

PART II  CT and MR Imaging of the Whole Body

B1

B2

B3

B4 FIG 29-6, cont’d B1-B4, Sequestered disk. Sagittal (B1) and axial (B2) T2-weighted images show a broadbased disk bulge at L5-S1, with focal left paracentral disk protrusion (arrows). There is also disk desiccation, grade I anterolisthesis, and narrowing at L4-L5, with uncovering of the disk and ligamentum flavum hypertrophy causing mild central canal stenosis. Right paramedian sagittal T2-weighted image (B3) and axial T2-weighted image (B4) inferior to the L5-S1 disk space demonstrates a disk fragment in the right subarticular zone without a clear attachment to an intervertebral disk, compatible with a sequestered disk fragment. Incidentally noted is close apposition between the L2-L3 and L3-L4 spinous processes, with subchondral sclerosis, edema, and interspinous bursa formation due to interspinous ligament degeneration (Baastrup’s disease).

http://radiologyebook.com

CHAPTER 29  Degenerative Disease

*

C2

C1

C3

*

D1

D2 FIG 29-6, cont’d C1-C3, Disk extrusion. Axial T1- (C1) and T2-weighted (C2) images through the L5-S1 disk space demonstrate a large left paracentral disk extrusion (arrow) narrowing the lateral recess and displacing the traversing left S1 nerve root (*). The sagittal T1-weighted image (C3) demonstrates caudal migration of the disk material behind the S1 vertebral body. D1-D2, Foraminal stenosis. Axial T1-weighted image (D1) shows a right foraminal extrusion (arrow) effacing the perineural fat and obscuring the exiting right L4 nerve root (note the normal exiting left L4 nerve root). Sagittal T1-weighted image (D2) shows the extruded disk material flattening and displacing the exiting L4 nerve root superiorly within the foramen (circled), with nearcomplete obliteration of the perineural fat (compare to normal fat surrounding the exiting L3 nerve root above).

http://radiologyebook.com

835

836

PART II  CT and MR Imaging of the Whole Body

agreement that a high-intensity zone is a marker for a posterior annular fissure, there is variable disagreement as to the significance of highintensity zones identified on MRI as an indicator of low back pain, because they are not always associated with reproducible pain response at diskography.16,89,92,95 However, it has been shown that annular tears exhibiting higher levels of T2 signal have greater specificity and stronger positive predictive value for pain production.64,81

Disk Herniation Nomenclature.  To address deficiencies in standardizing the language used in describing and interpreting the spectrum of degenerative changes and conditions affecting the lumbar spine, a multidisciplinary task force developed a consensus lexicon of nomenclature describing disk bulges and herniations.28,31 As discussed earlier, the following descriptions of intervertebral disk pathology do not imply a specific etiology of presenting symptoms and thus must be analyzed in the context of the clinical scenario. Abnormal disks can be classified as bulging or herniated. Disk herniation is defined as a localized displacement of disk material beyond the limits of the intervertebral disk space and includes nucleus pulposus material, cartilage, fragmented apophyseal bone, annular fissure, or a combination thereof. Herniated disks can be subclassified as protruded, extruded, or sequestered. Contour abnormalities of a disk in which the outer anulus extends in the horizontal (axial) plane beyond the margins of a vertebral body by more than 50% (>180 degrees) has been termed generalized displacement or bulging disk. Bulging disks are not considered to be a form of herniation and are unlikely to result in clinical symptoms. A disk herniation is a localized disk displacement of disk material in the horizontal (axial) plane of less than 50% of of the disk circumference. A focal herniation is considered less than 25% of disk circumference, whereas a broad-based herniation is 25% to 50% of disk circumference. Distinction between disk protrusion and extrusion is based upon shape of the displaced material. Protrusion refers to a herniated disk in which the width of the displaced material in any plane does not exceed the width of its base (see Fig. 29-5A1-A3). In other words, the base against the parent disk is broader than any other diameter of the herniation. Furthermore, the displaced nuclear material is contained by some intact annular fibers and the posterior longitudinal ligament. An extruded disk suggests complete disruption of the outer annular fibers, with disk material entering the epidural space. Extrusions are characterized by herniated disk material demonstrating a distance greater than its base of origin (i.e., the base is narrower than any other diameter of the herniation) (see Fig. 29-6A1-A4). Disk herniations  that have lost complete continuity with the disk of origin are characterized as sequestrations. Disk material displaced cranially or caudally from the site of extrusion, either remaining contiguous with the parent disk of origin or a sequestered fragment is known as migration (see Fig. 29-6B1-B4). Displaced disk material can be further classified into anatomic zones in the axial plane.28 The central zone, defined by the medial margin of the facets; the subarticular zone (lateral recess), defined as the medial facet margin to the medial pedicle margin; the foraminal zone, defined as the medial pedicle margin to the lateral pedicle margin; and the extraforaminal zone, which is defined as peripheral to the lateral pedicle margin. Over 90% of lumbar disk herniations occur at the L4-L5 level, with most occurring posterolaterally.94 Therefore most disk herniations affect the traversing roots within the lateral recess. For example, a posterolateral disk herniation at L5-S1 would contact/impinge upon the traversing S1 root and result in S1 radiculopathy (see Fig. 29-6C1-C3). Compromise of the exiting L4 nerve at the level of the L4-L5 disk level would be due to cephalad migration

of extruded disk material or disk displacement into the L4-L5 neural foramen (see Fig. 29-6D1-D2). Careful attention to the descriptions of disk herniation characterization and location, as well as imaging evaluation with knowledge of the presenting clinical context, allows for more accurate assessment, reporting, and interpretation of disk disease for the individual patient.

CERVICAL SPINE Although neck pain and cervical radiculopathy are commonly seen in clinical practice, recent studies have concluded that imaging may be overutilized.34 Similar to patients presenting with low back pain, physicians encountering the patient with neck pain are confronted with a diagnostic dilemma and the role of imaging. Multiple studies have concluded an association between previous injury and the development of neck pain, but they have also suggested that causation is likely multifactorial, including psychological, social, and stress-related factors.8,34,43,59,76,108 Chronic pain may also be seen in postoperative patients, as well as those presenting with ossification of the posterior longitudinal ligament. As in the lumbar spine, age-related changes in the cervical spine are common, asymptomatic, and increase with age. Likewise there is a basic specificity fault in cervical spine imaging that spondylosis, central canal stenosis and foraminal narrowing, and facet changes are ubiquitous in asymptomatic patients.37,38,62-64,79,107

Anatomy The cervical spine consists of seven vertebral bodies with three articulations, including the diskovertebral complex, zygapophyseal joints, and the uncovertebral joints. Anatomically the cervical uncovertebral joints, or joints of Luschka, are formed by the paired uncinate process, bony ridges directed superiorly and posteriorly from the posterolateral margin of the vertebral bodies, which articulate with notches along the posterolateral surfaces of the adjacent vertebral inferior end plate. The neural foramina are bordered anteromedially by the uncinate processes and posterolaterally by the facet joints, with the pedicles of adjacent vertebral bodies forming the superior and inferior borders. The  cervical zygapophyseal joints are more obliquely oriented and share equally in bearing axial compressive loads with the intervertebral disks11 (Fig. 29-7).

Imaging of Cervical Degenerative Disk Disease

Pathophysiology.  Analogous to lumbar disk herniations, the natural history of cervical disk herniation is spontaneous regression, with extrusions, migrated or sequestered material, and laterally oriented disk herniations more likely to undergo resolution.14,28,66 However, in the cervical spine, compressive lesions are more likely to have an osseous component rather than soft disk material, thus resolution is less likely compared to the lumbar spine. Osteophyte formation or squaring of the uncinate processes, as well as facet arthropathy and disk space height loss, all contribute to foraminal narrowing, resulting in cervical radicular symptoms (Fig. 29-8). The exiting cervical nerve roots travel along an inferior course through the foramen at about 45 degrees with respect to the coronal plane. Cervical nerves exit above the level of their respective disk level until C8, where is there is a transition to the nerves exiting below their respective disk level. Hence a foraminal extrusion compromising the C4-C5 foramen would impinge upon the exiting C5 nerve root, whereas a paracentral disk extrusion at C4-C5 will contact the traversing C6 nerve root. The most commonly affected levels are C5-C6 and C6-C7. Cervical intervertebral disks are structurally different than lumbar disks. Compared to lumbar intervertebral disks, the cervical disk

http://radiologyebook.com

CHAPTER 29  Degenerative Disease

*

837

D

B

A

D V

V

C FIG 29-7  Normal cervical anatomy. A, Sagittal T2-weighted fat suppression image demonstrates hyperintense T2 CSF, as well as inclusion of the cerebellar tonsils and upper thoracic levels. B, Axial gradient echo image at the level of the intervertebral disk (D) demonstrates the anterior and posterior divisions of the exiting nerve roots (arrows), as well as the posterolaterally positioned uncovertebral joints of Luschka (*). C, Corresponding axial CT image shows the paired facet joints (V).

anulus is not completely concentric, diminutive posteriorly and thicker anteriorly, functioning as an interosseous ligament rather than a constraint.11,65 The small nuclear component disappears early in life, leaving a residual fibrocartilage plate, leading to internal cracks, fissures, and increased mechanical stresses across the cartilaginous end plates, resulting in slowly developing disk-osteophyte complexes encroaching on the ventral canal and neural foramina. Increased loading of the uncovertebral joints results in hypertrophic change, compromising the lateral recesses and neural foramen. Changes occurring in the anterior column place increased stress across the facet joints, also resulting in hypertrophy and ligamentum flavum buckling anteriorly,63 further narrowing the central canal. Other than static

compression of the cervical cord, dynamic changes occurring during flexion and extension contribute to repetitive cord injury. During flexion, the spinal cord lengthens, allowing for contact along ventrally positioned osteophytes, whereas extension causes increased anterior flavum buckling and thus narrowing of the canal posteriorly.63,71,115 A central disk herniation/disk-osteophyte complex can encroach upon the ventral thecal sac and eventually progress to cord compression and cervical spondyolytic myelopathy (Fig. 29-9). In patients with neurologic symptoms, MRI readily depicts myelopathic changes in the cervical cord. Focal intramedullary T2 hyperintensity and corresponding T1 hypointensity may represent irreversible myelomalacia, including demyelination, gliosis, and cystic necrosis,2,63,77,88,95,105,106 whereas a

http://radiologyebook.com

838

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 29-8  Cervical disk-osteophyte complex. A, Sagittal T2-weighted image shows subchondral end-plate edema at C3-C4, with an associated posterior disk-osteophyte complex (arrow). The axial T2 (B) and T2* (C) images show the central and left paracentral posterior disk-osteophyte complex to efface the ventral thecal sac, as well as narrowing of the left lateral recess and neural foramen due to uncovertebral joint hypertrophy.

more faint-appearing or indistinct T2 hyperintensity is more likely to reflect reversible edema.60,63,95 Moreover, gadolinium enhancement20,63,82 and T1 hypointensity has been shown to be a poor prognostic indicator for recovery following decompression.2,5,32,60,70,74,102,104 Lastly, recent work evaluating diffusion tensor imaging has shown that reduced fractional anisotropy values reflect symptomatic myelopathy regardless of the presence of T2 hyperintensity and thus may prove more sensitive in identifying symptomatic and asymptomatic patients.13,46,47 In patients presenting with cervical radiculopathy, the North American Spine Society (2010) currently recommends that MRI be performed as the initial imaging modality. However, CT myelography may

be performed in patient populations with contraindications to MRI or in those situations where determining the nature of the compressive lesion—osseous versus disk as the cause of the symptoms (Fig. 29-10)—might perhaps alter the operative management.

THORACIC DISK DISEASE Degenerative disk disease of the thoracic spine is less common than both cervical and lumbar spine disease.109 Accurate diagnosis of thoracic spine pathology requires not only a strong clinical suspicion  but also appropriate diagnostic imaging examinations. Operations

http://radiologyebook.com

CHAPTER 29  Degenerative Disease

A

839

B FIG 29-9  Cervical disk herniation with severe stenosis. Sagittal (A) and axial (B) T2-weighted images show a large central disk extrusion arising from C5-C6, with cranial and caudal migration (arrows) effacing the ventral thecal sac and compressing the spinal cord. The sagittal image also shows focal hyperintense T2 signal within the cord (long arrow in A), compatible with contusion/myelomalacia.

resulting from symptomatic disk herniations in the thoracic spine constitute less than 1% to 2% of all disk surgeries.28,56,101 It has been estimated that the incidence of symptomatic disk herniations of the thoracic spine occur in about 1 in 1 million persons per year,18,109 resulting an incidence of approximately 0.25% to 0.75% each year. The occurrence in symptomatic herniations is greatest between the fourth and sixth decades of life, with peak incidence in the fifth decade. There is a slight male predominance, with symptomatic females presenting later in life. Symptomatic disk herniations may be seen in adolescents as well, typically in patients with Scheuermann’s disease. Similar to lumbar and cervical disk herniations, Wood and colleagues concluded that thoracic spine disk herniations rarely change over time or become symptomatic.114 Symptomatic thoracic spinal canal stenosis related to aging is less common compared to the cervical and lumbar spine, likely owing to the added stability of the thoracic cage, and when present more commonly affects the lower thoracic spine.63

Anatomy When compared with the cervical and lumbar spine, the thoracic spine is relatively rigid because of the thoracic rib cage. The facets of the T1 through T10 vertebral bodies are vertically oriented, with a slight medial angulation in the coronal plane. As a result of the orientation of these facets, the thoracic spine has significant stability during flexion and extension and allows greater movement in both rotational and lateral bending movements. Studies have shown that biomechanically the intervertebral disks of the thoracic spine are at greatest risk for injury when there is a combined torsion and bending force.109 The spinal canal has a circular morphology, and despite the cord having the smallest diameter in the thoracic spine, the thoracic cordto-canal ratio is 40% compared with 25% in the cervical spine.109 The

dentate ligament runs longitudinally between the spinal cord and nerve roots, limiting posterior displacement of the cord within the canal, thus placing the cord at increased risk for ventral compression due to an anterior disk or bony processes. Additionally the natural kyphosis of the thoracic spine causes the cord to rest directly on the posterior longitudinal ligament and posterior aspect of the diskovertebral complex.

Imaging of Thoracic Degenerative Disk Disease

Pathophysiology.  As commonly seen in the cervical and lumbar regions, a higher percentage of disk herniations occur in a central or paracentral location, thus causing myelopathic symptoms.109 The majority of symptomatic disk herniations involve the middle to lower thoracic spine, with most surgical interventions occurring between T9 and T1128,103 and 7% of these herniations resulting in intradural extension. Thoracic disk herniations may be calcified, which may cause adhesion between the disk fragment and dura,40,91 and previous thoracic spine surgery may contribute to intradural fragments.26 Of note, herniations at the level of the conus or high cauda equina can mimic lumbar disease symptoms, and thus all lumbar MRI studies must include the conus.28,58 Initial radiographs are useful in defining alignment, acute osseous injuries, and detecting intradiskal calcifications. Intervertebral disk calcifications are often identified in symptomatic thoracic disk herniations, ranging from 45% to 71%, but are only identified in 10% of asymptomatic herniations.93,109 CT myelography can accurately diagnose disk herniations and associated osseous pathology (Fig. 29-11). On MRI T1-weighted images, disk herniations appear as intermediate signal, and on T2-weighted images as low signal. A pseudomyelogram appearance is seen on T2-weighted images secondary to the surrounding CSF. Calcifications within the disk appear as low signal on both T1 and T2 images.

http://radiologyebook.com

840

PART II  CT and MR Imaging of the Whole Body

B

A

C

D FIG 29-10  Severe acquired cervical spinal stenosis. Axial (A, B) and sagittal (C) reformatted images from a CT myelogram demonstrate a large central disk extrusion at C4-C5 (arrows) causing severe cord compression. D, Axial CT image through the C6-C7 level shows a partially calcified disk-osteophyte complex, with uncovertebral joint hypertrophy severely narrowing the foramen and impinging upon the exiting right C7 nerve root.

http://radiologyebook.com

CHAPTER 29  Degenerative Disease

A

841

B

FIG 29-11  Thoracic disk herniation. Sagittal (A) and axial (B) reformatted images from a CT myelogram show multilevel disk bulges, with a focal right paracentral disk protrusion at T5-T6 (arrow). Note vacuum disk phenomenon and Schmorl’s nodes at multiple thoracic levels.

REFERENCES 1. Albert HB, Manniche C: Modic changes following lumbar disc herniation. Eur Spine J 16(7):977–982, 2007. 2. Al-Mefty O, Harkey LH, Middleton TH, et al: Myelopathic cervical spondylotic lesions demonstrated by magnetic resonance imaging.   J Neurosurg 68:217–222, 1988. 3. Aprill C, Bogduk N: High-intensity zone: A diagnostic sign of painful lumbar disc on magnetic resonance imaging. Br J Radiol 65(773): 361–369, 1992. 4. Autio RA, Karppinen J, Niinimaki J, et al: Determinants of spontaneous resorption of intervertebral disc herniations. Spine (Phila Pa 1976) 31(11):1247–1252, 2006. 5. Avadhani A, Rajasekaran S, Shetty AP: Comparison of prognostic value of different MRI classifications of signal intensity change in cervical spondylotic myelopathy. Spine J 10(6):475–485, 2010. 6. Bartynski WS, Lin L: Lumbar root compression in the lateral recess: MR imaging, conventional myelography, and CT myelography comparison with surgical confirmation. Am J Neuroradiol 24:348–360, 2003. 7. Bendix T, Kjaer P, Korsholm L: Burned-out discs stop hurting. Fact or fiction? Spine 33:E962–E967, 2008. 8. Binder AI: Cervical spondylosis and neck pain. BMJ 334(7592):527– 531, 2007. 9. Boden SD, Davis DO, Dina TS, et al: Abnormal magnetic-resonance scans of the lumbar spine in asymptomatic subjects. A prospective investigation. J Bone Joint Surg Am 72:403–408, 1990. 10. Boden SD, McCowin PR, Davis DG, et al: Abnormal magneticresonance scans of the cervical spine in asymptomatic subjects:   A prospective investigation. J Bone Joint Surg Am 72:1178–1184, 1990. 11. Bogduk N: Degenerative joint disease of the spine. Radiol Clin North Am 50(4):613–628, 2012. 12. Borenstein DG, O’Mara JW, Boden SD, et al: The value of magnetic resonance imaging of the lumbar spine to predict low-back pain in asymptomatic subjects (a seven-year follow-up study). J Bone Joint Surg Am 83A:1306–1311, 2001.

13. Budzik JF, Balbi V, Le Thuc V, et al: Diffusion tensor imaging and fibre tracking in cervical spondylotic myelopathy. Eur Radiol 21(2):426–433, 2011. 14. Bush K, Chaudhuri R, Hillier S, et al: The pathomorphologic changes that accompany the resolution of cervical radiculopathy: A prospective study with repeat magnetic resonance imaging. Spine 22:183–186, 1997. 15. Carragee E, Alamin T, Cheng I, et al: Are first-time episodes of serious LBP associated with new MRI findings? Spine J 6(6):624–635, 2006. 16. Carragee EJ, Paragioudakis SJ, Khurana S: Volvo Award winner in clinical studies (Lumbar high-intensity zone and discography in subjects without low back problems). Spine 25:2987–2992, 2000. 17. Carrino JA, Morrison WB: Imaging of lumbar degenerative disc disease. Semin Spine Surg 15(4):361–383, 2003. 18. Carson J, Gumpert J, Jefferson A: Diagnosis and treatment of thoracic intervertebral disc protrusions. J Neurol Neurosurg Psychiatry 34(1):68–77, 1971. 19. Cheung KM, Karpinnen J, Chan D, et al: Prevalence and pattern of lumbar magnetic resonance imaging changes in a population study of one thousand forty-three individuals. Spine 34:934–940, 2009. 20. Cho YE, Shin JJ, Kim K, et al: The relevance of intramedullary high signal intensity and gadolinium (Gd-DTPA) enhancement to the clinical outcome in cervical compressive myelography. Eur Spine J 20(12):2267–2274, 2011. 21. Chokshi FH, Quencer RM, Smoker WR: The “thickened” ligamentum flavum: Is it buckling or enlargement? AJNR Am J Neuroradiol 31(10):1813–1816, 2010. 22. Chou R, Deyo RA, Jarvik JG: Appropriate use of lumbar imaging for evaluation of low back pain. Radiol Clin North Am 50(4):569–585, 2012. 23. Chou R, Fu R, Carrino JA, et al: Imaging strategies for low-back pain: Systematic review and meta-analysis. Lancet 373(9662):463–472, 2009. 24. Cohen SP, Bogduk N, Dragovich A, et al: Randomized, double-blind, placebo-controlled, dose-response, and preclinical safety study of transforaminal epidural etanercept for the treatment of sciatica. Anesthesiology 110(5):1116–1126, 2009. 25. Cribb GL, Jaffray DC, Cassar-Pullicino VN: Observations on the natural history of massive lumbar disc herniation. J Bone Joint Surg Br 89(6):782–784, 2007.

http://radiologyebook.com

842

PART II  CT and MR Imaging of the Whole Body

26. D’Andrea G, Trillo G, Roperto R, et al: Intradural lumbar disc herniations: The role of MRI in preoperative diagnosis and review of the literature. Neurosurg Rev 27(2):75–80, discussion 81–82, 2004. 27. Patel ND, Broderick DF, Burns J, et al: ACR Appropriateness Criteria® Low Back Pain. Available at https://acsearch.acr.org/docs/69483/ Narrative/. American College of Radiology. Accessed January 16,   2016. 28. Del Grande F, Maus TP, Carrino JA: Imaging the intervertebral disk: Age-related changes, herniations, and radicular pain. Radiol Clin North Am 50(4):629–649, 2012. 29. Deyo RA, Weinstein JN: Low back pain. N Engl J Med 344:363–370, 2001. 30. Elias F: Roentgen findings in the asymptomatic cervical spine. N Y State J Med 58:3300–3303, 1958. 31. Fardon DF, Milette PC: Combined Task Forces of the North American Spine Society, American Society of Spine Radiology, and American Society of Neuroradiology. Nomenclature and classification of lumbar disc pathology. Recommendations of the Combined Task Forces of the North American Spine Society, American Society of Spine Radiology, and American Society of Neuroradiology. Spine (Phila Pa 1976) 26(5):E93–E113, 2001. 32. Fernandez de Rota JJ, Meschian S, Fernandez de Rota A, et al: Cervical spondylotic myelopathy due to chronic compression: The role of signal intensity changes in magnetic resonance images. J Neurosurg Spine 6:17–22, 2007. 33. Feydy A, Pluot E, Guerini H, et al: Osteoarthritis of the wrist and hand, and spine. Radiol Clin North Am 47(4):723–759, 2009. 34. Goode AP, Freburger J, Carey T: Prevalence, practice patterns, and evidence for chronic neck pain. Arthritis Care Res (Hoboken) 62(11):1594–1601, 2010. 35. Gore DR, Sepic SB, Gardner GM: Roentgenographic findings of the cervical spine in asymptomatic people. Spine 1:521–524, 1986. 36. Hammouri QM, Haims AH, et al: The utility of dynamic flexionextension radiographs in the initial evaluation of the degenerative lumbar spine. Spine (Phila Pa 1976) 32(21):2361–2364, 2007. 37. Hayashi H, Okada K, Hamada M, et al: Etiologic factors of myelopathy. A radiographic evaluation of the aging changes in the cervical spine. Clin Orthop Relat Res 214:200–209, 1987. 38. Heller CA, Stanley P, Lewis-Jones B, et al: Value of x-ray examinations of the cervical spine. Br Med J 287:1276–1278, 1983. 39. Jacobs DS: Degenerative diseases of the spine. Haaga: CT and MRI of the Whole Body, ed 5, St. Louis, 2008, Mosby, pp 755–799. 40. Jenkins LE, Bowman M, Cotler HB, et al: Intradural herniation of a lumbar intervertebral disc. J Spinal Disord 2(3):196–200, 1989. 41. Jensen MC, Brant-Zawadzki MN, Obuchowski N, et al: Magnetic resonance imaging of the lumbar spine in people without back pain.   N Engl J Med 331:69–73, 1994. 42. Jinkins JR: Acquired degenerative changes of the intervertebral segments at and suprajacent to the lumbosacral junction. A radioanatomic analysis of the nondiscal structures of the spinal column and perispinal soft tissues. Eur J Radiol 50(2):134–158, 2004. 43. Kaaria S, Laaksonen M, Rahkonen O, et al: Risk factors of chronic neck pain: A prospective study among middle-aged employees. Eur J Pain 16(6):911–920, 2012. 44. Kalichman L, Li L, Kim DH, et al: Facet joint osteoarthritis and low back pain in the community-based population. Spine 33:2560–2565, 2008. 45. Kang C, Kim Y, Lee S, et al: Can magnetic resonance imaging accurately predict concordant pain provocation during provocative disc injection? Skeletal Radiol 38:877–885, 2009. 46. Kara B, Celik A, Karadereler S, et al: The role of DTI in early detection of cervical spondylotic myelopathy: A preliminary study with 3-T MRI. Neuroradiology 53(8):609–616, 2011. 47. Keřkovský M, Bednařík J, Dušek L, et al: Magnetic resonance diffusion tensor imaging in patients with cervical spondylotic spinal cord compression: Correlations between clinical and electrophysiological findings. Spine (Phila Pa 1976) 37(1):48–56, 2012. 48. Khalil JG, Nassr A, Maus TP: Physiologic imaging of the spine. Radiol Clin North Am 50(4):599–611, 2012.

49. Kjaer P, Leboeuf C, Korsholm L, et al: Magnetic resonance imaging and low back pain in adults: A diagnostic imaging study of 40-year-old men and women. Spine 30:1173–1180, 2005. 50. Kotsenas AL: Imaging of posterior element axial pain generators: Facet joints, pedicles, spinous processes, sacroiliac joints, and transitional segments. Radiol Clin North Am 50(4):705–730, 2012. 51. Kuisma M, Karppinen J, Niinimaki J, et al: A three-year follow-up of lumbar spine endplate (Modic) changes. Spine (Phila Pa 1976) 31(15):1714–1718, 2006. 52. Kuisma M, Karppinen J, Niinimaki J, et al: Modic changes in endplates of lumbar vertebral bodies: Prevalence and association with low back and sciatic pain among middle-aged male workers. Spine (Phila Pa 1976) 32(10):1116–1122, 2007. 53. Kulisch SD, Ulstrom CL, Michael CJ: The tissue origin of low-back pain and sciatica: A report of pain response to tissue stimulation during operations on the lumbar spine using local anesthesia. Orthop Clin North Am 22:181–187, 1991. 54. Lam KS, Carlin D, Mulholland RC: Lumbar disc high-intensity zone (The value and significance of provocative discography in the determination of the discogenic pain source). Eur Spine J 9:36–41, 2000. 55. Lee MJ, Riew DK: The prevalence cervical facet arthrosis: An osseous study in a cadaveric population. Spine J 9:711–714, 2009. 56. Levi N, Gjerris F, Dons K: Thoracic disc herniation: Unilateral transpedicular approach in 35 consecutive patients. J Neurosurg Sci 43:37–42, 1999. 57. Luoma K, Vehmas T, Grönblad M, et al: Relationship of Modic type 1 change with disk degeneration: A prospective MRI study. Skeletal Radiol 38:237–244, 2009. 58. Lyu RK, Chang HS, Tang LM, et al: Thoracic disc herniation mimicking acute lumbar disc disease. Spine 24:416–418, 1999. 59. Makela M, Heliovaara M, Sievers K, et al: Prevalence, determinants, and consequences of chronic neck pain in Finland. Am J Epidemiol 134(11):1356–1367, 1991. 60. Mastronardi L, Elsawaf A, Roperto R, et al: Prognostic relevance of the post operative evolution of intramedullary spinal cord changes in signal intensity on magnetic resonance imaging after anterior decompression for cervical spondylotic myelopathy. J Neurosurg Spine 7(6):615–622, 2007. 61. Masui T, Yukawa Y, Nakamura S, et al: Natural history of patients with lumbar disc herniation observed by magnetic resonance imaging for minimum 7 years. J Spinal Disord Tech 18(2):121–126, 2005. 62. Matsumoto M, Fujimura Y, Suzuki N, et al: MRI of cervical intervertebral discs in asymptomatic subjects. J Bone Joint Surg Br 80(1):19–24, 1998. 63. Maus TP: Imaging of spinal stenosis: Neurogenic intermittent claudication and cervical spondylotic myelopathy. Radiol Clin North Am 50(4):651–679, 2012. 64. Maus TP, Aprill CN: Lumbar diskogenic pain, provocation diskography, and imaging correlates. Radiol Clin North Am 50(4):681–704, 2012. 65. Mercer S, Bogduk N: The ligaments and anulus fibrosus of human adult cervical intervertebral discs. Spine 24:619–626, 1999. 66. Mochida K, Komori H, Okawa A, et al: Regression of cervical disc herniation observed on magnetic resonance images. Spine 23:990–997, 1998. 67. Modic MT, Obuchowski NA, Ross JS, et al: Acute low back pain and radiculopathy: MR imaging findings and their prognostic role and effect on outcome. Radiology 237(2):597–604, 2005. 68. Modic MT, Steinberg PM, Ross JS, et al: Degenerative disk disease (Assessment of changes in vertebral body marrow with MR imaging). Radiology 166:193–199, 1988. 69. Monument MJ, Salo PT: Spontaneous regression of a lumbar disk herniation. CMAJ 183(7):823, 2011. 70. Morio Y, Teshima R, Nagashima H, et al: Correlation between operative outcomes of cervical compression myelography and MRI of the spinal cord. Spine 26(11):123, 2001. 71. Muhle C, Weinert D, Falliner A, et al: Dynamic changes of the spinal canal in patients with cervical spondylosis at flexion and extension using magnetic resonance imaging. Invest Radiol 33(8):444–449, 1998.

http://radiologyebook.com

CHAPTER 29  Degenerative Disease 72. Mulleman D, Mammou S, Griffoul I, et al: Pathophysiology of disk-related sciatica. I.–Evidence supporting a chemical component. Joint Bone Spine 73(2):151–158, 2006. 73. Mulleman D, Mammou S, Griffoul I, et al: Pathophysiology of disk-  related low back pain and sciatica. II. Evidence supporting treatment with TNF-alpha antagonists. Joint Bone Spine 73(3):270–277, 2006. 74. Mummaneni PV, Kaiser MG, Matz PG, et al: Preoperative patient selection with magnetic resonance imaging, computed tomography, and electroencephalography: Does the test predict outcome after cervical surgery? J Neurosurg Spine 11(2):119–129, 2009. 75. Newman JS, Weissman BN, Angevine PD, et al: ACR Appropriateness Criteria® Chronic Neck Pain. Available at https://acsearch.acr.org/ docs/69426/Narrative/. American College of Radiology. Accessed February 22, 2014. 76. Nolet PS, Cote P, Cassidy JD, et al: The association between a lifetime history of a neck injury in a motor vehicle collision and future neck pain: A population-based cohort study. Eur Spine J 19(6):972–981, 2010. 77. Ohshiro I, Hatayama A, Kaneda K, et al: Correlation between histopathologic features and magnetic resonance images of spinal cord lesions. Spine 18:1140–1149, 1993. 78. Ohtori S, Inoue G, Ito K, et al: Tumor necrosis factor-immunoreactive cells and PGP 9.5-immunoreactive nerve fibers in vertebral endplates of patients with discogenic low back pain and Modic type 1 or type 2 changes on MRI. Spine 31(9):1026–1031, 2006. 79. Okada E, Matsumoto M, Fujiwara H, et al: Disc degeneration of cervical spine on MRI in patients with lumbar disc herniation: Comparison study with asymptomatic volunteers. Eur Spine J 20(4):585–591, 2011. 80. Olmarker K, Rydevik B, Nordborg C: Autologous nucleus pulposus induces neurophysiologic and histologic changes in porcine cauda equina nerve roots. Spine 18:1425–1432, 1993. 81. O’Neill C, Kurganshy M, Kaiser J, et al: Accuracy of MRI for diagnosis of discogenic pain. Pain Physician 11:311–326, 2008. 82. Ozawa H, Sato T, Hyodo H, et al: Clinical significance of intramedullary Gd-DTPA enhancement in cervical myelopathy. Spinal Cord 48(5):415–422, 2010. 83. Peng B, Hao J, Hou S, et al: Possible pathogenesis of painful intervertebral disc degeneration. Spine 31:560–566, 2006. 84. Peng B, Hou S, Wu W: The pathogenesis and clinical significance of a high-intensity zone (HIZ) of lumbar intervertebral disc on MRI imaging in the patient with discogenic low back pain. Eur Spine J 15(5):583–587, 2006. 85. Peng B, Wu W, Hou S, et al: The pathogenesis of discogenic low back pain. J Bone Joint Surg Br 87:62–67, 2005. 86. Pengel LH, Herbert RD, Maher CG, et al: Acute low back pain: Systematic review of its prognosis. BMJ 327:323–327, 2003. 87. Rahme R, Moussa R: The Modic vertebral end plate and marrow changes: Pathologic significance and relation to low back pain and segmental instability of the lumbar spine. Am J Neuroradiol 29(5): 838–842, 2008. 88. Ramanauskas WL, Wilner HI, Metes JJ, et al: MR imaging of compressive myelomalacia. J Comput Assist Tomogr 13:399–404, 1989. 89. Rankine JJ, Gill KP, Hutchinson CE, et al: The clinical significance of the high-intensity zone on lumbar spine magnetic resonance imaging. Spine 24:1913–1919, 1999. 90. Rannou F, Ouanes W, Boutron I, et al: High-sensitivity C-reactive protein in chronic low back pain with vertebral end-plate Modic signal changes. Arthritis Rheum 57:1311–1315, 2007. 91. Reema C, Prabodhan P, Kshitij C, et al: MRI Diagnosis of intradural lumbar disc herniation. Report of three cases with review of literature. Internet J Orthop Surg 18:2, 2010. Available at http://ispub.com/IJOS/ 18/2/6425. Accessed March 8, 2014. 92. Ricketson R, Simmons JW, Hauser BO: The prolapsed intervertebral disc (The high-intensity zone with discography correlation). Spine 21:2758–2762, 1996. 93. Rogers MA, Crockard HA: Surgical treatment of the symptomatic herniated thoracic disk. Clin Orthop Relat Res 300:70–78, 1994.

843

94. Ross JS, Brant-Zawadzki M, Moore KR, et al: Diagnostic imaging spine, Salt Lake City, 2007, Amirys. 95. Saifuddin A, Braithwaite I, White J, et al: The value of lumbar spine magnetic resonance imaging in the demonstration of anular tears. Spine 23:453–457, 1998. 96. Savage RA, Whitehouse GH, Roberts N: The relationship between   the magnetic resonance imaging appearance of the lumbar spine and low back pain, age and occupation in males. Eur Spine J 6:106–114, 1997. 97. Schellhas KP, Pollei SR, Gundry CR, et al: Lumbar disc high-intensity zone (Correlation of magnetic resonance imaging and discography). Spine 21:79–86, 1996. 98. Schonstron NR, Hansson GH: Thickness of the lumbar ligamentum flavum as a function of load: An in vitro experimental study. Clin Biomech 6:19–24, 1991. 99. Schwarzer AC, Wang S, O’Driscoll D, et al: The ability of computed tomography to identify a painful zygapophysial joint in patients with chronic low back pain. Spine 20:907–912, 1995. 100. Siebert E, Prüss H, Klingebiel R, et al: Lumbar spinal stenosis: Syndrome, diagnostics and treatments. Nat Rev Neurol 5:392–403, 2009. 101. Singounas EG, Kypriades EM, Kellerman AJ, et al: Thoracic disc herniation analysis of 14 cases and review of the literature. Acta Neurochir (Wien) 116(1):49–52, 1992. 102. Smith ZA, Buchanon CC, Raphael D, et al: Ossification of the posterior longitudinal ligament: Pathogenesis, management, and current surgical approaches. A review. Neurosurg Focus 30(3):E10, 2011. 103. Stillerman CB, Chen TC, Douldwell WE, et al: Experience in the surgical management of 82 symptomatic herniated thoracic discs and review of the literature. J Neurosurg 88:623–633, 1998. 104. Suri A, Chabbra RP, Mehta VS, et al: Effect of intramedullary signal changes on the surgical outcome of patients with cervical spondylotic myelopathy. Spine J 3(1):33–45, 2003. 105. Takahashi M, Yamashita Y, Sakamoto Y, et al: Chronic cervical cord compression: Clinical significance of increased signal intensity on MR imaging. Radiology 173:219–224, 1989. 106. Taneichi H, Abumi K, Kaneda K, et al: Monitoring the evaluation of intramedullary lesions in cervical spinal cord injury. Qualitative and quantitative analysis with sequential MR imaging. Paraplegia 32:9–18, 1994. 107. Teresi LM, Lufkin RB, Reicher MA, et al: Asymptomatic degenerative disk disease and spondylosis of the cervical spine: MR imaging. Radiology 164(1):83–88, 1987. 108. van der Donk J, Schouten JS, Passchier J, et al: The associations of neck pain with radiological abnormalities of the cervical spine and personality traits in a general population. J Rheumatol 18(12):1884– 1889, 1991. 109. Vanichkachorn JS, Vaccaro AR: Thoracic disk disease: Diagnosis and treatment. J Am Acad Orthop Surg 8:159–169, 2000. 110. Van Rijn JC, Klemetso N, Reitsma JB, et al: Observer variation in the evaluation of lumbar herniated discs and root compression: Spiral CT compared with MRI. Br J Radiol 79:372–377, 2006. 111. Videman T, Battié MC, Gill K, et al: Magnetic resonance imaging findings and their relationships in the thoracic and lumbar spine. Insights into the etiopathogenesis of spinal degeneration. Spine 2:928–935, 1995. 112. Vroomen P, de Krom M, Knottnerus J: Predicting the outcome of sciatica at short-term follow-up. Br J Gen Pract 52:119–123, 2002. 113. Wang Y, Videman T, Battle MC: Modic changes: Prevalence, distribution patterns, and association with age in white men. Spine J 12(5):411–416, 2012. 114. Wood KB, Blair JM, Aepple DM, et al: The natural history of asymptomatic thoracic disc herniations. Spine 22(5):525–529, 1997. 115. Zhang L, Zeitoun D, Rangel A, et al: Preoperative evaluation of the cervical spondylotic myelopathy with flexion-extension magnetic resonance imaging: About a prospective study of fifty patients. Spine (Phila Pa 1976) 36(17):E1134–E1139, 2011.

http://radiologyebook.com

30  Spinal Tumors Fanny E. Morón, Alfred Delumpa, and Janio Szklaruk

33% to 53% of patients with NF2 have spinal cord ependymomas, which can be multiple in up to 58% of patients.39,51,52

INTRODUCTION Spinal neoplasms are divided into three groups based on their anatomic locations: extradural, extramedullary-intradural, and intramedullary tumors (Fig. 30-1). Extradural tumors include masses in the bones, disks, and paraspinal soft tissues. The majority of spinal tumors (60%) in adults are extradural tumors and are most commonly (90%) metastases of systemic cancer.62 Extramedullary-intradural tumors include masses originating from the dura/arachnoid and nerves. They are the second most common spinal tumors in adults (20%-30%) and most are either meningioma (50%) or peripheral nerve sheath tumors (50%).14 Intramedullary spinal cord tumors (IMSCTs) occur within the spinal cord. IMSCTs are the least common spinal tumor type in adults (4%-10%) but the most common spinal tumors in children (35%).14,79 IMSCTs primarily consist of gliomas (90%), of which ependymomas represent 60% and astrocytomas 30%14 (Fig. 30-2). Even though rare, IMSCTs are usually associated with significant neurologic dysfunction, morbidity, and mortality. The clinical presentation for spinal cord tumors often includes back pain, weakness, and/or insidious and progressive myelopathy. Torticollis and scoliosis can be seen in younger children. Patients with symptomatic spinal cord tumors often undergo a computed tomography (CT) examination of the spine as the first imaging modality. These images should be evaluated for the presence of intramedullary or epidural hyperdensity, which is always abnormal, usually reflecting trauma, neoplasm, or vascular malformation. Additionally CT helps identify matrix calcifications and extent of osseous destruction in cases of extradural tumors. However, magnetic resonance imaging (MRI) is the most appropriate imaging modality for any spinal cord and extramedullary-intradural pathology and for most of the extradural lesions.61 The contrast resolution of MRI is useful in identifying the anatomic location and soft tissue extent of all spinal tumors. Management and treatment of these tumors is highly dependent on the clinical presentation, tumor location, and histology. In this chapter we will review the epidemiology, clinical presentation, management, pathophysiology, and imaging findings of the most common spinal tumors.

INTRAMEDULLARY SPINAL CORD TUMORS Ependymoma Epidemiology.  Spinal ependymomas account for 1.8% of all central nervous system (CNS) neoplasms in adults.10 Ependymoma is the most common primary spinal cord tumor in adults (60%) and second most common in children.14 Most ependymomas are low-grade tumors.3,41 Approximately 50% of CNS ependymomas occur in the spinal cord, and those arising in the conus medullaris/cauda equina are generally of the myxopapillary type. Ependymomas are usually sporadic but can be associated with neurofibromatosis type 2 (NF2). Approximately

844

Presentation and Natural History.  The peak incidence for presentation for ependymomas is in the fourth to fifth decade of life. There is a slight male predominance for the intramedullary type. The cervical spinal cord is the most common location, followed by the thoracic and the conus medullaris.3,27,67 Patients commonly present with neck/back pain (95%) and have variable degrees of myelopathy with bilateral motor and sensory deficits for an average of 2 to 3 years prior to the time of diagnosis. The signs and symptoms usually spare the head and face.40 A significant number of patients with spinal cord ependymomas are asymptomatic, and surveillance with serial imaging and physical examination is advised when looking for disease progression.21 It is recommended that a safe surgical intervention be performed promptly in patients with a decline in neurologic function.21,52 Ependymomas usually have a plane of dissection, and gross total resection is possible and curative, with 5- and 10-year survival rates of 58% to 100%.21,41 Low-grade ependymomas have a low risk of local recurrence or metastatic disease. Residual tumor or rupture of the capsule during surgery increases the chances of local recurrence and development of metastases. Follow-up for ependymomas includes an immediate postoperative MRI and clinical examination, repeated at 6 and 12 months, and then annually for 10 years.

Pathophysiology.  Spinal cord ependymoma is a neoplasm that originates from the ependymal lining of the spinal cord central canal (see Fig. 30-2). The majority of ependymomas are of the cellular type, World Health Organization (WHO) grade II with low tendency to infiltrate the normal surrounding tissues. Generally there is a distinct plane between the tumor and adjacent normal cord, and a thin capsule may be present. The myxopapillary variant is a WHO grade I tumor and is found only at the conus medullaris (intramedullary) and filum terminale/cauda equina (extramedullary-intradural) (Fig. 30-3).

Imaging.  The most common imaging finding on CT and radiographs is scoliosis and nonaggressive bone remodeling. This is more often seen in ependymomas of the lower spinal canal and filum.26 CT and CT myelograms should be reserved only for patients who have contraindications for MRI.61 MRI is the most appropriate imaging modality for evaluating any patient suspected of spinal cord pathology.61 The MRI examination should include a combination of sagittal T1- and T2-  weighted imaging, short tau inversion recovery (STIR), susceptibilityweighted imaging (SWI), diffusion-weighted imaging (DWI); axial T1, T2, and SWI; and postgadolinium sagittal and axial T1 with fat suppression. Additional advanced imaging such as diffusion tensor imaging (DTI), magnetic resonance spectroscopy (MRS), and magnetic resonance perfusion (MRP) can help better characterize the lesions.

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

845

Dura Mater

FIG 30-1  Anatomic locations of intramedullary (green), extramedullaryintradural (blue), and extradural (red) spinal tumors. (Illustrated by David Bier, University of Texas MD Anderson Cancer Center, Houston, TX.)

A

B

C

FIG 30-3  Myxopapillary ependymoma of conus medullary (2 different cases). Case 1 A and B, Sagittal T1W and T2W images reveal a wellcircumscribed, encapsulated mass. Case 2, C, Patient with two wellcircumscribed and homogeneously enhancing lesions along the filum terminale/cauda equina consistent with distal drop metastatic myxopapillary ependymoma (arrow).

Dura Mater

FIG 30-2  Most common locations of the three most common intramedullary tumors: ependymoma is central (green), astrocytoma is eccentric (red), and hemangioblastoma is dorsal subpial (blue). (Illustrated by David Bier, University of Texas MD Anderson Cancer Center, Houston, TX.)

Most ependymomas are well circumscribed and originate from the centrally located ependymal cells and grow centrifugally, with symmetric expansion of the cord (see Fig. 30-2). They generally span an average of four vertebral segments, but accompanying cord edema extends beyond the tumor margins.17,67 Ependymomas are iso- to hypointense on T1-weighted images and hyperintense on T2 sequences. About 90% of ependymomas enhance after contrast administration and show well-defined margins with a homogeneous pattern (70%80%); rarely they will have little or no contrast enhancement.26,45 Most ependymomas do not restrict on DWI, and the apparent diffusion coefficient (ADC) is similar or higher compared to normal tissue.37,76 Ependymomas frequently present with hemorrhage (20%-60%) easily seen on T2* gradient echo and even more conspicuous on SWI images as signal loss from blood products within the tumor or in its rostral and caudal margins, which is known as the cap sign.26,47,73 About 60% to 90% of ependymomas have associated cysts,67 which are classified as tumoral or nontumoral. Tumoral cysts are usually located within the substance of the tumor and reflect necrosis-hemorrhage that shows inhomogeneous signal intensity and peripheral contrast enhancement on MRI. Nontumoral cysts are nonenhancing; their signal is cerebro-

A

B

FIG 30-4  Ependymoma. A, Sagittal T2W image shows an isointense to the cord mass with intratumoral areas of high signal and a non tumoral rostral polar cyst (white arrow). B, Post-contrast sagittal T1W image reveals a heterogeneously enhancing very well-circumscribed mass. A non-enhancing isointense to the CSF, non-tumoral superior polar cyst (black arrow). Syrinx distal to the tumor (arrow heads on A and B).

spinal fluid (CSF) equivalent and extends either rostrally or caudally beyond the tumor margins26,58 (Fig. 30-4). Nontumoral cysts are not removed during surgery but just aspirated, drained, or shunted, which allows for a smaller laminectomy needed to only resect the solid tumor portions.

http://radiologyebook.com

846

PART II  CT and MR Imaging of the Whole Body Astrocytoma

Other techniques such as DTI and its derived fractional anisotropy (FA) can be used in the evaluation of ependymomas. DTI can help depict the location, orientation, and anisotropy of white matter tracts adjacent to the tumor. Ependymomas tend to displace rather than destroy or infiltrate white matter tracts.64,71,74 In ependymomas there is a nonspecific tumoral and peritumoral reduction in FA values (FA = 0.48 ± 0.04), which is indicative of local extracellular edema and/or decrease in number of white matter fibers, with increased extracellular space12,35,37 (Fig. 30-5). On MRP the relative cerebral blood volume (rCBV) may be elevated in ependymomas.37 MRS shows elevated choline, which correlates with cell proliferation and thus tumor malignancy, and lipids, which are associated with hypoxia, apoptosis, and necrosis.29,76,77

A

G

B

H

C

Epidemiology.  Astrocytomas are the most common intramedullary tumors in the pediatric population and young adults (60%) and the second most common intramedullary tumor in adults (30%).14,30 There is roughly similar male-to-female predominance.8 Astrocytomas are more commonly located in the cervicothoracic or thoracic segments.41

Presentation and Natural History.  In adults the average age of onset is 30 years, which is earlier compared to ependymomas. The average duration of symptoms at the time of diagnosis is about 9 months but can be as long as 3 years. A shorter duration of symptoms is associated with higher tumor grades and poorer survival rates.8 The clinical presentation often consists of back pain and weakness.

D

E

F

I

FIG 30-5  Ependymomas (2) in NF2 patient. The sagittal T2W A and postcontrast T1 B, images show a well circumscribed heterogeneous mass with patchy enhancement and intratumoral cyst in the proximal cord (arrow head) and a distal smaller solid and homogeneously enhancing lesion (arrow). C, The sagittal DWI) without significant DWI hyperintensity. The ADC value D, in both lesions is higher or similar (proximal lesion, 1.050 × 10−3; distal lesion, 0.912 × 10−3) to the normal cord (0.907 × 10−3), confirming the unrestricted behavior of these lesions. E, The colored orientation DTI map shows continuity of the normal cord in blue displaced dorsally; and absence of color at the tumor sites (arrows). The DTI tractography F and G, confirms the noninfiltrative nature of the mass, with fibers being displaced and not entering the lesion. The fractional anisotrophy value H is decreased in the tumor (P = 0.138) compared to the normal cord (P = 0.628). The SWI I, shows subtle hypointense foci from prior hemorrhage, which is visible only on this sequence (arrow).

http://radiologyebook.com

CHAPTER 30  Spinal Tumors Torticollis and scoliosis are usually seen in younger children. Insidious myelopathy in children should prompt a spinal MRI study.8,79 Lowgrade astrocytomas have an 80% 5-year survival rate compared to 30% for high-grade astrocytomas.41,44 There is an association of spinal cord astrocytomas with neurofibromatosis type 1 (NF1). Pilocytic astrocytoma tends to have a more benign biological behavior when it occurs in patients with NF1. Benign well-circumscribed pilocytic astrocytoma has a plane of dissection, and gross total resection can be attempted. However, WHO grade II, III, and IV astrocytomas are infiltrative, and total resection without injuring the normal tissue is seldom accomplished; therefore conservative decompression, partial tumor removal, cyst decompression, and tissue diagnosis is commonly performed. Radiotherapy and chemotherapy are recommended for infiltrative residual or recurrent high-grade tumors.16,21

Pathophysiology.  Astrocytoma arises from astrocytic glial cells of the cord, with an affinity for the white matter tracts located in the periphery of the cord (see Fig. 30-2). Most are low-grade astrocytomas (70%-90%), either pilocytic astrocytoma (15% WHO grade I) or more commonly fibrillary astrocytoma (75% WHO grade II).3,40 On histo-

A

D

847

logic evaluation, fibrillary astrocytoma shows widespread parenchymal infiltration, whereas pilocytic astrocytoma displaces rather than infiltrates the cord. High-grade anaplastic astrocytoma (WHO grade III) and glioblastoma (WHO grade IV) are rare, accounting for about 10% to 30% of spinal cord astrocytomas.40,41 High-grade astrocytomas characteristically show mitotic activity and nuclear atypia. High-grade astrocytomas have a poor survival with higher recurrence and metastatic dissemination through the neuroaxis.30

Imaging.  MRI is the most appropriate imaging modality, and CT/CT myelograms should be reserved for patients with contraindications for MRI.61 Astrocytomas are infiltrating poorly defined neoplasms and tend to be eccentrically located with asymmetric and fusiform cord expansion (see Fig. 30-2). Astrocytomas span about four or fewer vertebral segments at the time of presentation. Holocord involvement is rare and usually seen in pilocytic astrocytoma, mainly in  children and adolescents.15,60 Astrocytomas are hypo- to isointense on T1 and hyperintense on T2 and STIR. Spinal astrocytomas usually enhance inhomogeneously in a nodular or patchy manner, and the enhancing tumor does not define the true tumor margins (Fig. 30-6).

B

C

E

FIG 30-6  Astrocytoma. The sagittal and axial T2-weighted images (A and D) show an ill-defined hyperintense mass with cord expansion and mild surrounding edema. B, Sagittal T1-weighted images show a slightly hypointense mass. C and E, Sagittal and axial T1-weighted postcontrast images show nodular inhomogeneously enhancing mass.

http://radiologyebook.com

848

PART II  CT and MR Imaging of the Whole Body

Approximately 18% to 30% of astrocytomas do not enhance.63 Most astrocytomas do not restrict on DWI, and their ADC values are not significantly decreased.76 They may have tumoral cysts with peripheral contrast enhancement, as well as nonenhancing nontumoral polar cysts and syrinxes.58 Hemorrhage is uncommon compared to ependymomas, and surrounding edema is variable 65 (Fig. 30-7). On DTI astrocytomas show diffuse infiltration of the cord or disrupted white matter tracts from tumor invasion13,71 (see Fig. 30-7). There is a reduction in FA values in astrocytomas, similar to the ones seen in ependymomas (0.48 ± 0.2).12,35,37 MRP shows low to moderate rCBV and increased vascular permeability on derived permeability maps, and one glioblastoma with increased rCBV has been reported.36,37 MRS obtained in the solid component of pilocytic astrocytoma demonstrates elevated choline/N-acetylaspartate (NAA) ratio, ranging from 1.80 to 3.40 compared with 0.53 to 0.75 for normal tissue.12,22,23,31 This finding implies hypercellularity.

Hemangioblastoma

Epidemiology.  Hemangioblastoma (HB) is the third most common tumor, accounting for 2% to 8% of all IMSCTs.38,78 It occurs sporadi-

cally in 70% of patients, but in 30% is associated with von HippelLindau (VHL) disease. VHL patients have 3p chromosomal mutations in the VHL gene.78 Multiple and usually small (<1 cm) HBs raise suspicion for VHL syndrome and require contrast-enhanced MRI of the entire neuraxis to evaluate for cerebellar and/or spinal HBs.6

Presentation and Natural History.  HBs usually present in the third or fourth decade of life; those in VHL patients usually present a decade earlier. There is a male predominance. HBs are slow-growing tumors with a long clinical course (mean duration of symptoms, 38 months), and sensory disturbances occur more commonly than motor dysfunction and pain.3,30 In patients with small tumors (<1 cm) symptoms are attributed to an accompanying syrinx, and in larger tumors are due to the tumor itself and/or a syrinx. HBs almost never cause subarachnoid hemorrhage or hematomyelia. The treatment of choice in symptomatic patients is microsurgical en bloc resection, which is curative. Once the tumor is removed, the syrinx resolves as well. Preoperative endovascular embolization reduces arterial flow and shrinks the draining veins, which decreases the risk of intraoperative bleeding. Radiosurgery and gamma knife are alternative treatment options for patients

C

A

F

B

E

D

G

H

FIG 30-7  Astrocytoma. A, The sagittal T2W image shows a heterogeneous signal solid and cystic and B and C, the post-contrast sagittal and axial T1W images show patchy enhancement and a superior polar peripherally enhancing tumoral cyst (arrows). D, The axial SWI shows conspicuous hypointense foci from prior hemorrhage, barely suggested on T2 (arrows). E and F, Sagittal DWI and ADC without significant DWI hyperintensity and ADC value similar to slightly higher than (1.094 × 10−3) the normal cerebellum (0.829 × 10−3) confirm the unrestricted behavior of the tumor. G and H, The DTI tractography shows that most of the cord fiber tracts are infiltrated by the tumor.

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

A

B

A

FIG 30-8  Hemangioblastomas of the spinal cord in a patient with von Hippel-Lindau syndrome. Sagittal T2W A, and post-contrast T1W B, images show extensive syrinx proximal and distal to multiple lesions, some with flow voids (arrow) and all of them homogeneously and intensely enhancing. The lesions are located in the midline of the dorsal surface of the cord.

849

B

FIG 30-9  Hemangioblastoma. Sagittal T2- (A) and T1-weighted (B) images show heterogeneous signal lesion (arrow) with low signal foci that correspond to flow voids from dilated vessels and intense enhancement.

with medical comorbidities that preclude surgery and in patients with recurrent or unresectable tumors.4,75 In patients with VHL, serial MRIs are recommended to evaluate for recurrent and de novo lesions.

Pathophysiology.  HBs are benign WHO grade I neoplasms. They are nonglial mesenchymal-, capillary-, and venous-rich tumors with foamy stromal cells.6,19 On gross pathologic examination they appear as well-circumscribed and encapsulated masses with dilated and tortuous vessels on their surface. Most are intramedullary (60%-75%), but they can be combined intra- and extramedullary (10%-15%), purely extramedullary, or even extradural in location.

Imaging.  On MRI the solid portions of HBs are seen as welldemarcated hyperenhancing nodular masses. Small HBs (<1 cm) are isointense on T1, hyperintense on T2, and enhance intensely and homogeneously after contrast administration. They are usually located subpially in the dorsal surface of the thoracic more common than cervical spinal cord (70%) (Fig. 30-8; see Fig. 30-2). A disproportionately large syrinx as compared to the size of the tumor is seen in up to 64% of them.6 Larger HBs are hypointense on T1, heterogeneous on T2 (from intratumoral hemorrhage), with heterogeneous hyperenhancement after contrast administration. HBs that are larger than 2.5 cm always have vascular flow voids on T1 and T2 arising from dilated feeding arteries and draining veins containing fast-flowing arterialized blood6 (Fig. 30-9). MRI underestimates the presence of these dilated vessels in lesions smaller than 1.5 cm. A cystic lesion with an enhancing mural nodule is characteristic, but it is a less common presentation (Fig. 30-10A). Superficial enhancement of the cord adjacent to the lesion may be present and may correspond to perimedullary veins. FA values are decreased (0.59) but less than in other spinal cord tumors. On DTI the white matter fibers are wrapped around the tumor without being infiltrated or destroyed.12

A

B

FIG 30-10  Hemangioblastoma. A, Postcontrast sagittal T1-weighted image shows a well-circumscribed cystic lesion with an intensely enhancing nodule. B, Spinal angiogram with left 10 thoracic intercostal artery injection and opacification of the spinal artery of Adamkiewicz (arrow) confirms the hypervascular mass with a prominent tumor blush (arrowhead).

Computed tomography angiography (CTA) and magnetic resonance angiography (MRA) have sufficient spatial resolution to demonstrate the abnormal vessels and even the arterial pedicles; however, they are unable to distinguish arterial feeders from draining veins and usually depict both simultaneously. Time-resolved CTA and MRA may

http://radiologyebook.com

850

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 30-11  Hemangioblastoma. A, Dyna CT depicts a blush (arrow) within the hypervascular tumor and the prominent supplying arteries that are selectively catheterized to embolize the lesion. B, Following embolization there is complete devascularization of the tumor.

help in this regard. As such, Dyna CT provides superior spatial and time resolution to distinguish the tumor and the feeding artery from draining veins, compared to three-dimensional (3D)-CTA with intravenous injection70 (Fig. 30-11). This is essential in planning surgical treatment. Dyna-CT angiography is performed with iodine contrast injected directly to an intercostal artery. At surgery arterial devascularization should be performed before coagulating the draining vein; otherwise cord damage from venous hypertension may develop. Conventional or 3D rotational digital subtraction angiography is also useful for diagnosis and surgical planning and shows a prominent dilated supplying artery, intense and prolonged vascular blush, and a dilated draining vein with early venous opacification if a shunt is present (see Fig. 30-10A and B). These abnormal vessels will be seen even in small tumors that do not demonstrate abnormal vasculature on MRI.25

Ganglioglioma

Epidemiology.  Gangliogliomas are very rare tumors in adults, representing only about 1% of all spinal cord tumors,33,49,50 but they account for up to 15% of intramedullary neoplasms in the pediatric population.54 In a case series of 56 patients, a male predominance was reported24; in a recent smaller group of patients (18) a slight female predominance was found.80 The cervicothoracic segment is the most common location, followed in order of decreasing frequency by the thoracic, cervicomedullary, cervical, and conus.24

Presentation and Natural History.  Symptoms include paraparesis, segmental pain, radiculopathy, and focal sensory or motor deficit.24 The mean duration of symptoms is 12 months (range, 2 weeks to 84 months).24 Gross total surgical excision is the treatment of choice. After resection, there is a 5- and 10-year survival rate of 89% and 80%, respectively.20,24,42 Tumor recurrence after surgery occurs in about 30% of patients; close follow-up with serial semiannual or annual MRI scan is advised.24

patient’s young age together with some MRI characteristic features help suggest the diagnosis. These findings are long tumor length (>four levels, extending over an average of eight vertebral bodies, or up to the whole cord), tumoral cysts, absence of edema, mixed signal intensity on T1 (due to dual cellular composition of the tumor), eccentric location within the cord, patchy tumor enhancement, and pial surface enhancement (Fig. 30-12A and B). Hemorrhage and calcifications are rare.24,50

Spinal Cord Lymphoma

Epidemiology.  Primary CNS lymphoma (CNSL) is rare and accounts for approximately 1% to 5% of all CNS tumors.18 Secondary involvement by systemic lymphoma is more commonly seen (via hematogenous or direct extension), and the incidence varies between 2% and 27% for non-Hodgkin’s lymphoma and less than 0.5% for Hodgkin’s lymphoma.46 Intramedullary lymphoma occurs in 1% to 3% of all primary CNSLs and usually carries a poor prognosis. Ninety percent of primary intramedullary lymphomas are large B-cell lymphomas, as are most CNS ones.34 Patients with acquired immunodeficiency disorders such as post–organ transplant treatment or human immunodeficiency virus (HIV) are at a higher risk to develop primary CNSL, which is linked to Epstein-Barr virus infections. In the immunocompromised patient the onset of disease is at an earlier age (30-40 years) than for the immunocompetent population (60 years).18,69 Treatment with highly active antiretroviral therapy (HAART) has decreased the incidence of CNSL in HIV patients.9

Presentation and Natural History.  Patients present with back pain

Pathophysiology.  Gangliogliomas are composed of a combination

and progressive spinal neuropathy and radiculopathy. Treatment generally involves administration of corticosteroids, methotrexate, and radiation because CNSLs are sensitive to chemotherapy and external radiation therapy, with remission seen in approximately 80% of cases. However, long-term survival is poor (≈36% in 2 years), with a high frequency of recurrence within the first 2 years after diagnosis (50%60%).18 Surgery is limited to tissue sampling, preferably with stereotactic biopsy.

of neoplastic ganglion cells and glial elements.42 They are slow growing, mostly benign, WHO grade I or II tumors.65

Pathophysiology.  Lymphomas are lymphoreticular neoplasms of

Imaging.  On imaging, scoliosis and bone remodeling are common but not specific. On MRI the findings are very variable; however, the

specific cellular differentiation. Primary CNSL has an angiocentric growth pattern with perivascular tumor infiltration by multiple cell layers and late vascular invasion, which ultimately compresses the

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

851

S

A

C

B

D

FIG 30-12  Ganglioglioma. A, Sagittal T2-weighted image shows an ill-defined slightly hyperintense mass with cord expansion and no surrounding edema. B, Noncontrast sagittal T1-weighted image shows a mixed low- and high-signal-intensity mass. C and D, Sagittal and coronal T1-weighted postcontrast images show patchy inhomogeneously enhancing mass.

blood vessels, resulting in necrosis. On gross pathology it demonstrates an ill-defined infiltrating tumor. Patients with secondary CNSL present with leptomeningeal spread in two thirds of cases and parenchymal lesions in one third of patients.

Imaging.  Contrast MRI is the modality of choice for diagnosis while on treatment and after therapy. Primary CNSL presents as a poorly circumscribed parenchymal mass that is iso- to hypointense relative to the cord on T1, a hyperintense mass on T2, and has surrounding edema; most lesions enhance after contrast administration (Fig. 30-13). The pattern of enhancement varies from homogeneous-intense to patchy, confluent, or infiltrating. There is leptomeningeal enhancement in 10% of primary CNSLs, which usually contacts a CSFcontaining surface.18,81 CNSL can present as a single or multifocal persistently enhancing lesion(s).34 The high tumor cellularity in lymphoma results in DWI hyperintensity and hypointensity on ADC maps, with a threshold value of 1.2 × 10−3 mm2/sec. This may help differentiate lymphoma from other tumors in the brain.1,76 Positive response to treatment results in increasing ADC values, whereas decreasing ADC values suggest tumor progression.11 The ADC value of lymphoma is lower than that of toxoplasmosis in immunocompromised patients.2 On DTI, increasing tumor cellularity in lymphomas has been  correlated with decreasing FA and increasing mean diffusivity.66,72 Only a modest increase in rCBV is seen (compared to double or  triple increases in high-grade gliomas), and it has been reported  that when there is too much cellularity or intravascular tumor infiltration, rCBV will decrease.32 Lymphoma shows a nonspecific MRS pattern, similar to high-grade gliomas, with decreased NAA, decreased

A

B

FIG 30-13  Lymphoma. Sagittal T2- (A) and postcontrast T1-weighted (B) images demonstrate a slightly iso- to hyperintense mass on T2 image with strong enhancement, cord expansion, and peritumoral edema.

http://radiologyebook.com

852

PART II  CT and MR Imaging of the Whole Body

creatine, increase choline, and lipid-lactate peaks. On SWI, lymphomas do not show hemorrhage as high-grade astrocytomas and ependymomas often do. Lymphomas are hypermetabolic fluorodeoxyglucose (FDG)-avid tumors on FDG positron emission tomography (PET), which has an 80% sensitivity in their diagnosis.28 Both PET and PET-CT can be used to stage and monitor treatment response. Single-photon emission CT (SPECT) with thallium 201 is useful to diagnose hypermetabolic lymphoma, as opposed to hypometabolic toxoplasmosis, in immunocompromised patients, with a sensitivity of 55% to 100% and specificity of 85% to 100%.43,48,55,82 These techniques can also distinguish hypermetabolic recurrent lymphoma from decreased or absent activity in treatment-induced necrosis.

Spinal Cord Metastases

Epidemiology.  Intramedullary spinal cord metastases (IMSCMs) are rare; their prevalence is 0.9% to 2.1% in postmortem series of patients with cancer.5,7,53 Lung cancer is the most common primary (close to 50%), followed by breast cancer (≈15%).56,57,59,68 Approximately 20% to 60% of these patients have additional neuraxis and other systemic metastases.56,59,68 Hematogenous dissemination is the most common route, either arterial or retrograde spread from the venous system. Less common paths are CSF dissemination through the spinal cord central canal or contiguous spread from carcinomatous meningitis.

Presentation and Natural History.  Patients with IMSCM present with rapid deterioration of motor and sensory symptoms (including hemicord syndromes) compared to the slower progression of symptoms in patients with primary cord neuroepithelial tumors.59,68 IMSCMs are usually associated with widespread systemic disease and imply an advanced stage and a dismal prognosis. The overall median survival is about 4 months in patients treated with chemotherapy and radiotherapy. Surgery in selected cases may improve symptoms and survival by 1 to 2 months.68 Entire neuraxis imaging should be performed when IMSCM is suspected.

A

B

C

FIG 30-14  Multiple intramedullary spinal cord metastasis from breast cancer. A, Sagittal T2-weighted image shows multiple well-defined isointense (to the cord) lesions with disproportionate peritumoral edema. B, Sagittal T1-weighted image shows multiple vertebrae with hypointense metastasis and diffuse hypointense cord edema. C, Postcontrast sagittal T1-weighted image shows multiple posterior fossae and spinal cord strongly enhancing masses.

Imaging.  MRI shows isointense lesion(s) on T1 and T2 hyperintense lesion(s) with disproportionate large T2/STIR hyperintense surrounding edema relative to the size of the lesion(s) (Figs. 30-14 and 30-15).56 IMSCMs are usually small, single or multiple, well-circumscribed, rounded, homogeneously or ring-enhancing lesions with mild cord expansion. Syrinx, cyst, and hemorrhage are rare. A more intense thin rim of peripheral enhancement and an ill-defined flame-shaped region of enhancement at the superior/inferior lesion margins have been described as common and specific for IMSCM when compared to primary spinal cord masses57 (see Fig. 30-15). On DTI the FA values are similar to other tumors (≈0.46), and DTI shows a pattern of fiber infiltration or destruction similar to that of astrocytomas.12

EXTRADURAL SPINAL TUMORS Metastatic Disease

*

Epidemiology.  The spine is the third most common site for metastatic disease, following the lung and liver.251 In fact, spinal metastases are the most common spinal tumor—20 times more common than primary spinal neoplasms.207 Metastatic disease to the spine can involve the bone, epidural space, leptomeninges, and the spinal cord. Extradural metastatic disease, which can seed the epidural space or the vertebra, accounts for 95% of spinal metastases. Purely epidural metastases are rare.90 The most common primary malignancies are breast, prostate, and lung cancer. Spinal metastases have the highest incidence between age 40 and 65 years.207 The most common location is the

A

B

FIG 30-15  Single intramedullary metastasis from lung cancer. (A) Sagittal T2W image shows poor definition of the lesion and disproportionate peritumoral edema. (B) Post-contrast sagittal T1W shows an enhancing mass with an ill-defined flame-shaped region of enhancement at the superior and to a lesser extent inferior margin of the lesion (arrows). Metastatic pathologic compression fracture with vertebra plana of T4 (*).

http://radiologyebook.com

CHAPTER 30  Spinal Tumors thoracic spine (60%-80%), followed by the lumbar spine (15%-30%), and the cervical spine (<10%).175

Presentation and Natural History.  Metastatic spinal tumor may be the first manifestation of malignancy in 10% or more of patients.178 Local neck or back pain is the major presenting symptom in the majority of patients. Progression of disease may later cause weakness, numbness, and sphincter dysfunction. Eventual progression to irreversible complete paralysis is inevitable unless the patient is treated.90,207 Generally, the prognosis of patients presenting with bone metastases is poor.216 Radiotherapy and surgery are important in the treatment of spinal metastatic disease.90,208 Percutaneous vertebral body augmentation, such as vertebroplasty or kyphoplasty, can be performed to treat painful vertebral body metastasis without neurologic compromise and can be used in conjunction with radiation therapy or as an alternative to surgery.122,175 Chemotherapeutic agents are sometimes administered with or after radiotherapy and/or surgery.175

Pathophysiology.  Metastases may spread arterially into the nourishing environment of the bone marrow. Metastases may be transmitted through the valveless venous channels of Batson’s venous plexus into the epidural space, generally via the canal for the basivertebral vein. Finally, paraspinal tumors may extend directly into the epidural space via venous channels or through the intervertebral foramen.207 When tumor cells invade bones, the normal balance of osteoblastic and osteoclastic activity is disturbed, which results in the osteolytic, sclerotic, or mixed appearance of bone metastases.83,213

Imaging.  A mass requires a 1-cm minimum diameter and 50% bone mineral loss at minimum for detection by radiographs; thus, up to 40%

A

B

C

853

of metastatic spinal lesions will be missed on plain films.219 Most metastatic spinal lesions manifest as osteolytic vertebral destruction.207 Breast, lung, kidney, thyroid, oropharyngeal, melanoma, adrenal, and uterine primaries usually produce osteolytic metastases. Prostate, bladder, nasopharyngeal, medulloblastoma, neuroblastoma, and carcinoid primaries often result in osteoblastic metastases. Mixed lyticsclerotic lesions are seen in breast, lung, ovarian, testicular, and cervical primaries.228 The most common finding on radiographs is pedicle erosion, followed by vertebral body compression fracture. Paraspinal line displacement and pathologic dislocation are less common findings.165 CT scans can recognize a bony metastatic lesion up to 6 months earlier than a radiograph.219 CT has excellent spatial resolution and can detect cortical destruction (Fig. 30-16A) and epidural soft tissue mass. Osteolytic metastases often involve the posterior vertebral cortex and pedicle. Osteoblastic metastases may present as nodular or mottled areas of sclerosis that also tend to erode the posterior vertebral body cortex and pedicle228 (Fig. 30-17A). MRI is excellent in detecting the early bone marrow deposits of metastatic disease and is more sensitive than CT in detecting osseous spinal metastases. MRI is essential in identifying epidural mass and spinal cord or nerve root compression. Unfortunately, conventional MRI often cannot differentiate changes due to treatment, fracture, and tumor. Spinal marrow signal that is hypointense to muscle or disks on T1-weighted images should be considered abnormal (see Figs. 30-16B and 30-17B). Hematopoietic marrow, which has more water content than fatty marrow, is hypointense to fat but hyperintense to normal muscle on T1-weighted images.111,255An understanding of the agerelated dynamic appearance of normal bone marrow is essential to

D

FIG 30-16  Osteolytic breast cancer metastases. A, Sagittal CT shows multilevel destructive vertebral body and posterior element lytic lesions with compression fractures. B, On sagittal T1-weighted image, the lesions are hypointense to intervertebral disk. C, Postcontrast T1-weighted image shows lesion enhancement. D, Axial T2-weighted image shows a focal right posterior vertebral body mass extending into the epidural space (arrow) but sparing the anterior epidural raphe.

http://radiologyebook.com

854

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 30-17  Osteoblastic breast cancer metastases. A, Sagittal CT shows multilevel areas of vertebral body and posterior element nodular and mottled sclerosis with associated compression fractures. B, On sagittal T1-weighted image, the lesions are hypointense to intervertebral disk. C, On sagittal T2-weighted fatsuppressed image, the lesions are hyperintense to marrow. Compression fractures at L1 and L3 are likely pathologic, given the multilevel lesions and signal abnormalities in their spinous processes.

distinguish hematopoietic marrow from spinal metastatic disease.227 On unenhanced fat-saturated T1-weighted sequences, spinal metastases have mixed to high signal intensity, whereas nonneoplastic  lesions have homogeneous low signal intensity. Postcontrast fatsuppressed T1 imaging is very helpful to identify enhancing tumor244 (see Fig. 30-16C). On T2-weighted images, vertebral metastases are usually hyperintense to bone marrow (see Fig. 30-17C). Strong indicators of metastases include a rim of T2 hyperintensity and diffuse T2 hyperintensity. Conversely, a focus of T2 hyperintensity in the center of an osseous lesion has been shown to be a specific sign of normal hematopoietic marrow.225 Blastic metastases are hypointense on all MR sequences. STIR imaging increases the conspicuity of T2 hyperintense vertebral metastases, but also increases imaging time—a concern in patients with pain, who cannot hold still for long periods. Together, STIR and unenhanced T1-weighted sequences are most useful for detecting malignant bone marrow abnormalities.228 Compression of the spinal cord or cauda equina are dreaded complications of metastatic spinal disease that occur secondary to pathologic compression fracture(s) or epidural mass(es). Symptomatic spinal cord compression is present in 10% to 20% of patients with metastatic spinal disease.159 In general, benign and pathologic vertebral compression fractures both result in marrow edema that is hypointense on T1- and hyperintense on T2-weighted images. Findings suggestive of a pathologic fracture include paravertebral mass, signal abnormality in the posterior elements, convex posterior vertebral body border, epidural mass, and other spinal metastases152 (see Fig. 30-17C). Pathologic compression fractures have been shown to be hyperintense to normal vertebral bodies on DWI sequences, whereas benign compression fractures  were hypo- to isointense.92 Furthermore, lower ADCs in pathologic fractures have been demonstrated.256 MRI findings suggestive of osteoporotic compression fractures include a low signal band on T1- and

T2-weighted images, normal marrow signal in the vertebral body, retropulsion of a posterior bone fragment, and multiple compression fractures.152 The MR fluid sign, a well-circumscribed intravertebral fluid collection, has also been described in avascular necrosis of the vertebral body (Kummel’s disease) and benign osteoporotic vertebral fractures. The intravertebral vacuum cleft sign, a gas-density cleft within a transverse separation of the vertebral body, has been described as a specific sign of vertebral osteonecrosis and not seen in other causes of pathologic fracture.176,183 Spinal perfusion MR studies show increased blood flow and volume in tumors. Compression of neural structures also results from epidural metastatic disease. Bone destruction at the level of epidural tumor involvement is present in most cases.158 Contrast-enhanced T1-weighted imaging is essential in delineating the extent of epidural tumor and cord compression. Epidural metastatic lesions are usually T1 hypointense, T2 hyperintense, and avidly enhancing. The anterior epidural raphe is often spared, resulting in the “draped curtain” appearance as vertebral body tumor extends on its sides into the epidural space228 (see Fig. 30-16D). Understanding the posttreatment imaging changes of vertebral metastases is essential in the management of patients. Progressive sclerosis of a lytic metastasis generally indicates a positive response to treatment. Evaluating the response of sclerotic metastases is difficult because progression of disease with osteolytic conversion may appear similar to a positive response to treatment. Persistent low T1 signal in bone marrow after treatment may represent residual tumor or fibrosis. In difficult cases such as this, DWI may play a role in evaluating treatment response; DWI may show decreased signal intensity of vertebral metastatic disease with successful therapy.108

Multiple Myeloma

Epidemiology.  Multiple myeloma is the most common primary malignancy of bone and primary vertebral tumor.184 It primarily

http://radiologyebook.com

CHAPTER 30  Spinal Tumors occurs between the ages of 40 and 80 years.168 The lesions of multiple myeloma most frequently occur in the axial and proximal appendicular skeleton.141 One large series of patients showed that the thoracolumbar spine (15%), cervical spine (13%), and pelvis (12%) were most commonly affected.130

Presentation and Natural History.  When multiple myeloma is symptomatic, it can be characterized by one or more clinical manifestations denoted by the acronym CRAB: calcium elevation, renal insufficiency related to multiple myeloma, anemia, and bone abnormalities such as lytic lesions and osteopenia. Multiple myeloma commonly presents with fatigue, back pain, or rib pain with associated fractures or infection. However, patients may be asymptomatic and only show anemia or hyperproteinemia. When multiple myeloma is suspected, a radiographic skeletal survey and bone marrow aspiration and biopsy are performed.84,202 Other forms of multiple myeloma include POEMS syndrome, which is characterized by polyneuropathy, organomegaly, endocrinopathy, high M-protein levels, and skin changes.141 Solitary plasmacytoma, accounting for less than 5% to 10% of plasma cell neoplasms, is associated with a better prognosis and may be cured with radiation therapy alone.157 Multiple myeloma patients are staged using the Durie-Salmon system, which estimates tumor burden. The system is based on amount of M protein, serum hemoglobin, serum calcium, renal function,  and number of lytic lesions on imaging.128,129 Prognosis of multiple myeloma is dependent on stage, with survival ranging from a few months to more than 10 years.89 New developments in chemotherapy and autologous stem cell transplant have led to improvement of patient survival.169 Percutaneous vertebroplasty and kyphoplasty are sometimes used to treat pain associated with spinal fracture.84

Pathophysiology.  Multiple myeloma is a low-grade proliferation of monoclonal malignant plasma cells that usually originate from bone marrow.122 These abnormal cells produce monoclonal protein (M protein), which can be detected in blood and urine.141 The exact cause of multiple myeloma is unknown; however, literature suggests that monoclonal malignant plasma cells arise from a late cell in B-cell maturation. Radiation exposure increases the risk of multiple myeloma.84 Multiple myeloma is part of the spectrum of plasma cell dyscrasias. A plasmacytoma, a related entity, is a solitary plasma cell tumor without

A

855

systemic spread. This solitary form mostly occurs in the vertebra and skull.157 When these malignant plasma cells invade bone marrow, they induce secretion of osteoclast-activating factors, leading to excessive bone resorption and decreased bone formation. The excessive osteoclastic activity promotes multiple myeloma clone proliferation, leading to a vicious cycle of tumor growth and bone destruction.212,240

Imaging.  Four radiologic presentations of multiple myeloma skeletal involvement have been described: (1) the solitary lesion (plasmacytoma), (2) diffuse skeletal involvement (myelomatosis), (3) diffuse skeletal osteopenia, and (4) sclerosing myeloma.139 Plasmacytomas are typically lytic lesions that can be found in the spine, ribs, sternum, and proximal appendages.173 Sclerotic lesions can be seen in POEMS syndrome.84,141 Most patients with multiple myeloma will have a positive radiographic finding; however, many lesions can be missed by radiograph.173 Multidetector CT (MDCT) is more sensitive than skeletal survey radiographs for evaluating the extent of bone disease.182 CT findings include punched-out lytic lesions, expansile lesions with soft tissue mass, diffuse osteopenia, fractures, and rarely osteosclerosis84 (Figs. 30-18A and 30-19A). The combination of MDCT and clinical follow-up is very useful in posttreatment evaluation.144 One major advantage of CT is distinct delineation of cortical bone to evaluate vertebral bodies at risk for pathologic fracture, which occurs in 55% to 75% of patients.172,182 A spinal lytic lesion with volume loss greater than 50% of the vertebral body is at risk for fracture182 (see Fig. 30-18A). MRI is more sensitive than radiographs in detecting multiple myeloma lesions in the spine and pelvis.171 STIR and T2-weighted imaging are the most sensitive MRI sequences for identifying marrow infiltration.141 Normal bone marrow signal is seen in 28% of cases, which corresponds to minimal plasma cell infiltration on histology (see Fig. 30-19B). Focal myeloma infiltration, seen in 30% of cases, is characterized by circumscribed lesions that are hyperintense on T2-weighted sequences (see Fig. 30-18C) but hypointense to intervertebral disks on T1-weighted images (see Fig. 30-19C). Diffuse marrow infiltration can also be seen. Intermediate-grade marrow infiltration may be difficult to identify on noncontrast T1-weighted images alone; in these cases contrast enhanced T1-weighted images may be helpful— greater than 40% enhancement is considered pathologic93 (see Fig. 30-19D). The same MRI criteria used to distinguish metastatic pathologic vertebral compression fractures from benign ones can also be applied to multiple myeloma.125

B

C

FIG 30-18  Plasmacytoma. A, Axial CT with bone windows shows an expansile destructive, lytic lumbar vertebral body mass. Greater than 50% of this vertebral body is involved; thus, this vertebral body is at risk for fracture. B, Postcontrast axial T1-weighted image further delineates the expansile soft tissue mass with paravertebral (arrow) and epidural (arrowhead) extension. C, On sagittal T2-weighted image, the mass is hyperintense to marrow.

http://radiologyebook.com

856

PART II  CT and MR Imaging of the Whole Body

A

C

B

D

FIG 30-19  Multiple myeloma. A, Sagittal CT shows expansile lytic destruction of L4 and L5 with soft tissue mass. B, On sagittal T2-weighted image, the soft tissue mass is iso- to hypointense to marrow. This may indicate low plasma cell infiltration. C, On sagittal T1-weighted image, the soft tissue mass is slightly hypointense to disk. D, On postcontrast sagittal T1-weighted fat-saturation image, vertebral body and intervertebral disk involvement is better delineated with contrast.

After treatment, patients with complete remission can show resolution of marrow signal abnormality or persistent marrow signal abnormality without enhancement or with residual peripheral rim enhancement.195 On T2 and proton density weighted images, lesions may have uniform high signal intensity or only rimlike high signal intensity as healing occurs.170 In partial response to therapy, diffuse marrow abnormality decreases in amount and converts to a variegated or focal pattern with persistent enhancement.195 In general, marrow lesions that return to baseline signal intensity on fluid-sensitive images and have fat signal intensity on T1-weighted images should be considered to contain no active disease. However, it may take as long as 5 years for marrow signal to return to normal. Conversely a patient may still have residual disease in the absence of imaging findings. Poor response to treatment appears as no change or worsening of previously identified lesions. In disease progression, focal infiltration can change into diffuse infiltration, mainly in the spine and pelvis.248

Primary Spinal Lymphoma

Epidemiology.  Spinal lymphoma is usually secondary to disseminated lymphoma.239 However, primary spinal lymphoma, which by definition occurs in the absence of extraspinal lymphomatous foci, is relatively rare. Osseous vertebral body and epidural lymphomas are the most common forms of spinal lymphoma, whereas intradural extramedullary and intramedullary deposits are less common.142 Primary vertebral lymphoma accounts for up to one third of all primary lymphomas of bone.209 Primary lymphoma of bone accounts for 5% of extranodal non-Hodgkin’s lymphoma and 7% of primary bone tumors.135 Primary spinal epidural lymphoma occurs in 0.9% to 6.5% of cases of non-Hodgkin’s lymphoma and accounts for 10% to 30% of epidural spinal tumors.106 Primary spinal epidural lymphoma accounts for 0.1% to 3.3% of all lymphomas.142 Primary spinal epidural lymphoma predominantly affects patients with a median age of 55 to 65 years, but has also been described in pediatric patients.191,192 One series of 94 cases of spinal extradural lymphoma showed that the thoracic spine was most commonly involved (69%), followed by the lumbar spine (27%), and the cervical spine (4%).140

Presentation and Natural History.  Signs and symptoms of spinal lymphoma include weakness, back or neck pain, sensory deficits, and incontinence. Cord compression can be seen in up to 7% of cases.180 The classic “B” symptoms of lymphoma (fever, weight loss, and night sweats) are less common in primary spinal lymphoma.140,191

Primary spinal lymphoma has a potentially favorable prognosis  if diagnosed and treated early, since the disease is very sensitive to radiation and chemotherapy.106,191 In fact, the prognosis of cord compression due to epidural lymphoma is far better than that of cord compression due to metastatic carcinoma.140,180 Combined modality treatment, including radiation therapy and chemotherapy, is the most efficient treatment for primary spinal epidural lymphoma, with reported rates of local control in 88% and 5-year overall survival in 69% of cases. The role of surgery has been reduced and limited to laminectomy or even biopsy only.191

Pathophysiology.  Lymphoma is a lymphoid neoplasm classified into two broad groups based on pathology: Hodgkin’s disease and non-Hodgkin’s lymphoma.163 Non-Hodgkin’s lymphoma is eight times more common than Hodgkin’s lymphoma.239 Several theories describe how epidural lymphoma may arise. It may come from lymphoid tissue in the epidural venous plexus or “lymphoid rests” in the epidural space.100,127,132 Hematogenous or lymphatic spread to the epidural space from an unidentified source may occur.109 Epidural extension from the vertebral body lymphoid tissue through cortical vascular channels may occur. Infiltrative growth through neural foramina and paravertebral extension are common growth patterns in spinal lymphoma. In fact, epidural infiltration rather than osseous disease is usually the cause of spinal cord compression.136,239,245 Infiltration of only the spinal epidural space, with multiple vertebral level involvement, is more frequent in lymphoma relative to metastatic disease.106 Knowledge of this tendency is helpful when characterizing epidural masses.

Imaging.  Radiographs are not sensitive in detecting vertebral and epidural lymphoma. Findings are nonspecific and include compression fractures and pedicle erosion. Vertebral body lesions may appear sclerotic, lytic, or mixed. Sclerosis (i.e., ivory vertebra) and mixed patterns are more common in Hodgkin’s disease.106,140 CT findings of spinal epidural lymphoma are nonspecific. A common finding on CT is epidural tumor with adjacent soft tissue or nodal mass. Epidural tumor is usually homogeneously hyperdense and enhances with intravenous contrast (Fig. 30-20A). Osseous involvement on CT is a relatively less common finding.180 MRI findings of vertebral lymphoma are also nonspecific. On T1-weighted images, lymphomatous infiltration of marrow is usually heterogeneous and diffuse. Focal marrow lesions are less frequent (Fig. 30-21A). The abnormal marrow is hyperintense on T2-weighted

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

*

*

* A

C

B

857

D

FIG 30-20  Epidural lymphoma. A, Postcontrast sagittal CT shows hyperdense infiltrative mass (asterisk) “capped” superiorly and inferiorly by relatively hypodense epidural fat. B, On sagittal T2- (B) and T1-weighted (C) images, the epidural mass (asterisk) is hypointense to fat. D, On postcontrast T1-weighted fat-saturation image, the epidural mass enhances.

primary spinal bone tumors, chordoma was the second most common tumor (multiple myeloma/plasmacytoma was most common), accounting for 22% of cases.156 The peak incidence is between ages 50 and 60. Men are affected twice as often as women.189 Chordomas have an almost equal distribution in the sacrococcygeal region, sphenooccipital region, and in the vertebrae.189 However, some studies have shown that sacrococcygeal chordomas are most common, accounting for up to 60% of all chordomas. In fact, chordomas constitute over 50% of primary tumors of the sacrum.113 In the mobile spine, the lumbar spine is the most common site of involvement, followed by the cervical, and then thoracic spine. The vertebral body is most commonly involved.101

Presentation and Natural History.  Chordomas are slow-growing

A

B

FIG 30-21  Vertebral lymphoma. A, On sagittal T1-weighted image, nonspecific multilevel diffuse marrow infiltration is seen. Focal marrow lesions are less frequent. B, On sagittal STIR image, the abnormal marrow is hyperintense.

images (see Fig. 30-21B) and enhances on contrast-enhanced T1-weighted images.194 Lymphomatous epidural infiltration is homogenous on MRI. Lesion contrast is greatest on T1-weighted images, where epidural lymphoma is hypointense to fat but slightly hyperintense to muscle (see Fig. 30-20C). On T2-weighted images, epidural lymphomas are iso- to hypointense to fat and hyperintense to muscle (see Fig. 30-20B). This may help distinguish lymphoma from metastatic disease, since most metastatic lesions are hyperintense to fat on T2-weighted images. Contrast-enhanced T1-weighted images highlight extraosseous soft tissue mass, paravertebral infiltration, and levels of spinal cord compromise203 (see Fig. 30-20D).

tumors that are locally aggressive and invasive; thus, they are usually large when discovered. Vertebral or sacral chordomas usually present with pain, often radicular. Advanced disease can lead to incontinence. Cervical chordoma can cause airway obstruction or dysphagia. Sacral chordomas can invade pelvic structures anteriorly and cause symptoms such as constipation or tenesmus.96,236 Median survival has been shown to be 6.3 years, with 5-year, 10-year, and 20-year survivals of 67.6%, 39.9%, and 13.1%, respectively. Up to 5% of chordomas have metastatic disease to the lungs, bone, skin, or brain at initial presentation. In advanced-stage chordomas, up to 65% of cases are metastatic.99,112 With current treatments, most chordomas will recur or progress. Recurrence is due to subtotal excision or violation of the tumor capsule with resultant seeding of   the surgical wound.153 Thus, en bloc resection is the mainstay treatment.145,234,243 The role of radiation therapy is unclear.103 Chordomas are insensitive to conventional chemotherapies.87,254

Pathophysiology.  In 1857, Virchow first described the physaliferous cells—cells with cytoplasm replaced by bubble-like vacuoles—that are characteristic of chordoma.250 Strong evidence suggests that chordomas arise from notochordal remnants located in vertebral bodies and throughout the axial skeleton.115,218,230,247 Chordomas have three histologic variants: classical (conventional), chondroid, or dedifferentiated. Distinguishing chondroid chordomas from chondrosarcomas histologically can sometimes be difficult.117

Chordoma

Imaging.  Chordomas are usually midline expansile lesions with

mas are the most common primary malignant bone tumor of the sacrum and mobile spine.189 In one series of benign and malignant

osseous destruction and soft tissue mass (Fig. 30-22). The vertebral body is most commonly involved in spinal chordomas. Radiographic findings include calcifications (30%-70%), osteosclerosis (43%-62%),

Epidemiology.  Excluding lymphoproliferative neoplasms, chordo-

http://radiologyebook.com

858

PART II  CT and MR Imaging of the Whole Body

invasion of adjacent intervertebral disks and vertebrae (11%-14%), and vertebral collapse.238 Chordomas are sharply marginated on CT (Fig. 30-23), which may be secondary to a fibrous pseudocapsule that can be seen pathologically. Spinal chordoma destroys the vertebral body, with associated soft tissue mass usually extending anterior or lateral to the spinal column. Amorphous calcifications scattered sparsely and randomly within the tumor may be present. Additionally, the soft tissue mass may contain solitary or multiple hypodense areas with septation, which may correspond to myxoid gelatinous material that can be seen on gross examination of tumor.121,190,235 On MRI, chordomas have low to intermediate signal intensity on T1- and very high signal on T2-weighted images owing to their high

water content. Most chordomas have low-signal fibrous septa that divide the gelatinous components of the tumor, seen on T2-weighted images.201 Heterogeneous enhancement with ring or arc pattern is characteristic (Fig. 30-24). Posterior extension of tumor into the epidural space is common. Multilevel vertebral involvement is possible.233 Recurrent chordomas have a tendency to involve the posterior elements.190

Chondrosarcoma

Epidemiology.  Chondrosarcoma is the second most common nonlymphoproliferative primary malignant spinal tumor in adults after chordoma. These tumors account for up to 25% of primary malignant spinal neoplasms and 7% to 12% of all primary spinal tumors.104,118

*

FIG 30-22  Sacral chordoma. Frontal radiograph shows loss of sacral

FIG 30-23  Sacral chordoma. Sagittal CT shows a sharply marginated

arcuate lines on the right with associated expansile soft tissue density (asterisk).

sacral mass with presacral, sacral canal, and posterior element extension.

S

A

S

B

S

C

FIG 30-24  Sacral chordoma. A, Sagittal T2-weighted image shows very high signal within the sacral chordoma due to high water content. Low-signal fibrous septa are also seen. B, On sagittal T1-weighted image, the tumor has low to intermediate signal intensity. C, On postcontrast sagittal T1-weighted image, this chordoma shows heterogeneous enhancement.

http://radiologyebook.com

CHAPTER 30  Spinal Tumors Patients are usually between age 33 and 51 years, with a 4 : 1 male predilection.104 Spinal chondrosarcoma most commonly occurs in the thoracic spine (60%), followed by the lumbar spine (20%-39%), and the cervical spine (19%-20%).116 They usually arise from the posterior elements.143,232 Sacral chondrosarcomas tend to involve the upper sacrum and can extend into the sacroiliac joint.177

Presentation and Natural History.  The most common symptom is focal pain that may gradually worsen. Up to half of patients present with radiculopathy, myelopathy, or cauda equina syndrome. Posterior element involvement may result in a palpable neck or back mass.116,253 The 10-year survival rates for grade I, grade II, and grade IV tumors are 90%, 65% to 80%, and 30% to 40%, respectively.116 Although technically difficult, complete en bloc surgical resection is the ideal treatment for chondrosarcoma. Local curettage almost always results in recurrence.150,253 Chondrosarcomas are generally resistant to chemotherapy and radiation therapy.188

Pathophysiology.  Most chondrosarcomas are primary; however, secondary chondrosarcomas can arise from a preexisting benign osseous or cartilaginous tumor (e.g., osteochondroma).220 The tumor produces cartilage that surrounds and invades the osseous matrix and fills marrow spaces. Chondrosarcoma histologic subtypes include conventional, mesenchymal, clear cell, and dedifferentiated chondrosarcomas. Most chondrosarcomas are of the conventional subtype, which are mostly low grade.118,177

Imaging.  Radiographs of spinal chondrosarcoma typically show lytic or expansile osseous destruction, usually arising from the posterior elements.177,188,197 Ring and arc calcifications of chondroid matrix mineralization are present in up to 70% of patients232 (Figs. 30-25 and 30-26A). Involvement of adjacent vertebral bodies through the disk or extension to adjacent ribs may be seen.197 On CT, these tumors are usually hypodense to skeletal muscle because of the water content of chondroid188 (see Fig. 30-26A). Sometimes the chondroid calcification may be amorphous, making some chondrosarcomas difficult to distinguish from osteoblastic tumors.177 The high water content of chondrosarcomas make them hypointense on T1- and hyperintense on T2-weighted images.154,188 Mineral-

859

ized cartilage appears as low signal foci. Low-signal septa may be seen on all MRI sequences. Various enhancement patterns (e.g., line or septa-like, diffuse, heterogeneous) have been described (see Fig. 30-26B-D). Since most chondrosarcomas are low grade, adjacent reactive bone marrow is rarely seen. Extradural extension with cord compression is often present. In some cases, chondrosarcoma can present as a dumbbell-shaped mass with neural foramina widening.177

Osteosarcoma

Epidemiology.  Spinal osteosarcoma may be the second or third most common malignant primary spinal tumor excluding lymphoproliferative neoplasms.156,215 Spinal osteosarcoma has a peak incidence in the fourth decade of life, slightly later than that of appendicular osteosarcoma,193 with a slight male predilection.215 The lumbosacral spine is the site of involvement in 60% to 70% of cases, with thoracic vertebral body being rare.238 However, one series showed equal involvement of the thoracic and lumbar spine.149 Primary vertebral body involvement is most common, accounting for up to 90% of cases.236

Presentation and Natural History.  Osteosarcomas almost always present with pain. Two thirds of patients have some neurologic impairment. Patients may also present with a palpable growing mass.155 One series reported that 28% of primary spinal osteosarcomas present with metastasis, usually to the lung.155,196 The prognosis of spinal osteosarcoma is poorer than appendicular osteosarcoma,155 with a median survival of 11 months.114,196 Wide en bloc excision is the most effective treatment. Neoadjuvant chemotherapy is given to decrease the possibility of local recurrence. Radiation therapy can be used in a neoadjuvant or adjuvant role.155,236

Pathophysiology.  The WHO classifies osteosarcoma into several histologic subtypes, all with the common feature of osteoid production by malignant osteoblasts. Most osteosarcomas are of the con­ ventional osteoblastic subtype, which are high grade.155 Secondary osteosarcomas can be associated with Paget’s disease, although malignant transformation of a spinal lesion is rare.197,238 Radiation-induced spinal osteosarcoma can also occur, with a latency of 5 to 20 years.155,197 The incidence of osteosarcoma is increased in several genetic disorders, such as hereditary retinoblastoma and Li-Fraumeni syndrome.155

Imaging.  The vertebral body is most commonly involved in spinal

FIG 30-25  Cervical chondrosarcoma. Frontal radiograph shows osseous destruction of the right C5-C6 facets with associated ill-defined soft tissue calcifications.

osteosarcoma. Most lesions have a sclerotic matrix, sometimes appearing as an “ivory” vertebral body. Lytic lesions are infrequent and classically associated with the telangiectatic subtype.149,155,197 A mixed pattern may be found in up to 20% of cases. Up to 5% of cases will have no radiographic abnormality.204 Cortical destruction, expansile remodeling, aggressive periosteal reaction, and pathologic fractures are better seen with CT. CT will show matrix mineralization in up to 80% of cases155,200 (Fig. 30-27A). Loss of vertebral body height with sparing of the adjacent disk is common.238 Multilevel vertebral body involvement can be seen.149 On MRI, these tumors have low to intermediate signal on T1- and high signal on T2-weighted images. Areas of mineralization are low signal on all sequences197 (see Fig. 30-27B and C). Heterogeneous enhancement is seen238 (see Fig. 30-27D and E). Invasion of the spinal canal is typical.149 However, if there is no prominent soft tissue mass, osteosarcoma may be indistinguishable from an osteoblastoma. Telangiectatic osteosarcoma may have a similar appearance to aneurysmal bone cyst. Thick nodular cyst walls, matrix mineralization, and a more aggressive growth pattern suggest a telangiectatic osteosarcoma over an aneurysmal bone cyst.149,200,241

http://radiologyebook.com

860

PART II  CT and MR Imaging of the Whole Body

B

A

C

D

FIG 30-26  Cervical chondrosarcoma. A, Axial CT shows an expansile, destructive, hypodense mass centered to the right of the cervical vertebrae, with classic ring and arc calcifications. B, On precontrast axial T1-weighted image, the mass is hypointense with an epidural component (arrow). C, On postcontrast axial T1-weighted image, the mass demonstrates linear/septa-like enhancement. The extent of epidural extension is better delineated with contrast (arrow). D, On sagittal T2-weighted image, the mass is very hyperintense.

Ewing’s Sarcoma and Primitive Neuroectodermal Tumor

Epidemiology.  Ewing’s sarcoma and primitive neuroectodermal tumor (PNET) have similar clinical, histologic, and radiologic characteristics and are thus, discussed together. They are the most common non-lymphoproliferative primary malignant tumors of the spine in children. Overall, primary spinal Ewing’s sarcoma and PNET are rare, accounting for 3% to 10% of all Ewing’s sarcomas and PNETs, respectively. Secondary spinal Ewing’s sarcoma or PNET is much more common. At the time of presentation, patients are usually between age 10 and 30 years, with a male predominance.133,197,238 The sacrum is most frequently involved (55% of cases), followed by the lumbar spine (25%), and cervical spine (3%).148 These tumors are typically centered in the vertebral body,133,197,238 although posterior element involvement is often seen.148

Presentation and Natural History.  Patients usually develop pain, paresthesia, weakness, and/or incontinence over several weeks to months. Associated fever and leukocytosis may be present.131,133,197,238 Long-term survival is dependent on tumor size at presentation. One series showed a median survival of 26 months.229

Aggressive chemotherapy followed by radiation therapy is the mainstay of treatment for spinal Ewing’s sarcoma and PNET.133 In nonsacral tumors, local control is achieved in nearly all cases. Longterm survival is achieved in over 70% of patients. Sacrococcygeal tumors have a worse prognosis, with only 60% local control and 25% long-term survival, likely related to larger tumor sizes at presentation.229 Patients with instability and neurologic compromise may need surgical decompression and stabilization.133,229

Pathophysiology.  Histologically, Ewing’s sarcoma and PNET consist of high-grade, undifferentiated, small, round blue cells with high nucleus-to-cytoplasm ratios. Both tumors have similar cytogenetic features, including chromosome 11 and 22 translocations. Immunohistochemical studies are needed to distinguish the two entities, with PNET showing neural differentiation. Ewing’s sarcoma is thought to be the primitive stage of PNET. Many of these tumors show vascular endothelial proliferation.131,133,197 Areas of osteonecrosis are often seen in spinal lesions.231

Imaging.  One large series showed that lytic bone destruction was the most common radiographic finding in spinal Ewing’s sarcomas (92%

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

861

B

A

C

E

D

FIG 30-27  Cervical osteosarcoma. A, Axial CT shows an expansile mass centered in the right lamina/facet with sclerotic mineralized matrix. B, On axial T1-weighted image, the mass has intermediate signal and a right epidural component. The spinal cord is displaced and compressed toward the left. C, On axial T2-weighted image, the mass has high signal. Areas of mineralization are hypointense on both T1 and T2 images. Sagittal pre- (D) and postcontrast (E) T1-weighted images, show heterogeneous enhancement within the mass.

A

C

B

FIG 30-28  Thoracic Ewing’s sarcoma. A, Axial CT shows prominent right paravertebral mass with associated right vertebral body, pedicle, and transverse process lytic destruction. The paraspinal component is larger than the osseous component. B, On axial T2-weighted image, the mass is heterogeneously hyperintense. C, On postcontrast axial T1-weighted image, the mass shows homogeneous enhancement.

of cases)148 (Fig. 30-28A). However, another smaller series showed that sclerosis was the most common finding (69% of cases), which was associated with osteonecrosis.231 Other radiographic findings include vertebra plana, ivory vertebra, and pseudohemangioma. Paraspinal soft tissue mass and spinal canal invasion, both common findings, can

be seen on CT. The paraspinal component is often larger than the osseous component.95,148 On MRI, these tumors usually have intermediate signal intensity on T1- and intermediate to high signal intensity on T2-weighted images (see Fig. 30-28B). These tumors can have heterogeneous signal

http://radiologyebook.com

862

PART II  CT and MR Imaging of the Whole Body

A

B

D

E

C

F

FIG 30-29  Classic lumbar hemangioma. A, Sagittal CT demonstrates a “corduroy” appearance of the vertebral body. B, Axial CT shows a “polka-dot” appearance of the vertebral body. On both axial T1- (C) and T2-weighted (D) images, this mass is hyperintense due to high fat content. Sagittal T2 (E) and STIR (F) images confirm that the mass follows fat signal.

due to hemorrhage, calcification, or necrosis.197 Marked homogeneous enhancement can be seen249 (see Fig. 30-28C). Young patients normally have abundant red marrow in the spine, which is problematic for imaging, because both tumor and red marrow have low signal on T1-weighted images. Thus, MRI may be inferior to CT in depicting lesions in young patients with abundant spinal red marrow. Gadolinium may help in these cases because tumor enhancement is higher than that of red marrow.133,227

Hemangioma

Epidemiology.  Hemangioma is the most common benign tumor of the spine in adults, accounting for up to 30% of primary vertebral tumors.114 The incidence of benign vertebral hemangiomas is estimated to be over 10% in autopsy studies.91 They can be found at any age, with peak incidence between the fourth and sixth decades.114 These tumors most frequently occur in the thoracic spine, followed by the lumbar spine.151,198 Vertebral body involvement is most common, although lesions can extend to the pedicles and rarely to the spinous processes.246 Multiple vertebral hemangiomas are seen in 25% to 30% of cases.237

Presentation and Natural History.  Most vertebral hemangiomas are small, asymptomatic, and found incidentally. Rarely, lesions become symptomatic by hemorrhage, vertebral body expansion, or neural compression. Myelopathy is the most common symptom, followed  by radiculopathy.151 Neck or back pain often precedes neurologic symptoms.126 Asymptomatic vertebral hemangiomas do not require treatment. Proven treatment modalities for aggressive, symptomatic vertebral hemangiomas include surgery, vertebroplasty, direct ethanol injection, transarterial embolization, radiation therapy, or a combination of these treatments.126,151 Aggressive vertebral hemangiomas behave like benign tumors, so treatment decisions should be focused on symptoms.88

Pathophysiology.  Hemangiomas are composed of endothelial-lined vascular spaces that replace cancellous bone and intersperse among the trabeculae, with remaining ones becoming thick. These tumors have variable amounts of fat. Aggressive vertebral hemangiomas can extend into the epidural or paravertebral space.126,134

Imaging.  On radiographs, classic vertebral hemangiomas present as an osteopenic vertebral body with coarse, thickened, vertical trabeculae—the “corduroy” appearance (Fig. 30-29A). The CT correlate of this corduroy appearance is the “polka-dot” appearance seen on axial images198 (see Fig. 30-29B). However, these findings may be absent in some cases.120 In most cases, CT shows fat-attenuation values in the marrow of affected vertebral bodies.133 The thickened trabeculae are low signal intensity on all MRI sequences. Areas of punctate signal abnormality are seen on this lowsignal background of trabecular thickening. Hemangiomas vary from having low to high signal on T1-weighted images, depending on the amount of fat present within the tumor. However, high signal on noncontrast T1-weighted images is highly suggestive of a hemangioma (see Fig. 30-29C); only a few other vertebral tumors may have this appearance (i.e., metastatic melanoma, exostosis). The vascular spaces of these lesions have high signal on T2-weighted images. On fatsuppression sequences, hemangiomas have variable signal depending on the fat and vascular composition of the tumor (Figs. 30-30 and 30-31; see Figs. 30-29C-F and 30-31A and B). These lesions enhance after contrast administration133,198 (see Fig. 30-31C and D).

Osteoid Osteoma

Epidemiology.  Osteoid osteomas and osteoblastomas are histologically similar benign tumors that produce osteoid and woven bone; they are the most common benign vertebral tumors in children, accounting for 10% to 12% of them.114 Vertebral osteoid osteomas usually occur between age 10 and 20 years, with up to a 3 : 1 male predilection.237 The

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

A

B

C

FIG 30-30  Small typical thoracic hemangiomas. Sagittal T1- (A) and T2-weighted images (B) show two focal marrow lesions in the same vertebral body, with high signal due to fat content. C, Sagittal STIR image confirms that these lesions follow fat signal.

A

B

C

D

FIG 30-31  Aggressive atypical lumbar hemangioma. A, Sagittal T1-weighted image shows a mildly hyperintense vertebral body hemangioma. B, However, on sagittal STIR image, the lesion retains high signal, indicating that there is relatively less fat content and more vascular component in this hemangioma. C, Postcontrast axial T1-weighted image shows the enhancing aggressive hemangioma extending into the epidural space but sparing the anterior epidural raphe. D, Angiogram with catheter positioned in a lumbar artery demonstrates the high vascularity of the hemangioma.

http://radiologyebook.com

863

864

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 30-32  Thoracic osteoid osteoma. A, Axial CT in bone windows demonstrates a lucent nidus (arrow) in the right lamina, with central calcification and mild surrounding sclerosis. B, On sagittal STIR image, the nidus (arrow) is hyperintense, with a focal central hypointense calcification and surrounding marrow and soft tissue edema. The nidus has intermediate signal on T1-weighted image (C) and enhances on postcontrast T1-weighted image (D). Mild enhancement is also seen in the soft tissues around the nidus (arrow).

most common location is the lumbar spine (59%), followed by the cervical, thoracic, and sacral vertebrae with frequencies of 27%, 12%, and 2%, respectively.137 The tumor most often involves the posterior elements (75%), namely the laminae (33%), articular facets (19%), and pedicles (15%). Isolated vertebral body involvement is rare.137

Presentation and Natural History.  The classic presentation is pain that is worse at night and relieved by aspirin. However, this is not always present in vertebral osteoid osteoma. These lesions can cause gait disturbance, radiculopathy, limb atrophy, and scoliosis.137 The goal of treatment is pain relief. Nonoperative management with nonsteroidal antiinflammatory drugs (NSAIDs) is a feasible management option. New therapies include percutaneous CT-guided removal or percutaneous ablation with radiofrequency electrode, laser, or alcohol. Indications for surgery include failure to control symptoms with NSAIDS, neurologic deficits, and requirement for pathologic diagnosis. The natural history of osteoid osteoma is variable, with symptomatic relief achieved in some cases.107,133

Pathophysiology.  On gross pathologic examination, the nidus of an osteoid osteoma is a small (<1.5 to 2 cm diameter), round mass of pink to red tissue owing to vascularity. Histologically, the nidus consists of trabecular bone with intertrabecular vascularized fibrous connective tissue.161 In most cases, this nidus is surrounded by a variable amount of reactive sclerotic bone. Osteoid osteomas usually do not grow.197

Imaging.  The classic radiographic appearance of osteoid osteoma is a small circumscribed area of osteolysis corresponding to the nidus, with surrounding reactive sclerosis. However, often only a sclerotic

pedicle is seen on radiographs. If scoliosis is present, the osteoid osteoma is usually at the concave portion of the curve apex. CT is the best modality for detecting osteoid osteoma. The nidus is a small, circumscribed, low-attenuation lesion that may or may not contain central calcification. The surrounding reactive sclerosis is further delineated with CT133,197,221,237 (Fig. 30-32A). On MRI, the nidus is usually low to intermediate signal on T1- and intermediate to high signal on T2-weighted images. Most cases will have marked edema in the surrounding bone marrow and soft tissue, best seen on T2-weighted images (see Fig. 30-32B and C). A heavily calcified nidus is low signal on all MRI sequences. Often, only edema is seen in and around the posterior elements without a clear nidus;  in these cases, CT is helpful in identifying the nidus. Marked  enhancement can be seen, especially if the nidus is not calcified.133,197,237 On MRI, the soft tissues around the lesion may also enhance after contrast administration, presumably secondary to inflammation (see Fig. 30-32D).

Osteoblastoma

Epidemiology.  The spine is the most common site of osteoblastomas, accounting for approximately 30% to 40% of all osteoblastomas.164 Osteoblastoma and osteoid osteoma, histologically similar benign tumors, are the most common benign vertebral tumors in children, accounting for 10% to 12%.114 However, osteoblastomas are four times less common than osteoid osteomas.107 Spinal osteoblastomas occur mostly in the second and third decades, with a 2 : 1 male predominance.164 These tumors are equally distributed in the cervical, thoracic, and lumbar spine.164 Posterior element involvement is most common.187

http://radiologyebook.com

CHAPTER 30  Spinal Tumors Presentation and Natural History.  Osteoblastomas often produce a dull localized pain, as opposed to the intense night pain seen with osteoid osteomas. Unlike osteoid osteomas, osteoblastomas often present with neurologic symptoms.197 The mainstay treatment of spinal osteoblastoma is surgical resection.197 Recurrence rates are 10% to 15%.164 Aggressive osteoblastoma has a greater recurrence rate that approaches 50%, probably because of their larger size.179 Recurrence can ultimately lead to death. Rarely, malignant transformation of an osteoblastoma to osteosarcoma and metastatic disease can occur.197

Pathophysiology.  Histologically, osteoblastoma is similar to osteoid osteoma, but less organized. Grossly, these tumors are larger than osteoid osteomas, usually greater than 1.5 to 2.0 cm in diameter. An aneurysmal bone cyst component may be seen in 10% to 15% of cases.224 Osteoblastomas typically grow slowly.197 However, a subgroup of aggressive osteoblastoma has been described, which is difficult to distinguish from osteosarcoma.186

Imaging.  Three radiographic patterns of osteoblastoma have been described and are best appreciated with CT. The first pattern, which is the most common, is that of an expansile lesion with multiple small

A

C

865

calcifications and a peripheral sclerotic rim. The second pattern is similar in appearance to that of an osteoid osteoma with a lucent nidus and surrounding sclerosis, except larger. The presence of osseous expansion and multifocal matrix mineralization instead of central calcification favors osteoblastoma over osteoid osteoma. The third pattern is that of an aggressive lesion with osseous expansion and destruction with soft tissue infiltration and matrix calcification (see Fig. 30-33A). This aggressive appearance may mimic a chondrosarcoma.133,197 Osteoblastomas also have a similar MRI appearance to osteoid osteomas. These tumors are low to intermediate signal on T1- and intermediate to high signal on T2-weighted images, with peritumoral marrow and soft tissue inflammation and edema. These tumors can homogeneously or heterogeneously enhance. MRI may also show extensive soft tissue infiltration and neural structure impingement133,237 (see Fig. 30-33B-E). The presence of soft tissue mass suggests an osteoblastoma over osteoid osteoma.133,197

Aneurysmal Bone Cyst

Epidemiology.  Spinal aneurysmal bone cyst (ABC) accounts for up to 30% of all ABCs160,162 and up to 20% of all primary spinal tumors.138 Patients are usually younger than age 30, with a slight female predominance.174,185 They most commonly arise in the lumbar spine, followed

B

E

D

FIG 30-33  Cervical osteoblastoma. A, Axial CT shows an expansile, lytic mass arising from the posterior elements, with multifocal matrix mineralization. On axial (B) and sagittal (C) T2-weighted images, the mass has intermediate to high signal intensity, with peritumoral edema. Central cord myelomalacia is present. D, On T1-weighted image, the mass has low to intermediate signal. On postcontrast T1-weighted image, the mass enhances with surrounding peritumoral enhancement.

http://radiologyebook.com

866

PART II  CT and MR Imaging of the Whole Body

A

B

D

E

C

FIG 30-34  Lumbar ABC. A, Radiograph shows destruction of the left pedicle and transverse process of L4 (arrow). Axial CT (B) and T2-weighted MRI (C) demonstrate an expansile mass with classic fluid-fluid levels. On the T2-weighted image a low signal rim is seen. On axial pre- (D) and postcontrast (E) T1-weighted images, the solid components of the ABC enhance. Areas on intinsic intratumoral T1 hyperintensity are compatible with hemorrhage. The ABC extends into the left lateral recess of the epidural space (arrow).

by equal involvement of the thoracic and cervical spine.119 Sacral involvement is rare.110 ABCs often involve the posterior elements, especially the lamina205; however, extension into the vertebral body is seen in most cases.160,162

Presentation and Natural History.  Pain and/or neurologic deficits are the most common presentation. Patients often have associated scoliosis or kyphosis.124 The clinical course is highly variable, with recurrence possible after all forms of treatment.105 Total surgical excision is the optimal treatment for local control of tumor and prevention of recurrence. Other treatments include curettage with or without bone grafting, arterial embolization, intralesional steroid or calcitonin injections, and radiation therapy. Vertebroplasty or kyphoplasty may be used to reinforce the bone after curettage. Recurrence occurs in 10% to 44% of cases, but is rare with complete excision.257

Pathophysiology.  ABC is a highly vascular benign, but locally aggressive lesion. Grossly, ABC has a characteristic appearance of a “blood-filled sponge” with multiloculated blood-filled cysts.162 The blood-filled spaces are not lined by endothelium, but surrounded by solid septations consisting of fibrous tissue, reactive bone, and giant cells.217 The solid component may predominate histologically in 5% to 7.5% of cases.97 In fact, this solid variant of ABC has a tendency to involve the spine.221 ABCs arise from a local circulatory disturbance and can be classified as primary or secondary. Secondary ABCs arise when an underlying neoplasm causes venous obstruction or arteriovenous fistulae. Giant cell tumor, osteoblastoma, chondroblastoma, and osteosarcoma can have associated secondary ABCs. Primary ABCs, which are the majority, do not have an underlying neoplasm and may be caused by trauma.162

Imaging.  Radiographs usually show a well-defined, radiolucent, expansile lesion surrounded by a thin sclerotic margin (Fig. 30-34A). A compression fracture may be seen. CT further delineates the expansile nature and soft tissue extent of ABC. The spinal lesions can extend into adjacent vertebral bodies, intervertebral disks, posterior ribs, epidural space, and paravertebral soft tissues. The multicystic architecture of ABC with fluid-fluid levels can be seen on CT in up to 35% of cases (see Fig. 30-34B). Fluid-fluid levels are a typical, but nonspecific finding for ABC.146,197,237 On MRI, ABCs are lobulated or septated with a thin well-defined low-signal rim on all sequences, which corresponds to periosteum. MRI is more sensitive than CT in detecting their multicystic appearance with fluid-fluid levels (see Fig. 30-34C). The signal characteristics of the cystic components vary depending on the age of hemorrhage within them. The solid components of ABC are usually high signal on T2-weighted images and enhance after contrast administration (see Fig. 30-34D and E). The imaging features of solid ABCs are nonspecific, except for the location.94,147,181,237

Giant Cell Tumor

Epidemiology.  Spinal giant cell tumor (GCT) accounts for up to 7% of all GCTs197 and 5% of all primary vertebral tumors.114 Patients are usually 20 to 40 years old, with a 2 : 1 female predilection. Most GCTs occur in the sacrum, followed by, in order of decreasing frequency, the thoracic, cervical, and lumbar vertebrae.193 GCT is the second most common non-lymphoproliferative sacral tumor after chordoma.215 Mobile spine GCT usually involves the vertebral body, although posterior element extension is common.133,197

Presentation and Natural History.  Patients usually present with pain or neurologic deficits. These tumors can increase in size during

http://radiologyebook.com

CHAPTER 30  Spinal Tumors pregnancy, likely secondary to hormonal stimulation.167,197 Spinal GCT has a poorer prognosis than GCT in tubular bones, because these tumors are larger and more difficult to excise completely. Recurrence rate is approximately 40% to 60%. En bloc resection is the optimal treatment, although it is often only possible in some sacral lesions. GCT not amenable to complete excision is usually treated with a combination of partial curettage and radiation therapy.210,226 Selective arterial embolization has been used before surgical excision or as the sole treatment in unresectable tumors. In one series, 12% of patients with spinal GCT developed pulmonary metastases. Local recurrence often precedes metastases.102 Malignant transformation is a concern in patients with GCT treated with radiation therapy. Studies of irradiated GCT in the spine, sacrum, and pelvis show a 17% rate of developing secondary sarcoma.222,223,242 Latency periods can vary from 6 months to 10 years or more.102

Pathophysiology.  GCTs are considered histologically benign, although they are locally aggressive and can metastasize102; thus, some classify GCT as a malignant tumor.114 Histologically, GCT consists of many multinucleated osteoclastic giant cells in a spindle-cell stroma.166 Malignant GCTs, seen in 5% to 10% of cases, have a sarcomatous stroma.211

A

Imaging.  Spinal GCT usually presents as a lytic lesion that causes expansion or collapse of a vertebral body.199 Sacral GCT is usually large, involves both sides of midline, and destroys sacral foraminal lines. However, some GCTs may be invisible on radiographs. On CT, GCT has soft tissue attenuation with a well-defined margin and thin sclerotic rim (Fig. 30-35A and 30-36A). Areas of hemorrhage and necrosis result in a heterogeneous appearance, with focal areas of low attenuation. Similar to appendicular GCT, spinal GCT shows no evidence of mineralized matrix. Extension into the posterior elements, across intervertebral disks, and through the sacroiliac joint can be seen.133,197 Epidural and paravertebral extension, which is very common in GCT, is well depicted on MRI.167 Regardless of the type of imaging sequence used, GCT usually has heterogeneous signal intensity. On T1-weighted images, these tumors have low to intermediate signal intensity. On T2-weighted images, most GCTs are hypo- or isointense to the normal spinal cord, which may be secondary to fibrous tissue and hemosiderin within them (Figs. 30-35B and 30-36B and C). This helps distinguish spinal GCTs from other spinal tumors, since many other tumors are hyperintense on T2-weighted images.85,211 Cystic components, similar to those seen in ABCs, with evidence of previous hemorrhage are common.197 Curvilinear low-signal areas within the vertebral body on both T1- and T2-weighted images have been

B

FIG 30-35  Sacral GCT. A, Axial CT shows a large, well-defined, lytic soft tissue sacral mass that crosses midline without mineralized matrix. The mass invades a right sacral foramen (arrow), which contains less fat relative to the left sacral foramen. B, On axial T2-weighted image, the mass is heterogeneously hypointense.

A

867

B

C

FIG 30-36  Lumbar GCT. A, Axial CT shows a small, well-defined, lytic soft tissue laminar mass (arrows) without mineralized matrix. On axial T2-weighted (B) and sagittal T2-weighted fat suppressed (C) images, the mass has intermediate signal intensity with prominent epidural component.

http://radiologyebook.com

868

PART II  CT and MR Imaging of the Whole Body

described, which may correspond to thickened trabeculae, fibrous septae, or hemosiderin deposition. GCTs are vascular and enhance.167

Eosinophilic Granuloma

Epidemiology.  Eosinophilic granuloma (EG) is a solitary lesion that most commonly occurs in bone and is the localized and most benign form of Langerhans cell histiocytosis.86 Vertebral involvement is seen in 7.8% to 25% of patients.252 EG accounts for approximately 12% to 25% of all primary spinal tumors.114 EG primarily affects children younger than age 15 and has a male predilection.98,206 Spinal EG most commonly affects the thoracic spine, followed by the lumbar,  and lastly, the cervical spine. The vertebral body is most commonly involved.86

Presentation and Natural History.  Patients usually have local pain that subsides with bed rest. Some spinal lesions may be asymptomatic. Neurologic complications are rare and usually mild. Other clinical findings include mild fever, elevated erythrocyte sedimentation  rate, slight eosinophilia, and leukocytosis. Given the clinical and  radiographic presentation, EG can be misdiagnosed as infectious spondylitis.252 Prognosis of spinal EG is generally excellent. Regardless of treatment method, most patients will have satisfactory, but not always complete reconstitution of vertebral body height. Treatment of EG is usually guided by the presence of neurologic deficits or spinal instability. Cases with typical clinical and radiologic findings can be managed conservatively with clinical and radiologic follow-up for several months. In these cases, biopsy is not advised. Most of the remaining normal tissues of a collapsed vertebral body are apophyseal plates, which may be damaged during biopsy, resulting in impaired reconstitution of vertebral body height. Patients with severe pain may be treated with bed rest and NSAIDs. If there is worsening of systemic

symptoms or concerning radiologic changes during observation, biopsy should be considered. Low-dose radiotherapy is recommended for patients who are paralyzed or have impending paralysis. Surgical decompression should be considered for cases in which neurologic compromise does not respond to radiotherapy or progresses rapidly.252

Pathophysiology.  Langerhans histiocytes are the dendritic antigenpresenting cells found in skin. Their clonal proliferation results in the various Langerhans cell histiocytosis syndromes, of which EG is the most localized and benign.214 Histologically, EG consists of sheets of Langerhans cells with a variable amount of lymphocytes, polymorphonuclear cells, and eosinophils. This proliferation of cells leads to lytic destruction of bone.133 Langerhans cells are characterized by the “tennis racquet–shaped” cytoplasmic Birbeck granule seen on electron microscopy.166

Imaging.  EG causes lytic destruction of the vertebral body that results in the classic vertebra plana appearance (Fig. 30-37A). Complete or incomplete collapse of the vertebral body is seen in almost all cases. Other sensitive radiologic findings include absence of an osteolytic area in the spinal lesion, preservation of the pedicles and posterior elements, preservation of the adjacent disks, absence of paravertebral soft tissue, and increased radiodensity of the collapsed vertebral body. Involvement of only one vertebral body is seen in approximately 50% of cases.214,252 Encroachment of EG into the spinal canal is well depicted with MRI. EG is isointense on T1- and hyperintense on T2-weighted images (see Fig. 30-37B and C). Marked enhancement is seen (see Fig. 30-37D), often more extensive than that of the normal red bone marrow. CT may be superior to MRI in detecting EG within bone marrow, given the prominence of spinal red marrow seen in children that can make detection of lesions relatively more difficult on MRI.133

A

B

C

D

FIG 30-37  Lumbar EG. A, Lateral radiograph shows the classic vertebra plana appearance. B, On sagittal T2-weighted image, an intermediate signal intensity focal anterior epidural mass is present at the level of the vertebra plana. The mass only spans one vertebral level. C, On sagittal T1-weighted image, the focal epidural mass has low signal intensity. D, The focal epidural mass enhances on postcontrast sagittal T1-weighted image.

http://radiologyebook.com

CHAPTER 30  Spinal Tumors Summary Spinal tumors are divided into three groups based on their anatomic locations: extradural, extramedullary-intradural, and intramedullary. Extradural spinal tumors are the most common spinal tumors in adults, accounting for 60% of spinal tumors. Metastatic disease from systemic cancer is the most common extradural spinal tumor in adults. Vertebral multiple myeloma is the most common primary extradural spinal tumor. Extradural spinal lymphoma usually arises secondary to disseminated lymphoma. These tumors are usually easily diagnosed on imaging, given the clinical history and laboratory data. Primary vertebral tumors, which constitute the majority of remaining extradural spinal tumors, are relatively more rare. Many of these primary vertebral tumors have almost specific imaging findings. Spinal chordomas are characterized by marked hyperintensity on T2-weighted MRI and their sacral predilection. Osseous destruction and matrix calcifications can suggest chondrosarcoma or osteosarcoma. Hemangiomas have a characteristic “corduroy” or “polka-dot” radiographic appearance with varying intratumoral fat. Osteoid osteomas are characterized by a calcified or noncalcified radiolucent nidus with surrounding reactive osseous sclerosis and soft tissue inflammation. Aneurysmal bone cysts have an expansile multicystic appearance with fluid-fluid levels. Giant cell tumors, which can be aggressive, are characteristically relatively hypointense on T2-weighted MRI. Eosinophilic granulomas classically present as vertebra plana in a pediatric patient. Knowledge of the characteristic imaging findings of primary vertebral tumors, along with age and clinical presentation of the patient, aids the radiologist in making a precise diagnosis. Extramedullary-intradural tumors are the second most common spinal tumors in adults, accounting for 20% to 30% of spinal tumors. In comparison to meningiomas, schwannomas demonstrate heterogeneous and peripheral enhancement and slightly more heterogeneous increased signal on T2-weighted MRI. Schwannomas are a common cause of vertebral scalloping and neuroforaminal widening. A distinct feature of meningioma is the presence of calcifications, a dural tail, and homogeneous enhancement. Intramedullary spinal tumors are the least common type of spinal tumor in adults, accounting for 4% to 10%. However, they are the most common spinal tumor in children, accounting for 35%. Ependymoma is the most common intramedullary spinal tumor in adults (associated with NF2), and astrocytoma is most common in children. On imaging, an ependymoma is a well-circumscribed hyperenhancing mass with hemorrhage and cystic components. Astrocytomas are usually infiltrative. The enhancement pattern of astrocytomas is less intense and illdefined compared to ependymomas; hemorrhage and cysts are rare. Hemangioblastomas (HBs) are associated with von Hippel-Lindau disease. They are subpial dorsal hypervascular lesions. HBs are commonly cystic with a hypervascular enhancing nodule or a solid hyperenhancing mass with flow voids. In HBs, the surrounding edema and associated syrinx could be very prominent. Spinal cord lymphoma is an infiltrating tumor that does not have an associated cystic or hemorrhagic component. The distinct imaging feature of spinal cord metastasis is the presence of single or multiple small focal hyperenhancing lesions with disproportionate extensive edema. The combination of these imaging features and presence of a primary malignancy are suggestive of spinal cord metastases.

REFERENCES Introduction/Intramedullary Spinal Cord Tumors 1. Al-Okaili RN, Krejza J, Woo JH, et al: Intraaxial brain masses: MR imaging-based diagnostic strategy–Initial experience. Radiology 243(2):539–550, 2007.

869

2. Camacho DL, Smith JK, Castillo M: Differentiation of toxoplasmosis and lymphoma in AIDS patients by using apparent diffusion coefficients. AJNR Am J Neuroradiol 24(4):633–637, 2003. 3. Chamberlain MC, Tredway TL: Adult primary intradural spinal cord tumors: A review. Curr Neurol Neurosci Rep 11(3):320–328, 2011. 4. Chang SD, Meisel JA, Hancock SL, et al: Treatment of hemangioblastomas in von Hippel-Lindau disease with linear accelerator-based radiosurgery. Neurosurgery 43(1):24–34, 1998. 5. Chason JL, Walker FB, Landers JW: Metastatic carcinoma in the central nervous system and dorsal root ganglia. A prospective autopsy study. Cancer 16(6):781–787, 1963. 6. Chu BC, Terae S, Hida K, et al: MR findings in spinal hemangioblastoma: Correlation with symptoms and with angiographic and surgical findings. AJNR Am J Neuroradiol 22(1):206–217, 2001. 7. Costigan DA, Winkelman MD: Intramedullary spinal cord metastasis. A clinicopathological study of 13 cases. J Neurosurg 62(2):227–233, 1985. 8. Crawford JR, Zaninovic A, Santi M, et al: Primary spinal cord tumors of childhood: Effects of clinical presentation, radiographic features, and pathology on survival. J Neurooncol 95(2):259–269, 2009. 9. Diamond C, Taylor TH, Aboumrad T, et al: Changes in acquired immunodeficiency syndrome-related non-Hodgkin lymphoma in   the era of highly active antiretroviral therapy. Cancer 106(1):128–135, 2006. 10. Dolecek TA, Propp JM, Stroup NE, et al: CBTRUS statistical report: Primary brain and central nervous system tumors diagnosed in the United States in 2005–2009. Neuro-Oncol 14(Suppl 5):v1–v49, 2012. 11. Dong Q, Welsh RC, Chenevert TL, et al: Clinical applications of diffusion tensor imaging. J Magn Reson Imaging 19(1):6–18, 2004. 12. Ducreux D, Fillard P, Facon D, et al: Diffusion tensor magnetic resonance imaging and fiber tracking in spinal cord lesions: Current and future indications. Neuroimaging Clin N Am 17(1):137–147, 2007. 13. Ducreux D, Lepeintre JF, Fillard P, et al: MR diffusion tensor imaging and fiber tracking in 5 spinal cord astrocytomas. AJNR Am J Neuroradiol 27(1):214–216, 2006. 14. Duong LM, McCarthy BJ, McLendon RE, et al: Descriptive epidemiology of malignant and nonmalignant primary spinal cord, spinal meninges, and cauda equina tumors, United States, 2004–2007. Cancer 118(17):4220–4227, 2012. 15. Ebner FH, Schittenhelm J, Roser F, et al: Management of holocord pilocytic astrocytomas in children and adolescents: An update. Pediatr Neurosurg 48(3):133–140, 2012. 16. Fakhreddine MH, Mahajan A, Penas-Prado M, et al: Treatment, prognostic factors, and outcomes in spinal cord astrocytomas. Neuro Oncol 15(4):406–412, 2013. 17. Fine MJ, Kricheff II, Freed D, et al: Spinal cord ependymomas: MR imaging features. Radiology 197(3):655–658, 1995. 18. Flanagan EP, O’Neill BP, Porter AB, et al: Primary intramedullary spinal cord lymphoma. Neurology 77(8):784–791, 2011. 19. Frank TS, Trojanowski JQ, Roberts SA, et al: A detailed immunohistochemical analysis of cerebellar hemangioblastoma:   An undifferentiated mesenchymal tumor. Mod Pathol 2(6):638–651, 1989. 20. Hamburger C: Ganglioglioma of the spinal cord: Report of two rare cases and review of the literature. Neurosurgery 41(6):1410–1415, discussion 1415–1416, 1997. 21. Harrop JS, Ganju A, Groff M, et al: Primary intramedullary tumors of the spinal cord. Spine 34(Suppl 22):S69–S77, 2009. 22. Horger M, Ritz R, Beschorner R, et al: Spinal pilocytic astrocytoma: MR imaging findings at first presentation and following surgery. Eur J Radiol 79(3):389–399, 2011. 23. Hwang JH, Egnaczyk GF, Ballard E, et al: spectroscopic characteristics of pediatric pilocytic astrocytomas. AJNR Am J Neuroradiol 19(3): 535–540, 1998. 24. Jallo GI: Spinal cord gangliogliomas: A review of 56 patients.   J Neurooncol 68(1):71–77, 2004. 25. Jung SH, Jung S, Moon KS, et al: Tailored surgical approaches for benign craniovertebral junction tumors. J Korean Neurosurg Soc 48(2):139–144, 2010.

http://radiologyebook.com

870

PART II  CT and MR Imaging of the Whole Body

26. Kahan H, Sklar EML, Judith Donovan Post M, et al: MR characteristics of histopathologic subtypes of spinal ependymoma. AJNR Am J Neuroradiol 17(1):143–150, 1996. 27. Kane PJ, El-Mahdy W: Spinal intradural tumours: Part II– intramedullary. Br J Neurosurg 13(6):558, 1999. 28. Karantanis D, O’Eill BP, Subramaniam RM, et al: 18F-FDG PET/CT in primary central nervous system lymphoma in HIV-negative patients. Nucl Med Commun 28(11):834–841, 2007. 29. Kim YG, Choi GH, Kim DH, et al: In vivo proton magnetic resonance spectroscopy of human spinal mass lesions. J Spinal Disord Tech 17(5):405–411, 2004. 30. Koeller KK, Rosenblum RS, Morrison AL: Neoplasms of the spinal cord and filum terminale: Radiologic-pathologic correlation. Radiographics 20(6):1721–1749, 2000. 31. Koeller KK, Rushing EJ: From the archives of the AFIP: Pilocytic astrocytoma: Radiologic-pathologic correlation. Radiographics 24(6):1693–1708, 2004. 32. Law M, Teicher N, Zagzag D, et al: Dynamic contrast enhanced perfusion MRI in mycosis fungoides. J Magn Reson Imaging 18(3): 364–367, 2003. 33. Lee JC: Ganglioglioma: Light and electron microscopic study. Neurochirurgia (Stuttg) 11(4):160–170, 1968. 34. Lee DK, Chung CK, Kim HJ, et al: Multifocal primary CNS T cell lymphoma of the spinal cord. Clin Neuropathol 21(4):149–155, 2002. 35. Lerner A, Mogensen MA, Kim PE, et al: Clinical applications of diffusion tensor imaging. World Neurosurg 82:96–109, 2014. 36. Liu X, Germin BI, Ekholm S: A case of cervical spinal cord glioblastoma diagnosed with MR diffusion tensor and perfusion imaging. J Neuroimaging 21(3):292–296, 2011. 37. Liu X, Tian W, Kolar B, et al: Advanced MR diffusion tensor imaging and perfusion weighted imaging of intramedullary tumors and tumor like lesions in the cervicomedullary junction region and the cervical spinal cord. J Neurooncol 2013. 38. Lonser RR, Weil RJ, Wanebo JE, et al: Surgical management of spinal cord hemangioblastomas in patients with von Hippel-Lindau disease.   J Neurosurg 98(1):106–116, 2003. 39. Mautner VF, Tatagiba M, Lindenau M, et al: Spinal tumors in patients with neurofibromatosis type 2: MR imaging study of frequency, multiplicity, and variety. AJR Am J Roentgenol 165(4):951–955, 1995. 40. Mechtler LL, Nandigam K: Spinal cord tumors: New views and future directions. Neurol Clin 31(1):241–268, 2013. 41. Milano MT, Johnson MD, Sul J, et al: Primary spinal cord glioma: A surveillance, epidemiology, and end results database study. J Neurooncol 98(1):83–92, 2010. 42. Miller DC: Central nervous system gangliogliomas. Part 1: Pathology.   J Neurosurg 79(6):859–866, 1993. 43. Miller RF, Hall-Craggs MA, Costa DC, et al: Magnetic resonance imaging, thallium-201 SPET scanning, and laboratory analyses for discrimination of cerebral lymphoma and toxoplasmosis in AIDS.   Sex Transm Infect 74(4):258–264, 1998. 44. Minehan KJ, Brown PD, Scheithauer BW, et al: Prognosis and treatment of spinal cord astrocytoma. Int J Radiat Oncol Biol Phys 73(3):727–733, 2009. 45. Miyazawa N, Hida K, Iwasaki Y, et al: MRI at 1.5 T of intramedullary ependymoma and classification of pattern of contrast enhancement. Neuroradiology 42(11):828–832, 2000. 46. Montoto S, Lister TA: Secondary central nervous system lymphoma: Risk factors and prophylaxis. Hematol Oncol Clin North Am 19(4):751– 763, 2005. 47. Nemoto Y, Inoue Y, Tashiro T, et al: Intramedullary spinal cord tumors: significance of associated hemorrhage at MR imaging. Radiology 182(3):793–796, 1992. 48. O’Malley JP, Ziessman HA, Kumar PN, et al: Diagnosis of intracranial lymphoma in patients with AIDS: Value of 201TI single-photon emission computed tomography. AJR Am J Roentgenol 163(2):417–421, 1994. 49. Park CK: Intramedullary spinal cord ganglioglioma: A report of five cases. Acta Neurochir (Wien) 142(5):547–552, 2000.

50. Patel U: MR of spinal cord ganglioglioma. AJNR Am J Neuroradiol 19(5):879–887, 1998. 51. Patronas NJ, Courcoutsakis N, Bromley CM, et al: Intramedullary and spinal canal tumors with neurofibromatosis 2: MR imaging findings and correlation with genotype. Radiology 218(2):434–442, 2001. 52. Plotkin SR, O’Donnell CC, Curry WT, et al: Spinal ependymomas in neurofibromatosis type 2: A retrospective analysis of 55 patients.   J Neurosurg Spine 14(4):543–547, 2011. 53. Raco A, Esposito V, Lenzi J, et al: Long-term follow-up of intramedullary spinal cord tumors: A series of 202 cases. Neurosurgery 56(5):972–981, discussion 972–981, 2005. 54. Rossi A: Tumors of the spine in children. Neuroimaging Clin N Am 17(1):17–35, 2007. 55. Ruiz A, Ganz WI, Post MJ, et al: Use of thallium-201 brain SPECT to differentiate cerebral lymphoma from toxoplasma encephalitis in AIDS patients. AJNR Am J Neuroradiol 15(10):1885–1894, 1994. 56. Rykken JB, Diehn FE, Hunt CH, et al: Intramedullary spinal cord metastases: MRI and relevant clinical features from a 13-year institutional case series. AJNR Am J Neuroradiol 34(10):2043–2049, 2013. 57. Rykken JB, Diehn FE, Hunt CH, et al: Rim and flame signs: Postgadolinium MRI findings specific for non-CNS intramedullary spinal cord metastases. AJNR Am J Neuroradiol 34(4):908–915, 2013. 58. Samii M, Klekamp J: Surgical results of 100 intramedullary tumors in relation to accompanying syringomyelia. Neurosurgery 35(5):865–873, discussion 873, 1994. 59. Schiff D, O’Neill BP: Intramedullary spinal cord metastases: Clinical features and treatment outcome. Neurology 47(4):906–912, 1996. 60. Schittenhelm J, Ebner FH, Tatagiba M, et al: Holocord pilocytic astrocytoma—Case report and review of the literature. Clin Neurol Neurosurg 111(2):203–207, 2009. 61. Seidenwurm DJ, Wippold Iii FJ, Cornelius RS, et al: ACR appropriateness criteria ® myelopathy. JACR Journal of the American College of Radiology 9(5):315–324, 2012. 62. Simmons ED, Zheng Y: Vertebral tumors: Surgical versus nonsurgical treatment. Clin Orthop Relat Res 443:233–247, 2006. 63. Seo HS, Kim JH, Lee DH, et al: Nonenhancing intramedullary astrocytomas and other MR imaging features: A retrospective study and systematic review. AJNR Am J Neuroradiol 31(3):498–503, 2010. 64. Setzer M, Murtagh RD, Murtagh FR, et al: Diffusion tensor imaging tractography in patients with intramedullary tumors: Comparison with intraoperative findings and value for prediction of tumor resectability.   J Neurosurg Spine 13(3):371–380, 2010. 65. Smith AB, Soderlund KA, Rushing EJ, et al: Radiologic-pathologic correlation of pediatric and adolescent spinal neoplasms: Part 1, intramedullary spinal neoplasms. AJR Am J Roentgenol 198(1):34–43, 2012. 66. Stadlbauer A, Ganslandt O, Buslei R, et al: Gliomas: Histopathologic evaluation of changes in directionality and magnitude of water diffusion at diffusion-tensor MR imaging. Radiology 240(3):803–810, 2006. 67. Sun B, Wang C, Wang J, et al: MRI features of intramedullary spinal cord ependymomas. J Neuroimaging 13(4):346–351, 2003. 68. Sung W-S, Sung M-J, Chan JH, et al: Intramedullary spinal cord metastases: A 20-year institutional experience with a comprehensive literature review. World Neurosurg 79(3–4):576–584, 2013. 69. Sutherland T, Yap K, Liew E, et al: Primary central nervous system lymphoma in immunocompetent patients: A retrospective review of MRI features. J Med Imaging Radiat Oncol 56(3):295–301, 2012. 70. Tani S, Ashida S, Yoneda H, et al: The Efficacy of CT angiography using the IVR-CT/angio system for surgical treatment of spinal hemangioblastoma. No Shinkei Geka 40(1):49–54, 2012. 71. Thurnher MM, Law M: Diffusion-weighted imaging, diffusion-tensor imaging, and fiber tractography of the spinal cord. Magn Reson Imaging Clin N Am 17(2):225–244, 2009. 72. Toh CH, Castillo M, Wong AM, et al: Primary cerebral lymphoma and glioblastoma multiforme: Differences in diffusion characteristics

http://radiologyebook.com

CHAPTER 30  Spinal Tumors

73. 74.

75.

76. 77.

78.

79. 80.

81.

82.

evaluated with diffusion tensor imaging. AJNR Am J Neuroradiol 29(3):471–475, 2008. Traul DE, Shaffrey ME, Schiff D, et al: Spinal-cord neoplasmsintradural neoplasms. Lancet Oncol 8(1):35–45, 2007. Vargas MI, Delavelle J, Jlassi H, et al: Clinical applications of diffusion tensor tractography of the spinal cord. Neuroradiology 50(1):25–29, 2008. Veeravagu A, Jiang B, Moss J, et al: Central nervous system hemangioblastomas: Part IV: Radiosurgery. Contemporary Neurosurgery 32(11):1–2, 2,3, 2010. Vertinsky AT: Cutting-edge imaging of the spine. Neuroimaging Clin N Am 17(1):117–136, 2007. Vicente J, Fuster-Garcia E, Tortajada S, et al: Accurate classification of childhood brain tumours by in vivo 1H MRS—A multi-centre study. Eur J Cancer 49(3):658–667, 2013. Wanebo JE, Lonser RR, Glenn GM, et al: The natural history of hemangioblastomas of the central nervous system in patients with von Hippel-Lindau disease. J Neurosurg 98(1):82–94, 2003. Wilson PE, Oleszek JL, Clayton GH: Pediatric spinal cord tumors and masses. J Spinal Cord Med 30(Suppl 1):S15–S20, 2007. Yang C: Intramedullary gangliogliomas: Clinical features, surgical outcomes, and neuropathic scoliosis. J Neurooncol 116(1):135–143, 2014. Yap KK, Sutherland T, Liew E, et al: Magnetic resonance features of primary central nervous system lymphoma in the immunocompetent patient: A pictorial essay. J Med Imaging Radiat Oncol 56(2):179–186, 2012. Young RJ, Ghesani MV, Kagetsu NJ, et al: Lesion size determines accuracy of thallium-201 brain single-photon emission tomography in differentiating between intracranial malignancy and infection in AIDS patients. AJNR Am J Neuroradiol 26(8):1973–1979, 2005.

Extradural Spinal Tumors 83. Abeloff MD, Armitage JO, Niederhuber JE, et al: Abeloff ’s Clinical Oncology, 2008. 84. Angtuaco EJ, Fassas AB, Walker R, et al: Multiple myeloma: Clinical review and diagnostic imaging. Radiology 231:11–23, 2004. 85. Aoki J, Tanikawa H, Ishii K, et al: MR findings indicative of hemosiderin in giant-cell tumor of bone: Frequency, cause, and diagnostic significance. AJR Am J Roentgenol 166:145–148, 1996. 86. Azouz EM, Saigal G, Rodriguez MM, et al: Langerhans’ cell histiocytosis: Pathology, imaging and treatment of skeletal involvement. Pediatr Radiol 35:103–115, 2005. 87. Azzarelli A, Quagliuolo V, Cerasoli S, et al: Chordoma: Natural history and treatment results in 33 cases. J Surg Oncol 37:185–191, 1988. 88. Bandiera S, Gasbarrini A, De Iure F, et al: Symptomatic vertebral hemangioma: The treatment of 23 cases and a review of the literature. Chir Organi Mov 87:1–15, 2002. 89. Barlogie B, Epstein J, Selvanayagam P, et al: Plasma cell myeloma–  New biological insights and advances in therapy. Blood 73:865–879, 1989. 90. Bartels RH, van der Linden YM, van der Graaf WT: Spinal extradural metastasis: Review of current treatment options. CA Cancer J Clin 58:245–259, 2008. 91. Barzin M, Maleki I: Incidence of vertebral hemangioma on spinal magnetic resonance imaging in Northern Iran. Pakistan journal of biological sciences: PJBS 12:542–544, 2009. 92. Baur A, Stabler A, Bruning R, et al: Diffusion-weighted MR imaging of bone marrow: Differentiation of benign versus pathologic compression fractures. Radiology 207:349–356, 1998. 93. Baur-Melnyk A, Buhmann S, Durr HR, et al: Role of MRI for the diagnosis and prognosis of multiple myeloma. Eur J Radiol 55:56–63, 2005. 94. Beltran J, Simon DC, Levy M, et al: Aneurysmal bone cysts: MR imaging at 1.5 T. Radiology 158:689–690, 1986. 95. Bemporad JA, Sze G, Chaloupka JC, et al: Pseudohemangioma of the vertebra: An unusual radiographic manifestation of primary Ewing’s sarcoma. AJNR Am J Neuroradiol 20:1809–1813, 1999.

871

96. Bergh P, Kindblom LG, Gunterberg B, et al: Prognostic factors in chordoma of the sacrum and mobile spine: A study of 39 patients. Cancer 88:2122–2134, 2000. 97. Bertoni F, Bacchini P, Capanna R, et al: Solid variant of aneurysmal bone cyst. Cancer 71:729–734, 1993. 98. Bertram C, Madert J, Eggers C: Eosinophilic granuloma of the cervical spine. Spine 27:1408–1413, 2002. 99. Bjornsson J, Wold LE, Ebersold MJ, et al: Chordoma of the mobile spine. A clinicopathologic analysis of 40 patients. Cancer 71:735–740, 1993. 100. Blakslee GA: Compression of the spinal cord in Hodgkin’s disease: A case thirteen years duration with recession of symptoms following Roentgen-ray therapy. Arch Neurol Psychiatr 20:130–137, 1928. 101. Boriani S, Bandiera S, Biagini R, et al: Chordoma of the mobile spine: Fifty years of experience. Spine 31:493–503, 2006. 102. Boriani S, Bandiera S, Casadei R, et al: Giant cell tumor of the mobile spine: A review of 49 cases. Spine 37:E37–E45, 2012. 103. Boriani S, Chevalley F, Weinstein JN, et al: Chordoma of the spine above the sacrum. Treatment and outcome in 21 cases. Spine (Phila Pa 1976) 21:1569–1577, 1996. 104. Boriani S, De Iure F, Bandiera S, et al: Chondrosarcoma of the mobile spine: Report on 22 cases. Spine 25:804–812, 2000. 105. Boriani S, De Iure F, Campanacci L, et al: Aneurysmal bone cyst of the mobile spine: Report on 41 cases. Spine 26:27–35, 2001. 106. Boukobza M, Mazel C, Touboul E: Primary vertebral and spinal epidural non-Hodgkin’s lymphoma with spinal cord compression. Neuroradiology 38:333–337, 1996. 107. Burn SC, Ansorge O, Zeller R, et al: Management of osteoblastoma and osteoid osteoma of the spine in childhood. J Neurosurg Pediatr 4:434–438, 2009. 108. Byun WM, Shin SO, Chang Y, et al: Diffusion-weighted MR imaging of metastatic disease of the spine: Assessment of response to therapy. AJNR Am J Neuroradiol 23:906–912, 2002. 109. Cagavi F, Kalayci M, Tekin IO, et al: Primary spinal extranodal Hodgkin’s disease at two levels. Clin Neurol Neurosurg 108:168–173, 2006. 110. Capanna R, Van Horn JR, Biagini R, et al: Aneurysmal bone cyst of the sacrum. Skeletal Radiol 18:109–113, 1989. 111. Carroll KW, Feller JF, Tirman PF: Useful internal standards for distinguishing infiltrative marrow pathology from hematopoietic marrow at MRI. J Magn Reson Imaging 7:394–398, 1997. 112. Chambers PW, Schwinn CP: Chordoma. A clinicopathologic study of metastasis. Am J Clin Pathol 72:765–776, 1979. 113. Cheng EY, Ozerdemoglu RA, Transfeldt EE, et al: Lumbosacral chordoma. Prognostic factors and treatment. Spine 24:1639–1645, 1999. 114. Chi JH, Bydon A, Hsieh P, et al: Epidemiology and demographics for primary vertebral tumors. Neurosurg Clin N Am 19:1–4, 2008. 115. Choi KS, Cohn MJ, Harfe BD: Identification of nucleus pulposus precursor cells and notochordal remnants in the mouse: Implications for disk degeneration and chordoma formation. Dev Dyn 237:3953– 3958, 2008. 116. Chow WA: Update on chondrosarcomas. Curr Opin Oncol 19:371–376, 2007. 117. Chugh R, Tawbi H, Lucas DR, et al: Chordoma: The nonsarcoma primary bone tumor. Oncologist 12:1344–1350, 2007. 118. Coon S: Pathology of tumors of the spinal cord, spine, and paraspinous soft tissue. In Dickman C, Fehlings M, Gokaslan Z, editors: Spinal cord and spinal column tumors: Principles and practice, New York, 2006, Thieme, pp 42–110. 119. Cottalorda J, Kohler R, Sales de Gauzy J, et al: Epidemiology of aneurysmal bone cyst in children: A multicenter study and literature review. J Pediatr orthop B 13:389–394, 2004. 120. Cross JJ, Antoun NM, Laing RJ, et al: Imaging of compressive vertebral haemangiomas. Eur Radiol 10:997–1002, 2000. 121. Dahlin DC: Bone tumors: General aspects and data on 6221 cases, Springfield, 1978, CC Thomas, pp 329–343. 122. Delorme S, Baur-Melnyk A: Imaging in multiple myeloma. Recent Results Cancer Res 183:133–147, 2011.

http://radiologyebook.com

872

PART II  CT and MR Imaging of the Whole Body

123. Deramond H, Depriester C, Galibert P, et al: Percutaneous vertebroplasty with polymethylmethacrylate. Technique, indications, and results. Radiol Clin North Am 36:533–546, 1998. 124. DiCaprio MR, Murphy MJ, Camp RL: Aneurysmal bone cyst of the spine with familial incidence. Spine 25:1589–1592, 2000. 125. Dimopoulos M, Terpos E, Comenzo RL, et al: International Myeloma Working Group consensus statement and guidelines regarding the current role of imaging techniques in the diagnosis and monitoring of multiple myeloma. Leukemia 23:1545–1556, 2009. 126. Doppman JL, Oldfield EH, Heiss JD: Symptomatic vertebral hemangiomas: Treatment by means of direct intralesional injection of ethanol. Radiology 214:341–348, 2000. 127. Drake RL: Lymphosarcoma involving the epidural space. J Kans Med Soc 42:212–222, 1941. 128. Durie BG: The role of anatomic and functional staging in myeloma: Description of Durie/Salmon plus staging system. Eur J Cancer 42:1539–1543, 2006. 129. Durie BG, Salmon SE: A clinical staging system for multiple myeloma. Correlation of measured myeloma cell mass with presenting clinical features, response to treatment, and survival. Cancer 36:842–854, 1975. 130. Dutoit JC, Vanderkerken MA, Verstraete KL: Value of whole body MRI and dynamic contrast enhanced MRI in the diagnosis, follow-up and evaluation of disease activity and extent in multiple myeloma. Eur J Radiol 82:1444–1452, 2013. 131. Ellis JA, Rothrock RJ, Moise G, et al: Primitive neuroectodermal tumors of the spine: A comprehensive review with illustrative clinical cases. Neurosurg Focus 30:E1, 2011. 132. Epelbaum R, Haim N, Ben-Shahar M, et al: Non-Hodgkin’s lymphoma presenting with spinal epidural involvement. Cancer 58:2120–2124, 1986. 133. Erlemann R: Imaging and differential diagnosis of primary bone tumors and tumor-like lesions of the spine. Eur J Radiol 58:48–67, 2006. 134. Fletcher CDM: Diagnostic histopathology of tumors, Edinburgh, NY, 1995, Churchill Livingstone. 135. Freeman C, Berg JW, Cutler SJ: Occurrence and prognosis of extranodal lymphomas. Cancer 29:252–260, 1972. 136. Fried G, Ben Arieh Y, Haim N, et al: Primary Hodgkin’s disease of the bone. Med Pediatr Oncol 24:204–207, 1995. 137. Gamba JL, Martinez S, Apple J, et al: Computed tomography of axial skeletal osteoid osteomas. AJR Am J Roentgenol 142:769–772, 1984. 138. Gasbarrini A, Cappuccio M, Donthineni R, et al: Management of benign tumors of the mobile spine. Orthop Clin North Am 40:9–19, v, 2009. 139. Greenspan A: Malignant bone tumors II. In Orthopedic radiology: A practical approach, Philadelphia, Pa, 2000, Lippincott Williams & Wilkins, pp 697–700. 140. Haddad P, Thaell JF, Kiely JM, et al: Lymphoma of the spinal extradural space. Cancer 38:1862–1866, 1976. 141. Hanrahan CJ, Christensen CR, Crim JR: Current concepts in the evaluation of multiple myeloma with MR imaging and FDG PET/CT. Radiographics 30:127–142, 2010. 142. Haque S, Law M, Abrey LE, et al: Imaging of lymphoma of the central nervous system, spine, and orbit. Radiol Clin North Am 46:339–361, ix, 2008. 143. Hirsh LF, Thanki A, Spector HB: Primary spinal chondrosarcoma with eighteen-year follow-up: Case report and literature review. Neurosurgery 14:747–749, 1984. 144. Horger M, Kanz L, Denecke B, et al: The benefit of using whole-body, low-dose, nonenhanced, multidetector computed tomography for follow-up and therapy response monitoring in patients with multiple myeloma. Cancer 109:1617–1626, 2007. 145. Hsieh PC, Xu R, Sciubba DM, et al: Long-term clinical outcomes following en bloc resections for sacral chordomas and chondrosarcomas: A series of twenty consecutive patients. Spine 34:2233–2239, 2009. 146. Hudson TM: Fluid levels in aneurysmal bone cysts: A CT feature. AJR Am J Roentgenol 142:1001–1004, 1984.

147. Hudson TM, Hamlin DJ, Fitzsimmons JR: Magnetic resonance imaging of fluid levels in an aneurysmal bone cyst and in anticoagulated human blood. Skeletal Radiol 13:267–270, 1985. 148. Ilaslan H, Sundaram M, Unni KK, et al: Primary Ewing’s sarcoma of the vertebral column. Skeletal Radiol 33:506–513, 2004. 149. Ilaslan H, Sundaram M, Unni KK, et al: Primary vertebral osteosarcoma: Imaging findings. Radiology 230:697–702, 2004. 150. Jacobs W, Fehlings M: 1. Primary vertebral column tumors. In Dickman C, Fehlings M, Gokaslan Z, editors: Spinal cord and spinal column tumors: Principles and practice, New York, 2006, Thieme, pp 369–386. 151. Jiang L, Liu XG, Yuan HS, et al: Diagnosis and treatment of vertebral hemangiomas with neurologic deficit: A report of 29 cases and literature review. Spine J 14:944–954, 2014. 152. Jung HS, Jee WH, McCauley TR, et al: Discrimination of metastatic from acute osteoporotic compression spinal fractures with MR imaging. Radiographics 23:179–187, 2003. 153. Kaiser TE, Pritchard DJ, Unni KK: Clinicopathologic study of sacrococcygeal chordoma. Cancer 53:2574–2578, 1984. 154. Katonis P, Alpantaki K, Michail K, et al: Spinal chondrosarcoma:   A review. Sarcoma 2011:378957, 2011. 155. Katonis P, Datsis G, Karantanas A, et al: Spinal osteosarcoma. Clin Med Insights Oncol 7:199–208, 2013. 156. Kelley SP, Ashford RU, Rao AS, et al: Primary bone tumours of the spine: A 42-year survey from the Leeds Regional Bone Tumour Registry. Eur Spine J 16:405–409, 2007. 157. Kilciksiz S, Karakoyun-Celik O, Agaoglu FY, et al: A review for solitary plasmacytoma of bone and extramedullary plasmacytoma. ScientificWorldJournal 2012:895765, 2012. 158. Kim RY, Spencer SA, Meredith RF, et al: Extradural spinal cord compression: Analysis of factors determining functional prognosis– Prospective study. Radiology 176:279–282, 1990. 159. Klimo P, Jr, Schmidt MH: Surgical management of spinal metastases. Oncologist 9:188–196, 2004. 160. Koci TM, Mehringer CM, Yamagata N, et al: Aneurysmal bone cyst of the thoracic spine: Evolution after particulate embolization. AJNR Am J Neuroradiol 16:857–860, 1995. 161. Kransdorf MJ, Stull MA, Gilkey FW, et al: Osteoid osteoma. Radiographics 11:671–696, 1991. 162. Kransdorf MJ, Sweet DE: Aneurysmal bone cyst: Concept, controversy, clinical presentation, and imaging. AJR Am J Roentgenol 164:573–580, 1995. 163. Krishnan A, Shirkhoda A, Tehranzadeh J, et al: Primary bone lymphoma: Radiographic-MR imaging correlation. Radiographics 23:1371–1383, discussion 1384–1377, 2003. 164. Kroon HM, Schurmans J: Osteoblastoma: Clinical and radiologic findings in 98 new cases. Radiology 175:783–790, 1990. 165. Krushelnycky BW, Perrin RG: Radiologic manifestations of spinal metastases, 1986. 166. Kumar V, Robbins SL: Robbins basic pathology, Philadelphia, PA, 2007, Saunders/Elsevier. 167. Kwon JW, Chung HW, Cho EY, et al: MRI findings of giant cell tumors of the spine. AJR Am J Roentgenol 189:246–250, 2007. 168. Kyle RA, Rajkumar SV: Epidemiology of the plasma-cell disorders. Best Pract Res Clin Haematol 20:637–664, 2007. 169. Kyle RA, Rajkumar SV: Multiple myeloma. Blood 111:2962–2972, 2008. 170. Lecouvet FE, Dechambre S, Malghem J, et al: Bone marrow transplantation in patients with multiple myeloma: Prognostic significance of MR imaging. AJR Am J Roentgenol 176:91–96, 2001. 171. Lecouvet FE, Malghem J, Michaux L, et al: Skeletal survey in advanced multiple myeloma: Radiographic versus MR imaging survey. Br J Haematol 106:35–39, 1999. 172. Lecouvet FE, Vande Berg BC, Maldague BE, et al: Vertebral compression fractures in multiple myeloma. Part I. Distribution and appearance at MR imaging. Radiology 204:195–199, 1997. 173. Lecouvet FE, Vande Berg BC, Malghem J, et al: Magnetic resonance and computed tomography imaging in multiple myeloma. Semin Musculoskelet Radiol 5:43–55, 2001.

http://radiologyebook.com

CHAPTER 30  Spinal Tumors 174. Leithner A, Windhager R, Lang S, et al: Aneurysmal bone cyst.   A population based epidemiologic study and literature review. Clin Orthop Relat Res 363:176–179, 1999. 175. Lewandrowski KU, Bell GR, McLain RF: Cancer of the spine: How big is the problem? In Totowa NJHP, editor: Cancer in the spine: Comprehensive care, 2006. 176. Libicher M, Appelt A, Berger I, et al: The intravertebral vacuum phenomenon as specific sign of osteonecrosis in vertebral compression fractures: Results from a radiological and histological study. Eur Radiol 17:2248–2252, 2007. 177. Liu G, Wu G, Ghimire P, et al: Primary spinal chondrosarcoma: Radiological manifestations with histopathological correlation in eight patients and literature review. Clin Imaging 37:124–133, 2013. 178. Livingston KE, Perrin RG: The neurosurgical management of spinal metastases causing cord and cauda equina compression. J Neurosurg 49:839–843, 1978. 179. Lucas DR, Unni KK, McLeod RA, et al: Osteoblastoma: Clinicopathologic study of 306 cases. Hum Pathol 25:117–134, 1994. 180. MacVicar D, Williams MP: CT scanning in epidural lymphoma. Clin Radiol 43:95–102, 1991. 181. Mahnken AH, Nolte-Ernsting CC, Wildberger JE, et al: Aneurysmal bone cyst: Value of MR imaging and conventional radiography. Eur Radiol 13:1118–1124, 2003. 182. Mahnken AH, Wildberger JE, Gehbauer G, et al: Multidetector CT of the spine in multiple myeloma: Comparison with MR imaging and radiography. AJR Am J Roentgenol 178:1429–1436, 2002. 183. Maldague BE, Noel HM, Malghem JJ: The intravertebral vacuum cleft: A sign of ischemic vertebral collapse. Radiology 129:23–29, 1978. 184. Manaster BJ: Tumors. In Manaster BJ, Disler DG, May DA, et al, editors: Musculoskeletal Imaging: The Requisites. ed 2. St Louis, MO, 2002, Mosby, pp 1–104. 185. Mankin HJ, Hornicek FJ, Ortiz-Cruz E, et al: Aneurysmal bone cyst:   A review of 150 patients. J Clin Oncol 23:6756–6762, 2005. 186. Mayer L: Malignant degeneration of so-called benign osteoblastoma. Bull Hosp Joint Dis 28:4–13, 1967. 187. McLeod RA, Dahlin DC, Beabout JW: The spectrum of osteoblastoma. AJR Am J Roentgenol 126:321–325, 1976. 188. McLoughlin GS, Sciubba DM, Wolinsky JP: Chondroma/ Chondrosarcoma of the spine. Neurosurg Clin N Am 19:57–63, 2008. 189. McMaster ML, Goldstein AM, Bromley CM, et al: Chordoma: Incidence and survival patterns in the United States, 1973–1995. Cancer Causes Control 12:1–11, 2001. 190. Meyer JE, Lepke RA, Lindfors KK, et al: Chordomas: Their CT appearance in the cervical, thoracic and lumbar spine. Radiology 153:693–696, 1984. 191. Monnard V, Sun A, Epelbaum R, et al: Primary spinal epidural lymphoma: Patients’ profile, outcome, and prognostic factors: A multicenter Rare Cancer Network study. Int J Radiat Oncol Biol Phys 65:817–823, 2006. 192. Mora J, Wollner N: Primary epidural non-Hodgkin lymphoma: Spinal cord compression syndrome as the initial form of presentation in childhood non-Hodgkin lymphoma. Med Pediatr Oncol 32:102–105, 1999. 193. Motamedi K, Ilaslan H, Seeger LL: Imaging of the lumbar spine neoplasms. Semin Ultrasound CT MR 25:474–489, 2004. 194. Moulopoulos LA, Dimopoulos MA: Magnetic resonance imaging   of the bone marrow in hematologic malignancies. Blood 90:2127–2147, 1997. 195. Moulopoulos LA, Dimopoulos MA, Alexanian R, et al: Multiple myeloma: MR patterns of response to treatment. Radiology 193:441– 446, 1994. 196. Mukherjee D, Chaichana KL, Adogwa O, et al: Association of extent of local tumor invasion and survival in patients with malignant primary osseous spinal neoplasms from the surveillance, epidemiology, and end results (SEER) database. World Neurosurg 76:580–585, 2011. 197. Murphey MD, Andrews CL, Flemming DJ, et al: From the archives of the AFIP. Primary tumors of the spine: Radiologic pathologic correlation. Radiographics 16:1131–1158, 1996.

873

198. Murphey MD, Fairbairn KJ, Parman LM, et al: From the archives of the AFIP. Musculoskeletal angiomatous lesions: Radiologic-pathologic correlation. Radiographics 15:893–917, 1995. 199. Murphey MD, Nomikos GC, Flemming DJ, et al: From the archives of AFIP. Imaging of giant cell tumor and giant cell reparative granuloma of bone: Radiologic-pathologic correlation. Radiographics 21:1283– 1309, 2001. 200. Murphey MD, wan Jaovisidha S, Temple HT, et al: Telangiectatic osteosarcoma: Radiologic-pathologic comparison. Radiology 229:545– 553, 2003. 201. Murphy JM, Wallis F, Toland J, et al: CT and MRI appearances of a thoracic chordoma. Eur Radiol 8:1677–1679, 1998. 202. Nau KC, Lewis WD: Multiple myeloma: Diagnosis and treatment. Am Fam Physician 78:853–859, 2008. 203. Negendank WG, al-Katib AM, Karanes C, et al: Lymphomas: MR imaging contrast characteristics with clinical-pathologic correlations. Radiology 177:209–216, 1990. 204. Ozaki T, Flege S, Liljenqvist U, et al: Osteosarcoma of the spine: Experience of the Cooperative Osteosarcoma Study Group. Cancer 94:1069–1077, 2002. 205. Papagelopoulos PJ, Currier BL, Shaughnessy WJ, et al: Aneurysmal bone cyst of the spine. Management and outcome. Spine 23:621–628, 1998. 206. Per H, Koc KR, Gumus H, et al: Cervical eosinophilic granuloma and torticollis: A case report and review of the literature. J Emerg Med 35:389–392, 2008. 207. Perrin RG, Laxton AW: Metastatic spine disease: Epidemiology, pathophysiology, and evaluation of patients. Neurosurg Clin N Am 15:365–373, 2004. 208. Peters LJ: The ESTRO Regaud lecture. Inherent radiosensitivity of tumor and normal tissue cells as a predictor of human tumor response. Radiother Oncol 17:177–190, 1990. 209. Ramadan KM, Shenkier T, Sehn LH, et al: A clinicopathological retrospective study of 131 patients with primary bone lymphoma: A population-based study of successively treated cohorts from the British Columbia Cancer Agency. Ann Oncol 18:129–135, 2007. 210. Randall RL: Giant cell tumor of the sacrum. Neurosurg Focus 15:E13, 2003. 211. Resnick D: Tumors and tumor-like diseases. In Diagnosis of bone and joint disorders, Philadelphia, PA, 1995, Saunders, pp 3785–3806. 212. Roodman GD: Cell biology of the osteoclast. Exp Hematol 27:1229– 1241, 1999. 213. Roodman GD: Mechanisms of bone metastasis. N Engl J Med 350:1655–1664, 2004. 214. Ropper AE, Cahill KS, Hanna JW, et al: Primary vertebral tumors:   A review of epidemiologic, histological, and imaging findings, Part I: benign tumors. Neurosurgery 69:1171–1180, 2011. 215. Ropper AE, Cahill KS, Hanna JW, et al: Primary vertebral tumors:   A review of epidemiologic, histological and imaging findings, part II: Locally aggressive and malignant tumors. Neurosurgery 70:211–219, discussion 219, 2012. 216. Rosenthal DI: Radiologic diagnosis of bone metastases. Cancer 80:1595–1607, 1997. 217. Saccomanni B: Aneurysmal bone cyst of spine: A review of literature. Arch Orthop Trauma Surg 128:1145–1147, 2008. 218. Salisbury JR, Deverell MH, Cookson MJ, et al: Three-dimensional reconstruction of human embryonic notochords: Clue to the pathogenesis of chordoma. J Pathol 171:59–62, 1993. 219. Salvo N, Christakis M, Rubenstein J, et al: The role of plain radiographs in management of bone metastases. J Palliat Med 12:195–198, 2009. 220. Sandberg AA: Cytogenetics and molecular genetics of bone and soft-tissue tumors. Am J Med Genet 115:189–193, 2002. 221. Sanerkin NG, Mott MG, Roylance J: An unusual intraosseous lesion with fibroblastic, osteoclastic, osteoblastic, aneurysmal and fibromyxoid elements. “Solid” variant of aneurysmal bone cyst. Cancer 51:2278– 2286, 1983. 222. Sanjay BK, Frassica FJ, Frassica DA, et al: Treatment of giant-cell tumor of the pelvis. J Bone Joint Surg Am 75:1466–1475, 1993.

http://radiologyebook.com

874

PART II  CT and MR Imaging of the Whole Body

223. Sanjay BK, Sim FH, Unni KK, et al: Giant-cell tumours of the spine.   J Bone Joint Surg Br 75:148–154, 1993. 224. Schajowicz F, Lemos C: Osteoid osteoma and osteoblastoma. Closely related entities of osteoblastic derivation. Acta Orthop Scand 41:272– 291, 1970. 225. Schweitzer ME, Levine C, Mitchell DG, et al: Bull’s-eyes and halos: Useful MR discriminators of osseous metastases. Radiology 188:249– 252, 1993. 226. Seider MJ, Rich TA, Ayala AG, et al: Giant cell tumors of bone: Treatment with radiation therapy. Radiology 161:537–540, 1986. 227. Shah LM, Hanrahan CJ: MRI of spinal bone marrow: Part I, techniques and normal age-related appearances. AJR Am J Roentgenol 197:1298– 1308, 2011. 228. Shah LM, Salzman KL: Imaging of spinal metastatic disease. Int J Surg Oncol 2011:769753, 2011. 229. Sharafuddin MJ, Haddad FS, Hitchon PW, et al: Treatment options in primary Ewing’s sarcoma of the spine: Report of seven cases and review of the literature. Neurosurgery 30:610–618, discussion 618–619, 1992. 230. Shen J, Li CD, Yang HL, et al: Classic chordoma coexisting with benign notochordal cell rest demonstrating different immunohistological expression patterns of brachyury and galectin-3. J Clin Neurosci 18:96–99, 2011. 231. Shirley SK, Gilula LA, Siegal GP, et al: Roentgenographic-pathologic correlation of diffuse sclerosis in Ewing sarcoma of bone. Skeletal Radiol 12:69–78, 1984. 232. Shives TC, McLeod RA, Unni KK, et al: Chondrosarcoma of the spine.   J Bone Joint Surg Am 71:1158–1165, 1989. 233. Smolders D, Wang X, Drevelengas A, et al: Value of MRI in the diagnosis of non-clival, non-sacral chordoma. Skeletal Radiol 32:343–350, 2003. 234. Stacchiotti S, Casali PG, Lo Vullo S, et al: Chordoma of the mobile spine and sacrum: A retrospective analysis of a series of patients surgically treated at two referral centers. Ann Surg Oncol 17:211–219, 2010. 235. Sundaresan N, Galicich JH, Chu FC, et al: Spinal chordomas.   J Neurosurg 50:312–319, 1979. 236. Sundaresan N, Rosen G, Boriani S: Primary malignant tumors of the spine. Orthop Clin North Am 40:21–36, v, 2009. 237. Theodorou DJ, Theodorou SJ, Sartoris DJ: An imaging overview of primary tumors of the spine: Part 1. Benign tumors. Clin Imaging 32:196–203, 2008. 238. Theodorou DJ, Theodorou SJ, Sartoris DJ: An imaging overview of primary tumors of the spine: Part 2. Malignant tumors. Clin Imaging 32:204–211, 2008. 239. Thomas AG, Vaidhyanath R, Kirke R, et al: Extranodal lymphoma from head to toe: Part 1, the head and spine. AJR Am J Roentgenol 197:350– 356, 2011.

240. Tricot G: New insights into role of microenvironment in multiple myeloma. Lancet 355:248–250, 2000. 241. Tsai JC, Dalinka MK, Fallon MD, et al: Fluid-fluid level: A nonspecific finding in tumors of bone and soft tissue. Radiology 175:779–782, 1990. 242. Turcotte RE, Biagini R, Sim FH, et al: Giant cell tumor of the spine and sacrum. Chir Organi Mov 75:104–107, 1990. 243. Tzortzidis F, Elahi F, Wright D, et al: Patient outcome at long-term follow-up after aggressive microsurgical resection of cranial base chordomas. Neurosurgery 59:230–237, discussion 230–237, 2006. 244. Uchida N, Sugimura K, Kajitani A, et al: MR imaging of vertebral metastases: Evaluation of fat saturation imaging. Eur J Radiol 17:91–94, 1993. 245. Verity GL: Neurologic manifestations and complications of lymphoma. Radiol Clin North Am 6:97–106, 1968. 246. Vilanova JC, Barcelo J, Smirniotopoulos JG, et al: Hemangioma from head to toe: MR imaging with pathologic correlation. Radiographics 24:367–385, 2004. 247. Vujovic S, Henderson S, Presneau N, et al: Brachyury, a crucial regulator of notochordal development, is a novel biomarker for chordomas. J Pathol 209:157–165, 2006. 248. Walker R, Barlogie B, Haessler J, et al: Magnetic resonance imaging in multiple myeloma: Diagnostic and clinical implications. J Clin Oncol 25:1121–1128, 2007. 249. Weinstein JB, Siegel MJ, Griffith RC: Spinal Ewing sarcoma: Misleading appearances. Skeletal Radiol 11:262–265, 1984. 250. Windeyer BW: Chordoma. Proc R Soc Med 52:1088–1100, 1959. 251. Witham TF, Khavkin YA, Gallia GL, et al: Surgery insight: Current management of epidural spinal cord compression from metastatic spine disease. Nat Clin Pract Neurol 2:87–94, quiz 116, 2006. 252. Yeom JS, Lee CK, Shin HY, et al: Langerhans’ cell histiocytosis   of the spine. Analysis of twenty-three cases. Spine 24:1740–1749, 1999. 253. York JE, Berk RH, Fuller GN, et al: Chondrosarcoma of the spine: 1954 to 1997. J Neurosurg 90:73–78, 1999. 254. York JE, Kaczaraj A, Abi-Said D, et al: Sacral chordoma: 40-year experience at a major cancer center. Neurosurgery 44:74–79, discussion 79–80, 1999. 255. Zhao J, Krug R, Xu D, et al: MRI of the spine: Image quality and normal-neoplastic bone marrow contrast at 3 T versus 1.5 T. AJR Am J Roentgenol 192:873–880, 2009. 256. Zhou XJ, Leeds NE, McKinnon GC, et al: Characterization of benign and metastatic vertebral compression fractures with quantitative diffusion MR imaging. AJNR Am J Neuroradiol 23:165–170, 2002. 257. Zileli M, Isik HS, Ogut FE, et al: Aneurysmal bone cysts of the spine. Eur Spine J 22:593–601, 2013.

http://radiologyebook.com

31  Spinal Infection Theeraphol Panyaping

INTRODUCTION Spinal infection is a common disease worldwide and is being recognized more often, especially since the introduction of more sophisticated imaging techniques. It may be a primary disease, starting locally as a spinal infection, or secondary to a systemic infection or an infection of the central nervous system. Most patients usually present with subtle and nonspecific clinical symptoms that are difficult to diagnose. Because many spinal infections are life threatening, early diagnosis leads to improved outcomes. Delayed diagnosis of spinal infection may lead to increased morbidity and mortality. Therefore imaging is increasingly used and plays an important role in diagnosis, management, and follow-up. In this chapter, we review the clinical manifestations, imaging findings, and treatment of spinal infections. Many pathogens can cause spondylitis, myelitis, and/or meningitis. Infection of the spine primarily involving the spinal cord and its coverings are discussed separately from those primarily involving the bones, paraspinal regions, and the epidural space. The focus of this chapter is on pyogenic spondylitis and tuberculous spondylitis, which are common spinal infections in routine clinical practice. Nowadays magnetic resonance imaging (MRI) is the diagnostic method of choice for spinal infections, owing to its high sensitivity and specificity. Here we emphasize the MRI findings in each different type of spinal infection, including clues in differential diagnosis, which allows the radiologist to distinguish them from other diseases.

PYOGENIC SPONDYLODISKITIS Pyogenic spondylodiskitis is caused by bacterial infection of the bony spinal column, the intervertebral disks, and/or extradural/intradural space. The most common cause of pyogenic spondylodiskitis is  hematogenous spread of infection from a remote site. Urinary tract infection is the most common source. The infection typically spreads via an arterial route, although the paravertebral venous plexus has also been accessed. Other etiologies include direct inoculation in a surgical setting, diskography, and therapeutic spinal injections, as well as contiguous spread from adjacent infected sites.13,18,27 Staphylococcus aureus is the most common causative organism, found in approximately one third of all patients.30,46,48 Peak incidences are in the sixth and seventh decades of life. Risk factors include age older than 50 years, diabetes, underlying chronic disease such as renal failure and cirrhosis, AIDS and other immunocompromising host infections, chronic steroid  use, and intravenous drug use. Diabetic patients, particularly in their 50s and 60s, are frequently affected, and overall spinal infections are more common in men. Patients with pyogenic spondylodiskitis typically present with fever, back pain, and associated muscle spasm with tenderness to palpation. Limited range of motion and/or rapidly

progressive neurologic deficits can also be present. However, patients may present with subtle, insidious, and nonspecific symptoms such as malaise and weight loss. The time course ranges from acute to chronic. Laboratory evaluation is not reliable, although elevated erythrocyte sedimentation rate (ESR) and elevated C-reactive protein (CRP) are often present. Leukocytosis is less common.9,17,21 The lumbar spine is most frequently affected (≈45% of cases), followed by the thoracic spine (35% of cases); the cervical spine is the least common site (≈15% of cases). The intervertebral disks in adults are nearly avascular owing to involution of the rich network of intraosseous anastomoses and the profuse capillary network at the vertebral margins of the disks during adolescence. Only the cancellous bone adjacent to the cartilaginous end plate is highly vascularized in adults and is prone to bloodborne organisms. In most patients the infection is limited to the disk and adjacent vertebral bodies. Multiple levels and/or skip lesions are also common. The most widely accepted pathophysiologic hypothesis is that infected emboli cause infarction and subsequent infection in the metaphyses of vertebral bodies. Infection typically begins in the anterior aspect of the vertebral bodies and spreads early to the adjacent disks. The remainder of vertebral bodies and opposite vertebral end plates are also involved late in the course of the disease. Spread of infection into the paraspinal and/or epidural spaces is also common. Infection may extend deeply into the subdural or subarachnoid spaces, causing subdural abscess or arachnoiditis. In contrast to adults, the primary site of infection in children is the disk because of its rich vascular supply, which allows infection to spread directly to the disk. Thus isolated diskitis can occur in children.

Imaging Evaluation In current routine clinical practice, plain radiographs are usually the first imaging modality in patients with back pain. However, plain radiographs are not sensitive for spondylodiskitis, particularly in the early phase of the disease. Radiographic findings are usually subtle and may not appear until late in the process. Therefore the hallmark radiographic findings of spondylodiskitis (Fig. 31-1), such as disk space narrowing, end-plate irregularity/destruction, and associated paraspinal soft tissue bulging, typically lag behind the clinical evolution and may not be present until several days or weeks after the initial infection occurred. In addition, plain radiographs cannot demonstrate the intraspinal complications of spondylodiskitis, especially epidural abscess, arachnoiditis, or any other lesion that results in mass effect on the spinal cord/nerve roots.30 In the chronic stage of spondylodiskitis, findings such as end-plate sclerosis and bony ankylosis can be seen. Computed tomography (CT) is more sensitive than plain films but also has limited evaluation of soft tissue changes, as well as any abnormality involving the spinal cord and nerve roots.

http://radiologyebook.com

875

876

PART II  CT and MR Imaging of the Whole Body

A

C

B

F

D

E

FIG 31-1  Pyogenic spondylodiskitis. This 5-year-old boy (known case of acute nonlymphoblastic leukemia post bone marrow transplant) presented with low back pain for 4 days without fever. Anteroposterior (A) and lateral (B) plain radiographs show narrowing of the L4-L5 disk space, with subtle irregularity of the adjacent end plates (arrows). C, T1-weighted sagittal MRI shows diffuse T1 hypointensity of the visualized vertebral bodies, with destruction of the L4 and L5 vertebral bodies (arrows). D, T2-weighted sagittal MRI demonstrates abnormal mild T2 hyperintensity in the L4-5 disk space and irregular destruction of the adjacent end plates, with ventral epidural and anterior paraspinal soft tissue (arrows). E, Postgadolinium fat-saturated T1-weighted sagittal MRI reveals diffuse marrow enhancement of the L4 and L5 vertebral bodies as well as inhomogeneous enhancement of the L4-L5 disk space (left arrow). Thick enhancing ventral epidural soft tissue (right arrow) and thin anterior paraspinal enhancement extending from lower L3 to upper S1 level is seen. F, Postgadolinium fat-saturated T1-weighted axial MRI confirms vertebral marrow enhancement and enhancing ventral epidural and paraspinal soft tissue. In this case there are a few small rim-enhancing portions (arrows) in the anterior epidural and paraspinal soft tissue, enhancement possibly representing small abscess formation.

http://radiologyebook.com

CHAPTER 31  Spinal Infection Currently MRI is the imaging of choice for spine infection, owing to its high sensitivity and specificity for detecting spinal infections early in their course. The sensitivity of MRI for spondylodiskitis is comparable to that of radionuclide studies, but its specificity is much higher. Even in the early process of spondylitis, MRI shows a sensitivity of 96%, a specificity of 92%, and accuracy of 92%.10,29 Regarding the MRI protocol, sagittal MRIs of the entire spinal axis are recommended to identify the extent of infection and possible skip lesions. In the presence of a spinal infection, there is increased T2 signal intensity in the vertebral marrow due to edema, whereas T1-weighted signal intensity is decreased due to replacement of marrow fat by edema. Gadolinium enhancement is easier seen with fat-suppressed T1-weighted images and helps to delineate intraosseous infection (Fig. 31-2; also see Fig. 31-1).8,26 Therefore fat-suppressed T2- weighted and fat-suppressed postgadolinium T1-weighted sequences are useful and should be performed in every case of spinal infection.17,21 Pathologically the infection

A

877

typically begins in the anterior metaphyseal region of the vertebral body and then spreads to involve the disk space and adjacent vertebral body. Therefore bone marrow signal changes typically begin at the vertebral end plates adjacent to the infected disk and appear as T1 hypointensity and T2 hyperintensity with contrast enhancement in the subchondral bone marrow. Unfortunately this change is nonspecific and can also be seen in some degenerative end-plate changes and in many other noninfectious conditions. End-plate erosion and vertebral body destruction are commonly evident. Disk involvement typically manifests as increased T2 signal, which is the most reliable sign of diskitis.10 Other MRI findings of diskitis include loss of the normal intradiskal band of decreased T2 signal that represents the internuclear cleft, disk height loss, and disk enhancement (see Figs. 31-1 and 31-2). Associated paraspinal soft tissue involvement usually presents as either solid-appearing enhancing soft tissues typical of phlegmon or T1 hypointense/T2 hyperintense fluid collections with peripheral contrast

B

C

D FIG 31-2  Pyogenic spondylodiskitis. This 73-year-old woman presented with back pain and mild weakness of both legs for 1 month. A, T1-weighted sagittal MRI shows diffuse T1 hypointensity (white arrows) in the L4 and L5 vertebral bodies, as well as some in the posterior aspect of L3 vertebral body. B, T2-weighted fat-saturated sagittal MRI demonstrates abnormal T2 hyperintensity in the narrowed L4-L5 disk space and T2 hyperintensity in the L4 and L5 vertebral bodies (left white arrows), as well as some in the posterior aspect of the L3 vertebral body (right white arrow). C, Postgadolinium fat-saturated T1-weighted sagittal MRI reveals marrow enhancement corresponding to the abnormal vertebral marrow signal. Evidence of disk enhancement, enhancing ventral epidural soft tissue, and thin anterior paraspinal enhancement extending from lower L3 to upper S1 level (white arrows) is seen. Enhancing cauda equina nerve roots are also seen, indicating radiculitis/arachnoiditis. Note a small nonenhancing portion in the anterior paravertebral region at the upper L5 level, representing small abscess formation. D, Postgadolinium fat-saturated T1-weighted axial MRI confirms the vertebral marrow enhancement, enhancing ventral epidural and paraspinal soft tissue (black arrows).

http://radiologyebook.com

878

PART II  CT and MR Imaging of the Whole Body

enhancement, representing frank abscess (see Figs. 31-1 and 31-2). The most important findings that have been suggested for the diagnosis of pyogenic spondylodiskitis include early disk space involvement manifested as T2 hyperintensity, disk space enhancement, vertebral marrow T1 hypointensity, and the presence of paraspinal or epidural inflammation. In contrast, findings with relatively low sensitivity for the diagnosis of pyogenic spondylodiskitis include T1 hypointensity in  the disk and disk-space height loss.11,25 Therefore if T2 hyperintensity and/or enhancement are present in both the disk and adjacent vertebral bodies, infectious spondylodiskitis should be considered. Pyogenic spondylodiskitis may have atypical imaging features, including lack of end-plate erosive changes and signal intensity abnormalities in the early course of the disease, involvement of two adjacent bodies with intervening disk sparing, only single vertebral body involvement, and involvement of one vertebral body and one disk only.19 Occasionally pyogenic spondylodiskitis may appear as solitary or multiple discrete enhancing bony lesions without gross abnormalities of the disks, mimicking metastatic disease or other marrow infiltrative diseases.16 MRI is not recommended for routine follow-up, because it may lag behind the clinical picture, especially of improvement and clinical response to antibiotics.6,15,16,22,49 MRI findings can even worsen despite clinical improvement. Abnormal MRI findings, particularly of the affected bone and disk as well as paraspinal soft tissues, may persist despite antibiotic therapy for 4 to 8 weeks22 and even months after a clinical cure has been achieved.15,31,49 The only radiographic sign correlating with the healing process at MRI follow-up is reappearance of normal marrow T1 hyperintensity, which is thought to be normal fatty replacement.16,19 MRI is also useful in the setting of postoperative spinal infection. Postoperative diskitis and/or osteomyelitis are relatively uncommon complications after lumbar spine surgery. MRI is less reliable in this setting because the postoperative spine may demonstrate disk or endplate signal changes/enhancement that are part of the process. MRI cannot reliably differentiate between pathology and postoperative changes until at least 6 months after surgery. One finding that may suggest postdiskectomy changes rather than infection is the presence of two parallel thin bands of enhancement in the disk space, which is different from the amorphous/patchy enhancement commonly seen throughout the disk with infection. If paravertebral enhancement is also present, it supports the diagnosis of infection, whereas the absence of T2 hyperintensity/enhancing end-plate changes or disk space enhancement makes infection unlikely.36 Regarding nuclear medicine imaging, despite relatively high sensitivity it is used only in selected patients, owing to its limited spatial resolution, low availability outside of industrialized countries, long examination time, and low specificity of positive findings. Indications for nuclear scanning include patients in whom MRI is contraindicated, equivocal MRI and CT findings, suspected multifocal infectious disease, or high clinical suspicion with negative MRI examination.

Treatment Parenteral antibiotic therapy is the mainstay of clinical management for at least 6 weeks.9,43 Surgery is performed in patients who have complications of spondylodiskitis, such as epidural abscess, cord or nerve root compression, chronic spinal deformity, or spinal instability due to collapse of vertebral bodies or disks. As mentioned previously, the clinical and imaging evolutions may be discordant. The most important predictive feature that indicates healing of spondylodiskitis is clinical improvement. Laboratory markers such as improvement of leukocytosis and ESR can also be helpful. Persistently elevated or

increased ESR and CRP at week 4 of antibiotic therapy suggest treatment failure.48

SPINAL EPIDURAL AND SUBDURAL ABSCESSES Spinal epidural abscesses commonly are due to direct extension from spondylodiskitis, spontaneously or related to hematogenous spread or iatrogenic inoculation from an invasive spinal procedure. Predisposing factors include immunosuppressive conditions and underlying chronic diseases, especially diabetes. As in pyogenic spondylodiskitis, S. aureus is the most common pathogen, recovered in approximately 70% of patients. The clinical presentation of patients with spinal epidural abscess is variable. Back pain is the most common symptom and is severe. Fever and neurologic deficits are also common, as well as other symptoms such as tenderness and limited range of motion. Leukocytosis and elevations of ESR and CRP are common. Spinal epidural abscesses are most common in the thoracic spine but can involve the entire length of the spine. Abscesses can be located either in the anterior or posterior epidural space. Rapid expansion of the abscess can cause permanent neurologic deficits resulting from direct cord or nerve root compression. Neurologic deficits can also be caused by thrombosis or thrombophlebitis.45

Imaging Evaluation MRI of the entire spinal axis is recommended when a spinal epidural abscess is clinically suspected. MRI typically reveals a T1 hypointense/ T2 hyperintense lesion in the epidural space. On postgadolinium T1-weighted imaging, this lesion typically enhances either homogeneously or heterogeneously when it is a phlegmon or peripherally with central nonenhancing fluid signal intensity in a mature abscess11,31 (Fig. 31-3). Patients with an epidural abscess always have meningeal thickening and enhancement. A ventral epidural abscess usually has a bilobed configuration on axial MRI images because of the firmly anchored midline raphe and the fixed lateral membranes giving rise to the so-called curtain sign.38 Engorgement of the epidural venous plexus and/or basivertebral veins may be present. Thecal sac, spinal cord, and/or cauda equina nerve root compression are common. Diffusion-weighted imaging may show restricted diffusion within the epidural abscess.14 Subdural abscess of the spine is extremely rare. In it, the pus collection is located in the space between the dura and arachnoid, which shows as a ring-enhancing fluid signal collection on axial postgadolinium T1-weighted MRI. The most common location is the lumbar spine. Risk factors and clinical features are similar to epidural abscess. S. aureus is again the most common causative organism.

Treatment Emergent surgical decompression and abscess drainage is the definitive treatment, followed by antibiotic therapy.2,12,34 Antibiotic therapy may be delayed in neurologically stable patients until emergent surgery has been performed. A phlegmon may require a widespread decompressive approach with laminectomy, whereas an abscess may be treated by limited laminotomies and irrigation.45

GRANULOMATOUS SPINAL INFECTIONS Tuberculous Spondylitis Tuberculous spondylitis (Pott’s disease) is the most common nonpyogenic infection of the spine and also the most common overall  cause of spinal infection in the developing world. Tuberculous infection is more insidious than pyogenic infection of the spine. The incidence of tuberculosis remains high in endemic regions, even in

http://radiologyebook.com

CHAPTER 31  Spinal Infection

A

C

B

D

879

E

FIG 31-3  Pyogenic spondylitis with epidural abscesses. This 53-year-old man presented with fever and back pain for 3 weeks. A, T1-weighted sagittal MRI. B, T2-weighted fat-saturated sagittal MRI. C, Postgadolinium fat-saturated T1-weighted sagittal MRI demonstrates abnormal T2 hyperintensity with some enhancement in the L5-S1 disk space as well as T2 hyperintensity with marrow enhancement in the L5-S1 end plates, indicating spondylodiskitis. There are enhancing anterior epidural and anterior paraspinal soft tissues extending from the upper L4 level down to the end of thecal sac. Note a few nonenhancing T2 hyperintense fluid collections (white arrows) within enhancing anterior epidural soft tissue, representing epidural abscesses. D and E, Postgadolinium fat-saturated T1-weighted axial MRIs also demonstrate two anterior epidural abscesses (black arrows) at the L5 and S1 levels compressing the thecal sac. An enhancing left-sided anterior paraspinal soft tissue phlegmon is noted.

developed countries, owing to increasing immigration from endemic areas, development of drug-resistant strains, and the human immunodeficiency virus (HIV) pandemic.41 Tuberculous spondylitis is more common in HIV-positive patients than normal hosts.28 The clinical presentation of tuberculous spondylitis include back pain, low-grade fever, nonspecific symptoms such as malaise or weight loss, and in the chronic stage, kyphotic deformity.32 The source of tuberculosis is commonly pulmonary by hematogenous dissemination, typically from a clinically quiescent primary focus.1 However, chest radiographs commonly do not demonstrate evidence of pulmonary tuberculosis in patients with tuberculous spondylitis.42 Tuberculous spondylitis most commonly involves the thoracic spine, particularly the thoracolumbar junction. The tuberculous  bacilli spread to the vertebral body hematogenously and lodge in the anterior subchondral regions of the vertebral body. Extension to  other vertebral bodies occurs beneath the anterior and posterior longitudinal ligaments and is commonly present at the time of diagnosis. When the infection spreads posteriorly, it can cause an epidural abscess,

intradural abscesses, tuberculous arachnoiditis, and spinal cord myelitis.1,13,23,28,32,41,42

Imaging Evaluation.  As with other spinal infections, MRI is the imaging modality of choice for evaluating tuberculous spondylitis. The classic imaging appearance is similar to that of pyogenic spondylodiskitis, including involvement of one or more adjacent vertebral levels preferentially at the vertebral end plates, involvement of the disk space, and adjacent paraspinal soft tissue inflammatory changes.32 However, disk-space involvement is often not present early in the disease and is delayed when compared with pyogenic spondylodiskitis. There are many atypical features of tuberculous spondylitis, such as isolated single vertebral lesions or multilevel disease (contiguous or noncontiguous) without disk involvement, mimicking metastases.32-34 Sparing of the disk space has been ascribed to the fact that Mycobacterium lacks proteolytic enzymes. The infection commonly spreads in a subligamentous fashion across one or more levels beneath the anterior or posterior longitudinal ligaments (Fig. 31-4). Spreading of the infection

http://radiologyebook.com

880

PART II  CT and MR Imaging of the Whole Body

A

B

D

E

C

F

FIG 31-4  Tuberculous spondylitis. This 40-year-old woman presented with chronic kyphosis and paraplegia for 2 months. Anteroposterior (A) and lateral (B) views of plain radiographs show mild kyphotic deformity of the upper thoracic spine, a mildly collapsed T6 vertebra, narrowing of the T6-T7 disk space with indistinct end plates, and bilateral paravertebral soft tissue lesions (black arrows). C, Postcontrast sagittal CT demonstrates the mildly collapsed T6 vertebra, involvement of the T6-T7 disk space, with large anterior epidural soft tissue (black arrow) compressing the thoracic cord. D, Postcontrast coronal CT confirms enhancing bilateral paravertebral soft tissues. E and F, Postcontrast axial CT at T6-T7 level demonstrates a large, thin, smooth rim-enhancing lesion (white arrows) at the right-sided paravertebral region, with internal small air bubble representing a tuberculous abscess.

into the epidural space may also occur but is less common than extension into the anterior paravertebral regions. Subligamentous spread may be much more extensive than the degree of vertebral involvement and can lead to skip lesions of involved bones/disks, characterized by intervening normal vertebral levels. Tuberculous abscess more commonly contain calcifications and are larger and have smoother and thinner walls when compared with pyogenic abscesses (see Fig. 31-4). The posterior elements are more commonly involved than in pyogenic disease, and isolated posterior element involvement has been reported in 15% of cases. Gibbus deformity is a classic radiographic finding of tuberculous spondylitis (Fig. 31-5; also see Fig. 31-4) because of the preferential involvement of the anterior vertebral column, causing collapse of the destroyed vertebral bodies. Similar to other spinal infections, radiographic findings of tuberculous spondylitis tend to lag behind the clinical setting.12,33

Treatment.  Antituberculous drugs are the main treatment, and the duration of treatment is generally 6 to 12 months to obtain a cure. Surgery may be performed in patients with neurologic complications or medical therapy failure, or occasionally for prevention of new or worsening deformity.41

Brucellar Spondylitis Brucellosis is a zoonosis endemic in Saudi Arabia and the Mediterranean countries. Persons at risk include farm workers, slaughterhouse workers, and veterinary personnel. Brucellosis is caused by a gramnegative bacillus and is usually transmitted by ingestion of uncooked meat or unpasteurized dairy products. The most common animals to harbor brucellosis are sheep, cattle, goats, pigs, and dogs.2,12 Blood cultures and serologic testing typically confirm the diagnosis.40 Treatment is with antibiotics for 3 to 6 months.41

http://radiologyebook.com

CHAPTER 31  Spinal Infection

A

B

D

881

C

E FIG 31-5  Tuberculous spondylodiskitis. This 47-year-old woman presented with back pain and weakness of both legs for 2 months A, T1 weighted sagittal MRI shows collapsed T6 and T7 vertebrae, with diffuse T1 hypointensity involving both vertebral bodies and posterior element. Kyphotic change or gibbus deformity at this level is seen. B, T2-weighted fat-saturated sagittal MRI demonstrates abnormal T2 hyperintensity in the T6 and T7 vertebral bodies in both vertebral bodies and posterior elements. Severe narrowing of the spinal canal with cord compression is seen. Marked narrowing of the T6-T7 disk space without definite T2 hyperintensity is noted. C, Postgadolinium fat-saturated T1-weighted sagittal MRI reveals diffuse bone marrow enhancement corresponding to the abnormal vertebral marrow T1 signal. There is subligamentous spread of infection beneath the anterior longitudinal ligament down to the T12 level (white arrows). Evidence of disk enhancement, enhancing ventral and dorsal epidural soft tissues, and thin anterior paraspinal contrast enhancement is seen. T2-weighted fat-saturated axial MRI (D) and postgadolinium fat-saturated T1-weighted axial MRI (E) confirm the vertebral marrow enhancement (black arrowhead in E), and enhancing paraspinal and epidural soft tissues (black arrows) encircle the thoracic cord, resulting in cord compression. In this case the epidural and paraspinal enhancement represents inflammatory phlegmon and/or venous engorgement without discrete abscess formation.

Imaging Evaluation.  Brucellosis commonly involves the lumbar and lumbosacral spine. The MRI findings of brucellosis may be indistinguishable from those of tuberculous spondylitis, but brucellosis tends to appear less aggressive than tuberculosis.39 Specific radiographic features that that suggest brucellosis over tuberculous spondylitis include relatively intact vertebral bodies and even normal disk spaces despite signal abnormality and contrast enhancement, sparing of the posterior

elements, and lack of associated spinal deformity. A relatively low incidence of abscess formation has also been reported.5

Fungal Spondylitis Fungal spondylodiskitis occurs in immunocompromised patients. Many fungal organisms can cause spine infection; most are Aspergillus and Candida species.37 The radiographic findings of fungal

http://radiologyebook.com

882

PART II  CT and MR Imaging of the Whole Body

spondylodiskitis are usually nonspecific and cannot be distinguished from those of pyogenic and tuberculous spondylodiskitis. One MRI finding that has been recently reported in fungal spondylitis is a lack of T2 hyperintensity in the disk spaces, in contrast to pyogenic spondylodiskitis where the disk is always T2 bright.47 The MRI findings of Aspergillus spondylitis are similar to those of tuberculous spondylitis, including involvement of multiple vertebral levels with skip lesions or subligamentous spread, serrated appearance of the vertebral end plates, and subchondral T2 hypointensity (Fig. 31-6). Therefore Aspergillus spondylitis should be suspected in immunocompromised patients

A

when either typical or atypical radiographic findings of spondylitis are present.24

SPINAL MENINGITIS Many organisms can access the leptomeningeal and subarachnoid spaces, resulting in spinal meningitis or arachnoiditis. These organisms can be bacteria, tuberculosis, viruses, and even parasites. The main transmission routes are hematogenous spread and contiguous  spread from a local site of infection. Spinal meningitis commonly

C

B

D FIG 31-6  Aspergillus spondylitis. This 60-year-old man with diabetes presented with neck pain and weakness of the upper extremities of 1 month’s duration. T1- (A) and T2-weighted fat-saturated (B) sagittal MRIs show diffuse T1 hypointensity/T2 hyperintensity extending from C6 to the T4 vertebral bodies. There are end-plate irregularities with subchondral low signal intensity (white arrows in A) adjacent to the disks and some hypointense foci in the anterior aspect of disks (white arrows in B). Abnormal T2 hyperintensity in C6-C7 down to T3-T4 disk spaces is also seen. C, Postgadolinium fat-saturated T1-weighted sagittal MRI reveals diffuse bone marrow enhancement (white arrows) at the C6 to T4 vertebrae, corresponding to the abnormal vertebral bone marrow signal. Evidence of disk enhancement and thin, enhancing, ventral epidural soft tissue and anterior paraspinal enhancement (black arrows) are seen. D, Postgadolinium fat-saturated T1-weighted axial MRI confirms vertebral marrow enhancement, enhancing ventral epidural soft tissue, and thin anterior paraspinal soft tissue.

http://radiologyebook.com

CHAPTER 31  Spinal Infection accompanies cerebral meningitis or spinal spondylodiskitis (see Fig. 31-2). Clinical presentation depends on the virulence of the organism, number of inoculated organisms, and immunologic status of the patient. Patients with spinal meningitis typically present with back and radicular pain, fever, and malaise. Cerebrospinal fluid (CSF) analysis and culture typically confirm the diagnosis. Pyogenic leptomeningitis is the most common bacterial infection of the spinal intradural axis. Neisseria meningitidis, S. aureus, and streptococci are the leading pathogens in adults. On MRI, postcontrast T1-weighted images are most useful and show contrast enhancement of the inflamed meninges on the surface of the spinal cord and/or nerve root sheaths. The severity of the contrast enhancement does not correlate with the severity of the disease, as is also the case in cerebral meningitis. MRI has low specificity in differentiating leptomeningitis from leptomeningeal tumor processes.44 Tuberculous leptomeningitis is frequently associated with spinal cord and nerve root involvement.7,44 Spinal tuberculous leptomeningitis occasionally causes a subarachnoid “block” preventing free move-

A

D

883

ment of CSF in the subarachnoid space. The imaging features of spinal tuberculous leptomeningitis include CSF loculations, obliteration of the spinal subarachnoid space with loss of outline of the spinal cord, and thickened, clumped cauda equina nerve roots.44 The cord and nerve roots may be swollen and the subarachnoid space may be isointense or hyperintense relative to the spinal cord on all MRI sequences. Extensive meningeal enhancement is frequently seen.35 Spinal viral meningitis is commonly accompanied by viral myelitis. Herpesvirus, poliovirus, Coxsackie genus, and HIV are the most common agents. MRI features of spinal viral meningitis are usually nonspecific and indistinguishable from other causes of leptomeningeal and/or nerve root enhancement (Figs. 31-7 and 31-8). The definitive diagnosis is made by isolation of the virus from the CSF.44

Intramedullary Cord Infection Infection of the spinal cord, or myelitis, is most commonly due to viral infection. The common viruses responsible are herpes zoster and HIV; occasionally a postviral immunologic process is involved.

B

C

E

FIG 31-7  CMV radiculitis. This 37-year-old man (known case of HIV and cytomegalovirus retinitis) presented with progressive leg weakness and bowel-bladder involvement for 1 month. T1-weighted (A) and T2-weighted (B) sagittal MRIs show clumping of the cauda equina nerve roots (white arrows). C, Postgadolinium T1-weighted sagittal MRI demonstrates diffuse enhancement of the clumped cauda equina nerve roots (white arrow). D and E, Postgadolinium T1-weighted axial MRIs again reveal diffuse enhancement, clumping and thickening of the cauda equina nerve roots (black arrows).

http://radiologyebook.com

884

PART II  CT and MR Imaging of the Whole Body

A

B

D

E

C

F

FIG 31-8  Viral myeloradiculitis. This 2-year-old girl (known case of hand-foot-mouth disease) developed acute paraparesis of both lower extremities (grade II) and a stiff neck. A and B, T2-weighted sagittal MRI shows extensive long-segment T2 hyperintensity (white arrows) in the brainstem and entire spinal cord down to conus medullaris. C, Postgadolinium T1-weighted sagittal MRI of cervicothoracic spine demonstrates leptomeningeal enhancement (black arrow) along the prepontine and perimedullary cisterns, as well as anterior to the cervical spinal cord. D and E, Postgadolinium T1-weighted sagittal MRIs of the thoracolumbar spine also reveal leptomeningeal enhancement along the distal thoracic cord and conus medullaris (black arrows in D), as well as diffuse mild thickening and enhancement of the cauda equina nerve roots (black arrows in E). No definite enhancement of the spinal cord is seen. F, Postgadolinium T1-weighted axial MRI at L2-L3 level demonstrates diffuse mild thickening and enhancement of the cauda equina nerve roots (black arrows).

Pyogenic abscess of the spinal cord is very rare and has been reported in patients with bacteremia.3 The spinal cord itself can sometimes be infected, and cord abscesses may form. Sources of spinal cord infection include hematogenous spread and direct spread, which is a known complication of congenital dermal sinuses. MRI is the imaging modality of choice to detect myelitis and typically shows a swollen and edematous cord with a variable area of T2 hyperintensity (see Fig.

31-8). The extent of abnormal cord T2 signal abnormality may range from one vertebral segment to extensive cord involvement. Ill-defined patchy enhancement in the involved cord segment is frequently observed.4,20 Cord abscesses typically demonstrate poorly defined ring contrast enhancement within the area of T2 hyperintensity. Cord abscesses are less well defined and less intensely enhanced than brain abscesses, owing to the relative hypovascularity of the cord compared

http://radiologyebook.com

CHAPTER 31  Spinal Infection to the brain.3,4,20 Imaging features of spinal cord infection are usually nonspecific and sometimes indistinguishable from other noninfectious causes of transverse myelitis, including demyelinating disease. Clinical correlation and further workup with CSF or laboratory evaluation is recommended.

SUMMARY Spinal infection is a common condition that has been increasing in incidence owing to immigration and immunosuppression. Nowadays imaging plays an important role in the diagnosis and follow-up of these potentially life-threatening conditions. MRI is the imaging modality of choice for patients with suspected spinal infection, and protocols should include fat-suppressed T2- and fat-suppressed contrast-enhanced T1-weighted sequences to better detect and delineate the extent of the spinal infection. Pyogenic spondylodiskitis should be considered when there is increased T2-weighted signal and/ or enhancement in the disk and subchondral regions of adjacent vertebral bodies. Spinal epidural and subdural abscesses generally accompany pyogenic spondylodiskitis. MRI is critical in diagnosing both spinal epidural and subdural abscesses, which need to be treated with emergent surgery. Tuberculous spondylitis is the most common spinal infection worldwide, and its characteristic imaging findings include disk space sparing, subligamentous spread, and large paraspinal abscesses. Brucellar spondylitis is commonly seen in endemic regions and can be indistinguishable from tuberculous spondylitis, but it tends to be less aggressive in appearance. Fungal spondylitis is a rare disease found in immunocompromised patients and has a nonspecific imaging appearance. In the setting of spinal infections, spinal meningitis and cord infection can occur spontaneously or accompany spondylodiskitis. Therefore the spinal cord and intradural compartments should be carefully evaluated in all patients with spondylodiskitis.

REFERENCES 1. Almeida A: Tuberculosis of the spine and spinal cord. Eur J Radiol 55(2):193–201, 2005. 2. Al-Sous MW, Bohlega S, Al-Kawi MZ, et al: Neurobrucellosis: clinical and neuroimaging correlation. AJNR Am J Neuroradiol 25(3):395–401, 2004. 3. Baby R, Jafer JJ, Hray PP, et al: Intramedullary abscess associated with spinal cord ependymoma. Neurosurgery 30:121–124, 1992. 4. Barakos JA, Mark AS, Dillon WP, et al: MR imaging of acute transverse myelitis and AIDS myelopathy. J Comput Assist Tomogr 14:45–50, 1990. 5. Bozgeyik Z, Ozdemir H, Demirdag K, et al: Clinical and MRI findings of brucellar spondylodiscitis. Eur J Radiol 67(1):153–158, 2008. 6. Carragee EJ: The clinical use of magnetic resonance imaging in pyogenic vertebral osteomyelitis. Spine (Phila Pa 1976) 22(7):780–785, 1997. 7. Chang KH, Han MH, Choi YW, et al: Tuberculous arachnoiditis of the spine: Findings on myelography, CT, and MR imaging. AJNR Am J Neuroradiol 10(6):1255–1262, 1989. 8. Cohen DB: Infectious origins of cauda equina syndrome. Neurosurg Focus 16(6):e2, 2004. 9. Cottle L, Riordan T: Infectious spondylodiscitis. J Infect 56(6):401–412, 2008. 10. Dagirmanjian A, Schils J, McHenry MC: MR imaging of spinal infections. Magn Reson Imaging Clin N Am 7:525–538, 1999. 11. Dagirmanjian A, Schils J, McHenry M, et al: MR imaging of vertebral osteomyelitis revisited. AJR Am J Roentgenol 167(6):1539–1543, 1996. 12. da Rocha AJ, Maia AC, Jr, Ferreira NP, et al: Granulomatous diseases of the central nervous system. Top Magn Reson Imaging 16(2):155–187, 2005. 13. Diehn FE: Imaging of spinal infection. Radiol Clin North Am 50(4): 777–798, 2012.

885

14. Eastwood JD, Vollmer RT, Provenzale JM: Diffusion weighted imaging in a patient with vertebral and epidural abscesses. AJNR Am J Neuroradiol 23(3):496–498, 2002. 15. Euba G, Narvaez JA, Nolla JM, et al: Long-term clinical and radiological magnetic resonance imaging outcome of abscess-associated spontaneous pyogenic vertebral osteomyelitis under conservative management. Semin Arthritis Rheum 38(1):28–40, 2008. 16. Gillams AR, Chaddha B, Carter AP: MR appearances of the temporal evolution and resolution of infectious spondylitis. AJR Am J Roentgenol 166(4):903–907, 1996. 17. Gouliouris T, Aliyu SH, Brown NM: Spondylodiscitis: Update on diagnosis and management. J Antimicrob Chemother 65(Suppl 3): iii11–iii24, 2010. 18. Hooten WM, Mizerak A, Carns PE, et al: Discitis after lumbar epidural corticosteroid injection: A case report and analysis of the case report literature. Pain Med 7(1):46–51, 2006. 19. Hsu CY, Yu CW, Wu MZ, et al: Unusual manifestations of vertebral osteomyelitis: Intraosseous lesions mimicking metastases. AJNR Am J Neuroradiol 29(6):1104–1110, 2008. 20. Isoda H, Ramsey RG: MR imaging of acute transverse myelitis (myelopathy). Radiat Med 16:179–186, 1998. 21. James SL, Davies AM: Imaging of infectious spinal disorders in children and adults. Eur J Radiol 58(1):27–40, 2006. 22. Kowalski TJ, Layton KF, Berbari EF, et al: Follow-up MR imaging in patients with pyogenic spine infections: Lack of correlation with clinical features. AJNR Am J Neuroradiol 28(4):693–699, 2007. 23. Kulkarni AG, Chu G, Fehlings MG: Pyogenic intradural abscess: A case report. Spine (Phila Pa 1976) 32(12):E354–E357, 2007. 24. Kwon JW, Hong SH, Choi SH, et al: MRI findings of Aspergillus spondylitis. AJR Am J Roentgenol 197(5):W919–W923, 2011. 25. Ledermann HP, Schweitzer ME, Morrison WB, et al: MR imaging findings in spinal infections: Rules or myths? Radiology 228(2):506–514, 2003. 26. Longo M, Granata F, Ricciardi GK: Contrast-enhanced MR imaging with fat suppression in adult-onset septic spondylodiscitis. Eur Radiol 13:626–637, 2003. 27. Lury K, Smith JK, Castillo M: Imaging of spinal infections. Semin Roentgenol 41(4):363–379, 2006. 28. Moon MS: Tuberculosis of the spine. Controversies and a new challenge. Spine (Phila Pa 1976) 22(15):1791–1797, 1997. 29. Muller EJ, Russe OJ, Muhr G: Osteomyelitis of the spine. Orthopade 33:305–315, 2004. 30. Mylona E, Samarkos M, Kakalou E, et al: Pyogenic vertebral osteomyelitis: A systematic review of clinical characteristics. Semin Arthritis Rheum 39(1):10–17, 2009. 31. Numaguchi Y, Rigamonti D, Rothman MI, et al: Spinal epidural abscess: Evaluation with gadolinium enhanced MR imaging. Radiographics 13(3):545–559, 1993. 32. Pande KC, Babhulkar SS: Atypical spinal tuberculosis. Clin Orthop Relat Res 398:67–74, 2002. 33. Pertuiset E, Beaudreuil J, Liote F, et al: Spinal tuberculosis in adults.   A study of 103 cases in a developed country, 1980-1994. Medicine (Baltimore) 78(5):309–320, 1999. 34. Polley P, Dunn R: Noncontiguous spinal tuberculosis: Incidence and management. Eur Spine J 18(8):1096–1101, 2009. 35. Provenzale JM, Jinkins JR: Brain and spine imaging findings in AIDS patients. Radiol Clin North Am 35(5):1127–1166, 1997. 36. Ross JS, Zepp R, Modic MT: The postoperative lumbar spine: Enhanced MR evaluation of the intervertebral disk. AJNR Am J Neuroradiol 17(2):323–331, 1996. 37. Saigal G, Donovan Post MJ, Kozic D: Thoracic intradural Aspergillus abscess formation following epidural steroid injection. AJNR Am J Neuroradiol 25(4):642–644, 2004. 38. Schellinger D: Patterns of anterior spinal canal involvement by neoplasms and infections. AJNR Am J Neuroradiol 17:953–959, 1996. 39. Sharif HS, Aideyan OA, Clark DC, et al: Brucellar and tuberculous spondylitis: Comparative imaging features. Radiology 171(2):419–425, 1989.

http://radiologyebook.com

886

PART II  CT and MR Imaging of the Whole Body

40. Sharif HS, Clark DC, Aabed MY, et al: Granulomatous spinal infections: MR imaging. Radiology 177(1):101–107, 1990. 41. Skaf GS, Kanafani ZA, Araj GF, et al: Non-pyogenic infections of the spine. Int J Antimicrob Agents 36(2):99–105, 2010. 42. Smith AS, Weinstein MA, Mizushima A, et al: MR imaging characteristics of tuberculous spondylitis vs vertebral osteomyelitis.   AJR Am J Roentgenol 153(2):399–405, 1989. 43. Sobottke R, Seifert H, Fatkenheuer G, et al: Current diagnosis and treatment of spondylodiscitis. Dtsch Arztebl Int 105(10):181–187, 2008. 44. Tali ET, Gultekin S: Spinal infections. Eur Radiol 15:599–607, 2005. 45. Tompkins M, Panuncialman I, Lucas P, et al: Spinal epidural abscess.   J Emerg Med 39(3):384–390, 2010.

46. Vural M, Arslantas A, Adapinar B, et al: Spinal subdural Staphylococcus aureus abscess: Case report and review of the literature. Acta Neurol Scand 112(5):343–346, 2005. 47. Williams RL, Fukui MB, Meltzer CC, et al: Fungal spinal osteomyelitis in the immunocompromised patient: MR findings in three cases. AJNR Am J Neuroradiol 20(3):381–385, 1999. 48. Yoon SH, Chung SK, Kim KJ, et al: Pyogenic vertebral osteomyelitis: Identification of microorganism and laboratory markers used to predict clinical outcome. Eur Spine J 19(4):575–582, 2010. 49. Zarrouk V, Feydy A, Salles F, et al: Imaging does not predict the clinical outcome of bacterial vertebral osteomyelitis. Rheumatology (Oxford) 46(2):292–295, 2007.

http://radiologyebook.com

32  Cystic Lesions Gustavo A. Tedesqui and Mauricio Castillo

A broad spectrum of cystic lesions can be identified on spine imaging. Multiple etiologies and different imaging characteristics of these lesions make the differential diagnosis a challenging issue for radiologists. Cysts may develop within the spinal canal and may be grouped according to their location as intramedullary, intradural extramedullary, or extradural lesions. This classification is useful in planning treatment, but it is less useful for narrowing the differential diagnosis because most of the lesions involve several compartments of the spine and even the paraspinal regions. Therefore separating these lesions from one another is best accomplished on the basis of multimodality imaging, including ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI). MRI is an excellent noninvasive technique for detection and evaluation of spinal cystic lesions. US is an important modality in pediatric patients, whereas CT is helpful in vertebral column lesions. Multiple types of spinal cysts have been described, each one with different clinical implications. Cystic lesions of the spine can be related to infection or be posttraumatic, postinterventional, or degenerative. The most common cystic spinal lesions include synovial/ganglion cysts, simple bone cysts, hydatid cysts, aneurysmal bone cysts,  diskal cysts, lateral spinal meningoceles, perineural root sleeve cysts (Tarlov’s cyst), spontaneous meningeal diverticula, occult intrasacral meningocele, neuroenteric cysts, arachnoid cysts, epidermoid cysts, dermoid cysts, spinal cysticercosis, ventriculus terminalis, syringohydromyelia, cystic myelomalacia, cystic hematoma, and cystic tumors.20,26,27 This chapter discusses the study of nonneoplastic spinal cystic lesions, emphasizing their imaging characteristics and concentrating on the three most common types: arachnoid cyst, perineural root sleeve cyst (also known as Tarlov’s cyst), and syringomyelia.

ARACHNOID CYST Spinal arachnoid cysts are expansile lesions located within the spinal canal and are also known as meningeal cysts, meningeal diverticula, or pouches.8,9,28 Spinal arachnoid cysts are uncommon, accounting for about 1% of all spinal masses.6 These lesions occur most frequently in the thoracic spine, followed by the lumbosacral spine, and lastly the thoracolumbar spine. Most are located posteriorly within the spinal canal. These cysts can occur in either gender and have been reported in patients as young as a few months old to nearly 80 years of age.7,22,25,55 Patients may have more than one spinal arachnoid cyst or both a spinal arachnoid cyst and a concurrent intracranial arachnoid cyst.4 The Nabors classification of spinal meningeal cysts differentiates three major categories25: Type I: extradural cyst without nerve root involvement Type IA: extradural arachnoid cyst Type IB: sacral meningocele

Type II: extradural cysts with nerve root involvement Type III: intradural cysts

Clinical Presentation Most patients with spinal arachnoid cysts are asymptomatic, and the lesions are incidentally found on MRI.22 The clinical presentation may include progressive signs and symptoms of spinal cord compression. Because a cyst may occur at any spinal level, the clinical presentation depends on location as well as on its extension. Cervical cysts may present as spastic tetraparesis and sensory disturbances. Thoracic cysts may cause progressive spastic paraparesis and sensory symptoms. Lumbar cysts typically present as low back pain, leg pain, weakness, and/or numbness. Less typical presentations include noncardiac chest pain, isolated gait difficulty, and isolated urinary urgency.7,18,22,25 Symptoms may be exacerbated by a change in posture and/or Valsalva maneuver.7,18,22,42,55 Clinical signs and symptoms develop over months to years, although approximately one third of patients show a fluctuating course.11 Furthermore, some patients may have long-term remissions extending over many years.51

Pathophysiology Explanations as to how spinal arachnoid cysts form are multiple.7,22,34 Fortuna et al. suggested that incarceration of arachnoid granulations may produce cerebrospinal fluid (CSF) that becomes trapped in arachnoid diverticula. These sequestered pockets of fluid lead to further disruption of normal pulsatile CSF flow and thus are capable of expanding or producing associated syringomyelia.13,14 However, histologic findings often fail to reveal active secretion and absorption of CSF by the wall of the cyst.45 Several mechanisms have been proposed to explain why spinal arachnoid cysts enlarge. Many reports support the presence of a valvelike mechanism. A unidirectional “valve” may allow fluid into but not out of the cyst.28,35,45 Another proposed cause of cyst expansion is an osmotic spinal gradient between the subarachnoid space and the cyst; pulsatile CSF dynamics may also play a role in cyst formation.15 Acquired or secondary spinal arachnoid cysts are most common and result from spinal trauma, postsurgical arachnoiditis, meningeal infection, and other insults that cause inflammation and subarachnoid adhesions.24,50 Congenital and primary arachnoid cysts are diagnosed when there is no antecedent history of trauma or other etiologies. These cysts are rare and have been described in children in association with neural tube defects.43 In adults these cysts have been described in association with deformities of the spinal column.1

Imaging Most primary spinal arachnoid cysts are dorsal to the cord in location. They occur most frequently in the thoracic spine, followed by the

http://radiologyebook.com

887

888

PART II  CT and MR Imaging of the Whole Body

lumbosacral spine, and the thoracolumbar spine. Secondary spinal arachnoid cysts can be located anywhere.1,14,43,49 Most radiologic articles report that MRI adds sensitivity and specificity to the diagnosis of spinal arachnoid cyst and recommend it as the diagnostic procedure of choice.44,49

Radiography.  Plain radiographs are not useful in the diagnosis of spinal cyst. Only indirect signs attributable to the mass effect of large cysts (e.g., spinal canal expansion, bone erosion, enlarged foramina, posterior vertebral scalloping) are visualized.

CT Myelography.  Intradural arachnoid cysts usually communicate with the subarachnoid space and opacify after intrathecal administration of contrast agents in both myelography and CT myelography. Cysts with a large communication to the subarachnoid space are rapidly opacified and may appear to have the same attenuation values as the subarachnoid space. However, some cysts may not show immediate filling after injection of contrast into the subarachnoid space, and instead the cyst may fill hours later, and thus delayed CT may be indicated. When a communication is absent the cyst remains of low attenuation on myelo-CT.49

MRI.  MRI is useful to identify cysts, determine their precise anatomic locations and extensions, and define their relationship to surrounding structures. On MRI arachnoid cysts are well-circumscribed oval or elongated lesions. Sometimes these cysts follow the signal intensity of CSF and their walls are generally invisible; they may not be identified unless the spinal cord is displaced. In such cases, constructive interference in steady state (CISS) pulse sequences may show the thin cyst wall that separates it from the surrounding subarachnoid space (Fig. 32-1).

Combined intraforaminal and extraforaminal cysts may assume a dumbbell shape. Epidural fat capping of the lesion (the “cap sign”) at its superior and inferior poles can be seen on sagittal T1-weighted imaging and suggests its extradural location. In all sequences the signal intensity within the cysts is isointense to CSF, hyperintense on T2, and hypointense on T1-weighted images. No contrast enhancement is seen (Fig. 32-2). No evidence of restricted diffusion is seen on diffusionweighted imaging (DWI). Varying degrees of cord atrophy and myelomalacia can also be seen by MRI. Determination of whether the arachnoid cyst communicates with the CSF-containing spaces is important in the preoperative evaluation, and this finding may not always be evident with only anatomic imaging. Therefore CSF flow studies such as cinematic MRI (cine-MRI) may improve diagnostic confidence in differentiating communicating and noncommunicating arachnoid cysts, detecting the specific communication site, and thus allowing treatment by selective closure of the dural defect.36

PERINEURAL ROOT SLEEVE CYST (TARLOV’S CYST) Tarlov’s cysts are isolated or multiple nerve root cysts usually occurring in the sacral spine.52,53 They were first described by Tarlov in 1938 as an incidental finding at autopsy.52 Since the original description of these sacral cysts, numerous articles have appeared in the literature that use different and often confusing terms to describe them. In 1988, Nabors et al. classified spinal cysts into three types35: extradural meningeal cysts without spinal nerve-root fibers (type I), spinal extradural meningeal cysts with spinal nerve root fibers (type II: Tarlov’s cysts), and spinal intradural meningeal cysts (type III). Regardless of the

A

B

FIG 32-1  Arachnoid cysts in a 6-year-old boy presenting with hydrocephalus. Parasagittal (A) and coronal (B) high-resolution CISS sequences show presence of multifocal arachnoid cysts (white arrows) anteriorly to the cervical spine and prepontine cistern. The cerebellar tonsils are inferiorly displaced.

http://radiologyebook.com

CHAPTER 32  Cystic Lesions

A

889

C

B

FIG 32-2  Arachnoid cyst in a 3-year-old child; MRI shows typical pattern of a thoracic arachnoid cyst. Sagittal T1- (A) and T2-weighted (B) images demonstrate a mass displacing the spinal cord posteriorly. The mass has the same signal intensity as CSF, and a sagittal T1 image (C) with contrast shows lack of enhancement.

classification system, the definition of a Tarlov’s cyst is histopathologic because it requires identification of of spinal nerve root fibers in the wall of the cyst or in its cavity.37 Tarlov’s cysts are typically located at the junction of the dorsal root ganglion and the posterior nerve root and usually develop between the endoneurium and perineurium.17 Tarlov’s cysts are not rare lesions, usually incidental, and occur in about 1% to 5% of patients.41

Pathophysiology

thy, sexual disorders, leg weakness, neurogenic claudication, abdominal pain, dyspareunia, and urinary incontinence.19,23 Symptoms are often exacerbated by standing, sitting, and walking, owing to increased CSF pressure leading to activation of the valve mechanism.33 Symptomatic relief can usually be achieved by recumbent positioning.23

Imaging

Radiography.  Radiographs are usually normal but may reveal a widened canal, enlarged foramina, and bone erosion.

Tarlov’s cysts are defined as CSF-filled saccular lesions located in the extradural space of the sacral spinal canal and are formed within the nerve root sheath at the dorsal root ganglion between the perineurium and endoneurium. Nerve fibers are always present inside or on the wall of the cyst that is usually formed by perineurium, neuronal, and fibrous connective tissue.27,35,40,52 There is no clear consensus on the pathogenesis of perineural cysts. Different hypotheses, such as inflammation, trauma, congenital origin, degenerative processes, and genetic ones, have been proposed to explain their development.19,40 Although their etiology is unclear, the mechanisms of cyst formation include microcommunications with the subarachnoid space at the dural sleeve of the nerve root working as valves, allowing CSF influx and restricting CSF efflux and causing formation of these cysts.35,37 Perineural cysts begin to fill and expand in size, compressing neighboring nerve fibers and resulting in neurologic symptoms.

Clinical Presentation Most individuals with Tarlov’s cysts are asymptomatic. The reported incidence of symptomatic cysts is about 1% or less of them all.23,39 Clinical presentation of Tarlov’s cysts is variable and similar to that of other lumbar spine disorders and include back pain, sacral radiculopa-

CT.  CT may demonstrate cystic masses isodense to CSF located at the foramina, which may be enlarged, and an enlarged sacral canal. Myelography showing filling of the cyst up to 1 hour after injection of contrast medium is suggestive of a perineural cyst.

MRI.  MRI is the gold standard to detect sacral perineural cysts, study their relationship with surrounding structures, and plan surgical or percutaneous treatment. The cysts demonstrate low signal on T1-weighted images and high signal on T2-weighted images, similar to CSF. Scalloping of the sacral vertebral body or posterior arches, caused by a slow increase in size of the cysts, can be seen on both MRI and CT. A Tarlov’s cyst does not enhance with gadolinium contrast agents (Figs. 32-3 and 32-4).

SYRINGOMYELIA Syringomyelia refers to a heterogeneous collection of conditions characterized by the presence of abnormal fluid-filled cavities within the spinal cord. The term syrinx is Greek for a cavity of tubular shape. Syringomyelia was suggested in 1827 by Ollivier d’Angers as a name for cystic cavitations of the spinal cord.38

http://radiologyebook.com

890

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 32-3  Perineural root sleeve cyst (Tarlov’s cyst). Sagittal T1- (A) and T2-weighted (B) images of the lumbosacral spine show CSF intensity and well-circumscribed Tarlov’s cyst (arrow) slightly remodeling the posterior aspect of S2.

A

B

FIG 32-4  Giant Tarlov’s cyst. Sagittal (A) and axial T2 with fat suppression (B) through the lumbar spine demonstrate a paraspinal hyperintense lesion that appears to communicate and have the same intensity as CSF, which proved to be a giant perineural root sleeve cyst (Tarlov’s cyst).

http://radiologyebook.com

CHAPTER 32  Cystic Lesions Specific terms are used when describing a cystic lesion within the spinal cord and include57: • Hydromyelia: cystic dilatation of the central spinal cord canal (the cavity wall is lined by ependymal) • Syringomyelia: a cystic cavity within the spinal cord that may or may not communicate with the central spinal cord canal (the cavity wall is not lined by ependyma) • Syringohydromyelia: term used for either of the above because differentiation between them by imaging is difficult; this term is often used to encompass both possibilities • Syrinx: a common term employed to describe any spinal cord cavity Syringomyelia affects mainly children and young adults and in most cases presents in late adolescence. No gender predilection is known. Estimates indicate a prevalence of 9 per 100,000 and an incidence of 0.44 cases per year.31,32 Syringomyelia may arise in association with Chiari malformations (both types 1 and 2).32 It is seen in about 50% of Chiari I patients, predominantly in the cervical region. When associated with Chiari II malformation it is most commonly found in the distal tethered spinal cord. However, in both malformations it may occur anywhere in the cord. Posttraumatic syringomyelia is a complication of spinal trauma developing in up to 21% to 28% of such patients. There is a significant male preponderance in this group, which follows the gender distribution in spinal cord injury.2,10,56

Clinical Presentation Clinical presentation varies with the location of the syrinx. Common initial symptoms include segmental pain and sensory loss. Symptomatic progression is usually gradual, although sudden deterioration from hemorrhage into a syrinx has been described. Other associated clinical presentations include asymmetric reduction in reflexes, hyperhidrosis, autonomic dysreflexia, dysphagia, neuropathic joints (especially shoulders), and cardiorespiratory dysfunction. In rare cases, the syrinx can extend into the brainstem and cause bulbar symptoms.10,21,47 Asymptomatic patients are now diagnosed readily with MRI and are likely to have smaller and shorter syrinxes than patients who are symptomatic.54 Other associated abnormalities include hydrocephalus, Chiari malformations, spinal dysraphism syndromes, tethered cord, scoliosis, arachnoid cyst, and rarely meningioma and degenerative changes.32,57 Because syringomyelia is a potential long-term complication of spinal trauma, late neurologic deterioration even years after spinal trauma needs careful evaluation to rule out syringomyelia.47,54

Pathophysiology The most common causes of syringomyelia are caudal protrusion of the posterior fossa contents into the upper cervical spinal canal, as in Chiari I and II malformations, followed by tethered cord.21,31 Most theories of the pathogenesis of syringomyelia in association with Chiari malformations emphasize obstruction of CSF flow at the foramen magnum level, altering the balance of hydrostatic forces between the extramedullary CSF within the subarachnoid space and the intraparenchymal forces within the spinal cord. These may cause mechanical stress on the patent central canal or on the surface of the spinal cord. Therefore spontaneous resolution of a syrinx may be observed in some Chiari I malformation patients and could be secondary to differential growth between the bone structures (skull and spine) and the central nervous system, with improvement in CSF flow.5,21,30 Additionally, syrinx resolution is commonly observed following Chiari I decompression, with reestablishment of normal or nearly normal CSF flow at the foramen magnum. The exact pathogenesis of posttraumatic syrinx formation remains unclear. The initial formation of a spinal cord cavity may involve

891

different mechanisms. An inflammatory response to trauma results in edema and sometimes in cyst formation. Absorption of a hematoma may leave behind a cystic cavity that forms the basis of a subsequent syrinx.12,16 Formation of the cavity may occur months or years after trauma. It is evident that syrinx enlargement results from an imbalance between fluid inflow and outflow. One commonly suggested mechanism is that respiratory- or arterial-induced CSF pulsations in the subarachnoid space increase pressure within the perivascular spaces and/or microcysts, which then dissect into cord tissue and coalesce, producing larger spaces filled with fluid.56 Syringohydromyelia is also observed in association with spinal cord tumors and spinal arachnoid cysts. Fluid filtrating from tumor vessels and secretion by tumor cells are generally thought to be the main cause of syrinx development in patients with intramedullary tumors.3,18,46

Imaging In the past, before the availability of MRI, syrinx was difficult to diagnose clinically, and these patients represented difficult diagnostic problems. Currently MRI is the investigational mode of choice to diagnose and monitor syringomyelia.

Ultrasonography.  A hypoechoic intramedullary cavitation or central canal dilatation can be seen. Intraoperative US is useful to localize the syrinx, reveal dural and subarachnoid adhesions and webs, define intrasyringal septations, determine the safest place to open the dura, and confirm shunt placement. After shunting and fenestration, intraoperative sonography reveals collapse of the cyst.

Radiography.  The only radiographic finding is widening of the spinal canal. Neuropathic changes in one or both shoulders and elbows should prompt a search for a syrinx.

CT.  CT shows a normal or widened osseous spinal canal. CT myelography can help reveal arachnoid adhesions or cysts that may not be depicted on conventional MRI and can be the cause of syrinx formation. Delayed filling of the syrinx may occur after intrathecal administration of iodinated contrast and may necessitate repeat CT 6 to 8 hours after the initial one.

MRI.  MRI is the best imaging modality to identify a syrinx and accompanying pathologic processes and demonstrates most associated abnormalities. MRI demonstrates that syrinxes are often asymmetric, loculated, and multiple. Typically a syrinx shows high signal on T2-weighted and low signal on T1-weighted imaging, similar to CSF (Fig. 32-5). Flow voids in CSF and other signs of fluid dephasing are thought to indicate growth potential and an actively expanding syrinx. These are better seen by turning off the flow-compensating gradients on T2 sequences. “Flame-shaped” areas of edema cephalad or caudal to the edges of a syrinx are thought to precede growth and have been termed a presyrinx stage. An idiopathic syrinx does not enhance with gadolinium contrast agents. T1 with gadolinium enhancement rather suggests the presence of inflammatory or neoplastic lesions. Cine phase-contrast MRI demonstrates obstruction of the craniocaudal CSF flow during cardiac systole at the level of the foramen magnum in patients with Chiari I malformations or arachnoid webs and syringomyelia. Postoperatively, cine-MRI plays an important role in demonstrating normalization of CSF flow following surgery for Chiari I, which has proven to be an important prognostic factor for symptom and syrinx resolution.29,56 Anatomic reformation of the cisterna magna is imperative for normalization of CSF flow at the foramen magnum. Cine-MRI may also help confirm spinal cord tethering and communication of spinal cord cysts with the subarachnoid spaces.48

http://radiologyebook.com

892

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 32-5  Syringomyelia. Sagittal T1- (A) and T2-weighted (B) images demonstrate a large septated syrinx extending through the cervical and thoracic cord, associated with tonsilar ectopia in a patient with Chiari I malformation. On the T2 image, note areas of darkness due to dephasing of signal secondary to fluid turbulence and motion, probably related to active growth of the syrinx.

REFERENCES 1. Alvisi C, Cerisoli M, Giulioni M, et al: Long-term results of surgically treated congenital intradural spinal arachnoid cysts. J Neurosurg 67:333–335, 1987. 2. Backe HA, Betz RR, Mesgarzadeh M, et al: Post-traumatic spinal cord cysts evaluated by magnetic resonance imaging. Paraplegia 29:607–612, 1991. 3. Baker KB, Moran CJ, Wippold FJ, II, et al: MR imaging of spinal hemangioblastomas. AJR Am J Roentgenol 174:377–382, 2000. 4. Bitaraf MA, Zeinalizadeh M, Meybodi AT, et al: Multiple extradural spinal arachnoid cysts: A case report and review of the literature. Cases J 2:7531–7542, 2009. 5. Castillo M, Wilson JD: Spontaneous resolution of a Chiari I malformation: MR demonstration. AJNR Am J Neuroradiol 16: 1158–1160, 1995. 6. Chae KH, Kim SW, Shin H: Septated extradural arachnoid cyst in thoracolumbar spine causing myelopathy. J Korean Neurosurg Soc 40:3813, 2006. 7. Choi JY, Kim SH, Lee WS, et al: Spinal extradural arachnoid cyst. Acta Neurochir (Wien) 148:579–585, 2006. 8. Cloward RB: Congenital spinal extradural cysts: Case report with review of literature. Ann Surg 168:851–864, 1968. 9. Dyck P: Myelopathy due to spinal arachnoid cyst. Spine 11:80, 1986. 10. Edgar R, Quail P: Progressive post-traumatic cystic and non-cystic myelopathy. Br J Neurosurg 8:7–22, 1994. 11. Elsberg CA, Dyke CG, Brewer ED: Symptoms and diagnosis of extradural cysts. Bull Neurol Inst NY 3:395–417, 1934. 12. Fischbein NJ, Dillon WP, Cobbs C, et al: The ‘‘presyrinx’’ state: A reversible myelopathic condition that may precede syringomyelia. AJNR Am J Neuroradiol 20:7–20, 1999. 13. Fortuna A, Mercuri S: Intradural spinal cysts. Acta Neurochir (Wein) 68:289–314, 1983.

14. Fortuna A, Torre EL, Ciapetta P: Arachnoid diverticula: A unitary approach to spinal cysts communicating with the subarachnoid space. Acta Neurochir (Wein) 39:259–268, 1977. 15. Gortvai P: Extradural cysts of the spinal canal. J Neurol Neurosurg Psychiatry 26:223–230, 1963. 16. Guizar-Sahagun G, Grijalva I, Madrazo I, et al: Development of post-traumatic cysts in the spinal cord of rats subjected to severe spinal cord contusion. Surg Neurol 41:241–249, 1994. 17. Guo D, Shu K, Chen R, et al: Microsurgical treatment of symptomatic sacral perineural cysts. Neurosurgery 60:1059–1066, 2007. 18. Holly LT, Batzdorf U: Syringomyelia associated with intradural arachnoid cysts. J Neurosurg Spine 5:111–116, 2006. 19. Hung MW, Wanigasiri UI, Ong CL: Perineural (Tarlov) cysts mimicking adnexal masses: A report of three cases. Ultrasound Obstet Gynecol 34(2):230–233, 2009. 20. Ishizaka S, Hayashi K, Otsuka M, et al: Syringomyelia and arachnoid cysts associated with spinal arachnoiditis following subarachnoid hemorrhage. Neurol Med Chir (Tokyo) 52(9):686–690, 2012. 21. Klekamp J, Samii M: Syringomyelia: Diagnosis and management, ed 1, Berlin, 2002, Springer, p 195. 22. Kumar K, Malik S, Schulte PA: Symptomatic spinal arachnoid cysts: Report of two cases with review of the literature. Spine 28:E25–E29, 2003. 23. Langdown AJ, Grundy JR, Birch NC: The clinical relevance of Tarlov cysts. J Spinal Disord Tech 18:29–33, 2005. 24. Lee T, Alameda G, Camilo E, et al: Surgical treatment of post-  traumatic myelopathy associated with syringomyelia. Spine 26: S119–S127, 2001. 25. Liu JK, Cole CD, Sherr GT, et al: Noncommunicating spinal extradural arachnoid cyst causing spinal cord compression in a child. J Neurosurg 103(3 Suppl):266–269, 2005. 26. Lohani S, Rodriguez D, Lidov H, et al: Intrasacral meningocele in the pediatric population. J Neurosurg Pediatr 11(6):615–622, 2013.

http://radiologyebook.com

CHAPTER 32  Cystic Lesions 27. Lucantoni C, Than KD, Wang AC, et al: Park Tarlov cysts: A controversial lesion of the sacral spine P. Neurosurg Focus 31(6):E14.3171, 2011. 28. McCrum C, Williams B: Spinal extradural arachnoid pouches. Report of two cases. J Neurosurg 57:849–852, 1982. 29. Menick BJ: Phase-contrast magnetic resonance imaging of cerebrospinal fluid flow in the evaluation of patients with Chiari I malformation. Neurosurg Focus 11:1–4, 2001. 30. Mikulis DJ, Diaz O, Egglin TK, et al: Variance of the position of the cerebellar tonsils with age: Preliminary report. Radiology 183:725–728, 1992. 31. Milhorat TH, Chou MW, Trinidad EM, et al: Chiari I malformation redefined: Clinical and radiographic findings for 364 symptomatic patients. Neurosurgery 44:1005–1017, 1999. 32. Moriwaka F, Tashiro K, Tachibana S, et al: Epidemiology of syringomyelia in Japan—The nationwide survey. Rinsho Shinkeigaku 35:1395–1397, 1995. 33. Murphy KJ, Nussbaum DA, Schnupp S, et al: Tarlov cysts: An overlooked clinical problem. Semin Musculoskelet Radiol 15:163–167, 2011. 34. Muthukumar N: Sacral extradural arachnoid cyst: A rare cause of low back and perineal pain. Eur Spine J 11:162–166, 2002. 35. Nabors MW, Pait TG, Byrd EB, et al: Updated assessment and current classification of spinal meningeal cysts. J Neurosurg 68:366–377, 1988. 36. Neo M, Koyama T, Sakamoto T, et al: Detection of a dural defect by cinematic magnetic resonance imaging and its selective closure as a treatment for a spinal extradural arachnoid cyst. Spine 29(19):E426– E430, 2004. 37. Neulen A, Kantelhardt SR, Pilgram-Pastor SM, et al: Microsurgical fenestration of perineural cysts to the thecal sac at the level of the distal dural sleeve. Acta Neurochir (Wien) 153:1427–1434, 2011. 38. Ollivier D’Angers CP: De le moelle epiniere et de ses maladies, ed 2, Paris, 1827, Chez Crevot. 39. Park HJ, Jeon YH, Rho MH, et al: Incidental findings of the lumbar spine at MRI during herniated intervertebral disk disease evaluation. AJR Am J Roentgenol 196:1151–1155, 2011. 40. Park HJ, Kim IS, Lee SW, et al: Two cases of symptomatic perineural cysts (Tarlov cysts) in one family: A case report. J Korean Neurosurg Soc 44(3):174–177, 2008. 41. Paulsen RD, Call GA, Murtagh FR: Prevalence and percutaneous drainage of cysts of the sacral nerve root sheath (Tarlov cysts). AJNR Am J Neuroradiol 15:293–299, 1994. 42. Prevo RL, Hageman G, Bruyn RP, et al: Extended extradural spinal arachnoid cyst: An unusual cause of progressive spastic paraparesis. Clin Neurol Neurosurg 101:260–263, 1999.

893

43. Rabb C, McComb J, Raffel C, et al: Spinal arachnoid cysts in the pediatric age group: An association with neural tube defects. J Neurosurg 77:369–372, 1992. 44. Rimmelin A, Clouet PL, Salatino S, et al: Imaging of thoracic and lumbar spinal extradural arachnoid cysts: Report of two cases. Neuroradiology 19:203–206, 1997. 45. Rohrer DC, Burchiel KJ, Gruber DP: Intraspinal extradural meningeal cyst demonstrating ball-valve mechanism of formation. Case report.   J Neurosurg 78:122–125, 1993. 46. Samii M, Klekamp J: Surgical results of 100 intramedullary tumors in relation to accompanying syringomyelia. Neurosurgery 35(5):865–873, 1994. 47. Schurch B, Wichmann W, Rossier AB: Post-traumatic syringomyelia (cystic myelopathy): A prospective study of 449 patients with spinal cord injury. J Neurol Neurosurg Psychiatry 60:61–67, 1996. 48. Schwartz ED, Falcone SF, Quencer RM, et al: Posttraumatic syringomyelia: Pathogenesis, imaging, and treatment. Am J Roentgenol 173:487–492, 1999. 49. Silbergleit R, Brunberg J, Patel S, et al: Imaging of spinal intradural arachnoid cysts: MRI, myelography, and CT. Neuroradiology 40:664–668, 1998. 50. Sklar E, Quencer R, Green B, et al: Acquired spinal subarachnoid cysts. AJNR 10:1097–1104, 1989. 51. Takahashi H, Taniguchi M, Ota T, et al: Congenital extradural cyst causing 30-year history of myelopathy with long-term remission.   No Shinkei Geka 21:443–447, 1993. 52. Tarlov IM: Perineurial cysts of the spinal nerve roots. Arch Neurol Psychiatr 40:1067–1074, 1938. 53. Tarlov IM: Spinal perineurial and meningeal cysts. J Neurol Neurosurg Psychiatry 33:833–843, 1970. 54. Wang D, Bodley R, Sett P, et al: A clinical magnetic resonance imaging study of the traumatised spinal cord more than 20 years following injury. Paraplegia 34:65–81, 1996. 55. Wang MY, Levi AD, Green BA: Intradural spinal arachnoid cysts in adults. Surg Neurol 60:49–56, 2003. 56. Williams B: Pathogenesis of post-traumatic syringomyelia. Br J Neurosurg 6:517–520, 1992. 57. Yasui K, Hashizume Y, Yoshida M, et al: Age-related morphologic changes of the central canal of the human spinal cord. Acta Neuropathol 97:253–259, 1999.

http://radiologyebook.com

33  Spinal Vascular Diseases Young-Jun Lee and Timo Krings

INTRODUCTION Spinal vascular diseases consist of spinal vascular malformations, spinal aneurysms, and spinal cord arterial ischemia. They are relatively rare but should be kept in the differential diagnosis for spinal tumors, demyelinations, or compressive myelopathies. As for spinal vascular malformations, there have been various nomenclatures or classifi­ cations. Spetzler proposed a classification based on anatomic and pathophysiologic characteristics, and he subdifferentiated extradural arteriovenous fistulae (AVF), intradural dorsal/ventral AVFs, extradural-  intradural arteriovenous malformations (AVMs), intramedullary AVMs, and conus medullaris AVMs.24 The Bicetre group proposed a classification based on genetic and hereditary points of view that com­ prised three main groups: genetic hereditary lesions, genetic nonhe­ reditary lesions, and single lesions reflecting an incomplete expression of a genetic condition.27 The majority of spinal vascular malformations fall into this third group and can be subclassified, depending on their feeding artery, into dural AVFs, spinal cord pial AVMs (nidus or fistu­ lous), and paravertebral AVFs. Clinical manifestations may be due to hemorrhage, space-occupying effects, venous congestion, and (rarely) arterial steal.15,19,32

SPINAL VASCULAR MALFORMATIONS Spinal Dural Arteriovenous Fistulae (SDAVFs) SDAVFs refer to a shunting lesion between dural branches of the seg­ mental artery and a radicular vein. They are the most common form (70%) of spinal vascular malformation.23,28 SDAVFs are presumed to be acquired lesions, similar to their cranial counterpart, the cranial dural AVF. SDAVFs are most frequently found in the thoracolumbar region of elderly men.20 Most SDAVFs develop in the lateral epidural space near the dural root sleeve. Venous drainage is directed retrogradely into perimedullary veins, which is responsible for the following hemody­ namic sequence: spinal venous pressure increase due to arterialization, spinal cord congestion, reduced tissue perfusion, and finally cord isch­ emia.22 Patients with SDAVFs usually present with progressive myelop­ athies, and additional autonomic symptoms (e.g., bladder, bowel, and erectile dysfunction) are seen in the later course of the disease.48 On computed tomography (CT), spinal cord swelling and dilated perimedullary venous enhancement may be noted; however, CT is not a comprehensive tool for the diagnosis of spinal vascular diseases. Magnetic resonance imaging (MRI) is the imaging method of choice because it can demonstrate the congestive edema of the spinal cord, breakdown of the blood–spinal cord barrier due to chronic venous ischemia, and the dilated perimedullary vessels. Thus on MRI the typical imaging findings are an enlargement of the spinal cord with central confluent T2 hyperintensity over multiple segments, patchy

894

contrast enhancement over multiple segments,11,18 and dilated peri­ medullary veins on the dorsal surface of the spinal cord, with multiple serpentine flow voids on T2 that enhance on postcontrast T1. Firstpass contrast-enhanced MR angiography may show the level of fistula and feeders, which is very useful guidance in catheter angiography.7,17 Most importantly, in equivocal cases this technique can identify with fairly high sensitivity and specificity the presence of a shunt due to early venous filling. Selective spinal catheter angiography is still the gold standard for evaluation of SDAVFs regarding confirmative diag­ nosis, identification of the exact level, precise ascertainment of the location of the anterior and posterior spinal arteries, and feasibility for embolization. Venous congestion can be demonstrated on angiogram as a delayed washout of contrast from the anterior spinal artery. MRI differential diagnosis includes other spinal vascular malformations (perimedullary and filum terminale AVFs may mimic dural AVFs), CSF pulsation artifact, hypervascularized spinal cord tumors (hemangio­ blastomas or paragangliomas), and serpentine nerve roots and venous engorgement related to spinal stenosis. Both endovascular embolization and open surgery are available for treating these lesions, with the aim being to obliterate the most distal part of the feeding artery and the proximal draining vein. In our insti­ tution, embolization is usually attempted before surgery, because the confirmatory diagnosis requires spinal angiography. Liquid embolic agents, mostly glue, are used as embolizing agents, and penetration of glue into the proximal vein is necessary for complete occlusion; oth­ erwise (i.e., in ligation types of embolization) the fistula will recur.28 In most cases, myelopathic symptoms stabilize after successful treat­ ment. However, despite the resolution of cord edema, neurologic improvement is not always complete, especially when the treatment was delayed, because treatment outcome depends on symptom dura­ tion. Autonomic dysfunctions rarely improve, even after successful treatment9 (Figs. 33-1 and 33-2).

Spinal Cord Arteriovenous Malformations (SCAVMs) SCAVMs are intradural shunting diseases comprising about 20% to 25% of all spinal vascular malformations.14,29 They can be subdivided into spinal cord (pial) perimedullary or intramedullary, nerve root, and filum terminale lesions according to their anatomic location and resultant feeding artery. The gross angioarchitecture at their transition from artery into vein can be either of a nidus or fistulous type.44,45 The AVMs are fed by radiculomedullary arteries or radiculopial arteries and drained by intramedullary spinal veins, perimedullary veins, and/or radicular veins.

Spinal Cord (Pial) Arteriovenous Malformations Perimedullary arteriovenous fistulae (PMAVFs).  PMAVFs are direct arteriovenous shunts without intervening nidus. Like their

http://radiologyebook.com

CHAPTER 33  Spinal Vascular Diseases

895

FIG 33-1  Spinal dural AVF in a 56-year-old man. Spinal cord edema with breakdown of blood–spinal cord barrier is demonstrated as hyperintensity on T2 (A) and patchy enhancement on postcontrast T1 (B). Dilated perimedullary veins on the dorsal surface of spinal cord can be seen as multiple serpentine flow voids (arrows on A and B) on T2 (A) and enhancements on postcontrast T1 (B).

A

B

D

A

B

C

E

FIG 33-2  Spinal dural AVF in a 53-year-old man. Contrast-enhanced spinal MR angiograms (A, B) can show the level and feeding artery of the fistula and thereby direct the subsequent catheter angiogram (C, D). The arrows mark the transition between the artery and the veins (i.e., the fistulous zone). After embolization with glue, the shunt was completely obliterated (E).

http://radiologyebook.com

896

PART II  CT and MR Imaging of the Whole Body

cranial counterparts, pial AVFs of the brain, these shunts are located superficially on the spinal cord and only rarely possess intramedullary compartments.25 The anterior or posterior spinal arteries are feeders, and venous drainage is through perimedullary and/or spinal veins. They can clinically present with either acute hemorrhage or progressive myelopathy.45,48 PMAVFs can be further categorized into micro- or macrofistulae according to the size of the feeding vessels, shunt volume, and size of the draining veins. Macrofistulae are typically encountered in children with a history of hereditary hemorrhagic telangiectasia. The majority of PMAVFs are microfistulae. If drainage is directed into the ventral spinal veins, these microfistulae may only present with cord edema because the ventral spinal veins are located subpially in the depth of the central sulcus of the cord and may therefore not be seen on MRI as dilated vessels. The posterior surface veins, on the other hand, are located in the subarachnoid space and are therefore more often seen when arterialized. Concerning treatment, microsurgery may be pre­ ferred if the feeding artery is too small for microcatheter navigation, whereas embolization can be chosen for cases feasible for endovascular access or having higher surgical risks due to extreme shunt flow with venous ectasias. With hemorrhagic presentation, hematoma in the subarachnoid space or spinal cord can be demonstrated on both CT and MRI. When there is spinal cord congestion, T2 hyperintensity with diffuse contrast enhancement in the central cord can be noted on MRI. Dilated drain­ ing veins can be seen as flow voids on T2. Contrast-enhanced MR angiography is useful for determination of the fistulous site and pre­ sumed feeders (Figs. 33-3 and 33-4). Intramedullary arteriovenous malformations (IMAVMs).  IMAVMs refer to nidus-type AVMs and may be considered counterparts of pial

A

brain AVMs. The nidus is usually located in the spinal cord, but a superficial compartment can also be exposed to the subarachnoid space. Clinical presentations are either due to acute hemorrhage or progressive myelopathy. In hemorrhagic presentation, the neurologic sequelae largely depend on the location of hemorrhage; intramedullary hematoma tends to cause more severe neurologic deficits than sub­ arachnoid hemorrhage. The nidus can be noted as a conglomeration of flow voids on T2, with mixed signal intensities with variable contrast enhancement on T1. Feeding arteries are frequently multiple because there are many anastomoses between the anterior and posterior intrinsic arterial system of the spine.25 Contrast-enhanced MR angiography is useful for identification of feeding artery and nidus, though catheter angiogra­ phy is mandatory to ascertain the feasibility of endovascular treatment. Draining veins are the intrinsic spinal cord veins and perimedullary veins, which may be demonstrated as serpentine flow voids on T2. Spinal cord congestion can be noted in patients with progressive myelopathic presentation as central T2 hyperintensity and patchy con­ trast enhancement. Embolization may sometimes be impossible or incomplete because of multiple feeders and difficulty in avoiding anterior spinal arterial compromise. Targeted embolization of the vulnerable lesions respon­ sible for hemorrhage or the major shunt responsible for cord conges­ tion may be effective in acutely symptomatic patients. Microsurgical resection with or without prior embolization may be another option for treatment (Fig. 33-5).

Filum Terminale Arteriovenous Fistulae (FTAVFs).  FTAVFs refer to a rare type of shunt between the artery of the filum terminale and the spinal vein. Anatomically the filum terminale is the extension of

B

FIG 33-3  PMAVF in a 33-year-old man. A, Acute subarachnoid hemorrhage with T2 dark signal intensity is seen, (arrow in A) as well as serpentine flow voids representing dilated draining veins on T2. B, Angiogram shows microfistula between tortuous anterior spinal artery and vein.

http://radiologyebook.com

CHAPTER 33  Spinal Vascular Diseases

T11

A

B

FIG 33-4  PMAVF in a 47-year-old man. A, Hyperintense spinal cord on T2 is suggestive of congestive edema (arrow). B, Angiogram shows macrofistula between the posterior spinal artery and vein.

A

B

C

D

FIG 33-5  IMAVM in a 20-year-old woman. Intramedullary nidus is noted as conglomeration of small flow voids with adjacent edema on T2 (A) and enhancement on postcontrast T1 (B). Feeder (straight arrow), nidus (arrowhead), and upwardly draining vein (curved arrow) are well demonstrated on contrast-enhanced MR angiogram (C) and verified by catheter angiograms (D).

http://radiologyebook.com

897

898

PART II  CT and MR Imaging of the Whole Body

the pia mater from the conus medullaris; therefore FTAVF can be regarded as an analogous lesion to PMAVF. The artery of the filum terminale is the continuation of the anterior spinal artery below the level of the cone and therefore does not supply the cord. The shunts are located in the lumbosacral region. They usually present clinically with progressive myelopathy or conus medullaris syndrome as a result of spinal cord congestion.21,31,33 In addition to supply from the artery of the filum terminale via the anterior spinal artery, additional feeders can also originate from the internal iliac arteries.21,31 Associations with lumbosacral congenital dysraphic malformations have been reported.10 Endovascular treatment is often difficult owing to the long and tortu­ ous course of the anterior spinal artery and the artery of filum termi­ nale; microsurgery is preferred in most cases (Fig. 33-6).

Radicular Arteriovenous Fistulae (RAVFs) RAVFs refer to the shunt between the radicular artery and radicular vein at the level of the intradural nerve roots.3 When a patient has an intramedullary AVM as well as RAVF, a metameric disease can be sug­ gested.39 The MR and angiographic findings of RAVF may resemble those of dural AVF. RAVFs in the cauda equina may be difficult to dif­ ferentiate from FTAVFs,41 whereas other radicular fistulae may mimic dural AV shunts. The major difference is that radicular shunts are present after the artery has pierced the dura, but dural AVFs are present at the level of the dura. Clinical manifestations and treatment are in keeping with other forms of spinal cord vascular malformations; however, radicular symptoms are often the first symptom in these types of lesions and are rarely seen in other vascular shunts. When the

A

anterior spinal artery arises from the same radicular feeder that  supplies the AVM, microsurgical treatment tends to be favored41 (Fig. 33-7).

Paravertebral Arteriovenous Fistulae (PVAVFs) PVAVFs (epidural AVFs) are relatively rare lesions and occur outside the dura mater between the radicular artery and epidural venous plexus. They typically present with benign clinical symptoms such as radiculopathy from a compressive mass effect on nerve roots; only very rarely does congestive myelopathy due to perimedullary reflux occur.26,42 The lesions are found more often in the lumbar region of elderly men.26 On T2 imaging, dilated epidural veins compressing the cord can be noted, with multiple flow voids in the paravertebral region. The spinal cord appears hyperintense, with dark serpentine perimedul­ lary veins when intradural venous reflux exists. Angiographic findings can be similar to dural AVF except for the presence of intradural reflux via the epidural venous plexus far from the primary fistula site in the epidural space.13 Endovascular treatment can be performed transarte­ rially, but transvenous embolization may be more effective when there are multiple feeders (Fig. 33-8).

Metameric Arteriovenous Malformations Metameric AVMs refer to multiple vascular malformations that affect more than one tissue derived from the same spinal metameric segment.38 They can be classified under the category of “genetic non­ hereditary” according to the Bicetre group. Metameric AVMs comprise less than 10% of all spinal cord AVMs and can occur in any spinal

B

C

FIG 33-6  FTAVF in a 47-year-old woman. A, On T2 a low-lying conus suggesting tethered cord is noted, as well as spinal cord congestion (arrow in A). Shunt is dually fed by anterior spinal artery (arrow in B) and lateral sacral artery (B), which is clearly demonstrated on contrast-enhanced MR angiogram (C). The arrow in C points to the anterior spinal artery.

http://radiologyebook.com

CHAPTER 33  Spinal Vascular Diseases

A

B

C

D

FIG 33-7  RAVF in a 64-year-old woman. A fistula along the fourth cervical nerve root can be noted on T2 (A) and contrast-enhanced MR angiogram (B). The fistula between the intradural radicular artery (straight arrow) and a radicular vein with drainage to the anterior spinal vein (arrowhead) is seen on both catheter angiogram and contrast-enhanced MR angiogram (C, D).

A

C

B

FIG 33-8  Paravertebral AVF in a 34-year-old woman. Ectatic veins compressing the dural sac and nerve roots are noted as flow voids on T2-weighted images (A, B) in the epidural space and paraspinal space (arrows in A and B). C, Angiogram from right thyrocervical trunk (arrow in C) shows significant AV shunting and ectatic epidural veins (arrowheads).

http://radiologyebook.com

899

900

PART II  CT and MR Imaging of the Whole Body

segment.4 Female predominance and earlier presentation than other spinal vascular malformations were reported in one series.38 Clinical presentation is manifold, including subarachnoid hemor­ rhage or hematomyelia, nonhemorrhagic presentations including pain or progressive myelopathy, acute thrombosis of a draining vein, venous hypertension, or mass effect from dilated venous structures. CT or MRI may show the extent of disease as well as the abnormality of the spinal cord. Angiographically the shunts are most commonly of the nidus type. Because of the complexity of the lesions, complete oblitera­ tion is rarely possible. Therefore treatments are mostly palliative, focused on preventing hemorrhage, preserving spinal cord function, and relieving pain. Although the long-term functional prognosis is poor, the neurologic deteriorations can be delayed and improved by performing targeted palliative embolization16,38 (Fig. 33-9).

lesion gets symptomatic, because it usually has a progressive course1 (Fig. 33-10).

SPINAL ANEURYSMS Aneurysms arising within the spinal canal are very rare. Most of them are thought to be associated with conditions that may increase blood flow in the spinal arteries, such as spinal AVMs, coarctation of the aorta, or renal transplantation.2,5,6,8 Therefore solitary spinal aneu­ rysms are even rarer, and postulated pathogenesis includes inflamma­ tory processes, dissection, and intimal defects.46 They usually present with hemorrhage but can sometimes manifest as a compressive lesion by mass effect. They are most commonly found intradurally at the cervical levels, arising from the anterior spinal artery.34 CT or MRI can show the hemorrhage, coexisting vascular malfor­ mations, or the aneurysmal sac itself. Catheter spinal angiogram is necessary to confirm diagnosis and plan treatment. Owing to their presumed dissecting nature, isolated spinal aneurysms tend to have a fusiform shape and do not occur at branching sites. No standard therapy has been established for ruptured solitary spinal aneurysm, and parent vessel sacrifice is commonly believed to be the most effec­ tive way to prevent rebleeding (Fig. 33-11).30

Spinal Cord Cavernomas Spinal cord cavernomas are nonshunting vascular malformations that are analogous to the ones in the brain. They consist of enlarged thinwalled capillaries and sinusoids without intervening parenchyma. They are located intramedullary and have no identifiable feeding artery or draining vein. They comprise about 5% of all spinal vascular mal­ formations and are usually found in the thoracic region in middle-aged women.47,49,50 Intramedullary hematoma with repeated episodes can cause acute, recurrent, and progressive myelopathy, and continued oozing of blood from the cavernoma can present with intramedullary edema.50 Calcifications are rarely detected on CT and only present in larger lesions. On MRI, cavernomas appear as well-defined circumscribed lesions with a T2 hypointense rim and central mixed signal intensity, representing hemorrhages in different stages of evolution. Although they are angiographically silent, spinal angiography is sometimes needed for acutely hemorrhaged cavernomas to rule out other spinal vascular malformations. Microsurgical resection is indicated when the

A

B

SPINAL CORD ARTERIAL INFARCT Spinal cord arterial infarct comprises 5% to 8% of all acute myelopa­ thies.37 It is mostly a single event, and the main etiologies are interven­ tions and pathologies affecting the aorta.12,36,43 It needs to be differentiated from other causes of acute noncompressive myelopa­ thies such as acute transverse myelitis or multiple sclerosis. The infarcts are most commonly found as an isolated pencil-like area of T2 hyper­ intensity at the centromedullary territory of the anterior spinal artery in the thoracolumbar region, often encompassing more than two

C

D

FIG 33-9  Spinal metameric AVM in a 27-year-old woman. Vascular malformation involving both intradural and extradural spaces is noted in the cervicothoracic region. Intramedullary hematoma is noted as dark signal intensity on T2-weighted image (arrow in A). Dilated perimedullary veins and epidural venous plexi are shown as flow voids on T2-weighted image (arrowheads in A) and enhancement on postcontrast T1-weighted image (arrowheads in B). Catheter angiograms show nidus-type spinal cord AVM (arrow in C) and epidural AVF (arrow in D).

http://radiologyebook.com

CHAPTER 33  Spinal Vascular Diseases

A

901

B

FIG 33-10  Spinal cavernoma in a 47-year-old woman. T1- (A) and T2-weighted images (B) show a welldefined circumscribed lesion with hypointense rim and central mixed signal intensity in the cervical cord (arrow in A and B).

A

B

C

D

FIG 33-11  Dissecting spinal aneurysm in a 72-year-old man. Subarachnoid hemorrhages are noted as hyperintensities on T1-weighted sagittal image (arrow in A). An oval-shaped dark intradural lesion on T2-weighted images (arrows in B, C) is confirmed to be a dissecting aneurysm from the radiculopial artery on catheter angiogram (arrow in D).

vertebral segments.40 Enhancement is seen in the affected area after 4 to 5 days and may persist for 3 to 4 weeks. Additional cauda equina enhancement can be observed. Diffusion-weighted imaging is useful for earlier detection of spinal cord arterial infarct, though there are several technical difficulties such as motion artifacts, susceptibility arti­ facts, and low signal-to-noise ratio.40 The clinical spectrum is broad

and will depend on the local collaterals, location of the ischemia, and the involved vessel. Clinical symptoms develop within the first 24 hours after symptom onset to their full clinical picture. Radicular pain is often the first clinical symptom. The functional long-term outcome depends mainly on the severity of initial neurologic deficits, especially of motor deficits35 (Fig. 33-12).

http://radiologyebook.com

902

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 33-12  Spinal cord arterial infarct in a 36-year-old man. T2-weighted sagittal (A) and axial (B) images show a relatively well-defined hyperintense lesion in the centromedial territory of the anterior spinal artery along the anterior column of the thoracic cord (arrows in A and B). Postcontrast T1 sagittal image (C) shows enhancement (arrow in C).

REFERENCES 1. Anson JA, Spetzler RF: Surgical resection of intramedullary spinal cord cavernous malformations. J Neurosurg 78:446–451, 1993. 2. Aoun SG, Soltanolkotabi M, Ansari SA, et al: Ruptured spinal artery aneurysm associated with coarctation of the aorta: Case report and literature review. World Neurosurg 2012. 3. Berenstein ALP, Terbrugge K: Classification of spinal cord arteriovenous malformations. In Surgical neuroangiography, ed 2, New York, 2004, Springer, pp 767–774. 4. Berenstein ALP, Terbrugge K: Spinal cord arteriovenous malformations. In Surgical neuroangiography, ed 2, Berlin, Germany, 2004, Spring-Verlag, pp 737–847. 5. Berlis A, Scheufler KM, Schmahl C, et al: Solitary spinal artery aneurysms as a rare source of spinal subarachnoid hemorrhage: Potential etiology and treatment strategy. AJNR Am J Neuroradiol 26:405–410, 2005. 6. Biondi A, Merland JJ, Hodes JE, et al: Aneurysms of spinal arteries associated with intramedullary arteriovenous malformations. I. Angiographic and clinical aspects. AJNR Am J Neuroradiol 13:913–922, 1992. 7. Bowen BC, Fraser K, Kochan JP, et al: Spinal dural arteriovenous fistulas: Evaluation with MR angiography. AJNR Am J Neuroradiol 16:2029–2043, 1995. 8. Cai X, Liu C, Hu K, et al: A case report of posterior spinal artery aneurysm. Chin Med J 111:1051–1053, 1998. 9. Cenzato M, Versari P, Righi C, et al: Spinal dural arteriovenous fistulae: Analysis of outcome in relation to pretreatment indicators. Neurosurgery 55:815–822, discussion 822–813, 2004. 10. Chanthanaphak E, Pongpech S, Jiarakongmun P, et al: Filum terminale arteriovenous fistulas: The role of endovascular treatment. J Neurosurg Spine 19:49–56, 2013. 11. Chen CJ, Chen CM, Lin TK: Enhanced cervical MRI in identifying intracranial dural arteriovenous fistulae with spinal perimedullary venous drainage. Neuroradiology 40:393–397, 1998.

12. Cheshire WP, Santos CC, Massey EW, et al: Spinal cord infarction: Etiology and outcome. Neurology 47:321–330, 1996. 13. Chul Suh D, Gon Choi C, Bo Sung K, et al: Spinal osseous epidural arteriovenous fistula with multiple small arterial feeders converging to a round fistular nidus as a target of venous approach. AJNR Am J Neuroradiol 25:69–73, 2004. 14. da Costa L, Dehdashti AR, terBrugge KG. Spinal cord vascular shunts: Spinal cord vascular malformations and dural arteriovenous fistulas. Neurosurg Focus 26:E6, 2009. 15. Detweiler PW, Porter RW, Spetzler RF: Spinal arteriovenous malformations. Neurosurg Clin N Am 10:89–100, 1999. 16. Ducruet AF, Crowley RW, McDougall CG, et al: Endovascular management of spinal arteriovenous malformations. J Neurointerv Surg 5:605–611, 2013. 17. Farb RI, Kim JK, Willinsky RA, et al: Spinal dural arteriovenous fistula localization with a technique of first-pass gadolinium-enhanced MR angiography: Initial experience. Radiology 222:843–850, 2002. 18. Gilbertson JR, Miller GM, Goldman MS, et al: Spinal dural arteriovenous fistulas: MR and myelographic findings. AJNR Am J Neuroradiol 16:2049–2057, 1995. 19. Grote EH, Voigt K: Clinical syndromes, natural history, and pathophysiology of vascular lesions of the spinal cord. Neurosurg Clin N Am 10:17–45, 1999. 20. Jellema K, Tijssen CC, van Gijn J: Spinal dural arteriovenous fistulas:   A congestive myelopathy that initially mimics a peripheral nerve disorder. Brain 129:3150–3164, 2006. 21. Jin YJ, Kim KJ, Kwon OK, et al: Perimedullary arteriovenous fistula of the filum terminale: Case report. Neurosurgery 66:E219–E220, discussion E220, 2010. 22. Kataoka H, Miyamoto S, Nagata I, et al: Venous congestion is a   major cause of neurological deterioration in spinal arteriovenous malformations. Neurosurgery 48:1224–1229, discussion 1229–1230, 2001. 23. Kendall BE, Logue V: Spinal epidural angiomatous malformations draining into intrathecal veins. Neuroradiology 13:181–189, 1977.

http://radiologyebook.com

CHAPTER 33  Spinal Vascular Diseases 24. Kim LJ, Spetzler RF: Classification and surgical management of spinal arteriovenous lesions: Arteriovenous fistulae and arteriovenous malformations. Neurosurgery 59:S195–S201, discussion S193–113, 2006. 25. Lee YJ, Terbrugge KG, Saliou G, et al: Clinical features and outcomes of spinal cord arteriovenous malformations: comparison between nidus and fistulous types. Stroke 45(9):2606–2612, 2014. 26. Kiyosue H, Tanoue S, Okahara M, et al: Spinal ventral epidural arteriovenous fistulas of the lumbar spine: Angioarchitecture and endovascular treatment. Neuroradiology 55:327–336, 2013. 27. Krings T: Vascular malformations of the spine and spinal cord: anatomy, classification, treatment. Clin Neuroradiol 20:5–24, 2010. 28. Krings T, Geibprasert S: Spinal dural arteriovenous fistulas. AJNR Am J Neuroradiol 30:639–648, 2009. 29. Krings T, Mull M, Gilsbach JM, et al: Spinal vascular malformations.   Eur Radiol 15:267–278, 2005. 30. Lavoie P, Raymond J, Roy D, et al: Selective treatment of an anterior spinal artery aneurysm with endosaccular coil therapy. Case report.   J Neurosurg Spine 6:460–464, 2007. 31. Lim SM, Choi IS, David CA: Spinal arteriovenous fistulas of the filum terminale. AJNR Am J Neuroradiol 32:1846–1850, 2011. 32. Lv X, Li Y, Yang X, et al: Endovascular embolization for symptomatic perimedullary AVF and intramedullary AVM: A series and a literature review. Neuroradiology 54:349–359, 2012. 33. Macht S, Chapot R, Bieniek F, et al: Unique sacral location of an arteriovenous fistula of the filum terminale associated with diastematomyelia and lowered spinal cords. Neuroradiology 54:517–519, 2012. 34. Madhugiri VS, Ambekar S, Roopesh Kumar VR, et al: Spinal aneurysms: Clinicoradiological features and management paradigms. J Neurosurg Spine 19:34–48, 2013. 35. Masson C, Pruvo JP, Meder JF, et al: Spinal cord infarction: Clinical and magnetic resonance imaging findings and short term outcome. J Neurol Neurosurg Psychiatry 75:1431–1435, 2004. 36. Messe SR, Bavaria JE, Mullen M, et al: Neurologic outcomes from high risk descending thoracic and thoracoabdominal aortic operations in the era of endovascular repair. Neurocrit Care 9:344–351, 2008. 37. Nedeltchev K, Loher TJ, Stepper F, et al: Long-term outcome of acute spinal cord ischemia syndrome. Stroke 35:560–565, 2004. 38. Niimi Y, Uchiyama N, Elijovich L, et al: Spinal arteriovenous metameric syndrome: Clinical manifestations and endovascular management. AJNR Am J Neuroradiol 34:457–463, 2013.

903

39. Nishio A, Ohata K, Takami T, et al: Spinal arteriovenous malformation associated with a radicular arteriovenous fistula suggested a metameric disease. A case report. Interv Neuroradiol 9:75–78, 2003. 40. Nogueira RG, Ferreira R, Grant PE, et al: Restricted diffusion in spinal cord infarction demonstrated by magnetic resonance line scan diffusion imaging. Stroke 43:532–535, 2012. 41. Ohtonari T, Ota S, Nishihara N, et al: Arteriovenous fistula in a nerve root of the cauda equina fed by a proximal radiculo-medullary artery:   a report of two cases. Interv Neuroradiol 17:217–223, 2011. 42. Ramanathan D, Levitt MR, Sekhar LN, et al: Management of spinal epidural arteriovenous fistulas: Interventional techniques and results.   J Neurointerv Surg 2013. 43. Restrepo L, Guttin JF: Acute spinal cord ischemia during aortography treated with intravenous thrombolytic therapy. Tex Heart Inst J 33:74–77, 2006. 44. Rodesch G, Hurth M, Alvarez H, et al: Classification of spinal cord arteriovenous shunts: Proposal for a reappraisal—The Bicetre experience with 155 consecutive patients treated between 1981 and 1999. Neurosurgery 51:374–379, discussion 379–380, 2002. 45. Rosenblum B, Oldfield EH, Doppman JL, et al: Spinal arteriovenous malformations: A comparison of dural arteriovenous fistulas and intradural AVM’s in 81 patients. J Neurosurg 67:795–802, 1987. 46. Son S, Lee SG, Park CW: Solitary ruptured aneurysm of the spinal artery of Adamkiewicz with subarachnoid hemorrhage. J Korean Neurosurg Soc 54:50–53, 2013. 47. Spetzger U, Gilsbach JM, Bertalanffy H: Cavernous angiomas of the spinal cord clinical presentation, surgical strategy, and postoperative results. Acta Neurochir (Wien) 134:200–206, 1995. 48. Van Dijk JM, TerBrugge KG, Willinsky RA, et al: Multidisciplinary management of spinal dural arteriovenous fistulas: Clinical presentation and long-term follow-up in 49 patients. Stroke 33:1578–1583, 2002. 49. Weinzierl MR, Krings T, Korinth MC, et al: MRI and intraoperative findings in cavernous haemangiomas of the spinal cord. Neuroradiology 46:65–71, 2004. 50. Zevgaridis D, Medele RJ, Hamburger C, et al: Cavernous haemangiomas of the spinal cord. A review of 117 cases. Acta Neurochir (Wien) 141:237–245, 1999.

http://radiologyebook.com

34  Systemic Diseases Affecting the Spine Romyll Garcia and Steven Falcone

A variety of systemic diseases affect the osseous skeleton and therefore the spine. Detection of bone abnormalities with imaging can be challenging with plain film, given that demineralization, a common element of many conditions, requires a 50% decrease in bone mass before it can be detected. Computed tomography (CT) has improved the detection of demineralization, and magnetic resonance imaging (MRI) can enhance the sensitivity and specificity of detecting marrow abnormalities. Bone scan is a tool used to detect osteoblastic activity, a feature typically representative of the reparative process, although a primary lesion might have intrinsic elevation of osteoblastic activity. Metabolic bone disease has a variety of causes and represents, as the name implies, conditions that result from some alteration of bone metabolism. In the normal adult there is a balance of osteoclastic and osteoblastic activity, resulting in neither a net decrease nor increase in the amount of bone. Thus an alteration in this balance results in abnormal bone resorption or bone deposition. In this chapter we will focus on some systemic diseases that have a characteristic appearance when they affect the spine. We will discuss Paget’s disease of bone (PDB), renal osteodystrophy, myelofibrosis, and extramedullary hematopoiesis (EMH).

PAGET’S DISEASE OF BONE PDB (osteitis deformans) was originally described by Sir James Paget in 1877. PDB primarily affects the skull, spine, pelvis, and femur. The pelvis and spine are the most commonly affected sites, involved in 70% and 55% of cases, respectively.4 PDB is polyostotic in 66% of cases.4 PDB has osteolytic and osteoblastic phases, with the lytic phase predominating early on in the course of the disease process. It is a chronic disorder characterized by excessive and abnormal remodeling of bone often leading to osseous deformities. The etiology is unclear but likely multifactorial. There is epidemiologic data to support autosomal dominant trait.10 There are also published data suggesting a possible causative role by paramyxoviral infection.10 Paget’s disease is a relatively common disease affecting the aging population, with affected individuals usually older than age 55 years at diagnosis. Patients younger than age 40 are rarely affected. The prevalence of PDB varies greatly between different countries; it is extremely rare in Japan and has approximately a 5.4% prevalence in the United Kingdom.3 Imaging findings typically include enlarged vertebral bodies with trabecular thickening. There is usually involvement of the anterior and posterior elements. Characteristic vertebral body findings have been described as “picture frame” appearance secondary to central osteopenia and a peripheral coarse and thickened trabecular pattern (Figs. 34-1 through 34-4). Findings may also include increased size and diffuse sclerosis of the vertebral body, referred to as “ivory” vertebra.

904

Vertebral enlargement could lead to neural foraminal and/or central canal stenosis with associated neurologic symptoms (see Fig. 34-3). Contrast-enhanced CT or MRI may demonstrate significant marrow enhancement in the active or lytic phase. MRI usually demonstrates associated low T1 and T2 signal intensity, correlating to sclerotic bone. Areas of active lysis demonstrate increased T2 signal intensity (see Fig. 34-2). There is preservation of the normal marrow in unaffected areas. Bone scan may aid with polyostotic disease. “Mickey mouse” is a term that has been classically used to describe the inverted triangular radiotracer uptake classically noted in the involved vertebral bodies and pedicles. There is usually increased radiotracer uptake during the active phase. Alkaline phosphatase, which reflects the level of new bone formation, is used as an adjunct in diagnosis and also plays a role in evaluating treatment response.2 However, it is nonspecific and values are normal in up to 15% of symptomatic patients with monostotic disease. Disease activity and response to therapy can be followed with measurements of alkaline phosphatase and other serum markers. Basilar invagination, kyphoscoliosis, spondylosis, and fractures are the result of bone softening and secondary complications of Paget’s disease of the spine.14,20,21 Sarcomatous degeneration involving the spine is rare, seen in less than 1% of cases and most commonly in patients with extensive disease.9 If there is interval change in imaging characteristics, soft tissue component, or increasing pain, further cross-sectional contrast-enhanced imaging should be performed to evaluate for rare sarcoma formation. Osteosarcoma is the most common type of sarcoma associated with Paget’s disease, representing approximately 50% of cases; however, fibrosarcoma and chondrosarcoma are also described, with an incidence of 20% to 25% and 10%, respectively.6 Sarcomatous transformation is highly fatal, with a 3-year mortality rate greater than 90%.16 Differential diagnosis for Paget’s disease involving the spine includes metastatic disease and vertebral hemangioma. Although PDB has a distribution similar to metastatic disease, greater involvement of the ribs has been described with the former. There is also typically no enlargement of the affected osseous structures with metastatic disease. However, sometimes the distinction is difficult and biopsy is necessary. Vertebral hemangiomas have a delicate trabecular pattern, in contrast to the coarse and enlarged trabecular pattern seen in pagetoid bone. Additionally, typical vertebral hemangiomas demonstrate high T1 and T2 signal intensity on MRI, with no associated enlargement of the vertebral body. Treatment of PDB is mostly targeted at symptomatic relief with analgesic medication. Medications that inhibit osteoclastic activity (e.g., bisphosphonate) also play a role.13,22 Surgical intervention is sometimes needed in symptomatic cases of neural foraminal or central canal stenosis.

http://radiologyebook.com

CHAPTER 34  Systemic Diseases Affecting the Spine

905

B

A

FIG 34-1  PDB. Sagittal and axial CT of the cervical spine demonstrates enlarged C7 vertebra with involvement of the anterior and posterior elements. There is associated loss of height of the vertebra due to bone softening. Note “picture frame” appearance of the spinous process.

B

A

FIG 34-2  PDB (same patient as Fig. 34-1). T1- and T2-weighted images of the cervical spine during active phase demonstrates low T1 and high T2 signal intensity of the C7 vertebra. There is an MR “picture frame” appearance, most evident in the spinous process.

RENAL OSTEODYSTROPHY Renal osteodystrophy refers to the skeletal manifestations of end-stage renal disease. Pathophysiology is complex and involves vitamin D deficiency secondary to 1α-hydroxylase impairment. In addition, impaired kidneys fail to excrete phosphate. Excess phosphate binds calcium,

leading to hypocalcemia and resulting in elevation of parathyroid hormone levels. Pathogenesis is also multifactorial; side effects of chronic hemodialysis and steroid use also lead to typical imaging findings. There are nearly 1 million individuals currently undergoing renal dialysis worldwide.18 Long-term hemodialysis can lead to secondary amyloidosis, which in turn can lead to erosive spondyloarthropathy.

http://radiologyebook.com

906

PART II  CT and MR Imaging of the Whole Body

Chronic steroid use by most renal failure patients can lead to avascular necrosis, immunosuppression, and osteomyelitis. End-stage renal disease develops in 1 out of 10,000 individuals in the United States each year. There is an increased incidence in African Americans and Native Americans. Imaging findings affecting the spine include osteopenia, osteosclerosis, spondyloarthropathy, and brown tumors. Osteosclerosis has

FIG 34-3  PDB (same patient as Fig. 34-1). Axial T2-weighted gradient echo image demonstrates severe bilateral neural foraminal stenosis and central canal encroachment secondary to pagetoid changes.

A

been reported in 10% to 20% of patients with chronic renal disease. When osteosclerosis predominantly involves the end plates, the result is the classically described “rugger jersey” spine (Fig. 34-5). MRI demonstrates areas of low T1 and T2 signal intensity correlating to sclerosis (Fig. 34-6). Erosive spondyloarthropathy, usually secondary to amyloid deposition, leads to disk-space narrowing with erosion of vertebral end plates and subarticular cysts. The anterosuperior and anteroinferior corners are usually affected early in the disease by erosive changes. The craniocervical junction can be involved, although the lower cervical spine is most commonly affected in erosive spondyloarthropathy. Brown tumors affecting the spine are rare but have been described.12 Brown tumors are histologically benign lesions associated with hyperparathyroidism. “Brown tumor” is a misnomer in that it does not represent a neoplastic process but develops secondary to bone resorption, leading to varying amounts of hemorrhage, fibrous tissue, and necrosis. Characteristic imaging features for brown tumors include a geographic nonsclerotic and expansile lytic appearance.8 Their expansile nature can lead to neural foraminal and central canal stenosis, resulting in nerve and spinal cord impingement. MRI findings usually vary, and T1 and T2 signal intensity depend on varying degrees of fibrous tissue, hemorrhage, and necrosis within the brown tumors. Brown tumors can develop in patients with renal osteodystrophy, but they most commonly develop in the setting of primary hyperparathyroidism. Bone scan in renal osteodystrophy may demonstrate metabolic “superscan” with diffuse intense uptake throughout the axial and appendicular skeleton. It can also demonstrate focal uptake in areas of fractures. Differential diagnosis for erosive spondyloarthropathy in the setting of end-stage renal disease includes osteomyelitis/diskitis and seronegative spondyloarthropathies. There is usually no postcontrast enhancement associated with spondyloarthropathies, and paraspinal soft tissues are preserved, in contrast to osteomyelitis/diskitis. Clinical history and laboratory results can aid in differentiating between the different causes of spondyloarthropathy. Bandlike sclerosis affecting the vertebral end plates can be seen in the setting of osteopetrosis

B

FIG 34-4  PDB. Coronal and sagittal CT of the lumbar spine demonstrates multiple enlarged vertebrae with involvement of the anterior and posterior elements. There is a coarse and sclerotic peripheral trabecular pattern with relative central osteopenia involving L2. There is also involvement of the iliac bones.

http://radiologyebook.com

CHAPTER 34  Systemic Diseases Affecting the Spine

A

B

FIG 34-5  Renal osteodystrophy. Coronal and sagittal CT of the spine demonstrates marked sclerosis of the vertebral bodies adjacent to the end plates, creating a striped appearance like a rugby jersey.

A

B FIG 34-6  Renal osteodystrophy (same patient as Fig. 34-5). T1- and T2-weighted MRI of the spine demonstrate decreased signal intensity in areas of sclerosis.

http://radiologyebook.com

907

908

PART II  CT and MR Imaging of the Whole Body

(“sandwich vertebra”). In contrast to rugger-jersey spine, sclerotic bands in osteopetrosis are denser and more sharply defined. There is no treatment for renal osteodystrophy, but a renal transplant can halt progression of the disease by restoring calcium, phosphate, and parathyroid hormone levels. In addition, discontinuation of hemodialysis eliminates associated complications.

MYELOFIBROSIS Myelofibrosis represents a spectrum of bone marrow disorders that result in replacement of red marrow with fibrotic tissue, and represents an uncommon form of chronic leukemia. Primary myelofibrosis (PMF) has an incidence of approximately 0.4 to 1.4 per 100,000 individuals.17 The exact etiology of the disease is unknown, but it is believed to arise from a somatic mutation of a pluripotent hematopoietic progenitor cell. The end result is a neoplastic process leading to abnormal proliferation of fibroblasts and collagen deposition. Janus kinase 2 (JAK2) mutation is believed to play a role in primary and secondary myelofibrosis.19 Secondary myelofibrosis is a known associated complication of polycythemia vera and essential thrombocythemia. JAK2 mutation has been reported in 65% to 97% of patients with polycythemia vera. Mutation is also found in patients with essential thrombocythemia and PMF. Imaging findings include diffuse increased density of the osseous structures without associated enlargement of the affected bones (Fig. 34-7). Increased density secondary to fibrotic changes is characterized by low T1 and T2 signal on MRI. Opposed phase imaging does not show a drop in signal as typically noted in normal red marrow. In- and out-of-phase MR sequence does demonstrate a drop in signal in areas of reconverted marrow. This can be helpful in differentiating sites of marrow reconversion from neoplastic infiltration. Extramedullary

FIG 34-7  EMH. T1 pre- and postcontrast MR sequences demonstrate isointense paraspinal soft tissue mass in patient with thalassemia. Notice minimal to mild enhancement of the paraspinal mass.

hematopoiesis is a common associated finding in the paraspinal soft tissue. The most common signs and symptoms include easy bruising, anemia, weight loss, and hepatosplenomegaly. Primary acute myelofibrosis is a rapidly progressing disease with a high mortality rate. Leukemic conversion can be seen in up to 20% of affected individuals.5 Replacement of bone marrow can lead to deficiency and related clinical manifestations of all blood cell lineages. Infection and hemorrhage can be severe and are often the presenting clinical complaint. Differential diagnosis for diffuse osseous sclerosis includes malignant, metabolic, hematologic, and congenital disorders. Clinical history and lack of extramedullary hematopoiesis in most of the other etiologies is helpful in narrowing the differential. Bone marrow and stem cell transplantation appear to improve prognosis of the disease, with some patients demonstrating long-term remission. Mild cases can be monitored and blood counts managed with medication or transfusions. JAK2 inhibitors appear to be a promising developing treatment option.11 Clinical trials have demonstrated a significant decrease in constitutional symptoms and splenomegaly.

EXTRAMEDULLARY HEMATOPOIESIS Extramedullary hematopoiesis (EMH) refers to the production of blood cells outside of the bone marrow in response to severe chronic anemia. Almost all body sites can be involved, including the spleen, liver, and spine. Nonhepatosplenic EMH most commonly affects the spine, described in up to 27% of cases. Extramedullary hematopoiesis is most commonly secondary to beta thalassemia but can be seen in the setting of other hemoglobinopathies, myoproliferative disorders, and neoplasms.1 The pathophysiology of EMH is poorly understood. It is unclear whether pluripotential cells migrate to site of EMH or are always locally present and activated by anemic states. Typical radiographic imaging characteristics include paraspinal soft tissue masses, most commonly involving the thoracic spine. In some cases there might be associated expansion of the ribs and/or clavicles. Paraspinal soft tissue masses are isointense to the spinal cord in MRI and demonstrate some degree of enhancement, usually mild (Fig. 34-8). MRI can be useful in cases where there is neural foraminal or central canal stenosis. In cases of EMH secondary to myelofibrosis there will be diffuse bony sclerosis. Sulfur colloid scan can aid the diagnosis, because radiotracer will be taken up throughout affected sites.15 Differential diagnosis includes lymphoma, metastatic disease, hematoma, nerve sheath tumors, and abscesses. Lymphoma usually demonstrates intense enhancement, in contrast to the mild enhancement noted in most cases of EMH.7 Metastatic disease usually involves the vertebral bodies, and in most cases there is a known primary malignancy. It is also rare to have isolated paraspinal soft tissue involvement with sparing of the vertebral bodies in metastatic disease. Nerve sheath tumors usually demonstrate some degree of T2 hyperintensity and most commonly involve a single level, with the exception of neurofibromatosis patients. Epidural hematomas most commonly develop acutely in the setting of trauma, and signal intensity and enhancement characteristics vary depending on the stage of the blood products. Paraspinal abscesses most commonly cause destructive changes in adjacent bone and inflammatory changes in surrounding soft tissues. Treatment depends on the complications of the disease. In cases of spinal cord compression, steroids might help reduce inflammation and improve symptoms. Surgical decompression is needed in refractory cases. Transfusions, radiotherapy, and hemoglobin-stimulating agents also play a role.

http://radiologyebook.com

CHAPTER 34  Systemic Diseases Affecting the Spine

A

909

B FIG 34-8  Myelofibrosis. Coronal CT demonstrates diffuse osseous sclerosis and splenomegaly in patient with myelofibrosis.

REFERENCES 1. Alorainy IA, Al-Asmi AR, del Carpio R: MRI features of epidural extramedullary hematopoiesis. Eur J Radiol 35(1):8–11, 2000. 2. Colina M, La Corte R, De Leonardis F, et al: Paget’s disease of bone: A review. Rheumatol Int 28(11):1069–1075, 2008. 3. Corral-Gudino L, Borao-Cengotita-Bengoa M, Del Pino-Montes J, et al: Epidemiology of Paget’s disease of bone: A systematic review and meta-analysis of secular changes. Bone 55(2):347–352, 2013. 4. Dell’Atti C, Cassar-Pullicino V, Lalam R, et al: The spine in Paget’s disease. Skeletal Radiol 36(7):609–626, 2007. 5. Etienne A, Gruson B, Chatelain D, et al: Myelofibrosis-associated lymphoproliferative disease: Retrospective study of 16 cases and literature review. Adv Hematol 2009:179847, 2009. 6. Hadjipavlou A, Lander P, Srolovitz H, et al: Malignant transformation in Paget disease of bone. Cancer 70(12):2802–2808, 1992. 7. Haidar R, Mhaidli H, Taher AT: Paraspinal extramedullary hematopoiesis in patients with thalassemia intermedia. Eur Spine J 19(6):871–878, 2010. 8. Hamouda M, Handous I, Dhia NB, et al: Brown tumors in dialyzed patients with secondary hyperparathyroidism: Report of 16 cases. Hemodial Int 16(4):497–503, 2012. 9. Hansen MF, Seton M, Merchant A: Osteosarcoma in Paget’s disease of bone. J Bone Miner Res 21(S2):P58–P63, 2006. 10. Jevtic V: Imaging of renal osteodystrophy. Eur J Radiol 46(2):85–95, 2003. 11. Mesa RA, Tefferi A: Emerging drugs for the therapy of primary and post essential thrombocythemia, post polycythemia vera myelofibrosis. Expert Opin Emerg Drugs 14(3):471–479, 2009.

12. Paderni S, Bandiera S, Boriani S: Vertebral localization of a brown tumor: Description of a case and review of the literature. Chir Organi Mov 88(1):83–91, 2003. 13. Ralston SH: Paget’s disease of bone. N Engl J Med 368(7):644–650, 2013. 14. Saifuddin A, Hassan A: Paget’s disease of the spine: Unusual features and complications. Clin Radiol 58(2):102–111, 2003. 15. Singh S, Gunatunga I, Bartley L, et al: Paraspinal extramedullary hematopoiesis diagnosed using tc-99m sulfur colloid. Clin Nucl Med 33(11):794–796, 2008. 16. Smith SE, Murphey MD, Motamedi K, et al: From the archives   of the AFIP. Radiologic spectrum of Paget disease of bone and its complications with pathologic correlation. Radiographics 22(5):1191– 1216, 2002. 17. Spieth ME, Kasner DL, Manor WF: Positron emission tomography and Paget disease: Hot is not necessarily malignant. Clin Nucl Med 28(9):773–774, 2003. 18. Sundaram M: Founders lecture 2007: Metabolic bone disease: What has changed in 30 years? Skeletal Radiol 38(9):841–853, 2009. 19. Tefferi A, Vardiman JW: Myelodysplastic syndromes. N Engl J Med 361(19):1872–1885, 2009. 20. Tessitore E, Luzi M, Lobrinus JA, et al: Cervical Paget disease of bone with spinal cord compression due to atlanto-axial instability: A case report and review of the literature. Spine (Phila Pa 1976) 33(3):E85–E89, 2008. 21. Whitehouse RW: Paget’s disease of bone. Semin Musculoskelet Radiol 6(04):313–322, 2002. 22. Whyte MP: Paget’s disease of bone. N Engl J Med 355(6):593–600, 2006.

http://radiologyebook.com

35  Congenital Abnormalities Manal Nicolas-Jilwan

INTRODUCTION Development of the spine and spinal cord is a complex process determined by an intricate chain of embryologic stages under precise genetic control. Although congenital abnormalities of spine development are frequently classified according to the specific derangement responsible for the anomaly, the pathogenesis of some abnormalities is not well understood. We review these anomalies while classifying them from clinical and imaging perspectives. These include three major categories: open spinal dysraphism (OSD), which presents clinically with a usually low lumbosacral skin defect with neural tissue exposed at this site; closed spinal dysraphism (CSD) with a subcutaneous mass that typically manifests clinically as a low lumbar skincovered mass situated above the gluteal cleft and extending asymmetrically into one buttock; and CSD without a subcutaneous mass, which may come to clinical attention secondary to spinal cord tethering or other neurologic symptoms, associated cutaneous stigmata, or congenital spinal deformities. Within each of these categories, we will organize the described malformations depending on their responsible embryologic derangements.

EMBRYOLOGY Spinal cord development occurs during gestational weeks 2 to 6 and follows three consecutive steps: gastrulation (weeks 2-3), primary neurulation (weeks 3-4), and secondary neurulation with retrogressive differentiation (weeks 5-6).15,37

Gastrulation (Fig. 35-1) During the first 10 to 14 days of life, the embryo consists of two cell layers: the ectoderm facing the amniotic cavity and the endoderm facing the yolk sac. Gastrulation occurs at weeks 2 to 3 of gestation and transforms the embryo into a trilayered structure with formation of an intervening third layer, the mesoderm. This happens as follows: the primitive streak is a strip of thickened ectoderm that forms in the midline of the caudal aspect of the dorsal surface of the embryo. The primitive streak is composed of totipotential cells and is an area of active mitotic activity. Its cranial end thickens and forms a knoblike structure, Hensen’s node, that contains a central depression called the primitive pit. Ectodermal cells migrate to the primitive streak and transit inward through the primitive pit. Once they reach the interface between ectoderm and endoderm, they migrate laterally to form the mesoderm.37 Subsequently the paired mesodermal anlages join in the midline to form the notochord. The notochord is a transient embryologic structure that is essential for patterning of surrounding tissues by secreting a signaling molecule called sonic hedgehog (shh).40 The notochord regulates development of ectoderm-derived structures (in

910

particular differentiation of the neural plate), patterning of mesodermderived structures including differentiation of somites (which form the vertebral column and paraspinal musculature), and development of the cardiovascular system and endoderm-derived organs including organs of the respiratory and gastrointestinal tract.6

Primary Neurulation (Fig. 35-2) At the beginning of the third gestational week, the notochord induces the differentiation of the neuroectoderm from the ectoderm. The neuroectoderm is originally flat and forms the neural plate, which is continuous laterally on both sides with the remainder of the ectoderm. The neural plate progressively invaginates along its central axis, forming a neural groove, with its margins forming neural folds that approximate each other in the midline. The neural folds finally fuse and form the neural tube. The process of neural tube closure was traditionally thought to start at the level of the craniocervical junction and proceed rostrally and caudally in a zipperlike fashion. Recent pathologic and epidemiologic evidence suggests that neural tube closure occurs at multiple sites rather than being a continuous bidirectional closure.21,38 This multisite closure theory better explains the pathogenesis of human neural tube defects. The caudal end of the neural tube (posterior neuropore) closes at 27 to 28 days of gestation— after the cranial end (anterior neuropore), which closes at about 24 to 25 days of gestation. Closure of the entire neural tube terminates the process of primary neurulation. The neural tube formed during primary neurulation will later develop into the brain and the upper 90% of the spinal cord.37 While the neural tube is closing, it detaches from the surface ectoderm, and the latter closes dorsal to the neural tube to provide a continuous skin covering of the neural tube. This process is called disjunction. The mesenchyme that migrates dorsally between the skin and neural tube forms the meninges, neural arches, and the paraspinal muscles. The cells at the bilateral margins between ectoderm and neuroectoderm form the neural crests. While the neural plate invaginates during its closure, the neural crests detach from the ectoderm and neural plate and fragment on both sides of the neural tube to form the sensory ganglia.

Secondary Neurulation and Retrogressive Differentiation (Fig. 35-3) Secondary neurulation is responsible for the formation of the tip of the conus medullaris and filum terminale. It refers to formation of a secondary neural tube at the caudal aspect of the posterior neuropore. This starts as a caudal cell mass of undifferentiated totipotential cells originating from the caudal end of the primitive streak. The caudal cell mass undergoes canalization with formation of a central canal. The

http://radiologyebook.com

911

CHAPTER 35  Congenital Abnormalities Amniotic cavity

Embryonic disc Primitive streak

Ectoderm

Endoderm

Yolk sac

Endoderm

Mesoderm

FIG 35-1  Illustration of gastrulation. Cross-sectional representation of cell migration at the primitive pit. The primitive pit forms at the rostral end of the primitive streak. Cells ingress from the ectoderm to the interface between ectoderm and endoderm, migrating laterally on both sides within this interface to form the mesoderm.

central canal of the caudal mass merges with the neural tube canal formed during primary neurulation. Retrogressive differentiation is the step that follows secondary neurulation. This is an apoptotic process where there is regression and differentiation of the distal segment formed during secondary neurulation, which eventually becomes the tip of the conus medullaris and the filum terminale. At this stage, the ventriculus terminalis is a focal expansion of the central canal present at the distal conus. It will subsequently regress in the majority of individuals.37 Formation of the anorectal and genitourinary tract from the cloaca occurs simultaneously and in close proximity to the caudal cell mass. The caudal spine should therefore always be evaluated in the presence of anomalies of the anorectal and/or lower genitourinary tract.39

Formation of the Bony Spinal Column (Figs. 35-4 and 35-5) The mesoderm forms a somite plate on each side of the neural tube and notochord. This paired plate becomes segmented into individual somites in a cyclic process that progresses from the cranial to the caudal end of the embryo. Each somite separates into a sclerotome, a dermatome, and a myotome. Sclerotomes migrate toward the midline to surround the notochord and overlying neural tube and form the vertebral bodies. Dermatomes and myotomes remain in place at the lateral aspect of the neural tube and notochord. The myotomes give rise to the paraspinal musculature. Each sclerotome separates into two halves. The caudal half of each sclerotome joins the cephalic half of the subjacent sclerotome to form a vertebral body. The intervertebral disk is formed in the space between the separated two halves of each sclerotome. Mesenchymal cells migrate in this space to form the anulus fibrosus, and the notochord remnants form the nucleus pulposus. Thus, formation of a normal vertebral body is the result of the intersegmental union between two adjacent sclerotomes. The portions of

the sclerotomes that surround the neural tube form the neural arches, which form the posterior bony elements of the spinal column. During the sixth week of gestation, signals from the notochord and neural tube lead to chondrification, which ultimately leads to ossification of the spinal column, at which point the notochord disintegrates, with its remnants forming the nucleus pulposus. Aberrations in this intersegmental formation of vertebral bodies are at the origin of vertebral segmentation anomalies, which will be discussed later in this chapter.22

OPEN SPINAL DYSRAPHISM Dysraphism refers to a defect of neural tube closure. Therefore, all dysraphic disorders consist of abnormalities of primary neurulation. In the OSD group of malformations, neural tissue is directly exposed at the skin surface and leakage of cerebrospinal fluid (CSF) is observed. OSDs include four types of malformations of different pathogenesis as follows: • Disorders of primary neurulation: myelomeningoceles and myeloceles • Disorders of gastrulation and primary neurulation: hemimyelomeningoceles and hemimyeloceles The vast majority of cases of OSD are myelomeningoceles (98.8% of cases).37 OSD is always associated with the presence of a placode, which corresponds to a segment of cord at the neural plate stage that did not undergo closure and failed to detach from the surface ectoderm, hence it is adherent to the skin surface. There is a midline skin defect where the placode is directly visible. As the mesenchyme fails to migrate behind the neural tube, bones, cartilage, and muscles develop anterolateral to the neural tissue and typically appear everted.28 A placode can be segmental—that is, at any level along the cord or terminal at the caudal end of the cord. It is most often terminal,  with the myelomeningocele being lumbosacral (Fig. 35-6). The ventral

http://radiologyebook.com

912

PART II  CT and MR Imaging of the Whole Body

Ectoderm

Neural plate border

Neural plate

Mesoderm

FIG 35-2  Primary neurulation, or closure of the neural tube, is induced by the notochord. The notochord forms in the midline by fusion of the bilateral mesoderm anlage. The neural plate differentiates from the ectoderm and folds in the midline. Its ends join at the neural plate borders, which are referred to as the neural crests. The bilateral neural crests (represented in blue) separate from ectoderm and neural plate during neural tube closure and fragment bilaterally to form spinal sensory ganglia and the peripheral nervous system. While the neural tube is closing, it detaches from the surface ectoderm, and the surface ectoderm closes dorsal to the neural tube to provide a continuous skin covering of the neural tube. This process is called disjunction.

Notochord Neural crest

Ectoderm

Neural crest cells

Neural tube

Epidermis Neural tube

Notochord, mesoderm-derived Spinal ganglion

Somites, mesoderm-derived

surface of the placode is covered by pia-arachnoid, and nerve roots originate from this surface, coursing through the subarachnoid space to reach their neural foramina. The occurrence of purely lumbar, thoracolumbar, and thoracic myelomeningoceles where the placode is segmental and the spinal cord distal to the placode normally neurulated cannot be explained by the traditional theory of primary neurulation, where closure of the neural tube is believed to proceed in a zipperlike pattern cranially and caudally in a continuous fashion. These cases of OSD can be explained by the concept of multisite closure of the neural tube suggested by Van Allen et al., which supports that the neural tube closes independently at five different sites.38 A myelomeningocele can clinically be differentiated from a myelocele. In cases of myelocele, the flat neural placode is level with the skin surface, whereas in myelomeningocele, the neural placode is pushed above skin level by the more anterior expansion of the CSF space.15 Hemimyelocele and hemimyelomeningocele refer to the very rare cases where OSD affects one of the hemicords in a split cord malformation (SCM). They therefore represent a more complex anomaly involving the gastrulation stage (resulting in the SCM) and primary neurulation (causing the OSD). Opposite to myelomeningocele and myelocele— where the diagnosis is clinically obvious—a hemimyelocele and a hemimyelomeningocele are diagnosed by imaging, which demonstrates the split cord anomaly.37 Almost all cases of OSD are associated with an Arnold-Chiari II malformation (see Fig. 35-6). It is believed that the chronic leakage of CSF at the neural placode during intrauterine life results in defective expansion of the developing rhombencephalic vesicle (developing fourth ventricle), which prevents normal development of the posterior fossa.18 The resultant Chiari II malformation has a spectrum of anomalies that include a small posterior fossa, hindbrain herniation with downward displacement of the cerebellar vermis, fourth ventricle, and medulla into the spinal canal, as well as several variably associated structural brain abnormalities. The posterior fossa malformation is often complicated by hydrocephalus. Diagnosis of OSD and Chiari II malformation is usually confirmed prenatally by fetal ultrasound and if needed by magnetic resonance imaging (MRI).4 In utero repair of the myelomeningocele before 26 weeks’ gestation is increasingly performed in an attempt to reduce the severity of the associated Chiari II malformation. Intrauterine repair reduces the need for shunting and improves mental development, motor outcome, and hindbrain herniation. This suggests a two-hit hypothesis where the final neurologic outcome depends on the failure of neural tube closure and spinal cord injury resulting from repetitive mechanical trauma and chemical trauma related to prolonged exposure of neural tissue to the amniotic fluid.1

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

1. Dispersed mesenchyme

2. Mesenchyme condensation

3. Formation of the caudal cell mass

4. Canalization of the caudal cell mass

913

5. Neural tube complete

FIG 35-3  Secondary neurulation is responsible for the formation of the tip of the conus medullaris and filum terminale. It refers to formation of a secondary neural tube at the caudal aspect of the primary neural tube. This starts by conglomeration of a caudal cell mass of undifferentiated totipotential cells originating from the caudal end of the primitive streak. The caudal cell mass then undergoes canalization with formation of a central canal. The central canal of the caudal mass merges with the neural tube canal formed during primary neurulation.

CLOSED SPINAL DYSRAPHISM WITH A SUBCUTANEOUS MASS 7

3 6

1

5 4 2

FIG 35-4  Somite separation. The bilateral mesoderm segments into individual somites. Somites separate into sclerotomes (5), dermatomes (7), and myotomes (6). Dermatomes and myotomes remain in place. The sclerotomes migrate to surround the notochord (4) and overlying neural tube (3) and form the vertebral bodies. The myotomes give rise to the paraspinal musculature. 1, ectoderm; 2, endoderm.

In the CSD group of malformations, the malformed neural tube is covered by skin (ectoderm) and subcutaneous fat (mesoderm). Midline lumbosacral skin lesions are frequently associated, including subcutaneous lipomas, skin tags, vascular nevi, dimples, and hair patches, which aid in orienting the clinician to the diagnosis.13,33 From a pathogenetic standpoint, this group includes three categories of disorders as follows: • Disorders of primary neurulation: lipomyeloschisis (or lipomyelocele) and lipomyelomeningocele • Disorders of secondary neurulation and retrogressive differentiation: terminal myelocystocele • Disorders of unknown etiology: meningocele These malformations most commonly occur in the lower lumbar region immediately above the gluteal cleft and present clinically as a midline subcutaneous mass extending asymmetrically into one buttock. This group includes two common conditions, lipomyeloschisis (or lipomyelocele) and lipomyelomeningocele, collectively referred to as lipomas with a dural defect; and two uncommon conditions, terminal myelocystocele and meningocele. Very rarely these malformations involve the cervical spine, where cervical myelocystoceles and cervical myelomeningoceles have been described (Fig. 35-7).

http://radiologyebook.com

914

PART II  CT and MR Imaging of the Whole Body Intervertebral disk

Terminal Myelocystocele

Sclerotome Vertebra

Annulus fibrosus

Nucleus pulposus Notochord

FIG 35-5  Intersegmental formation of the vertebral bodies. Each sclerotome separates into a cranial and a caudal portion that surround the notochord. Fusion of the caudal portion of a sclerotome with the cranial portion of the subjacent sclerotome forms a vertebral body. Intervertebral disks form between the rostral and caudal portions of each sclerotome. The notochord later regresses and only persists as the nucleus pulposus.

Lipomyeloschisis and Lipomyelomeningocele In both lipomyeloschisis and lipomyelomeningocele, a large subcutaneous lipoma extends into the spinal canal through a wide spina bifida and tethers the spinal cord.37 These lipomas do not represent lipomatous tumors but rather hamartomatous lesions that can grow along with changes in the fatty pool in the body (i.e., with weight gain or pregnancy).12 They are believed to be secondary to premature disjunction of the neural tube from the neuroectoderm before neural tube closure. This allows mesenchymal cells to gain access to the neural tube before its closure. The interaction of mesenchymal cells with the neural tube induces formation of a variable-sized lipoma.15 Depending on the size of the lipoma and the degree of expansion of the subarachnoid space, the interface lipoma-placode may be in the spinal canal (lipomyeloschisis) or outside the spinal canal within a meningocele when the subarachnoid space is expanded (lipomyelomeningocele).36 In lipomyeloschisis (Fig. 35-8) the placode is terminal and the placodelipoma interface may span multiple vertebral levels. The subarachnoid space ventral to the placode is not expanded. In lipomyelomeningocele the placode is usually segmental and the ventral subarachnoid space is expanded. The placode is most frequently rotated toward the lipoma on one side, so that the meninges herniate on the contralateral side. Spinal roots arising on the side of the meninges are longer and at risk of trauma, whereas those arising on the side of the lipoma are short and may cause tethering of the spinal cord.36

Terminal myelocystocele is an extremely rare form of CSD. It is, however, very important to recognize this malformation when present, because infants are usually neurologically normal at birth, and early surgical repair permits preservation of neurologic function. In this form of CSD, the subcutaneous mass represents an excessively dilated terminal ventricle lined by ependyma that herniates through a wide posterior spina bifida. The upper portion of the myelocystocele is a dilatation of the subarachnoid space, forming a meningocele lined by arachnoid and dura. The lower portion corresponds to the cystlike dilated terminal ventricle protruding beyond the confines of the thecal sac and covered by subcutaneous fat of variable thickness. This malformation is frequently associated with other abnormalities, mainly from the OEIS complex (omphalocele, cloacal exstrophy, imperforate anus, and spine defects). The pathogenesis is thought to be a deficient retrogressive differentiation, causing the terminal ventricle to be severely distended by CSF because its terminal end remains closed. It subsequently bulges posteriorly and ruptures the dorsal mesenchyme, preventing the posterior spinal elements from developing.5 There is a possible relationship of myelocystoceles to teratogens such as loperamide HCl and retinoic acid.26

Posterior Meningocele Posterior meningoceles (Fig. 35-9) are rare and seen most commonly in the lumbar or sacral region. They represent herniation of a CSFfilled sac lined by arachnoid and dura through a posterior spina bifida. The spinal cord is normal and does not herniate in the meningocele sac, although it can be tethered to its neck.

CLOSED SPINAL DYSRAPHISM WITHOUT A SUBCUTANEOUS MASS Simple Dysraphic States Simple dysraphic states represent the most common group of spinal dysraphisms and present clinically with different degrees of spinal cord tethering due to excessive tension on the cord.30,37 From an embryologic standpoint, they include the following two groups of disorders: • Disorders of primary neurulation: intradural lipomas • Disorders of secondary neurulation and retrogressive differentiation: lipoma of the filum terminale, tight filum terminale, and persistent terminal ventricle

Normal Position of the Conus Medullaris.  The conus medullaris position relative to the vertebral column ascends gradually during fetal life. It reaches its normal adult level by 40 postconceptional weeks, and no further ascent is observed thereafter.3 The conus medullaris should be positioned no lower than the inferior endplate of the L2 vertebra. Otherwise a tethering mass, bony spur, or thick filum terminale should be sought.2

Spina Bifida Occulta.  This represents absent fusion of the posterior vertebral neural arches, usually at L5 and/or S1, and is seen in 4% of the population.31 It is usually an incidental finding unless there are clinical symptoms or signs of spinal cord tethering, when it can point to the presence of another CSD. It should be noted that the L5 laminae may normally remain unfused until age 5 to 6 years.20

Persistent Terminal Ventricle and Filar Cysts.  A persistent terminal ventricle represents persistence of the normal terminal ventricle, a small cystic expansion of the central ependymal canal at the level of the conus that fails to undergo normal regression during secondary

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

A

B

C

* *

D

E

FIG 35-6  Lumbosacral myelomeningocele. A, Midline sagittal T2-weighted image demonstrates absent posterior elements at the lower lumbar and sacral levels. The spinal cord exits at the upper aspect of this dysraphic segment into a meningocele sac (upper arrow in lumbar region). This segment of the cord consists of a flat nonneurulated placode, which in this patient is adherent to the meningocele sac (lower arrow in lumbar region). The uppermost arrow in the upper cervical region indicates the pointed cerebellar tonsils, which are herniated into the upper spinal canal, consistent with the almost always (>98% of cases) associated Chiari II malformation. B, Sagittal T2-weighted image off midline shows the cauda equina nerves coursing from the neural placode anteriorly (arrows). C, Axial high-resolution T2-weighted image through the level indicated by the arrows in A shows the nonneurulated placode (white arrow) exiting the spinal canal and coursing in the meningocele sac toward its posterior wall (small black arrows). D, Axial high-resolution T2-weighted image at a lower level through the meningocele sac demonstrates nerves (black arrows) coursing from the placode anteriorly. E, Axial T2-weighted image at the level of the lateral ventricles showing the ventricular dilatation that frequently complicates the altered CSF dynamics as a result of the crowded craniocervical junction. The septum pellucidum is absent (*). This is a common finding in Chiari II malformation. It represents destruction of the septum pellucidum secondary to high intraventricular pressure.

http://radiologyebook.com

915

916

PART II  CT and MR Imaging of the Whole Body

* A

B

*

C

D

FIG 35-7  Cervical myelocystocele complicated by squamous cell carcinoma (an extremely rare occurrence). A 22-year-old presented with a mass in the back of his neck since birth, which started bleeding recently. A, Midline sagittal T1-weighted image demonstrates the typical appearance of a narrow posterior spinal dysraphic defect (arrow) through which neural elements and dura exit into a large meningocele sac. B, Axial T2 image shows that the cord remains in the spinal canal (short anterior arrow) and dysmorphic neural elements from the dorsal aspect of the cord are stretched into the meningocele sac (long posterior arrow). C, Postcontrast sagittal T1 image with fat saturation shows an enhancing mass (*) in the posterior subcutaneous tissues, pathologically confirmed to represent squamous cell carcinoma, a very rare complication of this very rare malformation. D, Sagittal T2 sequence after tumor resection better demonstrates the neural elements stretched from the dorsal cord into the meningocele sac (arrow). The cord itself remains in the spinal canal.

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

*

*

A

B

FIG 35-8  Lipomyeloschisis. A, Sagittal T1 image best demonstrates the T1-hyperintense large lipoma (*) extending through a wide spinal dysraphic defect (black arrow) to the subcutaneous fat. B, The terminal placode-lipoma interface lies within the spinal canal (white arrow). The subarachnoid space ventral to the placode is not expanded. Note the mild hydromyelia that not infrequently complicates spinal cord tethering.

* *

A

B

*

C

FIG 35-9  Posterior meningocele. One-day-old with prenatal diagnosis of lumbosacral meningocele. A, Sagittal T1 image shows a wide lumbosacral spinal dysraphic defect with a meningocele sac herniating at this site (*). The cord is low lying (arrow) and tethered at the site of attachment of the meningocele. B, Axial T2 at the level of the meningocele (*) demonstrates the low-lying conus in the spinal canal (arrow). C, Sagittal T2 image demonstrates a nerve (arrow) herniated in the meningocele (*) (surgically confirmed).

http://radiologyebook.com

917

918

PART II  CT and MR Imaging of the Whole Body

A

B

C

*

D FIG 35-10  Cervical intradural lipoma in a 20-year-old woman. A and B, Sagittal T1 and T2 images show a well-defined spindle-shaped intradural extramedullary tumor at the dorsal aspect of the cord, with signal characteristics consistent with fat, including high T1 signal and high signal on fast-spin echo T2 imaging. C, Postcontrast T1 with fat saturation confirms the fat content of the lesion, which suppresses signal on this sequence. No solid enhancing component is demonstrated. D, Axial T2 image shows the mass (*) displacing the cord to the left (arrow).

neurulation. They are usually asymptomatic and seen as a nonenhancing ovoid small cystic expansion of the central ependymal canal in a normally positioned conus. Symptomatic cases are associated with obstruction of the ventriculus terminalis, which loses its connection to the central ependymal canal. The obstructed ventriculus terminalis becomes expanded and results in thinning of the surrounding cord parenchyma.7 Filar cysts are simple cysts of the filum terminale, commonly seen on spinal sonography of infants. When isolated, they are generally asymptomatic and regress spontaneously at follow-up.16

Intradural Lipomas.  These are benign masses that are secondary to early disjunction of the neural tube from the ectoderm, allowing the mesenchyme to insinuate and adhere to the ependyma. This induces

the mesenchyme to differentiate into adipose tissue.37 These masses are subpial, spindle-shaped, well defined, and frequently situated dorsal or dorsolateral to the cord14 (Fig. 35-10).

Fibrolipoma of the Filum Terminale.  The normal filum terminale is a thin fibrous structure that extends from the tip of the conus medullaris through the thecal sac, pierces the lower tip of the thecal sac, and follows a short extradural course to insert on the dorsal aspect of the first coccygeal vertebra.9 The normal filum terminale measures no more than 1 mm at the L5-S1 level.2 Fibrolipomas of the filum terminale represent a disorder of secondary neurulation and result in excessive fibrous thickening and fatty infiltration of the filum terminale. Small amounts of fat can be seen along the filum in normal people and are reported in 4% to 5% of

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

*

* A

B

919

* C

FIG 35-11  Fibrolipoma of the filum terminale. Seven-month-old with tethered cord syndrome. A, Sagittal T1 image demonstrates fatty infiltration of the filum terminale, showing high T1 signal (*). B, Sagittal T2 image shows a low-lying conus at L3 (arrow) and redemonstrates the fatty filum. Note also the posterior sacral dysraphism. C, Axial T1 image through the filar lipoma shows the T1-hyperintense fatty mass (*). The central globular T1-hypointense component anteriorly (arrow) represents fibrous thickening of the filum.

patients at autopsy.15 They may, however, become symptomatic at any age.25 A tethered conus medullaris is frequently seen.27 Fat along the filum terminale is best depicted as high T1 signal on axial T1-weighted MRIs (Fig. 35-11).

Tight Filum Terminale.  This is a disorder of retrogressive differentiation resulting in a thick and short filum terminale that exerts downward traction on the lower cord. Because the dentate ligaments secure the spinal cord in position above T12, the tension from a short filum is predominantly transmitted to the conus medullaris, and symptoms are usually confined to the lower extremities, bladder, and bowel function.41 The conus is frequently low lying. The filum is thick but without fatty infiltration.

COMPLEX DYSRAPHIC STATES This group of disorders results from a defect during the gastrulation stage of embryonic development, with abnormal development of the notochord. Abnormal development of the notochord can consist of two different scenarios as follows: • Disorders of midline notochordal integration that result in longitudinal splitting: dorsal enteric fistula, neurenteric cyst, diastematomyelia, and dermal sinus • Failure of notochordal formation that results in absence of a given notochordal segment: caudal regression syndrome and segmental spinal dysgenesis The notochord not only induces formation of the spine and spinal cord but also several other organs in the body. Consequently, this category of spinal dysraphism is characterized by complex anomalies of the spine and other organ systems.29

Disorders of Midline Notochordal Integration

Split Cord Malformations or Diastematomyelia.  The neurenteric canal (Fig. 35-12) is a transient communication between the amniotic cavity and yolk sac that normally persists for only 48 hours during the third week of gestation and is obliterated afterward. Its transient formation is secondary to three stages of notochordal development: (1) canalization forming a hollow notochordal process, (2) intercalation forming a notochordal plate, and (3) excalation forming the solid final notochord. The notochordal process undergoes canalization, and its central cavity communicates with the amniotic cavity through the primitive pit. The canalized notochordal process transiently fuses with the endoderm to become a notochordal plate, a process called intercalation, which allows communication of the amnion with the yolk sac and formation of the primitive neurenteric canal. During excalation the notochord separates from the endoderm and its cavity is obliterated. The neurenteric canal is subsequently obliterated.8 SCMs result from adhesions forming between ectoderm and endoderm, leading to persistence of the neurenteric canal, which prevents fusion of the bilateral notochordal anlage in the midline. Each notochord subsequently induces the development of a hemineural plate.24 The primitive streak condenses around the neurenteric canal, forming an endomesenchymal tract. The endomesenchymal tract may form bone and cartilage, developing into an osteocartilaginous spur that separates two hemicords contained in separate dural sacs (type I SCM). If the endomesenchymal tract involutes completely or into a thin fibrous septum, the two hemicords are contained in a single dural sac (type II SCM). Type I SCM (Fig. 35-13).  A very reliable sign is a hair tuft high on the back of the patient. Patients present with symptoms of cord

http://radiologyebook.com

920

PART II  CT and MR Imaging of the Whole Body Notochordal canalization Neural plate Primitive pit

iotic

Notochordal process

Amn

sac

Notochordal canal k Yol

Intercalation

sac

Neurenteric canal

Notochordal/endodermal fusion

Excalation

Notochord

FIG 35-12  Formation of the neurenteric canal. Its transient formation is secondary to three stages of notochordal development: 1. Canalization forming a hollow notochordal process that communicates at the primitive pit with the amniotic cavity. 2. Intercalation: the canalized notochordal process transiently fuses with the endoderm to become a notochordal plate, which allows communication of the amnion with the yolk sac and formation of the primitive neurenteric canal. 3. Excalation forming the solid final notochord. During excalation, the notochord separates from the endoderm and its cavity is obliterated. The neurenteric canal is subsequently obliterated.

tethering usually associated with scoliosis and variable spinal segmentation-fusion anomalies such as hemivertebrae, block vertebrae, spina bifida, bifid laminae, and wide interpedicular distance. The hallmark of type I SCM is the osteocartilaginous spur, which most often is complete, extending from the neural arch to the posterior wall of the vertebral body; less often it is incomplete. It divides the spinal canal symmetrically or (less often) asymmetrically, resulting in two

dural sacs and hemicords of different sizes. The spur lies at the caudal aspect of the split. When the cleft is terminal, the filum is also split and generally thickened.29 Hydromyelia can affect the normal or split cord about the level of splitting. Type II SCM (Fig. 35-14).  It consists of a single dural sac containing both hemicords. There are three subtypes of gradually higher severity: partial splitting, split cord without a fibrous septum, and split

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

*

A

B

*

*

C

D

E FIG 35-13  SCM type I. A, Axial T2 at the inferior aspect of the split cord segment demonstrates a bony spur extending from the posterior wall of the vertebral body, separating two dural sacs each containing a hemicord (arrow). The left hemicord is slightly larger than the right. Note the associated spinal dysraphism (*). B, Axial T2 at a slightly higher level shows a single thecal sac with two hemicords. The osteocartilaginous septum and dural sac separation occur at the lower aspect of the split cord segment. C, Sagittal T2 demonstrates associated block vertebra (short arrow) and posterior spinal dysraphism (*). A segment of hydromyelia above the level of cord splitting (long arrow) is common. The conus terminates at L4. Sagittal T1 (D) and axial T1 (E) best delineate associated fatty infiltration of the filum terminale (arrows).

http://radiologyebook.com

921

922

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 35-14  SCM type II. Ten-year-old patient known to have split cord malformation since birth when she presented with a hairy patch over her lower thoracic spine. Axial T2 (A) and coronal T1 (C) demonstrate a segment of splitting of the distal cord into two hemicords of similar size within a single dural sac (arrows). Note the absence of an osteocartilaginous or fibrous spur or dural sac splitting, which are characteristic of SCM type I. Mild hydromyelia above the level of cord splitting (arrow in B) is common.

A

B

C

D

FIG 35-15  Dermal sinus tract in a 6-month-old with lower lumbar dimple. A, High-resolution CISS sequence best demonstrates the subcutaneous tract (vertical arrow), which pierces the dura and ascends to the dorsal surface of the conus (horizontal arrow). B, Axial CISS sequence also demonstrates a small cyst at the tip of the conus (arrow), which likely represents a tiny epidermoid or dermoid cyst. C, Sagittal T1 delineates well the subcutaneous dermal tract against the T1-hyperintense fat. D, Mild fatty infiltration of the filum is best depicted on axial T1 imaging (arrow).

cord with a fibrous septum. The presence of the fibrous septum can tether the cord. Vertebral anomalies are milder than in type I and often consist of butterfly vertebrae and posterior spina bifida. Hydromyelia can be present.

Neurenteric Cysts.  These represent heterotopic endodermal rests of mucin-secreting epithelium that resemble the gastrointestinal epithelium and are found as intradural cysts in the ventral spinal canal, with a predilection for the cervicomedullary junction. They can be associated with ventral vertebral anomalies such as hemivertebrae, block vertebrae, or a bifid clivus.19,32

Dorsal Enteric Fistula.  This is an extremely rare malformation that consists of a direct connection between endodermal derivatives and the neuroectoderm owing to failed integration of the notochord and persistence of the neurenteric canal. There is a cleft through the vertebral column and spinal cord, which are both split, connecting bowel or mesentery to the dorsal skin surface.

Dorsal Dermal Sinus.  A dorsal dermal sinus (Fig. 35-15) is a small persistent epithelium-lined tract between the neural tube and the skin surface. A small cutaneous dimple is usually noted at the site of the opening. The pathogenesis of dermal sinus tracts is not well

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

A

B

C

923

D

FIG 35-16  Type I caudal regression syndrome in a 5-year-old girl. A and B, Sagittal T2- and T1-weighted images show blunted contour of the conus (black arrow in A). Only the first sacral segment is formed (white arrow in A), and the thecal sac terminates high at L5 (arrow in B). C, Several vertebral segmentation fusion anomalies are present, including partial fusion of C2 and C3 (short arrow) and fusion of the T1 and T2 vertebrae (long arrow). D, Coronal T2 sequence best demonstrates sagittal clefts through the T7 and T8 vertebral bodies (arrows).

understood, but they are likely secondary to incomplete disjunction of the neuroectoderm from the surface ectoderm and are most commonly seen in the lumbar region.15 The tract can extend to the dura without entering the thecal sac (the dura appears tented at the site of insertion), to the subarachnoid space, conus, a thickened lipomatous filum, or an intradural lipoma adherent to the conus. Associated intradural dermoid or epidermoid cysts are not uncommon and usually seen near the conus. These cysts are best demonstrated on fluidattenuated inversion recovery (FLAIR) and diffusion imaging (epidermoid cysts exhibit restricted diffusion), because they are frequently of similar signal intensity to CSF on conventional T1- and T2-weighted images. The subcutaneous portion of the dermal sinus is well demonstrated on thin-section T1-weighted images as a low-signal, thin tract coursing through the T1 hyperintense fat. The intradural tract can best be delineated on heavily T2-weighted high-resolution threedimensional (3D) imaging, which can be reconstructed in three planes. Dermal sinus tracts can be complicated by infection, meningitis, and abscess formation, which are best evaluated with postcontrast imaging.

Failure of Notochordal Formation The location of each notochordal cell along the craniocaudal axis of the embryo is under the control of homeobox genes. Cells that are improperly placed along the craniocaudal axis of the embryo during gastrulation are destroyed by a process referred to as positional apoptosis.17 This segmental cell destruction results in a segmental defect that affects the spinal cord, spinal column, and other organs that depend on the notochord as their inductor. Most commonly the abnormality involves the caudal end of the embryo, affecting the caudal cell mass and resulting in the caudal regression syndrome. When an intermediate segment is affected, the result is a segmental spinal dysgenesis; in milder forms involving only the spinal column, segmental fusion/ segmentation anomalies of the spinal column are seen.35

Caudal Regression Syndrome (CRS).  This rare malformation is frequently associated with malformations of the hindgut and urogenital system. The lower extremities are always dysplastic,10,23 with the very rare and most severe cases presenting with fusion of the lower extremities (sirenomelia).34 CRS can be part of the OEIS complex (omphalocele, cloacal exstrophy, imperforate anus, and spinal deformities), the VATER syndrome (vertebral abnormality, anal imperforation, tracheoesophageal fistula, and renal-radial anomalies), or the Currarino triad (partial sacral agenesis, anorectal malformation, and sacrococcygeal teratoma). The level of weakness correlates well with the level of the bone anomaly, whereas sensory function is relatively preserved.23 This malformation has been linked to maternal diabetes. Type I CRS (Fig. 35-16).  This is the more severe form of CRS. The spine ends above S1 and the spinal cord ends high with an abrupt club-shaped termination lacking the normal distal tapering of the distal cord. Type II CRS (Fig. 35-17).  This is the milder form of dysgenesis with the spine formed down to S2 or even lower levels. Only the most distal conus medullaris is absent. The conus is stretched caudally and tethered by a tight filum, intradural lipoma, terminal myelocystocele, or lipomyelomeningocele.37

Segmental Spinal Dysgenesis.  This includes segmental dysgenesis or agenesis of the thoracic, lumbar, or sacral spine, with associated segmental anomaly of the cord and nerve roots. There is resultant paraplegia or paraparesis.35

SEGMENTATION ANOMALIES OF THE SPINAL COLUMN Spinal segmentation anomalies can be encountered in variable clinical settings, including as incidental findings on chest or abdominal radiographs, during the workup of chronic back pain, the workup of

http://radiologyebook.com

924

PART II  CT and MR Imaging of the Whole Body

*

*

A

*

*

B

C

D

FIG 35-17  CRS type II in a 22-year-old woman. A, Sagittal T2 shows mild sacrococcygeal dysgenesis, with three sacral segments formed and a sacralized L5 vertebra (* in A). B, The spinal cord is stretched inferiorly and terminates at the lower L4 level (arrow). A tiny syrinx is noted more superiorly in the cord. C and D, The cord is tethered by a fibrofatty filum, indicated by arrows in C (sagittal T2) and D (sagittal T1) images.

*

A

* *

B

C

D

FIG 35-18  Extensive spinal segmentation fusion anomalies in a 12-year-old girl. Fused lumbar vertebrae with resultant lumbar kyphosis, indicated by * on sagittal T2 image (A) and arrows on coronal T2 image (C). An anterior hemivertebra in the lower lumbar region (arrow in A) and lower thoracic butterfly vertebrae (arrows in B) are also present. Note the associated cross-fused renal ectopia (arrow in D).

progressive scoliosis, or in association with a wide spectrum of spinal dysraphism, syndromes, and chromosomal abnormalities. Severe cases can have significant implications such as resultant kyphosis, scoliosis, back pain, functional disability, and limitation of pulmonary function. Their occurrence has been linked to a large variety of maternal exposures during pregnancy, including alcohol use, valproic acid, hyperthermia, and diabetes mellitus.11 Failure of vertebral segmentation can lead to vertebral body fusion (block vertebra), pedicular fusion (pediculate bar), and neural arch fusion (Fig. 35-18). The resultant scoliotic deformity depends on the extent of involvement. In the case of a single-level abnormality, a sharp focal scoliosis is usually seen (Fig. 35-19). Associated rib fusion can

also be seen and is more likely to result in restriction of pulmonary function than isolated vertebral anomalies. Aberrant intersegmental fusion results in hemivertebrae (wedge-shaped vertebrae that are most common in the midthoracic spine and can also result in kyphoscoliosis) and in milder cases, butterfly vertebrae (a central incomplete cleft through the vertebral body due to persistence of the notochord, usually asymptomatic).

CONCLUSION Box 35-1 summarizes the spectrum of congenital spinal malformations that we reviewed in this chapter and categorizes them according to

http://radiologyebook.com

CHAPTER 35  Congenital Abnormalities

925

BOX 35-1  Classification of Spinal Dysraphism Open Spinal Dysraphism Anomalies of Primary Neurulation • Myelomeningocele • Myelocele Anomalies of Gastrulation and Primary Neurulation • Hemimyelomeningocele • Hemimyelocele Closed Spinal Dysraphism with a Subcutaneous Mass Anomalies of Primary Neurulation • Lipomyelomeningocele • Lipomyeloschisis

Anomalies of Secondary Neurulation and Retrogressive Differentiation • Persistent terminal ventricle and filar cyst • Lipoma of the filum terminale • Tight filum terminale Complex Dysraphic States Disorders of Midline Notochordal Integration • Dorsal enteric fistula • Neurenteric cyst • Diastematomyelia • Dermal sinus Disorders of Notochordal Formation • Caudal regression syndrome • Segmental spinal dysgenesis

Anomalies of Secondary Neurulation and Retrogressive Differentiation • Terminal myelocystocele Anomalies of Unknown Pathogenesis • Meningocele Closed Spinal Dysraphism Without a Subcutaneous Mass Simple Dysraphic States Anomalies of Primary Neurulation • Intradural lipoma

REFERENCES

A

B

FIG 35-19  A and B, Six-month-old infant with right hemivertebra at T4-T5 causing focal thoracic scoliosis centered at this level (arrows).

their pathogenesis and clinical presentation. Although the imaging features of the more complex malformations may be complicated and misleading, correlation with the patient’s clinical presentation and knowledge of the embryologic derangements involved allow an accurate diagnosis in the majority of cases.

1. Adzick NS, Thom EA, Spong CY, et al: A randomized trial of prenatal versus postnatal repair of myelomeningocele. N Engl J Med 364(11):993– 1004, 2011. 2. Barkovich A. James, Raybaud Charles (2012-02-20). Pediatric Neuroimaging. Philadelphia, PA: Lippincott, Williams & Wilkins. 3. Beek FJ, de Vries LS, Gerards LJ, et al: Sonographic determination of   the position of the conus medullaris in premature and term infants. Neuroradiology 38(Suppl 1):S174–S177, 1996. 4. Ben-Sira L, Garel C, Malinger G, et al: Prenatal diagnosis of spinal dysraphism. Childs Nerv Syst 29(9):1541–1552, 2013. 5. Byrd SE, Harvey C, Darling CF: MR of terminal myelocystoceles. Eur J Radiol 20(3):215–220, 1995. 6. Cleaver O, Krieg PA: Notochord patterning of the endoderm. Dev Biol 234(1):1–12, 2001. Review. 7. Coleman LT, Zimmerman RA, Rorke LB: Ventriculus terminalis of the conus medullaris: MR findings in children. AJNR Am J Neuroradiol 16(7):1421–1426, 1995. 8. Dias MS, Walker ML: The embryogenesis of complex dysraphic malformations: A disorder of gastrulation? Pediatr Neurosurg 18:229– 253, 1992. 9. Emery JL, Lendon RG: Lipomas of the cauda equina and other fatty tumours related to neurospinal dysraphism. Dev Med Child Neurol Suppl 20:62–70, 1969. 10. Estin D, Cohen AR: Caudal agenesis and associated caudal spinal cord malformations. Neurosurg Clin N Am 6(2):377–391, 1995. 11. Giampietro PF, Raggio CL, Blank RD, et al: Clinical, genetic and environmental factors associated with congenital vertebral malformations. Mol Syndromol 4(1-2):94–105, 2013. 12. Giudicelli Y, Pierre-Kahn A, Bourdeaux AM, et al: Are the metabolic characteristics of congenital intraspinal lipoma cells identical to, or different from normal adipocytes? Childs Nerv Syst 2(6):290–296, 1986. 13. Guggisberg D, Hadj-Rabia S, Viney C, et al: Skin markers of occult   spinal dysraphism in children: A review of 54 cases. Arch Dermatol 140(9):1109–1115, 2004.

http://radiologyebook.com

926

PART II  CT and MR Imaging of the Whole Body

14. Guglielmi G, Utomo S, Cascavilla M, et al: Imaging in congenital deformities of the spinal cord. Radiol Med 117(5):872–884, 2012. 15. Huisman TA, Rossi A, Tortori-Donati P: MR imaging of neonatal spinal dysraphia: What to consider? Magn Reson Imaging Clin N Am 20(1):45– 61, 2012. 16. Irani N, Goud AR, Lowe LH: Isolated filar cyst on lumbar spine sonography in infants: A case-control study. Pediatr Radiol 36(12):1283– 1288, 2006. 17. Maden M, Graham A, Gale E, et al: Positional apoptosis during vertebrate CNS development in the absence of endogenous retinoids. Development 124:2799–2805, 1997. 18. McLone DG, Dias MS: The Chiari II malformation: Cause and impact. Childs Nerv Syst 19(7–8):540–550, 2003. 19. Menezes AH, Traynelis VC: Spinal neurenteric cysts in the magnetic resonance imaging era. Neurosurgery 58(1):97–105, discussion 97–105, 2006. 20. Naidich TP, Zimmerman RA, McLone DG, et al: Congenital anomalies of the spine and spinal cord. In Atlas SW, editor: Magnetic Resonance Imaging of the Brain and Spine, ed 2, Philadelphia, 1996, LippincottRaven, pp 1265–1337. 21. Nakatsu T, Uwabe C, Shiota K: Neural tube closure in humans initiates at multiple sites: Evidence from human embryos and implications for the pathogenesis of neural tube defects. Anat Embryol (Berl) 201(6):455– 466, 2000. 22. O’Rahilly R: Human embryology and teratology, New York, 1996, John Wiley & Sons. 23. Pang D: Sacral agenesis and caudal spinal cord malformations. Neurosurgery 32(5):755–778, discussion 778–9, 1993. 24. Pang D, Dias MS, Ahab-Barmada M: Split cord malformation: Part I: A unified theory of embryogenesis for double spinal cord malformations. Neurosurgery 31(3):451–480, 1992. 25. Pang D, Wilberger JE, Jr: Tethered cord syndrome in adults. J Neurosurg 57(1):32–47, 1982. 26. Peacock WJ, Murovic JA: Magnetic resonance imaging in myelocystoceles. Report of two cases. J Neurosurg 70(5):804–807, 1989. 27. Raghavan N, Barkovich AJ, Edwards M, et al: MR imaging in the tethered spinal cord syndrome. AJR Am J Roentgenol 152(4):843–852, 1989.

28. Rossi A, Biancheri R, Cama A, et al: Imaging in spine and spinal cord malformations. Eur J Radiol 50(2):177–200, 2004. 29. Rossi A, Gandolfo C, Morana G, et al: Current classification and imaging of congenital spinal abnormalities. Semin Roentgenol 41(4):250–273, 2006. 30. Sanchez T, John RM: Early Identification of tethered cord syndrome: A clinical challenge. J Pediatr Health Care 2013 Aug 8. 31. Sattar MT, Bannister CM, Turnbull IW: Occult spinal dysraphism—The common combination of lesions and the clinical manifestations in 50 patients. Eur J Pediatr Surg 6(Suppl 1):10–14, 1996. 32. Savage JJ, Casey JN, McNeill IT, et al: Neurenteric cysts of the spine.   J Craniovertebr Junction Spine 1(1):58–63, 2010. 33. Schropp C, Sörensen N, Collmann H, et al: Cutaneous lesions in occult spinal dysraphism—Correlation with intraspinal findings. Childs Nerv Syst 22(2):125–131, 2006. 34. Stocker JT, Heifetz SA: Sirenomelia. A morphological study of   33 cases and review of the literature. Perspect Pediatr Pathol 10:7–50, 1987. 35. Tortori-Donati P, Fondelli MP, Rossi A, et al: Segmental spinal dysgenesis: Neuroradiologic findings with clinical and embryologic correlation. AJNR Am J Neuroradiol 20(3):445–456, 1999. 36. Tortori-Donati P, Rossi A, Biancheri R, et al: Magnetic resonance imaging of spinal dysraphism. Top Magn Reson Imaging 12(6):375–409, 2001. 37. Tortori-Donati P, Rossi A, Cama A: Spinal dysraphism: A review of neuroradiological features with embryological correlations and proposal for a new classification. Neuroradiology 42(7):471–491, 2000. 38. Van Allen MI, Kalousek DK, Chernoff GF, et al: Evidence for multi-site closure of the neural tube in humans. Am J Med Genet 47(5):723–743, 1993. 39. Warf BC, Scott RM, Barnes PD, et al: Tethered spinal cord in patients with anorectal and urogenital malformations. Pediatr Neurosurg 19(1):25–30, 1993. 40. Wilson L, Maden M: The mechanisms of dorsoventral patterning in the vertebrate neural tube. Dev Biol 282(1):1–13, 2005. 41. Yamada S, Knerium DS, Mandybur GM, et al: Pathophysiology of tethered cord syndrome and other complex factors. Neurol Res 26(7):722–726, 2004.

http://radiologyebook.com

36  Nonneoplastic Parenchymal Lung Disease Tomás Franquet

Although many lung disorders can be diagnosed from the plain chest radiograph, its limitations in providing a definitive diagnosis are widely understood. Complicated chest radiographs, the lack of radiographic abnormalities in symptomatic patients, and the limitations of portable chest studies make evaluation of conventional chest films sometimes difficult. Imaging evaluation of parenchymal lung disease is essentially confined to chest radiography, high-resolution computed tomography (HRCT),1 spiral CT, and multidetector CT (MDCT). These techniques serve as important adjuncts to the plain film for the purposes of detection, diagnosis, and characterization of parenchymal lung disease. Compared with conventional radiographic techniques, CT provides better spatial resolution, detects minor differences in radiographic contrast, and expands the ability of radiology to portray boundaries between different tissues.

IMAGING TECHNIQUES Imaging of the lung has changed completely during the past decade. With the improvement of CT imaging techniques, shorter examination times, thinner sections, and higher resolution are now possible without loosing image quality. The introduction of fast MDCT scanners further improved the ability to visualize small pulmonary nodules and pulmonary emboli and allowed for three-dimensional (3D) imaging of the lung.135,333,558

HRCT (Thin-Section CT) HRCT combines the use of thinnest-beam collimation (1-1.5 to 2 mm in thickness) with reconstruction algorithms designed to produce high spatial resolution.404-406 Spatial resolution in the axial plane is a function of pixel size, which is dependent on the image matrix (typically 512 × 512) and field of view. 406,659,404 Thin collimation decreases partial volume averaging and improves the ability of CT to demonstrate small pulmonary lesions.405,406 The thinnest collimation available on most scanners is 1 to 2 mm. This is closely matched to the size of the  anatomic structures of interest within the secondary pulmonary lobule.406,416,454,632 Use of thin collimation causes increased quantum image noise and decreased contrast sensitivity because fewer photons are available to reach the CT detector system. The anatomic detail provided by HRCT imaging sections allows a more detailed analysis of pathologic processes affecting the small airways, airspaces or alveolar walls, and interstitium. HRCT has become established as a useful technique for the detection and characterization of diffuse lung disease (DLD). Standard HRCT scanning with singleslice CT allows assessment of the lung at the scanned levels only and cannot exclude abnormalities in the lungs between these sections. HRCT technique is indicated in a variety of clinical settings (e.g.,

928

patients with normal chest radiography who have either chronic respiratory symptoms or abnormal pulmonary function tests). HRCT may also play an important role in detecting early morphologic changes in patients with suspected pulmonary disorders such as interstitial lung disease (ILD), emphysema, asbestosis, and bronchiectasis. Compared with standard CT radiographic technique, the use of higher kilovolt peak (kVp), higher milliamperes (mA), or longer scan duration increases the number of photons incident upon the CT detector system. Although the use of low-dose HRCT has been shown to be slightly less sensitive in detecting some instances of ground-glass opacity and interstitial lines, it has minimal compromise in diagnostic information.348,658 High-spatial-frequency algorithms, by decreasing the amount of smoothing of the image data, improve visualization of both normal and pathologic structures.406,440 This technique has been shown to be more sensitive than conventional CT and chest radiography for detection and characterization of diffuse infiltrative lung disease.432,439,454,515 HRCT can also precisely differentiate emphysema from other cystic lesions.74 Expiratory CT accurately identifies air trapping.588,591 However, assessment of air trapping can be difficult in obese patients or in those who are unable to suspend respiration, especially at endexpiration. Complementary expiratory CT maneuvers can be used to improve the visualization and extent of air trapping.199,365 Prone CT scans can be taken to differentiate focal dependent edema or microatelectasis, commonly seen in the lung bases of normal patients, from early pulmonary fibrosis. Window settings are important variables in HRCT technique. To view the lung parenchyma, window settings are in the range from 1000 to 2000 HU and level settings from −500 to −750 HU. Images of the mediastinum can be reconstructed using a standard soft tissue algorithm and viewed using mediastinal window settings of 410 HU and level settings of 35 HU. Additional window settings are adjusted manually at the time of imaging to enhance depiction of subtle ground-glass opacities or emphysema. Emphysema may be demonstrated by the use of low window levels of −800 to −900 HU and narrow widths of 500 HU.

Spiral or Helical CT Spiral CT technique requires a scanner with a continuously rotating x-ray tube capable of generating a series of consecutive scans during a single breath-hold maneuver. Unlike conventional CT, the patient is not scanned section by section but is translated through the scan plane at a uniform table speed during acquisition of raw data. The advantages of spiral CT arise from continuous data acquisition and short total scanning time. The operator chooses three basic parameters for spiral CT: section collimation, table feed per rotation, and reconstruction interval.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease The section collimation (0.5 mm, 1 mm, 2 mm, 3 mm, 5 mm, 7 mm, or 10 mm) determines the spatial resolution achieved along the z-axis. The effective slice thickness can be influenced by the table feed speed per tube rotation and the interpolation algorithm used for image reconstruction. Table speed is set to between 1 and 10 mm/second. Pitch is defined as the ratio as the table feed per tube rotation to the collimation. Pitch is usually best kept between 1.0 and 2.0. Whereas with a pitch value of 1.0 (table feed equal to collimation), longitudinal scan coverage is limited, scanning at a pitch less than 1.0 produces an overlapping scan pattern. Reconstruction of overlapping sections improves the quality of the 3D images. With single-detector-row scanners, routine spiral CT imaging in the thorax is performed with pitch values between 1.0 and 1.5, using collimation in the range of 8 to 10 mm and table movement speed of 8 to 15 mm/sec.

MDCT The advent of MDCT represents an enormous technical advance. The development of MDCT allows higher z-axis resolution and faster image acquisition than previously possible, and combined with PACS, greater diagnostic evaluation of the bronchi and vessels and improved visualization of parenchymal processes. It uses a multiple-row detector with narrow detector collimation in conjunction with a relatively rapid table translation to achieve faster scans with thinner slices than singledetector CT.558 MDCT results in a gapless volumetric acquisition that eliminates problems of slice misregistration caused by variations in the depth of inspiration.299,305,333,461,462,558,626 The resulting data set is more accurate, more reproducible from one study to the next, and more quickly acquired. The reduced scan time of MDCT reduces motion artifacts, particularly in patients with underlying lung disease.522 Knowledge of the capabilities and limitations of MDCT allows the radiologist to use MDCT appropriately and for the best possible patient care. Multislice HRCT may be used to acquire high-resolution images of the lung parenchyma.263,264,595 The image quality of multislice HRCT is equal to that on single-detector HRCT.289,595 Although spatial resolution is better on axial HRCT, multislice HRCT is appropriate for evaluating subtle lung abnormalities. Moreover, the image quality of coronal multiplanar reconstructions (MPRs) from isotropic voxel data obtained using 0.5-mm collimation, with or without overlapping reconstruction, is similar to that of direct coronal thin-section CT scans.265 Careful attention must be paid to technical parameters during scanning, and the dose needs to be tailored to the specific clinical situation. Although low-dose techniques can be used to reduce the radiation dose in patients in whom a volumetric HRCT is indicated, the advantages of accurate diagnosis may outweigh the increased radiation dose. With the latest generations of 64-slice MDCT and dual-source technology, the performance of CT scanning has dramatically improved. An important advance of MDCT scanners is that from the same data set, thin sections can be reconstructed to minimize partial volume averaging, and thicker sections can be reconstructed to decrease image noise if necessary. MDCT scanners can provide continuous and/or overlapping high-resolution thin sections throughout the lungs in a single breath hold. The provided additional information and its impact on clinical management has justified its use in the area of thoracic imaging. In combination with advanced image processing techniques, including MPRs, maximum intensity projections (MIPs), minimum intensity projections (MiniPs), external rendering with either 3D shaded-surface displays or volumetric rendering, and internal rendering (or so-called virtual bronchoscopy), CT has become a valuable tool for evaluating diseases affecting the airways (Fig. 36-1). Given the shorter acquisition times, correct scan timing after intravenous (IV) contrast medium administration is mandatory to achieve adequate enhancement of the organ or vessels within the area of

929

interest. Because of different MDCT scanners, the protocols need to be adjusted to match the suggested iodine doses and administration rates. A small to moderate amount of contrast medium (60-120 mL of 300 mg iodine[I]/mL) delivered at a slow injection rate (1.5-3 mL/sec), scan delay time of 25 seconds, and long injection duration (>30 seconds) results in sufficient opacification of thoracic vessels. Saline flushing is favorable to reduce perivenous artifacts. Moderate to large amounts of iodine (3-4 mL/sec) are required to study diverse abnormalities of pulmonary arteries such as acute pulmonary embolism (PE), chronic thromboembolic disease, pulmonary arteriovenous malformations (AVMs), pulmonary artery aneurysms, and arteriovenous fistulas. A slower injection is recommended in those patients with diminished cardiac output. For fast acquisitions (16- and 64-channel MDCT), thoracic and coronary CT angiography require moderate amounts of iodine at high iodine administration rates. The value of MDCT for evaluating patients with chest pain presenting to the emergency department has also been recently discussed.288,592,636,637,639 In addition to evaluation of the coronary arteries, CT angiography has long been used to evaluate patients for other dangerous causes of chest pain such as aortic dissection and pulmonary embolus (“triple rule-out”).554

Dual-Energy Computed Tomography Dual-energy computed tomography (DECT) is very helpful in the detection and diagnosis of thoracic abnormalities.325 Selective iodine mapping is made possible by the x-ray absorption characteristics of iodine, which has a high atomic number compared with elements that constitute air or soft tissue.609 A major advantage of dual-energy imaging over conventional radiography is its superior sensitivity for the detection of calcification within a pulmonary nodule.95 Other advantages of dual-energy imaging are related to identification of bone and pleural abnormalities, recognition of hilar and mediastinal masses, detection of tracheal narrowing and airway disease, and localization of stents, catheters, and other indwelling devices. DECT can also be used for visualization  of pulmonary microvascular contrast material distribution (DECT angiography) and detection of PE609 (Fig. 36-2).

Magnetic Resonance Imaging Despite the fact that magnetic resonance imaging (MRI) has a clear advantage over other imaging modalities in evaluating cardiac and vascular disorders of the chest, it continues to play a limited role in parenchymal lung evaluation. Reasons for the limited use of pulmonary MRI include respiratory motion and magnetic susceptibility effects caused by air-tissue interfaces. Assessment of chest wall disease and peridiaphragmatic processes, which do not suffer as much from respiratory or magnetic susceptibility artifacts, already benefits from the multiplanar capabilities of MRI and its improved soft tissue contrast. Hyperpolarized gas MRI of the lung provides high temporal and spatial resolution images of the airspaces of the lung and can be used to elucidate both lung ventilation and morphology. In patients with asthma and cystic fibrosis, areas of diminished gas ventilation from airflow obstruction are readily demonstrated and visualized as ventilation defects23,151 (Fig. 36-3).

FOCAL LUNG DISEASE A large number of nonneoplastic lung diseases may present with a solitary or multiple pulmonary nodules or masses. These include congenital abnormalities, inflammatory diseases, pulmonary infections, vascular pulmonary lesions, and miscellaneous other conditions. Chest radiography serves as the primary technique for detecting most focal

http://radiologyebook.com

930

PART II  CT and MR Imaging of the Whole Body

A

B

10 cm

C

D FIG 36-1  Normal CT anatomy of the tracheobronchial tree. A, Coronal MPR volume-rendered image through the carina shows a normal tracheobronchial tree. Coronal (B) and axial (C) MIP images show a normal pulmonary vasculature. MIP enhances visualization of the pulmonary vessels. D, Coronal MPR minimum intensity projection (MiniP) image shows a normal tracheobronchial tree. Images were reconstructed using 5 contiguous 1-mm sections obtained from a 16-detector CT scanner using 1-mm collimators.

lung disease. A major challenge of lung imaging is to differentiate common and benign focal abnormalities from malignant tumors such as metastases or bronchogenic carcinoma. Although MDCT and fluorine-18 fluorodeoxyglucose positron emission tomography (FDG PET) have changed the evaluation of focal lung disease, optimal radiographic evaluation and management of these cases remain challenging clinical problems.

Congenital Pulmonary Arteriovenous Malformations Congenital pulmonary AVMs are fistulous vascular communications between a pulmonary artery and vein (95%) or systemic artery and pulmonary vein, bypassing the capillary bed. About 50% of pulmonary AVMs are associated with Osler-Weber-Rendu syndrome (hereditary hemorrhagic telangiectasia). The diagnosis is usually straightforward on CT.341,342,638 The typical features are a smooth, lobulated, oval/round mass, or even serpiginous

nodule, less than 1 cm to several centimeters in size, associated with a visible enlarged feeding artery and an enlarged draining vein; the feeding artery is usually less than half the diameter of the nodule (Fig. 36-4). Complex AVMs having more than one feeding artery are rare.508,509 CT pulmonary angiography performed after a bolus injection of contrast material can be used to confirm the vascular nature of the lesion. Multiple AVMs can be confused with metastases if the enlarged feeding vessels are overlooked. MIP reconstructions can also be obtained to scrutinize the pulmonary parenchyma for malformations and depict the size and number of the supplying artery(ies)638 (Fig. 36-5). Pulmonary AVMs can be successfully treated with embolotherapy80,511 (Fig. 36-6).

Inflammatory Disease

Solitary Pulmonary Nodule.  The incidental finding of a pulmonary nodule on chest radiograph or CT is becoming an increasingly

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease frequent event.645 A solitary pulmonary nodule has been defined as a spherical intraparenchymal opacity 3 cm or less in diameter, completely surrounded by lung, without associated mediastinal lymphadenopathy or atelectasis.244 Lesions larger than 3 cm in diameter usually are defined as masses, and most represent bronchogenic carcinoma. The goals of imaging techniques in evaluating an undiagnosed pulmonary nodule are: (1) to make a specific diagnosis and (2) if a specific diagnosis cannot be made, to distinguish benign nodules from those

931

that are indeterminate or potentially malignant. In the early stage, bronchogenic carcinoma may also appear as a small nodule (70% in stage I). In patients younger than 30 years, the incidence of lung cancer is very low, so a solitary pulmonary nodule may not require further evaluation unless the patient has a known extrathoracic malignancy. New guidelines have been proposed by the Fleischner Society for follow-up and management of small pulmonary nodules detected on CT scans.378 Despite high false-positive rates, computer-aided detection has been shown to be useful in the detection of lung nodules on chest radiographs.45 Bone suppression on conventional chest radiographs will be a helpful technique in detecting more pulmonary nodules  by minimizing or eliminating the overprojection of the clavicles  and ribs.600 Enhancement of lung nodules.  Because of differences in the vascularity of benign and malignant nodules, noncalcified nodules may show some degree of enhancement after administration of IV contrast

FIG 36-2  Pulmonary embolism in a 62-year-old woman with deep venous thrombosis, developing severe dyspnea. CT image in angiographic window with superimposed color-coded DE perfusion map. Bilateral perfusion defects in the upper lobes due to acute PE are depicted in the color-coded maps by the lack of color coding (arrows). (Courtesy Marcelo Sanchez, MD, Barcelona, Spain.)

A

FIG 36-4  Simple pulmonary AVM. CT scan at the level of the lung bases shows a well-defined smooth round nodule. Note that the feeding vessel is about half the diameter of the fistula.

B FIG 36-3  Cystic fibrosis. A and B, 3He MRI coronal images show major ventilation defects in both upper lobes in addition to ventilation defects in the right middle lung zone (arrows).

http://radiologyebook.com

932

PART II  CT and MR Imaging of the Whole Body

material (Fig. 36-7). Malignant nodules tend to enhance more than benign inactive granulomas, which show very little or no enhancement. However, inflammatory nodules such as active granulomas resulting from fungal infection or tuberculosis (TB) may also show significant contrast enhancement.599,657 Although false-negative results have been reported, absence of significant lung nodule enhancement (≤15 HU) at CT is strongly predictive of benignity.599

Granuloma.  Postinflammatory granulomas due to TB or fungal infections are the most common cause of pulmonary nodules. Thinslice section CT is much more sensitive than conventional radiography for the detection of calcification within a nodule. On unenhanced CT,

a well-circumscribed calcified pulmonary nodule measuring 2 cm or less in diameter is usually a benign calcified granuloma270,306,569,570,598,599,656 (Fig. 36-8). The CT pattern of calcification in a solitary pulmonary nodule can help differentiate benign from malignant lesions. Patterns for benignity are homogeneous calcification throughout the nodule, laminated or concentric rings of calcification, and a “popcorn-like” appearance. Exception occurs in the specific clinical setting of metastatic osteogenic sarcoma; lung metastases from osteogenic sarcoma are typically calcified in a homogeneous pattern. Although lung cancers and other malignancies may contain small areas of calcification, they rarely demonstrate calcification in one of the benign patterns described previously.

Hamartoma and Other Fat-Containing Nodules.  Focal areas of fat

within a nodule with attenuation values ranging from −40 to −120 HU, sometimes associated with chondroid (“popcorn”) calcification, are a reliable indicator of a hamartoma (Fig. 36-9). In patients with chronic rhinitis and a history of mineral oil or oily nose drop use, the presence of focal areas of fat within a pulmonary mass or focal consolidation strongly suggest exogenous lipoid pneumonia (Fig. 36-10).

Vascular Lesions

Vessel-Related Disorders.  The feeding vessel sign is defined as a vessel coursing directly into a nodule. This sign indicates either that the lesion has a hematogenous origin or that the disease process occurs in close proximity to small pulmonary vessels. A number of vesselrelated nonneoplastic disorders of the lung produce focal disease, including pulmonary infarcts, septic emboli, and pulmonary vasculitis.8,44,158,267,280,330,337,402

Pulmonary Infarction.  Pulmonary infarction is associated with FIG 36-5  Multiple bilateral pulmonary AVMs. Noncontrast CT scan (lung window) at the level of lung bases shows multiple AVMs (arrows).

A

thromboembolic obstruction of a medium-sized pulmonary artery and implies necrosis of lung tissue.627 Classic radiographic findings

B

FIG 36-6  Complex AVM. A, Pretreatment right pulmonary angiogram shows simple angioarchitecture of complex AVM in the medial segment of the right lower lobe (arrows). B, Follow-up angiogram obtained after deposition of several coils (arrowhead) in large feeding artery confirm cessation of blood flow within the malformation (arrows). (Courtesy Jordi Villalba, MD, Barcelona, Spain.)

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

933

B

FIG 36-7  Nodule enhancement. A, Preenhancement CT scan shows a homogeneous noncalcified welldefined nodule in the right lower lobe, with a density of 49 HU (arrows). B, After administration of contrast material the nodule showed enhanced density (113 HU). This finding is highly consistent with malignancy.

FIG 36-9  Hamartoma. CT scan (mediastinal window) shows a sharply well-defined lobulated nodule showing typical popcorn calcifications in the center of the nodule (arrow).

FIG 36-8  Calcified granuloma. Nonenhanced CT scan (mediastinal window) shows a sharply well-defined calcified nodule in the right upper lobe (arrow).

include Westermark’s sign, an area of focal hypoperfusion, and Hampton’s hump, a peripheral wedge-shaped density above the diaphragm. Evidence of a dilated pulmonary artery is occasionally observed. CT findings have been reported in only 10% of cases of documented PE and include segmental and/or subsegmental atelectasis and focal illdefined ground-glass attenuation or parenchymal consolidation, often peripheral and triangular in distribution (Fig. 36-11). The reversed halo sign, characterized by a central ground-glass opacity surrounded by

denser airspace consolidation, was first described on HRCT as being specific for cryptogenic organizing pneumonia. Since then, it has been reported in association with a wide range of pulmonary diseases, including pulmonary infarction.221 Infarcts may completely resolve or persist over a period of weeks, resulting in a residual scar that may be mistaken for a pulmonary nodule or mass.117

Septic Embolism.  Septic PE generally presents with insidious onset of fever, cough, and pulmonary opacities. It is seen most commonly in patients with indwelling catheters, tricuspid valve endocarditis, alcoholism, skin infection, and in IV drug users; less common causes include pelvic thrombophlebitis, periodontal disease, and suppurative processes in the head and neck.108,158,280,566 Septic PE from the lesions

http://radiologyebook.com

934

PART II  CT and MR Imaging of the Whole Body

FIG 36-10  Exogenous lipoid pneumonia mimicking lung cancer. CT scan (lung window) shows a spiculated nodule in the medial segment of the right lower lobe, highly suspicious for lung carcinoma (arrow). Histologically an exogenous lipoid pneumonia with a high fibrotic component was demonstrated.

FIG 36-12  Septic emboli in a patient with endocarditis. Axial CT shows multiple nodules with cavities. Note that multiple feeding vessel are also seen (arrows).

for confirming the presence of pulmonary septic emboli even when conventional chest radiographs remain negative. The CT features of septic PE include multiple peripheral nodules varying in size from 0.5 to 3.5 cm (85%), a feeding vessel sign (60%-70%), cavitation (50%), wedge-shaped peripheral lesions abutting on the pleura (50%), and extension into the pleural space (39%)254 (Fig. 36-12). The nodules may be circumscribed or poorly defined and may be associated with patchy areas of consolidation.158,330 MPRs show that most of these vessels course around the nodule and that the others are pulmonary veins.158

Angioinvasive Pulmonary Aspergillosis.  Angioinvasive pulmo-

FIG 36-11  Pulmonary infarction in a patient with pulmonary embolism. CT section at the base of the left lower lobe shows various nonenhanced wedge-shaped opacities in the peripheral lung. Note that the opacities did not enhance after contrast administration.

of periodontitis should be included in the differential diagnosis of septic PE.254 The radiographic manifestations usually consist of bilateral nodular opacities, variable in size, which are frequently cavitated. Septic embolism is often complicated with empyema. CT is an important modality

nary aspergillosis,642 generally occurring in the neutropenic phase, is the most common opportunistic infection of respiratory complications following bone marrow transplantation. Angioinvasive pulmonary aspergillosis is associated with high morbidity and mortality, reaching 50% to 60% in severely immunocompromised patients. Major risk factors for the disease include severe or prolonged neutropenia (absolute neutrophil count < 500), prolonged corticosteroid therapy, graft-versus-host disease after allogeneic bone marrow transplantation, and late-stage AIDS.212,215,329 The diagnosis of angioinvasive aspergillosis is based on clinical, radiologic, and mycologic data. Affected patients report nonspecific symptoms of fever, cough, and dyspnea. The clinical diagnosis is very difficult owing to its nonspecific symptoms, low frequency of positive blood cultures, and high frequency of contamination. Thrombocytopenia of these patients may preclude invasive diagnostic procedures such as percutaneous or transbronchial biopsy. Cultures of bronchoalveolar lavage (BAL) fluid are positive in 30% to 68% of infected patients. Galactomannan and nucleic acid detection in serum or BAL fluid are useful in the early identification of invasive aspergillosis in the immunocompromised host; however, a definite diagnosis of invasive aspergillosis still requires demonstration of the fungus in tissue specimens.380

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

935

B FIG 36-13  Angioinvasive aspergillosis. A, PA radiograph shows multiple nodules in the lungs (arrows). B, CT section at the level of the intermediary bronchus shows a nodule surrounded by a halo of ground-glass attenuation (arrows).

Radiographically, angioinvasive pulmonary aspergillosis characteristically manifests as multiple ill-defined 1- to 3-cm peripheral nodules that gradually coalesce into larger masses or areas of consolidation. An early CT finding (best seen on thin-section images) is a rim of groundglass opacity surrounding the nodules (CT halo sign)329 (Fig. 36-13). The halo represents hemorrhage around a focal area of lung infarction caused by the Aspergillus organism invading pulmonary vessels. In the nonleukemic nonneutropenic patient, the CT halo sign is less specific. This finding has also been described in patients with other types of angioinvasive infections, hemorrhagic metastases, TB, mucormycosis, and Wegener’s granulomatosis.495 Cavitations are usually a late finding in these patients and represent the resolution phase of the infection. As the bone marrow recovers from aplasia and the white blood cell count returns to normal, the pulmonary lesions begin to cavitate and often have a distinctive radiographic appearance, the air crescent sign215,329,361 (Fig. 36-14). This finding, resulting from an intracavitary mass composed of  invasive pulmonary aspergillosis and necrotic lung, characteristically occurs 2 to 3 weeks after initiation of treatment and concomitant with resolution of the neutropenia; it usually indicates a good prognosis.162,215,329,643

FIG 36-14  Angioinvasive aspergillosis. CT section at the level of the lower trachea shows a consolidation with an eccentric cavitation and air crescent sign (arrows). This finding in this neutropenic patient is highly diagnostic of angioinvasive aspergillosis.

Pulmonary Vasculitis.  The vasculitides are disorders characterized by blood vessel inflammation leading to tissue or end-organ injury. In the Chapel Hill classification, large vessels refers to the aorta and its largest branches, medium-sized vessels to the main visceral arteries, and small vessels to the capillaries, venules, and arterioles.284 Thoracic involvement is most commonly seen with primary idiopathic largevessel vasculitides (Takayasu’s arteritis and giant cell arteritis) and primary antineutrophil cytoplasmic autoantibody (ANCA)-associated small-vessel vasculitides (Wegener’s granulomatosis, microscopic polyangiitis, Churg-Strauss syndrome).91 The radiologic findings in primary pulmonary vasculitis vary widely and can include vessel wall thickening, nodular or cavitary lesions, ground-glass opacities, and consolidations, among others. Diffuse alveolar hemorrhage usually results from primary small-vessel vasculitis, specifically microscopic polyangiitis and Wegener’s granulomatosis.91 Wegener’s granulomatosis is a multisystem disease characterized clinicopathologically by the combination of a granulomatous

necrotizing and ulcerative process of the respiratory tract and the internal organs, generalized necrotizing granulomatous vasculitis, and a usually focal necrotizing glomerulonephritis.112 A limited (nonrenal) form of Wegener’s granulomatosis affecting only the respiratory tract may also occur. Active vasculitis associated with Wegener’s granulomatosis may cause abnormalities of pulmonary perfusion at the subsegmental level, resulting in ischemia or frank infarctions, leading to parenchymal scars. The characteristic CT manifestations consist of bilateral subpleural or peribronchovascular nodules or masses (1-4 cm in diameter), sometimes spiculated (Fig. 36-15); in approximately 50% of cases, masses may cavitate.243,322,347,363,364,494 Other common CT features are diffuse bilateral ground-glass opacities, bilateral nonsegmental or peribronchovascular areas of airspace consolidation, peribronchial thickening, and small centrilobular nodules and branching linear structures (so-called tree-in-bud pattern).243,397 Ground-glass opacities and airspace consolidation correlate with focal or diffuse pulmonary

http://radiologyebook.com

936

PART II  CT and MR Imaging of the Whole Body

FIG 36-15  Wegener’s granulomatosis. CT scan shows bilateral spiculated nodules. Spiculation is uncommonly found in this disease.

hemorrhage (DPH). This occurs when pulmonary capillaries are primarily involved by the vasculitic process.389,397

Pulmonary Embolism.  PE is a common condition, representing the third most lethal disease in the United States. It results from a thrombus formed in the venous system. PE is a frequently missed clinical diagnosis that has absent or nonspecific radiographic findings.32,457,538 Chest radiography is of limited value in the diagnosis; radiographic findings include elevation of the diaphragm, subsegmental atelectasis, patchy opacities (due to associated edema, hemorrhage, or infarction), pleural effusion, and segmental or larger areas of oligemia. Radiographs are normal in 10% to 15% of cases with a proven diagnosis; their major importance lies in excluding other disease processes that can mimic PE (e.g., pneumonia, pneumothorax) and for interpretation of V/Q scintigraphy.117,118,548-551 Compared with single-detector CT scanners used in the 1990s, a new generation of CT scanners with multiple detector rows has dramatically changed the diagnostic accuracy of PE.216,538,549,568,622 MDCT has a high degree of accuracy in identifying central pulmonary emboli down to the second-order to fourth-order pulmonary arteries. Initial reports pointed out the suboptimal detection of PE in subsegmental arteries. Actually the use of narrow slice reconstruction has led to progressive improvement in visualization of subsegmental arteries. Numerous studies have corroborated significantly improved detection of subsegmental emboli (75%-95%) using MDCT, with overall sensitivities and specificities ranging between 90% and 100%, respectively.117,118,548,549-551 Advanced image processing techniques, such as MPR or MIP reconstructed from isotropic volumetric data sets, also facilitate the demonstration of subsegmental pulmonary emboli. Suboptimal visualization of peripheral emboli in subsegmental arteries will be of potential clinical significance only when embolic disease is limited to vessels of this caliber.549,551 The accuracy of MDCT for diagnosing PE depends on the size of the artery affected and the size of the emboli. Optimal assessment of the pulmonary vessels on MDCT requires careful attention to several parameters, including scan collimation, imaging volume, and contrast enhancement.488,551,568 The lung volume that is scanned should be large enough to include all segmental and subsegmental pulmonary arteries. This goal can be achieved by scanning from the top of the aortic arch to the dome of the diaphragm. Respiratory motion exacerbates partial volume artifacts of arteries in or near the scan plane, leading to loss of continuity of arteries in

FIG 36-16  Acute pulmonary embolism. Magnified coronal contrastenhanced CT reconstruction through the proximal arteries shows filling defect within the right lower lobe segmental branch indicative of pulmonary embolus (arrow).

the z-axis and resulting in interpretative difficulty. Although the scans can be performed in the craniocaudal direction, scanning in the caudal-to-cranial direction helps minimize motion artifacts (respiratory motion is most marked in the lung bases), particularly in patients unable to hold their breath for the duration of the scan. The current generation of MDCT devices (up to 64 detectors) can obtain thin sections up to 0.5 mm with a resolution down to 0.4 × 0.4 × 0.4 mm; thin sections and a high spatial resolution improve the depiction of small peripheral emboli down to sixth-order branches (subsegmental pulmonary emboli).218,219 A basic understanding of contrast medium dynamics is crucial for the rational design of contrast medium injection. Incorrect timing of image acquisition after the contrast bolus injection is a common cause of suboptimal studies. The diagnosis of acute PE on contrast-enhanced CT scan is based on direct visualization of partial or complete filling defects within the pulmonary arteries (Figs. 36-16 to 36-18). Whereas a central thrombus when seen in cross section shows a round intraluminal filling defect surrounded by contrast (the “doughnut” sign), when the thrombus is imaged along its axis shows a linear intraluminal filling defect outlined by contrast (the “railroad track” sign). Other useful CT signs of acute PE include acute angles with the vessel wall, complete cutoff of vascular opacification, and increased diameter of the occluded vessel.64,226,271,499 Occasionally, hyperdense pulmonary clots can be seen in nonenhanced CT scans.302 Chronic thrombi often appear as crescentic or laminar filling defects adherent to the walls of the pulmonary artery.64,315,516,653 Calcification may be identified in chronic clots in patients with longstanding pulmonary hypertension315 (Fig. 36-19). Pitfalls in the diagnosis of PE may include poor arterial opacification caused by

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

937

B FIG 36-17  Acute pulmonary embolism. Contrast-enhanced CT section through the right and left pulmonary arteries shows massive bilateral pulmonary emboli (arrows). (Courtesy Martine Remy-Jardin, MD, Lille, France).

FIG 36-18  Acute pulmonary embolism: saddle emboli. Contrastenhanced CT sections through the right and left pulmonary arteries shows a saddle embolus bridging the main pulmonary arteries (arrowheads).

incorrect bolus timing and the presence of nonopacified blood within the pulmonary vasculature arriving from the inferior vena cava and right atrium (transient interruption of contrast).225 The lymphatic and connective tissue located adjacent to the pulmonary arteries may also mimic the appearance of PE.

Ancillary CT findings that may suggest PE include pleural-based wedge-shaped images, linear parenchymal bands, and central pulmonary arterial enlargement.117 Although a triangular pleural-based parenchymal image may represent a pulmonary infarct, this finding is not specific, and other entities such as pneumonia, tumor, fibrosis, hemorrhage, or edema should also be considered.559 Pleural effusion is common but does not occur significantly more frequently in the presence of PE. In chronic thromboembolic disease, the lung parenchyma may demonstrate mosaic attenuation, with areas of decreased attenuation associated with “pruned” or attenuated vessels.315 Nonthrombotic pulmonary embolism.  Nonthrombotic PE is an uncommon condition but is sometimes associated with specific imaging findings, including discrete nodules with cavitation (septic embolism), widespread homogeneous and heterogeneous areas of increased opacity or attenuation that typically appear 12 to 24 hours after trauma (fat embolism), fine miliary nodules that subsequently coalesce into large areas of increased opacity or attenuation (talcosis), or opaque tubular and branching structures (methyl-methacrylate PE after percutaneous vertebroplasty).104,238 Fat embolism secondary to long bone fractures occurs in 1% to 3% of patients with simple tibial or femoral fractures but in up to 20% of individuals with more severe trauma. Imaging findings are similar to those in acute respiratory distress syndrome (ARDS).29,209,261,381,500,618 The radiographic findings resemble those in ARDS. Hydatid PE is a rare complication of cardiac or hepatic echinococcosis.196 Pulmonary tumor embolism.  Pulmonary tumor embolism and pulmonary tumor thrombotic microangiopathy are uncommon manifestations of neoplasms.191 Although pulmonary intravascular tumor emboli are seen in up to 26% of autopsies, its diagnosis is difficult to establish both clinically and on conventional radiographic studies.528,552 The major CT findings consist of multiple multifocal dilatation and beading of peripheral pulmonary arteries, primarily in a subsegmental distribution.486,563 A vascular cause of the tree-in-bud pattern on thin-section CT has been described in a patient with pulmonary tumor thrombotic

http://radiologyebook.com

938

PART II  CT and MR Imaging of the Whole Body

A

B FIG 36-19  Chronic pulmonary embolism. A, Coned-down noncontrast CT through the right main pulmonary artery. Note that the main pulmonary artery is enlarged (arrows), and its appearance is heterogeneous and contains irregular calcifications (arrowheads). B, Contrast-enhanced CT image at the same level as shown in A demonstrates a large organized thrombus along the lateral walls of the right pulmonary artery (arrows). Note linear calcification visible within the organized thrombus (arrowheads).

microangiopathy (PTTM)191 (Fig. 36-20). Histopathologically, PTTM is characterized by widespread fibrocellular intimal hyperplasia of small pulmonary arteries (so-called carcinomatous endarteritis) induced by tumor microemboli.191

Other Focal Lesions

Rounded Atelectasis.  Rounded atelectasis refers to a form of nonsegmental peripheral pulmonary collapse that may mimic a primary lung cancer.239 It is always found adjacent to an area of pleural thickening, and the process is thought to begin with an adhesive pleural effusion or pleural reaction that entraps the adjacent atelectatic lung.585 Although many cases of rounded atelectasis are the result of asbestosrelated pleural disease, other causes of pleural reaction may produce similar findings.415 Usually this pulmonary pseudotumor can be differentiated from malignancy using CT criteria such as: (1) a rounded or wedge-shaped mass that forms an acute angle with the adjacent thickened pleura, (2) vessels and bronchi emanating from the hilar region, swirling around and converging in a curvilinear fashion into the lower border of the mass, forming a comet’s tail or vacuum cleaner sign, (3) air bronchogram visible in the central portion of the mass, (4) homogeneous contrast enhancement of the atelectatic lung, and (5) volume loss in the ipsilateral hemithorax159,368,411 (Fig. 36-21). However, percutaneous needle biopsy, radiologic follow-up, or surgical resection are required for indeterminate cases.

DIFFUSE LUNG DISEASE DLD may result from a variety of acute and chronic causes, including ILD, airspace processes, airway diseases, and emphysema. There are over 200 clinical forms of DLD of known causes (e.g., collagen vascular disease, environmental or drug related), as well as disorders of unknown cause. The latter include idiopathic interstitial pneumonias (IIPs),

granulomatous lung disorders (e.g., sarcoidosis), and other forms of ILD, including lymphangioleiomyomatosis (LAM), Langerhans cell histiocytosis (LCH), and eosinophilic pneumonia. HRCT (1- to 2-mm collimation) is more sensitive than both conventional thick-section CT and chest radiography for the detection and characterization of ILD. With the advent of MDCT scanners it is also feasible to perform volumetric HRCT through the entire chest; this approach allows assessment of the entire lung parenchyma and therefore minimizes the risk of a false-negative examination. In addition, volumetric HRCT allows use of advanced imaging techniques such as MPRs and MIPs or MiniPs. There is no significant difference in overall image quality between volumetric HRCT and standard axial HRCT techniques.265,510,548 Follow-up CT studies can be performed by using low-dose technique whenever feasible.

Basic CT Anatomy of the Lung Parenchyma A basic knowledge of the anatomy of the interstitium and secondary lobule is essential for understanding pathologic alterations in DLDs and their radiographic appearances. In general, DLD can be divided into disease involving primarily the airspace, interstitium, or small airways.

Lung Attenuation.  Attenuation (density) on a CT scan is expressed in terms of Hounsfield units (HU); under normal circumstances, water is 0 HU and air is −1000 HU. CT scans of the lung are usually obtained during maintained full inspiration. Normally the attenuation of lung parenchyma is determined by its relative proportions of blood, gas, extravascular fluid, and pulmonary tissue. The normal lung parenchyma has a fairly homogeneous attenuation that is slightly greater than that of air. As lung gas volume is reduced, lung attenuation increases. A significant increase in attenuation of the lung parenchyma is seen on expiratory scans. Attenuation values are variable and

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

939

B

FIG 36-20  Pulmonary tumor thrombotic microangiopathy. A and B, Transverse thin-section CT images through the lower lobes shows small branching opacities (arrows) representing enlarged peripheral arteries. (From Franquet T et al: Thrombotic microangiopathy of pulmonary tumors: A vascular cause of tree-in-bud pattern on CT. AJR Am J Roentgenol 179:897–899, 2002.)

Secondary Pulmonary Lobule.  The secondary pulmonary lobule is

FIG 36-21  Rounded atelectasis. CT shows a rounded opacity in the peripheral portion of the right middle lobe extending to the pleural surface. Note the curvilinear configuration of the proximal vessels (comet tail sign) (arrows). Note the pleural thickening adjacent to atelectasis.

influenced by the pulmonary volume and different parameters such as kilovoltage, patient size, and the particular region of the lung being assessed. Appearance of CT images on the monitor or film (hard copy) depends on the display parameters used (window level ands window width) and is limited to 256 shades of gray. For clinical practice a window level of −600 to −700 HU and a window width of 1000 to 1500 HU are recommended. Window levels of 30 to 50 HU and window widths of 350 to 500 HU are the recommended parameters to evaluate mediastinum, hila, and pleura.

a fundamental unit of lung structure defined as the smallest portion of the lung that is surrounded by connective tissue septa, made up of 3 to 12 acini, and supplied by 3 to 5 terminal bronchioles.633 A pulmonary acinus is defined as the portion of the lung parenchyma distal to a terminal bronchiole and supplied by a first-order respiratory bronchiole or bronchioles and comprising respiratory bronchioles, alveolar ducts, alveolar sacs, and alveoli.506 Secondary pulmonary lobules are irregularly polyhedral in shape and somewhat variable in size (1-2.5 cm in length).632 An understanding of lobular anatomy is essential to interpretation of thin-section CT of the lung.65 The most typical secondary pulmonary lobules are recognized in the outermost part of the lung, where each lobule is bounded by the visceral pleura, interlobular septa, and pulmonary veins.258 The secondary lobule of the lung is separated incompletely by the interlobular septa from adjacent lobules.65,258,631 Structures in the centrilobular region include terminal and proximal respiratory bronchioles, accompanying pulmonary arteries, interstitial tissue enveloping these bronchovascular structures, and immediately adjacent alveolar ducts, sacs, and alveoli. The smallest pulmonary arteries and accompanying bronchioles are situated in the centrilobular area, but bronchioles cannot be seen in normal conditions because they are no more than 1 mm in diameter and their walls are less than 0.1 mm thick. Lambert’s canal connects airspace to bronchioles. The number of terminal bronchioles included in one lobule varies from 3 to approximately 30, depending upon the size of the lobule. The interlobular septa define the margins of the secondary pulmonary lobule, whose sides measure approximately 1 to 2.5 cm in length and contain pulmonary venules and lymphatics.630,631 Normal intralobular bronchioles and interlobular septa usually are not detected on CT.630 Their presence usually indicates abnormal thickening of the bronchiolar walls or septal interstitium.

http://radiologyebook.com

940

PART II  CT and MR Imaging of the Whole Body

CT Patterns of Abnormality and Anatomic Distribution The imaging findings associated with DLDs have been extensively described. A number of studies have shown that HRCT can demonstrate the presence, distribution, and extent of parenchymal abnormalities, reflecting the pathologic findings closely. The best approach to diagnosing DLD is recognition and analysis of the different patterns. In an attempt to arrive at a differential diagnosis, DLDs can be further characterized into two broad categories based on their CT appearances: (1) increased lung attenuation, including linear and reticular opacities, nodular opacities, ground-glass opacity, and airspace consolidation, and (2) decreased lung opacity, including cysts, cavities, emphysema, and mosaic perfusion.

Increased Lung Attenuation Linear and reticular opacities.  Interlobular septum marginates part of a secondary pulmonary lobule. Normally only a few septa  are seen.633 Thickening of interlobular septa (septal lines) is commonly seen in patients with a variety of ILDs and may be smooth, irregular, or nodular.630 Thickened septa can result from fibrosis, interstitial fluid, and infiltration by cells or other material.438,587 Septal lines may be smooth, nodular, or irregular. Smooth septal thickening is commonly seen in patients with pulmonary edema, hemorrhage, lymphatic spread of carcinoma, and amyloidosis; less common causes include lymphoma, leukemia, Churg-Strauss syndrome, acute lung rejection, lymphangiectasia, and metabolic storage diseases such as Niemann-Pick and Gaucher’s diseases. Irregular septal thickening, often accompanied by distortion of lung architecture, is characteristic of diseases causing interstitial pulmonary fibrosis. Nodular or beaded thickening of the interlobular septa (beaded septum sign) occurs in sarcoidosis, lymphangitic carcinomatosis, amyloidosis, and occasionally silicosis. As noted earlier, the combination of ground-glass opacity with a superimposed linear pattern is termed crazy paving.343 Although initially described as typical of alveolar proteinosis, once again, this appearance is nonspecific and may be seen in a variety  of acute (e.g., edema, hemorrhage, P. jiroveci pneumonia, ARDS) or chronic (alveolar proteinosis, exogenous lipoid pneumonia) lung diseases. A reticular pattern (intralobular linear opacities) represents thickening of the intralobular interstitium within the secondary pulmonary lobule.431,433 Although intralobular interstitial thickening is a common HRCT finding in patients with usual interstitial pneumonia (UIP) it may also be seen in patients with nonspecific interstitial pneumonia (NSIP). Intralobular interstitial thickening usually indicate fibrosis; however, it can also be due to infiltration of such structures by infection, malignant cells (lymphangitic carcinomatosis), pulmonary edema, an abnormal protein (alveolar proteinosis), or simply edema fluid from cardiac failure.62,129,185,294,431,433 Nodules and nodular opacities.  A nodular pattern refers to multiple round opacities ranging in diameter from 1 to 10 mm. Nodules can be characterized according to their margins (e.g., well defined or poorly defined), presence of associated findings, and distribution. Three main types of distribution can be identified on HRCT: centrilobular, perilymphatic, and random. Multiple small smooth or irregularly marginated nodules in a perilymphatic distribution are characteristic of sarcoidosis, lymphangitic carcinomatosis, silicosis, and coal worker’s pneumoconiosis (CWP).79,213,386,438,587 Centrilobular nodules are characterized on HRCT by their location within the secondary pulmonary lobule several millimeters away from the pleura, interlobar fissures, and interlobular septa. Morphologically, centrilobular nodules may have either a poorly defined ground-glass density or a tree-in-bud configuration; the tree-in-bud pattern is characteris-

tic of acute cellular bronchiolitis, representing dilated centrilobular bronchioles impacted with mucus, fluid, and/or pus and usually  associated with a peribronchiolar inflammation.26,123,163,194,272,537 Predominantly numerous small nodules of ground-glass opacity in a centrilobular distribution are characteristic of the subacute stage of extrinsic allergic alveolitis or respiratory bronchiolitis, and rarely cryptogenic organizing pneumonia (COP). Tree-in-bud opacities are characteristic of infectious bronchiolitis, bronchopneumonia, and endobronchial spread of TB or Mycobacterium avium-intracellulare complex (MAC) infection. 273,537 Miliary nodules as small as 1 mm in diameter can be seen with hematogenous spread of TB, fungal infection, or metastases.25,272 Nodules larger than 1 cm in diameter may result from conglomeration of small nodules, as occurs in sarcoidosis, silicosis, or talcosis.28,245,367,470,473 Larger nodules usually result from metastases, septic emboli, or Wegener’s granulomatosis. In immunocompromised patients, nodules surrounded with a halo of ground-glass attenuation are caused by invasive pulmonary aspergillosis; small nodules with halo have been described in diverse viral infections.205,227,329,495 Ground-glass opacity.  Ground-glass opacity is defined as a hazy increased attenuation of lung, with preservation of bronchial and vascular margins.244 The significance of ground-glass opacity depends on the clinical scenario. This finding can be made with HRCT while plain films are still negative. Ground-glass opacity may be caused by partial filling of airspaces, interstitial thickening, partial collapse of alveoli, normal expiration, or increased capillary blood volume. It should not be confused with consolidation, in which bronchovascular margins are obscured and may be associated with an air bronchogram. Because it is a nonspecific CT finding, a large number of diseases can be associated with this pattern.124,469,513,561 In acute DLD, ground-glass attenuation as a main pattern is suggestive of acute interstitial pneumonia, acute or subacute hypersensitivity pneumonitis (HP), pulmonary edema, pulmonary hemorrhage, drug-induced disease, and Pneumocystis jiroveci pneumonia in patients with AIDS.222,332,423,517,535 The most common causes of ground-glass opacity in chronic DLD include NSIP, desquamative interstitial pneumonia (DIP), and respiratory bronchiolitis–associated interstitial lung disease (RBILD).355,519 A “crazy-paving” pattern represents superimposition of a linear pattern in a background of ground-glass opacity.343,442 This pattern, initially described in alveolar proteinosis, can also be seen in a variety of pulmonary disorders including bronchioloalveolar cell carcinoma and exogenous lipoid pneumonia.119,189,213,292,343,441,536,597,602 Consolidation.  The term airspace consolidation is used when an increase in lung density stems from replacement of the air within the pulmonary acinus by liquid, blood, cells, or a combination of these.279,355 Occasionally, it may result from replacement of air in the alveoli by extensive interstitial disease as occur in patients with sarcoidosis (“alveolar sarcoidosis”) and NSIP.78,452,536,654,655 Airspace diseases manifest different CT appearances depending on the degree of involvement. The earliest finding, occurring when single acini are replaced by liquid or tissue, is poorly marginated small nodules approximately 7 mm in diameter, called acinar nodules.416 These are not true nodules but represent clusters of single acini filled with fluid, inflammatory cells, or blood, depending on the disease process. As adjacent acini become involved, acinar nodules coalesce and may progress to lobular, subsegmental, segmental, lobar, or multilobar areas of consolidation. On HRCT, unlike ground-glass opacity, consolidation results in an increase in lung opacity, with obscuration of underlying vessels.452,654,655 Air bronchograms are often visible if the bronchi are patent. Common diffuse diseases presenting with consolidation include bacterial and fungal pneumonia, ARDS, COP, HP, and P. jiroveci pneumonia in patients with AIDS.452,453,536,654

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease Decreased Lung Attenuation.  Decreased lung opacity may result from lung destruction, cystic lung disease, or decreased blood flow (mosaic perfusion). Cystic lung disease and abnormal airspaces.  Multiple lung cysts can be found in a wide variety of conditions. The term cyst refers to a thin-walled (<2 mm), well-defined and circumscribed, air- or fluid-containing lesion, 1 cm or more in diameter, surrounded by an epithelial or fibrous wall of variable thickness.334 Cystic structures identified on CT scans include blebs, bullae, cavities, pneumatoceles, “honeycombing,” dilated bronchi, and dilated terminal airways and alveolar sacs. Although the precise mechanism of cystic changes is not fully understood, they can be found whenever lung destruction occurs, as in end-stage fibrosis, emphysema, or infection. A “check-valve” mechanism has also been advocated in the pathogenesis of some cysts. Cystic lung disease encompasses a wide variety of pathologic  processes such as LAM, LCH, light chain deposition disease, lymphocytic interstitial pneumonia, amyloidosis, postinfectious pneumatoceles, cystic metastases, and extrinsic allergic alveolitis.90,193,556 Recently, lung cysts have also been reported in association with Birt-HoggDubé syndrome, an autosomal dominant inherited disorder characterized by fibrofolliculomas, renal tumors, pulmonary cysts, and pneumothorax.39,204,611 Honeycombing is a process characterized by the presence of cystic spaces having thick, clearly definable fibrous walls lined by bronchiolar epithelium; cystic spaces usually average 1 cm in diameter and their walls 1 to 3 mm in thickness. Honeycomb pattern represents the presence of terminal lung disease (“end-stage lung”). At this stage, no specific diagnosis can be reached by biopsy techniques. The determination of the presence or absence of honeycombing on HRCT in patients with IIP is of great importance.596 An additional and important function of HRCT is to guide the surgeon away from the honeycombed region and toward areas of ground-glass opacities where earlier stages of disease are more likely to be found. A confident diagnosis of UIP can be established on the basis of characteristic HRCT findings.41,372 Honeycombing may have an atypical distribution, particularly in asbestosis, sarcoidosis, NSIP, drug-related fibrosis, and chronic HP.580 Mosaic attenuation pattern.  Diseases that produce mosaic attenuation on CT scans fall into three basic categories: airway disease, vascular disease, and infiltrative disease.271,334,396,564,647 Bronchiolar involvement is the basis of airflow obstruction observed in some granulomatous interstitial pneumonias, including sarcoidosis, aspirative diseases (e.g., aspiration bronchiolitis), pneumoconioses (e.g., dust macules, cobalt/hard metal disease), and foreign material inhalation (drugs, organic dusts). Diffuse pulmonary neuroendocrine cell proliferations confined to the bronchial or bronchiolar epithelium may also cause mosaic attenuation or air trapping.60,148,420 Diffuse pulmonary neuroendocrine cell hyperplasia (DIPNECH) and carcinoid tumorlets may be diagnosed incidentally in asymptomatic patients and should be considered in older women with chronic respiratory symptoms and CT evidence of pulmonary micronodules and air trapping. Airway obstruction results in hypoxia of the underventilated lung, reflex vasoconstriction, and air trapping. This is reflected on HRCT by the presence of areas of decreased lung attenuation associated with decreased vascularity. Blood flow redistribution to uninvolved areas results in increased attenuation and increased number and size of vessels in these areas. The combination of areas of decreased attenuation and vascularity and areas of increased attenuation and vascularity is commonly referred to as mosaic perfusion. Although small airway disease can often be suspected by the presence of mosaic perfusion on standard HRCT scans performed at endinspiration, airway obstruction is more readily detected on images

941

performed at the end of maximal expiration. In bronchiolar disease, the lucent regions of lung seen at inspiration remain lucent at expiration owing to air trapping.590,591

Anatomic Distribution of Abnormalities DLD can be characterized as uniform (i.e., involving all parts of the lung diffusely) or having a zonal predominance (upper, middle, or lower; central/perihilar vs. peripheral). Anatomic distribution and CT patterns can be helpful in characterizing the nature of the lung disease and arriving at a limited differential diagnosis for the pulmonary process.228,432 Depending on the anatomic location of abnormalities within the secondary pulmonary lobule, DLD may present with a random distribution, with a distribution primarily along the bronchovascular bundles, or with a centrilobular/peribronchiolar or subpleural distribution.433,444

Bronchovascular Distribution.  Many of the diseases distributed primarily along the bronchovascular interstitium (e.g., sarcoidosis, pulmonary lymphangitic carcinomatosis, silicosis, lymphoproliferative disorders) show a predilection for lymphatic routes. Other entities  can affect the peribronchovascular interstitium without a predominant perilymphatic distribution (e.g., Wegener’s granulomatosis, COP, Kaposi’s sarcoma).93,327,331,471

Centrilobular and Peribronchiolar Distribution.  Diseases that cause this type of CT abnormality are often characterized histologically by peribronchiolar infiltration with inflammatory cells or lymphocytes or with granulomatous reaction. Bronchogenic spread of TB may  also produce a focal form of this pattern when expectorated caseous material from a TB cavity is spread through the airways, becomes inspissated in small bronchioles, and secondarily causes bronchiolar dilatation, thickening, and peribronchiolar infection and inflammation.439 This produces the tree-in-bud appearance on CT scans. TB and other mycobacterial infections often produce some of the most dramatic examples of tree-in-bud opacities,274 but this finding is not necessarily specific for mycobacterial infections and can be seen in other causes of bronchiolitis, both infectious and inflammatory.

Subpleural Distribution.  Diseases such as idiopathic pulmonary fibrosis (IPF), asbestosis, organizing pneumonia, eosinophilic pneumonia, and certain types of drug-induced lung disease may affect the periphery of the lung. CT abnormalities may consist of ground-glass opacity, reticulation, and honeycombing.

Clinical Categories of Diffuse Lung Disease DLD can be divided into 6 broad clinical categories: (1) primary disease related, (2) environmental exposure related, (3) drug-induced lung injury, (4) collagen vascular associated, (5) smoking related, and (6) IIP.

Primary Disease–Related DLD Sarcoidosis.  Sarcoidosis is a common idiopathic systemic disease characterized histologically by the formation of noncaseating granulomas. Although any organ system can be involved, the lung and mediastinal lymph nodes are especially affected. Pulmonary involvement in sarcoidosis is usually bilateral, although it may be asymmetric. A normal chest radiograph is seen in about 10% of patients with sarcoidosis.413 CT is more accurate than conventional radiography in detecting subtle lesions in the lung parenchyma and delineating the full extent of lung disease caused by sarcoidosis.237,328,530 Although the value of MRI in assessing chronic infiltrative lung disease is still controversial, pulmonary fibrosis can be imaged in high detail using MRI.110

http://radiologyebook.com

942

PART II  CT and MR Imaging of the Whole Body

A

B FIG 36-22  Mediastinal lymph node enlargement in sarcoidosis. A, HRCT at the level of the bronchus intermedius (lung window) shows bilateral hilar enlargement (arrows) and multiple small nodules in a perilymphatic distribution (arrowheads). B, CT scan (mediastinal window) at the same level as shown in A demonstrates symmetric hilar lymph node enlargement; enlarged nodes are also visible in the subcarinal space (arrowhead).

Lymphadenopathy is the most common intrathoracic manifestation of sarcoidosis, occurring in 75% to 80% of patients. Bilateral hilar and right paratracheal lymphadenopathy are typical radiographic findings of thoracic sarcoidosis229,317 (Fig. 36-22). In contrast to adenopathy caused by neoplastic disorders, enlarged lymph nodes resulting from sarcoidosis tend to remain discrete and maintain their rounded nodal shape rather than coalescing into amorphous nodal masses.237,327 Eggshell calcification is thought to develop only in affected nodes and is related to duration of disease.150,421 Other diseases that can produce eggshell calcifications in lymph nodes are TB, fungal infections, and silicosis. Parenchymal disease results from interstitial involvement by the granulomatous process to irreversible parenchymal changes related to fibrosis.413 CT findings include areas of ground-glass attenuation, subpleural and perivascular nodules, irregular thickening of broncho­ vascular bundles, and thickening of interlobular septa.150,421,468,478 A reticulonodular pattern results from a combination of nodules and thickening of the interlobular septa. Areas of ground-glass attenuation result from microscopic interstitial granulomas below the resolution of CT. Subpleural nodules, septal nodules, and perihilar peribronchovascular nodules are typical of sarcoidosis (Fig. 36-23). Nodules vary in size and frequently range from 1 to 10 mm, are well circumscribed, and tend to have irregular margins. Granulomas along the interlobular septa can produce an appearance resembling the beaded septum sign seen in lymphangitic carcinomatosis.328,530 These findings typically predominate in the middle to upper lung zones. Occasionally the coalescence of numerous noncaseating granulomas presents as multiple round masslike consolidations on chest films and CT. The distribution of the parenchymal abnormalities can be well seen on cross-sectional images, but the predominant upper lobe and perihilar distribution is particularly well depicted on coronal reformations (Fig. 36-24). Bronchovascular bundles, interlobular septa, major fissures, and pleural/ subpleural regions represent the typical perilymphatic distribution of pulmonary sarcoidosis.610 Collapse is a well recognized but uncommon manifestation of sarcoidosis, occurring in about 1% of cases; any lobe may be affected, most commonly the middle.317,530 Air trapping due to small airway obstruction also may be seen.149,245 Fibrotic conglomerate masses reminiscent of those seen in silicosis may also develop in advanced disease.327 The masses often demonstrate

FIG 36-23  Sarcoidosis. HRCT at the level of the carina shows a perilymphatic distribution of nodules involving the pleural surfaces (arrowheads) and a fissure (arrow).

air bronchograms, and this form of sarcoidosis has been called alveolar or acinar sarcoid.220,326,328,530 True cavities lined by noncaseating granulomas are extremely rare (<0.6%) in sarcoidosis.530 Superimposed infection of fibrocystic spaces, particularly with Aspergillus, is a frequent complication of sarcoidosis and accounts for significant morbidity and mortality. Progressive fibrosis and architectural distortion lead to honeycombing, fibrocystic changes, bullae, traction bronchiectasis, and retraction of the pulmonary hila327,470 (Fig. 36-25). Although mediastinal adenopathy can be massive, it rarely causes compression of great veins with subsequent superior vena cava syndrome. Fibrosing mediastinitis can be a serious complication of sarcoidosis.237 On CT scans, abnormal fibrotic tissue can be seen to narrow and obstruct the mainstem bronchi, mediastinal vessels, and pulmonary vessels. Pulmonary alveolar proteinosis.  Pulmonary alveolar proteinosis (PAP) is a rare disorder characterized by accumulation of surfactantderived phospholipoproteinaceous material into the alveolar spaces.9,122 PAP can be acquired, congenital, or secondary to other conditions such as malignancy (e.g., leukemia, lymphoma), immunodeficiency, lysinuric protein intolerance, and acute silicosis.493,557

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

943

A S1042

FIG 36-26  Crazy-paving pattern in alveolar proteinosis. CT image at the level of the aortic arch shows extensive bilateral ground-glass opacities with superimposed septal lines, resulting in a crazy-paving pattern.

R 1 6 0

B

L 1 6 9

S712

FIG 36-24  Sarcoidosis. HRCT performed on an MDCT scanner shows multiple nodules in a perilymphatic distribution predominating along the fissures and vessels (arrows). Some mediastinal lymph node enlargement is also visible (arrowhead). (Courtesy of Nestor L. Müller MD, Vancouver, BC).

FIG 36-25  Lung fibrosis in sarcoidosis. CT at the level of the aortic arch shows peribronchovascular masses of fibrous tissue (arrows) with central conglomeration and dilatation of bronchi. Note that a right pleural effusion is also visible.

FIG 36-27  Exogenous lipoid pneumonia. HRCT at the level of the lower lobes shows patchy areas of increased lung attenuation and thickened interlobular septa in the middle lobe, lingula, and lower lobes. Note characteristic crazy-paving pattern. (From Franquet T et al: The crazypaving pattern in exogenous lipoid pneumonia: CT-pathologic correlation. AJR Am J Roentgenol 170:315–317, 1998.)

Conventional radiography shows a typical “butterfly” appearance similar to the type of shadow seen in uremic lung. A characteristic CT appearance called the crazy-paving pattern was described as highly suggestive of this disorder222,346,442 (Fig. 36-26). Today it is well known that the crazy-paving appearance is a nonspecific finding seen in a variety of interstitial and airspace lung diseases (e.g., alveolar sarcoidosis, lipoid pneumonia, mucinous bronchoalveolar carcinoma)189,441,536 (Fig. 36-27). However, the various diseases that can cause this appearance can often be distinguished by their clinical manifestations. In PAP the extent of imaging findings is often disproportionately increased relative to the severity of the symptoms and physical findings. PAP may be complicated by infection with unusual organisms such as Aspergillus, Nocardia, mycobacteria, Cryptococcus neoformans, Histoplasma capsulatum, P. jiroveci, and viruses.1,54,111,557 Eosinophilic lung disease.  Eosinophilic lung disease includes a group of widely varied disorders characterized by accumulation of eosinophils in the interstitium and airspaces.21,140 Eosinophilic lung

http://radiologyebook.com

944

PART II  CT and MR Imaging of the Whole Body

disease can be classified as idiopathic or of known cause (allergic bronchopulmonary aspergillosis, bronchocentric granulomatosis, parasitic and fungal infections, drug reactions).21,426 Common idiopathic eosinophilic lung diseases include 5 major categories: (1) simple pulmonary eosinophilia (Löffler’s syndrome), (2) acute eosinophilic pneumonia, (3) chronic eosinophilic pneumonia or a chronic form of pulmonary opacities with eosinophilia, (4) hypereosinophilic syndrome, and (5) polyangiitis or vasculitis in association with pulmonary opacities, which are the predominant part of the disease. CT features of eosinophilic lung diseases have been well described.146,285,293 Acute eosinophilic pneumonia is a sudden and severe febrile illness characterized by acute respiratory insufficiency, hypoxemia, fever, diffuse radiographic infiltrates, and eosinophilia in BAL fluid or lung biopsies in the absence of infection, atopy, or asthma.20,608 HRCT findings are similar to those of pulmonary edema and include diffuse bilateral reticular densities, areas of ground-glass opacity, thickening of bronchovascular bundles, smooth septal thickening, and small pleural effusion103,146 (Fig. 36-28). Whereas simple pulmonary eosinophilia is characterized by mild or absent symptoms, blood eosinophilia, and transitory and/or migratory pulmonary consolidations, chronic eosinophilic pneumonia manifests with symptoms present for several months and characterized by cough with mucoid sputum, high fever, dyspnea, wheeze, malaise, and

A

weight loss.285,424 Simple pulmonary eosinophilia and chronic eosinophilic pneumonia can have similar CT findings, consisting of patchy areas of ground-glass opacity and/or airspace consolidation, often peripheral and patchy in distribution. Other features are large nodules with a halo sign, crazy-paving appearance, focal areas of consolidation, and bandlike opacities.285,293,310 Allergic bronchopulmonary aspergillosis (ABPA) or allergic pulmonary mycosis are terms used to describe a disease characterized by asthma, blood eosinophilia, radiographic pulmonary infiltrates, and allergy to antigens of Aspergillus spp. or other types of fungi.43 Clinical manifestations are wheeze, dyspnea, and productive cough sometimes associated with coughing up sputum plugs.382-384 In the acute phase, radiographic findings are transient and secondary to airway obstruction, including lobar or segmental collapse, mucoid impaction, and consolidation. Mucoid impaction and eosinophilic pneumonia are common pathologic findings in ABPA.75 On CT the impacted proximal bronchi present with either a linear, sharply demarcated, branched or unbranched shadow (“toothpaste shadow”) or with “V- and Y-shaded” opacities (“globed finger shadow”) (Fig. 36-29). In the chronic phase, findings are permanent and characterized by damaged and bronchiectatic airways; proximal bronchiectasis has been considered highly  specific for ABPA.308,425 Bronchocentric granulomatosis is not a disease but rather a descriptive pathologic diagnosis.607 Parenchymal lung involvement mainly

A

B

B FIG 36-28  Acute eosinophilic pneumonia. A, Anteroposterior chest radiograph shows multiple bilateral areas of consolidation (arrows). B, Corresponding coronal CT reconstruction confirms the peripheral distribution of opacities (arrows).

FIG 36-29  Allergic bronchopulmonary aspergillosis. A, HRCT at the level of the carina shows multiple central bronchiectasis (arrowheads) and a left parahilar mucous plug within a dilated central bronchus. B, CT scan obtained at a lower level shows bilateral central bronchiectasis and bronchial wall thickening (arrows).

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease takes the form of consolidation or masslike lesions, reflecting the presence of granuloma formation with or without associated bronchial obstruction.115,183,285,293,628 Hypereosinophilic syndrome is a multisystem disorder characterized by blood eosinophilia present for at least 6 months, associated with multiorgan tissue infiltration by mature eosinophils. HRCT manifestations include transient focal areas of ground-glass attenuation  or consolidation, bilateral small nodules with or without a halo of ground-glass attenuation, and small pleural effusions.42,300,310 Known causes of eosinophilic lung disease include drugs, parasites, and fungi. CT findings are nonspecific and consist of areas of ground-glass opacity, nodules, septal thickening, and thickening of bronchovascular bundles.183,285,293,449 Churg-Strauss syndrome is a systemic disease of unknown etiology characterized by peripheral eosinophilia, asthma, and systemic vasculitis. Radiologic findings consist of multifocal parenchymal nodules, bronchial wall thickening, and ground-glass opacities (Figs. 36-30   and 36-31).

945

Lymphangiomyomatosis.  LAM is a rare disease of unclear etiology exclusively confined to women of childbearing age479 (Fig. 36-32). Microscopically the lungs are characterized by cystic airspaces and a nodular proliferation of abnormal smooth muscle cells (LAM cells) in the walls of lymphatics, bronchioles, and blood vessels.203 Although not a neurocutaneous disorder, LAM shares many pulmonary features with pulmonary tuberous sclerosis (e.g., renal angiomyolipomas are also found in LAM). The earliest radiographic signs in LAM consist of fine nodular, reticular, or reticulonodular opacities.133 In severe cases, chest radiographs may demonstrate hyperinflation, honeycombing, irregular opacities, coarse reticulations, chylous pleural effusions, and pneumothoraces.203,351,432 The CT appearance of this entity is considered so suggestive as to be pathognomonic.35,36 Innumerable thin-walled cysts varying in size from a few millimeters to up to 5 cm can be seen scattered uniformly throughout the lung parenchyma351,432,479 (Fig. 36-33). These cysts are presumably due to airspace dilatation secondary to progressive airway obstruction. Intervening normal lung appears relatively normal, with the notable absence of any other CT sign of disease, such as reticulation, opacities, or septal prominence. Differentiation from emphysema may appear superficially difficult in some cases, but attention to the relationship of vessels to the cystic structures usually solves this problem. In LAM, vessels are always located at the periphery of cysts, whereas in emphysema vessels remain centrally located.

Environmental Exposure–Related DLD.  These diseases are clini-

FIG 36-30  Churg-Strauss syndrome. CT at the level of the carina shows multiple patchy bilateral subpleural consolidations (arrows).

A

cally divided into two groups based on the causative substance: pneumoconiosis (inorganic substances) and HP (organic substances).11 The term pneumoconiosis describes lung diseases secondary to inhalation of mineral dusts. Classic pneumoconiosis includes silicosis, CWP, asbestosis, and asbestos-related pleural disease.11 Silicosis and coal worker’s pneumoconiosis.  Whereas silicosis is a fibrotic lung disease caused by inhalation of dust containing  free crystalline silica during occupational exposure such as mining  and sandblasting, CWP results from inhalation of a dust that is a mixture of coal, kaolin, mica, and silica.586,617 Three clinicopathologic types of silicosis have been described: (1) chronic silicosis, which typically follows exposure to respirable dust over a period of decades,

B FIG 36-31  Churg-Strauss syndrome. A, AP chest radiograph shows a focal area of consolidation in the right upper lobe (arrows). B, Coronal reformation from MDCT scan shows ground-glass geographic opacity (arrows).

http://radiologyebook.com

946

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 36-32  Tuberous sclerosis with associated LAM. A, HRCT at the level of carina shows numerous bilateral thin-walled cysts randomly distributed throughout both lungs. B, CT scan at the level of the upper abdomen shows a left renal angiomyolipoma (arrows).

Pn

Pn

FIG 36-33  LAM with associated pneumothorax (Pn). HRCT at the level of the inferior pulmonary veins shows numerous bilateral thin-walled cysts randomly distributed throughout both lungs. A loculated right pneumothorax614 is demonstrated (arrows). (Courtesy Kyung Soo Lee, MD, Seoul, Korea.)

(2) accelerated silicosis, which follows shorter and heavier exposure, and (3) acute silicosis (silicoproteinosis), which follows intense exposure to fine silica dust, generally over a period of months rather than years.202 Simple silicosis and CWP are regarded as asymptomatic diseases. Caplan’s syndrome is a form of CWP associated with rheumatoid arthritis (RA). CT findings of simple silicosis and CWP include diffuse, randomly distributed, small, well-defined nodules less than 5 mm in diameter that may calcify (Fig. 36-34). Nodules are characteristically in a centrilobular distribution and are most prominent in the posterior aspects of the upper lung zones. Eggshell calcification of hilar or mediastinal lymph nodes can also occur in simple silicosis and CWP.11,13,55,57,63,520 Paired inspiratory and expiratory HRCT scans are useful in the assessment of airway obstruction associated with silicosis.28 With disease progression, coalescence of the nodules leads to development of progressive massive fibrosis (PMF), also known as complicated silicosis (Fig. 36-35). Clinical symptoms and signs of chronic silicosis are no different from those of focal lung fibrosis caused by other agents. The most prominent CT feature of PMF associated with

FIG 36-34  Silicosis. Axial HRCT performed on an MDCT scanner demonstrates numerous bilateral randomly distributed small nodules representing miliary silicotic nodules.

either silicosis or CWP is masslike consolidation (conglomerate masses) associated with parenchymal scarring and adjacent bullae (cicatricial emphysema), usually in the upper lobes.55,56,63 Conglomerate masses are typically oval and have irregular borders. Although usually bilateral, unilateral conglomerate masses may occur and be confused with carcinoma. New lesions, especially if cavitated, should be regarded as evidence of mycobacterial disease.4 PMF is always associated with a background of small nodules visible on HRCT. Punctate calcification in association with conglomerate masses is common.516 MRI has been used to evaluate PMF. The MRI appearance of PMF includes isointensity on T1- and hypointensity on T2-weighted images

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

947

B

A

FIG 36-35  Progressive massive fibrosis in silicosis. A, HRCT (lung window) shows bilateral upper-lobe masses associated with distortion of the adjacent lung parenchyma. B, HRCT at the same level as shown in A (mediastinal window) shows calcification within the masses; multiple calcified mediastinal lymph nodes with typical eggshell calcification (arrow) are also visible.

when compared with skeletal muscle; internal high SI areas on T2 suggest the presence of necrosis.399 Other causes of pneumoconiosis include inhalation of beryllium (berylliosis), hard metals (silicon carbide), polyvinyl chloride, iron oxide (siderosis), tin oxide (stannosis), barium dust (baritosis), and nylon flock.460,503 Asbestosis.  Asbestosis is a respiratory disease resulting from inhalation of asbestos fibers. There are two main groups and forms of asbestos fibers: amphiboles (crocidolite: stiff and straight) and serpentines (chrysotyle: curly and flexible). The toxicity of these fibers is determined by their physical and aerodynamic properties (e.g., smaller fibers cause more disease) as well as by their retention and biopersistence (crocidolite causes more disease than chrysotyle). Asbestosis is diagnosed on the basis of a history of exposure and the characteristic radiographic changes.128,621 Combined exposure to asbestos and cigarettes has an additive effect on lung cancer mortality (14-fold increased risk).46 Asbestos-related pleural disease includes pleural plaques, diffuse pleural thickening, benign asbestos-related pleural effusions, and malignant pleural mesothelioma.447 CT has become an important adjunct to conventional radiographs in the evaluation of asbestos-related diseases.105,128 It is more sensitive and specific than plain films for identifying and documenting the presence of pleural plaques, pleural thickening, and pleural calcifications, all of which are important markers of asbestos exposure17,18,152 (Fig. 36-36). CT shows inter- and intralobular septal thickening, groundglass opacities, and parenchymal bands in a subpleural and basal distribution. With advanced disease, honeycombing becomes more apparent. Fibrotic changes correspond both pathologically and radiographically to those described in UIP.16,130 Although nonspecific, when they are found in association with pleural plaques, they are compatible with asbestos-induced parenchymal lung damage.152 CT is also helpful in evaluating asbestos-related pleural disease, almost always present in patients with parenchymal disease.99 Hypersensitivity pneumonitis.  HP320 is an immunologically mediated inflammation of the lung parenchyma and airways in response to repeated inhalation of organic or low-molecular-weight inorganic antigens.428,456 Disease entities include farmer’s lung, birdfancier’s lung, mushroom worker’s lung, bagassosis, air conditioner

FIG 36-36  Asbestos-related pleural disease. CT shows geographic opacities and focal areas of pleural thickening, some of them with visible calcification (arrows).

lung, cheese washer’s lung, malt worker’s lung, maple bark disease, pituitary snuff taker’s lung, woodworker’s lung, and so forth.7,428,456 Three clinicopathologic types of HP have been described: (1) acute, which typically occurs 4 to 6 hours after antigen inhalation, (2) subacute, which occurs over a period of weeks to months after exposure, and (3) chronic, which follows long-term exposure to antigen.7,428,456 Chronic HP is thought to develop slowly after unrecognized persistent or intermittent exposure to very low levels of antigen during months or years, without discrete episodes of acute respiratory symptoms.287 The pathologic findings of HP depend on the stage of the disease at tissue sampling. Characteristic histologic findings consist of illdefined granulomas, bronchiolitis, and alveolitis.121,257 The alveolitis is characterized by a lymphoplasmacytic interstitial infiltrate that closely resembles the infiltrate seen in lymphocytic interstitial pneumonia. Imaging findings of HP vary with the stage of the disease. The most characteristic CT features are bilateral patchy areas of ground-glass opacity, poorly defined centrilobular nodules, and lobular areas of

http://radiologyebook.com

948

PART II  CT and MR Imaging of the Whole Body

FIG 36-37  Subacute hypersensitivity pneumonitis with air trapping. HRCT at the level of the lung bases shows diffuse centrilobular illdefined nodules. The nodules do not involve the fissures or peripheral pleural surface (arrowheads). Note the presence of patchy lobular areas of lucency representing lobular air trapping (arrows).

mosaic attenuation on inspiratory images and areas of air trapping on expiratory images. Although CT is usually not performed in acute HP, the characteristic CT findings of this stage consist of bilateral areas of airspace consolidation resembling pulmonary edema, and small (1-3 mm in diameter) poorly defined centrilobular nodules.574 The majority of CT examinations are performed in the subacute and chronic stages of the disease. In subacute HP, patchy ground-glass opacities are seen intermingled with ill-defined centrilobular nodules predominating in the middle and lower lung zones251,371,484,571,574 (Fig. 36-37). Another common manifestation of subacute HP is the presence of a mosaic attenuation pattern due to either patchy areas of groundglass attenuation or air trapping, or often a combination of both. Air trapping on expiratory CT scans, presumed to be due to bronchiolitis, is seen in many cases of subacute HP.241,247,484,513,571 The combination of geographic areas of ground-glass opacity, normal lung attenuation, and decreased attenuation (air trapping) result in an appearance that has been termed the head cheese sign.109 Sometimes thin-walled cysts can be seen in a small percentage of patients (10%) with subacute HP193 (Fig. 36-38). Chronic HP is characterized by the presence of fibrosis.113 The most common CT findings of chronic HP consist of irregular reticular opacities, patchy bilateral areas of ground-glass opacity, small poorly defined nodules, and honeycombing.113,228,370,371 An appearance similar to that of chronic HP can be seen in some patients who have IPF or fibrotic NSIP.113,370,571

Drug-Induced Lung Injury.  A large number of drugs, both cytotoxic and noncytotoxic, are known to adversely affect the lungs and have been associated with an acute or chronic pulmonary reaction.126,127,142,449,577,578 Diagnosis requires careful correlation of clinical symptoms, history of drug exposure, and imaging findings.321 Druginduced lung injury is thought to result from either a direct toxic  effect or an indirect effect by way of a hypersensitivity reaction.14,88 Cytotoxic drugs comprise the largest group of drugs associated with lung toxicity.88 They include cytotoxic antibiotics (e.g., bleomycin, mitomycin, sirolimus), alkylating agents (e.g., busulfan, cyclophosphamide, chlorambucil), nitrosoureas (e.g., BCNU), antimetabolites (e.g., methotrexate, gemcitabine), and new therapeutic agents (e.g., taxoid derivatives paclitaxel and docetaxel, gemcitabine, topotecan, vinorelbine). In addition, the list of noncytotoxic drugs associated with pulmonary toxicity is constantly enlarging and includes a wide range of

FIG 36-38  Subacute hypersensitivity pneumonitis with cysts. HRCT obtained at the level of the aortic arch shows subtle centrilobular nodular opacities and extensive bilateral ground-glass attenuation. Multiple thin-walled cysts are seen in both upper lobes (arrows). (From Franquet T et al: Lung cysts in subacute hypersensitivity pneumonitis. J Comput Assist Tomogr 27:475–478, 2003.)

therapeutic agents such as antimicrobials (e.g., nitrofurantoin, paraaminosalicylic acid, sulfasalazine), antiinflammatory drugs (e.g., gold salts, etanercept, infliximab), interferons (e.g., interferon alfa and gamma), antiarrhythmics (e.g., amiodarone), antihypertensives, psychotherapeutic drugs, anorexigens, and others. Pulmonary reactions to cytotoxic and noncytotoxic drugs have been well established and can take several forms of tissue reaction: (1) diffuse alveolar damage (e.g., busulfan, cyclophosphamide, BCNU, sirolimus, gemcitabine), (2) organizing pneumonia (e.g., cyclophosphamide, bleomycin, mitomycin, methotrexate, nitrofurantoin), (3) eosinophilic pneumonia (e.g., bleomycin, sulfasalazine), (4) NSIP (e.g., bleomycin, busulfan, cyclophosphamide, methotrexate, gemcitabine, amiodarone, gold salts), and (5) pulmonary hemorrhage (e.g., anticoagulants, amphotericin B, high-dose cyclophosphamide, mitomycin, cytarabine [ara-C], penicillamine).115,167,168,533,535,573 There is considerable overlap between the histologic responses to various drugs.115 Drug-induced lung injury142 comprises a wide variety of pathologies, and there are no characteristic CT findings to establish the diagnosis.164,535 NSIP and fibrosis result in ground-glass opacities, focal areas of consolidation, and irregular linear opacities that tend to involve the lower zones of the lungs.164,167,535 In diffuse alveolar damage, CT shows extensive bilateral parenchymal consolidation that may be more marked in the dependent lung regions. Organizing pneumonia results in peribronchial or subpleural areas of consolidation. CT findings of eosinophilic pneumonia consist of patchy areas of consolidation, ground-glass opacity, or pulmonary edema.

Reactions to specific drugs

Bleomycin.  The pulmonary toxicity of bleomycin has been well established and can take several forms, including interstitial pneumonitis, organizing pneumonia, and eosinophilic hypersensitivity. CT features of bleomycin lung toxicity include linear reticulations, fibrotic bands, and nodular areas of consolidation that may mimic pulmonary metastases or recurrent tumor.58,450,527 Diffuse airspace disease can also be seen with acute bleomycin toxicity (Fig. 36-39). Associated pleural effusions have been reported in a small number of patients. These changes often appear first in the periphery of the lung and at the lung bases, paralleling the pleural surfaces.450,527

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

949

Amiodarone.  Amiodarone, a triiodinated antiarrhythmic drug, can cause significant and potentially irreversible pulmonary toxicity in approximately 6% of patients who have been treated with high cumulative doses for a long period of time.394,395,529,644 Rarely, short-term amiodarone therapy can also induce pulmonary toxicity.349 Symptoms are nonspecific and include pleuritic chest pain, shortness of breath, malaise, and low-grade fever.529,644 Discontinuation of therapy is typically associated with a good prognosis. Associated tissue reactions include NSIP (cellular or fibrotic type), alveolar filling by foamy macrophages, organizing pneumonia, diffuse alveolar damage, and a pattern of UIP. Alveolar hemorrhage and prominent lymphoid and/or eosinophilic infiltrates may rarely appear to be an unusual drug-induced pulmonary toxicity associated with amiodarone.339,604 On unenhanced CT scans, amiodarone toxicity may be the only condition in which CT is diagnostic because of the high-attenuation values of amiodarone deposits in the lungs335,336 (Fig. 36-40). These opacities are usually peripheral in location.83,336,466,523,581 Increased attenuation is also frequently demonstrated in the liver, spleen, thyroid gland, and occasionally in the heart muscle. Other common HRCT findings are septal thickening and interstitial fibrosis.

Methotrexate.  Methotrexate is a folate antagonist used to treat a wide variety of neoplastic and nonneoplastic inflammatory conditions including psoriasis, primary biliary cirrhosis, inflammatory bowel diseases, and most commonly RA.34,276 Methotrexate-induced pulmonary drug toxicity occurs in 5% to 10% of patients. Histologic features associated with lung injury include diffuse  alveolar damage, HP, NSIP, and organizing pneumonia89,340 (Fig. 36-41). CT findings consist of patchy or diffuse areas of ground-glass opacity.115,475 There is no correlation between development of drug toxicity and duration of therapy or total cumulative dose.168,533,535 Sirolimus.  Sirolimus (rapamycin) is a new immunosuppressive drug used in solid organ transplantation to limit the nephrotoxicity of calcineurin inhibitors (cyclosporine and tacrolimus).211 Pulmonary toxicity has recently been recognized as a potentially serious complication associated with sirolimus therapy. Several distinct histologic features have been described, such as lymphocytic alveolitis, lymphocytic interstitial pneumonitis, organizing pneumonia, and pulmonary alveolar hemorrhage.489,625 Organizing pneumonia is the most common pathologic finding in sirolimus-associated pulmonary toxicity in cardiac transplantation359 (Fig. 36-42). CT findings consist of ground-glass opacities and consolidation, either alone or in combination.145,391 Illicit drugs.  The use of illicit drugs is increasing, and these drugs represent a common cause of drug-induced pulmonary disease.407,490,524 Heavy smoking of freebase (“crack”) cocaine may induce an acute respiratory distress associated with diffuse lung infiltrates (“crack lung”).22,141,318 Both cardiogenic and noncardiogenic pulmonary edema are common pulmonary complications associated with IV cocaine abuse and crack cocaine smoking.526 Pulmonary hemorrhage manifests radiologically as bilateral multifocal areas of increased opacity. On HRCT, these opacities have a ground-glass appearance that may be centrilobular with interlobular septal thickening443 (Fig. 36-43). IV injection of methylphenidate may result in talcosis (IV drug abuser’s lung) and severe panlobular emphysema.115 Initial HRCT findings consist of numerous high-attenuation nodules measuring less than 1 mm in diameter or diffuse ground-glass opacities.387

FIG 36-39  Organizing pneumonia pattern resulting from treatment

Collagen Vascular Disease–Associated DLD.  The connective tissue/collagen vascular diseases are a heterogeneous group of immunologically mediated disorders in which the lungs are an important

with bleomycin. CT through the upper lobes shows patchy areas of consolidation and ground-glass pattern.

A

B FIG 36-40  Amiodarone lung toxicity. Nonenhanced CT scan (mediastinal window) at the level of the lower lobes shows bilateral areas of airspace consolidation denser than soft tissues (arrows).

http://radiologyebook.com

950

PART II  CT and MR Imaging of the Whole Body

FIG 36-43  Noncardiogenic pulmonary edema resulting from IV cocaine injection. Chest radiograph shows bilateral and symmetric alveolar infiltrates. Note the normal heart size (arrows).

FIG 36-41  Nonspecific interstitial pneumonia (NSIP) pattern resulting from treatment with bleomycin. CT (lung window) at the level of the lung bases shows patchy areas of ground-glass opacity.

FIG 36-42  Organizing pneumonia pattern resulting from treatment with sirolimus. HRCT at the level of the lung bases shows bilateral irregular areas of consolidation, many occurring in relation to bronchi. There is sparing of the immediate right subpleural lung. Bilateral pleural effusion is also seen.

target organ owing to their abundant connective tissue.210,290,296,498 The most common form of diffuse pulmonary lung disease is a chronic interstitial pneumonitis presenting either with a UIP or NSIP pattern. HRCT is the imaging method of choice in evaluating patients with connective tissue diseases, demonstrating the presence, gross characteristics, and distribution of pulmonary disease with greater sensitivity than conventional chest radiographs. In certain clinical circumstances, HRCT findings can suggest a specific diagnosis.210,290,297,498,603 Rheumatoid arthritis.  RA is the most common of the connective tissue diseases, affecting about 1% of people worldwide.562,605 Females are affected two to three times more often than males. The frequency of pulmonary abnormalities found in association with RA has been

shown to vary widely and is likely to depend on multiple factors, such as the stage of disease in which patients are studied, the source of patient referral, and the parameters used to define disease.27 Pleural involvement, either pleural effusion or pleural thickening, is the most common thoracic manifestation of RA. Pleural disease (i.e., adhesions, thickening, or effusion) is found in different autopsy series in 38% to 73% of cases.40,492,562 The degree of ILD in RA does not necessarily correlate with the severity of underlying articular involvement. Radiographic findings are often subtle, and chest radiographs may be normal or show only fine interstitial opacities.201,210,259,269 Clear fibrosis and honeycombing are only detected in the more severe forms of the disease. Although the ILD associated with RA is histologically indistinguishable from UIP, lung fibrosis is usually slowly progressive504,650 (Fig. 36-44). NSIP may also occur in RA.304 Rheumatoid (necrobiotic) pulmonary nodules are pathologically identical to those found in subcutaneous tissue. The nodules typically measure between 0.5 and 5 cm in size and may cavitate. The reported prevalence of rheumatoid nodules is variable, ranging from low rates based on chest radiographs to 22% in a series of 77 patients evaluated by HRCT.134 Although rheumatoid nodules are often asymptomatic, spontaneous pneumothorax due to rupture of a cavitary necrobiotic nodule is a rare complication of RA.100,298 Caplan’s syndrome is characterized by RA with pulmonary rheumatic nodules in coal miners.565 Airways disease reflecting the presence of bronchiolitis obliterans or follicular bronchiolitis is being increasingly recognized in RA and other collagen vascular diseases.255,256 Small airways disease is difficult to diagnose at chest radiography because findings may be subtle or confusing. Focal areas of air trapping may be seen on images obtained at expiratory CT256,266,487,514 (Fig. 36-45). Systemic lupus erythematosus.  Systemic lupus erythematosus (SLE), which most commonly affects young women, is a chronic  multisystem disease of unknown origin characterized by the presence of autoantibodies against various cell nuclear antigens.48,312 Pleuropulmonary disease will occur in more than half of patients with SLE at some point during the course of their illness.10,43,49,63 Pleuritis is the most common thoracic manifestation of SLE. Other thoracic manifestations are inflammatory pneumonitis (acute lupus pneumonitis), interstitial pulmonary fibrosis, pulmonary hypertension, pericarditis,

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

951

B FIG 36-44  Rheumatoid arthritis with fine reticular opacities. A, PA chest radiograph shows reduced lung volumes and bilateral peripheral reticular opacities. B, HRCT shows a fine reticular pattern with peripheral and basal predominance. There is no evidence of honeycombing.

10

A

FIG 36-46  Acute lupus pneumonitis. CT shows patchy rounded areas of consolidation and ground-glass opacities in the left lung (arrows) and a multilobar consolidation involving middle and right lower lobes. A right pleural effusion is also visible (arrowheads).

B FIG 36-45  Obliterative bronchiolitis in a patient with rheumatoid arthritis. A, Inspiratory HRCT scan at the level of the upper lobes shows bilateral inhomogeneous lung attenuation representing mosaic attenuation. B, Expiratory CT at the same level as shown in A confirms the presence of air trapping. Note decreased vessel size in the lucent lung regions (arrows).

diaphragmatic dysfunction (“shrinking lung syndrome”), and phrenic nerve palsy.48,312,629 In contrast to many other collagen vascular diseases, SLE is not commonly associated with chronic diffuse ILD. Infection is the most common pulmonary complication in SLE, accounting for about 50% of pleuropulmonary manifestations.369 Usual bacterial and opportunistic pathogens, mycobacterial, and nocardial infections seem to be particularly important in patients with SLE429 (see Fig. 36-6). Acute lupus pneumonitis is an uncommon

life-threatening complication of SLE that is characterized by fever, severe hypoxemia, and diffuse pulmonary infiltrates (Fig. 36-46). It must be differentiated from pulmonary infection or hemorrhage. Histologically there is interstitial edema and inflammation with intraalveolar exudation and hyaline membrane formation, corresponding to diffuse alveolar damage. Diffuse alveolar hemorrhage70 is a rare but dramatic complication of SLE.652 Its radiographic appearance is that of nonspecific diffuse airspace consolidation. HRCT scans can show patchy bilateral and/or diffuse areas of ground-glass attenuation with superimposed interlobular septal thickening, giving an appearance of a crazy-paving pattern174,312 (Fig. 36-47). The so-called antiphospholipid antibody syndrome is a rare life-threatening complication characterized by an increased incidence of recurrent venous and/or arterial thrombosis.369 Progressive systemic sclerosis.  Progressive systemic sclerosis is an uncommon inflammatory fibrotic disease of the skin (scleroderma) and some internal organs. The CREST syndrome (subcutaneous calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia) is a limited form of scleroderma.498

http://radiologyebook.com

952

PART II  CT and MR Imaging of the Whole Body

FIG 36-47  Diffuse alveolar hemorrhage. HRCT at the level of the aortic arch shows bilateral ground-glass opacities with superimposed septal lines in the upper lobes (crazy-paving pattern).

ILD is the most common complication of progressive systemic sclerosis, occurring in up to 75% of cases. In the majority of patients, the interstitial lung pathology is subclinical, being completely asymptomatic in the early stages, as are most other ILDs.144 The HRCT findings of interstitial fibrosis include ground-glass attenuation, subpleural reticular opacities, traction bronchiectasis and bronchiolectasis, architectural distortion, pleural thickening or effusion, and centrilobular micronodules (follicular bronchiolitis) in a predominantly peripheral and basilar distribution.61,498,553 Progressive systemic sclerosis follows a less aggressive course and has a better prognosis than IPF; it is commonly associated with mediastinal lymph node enlargement, pulmonary vasculitis, and pulmonary hypertension. Esophageal dilatation is demonstrated on CT in 80% of patients with ILD68 (Fig. 36-48). Sjögren’s syndrome.  Sjögren’s syndrome is a chronic inflammatory autoimmune disorder characterized by the triad of keratoconjunctivitis sicca, xerostomia, and (in up to 50% of cases) a connective tissue disorder.84 Various forms of pulmonary complications have been reported in patients with primary Sjögren’s syndrome, including interstitial fibrosis, small airways disease, lymphocytic interstitial pneumonitis, and low-grade B-cell lymphoma of bronchus-associated lymphoid tissue (BALToma).190 Lymphocytic interstitial pneumonia is a benign lymphoproliferative disorder characterized by diffuse interstitial proliferation of small lymphocytes and plasma cells described in patients with Sjögren’s  syndrome, AIDS, autoimmune thyroid disease, and Castleman’s disease.290,294 HRCT findings include ground-glass attenuation, cystic lesions, poorly defined centrilobular nodules, and thickening of the interstitium along the lymphatic vessels290,296 (Fig. 36-49). Ankylosing spondylitis.  Ankylosing spondylitis most commonly affects young men between 20 and 40 years of age, 90% of whom are HLA-B27 positive.369,605 Although the exact prevalence of pulmonary disease is rare and remains to be determined, parenchymal abnormalities have been described in only 1% of patients.534 The earliest manifestation of the disease is thought to be bilateral apical scarring. These radiographic changes may mimic those of chronic TB, with apical fibrocavitary changes and cyst formation. Secondary superinfection of the cavities with Aspergillus has been described in some of these patients195,369,605 (Fig. 36-50). Additional abnormalities such as nonapical fibrosis, bronchiectasis, paraseptal emphysema, and tracheobronchomegaly have also been documented.173,498

FIG 36-48  Progressive systemic sclerosis. HRCT shows a fine reticular pattern with a peripheral and basal predominance. There is irregular dilatation of multiple bronchi (arrows). This appearance probably is due to NSIP. Dilatation of the lower esophagus is also visible (arrowheads).

FIG 36-49  Sjögren’s syndrome and lymphoid interstitial pneumonia. HRCT at the level of the bronchus intermedius shows diffuse groundglass opacity and bilateral ill-defined nodular opacities (arrows). A single cyst is visible in the right upper lobe (arrowhead).

Smoking-Related Lung Diseases.  Cigarette smoking has been recently related to several lung disorders, including DIP, RBILD, and pulmonary LCH. Desquamative interstitial pneumonia.  DIP is one of the rarest of the IIPs, accounting for less than 3% of ILDs and affecting young cigarette smokers (aged 30-40 years), with a male-to-female ratio of 2 : 1. Histologically DIP is characterized by large numbers of intraalveolar macrophages containing a characteristic finely granular light brown pigment, mild chronic inflammatory interstitial cell infiltrate, minor fibrotic changes, and preservation of lung architecture. Histologic findings of DIP may sometimes overlap with those of RBILD.649,651 A DIPlike reaction may also be observed in several disorders such as LCH, dust inhalation, and drug reactions (e.g., to nitrofurantoin, busulfan, sulfasalazine). The chest radiograph is usually normal. HRCT demonstrates bilateral diffuse ground-glass opacification with a predominant lower involvement (Figs. 36-51 and 36-52). Although interlobular septal thickening suggestive of fibrosis is seen in half of these patients, honeycombing is rarely observed. The CT differential diagnosis includes RBILD, NSIP, and acute or subacute HP.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease Respiratory bronchiolitis–associated interstitial lung disease.  RBILD is an incidental histopathologic finding found in virtually all smokers and characterized by the presence of pigmented intraluminal macrophages within first- and second-order respiratory bronchioles.253,541 It is rarely symptomatic. RBILD is the clinical manifestation

FIG 36-50  Ankylosing spondylitis. Coronal reformation from HRCT performed on an MDCT scanner demonstrates reticular opacities in the right lung apex, with significant volume loss (arrowheads). A large cavity containing an aspergilloma is present in the left upper lobe. The fungus ball is separated from the wall of the cavity by an airspace, resulting in the distinctive air crescent sign (arrows). Also note left upper lobe volume loss, bronchiectasis, and communication between the cavity and the left upper lobe bronchus. (This material has not been reviewed by European Respiratory Society prior to release; therefore the European Respiratory Society may not be responsible for any errors, omissions or inaccuracies, or for any consequences arising there from, in the content. Reproduced with permission of the European Respiratory Society©: European Respiratory Journal Mar 2004, 23(3):488–489; DOI: 10.1183/09031936.04.00042804.)

A

953

of ILD associated with the pathologic lesion of RB. This clinicopathologic entity is seen almost exclusively in current or former cigarette smokers aged 30 to 40 years and has a male-to-female ratio of 2 : 1. In terms of the differential diagnosis, there is a strong histopathologic overlap between RBILD and DIP.253,541,651 Unlike DIP, macrophages in RBILD have a predominantly peribronchiolar distribution.649,651 Histologically, areas of LCH, RBILD, and DIP may be found in the same patient.619 Peribronchiolar alveolar septal thickening by fibroblasts and collagen deposits can be present. Usually there is a chronic inflammatory cell infiltrate in the wall of the bronchiole and surrounding alveolar walls.186 Honeycombing has not been described. The discovery of RB on lung biopsy is not equivalent to a diagnosis of RBILD. Whereas respiratory bronchiolitis refers to the universal inflammatory reaction found in the bronchioles of cigarette smokers and is an extremely

FIG 36-51  Desquamative interstitial pneumonia. Magnified view from a HRCT shows patchy areas of ground-glass opacity in the left lower lobe and lingula (arrows).

B FIG 36-52  Desquamative interstitial pneumonia. A and B, HRCT at two different levels shows bilateral hazy ground-glass opacity in the peripheral lung.

http://radiologyebook.com

954

PART II  CT and MR Imaging of the Whole Body matous changes. Changes predominate in the upper and middle lung zones; costophrenic angles are generally spared. Lung volumes are usually normal but sometimes increased. HRCT demonstrates the extent and distribution of parenchymal lung damage in LCH more accurately than conventional radiography. Much of what appears to be reticulation on plain films can be shown to be discrete cystic airspaces with definable walls on HRCT.25 In the early stage of the disease, CT demonstrates ill-defined centrilobular micronodules (1-5 mm) mainly in a peribronchiolar distribution. Cavitated nodules are seen in about 10% of cases (Fig. 36-54). In a later stage, smooth cysts ranging from a few millimeters to several centimeters in diameter are frequently seen. Sometimes several cystic spaces coalesce to form bizarrely shaped airspaces on CT scans (Fig. 36-55). Their pathogenesis is uncertain and may be related to more than one mechanism, such as necrosis of the central portion of the cellular nodules or peripheral airspace dilatation due to bronchiolar obstruction. Sometimes cystic changes can be difficult to differentiate from emphysema. Lack of visible walls in emphysema may be the only discriminating feature. Later, cysts appear that vary in shape, size, and wall thickness. These cystic changes are produced by two mechanisms: one from cavitation of the nodules and the other from dilated bronchiolar structures surrounded by peribronchiolar granulomas. Sometimes cystic changes can be difficult to differentiate from emphysema. Lack of visible walls in emphysema may be the only discriminating feature. In advanced disease the lung may consist almost entirely of fibrous tissue, forming large areas of honeycombing and paracicatricial emphysema, more prominently in the upper lobes.

A

Idiopathic Interstitial Pneumonias.  IIPs are a group of diffuse

B FIG 36-53  Respiratory bronchiolitis–associated interstitial lung disease (RBILD). A, HRCT performed on an MDCT scanner at the level of the aortic arch shows patchy areas of hazy ground-glass opacity (arrows) in the peripheral lung. B, Coronal reformation from HRCT performed on an MDCT scanner shows similar findings in the lung apices (arrows).

common lesion, RBILD is an uncommon ILD that occurs in a small proportion of smokers. Chest radiographic abnormalities include thickening of the walls of the central and peripheral bronchi and diffuse bilateral reticulonodular opacities. The chest x-ray is normal in 20% of patients.517 Areas of ground-glass attenuation are the most common finding on HRCT, but fine centrilobular nodules may also be seen (Fig. 36-53). Although emphysematous changes may be present, subpleural honeycombing and traction bronchiectasis are absent. Similar but less extensive findings can be seen in asymptomatic smokers. These changes are attributable to respiratory bronchiolitis. The CT differential diagnosis of RBILD includes acute HP, DIP, and NSIP. There is considerable overlap in the appearances of RBILD and DIP on HRCT that could represent different degrees of severity of small airway and parenchymal reactions to cigarette smoke.619 Pulmonary langerhans cell histiocytosis.  The isolated pulmonary form of LCH is an uncommon disorder of young adult smokers characterized by accumulation of Langerhans cells on the epithelial surface of the respiratory tract. Most patients are smokers aged 30 to 40 years, with an even distribution between males and females. Spontaneous pneumothorax may develop in 15% of patients. Plain chest radiography demonstrates any combination of reticulation, nodules (2-5 mm), cysts (<1 cm), honeycombing, and emphyse-

parenchymal lung diseases that manifest varying histopathologic patterns of inflammation and fibrosis. In 2002 the American Thoracic Society/European Respiratory Society (ATS/ERS) classification of IIPs defined seven specific entities and provided standardized terminology and diagnostic criteria.24 These disorders are classified most accurately by the combination of histopathology on lung biopsy with radiographic and clinical findings (e.g., exertional dyspnea, restrictive lung defect, including decreased diffusing capacity of lung for carbon monoxide [Dlco] and abnormal P[A-a]o2 [alveolar-arterial difference in partial pressure of oxygen] at rest and/or with exertion).180,297,433 The diagnosis and classification of IIPs continue to be problematic areas for pathologists.206,464,465 Recently the international multidisciplinary classification of IIPs has been updated.614 Although the main entities are preserved, significant changes have been introduced.614 The current classification of IIPs include the following categories: • Major IIPs: IPF/UIP, idiopathic NSIP, RBILD, DIP, COP, and acute interstitial pneumonia • Rare IIPs: idiopathic lymphoid interstitial pneumonia (LIP) and idiopathic pleuroparenchymal fibroelastosis (PPFE) • Unclassifiable IIPs Idiopathic PPFE is a rare condition that consists of fibrosis involving the pleura and subpleural lung parenchyma, predominantly in the upper lobes.53,187 The distinction among IIPs has to be based on clinical, radiographic, and histopathologic criteria.5,181,182

Usual interstitial pneumonia and idiopathic pulmonary fibrosis.  UIP/IPF, the most common of the IIPs, is an idiopathic progres-

sive ILD with a mean survival from the onset of dyspnea of 3 to 6 years.547 The histologic findings of UIP/IPF include dense fibrosis causing remodeling of lung architecture, with frequent honeycomb changes, fibroblastic foci typically scattered at the edges of dense scars, and fibrotic zones with temporal heterogeneity (with dense acellular collagen and scattered fibroblastic foci).

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

955

B FIG 36-54  Pulmonary Langerhans cell histiocytosis. A and B, Coronal reformation from MDCT combined with MIP demonstrates numerous thick-walled irregular cysts scattered throughout the lungs, and small nodules, some of them cavitated (arrows). Note relative sparing of the lung bases.

B

A

FIG 36-55  Pulmonary Langerhans cell histiocytosis. A and B, HRCT performed on an MDCT scanner at two different levels shows numerous thin-walled irregular cysts (arrowhead) and small nodules (arrows) scattered throughout the lungs.

The most common HRCT findings consist of symmetric fine reticular attenuation with intralobular linear opacities, interlobular septal thickening, architectural distortion with associated traction bronchiectasis and bronchiolectasis, ground-glass opacity, and a honeycomb pattern.433 Honeycombing is more severe in the subpleural lung regions and lung bases (Fig. 36-56). In the majority of patients the extent of ground-glass is less than the extent of reticulation.374 The CT differential diagnosis of IPF includes UIP caused by asbestosis and collagen vascular disease, chronic HP, and sarcoidosis.6,130,294,367,478,521 Combined pulmonary fibrosis and emphysema.  The combination of emphysema and diverse fibrosing ILD is frequently observed in HRCT imaging and sometimes in conjunction with pathology. Characteristic imaging features of combined pulmonary fibrosis and

emphysema include evidence of emphysema in the upper zones (i.e., centrilobular and/or paraseptal emphysema in 90% of cases) and diffuse infiltrating fibrosing lung disease at the bases (subpleural reticular opacities, honeycombing, and traction bronchiectasis).532,641 This combination has been recently proposed as a new clinical entity in smokers or a different clinical phenotype in smokers developing IPF.414,539 Although the prevalence of combined pulmonary fibrosis and emphysema has been estimated at close to 30%, the reasons for this combination remain unknown.136,138,139,414 Clinically, patients with IPF and emphysema had a higher mortality rate compared with those with IPF without emphysema.414,539 Nonspecific interstitial pneumonia.  NSIP is a chronic ILD that has a more favorable prognosis than UIP. Patients present with

http://radiologyebook.com

956

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 36-56  Idiopathic pulmonary fibrosis. A and B, HRCT performed on an MDCT scanner shows extensive subpleural honeycombing, reticular opacities, and traction bronchiectasis in the lung periphery (arrows).

A

B FIG 36-57  Nonspecific interstitial pneumonia. A, HRCT at the level of the lung bases shows bilateral subpleural ground-glass opacity (arrows). B, Prone HRCT scan at the same level shows ground-glass opacity in the posterior lung bases (arrows).

subacute to chronic dyspnea, and the mean age of onset is 40 to 50 years. Histologically NSIP is temporally and spatially homogeneous, with uniform involvement of alveolar septa by varying degrees of inflammation or fibrosis.137,304 According to the relative amounts of inflammation and fibrosis, NSIP is divided into two morphologic subtypes: cellular NSIP and fibrotic NSIP.311 The active fibroblastic foci and microscopic honeycombing that define UIP are rare in NSIP.304,451 NSIP may be idiopathic or a manifestation encountered in association with other disorders such as connective tissue diseases, HP, or  drug exposure.167,278,295 The most common HRCT findings consists of ground-glass attenuation and a fine reticular pattern in the lower lung zones; ground-glass attenuation may be the sole feature in nearly one third of patients5,252 (Fig. 36-57). Honeycombing is seen in 27% of patients with NSIP; traction bronchiectasis and bronchiolectasis become increasingly prominent as fibrosis progresses.252,295,374 Acute interstitial pneumonitis.  Acute interstitial pneumonitis is an acute idiopathic ILD characterized by rapidly progressive diffuse pulmonary infiltrates, dyspnea, and hypoxemia. Its pathologic hallmark is diffuse alveolar damage525 on lung biopsy, manifested by injury

to alveolar epithelial and endothelial cells, pulmonary edema, and hyaline membrane formation, and later proliferative changes involving the alveolar and bronchial lining cells and interstitial cells.481 The commonest HRCT features consist of bilateral patchy densities that usually progress to a diffuse alveolar pattern.296 Acute exacerbation of UIP, also referred to as the accelerated phase of IPF, represents a condition in which patients with UIP (and occasionally other forms of fibrotic ILD) develop rapid respiratory failure accompanied by extensive radiologic infiltrates.147,314,323,483 Histologically, findings consist of diffuse alveolar damage superimposed on features typical of UIP or NSIP.114,481 Architectural distortion, traction bronchiectasis, and ground-glass opacity are prominent features on initial CT scans obtained more than 7 days after the onset of symptoms10 (Fig. 36-58). Organizing pneumonia.  Organizing pneumonia (OP), previously called bronchiolitis obliterans organizing pneumonia (BOOP), is histologically characterized by proliferation of granulation tissue within bronchioles, alveolar ducts, and adjacent alveolar airspaces.165,412 Although OP may be idiopathic (COP), most cases are seen in

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

FIG 36-58  Acute interstitial pneumonia. HRCT scan at the level of the lung bases shows diffuse ground-glass opacity with associated fine reticulation (crazy-paving appearance) and traction bronchiectasis (arrows) indicative of fibrosis.

A

association with a variety of conditions including connective tissue disorders, viral or bacterial infection, aspiration, airway obstruction, and drugs.59,98,131,132,412 Clinically patients present with a subacute onset of nonspecific symptoms, including cough, dyspnea, fever, and flulike illness.165 However, significant differences have been found between COP and secondary OP regarding severity of symptoms, signs, and laboratory and pulmonary function tests.19,362,482,620 The characteristic CT appearance of OP consists of patchy unilateral or bilateral areas of airspace consolidation5,165,437 (Fig. 36-59). Other CT features include ground-glass opacities, a focal organizing pneumonia pattern, a variety of nodular appearances, a bronchocentric distribution, bandlike opacities, a perilobular pattern, and a progressive fibrotic form.30,76,471,615 The reversed halo sign, a relatively specific CT finding associated with OP, has been reported in one fifth of patients with this disease313 (Fig. 36-60). Idiopathic lymphoid interstitial pneumonia.  The consensus classification sponsored by the ATS/ERS recognizes idiopathic lymphoid

B FIG 36-59  Cryptogenic organizing pneumonia. A, Axial HRCT performed on an MDCT scanner shows bilateral areas of consolidation in a predominantly peribronchial distribution (arrows). B, Coronal reformation demonstrates bilateral areas of consolidation and multifocal areas of ground-glass opacity (arrows).

A

957

B FIG 36-60  Cryptogenic organizing pneumonia (reversed halo sign). A and B, CT at the level of the lower lobes shows patchy areas of ground-glass opacity surrounded by a dense rim (arrows).

http://radiologyebook.com

958

PART II  CT and MR Imaging of the Whole Body

interstitial pneumonia as a rare ILD.614 Lymphoid interstitial pneumonia usually occurs in association with autoimmune diseases, most often Sjögren’s syndrome.196 The most common HRCT findings in both forms consist of ground-glass attenuation, cystic lesions, poorly defined centrilobular nodules, and thickening of the interstitium along the lymphatic vessels.94,196,297,290

chronic. CT findings consist of bilateral ground-glass opacities in association with septal thickening in a predominantly peripheral or dependent distribution, axial thickening, and pleural effusions (Fig. 36-61). Mediastinal lymphadenopathy and hazy mediastinal fat may also be seen in congestive heart failure.169,463,576

Other Specific Disease Processes

bilateral airspace disease that often has a ground-glass appearance on CT.385,496 Patients present with hemoptysis and dyspnea and may have iron deficiency anemia if episodes are recurrent. DPH can result from a variety of causes including Goodpasture’s disease (antiglomerular basement membrane disease), collagen vascular or autoimmune disorders (e.g., SLE, Wegener’s granulomatosis), idiopathic pulmonary hemosiderosis, bleeding disorders, drug reactions, infections, and tumors200,496 (Fig. 36-62). Pulmonary renal syndrome represents a severe syndrome characterized by a combination of DPH and glomerulonephritis.208 Goodpasture’s syndrome, vasculitis syndromes, and collagen vascular

Diffuse Pulmonary Hemorrhage.  DPH is characterized typically by

Pulmonary Edema.  Hydrostatic pulmonary edema can be defined as an abnormal increase in extravascular water secondary to elevated pressure in the pulmonary circulation, as in congestive heart failure or intravascular volume overload. The diagnosis is usually based on clinical information, conventional chest radiograph findings, and response to treatment. Edema can mimic other diseases or occur as an unsuspected finding in patients having HRCT for other indications.593 The differential diagnosis of ground-glass opacities with or without associated crazy-paving must be based on a consideration of clinical as well as HRCT findings and knowledge of whether symptoms are acute or

A

B FIG 36-61  Pulmonary edema. A, HRCT at the level of the bronchus intermedius shows patchy perihilar ground-glass opacity. Thickening of the fissures and small bilateral pleural effusions are seen. Note the enlarged main pulmonary artery (arrowheads). B, At a lower level, septal thickening (arrows) and patchy lobular ground-glass opacity is visible.

A

B FIG 36-62  Goodpasture’s syndrome with pulmonary hemorrhage. HRCT performed on an MDCT scanner shows multiple bilateral areas of consolidation and geographic areas of ground-glass opacity with a superimposed linear pattern.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease diseases may present with a combination of pulmonary hemorrhage and renal disease. Goodpasture’s syndrome is characterized by episodes of DPH, renal glomerulonephritis, and circulating antiglomerular basement membrane antibodies. The characteristic CT findings include bilateral perihilar airspace consolidation or ground-glass opacity that spares the apices and costophrenic angles; ill-defined centrilobular opacities or a crazy-paving pattern may also be seen.101,392,496 Diffuse alveolar hemorrhage also occurs in microscopic polyangiitis and Wegener’s granulomatosis.480,613

Acute Respiratory Distress Syndrome.  ARDS is characterized by diffuse lung injury with progressive dyspnea and hypoxemia.66,184 The patient with ARDS generally requires high positive-pressure mechanical ventilation and is at high risk for development of barotrauma. ARDS can be related to a variety of pathologic processes, either pulmonary (e.g., infection, inhalation or aspiration of toxic or irritating substances, embolic disease, drugs) or extrapulmonary (e.g., sepsis, pancreatitis, extrathoracic trauma, shock).66 Although CT has not been used extensively in patients with ARDS, it may play a role in the evaluation of acute and long-term complications in this setting.214,224 Differences in CT appearance between ARDS due to pulmonary disease and that due to extrapulmonary disease have been described.224 ARDS due to pulmonary disease tends to be asymmetric with a mix of consolidation and ground-glass opacification, whereas ARDS due to extrapulmonary disease has predominantly symmetric ground-glass opacification (Fig. 36-63). Pleural effusions and air bronchograms are common in both groups.224 Asymmetric distribution of ARDS due to compensatory hyperexpansion of the residual lung within the hemithorax operated has been described following thoracic surgery (lobectomy).476 CT findings may vary according to the evolutional stage of the disease over time. A reticular pattern with a striking anterior distribution is a frequent finding of follow-up CT in ARDS survivors and is most strongly related to the duration of pressure-controlled inverse-ratio ventilation.155,474 Barotrauma-related complications in ARDS may occur.377,642 The Macklin effect relates to a three-step pathophysiologic process: alveolar ruptures, air dissection along bronchovascular sheaths, and spreading of this blunt pulmonary interstitial emphysema into the mediastinum.376,377 CT is more accurate in identifying and localizing complications of barotrauma such as pneumothoraces, pneumomediastinum, and subpleural air cysts.262 This may become important in those

FIG 36-63  Acute respiratory distress syndrome (ARDS). CT shows bilateral extensive areas of patchy consolidation, ground-glass attenuation, and a focal area of crazy-paving in the dependent region of the right upper lobe (arrows). Multifocal areas of normal lung are visible.

959

selected patients who are not responding to treatment or deteriorate suddenly for unclear reasons.

Pulmonary Emphysema.  Emphysema is defined as a condition of the lung characterized by abnormal permanent enlargement of airspaces distal to terminal bronchioles, accompanied by destruction of their walls, without obvious fibrosis. Clinically emphysema is characterized by irreversible airflow obstruction, pulmonary hyperinflation, and dyspnea on exertion. The chest radiograph remains unreliable in the detection of emphysema, except in the advanced stage of the condition. Radiologic appearances of emphysema may closely resemble those seen in severe asthma. In severe emphysema, the diaphragm is low and flat, the retrosternal space is increased, and the heart is small and vertical, all signs of overinflation. However, loss of peripheral vessels and diminution in the size of the mid-lung vessels in emphysema are not as uniform throughout the lung as in asthma. CT is now widely used for evaluation of emphysema234 (Figs. 36-64 and 36-65). In terms of sensitivity, specificity, and assessment of disease extension, CT—particularly MDCT—represents a significant advance over chest radiography. CT lung densitometry is more sensitive than the results of pulmonary function tests in detecting the progression  of emphysema.234,357 Longitudinal CT studies in emphysema have shown that quantitative analyses of low-attenuation clusters may be helpful for not only the elucidation of emphysema progression but also clarification of relation with risk factors such as smoking357,400,401,544 (Fig. 36-66).

Airways Disease

Bronchiectasis.  Bronchiectasis is a condition that is characterized by permanent dilatation of bronchi, with destruction of elastic and muscular components of their walls, usually due to acute or chronic infection. Although most cases of bronchiectasis in adults are idiopathic, a variety of causes such as infection (bacterial, viral, fungal,

FIG 36-64  Centrilobular emphysema. Coronal reformation performed on an MDCT scanner shows diffuse areas of emphysema in the upper lobes.

http://radiologyebook.com

960

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 36-65  Paraseptal emphysema. HRCT at the level of the upper lobes shows extensive emphysema in the subpleural regions. The appearance is characteristic of paraseptal emphysema. Mild centrilobular emphysema is also visible.

A

B

FIG 36-66  Quantification of emphysema. Transverse CT section through the left lower lobe of a patient with emphysema (A) and corresponding density map of the same section (B). The transverse section (A) shows substantial parenchymal destruction caused by emphysema. The density map (B) gives the CT densities for each pixel. These densities are then color coded according to predefined ranges of densities. In this example, red and green correspond to the most severe parenchymal destruction, whereas dark pink indicates normal lung density. Note that parenchymal scars and larger vessels (white bands and spots) have been excluded by the software. (Courtesy Alex Bankier, MD, Boston, Mass.)

mycobacterial), hypogammaglobulinemia and other immune deficiencies (e.g., ABPA), dyskinetic cilia syndromes (Kartagener’s syndrome), bronchial obstruction (e.g., foreign body, broncholithiasis, neoplasm), genetic disorders (e.g., cystic fibrosis), congenital cartilage deficiency (e.g., Williams-Campbell syndrome), and pulmonary fibrosis should be considered in the differential diagnosis.283 Kartagener’s syndrome refers to the triad of situs inversus totalis, bronchiectasis, and either

nasal polyps or recurrent sinusitis occurring in approximately half of patients with primary ciliary dyskinesia.283,398 Although a chest radiograph must be the initial imaging examination in patients with suspected bronchiectasis, in most instances radiographic findings are either nonexistent or nonspecific.232,233 HRCT scanning of the chest is the best test to establish this diagnosis,  with sensitivity and specificity exceeding 90%.344 Morphologically,

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease bronchiectasis is described as cylindrical, varicose, or cystic in type. The diagnosis of bronchiectasis is generally established by recognition of characteristic CT signs. Bronchial wall thickness on thin-section CT scans should be evaluated with window centers between −250 and −700 HU and with window widths greater than 1000 HU. Other than window settings, notably window widths less than 1000 HU, can lead to substantial artificial thickening of bronchial walls.47 Cylindrical bronchiectasis is characterized by smooth uniformly dilated bronchi. When the dilated bronchus is sectioned lengthwise the parallel thickened airway walls resemble a railroad track or tram track, and sectioned crosswise they appear round or oval; because each bronchus is accompanied by a pulmonary artery, bronchiectasis commonly produces a signet ring sign (Fig. 36-67). Lack of bronchial tapering is also an important sign in the diagnosis of bronchial dilatation. When dilated bronchi are filled with inspissated material, mucoid impaction results. On a CT scan, mucous plugs are identified by recognizing the branching tubular nature of the opacity on multiple CT sections. The term tree-in-bud refers to the presence of nodular and branching structures, reflecting the presence of dilated mucus- or pus-filled centrilobular bronchioles (cellular bronchiolitis). Tree-in-bud and air trapping (on expiratory scans) are findings commonly seen in patients with bronchiectasis. Beaded dilatation of bronchi resembling a string of pearls describes the varicose bronchiectasis. Cystic bronchiectasis is characterized by cysts in clusters, resembling a cluster of grapes (Fig. 36-68). Air-fluid levels resulting from retained secretions or superimposed infection are commonly identified in cystically dilated bronchi. Bronchial postinflammatory changes during or immediately after an episode of acute pneumonia may simulate bronchiectasis; these changes (reversible bronchiectasis) revert to normal in 1 to 3 months.459 Traction bronchiectasis refers to irreversible dilatation of bronchi and bronchioles in areas of pulmonary fibrosis.154,635 Bronchiectasis related to ABPA are seen most commonly in patients with longstanding bronchial asthma. They are characterized pathologically by the presence of plugs of inspissated mucus containing Aspergillus organisms and eosinophils. Radiologic manifestations include homogeneous tubular finger-in-glove opacities in a bronchial distribution, usually involving predominantly or exclusively the upper lobes.425

A

961

These shadows are related to plugging of airways by hyphal masses with distal mucous impaction and can migrate from one region to another. Bronchiectasis may be varicose or cylindrical, involving predominantly the segmental and subsegmental bronchi of the upper lobes.458 Bronchiectasis in association with pulmonary nodules or tree-inbud appearance is characteristic of nontuberculous mycobacterial

FIG 36-67  Left lower lobe bronchiectasis. Magnified view from HRCT shows evidence of bronchiectasis (arrows) associated with bronchial wall thickening (arrowheads).

B FIG 36-68  Cystic bronchiectasis. A and B, HRCT performed on an MDCT scanner shows bilateral numerous thin-walled cystic bronchiectasis (arrows). Note the presence of a small air-fluid level, indicative of superimposed infection (arrowhead in A).

http://radiologyebook.com

962

PART II  CT and MR Imaging of the Whole Body

A

B FIG 36-69  Nontuberculous mycobacterial infection and bronchiectasis. A, HRCT performed on an MDCT scanner shows bilateral dilated thickwalled bronchi. B, Coronal reformation demonstrates numerous thickwalled and dilated bronchi along with mucous plugging (arrows).

infection due to MAC356 (Fig. 36-69). It typically occurs in immunocompetent elderly women (>60 years) with no significant smoking history or underlying pulmonary disease; this condition, usually limited to the right middle lobe or lingula, has been named Lady Windermere syndrome.156,356,505 Adult patients with cystic fibrosis typically have panlobular bronchiectasis.246 Mild reversible bronchial dilatation may be seen with severe chronic bronchitis and be associated with acute pneumonia. Patients with an established diagnosis of cystic fibrosis may have significant worsening of respiratory symptoms with little visible radiographic changes.

Chronic Bronchitis and Asthma.  Chronic bronchitis, one of the most common causes of lower airway disease, has been traditionally defined as cough and sputum expectoration occurring on most days for at least 3 months of the year and at least 2 consecutive years. In developed countries, cigarette smoking is responsible for 85% to 90% of cases of chronic bronchitis. Other airway diseases such as cystic fibrosis and asthma have clinical and histologic features that overlap

those of chronic bronchitis. The role of infection in chronic bronchitis remains unclear. Asthma is a chronic inflammatory disorder of the airways that is associated with recruitment of inflammatory cells. It affects 5% of the U.S. population and accounts for 2 million emergency department visits, 470,000 hospitalizations, and 4500 deaths annually.51 In a patient with chronic cough, asthma should always be considered as a potential etiology because it is a common condition with which cough is commonly associated. Different studies have shown that asthma is one of the most common etiologies of chronic cough (24%-29%) in adult nonsmokers.277,408 Clinically the patient experiences intermittent episodes of cough, dyspnea, and wheezing that may be associated with exposure to specific sensitizing substances (extrinsic asthma). Intrinsic asthma is characterized by the absence of atopy or specific external triggers of bronchoconstriction. Symptoms may reverse either spontaneously or as a result of treatment. A conventional chest radiograph is rarely used to make a diagnosis of asthma. On chest radiography, increased lung volume, increased lung lucency, mild bronchial wall thickening, and mild prominence of hilar vasculature are the most common findings seen in adults with asthma. It is not surprising that radiologic changes, when present, are often slight and that the chest radiograph is commonly normal. The reported frequency of hyperinflation in adults with acute asthma varies between 20% and 70%. Moreover, chest radiographs are useful to exclude the presence of a number of complications and associations with asthma such as consolidation, atelectasis with mucoid impaction, pneumothorax, pneumomediastinum, and ABPA. Consolidation in asthmatics is commonly infective but in some cases is due to eosinophilic consolidation. CT is not usually indicated in the routine assessment of patients with asthma. However, the majority of symptomatic patients have an abnormal CT showing both bronchial and parenchymal abnormalities. Characteristic CT findings are bronchial dilatation, bronchial wall thickening, mucoid impaction, cylindrical bronchiectasis, centrilobular bronchiolar abnormalities such as tree-in-bud, patchy areas of mosaic perfusion and regional areas of air trapping on expiratory scans, and thick linear opacities representing discoid atelectasis240,248 (Fig. 36-70). When air trapping is severe, they can be seen on the inspiration scans. In patients with severe asthma, a mosaic pattern of lung attenuation is indistinguishable from that seen in patients with constrictive bronchiolitis. In patients in whom bronchiolitis is suspected, a surgical lung biopsy should be performed when the combination of the clinical syndrome, physiology, and HRCT findings do not provide a confident diagnosis. Improvements in CT scanning techniques together with faster quantitative algorithms to measure airway wall and lumen areas and to quantify and localize air trapping, are being applied with increased frequency in research efforts to understand the changes in airways that occur in chronic obstructive lung diseases.

Bronchiolitis.  Several infectious and noninfectious processes may affect the small airways, causing either reversible or fixed obstruction.434,542 Some of these disorders exclusively affect the small airways, whereas others may also affect the larger bronchi or lung parenchyma. The introduction of HRCT has led to a considerable improvement in the diagnosis of small airway diseases.194,242 The pattern of abnormality on HRCT has been shown to correlate closely with histologic findings. HRCT therefore is helpful in the differential diagnosis of small airway diseases.279,439 Furthermore, small airway diseases may be evident on HRCT in patients with a normal chest radiograph.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

B FIG 36-70  Asthma. A, Inspiratory HRCT shows no significant abnormalities. B, Dynamic expiratory scan shows subtle geographic areas of air trapping (arrows) indicative of bronchial obstruction.

963

findings.248,366 Moreover, MiniP improves visualization of peripheral bronchi (MiniP images).67,455,512,518 Obliterative bronchiolitis (bronchiolitis obliterans).  Obliterative bronchiolitis (bronchiolitis obliterans, constrictive bronchiolitis) is characterized histologically by the presence of inflammation of the walls of membranous and respiratory bronchioles and concentric narrowing of the bronchiolar lumen caused by submucosal and peribronchiolar fibrosis.448 Some of the commonest causes and associations of OB are infections, allograft recipients, connective tissue disorders, toxic fumes inhalation, drug reactions, and bronchiectasis. OB has also been described in association with microcarcinoid (neuroendocrine) tumorlets.120,198,448,542 Viral infections and Mycoplasma pneumoniae are a common cause of OB in children. The abnormalities in these patients can be patchy or diffuse. Swyer-James syndrome represents a postinfectious OB.393,430 OB is a relatively common complication of lung, heart-lung, and bone marrow transplantation. OB is a major long-term complication of lung transplantation, typically occurring after the first postoperative year.125 The term bronchiolitis obliterans syndrome (BOS) is used to refer to a progressive deterioration of graft function secondary to airway disease but not explained by other factors such as infection, acute rejection, or anastomotic complications.345,353 Air trapping, as detected on expiratory CT, is the most sensitive and accurate radiologic indicator of OB in lung transplant recipients.345,353 OB is seen with increased frequency in patients with connective tissue disease, particularly RA.27,640 It also is relatively common in patients with inflammatory bowel disease such as ulcerative colitis and Crohn’s disease.87 Exposure to toxic fumes of such gases as sulfur dioxide, ammonia, phosgene, and oxides of nitrogen may also be associated with an acute alveolar injury and severe OB.412,434,120,198,542 Characteristic HRCT findings of OB consist of areas of decreased attenuation and vascularity, with blood flow redistribution resulting  in areas of increased lung attenuation and vascularity (mosaic perfusion pattern).153,588-590 On inspiratory CT scans the density differences that characterize the mosaic perfusion pattern may be subtle; in  these cases expiratory scans are especially effective in the diagnosis of mosaic perfusion due to air trapping secondary to airway diseases31 (Fig. 36-71).

Other bronchiolitides

Window levels appropriate for optimal visual assessment of the lung parenchyma (window width 1000 ± 1500 HU; window level −600 to −700 HU and ± 450 HU) are recommended. The usual protocol consists of a combination of 1-mm collimation scans at 10-mm intervals at end-inspiration from the lung apices to below the costophrenic angles and additional expiratory CT scans at five selected levels through the upper and lower lungs (from the top of the aortic arch to just above the right hemidiaphragm). Expiratory HRCT is most helpful in the detection of air trapping.588 At suspended full expiration, while normal lung parenchyma increases in attenuation, air trapping results in areas of decreased attenuation and vascularity.542,590,647 Assessment of air trapping can be difficult in patients who are unable to suspend respiration for the time required for CT scanning. In these cases lateral decubitus CT scanning and continuous expiratory helical CT have been proposed as comple­ mentary maneuvers to improve the conspicuity and extent of air trapping.199,365 The diagnostic accuracy of volumetric acquisition of HRCT images represents an important improvement over conventional CT for assessing thoracic abnormalities, including those that involve the small airways.491,542 Helical CT is more accurate than routine HRCT in the assessment of bronchiectasis in patients with equivocal HRCT

Diffuse panbronchiolitis.  Diffuse panbronchiolitis, a chronic inflammatory disease found most commonly in Asians, is now increasingly recognized elsewhere.12,15 It is a clinicopathologic entity characterized by symptoms of chronic cough, dyspnea, wheezing, and sputum production associated with abnormal pulmonary function tests (mild to moderate obstruction) and ill-defined nodular infiltrates.281 It is characterized histologically by inflammatory cell infiltration of the bronchiolar walls and surrounding interstitium.281 In early disease, CT findings include poorly defined nodules and distal branching bronchovascular structures (tree-in-bud pattern), peribronchiolar thickening, and peripheral areas of mosaic attenuation. These changes are strikingly diffuse, involving all lobes.12,15 In advanced disease, this inflammatory process causes bronchiolectasis, bronchial wall thickening, and bronchiectasis12,15 (Fig. 36-72). Although progression to respiratory failure may occur, nodules, branching structures, and mucus plugging are partially reversible with low-dose erythromycin therapy. Follicular bronchiolitis.  Follicular bronchiolitis, defined as lymphoid hyperplasia of the bronchus-associated lymphoid tissue (BALT), is a benign condition associated with connective tissue diseases.316,540,624,634 HRCT findings consist of multiple small nodules in a predominantly centrilobular, subpleural, and peribronchial dis­ tribution266 (Fig. 36-73). The differential diagnosis includes lymphocytic interstitial pneumonia, also seen with increased frequency in

http://radiologyebook.com

964

PART II  CT and MR Imaging of the Whole Body patients with connective tissue diseases, particularly Sjögren’s syndrome.286,294,572

Tracheobronchial Papillomatosis.  Laryngeal papillomatosis, a condition characterized by relapsing benign epithelial neoplastic polyps in the larynx, is caused by infection with the human papillomavirus (HPV).207 This disease entity is most common in children but may also occur in adults. Papillomas may be successfully treated by surgical resection or regress spontaneously at puberty. Rarely, distal spread of papillomas involving the tracheobronchial tree and lung parenchyma is seen in 2% to 5% of cases; this condition is referred as tracheobronchial papillomatosis.69,307 Radiographs or CT may show widely distributed round nodules, either solid or cystic231 (Figs. 36-74 and 36-75). Malignant degeneration of pulmonary lesions may occur, leading to squamous cell carcinoma.502 FIG 36-71  Obliterative bronchiolitis (mosaic attenuation pattern). Expiratory HRCT shows bilateral well-demarcated patchy areas of increased and decreased lung attenuation. Whereas hypoattenuated areas contain few and small pulmonary vessels, hyperattenuated areas contain enlarged pulmonary vessels (arrows), reflecting the pulmonary blood flow distribution toward normal ventilated areas.

A

FIG 36-72  Diffuse panbronchiolitis. HRCT at the level of the aortic arch shows bilateral bronchial dilatation, bronchial wall thickening, and decreased lung attenuation in the periphery of the lungs. (Courtesy Kyung Soo Lee, MD, Seoul, Korea.)

B FIG 36-74  Tracheobronchial papillomatosis. A, HRCT performed on an FIG 36-73  Follicular bronchiolitis. HRCT performed on an MDCT scanner at the level of the upper lobes shows ill-defined centrilobular nodules (arrows).

MDCT scanner at the level of the upper trachea shows nodular irregularity of the tracheal lumen (arrow). B, CT at a lower level (lung window); a small nodule (papilloma) is visible arising from the tracheal wall (arrowhead). A thin-walled cyst is visible in the right upper lobe (arrow).

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease Pulmonary Infections The spectrum of organisms known to cause respiratory infections  is broad and constantly increasing. In the last 2 decades important  new viral pathogens have been recognized, including hantaviruses, human metapneumovirus, avian influenza A viruses, and coronavirus associated with severe acute respiratory syndrome  (SARS).77,86,102,197,235,236,379,555 The chest radiograph represents an important initial examination in all patients suspected of having a pulmonary infection.188,217,606,623 CT is a useful adjunct to conventional radiography in detection, dif-

FIG 36-75  Tracheobronchial papillomatosis. Coronal reformation from an MDCT scanner shows multiple bilateral cystic lesions of different sizes. (Courtesy Alex Bankier, MD, Boston, Mass.)

ferential diagnosis, and management of patients with pulmonary infections.81,282,497,612 The most common CT features of pulmonary infection are foci of consolidation, ground-glass opacities, or reticulonodular opacities. Other less common findings include hilar and mediastinal lymphadenopathy, pleural effusion, cavitation, and chest wall invasion. Community-acquired pneumonia642 is a major healthcare problem because of its considerable incidence and mortality.467 The spectrum of causative organisms includes gram-positive bacteria such as Streptococcus pneumoniae (pneumococcus), Haemophilus influenzae, and Staphylococcus aureus, as well as atypical organisms such as Mycoplasma pneumoniae, Chlamydia pneumoniae, or Legionella pneumophila and viral agents such as influenza A and respiratory syncytial viruses. In community-acquired pneumonia, focal consolidation is the most common CT pattern and may be seen in association with either bacterial or viral infections. S. pneumoniae is by far the most common cause of complete lobar consolidation.85,157,303 Other causative agents that produce complete lobar consolidation include Klebsiella pneumoniae and other gram-negative bacilli, L. pneumophila, H. influenzae, and occasionally M. pneumoniae.3,85,157,303 Bacterial infections may produce multiple rounded pulmonary nodules or masses with or without cavitation. This may occur from Nocardia, Aspergillus, Legionella, Q fever, or M. tuberculosis infection.161,338,501 Diffuse bilateral interstitial and/or interstitial-alveolar (mixed) opacities are most commonly caused by viruses and M. pneumoniae.507 Alcoholic patients and those with poor oral hygiene are prone to develop pulmonary infections.192,390 Patients with advanced periodontal disease are at particular risk for development of either aspiration pneumonia or septic PE.108,566 CT findings include focal or patchy ill-defined areas of consolidation, multiple rounded opacities surrounded by a halo of ground-glass attenuation, and progressive abscess formation.108,566 Recently a newly identified human respiratory virus, human metapneumovirus (hMPV), has been reported.73 The HRCT manifestations of hMPV pneumonia consist of a mixture of patterns, including most commonly ground-glass attenuation and nodular opacities197 (Fig. 36-76). A new form of atypical pneumonia termed severe acute respiratory syndrome (SARS) by the World Health Organization was

* A

965

B FIG 36-76  Human metapneumovirus infection after HSCT. A, HRCT scan at level of carina shows poorly defined nodules in both lungs (arrows). Also seen are bilateral subpleural areas of ground-glass opacification. B, HRCT scan obtained at level of lung bases shows multiple centrilobular nodules (long arrows), few branching opacities (tree-in-bud pattern) (short arrows), and focal areas of consolidation (asterisk) with adjacent ground-glass opacification in left lower lobe. (From Franquet T et al: Human metapneumovirus infection in hematopoietic stem cell transplant recipients. J Comput Assist Tomogr 29:223–227, 2005.)

http://radiologyebook.com

966

PART II  CT and MR Imaging of the Whole Body

FIG 36-77  SARS. HRCT at the level of the lower pulmonary veins shows a focal area of consolidation in the medial segment of the left lower lobe (arrows) and bilateral ground-glass opacities in the lower lobes (arrowheads). (Courtesy Nestor L. Müller, MD, Vancouver, Canada.)

reported in the East Asian region beginning in early 2003.160,324 The diagnosis of SARS relies on a combination of clinical features and imaging findings. The most common radiographic findings consist of unilateral or bilateral ground-glass opacities or focal unilateral or bilateral areas of consolidation96,223,268,319,436,575 (Fig. 36-77). In hospitalized patients the abnormalities tend to progress to bilateral airspace consolidation.197,358,435,436,472 Primary TB occurs when the host immune response initially fails to overcome M. tuberculosis. Postpulmonary TB is usually associated with a reduction in immunocompetence. The most frequent manifestations of primary TB are parenchymal disease, lymphadenopathy, pleural effusion, miliary disease, or atelectasis.107,250,352,419 Cavitation indicates active disease. On CT scans, cavities can be thick walled or thin walled and are indistinguishable from those caused by other atypical mycobacterial species.352,419 Communication with bronchi enables the expectoration of tubercle bacilli and endobronchial spread, leading to the typical tree-in-bud CT appearance350,419 (Fig. 36-78). MAC infection can also strike otherwise healthy immunocompetent hosts. A distinct subtype of pulmonary MAC infection has been reported in middle-aged and elderly women presenting with a chronic persistent cough.505 CT findings consist of bronchiectasis and small nodules, predominantly located within the middle lobe and lingula.505

A

B FIG 36-78  Airway abnormalities due to Mycobacterium avium infection. A and B, CT sections at the level of the inferior pulmonary veins shows branching opacities (tree-in-bud) (arrows) representing small airways impacted by pus in the right middle lobe, lingula, and superior segment of the right lower lobe.

Immunocompromised Patients.  Patients with impaired immune function are susceptible to infections by a wide range of organisms.373,445,446 In the last 3 decades, the AIDS epidemic, advances in the treatment of cancer, organ transplantation, and immunosuppressive therapy have resulted in large numbers of patients who develop abnormalities in their immune system.143,178,260,403 AIDS.  In patients with AIDS the type and pattern of infection are influenced by the patient’s immune status.249 Patients who have more than 200 CD4 cells/mm3 are predisposed mainly to bronchial infections and bacterial pneumonia, whereas patients with fewer than 200 CD4 cells/mm3 are predisposed to opportunistic infections such as P. jiroveci.249,560 At CD4 lymphocyte counts below 100 cells/mm3, and especially at CD4 lymphocyte counts below 50 cells/mm3, infectious causative agents include M. tuberculosis, MAC, endemic fungi such as P. jiroveci, certain viruses, and many of the more common grampositive and gram-negative bacteria.52,71,92,230,427

Mycobacterial infection.  After decades of decreasing incidence, TB has reemerged as an important infection, its increased incidence since the the mid-1980s related to the AIDS epidemic.33,352 The incidence of TB in patients with AIDS is 200 to 500 times that of the general population.352,360 Also increased in these patients is the frequency of nontuberculous mycobacterial infections, most commonly MAC.145,179 Infection with M. tuberculosis or MAC can be acquired through primary infection or secondary to reactivation. The immu­ nosuppressed state associated with AIDS predisposes patients with latent TB to reactivate their disease. In the early state of human immunodeficiency virus (HIV) infection, TB presents in a more classic fashion with cavitary disease and bronchogenic spread. In the setting of more severe immunosuppression, TB presents in a manner more reminiscent of primary TB, primary progressive TB, or miliary326 TB

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

967

FIG 36-81  Pneumocystis jiroveci pneumonia with cyst formation. CT FIG 36-79  Miliary spread of Mycobacterium tuberculosis. HRCT at the level of the bronchus intermedius shows numerous bilateral randomly distributed small nodules representing miliary spread of TB.

FIG 36-80  Mediastinal lymphadenopathy due to Mycobacterium tuberculosis infection. Contrast-enhanced CT at the level of the aortic arch shows multiple low-attenuation lymph nodes (arrows).

(Fig. 36-79). Mediastinal and hilar adenopathy as a manifestation of TB in an AIDS patient is the rule rather than the exception (Fig. 36-80). Airway involvement has been reported in 10% to 20% of all patients with pulmonary TB.26,273,274,409 Centrilobular bronchial abnormalities appear as branching structures in a tree-in-bud appearance as a sign of endobronchial infection.26 Infection with MAC tends to occur in the late stage of AIDS when immunodeficiency is severe and CD4 counts fall below 50 cells/ mm3.179,375 In 10% to 20% of patients with AIDS, MAC infections are detected sometime during their life from blood, sputum, or bone marrow cultures. Approximately 20% of chest radiographs in patients with MACrelated pulmonary disease are normal.388 Radiologic appearances of MAC-related pulmonary disease are similar to TB, including multifocal patchy consolidation or ill-defined nodules that may cavitate.166,417 The most common CT manifestation of MAC infection is diffuse adenopathy.50 After administration of contrast material, tuberculous lymph nodes demonstrate characteristic low-attenuation necrotic centers and rim enhancement.50,275,352

at the level of upper lobes shows multiple cysts surrounded by groundglass attenuation (arrows).

Fungal infection.  Patients with AIDS are at risk of developing fungal infections, which require intact T-cell function for containment. Early in the course of infection, 10% of patients with documented P. jiroveci infection have normal or equivocal chest films. In this setting, CT can be used to confirm or exclude the presence of parenchymal lung abnormalities.37 Advances in prevention and treatment of P. jiroveci infection have been associated with an increased frequency of different radiographic presentations. A cystic form of Pneumocystis pneumonia associated with increased risk of spontaneous pneumothorax has become more prevalent in AIDS patients receiving prophylaxis with aerosolized pentamidine and trimethroprim-sulfamethoxazole106 (Fig. 36-81). Less common radiographic patterns of Pneumocystis pneumonia include parenchymal consolidation, mass lesions, multiple pulmonary nodules, and lymph node enlargement.72 Pleural effusion and pleural calcifications have been reported in disseminated Pneumocystis infection. Fungal pneumonias other than P. jiroveci have been increasingly reported in AIDS patients.582-584 Infections caused by Aspergillus occur in patients with AIDS but much less frequently than they occur in other immunocompromised patients, such as those with leukemia and prolonged neutropenia. Cryptococcus and Aspergillus have become the most frequent lung mycoses in AIDS.567 Obstructing bronchopulmonary aspergillosis is a descriptive term for the unusual pattern of a noninvasive form of aspergillosis characterized by massive intraluminal overgrowth of Aspergillus spp., usually Aspergillus fumigatus, in patients with AIDS.422,648 Other fungal infections including Histoplasma capsulatum and Coccidioides immitis are seen in endemic areas.545,546 Bacterial infection.  The incidence of pyogenic bacterial pneumonia in AIDS patients is five times greater than in the HIV-negative population.72,477 Recurrent bacterial infections have been included as an AIDS-defining illness in the revised Centers for Disease Control (CDC) criteria.2 Bacterial infections are mainly caused by S. pneumoniae, Staphylococcus pneumoniae, and H. influenzae. Nocardia is another organism that is increasingly being recognized in AIDS patients.176 A solitary or multiple pulmonary nodules that eventually develop into a cavitary mass or more extensive cavitary pneumonia are the most common CT findings.531,594 In recent years recurrent episodes of pyogenic airways disease and pneumonia, usually caused by S. pneumoniae, H. influenzae, Pseudomonas aeruginosa, Streptococcus viridans, and S. aureus, have been increasingly recognized.72,409,410,601 Non-AIDS immunocompromised patients.  The number of nonAIDS immunocompromised patients has significantly increased

http://radiologyebook.com

968

PART II  CT and MR Imaging of the Whole Body

FIG 36-82  Cytomegalovirus pneumonia. HRCT scan at the level of the bronchus intermedius shows multiple poorly defined bilateral nodules in a random distribution (arrows).

mainly because of increased numbers of hematopoietic stem cell or organ transplants, advances in the treatment of cancer, and immunosuppressive therapy.497 Hematopoietic stem cell transplant (HSCT) is currently the treatment of choice for many hematologic malignancies and severe congenital or acquired disorders of the hematopoietic or immune systems.354,579 Pulmonary complications occur in 40% to 60% of HSCT recipients and are a major cause of morbidity and mortality.170-172,646 Fungi are the most common cause of pulmonary infection during the neutropenic phase of HSCT (up to 3 weeks after transplantation).543,646 Conversely, cytomegalovirus (CMV) pneumonia most commonly occurs 3 weeks to 100 days after transplantation (early phase)543,646 (Fig. 36-82). Solid organ transplantation has become the treatment of choice for patients with end-stage diseases of the kidney, liver, heart, and lung. Transplant recipients are also at high risk for pulmonary infections.418 Bacterial pneumonia is the most common infectious complication of heart and heart-lung transplantation, occurring in up to 50% of cases.170-172,418 During the first month after heart transplantation, gramnegative bacterial pneumonia is particularly frequent because of prolonged intubation, pulmonary edema, and effects of surgery on lung mechanics.170-172 Gram-negative bacteria (Enterobacter and Pseudomonas) are also a common cause of infection in kidney and liver transplants.177 Bacterial pneumonias are not as lethal as viral and fungal infections.418 CMV is the most common viral pathogen encountered in solid organ transplant recipients. CMV infection typically emerges within the first 3 months after transplantation.170-172 Primary infection, the most serious, occurs in 50% to 100% of seronegative recipients who receive a graft from a seropositive donor.646 Aspergillus species commonly colonize the airways of lung transplant recipients, but only a minority of patients develop invasive disease.485

HIGH-ATTENUATION PARENCHYMAL ABNORMALITIES Diffuse high-attenuation pulmonary abnormalities can result from deposition of calcium or, less commonly, other high-attenuation material such as talc, amiodarone, iron, mercury, and barium sulfate.82,387

FIG 36-83  A 30-year-old woman with idiopathic pulmonary hemosiderosis. HRCT scan shows extensive bilateral centrilobular ground-glass opacities as well as poorly defined and well-defined centrilobular nodules. Some of the nodules have increased attenuation, but no calcification was evident on CT. The patient had repeated episodes of diffuse pulmonary hemorrhage since early childhood. HRCT findings had not changed appreciably over several years. (From Marchiori E et al: Diffuse high-attenuation pulmonary abnormalities: A pattern-oriented diagnostic approach on high-resolution CT. AJR Am J Roentgenol 184:273–282, 2005.)

Metastatic Pulmonary Calcification Metastatic pulmonary calcification is associated with a variety of benign and malignant disorders such as primary and secondary hyperparathyroidism, chronic renal failure, sarcoidosis, IV calcium therapy, multiple myeloma, and massive osteolysis caused by metastases.82,97,291 In chronic renal failure, HRCT shows multiple 3- to 10-mm calcified nodules or, more commonly, fluffy nodular opacities that mimic airspace nodules but contain foci of calcification291 (Fig. 36-83). A frequent associated finding is calcification in the vessels of the chest wall.

Primary Amyloidosis Primary amyloidosis results from abnormal production and excessive deposition of light chain (AL) protein secreted by B lymphocytes and plasma cells. Pulmonary amyloidosis may appear in three major forms: tracheobronchial, nodular, and diffuse parenchymal. The CT findings of the diffuse parenchymal form include small well-defined nodules (2-4 mm in diameter), abnormal reticular opacities, interlobular septal thickening, and subpleural confluent consolidations38,616 (Fig. 36-84). The nodular form shows solitary or multiple nodules of well-defined rounded or lobular borders. Approximately 50% of nodules calcify or ossify. In patients with Sjögren’s syndrome, the association of cysts and nodules may represent either lymphocytic interstitial pneumonia or pulmonary amyloidosis.49

Disseminated Dendriform Pulmonary Ossification Disseminated dendriform pulmonary ossification, defined as widespread heterotopic bone formation within the lungs, is an infrequent form of diffuse pulmonary ossification.97 It can be idiopathic or associated with a variety of pulmonary, cardiac, and systemic disorders. In chronic pulmonary fibrosis, branching spicules of bone extend through the lung interstitium in a racemose or dendriform manner.301 Disseminated dendriform pulmonary ossification has been associated with UIP.309

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease

A

969

B FIG 36-84  Nodular parenchymal amyloidosis. A, CT scan obtained using lung window settings shows multiple bilateral, randomly distributed pulmonary nodules (arrows). They range in diameter from 0.2 to 4 cm. B, CT scan obtained at different level from A using mediastinal window settings shows that many nodules are calcified. Note small calcified nodules in subpleural location (arrows). (From Marchiori E et al: Diffuse high-attenuation pulmonary abnormalities: A pattern-oriented diagnostic approach on high-resolution CT. AJR Am J Roentgenol 184:273–282, 2005.)

B

A

FIG 36-85  Dendriform pulmonary ossification. A, HRCT through the lower lung zones demonstrates bilateral fine reticular opacities in the subpleural lung regions (arrow). B, Corresponding mediastinal windows reveals multiple punctuate calcifications (arrows) within these opacities, representing dendritic calcification. (From Marchiori E et al: Diffuse high-attenuation pulmonary abnormalities: A pattern-oriented diagnostic approach on high-resolution CT. AJR Am J Roentgenol 184:273–282, 2005.)

Although usually invisible on chest radiographs, HRCT with appropriate windows settings shows tiny calcific opacities in the periphery of the lung82,97 (Fig. 36-85).

Pulmonary Alveolar Microlithiasis Pulmonary alveolar microlithiasis is a rare autosomal-recessive pulmonary disorder of unknown etiology characterized by intraalveolar

development and accumulation of spherical microliths of calcium phosphate.82 Characteristic CT findings consist of innumerable bilateral, small, parenchymal calcified micronodules measuring less than 1 mm in diameter (Fig. 36-86). Other CT features are calcifications along the interlobular septa that sometimes show a crazy-paving pattern and subpleural cysts.175 In patients with long-standing disease the numerous adjacent nodules result in areas of dense consolidations on CT.116

http://radiologyebook.com

970

PART II  CT and MR Imaging of the Whole Body

A

B FIG 36-86  Alveolar microlithiasis. A, HRCT (lung window) shows diffuse scattered micronodules through both lungs (arrows). There is a confluence of nodules in the dependent lung regions. B, HRCT scan at the same level (soft tissue windows) demonstrates numerous dense calcific areas of attenuation in the dependent lung regions, as well as calcific thickening of the interlobular septa (arrows). (From Marchiori E et al: Diffuse high-attenuation pulmonary abnormalities: A pattern-oriented diagnostic approach on high-resolution CT. AJR Am J Roentgenol 184:273–282, 2005.)

REFERENCES 1. Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 18-1988. A 30-year-old man with bilateral pulmonary consolidation and cavitation. 1988. 2. Proceedings of the 1992 International Symposium on Public Health Surveillance. Atlanta, Georgia, April 22-24, 1992. MMWR Morb Mortal Wkly Rep 41(Suppl):1–218, 1992. 3. Hospital-acquired pneumonia in adults: Diagnosis, assessment of severity, initial antimicrobial therapy, and preventive strategies. A consensus statement, American Thoracic Society, November 1995. Am J Respir Crit Care Med 153(5):1711–1725, 1996. 4. Adverse effects of crystalline silica exposure. American Thoracic Society Committee of the Scientific Assembly on Environmental and Occupational Health. Am J Respir Crit Care Med 155(2):761–768, 1997. 5. American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias. This joint statement of the American Thoracic Society (ATS), and the European Respiratory Society (ERS) was adopted by the ATS board of directors, June 2001 and by the ERS Executive Committee, June 2001. Am J Respir Crit Care Med 165(2):277–304, 2002. 6. Adler BD, Padley SP, Muller NL, et al: Chronic hypersensitivity pneumonitis: High-resolution CT and radiographic features in 16 patients. Radiology 185(1):91–95, 1992. 7. Agostini C, Trentin L, Facco M, et al: New aspects of hypersensitivity pneumonitis. Curr Opin Pulm Med 10(5):378–382, 2004. 8. Aguado JM, Arjona R, Ugarte P: Septic pulmonary emboli. A rare cause of bilateral pneumothorax in drug abusers. Chest 98(5):1302–1304, 1990. 9. Akin MR, Nguyen GK: Pulmonary alveolar proteinosis. Pathol Res Pract 200(10):693–698, discussion 699–700, 2004. 10. Akira M: Computed tomography and pathologic findings in fulminant forms of idiopathic interstitial pneumonia. J Thorac Imaging 14(2):76– 84, 1999. 11. Akira M: High-resolution CT in the evaluation of occupational and environmental disease. Radiol Clin North Am 40(1):43–59, 2002. 12. Akira M, Higashihara T, Sakatani M, et al: Diffuse panbronchiolitis: Follow-up CT examination. Radiology 189(2):559–562, 1993. 13. Akira M, Higashihara T, Yokoyama K, et al: Radiographic type p pneumoconiosis: High-resolution CT. Radiology 171(1):117–123, 1989.

14. Akira M, Ishikawa H, Yamamoto S: Drug-induced pneumonitis: Thin-section CT findings in 60 patients. Radiology 224(3):852–860, 2002. 15. Akira M, Kitatani F, Lee YS, et al: Diffuse panbronchiolitis: Evaluation with high-resolution CT. Radiology 168(2):433–438, 1988. 16. Akira M, Yamamoto S, Inoue Y, et al: High-resolution CT of asbestosis and idiopathic pulmonary fibrosis. AJR Am J Roentgenol 181(1):163– 169, 2003. 17. Akira M, Yamamoto S, Yokoyama K, et al: Asbestosis: High-resolution CT-pathologic correlation. Radiology 176(2):389–394, 1990. 18. Akira M, Yokoyama K, Yamamoto S, et al: Early asbestosis: Evaluation with high-resolution CT. Radiology 178(2):409–416, 1991. 19. Alasaly K, Muller N, Ostrow DN, et al: Cryptogenic organizing pneumonia. A report of 25 cases and a review of the literature. Medicine (Baltimore) 74(4):201–211, 1995. 20. Allen J: Acute eosinophilic pneumonia. Semin Respir Crit Care Med 27(2):142–147, 2006. 21. Allen JN, Magro CM, King MA: The eosinophilic pneumonias. Semin Respir Crit Care Med 23(2):127–134, 2002. 22. Allred RJ, Ewer S: Fatal pulmonary edema following intravenous “freebase” cocaine use. Ann Emerg Med 10(8):441–442, 1981. 23. Altes TA, de Lange EE: Applications of hyperpolarized helium-3 gas magnetic resonance imaging in pediatric lung disease. Top Magn Reson Imaging 14(3):231–236, 2003. 24. American Thoracic, S. and S. European Respiratory: American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias. This joint statement of the American Thoracic Society (ATS), and the European Respiratory Society (ERS) was adopted by the ATS board of directors, June 2001 and by the ERS Executive Committee, June 2001. Am J Respir Crit Care Med 165(2):277–304, 2002. 25. Andreu J, Mauleon S, Pallisa E, et al: Miliary lung disease revisited. Curr Probl Diagn Radiol 31(5):189–197, 2002. 26. Aquino SL, Gamsu G, Webb WR, et al: Tree-in-bud pattern: Frequency and significance on thin section CT. J Comput Assist Tomogr 20(4):594– 599, 1996. 27. Aquino SL, Webb WR, Golden J: Bronchiolitis obliterans associated with rheumatoid arthritis: Findings on HRCT and dynamic expiratory CT. J Comput Assist Tomogr 18(4):555–558, 1994. 28. Arakawa H, Gevenois PA, Saito Y, et al: Silicosis: Expiratory thin-section CT assessment of airway obstruction. Radiology 236(3):1059–1066, 2005.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 29. Arakawa H, Kurihara Y, Nakajima Y: Pulmonary fat embolism syndrome: CT findings in six patients. J Comput Assist Tomogr 24(1):24–29, 2000. 30. Arakawa H, Kurihara Y, Niimi H, et al: Bronchiolitis obliterans with organizing pneumonia versus chronic eosinophilic pneumonia: High-resolution CT findings in 81 patients. AJR Am J Roentgenol 176(4):1053–1058, 2001. 31. Arakawa H, Webb WR, McCowin M, et al: Inhomogeneous lung attenuation at thin-section CT: Diagnostic value of expiratory scans. Radiology 206(1):89–94, 1998. 32. Araoz PA, Gotway MB, Harrington JR, et al: Pulmonary embolism: Prognostic CT findings. Radiology 242(3):889–897, 2007. 33. Aronchick JM: Pulmonary infections in cancer and bone marrow transplant patients. Semin Roentgenol 35(2):140–151, 2000. 34. Aronchick JM, Gefter WB: Drug-induced pulmonary disorders. Semin Roentgenol 30(1):18–34, 1995. 35. Avila NA, Dwyer AJ, Rabel A, et al: CT of pleural abnormalities in lymphangioleiomyomatosis and comparison of pleural findings after different types of pleurodesis. AJR Am J Roentgenol 186(4):1007–1012, 2006. 36. Avila NA, Dwyer AJ, Rabel A, et al: Sporadic lymphangioleiomyomatosis and tuberous sclerosis complex with lymphangioleiomyomatosis: Comparison of CT features. Radiology 242(1):277–285, 2007. 37. Aviram G, Fishman JE, Boiselle PM: Thoracic infections in human immunodeficiency virus/acquired immune deficiency syndrome. Semin Roentgenol 42(1):23–36, 2007. 38. Aylwin AC, Gishen P, Copley SJ: Imaging appearance of thoracic amyloidosis. J Thorac Imaging 20(1):41–46, 2005. 39. Ayo DS, Aughenbaugh GL, Yi ES, et al: Cystic lung disease in BirtHogg-Dube syndrome. Chest 132(2):679–684, 2007. 40. Ayzenberg O, Reiff DB, Levin L: Bilateral pneumothoraces and pleural effusions complicating rheumatoid lung disease. Thorax 38(2):159–160, 1983. 41. Aziz ZA, Wells AU, Bateman ED, et al: Interstitial lung disease: Effects of thin-section CT on clinical decision making. Radiology 238(2):725– 733, 2006. 42. Bain GA, Flower CD: Pulmonary eosinophilia. Eur J Radiol 23(1):3–8, 1996. 43. Bains SN, Judson MA: Allergic bronchopulmonary aspergillosis. Clin Chest Med 33(2):265–281, 2012. 44. Balakrishnan J, Meziane MA, Siegelman SS, et al: Pulmonary infarction: CT appearance with pathologic correlation. J Comput Assist Tomogr 13(6):941–945, 1989. 45. Balkman JD, Mehandru S, DuPont E, et al: Dual energy subtraction digital radiography improves performance of a next generation computer-  aided detection program. J Thorac Imaging 25(1):41–47, 2010. 46. Balmes JR: Asbestos and lung cancer: What we know. Am J Respir Crit Care Med 188(1):8–9, 2013. 47. Bankier AA, Fleischmann D, Mallek R, et al: Bronchial wall thickness: Appropriate window settings for thin-section CT and radiologicanatomic correlation. Radiology 199(3):831–836, 1996. 48. Bankier AA, Kiener HP, Wiesmayr MN, et al: Discrete lung involvement in systemic lupus erythematosus: CT assessment. Radiology 196(3):835– 840, 1995. 49. Baqir M, Kluka EM, Aubry MC, et al: Amyloid-associated cystic lung disease in primary Sjogren’s syndrome. Respir Med 107(4):616–621, 2013. 50. Barnett SM: CT findings in tuberculous mediastinitis. J Comput Assist Tomogr 10(1):165–166, 1986. 51. Barrios RJ, Kheradmand F, Batts L, et al: Asthma: Pathology and pathophysiology. Arch Pathol Lab Med 130(4):447–451, 2006. 52. Barry SM, Lipman MC, Johnson MA, et al: Respiratory infections in immunocompromised patients. Curr Opin Pulm Med 5(3):168–173, 1999. 53. Becker CD, Gil J, Padilla ML: Idiopathic pleuroparenchymal fibroelastosis: An unrecognized or misdiagnosed entity? Mod Pathol 21(6):784–787, 2008.

971

54. Beeson PB: Nocardiosis as a complication of pulmonary alveolar proteinosis. Ann Intern Med 60:314, 1964. 55. Begin R, Bergeron D, Samson L, et al: CT assessment of silicosis in exposed workers. AJR Am J Roentgenol 148(3):509–514, 1987. 56. Begin R, Cantin A, Masse S: Recent advances in the pathogenesis and clinical assessment of mineral dust pneumoconioses: Asbestosis, silicosis and coal pneumoconiosis. Eur Respir J 2(10):988–1001, 1989. 57. Begin R, Ostiguy G, Groleau S, et al: Computed tomographic scanning of the thorax in workers at risk of or with silicosis. Semin Ultrasound CT MR 11(5):380–392, 1990. 58. Bellamy EA, Husband JE, Blaquiere RM, et al: Bleomycin-related lung damage: CT evidence. Radiology 156(1):155–158, 1985. 59. Bellomo R, Finlay M, McLaughlin P, et al: Clinical spectrum of cryptogenic organising pneumonitis. Thorax 46(8):554–558, 1991. 60. Benson RE, Rosado-de-Christenson ML, Martinez-Jimenez S, et al: Spectrum of pulmonary neuroendocrine proliferations and neoplasms. Radiographics 33(6):1631–1649, 2013. 61. Bergin C, Castellino RA: Mediastinal lymph node enlargement on CT scans in patients with usual interstitial pneumonitis. AJR Am J Roentgenol 154(2):251–254, 1990. 62. Bergin CJ, Coblentz CL, Chiles C, et al: Chronic lung diseases: Specific diagnosis by using CT. AJR Am J Roentgenol 152(6):1183–1188, 1989. 63. Bergin CJ, Muller NL, Vedal S, et al: CT in silicosis: Correlation with plain films and pulmonary function tests. AJR Am J Roentgenol 146(3):477–483, 1986. 64. Bergin CJ, Rios G, King MA, et al: Accuracy of high-resolution CT in identifying chronic pulmonary thromboembolic disease. AJR Am J Roentgenol 166(6):1371–1377, 1996. 65. Bergin C, Roggli V, Coblentz C, et al: The secondary pulmonary lobule: Normal and abnormal CT appearances. AJR Am J Roentgenol 151(1):21–25, 1988. 66. Bernard GR, Artigas A, Brigham KL, et al: The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 149(3 Pt 1):818–824, 1994. 67. Bhalla M, Naidich DP, McGuinness G, et al: Diffuse lung disease: Assessment with helical CT—Preliminary observations of the role of maximum and minimum intensity projection images. Radiology 200(2):341–347, 1996. 68. Bhalla M, Silver RM, Shepard JA, et al: Chest CT in patients with scleroderma: Prevalence of asymptomatic esophageal dilatation and mediastinal lymphadenopathy. AJR Am J Roentgenol 161(2):269–272, 1993. 69. Blackledge FA, Anand VK: Tracheobronchial extension of recurrent respiratory papillomatosis. Ann Otol Rhinol Laryngol 109(9):812–818, 2000. 70. Bodlet A, Maury G, Jamart J, et al: Influence of radiological emphysema on lung function test in idiopathic pulmonary fibrosis. Respir Med 107(11):1781–1788, 2013. 71. Boiselle PM, Crans CA, Jr, Kaplan MA: The changing face of Pneumocystis carinii pneumonia in AIDS patients. AJR Am J Roentgenol 172(5):1301–1309, 1999. 72. Boiselle PM, Tocino I, Hooley RJ, et al: Chest radiograph interpretation of Pneumocystis carinii pneumonia, bacterial pneumonia, and pulmonary tuberculosis in HIV-positive patients: Accuracy, distinguishing features, and mimics. J Thorac Imaging 12(1):47–53, 1997. 73. Boivin G, Abed Y, Pelletier G, et al: Virological features and clinical manifestations associated with human metapneumovirus: A new paramyxovirus responsible for acute respiratory-tract infections in all age groups. J Infect Dis 186(9):1330–1334, 2002. 74. Bonelli FS, Hartman TE, Swensen SJ, et al: Accuracy of high-resolution CT in diagnosing lung diseases. AJR Am J Roentgenol 170(6):1507–1512, 1998. 75. Bosken CH, Myers JL, Greenberger PA, et al: Pathologic features of allergic bronchopulmonary aspergillosis. Am J Surg Pathol 12(3):216– 222, 1988.

http://radiologyebook.com

972

PART II  CT and MR Imaging of the Whole Body

76. Bouchardy LM, Kuhlman JE, Ball WC, Jr, et al: CT findings in bronchiolitis obliterans organizing pneumonia (BOOP) with radiographic, clinical, and histologic correlation. J Comput Assist Tomogr 17(3):352–357, 1993. 77. Bouza E, Munoz P: Introduction: Infections caused by emerging resistant pathogens. Clin Microbiol Infect 11(s4):iv, 2005. 78. Brauner MW, Lenoir S, Grenier P, et al: Pulmonary sarcoidosis: CT assessment of lesion reversibility. Radiology 182(2):349–354, 1992. 79. Brichet A, Tonnel AB, Brambilla E, et al: Chronic interstitial pneumonia with honeycombing in coal workers. Sarcoidosis Vasc Diffuse Lung Dis 19(3):211–219, 2002. 80. Brillet PY, Dumont P, Bouaziz N, et al: Pulmonary arteriovenous malformation treated with embolotherapy: Systemic collateral supply at multidetector CT angiography after 2-20-year follow-up. Radiology 242(1):267–276, 2007. 81. Brown MJ, Miller RR, Muller NL: Acute lung disease in the immunocompromised host: CT and pathologic examination findings. Radiology 190(1):247–254, 1994. 82. Brown K, Mund DF, Aberle DR, et al: Intrathoracic calcifications: Radiographic features and differential diagnoses. Radiographics 14(6):1247–1261, 1994. 83. Butler S, Smathers RL: Computed tomography of amiodarone pulmonary toxicity. J Comput Assist Tomogr 9(2):375–376, 1985. 84. Cain HC, Noble PW, Matthay RA: Pulmonary manifestations of Sjogren’s syndrome. Clin Chest Med 19(4):687–699, viii, 1998. 85. Cameron DC, Borthwick RN, Philp T: The radiographic patterns of acute mycoplasma pneumonitis. Clin Radiol 28(2):173–180, 1977. 86. Cameron PA, Rainer TH: Update on emerging infections: News from the Centers for Disease Control and Prevention. Update: Outbreak of severe acute respiratory syndrome—Worldwide, 2003. Ann Emerg Med 42(1):110–112, 2003. 87. Camus P, Colby TV: The lung in inflammatory bowel disease. Eur Respir J 15(1):5–10, 2000. 88. Camus P, Fanton A, Bonniaud P, et al: Interstitial lung disease induced by drugs and radiation. Respiration 71(4):301–326, 2004. 89. Cannon GW: Methotrexate pulmonary toxicity. Rheum Dis Clin North Am 23(4):917–937, 1997. 90. Cantin L, Bankier AA, Eisenberg RL: Multiple cystlike lung lesions in the adult. AJR Am J Roentgenol 194(1):W1–W11, 2010. 91. Castaner E, Alguersuari A, Andreu M, et al: Imaging findings in pulmonary vasculitis. Semin Ultrasound CT MR 33(6):567–579, 2012. 92. Castaner E, Gallardo X, Mata JM, et al: Radiologic approach to the diagnosis of infectious pulmonary diseases in patients infected with the human immunodeficiency virus. Eur J Radiol 51(2):114–129, 2004. 93. Castaner E, Gallardo X, Pallardo Y, et al: Diseases affecting the peribronchovascular interstitium: CT findings and pathologic correlation. Curr Probl Diagn Radiol 34(2):63–75, 2005. 94. Cha SI, Fessler MB, Cool CD, et al: Lymphoid interstitial pneumonia: Clinical features, associations and prognosis. Eur Respir J 28(2):364– 369, 2006. 95. Chae EJ, Song JW, Krauss B, et al: Dual-energy computed tomography characterization of solitary pulmonary nodules. J Thorac Imaging 25(4):301–310, 2010. 96. Chan MS, Chan IY, Fung KH, et al: High-resolution CT findings in patients with severe acute respiratory syndrome: A pattern-based approach. AJR Am J Roentgenol 182(1):49–56, 2004. 97. Chan ED, Morales DV, Welsh CH, et al: Calcium deposition with or without bone formation in the lung. Am J Respir Crit Care Med 165(12):1654–1669, 2002. 98. Chandler PW, Shin MS, Friedman SE, et al: Radiographic manifestations of bronchiolitis obliterans with organizing pneumonia vs usual interstitial pneumonia. AJR Am J Roentgenol 147(5):899–906, 1986. 99. Chapman SJ, Cookson WO, Musk AW, et al: Benign asbestos pleural diseases. Curr Opin Pulm Med 9(4):266–271, 2003. 100. Chauhan MS, Tewari SC, Prakash MJ, et al: Necrobiotic pulmonary nodule leading to pyopneumothorax in a case of rheumatoid

101.

102.

103.

104.

105. 106.

107. 108. 109.

110.

111.

112. 113. 114.

115.

116. 117. 118.

119.

120. 121. 122. 123. 124. 125.

arthritis—A rare clinical presentation. Indian J Chest Dis Allied Sci 31(3):217–220, 1989. Cheah FK, Sheppard MN, Hansell DM: Computed tomography of diffuse pulmonary haemorrhage with pathological correlation. Clin Radiol 48(2):89–93, 1993. Cheney PR: Update on emerging infections from the Centers for Disease Control and Prevention. Hantavirus pulmonary syndrome— Colorado and New Mexico, 1998. Ann Emerg Med 33(1):121–123, 1999. Cheon JE, Lee KS, Jung GS, et al: Acute eosinophilic pneumonia: Radiographic and CT findings in six patients. AJR Am J Roentgenol 167(5):1195–1199, 1996. Choe DH, Marom EM, Ahrar K, et al: Pulmonary embolism of polymethyl methacrylate during percutaneous vertebroplasty and kyphoplasty. AJR Am J Roentgenol 183(4):1097–1102, 2004. Chong S, Lee KS, Chung MJ, et al: Pneumoconiosis: Comparison of imaging and pathologic findings. Radiographics 26(1):59–77, 2006. Chow C, Templeton PA, White CS: Lung cysts associated with Pneumocystis carinii pneumonia: Radiographic characteristics, natural history, and complications. AJR Am J Roentgenol 161(3):527–531, 1993. Choyke PL, Sostman HD, Curtis AM, et al: Adult-onset pulmonary tuberculosis. Radiology 148(2):357–362, 1983. Christensen PJ, Kutty K, Adlam RT, et al: Septic pulmonary embolism due to periodontal disease. Chest 104(6):1927–1929, 1993. Chung MH, Edinburgh KJ, Webb EM, et al: Mixed infiltrative and obstructive disease on high-resolution CT: Differential diagnosis and functional correlates in a consecutive series. J Thorac Imaging 16(2):69–75, 2001. Chung JH, Little BP, Forssen AV, et al: Proton MRI in the evaluation of pulmonary sarcoidosis: Comparison to chest CT. Eur J Radiol 82(12):2378–2385, 2013. Chung JH, Pipavath SJ, Myerson DH, et al: Secondary pulmonary alveolar proteinosis: A confusing and potentially serious complication of hematologic malignancy. J Thorac Imaging 24(2):115–118, 2009. Churg A: Pulmonary angiitis and granulomatosis revisited. Hum Pathol 14(10):868–883, 1983. Churg A, Muller NL, Flint J, et al: Chronic hypersensitivity pneumonitis. Am J Surg Pathol 30(2):201–208, 2006. Churg A, Muller NL, Silva CI, et al: Acute exacerbation (acute lung injury of unknown cause) in UIP and other forms of fibrotic interstitial pneumonias. Am J Surg Pathol 31(2):277–284, 2007. Cleverley JR, Screaton NJ, Hiorns MP, et al: Drug-induced lung disease: High-resolution CT and histological findings. Clin Radiol 57(4):292– 299, 2002. Cluzel P, Grenier P, Bernadac P, et al: Pulmonary alveolar microlithiasis: CT findings. J Comput Assist Tomogr 15(6):938–942, 1991. Coche EE, Muller NL, Kim KI, et al: Acute pulmonary embolism: Ancillary findings at spiral CT. Radiology 207(3):753–758, 1998. Coche E, Verschuren F, Hainaut P, et al: Pulmonary embolism findings on chest radiographs and multislice spiral CT. Eur Radiol 14(7):1241– 1248, 2004. Coche E, Weynand B, Noirhomme P, et al: Non-specific interstitial pneumonia showing a “crazy paving” pattern on high resolution CT.   Br J Radiol 74(878):189–191, 2001. Colby TV: Bronchiolitis. Pathologic considerations. Am J Clin Pathol 109(1):101–109, 1998. Coleman A, Colby TV: Histologic diagnosis of extrinsic allergic alveolitis. Am J Surg Pathol 12(7):514–518, 1988. Collard B, Cool M, Hertveldt K, et al: Pulmonary alveolar proteinosis. JBR-BTR 85(5):260–263, 2002. Collins J, Blankenbaker D, Stern EJ: CT patterns of bronchiolar disease: What is “tree-in-bud”? AJR Am J Roentgenol 171(2):365–370, 1998. Collins J, Stern EJ: Ground-glass opacity at CT: The ABCs. AJR Am J Roentgenol 169(2):355–367, 1997. Cooper JD, Billingham M, Egan T, et al: A working formulation for the standardization of nomenclature and for clinical staging of chronic dysfunction in lung allografts. International Society for Heart and Lung Transplantation. J Heart Lung Transplant 12(5):713–716, 1993.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 126. Cooper JA, Jr, White DA, Matthay RA: Drug-induced pulmonary disease. Part 1: Cytotoxic drugs. Am Rev Respir Dis 133(2):321–340, 1986. 127. Cooper JA, Jr, White DA, Matthay RA: Drug-induced pulmonary disease. Part 2: Noncytotoxic drugs. Am Rev Respir Dis 133(3):488–505, 1986. 128. Copley SJ, Lee YC, Hansell DM, et al: Asbestos-induced and smokingrelated disease: Apportioning pulmonary function deficit by using thin-section CT. Radiology 242(1):258–266, 2007. 129. Copley SJ, Padley SP: High-resolution CT of paediatric lung disease. Eur Radiol 11(12):2564–2575, 2001. 130. Copley SJ, Wells AU, Sivakumaran P, et al: Asbestosis and idiopathic pulmonary fibrosis: Comparison of thin-section CT features. Radiology 229(3):731–736, 2003. 131. Cordier JF: Bronchiolitis obliterans organizing pneumonia. Semin Respir Crit Care Med 21(2):135–146, 2000. 132. Cordier JF: Cryptogenic organizing pneumonia. Clin Chest Med 25(4):727–738, vi-vii, 2004. 133. Corrin B, Liebow AA, Friedman PJ: Pulmonary lymphangiomyomatosis. A review. Am J Pathol 79(2):348–382, 1975. 134. Cortet B, Flipo RM, Remy-Jardin M, et al: Use of high resolution computed tomography of the lungs in patients with rheumatoid arthritis. Ann Rheum Dis 54(10):815–819, 1995. 135. Costello P, Anderson W, Blume D: Pulmonary nodule: Evaluation with spiral volumetric CT. Radiology 179(3):875–876, 1991. 136. Cottin V, Cordier JF: The syndrome of combined pulmonary fibrosis and emphysema. Chest 136(1):1–2, 2009. 137. Cottin V, Donsbeck AV, Revel D, et al: Nonspecific interstitial pneumonia. Individualization of a clinicopathologic entity in a series of 12 patients. Am J Respir Crit Care Med 158(4):1286–1293, 1998. 138. Cottin V, Le Pavec J, Prevot G, et al: Pulmonary hypertension in patients with combined pulmonary fibrosis and emphysema syndrome. Eur Respir J 35(1):105–111, 2010. 139. Cottin V, Nunes H, Brillet PY, et al: Combined pulmonary fibrosis and emphysema: A distinct underrecognised entity. Eur Respir J 26(4):586– 593, 2005. 140. Crofton JW, Livingstone JL, Oswald NC, et al: Pulmonary eosinophilia. Thorax 7(1):1–35, 1952. 141. Cucco RA, Yoo OH, Cregler L, et al: Nonfatal pulmonary edema after “freebase” cocaine smoking. Am Rev Respir Dis 136(1):179–181, 1987. 142. Cullen MR, Barnett MJ, Balmes JR, et al: Predictors of lung cancer among asbestos-exposed men in the {beta}-carotene and retinol efficacy trial. Am J Epidemiol 161(3):260–270, 2005. 143. Cunningham I: Pulmonary infections after bone marrow transplant. Semin Respir Infect 7(2):132–138, 1992. 144. D’Angelo WA, Fries JF, Masi AT, et al: Pathologic observations in systemic sclerosis (scleroderma). A study of fifty-eight autopsy cases and fifty-eight matched controls. Am J Med 46(3):428–440, 1969. 145. Dabydeen DA, Jagannathan JP, Ramaiya N, et al: Pneumonitis associated with mTOR inhibitors therapy in patients with metastatic renal cell carcinoma: Incidence, radiographic findings and correlation with clinical outcome. Eur J Cancer 48(10):1519–1524, 2012. 146. Daimon T, Johkoh T, Sumikawa H, et al: Acute eosinophilic pneumonia: Thin-section CT findings in 29 patients. Eur J Radiol 2007. 147. Dallari R, Foglia M, Paci M, et al: Acute exacerbation of idiopathic pulmonary fibrosis. Eur Respir J 23(5):792, 2004. 148. Davies SJ, Gosney JR, Hansell DM, et al: Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia: An under-recognised spectrum of disease. Thorax 62(3):248–252, 2007. 149. Davies CW, Tasker AD, Padley SP, et al: Air trapping in sarcoidosis on computed tomography: Correlation with lung function. Clin Radiol 55(3):217–221, 2000. 150. Dawson WB, Muller NL: High-resolution computed tomography in pulmonary sarcoidosis. Semin Ultrasound CT MR 11(5):423–429, 1990. 151. de Lange EE, Altes TA, Patrie JT, et al: Evaluation of asthma with hyperpolarized helium-3 MRI: Correlation with clinical severity and spirometry. Chest 130(4):1055–1062, 2006.

973

152. De Vuyst P, Gevenois PA, Van Muylem A, et al: Changing patterns in asbestos-induced lung disease. Chest 126(3):999, 2004. 153. Desai SR, Hansell DM: Small airways disease: Expiratory computed tomography comes of age. Clin Radiol 52(5):332–337, 1997. 154. Desai SR, Wells AU, Rubens MB, et al: Traction bronchiectasis in cryptogenic fibrosing alveolitis: Associated computed tomographic features and physiological significance. Eur Radiol 13(8):1801–1808, 2003. 155. Desai SR, Wells AU, Rubens MB, et al: Acute respiratory distress syndrome: CT abnormalities at long-term follow-up. Radiology 210(1):29–35, 1999. 156. Dhillon SS, Watanakunakorn C: Lady Windermere syndrome: Middle lobe bronchiectasis and Mycobacterium avium complex infection due to voluntary cough suppression. Clin Infect Dis 30(3):572–575, 2000. 157. Dietrich PA, Johnson RD, Fairbank JT, et al: The chest radiograph in legionnaires’ disease. Radiology 127(3):577–582, 1978. 158. Dodd JD, Souza CA, Muller NL: High-resolution MDCT of pulmonary septic embolism: Evaluation of the feeding vessel sign. AJR Am J Roentgenol 187(3):623–629, 2006. 159. Doyle TC, Lawler GA: CT features of rounded atelectasis of the lung. AJR Am J Roentgenol 143(2):225–228, 1984. 160. Drosten C, Gunther S, Preiser W, et al: Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 348(20):1967–1976, 2003. 161. Eggli KD, Newman B: Nodules, masses, and pseudomasses in the pediatric lung. Radiol Clin North Am 31(3):651–666, 1993. 162. Einsele H, Quabeck K, Muller KD, et al: Prediction of invasive pulmonary aspergillosis from colonisation of lower respiratory tract before marrow transplantation. Lancet 352(9138):1443, 1998. 163. Eisenhuber E: The tree-in-bud sign. Radiology 222(3):771–772, 2002. 164. Ellis SJ, Cleverley JR, Muller NL: Drug-induced lung disease: Highresolution CT findings. AJR Am J Roentgenol 175(4):1019–1024, 2000. 165. Epler GR: Bronchiolitis obliterans organizing pneumonia: Definition and clinical features. Chest 102(1 Suppl):2S–6S, 1992. 166. Erasmus JJ, McAdams HP, Farrell MA, et al: Pulmonary nontuberculous mycobacterial infection: Radiologic manifestations. Radiographics 19(6):1487–1505, 1999. 167. Erasmus JJ, McAdams HP, Rossi SE: Drug-induced lung injury. Semin Roentgenol 37(1):72–81, 2002. 168. Erasmus JJ, McAdams HP, Rossi SE: High-resolution CT of druginduced lung disease. Radiol Clin North Am 40(1):61–72, 2002. 169. Erly WK, Borders RJ, Outwater EK, et al: Location, size, and distribution of mediastinal lymph node enlargement in chronic congestive heart failure. J Comput Assist Tomogr 27(4):485–489, 2003. 170. Ettinger NA, Trulock EP: Pulmonary considerations of organ transplantation. Part 2. Am Rev Respir Dis 144(1):213–223, 1991. 171. Ettinger NA, Trulock EP: Pulmonary considerations of organ transplantation. Part 3. Am Rev Respir Dis 144(2):433–451, 1991. 172. Ettinger NA, Trulock EP: Pulmonary considerations of organ transplantation. Part I. Am Rev Respir Dis 143(6):1386–1405, 1991. 173. Fenlon HM, Casserly I, Sant SM, et al: Plain radiographs and thoracic high-resolution CT in patients with ankylosing spondylitis. AJR Am J Roentgenol 168(4):1067–1072, 1997. 174. Fenlon HM, Doran M, Sant SM, et al: High-resolution chest CT in systemic lupus erythematosus. AJR Am J Roentgenol 166(2):301–307, 1996. 175. Ferreira Francisco FA, Pereira e Silva JL, Hochhegger B, et al: Pulmonary alveolar microlithiasis. State-of-the-art review. Respir Med 107(1):1–9, 2013. 176. Filice GA: Nocardiosis in persons with human immunodeficiency virus infection, transplant recipients, and large, geographically defined populations. J Lab Clin Med 145(3):156–162, 2005. 177. Fishman JE, Rabkin JM: Thoracic radiology in kidney and liver transplantation. J Thorac Imaging 17(2):122–131, 2002. 178. Fishman JA, Rubin RH: Infection in organ-transplant recipients. N Engl J Med 338(24):1741–1751, 1998.

http://radiologyebook.com

974

PART II  CT and MR Imaging of the Whole Body

179. Fishman JE, Schwartz DS, Sais GJ: Mycobacterium kansasii pulmonary infection in patients with AIDS: Spectrum of chest radiographic findings. Radiology 204(1):171–175, 1997. 180. Flaherty KR, Andrei AC, King TE, Jr, et al: Idiopathic interstitial pneumonia: Do community and academic physicians agree on diagnosis? Am J Respir Crit Care Med 175(10):1054–1060, 2007. 181. Flaherty KR, Toews GB, Travis WD, et al: Clinical significance of histological classification of idiopathic interstitial pneumonia. Eur Respir J 19(2):275–283, 2002. 182. Flaherty KR, Travis WD, Colby TV, et al: Histopathologic variability in usual and nonspecific interstitial pneumonias. Am J Respir Crit Care Med 164(9):1722–1727, 2001. 183. Flieder DB, Travis WD: Pathologic characteristics of drug-induced lung disease. Clin Chest Med 25(1):37–45, 2004. 184. Force ADT, Ranieri VM, Rubenfeld GD, et al: Acute respiratory distress syndrome: The Berlin definition. JAMA 307(23):2526–2533, 2012. 185. Forster BB, Muller NL, Mayo JR, et al: High-resolution computed tomography of experimental hydrostatic pulmonary edema. Chest 101(5):1434–1437, 1992. 186. Fraig M, Shreesha U, Savici D, et al: Respiratory bronchiolitis: A clinicopathologic study in current smokers, ex-smokers, and neversmokers. Am J Surg Pathol 26(5):647–653, 2002. 187. Frankel SK, Cool CD, Lynch DA, et al: Idiopathic pleuroparenchymal fibroelastosis: Description of a novel clinicopathologic entity. Chest 126(6):2007–2013, 2004. 188. Franquet T: Imaging of pneumonia: Trends and algorithms. Eur Respir J 18(1):196–208, 2001. 189. Franquet T, Gimenez A, Bordes R, et al: The crazy-paving pattern in exogenous lipoid pneumonia: CT-pathologic correlation. AJR Am J Roentgenol 170(2):315–317, 1998. 190. Franquet T, Gimenez A, Monill JM, et al: Primary Sjogren’s syndrome and associated lung disease: CT findings in 50 patients. AJR Am J Roentgenol 169(3):655–658, 1997. 191. Franquet T, Gimenez A, Prats R, et al: Thrombotic microangiopathy of pulmonary tumors: A vascular cause of tree-in-bud pattern on CT. AJR Am J Roentgenol 179(4):897–899, 2002. 192. Franquet T, Gimenez A, Roson N, et al: Aspiration diseases: Findings, pitfalls, and differential diagnosis. Radiographics 20(3):673–685, 2000. 193. Franquet T, Hansell DM, Senbanjo T, et al: Lung cysts in subacute hypersensitivity pneumonitis. J Comput Assist Tomogr 27(4):475–478, 2003. 194. Franquet T, Muller NL: Disorders of the small airways: High-resolution computed tomographic features. Semin Respir Crit Care Med 24(4):437–444, 2003. 195. Franquet T, Muller NL, Flint JD: A patient with ankylosing spondylitis and recurrent haemoptysis. Eur Respir J 23(3):488–491, 2004. 196. Franquet T, Plaza V, Llauger J, et al: Hydatid pulmonary embolism from a ruptured mediastinal cyst: High-resolution computed tomography, angiographic, and pathologic findings. J Thorac Imaging 14(2):138–141, 1999. 197. Franquet T, Rodriguez S, Martino R, et al: Human metapneumovirus infection in hematopoietic stem cell transplant recipients: Highresolution computed tomography findings. J Comput Assist Tomogr 29(2):223–227, 2005. 198. Franquet T, Stern EJ: Bronchiolar inflammatory diseases: Highresolution CT findings with histologic correlation. Eur Radiol 9(7):1290–1303, 1999. 199. Franquet T, Stern EJ, Gimenez A, et al: Lateral decubitus CT: A useful adjunct to standard inspiratory-expiratory CT for the detection of air-trapping. AJR Am J Roentgenol 174(2):528–530, 2000. 200. Frazier AA, Rosado-de-Christenson ML, Galvin JR, et al: Pulmonary angiitis and granulomatosis: Radiologic-pathologic correlation. Radiographics 18(3):687–710, quiz 727, 1998. 201. Fujii M, Adachi S, Shimizu T, et al: Interstitial lung disease in rheumatoid arthritis: Assessment with high-resolution computed tomography. J Thorac Imaging 8(1):54–62, 1993. 202. Fujimura N: Pathology and pathophysiology of pneumoconiosis. Curr Opin Pulm Med 6(2):140–144, 2000.

203. Fukuda Y, Kawamoto M, Yamamoto A, et al: Role of elastic fiber degradation in emphysema-like lesions of pulmonary lymphangiomyomatosis. Hum Pathol 21(12):1252–1261, 1990. 204. Furuya M, Nakatani Y: Birt-Hogg-Dube syndrome: Clinicopathological features of the lung. J Clin Pathol 66(3):178–186, 2013. 205. Gaeta M, Blandino A, Scribano E, et al: Computed tomography halo sign in pulmonary nodules: Frequency and diagnostic value. J Thorac Imaging 14(2):109–113, 1999. 206. Gal AA, Staton GW, Jr: Current concepts in the classification of interstitial lung disease. Am J Clin Pathol 123(Suppl):S67–S81, 2005. 207. Gale N, Poljak M, Kambic V, et al: Laryngeal papillomatosis: Molecular, histopathological, and clinical evaluation. Virchows Arch 425(3):291– 295, 1994. 208. Gallagher H, Kwan JT, Jayne DR: Pulmonary renal syndrome:   A 4-year, single-center experience. Am J Kidney Dis 39(1):42–47, 2002. 209. Gallardo X, Castaner E, Mata JM, et al: Nodular pattern at lung computed tomography in fat embolism syndrome: A helpful finding.   J Comput Assist Tomogr 30(2):254–257, 2006. 210. Gamsu G: Radiographic manifestations of thoracic involvement by collagen vascular diseases. J Thorac Imaging 7(3):1–12, 1992. 211. Garrean S, Massad MG, Tshibaka M, et al: Sirolimus-associated interstitial pneumonitis in solid organ transplant recipients. Clin Transplant 19(5):698–703, 2005. 212. Gasparetto EL, Souza CA, Tazoniero P, et al: Angioinvasive pulmonary aspergillosis after allogeneic bone marrow transplantation: Clinical and high-resolution computed tomography findings in 12 cases. Braz J Infect Dis 11(1):110–113, 2007. 213. Gasparetto EL, Tazoniero P, Escuissato DL, et al: Pulmonary alveolar microlithiasis presenting with crazy-paving pattern on high resolution CT. Br J Radiol 77(923):974–976, 2004. 214. Gattinoni L, Caironi P, Pelosi P, et al: What has computed tomography taught us about the acute respiratory distress syndrome? Am J Respir Crit Care Med 164(9):1701–1711, 2001. 215. Gefter WB, Albelda SM, Talbot GH, et al: Invasive pulmonary aspergillosis and acute leukemia. Limitations in the diagnostic utility of the air crescent sign. Radiology 157(3):605–610, 1985. 216. Ghanima W, Nielssen BE, Holmen LO, et al: Multidetector computed tomography (MDCT) in the diagnosis of pulmonary embolism: Interobserver agreement among radiologists with varied levels of experience. Acta Radiol 48(2):165–170, 2007. 217. Gharib AM, Stern EJ: Radiology of pneumonia. Med Clin North Am 85(6):1461–1491, x, 2001. 218. Ghaye B: Peripheral pulmonary embolism on multidetector CT pulmonary angiography. JBR-BTR 90(2):100–108, 2007. 219. Ghaye B, Szapiro D, Mastora I, et al: Peripheral pulmonary arteries: How far in the lung does multi-detector row spiral CT allow analysis? Radiology 219(3):629–636, 2001. 220. Glazer HS, Levitt RG, Shackelford GD: Peripheral pulmonary infiltrates in sarcoidosis. Chest 86(5):741–744, 1984. 221. Godoy MC, Viswanathan C, Marchiori E, et al: The reversed halo sign: Update and differential diagnosis. Br J Radiol 85(1017):1226–1235, 2012. 222. Godwin JD, Muller NL, Takasugi JE: Pulmonary alveolar proteinosis: CT findings. Radiology 169(3):609–613, 1988. 223. Goh JS, Tsou IY, Kaw GJ: Severe acute respiratory syndrome (SARS): Imaging findings during the acute and recovery phases of disease.   J Thorac Imaging 18(3):195–199, 2003. 224. Goodman LR, Fumagalli R, Tagliabue P, et al: Adult respiratory distress syndrome due to pulmonary and extrapulmonary causes: CT, clinical, and functional correlations. Radiology 213(2):545–552, 1999. 225. Gosselin MV, Rassner UA, Thieszen SL, et al: Contrast dynamics during CT pulmonary angiogram: Analysis of an inspiration associated artifact. J Thorac Imaging 19(1):1–7, 2004. 226. Greaves SM, Hart EM, Brown K, et al: Pulmonary thromboembolism: Spectrum of findings on CT. AJR Am J Roentgenol 165(6):1359–1363, 1995.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 227. Greene RE, Schlamm HT, Oestmann JW, et al: Imaging findings in acute invasive pulmonary aspergillosis: Clinical significance of the halo sign. Clin Infect Dis 44(3):373–379, 2007. 228. Grenier P, Valeyre D, Cluzel P, et al: Chronic diffuse interstitial lung disease: Diagnostic value of chest radiography and high-resolution CT. Radiology 179(1):123–132, 1991. 229. Grossman H, Merten DF, Spock A, et al: Radiographic features of sarcoidosis in pediatric patients. Semin Roentgenol 20(4):393–399, 1985. 230. Gruden JF, Huang L, Turner J, et al: High-resolution CT in the evaluation of clinically suspected Pneumocystis carinii pneumonia in AIDS patients with normal, equivocal, or nonspecific radiographic findings. AJR Am J Roentgenol 169(4):967–975, 1997. 231. Gruden JF, Webb WR, Sides DM: Adult-onset disseminated tracheobronchial papillomatosis: CT features. J Comput Assist Tomogr 18(4):640–642, 1994. 232. Gudbjerg CE: Roentgenologic diagnosis of bronchiectasis; an analysis of 112 cases. Acta Radiol 43(3):210–226, 1955. 233. Gudbjerg CE: Bronchiectasis; radiological diagnosis and prognosis after operative treatment. Acta Radiol Suppl 143:1–146, 1957. 234. Gurney JW, Jones KK, Robbins RA, et al: Regional distribution of emphysema: Correlation of high-resolution CT with pulmonary function tests in unselected smokers. Radiology 183(2):457–463, 1992. 235. Hamelin ME, Abed Y, Boivin G: Human metapneumovirus: A new player among respiratory viruses. Clin Infect Dis 38(7):983–990, 2004. 236. Hammel JM, Chiang WK: Update on emerging infections: News from the Centers for Disease Control and Prevention. Outbreaks of avian influenza A (H5N1) in Asia and interim recommendations for evaluation and reporting of suspected cases—United States, 2004. Ann Emerg Med 45(1):88–92, 2005. 237. Hamper UM, Fishman EK, Khouri NF, et al: Typical and atypical CT manifestations of pulmonary sarcoidosis. J Comput Assist Tomogr 10(6):928–936, 1986. 238. Han D, Lee KS, Franquet T, et al: Thrombotic and nonthrombotic pulmonary arterial embolism: Spectrum of imaging findings. Radiographics 23(6):1521–1539, 2003. 239. Hanke R, Kretzschmar R: Round atelectasis. Semin Roentgenol 15(2):174–182, 1980. 240. Hansell DM: Bronchiectasis. Radiol Clin North Am 36(1):107–128, 1998. 241. Hansell DM: HRCT of obliterative bronchiolitis and other small airways diseases. Semin Roentgenol 36(1):51–65, 2001. 242. Hansell DM: Small airways diseases: Detection and insights with computed tomography. Eur Respir J 17(6):1294–1313, 2001. 243. Hansell DM: Small-vessel diseases of the lung: CT-pathologic correlates. Radiology 225(3):639–653, 2002. 244. Hansell DM, Bankier AA, MacMahon H, et al: Fleischner Society: Glossary of terms for thoracic imaging. Radiology 246(3):697–722, 2008. 245. Hansell DM, Milne DG, Wilsher ML, et al: Pulmonary sarcoidosis: Morphologic associations of airflow obstruction at thin-section CT. Radiology 209(3):697–704, 1998. 246. Hansell DM, Strickland B: High-resolution computed tomography in pulmonary cystic fibrosis. Br J Radiol 62(733):1–5, 1989. 247. Hansell DM, Wells AU, Padley SP, et al: Hypersensitivity pneumonitis: Correlation of individual CT patterns with functional abnormalities. Radiology 199(1):123–128, 1996. 248. Hansell DM, Wells AU, Rubens MB, et al: Bronchiectasis: Functional significance of areas of decreased attenuation at expiratory CT. Radiology 193(2):369–374, 1994. 249. Haramati LB, Jenny-Avital ER, Alterman DD: Effect of HIV status on chest radiographic and CT findings in patients with tuberculosis. Clin Radiol 52(1):31–35, 1997. 250. Harisinghani MG, McLoud TC, Shepard JA, et al: Tuberculosis from head to toe. Radiographics 20(2):449–470, quiz 528-449, 532, 2000. 251. Hartman TE: The HRCT features of extrinsic allergic alveolitis. Semin Respir Crit Care Med 24(4):419–426, 2003. 252. Hartman TE, Swensen SJ, Hansell DM, et al: Nonspecific interstitial pneumonia: Variable appearance at high-resolution chest CT. Radiology 217(3):701–705, 2000.

975

253. Hartman TE, Tazelaar HD, Swensen SJ, et al: Cigarette smoking: CT and pathologic findings of associated pulmonary diseases. Radiographics 17(2):377–390, 1997. 254. Hatani T, Takemura M, Inoue D, et al: Septic pulmonary embolism due to periodontal disease. Respirology 18(2):308–312, 2013. 255. Hayakawa H, Sato A, Imokawa S, et al: Diffuse panbronchiolitis and rheumatoid arthritis-associated bronchiolar disease: Similarities and differences. Intern Med 37(6):504–508, 1998. 256. Hayakawa H, Sato A, Imokawa S, et al: Bronchiolar disease in rheumatoid arthritis. Am J Respir Crit Care Med 154(5):1531–1536, 1996. 257. Hayakawa H, Shirai M, Sato A, et al: Clinicopathological features of chronic hypersensitivity pneumonitis. Respirology 7(4):359–364, 2002. 258. Heitzman ER, Markarian B, Berger I, et al: The secondary pulmonary lobule: A practical concept for interpretation of chest radiographs. II. Application of the anatomic concept to an understanding of roentgen pattern in disease states. Radiology 93(3):513–519, 1969. 259. Helmers R, Galvin J, Hunninghake GW: Pulmonary manifestations associated with rheumatoid arthritis. Chest 100(1):235–238, 1991. 260. Herman SJ: Radiologic assessment after lung transplantation. Radiol Clin North Am 32(4):663–678, 1994. 261. Heyneman LE, Muller NL: Pulmonary nodules in early fat embolism syndrome: A case report. J Thorac Imaging 15(1):71–74, 2000. 262. Ho AS, Ahmed A, Huang JS, et al: Multidetector computed tomography of spontaneous versus secondary pneumomediastinum in 89 patients: Can multidetector computed tomography be used to reliably distinguish between the 2 entities? J Thorac Imaging 27(2):85–92, 2012. 263. Honda O, Johkoh T, Ichikado K, et al: Comparison of high resolution CT findings of sarcoidosis, lymphoma, and lymphangitic carcinoma: Is there any difference of involved interstitium? J Comput Assist Tomogr 23(3):374–379, 1999. 264. Honda O, Johkoh T, Tomiyama N, et al: High-resolution CT using multidetector CT equipment: Evaluation of image quality in 11 cadaveric lungs and a phantom. AJR Am J Roentgenol 177(4):875–879, 2001. 265. Honda O, Johkoh T, Yamamoto S, et al: Comparison of quality of multiplanar reconstructions and direct coronal multidetector CT scans of the lung. AJR Am J Roentgenol 179(4):875–879, 2002. 266. Howling SJ, Hansell DM, Wells AU, et al: Follicular bronchiolitis: Thin-section CT and histologic findings. Radiology 212(3):637–642, 1999. 267. Huang RM, Naidich DP, Lubat E, et al: Septic pulmonary emboli: CT-radiographic correlation. AJR Am J Roentgenol 153(1):41–45, 1989. 268. Hui JY, Cho DH, Yang MK, et al: Severe acute respiratory syndrome: Spectrum of high-resolution CT findings and temporal progression of the disease. AJR Am J Roentgenol 181(6):1525–1538, 2003. 269. Hunninghake GW, Fauci AS: Pulmonary involvement in the collagen vascular diseases. Am Rev Respir Dis 119(3):471–503, 1979. 270. Huston J, 3rd, Muhm JR: Solitary pulmonary nodules: Evaluation with a CT reference phantom. Radiology 170(3 Pt 1):653–656, 1989. 271. Im JG, Choi YW, Kim HD, et al: Thin-section CT findings of the lungs: Experimentally induced bronchial and pulmonary artery obstruction in pigs. AJR Am J Roentgenol 167(3):631–636, 1996. 272. Im JG, Itoh H, Han MC: CT of pulmonary tuberculosis. Semin Ultrasound CT MR 16(5):420–434, 1995. 273. Im JG, Itoh H, Lee KS, et al: CT-pathology correlation of pulmonary tuberculosis. Crit Rev Diagn Imaging 36(3):227–285, 1995. 274. Im JG, Itoh H, Shim YS, et al: Pulmonary tuberculosis: CT findings— Early active disease and sequential change with antituberculous therapy. Radiology 186(3):653–660, 1993. 275. Im JG, Song KS, Kang HS, et al: Mediastinal tuberculous lymphadenitis: CT manifestations. Radiology 164(1):115–119, 1987. 276. Imokawa S, Colby TV, Leslie KO, et al: Methotrexate pneumonitis: Review of the literature and histopathological findings in nine patients. Eur Respir J 15(2):373–381, 2000. 277. Irwin RS, Curley FJ, French CL: Chronic cough. The spectrum and frequency of causes, key components of the diagnostic evaluation, and outcome of specific therapy. Am Rev Respir Dis 141(3):640–647, 1990.

http://radiologyebook.com

976

PART II  CT and MR Imaging of the Whole Body

278. Ito I, Nagai S, Kitaichi M, et al: Pulmonary manifestations of primary Sjogren’s syndrome: A clinical, radiologic, and pathologic study. Am J Respir Crit Care Med 171(6):632–638, 2005. 279. Itoh H, Murata K, Konishi J, et al: Diffuse lung disease: Pathologic basis for the high-resolution computed tomography findings. J Thorac Imaging 8(3):176–188, 1993. 280. Iwasaki Y, Nagata K, Nakanishi M, et al: Spiral CT findings in septic pulmonary emboli. Eur J Radiol 37(3):190–194, 2001. 281. Iwata M, Colby TV, Kitaichi M: Diffuse panbronchiolitis: Diagnosis and distinction from various pulmonary diseases with centrilobular interstitial foam cell accumulations. Hum Pathol 25(4):357–363, 1994. 282. Janzen DL, Padley SP, Adler BD, et al: Acute pulmonary complications in immunocompromised non-AIDS patients: Comparison of diagnostic accuracy of CT and chest radiography. Clin Radiol 47(3):159–165, 1993. 283. Javidan-Nejad C, Bhalla S: Bronchiectasis. Radiol Clin North Am 47(2):289–306, 2009. 284. Jennette JC, Falk RJ, Andrassy K, et al: Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 37(2):187–192, 1994. 285. Jeong YJ, Kim KI, Seo IJ, et al: Eosinophilic lung diseases: A clinical, radiologic, and pathologic overview. Radiographics 27(3):617–637, discussion 637-619, 2007. 286. Jeong YJ, Lee KS, Chung MP, et al: Amyloidosis and lymphoproliferative disease in Sjogren syndrome: Thin-section computed tomography findings and histopathologic comparisons. J Comput Assist Tomogr 28(6):776–781, 2004. 287. Jeong YJ, Lee KS, Chung MP, et al: Chronic hypersensitivity pneumonitis and pulmonary sarcoidosis: Differentiation from usual interstitial pneumonia using high-resolution computed tomography. Semin Ultrasound CT MR 35(1):47–58, 2014. 288. Jeudy J, White CS: Evaluation of acute chest pain in the emergency department: Utility of multidetector computed tomography. Semin Ultrasound CT MR 28(2):109–114, 2007. 289. Johkoh T, Honda O, Yamamoto S, et al: Evaluation of image quality and spatial resolution of low-dose high-pitch multidetector-row helical high-resolution CT in 11 autopsy lungs and a wire phantom. Radiat Med 19(6):279–284, 2001. 290. Johkoh T, Ichikado K, Akira M, et al: Lymphocytic interstitial pneumonia: Follow-up CT findings in 14 patients. J Thorac Imaging 15(3):162–167, 2000. 291. Johkoh T, Ikezoe J, Nagareda T, et al: Metastatic pulmonary calcification: Early detection by high-resolution CT. J Comput Assist Tomogr 17(3):471–473, 1993. 292. Johkoh T, Itoh H, Muller NL, et al: Crazy-paving appearance at thin-section CT: Spectrum of disease and pathologic findings. Radiology 211(1):155–160, 1999. 293. Johkoh T, Muller NL, Akira M, et al: Eosinophilic lung diseases: Diagnostic accuracy of thin-section CT in 111 patients. Radiology 216(3):773–780, 2000. 294. Johkoh T, Muller NL, Cartier Y, et al: Idiopathic interstitial pneumonias: Diagnostic accuracy of thin-section CT in 129 patients. Radiology 211(2):555–560, 1999. 295. Johkoh T, Muller NL, Colby TV, et al: Nonspecific interstitial pneumonia: Correlation between thin-section CT findings and pathologic subgroups in 55 patients. Radiology 225(1):199–204, 2002. 296. Johkoh T, Muller NL, Pickford HA, et al: Lymphocytic interstitial pneumonia: Thin-section CT findings in 22 patients. Radiology 212(2):567–572, 1999. 297. Johkoh T, Muller NL, Taniguchi H, et al: Acute interstitial pneumonia: Thin-section CT findings in 36 patients. Radiology 211(3):859–863, 1999. 298. Jones JS: An account of pleural effusions, pulmonary nodules and cavities attributable to rheumatoid disease. Br J Dis Chest 72(1):39–56, 1978. 299. Kalender WA, Seissler W, Klotz E, et al: Spiral volumetric CT with single-breath-hold technique, continuous transport, and continuous scanner rotation. Radiology 176(1):181–183, 1990.

300. Kang EY, Shim JJ, Kim JS, et al: Pulmonary involvement of idiopathic hypereosinophilic syndrome: CT findings in five patients. J Comput Assist Tomogr 21(4):612–615, 1997. 301. Kanne JP, Godwin JD, Takasugi JE, et al: Diffuse pulmonary ossification. J Thorac Imaging 19(2):98–102, 2004. 302. Kanne JP, Gotway MB, Thoongsuwan N, et al: Six cases of acute central pulmonary embolism revealed on unenhanced multidetector CT of the chest. AJR Am J Roentgenol 180(6):1661–1664, 2003. 303. Kantor HG: The many radiologic facies of pneumococcal pneumonia. AJR Am J Roentgenol 137(6):1213–1220, 1981. 304. Katzenstein AL, Fiorelli RF: Nonspecific interstitial pneumonia/fibrosis. Histologic features and clinical significance. Am J Surg Pathol 18(2):136–147, 1994. 305. Kavuru M, Ney D, Fishman EK, et al: Three-dimensional imaging of the lung in vivo: Work in progress. J Digit Imaging 4(3):137–142, 1991. 306. Khan A, Herman PG, Vorwerk P, et al: Solitary pulmonary nodules: Comparison of classification with standard, thin-section, and reference phantom CT. Radiology 179(2):477–481, 1991. 307. Khan MA, Mathur AP, Allen A, et al: Tracheobronchial papillomatosis. N Y State J Med 77(13):2073–2077, 1977. 308. Kiely JL, Spense L, Henry M, et al: Chest radiographic staging in allergic bronchopulmonary aspergillosis: Relationship with immunological findings. Eur Respir J 12(2):453–456, 1998. 309. Kim TS, Han J, Chung MP, et al: Disseminated dendriform pulmonary ossification associated with usual interstitial pneumonia: Incidence and thin-section CT-pathologic correlation. Eur Radiol 15(8):1581–1585, 2005. 310. Kim Y, Lee KS, Choi DC, et al: The spectrum of eosinophilic lung disease: Radiologic findings. J Comput Assist Tomogr 21(6):920–930, 1997. 311. Kim EY, Lee KS, Chung MP, et al: Nonspecific interstitial pneumonia with fibrosis: Serial high-resolution CT findings with functional correlation. AJR Am J Roentgenol 173(4):949–953, 1999. 312. Kim JS, Lee KS, Koh EM, et al: Thoracic involvement of systemic lupus erythematosus: Clinical, pathologic, and radiologic findings. J Comput Assist Tomogr 24(1):9–18, 2000. 313. Kim SJ, Lee KS, Ryu YH, et al: Reversed halo sign on high-resolution CT of cryptogenic organizing pneumonia: Diagnostic implications. AJR Am J Roentgenol 180(5):1251–1254, 2003. 314. Kim DS, Park JH, Park BK, et al: Acute exacerbation of idiopathic pulmonary fibrosis: Frequency and clinical features. Eur Respir J 27(1):143–150, 2006. 315. King MA, Ysrael M, Bergin CJ: Chronic thromboembolic pulmonary hypertension: CT findings. AJR Am J Roentgenol 170(4):955–960, 1998. 316. Kinoshita M, Higashi T, Tanaka C, et al: Follicular bronchiolitis associated with rheumatoid arthritis. Intern Med 31(5):674–677, 1992. 317. Kirks DR, V., McCormick D, Greenspan RH: Pulmonary sarcoidosis. Roentgenologic analysis of 150 patients. Am J Roentgenol Radium Ther Nucl Med 117(4):777–786, 1973. 318. Kline JN, Hirasuna JD: Pulmonary edema after freebase cocaine smoking–Not due to an adulterant. Chest 97(4):1009–1010, 1990. 319. Ko SF, Lee TY, Huang CC, et al: Severe acute respiratory syndrome: Prognostic implications of chest radiographic findings in 52 patients. Radiology 233(1):173–181, 2004. 320. Kobayashi Y, Sakao Y, Deshpande GA, et al: The association between baseline clinical-radiological characteristics and growth of pulmonary nodules with ground-glass opacity. Lung Cancer 83(1):61–66, 2014. 321. Koenig SM: Drug-induced respiratory disease: “The great mimicker”. South Med J 100(9):863–864, 2007. 322. Koldingsnes W, Jacobsen EA, Sildnes T, et al: Pulmonary function and high-resolution CT findings five years after disease onset in patients with Wegener’s granulomatosis. Scand J Rheumatol 34(3):220–228, 2005. 323. Kondoh Y, Taniguchi H, Kawabata Y, et al: Acute exacerbation in idiopathic pulmonary fibrosis. Analysis of clinical and pathologic findings in three cases. Chest 103(6):1808–1812, 1993.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 324. Ksiazek TG, Erdman D, Goldsmith CS, et al: A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 348(20):1953–1966, 2003. 325. Kuhlman JE, Collins J, Brooks GN, et al: Dual-energy subtraction chest radiography: What to look for beyond calcified nodules. Radiographics 26(1):79–92, 2006. 326. Kuhlman JE, Deutsch JH, Fishman EK, et al: CT features of thoracic mycobacterial disease. Radiographics 10(3):413–431, 1990. 327. Kuhlman JE, Fishman EK, Hamper UM, et al: The computed tomographic spectrum of thoracic sarcoidosis. Radiographics 9(3):449– 466, 1989. 328. Kuhlman JE, Fishman EK, Hruban RH, et al: Diseases of the chest in AIDS: CT diagnosis. Radiographics 9(5):827–857, 1989. 329. Kuhlman JE, Fishman EK, Siegelman SS: Invasive pulmonary aspergillosis in acute leukemia: Characteristic findings on CT, the CT halo sign, and the role of CT in early diagnosis. Radiology 157(3):611– 614, 1985. 330. Kuhlman JE, Fishman EK, Teigen C: Pulmonary septic emboli: Diagnosis with CT. Radiology 174(1):211–213, 1990. 331. Kuhlman JE, Hruban RH, Fishman EK: Wegener granulomatosis: CT features of parenchymal lung disease. J Comput Assist Tomogr 15(6):948–952, 1991. 332. Kuhlman JE, Kavuru M, Fishman EK, et al: Pneumocystis carinii pneumonia: spectrum of parenchymal CT findings. Radiology 175(3):711–714, 1990. 333. Kuhlman JE, Ney DR, Fishman EK: Two-dimensional and threedimensional imaging of the in vivo lung: Combining spiral computed tomography with multiplanar and volumetric rendering techniques.   J Digit Imaging 7(1):42–47, 1994. 334. Kuhlman JE, Reyes BL, Hruban RH, et al: Abnormal air-filled spaces in the lung. Radiographics 13(1):47–75, 1993. 335. Kuhlman JE, Scatarige JC, Fishman EK, et al: CT demonstration of high attenuation pleural-parenchymal lesions due to amiodarone therapy.   J Comput Assist Tomogr 11(1):160–162, 1987. 336. Kuhlman JE, Teigen C, Ren H, et al: Amiodarone pulmonary toxicity: CT findings in symptomatic patients. Radiology 177(1):121–125, 1990. 337. Kwon WJ, Jeong YJ, Kim KI, et al: Computed tomographic features of pulmonary septic emboli: Comparison of causative microorganisms.   J Comput Assist Tomogr 31(3):390–394, 2007. 338. Kwong JS, Muller NL, Godwin JD, et al: Thoracic actinomycosis: CT findings in eight patients. Radiology 183(1):189–192, 1992. 339. Larsen BT, Vaszar LT, Colby TV, et al: Lymphoid hyperplasia and eosinophilic pneumonia as histologic manifestations of amiodaroneinduced lung toxicity. Am J Surg Pathol 36(4):509–516, 2012. 340. Lateef O, Shakoor N, Balk RA: Methotrexate pulmonary toxicity. Expert Opin Drug Saf 4(4):723–730, 2005. 341. Lawler LP, Fishman EK: Multi-detector row CT of thoracic disease with emphasis on 3D volume rendering and CT angiography. Radiographics 21(5):1257–1273, 2001. 342. Lawler LP, Fishman EK: Arteriovenous malformations and systemic lung supply: Evaluation by multidetector CT and three-dimensional volume rendering. AJR Am J Roentgenol 178(2):493–495, 2002. 343. Lee CH: The crazy-paving sign. Radiology 243(3):905–906, 2007. 344. Lee PH, Carr DH, Rubens MB, et al: Accuracy of CT in predicting the cause of bronchiectasis. Clin Radiol 50(12):839–841, 1995. 345. Lee ES, Gotway MB, Reddy GP, et al: Early bronchiolitis obliterans following lung transplantation: Accuracy of expiratory thin-section CT for diagnosis. Radiology 216(2):472–477, 2000. 346. Lee KN, Levin DL, Webb WR, et al: Pulmonary alveolar proteinosis: High-resolution CT, chest radiographic, and functional correlations. Chest 111(4):989–995, 1997. 347. Lee KS, Kim TS, Fujimoto K, et al: Thoracic manifestation of Wegener’s granulomatosis: CT findings in 30 patients. Eur Radiol 13(1):43–51, 2003. 348. Lee KS, Primack SL, Staples CA, et al: Chronic infiltrative lung disease: comparison of diagnostic accuracies of radiography and low- and conventional-dose thin-section CT. Radiology 191(3):669–673, 1994.

977

349. Lee W, Ryu DR, Han SS, et al: Very early onset of amiodarone-  induced pulmonary toxicity. Korean Circ J 43(10):699–701, 2013. 350. Lee KS, Song KS, Lim TH, et al: Adult-onset pulmonary tuberculosis: Findings on chest radiographs and CT scans. AJR Am J Roentgenol 160(4):753–758, 1993. 351. Lenoir S, Grenier P, Brauner MW, et al: Pulmonary lymphangiomyomatosis and tuberous sclerosis: Comparison of radiographic and thin-section CT findings. Radiology 175(2):329–334, 1990. 352. Leung AN: Pulmonary tuberculosis: The essentials. Radiology 210(2):307–322, 1999. 353. Leung AN, Fisher K, Valentine V, et al: Bronchiolitis obliterans after lung transplantation: Detection using expiratory HRCT. Chest 113(2):365–370, 1998. 354. Leung AN, Gosselin MV, Napper CH, et al: Pulmonary infections after bone marrow transplantation: Clinical and radiographic findings. Radiology 210(3):699–710, 1999. 355. Leung AN, Miller RR, Muller NL: Parenchymal opacification in chronic infiltrative lung diseases: CT-pathologic correlation. Radiology 188(1):209–214, 1993. 356. Levin DL: Radiology of pulmonary Mycobacterium avium-intracellulare complex. Clin Chest Med 23(3):603–612, 2002. 357. Ley-Zaporozhan J, Ley S, Weinheimer O, et al: Quantitative analysis of emphysema in 3D using MDCT: Influence of different reconstruction algorithms. Eur J Radiol 65(2):228–234, 2008. 358. Li AM, So HK, Chu W, et al: Radiological and pulmonary function outcomes of children with SARS. Pediatr Pulmonol 38(6):427–433, 2004. 359. Lindenfeld JA, Simon SF, Zamora MR, et al: BOOP is common in cardiac transplant recipients switched from a calcineurin inhibitor to sirolimus. Am J Transplant 5(6):1392–1396, 2005. 360. Logan PM, Finnegan MM: Pulmonary complications in AIDS: CT appearances. Clin Radiol 53(8):567–573, 1998. 361. Logan PM, Primack SL, Miller RR, et al: Invasive aspergillosis of the airways: Radiographic, CT, and pathologic findings. Radiology 193(2):383–388, 1994. 362. Lohr RH, Boland BJ, Douglas WW, et al: Organizing pneumonia. Features and prognosis of cryptogenic, secondary, and focal variants. Arch Intern Med 157(12):1323–1329, 1997. 363. Lohrmann C, Uhl M, Kotter E, et al: Pulmonary manifestations of Wegener granulomatosis: CT findings in 57 patients and a review of the literature. Eur J Radiol 53(3):471–477, 2005. 364. Lohrmann C, Uhl M, Schaefer O, et al: Serial high-resolution computed tomography imaging in patients with Wegener granulomatosis: Differentiation between active inflammatory and chronic fibrotic lesions. Acta Radiol 46(5):484–491, 2005. 365. Lucidarme O, Grenier PA, Cadi M, et al: Evaluation of air trapping at CT: Comparison of continuous versus suspended-expiration CT techniques. Radiology 216(3):768–772, 2000. 366. Lucidarme O, Grenier P, Coche E, et al: Bronchiectasis: Comparative assessment with thin-section CT and helical CT. Radiology 200(3):673– 679, 1996. 367. Lynch JP, 3rd: Computed tomographic scanning in sarcoidosis. Semin Respir Crit Care Med 24(4):393–418, 2003. 368. Lynch DA, Gamsu G, Ray CS, et al: Asbestos-related focal lung masses: Manifestations on conventional and high-resolution CT scans. Radiology 169(3):603–607, 1988. 369. Lynch JP, 3rd, Hunninghake GW: Pulmonary complications of collagen vascular disease. Annu Rev Med 43:17–35, 1992. 370. Lynch DA, Newell JD, Logan PM, et al: Can CT distinguish hypersensitivity pneumonitis from idiopathic pulmonary fibrosis?   AJR Am J Roentgenol 165(4):807–811, 1995. 371. Lynch DA, Rose CS, Way D, et al: Hypersensitivity pneumonitis: Sensitivity of high-resolution CT in a population-based study. AJR Am J Roentgenol 159(3):469–472, 1992. 372. Lynch DA, Travis WD, Muller NL, et al: Idiopathic interstitial pneumonias: CT features. Radiology 236(1):10–21, 2005.

http://radiologyebook.com

978

PART II  CT and MR Imaging of the Whole Body

373. Lyon R, Haque AK, Asmuth DM, et al: Changing patterns of infections in patients with AIDS: A study of 279 autopsies of prison inmates and nonincarcerated patients at a university hospital in eastern Texas, 1984-1993. Clin Infect Dis 23(2):241–247, 1996. 374. MacDonald SL, Rubens MB, Hansell DM, et al: Nonspecific interstitial pneumonia and usual interstitial pneumonia: Comparative appearances at and diagnostic accuracy of thin-section CT. Radiology 221(3):600– 605, 2001. 375. MacGregor RR: Tuberculosis: From history to current management. Semin Roentgenol 28(2):101–108, 1993. 376. Macklin CC: Histological indications of the sites of air leakage from the lung alveoli into the vascular sheaths during local overinflation of the living cat’s lung. Can Med Assoc J 38(4):401–402, 1938. 377. Macklin CC: Transport of air along sheats of blood vessels from alveoli to mediastinum: Clinical implications. Arch Intern Med 64:913–926, 1939. 378. MacMahon H, Austin JH, Gamsu G, et al: Guidelines for management of small pulmonary nodules detected on CT scans: A statement from the Fleischner Society. Radiology 237(2):395–400, 2005. 379. Madhi SA, Ludewick H, Abed Y, et al: Human metapneumovirusassociated lower respiratory tract infections among hospitalized human immunodeficiency virus type 1 (HIV-1)-infected and HIV-1-uninfected African infants. Clin Infect Dis 37(12):1705–1710, 2003. 380. Maertens J, Verhaegen J, Lagrou K, et al: Screening for circulating galactomannan as a noninvasive diagnostic tool for invasive aspergillosis in prolonged neutropenic patients and stem cell transplantation recipients: A prospective validation. Blood 97(6):1604– 1610, 2001. 381. Malagari K, Economopoulos N, Stoupis C, et al: High-resolution CT findings in mild pulmonary fat embolism. Chest 123(4):1196–1201, 2003. 382. Malo JL, Hawkins R, Pepys J: Studies in chronic allergic bronchopulmonary aspergillosis. 1. Clinical and physiological findings. Thorax 32(3):254–261, 1977. 383. Malo JL, Longbottom J, Mitchell J, et al: Studies in chronic allergic bronchopulmonary aspergillosis. 3. Immunological findings. Thorax 32(3):269–274, 1977. 384. Malo JL, Pepys J, Simon G: Studies in chronic allergic bronchopulmonary aspergillosis. 2. Radiological findings. Thorax 32(3):262–268, 1977. 385. Marasco WJ, Fishman EK, Kuhlman JE, et al: Acute pulmonary hemorrhage. CT evaluation. Clin Imaging 17(1):77–80, 1993. 386. Marchiori E, Ferreira A, Muller NL: Silicoproteinosis: High-resolution CT and histologic findings. J Thorac Imaging 16(2):127–129, 2001. 387. Marchiori E, Souza AS, Jr, Franquet T, et al: Diffuse high-attenuation pulmonary abnormalities: A pattern-oriented diagnostic approach on high-resolution CT. AJR Am J Roentgenol 184(1):273–282, 2005. 388. Marinelli DL, Albelda SM, Williams TM, et al: Nontuberculous mycobacterial infection in AIDS: Clinical, pathologic, and radiographic features. Radiology 160(1):77–82, 1986. 389. Mark EJ, Ramirez JF: Pulmonary capillaritis and hemorrhage in patients with systemic vasculitis. Arch Pathol Lab Med 109(5):413–418, 1985. 390. Marom EM, McAdams HP, Erasmus JJ, et al: The many faces of pulmonary aspiration. AJR Am J Roentgenol 172(1):121–128, 1999. 391. Maroto JP, Hudes G, Dutcher JP, et al: Drug-related pneumonitis in patients with advanced renal cell carcinoma treated with temsirolimus. J Clin Oncol 29(13):1750–1756, 2011. 392. Marten K, Schnyder P, Schirg E, et al: Pattern-based differential diagnosis in pulmonary vasculitis using volumetric CT. AJR Am J Roentgenol 184(3):720–733, 2005. 393. Marti-Bonmati L, Ruiz Perales F, Catala F, et al: CT findings in Swyer-James syndrome. Radiology 172(2):477–480, 1989. 394. Martin WJ, 2nd, Rosenow EC, 3rd: Amiodarone pulmonary toxicity. Recognition and pathogenesis (Part I). Chest 93(5):1067–1075, 1988. 395. Martin WJ, 2nd, Rosenow EC, 3rd: Amiodarone pulmonary toxicity. Recognition and pathogenesis (Part 2). Chest 93(6):1242–1248, 1988.

396. Martin KW, Sagel SS, Siegel BA: Mosaic oligemia simulating pulmonary infiltrates on CT. AJR Am J Roentgenol 147(4):670–673, 1986. 397. Martinez F, Chung JH, Digumarthy SR, et al: Common and uncommon manifestations of Wegener granulomatosis at chest CT: Radiologicpathologic correlation. Radiographics 32(1):51–69, 2012. 398. Martinez S, Heyneman LE, McAdams HP, et al: Mucoid impactions: Finger-in-glove sign and other CT and radiographic features. Radiographics 28(5):1369–1382, 2008. 399. Matsumoto S, Mori H, Miyake H, et al: MRI signal characteristics of progressive massive fibrosis in silicosis. Clin Radiol 53(7):510–514, 1998. 400. Matsuoka S, Kurihara Y, Yagihashi K, et al: Morphological progression of emphysema on thin-section CT: Analysis of longitudinal change in the number and size of low-attenuation clusters. J Comput Assist Tomogr 30(4):669–674, 2006. 401. Matsuoka S, Kurihara Y, Yagihashi K, et al: Quantitative thin-section CT analysis of the enlargement and coalescence of low-attenuation clusters in patients with emphysema. Respiration 74(2):136–141, 2007. 402. Mattar CS, Keith RL, Byrd RP, Jr, et al: Septic pulmonary emboli due to periodontal disease. Respir Med 100(8):1470–1474, 2006. 403. Maurer JR, Tullis DE, Grossman RF, et al: Infectious complications following isolated lung transplantation. Chest 101(4):1056–1059, 1992. 404. Mayo JR: High resolution computed tomography. Technical aspects. Radiol Clin North Am 29(5):1043–1049, 1991. 405. Mayo JR: The high-resolution computed tomography technique. Semin Roentgenol 26(2):104–109, 1991. 406. Mayo JR, Webb WR, Gould R, et al: High-resolution CT of the lungs: An optimal approach. Radiology 163(2):507–510, 1987. 407. McCarroll KA, Roszler MH: Lung disorders due to drug abuse. J Thorac Imaging 6(1):30–35, 1991. 408. McGarvey LP, Heaney LG, Lawson JT, et al: Evaluation and outcome of patients with chronic non-productive cough using a comprehensive diagnostic protocol. Thorax 53(9):738–743, 1998. 409. McGuinness G: Changing trends in the pulmonary manifestations of AIDS. Radiol Clin North Am 35(5):1029–1082, 1997. 410. McGuinness G, Gruden JF, Bhalla M, et al: AIDS-related airway disease. AJR Am J Roentgenol 168(1):67–77, 1997. 411. McHugh K, Blaquiere RM: CT features of rounded atelectasis. AJR Am J Roentgenol 153(2):257–260, 1989. 412. McLoud TC, Epler GR, Colby TV, et al: Bronchiolitis obliterans. Radiology 159(1):1–8, 1986. 413. McLoud TC, Epler GR, Gaensler EA, et al: A radiographic classification for sarcoidosis: Physiologic correlation. Invest Radiol 17(2):129–138, 1982. 414. Mejia M, Carrillo G, Rojas-Serrano J, et al: Idiopathic pulmonary fibrosis and emphysema: Decreased survival associated with severe pulmonary arterial hypertension. Chest 136(1):10–15, 2009. 415. Menzies R, Fraser R: Round atelectasis. Pathologic and pathogenetic features. Am J Surg Pathol 11(9):674–681, 1987. 416. Meziane MA, Hruban RH, Zerhouni EA, et al: High resolution CT of the lung parenchyma with pathologic correlation. Radiographics 8(1):27–54, 1988. 417. Miller WT, Jr: Spectrum of pulmonary nontuberculous mycobacterial infection. Radiology 191(2):343–350, 1994. 418. Miller WT, Jr: Pulmonary infections in patients who have received solid organ transplants. Semin Roentgenol 35(2):152–170, 2000. 419. Miller WT, Miller WT, Jr: Tuberculosis in the normal host: Radiological findings. Semin Roentgenol 28(2):109–118, 1993. 420. Miller RR, Muller NL: Neuroendocrine cell hyperplasia and obliterative bronchiolitis in patients with peripheral carcinoid tumors. Am J Surg Pathol 19(6):653–658, 1995. 421. Miller BH, Rosado-de-Christenson ML, McAdams HP, et al: Thoracic sarcoidosis: Radiologic-pathologic correlation. Radiographics 15(2):421– 437, 1995. 422. Miller WT, Jr, Sais GJ, Frank I, et al: Pulmonary aspergillosis in patients with AIDS. Clinical and radiographic correlations. Chest 105(1):37–44, 1994.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 423. Miller WT, Jr, Shah RM: Isolated diffuse ground-glass opacity in thoracic CT: causes and clinical presentations. AJR Am J Roentgenol 184(2):613–622, 2005. 424. Minakuchi M, Niimi A, Matsumoto H, et al: Chronic eosinophilic pneumonia: Treatment with inhaled corticosteroids. Respiration 70(4):362–366, 2003. 425. Mintzer RA, Rogers LF, Kruglik GD, et al: The spectrum of radiologic findings in allergic bronchopulmonary aspergillosis. Radiology 127(2):301–307, 1978. 426. Mochimaru H, Kawamoto M, Fukuda Y, et al: Clinicopathological differences between acute and chronic eosinophilic pneumonia. Respirology 10(1):76–85, 2005. 427. Moe AA, Hardy WD: Pneumocystis carinii infection in the HIVseropositive patient. Infect Dis Clin North Am 8(2):331–364, 1994. 428. Mohr LC: Hypersensitivity pneumonitis. Curr Opin Pulm Med 10(5):401–411, 2004. 429. Mok CC, Yuen KY, Lau CS: Nocardiosis in systemic lupus erythematosus. Semin Arthritis Rheum 26(4):675–683, 1997. 430. Moore AD, Godwin JD, Dietrich PA, et al: Swyer-James syndrome:   CT findings in eight patients. AJR Am J Roentgenol 158(6):1211–1215, 1992. 431. Muller NL: High-resolution computed tomography of diffuse lung disease. Curr Opin Radiol 1(1):5–8, 1989. 432. Muller NL: Clinical value of high-resolution CT in chronic diffuse lung disease. AJR Am J Roentgenol 157(6):1163–1170, 1991. 433. Muller NL, Coiby TV: Idiopathic interstitial pneumonias: Highresolution CT and histologic findings. Radiographics 17(4):1016–1022, 1997. 434. Muller NL, Miller RR: Diseases of the bronchioles: CT and histopathologic findings. Radiology 196(1):3–12, 1995. 435. Muller NL, Ooi GC, Khong PL, et al: Severe acute respiratory syndrome: Radiographic and CT findings. AJR Am J Roentgenol 181(1):3–8, 2003. 436. Muller NL, Ooi GC, Khong PL, et al: High-resolution CT findings of severe acute respiratory syndrome at presentation and after admission. AJR Am J Roentgenol 182(1):39–44, 2004. 437. Muller NL, Staples CA, Miller RR: Bronchiolitis obliterans organizing pneumonia: CT features in 14 patients. AJR Am J Roentgenol 154(5):983–987, 1990. 438. Munk PL, Muller NL, Miller RR, et al: Pulmonary lymphangitic carcinomatosis: CT and pathologic findings. Radiology 166(3):705–709, 1988. 439. Murata K, Itoh H, Todo G, et al: Centrilobular lesions of the lung: Demonstration by high-resolution CT and pathologic correlation. Radiology 161(3):641–645, 1986. 440. Murata K, Khan A, Rojas KA, et al: Optimization of computed tomography technique to demonstrate the fine structure of the lung. Invest Radiol 23(3):170–175, 1988. 441. Murayama S, Murakami J, Yabuuchi H, et al: “Crazy paving appearance” on high resolution CT in various diseases. J Comput Assist Tomogr 23(5):749–752, 1999. 442. Murch CR, Carr DH: Computed tomography appearances of pulmonary alveolar proteinosis. Clin Radiol 40(3):240–243, 1989. 443. Murray RJ, Albin RJ, Mergner W, et al: Diffuse alveolar hemorrhage temporally related to cocaine smoking. Chest 93(2):427–429, 1988. 444. Murray K, Gosselin M, Anderson M, et al: Distribution of lung disease. Semin Ultrasound CT MR 23(4):352–371, 2002. 445. Murray JF, Mills J: Pulmonary infectious complications of human immunodeficiency virus infection. Part I. Am Rev Respir Dis 141(5 Pt 1):1356–1372, 1990. 446. Murray JF, Mills J: Pulmonary infectious complications of human immunodeficiency virus infection. Part II. Am Rev Respir Dis 141(6):1582–1598, 1990. 447. Myers R: Asbestos-related pleural disease. Curr Opin Pulm Med 18(4):377–381, 2012. 448. Myers JL, Colby TV: Pathologic manifestations of bronchiolitis, constrictive bronchiolitis, cryptogenic organizing pneumonia, and diffuse panbronchiolitis. Clin Chest Med 14(4):611–622, 1993.

979

449. Myers JL, Limper AH, Swensen SJ: Drug-induced lung disease: A pragmatic classification incorporating HRCT appearances. Semin Respir Crit Care Med 24(4):445–454, 2003. 450. Nachman JB, Baum ES, White H, et al: Bleomycin-induced pulmonary fibrosis mimicking recurrent metastatic disease in a patient with testicular carcinoma: Case report of the CT scan appearance. Cancer 47(2):236–239, 1981. 451. Nagai S, Kitaichi M, Itoh H, et al: Idiopathic nonspecific interstitial pneumonia/fibrosis: Comparison with idiopathic pulmonary fibrosis and BOOP. Eur Respir J 12(5):1010–1019, 1998. 452. Naidich DP, Zerhouni EA, Hutchins GM, et al: Computed tomography of the pulmonary parenchyma. Part 1: Distal air-space disease. J Thorac Imaging 1(1):39–53, 1985. 453. Nakamura Y, Chida K, Suda T, et al: Nonspecific interstitial pneumonia in collagen vascular diseases: Comparison of the clinical characteristics and prognostic significance with usual interstitial pneumonia. Sarcoidosis Vasc Diffuse Lung Dis 20(3):235–241, 2003. 454. Nakata H, Kimoto T, Nakayama T, et al: Diffuse peripheral lung disease: Evaluation by high-resolution computed tomography. Radiology 157(1):181–185, 1985. 455. Napel S, Rubin GD, Jeffrey RB, Jr: STS-MIP: A new reconstruction technique for CT of the chest. J Comput Assist Tomogr 17(5):832–838, 1993. 456. Navarro C, Mejia M, Gaxiola M, et al: Hypersensitivity pneumonitis: A broader perspective. Treat Respir Med 5(3):167–179, 2006. 457. Nchimi A, Ghaye B, Noukoua CT, et al: Incidence and distribution of lower extremity deep venous thrombosis at indirect computed tomography venography in patients suspected of pulmonary embolism. Thromb Haemost 97(4):566–572, 2007. 458. Neeld DA, Goodman LR, Gurney JW, et al: Computerized tomography in the evaluation of allergic bronchopulmonary aspergillosis. Am Rev Respir Dis 142(5):1200–1205, 1990. 459. Nelson SW, Christoforidis A: Reversible bronchiectasis. Radiology 71(3):375–382, 1958. 460. Newman LS, Buschman DL, Newell JD, Jr, et al: Beryllium disease: Assessment with CT. Radiology 190(3):835–840, 1994. 461. Ney DR, Fishman EK, Magid D, et al: Three-dimensional volumetric display of CT data: Effect of scan parameters upon image quality.   J Comput Assist Tomogr 15(5):875–885, 1991. 462. Ney DR, Kuhlman JE, Hruban RH, et al: Three-dimensional CT-volumetric reconstruction and display of the bronchial tree. Invest Radiol 25(6):736–742, 1990. 463. Ngom A, Dumont P, Diot P, et al: Benign mediastinal lymphadenopathy in congestive heart failure. Chest 119(2):653–656, 2001. 464. Nicholson AG, Colby TV, Wells AU: Histopathological approach to patterns of interstitial pneumonia in patient with connective tissue disorders. Sarcoidosis Vasc Diffuse Lung Dis 19(1):10–17, 2002. 465. Nicholson AG, Fulford LG, Colby TV, et al: The relationship between individual histologic features and disease progression in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 166(2):173–177, 2002. 466. Nicholson AA, Hayward C: The value of computed tomography in the diagnosis of amiodarone-induced pulmonary toxicity. Clin Radiol 40(6):564–567, 1989. 467. Niederman MS, McCombs JS, Unger AN, et al: The cost of treating community-acquired pneumonia. Clin Ther 20(4):820–837, 1998. 468. Niimi H, Kang EY, Kwong JS, et al: CT of chronic infiltrative lung disease: Prevalence of mediastinal lymphadenopathy. J Comput Assist Tomogr 20(2):305–308, 1996. 469. Nowers K, Rasband JD, Berges G, et al: Approach to ground-glass opacification of the lung. Semin Ultrasound CT MR 23(4):302–323, 2002. 470. Nunes H, Brillet PY, Valeyre D, et al: Imaging in sarcoidosis. Semin Respir Crit Care Med 28(1):102–120, 2007. 471. Oikonomou A, Hansell DM: Organizing pneumonia: The many morphological faces. Eur Radiol 12(6):1486–1496, 2002. 472. Ooi GC, Daqing M: SARS: Radiological features. Respirology 8(Suppl):S15–S19, 2003.

http://radiologyebook.com

980

PART II  CT and MR Imaging of the Whole Body

473. Ooi GC, Tsang KW, Cheung TF, et al: Silicosis in 76 men: Qualitative and quantitative CT evaluation—Clinical-radiologic correlation study. Radiology 228(3):816–825, 2003. 474. Owens CM, Evans TW, Keogh BF, et al: Computed tomography in established adult respiratory distress syndrome. Correlation with lung injury score. Chest 106(6):1815–1821, 1994. 475. Padley SP, Adler B, Hansell DM, et al: High-resolution computed tomography of drug-induced lung disease. Clin Radiol 46(4):232–236, 1992. 476. Padley SP, Jordan SJ, Goldstraw P, et al: Asymmetric ARDS following pulmonary resection: CT findings initial observations. Radiology 223(2):468–473, 2002. 477. Padley SP, King LJ: Computed tomography of the thorax in HIV disease. Eur Radiol 9(8):1556–1569, 1999. 478. Padley SP, Padhani AR, Nicholson A, et al: Pulmonary sarcoidosis mimicking cryptogenic fibrosing alveolitis on CT. Clin Radiol 51(11):807–810, 1996. 479. Pallisa E, Sanz P, Roman A, et al: Lymphangioleiomyomatosis: Pulmonary and abdominal findings with pathologic correlation. Radiographics 22(Spec No):S185–S198, 2002. 480. Papiris SA, Manoussakis MN, Drosos AA, et al: Imaging of thoracic Wegener’s granulomatosis: The computed tomographic appearance.   Am J Med 93(5):529–536, 1992. 481. Parambil JG, Myers JL, Ryu JH: Histopathologic features and   outcome of patients with acute exacerbation of idiopathic pulmonary fibrosis undergoing surgical lung biopsy. Chest 128(5):3310–3315, 2005. 482. Pardo J, Panizo A, Sola I, et al: Prognostic value of clinical, morphologic, and immunohistochemical factors in patients with bronchiolitis obliterans-organizing pneumonia. Hum Pathol 44(5):718– 724, 2013. 483. Park IN, Kim DS, Shim TS, et al: Acute exacerbation of interstitial pneumonia other than idiopathic pulmonary fibrosis. Chest 2007. 484. Patel RA, Sellami D, Gotway MB, et al: Hypersensitivity pneumonitis: Patterns on high-resolution CT. J Comput Assist Tomogr 24(6):965–970, 2000. 485. Patterson JE, Peters J, Calhoon JH, et al: Investigation and control of aspergillosis and other filamentous fungal infections in solid organ transplant recipients. Transpl Infect Dis 2(1):22–28, 2000. 486. Pena E, Dennie C, Franquet T, et al: Nonthrombotic pulmonary embolism: A radiological perspective. Semin Ultrasound CT MR 33(6):522–534, 2012. 487. Perez T, Remy-Jardin M, Cortet B: Airways involvement in rheumatoid arthritis: Clinical, functional, and HRCT findings. Am J Respir Crit Care Med 157(5 Pt 1):1658–1665, 1998. 488. Perrier A, Roy PM, Sanchez O, et al: Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med 352(17):1760–1768, 2005. 489. Pham PT, Pham PC, Danovitch GM, et al: Sirolimus-associated pulmonary toxicity. Transplantation 77(8):1215–1220, 2004. 490. Pietra GG: Pathologic mechanisms of drug-induced lung disorders.   J Thorac Imaging 6(1):1–7, 1991. 491. Pipavath SJ, Lynch DA, Cool C, et al: Radiologic and pathologic features of bronchiolitis. AJR Am J Roentgenol 185(2):354–363, 2005. 492. Portner MM, Gracie WA, Jr: Rheumatoid lung disease with cavitary nodules, pneumothorax and eosinophilia. N Engl J Med 275(13):697– 700, 1966. 493. Presneill JJ, Nakata K, Inoue Y, et al: Pulmonary alveolar proteinosis. Clin Chest Med 25(3):593–613, viii, 2004. 494. Pretorius ES, Stone JH, Hellman DB, et al: Wegener’s granulomatosis: CT evolution of pulmonary parenchymal findings in treated disease. Crit Rev Comput Tomogr 45(1):67–85, 2004. 495. Primack SL, Hartman TE, Lee KS, et al: Pulmonary nodules and the CT halo sign. Radiology 190(2):513–515, 1994. 496. Primack SL, Miller RR, Muller NL: Diffuse pulmonary hemorrhage: Clinical, pathologic, and imaging features. AJR Am J Roentgenol 164(2):295–300, 1995.

497. Primack SL, Muller NL: High-resolution computed tomography in acute diffuse lung disease in the immunocompromised patient. Radiol Clin North Am 32(4):731–744, 1994. 498. Primack SL, Muller NL: Radiologic manifestations of the systemic autoimmune diseases. Clin Chest Med 19(4):573–586, vii, 1998. 499. Primack SL, Muller NL, Mayo JR, et al: Pulmonary parenchymal abnormalities of vascular origin: High-resolution CT findings. Radiographics 14(4):739–746, 1994. 500. Prologo JD, Dogra V, Farag R: CT diagnosis of fat embolism. Am J Emerg Med 22(7):605–606, 2004. 501. Quagliano PV, Das Narla L: Legionella pneumonia causing multiple cavitating pulmonary nodules in a 7-month-old infant. AJR Am J Roentgenol 161(2):367–368, 1993. 502. Rahman A, Ziment I: Tracheobronchial papillomatosis with malignant transformation. Arch Intern Med 143(3):577–578, 1983. 503. Raymond LW: Flock worker’s lung. Ann Intern Med 130(7):615–616, 1999. 504. Rees JH, Woodhead MA, Sheppard MN, et al: Rheumatoid arthritis   and cryptogenic organising pneumonitis. Respir Med 85(3):243–246, 1991. 505. Reich JM, Johnson RE: Mycobacterium avium complex pulmonary disease presenting as an isolated lingular or middle lobe pattern. The Lady Windermere syndrome. Chest 101(6):1605–1609, 1992. 506. Reid L, Simon G: The peripheral pattern in the normal bronchogram and its relation to peripheral pulmonary anatomy. Thorax 13(2):103– 109, 1958. 507. Reittner P, Muller NL, Heyneman L, et al: Mycoplasma pneumoniae pneumonia: Radiographic and high-resolution CT features in 28 patients. AJR Am J Roentgenol 174(1):37–41, 2000. 508. Remy J, Remy-Jardin M, Giraud F, et al: Angioarchitecture of pulmonary arteriovenous malformations: Clinical utility of threedimensional helical CT. Radiology 191(3):657–664, 1994. 509. Remy J, Remy-Jardin M, Wattinne L, et al: Pulmonary arteriovenous malformations: Evaluation with CT of the chest before and after treatment. Radiology 182(3):809–816, 1992. 510. Remy-Jardin M, Campistron P, Amara A, et al: Usefulness of coronal reformations in the diagnostic evaluation of infiltrative lung disease.   J Comput Assist Tomogr 27(2):266–273, 2003. 511. Remy-Jardin M, Dumont P, Brillet PY, et al: Pulmonary arteriovenous malformations treated with embolotherapy: Helical CT evaluation of long-term effectiveness after 2-21-year follow-up. Radiology 239(2):576–585, 2006. 512. Remy-Jardin M, Remy J, Artaud D, et al: Diffuse infiltrative lung disease: Clinical value of sliding-thin-slab maximum intensity projection CT scans in the detection of mild micronodular patterns. Radiology 200(2):333–339, 1996. 513. Remy-Jardin M, Remy J, Boulenguez C, et al: Morphologic effects of cigarette smoking on airways and pulmonary parenchyma in healthy adult volunteers: CT evaluation and correlation with pulmonary function tests. Radiology 186(1):107–115, 1993. 514. Remy-Jardin M, Remy J, Cortet B, et al: Lung changes in rheumatoid arthritis: CT findings. Radiology 193(2):375–382, 1994. 515. Remy-Jardin M, Remy J, Deffontaines C, et al: Assessment of diffuse infiltrative lung disease: Comparison of conventional CT and high-resolution CT. Radiology 181(1):157–162, 1991. 516. Remy-Jardin M, Remy J, Farre I, et al: Computed tomographic evaluation of silicosis and coal workers’ pneumoconiosis. Radiol Clin North Am 30(6):1155–1176, 1992. 517. Remy-Jardin M, Remy J, Giraud F, et al: Computed tomography assessment of ground-glass opacity: Semiology and significance.   J Thorac Imaging 8(4):249–264, 1993. 518. Remy-Jardin M, Remy J, Gosselin B, et al: Sliding thin slab, minimum intensity projection technique in the diagnosis of emphysema: Histopathologic-CT correlation. Radiology 200(3):665–671, 1996. 519. Remy-Jardin M, Remy J, Wallaert B, et al: Subacute and chronic bird breeder hypersensitivity pneumonitis: Sequential evaluation with CT and correlation with lung function tests and bronchoalveolar lavage. Radiology 189(1):111–118, 1993.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 520. Remy-Jardin M, Remy J, Wattinne L, et al: Central pulmonary thromboembolism: Diagnosis with spiral volumetric CT with the single-breath-hold technique—Comparison with pulmonary angiography. Radiology 185(2):381–387, 1992. 521. Remy-Jardin M, Sobaszek A, Duhamel A, et al: Asbestos-related pleuropulmonary diseases: Evaluation with low-dose four-detector row spiral CT. Radiology 233(1):182–190, 2004. 522. Remy-Jardin M, Tillie-Leblond I, Szapiro D, et al: CT angiography of pulmonary embolism in patients with underlying respiratory disease: Impact of multislice CT on image quality and negative predictive value. Eur Radiol 12(8):1971–1978, 2002. 523. Ren H, Kuhlman JE, Hruban RH, et al: CT-pathology correlation of amiodarone lung. J Comput Assist Tomogr 14(5):760–765, 1990. 524. Restrepo CS, Carrillo JA, Martinez S, et al: Pulmonary complications from cocaine and cocaine-based substances: Imaging manifestations. Radiographics 27(4):941–956, 2007. 525. Revel MP, Triki R, Chatellier G, et al: Is It possible to recognize pulmonary infarction on multisection CT images? Radiology 244(3):875–882, 2007. 526. Riezzo I, Fiore C, De Carlo D, et al: Side effects of cocaine abuse: Multiorgan toxicity and pathological consequences. Curr Med Chem 19(33):5624–5646, 2012. 527. Rimmer MJ, Dixon AK, Flower CD, et al: Bleomycin lung: Computed tomographic observations. Br J Radiol 58(695):1041–1045, 1985. 528. Roberts KE, Hamele-Bena D, Saqi A, et al: Pulmonary tumor embolism: A review of the literature. Am J Med 115(3):228–232, 2003. 529. Roca J, Heras M, Rodriguez-Roisin R, et al: Pulmonary complications after long term amiodarone treatment. Thorax 47(5):372–376, 1992. 530. Rockoff SD, Rohatgi PK: Unusual manifestations of thoracic sarcoidosis. AJR Am J Roentgenol 144(3):513–528, 1985. 531. Rodriguez JL, Barrio JL, Pitchenik AE: Pulmonary nocardiosis in the acquired immunodeficiency syndrome. Diagnosis with bronchoalveolar lavage and treatment with non-sulphur containing drugs. Chest 90(6):912–914, 1986. 532. Rogliani P, Mura M, Mattia P, et al: HRCT and histopathological evaluation of fibrosis and tissue destruction in IPF associated with pulmonary emphysema. Respir Med 102(12):1753–1761, 2008. 533. Rosenow EC, 3rd, Myers JL, Swensen SJ, et al: Drug-induced pulmonary disease. An update. Chest 102(1):239–250, 1992. 534. Rosenow E, Strimlan CV, Muhm JR, et al: Pleuropulmonary manifestations of ankylosing spondylitis. Mayo Clin Proc 52(10):641– 649, 1977. 535. Rossi SE, Erasmus JJ, McAdams HP, et al: Pulmonary drug toxicity: Radiologic and pathologic manifestations. Radiographics 20(5):1245– 1259, 2000. 536. Rossi SE, Erasmus JJ, Volpacchio M, et al: “Crazy-paving” pattern at thin-section CT of the lungs: Radiologic-pathologic overview. Radiographics 23(6):1509–1519, 2003. 537. Rossi SE, Franquet T, Volpacchio M, et al: Tree-in-bud pattern at thin-section CT of the lungs: Radiologic-pathologic overview. Radiographics 25(3):789–801, 2005. 538. Rubini G, Niccoli Asabella A, Stabile Ianora AA, et al: Acute pulmonary embolism: Comparison and integration of perfusion lung scintigraphy with multislice spiral CT. Radiol Med (Torino) 112(2):174–184, 2007. 539. Ryerson CJ, Hartman T, Elicker BM, et al: Clinical features and outcomes in combined pulmonary fibrosis and emphysema in idiopathic pulmonary fibrosis. Chest 144(1):234–240, 2013. 540. Ryu JH: Classification and approach to bronchiolar diseases. Curr Opin Pulm Med 12(2):145–151, 2006. 541. Ryu JH, Colby TV, Hartman TE, et al: Smoking-related interstitial lung diseases: A concise review. Eur Respir J 17(1):122–132, 2001. 542. Ryu JH, Myers JL, Swensen SJ: Bronchiolar disorders. Am J Respir Crit Care Med 168(11):1277–1292, 2003. 543. Sable CA, Donowitz GR: Infections in bone marrow transplant recipients. Clin Infect Dis 18(3):273–281, quiz 282-274, 1994. 544. Sakai F, Gamsu G, Im JG, et al: Pulmonary function abnormalities in patients with CT-determined emphysema. J Comput Assist Tomogr 11(6):963–968, 1987.

981

545. Sarosi GA, Johnson PC: Progressive disseminated histoplasmosis in the acquired immunodeficiency syndrome: A model for disseminated disease. Semin Respir Infect 5(2):146–150, 1990. 546. Sarosi GA, Johnson PC: Disseminated histoplasmosis in patients infected with human immunodeficiency virus. Clin Infect Dis 14(Suppl 1):S60–S67, 1992. 547. Scadding JG, Hinson KF: Diffuse fibrosing alveolitis (diffuse interstitial fibrosis of the lungs). Correlation of histology at biopsy with prognosis. Thorax 22(4):291–304, 1967. 548. Schoepf UJ, Bruening R, Konschitzky H, et al: Pulmonary embolism: Comprehensive diagnosis by using electron-beam CT for detection of emboli and assessment of pulmonary blood flow. Radiology 217(3):693–700, 2000. 549. Schoepf UJ, Costello P: Multidetector-row CT imaging of pulmonary embolism. Semin Roentgenol 38(2):106–114, 2003. 550. Schoepf UJ, Costello P: CT angiography for diagnosis of pulmonary embolism: State of the art. Radiology 230(2):329–337, 2004. 551. Schoepf UJ, Helmberger T, Holzknecht N, et al: Segmental and subsegmental pulmonary arteries: Evaluation with electron-beam versus spiral CT. Radiology 214(2):433–439, 2000. 552. Schriner RW, Ryu JH, Edwards WD: Microscopic pulmonary tumor embolism causing subacute cor pulmonale: A difficult antemortem diagnosis. Mayo Clin Proc 66(2):143–148, 1991. 553. Schurawitzki H, Stiglbauer R, Graninger W, et al: Interstitial lung disease in progressive systemic sclerosis: High-resolution CT versus radiography. Radiology 176(3):755–759, 1990. 554. Schussler JM, Smith ER: Sixty-four-slice computed tomographic coronary angiography: Will the “triple rule out” change chest pain evaluation in the ED? Am J Emerg Med 25(3):367–375, 2007. 555. Schwartz DA, Bryan RT, Hughes JM: Pathology and emerging infections—Quo vadimus? Am J Pathol 147(6):1525–1533, 1995. 556. Seaman DM, Meyer CA, Gilman MD, et al: Diffuse cystic lung disease at high-resolution CT. AJR Am J Roentgenol 196(6):1305–1311, 2011. 557. Seymour JF, Presneill JJ: Pulmonary alveolar proteinosis: Progress in the first 44 years. Am J Respir Crit Care Med 166(2):215–235, 2002. 558. Shaffer K, Pugatch RD: Small pulmonary nodules: Dynamic CT with a single-breath technique. Radiology 173(2):567–568, 1989. 559. Shah AA, Davis SD, Gamsu G, et al: Parenchymal and pleural findings in patients with and patients without acute pulmonary embolism detected at spiral CT. Radiology 211(1):147–153, 1999. 560. Shah RM, Kaji AV, Ostrum BJ, et al: Interpretation of chest radiographs in AIDS patients: Usefulness of CD4 lymphocyte counts. Radiographics 17(1):47–58, discussion 59-61, 1997. 561. Shah RM, Miller W, Jr: Widespread ground-glass opacity of the lung in consecutive patients undergoing CT: Does lobular distribution assist diagnosis? AJR Am J Roentgenol 180(4):965–968, 2003. 562. Shannon TM, Gale ME: Noncardiac manifestations of rheumatoid arthritis in the thorax. J Thorac Imaging 7(2):19–29, 1992. 563. Shepard JA, Moore EH, Templeton PA, et al: Pulmonary intravascular tumor emboli: Dilated and beaded peripheral pulmonary arteries at CT. Radiology 187(3):797–801, 1993. 564. Sherrick AD, Swensen SJ, Hartman TE: Mosaic pattern of lung attenuation on CT scans: Frequency among patients with pulmonary artery hypertension of different causes. AJR Am J Roentgenol 169(1):79– 82, 1997. 565. Shida H, Chiyotani K, Honma K, et al: Radiologic and pathologic characteristics of mixed dust pneumoconiosis. Radiographics 16(3):483– 498, 1996. 566. Shiota Y, Arikita H, Horita N, et al: Septic pulmonary embolism associated with periodontal disease: Reports of two cases and review of the literature. Chest 121(2):652–654, 2002. 567. Sider L, Westcott MA: Pulmonary manifestations of cryptococcosis in patients with AIDS: CT features. J Thorac Imaging 9(2):78–84, 1994. 568. Siegel MD: Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med 353(6):630–631, author reply 630-631, 2005. 569. Siegelman SS, Khouri NF, Leo FP, et al: Solitary pulmonary nodules:   CT assessment. Radiology 160(2):307–312, 1986.

http://radiologyebook.com

982

PART II  CT and MR Imaging of the Whole Body

570. Siegelman SS, Khouri NF, Scott WW, Jr, et al: Pulmonary hamartoma: CT findings. Radiology 160(2):313–317, 1986. 571. Silva CI, Churg A, Muller NL: Hypersensitivity pneumonitis: Spectrum of high-resolution CT and pathologic findings. AJR Am J Roentgenol 188(2):334–344, 2007. 572. Silva CI, Flint JD, Levy RD, et al: Diffuse lung cysts in lymphoid interstitial pneumonia: High-resolution CT and pathologic findings.   J Thorac Imaging 21(3):241–244, 2006. 573. Silva CI, Muller NL: Drug-induced lung diseases: Most common reaction patterns and corresponding high-resolution CT manifestations. Semin Ultrasound CT MR 27(2):111–116, 2006. 574. Silver SF, Muller NL, Miller RR, et al: Hypersensitivity pneumonitis: Evaluation with CT. Radiology 173(2):441–445, 1989. 575. Singh K, Eong OE, Kumarsil B, et al: Severe acute respiratory syndrome without respiratory symptoms or abnormal chest radiograph findings. Clin Infect Dis 38(4):585–586, 2004. 576. Slanetz PJ, Truong M, Shepard JA, et al: Mediastinal lymphadenopathy and hazy mediastinal fat: New CT findings of congestive heart failure. AJR Am J Roentgenol 171(5):1307–1309, 1998. 577. Sostman HD, Matthay RA, Putman CE: Cytotoxic drug-induced lung disease. Am J Med 62(4):608–615, 1977. 578. Sostman HD, Putman CE, Gamsu G: Review: Diagnosis of chemotherapy lung. AJR Am J Roentgenol 136(1):33–40, 1981. 579. Soubani AO, Miller KB, Hassoun PM: Pulmonary complications of bone marrow transplantation. Chest 109(4):1066–1077, 1996. 580. Souza CA, Muller NL, Flint J, et al: Idiopathic pulmonary fibrosis: Spectrum of high-resolution CT findings. AJR Am J Roentgenol 185(6):1531–1539, 2005. 581. Standertskjold-Nordenstam CG, Wandtke JC, Hood WB, Jr, et al: Amiodarone pulmonary toxicity. Chest radiography and CT in asymptomatic patients. Chest 88(1):143–145, 1985. 582. Stansell JD: Fungal disease in HIV-infected persons: Cryptococcosis, histoplasmosis, and coccidioidomycosis. J Thorac Imaging 6(4):28–35, 1991. 583. Stansell JD: Pulmonary fungal infections in HIV-infected persons. Semin Respir Infect 8(2):116–123, 1993. 584. Stansell JD, Osmond DH, Charlebois E, et al: Predictors of Pneumocystis carinii pneumonia in HIV-infected persons. Pulmonary Complications of HIV Infection Study Group. Am J Respir Crit Care Med 155(1):60–66, 1997. 585. Stark P: Round atelectasis: Another pulmonary pseudotumor. Am Rev Respir Dis 125(2):248–250, 1982. 586. Stark P, Jacobson F, Shaffer K: Standard imaging in silicosis and coal worker’s pneumoconiosis. Radiol Clin North Am 30(6):1147–1154, 1992. 587. Stein MG, Mayo J, Muller N, et al: Pulmonary lymphangitic spread of carcinoma: Appearance on CT scans. Radiology 162(2):371–375, 1987. 588. Stern EJ, Frank MS: Small-airway diseases of the lungs: Findings at expiratory CT. AJR Am J Roentgenol 163(1):37–41, 1994. 589. Stern EJ, Muller NL, Swensen SJ, et al: CT mosaic pattern of lung attenuation: Etiologies and terminology. J Thorac Imaging 10(4):294– 297, 1995. 590. Stern EJ, Swensen SJ, Hartman TE, et al: CT mosaic pattern of lung attenuation: Distinguishing different causes. AJR Am J Roentgenol 165(4):813–816, 1995. 591. Stern EJ, Webb WR: Dynamic imaging of lung morphology with ultrafast high-resolution computed tomography. J Thorac Imaging 8(4):273–282, 1993. 592. Stillman AE, Oudkerk M, Ackerman M, et al: Use of multidetector computed tomography for the assessment of acute chest pain: A consensus statement of the North American Society of Cardiac Imaging and the European Society of Cardiac Radiology. Int J Cardiovasc Imaging 2007. 593. Storto ML, Kee ST, Golden JA, et al: Hydrostatic pulmonary edema: High-resolution CT findings. AJR Am J Roentgenol 165(4):817–820, 1995. 594. Subhash HS, Christopher DJ, Roy A, et al: Pulmonary nocardiosis in human immunodeficiency virus infection: A tuberculosis mimic.   J Postgrad Med 47(1):30–32, 2001.

595. Sumikawa H, Johkoh T, Koyama M, et al: Image quality of highresolution CT with 16-channel multidetector-row CT: Comparison between helical scan and conventional step-shoot scan. Radiat Med 23(8):539–544, 2005. 596. Sumikawa H, Johkoh T, Yamamoto S, et al: Quantitative analysis for computed tomography findings of various diffuse lung diseases using volume histogram analysis. J Comput Assist Tomogr 30(2):244–249, 2006. 597. Sverzellati N, Poletti V, Chilosi M, et al: The crazy-paving pattern in granulomatous mycosis fungoides: High-resolution computed tomography-pathological correlation. J Comput Assist Tomogr 30(5):843–845, 2006. 598. Swensen SJ, Harms GF, Morin RL, et al: CT evaluation of solitary pulmonary nodules: Value of 185-H reference phantom. AJR Am J Roentgenol 156(5):925–929, 1991. 599. Swensen SJ, Morin RL, Schueler BA, et al: Solitary pulmonary nodule: CT evaluation of enhancement with iodinated contrast material—A preliminary report. Radiology 182(2):343–347, 1992. 600. Szucs-Farkas Z, Schick A, Cullmann JL, et al: Comparison of dualenergy subtraction and electronic bone suppression combined with computer-aided detection on chest radiographs: Effect on human observers’ performance in nodule detection. AJR Am J Roentgenol 200(5):1006–1013, 2013. 601. Tamm M, Traenkle P, Grilli B, et al: Pulmonary cytomegalovirus infection in immunocompromised patients. Chest 119(3):838–843, 2001. 602. Tan RT, Kuzo RS: High-resolution CT findings of mucinous bronchioloalveolar carcinoma: A case of pseudopulmonary alveolar proteinosis. AJR Am J Roentgenol 168(1):99–100, 1997. 603. Tanaka N, Newell JD, Brown KK, et al: Collagen vascular disease-related lung disease: High-resolution computed tomography findings based on the pathologic classification. J Comput Assist Tomogr 28(3):351–360, 2004. 604. Tanawuttiwat T, Harindhanavudhi T, Hanif S, et al: Amiodaroneinduced alveolar haemorrhage: A rare complication of a common medication. Heart Lung Circ 19(7):435–437, 2010. 605. Tanoue LT: Pulmonary manifestations of rheumatoid arthritis. Clin Chest Med 19(4):667–685, viii, 1998. 606. Tarver RD, Teague SD, Heitkamp DE, et al: Radiology of communityacquired pneumonia. Radiol Clin North Am 43(3):497–512, viii, 2005. 607. Tazelaar HD, Baird AM, Mill M, et al: Bronchocentric mycosis occurring in transplant recipients. Chest 96(1):92–95, 1989. 608. Tazelaar HD, Linz LJ, Colby TV, et al: Acute eosinophilic pneumonia: Histopathologic findings in nine patients. Am J Respir Crit Care Med 155(1):296–302, 1997. 609. Thieme SF, Johnson TR, Reiser MF, et al: Dual-energy lung perfusion computed tomography: A novel pulmonary functional imaging method. Semin Ultrasound CT MR 31(4):301–308, 2010. 610. Thomas PD, Hunninghake GW: Current concepts of the pathogenesis of sarcoidosis. Am Rev Respir Dis 135(3):747–760, 1987. 611. Tobino K, Gunji Y, Kurihara M, et al: Characteristics of pulmonary cysts in Birt-Hogg-Dube syndrome: Thin-section CT findings of the chest in 12 patients. Eur J Radiol 77(3):403–409, 2011. 612. Tomiyama N, Muller NL, Johkoh T, et al: Acute parenchymal lung disease in immunocompetent patients: Diagnostic accuracy of high-resolution CT. AJR Am J Roentgenol 174(6):1745–1750, 2000. 613. Travis WD, Carpenter HA, Lie JT: Diffuse pulmonary hemorrhage. An uncommon manifestation of Wegener’s granulomatosis. Am J Surg Pathol 11(9):702–708, 1987. 614. Travis WD, Costabel U, Hansell DM, et al: An official American Thoracic Society/European Respiratory Society statement: Update of the international multidisciplinary classification of the idiopathic interstitial pneumonias. Am J Respir Crit Care Med 188(6):733–748, 2013. 615. Ujita M, Renzoni EA, Veeraraghavan S, et al: Organizing pneumonia: Perilobular pattern at thin-section CT. Radiology 232(3):757–761, 2004. 616. Urban BA, Fishman EK, Goldman SM, et al: CT evaluation of amyloidosis: Spectrum of disease. Radiographics 13(6):1295–1308, 1993.

http://radiologyebook.com

CHAPTER 36  Nonneoplastic Parenchymal Lung Disease 617. Vallyathan V, Landsittel DP, Petsonk EL, et al: The influence of   dust standards on the prevalence and severity of coal worker’s pneumoconiosis at autopsy in the United States of America. Arch Pathol Lab Med 135(12):1550–1556, 2011. 618. Van den Brande FG, Hellemans S, De Schepper A, et al: Post-traumatic severe fat embolism syndrome with uncommon CT findings. Anaesth Intensive Care 34(1):102–106, 2006. 619. Vassallo R, Jensen EA, Colby TV, et al: The overlap between respiratory bronchiolitis and desquamative interstitial pneumonia in pulmonary Langerhans cell histiocytosis: High-resolution CT, histologic, and functional correlations. Chest 124(4):1199–1205, 2003. 620. Vasu TS, Cavallazzi R, Hirani A, et al: Clinical and radiologic distinctions between secondary bronchiolitis obliterans organizing pneumonia and cryptogenic organizing pneumonia. Respir Care 54(8):1028–1032, 2009. 621. Vehmas T, Kivisaari L, Huuskonen MS, et al: Scoring CT/HRCT findings among asbestos-exposed workers: Effects of patient’s age, body mass index and common laboratory test results. Eur Radiol 15(2):213– 219, 2005. 622. Vigo M, Pesavento R, Bova C, et al: The value of four-detector row spiral computed tomography for the diagnosis of pulmonary embolism. Semin Thromb Hemost 32(8):831–837, 2006. 623. Vilar J, Domingo ML, Soto C, et al: Radiology of bacterial pneumonia. Eur J Radiol 51(2):102–113, 2004. 624. Visscher DW, Myers JL: Bronchiolitis: The pathologist’s perspective. Proc Am Thorac Soc 3(1):41–47, 2006. 625. Vlahakis NE, Rickman OB, Morgenthaler T: Sirolimus-associated diffuse alveolar hemorrhage. Mayo Clin Proc 79(4):541–545, 2004. 626. Vock P, Soucek M, Daepp M, et al: Lung: spiral volumetric CT with single-breath-hold technique. Radiology 176(3):864–867, 1990. 627. Wagenvoort CA: Pathology of pulmonary thromboembolism. Chest 107(1 Suppl):10S–17S, 1995. 628. Ward S, Heyneman LE, Flint JD, et al: Bronchocentric granulomatosis: Computed tomographic findings in five patients. Clin Radiol 55(4):296–300, 2000. 629. Warrington KJ, Moder KG, Brutinel WM: The shrinking lungs syndrome in systemic lupus erythematosus. Mayo Clin Proc 75(5):467– 472, 2000. 630. Webb WR: High resolution lung computed tomography. Normal anatomic and pathologic findings. Radiol Clin North Am 29(5):1051– 1063, 1991. 631. Webb WR: Thin-section CT of the secondary pulmonary lobule: Anatomy and the image—The 2004 Fleischner lecture. Radiology 239(2):322–338, 2006. 632. Webb WR, Stein MG, Finkbeiner WE, et al: Normal and diseased isolated lungs: High-resolution CT. Radiology 166(1 Pt 1):81–87, 1988. 633. Weibel ER: Fleischner Lecture. Looking into the lung: What can it tell us? AJR Am J Roentgenol 133(6):1021–1031, 1979. 634. Wells AU, du Bois RM: Bronchiolitis in association with connective tissue disorders. Clin Chest Med 14(4):655–666, 1993. 635. Westcott JL, Cole SR: Traction bronchiectasis in end-stage pulmonary fibrosis. Radiology 161(3):665–669, 1986. 636. White CS: Chest pain in the emergency department: Potential role of multidetector CT. J Thorac Imaging 22(1):49–55, 2007. 637. White CS, Kuo D, Kelemen M, et al: Chest pain evaluation in the emergency department: Can MDCT provide a comprehensive evaluation? AJR Am J Roentgenol 185(2):533–540, 2005. 638. White RI, Jr, Pollak JS: Pulmonary arteriovenous malformations: Diagnosis with three-dimensional helical CT—A breakthrough without contrast media. Radiology 191(3):613–614, 1994.

983

639. White C, Read K, Kuo D: Assessment of chest pain in the emergency room: What is the role of multidetector CT? Eur J Radiol 57(3):368– 372, 2006. 640. White ES, Tazelaar HD, Lynch JP, 3rd: Bronchiolar complications of connective tissue diseases. Semin Respir Crit Care Med 24(5):543–566, 2003. 641. Wiggins J, Strickland B, Turner-Warwick M: Combined cryptogenic fibrosing alveolitis and emphysema: The value of high resolution computed tomography in assessment. Respir Med 84(5):365–369, 1990. 642. Wintermark M, Wicky S, Schnyder P, et al: Blunt traumatic pneumomediastinum: Using CT to reveal the Macklin effect. AJR Am J Roentgenol 172(1):129–130, 1999. 643. Won HJ, Lee KS, Cheon JE, et al: Invasive pulmonary aspergillosis: Prediction at thin-section CT in patients with neutropenia—A prospective study. Radiology 208(3):777–782, 1998. 644. Wood DL, Osborn MJ, Rooke J, et al: Amiodarone pulmonary toxicity: Report of two cases associated with rapidly progressive fatal adult respiratory distress syndrome after pulmonary angiography. Mayo Clin Proc 60(9):601–603, 1985. 645. Wormanns D, Diederich S: Characterization of small pulmonary nodules by CT. Eur Radiol 14(8):1380–1391, 2004. 646. Worthy SA, Flint JD, Muller NL: Pulmonary complications after bone marrow transplantation: High-resolution CT and pathologic findings. Radiographics 17(6):1359–1371, 1997. 647. Worthy SA, Muller NL, Hartman TE, et al: Mosaic attenuation pattern on thin-section CT scans of the lung: Differentiation among infiltrative lung, airway, and vascular diseases as a cause. Radiology 205(2):465– 470, 1997. 648. Yano S, Usui N, Asai O, et al: Septic intramuscular embolism in a neutropenic patient with myelodysplastic syndrome accompanied by asymptomatic septic pulmonary emboli. Intern Med 44(10):1100–1102, 2005. 649. Yousem SA: Respiratory bronchiolitis-associated interstitial lung disease with fibrosis is a lesion distinct from fibrotic nonspecific interstitial pneumonia: A proposal. Mod Pathol 19(11):1474–1479, 2006. 650. Yousem SA, Colby TV, Carrington CB: Lung biopsy in rheumatoid arthritis. Am Rev Respir Dis 131(5):770–777, 1985. 651. Yousem SA, Colby TV, Gaensler EA: Respiratory bronchiolitisassociated interstitial lung disease and its relationship to desquamative interstitial pneumonia. Mayo Clin Proc 64(11):1373–1380, 1989. 652. Zamora MR, Warner ML, Tuder R, et al: Diffuse alveolar hemorrhage and systemic lupus erythematosus. Clinical presentation, histology, survival, and outcome. Medicine (Baltimore) 76(3):192–202, 1997. 653. Zeman RK, Silverman PM, Vieco PT, et al: CT angiography. AJR Am J Roentgenol 165(5):1079–1088, 1995. 654. Zerhouni E: Computed tomography of the pulmonary parenchyma. An overview. Chest 95(4):901–907, 1989. 655. Zerhouni EA, Naidich DP, Stitik FP, et al: Computed tomography of the pulmonary parenchyma. Part 2: Interstitial disease. J Thorac Imaging 1(1):54–64, 1985. 656. Zerhouni EA, Stitik FP, Siegelman SS, et al: CT of the pulmonary nodule: A cooperative study. Radiology 160(2):319–327, 1986. 657. Zhang M, Kono M: Solitary pulmonary nodules: Evaluation of blood flow patterns with dynamic CT. Radiology 205(2):471–478, 1997. 658. Zwirewich CV, Mayo JR, Muller NL: Low-dose high-resolution CT of lung parenchyma. Radiology 180(2):413–417, 1991. 659. Zwirewich CV, Terriff B, Muller NL: High-spatial-frequency (bone) algorithm improves quality of standard CT of the thorax. AJR Am J Roentgenol 153(6):1169–1173, 1989.

http://radiologyebook.com

37  Neoplastic Disease of the Lung Jeremy J. Erasmus and Santiago E. Rossi

Although most primary pulmonary tumors are carcinomas, a large histologic spectrum of benign and malignant tumors of the lung exist. This chapter reviews the more common neoplasms according to the classification proposed by the World Health Organization (WHO)22,301 (Box 37-1). Additionally, to address advances in oncology, molecular biology, pathology, radiology, and surgery this chapter includes the modifications in the histologic classification of lung cancer as proposed by Travis et al.296,297 This chapter also emphasizes the radiologic manifestations of thoracic neoplasms and the use of imaging in diagnosis and management.

MALIGNANT NEOPLASMS OF THE LUNG PREINVASIVE LESIONS Atypical Adenomatous Hyperplasia Atypical adenomatous hyperplasia (AAH), a localized proliferation of atypical type II pneumocytes and/or Clara cells lining alveoli and respiratory bronchioles, is typically less than 5 mm in diameter. AAH is considered a precursor of adenocarcinoma, based on a variety of molecular findings that demonstrate a relationship to lung adenocarcinoma, including clonality, KRAS mutation, EGFR mutation, p53 expression, loss of heterozygosity, methylation, and epigenetic alterations in the Wnt pathway.149,199,201,296 The frequency of AAH is unknown but has been reported as an incidental finding in 2% to 3% of patients without a primary lung cancer at postmortem examination.282,334 However, AAH is most frequently diagnosed in patients with primary non–small cell carcinoma, especially adenocarcinoma. In this regard, AAH occurs in the adjacent lung parenchyma in 5% to 23% of resected lung adenocarcinomas.41,191,205,206,320 It has also been reported that AAH lesions are not significantly correlated with gender, age, smoking status, familial history of malignancy, or preceding malignancy.205 Radiologically, AAH manifests as a small well-circumscribed solitary or multifocal ground-glass nodular opacity—usually less than 1 cm but ranging in size from a few millimeters to approximately 2 cm—that typically remains stable for several months to years.148,170,225 Because AAH and adenocarcinoma can have similar radiologic manifestations, computed tomography (CT) differentiation is often difficult, and resection is usually required for a definitive diagnosis.

Adenocarcinoma In Situ A major change in the classification proposal for lung adenocarcinoma by Travis et al. is the recognition of adenoma in situ (AIS) as a preinvasive lesion.296 AIS was formerly classified as a bronchioloalveolar cell carcinoma (BAC) according to the 2004 WHO classification. AIS is a small (≤3 cm) localized adenocarcinoma that has no stromal, vascular, or pleural invasion and demonstrates lepidic growth. Most AISs are

984

nonmucinous and consist of type II pneumocytes and/or Clara cells with no nuclear atypia. The rare cases of mucinous AIS consist of tall columnar cells with basal nuclei and abundant cytoplasmic mucin.296 Radiologically, nonmucinous AIS typically manifests on CT as a pure ground-glass nodule (GGN) but can occasionally manifest as a part solid (due to focal collapsed alveoli or focal thickened alveolar septa) or solid nodule. Mucinous AIS can manifest as a solid nodule or consolidation.

Diffuse Idiopathic Pulmonary Neuroendocrine Cell Hyperplasia Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia (DIPNECH) is a widespread proliferation of pulmonary neuroendocrine cells or neuroendocrine bodies within the epithelium of the distal bronchi or terminal bronchioles.62,112 DIPNECH is regarded as a precursor lesion of neuroendocrine lung tumors, specifically carcinoid tumors.62,112 This proliferation is uncommon and can be confined to the bronchial and bronchiolar epithelium, or there can be extension beyond the basement membrane, forming small localized or diffuse aggregates called tumorlets (<0.5 cm in diameter) or carcinoids (>0.5 cm).62,96,112 An important histologic feature occurring in a third of patients is fibrous obliterative bronchiolitis and peribronchiolar fibrosis of the involved airways.5,192 DIPNECH occurs more frequently in women, and most patients present in their fifth or sixth decades (mean age, 58 years), although the age of distribution is wide (22-79 years).25,57,80,258 Patients with DIPNECH typically have a chronic history of an insidious onset of progressive dyspnea and nonproductive cough. Radiologically, DIPNECH manifests as bronchial wall thickening and a mosaic attenuation pattern (ground-glass opacities and hyperlucent areas) due to air trapping.25,96 Solitary or multiple nodules ranging in size from 2 to 20 mm in diameter are an associated feature25,57,169 (Fig. 37-1).

LUNG CANCER Epidemiology Lung cancer is a common malignancy. The American Cancer Society (ACS) estimates that 224,210 new cases will be diagnosed in the United States in 2014.272 Lung cancer incidence rates have declined since the mid-1980s in men and late 1990s in women as a result of reduced smoking prevalence.272 The number of new cases in men has been decreasing at 1.9% per year (1992-2010) and at 1.2% (2005-2010) in woman. The annual mortality rate in men has decreased since the 1990s, and in 2005-2010 this decrease was 2.9% per year, whereas the rate in women, after continuously increasing for several decades, plateaued in 2004 and is now decreasing by 1.4% per year (2004-2010).248,272 However, lung cancer remains the leading cause of cancer-related

http://radiologyebook.com

CHAPTER 37  Neoplastic Disease of the Lung

985

BOX 37-1  Histologic Classification of Lung Tumors I. Epithelial tumors A. Benign Papillomas Adenomas B. Malignant Preinvasive lesions Squamous carcinoma in situ Atypical adenomatous hyperplasia Adenocarcinoma in situ (formerly solitary bronchioloalveolar cell carcinoma [BAC]) Nonmucinous Mucinous Mixed mucinous/nonmucinous Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia Squamous cell carcinoma Variants: papillary, clear cell, small cell, basaloid Adenocarcinoma Minimally invasive adenocarcinoma (≤3 cm lepidic-predominant tumor with ≤ 5 mm invasion) Nonmucinous Mucinous Mixed mucinous/nonmucinous mixed subtype Invasive adenocarcinoma Lepidic predominant (formerly nonmucinous BAC pattern with > 5-mm invasion) Acinar predominant Papillary predominant Micropapillary predominant Solid predominant with mucin production Variants of invasive adenocarcinoma Invasive mucinous adenocarcinoma (including formerly mucinous BAC) Variants: fetal, colloid, enteric Large cell carcinoma Variants: large cell neuroendocrine, basaloid, lymphoepitheliomalike, clear cell, large cell with rhabdoid phenotype

Adenosquamous carcinoma Small cell carcinoma Variant: combined small cell carcinoma Sarcomatoid carcinoma Variants: pleomorphic, spindle cell, giant cell, carcinosarcoma, pulmonary blastoma (pleomorphic carcinoma, spindle cell carcinoma, giant cell carcinoma) Carcinoid tumor Typical carcinoid Atypical carcinoid Salivary gland tumors Mucoepidermoid carcinoma Adenoid cystic carcinoma Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia Mesenchymal tumors Epithelioid hemangioendothelioma Angiosarcoma Pleuropulmonary blastoma Chondroma Congenital peribronchial myofibroblastic tumor Diffuse pulmonary lymphangiomatosis Inflammatory myofibroblastic tumor Lymphangioleiomyomatosis Synovial sarcoma Pulmonary artery/vein sarcoma II. Lymphoproliferative tumors Marginal zone B-cell lymphoma of the mucosa-associated lymphoid tissue (MALT) Diffuse large B-cell lymphoma Lymphomatoid granulomatosis Langerhans cell histiocytosis III. Miscellaneous tumors Hamartoma Sclerosing hemangioma Clear cell tumor Germ cell tumor

Modified from Travis WD, et al: Pathology and genetics: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press.301

deaths in both men and women in the United States, and the ACS estimates that in 2014 it will account for 28% (86,930) of all cancer deaths in men and 26% (72,330) in women.272

Etiology The carcinogenic effect of tobacco smoke was established in 1950 when Doll and Hill conducted the first large epidemiologic study.69 Cigarette smoking is now recognized as the strongest risk factor for the development of lung cancer; an estimated 85% to 90% of lung cancers in men and 80% in women are attributable to smoking.21,82,243,305 The risk increases with early age at initiation and the length of time and number and type of cigarettes smoked.86,180,266,290 Nearly all cases of squamous cell and small cell carcinoma are related to cigarette smoking.329 Although squamous cell and small cell lung carcinomas have been considered to have the highest association with smoking, it has recently been reported that adenocarcinomas are more strongly associated with tobacco smoke exposure than previously recognized.86,150,329 The change in smoking habits (use of filter tips, decrease in tar yield) has been postulated to account for the increase in incidence of adenocarcinomas in cigarette smokers.66,86,300,325 Additionally, because adenocarcinomas

are more likely to be caused by other factors besides smoking (e.g., passive smoke inhalation), adenocarcinomas are the predominant cell type in nonsmokers.86,329 Involuntary smoke exposure (passive smoking) has been reported to increase cotinine, the main metabolite of nicotine, and is generally considered to be associated with an increased risk of lung cancer.20 According to the International Agency for Research on Cancer, involuntary smoking causes lung cancer in never-smokers, with an excess risk of 20% for women and 30% for men.137 In the workplace, involuntary smoke exposure is reported to be associated with an increase of 16% to 19% in the risk of lung cancer among never-smokers.137 This conclusion is supported by a meta-analysis comprising 22 studies that showed a 24% increase in lung cancer risk among workers exposed to environmental tobacco smoke.280 Environmental and occupational exposures to particulate and chemical substances are additional risk factors.281 Exposure to the naturally occurring radioactive gas radon, both in homes and in mines, is an important risk factor for lung cancer.44,315 In fact radon is the second leading cause of lung cancer death among smokers and nonsmokers.54 It is estimated that 7000 to 30,000 lung cancer deaths

http://radiologyebook.com

986

PART II  CT and MR Imaging of the Whole Body Histologic Classification Lung cancer is divided by the 2004 WHO Classification into two major histologic categories: non–small cell lung carcinoma (NSCLC) and small cell lung carcinoma (SCLC)298,299,302 (see Box 37-1). NSCLC is further subdivided into histologic types such as squamous cell carcinoma, adenocarcinoma, and large cell carcinoma according to the most differentiated portion of the tumor. However, many tumors are composed of more than one histologic type; a definitive diagnosis may require a resected specimen. NSCLCs are also graded as well differentiated, moderately differentiated, or poorly differentiated according to the least differentiated feature.298 Additionally, some NSCLCs have immunohistochemical and/or ultrastructural features of neuroendocrine differentiation and are collectively referred to NSCLCs with neuroen­ docrine differentiation.302

FIG 37-1  Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia (DIPNECH) in a 63-year-old woman with a chronic history of progressive cough and dyspnea on minimal exertion, including laughing. CT scan shows multiple small bilateral pulmonary nodules (arrows). Wedge resection biopsy of the middle lobe revealed carcinoid tumor (5 mm) and multiple carcinoid tumorlets (<5 mm).

occur annually in the United States as a result of residential exposure to radon.54,315 Additional risk factors for development of lung cancer include exposure to arsenic, chloromethyl ethers, chromium, isopropyl oil, mustard gas, nickel, chloroprene, vinyl chloride, and various smelting byproducts such as lead and copper.281 Although exposure to beryllium has been reported to cause an increased risk for lung cancer on the order of 20% or higher, a recent study indicates a lack of increased lung cancer risk among workers employed in the modern beryllium industry with exposure restricted to insoluble beryllium.31 Asbestos is also a carcinogen, and although there is an unequivocal association between lung cancer and asbestos exposure, the magnitude of the risk is frequently overestimated (most cohort studies show less than a twofold increase in risk).104,131 Although the risk of lung cancer in workers exposed to asbestos may depend on individual occupational exposure characteristics (e.g., duration, concentration, fiber type), the increased risk may be largely limited to those with radiologic evidence of asbestosis or to cigarette smokers.26,104,253,304,318 It has been estimated that as many as 33% of lung cancers that occur in smokers exposed to asbestos are the result of the synergistic effect of the two carcinogens.75 Other factors such as focal or diffuse pulmonary fibrosis and ionizing radiation have been reported to increase the risk of lung cancer. However, epidemiologic data do not clearly establish a cause-andeffect relationship except in some patients with Hodgkin’s disease treated with radiation.178 Finally, although not clearly understood, it is likely that genetic susceptibility to lung cancer may be important independent of smoking history or exposure to environmental cigarette smoke.19,163,267,322 A number of different gene mutations (e.g., p53, KRAS) are common in lung cancer, and there is a strong association between KRAS mutations in adenocarcinoma and smoking.164,167 In addition, mutations of the epidermal growth factor receptor gene (EGFR) appears to have a strong association with adenocarcinoma and effectiveness of therapy.50,164,202,263,324,332 Additionally, although multigenic factors that influence carcinogen metabolism seem likely to play a role, it appears that a region on chromosome 6q may be important in familial lung cancer, and there is compelling evidence that a locus at 15q25 predisposes to lung cancer.12,19,134

Non–Small Cell Lung Carcinoma Squamous cell carcinoma.  Squamous cell carcinomas have been decreasing in relative incidence and now constitute approximately 30% of all lung cancers8,314 Histologically, squamous cell carcinoma grows in nests within a desmoplastic stroma, and the tumor cells demonstrate keratinization, intercellular bridge formation, and squamous pearls. They typically occur in central bronchi and frequently manifest as postobstructive pneumonia or atelectasis254,271 (Fig. 37-2). Mucoid impaction, bronchiectasis, and hyperinflation are uncommon radiologic manifestations.271 Approximately one third of squamous cell carcinomas occur beyond the segmental bronchi and usually range in size from 1 to 10 cm271 (Fig. 37-3). Squamous cell carcinomas are more likely to cavitate than the other histologic cell types of lung cancer.271 Cavitation occurs in 10% to 30% and is more common in large peripheral masses and poorly differentiated tumors. Cavitation is typically eccentric with thick irregular walls, although thin walls may occur in rare circumstances.271 Most squamous cell carcinomas grow slowly, and extrathoracic metastases tend to occur late.271 Adenocarcinoma.  The increasing incidence of adenocarcinoma since the 1970s has resulted in a notable shift in the incidence rates of lung cancer according to histologic type. Although the rate of increase in the incidence of adenocarcinomas has slowed in the United States since 1999, adenocarcinoma has replaced squamous cell carcinoma as the most frequent histologic form of lung cancer and is now the most common cell type (≈50% of all lung cancers).45 Significant changes in pathologic classification of lung adenocarcinoma were proposed in 2011 as a result of the recommendations of an international multidisciplinary core panel of experts sponsored by the International Association for the Study of Lung Cancer (IASLC), American Thoracic Society, and European Respiratory Society.296 The proposed changes in classification address advances in oncology, molecular biology, pathology, radiology, and surgery and, by improving the diagnosis and classification of lung adenocarcinoma, should result in improved therapeutic outcomes. This new adenocarcinoma classification is primarily based on histology and provides uniform terminology and criteria that define pathologic entities, as well as information relevant to patient management. Unlike previous WHO classifications where the primary diagnostic criteria were based on hematoxylin and eosin examination, this classification emphasizes the use and integration of immunohistochemical (i.e., thyroid transcription factor, p63 staining), histochemical (i.e., mucin staining), and molecular studies.296 The new adenocarcinoma classification taxonomy now includes the term minimally invasive adenocarcinoma. Minimally invasive adenocarcinoma (MIA) is a solitary adenocarcinoma (≤3 cm) with 5-mm or less invasion in any one focus and a predominantly lepidic pattern of

http://radiologyebook.com

CHAPTER 37  Neoplastic Disease of the Lung

*

987

M

B

A

FIG 37-2  Squamous cell lung cancer manifesting as a central endobronchial mass. A, Posteroanterior chest radiograph shows complete atelectasis of the left upper lobe. Convexity in the lower portion of the atelectatic lung (arrow) is the result of a central mass. B, CT confirms an endobronchial mass (M) that occludes the left upper lobe bronchus and causes complete atelectasis of the left upper lobe. Note the enlarged subcarinal node (*) due to metastasis.

A

B FIG 37-3  Squamous cell lung cancer manifesting as a peripheral mass. A, Posteroanterior chest radiograph shows a mass in the right lower lobe and infrahilar and paratracheal adenopathy (arrows). B, CT confirms the mass in the right lower lobe and infrahilar adenopathy (arrow) and reveals variable attenuation of the mass consistent with necrosis.

growth.296 MIA is usually nonmucinous but rarely can be mucinous. The invasive component can be acinar, papillary, micropapillary, and/ or solid. MIA is excluded if the tumor invades lymphatics, blood vessels, or pleura or contains necrosis. Accordingly the diagnosis of MIA requires entire histologic sampling of the tumor. Invasive adenocarcinomas are composed of a complex heterogeneous mixture of histologic subtypes, and the proposed classification by Travis et al. classifies lung adenocarcinomas according to the most predominant subtype.296 The most predominant histologic subtype (e.g., lepidic,

acinar, papillary, micropapillary, or solid) is now added as a descriptor to invasive adenocarcinoma; this replaces use of the term adeno­ carcinoma, mixed subtype. For example, lepidic-predominant adenocarcinoma (LPA), consisting of bland pneumocytic cells (type II pneumocytes or Clara cells) growing along the surface of alveolar walls, is a nonmucinous adenocarcinoma that can have necrosis, invasion of lymphatics or blood vessels, and at least one focus of invasion larger than 5 mm. Acinar-predominant adenocarcinoma shows a majority component of glands with a central luminal space surrounded by

http://radiologyebook.com

988

PART II  CT and MR Imaging of the Whole Body

tumor cells.296 Papillary-predominant adenocarcinoma shows a major component of a growth of glandular cells along central fibrovascular cores, and micropapillary-predominant adenocarcinoma has tumor cells growing in papillary tufts that lack fibrovascular cores.296 Solidpredominant adenocarcinoma has a major component of polygonal tumor cells forming sheets that lack recognizable patterns of acinar, papillary, micropapillary, or lepidic growth. For mucinous malignancies the term invasive mucinous adenocar­ cinoma has replaced “mucinous BAC.” Invasive mucinous adenocarcinoma has a distinctive histologic appearance, with tumor cells having a goblet or columnar cell morphology with abundant intracytoplasmic mucin. These tumors can show the same heterogeneous mixture of lepidic, acinar, papillary, micropapillary, and solid growth as nonmucinous tumors. However, these tumors differ from mucinous AIS and MIA by being larger than 3 cm and/or showing invasion greater than 0.5 cm and/or multiple nodules, or lack of a circumscribed border with miliary spread into adjacent lung parenchyma. Invasive mucinous adenocarcinoma typically manifests as solid nodules or consolidative opacities. Adenocarcinomas typically manifest radiologically as peripheral solitary pulmonary nodules. Historically, nodules have been described as having soft tissue attenuation and an irregular or spiculated margin as a result of parenchymal invasion and an associated fibrotic response271,291 (Fig. 37-4). However, with the increasing use of CT and recent interest in screening for lung cancer, adenocarcinomas manifesting as purely ground-glass or part-solid attenuation are being detected with increasing frequency (Fig. 37-5). In this regard nonmucinous minimally invasive and lepidic-predominant adenocarcinomas typically manifest as solitary GGNs or part-solid nodules (PSNs). Acinar-, papillary-, micropapillary-, or solid-predominant patterns and invasive mucinous adenocarcinomas manifest as solid nodules, although invasive mucinous adenocarcinomas can also be consolidative opacities and are frequently multifocal (Figs. 37-6, 37-7, and 37-8). To date there has not been a large series validating the radiologicpathologic correlation of the proposed classification as it pertains to the size of the invasive histologic component and solid component detected on CT in MIA and LPA. However, Suzuki et al. characterized the presence and distribution of soft tissue within 349 stage I

adenocarcinomas that were 2 cm or less in size.286 The tumors ranged from pure GGNs to PSNs on CT. A direct correlation was reported between the presence and size of the solid component of PSNs with the degree of invasion of adenocarcinomas.286 Additionally, the highresolution (HR)CT appearance of these opacities has been reported to have a correlation with a classification proposed by Noguchi et al., whereby small (≤2 cm) peripheral adenocarcinomas are classified into six types based on tumor growth patterns: type A, localized BAC; type B, localized BAC with foci of structural collapse of alveoli; type C, localized BAC with active fibroblastic proliferation; type D, poorly differentiated adenocarcinoma; type E, tubular adenocarcinoma; and type F, papillary adenocarcinoma with a compressive growth pattern.154,215,289 Ground-glass attenuation of nodular opacities has been reported to be more frequent in types A to C than types D to F,

FIG 37-5  Adenocarcinoma manifesting as a pulmonary nodule with ground-glass attenuation. CT scan shows a well-marginated groundglass opacity in the left upper lobe (arrow). Surgical clips are present because of prior left lower lobe lobectomy. Note nonmucinous minimally invasive and lepidic-predominant nonmucinous adenocarcinomas typically manifest as solitary ground-glass or part-solid nodules.

FIG 37-4  Adenocarcinoma manifesting as a pulmonary nodule. CT scan shows a solid nodule in the right upper lobe (arrow). The spiculated margin is typical of lung cancer. Note acinar, papillary, micropapillary, or solid-predominant patterns and invasive mucinous adenocarcinomas manifest as solid nodules.

FIG 37-6  Adenocarcinoma manifesting as multiple pulmonary nodules. CT shows small pulmonary nodules and ground-glass opacities. Cavitary nodules with central areas of cavitation are occasionally referred to as the Cheerio sign.

http://radiologyebook.com

CHAPTER 37  Neoplastic Disease of the Lung

A

989

B FIG 37-7  Adenocarcinoma mimicking pneumonia. A, Posteroanterior chest radiograph shows poorly marginated homogeneous consolidative opacity in the right lower lobe. B, CT confirms right lower lobe consolidation and reveals ground-glass opacities in the right lower lobe. Note invasive mucinous adenocarcinoma can manifest as consolidative opacities and are frequently multifocal.

carcinoma (3% of lung cancers in surgical series).82,254,291,303 Most are peripheral, poorly marginated masses greater than 7 cm in diameter254 (Fig. 37-10). Although growth is typically rapid, cavitation is uncommon. Hilar and mediastinal adenopathy occurs in up to one third of patients at presentation, and early extrathoracic metastases are common.271

Small Cell Lung Carcinoma.  SCLCs comprise 13% to 15% of all

FIG 37-8  Adenocarcinoma manifesting as nodules and consolidation. CT scan shows an opacity in the right upper lobe resembling pneumonia and numerous small solid and ground-glass nodules in the right lung.

whereas soft tissue attenuation is more frequent in types B to F.154 The soft tissue attenuation component tends to be absent or less than a third of the opacity with type A and greater in extent (more than two thirds) in types D to F. Additionally the likelihood of invasive adenocarcinoma and more advanced stage of lung cancer has been reported to be more likely with mixed and solid opacities.156,286 Lymphangitic carcinomatosis, although uncommon at presentation, occurs more frequently with adenocarcinomas and typically manifests radiologically as thickening of interlobular septa or multiple small pulmonary nodules (Fig. 37-9). Intrathoracic metastases to hilar and mediastinal nodes are present in 18% to 40% and in 2% to 27% of patients, respectively, and tend to occur more often with more centrally located adenocarcinomas.254,291,323 Large cell carcinoma.  These lesions make up 10% to 20% of all lung cancers and include the variant large cell neuroendocrine lung

lung cancers and have been decreasing in incidence in the United States over the last few decades.110,246,294 SCLC is characterized histologically by small cells with scant cytoplasm, marked nuclear atypia, high mitotic rate (>10 mitoses/10 high-power field [HPF]) and extensive necrosis. There are two subtypes: pure SCLC (≈70%) and combined SCLC (≈30%).246,294 Immunohistochemical stains including chromogranin and synaptophysin, CD56, TTF-1, and Ki-67, together with pancytokeratin antibodies, are used to differentiate SCLC from other lung neoplasms. The Ki-67 proliferation index is typically high (80%100%) and is useful in separating SCLC from carcinoid tumors.246,294 The primary tumor is typically small, central in location, and associated with marked hilar and mediastinal adenopathy and distant metastases to liver, bone marrow, adrenals, and brain254,303 (Fig. 37-11). Pleural effusions occur in 5% to 40% of patients. Approximately 5% of SCLCs manifest as small, peripheral, solitary pulmonary nodules without intrathoracic adenopathy, disseminated extrathoracic disease, or pleural effusions.109,303

Clinical Manifestations Most patients are in their fifth and sixth decades of life, and approximately three quarters of patients are symptomatic at presentation.220 Symptoms are variable and depend on the local effects of the primary mass, presence of regional or distant metastases, and coexistence of paraneoplastic syndromes. Although solitary peripherally located tumors tend to be asymptomatic, central endobronchial carcinomas can manifest with fever, dyspnea, hemoptysis, and cough.111 Of these, cough is the most common presentation and may result from endobronchial irritation, parenchymal infiltration, or postobstructive pneumonia.220 Cough productive of copious amounts of watery sputum (bronchorrhea)

http://radiologyebook.com

A

B FIG 37-9  Adenocarcinoma of the lung manifesting as lymphangitic carcinomatosis. A and B, Thin-section CT scan shows thickening of bronchial walls and interlobular septa (arrowheads) in the left lung. Nodularity of the septa is suggestive of malignancy.

A

B FIG 37-10  Large cell lung cancer manifesting as a large mass. A, Posteroanterior chest radiograph shows a large mass in the right upper lobe. B, CT shows a well-circumscribed mass. Note heterogeneous attenuation of the mass is consistent with necrosis and nonenlarged paratracheal lymph nodes (arrow).

*

A

B FIG 37-11  SCLC manifesting as extensive hilar and mediastinal adenopathy. A, Posteroanterior chest radiograph shows a large poorly marginated right mass, multicompartmental mediastinal adenopathy, and obstructive consolidation and atelectasis of the right lung. B, CT scan confirms marked hilar and mediastinal adenopathy and reveals marked narrowing of the bronchus intermedius (*). Note consolidative and atelectatic opacities distal to the large right hilar mass and a small right effusion. http://radiologyebook.com

CHAPTER 37  Neoplastic Disease of the Lung

991

AA

* DA

A

B FIG 37-12  SCLC in a 65-year-old man presenting with hoarseness caused by involvement of the recurrent laryngeal nerve. A, Posteroanterior chest radiograph shows a left upper lobe mass extending to the perihilar region (arrows) and elevation of the left hemidiaphragm. Note diaphragmatic paralysis is due to phrenic nerve involvement. B, CT reveals mediastinal invasion with extension of the mass into the aortopulmonary window (*) (the anatomic location of recurrent laryngeal nerve). Note heterogeneous attenuation of the mass is consistent with necrosis. AA, ascending aorta; DA, descending aorta.

A

B FIG 37-13  NSCLC in a 63-year-old woman with a superior sulcus (Pancoast) tumor and a history of shoulder pain. A, Posteroanterior chest radiograph shows a poorly marginated right apical mass. B, CT 6 months after A confirms the superior sulcus tumor and shows destruction of the first and second ribs (arrowheads).

typically also occurs in patients with invasive mucinous adenocarcinoma and extensive parenchymal disease, but this is rarely seen. Symptoms that can result from local growth and invasion of adjacent nerves, vessels, and mediastinal structures include: 1. Chest pain (peribronchial nerve involvement), vocal cord paralysis, and hoarseness (recurrent laryngeal nerve involvement) (Fig. 37-12). Recurrent laryngeal nerve paralysis more commonly involves the left vocal cord because the right-side recurrent laryngeal nerve does not traverse extensively into the chest. Other symptoms are dyspnea due to diaphragmatic paralysis (phrenic nerve involvement) (see Fig. 37-12) and Horner’s

syndrome (ptosis [narrowing of the palpebral fissure], miosis [pupillary constriction], anhidrosis [absence of sweating on one side of the face]) due to sympathetic chain and stellate ganglion involvement by superior sulcus tumors (Fig. 37-13). Non–small cell carcinomas of the superior pulmonary sulcus represent 3% of all lung cancers and can also manifest as arm pain and atrophy of the intrinsic muscles of the hand.38,39,97 2. Facial and upper truncal edema, headaches, neck vein distention, and enlarged collateral chest wall vessels (superior vena cava obstruction). The clinical manifestations vary with the rapidity with which obstruction of the superior cava develops

992

PART II  CT and MR Imaging of the Whole Body

and the location of the occlusion (i.e., occlusion proximal to the juncture of the azygous vein and superior vena cava is associated with more marked clinical manifestations). 3. Dysphagia (esophageal involvement). Esophageal involvement is usually due to extrinsic compression by the tumor or associated mediastinal lymphadenopathy. Formation of a tracheo- or bronchoesophageal fistula is uncommon (frequency of 0.16%) and manifests as cough and recurrent aspiration pneumonia.111 Many patients present with nonspecific systemic manifestations of malignancy, including anorexia, weight loss, or fatigue.78 Additionally the symptoms can be related to the site of metastatic disease, with bone pain and symptoms of brain metastases (e.g., headache, nausea, vomiting, seizures, mental status changes) being most common.220 Clinical signs and symptoms can also be caused by tumor excretion of a bioactive substance or hormone, or as a result of immune-mediated neural tissue destruction caused by antibody- or cell-mediated immune responses. These paraneoplastic syndromes occur in 10% to 20% of lung cancer patients and are usually associated with SCLC.56,94,220 Antidiuretic and adrenocorticotropin hormones are the more frequently excreted hormones and can result in hyponatremia and serum

hypoosmolarity and in Cushing’s syndrome (central obesity, hyper­ tension, glucose intolerance, plethora, hirsutism), respectively.94,220 Other hormones that can be elevated are calcitonin, growth hormone, human chorionic gonadotropin, and (rarely) prolactin and serotonin. Neurologic paraneoplastic syndromes (Lambert-Eaton myasthenic syndrome—clinically characterized by craniocaudally progressive proximal muscle weakness that occurs predominantly at the hip girdle; anti-Hu syndrome—limbic encephalitis, brainstem encephalitis, cerebellar degeneration, opsoclonus myoclonus, myelopathy, cranial nerve palsy, sensory neuropathy; and anti-Yo syndrome—primarily brainstem abnormalities and paraneoplastic cerebellar degeneration) are rare and usually associated with SCLC.56,174,220 The neurologic symptoms typically precede the diagnosis of lung cancer by up to 2 years, are incapacitating, and progress rapidly, although improvement can occur after treatment of the lung cancer.56,311 Miscellaneous paraneoplastic syndromes associated with lung cancer include acanthosis nigricans, dermatomyositis, disseminated intravascular coagulation, and hypertrophic pulmonary osteoarthropathy (HPO). Of these, HPO is the most common, occurring in up to 17% of patients with adenocarcinoma200 (Fig. 37-14).

A

FIG 37-14  NSCLC in an asymptomatic 64-year-old woman. A, CT scan shows a large right upper lobe mass. B, Technetium 99m (99mTc)-labeled methylene diphosphonate (MDP) bone scintigraphy shows linear areas of increased radiotracer uptake in the femurs and tibias. This appearance is characteristic of hypertrophic pulmonary osteoarthropathy.

B

CHAPTER 37  Neoplastic Disease of the Lung

993

Radiologic Evaluation Imaging has an important role in screening for malignancy as well as in diagnosis, staging, and monitoring of patients with lung cancer.

Screening.  Because diagnosis of lung cancer at an early stage is associated with an improved prognosis, screening has been advocated to detect lung cancer before clinical presentation. During the 1990s, multiinstitutional single-arm low-dose CT screening studies in the United States, as well as other small trials in Japan and Germany, were performed to examine the role of screening in lung cancer management and showed higher-than-expected lung cancer detection at early stages.68,123,130,144,276 Following these results, an international single-arm study was initiated, the International Early Lung Cancer Action Program, and reported a high percentage of clinical stage I lung cancer detected at baseline and annual screening.124 However, to overcome the various biases inherent in the results of single-arm trials, randomized controlled trials have been conducted.2,136,261 The National Lung Screening Trial (NLST), a randomized study comparing the effectiveness of low-dose CT (LDCT) to chest radiography in more than 50,000 participants, reported significant reductions in lung cancer (20%) and all-cause mortality (6.7%).2 Accordingly, LDCT is currently advocated as a screening tool for lung cancer. However, optimal screening strategies have not been determined. In this regard, it is uncertain whether younger patients and/or smokers with less pack-years’ smoking history than the eligibility criteria used for those examined in the NLST (55- to 74-year-old patients with at least 30 pack-years’ smoking history who were still active smokers, or former smokers who stopped smoking within the previous 15 years) will also benefit from screening. It is hoped that contentious issues including overdiagnosis bias, whether detection of preclinical disease affects survival, and the costeffectiveness of screening programs will be resolved in the coming years.64,190,228

Diagnosis Because most patients with lung cancer have advanced disease at presentation, diagnosis is usually not difficult. Nevertheless, 20% to 30% of patients with lung cancer present with a solitary pulmonary nodule that may be difficult to differentiate from a benign nodule.154 Certain morphologic and physiologic features, however, may suggest a diagnosis of lung cancer: 1. Size. The larger the nodule, the more likely it is to be malignant.114,341 Small size, however, cannot be used to reliably exclude lung cancer, since many resected lung cancers are less than 2 cm in diameter. Additionally the widespread use and improvements in CT technology, coupled with a recent interest in CT screening for lung cancer, has resulted in frequent and incidental detection of small nodules (1-5 mm).123,124,144,276,287 Although the majority of these nodules are benign, studies of resected small nodules have shown that a considerable number are either primary or secondary pulmonary malignancies.98,154,203,207 2. Margins. Lung cancers typically have irregular or spiculated margins.114,271,341 Although suggestive of lung cancer, these findings can occasionally be seen with benign nodules.114,341 Furthermore, a smooth margin, a feature typical of benign nodules, cannot be used to exclude lung cancer, because 20% of malignant nodules have this appearance.154 3. Internal morphology. Except for fat (attenuation −40 to −120 Hounsfield units [HU]) and calcification within a nodule, internal morphology is unreliable in distinguishing lung cancer from a benign nodule.114,341 Calcification occurs histologically in up to 14% of lung cancers and may be detected by CT.183 It is

FIG 37-15  NSCLC manifesting as a calcified mass. CT shows a large right upper lobe mass with mediastinal invasion. Scattered areas of amorphous calcification in the mass are typical of (but not diagnostic for) malignancy.

typically amorphous in appearance (Fig. 37-15), unlike the diffuse solid, central punctate, or laminated calcifications that are diagnostic of benign nodules. Amorphous calcification, however, is not diagnostic of lung cancer, because similar calcifications are occasionally detected in benign lesions. Widespread use of thin collimation CT images has increased the detection of soft tissue nodules with low attenuation (GGNs, PSNs)100,103,154 (Fig. 37-16). Henschke et al. have reported that part-solid nodular opacities have the highest incidence of malignancy (63%) as opposed to 18% for ground-glass opacities and 7% malignancy rate for solid nodules.125 Importantly, with these nodules the likelihood of malignancy varies according to the degree of soft tissue attenuation. However, the soft tissue component of PSNs can represent the invasive component and/or fibrosis and alveolar collapse. The degree of invasion is reported to correlate directly with the size of the soft tissue component on CT.16,132 To improve characterization of the degree of invasion of part-solid adenocarcinomas, Yamada et al. measured CT attenuation numbers of the solid part of PSNs and reported that CT attenuation number was significantly higher in the group with invasion compared to the no-invasion and microinvasion groups.328 In addition the presence of notches in the nodule margin and pleural tags are more frequent in invasive adenocarcinoma as compared to MIA and AIS.132 The frequency of lobulation, spiculation, well-defined but coarse interface, and pseudocavitation (small focal lucencies) have also been reported to be significantly higher in malignant PSNs than benign nodules.77 However, there is considerable overlap, and infectious and inflammatory etiologies can also manifest as pseudocavitation. Cavitation occurs in benign nodules and lung cancer. Malignant nodules typically have thick irregular walls, whereas benign nodules have smooth thin walls.341 These findings, however, are not specific, and wall thickness cannot be used to confidently differentiate benign nodules from lung cancer. 4. Growth. Based on radiographic analysis, lung cancers typically double in volume (an increase of 26% in diameter) between 30 and 400 days (average, 240 days).176 However, it is difficult to reliably detect growth in small nodules, because the small change in diameter associated with a doubling of volume may not be visible on radiographs.176 The use of CT can improve the accuracy of growth assessment, and growth can be detected in lung cancers as small as 5 mm when CT imaging is repeated within 30 days.330 In a screening study analyzing the growth rates of small lung cancers using CT, Hasegawa et al. reported that

994

PART II  CT and MR Imaging of the Whole Body BOX 37-2  Recommendations for

Management of Incidentally Detected Solid Nodules

*

FIG 37-16  Adenocarcinoma manifesting as a part ground-glass and solid nodule. CT shows a nodule with ground-glass and solid tissue (*) attenuation (arrowheads). Part-solid nodular opacities have the highest incidence of malignancy when compared to pure ground-glass opacities and solid nodules.

doubling in volume ranged from 52 to 1733 days (mean, 452 days), and approximately 20% (12 of 61) of these opacities had a volumetric doubling time greater than 2 years.119 These nodular opacities were typically well-differentiated adenocarcinomas. In this regard, the use of stability in size for 2 years to infer benignity has been questioned by Yankelevitz et al., and it is recommended that this criterion be used with caution.331 In addition, when determining growth of a PSN, the limitations in assessing growth are compounded because these lesions are typically small, poorly defined, and growth can be indolent and difficult to perceive.119 Also, in contrast to growth in solid nodules based only on size, growth of PSNs can manifest as increase in size and/or increase in attenuation and/or development or increase in size of a solid component. These imaging features of growth are suspicious for increased risk of malignancy.142 It is important to emphasize that in the serial evaluation of GGNs or PSNs to determine growth, 37% to 70% of screen-detected nodules resolve on follow-up imaging.101,102 Measurement of serial volumes rather than diameters, and computer-calculated doubling of volume of small nodules have been suggested to be an accurate and potentially useful method to assess growth.141,247 In a study by Jennings et al. analyzing the growth rates of stage I lung cancers with serial volumetric CT measurements, the median time to double in volume was 207 days (50% of the tumors doubled in volume at 8 weeks and 75% at 14 weeks).141 However, there was a wide variability in growth rates, and 21 of 149 tumors did not increase or decreased in volume between examinations, and two tumors doubled in volume after 24 months. Recently nodule mass, defined as the combination of nodule volume and density, has been proposed

A.  Low-Risk Populations (Little or No History of Smoking and No Other Risk Factors) i. Nodule ≤ 4 mm. Very small likelihood of malignancy, and no reassessment is necessary. ii. Nodule > 4 mm but ≤ 6 mm. Reassessment CT at 12 months, and if stable no further evaluation is required. The exception is the nonsolid or partially solid nodule, for which reassessment may need to be continued to exclude risk of an indolent adenocarcinoma. iii. Nodule > 6 mm but ≤ 8 mm. Reassessment CT at 6 to 12 months, and if stable again at 18 to 24 months. iv. Nodule > 8 mm. Either reassessment CT scans at 3, 9, and 24 months to assess for stability in size or further evaluation with contrast-enhanced CT, CT-PET, or biopsy or resection. B.  High-Risk Populations (History of Smoking or Other Exposure or Risk Factor) i. Nodule ≤ 4 mm. Reassessment at 12 months, and if stable no further evaluation is required. The exception is the nonsolid or partially solid nodule, for which reassessment may need to be continued to exclude risk of an indolent adenocarcinoma. ii. Nodule > 4 mm but ≤ 6 mm. Reassessment CT at 6 to 12 months and if stable again at 18 to 24 months. iii. Nodule > 6 mm but ≤ 8 mm. Reassessment CT at 3 to 6 months and if stable again at 9 to 12 months and 24 months. iv. Nodule > 8 mm. Either reassessment CT at 3, 9, and 24 months to assess stability or perform contrast-enhanced CT, CT-PET, or biopsy or resection. From MacMahon H, et al: Guidelines for management of small pulmonary nodules detected on CT scans: A statement from the Fleischner Society. Radiology 237:395–400, 2005.

as a more accurate determination of growth of PSNs.58 By multiplying nodule volume and density, mass measurements allow detection of growth of PSNs earlier and are subject to less variability than volume or diameter measurements. At present the determination of when and how to observe and image small nodules in the assessment of tumor growth rate has not been resolved. However, the Fleischner Society has guidelines for evaluation of an incidentally discovered solid nodule in an adult patient that integrate lesion morphology, growth rate, patient age, and smoking history (Box 37-2). In addition, to complement the recommendations for incidentally detected solid nodules, the Society has recently published recommendations specifically aimed at the management of GGNs and PSNs181,204 (Box 37-3). Importantly for GGNs and PSNs, risk factors such as smoking history, familial history of lung cancer, or exposure to carcinogenic agents are not considered in the current guidelines owing to a lack of sufficient data. In addition, other issues to be aware of are that a slight temporary decrease in size can be seen, with adenocarcinomas manifesting as GGNs or PSNs owing to fibrosis or atelectasis; enlargement and/or increasing attenuation with or without the new appearance of a solid component during follow-up should be managed with a high degree of suspicion.204 5. Blood supply. Blood supply to malignant pulmonary nodules is qualitatively and quantitatively different from the blood supply

CHAPTER 37  Neoplastic Disease of the Lung to benign nodules. Contrast-enhanced CT can be used to image this difference by determining nodule enhancement. Typically, malignant nodules enhance more than 20 HU, whereas benign nodules enhance less than 15 HU288 (Fig. 37-17). A computeraided design (CAD) system using quantitative features to

describe the nodule’s size, shape, attenuation, and enhancement properties has been reported to assist in the differentiation of benign and malignant nodules.269 The study showed that CAD using volumetric and contrast-enhanced data to evaluate solitary pulmonary nodules with a mean diameter of 25 mm (range, 6-54 mm) was useful in assisting in the differentiation of benign and malignant solitary pulmonary opacities.269 6. Metabolism. Metabolism of glucose is typically increased in lung cancer cells. Positron emission tomography (PET) using a d-glucose analog, fluorine-18 fluorodeoxyglucose (FDG), can be used as an additional study to improve the characterization of focal pulmonary opacities. However, because the intrinsic spatial resolution of a PET scanner is on the order of 5 to 7 mm, false-negative studies can result when nodules smaller than 1 cm in diameter are evaluated.126,216 FDG uptake can be assessed visually on PET images by comparing the activity of the lesion with that of the background, or the FDG uptake can be quantified by using several parameters including standardized uptake value (SUV) or local metabolic rate of glucose (Ki = influx constant). SUV of greater than 2.5 is generally considered to be indicative of malignancy, but this number is typically not used as a rigid cutoff to differentiate between a benign and a malignant nodule. In a meta-analysis of 40 studies, FDG PET was found to operate at a point on the summary receiver operating characteristic curve corresponding to a sensitivity of 96.8% and specificity of 78%.108 Although there was no difference in diagnostic accuracy according to size, few data exist for nodules smaller than 1 cm in diameter.108 It is important to emphasize that the published data regarding FDG PET evaluation of solitary pulmonary opacities indicating high sensitivity, specificity, and accuracy mostly pertains to nodules that are both solid and 1 cm or greater in diameter. Because the probability of malignancy is high when a solid nodule of 1 cm or greater has increased FDG uptake (Fig. 37-18), these nodules should be biopsied or resected, although close radiologic

BOX 37-3  Recommendations for

Management of Subsolid Pulmonary Nodules Detected at CT A.  Recommendations for Management of Incidentally Detected Solitary Ground-Glass and Part-Solid Nodules i. Solitary pure ground-glass nodules (GGNs) ≤ 5 mm require no CT follow-up. ii. Nodule > 5 mm. Initial follow-up CT at 3 months to confirm persistence, then annual surveillance CT for a minimum of 3 years. iii. Solitary part-solid nodules (PSNs). Initial follow-up CT at 3 months to confirm persistence. If persistent and solid component < 5 mm, then yearly surveillance CT for a minimum of 3 years. If persistent and solid component > 5 mm, then biopsy or surgical resection. Consider PET-CT for PSNs > 10 mm. B.  Recommendations for Management of Incidentally Detected Multiple Ground-Glass and Part-Solid Nodules i. Pure GGNs < 5 mm. Obtain follow-up CT at 2 and 4 years. ii. Pure GGNs > 5 mm without a dominant lesion(s). Initial follow-up CT at 3 months to confirm persistence, and then annual surveillance CT for a minimum of 3 years. iii. Dominant nodule(s) with part-solid or solid component. Initial follow-up CT at 3 months to confirm persistence. If persistent, biopsy or surgical resection is recommended, especially for lesions with > 5 mm solid component. From Naidich DP, et al: Recommendations for the management of subsolid pulmonary nodules detected at CT: A statement from the Fleischner Society. Radiology 266:304–317, 2013.

A

995

B FIG 37-17  NSCLC appearing as an enhancing nodule after administration of contrast material. A, Non– contrast-enhanced CT scan shows a left lung nodule with an attenuation value of 24 HU. B, Contrastenhanced CT scan shows nodule enhancement of 70 HU and central necrosis. Enhancement of more than 20 HU suggests malignancy. Resection revealed non–small cell cancer. (Courtesy Tom Hartman, MD, Mayo Clinic, Rochester, Minn.)

996

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 37-18  NSCLC manifesting as a hypermetabolic nodule on an integrated CT-PET scan with FDG. A, CT scan shows a spiculated nodule in the right upper lobe. The irregular margin suggests malignancy. Note diffuse emphysema. B, Axial fused CT-PET image with FDG shows increased uptake within the nodule when compared to adjacent mediastinum; findings are suspicious for malignancy. Transthoracic needle aspiration biopsy revealed adenocarcinoma.

surveillance is an option if the clinical likelihood of infection is high. Although FDG PET can reduce the number of benign nodules resected, benign neoplasms and nodules due to infection and inflammation (e.g., rheumatoid nodules, tuberculosis, histoplasmosis) can result in increased FDG uptake. When FDG uptake by a solid nodule 1 cm or greater in diameter is low, the likelihood of malignancy is generally considered to be low. However, in a recent study of 360 patients with lung nodules evaluated by FDG PET, 43 patients had solid nodules with an SUV less than 2.5, 16 of which were malignant.120 In this study visual analysis was as accurate as semiquantitative evaluation, and the probability of malignancy was very low if FDG uptake was visually absent and 60% if visually present. False-negative PET results also tend to occur with carcinoid tumors and mucinous adenocarcinomas.74,129 False-negative PET results can also occur in malignant nodules that are smaller than 1 cm in diameter or have ground-glass or part-solid morphology. Nomori et al. reported that 9 of 10 well-differentiated adenocarcinomas manifesting as ground-glass nodular opacities were falsely negative on FDG PET, while 4 of 5 benign ground-glass nodular opacities were falsely positive.216 The sensitivity (10%) and specificity (20%) for ground-glass opacities in this study were significantly lower than those for solid nodules (90% and 71%, respectively).

Staging Because treatment and prognosis depend on the anatomic extent of NSCLC at initial presentation, accurate assessment is important. The seventh edition of the American Joint Committee on Cancer (AJCC) is used to describe the extent of NSCLC in terms of the primary tumor (T descriptor), lymph nodes (N descriptor), and metastases (M descriptor).233,238,239,260 The TNM descriptors can be determined clinically (history, physical examination, radiologic imaging) or by pathologic analysis of samples obtained by biopsy or surgery. In general the clinical stage underestimates the extent of disease when compared with the pathologic stage. In terms of the descriptors, radiologic imaging is usually directed at detecting nonresectable disease (N3 or M1). The detection of contralateral nodal (N3) and/or metastases (M1a, M1b) are important because these typically preclude surgical resection or require additional chemotherapy or radiotherapy. How­ever, there is

currently little consensus on the imaging that should be performed for appropriate staging evaluation in patients presenting with NSCLC. The American Society of Clinical Oncology (ASCO) has published evidence-based guidelines for the diagnostic evaluation of patients with NSCLC.230 In the staging of locoregional disease these guidelines recommend that a chest radiograph and contrast-enhanced chest CT that includes the liver and adrenals should be performed. In addition, whole-body FDG PET is an integral component of NSCLC staging; it improves detection of nodal and distant metastases and frequently alters patient management.83,165,244,310,312 FDG PET complements radiologic findings, and the 2003 ASCO recommendations are that FDG PET should be performed when there is no evidence of distant metastatic disease on CT.230 Presently, FDG PET is usually used together with CT because the relatively poor spatial resolution of FDG PET limits its utility in the evaluation of the primary tumor (T descriptor) and in the determination of the precise anatomic location of regions of focal increased FDG uptake (N and M descriptors).

Primary Tumor (T) Status.  The T status defines the size, location, and extent of the primary tumor (Box 37-4). There are numerous changes to the T descriptor in the seventh edition of the TNM classification of lung cancer that are based on differences in survival: (1) T1 is now subclassified as T1a (≤2 cm) or T1b (>2 cm to ≤ 3 cm); (2) T2 is now subclassified as T2a (>3 cm to ≤ 5 cm or T2 by other factor and ≤ 5 cm) or T2b (>5 cm to ≤ 7 cm); (3) T2 tumors larger than 7 cm are reclassified as T3; (4) T4 tumors by additional nodule(s) in the lung (primary lobe) are reclassified as T3; (5) M1 by additional nodule(s) in the ipsilateral lung (different lobe) is reclassified as T4; and (6) T4 pleural dissemination (malignant pleural effusions, pleural nodules) is reclassified as M1.238 Because the extent of the primary tumor determines therapeutic management (surgical resection or palliative radiotherapy or chemotherapy), imaging is often performed to assess the degree of pleural, chest wall, and mediastinal invasion. Invasion of the chest wall by a primary NSCLC tumor of any size is designated T3, and en bloc resection (removal of lung parenchyma in continuity with a portion of the adjacent parietal pleura and chest wall) is considered the treatment of choice.63,160,184 In this regard, NSCLC invading the chest wall accounts for 5% to 8% of resected cases. The main objectives of preoperative imaging in patients presenting with a peripheral NSCLC abutting the

CHAPTER 37  Neoplastic Disease of the Lung

997

BOX 37-4  International Association for the Study of Lung Cancer TNM Descriptors T—Primary Tumor TX Primary tumor cannot be assessed, or tumor proven by the presence of malignant cells in sputum or bronchial washings but not visualized by imaging or bronchoscopy T0 No evidence of primary tumor Tis Carcinoma in situ T1 Tumor ≤ 3 cm in greatest dimension, surrounded by lung or visceral pleura, without bronchoscopic evidence of invasion more proximal than the lobar bronchus (i.e., not in the main bronchus) T1a Tumor ≤ 2 cm in greatest dimension1 T1b Tumor > 2 cm but not > 3 cm in greatest dimension T2 Tumor > 3 cm but not > 7 cm; or tumor with any of the following features2: • Involves main bronchus 2 cm or more distal to the carina • Invades visceral pleura • Associated with atelectasis or obstructive pneumonitis that extends to the hilar region but does not involve the entire lung T2a Tumor > 3 cm but not > 5 cm in greatest dimension T2b Tumor > 5 cm but not > 7 cm in greatest dimension T3 Tumor > 7 cm or one that directly invades any of the following: chest wall (including superior sulcus tumors), diaphragm, phrenic nerve, mediastinal pleura, parietal pericardium; or tumor in the main bronchus < 2 cm distal to the carina1 but without involvement of the carina; or associated atelectasis or obstructive pneumonitis of the entire lung or separate tumor nodule(s) in the same lobe as the primary T4 Tumor of any size that invades any of the following: mediastinum, heart, great vessels, trachea, recurrent laryngeal nerve, esophagus, vertebral body, carina; separate tumor nodule(s) in a different ipsilateral lobe to that of the primary

N—Regional Lymph Nodes NX Regional lymph nodes cannot be assessed. N0 No regional lymph node metastasis N1 Metastasis in ipsilateral peribronchial and/or ipsilateral hilar lymph nodes and intrapulmonary nodes, including involvement by direct extension N2 Metastasis in ipsilateral mediastinal and/or subcarinal lymph node(s) N3 Metastasis in contralateral mediastinal, contralateral hilar, ipsilateral or contralateral scalene, or supraclavicular lymph node(s) M—Distant Metastasis M0 No distant metastasis M1 Distant metastasis M1a Separate tumor nodule(s) in a contralateral lobe; tumor with pleural nodules or malignant pleural or pericardial effusion3 M1b Distant metastasis Notes 1 The uncommon superficial spreading tumor of any size with its invasive component limited to the bronchial wall, which may extend proximal to the main bronchus, is also classified as T1a. 2 T2 tumors with these features are classified T2a if ≤ 5 cm or if size cannot be determined, and T2b if > 5 cm but not > 7 cm. 3 Most pleural (pericardial) effusions with lung cancer are due to tumor. In a few patients, however, multiple microscopic examinations of pleural (pericardial) fluid are negative for tumor, and the fluid is nonbloody and is not an exudate. Where these elements and clinical judgment dictate that the effusion is not related to the tumor, the effusion should be excluded as a staging element and the patient should be classified as M0.

Reprinted with permission courtesy of the International Association for the Study of Lung Cancer. Copyright © 2009 IASLC.

pleura are to determine whether chest wall resection is required and to evaluate for mediastinal nodal metastasis, because this generally precludes resection.184 Besides nodal metastasis, other factors that influence survival are completeness of resection and the extent of chest wall invasion.160 CT is useful in confirming gross chest wall invasion (Fig. 37-19) but is inaccurate in differentiating between anatomic contiguity and subtle invasion. Findings suggestive of chest wall invasion include63,155,241: • Tumor-pleura contact extending over more than 3 cm • An obtuse angle at the tumor-pleura interface • Thickening of the pleura or increased attenuation of the extrapleural fat adjacent to the tumor Obliteration of the extrapleural adipose plane has been reported to have the best sensitivity (85%) and specificity (87%) for chest wall invasion by a peripheral lung cancer, whereas the length of the tumorpleura contact has a sensitivity of 83% and specificity of 80%.241 Although magnetic resonance imaging (MRI) offers superior soft tissue contrast resolution, the sensitivity (63%-90%) and specificity (84%-86%) of MRI in identifying chest wall invasion is similar to that of CT155,222,237,317 (Fig. 37-20). Respiratory dynamic MRI has been proposed as a complementary imaging modality to CT in determining chest wall invasion (sensitivity 100%, specificity 83%) but is overall not widely used in the evaluation of subtle chest wall invasion.7 However, MRI is particularly useful in the evaluation of superior sulcus (Pancoast) tumors and can be used to assess invasion of the brachial plexus, subclavian vessels, and vertebral bodies38 (Fig. 37-21).

Imaging with CT or MRI is useful in confirming gross invasion of the mediastinum (Figs. 37-22 and 37-23), but these modalities are inaccurate in determining subtle invasion (56%-89% and 50%-93%, respectively).155,317 CT and MRI findings that can be useful in suggesting subtle mediastinal invasion include127,155: • Tumor-mediastinal contact extending over more than 3 cm • Obliteration of the fat plane between the mediastinum and tumor • Tumor contacting more than 90 degrees of the aortic circumference The International Staging System for Lung Cancer combines the T descriptors with the N and M descriptors into subsets or stages that have similar treatment options and prognosis (see Table 37-1). However, it is important to realize that although a T4 descriptor generally precludes resection, patients with cardiac, tracheal, and vertebral body invasion are designated in the seventh edition staging system as being potentially resectable (stage IIIa) in the absence of N2 and/or N3 disease.92,135,160 The largest surgical experience for T4 involvement is in patients with carinal invasion, with less surgical resection experience with tumors invading the atrium, superior vena cava, aorta, and vertebral bodies.30,46,60,76,160,195,240,245 Surprisingly, complete surgical resection has been reported to be possible in many of these patients that on initial staging would have been denied surgery. Besides an improvement in locoregional control, patients who could be completely resected also showed improvement in long-term survival rates that were unforeseen considering their initial clinical staging.160

998

PART II  CT and MR Imaging of the Whole Body

M

* A

B FIG 37-19  NSCLC and chest wall invasion. A, Posteroanterior chest radiograph shows a large cavitary right lower lobe lung mass (arrows) and destruction of the posterior aspect of the right seventh rib (arrowheads). B, CT confirms the cavitary mass (M) and destruction of the adjacent rib (*). Note extension of the mass into the chest wall (arrowheads).

* *

B

A

FIG 37-20  SCLC and chest wall invasion. A, CT shows a large, well-circumscribed, peripheral lung mass with invasion of the chest wall. B, Coronal fast-spin echo MRI shows mass abutting the chest wall and the surrounding ribs (*) and extending through the intercostal space into the soft tissues of the chest wall (arrow).

Regional Lymph Node (N) Status.  The presence and location of nodal metastasis are of major importance in determining management and prognosis in patients with NSCLC.259 To enable a consistent and standardized description of N status, nodal stations are defined by the relation to anatomic structures or boundaries that can be identified before and during thoracotomy (Fig. 37-24). The N descriptors in the seventh edition of the TNM classification of lung cancer have been

maintained because there were no significant survival differences in analysis by station.260 However, it proposed that lymph node stations could be grouped together in 6 zones within the current N1 and N2 patient subsets for further evaluation. Zones were defined as peripheral (stations 12, 13, 14) or hila (stations 10, 11) for N1 and upper mediastinal (stations 1, 2, 3, 4), lower mediastinal (stations 8, 9), aortopulmonary (stations 5, 6), and subcarinal (station 7) for N2 nodes (see

CHAPTER 37  Neoplastic Disease of the Lung

999

A M

R

*

M

PA

LA

FIG 37-23  NSCLC invading the left atrium. Double inversion recovery FIG 37-21  Superior sulcus (Pancoast) tumor with chest wall invasion and involvement of the brachial plexus. Coronal T1-weighted MRI shows a mass (M) in the apex of the right hemithorax, with invasion into the neck. The mass surrounds the subclavian artery (*) and brachial plexus. R, first rib.

FIG 37-22  NSCLC with mediastinal invasion. CT scan shows a large left upper lobe mass with invasion of the anterior mediastinum. Broad abutment and loss of soft tissue plane between the mass and the transverse aorta suggest vascular invasion.

Fig. 37-24).260 This proposal is based on survival analysis by anatomic location (or zone) of involved nodes, the number of zones involved, and presence of skip metastases. Size is the only criterion used to diagnose nodal metastases, with nodes greater than 10 mm in short-axis diameter considered abnormal. However, an inherent limitation is that lymph node size is not a reliable parameter for the evaluation of nodal metastatic disease in patients with NSCLC.234,293 Prenzel et al. recently reported that in 2891 resected hilar and mediastinal nodes obtained from 256 patients with NSCLC, 77% of the 139 patients with no nodal metastases had at least

contrast-enhanced single breath-hold coronal MRI shows an enhancing right upper lobe mass (M) that extends via the right superior pulmonary vein into the left atrium. A, aorta; LA, left atrium; PA, right main pulmonary artery.

one node greater than 1 cm in diameter.234 Furthermore, 12% of the 127 patients with nodal metastases had no nodes greater than 1 cm. A meta-analysis of 20 studies (3438 patients) evaluating CT accuracy for staging the mediastinum yielded a pooled sensitivity of 57% and specificity of 82%.293 Thus, given the limitations of mediastinal nodal staging by CT, biopsy is recommended for confirmation of locoregional nodal metastases when the confirmation of nodal metastasis will have an impact on patient management. Because surgical resection and potential use of adjuvant therapy are dependent on the patient’s N descriptor, attempts have been made to improve the accuracy of detection of nodal metastases. FDG PET complements CT findings and provides information on locoregional nodal staging that affects management14,28,107,244 (Fig. 37-25). In a metaanalysis (17 studies, 833 patients) comparing PET and CT in nodal staging in patients with NSCLC, the sensitivity and specificity of FDG PET for detecting mediastinal lymph node metastases ranged from 66% to 100% (overall 83%) and 81% to 100% (overall 92%), respectively, compared to sensitivity and specificity of CT of 20% to 81% (overall 59%) and 44% to 100% (overall 78%), respectively.28 Because of the improvements of nodal staging when PET-CT is incorporated into the imaging algorithm of those patients with potentially resectable NSCLC, PET-CT should be performed in all patients without CT findings of distant metastasis, regardless of mediastinal nodal size, to direct nodal sampling as well as to detect distant occult metastasis. Although FDG PET for nodal staging is cost-effective and can reduce the likelihood that a patient with mediastinal nodal metastases (N3) that would preclude surgery will undergo attempted resection, the number of false-positive results due to infectious or inflammatory etiologies are too high to preclude mediastinoscopy. In an attempt to determine the necessity for invasive sampling after PET and CT imaging, a meta-analysis evaluated the association between the size of mediastinal nodes and the probability of malignancy.59 The authors reported a 5% posttest probability for N2 disease in patients with a negative FDG PET if the mediastinal nodes were 10 to 15 mm on CT and suggested these patients should proceed directly to thoracotomy.

1000

PART II  CT and MR Imaging of the Whole Body

In patients with a negative FDG PET and lymph nodes 16 mm or larger on CT, the posttest probability for N2 disease was 21%, suggesting that these patients should have mediastinoscopy prior to thoracotomy. In patients with a positive FDG PET the posttest probability of malignancy was 62% if the nodes were 10 to 15 mm on CT and 90% if the nodes were 16 mm or larger. Although the authors did not suggest any management strategies regarding histologic confirmation of nodal metastasis when FDG PET is positive, N2/N3 should be histologically confirmed in patients potentially eligible for resection or adjuvant therapy.

Metastatic Disease (M) Status.  The M descriptor in the seventh edition of the TNM classification of lung cancer is subclassified into M1a (additional nodules in the contralateral lung, malignant pleural effusion, pleural nodule[s], pericardial nodule[s]) and M1b (extrathoracic metastasis), and these metastases are common in patients with NSCLC at presentation.233,239 Up to 33% of patients with NSCLC have pleural metastases (M1a) at presentation. Pleural (or pericardial)

involvement (either multiple implants or a malignant effusion) is classified as M1a because of slightly better survival than for distant metastatic sites and worse survival than for other categories of T4.233 The diagnosis of pleural metastases or malignant effusion, however, can be difficult to confirm. Pleural thickening and nodularity on CT scans suggests metastatic pleural disease (Fig. 37-26), but these abnormalities may not be present in association with a malignant effusion.249 Furthermore, cytologic evaluation has a mean sensitivity of 72% (range, 49%-91%) when at least two pleural fluid specimens are analyzed.249 If the first pleural fluid analysis is nondiagnostic, a second specimen yields a diagnosis of malignancy in about 25% to 28% of patients.249 Although a nodule in the contralateral lung is potentially a metastasis (M1a), most (≈75%) of additional pulmonary nodules on CT imaging in patients with potentially operable clinical stage I to IIIa lung cancer are benign.65,106 An additional nodule may be a synchronous second primary lung cancer (incidence ≈ 1.5%-2% per patient per year).65 The AJCC rules are confusing with regard to stage classification but indicate that when a patient has simultaneous bilateral cancers in paired

FIG 37-24  A, International Association for the Study of Lung Cancer Nodal Chart with stations and zones.

CHAPTER 37  Neoplastic Disease of the Lung

1001

FIG 37-24, cont’d  B, Nodal definitions. (Reprinted with approval by the International Association for Study of Lung Cancer and permission by Memorial Sloan Kettering, © 2009.)

organs, the tumors are classified separately as independent tumors in different organs.11 To clarify the staging confusion, the American College of Chest Physicians (ACCP) recommends that when two lung cancers are deemed to be synchronous primary cancers, they be classified with a TNM descriptor for each tumor.65 The role of imaging in detecting M1b metastases is not clearly defined. Because staging performed on the basis of clinical findings and conventional radiologic imaging incorrectly stage some patients with NSCLC, whole-body FDG PET is increasingly being used to improve the accuracy of staging. In fact the ACCP recommends PET imaging in patients with a normal clinical evaluation and no suspicious extrathoracic abnormalities on chest CT being considered for curative-intent treatment to evaluate for metastases (except the brain).161 In this regard, FDG PET has a higher sensitivity and

specificity than CT in detecting metastases to the adrenals, bones, and extrathoracic lymph nodes. For instance, the results of the American College of Surgeons Oncology Trial reports a sensitivity, specificity, positive predictive value, and negative predictive value of 83%, 90%, 36%, and 99%, respectively, for M1 disease.244 Whole-body PET imaging stages intra- and extrathoracic disease in a single study and detects occult extrathoracic metastases in up to 24% of patients selected for curative resection.182,231,244,310 The incidence of detection of occult metastases has been reported to increase as the staging T and N descriptors increase (i.e., 7.5% in early-stage disease to 24% in advanced disease).182 A randomized controlled trial of the role of PET in earlystage lung cancer (>90% of patients with T1-2/N0) showed that distant metastases were rarely detected (<2.5%), although PET improved the accuracy of staging and led to more age-specific therapy.313 In addition

1002

PART II  CT and MR Imaging of the Whole Body

*

B

A

FIG 37-25  NSCLC and intrathoracic nodal metastases manifesting as hypermetabolic nodes on PET-CT with FDG. A, CT shows enlarged left hilar (arrows) and subcarinal lymph nodes (*). B, Axial fused FDG CT-PET shows increased FDG uptake in mediastinal nodes suspicious for N2 metastasis. Mediastinoscopy confirmed N2 metastatic disease.

FIG 37-26  NSCLC and pleural metastases. Contrast-enhanced CT shows a large right pleural effusion and enhancing nodular pleural metastases (arrowheads). Note compressive atelectasis of the right lower lobe and a metastatic left rib lesion (arrow).

a subset analysis of the patients with clinical stage I tumors in the American College of Surgeons Oncology Group PET scanning study found that PET detected N2, N3, or M1 in 7% of patients.161 However, the false-positive rate for N2, N3, and M1 was 14%. Two studies with a higher proportion of more advanced lung cancers considered resectable by standard clinical staging showed that PET imaging prevented nontherapeutic surgery in 1 in 5 patients.244,310 Metastases to the adrenal glands are common and detected in up to 20% of patients at presentation155,212,219,223,224 (Fig. 37-27). A small (<3 cm) adrenal mass, however, is more likely to be benign in the absence of other extrathoracic metastases.155 CT and MRI can be useful in the evaluation of adrenal masses. CT and MRI features favoring malignancy include:185 • Size greater than 3 cm • Poorly defined margins

• Irregularly enhancing rim • Invasion of adjacent structures • High signal intensity on T2-weighted sequences A confident diagnosis of benignity can be made if an adrenal mass has an attenuation value less than 10 HU on a non–contrast-enhanced CT scan. A meta-analysis of 10 studies to determine an optimal threshold for differentiating benign from malignant lesions yielded a sensitivity of 71% and a specificity of 98% for characterizing adrenal masses with a threshold of 10 HU.35 Although the finding of low attenuation is useful to characterize an adenoma, up to 30% of adenomas do not contain sufficient lipid to demonstrate low attenuation at CT.229 In these cases, delayed contrast-enhanced CT has been shown to be of use. Two features can be assessed at delayed CT: the attenuation value of the adrenal and the washout of contrast media. At 10 minutes after injection of contrast, an attenuation value of less than 30 HU is diagnostic of an adenoma.33 A decrease of the attenuation value of the adrenal mass by 50% at delayed CT is also specific for an adenoma. MRI using chemical shift analysis and dynamic gadolinium enhancement can also be used to determine whether an adrenal mass is benign34,221,268 (Fig. 37-28). FDG PET is useful in distinguishing benign from malignant adrenal masses detected on CT and is particularly advocated when the adrenal mass is small. A meta-analysis of 21 studies, 1391 lesions in which 5 reports were specifically in patients with lung cancer, reported a combined sensitivity of 94% and specificity of 82% of FDG PET in detecting adrenal metastasis in patients with lung cancer.32 PET has changed the imaging algorithm used to evaluate an indeterminate adrenal mass detected on CT and is now often used as the definitive imaging modality rather than MRI, particularly when the adrenal mass is small. In fact adrenal masses can be characterized by FDG PET, and subsequent imaging is usually unnecessary32 (see Fig. 37-27). In this regard, if FDG uptake by an adrenal mass is normal in a patient with potentially resectable NSCLC, curative resection should be considered without further evaluation. Patients with NSCLC and an isolated adrenal mass with increased FDG uptake should have the adrenal lesion biopsied before being denied surgery. Unfortunately the status of some adrenal lesions remains indeterminate after extensive imaging evaluation; biopsy is required.

CHAPTER 37  Neoplastic Disease of the Lung

A

B

C FIG 37-27  NSCLC and adrenal metastasis. A, CT shows a nodule in the right upper lobe (arrow). B, CT of the abdomen reveals a left adrenal nodule (arrow). The presence of adipose tissue is suggestive of an adenoma. C, Axial fused CT-PET with FDG shows increased FDG uptake in the adrenal nodule. Resection of the left adrenal confirmed a metastasis and also revealed an incidental adjacent adenoma.

S

A

B FIG 37-28  NSCLC and incidental adrenal adenoma evaluated by chemical shift MRI. A, In-phase (TR = 200 msec, TE = 4.2 msec) spoiled gradient recalled echo (SPGR) MRI shows a left adrenal mass (arrow) that is slightly hyperintense compared with the spleen (S). B, Opposed-phase (TR = 200 msec, TE = 1.8 msec) SPGR MRI shows that the left adrenal mass is now markedly hypointense. Findings are typical of those for adenoma. The patient underwent resection of the primary lung malignancy.

1003

1004

PART II  CT and MR Imaging of the Whole Body

A

B FIG 37-29  NSCLC and brain metastasis at presentation manifesting as cough, dizziness, and ataxia. A, CT shows a 3.5-cm spiculated left upper lobe mass. B, Axial contrast-enhanced cranial MRI shows an enhancing metastasis in the cerebellum (arrow).

Central nervous system (CNS) metastases are common and detected in up to 18% of patients at presentation133,194 (Fig. 37-29). Detection of brain metastasis in patients with NSCLC and a negative clinical examination by CT/MRI is reported to be 0% to 10%.161 To date, the use of routine MRI in staging patients with NSCLC and a negative clinical evaluation has not been studied adequately. However, Earnest et al. have reported that occult brain metastases were identified in 6 of 27 (22%) patients with potentially resectable NSCLC (excluding earlystage lung cancer) on contrast-enhanced MRI.70 Additionally there is a substantial early postoperative recurrence of tumor in the brain, particularly in patients with non–squamous cell histology and stage of disease higher than T1/N0, suggesting that undetected metastases were present at the time of surgical resection of the primary lung malignancy.251,333 Consequently there may be a limited role for imaging in the detection of occult metastases. In particular, imaging of the brain may be indicated for exclusion of brain metastases in patients with clinically resectable locally advanced NSCLC with nonsquamous histology. Furthermore the ACCP recommends that patients with clinical stage III or IV NSCLC have routine imaging of the brain with MRI (or CT if MRI is not available) even if the clinical evaluation is negative for brain metastases.161 Skeletal metastases, although often symptomatic, can be asymptomatic. It is recommended that bone radiographs, MRI, and technetium 99m methylene diphosphonate (99mTc-MDP) bone scintigraphy be performed only to evaluate a history of focal bone pain or elevated alkaline phosphatase.274 However, FDG PET is a substitute for 99mTcMDP bone scan imaging in the detection of skeletal metastases and has utility in detecting occult osseous metastases. In this regard a recent meta-analysis of 17 studies reported a pooled sensitivity and specificity of 92% and 98%, respectively, for FDG PET-CT and 86% and 88%, respectively, for bone scintigraphy in the detection of bone metastasis.236 In fact FDG PET-CT has to a large extent replaced 99mTc-MDP

bone scintigraphy in the evaluation of possible bone metastasis in patients with NSCLC179,193,236 (Fig. 37-30). Specifically, 99mTc-MDP now has no additional utility in NSCLC patients if FDG PET is performed as part of the staging algorithm. In fact discordant findings of skeletal metastasis between 99mTc-MDP scintigraphy and FDG PET-CT are reported to occur in 20% of patients with NSCLC.6 This discordance is in large part due to the ability of FDG PET to detect early bone metastasis and the failure of 99mTc-MDP scintigraphy to detect early neoplastic infiltration of bone marrow. Stage grouping: TNM subsets.  The International Staging System for Lung Cancer combines the TNM subsets into categories, or stages, that have similar treatment options and prognosis (Table 37-1). Despite the recognition of many new subdivisions of the T and M descriptors, the stage grouping has no new subdivisions. However, the definition of the stage groups has become more complex because of the additional T and M descriptor subgroups.65

Staging Small Cell Lung Carcinoma.  SCLC is generally staged according to the Veterans Administration Lung Study Group (VALSG) recommendations as limited disease (LD) or extensive disease (ED).55 LD defines tumor confined to a hemithorax and the regional lymph nodes. Unlike the TNM classification for NSCLC, metastases to the ipsilateral supraclavicular, contralateral supraclavicular, and mediastinal lymph nodes as well as ipsilateral pleural effusions are considered local disease. ED includes tumor with noncontiguous metastases to the contralateral lung and distant metastases.55,283 Most patients with SCLC have ED at presentation140,155 (Fig. 37-31). Common sites of metastatic disease include the liver, bone, bone marrow, brain, and retroperitoneal lymph nodes.140 The seventh edition AJCC TNM staging system is applicable to SCLC but is seldom used in clinical practice, because this staging has relied historically on surgical confirmation for accuracy and patients

CHAPTER 37  Neoplastic Disease of the Lung

1005

M

A

B FIG 37-30  NSCLC and bone metastases. A, Posteroanterior chest radiograph shows a left lower lobe mass. B, Coronal whole-body PET-CT with FDG shows increased FDG uptake in the primary lung mass (M) and vertebral bodies (arrows). Biopsy confirmed bone metastasis.

TABLE 37-1  International Association for the Study of Lung Cancer Stage Grouping: Tumor-

Node-Metastasis (TNM) Subsets

N0

N1

N2

N3

6th edition TNM

T and M 7th edition TNM

Stage

Stage

Stage

Stage

T1(<3 cm)

T1a (<2 cm) T1b (>2-3 cm) T2a (>3-5 cm) T2b (>5-7 cm) T3 (>7 cm) T3 T3 T4 T4 M1a M1a M1b

IA IA IB IIA (IB) IIB (IA) IIB IIB (IIIB) IIIA (IIIB) IIIA (IV) IV (IIIB) IV IV

IIA IIA IIA (IIB) IIB IIIA (IIB) IIIA IIIA (IIIB) IIIA (IIIB) IIIA (IV) IV (IIIB) IV IV

IIIA IIIA IIIA IIIA IIIA IIIA IIIA (IIIB) IIIB IIIB (IV) IV (IIIB) IV IV

IIIB IIIB IIIB IIIB IIIB IIIB IIIB IIIB IIIB (IV) IV (IIIB) IV IV

T2 (>3 cm)

T3 invasion T4 (same lobe nodules) T4 (extension) M1 (ipsilateral lung) T4 (pleural effusion) M1 (contralateral lung) M1 (distant)

Shaded boxes highlight 7th edition changes in classification, with 6th edition classification in parentheses. Reprinted with permission courtesy of the International Association for the Study of Lung Cancer. Copyright © 2009 IASLC.

with SCLC seldom present at a stage for which surgery is appropriate. Nevertheless, small published series of resected SCLC have suggested that TNM pathologic staging correlates with survival of resected patients. The TNM system has demonstrated the ability to identify potential surgical candidates and may result in improved consistency in radiation treatment planning.143 Accordingly the IASLC has recommended that the TNM staging system replace the VALSG system for staging of SCLC.307

Although there is no consensus regarding the imaging and invasive procedures that should be performed in the staging evaluation of patients with SCLC, MRI has been advocated to assess the liver, adrenals, brain, and axial skeleton in a single study.140 Whole-body PET imaging improves the accuracy of staging of patients with SCLC.17,84,162,213 In this regard, Niho et al. performed FDG PET imaging in 63 patients diagnosed by conventional staging procedures as having limited-stage SCLC.213 Therapeutic management was changed in 5 patients (8%)

1006

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 37-31  SCLC with disseminated metastases at presentation. A, Posteroanterior chest radiograph shows a lobular 3-cm right upper lobe nodule (arrow) and hilar adenopathy (arrowhead). B, CT shows left adrenal metastasis (arrow). C, Axial contrast-enhanced cranial MRI shows an enhancing cerebral metastasis (arrow).

owing to detection of unsuspected distant metastases and change in stage from LD to ED. Evaluation of extrathoracic metastatic disease usually includes the following: 1. Bone marrow aspiration, 99mTc-MDP bone scintigraphy, and MRI. Patients with bone (30%) and bone marrow (17%-34%) metastases are often asymptomatic, and blood alkaline phosphatase levels are frequently normal.55,278,279 Because isolated bone and bone marrow metastases are uncommon, however, routine bone marrow aspiration and radiologic imaging for occult metastases are usually performed only if there are other findings of ED. 2. Brain MRI. CNS metastases are common (10%-27%) at presentation, and approximately 5% of patients are asymptomatic.3,55 Because therapeutic CNS radiation and chemotherapy can decrease morbidity and improve prognosis, routine MRI of the brain is recommended in patients with SCLC.55,72,308 3. CT or MRI of the abdomen. Metastases to the liver (30%) and retroperitoneal nodes (11%) are common at presentation.3,55 Because patients are often asymptomatic and liver function tests can be normal, staging evaluation routinely includes CT or MRI of the abdomen.

Other Lung Malignancies

Sarcomatoid Carcinoma.  Sarcomatoid carcinomas of the lung are poorly differentiated carcinomas containing carcinomatous and sarcomatous components.208,295,321 Molecular evidence supports the concept that these are carcinomas of the lung with a clonal origin from pluripotent stem cells capable of divergent differentiation into carcinomatous and sarcomatous components.295 They represent a clinicopathologic continuum, and five subtypes are recognized: pleomorphic carcinoma, spindle cell carcinoma, giant cell carcinoma, carcinosarcoma, and pulmonary blastoma.295,321 Carcinomas with spindle or giant cells.  Sarcomatoid carcinomas of the lung (pleomorphic carcinoma, spindle cell carcinoma, giant cell carcinoma, carcinosarcoma) are rare, comprising 0.3% to 1% of malignant lung neoplasms.208,295 They are a heterogeneous group of NSCLCs containing a sarcomalike component that histogenetically may represent a malignant epithelial neoplasm undergoing divergent tissue differentiation originating from a single clone.29,256 Most patients are men, and mean age at presentation is 65 years (range, 44-78 years).29,208 Most patients present with cough, dyspnea, hemoptysis, chest pain, or weight loss.208,256 Although sarcomatoid carcinomas are usually localized at presentation, distant metastases occur frequently and the prognosis is poor.208,321

CHAPTER 37  Neoplastic Disease of the Lung Radiologically these neoplasms can manifest either as large peripheral masses or as polypoid endobronchial lesions with atelectasis or postobstructive pneumonia.151,208,295,321 Calcification and cavitation are uncommon, but necrosis and hemorrhage can manifest as heterogeneous attenuation on CT.151,208,321 Hilar or mediastinal adenopathy is uncommon.208 Pleural effusion can occur as a result of local invasion.151 Metastases involve sites similar to those of lung cancer (lung, liver, bones, adrenals, brain).321 Pulmonary blastoma.  Pulmonary blastoma is a rare malignancy that makes up an estimated 0.25% to 0.5% of primary lung tumors.87,166,173,309 The tumor derives its name from its histologic resemblance to fetal lung tissue. Pulmonary blastomas, however, are thought to arise from primitive pluripotential stem cells and may represent a variant of carcinosarcoma.321 Pulmonary blastomas are subdivided in three categories: well-differentiated fetal adenocarcinoma, classic biphasic pulmonary blastoma and pleuropulmonary blastoma, which is currently regarded as a separate entity. Welldifferentiated fetal adenocarcinoma is categorized as a variant of adenocarcinoma of the lung.296 Although the age range at presentation is wide (0-80 years), these tumors typically occur in adults, with a peak incidence between 40 and 60 years of age.87,166,173 However, pleuropulmonary blastomas usually occurs exclusively in children less than 15 years of age, with a median age of 3 years and one-fourth are hereditary.309 Patients are often symptomatic at presentation; cough, hemoptysis, and chest pain are frequent manifestations.4,159 Pediatric age patients typically present with fever and respiratory distress.36,309 The behavior of pulmonary blastomas is aggressive and the outcome is poor due to frequent relapses and metastases.36,51,87 Radiologically, pulmonary blastomas typically manifest as large (range, 2.5-26 cm), well-marginated masses located peripherally in the lung36,177,309 (Fig. 37-32). Multiple masses, cavitation, and calcification are rare. Local invasion of the mediastinum and pleura occurs in 8%

and 25% of cases, respectively.87 Metastases to hilar and mediastinal lymph nodes are present in 30% of resected cases.87 Extrathoracic metastases are common and have a distribution similar to that of lung cancer.87,173

Pulmonary Neuroendocrine Neoplasms.  Four major types of neuroendocrine neoplasms are recognized by the 2004 WHO Classification of Tumors and are grouped into three histologic grades. Typical carcinoid is characterized as low-grade malignant neoplasm, atypical carcinoid as intermediate grade, and large cell neuroendocrine lung carcinoma and SCLC as high-grade malignancies. Carcinoid tumors.  Primary pulmonary carcinoid tumors are lowgrade malignancies that constitute 1% to 2% of primary lung tumors.246,294 They are classified histologically as typical (80%-90%) or atypical (10%-20%) tumors depending on the degree of cellular atypia.23,246,294 Typical carcinoid is a well-differentiated neoplasm with neuroendocrine histologic features of 5 mm or greater size with less than 2 mitoses per 10 HPF and no necrosis, whereas atypical carcinoid has 2 to 10 mitoses per 10 HPF or necrosis. The most useful immunohistochemical markers for identifying neuroendocrine neoplasms are chromogranin, CD56, and synaptophysin; Ki-67 is useful in differentiating typical carcinoid and atypical carcinoid from large cell neuroendocrine lung carcinoma and SCLC.246,294 Typical carcinoid tumors occur with equal frequency in men and women; the mean age at diagnosis is 35 to 50 years.275 Tumors usually arise in lobar, segmental, or proximal subsegmental bronchi and are generally 1 to 4 cm in size.95 However, increased detection of peripheral carcinoid tumors more recently has been attributed to widespread use of CT.189 Typical carcinoid tumors rarely metastasize to regional nodes or beyond the thorax. Atypical carcinoid tumors are usually discovered at a slightly older age (mean, 53-60 years), are often larger, and tend to occur more frequently in the peripheral aspect of the lungs.24,95,275

*

A

1007

B FIG 37-32  Pulmonary blastoma manifesting as a large pulmonary mass. A, Posteroanterior chest radiograph shows complete homogeneous opacification of the left hemithorax and displacement of the mediastinum to the right. B, CT reveals a large lung mass and shows heterogeneous attenuation consistent with necrosis. There is subcarinal adenopathy due to metastatic disease (*). Note the complete compressive atelectasis of the left lung.

1008

PART II  CT and MR Imaging of the Whole Body

They behave more aggressively than typical carcinoid tumors and frequently metastasize to regional nodes, lung, liver, and bone.95,242 In a published series of 661 resected carcinoid tumors (569 typical carcinoid, 92 atypical carcinoid) the 5-year survival was 97% and 78%, respectively. The 5-year survival of patients with atypical carcinoid tumors decreased to 60% if there was nodal involvement. In fact the reported difference in 5-year survival between typical carcinoid tumors and atypical carcinoid tumors may be much more related to the N status than to the histologic subtype, with N2 being the most important prognostic factor.40 Clinical manifestations depend on the histologic type and location of the carcinoid tumor. Peripheral tumors are usually asymptomatic, whereas central neoplasms can manifest as cough, hemoptysis, or recurrent infection (Fig. 37-33). Paraneoplastic manifestations such as

A

carcinoid syndrome (cutaneous flushing, bronchospasm, chronic diarrhea, and valvular heart disease) and Cushing’s syndrome are rare and more common with atypical carcinoid tumors.95,115,294 Some 40% to 50% of patients with carcinoid syndrome develop carcinoid heart disease, characterized by right-sided heart failure caused by fibrotic endocardial plaques.85,232 Carcinoid tumors most commonly manifest radiographically or on CT as central endobronchial masses with or without atelectasis or consolidation25,95 (Fig. 37-34). A peripheral well-marginated pulmonary nodule is a less common manifestation (Fig. 37-35). The tumors are usually less than 3 cm in size, although occasionally they may be as large as 10 cm in diameter.95,306 Calcification is detected by CT in approximately 25% of carcinoid tumors340 (Fig. 37-36). Hilar and mediastinal adenopathy and extrathoracic metastases are uncommon

B FIG 37-33  Typical carcinoid tumor manifesting as recurrent pulmonary infections. A and B, CT shows a mass in the segmental bronchus of the left lower lobe (arrows in A), left lower lobe volume loss, and distal bronchiectasis in the atelectatic lung (arrowheads in B).

M

A

B FIG 37-34  Typical carcinoid appearing as a central endobronchial lesion. A, Posteroanterior chest radiograph shows complete atelectasis of right lower lobe (arrowheads) with compensatory hyperinflation of the left lung. Note displacement of the anterior junction line (arrows). B, CT confirms atelectasis of the right lower lobe and reveals an endobronchial mass (M) with marked narrowing of the bronchus intermedius (arrow).

CHAPTER 37  Neoplastic Disease of the Lung

1009

T

FIG 37-35  Typical carcinoid manifesting as a solitary peripheral nodule. CT shows a well-circumscribed small nodule in the left lower lobe (arrow).

FIG 37-37  Adenoid cystic carcinoma manifesting as a tracheal mass. CT shows a circumferential soft tissue trachea mass (T) and narrowing of the trachea.

FIG 37-36  Atypical carcinoid tumor with rib and hepatic metastases at presentation. CT shows a large endobronchial mass with dense punctate calcification.

at presentation in patients with typical carcinoid tumors and occur more frequently in patients with atypical carcinoid tumors.24,95 Beasley et al. reported that regional and distal nodal metastases are present in 40% to 50% of patients at presentation, and 10% will have distant metastases.24 Although routine use of scintigraphy with somatostatin analogues (indium 111, octreotide, or lanreotide) can be useful for localization of occult tumors, this imaging is usually reserved for staging or to assess response to therapy.115,232 On FDG PET imaging, pulmonary carcinoid tumors usually have lower FDG uptake than expected for malignant tumors.47,74

Salivary Gland–Type Carcinomas Mucoepidermoid carcinoma.  Mucoepidermoid carcinoma (formerly categorized as bronchial adenoma) is a rare tracheobronchial tumor of low malignant potential. It is of bronchial gland origin and composed of distinct areas of squamoid cells, mucus-secreting cells, and cells of intermediate type.335 Although patients vary widely in age at presentation, the tumors are more common in adults, with a peak incidence between 35 and 45 years.270,336 They typically occur in the main or lobar bronchi but in rare instances may be located in the trachea and periphery of the lung.336 They are usually slow-growing, low-grade neoplasms with a benign clinical course.326 Although

occasionally they exhibit aggressive local behavior, metastases are uncommon.122,326,336 Radiologically the tumor usually manifests as a central endobronchial mass and less commonly as a polypoid intraluminal tracheal or peripheral lung nodule or a mass. Adenoid cystic carcinoma (cylindroma).  Adenoid cystic carcinoma (formerly categorized as a bronchial adenoma) is an uncommon primary tumor of the lung. It occurs most often in the trachea (Fig. 37-37) and main bronchi, although 10% to 15% are located peripherally in the lung.10,27,53 Adenoid cystic carcinomas occur with equal frequency in men and women; the mean age at diagnosis is 45 to 51 years.186,198 Although review of the literature provides limited information regarding their biological behavior in lung, they typically exhibit slow progressive growth. Metastases to regional lymph nodes, lung, bone, liver, and brain are common but tend to occur late in the disease.9,27 Radiologically the tumor usually manifests as an endotracheal or endobronchial mass that is typically lobulated or polypoid and encroaches on the airway lumen. Masses can be circumferential and may manifest as diffuse stenosis.188 A less common manifestation is a peripheral lung nodule or mass.93

Mesenchymal Tumors Primary pulmonary sarcomas.  Primary lung sarcomas with a vascular origin (angiosarcomas, epithelioid hemangioendotheliomas) are rare primary tumors of lung.18,285,319 Most angiosarcomas are lung metastases, and the existence of primary pulmonary angiosarcoma has been questioned.277 The tumor usually occurs in young adults, and the most frequent radiologic finding is multiple bilateral nodules.285 Epithelioid hemangioendothelioma is typically seen in women younger than age 40 (range, 7-76 years).18,42,252,255,319 Most patients are asymptomatic at presentation; complaints include weight loss, dyspnea, chest pain, and cough. Epithelioid hemangioendotheliomas are usually indolent, with survival reported up to 24 years following resection.52 However, loss of weight, anemia, pulmonary symptoms, and hemorrhagic pleural effusions are significant factors of poor prognosis, with a median survival of less than 1 year.18 Pulmonary

1010

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 37-38  Epithelioid hemangioendothelioma appearing as multiple pulmonary nodules. A, Posteroanterior chest radiograph shows numerous small bilateral pulmonary nodules. Note the absence of hilar and mediastinal adenopathy. B, CT confirms the numerous small, bilateral, well-circumscribed nodules. C, CT scan of the abdomen reveals small, bilateral, low-attenuation hepatic epithelioid hemangioendotheliomas.

epithelioid hemangioendothelioma usually manifests radiologically as multiple 1- to 2-cm bilateral pulmonary nodules, although single nodules and unilateral distribution have been reported in approximately 25% of patients18,52,255 (Fig. 37-38). Irregular thickening of the bronchovascular bundles and perilobular structures due to lymphangitic spread, and associated multiple pulmonary nodules as well as multiple small peripheral nodules bilaterally have been described on HRCT as rare manifestations of pulmonary epithelioid hemangioendothelioma.262 Calcification is rarely detected but is common histologically.52,255 Hilar adenopathy and pleural effusions occur in 9% of patients.255 Multiorgan involvement, most commonly the liver, occurs occasionally and may represent metastatic disease or multicentric origin of the tumor.37,43,71,73,257 Primary lung sarcomas with a spindle origin are rare, accounting for fewer than 0.5% of primary lung malignancies.15,285 Spindle cell sarcomas (malignant fibrous histiocytoma, hemangiopericytoma, fibrosarcoma, leiomyosarcoma, synovial sarcoma) are the most common primary pulmonary sarcomas.145,197,285,338 They are a heterogeneous group of tumors with morphologic similarities to their extrathoracic counterparts.15 The diagnosis is established only after metastatic disease and sarcomalike primary lung malignancies (sarcomatoid carcinomas) have been excluded.15,87,138,208 Pulmonary sarcomas have a peak incidence in the fifth and sixth decades.87,139,197,285,337 The tumors are usually slowly growing with late metastases, and the prognosis is generally better than that with lung cancer.285

Radiologically, pulmonary sarcomas are more commonly located peripherally, although central and endobronchial masses are reported.15,139,197,285,337 Lesions range in size from 0.6 to 25 cm; they are typically sharply marginated and occasionally calcified15,117,145,197,285 (Figs. 37-39 and 37-40). Cavitation is uncommon, although CT can show heterogeneous attenuation resulting from necrosis within the mass.117,145 Diagnostic angiographic features have been reported for hemangiopericytomas in the soft tissues and bone.117,327 The pathognomonic hypervascularity that occurs as a result of the numerous vascular spaces within the tumor, however, are seldom seen on CT scans or MRIs of primary pulmonary hemangiopericytomas (Fig. 37-41).

LYMPHOPROLIFERATIVE DISORDERS Primary Lymphomas Primary pulmonary lymphomas account for fewer than 1% of all lymphomas.79,292 The diagnosis is generally considered if the lymphoid proliferation is monoclonal and there are no sites of extrathoracic lymphoma at presentation or for at least 3 months after diagnosis.210,292 However, criteria used to define primary pulmonary lymphomas are variable; some authors restrict the diagnosis to pulmonary parenchymal disease only, but others include hilar adenopathy with or without mediastinal adenopathy. Primary pulmonary lymphomas encompass a histologic spectrum of malignant lymphomas including non-Hodgkin’s lymphoma,

CHAPTER 37  Neoplastic Disease of the Lung

1011

A

M M

P

P L

B

L

C

FIG 37-39  Synovial carcinoma of left lung manifesting as a large mass. A, CT shows a large mass in the left lower lobe with a small focal calcification (arrow). A small pleural effusion can be seen. B and C, Coronal T1-weighted (B) and fast-spin echo (C) MRIs show a large heterogeneous mass abutting the diaphragm without signs of invasion. There is a small pleural effusion (P). L, liver; M; mass.

Hodgkin’s lymphoma, and lymphoproliferative disorders associated with immunodeficiency states (e.g., posttransplant lymphoproliferative disorders, acquired immunodeficiency syndrome).48,116 Primary pulmonary lymphomas are typically non-Hodgkin’s lymphomas of B-cell immunophenotype composed of monoclonal lymphocytic cells with low histologic grade, and they are classified as extranodal marginal zone lymphomas.81,158,175,210 Most primary extranodal lymphomas arise from mucosa-associated lymphoid tissue, and the Revised European-American Lymphoma classification recommends that those lymphomas with low-grade features be referred to as marginal zone

B-cell lymphoma of the mucosa-associated lymphoid tissue type, whereas those with high-grade features be referred to as diffuse large B-cell nonHodgkin’s lymphoma.210,211,316 Primary pulmonary lymphomas tend to remain localized to the lung, although recurrence after treatment (often at extrapulmonary sites) occurs in up to 50%.49,81,158,175 In the lung the neoplastic cells typically infiltrate the bronchiolar mucosal epithelium, forming lymphoepithelial lesions.210 High-grade non-Hodgkin’s lymphomas, which constitute approximately 13% of primary pulmonary lymphomas, are usually B-cell tumors with aggressive behavior and a poor 5-year

1012

PART II  CT and MR Imaging of the Whole Body

A

B FIG 37-40  Malignant fibrous histiocytoma manifesting as a large lung mass. A, Posteroanterior chest radiograph shows a left upper lobe mass. B, CT shows a large lobular mass and reveals narrowing of the left main pulmonary artery (arrowheads).

A

B

C FIG 37-41  Hemangiopericytoma appearing as a large hypervascular mass. A, Contrast-enhanced CT shows a large right lower lobe mass. Large vessels can be seen within the mass. B and C, Axial T1-weighted (B) and fast-spin echo (C) MRIs show a heterogeneous mass in the right hemithorax. Flow voids can be seen within the intratumoral vessels.

CHAPTER 37  Neoplastic Disease of the Lung

A

1013

B FIG 37-42  Primary pulmonary lymphoma manifesting as bilateral pulmonary opacities. A and B, CT shows focal poorly marginated nodular opacities in both lungs. Air bronchograms within opacities are suggestive of the diagnosis.

survival rate.49,81,158,175 Most patients are 55 to 60 years old, although the age range is wide.175 Patients with low-grade lymphomas are usually asymptomatic at presentation, whereas patients with high-grade tumors usually present with cough, fever, or weight loss.49 Radiologic manifestations include a solitary nodule or mass, multiple nodules or masses, focal or multifocal consolidation, reticulonodular opacities, and atelectasis49,81,153,171,217 (Fig. 37-42). Hilar adenopathy is rare, and pleural effusions occur in 7% to 25% of patients.49,158

BENIGN NEOPLASMS OF THE LUNG HAMARTOMA Hamartomas constitute 0.25% of all primary lung tumors; although uncommon they are the most common benign tumor of the lung.265 The term hamartoma was initially used to describe lesions with an abnormal composition or disorganized arrangement of normal lung tissue. However, hamartomas are now considered true neoplasms containing a mixture of epithelial and mesenchymal tissues.273 Pulmonary hamartomas typically occur in patients older than 30 years, with a peak incidence in the sixth decade (range, 0-76 years).118,218,235,265 Most patients are asymptomatic; symptoms are typically present with central endobronchial lesions and include hemoptysis, recurrent pneumonia, and dyspnea.88,99,265 Hamartomas are typically solitary, well-marginated, slightly lobular nodules or masses that are less than 4 cm in size (range, 1-30 cm)88,99,235,265 (Fig. 37-43). Most are located peripherally within the lung. Endobronchial hamartomas are less common (up to 20% of cases).88,118 Calcification has been reported in up to 50% but is probably present in fewer than 5% of hamartomas.168,273 Fat (CT attenuation, −40 to −120 HU) occurs in up to 50% of cases and is a diagnostic feature.273 Rare radiographic manifestations include cavitation and multiple pulmonary hamartomas.152

GRANULAR CELL TUMOR Granular cell tumors (formerly granular cell myeloblastomas) are uncommon benign mesenchymal neoplasms that are thought to arise from Schwann cells and usually present as a small solitary skin or submucosal nodule.61,187 Pulmonary granular cell tumors are rare and constitute 6% to 10% of all granular cell tumors.1 Patients are usually adults (peak incidence, age 30-50 years; range, 0-59 years), and there is a higher incidence in African Americans.61,157,226,227,250 Most pulmo-

FIG 37-43  Hamartoma manifesting as a large mass. CT shows a right lower mass that has focal calcified regions as well as focal regions of low attenuation (attenuation, −60 HU) consistent with fat. The diagnosis is suggested by small areas of fat within the mass.

nary granular cell tumors manifest as small, central endobronchial masses (range, 0.3-6.5 cm).1,61,128 Multiple lesions, typically in the larger bronchi, are found in up to 25% of cases.61,128 Common radiologic findings include atelectasis and obstructive pneumonia, although slow-growing solitary pulmonary nodules or masses occur in 12% of cases.61,128,196

SCLEROSING HEMANGIOMA Primary pulmonary sclerosing hemangioma (formerly pneumocytoma, papillary pneumocytoma) is a benign tumor consisting of thin-walled vessels and connective tissue.67,214 Originally thought to represent a vascular tumor, it is now considered to arise from primitive respiratory epithelium.67,121 Most patients are asymptomatic women between 30 and 50 years of age (range, 15-77 years).146,284 However, cough and hemoptysis can occur.67 Radiologically, pulmonary hemangiomas usually manifest as peripheral, well-marginated, solitary pulmonary

1014

PART II  CT and MR Imaging of the Whole Body

FIG 37-44  Clear cell tumor manifesting as a solitary pulmonary nodule. CT shows a well-circumscribed left upper lobe nodule with heterogeneous enhancement.

nodules or masses 0.4 to 8 cm in diameter (average, 3 cm).113,209 Multiple pulmonary nodules and calcification are rare.146,172 Marked contrast enhancement is typical on CT, and tumors typically are heterogeneous in attenuation.209 On pathologic examination this is due to the angiomatous, solid and sclerotic, and cystic composition of the tumor. The MRI appearance of the tumor is variable and has been reported to be both homogeneous and heterogeneous in signal intensity.89,113 However, MRI can be useful in suggesting the diagnosis when regions of different signal intensity are present in the tumor.89 On T1-weighted images, areas of high signal intensity correspond to regions of abundant clear cells, whereas on T2-weighted images, areas of low signal intensity correspond to fibrotic or hemorrhagic regions within the tumor. The areas of high signal intensity on T2-weighted images correspond to the hemangiomatous portions of the tumor. There is also variable contrast enhancement, with the areas demonstrating contrast enhancement corresponding to the hemangiomatous portions of the tumor.89,209

CLEAR CELL TUMOR Clear cell tumor is a rare neoplasm of the lung.91 Most patients are asymptomatic and between 50 and 60 years of age (range, 8-70 years).91,105 Almost all these tumors behave in a benign nature, although extrathoracic metastases have been reported.264 Clear cell tumors typically manifest radiologically as well-marginated solitary pulmonary nodules usually less than 3 cm in diameter (range, 0.7-12 cm)13,90,91,147,264,339 (Fig. 37-44).

REFERENCES 1. Abdulhamid I, Rabah R: Granular cell tumor of the bronchus. Pediatr Pulmonol 30:425–428, 2000. 2. Aberle DR, Berg CD, Black WC, et al: The National Lung Screening Trial: Overview and study design. Radiology 258:243–253, 2011. 3. Abrams J, Doyle LA, Aisner J: Staging, prognostic factors, and special considerations in small cell lung cancer. Semin Oncol 15:261–277, 1988. 4. Adluri RK, Boddu SR, Martin-Ucar A, et al: Pulmonary blastoma—A rare tumor with variable presentation. Eur J Cardiothorac Surg 29:236–239, 2006. 5. Aguayo SM, Miller YE, Waldron JA, Jr, et al: Brief report: Idiopathic diffuse hyperplasia of pulmonary neuroendocrine cells and airways disease. N Engl J Med 327:1285–1288, 1992.

6. Ak I, Sivrikoz MC, Entok E, et al: Discordant findings in patients with non-small-cell lung cancer: Absolutely normal bone scans versus disseminated bone metastases on positron-emission tomography/ computed tomography. Eur J Cardiothorac Surg 37:792–796, 2010. 7. Akata S, Kajiwara N, Park J, et al: Evaluation of chest wall invasion by lung cancer using respiratory dynamic MRI. J Med Imaging Radiat Oncol 52:36–39, 2008. 8. Alberg AJ, Brock MV, Samet JM: Epidemiology of lung cancer: Looking to the future. J Clin Oncol 23:3175–3185, 2005. 9. Albers E, Lawrie T, Harrell JH, et al: Tracheobronchial adenoid cystic carcinoma: A clinicopathologic study of 14 cases. Chest 125:1160–1165, 2004. 10. Allen MS: Malignant tracheal tumors. Mayo Clin Proc 68:680–684, 1993. 11. American Joint Committee on Cancer: AJCC cancer staging manual, ed 7, New York, NY, 2009, Springer. 12. Amos CI, Xu W, Spitz MR: Is there a genetic basis for lung cancer susceptibility? Recent Results Cancer Res 151:3–12, 1999. 13. Andrion A, Mazzucco G, Gugliotta P, et al: Benign clear cell (sugar) tumor of the lung. A light microscopic, histochemical, and ultrastructural study with a review of the literature. Cancer 56:2657– 2663, 1985. 14. Antoch G, Stattaus J, Nemat AT, et al: Non-small cell lung cancer: Dual-modality PET/CT in preoperative staging. Radiology 229:526–533, 2003. 15. Attanoos RL, Appleton MA, Gibbs AR: Primary sarcomas of the lung: A clinicopathological and immunohistochemical study of 14 cases. Histopathology 29:29–36, 1996. 16. Austin JH, Garg K, Aberle D, et al: Radiologic implications of the 2011 classification of adenocarcinoma of the lung. Radiology 266:62–71, 2013. 17. Azad A, Chionh F, Scott AM, et al: High impact of 18F-FDG-PET on management and prognostic stratification of newly diagnosed small cell lung cancer. Mol Imaging Biol 12:443–451, 2010. 18. Bagan P, Hassan M, Le Pimpec Barthes F, et al: Prognostic factors and surgical indications of pulmonary epithelioid hemangioendothelioma: A review of the literature. Ann Thorac Surg 82:2010–2013, 2006. 19. Bailey-Wilson JE, Amos CI, Pinney SM, et al: A major lung cancer susceptibility locus maps to chromosome 6q23-25. Am J Hum Genet 75:460–474, 2004. 20. Baltar VT, Xun WW, Chuang SC, et al: Smoking, secondhand smoke, and cotinine levels in a subset of EPIC cohort. Cancer Epidemiol Biomarkers Prev 20:869–875, 2011. 21. Barbone F, Bovenzi M, Cavallieri F, et al: Cigarette smoking and histologic type of lung cancer in men. Chest 112:1474–1479, 1997. 22. Beasley MB, Brambilla E, Travis WD: The 2004 World Health Organization classification of lung tumors. Semin Roentgenol 40:90–97, 2005. 23. Beasley MTF, Haselton P, et al: Carcinoid tumour. In Travis WDBE, Muller-Hermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 59–62. 24. Beasley MB, Thunnissen FB, Brambilla E, et al: Pulmonary atypical carcinoid: Predictors of survival in 106 cases. Hum Pathol 31:1255– 1265, 2000. 25. Benson RE, Rosado-de-Christenson ML, Martinez-Jimenez S, et al: Spectrum of pulmonary neuroendocrine proliferations and neoplasms. Radiographics 33:1631–1649, 2013. 26. Berman DW, Crump KS: A meta-analysis of asbestos-related cancer risk that addresses fiber size and mineral type. Crit Rev Toxicol 38(Suppl 1): 49–73, 2008. 27. Bhattacharyya T, Bahl A, Kapoor R, et al: Primary adenoid cystic carcinoma of lung: A case report and review of the literature. J Cancer Res Ther 9:302–304, 2013. 28. Birim O, Kappetein AP, Stijnen T, et al: Meta-analysis of positron emission tomographic and computed tomographic imaging in detecting mediastinal lymph node metastases in nonsmall cell lung cancer. Ann Thorac Surg 79:375–382, 2005.

CHAPTER 37  Neoplastic Disease of the Lung 29. Blaukovitsch M, Halbwedl I, Kothmaier H, et al: Sarcomatoid carcinomas of the lung—Are these histogenetically heterogeneous tumors? Virchows Arch 449:455–461, 2006. 30. Bobbio A, Carbognani P, Grapeggia M, et al: Surgical outcome of combined pulmonary and atrial resection for lung cancer. Thorac Cardiovasc Surg 52:180–182, 2004. 31. Boffetta P, Fordyce T, Mandel JS: A mortality study of workers exposed to insoluble forms of beryllium. Eur J Cancer Prev 2014. 32. Boland GW, Dwamena BA, Jagtiani Sangwaiya M, et al: Characterization of adrenal masses by using FDG PET: A systematic review and meta-analysis of diagnostic test performance. Radiology 259:117–126, 2011. 33. Boland GW, Hahn PF, Pena C, et al: Adrenal masses: Characterization with delayed contrast-enhanced CT. Radiology 202:693–696, 1997. 34. Boland GW, Lee MJ: Magnetic resonance imaging of the adrenal gland. Crit Rev Diagn Imaging 36:115–174, 1995. 35. Boland GW, Lee MJ, Gazelle GS, et al: Characterization of adrenal masses using unenhanced CT: An analysis of the CT literature. AJR 171:201–204, 1998. 36. Boldrini R, Devito R, Diomedi-Camassei F, et al: Pulmonary blastomas of childhood: Histologic, immunohistochemical, ultrastructural aspects and therapeutic considerations. Ultrastruct Pathol 29:493–501, 2005. 37. Bollinger BK, Laskin WB, Knight CB: Epithelioid hemangioendothelioma with multiple site involvement. Literature review and observations. Cancer 73:610–615, 1994. 38. Bruzzi JF, Komaki R, Walsh GL, et al: Imaging of non-small cell lung cancer of the superior sulcus: Part 1: Anatomy, clinical manifestations, and management. Radiographics 28:551–560, quiz 620, 2008. 39. Bruzzi JF, Komaki R, Walsh GL, et al: Imaging of non-small cell lung cancer of the superior sulcus: Part 2: Initial staging and assessment of resectability and therapeutic response. Radiographics 28:561–572, 2008. 40. Cardillo G, Sera F, Di Martino M, et al: Bronchial carcinoid tumors: Nodal status and long-term survival after resection. Ann Thorac Surg 77:1781–1785, 2004. 41. Carey FA, Wallace WA, Fergusson RJ, et al: Alveolar atypical hyperplasia in association with primary pulmonary adenocarcinoma: A clinicopathological study of 10 cases. Thorax 47:1041–1043, 1992. 42. Carter EJ, Bradburne RM, Jhung JW, et al: Alveolar hemorrhage   with epithelioid hemangioendothelioma. A previously unreported manifestation of a rare tumor. Am Rev Respir Dis 142:700–701, 1990. 43. Celikel C, Yumuk PF, Basaran G, et al: Epithelioid hemangioendothelioma with multiple organ involvement. APMIS 115:881–888, 2007. 44. Chen J: Estimated risks of radon-induced lung cancer for different exposure profiles based on the new EPA model. Health Phys 88:323– 333, 2005. 45. Chen F, Bina WF, Cole P: Declining incidence rate of lung adenocarcinoma in the United States. Chest 131:1000–1005, 2007. 46. Chen F, Takahashi A, Omasa M, et al: En bloc total vertebrectomy for lung cancer invading the spine. Lung Cancer 61:137–139, 2008. 47. Chong S, Lee KS, Kim BT, et al: Integrated PET/CT of pulmonary neuroendocrine tumors: Diagnostic and prognostic implications. AJR Am J Roentgenol 188:1223–1231, 2007. 48. Cooksley N, Judge DJ, Brown J: Primary pulmonary Hodgkin’s lymphoma and a review of the literature since 2006. BMJ Case Rep 2014:2014. 49. Cordier JF, Chailleux E, Lauque D, et al: Primary pulmonary lymphomas: A clinical study of 70 cases of nonimmunocompromised patients. Chest 103:201–208, 1994. 50. Costa DB, Kobayashi S, Tenen DG, et al: Pooled analysis of the prospective trials of gefitinib monotherapy for EGFR-mutant non-small cell lung cancers. Lung Cancer 58:95–103, 2007. 51. Cutler CS, Michel RP, Yassa M, et al: Pulmonary blastoma: Case report of a patient with a 7-year remission and review of chemotherapy experience in the world literature. Cancer 82:462–467, 1998. 52. Dail DH, Liebow AA, Gmelich JT, et al: Intravascular, bronchiolar, and alveolar tumor of the lung (IVBAT). An analysis of twenty cases of a peculiar sclerosing endothelial tumor. Cancer 51:452–464, 1983.

1015

53. Dalton ML, Gatling RR: Peripheral adenoid cystic carcinoma of the lung. South Med J 83:577–579, 1990. 54. Darby S, Hill D, Auvinen A, et al: Radon in homes and risk of lung cancer: Collaborative analysis of individual data from 13 European case-control studies. BMJ 330:223, 2005. 55. Darling GE: Staging of the patient with small cell lung cancer. Chest Surg Clin N Am 7:81–94, 1997. 56. Darnell RB, Posner JB: Paraneoplastic syndromes affecting the nervous system. Semin Oncol 33:270–298, 2006. 57. Davies SJ, Gosney JR, Hansell DM, et al: Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia: An under-recognised spectrum of disease. Thorax 62:248–252, 2007. 58. de Hoop B, Gietema H, van de Vorst S, et al: Pulmonary ground-glass nodules: Increase in mass as an early indicator of growth. Radiology 255:199–206, 2010. 59. de Langen AJ, Raijmakers P, Riphagen I, et al: The size of mediastinal lymph nodes and its relation with metastatic involvement: A metaanalysis. Eur J Cardiothorac Surg 29:26–29, 2006. 60. de Perrot M, Fadel E, Mercier O, et al: Long-term results after carinal resection for carcinoma: Does the benefit warrant the risk? J Thorac Cardiovasc Surg 131:81–89, 2006. 61. Deavers M, Guinee D, Koss MN, et al: Granular cell tumors of the   lung. Clinicopathologic study of 20 cases. Am J Surg Pathol 19:627–635, 1995. 62. den Bakker MA, Thunnissen FB: Neuroendocrine tumours—Challenges in the diagnosis and classification of pulmonary neuroendocrine tumours. J Clin Pathol 66:862–869, 2013. 63. Deslauriers J, Tronc F, Fortin D: Management of tumors involving the chest wall including Pancoast tumors and tumors invading the spine. Thorac Surg Clin 23:313–325, 2013. 64. Detterbeck FC: Overdiagnosis during lung cancer screening: Is it an overemphasised, underappreciated, or tangential issue? Thorax 69:407–408, 2014. 65. Detterbeck FC, Postmus PE, Tanoue LT: The stage classification of lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 143:e191S–210S, 2013. 66. Devesa SS, Bray F, Vizcaino AP, et al: International lung cancer trends by histologic type: Male:female differences diminishing and adenocarcinoma rates rising. Int J Cancer 117:294–299, 2005. 67. Devouassoux-Shisheboran MNA, Leslie K, Niho S: Sclerosing haemangioma. In Travis WDBE, Muller-Hermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 115–117. 68. Diederich S, Wormanns D, Semik M, et al: Screening for early lung cancer with low-dose spiral CT: Prevalence in 817 asymptomatic smokers. Radiology 222:773–781, 2002. 69. Doll R, Hill AB: Smoking and carcinoma of the lung: Preliminary report. Br Med J 2:739–748, 1950. 70. Earnest Ft, Ryu JH, Miller GM, et al: Suspected non-small cell lung cancer: Incidence of occult brain and skeletal metastases and effectiveness of imaging for detection—Pilot study. Radiology 211:137–145, 1999. 71. Echevarria RA: Angiogenic nature of ‘intravascular bronchioloalveolar tumor. Arch Pathol Lab Med 105:627–628, 1981. 72. Elias AD: Small cell lung cancer: State-of-the-art therapy in 1996. Chest 112:251S–258S, 1997. 73. Emery RW, Fox AL, Raab DE: Intravascular bronchioloalveolar tumour. Thorax 37:472–473, 1982. 74. Erasmus JJ, McAdams HP, Patz JEF, et al: Evaluation of primary pulmonary carcinoid tumors using FDG PET. Am J Roentgenol 170:1369–1373, 1998. 75. Erren TC, Jacobsen M, Piekarski C: Synergy between asbestos and smoking on lung cancer risks. Epidemiology 10:405–411, 1999. 76. Fadel E, Missenard G, Court C, et al: Long-term outcomes of en bloc resection of non-small cell lung cancer invading the thoracic inlet and spine. Ann Thorac Surg 92:1024–1030, discussion 30, 2011.

1016

PART II  CT and MR Imaging of the Whole Body

77. Fan L, Liu SY, Li QC, et al: Multidetector CT features of pulmonary focal ground-glass opacity: Differences between benign and malignant. Br J Radiol 85:897–904, 2012. 78. Feinstein AR, Wells CK: A clinical-severity staging system for patients with lung cancer. Medicine (Baltimore) 69:1–33, 1990. 79. Ferraro P, Trastek VF, Adlakha H, et al: Primary non-Hodgkin’s lymphoma of the lung. Ann Thorac Surg 69:993–997, 2000. 80. Fessler MB, Cool CD, Miller YE, et al: Idiopathic diffuse hyperplasia   of pulmonary neuroendocrine cells in a patient with acromegaly. Respirology 9:274–277, 2004. 81. Fiche M, Caprons F, Berger F, et al: Primary pulmonary non-Hodgkin’s lymphomas. Histopathology 26:529–537, 1995. 82. Filderman AE, Shaw C, Matthay RA: Lung cancer. Part I: Etiology, pathology, natural history, manifestations, and diagnostic techniques. Invest Radiol 21:80–90, 1986. 83. Fischer BM, Lassen U, Hojgaard L: PET-CT in preoperative staging of lung cancer. N Engl J Med 364:980–981, 2011. 84. Fischer BM, Mortensen J, Langer SW, et al: A prospective study of PET/ CT in initial staging of small-cell lung cancer: Comparison with CT, bone scintigraphy and bone marrow analysis. Ann Oncol 18:338–345, 2007. 85. Fox DJ, Khattar RS: Carcinoid heart disease: Presentation, diagnosis, and management. Heart 90:1224–1228, 2004. 86. Franceschi S, Bidoli E: The epidemiology of lung cancer. Ann Oncol 10(Suppl 5):S3–S6, 1999. 87. Francis D, Jacobsen M: Pulmonary blastoma. Curr Top Pathol 73:165–294, 1998. 88. Fudge TL, Ochsner JL, Mills NL: Clinical spectrum of pulmonary hamartomas. Ann Thorac Surg 30:36–39, 1980. 89. Fujiyoshi F, Ichinari N, Fukukura Y, et al: Sclerosing hemangioma of the lung: MR findings and correlation with pathological features. J Comput Assist Tomogr 22:1006–1008, 1998. 90. Gaffey MJ, Mills SE, Askin FB, et al: Clear cell tumor of the lung. A clinicopathologic, immunohistochemical, and ultrastructural study of eight cases. Am J Surg Pathol 14:248–259, 1990. 91. Gaffey MJ, Mills SE, Ritter JH: Clear cell tumors of the lower respiratory tract. Semin Diagn Pathol 14:222–232, 1997. 92. Galetta D, Cesario A, Margaritora S, et al: Enduring challenge in the treatment of nonsmall cell lung cancer with clinical stage IIIB: Results of a trimodality approach. Ann Thorac Surg 76:1802–1808, discussion 8–9, 2003. 93. Gallagher CG, Stark R, Teskey J, et al: Atypical manifestations of pulmonary adenoid cystic carcinoma. Br J Dis Chest 80:396–399, 1986. 94. Gandhi L, Johnson BE: Paraneoplastic syndromes associated with small cell lung cancer. J Natl Compr Canc Netw 4:631–638, 2006. 95. Garcia-Yuste M, Matilla JM, Cueto A, et al: Typical and atypical carcinoid tumours: Analysis of the experience of the Spanish Multicentric Study of Neuroendocrine Tumours of the Lung. Eur J Cardiothorac Surg 31:192–197, 2007. 96. Ge Y, Eltorky MA, Ernst RD, et al: Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia. Ann Diagn Pathol 11:122–126, 2007. 97. Ginsberg RJKM, Armstrong JG: Cancer of the Lung. In DeVita VTHS, Rosenberg SA, editors: Cancer: Principles and practice of oncology, ed 4, Philadelphia, 1994, Lippincott, pp 673–722. 98. Ginsberg MS, Griff SK, Go BD, et al: Pulmonary nodules resected at video-assisted thoracoscopic surgery: Etiology in 426 patients. Radiology 213:277–282, 1999. 99. Gjevre JA, Myers JL, Prakash UB: Pulmonary hamartomas. Mayo Clin Proc 71:14–20, 1996. 100. Godoy MC, Naidich DP: Subsolid pulmonary nodules and the spectrum of peripheral adenocarcinomas of the lung: Recommended interim guidelines for assessment and management. Radiology 253:606–622, 2009. 101. Godoy MC, Naidich DP: Overview and strategic management of subsolid pulmonary nodules. J Thorac Imaging 27:240–248, 2012. 102. Godoy MC, Sabloff B, Naidich DP: Subsolid pulmonary nodules: Imaging evaluation and strategic management. Curr Opin Pulm Med 18:304–312, 2012.

103. Godoy MC, Truong MT, Sabloff B, et al: Subsolid pulmonary nodule management and lung adenocarcinoma classification: State of the art and future trends. Semin Roentgenol 48:295–307, 2013. 104. Goodman M, Morgan RW, Ray R, et al: Cancer in asbestos-exposed occupational cohorts: A meta-analysis. Cancer Causes Control 10:453–465, 1999. 105. Gora-Gebka M, Liberek A, Bako W, et al: The “sugar” clear cell tumor of the lung-clinical presentation and diagnostic difficulties of an unusual lung tumor in youth. J Pediatr Surg 41:e27–e29, 2006. 106. Gould MK, Donington J, Lynch WR, et al: Evaluation of individuals with pulmonary nodules: When is it lung cancer? Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 143:e93S– 120S, 2013. 107. Gould MK, Kuschner WG, Rydzak CE, et al: Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: A meta-analysis. Ann Intern Med 139:879–892, 2003. 108. Gould MK, Maclean CC, Kuschner WG, et al: Accuracy of positron emission tomography for diagnosis of pulmonary nodules and mass lesions: A meta-analysis. JAMA 285:914–924, 2001. 109. Govindan R, Ihde DC: Practical issues in the management of the patient with small cell lung cancer. Chest Surg Clin N Am 7:167–181, 1997. 110. Govindan R, Page N, Morgensztern D, et al: Changing epidemiology   of small-cell lung cancer in the United States over the last 30 years: Analysis of the surveillance, epidemiologic, and end results database.   J Clin Oncol 24:4539–4544, 2006. 111. Grippi MA: Clinical aspects of lung cancer. Semin Roentgenol 25:12–24, 1990. 112. Grosney JRT: Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia. In Travis WDBE, Muller-Hermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 76–77. 113. Guibaud L, Pracros JP, Rode V, et al: Sclerosing hemangioma of the lung: Radiological findings and pathological diagnosis. Pediatr Radiol 25(Suppl 1):S207–S208, 1995. 114. Gurney JW, Lyddon DM, McKay JA: Determining the likelihood of malignancy in solitary pulmonary nodules with bayesian analysis. Part II. Application. Radiology 186:415–422, 1993. 115. Gustafsson BI, Kidd M, Chan A, et al: Bronchopulmonary neuroendocrine tumors. Cancer 113:5–21, 2008. 116. Habermann TM, Ryu JH, Inwards DJ, et al: Primary pulmonary lymphoma. Semin Oncol 26:307–315, 1999. 117. Halle M, Blum U, Dinkel E, et al: CT and MR features of primary pulmonary hemangiopericytomas. J Comput Assist Tomogr 17:51–55, 1993. 118. Hansen CP, Holtveg H, Francis D, et al: Pulmonary hamartoma.   J Thorac Cardiovasc Surg 104:674–678, 1992. 119. Hasegawa M, Sone S, Takashima S, et al: Growth rate of small lung cancers detected on mass CT screening. Br J Radiol 73:1252–1259, 2000. 120. Hashimoto Y, Tsujikawa T, Kondo C, et al: Accuracy of PET for diagnosis of solid pulmonary lesions with 18F-FDG uptake below   the standardized uptake value of 2.5. J Nucl Med 47:426–431, 2006. 121. Heikkila P, Salminen US: Papillary pneumocytoma of the lung: An immunohistochemical and electron microscopic study. Pathol Res Pract 190:194, 1993. 122. Heitmiller RF, Mathisen DJ, Ferry JA, et al: Mucoepidermoid lung tumors. Ann Thorac Surg 47:394–399, 1989. 123. Henschke CI, McCauley DI, Yankelevitz DF, et al: Early lung cancer action project: Overall design and findings from baseline screening. Lancet 354:99–105, 1999. 124. Henschke CI, Yankelevitz DF, Miettinen OS, et al: Computed tomographic screening for lung cancer: The relationship of disease stage to tumor size. Arch Intern Med 166:321–325, 2006.

CHAPTER 37  Neoplastic Disease of the Lung 125. Henschke CI, Yankelevitz DF, Mirtcheva R, et al: CT screening for lung cancer: Frequency and significance of part-solid and nonsolid nodules. AJR Am J Roentgenol 178:1053–1057, 2002. 126. Herder GJ, Golding RP, Hoekstra OS, et al: The performance of (18) F-fluorodeoxyglucose positron emission tomography in small solitary pulmonary nodules. Eur J Nucl Med Mol Imaging 31:1231–1236, 2004. 127. Herman SJ, Winton TL, Weisbrod GL, et al: Mediastinal invasion by bronchogenic carcinoma: CT signs. Radiology 190:841–846, 1994. 128. Hernandez OG, Haponik EF, Summer WR: Granular cell tumour of the bronchus: Bronchoscopic and clinical features. Thorax 41:927–931, 1986. 129. Higashi K, Ueda Y, Seki H, et al: Fluorine-18-FDG PET imaging is negative in bronchioloalveolar lung carcinoma. J Nucl Med 39:1016– 1020, 1998. 130. Hillman BJ, Schnall MD: American College of Radiology Imaging Network: Future clinical trials. Radiology 227:631–632, 2003. 131. Hodgson JT, Darnton A: The quantitative risks of mesothelioma and lung cancer in relation to asbestos exposure. Ann Occup Hyg 44:565– 601, 2000. 132. Honda T, Kondo T, Murakami S, et al: Radiographic and pathological analysis of small lung adenocarcinoma using the new IASLC classification. Clin Radiol 68:e21–e26, 2013. 133. Hooper RG, Tenholder MF, Underwood GH, et al: Computed tomographic scanning of the brain in initial staging of bronchogenic carcinoma. Chest 85:774–776, 1984. 134. Hung RJ, McKay JD, Gaborieau V, et al: A susceptibility locus for lung cancer maps to nicotinic acetylcholine receptor subunit genes on 15q25. Nature 452:633–637, 2008. 135. Ichinose Y, Fukuyama Y, Asoh H, et al: Induction chemoradiotherapy and surgical resection for selected stage IIIB non-small-cell lung cancer. Ann Thorac Surg 76:1810–1814, discussion 5, 2003. 136. Infante M, Cavuto S, Lutman FR, et al: A randomized study of lung cancer screening with spiral computed tomography: Three-year results from the DANTE trial. Am J Respir Crit Care Med 180:445–453, 2009. 137. International Agency for Research on Cancer: Tobacco smoke and involuntary smoking. IARC Monograph 83, Lyon, France, 2004, International Agency for Research on Cancer. 138. Ishida T, Tateishi M, Kaneko S, et al: Carcinosarcoma and spindle cell carcinoma of the lung. Clinicopathologic and immunohistochemical studies. J Thorac Cardiovasc Surg 100:844–852, 1990. 139. Janssen JP, Mulder JJ, Wagenaar SS, et al: Primary sarcoma of the lung: A clinical study with long-term follow-up. Ann Thorac Surg 58:1151– 1155, 1994. 140. Jelinek JS, Redmond J, 3rd, Perry JJ, et al: Small cell lung cancer: Staging with MR imaging. Radiology 177:837–842, 1990. 141. Jennings SG, Winer-Muram HT, Tann M, et al: Distribution of stage I lung cancer growth rates determined with serial volumetric CT measurements. Radiology 241:554–563, 2006. 142. Kakinuma R, Ohmatsu H, Kaneko M, et al: Progression of focal   pure ground-glass opacity detected by low-dose helical computed tomography screening for lung cancer. J Comput Assist Tomogr 28:17–23, 2004. 143. Kalemkerian GP, Gadgeel SM: Modern staging of small cell lung cancer. J Natl Compr Canc Netw 11:99–104, 2013. 144. Kaneko M, Eguchi K, Ohmatsu H, et al: Peripheral lung cancer: Screening and detection with low-dose spiral CT versus radiography. Radiology 201:798–802, 1996. 145. Katz DS, Lane MJ, Leung AN, et al: Primary malignant pulmonary hemangiopericytoma. Clin Imaging 22:192–195, 1998. 146. Katzenstein AL, Gmelich JT, Carrington CB: Sclerosing hemangioma   of the lung: A clinicopathologic study of 51 cases. Am J Surg Pathol 4:343–356, 1980. 147. Kavunkal AM, Pandiyan MS, Philip MA, et al: Large clear cell tumor of the lung mimicking malignant behavior. Ann Thorac Surg 83:310–312, 2007. 148. Kawakami S, Sone S, Takashima S, et al: Atypical adenomatous hyperplasia of the lung: Correlation between high-resolution CT findings and histopathologic features. Eur Radiol 11:811–814, 2001.

1017

149. Kerr KMFAE, Pugatch B, et al: Atypical adenomatous hyperplasia. In Travis WDBE, Muller-Hermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 73–75. 150. Khuder SA: Effect of cigarette smoking on major histological types of lung cancer: A meta-analysis. Lung Cancer 31:139–148, 2001. 151. Kim K, Flint JDA, Müller NL: Pulmonary carcinosarcoma: Radiologic and pathologic findings in three patients. Am J Roentgenol 169:691–694, 1997. 152. King TE, Jr, Christopher KL, Schwarz MI: Multiple pulmonary chondromatous hamartomas. Hum Pathol 13:496–497, 1982. 153. King LJ, Padley SP, Wotherspoon AC, et al: Pulmonary MALT lymphoma: Imaging findings in 24 cases. Eur Radiol 10:1932–1938, 2000. 154. Kishi K, Homma S, Kurosaki A, et al: Small lung tumors with the size of 1cm or less in diameter: Clinical, radiological, and histopathological characteristics. Lung Cancer 44:43–51, 2004. 155. Klein JS, Webb WR: The radiologic staging of lung cancer. J Thorac Imaging 7:29–47, 1991. 156. Kondo T, Yamada K, Noda K, et al: Radiologic-prognostic correlation   in patients with small pulmonary adenocarcinomas. Lung Cancer 36:49–57, 2002. 157. Korompai FL, Awe RJ, Beall AC, et al: Granular cell myoblastoma of   the bronchus: A new case, 12-year followup report, and review of the literature. Chest 66:578–580, 1974. 158. Koss MN: Pulmonary lymphoid disorders. Semin Diagn Pathol 12:158–171, 1995. 159. Koss MN, Hochholzer L, O’Leary T: Pulmonary blastomas. Cancer 67:2368–2381, 1991. 160. Kozower BD, Larner JM, Detterbeck FC, et al: Special treatment issues in non-small cell lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 143:e369S–e399S, 2013. 161. Kozower BD, Meyers BF, Reed CE, et al: Does positron emission tomography prevent nontherapeutic pulmonary resections for clinical stage IA lung cancer? Ann Thorac Surg 85:1166–1169, discussion 9–70, 2008. 162. Kut V, Spies W, Spies S, et al: Staging and monitoring of small cell lung cancer using [18F]fluoro-2-deoxy-D-glucose-positron emission tomography (FDG-PET). Am J Clin Oncol 30:45–50, 2007. 163. Lam WK, Watkins DN: Lung cancer: Future directions. Respirology 12:471–477, 2007. 164. Langer CJ: Roles of EGFR and KRAS Mutations in the treatment of patients with non-small-cell lung cancer. P T 36:263–279, 2011. 165. Lardinois D, Weder W, Hany TF, et al: Staging of non-small-cell lung cancer with integrated positron-emission tomography and computed tomography. N Engl J Med 348:2500–2507, 2003. 166. Larsen H, Sorensen JB: Pulmonary blastoma: A review with special emphasis on prognosis and treatment. Cancer Treat Rev 22:145–160, 1996. 167. Le Calvez F, Mukeria A, Hunt JD, et al: TP53 and KRAS mutation load and types in lung cancers in relation to tobacco smoke: Distinct patterns in never, former, and current smokers. Cancer Res 65:5076– 5083, 2005. 168. Ledor K, Fish B, Chaise L, et al: CT diagnosis of pulmonary hamartomas. CT. J Comput Tomogr 5:343–344, 1981. 169. Lee JS, Brown KK, Cool C, et al: Diffuse pulmonary neuroendocrine cell hyperplasia: Radiologic and clinical features. J Comput Assist Tomogr 26:180–184, 2002. 170. Lee HJ, Goo JM, Lee CH, et al: Nodular ground-glass opacities on thin-section CT: Size change during follow-up and pathological results. Korean J Radiol 8:22–31, 2007. 171. Lee KS, Kim Y, Primack SL: Imaging of pulmonary lymphomas. Am J Roentgenol 168:339–345, 1997. 172. Lee ST, Lee YC, Hsu CY, et al: Bilateral multiple sclerosing hemangiomas of the lung. Chest 101:572–573, 1992. 173. LeMense GP, Reed CE, Silvestri GA: Pulmonary blastoma: A rare lung malignancy. Lung Cancer 15:233–237, 1996.

1018

PART II  CT and MR Imaging of the Whole Body

174. Levin KH: Paraneoplastic neuromuscular syndromes. Neurol Clin 15:597–614, 1997. 175. Li G, Hansmann M-L, Zwingers T, et al: Primary lymphoma of the lung: Morphological, immunohistochemical and clinical features. Histopathology 16:519–531, 1990. 176. Lillington GA, Caskey CI: Evaluation and management of solitary and multiple pulmonary nodules. Clin Chest Med 14:111–119, 1993. 177. Liman ST, Altinok T, Topcu S, et al: Survival of biphasic pulmonary blastoma. Respir Med 100:1174–1179, 2006. 178. List AF, Doll DC, Greco FA: Lung cancer in Hodgkin’s disease: Association with previous radiotherapy. J Clin Oncol 3:215–221, 1985. 179. Liu N, Ma L, Zhou W, et al: Bone metastasis in patients with non-small cell lung cancer: The diagnostic role of F-18 FDG PET/CT. Eur J Radiol 74:231–235, 2010. 180. Lubin JH, Alavanja MC, Caporaso N, et al: Cigarette smoking and cancer risk: Modeling total exposure and intensity. Am J Epidemiol 166:479–489, 2007. 181. MacMahon H, Austin JH, Gamsu G, et al: Guidelines for management of small pulmonary nodules detected on CT scans: A statement from the Fleischner Society. Radiology 237:395–400, 2005. 182. MacManus MP, Hicks RJ, Matthews JP, et al: High rate of detection of unsuspected distant metastases by PET in apparent stage III non-smallcell lung cancer: Implications for radical radiation therapy. Int J Radiat Oncol Biol Phys 50:287–293, 2001. 183. Mahoney MC, Shipley RT, Corcoran HL, et al: CT demonstration of calcification in carcinoma of the lung. AJR Am J Roentgenol 154:255– 258, 1990. 184. Matsuoka H, Nishio W, Okada M, et al: Resection of chest wall invasion in patients with non-small cell lung cancer. Eur J Cardiothorac Surg 26:1200–1204, 2004. 185. Mayo-Smith WW, Boland GW, Noto RB, et al: State-of-the-art adrenal imaging. Radiographics 21:995–1012, 2001. 186. Maziak DE, Todd TR, Keshavjee SH, et al: Adenoid cystic carcinoma of the airway: Thirty-two-year experience. J Thorac Cardiovasc Surg 112:1522–1531, discussion 31–32, 1996. 187. Mazur MT, Shultz JJ, Myers JL: Granular cell tumor. Immunohistochemical analysis of 21 benign tumors and one malignant tumor. Arch Pathol Lab Med 114:692–696, 1990. 188. McCarthy MJ, Rosado-de-Christenson ML: Tumors of the trachea.   J Thorac Imaging 10:180–198, 1995. 189. Meisinger QC, Klein JS, Butnor KJ, et al: CT features of peripheral pulmonary carcinoid tumors. AJR Am J Roentgenol 197:1073–1080, 2011. 190. Meza R, Ten Haaf K, Kong CY, et al: Comparative analysis of 5 lung cancer natural history and screening models that reproduce outcomes of the NLST and PLCO trials. Cancer 2014. 191. Miller RR: Bronchioloalveolar cell adenomas. Am J Surg Pathol 14:904–912, 1990. 192. Miller RR, Muller NL: Neuroendocrine cell hyperplasia and obliterative bronchiolitis in patients with peripheral carcinoid tumors. Am J Surg Pathol 19:653–658, 1995. 193. Min JW, Um SW, Yim JJ, et al: The role of whole-body FDG PET/CT, Tc 99m MDP bone scintigraphy, and serum alkaline phosphatase in detecting bone metastasis in patients with newly diagnosed lung cancer. J Korean Med Sci 24:275–280, 2009. 194. Mintz BJ, Tuhrim S, Alexander S, et al: Intracranial metastases in   the initial staging of bronchogenic carcinoma. Chest 86:850–853, 1984. 195. Misthos P, Papagiannakis G, Kokotsakis J, et al: Surgical management of lung cancer invading the aorta or the superior vena cava. Lung Cancer 56:223–227, 2007. 196. Miyake M, Tateishi U, Maeda T, et al: Bronchial granular cell tumor:   A case presenting secondary obstructive changes on CT. Radiat Med 24:154–157, 2006. 197. Moran CA, Suster S, Abbondanzo SL, et al: Primary leiomyosarcomas of the lung: A clinicopathologic and immunohistochemical study of 18 cases. Mod Pathol 10:121–128, 1997.

198. Moran CA, Suster S, Koss MN: Primary adenoid cystic carcinoma of the lung. A clinicopathologic and immunohistochemical study of 16 cases. Cancer 73:1390–1397, 1994. 199. Morandi L, Asioli S, Cavazza A, et al: Genetic relationship among atypical adenomatous hyperplasia, bronchioloalveolar carcinoma and adenocarcinoma of the lung. Lung Cancer 56:35–42, 2007. 200. Morgan B, Coakley F, Finlay DB, et al: Hypertrophic osteoarthropathy in staging skeletal scintigraphy for lung cancer. Clin Radiol 51:694–697, 1996. 201. Mori M, Rao SK, Popper HH, et al: Atypical adenomatous hyperplasia of the lung: A probable forerunner in the development of adenocarcinoma of the lung. Mod Pathol 14:72–84, 2001. 202. Mounawar M, Mukeria A, Le Calvez F, et al: Patterns of EGFR, HER2, TP53, and KRAS mutations of p14arf expression in non-small cell   lung cancers in relation to smoking history. Cancer Res 67:5667–5672, 2007. 203. Munden RF, Pugatch RD, Liptay MJ, et al: Small pulmonary lesions detected at CT: Clinical importance. Radiology 202:105–110, 1997. 204. Naidich DP, Bankier AA, MacMahon H, et al: Recommendations   for the management of subsolid pulmonary nodules detected at CT:   a statement from the Fleischner Society. Radiology 266:304–317, 2013. 205. Nakahara R, Yokose T, Nagai K, et al: Atypical adenomatous hyperplasia of the lung: A clinicopathological study of 118 cases including cases with multiple atypical adenomatous hyperplasia. Thorax 56:302–305, 2001. 206. Nakanishi K: Alveolar epithelial hyperplasia and adenocarcinoma of the lung. Arch Pathol Lab Med 114:363–368, 1990. 207. Nakata M, Sawada S, Saeki H, et al: Prospective study of thoracoscopic limited resection for ground-glass opacity selected by computed tomography. Ann Thorac Surg 75:1601–1605, discussion 5–6, 2003. 208. Nappi O, Glasner SD, Swanson PE, et al: Biphasic and monophasic sarcomatoid carcinomas of the lung. A reappraisal of ‘carcinosarcomas’ and ‘spindle-cell carcinomas’. Am J Clin Pathol 102:331–340, 1994. 209. Neuman J, Rosioreanu A, Schuss A, et al: Radiology-pathology conference: Sclerosing hemangioma of the lung. Clin Imaging 30:409–412, 2006. 210. Nicholson AH: N. Marginal zone B-cell lymphoma of the mucosaassociated lymphoid tissue (MALT) type. In Travis WDBE, MullerHermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 88–90. 211. Nicholson AH: N. Primary pulmonary diffuse large cell B-lymphoma. In Travis WDBE, Muller-Hermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 91–94. 212. Nielsen ME, Jr, Heaston DK, Dunnick NR, et al: Preoperative CT evaluation of adrenal glands in non-small cell bronchogenic carcinoma. AJR Am J Roentgenol 139:317–320, 1982. 213. Niho S, Fujii H, Murakami K, et al: Detection of unsuspected distant metastases and/or regional nodes by FDG-PET [corrected] scan in apparent limited-disease small-cell lung cancer. Lung Cancer 57:328– 333, 2007. 214. Niho S, Suzuki K, Yokose T, et al: Monoclonality of both pale cells and cuboidal cells of sclerosing hemangioma of the lung. Am J Pathol 152:1065–1069, 1998. 215. Noguchi M, Morikawa A, Kawasaki M, et al: Small adenocarcinoma of the lung. Histologic characteristics and prognosis. Cancer 75:2844– 2852, 1995. 216. Nomori H, Watanabe K, Ohtsuka T, et al: Evaluation of F-18 fluorodeoxyglucose (FDG) PET scanning for pulmonary nodules less than 3 cm in diameter, with special reference to the CT images. Lung Cancer 45:19–27, 2004. 217. O’Donnell PG, Jackson SA, Tung KT, et al: Radiological appearances of lymphomas arising from mucosa-associated lymphoid tissue (MALT) in the lung. Clin Radiol 53:258–263, 1998. 218. Oldham HN, Jr, Young WG, Jr, Sealy WC: Hamartoma of the lung.   J Thorac Cardiovasc Surg 53:735–742, 1967.

CHAPTER 37  Neoplastic Disease of the Lung 219. Oliver TW, Jr, Bernardino ME, Miller JI, et al: Isolated adrenal masses in nonsmall-cell bronchogenic carcinoma. Radiology 153:217–218, 1984. 220. Ost DE, Yeung SC, Tanoue LT, et al: Clinical and organizational factors in the initial evaluation of patients with lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 143:e121S– e141S, 2013. 221. Outwater EK, Siegelman ES, Huang AB, et al: Adrenal masses: Correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposed-phase sequences. Radiology 200:749–752, 1996. 222. Padovani B, Mouroux J, Seksik L, et al: Chest wall invasion by bronchogenic carcinoma: Evaluation with MR imaging. Radiology 187:33–38, 1993. 223. Pagani JJ: Normal adrenal glands in small cell lung carcinoma: CT-guided biopsy. AJR Am J Roentgenol 140:949–951, 1983. 224. Pagani JJ: Non-small cell lung carcinoma adrenal metastases. Computed tomography and percutaneous needle biopsy in their diagnosis. Cancer 53:1058–1060, 1984. 225. Park CM, Goo JM, Lee HJ, et al: CT findings of atypical adenomatous hyperplasia in the lung. Korean J Radiol 7:80–86, 2006. 226. Park SH, Kim TJ, Chi JG: Congenital granular cell tumor with systemic involvement. Immunohistochemical and ultrastructural study. Arch Pathol Lab Med 115:934–938, 1991. 227. Paskin DL, Hull JD, Cookson PJ: Granular cell myoblastoma:   A comprehensive review of 15-years experience. Ann Surg 175:501–504, 1972. 228. Patz EF, Jr, Pinsky P, Gatsonis C, et al: Overdiagnosis in low-dose computed tomography screening for lung cancer. JAMA Intern Med 174:269–274, 2014. 229. Pena CS, Boland GW, Hahn PF, et al: Characterization of indeterminate (lipid-poor) adrenal masses: Use of washout characteristics at contrast-enhanced CT. Radiology 217:798–802, 2000. 230. Pfister DG, Johnson DH, Azzoli CG, et al: American Society of Clinical Oncology treatment of unresectable non-small-cell lung cancer guideline: Update 2003. J Clin Oncol 22:330–353, 2004. 231. Pieterman RM, van Putten JW, Meuzelaar JJ, et al: Preoperative staging of non-small-cell lung cancer with positron emission tomography.   N Engl J Med 343:254–261, 2000. 232. Pinchot SN, Holen K, Sippel RS, et al: Carcinoid tumors. Oncologist 13:1255–1269, 2008. 233. Postmus PE, Brambilla E, Chansky K, et al: The IASLC Lung Cancer Staging Project: Proposals for revision of the M descriptors in the forthcoming (seventh) edition of the TNM classification of lung cancer. J Thorac Oncol 2:686–693, 2007. 234. Prenzel KL, Monig SP, Sinning JM, et al: Lymph node size and metastatic infiltration in non-small cell lung cancer. Chest 123:463–467, 2003. 235. Prohm P, Winter J, Schmucker P: Pulmonary hamartoma. Thorac Cardiovasc Surg 30:302–305, 1982. 236. Qu X, Huang X, Yan W, et al: A meta-analysis of (1)(8)FDG-PET-CT, (1)(8)FDG-PET, MRI and bone scintigraphy for diagnosis of bone metastases in patients with lung cancer. Eur J Radiol 81:1007–1015, 2012. 237. Quint LE, Francis IR, Wahl RL, et al: Preoperative staging of non-smallcell carcinoma of the lung: Imaging methods. Am J Roentgenol 164:1349–1359, 1995. 238. Rami-Porta R, Ball D, Crowley J, et al: The IASLC Lung Cancer   Staging Project: Proposals for the revision of the T descriptors in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol 2:593–602, 2007. 239. Rami-Porta R, Bolejack V, Goldstraw P: The new tumor, node, and metastasis staging system. Semin Respir Crit Care Med 32:44–51, 2011. 240. Ratto GB, Costa R, Vassallo G, et al: Twelve-year experience with left atrial resection in the treatment of non-small cell lung cancer. Ann Thorac Surg 78:234–237, 2004.

1019

241. Ratto GB, Piacenza G, Frola C, et al: Chest wall involvement by lung cancer: Computed tomographic detection and results of operation.   Ann Thorac Surg 51:182–188, 1991. 242. Rea F, Rizzardi G, Zuin A, et al: Outcome and surgical strategy in bronchial carcinoid tumors: Single institution experience with 252 patients. Eur J Cardiothorac Surg 31:186–191, 2007. 243. Reducing the Health Consequences of Smoking: 25 Years of Progress:   A Report of the Surgeon General. In Services UDoHaH, editor: Centers for Disease Control (CDC), Washington, DC, 1989, Government Printing Office. MMWR Morb Mortal Wkly Rep Mar 24;38(Suppl 2):1–32, 1989. 244. Reed CE, Harpole DH, Posther KE, et al: Results of the American College of Surgeons Oncology Group Z0050 trial: The utility of positron emission tomography in staging potentially operable non-small cell lung cancer. J Thorac Cardiovasc Surg 126:1943–1951, 2003. 245. Regnard JF, Perrotin C, Giovannetti R, et al: Resection for tumors with carinal involvement: Technical aspects, results, and prognostic factors. Ann Thorac Surg 80:1841–1846, 2005. 246. Rekhtman N: Neuroendocrine tumors of the lung: An update. Arch Pathol Lab Med 134:1628–1638, 2010. 247. Revel MP, Merlin A, Peyrard S, et al: Software volumetric evaluation of doubling times for differentiating benign versus malignant pulmonary nodules. AJR Am J Roentgenol 187:135–142, 2006. 248. Rivera MP: Lung cancer in women: Differences in epidemiology, biology, histology, and treatment outcomes. Semin Respir Crit Care Med 34:792–801, 2013. 249. Rivera MP, Mehta AC, Wahidi MM: Establishing the diagnosis of lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 143:e142S–e165S, 2013. 250. Robinson JM, Knoll R, Henry DA: Intrathoracic granular cell myoblastoma. South Med J 81:1453–1457, 1988. 251. Robnett TJ, Machtay M, Stevenson JP, et al: Factors affecting the   risk of brain metastases after definitive chemoradiation for locally advanced non-small-cell lung carcinoma. J Clin Oncol 19:1344–1349, 2001. 252. Rock MJ, Kaufman RA, Lobe TE, et al: Epithelioid hemangioendothelioma of the lung (intravascular bronchioloalveolar tumor) in a young girl. Pediatr Pulmonol 11:181–186, 1991. 253. Roggli VL, Hammar SP, Pratt PC, et al: Does asbestos or asbestosis cause carcinoma of the lung? Am J Ind Med 26:835–838, 1994. 254. Rosado-de-Christenson ML, Templeton PA, Moran CA: Bronchogenic carcinoma: Radiologic-pathologic correlation. Radiographics 14:429– 446, quiz 47–48, 1994. 255. Ross GJ, Violi L, Friedman AC, et al: Intravascular bronchioloalveolar tumor: CT and pathologic correlation. J Comput Assist Tomogr 13:240–243, 1989. 256. Rossi G, Cavazza A, Sturm N, et al: Pulmonary carcinomas with pleomorphic, sarcomatoid, or sarcomatous elements: A clinicopathologic and immunohistochemical study of 75 cases. Am J Surg Pathol 27:311–324, 2003. 257. Ruebner BH, Eggleston JC: What is new in epithelioid hemangioendothelioma of the liver? Pathol Res Pract 182:110–112, 1987. 258. Ruffini E, Bongiovanni M, Cavallo A, et al: The significance of associated pre-invasive lesions in patients resected for primary lung neoplasms. Eur J Cardiothorac Surg 26:165–172, 2004. 259. Rusch VW, Asamura H, Watanabe H, et al: The IASLC lung cancer staging project: A proposal for a new international lymph node map in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 4:568–577, 2009. 260. Rusch VW, Crowley J, Giroux DJ, et al: The IASLC Lung Cancer   Staging Project: Proposals for the revision of the N descriptors in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 2:603–612, 2007. 261. Saghir Z, Dirksen A, Ashraf H, et al: CT screening for lung cancer brings forward early disease. The randomised Danish Lung Cancer

1020

262.

263.

264.

265.

266. 267. 268.

269.

270. 271. 272. 273. 274.

275. 276.

277. 278.

279.

280.

281. 282.

283. 284.

285. 286.

287.

PART II  CT and MR Imaging of the Whole Body

Screening Trial: Status after five annual screening rounds with low-dose CT. Thorax 67:296–301, 2012. Sakamoto N, Adachi S, Monzawa S, et al: High resolution CT findings of pulmonary epithelioid hemangioendothelioma: Unusual manifestations in 2 cases. J Thorac Imaging 20:236–238, 2005. Sakamoto H, Shimizu J, Horio Y, et al: Disproportionate representation of KRAS gene mutation in atypical adenomatous hyperplasia, but even distribution of EGFR gene mutation from preinvasive to invasive adenocarcinomas. J Pathol 212:287–294, 2007. Sale GE, Kulander BG: “Benign” clear-cell tumor (sugar tumor) of the lung with hepatic metastases ten years after resection of pulmonary primary tumor. Arch Pathol Lab Med 112:1177–1178, 1988. Salminen US: Pulmonary hamartoma. A clinical study of 77 cases   in a 21-year period and review of literature. Eur J Cardiothorac Surg 4:15–18, 1990. Samet JM: The epidemiology of lung cancer. Chest 103:20S–29S, 1993. Schwartz AG: Lung cancer: family history matters. Chest 130:936–937, 2006. Schwartz LH, Ginsberg MS, Burt ME, et al: MRI as an alternative to CT-guided biopsy of adrenal masses in patients with lung cancer. Ann Thorac Surg 65:193–197, 1998. Shah SK, McNitt-Gray MF, Rogers SR, et al: Computer aided characterization of the solitary pulmonary nodule using volumetric and contrast enhancement features. Acad Radiol 12:1310–1319, 2005. Shen C, Che G: Clinicopathological analysis of pulmonary mucoepidermoid carcinoma. World J Surg Oncol 12:33, 2014. Sider L: Radiographic manifestations of primary bronchogenic carcinoma. Radiol Clin North Am 28:583–597, 1990. Siegel R, Ma J, Zou Z, et al: Cancer statistics, 2014. CA Cancer J Clin 64:9–29, 2014. Siegelman SS, Khouri NF, Scott JWW, et al: Pulmonary hamartoma: CT findings. Radiology 160:313–317, 1986. Silvestri GA, Littenberg B, Colice GL: The clinical evaluation for detecting metastatic lung cancer. A meta- analysis. Am J Respir Crit Care Med 152:225–230, 1995. Soga J: Carcinoids and their variant endocrinomas. An analysis of 11842 reported cases. J Exp Clin Cancer Res 22:517–530, 2003. Sone S, Takashima S, Li F, et al: Mass screening for lung cancer with mobile spiral computed tomography scanner. Lancet 351:1242–1245, 1998. Spencer H: Pathology of the lung, Oxford, 1985, Pergamon. Stahel RA, Ginsberg R, Havemann K, et al: Staging and prognostic factors in small cell lung cancer: A consensus report. Lung Cancer 5:119–126, 1989. Stahel RA, Mabry M, Skarin AT, et al: Detection of bone marrow metastasis in small-cell lung cancer by monoclonal antibody. J Clin Oncol 3:455–461, 1985. Stayner L, Bena J, Sasco AJ, et al: Lung cancer risk and workplace exposure to environmental tobacco smoke. Am J Public Health 97:545–551, 2007. Steenland K, Loomis D, Shy C, et al: Review of occupational lung carcinogens. Am J Ind Med 29:474–490, 1996. Sterner DJ, Mori M, Roggli VL, et al: Prevalence of pulmonary atypical alveolar cell hyperplasia in an autopsy population: A study of 100 cases. Mod Pathol 10:469–473, 1997. Stitik FP: The new staging of lung cancer. Radiol Clin North Am 32:635–647, 1994. Sugio K, Yokoyama H, Kaneko S, et al: Sclerosing hemangioma of the lung: Radiographic and pathological study. Ann Thorac Surg 53:295– 300, 1992. Suster S: Primary sarcomas of the lung. Semin Diagn Pathol 12:140– 157, 1995. Suzuki K, Kusumoto M, Watanabe S, et al: Radiologic classification   of small adenocarcinoma of the lung: Radiologic-pathologic   correlation and its prognostic impact. Ann Thorac Surg 81:413–419, 2006. Swensen SJ, Jett JR, Hartman TE, et al: CT screening for lung cancer: Five-year prospective experience. Radiology 235:259–265, 2005.

288. Swensen SJ, Viggiano RW, Midthun DE, et al: Lung nodule enhancement at CT: Multicenter study. Radiology 214:73–80, 2000. 289. Takashima S, Li F, Maruyama Y, et al: Discrimination of subtypes of small adenocarcinoma in the lung with thin-section CT. Lung Cancer 36:175–182, 2002. 290. The Health Benefits of Smoking Cessation: A Report of the Surgeon General. In Services UDoHaH, editor: Samet JM, Washington, DC, 1990, Government Printing Office. Am Rev Respir Dis 142:993–994, 1990. 291. Theros EG: 1976 Caldwell Lecture: Varying manifestation of peripheral pulmonary neoplasms: A radiologic-pathologic correlative study. AJR Am J Roentgenol 128:893–914, 1977. 292. Toh HC, Ang PT: Primary pulmonary lymphoma—Clinical review   from a single institution in Singapore. Leuk Lymphoma 27:153–163, 1997. 293. Toloza EM, Harpole L, Detterbeck F, et al: Invasive staging of non-small cell lung cancer: A review of the current evidence. Chest 123:157S– 166S, 2003. 294. Travis WD: Advances in neuroendocrine lung tumors. Ann Oncol 21(Suppl 7):vii65–vii71, 2010. 295. Travis WD: Sarcomatoid neoplasms of the lung and pleura. Arch Pathol Lab Med 134:1645–1658, 2010. 296. Travis WD, Brambilla E, Noguchi M, et al: International   Association for the Study of Lung Cancer/American Thoracic   Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 6:244–285, 2011. 297. Travis WD, Brambilla E, Riely GJ: New pathologic classification of lung cancer: Relevance for clinical practice and clinical trials. J Clin Oncol 31:992–1001, 2013. 298. Travis WD, Colby TV, Corrin B, et al: Histological typing of lung and pleural tumours, ed 3, 1999, Springer-Verlag. 299. Travis WDCTV, Corrin B, et al: Histological Typing of Lung and Pleural Tumours, ed 3, Berlin, 1999, Springer. 300. Travis WD, Lubin J, Ries L, et al: United States lung carcinoma incidence trends: Declining for most histologic types among males, increasing among females. Cancer 77:2464–2470, 1996. 301. Travis WDBE, Muller-Hermelink HK, Harris CC: Pathology and genetics of tumours of the lung pleura, thymus and heart, Lyon, France, 2004, International Agency for Research on Cancer. 302. Travis WDBE, Muller-Hermelink HK, Harris CC: Pathology and Genetics: Tumours of the Lung, Pleura, Thymus and Heart, Lyon, France, 2004, IARC Press. 303. Urschel JD: Surgical treatment of peripheral small cell lung cancer. Chest Surg Clin N Am 7:95–103, 1997. 304. Vainio H, Boffetta P: Mechanisms of the combined effect of asbestos and smoking in the etiology of lung cancer. Scand J Work Environ Health 20:235–242, 1994. 305. Valanis BG: Epidemiology of lung cancer: A worldwide epidemic. Semin Oncol Nurs 12:251–259, 1996. 306. Valli M, Fabris GA, Dewar A, et al: Atypical carcinoid tumour of the lung: A study of 33 cases with prognostic features. Histopathology 24:363–369, 1994. 307. Vallieres E, Shepherd FA, Crowley J, et al: The IASLC Lung Cancer Staging Project: Proposals regarding the relevance of TNM in the pathologic staging of small cell lung cancer in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol 4:1049–1059, 2009. 308. van de Pol M, van Oosterhout AGM, Wilmink JT, et al: MRI in detection of brain metastases at initial staging of small-cell lung cancer. Neuroradiology 38:207–210, 1996. 309. Van Loo S, Boeykens E, Stappaerts I, et al: Classic biphasic pulmonary blastoma: A case report and review of the literature. Lung Cancer 73:127–132, 2011. 310. van Tinteren H, Hoekstra OS, Smit EF, et al: Effectiveness of positron emission tomography in the preoperative assessment of patients with suspected non-small-cell lung cancer: The PLUS multicentre randomised trial. Lancet 359:1388–1393, 2002.

CHAPTER 37  Neoplastic Disease of the Lung 311. Vedeler CA, Antoine JC, Giometto B, et al: Management of paraneoplastic neurological syndromes: Report of an EFNS Task Force. Eur J Neurol 13:682–690, 2006. 312. Verhagen AF, Bootsma GP, Tjan-Heijnen VC, et al: FDG-PET in staging lung cancer: How does it change the algorithm? Lung Cancer 44:175– 181, 2004. 313. Viney RC, Boyer MJ, King MT, et al: Randomized controlled trial of the role of positron emission tomography in the management of stage I and II non-small-cell lung cancer. J Clin Oncol 22:2357–2362, 2004. 314. Wahbah M, Boroumand N, Castro C, et al: Changing trends in the distribution of the histologic types of lung cancer: A review of 4,439 cases. Ann Diagn Pathol 11:89–96, 2007. 315. Wang Y, Ju C, Stark AD, et al: Radon mitigation survey among New York State residents living in high radon homes. Health Phys 77:403– 409, 1999. 316. Wannesson L, Cavalli F, Zucca E: Primary pulmonary lymphoma: Current status. Clin Lymphoma Myeloma 6:220–227, 2005. 317. Webb WR, Gatsonis C, Zerhouni EA, et al: CT and MR imaging in staging non-small cell bronchogenic carcinoma: Report of the Radiologic Diagnostic Oncology Group. Radiology 178:705–713, 1991. 318. Weiss W: Asbestosis: A marker for the increased risk of lung cancer among workers exposed to asbestos. Chest 115:536–549, 1999. 319. Weiss SW, Ishak KG, Dail DH, et al: Epithelioid hemangioendothelioma and related lesions. Semin Diagn Pathol 3:259–287, 1986. 320. Weng S, Tsuchiya E, Satoh Y, et al: Multiple atypical adenomatous hyperplasia of type II pneumonocytes and bronchiolo-alveolar carcinoma. Histopathology 16:101–103, 1990. 321. Wick MR, Ritter JH, Humphrey PA: Sarcomatoid carcinomas of the lung. Am J Clin Pathol 108:40–53, 1997. 322. Wislez M, Beer DG, Wistuba I, et al: Molecular biology, genomics, and proteomics in bronchioloalveolar carcinoma. J Thorac Oncol 1:S8–S12, 2006. 323. Woodring JH, Stelling CB: Adenocarcinoma of the lung: A tumor   with a changing pleomorphic character. Am J Roentgenol 140:657–664, 1983. 324. Wu YL, Zhong WZ, Li LY, et al: Epidermal growth factor receptor mutations and their correlation with gefitinib therapy in patients   with non-small cell lung cancer: A meta-analysis based on updated individual patient data from six medical centers in mainland China.   J Thorac Oncol 2:430–439, 2007. 325. Wynder EL, Hoffmann D: Smoking and lung cancer: Scientific challenges and opportunities. Cancer Res 54:5284–5295, 1994.

1021

326. Xi JJ, Jiang W, Lu SH, et al: Primary pulmonary mucoepidermoid carcinoma: An analysis of 21 cases. World J Surg Oncol 10:232, 2012. 327. Yaghmai I: Angiographic manifestations of soft-tissue and osseous hemangiopericytomas. Radiology 126:653–659, 1978. 328. Yamada N, Kusumoto M, Maeshima A, et al: Correlation of the solid part on high-resolution computed tomography with pathological scar in small lung adenocarcinomas. Jpn J Clin Oncol 37:913–917, 2007. 329. Yang P, Cerhan JR, Vierkant RA, et al: Adenocarcinoma of the lung is strongly associated with cigarette smoking: Further evidence from a prospective study of women. Am J Epidemiol 156:1114–1122, 2002. 330. Yankelevitz DF, Gupta R, Zhao B, et al: Small pulmonary nodules: Evaluation with repeat CT–Preliminary experience. Radiology 212:561–566, 1999. 331. Yankelevitz DF, Henschke CI: Does 2-year stability imply that pulmonary nodules are benign? Am J Roentgenol 168:325–328, 1997. 332. Yatabe Y, Mitsudomi T: Epidermal growth factor receptor mutations in lung cancers. Pathol Int 57:233–244, 2007. 333. Yokoi K, Kamiya N, Matsuguma H, et al: Detection of brain metastasis in potentially operable non-small cell lung cancer. A comparison of CT and MRI. Chest 115:714–719, 1999. 334. Yokose T, Doi M, Tanno K, et al: Atypical adenomatous hyperplasia of the lung in autopsy cases. Lung Cancer 33:155–161, 2001. 335. Yousem SN: A. Mucoepidermoid carcinoma. In Travis WDBE, Muller-Hermelink HK, Harris CC, editors: World Health Organization classification of tumours: Tumours of the lung, pleura, thymus and heart, Lyon, France, 2004, IARC Press, pp 63–64. 336. Yousem SA, Hochholzer L: Mucoepidermoid tumors of the lung. Cancer 60:1346–1352, 1987. 337. Yousem SA, Hochholzer L: Primary pulmonary hemangiopericytoma. Cancer 59:549–555, 1987. 338. Zern H, Moran CA, Suster S, et al: Primary pulmonary sarcomas with features of monophasic synovial sarcoma: A clinicopathological, immunohistochemical, and ultrastructural study of 25 cases. Hum Pathol 26:474–480, 1995. 339. Zolliker A, Jacques J, Goldstein AS: Benign clear cell tumor of the lung. Arch Pathol Lab Med 103:526–530, 1979. 340. Zwiebel BR, Austin JH, Grimes MM: Bronchial carcinoid tumors: Assessment with CT of location and intratumoral calcification in 31 patients. Radiology 179:483–486, 1991. 341. Zwirewich CV, Vedal S, Miller RR, et al: Solitary pulmonary nodule: High-resolution CT and radiologic-pathologic correlation. Radiology 179:469–476, 1991.

38  Mediastinal Disease Jared D. Christensen, Danielle M. Seaman, and H. Page McAdams

There are many excellent textbooks and articles on the mediastinum.86,93,105,126,128,237,353 This chapter references these articles and a review of the recent literature as the basis for a discussion of the anatomy, pathology, and radiologic manifestations of mediastinal diseases.

NORMAL ANATOMY Comprehensive knowledge of cross-sectional anatomy is required to accurately evaluate mediastinal abnormalities. Interpretation of mediastinal anatomy can be assisted by analyzing computed tomography (CT) images at specific levels within the chest that are easily identifiable because of their characteristic anatomic landmarks and appearance.

Axial Plane

Thoracic Inlet Level At the junction of the neck and the thorax, most of the mediastinal structures are vascular. The two brachiocephalic veins are formed as the internal jugular veins join the subclavian veins and are located posterior to the clavicular heads (Fig. 38-1). These veins are the most anterior and lateral of the six major vessels at this level. More medial are the two common carotid arteries, and just posterior to these are the subclavian arteries. The subclavian arteries and veins exit the mediastinum to enter the axilla after crossing over the first rib. The esophagus is posterior or posterolateral to the trachea.

Left Brachiocephalic Vein Level The left brachiocephalic vein crosses the midline anterior to the arterial branches of the aorta and joins the more vertically orientated right brachiocephalic vein to form the superior vena cava (SVC) (Fig. 38-2). The internal mammary veins can often be identified coursing posteriorly from a parasternal location to join the brachiocephalic veins. The innominate, left common carotid, and left subclavian arteries are located posterior to the left brachiocephalic vein and anterior to the trachea. The innominate artery is the more centrally located artery; the left common carotid artery (smallest of the three arteries) and the left subclavian artery are located to the left of the midline in a more lateral and posterior position. The esophagus maintains its position posterior or posterolateral to the trachea and anterior to the spine.

Aortic Arch Level The transverse arch crosses the mediastinum anterior to the trachea, coursing obliquely from right to left and from anterior to posterior (Fig. 38-3). The SVC is located adjacent to the anterior aspect of the transverse aorta to the right of the trachea. The fat-filled region

1022

posterior to the SVC, anterior to the trachea, and lateral to the aorta is the pretracheal or anterior paratracheal space. Anterior to the transverse aorta is the fat-filled prevascular space, a compartment of the anterior mediastinum that extends cephalad anterior to the great vessels of the aorta and the brachiocephalic veins. If present, the thymus gland is located in this space. Both of these spaces often contain a few small lymph nodes. The esophagus maintains its position posterior or posterolateral to the trachea and anterior to the spine.

Azygous Arch–Aortopulmonary Window Level The azygos vein, located anterior and slightly to the right of the spine, arches anteriorly and joins the SVC at this level (Figs. 38-4 and 38-5). The azygos arch may not be seen in its entirety on one slice and may be confused with lymphadenopathy. The ascending and descending aortas are visible as separate structures at this level. The ascending aorta is anterior to the trachea, and the descending aorta is posterolateral to the trachea on the left. Typically the ascending aorta (mean diameter 3.5 cm) is larger than the descending aorta (mean diameter 2.5 cm). The aortopulmonary window is a space located between the inferior aspect of the aortic arch and the superior aspect of the left main pulmonary artery. In some patients the close anatomic contiguity of these two structures prevents visualization of this space. The space is otherwise fat-filled and contains a few small lymph nodes, the left recurrent laryngeal nerve, and the ligamentum arteriosum. The esophagus maintains its position posterior to the trachea, anterior to the spine, and medial to the descending aorta.

Left Pulmonary Artery Level50,340 The main pulmonary artery is anterior and to the left of the ascending aorta (Figs. 38-6 and 38-7). The left pulmonary artery curves posteriorly from its origin from the main pulmonary artery and is located anterolateral to the left main bronchus at the level of the carina. The left superior pulmonary veins are located lateral to the posterior portion of the left pulmonary artery. On the right at the level of the carina is the origin of the right upper lobe bronchus. Anterior to the right upper lobe bronchus lies the right upper lobe pulmonary artery, or the truncus anterior. The right superior pulmonary veins are located anterior and lateral to the truncus anterior. The azygos vein is located posterior and to the right of the esophagus, while the hemiazygos vein parallels the course of the azygos but is to the left of the spine. The superior pericardial recess, a crescent-shaped extension of the pericardial space, is contiguous to the posterior aspect of the ascending aorta. The space often contains a small amount of fluid and may occasionally be confused with a lymph node; however, its characteristic location, shape, and low attenuation allow confident identification.

CHAPTER 38  Mediastinal Disease

1023

RCC LCC A

T LV

RV

S T AZ

RSC E

D

LSC E

FIG 38-1  Axial plane—thoracic inlet level. The trachea (T) and esophagus (E) are midline and separate the right and left vascular structures. LCC, left common carotid artery; LSC, left subclavian artery; LV, left brachiocephalic vein; RCC, right common carotid artery; RSC, right subclavian artery; RV, right brachiocephalic vein.

FIG 38-4  Axial plane—arch of the azygos vein or aortopulmonary window level. The azygos vein (AZ) arches from a posterior position along the midesophagus to join the superior vena cava (S), crossing over the right upper lobe bronchus. Between the ascending (A) and descending (D) aorta is the aortopulmonary window region. E, esophagus; T, trachea.

LV BC LCC

T

A

RV LSC E

S D

AZ E

FIG 38-2  Axial plane—left brachiocephalic vein level. The right (RV) and left (LV) brachiocephalic veins are anterior to the right brachiocephalic artery (BC), left common carotid artery (LCC), left subclavian artery (LSC), trachea (T), and esophagus (E).

FIG 38-5  Axial plane—aortopulmonary window level. The aortopulmonary window is between the ascending (A) and descending (D) aorta and the superior vena cava (S). AZ, azygos vein; E, esophagus.

ARCH S T

E

FIG 38-3  Axial plane—aortic arch level. The superior vena cava (S) and aortic arch (ARCH) lie anterior to the trachea (T) and esophagus (E). The thymus, which has undergone fatty infiltration, is anterior to the arch.

Occasionally the superior pericardial recess extends more cephalad to the level of the brachiocephalic vessels and may be mistaken for a mass or mediastinal cyst. The use of thin sections can be helpful to distinguish this so-called high-riding pericardial recess from pathologic conditions by showing that the “mass” is of water attenuation and by demonstrating continuity with the superior pericardial recess at the level of the ascending aorta. The concave extension of the right lung into the mediastinum anterior to the spine is the azygoesophageal recess, which extends inferiorly from the subcarinal region to the level of the diaphragm.

Right Pulmonary Artery Level The main pulmonary artery is anterior and to the left of the ascending aorta (Figs. 38-8 and 38-9). The right pulmonary artery extends posteriorly and to the right from the main pulmonary artery, passing anterior to the bronchus intermedius and posterior to the SVC. The right superior pulmonary vein is to the right of and lateral to the

1024

PART II  CT and MR Imaging of the Whole Body

LPA

A

S

PA

A

TA

LSPV RPA SPR D

AZE

D

FIG 38-6  Axial plane—left pulmonary artery level. To the left of the left pulmonary artery (LPA) lie the left superior pulmonary veins. The ascending (A) and descending (D) aorta and superior vena cava maintain their relative positions from the level above. The truncus anterior (TA), or right upper lobe pulmonary artery, is anterior to the right upper lobe bronchus. The azygoesophageal recess (AZE) is a concavity anterior to the spine. The superior pericardial recess (SPR) is an extension of the pericardium.

S

FIG 38-8  Axial plane—right pulmonary artery level. The main pulmonary artery (PA) divides into the left (LPA) and right (RPA) pulmonary arteries. The right pulmonary artery passes anterior to the right main bronchus, and the left pulmonary artery passes over the left main bronchus. A, ascending aorta; D, descending aorta; LSPV, left superior pulmonary vein; S, superior vena cava.

S

MPA

A

LPA

A

LPA

LSPV RPA

D

D

FIG 38-7  Axial plane—left pulmonary artery level. A, ascending aorta; D, descending aorta; LPA, left pulmonary artery; S, superior vena cava.

intrapulmonary portion of the right pulmonary artery. Anterior to the left main and upper lobe bronchi is the left superior pulmonary vein, and posterior to the left upper lobe bronchus is the left lower lobe pulmonary artery.

Left Atrial Level The anatomy of the mediastinum is complex at this level (Figs. 38-10 and 38-11). Anterior to the SVC is the right atrial appendage, curving laterally around the ascending aorta. The aorta is located centrally within the mediastinum, and anterior and to the left of the aorta is the pulmonary outflow tract. Posterolateral and to the left of the pulmonary outflow tract is the left atrial appendage. Within the fat between the left atrial appendage and the aortic root is the left main coronary artery. The left superior pulmonary vein enters the left atrium imme-

FIG 38-9  Axial plane—right pulmonary artery level. The main pulmonary artery (MPA) divides into the right pulmonary artery (RPA) and the left pulmonary artery, which is more superior and not visible. The left superior pulmonary vein (LSPV) is anterior to the left main bronchus, and the left pulmonary artery is posterior. A, ascending aorta; D, descending aorta; S, superior vena cava.

diately posterior to the left atrial appendage. The right superior pulmonary veins enter the left atrium just posterior to the SVC. The esophagus position at this level is variable and may be midline or to the right or left of midline along the posterior wall of the left atrium.

Four-Chamber Level All four chambers of the heart are identified at this axial level (Figs. 38-12 and 38-13) and should not be confused with a four-chamber cardiac reconstructed image. The posteriorly located left atrium is the most cranial of all the chambers. The right and left inferior pulmonary

CHAPTER 38  Mediastinal Disease

RV

RAP S

1025

P

RA

LV

A A LAP

LA

LA

RPA D

LPA

FIG 38-10  Axial plane—left atrial level. The left atrium (LA) is the most superior and posterior chamber of the heart. Superior pulmonary veins enter the anterosuperior portion of the left atrium. The left atrial appendage (LAP) is situated anterior and to the left of the left atrium, adjacent to the main pulmonary artery (P). The right atrial appendage (RAP) is anterior to the superior vena cava (S) and adjacent to the ascending aorta (A). D, descending aorta; LPA, left pulmonary artery; RPA, right pulmonary artery.

D

FIG 38-12  Axial plane—four-chamber level. A, aortic root; D, descending aorta; LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.

R

LAD RV

RA

C

A

IPV

LA RPA D

LPA

FIG 38-13  Axial plane—calcified coronary arteries at the four-chamber FIG 38-11  Axial plane—left atrial level. A, aortic root; D, descending aorta; IPV, left inferior pulmonary vein; LA, left atrium; LPA, left lower lobe pulmonary arteries; RA, right atrium; RPA, right lower lobe pulmonary artery; RV, right ventricle.

veins enter the posterolateral aspect of the left atrium. Anterior and to the right is the right atrium. To the left and anterior to the right atrium is the right ventricle, located behind the sternum. The left ventricle is to the left and posterior to the right ventricle. The coronary arteries can be detected if they are calcified or if there is fat surrounding the heart. The right coronary artery lies in the right atrioventricular groove. The circumflex coronary artery lies in the left atrioventricular groove, and the left anterior descending coronary artery lies in the interventricular groove between the left and right ventricles.

level. C, circumflex coronary artery; LAD, left anterior descending coronary artery; R, right coronary artery.

Three-Chamber Level The ventricles and right atrium are identified at this level (Fig. 38-14); the left atrium, because of its more cranial position, is not visible. The right atrium is located to the right. To the left and anteriorly, the thinwalled right ventricle is located just beneath the sternum. The thickwalled left ventricle makes up the posterolateral left portion of the mediastinum. Between the left ventricle and the inferior vena cava’s (IVC’s) opening into the right atrium is the coronary sinus.

Coronal and Sagittal Planes As CT imaging and workstation technologies have advanced, the use of multiplanar imaging reconstructions for interpretation has become routine. It is important for radiologists to recognize the normal

1026

PART II  CT and MR Imaging of the Whole Body T BC

RV LV

RA

RV

S

LCC A P

I CS AZ

D

LV

FIG 38-14  Axial plane—three-chamber level. The left atrium is more cranial to this level and not visible. The three visualized chambers are the right atrium (RA), right ventricle (RV), and left ventricle (LV). The coronary sinus (CS) is between the inferior vena cava (I) and the left ventricle. AZ, azygos vein; D, descending aorta.

FIG 38-16  Coronal plane—ascending aorta level. A, aorta; BC, brachiocephalic artery; LCC, left common carotid artery; LV, left ventricle; P, pulmonary artery; RV, right brachiocephalic vein; S, superior vena cava; T, trachea.

T RV LB

A

S

LCC A P

P

RAP

LA RA

LV

FIG 38-15  Coronal plane—aortic outflow tract level. A, aorta; LB, left

LV

FIG 38-17  Coronal plane—pulmonary artery level. A, aorta; LA, left

brachiocephalic vein; LV, left ventricle; P, pulmonary artery; RA, right atrium; RAP, right atrial appendage.

atrium; LCC, left common carotid artery; LV, left ventricle; P, pulmonary artery; RV, right brachiocephalic vein; S, superior vena cava; T, trachea.

mediastinal anatomy in the coronal and sagittal planes. Given the long axis of several structures in the mediastinum, axial and sagittal planes are particularly useful when assessing the aorta, great vessels, vena cava, trachea, esophagus, and heart. Coronal (Figs. 38-15 to 38-19) and sagittal (Figs. 38-20 to 38-24) reformations at different levels are provided for correlation. The same nomenclature as in the axial plane is used, along with the same abbreviations to identify the different structures.

anatomic regions or compartments is still important, it is not essential, because CT and magnetic resonance imaging (MRI) can accurately localize and in many instances characterize masses. Despite these advances, however, mediastinal masses are still discussed, classified, and grouped by their position on chest radiographs. In this chapter we use the widely accepted division of the mediastinum into superior, anterior, middle, and posterior compartments. The superior mediastinum is the space between the thoracic inlet and the superior aspect of the aortic arch (i.e., all mediastinal structures above an imaginary line drawn between the sternal angle and the fourth intervertebral disk on the lateral chest radiograph). The anterior mediastinum is the space anterior to the heart and great vessels on the lateral chest radiograph. It is bordered anteriorly by the sternum and posteriorly by the pericardium. Using this designation, normal

Mediastinal Divisions The mediastinum extends craniocaudad from the thoracic inlet to the diaphragm, and historically it has been divided into compartments (on the chest radiograph) to facilitate lesion localization and aid in the differential diagnosis. Although this division of the mediastinum into

CHAPTER 38  Mediastinal Disease

1027

LSC RV A

T

LPA

RAP S

RPA LA

RPA RA

FIG 38-18  Coronal plane—left atrial level. A, aorta; LA, left atrium; LPA, left pulmonary artery; LSC, left subclavian artery; RPA, right pulmonary artery; T, trachea.

FIG 38-20  Sagittal plane—superior vena cava level. RA, right atrium; RAP, right atrial appendage; RPA, right pulmonary artery; RV, right brachiocephalic vein; S, superior vena cava.

A

T

A

D

RPA

AZ

LA

FIG 38-19  Coronal plane—descending aorta level. A, aortic arch; D, descending aorta.

anterior mediastinal structures include the thymus, branches of the internal mammary artery and vein, lymph nodes, and fat. There is controversy concerning the division between the middle and posterior mediastinum. Some authors believe the dividing line should be the posterior aspect of the pericardium, whereas others divide the two by an imaginary line drawn 1 cm posterior to the anterior border of the vertebral bodies. The latter method places most aortic and esophageal lesions in the middle mediastinum and reserves the posterior mediastinum for neurogenic or other paraspinal lesions. Other methods divide the middle and posterior mediastinum based on the azygos arch and aorta. For purposes of this discussion, we define the middle mediastinum as the space between the anterior border of the pericardium and an imaginary line drawn 1 cm posterior to the anterior border of the vertebral bodies on the lateral radiograph. As such, it contains the heart, aorta, SVC, IVC, brachiocephalic arteries and veins, pulmonary

RA

FIG 38-21  Sagittal plane—right atrial level. A, ascending aorta; AZ, azygos vein; LA, left atrium; RA, right atrium; RPA, right pulmonary artery; T, trachea.

arteries and veins, thoracic duct, azygos and hemiazygos veins, phrenic and vagus nerves, trachea and proximal bronchi, esophagus, mediastinal fat, and lymph nodes. The posterior mediastinum is defined as the space between the imaginary line drawn 1 cm posterior to the anterior border of the vertebral bodies and the posterior paravertebral gutters. Structures in the posterior mediastinum thus include nerves, fat, vertebral column, and lymph nodes.

1028

PART II  CT and MR Imaging of the Whole Body

BC T

LPA

RPA A LA RV

RV

LV

D

FIG 38-22  Sagittal plane—aortic outflow tract level. A, ascending aorta; BC, brachiocephalic artery; D, descending aorta; LA, left atrium; RPA, right pulmonary artery; RV, right ventricle; T, trachea.

LCC LSC A

P

AT

LA

RV D

FIG 38-23  Sagittal plane—pulmonary artery outflow tract level. A, ascending aorta; AT, aortic outflow tract; D, descending aorta; LA, left atrium; LCC, left common carotid artery; LSC, left subclavian artery; P, pulmonary artery (outflow tract); RV, right ventricle.

IMAGING TECHNIQUES Computed Tomography Although the chest radiograph is usually the initial study that reveals a mediastinal abnormality, CT is used to assess the location and extent of mediastinal disease; because of its superior contrast resolution, it is also used to characterize the tissue components of masses (Fig. 38-25).

FIG 38-24  Sagittal plane—left ventricular level. LPA, left pulmonary artery; LV, left ventricle; RV, right ventricle.

CT is useful for distinguishing vascular variants or benign processes of the mediastinum (e.g., lipomatosis) from true pathologic conditions. Although CT is usually used to further evaluate an abnormality detected on the chest radiograph, in certain clinical situations CT may be performed when the chest radiograph is normal, such as in patients with myasthenia gravis, owing to the strong association between that disease and thymoma (Fig. 38-26). Also, certain malignancies such as lung cancer have a marked predilection to metastasize to mediastinal lymph nodes. Because these lymph node metastases may not be visible on chest radiographs, CT is used by most thoracic surgeons and oncologists to assess the mediastinal nodes in patients with lung cancer. Recent advances in CT imaging, such as multidetector and dualsource technologies, have improved its ability to image the mediastinum. Significant shortening of scanning time limits respiratory and cardiac motion artifacts, and in some instances the total dose of iodinated contrast material can be reduced. Dual-energy CT applications for imaging the mediastinum are not yet well established but may be helpful in quantifying perfusion, a potentially helpful marker in differentiating benign from malignant processes.182 Volumetric CT data sets can be effectively reconstructed in a variety of nonaxial planes, often facilitating interpretation of mediastinal abnormalities. Application of nonaxial two- and three-dimensional (2D and 3D) reconstruction techniques has proved most useful for imaging abnormalities of the central airways and great vessels (Figs. 38-27 and 38-28). For example, in the evaluation of stenoses in obliquely oriented bronchi, these reconstruction techniques can improve diagnostic accuracy and confidence in the interpretation. It should be emphasized, however, that axial CT images are usually sufficient for diagnosis. Nevertheless, because 2D and 3D images present anatomic information in a context more familiar to clinicians, these methods may demonstrate the location and extent of an abnormality in a way that radiologic reports and axial CT images often do not and can be helpful in treatment planning. Most routine CT scans of the mediastinum can be performed without intravenous (IV) contrast enhancement unless the primary

CHAPTER 38  Mediastinal Disease

1029

* *

A

A

B FIG 38-25  Coronary artery pseudoaneurysm post ascending aortic repair. A, Posteroanterior chest radiograph shows a mediastinal mass (arrows) superimposed on the right heart border and hilum. B, Contrastenhanced CT demonstrates the mass to represent a large pseudoaneurysm (arrows) with marginal thrombus (asterisks) arising from the right coronary artery anastomosis with the ascending aortic graft (A). Descending aortic dissection (arrowhead) is also present. A, ascending aortic graft.

S A

FIG 38-26  Thymoma in a woman with myasthenia gravis. The chest radiograph (not shown) was normal. Because of the association between myasthenia gravis and thymoma, contrast-enhanced chest CT was performed, revealing a small soft tissue mass in the anterior mediastinum (arrow). Resection confirmed thymoma. A, aorta; S, superior vena cava.

indication is a suspected vascular abnormality (Fig. 38-29). Interpretation of non–contrast-enhanced (NCE) CT scans can be more difficult and time-consuming than interpretation of contrast-enhanced scans and requires a thorough knowledge of mediastinal anatomy and normal variants. There are, however, several advantages to NCE CT: (1) scan time is decreased because there is no need to establish IV access, set up the contrast injector, or delay the scan as the bolus is administered; (2) there is no risk of contrast reaction; and (3) the cost is lower. When needed, IV contrast is given through a large forearm or antecubital vein using a power injector. The contrast injection rate and total injection volume depend on the indication. For routine purposes a rate of 3 mL/sec for a total volume of 100 mL is sufficient. For vascular imaging (aorta and pulmonary artery), higher injection rates (4-6 mL/sec) are often necessary. In such cases, injection through a relatively large-bore IV catheter (18 or 20 gauge) into a large antecubital vein is mandatory. Timing of the bolus in relationship to the scan also varies and depends on both the indication (routine vs. arterial imaging) and the time required to complete the scan. CT is usually performed without esophageal contrast; however, oral contrast can be helpful in identifying the relationship between the esophagus and other mediastinal structures and occasionally in depicting intrinsic esophageal abnormalities (Fig. 38-30). Commercial preparations of esophageal paste made of a dilute barium mixture are available. The patient must swallow several teaspoons of the paste and refrain from further swallowing during the examination. CT scans of the mediastinum are usually reconstructed and interpreted in an axial format. Until recently, sagittal, coronal, and off-axis reconstructions of the thorax were not routinely performed because of slice misregistration and respiratory motion artifacts. However, with

1030

PART II  CT and MR Imaging of the Whole Body S T

R R

L

U

L M

L

D

B

A

FIG 38-27  Fibrosing mediastinitis. A, Axial CT shows marked narrowing of the bronchus intermedius (arrowhead) by a soft tissue attenuation mass (M, arrows). Note the extensive subcarinal calcification. B, Volumerendered shaded-surface display shows long-segment irregular narrowing of the bronchus intermedius (arrowheads). 3D reconstructions can facilitate assessment and treatment of airway stenoses. D, descending aorta; L, left main bronchus; R, right main bronchus; S, superior vena cava; T, trachea; U, right upper lobe bronchus.

A

B FIG 38-28  Traumatic aortic pseudoaneurysm. A, Axial contrast-enhanced CT demonstrates a focal contour abnormality of the medial aortic arch (arrow). B, Volume surface-rendered reconstruction in a posterior coronal plane better demonstrates the three-dimensional nature of the pseudoaneurysm (arrow) and its relationship to other anatomic structures. 3D reconstructions of CT angiograms display anatomy in a more familiar perspective to clinicians than do axial CT images.

the advent of multiplanar scanning, continuous volume data sets can be acquired during a single breath hold, and excellent nonaxial 2D and 3D reconstructed images of mediastinal vascular structures and airways can be generated (see Figs. 38-3 and 38-4). The different reconstruction methods, such as multiplanar reformat imaging, multiplanar volume reformat imaging, and external and internal 3D renderings, vary significantly in computational complexity and time required to generate the images.

Interpretation of mediastinal CT images is most commonly performed at a PACS (picture archiving and communication system) or workstation that permits dynamic adjustment of window and level settings for thorough assessment of the mediastinum; a wide range of attenuation values (−800 Hounsfield units [HU] for lung, 600 HU for bone) are encountered. Appropriate image settings are essential for accurate depiction of air, fat, fluid, calcification, IV contrast, and bone. If hard-copy (film) format is used, care should be taken to ensure

CHAPTER 38  Mediastinal Disease

1031

adequate window and level settings are used. When filming CT studies of the chest, lung settings should have a wide width to encompass airfilled lung and soft tissue and are usually viewed at a level of −600 and a width of 1200. Because of the smaller density range among fat, soft tissue, bone, and contrast-enhanced vessels, mediastinal settings are usually viewed at a level of 50 and a width of 350. These levels and widths can be varied to aid in evaluation of any lesion that cannot be adequately differentiated on standard settings. Bone windows, with a wide width and a level of near 400 HU, are often needed to evaluate for possible bone metastases.

Magnetic Resonance Imaging

A

*

Although CT imaging usually provides the requisite information in most patients with mediastinal abnormalities, MRI, because of its multiplanar capability and high contrast resolution, is occasionally used to evaluate the location and extent of disease. MRI is the modality of choice for imaging neurogenic tumors, because it not only demonstrates the number and nature of the lesions but also depicts intraspinal extension. Additionally MRI is useful in confirming the cystic nature of mediastinal lesions that appear solid on CT (Fig. 38-31) and demonstrating vascular structures in patients in whom use of iodinated IV contrast is contraindicated (Fig. 38-32). The role of MRI in characterizing mediastinal pathology is evolving, with recent literature suggesting potential applications in differentiating benign from malignant nodal and thymic disease. Two potential disadvantages of MRI of mediastinal abnormalities, compared with CT, are its poor demonstration of calcification and comparatively poor spatial resolution.

Positron Emission Tomography

B

C FIG 38-29  Pseudoaneurysm of the proximal descending aorta. A, Posteroanterior chest radiograph shows a soft tissue mass (arrowheads) superimposed on the left hemithorax, with preserved arch and descending aortic interfaces. B, Noncontrast chest CT was performed, revealing a homogeneous mass (asterisk) of the middle and anterior mediastina. Percutaneous biopsy was requested by the primary care team for tissue diagnosis. C, In lieu of biopsy, a contrast-enhanced CT was performed, revealing the mass to represent a large pseudoaneurysm (arrows) arising at the isthmus of the proximal descending thoracic aorta.

Whereas CT and MRI provide morphologic information, positron emission tomography (PET) and integrated PET/CT provide important functional information in the evaluation and management  of mediastinal disease. Several studies have shown the value of fluorine-18 fluorodeoxyglucose (18F FDG) PET in differentiating benign from malignant mediastinal masses. FDG uptake is associated with thymoma, thymic carcinoma, carcinoid, lymphoma, esophageal carcinoma, and metastatic disease among other entities. Of particular interest in oncologic imaging is the identification of mediastinal nodal disease for staging purposes, which affects treatment decisions and potential resectability of the primary tumor. On CT, lymph  node size is the primary determinant of disease, with a short-axis dimension greater than 1  cm indicative of pathology. Unfortunately, metastatic involvement can be associated with normal-sized lymph nodes and go undetected at CT imaging. FDG PET, however, is highly sensitive and specific for detection of nodal disease relative to  CT alone. Accuracy, sensitivity and specificity improves even more when PET is performed in conjunction with CT (PET/CT). Fur­ thermore, PET is superior to CT in detecting distant metastases, which has been shown to improve staging accuracy and reduce futile treatments. FDG PET also plays an important role in monitoring treatment response following initiation of therapy. PET has also been found to provide prognostic information in certain mediastinal pathology; for example, a negative PET obtained following treatment for lymphoma is associated with a lower risk of recurrent disease and longer progression-free survival than in patients with a positive posttreatment PET examination. Although PET is primarily used in the evaluation of suspected malignancy, FDG uptake can also be associated with infectious and inflammatory conditions of the mediastinum, such as acute mediastinitis or sarcoidosis. Examples of the utility of FDG PET in the

1032

PART II  CT and MR Imaging of the Whole Body

L

*

M M

A

A

A

B FIG 38-30  Esophageal leiomyoma. A, CT shows a large middle mediastinal mass with focal punctate calcifications (arrowheads). Note displacement of the esophagus anteriorly (arrow). The CT appearance is nonspecific, and the origin of the mass is uncertain. A, aorta; L, liver; M, mass. B, CT following oral ingestion of barium reveals distortion of the esophageal lumen (asterisk), consistent with a mass (M) arising within the wall of the esophagus. Resection confirmed leiomyoma.

P A

A

P

A

B

C FIG 38-31  Mediastinal cyst. A, CT shows a well-circumscribed homogeneous mass in the anterior mediastinum (arrowheads). Although of lower attenuation than vascular structures, the appearance is not diagnostic for mediastinal cyst. A, ascending aorta; P, main pulmonary artery. B and C, Axial T1- and T2-weighted MRIs show the mass to be of homogeneous low signal intensity on the T1-weighted image (arrowheads in B) and high signal intensity on the T2-weighted image (C), consistent with a cyst.

CHAPTER 38  Mediastinal Disease

1033

TABLE 38-1  Mediastinal Abnormalities

by Compartment A

D

Location

Lesion

Superior mediastinum Anterior mediastinum

Thyroid goiter, tortuous great vessels Thymic lesions, germ cell tumors, parathyroid adenoma, lymphatic malformations, hemangioma Esophageal lesions, airway lesions, foregut cysts, pericardial cysts Neurogenic tumors, paraspinal abscess, extramedullary hematopoiesis Mediastinitis, lipomatosis, lymphadenopathy (lymphoma, metastases, Castleman’s disease, infection), mesenchymal tumors, vascular abnormalities and anomalies, diaphragmatic hernias

Middle mediastinum Posterior mediastinum Multiple compartments

A

A

D

P

B FIG 38-32  Non-Hodgkin’s lymphoma in a patient with renal insufficiency. A, CT shows a homogeneous middle and posterior mediastinal mass encasing the descending aorta (D), anteriorly displacing the tracheal carina (arrows) and extending into the paravertebral region bilaterally (arrowheads). Note small bilateral pleural effusions. A, ascending aorta. B, Cine gradient-recalled echo MRI performed to evaluate possible vascular invasion confirms flow within the descending aorta (D). Note the low-signal-intensity mass (arrowheads) and the high-signalintensity pleural effusion (P). A, ascending aorta.

evaluation of mediastinal processes are provided in reference to specific disease entities throughout this chapter.

MEDIASTINAL ABNORMALITIES A useful approach for discussing mediastinal abnormalities is to classify them into processes that predominantly affect specific compartments (superior, anterior, middle, or posterior) and those that can affect multiple compartments (Table 38-1).

Superior Mediastinal Abnormalities Common superior mediastinal abnormalities include intrathoracic extension of thyroid goiters or masses and tortuous great vessels.

Thyroid Goiter20,102,109,129,131,237,305 On CT the normal thyroid gland is typically visible at or just below the level of the cricoid cartilage (Fig. 38-33A). It appears on NCE CT

as two wedge-shaped structures of homogeneous attenuation on  either side of the trachea, separated by a narrow anterior isthmus. It typically enhances homogeneously following administration of IV contrast. On MRI T1-weighted images the normal thyroid gland usually has intermediate signal intensity, slightly greater than that of adjacent muscle, and on T2-weighted images it has higher signal intensity (see Fig. 38-33B). Most thyroid masses in the mediastinum are caused by intrathoracic extension of thyroid goiters; these account for up to 10% of mediastinal masses resected at thoracotomy. True ectopic thyroid masses in the mediastinum are rare. Typically a thyroid goiter extends into the thyropericardiac space anterior to the recurrent laryngeal nerve and brachiocephalic vessels, although posterior extension adjacent to the trachea occurs in 20% of cases. Rarely a thyroid mass extends behind the esophagus and presents as a posterior mediastinal mass. A mediastinal goiter is usually detected on chest radiographs as a thoracic inlet or superior mediastinal mass that deviates and occasionally narrows the trachea. Although scintigraphy can detect mediastinal goiters, uptake of technetium or iodine is variable and may not be visible. The CT appearance of a mediastinal goiter is variable, but it can be confidently diagnosed when lesion continuity with the thyroid gland is visible. Additional useful features to establish the diagnosis are heterogeneous attenuation with areas of both high and low attenuation on NCE scans, marked enhancement following administration of IV contrast, and focal punctate or curvilinear calcifications (Figs. 38-34 and 38-35). The cystic and adenomatoid composition of goiters results in a heterogeneous appearance on both T1- and T2-weighted MRIs (Fig. 38-36). On T1-weighted images goiters may have a lower signal intensity than normal thyroid, but high signal intensity may be seen in areas of subacute hemorrhage, colloid cysts, and adenomas. Goiters have high signal intensity on T2-weighted images.

Tortuous Great Vessels Tortuosity of the great vessels is a common cause of superior mediastinal abnormality on chest radiographs. The most common location of the radiographic abnormality is in the right paratracheal region; this is true because in older patients the right brachiocephalic artery often has a tortuous course before giving rise to the right subclavian and common carotid arteries. NCE CT can often determine that tortuous vessels are the cause of the apparent mass. Calcification within vessel

1034

PART II  CT and MR Imaging of the Whole Body

T

T

B

A

FIG 38-33  Normal thyroid gland. A, Contrast CT shows the typical appearance of the thyroid gland (arrows). Note the homogeneous enhancement. T, trachea. B, T1-weighted axial MRI shows homogeneous intermediate signal intensity, typical of a normal thyroid gland (arrows). T, trachea.

walls allows confident identification of tortuous arteries. If any doubt exists as to the vascular nature of the mass, IV contrast enhancement readily identifies the vessels (Fig. 38-37).

Anterior Mediastinal Abnormalities38,52,325,339,353 Approximately half of all mediastinal masses occur in the anterior mediastinum. Lesions that occur preferentially in the anterior mediastinum include thymic lesions (cysts, thymolipoma, thymoma, thymic carcinoma), germ cell tumors, parathyroid adenoma, lymphatic malformations, and hemangioma. Lymphoma is another common cause of an anterior mediastinal mass, but it is discussed in the section on diffuse mediastinal disease.

Thymic Lesions225,250,327 Normal thymus.17,64,87,92,100,237,246,318,355  The thymus is a bilobed triangular gland that occupies the thyropericardiac space of the anterior mediastinum and extends inferiorly to the heart. The normal morphology and size of the thymus change markedly with age. There is wide variation in the normal size of the thymus, particularly in children and young adults. In newborns the thymus gland is often larger than the heart. Thymic size decreases with age as the gland undergoes fatty infiltration. An atrophied thymus is often visualized on CT in patients in the fourth decade of life but is seen in less than 50% of patients older than 40 years (Fig. 38-38). The most useful measurement is the thickness of the lobes, measured perpendicular to the long axis of the gland. The normal maximal thickness before age 20 is 18 mm; it is 13 mm in older individuals. Although these measurements are useful indicators of thymic abnormality, thymic shape is also important; focal contour abnormality of the normal thymus gland suggests an underlying abnormality. On CT the normal thymus appears as a homogeneous bilobed structure of soft tissue attenuation in the anterior mediastinum (Fig. 38-39). It is usually seen at the level of the aortic arch and the origin of the great vessels. The left lobe is usually slightly larger than the right. Rarely a lobe is congenitally absent. The normal thymus is easily demonstrated on MRI (Fig. 38-40). Characteristically the thymus is homogeneous with intermediate signal intensity (less than that of fat) on T1-weighted images. Because the thymus begins to involute at puberty and is replaced by fat in older patients, the T1-weighted signal intensity of the thymus increases with age. On T2-weighted images the thymus has high signal intensity similar to fat in all age groups; this can make

identification of the thymus difficult in patients with abundant mediastinal fat. The normal thymus in patients younger than 20 years typically has diffuse increased FDG uptake on PET. After that age, significant FDG accumulation is less common. Nevertheless, significant FDG uptake in the thymus, sometimes simulating that seen in malignancy, is occasionally seen in adults (see Fig. 38-39). A standardized uptake value (SUV) of less than 3.1 is typical for normal thymic tissue, whereas SUV greater than 3.4 is suspicious for malignancy. Thymic masses may sometimes be difficult to distinguish from normal thymus. As a rule, however, thymic masses usually manifest on CT or MRI as round masses and do not conform to the shape of the normal thymus. In addition, thymic masses are usually of heterogeneous attenuation on CT, with areas of decreased attenuation and possibly calcification. Thymic hyperplasia.3,22,76,103,121,137,148,195,196  Thymic hyperplasia can occur in association with hyperthyroidism, acromegaly, Addison’s disease, or stress, as well as in patients receiving chemotherapy or radiation therapy and in those with myasthenia gravis. Differentiating between hyperplasia and other causes of thymic enlargement on CT can be difficult because there are no specific attenuation characteristics associated with hyperplasia. Because thymic rebound or hyperplasia occurs in patients after chemotherapy or radiation therapy, it can be difficult to distinguish this enlargement from recurrent tumor. Generally, however, the thymus in rebound hyperplasia has a normal shape, with homogeneous attenuation on CT (Figs. 38-41 and 38-42). Asymmetry of the lobes, focal contour abnormalities, and heterogeneity suggest tumor. MRI may be beneficial in distinguishing normal thymus and thymic hyperplasia from thymic neoplasms. Unlike thymic tumors, normal and hyperplastic thymic tissue suppresses on out-of-phase gradient echo imaging because of the presence of interspersed microscopic fat. In addition, thymic rebound or hyperplasia is typically of homogeneous normal intensity on MRI, whereas tumors demonstrate more heterogeneous signal properties and enhancement. Preliminary data suggest that FDG PET imaging may be helpful for differentiating thymic hyperplasia from thymoma in patients with myasthenia gravis. Patients with thymic hyperplasia show homogeneous FDG accumulation in the thymus (see Fig. 38-42), whereas patients with thymoma tend to have more focal FDG uptake. The degree of uptake tends to be greater in thymoma than in thymic hyperplasia.

1035

CHAPTER 38  Mediastinal Disease

T

G

T LSC

A LCC

T

A G

T

G A

B FIG 38-34  Thyroid goiter. A, CT shows a heterogeneous superior mediastinal mass (arrowheads) with punctate calcification. Images at the thoracic inlet (not shown) confirmed the mass’s contiguity with the left thyroid lobe, consistent with an intrathoracic extension of goiter. T, trachea. B, Contrast CT shows heterogeneous enhancement of the mass. T, trachea.

B FIG 38-35  Mediastinal goiter. A, Contrast CT in the axial plane shows a large goiter (G) displacing the trachea (T) to the right and displacing the left common carotid (LCC) and left subclavian arteries (LSC) to the left. Note the heterogeneous enhancement and calcification. B, In a different patient, contrast CT in the coronal plane shows intrathoracic extension of a goiter (G) and rightward displacement of the trachea (T). Note heterogeneous enhancement and calcification. A, aorta.

1036

PART II  CT and MR Imaging of the Whole Body

T T

FIG 38-36  Mediastinal goiter. Coronal MRI shows mediastinal thyroid tissue (T) on both sides of the trachea. The trachea is deviated to the right at the thoracic inlet.

Thymic cyst.3,18,51,62,115,139,146,195,219,298,325,331  Thymic cysts account for approximately 3% of all anterior mediastinal masses and are either congenital or secondary to inflammation or malignancy. Congenital thymic cysts most likely arise from remnants of the thymopharyngeal duct and are usually thin-walled unilocular lesions less than 6 cm in diameter. On CT they are homogeneous, of water attenuation, with very thin or imperceptible walls. On MRI they manifest as well-circumscribed anterior mediastinal masses of high signal intensity on T2-weighted images. Signal intensity usually increases with increasing repetition time (TR). Occasionally, thymic cysts appear as solid masses on CT because they are filled with proteinaceous fluid. In these cases, MRI can be useful to confirm the cystic nature of the lesion (Fig. 38-43). Acquired thymic cysts occur in the setting of inflammation (e.g., Sjögren’s syndrome, aplastic anemia, myasthenia gravis) or malignancy. Associated malignancies include Hodgkin’s lymphoma, seminoma, thymoma, and thymic carcinoma (Fig. 38-44). These cysts usually have walls of variable thickness, are multilocular, and range in size from 3 to 17 cm in diameter. These cysts can contain hemorrhage or calcification. As such, the CT and MRI features of acquired thymic cysts are more variable than those of congenital thymic cysts. Because of their association with thymic neoplasia, care must be taken when interpreting cystic lesions of the anterior mediastinum. If a thymic cyst is multilocular, heterogeneous, thick walled, or associated with a soft tissue component, it must be evaluated further (biopsy or resection) to exclude malignancy (see Fig. 38-44). Furthermore, multiloculated thymic cysts can exhibit FDG uptake on PET/CT, although typically less than that of thymoma or thymic carcinoma (Fig. 38-45). Thymolipoma.195,223,287,311,358  Thymolipomas are rare, benign, slowly growing neoplasms of the anterior mediastinum that can simulate cardiomegaly on the chest radiograph. They occur with equal incidence in men and women, and although there is a wide age range at presentation, most are diagnosed in young adults. On CT these masses are typically large and heterogeneous with a mixture of fat elements and soft tissue attenuation, conforming to adjacent structures. They are not locally invasive, although compression of adjacent structures occurs in approximately 50% of patients. Because they consist of fat

and residual thymic tissue, thymolipomas usually have high signal intensity (similar to fat) with interspersed areas of intermediate signal intensity on both T1- and T2-weighted MRIs. Thymic epithelial neoplasms.*  The classification of thymic epithelial neoplasms is continuing to evolve. According to the traditional classification, neoplasms of the thymus were termed thymoma, invasive thymoma, and thymic carcinoma. However, in the most recent 2004 World Health Organization (WHO) classification, they are designated thymoma (subtypes A and B) and thymic carcinoma. There are inherent problems with the WHO classification, including poor inter- and intraobserver reproducibility and inconsistencies between histologic grade and clinical outcomes. Despite these limitations, the WHO histologic subtype classification reportedly closely correlates with the clinicopathologic staging classification of thymomas devised by Masaoka and coworkers.205,168 The Masaoka-Koga staging system is most widely used today and is based on the presence of capsular invasion: Stage I: macroscopically encapsulated Stage IIa: microscopic capsular invasion Stage IIb: macroscopic invasion into surrounding fatty tissue of mediastinal pleura Stage III: macroscopic invasion into a neighboring organ Stage IVa: pleural or pericardial dissemination Stage IVb: lymphatic or hematogenous metastasis More recently, Suster and Moran proposed that thymic neoplasms be classified as well-differentiated (WHO thymoma types A, AB, B1, and B2), moderately-differentiated (WHO thymoma type B3), and poorlydifferentiated thymic carcinoma (WHO thymic carcinoma).329,330 Owing to the complexities of histopathologic classification and the lack of correlation with prognosis, the treatment of thymic epithelial tumors is primarily based upon clinical staging and completeness of surgical resection. These two variables have been found to best correlate with progression-free and overall survival. Thymoma.  Thymomas (well- and moderately-differentiated thymic carcinoma or thymoma types A, AB, B1, B2, and B3 in the WHO classification) account for approximately 20% of all mediastinal tumors and are the most common primary tumor of the anterior mediastinum. Seventy-five percent of thymomas occur in the anterior mediastinum, 15% occur in both the anterior and superior mediastina, and 6% occur in the superior mediastinum. Another 4% occur ectopically, with the posterior mediastinum being the least common location. Although thymomas can occur in children, most patients are older than 40 years at presentation; 70% of affected patients present in the fifth and sixth decades of life. There is an equal incidence in men and women. Most affected patients are asymptomatic at presentation, with the lesion detected on routine chest radiographs. Chest pain, cough, and symptoms related to compression of adjacent structures occur in approximately 33% of patients. Paraneoplastic syndromes are common and include myasthenia gravis (50%), hypogammaglobulinemia (10%), and pure red cell aplasia (5%). Although up to 50% of patients with thymoma have myasthenia gravis, only 10% to 20% of patients with myasthenia gravis have a thymoma; 65% of these patients do, however, have lymphoid hyperplasia of the thymus. Because of the strong association between myasthenia gravis and thymoma, CT is often performed in patients with symptoms of myasthenia gravis even if the chest radiograph is normal. Text continued on p. 1042

*References 1, 3, 25, 77, 78, 80, 120, 147, 150, 168, 172, 193, 195, 202, 203, 205, 226, 231-233, 268, 286, 292, 294, 298, 325, 328-330, 336, 343, 354.

CHAPTER 38  Mediastinal Disease

A

A

B FIG 38-37  Tortuous great vessels. A, Cone view from a posteroanterior chest radiograph shows a masslike opacity along the right superior mediastinum (arrow). B, Noncontrast CT shows tortuous right brachiocephalic and subclavian arteries (arrows) causing a right paratracheal mass. A, aorta.

1037

1038

PART II  CT and MR Imaging of the Whole Body

S S

A

A

A

B FIG 38-38  Thymus gland in a 42-year-old woman with breast carcinoma. A, Contrast CT shows fatty replacement of the thymus gland. Residual thymic tissue manifests as linear areas of soft tissue in anterior mediastinal fat (arrow). A, aorta; S, superior vena cava. B, Noncontrast CT 6 months after bone marrow transplantation shows increased soft tissue in the anterior mediastinum, consistent with thymic hyperplasia (arrow). A, aorta; S, superior vena cava.

A

B FIG 38-39  Normal thymus in a 20-year-old man. A and B, Contrast-enhanced PET/CT shows soft tissue (arrows) in the anterior mediastinum that conforms to the normal shape of the thymus gland. Note homogeneous FDG uptake in normal thymic tissue.

CHAPTER 38  Mediastinal Disease

S

A

FIG 38-40  Normal thymus in a 9-month-old child with coarctation of the aorta. Axial T1-weighted MRI shows the thymus gland (arrowheads) in the mediastinum, anterior to the aorta (A), with homogeneous intermediate signal intensity less than that of fat. Note the aortic coarctation (arrow). S, superior vena cava. (Case courtesy Donald Frush, MD, Duke University Medical Center, Durham, NC.)

T

T

A

B FIG 38-41  Thymic rebound. A, CT at the time of chemotherapy shows thymus (T) is small. B, CT after cessation of chemotherapy shows thymus (T) has increased in size.

A SVC

P

A

A

B FIG 38-42  Thymic hyperplasia in a young patient with chest pain. A, Contrast-enhanced CT demonstrates a prominent thymus without a discrete mass. A, aorta; P, pulmonary artery; SVC, superior vena cava. B, PET/ CT shows homogeneous FDG uptake within the thymus.

1039

1040

PART II  CT and MR Imaging of the Whole Body

* S TA T

A

A P

B

A

P

C FIG 38-43  Congenital thymic cyst. A, Contrast CT shows a homogeneous water-attenuation cyst (asterisk) in the anterior mediastinum. Note absence of perceptible walls and a residual right thymic lobe (arrow). S, superior vena cava; T, trachea; TA, transverse aorta. B and C, MRIs obtained more caudally show that the mass (arrowheads) is homogeneous and of low signal intensity on the T1-weighted image (B) and of high signal intensity on the T2-weighted image (C). This appearance is typical of a cyst. A, ascending aorta; P, main pulmonary artery.

CHAPTER 38  Mediastinal Disease

T

B

A

FIG 38-44  Seminoma associated with a thymic cyst. A, Contrast CT shows a well-circumscribed waterattenuation mass in the anterior mediastinum. Note the thin but perceptible wall (arrowheads). B, A more cephalic image reveals the soft tissue component of the mass (arrow). Resection revealed seminoma. T, trachea.

A

B

C

D FIG 38-45  Acquired multilocular thymic cyst in an asymptomatic 64-year-old man who presented with an anterior mediastinal mass noted on a preoperative chest radiograph. A and B, Axial CT images demonstrate a complex multicystic mass in the anterior mediastinum. C and D, Axial PET/CT shows increased FDG uptake within the cyst walls.

1041

1042

PART II  CT and MR Imaging of the Whole Body

A

B FIG 38-46  Thymoma in a patient with myasthenia gravis. A, Posteroanterior chest radiograph shows a mass in the right cardiophrenic angle (arrowheads). B, Contrast-enhanced CT confirms a well-circumscribed heterogeneous cardiophrenic angle mass (arrowheads). Resection revealed thymoma.

Thymomas are typically well-marginated smooth or lobulated mediastinal masses, 5 to 10 cm in diameter, that characteristically arise from one lobe of the thymus. They are usually located anterior to the aortic arch but can occur in the cardiophrenic angle (Fig. 38-46). Most lesions are homogeneous, but necrosis and hemorrhage occur in up to one third. On CT or MRI, thymomas most commonly manifest as smooth or lobulated masses that distort the normal contour of the thymus (Fig. 38-47). They are typically unilateral masses, although bilateral mediastinal involvement can occur. On CT they can manifest as homogeneous or heterogeneous soft tissue–attenuation masses. Intratumoral cysts or areas of necrosis may be seen. Enhancement following IV administration of contrast is variable. Calcification, seen in up to 7% of cases, is usually thin and linear and located in the capsule. On MRI, thymomas manifest with low to intermediate signal intensity (similar to skeletal muscle) on T1-weighted images and with high signal intensity on T2-weighted images. Because up to 33% of thymomas have focal areas of necrosis, hemorrhage, and cystic change, the masses may have heterogeneous signal intensity. MRI occasionally reveals fibrous septa within the masses. The role of MRI in evaluation of thymoma is evolving. Recent studies suggest that diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) values may be beneficial in differentiating low-grade thymomas from high-grade thymomas and thymic carcinomas. In addition to diagnosis, imaging plays an important role in thymoma staging. Patients with invasive thymoma (Masaoka-Koga stage III or greater) benefit from neoadjuvant chemotherapy. Imaging findings that suggest invasive thymoma on CT or MRI include (1) tumor size 7 cm or greater, (2) lobulated contours, (3) poorly defined or infiltrative margins, (3) definite vascular or chest wall invasion, (4) irregular interface with adjacent lung, and (5) evidence of spread to pleura (Figs. 38-48 to 38-50). Pleural spread manifests either as isolated pleural nodules (“drop metastases”) or as a contiguous pleural mass, often mimicking mesothelioma. Transdiaphragmatic spread may occur; hence, staging MRI or CT studies should include the upper abdomen. Pleural effusion is uncommon despite the often extensive

T

A

B FIG 38-47  Thymoma in a patient with myasthenia gravis. A, Axial chest CT shows a well-circumscribed homogeneous anterior mediastinal mass (T). B, PET/CT depicts heterogeneous FDG uptake. Resection revealed thymoma.

CHAPTER 38  Mediastinal Disease pleural metastases. Several CT and MRI findings are associated with a high rate of recurrence and metastasis as well as a poorer prognosis: tumor larger than 10 cm in diameter, with a lobulated or irregular contour; mediastinal fat or great vessel invasion; and pleural seeding. Typically, thymomas are FDG avid on PET/CT (see Fig. 38-47). Several recent studies have highlighted the potential use of FDG PET to differentiate noninvasive from invasive thymoma by the presence of confined uptake in the former and multiple discrete foci in the latter. Thymic carcinoma.  Thymic carcinomas account for 20% of thymic tumors. They are aggressive malignancies that often exhibit marked local invasion and early dissemination to regional lymph nodes and distant sites. Distant metastases (lung, liver, brain, bone) are detected in 50% to 65% of patients at presentation. Typical histologies

A

FIG 38-48  Invasive thymoma in a patient with symptoms of superior vena cava obstruction. Axial T1-weighted MRI shows an infiltrative soft tissue mass (arrowheads) of intermediate signal intensity encasing the aorta (A) and extending into the superior vena cava (arrow). (From Erasmus JJ, et al: MR imaging of mediastinal masses. Magn Reson Imaging Clin N Am 8:59–89, 2000.)

*

A

1043

include squamous cell carcinoma and lymphoepithelioma-like carcinoma. Symptoms (chest pain, dyspnea, cough, SVC syndrome) are usually due to compression or invasion of adjacent structures; unlike with thymomas, paraneoplastic syndromes are rare. Radiologically, thymic carcinomas commonly manifest as large, poorly marginated, anterior mediastinal masses frequently associated with intrathoracic lymphadenopathy and pleural and pericardial effusions. Focal pleural implants are uncommon. On CT they are usually of heterogeneous attenuation and have poorly defined infiltrative margins (Fig. 38-51). On MRI they typically have intermediate signal intensity (slightly higher than that of skeletal muscle) on T1-weighted images and high signal intensity on T2-weighted images. Signal intensity may be heterogeneous because of hemorrhage and necrosis within the masses. MRI can be helpful for revealing local soft tissue and vascular invasion. The presence of an irregular contour, necrotic or cystic component, heterogeneous enhancement, lymphadenopathy, or great vessel invasion on CT or MRI is strongly suggestive of   thymic carcinoma and frequently associated with a poor prognosis. Typically, thymic carcinoma exhibits significant FDG uptake on  PET/CT that is usually higher than that in thymoma and thymic hyperplasia (Fig. 38-52). Neuroendocrine tumors of the thymus.25,38,39,117,165,166,261,285  Neuroendocrine tumors are uncommon lesions of the anterior mediastinum. These tumors have a varying malignant potential that ranges from relatively benign (thymic carcinoid) to highly malignant (small cell carcinoma of the thymus). Thymic carcinoid tumor is the most common of this group of tumors. Affected patients are typically in  the fourth or fifth decade of life; there is a male predominance. Up to 50% of affected patients have hormonal abnormalities, and up to  35% have Cushing’s syndrome due to tumoral production of adrenocorticotropic hormone. Nonfunctioning thymic carcinoids may be seen in association with multiple endocrine neoplasia (MEN) syndrome type 1. Radiologically these tumors typically manifest as large masses with a propensity for local invasion. Focal areas of necrosis and punctate calcification may be present. On CT or MRI the masses are usually of heterogeneous attenuation or signal intensity, respectively (Fig. 38-53).

L

Ao

B FIG 38-49  Invasive thymoma. A and B, Contrast CT shows a heterogeneous anterior mediastinal mass with focal areas of calcification. Note that the mass extends between the superior vena cava (asterisk) and the ascending aorta (Ao); there is contiguous extension to the hemidiaphragm (arrowheads). L, liver.

1044

PART II  CT and MR Imaging of the Whole Body

A

B FIG 38-50  Invasive thymoma with pleural metastases. A, Contrast CT shows a lobular heterogeneous mass (arrowheads) in the anterior mediastinum. B, Noncontiguous pleural metastases (“drop metastases”; arrowheads) are present in the left hemithorax.

S

A

A FIG 38-51  Thymic carcinoma. Contrast CT shows a heterogeneous mass in the anterior mediastinum. Low-attenuation regions are consistent with necrosis. A, aorta; S, superior vena cava.

FDG uptake by carcinoid tumors is variable, with a tendency to be low. However, increased uptake has been reported to correlate with aggressive behavior such as local invasion and distal metastases.81,115

Germ Cell Tumors66,183,191,192,248,249,288,325,339,347 Germ cell tumors (teratomas, seminomas, embryonal carcinomas, endodermal sinus tumors, and choriocarcinomas) are thought to arise from mediastinal remnants of embryonal cell migration. They constitute 10% to 15% of anterior mediastinal tumors. The mediastinum is the most common extragonadal primary site, and mediastinal lesions account for 60% of all germ cell tumors in adults. Germ cell tumors usually occur in young adults (mean age, 27 years). Most malignant germ cell tumors (>90%) occur in men, whereas benign lesions (mature teratomas) occur with equal incidence in men and women.

B FIG 38-52  Thymic squamous cell carcinoma. A, Contrast CT in the axial plane shows an anterior mediastinal mass (arrow) with focal calcification. There is stranding of the mediastinal fat anterior to the mass, suggesting invasion of the mediastinal fat. B, PET/CT in the axial plane depicts marked FDG uptake within the mass (arrow).

CHAPTER 38  Mediastinal Disease

A S

1045

A

D

FIG 38-54  Mature teratoma. Contrast CT shows a large anterior mediFIG 38-53  Thymic carcinoid. Contrast CT shows a large anterior mediastinal mass. Note the heterogeneous appearance of the mass and infiltration of the mediastinum, with extension of the mass between vascular structures. A, aorta; S, superior vena cava.

astinal mass (arrowheads). The mass is heterogeneous in attenuation and is composed of fat, calcium, and fluid. This CT appearance is typical of a mature teratoma. A, ascending aorta; D, descending aorta.

Mediastinal teratoma.48,49,61,95,133,192,200,222,223,248,257,291,297,321,325  Teratomas, the most common mediastinal germ cell tumors, are composed of elements that arise from more than one of the three primitive germ cell layers. Teratomas may broadly be classified as benign or malignant. A histologic classification system for extragonadal germ cell tumors by Gonzalez-Crussi categorizes teratomas as mature, immature-benign, immature-malignant, and malignant. Moran and Suster proposed a modified classification as mature, immature, and teratoma with additional malignant components. Regardless of the system, classification is based upon tumor cellular composition. Mature or benign teratomas are composed of different tissue types (ectoderm, endoderm, mesoderm), with ectodermal derivatives predominating. The term dermoid cyst is commonly applied to tumors in which ectodermal components predominate. Mature teratomas are common, accounting for 70% of germ cell tumors in childhood and 60% of mediastinal germ cell tumors in adults. Mature teratomas occur most frequently in children and young adults. Patients may be asymptomatic, but chest pain, dyspnea, and cough due to compression of adjacent structures are common. By definition, patients with mature teratoma have normal serum levels of β-human chorionic gonadotropin hormone (β-hCG) and α-fetoprotein (AFP); elevation of these markers implies a malignant component. Complete resection is the treatment for teratomas and results in complete cure. Despite a benign histology, these tumors may be difficult to remove because they are adherent to local structures. Whereas mature teratomas are benign, immature teratomas are potentially malignant and are composed of more than 10% embryonic immature elements at histopathologic evaluation, with varying degrees of differentiation. Immature-malignant teratomas are those that would otherwise be classified histologically as benign but sub­ sequently become metastatic after diagnosis. Malignant teratomas  have immature histologic features but contain areas of malignancy (e.g., differentiated germ cell tumor, sarcoma, carcinoma), hence the nomenclature—immature teratoma with malignant components. These tumors are frequently metastatic at the time of diagnosis. Teratomas typically occur in the anterior mediastinum, although occurrence at other sites, including the middle mediastinum and posterior mediastinum, accounts for up to 20% of cases. Mature teratomas manifest on CT or MRI as smooth or lobulated mediastinal masses

A S

FIG 38-55  Mature teratoma. CT shows a large, predominantly waterattenuation anterior mediastinal mass with a small focal area of fat (arrow). Note the mass effect on the superior vena cava (S) and superior pulmonary vein (arrowheads). A, ascending aorta.

that typically have cystic and solid components; in contrast, malignant teratomas are usually poorly marginated masses containing areas of necrosis. The combination of fluid, soft tissue, calcium, and fat is diagnostic of teratoma (Fig. 38- 54). The finding of a fat-fluid level within a mass on CT or MRI is also diagnostic of teratoma. Fat occurs in up to 75% of mature teratomas and in up to 40% of malignant teratomas. However, only 17% to 39% of mature teratomas have all tissue components, and approximately 15% of mature teratomas manifest only a unilocular or multilocular cystic component (Fig. 38-55). Because of the varying composition of soft tissue, fat, calcium, and hemorrhage, MRI typically demonstrates heterogeneous signal intensity; this finding can be useful in differentiating teratomas from thymomas and lymphomas. Mature teratomas have been reported to rupture into the lung, pleural space, and pericardium in up to 33% of patients, and CT or MRI can be useful in detecting fat within these regions. The presence of nonhomogeneous internal components and other ancillary changes (e.g., adjacent parenchymal consolidation or atelectasis, pleural effusion, pericardial effusion) are signs of rupture

1046

PART II  CT and MR Imaging of the Whole Body Nonseminomatous germ cell malignancies.167,191,200,228,237  Non-

*

A

seminomatous germ cell tumors of the mediastinum include embryonal cell carcinoma, endodermal sinus tumor, choriocarcinoma, and mixed germ cell tumors, with the latter being the most common within this group. Most patients with malignant nonseminomatous germ cell tumors are symptomatic at presentation, with chest pain, cough, dyspnea, weight loss, and fever. Up to 80% of affected patients have elevated levels of AFP, and 54% have elevated levels of β-hCG. There is an association between malignant nonseminomatous germ cell tumors of the mediastinum and hematologic malignancies. Up to 20% of affected patients have Klinefelter’s syndrome. On CT or MRI these tumors manifest as large poorly marginated masses of heterogeneous attenuation or signal intensity (Fig. 38-59). Invasion of the adjacent mediastinal structures, chest wall, and lung, as well as metastases to the regional lymph nodes and distant sites, is common. FDG PET is useful in staging and for detection of recurrent or progressive disease, which usually appears as areas of increased uptake.

Parathyroid Adenoma8,41,82,141-143,151,160,171,187,188, 251,299,319,322

B FIG 38-56  Ruptured teratoma. A, Contrast-enhanced CT shows a heterogeneous anterior mediastinal mass (arrow). Note areas of enhancement and fat attenuation. There is stranding of the mediastinal fat (asterisk) and a small left-sided pleural effusion (arrowhead). B, Chest CT, lung window, demonstrates ground-glass opacities (arrow) surrounding the mass, suggesting rupture. (Case courtesy S.H. Paik, MD, Seoul, Republic of Korea.)

(Fig. 38-56). Malignant teratomas can be difficult to differentiate from benign mature teratoma at imaging unless invasion or metastatic disease is present (Fig. 38-57). FDG PET is of limited utility in evaluating mature teratoma (owing to the tumor’s lack of FDG avidity) but may be beneficial in identifying malignant immature components. Seminoma.10,183,200,228,230,310,325  Seminoma is the most common pure histologic type of malignant mediastinal germ cell tumor in men and accounts for 40% of such tumors. Affected patients are usually in the third or fourth decade of life and are symptomatic at presentation. Levels of β-hCG and AFP are normal; elevation of AFP indicates a nonseminomatous component of the tumor. On CT or MRI, seminomas manifest as large lobulated masses of homogeneous attenuation or signal intensity in the anterior mediastinum (Fig. 38-58). Cysts or areas of necrosis may also be seen in association with mediastinal seminoma. Invasion of adjacent structures is uncommon and calcification is rare. Metastases to regional nodes may occur. FDG PET is useful in staging and for detection of recurrent or progressive disease, usually seen as areas of increased uptake (see Fig. 38-58).

Because neck exploration for parathyroid gland removal is curative in more than 90% of patients with primary hyperparathyroidism, surgeons often do not obtain preoperative imaging studies to localize the parathyroid glands. However, 10% of parathyroid adenomas are ectopic in location, and the majority of ectopic adenomas are located either in the region of the tracheoesophageal groove or in the anterior mediastinum. Such ectopic adenomas can be missed at surgical exploration. Preoperative localization of ectopic parathyroid adenomas can reduce operative time, postoperative morbidity, and the need for repeat surgery. Imaging techniques for localizing abnormal parathyroid glands include sonography, radionuclide imaging (technetium 99m methoxyisobutylisonitrile [99mTc MIBI], 99mTc tetrofosmin), CT, and MRI. On CT, ectopic mediastinal parathyroid glands manifest as 1- to 2-cm rounded masses that may resemble lymph nodes (Fig. 38-60). CT or MRI can detect these glands, but careful evaluation is required so that small abnormalities are not overlooked. Whereas FDG does not accumulate in parathyroid adenomas, they are frequently avid for 11C methionine. MRI is accurate in identifying abnormal parathyroid glands in ectopic locations. Functioning parathyroid adenomas are intermediate in signal intensity on T1-weighted images and usually have markedly increased signal intensity on T2-weighted images (Fig. 38-61). Similar findings are seen in cases of parathyroid hyperplasia and carcinoma. However, up to 13% of abnormal glands do not have high signal intensity on T2-weighted images, owing to fibrosis or hemorrhage. Although MRI is comparable or superior to other imaging modalities for detecting parathyroid pathology (sensitivity, 78%; specificity, 90%; accuracy, 90%), it is most appropriate as an adjunct to 99mTc MIBI radionuclide imaging. In this setting, MRI provides accurate anatomic localization of the adenoma and can be useful in predicting the need for mediastinotomy or a lateral cervical incision. Although the precise role of imaging in the evaluation of a virgin neck for hyperparathyroidism may be debated, imaging is useful in second-look procedures and in those patients considered to be at high risk for surgery. In this population, the success rate for surgery is lower, ranging from 64% to 90%, and the combined use of 99mTc MIBI and MRI has been shown to be 89% sensitive and 95% specific for preoperative localization.

Lymphatic Malformations59,116,156,243,264,303,309,312,359 Lymphatic malformations are rare benign congenital lesions that account for 0.7% to 4.5% of all mediastinal masses. Approximately 75% occur in the neck, 20% in the axillary region, and 5% in the

CHAPTER 38  Mediastinal Disease

1047

* M

S

A

B FIG 38-57  Malignant teratoma. A, Contrast-enhanced CT shows a large heterogeneous mediastinal mass (M) with areas of fluid (asterisk) and fat attenuation (arrow). The superior vena cava (S) is compressed and displaced. B, More inferiorly the mass contains marginal enhancing soft tissue (arrows) that invades and disseminates within the pericardium (arrowheads). Histopathology confirmed a high-grade immature teratoma with additional malignant components of embryonal carcinoma. S, superior vena cava.

M

A

mediastinum. Primary mediastinal lymphatic malformations are rare, and most mediastinal lesions are due to extension from the neck. Lymphatic malformations are usually located in the superior mediastinum and anterior mediastinum, although they can occur in any mediastinal compartment. They usually occur in patients younger than 2 years, and there is a male predominance. In adults, lymphatic malformations are usually located in the mediastinum and are often due to recurrence of an incompletely resected childhood lesion that increases in size as lymphatic channels dilate with reaccumulated lymph. Lymphatic malformations are currently classified as macrocystic or microcystic, replacing the corresponding outdated terms cystic hygroma and lymphangioma, respectively. On CT these lesions manifest as lobular multicystic tumors that surround and infiltrate adjacent mediastinal structures (Fig. 38-62). They can appear solid on CT because of protein or hemorrhage  within the cysts. The cysts are typically 1 to 2 cm in diameter, and the septa may enhance following administration of IV contrast. MRI  can be useful to confirm the cystic nature of these lesions; lymphatic malformations usually have markedly increased signal intensity on T2-weighted images. Their appearance on T1-weighted sequences is more variable, although most are generally of low to intermediate signal intensity (similar to skeletal muscle) and can contain focal areas of signal intensity greater than that of muscle. Occasionally, lymphatic malformations have high signal intensity similar to that of fat on T1-weighted images. Because surgical resection is the treatment of choice, multiplanar MRI can be useful in preoperative evaluation of local invasion and determination of the anatomic extent of tumor.

Hemangioma144,152,213,295,301,302

B FIG 38-58  Seminoma. A, CT shows an anterior mediastinal mass (M) with several small central areas of low attenuation, consistent with necrosis. B, PET/CT in the coronal plane demonstrates marked FDG activity. Note relationships to adjacent mediastinal structures.

Hemangiomas are rare mediastinal tumors that account for less than 0.5% of all mediastinal masses. Mediastinal hemangiomas usually occur in the anterior (68%) or posterior (22%) mediastinum, although multicompartment involvement is present in up to 14% of cases. Most mediastinal lesions are cavernous hemangiomas and are composed of large interconnecting vascular spaces with varying amounts of interposed stromal elements such as fat and fibrous tissue. Focal areas of organized thrombus can calcify as phleboliths. On radiographs, hemangiomas appear as sharp, smoothly marginated mediastinal masses. Phleboliths are seen in less than 10% of cases.

1048

PART II  CT and MR Imaging of the Whole Body

*

B

A

FIG 38-59  Nonseminomatous germ cell malignancy (yolk sac tumor) that presented with superior vena cava syndrome. A, CT shows a large mixed-attenuation anterior mediastinal mass. There is tumoral invasion of the superior vena cava (arrow). B, PET/CT shows marked FDG activity, with several photopenic areas consistent with necrosis. Note FDG activity within the superior vena cava thrombus (asterisk).

A

B FIG 38-60  Ectopic parathyroid adenoma. A, Axial noncontrast CT shows a small soft tissue nodule (arrow) within the anterior mediastinum in a patient with persistently elevated serum calcium levels after undergoing neck dissection for hyperparathyroidism. This was thought to represent a small lymph node; however, the clinical history prompted further evaluation. B, 99mTc MIBI scan was positive (arrow), with pathology confirming an ectopic parathyroid adenoma.

CHAPTER 38  Mediastinal Disease

T

1049

T

V V

A

B FIG 38-61  Ectopic parathyroid adenoma in a woman with persistent hyperparathyroidism after surgical neck exploration. T1- (A) and T2-weighted (B) MRIs show a 1-cm mass (arrowheads) in the superior mediastinum, with intermediate signal intensity (similar to muscle) on the T1-weighted image and high signal intensity on the T2-weighted image. This appearance is typical of parathyroid adenoma. T, trachea; V, vertebral body. (From Erasmus JJ, et al: MR imaging of mediastinal masses. Magn Reson Imaging Clin N Am 8:59–89, 2000.)

S T

A

B FIG 38-62  Lymphatic malformation. A and B, Contrast CT shows a homogeneous water-attenuation mass (arrows) in the superior mediastinum that arises in the neck. The mass diffusely infiltrates the mediastinum and extends into the right hilum (B). S, superior vena cava; T, trachea.

CT typically reveals a heterogeneous mass with intense central and peripheral rimlike enhancement after administration of IV contrast (Fig. 38-63). Hemangiomas typically have heterogeneous signal intensity on T1-weighted images. In lesions with significant stromal fat, linear areas of increased signal intensity on T1-weighted images can occasionally be identified. The central vascular lakes typically become markedly hyperintense on T2-weighted images, a potentially diagnostic feature (Fig. 38-64).

Middle Mediastinal Abnormalities Lesions that occur primarily in the middle mediastinum include esophageal lesions, airway lesions, foregut duplication cysts, and pericardial cysts.

Esophageal Lesions Esophageal cancer.60,96,119,134,162,184,185,190,195,234,263,269,279,282,334,341,349  CT is used to help stage esophageal carcinoma by demonstrating the length

of the lesion, thickness of the esophageal wall, involvement of adjacent structures (e.g., airway, aorta, pericardium, spine), and nodal spread to regional mediastinal, celiac, and gastrohepatic ligament nodes  (Fig. 38-65). Endoscopy, endoscopic ultrasonography, and esophagography are also helpful in staging esophageal carcinoma. Invasion of local structures is common because the esophagus lacks a serosa. CT and MRI can over- or underestimate the degree of local invasion, because the lack of normal fat planes between the esophagus and adjacent structures makes invasion difficult to determine. Anterior displacement of the carina or left main bronchus suggests airway invasion; contact with the aorta that exceeds one quarter of the aortic circumference suggests aortic invasion. All imaging findings must be interpreted in conjunction with the patient’s clinical status when planning treatment. CT is also useful in planning radiation treatment ports and for imaging complications resulting from esophagectomy. Although esophageal malignancies can occasionally be detected by conventional chest CT after IV contrast (venous phase), acquisition of

1050

PART II  CT and MR Imaging of the Whole Body

A

A

A

B

FIG 38-63  Hemangioma. A, Noncontrast CT shows a small homogeneous anterior mediastinal mass (arrowheads). A, ascending aorta. B, Contrast CT shows marked enhancement, typical of hemangioma (arrowheads). A, ascending aorta.

images during an earlier arterial phase can increase the sensitivity of detection.341 Esophageal cancer is usually FDG avid, with the exception of small and early T-stage tumors. Importantly, the detection of nodal and distant metastases by FDG PET increases the accuracy of staging and, by identifying unsuspected metastases, improves patient management (see Fig. 38-65). The role of FDG PET in evaluating the primary tumor’s response to neoadjuvant therapy in patients with esophageal cancer is currently being evaluated. Although initially promising results have been reported, the presence of therapy-induced inflammation and ulceration can lead to false-positive results in patients with a complete response to therapy. Additionally, because of the inability to detect residual microscopic disease, a negative PET scan does not preclude surgical resection. The normal esophagus is not optimally demonstrated on MRI because of peristaltic motion and relatively poor spatial resolution. Distinguishing small mucosal tumors from normal esophagus is difficult because both have intermediate signal intensity on T1-weighted images and high signal intensity on T2-weighted images. The role of MRI for intrathoracic staging of esophageal neoplasms is limited, although some early reports on new techniques (e.g., surface coil and endoscopic MRI) have shown potential benefits. Some authors have reported high intrathoracic staging accuracy for MRI, but most reports suggest that MRI is no better than CT in this regard. More recently, combined PET/MR has been found to improve staging accuracy over MRI alone, with reported rates similar to endoscopic ultrasound, though the role of this modality remains to be validated. At present the primary utility of MRI in the evaluation of esophageal cancer is as a problem-solving modality, such as when invasion of the pericardium or heart is suspected on CT or when the patient is allergic to iodinated contrast material.

Esophageal dilatation.68,270  Esophageal dilatation can be easily demonstrated on CT. Focal dilatation occurs in Zenker’s diverticulum (upper esophagus), traction diverticula due to granulomatous disease (middle esophagus), and epiphrenic diverticula (lower right esophagus) (Fig. 38-66). These focal dilatations or diverticula are clearly demonstrated on CT and can be a source of confusion unless oral contrast material is administered. Diffuse dilatation of the esophagus can occur as a result of motility disorders (achalasia, postvagotomy syndrome, Chagas disease, scleroderma, systemic lupus erythematosus, presbyesophagus, diabetic neuropathy, esophagitis) or as a result of distal obstruction (carcinoma, stricture, extrinsic compression) (Fig. 38-67).

Airway Lesions11,28,47,163,176,206,207,214,216,256,260,285 Tumors of the trachea or proximal bronchi can manifest as middle mediastinal masses on chest radiographs or CT. Malignant neoplasms account for 90% of primary tracheal tumors. The most common primary malignancies of the trachea are squamous cell and adenoid cystic carcinoma. Benign neoplasms such as papilloma, adenoma, hamartoma, chondroma, leiomyoma, and granular cell myoblastoma account for less than 10% of primary tracheal tumors. The most common primary malignancy of the major bronchi is non–small cell lung cancer. Carcinoid tumors and mucoepidermoid carcinoma are less common bronchial malignancies that occur in younger patients. These lesions are discussed more fully in Chapter 37.

Foregut Cysts313,314 Foregut cysts, which account for approximately 20% of all mediastinal masses, most likely arise as a result of maldevelopment of the primitive foregut. Bronchogenic cysts are the most common mediastinal foregut cysts; esophageal duplication and neurenteric cysts are less common. It is important not to confuse these cystic masses with fluid-filled

CHAPTER 38  Mediastinal Disease

b T

1051

*

A A T

B

C

D

E

H

B

b

*

T

C FIG 38-64  Hemangioma. A, Contrast CT shows a heterogeneous mass (arrowheads) in the anterior mediastinum. Note punctate calcification, consistent with a phlebolith within the mass (arrow). Asterisk, carotid artery; b, brachiocephalic artery; T, trachea. B, Coronal T1-weighted MRI shows a mass of intermediate signal intensity (arrowheads) in the anterior and superior mediastinum. Note septa within the mass and invasion of the mediastinum. H, heart; T, trachea. C, Axial T2-weighted MRI shows high signal intensity within the mass (arrowheads). This appearance is typical for hemangioma. Asterisk, carotid artery; b, brachiocephalic artery; T, trachea.

FIG 38-65  Esophageal malignancy. A, Noncontrast CT shows concentric thickening of the distal esophagus (arrow). B and C, Enlarged lymph nodes are depicted in the superior mediastinum (arrows). D and E, PET/ CT, coronal fusion, demonstrates avid uptake within the tumor and the mediastinal lymph nodes, consistent with metastases.

1052

PART II  CT and MR Imaging of the Whole Body

structures within the mediastinum such as fluid-filled pericardial recesses, loculated pleural fluid, extension of ascites through the esophageal hiatus, mediastinal extension of pancreatic pseudocyst, or a fluidfilled and dilated esophagus (Figs. 38-68 and 38-69). Bronchogenic cyst.14,30,59,62,114,154,180,181,189,195,197,201,211,236,239,241,326  Bronchogenic cysts are thought to arise from abnormal budding of the

FIG 38-66  Epiphrenic diverticulum. CT after oral administration of barium demonstrates a lateral pouch (arrow) partially filled with contrast medium. The diverticulum was surgically removed.

ventral foregut. Histopathologically they are lined by ciliated respiratory epithelium, cartilage, smooth muscle, fibrous tissue, and mucous glands. Most bronchogenic cysts arise in the middle mediastinum, typically in the subcarinal or right paratracheal region, but may also occur in the anterior or posterior mediastinum. Bronchogenic cysts are found in unusual locations in up to 15% of cases, including the pleural space, diaphragm, pericardium, lung, or abdomen. Bronchogenic cysts may also rarely be pedunculated and can be mobile within the pleural space. On CT, bronchogenic cysts typically manifest as round or spherical, sharply marginated, homogeneous masses. Approximately half of bronchogenic cysts are of water attenuation on CT (Fig. 38-70). The remainder have increased attenuation, likely secondary to proteinaceous debris or hemorrhage within the lesions (Fig. 38-71). A small percentage have calcified walls or contain calcium suspended within the fluid. MRI shows lesions that are typically of low to intermediate signal intensity on T1-weighted images and of markedly increased signal intensity on T2-weighted images. The lesions can be heterogeneous on T1-weighted images but are typically homogeneous on T2-weighted images. These MRI characteristics can be useful for differentiating cysts that appear solid on CT from solid neoplasms or lymphadenopathy. Gadolinium administration can also help distinguish mediastinal cysts from solid neoplasms by demonstrating lack of central enhancement and occasionally rim enhancement. Because many cardiac-gated T1-weighted pulse sequences have  a relatively long TR, fluid-filled lesions can be of intermediate  signal intensity. Cysts with proteinaceous, mucinous, or hemorrhagic

B

A

C

FIG 38-67  Achalasia. A, Barium swallow demonstrates marked dilatation of the esophagus. B and C, CT shows a dilated esophagus and abundant intraluminal debris. Note anterior displacement of the trachea.

CHAPTER 38  Mediastinal Disease contents can have a further increase in signal intensity on these sequences (Fig. 38-72). However, these lesions usually have markedly increased signal intensity on T2-weighted images, suggesting their true cystic nature. Bronchogenic cysts do not exhibit FDG uptake on PET (see  Fig. 38-71).

attenuation due to intracystic hemorrhage or proteinaceous debris.  On MRI they have signal intensity characteristics similar to those of bronchogenic cysts: variable signal intensity on T1-weighted images (depending on the intracystic contents) and markedly increased signal intensity on T2-weighted images (Fig. 38-74). Neurenteric cysts result from incomplete separation of the endoderm and notochord, resulting in a diverticulum of the endoderm. Neurenteric cysts, which are pathologically identical to esophageal duplication cysts, may have either a fibrous connection to the spine or an intraspinal component. These cysts are typically associated with vertebral body anomalies, most commonly a sagittal cleft, occurring at or above the level of the cyst. Most neurenteric cysts occur in the posterior rather than middle mediastinum, usually above the level of the carina. The CT and MRI appearance of these lesions is similar to that of other foregut cysts. MRI is optimal for demonstrating the extent of spinal abnormality and the degree of intraspinal involvement.

Esophageal duplication and neurenteric cysts.88,164,180,236,313,326 

Duplication of the esophagus is thought to represent a diverticulum of the primitive foregut or an aberrant recanalization of the gut in embryogenesis. These cysts are located in the middle or posterior mediastinum and may be indistinguishable from bronchogenic cysts. They are composed of a muscular coat and mucosa that may resemble the esophagus, stomach, or small intestine, although it is usually ciliated. Esophageal duplication cysts usually occur within the wall of the esophagus or are adherent to it. They are either spherical or tubular and are usually located distally along the lateral aspect of the esophagus. On CT they typically manifest as spherical or tubular masses near the esophageal wall. They are usually homogeneous and of water attenuation (Fig. 38-73). Like bronchogenic cysts, they may be of soft tissue

Pericardial Cyst15,85,235,236,273,275,280,296,326 Pericardial cysts are unilocular mesothelium-lined cysts that arise from congenital defects related to the ventral and parietal pericardial recesses. True pericardial cysts contain all layers of the pericardium and do not communicate with the pericardial space. They are usually found in asymptomatic adults, often discovered incidentally on chest radiographs. They are typically unilocular cystic lesions with clear fluid contents and thin walls. Pericardial cysts are variable in size and shape and most commonly occur in the right (70%) or left (22%) cardiophrenic angle. CT shows a homogeneous, nonenhancing, waterattenuation, rounded mass adjacent to the pericardium (Fig. 38-75). The cyst wall may calcify. Pericardial cysts are usually homogeneous on MRI, with low signal intensity on T1-weighted images and markedly increased signal intensity on T2-weighted sequences (Fig. 38-76).

A

*

1053

* D

Posterior Mediastinal Lesions157 Neurogenic Tumors12,62,195,242,326,356

Neurogenic tumors account for 20% of all adult and 35% of all pediatric mediastinal neoplasms. Most neurogenic tumors occur within  the posterior mediastinum, and neurogenic tumors account for  the majority of posterior mediastinal neoplasms. These lesions can  be classified as tumors of the peripheral nerves (neurofibromas, schwannomas, malignant tumors of nerve sheath origin), sympathetic

FIG 38-68  Thoracic lymphocele after esophageal resection with posterior mediastinal gastric conduit. Contrast CT shows a large homogeneous fluid collection (arrowhead) in the middle-posterior mediastinum, displacing the tracheal carina (asterisk) anteriorly and the gastric conduit anterolaterally. A, ascending aorta; D, descending aorta.

H

C

A

A

B FIG 38-69  Mediastinal pseudocyst in a patient with chronic pancreatitis. A, Contrast CT shows a homogeneous water-attenuation mass (C) in the middle mediastinum, with a thin enhancing rim (arrowheads). Note small left pleural effusion. A, aorta; H, heart. B, More caudal CT image confirms the origination of the mass (arrowheads) from the tail of the pancreas. (Case courtesy May Lesar, Bethesda, MD.)

1054

PART II  CT and MR Imaging of the Whole Body

A

P A

P

T

A

B FIG 38-70  Bronchogenic cyst. A, Contrast-enhanced CT shows a well-circumscribed water-attenuation middle mediastinal mass (arrow) with an imperceptible wall. The mass is adjacent to the trachea (T) and aorta (A). B, CT coronal reformation depicts the relationship of the mass (arrow) with adjacent structures, such as the superior vena cava (arrowhead in A and B), aorta (A), and pulmonary arteries (P).

BC

M

A

B FIG 38-71  Bronchogenic cyst. A, Contrast-enhanced CT shows a large homogeneous mass (M) of soft tissue attenuation in the subcarinal region. B, PET/CT was performed, which shows no FDG uptake within the lesion (BC), a characteristic feature of bronchogenic cysts. BC, bronchogenic cyst.

CHAPTER 38  Mediastinal Disease

1055

T

A

FIG 38-72  Bronchogenic cyst. Coronal T1-weighted MRI shows a subcarinal mass with high signal intensity due to viscous cyst fluid. Note displacement of both main bronchi. A subcarinal location is typical for bronchogenic cysts. T, trachea.

*

FIG 38-74  Esophageal duplication cyst. Axial T1-weighted MRI shows a well-circumscribed mass (arrow) with high signal intensity adjacent to the descending aorta (A). This appearance, atypical for a simple cyst, is due to proteinaceous fluid within the cyst.

C A

FIG 38-73  Esophageal duplication cyst. Noncontrast CT shows a wellmarginated low-attenuation middle mediastinal mass (arrow) adjacent to or possibly contiguous with the right lateral wall of the esophagus (asterisk). A, descending aorta.

ganglia (ganglioneuromas, ganglioneuroblastomas, neuroblastomas), or parasympathetic ganglia (paraganglioma, pheochromocytoma). Peripheral nerve tumors are more common in adults, and sympathetic ganglia tumors are more common in children. CT is usually performed in the initial evaluation of a suspected neurogenic tumor and is helpful in identifying intratumoral calcification and assessing associated bone erosion or destruction. Because neurogenic tumors usually arise in a paravertebral location, intraspinal extension is common. MRI is the preferred imaging modality for evaluating neurogenic tumors; it allows simultaneous assessment of intraspinal extension, spinal cord abnormalities, longitudinal extent of tumor, and extradural extension. Although the various types of neurogenic tumors can have similar radiologic appearances, there are certain features that aid in the differential diagnosis. Peripheral nerve tumors typically manifest in

FIG 38-75  Pericardial cyst. Contrast CT shows a large, homogeneous, water-attenuation mass (C) in the right cardiophrenic angle. Note the imperceptible wall; enhancement at the periphery (arrows) is due to an atelectatic lung. The location and appearance are typical of pericardial cysts.

adults as round masses oriented along the axis of a peripheral nerve, and there may be an intraspinal component (“dumbbell” lesion). Tumors of the sympathetic ganglia typically manifest in children as oval masses that are elongated along the spine and may contain calcifications on CT. Benign peripheral nerve tumors.*  Benign peripheral nerve tumors (schwannomas, neurofibromas) are slow-growing neoplasms and the most common neurogenic tumors of the mediastinum. Schwannomas are encapsulated neoplasms that arise from the nerve *References 9, 27, 40, 89, 97, 122, 174, 195, 199, 242, 278, 289, 326, 332, 356.

1056

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 38-76  Pericardial cyst. A, CT shows a small, low-attenuation, well-circumscribed mass in the right cardiophrenic angle (arrow). B and C, MRIs show that the mass is homogeneous, with low signal intensity on the T1-weighted image (B) and high signal intensity on the T2-weighted image (C). This MR appearance is typical of pericardial cysts.

sheath and typically have areas of cystic degeneration, hemorrhage, and small focal areas of calcification. Schwannomas grow lateral to the parent nerve and cause symptoms by compressing the nerve. Neurofibromas differ from schwannomas in that they are unencapsulated and result from proliferation of all nerve elements, including Schwann cells, nerve fibers, and fibroblasts. They grow by diffusely expanding the parent nerve. This type of neural tumor is found in neurofibromatosis (NF)1 (von Recklinghausen’s disease). Plexiform neurofibromas are variants that infiltrate along nerve trunks or plexuses. Schwannomas and neurofibromas typically occur with equal frequency in men and women, most commonly in the third and fourth decades of life. Between 30% and 45% of neurofibromas occur in patients with NF. Multiple neurogenic tumors or a single plexiform neurofibroma is pathognomonic of the disease. On CT, schwannomas and neurofibromas appear as sharply marginated, unilateral, spherical or lobular posterior mediastinal masses (Fig. 38-77). Pressure erosion of adjacent ribs or vertebral bodies or enlargement of the neural foramen occurs in up to 50% of cases. Punctate intralesional calcification occurs occasionally. On NCE CT, schwannomas are often of lower attenuation than skeletal muscle owing to their high lipid content, interstitial fluid, and areas of cystic

degeneration. Neurofibromas are often more homogeneous and of higher attenuation than schwannomas because they have fewer of these histologic features. These lesions may heterogeneously enhance following administration of IV contrast. In several small series, schwannomas have been reported to be FDG avid on PET imaging. Neurofibromas can also be FDG avid. Recent studies have reported that malignant degeneration is usually associated with increased FDG uptake. On MRI, schwannomas and neurofibromas are of variable signal intensity on T1-weighted images but typically have similar signal intensity to the spinal cord. On T2-weighted images these neoplasms characteristically have high signal intensity peripherally and low signal intensity centrally (target sign) owing to collagen deposition. This feature, when present, helps distinguish neurofibromas from other mediastinal tumors. Also, areas of cystic degeneration within the lesions may result in foci of increased signal intensity on T2-weighted images. Although the high signal intensity of schwannomas and neurofibromas on T2-weighted images can facilitate differentiation of tumors from spinal cord, the tumors may be obscured by the high signal intensity of cerebrospinal fluid. Schwannomas and neurofibromas, however, enhance with gadolinium, and this feature can be useful in detecting and determining the degree of intradural

CHAPTER 38  Mediastinal Disease

A

B

C

D

1057

FIG 38-77  Schwannoma. A, Contrast-enhanced axial CT shows a large heterogeneous paraspinal mass (arrow). The rounded shape is typical of a peripheral nerve tumor. B, Coronal T1-weighted MRI shows that the mass (arrow) is homogeneous and isointense to skeletal muscle. C, Coronal T2-weighted MRI shows that the mass (arrow) is heterogeneously hyperintense, with internal septations. D, Coronal contrastenhanced T1-weighted MRI shows heterogeneous enhancement within the mass (arrow). Areas remaining hypointense represent focal degeneration.

extension of these tumors (Fig. 38-78). Ten percent of paravertebral neurofibromas and schwannomas extend into the spinal canal and appear as dumbbell-shaped masses, with widening of the affected neural foramen. CT of plexiform neurofibromas demonstrates low-attenuation infiltrative masses along the mediastinal nerves and sympathetic chains, which may occur in any mediastinal compartment (Fig. 38-79). MRI of plexiform neurofibromas also demonstrates the infiltrative nature of the tumors, and the masses have low signal intensity on both T1- and T2-weighted images owing to their fibrous nature.

Malignant tumor of nerve sheath origin.*  Malignant tumors of nerve sheath origin (also termed malignant neurofibromas, malignant schwannomas, or neurofibrosarcomas) are rare neoplasms that typically develop from solitary or plexiform neurofibromas in the third to fifth decades of life. Up to 50% occur in patients with NF1; in such patients, these tumors occur at an earlier age (typically during adolescence) and with a higher incidence than among the general population. Because *References 9, 27, 36, 72, 104, 135, 174, 199, 204, 242, 258, 326, 332.

1058

PART II  CT and MR Imaging of the Whole Body

V

V

A

B FIG 38-78  Neurofibroma. A, Axial T1-weighted MRI shows a lobular mass (arrowheads) extending into the intervertebral foramina. V, vertebral body. B, Axial gadolinium-enhanced MRI shows homogeneous enhancement within the mass (arrowheads). V, vertebral body.

FIG 38-79  Plexiform neurofibromas in a woman with type I neurofibromatosis. Coronal noncontrast CT shows plexiform neurofibromas along the course of nerves within the posterior mediastinum (arrowheads).

most benign neurogenic tumors are asymptomatic, the development of pain often indicates malignant transformation. On CT or MRI, malignant tumors of nerve sheath origin typically manifest as posterior mediastinal masses larger than 5 cm in diameter. Although benign and malignant neurogenic tumors cannot be differentiated with certainty, findings that suggest malignancy include a sudden change in size of a preexisting mass or the development of heterogeneous signal intensity on MRI (caused by necrosis and

hemorrhage). The presence of multiple target signs throughout the lesion on MRI favors the diagnosis of a plexiform neurofibroma rather than a malignant tumor of nerve sheath origin. Neurofibrosarcomas are often FDG avid, and when the SUV is greater than 3, shorter mean survival time has been reported. Sympathetic ganglia tumors.4,19,46,63,97,242,293,308,320,326,332,344  Sympathetic ganglia tumors (ganglioneuromas, ganglioneuroblastomas, neuroblastomas) are rare neoplasms that originate from nerve cells. Ganglioneuromas and ganglioneuroblastomas usually arise from the sympathetic ganglia in the posterior mediastinum. Ganglioneuromas are benign neoplasms that usually occur in children and young adults. Ganglioneuroblastomas, which exhibit varying degrees of malignancy, usually occur in patients younger than 10 years. The posterior mediastinum is also the most common extraabdominal location of neuroblastomas, with up to 30% of these tumors occurring in this region. Neuroblastomas are highly malignant tumors that typically occur in children younger than 5 years. A posterior mediastinal mass in this age group should be considered a neuroblastoma until proved otherwise. Radiologically, ganglioneuromas and ganglioneuroblastomas usually manifest as well-marginated, elliptical, posterior mediastinal masses that extend vertically over three to five vertebral bodies. They are usually located lateral to the spine and may cause pressure erosion on adjacent vertebral bodies. On CT they are typically heterogeneous and may contain stippled or punctate calcification (Fig. 38-80). On T1- and T2-weighted MRIs they are usually homogeneous and of intermediate signal intensity (Fig. 38-81). Occasionally these lesions are heterogeneous and of high signal intensity on T2-weighted images. Ganglioneuroblastomas are typically larger and more aggressive than ganglioneuromas, with evidence of local and intraspinal invasion. On CT, neuroblastomas manifest as paraspinal masses of heterogeneous predominantly soft tissue attenuation. The lesions usually contain areas of hemorrhage, necrosis, cystic generation, and calcium (30%). On MRI the lesions typically demonstrate heterogeneous signal intensity on all pulse sequences and show heterogeneous enhancement following gadolinium administration. Neuroblastomas often exhibit widespread local invasion and have irregular margins, although many of these lesions are well marginated on CT or MRI. Neuroblastomas also have a tendency to cross the midline.

CHAPTER 38  Mediastinal Disease

1059

LA

FIG 38-80  Ganglioneuroma. Contrast CT shows a well-circumscribed low-attenuation paraspinal mass containing punctate calcification (arrows). LA, left atrium.

FIG 38-82  Paraganglioma. Contrast CT shows a well-defined round mass occupying the aortopulmonary window (arrow). Note the marked enhancement.

(Fig. 38-83). Flow voids in the lesion are sometimes seen, indicating their hypervascular nature. The lesions may be well circumscribed or show invasion of surrounding mediastinal structures. The role of PET in paraganglioma imaging is not well established, owing to lack of specificity and variable FDG uptake.

Lateral Thoracic Meningocele9,140,240,242,272,315

L S

FIG 38-81  Ganglioneuroma. Coronal T1-weighted MRI shows an elliptical paraspinal mass (arrowheads) of intermediate signal intensity. The vertical orientation and signal characteristics are typical of sympathetic ganglia tumors. L, liver; S, spleen.

Parasympathetic

ganglion

tumors.93,227,242,333  Paragangliomas

(chemodectomas) are rare neural tumors of the extraadrenal parasympathetic system. These tumors can be histologically identical to pheochromocytomas and can be functional or nonfunctional. Mediastinal paragangliomas occur in one of two locations: (1) in the middle mediastinum in close association with the origins of the aorta and pulmonary artery (aorticopulmonary paraganglioma) or (2) in the posterior mediastinum (paravertebral paraganglioma). On NCE CT the lesions are typically heterogeneous and enhance intensely after administration of IV contrast (Fig. 38- 82). On MRI the lesions are usually hypointense on T1-weighted images and hyperintense on T2-weighted images

Because of the pressure difference between the thorax and subarachnoid space, pulsion diverticula called lateral meningoceles can develop and protrude through the adjacent neural foramina. They occur most commonly in patients with NF1. Lateral meningoceles manifest radiographically as well-circumscribed paravertebral masses that usually occur on the convex side of a scoliosis. On CT, meningoceles appear as well-circumscribed paraspinal masses of water attenuation. The adjacent neural foramen is typically enlarged. These lesions can be confused with low-attenuation neurofibromas. The diagnosis is established by demonstrating communication with the subarachnoid space by CT myelography. MRI can also be helpful in differentiating meningoceles from neurofibromas, because meningoceles typically have low signal intensity on T1-weighted images and high signal intensity on T2-weighted images and do not enhance after gadolinium administration. Cardiac-gated MR cine images of meningoceles can reveal pulsatile motion owing to communication with the subarachnoid space.

Extramedullary Hematopoiesis33,107,118,156,223,242,300 Extramedullary hematopoiesis is a compensatory phenomenon that occurs when erythrocyte production is diminished or destruction is accelerated. Extramedullary hematopoiesis is usually seen in patients with chronic hemolytic disorders such as thalassemia, hereditary 

1060

PART II  CT and MR Imaging of the Whole Body

T

A

A

D

A

B FIG 38-83  Paraganglioma. A, CT shows a homogeneous soft tissue mass in the aortopulmonary window (arrowheads). Note that the mass is difficult to distinguish from the aorta. A, ascending aorta; D, descending aorta. B, Coronal T1-weighted MRI shows an aortopulmonary window mass (arrowheads) of intermediate signal intensity. Biopsy revealed paraganglioma. A, aorta; T, trachea.

spherocytosis, sickle cell disease, or extensive bone marrow replacement secondary to myelofibrosis. Extramedullary hematopoiesis is usually microscopic and commonly involves the liver, spleen, and lymph nodes. Thoracic manifestations are less common and consist of paravertebral soft tissue masses. The masses represent extrusion of the marrow through the thinned cortex of the posterior ribs. Histologically the masses resemble splenic tissue, with hematopoietic elements mixed with fat. The masses are usually bilateral, contain no calcification, and cause no rib destruction. Additional masses can be found along the lateral margins of the ribs. On CT, extramedullary hematopoiesis manifests as a heterogeneous mass or masses, often with focal areas of fat within the lesion (Fig. 38-84). The typical MRI appearance consists of well-marginated, smooth or lobulated, unilateral or bilateral paravertebral masses. T1- and T2-weighted MRIs typically show bilateral heterogeneous masses with increased signal intensity on T1-weighted images because of fat within the masses.

Paraspinal Inflammation5,6,58,220,255 Vertebral osteomyelitis can result in a paraspinal abscess. Causative organisms include Mycobacterium tuberculosis, Staphylococcus aureus, and anaerobic organisms. On CT a paraspinal abscess appears as a mass of heterogeneous attenuation. Rim enhancement following administration of IV contrast is characteristic. Imaging features that suggest paraspinal abscess include narrowing of the adjacent intervertebral disk and destruction of two or more contiguous vertebral bodies. Spondylitis and vertebral osteomyelitis are often accompanied by epidural or paraspinal inflammatory masses or abscesses (Fig. 38-85). MRI is especially well suited for identifying and demonstrating the extent of paraspinal infections because of its ability to depict the disk space, the spinal canal and its contents, and the paraspinal regions. T1-weighted images show inflammation as low signal intensity, and T2-weighted images show it as high signal intensity. The addition of gadolinium is helpful in demonstrating the extent of disease.

Diffuse Mediastinal Abnormalities Numerous lesions can manifest as diffuse or multicompartmental mediastinal disease, including lymphadenopathy, mediastinitis, lymphoproliferative disorders, and metastatic disease.

Lymphadenopathy21,31,69,98,101,111,132,195,259,265,342,346 Lymph nodes are common in all regions of the mediastinum but are most numerous around the tracheobronchial tree in the middle mediastinum (Fig. 38-86). Though still a subject of some controversy, the generally accepted upper limit of normal for short-axis lymph node diameter is 1 cm. Although enlargement is the primary CT or MR criterion for establishing the presence of lymph node disease, attenuation is also important. For instance, diffuse calcification is typical of prior granulomatous infection (e.g., tuberculosis, histoplasmosis), sarcoidosis, silicosis, calcifying or ossifying metastases, and treated lymphoma (Fig. 38-87). Nodes with low-attenuation centers and rim enhancement are often seen in patients with active infection (e.g., tuberculosis, nontuberculous mycobacterial disease) and in patients with metastatic disease to the lymph nodes (e.g., lung cancer, testicular germ cell malignancy). Diffuse intense nodal enhancement after administration of IV contrast material is typical of Castleman’s disease and some metastatic processes (e.g., renal cell carcinoma). Causes of mediastinal lymph node enlargement may be classified as follows: • Primary neoplasms of lymph nodes (e.g., lymphoma, leukemia) • Metastases from intrathoracic or extrathoracic primary malignancies • Nonlymphomatous lymphoid disorders such as Castleman’s disease • Infection (e.g., tuberculosis, fungal infection) CT and MRI are generally considered to be equivalent in their ability to detect mediastinal lymph node enlargement (Fig. 38-88). MRI is limited in its ability to detect calcification within nodes, a finding useful for distinguishing benign from malignant lymphadenopathy. Furthermore there is considerable overlap in signal intensity characteristics of benign and malignant lymphadenopathy on both T1- and T2-weighted images; thus CT remains the primary modality for diagnosis and characterization of the morphologic features of mediastinal lymphadenopathy. Whereas CT is limited to morphologic information, FDG PET provides quantitative physiologic information as a surrogate marker for malignancy. The role of FDG PET and coregistered PET/CT are well

CHAPTER 38  Mediastinal Disease

A A

A B

A

C FIG 38-84  Extramedullary hematopoiesis. A, Axial noncontrast CT shows well-circumscribed bilateral paravertebral soft tissue masses (arrowheads). B, A more superior level demonstrates a soft tissue mass (arrowhead) with medullary expansion of the bilateral posterior ribs (arrows), a common associated finding. C, Contrast-enhanced CT in a different patient also demonstrates bilateral paravertebral soft tissue masses (arrowheads) containing macroscopic areas of fat, typical for extramedullary hematopoiesis. A, descending aorta.

1061

1062

PART II  CT and MR Imaging of the Whole Body

A

A

C

A

*

B

*

D

FIG 38-85  Paraspinal abscess with spondylodiskitis due to Mycobacterium tuberculosis. A, Axial contrastenhanced CT shows bilateral enhancing paraspinal masses (arrows) with associated vertebral body destruction. M. tuberculosis was cultured from needle aspirate. B, CT coronal reconstruction better depicts vertebral body collapse and destruction with adjacent paraspinal collections (asterisk). C, Axial T1-weighted postcontrast MRI defines the extent of paravertebral inflammatory changes. D, Sagittal T2-weighted MRI shows associated cord compression (arrow) resulting from spondylodiskitis. A, descending aorta.

CHAPTER 38  Mediastinal Disease

B

1063

A S TA

T

D

A

B

A

P

A

S

S

D

C

D

D FIG 38-86  Diffuse mediastinal adenopathy in a patient with chronic lymphocytic leukemia. Contrast CT shows marked intrathoracic adenopathy in the paratracheal and prevascular mediastinum (A), precarinal region and aortopulmonary window (B), left and right hilum (C), and subcarinal region (D). Note the axillary adenopathy (arrowheads in A). A, ascending aorta; B, brachiocephalic vein; D, descending aorta; P, main pulmonary artery; S, superior vena cava; T, trachea; TA, transverse aorta.

A

P

P

FIG 38-87  Osteosarcoma metastasis to the anterior mediastinum. Contrast CT shows a soft tissue–attenuation mass with extensive osteoid (arrowheads) in the anterior mediastinum. A, aorta; P, pulmonary artery.

established in mediastinal lymph node staging; however, false positives can be seen in the setting of inflammatory and infectious diseases, and false negatives can occur in relation to technique and with certain malignancies. Like PET/CT, recent advances in MRI technology have greatly improved the imaging properties of mediastinal lymph nodes

beyond size assessment. Several studies indicate that MR sequences (e.g., short tau inversion recovery [STIR] turbo-spin echo, DWI) can differentiate benign from malignant nodes with similar or improved accuracy as compared to FDG PET. Although these small series are promising, the utility of DWI for lymph node characterization has yet to be validated in large randomized controlled trials. Lymphoma.80,90,123,158,179,194,200,237,304,324  Lymphomas are common mediastinal neoplasms that can be either focal or diffuse. Lymphomas are classified as either Hodgkin’s disease (HD) or non-Hodgkin’s lymphoma (NHL). HD is the most common mediastinal lymphoma. Of the four types of HD, nodular sclerosing is the most common type and has a unique predilection for the anterior mediastinum. Other types of HD typically manifest with diffuse mediastinal lymphadenopathy. NHL that involves the mediastinum usually manifests with diffuse lymphadenopathy involving anterior, middle, and occasionally posterior mediastinal nodal groups. However, large B-cell or lymphoblastic lymphoma can manifest as an isolated anterior mediastinal mass. These large tumors can obstruct the SVC, compress the airway, or invade the chest wall or adjacent structures. Hodgkin’s disease.  HD accounts for 0.75% of all cancers diagnosed in the United States each year. The median age at diagnosis  is 26 years, and there is a slight male predominance. The incidence of HD has a bimodal distribution with peaks between 25 and 30 years and 75 and 80 years. The characteristic Reed-Sternberg cell is  the diagnostic hallmark of HD. There are four histologic subtypes  of HD:

1064

PART II  CT and MR Imaging of the Whole Body

S

A

T T

B

A

A S

C FIG 38-88  Metastatic mediastinal adenopathy in a patient with melanoma. A to C, Axial T1-weighted images of the chest show lymph nodes of intermediate signal intensity in the high paratracheal (A), low paratracheal (B), and precarinal and aortopulmonary window (C) regions (white arrows). Note the small prevascular lymph node (black arrow in B). A, aorta; S, superior vena cava; T, trachea.

• Lymphocyte predominance (5% of cases): the best prognosis • Nodular sclerosing, the most common type (78%): the second most favorable prognosis • Mixed cellularity (17%): the third most favorable prognosis • Lymphocyte depletion (1%): the worst prognosis HD is staged using the Ann Arbor staging system (Table 38-2). The absence or presence of fever or night sweats or the unexplained loss of 10% or more body weight 6 months before diagnosis is denoted by the suffix A (designating absence of these symptoms) or B (designating presence of these symptoms). Stages I and II are treated with radiation alone, and stages III and IV are treated with a combination of radiation and chemotherapy or with chemotherapy alone. Survival is greater than 90% in stages I, II, and IIIA; 80% in stage IIIB; and 75% in stage IV. Non-Hodgkin’s lymphoma.  The NHLs are a heterogeneous group of diseases with different histologies, treatments, and prognoses but with enough similarities that they are considered collectively. Although there is no known cause of NHL, patients with impaired immune systems have a higher risk of developing this malignancy. NHL accounts for up to 3% of all newly diagnosed cancers in the United States and is four times more common than HD. NHL is the third most common childhood cancer, although the median age at

TABLE 38-2  Ann Arbor Classification

Staging System for Hodgkin’s Disease Stage

Description

I II

Involvement of a single lymph node region Involvement of two or more lymph node regions on same side of diaphragm Localized involvement of an extralymphatic organ or site and one or more lymph node regions on same side of diaphragm Involvement of lymph node regions on both sides of diaphragm Involvement of lymph node regions on both sides of diaphragm accompanied by involvement of spleen Involvement of lymph node regions on both sides of diaphragm accompanied by localized involvement of an extralymphatic organ or site Involvement of lymph node regions on both sides of diaphragm accompanied by both involvement of spleen and localized involvement of an extralymphatic organ or site Diffuse or disseminated involvement of one or more extralymphatic organs or tissues, with or without associated lymph node involvement

  IIE III   IIIS   IIIE

  IIIES

IV

CHAPTER 38  Mediastinal Disease

1065

TABLE 38-3  Histologic Classification of

Non-Hodgkin’s Lymphoma Grade

Description

Low

Small lymphocytic Follicular with predominantly small cleaved cell Follicular with mixed small and large cell Follicular with predominantly large cell Diffuse with small cleaved cell Diffuse with mixed small and large cell Diffuse with large cleaved cell Noncleaved Diffuse large cell and immunoblastic Small noncleaved cell and lymphoblastic

Intermediate

High

the time of diagnosis is 55 years. There is a male predominance (1.4 : 1). The histologic classification of NHL is complex, with 10 different cell descriptions divided into three different grades of tumor (Table 38-3). The two most common types of mediastinal NHL are diffuse large B-cell lymphoma and T-cell lymphoblastic lymphoma. Treatment is complex and involves radiation for the lower-grade tumors and chemotherapy for the higher-grade tumors. Survival is 50% to 70% for low-grade NHL, 35% to 45% for intermediate-grade NHL, and 20% to 30% for high-grade NHL. Several unique observations are seen in NHL: low-grade tumors may spontaneously regress, recur, or transform into higher grade tumors. The incidence of NHL is higher in patients with severe immune compromise, including congenital immune disorders, transplant immunosuppression, and HIV. NHL in these patients is often more aggressive and involves extranodal sites such as the central nervous system, lung parenchyma, and gastrointestinal tract. Imaging of lymphoma.*  CT is the most readily available and historically used modality for identifying and staging HD and NHL. CT provides accurate measurement of initial tumor size and extent, as well as a means of following the response to therapy (Figs. 38-89 to 38-96). Chest CT alters the initial treatment plan in approximately 10% of patients with HD by detecting more extensive disease. The anterior mediastinal, pretracheal, and hilar nodes are most commonly involved with HD. Paracardiac, subcarinal, superior diaphragmatic, internal mammary, and axillary nodes are less frequently involved. Calcification within nodes is usually a consequence of radiation therapy but is occasionally detected before treatment. Central areas of low density represent areas of necrosis, which does not appear to alter the treatment response or survival. Lymphadenopathy varies in size and extent and can manifest as a solitary large mass or discrete nodes within masses of matted nodes. Bulky mediastinal lymphadenopathy, defined as the diameter of the nodal mass exceeding one third of the thoracic diameter, is an important distinction because it may alter prognosis and therapy. When presenting as a dominant anterior mediastinal mass, features of HD include irregular contours, surface lobulation, and absence of vascular involvement. In comparison, NHL typically has regular contours.

*References 2, 7, 13, 24, 37, 42, 45, 57, 99, 100, 149, 195, 209, 238, 247, 252, 271, 317, 338, 339, 360.

A

A

A

* *

B FIG 38-89  Hodgkin’s lymphoma in a young woman. A, CT shows a homogeneous anterior mediastinal mass (arrows). The arrowhead points to the superior vena cava. A, ascending aorta. B, PET/CT in the coronal plane. The mass shows marked FDG activity (arrows). Note the associated left neck mass (arrowhead). The left ventricle activity (asterisk) is normal. A, ascending aorta.

Intrathoracic disease is noted in only 40% to 50% of patients with NHL at presentation, compared with 85% of those with HD. Whereas the most common nodal sites of involvement in HD are the anterior and superior mediastinal nodes, NHL typically involves other nodal sites, lung parenchyma, pleura, and pericardium. HD has a propensity to spread via contiguous nodal groups compared to NHL, which has multifocal nodal involvement. FDG PET has become the primary imaging modality in the staging and management of patients with lymphoma, replacing gallium-67 scintigraphy. PET/CT has been shown to have better accuracy in demonstrating nodal involvement as well as extranodal sites, thereby improving accuracy of initial staging compared to CT alone. Metabolic activity on initial PET has been shown to be predictive of tumor grade and prognosis, with higher uptake being a sign of more aggressive lymphoma. In addition, FDG PET has high diagnostic accuracy in the early prediction of response to chemotherapy, is useful for restaging lymphoma after initial treatment, and aids in the detection of early

1066

PART II  CT and MR Imaging of the Whole Body

A

S A

* A FIG 38-90  Hodgkin’s lymphoma. Contrast CT shows a large anterior mediastinal mass with a large cystic component. Note posterior displacement of the right main bronchus (asterisk). A, ascending aorta. A

S

relapse after completion of therapy. Of particular interest is the ability of FDG PET to detect persistent posttherapy lymphoma within residual masses discovered on conventional imaging studies that may warrant further treatment (Fig. 38-97). PET has also been used to help distinguish normal thymus and thymic rebound from primary or recurrent mediastinal lymphoma. Despite its utility, one primary  limitation of PET/CT is that certain subtypes of lymphoma, such as extranodal marginal zone B-cell lymphoma and mucosa-associated lymphoid tissue (MALT), are frequently not FDG avid. Although MRI can reveal additional information in as many as 15% of cases, it is more often used to assess suspected SVC obstruction, vascular invasion, and chest wall or mediastinal invasion. MRI has also been used to monitor and evaluate response to therapy, differentiate fibrosis from residual tumor, and detect recurrent lymphoma. Residual masses are seen in the immediate follow-up period in up to 88% of patients with HD and 40% of patients with NHL. These residual masses typically resolve over 12 to 18 months. The presence of residual mediastinal abnormalities is of concern because lymphoma eventually recurs in half of these patients, usually at the site of the original mass. Distinction of a residual fibrotic mass from persistent or recurrent tumor can be difficult by conventional imaging. MRI and PET/CT can be useful for this assessment. Because of its high water content, untreated lymphoma is typically homogeneous with low to intermediate signal intensity on T1-weighted images and increased signal intensity on T2-weighted images. Patients with nodular sclerosing HD occasionally have focal regions of low signal intensity within the mass on pretreatment T2-weighted images because of intralesional fibrosis. During and immediately after completion of therapy, the lesions typically become heterogeneous on MRI, and signal intensity on T2-weighted images becomes more variable. Over the subsequent 4 to 6 months, there is a tendency for the residual mass to decrease in size and signal intensity on T2-weighted images. Six months after therapy, residual masses due to fibrosis should be homogeneous, with low signal intensity on both T1- and T2-weighted images. This homogeneous low signal intensity is typical of treated inactive lymphoma. If the mass remains heterogeneous with focal regions of high signal intensity on T2-weighted images, recurrent or residual lymphoma is suggested. Based on these findings, MRI can detect persistent or recurrent tumor in treated patients as early as 8 to 12 weeks before the onset of clinical symptoms. The presence of fat

B

S

A

C FIG 38-91  Nodular sclerosing Hodgkin’s lymphoma. A, T1-weighted MRI shows a lobular heterogeneous anterior mediastinal mass with intermediate signal intensity. A, aorta; S, superior vena cava. B and C, T2-weighted MRIs show that the mass has heterogeneous signal intensity. On the cephalic image (B) there are areas of decreased signal intensity, consistent with intratumoral fibrosis. On the caudal image (C) the mass is of predominantly high signal intensity, similar to subcutaneous fat. A, aorta; S, superior vena cava. (From Erasmus JJ, et al: MR imaging of mediastinal masses. Magn Reson Imaging Clin N Am 8:59–89, 2000.)

mixed with residual fibrotic tissue, however, represents a pitfall of MRI. This potential misinterpretation can be avoided by realizing that regions of high signal intensity on T2-weighted images correlate with high-signal-intensity fat on T1-weighted images (rather than the low signal intensity of recurrent tumor). Fat-suppression pulse sequences may also be helpful for distinguishing interspersed fat from areas of residual tumor.

CHAPTER 38  Mediastinal Disease

1067

Castleman’s disease.32,44,67,79,127,155,159,170,186,212,215,224,253,262,277,316  Cas-

A

P

tleman’s disease is a benign lymphoproliferative disorder also known as angiofollicular hyperplasia or giant lymph node hyperplasia. Based on pathologic characteristics, the disease is currently classified into three major subgroups: hyaline vascular variant (HVV), plasma cell variant (PCV), and plasmablastic multicentric Castleman’s disease. PCV is further subclassified as unicentric or multicentric. Approximately 70% of cases occur in the chest, 15% in the abdomen and pelvis, and 15% in the neck.

Unicentric hyaline vascular variant and unicentric plasma cell variant.  Unicentric HVV is the most common form of CastleA

T

B

L

H

C FIG 38-92  Non-Hodgkin’s lymphoma manifesting as a mediastinal mass with chest wall involvement. A to C, Contrast CT shows a diffuse anterior mediastinal mass with chest wall invasion. Note enlarged axillary nodes (arrowheads in A). The mass extends into the neck and surrounds vessels (arrow in B). More caudally the mass manifests as a mantle of soft tissue encasing the heart (C). A, ascending aorta; H, heart; L, liver; P, main pulmonary artery; T, trachea.

man’s disease and accounts for approximately 70% of all cases. Although all age groups are affected, there is a peak incidence in the fourth decade of life. There is no gender predilection, and patients are usually asymptomatic. Unicentric HVV typically manifests as enlargement of either a single lymph node or a single chain of lymph nodes in the neck, mediastinum, abdomen, and axilla. Diagnosis usually requires excisional lymph node biopsy rather than fine-needle aspiration. Unicentric PCV accounts for about 20% of all cases and usually affects a single chain of lymph nodes rather than a solitary node. The peak incidence is in the third decade of life; like HVV, there is no gender predilection. The most common location is the abdomen. Clinically, unicentric PCV presents with constitutional symptoms, anemia, and elevated sedimentation rates. Excisional lymph node biopsy is usually required to confirm the diagnosis. A mixed form of Castleman’s has also been described when HVV and PCV disease occur in the same patient. Imaging studies show one of three major morphologic patterns: solitary mass (50%), dominant infiltrative mass with associated lymphadenopathy (40%), or diffuse lymphadenopathy confined to a single mediastinal compartment (10%). Identification of the first pattern suggests that complete surgical resection is likely. Identification of the second or third pattern suggests that complete excision may be difficult or impossible. On NCE CT, localized HVV Castleman’s disease manifests as a homogeneous or heterogeneous mass of soft tissue attenuation. Calcification is uncommon (5%-10%) and is typically coarse and central in location. HVV typically enhances intensely following the IV administration of iodinated contrast (Fig. 38-98). Prominent feeding vessels may also be shown on contrast-enhanced CT. Rarely Castleman’s disease can present as a well-defined pleural or pericardial mass or pleural effusion. On MRI, Castleman’s lesions are typically heterogeneous and of increased signal intensity (compared with skeletal muscle) on T1-weighted sequences; they become markedly hyperintense on T2-weighted sequences. Low-signal-intensity septa are occasionally visible. In larger lesions, flow voids in and around the mass may be identified and are important clues to the hypervascular nature of the lesions. Because of this hypervascularity, diffuse enhancement following administration of IV gadolinium is common (Fig. 38-99).

Multicentric plasma cell variant and plasmablastic multicentric Castleman’s disease.  The multicentric PCV of Castleman’s DWI is emerging as a potential alternative to PET/CT for the initial staging of lymphoma. Lymphomatous nodes show restricted diffusion on DWI, with corresponding low ADC values. This applies to both the more common lymphoma subtypes and the less common subtypes that are typically not FDG avid and therefore difficult to stage with PET/CT. Early studies show good agreement between MR DWI and PET/CT, suggesting that MRI may prove to be a particularly good alternative in children and patients requiring long-term follow-up for lymphoma in order to minimize radiation dose.

disease exists in two forms, depending on whether the patient has been infected with human herpesvirus type 8 (HHV-8). Multicentric Castleman’s disease with prior HHV-8 infection is classically referred to as multicentric PCV. Multicentric PCV accounts for 10% of all cases, and the disease typically affects multiple nodes or chains of nodes. There is no gender predilection, the median age at presentation is in the fourth to sixth decades, and patients usually have systemic symptoms such as fever, weight loss, and night sweats. HHV-8-positive multicentric Castleman’s disease occurs in both HIV-positive and HIV-negative patients. HIV-positive patients are almost always HHV-8 positive;

1068

PART II  CT and MR Imaging of the Whole Body

A

A P

A

P

B

C FIG 38-93  Non-Hodgkin’s lymphoma manifesting as multifocal disease. A, Noncontrast CT shows a large infiltrative anterior mediastinal mass (arrows). A, aorta; P, left pulmonary artery. B, PET/CT in the axial plane. The mass shows intense FDG activity. A more photopenic area in the center represents necrosis. C, PET/ CT in the coronal plane. Arrows identify the anterior mediastinal mass. Additional foci of disease are identified in the neck and abdomen (arrowheads).

CHAPTER 38  Mediastinal Disease

1069

Castleman’s disease is usually moderately FDG avid on PET but less than that typically seen with mediastinal malignancy (e.g., lymphoma, thymoma, metastatic nodes). SUV values tend to be higher in multicentric disease compared to unicentric disease. Detection of additional foci of disease is helpful for staging, evaluation of treatment, and recognition of recurrence.

Mediastinitis26,52,91,136,281,290 Acute mediastinitis.29,35,113,153,161,175,221,283,337,352  Acute bacterial medi-

A

A

B FIG 38-94  Non-Hodgkin’s lymphoma manifesting as multifocal disease. A, CT shows a right paravertebral soft tissue mass (arrowhead). Axillary adenopathy can be seen (arrows). B, CT at a more caudal level shows bilateral paravertebral masses (arrowheads). Small pleural effusions can be visualized bilaterally. A, aorta.

clinically the disease has a higher prevalence of pulmonary symptoms and a stronger association with Kaposi’s sarcoma. Among multicentric PCV patients who are HIV negative, the prevalence of HHV-8 infection is approximately 40%. Multicentric PCV is also associated with POEMS syndrome:  polyneuropathy, organomegaly, endocrinopathy, M protein, and skin changes. A new variant of multicentric Castleman’s disease has been described in Japan and is referred to as Castleman-Kojima disease or TAFRO syndrome (thrombocytopenia, anasarca, fever, reticulin fibrosis, organomegaly). This variant is also characterized by ascites, microcytic anemia, myelofibrosis, and renal dysfunction. Lymph nodes in these patients are typically smaller (<1.5 cm) and have mixed- or hyaline-vascular-type Castleman’s disease on pathology. Multicentric thoracic Castleman’s disease often manifests with an acute onset and has a poor prognosis due to the development of disseminated infection and malignancy, usually plasmablastic lymphoma, although other neoplasms have also been described (e.g., Kaposi’s sarcoma and HD). Multicentric thoracic Castleman’s disease usually manifests on chest radiographs as mediastinal widening. Focal mediastinal masses are rare. The anterior mediastinum is most commonly affected. On CT, diffuse lymphadenopathy involving multiple mediastinal compartments is common (Fig. 38-100). The nodes typically range in size from 1 to 6 cm in diameter and are of homogeneous attenuation on NCE CT scans. Diffuse centrilobular nodules in cases of multicentric PCV are often related to associated lymphocytic interstitial pneumonia (LIP). Associated findings such as splenomegaly or ascites are common.

astinitis is a life-threatening condition that requires prompt diagnosis and treatment. Spontaneous or iatrogenic esophageal rupture is the most common cause, accounting for up to 90% of cases. Other causes include direct spread from retropharyngeal or pleural infection and mediastinitis after cardiac surgery. Findings of esophageal perforation on CT include periesophageal fluid collections (100%), extraluminal mediastinal air (100%), esophageal wall thickening (82%), and pleural effusion (82%) (Fig. 38-101). Pleural effusion tends to be right-sided in patients with iatrogenic mid-esophageal perforation and left-sided in patients with spontaneous distal esophageal perforation (Boerhaave’s syndrome). The site of perforation is uncommonly visualized on CT (18%). Descending cervical mediastinitis is an uncommon but lifethreatening cause of mediastinitis. Such infections begin in the head and neck region and spread via fascial planes into the mediastinum, usually in the prevertebral space, where infection spreads into the middle and posterior mediastina. Typical causes include odontogenic infection, suppurative tonsillitis and adenitis, and retropharyngeal abscess. CT shows fluid collections in the mediastinum that may be contiguous with fluid collections in the cervical region (Fig. 38-102). CT is essential to confirm the diagnosis, assist in fluid aspiration to verify infection, and monitor response to therapy. Mediastinitis after cardiac surgery is uncommon, occurring in only 0.5% to 1% of patients. CT is frequently performed in patients with clinically suspected mediastinitis, but it can be difficult to interpret because fluid and air collections, hematomas, pleural effusions, and increased attenuation of anterior mediastinal fat—all potential findings of mediastinitis—are common and expected findings in the immediate postoperative period (Fig. 38-103). These findings generally resolve, however, in the days and weeks after median sternotomy. Needle aspiration of fluid collections may be necessary to rule out infection when mediastinitis is suspected (Fig. 38-104). CT is most useful for distinguishing patients with significant retrosternal fluid collections that require open drainage from those with only superficial wound infections. CT has limited ability to detect the early changes of sternal osteomyelitis. Histoplasmosis and tuberculosis can cause acute pericarditis and mediastinitis. As in mediastinitis of any cause, the CT appearance is that of increased attenuation of mediastinal fat, focal fluid collections, or abscesses. Infected lymph nodes can have low-attenuation centers and peripheral enhancing rims. Fibrosing mediastinitis.52,80,84,136,210,218,281,284,290,306,348  Mediastinal fibrosis (sclerosing or fibrosing mediastinitis) is a rare benign disorder caused by proliferation of acellular collagen and fibrous tissue within the mediastinum. An abnormal immunologic response to Histoplasma capsulatum antigens is thought to be the most common cause in the United States. Other causes include M. tuberculosis infection, autoimmune disease, radiation therapy, trauma, and drugs such as methysergide. A rare familial form associated with retroperitoneal fibrosis, sclerosing cholangitis, Riedel’s thyroiditis, and pseudotumor of the orbit is also reported. Affected patients usually present in the second through fifth decades of life with signs and symptoms due to obstruction and compression

1070

PART II  CT and MR Imaging of the Whole Body

RV

LV

LV LA

A

B

RV

RA LV

C FIG 38-95  Non-Hodgkin’s lymphoma with intracardiac involvement. A, Steady-state free precession cardiac MRI shows a homogeneous soft tissue mass involving the pericardium, right atrium, and right ventricle (arrowheads). B, Delayed enhanced MRI following contrast administration demonstrates heterogeneous enhancement throughout the mass (arrowheads). Mediastinal biopsy confirmed non-Hodgkin’s lymphoma with cardiac invasion. C, Posttreatment steady-state free precession cardiac MRI image documents complete remission. LV, left ventricle; LA, left atrium; RA, right atrium; RV, right ventricle.

CHAPTER 38  Mediastinal Disease

S A

P

A P

A B

S

A P

C FIG 38-96  Non-Hodgkin’s lymphoma. A, Contrast CT shows a large heterogeneous anterior mediastinal mass. A, ascending aorta; P, main pulmonary artery. B, T1-weighted image following therapy shows that the mass, of intermediate signal intensity, has shrunk. A, ascending aorta; P, main pulmonary artery; S, superior vena cava. C, T2-weighted image shows that the mass has lower signal intensity than in B. This decrease in signal intensity suggests that the residual mass is fibrotic. Biopsy confirmed fibrosis and the absence of residual tumor. A, ascending aorta; P, main pulmonary artery; S, superior vena cava.

A

B FIG 38-97  Hodgkin’s lymphoma, nodular sclerosing type, in a young male patient after completion of treatment. A, Contrast-enhanced CT in the axial plane demonstrates soft tissue in the anterior mediastinum (arrow). B, PET/CT in the axial plane. The mass (arrow) exhibits no FDG activity, consistent with residual fibrous tissue after treatment. The patient has remained in remission.

1071

1072

PART II  CT and MR Imaging of the Whole Body of the SVC, pulmonary veins or arteries, central airways, or esophagus. Patients most commonly present with cough, recurrent pulmonary infection, hemoptysis, or chest pain. Pulmonary venous obstruction may result in symptoms that mimic those of mitral stenosis. The chest radiographic findings of fibrosing mediastinitis are nonspecific and often underestimate the extent of mediastinal disease. CT or MRI is thus useful in the evaluation of this disorder (Figs. 38-105 to 38-108). On CT, fibrosing mediastinitis typically manifests as an infiltrative, often calcified, hilar or mediastinal process. CT is also useful for demonstrating airway and pulmonary artery and vein involvement. On MRI the process is typically of heterogeneous signal intensity on T1- and T2-weighted images. Markedly decreased signal intensity on T2-weighted images is occasionally seen, suggesting the fibrotic nature of the process. MRI may be superior to CT for defining the full extent of disease, which is important in terms of preoperative planning. However, CT is better for demonstrating calcification within the lesion, a finding that is critical for differentiating fibrosing mediastinitis from other infiltrative disorders of the mediastinum (e.g., metastatic carcinoma, lymphoma). In the few published series and case reports, FDG uptake in fibrosing mediastinitis is variable.

A

Hemorrhage

B FIG 38-98  Hyaline-vascular Castleman’s disease. A, Prone CT for biopsy shows a heterogeneous soft tissue attenuation mass in the left posterior mediastinum (arrowheads). Note biopsy needle (arrow), central linear calcification, and left pleural effusion. B, Contrast CT shows intense enhancement within the mass, as well as a small associated pretracheal node (arrow).

Bleeding within the mediastinum can be due to trauma, surgery, or line placement; occur after endoscopy; or be a result of abnormal clotting factors. Hemorrhage may extend from the neck or retroperitoneum into the mediastinum. Hemorrhage can manifest as diffuse areas of increased density or as focal areas of increased density within the mediastinal fat (Fig. 38-109).

Mediastinal Lipomatosis83,145,157,173,208,223 Mediastinal lipomatosis is a benign accumulation of fat within the mediastinum that can manifest as widening of the mediastinum on

T

T

A

B FIG 38-99  Hyaline-vascular Castleman’s disease. A, Coronal T1-weighted MRI shows a large homogeneous mass (arrowheads) of intermediate signal intensity in the subcarinal region. Note mass effect on carina. T, trachea. B, Coronal T1-weighted MRI following administration of gadolinium-based contrast material demonstrates diffuse enhancement within the mass, typical of hyaline-vascular Castleman’s disease. T, trachea.

D

A

B FIG 38-100  Disseminated Castleman’s disease. A, CT shows marked subcarinal adenopathy (arrows). There was also extensive paratracheal and prevascular adenopathy (not shown). D, descending aorta. B, Contrast CT shows marked enhancement of subcarinal and aortopulmonary window (arrowhead) lymph nodes. Note right pleural effusion.

A

*

A

P

E A

*

*

FIG 38-101  Acute mediastinitis, mediastinal abscess, and empyema after esophageal perforation due to Boerhaave’s syndrome. CT performed after administration of oral contrast material shows contrast extravasation (arrow) from the esophagus (E) into the right pleural space (asterisks). Soft tissue surrounding the esophagus represents abscess and acute mediastinitis. A, aorta.

AB

FIG 38-102  Descending cervical mediastinitis. CT shows an aircontaining mediastinal abscess (AB) that extended from an abscess in the neck.

FIG 38-103  Mediastinal hematoma mimicking mediastinal abscess after a coronary artery bypass procedure. Contrast CT shows a wellcircumscribed soft tissue abnormality in the fat of the anterior mediastinum (arrowheads). Note small focal adjacent air collections (arrow) and surgical clips anterior to the ascending aorta (A). Needle aspirate confirmed hematoma, and culture was negative for organisms. P, main pulmonary artery.

1074

PART II  CT and MR Imaging of the Whole Body

S

B

T

FIG 38-104  Mediastinitis manifesting as mediastinal abscess. CT shows a poorly marginated heterogeneous soft tissue mass (arrowheads) in the anterior mediastinum. Note absence of air within the abscess. B, brachiocephalic artery; S, superior vena cava.

FIG 38-105  Mediastinal granuloma due to histoplasmosis. CT shows a soft tissue attenuation mass with punctate calcification (arrowheads) in the superior mediastinum. Biopsy confirmed histoplasmosis. T, trachea.

A

A

P

A

P

B FIG 38-106  Fibrosing mediastinitis in a 45-year-old woman with chest pain. A, Chest CT demonstrates an infiltrating right hilar mass (arrowheads) with punctate calcification and right posterior pleural thickening. A, ascending aorta; P, pulmonary artery. B, Axial T1-weighted MRI shows a mass of intermediate signal intensity (arrowheads) and obstruction of the right pulmonary artery (P). The MR appearance is nonspecific; presence of extensive calcification on CT suggests the diagnosis. A, ascending aorta. (From Erasmus JJ, et al: MR imaging of mediastinal masses. Magn Reson Imaging Clin N Am 8:59–89, 2000.)

CHAPTER 38  Mediastinal Disease

1075

B

A

FIG 38-107  Fibrosing mediastinitis with pulmonary infarct. A, Axial noncontrast CT shows a heterogeneous soft tissue mass with calcifications (arrowheads) within the middle mediastinum, encasing the right mainstem bronchus. There is complete occlusion of the SVC (arrow). B, Axial CT on lung windows more inferiorly demonstrates mixed ground-glass and consolidative opacities of the right lower lobe (arrowheads), a finding associated with both venous congestion from impaired central pulmonary venous return and infarction from pulmonary arterial obstruction.

D

FIG 38-109  Diffuse mediastinal hematoma secondary to placement of a central line. CT shows diffuse mediastinal stranding (arrows) representing mediastinal hematoma.

FIG 38-108  Posterior mediastinal fibrosis. CT shows a diffuse soft tissue attenuation mass encasing the distal aorta (D) and extending into the pleural space (arrowhead) bilaterally. Note extensive calcification (arrow) in the right pleural space.

(lipomatous hypertrophy of the interatrial septum) is a common normal variant that can be FDG avid and should not be mistaken for malignancy (Fig. 38-112).

Mesenchymal Tumors43,70,75,138,229,245,254,335 chest radiographs. Mediastinal lipomatosis may be associated with obesity, steroid use, or Cushing’s syndrome, but there is often no predisposing factor. Accumulation of fat is more pronounced in the upper mediastinum. Mediastinal lipomatosis appears as uniform fat density of approximately −100 HU on CT (Fig. 38-110). If the fat has a higher density or strands of soft tissue density within it, other diagnoses such as liposarcoma, hemorrhage, tumor infiltration, or mediastinitis should be considered. Occasionally a prominent epicardial fat pad can result in a mediastinal mass (Fig. 38-111). Deposition of fat in the interatrial septum

A variety of mesenchymal tumors can arise in the mediastinum and affect one or multiple compartments. These include malignancies such as leiomyosarcoma and liposarcoma, as well as benign desmoid tumors of the mediastinum. Liposarcomas typically arise in the anterior mediastinum, occur in middle-aged adults, and present with symptoms of chest pain and dyspnea. Masses are typically large and diffusely infiltrate the mediastinum. Low-grade tumors can have large amounts of fat demonstrable on CT or MRI; high-grade tumors typically manifest as heterogeneous soft tissue masses on CT or MRI (Fig. 38-113).

1076

PART II  CT and MR Imaging of the Whole Body

A P S

D

FIG 38-110  Mediastinal lipomatosis. CT shows bilateral mediastinal widening due to mediastinal fat deposition (arrows). There is also fat deposition in paraspinal and posterior pleural regions bilaterally (arrowheads). The distribution is typical of mediastinal lipomatosis. A, ascending aorta; D, descending aorta; P, pulmonary artery; S, superior vena cava.

Leiomyomas and leiomyosarcomas usually arise from the esophagus, although rarely they arise in the mediastinum independent of the esophagus. Affected patients typically present with dysphagia. The tumors appear as large middle mediastinal masses of soft tissue attenuation or signal intensity on CT or MRI. Other less common mesenchymal sarcomas that may arise within the mediastinum include rhabdomyosarcoma, fibrosarcoma, chondrosarcoma, and osteosarcoma (Fig. 38-114). Desmoid tumors are uncommon benign proliferations of fibroblastic cells that typically arise from muscular fascia. Although histologically benign, they can exhibit aggressive behavior with a propensity for local invasion and recurrence after resection. They uncommonly arise within the mediastinum. Most tumors that affect the mediastinum do so by extension from chest wall or head and neck lesions. On CT desmoid tumors are usually homogeneous and of slightly increased attenuation relative to skeletal muscle. On T1-weighted MRIs they are of low signal intensity; on T2-weighted images they are of variable signal intensity. Although the tumors are usually localized, extensive invasion of mediastinal structures is occasionally detected. Overall, FDG PET has become a powerful tool to diagnose and stage mesenchymal tumors and to evaluate recurrence and tumor response. However, it lacks the ability to differentiate benign from lowgrade malignant tumors (e.g., lipoma vs. low-grade liposarcoma).

Metastatic Malignancy Metastatic malignancy can occasionally result in a diffuse mediastinal abnormality. This most likely results from extranodal extension in patients with metastatic disease to the mediastinal lymph nodes. Metastatic mucinous tumors to the mediastinum (ovarian, colon cancer) can occur in young patients, and occasionally this happens before diagnosis of the primary malignancy. Metastatic malignancy manifests on CT or MRI as an infiltrating heterogeneous mass that invades and occasionally obliterates adjacent mediastinal structures (Fig. 38-115). It can be difficult to differentiate metastatic malignancy from fibrosing mediastinitis. Features that suggest fibrosing mediastinitis rather than metastatic malignancy include diffuse calcification within

A

B FIG 38-111  Right cardiophrenic angle mass due to fat. A, Core view from a posteroanterior chest radiograph shows a mass occupying the cardiophrenic angle (arrow). B, CT shows a large right cardiophrenic angle mass (arrow) of homogeneous fat attenuation, consistent with an enlarged fat pad.

the lesion, young patient age, and no history of extrathoracic primary malignancy.

Vascular Abnormalities and Anomalies54,55,124,125,350,351 A variety of vascular anomalies may simulate a mediastinal mass. These include but are not limited to dilatation of the pulmonary arteries, dilatation of the vena cava, persistence of a left SVC, persistent vertical vein, SVC syndrome, dilatation of the azygos or hemiazygos vein, mediastinal varices, aneurysms and pseudoaneurysms of the aorta, right-sided and double aortic arches, aberrant subclavian arteries, and tortuous great vessels. In all cases, CT (usually with IV contrast) or MRI is essential to confirm the diagnosis (Fig. 38-116). Azygous continuation of the inferior vena cava.73,74,345  Interruption of the IVC is a congenital abnormality that can be associated with polysplenia. In this anomaly the infrahepatic suprarenal segment of the IVC is absent. Venous blood from the lower extremity and abdomen is diverted into the azygos system or (less commonly) the hemiazygos

CHAPTER 38  Mediastinal Disease

A

B

FIG 38-112  Lipomatous hypertrophy of the interatrial septum. A, Chest CT in the axial plane demonstrates a fat-attenuation mass (arrow) along the interatrial septum. B, PET/CT depicts marked FDG activity within the mass (arrow), consistent with brown or immature fat.

A

P

A

A

B FIG 38-113  Mediastinal liposarcoma. A, Contrast-enhanced CT in the axial plane shows a large heterogeneous mediastinal mass involving multiple compartments. Note enhancing septa (arrow) and extensive fat attenuation within the mass. The arrowhead points to the superior vena cava. A, aorta; P, pulmonary artery. B, CT in the sagittal plane better delineates the craniocaudal extension of the mass. Arrows identify soft tissue septa.

1077

1078

PART II  CT and MR Imaging of the Whole Body

A

*

P

A

A R

D

A

FIG 38-115  Diffuse mediastinal metastases. Contrast CT shows a diffuse infiltrating mediastinal mass due to metastatic breast carcinoma. Note narrowing of the superior vena cava (*), chest wall metastasis (arrowhead), and large right pleural effusion. A, ascending aorta; D, descending aorta; R, right main bronchus.

A

LV

enhanced chest CT in the axial plane shows a large homogenous anterior mediastinal mass (arrows). Note compression of the pulmonary artery (P) and left pleural effusion. A, aorta. B, Contrast-enhanced chest CT in the oblique coronal plane better demonstrates the relationship of the mass to the aorta (A) and left pulmonary artery (arrow). LV, left ventricle.

a round density just to the left of the aortic arch. If contrast is administered via the left arm, the vein will enhance (Fig. 38-119). The left SVC extends inferiorly to the region of the left superior pulmonary veins and then connects to the coronary sinus (see Fig. 38-119). Rarely the vein may connect to the left atrium, giving rise to a small right-toleft shunt. The persistent left SVC should not be confused with the left superior intercostal vein. The superior intercostal vein connects the left brachiocephalic vein with the hemiazygos system. It runs parallel to the aortic arch, is often seen in the fat adjacent to the arch, and normally measures no more than 4 mm in diameter. It can become enlarged owing to collateral flow in patients with congenital or acquired obstruction of the vena cava.

system, resulting in dilatation of these veins. Chest radiographs usually show an enlarged azygos vein, widened paraspinal reflections, and nonvisualization of the IVC on the lateral radiograph. CT or MRI shows enlargement of the azygos or hemiazygos vein and marked dilatation of the azygos arch or left superior intercostal vein (Figs. 38-117 and 38-118). Persistent left superior vena cava.56,65  A persistent left SVC occurs in about 1 in 200 adults with normal hearts and in up to 5% of children with congenital heart disease. Most patients with this anomaly have both a right and a left SVC; occasionally a persistent brachiocephalic vein connects the two. On CT the left SVC is seen as

be due to malignancy, infection, or iatrogenic causes. Lung carcinoma is the most common malignant cause, histoplasmosis-induced fibrosing mediastinitis is the most common infectious cause (in the midwestern United States), and central venous catheters are the most common iatrogenic cause. Superior vena cava syndrome is a clinical disorder of SVC obstruction resulting in signs and symptoms such as upper extremity or facial edema, dyspnea, dysphagia, headache, and cognitive dysfunction due to venous congestion. Most patients present with symptoms prior to diagnosis. In the past, upper extremity venography was used to diagnose SVC or brachiocephalic venous obstruction. Contrast CT utilizing a delayed venous protocol is now the most common imaging modality used (Fig. 38-120). MRI can also be useful

B FIG 38-114  Primary mediastinal synovial sarcoma. A, Contrast-

Superior vena cava or brachiocephalic vein obstruction.23,177,198,307,357  Obstruction of the SVC or brachiocephalic vein can

* *

*

*

* D

*

*

E

A

B FIG 38-116  Esophageal varices. A, Contrast CT shows multiple tubular contrast-enhancing structures (asterisks) surrounding the distal esophagus (E), compatible with esophageal varices. Note the nodularity of the liver (arrows). D, descending aorta. B, Contrast-enhanced CT in the coronal plane also demonstrates varices (asterisks).

S

A

B

A

L

* *

*

C FIG 38-117  Azygos continuation of the interrupted inferior vena cava. A, CT shows a dilated azygos vein (arrowheads) as it enters the posterior wall of the superior vena cava (S). B, Caudal CT shows a dilated azygos vein (arrow) without evidence of collaterals. C, CT through the upper abdomen demonstrates heterotaxy syndrome with situs ambiguous; the liver (L) is left predominant with right-sided polysplenia (asterisks). There is no intrahepatic IVC. A, descending aorta.

1080

PART II  CT and MR Imaging of the Whole Body

A

R L

L

AA T

T

V

e D

A

B

P A

D H

C FIG 38-118  Complex venous anomalies in a patient with polysplenia. A, Contrast CT shows the left (L) and right (R) superior vena cavae. AA, aortic arch; e, esophagus; T, trachea. B, A more caudal image shows a dilated superior intercostal vein (V) along the lateral aspect of the proximal descending aorta (D). Note the anterior communication with the left superior vena cava (L). A, ascending aorta. T, trachea. C, An even more caudal image shows a dilated hemiazygos vein (H), suggesting interruption of the inferior vena cava with hemiazygos continuation, which was confirmed on abdominal images (not shown). A, ascending aorta; D, descending aorta; P, main pulmonary artery; T, trachea.

CHAPTER 38  Mediastinal Disease

1081

P

A A

A

A

B

A

P P

RA CS

A RA

C

D FIG 38-119  Persistent left superior vena cava (SVC). A, Contrast-enhanced CT shows a persistent left SVC (arrow) adjacent to the transverse aorta (A). B and C, Contrast-enhanced CT images at more caudal levels show characteristic anatomic findings of a persistent left SVC. Note the two vascular structures anterior to the left main bronchus. The left SVC (arrow in B) is anterior to the left upper lobe bronchus, adjacent and medial to the left superior pulmonary vein. Drainage of the left SVC into the coronary sinus (CS in C) typically results in dilatation of the sinus. A, aorta; P, pulmonary arteries; RA, right atrium. D, Contrast-enhanced CT in the oblique coronal plane demonstrates the course of the left SVC (arrows) along the left mediastinum, as well as the relationship to the aorta (A) and pulmonary arteries (P). RA, right atrium.

in patients with contraindications to administration of iodinated contrast material, but it is occasionally difficult to differentiate complete obstruction from marked narrowing with very slow flow. Treatment options vary and are dependent upon the underlying cause of obstruction, which can often be elucidated by contrast  CT. For in situ thrombus, thrombolytics or—in extreme cases— thrombectomy may be performed to relieve obstruction. In the setting of extrinsic compression due to malignancy, radiation or chemotherapy are the primary management options, with CT for surveillance in monitoring response. Endovascular stent grafting may be performed prior to or in conjunction with other treatments to provide immediate

symptomatic relief; however, obstruction often recurs in the absence of definitive treatment. Aberrant right subclavian artery.34,53,178,266,267,276,323  An aberrant right subclavian artery is a common vascular abnormality that occurs in approximately 1 in 200 people. In younger patients it is usually not obvious on a chest radiograph, but in older patients it can manifest as a right paratracheal mass or, on the lateral radiograph, as a retrotracheal masslike opacity. On CT the aberrant right subclavian artery arises as the last and most posterior vessel off the aortic arch, crosses the mediastinum behind the trachea and esophagus, and continues into the neck and axilla along the right side of the trachea (Figs. 38-121

1082

PART II  CT and MR Imaging of the Whole Body

H

*

L

S

A

B FIG 38-120  Collateral venous channels in a patient with thrombosis of the left subclavian vein due to an indwelling catheter. A, Contrast CT shows narrowing of the left subclavian vein (arrows), as well as numerous enlarged veins in the soft tissues of the chest (black arrowheads) and paravertebral region. Note the enlarged right superior intercostal vein (white arrowhead) and paratracheal adenopathy (asterisk). B, Contrast CT shows collateral venous drainage in the recanalized umbilical vein (arrow). Note bilateral pleural effusions. H, heart; L, liver; S, spleen.

A T

A

A

A

B

C

D

FIG 38-121  Aberrant right subclavian artery with left aortic arch. A and B, Contrast-enhanced CT in the axial (A) and coronal (B) planes shows that the aberrant right subclavian artery (arrows) is retroesophageal (arrowhead in A indicates the esophagus) and retrotracheal. A, aorta; T, trachea. C and D, 3D shaded-surface rendering in anterior (C) and posterior (D) views. Note the origin of the aberrant subclavian artery (arrows) as the last branch of the aortic arch.

CHAPTER 38  Mediastinal Disease

c

1083

A

T T

c e

L

A

B FIG 38-122  Aneurysm of aberrant right subclavian artery. A, CT shows a homogeneous soft tissue mass (arrows) with rim calcification posterior to the trachea (T). c, left and right common carotid arteries; e, esophagus; L, left subclavian artery. B, A more caudal image confirms that the mass is an aneurysm of the aberrant right subclavian artery, with its origin from the proximal descending aorta (arrowheads). A, common origin of carotid arteries from the aorta; T, trachea.

and 38-122). Frequently a diverticulum (Kommerell’s) is seen at the origin of the aberrant right subclavian artery from the aorta. Although usually inconsequential, Kommerell’s diverticula may become aneurysmal, with compression of the esophagus and resultant dysphagia. Rupture is a rare complication, but the incidence increases beyond 3 cm in size, at which point treatment should be considered. Right aortic arch.16,178,217,266,276,323  A right aortic arch is commonly mistaken for a mediastinal mass on chest radiographs, and patients are often referred for CT evaluation of a right paratracheal mass. Most asymptomatic adults with a right-sided aortic arch have an aberrant left subclavian artery (≈1 in 200 people). Other anatomic configurations such as mirror-image branching of the great vessels are less common and more often associated with severe congenital heart disease. CT or MRI easily identifies the aortic arch to the right of the trachea. The aberrant right subclavian artery arises as the last and most posterior vessel off the right-sided aortic arch, crosses the mediastinum behind the trachea and esophagus, and continues into the neck and axilla along the left side of the trachea (Fig. 38-123). A diverticulum can be associated with the origin of the aberrant left subclavian artery from the aorta and when large can simulate a double aortic arch on chest radiography. Although Kommerell’s diverticulum was originally described in association with aberrant right subclavian artery and a left aortic arch, the eponym is more generally applied to a diverticulum associated with either right or left aberrant subclavian arteries. Double aortic arch.178,217,266,276,323  Double aortic arches typically manifest in childhood with stridor. Rarely the lesion manifests in adulthood as an asymptomatic mediastinal mass. On CT or MRI the double aortic arch is obvious as two large vascular structures encircling the trachea. Each arch gives rise to a subclavian and a carotid artery. The descending aorta is usually on the left, and the right arch is often higher and slightly larger than the left arch. Left ventricular aneurysm and pseudoaneurysm.130,169,274  Aneurysms and pseudoaneurysms of the left ventricle often present as

mediastinal masses. True left ventricular aneurysms are caused by myocardial ischemia and infarction, resulting in focal thinning, dilatation, and paradoxical motion of the affected portion of the ventricle. True aneurysms usually occur in the distribution of the left anterior descending coronary artery (along the anterior and apical walls of the left ventricle) and frequently calcify. On CT they manifest as broadbased areas of thinning of the anterior or apical walls of the left ventricle, often with associated calcification (Figs. 38-124 and 38-125). Mural thrombus may be identified adjacent to the aneurysm. Cine MRI can be useful to demonstrate paradoxical wall motion in the region of the aneurysm. Left ventricular pseudoaneurysms result from perforation of the ventricular wall after myocardial infarction. The perforation is contained by pericardium, resulting in pseudoaneurysm formation. Pseudoaneurysms more commonly occur along the posterior and inferior surfaces of the heart and infrequently calcify. On contrast-enhanced CT they manifest as focal enhancing masses along the posterior or inferior surface of the ventricle, with a narrow communication (neck) with the left ventricular chamber (Fig. 38-126). It is important to differentiate pseudoaneurysms from true aneurysms; the former are prone to delayed rupture, whereas the latter are not.

Diaphragmatic Hernia71,93,106,108,223,244 Abdominal organs or retroperitoneal fat can herniate through areas of congenital or acquired diaphragmatic weakness or tears and manifest as a mediastinal mass on chest radiographs. Common nontraumatic sites of herniation include the anterior parasternal hiatus, the esophageal hiatus, and the posterior pleuroperitoneal hiatus. Hiatal hernias are a common cause of middle mediastinal masses in the lower thorax. The diagnosis is easily made if there is air or contrast within the hernia (Fig. 38-127). Herniation through the anterior parasternal hiatus (foramen of Morgagni) is uncommon and usually manifests as a rightsided mediastinal or cardiophrenic angle mass (Fig. 38- 128). These defects are more commonly identified in middle-aged to older adults

1084

PART II  CT and MR Imaging of the Whole Body

T

RV LV

R D

FIG 38-124  Calcified left ventricular aneurysm. CT shows a thin curvilinear calcification in the left ventricular apex, consistent with a left ventricular aneurysm (arrow) due to prior infarction. LV, left ventricle; RV, right ventricle.

A

BC

S

R

T

**

LS

B FIG 38-123  Right aortic arch with aberrant left subclavian artery. A, Coronal cardiac-gated black blood sequence MRI shows a right aortic arch (R) and retroesophageal aortic diverticulum (D) giving rise to an aberrant left subclavian artery. T, trachea. B, Axial cardiac-gated MRI shows an aberrant left subclavian artery (LS) in a retrotracheal location. Asterisks, esophagus; BC, brachiocephalic vein; R, right aortic arch; S, superior vena cava; T, trachea.

FIG 38-125  Calcified left ventricular aneurysm with apical thrombus. Contrast CT shows a thin curvilinear calcification (white arrowhead) in the left ventricular apex, consistent with a left ventricular aneurysm. Note the focal wall thinning and adjacent apical thrombus (black arrowheads).

CHAPTER 38  Mediastinal Disease

1085

* LV

A

FIG 38-126  Left ventricular pseudoaneurysm. Contrast CT shows a large pseudoaneurysm arising from the posterior wall of the left ventricle. Note the narrow neck (arrowheads) between the pseudoaneurysm (A) and left ventricular cavity (LV), as well as peripheral thrombus.

A

B FIG 38-128  Foramen of Morgagni hernia. A, Axial CT shows herniation of omental fat into the right cardiophrenic angle (arrowheads). Note the presence of mesenteric vessels and air-containing bowel loops (asterisk). B, Coronal CT demonstrates the anterior diaphragmatic defect (arrowheads) through which fat, vessels, and bowel herniate.

HH

A

B FIG 38-127  Hiatal hernia. A, Lateral chest radiograph shows a middle mediastinal mass containing an airfluid level (arrows). B, Axial CT with both oral and IV contrast demonstrates herniation of the gastric fundus (HH) through the diaphragmatic esophageal hiatus. HH, hiatal hernia.

1086

PART II  CT and MR Imaging of the Whole Body

REFERENCES

*

*

FIG 38-129  Bochdalek’s hernia. CT shows a homogeneous fat- attenuation mass in a right paraspinal location (asterisk). This location and CT appearance are diagnostic for Bochdalek’s hernia. Also note the large hiatal hernia (arrows).

TABLE 38-4  Classification of Mediastinal

Abnormalities Based on CT Attenuation Characteristics CT Attenuation

Lesion

Fat

Lipoma, lipomatosis, thymolipoma, liposarcoma, teratoma, epicardial fat pad, hernia Thymic cyst, teratoma, pericardial cyst, bronchogenic cyst, esophageal duplication cyst, meningocele, neurenteric cyst Thymoma, thymic carcinoma, germ cell tumor, esophageal neoplasm, neurogenic tumor, lymphoma, lymphadenopathy Calcified nodes (histoplasmosis, tuberculosis, sarcoidosis, silicosis), calcified neoplasm (thymoma, teratoma, treated lymphoma, neurogenic tumor, metastases), fibrosing mediastinitis, hemorrhage, goiter, vascular Vascular, goiter, Castleman’s disease, hemangioma, paraganglioma

Water

Soft tissue

High attenuation

Contrast enhancing

and are typically asymptomatic. Rarely defects occur on the left or  into the pericardium. Herniation through the posterior pleuroperitoneal hiatus (foramen of Bochdalek) is the most common cause of congenital diaphragmatic hernia in infants. In adults the lesions are usually small and asymptomatic and occur more commonly on the  left (65%) than on the right (35%). The incidence of small, posterior, fat-containing defects increases with age and degree of emphysema (Fig. 38-129).

CLASSIFICATION OF MASSES BY CT ATTENUATION An alternative method of classifying mediastinal masses is based on their predominant attenuation characteristics on CT rather than by their location110,112 (Table 38-4).

1. Abdel Razek AA, Khairy M, Nada N: Diffusion-weighted MR imaging in thymic epithelial tumors: Correlation with World Health Organization classification and clinical staging. Radiology 273:268–275, 2014. 2. Abrahamsen AF, Lien HH, Aas M, et al: Magnetic resonance imaging and 67gallium scan in mediastinal malignant lymphoma: A prospective pilot study. Ann Oncol 5:433–436, 1994. 3. Ackman JB, Wu CC: MRI of the thymus. AJR Am J Roentgenol 197:W15–W20, 2011. 4. Adam A, Hochholzer L: Ganglioneuroblastoma of the posterior mediastinum: A clinicopathologic review of 80 cases. Cancer 47:373– 381, 1981. 5. al-Mulhim FA, Ibrahim EM, el-Hassan AY, et al: Magnetic resonance imaging of tuberculous spondylitis. Spine 20:2287–2292, 1995. 6. Arizono T, Oga M, Shiota E, et al: Differentiation of vertebral osteomyelitis and tuberculous spondylitis by magnetic resonance imaging. Int Orthop 19:319–322, 1995. 7. Aronberg DJ, Glazer HS, Sagel SS: MRI and CT of the mediastinum: Comparisons, controversies, and pitfalls. Radiol Clin North Am 23:439–448, 1985. 8. Auffermann W, Gooding GA, Okerlund MD, et al: Diagnosis of recurrent hyperparathyroidism: Comparison of MR imaging and other imaging techniques. AJR Am J Roentgenol 150:1027–1033, 1988. 9. Aughenbaugh GL: Thoracic manifestations of neurocutaneous diseases. Radiol Clin North Am 22:741–756, 1984. 10. Aygun C, Slawson RG, Bajaj K, et al: Primary mediastinal seminoma. Urology 23:109–117, 1984. 11. Azar T, Abdul-Karim FW, Tucker HM: Adenoid cystic carcinoma of the trachea. Laryngoscope 108:1297–1300, 1998. 12. Azarow KS, Pearl RH, Zurcher R, et al: Primary mediastinal masses. A comparison of adult and pediatric populations. J Thorac Cardiovasc Surg 106:67–72, 1993. 13. Baba S, Abe K, Isoda T, et al: Impact of FDG-PET/CT in the management of lymphoma. Ann Nucl Med 25:701–716, 2011. 14. Barakos JA, Brown JJ, Brescia RJ, et al: High signal intensity lesions of the chest in MR imaging. J Comput Assist Tomogr 13:797–802, 1989. 15. Barakos JA, Brown JJ, Higgins CB: MR imaging of secondary cardiac and paracardiac lesions. AJR Am J Roentgenol 153:47–50, 1989. 16. Baron RL, Gutierrez FR, Sagel SS, et al: CT of anomalies of the mediastinal vessels. AJR Am J Roentgenol 137:571–576, 1981. 17. Baron RL, Lee JK, Sagel SS, et al: Computed tomography of the normal thymus. Radiology 142:121–125, 1982. 18. Baron RL, Sagel SS, Baglan RJ: Thymic cysts following radiation therapy for Hodgkin disease. Radiology 141:593–597, 1981. 19. Bar-Ziv J, Nogrady MB: Mediastinal neuroblastoma and ganglioneuroma. The differentiation between primary and secondary involvement on the chest roentgenogram. Am J Roentgenol Radium Ther Nucl Med 125:380–390, 1975. 20. Bashist B, Ellis K, Gold RP: Computed tomography of intrathoracic goiters. AJR Am J Roentgenol 140:455–460, 1983. 21. Batra P, Brown K, Steckel RJ, et al: MR imaging of the thorax: A comparison of axial, coronal, and sagittal imaging planes. J Comput Assist Tomogr 12:75–81, 1988. 22. Batra P, Herrmann C, Jr, Mulder D: Mediastinal imaging in myasthenia gravis: Correlation of chest radiography, CT, MR, and surgical findings. AJR Am J Roentgenol 148:515–519, 1987. 23. Bechtold RE, Wolfman NT, Karstaedt N, et al: Superior vena caval obstruction: Detection using CT. Radiology 157:485–487, 1985. 24. Bendini M, Zuiani C, Bazzocchi M, et al: Magnetic resonance imaging and 67Ga scan versus computed tomography in the staging and in the monitoring of mediastinal malignant lymphoma: A prospective pilot study. MAGMA 4:213–224, 1996. 25. Benveniste MF, Moran CA, Mawlawi O, et al: FDG PET-CT aids in the preoperative assessment of patients with newly diagnosed thymic epithelial malignancies. J Thorac Oncol 8:502–510, 2013. 26. Berry DF, Buccigrossi D, Peabody J, et al: Pulmonary vascular occlusion and fibrosing mediastinitis. Chest 89:296–301, 1986.

CHAPTER 38  Mediastinal Disease 27. Bhargava R, Parham DM, Lasater OE, et al: MR imaging differentiation of benign and malignant peripheral nerve sheath tumors: Use of the target sign. Pediatr Radiol 27:124–129, 1997. 28. Bhosale P, Shah A, Wei W, et al: Carcinoid tumours: Predicting the location of the primary neoplasm based on the sites of metastases. Eur Radiol 23:400–407, 2013. 29. Bitkover CY, Cederlund K, Aberg B, et al: Computed tomography of the sternum and mediastinum after median sternotomy. Ann Thorac Surg 68:858–863, 1999. 30. Biyyam DR, Chapman T, Ferguson MR, et al: Congenital lung abnormalities: Embryologic features, prenatal diagnosis, and postnatal radiologic-pathologic correlation. Radiographics 30:1721–1738, 2010. 31. Boiselle PM, Patz EF, Jr, Vining DJ, et al: Imaging of mediastinal   lymph nodes: CT, MR, and FDG PET. Radiographics 18:1061–1069, 1998. 32. Bonekamp D, Horton KM, Hruban RH, et al: Castleman disease: The great mimic. Radiographics 31:1793–1807, 2011. 33. Boyacigil S, Ardic S, Tokoglu F, et al: Intrathoracic extramedullary haemopoiesis. Australas Radiol 40:179–181, 1996. 34. Branscom JJ, Austin JH: Aberrant right subclavian artery. Findings seen on plain chest roentgenograms. Am J Roentgenol Radium Ther Nucl Med 119:539–542, 1973. 35. Breatnach E, Nath PH, Delany DJ: The role of computed tomography in acute and subacute mediastinitis. Clin Radiol 37:139–145, 1986. 36. Brenner W, Friedrich RE, Gawad KA, et al: Prognostic relevance of FDG PET in patients with neurofibromatosis type-1 and malignant peripheral nerve sheath tumours. Eur J Nucl Med Mol Imaging 33:428–432, 2006. 37. Brice P, Rain JD, Frija J, et al: Residual mediastinal mass in malignant lymphoma: Value of magnetic resonance imaging and gallium scan. Nouv Rev Fr Hematol 35:457–461, 1993. 38. Brown LR, Aughenbaugh GL: Masses of the anterior mediastinum: CT and MR imaging. AJR Am J Roentgenol 157:1171–1180, 1991. 39. Brown LR, Aughenbaugh GL, Wick MR, et al: Roentgenologic diagnosis of primary corticotropin-producing carcinoid tumors of the mediastinum. Radiology 142:143–148, 1982. 40. Burk DL, Jr, Brunberg JA, Kanal E, et al: Spinal and paraspinal neurofibromatosis: Surface coil MR imaging at 1.5 T1. Radiology 162:797–801, 1987. 41. Caldarella C, Treglia G, Isgro MA, et al: Diagnostic performance of positron emission tomography using (1)(1)C-methionine in patients with suspected parathyroid adenoma: A meta-analysis. Endocrine 43:78–83, 2013. 42. Canini R, Battista G, Monetti N, et al: Bulky mediastinal lymphomas: Role of magnetic resonance and SPECT-Ga- 67 in the evaluation of residual masses. Radiol Med (Torino) 90:448–456, 1995. 43. Casillas J, Sais GJ, Greve JL, et al: Imaging of intra- and extraabdominal desmoid tumors. Radiographics 11:959–968, 1991. 44. Casper C: The aetiology and management of Castleman disease at 50 years: Translating pathophysiology to patient care. Br J Haematol 129:3–17, 2005. 45. Castellino RA: Diagnostic imaging studies in patients with newly diagnosed Hodgkin’s disease. Ann Oncol 3(Suppl 4):45–47, 1992. 46. Caty MG, Shamberger RC: Abdominal tumors in infancy and childhood. Pediatr Clin North Am 40:1253–1271, 1993. 47. Chao MW, Smith JG, Laidlaw C, et al: Results of treating primary tumors of the trachea with radiotherapy. Int J Radiat Oncol Biol Phys 41:779–785, 1998. 48. Cho A, Kim SW, Choi J, et al: The additional value of attenuation correction CT acquired during 18F-FDG PET/CT in differentiating mature from immature teratomas. Clin Nucl Med 39:e193–e196, 2014. 49. Choi SJ, Lee JS, Song KS, et al: Mediastinal teratoma: CT differentiation of ruptured and unruptured tumors. AJR Am J Roentgenol 171:591–594, 1998. 50. Choi YW, McAdams HP, Jeon SC, et al: The “High-Riding” superior pericardial recess: CT findings. AJR Am J Roentgenol 175:1025–1028, 2000.

1087

51. Choi YW, McAdams HP, Jeon SC, et al: Idiopathic multilocular thymic cyst: CT features with clinical and histopathologic correlation. AJR Am J Roentgenol 177:881–885, 2001. 52. Chong S, Kim TS, Kim BT, et al: Fibrosing mediastinitis mimicking malignancy at CT: Negative FDG uptake in integrated FDG PET/CT imaging. Eur Radiol 17:1644–1646, 2007. 53. Cina CS, Althani H, Pasenau J, et al: Kommerell’s diverticulum and right-sided aortic arch: A cohort study and review of the literature. J Vasc Surg 39:131–139, 2004. 54. Cohen MC, Hartnell GG, Finn JP: Magnetic resonance angiography of congenital pulmonary vein anomalies. Am Heart J 127:954–955, 1994. 55. Cole TJ, Henry DA, Jolles H, et al: Normal and abnormal vascular structures that simulate neoplasms on chest radiographs: Clues to the diagnosis. Radiographics 15:867–891, 1995. 56. Cormier MG, Yedlicka JW, Gray RJ, et al: Congenital anomalies of the superior vena cava: A CT study. Semin Roentgenol 24:77–83, 1989. 57. Cronin CG, Swords R, Truong MT, et al: Clinical utility of PET/CT in lymphoma. AJR Am J Roentgenol 194:W91–W103, 2010. 58. Dagirmanjian A, Schils J, McHenry M, et al: MR imaging of vertebral osteomyelitis revisited [see comments]. AJR Am J Roentgenol 167:1539– 1543, 1996. 59. Daltro P, Werner H, Gasparetto TD, et al: Congenital chest malformations: A multimodality approach with emphasis on fetal MR imaging. Radiographics 30:385–395, 2010. 60. Dave UR, Williams AD, Wilson JA, et al: Esophageal cancer staging with endoscopic MR imaging: Pilot study. Radiology 230:281–286, 2004. 61. Davidson AJ, Hartman DS, Goldman SM: Mature teratoma of the retroperitoneum: Radiologic, pathologic, and clinical correlation. Radiology 172:421–425, 1989. 62. Davis RD, Jr, Oldham HN, Jr, Sabiston DC, Jr: Primary cysts and neoplasms of the mediastinum: Recent changes in clinical presentation, methods of diagnosis, management, and results. Ann Thorac Surg 44:229–237, 1987. 63. Davis S, Rogers MA, Pendergrass TW: The incidence and epidemiologic characteristics of neuroblastoma in the United States. Am J Epidemiol 126:1063–1074, 1987. 64. de Geer G, Webb WR, Gamsu G: Normal thymus: Assessment with MR and CT. Radiology 158:313–317, 1986. 65. Demos TC, Posniak HV, Pierce KL, et al: Venous anomalies of the thorax. AJR Am J Roentgenol 182:1139–1150, 2004. 66. Dexeus FH, Logothetis CJ, Chong C, et al: Genetic abnormalities in men with germ cell tumors. J Urol 140:80–84, 1988. 67. Dispenzieri A, Gertz MA: Treatment of Castleman’s disease. Curr Treat Options Oncol 6:255–266, 2005. 68. Donner MW, Saba GP, Martinez CR: Diffuse diseases of the esophagus: A practical approach. Semin Roentgenol 16:198–213, 1981. 69. Dooms GC, Hricak H, Moseley ME, et al: Characterization of lymphadenopathy by magnetic resonance relaxation times: preliminary results. Radiology 155:691–697, 1985. 70. Dosios TJ, Angouras DC, Floros DG: Primary desmoid tumor of the posterior mediastinum. Ann Thorac Surg 66:2098–2099, 1998. 71. Dreizin D, Borja MJ, Danton GH, et al: Penetrating diaphragmatic injury: Accuracy of 64-section multidetector CT with trajectography. Radiology 268:729–737, 2013. 72. Ducatman BS, Scheithauer BW, Piepgras DG, et al: Malignant peripheral nerve sheath tumors. A clinicopathologic study of 120 cases. Cancer 57:2006–2021, 1986. 73. Dudiak CM, Olson MC, Posniak HV: Abnormalities of the azygos system: CT evaluation. Semin Roentgenol 24:47–55, 1989. 74. Dudiak CM, Olson MC, Posniak HV: CT evaluation of congenital and acquired abnormalities of the azygos system. Radiographics 11:233–246, 1991. 75. Eisenstat R, Bruce D, Williams LE, et al: Primary liposarcoma of the mediastinum with coexistent mediastinal lipomatosis. AJR Am J Roentgenol 174:572–573, 2000. 76. El-Bawab H, Al-Sugair AA, Rafay M, et al: Role of flourine-18 fluorodeoxyglucose positron emission tomography in thymic pathology. Eur J Cardiothorac Surg 31:731–736, 2007.

1088

PART II  CT and MR Imaging of the Whole Body

77. Endo M, Adachi S, Kusumoto M, et al: A study of the utility of the MR image for the diagnosis of thymic tumors–Imaging and pathologic correlation. Nippon Igaku Hoshasen Gakkai Zasshi 53:1–10, 1993. 78. Engels EA: Epidemiology of thymoma and associated malignancies.   J Thorac Oncol 5:S260–S265, 2010. 79. Enomoto K, Nakamichi I, Hamada K, et al: Unicentric and multicentric Castleman’s disease. Br J Radiol 80:e24–e26, 2007. 80. Erasmus JJ, McAdams HP, Donnelly LF, et al: MR imaging of mediastinal masses. Magn Reson Imaging Clin N Am 8:59–89, 2000. 81. Erasmus JJ, McAdams HP, Patz EF, Jr, et al: Evaluation of primary pulmonary carcinoid tumors using FDG PET. AJR Am J Roentgenol 170:1369–1373, 1998. 82. Eslamy HK, Ziessman HA: Parathyroid scintigraphy in patients with primary hyperparathyroidism: 99mTc sestamibi SPECT and SPECT/ CT. Radiographics 28:1461–1476, 2008. 83. Fan CM, Fischman AJ, Kwek BH, et al: Lipomatous hypertrophy of the interatrial septum: Increased uptake on FDG PET. AJR Am J Roentgenol 184:339–342, 2005. 84. Farmer DW, Moore E, Amparo E, et al: Calcific fibrosing mediastinitis: Demonstration of pulmonary vascular obstruction by magnetic resonance imaging. AJR Am J Roentgenol 143:1189–1191, 1984. 85. Feigin DS, Fenoglio JJ, McAllister HA, et al: Pericardial cysts. A radiologic-pathologic correlation and review. Radiology 125:15–20, 1977. 86. Felson B: Chest roentgenology, Philadelphia, 1973, W. B. Saunders. 87. Ferdinand B, Gupta P, Kramer EL: Spectrum of thymic uptake at 18F-FDG PET. Radiographics 24:1611–1616, 2004. 88. Fernandes ET, Custer MD, Burton EM, et al: Neurenteric cyst: Surgery and diagnostic imaging. J Pediatr Surg 26:108–110, 1991. 89. Ferner RE, Lucas JD, O’Doherty MJ, et al: Evaluation of (18) fluorodeoxyglucose positron emission tomography ([18]FDG PET) in the detection of malignant peripheral nerve sheath tumours arising from within plexiform neurofibromas in neurofibromatosis 1. J Neurol Neurosurg Psychiatry 68:353–357, 2000. 90. Filly R, Bland N, Castellino RA: Radiographic distribution of intrathoracic disease in previously untreated patients with Hodgkin’s disease and non-Hodgkin’s lymphoma. Radiology 120:277–281, 1976. 91. Francis IR, Dorovini-Zis K, Glazer GM, et al: The fibromatoses: CT-pathologic correlation. AJR Am J Roentgenol 147:1063–1066, 1986. 92. Francis IR, Glazer GM, Bookstein FL, et al: The thymus: Reexamination of age-related changes in size and shape. AJR Am J Roentgenol 145:249–254, 1985. 93. Fraser RS, Muller NL, Colman N, et al, editors: Fraser and Paré’s diagnosis of diseases of the chest, ed 4, Philadelphia, 1999, WB Saunders. 94. Reference deleted in proofs. 95. Fulcher AS, Proto AV, Jolles H: Cystic teratoma of the mediastinum: Demonstration of fat/fluid level. AJR Am J Roentgenol 154:259–260, 1990. 96. Furukawa H: Magnetic resonance (MR) imaging for the detection of the invasion into neighboring structures in esophageal cancers. Nippon Geka Gakkai Zasshi 92:636–644, 1991. 97. Gale AW, Jelihovsky T, Grant AF, et al: Neurogenic tumors of the mediastinum. Ann Thorac Surg 17:434–443, 1974. 98. Gamsu G, Stark DD, Webb WR, et al: Magnetic resonance imaging of benign mediastinal masses. Radiology 151:709–713, 1984. 99. Gasparini MD, Balzarini L, Castellani MR, et al: Current role of gallium scan and magnetic resonance imaging in the management of mediastinal Hodgkin lymphoma. Cancer 72:577–582, 1993. 100. Gawande RS, Khurana A, Messing S, et al: Differentiation of normal thymus from anterior mediastinal lymphoma and lymphoma recurrence at pediatric PET/CT. Radiology 262:613–622, 2012. 101. Gefter WB: Magnetic resonance imaging in the evaluation of lung cancer. Semin Roentgenol 25:73–84, 1990. 102. Gefter WB, Spritzer CE, Eisenberg B, et al: Thyroid imaging with high-field-strength surface-coil MR. Radiology 164:483–490, 1987.

103. Gelfand DW, Goldman AS, Law EJ, et al: Thymic hyperplasia in children recovering from thermal burns. J Trauma 12:813–817, 1972. 104. Ghosh BC, Ghosh L, Huvos AG, et al: Malignant schwannoma. A clinicopathologic study. Cancer 31:184–190, 1973. 105. Gibbs JM, Chandrasekhar CA, Ferguson EC, et al: Lines and stripes: Where did they go?—From conventional radiography to CT. Radiographics 27:33–48, 2007. 106. Gilkeson RC, Basile V, Sands MJ, et al: Chest case of the day. Morgagni’s hernia. AJR Am J Roentgenol 169:266, 268–270, 1997. 107. Gilkeson RC, Basile V, Sands MJ, et al: Chest case of the day. Extramedullary hematopoiesis (EMH). AJR Am J Roentgenol 169:267, 270–263, 1997. 108. Glasser DL, Shanmuganathan K, Mirvis SE: General case of the day. Acute intrapericardial diaphragmatic hernia. Radiographics 18:799–801, 1998. 109. Glazer GM, Axel L, Moss AA: CT diagnosis of mediastinal thyroid. AJR Am J Roentgenol 138:495–498, 1982. 110. Glazer HS, Molina PL, Siegel MJ, et al: High-attenuation mediastinal masses on unenhanced CT. AJR Am J Roentgenol 156:45–50, 1991. 111. Glazer GM, Orringer MB, Chenevert TL, et al: Mediastinal lymph nodes: Relaxation time/pathologic correlation and implications in staging of lung cancer with MR imaging [see comments]. Radiology 168:429–431, 1988. 112. Glazer HS, Wick MR, Anderson DJ, et al: CT of fatty thoracic masses [see comments]. AJR Am J Roentgenol 159:1181–1187, 1992. 113. Goodman LR, Kay HR, Teplick SK, et al: Complications of median sternotomy: Computed tomographic evaluation. AJR Am J Roentgenol 141:225–230, 1983. 114. Gorkem SB, Coskun A, Yikilmaz A, et al: Evaluation of pediatric thoracic disorders: Comparison of unenhanced fast-imaging-sequence 1.5-T MRI and contrast-enhanced MDCT. AJR Am J Roentgenol 200:1352–1357, 2013. 115. Graeber GM, Thompson LD, Cohen DJ, et al: Cystic lesion of the thymus. An occasionally malignant cervical and/or anterior mediastinal mass. J Thorac Cardiovasc Surg 87:295–300, 1984. 116. Greene AK, Perlyn CA, Alomari AI: Management of lymphatic malformations. Clin Plast Surg 38:75–82, 2011. 117. Groves AM, Mohan HK, Wegner EA, et al: Positron emission tomography with FDG to show thymic carcinoid. AJR Am J Roentgenol 182:511–513, 2004. 118. Gumbs RV, Higginbotham-Ford EA, Teal JS, et al: Thoracic extramedullary hematopoiesis in sickle-cell disease. AJR Am J Roentgenol 149:889–893, 1987. 119. Halvorsen RA, Jr, Thompson WM: CT of esophageal neoplasms. Radiol Clin North Am 27:667–685, 1989. 120. Han HJ, Park SJ, Min KH, et al: Whole-body magnetic resonance imaging for staging metastatic thymic carcinoma. Am J Respir Crit Care Med 183:1573–1574, 2011. 121. Hara M, McAdams HP, Vredenburgh JJ, et al: Thymic hyperplasia after high-dose chemotherapy and autologous stem cell transplantation: Incidence and significance in patients with breast cancer. AJR Am J Roentgenol 173:1341–1344, 1999. 122. Harkin JC, Reed RJ: Atlas of tumor pathology: Tumors of the peripheral nervous system. Second series, Fascicle 3, Supplement, Washington, DC, 1982, Armed Forces Institute of Pathology. 123. Harris NL, Jaffe ES, Stein H, et al: A revised European-American classification of lymphoid neoplasms: A proposal from the International Lymphoma Study Group [see comments]. Blood 84:1361–1392, 1994. 124. Hartnell GG: Magnetic resonance angiography of systemic thoracic   and abdominal veins. Magn Reson Imaging Clin N Am 1:281–294, 1993. 125. Hartnell GG, Hughes LA, Finn JP, et al: Magnetic resonance angiography of the central chest veins. A new gold standard? Chest 107:1053–1057, 1995. 126. Heitzman ER: The mediastinum: Radiologic correlations with anatomy and pathology. Completely rev, ed 2, Berlin; New York, 1988, Springer-Verlag.

CHAPTER 38  Mediastinal Disease 127. Hengge UR, Ruzicka T, Tyring SK, et al: Update on Kaposi’s sarcoma and other HHV8 associated diseases. Part 2: pathogenesis, Castleman’s disease, and pleural effusion lymphoma. Lancet Infect Dis 2:344–352, 2002. 128. Herold CJ, Zerhouni EA: The mediastinum and lungs. In Higgins CB, Hricak H, Helms CA, editors: Magnetic resonance imaging of the body, ed 2, New York, 1992, Raven Press. 129. Higgins CB, Auffermann W: MR imaging of thyroid and parathyroid glands: A review of current status. AJR Am J Roentgenol 151:1095–1106, 1988. 130. Higgins C, Caputo G: Acquired heart disease. In Higgins CBHH, Helms CA, editors: Magnetic resonance imaging of the body, ed 3, New York, 1996, Lippincott Williams & Wilkins. 131. Higgins CB, McNamara MT, Fisher MR, et al: MR imaging of the thyroid. AJR Am J Roentgenol 147:1255–1261, 1986. 132. Hochhegger B, Marchiori E, Sedlaczek O, et al: MRI in lung cancer:   A pictorial essay. Br J Radiol 84:661–668, 2011. 133. Hoffman OA, Gillespie DJ, Aughenbaugh GL, et al: Primary mediastinal neoplasms (other than thymoma). Mayo Clin Proc 68:880–891, 1993. 134. Holscher AH, Dittler HJ, Siewert JR: Staging of squamous esophageal cancer: Accuracy and value. World J Surg 18:312–320, 1994. 135. Hope DG, Mulvihill JJ: Malignancy in neurofibromatosis. Adv Neurol 29:33–56, 1981. 136. Imran MB, Kubota K, Yoshioka S, et al: Sclerosing mediastinitis: Findings on fluorine-18 fluorodeoxyglucose positron emission tomography. Clin Nucl Med 24:305–308, 1999. 137. Inaoka T, Takahashi K, Mineta M, et al: Thymic hyperplasia and thymus gland tumors: Differentiation with chemical shift MR imaging. Radiology 243:869–876, 2007. 138. Inase N, Ichioka M, Akamatsu H, et al: Mediastinal fibromatosis presenting with superior vena cava syndrome. Respiration 66:464–466, 1999. 139. Indeglia RA, Shea MA, Grage TB: Congenital cysts of the thymus gland. Arch Surg 94:149–152, 1967. 140. Ishaque MA, Crockard HA, Stevens JM: Ossified pseudomeningocoele following laminectomy: Case reports and review of the literature. Eur Spine J 6:430–432, 1997. 141. Ishibashi M, Nishida H, Hiromatsu Y, et al: Comparison of technetium99m-MIBI, technetium-99m-tetrofosmin, ultrasound and MRI for localization of abnormal parathyroid glands. J Nucl Med 39:320–324, 1998. 142. Ishibashi M, Nishida H, Hiromatsu Y, et al: Localization of ectopic parathyroid glands using technetium-99m sestamibi imaging: Comparison with magnetic resonance and computed tomographic imaging. Eur J Nucl Med 24:197–201, 1997. 143. Ishibashi M, Nishida H, Kumabe T, et al: Tc-99m tetrofosmin. A new diagnostic tracer for parathyroid imaging. Clin Nucl Med 20:902–905, 1995. 144. Ishii K, Maeda K, Hashihira M, et al: MRI of mediastinal cavernous hemangioma. Pediatr Radiol 20:556–557, 1990. 145. James OG, Christensen JD, Wong TZ, et al: Utility of FDG PET/CT in inflammatory cardiovascular disease. Radiographics 31:1271–1286, 2011. 146. Jaramillo D, Perez-Atayde A, Griscom NT: Apparent association between thymic cysts and prior thoracotomy. Radiology 172:207–209, 1989. 147. Jeong YJ, Lee KS, Kim J, et al: Does CT of thymic epithelial tumors enable us to differentiate histologic subtypes and predict prognosis? AJR Am J Roentgenol 183:283–289, 2004. 148. Jerushalmi J, Frenkel A, Bar-Shalom R, et al: Physiologic thymic uptake of 18F-FDG in children and young adults: A PET/CT evaluation of incidence, patterns, and relationship to treatment. J Nucl Med 50:849–853, 2009. 149. Jochelson M, Mauch P, Balikian J, et al: The significance of the residual mediastinal mass in treated Hodgkin’s disease. J Clin Oncol 3:637–640, 1985. 150. Kaira K, Sunaga N, Ishizuka T, et al: The role of [(18)F] fluorodeoxyglucose positron emission tomography in thymic epithelial tumors. Cancer Imaging 11:195–201, 2011.

1089

151. Kang YS, Rosen K, Clark OH, et al: Localization of abnormal parathyroid glands of the mediastinum with MR imaging [see comments]. Radiology 189:137–141, 1993. 152. Kaplan PA, Williams SM: Mucocutaneous and peripheral soft-tissue hemangiomas: MR imaging. Radiology 163:163–166, 1987. 153. Katabathina VS, Restrepo CS, Martinez-Jimenez S, et al: Nonvascular, nontraumatic mediastinal emergencies in adults: A comprehensive review of imaging findings. Radiographics 31:1141–1160, 2011. 154. Kataoka K, Matsuura M, Seno N: Study of cystic mediastinal lesions— Diagnostic usefulness of magnetic resonance imaging. Nippon Kyobu Geka Gakkai Zasshi 43:438–445, 1995. 155. Kawabata H, Takai K, Kojima M, et al: Castleman-Kojima disease (TAFRO syndrome): A novel systemic inflammatory disease characterized by a constellation of symptoms, namely, thrombocytopenia, ascites (anasarca), microcytic anemia, myelofibrosis, renal dysfunction, and organomegaly: A status report and summary of Fukushima (6 June, 2012) and Nagoya meetings (22 September, 2012). J Clin Exp Hematop 53:57–61, 2013. 156. Kawashima A, Fishman EK, Kuhlman JE: CT and MR evaluation of posterior mediastinal masses. Crit Rev Diagn Imaging 33:311–367, 1992. 157. Kawashima A, Fishman EK, Kuhlman JE, et al: CT of posterior mediastinal masses. Radiographics 11:1045–1067, 1991. 158. Keller AR, Castleman B: Hodgkin’s disease of the thymus gland. Cancer 33:1615–1623, 1974. 159. Keller AR, Hochholzer L, Castleman B: Hyaline-vascular and plasmacell types of giant lymph node hyperplasia of the mediastinum and other locations. Cancer 29:670–683, 1972. 160. Kelly JD, Forster AM, Higley B, et al: Technetium-99m-tetrofosmin as a new radiopharmaceutical for myocardial perfusion imaging. J Nucl Med 34:222–227, 1993. 161. Kiernan PD, Hernandez A, Byrne WD, et al: Descending cervical mediastinitis [see comments]. Ann Thorac Surg 65:1483–1488, 1998. 162. Kim TJ, Kim HY, Lee KW, et al: Multimodality assessment of esophageal cancer: Preoperative staging and monitoring of response to therapy. Radiographics 29:403–421, 2009. 163. Kim TS, Lee KS, Han J, et al: Mucoepidermoid carcinoma of the tracheobronchial tree: Radiographic and CT findings in 12 patients. Radiology 212:643–648, 1999. 164. Kirwan WO, Walbaum PR, McCormack RJ: Cystic intrathoracic derivatives of the foregut and their complications. Thorax 28:424–428, 1973. 165. Klemm KM, Moran CA: Primary neuroendocrine carcinomas of the thymus. Semin Diagn Pathol 16:32–41, 1999. 166. Klemm KM, Moran CA, Suster S: Pigmented thymic carcinoids: A clinicopathological and immunohistochemical study of two cases. Mod Pathol 12:946–948, 1999. 167. Knapp RH, Hurt RD, Payne WS, et al: Malignant germ cell tumors of the mediastinum. J Thorac Cardiovasc Surg 89:82–89, 1985. 168. Koga K, Matsuno Y, Noguchi M, et al: A review of 79 thymomas: Modification of staging system and reappraisal of conventional division into invasive and non-invasive thymoma. Pathol Int 44:359–367, 1994. 169. Konen E, Merchant N, Gutierrez C, et al: True versus false left ventricular aneurysm: Differentiation with MR imaging–Initial experience. Radiology 236:65–70, 2005. 170. Krishnan J, Danon AD, Frizzera G: Reactive lymphadenopathies and atypical lymphoproliferative disorders. Am J Clin Pathol 99:385–396, 1993. 171. Krubsack AJ, Wilson SD, Lawson TL, et al: Prospective comparison of radionuclide, computed tomographic, sonographic, and magnetic resonance localization of parathyroid tumors. Surgery 106:639–644, discussion 644–646, 1989. 172. Kubota K, Yamada S, Kondo T, et al: PET imaging of primary mediastinal tumours. Br J Cancer 73:882–886, 1996. 173. Kuester LB, Fischman AJ, Fan CM, et al: Lipomatous hypertrophy of the interatrial septum: Prevalence and features on fusion 18F fluorodeoxyglucose positron emission tomography/CT. Chest 128:3888–3893, 2005.

1090

PART II  CT and MR Imaging of the Whole Body

174. Kumar AJ, Kuhajda FP, Martinez CR, et al: Computed tomography of extracranial nerve sheath tumors with pathological correlation. J Comput Assist Tomogr 7:857–865, 1983. 175. Kushihashi T, Munechika H, Motoya H, et al: CT and MR findings in tuberculous mediastinitis. J Comput Assist Tomogr 19:379–382, 1995. 176. Kwong JS, Muller NL, Miller RR: Diseases of the trachea and mainstem bronchi: Correlation of CT with pathologic findings. Radiographics 12:645–657, 1992. 177. Lanciego C, Pangua C, Chacon JI, et al: Endovascular stenting as the first step in the overall management of malignant superior vena cava syndrome. AJR Am J Roentgenol 193:549–558, 2009. 178. Latson L, Jr, Levsky JM, Haramati LB: Adult congenital heart disease: A practical approach. J Thorac Imaging 28:332–344, quiz 345–346, 2013. 179. Lazzarino M, Orlandi E, Paulli M, et al: Primary mediastinal B-cell lymphoma with sclerosis: An aggressive tumor with distinctive clinical and pathologic features [see comments]. J Clin Oncol 11:2306–2313, 1993. 180. LeBlanc J, Guttentag AR, Shepard JA, et al: Imaging of mediastinal foregut cysts. Can Assoc Radiol J 45:381–386, 1994. 181. Lee EY, Boiselle PM, Cleveland RH: Multidetector CT evaluation of congenital lung anomalies. Radiology 247:632–648, 2008. 182. Lee SH, Hur J, Kim YJ, et al: Additional value of dual-energy CT to differentiate between benign and malignant mediastinal tumors: an initial experience. Eur J Radiol 82:2043–2049, 2013. 183. Lee KS, Im JG, Han CH, et al: Malignant primary germ cell tumors of the mediastinum: CT features. AJR Am J Roentgenol 153:947–951, 1989. 184. Lee G, I H, Kim SJ, et al: Clinical Implication of PET/MR imaging in preoperative esophageal cancer staging: Comparison with PET/CT, endoscopic ultrasonography, and CT. J Nucl Med 2014, 1989. 185. Lee T, Lee SK, Chen CL, et al: Role of MR imaging of middle third esophageal carcinoma in determining resectability. Chung Hua I Hsueh Tsa Chih (Taipei) 52:363–367, 1993. 186. Lee ES, Paeng JC, Park CM, et al: Metabolic characteristics of Castleman disease on 18F-FDG PET in relation to clinical implication. Clin Nucl Med 38:339–342, 2013. 187. Lee VS, Spritzer CE: MR imaging of abnormal parathyroid glands. AJR Am J Roentgenol 170:1097–1103, 1998. 188. Lee VS, Spritzer CE, Coleman RE, et al: The complementary roles of fast spin-echo MR imaging and double-phase 99m Tc-sestamibi scintigraphy for localization of hyperfunctioning parathyroid glands. AJR Am J Roentgenol 167:1555–1562, 1996. 189. Lee EY, Tracy DA, Mahmood SA, et al: Preoperative MDCT evaluation of congenital lung anomalies in children: Comparison of axial, multiplanar, and 3D images. AJR Am J Roentgenol 196:1040–1046, 2011. 190. Lehr L, Rupp N, Siewert JR: Assessment of resectability of esophageal cancer by computed tomography and magnetic resonance imaging. Surgery 103:344–350, 1988. 191. Levitt RG, Husband JE, Glazer HS: CT of primary germ-cell tumors of the mediastinum. AJR Am J Roentgenol 142:73–78, 1984. 192. Lewis BD, Hurt RD, Payne WS, et al: Benign teratomas of the mediastinum. J Thorac Cardiovasc Surg 86:727–731, 1983. 193. Lewis JE, Wick MR, Scheithauer BW, et al: Thymoma. A clinicopathologic review. Cancer 60:2727–2743, 1987. 194. Lichtenstein AK, Levine A, Taylor CR, et al: Primary mediastinal lymphoma in adults. Am J Med 68:509–514, 1980. 195. Link KM, Samuels LJ, Reed JC, et al: Magnetic resonance imaging of the mediastinum. J Thorac Imaging 8:34–53, 1993. 196. Liu RS, Yeh SH, Huang MH, et al: Use of fluorine-18 fluorodeoxyglucose positron emission tomography in the detection of thymoma: a preliminary report. Eur J Nucl Med 22:1402–1407, 1995. 197. Lyon RD, McAdams HP: Mediastinal bronchogenic cyst: Demonstration of a fluid-fluid level at MR imaging [see comments]. Radiology 186:427–428, 1993. 198. Maleux G, Gillardin P, Fieuws S, et al: Large-bore nitinol stents for malignant superior vena cava syndrome: Factors influencing outcome. AJR Am J Roentgenol 201:667–674, 2013. 199. Marchevsky AM: Mediastinal tumors of peripheral nervous system origin. Semin Diagn Pathol 16:65–78, 1999.

200. Marchevsky AM, Wick MR: Pathology of the mediastinum, Cambridge, 2014, Cambridge University Press. 201. Marin ML, Romney BM, Franco K, et al: Bronchogenic cyst: A case report emphasizing the role of magnetic resonance imaging. J Thorac Imaging 6:43–46, 1991. 202. Marom EM: Advances in thymoma imaging. J Thorac Imaging 28:69–80, quiz 81–83, 2013. 203. Marom EM, Milito MA, Moran CA, et al: Computed tomography findings predicting invasiveness of thymoma. J Thorac Oncol 6:1274– 1281, 2011. 204. Martin G, Kleinsasser O: Neurogenic sarcomas of the neck in neurofibromatosis. Arch Otorhinolaryngol 232:273–283, 1981. 205. Masaoka A, Monden Y, Nakahara K, et al: Follow-up study of thymomas with special reference to their clinical stages. Cancer 48:2485–2492, 1981. 206. Mathisen DJ: Tracheal tumors. Chest Surg Clin N Am 6:875–898, 1996. 207. Mathisen DJ: Primary tracheal tumor management. Surg Oncol Clin N Am 8:307, 1999. 208. Maurer AH, Burshteyn M, Adler LP, et al: How to differentiate benign versus malignant cardiac and paracardiac 18F FDG uptake at oncologic PET/CT. Radiographics 31:1287–1305, 2011. 209. Mayerhoefer ME, Karanikas G, Kletter K, et al: Evaluation of diffusionweighted MRI for pretherapeutic assessment and staging of lymphoma: Results of a prospective study in 140 patients. Clin Cancer Res 20:2984–2993, 2014. 210. McAdams HP: Chest case of the day. Fibrosing mediastinitis. AJR Am J Roentgenol 165:189–190, 1995. 211. McAdams HP, Kirejczyk WM, Rosado-de-Christenson ML, et al: Bronchogenic cyst: Imaging features with clinical and histopathologic correlation. Radiology 217:441–446, 2000. 212. McAdams HP, Rosado-de-Christenson M, Fishback NF, et al: Castleman disease of the thorax: Radiologic features with clinical and histopathologic correlation. Radiology 209:221–228, 1998. 213. McAdams HP, Rosado-de-Christenson ML, Moran CA: Mediastinal hemangioma: Radiographic and CT features in 14 patients. Radiology 193:399–402, 1994. 214. McCarthy MJ, Rosado-de-Christenson ML: Tumors of the trachea. J Thorac Imaging 10:180–198, 1995. 215. McCarty MJ, Vukelja SJ, Banks PM, et al: Angiofollicular lymph node hyperplasia (Castleman’s disease). Cancer Treat Rev 21:291–310, 1995. 216. McElaney B, McAdams HP: Chest case of the day. Metastatic bronchial carcinoid. AJR Am J Roentgenol 165:186–187, 1995. 217. McLoughlin MJ, Weisbrod G, Wise DJ, et al: Computed tomography in congenital anomalies of the aortic arch and great vessels. Radiology 138:399–403, 1981. 218. McNeeley MF, Chung JH, Bhalla S, et al: Imaging of granulomatous fibrosing mediastinitis. AJR Am J Roentgenol 199:319–327, 2012. 219. Merine DS, Fishman EK, Zerhouni EA: Computed tomography and magnetic resonance imaging diagnosis of thymic cyst. J Comput Tomogr 12:220–222, 1988. 220. Meyer CA, Vagal AS, Seaman D: Put your back into it: pathologic conditions of the spine at chest CT. Radiographics 31:1425–1441, 2011. 221. Misawa Y, Fuse K, Hasegawa T: Infectious mediastinitis after cardiac operations: Computed tomographic findings. Ann Thorac Surg 65:622–624, 1998. 222. Moeller KH, Rosado-de-Christenson ML, Templeton PA: Mediastinal mature teratoma: Imaging features. AJR Am J Roentgenol 169:985–990, 1997. 223. Molinari F, Bankier AA, Eisenberg RL: Fat-containing lesions in adult thoracic imaging. AJR Am J Roentgenol 197:W795–W813, 2011. 224. Moon WK, Im JG, Han MC: Castleman’s disease of the mediastinum: MR imaging features. Clin Radiol 49:466–468, 1994. 225. Moore AV, Korobkin M, Olanow W, et al: Age-related changes in the thymus gland: CT-pathologic correlation. AJR Am J Roentgenol 141:241–246, 1983. 226. Moore AV, Korobkin M, Powers B, et al: Thymoma detection by mediastinal CT: Patient with myasthenia gravis. AJR Am J Roentgenol 138:217–222, 1982.

CHAPTER 38  Mediastinal Disease 227. Moran CA, Suster S, Fishback N, et al: Mediastinal paragangliomas. A clinicopathologic and immunohistochemical study of 16 cases. Cancer 72:2358–2364, 1993. 228. Moran CA, Suster S, Koss MN: Primary germ cell tumors of the mediastinum: III. Yolk sac tumor, embryonal carcinoma, choriocarcinoma, and combined nonteratomatous germ cell tumors of the mediastinum—A clinicopathologic and immunohistochemical study of 64 cases. Cancer 80:699–707, 1997. 229. Moran CA, Suster S, Perino G, et al: Malignant smooth muscle tumors presenting as mediastinal soft tissue masses. A clinicopathologic study of 10 cases. Cancer 74:2251–2260, 1994. 230. Moran CA, Suster S, Przygodzki RM, et al: Primary germ cell tumors of the mediastinum: II. Mediastinal seminomas—A clinicopathologic and immunohistochemical study of 120 cases. Cancer 80:691–698, 1997. 231. Moran CA, Walsh G, Suster S, et al: Thymomas II: A clinicopathologic correlation of 250 cases with a proposed staging system with emphasis on pathologic assessment. Am J Clin Pathol 137:451–461, 2012. 232. Moran CA, Weissferdt A, Kalhor N, et al: Thymomas I: A clinicopathologic correlation of 250 cases with emphasis on the World Health Organization schema. Am J Clin Pathol 137:444–450, 2012. 233. Morgenthaler TI, Brown LR, Colby TV, et al: Thymoma. Mayo Clin Proc 68:1110–1123, 1993. 234. Moss AA, Schnyder P, Thoeni RF, et al: Esophageal carcinoma: Pretherapy staging by computed tomography. AJR Am J Roentgenol 136:1051–1056, 1981. 235. Motwani M, Kidambi A, Herzog BA, et al: MR imaging of cardiac tumors and masses: A review of methods and clinical applications. Radiology 268:26–43, 2013. 236. Murayama S, Murakami J, Watanabe H, et al: Signal intensity characteristics of mediastinal cystic masses on T1- weighted MRI.   J Comput Assist Tomogr 19:188–191, 1995. 237. Naidich DP, Srichai MB: Computed tomography and magnetic resonance of the thorax, ed 4, Philadelphia, 2007, Wolters Kluwer/Lippincott Williams & Wilkins. 238. Naidich DP, Webb WR, Muller NL, et al: Computed tomography and magnetic resonance imaging of the thorax, ed 3, Philadelphia, 1999, Lippincott-Raven. 239. Nakagawa M, Hara M, Oshima H, et al: Pleural bronchogenic cysts: Imaging findings. J Thorac Imaging 23:284–288, 2008. 240. Nakasu Y, Minouchi K, Hatsuda N, et al: Thoracic meningocele in neurofibromatosis: CT and MR findings. J Comput Assist Tomogr 15:1062–1064, 1991. 241. Nakata H, Egashira K, Watanabe H, et al: MRI of bronchogenic cysts.   J Comput Assist Tomogr 17:267–270, 1993. 242. Nakazono T, White CS, Yamasaki F, et al: MRI findings of mediastinal neurogenic tumors. AJR Am J Roentgenol 197:W643–W652, 2011. 243. Nanson EM: Lymphangioma (cystic hygroma) of the mediastinum.   J Cardiovasc Surg (Torino) 9:447–452, 1968. 244. Nason LK, Walker CM, McNeeley MF, et al: Imaging of the diaphragm: Anatomy and function. Radiographics 32:E51–E70, 2012. 245. Nasseri F, Chen GJ, Nachiappan AC: Case 195: Chondrosarcoma of the posterior mediastinum. Radiology 268:299–303, 2013. 246. Nasseri F, Eftekhari F: Clinical and radiologic review of the normal and abnormal thymus: Pearls and pitfalls. Radiographics 30:413–428, 2010. 247. Negendank WG, al-Katib AM, Karanes C, et al: Lymphomas: MR imaging contrast characteristics with clinical-pathologic correlations [see comments]. Radiology 177:209–216, 1990. 248. Nichols CR: Mediastinal germ cell tumors. Clinical features and biologic correlates. Chest 99:472–479, 1991. 249. Nichols CR: Mediastinal germ cell tumors. Semin Thorac Cardiovasc Surg 4:45–50, 1992. 250. Nishino M, Ashiku SK, Kocher ON, et al: The thymus: A comprehensive review. Radiographics 26:335–348, 2006. 251. Norris EH: The parathyroid adenoma: A study of 322 cases. Int Abst Surg 84:1–41, 1947. 252. Nyman R, Forsgren G, Glimelius B: Long-term follow-up of residual mediastinal masses in treated Hodgkin’s disease using MR imaging. Acta Radiol 37:323–326, 1996.

1091

253. Oksenhendler E, Duarte M, Soulier J, et al: Multicentric Castleman’s disease in HIV infection: A clinical and pathological study of 20 patients. AIDS 10:61–67, 1996. 254. O’Regan KN, Jagannathan J, Krajewski K, et al: Imaging of liposarcoma: Classification, patterns of tumor recurrence, and response to treatment. AJR Am J Roentgenol 197:W37–W43, 2011. 255. O’Sullivan R, McKenzie A, Hennessy O: Value of CT scanning in assessing location and extent of epidural and paraspinal inflammatory conditions. Australas Radiol 32:203–206, 1988. 256. Park CM, Goo JM, Lee HJ, et al: Tumors in the tracheobronchial tree: CT and FDG PET features. Radiographics 29:55–71, 2009. 257. Patel IJ, Hsiao E, Ahmad AH, et al: AIRP best cases in radiologicpathologic correlation: Mediastinal mature cystic teratoma. Radiographics 33:797–801, 2013. 258. Patel NB, Stacy GS: Musculoskeletal manifestations of neurofibromatosis type 1. AJR Am J Roentgenol 199:W99–W106, 2012. 259. Patterson GA, Ginsberg RJ, Poon PY, et al: A prospective evaluation of magnetic resonance imaging, computed tomography, and mediastinoscopy in the preoperative assessment of mediastinal node status in bronchogenic carcinoma. J Thorac Cardiovasc Surg 94:679– 684, 1987. 260. Perelman MI, Koroleva N, Birjukov J, et al: Primary tracheal tumors. Semin Thorac Cardiovasc Surg 8:400–402, 1996. 261. Peters R, Peters O, Braak S, et al: Pathology of the thymus on CT-imaging. JBR-BTR 95:281–288, 2012. 262. Peterson BA, Frizzera G: Multicentric Castleman’s disease. Semin Oncol 20:636–647, 1993. 263. Picus D, Balfe DM, Koehler RE, et al: Computed tomography in the staging of esophageal carcinoma. Radiology 146:433–438, 1983. 264. Pilla TJ, Wolverson MK, Sundaram M, et al: CT evaluation of cystic lymphangiomas of the mediastinum. Radiology 144:841–842, 1982. 265. Poon PY, Bronskill MJ, Henkelman RM, et al: Mediastinal lymph node metastases from bronchogenic carcinoma: Detection with MR imaging and CT. Radiology 162:651–656, 1987. 266. Predey TA, McDonald V, Demos TC, et al: CT of congenital anomalies of the aortic arch. Semin Roentgenol 24:96–113, 1989. 267. Proto AV, Cuthbert NW, Raider L: Aberrant right subclavian artery: Further observations. AJR Am J Roentgenol 148:253–257, 1987. 268. Qu YJ, Liu GB, Shi HS, et al: Preoperative CT findings of thymoma are correlated with postoperative Masaoka clinical stage. Acad Radiol 20:66–72, 2013. 269. Quint LE, Glazer GM, Orringer MB: Esophageal imaging by MR and CT: Study of normal anatomy and neoplasms. Radiology 156:727–731, 1985. 270. Rabushka LS, Fishman EK, Kuhlman JE: CT evaluation of achalasia.   J Comput Assist Tomogr 15:434–439, 1991. 271. Radford JA, Cowan RA, Flanagan M, et al: The significance of residual mediastinal abnormality on the chest radiograph following treatment for Hodgkin’s disease [see comments]. J Clin Oncol 6:940–946, 1988. 272. Rainov NG, Heidecke V, Burkert W: Thoracic and lumbar meningocele in neurofibromatosis type 1. Report of two cases and review of the literature. Neurosurg Rev 18:127–134, 1995. 273. Rajiah P: Cardiac MRI: Part 2, pericardial diseases. AJR Am J Roentgenol 197:W621–W634, 2011. 274. Rajiah P, Desai MY, Kwon D, et al: MR imaging of myocardial infarction. Radiographics 33:1383–1412, 2013. 275. Rajiah P, Kanne JP: Computed tomography of the pericardium and pericardial disease. J Cardiovasc Comput Tomogr 4:3–18, 2010. 276. Ramos-Duran L, Nance JW, Jr, Schoepf UJ, et al: Developmental aortic arch anomalies in infants and children assessed with CT angiography. AJR Am J Roentgenol 198:W466–W474, 2012. 277. Reddy MP, Graham MM: FDG positron emission tomographic   imaging of thoracic Castleman’s disease. Clin Nucl Med 28:325–326, 2003. 278. Reed JC, Hallet KK, Feigin DS: Neural tumors of the thorax: Subject review from the AFIP. Radiology 126:9–17, 1978. 279. Reinig JW, Stanley JH, Schabel SI: CT evaluation of thickened esophageal walls. AJR Am J Roentgenol 140:931–934, 1983.

1092

PART II  CT and MR Imaging of the Whole Body

280. Restrepo CS, Vargas D, Ocazionez D, et al: Primary pericardial tumors. Radiographics 33:1613–1630, 2013. 281. Rholl KS, Levitt RG, Glazer HS: Magnetic resonance imaging of fibrosing mediastinitis. AJR Am J Roentgenol 145:255–259, 1985. 282. Riddell AM, Hillier J, Brown G, et al: Potential of surface-coil MRI for staging of esophageal cancer. AJR Am J Roentgenol 187:1280–1287, 2006. 283. Robicsek F: Postoperative sterno-mediastinitis. Am Surg 66:184–192, 2000. 284. Rodriguez E, Soler R, Pombo F, et al: Fibrosing mediastinitis: CT and MR findings. Clin Radiol 53:907–910, 1998. 285. Rosado de Christenson ML, Abbott GF, Kirejczyk WM, et al: Thoracic carcinoids: Radiologic-pathologic correlation. Radiographics 19:707– 736, 1999. 286. Rosado-de-Christenson ML, Galobardes J, Moran CA: Thymoma: Radiologic-pathologic correlation. Radiographics 12:151–168, 1992. 287. Rosado-de-Christenson ML, Pugatch RD, Moran CA, et al: Thymolipoma: Analysis of 27 cases. Radiology 193:121–126, 1994. 288. Rosado-de-Christenson ML, Templeton PA, Moran CA: From the archives of the AFIP. Mediastinal germ cell tumors: Radiologic and pathologic correlation. Radiographics 12:1013–1030, 1992. 289. Rossi SE, Erasmus JJ, McAdams HP, et al: Thoracic manifestations of neurofibromatosis-I. AJR Am J Roentgenol 173:1631–1638, 1999. 290. Rossi SE, McAdams HP, Rosado-de-Christenson ML, et al: Fibrosing mediastinitis. Radiographics 21:737–757, 2001. 291. Ruzal-Shapiro C, Abramson SJ, Berdon WE: Posterior mediastinal cystic teratoma surrounded by fat in a 13-month-old boy. Value of magnetic resonance imaging. Pediatr Radiol 20:107–109, 1989. 292. Sadohara J, Fujimoto K, Muller NL, et al: Thymic epithelial tumors: Comparison of CT and MR imaging findings of low-risk thymomas, high-risk thymomas, and thymic carcinomas. Eur J Radiol 60:70–79, 2006. 293. Sakai F, Sone S, Kiyono K, et al: Intrathoracic neurogenic tumors: MR-pathologic correlation. AJR Am J Roentgenol 159:279–283, 1992. 294. Sakai F, Sone S, Kiyono K, et al: MR imaging of thymoma: Radiologicpathologic correlation. AJR Am J Roentgenol 158:751–756, 1992. 295. Sakurai K, Hara M, Ozawa Y, et al: Thoracic hemangiomas: Imaging via CT, MR, and PET along with pathologic correlation. J Thorac Imaging 23:114–120, 2008. 296. Salyer DC, Salyer WR, Eggleston JC: Benign developmental cysts of the mediastinum. Arch Pathol Lab Med 101:136–139, 1977. 297. Sasaka K, Kurihara Y, Nakajima Y, et al: Spontaneous rupture: A complication of benign mature teratomas of the mediastinum. AJR Am J Roentgenol 170:323–328, 1998. 298. Sasaki M, Kuwabara Y, Ichiya Y, et al: Differential diagnosis of thymic tumors using a combination of 11C-methionine PET and FDG PET.   J Nucl Med 40:1595–1601, 1999. 299. Satava RM, Jr, Beahrs OH, Scholz DA: Success rate of cervical exploration for hyperparathyroidism. Arch Surg 110:625–628, 1975. 300. Savader SJ, Otero RR, Savader BL: MR imaging of intrathoracic extramedullary hematopoiesis. J Comput Assist Tomogr 12:878–880, 1988. 301. Schurawitzki H, Stiglbauer R, Klepetko W, et al: CT and MRI in benign mediastinal haemangioma. Clin Radiol 43:91–94, 1991. 302. Seline TH, Gross BH, Francis IR: CT and MR imaging of mediastinal hemangiomas. J Comput Assist Tomogr 14:766–768, 1990. 303. Shaffer K, Rosado-de-Christenson ML, Patz EF, Jr, et al: Thoracic lymphangioma in adults: CT and MR imaging features. AJR Am J Roentgenol 162:283–289, 1994. 304. Shaffer K, Smith D, Kirn D, et al: Primary mediastinal large-B-cell lymphoma: Radiologic findings at presentation. AJR Am J Roentgenol 167:425–430, 1996. 305. Shahian DM, Rossi RL: Posterior mediastinal goiter. Chest 94:599–602, 1988. 306. Sherrick AD, Brown LR, Harms GF, et al: The radiographic findings of fibrosing mediastinitis. Chest 106:484–489, 1994. 307. Sheth S, Ebert MD, Fishman EK: Superior vena cava obstruction evaluation with MDCT. AJR Am J Roentgenol 194:W336–W346, 2010.

308. Shields TW, Reynolds M: Neurogenic tumors of the thorax. Surg Clin North Am 68:645–668, 1988. 309. Shin MS, Berland LL, Ho KJ: Mediastinal cystic hygromas: CT characteristics and pathogenetic consideration. J Comput Assist Tomogr 9:297–301, 1985. 310. Shin MS, Ho KJ: Computed tomography of primary mediastinal seminomas. J Comput Assist Tomogr 7:990–994, 1983. 311. Shirkhoda A, Chasen MH, Eftekhari F, et al: MR imaging of mediastinal thymolipoma. J Comput Assist Tomogr 11:364–365, 1987. 312. Silverman PM, Korobkin M, Moore AV: CT diagnosis of cystic hygroma of the neck. J Comput Assist Tomogr 7:519–520, 1983. 313. Sirivella S, Ford WB, Zikria EA, et al: Foregut cysts of the mediastinum. Results in 20 consecutive surgically treated cases. J Thorac Cardiovasc Surg 90:776–782, 1985. 314. Snyder ME, Luck SR, Hernandez R, et al: Diagnostic dilemmas of mediastinal cysts. J Pediatr Surg 20:810–815, 1985. 315. So CB, Li DK: Anterolateral cervical meningocele in association with neurofibromatosis: MR and CT studies. J Comput Assist Tomogr 13:692–695, 1989. 316. Soulier J, Grollet L, Oksenhendler E, et al: Kaposi’s sarcoma-associated herpesvirus-like DNA sequences in multicentric Castleman’s disease. Blood 86:1276–1280, 1995. 317. Spiers AS, Husband JE, MacVicar AD: Treated thymic lymphoma: Comparison of MR imaging with CT. Radiology 203:369–376, 1997. 318. Spritzer C, Gamsu G, Sostman HD: Magnetic resonance imaging of the thorax: Techniques, current applications, and future directions. J Thorac Imaging 4:1–18, 1989. 319. Spritzer CE, Gefter WB, Hamilton R, et al: Abnormal parathyroid glands: High-resolution MR imaging. Radiology 162:487–491, 1987. 320. Stark DD, Moss AA, Brasch RC, et al: Neuroblastoma: Diagnostic imaging and staging. Radiology 148:101–105, 1983. 321. Stephens AW, Gonin R, Hutchins GD, et al: Positron emission tomography evaluation of residual radiographic abnormalities in postchemotherapy germ cell tumor patients. J Clin Oncol 14:1637–1641, 1996. 322. Stevens SK, Chang JM, Clark OH, et al: Detection of abnormal parathyroid glands in postoperative patients with recurrent hyperparathyroidism: Sensitivity of MR imaging. AJR Am J Roentgenol 160:607–612, 1993. 323. Stojanovska J, Cascade PN, Chong S, et al: Embryology and imaging review of aortic arch anomalies. J Thorac Imaging 27:73–84, 2012. 324. Strickler JG, Kurtin PJ: Mediastinal lymphoma. Semin Diagn Pathol 8:2–13, 1991. 325. Strollo DC, Rosado de Christenson ML, Jett JR: Primary mediastinal tumors. Part 1: Tumors of the anterior mediastinum. Chest 112:511– 522, 1997. 326. Strollo DC, Rosado-de-Christenson ML, Jett JR: Primary mediastinal tumors: Part II. Tumors of the middle and posterior mediastinum. Chest 112:1344–1357, 1997. 327. Strollo DC, Rosado-de-Christenson ML: Tumors of the thymus.   J Thorac Imaging 14:152–171, 1999. 328. Sung YM, Lee KS, Kim BT, et al: 18F-FDG PET/CT of thymic epithelial tumors: Usefulness for distinguishing and staging tumor subgroups.   J Nucl Med 47:1628–1634, 2006. 329. Suster S, Moran CA: Thymoma classification: Current status and future trends. Am J Clin Pathol 125:542–554, 2006. 330. Suster S, Moran CA: Histologic classification of thymoma: The World Health Organization and beyond. Hematol Oncol Clin North Am 22:381–392, 2008. 331. Suster S, Rosai J: Multilocular thymic cyst: An acquired reactive process. Study of 18 cases [see comments]. Am J Surg Pathol 15:388–398, 1991. 332. Swanson PE: Soft tissue neoplasm of the mediastinum. Semin Diagn Pathol 8:14–34, 1991. 333. Taieb D, Neumann H, Rubello D, et al: Modern nuclear imaging for paragangliomas: Beyond SPECT. J Nucl Med 53:264–274, 2012. 334. Takashima S, Takeuchi N, Shiozaki H, et al: Carcinoma of the esophagus: CT vs MR imaging in determining resectability. AJR Am J Roentgenol 156:297–302, 1991.

CHAPTER 38  Mediastinal Disease 335. Tam CG, Broome DR, Shannon RL: Desmoid tumor of the anterior mediastinum: CT and radiologic features. J Comput Assist Tomogr 18:499–501, 1994. 336. Tan A, Holdener GP, Hecht A, et al: Malignant thymoma in an ectopic thymus: CT appearance. J Comput Assist Tomogr 15:842–844, 1991. 337. Templeton PA, Fishman EK: CT evaluation of poststernotomy complications. AJR Am J Roentgenol 159:45–50, 1992. 338. Tesoro-Tess JD, Balzarini L, Ceglia E, et al: Magnetic resonance imaging in the initial staging of Hodgkin’s disease and non-Hodgkin lymphoma. Eur J Radiol 12:81–90, 1991. 339. Tomiyama N, Honda O, Tsubamoto M, et al: Anterior mediastinal tumors: Diagnostic accuracy of CT and MRI. Eur J Radiol 69:280–288, 2009. 340. Truong MT, Erasmus JJ, Gladish GW, et al: Anatomy of pericardial recesses on multidetector CT: Implications for oncologic imaging. AJR Am J Roentgenol 181:1109–1113, 2003. 341. Umeoka S, Koyama T, Togashi K, et al: Esophageal cancer: Evaluation with triple-phase dynamic CT–initial experience. Radiology 239:777– 783, 2006. 342. Usuda K, Zhao XT, Sagawa M, et al: Diffusion-weighted imaging is superior to positron emission tomography in the detection and nodal assessment of lung cancers. Ann Thorac Surg 91:1689–1695, 2011. 343. Verstandig AG, Epstein DM, Miller WT, Jr, et al: Thymoma—Report of 71 cases and a review. Crit Rev Diagn Imaging 33:201–230, 1992. 344. Wang YM, Li YW, Sheih CP, et al: Magnetic resonance imaging of neuroblastoma, ganglioneuroblastoma, and ganglioneuroma. Chung Hua Min Kuo Hsiao Erh Ko I Hsueh Hui Tsa Chih 36:420–424, 1995. 345. Webb WR, Gamsu G, Speckman JM, et al: Computed tomographic demonstration of mediastinal venous anomalies. AJR Am J Roentgenol 139:157–161, 1982. 346. Webb WR, Gamsu G, Stark DD, et al: Magnetic resonance imaging of the normal and abnormal pulmonary hila. Radiology 152:89–94, 1984. 347. Weidner N: Germ-cell tumors of the mediastinum. Semin Diagn Pathol 16:42–50, 1999.

1093

348. Weinstein JB, Aronberg DJ, Sagel SS: CT of fibrosing mediastinitis: Findings and their utility. AJR Am J Roentgenol 141:247–251, 1983. 349. Westerterp M, van Westreenen HL, Reitsma JB, et al: Esophageal cancer: CT, endoscopic US, and FDG PET for assessment of response to neoadjuvant therapy—Systematic review. Radiology 236:841–851, 2005. 350. White CS: MR imaging of thoracic veins. Magn Reson Imaging Clin N Am 8:17–32, 2000. 351. White CS, Baffa JM, Haney PJ, et al: MR imaging of congenital anomalies of the thoracic veins. Radiographics 17:595–608, 1997. 352. White CS, Templeton PA, Attar S: Esophageal perforation: CT findings [see comments]. AJR Am J Roentgenol 160:767–770, 1993. 353. Whitten CR, Khan S, Munneke GJ, et al: A diagnostic approach to mediastinal abnormalities. Radiographics 27:657–671, 2007. 354. Wilkins EW, Jr, Edmunds LH, Jr, Castleman B: Cases of thymoma at the Massachusetts General Hospital. J Thorac Cardiovasc Surg 52:322–330, 1966. 355. Wittram C, Fischman AJ, Mark E, et al: Thymic enlargement and FDG uptake in three patients: CT and FDG positron emission tomography correlated with pathology. AJR Am J Roentgenol 180:519–522, 2003. 356. Wychulis AR, Payne WS, Clagett OT, et al: Surgical treatment of mediastinal tumors: A 40 year experience. J Thorac Cardiovasc Surg 62:379–392, 1971. 357. Yedlicka JW, Schultz K, Moncada R, et al: CT findings in superior vena cava obstruction. Semin Roentgenol 24:84–90, 1989. 358. Yeh HC, Gordon A, Kirschner PA, et al: Computed tomography and sonography of thymolipoma. AJR Am J Roentgenol 140:1131–1133, 1983. 359. Zadvinskis DP, Benson MT, Kerr HH, et al: Congenital malformations of the cervicothoracic lymphatic system: Embryology and pathogenesis. Radiographics 12:1175–1189, 1992. 360. Zinzani PL, Zompatori M, Bendandi M, et al: Monitoring bulky mediastinal disease with gallium-67, CT-scan and magnetic resonance imaging in Hodgkin’s disease and high-grade non-Hodgkin’s lymphoma. Leuk Lymphoma 22:131–135, 1996.

39  Disease of the Pleura, Chest Wall, and Diaphragm Asha Sarma and Ritu R. Gill

NORMAL ANATOMY Pleura and Chest Wall The pleura is a thin, two-layered serous membrane that lines the lungs and thoracic cavity. The visceral pleura lines the lungs and invaginates into a double layer to form the major and minor fissures, and the parietal pleura lines the chest wall, mediastinum, and diaphragmatic surface (sparing the hila).52 The pleural space is a potential space between the parietal and visceral pleurae. Under physiologic conditions, the pleural space of each hemithorax contains approximately 8 mL of acellular, clear fluid that lubricates and thus facilitates movement of the lungs in respiration.52,65 Normally, there is no communication between the pleural spaces of the right and left hemithoraces. The visceral and parietal pleura have different nodal drainage patterns in malignant states.52 External to the parietal pleura is a variable amount of extrapleural fat that separates the visceral pleura and the endothoracic fascia that covers the surface of the intercostal muscles and ribs (Fig. 39-1).

Diaphragm The diaphragm is a complex muscle that separates the thorax from the abdomen. It is the primary muscle of respiration. The four major embryologic structures that develop into the diaphragm are the paired pleuroperitoneal folds, the esophageal mesentery, the transverse septum, and the muscular body wall. The diaphragm attaches to numerous sites on the body wall. The two crura are attached posteriorly to the upper lumbar spine and are connected by the median arcuate ligament. Anteriorly and laterally, the diaphragm is attached to the sternum, xiphoid process, lower six ribs, and costal cartilages. The three openings in the diaphragm that allow passage of structures from thorax to abdomen include the inferior vena caval hiatus at the level of T8, the esophageal hiatus at T10 that contains the esophagus and vagus nerve, and the aortic hiatus at T12, which contains the aorta, thoracic duct, and azygos and hemiazygous veins.28,94

PLEURA Approach to Imaging the Pleura Radiography.  In general, when thoracic or pleural disease is suspected, chest radiography is the initial imaging study of choice. In normal patients, the pleura is seen on chest radiographs only where the visceral pleura has infolded to produce a double layer at the major or oblique fissures of both hemithoraces and the minor or horizontal fissure of the right hemithorax. Accessory fissures are variably present, the most common of which are the azygos fissure of the right upper lobe, the inferior accessory fissure of either lower lobe, the superior accessory fissure of either lower lobe, and the left minor fissure. In cases of pneumothorax, a thin “visceral pleural line” demarcates the superior

1094

or lateral edge of the collapsed lung. Fissures are visible only where they are tangential to the x-ray beam, and they may be “incomplete” in normal variants where they do not extend all the way to the hilum or chest wall.50,123 The presence of the incomplete border sign is a clue that a mass is extrapulmonary in the pleura or chest wall. The inner border of the radiopaque mass is tangential to the x-ray beam and is well defined by the adjacent lucent lung. The outer margin is ill defined because it is seen en face or partially en face33 (Fig. 39-2).

Ultrasound.  Ultrasound is frequently used to distinguish pleural thickening from effusion, characterize and quantify effusion, and guide procedures such as thoracentesis. As in other applications, ultrasound is inexpensive and portable, making it useful in a wide variety of venues including outpatient, inpatient, procedural, and critical care settings. The normal visceral pleura is seen as a bright specular reflector between the chest wall and the air-filled lung. Horizontal repetitions of the pleural line, or reverberation artifacts called A-lines, are seen deep to the visceral pleural line.53 Small inhomogeneities in the pleura produce “comet tails” or moving reverberation artifacts that extend deep from the pleural surface during normal respiratory motion. Lung sliding, or a shimmering movement at the pleural stripe–chest wall interface, is seen during normal respiratory movement and disappears along with comet tails in cases of pneumothorax.71 Like the liver, the lung has a dual arterial supply and thus can be imaged using contrast-enhanced sonography (CES).49 Processes that have a pulmonary arterial vascular supply show short time to enhancement and marked parenchymal enhancement (e.g., pneumonia), whereas lesions supplied by the bronchial arteries (e.g., malignancy) show delayed time to enhancement and reduced enhancement during the arterial and parenchymal phases. In pulmonary embolism, there is decreased pulmonary arterial vascularity.49 In imaging the chest wall and pleura, higher-frequency linear array transducers (e.g., 7.5 MHz) provide the greatest spatial resolution; however, in larger patients or in patients with large-volume pleural disease, lower-frequency transducers are able to provide greater penetration. Variable-frequency 3.5- to 5-MHz sector transducers with small footprints are most versatile because they fit between the ribs and may be used to guide interventional procedures while still providing excellent image quality (Fig. 39-3).

Computed Tomography.  Following chest radiography findings that raise suspicion for pleural disease requiring further evaluation, computed tomography (CT) is usually the study choice owing to its wide availability, relatively low cost, efficiency, and ability to image the upper abdomen and lung parenchyma.52 Spatial resolution of CT is superior to that of magnetic resonance imaging (MRI). Unlike ultrasound and

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

1095

B FIG 39-1  A, Inverted coronal CT image showing the fissures (arrows) and the diaphragm (arrowheads). B, Parietal pleura (thick arrow) and visceral pleura (thin arrow).

B

A

FIG 39-2  Chest PA (A) and lateral (B) radiographs showing a right upper pleural mass (arrows). Note that the inner border of the mass is tangential to the x-ray beam and is well defined by the adjacent lucent lung; the outer margin is ill defined.

MRI, CT exposes the patient to ionizing radiation, which may be a relative drawback in young patients, especially those who are expected to incur high cumulative radiation doses from serial imaging. Multidetector CT (MDCT) replaces the linear array of detector elements in a conventional CT scanner with a two-dimensional (2D) array, allowing acquisition of data from multiple sections simultaneously; this faster acquisition decreases respiratory motion artifact and improves image quality.

In evaluation of pleural pathology, coronal and sagittal reformations from thin-section (3 mm) axial images provide valuable information. MDCT acquisitions allow for application of advanced three-dimensional (3D) postprocessing techniques, which may be valuable for applications such as assessment of violation of critical structures by tumor, volumetric tumor quantification using serial segmentation, and surgical planning. Maximum intensity projection (MIP), which displays the highest-attenuation voxels in a given

1096

PART II  CT and MR Imaging of the Whole Body

TRANS

CB/SBA OR

FIG 39-3  Ultrasound image of the pleura in a patient with mesothelioma; arrow points to the parietal pleura. Note the horizontal repetitions of the pleural line, seen deep to the visceral pleura.

the study objectives. T1-weighted images are known for their excellent anatomic resolution and ability to resolve between the pleural space and extrapleural fat. T2-weighted images allow differentiation of pleural thickening and nodularity from pleural fluid and help differentiate tumor from adjacent muscle. Post-gadolinium T1-weighted images with fat saturation increase tissue contrast and are often helpful in identifying the borders between lesions and normal structures. Short tau inversion recovery (STIR) images are sensitive for bone involvement by tumor or infection, denoted by fluid signal in place of normal dark to intermediate marrow signal and breach of the normal dark cortical line.52 Parallel MRI enables assessment of tumor mobility and local lung motion. This, and sequences such as steady-state free precession (SSFP) that allow for cine imaging with high temporal resolution, may reveal subtle invasion of mediastinal structures and the chest wall. Additional sequences that have utility in selected cases of tumor assessment include diffusion-weighted imaging (DWI) and DCE (perfusion) MRI, which are subsequently explained in depth (Fig. 39-5).

Positron Emission Tomography.  PET with the radiotracer 18

selection, are helpful in evaluating high-density structures such as contrast-enhanced vessels and stents, whereas minimum intensity projections (MiniP) that display the lowest-attenuation voxels are useful for assessing air-filled structures such as the lung parenchyma and airways40,45,46 (Fig. 39-4). Intravenous (IV) contrast is vital in assessing pleural pathology other than pleural plaques and pneumothoraces. For example, without IV contrast, it is difficult to distinguish pleural malignancy or pleura thickened by empyema from adjacent consolidated lung or pleural effusion.102 Imaging during the late venous or tissue parenchymal phase (20-60 seconds after peripheral venous injection) is ideal.51,52 Dynamic contrast-enhanced (DCE; perfusion) CT to assess blood flow dynamics and capillary permeability may be useful in some oncologic evaluations.45 Mediastinal window settings are optimal for evaluation for pleural and chest wall disease. It is important to evaluate the images in lung windows to assess for associated pulmonary pathology, and bone windows are essential for detecting rib destruction. A sharp convolution kernel reconstruction algorithm should be used for lung and bone evaluation. High-resolution CT for evaluation of the lung parenchyma in the context of asbestos-related pleural disease has been largely replaced by MDCT.

Magnetic Resonance Imaging.  MRI provides superior contrast resolution to CT and may be used for more refined characterization of tissues. It is the modality of choice for evaluating invasion of the chest wall, diaphragm, heart, vessels, airways, and other critical structures by tumor or infection. MRI is not useful to evaluate the air-filled lung parenchyma, and acquisition times are substantially longer than those for CT. Medically unstable patients or those with implanted devices such as pacemakers may not be able to undergo MRI. Unlike CT and positron emission tomography (PET), MRI does not expose the patient to ionizing radiation, and MRI may be useful in patients with allergies to iodinated contrast. In the thorax, limitations of MRI include respiratory and cardiac motion artifacts and susceptibility artifact from gas-containing structures, though many tools that mitigate these factors (e.g., cardiac gating) are available. Large field-of-view scout images are obtained using a body coil, and specialized coils are selected for further imaging. Pulse sequences, imaging planes, and parameters such as slice thickness and interslice gap are specified according to the anatomic range being imaged and

F-fluorodeoxyglucose (FDG) complements CT and MRI in evaluation of pleural and chest wall malignancies. PET allows for evaluation of metabolic activity of tumor, which concentrates FDG to a greater degree than normal tissue. This type of assessment is useful in the context of pretreatment planning, staging, and follow-up. Furthermore, different tumor subtypes have shown differences in FDG avidity—for example, more indolent epithelioid mesothelioma tends to be less metabolically active than more aggressive sarcomatoid mesothelioma.45 Poor spatial resolution of PET is largely compensated for by co-registration with CT (PET-CT). Hybrid PET-CT systems that are capable of performing PET and CT in a single session have come into widespread use.45 Integrated PET-CT is superior to PET and CT alone in the diagnosis and staging of thoracic malignancies. Importantly, PET-CT is superior in identifying patients with occult metastases and unresectable tumors.45 PET-CT is also very useful for distinguishing malignancy from atelectasis, infection, and other benign processes45 (Fig. 39-6).

Pathology

Apical Pleural Cap.  Apical opacities, or apical caps, are a common nonspecific finding that may be idiopathic or seen in various disease states including infection, inflammatory conditions, reaction to radiotherapy, neoplasms such as Pancoast tumor, mediastinal hemorrhage, mediastinal lipomatosis, and peripheral upper lobe collapse.55 On chest radiographs, idiopathic asymptomatic apical pleural caps are irregular homogeneous densities less than 5 mm in thickness over the lung apex. The undersurface is usually sharply marginated.81 Apical pleural caps are as commonly unilateral as bilateral and are increasingly common with aging; they are found in as many as 16% of patients over age 45.103 When they are bilateral, apical pleural caps are usually asymmetric.103 Apical opacities related to tuberculosis tend to be thicker than idiopathic scars, measuring up to 2.5 cm in thickness. CT has shown these to be composed of thickened extrapleural fat.55 On CT, apical pleural caps may appear more irregular than on radiographs and are occasionally difficult to distinguish from spiculated lung cancers.50 Compared with tumors, however, idiopathic caps have been shown to be smaller in craniocaudal than axial dimensions.126 Superior sulcus or Pancoast tumor should be considered in the setting of local pain or underlying bone destruction.81

Pneumothorax.  Pneumothorax is defined as gas within the pleural space, which may or may not be under tension. Pneumothorax occurs

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A B

R C D

E FIG 39-4  A, Axial CT image in a normal individual. B, Coronal CT image with left apical pleural tumor. C, Sagittal CT image with fissural involvement. D, Volume-rendered image shows left-sided pleural tumor and its relationship to the left internal mammary vessels. E, Volumetric assessment using 3D volumerendering technique.

1097

1098

PART II  CT and MR Imaging of the Whole Body

A

B

D

E

C

FIG 39-5  MRIs showing pleural tumor. A, Postcontrast axial VIBE image showing enhancing left-sided mesothelioma. B, Axial postcontrast FAME image showing tumor abutting the descending aorta without invasion. Note the intramural thrombus involving the descending aorta and also left rib metastases. C, Coronal multiparametric map of a left pleural tumor. D, ADC map of a left-sided pleural tumor showing restricted diffusion and an ADC value of 1.01 × 10−3 mm2/sec. E, Coronal image showing transmural diaphragmatic extension of the right pleural tumor.

A

Fused coronals

B

FIG 39-6  A, Coronal fused FDG PET-CT image with localized tumor within the left hemithorax. B, Tumor extending into the abdomen and mediastinal node involvement.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm spontaneously in otherwise healthy individuals and in trauma and iatrogenic injury (e.g., from central line placements, biopsies, and mechanical ventilation). Predisposing conditions include emphysema, pulmonary fibrosis, cystic fibrosis, cavitary pneumonia, and cystic interstitial lung diseases such as Langerhans cell histiocytosis and lymphangiomyomatosis.97 Chest radiography is the first-line imaging test in patients with suspected pneumothorax. The classic chest radiography finding of pneumothorax is a thin visceral pleural line beyond which normal lung markings do not extend. In supine patients, gas may collect at the lung bases, including the costophrenic angles (deep sulcus sign), or adjacent to the heart in the cardiophrenic recess.97 At times, it may be difficult to differentiate bullae, the medial border of the scapula, skin folds, or objects external to the patient from pneumothorax, but knowledge of these potential confounders is helpful in radiographic interpretation. Expiratory and decubitus radiographs accentuate small pneumothoraces97 (Fig. 39-7). Ultrasound is a useful adjunct to chest radiography, particularly in critically ill patients who may require evaluation at the bedside.70 Ultrasound has been shown to be more sensitive for pneumothorax than chest radiography in postbiopsy patients. Typical signs of pneumothorax on ultrasound include loss of the normal comet tail and lung sliding signs. False positives on ultrasound may be seen in patients with impaired lung function. It is not possible to quantify the volume of pneumothorax using ultrasound, and comprehensive assessment of the entire thorax, as is possible with radiographs and cross-sectional imaging, is not feasible. CT detects pneumothoraces that are below the limit of detection of chest radiography and is helpful to distinguish pneumothorax from bullous disease. Furthermore, CT may be used to assess the degree of loculation in complicated pleural collections, guide drainage of pneu-

A

1099

mothoraces, and assess positioning of drainage catheters. Multiplanar reformations and thin-section technique are often useful. IV contrast is not necessary for evaluation of pneumothorax. Given susceptibility artifact produced by gas, MRI has no significant role in the evaluation of pneumothoraces.

Pleural Effusion.  The normal volume of fluid in each pleural space is approximately 8 mL.95 Pleural fluid volumes increase in myriad pathologic processes. Broadly, pleural effusions are divided into transudates and exudates. Transudates are caused by imbalanced hydrostatic or oncotic forces between lymphovascular structures and the pleural space, whereas exudates are caused by altered permeability of the lymphovascular structures themselves. Table 39-1 lists a selection of conditions that cause exudative and transudative effusions. Biochemical pleural fluid evaluation (e.g., using Light’s criteria; Box 39-1) helps differentiate exudate from transudate.73 When diagnostic thoracentesis is performed for a pleural fluid collection of unknown etiology, typical basic laboratory evaluation includes Gram stain, culture, cell count and differential, lactate dehydrogenase (LDH), glucose, pH, specific gravity, and protein. Fluid should be sent for cytologic evaluation if malignancy is in the differential diagnosis. Cases that may call for specialized lab testing include chylothorax, pancreatitis-related effusion, esophageal rupture, and tuberculous infection. Malignant effusions and empyema will be discussed in greater depth in subsequent sections. Imaging of pleural effusion.  In patients with suspected pleural effusion, chest radiography is the first-line test and ultrasound is a useful adjunct. Many cases can be managed using radiography and ultrasound alone. CT is useful in cases requiring comprehensive assessment of the entire pleura and extrapleural structures. The roles of

B FIG 39-7  Inspiration (A) and expiration (B) chest radiographs showing a right tension pneumothorax. Note shift of mediastinum to the left with a large right pneumothorax and complete atelectasis of the right lung.

1100

PART II  CT and MR Imaging of the Whole Body

TABLE 39-1  Some Conditions That Cause Transudative and Exudative Effusions (or Either) Transudate

Exudate

Either Exudate or Transudate

Congestive heart failure Cirrhosis/hypoalbuminemia Nephrotic syndrome Cardiac tamponade Pregnancy

Pneumonia Empyema Malignancy (may or may not contain malignant cells) Autoimmune serositis Pulmonary embolism Vasculitis Dressler syndrome (often bloody) Medications Uremia (sometimes bloody)

Asbestos exposure Intraabdominal abscess

BOX 39-1  Light’s Criteria for

TABLE 39-2  Role of Imaging Modalities in

Pleural fluid protein–to–serum protein ratio > 0.5 Pleural fluid LDH–to–serum LDH ratio > 0.6 Pleural fluid LDH > 2/3 times the upper limit of normal for serum

Radiography

Defining Exudate

various modalities in the diagnosis and treatment of pleural effusion are summarized in Table 39-2. Radiography and ultrasound.  Typical radiographic findings include unilateral or bilateral blunting of the costophrenic or cardiophrenic sulci with a meniscus sign, diaphragmatic silhouetting, fluid in the fissures, and mediastinal shift in cases of large effusion. Detection of pleural fluid on radiographs requires at least 175 mL on upright frontal radiographs, 75 mL on lateral views (in the posterior costophrenic sulcus), and as little as 10 mL on lateral decubitus views, which are most sensitive.123 Identification of hydropneumothorax or empyema (loculated effusion) is often possible with radiography (Fig. 39-8). Whether the effusion is unilateral (more likely exudative) or bilateral (more likely transudative) may be helpful in determining the character of an effusion in a given clinical context, though there is considerable overlap; biochemical analysis of pleural fluid is usually necessary for definitive assessment. After thoracentesis, radiography is routinely used to evaluate for complications such as pneumothorax. After evacuation of large pleural effusions, postthoracentesis pneumothorax ex vacuo, or “trapped lung,” may occur and should not be confused with iatrogenic pneumothorax caused by pulmonary puncture.11 The appearance of pneumothorax ex vacuo is caused by air in the pleural space, with lack of reexpansion of the chronically atelectatic or fibrotic lung (Fig. 39-9). Ultrasound is more accurate than radiography for estimating pleural fluid volume (as little as 5 mL may be detected) and more sensitive than CT in identifying septations.51,61 Importantly, ultrasound is useful in guiding interventions such as diagnostic and therapeutic thoracentesis, particularly in cases of small or loculated effusion, and pleural mass biopsy.34,77 Sonography has also been used to guide chest tube drainage and thoracoscopy in patients with parapneumonic effusion and empyema.15 CT.  CT is more sensitive than radiography for detection of small pleural effusions.34 The capability of CT to evaluate the entire pleural surface, lung parenchyma, mediastinum, and chest wall makes it the ideal modality for gauging the extent and nature of pleural fluid collections.51 Small simple effusions are seen as meniscoid low-attenuation collections that layer dependently, whereas loculated effusion is most often seen as lenticular fluid-attenuation masses along the dependent

Cases of Pleural Effusion

Ultrasound

CT

MRI

• First-line modality for detecting effusion • Often useful in detecting loculated effusion or hydropneumothorax • Whether effusion is unilateral or bilateral may suggest transudate or exudate and guide differential diagnosis, though biochemical analysis is usually necessary for definitive assessment. • Used to assess for complications after thoracentesis • Can detect smaller amounts of fluid than radiography • Whether effusion is anechoic or complex may suggest transudate or exudate and guide differential diagnosis, though biochemical analysis is usually necessary for definitive assessment. • More sensitive than CT for identifying septations • Often used to guide procedures • Portability an asset in assessment of critically ill patients • Comprehensive assessment of the entire pleura and adjacent structures in three dimensions • Higher contrast resolution than both radiography and ultrasound helpful for characterizing pleural fluid content by attenuation and enhancement characteristics • Useful for guiding intervention in selected cases not amenable to ultrasound • Limited role in evaluation of nonmalignant pleural effusions • May be useful in cases of suspected tumoral rib or chest wall invasion because of superior contrast resolution • DWI may help discriminate exudate (higher intensity on DWI, lower ADC) from transudate, but this may not be practical in routine clinical contexts.

posterior pleural surface.51 CT readily diagnoses loculated effusion in the major or minor fissure, which produces a pseudotumor appearance on plain radiographs. At times it may be difficult to differentiate pleural effusion from ascites using CT in a supine patient. Table 39-3 lists some signs that may be helpful in making this distinction. Attenuation values (Hounsfield units [HU]) are sometimes helpful in determining the content of a pleural effusion. Acute hemothorax is seen as a high-attenuation effusion. Whether attenuation value is 

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B

C

D FIG 39-8  A, Axial CT image showing a right-sided transudative effusion in a patient with chronic renal failure. B, Loculated small right pleural effusion with the “split pleura sign” in a patient with empyema. C, Fissural pleural thickening in a patient with right pleural mesothelioma. D, Large left hemothorax after resection of a left-sided bronchogenic cyst.

A

B FIG 39-9  Chest PA views before (A) and after (B) right-sided thoracentesis, showing lucency in the right lower hemithorax (pneumothorax ex vacuo, or “trapped lung”).

1101

1102

PART II  CT and MR Imaging of the Whole Body

TABLE 39-3  Signs Useful for

Distinguishing Pleural Effusion from Ascites on Supine CT “Displaced crus” sign32

“Interface” sign116

“Diaphragm” sign2 Sparing of the “bare area” of the liver90

Anterior displacement of the diaphragmatic crus by pleural fluid adjacent to the vertebral bodies, not seen in ascites Hazy rather than sharp interface between pleural fluid and the abdominal viscera created by diaphragmatic interposition; ascites has a sharp interface with the liver or spleen owing to lack of diaphragmatic interposition. Pleural fluid lies peripheral to the diaphragm rather than ascites, which lies medially. Ascites cannot interpose between the posterior right lobe of the liver and the abdominal wall because this portion is attached directly to the abdominal wall; fluid posterior to the right lobe of the liver must be in the pleural space.

reliable in differentiating exudate from transudate remains controversial. In their study of 100 patients, Abramowitz et al.1 concluded that attenuation values of pleural fluid do not reliably differentiate exudate from transudate. Additional features such as fluid loculation, pleural thickening, and pleural nodules were more commonly present with exudative effusion, but these findings were not statistically significant. Yildiz et al.125 asserted that this study’s conclusions may have been confounded by relatively low numbers of organizing consolidated effusions and inadequate statistical power. A study of 106 patients by Cullu et al.24 reached a differing conclusion, that exudates had a significantly higher median attenuation value (12.5 HU) than transudates (5 HU). Though there was significant overlap between exudates and transudates, it was concluded that a discriminatory value of 15 HU or greater is useful for identifying exudates. Aquino et al.3 determined that parietal pleural thickening is 96% specific for exudate in the setting of pleural effusion and that exudate without parietal pleural thickening is most common in malignancy and uncomplicated parapneumonic effusion. Outside the realm of exudative versus transudative effusion, CT is capable of identifying blood (hemothorax), enhancing septations or nodules (infection, malignancy), and gas (instrumentation, esophageal rupture, bronchopleural fistula [BPF], infection). The presence of small locules of gas in the pleural space is expected after pleural intervention but is specific for infection in other contexts.51 MRI.  MRI plays a limited role in the evaluation of pleural effusions, especially in benign disease, but may add value in cases where rib or chest wall invasion by the underlying process is suspected.51 Subacute or chronic hemothorax has high signal intensity on T1- and T2-weighted images. In chronic hemorrhagic effusion, there may be a concentric ring sign caused by peripheral hemosiderin (dark on T1- and T2-weighted images) and central methemoglobin (bright on T1- and T2-weighted images).79 Though chylothorax usually appears as a simple pleural effusion on all modalities, there have been reports of high signal intensity on T1-weighted MRI.79 MRI is sensitive for pleural septations and may be useful when cross-sectional imaging is needed but iodinated contrast is contraindicated. Simple fluid demonstrates low signal intensity on T1-weighted imaging and high intensity on T2-weighted imaging. Triple-echo pulse sequences have been used to differentiate complex exudate (containing

malignant cells or infection, highest signal) from simple exudate (no malignant cells or infection, high signal) and transudate (low signal).26 In their evaluation of 57 patients with pleural effusion using DWI, Baysal et al.8 found that the mean apparent diffusion coefficient (ADC) value of exudative effusions was lower than that of transudative effusions, with an optimum cutoff value of 3.38 × 10−3 mm2/sec, yielding 91% sensitivity and 89% specificity. In their study of 58 pleural effusions, Inan et al.56 reached similar conclusions. With current technology, however, routine use of MRI for this application is not feasible owing to factors including expense and technical limitations in scanning tachycardic, dyspneic, and critically ill patients.51 Pleurodesis.  Pleurodesis is therapeutic sealing of the pleural space by inciting an organizing inflammatory reaction that causes the visceral and parietal pleura to adhere to one another. The usual indications are recurrent pneumothorax and recurrent malignant effusion. Talc has been found to have higher success rates and a lower rate of systemic complications than the chemical agents bleomycin and tetracycline (doxycycline is the most commonly used agent in the tetracycline class).29,48 The agent is applied as an aerosolized powder at thoracoscopy or thoracotomy (poudrage) or at the bedside as slurry instilled via a chest tube.27,48,76 The major complications of talc pleurodesis are acute pneumonitis, acute respiratory distress syndrome (ARDS), and pulmonary edema.13,27,50 Following pleurodesis, the organizing phase is characterized on radiographs as pleural thickening and loculation.123 Over half of patients have visible pleural thickening on long-term follow-up imaging.80 Following talc pleurodesis, CT usually shows pleural thickening and nodularity, with some residual loculated effusion and dependently located high-attenuation foci along both the visceral and parietal pleura that may mimic pleural calcification.93 The highattenuation foci may also be located in the fissures.89 Areas correlating with these high-attenuation foci may have markedly increased FDG uptake on PET imaging that can simulate malignancy and persist for many years66 (Fig. 39-10). Aggressive antibiotic therapy is necessary in empyema. Surgical decortication may be helpful in patients who do not respond to antibiotics or in cases of moderate or large empyema and may include a Clagett window procedure, in which a small thoracotomy is left open for ongoing irrigation with antibiotic solution and drainage. When infection has resolved and granulation tissue has formed, antibiotic solution is instilled into the pleural space and the window may be closed using muscular flaps from the latissimus dorsi (Fig. 39-11).

Pleural Infection: Parapneumonic Effusions and Empyema.  Although approximately 40% of patients with pneumonia develop  an associated parapneumonic effusion, only about 10% develop an empyema, which is defined as a purulent pleural collection. Pneumonia is the most common cause of empyema. Other causes include lung abscess, BPF, esophageal perforation, complications of surgery, and trauma.78,127 The evolution of empyema in three stages is summarized in Table 39-4.64,72,127 If each stage is not treated adequately, there is progression to the next stage. Imaging of empyema.  As previously discussed, radiography and ultrasound are the modalities of choice for initial evaluation of pleural effusion. On plain radiographs, lenticular biconvex shadows rather than meniscoid shape suggests loculation and empyema rather than layering effusion. Loculated effusions do not layer on supine or decubitus views. Contrast-enhanced CT is a vital adjunct to pleural fluid analysis in managing patients with infected effusions or empyema and is highly accurate in differentiating empyema from lung parenchymal abscess abutting the pleura.64

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B FIG 39-10  A, Axial CT image in a patient with left-sided talc pleurodesis shows subtle high density within the left-sided pleural thickening. B, Axial fused PET-CT image showing diffuse increased FDG uptake, which is more likely to be related to post-pleurodesis inflammation than tumor. It is important to note that radiotracer uptake due to inflammation cannot be definitively differentiated from uptake by tumor.

A

C

B FIG 39-11  A, Axial CT image showing a bronchopleural fistula in the left upper thorax and a nonresolving left upper lobe pneumonia. B, Postsurgical débridement and a Clagett window and (C) post closure of Clagett window with a latissimus dorsi flap after the infection healed.

1103

1104

PART II  CT and MR Imaging of the Whole Body

A

B FIG 39-12  A, Axial CT image in a patient with right upper lobe necrotizing pneumonia with evidence of irregular cavitation within a right upper lobe consolidation. B, Another patient with an evolving right lower lobe pneumonia and a right hydropneumothorax with subtle pleural thickening, suggestive of an empyema.

TABLE 39-4  Three Stages of Evolution of

TABLE 39-5  CT Features Differentiating

1. First or “exudative” stage

Empyema

Parenchymal Abscess

Thin, uniform-thickness walls Smooth luminal and exterior margin Split pleura sign present Obtuse angle with chest wall Lenticular shape Compression rather than obliteration of lung structures Increased attenuation of extrapleural fat

Thick walls of nonuniform width Irregular luminal and exterior margin

a Parapneumonic Effusion into Empyema

2. Second or “fibropurulent stage”

3. Third or “pleural peel”

• Exudative sterile effusion due to lymphovascular leakage from the pulmonary interstitium into the pleural space • Appropriate treatment includes antibiotics and drainage. • Effusion becomes infected (complicated effusion) and loculations develop. Effusion containing neutrophils is known as empyema. • Appropriate treatment includes antibiotics and drainage. • Fibroblasts migrate from the pleura into the empyema and create a fibrous pleural peel that may restrict lung movement. • Appropriate treatment includes antibiotics and open thoracotomy or video-assisted thorascopic surgery (VATS) decortication.

Differentiation of abscess from empyema is important because attempted drainage of parenchymal abscess may seed the sterile pleural space with bacteria, whereas drainage is a mainstay of therapy for empyema. CT features differentiating empyema from parenchymal abscess are summarized in Table 39-56,79,111,122 (Fig. 39-12). A classic contrast-enhanced CT sign of empyema is the split pleura sign, which describes thickened, enhancing visceral and parietal pleural layers separated by low-attenuation effusion.64 Pleural thickening and enhancement characterizes 86% to 100% of cases of empyema as opposed to only 60% of parapneumonic effusion.3,6 Increased attenuation of the extrapleural fat (>50 HU greater than subcutaneous fat) is an associated finding that may persist for a period of time following resolution of infection.6,59,122

Empyema from Parenchymal Abscess

Split pleura sign absent Acute angle with chest wall Round shape Abrupt vessel cutoff and presence of bronchi at interface with normal lung Normal attenuation of extrapleural fat

Ultrasound is more sensitive than CT in detecting septations and is often used to guide intervention; CT is capable of detecting separate loculations of pleural fluid (inferred by trapped gas pockets within the collection rather than by visualization of the septa themselves) in approximately 20% of cases.25 CT is useful to guide thoracentesis or drainage catheter placement when ultrasound is not feasible because of large patient body habitus or poor visualization of critical structures owing to air in the lung parenchyma or adjacent bony structures.127 Radiography and CT play complementary roles in assessing appropriate catheter positioning after placement and in cases where catheter drainage has ceased or the patient is not responding to therapy. MRI is rarely needed for routine evaluation of empyema. Relative weaknesses as compared with CT are inferior spatial resolution and susceptibility to artifacts; however, MRI is sensitive for demonstrating septations within pleural collections and may be a useful alternative  in patients with contrast allergies or in contexts where avoidance of ionizing radiation exposure is desired. Rarely malignancies such as non-Hodgkin’s lymphoma, squamous cell carcinoma, sarcoma, and mesothelioma arise in chronic empyema cavities. MRI, because of to

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

1105

B FIG 39-13  A, Axial CT image showing bilateral pleural plaques—both calcified and noncalcified—in a patient with asbestos exposure. B, Diffuse pleural thickening with dystrophic calcification in a patient with left-sided pleural metastasis from osteosarcoma.

its excellent contrast resolution, is the modality of choice for differentiating neoplasm from fibrotic tissue.86

Bronchopleural Fistula.  BPF is defined as a direct communication between the bronchial tree and the pleural space that may occur spontaneously, as in cases of necrotizing pleural infection or malignancy, or as a result of trauma or iatrogenic injury. BPF is associated with high morbidity and mortality. Central BPF is a connection between a central large bronchus and the pleural space, whereas peripheral BPF is a connection between a peripheral airway and the pleural space. The most common causes of central BPF are dehiscence of a bronchial stump after lung transplantation or pneumonectomy, and trauma. Peripheral BPF is caused by necrotizing pneumonia, empyema, trauma, ruptured bullae, and iatrogenic injury from radiation therapy or thoracentesis. On radiography, increasing volume of pleural air (without a history of instrumentation), mediastinal shift, and loss of fluid in a previously filled or filling postpneumonectomy space are indirect signs of BPF. New gas within an empyema may also suggest the diagnosis. CT with IV contrast is the modality of choice in suspected BPF, given its ability to depict the vasculature and evaluate for the underlying cause.  Indirect findings on CT are analogous to the previously described radiographic findings. Direct findings of BPF include the fistulous connection itself or extraluminal air bubbles adjacent to the bronchial stump. Multiplanar reformation and advanced postprocessing techniques (e.g., volume rendering, virtual bronchoscopy, MIP, MiniP imaging) are valuable in the evaluation of BPF.40 Demonstration of inhaled radiotracer (e.g., xenon) in the pleural space on ventilation scintigraphy may be useful in cases of occult BPF.40 Treatment of central BPF includes broad-spectrum antibiotics, protection of the contralateral lung from spilled pleural fluid by appropriate patient positioning, drainage or surgical washout of the pleural space, and primary repair of the bronchial stump in postpneumonectomy patients (sometimes with reinforcement using omentum or other vascularized soft tissue). The goal of interventional therapy for peripheral BPF is pleural apposition, which is accomplished by thoracostomy, decortication, pleurodesis, and open drainage, or transbronchial occlusion using muscle flaps, fibrin sealants, coils, stents, and one-way valves.40

Pleural Calcifications.  Pleural calcifications may be seen in numerous conditions, including asbestos-related pleural disease, resolved

hemothorax, and pyogenic or mycobacterial infection. Prior talc pleurodesis should also be considered in the differential diagnosis for high-density material in the pleural space (Fig. 39-13).

Diffuse Pleural Thickening.  Diffuse fibrous pleural thickening, likely a reaction to chronic inflammation of the visceral pleura that results in fusion of the parietal and visceral pleura, is seen in a number of disease states. It is especially common in diseases that cause exudative pleural effusion.104 Tuberculous or bacterial empyema, hemothorax (especially posttraumatic), talc pleurodesis, asbestos exposure, radiation therapy, and some medications are the most common causes. Pleural thickening related to asbestos exposure and malignancy is discussed subsequently. On radiographs, pleural thickening may be indistinguishable from effusion. Typical findings include blunting of the costophrenic sulci, a several millimeter-thick soft tissue density stripe separating the lungs from the adjacent chest wall, thickened soft tissue internal to the ribs in patients with pneumothorax, pleural calcification, and asymmetric increase in low-attenuation extrapleural fat.123 When radiographs raise suspicion for diffuse pleural thickening, CT is useful for distinguishing thickened pleural from prominent extrapleural fat and malignant disease. On CT, pleural thickening may be seen as a soft tissue density stripe at least 1 mm in thickness internal to the ribs, innermost intercostal muscles, and paravertebral regions.123 The appearance of the thickened pleura itself is nonspecific, and evaluation of patient exposure history (e.g., tuberculosis, asbestos) and assessment of the chest wall and lungs by CT is often necessary to narrow the differential diagnosis. Pleural thickening should be evaluated at several levels and outside the paraspinal region, where intercostal vessels and normal thickened fat may be difficult to distinguish from the pleura.91 Mimics of pleural thickening include the normal subcostalis and transverse thoracis muscles, and intercostal veins.123 In patients without a history of asbestos exposure, ancillary findings in the chest wall and lungs may permit accurate diagnosis. For example, presence of rib fractures and normal lung parenchyma may suggest pleural thickening due to prior hemothorax, whereas post­ infectious (including posttuberculous) pleural thickening may have volume loss, extrapleural fat thickening, or parenchymal scarring. After talc pleurodesis, high-attenuation talc between the thickened visceral and parietal pleura (usually in the posterior thorax) creates the talc sandwich sign.89

1106

PART II  CT and MR Imaging of the Whole Body

Erdheim-Chester disease (ECD) is a rare idiopathic form of non– Langerhans cell histiocytosis that may cause diffuse pleural thickening. The most typical presentation is bone pain with symmetric osteosclerosis of the metadiaphyses of the long bones.14 Thoracic involvement is seen in about 15% to 30% of cases.30,121 In their study of 40 patients with thoracic manifestations of ECD, Brun et al.14 reported perivascular and mediastinal infiltration, pericardial effusion and thickening, ground-glass opacities and poorly defined pulmonary nodules, pulmonary cysts, pleural effusions, and pleural thickening.

Benign Asbestos-Related Pleural Disease.  Asbestos is a naturally occurring substance with fire-resistant properties that has been used in many industrial applications including building insulation, manufacturing, and shipbuilding. Inhalation of asbestos fibers has been associated with benign and malignant pleuropulmonary diseases such as formation of fibrous pleural effusions, pleural plaques, diffuse pleural thickening, round atelectasis, lung cancer, and malignant mesothelioma.104 Many pathologic entities related to asbestos exposure tend to manifest in a delayed fashion, with latency of up to  60 years.

Asbestos-Related Pleural Effusion.  Benign hemorrhagic exudative pleural effusions are the earliest manifestation of asbestos-related pleural disease, appearing within 10 years of exposure. Risk of pleural effusion appears to be dose dependent. Asbestos-related effusions may be asymptomatic. They tend to resolve over months but may persist or recur. Development of diffuse pleural thickening often follows.104

Pleural Plaques.  Pleural plaques are discrete areas of fibrosis that arise more commonly from the parietal than visceral pleura, appearing 20 to 30 years after exposure. They are the most common manifestation of asbestos-related pleural disease and are usually asymptomatic.104 Asbestos fibers are long and thin; therefore they are not cleared by macrophages or lymphatics. They tend to remain in the lower lung zones where ventilation is higher than in the upper lung zones. Thus, distribution of asbestos-related lung disease is predominantly basal. The anterior and lateral chest wall have greater excursion with respiration than the posterior chest wall, which is relatively fixed to the spine.38 Together, these attributes have led some to speculate that the reason asbestos-related pleural plaques form preferentially in the basal anterior and posterolateral chest is because of greater chest wall motion and hence more friction, leading to more irritation of the pleura by asbestos fibers. On radiographs, pleural plaques appear as opacities that demonstrate the incomplete border sign. They are most commonly identified at the posterolateral chest wall and against the mediastinum and diaphragm, where thick flat opacities are virtually pathognomonic for asbestos-related pleural plaques. The apices and costophrenic sulci tend to be spared, unlike in diffuse pleural thickening. Calcification, present in 10% to 15% of cases, increases conspicuity. The holly leaf sign describes the appearance of pleural plaques as irregular in shape with angular outer contours.23,104 CT is more sensitive than radiography and reveals additional anterior and paravertebral plaques that may not be detected on radiographs. CT is also helpful in distinguishing pleural disease from extrapleural fat. Plaques involving the visceral pleura are sometimes called hairy plaques because of their association with short interstitial linear opacities in the adjacent lung parenchyma that appear to radiate from the plaque104 (Fig. 39-14).

Diffuse Pleural Thickening from Asbestos.  In patients with a history of asbestos exposure, diffuse pleural thickening is less common

than focal pleural plaques, occurring in less than 10%. Incidence is, on average, 20 to 40 years after initial exposure. A dose-dependent relationship has been described.57,84 Presenting symptoms include progressive dyspnea and decreased exercise tolerance. Extensive pleural thickening is associated with restrictive physiology, which is reflected by pulmonary function testing in affected patients. Pulmonary function testing abnormality has been shown to correlate with the degree of pleural thickening.21 For the diagnosis of diffuse pleural thickening, CT is more sensitive and specific than radiography.104 The CT appearance of diffuse pleural thickening in patients with a history of asbestos exposure has been described by McLoud et al.82 as smooth, uninterrupted pleural opacity extending over at least one quarter of the chest wall, with or without obliteration of the costophrenic angle. Calcification is more rare in diffuse pleural thickening than in focal pleural plaques, and diffuse pleural thickening more often involves the apices and costophrenic angles. Unlike diffuse pleural thickening, nonconfluent plaques do not extend over more than four interspaces, and diffuse pleural thickening may involve the interlobar fissures, whereas plaques typically do not.104 Reactive proliferation of extrapleural fat is a common associated finding. Ninety percent of patients with diffuse pleural thickening due to asbestos exposure have associated parenchymal abnormalities. Diffuse pleural thickening due to asbestos exposure is more often bilateral than pleural malignancy and lacks infiltration and nodularity.

Round Atelectasis.  Round atelectasis associated with asbestosrelated disease, also known as asbestos pseudotumor, is thought to occur as a result of pleural inflammation and fibrosis that results  in infolding into the lung parenchyma, which becomes trapped and atelectatic.104 The classic radiographic appearance on chest radiographs is a masslike opacity with a broad-based interface with thickened pleura, associated with bronchovascular comet tails. The presence of volume loss rather than mass effect may favor this diagnosis over true masses such as lung cancer, the primary and most feared differential diagnosis. Contrast-enhanced CT is helpful for further characterization and shows uniform enhancement of the atelectatic lung (though this is not a reliable feature for distinguishing atelectasis from lung cancer). Given the sometimes nonspecific appearance of round atelectasis and the primary differential consideration of lung cancer, biopsy or demonstration of stability or shrinkage on serial interval follow-up may be necessary to confirm the diagnosis.104 On T1-weighted MRI, atelectasis may appear as a mass with signal characteristics similar to those of liver tissue. MRI may also demonstrate the relationship of the pseudomass with bronchovascular structures. Curvilinear hypointense lines are thought to represent thickened indentations of visceral pleura.104,120

Pleural Masses and Malignant Pleural Disease Distinguishing malignant from benign disease

CT.  Contrast-enhanced CT has been shown to be a moderately sensitive and highly specific modality for differentiating benign from malignant pleural disease. It is useful for searching for primary malignancy in cases of metastasis of unknown primary and assessing for locoregional or extrapleural metastatic spread. CT features that favor malignant over benign pleural disease with their respective test characteristics are summarized in Table 39-6.69,83 In their study of 215 patients, Metintas et al.83 noted that invasion of the pericardium, chest wall, diaphragm, or mediastinum with pleural disease and nodular involvement of fissures was not seen in any cases

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B

C

D

1107

FIG 39-14  PA (A) and lateral (B) radiographs showing bilateral pleural plaques. C and D, Axial CT images showing consolidative opacities in the right lower lobe and left upper lobe that are characterized by round, masslike shape, associated bronchovascular “comet tails”, and broad-based interfaces with thickened adjacent pleura.

TABLE 39-6  CT Morphologic

Characteristics of Malignant Pleural Disease CT Feature Pleural thickening >1 cm Involvement of mediastinal pleura Pleural nodularity “Rindlike” or circumferential pleural thickening

Sensitivity

Specificity

36%-47% 56%-70% 38%-51% 41%-54%

64%-94% 83%-88% 94%-96% 95%-100%

of benign disease. The presence of calcification, for which CT is more sensitive than MRI, has been shown to be highly specific for benign disease.54,69 PET-CT.  FDG PET-CT has been shown to be highly accurate in differentiating between malignant and benign pleural disease;

however, false-positive results may be seen in cases of infection, uremic pleuritis, and pleurodesis, and false-negative results may occur in cases of solitary fibrous tumor of the pleura and low-grade lymphoma.31,44,63 MRI.  Given the excellent discriminating ability of CT, MRI is considered a problem-solving modality for distinguishing malignant from benign pleural disease. In their study of 42 patients, Hierholzer et al.54 showed MRI combining signal and morphologic characteristics to be highly sensitive (100%) and specific (93%) for malignant pleural disease. MRI is useful in surgical planning, given its excellent ability to detect diaphragmatic and chest wall invasion, which are reliable indicators of malignancy even in the absence of the classic morphologic characteristics described above.44,54 Morphologic characteristics that predict malignancy are similar between CT and MRI.54 Hypointensity of pleural disease relative to the intercostal muscles on T2-weighted sequences has been shown  to reliably predict benignity, whereas malignant disease generally

1108

PART II  CT and MR Imaging of the Whole Body

demonstrates T2 prolongation.35,36 Contrast-enhanced T1-weighted images are best for detecting chest wall, abdominal, and diaphragmatic invasion.44,54 In their study comparing PET-CT and DWI-MRI for differentiating malignant from benign pleural disease, in 31 patients, Coolen et al.20 reported an optimal ADC threshold of 1.52 × 10−3 mm2/ sec. DCE MRI raised both sensitivity and specificity.

Benign pleural masses

Solitary fibrous tumor of the pleura.  Solitary (or localized) fibrous tumor of the pleura (SFTP) is a rare neoplasm of mesenchymal origin that may be benign (≈80%) or malignant (≈20%).105,107 Histologically identical tumors are also found in extrathoracic sites. Small SFTPs are usually discovered incidentally in asymptomatic patients, but larger lesions may present with chest pain, cough, dyspnea, or a pressure sensation. Hypertrophic pulmonary osteoarthropathy is the most common paraneoplastic syndrome, seen in up to 22% of affected patients.105 Rarely these tumors express insulinlike growth factor (IGF)-2 and thus are associated with hypoglycemia.44 The imaging appearance of SFTP depends on size. Radiographically, these lesions are usually smooth or lobulated. When they abut the diaphragm, their appearance may simulate diaphragmatic eventration. In 40% of cases, lesions are pedunculated and hence mobile. Small effusions may be present.106 The typical CT appearance is of a homogeneous, well-circumscribed, noninvasive mass of soft tissue attenuation. Large lesions may have areas of myxoid change, hemorrhage, necrosis, or cystic degeneration, and many of these contain calcifications.105 CT may show the lesion to be within a fissure. Small lesions typically form obtuse angles with the pleural surface; large lesions may form acute angles and mimic a subpleural pulmonary mass. Regardless of whether the angle is acute or obtuse, the lesion usually has a smooth, tapering margin with the pleura. If an SFTP originates from the mediastinal pleura, it may mimic a mediastinal mass.75 Malignant lesions are more often sessile and atypically located, with necrosis and hemorrhage.105 On MRI, SFTP generally follows a pattern consistent with fibrous tissue—lesions tend to be low to intermediate signal on T1-weighted, T2-weighted, and proton density images. Areas of necrosis or myxoid change may be denoted by high signal on T2-weighted images. Intense homogeneous enhancement is typical, and the lesion may have a rim of low signal on T2-weighted images.18,44 On PET imaging, FDG uptake tends to be minimal.105 Core needle biopsy often does not yield sufficient tissue for definitive diagnosis107; thus excision is the treatment of choice for both malignant and benign SFTP105 (Figs. 39-15 and 39-16). Lipoma.  Lipomas are rare in the pleura and show characteristics of fat on all imaging modalities. Pleural lipomas are minimally complex, homogeneous, circumscribed, fat-attenuation lesions on CT. On MRI they are hyperintense on T1- and moderately intense on T2-weighted images, with suppression on fat-saturated sequences  and chemical shift artifact at their interfaces with fluid-rich tissues (Fig. 39-17).44 Pleural metastases and malignant effusions.  Metastatic disease is more common than primary pleural malignancy. Approximately 25% of all pleural effusions in elderly patients are malignant, and effusion is the most common finding in patients with pleural metastasis.12 The most common primary tumors that produce pleural metastases are lung, breast, ovarian, and gastric carcinomas, and lymphoma. Renal cell carcinoma and melanoma also have a propensity to metastasize to the pleura.12 Spread to the pleura may be direct (e.g., in cases of peripheral lung cancer), lymphatic, or hematogenous. Rarely thymic carcinoma and lymphoma may spread directly to the pleural space from the anterior mediastinum.44 Metastatic disease may present as solitary implants on the costal, diaphragmatic, or mediastinal pleural surface or in the interlobar fis-

sures. Contrarily it may be a circumferential rind that may be indistinguishable from mesothelioma12 (Figs. 39-18 and 39-19). Malignant pleural mesothelioma.  Malignant pleural mesothelioma (MPM) is an aggressive tumor arising from the parietal pleural surface. Though MPM was once a rare malignancy, worldwide incidence is increasing, most likely due to exposure to carcinogenic asbestos fibers, a known risk factor. Incidence is expected to peak from 2015 to 2020. The most common clinical presentation is an elderly male with pleural effusion and pleuritic chest pain who has a history of exposure to asbestos.106 Focal mass and advanced disease indistinguishable from advanced lung cancer are more rare presentations. Disease is difficult to manage and prognosis is poor, with median survival of 9 to 17 months after diagnosis.117 There are three histologic subtypes of MPM with different biological behaviors and natural histories. Epithelioid mesothelioma is the least aggressive and has the best prognosis; sarcomatoid and mixed (“biphasic”) tumors are often metastatic at the time of presentation.45 Histologically, sarcomatoid MPM possesses higher cellular and microvascular density and larger nucleus-to-cytoplasmic ratios than epithelioid subtypes.47 The risk of tumor seeding with nonexcisional biopsy has prompted development of imaging tests that can predict tumor characteristics noninvasively20 (Fig. 39-20).

Imaging in the diagnosis and staging of malignant pleural mesothelioma Radiography.  The radiographic appearance of MPM is nonspecific. Findings include pleural effusion, lobulated pleural thickening, or frank pleural masses. Coexistent pleural plaques may limit radiographic evaluation.45 Ultrasound.  Ultrasound may be useful in the diagnosis of MPM by guiding and thus increasing the diagnostic yield of biopsy, but it is not useful for the more global evaluation afforded by cross-sectional imaging.45 CT.  For diagnosis, staging, and monitoring of response to therapy in MPM, CT is the modality of choice. CT findings include unilateral pleural effusion (74%), circumferential nodular pleural thickening (92%), and thickening of the interlobular septa (86%).34 Tumor tends to develop at the diaphragmatic surface and grow upward.45 In some cases MPM presents with localized pleural masses rather than circumferential pleural thickening.106 There is often volume loss in the affected hemithorax, which has been termed the shrinking lung sign.4 In advanced disease, tumor may invade adjacent structures or metastasize to the lymph nodes, bones, lungs, or distant sites. Volume-rendered 3D CT images may show the relationship of tumor to critical structures and are often used for patient education and surgical planning. Other advanced CT techniques are also coming into widespread use, including volumetric analysis with serial segmentation and perfusion CT.45,100 DCE “perfusion” CT can be used to measure blood flow dynamics and capillary permeability, thereby evaluating microvascularity of tumors before and after treatment.45 CT does not assess nodal disease accurately, because nodal size does not correlate with tumor involvement.5 The most commonly used tumor, node, metastasis (TNM) staging system for CT and MRI was proposed by the International Mesothelioma Interest Group (IMIG) in 1995. Determination of resectability varies among institutions and is often considered on a case-by-case basis. Features that suggest unresectability include extensive chest wall invasion, mediastinal invasion, mediastinal lymph nodal involvement, and intraabdominal or distant metastatic disease.45 Other staging paradigms, including the Brigham and Women’s Hospital staging system, are also in use.113 Definitive staging of T1- and T2-stage disease by the IMIG TNM system using current imaging technology is not possible given that  the visceral and parietal pleurae cannot be distinguished, and all

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B

C

D

E FIG 39-15  A, PA view of the chest in a patient with fibrous tumors of the pleura showing large, lobulated left chest masses. B, Axial CT image showing a mass abutting the left posterior pleura and a partially imaged fissural mass. C and D, Axial T1- and T2-weighted images showing that the masses are isointense to muscle and (E) showing heterogeneous enhancement of the fissural mass and peripheral enhancement and central necrosis of the posterior pleural mass.

1109

1110

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 39-16  MRI images in a patient with malignant fibrous tumor of the right pleura. A, Axial T1-weighted image shows right fissural tumor isointense to muscle. B and D, The mass enhances. C, Coronal T2-weighted image shows the mass to be hyperintense to muscle.

modalities are relatively insensitive for nodal disease.12,101 Cervical mediastinoscopy with biopsy has been shown to be more sensitive and specific for nodal disease than CT and has been recommended when radical surgery is being considered.108,119 Currently, the International Association for the Study of Lung Cancer (IASLC) is developing a new staging system that accounts for factors such as the differing biological behavior of epithelioid and nonepithelioid cancers118 (Figs. 39-21 and 39-22). MRI.  As previously discussed, MRI has advantages over CT in differentiating malignant from benign pleural disease. In addition, MRI is useful for confirmation and further evaluation of CT findings.44 Given its higher signal-to-noise ratio, MRI is superior in detecting chest wall and diaphragmatic involvement, particularly with contrastenhanced T1-weighted sequences with fat suppression.4,45 MRI may reclassify about 30% of surgical candidates into an inoperable stage.112 MPM is iso- to slightly hyperintense relative to adjacent chest wall musculature on T1-weighted images and moderately hyperintense on T2-weighted images. IV gadolinium increases conspicuity of tumor.44

Sagittal and coronal images are most useful for delineating tumor breach of the diaphragm and mediastinal and chest wall involvement— factors that decrease resectability.45 DCE (perfusion) MRI may allow noninvasive distinction among histopathologic subtypes and evaluate microvascularity in a manner analogous to CT perfusion.4,45 Perfusion MRI has been shown to be capable of predicting therapeutic efficacy of chemotherapy in unresectable disease.45 Cine imaging techniques with high temporal resolution evaluate tumor mobility and lung motion (e.g., parallel MRI using fast imaging with SSFP), permitting identification of subtle mediastinal and chest wall invasion.44 DWI with ADC correlation has been shown to reflect the histologic differences between epithelioid and nonepithelioid subtypes, with higher intensity on DWI and lower ADC values found in sarcomatoid tissue, denoting edema, higher cellularity, and higher nuclear-to-cytoplasmic ratios.47 The pointillism sign, named for the postimpressionist painting style, describes the appearance of discrete high-intensity foci on high b-value DWI that likely represent multifocal tumor deposition. This sign has been shown to be highly sensitive (92%) and specific (86%)

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B FIG 39-17  A, PA view of the chest showing a pleural lesion in the right upper chest incidentally seen in a patient with right lower lobe pneumonia. B, Axial CT image showing a fat-containing right upper pleural lesion, consistent with lipoma.

A

B FIG 39-18  A, Coronal postcontrast VIBE image showing a solitary right pleural metastasis from melanoma. B, Coronal T2-weighted MRI showing diffuse right pleural metastases from lung adenocarcinoma, with nodular pleural thickening and a large loculated right pleural effusion.

1111

A

B FIG 39-19  A, Diffuse metastases from osteosarcoma, with dystrophic calcification. B, Diffuse enhancing left-sided pleural tumor (metastatic thymoma).

A

B

C FIG 39-20  A, Coronal postcontrast VIBE image showing left circumferential biphasic mesothelioma with apical involvement (arrow). Enhancement in the left lateral chest wall is due to recent surgical biopsy. B, Sagittal image showing the relationship of tumor to the diaphragm and fissures; the patient underwent successful resection of mesothelioma. C, Volume-rendered image showing the resected gross pathological specimen.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A B

C D

E

F FIG 39-21  A-F, Axial and coronal CT images in a patient with right-sided epithelioid mesothelioma with extensive nodal involvement. Note enlarged right and left paratracheal lymph nodes, subcarinal node, right internal mammary node, and pericardial and diaphragmatic nodes. Malignant and benign adenopathy from mesothelioma cannot be reliably differentiated by imaging, and sampling by cervical mediastinoscopy may be needed for accurate preoperative nodal staging.

1113

1114

PART II  CT and MR Imaging of the Whole Body

A

B FIG 39-22  A, Axial CT image showing focal invasion of the endothoracic fascia. B, Coronal image showing invasion of mediastinal fat medially and endothoracic fascia laterally.

for MPM and may be useful for guiding thorascopic evaluation4 (Figs. 39-23 through 39-27). PET-CT.  Hybrid PET-CT has been used in the management of MPM to determine metabolic behavior and extent of tumor. The chief strength of PET-CT is its ability to detect occult metastatic disease that would not be apparent by other modalities but changes staging.4 FDG uptake may be moderate to high, depending on histologic subtype, with epithelioid tumors being less FDG avid than their sarcomatoid or biphasic counterparts.42,45 Primary tumors tend to be more FDG avid in higher-stage disease.41 Higher metabolic activity has been shown to portend poorer prognosis with shorter survival.4 PET may allow identification of nodal disease not evident by CT and allow distinction of tumor from fibrosis, necrosis, and atelectasis45,106 (Fig. 39-28).

Treatment, assessment of treatment response, and postsurgical evaluation in MPM Treatment.  Treatment decisions are tailored to each patient on the basis of tumor staging and operative risk. Surgical considerations in the management of MPM remain controversial, with a major distinction between palliative surgery and surgery with curative intent (maximal debulking surgery).118 Maximal debulking surgery can be subdivided into: (1) extrapleural pneumonectomy (EPP), in which the entire pleura, ipsilateral lung, and (if necessary) the pericardium and diaphragm are removed, with subsequent soft tissue patch reconstruction; and (2) pleurectomy/decortication (P/D), in which the ipsilateral lung is left in place.118 A recent study of 101 patients demonstrated that neoadjuvant chemotherapy, P/D, and adjuvant chemotherapy were superior to neoadjuvant chemotherapy, EPP, and adjuvant radiotherapy in terms of overall survival.67 Modern multimodality therapy may include combinations of surgery, systemic or heated intraoperative pleural chemotherapy, photodynamic therapy (in which the tumor is sensitized to visible light by a systemically distributed agent and subsequently exposed to light), intrapleural immunotherapy, and radiotherapy.39,114,118 Many clinical trials are ongoing. Assessment of treatment response.  Volumetric analysis may be more useful in MPM than the Response Evaluation Criteria in Solid

Tumors (RECIST) and modified RECIST criteria in evaluating treatment response in unresectable disease, given that MPM is nonspherical and grows along pleural reflections, making it difficult to measure in orthogonal planes.45 Assessment for recurrence or progressive metastasis is generally with CT or PET-CT. CT findings of recurrence include enlarging soft tissue mass along the resection margin, ascites or peritoneal thickening indicative of abdominal disease, pulmonary nodules, and mediastinal lymphadenopathy.45 Postoperative granulation tissue along the resection margin is usually more irregular and nodular than recurrent tumor but is sometimes concerning. PET-CT is useful for distinguishing granulation tissue from recurrent tumor. Whereas FDG uptake in granulation tissue tends to decrease over time, tumor tends to have unchanging or increasing FDG uptake over serial studies.9 PET-CT is also used for semiquantitative comparison of pre- and posttreatment FDG uptake in tumor masses.45 Decrease in tumor microcirculation as measured by perfusion MRI has been shown to be linked to early clinical response and survival in MPM patients on chemotherapy43 (Fig. 39-29). Postsurgical evaluation.  Following surgery, radiography is used routinely; CT is used when complications are suspected.45 After pneumonectomy, the pneumonectomy space fills with fluid at a predictable rate, with opacification of approximately one additional intercostal space every 7 days by chest radiography. If the rate of filling is excessive, hemorrhage or chyle leak due to damage to the thoracic duct should be suspected. MRI can be used to evaluate the thoracic duct prior to attempted corrective embolization in cases of chyle leak. Contrarily, a decrease in accumulated fluid suggests development of a BPF or leakage into the abdomen via a defect in the reconstructed diaphragm.45 BPF tends to present within 1 week to 3 months after surgery.40 Management of BPF in the contexts of pneumonectomy and empyema is similar (Fig. 39-30). Herniation of the stomach into the thorax is a rare complication of left-sided pneumonectomy that may be complicated by strangulation.45 In the rare postpneumonectomy syndrome, extreme rotation Text continued on p. 1119

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B

C

D

E FIG 39-23  A-C, The diaphragm is clearly demonstrated as a black line on HASTE and an enhanced line on FS-T1-weighted images (arrows). A preserved fat plane is also clearly shown as a suppressed line. D, Epithelioid MPM with extension across the diaphragm. The line of the diaphragm is interrupted (white arrows). The subdiaphragmatic fat plane is also obscured by the mass (black arrows), indicating intraabdominal extension. E, Sagittal contrast-enhanced FS-T1-weighted image demonstrates hepatic involvement (small black arrow).

1115

1116

PART II  CT and MR Imaging of the Whole Body

FIG 39-24  Coronal subtraction image showing an irregular lateral contour of a right-sided mesothelioma, consistent with multifocal endothoracic fascial involvement.

A

B

C

FIG 39-25  Comparison of CT and MRI images in a patient with left-sided sarcomatoid mesothelioma with left chest wall extension, left rib metastases, and intramural thrombus in the descending aorta. MRI is superior for detecting chest wall invasion, involvement of vascular structures, and bone involvement.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

B A

D

C

E FIG 39-26  Patient with left sided Mesothelioma. A, T1-weighted axial image showing anterior chest wall invasion by the tumor, B, axial B1000 image, and C, an ADC map, with an ADC value of 0.883 × 10−3 mm2/ sec suggestive of sarcomatoid subtype. D, T2-weighted axial image and E, axial B500 image depicting the “pointillism sign”.

1117

1118

PART II  CT and MR Imaging of the Whole Body

AUC

A

B KTRANS

C

Ve

D FIG 39-27  Perfusion-weighted imaging in mesothelioma showing multiparametric maps. A, AUC (area under the curve); B, enhancement curve; C, Ktrans (transport coefficient); and D, Ve (elimination coefficient).

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B FIG 39-28  A and B, Fused 18F FDG PET-CT images in a patient with advanced epithelioid MPM, with avid left supraclavicular nodes, mediastinal nodes, and intraabdominal extension (arrows).

and shift of the mediastinum results in symptomatic central airway obstruction109 (Fig. 39-31). Primary pleural sarcoma.  Primary sarcomas that involve the pleura are rare and include liposarcoma, synovial sarcoma, and undifferentiated pleomorphic sarcoma, previously known as malignant fibrous histiocytoma. For all types, treatment includes resection, often with chemotherapy and/or radiation. Liposarcoma.  Pleural liposarcomas are often asymptomatic and appear on CT as large infiltrative masses that usually contain fat, with prominent soft tissue components such as enhancing nodules or thick septa. On MRI, liposarcomas appear hyperintense on T2-weighted sequences owing to myxoid degeneration, which is common. On T1-weighted images, they tend to be hypointense, with variable enhancement after administration of contrast12,44 (Fig. 39-32). Synovial sarcoma.  Thoracic synovial sarcoma is more rare than its extrathoracic counterpart and may arise in the chest wall, pleura, mediastinum, heart, or lung. Ipsilateral pleural effusions are common.

1119

On CT, pleural synovial sarcomas are noncalcified well-circumscribed masses that do not involve bone. The enhancement pattern is typically heterogeneous, and the tumor often contains areas of hemorrhage and necrosis. On MRI, nodular soft tissue with multiloculated fluid-filled components is typical. Rim enhancement is seen following administration of IV contrast. Tissue heterogeneity is more evident by MRI than CT. In reference to synovial sarcoma the triple sign describes the presence of hyper-, iso-, and hypointense areas on T2-weighted MRI87 (Fig. 39-33). Undifferentiated pleomorphic sarcoma.  Undifferentiated pleomorphic sarcoma (UPS) is a malignancy of fibrohistiocytic origin that rarely involves the pleura. On CT, UPS is a somewhat nonspecificappearing heterogeneously-enhancing mass. Calcification is unusual prior to treatment. The MRI appearance of UPS is a mass with intermediate to low signal on T1-weighted sequences and intermediate to high signal on T2-weighted sequences. Variably present fibrous tissue components include collagen (hypointense on all pulse sequences), myxoid stroma (hypointense on T1-weighted images and hyperintense on T2-weighted images), and hemorrhage92 (Fig. 39-34). Epithelioid hemangioendothelioma.  Epithelioid hemangioendothelioma (EH) is a rare low- to intermediate-grade aggressive neoplasm that rarely arises in the pleura. It is more common in the bone, liver, soft tissue, and lung. The pleural form more commonly affects males than females. Presenting symptoms are somewhat nonspecific and include chest pain, dyspnea, fever, and cough. Patients tend to present with advanced disease. Radiographic findings include pleural effusion and volume loss in the affected hemithorax. CT shows smooth and nodular pleural thickening, with or without invasion of the diaphragm or mediastinum. Pulmonary nodules or interstitial thickening indicate metastasis. As in other entities, MRI is superior for demonstrating invasion of the diaphragm and mediastinum, and PET-CT is useful for evaluating for occult metastasis. Overall the imaging findings are similar to those of MPM and pleural carcinomatosis. Given the rarity of the condition, no treatment regimen has been rigorously tested; to date, patients with EH have been treated with surgery and chemoradiation. The prognosis of pleural EH is poor (much worse than that of pulmonary EH), with a median survival time of 10 months after diagnosis.7,22,62 Pancoast tumors.  Apical lung cancers are also known as Pancoast or superior sulcus tumors. The term Pancoast syndrome refers to shoulder and arm pain with or without ipsilateral Horner’s syndrome due to involvement of the lower cords of the brachial plexus and sympathetic chain by tumor.16 Chest wall invasion with rib destruction is common. Other critical structures that may be involved are the spine, subclavian vessels, and inferior neck.37,50 Pancoast tumors may be difficult to identify on radiographs because of obscuration by the overlying ribs and clavicles and resemblance to the virtually ubiquitous benign apical pleural cap. Bone destruction, present in approximately one third of cases, may be difficult to identify on plain films.96 The finding of an asymmetric or enlarging pleural cap should be regarded with suspicion, especially in a symptomatic patient.50 CT is helpful in differentiating the intrapulmonary mass from an apical pleural cap and identifying bone destruction and chest wall invasion.17,50 CT is also essential for staging; mediastinal adenopathy portends a poor prognosis.37 CT-guided fine needle biopsy is successful in establishing the diagnosis in more than 90% of cases.37 MRI is superior for evaluating the extent of tumor and involvement of the brachial plexus, subclavian vessels, neural foramina, spinal canal, and bone marrow. Coronal and sagittal T1-weighted and STIR sequences obtained with thin sections and surface coils are optimal. MR angiography is particularly useful for vascular assessment.37 Interruption of the normal extrapleural fat line over the lung apex indicates tumoral invasion.50

1120

PART II  CT and MR Imaging of the Whole Body s

B

A

Fused coronals

FIG 39-29  Coronal fused F FDG PET-CT images in a patient (A) before and (B) after neoadjuvant chemo18

therapy. Though the anatomic distribution of tumor is unchanged, significant decrease in radiotracer uptake indicates treatment response.

A

B FIG 39-30  Axial CT images in patient who has undergone extrapleural pneumonectomy. A, Near-complete filling of the pneumonectomy space with fluid soon after surgery. B, Emptying of the space 7 days later secondary to a central bronchopleural fistula.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

C

B

D FIG 39-31  Radiograph (A) and coronal CT image (B) showing gastric herniation into the left chest after left extrapleural pneumonectomy. C and D, In a different patient who has undergone right-sided pneumonectomy, C, demonstrates rightward shift of the mediastinum into the pneumonectomy space. D, demonstrates a radiodense implant subsequently placed in the right hemithorax to displace the mediastinum toward the left.

A

B FIG 39-32  A, In a patient with biopsy-proven dedifferentiated liposarcoma, an axial CT image shows a heterogeneous mass filling the right hemithorax. B, Coronal T2-weighted MRI shows the mass to be predominantly isointense relative to fat.

1121

1122

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 39-33  Axial T1-weighted (A), T2-weighted (B), and postcontrast T1-weighted (C) MRI images and (D) coronal postcontrast T1-weighted image in a patient with pleural synovial sarcoma showing nodular soft tissue with multiloculated fluid-filled components. There is rim enhancement following administration of IV contrast, and tissue heterogeneity exemplifying the “triple sign” of hyper-, iso-, and hypointense areas on T2-weighted MRI.

Neoadjuvant chemoradiotherapy and en bloc resection have improved survival16,17,50 (Fig. 39-35).

CHEST WALL Approach to Imaging the Chest Wall Radiography.  Though many chest wall masses are evaluable on physical exam, radiography may be useful to delineate the size, location, and growth rate of the mass. Furthermore, radiographs may characterize soft tissue calcification, tumor matrix, ossification, and bone involvement. Low-kVp imaging (as used in bone radiography) is superior to high-kVp imaging (as used in chest radiography) for defining calcifications and soft tissue planes in fat-containing tumors such as lipomas.115 Dedicated breast imaging is indicated for lesions involving the breasts.

Ultrasound.  Ultrasound using a high-frequency transducer (e.g., 7.5-10 MHz) is useful for evaluating superficial chest wall and breast masses and is commonly used to guide biopsy.88,115

CT.  CT provides detailed information about tumor morphology, tissue characteristics, location, and extent. Enhancement characteris-

tics may refine differential diagnosis. Performing the CT scan in a modified scan plane and using multiplanar reformations may be useful when the lesion is supraclavicular or oriented parallel to the ribs.115 Particular strengths of CT with respect to chest wall masses are its ability to evaluate lesional calcification and bony involvement.88 CT may be useful to guide biopsy in cases of osteomyelitis involving the chest wall.19 In cases of destructive rib lesions, CT is superior to MRI in assessing cortical bone destruction. Several benign processes may simulate malignancy; for example, a subacute or old rib fracture may mimic a chondroid lesion or expansile neoplasm. Also, nonmalignant rib erosion may occur with benign processes such as granulomatous disease, radiation osteitis, and fibrous dysplasia.

MRI.  MRI refines tissue characterization, allowing identification of fat, vascularity, and other specific features that narrow differential diagnosis, and is superior for evaluating transspatial lesions and neurovascular involvement.88 It is considered the imaging modality of choice for evaluation of chest wall lesions by some authors.115 Compared to CT, MRI is more accurate in assessing the extent of chest wall masses and bone marrow infiltration. Superficial coils are recommended for evaluation of chest wall lesions; torso coils are recommended for evaluation of lesions that involve deeper portions of the

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

1123

A

C

B FIG 39-34  Coronal CT (A), MRI (B), and PET-CT (C) images in a patient with undifferentiated pleomorphic sarcoma, showing a heterogeneous pleural and parenchymal mass within the right chest, with enhancement and diffuse and heterogeneous FDG uptake.

thorax.115 Fat suppression highlights focal enhancement on postcontrast T1-weighted images.

PET.  FDG-PET provides information about the metabolic activity of a lesion, which is useful in assessment of metastatic or locally invasive lesions and guides biopsy to higher-yield cellular areas.88

Congenital Chest Wall Anomalies Congenital chest wall anomalies are briefly summarized in Table 39-7.58 Radiographs and CT are commonly indicated for evaluation of these disorders (Figs. 39-36 through 39-38).

Infectious Conditions of the Chest Wall Primary chest wall infection is rare. Physical examination may underestimate extent, especially in immunocompromised patients. Risk

factors include diabetes mellitus, immunocompromise, trauma, and surgical instrumentation.19 Klebsiella pneumoniae and Staphylococcus aureus are the most common bacterial agents. In chest wall infection, ultrasound, CT, and MRI are complementary modalities. MRI is the modality of choice for specific assessment of the soft tissues and can discern infected tissue from adjacent inflammation even without IV contrast, whereas CT is preferred for evaluating bony destruction. Multiplanar reformatted images in both modalities add valuable information about extent of infection and interspace communication within loculated collections and aid in surgical planning. Chest wall abscesses demonstrate low signal intensity on T1-weighted images and high signal intensity on STIR and T2-weighted images, with rim enhancement on postcontrast fat-saturated T1-weighted images. T1-weighted sequences are key for anatomic

1124

PART II  CT and MR Imaging of the Whole Body

B

C

A

FIG 39-35  A, PA chest radiograph in a patient presenting with Horner’s syndrome, miosis, and right arm weakness, showing subtle right apical pleural thickening. B and C, Axial CT and MRI showing a superior sulcus mass with encasement of the right brachial plexus and blood vessels, consistent with an unresectable Pancoast tumor.

TABLE 39-7  Congenital Chest

Wall Anomalies Pectus excavatum (funnel chest)

Pectus carinatum (pigeon breast) Cervical rib

Cleidocranial dysostosis Poland syndrome

• Depression of the sternum thought to be due to misdirected growth of the lower costal cartilages • Quantified using the Haller or pectus index on CT imaging—calculated by dividing the transverse by the AP diameter of the chest. Normal value is ≈ 2.56, and a value > 3.25 necessitates surgical correction. • Outward bowing of the sternum • Sometimes isolated, commonly seen in patients with congenital heart disease • Supernumerary rib articulating with the lowest cervical vertebra • The true second rib articulates with the sternomanubrial joint. • Asymptomatic in 90% of cases but may be associated with thoracic outlet or PagetSchroetter syndrome. Dedicated vascular imaging with multiplanar reformatted images may be useful. • Defective development of many bones (pubic bones, long bones, vertebrae) including the clavicles, which are variably hypoplastic • Complete or partial absence of the pectoralis major • Associated with ipsilateral syndactyly, rib hypoplasia, and aplasia of the ipsilateral breast or nipple

assessment because of their superior spatial resolution; STIR and T2-weighted images add contrast resolution that is helpful for assessing abscess extent and location. CT is more sensitive for detecting gas and calcification and superior to MRI in differentiating among substances that cause susceptibility artifact (e.g., calcium, gas, blood products, and metallic surgical materials). Osteomyelitis is most commonly due to contiguous spread of pulmonary infection (tuberculosis, fungal infection) or empyema (termed empyema necessitatis). S. aureus and Pseudomonas aeruginosa are the most common causal organisms. Bone destruction is generally not evident by radiography for 1 to 2 weeks; CT and MRI are capable of identifying commonly associated soft tissue masses, effacement of deep soft tissue planes, and periosteal elevation.58 MRI is highly sensitive and specific and thus considered the modality of choice for osteomyelitis involving the thoracic bones. Typical MRI findings include hypointensity against the high background signal of fat on T1-weighted images without fat suppression, corresponding high signal edema on fluid-sensitive sequences, and focal enhancement. IV drug use, immunosuppression, and chronic renal failure predispose to sternoclavicular osteomyelitis.58 Subclavian venous catheter-associated infection by S. aureus is a common scenario in this type of infection, with standard treatment being removal of the infected catheter and IV antibiotics. Surgical resection is reserved for cases in which infection is shown to extend beyond the joint on imaging. Nonpyogenic infection of the chest wall is more unusual. Mycobacterium tuberculosis rarely involves the chest wall. Hematogenous seeding is more common than contiguous spread. Chest wall abscess and sinus tracts are seen in about 25% of cases. Destructive chest wall masses with calcification and rim enhancement on CT are typical.58 Nocardia asteroides and Actinomyces israelii are bacterial species that infect the chest wall by tracking through the tissue planes separating

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B

C

E

D

F

G

H FIG 39-36  Axial (A), coronal (B), and volume-rendered images (C and D) of a patient with pectus excavatum. Axial (E), coronal (F), and volume-rendered images (G and H) after repair of pectus excavatum. Note the improvement in the chest deformity and displacement of the heart.

1125

1126

PART II  CT and MR Imaging of the Whole Body

FIG 39-37  PA radiograph of the chest showing cervical ribs (arrow).

the chest wall from the pleural space (empyema necessitatis). On crosssectional imaging, findings include soft tissue thickening and nonspecific inflammatory changes, fluid collections, draining sinus tracts, wavy periostitis, and rib destruction.58,60 Aspergillus fumigatus is a fungal species that produces multiple pulmonary manifestations. Chest wall invasion may occur, with osteolysis and fistula formation58 (Fig. 39-39).

Chest Wall Masses Benign chest wall lesions typically present as slow-growing, minimally symptomatic palpable masses. Bone erosion without destruction, resulting from relatively long-standing pressure effects, suggests benignity.

Benign Chest Wall Masses.  A convenient paradigm for evaluating benign chest wall masses classifies lesions by tissue origin. Lesions of bone and cartilaginous origin.  Benign lesions of bone and cartilage that occur in the chest wall include fibrous dysplasia, enchondroma, osteochondroma, chondroblastoma, aneurysmal bone cyst, and giant cell tumor.58 These lesions are discussed in depth elsewhere. Tumors of adipocyte origin.  Lipomas are the overall most common chest wall soft tissue tumors and also the most common benign chest wall mass.58 They are most frequently found in patients over age 50 years and are more common in obese patients. Chest wall lipomas tend to be deep seated, involving deep muscular and intermuscular layers, and less well circumscribed than lipomas elsewhere. In two thirds of cases, lipomas are composed solely of adipose tissue, but nonadipose tissue such as connective tissue septa and calcification is present in the remainder.

On CT, lipomas are circumscribed fat-attenuation lesions that lack enhancement. On MRI, lipomas follow fat signal intensity on all sequences, with suppression on fat-saturated sequences including STIR. There may be sparse fibrous septa that demonstrate low signal intensity on T1- and T2-weighted images and faint enhancement (n.b., thick, irregular, avidly enhancing septa or avidly enhancing nodules are concerning for malignancy).58,60 Chemical shift artifact at interfaces with more fluid-rich tissues is characteristic. A rare variant is spindle cell lipoma, which arises in the subcutaneous tissues of the neck or shoulder in middle aged to elderly men, appearing as a 3- to 5-cm mass that is heterogeneous on both T1- and T2-weighted sequences.58 Evaluation of extent, local infiltration, and relationship with neurovascular structures by MRI is essential for surgical planning (Fig. 39-40). Vascular and lymphatic lesions.  The term hemangioma encompasses many entities that can be classified according to the predominant vessel type (arteriovenous, capillary, mixed vascular, cavernous [composed of dilated, tortuous, thin walled vessels], and venous). Many lesions include nonvascular tissues including fat, smooth muscle, thrombus, fibrous tissue, and hemosiderin. Phleboliths are most commonly seen in cavernous hemangiomas but may be seen in other lesions.92 Hemangiomas are rare in the chest wall and may be found in the synovium or musculature. These lesions usually present in the pediatric population. On CT, most cavernous hemangiomas appear as heterogeneously attenuating soft tissue masses. The heterogeneity is due to the presence of fat (which may represent a substantial portion of the lesion) and vascular or fibrous components. Background attenuation is similar to skeletal muscle.92 Slow-flowing blood in dilated vessels leads to the characteristic MRI appearance of poorly defined curvilinear structures that are low in signal on T1-weighted sequences, with high intensity on T2-weighted images. Enhancement is marked.92 Muscle atrophy may be observed with intramuscular hemangiomas. As they are elsewhere, arteriovenous malformations (AVMs) of the chest wall are characterized by tubular flow voids. Glomus tumors most often present as painful solitary masses in adult patients. Less commonly, tumors are multiple and asymptomatic; rarely, glomus tumors are malignant. The neoplastic cells that comprise glomus tumors resemble smooth muscle cells of the normal glomus body. Lesions are often intramuscular. The typical appearance on CT is a soft tissue mass with adjacent osseous erosion. On MRI, the mass displaces major vessels and is surrounded by tortuous vessels arising from a vascular pedicle. Heterogeneous signal intensity and sharp interfaces with adjacent tissues after IV contrast administration are typical features.115 Lymphangiomas are uncommon hamartomatous lesions composed of dilated lymphatic vessels that most often present in early childhood as a neck mass that extends into the chest wall or axilla.92 Lesions are commonly present at the cervicothoracic junction in children and in the mediastinum in adults (probably because they are occult on physical exam).58,92 The constituent lymphatic channels are sequestered from the lymphatic system. MRI is the preferred imaging modality because CT appearance is nonspecific. Lymphangiomas follow simple fluid signal on all sequences, with thin peripheral and septal enhancement115 (Fig. 39-41). Peripheral nerve and nerve sheath tumors.  Spinal nerve roots, intercostal nerves, and distal brachial plexus branches may give rise to neurogenic chest wall tumors. The appearance of these tumors in the chest wall is similar to their appearance in other locations. There is overlap in the imaging appearances of neurofibroma and schwannoma, but the imaging features discussed below are often helpful in distinguishing them from one another.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

B

C FIG 39-38  Axial CT images (A and B) and volume-rendered image (C) showing absence of left pectoralis major muscle in a patient with Poland syndrome.

A

B FIG 39-39  A and B, Axial CT images showing a right chest wall mycobacterial abscess and a sinus tract in the left anterior chest wall in an immune-suppressed patient.

1127

1128

PART II  CT and MR Imaging of the Whole Body

C

B

A

FIG 39-40  Prone axial CT image (A), coronal T2-weighted image (B), and coronal T2-weighted image with fat suppression (C) showing a right-sided elastofibroma dorsi.

B A

C FIG 39-41  Axial T1-weighted MRI pre- (A) and post contrast (B) and T2-weighted MRI (C) showing nonenhancing, tubular, dilated lymphatics with minimal peripheral enhancement in a patient with a slow-flow right chest wall lymphatic malformation. Incidentally, a lesion is noted in the right lobe of the liver.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

1129

C

B

E

D

F FIG 39-42  A, PA radiograph of a patient with a neurogenic tumor in the left posterior mediastinum. Axial CT (B) and PET image (C) showing avid posterior mediastinal mass; D-E, axial T1-weighted pre- and postcontrast images showing heterogeneous enhancement and F, Coronal T2-weighted image showing high signal representing myxoid degeneration.

Schwannomas are benign nerve sheath tumors that arise from spinal nerve roots or intercostal nerves. They tend to be slow growing and are most commonly seen in young to middle-aged adult patients. Small lesions are well circumscribed, but larger lesions may be more lobulated. Scalloping rather than erosion of adjacent bones is typical. The usual CT appearance of schwannoma is a homogeneous mass  that is iso- or slightly hypoattenuating to muscle. A less common appearance is a cyst with fibrous walls, containing smaller nodules. Schwannomas tend to be isointense on T1- and markedly hyperintense on T2-weighted images, with avid enhancement. Tumors may

demonstrate the fascicular sign, with heterogeneous low signal intensity in a ringlike pattern. The nerve of origin may be identifiable. Degeneration—manifested as myxoid cystic change, hemorrhage, calcification, and fibrosis if present—is more common in schwannomas than in neurofibromas. Biopsy is painful owing to intrinsic innervation115 (Fig. 39-42). Neurofibromas may occur in isolation or in neurofibromatosis type 1 (NF1). Most affected patients present between the ages of 20 and 40. Typically, neurofibromas originate from spinal nerve roots and may be localized (most common), diffuse, or plexiform. Plexiform tumors

1130

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 39-43  Axial T1-weighted (A), T2-weighted (B), and postcontrast T1-weighted (C) MRI showing a hypointense, bilobed chest wall mass with intense enhancement post IV contrast administration, consistent with an fibroma dorsi.

involve long segments of nerves or multiple nerve branches. Like other slow-growing peripheral nerve tumors, they tend to remodel without eroding bone—for instance, expanding neural foramina when they arise from spinal nerve roots. Malignant degeneration occurs rarely  but is more common in cases of NF1. Malignant degeneration may  be suspected in cases of rapid growth.92 The CT appearance of neurofibroma is a low-attenuation heterogeneously enhancing lesion. On T2-weighted and postgadolinium T1-weighted images, a target appearance is typical. The center of the lesion is composed of fibrous stroma that is hypointense on T2-weighted images and demonstrates enhancement, and the periphery is composed of myxoid material that does not enhance significantly. Ganglioneuromas and paragangliomas arise from sympathetic nervous elements. Ganglioneuromas appear as sharply marginated, homogeneous or heterogeneous paravertebral masses that may be calcified. Curvilinear bands (fibrous tissue) that are hypointense on both T1- and T2-weighted images may give these lesions a whorled or septated appearance. Paragangliomas are more commonly right-sided paravertebral masses that occur in adolescents and young adults. They tend to be homogeneous and avidly-enhancing on cross-sectional imaging.92 Fibroblastic-myofibroblastic tumors.  Elastofibroma dorsi is a relatively common muscular pseudotumor that occurs characteristically at the inferior angle of the scapula. These tumors are much more common in women and tend to occur in middle-aged to elderly patients; lesions are often asymptomatic and are commonly bilateral. These lesions share attenuation and intensity characteristics with muscle on CT and MRI, respectively, with interspersed fat giving a layered appearance.92 With typical imaging characteristics, biopsy may be avoided92,115 (Fig. 39-43). Fibromatosis, also known as desmoid tumor, is a benign true neoplasm composed of fibroblasts and collagenous matrix that arises from connective tissue or posttraumatic or surgical scars. Typically, fibromatosis is infiltrative and locally aggressive. Attenuation and enhancement characteristics are variable at CT. The typical MRI appearance is a heterogeneous lesion that enhances moderately to avidly and contains bundles of collagen that are hypointense on

T2-weighted images and do not enhance.98 Recurrence is common after surgical resection.

Malignant Chest Wall Masses.  More than half of malignant chest wall masses are metastatic lesions or locally invasive breast, lung, or pleural tumors. The chest wall is an uncommon site for metastatic disease, and is usually involved only in the context of widespread metastasis.92,115 The most common primary chest wall malignancies are sarcomas, approximately half of which originate from the soft tissues and half of which are of cartilaginous or bony origin. Sarcomas.  Chondrosarcomas are the most common primary chest wall malignancy. These tumors are most often anteriorly located, arising from the costochondral junctions or sternum. Up to 20% arise from the ribs themselves. In about 10% of cases, chondrosarcoma is due to malignant transformation of a benign lesion such as enchondroma. Typically, chondrosarcomas are irregularly shaped, with bone destruction and variable calcification. Chondroid mineralization in a ring-and-arc configuration is characteristic and best demonstrated by CT. On MRI, chondrosarcomas are lobulated heterogeneously enhancing masses that are isointense to muscle on T1-weighted sequences and iso- or hyperintense relative to fat on T2-weighted images. Osteosarcomas are rare in the thorax and are most often found in the ribs, scapula, and clavicle in young adult patients. Metastatic involvement of the lungs and lymph nodes is more common with chest wall osteosarcoma than with peripheral osteosarcoma. On CT there is varied calcification that is greatest centrally and decreases toward the periphery. T1-weighted images show the lesion to be hyperintense relative to muscle, and T2-weighted images demonstrate high signal intensity. Enhancement is heterogeneous. Liposarcomas are classified by the World Health Organization into five groups based on histologic composition: well-differentiated (most common), dedifferentiated, myxoid, pleomorphic, and mixed. Approximately 50% to 75% of a typical well-differentiated liposarcoma is fat. Features that favor well-differentiated liposarcoma over lipoma on MRI include septa greater than 2 mm thick and nodular nonadipose components; however, in 4% to 9% of cases, well-differentiated

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

C

B

D

1131

FIG 39-44  Chest PA (A) and lateral (B) radiographs and CT scans (C and D) in a patient with a left upper chest wall Askin’s tumor.

liposarcoma and lipoma may be indistinguishable. Other histologic types of liposarcoma contain less fat (usually <25% by volume). Pleomorphic liposarcoma may be totally devoid of fat.92,99 Undifferentiated pleomorphic sarcoma (UPS) rarely involves the chest wall and demonstrates similar imaging features as in the pleural form (discussed above). Ewing’s sarcoma is a small round blue cell tumor of soft tissue origin that most commonly presents in adolescent boys as a painful, rapidly enlarging mass. Masses are usually solitary and most commonly occur in the ribs, scapulae, clavicles, or sternum. On CT, Ewing’s sarcoma usually appears as a large, poorly defined, heterogeneous mass that may contain calcification. MRI findings are similar in Ewing’s and non-Ewing’s sarcomas.

Nonsarcomatous primary chest wall malignancy.  Askin’s tumor is an uncommon primitive neuroectodermal tumor that is histologically similar to rhabdomyosarcoma, neuroblastoma, Ewing’s sarcoma, lymphoma, and small cell osteosarcoma. Askin’s tumors generally arise in the chest wall or in a paravertebral location in young adults. CT and MRI show a mass that causes erosion of associated  ribs or vertebral bodies. On MRI the tumor is heterogeneous because of focal hemorrhage and tumor necrosis, hyperintense to muscle  on T1-weighted images, and hyperintense on T2-weighted images  (Fig. 39-44). Uncommon primary tumors that may affect the chest wall include chondroblastomas, osteoblastomas, melanomas, lymphomas, rhabdomyosarcomas, and lymphangiosarcomas.

1132

PART II  CT and MR Imaging of the Whole Body radiograph, inference of thoracic spinal fracture may be made by the presence of signs of associated hematoma (lifting of the paraspinal lines, apical pleural cap), decreased vertebral body height, lateral offset of the vertebral bodies, increased interpediculate and interspinous distances, and rib disarticulation.68 Posterior mediastinal or paravertebral hematoma is an ancillary finding on CT.

DIAPHRAGM Imaging the Diaphragm

FIG 39-45  Axial CT image in patient involved in a motor vehicle accident, showing transection of the aorta and left rib fractures.

Chest Wall Trauma

Rib Fractures.  Rib fractures are common. Unless they are multiple (as in flail chest, discussed below), or if there is associated damage to the pleura or lung, they are of minimal clinical significance.85 Therefore a frontal chest radiograph without dedicated rib studies usually suffices for initial evaluation.50 Rib fractures may be missed on more than half of chest radiographs. CT is much more sensitive (though not necessary in all cases) and has been cited as a key factor in the decreasing morbidity and mortality associated with rib fractures.74 Fractures of the first through third ribs or multiple ribs suggests high-velocity trauma and underscores the need for thorough evaluation of the rest of the thorax, including the great vessels.50,110 Fractures of the tenth through twelfth ribs should prompt evaluation of the liver, kidneys, and spleen50 (Fig. 39-45). Uncomplicated rib fractures may cause pain and respiratory splinting; the treatment of choice is analgesia.

Flail Chest.  The term flail chest has been defined variously. A common definition is fracture of multiple contiguous ribs, with each broken in two or more places (segmental fracture). Compromised (sometimes paradoxical) respiratory excursion, due to flail chest has been shown to result in significant morbidity and mortality.50 Patients often require mechanical ventilation and surgical fixation.110 Thin-section CT with multiplanar reformatted images is often helpful in evaluating the flail segment prior to surgery.124

Diaphragmatic abnormalities are often initially detected on chest radiographs, which may be followed by CT, ultrasound, or MRI.  The fluoroscopic “sniff ” test is a dynamic test for diaphragmatic dysfunction.88 Normally the right hemidiaphragm is slightly higher than the left, and the anterior and medial portions are higher than the posterior and lateral portions. The diaphragm may be elevated, depressed, or distorted in a number of pathologic states. For instance, pulmonary volume loss, increased intraabdominal pressure, or phrenic nerve disease may elevate the diaphragm, whereas emphysema or cystic lung disease may depress it. Pleural thickening may result in tethering of the diaphragm to the chest wall, which produces elevation and contour abnormality (Fig. 39-46). Subpulmonic effusion is a common mimic of hemidiaphragmatic elevation.58,92 Real-time imaging using ultrasound or fluoroscopy (the sniff test) may be used to assess diaphragmatic motion with respiration.94 CT is the modality of choice for diaphragmatic trauma and is discussed subsequently. MRI plays a complementary role in evaluation of diaphragmatic trauma and is the modality of choice for assessment of tumoral invasion of the diaphragm. MRI for dynamic assessment of diaphragmatic movement has been described but is not in widespread use because of the low cost and widespread availability of other modalities.94

Congenital Diaphragmatic Hernias

Morgagni Hernias.  Morgagni hernias represent less than 10% of diaphragmatic hernias. These are anterior hernias that result from failure of fusion of the transverse septum (which later becomes the central tendon of the diaphragm) to the lateral body wall.99

Bochdalek Hernias.  Bochdalek hernias are posterior hernias that result from abnormal development of the pleuroperitoneal folds or failure of fusion of the folds and transverse septum to the intercostal muscles. These represent 90% of congenital diaphragmatic hernias and are usually left-sided.92

Diaphragmatic Paralysis Unilateral diaphragmatic paresis or paralysis is usually asymptomatic and appears as elevation of one hemidiaphragm. Bilateral paresis or paralysis, however, often leads to respiratory failure. Dysfunction anywhere along the neuromuscular axis from the central nervous system to the muscle itself may lead to diaphragmatic paralysis.94

Sternal Fracture.  Sternal fracture is common in high-velocity

Diaphragmatic Eventration

trauma and is more apparent on lateral than frontal radiographs. The presence of anterior mediastinal hematoma is a secondary sign. Costochondral separation may occur in younger patients. Like fracture of the upper ribs, sternal fracture should prompt close inspection of the critical intrathoracic structures.50

Eventration is congenital thinning of the diaphragm that leads to focal bulging. The anteromedial portion of the right hemidiaphragm is  the most common site of involvement. On imaging there is elevation of only a portion of a hemidiaphragm (in contradistinction to diaphragmatic paresis or paralysis), with a sharp transition at the edge of the eventration. Unlike in congenital diaphragmatic hernia, a thin layer of diaphragm remains between the abdominal structures and the thorax.

Thoracic Spine Fracture.  Thoracic spine injuries are much more readily detectable by CT than plain radiographs. On the frontal chest

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm

A

1133

B

FIG 39-46  A, Sagittal CT image in patient with focal herniation of the liver through the diaphragm mimicking a fibrous tumor of the pleura, confirmed with Eovist coronal MRI (B).

Diaphragmatic Trauma Traumatic diaphragmatic rupture is an underrecognized entity because of many factors, including the frequent presence of distracting injuries in severe polytrauma (e.g., hemothorax, pneumothorax, rib fracture, splenic and hepatic injury, long bone fracture, and head trauma)  and the variable imaging appearance of the normal and injured  diaphragm.10,28 However, when diaphragmatic rupture goes unrecognized, complications such as herniation of abdominal organs with incarceration and ischemia may occur. When structures near the diaphragm are injured, diaphragmatic injury should be specially sought to avoid missed diagnosis.10 It has been estimated that diaphragmatic injury occurs in 0.8%  to 8% of patients with blunt abdominal trauma. Prevalence of diaphragmatic injury from blunt or penetrating injury depends on the patient’s geographic location and socioeconomic status; however, given that penetrating injury is more commonly treated with surgical exploration, the majority of missed diagnoses are in cases of blunt trauma.28 The most common mechanism for blunt diaphragmatic rupture is high-kinetic-energy frontal or lateral impacts (e.g., motor vehicle collisions) that cause sudden deformation of the rib cage or elevation of intraabdominal pressure.10 Diaphragmatic defects tend to be smaller in penetrating trauma (<2 cm in 85% of cases), whereas injuries in blunt trauma are larger (10-15 cm). The higher reported prevalence of injury to the left hemidiaphragm than the right (≈3 : 1) is thought to be due to multiple factors, including congenital weakness at the fusion point of the lumbar and costal diaphragm (the so-called lumbar triangle), the presence of the esophageal hiatus, the protective effect of the liver on the right side that serves to dampen transmission of elevated intraabdominal pressure, and difficulty in identifying right-sided injury because of

the close apposition of the diaphragm and liver. Indeed, right-sided diaphragmatic injury may be underdiagnosed.10 Initial radiographs are normal or nonspecific in 20% to 50% of cases.10 Radiographic signs of diaphragmatic injury include herniation of abdominal structures into the thorax, with or without constriction of the herniated organs at the herniation sign (collar sign), identification of the nasogastric tube tip in a supradiaphragmatic position, and elevation of the affected hemidiaphragm (though this sign is somewhat nonspecific). Mechanical ventilation may lower sensitivity of chest radiography by increasing the intrathoracic pressure and thus depressing the affected hemidiaphragm.10,28 Ultrasound and fluoroscopic gastrointestinal studies also play a limited role but may be considered in selected cases. CT is considered the gold standard for evaluation of diaphragmatic injury. Advancements in CT technology (e.g., helical CT with highquality axial images and multiplanar reformations) have significantly improved its diagnostic ability since initial reports. Numerous CT signs of blunt diaphragmatic injury have been reported in the literature, including in excellent review articles by Bocchini et al.10 and Desir et al.28; a selection is summarized in Table 39-8. Broadly, these signs have been divided into direct (direct visualization of the diaphragmatic defect) and indirect (inference of a diaphragmatic defect by identification of ancillary findings) categories. Usually the diagnosis is made on the basis of multiple signs (Fig. 39-47). MRI plays a limited role in the evaluation of acutely injured patients but may be useful when CT is equivocal or in long-term follow-up. Cardiac and respiratory gating is critical to minimize artifacts. On T1-weighted imaging, the normal diaphragm is a continuous hypointense band against normal relatively hyperintense organs and fat in the abdomen and mediastinum. IV contrast and fat suppression may improve diagnostic accuracy.10,28

1134

PART II  CT and MR Imaging of the Whole Body

TABLE 39-8  Selected CT Signs of Blunt Diaphragmatic Injury10,28 CT Sign Direct CT Signs Segmental diaphragmatic defect

Dangling diaphragm Absent diaphragm Indirect CT Signs Collar sign (variants are the “hump” and “band” signs)

Findings

Potential Mimics

• Abrupt focal loss of continuity of the diaphragm • Edge of defect may appear thickened secondary to edema or hemorrhage • Sagittal and coronal MPR helpful • Free edge of torn diaphragm visible as curvilinear soft tissue structure perpendicular to chest wall • At fat or air interfaces where clear visualization is expected • Usually associated with a large hernia

• Normal variant in the uninjured population

• Waist-like narrowing of herniated intraabdominal structures • May appear as focal subtle indentation of involved organ(s) • MPR helpful

• Traumatic hepatic fracture or normal diaphragmatic slips may produce FP • Also seen in congenital or atraumatic acquired hernias • Also seen in congenital or atraumatic acquired hernias • Congenital or hiatal hernia may produce FP

Herniated viscera

• Abdominal contents in the thorax

Dependent viscera

• Upper third of liver or bowel (on the right) and stomach, spleen, or bowel (on the left) in direct contact with posterior ribs in supine patient; no normal separation by lung in the costophrenic sulcus

• FP where the diaphragm apposes fluid or soft tissue

FP, false positives; MPR, multiplanar reformatted imaging.

A

B FIG 39-47  Patient involved in trauma 6 months prior now presenting with diaphragmatic rupture. A, Coronal CT scan shows that the intraabdominal contents are within the chest. B, Coronal CT scan post repair.

REFERENCES 1. Abramowitz Y, Simanovsky N, Goldstein MS, et al: Pleural effusion: Characterization with CT attenuation values and CT appearance. AJR Am J Roentgenol 192(3):618–623, 2009. doi: 10.2214/AJR.08.1286. 2. Alexander ES, Proto AV, Clark RA: CT differentiation of subphrenic abscess and pleural effusion. AJR Am J Roentgenol 140(1):47–51, 1983. doi: 10.2214/ajr.140.1.47. 3. Aquino SL, Webb WR, Gushiken BJ: Pleural exudates and transudates: Diagnosis with contrast-enhanced CT. Radiology 192(3):803–808, 1994. doi: 10.1148/radiology.192.3.8058951.

4. Armato SG, Labby ZE, Coolen J, et al: Imaging in pleural mesothelioma: A review of the 11th International Conference of the International Mesothelioma Interest Group. Lung Cancer 82(2):190– 196, 2013. doi: 10.1016/j.lungcan.2013.08.005. 5. Attanoos RL, Galateau-Salle F, Gibbs AR, et al: Advances in surgical pathology: Mesothelioma, Philadelphia, 2013, Lippincott Williams & Wilkins. 6. Baber CE, Hedlund LW, Oddson TA, et al: Differentiating empyemas and peripheral pulmonary abscesses: The value of computed tomography. Radiology 135(3):755–758, 1980. doi: 10.1148/ radiology.135.3.7384467.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm 7. Bansal A, Chawla M, Cohen PJ, et al: Pleural epithelioid hemangioendothelioma. Lung 190(4):469–470, 2012. doi: 10.1007/ s00408-012-9377-9. 8. Baysal T, Bulut T, Gökirmak M, et al: Diffusion-weighted MR imaging of pleural fluid: Differentiation of transudative vs exudative pleural effusions. Eur Radiol 14(5):890–896, 2004. doi: 10.1007/ s00330-003-1995-4. 9. Bénard F, Sterman D, Smith RJ, et al: Metabolic imaging of malignant pleural mesothelioma with fluorodeoxyglucose positron emission tomography. Chest 114(3):713–722, 1998. 10. Bocchini G, Guida F, Sica G, et al: Diaphragmatic injuries after blunt trauma: Are they still a challenge? Reviewing CT findings and integrated imaging. Emerg Radiol 19(3):225–235, 2012. doi: 10.1007/ s10140-012-1025-4. 11. Boland GW, Gazelle GS, Girard MJ, et al: Asymptomatic hydropneumothorax after therapeutic thoracentesis for malignant pleural effusions. AJR Am J Roentgenol 170(4):943–946, 1998. doi: 10.2214/ajr.170.4.9530040. 12. Bonomo L, Feragalli B, Sacco R, et al: Malignant pleural disease. Eur J Radiol 34(2):98–118, 2000. 13. Brant A, Eaton T: Serious complications with talc slurry pleurodesis. Respirology 6(3):181–185, 2001. 14. Brun A-L, Touitou-Gottenberg D, Haroche J, et al: Erdheim-Chester disease: CT findings of thoracic involvement. Eur Radiol 20(11):2579– 2587, 2010. doi: 10.1007/s00330-010-1830-7. 15. Brutsche MH, Tassi G-F, Györik S, et al: Treatment of sonographically stratified multiloculated thoracic empyema by medical thoracoscopy. Chest 128(5):3303–3309, 2005. doi: 10.1378/chest.128.5.3303. 16. Bruzzi JF, Komaki R, Walsh GL, et al: Imaging of non-small cell lung cancer of the superior sulcus: Part 1: Anatomy, clinical manifestations, and management. Radiographics 28(2):551–560, quiz 620, 2008. doi: 10.1148/rg.282075709. 17. Bruzzi JF, Komaki R, Walsh GL, et al: Imaging of non-small cell lung cancer of the superior sulcus: Part 2: Initial staging and assessment of resectability and therapeutic response. Radiographics 28(2):561–572, 2008. doi: 10.1148/rg.282075710. 18. Cardinale L, Ardissone F, Garetto I, et al: Imaging of benign solitary fibrous tumor of the pleura: A pictorial essay. Rare Tumors 2(1):e1, 2010. doi: 10.4081/rt.2010.e1. 19. Chelli Bouaziz M, Jelassi H, Chaabane S, et al: Imaging of chest wall infections. Skeletal Radiol 38(12):1127–1135, 2009. doi: 10.1007/ s00256-008-0636-z. 20. Coolen J, De Keyzer F, Nafteux P, et al: Malignant pleural disease: Diagnosis by using diffusion-weighted and dynamic contrast-enhanced MR imaging—Initial experience. Radiology 263(3):884–892, 2012. doi: 10.1148/radiol.12110872. 21. Copley SJ, Lee YCG, Hansell DM, et al: Asbestos-induced and smoking-related disease: Apportioning pulmonary function deficit by using thin-section CT. Radiology 242(1):258–266, 2007. doi: 10.1148/ radiol.2421051167. 22. Crotty EJ, McAdams HP, Erasmus JJ, et al: Epithelioid hemangioendothelioma of the pleura: Clinical and radiologic features. AJR Am J Roentgenol 175(6):1545–1549, 2000. doi: 10.2214/ ajr.175.6.1751545. 23. Cugell DW, Kamp DW: Asbestos and the pleura: A review. Chest 125(3):1103–1117, 2004. 24. Cullu N, Kalemci S, Karakaş O, et al: Efficacy of CT in diagnosis of transudates and exudates in patients with pleural effusion. Diagn Interv Radiol 2013. doi: 10.5152/dir.2013.13066. 25. Davies CWH, Gleeson FV, Davies RJO, et al: BTS guidelines for   the management of pleural infection. Thorax 58(Suppl 2):ii18–ii28, 2003. 26. Davis SD, Henschke CI, Yankelevitz DF, et al: MR imaging of pleural effusions. J Comput Assist Tomogr 14(2):192–198, 1990. 27. de Campos JR, Vargas FS, de Campos Werebe E, et al: Thoracoscopy talc poudrage: A 15-year experience. Chest 119(3):801–806, 2001. 28. Desir A, Ghaye B: CT of blunt diaphragmatic rupture. Radiographics 32(2):477–498, 2012. doi: 10.1148/rg.322115082.

1135

29. DeVita VT, Lawrence TS, Rosenberg SA: DeVita, Hellman, and Rosenberg’s cancer, Philadelphia, 2008, Lippincott Williams & Wilkins. 30. Devouassoux G, Lantuejoul S, Chatelain P, et al: Erdheim-Chester disease: A primary macrophage cell disorder. Am J Respir Crit Care Med 157(2):650–653, 1998. doi: 10.1164/ajrccm.157.2.97-03109. 31. Duysinx B, Nguyen D, Louis R, et al: Evaluation of pleural disease with 18-fluorodeoxyglucose positron emission tomography imaging. Chest 125(2):489–493, 2004. 32. Dwyer A: The displaced crus: a sign for distinguishing between pleural fluid and ascites on computed tomography. J Comput Assist Tomogr 2(5):598–599, 1978. 33. Ellis R: Incomplete border sign of extrapleural masses. JAMA 237(25):2748, 1977. 34. Evans AL, Gleeson FV: Radiology in pleural disease: State of the art. Respirology 9(3):300–312, 2004. doi: 10.1111/j.1440-1843.2004.00599.x. 35. Falaschi F, Battolla L, Mascalchi M, et al: Usefulness of MR signal intensity in distinguishing benign from malignant pleural disease. AJR Am J Roentgenol 166(4):963–968, 1996. doi: 10.2214/ajr.166.4.8610582. 36. Falaschi F, Battolla L, Zampa V, et al: [Comparison of computerized tomography and magnetic resonance in the assessment of benign and malignant pleural diseases]. Radiol Med 92(6):713–718, 1996. 37. Foroulis CN, Zarogoulidis P, Darwiche K, et al: Superior sulcus (Pancoast) tumors: Current evidence on diagnosis and radical treatment. J Thorac Dis 5(Suppl 4):S342–S358, 2013. doi: 10.3978/j. issn.2072-1439.2013.04.08. 38. Freundlich IM, Bragg DG: A radiologic approach to diseases of the chest, Baltimore, 1997, Lippincott Williams & Wilkins. 39. Friedberg JS, Mick R, Culligan M, et al: Photodynamic therapy and the evolution of a lung-sparing surgical treatment for mesothelioma.   Ann Thorac Surg 91(6):1738–1745, 2011. doi: 10.1016/j. athoracsur.2011.02.062. 40. Gaur P, Dunne R, Colson YL, et al: Bronchopleural fistula and the   role of contemporary imaging. J Thorac Cardiovasc Surg 148(1):341– 347, 2013. doi: 10.1016/j.jtcvs.2013.11.009. 41. Gerbaudo VH, Britz-Cunningham S: Primary lesions were more 18F-FDG avid in patients with distant spread than in those without extensive disease. Mol Imaging Biol 11(6):487, author reply 488, 2009. 42. Gerbaudo VH, Sugarbaker DJ, Britz-Cunningham S, et al: Assessment of malignant pleural mesothelioma with (18)F-FDG dual-head gamma-camera coincidence imaging: Comparison with histopathology. J Nucl Med 43(9):1144–1149, 2002. 43. Giesel FL, Bischoff H, von Tengg-Kobligk H, et al: Dynamic contrastenhanced MRI of malignant pleural mesothelioma: A feasibility study of noninvasive assessment, therapeutic follow-up, and possible predictor of improved outcome. Chest 129(6):1570–1576, 2006. doi: 10.1378/chest.129.6.1570. 44. Gill RR, Gerbaudo VH, Jacobson FL, et al: MR imaging of benign and malignant pleural disease. Magn Reson Imaging Clin N Am 16(2):319– 339, x, 2008. doi: 10.1016/j.mric.2008.03.004. 45. Gill RR, Gerbaudo VH, Sugarbaker DJ, et al: Current trends in radiologic management of malignant pleural mesothelioma. Semin Thorac Cardiovasc Surg 21(2):111–120, 2009. doi: 10.1053/j. semtcvs.2009.06.011. 46. Gill RR, Poh AC, Camp PC, et al: MDCT evaluation of central airway and vascular complications of lung transplantation. AJR Am J Roentgenol 191(4):1046–1056, 2008. doi: 10.2214/AJR.07.2691. 47. Gill RR, Umeoka S, Mamata H, et al: Diffusion-weighted MRI of malignant pleural mesothelioma: Preliminary assessment of apparent diffusion coefficient in histologic subtypes. AJR Am J Roentgenol 195(2):W125–W130, 2010. doi: 10.2214/AJR.09.3519. 48. Glazer M, Berkman N, Lafair JS, et al: Successful talc slurry pleurodesis in patients with nonmalignant pleural effusion. Chest 117(5):1404– 1409, 2000. 49. Görg C, Bert T, Görg K: Contrast-enhanced sonography for differential diagnosis of pleurisy and focal pleural lesions of unknown cause. Chest 128(6):3894–3899, 2005. doi: 10.1378/chest.128.6.3894. 50. Hansell DM, Lynch DA, McAdams HP, et al: Imaging of diseases of the chest, St. Louis, 2009, Elsevier.

1136

PART II  CT and MR Imaging of the Whole Body

51. Heffner JE, Klein JS, Hampson C: Diagnostic utility and clinical application of imaging for pleural space infections. Chest 137(2):467– 479, 2010. doi: 10.1378/chest.08-3002. 52. Helm EJ, Matin TN, Gleeson FV: Imaging of the pleura. J Magn Reson Imaging 32(6):1275–1286, 2010. doi: 10.1002/jmri.22372. 53. Hew M, Heinze S: Chest ultrasound in practice: A review of utility   in the clinical setting. Intern Med J 42(8):856–865, 2012. doi: 10.1111/j.1445-5994.2012.02816.x. 54. Hierholzer J, Luo L, Bittner RC, et al: MRI and CT in the differential diagnosis of pleural disease. Chest 118(3):604–609, 2000. 55. Im JG, Webb WR, Han MC, et al: Apical opacity associated with pulmonary tuberculosis: High-resolution CT findings. Radiology 178(3):727–731, 1991. doi: 10.1148/radiology.178.3.1994409. 56. Inan N, Arslan A, Akansel G, et al: Diffusion-weighted MRI in the characterization of pleural effusions. Diagn Interv Radiol 15(1):13–18, 2009. 57. Jeebun V, Stenton SC: The presentation and natural history of asbestos-induced diffuse pleural thickening. Occup Med (Lond) 62(4):266–268, 2012. doi: 10.1093/occmed/kqs028. 58. Jeung MY, Gangi A, Gasser B, et al: Imaging of chest wall disorders. Radiographics 19(3):617–637, 1999. 59. Jiménez Castro D, Díaz G, Pérez-Rodríguez E, et al: Prognostic features of residual pleural thickening in parapneumonic pleural effusions. Eur Respir J 21(6):952–955, 2003. 60. Kanne JP, Yandow DR, Mohammed T-LH, et al: CT findings of pulmonary nocardiosis. AJR Am J Roentgenol 197(2):W266–W272, 2011. doi: 10.2214/AJR.10.6208. 61. Kearney SE, Davies CW, Davies RJ, et al: Computed tomography and ultrasound in parapneumonic effusions and empyema. Clin Radiol 55(7):542–547, 2000. doi: 10.1053/crad.1999.0480. 62. Kim EY, Kim TS, Han J, et al: Thoracic epithelioid hemangioendothelioma: Imaging and pathologic features. Acta Radiol 52(2):161–166, 2011. doi: 10.1258/ar.2010.100292. 63. Kramer H, Pieterman RM, Slebos D-J, et al: PET for the evaluation of pleural thickening observed on CT. J Nucl Med 45(6):995–998, 2004. 64. Kraus GJ: The split pleura sign. Radiology 243(1):297–298, 2007. doi: 10.1148/radiol.2431041658. 65. Kumar V, Abbas AK, Fausto N, et al: Robbins & Cotran pathologic basis of disease, St. Louis, 2009, Elsevier. 66. Kwek BH, Aquino SL, Fischman AJ: Fluorodeoxyglucose positron emission tomography and CT after talc pleurodesis. Chest 125(6):2356– 2360, 2004. 67. Lang-Lazdunski L, Bille A, Lal R, et al: Pleurectomy/decortication is superior to extrapleural pneumonectomy in the multimodality management of patients with malignant pleural mesothelioma. J Thorac Oncol 7(4):737–743, 2012. doi: 10.1097/JTO.0b013e31824ab6c5. 68. Lawrason JN, Novelline RA, Rhea JT, et al: Early detection of thoracic spine fracture in the multiple trauma patient: Role of the initial portable chest radiograph. Emerg Radiol 4(5):309–319, 1997. 69. Leung AN, Müller NL, Miller RR: CT in differential diagnosis of   diffuse pleural disease. AJR Am J Roentgenol 154(3):487–492, 1990. doi: 10.2214/ajr.154.3.2106209. 70. Lichtenstein DA: Lung ultrasound in the critically ill. Ann Intensive Care 4(1):1, 2014. doi: 10.1186/2110-5820-4-1. 71. Lichtenstein DA, Menu Y: A bedside ultrasound sign ruling out pneumothorax in the critically ill. Lung sliding. Chest 108(5):1345– 1348, 1995. 72. Light RW: Parapneumonic effusions and empyema. Proc Am Thorac Soc 3(1):75–80, 2006. doi: 10.1513/pats.200510-113JH. 73. Light RW, Macgregor MI, Luchsinger PC, et al: Pleural effusions: The diagnostic separation of transudates and exudates. Ann Intern Med 77(4):507–513, 1972. 74. Livingston DH, Shogan B, John P, et al: CT diagnosis of rib fractures and the prediction of acute respiratory failure. J Trauma 64(4):905–911, 2008. doi: 10.1097/TA.0b013e3181668ad7. 75. Luciano C, Francesco A, Giovanni V, et al: CT signs, patterns and differential diagnosis of solitary fibrous tumors of the pleura. J Thorac Dis 2(1):21–25, 2010.

76. Marom EM, Patz EF, Erasmus JJ, et al: Malignant pleural effusions: Treatment with small-bore-catheter thoracostomy and talc pleurodesis. Radiology 210(1):277–281, 1999. doi: 10.1148/radiology.210.1 .r99dc04277. 77. Maskell NA, Butland RJA: Pleural Diseases Group, Standards of Care Committee, British Thoracic Society. BTS guidelines for the investigation of a unilateral pleural effusion in adults. Thorax 58(Suppl 2):ii8–ii17, 2003. 78. Mason RJ, Broaddus VC, Martin T, et al: Murray and Nadel’s textbook of respiratory medicine, St. Louis, 2010, Elsevier. 79. McLoud TC, Flower CD: Imaging the pleura: Sonography, CT, and MR imaging. AJR Am J Roentgenol 156(6):1145–1153, 1991. doi: 10.2214/ ajr.156.6.2028857. 80. McLoud TC, Isler R, Head J: The radiologic appearance of chemical pleurodesis. Radiology 135(2):313–317, 1980. doi: 10.1148/ radiology.135.2.7367618. 81. McLoud TC, Isler RJ, Novelline RA, et al: The apical cap. AJR Am J Roentgenol 137(2):299–306, 1981. doi: 10.2214/ajr.137.2.299. 82. McLoud TC, Woods BO, Carrington CB, et al: Diffuse pleural thickening in an asbestos-exposed population: Prevalence and causes. AJR Am J Roentgenol 144(1):9–18, 1985. doi: 10.2214/ajr.144.1.9. 83. Metintas M, Ucgun I, Elbek O, et al: Computed tomography features in malignant pleural mesothelioma and other commonly seen pleural diseases. Eur J Radiol 41(1):1–9, 2002. 84. Miles SE, Sandrini A, Johnson AR, et al: Clinical consequences of asbestos-related diffuse pleural thickening: A review. J Occup Med Toxicol 3:20, 2008. doi: 10.1186/1745-6673-3-20. 85. Miller LA: Chest wall, lung, and pleural space trauma. Radiol Clin North Am 44(2):213–224, viii, 2006. doi: 10.1016/j.rcl.2005.10.006. 86. Minami M, Kawauchi N, Yoshikawa K, et al: Malignancy associated with chronic empyema: Radiologic assessment. Radiology 178(2):417– 423, 1991. doi: 10.1148/radiology.178.2.1987602. 87. Mirzoyan M, Muslimani A, Setrakian S, et al: Primary pleuropulmonary synovial sarcoma. Clin Lung Cancer 9(5):257–261, 2008. doi: 10.3816/ CLC.2008.n.040. 88. Mullan CP, Madan R, Trotman-Dickenson B, et al: Radiology of   chest wall masses. AJR Am J Roentgenol 197(3):W460–W470, 2011. doi: 10.2214/AJR.10.7259. 89. Murray JG, Patz EF, Erasmus JJ, et al: CT appearance of the pleural space after talc pleurodesis. AJR Am J Roentgenol 169(1):89–91, 1997. doi: 10.2214/ajr.169.1.9207506. 90. Naidich DP, Megibow AJ, Hilton S, et al: Computed tomography of the diaphragm: Peridiaphragmatic fluid localization. J Comput Assist Tomogr 7(4):641–649, 1983. 91. Naidich DP, Müller NL, Webb WR: Computed tomography and magnetic resonance of the thorax, Philadelphia, 2007, Lippincott Williams & Wilkins. 92. Nam SJ, Kim S, Lim BJ, et al: Imaging of primary chest wall tumors with radiologic-pathologic correlation. Radiographics 31(3):749–770, 2011. doi: 10.1148/rg.313105509. 93. Narayanaswamy S, Kamath S, Williams M: CT appearances of   talc pleurodesis. Clin Radiol 62(3):233–237, 2007. doi: 10.1016/ j.crad.2006.10.007. 94. Nason LK, Walker CM, McNeeley MF, et al: Imaging of the diaphragm: Anatomy and function. Radiographics 32(2):E51–E70, 2012. doi: 10.1148/rg.322115127. 95. Noppen M, De Waele M, Li R, et al: Volume and cellular content of normal pleural fluid in humans examined by pleural lavage. Am J Respir Crit Care Med 162(3 Pt 1):1023–1026, 2000. 96. O’Connell RS, McLoud TC, Wilkins EW: Superior sulcus tumor: Radiographic diagnosis and workup. AJR Am J Roentgenol 140(1):25– 30, 1983. doi: 10.2214/ajr.140.1.25. 97. O’Connor AR, Morgan WE: Radiological review of pneumothorax. BMJ 330(7506):1493–1497, 2005. doi: 10.1136/bmj.330.7506.1493. 98. O’Sullivan P, O’Dwyer H, Flint J, et al: Soft tissue tumours and mass-like lesions of the chest wall: A pictorial review of CT and   MR findings. Br J Radiol 80(955):574–580, 2007. doi: 10.1259/ bjr/16591964.

CHAPTER 39  Disease of the Pleura, Chest Wall, and Diaphragm 99. O’Sullivan P, O’Dwyer H, Flint J, et al: Malignant chest wall neoplasms of bone and cartilage: A pictorial review of CT and MR findings. Br J Radiol 80(956):678–684, 2007. doi: 10.1259/bjr/82228585. 100. Pass HI, Temeck BK, Kranda K, et al: Preoperative tumor volume is associated with outcome in malignant pleural mesothelioma. J Thorac Cardiovasc Surg 115(2):310–317, discussion 317-8, 1998. 101. Patz EF, Rusch VW, Heelan R: The proposed new international TNM staging system for malignant pleural mesothelioma: Application to imaging. AJR Am J Roentgenol 166(2):323–327, 1996. doi: 10.2214/ ajr.166.2.8553940. 102. Raj V, Kirke R, Bankart MJ, et al: Multidetector CT imaging of pleura: Comparison of two contrast infusion protocols. Br J Radiol 84(1005):796–799, 2011. doi: 10.1259/bjr/55980445. 103. Renner RR, Pernice NJ: The apical cap. Semin Roentgenol 12(4):299– 302, 1977. 104. Roach HD, Davies GJ, Attanoos R, et al: Asbestos: When the dust settles—An imaging review of asbestos-related disease. Radiographics 22(Spec No):S167–S184, 2002. doi: 10.1148/radiographics.22.suppl_1. g02oc10s167. 105. Robinson LA: Solitary fibrous tumor of the pleura. Cancer Control 13(4):264–269, 2006. 106. Robinson BWS, Lake RA: Advances in malignant mesothelioma. N Engl J Med 353(15):1591–1603, 2005. doi: 10.1056/NEJMra050152. 107. Rosado-de-Christenson ML, Abbott GF, McAdams HP, et al: From the archives of the AFIP: Localized fibrous tumor of the pleura. Radiographics 23(3):759–783, 2003. doi: 10.1148/rg.233025165. 108. Schouwink JH, Kool LS, Rutgers EJ, et al: The value of chest computer tomography and cervical mediastinoscopy in the preoperative assessment of patients with malignant pleural mesothelioma. Ann Thorac Surg 75(6):1715–1718, discussion 1718-9, 2003. 109. Shen KR, Wain JC, Wright CD, et al: Postpneumonectomy syndrome: Surgical management and long-term results. J Thorac Cardiovasc Surg 135(6):1210–1216, discussion 1216-9, 2008. doi: 10.1016/j. jtcvs.2007.11.022. 110. Sirmali M, Türüt H, Topçu S, et al: A comprehensive analysis of traumatic rib fractures: Morbidity, mortality and management. Eur J Cardiothorac Surg 24(1):133–138, 2003. 111. Stark DD, Federle MP, Goodman PC, et al: Differentiating lung abscess and empyema: Radiography and computed tomography. AJR Am J Roentgenol 141(1):163–167, 1983. doi: 10.2214/ajr.141.1.163. 112. Stewart D, Waller D, Edwards J, et al: Is there a role for pre-operative contrast-enhanced magnetic resonance imaging for radical surgery in malignant pleural mesothelioma? Eur J Cardiothorac Surg 24(6):1019– 1024, 2003. 113. Sugarbaker DJ, Flores RM, Jaklitsch MT, et al: Resection margins, extrapleural nodal status, and cell type determine postoperative

114.

115.

116.

117. 118.

119.

120.

121.

122.

123. 124.

125.

126.

127.

1137

long-term survival in trimodality therapy of malignant pleural mesothelioma: Results in 183 patients. J Thorac Cardiovasc Surg 117(1):54–63, discussion 63-5, 1999. Sugarbaker DJ, Gill RR, Yeap BY, et al: Hyperthermic intraoperative pleural cisplatin chemotherapy extends interval to recurrence and survival among low-risk patients with malignant pleural mesothelioma undergoing surgical macroscopic complete resection. J Thorac Cardiovasc Surg 145(4):955–963, 2013. doi: 10.1016/j.jtcvs.2012.12.037. Tateishi U, Gladish GW, Kusumoto M, et al: Chest wall tumors: Radiologic findings and pathologic correlation: Part 1. Benign tumors. Radiographics 23(6):1477–1490, 2003. doi: 10.1148/rg.236015526. Teplick JG, Teplick SK, Goodman L, et al: The interface sign: A computed tomographic sign for distinguishing pleural and intraabdominal fluid. Radiology 144(2):359–362, 1982. doi: 10.1148/ radiology.144.2.7089290. Tsao AS, Wistuba I, Roth JA, et al: Malignant pleural mesothelioma.   J Clin Oncol 27(12):2081–2090, 2009. doi: 10.1200/JCO.2008.19.8523. Van Schil PE, Opitz I, Weder W, et al: Multimodal management of malignant pleural mesothelioma: Where are we today? Eur Respir J 2014. doi: 10.1183/09031936.00207213. van Zandwijk N, Reid G, Linton A, et al: Radical surgery for malignant pleural mesothelioma: Have we identified the appropriate selection tools? Ann Cardiothorac Surg 1(4):481–486, 2012. doi: 10.3978/ j.issn.2225-319X.2012.10.01. Verschakelen JA, Demaerel P, Coolen J, et al: Rounded atelectasis of the lung: MR appearance. AJR Am J Roentgenol 152(5):965–966, 1989. doi: 10.2214/ajr.152.5.965. Veyssier-Belot C, Cacoub P, Caparros-Lefebvre D, et al: ErdheimChester disease. Clinical and radiologic characteristics of 59 cases. Medicine (Baltimore) 75(3):157–169, 1996. Waite RJ, Carbonneau RJ, Balikian JP, et al: Parietal pleural changes   in empyema: Appearances at CT. Radiology 175(1):145–150, 1990. doi: 10.1148/radiology.175.1.2315473. Webb WR, Higgins CB: Thoracic imaging, Philadelphia, 2011, Lippincott Williams & Wilkins. Weyant MJ, Bleier JI, Naama H, et al: Severe crushed chest injury with large flail segment: Computed tomographic three-dimensional reconstruction. J Trauma 52(3):605, 2002. Yildiz S, Cece H, Turksoy O: Discriminative role of CT in exudative and transudative pleural effusions. AJR Am J Roentgenol 195(4):W305, author reply W306, 2010. doi: 10.2214/AJR.10.4437. Yousem SA: Pulmonary apical cap: A distinctive but poorly recognized lesion in pulmonary surgical pathology. Am J Surg Pathol 25(5):679– 683, 2001. Yu H: Management of pleural effusion, empyema, and lung abscess. Semin Intervent Radiol 28(1):75–86, 2011. doi: 10.1055/s-0031-1273942.

40  Airway Phillip M. Boiselle

NORMAL ANATOMY Trachea and Main Bronchi The trachea comprises 16 to 22 C-shaped cartilages that are linked longitudinally by annular ligaments of fibrous and connective tissue.30 The cartilages are connected posteriorly by the membranous wall of the trachea, which lacks cartilage and is supported by the trachealis muscle, which is composed of transverse smooth muscle fibers that narrow the tracheal lumen on contraction.30 During forced expiration, this narrowing results in normal anterior bulging of the posterior membrane, resulting in a mean decrease in anteroposterior dimension of 32%67 (Fig. 40-1). Excessive expiratory collapsibility, also referred to as tracheobronchomalacia and generally defined as greater than 50% expiratory reduction in cross-sectional area of the lumen, may occur secondary to weakening or deficiency of the cartilaginous or membranous components of the airway wall. The trachea extends from the inferior margin of the cricoid cartilage to the carina, a ridge that marks the origin of the main bronchi30 (Fig. 40-2). Its length averages 10 to 11 cm in adults, including both the extrathoracic (2-4 cm) and intrathoracic (6-9 cm) portions.30 Tracheal length is dynamic, however, and changes with respiration and neck flexion and extension.30 The trachea is generally midline in position, but it is often displaced slightly to the right at the level of the aortic arch, with a greater degree of displacement in the setting of a markedly tortuous atherosclerotic aorta. Thyroid enlargement may also result in displacement of the trachea, often in association with luminal narrowing. The proximal trachea lies close to the skin surface, but the trachea angles posteriorly as it courses inferiorly in the thorax, eventually achieving a midcoronal location at the level of the carina (Fig. 40-3). Because of the angled course of the trachea, axial computed tomography (CT) images do not provide a true perpendicular cross section of its lumen. Tracheal diameter in men averages 19.5 mm and ranges from 13 to 25 mm in the coronal plane and 13 to 27 mm in the sagittal plane.12,20,30,72 In women, tracheal diameter averages 17.5 mm and ranges from 10 to 21 mm in the coronal dimension and 10 to 23 mm in the sagittal dimension.12,20,30,72 Tracheomegaly is defined in women as tracheal diameter greater than 21 mm in the coronal dimension and 23 mm in the sagittal dimension, and in men as tracheal diameter greater than 25 mm in the coronal dimension and 27 mm in the sagittal dimension.75 On axial CT images, the normal tracheal lumen most commonly demonstrates an oval, round (see Fig. 40-1A), or horseshoe shape.20 The tracheal index can be calculated by dividing the coronal diameter (in millimeters) by the sagittal diameter (in millimeters). The normal value is approximately 1.12,20,30,72 A saber sheath trachea (Fig. 40-4) refers

1138

to a configuration in which there is marked coronal narrowing and accentuation of the sagittal diameter (sagittal-to-coronal ratio > 2).24,70 This finding is frequently associated with chronic obstructive lung disease. In contrast, a “lunate” trachea (Fig. 40-5A) refers to accentuation of the coronal diameter, with a relative narrowing of the sagittal diameter (coronal-to-sagittal ratio > 1).42 This finding is frequently associated with airway malacia (excessive expiratory collapsibility of the airway lumen) (see Fig. 40-5B). The tracheal wall is composed of several layers, including an inner mucosa layer, submucosa, cartilage or muscle, and an outer adventitial layer.73 On axial CT images, the tracheal wall is usually visible as a 1- to 3-mm soft tissue density stripe, demarcated internally by the air-filled tracheal lumen and externally by the adjacent fat density of the mediastinum73 (Fig. 40-6). The posterior wall is typically thinner than the anterior and lateral walls. Cartilage within the tracheal wall may normally appear slightly denser than surrounding soft tissue and fat.73 Calcification of cartilage may be observed in older patients, especially women.73 Calcified thickening of the tracheal wall is also associated with tracheal pathology, particularly relapsing polychondritis. Thickening of the airway wall (with or without calcification) is an important sign of tracheal pathology (Fig. 40-7). Axial images are the reference standard for assessing tracheal wall thickening, a finding  that may be overlooked at conventional bronchoscopy and threedimensional (3D) reconstruction images. The distribution of wall thickening can be helpful in limiting the differential diagnosis of the various causes of tracheal stenosis. For example, disorders of cartilage (e.g., relapsing polychondritis [see Fig. 40-7], tracheobronchopathia osteochondroplastica) will spare the posterior membranous wall of the trachea, whereas other disorders generally result in circumferential thickening (Fig. 40-8). The main bronchi arise from the trachea at the level of the carina and course obliquely to the axial plane. Owing to the limitations of axial images for assessing structures that course obliquely, multiplanar and 3D reconstruction images are particularly helpful for evaluating caliber changes in the mainstem bronchi.

Lobar, Segmental, and Subsegmental Bronchi Similar to the pulmonary vasculature, the airways divide by dichotomous branching, with roughly 23 generations of branches between the trachea and alveoli.51 Bronchial walls are composed of both cartilaginous and fibromuscular elements. For bronchi with luminal diameters less than 5 mm, the bronchial wall typically measures between one sixth and one tenth of the airway lumen.51 Thus because identification of a bronchus on CT depends on visualization of the thin airway wall, typically only the first eight generations of bronchi can be visualized with thin-section high-resolution CT (HRCT) techniques. As a general

CHAPTER 40  Airway

1139

A

FIG 40-3  Normal trachea. Minimum intensity projection, sagittal oblique reformation CT image of the normal trachea (arrow) shows the typical angled course of the trachea from a proximal anterior location to a midcoronal position at the level of the carina.

B FIG 40-1  Normal tracheal dynamics. End-inspiratory image (A) shows a round configuration of the tracheal lumen. At end-expiration (B), the tracheal lumen has narrowed slightly, with flattening of the posterior membranous wall (arrow).

FIG 40-4  Saber sheath trachea. Axial CT image demonstrates an elongated sagittal dimension of the trachea with relative narrowing coronally, consistent with a saber sheath configuration.

FIG 40-2  Normal trachea. Minimum intensity projection, coronal oblique reformation CT image of the normal trachea and bronchi. Main bronchi originate from the inferior trachea at the level of the carina (arrow).

rule the limit of visibility of normal bronchi on CT is generally a 2-mm-diameter bronchus that is typically located about 2 cm from the pleural surface.51 On CT the appearance of bronchi depends on their course with respect to the axial plane. For example, bronchi that course horizontally within the axial plane (e.g., right and left upper lobe bronchi and anterior upper lobe segmental bronchi; superior segment lower lobe bronchi) will be viewed along their long axes (Fig. 40-9), whereas those that course perpendicularly (e.g., bronchus intermedius, lower lobe bronchi and basilar segments, upper lobe apical and apical-posterior segments) will be viewed as small circular lucencies18 (Fig. 40-10). Bronchi that course obliquely to the axial plane are more difficult to evaluate with CT than other bronchi are. On axial images, they are typically seen as oval lucencies that can be connected by scrolling a series of axial slices. These bronchi include the distal right middle lobe bronchus and its medial and lateral segmental branches (Fig. 40-11)

1140

PART II  CT and MR Imaging of the Whole Body

A

B FIG 40-5  Lunate trachea. Axial end-inspiratory CT image (A) shows an elongated coronal dimension of the trachea. Axial dynamic expiratory CT image (B) shows severe malacia with complete collapse of the tracheal lumen (arrow).

FIG 40-6  Normal trachea. Axial CT image shows normal thickness of the tracheal wall. Note the paper-thin quality of the posterior membranous wall (white arrow) and the slightly thicker anterior and lateral walls (black arrows).

FIG 40-8  Postintubation stenosis. Axial CT image shows circumferential wall thickening of the subglottic airway lumen.

and the lingular bronchus and its superior and inferior segmental divisions. Bronchi and pulmonary arteries have a parallel course, and the bronchial lumen is generally about the same diameter as the lumen of its accompanying pulmonary artery. This relationship can be helpful for identifying bronchiectasis, a term that refers to irreversible dilation of a bronchus. When a bronchus is dilated, its lumen appears larger than that of the accompanying pulmonary artery. When viewed in cross section, this mimics the appearance of a signet ring (the “signet ring sign”) (Fig. 40-12). Identification of a dilated bronchus that courses horizontally is usually based on a lack of normal tapering of the bronchus as it courses from proximal to distal. Subsegmental bronchi show considerable variability in their branching patterns, and there is currently no widely accepted nomenclature for their description.51 FIG 40-7  Relapsing polychondritis. Axial CT image shows thickening of the anterolateral tracheal walls, with focal calcification (arrow). Note the characteristic sparing of the posterior membranous wall.

Bronchioles Bronchioles are small airways that do not contain cartilage and  glands.18 They are composed of membranous (lobular and terminal)

CHAPTER 40  Airway

A

1141

B FIG 40-9  CT appearance of bronchi that course horizontally within the axial plane. A, Axial CT image shows the right upper lobe bronchus (arrow). B, Axial CT image shows the superior segment right lower lobe bronchus (arrow).

A

B FIG 40-10  CT appearance of bronchi that course perpendicularly to the axial plane. A, Axial CT image shows bronchus intermedius (arrow). B, Axial CT image shows lower lobe segmental bronchi (arrows).

FIG 40-11  CT appearance of bronchi that course obliquely to the axial plane. Axial CT image shows the origin of the right middle lobe medial and lateral segmental bronchi (arrows).

FIG 40-12  Signet ring sign. The lumen of the dilated bronchus (arrow) is larger than its accompanying pulmonary artery, mimicking a signet ring.

1142

PART II  CT and MR Imaging of the Whole Body

A

B FIG 40-13  Subglottic stenosis. A, Axial CT image shows subtle asymmetric wall thickening (arrow) that could easily be overlooked or underestimated. B, Minimum intensity projection multiplanar reformation image of the trachea shows moderate tracheal stenosis (arrows).

bronchioles and respiratory bronchioles. Because of their small size (internal diameter ≤ 2 mm), normal bronchioles are below the resolution of HRCT. However, in the setting of bronchiolar disease, both direct and indirect signs of disease may be visualized on HRCT. Terminal bronchioles and their accompanying centrilobular pulmonary arterial branches are located in the center of the secondary pulmonary lobule.18 This location accounts for the characteristic centrilobular distribution of bronchiolar abnormalities on CT, which are reviewed further in the final section of this chapter.

IMAGING TECHNIQUES With the latest generation of multidetector-row CT (MDCT) scanners, thin-section images of the entire central airways can be obtained in only a few seconds, creating an isotropic data set in which the resolution is the same in the axial, coronal, and sagittal planes. Compared to standard helical CT scanners, MDCT provides higher spatial resolution, faster speed, greater anatomic coverage, and higher-quality multiplanar reformation and 3D reconstruction images. Together, these features have combined to revolutionize CT imaging of the airways. Axial CT images provide anatomic information regarding the airway lumen, airway wall, and adjacent mediastinal and lung structures. Although axial CT images are considered the reference standard for airway imaging, it is important to be aware of the limitations of the axial plane for assessing the larger airways, including limited ability to detect subtle airway stenoses, underestimation of the craniocaudal extent of disease, difficulty displaying complex 3D relationships of the airway, and inadequate representation of airways oriented obliquely to the axial plane.9,11,49,59,60,62,64 These limitations have important implications for the assessment of certain airway disorders such as airway stenoses and complex congenital airway abnormalities. Multiplanar and 3D reconstruction images help overcome these limitations by providing a more anatomically meaningful display of complex structures

such as the airways.9,11,49,59,60,62,64 These images have been shown to enhance detection of airway stenoses (Fig. 40-13), aid assessment of the craniocaudal extent of stenoses, and clarify complex congenital airway abnormalities.54 They have also been shown to improve diagnostic confidence of interpretation, enhance preprocedural planning for bronchoscopy and surgery, and improve communication among radiologists, clinicians, and patients.9,11,64 For detection of bronchial and bronchiolar diseases, thin-section (1-1.5 mm) HRCT images should be obtained. Although this technique has traditionally been performed by obtaining noncontiguous images acquired at intervals of 1 or 2 cm, MDCT allows for a volumetric spiral acquisition with retrospective reconstruction of both thinsection and thick-section data sets. Such volumetric acquisition offers several advantages, including the enhanced ability to detect bronchiectasis44 (Fig. 40-14) and the possibility of reconstructing multiplanar or postprocessed images to facilitate detection and characterization of airway-related abnormalities. For example, it has been shown that contiguous 3-mm-collimation CT images are superior to noncontiguous HRCT images for detection of bronchiectasis (Fig. 40-15) and have significantly better interobserver agreement for the presence and extent of bronchiectasis.44 Anatomic imaging of the airways is performed at end-inspiration. However, several airways diseases, including tracheobronchomalacia (see Fig. 40-5B) and constrictive bronchiolar diseases, will be either undetected or underestimated if only end-inspiratory imaging is performed. Thus expiratory imaging should be performed as a complement to end-inspiratory imaging for suspected airway malacia and bronchiolar disorders. Additionally, expiratory imaging facilitates a comprehensive assessment of diffuse lung diseases that have an obstructive airways component, such as hypersensitivity pneumonitis (Fig. 40-16) and sarcoid. A variety of expiratory methods are available. For assessment of tracheomalacia, dynamic expiratory imaging (imaging during a forced expiration) has been shown to be superior to end-expiratory imaging

CHAPTER 40  Airway (imaging after exhalation).4 However, end-expiratory imaging is typically the standard technique for evaluating air trapping in the lung parenchyma, an indirect sign of constrictive bronchiolar disease. Similar to end-inspiratory HRCT imaging, end-expiratory HRCT was traditionally performed at selected noncontiguous levels. More recently, however, advances in CT technology have allowed for volumetric end-expiratory HRCT imaging.53 This method has potential

1143

advantages over noncontiguous axial HRCT in terms of enhanced visualization of the airways conducting to regions of air trapping and in the ability to create multiplanar reformation images, which have the potential to enhance the clarity of borders of air trapping compared to axial images.53 Although end-expiratory imaging is routinely applied to detect air trapping, other methods of expiratory imaging have also been reported.15,23,43 For example, free breathing methods and lateral decubitus positioning have been reported for evaluating pediatric patients who may not be able to cooperate with standard end-expiratory breathing maneuvers.

PATHOLOGY Tracheobronchial Stenosis

FIG 40-14  Bronchiectasis. Coronal reformation image demonstrates bronchiectasis and bronchial wall thickening in the left lower lobe. Coronal images enhance the continuous display of bronchi that course perpendicularly to the axial plane.

A

Tracheobronchial stenosis is defined as focal or diffuse narrowing of the tracheal lumen and may occur secondary to a wide variety of benign and malignant causes (Box 40-1). In the following paragraphs, imaging methods and a general approach to recognition and characterization of tracheobronchial stenoses are reviewed. Axial CT images provide exquisite anatomic display of the tracheal wall, tracheal lumen, and adjacent mediastinal and lung structures. Use of narrow collimation (≤3 mm) enhances the ability to detect and characterize tracheal wall thickening (defined as wall thickness > 3 mm) and calcification. However, axial images have a limited ability to detect subtle airway stenoses and frequently underestimate the craniocaudal extent of stenoses.9,11,49,58,60,64,69 By providing a continuous anatomic display of the airways, multiplanar and 3D reconstruction images help overcome these limitations.11 In settings where such reconstructions are not possible, careful review of contiguous thin-section axial images can help prevent the radiologist from overlooking stenoses. Use of thin-section images is particularly advantageous in the subglottic region, where stenoses are often focal and may be easily overlooked or underestimated on 5-mm sections (see Fig. 40-13). CT evaluation of suspected tracheal stenosis is routinely performed at end-inspiration during a single breath hold. An additional sequence during dynamic expiration may be helpful to assess for

B FIG 40-15  Enhanced detection of bronchiectasis with helical CT. A, Axial, prone, noncontiguous highresolution CT image demonstrates a small cystic space (arrow) in the right middle lobe with no discernible connection to the airway. B, Axial helical CT image identifies the connecting airway (arrow), allowing for a confident diagnosis of bronchiectasis.

1144

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 40-16  Hypersensitivity pneumonitis. A, Axial end-inspiratory CT image demonstrates mosaic lung attenuation with diffuse areas of ground-glass attenuation and a subtle focus (arrow) of reduced lung attenuation. B, Axial end-expiratory CT image demonstrates multiple foci of lobular air trapping (arrows) consistent with small airways disease.

BOX 40-1  Causes of Tracheobronchial

Stenosis

Iatrogenic Post intubation* Lung transplantation Idiopathic Infection Laryngotracheal papillomatosis† Rhinoscleroma Tuberculosis Neoplasm Primary tracheal neoplasm (squamous carcinoma, adenoid cystic carcinoma) Direct invasion (lung, esophagus, thyroid) Secondary neoplasm (breast, renal, melanoma, thyroid) Saber sheath deformity Systemic diseases Amyloidosis Inflammatory bowel disease Relapsing polychondritis Sarcoidosis Wegener’s granulomatosis Tracheobronchopathia osteochondroplastica Trauma *Most common etiology. † Although papillomas represent growth of new tissue, laryngotracheal papillomatosis is most frequently categorized as a nonneoplastic disease.

tracheomalacia, which may coexist with tracheobronchial stenoses, especially among patients with relapsing polychondritis, postintubation stenosis, and airway narrowing secondary to long-standing extrinsic compression (e.g., thyroid enlargement or mediastinal aortic vascular anomalies).8,13 It is important to accurately assess the location, length, and distribution of the stenosis as well as the presence, distribution, and type of

FIG 40-17  Amyloidosis. Axial CT image shows marked thickening of anterolateral tracheal walls (arrow) with foci of calcification. Note sparing of the posterior membranous wall. An identical appearance can be seen in relapsing polychondritis.

wall thickening. Consideration of these factors in combination with ancillary thoracic findings and pertinent clinical and laboratory data will allow the radiologist to effectively narrow the broad differential diagnosis of tracheobronchial stenosis to a few likely entities. In certain cases, such as relapsing polychondritis or tracheobronchopathia osteochondroplastica, a confident diagnosis can be made on the basis of imaging findings alone. As reviewed earlier in this chapter, the posterior membranous  wall of the trachea is devoid of cartilage. Thus diseases such as  relapsing polychondritis (see Fig. 40-7) or tracheobronchopathia osteochondroplastica—which target the cartilage—will characteristically spare the posterior tracheal wall,57,73 whereas most other causes of stenosis will result in circumferential tracheal wall involvement (see Fig. 40-13). Interestingly, amyloidosis may result in circumferential thickening or may spare the posterior wall (Fig. 40-17).

CHAPTER 40  Airway BOX 40-2  Differentiating Features of

BOX 40-3  Malignant and Benign

Sparing of Posterior Membranous Wall Relapsing polychondritis Tracheobronchopathia osteochondroplastica ± Amyloidosis (may also be circumferential)

Malignant Neoplasms Epithelial Squamous cell carcinoma Adenoid cystic carcinoma Carcinoid Mucoepidermoid carcinoma Adenocarcinoma Small cell undifferentiated carcinoma Large cell carcinoma Acinic cell carcinoma Malignant salivary gland–type mixed tumors Carcinomas with pleomorphic, sarcomatoid, or sarcomatous elements

Tracheobronchial Stenosis

“Hourglass” Configuration Post intubation Calcification Amyloid Relapsing polychondritis Tracheobronchopathia osteochondroplastica Tuberculosis Tracheomalacia (Coexisting with Stenosis) Post intubation Relapsing polychondritis Saber sheath trachea Long-standing extrinsic compression

Although there is significant overlap in the imaging features  of many causes of tracheobronchial stenosis, several key imaging features can help to effectively narrow this broad differential diagnosis. As listed in Box 40-2, these features include sparing of the posterior membranous wall, hourglass configuration, calcification, and coexisting tracheomalacia. Additionally, systemic disorders (e.g., amyloid, inflammatory bowel disease, polychondritis, sarcoid, tuberculosis, Wegener’s granulomatosis) are often associated with characteristic ancillary imaging and laboratory findings that may suggest their diagnosis. Postintubation stenosis is by far the most common cause of acquired tracheal stenosis.19 It typically occurs secondary to injury of the trachea from the high pressure of an endotracheal tube balloon against the wall of the trachea.19,57 This initially results in mucosal necrosis, followed by scarring and stenosis.19,57 A study comparing MDCT with the gold standard of bronchoscopy showed a high sensitivity (92%) and specificity (100%) for detection of postintubation stenosis.68 The most common CT finding in postintubation stenosis is a focal area of proximal tracheal luminal narrowing measuring approximately 2 cm in craniocaudal length19,57,73 (see Fig. 40-13). The focal nature and circumferential narrowing may produce a characteristic hourglass configuration. Less common findings include a thin membrane projecting into the tracheal lumen or a long segment of eccentric soft tissue thickening.19,57

Tracheobronchial Neoplasms Tracheobronchial neoplasms are rare; it has been estimated that a primary tracheal tumor is roughly 180 times less common than a primary lung cancer.31 A neoplasm within the central airways is more likely due to direct airway invasion by an adjacent secondary neoplasm originating from the thyroid, lung, or esophagus.16 The majority of primary tracheal neoplasms in adults are malignant.19,46 Although the differential diagnosis of central airway neoplasms is broad (Box 40-3), only five histologies comprise the majority of primary tracheobronchial tumors: squamous cell carcinoma, adenoid cystic carcinoma (Fig. 40-18), carcinoid, mucoepidermoid carcinoma, and squamous cell papilloma.22 The most common cell

1145

Tracheobronchial Neoplasms

Mesenchymal Fibrosarcoma Rhabdomyosarcoma Angiosarcoma Kaposi’s sarcoma Liposarcoma Osteosarcoma Leiomyosarcoma Chondrosarcoma Paraganglioma Spindle cell sarcoma Lymphoma Malignant fibrous histiocytoma

Benign Neoplasms Epithelial Squamous cell papilloma Papillomatosis Pleomorphic adenoma Glandular papilloma Adenomas of salivary gland type Mucous gland adenoma Monomorphic adenoma Oncocytoma Mesenchymal Hamartoma Neurofibroma Chondroma Fibroma Hemangioma Granular cell tumor Schwannoma Fibrous histiocytoma Pseudosarcoma Hemangioendothelioma Leiomyoma Chondroblastoma Lipoma Glomus tumor

type is squamous cell carcinoma, which is associated with cigarette smoking.19,46 The second most common cell type is adenoid cystic carcinoma, which is not smoking related.19,46 It has a better prognosis than squamous cell carcinoma, but late recurrence is relatively common. The most common benign tracheal neoplasm is squamous cell papilloma, which is also associated with cigarette smoking.19,46 Carcinoid tumors are neuroendocrine neoplasms that most commonly arise within the mainstem and lobar bronchi, but about 15% develop within the segmental bronchi or lung periphery.54,56 They are not associated with cigarette smoking. MDCT is the imaging modality of choice for detection and staging of central airway neoplasms. Multiplanar reformation and 3D reconstruction images complement conventional axial images by providing a more anatomically meaningful display of the neoplasm and its relationship to adjacent structures, and by accurately determining the craniocaudad extent of disease.11 Virtual bronchoscopy images provide a unique intraluminal perspective of the tumor and adjacent proximal and distal airways (see Fig. 40-18), which may not be accessible with conventional bronchoscopy in cases in which there is high-grade luminal narrowing or obstruction. Moreover, this method can provide a more global perspective of an endoluminal lesion than conventional bronchoscopy can. Thus it provides complementary information to conventional bronchoscopy. MDCT provides important information for preprocedural planning. If they are resectable, surgery is the optimal therapy for both benign and malignant airway neoplasms.31 Radiation therapy may be

1146

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 40-18  Adenoid cystic carcinoma. A, Axial CT scan shows a lobulated intraluminal mass (black arrow) at the level of the carina. An adjacent tubular structure represents a normal partially opacified azygos vein (white arrow). B, Virtual bronchoscopic image provides an intraluminal perspective of the mass (arrows).

FIG 40-19  Right lower lobe atelectasis (arrow) secondary to obstructing carcinoid tumor. The obstructing lesion is shown in Figure 40-20.

administered to patients with unresectable disease or as an adjunct to surgery, particularly in patients with adenoid cystic carcinoma.35 MDCT can determine whether a tumor is amenable to complete surgical resection, as well as the approach, type, and extent of surgical resection. MDCT provides critical information for the surgeon, including precise delineation of the 3D size of the neoplasm (e.g., craniocaudal length) and its relationship with other vital (e.g., vascular) structures. MDCT accurately defines the intraluminal and extraluminal extension of tumor, as well as postobstructive complications such as atelectasis (Fig. 40-19), pneumonia, and mucus plugging. Although MDCT can reliably detect the presence of lymphadenopathy, it cannot distinguish between hyperplastic and malignant nodes (Fig. 40-20). Hyperplastic nodes are frequently seen in the setting of postobstructive pneumonitis from an endoluminal neoplasm. Additionally, MDCT does not reliably detect microscopic mediastinal invasion or neural

FIG 40-20  Benign hyperplastic lymph nodes. Coronal oblique MIP image shows an enhancing right hilar mass (white arrow) with obstruction of the bronchus intermedius and distal right lower lobe collapse. Enlarged subcarinal lymph nodes (black arrow) proved to be benign hyperplastic nodes at biopsy.

invasion.37 MDCT may detect pulmonary and extrapulmonary metastases, thereby enhancing tumor staging. Finally, for patients who are not surgical candidates, MDCT can evaluate patency of the airways distal to the neoplasm, thereby determining suitability of the airways for palliative stent placement. When a discrete tracheal mass is identified on imaging studies, the diagnosis of a primary tracheal neoplasm can usually be made with a

CHAPTER 40  Airway high degree of confidence. CT can frequently suggest whether a tracheal lesion is malignant or benign.35 Although some overlap exists, benign lesions are typically less than 2 cm in diameter, with welldefined smooth borders and without evidence of contiguous tracheal thickening or mediastinal invasion. In contrast, malignant lesions usually vary in size between 2 and 4 cm in diameter, with a flat or polypoid shape and irregular or lobulated borders. Contiguous tracheal wall thickening and mediastinal invasion are frequently observed. Although CT usually cannot distinguish between neoplastic cell types, detection of fat within a lesion on CT is nearly pathognomonic for a hamartoma or lipoma, and identification of calcification within a lesion is highly suggestive of a chondroid tumor (chondroma, chondrosarcoma). Vigorous enhancement of a bronchial neoplasm is typical of a carcinoid tumor. In a minority of cases, tracheal neoplasms will present as eccentric or circumferential tracheal wall thickening rather than as a discrete mass (Fig. 40-21). In such cases the differential diagnosis includes tracheal stenosis from a variety of nonmalignant entities. In general the presence of marked irregularity of tracheal wall thickening and the presence of extratracheal extension favor a primary tracheal neoplasm, but biopsy is usually necessary to establish the diagnosis.

1147

Although bronchoscopy with functional maneuvers can reliably detect tracheobronchomalacia, it is not clinically feasible or desirable to perform this invasive test in all patients who present with chronic cough and other nonspecific respiratory symptoms. Fortunately, recent advances in MDCT afford the opportunity to noninvasively diagnose tracheobronchomalacia with sensitivity that is similar to that of conventional bronchoscopy.22,40,76 Changes in size of malacic trachea and bronchi depend on the difference between the intraluminal pressure inside the airways and the pleural (intrathoracic) pressure outside.4,13 Pleural pressure depends mostly on respiratory muscles and is high during expiratory efforts. In contrast, intraluminal pressures are highly variable and depend on airflow. When airflow is zero, intraluminal pressure equals alveolar

Tracheobronchomalacia Tracheobronchomalacia refers to weakness of the airway walls and/or supporting cartilage and is characterized by excessive expiratory collapse8,13,29,33,34 (Fig. 40-22; see Fig. 40-5). Although first reported in the 1930s and 1940s, this condition has only recently been recognized as a relatively common and potentially treatable cause of chronic cough, dyspnea, and recurrent infections.13 Tracheobronchomalacia may be either congenital or acquired. The acquired form is associated with a variety of risk factors and comorbidities, most notably chronic obstructive pulmonary disease8,13,33 (Box 40-4). Because it cannot be detected with routine end-inspiratory imaging studies, tracheobronchomalacia is widely considered an underdiagnosed condition; bronchoscopic series report a prevalence ranging from 4.5% to 23%.32,33,55 Recent CT studies also support the concept that this is an underdiagnosed condition,28,39 with a prevalence of 10% in a recent CT angiography study of patients with nonspecific respiratory symptoms referred for suspected pulmonary embolism.28

A

FIG 40-21  Adenoid cystic carcinoma. Axial CT image demonstrates eccentric circumferential wall thickening in subglottic airway (arrows).

B FIG 40-22  Tracheomalacia in patient with relapsing polychondritis. End-inspiratory image (A) of proximal trachea shows partially calcified anterolateral wall thickening; at dynamic-expiration (B), there is marked malacia (arrow). (The end-inspiratory image is shown in a soft tissue window setting to enhance display of wall thickening.)

1148

PART II  CT and MR Imaging of the Whole Body

BOX 40-4  Risk Factors for Acquired

Tracheomalacia

Chronic obstructive lung disease Posttraumatic and iatrogenic factors Post intubation Post tracheostomy Radiation therapy External chest trauma Post lung transplantation Chronic infection/bronchitis Chronic inflammation Relapsing polychondritis Chronic external compression of the trachea Paratracheal neoplasms (benign and malignant) Paratracheal masses (e.g., goiter, congenital cyst) Aortic aneurysms and vascular rings Skeletal abnormalities (e.g., pectus, scoliosis) FIG 40-23  Frown sign. Expiratory CT image demonstrates frownlike

pressure and differs from pleural pressure only by the elastic recoil pressure of the lung, which depends on lung volume. At maximal lung volume with no flow (end-inspiration), the intraluminal pressure is 20 to 30 cm H2O greater than pleural pressure, and the pressure difference expands the trachea. At low lung volumes with no flow (endexpiration), the intraluminal pressure is nearly equal to pleural pressure and the trachea is unstressed. The trachea is most compressed during cough and dynamic expiration at low lung volume when pleural pressure is high (≈100 cm H2O) and expiratory flow limitation in the small airways prevents transmission of the high alveolar pressures to the central airways. Under these conditions, intraluminal pressure is nearly atmospheric and the large transmural pressure causes tracheal collapse.74 Baroni and colleagues directly compared the ability of endexpiratory and dynamic expiratory CT imaging methods to elicit tracheobronchomalacia.5 Consistent with the principles of respiratory physiology, this study showed that dynamic expiratory CT elicited a significantly greater degree of tracheal collapse than end-expiratory CT did. Tracheobronchial collapse is thus ideally measured during forceful exhalation. Historically, a diagnosis of tracheobronchomalacia has been made using a threshold of 50% or more expiratory reduction in the airway lumen when using either bronchoscopy or CT.40 Recently, however, studies of healthy volunteers with normal pulmonary function have reported mean levels of expiratory collapse in the trachea and right and left main bronchi of 54%, 67%, and 61%, respectively.10,41,61 Moreover, in a minority of healthy individuals, the bronchus intermedius may show complete expiratory collapse. Although future studies are necessary to precisely determine a more rigorous threshold for diagnosing tracheobronchomalacia, it is likely that such a threshold will be defined as at least 70% expiratory collapse. In the meantime, to avoid overdiagnosis of this condition, identification of excessive airway collapse on CT or bronchoscopy should be closely correlated with clinical symptoms, risk factors, and pulmonary function testing.61 Interpretation of CT images for the diagnosis of tracheobronchomalacia requires careful review and comparison of both end-inspiratory and dynamic expiratory images. End-inspiratory images provide important anatomic information about tracheal size and shape, thickness of the tracheal wall, and presence or absence of extrinsic masses compressing the trachea. The tracheal lumen is almost always normal

configuration of the proximal trachea owing to excessive anterior displacement of the posterior membranous wall that parallels the convex contour of the anterior wall, resulting in a characteristic crescentic lucency (arrow) that mimics a frown.

in appearance on end-inspiration CT in patients with tracheobronchomalacia.6 Several exceptions include patients with relapsing polychondritis, who may demonstrate characteristic wall thickening and calcification that spares the posterior membranous wall of the trachea; patients with lunate tracheal shape (coronal > sagittal dimension), which is frequently associated with tracheomalacia; and patients with extrinsic tracheal compression from adjacent vascular anomalies or thyroid masses, in whom long-standing compression has been complicated by tracheomalacia. The most accurate means for diagnosing malacia on CT is to use an electronic tracing tool to calculate the cross-sectional area of the airway lumen on images at the same anatomic level obtained at inspiration and dynamic expiration.8 Such tools can be found on commercially available PACS (picture archiving and communication system) stations as well as with 3D workstations. Care should be taken to ensure that the same anatomic level is compared between the two sequences by comparing vascular structures and other anatomic landmarks. In the setting of severe malacia in which there is near-complete collapse of the airway lumen during expiration, the diagnosis can be confidently made based on visual analysis of the images (see Fig. 40-5B). Interestingly, about half of patients with acquired tracheomalacia will demonstrate an expiratory “frownlike” configuration, in which the posterior membranous wall is excessively bowed forward and parallels the convex contour of the anterior wall, with less than 6 mm distance between the anterior and posterior walls6 (Fig. 40-23). This appearance, which has been termed the frown sign, is highly suggestive of tracheomalacia and has the potential to aid detection of tracheomalacia when patients inadvertently breathe during routine CT scans.6 Ideally, however, the diagnosis of tracheomalacia should be confirmed and quantified by a dedicated CT trachea study. When interpreting CT scans of patients with tracheomalacia, it  is important to report the severity, distribution, and morphology. These factors have an important impact on treatment decisions,  which are based on a combination of symptoms, severity and distri­ bution of disease, and underlying cause of tracheomalacia.48 Patients with severely symptomatic diffuse tracheomalacia characterized by

CHAPTER 40  Airway excessive mobility and weakening of the posterior membranous wall may benefit from tracheoplasty, a surgical procedure that reinforces the posterior wall of the trachea with a Marlex graft.4

1149

Bronchial Agenesis.  Bronchial agenesis occurs in the setting of

A variety of congenital anomalies may affect the trachea and bronchi, including branching anomalies (displaced and supernumerary bronchi), bronchial agenesis and atresia, congenital stenosis, congenital malacia, congenital tracheobronchomegaly, and congenital diverticula.

congenital absence of a lung (pulmonary agenesis), including its vascular and bronchial supply.21,77 This condition has a high mortality rate because of other associated cardiopulmonary abnormalities. It usually presents in neonates with respiratory symptoms, but it can rarely be an incidental finding in older children and adults. The radiographic appearance simulates pneumonectomy, with opacification of the hemithorax and ipsilateral mediastinal shift. CT confirms unilateral absence of a lung, its bronchi, and vascular supply and may also detect associated congenital heart disease and spinal anomalies.

Displaced Bronchi.  Tracheal bronchus refers to an anomalous bron-

Bronchial Atresia.  Bronchial atresia refers to atresia or obstruction

chial origin from the trachea, carina, or main bronchi, usually located within 2 cm of the carina.10,21,51 Most commonly this occurs as a displaced bronchus arising from the lower trachea and supplying the right upper lobe apical segment (Fig. 40-24). Less commonly the entire right upper lobe bronchus may be displaced, a configuration that is also referred to as a “pig” bronchus. Although usually an asymptomatic  and incidentally detected finding, impaired drainage may result in recurrent infections in some cases. Additionally, following endotracheal intubation, atelectasis may occur in the portion of lung supplied by the aberrant bronchus, owing to inadvertent obstruction by the balloon cuff. Other types of displaced bronchi include proximal or distal displacement of segmental or subsegmental bronchi in the same lobe, bridging bronchus (the displaced bronchus arises from the left main bronchus and crosses the midline to supply the right lower lobe), and esophageal bronchus (the displaced bronchus arises from the esophagus and is directed to a lower lobe).21,51,77

of a segmental or subsegmental bronchus, with normal development of the distal airway.21,77 Typically mucus accumulates within the patent bronchus distal to the atretic segment and the lung hyperinflates beyond the obstruction (Fig. 40-25), most commonly in the apicoposterior segmental bronchus of the left upper lobe. CT findings include a central masslike opacity with a tubular configuration. The distal segmental branches are dilated and contain secretions. The peripheral lung is hyperexpanded, with decreased attenuation and reduced vasculature. This condition is usually asymptomatic and requires no treatment. Surgery may be indicated for cases complicated by recurrent infections.

Congenital Anomalies

Congenital Tracheal Stenosis.  Congenital tracheal stenosis is a rare disorder characterized by complete tracheal rings associated with absent or deficient tracheal membranes.2,7,15 It ranges in length from short- to long-segment stenosis. It typically presents during the first year of life and is associated with a high mortality rate. Symptoms

Supernumerary Bronchi.  Cardiac bronchus refers to a supernumerary bronchus arising from the medial wall of the bronchus intermedius.21,47,51,77 This may be a blind ending or may course 1 to 5 cm caudad toward the heart. It may ventilate a hypoplastic lobule and can be associated with a discrete soft tissue mass due to vascularized bronchial or vestigial parenchymal tissue. Although usually asymptomatic, affected patients may present with infection or hemoptysis. Surgical excision is recommended for symptomatic patients.

FIG 40-24  Tracheal “pig” bronchus. Axial CT image shows the anoma-

FIG 40-25  Bronchial atresia. Coned-down view of the left lung from

lous origin of the entire right upper lobe bronchus (arrow) from the lower trachea.

an axial CT image demonstrates a central tubular structure (arrow) with hyperlucency of the apicoposterior segment of the left upper lobe.

1150

PART II  CT and MR Imaging of the Whole Body

FIG 40-26  Congenital tracheal stenosis. A 3D external rendering of the trachea demonstrates a short-segment stenosis of the lower trachea (arrow).

FIG 40-27  Congenital tracheobronchomegaly. Axial CT image shows massive enlargement of main bronchi and a small bronchial diverticulum (black arrow). Also note mild bronchiectasis (white arrows).

include stridor, wheezing, cyanosis, and recurrent pneumonia. Adult presentation is rare but is more common with short-segment stenosis7 (Fig. 40-26). CT is highly sensitive for evaluating the length and extent of narrowing, as well as for identifying associated cardiopulmonary anomalies, including pulmonary artery sling. On axial CT images there is narrowing of the tracheal lumen, with an O-shaped configuration and absence of airway wall thickening. Virtual bronchoscopy is diagnostic by showing concentric rings extending along the posterior wall of the trachea. Surgical treatment usually involves resection of the involved segment or a slide tracheoplasty.7 Asymptomatic pediatric patients may be followed by CT to noninvasively monitor tracheal growth.63

Congenital Tracheomalacia.  Congenital tracheomalacia refers to

FIG 40-28  Tracheal diverticulum. Axial CT image demonstrates a small

softening of the trachea secondary to weakening or deficiency of cartilage, resulting in increased tracheal compliance with excessive expiratory collapse.13 This condition is most commonly identified in premature infants, likely related to inadequate maturation of the tracheobronchial cartilage. It may also be associated with diseases involving cartilage (e.g., polychondritis, chondromalacia, Hunter’s and Hurler’s syndromes).13 Symptoms include expiratory stridor and a barking cough. Intervention is unnecessary in most cases because tracheal cartilage strengthens and stiffens as the child grows. Patients who do not spontaneously improve may respond to continuous positive airway pressure or surgical intervention (e.g., aortopexy or tracheoplasty).

single or multiple invaginations of the tracheal wall. They commonly arise 4 to 5 cm below vocal cords or 2 to 3 cm above the carina on the right lateral aspect of the trachea (Fig. 40-28). In the absence of symptoms, no treatment is necessary.

Congenital Tracheomegaly.  Congenital tracheomegaly is also referred to as Mounier-Kuhn syndrome and is characterized by dilation of the trachea and main bronchi owing to severe atrophy of longitudinal elastic fibers and thinning of the muscularis mucosa.65,75 Affected patients typically present during the third and fourth decades with recurrent respiratory infections. On imaging studies there is dilation of the tracheobronchial lumen, often associated with a corrugated appearance of the tracheal wall and frequent diverticula (Fig. 40-27). Tracheobronchomalacia is commonly identified at expiratory imaging.

Congenital Tracheal Diverticula.  Congenital tracheal diverticula are less common than the acquired form.66 They are characterized by

diverticulum (arrow) arising from the junction of the right lateral and posterior tracheal walls.

Bronchiectasis Bronchiectasis, defined as irreversible dilation of the bronchi, may occur secondary to a wide variety of congenital and acquired causes25,36,50 (Box 40-5). Infection is the most common cause.50 Affected patients may present with chronic cough, expectoration of purulent sputum, recurrent infections, and hemoptysis. The diagnosis is made on CT by identifying dilation of the bronchi, with or without accompanying bronchial wall thickening. Bronchiectasis may be classified as cylindrical, varicose, or cystic.25,36,50 Diagnostic criteria for cylindrical bronchiectasis include a bronchial diameter greater than its accompanying artery (signet ring sign) (see Fig. 40-12), lack of bronchial tapering (Fig. 40-29), and identification of a bronchus within 1 cm of the costal pleura or abutting the mediastinal pleura.25,36,50 Varicoid bronchiectasis is characterized by a beaded appearance mimicking a “string of pearls” (Fig. 40-30). Cystic bronchiectasis is characterized by thin-walled cystic

CHAPTER 40  Airway

1151

BOX 40-5  Causes of Bronchiectasis Acquired bronchial obstruction (neoplasm, foreign body, broncholith) Asthma Extrinsic bronchial obstruction (lymph node enlargement, neoplasm) Idiopathic Immune deficiency (e.g., IgG deficiency, human immunodeficiency virus infection) Immunologic disorders (allergic bronchopulmonary aspergillosis, collagenvascular disorders, inflammatory bowel disease, graft-versus-host disease following stem cell transplant) Infection* (bacteria, mycobacteria, fungus, virus) Inhalational injury (aspiration, toxic gases) Inherited molecular and cellular defects (cystic fibrosis, α1-antitrypsin deficiency) Inherited bronchial structural deficiencies (bronchial atresia, congenital tracheobronchomegaly, Williams-Campbell syndrome) Primary ciliary dyskinesia (Kartagener’s syndrome) Pulmonary fibrosis (results in traction bronchiectasis) *Most common cause.

FIG 40-30  Bronchiectasis. Axial CT image demonstrates a beaded appearance of dilated bronchi (arrow) in the right upper lobe, consistent with varicoid bronchiectasis.

FIG 40-29  Bronchiectasis. Axial CT image shows lack of tapering of left lower lobe bronchi (arrow), with peripheral dilation.

spaces that connect with the proximal airways, with or without fluid levels (Figs. 40-31 and 40-32). Although volumetric axial images have a high sensitivity for detection of bronchiectasis, the additional review of coronal multiplanar reformation images has been shown to improve detection rate and diagnostic confidence in determining the distribution and type of bronchiectasis.69 CT has limited ability to determine the cause of bronchiectasis,50 but the distribution of disease can help suggest the most likely causes. For example, a bilateral upper lobe distribution is most commonly observed in cystic fibrosis and allergic bronchopulmonary aspergillosis, a unilateral upper lobe distribution is most common in tuberculosis, and a lower lobe distribution is most common in patients with bronchiectasis secondary to childhood viral infections.14 Right middle lobe and lingular distribution of bronchiectasis is often seen in association with the bronchiectatic form of atypical mycobacterial infection45 (Fig. 40-33). Affected patients are typically elderly white women who present with a chronic cough and CT findings of cylindrical bronchiectasis, centrilobular nodules, and tree-in-bud opacities. Impaction of bronchiectatic airways characteristically results in tubular opacities with a Y- or V-shaped configuration, often mimicking

FIG 40-31  Bronchiectasis. Axial CT image demonstrates cystic dilation of bronchi, some of which have small air-fluid levels (arrow).

a “gloved finger” (Fig. 40-34). Impaction of a single bronchus may mimic a parenchymal mass, but careful inspection typically reveals a tubular rather than round configuration (Fig. 40-35). For airways coursing perpendicular to the axial plane, scrolling through a series of axial images or assessment with multiplanar reformations can be helpful to avoid misdiagnosis as a parenchymal mass.

1152

PART II  CT and MR Imaging of the Whole Body

FIG 40-32  Bronchiectasis. Coronal reformation CT image shows multiple foci of cystic bronchiectasis, with a few air-fluid levels (arrow). Also note paraseptal emphysema, most marked at the right apex.

FIG 40-34  Allergic bronchopulmonary aspergillosis. Coronal reformation CT image demonstrates impacted bronchi in the left upper lobe (arrow) producing a “gloved finger” appearance.

FIG 40-33  Bronchiectatic form of chronic atypical mycobacterial infection. Axial CT scan shows extensive bronchiectasis, bronchial wall thickening, and centrilobular nodules, most severe in the middle lobe and lingula.

Bronchiolitis As discussed earlier in this chapter, normal bronchioles are below  the resolution of HRCT. However, in the setting of bronchiolar  disease, both direct and indirect signs of disease may be visualized  on HRCT. The pathologic classification includes two main categories of bronchiolitis: cellular bronchiolitis and constrictive (obliterative) bronchiolitis.1 In the setting of cellular bronchiolitis, one may identify poorly defined centrilobular nodules and/or a combination of linear and nodular branching opacities (tree-in-bud sign). Although these findings are usually well visualized on thin-section CT images, the creation of additional maximum intensity projection (MIP) reconstructions

may be helpful in selected cases to facilitate their recognition1 (Fig. 40-36). The tree-in-bud sign, which reflects bronchiolar wall thickening and distention with intraluminal secretions, is most characteristic of an acute or chronic infectious etiology (Box 40-6).52,70 Ancillary findings may provide a clue to the responsible infectious organism.1 For example, the presence of cavitary lesions is highly suggestive of tuberculosis, but it may also be seen in atypical mycobacterial and fungal infections.1 Additionally, the identification of a chronic tree-in-bud pattern in association with bronchiectasis is suggestive of atypical mycobacterial infection, especially when the airway disease is most severe in the middle lobe and lingula. Common noninfectious etiologies for the tree-in-bud pattern include aspiration bronchiolitis and bronchiolitis associated with chronic bronchial diseases. Neoplasm is a rare cause of this pattern and may result from pulmonary tumor embolism or thrombotic microangiopathy of pulmonary neoplasms.71 The presence of poorly defined ground-glass centrilobular nodules without associated tree-in-bud opacities usually represents peribronchiolar inflammation without intraluminal secretions (see Box 40-6).52 This pattern is commonly seen in the setting of subacute hypersensitivity pneumonitis and respiratory bronchiolitis. Information about a patient’s smoking history may be helpful in making this distinction; the former condition is very uncommon in cigarette smokers, whereas the latter is associated with cigarette smoking. Follicular bronchiolitis may also result in this pattern and should be considered when a patient has a history of collagen vascular disease (e.g., rheumatoid arthritis). In the setting of constrictive bronchiolitis, the findings are more subtle and indirect than those associated with cellular bronchiectasis. On end-inspiratory CT images there may be areas of decreased lung attenuation (mosaic perfusion) due to reflex vasoconstriction induced by regional underventilation, with accentuation on end-expiratory CT images due to air trapping52 (Fig. 40-37). Constrictive bronchiolitis may be due to a variety of causes (Box 40-7). Additionally, cellular bronchiolitis may sometimes progress to constrictive bronchiolitis.

CHAPTER 40  Airway

1153

B

A

FIG 40-35  Bronchial impaction. A, Axial CT image shows a tubular structure (arrow) in the superior segment of the right lower lobe with no visible aerated bronchus. B, Axial CT image following expectoration of a large mucus plug shows underlying bronchiectasis (arrow).

BOX 40-6  Causes of Cellular Bronchiolitis Tree-in-Bud Pattern Acute infectious bronchiolitis (especially viral and mycoplasma) Aspiration bronchiolitis Chronic inflammatory conditions of bronchi (asthma, chronic bronchitis, bronchiectasis) Diffuse panbronchiolitis Neoplasm (rare) Centrilobular Nodules without Tree-in-Bud Pattern Hypersensitivity pneumonitis Respiratory bronchiolitis Follicular bronchiolitis

BOX 40-7  Causes of Constrictive

Bronchiolitis

FIG 40-36  Viral bronchiolitis. MIP CT image of the right lower lung demonstrates diffuse centrilobular nodular and branching opacities with tree-in-bud configuration.

Prior infection (especially viral) Toxic fume inhalation Connective tissue disorders Drugs Chronic transplant rejection Neuroendocrine cell hyperplasia Inflammatory bowel disease

1154

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 40-37  Constrictive bronchiolitis. A, Axial end-inspiratory HRCT scan is normal. B, Axial end-expiratory HRCT image shows multiple lobular foci of air trapping (arrows). Coronal end-inspiratory (C) and endexpiratory (D) images provide global illustration of lungs with better appreciation of the extent of air trapping (arrows).

REFERENCES 1. Abbott GF, Rosado-de-Christenson ML, Rossi SE, et al: Imaging of small airways disease. J Thorac Imaging 24:285–298, 2009. 2. Ali MI, Brunson CD, Mayhew JF: Failed intubation secondary to complete tracheal rings: A case report and literature review. Paediatr Anaesth 15:890–892, 2005. 3. Aquino SL, Shepard JA, Ginns LC, et al: Acquired tracheomalacia: Detection by expiratory CT scan. J Comput Assist Tomogr 25:394–399, 2001. 4. Baroni RH, Ashiku S, Boiselle PM: Dynamic-CT evaluation of the central airways in patients undergoing tracheoplasty for tracheobronchomalacia. AJR Am J Roentgenol 184:1444–1449, 2005. 5. Baroni R, Feller-Kopman D, Nishino M, et al: Tracheobronchomalacia: Comparison between end-expiratory and dynamic-expiratory CT methods for evaluation of central airway collapse. Radiology 2:635–641, 2005. 6. Boiselle PM, Ernst A: Tracheal morphology in patients with tracheomalacia: Prevalence of inspiratory “lunate” and expiratory “frown” shapes. J Thorac Imaging 21:190–196, 2006.

7. Boiselle PM, Ernst A, DeCamp M: CT diagnosis of complete   tracheal rings in an adult patient. J Thorac Imaging 22:169–171, 2007. 8. Boiselle PM, Feller-Kopman D, Ashiku S, et al: Tracheobronchomalacia: Evolving role of multislice helical CT. Radiol Clin N Am 41:627–636, 2003. 9. Boiselle PM, Lee KS, Ernst A: Multidetector CT of the central airways.   J Thorac Imaging 20:186–195, 2005. 10. Boiselle PM, O’Donnell CR, Bankier AA, et al: Tracheal   collapsibility in healthy volunteers during forced expiration:   Assessment with multidetector CT. Radiology 252:255–262, 2009. 11. Boiselle PM, Reynolds KF, Ernst A: Multiplanar and three-dimensional imaging of the central airways with multidetector CT. AJR Am J Roentgenol 179:301–308, 2002. 12. Breatnach E, Abbott GC, Fraser RG: Dimensions of the normal human trachea. AJR Am J Roentgenol 141:903–906, 1983. 13. Carden K, Boiselle PM, Waltz D, et al: Tracheomalacia and tracheobronchomalacia in children and adults: An in-depth review of a common disorder. Chest 127:984–1005, 2005.

CHAPTER 40  Airway 14. Cartier Y, Kavanagh PV, Johkoh T, et al: Bronchiectasis: Accuracy of high-resolution CT in the differentiation of specific diseases. AJR Am J Roentgenol 173:47–52, 1999. 15. Choi SJ, Choi BK, Kim HJ, et al: Lateral decubitus HRCT: A simple technique to replace expiratory CT in children with air trapping. Pediatr Radiol 32:179–182, 2002. 16. Dennie CJ, Coblentz CL: The trachea: Pathologic conditions and trauma. Can Assoc Radiol J 44:157–167, 1993. 17. Faust RA, Stroh B, Rimell F: The near complete tracheal ring deformity. Int J Pediatr Otolaryngol 45:171–176, 1998. 18. Fraser RS, Colman N, Muller NL, et al: The airways and pulmonary ventilation. In Fraser RS, Colman N, Muller NL, et al, editors: Fraser and Pare’s diagnosis of diseases of the chest, ed 4, Philadelphia, 1999, WB Saunders, pp 3–70. 19. Fraser RS, Colman N, Müller NL, et al: Upper airway obstruction. In Fraser RS, Colman N, Müller NL, et al, editors: Fraser and Pare’s diagnosis of diseases of the chest, ed 4, Philadelphia, 1999, WB Saunders, pp 2033–2036. 20. Gamsu G, Webb WR: Computed tomography of the trachea: Normal and abnormal. AJR Am J Roentgenol 139:321–326, 1982. 21. Ghaye B, Szapiro D, Fanchamps JM, et al: Congenital bronchial abnormalities revisited. Radiographics 21:105–119, 2001. 22. Gilkeson RC, Ciancibello LM, Hejal RB, et al: Tracheobronchomalacia: Dynamic airway evaluation with multidetector CT. AJR Am J Roentgenol 176:205–210, 2001. 23. Goo HW, Kim HJ: Detection of air trapping on inspiratory and expiratory phase images obtained by 0.3-second cine CT in the lungs of free-breathing young children. AJR Am J Roentgenol 187:1019–1023, 2006. 24. Greene R, Lechner GL: “Saber-sheath” trachea: A clinical and functional study of marked coronal narrowing of the intrathoracic trachea. Radiology 115:265–268, 1975. 25. Grenier PA, Beigelman-Aubry C, Fetita C, et al: New frontiers in CT imaging of airway disease. Eur Radiol 12:1022–1044, 2002. 26. Grillo HC, Mathisen DJ: Primary tracheal tumors: Treatment and results. Ann Thorac Surg 49:69–77, 1990. 27. Hansell DM: Small airways diseases: Detection and insights with computed tomography. Eur Respir J 17:1294–1313, 2001. 28. Hasegawa I, Boiselle PM, Hatabu H: Bronchial artery visualization on multislice CT in patients with acute PE: Comparison with chronic or recurrent PE. AJR Am J Roentgenol 182:67–72, 2004. 29. Heussel CP, Hafner B, Lill J, et al: Paired inspiratory/expiratory spiral   CT and continuous respiration cine CT in the diagnosis of tracheal instability. Eur Radiol 11:982–989, 2001. 30. Holbert JM, Strollo DC: Imaging of the normal trachea. J Thorac Imaging 10:171–179, 1995. 31. Houston HE, Payne WS, Harrison EG, Jr, et al: Primary cancers of the trachea. Arch Surg 99:132–140, 1969. 32. Ikeda S, Hanawa T, Konishi T, et al: [Diagnosis, incidence, clinicopathology and surgical treatment of acquired tracheobronchomalacia] Japanese. Nihon Kyobu Shikkan Gakkai Zasshi 30:1028–1103, 1992. 33. Johnson TH, Mikita JJ, Wilson RJ, et al: Acquired tracheomalacia. Radiology 109:576–580, 1973. 34. Jokinen K, Palva T, Sutinen S, et al: Acquired tracheobronchomalacia. Ann Clin Res 9:52–57, 1977. 35. Kaminski JM, Langer CJ, Movsas B: The role of radiation therapy and chemotherapy in the management of airway tumors other than small-cell carcinoma and non-small-cell carcinoma. Chest Surg Clin N Am 13:149–167, 2003. 36. Kim JS, Muller NL, Park CS, et al: Cylindrical bronchiectasis: Diagnostic findings on thin-section CT. AJR Am J Roentgenol 168:751–754, 1997. 37. Kwak SH, Lee KS, Chung MJ, et al: Adenoid cystic carcinoma of the airways: Helical CT and histopathologic correlation. AJR Am J Roentgenol 183:277–281, 2004. 38. Kwong JS, Adler BD, Padley SPG, et al: Diagnosis of diseases of the trachea and main bronchi: Chest radiography vs CT. AJR Am J Roentgenol 161:519–522, 1993.

1155

39. Lee KS, Ernst A, Trentham D, et al: Prevalence of functional airway abnormalities in relapsing polychondritis. Radiology 240:565–573, 2006. 40. Lee KS, Sun ME, Ernst A, et al: Comparison of dynamic expiratory CT with bronchoscopy in diagnosing airway malacia. Chest 131:758–764, 2007. 41. Litmanovich D, O’Donnell CR, Bankier AA, et al: Bronchial collapsibility at forced expiration in healthy volunteers: Assessment with multidetector CT. Radiology 257:560–567, 2010. 42. Lomasney L, Bergin CJ, Lomasney J, et al: CT appearance of lunate trachea. JCAT 13:520–522, 1989. 43. Long FR, Williams RS, Adler BH, et al: Comparison of quiet breathing and controlled ventilation in the high-resolution CT assessment of airway disease in infants with cystic fibrosis. Pediatr Radiol 35:1075– 1080, 2005. 44. Lucidarme O, Grenier PA, Coche E, et al: Bronchiectasis: Comparative assessment with thin-section CT and helical CT. Radiology 200:673–679, 1996. 45. Martinez S, McAdams HP, Batchu CS: The many faces of pulmonary nontuberculosis mycobacterial infection. AJR Am J Roentgenol 189:177– 186, 2007. 46. McCarthy MJ, Rosado-de-Christenson ML: Tumors of the trachea.   J Thorac Imaging 10:180–198, 1995. 47. McGuinness G, Naidich DP, Garay SM, et al: Accessory cardiac bronchus: Computed tomographic features and clinical significance. Radiology 189:563–566, 1993. 48. Murgu SD, Colt HG: Recognizing tracheobronchomalacia. J Respir Dis 27:327–335, 2006. 49. Naidich DP, Gruden JF, McGuiness GM, et al: Volumetric (helical/spiral) CT (VCT) of the airways. J Thorac Imaging 12:11–28, 1997. 50. Naidich DP, Webb WR, Grenier PA, et al: Bronchiectasis. In Naidich DP, Webb WR, Grenier PA, et al, editors: Imaging of the airways: Functional and radiologic correlations, Philadelphia, 2005, Lippincott Williams and Wilkins, pp 106–145. 51. Naidich DP, Webb WR, Grenier PA, et al: Introduction to imaging methodology and airway anatomy. In Naidich DP, Webb WR, Grenier PA, et al, editors: Imaging of the airways: Functional and radiologic correlations, Philadelphia, 2005, Lippincott Williams and Wilkins, pp 1–28. 52. Naidich DP, Webb WR, Grenier PA, et al: Small airway diseases. In Naidich DP, Webb WR, Grenier PA, et al, editors: Imaging of the airways: Functional and radiologic correlations, Philadelphia, 2005, Lippincott Williams and Wilkins, pp 146–176. 53. Nishino M, Hatabu H: Volumetric expiratory HRCT imaging with MSCT. J Thorac Imaging 20:176–185, 2005. 54. Okike N, Bernatz PE, Woolner LB: Carcinoid tumors of the lung. Ann Thorac Surg 22:270–277, 1976. 55. Palombini BC, Villanova CA, Araujo E, et al: A pathogenic triad in chronic cough: Asthma, postnasal drip syndrome and gastroesophageal reflux disease. Chest 116:279–284, 1999. 56. Parsons RB, Milestone BN, Adler LP: Radiographic assessment of airway tumors. Chest Surg Clin N Am 13:63–77, 2003. 57. Prince JS, Duhamel DR, Levin DL, et al: Nonneoplastic lesions of the tracheobronchial wall: Radiographic findings with bronchoscopic correlation. Radiographics 22:S215–S230, 2002. 58. Remy-Jardin M, Remy J, Artaud D, et al: Tracheobronchial tree: Assessment with volume rendering—Technical aspects. Radiology 208:393–398, 1998. 59. Remy-Jardin M, Remy J, Artaud D, et al: Volume rendering of the tracheobronchial tree: Clinical evaluation of bronchographic images. Radiology 208:761–770, 1998. 60. Remy-Jardin M, Remy J, Deschildre F, et al: Obstructive lesions of the central airways: Evaluation by using spiral CT with multiplanar and three-dimensional reformations. Eur Radiol 6:807–816, 1996. 61. Ridge CA, O’Donnell CR, Lee EY, et al: Tracheobronchomalacia: Current concepts and controversies. J Thorac Imaging 26:278–289, 2011. 62. Rubin GD: Data explosion: The challenge of multidetector-row CT. Eur J Radiol 36:74–80, 2000.

1156

PART II  CT and MR Imaging of the Whole Body

63. Rutter MJ, Willging JP, Cotton RT: Nonoperative management of complete tracheal rings. Arch Otolaryngol Head Neck Surg 130:450–452, 2004. 64. Salvolini L, Secchi EB, Costarelli L, et al: Clinical applications of 2D and 3D CT imaging of the airways—A review. Eur J Radiol 34:9–25, 2000. 65. Shin MS, Jackson RM, Ho K: Tracheobronchomegaly (Mounier-Kuhn syndrome): CT diagnosis. AJR Am J Roentgenol 150:777–778, 1988. 66. Soto-Hurtado EJ, Penuela-Ruiz L, Rivera-Sanchez I, et al: Tracheal diverticulum: A review of the literature. Lung 184:303–307, 2006. 67. Stern EJ, Graham CM, Webb WR, et al: Normal trachea during forced expiration: Dynamic CT measurements. Radiology 187:27–31, 1993. 68. Sun M, Ernst A, Boiselle PM: MDCT of the central airways: Comparison with bronchoscopy in the evaluation of complications of endotracheal and tracheostomy tubes. J Thorac Imaging 22:136–142, 2007. 69. Sung YM, Yi CA, Yoon YC, et al: Additional coronal images using low-milliamperage multidetector row CT: Effectiveness in the diagnosis of bronchiectasis. J Comput Assist Tomogr 27:490–495, 2003. 70. Trigaux JP, Hermes G, Dubois P, et al: CT of saber-sheath trachea: Correlation with clinical, chest radiographic and functional findings. Acta Radiol 35:247–250, 1994.

71. Verma N, Chung JH, Mohammed TL: Tree-in-bud sign. J Thorac Imaging 27:W27, 2012. 72. Vock P, Spiegel T, Fram EK, et al: CT assessment of the adult intrathoracic cross section of the trachea. JCAT 8:1076–1082, 1984. 73. Webb EM, Elicker BM, Webb WR: Using CT to diagnose nonneoplastic tracheal abnormalities. Appearance of the tracheal wall. AJR Am J Roentgenol 174:1315–1321, 2000. 74. Wilson TA, Rodarte JR, Butler JP: Wave speed and viscous flow limitation. In Macklem PT, Mead J, editors: Handbook of physiology, The respiratory system, vol 3. Mechanics of Breathing, Part 1, Bethesda, MD, 1986, The American Physiological Society, pp 55–61. 75. Woodring JH, Smith Howard R, Rehn SR: Congenital tracheobronchomegaly (Mounier-Kuhn) syndrome: A report of 10 cases and review of the literature. J Thorac Imaging 6:1–10, 1991. 76. Zhang J, Hasegawa I, Feller-Kopman D, et al: Dynamic expiratory volumetric CT imaging of the central airways: Comparison of standarddose and low-dose techniques. Acad Radiol 10:719–724, 2003. 77. Zylak CJ, Eyler WR, Spizarny DL, et al: Developmental lung anomalies in the adult: Radiologic-pathologic correlation. Radiographics 22:S25–S43, 2002.

41  Chest Imaging in the Pediatric Patient Maryam Etesami, Apeksha Chaturvedi, and Prabhakar Rajiah

INTRODUCTION Plain film radiography is the first-line imaging modality in the evaluation of pediatric chest disorders. However, computed tomography (CT) and magnetic resonance imaging (MRI) are increasingly being used and are more valuable for further characterization of pediatric chest abnormalities. Multidetector CT (MDCT) is the most important imaging modality and is more sensitive than radiographs. CT has several advantages including high spatial resolution, good temporal resolution (for cardiovascular disorders), wide field of view (FOV), isotropic resolution, and multiplanar reconstruction (MPR) capabilities, all of which make it an excellent modality for a comprehensive evaluation of pulmonary, airway, mediastinal, cardiovascular, and musculoskeletal abnormalities. Modern scanners can scan in a few seconds because of high z-axis coverage and fast gantry rotation time. This is well suited for children and obviates the need for sedation in the majority of cases. The main disadvantage of CT is ionizing radiation, which is associated with stochastic effects, including cancer. Owing to these radiation risks, CT should only be performed for a definite clinical indication if no alternative nonradiation imaging options exist. The radiation dose should be kept as low as reasonably achievable (ALARA) using several available dose-reduction strategies such as low tube voltage (80 or 100 kVp), low tube current adapted  to body weight/size, automated tube current modulation, iterative reconstruction algorithms, and filters to reduce overranging effect in spiral acquisitions. The scan should be confined to the area of clinical interest. Because of rapid image acquisition, children older than 3 to 4 years are generally compliant for CT scan, particularly if they are aware and prepared. Small children (<15 kg body weight) and those who are noncompliant are sedated for CT, typically with chloral hydrate (50 mg/kg). Ketamine, midazolam, and propofol are alternatives. General anesthesia is rarely required, particularly in noncompliant children. Older children are compliant with breath-hold instructions and are able to hold it in inspiration. Suboptimal inspiration can cause increased density in lungs. In sedated children, controlled ventilation can be used for obtaining images at optimal lung capacity. In noncompliant children, an alternative is to use respiratory gating. Intravenous (IV) contrast is necessary in several situations, particularly cardiovascular disorders. In our institution, 1 to 2 mL/kg of ioversol (Optiray 350) or iopamidol (Isovue 370) is injected, followed by 10 mL of saline bolus. Venous access is challenging in smaller children; contrast is hand injected in this subset of patients. With hand injection, the scan acquisition is initiated after the contrast bolus is administered. Power injector is preferred over hand injection to obtain homogeneous vascular opacification. An empirical delay of 12 to 15 seconds after contrast initiation for children less than 10 kg or 20 to 25 seconds for larger children can be used for most studies. Bolus tracking can be used to

initiate image acquisition, particularly with cardiovascular studies. The scan acquisition is initiated when the attenuation in the structure of interest reaches a suitable threshold. However, bolus tracking is associated with higher radiation dose. CT protocols are tailored to the specific clinical indication. Typically, lung CT is performed using a volumetric spiral acquisition at end inspiration, following administration of IV contrast. Images are reconstructed at soft tissue and bone algorithms, with thin slices (0.6 mm) and overlapping or contiguous slice reconstruction for evaluation of the mediastinum and lung. High-resolution CT (HRCT) protocols for interstitial and airway diseases include non–contrastenhanced inspiratory acquisition (1-mm slices, 10-20 mm gap) and three expiratory images (apex, carina, base) reconstructed with highspatial-frequency bone algorithm. In children who are unable to hold their breath, two slices are taken in right and left decubitus positions, with the dependent lung behaving in apparent expiration, whereas the nondependent lung is in apparent inspiration. Evaluation of airway abnormalities is performed using dynamic volumetric acquisition, both in inspiration and expiration. Vascular studies can be performed without any electrocardiographic (ECG) gating, but cardiac studies will require ECG gating to minimize motion artifact. Prospective ECG-gated sequences are preferred because of significantly lower radiation dose than retrospective ECG-gated images. MPRs in coronal and sagittal planes, maximum intensity projections (MIPs), minimal intensity projections (MiniP), and three-dimensional (3D) volume rendering and surface shading are the postprocessing tools used in the evaluation of chest disorders. MPRs with isotropic resolution enable visualization of structures and abnormalities in different planes, giving a better depiction for presurgical planning. MIPs are useful in evaluation of vasculature and detecting small pulmonary nodules. MiniP is useful in evaluation and quantification of air trapping and airway disorders. 3D volume rendering or surface shading is useful in evaluation of cardiovascular anatomy. Virtual bronchoscopic images are useful in the evaluation of several airway disorders, mimicking a real bronchoscopy. MRI is used as a problem-solving tool in the evaluation of pediatric chest disorders. The advantages of MRI include wide FOV, multiplanar acquisition capabilities, good spatial and temporal resolution, inherent soft tissue contrast augmented with the use of gadolinium-based contrast agents, and the absence of ionizing radiation. MRI is well suited for the evaluation of cardiovascular, mediastinal, and pleural abnormalities. However, MRI is not widely available and acquisition times are long, as a result of which deep sedation/anesthesia is required in young children, who will not be able to lie still for prolonged periods. Evaluation of lung parenchymal and airway abnormalities is limited. MRI cannot be performed in patients with contraindications including claustrophobia. MRI protocol should be optimized for the clinical indication, and the acquisition times should be minimized to ensure

1157

1158

PART II  CT and MR Imaging of the Whole Body

patient compliance. Evaluation of the mediastinum and pleura is performed using T1-, T2-, and proton density–weighted, fat-saturated, steady-state free precession (SSFP), and diffusion sequences. ECG and respiratory gating may be required for obtaining images without respiratory and cardiac motion. Gradient echo T1-weighted images with in-phase and out-of-phase are obtained for characterizing thymic lesions. 3D volumetric contrast-enhanced T1 images (e.g., volumetric interpolated breath-hold examination [VIBE], T1 high-resolution isotropic volume excitation [THRIVE], liver acquisition volume acceleration [LAVA]) are performed to evaluate enhancement. Dynamic multiphasic contrast-enhanced sequences may help in lesion characterization. Cardiovascular studies include SSFP, perfusion, and delayed enhancement sequences. MR angiography (MRA) can be performed after administration of IV contrast (0.1-0.2 mmol/kg) and using a volumetric T1-weighted spoiled gradient echo sequence. Noncontrast MRA can be performed using volumetric SSFP, time of flight, or phasecontrast sequences. In this chapter, we review the CT and MRI appearances of common pediatric thoracic conditions.

A

CONGENITAL, DEVELOPMENTAL, AND GENETIC ABNORMALITIES Congenital Cystic Adenomatoid Malformation (Congenital Pulmonary Airway Malformation) Congenital cystic adenomatoid malformation (CCAM) is a spectrum of malformations consisting of proliferation of distal bronchioles with suppression of alveolar development. The incidence is 1 : 25,000 to 1 : 35,000 of pregnancies.179 The new term congenital pulmonary airway malformation (CPAM) has been recommended, and a new classification is also proposed. There are five types of CPAMs based on pathologic classification (Stocker classification).154 Type 0 has a tracheal or bronchial origin and is really acinar dysgenesis or dysplasia, type 1 has a bronchial or bronchiolar origin (large cystic lesion), type 2 has a bronchiolar origin (small cystic lesion), type 3 has a bronchiolar– alveolar duct origin (adenomatoid type), and type 4 has a distal acinar origin (the “unlined” cystic lesion). The four types of CPAMs seen in imaging are large-cyst CPAM (type 1), small-cyst CPAM (type 2), microcystic (cysts < 5 mm) , solid type (type 3), and a large cyst causing mass effect, lobar expansion, and mediastinal shift that is indistinguishable from a predominantly cystic pleuropulmonary blastoma (type 4).19 CPAM is usually diagnosed prenatally on fetal ultrasound or MRI or early after birth. On pathology, type 1 CPAM reveals a large (>2-cm) cyst and sometimes multiple or multiloculated cysts, occasionally with small cysts that merge with the adjacent lung. At microscopic examination, these cysts are lined by ciliated epithelium, and some contain mucigenic epithelium, highlighting the embryologic origin of the lung as a foregut derivative.93 Fetal ultrasound of a large-cyst CPAM (type 1) demonstrates numerous variable-sized anechoic spaces intermixed with echogenic soft tissue. On fetal MRI, large-cyst CPAMs manifest as hyperintense unilocular or multilocular lesions with discrete walls on T2-weighted images.36 Postnatal radiography shows variable density in the region of the mass depending on the fluid contents of the cysts and possibly mediastinal shift, depending on the size of the CPAM. In the neonatal period a large-cyst CPAM may be seen as a round soft tissue mass that gradually becomes filled with air, since there is delayed clearance of fetal lung fluid from the cysts through the abnormal airway. It may be seen as a solitary well-defined air-filled cyst with thin walls or as multiple cysts of varying size (Fig. 41-1A). Air-fluid levels may be identified. Postnatally, CT is optimal for detailed depiction of the lung

B

C FIG 41-1  Cystic adenomatoid malformation. A, Axial CT shows large cystic spaces filled partially with fluid in the medial aspect of the right upper lobe. B, Axial CT shows multiple small cystic lesions in the left lower lobe in type 2 CPAM. C, Coronal CT in same patient shows multiple small cystic lesions in the left lower lobe.

CHAPTER 41  Chest Imaging in the Pediatric Patient parenchyma and airway, and these lesions are readily detectable as well-defined air-filled spaces.19 On pathology, type 2 CPAM is seen as focal areas of microcystic maldevelopment of the lung parenchyma. On histologic analysis, they consist of closely apposed bronchiole-like structures with variable amounts of intervening lung parenchyma. On fetal ultrasound an echogenic mass with multiple small cysts is seen, whereas on MRI it has a variable T2 appearance that depends on the cystic and solid components. Postnatal radiography may demonstrate a heterogeneous intrapulmonary mass with small air-filled cystic areas.180 Some lesions can be subtle and may not be detected at postnatal radiography. CT may allow better evaluation, demonstrating a heterogeneous mass with small cysts (see Fig. 41-1B and C) and variable surrounding consolidation.19 The differential diagnosis for type 1 and 2 CPAM includes congenital diaphragmatic hernia, pulmonary sequestration, bronchogenic cyst, congenital lobar emphysema, and other bronchopulmonary foregut malformations. A microcystic (predominantly solid-appearing) CPAM (type 3) is homogeneously echogenic on prenatal ultrasound, whereas on T2-weighted MRI it manifests as a homogeneously hyperintense solid mass with normal adjacent lung parenchyma.36 Postnatal CT demonstrates an ill-defined area of increased attenuation.19 CPAMs communicate with the bronchial tree but do not typically have a systemic arterial supply. However, occasionally, systemic arterial supply may be seen, particularly in small-cyst CPAMs, and these lesions are called hybrid lesions since they demonstrate histologic and imaging features of both a CPAM and bronchopulmonary sequestration.29,93 A fastgrowing CPAM may cause mediastinal shift and subsequent development of polyhydramnios and hydrops.2,3 Indicators of a poor prognosis in CPAM include large lesions, bilateral lung involvement, and hydrops.2,9,37,42,167 Quantitative measurements of mass size that help predict development of hydrops and poor outcome include a mass-tothorax ratio of more than 0.56167 or a CPAM volume ratio (i.e., volume of the mass divided by head circumference) greater than 1.6.9 Expectant management is appropriate for the nonhydropic fetus, whereas the survival of fetuses with hydrops improves only with fetal intervention, including aspiration of the cyst, thoracoamniotic shunt, injection of sclerosing agent, or in some cases open fetal surgery.46,117,167 After 32 weeks’ gestation, surgical resection of the mass may be possible while the fetus is maintained on placental circulation, using an ex utero intrapartum treatment (“EXIT”) procedure.9 When fetal distress is not evident, the mortality rate may be as low as 5% or less.37 If not recognized antenatally, CPAMs are usually discovered between the neonatal period and 2 years of age, manifesting as respiratory difficulty or infection. Symptomatic infants who are diagnosed postnatally are treated with surgical resection, which generally consists of lobectomy or segmental resection.9 Surgical intervention for asymptomatic CPAMs remains controversial. Recurrent infection and a questionable small risk of malignancy have been cited as reasons for elective resection.9

Congenital Lobar Emphysema/Overinflation Congenital lobar emphysema (CLE) is characterized by progressive lobar overexpansion, usually associated with compression of the remaining lung. CLE can either be due to a primary intrinsic cartilaginous abnormality with weak/absent bronchial walls or secondary to intrinsic/extrinsic bronchial obstruction (e.g., large pulmonary artery or a bronchogenic cyst). In either case the collapsed airway acts as a one-way valve, resulting in air trapping.19 The term congenital lobar overinflation (CLO) is suggested as a substitute because in this process the alveoli expand but are not destroyed. The most commonly involved lobe is the left upper lobe (42%), followed by the right middle lobe

1159

(35%) and right upper lobe (20%). Involvement of the lower lobes is uncommon (1%).151 Males are more frequently affected than females. It presents with respiratory symptoms in the neonatal period but can also be detected incidentally on fetal ultrasound as a homogeneously hyperechoic mass. In the neonatal period, CLE is seen as a radiopaque mass on  radiographs, since it is filled with fluid. Gradually the fluid will be absorbed and the lesion becomes lucent.121 Subsequently the affected lobe becomes hyperinflated, with hyperlucency demonstrated on  conventional radiography and CT scan. The vascular markings are diminished in the affected lobe (Fig. 41-2). Associated findings are compressive atelectasis in the ipsilateral or even contralateral lobes, mediastinal shift, hemidiaphragm depression, and widening of the rib spaces.125,127,129,151 CT can also be helpful in identifying an abnormal bronchus or focal obstruction as an underlying etiology, as well as evaluating any associated mediastinal or cardiac abnormality.127 Differential diagnosis includes pneumothorax, pulmonary cysts, CPAM, and pulmonary interstitial emphysema (PIE). The pulmonary markings passing through the overinflated lung differentiate it from pneumothorax and pulmonary cysts. In CLE the septae and vascular structures are located at the periphery of expanded air spaces, whereas they are seen in the center in PIE, in which the abnormal air collection occurs in the interstitium. No associated soft tissue is seen in CLE, unlike in CPAM.47 Asymptomatic or mildly symptomatic patients are managed conservatively with low-volume and low-pressure ventilatory support given as needed. In patients with more severe respiratory distress, emergent surgical resection of the affected lobe might be needed.49

Congenital Diaphragmatic Hernia Congenital diaphragmatic hernia (CDH) is an uncommon congenital anomaly of the diaphragm with an incidence of 1 per 2000 to 3000 births.92 Normally the diaphragm starts to develop at approximately 4th week of gestation and is fully formed by the 12th week of gestation. In CDH, one of the components of the diaphragm does not form properly, creating a defect that allows abdominal viscera to enter into the thoracic cavity. This impedes normal development of the lungs, resulting in severe respiratory failure soon after birth because of pulmonary hypoplasia and the presence of pulmonary hypertension.87,101 CDH can be classified based on the anatomic position of the defect into posterolateral, anterior, and central defects. The posterolateral defect (Bochdalek hernia) occurs in 70% to 75% of cases, anterior defects (Morgagni hernia) in 23% to 28%, and central defects in only 2% to 7% of cases. The posterolateral defect most often occurs on the left side (85%) but can occur on the right (13%) or even bilaterally (2%).101,166 CDH is associated with other congenital abnormalities in approximately 40% of cases.131 After pulmonary anomalies, cardiovascular anomalies are the most common group of congenital anomalies present in CDH infants.105 It has been shown to be associated with genetic syndromes such as Beckwith-Wiedemann syndrome and CHARGE syndrome and chromosomal abnormalities including trisomies of 13, 18, 21, and Turner’s syndrome.54,73,131 CDH can be diagnosed prenatally by ultrasonography (60% of cases) or postnatally soon after birth.62 Milder forms of CDH may present with mild respiratory or gastrointestinal symptoms after several months or even years.53 Postnatal chest and abdominal radiographs are usually diagnostic and will show an opacified hemithorax with mass effect and contralateral shift of the mediastinum  with stomach and gas-filled loops of bowel in the chest.160 Crosssectional imaging can be helpful in cases where the diagnosis is still not clear and to evaluate the spectrum of associated anomalies (Fig. 41-3). CT is useful to demonstrate any associated lung masses and

1160

PART II  CT and MR Imaging of the Whole Body

A

B FIG 41-2  Congenital lobar emphysema. A, Axial CT shows hyperinflation of the posterior segment of the right lower lobe in a patient with congenital lobar emphysema. B, Sagittal CT in the same patient shows lucency and hyperinflation of the posterior segment of the right lower lobe.

bronchopulmonary foregut malformations. CT angiography is useful for depiction of the vascular supply of lung lesions. However, oral contrast material should be avoided because it can increase the mediastinal shift and subsequent respiratory compromise. If the patient is stable, MRI can be used in the evaluation of complex lesions as well. In recent years there has been an increase in the use of fetal surgery as an experimental treatment approach. The postnatal management of CDH has evolved in recent decades and now includes gentle ventilation with permissive hypercapnia and delayed surgical repair after stabilization. Prognosis of CDH is related to the degree of pulmonary hypoplasia and need for extracorporeal membrane oxygenation (ECMO) in the neonatal period.57,165 Follow-up chest imaging after surgical repair might be useful to evaluate the functional consequences of the underlying pulmonary hypoplasia and its complications. Pulmonary hypoplasia will manifest as a smaller lung with diminished pulmonary vascularity.

Pulmonary Sequestration

FIG 41-3  Congenital diaphragmatic hernia. Single-shot coronal T2weighted MRI of an approximately 28-week fetus reveals nondilated bowel filling the left hemithorax. The left lung was hypoplastic. Mediastinum was shifted rightward. Patient received surgery soon after birth.

Pulmonary sequestration is defined as a segment of lung tissue that is separate from the tracheobronchial tree and receives its blood supply from a systemic artery.61,153,152 There are two types of pulmonary sequestration: (1) intralobar, which is covered by pleura of the adjacent lung (Fig. 41-4), and (2) extralobar, which has its own separate pleural covering (Fig. 41-5).153 Intralobar sequestrations account for 75% of cases of sequestration. Both intra- and extralobar forms receive systemic arterial supply. In 80% to 90% of sequestration, the arterial supply is from the thoracic aorta (see Fig. 41-4B and C) coursing through the inferior pulmonary ligament, and in the remaining cases it is from the abdominal aorta. Venous drainage of the intralobar form is to the pulmonary vein (see Fig. 41-4C), whereas the extralobar form drains into systemic veins (see Fig. 41-5E). The extralobar form is most commonly diagnosed in the prenatal/neonatal period, and is often

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1161

associated with other congenital systemic anomalies, such as CDH, cardiac abnormalities, pulmonary hypoplasia, or foregut duplication cysts.119 The intralobar form is usually diagnosed later in childhood or adulthood,19 usually with recurrent infections or hemoptysis. Association of intralobar sequestration with other congenital anomalies is rare.108,109 The intralobar form typically affects the posterior basal segment of the left lung in 60% of cases.126 Rarely upper lobe and bilateral lesions have been reported. The radiographic appearance is nonspecific and depends on the degree of aeration and the presence of infection within the sequestered lung. Most pulmonary sequestrations appear as a triangular or oval mass in the medial and basal portion of a lung, more commonly on the left (see Fig. 41-5A, B and C).109 The lack of communication of the sequestered lung with the normal tracheobronchial tree is associated with continued accumulation of secretions accompanied by acute inflammation, infection, and air trapping.59 Repeated infection may establish an abnormal communication with the tracheobronchial tree and cystic bronchiectasis may ensue. In older patients the sequestration will show the effects of recurrent infections. Occasionally, calcifications occur in pulmonary sequestration and are readily detected by CT.164 MDCT angiography and MRA are the imaging methods of choice in the noninvasive assessment of pulmonary sequestration, showing in excellent detail the anomalous systemic vessel.31,43,116 Systemic arterial supply originates from the lower thoracic/upper abdominal aorta and is an essential feature of pulmonary sequestration that differentiates it from other focal lesions with similar characteristics, such as CPAM (see Fig. 41-5D).5,43,116 Venous drainage is also identified, with the intralobar form draining into pulmonary veins (see Fig. 41-4C) and the extralobar form draining into systemic veins (see Fig. 41-5E).

Scimitar Syndrome (Hypogenetic Lung Syndrome) B

C FIG 41-4  Intralobar sequestration. A, Contrast CT of chest in a 12-month-old reveals an airspace opacity in the posteromedial base of right lung (arrow). B, Axial CT also shows an obliquely coursing artery arising off the aorta, feeding this abnormal portion of lung (arrow). C, Another serpiginous contrast-enhancing structure arises from this portion and connects with a right inferior pulmonary vein (arrow). Findings suggest an intralobar sequestration.

Scimitar syndrome (hypogenetic lung syndrome, congenital venolobar syndrome) is characterized by partial anomalous pulmonary venous drainage of the lung associated with hypoplasia of the lung and pulmonary artery. It is caused by failure of integration of the primitive pulmonary venous plexus with the common pulmonary vein, resulting in communication with cardinal or umbilical veins; it is almost always seen on the right. The anomalous vein drains part or the entire right lung into the inferior vena cava (Fig. 41-6), coronary sinus, right atrium, azygos vein, portal vein, or hepatic vein. Cardiac dextroposition and systemic arterial supply to the right lung are also seen. Occasionally not all these features are present in an individual patient. Anomalies of bronchial and pulmonary arterial branching, lobes, and fissures are commonly seen, including horseshoe or crossover lung. Other associations include pulmonary sequestration, aortopulmonary collateral vessels, congenital cardiac defects (e.g., atrial septal defect, hypoplastic left heart), CDH, and musculoskeletal abnormalities. Scimitar syndrome results in a left-to-right shunt, which if large or associated with pulmonary hypertension may require further treatment with surgery or embolization. On radiographs a characteristic scimitar sign is seen, which is caused by the anomalous pulmonary vein curving medially toward the diaphragm and inferior vena cava and progressively widening, giving the appearance of a scimitar (Turkish/Persian sword with a curved blade). CT and MRA are useful in accurate depiction of this anomalous venous drainage (see Fig. 41-6). MRI also helps quantification of the left-to-right shunt using velocity-encoded phase-contrast imaging. Abnormalities of the lung, systemic arterial supply, and other associated abnormalities can also be evaluated using CT and MRI. Scimitar sign can also be seen in another entity called anomalous single pulmonary vein (ASPV) (anomalous unilateral single pulmonary vein, meandering pulmonary vein, scimitar variant, pseudoscimitar syndrome, or

A

B

C

D E FIG 41-5  Extralobar sequestration. A, Axial CT shows a fluid-containing lesion in the left lower lobe adjacent to the diaphragmatic dome, consistent with extralobar pulmonary sequestration (arrow). B, Coronal CT in the same patient shows the fluid-containing lesion in the left lower lobe adjacent to the left dome of the diaphragm (arrow). C, T2-weighted coronal MRI in a 6-month-old boy shows a mass in the left lower lobe adjacent to the diaphragm; this suggests an extralobar sequestration (arrow). D, Coronal time-resolved MRA image shows arterial supply to this left lower lung mass arising directly from the upper abdominal aorta (arrow). E, Venous phase image from the same MRA shows the vertically coursing draining vein (arrow), which ultimately joins the left renal vein (arrowhead).

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1163

B FIG 41-6  Scimitar syndrome. A, Coronal MIP image from a CT scan shows a large right pulmonary vein that is extending into the right lower lobe (arrow) parallel to the right heart border and extending below the diaphragm to drain into the inferior vena cava. The right lung is hypoplastic. B, Coronal MRA image in another patient shows an anomalous pulmonary vein (arrow) that is extending below the diaphragm to drain into the portal vein (not shown here) in this patient with scimitar syndrome.

pulmonary varix), in which the entire lung is drained by a single pulmonary vein that has an anomalous course, but the main difference from scimitar syndrome is that this vein drains normally into the left atrium. It shares several common features with scimitar syndrome, including right sidedness, right lung hypoplasia, cardiac dextroposition, systemic arterial supply, branching abnormalities, and other congenital defects. ASPV is caused by hypoplasia/atresia of one of the pulmonary veins before pulmonary segmentation, as a result of which drainage of the entire lung occurs through the remaining vein. Since ASPV is not a shunt, it does not require further investigation or treatment. Other differential diagnoses for scimitar sign include pulmonary sequestration, pulmonary varix, and pulmonary nodule/mass.

Pulmonary Arteriovenous Malformation Pulmonary arteriovenous malformations (AVMs) are vascular communications between a pulmonary artery and vein (95%) bypassing the capillary bed. Pulmonary AVMs are usually congenital but may also be idiopathic, occur as a result of trauma and infection, or be secondary to hepatopulmonary syndrome and bidirectional cavopulmonary shunting. About 60% of pulmonary AVMs are associated with RenduOsler-Weber syndrome (hereditary hemorrhagic telangiectasia),44 and 15% of hereditary hemorrhagic telangiectasia patients have pulmonary AVM.66 This is an autosomal dominant disorder caused by genes that encode transforming growth factor binding protein,114 with associated abnormalities in liver, skin, brain, and gastrointestinal tract. Owing to this high association, screening for pulmonary AVM is indicated in this subgroup. Physiologic consequences of pulmonary AVM depend on the degree of right-to-left shunt and include hypoxemia, dyspnea, and cyanosis. Hereditary pulmonary AVMs tend to increase in size over time, with clinical manifestations usually presenting by the fifth decade. These patients are also prone to paradoxical embolization complications including stroke and brain abscess.28,146 The typical CT features of pulmonary AVM are: smooth lobulated oval/round mass (Fig. 41-7A) or even serpiginous nodule, less than 1 cm to several centimeters in size, associated with a visible enlarged feeding artery and an enlarged draining vein (see Fig. 41-7B). The feeding artery usually has less than half the diameter of the nodule.95,96,175 Complex AVMs having more than one feeding artery are rare.135,136 The

most common location is the lower lobe.44 CT or MR pulmonary angiography can be used to confirm the vascular nature of the lesion. Multiple AVMs can be confused with metastases if the enlarged feeding vessels are overlooked. MIP reconstructions can also be obtained to scrutinize the pulmonary parenchyma for malformations and depict the size and number of the supplying arteries.175 Pulmonary AVMs can be successfully treated with embolotherapy.21,58,134

Bronchial Atresia Congenital bronchial atresia is a rare anomaly resulting from focal obliteration of a segmental (less commonly a subsegmental) or lobar bronchus that lacks communication with the central airways.16,93,122,178 The bronchi distal to the stenosis are dilated and filled with mucus, forming a bronchocele. The alveoli supplied by these bronchi are ventilated by collateral pathways and show features of mild air trapping, resulting in a region of hyperinflation around the dilated bronchi.33 The upper lobe bronchi are more frequently affected, particularly the apicoposterior segment of the left upper lobe; middle and lower lobes are rarely affected.178 In some cases, bronchial atresia may be acquired postnatally owing to traumatic or postinflammatory insult to the bronchus.83 Bronchial atresia is usually asymptomatic and is found incidentally in chest radiographs; however, dyspnea, pneumonia, and bronchial asthma have been reported.16 In bronchial atresia the airway is occluded rather than narrowed; consequently there is no ball-valve effect as seen in CLE. Hence the lobe or segment does not become nearly as hyperinflated as with CLE. The characteristic chest radiographic findings consist of a bronchocele, seen as rounded branching opacities radiating from the hilum, that may contain an air-fluid level. The distal lung is emphysematous and produces an area of hyperlucency around the affected bronchi. In newborns the affected segment contains fetal lung liquid trapped behind the atresia and may be seen as a fluid-filled mass.16 CT is a sensitive modality for demonstrating the typical features of bronchial atresia. It shows a round opacity at the site of the atresia, medial to the air trapping caused by a mucus plug, which can be clearly depicted by performing expiratory CT15 (Fig. 41-8). CT is also excellent for excluding the presence of a hilar mass and precisely determining the location of lesions. In doubtful cases, multiplanar reformation helps distinguish

1164

PART II  CT and MR Imaging of the Whole Body

A

B FIG 41-7  Pulmonary arteriovenous malformation. A, Axial CT shows an irregular enhancing mass in the periphery of the left lung (arrow); appearances consistent with arteriovenous malformation. B, Coronal MIP of the same patient demonstrates post-embolization appearances of this arteriovenous malformation (arrow).

FIG 41-8  Bronchial atresia. Axial CT shows an area of mucoid plugging in the left lower lobe (white arrow), and distal to this there is an area of hyperlucency (black arrow), consistent with bronchial atresia.

mucoid impaction from nodular lesions. CT also shows the obstruction, associated segmental hypoattenuation, decreased vascularity, and possible proximal bronchocele or broncholith. When the location is basal, spiral CT with contrast can easily exclude a vascular component that could indicate pulmonary sequestration.15 MRI shows the bronchocele as a branching structure radiating from the hilum, with high signal intensity on T1- and T2-weighted images; however, MRI cannot depict regional air trapping.16 The major indication for surgery in these patients is recurrent infection.19

Cystic Fibrosis Cystic fibrosis (CF) is an autosomal recessive disorder caused by mutations in the CFTR gene and the most common lethal genetic disease in Caucasians. CF is a multisystem disorder of exocrine gland function,

with resultant abnormal viscosity of mucus and decreased mucociliary clearance. In the respiratory tract these conditions result in recurrent lung infection and chronic sinusitis with nasal polyps.124,171 Lung involvement is the main cause of morbidity among patients with CF.124 Improvements in therapeutic options have substantially increased the life expectancy of CF patients, with a current median survival of approximately 40 years, which is expected to increase even further.51,64,150 Until 2 decades ago, chest radiograph was the most widely used imaging to monitor morphologic changes in the CF lung. Although chest radiograph correlates significantly with pulmonary function test parameters, it has a low sensitivity to detect early changes in the CF lung.161,162 Currently, HRCT is recognized as the gold standard for assessment of morphologic changes of the airways and lung parenchyma for both early diagnosis and monitoring of CF lung disease38,39,148,163 with higher sensitivity than chest radiograph and lung function testing.17,22,23,41,80 Therefore, multiple disease-scoring systems based on HRCT findings have been proposed.17,25,69,144 An HRCT finding in early stages of lung involvement in CF is a mosaic attenuation pattern resulting from air trapping related to small airway disease.22 Other typical features that may be seen during the course of the disease include bronchiectasis (Fig. 41-9) and peribronchial thickening, mainly involving the upper lobes, and centrilobular nodules and a tree-in-bud pattern secondary to mucous plugs.22,23 Development of bronchiectasis in CF is almost universal, detected in approximately 30% of children with normal findings at lung function testing.22 Bronchiectasis alone or in combination with mucus plugging is given the highest weighting in most scoring systems and is thus commonly considered to be the most significant finding when evaluating CF lung disease.168 Pneumothorax is a rare complication with an incidence of 0.64% per year among CF patients.120 Considering the necessity of lifelong repeated imaging studies, MRI has been proposed as a radiation-free alternative to CT.27 Although spatial resolution is lower as compared to CT, MRI has the advantage of higher soft tissue characterization and visualizing different regional functional aspects of the lung parenchyma (pulmonary hemodynamics, perfusion, and ventilation).51,158,177 Using common MRI sequences, it is possible to visualize bronchiectasis/bronchial wall thickening, mucus plugging, air-fluid levels, consolidation, and segmental/lobar destruction.132

CHAPTER 41  Chest Imaging in the Pediatric Patient

1165

older adolescents, Mycoplasma is the commonest, but C. pneumoniae and fungal infections can be seen. Pneumonia is seen on radiographs, but CT is performed when (1) there are persistent signs, symptoms, and radiographic findings not responding to treatment; (2) to exclude underlying lesions such as sequestration; and (3) to evaluate complications such as pleural effusion, abscess, and empyema. CT is performed when radiographic findings are normal, equivocal, or nonspecific. CT is more sensitive than radiographs in the detection of pneumonia. CT findings of pneumonia include air space disease (Fig. 41-10), consolidation, ground-glass opacities, air bronchograms, air space nodules, and centrilobular or perilobular distribution.

Round Pneumonia.  Round pneumonia is a community-acquired

A

B FIG 41-9  Cystic fibrosis. A, Coronal CT in a patient with cystic fibrosis shows extensive cystic bronchiectatic changes predominantly seen in the upper lobes, along with bronchial thickening. The lungs are hyperinflated. B, Axial CT in the same patient shows extensive bilateral upper lobe–predominant bronchiectatic changes along with bronchial thickening and areas of mucoid plugging.

INFECTIONS Pneumonia Pediatric pneumonia is most commonly caused by Streptococcus pneumoniae. Other organisms include Mycoplasma pneumoniae and Chlamydophila pneumoniae and viruses. In newborns, group B Streptococcus, Listeria monocytogenes, or gram-negative rods are common bacterial agents. In children aged 1 to 3 years, the most common organisms are S. pneumoniae, Staphylococcus aureus, and Haemophilus influenzae. In infants, toddlers, and preschool-aged children, viruses are the most common pathogens, with respiratory syncytial virus (RSV) followed by parainfluenza type 1, 2, 3 and influenza A and B. S. pneumoniae is the most common bacterial pathogen. In school-aged children and young adolescents, Mycoplasma is the most common organism, with others being S. pneumoniae, S. aureus, and Streptococcus pyogenes. In

pneumonia manifesting as a rounded opacity on imaging.138 It is more common in children younger than 12 years.85 It is usually caused by bacterial infection, most commonly S. pneumoniae (90%). In older children, especially those with immunodeficiency, fungal or mycobacterial infections are occasionally encountered.170 Smaller alveoli, closely apposed septa, and underdeveloped interalveolar connections constrain the spread of infection to adjacent alveolar sacs in young children, resulting in a round opacity on imaging. Patients present with clinical signs and symptoms of pneumonia. The characteristic finding is a solitary circumscribed spherical opacity, with predilection for the posterior segment of the lower lobes. The lesion is usually in contact with the pleura or hilum. Air bronchograms are seen in 20% of cases. Satellite lesions are sometimes present, but multiple round pneumonias are rare (1%). Calcification, cavitation, lymphadenopathy, or pleural effusion are not characteristically present.137 In a child presenting with clinical symptoms of pneumonia, a rounded opacity on radiograph or CT scan (Fig. 41-11) is indicative of round pneumonia. A rounded opacity in adults could be neoplastic, but in children it is almost always benign.169 There are other entities mimicking round pneumonia, such as inflammatory masses (e.g., lung abscess, tuberculosis, fungal infection), congenital anomalies, neoplasms, and pseudomasses.35,50,70,142 CT should be considered for further evaluation of a radiographic round opacity in the following settings: clinical features not consistent with a pneumonic process, persistent opacity after appropriate antibiotic treatment, radiographic signs of nonpulmonary origin, or immunocompromised patient with suspicion of complicated infection.137

Pneumonia with Cavitary Necrosis.  Cavitary necrosis (necrotizing pneumonia) is one of the parenchymal complications of severe pneumonia. The spectrum of parenchymal complications of pneumonia also includes lung abscess, pneumatocele, and bronchopleural fistula.48 In adults, necrosis and cavitation are usually associated with S. aureus, Pseudomonas aeruginosa, Klebsiella, and anaerobic infection, but in children S. aureus or S. pneumoniae are most commonly involved. Cavitary necrosis is demonstrated as a dominant area of nonenhancing necrosis surrounded by a thin wall or without a wall within a parenchymal consolidation. Lung abscess is characterized as a solitary cavity with a thick enhancing wall (Fig. 41-12).72 Pneumatocele presents at later stages of healing necrosis. Although children with necrotizing pneumonia typically have a prolonged and intense course of illness, most will ultimately recover without the need for surgical intervention.76

Parapneumonic

Effusion/Empyema.  Parapneumonic effusion (PPE) or empyema complicates pneumonia in 28% to 53% of pediatric cases.24 Although the overall incidence of bacterial pneumonia has been declining in children following the introduction of pneumococcal conjugate vaccine, the incidence of PPE and empyema has

1166

PART II  CT and MR Imaging of the Whole Body

A

B FIG 41-10  Pneumonia. A, Axial CT shows an area of consolidation with air bronchogram in the left lower lobe (arrow) consistent with pneumonia. B, Coronal CT in the same patient shows the consolidation with air bronchogram in the left lower lobe.

A

B FIG 41-11  Round pneumonia. A, Axial CT shows a well-defined round area of consolidation in the periphery of the right lower lobe (arrow) in a 7-year-old boy. B, Coronal CT in the same patient shows the welldefined round pneumonia in the right lower lobe (arrow).

increased.78,79,65,103 S. pneumoniae remains the leading pathogen, but the incidence of empyema caused by other pathogens such as Staphylococcus has increased.65 Despite the low mortality and favorable outcome in most cases as compared to adults,12 complicated PPE causes significant morbidity in children. Four progressively complex stages characterize parapneumonic pleural disease: (1) precollection stage (pleuritic), (2) exudative stage with free-flowing pleural fluid and low white count, (3) fibrinopurulent stage or complicated PPE with deposition of fibrin and purulent material, (4) organizational stage with thick pleural peel.67 Chest radiograph is often the initial imaging modality, but in many cases it cannot differentiate pulmonary parenchymal disease from pleural disease. Ultrasound and CT can reliably differentiate these two entities.78 CT scan is reserved for special cases such as where characterization of the extent of the parenchymal disease and/or the presence and location of lung abscesses will affect presurgical planning.78 Neither ultrasound nor CT can reliably differentiate between different stages of the PPE (Fig. 41-13); a pleural tap is required to identify complicated PPE or empyema. On CT, PPE or empyema is seen as thickened and enhancing pleural layers surrounding a pleural effusion, resulting in the split pleura sign (Fig. 41-14). Pleural thickening and enhancement are seen in 86% to 100% of cases of empyema, compared with 60% of parapneumonic effusions. Although pleural thickness appears to increase on CT with advancing disease, it indicates neither a poor response to therapy nor the need for surgical intervention. Empyema is also associated with thickened extrapleural soft tissue and increased attenuation of extrapleural fat (an increase of >50 HU compared with fat posterior to the spine). These changes may persist for some time after resolution of the infection. Nodal enlargement of less than 2 cm is common in empyema and tends to be seen in ipsilateral subcarinal and paratracheal nodes, more often on the right. The amount of nodal involvement is not associated with more advanced disease. CT demonstrates separate loculations of infected pleural fluid in about 20% of cases, which helps in deciding the optimal placement of a chest drainage catheter, particularly in patients who are not responding to conventional therapy or in whom drainage has ceased. Septations within the same fluid  collection may be demonstrated on CT, signified by the presence  of trapped pockets of air within a fluid collection. Up to 30% of patients with empyema fail conservative medical therapy (antibiotics

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

C

B

D

1167

FIG 41-12  Pneumonia with lung abscess formation. A, Axial CT shows a large irregular gas-containing cavitary lesion in the left lower lobe (arrow) consistent with a lung abscess. B, Coronal CT shows a large thickwalled abscess cavity in the left lower lobe (arrow) with surrounding areas of consolidation. C, Axial CT in another patient shows a large fluid-containing lesion in the right lung (arrow) with thick walls and surrounding consolidation, consistent with lung abscess. D, Coronal CT shows a large thick-walled cavitary lesion with gas bubbles (arrow) consistent with a lung abscess.

and intercostal drainage) and require surgery. CT is useful in surgical planning of these patients. Distinguishing empyema from a pulmonary abscess abutting the pleural surface is important because empyema is treated with drainage, whereas an abscess is generally not drained to avoid the risk of pleural spread. On contrast CT, empyema is lenticular, thin walled with compression of adjacent lung structures, and increased attenuation of extrapleural fat. Abscesses are spherical thick-walled lesions associated with abrupt vessel cutoff and the presence of bronchi at the interface of the abscess and normal lung. MRI is not routinely used for assessment of patients with pleural infection owing to its poor spatial resolution and higher likelihood of motion artifact compared with CT. However, MRI is a sensitive technique for demonstrating septations within a pleural collection and  may be used if contrast administration is contraindicated or if there is concern about radiation exposure. Rarely malignancies may arise in the wall of chronic empyema cavities and are detected by MRI, which allows differentiation of fibrous from neoplastic tissue. Malignancies associated with chronic empyema include non-Hodgkin’s lymphoma, squamous cell carcinoma, mesothelioma, and sarcoma.

TRAUMA Blunt Trauma Rib fractures can be seen following trauma (Fig. 41-15). Pulmonary contusion is the most common lung injury from blunt chest trauma, occurring at the time of injury. On CT, patchy airspace opacities or consolidation with ill-defined margins is seen in a nonsegmental distribution with subpleural clearing (Fig. 41-16A). Contusion appears immediately, resolution begins in 24 to 48 hours, and complete clearing takes 3 to 10 days.81 Pulmonary laceration is a tear in the lung, with a resulting cavity. The cavity results from normal lung recoiling back from the laceration. On CT, single or multiple, unilocular or multilocular, round or oval cavities may be seen. The cavity may be filled with air (pneumatocele) (see Fig. 41-16B), blood (hematocele), or both. Lacerations heal slowly and take months.81 There are four types of lacerations based on mechanism, CT appearances, and location of rib fractures. Type I is caused by compressive forces causing lacerations in deep portions of lung, type 2 by compression shear injury with sudden shift of lower lobes across the spine, type 3 by a rib fracture

1168

PART II  CT and MR Imaging of the Whole Body

FIG 41-15  Traumatic rib injury. Axial CT shows a nondisplaced fracture of the anterior aspect of the left first rib (arrow).

A

penetrating the periphery of the lung, and type 4 is seen at the site of a preexisting pleuropulmonary adhesion.81 Pneumothorax can be caused by alveolar rupture due to sudden increase of intrathoracic pressure or blunt crushing force or deceleration.

Nonaccidental Trauma

B FIG 41-13  Parapneumonic effusion. A, Axial CT shows an area of consolidation (black arrow) in the left lower lobe, with associated parapneumonic effusion (white arrow). B, Coronal CT in a 21-month-old boy shows a large consolidation in the right lung (black arrow) with associated parapneumonic effusion (white arrow).

Nonaccidental trauma (battered baby syndrome, shaken infant syndrome) in infants and young children is due to abuse. The classic pattern of injuries is seen when the child is violently shaken while being held around the chest. In children with suspected nonaccidental injury (NAI) a skeletal survey is typically performed to evaluate all the injuries. However, sometimes the radiographic findings may be subtle or negative, warranting additional evaluation with CT. Occasionally these injuries may also be incidentally seen on CT and MRI. In such scenarios, maintaining a high level of suspicion, particularly when the injuries could not be explained by the reported mechanism, is essential for making a prompt diagnosis and alerting authorities. Owing to plasticity of ribs, fractures are generally uncommon in infants and children younger than 3 years, even in accidents and after cardiopulmonary resuscitation. Rib fractures, particularly those of the posterior (Fig. 41-17) and costochondral junctions, are highly specific for NAI. Rib fractures are the most common skeletal injury in NAI and are seen in 80% of these children181; in 29%, rib fracture may be the only skeletal abnormality. Although rib fractures are seen on radiographs, CT is performed when the radiograph is negative/subtle because of acute nondisplaced, incomplete, oblique fractures, which may be obscured by superimposed structures. CT has been shown to be more sensitive than radiographs in such situations.181 Fractures of the sternum and scapula are also suggestive of NAI.

NEOPLASMS Primary lung tumors are rare in children, with most lung masses likely benign developmental or reactive, which are 10 times more common.45 Of the neoplasms, metastasis is more common than primary lung tumors, and malignancies are three times as common as benign lesions. Inflammatory myofibroblastic tumor is the most common benign tumor, and pleuropulmonary blastoma and carcinoids are the most common primary malignancies.45

Pulmonary Blastoma FIG 41-14  Empyema. Axial CT shows a thick-walled left pleural fluid collection showing the so-called split pleura sign of thickened and enhancing pleural layers, consistent with empyema (arrow).

Pulmonary blastoma is a rare malignancy that accounts for 0.25% to 0.5% of primary lung tumors.60,89,94,102 The tumor derives its name from its histologic resemblance to fetal lung tissue. Pulmonary blastomas, however, are thought to arise from primitive pluripotential stem cells and may represent a variant of carcinosarcoma.176 Although the age

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1169

B FIG 41-16  Contusion and pneumatocele. A, Axial CT in a 12-year-old boy with a history of trauma shows an area of ground-glass opacity in the left lower lobe (arrow), consistent with pulmonary contusion. Note also a small left pneumothorax. B, Axial CT in the same boy shows that there is an air-containing focus (arrow) within the ground-glass opacity, consistent with a pneumatocele.

range at presentation is wide (0-80 years), these tumors typically  occur in adults, with a peak incidence between the age of 40 and  60 years.60,89,94,102 Patients are often symptomatic at presentation; cough, hemoptysis, and chest pain are frequent manifestations.89,102 Pediatric patients typically present with fever and respiratory distress.20 The behavior of pulmonary blastoma is aggressive, and the outcome is poor owing to frequent relapses and metastases.18,20,34 Radiologically, pulmonary blastomas typically manifest as large (2.5-26 cm) well-marginated masses located peripherally in the lung20,104,173 (Fig. 41-18). Multiple masses, cavitation, and calcification are rare.173 Local invasion of the mediastinum and pleura occurs in 8% and 25% of cases, respectively.60 Metastases to hilar and mediastinal lymph nodes are present in 30% of resected cases.60 Extrathoracic metastases are common and have a distribution similar to that of lung cancer.60,102

Pulmonary Inflammatory Pseudotumor Inflammatory pseudotumor (IPT) or inflammatory myofibroblastic tumor (plasma cell granuloma) is characterized by a proliferation of fibroblasts and myofibroblasts mixed with varying numbers of plasma cells, lymphocytes, macrophages, and other inflammatory cells.10,184 Historically, IPT has been believed to be a nonneoplastic reactive lesion simulating a neoplasm; however, over the past 15 years or so, almost all IPTs have been regarded as true neoplasms.149 Currently they are generally considered as a soft tissue mesenchymal tumors with indeterminate or low malignant potential.32 In fact the current World Health Organization classification of lung tumors adopted the name inflammatory myofibroblastic tumor (IMT) as the preferred term for these lesions. Other synonyms used for IPT are plasma cell granuloma, plasma cell pseudotumor, inflammatory myofibrohistiocytic proliferation, xanthofibroma, and omental mesenteric myxoid hamartoma. IPT was first described in the lungs, but several reports have described them in almost all other organs. Based on major histopathologic features, IPT can be divided to three categories: organizing pneumonia type, fibrous histiocytoma type, and lymphoplasmacytic type.110 Pulmonary IPT is a rare disease.30 It affects both sexes at any ages, with a predominance in children and young adults.68 Patients are afflicted with nonspecific symptoms such as cough, dyspnea, hemoptysis, chest pain, fever, and fatigue. Weight loss and anorexia are rare. However, most patients are asymptomatic and the tumor is discovered

incidentally on a chest x-ray performed for another reason. Radiologic aspects are variable and nonspecific. Most patients have a parenchymal pulmonary mass or nodule (85%). IPT can also present as endobronchial (11%) or endotracheal (4%) polypoid lesions.84,86 They are mostly solitary, rarely multiple (5%), 1 to 6 cm in diameter, well circumscribed, smooth, or bumpy (Fig. 41-19). Parenchymal lesions are predominantly located on the periphery, with a predilection for the lower lobes.4 CT shows the presence of a heterogeneous nodule or mass with variable contrast enhancement. CT specifies the solid or cystic nature of the tumor, its vascular behavior, and assesses the locoregional extension. Calcifications (15%) and cavitations are uncommon.82,110,115 MRI shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging. Various contrast enhancement patterns have been described.4,159 Pleural effusion is seen in less than 10% and atelectasis in 8% of cases. IPT is usually a slow-growing tumor; however, case reports of fast-growing cases are also present.182 Sometimes the tumor can extend toward the hilum, mediastinum, pleura, or diaphragm. Because of the diversity of clinical and radiologic manifestations of IPT, diagnosis is based on histologic examination of the surgically resected specimen. Currently the principal management of IPT—for diagnostic and therapeutic reasons—is a complete resection. Incomplete resection increases the risk of recurrence. Corticosteroids are generally not useful in adults, but good results have been reported in children in cases of unresectable tumors or hilar and mediastinal invasion. Chemotherapy is useful in cases of multifocal invasive lesions or in cases of local recurrence.68

Metastasis Metastasis accounts for 80% of lung tumors in children.45 Tumors that spread to lung are osteosarcoma, Wilms’ tumor, neuroblastoma, Ewing’s sarcoma, rhabdomyosarcoma, other soft tissue sarcomas, hepatoblastoma, thyroid carcinoma, and adrenocortical carcinoma. On CT, pulmonary metastasis is seen as multiple soft tissue nodules of variable sizes, more on the periphery of the lung, usually in lower lobes in those with hematogenous spread (Fig. 41-20). Solitary pulmonary nodule, miliary nodules, cavitation, calcification, and ground-glass opacity are atypical findings. A reticular or miliary pattern is seen in lymphangitic spread. Cavitation and pneumothorax is associated with osteosarcoma, Wilms’ tumor, and Hodgkin’s lymphoma.45

1170

PART II  CT and MR Imaging of the Whole Body

A A

B

B FIG 41-17  Nonaccidental injury. A, Axial CT shows a nondisplaced fracture of the left posterior sixth rib (arrow). B, Coronal CT in another patient shows multiple nondisplaced posterior rib fractures (arrows).

AIRWAY ABNORMALITIES Tracheal Bronchus Tracheal bronchus (so-called pig bronchus, or bronchus suis) is a congenital branching anomaly in which the upper lobe bronchus arises directly from the trachea and above the level of the carina16,63,123 (Fig. 41-21). The incidence of tracheal bronchus ranges between 0.1% and 5% in the general population.91 It is most common on the right side, with the right upper lobe bronchus originating from the right tracheal wall. Tracheal bronchus is usually diagnosed incidentally during bronchoscopy or CT performed for evaluation of various causes of respiratory symptoms. However, tracheal bronchus may also be the cause of recurrent atelectasis or pneumonia in the upper lobe in symptomatic pediatric patients.16 Treatment is not required in asymptomatic

C FIG 41-18  Pulmonary blastoma. A, Axial CT in an 18-month-old boy shows a large tumor almost completely replacing the left lung, which shows heterogeneous contrast enhancement (arrow) and displacement of the mediastinum to the right. B, Axial CT in the same patient at a higher level shows the heterogeneously enhancing mass in the left lung (arrow). C, PET-CT image of the same patient showing intense FDG uptake within the mass (arrow), which was shown to be pulmonary blastoma.

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1171

B FIG 41-19  Pulmonary inflammatory pseudotumor. A, Axial CT shows a heterogeneously enhancing peripheral mass in the right lower lobe (arrow). B, Coronal CT shows a heterogeneously enhancing peripheral mass in the right lower lobe (arrow), which was shown to be a pseudotumor.

A

B FIG 41-20  Metastasis. A, Axial CT shows multiple solid metastatic lesions along with a large cystic component in the left lower lobe in a patient with sarcoma of the soft tissue. B, Coronal CT in the same patient shows multiple solid metastatic lesions in both lungs.

patients, but in symptomatic patients, surgical resection of the aberrant bronchus and its lobe is currently recommended.99

tracheal stenosis is caused by extrinsic compression by vascular ring/ sling or a dilated esophagus.

Tracheal Stenosis

Tracheobronchomalacia

Tracheal stenosis in children can be congenital, caused by replacement of the posterior tracheal membrane with cartilages. This may affect over 50% of the tracheal lumen and may involve the bronchi as well (Fig. 41-22). On CT, an O-shaped trachea is seen. Associated anomalies may be seen in the cardiovascular or musculoskeletal systems. Acquired

Tracheobronchomalacia refers to expiratory collapse of the tracheobronchial tree. It can be caused by congenital softening of the cartilaginous rings in the anterior two thirds of the tracheobronchial wall. It is also associated with CF, Mounier-Kuhn syndrome, Marfan’s syndrome, Ehlers-Danlos syndrome, and Cornelia de Lange syndrome. It

1172

PART II  CT and MR Imaging of the Whole Body

FIG 41-21  Tracheal bronchus. Coronal CT shows the tracheal bronchus (bronchus suis) (arrow) originating from the trachea and supplying the right upper lobe.

can be caused by extrinsic compression by vascular rings/slings. It may be asymptomatic or associated with dyspnea, cough, and recurrent respiratory infections. CT with inspiratory and expiratory acquisitions is required for diagnosis. On the expiratory image, over 50% collapse of the airway compared to inspiration is diagnostic of tracheomalacia. A cutoff of 70% is more specific in children (Fig. 41-23). Dynamic volumetric acquisition is more sensitive in direct visualization of the airway collapse. On inspiratory images, a dilated trachea with posterior bowing may be seen.

Bronchiectasis Bronchiectasis is characterized by irreversible bronchial dilation and airflow obstruction (Fig. 41-24) and can be focal or diffuse. Congenital causes of bronchiectasis include CF, ciliary dyskinesia (see Fig. 41-24C), Young’s syndrome, Marfan’s syndrome, agammaglobulinemia, immunoglobulin deficiencies, Mounier-Kuhn syndrome, and WilliamsCampbell syndrome. Acquired causes include infection (see Fig. 41-24B), chronic aspiration, and airway obstruction. Morphologically, bronchiectasis can be cylindrical (straight, uniform), varicose (bulbous, alternating dilation and constriction), or cystic (saccular; ballooned appearance with neovascularization and ulceration). On CT, bronchiectasis is seen as dilation of the bronchi (>1.5 times the corresponding vessel) resulting in a signet ring sign, grapelike clusters of dilated large airways, and airways seen up to the periphery of the lung (see Fig. 41-24A). Air trapping is seen with airway obstruction.

Bronchial Foreign Body Foreign body aspiration is a frequent cause of acute respiratory distress in children55,99 and is the fifth most common cause of death in infants.99 It is more common between 6 months and 3 years of age.98,100 Common locations of foreign body lodgment are the right main bronchus (52%), left main bronchus (18%), and trachea (13%).55 The right main bronchus is more commonly affected because it is more directly in line with the trachea in an upright position, and the left main bronchus  is smaller in caliber in comparison. Affected children frequently  present with coughing, wheezing, and stridor after an acute choking episode.55,133,139,145 However, children with foreign body aspiration may be asymptomatic; therefore thorough investigation is necessary in

infants and children with a reliable history and high clinical suspicion of foreign body aspiration. Approximately 10% of aspirated foreign bodies within the trachea and bronchi are radiopaque and can be detected on chest radiographs.145 The sensitivity and specificity of correctly diagnosing foreign body aspiration with chest radiographs in pediatric patients are 68% to 74% and 45% to 67%, respectively.147,157 Therefore, pediatric patients who present with typical clinical symptoms and high clinical suspicion of foreign body aspiration should be further evaluated with either CT or conventional bronchoscopy after negative chest radiographs.11,90,98,100 Although CT is currently the most sensitive imaging modality for detecting foreign body aspiration (accuracy close to 100% in pediatric patients), judicious use of CT is advised and only recommended for cases where chest radiographic findings are negative or equivocal.98 On CT, foreign body aspiration can be diagnosed or suspected when there is an endoluminal mass with various attenuation within the airways.98,100 Secondary CT imaging findings that can further support the diagnosis include postobstructive air trapping, atelectasis, and consolidation.98,100

MEDIASTINAL LESIONS Several mediastinal lesions occur in children. Anterior mediastinal masses include thymic hyperplasia, thymolipoma, lymphoma, germ cell tumor, and lymphangioma. Middle mediastinal masses include foregut duplication cysts, lymph nodal lesions such as granulomas, fungus, tuberculosis, metastasis, and lymphoma/leukemia. Posterior mediastinal masses include neurogenic tumors (95%), which involve the ganglion cells (neuroblastoma, ganglioneuroblastoma, ganglioneuroma) or nerve sheath (neurofibroma, schwannoma), and extramedullary hematopoiesis.

Anterior Mediastinal Lesions Thymic hyperplasia is characterized by an increased volume (>50%) of the thymus. Maximal and mean normal thymic thickness on the inspiratory phase of CT are 1.8 and 1.1 cm, respectively, between ages 6 and 19 years, and 1.3 and 0.5 to 0.8 cm, respectively, in adults older than 20 years.1 Thymic hyperplasia is seen in hyperthyroidism, acromegaly, Addison’s disease, stress, chemotherapy, radiotherapy, and myasthenia gravis. Distinguishing hyperplasia from a neoplasm is critical, particularly in patients who have received chemotherapy and radiation. Hyperplastic thymus maintains normal shape and has homogeneous attenuation (Fig. 41-25A) and signal. Asymmetry of the lobes, focal contour abnormalities, round soft tissue mass larger than 7 cm, and heterogeneity on MRI or CT suggest neoplasm. Signal drop in opposedphase images as compared to in-phase images (see Fig. 41-25B and C) indicates normal thymus or thymic hyperplasia due to the presence of interspersed fat, whereas neoplasms do not show this signal drop. Chemical shift ratio (CSR) is calculated as (thymus SI OP/paraspinal muscle SI OP)/(thymus SI IP/paraspinal muscle IP). Thymic hyperplasia has a CSR of 0.5 to 0.6, whereas thymic neoplasms have CSRs of 0.9 to 1.0.74 Patients with thymic hyperplasia have homogeneous fluorodeoxyglucose (FDG) uptake, whereas patients with thymoma have more focal FDG uptake, which is also higher than in hyperplasia (see Fig. 41-25D). Progressive decline in size of the thymus over 3 to 6 months indicates hyperplasia. Absence of active disease is another indication. Thymic cyst accounts for 3% of anterior mediastinal masses.156 It can be congenital, caused by a remnant of the thymopharyngeal duct,183 or secondary to inflammation/neoplasm.14,75 On CT, thymic cysts are less than 6 cm, homogeneous with water attenuation, and thin walled without contrast enhancement. Proteinaceous content may give

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

C

B FIG 41-22  Tracheal stenosis. A, Axial CT shows severe narrowing of the trachea at the level of thoracic inlet (arrow) approximately 7 cm from the glottis. B, Coronal MiniP image shows severe narrowing (arrow) at the level of thoracic inlet. C, A 3D volume-rendered image shows severe narrowing of the trachea at the level of thoracic inlet (arrow).

A

B FIG 41-23  Tracheobronchomalacia. A, Axial image at the level of thoracic inlet shows normal caliber of the tracheal lumen. B, Axial image at the same level obtained at expiration shows severe narrowing of the lumen—90% greater than that in inspiration—consistent with tracheomalacia.

1173

1174

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 41-24  Bronchiectasis. A, Axial CT shows multiple dilated cylindrical bronchiectasis in a 7-year-old patient. B, Axial CT in another patient shows a focal area of bronchiectasis in the medial right lower lobe (arrow), which is a sequel of previous infection. C, Axial CT shows extensive cystic bronchiectasis in the lingula (arrow), with associated consolidation and also dextrocardia, which is seen with immotile cilia syndrome.

a hyperdense appearance.7 On MRI, it has low signal in T1-weighted and high signal in T2-weighted images. High T1 may be due to hemorrhage and protein. No contrast enhancement is seen. Thymolipoma is a fat-containing tumor in the thymus (Fig. 41-26). It does not have a capsule. It has soft tissue and fat elements, with heterogeneous enhancement. There is no calcification and mass effect. Lipoma may also originate from the mediastinum or extend from the neck into the mediastinum. This contains fat attenuation in CT and signal characteristics of fat in all sequences (Fig. 41-27). Thyroid lesions may also extend into the superior mediastinum (Fig. 41-28). Lymphoma is the most common mediastinal mass in children, with Hodgkin’s lymphoma more common than non-Hodgkin’s lymphoma. It is more common in the second decade. Some 70% of children with Hodgkin’s disease have thoracic disease, with the most common pattern being involvement of the mediastinal lymph nodes (95%) followed by involvement of thymus (90%), hilar nodes (25%), and  lung (10%). On CT and MRI, discrete enlarged lymph nodes or a

conglomerate mass may be seen, with minimal contrast enhancement and compression of adjacent structures such as airways and vasculature (Fig. 41-29). Non-Hodgkin’s disease occurs in the first and second decades. It most commonly involves the nodes and is not contiguous. It often spares the thymus. Germ cell tumor is the second most common cause of anterior mediastinal mass in children and the common cause of a fat-containing anterior mediastinal lesion. The majority (90%) of these lesions are benign and originate from the thymus. Teratoma has well-defined margins and is cystic with variable amounts of water, calcium, fat, and soft tissue. Size is variable and not an indicator of malignancy. Mature teratomas can also be large but are usually benign (Fig. 41-30). Malignant teratomas account for 10% of teratomas and have more soft tissue component, nodular component, invasive margins, invasion of adjacent structures, and metastatic lesions. Nonteratomatous germ cell tumors are choriocarcinoma, yolk sac tumor, and embryonal cell cancer.

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

B

C

D FIG 41-25  Thymic hyperplasia. A, Axial CT shows an enlarged thymus that has sharp and smooth margins (arrow). B, In-phase T1-weighted MRI shows intermediate to high signal in the thymus (arrow). C, Opposedphase MRI shows dropout of signal in the thymus, indicating hyperplasia (arrow). D, FDG PET shows homogeneous FDG uptake in the hyperplastic thymus (arrow).

A

B FIG 41-26  Thymolipoma. A, Axial CT shows a large fat-containing lesion in the anterior mediastinum (arrow). B, Coronal CT shows a huge fat-containing mass in the anterior mediastinum, proven on biopsy to be a thymolipoma (arrow).

1175

1176

PART II  CT and MR Imaging of the Whole Body

Lymphatic malformation (previously called lymphangioma) is a benign mediastinal lesion; 75% of these occur in the neck, 20% in the axillary region, and 5% in mediastinum,130 with a majority of them extending from the neck usually into the superior or anterior mediastinum. It is more common in males younger than age 2. On CT it is seen as a lobular multicystic mass with 1- to 2-cm cysts that infiltrates adjacent structures (Figs. 41-31). A solid appearance may be seen owing to protein/hemorrhage. Enhancement may be seen. On MRI, it has variable T1 signal, usually low to intermediate, but can contain focal areas of high signal intensity similar to fat. High signal is seen on T2-weighted images.143 Hemangioma is rare in the mediastinum (<0.5%)112; it occurs most commonly in the anterior mediastinum (68%) and is usually the cavernous type. On CT it is seen as a heterogeneous mass (see Fig. 41-32A)

with intense central and peripheral rimlike enhancement. Phleboliths may be seen.112 On MRI it has heterogeneous signal on T1-weighted images, with stromal fat seen as linear areas of high signal and high signal in T2-weighted images77,140,141 (see Fig. 41-32B and C).

Middle Mediastinal Lesions Bronchogenic cyst is a thin-walled mucous or serous cyst that is lined by ciliated respiratory epithelium, cartilage, smooth muscle, fibrous tissue, and mucous glands. It is thought to originate from abnormal budding of embryonic ventral foregut. The majority of bronchogenic cysts are seen in the middle mediastinum, either in the subcarinal or right paratracheal region, but 15% can be seen in the lung, pleura, pericardium, or diaphragm.40 On CT, bronchogenic cyst is well marginated, round or spherical, homogeneous with water attenuation, and shows no contrast enhancement (Fig. 41-33A). High attenuation due to protein or hemorrhage may be seen. Calcified wall/calcium in fluid may be seen. Contrast enhancement is not typically present but may be seen with infection, which also has an air-fluid level. Distal collapse/ overinflation of the lung may be seen. On MRI, bronchogenic cyst has low to intermediate T1 and high T2 signal, usually heterogeneous in T1 and homogeneous on T2 (see Fig. 41-33B and C). Intermediate signal may be seen in cardiac-gated T1-weighted sequences owing to long repetition time. Hemorrhage or protein also has high T1 signal.13,118,155 No contrast enhancement is seen typically, but rim enhancement may be seen with infection.97,106,111

Posterior Mediastinal Lesions

FIG 41-27  Lipoma. Axial CT in a 13-year-old patient shows a fatcontaining mass in the superior mediastinum (arrow), consistent with a lipoma.

A

Neurogenic lesions account for 95% of posterior mediastinal lesions. Lesions in the first decade are often malignant, typically neuroblastoma. In the second decade, these lesions are more commonly benign, particularly neurofibroma, schwannoma, and ganglioneuroma. Neuroblastoma is a malignant tumor that involves the mediastinum in 10% of cases. It can involve multiple rib spaces and frequently extends to the spinal canal. It has a fusiform shape, soft tissue attenuation (Fig. 41-34), and calcification in 50% of cases. Ganglioneuroma, neurofibroma, and schwannoma are seen in the second decade. They are smaller than ganglion cell tumors, are round or oval, and involve one

B FIG 41-28  Goiter. A, Axial CT shows a large heterogeneously enhancing mass in the superior mediastinum (arrow). B, Coronal CT in the same patient shows that the mass is contiguous with the thyroid and retrosternal extension of a thyroid goiter (arrow).

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1177

B FIG 41-29  Lymphoma. A, Axial CT shows a large heterogeneously enhancing mass in the anterior mediastinum, indenting on the aorta and pulmonary artery (arrow), which was shown to be a lymphoma. B, Axial CT at a lower level shows heterogeneously enhancing mass that is due to conglomerate lymph nodes (arrow).

or two vertebra. Erosion of adjacent rib and erosion of spinal canal may be seen. Neurenteric cysts contain both neural and gastrointestinal components. They are seen in the posterior mediastinum. They have well-defined margins with homogeneous attenuation of fluid, both in CT and MRI. There is no communication with cerebrospinal fluid. Extramedullary hematopoiesis is a rare condition of the posterior mediastinum characterized by formation of hematopoietic elements outside bone marrow. It is seen in severe anemias such as hemolytic anemia or sickle cell anemia. On CT and MRI, paravertebral masses are seen, with expansion and coarse trabeculation of adjacent ribs. Foregut enteric duplication cysts are cystic lesions that are lined by gastrointestinal mucosa. These are located in posterior or middle position and seen in the paraspinal region adjacent to the esophagus (Fig. 41-35). Achalasia cardia results in a dilated fluid/debris-filled esophagus in the posterior mediastinum (Fig. 41-36).

VASCULAR RINGS AND SLINGS A

B FIG 41-30  Germ cell tumor. A, Axial CT shows a large mass with homogeneous low attenuation in the anterior mediastinum (arrow). B, Coronal CT shows the large hypoattenuating anterior mediastinal mass (arrow), which was shown to be a germ cell tumor.

Vascular rings and slings are congenital vascular/ligamentous anomalies that cause a variable extent of encirclement of the trachea and esophagus. These anomalies may be asymptomatic or present with feeding difficulties, weight loss, and failure to thrive. Wheezing, stridor, dyspnea, tachypnea, and subcostal retractions are the clinical features.56 CT and MRI are important in the evaluation of vascular rings and defining a roadmap for surgery, which is typically performed in the first year of life for symptomatic rings. MRI has limited availability and involves higher cost, long acquisition times, need for deep sedation/ general anesthesia in young children, use of gadolinium-based contrast agents, and requires additional sequences for airway anomalies. CT can be performed faster, thus obviating the need for deep sedation/general anesthesia, high spatiotemporal resolution, MPR capabilities, and isotropic voxels, although there is a risk of radiation and use of potentially nephrotoxic iodinated contrast media.

Double Aortic Arch Double aortic arch is characterized by the presence of two aortic arches and is the most common type of symptomatic vascular ring (50%-60% of these cases).52,56 It is caused by persistence of the right and left fourth aortic arches and dorsal aorta.6 On CT or MRI there are two aortic

1178

PART II  CT and MR Imaging of the Whole Body

A

B FIG 41-31  Lymphatic malformation. A, Axial CT shows large predominantly cystic mass (arrow) containing calcifications in the anterior mediastinum. B, Coronal CT shows the large cystic mass indenting on adjacent structures (arrow).

A B

C FIG 41-32  Hemangioma. A, Coronal CT shows a large infiltrative mass with heterogeneous contrast enhancement in the lower neck and superior mediastinum (arrow). B and C, Axial fat-suppressed MR images of the upper thorax in a 6-month-old boy reveal extensive hyperintensity (arrow) surrounding—but not significantly displacing—mediastinal vascular structures, trachea, and esophagus. There is contiguous extension into the left axilla and posteriorly surrounding bilateral lung apices. This lesion is the inferiorly extending portion of an extensive hemangioma in the neck; patient had known PHACES syndrome.

A

B

C FIG 41-33  Bronchogenic cyst. A, Axial CT shows a well-defined cystic lesion in the posterior mediastinum (arrow), consistent with a bronchogenic cyst. There is also pericardial and left pleural effusion. B, Axial T2-weighted MRI in another patient shows a T2-hyperintense lesion in the subcarinal region (arrow), consistent with bronchogenic cyst. C, Coronal T2-weighted image in the same patient shows a hyperintense lesion (arrow) splaying the carina.

A

B FIG 41-34  Neuroblastoma. A, Axial contrast-enhanced CT shows an invasive posterior mediastinal mass in the left paraspinal region (arrow) in a patient with neuroblastoma. B, Axial STIR MRI in the same patient shows hyperintense lesion in the left paraspinal region (straight arrow) with extension to the spinal canal (curved arrow) and the adjacent rib, consistent with a neuroblastoma.

1180

PART II  CT and MR Imaging of the Whole Body

A

B FIG 41-35  Foregut enteric duplication cyst. A, Axial CT shows a mass indenting the midesophagus (arrow). B, Axial T2-weighted MRI shows a hyperintense lesion adjacent to the midesophagus (arrow), consistent with an esophageal duplication cyst.

ring. Using coronal reconstructions and volume rendering, the dominant arch can be identified, and typically surgical resection is performed with a thoracotomy on the side of the nondominant arch.

Right Aortic Arch with Retroesophageal Left Subclavian Artery

FIG 41-36  Achalasia cardia. Axial CT shows a severely dilated esophagus (arrow) filled with fluid and food debris secondary to achalasia cardia.

arches originating from the ascending aorta, which cross on either side of the trachea and esophagus to continue as a single descending aorta (Fig. 41-37). The right common carotid and subclavian arteries originate from the right arch, and the left common carotid and subclavian arteries originate from the left arch. The most common pattern is a right-dominant arch (55%-70%).113 Left-dominant and codominant patterns are less common. In 80%, the descending aorta is located on the left. The right arch is more posterior than the left arch in a leftsided descending aorta, but if there is a right-sided descending aorta, the left arch is more posterior than the right. The ductus is typically located on the left. The characteristic CT and MRI appearances at the level of the thoracic inlet are a four-artery sign, which is due to a characteristic trapezoidal arrangement of the four arch branch vessels. The arches encircle the trachea and esophagus to form a complete vascular

Right-sided aortic arch is caused by persistence of the right fourth arch and the right dorsal aorta, with involution of the left fourth arch and dorsal aorta, which is a reverse of the normal situation. It is seen in 0.1% of the normal population.107 In this anomaly the ascending aorta is located on the right as usual, but the arch is located on the right as well. The descending aorta is also located on the right but gradually extends to the left, particularly at the level of the diaphragmatic hiatus. In 84% of patients, there is a mirror-image branching pattern that is associated with cardiac anomaly in 90% of instances, mostly with tetralogy of Fallot.88 The ductus is typically located on the right, which does not cause a vascular ring. In 14% of right arches, there is an aberrant left subclavian artery, which originates as the last branch from the arch and courses behind the esophagus to reach the left side. This is caused by a persistent right fourth arch and regression of the left fourth arch between the left common carotid and left subclavian arteries. Although in 10% of these situations the ductus is located on the right, in 90% of cases it is located on the left. This will result in a ring that accounts for 30% of vascular rings, making it the second most common cause of a symptomatic vascular ring.128 A Kommerell’s diverticulum is seen, which is a bulbous diverticulum from the proximal descending aorta at the level of origin of the aberrant left subclavian artery.172 On CTA and MRA, the first branch arising from the arch is the left common carotid. It is followed by right common carotid, right subclavian, and lastly the aberrant left subclavian artery. The aberrant left subclavian artery has an abrupt change in caliber from the dilated proximal Kommerell’s diverticulum to the level of the retroesophageal segment; this is caused by a left-sided ductus arteriosus that extends from the base of the diverticulum from the descending thoracic  aorta to the left pulmonary artery (Fig. 41-38). The left ductus causes tracheal compression and an enlarged diverticulum/aneurysm can

CHAPTER 41  Chest Imaging in the Pediatric Patient

A B

C

D

E FIG 41-37  Double aortic arch. A, Axial CT at the level of the thoracic inlet shows the four-vessel sign of double aortic arch. The right branch is the right subclavian artery, the second is the right common carotid, the third is the left common carotid, and the fourth is the left subclavian branch. B, Axial MIP image shows two aortic arches, right and left (arrows), encircling the trachea and esophagus. C, Coronal MIP image shows bilateral aortic arches encircling the trachea and esophagus. The right aortic arch is larger than the left. D, Volume-rendered 3D image with posterior view shows the two aortic arches, right and left, with the right arch giving off the right common carotid and right subclavian branches and the left arch giving off the left common carotid and left subclavian branches. E, Coronal volume-rendered image of the airway shows compression of the lower trachea by the vascular ring of double aortic arch (arrow).

1181

1182

PART II  CT and MR Imaging of the Whole Body

A

B

D

C FIG 41-38  Right aortic arch with retroesophageal left subclavian artery. A, Axial CT shows a right-sided aortic arch with a large Kommerell’s diverticulum (arrow) compressing the posterior esophagus. Coronal MIP (B) and volume-rendered 3D CT reconstruction (C) images demonstrate a right aortic arch and prominent Kommerell’s diverticulum (arrows) in the proximal descending thoracic aorta. D, Coronal CT shows a rightsided aortic arch (arrow) compressing the trachea.

cause both tracheal and esophageal compression. This is treated with left thoracotomy and division of ligamentum arteriosum with oversewing of the stumps.26

Right Aortic Arch with Mirror-Image Branching and Intact Retroesophageal Left Ligamentum Arteriosum Right aortic arch with mirror-image branching is characterized by the first artery being the left brachiocephalic artery, which divides into the left subclavian and left common carotid artery, followed by the right common carotid and then the right subclavian artery. This is caused by a persistent right fourth arch and partial regression of the fourth arch between the left subclavian artery and the dorsal aorta. In 25% of

these cases, the ductus is located on the right and there is no ring, but in 75% the ductus is located on the left.71 When this ductus extends from the proximal descending aorta and has a retroesophageal course to reach the left to connect the left pulmonary artery, it results in a complete vascular ring (Fig. 41-39). This is associated with cardiac anomalies in 90%, usually tetralogy of Fallot. Occasionally the left ductus extends from the left brachiocephalic artery to the left pulmonary artery, which is typically not associated with a ring.71,88

Circumflex Retroesophageal Aortic Arch Circumflex retroesophageal aortic arch is the third most common type of vascular ring and is characterized by the arch (either right or left)

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1183

B FIG 41-39  Right aortic arch with mirror-image branching and intact retroesophageal left ligamentum arteriosum. Coronal reconstructed MIP (A) and volume-rendered 3D CT reconstruction (B) images show a rightsided aortic arch with mirror-image branching pattern and a patent ductus arteriosus (arrow) connecting with the left pulmonary artery, resulting in a ring.

coursing behind the esophagus, with the ascending and descending aortic segments located on either side of the thoracic spine (Fig. 41-40). In a person with a right-sided aortic arch, the arch runs to the right of the trachea, after which it courses behind the esophagus up to the level of the carina, beyond which it continues as the descending aorta. The ductus/ligamentum extends from the left descending aorta to the left pulmonary artery. With a left aortic arch, the arch courses to the left of the trachea, extends behind the esophagus to reach the right, and continues as the right descending aorta. A right ductus/ ligamentum connects the right descending aorta to the right pulmonary artery. This condition is treated with median sternotomy, cardiopulmonary bypass and relocation of the arch to anterior to the airway, and end-to-end anastomosis.26

Left Aortic Arch with Aberrant Right Subclavian Artery Left aortic arch with aberrant right subclavian artery is the most common type of aortic arch abnormality, where the first branch of the arch is the right common carotid artery, followed by the left common carotid artery, left subclavian artery, and then the right subclavian artery, which then courses behind the esophagus to reach the right side (Fig. 41-41). This is caused by regression of the right arch between the right subclavian and right common carotid arteries. The origin of the right subclavian artery is dilated to form a Kommerell’s diverticulum. Typically the ductus is located on the left and a vascular ring is not produced. However, rarely a ring is produced when there is a rightsided ductus and a circumflex right descending thoracic aorta.71

Brachiocephalic Artery Compression An anomalous origin of the right brachiocephalic artery from the arch, more posterior to the left, will result in the artery coursing in a right superior and posterior approach to continue as a right subclavian artery, which may result in anterior tracheal indentation or compres-

FIG 41-40  Circumflex retroesophageal aortic arch. Axial CT in a patient with circumflex retroesophageal aortic arch shows the right-side arch extending behind the esophagus (arrow) to reach the left.

sion8 (Fig. 41-42). This ring is located much higher than the earlierdescribed vascular rings. A significant compression occurs when there is narrowing of 70% to 80% of tracheal lumen.26 This anomaly is repaired surgically using a right anterolateral thoracotomy, elevation of the brachiocephalic artery from the trachea, and resuspension to the posterior aspect of the sternum.26

1184

PART II  CT and MR Imaging of the Whole Body

A

B FIG 41-41  Left aortic arch with aberrant right subclavian artery. A, Axial MIP image shows a left-sided aortic arch and an aberrant right subclavian artery (arrow) that originates as the last branch from the arch and courses behind the esophagus to reach the right. B, Coronal MIP image shows the aberrant right subclavian artery (arrow) originating from the proximal descending aorta and coursing to the right.

MPA

RPA

LPA

FIG 41-42  Brachiocephalic artery compression. Axial CT shows an anomalous course of the right brachiocephalic artery (arrow) that is causing anterior indentation of the upper trachea.

FIG 41-43  Pulmonary artery sling. Axial CT shows the left pulmonary

Pulmonary Artery Sling

artery (LPA) originating from the right pulmonary artery (RPA) instead of the main pulmonary artery (MPA) and coursing between the trachea and esophagus (arrow) to reach the left.

Pulmonary artery sling refers to the left pulmonary artery originating from the right pulmonary artery and subsequently coursing over the right main bronchus between the trachea and esophagus to reach the left hilum, thus forming a sling that compresses the trachea and esophagus (Fig. 41-43). This is caused by developmental obliteration of the left sixth aortic arch when the developing left lung bud captures its vascular supply from the right sixth arch, caudal to the developing tracheobronchial tree. This anomaly is associated with several airway anomalies, based on which it is classified.174 Type I, which has lower morbidity and mortality is associated with airway compression by the anomalous artery, but airway branching is normal. Type IA is associated with no airway anomaly, whereas type IB is associated with tracheal bronchus, tracheobronchomalacia, and unilateral pulmonary hyperinflation. Type II is associated with a more caudal anomalous pulmonary artery with associated long-segment tracheobronchial stenosis, branching abnormalities, left intermediate bronchus, right

bridging bronchus, T-shaped carina, complete tracheal rings, and bilateral pulmonary hyperinflation.174 In complete tracheal rings, the posterior membranous portion of the trachea is absent. In bridging bronchus, the right middle and lower lobes are supplied by a bronchus that originates from the left main bronchus, which crosses the mediastinum to reach the right side. The right upper lobe is supplied by a bronchus originating from the trachea at the level of the carina. Pulmonary artery sling should be distinguished from a right pulmonary agenesis, where the right pulmonary artery is absent but the left pulmonary artery originates from the main pulmonary artery. Pulmonary sling is treated by median sternotomy and cardiopulmonary bypass, with ligation and division of the anomalous artery and reimplantation into the main pulmonary artery anterior to the trachea.26

CHAPTER 41  Chest Imaging in the Pediatric Patient

A

1185

B FIG 41-44  Bronchopulmonary dysplasia (BPD). A, Axial CT in a patient with BPD shows heterogeneous attenuation containing high attenuation and also areas of low attenuation. B, Axial CT shows bilateral multiple small cystic lesions in lungs of a patient with BPD.

MISCELLANEOUS Bronchopulmonary Dysplasia Bronchopulmonary dysplasia (BPD) is a chronic pulmonary disorder that results from the use of high positive-pressure ventilation and highconcentration oxygen in neonates with respiratory distress syndrome. BPD is the most common cause of chronic lung disease in children. It is seen in preterm infants, in 10% of those weighing less than 1500 g and 20% weighing less than 1000 g. It is uncommon after 30 weeks’ gestation or weight over 1200 g. Pathologically it is characterized by inflammation, mucosal necrosis, fibrosis, and smooth muscle hypertrophy of the airways. There are four stages in the development of BPD, which may not necessarily be seen in the same patient. In stage 1 (1-3 days), hyaline membranes, atelectasis, vascular hyperemia, and lymphatic dilation are seen, similar to hyaline membrane disease. In stage 2 (4-10 days), lung destruction, bronchial obstruction, bronchial necrosis, fibrosis, and obliterative bronchiolitis are seen. In stage 3 (11-20 days) there is progressive repair of alveoli with hypertrophy of alveoli, bronchial wall muscle and glands, and presence of regenerating cells, air trapping, and pulmonary hyperinflation. In stage 4 (>1 month), emphysematous alveoli are seen. Pulmonary hypertension and cor pulmonale may be seen. On CT scan in stage 1, changes of respiratory distress syndrome, pneumothorax, and PIE are seen. In stage 2, interlobar septal thickening is the common finding. Expiratory air trapping (Fig. 41-44), hyperinflation, atelectasis, consolidation, prominent pulmonary arteries, and enlarged heart are seen. In severe cases there is marked distortion of lung parenchyma, large cysts, coarse fibrosis, and pulmonary hypertension. In stage 3, linear atelectasis and fibrotic changes are seen, as well as gradual clearing of areas of abnormal attenuation. In chronic stages, air trapping, fibrotic changes, triangular opacities, pulmonary architectural distortion, decreased bronchial diameters, bronchiectasis, bronchial wall thickening, tracheomegaly, tracheomalacia, enlarged pulmonary arteries, right ventricular enlargement, and pulmonary edema are seen. Differential diagnosis includes chronic lung disease, persistent pulmonary hypertension of the newborn, and Wilson-Mikity syndrome. Wilson-Mikity syndrome is seen in children between 1 and 2 months of age. History may include preterm delivery and low birth weight but typically no

history of high-pressure ventilation or high-concentration oxygen. Imaging is normal in the first week but later similar to BPD, with hyperinflation, stranding, streaky infiltrates, and cystic changes. These changes may persist for a few months to years.

CONCLUSION CT and MRI are valuable in the evaluation of several pediatric thoracic abnormalities. CT is highly sensitive in the evaluation of chest abnormalities and is ideally suited for children owing to rapid acquisition times. The radiation dose should be kept as low as possible by performing only justifiable studies and optimizing the protocols. MRI is used as a problem-solving tool in the evaluation of several cardiovascular, mediastinal, and pleural abnormalities.

REFERENCES 1. Ackman JB, Wu CC: MRI of the thymus. AJR Am J Roentgenol 197(1):W15–W20, 2011. 2. Adzick NS, Harrison MR, Crombleholme TM, et al: Fetal lung lesions: Management and outcome. Am J Obstet Gynecol 179(4):884–889, 1998. 3. Adzick NS, Harrison MR, Glick PL, et al: Fetal cystic adenomatoid malformation: Prenatal diagnosis and natural history. J Pediatr Surg 20(5):483–488, 1985. 4. Agrons GA, Rosado-de-Christenson ML, Kirejczyk WM, et al: Pulmonary inflammatory pseudotumor: Radiologic features. Radiology 206(2):511–518, 1998. 5. Ahmed M, Jacobi V, Vogl TJ: Multislice CT and CT angiography for non-invasive evaluation of bronchopulmonary sequestration. Eur Radiol 14(11):2141–2143, 2004. 6. Alsenaidi K, Gurofsky R, Karamlou T, et al: Management and outcomes of double aortic arch in 81 patients. Pediatrics 118(5):e1336–e1341, 2006. 7. Araki T, Sholl LM, Gerbaudo VH, et al: Intrathymic cyst: Clinical and radiological features in surgically resected cases. Clin Radiol 69(7):732– 738, 2014. 8. Ardito JM, Ossoff RH, Tucker GF, Jr, et al: Innominate artery compression of the trachea in infants with reflex apnea. Ann Otol Rhinol Laryngol 89(5 Pt 1):401–405, 1980.

1186

PART II  CT and MR Imaging of the Whole Body

9. Azizkhan RG, Crombleholme TM: Congenital cystic lung disease: Contemporary antenatal and postnatal management. Pediatr Surg Int 24(6):643–657, 2008. 10. Bahadori M, Liebow AA: Plasma cell granulomas of the lung. Cancer 31(1):191–208, 1973. 11. Bai W, Zhou X, Gao X, et al: Value of chest CT in the diagnosis and management of tracheobronchial foreign bodies. Pediatr Int 53(4):515– 518, 2011. 12. Balfour-Lynn IM, Abrahamson E, Cohen G, et al: BTS guidelines for the management of pleural infection in children. Thorax 60(Suppl 1):i1–i21, 2005. 13. Barakos JA, Brown JJ, Brescia RJ, et al: High signal intensity lesions   of the chest in MR imaging. J Comput Assist Tomogr 13(5):797–802, 1989. 14. Baron RL, Sagel SS, Baglan RJ: Thymic cysts following radiation therapy for Hodgkin disease. Radiology 141(3):593–597, 1981. 15. Beigelman C, Howarth NR, Chartrand-Lefebvre C, et al: Congenital anomalies of tracheobronchial branching patterns: Spiral CT aspects in adults. Eur Radiol 8(1):79–85, 1998. 16. Berrocal T, Madrid C, Novo S, et al: Congenital anomalies of the tracheobronchial tree, lung, and mediastinum: Embryology, radiology, and pathology. Radiographics 24(1):e17, 2004. 17. Bhalla M, Turcios N, Aponte V, et al: Cystic fibrosis: Scoring system with thin-section CT. Radiology 179(3):783–788, 1991. 18. Bisogno G, Brennan B, Orbach D, et al: Treatment and prognostic factors in pleuropulmonary blastoma: An EXPeRT report. Eur J Cancer 50(1):178–184, 2014. 19. Biyyam DR, Chapman T, Ferguson MR, et al: Congenital lung abnormalities: Embryologic features, prenatal diagnosis, and postnatal radiologic-pathologic correlation. Radiographics 30(6):1721–1738, 2010. 20. Boldrini R, Devito R, Diomedi-Camassei F, et al: Pulmonary blastomas of childhood: Histologic, immunohistochemical, ultrastructural aspects and therapeutic considerations. Ultrastruct Pathol 29(6):493–501, 2005. 21. Brillet PY, Dumont P, Bouaziz N, et al: Pulmonary arteriovenous malformation treated with embolotherapy: Systemic collateral supply at multidetector CT angiography after 2-20-year follow-up. Radiology 242(1):267–276, 2007. 22. Brody AS, Klein JS, Molina PL, et al: High-resolution computed tomography in young patients with cystic fibrosis: Distribution of abnormalities and correlation with pulmonary function tests. J Pediatr 145(1):32–38, 2004. 23. Brody AS, Tiddens HA, Castile RG, et al: Computed tomography in the evaluation of cystic fibrosis lung disease. Am J Respir Crit Care Med 172(10):1246–1252, 2005. 24. Byington CL, Spencer LY, Johnson TA, et al: An epidemiological investigation of a sustained high rate of pediatric parapneumonic empyema: Risk factors and microbiological associations. Clin Infect Dis 34(4):434–440, 2002. 25. Calder AD, Bush A, Brody AS, et al: Scoring of chest CT in children with cystic fibrosis: State of the art. Pediatr Radiol 44:1496–1506, 2014. 26. Backer CL, Mavroudis C: Vascular rings and pulmonary artery sling. In Mavroudis C, Backer C, Idriss RF, editors: Pediatric cardiac surgery, ed 4, Oxford, UK, 2012, Blackwell Publishing Ltd, pp 234–255. 27. Carr DH, Oades P, Trotman-Dickenson B, et al: Magnetic resonance scanning in cystic fibrosis: Comparison with computed tomography. Clin Radiol 50(2):84–89, 1995. 28. Cartin-Ceba R, Swanson KL, Krowka MJ: Pulmonary arteriovenous malformations. Chest 144(3):1033–1044, 2013. 29. Cass DL, Crombleholme TM, Howell LJ, et al: Cystic lung lesions   with systemic arterial blood supply: A hybrid of congenital cystic adenomatoid malformation and bronchopulmonary sequestration.   J Pediatr Surg 32(7):986–990, 1997. 30. Cerfolio RJ, Allen MS, Nascimento AG, et al: Inflammatory pseudotumors of the lung. Ann Thorac Surg 67(4):933–936, 1999. 31. Cho SY, Kim HC, Bae SH, et al: Case report. Demonstration of blood supply to pulmonary sequestration by MR and CT angiography.   J Comput Assist Tomogr 20(6):993–995, 1996.

32. Coffin CM, Humphrey PA, Dehner LP: Extrapulmonary inflammatory myofibroblastic tumor: A clinical and pathological survey. Semin Diagn Pathol 15(2):85–101, 1998. 33. Culiner MM, Grimes OF: Localized emphysema in association with bronchial cysts or mucoceles. J Thorac Cardiovasc Surg 41:306–313, 1961. 34. Cutler CS, Michel RP, Yassa M, et al: Pulmonary blastoma: Case   report of a patient with a 7-year remission and review of   chemotherapy experience in the world literature. Cancer 82(3):462–467, 1998. 35. Daltro P, Fricke BL, Kuroki I, et al: CT of congenital lung lesions in pediatric patients. AJR Am J Roentgenol 183(5):1497–1506, 2004. 36. Daltro P, Werner H, Gasparetto TD, et al: Congenital chest malformations: A multimodality approach with emphasis on fetal MR imaging. Radiographics 30(2):385–395, 2010. 37. Davenport M, Warne SA, Cacciaguerra S, et al: Current outcome of antenally diagnosed cystic lung disease. J Pediatr Surg 39(4):549–556, 2004. 38. Davis SD, Brody AS, Emond MJ, et al: Endpoints for clinical trials in young children with cystic fibrosis. Proc Am Thorac Soc 4(4):418–430, 2007. 39. Davis SD, Fordham LA, Brody AS, et al: Computed tomography reflects lower airway inflammation and tracks changes in early cystic fibrosis. Am J Respir Crit Care Med 175(9):943–950, 2007. 40. Davis RD, Oldham HN, Sabiston DC: Primary cysts and neoplasms of the mediastinum: Recent changes in clinical presentation, methods of diagnosis, management, and results. Ann Thorac Surg 44(3):229–237, 1987. 41. de Jong PA, Nakano Y, Lequin MH, et al: Progressive damage on high resolution computed tomography despite stable lung function in cystic fibrosis. Eur Respir J 23(1):93–97, 2004. 42. De Santis M, Masini L, Noia G, et al: Congenital cystic adenomatoid malformation of the lung: Antenatal ultrasound findings and fetal-neonatal outcome. Fifteen years of experience. Fetal Diagn Ther 15(4):246–250, 2000. 43. Deguchi E, Furukawa T, Ono S, et al: Intralobar pulmonary sequestration diagnosed by MR angiography. Pediatr Surg Int 21(7):576–577, 2005. 44. Dines DE, Seward JB, Bernatz PE: Pulmonary arteriovenous fistulas. Mayo Clin Proc 58(3):176–181, 1983. 45. Dishop MK, Kuruvilla S: Primary and metastatic lung tumors in the pediatric population: A review and 25-year experience at a large children’s hospital. Arch Pathol Lab Med 132(7):1079–1103, 2008. 46. Dommergues M, Louis-Sylvestre C, Mandelbrot L, et al: Congenital adenomatoid malformation of the lung: When is active fetal therapy indicated? Am J Obstet Gynecol 177(4):953–958, 1997. 47. Donnelly LF, Frush DP: Localized radiolucent chest lesions in neonates: Causes and differentiation. AJR Am J Roentgenol 172(6):1651–1658, 1999. 48. Donnelly LF, Klosterman LA: Pneumonia in children: Decreased parenchymal contrast enhancement—CT sign of intense illness and impending cavitary necrosis. Radiology 205(3):817–820, 1997. 49. Eber E: Antenatal diagnosis of congenital thoracic malformations: Early surgery, late surgery, or no surgery? Semin Respir Crit Care Med 28(3):355–366, 2007. 50. Eggli KD, Newman B: Nodules, masses, and pseudomasses in the pediatric lung. Radiol Clin North Am 31(3):651–666, 1993. 51. Eichinger M, Heussel CP, Kauczor HU, et al: Computed tomography and magnetic resonance imaging in cystic fibrosis lung disease. J Magn Reson Imaging 32(6):1370–1378, 2010. 52. Ekstrom G, Sandblom P: Double aortic arch. Acta Chir Scand 102(3):183–202, 1951. 53. Elhalaby EA, Abo Sikeena MH: Delayed presentation of congenital diaphragmatic hernia. Pediatr Surg Int 18(5–6):480–485, 2002. 54. Enns GM, Cox VA, Goldstein RB, et al: Congenital diaphragmatic defects and associated syndromes, malformations, and chromosome anomalies: A retrospective study of 60 patients and literature review. Am J Med Genet 79(3):215–225, 1998.

CHAPTER 41  Chest Imaging in the Pediatric Patient 55. Eren S, Balci AE, Dikici B, et al: Foreign body aspiration in children: Experience of 1160 cases. Ann Trop Paediatr 23(1):31–37, 2003. 56. Etesami M, Ashwath R, Kanne J, et al: Computed tomography in the evaluation of vascular rings and slings. Insights Imaging 5(4):507–521, 2014. 57. Falconer AR, Brown RA, Helms P, et al: Pulmonary sequelae in survivors of congenital diaphragmatic hernia. Thorax 45(2):126–129, 1990. 58. Faughnan ME, Thabet A, Mei-Zahav M, et al: Pulmonary arteriovenous malformations in children: Outcomes of transcatheter embolotherapy.   J Pediatr 145(6):826–831, 2004. 59. Felker RE, Tonkin IL: Imaging of pulmonary sequestration. AJR Am J Roentgenol 154(2):241–249, 1990. 60. Francis D, Jacobsen M: Pulmonary blastoma. Curr Top Pathol 73:265–294, 1983. 61. Frazier AA, Rosado de Christenson ML, Stocker JT, et al: Intralobar sequestration: Radiologic-pathologic correlation. Radiographics 17(3):725–745, 1997. 62. Garne E, Haeusler M, Barisic I, et al: Congenital diaphragmatic hernia: Evaluation of prenatal diagnosis in 20 European regions. Ultrasound Obstet Gynecol 19(4):329–333, 2002. 63. Ghaye B, Szapiro D, Fanchamps JM, et al: Congenital bronchial abnormalities revisited. Radiographics 21(1):105–119, 2001. 64. Gibson RL, Burns JL, Ramsey BW: Pathophysiology and management of pulmonary infections in cystic fibrosis. Am J Respir Crit Care Med 168(8):918–951, 2003. 65. Grijalva CG, Nuorti JP, Zhu Y, et al: Increasing incidence of empyema complicating childhood community-acquired pneumonia in the United States. Clin Infect Dis 50(6):805–813, 2010. 66. Guttmacher AE, Marchuk DA, White RI, Jr: Hereditary hemorrhagic telangiectasia. N Engl J Med 333(14):918–924, 1995. 67. Hamm H, Light RW: Parapneumonic effusion and empyema. Eur Respir J 10(5):1150–1156, 1997. 68. Hammas N, Chbani L, Rami M, et al: A rare tumor of the lung: Inflammatory myofibroblastic tumor. Diagn Pathol 7:83, 2012. 69. Helbich TH, Heinz-Peer G, Eichler I, et al: Cystic fibrosis: CT assessment of lung involvement in children and adults. Radiology 213(2):537–544, 1999. 70. Hernanz-Schulman M: Cysts and cystlike lesions of the lung. Radiol Clin North Am 31(3):631–649, 1993. 71. Hernanz-Schulman M: Vascular rings: A practical approach to imaging diagnosis. Pediatr Radiol 35(10):961–979, 2005. 72. Hodina M, Hanquinet S, Cotting J, et al: Imaging of cavitary necrosis   in complicated childhood pneumonia. Eur Radiol 12(2):391–396, 2002. 73. Holder AM, Klaassens M, Tibboel D, et al: Genetic factors in congenital diaphragmatic hernia. Am J Hum Genet 80(5):825–845, 2007. 74. Inaoka T, Takahashi K, Mineta M, et al: Thymic hyperplasia and thymus gland tumors: Differentiation with chemical shift MR imaging. Radiology 243(3):869–876, 2007. 75. Indeglia RA, Shea MA, Grage TB: Congenital cysts of the thymus gland. Arch Surg 94(1):149–152, 1967. 76. Isaacs RD: Necrotizing pneumonia in bacteraemic pneumococcal infection. Br J Dis Chest 80(3):295–296, 1986. 77. Ishii K, Maeda K, Hashihira M, et al: MRI of mediastinal cavernous hemangioma. Pediatr Radiol 20(7):556–557, 1990. 78. Islam S, Calkins CM, Goldin AB, et al: The diagnosis and management of empyema in children: A comprehensive review from the APSA Outcomes and Clinical Trials Committee. J Pediatr Surg 47(11):2101– 2110, 2012. 79. Jaffe A, Calder AD, Owens CM, et al: Role of routine computed tomography in paediatric pleural empyema. Thorax 63(10):897–902, 2008. 80. Jiménez S, Jiménez JR, Crespo M, et al: Computed tomography in children with cystic fibrosis: A new way to reduce radiation dose. Arch Dis Child 91(5):388–390, 2006. 81. Kaewlai R, Avery LL, Asrani AV, et al: Multidetector CT of blunt thoracic trauma. Radiographics 28(6):1555–1570, 2008.

1187

82. Kaufman RA: Calcified postinflammatory pseudotumor of the lung: CT features. J Comput Assist Tomogr 12(4):653–655, 1988. 83. Keslar P, Newman B, Oh KS: Radiographic manifestations of anomalies of the lung. Radiol Clin North Am 29(2):255–270, 1991. 84. Kim JH, Cho JH, Park MS, et al: Pulmonary inflammatory pseudotumor—A report of 28 cases. Korean J Intern Med 17(4):252– 258, 2002. 85. Kim YW, Donnelly LF: Round pneumonia: Imaging findings in a large series of children. Pediatr Radiol 37(12):1235–1240, 2007. 86. Kim TS, Han J, Kim GY, et al: Pulmonary inflammatory pseudotumor (inflammatory myofibroblastic tumor): CT features with pathologic correlation. J Comput Assist Tomogr 29(5):633–639, 2005. 87. Kitagawa M, Hislop A, Boyden EA, et al: Lung hypoplasia in congenital diaphragmatic hernia. A quantitative study of airway, artery, and alveolar development. Br J Surg 58(5):342–346, 1971. 88. Knight L, Edwards JE: Right aortic arch. Types and associated cardiac anomalies. Circulation 50(5):1047–1051, 1974. 89. Koss MN, Hochholzer L, O’Leary T: Pulmonary blastomas. Cancer 67(9):2368–2381, 1991. 90. Koşucu P, Ahmetoğlu A, Koramaz I, et al: Low-dose MDCT and virtual bronchoscopy in pediatric patients with foreign body aspiration. AJR Am J Roentgenol 183(6):1771–1777, 2004. 91. Landing BH, Dixon LG: Congenital malformations and genetic disorders of the respiratory tract (larynx, trachea, bronchi, and lungs). Am Rev Respir Dis 120(1):151–185, 1979. 92. Langham MR, Kays DW, Ledbetter DJ, et al: Congenital diaphragmatic hernia. Epidemiology and outcome. Clin Perinatol 23(4):671–688, 1996. 93. Langston C: New concepts in the pathology of congenital lung malformations. Semin Pediatr Surg 12(1):17–37, 2003. 94. Larsen H, Sorensen JB: Pulmonary blastoma: A review with special emphasis on prognosis and treatment. Cancer Treat Rev 22(3):145–160, 1996. 95. Lawler LP, Fishman EK: Multi-detector row CT of thoracic disease with emphasis on 3D volume rendering and CT angiography. Radiographics 21(5):1257–1273, 2001. 96. Lawler LP, Fishman EK: Arteriovenous malformations and systemic lung supply: Evaluation by multidetector CT and three-dimensional volume rendering. AJR Am J Roentgenol 178(2):493–495, 2002. 97. LeBlanc J, Guttentag AR, Shepard JA, et al: Imaging of mediastinal foregut cysts. Can Assoc Radiol J 45(5):381–386, 1994. 98. Lee EY, Greenberg SB, Boiselle PM: Multidetector computed tomography of pediatric large airway diseases: State-of-the-art. Radiol Clin North Am 49(5):869–893, 2011. 99. Lee EY, Restrepo R, Dillman JR, et al: Imaging evaluation of pediatric trachea and bronchi: Systematic review and updates. Semin Roentgenol 47(2):182–196, 2012. 100. Lee EY, Siegel MJ: MDCT of tracheobronchial narrowing in pediatric patients. J Thorac Imaging 22(3):300–309, 2007. 101. Leeuwen L, Fitzgerald DA: Congenital diaphragmatic hernia. J Paediatr Child Health 50(9):667–673, 2014. 102. LeMense GP, Reed CE, Silvestri GA: Pulmonary blastoma: A rare lung malignancy. Lung Cancer 15(2):233–237, 1996. 103. Li ST, Tancredi DJ: Empyema hospitalizations increased in US children despite pneumococcal conjugate vaccine. Pediatrics 125(1):26–33, 2010. 104. Liman ST, Altinok T, Topcu S, et al: Survival of biphasic pulmonary blastoma. Respir Med 100(7):1174–1179, 2006. 105. Lin AE, Pober BR, Adatia I: Congenital diaphragmatic hernia and associated cardiovascular malformations: Type, frequency, and impact on management. Am J Med Genet C Semin Med Genet 145C(2):201– 216, 2007. 106. Link KM, Samuels LJ, Reed JC, et al: Magnetic resonance imaging of the mediastinum. J Thorac Imaging 8(1):34–53, 1993. 107. Maldonado JA, Henry T, Gutierrez FR: Congenital thoracic vascular anomalies. Radiol Clin North Am 48(1):85–115, 2010. 108. Mata JM, Cáceres J: The dysmorphic lung: Imaging findings. Eur Radiol 6(4):403–414, 1996. 109. Mata JM, Cáceres J, Lucaya J, et al: CT of congenital malformations of the lung. Radiographics 10(4):651–674, 1990.

1188

PART II  CT and MR Imaging of the Whole Body

110. Matsubara O, Tan-Liu NS, Kenney RM, et al: Inflammatory pseudotumors of the lung: Progression from organizing pneumonia to fibrous histiocytoma or to plasma cell granuloma in 32 cases. Hum Pathol 19(7):807–814, 1988. 111. McAdams HP, Kirejczyk WM, Rosado-de-Christenson ML, et al: Bronchogenic cyst: Imaging features with clinical and histopathologic correlation. Radiology 217(2):441–446, 2000. 112. McAdams HP, Rosado-de-Christenson ML, Moran CA: Mediastinal hemangioma: Radiographic and CT features in 14 patients. Radiology 193(2):399–402, 1994. 113. McFaul R, Millard P, Nowicki E: Vascular rings necessitating right thoracotomy. J Thorac Cardiovasc Surg 82(2):306–309, 1981. 114. Meyer CA, White CS, Sherman KE: Diseases of the hepatopulmonary axis. Radiographics 20(3):687–698, 2000. 115. Michaelides SA, Passalidou E, Bablekos GD, et al: Cavitating lung lesion as a manifestation of inflammatory tumor (pseudotumor) of the lung: A case report and literature review. Am J Case Rep 15:258–265, 2014. 116. Miller PA, Williamson BR, Minor GR, et al: Pulmonary sequestration: Visualization of the feeding artery by CT. J Comput Assist Tomogr 6(4):828–830, 1982. 117. Min JY, Won HS, Lee MY, et al: Intrauterine therapy for macrocystic congenital cystic adenomatoid malformation of the lung. Obstet Gynecol Sci 57(2):102–108, 2014. 118. Nakata H, Egashira K, Watanabe H, et al: MRI of bronchogenic cysts.   J Comput Assist Tomogr 17(2):267–270, 1993. 119. Newman B: Congenital bronchopulmonary foregut malformations: Concepts and controversies. Pediatr Radiol 36(8):773–791, 2006. 120. Ng MY, Flight W, Smith E: Pulmonary complications of cystic fibrosis. Clin Radiol 69(3):e153–e162, 2014. 121. Odev K, Guler I, Altinok T, et al: Cystic and cavitary lung lesions in children: Radiologic findings with pathologic correlation. J Clin Imaging Sci 3:60, 2013. 122. Oh KS, Dorst JP, White JJ, et al: The syndrome of bronchial atresia or stenosis with mucocele and focal hyperinflation of the lung. Johns Hopkins Med J 138(2):48–53, 1976. 123. O’Sullivan BP, Frassica JJ, Rayder SM: Tracheal bronchus: A cause of prolonged atelectasis in intubated children. Chest 113(2):537–540, 1998. 124. O’Sullivan BP, Freedman SD: Cystic fibrosis. Lancet 373(9678):1891– 1904, 2009. 125. Ozçelik U, Göçmen A, Kiper N, et al: Congenital lobar emphysema: Evaluation and long-term follow-up of thirty cases at a single center. Pediatr Pulmonol 35(5):384–391, 2003. 126. Panicek DM, Heitzman ER, Randall PA, et al: The continuum of pulmonary developmental anomalies. Radiographics 7(4):747–772, 1987. 127. Pardes JG, Auh YH, Blomquist K, et al: CT diagnosis of congenital lobar emphysema. J Comput Assist Tomogr 7(6):1095–1097, 1983. 128. Park SC: Symposium on pediatric otolaryngology. Vascular abnormalities. Pediatr Clin North Am 28(4):949–955, 1981. 129. Paterson A: Imaging evaluation of congenital lung abnormalities in infants and children. Radiol Clin North Am 43(2):303–323, 2005. 130. Pilla TJ, Wolverson MK, Sundaram M, et al: CT evaluation of cystic lymphangiomas of the mediastinum. Radiology 144(4):841–842, 1982. 131. Pober BR: Genetic aspects of human congenital diaphragmatic hernia. Clin Genet 74(1):1–15, 2008. 132. Puderbach M, Eichinger M, Gahr J, et al: Proton MRI appearance of cystic fibrosis: Comparison to CT. Eur Radiol 17(3):716–724, 2007. 133. Reilly JS, Cook SP, Stool D, et al: Prevention and management of aerodigestive foreign body injuries in childhood. Pediatr Clin North Am 43(6):1403–1411, 1996. 134. Remy-Jardin M, Dumont P, Brillet PY, et al: Pulmonary arteriovenous malformations treated with embolotherapy: Helical CT evaluation of long-term effectiveness after 21-year follow-up. Radiology 239(2):576– 585, 2006. 135. Remy J, Remy-Jardin M, Giraud F: Wattinne L. Angioarchitecture of pulmonary arteriovenous malformations: Clinical utility of threedimensional helical CT. Radiology 191(3):657–664, 1994.

136. Remy J, Remy-Jardin M, Wattinne L, et al: Pulmonary arteriovenous malformations: Evaluation with CT of the chest before and after treatment. Radiology 182(3):809–816, 1992. 137. Restrepo R, Palani R, Matapathi UM, et al: Imaging of round pneumonia and mimics in children. Pediatr Radiol 40(12):1931–1940, 2010. 138. Rose RW, Ward BH: Spherical pneumonias in children simulating pulmonary and mediastinal masses. Radiology 106(1):179–182, 1973. 139. Rovin JD, Rodgers BM: Pediatric foreign body aspiration. Pediatr Rev 21(3):86–90, 2000. 140. Schurawitzki H, Stiglbauer R, Klepetko W, et al: CT and MRI in benign mediastinal haemangioma. Clin Radiol 43(2):91–94, 1991. 141. Seline TH, Gross BH, Francis IR: CT and MR imaging of mediastinal hemangiomas. J Comput Assist Tomogr 14(5):766–768, 1990. 142. Shady K, Siegel MJ, Glazer HS: CT of focal pulmonary masses in childhood. Radiographics 12(3):505–514, 1992. 143. Shaffer K, Rosado-de-Christenson ML, Patz EF, Jr, et al: Thoracic lymphangioma in adults: CT and MR imaging features. AJR Am J Roentgenol 162(2):283–289, 1994. 144. Shah RM, Sexauer W, Ostrum BJ, et al: High-resolution CT in the acute exacerbation of cystic fibrosis: Evaluation of acute findings, reversibility of those findings, and clinical correlation. AJR Am J Roentgenol 169(2):375–380, 1997. 145. Shin SM, Kim WS, Cheon JE, et al: CT in children with suspected residual foreign body in airway after bronchoscopy. AJR Am J Roentgenol 192(6):1744–1751, 2009. 146. Shovlin CL, Jackson JE, Bamford KB, et al: Primary determinants of ischaemic stroke/brain abscess risks are independent of severity of pulmonary arteriovenous malformations in hereditary haemorrhagic telangiectasia. Thorax 63(3):259–266, 2008. 147. Silva AB, Muntz HR, Clary R: Utility of conventional radiography in the diagnosis and management of pediatric airway foreign bodies. Ann Otol Rhinol Laryngol 107(10 Pt 1):834–838, 1998. 148. Sly PD, Brennan S, Gangell C, et al: Lung disease at diagnosis in infants with cystic fibrosis detected by newborn screening. Am J Respir Crit Care Med 180(2):146–152, 2009. 149. Snyder CS, Dell’Aquila M, Haghighi P, et al: Clonal changes in inflammatory pseudotumor of the lung: A case report. Cancer 76(9):1545–1549, 1995. 150. Stern M, Wiedemann B, Wenzlaff P, et al: From registry to quality management: The German Cystic Fibrosis Quality Assessment project 1995 2006. Eur Respir J 31(1):29–35, 2008. 151. Stigers KB, Woodring JH, Kanga JF: The clinical and imaging spectrum of findings in patients with congenital lobar emphysema. Pediatr Pulmonol 14(3):160–170, 1992. 152. Stocker JT, Kagan-Hallet K: Extralobar pulmonary sequestration: Analysis of 15 cases. Am J Clin Pathol 72(6):917–925, 1979. 153. Stocker JT: Sequestrations of the lung. Semin Diagn Pathol 3(2):106– 121, 1986. 154. Stocker J: Congenital pulmonary airway malformation: A new name for and an expanded classification of congenital cystic adenomatoid malformation of the lung. Histopathology 41(Suppl 2):424–430, 2002. 155. Strollo DC, Rosado-de-Christenson ML, Jett JR: Primary mediastinal tumors: Part II. Tumors of the middle and posterior mediastinum. Chest 112(5):1344–1357, 1997. 156. Strollo DC, Rosado de Christenson ML, Jett JR: Primary mediastinal tumors. Part 1: tumors of the anterior mediastinum. Chest 112(2):511– 522, 1997. 157. Svedström E, Puhakka H, Kero P: How accurate is chest radiography in the diagnosis of tracheobronchial foreign bodies in children? Pediatr Radiol 19(8):520–522, 1989. 158. Swift AJ, Woodhouse N, Fichele S, et al: Rapid lung volumetry using ultrafast dynamic magnetic resonance imaging during forced vital capacity maneuver: Correlation with spirometry. Invest Radiol 42(1):37–41, 2007. 159. Takayama Y, Yabuuchi H, Matsuo Y, et al: Computed tomographic and magnetic resonance features of inflammatory myofibroblastic tumor of the lung in children. Radiat Med 26(10):613–617, 2008.

CHAPTER 41  Chest Imaging in the Pediatric Patient 160. Taylor GA, Atalabi OM, Estroff JA: Imaging of congenital diaphragmatic hernias. Pediatr Radiol 39(1):1–16, 2009. 161. Terheggen-Lagro SW, Arets HG, van der Laag J, et al: Radiological and functional changes over 3 years in young children with cystic fibrosis. Eur Respir J 30(2):279–285, 2007. 162. Terheggen-Lagro S, Truijens N, van Poppel N, et al: Correlation of six different cystic fibrosis chest radiograph scoring systems with clinical parameters. Pediatr Pulmonol 35(6):441–445, 2003. 163. Tiddens HA: Chest computed tomography scans should be considered as a routine investigation in cystic fibrosis. Paediatr Respir Rev 7(3):202–208, 2006. 164. Van Dyke JA, Sagel SS: Calcified pulmonary sequestration: CT demonstration. J Comput Assist Tomogr 9(2):372–374, 1985. 165. Van Meurs KP, Newman KD, Anderson KD, et al: Effect of extracorporeal membrane oxygenation on survival of infants with congenital diaphragmatic hernia. J Pediatr 117(6):954–960, 1990. 166. Veenma DC, de Klein A, Tibboel D: Developmental and genetic aspects of congenital diaphragmatic hernia. Pediatr Pulmonol 47(6):534–545, 2012. 167. Vu L, Tsao K, Lee H, et al: Characteristics of congenital cystic adenomatoid malformations associated with nonimmune hydrops and outcome. J Pediatr Surg 42(8):1351–1356, 2007. 168. Vult von Steyern K, Björkman-Burtscher IM, Geijer M: Radiography, tomosynthesis, CT and MRI in the evaluation of pulmonary cystic fibrosis: An untangling review of the multitude of scoring systems. Insights Imaging 4(6):787–798, 2013. 169. Wagner AL, Szabunio M, Hazlett KS, et al: Radiologic manifestations of round pneumonia in adults. AJR Am J Roentgenol 170(3):723–726, 1998. 170. Wan YL, Kuo HP, Tsai YH, et al: Eight cases of severe acute respiratory syndrome presenting as round pneumonia. AJR Am J Roentgenol 182(6):1567–1570, 2004. 171. Weber SA, Ferrari GF: Incidence and evolution of nasal polyps in children and adolescents with cystic fibrosis. Braz J Otorhinolaryngol 74(1):16–20, 2008.

1189

172. Weinberg PM: Aortic arch anomalies. J Cardiovasc Magn Reson 8(4):633–643, 2006. 173. Weisbrod GL, Chamberlain DW, Tao LC: Pulmonary blastoma, report of three cases and a review of the literature. Can Assoc Radiol J 39(2):130–136, 1988. 174. Wells TR, Gwinn JL, Landing BH, et al: Reconsideration of the anatomy of sling left pulmonary artery: The association of one form with bridging bronchus and imperforate anus. Anatomic and diagnostic aspects. J Pediatr Surg 23(10):892–898, 1988. 175. White RI, Pollak JS: Pulmonary arteriovenous malformations: Diagnosis with three-dimensional helical CT—A breakthrough without contrast media. Radiology 191(3):613–614, 1994. 176. Wick MR, Ritter JH, Humphrey PA: Sarcomatoid carcinomas of the lung: A clinicopathologic review. Am J Clin Pathol 108(1):40–53, 1997. 177. Wielpütz MO, Eichinger M, Puderbach M: Magnetic resonance imaging of cysti fibrosis lung disease. J Thorac Imaging 28(3):151–159, 2013. 178. Williams AJ, Schuster SR: Bronchial atresia associated with a bronchogenic cyst. Evidence of early appearance of atretic segments. Chest 87(3):396–398, 1985. 179. Wilson RD, Hedrick HL, Liechty KW, et al: Cystic adenomatoid malformation of the lung: Review of genetics, prenatal diagnosis, and in utero treatment. Am J Med Genet A 140(2):151–155, 2006. 180. Winters WD, Effmann EL: Congenital masses of the lung: Prenatal and postnatal imaging evaluation. J Thorac Imaging 16(4):196–206, 2001. 181. Wootton-Gorges SL, Stein-Wexler R, Walton JW, et al: Comparison of computed tomography and chest radiography in the detection of rib fractures in abused infants. Child Abuse Negl 32(6):659–663, 2008. 182. Yano Y, Mori M, Kagami S, et al: Inflammatory pseudotumor of the lung with rapid growth. Intern Med 48(15):1279–1282, 2009. 183. Yasufuku M, Maeda K, Takano Y: Thymopharyngeal duct cyst: An unusual cause of respiratory compromise. Pediatr Surg Int 25(9):807– 809, 2009. 184. Yi E, Aubry MC: Pulmonary pseudoneoplasms. Arch Pathol Lab Med 134(3):417–426, 2010.

42  Biliary Tract and Gallbladder Jae Hoon Lim, Kyoung Won Kim, and Dong-il Choi

BILIARY TRACT NORMAL ANATOMY AND VARIANTS On high-resolution computed tomography (CT), normal intrahepatic bile ducts (IHDs) appear as linear water-density structures accompanying the portal vein branches.177 Normal IHDs measure less than 3 mm. They appear to be randomly scattered throughout the liver but are confluent toward the hilum. The IHDs from each lobe unite to form the right and left main hepatic ducts, which are located anterior to the portal veins (Fig. 42-1). Periportal lymphedema can mimic a dilated IHD but shows a low-density area completely surrounding the portal vein.306 A common anomaly is an aberrant IHD that drains a circumscribed portion of the liver, such as an anterior or posterior segment right lobe duct that drains into the left rather than the right main hepatic duct. Surgical difficulties may arise when the cystic duct enters the right main hepatic duct before joining the left main hepatic duct. The right and left main hepatic ducts unite in the hilum to form the common hepatic duct (CHD). The CHD usually courses along a 45-degree oblique plane with reference to the midline sagittal plane, which lies to the right and lateral to the proper hepatic artery. The right hepatic artery typically branches off the proper hepatic artery and passes between the CHD and the main portal vein in about 80% of subjects.342 The common bile duct (CBD) forms when the cystic duct joins the CHD. This union occurs at varying levels, from high in the porta hepatis to near the ampulla of Vater. Because the union is usually not demonstrated on CT, the term common duct is used when the CHD and CBD cannot be differentiated. Duplications of the CBD and cystic duct are rarely observed. On CT the CHD usually measures 3 to 6 mm in short-axis diameter, and the CBD measures up to 8 mm.12 The diameter should be measured in the short axis because the duct often courses obliquely through the transverse image.96 In postcholecystectomy patients and some older adults, the CBD may measure up to 10 mm without obstruction.12 The wall of the CHD and CBD can normally be demonstrated and measures less than 1.5 mm.298 The wall may enhance normally and be brighter than the adjacent pancreas on contrastenhanced CT. The CBD enters the pancreas and typically lies along the posterior and lateral aspect of the pancreatic head. The distal CBD and main pancreatic duct come into contact on the medial side of the descending part of the duodenum. The two ducts pass separately through the wall of the duodenum and unite to form a short dilated tube—the ampulla of Vater. The sphincter of Oddi is the circular muscle complex around the CBD, pancreatic duct, and ampulla of Vater; it consists of the sphincter choledochus, sphincter pancreaticus, and sphincter ampullae. On endoscopic retrograde cholangiopancrea-

1192

tography (ERCP), the ampullary segment is usually not visualized because of the contraction of the sphincter of Oddi (Fig. 42-2). The ampulla of Vater is also responsible for the potential reflux of pancreatic juice into the CBD.

CONGENITAL BILIARY ANOMALIES Biliary Atresia In biliary atresia, ultrasonography (US), CT, and a DISIDA (diisopropyl iminodiacetic acid) scan are the initial imaging studies to evaluate the biliary tree. In more than 10% of patients with biliary atresia, other anomalies are present, such as polysplenia, bilateral bilobed lungs, azygos continuation of the inferior vena cava, intestinal malrotation, and situs inversus.76 CT can identify the development of cirrhosis, varices, splenomegaly, and ascites and plays an important role in the evaluation of such patients before liver transplantation.

Anomalous Pancreaticobiliary Ductal Junction Anomalous pancreaticobiliary ductal junction is an uncommon congenital anomaly in which the main pancreatic duct and CBD are joined outside the duodenal wall and form a long common channel (usually >15 mm).181,322 Although this anomaly is found in almost all patients with congenital choledochal cyst, it is also observed in patients without choledochal cyst.181 Because an anomalous pancreaticobiliary ductal junction is frequently associated with cholangiocarcinoma, early diagnosis is important.181 ERCP is the most reliable method for detecting anomalous pancreaticobiliary ductal junction, but it is an invasive procedure (Fig. 42-3A). Recently magnetic resonance (MR) cholangiopancreatography (MRCP) and multidetector CT (MDCT) with multiplanar reconstruction of images have been used to diagnose this anomaly (see Fig. 42-3B and C).146,237

Choledochal Cyst Choledochal cyst involves congenital cystic dilatation of any portion of the extrahepatic bile ducts, most commonly the main portion of the CBD.70,80,108,187,357 It is classified based on the spectrum of morphologic changes in the bile ducts. The most common classification scheme is the Todani modification of the Alonso-Lej classification326: Type I: single cystic dilatation of the CBD, CHD, or both (80%-90% of cases) (Fig. 42-4) Type II: true diverticulum of the CBD (3%) Type III: choledochocele (5%) Type IV: any combination of cysts, which may include intrahepatic cystic dilatation (10%). Type IV-A involves multiple cysts of the common duct and IHDs, and type IV-B involves multiple cysts of the common duct. Type V: cystic dilatation of only the IHDs—so-called Caroli’s disease (rare)

CHAPTER 42  Biliary Tract and Gallbladder

A

1193

B FIG 42-1  A, High-resolution CT shows normal intrahepatic bile ducts (IHDs) (arrows) as linear water-density structures accompanying the portal vein branches. B, T1-weighted MRI after administration of gadobenate dimeglumine demonstrates IHDs (arrows) with biliary excretion of contrast material located anterior to the portal veins.

FIG 42-2  Normal ERCP image shows the bile ducts, gallbladder, and main pancreatic duct. The distal part of the common bile duct (arrow) is not seen owing to contraction of the sphincter of Oddi.

Among these five types, types III and V are not cystic dilatation of the CBD and thus are not considered true choledochal cysts. The cause of this condition is not clear, but it may be multifactorial. It has been suggested that choledochal cyst, biliary atresia, and neonatal hepatitis are part of the same spectrum of disease.194 This condition is also associated with an increased incidence of gallbladder anomalies and congenital hepatic fibrosis. It may result from an anomalous pancreaticobiliary ductal junction, with free reflux of pancreatic enzymes into the CBD. Choledochal cysts are more common in females. Although they can present from birth to old age, about 60% of choledochal cysts are found before adolescence. Newborns and infants present with obstructive jaundice; older children and adults may have right upper quadrant pain, mass, and intermittent jaundice. The diagnosis is easily made by

detecting cystic dilatation of the bile duct on cross-sectional imaging. CT and magnetic resonance imaging (MRI) can make the diagnosis if direct communication with the cyst and the biliary tree is demonstrated (see Figs. 42-3 and 42-4). Cholangiography can confirm the diagnosis. Mild secondary dilatation of IHDs, stones, or sludge may be found as well (Fig. 42-5). Complications of choledochal cysts in adults include rupture with bile peritonitis, stone formation, cholangitis, hemorrhage, portal vein thrombosis, biliary cirrhosis, liver abscess, pancreatitis, and cholangiocarcinoma.108,284 An increased incidence (4%-28%) of malignancy involving the hepatobiliary system has been reported in patients with choledochal cysts, compared with normal subjects.188,325-327,354 Malignancy developing in patients with choledochal cyst may be caused by chronic mucosal irritation; it usually occurs in the choledochal cyst but can develop anywhere in the biliary tree, including the gallbladder354 (Fig. 42-6). Cholangiocarcinoma arising in a choledochal cyst appears as a soft tissue–density mass or irregular thickening of the cyst wall. The principal management of choledochal cysts is surgery, with excision of all cyst tissue and reconstruction of continuity between the liver and the gut.284,325

Choledochocele Choledochocele is a rare anatomic anomaly of unknown cause. It is a protrusion of a dilated intramural segment of the distal CBD into the duodenum, analogous to a ureterocele.61,296,361 On CT a round waterdensity structure is found in or medial to the pancreatic head or occasionally within the lumen of the duodenum. Differential diagnostic considerations include an intraluminal duodenal diverticulum, small pseudocyst, duodenal duplication cyst, or small cystic neoplasm of the pancreas. Cholangiography shows smooth saccular dilatation of the intramural segment of the CBD (Fig. 42-7). Stones and sludge are common, and patients sometimes have episodes of biliary colic, intermittent jaundice, and pancreatitis as well.61,296

Caroli’s Disease Caroli’s disease is also known as communicating cavernous ectasia of the IHDs. The uncommon pure form of this entity involves congenital saccular dilatation of only the IHDs. The more common classic form consists of saccular dilatation of the IHDs with congenital hepatic

C

B

A

C

FIG 42-3  Anomalous pancreaticobiliary ductal junction. A, ERCP shows a long common channel measuring more than 15 mm (arrowheads). B, MRCP demonstrates an anomalous pancreaticobiliary ductal junction (arrow) and a choledochal cyst (C). Note the long common channel (arrowheads). C, Coronal MDCT with multiplanar reconstruction demonstrates a long common channel (arrowheads) and choledochal cyst (C).

C

C C

FIG 42-4  Choledochal cyst (C). MRCP demonstrates a markedly dilated common bile duct (choledochal cyst) and aberrant insertion of the common duct into the pancreatic duct (arrow).

FIG 42-5  Choledochal cyst (C) with a small stone (arrow) on MRCP.

CHAPTER 42  Biliary Tract and Gallbladder

C

B

A

FIG 42-6  Gallbladder cancer associated with choledochal cyst (C). A, MRCP shows a lobulated filling-defect lesion (arrows) in the gallbladder. B, T2-weighted transverse MRI demonstrates gallbladder cancer (arrows).

C

A

B

C FIG 42-7  Choledochocele. A, MRCP shows a choledochal cyst (C) and choledochocele (arrow). T2-weighted transverse MRI (B) and MDCT (C) show a small cholangiocarcinoma (arrow) within the choledochocele.

1195

1196

PART II  CT and MR Imaging of the Whole Body

FIG 42-8  Caroli’s disease. MDCT shows cystic dilatation of the intrahepatic bile ducts with stones (white arrows). The black arrow indicates the so-called central dot sign.

fibrosis.56,234,314 The common association of congenital CBD dilatation suggests that Caroli’s disease may be part of the spectrum of choledochal cyst. In addition this disease is frequently associated with renal tubular ectasia and other renal cystic disease.234 Clinically Caroli’s disease usually manifests in adulthood; however, it can be observed in newborns and infants. Adult patients generally present with recurrent attacks of cholangitis. Infants and children may present with hematemesis caused by hepatic fibrosis and portal hypertension.77,158,234 On CT the saccular dilatation of the IHDs may mimic multiple hepatic cysts. The cystic areas of Caroli’s disease can be shown to communicate directly with the IHDs.12 The distribution of cystic dilatation is often segmental (Fig. 42-8). CT may also show tiny dots with strong contrast enhancement within dilated IHDs—the so-called central dot sign. These dots represent the enhancing portal veins.56 Caroli’s disease is best demonstrated by cholangiography, which shows saccular dilatations of the IHDs. Complications of Caroli’s disease include stone formation, cholangitis, hepatic abscess, hepatic amyloidosis, pancreatitis, and biliary malignancy.133,222

PATHOLOGIC CONDITIONS Choledocholithiasis (Bile Duct Stones) Stones in the extrahepatic duct may originate from the gallbladder or primarily from within the bile ducts. Primary stones are formed primarily in the extrahepatic bile ducts. Clearly, stones in the extrahepatic ducts in patients with cholecystectomy or congenital absence of the gallbladder are primary.27 Extrahepatic duct stones that originate in the gallbladder and later pass through the cystic duct into the CBD are called secondary stones.27 In Western countries, stones in the CBD are composed of cholesterol in about 80% of cases and calcium bilirubinate in 20%. Cholesterol stones arise exclusively in the gallbladder; these stones contain more than 50% crystalline cholesterol monohydrate.6 Calcium bilirubinate stones are composed predominantly of bilirubin calcium  salt; these fragile stones are also called pigment stones.6 About 95% of patients with choledocholithiasis in Western countries also have gallbladder stones228 (Fig. 42-9); when they do not, the gallbladder may show acute or chronic inflammatory changes, suggesting that it previously contained stones27 (Fig. 42-10).

FIG 42-9  Three small stones in the gallbladder and one stone in the common bile duct, which migrated from the gallbladder.

FIG 42-10  A small stone (arrow) in the common bile duct originated in the gallbladder. The wall of the gallbladder is thickened and enhanced, suggesting acute cholecystitis.

In Asia, pigment stones account for more than 50% of stones, and they are formed primarily in the bile ducts6 (Fig. 42-11). Unlike cholesterol stones, pigment stones can crumble or crush easily.6 Some stones are very friable or muddy, like earth, floating in the bile as debris. The formation of pigment stones is related to a high prevalence of recurrent pyogenic cholangitis. Infection of the biliary tract, such  as with Escherichia coli, Clonorchis sinensis, Opisthorchis viverrini, or Ascaris lumbricoides, increases the likelihood of pigment stone formation by increasing the concentration of insoluble calcium salts of unconjugated bilirubin.239

Intrahepatic Stones.  Intrahepatic biliary stones, which are formed primarily within the IHDs, are rare in Western countries.183 These stones occur in patients with underlying biliary anomalies such as Caroli’s disease, biliary atresia, or cystic fibrosis or previous biliary surgery.92 In some parts of Asia, intrahepatic stones are common in patients with recurrent pyogenic cholangitis. The stones are found mostly in the large-caliber ducts, such as the left hepatic lobe (Fig. 42-12) or the posterior segment of the right hepatic lobe. The stones

CHAPTER 42  Biliary Tract and Gallbladder migrate downward to the CBD or reside in the liver when there is a stricture. The bile duct wall shows chronic proliferative cholangitis, suppurative cholangitis, and chronic granulomatous cholangitis, resulting in severe fibrosis and inflammatory cell infiltration.183

Natural History.  The natural history of choledocholithiasis is unpredictable. Small stones may pass spontaneously into the duodenum without causing symptoms. Clinical and autopsy series suggest that nearly half of stones do not produce clinical manifestations such as cholangitis or jaundice.359 Spontaneous resolution of jaundice occurs as the stones pass through the duodenal papilla or if the stones float back up to the proximal bile duct away from the narrow distal end.27 Passage of a large stone through the narrow duodenal papilla can induce papillitis or papillary stenosis and incomplete bile duct obstruction254 (Fig. 42-13). Stones that do not pass and reside in the bile duct may induce obstructive jaundice and cholangitis or pancreatitis (Fig. 42-14). When

1197

a stone obstructs the bile duct in the absence of infected bile, asymptomatic fluctuating jaundice often ensues. When a stone obstructs the CBD and the bile becomes infected, acute suppurative cholangitis ensues (Fig. 42-15). The classic triad of symptoms is fever with chills, pain, and jaundice. In patients with recurrent pyogenic cholangitis, stones are continually formed in the bile ducts and repeatedly pass through the ampulla of Vater and duodenal papillary orifice; this can cause papillary stenosis, presenting with recurrent bile duct obstruction and cholangitis.272 Typically patients experience one or two attacks of acute suppurative cholangitis a year.

Imaging Findings.  The detectability of a bile duct stone on CT depends on the calcium content. Although many stones are cholesterol stones, some have sufficient bilirubinate and calcium to exhibit high attenuation on CT images6 (Fig. 42-16). About 20% of stones are homogeneously calcified and easily seen; in others the calcium content is variable.14 Depending on the composition, stones may show soft tissue attenuation, near-water attenuation, or fat or air attenuation, with peripheral calcification around the stone.14 Precontrast CT is

FIG 42-11  Multiple stones formed in the common bile duct. There is a tight stricture at the confluence of the bile ducts (arrow). This patient underwent cholecystectomy 10 years ago, so the stones formed in the common bile duct.

A

FIG 42-12  Multiple calcified intrahepatic stones formed in the liver in a patient with recurrent pyogenic cholangitis.

B FIG 42-13  Duodenal papillitis due to a passed stone. The duodenal papilla is swollen (arrow) on CT (A) and cholangiogram (B), and the bile ducts are dilated.

1198

PART II  CT and MR Imaging of the Whole Body

FIG 42-14  Gallstone pancreatitis. The pancreas is enlarged and there is a peripancreatic fluid collection. The common bile duct (arrow) is dilated and the wall is thick. Note the tiny stones in the gallbladder and the thickened gallbladder wall.

necessary to visualize calcified stones because postcontrast CT masks them264 (Fig. 42-17). Some stones with mixed composition are isoattenuating to bile or adjacent soft tissue and are therefore difficult to detect on CT (Fig. 42-18). Pure cholesterol stones are rare but exhibit low attenuation resembling that of soft tissue (Fig. 42-19) or bile (Fig. 42-20). Stones containing fat or air can be detected because of the difference in attenuation from bile (Fig. 42-21). For more accurate detection of intrahepatic ductal stones, thin section with narrow collimation (<2.5 mm) is necessary. The presence of bile around a stone results in a ring (bull’s-eye or target sign) or a crescent of water attenuation between the stone and the bile duct wall11,14 (Fig. 42-22). A concentrically thickened bile duct wall with enhancement may be seen around a stone, representing inflammation and fibrosis due to irritation by the stone. On cholangiograms or MR cholangiograms, stones are depicted as single or multiple filling defects (Fig. 42-23). Muddy stones, small fragments of stones, and gravel are difficult to visualize on CT (see Fig. 42-20), MR, or even cholangiography. Bile ducts may not be dilated despite the presence of stones in the common duct in 24% to 36% of cases12,65 (Fig. 42-24). In acute obstruction due to stone impaction, the bile duct does not become dilated until some time has passed. Bile ducts remain dilated after passage of

FIG 42-15  Suppurative cholangitis caused by gallbladder stone migration into the common bile duct. Coronal CT shows slightly dilated bile ducts with severe thickening of the walls of the extrahepatic duct and gallbladder.

A

FIG 42-16  Calcified stone (arrow) in the distal common bile duct.

B FIG 42-17  A, Heavily calcified left intrahepatic stones (arrows) are seen on precontrast CT. B, The stones are not clearly delineated on postcontrast CT because the adjacent liver parenchyma enhances.

CHAPTER 42  Biliary Tract and Gallbladder

1199

A A

B FIG 42-18  Innumerable small intrahepatic stones in a patient with recurrent pyogenic cholangitis. Because attenuation of the stones and liver is similar, CT (A) shows several weakly calcified intrahepatic stones (arrows), whereas many stones are visible on cholangiogram (B).

B FIG 42-20  Common bile duct stone (arrow) with water attenuation on CT (A) and cholangiogram (B).

FIG 42-19  Common bile duct stone with soft tissue attenuation (arrow).

FIG 42-21  Common bile duct stone (arrow) containing air presenting with the “Mercedes-Benz” sign.

1200

PART II  CT and MR Imaging of the Whole Body

a stone through the papilla of Vater because the duodenal papillary orifice becomes tight owing to papillitis and papillary edema (see  Fig. 42-13). Small stones and gravel repeatedly passing through the duodenal papillary orifice can cause papillitis and eventually papillary stenosis.92,359 CT or MRI demonstrates a dilated and thickened bile duct and an enlarged duodenal papilla164,275 (see Fig. 42-13) but no actual stone in the bile duct. The ability to detect a biliary stone by imaging varies depending on the size, location, and components of the stone.90,142,164,209,275,282 Each modality has advantages and disadvantages. With CT, detection depends on the stone’s calcium content and bile duct dilatation. Detection of stones by MRI is hampered by artifacts caused by pulsation of the adjacent vessels. During cholangiography, the technique of cannulation and bile duct opacification must be optimal for visualization of small stones.

Sometimes a stone that is not detected by one imaging modality is demonstrated by another (Fig. 42-25; see Fig. 42-24). Thus two or more imaging methods may be necessary for the diagnosis of bile duct stones.

Cholangitis

Suppurative Cholangitis.  Acute suppurative cholangitis is a bacterial infection of the bile ducts occurring mostly in patients with biliary obstruction caused by either stones or a tumor. Most stones originate from the gallbladder (Fig. 42-26). The tumor may be a carcinoma of the bile ducts, head of the pancreas, or ampulla of Vater or (uncommonly) a benign tumor. The bile ducts may be normal or dilated, depending on the cause, duration, and degree of obstruction. In acute obstruction caused by a stone, the bile ducts may not be dilated for some time, whereas in obstruction caused by cancer, the bile ducts are usually dilated.

FIG 42-23  Small stone (arrows) in the common bile duct on MR FIG 42-22  Common bile duct stone with rim calcification (arrow) mim-

cholangiogram.

icking a bull’s-eye.

B

A

FIG 42-24  Small stone in the common bile duct. Intraductal sonogram discloses a stone (A), but there is no stone visible on cholangiogram (B).

CHAPTER 42  Biliary Tract and Gallbladder

A

1201

B FIG 42-25  Small stones in the common bile duct (arrows) are visualized on MR cholangiogram (A) but not on cholangiogram (B).

FIG 42-26  Acute suppurative cholangitis caused by a stone originating from the gallbladder. The common bile duct is slightly dilated, and the wall is thick and enhanced as a concentric ring (arrow). The gallbladder contains a small stone and its wall is thick and enhanced owing to acute cholecystitis.

Obstruction of the bile ducts causes bile stasis and a predisposition to infection. Gram-negative organisms such as E. coli are usually responsible. It is believed that bacteria within the intestinal tract ascend through the ampulla of Vater; for this reason the disease is also called ascending cholangitis. Bacterial colonization may also occur through portal venous bacteremia. The combination of obstruction and colonization leads to cholangitis, but neither obstruction nor colonization by itself is sufficient to cause disease.223 Patients present with acute abdominal pain, fever, chills, and sometimes jaundice. In some patients with bile duct stones or carcinoma along the bile ducts, the initial symptoms and signs may be similar to those of acute suppurative cholangitis.

FIG 42-27  Suppurative cholangitis with a common bile duct stone in a patient with recurrent pyogenic cholangitis. Note concentric thickening of the bile duct wall with enhancement.

Other than bile duct dilatation, imaging finding are usually normal.14 The presence of purulent bile may result in increased attenuation of bile (greater than that of water) on CT, but it is usually not clearly depicted. Depending on the calcium content, stones may be clearly depicted as calcified high-attenuating foci or as noncalcified intraluminal material (see Figs. 42-19 and 42-20). The wall of the bile ducts may be thickened concentrically and diffusely, with dense contrast enhancement (Fig. 42-27; see Fig. 42-26); this is well delineated with thin-section CT.14 In contrast to periductalinfiltrating bile duct cancer, the thickness of the bile ducts is uniform and less than 1 mm. In patients with suppurative cholangitis caused by a cancer along the bile ducts, a nodular mass or periductalinfiltrating cholangiocarcinoma is depicted on US, CT (Fig. 42-28), and MRI.

1202

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-28  A and B, Acute suppurative cholangitis caused by carcinoma of the common hepatic duct (arrow in B). Note the dilated bile ducts and concentric thickening of the wall in A.

One complication of suppurative cholangitis is the occurrence of multiple intrahepatic abscesses, and there may be communication with the biliary tree. In a fulminating infection of gas-forming organisms, pneumobilia can be seen on CT.

Recurrent Pyogenic Cholangitis Incidence.  Recurrent pyogenic cholangitis, also known as Oriental cholangitis, Oriental cholangiohepatitis, or intrahepatic pigment stone disease, is characterized by recurrent attacks of acute bacterial cholangitis presenting with abdominal pain, fever, chills, and jaundice.33,64,89,269,299 A bile duct stone is an integral part of the disease, and the symptom complex repeats once or twice a year. The disease is prevalent in the Far East and Southeast Asia (thus its alternative names); it occurs infrequently in non-Asians. The stones in patients with recurrent pyogenic cholangitis are mainly bilirubinate pigment stones formed in the bile ducts, whereas the most common stones in Caucasian people are cholesterol stones in the CBD that originated from the gallbladder.297 Many patients with recurrent pyogenic cholangitis also have muddy stones or tiny gravel.89 Chronic infection of the biliary tree by C. sinensis or O. viverrini or inadvertent biliary Ascaris infection is a possible cause of the disease.209,297 A long-standing partial obstruction due to parasites in the bile ducts, bile stasis, increased secretion of mucin, or eggs or excreta of parasites predisposes the patient to form bile duct stones, with the eggs or debris being the stone’s nidus.209 Alternatively, portal septicemia due to poor eating habits and poor nutrition has been proposed as a cause of bile duct infection and stone formation.57,239 Pathology and imaging findings.  Because of repeated inflammation, the bile duct walls are thickened diffusely owing to inflammatory cell infiltration and fibrosis. Fibrosis causes the bile ducts to become rigid and straight, lacking gentle dichotomous branching (Fig. 42-29). In severe infection, adjacent hepatic parenchyma is involved, resulting in cholangiohepatitis.57,64,89,269 Sometimes this inflammation becomes suppurative and the bile ducts fill with pus. Larger IHDs and extrahepatic ducts become dilated because of a stricture of the duodenal papilla caused by mechanical irritation and inflammation due to the repeated passage of stones or sharp gravel.209 However, the small IHDs are not dilated because of ductal fibrosis. Thus the biliary tree has a “pruned tree” appearance because of the notable dilatation of the larger bile ducts, decreased arborization, and nonvisualization of the

FIG 42-29  Recurrent pyogenic cholangitis. ERCP shows markedly dilated intrahepatic bile ducts with rigidity, decreased arborization, rightangle branching, acute tapering, and the “arrowhead” sign. The extrahepatic duct is also dilated, with proximal narrowing due to segmental fibrous stricture (arrows). Note the large stones (arrowheads) in the distal common bile duct.

peripheral branches164,195,209,220,339 (Fig. 42-30; see Fig. 42-29). Owing to long-standing inflammation and fibrosis, peripheral bile ducts appear rigid and taper abruptly, producing an “arrowhead” appearance (see Fig. 42-29).275 Imaging findings include stones, a dilated bile duct, wall thickening, and duodenal papillary stenosis. Stones are visualized in 70% to 80%

CHAPTER 42  Biliary Tract and Gallbladder

FIG 42-30  Recurrent pyogenic cholangitis. The extrahepatic ducts and lobar and segmental bile ducts are dilated, but the peripheral ducts are obliterated rapidly. Note multiple small stones in the extrahepatic ducts. The gallbladder was resected.

of patients in the intrahepatic or extrahepatic ducts or both (see Fig. 42-30).41,208,209 For patients with intrahepatic stones, the lateral segment of the left hepatic lobe and posterior segment of the right hepatic lobe are frequent sites of stone lodging.89,209 The characteristic bile duct involvement patterns persist even when the stone passes into the duodenum. This is true because repeated passage of stones causes papillitis, fibrosis, and papillary stenosis. In some cases, bile duct dilatation is limited to the hepatic segmental or lobar bile ducts. This dilatation occurs as intrahepatic stones lodge at the branching points of large-caliber bile ducts, causing mechanical irritation, chronic inflammation, and fibrosis and a resultant stricture209,275 (Fig. 42-31). Frequent sites of focal stricture are the branching points of the medial and lateral segments of the left hepatic lobe (Fig. 42-32), posterior segmental duct of the right hepatic lobe, and bifurcation of the right and left hepatic lobes (see Fig. 42-29).164,209 With acute and chronic inflammation, the wall of the bile duct becomes thickened. On high-resolution contrast CT, the thickened wall is clearly visualized and enhanced (Fig. 42-33). The thickening of the bile duct in patients with recurrent pyogenic cholangitis is more severe than in those with acute suppurative cholangitis but less severe than in patients with periductal-infiltrating cholangiocarcinoma. During the stage of acute exacerbation, the periductal hepatic parenchyma is enhanced on contrast-enhanced CT, reflecting peri­ ductal hepatic parenchymal inflammation (cholangiohepatitis) (Fig. 42-34). Sometimes the hepatic lobe or segments are enhanced in the arterial phase, probably owing to severe cholangiohepatitis and arterial hyperemia.209 Hepatic abscesses occur as a complication (Fig. 42-35). In patients with stricture of the lobar or segmental bile ducts, the corresponding hepatic lobe or segment becomes atrophic because of longstanding obstruction of the bile ducts.209 Periportal inflammation and portal venous encroachment may play an additional role. Duodenal papillitis and papillary stenosis are an integral part of recurrent pyogenic cholangitis. Impaction of stones and repeated passage of small stones or sharp gravel through the papillary orifice cause irritation, inflammation, edema, fibrosis, and a resultant papillary stricture.91,254,272 A swollen duodenal papilla can be demonstrated on CT (Fig. 42-36) and MR cholangiography. Papillary stenosis is difficult to

1203

FIG 42-31  ERCP in a patient with common bile duct obstruction caused by a passed stone originating from the gallbladder shows simple dilatation.

FIG 42-32  Stones in the lateral segmental bile ducts with severe dilatation due to stricture (arrow).

demonstrate on sonography, CT, cholangiography, and MR cholangiography because the structure of the ampulla-papilla is small, and papillary stenosis caused by fibrosis cannot be visualized on imaging40 (Fig. 42-37). Complications.  Biliary cirrhosis is a possible complication and usually follows long-standing severe, repeated disease.89 Stones at the lower end of the CBD may cause acute pancreatitis89,204,294 and a choledochoduodenal fistula.89 Intrahepatic cholangiocarcinoma may develop in a stone-harboring bile duct because of long-standing inflammation, hyperplasia, and dysplasia.167,186,261 The reported prevalence ranges from 1.5% to 11%.167

Primary Sclerosing Cholangitis Incidence.  Primary sclerosing cholangitis (PSC) is a chronic cholestatic liver disease of unknown cause characterized by diffuse inflammation and fibrosis of the biliary tree. PSC is a progressive disease that can lead to obliteration of the bile ducts and subsequently biliary  cirrhosis; it may eventually progress to end-stage liver disease and

1204

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-33  Recurrent pyogenic cholangitis. A, Cholangiogram shows severe dilatation and acute peripheral tapering of the intrahepatic bile ducts with the “arrowhead” sign. There are multiple large stones in the extrahepatic duct. B, CT shows a dilated common bile duct and stones within it. Note the thickened wall of the common bile duct (arrows).

FIG 42-34  Recurrent pyogenic cholangiohepatitis. CT obtained during the arterial phase shows thickening and enhancement of the bile ducts (arrows) and stones in the common bile duct. Note the heterogeneous enhancement of the hepatic parenchyma and periductal enhancement.

cholangiocarcinoma. Liver transplantation is the only effective lifeextending treatment for advanced PSC. The disease is much more common in northern Europe and the United States than elsewhere, with the age-adjusted incidence in the United States being 1.25 per 100,000 men and 0.54 per 100,000 women10; it is exceedingly rare in Asian populations. Approximately 70% of the patients are male, and 70% have ulcerative colitis.

Pathology.  The pathology of sclerosing cholangitis is characterized by a fibrosing inflammation of the biliary tree resulting in diffuse thickening of the bile duct wall, which if severe enough leads to obstruction of the lumen.32,236,345 There are alternating areas of focal narrowing and ductal dilatation.94,197,227 On microscopic examination there is dense fibrosis, sparse mixed inflammatory cell infiltration, and a relatively intact epithelium. These findings are nonspecific, and the diagnosis cannot be based solely on a biopsy specimen.197 Because of periductal fibrosis, ductal dilatation is not prominent even with highdegree obstruction. PSC involves the intrahepatic and extrahepatic ducts. In 20% of patients, the disease is confined to the intrahepatic and hilar ducts. The gallbladder is involved in 40% of patients. Imaging findings.  Cholangiogram has long been the gold standard for the diagnosis of PSC. Multiple segmental strictures involving the intra- and extrahepatic bile ducts are hallmarks of the disease and characteristic findings (Fig. 42-38). The abnormality is not uniform along the bile ducts. A normal or less-involved segment alternating with strictures produces the classic “beaded” appearance (Fig. 42-39). The length of a stricture can range from 1 to 2 mm to several centimeters. When the peripheral bile ducts are obliterated, the bile ducts have a pruned-tree appearance (see Fig. 42-38). The bile duct wall is irregular, from a fine brush border to a coarse, shaggy, or nodular appearance (see Fig. 42-39). Sometimes there are small diverticula. Thus the combination of short focal strictures and dilatations, beading, pruning, and mural irregularities are typical for PSC. MRCP has merit in depicting bile duct involvement patterns of alternating strictures and dilatations of the entire biliary tree (see  Fig. 42-39).105,145 In addition to the advantage of noninvasiveness, MR cholangiography can demonstrate abnormal bile ducts proximal to  the completely obstructed segment that cannot be demonstrated by cholangiography.

CHAPTER 42  Biliary Tract and Gallbladder

A

1205

B FIG 42-35  A, Multiple liver abscesses (arrows) in a patient with recurrent pyogenic cholangitis. B, In addition to the liver abscesses (arrows), note the small stone in the dilated segmental bile duct (arrowhead).

A

B FIG 42-36  Axial (A) and coronal (B) CT shows a markedly enlarged ampulla of Vater containing a stone that is bulged into the duodenum (arrow in B).

Sonography and contrast-enhanced CT and MRI may show  thickening (Fig. 42-40) and contrast enhancement (Fig. 42-41) of the  bile duct wall. The thickness varies from 1 to 3 mm. Thus these  imaging techniques are helpful to determine the severity and extent of disease. In some instances, cholangiographic “beading” can be seen on  contrast-enhanced CT (see Fig. 42-39). Like MRCP, CT has the advantage of being able to visualize an obstructed segment that cannot be demonstrated on cholangiography. Other imaging findings include alterations of the hepatic parenchyma such as biliary cirrhosis, hepatic segmental atrophy and hypertrophy, portal hypertension, a dominant biliary stricture resulting in obstructive jaundice, calculi in the gallbladder and CBD (25%-30%), and a complicating cholangiocarcinoma (8%-15%).310 Clinical course.  There is an increased incidence of cholangiocarcinoma in patients with PSC.310 In a series of patients undergoing liver transplantation for PSC, 10% were found to have cholangiocarcinoma in their native livers.129 When the two diseases coexist, the ductal abnormality of PSC can easily mask the presence of cholangiocarci-

noma.231 Cholangiocarcinoma usually causes more complete obstruction than PSC (dominant stricture), and there may be markedly dilated bile ducts proximal to the bile duct segment involved by the cholangiocarcinoma. A rapidly progressing segmental stricture and proximal dilatation on serial images associated with clinical deterioration favor the presence of cholangiocarcinoma. In this regard, thin-section CT and MRI are helpful to assess the clinical course.

AIDS Cholangitis.  Inflammation of the bile ducts can occur in patients with acquired immunodeficiency syndrome (AIDS), owing to an opportunistic infection by cytomegalovirus, Mycobacterium aviumintracellulare, Candida albicans, Cryptosporidium species, and Klebsiella pneumoniae.35,289 Inflammation may involve the intrahepatic and extrahepatic ducts as well as the duodenal papillae, resulting in papillary stenosis. Cholangiogram shows irregularities and stricture of the intrahepatic and extrahepatic bile ducts with associated bile duct dilatation. Papillary stenosis often occurs in conjunction with proximal duct stricture. Tiny polypoid intraductal defects due to granulation

1206

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-37  Papillary stenosis due to repeated passage of stones. A, ERCP shows dilatation of the extrahepatic ducts. There is normal contraction of the sphincter of Oddi at the end of the common bile duct (arrow). B, CT shows a normal-sized duodenal papilla (arrow).

A

B FIG 42-38  Primary sclerosing cholangitis. A, ERCP shows multisegmental or diffuse narrowing and proximal dilatation of the intra- and extrahepatic bile ducts. Peripheral ducts have a “pruned tree” appearance. B, Contrast-enhanced CT shows diffuse thickening of the wall of the intra- and extrahepatic bile ducts (arrows). Also note thickening of the gallbladder wall.

CHAPTER 42  Biliary Tract and Gallbladder

1207

A

C

B

FIG 42-39  Primary sclerosing cholangitis. A, MR cholangiogram shows innumerable foci of focal or segmental narrowing or dilatation of the intrahepatic bile ducts, with a “beaded” appearance and obliteration of the extrahepatic duct (arrow). B and C, CT shows peripheral bile duct narrowing and dilatation, with scattered segments of peripheral duct dilatation. Note obliteration of the lumen of the extrahepatic duct due to severe thickening of the wall (arrow in C).

tissue can also be observed.62,241 CT and sonography may demonstrate wall thickening of the bile duct and gallbladder. A small nodule may be seen at the distal end of the CBD, reflecting inflammation and edema of the papilla of Vater.73

Parasitic Diseases

FIG 42-40  Primary sclerosing cholangitis. Sonogram shows the thickened wall of the extrahepatic bile ducts (arrows).

Biliary parasitic diseases include clonorchiasis, opisthorchiasis, fascioliasis, and ascariasis. These diseases are prevalent in East and Southeast Asia but rare in the United States and Western Europe. C. sinensis and O. viverrini are human liver flukes acquired through ingestion of raw freshwater fish harboring metacercariae; Fasciola hepatica is a sheep liver fluke and humans are infected accidentally. Adult flukes live in the bile ducts and produce a mechanical obstruction, mucosal inflammation, adenomatous hyperplasia, and periductal fibrosis (Fig. 42-42). Infection with C. sinensis and O. viverrini produces diffuse and uniform dilatation of the peripheral IHD, with no or minimal dilatation of the extrahepatic ducts and without a focal obstructing lesion along the bile ducts208,215 (Fig. 42-43). This occurs because the flukes

1208

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-41  A and B, Primary sclerosing cholangitis involving segmental intrahepatic bile ducts and extrahepatic ducts, with thickening of the bile duct wall and intense enhancement (arrow in B). The gallbladder wall is also thickened and enhanced.

FIG 42-42  Clonorchis sinensis in the bile duct. Photomicrograph from a rabbit liver experimentally infected with C. sinensis shows adult flukes in the bile duct. Note adenomatous hyperplasia of the mucosa (short arrows) and periductal fibrosis (long arrows). (Hematoxylin-eosin; ×40.)

reside in the small IHDs. This characteristic pattern of the biliary tree—apparent on cholangiography, CT, and MR cholangiography—  is not encountered in other bile duct diseases such as cholangitis,  bile duct stone, or tumor.48,49,55,71,72,208,215 Adult flukes in the bile ducts can be demonstrated on cholangiography (Fig. 42-44) and MR cholangiography (Fig. 42-45) as small, linear, elliptical or leaflike filling defects measuring 2 to 10 mm long.208,215 The worms and their aggregates in the bile duct lumen, in association with the inflammatory reaction,55,72,208,215 impede contrast opacification of the peripheral ducts (see Fig. 42-44). Because of long-standing inflammation and fibrosis, the bile duct wall is thickened. CT may show the enhancing wall of the bile ducts. The gallbladder may be distended. In heavy infection the branch ducts of the tail of the pancreas are dilated, as in the liver.216 Once infection occurs the flukes can live in the biliary tree for decades. Peripheral intrahepatic dilatation persists for a long time, even after the flukes are eradicated spontaneously or by the administration of anthelmintic drugs. Recurrent pyogenic cholangitis and bile duct

stone formation are frequent complications of clonorchiasis. Intraductal inflammation, increased mucin, and mechanical obstruction by the flukes and their eggs cause bile stasis, which provides a favorable environment for bacterial infection. The eggs and bacteria serve as nidi for pigment stone formation.208 A carcinoma may arise from the intrahepatic and extrahepatic bile ducts as a complication of long-standing fluke infection (Figs. 42-46 and 42-47). Chronic inflammatory changes and adenomatous hyperplasia of the biliary mucosal epithelium may transform into dysplasia and cholangiocarcinoma.46,53,319 F. hepatica is a trematode that infects cattle and sheep; humans are accidental hosts. Infection is caused by ingestion of raw aquatic plants contaminated with metacercariae, which later penetrate the intestinal wall and migrate through the peritoneal cavity to the liver. After penetrating Glisson’s capsule, the flukes migrate randomly through the liver parenchyma (hepatic stage) and enter the bile ducts, where they mature and release eggs (biliary stage). Hepatic fascioliasis manifests as clusters of microabscesses, usually in the subcapsular regions, arranged in a characteristic serpentine fashion resembling “tunnels and caves”36,121,270,156,281 (Fig. 42-48). A large cystlike necrotic lesion may also be observed. Imaging findings of biliary fascioliasis are more or less the same as for clonorchiasis, with bile duct dilatation and intraductal flukes appearing as echogenic foci on ultrasonograms and as filling defects on cholangiograms (Fig. 42-49) and MR cholangiograms. The intestinal roundworm Ascaris lumbricoides is 15 to 30 cm long and 3 to 5 mm thick and resides in the human small intestine. Occasionally worms migrate through the papilla of Vater into the biliary tree, resulting in bile duct obstruction, cholangitis, and pancreatitis. The long worm is seen as a filling defect on cholangiograms or MR cholangiograms137,191 (Fig. 42-50).

Bile Duct Strictures A benign bile duct stricture may affect the intrahepatic or extrahepatic ducts or both. Common causes of benign strictures are inflammatory and iatrogenic. PSC and recurrent pyogenic cholangitis are the most common causes of an inflammatory stricture. An iatrogenic bile duct injury is most often caused by surgery such as a cholecystectomy, CBD exploration, T-tube indwelling in the CBD, endoscopic papillotomy, or liver transplantation.150 A laparoscopic cholecystectomy predisposes to bile duct injury because the CBD may be mistaken for the cystic duct. Thermal injury to the bile duct can occur during radiofrequency

CHAPTER 42  Biliary Tract and Gallbladder

A

1209

B FIG 42-43  Mild clonorchiasis infection. A and B, CT shows mild dilatation of the peripheral intrahepatic bile ducts, without dilatation of the central duct (arrow in B).

FIG 42-44  Clonorchiasis. Endoscopic retrograde cholangiogram shows innumerable small elliptical filling defects in the bile ducts, representing adult Clonorchis sinensis. Filling of the peripheral duct is impeded by flukes (arrowheads). The intrahepatic ducts are dilated, but the extrahepatic bile ducts are not.

FIG 42-46  Clonorchiasis and cholangiocarcinoma of the common hepatic duct. Endoscopic retrograde cholangiogram shows markedly dilated intrahepatic bile ducts containing multiple elliptical filling defects due to adult flukes of Clonorchis sinensis. Arrows point to the segment of cholangiocarcinoma.

FIG 42-45  Severe clonorchiasis. MRI shows severe dilatation of the intrahepatic bile ducts up to the periphery of the liver. Adult flukes are visible in the bile ducts (arrows).

1210

PART II  CT and MR Imaging of the Whole Body

FIG 42-47  Peripheral cholangiocarcinoma (arrow) in a patient with clonorchiasis, which has caused dilatation of the peripheral bile ducts.

A

B FIG 42-50  Biliary ascariasis. Endoscopic retrograde cholangiogram FIG 42-48  Hepatic fascioliasis. CT shows multiple small, clustered, necrotic lesions in the subcapsular area arranged as “tunnels and caves.”

FIG 42-49  Biliary fascioliasis. Endoscopic retrograde cholangiogram shows two elliptical filling defects in the common bile duct representing adult Fasciola hepatica. (Courtesy Cheol Min Park, MD, Korea University Medical School, Seoul, Republic of Korea.)

(A) and MRCP (B) show a long convoluted filling defect in the extrahepatic ducts representing adult Ascaris lumbricoides. (Courtesy So Yeon Kim, MD, Ulsan University Medical School, Seoul, Republic of Korea.)

ablation for hepatic tumors173 (Fig. 42-51). Blunt abdominal trauma may result in injury to the duodenum and extrahepatic bile ducts. For a biliary stricture with a benign cause, the stenotic segment shows minimal wall thickening, making it difficult to see on CT or MRI (Fig. 42-52). The stenotic segment is abrupt and short, measuring less than 1 cm. In contrast, for periductal-infiltrating cholangiocarcinoma, the involved bile duct wall is thicker than 2 mm and enhances intensely after contrast administration. In PSC, there are multiple sites of stricture and normal or dilated segments, resulting in a beaded appearance. In recurrent pyogenic cholangitis, branching points of lobar and segmental bile ducts are  the common sites of stricture; this is where the stones tend to lodge, causing irritation and inflammation at the branching points, resulting in fibrosis and stricture. Common sites are the lateral segmental bile duct of the left hepatic lobe (Fig. 42-53), posterior segmental bile duct of the right hepatic lobe (Fig. 42-54), and bifurcation of the right and left hepatic ducts (see Figs. 42-11 and 42-29). A focal stricture of the CBD occurs in approximately 0.5% of cholecystectomy patients.128 This complication occurs when the cystic duct is clamped and ligated and the CBD is inadvertently incorporated into

CHAPTER 42  Biliary Tract and Gallbladder

1211

the ligatures, causing injury.66 Clinical manifestations including jaundice may appear within a week, months, or years, depending on the severity of the bile duct damage. CT reveals severely dilated IHDs, with the degree of dilatation depending on the severity of the bile duct obstruction. At the hepatic hilum there is an abrupt transition from the dilated bile ducts to the normal extrahepatic ducts, where there is a short segmental tight stricture (Fig. 42-55). A coronal reconstruction image better shows the stenotic segment. A biliary stricture caused by injury to the anastomotic site during liver transplantation typically occurs at the proximal part of the extrahepatic duct. The degree of stenosis varies from mild stenosis without bile flow impairment  (Fig. 42-56) to a tight stricture and complete obstruction of bile flow (Fig. 42-57). A biliary stricture caused by blunt abdominal trauma occurs in  the suprapancreatic portion of the CBD.352 The CBD is susceptible to disruption from deceleration injuries, usually at the level of the pancreaticoduodenal junction, where it is relatively fixed compared with

the more proximal duct. A delayed common duct stricture has been reported after blunt abdominal trauma. The injury results in a short segmental tight stricture with concentric or eccentric involvement and proximal dilatation.

FIG 42-51  Bile duct stricture caused by thermal injury during radiofrequency ablation for hepatocellular carcinoma. CT shows severe dilatation of the posterior segmental bile ducts abutting the necrotic nodule.

FIG 42-53  Tight stricture (arrow) at the left hepatic duct in a patient with recurrent pyogenic cholangitis. Note stones in the left hepatic duct proximal to the stricture site.

A

Neoplasms of the Biliary System Benign Tumors of the Bile Ducts

Bile duct hamartoma.  Bile duct hamartoma, also known as von Meyenburg’s complex, is a focal disorderly collection of bile ducts and ductules surrounded by abundant fibrous stroma.60,131 There is usually a centrally located small cystic lesion lined by a single layer of biliary epithelium that does not communicate with the biliary tree.256 The lesions are 0.1 to 1.5 cm in diameter and are scattered diffusely in the liver parenchyma. CT shows multiple or innumerable hypoattenuating cystlike hepatic nodules, typically less than 1.5 cm in diameter, occurring throughout both lobes of the liver256 (Fig. 42-58). MRI shows tiny

B FIG 42-52  Bile duct stricture due to recurrent pyogenic cholangitis and fibrosis. A, Cholangiogram shows a tight stricture at the bifurcation (arrow). Note several stones in the extrahepatic duct. B, CT shows tight narrowing and wall thickening at the bifurcation (arrow).

1212

PART II  CT and MR Imaging of the Whole Body

hypointense lesions on T1-weighted images and strongly hyperintense lesions on T2-weighted images (Fig. 42-59). Contrast-enhanced CT or MRI may show homogeneous enhancement of the lesions256,300,309 (see Fig. 42-59). Bile duct adenoma.  A solitary adenoma appears as a wellcircumscribed wedgelike mass less than 1 cm in diameter. The gallbladder is the most frequent site; these tumors are rare along the bile ducts. A villous adenoma can occur at the duodenal papilla (Fig. 42-60). Biliary papillomatosis.  Biliary papillomatosis is a papillary or villous tumor in the intrahepatic or extrahepatic bile ducts showing cytologic and histologic atypia, though not enough for a diagnosis of malignancy.268 Papillary fragile tumors are seen on the inner surface of the biliary tree; there may be multiple tumors or they may spread diffusely along the biliary tree261 (Figs. 42-61 and 42-62). The tumor does not cause biliary obstruction; although it spreads superficially for a

FIG 42-54  Stricture of bile duct in recurrent pyogenic cholangitis. The posterior segmental branch of the right hepatic duct is missing (“missing duct” sign; long arrow). Note two stones in the anterior segmental bile duct (short arrows).

A

varying length, it produces a large amount of mucus, resulting in nonobstructive dilatation of the entire biliary tree.7 This tumor has great potential for malignant transformation (see Fig. 42-61).261 Biliary cystadenoma and cystadenocarcinoma.  Benign and malignant cystic tumors of bile duct origin can occur in the IHDs of the liver or rarely in the extrahepatic ducts. These tumors occur in adults, with a strong middle-aged female predominance. The usually single tumor is covered by a fibrous capsule and is unilocular or  multilocular, containing mucinous or serous fluid.18 A cystadenoma is lined by a single layer of cuboidal or tall columnar mucin-producing cells, and a cystadenocarcinoma is lined by intestinal-type mucosa (including goblet cells and Paneth cells) associated with varying degrees of atypia and mitotic activity.18 Benign and malignant components may coexist. There is usually no communication between the cystic tumor and the bile ducts. A biliary cystadenoma appears as a solitary cystic mass possessing a well-defined fibrous capsule containing clear fluid. The fibrous capsule may not be evident when the mass is covered by the liver

FIG 42-56  Mild bile duct stricture (arrow) at the site of anastomosis after liver transplantation. There are many stones and debris in the bile ducts proximal to the stricture.

B FIG 42-55  Iatrogenic stricture of the common hepatic duct after cholecystectomy. A, CT shows intrahepatic bile duct dilatation and abrupt stricture at the confluence of the rigid hepatic duct (arrow). B, Cholangiogram shows complete obstruction at the bifurcation (arrow).

CHAPTER 42  Biliary Tract and Gallbladder

A

1213

B

C FIG 42-57  Bile duct stricture at the site of anastomosis in a liver transplant patient. A and B, CT shows fibrous thickening of the anastomosis site (arrow) between the dilated intrahepatic ducts and the extrahepatic duct. C, ERCP shows the tight stricture at the anastomosis site (arrow).

FIG 42-58  Biliary hamartomas. CT shows many small cystic lesions scattered throughout the liver parenchyma.

parenchyma. Usually the cystic tumor is multiloculated, divided by thin uniform curvilinear septa, and measures 1 mm or less. Sometimes it is barely visible30,31,52,98,101,189,256,259 (Fig. 42-63). There may be multiple daughter cysts inside (Fig. 42-64). When the septa are very thin, sonography is better than CT or MRI for depicting them. In some cystadenomas, there is no septum or daughter cyst and differentiation from a simple hepatic cyst is impossible (Fig. 42-65). There may be calcium deposits in the fibrous capsule or septa. The contents of the cystic tumor can be serous or mucinous fluid, thus attenuation on CT and signal intensity on MRI may be different depending on the contents of the individual loculations (see Fig. 42-64). Papillary areas or polypoid mural nodules in the cyst wall or septa have been reported256 but are rare with cystadenoma. A biliary cystadenocarcinoma presents as a single cystic mass covered by a thick capsule; it contains multiple thick septa, mural nodules, and polypoid or fungating tumors attached to the cyst wall or septa213 (Fig. 42-66). Coarse thick septal or capsular calcifications are more frequent in cystadenocarcinoma189 than in cystadenoma

1214

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 42-59  Biliary hamartomas (von Meyenburg’s complex). A, T2-weighted image shows innumerable tiny high-signal-intensity nodules and cystic lesions in the liver. On T1-weighted (B) and gadolinium-enhanced (C) images, the number of lesions is markedly decreased, suggesting that many of the nodules are enhanced and isoattenuating to the liver parenchyma.

FIG 42-60  Adenoma of the ampulla of Vater. CT shows a 1-cm mass (arrow) at the ampulla of Vater that has bulged into the duodenum.

FIG 42-61  Biliary papillomatosis. MRI shows innumerable tiny papillary tumors in the surface of the dilated intrahepatic bile ducts (arrows). Foci of carcinoma in situ were found on pathologic examination.

CHAPTER 42  Biliary Tract and Gallbladder

1215

(Fig. 42-67). The septa and mural nodules are typically enhanced on contrast-enhanced CT or MRI. Thin or thick septa are clearly visible on sonography. Various CT attenuation or MR signal intensities are seen on both T1- and T2-weighted images, depending on the presence of solid components, hemorrhage, and protein content.256 A cystadenoma arising from the extrahepatic ducts is rare, accounting for less than 10% of biliary cystadenomas. The most common presentation is biliary obstruction and jaundice. As with an intrahepatic cystadenoma, an extrahepatic cystadenoma or cystadenocarcinoma appears as a uniloculated or multiloculated cystic mass with internal septa and mural nodules in the extrahepatic ducts180,273 (Fig. 42-68). A choledochal cyst, especially a choledochal diverticulum or choledochocele, can be differentiated based on the presence of septa or a multiloculated shape. There is no communication between the cystic tumor and the bile duct.273

Other benign tumors.  A plexiform neurofibroma may involve the biliary tract, forming a branching mass along the intrahepatic and extrahepatic bile ducts, surrounding the bile ducts and portal  veins, or sometimes forming an intraductal mass159,206 (Fig. 42-69). Hamartomas, granulosa cell tumors, heterotopic gastric or pancreatic mucosal rests, and adenomyomas (Fig. 42-70) may appear as intraluminal or intramural masses.163 Traumatic neuroma of the bile duct is not a true neoplasm but a reactive proliferation of pericholangial nerve tissue caused by cholecystectomy (Fig. 42-71), liver transplantation, or blunt trauma, resulting in bile duct stricture and proximal dilatation.134,244,304

FIG 42-62  Biliary papillomatosis. Cholangiogram shows multiple small

FIG 42-63  Biliary cystadenoma. CT shows a large oval cyst containing

flat papillary tumors along the proximal extrahepatic duct.

barely visible septa and small daughter cysts.

A

Malignant Tumors of the Bile Ducts Cholangiocarcinoma.  Cholangiocarcinoma is the most common malignant tumor arising from the epithelium of the bile ducts. It is much less common than hepatocellular carcinoma, accounting for 5% to 30% of all primary hepatic malignancies.290 The peak age is in the sixth to seventh decades. Hilar cholangiocarcinoma is the most common lesion, accounting for 53% to 67% of cases, with peripheral intrahepatic cholangiocarcinoma accounting for 6% to 25% of cases

B FIG 42-64  Biliary cystadenoma. A and B, CT shows curvilinear thin septa and daughter cysts. The attenuation value of cysts differs depending on the hemorrhage and protein content.

1216

PART II  CT and MR Imaging of the Whole Body

FIG 42-65  Biliary cystadenoma. CT shows a unilocular cyst with compression of a segmental bile duct.

and extrahepatic cholangiocarcinoma accounting for 22% to 27% of cases.262 The majority of cholangiocarcinomas are adenocarcinomas, regardless of location.151,340 This tumor has a tendency to spread along the bile duct walls,17 and periductal tissue invasion is common. Mucin stains are nearly always positive, with mucin production often being abundant.17,260,340 Predisposing conditions include PSC, biliary parasitic diseases such as long-standing C. sinensis or O. viverrini infection, recurrent pyogenic cholangitis with intrahepatic stones, Thorotrast exposure, and congenital biliary anomalies such as choledochal cyst and Caroli’s disease.17,151 Based on growth characteristics, cholangiocarcinomas are classified as mass-forming, periductal-infiltrating, and intraductal-growing types210,218,224 (Figs. 42-72 and 42-73). A mass-forming cholangiocarcinoma arises from the mucosa of the bile ducts, grows into the lumen, and invades and penetrates the bile duct wall; it thus grows three dimensionally to a certain size. As the intraluminal tumor occludes the lumen, the bile duct is obstructed. In contrast to the mass-forming type, a periductal-infiltrating cholangiocarcinoma tends to spread along the bile duct wall and periductal

A

B

C

D FIG 42-66  Biliary cystadenocarcinoma. CT (A), MRIs (B and C), and a sonogram (D) show a large unilocular cyst containing thick septa and multiple daughter cysts and solid components. Note the thick irregular capsule.

CHAPTER 42  Biliary Tract and Gallbladder

A

B

C FIG 42-67  Biliary cystadenocarcinoma. Sonogram (A), CT (B), and MRI (C) show an oval unilocular cyst containing a solid mass with an irregular surface. Note the punctate calcific focus and thick irregular wall (arrows in B).

1217

1218

A

PART II  CT and MR Imaging of the Whole Body

B

C FIG 42-68  Cystadenoma of the extrahepatic duct. Sonogram (A), MR cholangiogram (B), and cholangiogram (C) show an ovoid cystic lesion producing balloon-like expansion of the extrahepatic duct. Note the thin wall between the cyst and the bile duct (arrow in A); there is no communication between them. (Courtesy Dong Ho Lee, MD, Kyung Hee University Medical School, Seoul, Republic of Korea.)

CHAPTER 42  Biliary Tract and Gallbladder

A

B FIG 42-69  Neurofibroma of the bile ducts in a patient with neurofibromatosis. CT (A) and MR cholangiogram (B) show an intraductal mass (arrow) causing obstruction of the right and left hepatic ducts. (Courtesy Jun Woo Lee, MD, Pusan National University Medical School, Pusan, Republic of Korea.)

A

B FIG 42-70  Adenomyoma of the distal common bile duct. A, CT shows a well-marginated mass in the ampullary region (arrow). B, Cholangiogram shows a mass between the common bile duct and duodenum (arrows).

A

B FIG 42-71  Traumatic neuroma of the common bile duct after cholecystectomy. Coronal CT (A) and MR cholangiogram (B) show an eccentric mass in the common bile duct at the site of resection of the cystic duct (arrow). (Courtesy Yong Yeon Jeong, MD, Chonnam University Medical School, Guangju, Republic of Korea.)

1219

1220

PART II  CT and MR Imaging of the Whole Body

tissue.210,218,224,260,340 Occasionally a substantial tumor extends beneath the intact mucosal epithelium.340 Thus the tumor grows longitudinally and extends along the axis of the bile duct like the branch of a tree.217 The bile duct lumen is diffusely narrowed or completely obliterated. An extrahepatic periductal-infiltrating tumor is characterized 

CHOLANGIOCARCINOMA

A

by concentric thickening of the bile duct wall for a variable length17,210,217,218,340; it does not produce a sizable mass. Intraductal-growing cholangiocarcinoma grows into the lumen then spreads superficially along the mucosal layer.122 Tumor cells are confined to the mucosal layer and do not invade deep to the  submucosal layer. In contrast to mass-forming and periductalinfiltrating cholangiocarcinomas, the bile duct lumen is not completely obstructed.210,218 The prognosis of mass-forming and periductal-infiltrating chol­ angiocarcinomas is generally unfavorable; intraductal-growing  chol­angiocarcinoma has a much better prognosis after surgical resection.140,210,295,349 Precise information about tumor location, extent, growth pattern, and staging is mandatory for optimal treatment planning and for a prediction of prognosis. Surgical resection should be individually tailored depending on the morphologic type and the stage of the tumor.

Peripheral intrahepatic and hilar cholangiocarcinomas. 

B

C FIG 42-72  Morphologic classification of cholangiocarcinoma. Tubules represent bile ducts. Drawings show mass-forming (A), periductalinfiltrating (B), and intraductal-growing (C) cholangiocarcinomas. (From Kimura K, et al: Association of gallbladder carcinoma and anomalous pancreaticobiliary ductal union. Gastroenterology 89:1258–1265, 1985.)

Intrahepatic cholangiocarcinomas include peripheral and hilar cholangiocarcinomas. The vast majority of peripheral cholangiocarci­ nomas are the mass-forming type, whereas the majority of hilar  cholangiocarcinomas are the periductal-infiltrating type.50 Clinically, intrahepatic peripheral cholangiocarcinoma usually presents with nonspecific abdominal symptoms, whereas hilar cholangiocarcinoma presents with jaundice because of large bile duct obstruction. Mass-forming type.  A mass-forming intrahepatic cholangiocarcinoma is usually sizable, and the tumor does not cause clinical symptoms in its early stages. Grossly the tumor is firm and whitish gray because of abundant fibrous stroma.17 The margin is well circumscribed and wavy or lobulated. There may be central necrosis.290 Satellite tumors, especially around the main tumor, are common. These satellite tumors occur owing to the tumor’s propensity to invade the adjacent peripheral branches of the portal vein.17,210,217,224 The most common appearance of a peripheral intrahepatic cholangiocarcinoma on imaging is a well-defined single hypovascular mass with wavy borders50,54,132,140,210,218,340 (Fig. 42-74). Owing to intrahepatic metastasis via the portal vein, satellite or daughter nodules are frequent

A

B

C

D

FIG 42-73  Mode of spread of cholangiocarcinoma. Drawings show mucosal (A), mass-forming (B), periductalinfiltrating (C), and intraductal-growing (D) cholangiocarcinoma.

CHAPTER 42  Biliary Tract and Gallbladder

A

1221

B FIG 42-74  Peripheral cholangiocarcinoma. A, CT during the arterial phase shows a lobulated mass with a well-enhancing periphery and a heterogeneously enhancing center. Note a small satellite tumor (arrow). B, CT during the delayed phase shows the same mass, but it appears much smaller because of gradual enhancement from the periphery.

FIG

42-75  Multiple

intrahepatic

cholangiocarcinomas

due

to

metastases.

and vary in size165,210 (Fig. 42-75). On contrast-enhanced CT or MRI, thin or thick rimlike enhancement is frequently seen around the periphery of the tumor on arterial-phase images, and there is gradual centripetal enhancement on delayed-phase images132,165,316 (Fig. 42-76; see Figs. 42-74 and 42-75). The entire mass may be enhanced only on delayed-phase images some hours after contrast administration (Fig. 42-77). These findings reflect the nature of the tumor, which is mainly desmoplastic. Depending on the cellular and fibrous components, the central part of the mass may be enhanced homogeneously or heterogeneously. Sometimes there is septum-like enhancement.165 Occasionally the hepatic vein traverses the central part of the mass undisturbed.165 Capsular retraction is relatively frequent (see Fig. 42-76).122,165 Bile duct dilatation peripheral to the tumor is common. However, the bile ducts may not be dilated even if the tumor arises from one of the IHDs.165,210 On MRI the tumor is hypointense on T1-weighted images and hyperintense on T2-weighted images (see Fig. 42-76), without a capsule. Central hypointensity may be seen on T2-weighted images,

reflecting fibrosis; this is helpful in making the distinction from a metastatic tumor.47,50,119 On dynamic MRI the enhancement pattern is the same as on dynamic CT. Hilar and peripheral mass-forming cholangiocarcinomas may appear the same on imaging, except for the presence of IHD dilatation.51,317 Most hilar cholangiocarcinomas result in obstruction of the right and left hepatic ducts, and there may be a mass that overrides and separates the bile ducts.51,123,317 Periductal-infiltrating type.  Periductal-infiltrating cholangiocarcinoma grows along the bile ducts and produces diffuse thickening of them (Fig. 42-78); it is therefore difficult or impossible to depict the discrete tumor mass.50,51,120,123,210,218,317 The involved bile duct becomes narrow for a varying length, resulting in eventual obstruction. The majority of hilar cholangiocarcinomas are of the periductal-infiltrating type pathologically, but this type of lesion is not commonly seen in the periphery of the liver. On imaging, the involved bile ducts are obliterated or diffusely narrow, whereas the bile ducts proximal to the cholangiocarcinoma  are dilated210,218 (Fig. 42-79). Nonunion of the right and left hepatic ducts with or without a visibly thickened wall is a typical finding of  an infiltrating hilar cholangiocarcinoma (see Fig. 42-79).51,123,317 On cholangiography the lumen may be completely obstructed or markedly narrowed. A stringlike severely narrowed bile duct may be visualized.120 Hilar cholangiocarcinoma is often associated with lobar or segmental hepatic parenchymal atrophy, probably because of portal vein invasion and occlusion and decreased portal flow or because of long-standing dilatation of the bile ducts and diversion of the portal flow.34,323 There are combined mass-forming and periductal-infiltrating cholangiocarcinomas353 (Fig. 42-80). Intraductal-growing type.  Intraductal tumors are usually nodular or castlike, often spreading superficially along the mucosal surface and resulting in multiple tumors (papillomatosis or papillary carcinomatosis) along varying segments of the bile ducts.122,174,200,210,218,219,261 The intraductal-growing type accounts for 8% to 18% of cholangiocarcinomas.200 The majority of intraductal-growing cholangiocarcinomas are papillary adenocarcinomas,174,200,219,353 comprising innumerable frondlike infoldings of the proliferated columnar epithelial cells and slender fibrovascular cores.219 Infrequently an intraductal tumor may be papillotubular or purely tubular. The bile ducts are dilated

1222

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 42-76  MRI of peripheral cholangiocarcinoma. There is a well-defined lobulated mass with a dimple at the surface (arrow). The mass is of low signal intensity on the T1-weighted image (A) and of high signal intensity on the T2-weighted image (B). The central part is enhanced in the equilibrium phase (C), reflecting the presence of massive fibrosis.

CHAPTER 42  Biliary Tract and Gallbladder

A

B

C

D

1223

FIG 42-77  Contrast enhancement of peripheral cholangiocarcinoma. T1-weighted images before gadolinium administration (A) and immediately (B), 5 minutes (C), and 3 hours (D) after gadolinium injection. In the arterial phase the peripheral part is enhanced. In the delayed phase the central part is gradually enhanced and eventually greatly enhanced.

FIG

42-78  Periductal-infiltrating intrahepatic cholangiocarcinoma. Cancer has spread along the intrahepatic bile ducts (arrows).

secondary to incomplete obstruction by tumor, by sloughed tumor debris, or by an excessive amount of mucin.174,219 The lobar or segmental bile ducts are dilated diffusely, somewhat reminiscent of bronchiectasis in the lung, or aneurysmally, depending on bile duct obstruction and the amount of mucin production.122 On imaging, the bile ducts of the involved hepatic segment or lobe are dilated.122 An intraductal tumor can appear as a polypoid mass in the lumen of the dilated bile ducts (Fig. 42-81). Frequently intraductal tumors fill multiple segmental or subsegmental bile ducts. The mass is confined within the bile ducts, leaving the wall of the bile duct intact. The outer margin of the thickened bile duct wall is smooth and clear. The tumor may not be seen on imaging when it is small and isoattenuating to the adjacent hepatic parenchyma or when the complex orientation of the dilated bile ducts obscures the presence of the mass. On ERCP or percutaneous transhepatic cholangiography, the involved biliary tree is dilated secondary to partial obstruction and there are filling defects due to the papillary tumors. There may be fine irregularities, a velvety or serrated contour along the bile ducts, representing the papillary surface of the tumor.174,219

1224

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-79  A and B, Periductal-infiltrating cholangiocarcinoma involving the intra- and extrahepatic ducts. Note the ill-defined mass involving the right and left hepatic ducts (arrows in A) and the encircling and thickening of the extrahepatic ducts (arrows in B), obliterating the lumen of the bile duct.

B

A

FIG 42-80  A and B, Combined mass-forming and periductal-infiltrating cholangiocarcinoma showing a large ill-defined irregular mass (arrows) involving the hilum of the liver and wall thickening along the extrahepatic duct.

FIG 42-81  Intraductal-growing papillary neoplasm. CT shows an intraductal polypoid mass (arrow) with bile duct dilatation.

Extrahepatic cholangiocarcinoma.  Cholangiocarcinoma can develop in any part of the extrahepatic duct, with 50% to 75% of reported cases occurring in the upper third of the duct, including the hepatic hilum, 10% to 25% of cases occurring in the middle third of the duct, and 10% to 20% of cases occurring in the lower third of the duct.4,84 Approximately 95% of patients have extrahepatic bile duct obstruction at the time of diagnosis.79 Involvement of the periductal lymphatics, nerves, and perineural tissue is frequent.17 Infiltration to the adjacent arteries and veins and lymph node metastasis are relatively common. Mass-forming or nodular type.  The tumor forms a nodule that is usually 1 to 2 cm in diameter210,217,218 (Fig. 42-82). The tumor obstructs the bile duct lumen, penetrates the wall, and invades the periductal tissue.340 The tumor mass is usually small at the time of presentation because the bile ducts are occluded in the early stage of the disease, and patients present with jaundice. Approximately 25%  of extrahepatic cholangiocarcinomas are of this type. A mass-forming extrahepatic cholangiocarcinoma can be easily detected on imaging because the tumor occludes the bile duct, resulting in dilated proximal

CHAPTER 42  Biliary Tract and Gallbladder

A

1225

B FIG 42-82  Nodular or mass-forming cholangiocarcinoma of the distal common bile duct (arrow) on CT (A) and cholangiogram (B).

A

B FIG 42-83  Periductal-infiltrating cholangiocarcinoma. Axial (A) and coronal (B) CT images show concentric thickening of the common hepatic duct (arrow in A) involving the upper half of the extrahepatic duct (arrow in B).

bile ducts.13,123 However, because the mass is usually small, images should be carefully scrutinized. Periductal-infiltrating type.  This type of tumor constitutes more than 60% of extrahepatic cholangiocarcinomas. The tumor appears as a focal or segmental concentric thickening of the extrahepatic ducts, with almost complete obstruction of the lumen.210,217,340 The thickened wall (measuring up to 1 cm) is firm and grayish white.340 The extent of the tumor varies, ranging from 0.5 to 6 cm in length; it may involve all the extrahepatic ducts and extend proximally as far as the IHDs. On CT or MRI, the thickened bile ducts can be visualized as an enhancing ring or spot (Fig. 42-83). It is often difficult to visualize the lesion on CT or MRI because of the absence of distinct tumor formation; these imaging studies may show only focal or diffuse thickening of the bile duct.13,210,217,218 At the site of bile duct obstruction, the tumor border can be demonstrated as a symmetrically or asymmetrically thickened bile duct wall, constituting a transition zone.298 On cholangiography or MR cholangiography, the involved segment may not be

opacified in cases of complete obstruction, or it may appear to be stringlike when the lumen is not completely obstructed210 (Fig. 42-84). Differentiation from acute or recurrent pyogenic cholangitis may be problematic. The involved bile duct wall is thicker in cases of cholangiocarcinoma (usually >1 mm), whereas in cholangitis the bile duct wall is thinner than 1 mm. Furthermore, the involved segment of the bile duct is narrowed or obliterated in cholangiocarcinoma but normal or dilated in cholangitis. Intraductal-growing type.  The intraductal tumor may be polypoid, sessile, or superficially spreading along the lumen.174,210,217,218 There may also be multiple discrete tumors (papillary cholangiocarcinomatosis) along the inner surface of the bile ducts. Usually the tumor is limited to the mucosa and invades the wall and surrounding tissue in the very late phase.210,217,218 An intraductal papillary cholangiocarcinoma is friable and sloughs easily by touching or rubbing at the time of surgery, or it may slough spontaneously and occlude the bile ducts, simulating a bile duct stone214 (Fig. 42-85). The bile ducts are dilated and incompletely obstructed by the tumor itself or by viscous mucin.

1226

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-84  Periductal-infiltrating cholangiocarcinoma. MR cholangiogram (A) and source image (B) show segmental narrowing of the proximal extrahepatic duct (arrows in B).

A B

C FIG 42-85  Intraductal-growing cholangiocarcinoma of the extrahepatic duct. Axial (A) and coronal (B) CT images show a large castlike intraductal tumor (arrows). ERCP (C) shows a large intraductal tumor with an irregular papillary surface.

CHAPTER 42  Biliary Tract and Gallbladder On imaging, the bile ducts proximal to the tumor are dilated; the degree of dilatation depends on the degree of obstruction and the amount of mucin. The intraductal tumor is depicted as a sessile or castlike mass on sonography, CT, or MR cholangiography (see Fig. 42-85).174,214,219,221 The tumor is usually small and flat but may occasionally be large. The tumor tends to spread superficially along the lumen for a variable length and sometimes implants along the inner surface of the bile ducts, resulting in multiple discrete tumors.179,210,218,220 Although an intraductal tumor may be large, elongated, or castlike in shape, the tumor-harboring bile duct is typically not completely obstructed; bile flow is maintained through the space between the tumor surface and the bile duct wall (see Fig. 42-85).214,221 Cholangiography shows a serrated or velvety surface of the intraductal tumor and a smooth normal bile duct wall (see Fig. 42-85).221 Early bile duct carcinoma, which is defined as carcinoma limited to the mucosa or fibromuscular layer of the bile duct, may have an appearance similar to that of intraductal-growing tumor. On CT the bile ducts harboring the tumor mass are clearly delineated as a sharply defined outer margin or a thin enhancing ring that may be more densely enhanced than the tumor mass.212 The thickened fibromuscular layer of the tumor-bearing bile duct wall seems to be responsible for this appearance. The periductal fat is clear and not infiltrated. If

the tumor is resected, the patient has an excellent prognosis; cure rates range from 80% to 100%.38,212

Intraductal papillary mucinous tumors of the bile ducts.  Some papillary tumors of the bile ducts produce a large amount of mucin122,201,219,353 and the mucin may impede the flow of bile, resulting in obstructive jaundice. Endoscopy may show mucin protruding from the patulous orifice of the duodenal papilla. This tumor has a striking similarity to an intraductal papillary tumor of the pancreas in terms of its histopathology, production of excessive mucin, and pathophysiology.170,271 The bile ducts show severe dilatation; bile ducts both proximal and distal to the tumor are dilated because the viscous mucin does not easily pass the papilla of Vater. On imaging the tumor may appear as a nodular or castlike mass filling the bile ducts for a varying length (Fig. 42-86). Some intrahepatic papillary mucinous tumors produce aneurysmlike cystic dilatation of the bile ducts,122,210,213,219,360 which harbor multiple fungating intraductal papillary tumors (Fig. 42-87). Some of   the involved bile ducts dilate cystically, and others dilate diffusely and proportionally. Rarely a biliary intraductal papillary neoplasm presents only as dilatation of the lobar or segmental bile ducts, with no mass visible on imaging or evident on pathology studies.211 The bile ducts are dilated

A

B

1227

C FIG 42-86  A and B, Intraductal papillary mucinous adenocarcinoma (arrow) in the dilated intrahepatic ducts. The extrahepatic ducts are markedly dilated with mucin. C, Cholangiogram shows a large amount of mucin in the extrahepatic duct (arrows).

1228

PART II  CT and MR Imaging of the Whole Body

FIG 42-87  Cystic intraductal papillary mucinous adenocarcinoma containing innumerable small papillary tumors. Note dilated intrahepatic bile ducts and communication with the large cystic tumor.

and lined by papillary mucinous epithelial cells with a varying degree of dysplasia and malignant change. Because of abundant mucin production by the lining cells, the bile ducts are markedly dilated and the involved hepatic parenchyma becomes markedly atrophic. In severe cases the bile ducts are crowded and the hepatic parenchyma is virtually absent (Fig. 42-88).

Imaging techniques for diagnosis and staging of cholangiocarcinoma.  Surgical resection with complete removal of all cancerous tissue offers patients the only chance for cure or long-term survival. Advances in surgical techniques and devices have improved the results of aggressive surgery for cholangiocarcinoma.152,203,207,229,230,257 Advanced cholangiocarcinoma that was once considered inoperable has now become resectable. The criteria for tumor resectability are constantly changing. Accurate preoperative staging and in-depth knowledge of the anatomy of the liver and bile duct are prerequisites for curative resection. With improved imaging techniques, detailed information about the pathologic anatomy is now available for preoperative planning.  In the preoperative assessment of an intrahepatic cholangiocarcinoma, the primary concern is whether an apparently tumor-free segment of the right and left lobe is involved, thus permitting the obviously involved segment to be excised.50,261 For hilar cholangiocarcinoma the preoperative evaluation must address the extent of tumor along the bile ducts (Fig. 42-89). In periductal-type hilar or extrahepatic cholangiocarcinoma, information about the extent of tumor along the bile ducts (including the cystic duct), direct invasion into the adjacent arteries and portal vein, lymph node metastasis, and direct invasion of or metastasis to the liver is essential (Figs. 42-90 and 42-91). In an intraductal tumor, which frequently spreads superficially along the mucosal surface, the true extent of the tumor may be difficult to determine even with a high-quality cholangiogram (Fig. 42-92). CT or MRI can demonstrate thickening of the bile ducts, but the thickening might be caused by chronic obstruction without tumor involvement. Therefore cholangioscopy and a biopsy may be necessary before surgical resection.201 MDCT has excellent image quality, providing valuable information about the pathologic anatomy of the biliary tree, other abdominal organs, and peritoneal cavity. The use of advanced image processing, including maximum intensity projection and multiplanar reconstruction (especially coronal or sagittal reformatted images), allows superior

FIG 42-88  Intraductal papillary mucinous adenoma with tumor cells lining the bile ducts but no visible mass formation. CT shows severely dilated left intrahepatic bile ducts without a visible mass. Liver parenchyma is virtually absent owing to severe pressure atrophy. (From Lim JH, Jang KT, Choi D: Biliary intraductal papillary-mucinous neoplasm manifesting only as dilatation of the hepatic lobar or segmental bile ducts: Imaging features in six patients. AJR Am J Roentgenol 191:778– 782, 2008.)

three-dimensional visualization of the biliary tree and vascular structures.3,37,44,147,168,358 Even without the use of a contrast agent, such images are considered comparable to conventional cholangiography. MR cholangiography is accurate in identifying the presence and level of a cholangiocarcinoma. MR cholangiography in conjunction with conventional abdominal MRI and MR angiography yields a  comprehensive examination that permits the diagnosis and staging  of cholangiocarcinoma. An MR cholangiogram provides information about the extent of tumor along the bile duct, and MRI and MR angiography provide information about lymph node metastasis and vessel involvement.44,68,225,235,351 Cholangiography, including endoscopic retrograde cholangiography and percutaneous cholangiography via a drainage catheter, can be used to assess the extent of the tumor, but the depth of bile duct invasion cannot be determined (see Fig. 42-92). In addition the procedure is invasive, with a major complication rate of 1.4% to 5%, and technical failure to visualize the bile ducts or inadequate visualization occurs in about 10% to 20% of all attempts.104 Also, when the bile duct is completely obstructed, the extent of cancer cannot be evaluated in the duct proximal to the obstruction. Fluorodeoxyglucose positron emission tomography (FDG PET) is valuable for evaluating suspicious lesions (Figs. 42-93 to 42-95) and detecting unsuspected distant metastases in patients with peripheral cholangiocarcinoma that might preclude surgical therapy and result  in a change in their oncologic management.6,179,279 In addition, it is possible to evaluate cholangiocarcinoma in the setting of sclerosing cholangitis.6 However, FDG PET has a high false-negative rate for the detection of periductal-infiltrating cholangiocarcinoma6,179,279 (Fig. 42-96). Even though each imaging modality has a role, the use of several modalities may be necessary for accurate diagnosis and staging.44,68,202,291 Other malignant tumors of the bile ducts.  Hepatocellular carcinoma may present with a biliary intraductal tumor consisting of 

CHAPTER 42  Biliary Tract and Gallbladder

A

B

C

D FIG 42-89  Hilar cholangiocarcinoma, Bismuth type IV. A and B, CT shows concentric thickening of the wall of the right and left hepatic ducts and common hepatic duct. Adjacent periductal fat and vessels are intact. C and D, Contrast-enhanced MRIs show concentric thickening of the bile duct wall and enhancement. There is no evidence of tumor extension out of the bile duct.

FIG 42-90  Periductal-infiltrating cholangiocarcinoma producing circumferential thickening of the bile duct wall (arrows). Note that the tumor has invaded the portal vein and right and left hepatic arteries.

1229

A

B

C

D FIG 42-91  Periductal-infiltrating hilar cholangiocarcinoma, Bismuth type IIIA. A to C, CT images show concentric thickening of the right hepatic duct, common hepatic duct, and cystic duct (short arrows) and the common bile duct. Note the metastatic lymph node (long arrow in B) invading the replaced right hepatic artery from the superior mesenteric artery. The common hepatic artery and portal vein have not been invaded. D, Cholangiogram shows a cholangiocarcinoma of the common hepatic duct and right hepatic duct (arrow).

A

B FIG 42-92  Hilar cholangiocarcinoma, Bismuth type II. A, CT shows concentric thickening of the wall of the common hepatic duct and right and left hepatic ducts (arrow). B, ERCP shows normal right and left hepatic ducts (arrows) with tumor limited to the common hepatic duct.

CHAPTER 42  Biliary Tract and Gallbladder

A

B FIG 42-93  Carcinoma of the ampulla of Vater. A, CT shows a small suspicious mass at the duodenal papilla (arrow). The distal common bile duct is dilated. B, FDG PET shows a lesion at the ampulla of Vater (arrow).

A

B FIG 42-94  Cholangiocarcinoma of the distal common bile duct. A, CT shows a nodular mass (arrow) in the bile duct. B, FDG PET shows increased uptake (arrow) indicating cancer.

A

B FIG 42-95  Recurrence of cholangiocarcinoma. A, CT 2 years after surgery for hilar cholangiocarcinoma shows a suspicious nodule at the resection margin (arrow). B, FDG PET shows intense uptake at the resection margin (arrow) indicating recurrent cancer.

1231

1232

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 42-96  Cholangiocarcinoma. A, CT shows a small nodule in the dilated common bile duct (arrow). B, Cholangiogram shows complete obstruction of the common bile duct (arrow). C, FDG PET discloses no lesion.

either a polypoid or a pedunculated mass, resulting in obstructive jaundice.335,338 A hepatic parenchymal tumor may invade the bile ducts and grow within them like a thrombus. Rarely hepatocellular carcinoma may arise primarily within the bile ducts, with no tumor in the adjacent liver parenchyma; the bile duct wall is intact in this type of tumor. Like ordinary parenchymal hepatocellular carcinoma, intraductal hepatocellular carcinoma is well enhanced on contrast-enhanced CT or MRI (Fig. 42-97). Underlying liver cirrhosis may be helpful in differentiating hepatocellular carcinoma from cholangiocarcinoma. Embryonal rhabdomyosarcoma is the most common malignant tumor in the pediatric population, occurring mainly in infants and preschool-aged children. Most of these tumors arise in the CBD and present as polypoid masses obstructing the bile duct lumen.74,315 In adults, lymphoma, leiomyosarcoma, and carcinoid tumor may occur. Leukemic involvement of the biliary tree may produce segmental thickening of the bile duct wall and obliteration of the lumen, mimicking hilar cholangiocarcinoma45 (Fig. 42-98). Metastatic tumors from gastric and colon cancer may produce diffuse or segmental thickening of the extrahepatic bile duct wall (Figs. 42-99 and 42-100). Periductal lymph node metastasis produces bile duct compression and invasion and causes bile duct obstruction.

FIG 42-97  Hepatocellular carcinoma in the common hepatic duct. Contrast-enhanced CT shows an enhancing nodular mass filling the common hepatic duct like a thrombus (arrow). The proximal bile ducts are dilated.

CHAPTER 42  Biliary Tract and Gallbladder

1233

B

A

FIG 42-98  Leukemic infiltration along the common hepatic duct. A, CT shows severe wall thickening (arrows) around the biliary drainage catheter. B, ERCP shows severe segmental thickening of the common hepatic duct and dilated intrahepatic duct.

FIG 42-99  Extrahepatic duct metastasis with disseminated metastatic tumors in a patient with colon cancer. Note the tumor mass inside the bile duct (arrow).

Tumors of the Ampulla of Vater and Duodenal Papilla.  The ampulla of Vater and duodenal papilla are small continuous structures at the distal end of the CBD. The ampulla of Vater is within the wall of the duodenum and the papilla is in the lumen of the duodenum. A tumor may arise from either one. An ampullary tumor is defined as a tumor confined to the ampulla without significant extension into the duodenum, whereas a tumor of the major papilla involves primarily the duodenal mucosa at the papilla. An ampullary tumor can easily involve the duodenal papilla and vice versa. Differentiation is important because a benign tumor of the duodenal papilla can be treated by endoscopic resection. Villous adenoma is the most common tumor arising from the duodenal papilla. Not infrequently, adenocarcinoma may develop from a preexisting villous adenoma of the papilla.157 The adenocarcinoma may arise from the ampulla of Vater (Fig. 42-101) or the duodenal

FIG 42-100  Extrahepatic duct metastasis in a patient with rectal cancer. The bile duct wall is concentrically thickened, with intense mucosal enhancement (arrow).

papilla (Fig. 42-102) and produce biliary obstruction. The tumor is depicted by CT or MR cholangiography as a mass at the distal end of the CBD or in the duodenal lumen.29,139,166 Although the bile ducts are dilated secondary to obstruction, the main pancreatic duct is usually not dilated because of the presence of the accessory pancreatic duct. A small intraampullary mass is usually not visualized on duodenal endoscopy, because the mass is inside the ampulla of Vater. Large ampullary tumors with duodenal papillary bulging are more easily identified with endoscopy and imaging such as CT and MR cholangiography. Likewise, tumors arising from the distal CBD and the head of the pancreas may be difficult to differentiate.166 However, owing to recent progress in MDCT, MR cholangiography, and endoscopic sonography, differentiation of these tumors has been simplified. Duodenal endoscopy is a helpful adjunct in the diagnosis of tumors arising from the duodenal papilla and ampulla of Vater region.

1234

A

PART II  CT and MR Imaging of the Whole Body

B

C FIG 42-101  Carcinoma of the ampulla of Vater. Axial (A) and coronal (B) CT and MRCP (C) show a round mass at the ampulla of Vater (arrow) resulting in obstruction of the bile ducts and pancreatic duct.

A

B FIG 42-102  Adenocarcinoma of the duodenal papilla. A large lobulated mass (arrows) is evident within the duodenum on CT (A) and MRCP (B), causing bile duct obstruction.

CHAPTER 42  Biliary Tract and Gallbladder

1235

GALLBLADDER NORMAL ANATOMY The gallbladder is a blind pouch lying along the undersurface of the liver. The gallbladder fossa is located in the plane of the interlobar fissure, which lies between the right and left hepatic lobes. On  transverse CT images, the gallbladder is a rounded structure with a maximum diameter of less than 4 to 5 cm in the distended state.127 When contracted the gallbladder is a smaller tubular structure. Visualization of the gallbladder wall depends on the degree of gallbladder distention and the presence of abnormality. The normal gallbladder wall thickness ranges from 1 to 3.5 mm.311,312,343 On US, 3 mm might be a reasonable upper limit of normal. Enhancement of the gallbladder wall on CT and MRI is normal after the intravenous administration of contrast medium.321,343 The density of the gallbladder lumen is generally that of water (0-20 Hounsfield units [HU]).127 After intravenous contrast administration, an increase in density is observed on CT. In general a small fraction of the contrast medium is excreted into the biliary tree. This amount is usually too small to be seen on plain films.321,332,333

FIG 42-103  Phrygian cap. Transverse MDCT shows a phrygian cap with a thickened wall (arrow) containing sandy stones.

CONGENITAL VARIANTS AND ANOMALIES Agenesis of the Gallbladder Gallbladder agenesis results from failure of development of the caudal division of the primitive hepatic diverticulum or from failure of vacuolization after the solid phase of embryonic development. Also, atresia or hypoplasia of the gallbladder represents aborted development of the organ.5,67,117 This malformation is reported in 0.013% to 0.155% of autopsy series, but the surgical incidence of gallbladder agenesis is approximately 0.02%.5,22 Nearly two thirds of adult patients with agenesis of the gallbladder have biliary tract symptoms, and gallstones in the extrahepatic bile duct are observed in 25% to 50% of these patients.16,82 Although US, CT, or MRI may suggest the diagnosis, preoperative diagnosis of gallbladder agenesis is difficult.125,126 This disorder is generally diagnosed during intraoperative cholangiography. FIG 42-104  Multiseptate gallbladder. The thickened wall of the fundal

Duplication of the Gallbladder 113

Gallbladder duplication occurs in 0.025% of the population. This anomaly is caused by incomplete revacuolization of the primitive gallbladder, resulting in a persistent longitudinal septum. Each separate gallbladder cavity must have its own cystic duct. These duplicated cystic ducts may enter the common duct separately or unite before a common entrance.331 Most reported cases of gallbladder duplication include a clinical picture of cholecystitis with stones. A folded gallbladder, bilobed gallbladder, choledochal cyst, pericholecystic fluid, gallbladder diverticulum, vascular band across the gallbladder, and focal adenomyomatosis can mimic a double gallbladder.99,110,258 Complications in double gallbladder include torsion, papilloma, carcinoma, common duct obstruction, and biliary cirrhosis. Triple and quadruple gallbladders have also been reported.99

Phrygian Cap The most common anomaly of the entire biliary tree is a septation in the distal fundus of the gallbladder, which results in the configuration called a phrygian cap (Fig. 42-103). It occurs in 1% to 6% of the population.83 This deformity is characterized by a fold or septum of the gallbladder between the body and fundus. Two variations of this anomaly have been reported. In the retroserosal or concealed type, the mucosal fold projecting into the lumen may not be visible externally. In the serosal or visible type, the peritoneum follows the bend

segment (arrow) may represent chronic segmental cholecystitis.

in the fundus and then reflects on itself as the fundus overlies the body. This anomaly can be mistaken for stones or hyperplastic cholecystosis.83,99,110,258

Multiseptate Gallbladder Septations occur elsewhere in the gallbladder but are less common and may be multiple.243 The multiseptate gallbladder is a solitary gallbladder with multiple septa of various sizes internally.328 The gallbladder is usually normal in size and position, and the chambers communicate with one another by one or more orifices from the fundus to the cystic duct. These septations lead to stasis of bile and gallstones1 (Fig. 42-104). The differential diagnoses include desquamated gallbladder mucosa and hyperplastic cholecystosis.

Diverticula Gallbladder diverticula can occur anywhere in the gallbladder; they are usually single and can vary greatly in size.184 Congenital diverticula are true diverticula and contain all the muscle layers. In contrast the pseudodiverticula of adenomyomatosis have little or no smooth muscle in their walls.

1236

PART II  CT and MR Imaging of the Whole Body

Wandering Gallbladder The gallbladder may have a long mesentery that allows it to lie in an abnormal location, such as the pelvis243—the so-called wandering or floating gallbladder. It may also wander in front of the spine or to the left of the abdomen and can rarely herniate through the foramen of Winslow into the lesser peritoneal sac. These wandering gallbladders may be prone to torsion or volvulus.

Ectopic Gallbladder The gallbladder can be located in various positions (Fig. 42-105). The intrahepatic gallbladder is completely surrounded by hepatic parenchyma and typically lies in a subcapsular location along the anteroinferior aspect of the right hepatic lobe. It may complicate the clinical diagnosis of acute cholecystitis because there are few peritoneal signs. With hypoplasia, aplasia, atrophy, or hypertrophy of one of the hepatic lobes, the gallbladder may be displaced or rotated abnormally to one side. In patients with cirrhosis and associated atrophy of the right hepatic lobe or chronic obstructive pulmonary disease, the gallbladder together with the colon is often interposed between the liver and the diaphragm.111,263 A suprahepatic gallbladder has been described, with hypoplasia of the right hepatic lobe and eventration of the diaphragm.39 Patients with disease in an ectopic gallbladder may present diagnostic challenges because their symptoms are atypical with regard to the location of pain.

Cholecystomegaly and Microgallbladder Enlargement of the gallbladder has been reported in patients with diabetes and after vagotomy. The gallbladder also becomes larger than normal during pregnancy, in patients with sickle hemoglobinopathy, and in extremely obese people.15,87,251,336 In patients with cystic fibrosis, the gallbladder is typically small, trabeculated, contracted, and poorly functioning.242,346 Microgallbladder often contains echogenic bile, sludge, and cholesterol gallstones.

PATHOLOGIC CONDITIONS When renal function is abnormal, hepatobiliary excretion of contrast agent may be pronounced, with increased density of the gallbladder lumen on CT and plain films.302,321 Abnormal density of the gallbladder

FIG 42-105  Ectopic gallbladder. Transverse MDCT shows the ectopic gallbladder (arrow) between the stomach and left hepatic lobe.

lumen on CT can occur in various pathologic conditions. Hemobilia can increase the density of bile toward that of blood (50-60 HU).190 Innumerable tiny stones may appear simply as increased bile density in the gallbladder lumen when the stones are too small to be identified individually. However, the most common cause of increased bile density in the gallbladder lumen is cystic duct obstruction resulting from a stone or tumor. In acute and chronic cholecystitis, the density of the gallbladder lumen can be increased owing to inflammatory debris.332,333 Milk of calcium bile often makes a fluid-fluid level, with the denser calcium salt suspension layering dependently. It may have a density as high as 1000 HU.141 The cystic duct joining with the CHD forms the CBD at varying levels from high in the porta hepatis to near the ampulla of Vater (Fig. 42-106).

Gallstones Several factors affect the appearance of gallstones on CT, but the major determinant is composition: bile pigment, cholesterol, or calcium. Most gallstones are admixtures of these three elements, and their density depends on the amount of each element present. In general, calcium increases the density of the stone; up to 60% of gallstones attributed to calcium are dense on CT (Fig. 42-107). Mixed-composition stones with less calcium may have a density equal to that of bile, making them difficult to visualize on CT (Fig. 42-108). Pure cholesterol stones are rare but may have a lower density than bile (−100 HU in vitro).28,97 Also, the pattern in which calcium is distributed affects the appearance of gallstones. Whereas calcium bilirubinate stones  may have a homogeneously dense appearance, mixed-composition stones may have a low-density cholesterol center and a dense peripheral rim of calcium, producing a laminated appearance with alternating layers of calcium or punctate foci of calcium scattered throughout (see Fig. 42-107). Gallstones may have central clefts in their matrix that contain gas consisting mostly of nitrogen, producing the so-called Mercedes-Benz, crow’s feet, or seagull sign. A gaseous cleft may contribute to the floating of gallstones, aiding in their detection on CT when they are isodense to bile (Fig. 42-109). It has been reported that CT was 78% sensitive prospectively and 94% sensitive retrospectively in the detection of gallstones.127 Again, the composition of the gallstone is the most important factor in the detection rate. Calcified gallstones that are hyperdense compared with bile are more readily detected than isodense mixed-composition stones. In addition the size, number, and shape of gallstones, as well as bile density and scanning technique, can affect the ability to detect gallstones on CT. On MRI, gallstones tend to have low signal intensity on both T1and T2-weighted images. T2-weighted sequences are considered superior to T1-weighted sequences, because gallstones are easily contrasted against the high signal intensity of the surrounding bile39 (Fig. 42-110). Although cholesterol gallstones usually have low signal intensity on T1-weighted images, pigment gallstones, seen as brown to black on cross section, tend to have high signal intensity, and the intensity ratio of gallstone to bile is significantly higher for pigment gallstones than for cholesterol gallstones329 (Fig. 42-111). Sometimes a high-signalintensity center is seen on T2-weighted images, which corresponds to a fluid-filled cleft on cross section334 (Fig. 42-112). MRI should be carefully interpreted for the presence of gallstones in patients with pneumobilia, because the signal void from an air bubble can mimic a gallstone on T2-weighted images (Fig. 42-113). The differentiation of cholesterol and pigment gallstones was important when patients were treated with nonsurgical dissolution therapy, because the choice of the solvent and the success rate depended on the composition of the gallstones. In the era of laparoscopic

CHAPTER 42  Biliary Tract and Gallbladder

A

1237

B

C FIG 42-106  MDCT of the gallbladder and cystic duct. A, Transverse scan shows the cystic duct (single arrow) and two gallstones (two arrows) in the gallbladder neck. B, Transverse scan 15 mm above A shows the common hepatic duct (arrow) and right bile duct (arrowheads). C, Coronal image shows the common bile duct (black arrow) and cystic duct (white arrows) with spiral folds.

cholecystectomy, however, the assessment of gallstone composition has become less important.

Acute Cholecystitis Acute cholecystitis is caused by obstruction of the cystic duct or gallbladder neck, resulting in inflammation of the gallbladder wall and increased intraluminal pressure.20 Although most patients present with right upper quadrant pain and tenderness, the diagnosis may be a challenge in the setting of severe advanced inflammation or in older patients or those with systemic diseases such as diabetes. The clinical signs and symptoms of acute cholecystitis are nonspecific, and in 60% to 85% of patients evaluated for cholecystitis, the symptoms are found to result from other causes.276 The majority of cases of acute cholecystitis (≈90%) are caused by gallstones. The precipitating event for the development of acute calculous cholecystitis appears to be occlusion of the gallbladder neck or cystic duct by a gallstone. However, outflow obstruction does not always cause acute cholecystitis, and animal models in which the cystic duct was ligated or obliterated showed only shrinkage of the gallbladder, without the development of acute cholecystitis.293 Therefore mucosal ischemia or injury may be another contributing factor in the development of acute cholecystitis.308 US is generally the screening method of choice in a patient with suspected acute cholecystitis. CT is usually performed for evaluation

of suspected complications of acute cholecystitis such as gallbladder perforation (discussed later). However, CT may be used as an initial imaging modality if the cause of symptoms is unclear or if US findings are equivocal. The CT features of acute calculous cholecystitis include gallstones, thickening of the gallbladder wall, pericholecystic inflammatory change such as fluid collection and fat stranding, highattenuation bile, blurring of the interface between the gallbladder and the liver, and a transient increase in attenuation of the liver adjacent to the gallbladder20,144,238,248,276,350 (Figs. 42-114 and 42-115). Of all these findings, pericholecystic inflammatory change is believed to be the most specific20,93,276 because other findings such as gallbladder wall thickening and distention do not necessarily indicate the presence of inflammation. Only about 75% of stones are detected with CT.276 Mirvis and coworkers divided the CT findings of acute cholecystitis into major criteria—including gallstones, thickened gallbladder wall, pericholecystic fluid, and subserosal edema—and minor criteria— gallbladder distention and sludge.248 They concluded that the diagnosis of acute cholecystitis can be made if one major and two minor criteria are present. Bennett and colleagues21 reported that the overall sensitivity, specificity, and accuracy of CT for the diagnosis of acute cholecystitis is 91.7%, 99.1%, and 94.3%, respectively. Although MRI is usually performed only in cases of ambiguous clinical features, it is effective in depicting abnormalities of the gallbladder, especially MR cholangiography using heavily T2-weighted

1238

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 42-107  Various appearances of radiopaque gallstones on nonenhanced CT. A, Small calcified gallstone with a homogeneous dense appearance. B, Gallstone with a low-density cholesterol center and a dense peripheral rim of calcium. C, Laminated appearance of a gallstone with alternating layers of calcium. D, Gallstone with punctate foci of calcium scattered throughout.

FIG 42-108  Radiolucent gallstone. Nonenhanced CT obtained after

FIG 42-109  Radiolucent gallstones with gaseous central clefts (arrows)

ERCP shows a radiolucent gallstone (arrow). It is seen as a negative filling defect because of the positive contrast agent; otherwise it would be isodense with bile.

that contribute to their floating on nonenhanced CT.

A

B

FIG 42-110  A, Gallstones with low-density centers and subtle

C

A

peripheral calcification (arrows) are seen on nonenhanced CT. B and C, The gallstones have low signal intensity on both T1(B) and T2-weighted (C) MRIs.

B

FIG 42-111  A, Gallstone with a low-density center and dense

C

peripheral rim of calcium is seen on nonenhanced CT. B and C, The gallstone has low signal intensity on T2-weighted MRI (B) but high signal intensity on the T1-weighted image (C).

1240

A

PART II  CT and MR Imaging of the Whole Body

B FIG 42-112  A, T2-weighted MRI shows a gallstone with low signal intensity and a central cleft with high signal intensity. B, T1-weighted image shows the gallstone with a low-signal-intensity center and a highsignal-intensity periphery.

A

B FIG 42-113  A, Coronal T2-weighted MRI shows a suspicious gallstone, seen as a round signal void (arrow). B, On the axial T2-weighted image, the signal voids turns out to come from an air bubble (arrow) in the gallbladder.

A

B FIG 42-114  Acute calculous cholecystitis. Nonenhanced (A) and contrast-enhanced (B) CT shows gallbladder wall thickening, high-attenuation bile, a small gallstone obstructing the cystic duct (arrows), and pericholecystic fat stranding suggestive of acute cholecystitis.

CHAPTER 42  Biliary Tract and Gallbladder sequences. MR findings of acute cholecystitis are similar to those reported with CT, but MRI is superior because it can demonstrate radiolucent gallstones. US is operator dependent in many circumstances, and MR cholangiography is superior to US in depicting gallstones located in the cystic duct and gallbladder neck275 (Fig. 42-116). With the use of a hepatobiliary contrast agent such as mangafodipir trisodium, enhanced T1-weighted MR cholangiography provides functional information about cystic duct patency that is identical to that provided by hepatobiliary scintigraphy.178

Emphysematous Cholecystitis.  Emphysematous cholecystitis is an uncommon variant of acute cholecystitis. Vascular compromise of   the cystic artery is believed to play a role in the development of this condition, with proliferation of gas-forming organisms in an anaerobic environment and penetration of gas into the gallbladder wall.148,240 Commonly isolated gas-forming organisms are Clostridium perfringens

1241

and E. coli.240 Emphysematous cholecystitis has distinct differences from ordinary acute cholecystitis, suggesting a separate pathogenesis245; these include a male preponderance (up to 71%), frequent occurrence in diabetic patients (up to 50%), lack of gallstones in up to  one third of patients, and higher risk of gallbladder gangrene and perforation. On pathologic review, gallbladders with emphysematous cholecystitis have a higher incidence of endarteritis obliterans, supporting vascular insufficiency as a causative factor. The mortality rate for emphysematous cholecystitis is 15%, versus 4% in uncomplicated acute cholecystitis,114,245 and prompt surgical intervention is required. CT is more sensitive and specific than US or plain radiographs  for the diagnosis of emphysematous cholecystitis.109,114 Therefore CT should be performed if emphysematous cholecystitis is suspected on US, if there is a high clinical index of suspicion, or if the gallbladder cannot be visualized adequately by US in patients with suspected acute cholecystitis. In addition to the findings of acute cholecystitis, CT may show gas within the gallbladder lumen or wall114 and extension into the pericholecystic tissues in patients with emphysematous cholecystitis (Fig. 42-117). If free air is identified, this indicates perforation of the gallbladder, and urgent surgery is required.

Complications of Acute Cholecystitis.  As the pathologic process

FIG 42-115  Acute calculous cholecystitis. Contrast-enhanced CT shows gallbladder wall thickening, a small stone in the gallbladder neck (arrow), blurred interface between the gallbladder and liver, and increased attenuation of the liver adjacent to the gallbladder (arrowheads) suggestive of acute cholecystitis.

A

evolves, acute cholecystitis can be complicated by hemorrhage, gangrene, and perforation. Hemorrhagic cholecystitis.  Hemorrhagic cholecystitis is an uncommon complication of acute cholecystitis.153,301 In the presence of gallstones, it is presumed that hemorrhage results from inflammation causing mucosal ulceration and necrosis. Clinically patients with  hemorrhagic cholecystitis may present with biliary colic, jaundice, hematemesis, and melena. Rarely patients may present with massive upper gastrointestinal bleeding, hemoperitoneum, or obstruction of the CBD.280 In patients with acalculous cholecystitis, possible causes of gallbladder hemorrhage are tumor, vascular abnormalities (e.g., aneurysm), trauma, anticoagulation, or ectopic pancreatic or gastric mucosa. In addition to the findings of acute cholecystitis, CT in patients with hemorrhagic cholecystitis may show intraluminal hematoma or increased density of bile with a fluid-fluid level (Fig. 42-118). However, this must be differentiated from other conditions that can cause

B FIG 42-116  Acute calculous cholecystitis. A, T1-weighted MRI shows gallbladder distention, high-signalintensity bile, and multiple small fillings defects (arrows) suggestive of gallstones. Note, however, that the bile in the cystic duct has low signal intensity (arrowheads). B, T2-weighted image better demonstrates small gallstones in the cystic duct (arrow). Also shown are the different signal intensities of bile in the gallbladder and cystic duct.

1242

A

PART II  CT and MR Imaging of the Whole Body

B FIG 42-117  Emphysematous cholecystitis. Nonenhanced (A) and contrast-enhanced (B) CT shows gas within the gallbladder wall, along with the findings of simple cholecystitis: distended gallbladder, obstructing stone in the cystic duct (arrow in A), pericholecystic fat stranding, and pericholecystic fluid collection.

A

B

C FIG 42-118  Hemorrhagic cholecystitis. A, Contrast-enhanced CT shows mild thickening of the gallbladder wall and intraluminal hematoma (arrows) contrasted against the low-attenuation bile. B, ERCP demonstrates an amorphous filling defect (arrows) in the common bile duct, suggestive of hemobilia. C, Nonenhanced CT obtained after ERCP shows intraluminal hematoma, seen as a negative filling defect (arrows) owing to the high attenuation of contrast agent.

CHAPTER 42  Biliary Tract and Gallbladder

1243

B

A

FIG 42-119  Hemorrhagic cholecystitis. A, T1-weighted MRI shows intraluminal hematoma, seen as a highsignal-intensity lesion contrasted against the low signal intensity of bile. Also noted is a focal defect of the lateral gallbladder wall (arrowheads) suggestive of gangrenous change and a perforated gallbladder. B, T2-weighted image also shows intraluminal hematoma as a low-signal-intensity lesion and a focal defect of the lateral gallbladder wall (arrowheads).

high-density bile, such as recently administered intravenous contrast or milk of calcium. Hemoperitoneum or dilated IHD may be noted if hemorrhagic cholecystitis is associated with gallbladder perforation  or obstruction of the CBD due to hemobilia. On MRI, subacute  hemorrhage is typically bright on T1-weighted images because of the T1-shortening effects of methemoglobin. On T2-weighted images, a dark rim due to hemosiderin deposition may be appreciated along the inner surface of the gallbladder85 (Fig. 42-119). Gangrenous cholecystitis.  Acute cholecystitis may result in gangrene of the gallbladder, with an incidence ranging from 2% to 29.6% in surgical series.2,135,255,348 It is more common in men, patients of advanced age, and those with coexisting cardiovascular disease. Chemical inflammatory reaction of the mucosa with bile salts and progressive distention of the gallbladder lumen leading to vascular compromise are possible mechanisms.232,320 Once gangrenous cholecystitis is suspected, patients generally undergo emergency cholecystectomy to avoid further life-threatening complications, with a higher conversion rate to open cholecystectomy than in uncomplicated acute cholecystitis.149,246,307 CT findings of gangrenous cholecystitis include intraluminal membranes, hemorrhage into the lumen, irregular or absent wall, pericholecystic abscess, and gas in the gallbladder lumen or wall.21,93,196,276 CT with the administration of intravenous contrast is recommended to allow evaluation of gallbladder wall enhancement and mural contour. Bennett and colleagues reported that CT had high specificity (96%) but low sensitivity (29.3%) for identifying acute gangrenous cholecystitis, with an overall accuracy of 86.7%.21 In their series the CT findings with the highest specificity for gangrenous cholecystitis were gas in the wall or lumen (100%), intraluminal membranes (99.5%), irregularity or absence of the gallbladder wall (97.6%), and pericholecystic abscess (96.6%) (Fig. 42-120). Among other findings of acute cholecystitis, the presence of pericholecystic fluid, greater degree of gallbladder distention in the short axis, and mural thickening were correlated with a higher likelihood of gangrenous cholecystitis. Gallbladder perforation.  Gangrene of the gallbladder may eventually result in gallbladder perforation, with an incidence ranging from 8.3% to 11.9%.81,320 Other causes of gallbladder perforation include chronic cholecystitis, cholelithiasis, trauma, neoplasm, steroids, and vascular compromise. The fundus is the most frequent site of

*

FIG 42-120  Gangrenous cholecystitis. Contrast-enhanced CT shows an intraluminal membrane, gas in the gallbladder lumen and wall (arrows), a focal defect in the gallbladder wall (arrowheads), and a pericholecystic abscess (asterisk) suggestive of gangrenous cholecystitis with perforation.

perforation because of the relatively poor blood supply in this area. Niemeier divided gallbladder perforation into three types: (1) acute free perforation into the peritoneal cavity, (2) subacute perforation with pericholecystic abscess, and (3) chronic perforation with cholecystoenteric fistula.265 In most series, subacute perforation with pericholecystic abscess is the most common type as a complication of acute cholecystitis.95 A wide range of mortality rates—6% to 70%—has been reported in patients with gallbladder perforation.69,81,95,130,265 In one series there was a 40% mortality rate associated with type 1 free perforation and a 4% mortality rate for type 2.95 Prompt surgical intervention is warranted to decrease morbidity and mortality in patients with gallbladder perforation. Because clinical signs and symptoms are frequently nonspecific and indistinguishable from those of uncomplicated acute cholecystitis, radiologic studies play a vital role, and it is generally agreed that CT is superior to US for

1244

PART II  CT and MR Imaging of the Whole Body

this purpose.172,253,277 Interruption of the gallbladder wall or a focal defect on CT may indicate gallbladder perforation. A gallstone may  be noted outside of the gallbladder lumen.253 Pericholecystic abscess formation may be confined to the gallbladder fossa or rarely extend into the liver42,277 (Fig. 42-121). There may also be free perforation associated with free fluid corresponding to bile (Fig. 42-122). CT  is useful for further evaluation or to better determine the extent of inflammation.172,277

Acalculous Cholecystitis

clinical findings are nonspecific. In many circumstances, US is equivocal because these patients have distended thick-walled gallbladders even in the absence of inflammation, and the sonographic Murphy’s sign is not reliable in these patients. The CT findings are similar to those of acute calculous cholecystitis, with the exception of the absence of gallstones (Figs. 42-123). CT may be more specific when US findings are nondiagnostic.26,248,249,276 Earlier surgical intervention is considered in patients with acalculous cholecystitis, because they tend to have an increased complication rate and a higher overall mortality.338

Acalculous cholecystitis, an infrequent but clinically serious entity, is found in approximately 5% of all patients undergoing cholecystectomy.8 It usually occurs in the setting of prolonged illness, such as patients with significant trauma or long stays in the intensive care unit.20 An increase in bile viscosity from stasis, with subsequent cystic duct obstruction and mucosal ischemia, may be contributing factors. It is often difficult to diagnose acalculous cholecystitis, because the

FIG 42-121  Gallbladder perforation. Contrast-enhanced CT shows a focal defect in the gallbladder wall (arrowheads) and pericholecystic abscess (arrows).

A

FIG 42-122  Gallbladder perforation. Coronal contrast-enhanced CT shows a focal wall defect in the gallbladder fundus (arrowheads), a large amount of peritoneal fluid, and diffuse enhancement of thickened peritoneum, suggesting bile peritonitis from gallbladder perforation.

B FIG 42-123  Acute acalculous cholecystitis. Nonenhanced (A) and contrast-enhanced (B) CT shows a distended gallbladder with mild wall thickening, high-attenuation bile, pericholecystic fat stranding, fluid collection in the pericholecystic and perihepatic spaces, and absence of gallstones, suggestive of acute acalculous cholecystitis.

CHAPTER 42  Biliary Tract and Gallbladder Chronic Cholecystitis Chronic cholecystitis is almost always associated with gallstones,  but the pathogenesis of this common disease is poorly understood. Although it has been proposed that chronic cholecystitis may develop from recurrent attacks of mild acute cholecystitis, most patients lack a clinical history to support this hypothesis. Although repetitive mucosal trauma by gallstones may explain in part the inflammatory and reparative changes of chronic cholecystitis, there is a poor correlation between the severity of the inflammatory response and the number and size of gallstones. A genetically determined florid inflammatory response of the gallbladder or an abnormal composition of the bile leading to chemical mucosal injury and stone formation is a possible but currently unproved hypothesis.154 Chronic cholecystitis may have CT and MRI findings similar to those of acute cholecystitis, such as gallstones, wall thickening, and wall enhancement (Fig. 42-124). The gallbladder is sometimes contracted rather than distended (Fig. 42-125). However, gallbladder wall thickening and enhancement in chronic cholecystitis frequently masquerade as gallbladder carcinoma, especially in the absence of gallstones (Fig. 42-126). Conversely there may be minimal abnormality on CT except for the presence of gallstones, which is neither necessary nor sufficient for the diagnosis of chronic cholecystitis. In such cases, correlation with the clinical findings is critical, and diagnosis often requires nuclear medicine techniques, particularly to evaluate the gallbladder ejection fraction.

Xanthogranulomatous Cholecystitis.  Xanthogranulomatous cholecystitis is an uncommon form of chronic cholecystitis; it is nearly always associated with gallstones and is frequently associated with variable degrees of fibrosis. Although the pathogenesis of this condition is uncertain, it has been suggested that mucosal ulcers or ruptured Rokitansky-Aschoff sinuses, together with outflow obstruction by gallstones and infection, result in the penetration of bile into the gallbladder wall, forming xanthogranulomas.154 On CT, gallstones and masslike irregular thickening of the gallbladder wall are the most common abnormalities (Fig. 42-127). Several studies have reported that intramural hypoattenuated nodules on CT represent a xanthogranulomatous lesion, an abscess, or a combination

A

1245

of the two59,171,305 (Figs. 42-128 and 42-129). Chun and associates reported that continuous enhancement of the mucosal line (64%) on postcontrast CT indicated that the main lesion was located in the intramural portion and that the mucosal surface overlying the lesion was either intact or focally denuded.59 Similarly, Shuto and colleagues reported that luminal surface enhancement (70%) corresponded with the presence of an epithelial layer (see Fig. 42-129).305 MRI findings of xanthogranulomatous cholecystitis have rarely been reported, but it is not surprising that MRI may be helpful in depicting the pathologic processes of this entity because of its superb tissue contrast resolution (see Figs. 42-127 and 42-129). Furuta and coworkers described the signal intensity of the gallbladder wall as low on T1-weighted images and moderately high on T2-weighted images.106 They also noted multiple intramural hyperintense nodules on T2-  weighted images that were nonenhanced on postcontrast T1-weighted images. In their cases, histologic examination of specimens showed several abscesses with numerous foam cells in the gallbladder wall. Similarly, Shuto’s group found intramural nodules with very high signal intensity on T2-weighted images in four of five patients; these nodules were not enhanced on postcontrast studies and consisted of necrosis or abscess histopathologically.305 In their series the slightly high-intensity areas in the gallbladder wall on T2-weighted images were strongly enhanced on postcontrast T1-weighted images during the late phase and consisted of abundant proliferation of foamy  cells. It is important for the radiologist to distinguish xanthogranulomatous cholecystitis from gallbladder cancer. The characteristic CT and MRI findings of xanthogranulomatous cholecystitis—including a patent mucosal line or luminal surface enhancement on CT, and intramural nodules with hypoattenuation on CT or very high signal intensity on T2-weighted images—can be helpful in differentiating the two conditions.

Complications of Chronic Cholecystitis Mirizzi’s syndrome.  Mirizzi’s syndrome, typically a complication of long-standing calculous cholecystitis, is defined as narrowing of the extrahepatic bile duct, causing obstructive jaundice by impaction of gallstones within the cystic duct and subsequent extrinsic compression or inflammation.247 A long or low insertion of the cystic duct into the

B FIG 42-124  Chronic cholecystitis. A, Contrast-enhanced CT shows a small gallstone (arrow) and mild thickening of the gallbladder wall. B, MR cholangiogram better demonstrates the irregularity of the gallbladder wall at the fundus (arrowheads), as well as gallstones (long arrows). There is also a small stone in the common bile duct (short arrow).

1246

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 42-125  Chronic cholecystitis. Nonenhanced (A) and contrast-enhanced (B) CT shows a contracted gallbladder with mild wall thickening. Axial T2-weighted MRI (C) and MR cholangiography (D) better demonstrate the small gallstone (arrow) and irregularity of the gallbladder wall at the fundus.

FIG 42-126  Chronic cholecystitis. Contrast-enhanced CT shows irregular thickening of the gallbladder wall, without a definite stone.

CBD may predispose to this condition.185,278 Preoperative diagnosis of Mirizzi’s syndrome can be challenging because of the inflammation, fibrosis, and distortion of structures at the level of obstruction. In a recent surgical series, this syndrome was found in 1.13% of all cholecystectomies, but the diagnosis could be made preoperatively in only 56% of patients.162 Mirizzi’s syndrome may be suggested on CT or MRI by identification of a stone within the cystic duct, with concurrent biliary ductal dilatation at the level of the cystic duct176,330 (Fig. 42-130). However, on CT, Mirizzi’s syndrome may mimic carcinoma of the CHD or cystic duct if the gallstones are radiolucent (Fig. 42-131). In such cases it is important to ascertain that there is no definite mass or lymphadenopathy, as would be expected with bile duct cancer.25 Sometimes by means of pressure necrosis, impacted gallstones in the gallbladder neck or cystic duct may erode into the adjacent CHD or the surrounding tissue of the hepatoduodenal ligament. Gallstone ileus.  Gallstone ileus is rare, occurring in 0.3% to 0.5% of all cases of cholelithiasis.75 It is more prevalent in women than in men, with a ratio of 3 : 1 to 16 : 1.287 The majority of obstructing stones

CHAPTER 42  Biliary Tract and Gallbladder

A

B

C

D FIG 42-127  Xanthogranulomatous cholecystitis. Nonenhanced (A) and contrast-enhanced (B) CT shows a large gallstone, masslike irregular thickening of the gallbladder wall, and increased enhancement of the surrounding liver. The thickened gallbladder wall has low signal intensity on the T1-weighted MRI (C) and moderately high signal intensity on the T2-weighted image (D).

A

B FIG 42-128  Xanthogranulomatous cholecystitis. A, Contrast-enhanced CT shows masslike irregular thickening of the gallbladder wall and multiple intramural hypoattenuated nodules (arrows). Also note the findings of cirrhosis: surface nodularity and fibrotic low attenuation in the liver periphery, as well as a large amount of ascites. B, Photograph of the cut surface demonstrates multiple yellowish intramural nodules and hemorrhage (arrows) in the irregularly thickened gallbladder wall.

1247

1248

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 42-129  Xanthogranulomatous cholecystitis. A, Contrast-enhanced CT shows a gallstone, irregular thickening of the gallbladder wall, and multiple intramural hypoattenuated nodules (arrows). Also noted is the presence of luminal surface enhancement and focal discontinuity (arrowheads). B and C, Intraluminal nodules (arrows) exhibit variable signal intensity on the T1-weighted MRI (B) and bright high signal intensity on the T2-weighted image (C). Focal discontinuity of the luminal surface (arrowheads) is also appreciated on both T1- and T2-weighted images. D, Photograph of cut surface demonstrates multiple xanthogranulomas seen as yellowish intramural nodules (arrows) in the irregularly thickened gallbladder wall.

A

B FIG 42-130  Mirizzi’s syndrome. A, Contrast-enhanced CT shows wall thickening in the cystic duct containing the gallstone (arrows) and a compressed common hepatic duct, appearing as a triangular shape (arrowhead). B, ERCP shows nonopacification of the cystic duct and marked narrowing of the common hepatic duct by smooth extrinsic indentation (arrows) at the level of cystic duct insertion, suggestive of Mirizzi’s syndrome.

CHAPTER 42  Biliary Tract and Gallbladder

A

B

C

D

1249

FIG 42-131  Mirizzi’s syndrome. A, Contrast-enhanced CT shows dilated intrahepatic bile duct and an illdefined soft tissue masslike lesion (arrows) around the cystic duct. B, T2-weighted MRI demonstrates a small gallstone (arrow) seen as a dark filling defect contrasted against the high signal intensity of bile in the center of the ill-defined lesion. MR cholangiography (C) and ERCP (D) reveal dilated intrahepatic bile duct and narrowing of the common hepatic duct by smooth extrinsic indentation (arrows) at the level of cystic duct insertion, suggestive of Mirizzi’s syndrome.

in gallstone ileus originate in the gallbladder, but this does not account for all cases; this condition has also been reported in patients who had undergone prior cholecystectomy.75,199 When it occurs as a complication of chronic cholecystitis, the gallstone erodes through the gallbladder wall and into the adjacent bowel through a cholecystoenteric fistula, most commonly between the gallbladder and duodenum.9,75 Patients may have had a recent acute biliary episode preceding the onset of gallstone ileus. Once the stone enters the alimentary tract, the site of obstruction is usually the terminal ileum because it is the narrowest portion of the small bowel; however, less than half of these stones cause obstruction, and many are evacuated uneventfully in the stool.75,138,199 In general, gallstones causing ileus are single, large, and faceted.344 Although plain radiographs may demonstrate the traditional triad of small bowel obstruction, pneumobilia, and ectopic gallstone, not all gallstones are radiopaque on plain radiographs, and pneumobilia may not be appreciated. CT is more accurate in diagnosing gallstone ileus by demonstrating a stone within the bowel lumen (usually the distal ileum), dilated small bowel loops proximal to the stone, and air in the bile duct and gallbladder.78,115,198,286 CT may also demonstrate a

distorted gallbladder in direct continuity with a thickened duodenal wall226,286 (Fig. 42-132). Porcelain gallbladder.  Dystrophic calcification can be associated with fibrous tissue in the gallbladder wall in chronic cholecystitis, resulting in the so-called porcelain gallbladder. The term porcelain emphasizes the blue discoloration and brittle consistency of the  gallbladder.23,161 Calcium may be found either as a broad continuous plaque in the muscle layer or as punctate foci in the mucosal layer (Fig. 42-133). CT is the best method for depicting dystrophic calcification in the gallbladder wall, although plain radiographs and US may also demonstrate this finding. Sometimes the calcified rim of a large stone filling the gallbladder lumen or a deposit of lipiodol in the gallbladder wall after transcatheter arterial chemoembolization for hepatic malignancy may mimic porcelain gallbladder. For unknown reasons, porcelain gallbladder is frequently associated with carcinoma of the gallbladder23 (Fig. 42-134). Therefore special attention to any sign of carcinoma is mandatory when reading CT scans in these patients. Even when the patient has few or no gallbladder symptoms and there is no evidence of gallbladder carcinoma, prophylactic cholecystectomy is recommended.

1250

PART II  CT and MR Imaging of the Whole Body

*

A

B

C

D

E FIG 42-132  Gallstone ileus. A, Contrast-enhanced CT shows a large stone (arrow) in the gallbladder neck and masslike thickening of the gallbladder wall adherent to the adjacent duodenum (asterisk). B, The patient presented with sudden abdominal pain 3 months later. Plain abdominal radiograph shows small bowel ileus, but neither pneumobilia nor ectopic gallstone is evident. C, Contrast-enhanced CT shows a large fistula (arrowhead) between the gallbladder and duodenum, with disappearance of the gallstone. D, Nonenhanced CT reveals that the gallstone (arrow) has migrated to the ileum, causing small bowel ileus. E, Follow-up CT 6 days later shows migration of the gallstone (arrow) to the rectum and partial improvement of small bowel ileus.

CHAPTER 42  Biliary Tract and Gallbladder

1251

Hyperplastic Cholecystosis The generic term hyperplastic cholecystosis was first introduced by Jutras155 to describe a group of noninflammatory conditions of the gallbladder, including cholesterolosis and adenomyomatosis, in which benign proliferation of normal tissue is the important feature. However, except that neither is an inflammatory or a neoplastic condition, cholesterolosis and adenomyomatosis have little similarity in terms of cause and clinical features.24

Cholesterolosis.  Cholesterolosis consists of multiple deposits of cholesterol-laden macrophages in the lamina propria of the gallbladder wall, forming cholesterol polyps.24,155 These are generally too small to be seen on CT and MRI but are better appreciated on US (Fig. 42-135).

Adenomyomatosis.  Adenomyomatosis is characterized by hyperFIG 42-133  Porcelain gallbladder. Nonenhanced CT shows porcelain gallbladder with punctate foci of calcification.

A

plastic muscular wall thickening, mucosal overgrowth, and intramural diverticula, crypts, or sinus tracts—so-called Rokitansky-Aschoff sinuses.24,63,100,155,347 There are three different morphologic types of

B FIG 42-134  Porcelain gallbladder. A, Nonenhanced CT shows irregular calcification in the gallbladder wall, suggestive of porcelain gallbladder. B, Contrast-enhanced CT shows an irregular mass (arrows) engulfing the calcification, suggestive of gallbladder carcinoma. Also noted are multiple lymphadenopathies (arrowheads) in the hepatoduodenal ligament and around the left gastric artery, suggestive of metastases, and a dilated intrahepatic bile duct.

A

B FIG 42-135  Cholesterolosis. A, Nonenhanced CT shows mild thickening of the gallbladder wall. B, Photograph of the pathologic specimen shows multiple yellowish flecks of cholesterol deposits.

1252

PART II  CT and MR Imaging of the Whole Body

adenomyomatosis: diffuse (generalized), segmental (annular), and localized. Diffuse adenomyomatosis causes thickening of the muscle layer (often three to five times thicker than normal), irregularity of the mucosal surface, and glandlike or cystlike structures in the wall, which prompted the terms cholecystitis glandularis proliferans and cystic cholecystitis.182 In the segmental form, annular thickening of the gallbladder wall may cause focal stricture, dividing the lumen into separate interconnected compartments.19 The localized variety is most commonly seen in the fundus and is often called adenomyoma.155 The CT and MRI abnormalities of adenomyomatosis attributable to the characteristic macroscopic pathologic features have been the concern of many radiologic investigators.43,107 Besides the nonspecific finding of diffuse or focal gallbladder wall thickening, identification of intramural diverticula (Rokitansky-Aschoff sinuses) is a prerequisite for the diagnosis of adenomyomatosis. Although CT may be limited in detecting adenomyomatosis with minimal abnormalities and in differentiating adenomyomatosis from gallbladder carcinoma,43,107 the diagnosis of adenomyomatosis can be made with reasonable accuracy when a thickened gallbladder wall in the diseased segment appears to contain small cystlike spaces,43,136 corresponding to intramural diverticula pathologically136 (Figs. 42-136 to 42-138). T2-weighted MRIs and MR cholangiography can better demonstrate these intramural cystlike spaces, which appear as bright high-signal-intensity areas in

A

the thickened gallbladder wall—the “pearl necklace” sign (see Fig. 42-136).124,175,355 Yoshimitsu et al.355 also reported that on dynamic MRI, diffuse-type adenomyomatosis typically shows early mucosal  and subsequent serosal enhancement over the Rokitansky-Aschoff sinus, and localized adenomyomatosis exhibits homogeneous enhancement, showing smooth continuity of the surrounding gallbladder epithelium.

Gallbladder Neoplasms

Benign Gallbladder Neoplasms.  Except for adenomas, benign neoplasms are rare in the gallbladder, although a variety of epithelial and stromal tumors have been reported. Adenomas are often small and asymptomatic and are usually discovered incidentally during cholecystectomy, with an incidence of 0.3% to 0.5%.154 When it is large enough, an adenoma may appear as a small polypoid lesion along the gallbladder wall on CT or MRI (Fig. 42-139).

Gallbladder Carcinoma.  Most neoplasms that arise from the gallbladder are carcinoma. Adenocarcinoma is the most common histologic type, accounting for 85% to 90% of cases. Squamous or adenosquamous cell carcinoma, neuroendocrine carcinoma (e.g., small cell carcinoma), and undifferentiated carcinoma including sarcomatoid carcinoma have also been reported.169 The peak

B

C FIG 42-136  Adenomyomatosis. A, Contrast-enhanced CT shows localized thickening of the gallbladder wall at the fundus, with multiple intramural cystlike structures (arrows). Axial T2-weighted MRI (B) and MR cholangiogram (C) better demonstrate intramural cystlike spaces appearing as bright high-signal-intensity areas in the thickened gallbladder wall—the so-called pearl necklace sign (arrows).

CHAPTER 42  Biliary Tract and Gallbladder

A

B FIG 42-137  Localized adenomyomatosis. A, Contrast-enhanced CT shows localized thickening of the gallbladder wall at the fundus and a few cystlike intramural lesions (arrows). B, Photograph of the gross specimen demonstrates intramural cysts (arrows) at the gallbladder fundus.

A

B

C FIG 42-138  Diffuse adenomyomatosis. A, Contrast-enhanced CT shows a markedly thickened gallbladder wall and profuse mucosal enhancement radiating to the serosal side. B, MR cholangiogram demonstrates multiple intramural cysts (arrows) appearing as bright high-signal-intensity areas in the thickened gallbladder wall, suggestive of diffuse adenomyomatosis. C, Photograph of the gross specimen shows a markedly thickened gallbladder wall consisting of scattered glandular tissue with fibrosis and smooth muscle microscopically, consistent with diffuse adenomyomatosis.

1253

1254

PART II  CT and MR Imaging of the Whole Body

A

B FIG 42-139  Gallbladder adenoma (polypoid type). A, Contrast-enhanced CT shows an intraluminal polypoid lesion (arrow) with intense contrast enhancement. B, Coronal T2-weighted MRI better demonstrates the papillary surface of the lesion (arrow).

occurrence of gallbladder carcinoma is in the sixth and seventh decades of life, and there is a female predilection of 2 : 1 to 3 : 1.154 The four most important factors associated with the development of gallbladder carcinoma are genetic factors, gallstones, congenital abnormal pancreaticobiliary ductal union, and porcelain gallbladder.118,154,205,288,303,313 Patients with gallbladder carcinoma may have a mutation of the TP53 gene, overexpression of the c-erbB-2 gene, and decreased expression of the nm23 gene.58,102,103 The risk of gallbladder carcinoma is increased in patients with gallstones; 75% of patients  with gallbladder carcinoma have gallstones.118,303,313 The female predominance of gallbladder carcinoma parallels the high incidence of gallstones in women.154 An association between gallbladder carcinoma and anomalous pancreaticobiliary ductal union has been suggested; as a result of the reflux of pancreatic juice into the CBD, the mixture of pancreatic juice and bile may lead to chronic inflammation of the gallbladder, with subsequent metaplasia, dysplasia, and carcinoma.181,252 As mentioned earlier, porcelain gallbladder is frequently associated with carcinoma of the gallbladder, for unknown reasons  (see Fig. 42-134).23 The clinical features of gallbladder carcinoma are often nonspecific and overlap those seen in cholelithiasis and cholecystitis. Symptoms and signs of malignancy may not appear until the carcinoma has spread to other sites. Sometimes patients with carcinoma involving the neck of the gallbladder present with secondary acute cholecystitis due to outflow obstruction. Occasionally gallbladder carcinoma is an incidental finding on imaging studies. The typical CT findings of gallbladder carcinoma include three patterns: a mass replacing the gallbladder fossa, an intraluminal polypoid mass, and gallbladder wall thickening. A mass replacing the gallbladder fossa is the most common type in many series; the mass fills most of the enlarged and deformed gallbladder, obscuring the lumen.143,192,318,341 These masses typically have low attenuation with variable enhancement, most commonly irregular peripheral enhancement in the early phase and progressive enhancement in the late phase. This type of gallbladder carcinoma often invades adjacent structures, including the liver (Figs. 42-140 and 42-141). An intraluminal polypoid mass is less common; differentiation of polypoid gallbladder carcinoma from benign polyps is based on size, with polypoid gallbladder

carcinoma typically larger than 1 cm112 (Figs. 42-142 and 42-143). Most benign polyps (e.g., cholesterol polyps) and adenomas are small, measuring less than 1 cm. The gallbladder wall-thickening type of carcinoma may be difficult to distinguish from adenomyomatosis and cholecystitis (especially chronic cholecystitis and xanthogranulomatous cholecystitis) (Fig. 42-144). It has been reported that FDG PET may help distinguish between benign and malignant gallbladder lesions (see Figs. 42-143 and 42144).266,267 Using dual–time point FDG PET, Nishiyama and coworkers proposed that a delayed scan is more helpful than an early scan for evaluating malignant lesions because of increased lesion uptake and increased lesion-to-background contrast266; they also indicated that the diagnostic performance depends on C-reactive protein levels. Falsepositive diagnoses of gallbladder carcinoma are still possible with FDG PET because benign lesions such as adenomyomatosis and tuberculosis may appear as hot lesions.233,283 On MRI, gallbladder carcinomas usually have increased signal intensity and poorly delineated contours on T2-weighted images and either isointense or hypointense signal on T1-weighted images compared with the liver.292 MR cholangiography is especially useful for gallbladder carcinoma associated with anomalous pancreaticobiliary ductal union (see Fig. 42-141). Otherwise the enhancement pattern  of gallbladder carcinoma on dynamic MRI is similar to that reported with CT. Gallbladder carcinoma commonly spreads by direct extension into adjacent structures such as the liver, hepatoduodenal ligament, colon, and duodenum (Figs. 42-145 and 42-146). The liver is the most common site of invasion because of its proximity and the lack of serosa in the gallbladder wall where it is adjacent to the liver. The incidence of liver invasion at the time of presentation varies from 34% to 89%.311 Lymphatic spread to regional and distant lymph nodes is also common in gallbladder carcinoma. Engels et al.86 reported that the foramen of Winslow lymph node and the superior pancreaticoduodenal node are the most frequent sites of nodal metastases (see Fig. 42-145). The other pathways of tumor spread in gallbladder carcinomas are hematogenous metastasis to the liver, intraductal tumor spread, and peritoneal seeding (see Fig. 42-145).303 Biliary obstruction is present in about 50% of cases of gallbladder carcinoma, which is attributed to direct invasion

CHAPTER 42  Biliary Tract and Gallbladder

A

1255

B

C FIG 42-140  Gallbladder carcinoma (mass-replacing type). A and B, Contrast-enhanced CT shows a mass replacing the gallbladder lumen, with irregular stippled enhancement in the early phase (A) and diffuse enhancement in the late phase (B). Invasion of the adjacent liver is also suggested (arrowheads). C, Photograph of the gross specimen shows a large irregular mass that had replaced the gallbladder and invaded the liver (arrowheads).

of the bile duct in the porta hepatis, to metastatic lymphadenopathy, or to intraductal spread of tumor (see Figs. 42-145 and 42-146).192,324 CT plays a key role in the staging of gallbladder carcinoma. Overall accuracy for assessing resectability has been reported as 85% to 93% with dual-phase helical CT or MDCT.160,193 However, the ability to identify early-stage disease on CT remains disappointing. Yoshimitsu and associates correlated the performance of helical CT with pathologic findings according to the tumor-node-metastasis (TNM) system and concluded that helical CT provides acceptable sensitivity and specificity for T2 and more advanced lesions but poor sensitivity for T1 lesions (33%).356

Other Malignant Gallbladder Neoplasms.  Nonepithelial malignant tumors such as Kaposi’s sarcoma, leiomyosarcoma, malignant fibrous histiocytoma, angiosarcoma, rhabdomyosarcoma, and lymphoma may occur in the gallbladder. The radiologic appearance of the gallbladder sarcoma has rarely been reported. Although gallbladder

sarcoma can present as a bulky mass with central necrosis,169 it may be difficult to distinguish from gallbladder carcinoma (Fig. 42-147). Primary lymphoma involving the gallbladder is extremely rare. Secondary lymphoma of the gallbladder is less rare but still uncommon and has been described as a feature of disseminated lymphoma.250 Lymphoma may appear as an intraluminal polypoid lesion or as wall thickening on CT169 (Fig. 42-148). Metastases to the gallbladder are rare but can originate from virtually any source. Malignant melanoma is the most common cause  of metastatic tumors of the gallbladder, accounting for more than  50% of all cases of metastases found in the gallbladder.116 Although the majority of metastatic lesions are located on the serosal surface owing to peritoneal implantation, some metastases present as intraluminal polypoid masses169 (Fig. 42-149). It has been reported that 18 FDG PET can enhance the detection of gallbladder metastases from melanoma.285 Text continued on p. 1260

1256

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F FIG 42-141  Gallbladder carcinoma (thickened-wall type). A and B, Contrast-enhanced CT shows a mass partly obscuring the gallbladder lumen, with irregular peripheral enhancement in the early phase (A) and progressive enhancement in the late phase (B). C, On a T2-weighted MRI the mass has increased signal intensity and a poorly delineated contour. Dynamic contrast-enhanced early-phase (D) and late-phase (E) MRI show an enhancement pattern similar to that on CT. F, MR cholangiography demonstrates a long common channel of the common bile duct and main pancreatic duct (arrow), suggesting an associated anomalous pancreaticobiliary ductal union.

CHAPTER 42  Biliary Tract and Gallbladder

A

B FIG 42-142  Gallbladder carcinoma (polypoid type). A, Contrast-enhanced CT shows a large intraluminal polypoid mass with diffuse contrast enhancement. B, T2-weighted MRI better demonstrates the papillary surface of the mass.

A

B FIG 42-143  Gallbladder carcinoma (polypoid type). A, Contrast-enhanced CT shows a small intraluminal polypoid lesion with subtle contrast enhancement. B, Increased lesion uptake is noted on FDG PET, suggesting malignancy.

A

B FIG 42-144  Gallbladder carcinoma (thickened-wall type). A, Contrast-enhanced CT shows focal thickening of the gallbladder wall at the fundus. B, Increased lesion uptake is noted on FDG PET, suggesting malignancy.

1257

1258

PART II  CT and MR Imaging of the Whole Body

*

B

A

FIG 42-145  Gallbladder carcinoma with metastasis to the liver and lymph nodes. A and B, Contrastenhanced CT shows a mass replacing the gallbladder fossa with irregular peripheral enhancement, suggestive of gallbladder carcinoma. Direct liver invasion (arrowheads), hilar invasion (asterisk), multiple lymph node metastases (large arrows), and liver metastases (small arrows) are also suggested. Bile duct dilatation is presumed to be caused by hilar invasion.

*

A

B

*

C FIG 42-146  Gallbladder carcinoma (squamous cell type). A, Contrast-enhanced CT shows a mass replacing the gallbladder fossa, with irregular peripheral enhancement suggestive of gallbladder carcinoma. Pathologic diagnosis of percutaneous needle biopsy was squamous cell carcinoma of the gallbladder, and the patient underwent chemotherapy. B, Follow-up CT 3 months later shows increased size and cavitary change of the mass, which contains a large amount of air and fluid (asterisk). Liver metastasis (arrowhead) and bile duct dilatation are also noted. C, CT at a lower level reveals a large fistula (arrows) in the mass replacing the gallbladder and duodenum. Asterisk shows air and fluid.

CHAPTER 42  Biliary Tract and Gallbladder

A

B

C

D

E FIG 42-147  Malignant fibrous histiocytoma of the gallbladder. A, Contrast-enhanced CT shows a mass replacing the gallbladder and mild dilatation of the intrahepatic bile duct. B, T2-weighted MRI shows the mass with moderately high signal intensity and a poorly delineated contour. C, T1-weighted MRI shows the low signal intensity of the mass. D and E, Dynamic contrast-enhanced MRIs show the mass obscuring the gallbladder with irregular peripheral enhancement in the early phase (D) and progressive enhancement in the late phase (E).

1259

1260

PART II  CT and MR Imaging of the Whole Body

FIG 42-148  Gallbladder lymphoma. Contrast-enhanced CT shows irregular thickening and enhancement of the gallbladder wall. Pathologic diagnosis of diffuse large B-cell lymphoma was made after cholecystectomy.

FIG 42-149  Metastatic melanoma to the gallbladder. Contrastenhanced CT shows a large intraluminal mass in the gallbladder, with marked hypoattenuation and a poorly delineated contour. Mild enhancement at the base of the lesion (arrows) and a bile cleft between the mass and the adjacent gallbladder wall can be seen.

REFERENCES 1. Adear H, Barki Y: Multiseptate gallbladder in a child: Incidental diagnosis on sonography. Pediatr Radiol 20:192, 1990. 2. Ahmad MM, Macon WL: Gangrene of the gallbladder. Am Surg 49:155–158, 1983. 3. Ahmetoglu A, Kosucu P, Kul S, et al: MDCT cholangiography with volume rendering for the assessment of patients with biliary obstruction. AJR Am J Roentgenol 183:1327–1332, 2004. 4. Alexander F, Rossi RL, O’Bryan M, et al: Biliary carcinoma: A review of 109 cases. Am J Surg 147:503–509, 1984. 5. Al-Fallouji MA: Perforated posterior peptic ulcer associated with gallbladder agenesis and midgut malrotation. Br J Clin Pract 37:353– 356, 358, 1983. 6. Anderson CD, Rice MH, Pinson CW, et al: Fluorodeoxyglucose   PET imaging in the evaluation of gallbladder carcinoma and cholangiocarcinoma. J Gastrointest Surg 8:90–97, 2004. 7. Aoki S, Okayama Y, Kitajima Y, et al: Intrahepatic biliary papilloma morphologically similar to biliary cystadenoma. J Gastroenterol Hepatol 20:321–324, 2005.

8. Babb RR: Acute acalculous cholecystitis: A review. J Clin Gastroenterol 15:238–241, 1992. 9. Balthazar EJ, Schechter LS: Air in gallbladder: A frequent finding in gallstone ileus. AJR Am J Roentgenol 131:219–222, 1978. 10. Bambha K, Kim WR, Talwalkar J, et al: Incidence, clinical spectrum, and outcomes of primary sclerosing cholangitis in a United States community. Gastroenterology 125:1364–1369, 2003. 11. Baron RL: Common bile duct stones: Reassessment of criteria for CT diagnosis. Radiology 162:419–424, 1987. 12. Baron RL: Computed tomography of the biliary tree. Radiol Clin North Am 29:1235–1250, 1991. 13. Baron RL, Stanley RJ, Lee JK, et al: A prospective comparison of the evaluation of biliary obstruction using computed tomography and ultrasonography. Radiology 145:91–98, 1982. 14. Baron RL, Tublin ME, Peterson MS: Imaging the spectrum of biliary tract disease. Radiol Clin North Am 40:1325–1354, 2002. 15. Bartoli E, Calonaci N, Nenci R: Ultrasonography of the gallbladder in pregnancy. Gastrointest Radiol 9:35–38, 1984. 16. Bartone NF, Grieco RV: Absent gallbladder and cystic duct. Am J Roentgenol Radium Ther Nucl Med 110:252–255, 1970. 17. Barwick KW, Rosai J: Gallbaladder and extrahepatic bile ducts. In Rosai J, editor: Ackerman’s surgical pathology, ed 8, St Louis, 1996, Mosby, p 960. 18. Barwick KW, Rosai J: Liver. In Rosai J, editor: Ackerman’s surgical pathology, ed 8, St Louis, 1996, Mosby, pp 913–915. 19. Bean BC, Berens DL, Culver GJ: The relationship of stenosis to Rokitansky Aschoff sinuses of the gallbladder. Am J Roentgenol Radium Ther Nucl Med 77:47–54, 1957. 20. Bennett GL, Balthazar EJ: Ultrasound and CT evaluation of emergent gallbladder pathology. Radiol Clin North Am 41:1203–1216, 2003. 21. Bennett GL, Rusinek H, Lisi V, et al: CT findings in acute gangrenous cholecystitis. AJR Am J Roentgenol 178:275–281, 2002. 22. Bennion RS, Thompson JE, Jr, Tompkins RK: Agenesis of the gallbladder without extrahepatic biliary atresia. Arch Surg 123:1257– 1260, 1988. 23. Berk RN, Armbuster TG, Saltzstein SL: Carcinoma in the porcelain gallbladder. Radiology 106:29–31, 1973. 24. Berk RN, van der Vegt JH, Lichtenstein JE: The hyperplastic cholecystoses: Cholesterolosis and adenomyomatosis. Radiology 146:593–601, 1983. 25. Berland LL, Lawson TL, Stanley RJ: CT appearance of Mirizzi syndrome. J Comput Assist Tomogr 8:165–166, 1984. 26. Blankenberg F, Wirth R, Jeffrey RB, Jr, et al: Computed tomography as an adjunct to ultrasound in the diagnosis of acute acalculous cholecystitis. Gastrointest Radiol 16:149–153, 1991. 27. Blumgart LH: Stones in the common bile duct: Clinical features and open surgical approaches. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, pp 528–547. 28. Brink JA, Ferrucci JT: Use of CT for predicting gallstone composition: A dissenting view. Radiology 178:633–634, 1991. 29. Buck JL, Elsayed AM: Ampullary tumors: Radiologic-pathologic correlation. Radiographics 13:193–212, 1993. 30. Buetow PC, Buck JL, Pantongrag-Brown L, et al: Biliary cystadenoma and cystadenocarcinoma: Clinical-imaging-pathologic correlations with emphasis on the importance of ovarian stroma. Radiology 196:805–810, 1995. 31. Buetow PC, Midkiff RB: MR imaging of the liver: Primary malignant neoplasms in the adult. Magn Reson Imaging Clin N Am 5:289–318, 1997. 32. Cameron JL, Gayler BW, Sanfey H, et al: Sclerosing cholangitis: Anatomical distribution of obstructive lesions. Ann Surg 200:54–60, 1984. 33. Carmona RH, Crass RA, Lim RC, Jr, et al: Oriental cholangitis.   Am J Surg 148:117–124, 1984. 34. Carr DH, Hadjis NS, Banks LM, et al: Computed tomography of   hilar cholangiocarcinoma: A new sign. AJR Am J Roentgenol 145:53–56, 1985.

CHAPTER 42  Biliary Tract and Gallbladder 35. Cello JP: Acquired immunodeficiency syndrome cholangiopathy: Spectrum of disease. Am J Med 86:539–546, 1989. 36. Cevikol C, Karaali K, Senol U, et al: Human fascioliasis: MR imaging findings of hepatic lesions. Eur Radiol 13:141–148, 2003. 37. Cha JH, Han JK, Kim TK, et al: Preoperative evaluation of Klatskin tumor: Accuracy of spiral CT in determining vascular invasion as a sign of unresectability. Abdom Imaging 25:500–507, 2000. 38. Cha JM, Kim MH, Lee SK, et al: Clinicopathological review of 61 patients with early bile duct cancer. Clin Oncol (R Coll Radiol) 18:669–677, 2006. 39. Chan YL, Lam WW, Metreweli C, et al: Detectability and appearance of bile duct calculus on MR imaging of the abdomen using axial T1- and T2-weighted sequences. Clin Radiol 52:351–355, 1997. 40. Chang S, Lim JH, Choi D, et al: Differentiation of ampullary tumor from benign papillary stricture by thin-section multidetector CT. Abdom Imaging 33:457–462, 2008. 41. Chau EM, Leong LL, Chan FL: Recurrent pyogenic cholangitis: Ultrasound evaluation compared with endoscopic retrograde cholangiopancreatography. Clin Radiol 38:79–85, 1987. 42. Chen JJ, Lin HH, Chiu CT, et al: Gallbladder perforation with intrahepatic abscess formation. J Clin Ultrasound 18:43–45, 1990. 43. Ching BH, Yeh BM, Westphalen AC, et al: CT differentiation of adenomyomatosis and gallbladder cancer. AJR Am J Roentgenol 189:62–66, 2007. 44. Cho ES, Park MS, Yu JS, et al: Biliary ductal involvement of hilar cholangiocarcinoma: Multidetector computed tomography versus magnetic resonance cholangiography. J Comput Assist Tomogr 31:72–78, 2007. 45. Choi EK, Byun JH, Lee SJ, et al: Imaging findings of leukemic involvement of the pancreaticobiliary system in adults. AJR Am J Roentgenol 188:1589–1595, 2007. 46. Choi BI, Han JK, Hong ST, et al: Clonorchiasis and cholangiocarcinoma: Etiologic relationship and imaging diagnosis.   Clin Microbiol Rev 17:540–552, 2004. 47. Choi BI, Han JK, Shin YM, et al: Peripheral cholangiocarcinoma: Comparison of MRI with CT. Abdom Imaging 20:357–360, 1995. 48. Choi BI, Kim HJ, Han MC, et al: CT findings of clonorchiasis. AJR Am J Roentgenol 152:281–284, 1989. 49. Choi BI, Kim TK, Han JK: MRI of clonorchiasis and cholangiocarcinoma. J Magn Reson Imaging 8:359–366, 1998. 50. Choi BI, Lee JM, Han JK: Imaging of intrahepatic and hilar cholangiocarcinoma. Abdom Imaging 29:548–557, 2004. 51. Choi BI, Lee JH, Han MC, et al: Hilar cholangiocarcinoma: Comparative study with sonography and CT. Radiology 172:689–692, 1989. 52. Choi BI, Lim JH, Han MC, et al: Biliary cystadenoma and cystadenocarcinoma: CT and sonographic findings. Radiology 171:57–61, 1989. 53. Choi D, Lim JH, Lee KT, et al: Cholangiocarcinoma and Clonorchis sinensis infection: A case-control study in Korea. J Hepatol 44:1066– 1073, 2006. 54. Choi BI, Park JH, Kim YI, et al: Peripheral cholangiocarcinoma and clonorchiasis: CT findings. Radiology 169:149–153, 1988. 55. Choi TK, Wong KP, Wong J: Cholangiographic appearance in clonorchiasis. Br J Radiol 57:681–684, 1984. 56. Choi BI, Yeon KM, Kim SH, et al: Caroli disease: Central dot sign in CT. Radiology 174:161–163, 1990. 57. Chou ST, Chan CW: Recurrent pyogenic cholangitis: A necropsy study. Pathology 12:415–428, 1980. 58. Chow NH, Huang SM, Chan SH, et al: Significance of c-erbB-2 expression in normal and neoplastic epithelium of biliary tract. Anticancer Res 15:1055–1059, 1995. 59. Chun KA, Ha HK, Yu ES, et al: Xanthogranulomatous cholecystitis: CT features with emphasis on differentiation from gallbladder carcinoma. Radiology 203:93–97, 1997. 60. Chung EB: Multiple bile-duct hamartomas. Cancer 26:287–296, 1970. 61. Cohen F, Bernstein JR: Choledochocele presenting with jaundice: Diagnosis by percutaneous cholangiogram. Dig Dis Sci 26:667–670, 1981.

1261

62. Collins CD, Forbes A, Harcourt-Webster JN, et al: Radiological and pathological features of AIDS-related polypoid cholangitis. Clin Radiol 48:307–310, 1993. 63. Colquhoun J: Adenomyomatosis of the gall-bladder (intramural diverticulosis). Br J Radiol 34:101–112, 1961. 64. Cook J, Hou PC, Ho HC, et al: Recurrent pyogenic cholangeitis.   Br J Surg 42:188–203, 1954. 65. Cossi AF, Scholz FJ, Aretz HT, et al: Computed tomography of xanthogranulomatous cholecystitis. Gastrointest Radiol 12:154–155, 1987. 66. Cotton PB: Retained bile duct stones: T-tube in place, percutaneous or endoscopic management? Am J Gastroenterol 85:1075–1078, 1990. 67. Coughlin JP, Rector FE, Klein MD: Agenesis of the gallbladder in duodenal atresia: Two case reports. J Pediatr Surg 27:1304, 1992. 68. Courbiere M, Pilleul F, Henry L, et al: Value of magnetic resonance cholangiography in benign and malignant biliary stenosis: Comparative study with direct cholangiography. J Comput Assist Tomogr 27:315–320, 2003. 69. Cowley LL, Harkins HN: Perforation of the gallbladder. Surg Gynecol Obstet 77:661–668, 1943. 70. Crittenden SL, McKinley MJ: Choledochal cyst—Clinical features and classification. Am J Gastroenterol 80:643–647, 1985. 71. Damrongsak D, Damrongsak C, Bhothisuwan W, et al: Computed tomography in opisthorchiasis. Comput Radiol 8:379–385, 1984. 72. Dao AH, Barnwell SF, Adkins RB, Jr: A case of opisthorchiasis diagnosed by cholangiography and bile examination. Am Surg 57:206–209, 1991. 73. Da Silva F, Boudghene F, Lecomte I, et al: Sonography in AIDS-related cholangitis: Prevalence and cause of an echogenic nodule in the distal end of the common bile duct. AJR Am J Roenetgenol 160:1205–1207, 1993. 74. Davis GL, Kissane JM, Ishak KG: Embryonal rhabdomyosarcoma (sarcoma botryoides) of the biliary tree: Report of five cases and a review of the literature. Cancer 24:333–342, 1969. 75. Day EA, Marks C: Gallstone ileus: Review of the literature and presentation of thirty-four new cases. Am J Surg 129:552–558, 1975. 76. Day DL, Mulcahy PF, Dehner LP, et al: Post-operative abdominal CT scanning in extrahepatic biliary atresia. Pediatr Radiol 19:379–382, 1989. 77. Dayton MT, Longmire WP, Jr, Tompkins RK: Caroli’s disease: A premalignant condition? Am J Surg 145:41–48, 1983. 78. Delabrousse E, Bartholomot B, Sohm O, et al: Gallstone ileus: CT findings. Eur Radiol 10:938–940, 2000. 79. Desa LA, Akosa AB, Lazzara S, et al: Cytodiagnosis in the management of extrahepatic biliary stricture. Gut 32:1188–1191, 1991. 80. De Wilde VG, Elewaut AG, De Vos MM, et al: Choledochal cysts in the adult. Endoscopy 23:4–7, 1991. 81. Diffenbaugh WG: Gangrenous perforation of the gallbladder. Arch Surg 59:743–749, 1949. 82. Duckler L: Agenesis of gallbladder with common duct stones. Am J Dig Dis 17:1127–1130, 1972. 83. Edell S: A comparison of the “phrygian cap” deformity with bistable and gray scale ultrasound. J Clin Ultrasound 6:34–35, 1978. 84. Elias E, Hamlyn AN, Jain S, et al: A randomized trial of percutaneous transhepatic cholangiography with the Chiba needle versus endoscopic retrograde cholangiography for bile duct visualization in jaundice. Gastroenterology 71:439–443, 1976. 85. Elsayes KM, Oliveira EP, Narra VR, et al: Magnetic resonance imaging of the gallbladder: Spectrum of abnormalities. Acta Radiol 48:476–482, 2007. 86. Engels JT, Balfe DM, Lee JK: Biliary carcinoma: CT evaluation of extrahepatic spread. Radiology 172:35–40, 1989. 87. Everson GT, Nemeth A, Kourourian S, et al: Gallbladder function is altered in sickle hemoglobinopathy. Gastroenterology 96:1307–1316, 1989. 88. Reference deleted in proofs. 89. Fan ST, Wong J: Recurrent pyogenic cholangitis. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, pp 971–988.

1262

PART II  CT and MR Imaging of the Whole Body

90. Federle MP, Cello JP, Laing FC, et al: Recurrent pyogenic cholangitis in Asian immigrants: Use of ultrasonography, computed tomography, and cholangiography. Radiology 143:151–156, 1982. 91. Fenoglio-Preiser CM, Noffsinger AE, Stemmerrmann GN, et al: Nonneoplastic lesions of the small intestine. In Gastrointestinal pathology: An atlas and text, Philadelphia, 1999, Lippincott-Raven, pp 331–333. 92. Ferrucci JT: Advances in hepatobiliary radiology, St Louis, 1990, Mosby. 93. Fidler J, Paulson EK, Layfield L: CT evaluation of acute cholecystitis: Findings and usefulness in diagnosis. AJR Am J Roentgenol 166:1085– 1088, 1996. 94. Fleming KA: Hepatobiliary pathology of primary sclerosing cholangitis. Eur J Gastroenterol Hepatol 4:266–271, 1992. 95. Fletcher AG, Jr, Ravdin IS: Perforation of the gallbladder. Am J Surg 81:178–185, 1951. 96. Foley WD, Wilson CR, Quiroz FA, et al: Demonstration of the normal extrahepatic biliary tract with computed tomography. J Comput Assist Tomogr 4:48–52, 1980. 97. Fork FT, Nyman U, Sigurjonsson S: Recognition of gas in gallstones in routine computed tomograms of the abdomen. J Comput Assist Tomogr 7:805–809, 1983. 98. Forrest ME, Cho KJ, Shields JJ, et al: Biliary cystadenomas: Sonographic-angiographic-pathologic correlations. AJR Am J Roentgenol 135:723–727, 1980. 99. Foster DR: Triple gall bladder. Br J Radiol 54:817–818, 1981. 100. Fotopoulos JP, Crampton AR: Adenomyomatosis of the gallbladder. Med Clin North Am 48:9–36, 1964. 101. Frick MP, Feinberg SB: Biliary cystadenoma. AJR Am J Roentgenol 139:393–395, 1982. 102. Fujii K, Yasui W, Shimamoto F, et al: Immunohistochemical analysis of nm23 gene product in human gallbladder carcinomas. Virchows Arch 426:355–359, 1995. 103. Fujii K, Yokozaki H, Yasui W, et al: High frequency of p53 gene mutation in adenocarcinomas of the gallbladder. Cancer Epidemiol Biomarkers Prev 5:461–466, 1996. 104. Fulcher AS, Turner MA: MR cholangiopancreatography. Radiol Clin North Am 40:1363–1376, 2002. 105. Fulcher AS, Turner MA, Capps GW, et al: Half-Fourier RARE MR cholangiopancreatography: Experience in 300 subjects. Radiology 207:21–32, 1998. 106. Furuta A, Ishibashi T, Takahashi S, et al: MR imaging of xanthogranulomatous cholecystitis. Radiat Med 14:315–319, 1996. 107. Gerard PS, Berman D, Zafaranloo S: CT and ultrasound of gallbladder adenomyomatosis mimicking carcinoma. J Comput Assist Tomogr 14:490–491, 1990. 108. Ghahremani GG, Lu CT, Woodlief RM, et al: Choledochal cyst in adults: A clinical and radiological study in ten cases. Gastrointest Radiol 1:305–313, 1977. 109. Gill KS, Chapman AH, Weston MJ: The changing face of emphysematous cholecystitis. Br J Radiol 70:986–991, 1997. 110. Goiney RC, Schoenecker SA, Cyr DR, et al: Sonography of gallbladder duplication and differential considerations. AJR Am J Roentgenol 145:241–243, 1985. 111. Gore RM, Ghahremani GG, Joseph AE, et al: Acquired malposition of the colon and gallbladder in patients with cirrhosis: CT findings and clinical implications. Radiology 171:739–742, 1989. 112. Grand D, Horton KM, Fishman EK: CT of the gallbladder: Spectrum of disease. AJR Am J Roentgenol 183:163–170, 2004. 113. Granot E, Deckelbaum RJ, Gordon R, et al: Duplication of the gallbladder associated with childhood obstructive biliary disease and biliary cirrhosis. Gastroenterology 85:946–950, 1983. 114. Grayson DE, Abbott RM, Levy AD, et al: Emphysematous infections of the abdomen and pelvis: A pictorial review. Radiographics 22:543–561, 2002. 115. Grumbach K, Levine MS, Wexler JA: Gallstone ileus diagnosed   by computed tomography. J Comput Assist Tomogr 10:146–148, 1986.

116. Guida M, Cramarossa A, Gentile A, et al: Metastatic malignant melanoma of the gallbladder: A case report and review of the literature. Melanoma Res 12:619–625, 2002. 117. Haller JO: Sonography of the biliary tract in infants and children. AJR Am J Roentgenol 157:1051–1058, 1991. 118. Hamrick RE, Jr, Liner FJ, Hastings PR, et al: Primary carcinoma of the gallbladder. Ann Surg 195:270–273, 1982. 119. Hamrick-Turner J, Abbitt PL, Ros PR: Intrahepatic cholangiocarcinoma: MR appearance. AJR Am J Roentgenol 158:77–79, 1992. 120. Han JK, Choi BI, Kim TK, et al: Hilar cholangiocarcinoma: Thinsection spiral CT findings with cholangiographic correlation. Radiographics 17:1475–1485, 1997. 121. Han JK, Jang HJ, Choi BI, et al: Experimental hepatobiliary fascioliasis in rabbits: A radiology-pathology correlation. Invest Radiol 34:99–108, 1999. 122. Han JK, Lee JM: Intrahepatic intraductal cholangiocarcinoma. Abdom Imaging 29:558–564, 2004. 123. Hann LE, Greatrex KV, Bach AM, et al: Cholangiocarcinoma at the hepatic hilus: Sonographic findings. AJR Am J Roentgenol 168:985–989, 1997. 124. Haradome H, Ichikawa T, Sou H, et al: The pearl necklace sign: An imaging sign of adenomyomatosis of the gallbladder at MR cholangiopancreatography. Radiology 227:80–88, 2003. 125. Hatfield PM, Scholz FJ, Wise RE: Congenital diseases of the gallbladder and bile ducts. Semin Roentgenol 11:235–243, 1976. 126. Hatfield PM, Wise RE: Anatomic variation in the gallbladder and bile ducts. Semin Roentgenol 11:157–164, 1976. 127. Havrilla TR, Reich NE, Haaga JR, et al: Computed tomography of the gallbladder. AJR Am J Roentgenol 130:1059–1067, 1978. 128. Heloury Y, Leborgne J, Rogez JM, et al: Radiological anatomy of the bile ducts based on intraoperative investigation in 250 cases. Anat Clin 7:93–102, 1985. 129. Herbener T, Zajko AB, Koneru B, et al: Recurrent cholangiocarcinoma in the biliary tree after liver transplantation. Radiology 169:641–642, 1988. 130. Heuer GJ: The factors leading to death in operations upon the gallbladder and bile ducts. Ann Surg 99:881–892, 1934. 131. Honda N, Cobb C, Lechago J: Bile duct carcinoma associated with multiple von Meyenburg complexes in the liver. Hum Pathol 17:1287– 1290, 1986. 132. Honda H, Onitsuka H, Yasumori K, et al: Intrahepatic peripheral cholangiocarcinoma: Two-phased dynamic incremental CT and pathologic correlation. J Comput Assist Tomogr 17:397–402, 1993. 133. Hopper KD: The role of computed tomography in the evaluation of Caroli’s disease. Clin Imaging 13:68–73, 1989. 134. Hotta T, Kobayashi Y, Taniguchi K, et al: A traumatic neuroma of the bile duct: A case report. Hepatogastroenterology 51:39–42, 2004. 135. Hunt DR, Chu FC: Gangrenous cholecystitis in the laparoscopic era. Aust N Z J Surg 70:428–430, 2000. 136. Hwang JI, Chou YH, Tsay SH, et al: Radiologic and pathologic correlation of adenomyomatosis of the gallbladder. Abdom Imaging 23:73–77, 1998. 137. Hwang CM, Kim TK, Ha HK, et al: Biliary ascariasis: MR cholangiography findings in two cases. Korean J Radiol 2:175–178, 2001. 138. Ihara E, Ochiai T, Yamamoto K, et al: A case of gallstone ileus with a spontaneous evacuation. Am J Gastroenterol 97:1259–1260, 2002. 139. Irie H, Honda H, Shinozaki K, et al: MR imaging of ampullary carcinomas. J Comput Assist Tomogr 26:711–717, 2002. 140. Isaji S, Kawarada Y, Taoka H, et al: Clinicopathological features and outcome of hepatic resection for intrahepatic cholangiocarcinoma in Japan. J Hepatobiliary Pancreat Surg 6:108–116, 1999. 141. Itai Y: Computed tomographic evaluation of gallbladder disease. Crit Rev Diagn Imaging 27:113–152, 1987. 142. Itai Y, Araki T, Furui S, et al: Computed tomography and ultrasound in the diagnosis of intrahepatic calculi. Radiology 136:399–405, 1980. 143. Itai Y, Araki T, Yoshikawa K, et al: Computed tomography of gallbladder carcinoma. Radiology 137:713–718, 1980.

CHAPTER 42  Biliary Tract and Gallbladder 144. Ito K, Awaya H, Mitchell DG, et al: Gallbladder disease: Appearance of associated transient increased attenuation in the liver at biphasic, contrast-enhanced dynamic CT. Radiology 204:723–728, 1997. 145. Ito K, Mitchell DG, Outwater EK, et al: Primary sclerosing cholangitis: MR imaging features. AJR Am J Roentgenol 172:1527–1533, 1999. 146. Itoh S, Fukushima H, Takada A, et al: Assessment of anomalous pancreaticobiliary ductal junction with high-resolution multiplanar reformatted images in MDCT. AJR Am J Roentgenol 187:668–675, 2006. 147. Itoh S, Ikeda M, Ota T, et al: Assessment of the pancreatic and intrapancreatic bile ducts using 0.5-mm collimation and multiplanar reformatted images in multislice CT. Eur Radiol 13:277–285, 2003. 148. Jacob H, Appelman R, Stein HD: Emphysematous cholecystitis. Am J Gastroenterol 71:325–330, 1979. 149. Jacobs M, Verdeja JC, Goldstein HS: Laparoscopic cholecystectomy in acute cholecystitis. J Laparoendoscop Surg 1:175–177, 1991. 150. Jarnagin WR, Blumgart LH: Benign biliary strictures. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, pp 628–654. 151. Jarnagin WR, D’Angellica M, Blumgart LH: Intrahepatic and extrahepatic biliary cancer. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, pp 782–792. 152. Jarnagin WR, Fong Y, DeMatteo RP, et al: Staging, resectability, and outcome in 225 patients with hilar cholangiocarcinoma. Ann Surg 234:507–517, 2001. 153. Jenkins M, Golding RH, Cooperberg PL: Sonography and computed tomography of hemorrhagic cholecystitis. AJR Am J Roentgenol 140:1197–1198, 1983. 154. Jessurun J, Albores-Saavedra J: Diseases of the gallbladder. In Burt A, Portmann B, Ferrell L, editors: MacSween’s pathology of the liver, Philadelphia, 2007, Churchill Livingstone, pp 583–612. 155. Jutras JA: Hyperplastic cholecystoses: Hickey lecture, 1960. Am J Roentgenol Radium Ther Nucl Med 83:795–827, 1960. 156. Kabaalioglu A, Cubuk M, Senol U, et al: Fascioliasis: US, CT, and MRI findings with new observations. Abdom Imaging 25:400–404, 2000. 157. Kaiser A, Jurowich C, Schonekas H, et al: The adenoma-carcinoma sequence applies to epithelial tumours of the papilla of Vater. Z Gastroenterol 40:913–920, 2002. 158. Kaiser JA, Mall JC, Salmen BJ, et al: Diagnosis of Caroli disease by computed tomography: Report of two cases. Radiology 132:661–664, 1979. 159. Kakitsubata Y, Kakitsubata S, Sonoda T, et al: Neurofibromatosis type 1 involving the liver: Ultrasound and CT manifestations. Pediatr Radiol 24:66–67, 1994. 160. Kalra N, Suri S, Gupta R, et al: MDCT in the staging of gallbladder carcinoma. AJR Am J Roentgenol 186:758–762, 2006. 161. Kane RA, Jacobs R, Katz J, et al: Porcelain gallbladder: Ultrasound and CT appearance. Radiology 152:137–141, 1984. 162. Karademir S, Astarcioglu H, Sokmen S, et al: Mirizzi’s syndrome: Diagnostic and surgical considerations in 25 patients. J Hepatobiliary Pancreat Surg 7:72–77, 2000. 163. Kayahara M, Ohta T, Kitagawa H, et al: Adenomyomatosis of the papilla of Vater: A case illustrating diagnostic difficulties. Dig Surg 18:139–142, 2001. 164. Kim MJ, Cha SW, Mitchell DG, et al: MR imaging findings in recurrent pyogenic cholangitis. AJR Am J Roentgenol 173:1545–1549, 1999. 165. Kim TK, Choi BI, Han JK, et al: Peripheral cholangiocarcinoma of the liver: Two-phase spiral CT findings. Radiology 204:539–543, 1997. 166. Kim JH, Kim MJ, Chung JJ, et al: Differential diagnosis of periampullary carcinomas at MR imaging. Radiographics 22:1335–1352, 2002. 167. Kim JH, Kim TK, Eun HW, et al: CT findings of cholangiocarcinoma associated with recurrent pyogenic cholangitis. AJR Am J Roentgenol 187:1571–1577, 2006. 168. Kim HJ, Kim AY, Hong SS, et al: Biliary ductal evaluation of hilar cholangiocarcinoma: Three-dimensional direct multi-detector row CT cholangiographic findings versus surgical and pathologic results— feasibility study. Radiology 238:300–308, 2006.

1263

169. Kim MJ, Kim KW, Kim HC, et al: Unusual malignant tumors of the gallbladder. AJR Am J Roentgenol 187:473–480, 2006. 170. Kim HJ, Kim MH, Lee SK, et al: Mucin-hypersecreting bile duct tumor characterized by a striking homology with an intraductal papillary mucinous tumor (IPMT) of the pancreas. Endoscopy 32:389–393, 2000. 171. Kim PN, Lee SH, Gong GY, et al: Xanthogranulomatous cholecystitis: Radiologic findings with histologic correlation that focuses on intramural nodules. AJR Am J Roentgenol 172:949–953, 1999. 172. Kim PN, Lee KS, Kim IY, et al: Gallbladder perforation: Comparison of US findings with CT. Abdom Imaging 19:239–242, 1994. 173. Kim SH, Lim HK, Choi D, et al: Changes in bile ducts after radiofrequency ablation of hepatocellular carcinoma: Frequency and clinical significance. AJR Am J Roentgenol 183:1611–1617, 2004. 174. Kim YS, Myung SJ, Kim SY, et al: Biliary papillomatosis: Clinical, cholangiographic and cholangioscopic findings. Endoscopy 30:763–767, 1998. 175. Kim MJ, Oh YT, Park YN, et al: Gallbladder adenomyomatosis: Findings on MRI. Abdom Imaging 24:410–413, 1999. 176. Kim PN, Outwater EK, Mitchell DG: Mirizzi syndrome: Evaluation by MRI imaging. Am J Gastroenterol 94:2546–2550, 1999. 177. Kim HC, Park SH, Park SI, et al: Three-dimensional reconstructed images using multidetector computed tomography in evaluation of the biliary tract. Abdom Imaging 29:472–478, 2004. 178. Kim KW, Park MS, Yu JS, et al: Acute cholecystitis at T2-weighted and manganese-enhanced T1-weighted MR cholangiography: Preliminary study. Radiology 227:580–584, 2003. 179. Kim YJ, Yun M, Lee WJ, et al: Usefulness of 18F-FDG PET in intrahepatic cholangiocarcinoma. Eur J Nucl Med Mol Imaging 30:1467–1472, 2003. 180. Kimura H, Kagawa K, Deguchi T, et al: Extrahepatic biliary cystadenocarcinoma arising from the left hepatic duct. J Gastroenterol 33:895–898, 1998. 181. Kimura K, Ohto M, Saisho H, et al: Association of gallbladder carcinoma and anomalous pancreaticobiliary ductal union. Gastroenterology 89:1258–1265, 1985. 182. King E, MacCallum P: Cholecystitis glandularis proliferans. Br J Surg 19:310–323, 1931. 183. Kmiya J, Kitagawa Y, Nimura Y: Intrahepatic stones. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, pp 586–596. 184. Kochhar R, Nagi B, Mehta SK, et al: ERCP diagnosis of a gallbladder diverticulum. Gastrointest Endosc 34:150–151, 1988. 185. Koehler RE, Melson GL, Lee JK, et al: Common hepatic duct obstruction by cystic duct stone: Mirizzi syndrome. AJR Am J Roentgenol 132:1007–1009, 1979. 186. Koga A, Ichimiya H, Yamaguchi K, et al: Hepatolithiasis associated with cholangiocarcinoma: Possible etiologic significance. Cancer 55:2826– 2829, 1985. 187. Komi N, Takehara H, Kunitomo K: Choledochal cyst: Anomalous arrangement of the pancreaticobiliary ductal system and biliary malignancy. J Gastroenterol Hepatol 4:63–74, 1989. 188. Komi N, Tamura T, Miyoshi Y, et al: Nationwide survey of cases of choledochal cyst: Analysis of coexistent anomalies, complications and surgical treatment in 645 cases. Surg Gastroenterol 3:69–73, 1984. 189. Korobkin M, Stephens DH, Lee JK, et al: Biliary cystadenoma and cystadenocarcinoma: CT and sonographic findings. AJR Am J Roentgenol 153:507–511, 1989. 190. Krudy AG, Doppman JL, Bissonette MB, et al: Hemobilia: Computed tomographic diagnosis. Radiology 148:785–789, 1983. 191. Kubaska SM, Chew FS: Biliary ascariasis. AJR Am J Roentgenol 169:492, 1997. 192. Kumar A, Aggarwal S: Carcinoma of the gallbladder: CT findings in 50 cases. Abdom Imaging 19:304–308, 1994. 193. Kumaran V, Gulati S, Paul B, et al: The role of dual-phase helical CT in assessing resectability of carcinoma of the gallbladder. Eur Radiol 12:1993–1999, 2002. 194. Laing FC, Jeffrey RB, Jr, Wing VW, et al: Biliary dilatation: Defining the level and cause by real-time US. Radiology 160:39–42, 1986.

1264

PART II  CT and MR Imaging of the Whole Body

195. Lam SK, Wong KP, Chan PK, et al: Recurrent pyogenic cholangitis: A study by endoscopic retrograde cholangiography. Gastroenterology 74:1196–1203, 1978. 196. Lamki N, Raval B, St Ville E: Computed tomography of complicated cholecystitis. J Comput Tomogr 10:319–324, 1986. 197. Larusso NF: The Fifth Carlos E. Rubio Memorial Lecture. Sclerosing cholangitis: Pathogenesis, pathology, and practice. P R Health Sci J 18:11–17, 1999. 198. Lassandro F, Gagliardi N, Scuderi M, et al: Gallstone ileus: Analysis of radiological findings in 27 patients. Eur J Radiol 50:23–29, 2004. 199. Lassandro F, Romano S, Ragozzino A, et al: Role of helical CT in diagnosis of gallstone ileus and related conditions. AJR Am J Roentgenol 185:1159–1165, 2005. 200. Lee JW, Han JK, Kim TK, et al: CT features of intraductal intrahepatic cholangiocarcinoma. AJR Am J Roentgenol 175:721–725, 2000. 201. Lee SS, Kim MH, Lee SK, et al: MR cholangiography versus cholangioscopy for evaluation of longitudinal extension of hilar cholangiocarcinoma. Gastrointest Endosc 56:25–32, 2002. 202. Lee HY, Kim SH, Lee JM, et al: Preoperative assessment of resectability of hepatic hilar cholangiocarcinoma: Combined CT and cholangiography with revised criteria. Radiology 239:113–121, 2006. 203. Lee SG, Lee YJ, Park KM, et al: One hundred and eleven liver   resections for hilar bile duct cancer. J Hepatobiliary Pancreat Surg 7:135–141, 2000. 204. Lee SP, Nicholls JF, Park HZ: Biliary sludge as a cause of acute pancreatitis. N Engl J Med 326:589–593, 1992. 205. Levin B: Gallbladder carcinoma. Ann Oncol 10(Suppl 4):129–130, 1999. 206. Levy AD, Patel N, Dow N, et al: From the archives of the AFIP: Abdominal neoplasms in patients with neurofibromatosis type 1: Radiologic-pathologic correlation. Radiographics 25:455–480, 2005. 207. Lillemoe KD, Cameron JL: Surgery for hilar cholangiocarcinoma: The Johns Hopkins approach. J Hepatobiliary Pancreat Surg 7:115–121, 2000. 208. Lim JH: Radiologic findings of clonorchiasis. AJR Am J Roentgenol 155:1001–1008, 1990. 209. Lim JH: Oriental cholangiohepatitis: Pathologic, clinical, and radiologic features. AJR Am J Roentgenol 157:1–8, 1991. 210. Lim JH: Cholangiocarcinoma: Morphologic classification according to growth pattern and imaging findings. AJR Am J Roentgenol 181:819– 827, 2003. 211. Lim JH, Jang KT, Choi D: Biliary intraductal papillary mucinous neoplasm presenting only as dilatation of the hepatic lobar or segmental bile duct: Imaging features in six patients. AJR Am J Roentgenol 191:778–782, 2008. 212. Lim JH, Jang KT, Choi D, et al: Early bile duct carcinoma: Comparison of imaging features with pathologic findings. Radiology 238:542–548, 2006. 213. Lim JH, Jang KT, Rhim H, et al: Biliary cystic intraductal papillary mucinous tumor and cystadenoma/cystadenocarcinoma: Differentiation by CT. Abdom Imaging 32:644–651, 2007. 214. Lim JH, Kim MH, Kim TK, et al: Papillary neoplasms of the bile duct that mimic biliary stone disease. Radiographics 23:447–455, 2003. 215. Lim JH, Kim SY, Park CM: Parasitic diseases of the biliary tract. AJR Am J Roentgenol 188:1596–1603, 2007. 216. Lim JH, Ko YT: Clonorchiasis of the pancreas. Clin Radiol 41:195–198, 1990. 217. Lim JH, Lee WJ, Takehara Y, et al: Imaging of extrahepatic cholangiocarcinoma. Abdom Imaging 29:565–571, 2004. 218. Lim JH, Park CK: Pathology of cholangiocarcinoma. Abdom Imaging 29:540–547, 2004. 219. Lim JH, Yi CA, Lim HK, et al: Radiological spectrum of intraductal papillary tumors of the bile ducts. Korean J Radiol 3:57–63, 2002. 220. Lim JH, Yoon YK, Kim SY, et al: Endoscopic retrograde cholangiographic findings in choledocholithiasis. J Korean Radiol Soc 18:116–124, 1982. 221. Lim JH, Yoon KH, Kim SH, et al: Intraductal papillary mucinous tumor of the bile ducts. Radiographics 24:53–66, 2004.

222. Lindsay KA, Hall JR, Chapman AH: Radiology of excessive mucus production as a complication of congenital bile duct abnormalities. Clin Radiol 46:43–45, 1992. 223. Lipsett PA, Pitt HA: Cholangitis. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, pp 917–926. 224. Liver Cancer Study Group of Japan: The general rules for the clinical and pathological study of primary liver cancer, ed 4, Tokyo, 2000, Kanehara. 225. Lopera JE, Soto JA, Munera F: Malignant hilar and perihilar biliary obstruction: Use of MR cholangiography to define the extent of biliary ductal involvement and plan percutaneous interventions. Radiology 220:90–96, 2001. 226. Loren I, Lasson A, Nilsson A, et al: Gallstone ileus demonstrated by CT. J Comput Assist Tomogr 18:262–265, 1994. 227. Ludwig J, MacCarty RL, LaRusso NF, et al: Intrahepatic cholangiectases and large-duct obliteration in primary sclerosing cholangitis. Hepatology (Baltimore) 6:560–568, 1986. 228. Lygidakis NJ: Incidence and significance of primary stones of the common bile duct in choledocholithiasis. Surg Gynecol Obstet 157:434–436, 1983. 229. Lygidakis NJ, Sgourakis GJ, Dedemadi GV, et al: Long-term results following resectional surgery for Klatskin tumors: A twenty-year personal experience. Hepatogastroenterology 48:95–101, 2001. 230. Lygidakis NJ, Singh G, Bardaxoglou E, et al: Changing trends in the management of Klatskin tumor. Hepatogastroenterology 51:689–696, 2004. 231. MacCarty RL, LaRusso NF, May GR, et al: Cholangiocarcinoma complicating primary sclerosing cholangitis: Cholangiographic appearances. Radiology 156:43–46, 1985. 232. Madrazo BL, Francis I, Hricak H, et al: Sonographic findings in perforation of the gallbladder. AJR Am J Roentgenol 139:491–496, 1982. 233. Maldjian PD, Ghesani N, Ahmed S, et al: Adenomyomatosis of the gallbladder: Another cause for a “hot” gallbladder on 18F-FDG PET. AJR Am J Roentgenol 189:36–38, 2007. 234. Mall JC, Ghahremani GG, Boyer JL: Caroli’s disease associated with congenital hepatic fibrosis and renal tubular ectasia: A case report. Gastroenterology 66:1029–1035, 1974. 235. Manfredi R, Masselli G, Maresca G, et al: MR imaging and MRCP of hilar cholangiocarcinoma. Abdom Imaging 28:319–325, 2003. 236. Martin M: Primary sclerosing cholangitis. Annu Rev Med 44:221–227, 1993. 237. Matos C, Nicaise N, Deviere J, et al: Choledochal cysts: Comparison of findings at MR cholangiopancreatography and endoscopic retrograde cholangiopancreatography in eight patients. Radiology 209:443–448, 1998. 238. Matsui O, Takashima T, Kadoya M, et al: Staining in the liver surrounding gallbladder fossa on hepatic arteriography caused by increased cystic venous drainage. Gastrointest Radiol 12:307–312, 1987. 239. Matsushiro T, Suzuki N, Sato T, et al: Effects of diet on glucaric acid concentration in bile and the formation of calcium bilirubinate gallstones. Gastroenterology 72:630–633, 1977. 240. May RE, Strong R: Acute emphysematous cholecystitis. Br J Surg 58:453–458, 1971. 241. McCarty M, Choudhri AH, Helbert M, et al: Radiological features of AIDS related cholangitis. Clin Radiol 40:582–585, 1989. 242. McHugo JM, McKeown C, Brown MT, et al: Ultrasound findings in children with cystic fibrosis. Br J Radiol 60:137–141, 1987. 243. Meilstrup JW, Hopper KD, Thieme GA: Imaging of gallbladder variants. AJR Am J Roentgenol 157:1205–1208, 1991. 244. Mentha G, Rubbia-Brandt L, Orci L, et al: Traumatic neuroma with biliary duct obstruction after orthotopic liver transplantation. Transplantation 67:177–179, 1999. 245. Mentzer RM, Jr, Golden GT, Chandler JG, et al: A comparative appraisal of emphysematous cholecystitis. Am J Surg 129:10–15, 1975. 246. Merriam LT, Kanaan SA, Dawes LG, et al: Gangrenous cholecystitis: Analysis of risk factors and experience with laparoscopic cholecystectomy. Surgery 126:680–685, 1999.

CHAPTER 42  Biliary Tract and Gallbladder 247. Mirizzi PCJ: Sindrome del conducto hepatico. Int Chir 8:731–777, 1948. 248. Mirvis SE, Vainright JR, Nelson AW, et al: The diagnosis of acute acalculous cholecystitis: A comparison of sonography, scintigraphy, and CT. AJR Am J Roentgenol 147:1171–1175, 1986. 249. Mirvis SE, Whitley NO, Miller JW: CT diagnosis of acalculous cholecystitis. J Comput Assist Tomogr 11:83–87, 1987. 250. Mitropoulos FA, Angelopoulou MK, Siakantaris MP, et al: Primary non-Hodgkin’s lymphoma of the gall bladder. Leuk Lymphoma 40:123–131, 2000. 251. Mitsukawa T, Takemura J, Ohgo S, et al: Gallbladder function and plasma cholecystokinin levels in diabetes mellitus. Am J Gastroenterol 85:981–985, 1990. 252. Miyazaki K, Date K, Imamura S, et al: Familial occurrence of anomalous pancreaticobiliary duct union associated with gallbladder neoplasms. Am J Gastroenterol 84:176–181, 1989. 253. Mocharla R, Shah HR: Gallbladder perforation with contained empyema diagnosed by CT but missed by sonography and hepatobiliary scintigraphy. AJR Am J Roentgenol 168:283–284, 1997. 254. Moody FG, Calabuig R, Vecchio R, et al: Stenosis of the sphincter of Oddi. Surg Clin North Am 70:1341–1354, 1990. 255. Morfin E, Ponka JL, Brush BE: Gangrenous cholecystitis. Arch Surg 96:567–573, 1968. 256. Mortele KJ, Ros PR: Cystic focal liver lesions in the adult: Differential CT and MR imaging features. Radiographics 21:895–910, 2001. 257. Munoz L, Roayaie S, Maman D, et al: Hilar cholangiocarcinoma involving the portal vein bifurcation: Long-term results after resection. J Hepatobiliary Pancreat Surg 9:237–241, 2002. 258. Murayama S, Mizushima A, Russell WJ, et al: Sonographic diagnosis of bicameral gallbladders: A report of three cases. J Ultrasound Med 4:539–543, 1985. 259. Murphy BJ, Casillas J, Ros PR, et al: The CT appearance of cystic masses of the liver. Radiographics 9:307–322, 1989. 260. Nakajima T, Kondo Y, Miyazaki M, et al: A histopathologic study of 102 cases of intrahepatic cholangiocarcinoma: Histologic classification and modes of spreading. Hum Pathol 19:1228–1234, 1988. 261. Nakanuma Y, Sasaki M, Ishikawa A, et al: Biliary papillary neoplasm of the liver. Histol Histopathol 17:851–861, 2002. 262. Nakeeb A, Pitt HA, Sohn TA, et al: Cholangiocarcinoma: A spectrum of intrahepatic, perihilar, and distal tumors. Ann Surg 224:463–473, 1996. 263. Nardi PM, Yaghoobian J, Ruchman RB: CT demonstration of retrohepatic gallbladder in severe cirrhosis. J Comput Assist Tomogr 12:968–970, 1988. 264. Neitlich JD, Topazian M, Smith RC, et al: Detection of choledocholithiasis: Comparison of unenhanced helical CT and endoscopic retrograde cholangiopancreatography. Radiology 203:753– 757, 1997. 265. Niemeier OW: Acute free perforation of the gallbladder. Ann Surg 99:922–924, 1934. 266. Nishiyama Y, Yamamoto Y, Fukunaga K, et al: Dual-time-point 18F-FDG PET for the evaluation of gallbladder carcinoma. J Nucl Med 47:633–638, 2006. 267. Oe A, Kawabe J, Torii K, et al: Distinguishing benign from malignant gallbladder wall thickening using FDG-PET. Ann Nucl Med 20:699–703, 2006. 268. Okulski EG, Dolin BJ, Kandawalla NM: Intrahepatic biliary papillomatosis. Arch Pathol Lab Med 103:647–649, 1979. 269. Ong GB: A study of recurrent pyogenic cholangitis. Arch Surg 84:199–225, 1962. 270. Ooms H, Puylaert J, van der Werf S: Biliary fascioliasis: US and endoscopic retrograde cholangiopancreatography findings. Eur Radiol 5:196–199, 1995. 271. Oshikiri T, Kashimura N, Katanuma A, et al: Mucin-secreting bile duct adenoma: Clinicopathological resemblance to intraductal papillary mucinous tumor of the pancreas. Dig Surg 19:324–327, 2002. 272. Park JS, Kim MH, Lee SK, et al: The clinical significance of papillitis of the major duodenal papilla. Gastrointest Endosc 55:877–882, 2002. 273. Park JH, Lee DH, Kim HJ, et al: Unilocular extrahepatic biliary cystadenoma mimicking choledochal cyst: A case report. Korean J Radiol 5:287–290, 2004.

1265

274. Reference deleted in proofs. 275. Park MS, Yu JS, Kim KW, et al: Recurrent pyogenic cholangitis: Comparison between MR cholangiography and direct cholangiography. Radiology 220:677–682, 2001. 276. Paulson EK: Acute cholecystitis: CT findings. Semin Ultrasound CT MR 21:56–63, 2000. 277. Peer A, Witz E, Manor H, et al: Intrahepatic abscess due to gallbladder perforation. Abdom Imaging 20:452–455, 1995. 278. Pemberton M, Wells AD: The Mirizzi syndrome. Postgrad Med J 73:487–490, 1997. 279. Petrowsky H, Wildbrett P, Husarik DB, et al: Impact of integrated positron emission tomography and computed tomography on staging and management of gallbladder cancer and cholangiocarcinoma.   J Hepatol 45:43–50, 2006. 280. Polse S, Stoney RJ, Baldwin JN: Hemorrhage from the gallbladder: A report of three cases. Calif Med 107:51–53, 1967. 281. Pulpeiro JR, Armesto V, Varela J, et al: Fascioliasis: Findings in 15 patients. Br J Radiol 64:798–801, 1991. 282. Ralls PW, Colletti PM, Quinn MF, et al: Sonography in recurrent Oriental pyogenic cholangitis. AJR Am J Roentgenol 136:1010–1012, 1981. 283. Ramia JM, Muffak K, Fernandez A, et al: Gallbladder tuberculosis: False-positive PET diagnosis of gallbladder cancer. World J Gastroenterol 12:6559–6560, 2006. 284. Rattner DW, Schapiro RH, Warshaw AL: Abnormalities of the pancreatic and biliary ducts in adult patients with choledochal cysts. Arch Surg 118:1068–1073, 1983. 285. Rehani B, Strohmeyer P, Jacobs M, et al: Gallbladder metastasis from malignant melanoma: Diagnosis with FDG PET/CT. Clin Nucl Med 31:812–813, 2006. 286. Reimann AJ, Yeh BM, Breiman RS, et al: Atypical cases of gallstone ileus evaluated with multidetector computed tomography. J Comput Assist Tomogr 28:523–527, 2004. 287. Reisner RM, Cohen JR: Gallstone ileus: A review of 1001 reported cases. Am Surg 60:441–446, 1994. 288. Roa I, Araya JC, Villaseca M, et al: Gallbladder cancer in a high risk area: Morphological features and spread patterns. Hepatogastroenterology 46:1540–1546, 1999. 289. Romano AJ, van Sonnenberg E, Casola G, et al: Gallbladder and bile duct abnormalities in AIDS: Sonographic findings in eight patients. AJR Am J Roentgenol 150:123–127, 1988. 290. Ros PR, Buck JL, Goodman ZD, et al: Intrahepatic cholangiocarcinoma: Radiologic-pathologic correlation. Radiology 167:689–693, 1988. 291. Rosch T, Meining A, Fruhmorgen S, et al: A prospective comparison of the diagnostic accuracy of ERCP, MRCP, CT, and EUS in biliary strictures. Gastrointest Endosc 55:870–876, 2002. 292. Sagoh T, Itoh K, Togashi K, et al: Gallbladder carcinoma: Evaluation with MR imaging. Radiology 174:131–136, 1990. 293. Salomonowitz E, Frick MP, Simmons RL, et al: Obliteration of the gallbladder without formal cholecystectomy: A feasibility study. Arch Surg 119:725–729, 1984. 294. Saraswat VA, Sharma BC, Agarwal DK, et al: Biliary microlithiasis   in patients with idiopathic acute pancreatitis and unexplained biliary pain: Response to therapy. J Gastroenterol Hepatol 19:1206–1211, 2004. 295. Sasaki A, Aramaki M, Kawano K, et al: Intrahepatic peripheral cholangiocarcinoma: Mode of spread and choice of surgical treatment. Br J Surg 85:1206–1209, 1998. 296. Scholz FJ, Carrera GF, Larsen CR: The choledochocele: Correlation of radiological, clinical and pathological findings. Radiology 118:25–28, 1976. 297. Schulman A: Non-Western patterns of biliary stones and the role of ascariasis. Radiology 162:425–430, 1987. 298. Schulte SJ, Baron RL, Teefey SA, et al: CT of the extrahepatic bile ducts: Wall thickness and contrast enhancement in normal and abnormal ducts. AJR Am J Roentgenol 154:79–85, 1990. 299. Seel DJ, Park YK: Oriental infestational cholangitis. Am J Surg 146:366–370, 1983.

1266

PART II  CT and MR Imaging of the Whole Body

300. Semelka RC, Hussain SM, Marcos HB, et al: Biliary hamartomas: Solitary and multiple lesions shown on current MR techniques including gadolinium enhancement. J Magn Reson Imaging 10:196–201, 1999. 301. Shah VR, Clegg JF: Haemorrhagic cholecystitis. Br J Surg 66:404–405, 1979. 302. Shea TE, Pfister RC: Opacification of the gallbladder by urographic contrast media: Reflection of an alternate excretory pathway. Am J Roentgenol Radium Ther Nucl Med 107:763–768, 1969. 303. Shieh CJ, Dunn E, Standard JE: Primary carcinoma of the gallbladder: A review of a 16-year experience at the Waterbury Hospital Health Center. Cancer 47:996–1004, 1981. 304. Shimura K, Tamada K, Asada M, et al: Intraductal ultrasonography   of traumatic neuroma of the bile duct. Abdom Imaging 26:632–634, 2001. 305. Shuto R, Kiyosue H, Komatsu E, et al: CT and MR imaging findings of xanthogranulomatous cholecystitis: Correlation with pathologic findings. Eur Radiol 14:440–446, 2004. 306. Siegel MJ, Herman TE: Periportal low attenuation at CT in childhood. Radiology 183:685–688, 1992. 307. Singer JA, McKeen RV: Laparoscopic cholecystectomy for acute or gangrenous cholecystitis. Am Surg 60:326–328, 1994. 308. Sjodahl R, Wetterfors J: Lysolecithin and lecithin in the gallbladder wall and bile: Their possible roles in the pathogenesis of acute cholecystitis. Scand J Gastroenterol 9:519–525, 1974. 309. Slone HW, Bennett WF, Bova JG: MR findings of multiple biliary hamartomas. AJR Am J Roentgenol 161:581–583, 1993. 310. Small AJ, Larusso NF, Lazaridis KN: Primary sclerosing cholangitis. In Blumgart LH, editor: Surgery of the liver, biliary tract and pancreas, ed 4, Philadelphia, 2007, Saunders, p 619. 311. Smathers RL, Lee JK, Heiken JP: Differentiation of complicated cholecystitis from gallbladder carcinoma by computed tomography. AJR Am J Roentgenol 143:255–259, 1984. 312. Solomon A, Kreel L, Pinto D: Contrast computed tomography in the diagnosis of acute cholecystitis. J Comput Assist Tomogr 3:585–588, 1979. 313. Sons HU, Borchard F, Joel BS: Carcinoma of the gallbladder: Autopsy findings in 287 cases and review of the literature. J Surg Oncol 28:199–206, 1985. 314. Sood GK, Mahapatra JR, Khurana A, et al: Caroli disease: Computed tomographic diagnosis. Gastrointest Radiol 16:243–244, 1991. 315. Soper RT, Dunphy DL: Sarcoma botryoides of the biliary tree. Surgery 63:1005–1011, 1968. 316. Soyer P, Bluemke DA, Reichle R, et al: Imaging of intrahepatic cholangiocarcinoma. 1. Peripheral cholangiocarcinoma. AJR Am J Roentgenol 165:1427–1431, 1995. 317. Soyer P, Bluemke DA, Reichle R, et al: Imaging of intrahepatic cholangiocarcinoma. 2. Hilar cholangiocarcinoma. AJR Am J Roentgenol 165:1433–1436, 1995. 318. Soyer P, Gouhiri M, Boudiaf M, et al: Carcinoma of the gallbladder: Imaging features with surgical correlation. AJR Am J Roentgenol 169:781–785, 1997. 319. Sripa B: Pathobiology of opisthorchiasis: An update. Acta Trop 88:209–220, 2003. 320. Stohl EL, Diffenbaugh WG, Baker JH, et al: Collective reviews: Gangrene, and perforation of the gallbladder. Int Abst Surg 114:1–7, 1962. 321. Strax R, Toombs BD, Kam J, et al: Gallbladder enhancement following angiography: A normal CT finding. J Comput Assist Tomogr 6:766–768, 1982. 322. Sugiyama M, Haradome H, Takahara T, et al: Anomalous pancreaticobiliary junction shown on multidetector CT. AJR Am J Roentgenol 180:173–175, 2003. 323. Takayasu K, Muramatsu Y, Shima Y, et al: Hepatic lobar atrophy following obstruction of the ipsilateral portal vein from hilar cholangiocarcinoma. Radiology 160:389–393, 1986. 324. Thorsen MK, Quiroz F, Lawson TL, et al: Primary biliary carcinoma: CT evaluation. Radiology 152:479–483, 1984.

325. Todani T, Watanabe Y, Mizuguchi T, et al: Hepaticoduodenostomy at the hepatic hilum after excision of choledochal cyst. Am J Surg 142:584–587, 1981. 326. Todani T, Watanabe Y, Narusue M, et al: Congenital bile duct cysts: Classification, operative procedures, and review of thirty-seven cases including cancer arising from choledochal cyst. Am J Surg 134:263–269, 1977. 327. Todani T, Watanabe Y, Toki A, et al: Carcinoma related to choledochal cysts with internal drainage operations. Surg Gynecol Obstet 164:61–64, 1987. 328. Toombs BD, Foucar E, Rowlands BJ, et al: Multiseptate gallbladder. South Med J 75:610–612, 1982. 329. Tsai HM, Lin XZ, Chen CY, et al: MRI of gallstones with different compositions. AJR Am J Roentgenol 182:1513–1519, 2004. 330. Turner MA, Fulcher AS: The cystic duct: Normal anatomy and disease processes. Radiographics 21:3–22, 288-294, 2001. 331. Udelsman R, Sugarbaker PH: Congenital duplication of the gallbladder associated with an anomalous right hepatic artery. Am J Surg 149:812–815, 1985. 332. Ueda J, Hara K, Ohishi H, et al: High density bile in the gallbladder observed by computed tomography. J Comput Assist Tomogr 7:801–804, 1983. 333. Ueda J, Kobayashi Y, Nishida T: Computed tomography evaluation   of high-density bile in the gallbladder. Gastrointest Radiol 15:22–26, 1990. 334. Ukaji M, Ebara M, Tsuchiya Y, et al: Diagnosis of gallstone composition in magnetic resonance imaging: In vitro analysis. Eur J Radiol 41:49–56, 2002. 335. van Sonnenberg E, Ferrucci JT, Jr: Bile duct obstruction in hepatocellular carcinoma (hepatoma): Clinical and cholangiographic characteristics. Report of 6 cases and review of the literature. Radiology 130:7–13, 1979. 336. Vezina WC, Paradis RL, Grace DM, et al: Increased volume and decreased emptying of the gallbladder in large (morbidly obese, tall normal, and muscular normal) people. Gastroenterology 98:1000–1007, 1990. 337. Reference deleted in proofs. 338. Wang AJ, Wang TE, Lin CC, et al: Clinical predictors of severe gallbladder complications in acute acalculous cholecystitis. World J Gastroenterol 9:2821–2823, 2003. 339. Wastie ML, Cunningham IG: Roentgenologic findings in recurrent pyogenic cholangitis. Am J Roentgenol Radium Ther Nucl Med 119:71–77, 1973. 340. Weinbren K, Mutum SS: Pathological aspects of cholangiocarcinoma.   J Pathol 139:217–238, 1983. 341. Weiner SN, Koenigsberg M, Morehouse H, et al: Sonography and computed tomography in the diagnosis of carcinoma of the gallbladder. AJR Am J Roentgenol 142:735–739, 1984. 342. Weinstein JB, Heiken JP, Lee JK, et al: High resolution CT of the   porta hepatis and hepatoduodenal ligament. Radiographics 6:55–74, 1986. 343. Weissleder R, Stark DD, Compton CC, et al: Cholecystitis: Diagnosis by MR imaging. Magn Reson Imaging 6:345–348, 1988. 344. Whitesell FB, Jr: Gallstone ileus. Am Surg 36:317–322, 1970. 345. Wiesner RH, LaRusso NF: Clinicopathologic features of the syndrome of primary sclerosing cholangitis. Gastroenterology 79:200–206, 1980. 346. Willi UV, Reddish JM, Teele RL: Cystic fibrosis: Its characteristic appearance on abdominal sonography. AJR Am J Roentgenol 134:1005– 1010, 1980. 347. Williams I, Slavin G, Cox A, et al: Diverticular disease (adenomyomatosis) of the gallbladder: A radiological-pathological survey. Br J Radiol 59:29–34, 1986. 348. Wilson AK, Kozol RA, Salwen WA, et al: Gangrenous cholecystitis in an urban VA hospital. J Surg Res 56:402–404, 1994. 349. Yamamoto M, Takasaki K, Yoshikawa T, et al: Does gross appearance indicate prognosis in intrahepatic cholangiocarcinoma? J Surg Oncol 69:162–167, 1998.

CHAPTER 42  Biliary Tract and Gallbladder 350. Yamashita K, Jin MJ, Hirose Y, et al: CT finding of transient focal increased attenuation of the liver adjacent to the gallbladder in acute cholecystitis. AJR Am J Roentgenol 164:343–346, 1995. 351. Yeh TS, Jan YY, Tseng JH, et al: Malignant perihilar biliary obstruction: Magnetic resonance cholangiopancreatographic findings. Am J Gastroenterol 95:432–440, 2000. 352. Yoon KH, Ha HK, Kim MH, et al: Biliary stricture caused by blunt abdominal trauma: Clinical and radiologic features in five patients. Radiology 207:737–741, 1998. 353. Yoon KH, Ha HK, Kim CG, et al: Malignant papillary neoplasms of the intrahepatic bile ducts: CT and histopathologic features. AJR Am J Roentgenol 175:1135–1139, 2000. 354. Yoshida H, Itai Y, Minami M, et al: Biliary malignancies occurring in choledochal cysts. Radiology 173:389–392, 1989. 355. Yoshimitsu K, Honda H, Jimi M, et al: MR diagnosis of adenomyomatosis of the gallbladder and differentiation from gallbladder carcinoma: Importance of showing Rokitansky-Aschoff sinuses. AJR Am J Roentgenol 172:1535–1540, 1999. 356. Yoshimitsu K, Honda H, Shinozaki K, et al: Helical CT of the local spread of carcinoma of the gallbladder: Evaluation according to the

357.

358.

359.

360.

361.

1267

TNM system in patients who underwent surgical resection. AJR Am J Roentgenol 179:423–428, 2002. Young WT, Thomas GV, Blethyn AJ, et al: Choledochal cyst and congenital anomalies of the pancreatico-biliary junction: The clinical findings, radiology and outcome in nine cases. Br J Radiol 65:33–38, 1992. Zech CJ, Schoenberg SO, Reiser M, et al: Cross-sectional imaging of biliary tumors: Current clinical status and future developments. Eur Radiol 14:1174–1187, 2004. Zeman RK: Cholelithiasis and cholecystitis. In Gore RM, Levine MS, editors: Textbook of gastrointestinal radiology, ed 2, Philadelphia, 2000, Saunders, p 1336. Zen, Fujii T, Itatsu K, et al: Biliary cystic tumors with bile duct communication: A cystic variant of intraductal papillary neoplasm of the bile duct. Mod Pathol 19:1243–1254, 2006. Zimmon DS, Falkenstein DB, Manno BV, et al: Choledochocele: Radiologic diagnosis and endoscopic management. Gastrointest Radiol 3:348–351, 1978.

43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases Achille Mileto and Daniel T. Boll

“Is life worth living?—It all depends on the liver!” —William James, 1842-1910 Over more than 3 millennia the liver held a central and unsurpassed role among all human organs, manifested in coeval theogony, poetry, and fairy tales. The liver was chosen to be the seat of the soul, intelligence, and passion, equipped with particular divine protection and hoped to be indestructible, reflecting the prodigious recuperative powers of hepatic parenchyma.33 In Egyptian mythology, Isis’s envoy Imsety (18th Dynasty, 15701350 BC) was the guardian of the liver of the deceased on its journey to the afterlife. Later, Greek mythology emphasized the particular physical and psychological qualities of torture when the liver is being targeted. According to Hesiod (≈700 BC) an eagle devoured Prometheus’s liver after he stole fire from Zeus. Because of the recuperative powers of hepatic parenchyma, his torture had to be endured for 13 generations until Prometheus was saved by Heracles. Unfortunately, Achilles and Tityus later met the same fate. According to Galen (129-199) the liver is of vital physical and psychological relevance to all human beings. The four bodily humors— blood, phlegm, yellow bile, and black bile—were thought to give rise to the sanguine, phlegmatic, choleric, and melancholic temperaments, respectively; health thereafter was defined as equilibrium in these four humors. Physical and psychological distress was a symptom of imbalance in these humors, as phrased by Horace (65-8 BC): “My liver swells with bile difficult to repress.” Renaissance poetry emphasized the ancient belief that the liver is the seat of love, as seen in Shakespeare’s (1564-1616) rhapsody Love’s Labour’s Lost: “This is the liver-vein, which makes flesh a deity.” Fairy tales of more recent centuries picked up the basic principle of the liver being the seat of the soul and intelligence. According to the Brothers Grimm (1785/86-1863/59, respectively), Snow White’s stepmother ordered the woodsman not just to kill her but to cut out her liver to ensure her physical and psychological disappearance. It all depended on the liver!

NORMAL LIVER ANATOMY AND VARIANTS Normal Anatomy The liver, as the largest of all abdominal organs, commands the right upper abdominal quadrant, though it can extend into the epigastrium and as far as the left upper abdominal quadrant, abutting the splenic contour. Taking variations of liver volume and dimensions due to ethnicity into account, in the male the liver weighs from 1.4 to 1.7 kg and in the female 1.2 to 1.5 kg. Its greatest transverse measurement is from 20 to 26 cm; vertically near its lateral surface the liver measures 15 to 21 cm, and its greatest anteroposterior diameter (determined at

1268

the level of the upper right kidney) measures 7 to 12 cm. The hepatic parenchyma is surrounded by a dense layer of connective tissue forming the liver capsule.

Surface Anatomy.  The liver’s external structure features convex diaphragmatic and concave visceral surfaces. The diaphragmatic surface is perfectly fitted and affixed by intervening connective tissue to a triangular section of the undersurface of the diaphragm termed the bare area. Other than the bare area, only the fossa of the gallbladder, the fossa of the inferior vena cava (IVC), and the suprarenal impression are not covered by peritoneum. The bare area’s upper and lower margins are demarcated by peritoneal reflections forming the anterior and posterior layers of the coronary ligament, respectively. The posterior layer of the coronary ligament continues with the right layer of the lesser omentum as the hepatorenal ligament. The anterior and posterior layers of the hepatic coronary ligament converge on the right and left margins of the bare area, forming the right and left triangular ligaments. The right triangular ligament is a small peritoneal fold that extends toward the diaphragm. The left triangular ligament on the other hand is of considerable size and gives off connective tissue tracts extending to the diaphragm as well as portions of the falciform ligament. The main anterior portion of the left triangular ligament forms the left layer of the falciform ligament, whereas the right layer of the falciform ligament is a continuation of the anterior layer of the coronary ligament. The falciform ligament thereby forms a broad and thin peritoneal fold that dissects the liver into its right and left hepatic lobes and is in contact with the peritoneum posterior to the rectus abdominis musculature and the diaphragm. Intraabdominal fixation of the liver to the undersurface of the diaphragm is therefore realized by intervening connective tissue of the bare area as well as by means of connective tissue tracts extending from the coronary and triangular ligaments; the rather lax falciform ligament attachment to the peritoneum only limits lateral displacement. At the liver’s visceral surface, predominantly fossae and fissures define the lobar anatomy. Two fossae extending sagittally are joined by a transverse fissure, forming a letter H–like structure. The left limb of the letter H is known as the left sagittal fossa, a deep groove extending from the anterior surface to the posterior contour of the organ, thereby marking the division of the liver into the right and left hepatic lobes. The anterior part of the left sagittal fossa houses the umbilical vein during fetal circulation and its adult remains, the ligamentum teres, thereafter.135 The ligamentum teres itself extends toward the base of the falciform ligament and runs together with the paraumbilical veins between its right and left layers from the umbilicus to the left portal vein. The posterior part of the left sagittal fossa contains the ductus venosus, which during fetal circulation forms a bypass between the left portal vein and the left hepatic vein, functioning as a left-sided

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases portosystemic shunt, and during adulthood holds its remains, the ligamentum venosum.135 The anterior and posterior parts of the left sagittal fossa can be separated by the pons hepatis. The right limb of the letter H consists anteriorly of the fossa for the gallbladder and posteriorly of the fossa for the IVC. The latter usually appears as a short and deep impression; occasionally liver parenchyma can surround the IVC and thereby form a complete canal. Three veins—the left, middle, and right hepatic veins—perforate the floor of the fossa and drain into the IVC. These two fossae are free of peritoneal lining and are separated by the caudate process of the liver. The bar connecting the two limbs of the letter H is termed the transverse fissure. Anterior to the transverse fissure lies the quadrate lobe; posterior, the caudate lobe of the liver. The transverse fissure represents the porta hepatis, a 5-cm-deep groove containing the hepatoduodenal ligament, which transmits anteriorly the hepatic duct on the right and the hepatic artery on the left; posteriorly and partially in between the two resides the portal vein as well as hepatic nerves and lymphatics. The hepatic artery carries oxygenated blood to the liver, constituting about 20% to 25% of the total blood volume to the organ. The portal vein on the other hand carries nutrient-rich blood from the digestive system, comprising 75% to 80% of total hepatic blood volume. The hepatic artery and portal vein ramify in parallel to one another. In healthy persons, the caliber of the portal venous branches is larger than those of corresponding arterial branches. Venous return is realized through the three main hepatic veins that drain directly into the IVC. The visceral undersurface of the liver is uneven owing to its proximity to the antral esophagus, stomach, right colonic flexure, upper pole of the right kidney, right adrenal gland, and descending portion of the duodenum. Each of the organs forms characteristic impressions. Additionally a rounded eminence, the tuber omentale, abuts the gastric concavity of the lesser curvature in front of the lesser omentum. The diaphragmatic surface can be riddled with furrows from prominent muscular bundles forming the diaphragmatic crura, seen particularly in patients with chronic lung diseases.

Functional Segmental Anatomy.  Advances in hepatic surgery have made the lobar segmentation of the liver into right, left, quadrate, and caudate lobes obsolete. This division has yielded to a more functionally oriented classification primarily based on the surgical definition of feasible intrahepatic boundaries of resection, which in turn is based on segmental hepatic configuration.17,18,21,47,61 Centrally located in each of the hepatic segments is a segmental branch of the portal vein and hepatic artery as well as a segmental bile duct. The solitary distal hepatic veins lie between the individual segments. Several systems have been suggested for classifying the segments of the liver, many of them quite similar; the most common is the nomenclature proposed by Bismuth-Couinaud, which will now be considered in detail. In the Bismuth-Couinaud system, segment I corresponds to the caudate lobe. This segment is in the unique position of receiving branches from both the main portal vein and its right and left branches, termed the portal trinity. Furthermore, it does not drain into the hepatic veins but directly into the IVC. All remaining liver segments (II to VIII) are defined by their positions relative to branches of the portal and hepatic veins. The portal venous plane is defined as the transverse plane dissecting the liver at the level of the portal venous bifurcation into superior and inferior hepatic segments. The portal vein divides early into left and right portal venous branches, with the left portal pedicle supplying segments II through IV, thereby forming the segmental left hepatic lobe, and the right portal pedicle (segments V through VIII) thus constituting the segmental right hepatic lobe.

1269

The left portal vein, after giving off a caudate branch, divides into its terminal branches, the left lateral and left medial portal venous branches. The left lateral portal venous branch supplies superior segment II, located lateral to the left hepatic vein and above the portal venous plane. The left medial portal venous branch supplies inferior segment III located laterally to the left hepatic vein and beneath the portal venous plane as well as segment IV. Segment IV is delineated medially by the middle hepatic vein and laterally by the left hepatic vein and can be further subdivided into a superior segment IVa and an inferior segment IVb in regard to the portal venous plane. The right portal vein divides into its right anterior and right posterior terminal branches. The right anterior portal venous branch supplies the anterior-inferior segment V located between the middle and right hepatic veins and below the portal venous plane, whereas the anterior-superior segment VIII can be found contiguous to segment V but directly above the portal venous plane. The right posterior portal venous branch perfuses the posterior-inferior segment VI located posterolaterally to the right hepatic vein but beneath the portal venous plane, while the posterior-superior segment VII can be found adjacent to segment VI directly above the portal venous plane (Fig. 43-1).

Liver Anatomy Variants A significant number of normal variants in liver morphology and hepatic vascularity occur. Knowledge of the most common anatomic variants of the liver is essential, in particular to diagnose generalized hepatomegaly versus lobar hypertrophy, as well as to identify the spatial hepatic location of focal lesions based on the functional segmentation nomenclature proposed by Bismuth-Couinaud.

Variants in Morphology.  Common normal variants in liver morphology include the horizontal elongation of the lateral segment (Bismuth-Couinaud segment II) of the left hepatic lobe, which can extend into the left upper abdominal quadrant and eventually abut or even wrap around the splenic contour. This anatomic variant is more common in women. However, because of the different computed tomography (CT), magnetic resonance imaging (MRI), and perfusion characteristics of hepatic and splenic parenchyma, this variant usually does not impose problems during radiologic interpretation. Another normal variant of liver size and shape is vertical elon­ gation of the right lobe, termed a Riedel lobe (Fig. 43-2). The presence of a Riedel lobe usually results from a prominent inferiorly positioned narrow right lobe of the liver that significantly extends  the expected confines of the liver. A Riedel lobe is more frequently found in women.180 It has to be differentiated from extracapsular extension caused by a liver tumor, in particular hepatic adenoma or metastasis. A prominent caudate lobe can show a pronounced extension, which is then termed a hepatic papillary process (Fig. 43-3). This normal variant can simulate a mass in the portocaval region.9,72 Differentiation between an extrahepatic portocaval mass and a papillary process can be facilitated by recognizing the continuity of this process with the caudate lobe on contiguous axial sections, as well as multiplanar reformations (MPR), preferably in the coronal plane. Following a partial hepatectomy, rapid hypertrophy of the remaining liver segments can occur. This hypertrophy results in distortion of the expected anatomy as the major vessels become displaced by the enlarging segments. Furthermore, segmental hepatic resections have to be differentiated from congenital hypoplasia of the left lobe. Assessment of the clinical history of the patient and direct patient communication remain of paramount importance when interpreting complex anatomy.

1270

PART II  CT and MR Imaging of the Whole Body

4a

mb-LPV

2

ab-RPV

8

RHP

pb-RPV

7

A

B

4a FLV

8

LHV

MHP

Ib-LPV

C

LHVs

mb-LPV

2

MHP

Ib-LPV

ab-RPV

RHPs pb-RPV

7

D

E

F

FLT

FLT

4b

8/5

3 LHA RHP 1

mb-LPV ab-RPV

FLT t-LHP

Ib-LPV t-MHP

RPV

pb-RPV

t-RHP

7/6

G

H

4b 5

3

I

mb-LPV

Ib-LPV

t-LHP t-MHP

ab-RPV

6

t-RHP pb-RPV

J

K

L

FIG 43-1  Triple-phase contrast-enhanced MDCT imaging during the hepatic arterial phase (A, D, G, and J), the portal venous phase (B, E, H, and K) and the delayed hepatic phase (C, F, I, and L) at different levels in regard to the portal venous plane: upper superior plane (A to C), lower superior plane (D to F), portal venous plane (G to I), and inferior plane (J to L). A to C, In the upper superior hepatic plane the hepatocentric venous branches of the right hepatic vein (RHV), middle hepatic vein (MHV), and left hepatic vein (LHV) are clearly seen, differentiating the superior hepatic segments (2, 4a, 8, and 7) that are supplied by the lateral branch (lb-LPV) and medial branch (mb-LPV) of the left portal vein as well as the anterior branch (ab-RPV) and posterior branch (pb-RPV) of the right portal vein, respectively. D to F, The lower superior hepatic plane shows the more intrahepatic segments of right hepatic vein (RHV), middle hepatic vein (MHV), and left hepatic vein (LHV) as well as lateral branch (lb-LPV) and medial branch (mb-LPV) of the left portal vein and anterior branch (abRPV) and posterior branch (pb-RPV) of the right portal vein. The fissure for the ligamentum venosum (FLV) separates the upper caudate. G to I, In the portal venous plane the hepatocentric right portal vein (RPV) and left portal vein (LPV) can be seen with their corresponding lateral branch (lb-LPV) and medial branch (mb-LPV), as well as anterior branch (ab-RPV) and posterior branch (pb-RPV), respectively. Hepatocentric left hepatic and right hepatic arteries can be seen. Segment 1, the caudate lobe, is also visible. The terminal left hepatic vein (t-LHP) can bee seen to the left of the fissure of the ligamentum teres (FLT); the terminal terminal middle (t-MHP) and right (t-RHP) are on its right side. J to L, In the inferior hepatic plane the inferior hepatic segments (3, 4b, 6, and 5) and their analogously supplying distal portal venous branches can be found.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

1271

Variants in Vascularity.  Common normal variants in hepatic vas-

FIG 43-2  Precontrast T1-weighted hepatic MRI in a 77-year-old woman with Riedel’s lobe (arrows) resulting from a prominent inferiorly positioned narrow right lobe of the liver that significantly extends the expected confines of the liver.

cularity have to be identified prior to implementation of the BismuthCouinaud system to describe the location of hepatic focal lesions. Prior knowledge of replaced hepatic arteries is of importance in clinical scenarios of planned partial hepatectomies and extensive pancreatic surgery. Interindividual frequency and degree of variability of the hepatic arterial and venous as well as portal venous vasculature, particularly in the right liver lobe, is significant. Variants in arterial hepatic supply exist as accessory or replaced right hepatic arteries, commonly originating from the superior mesenteric artery, then located posterior in regard to the portal vein. In up to 20% of all patients, a duplication of the right hepatic vein may be found. High-resolution imaging of the hepatic veins and the portal venous system has shown that only the anterosuperior and posteroinferior sectors are visualized without significant variation in vascularity.36,70,112,215 It was shown that portal venous branches often cross the plane of the right hepatic vein, and no portal venous plane can be identified separating the anterosuperior and posteroinferior sectors; hence the right hepatic vein cannot serve as a reliable landmark for dividing the anteromedial and posterolateral segments of the right hepatic lobe. The left liver lobe presents fewer variations in vascular anatomy. Analogously, variants in arterial hepatic supply exist as accessory or replaced left hepatic arteries, commonly originating from the left gastric artery and then always coursing within the fissure of the venous ligament. Duplication of the middle hepatic vein is seen in about 5% of patients, of the left in about 15%. There may also be some variation in arterial structure, particularly in segment IV. In about 8% of all patients the dorsal portion of segment IV is supplied by the right hepatic artery, hence from the functionally contralateral side.105,131,141,206 In about 30% of patients, accessory portal segments analogous to the caudate lobe (Bismuth-Couinaud segment I) may be observed, each of which is supplied directly from the main portal venous trunk or from its right branch.10,35,150,215 Recognizing the significant interindividual variation in hepatic vascularity, advanced and less radical hepatic surgery aims to reduce the reliance on a rigid schematic division of the liver into segments and considers instead the patient’s individual vascular situation in surgeries involving segments or even subsegments of the liver.67,92,96,188 A liver segment is defined solely by its vascular supply. Based on highresolution multidetector (MD)CT and MR data sets, it is possible to construct three-dimensional (3D) hepatic maps of vascular anatomy to assist surgeons in exact and effective preoperative planning of the procedure.

HEPATIC IMAGING TECHNIQUES Multidetector CT Imaging of the Liver

FIG 43-3  T1-weighted contrast-enhanced MRI of a 55-year-old man with hypertrophy and extension of the caudate lobe (arrows), also known as a papillary hepatic process. Note the simple cyst along the fissure for the ligamentum teres (arrowhead).

MDCT remains the predominant imaging modality to visualize the liver. Besides the general availability of this modality, the role of MDCT is primarily defined by its excellent morphologic visualization capabilities, in particular of diffuse or focal intrahepatic lesions as well as of anatomic relationships between the liver and adjacent organs. These imaging capabilities have been further enhanced since overall detector row numbers have increased from 4-slice to 16-slice, and of late to 64-slice detector arrays that allow partial and multisegmental raw-data reconstructions and fundamentally affect imaging protocols. MDCT imaging in particular benefited from acquisition of high-resolution volume image data sets with drastically reduced scan times due to significantly shorter gantry rotation periods (as short as 300 milliseconds), which can be further decreased (to about 80 milliseconds) by employing a dual x-ray tube approach.54,92,188

1272

PART II  CT and MR Imaging of the Whole Body

Normal CT attenuation of the liver in unenhanced imaging studies varies interindividually between 55 and 65 Hounsfield units (HU). Usually the normal liver appears homogeneous at CT visualization, and its attenuation exceeds that of the spleen in healthy individuals by about 10 HU. The relatively large interindividual range in attenuation values is due to the varying fat and glycogen content of the organ. Increased diffuse deposition of fat leads to a reduction in attenuation, whereas increased glycogen is reflected as an increase in CT measured density.48,225 Blood circulation in the liver comprises two major components, the hepatic artery and the portal vein. Because of this dual-source hepatic blood supply, the pattern of extracellular parenchymal contrast uptake and the associated changes in tissue attenuation over time follow a complex multicompartmental perfusion model. The overall hepatic perfusion cycle can be differentiated into three idealized hepatic perfusion phases: (1) arterial phase, (2) redistribution/portal venous phase, and (3) equilibrium/hepatic venous phase. Because of the temporal proximity and relatively short duration of each of these perfusion phases, and owing to multiple extrahepatic factors that influence systemic and portal venous perfusion, an optimized and individualized regimen of intravenous contrast agent and saline chaser application has to be applied: • Application of contrast agent and saline chaser by means of mechanical power injectors is mandatory to realize homogenous flow rates of 1 to 5 mL/sec. • Venous access is usually supplied by a 16- to 20-gauge indwelling venous catheter placed in a right cubital or antecubital vein. • By choosing a nonionic and more highly concentrated contrast agent solution containing at least 370 mg iodine/mL, an overall contrast volume of 80 mL is sufficient to perfuse the liver with 30 g of iodine, resulting in a 25% increase in iodine application during the arterial phase. This reduces the overall injected iodine dose significantly compared to less-concentrated contrast agent solutions at fixed transit times from the cubital/ antecubital vein to the liver.11,123,197 • By taking into account the individual variations of cardiac output, blood volume, and visceral perfusion, the need for bolus tracking techniques as supplied by all MDCT imaging systems to determine the arrival time of the contrast agent bolus is a mandatory prerequisite. Bolus tracking relies on sequential lowdose scans obtained at the level of the upper abdominal aorta; aortic enhancement values are immediately displayed. When aortic enhancement reaches a predefined threshold of approximately 150 HU, hepatic scanning is initiated. The delay time between adequate bolus detection in the upper abdominal aorta, CT table motion, and subsequent initiation of scanning is as low as 5 seconds on current MDCT imagers. Approximately 10 seconds after contrast threshold–based scanning initiation during the early arterial phase, contrast enhancement of the abdominal aorta and hepatic artery is observed without admixture of enhanced portal venous blood. The late arterial phase at approximately 20 seconds after scanning initiation leads to a clear depiction of the hepatic artery and its branches, owing to distinctive contrast enhancement. However, a minimal admixture of enhanced portal venous blood may already have occurred. The redistribution/portal venous inflow phase, imaged about 30 seconds after scan initiation, allows early visualization of the portal vein and its intrahepatic branches, in contrast to the still unenhanced hepatic veins. Maximum contrast enhancement of the portal vein and its intrahepatic branches is reached after approximately 40 seconds. If monoslice CT scanners are used to perform biphasic liver scanning, all

the contrast-enhanced phases previously described would be combined in one hepatic arterial/portal venous phase.55,65 The hepatic venous phase can be acquired 60 seconds after scan initiation. Simultaneous enhancement of hepatic and portal veins will thus be visualized. Because of retained contrast agent, focal liver lesions such as cholangiocarcinoma benefit from further delayed hepatic imaging, after the parenchymal contrast has mostly been drained from the liver.

MDCT Imaging Strategies.  The multitude of clinical scenarios that require or include hepatic MDCT imaging cannot be appropriately addressed with one comprehensive imaging protocol. Instead a manageable list of MDCT imaging protocols to answer specific clinical situations have been defined throughout medical institutions. Furthermore, each of these imaging protocols is tailored to the individual patient, taking individual descriptors and parameters such as body habitus, body weight, heart rate, and limitations in the patient’s range of motion and positioning into account. Valid indications for non–contrast-enhanced (NCE) hepatic MDCT examinations include (but are not limited to) the search for calcifications in settings such as hemangioendothelioma, calcified metastases of mucinous adenocarcinoma, and postinflammatory calcifications in cases of alveolar echinococcosis. Furthermore, unenhanced MDCT can help detect intrahepatic CT-opaque bile duct concretions. Subcapsular or parenchymal hemorrhage can also be visualized. Finally, postprocedural visualization of the lipiodol distribution pattern following intraarterial embolization of liver tumors is another valid indication for NCE hepatic MDCT imaging. Whenever contrast-enhanced MDCT imaging of the liver is part  of a more comprehensive imaging protocol—for example, including the entire chest, abdomen, and pelvis in the imaging field of view (FOV) without suspected specific hepatic pathology—usually only one hepatic contrast phase is acquired. In this contrast monophase imaging scenario, the portal venous imaging phase will be preferred. Dedicated multiphase hepatic imaging can be performed subsequently to further evaluate suspected hepatic findings. In clinical scenarios with a known or suspected hypervascular primary neoplasm outside the liver for which there are suspected hypervascular hepatic metastases, a dedicated hepatic dual-phase imaging protocol should be performed. This protocol includes MDCT imaging during both the hepatic late arterial and the portal venous phase but does not include an NCE phase. Confirmed or suspected diagnoses of breast carcinoma, renal cell carcinoma, melanoma, and neuroendocrine tumors (e.g., islet cell tumors, carcinoid and thyroid carcinomas) would call for the dual-phase hepatic imaging protocol. Known or suspected cirrhosis or hepatocellular carcinoma as well as suspected benign primary liver lesions (e.g., focal nodular hyperplasia or hepatic adenoma) would be imaged by means of MDCT employing a triple-phase imaging protocol. This protocol includes a noncontrast imaging phase in addition to hepatic arterial and portal venous imaging phases; it also calls for a larger volume of contrast agent and higher infusion rates compared to the dual-phase protocol. Oftentimes a delayed phase acquired 10 to 15 minutes after initiation of contrast is performed as well. In patients with known or suspected cholangiocarcinoma another triple-phase imaging protocol (unenhanced, portal venous, and delayed) addresses a pathophysiologic phenomenon of delayed contrast agent washout and subsequent relative hyperattenuation of cholangiocarcinomas in comparison to surrounding hepatic parenchyma. For preoperative planning in anticipation of a partial hepatic resection, high-resolution imaging of the celiac axis, proximal superior

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases mesenteric artery, and porta hepatis is mandatory. A hepatic resection protocol combines high-resolution imaging of the arterial vasculature with a limited imaging FOV followed by a standard portal venous imaging phase.

MDCT Imaging Protocols.  The family of available CT imagers covers a huge bandwidth of detector configurations from single- to 128-detector arrays, which allow helical scanning at various temporal and spatial resolutions. Fundamental differences in detector configurations of equally sized and equally spaced detector arrays from asymmetric detector configurations of detector sizes and arrangements result in MDCT imaging protocols with unique parameters for almost each available MDCT imaging system. The advent of dual-source and/ or dual-detector MDCT imagers that allow for simultaneous dualenergy examinations will increase possible combinations of imaging parameters even further. This overview of MDCT imaging protocols attempts to cover a representative selection of widely used equally sized and equally spaced 16-detector and 128-detector MDCT images. Adaptations according to individual scanner configurations, however, can be performed using these imaging protocols as groundwork (Tables 43-1 to 43-6) (see  Fig. 43-1). Dual-energy CT applications for liver imaging.  Dual-energy MDCT allows for spectral analyses and thus tissue characterization based on using two x-ray beams of distinct and known energy distributions tuned to the characteristic absorption profiles of different tissues and injected contrast materials. This has the potential for enhancing inherent tissue differences and enabling selective contrast material detection.125 The identification, isolation, and quantification of iodinecontaining voxels throughout the liver enable the creation of iodine

maps from contrast-enhanced dual-energy MDCT studies (DECT). Alternatively, virtual nonenhanced image series might be obtained by subtracting the iodine contribution from the contrast-enhanced image data sets. The latter approach has the potential for substantially reducing the radiation dose to patients for multiphase MDCT protocols of the liver125 (Fig. 43-4). A more recent promising application for DECT imaging of the liver is based on material density analysis. Material density images are

TABLE 43-2  Hepatic Monophase Contrast

MDCT Protocol: Portal Venous Phase Parameter

Protocol

Collimation

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp 30 sec

Slice thickness Pitch Table speed/gantry rotation Tube current Tube voltage Delay from initiation of bolus

TABLE 43-3  Hepatic Dual-Phase Contrast

MDCT Protocol

PROTOCOL Parameter

Arterial Phase

Collimation

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec

TABLE 43-1  Hepatic NCE MDCT Protocol Parameter

Protocol

Collimation

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp Not applicable

Slice thickness Pitch Table speed/gantry rotation Tube current Tube voltage Delay from initiation of bolus

1273

Slice thickness Pitch Table speed/gantry rotation Tube current

Tube voltage Delay from initiation of bolus

Automatic adaptation 100-700 mA/ 340-380 mA 120 kVp Arterial bolus tracking

Portal Venous Phase 64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/ 340-380 mA 120 kVp 30 sec

TABLE 43-4  Hepatic Triple-Phase Contrast MDCT Protocol PROTOCOL Parameter

Non–Contrast-Enhanced

Arterial Phase

Portal Venous Phase

Collimation

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp Not applicable

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp Arterial bolus tracking

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp 30 sec

Slice thickness Pitch Table speed/gantry rotation Tube current Tube voltage Delay from initiation of bolus

1274

PART II  CT and MR Imaging of the Whole Body

TABLE 43-5  Cholangiocarcinoma MDCT Protocol PROTOCOL Parameter Collimation

Slice thickness Pitch Table speed/gantry rotation Tube current Tube voltage Delay from initiation of bolus

Non–ContrastEnhanced

Arterial Phase

Portal Venous Phase

Delayed Phase

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp N/A

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp Arterial bolus tracking

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp 30 sec

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.7 55.0 mm/up to 0.28 sec Automatic adaptation 100-700 mA/340-380 mA 120 kVp 15 min

TABLE 43-6  The Partial Liver Resection

MDCT Protocol

PROTOCOL Parameter

Arterial Phase

Portal Venous Phase

Collimation

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 1.25 mm 1.375/1.75 55.0 mm/up to 0.28 sec

64 × 1.5 mm or 64 × 0.6 mm (=128 z-flying spot)/32 × 1.5 mm 3-5 mm 1.375/1.75 55.0 mm/up to 0.28 sec

Automatic adaptation 100-700 mA/ 340-380 mA 120 kVp Arterial bolus tracking

Automatic adaptation 100-700 mA/ 340-380 mA 120 kVp 30 sec

Slice thickness Pitch Table speed/gantry rotation Tube current

Tube voltage Delay from initiation of bolus

obtained using a two-material decomposition algorithm, which works by first establishing a basis pair and identifying associated attenuation characteristics of each material in the basis pair.13 Then each voxel is decomposed proportionally into the two-material basis pairs according to the attenuation characteristics of that specific voxel as measured at each energy level (typically 80 and 140 kVp). Material density images represent the relative distribution of densities in a given material, which produce the observed projections at the high and low energies.13 Therefore the values obtained on material density images do not represent absolute density measurements, but rather they reflect the true proportion or relative quantity of material under investigation.13 As an example, for a material basis pair composed of fat and iodine, the effective fat density image is represented by the fat-iodine data set, whereas the effective iodine density image is represented by the iodinefat data set. A region-of-interest measurement obtained on a fat-iodine image data set represents the relative effective density of fat without the iodine contribution. Similarly, an ROI measurement drawn over an iodine-fat image depicts the relative effective density of iodine without fat within a given target area.13

Magnetic Resonance Imaging MRI of the liver has been proven to be a competitive and comprehensive modality to assess morphology and functional characteristics of the liver in clinical scenarios of diffuse and focal liver diseases. Current clinical abdominal MRI is performed at magnetic field strengths from 0.5 to 3.0 tesla (T). Concurrent technical improvements such as the development of more powerful gradient systems and phased-array body coils, as well as implementation of advanced imaging sequence designs such as respiratory-triggered 3D data acquisition and sparse k-space sampling schemes, permit high-quality examination of the liver with both T1- and T2-weighted pulse sequences. Timing parameters of T1- and T2-weighted pulse sequences for dedicated hepatic imaging (e.g., repetition time [TR], echo time [TE]) are in part based on liver tissue–specific relaxation times, T1 and T2, which in turn show a strong dependency on the surrounding magnetic field strength. At 0.5 T magnetic field strength, liver parenchyma has a T1 relaxation time of approximately 327 to 518 milliseconds and a tissue-specific T2 relaxation time of 55 to 62 milliseconds. At 1.5 T, hepatic tissue’s relaxation time has significantly increased for T1 to 547 to 568 milliseconds and minimally decreased for T2 to 51 to 56 milliseconds. At 3.0 T, liver-specific T1 relaxation time has increased further to 809 milliseconds, while liver-specific T2 relaxation time is further reduced to approximately 45 to 50 milliseconds.43,63,216 The underlying principle of tissue-specific T1-weighted imaging is to use the shortest possible TE to maximize hepatic contrast and the number of slices attainable performing a breath-hold examination. Hepatic T2-weighted imaging on the other hand requires a TR in excess of at least four times the tissue’s specific T1 relaxation time and a TE that approximates the tissue’s specific T2 relaxation time. Therefore sequence designs of T1- and T2-weighted pulse sequences have to be partially redesigned for each given field strength to allow high-quality hepatic imaging in reasonable time frames. A comprehensive evaluation of the liver by means of MRI frequently involves intravenous contrast agents. As opposed to intravenous contrast in hepatic CT imaging, which is based on nonionic extracellular iodine-containing solutions with various concentrations of iodine, contrast-enhanced MRI of the liver uses a cluster of contrast agents that include nonspecific agents with an extracellular distribution, materials that are taken up specifically by hepatocytes and are partially excreted through the biliary system, as well as solutions that are targeted specifically to the Kupffer cells as part of the reticuloendothelial system (RES).

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

B

C

D

1275

FIG 43-4  Dual-energy MDCT of the liver in a 52-year-old man. Dual-energy MDCT workflow with conventional grayscale data set (A), water-density (B), iodine-density (C) and color-coded iodine overlay map (D).

Nonspecific gadolinium chelates without tissue-specific biodis­ tribution confined to extracellular spaces have been commercially available since 1986. Paramagnetic gadolinium shortens tissue-specific relaxation times, leading to an increase in hepatic tissue signal inten­ sities on T1-weighted sequences. Rapid redistribution of gadolinium chelates from intravascular to extracellular spaces requires a small-volume bolus injection of this contrast agent at up to 2 mL/sec, with an individual-adapted contrast dose of 0.1 to 0.2 mmol/kg  of body weight. These gadolinium chelates are available in various compositions as gadopentetate dimeglumine (Gd-DTPA), gadoteridol (Gd-HP-DO3A), gadodiamide (Gd-DTPA-BMA), gadoversetamide (Gd-DTPA-BMEA), gadoterate meglumine (Gd-DOTA), and gadobutrol (Gd-BT-DO3A), with varying approval from governmental healthcare authorities, in particular the latter not having been approved by the U.S. Food and Drug Administration (FDA). Gadofosveset trisodium represents a new class of blood-pool contrast agents specifically designed for visualization of abdominal or limb vessels in patients with known or suspected vascular diseases. Whereas in the past, extremely attractive safety profiles have been reported using these contrast agents, of late reports have been made of a possible association between gadolinium chelates and development of nephrogenic systemic fibrosis in patients with impaired renal function. Until further scientific evalua-

tions of this phenomenon have been conducted, patients with a glomerular filtration rate of less than 30 mL/min/1.73 m2 should not receive gadolinium-containing contrast agents.23,120,124,169,198 Hepatocyte-targeted contrast agents undergo active transport into the intracellular space of the hepatocytes, where they are further metabolized. They are partly eliminated through the biliary system and subsequently allow parenchymal and biliary assessment employing T1-weighted pulse sequences. Hepatocyte-selective contrast agents are either manganese based, as mangafodipir trisodium (Mn-DPDP), or gadolinium based, as gadoxetic acid disodium (Gd-EOB-DTPA) or gadobenate dimeglumine (Gd-BOPTA), the latter two combining the properties of nonspecific extracellular with tissue-specific intracellular contrast agents. The pharmacodynamics of these contrast agents differ significantly. Mangafodipir trisodium is administered as a slow intravenous infusion over 2 minutes; the overall dose of 5 to 10 mmol/kg of body weight is less than 10% of that for gadolinium-based agents. Therefore dynamic imaging is not possible, and imaging of mangafodipir trisodium in its short-term extracellular distribution does not add valuable information. In patients with pronounced cirrhosis and subsequent hepatic impairment, an association with an elevated risk to experience neurologic complications has been described. Gadoxetic acid disodium and gadobenate dimeglumine on the other hand allow

1276

PART II  CT and MR Imaging of the Whole Body

for imaging during the arterial phase, redistribution/portal venous phase, and equilibrium/hepatic venous phase, comparable to the other gadolinium chelates; however, they add an excretory phase to the contrast-enhanced liver imaging protocol. Because Gd-EOB-DTPA and Gd-BOPTA feature a T1 relaxivity slightly higher than conventional gadolinium-based agents, an overall contrast dose of 0.025  to 0.1 mmol/kg of body weight is sufficient to acquire the same information.40,52,170,223,224 RES-specific contrast agents are selectively incorporated by the Kupffer cells, which comprise 2% of the liver volume. These RESspecific contrast agents consist of small iron particles, which when larger than 50 nm are termed superparamagnetic iron particles (SPIO) and when smaller than 50 nm are called ultrasmall superparamagnetic iron particles (USPIO). Ferumoxides diluted in 5% dextrose solution are administered intravenously at least 30 minutes prior to imaging at a dosage of 10 µmol/kg of body weight, but imaging can also occur as long as 4 hours after administration. Approximately 80% of injected ferumoxides are taken up by macrophages in the Kupffer cells, and the remaining 20% are stored in the spleen and bone marrow. The superparamagnetic effect is based on susceptibility artifacts generated by the incorporated ferumoxides and subsequent shortening of tissue-specific T2 relaxation times. T2-weighted pulse sequences are conventionally used for parenchymal postcontrast imaging of RES-specific contrast agents. As with all tissue-specific contrast agents, lack of this particular tissue will lead to lack of uptake of contrast material; in the case of the negative SPIO and USPIO contrast agents, the surrounding liver will appear hypointense. Positive contrast agents, such as mangafodipir trisodium and gadobenate dimeglumine, will lead to a hyperintense visualization of hepatic parenchyma. Imaging characteristics of hepatic lesions will be based on the presence or absence of liver-specific Kupffer cells or hepatocytes.57,101,102,147,201 On NCE pulse sequences, physiologic liver parenchyma shows a higher T1 signal intensity and a lower T2 signal intensity compared to the spleen. The hyperintensity of hepatic parenchyma on T1-weighted pulse sequences is due to the large amount of hepatic protein and rough endoplasmic reticulum within the hepatocytes. Sharply demarcated from the hepatic parenchyma is the more hypointense biliary tree as well as portal and hepatic veins; hyperintense inflow artifacts, however, can be observed on individual images. Therefore NCE T1-weighted pulse sequences should not be evaluated for the patency of hepatic or portal venous vasculature. Bile throughout the biliary tree appears hyperintense on T2-  weighted pulse sequences in contrast to the hypointense visualization of the surrounding liver parenchyma. Heavily T2-weighted pulse sequences are subsequently used for acquisition of MR cholangiopancreatography (MRCP) images, in which only the biliary tree is visualized against a hypointense background. Indications for MRI of the liver include: • Characterization of diffuse or focal hepatic lesions of uncertain etiology • Determination of the extent and segment localization of hepatic malignancies prior to planned partial liver resection • Follow-up in cases of known primary or secondary hepatic malignancies

MRI Strategies.  MR sequences for hepatic imaging continue to evolve at a fast rate, whereas the clinical applications of each of these sequences have changed little over time. Unchanged remain the three basic considerations if MRI has been chosen for hepatic imaging: to improve parenchymal contrast, to suppress respiratory motion artifact, and to ensure complete anatomic coverage. Guaranteeing these three basic considerations in order to assess various clinical scenarios of

differing hepatic diseases using the most advanced designs of T1-weighted and T2-weighted imaging sequences remains the greatest challenge in hepatic MRI. A major recent development—particularly in abdominal MRI— was implementation of parallel imaging techniques using spatial information from single coils of a phased-array set to perform a portion of the spatial encoding normally accomplished by time-consuming gradients and radiofrequency (RF) pulses, thereby significantly reducing acquisition times. Different approaches such as the simultaneous acquisition of spatial harmonics (SMASH) technique or the sensitivity encoding (SENSE) procedure have been proposed, realizing spatial encoding with multiple spatial-distinct receiver coils for a sparse sampling of k-space. Subsequently, linear combinations of component coil signals are used to emulate the effects of phase encoding gradients. However, basic parallel imaging techniques require detailed and time-consuming coil mapping procedures for calibration purposes to prevent image distortions. Faster and more sophisticated calibration techniques acquire a block of extra lines in the center of k-space and subsequently perform a sliding block reconstruction. Unlike the internal autocalibration functions of SMASH techniques (AutoSMASH), data from multiple lines from all coils are used to fit an autocalibration signal line in one single coil. This fit gives the weights that can be used to generate the missing lines from that coil. Once all of the lines are reconstructed for a particular coil, a Fourier transformation can be used to generate the uncombined image for that coil. This process is repeated for each coil of the array, and finally the full set of images can be combined using a sum-of-squares reconstruction. This approach is also known as the generalized autocalibrating partially parallel acquisition (GRAPPA) technique.66,189,190 Acceleration of individual imaging sequences and use of respiratory triggering techniques (e.g., respiratory bellows or implemented diaphragmatic navigators) ensure that T1-weighted and T2-weighted pulse sequences can be acquired in excellent quality. Whereas T1-weighted imaging is currently performed during a single breath hold—particularly important for dynamic contrast-enhanced (DCE) imaging—T2-weighted pulse sequences are obtained in a seamless multisegmental end-expiratory fashion using respiratory triggering. Developments and improvements in magnetic field strengths, gradient technology, and receiver coil design, as well as implementation of parallel imaging techniques and respiratory triggering procedures, allow acquisition of a comprehensive hepatic MR exam in less than 30 minutes. For hepatic MRI the following sequence types are used: single-shot fast-spin echo (SSFSE), balanced steady-state free precession (SSFP), dual gradient echo in-phase (IP) and opposed-phase (OP) imaging, dynamic multiphase 3D spoiled gradient echo, diffusionweighted imaging (DWI), and MRI elastography. Single-shot fast-spin echo.  SSFSE acquisition schemes (e.g., half-Fourier acquisition single-shot turbo-spin echo [HASTE]) are based on a spin echo sequence with data acquisition after an initial preparation pulse for contrast enhancement, further employing a long echo train, with each echo being individually phase-encoded. Additionally a sparse k-space sampling scheme is used, acquiring only 56% of the raw-data matrix; partial Fourier reconstruction results in T2-weighted hepatic MRIs. An interleaved acquisition scheme is usually chosen, with long TRs to allow the spin system to recover between excitation pulses. The parenchymal contrast on SSFSE acquisition schemes may be improved with spectral fat-suppression techniques or with an inversion recovery pulse (short tau inversion recovery [STIR]) that provides additional contrast.78,80,95,200 FSE sequences can be combined with respiratory triggering procedures to allow 3D data sampling on heavily T2-weighted hepatic MRIs,

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases resulting in high-resolution MRCP image data sets. These MRCP data sets can be further postprocessed with maximum intensity projections (MIPs). Balanced steady-state free precession.  SSFP acquisition schemes (FIESTA, trueFISP) are based on a gradient echo sequence with a steady state developing for transverse and longitudinal magnetization, as well as a chosen TR that is shorter than hepatic T1 and T2 relaxivity parameters. Balanced gradients in all spatial directions are used between any RF excitation pulses to return the spin system to the same phase it had before RF pulses were applied. In this way, contrast as well as spatial and temporal resolutions allow this sequence type  to become a practical whole-body application. Acquired images  show T2* over T1 imaging characteristics. A major advantage of this sequence type is its insensitivity to motion artifacts. This sequence type is mainly used as a bright blood technique for unenhanced imaging of the portal vein. Dual gradient echo in-phase and opposed-phase imaging.  This incoherent gradient echo sequence (e.g., multiplanar gradient recalled [MPGR], fast low-angle shot [FLASH]) is characterized by low flip angles and uses a semirandom spoiler gradient after each echo readout to spoil the steady-state situation to destroy any remaining transverse magnetization by causing a spatially dependent phase shift. This spoiler gradient allows extremely short TRs, resulting in heavily T1-weighted image series. TEs have to be kept as short as possible to avoid susceptibility artifacts. This sequence type is also suitable for bolus timing; a single transverse slice positioned at the level of the upper abdominal aorta will detect the injected contrast bolus and, by improving the widely used “guesstimation” approach, initiate DCE MRI thereafter. However, by adjusting the TE appropriately, scans can be obtained with water and lipids in either an IP or OP state. On IP images, the signal from both water and fat protons contributes to tissue signal intensity. On OP images, signal from fat protons cancels that of water protons. This technique is useful in evaluating focal or diffuse fatty infiltration and for tissue characterization, such as in the case of incidentalomas in the adrenal glands. Usually the boundaries between tissues with different water or fat content appear black on OP images, also known as India ink artifact or chemical shift artifact of the second kind. Choice of TE for IP and OP images depends on field strength. High-field-strength MR imagers with implemented strong gradient systems allow acquisition of different IP and OP pairs; however, to determine the underlying phenomenon of signal intensity variation, either due to IP and OP spin vectors or free induction decay, radiologists must be aware of the TE pair selected.136,138,139,161,165,181 Dynamic multiphase 3D spoiled gradient echo.  The advent of high-performance gradient systems and implementation of parallel imaging techniques (e.g., liver acquisition with volume acceleration [LAVA], volumetric interpolated breath-hold examination [VIBE], driven equilibrium [DRIVE]) have rendered 3D acquisition schemes with zero-fill interpolation, overlapping thin sections, and effective spectral fat saturation a superior method compared to previously established two-dimensional (2D) and multisectional DCE T1-weighted imaging. Detection and characterization of diffuse and focal liver diseases is significantly enhanced when the entire liver can be imaged during a single suspended respiration. Specifically, to optimize arterialphase timing, a single-breath-hold triple-arterial-phase acquisition— three successive acquisitions in the same breath hold—is increasingly adopted (Fig. 43-5). This technique has a number of advantages over a single arterial-phase acquisition technique. Indeed, a triple-arterialphase acquisition allows for increasing the likelihood of acquiring at least one precisely timed late hepatic arterial-phase image set while also

1277

mitigating the effect of transient artifacts by isolating them to one or two of the image sets. In addition a 2D parallel acceleration technique (controlled aliasing in parallel imaging results in higher acceleration [CAIPIRINHA]) combined with a 3D T1-weighted spoiled gradient recalled (SPGR) echo pulse sequence (e.g., volumetric interpolated breath-hold examination [VIBE]) more recently enables whole-liver coverage with a short temporal footprint. Further developments of this sequence type showed that 3D spoiled gradient echo acquisition schemes without spectral fat saturation  can be used to obtain high-resolution data sets of T1-weighted IP  and OP imaging, which allow calculation of fat-only and water-only data sets using Dixon methods of data summation or subtraction, respectively.111,164,222 Diffusion-weighted imaging.  Over the past decade, DWI has become a routine component of liver MR sequence protocols. DWI offers information based on hepatic water proton movements, which complements the morphologic information obtained with conventional MRI. DWI is routinely adopted either to detect or characterize hepatic disease, both benign and malignant hepatic lesions, with robust data showing the usefulness of this technique.19 DWI is carried out by supplying sequences with additional gradients implemented during the acquisition. Specifically, T2-weighted sequences, compared with conventional spin echo sequences, have an additional symmetric pair of diffusion-sensitive gradients that are routinely employed. As a result of the first diffusion gradient, water protons undergo an initial phase shift. This shift will be completely rephased by the second diffusion gradient, without the remaining net phase shift present for protons showing restricted diffusion. By comparison, water molecules without substantial diffusion restriction acquire different phase information between the first and second gradient; their signal will not be completely rephased by the second gradient, thus leading to a signal loss.19 The degree of water motion has been found to be inversely related to the degree of signal attenuation. The sensitivity of the DWI sequence to water motion can be varied by changing the amplitude of the diffusion-weighted gradient, the duration of the applied gradient, and the time interval between the paired gradients. These variations in gradient characteristics are expressed in one numeric parameter generally referred to as the b value. In everyday clinical practice the diffusion sensitivity can be modified by changing the b value, which is usually given in seconds per square millimeter. When the b value is changed, usually the gradient amplitude rather than the duration or time intervals between gradients is altered. For example, hepatic hemangiomas or cysts appear hyperintense on standard T2-weighted images and may also appear hyperintense on DWI, because the sequences are based on T2-weighted spin echo sequence designs (so-called T2 shine-through effect) (Fig. 43-6). To eliminate the impact of T1, T2, TR, and TE, apparent diffusion coefficient (ADC) maps can be calculated to obtain pure diffusion information. To calculate ADC maps for each pixel in a DWI series, at least two DWI data sets with differing b values (usually one b value in the range of 0-100 and a second b value ≥400) are postprocessed. The ADC maps represent grayscale-encoded ratios of logarithmized differences in signal intensities as well as differences in b values of the two sequential DWI sequences. Areas of true restricted diffusion demonstrate high signal intensity on DWI, with corresponding low signal intensity on the ADC map. Areas of relatively free  diffusion also appear hyperintense on DWI, corresponding to T2 shine-through effects, but will remain hyperintense on the ADC map.19 MRI elastography.  Accumulating evidence has shown that measurements of elasticity of liver parenchyma show a strong correlation with degree of hepatic fibrosis.71 Initially this theoretic concept was exploited by ultrasonography-based imaging techniques, where 

1278

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 43-5  T1-weighted contrast-enhanced MRIs of the liver in a 61-year-old man, showing early arterial (A), midarterial (B), and late arterial (C) phases of DCE hepatic imaging.

A

B FIG 43-6  MRI of the liver for suspected hepatic lesions in a 46-year-old woman. DWI obtained with a b value of 50 (A) and corresponding ADC maps (B) allow for confidently posing a diagnosis of nonenhancing hepatic cyst (arrow).

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases external compressional impulses are visualized with Doppler ultrasonography as shear-wave transients propagating through liver parenchyma and directly depending on the elasticity of liver parenchyma itself.71 By comparison, MRI elastography techniques combine the application to the liver parenchyma of an external sinusoidal vibration, with a 2D gradient echo MR elastography sequence. The latter applies motion-encoding gradients along the section-selection direction to detect cyclic motion in the through-plane direction, ensuing in colorcoded maps representing regional elastic properties of the liver parenchyma.19

MRI Sequences.  Numerous available MR imagers operating at magnetic low field strengths of 0.5 T, midfield strengths of 1.5 T, and high strengths of 3 T operate a variety of vendor-specific imaging sequences based on the prior introduced basic MR acquisition schemes. Attempts to unify at least the terminology of MR sequences and parameters have been unsuccessful up to now, and so has the ability to reproduce comparable and clearly manageable listings of sequence chronologies for dedicated MR examinations. The advent of parallel imaging capabilities and subsequent multichannel technologies of receiver coils have further complicated this endeavor. Despite these complicating factors, we intend to include a potpourri of MR sequences and their parameters, which allow a comprehensive hepatic MR workup in a time frame of less than 20 minutes. This imaging protocol includes IP, OP, water-only, and fat-only T1-weighted imaging of the entire liver in one breath hold, respiratorytriggered T2-weighted imaging using fat-suppressed T2-weighted and high-resolution 3D MRCP sequences, and contrast-enhanced fatsaturated T1-weighted imaging with three arterial perfusion phases acquired during one breath hold, one portal venous phase, and one equilibrium phase after administration of a biliary excreted gadoliniumbased contrast agent. Delayed hepatic imaging can be performed optionally 2 hours after initial contrast application (Fig. 43-7). Table 43-7 is a comprehensive fast liver MRI protocol optimized for 3-T scanners.

DIFFUSE PARENCHYMAL DISEASES OF THE LIVER The liver plays several complex roles in the metabolism of amino acids, carbohydrates, and lipids, as well as in the synthesis of a multitude of proteins. The liver is also one of the key organs in the multifarious systems of hematologic coagulation and biliary recycling. The basic pathophysiology of all diffuse parenchymal hepatic diseases is a failure in one of the pathways of these complex hepatic metabolic functions. A wide spectrum of extrinsic agents, organisms, viruses, and intrinsic factors (e.g., enzyme defects, disorders of arterial or portal venous supply and hepatic venous drainage, primary specific inflammatory processes) have been identified as activators, catalysts, or triggers of diffuse parenchymal diseases of the liver. Patients’ medical history and hepatic laboratory results derive from these hepatic disturbances and subsequent impairments; hepatic MDCT and MRI can help substantiate clinical diagnoses and verify success of treatment regimens. Depending on the predominant pathophysiologic processes, the diffuse parenchymal hepatic diseases discussed in this chapter have been categorized as storage diseases, vascular diseases, inflammatory parenchymal diseases, and cirrhosis.

Storage Diseases

Fatty Liver Disease.  Hepatic steatosis is the result of excessive accumulation of triglycerides within hepatocytes, a phenomenon that can affect hepatic parenchyma focally or diffusely. On a cellular level, three

1279

pathophysiologic phenomena have been identified that contribute to fatty infiltration of liver parenchyma: decreased mitochondrial fatty acid β-oxidation, increased endogenous fatty acid synthesis or enhanced alimentary delivery of fatty acids, and deficient incorporation or export of triglycerides as very low-density lipoproteins (VLDL).2,3,46 Fatty liver disease can cover a spectrum from dormant noninflammatory fatty liver, termed steatosis, to steatohepatitis with inflammation, fibrosis, and eventually cirrhosis. The term alcohol-related fatty liver denotes fatty liver disease with its entire spectrum of dormant to active inflammatory variants associated with alcohol use—as little as 300 mL ingested per week. In the absence of alcohol consumption, nonalcoholic fatty liver disease (NAFLD) is most commonly associated with obesity, type 2 diabetes mellitus, and dyslipidemia. It can present with an aggressive subtype called nonalcoholic steatohepatitis (NASH), characterized by hepatocyte ballooning and necrosis with and without Mallory’s hyaline and fibrosis. Several other clinical disorders and scenarios including malnutrition, severe hepatitis, corticosteroid use, pregnancy, drug toxicity, malaria, kwashiorkor, Reye’s syndrome, and chemotherapy can also lead to fatty liver disease.5,27,39,176 On unenhanced hepatic MDCT, fatty infiltration results in lowering of the attenuation of the liver parenchyma. In normal adults the attenuation value of the liver is consistently higher than that of the spleen, with a mean difference of 8 HUs.121,160,195 Mild degrees of diffuse fatty infiltration can be diagnosed when the attenuation value of liver parenchyma is less than that of the spleen. Marked steatosis hepatis leads to a characteristic MDCT appearance in which the attenuation of the liver is even lower than that of the blood vessels, in particular the central portal venous branches and IVC. Although fatty infiltration can be depicted on contrast-enhanced MDCT examinations, contrasted images are less reliable and less specific for detecting steatosis hepatis. Following contrast administration, significant fatty infiltration can be suggested if liver attenuation is less than that of muscle. Both focal fatty infiltration and residual foci of unaffected liver parenchyma surrounded by fatty infiltration (fatty sparing) may be confused with neoplastic lesions on MDCT examinations: • Focal fatty infiltration commonly has a segmental or lobar distribution, often with a straight-line margin providing a clue  to diagnosis. These large and often irregular-shaped lesions typically extend to the liver capsule, usually without associated bulging of the liver contour. In addition, vessels coursing through the area of abnormality are not displaced. Common locations for nodular fatty infiltration are in the anterior part of segment IV adjacent to the falciform ligament and the fissure for the ligamentum teres. • Focal areas of higher attenuation within a region of fatty change can be caused by neoplastic masses as well as regions of fatty sparing. The diagnosis of focal hepatic sparing is suggested when a geographic area of higher attenuation is identified in a typical location such as along the gallbladder fossa adjacent to the interlobar fissure, next to the ligamentum teres, or in a subcapsular location. Dual-energy MDCT is another promising imaging technique for grading fatty changes of the liver. Preliminary experiences have shown that a significant correlation exists between fat material density values and amount of fatty deposition69,76,98,153 (Figs. 43-8 to 43-12; see Fig. 43-1). When evaluating the hepatic parenchyma with conventional T1-weighted spin echo sequences, only subtle differences in signal intensity (range, 5%-10%) are seen between physiologic and fatty parenchyma. This difference is only present when affected parenchyma

1280

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F FIG 43-7  Multiphase contrast-enhanced MRI of the liver using the comprehensive fast liver imaging protocol. Coronal (A) and transverse (B) SSFSE T2-weighted imaging of the entire abdomen. C to F, Parallel precontrast T1-weighted imaging using 3D acquisition schemes with zero-fill interpolation and overlapping thin sections, allowing IP and OP as well as calculation of fat-only and water-only data sets using Dixon methods.

G

H

I

J

K

L

M

N

FIG 43-7, cont’d  G, Respiratory-triggered T2-weighted STIR sequences are then acquired. H to J, Contrastenhanced hepatic MRI is then performed by initially acquiring hepatic arterial fat-saturated parallel T1-weighted sequences using 3D acquisition schemes in a multiphasic fashion. K, Portal venous-phase T1-weighted imaging follows. A T1-weighted fat-saturated hepatic equilibrium phase (L) is visualized after respiratorytriggered T2-weighted MRCP images (M) are acquired. N, Optionally a delayed T1-weighted fat-saturated imaging sequence can be acquired 2 hours after contrast application.

1282

PART II  CT and MR Imaging of the Whole Body

TABLE 43-7  Comprehensive Fast Liver MRI Protocol Optimized for 3-T Scanners

Sequence

T2W

T1W IP, OP, Water-Only, Fat-Only

Acquisition scheme Coverage Triggering/breath hold TR TE Pixel bandwidth Slice thickness Field of view Imaging plane

SSFSE

SGE

2D (Non) Breath hold 1339 msec 141.312 msec 325.5 Hz 10 mm 480 mm Transverse and coronal 384 × 128 2 × 20 sec

3D Breath hold

Matrix Acquisition time

T2W Fat-Sat FSE

2D Respiratory triggered 5.026 msec 8571.4 msec 1.336/2.622 msec 90.22 msec 651.1 Hz 244.1 Hz 5 mm 5 mm 340 mm 340 mm Transverse Transverse

320 × 192 24 sec

384 × 224 3-5 min

T1W Fat-Sat T1W Fat-Sat T2W Triple Arterial Portal Venous Fat-Sat Phase Phase MRCP Dynamic SGE multiphase SGE 3D 3D Breath hold Breath hold 3.536 msec 1.72 msec 651 Hz 6 mm 340 mm Transverse

4.364 msec 2.1 msec 244.1 Hz 4.4 mm 340 mm Transverse and coronal 288 × 192 25 sec

256 × 128 3 × 8 sec

T1W Fat-Sat T1W Fat-Sat 2-Hour Delay Equilibrium Phase, Phase Optional

SSFSE

SGE

SGE

3D Respiratory triggered 3529.4 msec 810.92 msec 162.8 Hz 1.6 mm 340 mm Coronal

3D Breath hold

3D Breath hold

4.364 msec 2.1 msec 244.1 Hz 4.4 mm 340 mm Transverse

288 × 288 2-5 min

288 × 192 25 sec

4.364 msec 2.1 msec 244.1 Hz 4.4 mm 340 mm Transverse and coronal 288 × 192 25 sec

Fat-Sat, fat-saturated; FSE, fast-spin echo; IP, in-phase; MRCP, MR cholangiopancreatography; OP, opposed-phase; SGE, spoiled gradient echo; SSFSE, single-shot fast-spin echo; T1W, T1-weighted; T2W, T2-weighted; TE, echo time; TR, repetition time.

A

B FIG 43-8  Geographic fatty infiltration of the liver in this 45-year-old man can be seen on both noncontrast (A) and late hepatic arterial phase (B) images.

contains at least 10% triglycerides. T2-weighted sequences are especially insensitive to diagnose fatty hepatic parenchyma. Gradient echo phase shift imaging, however, has proven to be  an excellent tool for differentiating nodular fatty infiltration from  neoplasms. Employing IP and OP images, fatty infiltration can be readily diagnosed by comparing signal intensity. On OP images, focal fatty infiltration appears as a region of decreased signal intensity compared to the signal intensity of the corresponding area on the IP image81,108,114,155,158 (Fig. 43-13). In a clinical scenario of a liver lipoma containing only fat, no significant changes in signal intensity on IP and OP images occurs; in this clinical situation a fat-suppressed T1-weighted sequence will be able to confirm the diagnosis, showing a significant drop in lesional signal intensity.

Hemochromatosis.  In HLA-associated inherited primary hemochromatosis, a defect in the intestinal mucosa leads to increased

gastrointestinal resorption of iron despite already adequate iron storage. As a consequence, iron is deposited first in the hepatocytes and then, as the disease progresses, in other tissues such as the pancreas, gastrointestinal tract, kidneys, heart, joints, and endocrine glands (e.g., pituitary), with resulting destruction of these tissues. Iron is a direct hepatotoxin; subsequent thin fibrous septations develop slowly. The disease usually manifests during the fourth or fifth decade with hepatomegaly, ultimately leading to micronodular cirrhosis of the organ. Clinical symptoms include diabetes, arthralgia, cardiac insufficiency, and hypogonadism. Hepatocellular carcinoma is a late complication and occurs in up to 30% of cases, making it important to diagnose this inherited type of hemochromatosis.12,50 Secondary hemochromatosis presents with the same symptoms but is encountered in patients with chronic anemias who have received multiple blood transfusions. Additional disease forms occur with congenital transferrin deficiency, porphyria cutanea tarda, excessive iron ingestion, and complications of portocaval shunting.91,175

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

1283

B

C FIG 43-9  Dual-energy MDCT of the liver in a 65-year-old man with chemotherapy-induced mild liver steatosis. A mild fatty infiltration (arrow) of the left lower lobe can be suspected on conventional imaging (A). The latter suspicion is confirmed by either grayscale (B) or color-coded fat-iodine (C) maps.

FIG 43-10  NCE MDCT shows generalized steatosis hepatis and an area of focal fatty sparing (arrows) in this 42-year-old man. Note the nonobstructing renal stone on the left.

FIG 43-11  NCE MDCT shows focal fatty infiltrations (arrows) in the caudate lobe and hepatocentric and posterior right hepatic lobe in this 59-year-old woman. Note ascites surrounding the liver.

1284

PART II  CT and MR Imaging of the Whole Body As discussed earlier, NCE MDCT at best is able to establish the presence of a hepatic storage abnormality. Differentiating between hemochromatosis, in which there is an iron uptake of the hepatocytes, and hemosiderosis, in which iron is deposited in the Kupffer cells  of the hepatic reticuloendothelial system or the splenic monocyte/ macrophages system, is not possible using MDCT alone (Fig. 43-17). Abdominal MRI, on the other hand, is capable of distinguishing these disease entities. In hemosiderosis, low signal intensities are seen in the liver and spleen, similar in appearance to intravenously administered superparamagnetic iron oxide–based contrast agent. In hemochromatosis, hyperintense splenic signal and hypointense pancreatic signal characteristics are observed106,116,174 (Fig. 43-18).

Wilson’s Disease.  Impaired biliary excretion of copper and its sub-

FIG 43-12  Contrast-enhanced hepatic MDCT showing histologyproven nonalcoholic steatohepatitis in a 41-year-old man.

Hemochromatosis of the liver can be detected with NCE MDCT, but with moderately limited sensitivity and significantly reduced specificity. Increased attenuation within the hepatic parenchyma to levels in excess of 70 HU on unenhanced MDCT examinations is considered characteristic for iron overload.32,34,68,166 On the other hand, mild forms of iron overload or cases with coexistent steatosis hepatis may appear normal at MDCT. In addition a hyperdense appearance of the hepatic parenchyma is not only seen in hemochromatosis but can also be observed in hemosiderosis, glycogen storage disease, Wilson’s disease with increased copper deposition in the liver and brain, patients with cardiac arrhythmias taking amiodarone, and patients with a history of Thorotrast contrast medium exposure. Also, colloidal gold (used in arthritis treatment) and cisplatin therapy have been reported to increase liver attenuation.42,64,126,156,159,184 MRI has been shown to be significantly more sensitive in detecting milder forms of iron storage diseases due to hemochromatosis and has also been proven more specific than unenhanced MDCT scans. On MRI, affected hepatic parenchyma demonstrates signal reduction due to the paramagnetic effect of the ferric ions, Fe3+, which cause shortening of the T1 and T2 relaxation times of neighboring protons. T2 shortening is the predominant effect and accounts for the reduction in hepatic signal intensity, which is seen best on gradient echo T2-weighted imaging; gradient echo sequences lack the 180-degree refocusing pulse, and Fe3+-induced magnetic inhomogeneities are therefore emphasized. Furthermore, by acquiring multiple T2-weighted sequences with varying TEs, individual T2 relaxation parameters  can be calculated and used for evaluation of treatment efficacy25,193 (Figs. 43-14 to 43-16).

Hemosiderosis.  Aplastic and hemolytic anemias require frequent blood transfusions, with subsequent increased storage of ferric irons in the liver, spleen, lymph nodes, and bone marrow. Whereas the hepatic iron overload in hemochromatosis leads to increased uptake of iron into the hepatocytes, hemosiderosis is characterized by iron deposition within the reticuloendothelial system. Analogous to hemochromatosis, clinical presentations of hemosiderosis include the final cirrhotic state of the liver, highlighting the importance of early diagnosis of hemosiderosis and, for treatment options, its correct differentiation from the inherited type of hemochromatosis.

sequent accumulation in tissues, predominantly the liver and basal ganglia, characterizes this autosomal recessive disease, also referred to as hepatolenticular degeneration owing to deposition of copper ions in basal ganglia. Whereas no malignant transformation from hepatic copper deposition to hepatocellular carcinoma is known, Wilson’s disease can present as acute and even fulminant hepatitis with a rapid progression into mostly macronodular-type cirrhosis. Clinically Wilson’s disease presents with very low levels of ceruloplasmin. MDCT imaging demonstrates a nonspecific hyperdense hepatic attenuation profile due to copper deposition in the liver (Fig. 43-19). As of yet the MRI features of Wilson’s disease do not show characteristic features that distinguish it from other forms of chronic hepatic disease. In clinical scenarios of Wilson’s-induced fulminant hepatitis and subsequent liver failure, early patchy contrast enhancement with significantly delayed stromal gadolinium uptake have been observed; however, these are nonspecific imaging findings.133,178,221 Nevertheless, accumulation of copper ions in the periportal regions and along the hepatic sinusoids may be the hallmark of the hepatic Wilson’s disease, with clinical episodes of acute hepatitis with fatty changes in the liver and periportal inflammation.133,178,221

Reye’s Syndrome.  In the pediatric population, fatty degeneration of the liver occurring in concert with an encephalopathy has been described after exposure to influenza B and varicella viruses and  subsequent treatment with acetylsalicylic acid. This treatment causes a limited hepatomegaly with intermittent steatosis hepatis. Clinically the only decisive prognostic indicator is the extent of neurologic symptoms.16,31,38,162 MDCT and MRI of the liver depict mild hepatomegaly with accompanying steatosis hepatis, which are nonspecific imaging findings  (Fig. 43-20).

Glycogen Storage Disease.  All known forms of glycogen storage disease are inherited autosomal recessive disorders and are characterized by absence or deficiency of one of the enzymes responsible for producing or metabolizing glycogen. The enzyme deficiency causes either abnormal tissue concentrations of glycogen or abnormally formed glycogen. There are 11 known types of glycogen storage disease involving the liver, musculature, hematopoietic system, myocardium, and kidneys. All known variants of glycogen storage diseases—in particular types Ia (von Gierke’s), Ib, II (Pompe’s), III (Forbes’), and IV (Anderson’s)—as well as VI and IX, have the cardinal symptom of hepatomegaly. Types Ia and Ib have the potential to undergo malignant transformation to form hepatocellular carcinomas, whereas types III and IV progress to develop cirrhosis. In types Ia, Ib, and III an increased incidence of hepatic adenomas has been observed.143,167 Extrahepatic findings include but are not limited to glomerulonephritis, amyloidosis, hypoglycemia, and osteoporosis (types Ia and Ib) as well as

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

B

C

D

1285

FIG 43-13  Hepatic MRI shows diffuse steatosis hepatis, predominantly in the right lobe, in this 66-year-old woman. Note the dropout of hepatic signal intensity on OP imaging (A) compared to IP imaging (B). Fat-only (C) and water-only (D) imaging reemphasize the fatty infiltration.

myocardial (types II and IV) and striated muscular insufficiency/ weakness (type III). In all children in whom hepatomegaly is detected in combination with hypoglycemia, growth retardation, a disproportional distribution of body fat, and increased transaminases, the diagnosis of glycogen storage disease should be considered (Figs. 43-21 and 43-22).

Galactosemia and Hereditary Tyrosinemia.  Infants with an inherited autosomal recessive defect in the enzyme galactose 1-phosphate uridyltransferase present with clinical symptoms such as growth retardation, nausea, vomiting, diarrhea, and jaundice, usually after their first exposure to galactose in milk. Delayed symptoms include lethargy, irritability, and convulsions. Subtypes of galactosemia have been identified and are based on deficiencies of galactose kinase and galactose 6-phosphate epimerase.86,186 Initial imaging findings of cirrhotic changes on hepatic MRI within 6 months after onset of symptoms have to be considered an already

advanced stage of the disease. Extrahepatic manifestations include mental retardation, cataract formation, ascites, and renal failure. Hereditary tyrosinemia, also referred to as tyrosinosis, is an autosomal recessive enzymatic disorder that causes progressive liver disease with either micro- or macronodular cirrhosis. Patients with cirrhosis due to tyrosinemia may display high-attenuating hepatic nodules that may be challenging to differentiate from hepatocellular carcinoma lesions.41

Lipid Storage Disease and Mucopolysaccharidoses.  The spectrum of lipid storage disorders subsumes a diverse family of individual diseases related by their molecular pathology, in particular  an enzymatic deficiency of a lysosomal hydrolase. This enzymatic  deficiency leads to lysosomal accumulation of the enzyme’s specific substrate, sphingolipid. Because glycosphingolipids are one of the essential components of all cell membranes, they can be found  ubiquitously throughout the human body. The inability to properly 

1286

PART II  CT and MR Imaging of the Whole Body

FIG 43-15  T2-weighted hepatic imaging using SSFSE acquisition with FIG 43-14  T2-weighted hepatic imaging using SSFSE acquisition and

a long echo train shows hemochromatosis in this 33-year-old man.

a sparse k-space sampling shows hemochromatosis in this 29-year-old woman. Note Gamna-Gandy bodies in the spleen.

FIG 43-17  NCE MDCT in a 45-year-old man with hemolytic anemia FIG 43-16  T2-weighted hepatic imaging using STIR sequence shows

who required frequent blood transfusions and subsequently developed hemosiderosis.

hemochromatosis in this 39-year-old man.

A

B FIG 43-18  Hemosiderosis in a 51-year-old woman as seen on precontrast T1-weighted hepatic images using OP (A) and IP (B) imaging. Note the dropout of hepatic signal intensity due to induced magnetic field inhomogeneities in the sequence with the longer TE.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

1287

FIG 43-21  Contrast-enhanced MDCT in an 11-year-old girl with glycogen storage disease type Ia (von Gierke’s disease). Note the round hepatic adenoma in the left hepatic lobe.

FIG 43-19  NCE MDCT in a 41-year-old man with histology-proven Wilson’s disease.

FIG 43-22  Contrast-enhanced hepatic MRI in a 12-year-old girl with glycogen storage disease type III (Forbes’ disease). Note the small round hepatic adenoma in the right hepatic lobe.

FIG 43-20  Contrast-enhanced MDCT in a 10-year-old boy with Reye’s syndrome.

degrade glycosphingolipids leads to pathologic accumulation of enzymatic substrates, resulting in micropathologic and morphologic alterations that have characteristic clinical manifestations. Sphingolipid substrate storage in visceral cells can lead to organomegaly, skeletal abnormalities, and inflammatory infiltrations. The main target organ, however, is the central nervous system, which is responsible for the predominantly neurodegenerative course of lipid storage diseases. Depending on the enzymatic deficiency, the following subtypes  of lipid storage diseases have been identified: GM1 gangliosidosis (β-galactosidase), GM2 gangliosidosis (β-hexosaminidase), Gaucher’s disease (glucocerebrosidase), Niemann-Pick’s disease (sphingomye­ linase), Fabry’s disease (α-galactosidase A), Schindler’s disease (αgalactosidase B), metachromatic leukodystrophy (arylsulfatase A), Krabbe’s disease (galactosylceramide-β-galactosidase), Farber’s disease (ceramidase), and Wolman’s disease (cholesteryl-ester-hydrolase). In particular Gaucher’s disease, Niemann-Pick’s disease, Farber’s disease, Wolman’s disease and the gangliosidoses present with an enlarged liver

due to lipid and cholesterol deposition, which leads to cirrhosis and chronic liver failure before adulthood.1,26,28,29,51,60,137,210 MDCT and MRI of the liver show hepatomegaly with accompanying steatosis hepatis as nonspecific imaging findings (Fig. 43-23). Deficient activity of the enzymes regulating catabolism of glucosaminoglycans leads to accumulation of excess mucopolysaccharides in the somatic and visceral tissues, as well as excretion of specific metabolites in the urine. In addition to mucopolysaccharides, gangliosides also accumulate. Of the mucopolysaccharidoses, type I (Hurler’s), type II (Hunter’s), type III (Sanfilippo’s), type IV (Maroteaux-Lamy), and type VIIb have hepatic manifestations. Clinically hepatomegaly, diffuse fibrosis with heavy deposits of collagen bundles, and microdissection of parenchyma into nodules, as well as micronodular or macronodular cirrhosis can be observed.14,226 MDCT and MRI of the liver show cirrhotic changes, a nonspecific imaging finding (Fig. 43-24).

Amyloidosis.  Diffuse hepatic disease can be seen in either primary or secondary amyloidosis, which in both cases ensues in excessive deposition of fibrils of protein-mucopolysaccharide complexes.134 In

1288

PART II  CT and MR Imaging of the Whole Body

A

B FIG 43-23  Precontrast T1-weighted (A) and T2-weighted (B) hepatic MRIs in a 15-year-old boy with Gaucher’s disease.

A

B

C

D FIG 43-24  Mucopolysaccharidosis type I (Hurler’s syndrome) in a 19-year-old man as seen on precontrast T1-weighted OP (A) and IP (B), as well as contrast-enhanced (C) sequences and T2-weighted (D) hepatic MRI. Note the extended cirrhosis and area of fibrosis (arrows). Ascites and splenomegaly is also present.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

1289

B FIG 43-25  59-year-old male with non-occlusive thrombus in the main portal vein (A) and splenic vein (B) as seen on contrast-enhanced MDCT acquired during the portal venous contrast phase.

primary amyloidotic involvement of the liver, a mild to severe enlargement of the liver is observed owing to deposition of amyloid fibrils along the hepatic sinusoids and within Disse’s spaces.134 By comparison, secondary amyloidosis is often characterized by a relative sparing of sinusoidal spaces, with a more prominent involvement of the hepatic artery walls. In secondary amyloidosis, there is usually a concomitant involvement of the spleen and kidneys that may appear enlarged as well.134

Vascular Diseases

Portal Venous Thrombosis—Cavernous Transformation of the Portal Vein.  Virchow’s triad defines the three main mechanisms that lead to thrombosis as hypercoagulability, stasis, and disruption of the vascular endothelium. Hypercoagulable states usually result from systemic diseases such polycythemia vera, antithrombin III deficiency, idiopathic thrombocytosis, chronic myeloid leukemia, protein C deficiency, and paroxysmal nocturnal hemoglobinuria. In contrast, portal venous stasis is often due to local abnormalities such as cirrhosis, large hepatocellular carcinoma, and pancreatic carcinoma; it is also a known complication following splenectomy. Similarly, disruption of the portal venous endothelium has been observed in the clinical  scenario of sepsis, direct trauma, schistosomiasis, chronic inflammatory bowel disease, pylephlebitis, and as a complication after direct catheterization. Portal venous thrombosis has been described as a transient disease; however, spontaneous revascularization is rare once the intravascular thrombus has reached smaller portal venous branches. If no spontaneous resolution of the intraluminal thrombus occurs, portoportal venous channels may form to maintain portal perfusion. These portal pathways can be found not only in the porta hepatis but also within the liver. In the porta hepatis a mesh of fine to markedly enlarged serpiginous vessels can be observed in place of the occluded main portal vein. Intrahepatic portal venous shunts form, extending from one segmental portal vein to another. Simultaneously, portosystemic collateral channels are forming, suggesting that despite extensive hemodynamic adaptations, portal hypertension is present. Chronic occlusion of the portal vein or one of its main segmental branches may be accompanied by segmental atrophy of the effected hepatic segments as well as compensatory hypertrophy of the other segments. In livers

with preexisting diffuse fatty infiltration, thrombosis of intrahepatic segmental portal venous branches can result in focal fatty sparing, because fat delivery correlates with portal blood flow. MDCT and MRI of the liver are both capable of visualizing portal venous thrombosis and cavernous transformations of the portal vein during the portal venous contrast phase. Furthermore these crosssectional imaging modalities can contribute to the planning of shunt surgery and monitoring of shunt patency in a clinical setting of portal venous thrombosis. In rare cases, cavernous transformation of the portal vein can appear as a subhepatic spongelike mass. This appearance has also been reported in cases of pancreatic hemangiosarcoma and bile duct varices. Hepatic MRI with multiple arterial and portal venous contrast phase acquisitions in combination with biliary high-resolution MRCP imaging will contribute important information in the diagnostic workup. Portal vein patency can be confirmed or excluded by acquiring flow-sensitive gradient echo sequences such as steady state with free precession6,44,49,89,115,122,149,183,199,207 (Figs. 43-25 to 43-28).

Obstruction of Smaller Portal Venous Branches.  Thrombosis and subsequent fibrosis of smaller portal venous branches can result from hepatopedal extension from an existing portal venous thrombosis. Isolated peripheral occlusions of smaller portal venous branches have been reported to occur in systemic diseases of vasculitic and rheumatoid origin and are an early manifestation of primary sclerosing cholangitis (PSC), primary biliary cholangitis (PBC), and sarcoidosis. Thrombosis of smaller portal venous branches as a complication of a local inflammatory process can occur in the clinical scenario of schistosomiasis.4,45,93,117,146,163,229 Analogous to imaging of portal venous thrombosis and cavernous transformation of the portal vein, hepatic MDCT and MRI to detect thrombosis and subsequent fibrosis of smaller portal venous branches is best achieved during the portal venous perfusion phase (Figs. 43-29 and 43-30).

Portal Venous Air.  Portal venous air is recognized not as a specific disease entity but as a diagnostic clue in patients with acute abdominal pathology. The pathogenesis of this imaging sign is still not fully understood.

A

B FIG 43-26  Contrast-enhanced fat-saturated T1-weighted MRI in a 62-year-old woman with near-occlusive thrombus in the main portal vein (arrow in A). Note the layering phenomenon/flow-void artifacts of the intraluminal thrombus formation (arrow in B).

FIG 43-28  62-year-old male with mass-like cavernous transformation FIG 43-27  Contrast-enhanced MDCT acquired during the portal venous contrast phase shows cavernous transformation of the portal vein following long-standing portal venous thrombosis in this 49-year-old man. Ascites and status post splenectomy are also present.

of the portal vein following long-standing portal venous thrombosis (arrows) as seen on contrast-enhanced fat-saturated T1-weighted MRI. Note the flow void artifacts within the mass-like cavernous transformation (arrowhead).

FIG 43-29  Contrast-enhanced MDCT acquired during the portal venous

FIG 43-30  Contrast-enhanced MDCT acquired during the portal venous

contrast phase shows thrombosis of the right posterior portal venous branch in this 41-year-old man.

contrast phase shows thrombosis of one of the distal portal venous branches off the right anterior portal vein following cholecystectomy in this 43-year-old woman.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases Two possible pathophysiologic mechanisms have been proposed. According to the first theory, portal venous air can be found after breakdown of the mucosal lining of bowel in concert with increased barometric and fluid pressure in the bowel lumen. In this scenario the gas seen in pneumatosis intestinalis rises through the venous intestinal arcades and mesenteric veins cranially to reach the portal confluence. Within the liver the gas is transported to the small peripheral branches of the subsegmental portal veins by hepatofugal portal venous blood flow. The second theory is based on the presence of gas-forming bacteria in the intestinal venous and portal venous systems. Cranial ascension of the intravascular gas ceases in the most cranially located peripheral branches of the subsegmental portal veins. Portal venous air, usually visualized as branching lucencies or susceptibility artifacts located within 2 cm of the liver capsule on MDCT and MR, has to be distinguished from biliary gas. Gas within the biliary tree is usually found within the central portions of the liver, more than 2 cm distant from the liver capsule because of the hepatocentric flow of bile. A wide variety of underlying diseases have been identified as presenting with portal venous air; it has been described in clinical scenarios of gastric ulcers, ulcerative colitis, Crohn’s disease, a complication of endoscopic procedures, intestinal abscesses and tumors, following umbilical vein catheterization, diverticulitis, intraabdominal sepsis, pneumatosis intestinalis, hemorrhagic pancreatitis, superior mesenteric artery thrombosis with subsequent intestinal ischemia, intestinal volvulus, necrotizing gastritis, colonic fistulas, enteric necrosis and radiation-induced vasculitis. Most frequently, portal venous gas is caused by mesenteric vascular occlusion and subsequent bowel necrosis, and therefore it is considered an ominous sign with an overall high mortality of up to 75%. For the remaining wide spectrum of possible underlying diseases, however, the mortality drops to 39%.79,103,118,154,179 Detection of portal venous gas on MDCT imaging usually does not pose diagnostic difficulties. On hepatic MRI, however, the air-induced susceptibility artifacts have to be distinguished from vascular flow voids and possible intrahepatic calcifications. T1-weighted sequences acquired with different TEs will help improve diagnostic precision because the size of air-induced susceptibility artifacts increases with longer TEs (Figs. 43-31 to 43-34).

FIG 43-31  Contrast-enhanced MDCT acquired during the portal venous contrast phase shows air within the main portal vein (arrow) in this 67-year-old woman.

1291

Budd-Chiari Syndrome.  The classic Budd-Chiari syndrome was described in 1845 as a combination of portal venous hypertension and hepatomegaly due to an obstruction of hepatic venous drainage. The initiating factor of Budd-Chiari syndrome is an obstruction of venous outflow from the sinusoidal bed of the liver, with subsequent development of slowly increasing portal venous hypertension. Only in rare cases does this lead to fulminant liver failure and hepatic encephalopathy. Budd-Chiari syndrome type I involves occlusion of the IVC, with or without secondary occlusion of the hepatic veins; type II consists of occlusion of the major hepatic veins; and type III is venoocclusive disease of the liver, or a progressive thrombotic occlusion of small centrilobular veins. Stasis of the hepatic venous outflow may be related to a membranous obstruction of the IVC, constrictive pericarditis, atrial myxoma, cardiac insufficiency, Hodgkin’s disease, cirrhosis, hepatocellular carcinoma, hematoma or abscess, and adrenal or renal carcinomas. Other contributing factors that are not specific to hepatic diseases include hypercoagulability and vascular injury as identified in Virchow’s triad.

FIG 43-32  Contrast-enhanced MDCT acquired during the portal venous contrast phase shows air within the main portal vein (arrow) as well as gastric intramural air (arrowheads) in this 65-year-old man.

FIG 43-33  Contrast-enhanced MDCT acquired during the portal venous contrast phase shows air within the medial and lateral branches of the left portal vein (arrow) as well as gastric intramural air (arrowheads) in this 55-year-old man.

1292

PART II  CT and MR Imaging of the Whole Body

The role of MDCT and MRI of the liver in the specific settings  of suspected confirmed or treated Budd-Chiari syndrome includes identification of vascular anatomy, detection of intrahepatic or extrahepatic masses, parenchymal evaluation, and assessment of shunt patencies in cavoatrial, mesocaval, and mesoatrial vascular bypasses. Cross-sectional imaging modalities are able to quantify any reduction in vascular caliber of the hepatic veins or detect missing connections to the IVC or their complete absence. In addition they allow depiction of newly formed but less competent intrahepatic collateral veins. Alteration of venous drainage ultimately leads to peripheral atrophy of hepatic parenchyma and an atypical parenchymal enhancement. Compensatory hypertrophy of both central parenchymal segments and the caudate lobe—which have improved venous drainage and different vascular anatomy, respectively—can be observed. In patients with polycythemia vera or paroxysmal nocturnal hemoglobinuria as a

reason for their hypercoagulable state, the findings of diffusely increased red bone marrow/splenomegaly or increased renal parenchymal density/decreased intensity on T1-weighted sequences, respectively, are typical cross-sectional imaging characteristics. Invasive treatment options include redirection of either hepatic venous or portal venous blood in an effort to decrease liver congestion or portal hypertension, respectively. Cavoatrial shunts redirect blood flow directly into the right atrium in clinical scenarios of solitary occlusion of the IVC. Mesocaval shunts lower the pressure in the portal venous system by bypassing the solitarily occluded hepatic vein, connecting the superior mesenteric vein to the IVC. Finally, mesoatrial shunts bypass thrombosed or occluded hepatic veins and the IVC to redirect blood flow from the superior mesenteric vein to the right atrium. MDCT allows direct visualization of the thrombi that can occlude or narrow the shunts, whereas hepatic MRI flow-void analysis and phase-contrast angiography techniques allow depiction and quantification of flow, including detection of flow reversal8,96,128,130,140,185,191,213,217 (Figs. 43-35 and 43-36).

Hepatic Venoocclusive Disease.  Thrombotic and subsequent

FIG 43-34  Contrast-enhanced MDCT acquired during the portal venous contrast phase shows air within one of the venous branches of the superior mesenteric vein (arrows) in a clinical scenario of pneumatosis intestinalis in a 79-year-old man.

A

fibrotic obstruction of submillimeter hepatic veins initially leads to congestion of the sinusoids, with resultant hepatomegaly. Eventually fibrosis of the liver will occur. Medications and treatments associated with bone marrow transplantation (e.g., azathioprine, pyrrolizidine alkaloids, cysteamine, whole-body radiation) are known to induce hepatic venoocclusive disease. Distinguishing hepatic venoocclusive disease from Budd-Chiari syndrome types I and II is a key point because in venoocclusive disease the IVC and main hepatic veins are patent, but the terminal hepatic venules and sinusoids are partially or entirely destroyed. In comparison to Budd-Chiari syndrome, hepatic venoocclusive disease is associated with early-stage liver failure with jaundice and thrombocytopenia. DCE MRI may indicate increased arterial perfusion of the affected hepatic segments as well as a prolonged liver transit time of the contrast agent. However, confirmation of the diagnosis should be attempted with direct histopathologic examination or direct wedge hepatic venography.24,192,203 In the context of the hepatic venoocclusive disease spectrum, chronic graft-versus-host liver disease and liver allograft rejection should be mentioned. Owing to a substantial overlap in imaging 

B FIG 43-35  Contrast-enhanced MDCT acquired during the portal venous contrast phase in a 77-year-old man with acute Budd-Chiari syndrome (A) in combination with acute thrombosis of the portal vein (arrow in B).

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

1293

B

C FIG 43-36  Chronic Budd-Chiari syndrome in a 65-year-old woman as seen on T2-weighted (A), precontrast T1-weighted (B), and contrast-enhanced fat-saturated T1-weighted (C) hepatic MRI acquired during the portal venous contrast phase.

findings, both clinical history and pathology are of paramount importance for making a differential diagnosis among these entities196 (Fig. 43-37).

Liver Disease in Congestive Heart Failure.  Hepatic congestion due to chronic failure of the right heart leads to widening and splaying of the central hepatic veins and overall hepatomegaly. The combination of decreased venous blood flow, increased blood pressure, and hypoxemia leads to diffuse cellular necrosis. Long-standing diffuse cellular necrosis induces diffuse hepatic fibrosis with a micronodular cirrhotic pattern known as nutmeg liver. Macroscopically, hepatic MDCT and MRI are able to depict cirrhotic changes as a nonspecific imaging finding, but microscopically the hepatic architecture remains intact77,100,132,144 (Fig. 43-38).

Hepatic Arterial Obstruction.  In comparison to portal venous or hepatic venous obstruction or occlusion, significant stenosis or occlusion of the hepatic arteries is a much rarer entity, commonly associated

FIG 43-37  T2-weighted hepatic MRI shows hepatic venoocclusive disease following treatment with pyrrolizidine alkaloids in this 61-yearold woman. Note the mild hepatomegaly.

1294

PART II  CT and MR Imaging of the Whole Body

A

B FIG 43-38  Contrast-enhanced MDCT shows a hemodynamic-relevant right cardiac mass (arrows in A) leading to enlargement of the suprahepatic inferior vena cava, as well as hepatic congestion (B), in this 59-year-old man.

with liver transplantation. The arterial anastomosis of the donor liver to the native vessel trunk can initially be significantly narrowed and subsequently thrombose because of the changes in hemodynamic flow. In patients without prior liver transplantation, an embolic occlusion is the most common cause of hepatic arterial compromise. The hepatic artery constitutes approximately 25% of the blood flow to the liver, supplying mainly the endothelial lining of the intrahepatic bile ducts and the fibrous hepatic structures. Contrast-enhanced MDCT and MRI show a diffusely diminished enhancement of the hepatic parenchyma in cases of severe hepatic artery stenosis or occlusion. High-resolution MDCT and MRI of the porta hepatis can identify the exact location of the culprit intraluminal thrombus or stenosis20,22,85,127,202 (Fig. 43-39).

Intrahepatic Arteriovenous Fistulas—Intrahepatic Shunts.  Intrahepatic arteriovenous fistulas can occur as vascular connections between hepatic arteries and hepatic veins, as well as between hepatic arteries and the portal venous system. A diffuse intrahepatic distribution is seen with hereditary hemorrhagic telangiectasia, also known as Osler’s disease; solitary occurrences of arteriovenous fistulas are most often related to liver biopsies or surgical procedures. In Osler’s disease an enlarged hepatic artery is usually seen in combination with tortuous vasculature in the porta hepatis. Functionally the increased arterial perfusion of the liver tissue leads to a secondary hypertrophy. Severe cases of Osler’s disease can subsequently lead to right-sided cardiac failure. In solitary iatrogenic postprocedural intrahepatic arteriovenous fistulas, there is significant dilatation of the draining vein. In the case of a vascular connection between hepatic arteries and the portal venous system, the resultant increased portal venous pressure leads to signs of portal venous hypertension and oftentimes reversed flow in the affected portal venous branch. Both MDCT and DCE MRI are capable of depicting the early venous drainage occurring during the hepatic arterial phase and can help estimate the functional significance of the fistula15,58,84,168,227 (Figs. 43-40 and 43-41).

FIG 43-39  Contrast-enhanced fat-saturated T1-weighted MRI in a 41-year-old man status post liver transplant shows occlusive thrombus formation within the hepatic artery at the level of the vascular anastomosis (arrow) and subsequent development of left hepatic cholangitis.

Liver Infarction.  Hepatic infarction is rare, in large part because of the dual-source blood supply provided by the hepatic artery and portal vein. Significant infarcts of hepatic parenchyma are therefore only possible if both sources of hepatic blood supply are significantly compromised, as in acute shock trauma and hypercoagulability and in clinical scenarios such as preeclampsia or postpartum patients with HELLP syndrome (hemolytic anemia, elevated liver enzymes, low platelets). A particular variant of hepatic ischemia is pseudoinfarction of the liver, also referred to as a Zahn infarct, consisting of an area of congestion with parenchymal atrophy but without necrosis, usually due to obstruction of a branch of the portal vein. Zahn infarcts are unique in that there is collateral congestion of liver sinusoids that does not include areas of anoxia seen in most infarcts. Fibrotic tissue may develop in the area of the infarct.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

1295

The typical MDCT appearance of liver infarction is a peripheral wedge-shaped area of low attenuation on contrast-enhanced MDCT. Larger infarcts may have a more geographic distribution. At MRI, lesions are typically hypointense on T1-weighted images but hyperintense on T2-weighted images. Although the diagnosis of liver infarction is evident in most cases, it may be confused with focal fatty infiltration, abscess, or even a tumor. It is of note that hepatic infarction does not demonstrate either mass effect or increased contrast enhancement. Also, vessels coursing through the lesion, although typically seen in focal fatty infiltration, are usually not seen in hepatic infarction62,75,109,113,187,230 (Figs. 43-42 to 43-44).

Liver Trauma.  In acceleration/deceleration traumas such as motor

FIG 43-40  Contrast-enhanced fat-saturated T1-weighted hepatic MRI shows an intrahepatic fistula (arrow) between a right posterior portal venous branch and a hepatic venous branch draining into the inferior vena cava (arrowheads) in this 54-year-old woman.

A

vehicle accidents and direct penetrating traumas of the abdomen, the liver and spleen are the most frequently injured organs. Traumatization of the liver and spleen is often associated with an overall unstable hemodynamic situation. The right hepatic lobe is most commonly affected, with subcapsular hematoma formation and direct parenchymal compression, parenchymal contusions or lacerations, or rupture of the biliary tree with subsequent formation of bilomas.

B FIG 43-41  Contrast-enhanced fat-saturated T1-weighted hepatic MRI in a 44-year-old man shows intrahepatic fistulas between the right hepatic arterial branches (arrow in A) and portal venous branches, with early contrast enhancement of the portal vein (arrows in B) during the arterial contrast phase.

A

B FIG 43-42  Acute infarction of the liver parenchyma in a 78-year-old man as seen on contrast-enhanced MDCT (arrows in A) and T2-weighted hepatic MRI (arrows in B).

1296

PART II  CT and MR Imaging of the Whole Body

A

B FIG 43-43  Chronic infarction and subsequent atrophy of the liver parenchyma in a 69-year-old man as seen on contrast-enhanced MDCT (arrows in A) and T2-weighted hepatic MRI (arrows in B).

A

B FIG 43-44  This 72-year-old woman had multiorgan failure and infarctions involving liver, spleen, and bilateral kidneys.

Hemorrhage into the biliary tree can be identified due to the typical imaging appearance of blood on MDCT and MRI. Multiplanar reformations of high-resolution, cross-sectional imaging data sets are essential to detect complex parenchymal injuries, in particular parenchymal tears originating in the region of the bare area can be difficult to detect, as they follow the hepatic venous distribution7,30,104,110,148 (Fig. 43-45).

Inflammatory Parenchymal Diseases and Cirrhosis

Viral Hepatitis.  The histopathologic patterns of hepatitides of various causes show cellular alterations with differing penetration  of periportal hepatocellular necrosis, Kupffer cell mobilization, and portal inflammation of plasma cells, depending on the underlying infectious, toxic, or autoimmune etiology. Inflammatory conditions  of the hepatic parenchyma can be self-limiting, progress to scarring of liver segments, or end in an overall cirrhotic state. The acute form of hepatitis lasts less than 6 months; chronic hepatitis is defined as any inflammatory condition of the hepatic parenchyma that does not show signs of regression for time periods longer than 6 months.

Viral hepatitides are still the most common causes of liver disease worldwide. Six hepatotropic viruses, in particular picornavirus (hepatitis A), hepadnavirus (hepatitis B), flavivirus (hepatitis C), hepatitis-E virus, togavirus (hepatitis F), and GBV-C (hepatitis G) have been identified. The acute forms of viral hepatitis usually cause hepatomegaly with an edematous liver capsule and distinct necrotic areas that result in irregularities of the liver contour. In its fulminant presentation, acute hepatitis causes extensive necrosis that leads to shrinkage of the entire organ and significant loss of parenchymal volume. Necrotic areas transform into scars during the phase of restitution, which can take on an imaging appearance similar to cirrhotic liver segments. Chronic hepatitis is a necroinflammatory condition dominated by lymphocytes. Long-standing edema of the hepatic parenchyma progresses to heterogeneous patterns of inflammation and hepatocyte regeneration, with subsequent fibrosis and possible cirrhosis of the liver. Ascites and splenomegaly can accompany any state of chronic hepatitis.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

1297

B

C FIG 43-45  Contrast-enhanced MDCT during the portal venous contrast phase shows hepatic trauma with differing degrees of parenchymal injury. A, A 19-year-old man with laceration of the posterior liver contour and accompanying subcapsular and extracapsular hematoma. B, A 22-year-old woman with contusion and laceration of the hepatic parenchyma. C, A 29-year-old woman with massive traumatic injury to the hepatic and splenic parenchyma.

Acute and chronic hepatitides are primarily diagnosed by clinical and serologic examinations; hepatic MDCT and MRI can, however, contribute to the diagnosis. In particular these imaging modalities can be used to determine the extent and severity of compromised hepatic parenchyma. MDCT imaging of acute and fulminant courses of hepatitis is capable of identifying general hepatomegaly combined with edema of the liver capsule. Furthermore, NCE MDCT imaging can show heterogeneous attenuation of the hepatic parenchyma, with multifocal or confluent regions in the periphery. Overall hepatic parenchymal attenuation is equal to or less than that of the spleen. Contrast-enhanced MDCT can demonstrate irregular perfusion with heterogeneous  attenuation throughout the hepatic parenchyma; nodular areas of regeneration have higher attenuation values compared to surrounding edematous or partially necrotic parenchymal segments. The highly variable MDCT imaging patterns of chronic hepatitis are discussed later with cirrhosis. Hepatic MRI of acute viral hepatitis, analogously to MDCT imaging, shows general hepatomegaly combined with edema of the liver capsule. Additionally, regions of high signal intensity can be seen on T2-weighted image sequences surrounding the portal venous branches. A similar contrast enhancement pattern to that observed on MDCT imaging can be detected on DCE and multiphase contrastenhanced MRI (Fig. 43-46). During the course of chronic hepatitis, significant lymph node enlargement can occur, in particularly in the

porta hepatis. As opposed to nodal enlargement caused by metastases, the portal vein remains patent and maintains its shape. MRI patterns of chronic hepatitis are discussed later (see “Cirrhosis—Portal Venous Hypertension”73,87,97,99,129,145,157,214 (Figs. 43-47 and 43-48).

Radiation-Induced Hepatitis.  Given the large size and anatomic location of the liver, it is frequently affected by radiation therapy of extrahepatic malignancies. Whereas the histopathologic and MDCT/ MRI findings in radiation-induced hepatitis at least partially overlap with histopathologic analysis and imaging of hepatitides of other causes, the anatomic location and margination of radiation-induced hepatitis within the hepatic parenchyma can help correctly identify this entity. In particular there is often a straight linear margin to the parenchymal abnormality that corresponds to the radiation portal. Reduced portal venous flow in areas with radiation injury has been observed as well.59,212,228

Hepatic Sarcoidosis.  Noncaseating epithelioid granulomas scattered throughout the liver is the most common hepatic presentation of multisystem sarcoidosis, also known as Boeck’s sarcoid. Although involvement of abdominal organs is frequent, its abdominal manifestations are usually documented after the initial diagnosis has been made based on thoracic findings, namely bihilar lymphadenopathy. The underlying pathophysiology that leads to formation of noncaseating epithelioid granulomas remains unknown, but an association with

1298

PART II  CT and MR Imaging of the Whole Body

A

B FIG 43-46  Acute viral hepatitis in a 39-year-old man as seen on contrast-enhanced MDCT acquired during the arterial phase (A) and the portal venous phase (B). Note the heterogeneous contrast enhancement in the edematous enlarged liver.

A

B FIG 43-47  Precontrast OP (A) and contrast-enhanced fat-saturated (B) T1-weighted hepatic MRI in a 54-yearold man with chronic viral hepatitis.

primary sclerosing cholangitis and coexistence with a wide range of other autoimmune diseases has been revealed. The most common sites of involvement include the lymph nodes, lungs, liver, spleen, eyes, skin, salivary glands, nervous system, bones, and heart. Hepatic noncaseating epithelioid granulomas present without specific features and range from submillimeter to 1 to 2 cm in size. Larger granulomas contain greater amounts of reticulin and are surrounded by fibrotic hepatic parenchyma, reflecting a more vigorous immunologic response to the unknown causative agent. Acute complications include hepatic failure due to intrahepatic cholestasis and portal hypertension; a multisystem sarcoidosis mortality of 10% has been described and is due to cor pulmonale, lung fibrosis, and liver cirrhosis.56,74,94,171,204

NCE MDCT imaging shows multiple intrahepatic and intrasplenic hypodense lesions accompanied by nonspecific hepatosplenomegaly. These lesions become rapidly isodense to liver parenchyma after intravenous contrast administration (Figs. 43-49 and 43-50). Hepatic MRI demonstrates the noncaseating epithelioid granulomas as hypointense intrahepatic and intrasplenic nodules on T1- and T2-weighted imaging sequences. Advanced-stage hepatic sarcoidosis may present as liver cirrhosis (Fig. 43-51); MDCT and MRI patterns of cirrhosis are discussed in “Cirrhosis—Portal Venous Hypertension.”

Hepatic Candidiasis.  A fungal infection with Candida albicans leads to development of multiple well-defined, rounded intrahepatic

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

FIG 43-48  Portal venous phase imaging shows heterogeneously enhancing hepatic parenchyma as well as hepatomegaly in a 49-yearold man with acute viral hepatitis.

1299

FIG 43-50  Contrast-enhanced MDCT shows multisystemic sarcoidosis of the liver and spleen in this 75-year-old man. Note the subtle hepatic lesions (arrow) in contrast to the more clearly seen splenic lesion.

FIG 43-51  Contrast-enhanced MRI in a 64-year-old woman with multisystemic sarcoidosis of the liver and spleen.

FIG 43-49  Contrast-enhanced MDCT shows multisystemic sarcoidosis of the liver and spleen in this 55-year-old man. Note disseminated distribution of innumerable hepatic lesions.

microabscesses. Hepatic candidiasis is a common fungal infection in immunocompromised patients. Candida tropicalis infections are predominantly seen in tropical countries. Hepatic candidiasis results from intrahepatic spread of intestinal fungal seeding via portal venous  and hepatic venous circulation of Candida strains. Complications of hepatic candidiasis arise in cases of microabscess rupture and/or accompanying cholangitis. MDCT imaging will reveal innumerable hypodense liver lesions of less than 1 cm in diameter and without contrast enhancement. The intrahepatic periportal fields may show a slight increase in attenuation due to fibrosis. During the healing phase, scattered calcifications may be observed (Figs. 43-52 to 43-54). Hepatic MRI will show innumerable small hypointense lesions on T1-weighted imaging sequences, whereas on T2, the same lesions will

be hyperintense. No contrast enhancement is seen within the lesions. T1-weighted gradient echo sequences have shown greater sensitivity in detecting the small hypointense lesions compared to spin echo acquisition schemes.172,173,219

Hepatic Schistosomiasis.  Two Schistosoma species, Schistosoma japonicum and Schistosoma mansoni, are held primarily responsible for significant hepatic disease. For both species, infections results from water contact that is sustained by free-swimming cercariae burrowing themselves into the skin.142 The organisms thus enter the bloodstream, where they reside for a while to carry out the maturation process for then migrating in the mesenteric-portal venous system.142 Pathologic and imaging findings of hepatic disease by Schistosoma species can be explained by schistosome egg embolization to the terminal venous branches of the portal system, ensuing in chronic granulomatous reaction, extensive fibrosis, and calcification.142

1300

PART II  CT and MR Imaging of the Whole Body

FIG 43-52  Contrast-enhanced MDCT in a 41-year-old woman with microbiologically proven candidiasis of the liver. Note splenic involvement.

FIG 43-54  Contrast-enhanced MDCT in a 55-year-old woman with microbiologically proven candidiasis. Note disseminated distribution of innumerable submillimeter hepatic and splenic lesions.

FIG 43-55  Contrast-enhanced fat-saturated T1-weighted hepatic MRI FIG 43-53  Contrast-enhanced MDCT in a 49-year-old man with microbiologically proven candidiasis of the liver. Note disseminated distribution of innumerable submillimeter hepatic lesions.

Peliosis Hepatis.  Microscopically, peliosis hepatis shows dilatation of the sinusoids in concert with multiple blood-filled lacunar spaces. The pathogenesis remains unknown; general outflow obstruction  of the sinusoids with breakdown of the sinusoidal barriers has been suggested as an initiating pathomechanism. Agents associated with development of peliosis hepatis are anabolic steroids, corticosteroids, tamoxifen, diethylstilbestrol, azathioprine, and oral contraceptives. Macroscopically, hepatomegaly may be present. Hepatic MDCT shows multiple small hypodense lesions less than 1 cm in diameter on noncontrast images; centrifugal contrast uptake during the portal venous imaging phase can be seen. Thrombosed venous branches remain hypodense during parenchymal phase imaging. Hepatic MRI will reveal small hypointense lesions on T1-weighted imaging sequences, whereas on T2 the same lesions appear hyperintense. Analogous centrifugal contrast uptake is noted. Larger lesions

shows peliosis hepatis (arrows) that developed in this 59-year-old woman while she was being treated with tamoxifen.

can show imaging characteristics of blood at different phases of organization53,83,194,209,219 (Figs. 43-55 and 43-56).

Cirrhosis—Portal Venous Hypertension Cirrhosis as multisystem disease.  Liver cirrhosis represents an irreversible hepatic fibrosis bridging the portal tracts. Conditions leading to cirrhosis include chronic hepatitis, alcohol abuse, hemochromatosis, drug toxicity, chronic biliary obstruction as seen in primary sclerosing cholangitis and primary biliary cirrhosis, as well as inflammatory parenchymal processes. If the inciting cause is unknown the process is referred to as cryptogenic cirrhosis. Cirrhosis affects the contour of the organ and the size of the lobes and segments and may cause benign as well as malignant focal lesions to arise within the liver. Processes associated with cirrhosis include ascites and portal hypertension. Splenomegaly in turn is the most common finding of portal hypertension. Another common entity associated with portal venous hypertension are collateral vessels or varices; these can be seen as enhancing tortuous vessels in the paraesophageal and gastric regions, porta

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

FIG 43-56  Contrast-enhanced fat-saturated T1-weighted hepatic MRI shows peliosis hepatis (arrows) that developed in this 51-year-old woman while she was being treated with steroids.

hepatis, peritoneal cavity, retroperitoneum, and even coursing through the liver via paraumbilical collaterals. Furthermore, mesenteric, omental, and retroperitoneal edema can be demonstrated in clinical scenarios of cirrhosis. The edema is usually mild but may be severe in approximately 25% of patients. Cirrhosis markedly distorts the hepatic parenchyma, producing either a smooth, nodular, or lobular appearance of the hepatic capsule in the majority of cases. Margins that are either smooth or deformed by multiple small nodules are typically seen in micronodular cirrhosis. These micronodules underlying the surface of the liver are often not visualized at MDCT or MRI. Coarse nodularity of the hepatic margin is the result of macronodular cirrhosis. Tremendous overlap in the frequency of smooth and nodular hepatic morphology among different causes of cirrhosis usually prevents distinguishing the etiology based on imaging appearance. However, a lobular liver margin is more frequently seen in patients with primary sclerosing cholangitis. Adenopathy can be seen in the porta hepatis and peripancreatic region in some cases of cirrhosis. Occasionally these nodes can be large and bulky, mimicking neoplastic involvement. These lymph nodes are most prominent in cases of primary biliary cirrhosis. Fibrosis can appear as a confluent mass replacing the cirrhotic regenerating parenchyma in approximately 30% of patients. Focal confluent fibrosis may be seen in all types of cirrhosis but occurs most commonly in primary sclerosing cholangitis and least commonly in primary biliary cirrhosis. Essential imaging features include focal retraction of the liver capsule over the area of fibrosis and incomplete involvement of liver segments, with accompanying atrophy and characteristic location in segments IVa, IVb, VII, and VIII. Focal confluent hepatic fibrosis appears as a wedge-shaped structure radiating from the porta hepatis. On MDCT and MRI, cirrhosis causes a patchy intrahepatic attenuation or signal intensity in about 65% and a diffusely heterogeneous appearance in another third of patients. Approximately 25% of endstage cirrhotic livers appear normal in size and shape; another third may be diffusely atrophic. The majority of end-stage cirrhotic livers exhibit a combination of segmental hypertrophy and atrophy, with the predominant finding being hypertrophy of liver segments I-III. Focal atrophy is most common in segments IV-VIII and may range from mild to complete involution of the affected segments. Occasionally the segments II/III and VI/VII may be atrophic; this is most often seen in

1301

patients with primary sclerosing cholangitis. Another rare finding  in patients with primary sclerosing cholangitis is a higher attenuation in segment I, mimicking a pseudotumor in the caudate lobe.106 Approximately 25% of all patients with end-stage cirrhosis show a heterogeneous appearance in all parts of the liver on unenhanced MDCT and T2-weighted MRI.129 Histopathologic analysis reveals three causes for this appearance: diffuse fibrosis, fat deposition, and iron deposition. On MDCT imaging, four different patterns of diffuse fibrosis may be seen: (1) patchy and poorly defined regions of low attenuation on unenhanced MDCT, (2) thin perilobular bands of low attenuation on unenhanced MDCT, (3) low-attenuation thick fibrotic bridging surrounding regenerative nodules, and (4) diffuse fibrosis causing a highattenuation perivascular cuffing. The latter two are most commonly observed in patients with primary biliary cirrhosis. Owing to the reduction in portal venous flow in patients with end-stage cirrhosis, contrast-enhanced MDCT during the portal venous phase is usually not helpful in demonstrating fibrosis. On unenhanced MDCT, focal confluent fibrosis most often appears as a lesion with low attenuation. Following contrast medium administration, approximately 50% of the lesions become isoattenuating with the surrounding parenchyma. Contrast enhancement of the fibrotic area may rarely simulate a neoplastic process, which typically requires percutaneous biopsy (Figs. 43-57 to 43-59). On T2-weighted MRI, fibrosis appears as an area of high signal intensity. On unenhanced T1-weighted imaging, fibrosis usually cannot be visualized. Following contrast medium administration, however, in cases with thin perilobular bands or perivenous cuffing, the fibrous tissue may show mild enhancement on T1. Focal confluent fibrosis appears bright on T2 and hypointense on T1, with a characteristic enhancement pattern most pronounced on equilibrium phase imaging37,88,90,151,152,182,208,211,218,220 (see Fig. 43-59). Cirrhosis and its hepatospecific presentation.  Another characteristic sign of liver cirrhosis is development of a spectrum of focal liver lesions ranging from benign regenerative nodules to malignant hepatocellular carcinoma. Regenerative nodules are present in all cirrhotic livers; however, these nodules are visualized in only 25% of unenhanced MDCT scans and approximately 50% of MRIs. These nodules usually present as high-attenuation lesions on unenhanced MDCT, owing to their iron and/or glycogen content or simply because they are surrounded by lower-attenuation fibrosis. On portal venous phase contrast-enhanced imaging, these nodules are rarely seen. On MRI these focal lesions can appear hypo-, iso-, or hyperintense on T1-weighted images; an isointense appearance is most common and a hyperintense appearance least common. On T2-weighted images, they typically appear as hypointense nodules; the iron they contain causes local magnetic field inhomogeneities with decreased liver signal intensity due to shortening of T2 relaxation time (Fig. 43-60). Dysplastic nodules are premalignant lesions that represent an intermediate step in the pathway of hepatocarcinogenesis in cirrhotic livers. These lesions are present in approximately 10% of all end-stage cirrhotic livers. Dysplastic nodules appear slightly hyperdense or isodense on unenhanced MDCT imaging and isodense or hypodense on all contrast-enhanced scans (e.g., arterial phase, portal phase, and delayed scans). On MRI, dysplastic nodules appear hyperintense on T1 and iso- to hypointense on T2 and show varying amounts of contrast uptake213 (Figs. 43-61 and 43-62). The incidence of hepatocellular carcinoma in end-stage cirrhosis is high, ranging from 3% to 11%, most often in cases of cirrhosis caused by hepatitis. Carcinogenesis of hepatocellular tumors changes the Text continued on p. 1306

1302

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 43-57  Contrast-enhanced MDCT shows the spectrum of appearances of cirrhosis. A, A 51-year-old woman with prominence of the caudate lobe (arrow) and mild cirrhotic nodularity imaged during the arterial contrast phase. B, A 55-year-old woman with splenomegaly (arrows) as sign of portal hypertension, adenopathy (arrowheads), and irregularities of the hepatic contours imaged during the portal venous contrast phase. C, A 61-year-old man with increased nodularity of the hepatic parenchyma imaged during the portal venous contrast phase. D, A 63-year-old man with hepatic atrophy and ascites imaged during a delayed contrast phase.

A

B FIG 43-58  A and B, Contrast-enhanced MDCT during the parenchymal contrast phase shows confluent fibrosis (arrows) within a cirrhotic liver in this 52-year-old woman.

A

B FIG 43-59  Confluent fibrosis (arrows) within the cirrhotic liver of a 69-year-old man as seen on precontrast OP T1w (A) and T2-weighted (B) hepatic MRI.

A

B

C

D

FIG 43-60  Cirrhosis and multiple regenerating nodules dissemi-

E

nated throughout the liver of a 78-year-old man as seen on contrast-enhanced MDCT during the portal venous contrast phase (arrows in A and B) and on hepatic MRI using T2-weighted (C), T1-weighted precontrast (D), and T1-weighted fat-saturated postcontrast (E) sequences. Note the subtle delineation of the regenerating nodules in postcontrast MDCT and T1, the hypointense imaging characteristics on T2, and the lack of contrast uptake on T1 fat-saturated postcontrast sequences.

1304

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F FIG 43-61  Cirrhosis and a prominent dysplastic nodule in the right hepatic lobe (arrows) of a 79-year-old man as seen on arterial (A) and portal venous (B) MDCT imaging as well as on T2-weighted (C), precontrast T1-weighted (D), fat-saturated early arterial T1-weighted (E), and portal venous T1-weighted (F) hepatic MRI. Note the hypodense to isodense imaging characteristics on MDCT and the hyperintense visualization on precontrast T1-weighted hepatic MRI.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases

A

B

C

D

E FIG 43-62  Cirrhosis and dysplastic nodule in the right hepatic lobe (arrows) of a 51-year-old man. T2-weighted (A) hepatic MRI, shows an additional bright central focus suspicious for malignant transformation into a hepatocellular carcinoma, known as nodule-in-nodule sign. Precontrast T1-weighted image (B) shows a hyperintense lesion. Fat-saturated early arterial T1-weighted (C), late arterial (D), and delayed imaging (E) show an initial focal contrast uptake (arrows).

1305

1306

PART II  CT and MR Imaging of the Whole Body

A

B FIG 43-63  Biopsy-proven hepatocellular carcinoma in the right hepatic lobe (arrows) of a 59-year-old man as seen on arterial (A) and delayed (B) MDCT imaging. Note subtle demarcation of the pseudocapsule on the delayed image.

hemodynamic properties and vascular supply of the tumor nodules, resulting in different enhancement patterns in early and advanced hepatocellular carcinomas. In early hepatocellular carcinoma the combination of reduced hepatic arterial perfusion and preserved portal venous flow results in low attenuation on MDCT arteriography and isoattenuation on portal venous MDCT. In advanced hepatocellular carcinoma the combination of neoplastic arterial perfusion induced by mechanisms of angioneogenesis and obliteration of portal veins results in high attenuation on MDCT arteriography and low attenuation on portal venous MDCT. Employing hepatic MRI, various appearances may be seen on T1-weighted imaging, ranging from hypointense to hyperintense lesions. However, almost all hepatocellular carcinomas appear bright on T2, allowing for differentiation of dysplastic nodules from hepatocellular carcinomas. Small hepatocellular carcinomas may be seen on early DCE MRI only, on which they appear as hyperintense enhancing nodules. Another characteristic finding in hepatocellular carcinoma is the presence of a pseudocapsule consisting of connective fibrous tissue that appears hypointense on T1 and T2 sequences. Following administration of contrast agents, this pseudocapsule usually shows early enhancement but becomes more prominent on portal venous phase imaging82,107,119,177,205 (Fig. 43-63; see Fig. 43-62).

REFERENCES 1. Abenoza P, Sibley RK: Farber’s disease: A fine structural study. Ultrastruct Pathol 11:397–403, 1987. 2. Adams LA, Lymp JF, St SJ, et al: The natural history of nonalcoholic fatty liver disease: A population-based cohort study. Gastroenterology 129:113–121, 2005. 3. Adams LA, Zein CO, Angulo P, et al: A pilot trial of pentoxifylline   in nonalcoholic steatohepatitis. Am J Gastroenterol 99:2365–2368, 2004. 4. Andrade ZA, Peixoto E, Guerret S, et al: Hepatic connective tissue changes in hepatosplenic schistosomiasis. Hum Pathol 23:566–573, 1992. 5. Angulo P: Nonalcoholic fatty liver disease. N Engl J Med 346:1221– 1231, 2002. 6. Arai K, Matsui O, Takashima T, et al: Focal spared areas in fatty liver caused by regional decreased portal flow. AJR Am J Roentgenol 151:300–302, 1988.

7. Arkovitz MS, Liang H, Pachter HL, et al: Complete biliary avulsion from blunt compression injury. J Pediatr Surg 34:1559–1562, 1999. 8. Attwell A, Ludkowski M, Nash R, et al: Treatment of Budd-Chiari syndrome in a liver transplant unit, the role of transjugular intrahepatic porto-systemic shunt and liver transplantation. Aliment Pharmacol Ther 20:867–873, 2004. 9. Auh YH, Rosen A, Rubenstein WA, et al: CT of the papillary process of the caudate lobe of the liver. AJR Am J Roentgenol 142:535–538, 1984. 10. Auh YH, Rubenstein WA, Zirinsky K, et al: Accessory fissures of the liver: CT and sonographic appearance. AJR Am J Roentgenol 143:565– 572, 1984. 11. Awai K, Inoue M, Yagyu Y, et al: Moderate versus high concentration of contrast material for aortic and hepatic enhancement and tumor-toliver contrast at multi-detector row CT. Radiology 233:682–688, 2004. 12. Bassett ML, Halliday JW, Ferris RA, et al: Diagnosis of hemochromatosis in young subjects: Predictive accuracy of biochemical screening tests. Gastroenterology 87:628–633, 1984. 13. Bazalova M, Carrier JF, Beaulieu L, et al: Dual-energy CT-based material extraction for tissue segmentation in Monte Carlo dose calculations. Phys Med Biol 53:2439–2456, 2008. 14. Beck M: Mucopolysaccharidosis. Nosology—Clinical aspects— Therapeutic approaches. Monatsschr Kinderheilkd 139:120–127, 1991. 15. Begbie ME, Wallace GM, Shovlin CL: Hereditary haemorrhagic telangiectasia (Osler-Weber-Rendu syndrome): A view from the 21st century. Postgrad Med J 79:18–24, 2003. 16. Bhutta AT, Van SH, Schexnayder SM: Reye’s syndrome: Down but not out. South Med J 96:43–45, 2003. 17. Bismuth H: Surgical anatomy of the liver. Recent Results Cancer Res 100:179–184, 1986. 18. Bismuth H: Anatomy of the liver and hepatectomy techniques. Ann Chir 52:61–63, 1998. 19. Boll DT, Merkle EM: Diffuse liver disease: Strategies for hepatic CT and MR imaging. Radiographics 29:1591–1614, 2009. 20. Boraschi P, Donati F, Cossu MC, et al: Multi-detector computed tomography angiography of the hepatic artery in liver transplant recipients. Acta Radiol 46:455–461, 2005. 21. Botero AC, Strasberg SM: Division of the left hemiliver in man— Segments, sectors, or sections. Liver Transpl Surg 4:226–231, 1998. 22. Boyvat F, Aytekin C, Karakayali H, et al: Stent placement in pediatric patients with hepatic artery stenosis or thrombosis after liver transplantation. Transplant Proc 38:3656–3660, 2006. 23. Broome DR, Girguis MS, Baron PW, et al: Gadodiamide-associated nephrogenic systemic fibrosis: Why radiologists should be concerned. AJR Am J Roentgenol 188:586–592, 2007.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases 24. Brown BP, bu-Yousef M, Farner R, et al: Doppler sonography: A noninvasive method for evaluation of hepatic venoocclusive disease. AJR Am J Roentgenol 154:721–724, 1990. 25. Brown DW, Henkelman RM, Poon PY, et al: Nuclear magnetic resonance study of iron overload in liver tissue. Magn Reson Imaging 3:275–282, 1985. 26. Browne M, Somers G, Savoia H, et al: Wolman’s disease in an infant.   Br J Haematol 122:522, 2003. 27. Brunt EM: Nonalcoholic steatohepatitis. Semin Liver Dis 24:3–20, 2004. 28. Bryant KG, Horns KM, Longo N, et al: A primer on newborn screening. Adv Neonatal Care 4:306–317, 2004. 29. Burton BK: Inborn errors of metabolism: The clinical diagnosis in early infancy. Pediatrics 79:359–369, 1987. 30. Carrillo EH, Spain DA, Miller FB, et al: Intrahepatic vascular clamping in complex hepatic vein injuries. J Trauma 43:131–133, 1997. 31. Casteels-Van DM, Wouters C, Van GC, et al: Reye’s syndrome revisited. Outdated concept of Reye’s syndrome was used. BMJ 324:546, 2002. 32. Cecchin E, De MS, Querin F, et al: Efficacy of hepatic computed tomography to detect iron overload in chronic hemodialysis. Kidney Int 37:943–950, 1990. 33. Chen TS, Chen PS: The myth of Prometheus and the liver. J R Soc Med 87:754–755, 1994. 34. Chezmar JL, Nelson RC, Malko JA, et al: Hepatic iron overload: Diagnosis and quantification by noninvasive imaging. Gastrointest Radiol 15:27–31, 1990. 35. Cho A, Okazumi S, Makino H, et al: Anterior fissure of the right liver—The third door of the liver. J Hepatobiliary Pancreat Surg 11:390–396, 2004. 36. Cho A, Okazumi S, Takayama W, et al: Anatomy of the right anterosuperior area (segment 8) of the liver: Evaluation with helical CT during arterial portography. Radiology 214:491–495, 2000. 37. Chopra S, Dodd GD, III, Chintapalli KN, et al: Mesenteric, omental, and retroperitoneal edema in cirrhosis: Frequency and spectrum of CT findings. Radiology 211:737–742, 1999. 38. Chow EL, Cherry JD, Harrison R, et al: Reassessing Reye syndrome. Arch Pediatr Adolesc Med 157:1241–1242, 2003. 39. Clark JM, Diehl AM: Nonalcoholic fatty liver disease: An underrecognized cause of cryptogenic cirrhosis. JAMA 289:3000–3004, 2003. 40. Coffin CM, Diche T, Mahfouz A, et al: Benign and malignant hepatocellular tumors: Evaluation of tumoral enhancement after mangafodipir trisodium injection on MR imaging. Eur Radiol 9:444–449, 1999. 41. Day DL, Letourneau JG, Allan BT, et al: Hepatic regenerating nodules in hereditary tyrosinemia. AJR Am J Roentgenol 149:391–393, 1987. 42. De MM, De SG, Laconi A, et al: Gold storage in the liver: Appearance on CT scans. Radiology 159:355–356, 1986. 43. de Bazelaire CM, Duhamel GD, Rofsky NM, et al: MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: Preliminary results. Radiology 230:652–659, 2004. 44. De Gaetano AM, Lafortune M, Patriquin H, et al: Cavernous transformation of the portal vein: Patterns of intrahepatic and splanchnic collateral circulation detected with Doppler sonography. AJR Am J Roentgenol 165:1151–1155, 1995. 45. Devaney K, Goodman ZD, Epstein MS, et al: Hepatic sarcoidosis. Clinicopathologic features in 100 patients. Am J Surg Pathol 17:1272– 1280, 1993. 46. Diehl AM: Nonalcoholic steatohepatitis. Semin Liver Dis 19:221–229, 1999. 47. Dodd GD, III: An American’s guide to Couinaud’s numbering system. AJR Am J Roentgenol 161:574–575, 1993. 48. Dwyer A, Doppman JL, Adams AJ, et al: Influence of glycogen on liver density: Computed tomography from a metabolic perspective. J Comput Assist Tomogr 7:70–73, 1983. 49. Edelman RR, Zhao B, Liu C, et al: MR angiography and dynamic flow evaluation of the portal venous system. AJR Am J Roentgenol 153:755– 760, 1989.

1307

50. Edwards CQ: Early detection of hereditary hemochromatosis. Ann Intern Med 101:707–708, 1984. 51. Ellaway CJ, Wilcken B, Christodoulou J: Clinical approach to inborn errors of metabolism presenting in the newborn period. J Paediatr Child Health 38:511–517, 2002. 52. Eser G, Karabacakoglu A, Karakose S, et al: Mangafodipir trisodiumenhanced magnetic resonance imaging for evaluation of pancreatic mass and mass-like lesions. World J Gastroenterol 12:1603–1606, 2006. 53. Ferrozzi F, Tognini G, Zuccoli G, et al: Peliosis hepatis with pseudotumoral and hemorrhagic evolution: CT and MR findings. Abdom Imaging 26:197–199, 2001. 54. Flohr TG, McCollough CH, Bruder H, et al: First performance evaluation of a dual-source CT (DSCT) system. Eur Radiol 16:256–268, 2006. 55. Foley WD, Mallisee TA, Hohenwalter MD, et al: Multiphase hepatic CT with a multirow detector CT scanner. AJR Am J Roentgenol 175:679– 685, 2000. 56. Folz SJ, Johnson CD, Swensen SJ: Abdominal manifestations of sarcoidosis in CT studies. J Comput Assist Tomogr 19:573–579, 1995. 57. Fretz CJ, Elizondo G, Weissleder R, et al: Superparamagnetic iron oxide-enhanced MR imaging: Pulse sequence optimization for detection of liver cancer. Radiology 172:393–397, 1989. 58. Garcia-Tsao G, Korzenik JR, Young L, et al: Liver disease in patients with hereditary hemorrhagic telangiectasia. N Engl J Med 343:931–936, 2000. 59. Garra BS, Shawker TH, Chang R, et al: The ultrasound appearance of radiation-induced hepatic injury. Correlation with computed tomography and magnetic resonance imaging. J Ultrasound Med 7:605–609, 1988. 60. Garver WS, Francis GA, Jelinek D, et al: The National Niemann-Pick C1 disease database: Report of clinical features and health problems. Am J Med Genet A 143A:1204–1211, 2007. 61. Gazelle GS, Haaga JR: Hepatic neoplasms: surgically relevant segmental anatomy and imaging techniques. AJR Am J Roentgenol 158:1015–1018, 1992. 62. Giovine S, Pinto A, Crispano S, et al: Retrospective study of 23 cases of hepatic infarction: CT findings and pathological correlations. Radiol Med (Torino) 111:11–21, 2006. 63. Goldberg MA, Hahn PF, Saini S, et al: Value of T1 and T2 relaxation times from echoplanar MR imaging in the characterization of focal hepatic lesions. AJR Am J Roentgenol 160:1011–1017, 1993. 64. Goldman IS, Winkler ML, Raper SE, et al: Increased hepatic density and phospholipidosis due to amiodarone. AJR Am J Roentgenol 144:541–546, 1985. 65. Goshima S, Kanematsu M, Kondo H, et al: MDCT of the liver and hypervascular hepatocellular carcinomas: Optimizing scan delays for bolus-tracking techniques of hepatic arterial and portal venous phases. AJR Am J Roentgenol 187:W25–W32, 2006. 66. Griswold M, Jakob P, Heidemann R, et al: Push-button PPA Reconstructions: GRAPPA. Proc Intl Soc Mag Reson Med 1:8, 2001. 67. Guiney MJ, Kruskal JB, Sosna J, et al: Multi-detector row CT of relevant vascular anatomy of the surgical plane in split-liver transplantation. Radiology 229:401–407, 2003. 68. Guyader D, Gandon Y, Deugnier Y, et al: Evaluation of computed tomography in the assessment of liver iron overload. A study of 46 cases of idiopathic hemochromatosis. Gastroenterology 97:737–743, 1989. 69. Hamer OW, Aguirre DA, Casola G, et al: Imaging features of perivascular fatty infiltration of the liver: Initial observations. Radiology 237:159–169, 2005. 70. Hata F, Hirata K, Murakami G, et al: Identification of segments VI and VII of the liver based on the ramification patterns of the intrahepatic portal and hepatic veins. Clin Anat 12:229–244, 1999. 71. Hernandez-Guerra M, Garcia-Pagan JC, Bosch J: Increased hepatic resistance: A new target in the pharmacologic therapy of portal hypertension. J Clin Gastroenterol 39:S131–S137, 2005.

1308

PART II  CT and MR Imaging of the Whole Body

72. Hines J, Katz DS, Goffner L, et al: Fat collection related to the intrahepatic inferior vena cava on CT. AJR Am J Roentgenol 172:409– 411, 1999. 73. Hinterberger W: Differential chronic hepatitis diagnosis. Wien Med Wochenschr 150:476–480, 2000. 74. Hoeffel C, Bokemeyer C, Hoeffel JC, et al: CT hepatic and splenic appearances with sarcoidosis. Eur J Radiol 23:94–96, 1996. 75. Holbert BL, Baron RL, Dodd GD, III: Hepatic infarction caused by arterial insufficiency: Spectrum and evolution of CT findings. AJR Am J Roentgenol 166:815–820, 1996. 76. Holland AE, Hecht EM, Hahn WY, et al: Importance of small (< or = 20-mm) enhancing lesions seen only during the hepatic arterial phase at MR imaging of the cirrhotic liver: Evaluation and comparison with whole explanted liver. Radiology 237:938–944, 2005. 77. Holley HC, Koslin DB, Berland LL, et al: Inhomogeneous enhancement of liver parenchyma secondary to passive congestion: Contrastenhanced CT. Radiology 170:795–800, 1989. 78. Hori M, Murakami T, Kim T, et al: Single breath-hold T2-weighted MR imaging of the liver: Value of single-shot fast spin-echo and multishot spin-echo echoplanar imaging. AJR Am J Roentgenol 174:1423–1431, 2000. 79. Hou SK, Chern CH, How CK, et al: Hepatic portal venous gas: Clinical significance of computed tomography findings. Am J Emerg Med 22:214–218, 2004. 80. Huang J, Raman SS, Vuong N, et al: Utility of breath-hold fast-recovery fast spin-echo T2 versus respiratory-triggered fast spin-echo T2 in clinical hepatic imaging. AJR Am J Roentgenol 184:842–846, 2005. 81. Hussain HK, Chenevert TL, Londy FJ, et al: Hepatic fat fraction: MR imaging for quantitative measurement and display—Early experience. Radiology 237:1048–1055, 2005. 82. Hussain HK, Syed I, Nghiem HV, et al: T2-weighted MR imaging in the assessment of cirrhotic liver. Radiology 230:637–644, 2004. 83. Iannaccone R, Federle MP, Brancatelli G, et al: Peliosis hepatis: Spectrum of imaging findings. AJR Am J Roentgenol 187:W43–W52, 2006. 84. Ianora AA, Memeo M, Sabba C, et al: Hereditary hemorrhagic telangiectasia: Multi-detector row helical CT assessment of hepatic involvement. Radiology 230:250–259, 2004. 85. Ishigami K, Stolpen AH, Al-Kass FM, et al: Diagnostic value of gadolinium-enhanced 3D magnetic resonance angiography in patients with suspected hepatic arterial complications after liver transplantation. J Comput Assist Tomogr 29:464–471, 2005. 86. Isselbacher KJ: Galactose metabolism and galactosemia. Am J Med 26:715–723, 1959. 87. Itai Y: CT findings of fulminant hepatitis: Terminology and distribution of massive necrosis. Radiology 200:872, 1996. 88. Itai Y, Kurosaki Y, Saida Y, et al: CT and MRI in detection of intrahepatic portosystemic shunts in patients with liver cirrhosis. J Comput Assist Tomogr 18:768–773, 1994. 89. Itai Y, Ohtomo K, Furui S, et al: Lobar intensity differences of the liver on MR imaging. J Comput Assist Tomogr 10:236–241, 1986. 90. Johnson KJ, Olliff JF, Olliff SP: The presence and significance of lymphadenopathy detected by CT in primary sclerosing cholangitis. Br J Radiol 71:1279–1282, 1998. 91. Johnston DL, Rice L, Vick GW, III, et al: Assessment of tissue iron overload by nuclear magnetic resonance imaging. Am J Med 87:40–47, 1989. 92. Kamel IR, Kruskal JB, Pomfret EA, et al: Impact of multidetector CT on donor selection and surgical planning before living adult right lobe liver transplantation. AJR Am J Roentgenol 176:193–200, 2001. 93. Kardorff R, Mugashe C, Gabone RM, et al: Diagnostic value of connective tissue metabolites in Schistosoma mansoni related liver disease. Acta Trop 73:153–164, 1999. 94. Kataoka M, Nakata Y, Hiramatsu J, et al: Hepatic and splenic sarcoidosis evaluated by multiple imaging modalities. Intern Med 37:449–453, 1998. 95. Katayama M, Masui T, Kobayashi S, et al: Fat-suppressed T2-weighted MRI of the liver: Comparison of respiratory-triggered fast spin-echo,

96.

97.

98.

99. 100.

101.

102.

103.

104.

105.

106.

107.

108.

109. 110. 111.

112.

113.

114.

115. 116.

breath-hold single-shot fast spin-echo, and breath-hold fast-recovery fast spin-echo sequences. J Magn Reson Imaging 14:439–449, 2001. Kavanagh PM, Roberts J, Gibney R, et al: Acute Budd-Chiari syndrome with liver failure: The experience of a policy of initial interventional radiological treatment using transjugular intrahepatic portosystemic shunt. J Gastroenterol Hepatol 19:1135–1139, 2004. Kawamoto S, Soyer PA, Fishman EK, et al: Nonneoplastic liver disease: Evaluation with CT and MR imaging. Radiographics 18:827–848, 1998. Kawashima A, Suehiro S, Murayama S, et al: Focal fatty infiltration of the liver mimicking a tumor: Sonographic and CT features. J Comput Assist Tomogr 10:329–331, 1986. Kelly DA: Acute liver failure. Indian J Pediatr 66:S104–S109, 1999. Kim BS, Kim TK, Kim JS, et al: Hepatic venous congestion after living donor liver transplantation with right lobe graft: Two-phase CT findings. Radiology 232:173–180, 2004. Kim YK, Kim CS, Kwak HS, et al: Three-dimensional dynamic liver MR imaging using sensitivity encoding for detection of hepatocellular carcinomas: Comparison with superparamagnetic iron oxide-enhanced MR imaging. J Magn Reson Imaging 20:826–837, 2004. Kim YK, Kim CS, Lee YH, et al: Comparison of superparamagnetic iron oxide-enhanced and gadobenate dimeglumine-enhanced dynamic MRI for detection of small hepatocellular carcinomas. AJR Am J Roentgenol 182:1217–1223, 2004. Kinoshita H, Shinozaki M, Tanimura H, et al: Clinical features and management of hepatic portal venous gas: Four case reports and cumulative review of the literature. Arch Surg 136:1410–1414, 2001. Knudson MM, Maull KI: Nonoperative management of solid organ injuries. Past, present, and future. Surg Clin North Am 79:1357–1371, 1999. Kobayashi K, Matsui O, Yoshikawa J, et al: Right hepatic arterial   supply to the posterior aspect of segment IV of the liver: Analysis   by CT during hepatic arteriography. Abdom Imaging 24:591–593, 1999. Kolnagou A, Economides C, Eracleous E, et al: Low serum ferritin levels are misleading for detecting cardiac iron overload and increase the risk of cardiomyopathy in thalassemia patients. The importance of cardiac iron overload monitoring using magnetic resonance imaging T2 and T2*. Hemoglobin 30:219–227, 2006. Krinsky GA, Lee VS, Theise ND: Focal lesions in the cirrhotic liver: High resolution ex vivo MRI with pathologic correlation. J Comput Assist Tomogr 24:189–196, 2000. Kroncke TJ, Taupitz M, Kivelitz D, et al: Multifocal nodular fatty infiltration of the liver mimicking metastatic disease on CT: Imaging findings and diagnosis using MR imaging. Eur Radiol 10:1095–1100, 2000. Kronthal AJ, Fishman EK, Kuhlman JE, et al: Hepatic infarction in preeclampsia. Radiology 177:726–728, 1990. Lee SK, Carrillo EH: Advances and changes in the management of liver injuries. Am Surg 73:201–206, 2007. Lee VS, Lavelle MT, Rofsky NM, et al: Hepatic MR imaging with a dynamic contrast-enhanced isotropic volumetric interpolated breath-hold examination: Feasibility, reproducibility, and technical quality. Radiology 215:365–372, 2000. Lee VS, Morgan GR, Teperman LW, et al: MR imaging as the sole preoperative imaging modality for right hepatectomy: A prospective study of living adult-to-adult liver donor candidates. AJR Am J Roentgenol 176:1475–1482, 2001. Lev-Toaff AS, Friedman AC, Cohen LM, et al: Hepatic infarcts: New observations by CT and sonography. AJR Am J Roentgenol 149:87–90, 1987. Levenson H, Greensite F, Hoefs J, et al: Fatty infiltration of the liver: Quantification with phase-contrast MR imaging at 1.5 T vs biopsy. AJR Am J Roentgenol 156:307–312, 1991. Levy HM, Newhouse JH: MR imaging of portal vein thrombosis. AJR Am J Roentgenol 151:283–286, 1988. Ley TJ, Griffith P, Nienhuis AW: Transfusion haemosiderosis and chelation therapy. Clin Haematol 11:437–464, 1982.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases 117. Li Y, Ayata G, Baker SP, et al: Cholangitis: A histologic classification based on patterns of injury in liver biopsies. Pathol Res Pract 201:565– 572, 2005. 118. Liebman PR, Patten MT, Manny J, et al: Hepatic-portal venous gas in adults: Etiology, pathophysiology and clinical significance. Ann Surg 187:281–287, 1978. 119. Lim JH, Choi BI: Dysplastic nodules in liver cirrhosis: Imaging. Abdom Imaging 27:117–128, 2002. 120. Lin SP, Brown JJ: MR contrast agents: Physical and pharmacologic basics. J Magn Reson Imaging 25:884–899, 2007. 121. Loh YH, Dunn GD: Diffuse fatty infiltration of the liver: Pitfalls   in computed tomography diagnosis. Australas Radiol 41:383–386, 1997. 122. Lorigan JG, Charnsangavej C, Carrasco CH, et al: Atrophy with compensatory hypertrophy of the liver in hepatic neoplasms: Radiographic findings. AJR Am J Roentgenol 150:1291–1295, 1988. 123. Luboldt W, Straub J, Seemann M, et al: Effective contrast use in CT angiography and dual-phase hepatic CT performed with a subsecond scanner. Invest Radiol 34:751–760, 1999. 124. Marckmann P, Skov L, Rossen K, et al: Nephrogenic systemic fibrosis: Suspected causative role of gadodiamide used for contrast-enhanced magnetic resonance imaging. J Am Soc Nephrol 17:2359–2362, 2006. 125. Marin D, Boll DT, Mileto A, et al: State of the art: Dual-energy CT of the abdomen. Radiology 271:327–342, 2014. 126. Markos J, Veronese ME, Nicholson MR, et al: Value of hepatic computerized tomographic scanning during amiodarone therapy. Am J Cardiol 56:89–92, 1985. 127. Marubashi S, Dono K, Asaoka T, et al: Risk factors for graft dysfunction after adult-to-adult living donor liver transplantation. Transplant Proc 38:1407–1410, 2006. 128. Mathieu D, Vasile N, Menu Y, et al: Budd-Chiari syndrome: Dynamic CT. Radiology 165:409–413, 1987. 129. Matsui O, Kadoya M, Kameyama T, et al: Adenomatous hyperplastic nodules in the cirrhotic liver: Differentiation from hepatocellular carcinoma with MR imaging. Radiology 173:123–126, 1989. 130. Matsui O, Kadoya M, Takashima T, et al: Intrahepatic periportal abnormal intensity on MR images: An indication of various hepatobiliary diseases. Radiology 171:335–338, 1989. 131. Matsui O, Kadoya M, Yoshikawa J, et al: Posterior aspect of hepatic segment IV: Patterns of portal venule branching at helical CT during arterial portography. Radiology 205:159–162, 1997. 132. Mauro MA, Stackhouse DJ, Parker LA, et al: Computed tomography of hepatic venous hypertension: The reticulated-mosaic pattern. Gastrointest Radiol 15:35–38, 1990. 133. McCullough AJ, Fleming CR, Thistle JL, et al: Diagnosis of Wilson’s disease presenting as fulminant hepatic failure. Gastroenterology 84:161–167, 1983. 134. Mergo PJ, Ros PR, Buetow PC, et al: Diffuse disease of the liver: Radiologic-pathologic correlation. Radiographics 14:1291–1307, 1994. 135. Merkle EM, Gilkeson RC: Remnants of fetal circulation: Appearance on MDCT in adults. AJR Am J Roentgenol 185:541–549, 2005. 136. Merkle EM, Nelson RC: Dual gradient-echo in-phase and opposedphase hepatic MR imaging: A useful tool for evaluating more than fatty infiltration or fatty sparing. Radiographics 26:1409–1418, 2006. 137. Mistry PK, Smith SJ, Ali M, et al: Genetic diagnosis of Gaucher’s disease. Lancet 339:889–892, 1992. 138. Mitchell DG: Chemical shift magnetic resonance imaging: Applications in the abdomen and pelvis. Top Magn Reson Imaging 4:46–63, 1992. 139. Mitchell DG: Focal manifestations of diffuse liver disease at MR imaging. Radiology 185:1–11, 1992. 140. Mitchell MC, Boitnott JK, Kaufman S, et al: Budd-Chiari syndrome: Etiology, diagnosis and management. Medicine (Baltimore) 61:199–218, 1982. 141. Miyayama S, Matsui O, Taki K, et al: Arterial blood supply to the posterior aspect of segment IV of the liver from the caudate branch: Demonstration at CT after iodized oil injection. Radiology 237:1110– 1114, 2005.

1309

142. Monzawa S, Uchiyama G, Ohtomo K, et al: Schistosomiasis japonica of the liver: Contrast-enhanced CT findings in 113 patients. AJR Am J Roentgenol 161:323–327, 1993. 143. Moses SW: Pathophysiology and dietary treatment of the glycogen storage diseases. J Pediatr Gastroenterol Nutr 11:155–174, 1990. 144. Moulton JS, Miller BL, Dodd GD, III, et al: Passive hepatic congestion in heart failure: CT abnormalities. AJR Am J Roentgenol 151:939–942, 1988. 145. Murakami T, Baron RL, Peterson MS: Liver necrosis and regeneration after fulminant hepatitis: Pathologic correlation with CT and MR findings. Radiology 198:239–242, 1996. 146. Nakanuma Y, Kouda W, Harada K, et al: Hepatic sarcoidosis with vanishing bile duct syndrome, cirrhosis, and portal phlebosclerosis. Report of an autopsy case. J Clin Gastroenterol 32:181–184, 2001. 147. Nasu K, Kuroki Y, Nawano S, et al: Hepatic metastases: Diffusionweighted sensitivity-encoding versus SPIO-enhanced MR imaging. Radiology 239:122–130, 2006. 148. Nathan M, Gates J, Ferzoco SJ: Hepatic duct confluence injury in blunt abdominal trauma: Case report and synopsis on management. Surg Laparosc Endosc Percutan Tech 13:350–352, 2003. 149. Nelson RC, Lovett KE, Chezmar JL, et al: Comparison of pulsed Doppler sonography and angiography in patients with portal hypertension. AJR Am J Roentgenol 149:77–81, 1987. 150. Ortale JR, Borges Keiralla LC: Anatomy of the portal branches and the hepatic veins in the caudate lobe of the liver. Surg Radiol Anat 26:384–391, 2004. 151. Outwater E, Kaplan MM, Bankoff MS: Lymphadenopathy in primary biliary cirrhosis: CT observations. Radiology 171:731–733, 1989. 152. Outwater E, Kaplan MM, Bankoff MS: Lymphadenopathy in sclerosing cholangitis: Pitfall in the diagnosis of malignant biliary obstruction. Gastrointest Radiol 17:157–160, 1992. 153. Patel BN, Kumbla RA, Berland LL, et al: Material density hepatic steatosis quantification on intravenous contrast-enhanced rapid kilovolt (peak)-switching single-source dual-energy computed tomography. J Comput Assist Tomogr 37:904–910, 2013. 154. Peloponissios N, Halkic N, Pugnale M, et al: Hepatic portal gas in adults: Review of the literature and presentation of a consecutive series of 11 cases. Arch Surg 138:1367–1370, 2003. 155. Pilleul F, Chave G, Dumortier J, et al: Fatty infiltration of the liver. Detection and grading using dual T1 gradient echo sequences on clinical MR system. Gastroenterol Clin Biol 29:1143–1147, 2005. 156. Platt JF, Francis IR, Ellis JH, et al: Liver metastases: Early detection based on abnormal contrast material enhancement at dual-phase helical CT. Radiology 205:49–53, 1997. 157. Proujansky R, Vinton N: Acute hepatitis. Adolesc Med 6:437–446, 1995. 158. Qayyum A, Goh JS, Kakar S, et al: Accuracy of liver fat quantification at MR imaging: Comparison of out-of-phase gradient-echo and fat-saturated fast spin-echo techniques—Initial experience. Radiology 237:507–511, 2005. 159. Rao BK, Brodell GK, Haaga JR, et al: Visceral CT findings associated with Thorotrast. J Comput Assist Tomogr 10:57–61, 1986. 160. Raptopoulos V, Karellas A, Bernstein J, et al: Value of dual-energy CT in differentiating focal fatty infiltration of the liver from low-density masses. AJR Am J Roentgenol 157:721–725, 1991. 161. Reimer P, Saini S, Hahn PF, et al: Techniques for high-resolution echo-planar MR imaging of the pancreas. Radiology 182:175–179, 1992. 162. Reye RD, Morgan G, Baral J: Encephalopathy and fatty degeneration of the viscera. A disease entity in childhood. Lancet 91:749–752, 1963. 163. Reynolds WJ, Wanless IR: Nodular regenerative hyperplasia of the liver in a patient with rheumatoid vasculitis: A morphometric study suggesting a role for hepatic arteritis in the pathogenesis. J Rheumatol 11:838–842, 1984. 164. Rofsky NM, Lee VS, Laub G, et al: Abdominal MR imaging with a volumetric interpolated breath-hold examination. Radiology 212:876– 884, 1999. 165. Rofsky NM, Weinreb JC, Ambrosino MM, et al: Comparison between in-phase and opposed-phase T1-weighted breath-hold FLASH

1310

166.

167. 168.

169.

170.

171. 172.

173. 174.

175.

176. 177.

178. 179. 180. 181. 182. 183.

184.

185. 186. 187.

188.

189. 190.

PART II  CT and MR Imaging of the Whole Body

sequences for hepatic imaging. J Comput Assist Tomogr 20:230–235, 1996. Roudot-Thoraval F, Halphen M, Larde D, et al: Evaluation of liver iron content by computed tomography: Its value in the follow-up of treatment in patients with idiopathic hemochromatosis. Hepatology 3:974–979, 1983. Ryman BE: The glycogen storage diseases. J Clin Pathol Suppl (R Coll Pathol) 8:106–121, 1974. Sadick H, Sadick M, Gotte K, et al: Hereditary hemorrhagic telangiectasia: An update on clinical manifestations and diagnostic measures. Wien Klin Wochenschr 118:72–80, 2006. Sadowski EA, Bennett LK, Chan MR, et al: Nephrogenic systemic fibrosis: Risk factors and incidence estimation. Radiology 243:148–157, 2007. Sahani DV, O’Malley ME, Bhat S, et al: Contrast-enhanced MRI of the liver with mangafodipir trisodium: Imaging technique and results.   J Comput Assist Tomogr 26:216–222, 2002. Sakai T, Maeda M, Takabatake M, et al: MR imaging of hepatosplenic sarcoidosis. Radiat Med 13:39–41, 1995. Sallah S, Semelka R, Kelekis N, et al: Diagnosis and monitoring response to treatment of hepatosplenic candidiasis in patients with acute leukemia using magnetic resonance imaging. Acta Haematol 100:77–81, 1998. Sallah S, Semelka RC, Wehbie R, et al: Hepatosplenic candidiasis in patients with acute leukaemia. Br J Haematol 106:697–701, 1999. Salo S, Alanen A, Leino R, et al: The effect of haemosiderosis and blood transfusions on the T2 relaxation time and 1/T2 relaxation rate of liver tissue. Br J Radiol 75:24–27, 2002. Schafer AI, Cheron RG, Dluhy R, et al: Clinical consequences of acquired transfusional iron overload in adults. N Engl J Med 304:319– 324, 1981. Schafer DF, Sorrell MF: Hepatocellular carcinoma. Lancet 353:1253– 1257, 1999. Scharitzer M, Schima W, Schober E, et al: Characterization of hepatocellular tumors: Value of mangafodipir-enhanced magnetic resonance imaging. J Comput Assist Tomogr 29:181–190, 2005. Schilsky ML, Sternlieb I: Overcoming obstacles to the diagnosis of Wilson’s disease. Gastroenterology 113:350–353, 1997. Schindera ST, Triller J, Vock P, et al: Detection of hepatic portal venous gas: Its clinical impact and outcome. Emerg Radiol 12:164–170, 2006. Sherlock S, Dooley J: Diseases of the liver and the biliary system, ed 10, 1997. Siegelman ES: MR imaging of diffuse liver disease. Hepatic fat and iron. Magn Reson Imaging Clin N Am 5:347–365, 1997. Silverman PM, Baker ME, Cooper C, et al: CT appearance of diffuse mesenteric edema. J Comput Assist Tomogr 10:67–70, 1986. Silverman PM, Patt RH, Garra BS, et al: MR imaging of the portal venous system: Value of gradient-echo imaging as an adjunct to spin-echo imaging. AJR Am J Roentgenol 157:297–302, 1991. Silverman PM, Ram PC, Korobkin M: CT appearance of abdominal thorotrast deposition and Thorotrast-induced angiosarcoma of the liver. J Comput Assist Tomogr 7:655–658, 1983. Slakey DP, Klein AS, Venbrux AC, et al: Budd-Chiari syndrome: Current management options. Ann Surg 233:522–527, 2001. Smetana HF, Olen E: Hereditary galactose disease. Am J Clin Pathol 38:3–25, 1962. Smith GS, Birnbaum BA, Jacobs JE: Hepatic infarction secondary to arterial insufficiency in native livers: CT findings in 10 patients. Radiology 208:223–229, 1998. Smith PA, Klein AS, Heath DG, et al: Dual-phase spiral CT angiography with volumetric 3D rendering for preoperative liver transplant evaluation: Preliminary observations. J Comput Assist Tomogr 22:868–874, 1998. Sodickson DK: Tailored SMASH image reconstructions for robust in vivo parallel MR imaging. Magn Reson Med 44:243–251, 2000. Sodickson DK, Manning WJ: Simultaneous acquisition of spatial harmonics (SMASH): Fast imaging with radiofrequency coil arrays. Magn Reson Med 38:591–603, 1997.

191. Stanley P: Budd-Chiari syndrome. Radiology 170:625–627, 1989. 192. Stark DD, Goldberg HI, Moss AA, et al: Chronic liver disease: Evaluation by magnetic resonance. Radiology 150:149–151, 1984. 193. Stark DD, Moseley ME, Bacon BR, et al: Magnetic resonance imaging and spectroscopy of hepatic iron overload. Radiology 154:137–142, 1985. 194. Steinke K, Terraciano L, Wiesner W: Unusual cross-sectional imaging findings in hepatic peliosis. Eur Radiol 13:1916–1919, 2003. 195. Stephens DH, Sheedy PF, Hattery RR, et al: Computed tomography of the liver. AJR Am J Roentgenol 128:579–590, 1977. 196. Strasser SI, Shulman HM, Flowers ME, et al: Chronic graft-versus-host disease of the liver: Presentation as an acute hepatitis. Hepatology 32:1265–1271, 2000. 197. Sultana S, Awai K, Nakayama Y, et al: Hypervascular hepatocellular carcinomas: Bolus tracking with a 40-detector CT scanner to time arterial phase imaging. Radiology 243:140–147, 2007. 198. Swartz RD, Crofford LJ, Phan SH, et al: Nephrogenic fibrosing dermopathy: A novel cutaneous fibrosing disorder in patients with renal failure. Am J Med 114:563–572, 2003. 199. Tanaka K, Numata K, Okazaki H, et al: Diagnosis of portal vein thrombosis in patients with hepatocellular carcinoma: Efficacy of color Doppler sonography compared with angiography. AJR Am J Roentgenol 160:1279–1283, 1993. 200. Tang Y, Yamashita Y, Namimoto T, et al: Liver T2-weighted MR imaging: Comparison of fast and conventional half-Fourier single-shot turbo spin-echo, breath-hold turbo spin-echo, and respiratory-triggered turbo spin-echo sequences. Radiology 203:766–772, 1997. 201. Tanimoto A, Mukai M, Kuribayashi S: Evaluation of superparamagnetic iron oxide for MR imaging of liver injury: Proton relaxation mechanisms and optimal MR imaging parameters. Magn Reson Med Sci 5:89–98, 2006. 202. Tarhan NC, Firat A, Coskun M, et al: Diagnosis of complications in auxiliary heterotopic partial-liver transplant recipients: Spiral CT findings. Turk J Gastroenterol 13:192–197, 2002. 203. Teefey SA, Brink JA, Borson RA, et al: Diagnosis of venoocclusive disease of the liver after bone marrow transplantation: Value of duplex sonography. AJR Am J Roentgenol 164:1397–1401, 1995. 204. Thanos L, Zormpala A, Brountzos E, et al: Nodular hepatic and splenic sarcoidosis in a patient with normal chest radiograph. Eur J Radiol 41:10–11, 2002. 205. Theise ND: Macroregenerative (dysplastic) nodules and hepatocarcinogenesis: Theoretical and clinical considerations. Semin Liver Dis 15:360–371, 1995. 206. Tohma T, Cho A, Okazumi S, et al: Communicating arcade between the right and left hepatic arteries: Evaluation with CT and angiography during temporary balloon occlusion of the right or left hepatic artery. Radiology 237:361–365, 2005. 207. Torres WE, Gaylord GM, Whitmire L, et al: The correlation between MR and angiography in portal hypertension. AJR Am J Roentgenol 148:1109–1113, 1987. 208. Torres WE, Whitmire LF, Gedgaudas-McClees K, et al: Computed tomography of hepatic morphologic changes in cirrhosis of the liver. J Comput Assist Tomogr 10:47–50, 1986. 209. Toyoda S, Takeda K, Nakagawa T, et al: Magnetic resonance imaging of peliosis hepatis: A case report. Eur J Radiol 16:207–208, 1993. 210. Tsuji S, Choudary PV, Martin BM, et al: A mutation in the human glucocerebrosidase gene in neuronopathic Gaucher’s disease. N Engl J Med 316:570–575, 1987. 211. Tyrrel RT, Montemayor KA, Bernardino ME: CT density of mesenteric, retroperitoneal, and subcutaneous fat in cirrhotic patients: Comparison with control subjects. AJR Am J Roentgenol 155:73–75, 1990. 212. Unger EC, Lee JK, Weyman PJ: CT and MR imaging of radiation hepatitis. J Comput Assist Tomogr 11:264–268, 1987. 213. Valla DC: Prognosis in Budd Chiari syndrome after re-establishing hepatic venous drainage. Gut 55:761–763, 2006. 214. Valla D, Flejou JF, Lebrec D, et al: Portal hypertension and ascites in acute hepatitis: Clinical, hemodynamic and histological correlations. Hepatology 10:482–487, 1989.

CHAPTER 43  Liver: Normal Anatomy, Imaging Techniques, and Diffuse Diseases 215. van Leeuwen MS, Fernandez MA, van Es HW, et al: Variations in venous and segmental anatomy of the liver: Two- and threedimensional MR imaging in healthy volunteers. AJR Am J Roentgenol 162:1337–1345, 1994. 216. Van Lom KJ, Brown JJ, Perman WH, et al: Liver imaging at 1.5 tesla: Pulse sequence optimization based on improved measurement of tissue relaxation times. Magn Reson Imaging 9:165–171, 1991. 217. Van BB, Pringot J, Trigaux JP, et al: Hepatic heterogeneity on CT in Budd-Chiari syndrome: Correlation with regional disturbances in portal flow. Gastrointest Radiol 13:61–66, 1988. 218. Vignaux O, Legmann P, Coste J, et al: Cirrhotic liver enhancement on dual-phase helical CT: Comparison with noncirrhotic livers in 146 patients. AJR Am J Roentgenol 173:1193–1197, 1999. 219. Vignaux O, Legmann P, de Pinieux G, et al: Hemorrhagic necrosis due to peliosis hepatis: Imaging findings and pathological correlation. Eur Radiol 9:454–456, 1999. 220. Vitellas KM, Tzalonikou MT, Bennett WF, et al: Cirrhosis: Spectrum of findings on unenhanced and dynamic gadolinium-enhanced MR imaging. Abdom Imaging 26:601–615, 2001. 221. Vogl TJ, Steiner S, Hammerstingl R, et al: MRT of the liver in Wilson’s disease. Rofo 160:40–45, 1994. 222. Vogt FM, Antoch G, Hunold P, et al: Parallel acquisition techniques for accelerated volumetric interpolated breath-hold examination magnetic resonance imaging of the upper abdomen: Assessment of image quality and lesion conspicuity. J Magn Reson Imaging 21:376–382, 2005.

1311

224. Wang C: Mangafodipir trisodium (MnDPDP)-enhanced magnetic resonance imaging of the liver and pancreas. Acta Radiol Suppl 415:1–31, 1998. 225. Wang C, Ahlstrom H, Ekholm S, et al: Diagnostic efficacy of MnDPDP in MR imaging of the liver. A phase III multicentre study. Acta Radiol 38:643–649, 1997. 223. Wang B, Gao Z, Zou Q, et al: Quantitative diagnosis of fatty liver with dual-energy CT. An experimental study in rabbits. Acta Radiol 44:92–97, 2003. 226. Warzok R, Seidlitz G: Mucopolysaccharidoses. Genetics, clinical pathology, therapeutic regimes. Zentralbl Pathol 138:226–234, 1992. 227. Willinek WA, Hadizadeh D, von Falkenhausen M, et al: Magnetic resonance (MR) imaging and MR angiography for evaluation and follow-up of hepatic artery banding in patients with hepatic involvement of hereditary hemorrhagic telangiectasia. Abdom Imaging 31:694–700, 2006. 228. Yankelevitz DF, Knapp PH, Henschke CI, et al: MR appearance of radiation hepatitis. Clin Imaging 16:89–92, 1992. 229. Young ID, Segura J, Ford PM, et al: The pathogenesis of nodular regenerative hyperplasia of the liver associated with rheumatoid vasculitis. J Clin Gastroenterol 14:127–131, 1992. 230. Zissin R, Yaffe D, Fejgin M, et al: Hepatic infarction in preeclampsia as part of the HELLP syndrome: CT appearance. Abdom Imaging 24:594–596, 1999.

44  Liver: Focal Hepatic Mass Lesions Satoshi Kobayashi and Osamu Matsui

Focal hepatic mass lesions can be divided into: (1) pseudolesions or pseudotumors, (2) nontumorous mass lesions, including cysts and inflammatory masses, (3) benign tumors, and (4) primary and secondary malignant tumors. Ultrasonography plays a primary role in detection of focal hepatic mass lesions; for further characterization, computed tomography (CT) and magnetic resonance imaging (MRI) provide valuable information. This chapter discusses the CT and MRI features of hepatic mass lesions and their pathophysiology.

TYPES OF LESIONS

portion of the portal vein through the peribiliary vascular plexus, the obstructed segment shows early enhancement on dynamic CT and MRI (segmental staining) (Fig. 44-2). This segmental staining is often observed in HCC or other malignant tumors with portal vein invasion and makes it difficult to detect the underlying tumor itself.100,184 When Zahn’s infarction occurs, the obstructed area shows hypodensity on precontrast CT, hypointensity on T1-weighted images, and hyperintensity on T2-weighted images. In a fatty liver this area often demonstrates focal sparing, and the causative disease can be overlooked.61 Intrahepatic hepatic vein obstruction results in similar findings based on the segmental distribution of the hepatic veins.114

Pseudolesions and Pseudotumors

Pseudolesion and Pseudotumor Due to Third Inflow.  Several

Various kinds of pseudolesions (focal masslike findings seen only on imaging) or pseudotumors (focal masslike parenchymal changes) are seen in the liver. As a first step in the imaging diagnosis of focal hepatic mass lesions, exclusion of these entities is important. Pseudolesions and pseudotumors are often associated with localized intrahepatic portal flow abnormalities.

kinds of direct venous inflow into the liver parenchyma from outside the main portal trunk have been analyzed by anatomic and imaging studies.27,99,183,185,274 These are called third inflow. The areas receiving third inflow are well demonstrated by CT during arterial portography (CTAP) as portal perfusion defects in an otherwise normal liver.99,183,274 According to CTAP-based analysis, third inflow includes flow from an aberrant right gastric vein (or pancreaticoduodenogastric vein),183 cystic veins,185,340 veins of Sappey,137,229,259 and aberrant left gastric vein.294 These veins usually connect directly to the intrahepatic portal venules. The areas receiving third inflow often show pseudolesions or pseudotumors. To differentiate these entities from true hepatic mass lesions, visualization of these veins by color Doppler ultrasonography63,136 or dynamic CT is useful (Fig. 44-3). The cystic veins drain into the gallbladder fossa and communicate with the peripheral portion of the intrahepatic portal veins in segment IV or V.185,340 The drainage area often shows a tiny wedge-shaped area of early staining on dynamic CT and MRI because of earlier venous return relative to the surrounding liver (Fig. 44-4). On nonenhanced and postcontrast CT and MRI, no definite abnormality is demonstrated. This staining can be better visualized when blood flow is increased by gallbladder cancer or cystitis. This inflow through the cystic veins frequently causes focal sparing in a fatty liver.300 An aberrant right gastric vein accumulates blood from the gastric antrum, duodenum, and pancreas head; it is also referred to as the pancreaticoduodenogastric vein.37,183 The right gastric vein usually drains into the main portal trunk; however, it drains directly into the posterior aspect of segment IV of the liver in 6% to 14% of the general population and rarely into segment I, II, or III.182,183,274,335 The area drained by this vein occasionally demonstrates early enhancement on the arterial phase of dynamic CT and MRI (see Fig. 44-4) owing to relatively early venous return, focal sparing in a fatty liver (Fig. 44-5), focal fatty liver (Fig. 44-6), and hyperplastic change in a cirrhotic  liver13,63,104,119,180,182,183,306 (Fig. 44-7). On MRI this change can be seen in around 5% of cirrhotic patients.306 For purposes of differentiation,

Arterioportal Shunt.  In arterioportal (AP) shunt the hepatic arterial blood flows directly into the portal vein. AP communications are abundant through the peribiliary vascular plexus in the liver,41,42 and an AP shunt may form. However, as a result of biopsy or tumor, a direct fistula between the artery and portal vein can be created. On dynamic CT and MRI, a wedge-shaped early enhancement is seen in the arterial phase, with occasional visualization of the portal venules in the center of the lesion; this area becomes isodense in the equilibrium phase (Fig. 44-1). Coronal or sagittal multidetector CT images are occasionally valuable for demonstrating these features. Because AP shunt is usually not visualized on superparamagnetic  iron oxide (SPIO)-enhanced and hepatobiliary-phase gadoliniumethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)– enhanced MRI, these methods are effective in differentiating AP shunt from small hypervascular tumors or other true lesions.204,280 In hypervascular hepatocellular carcinoma (HCC), corona enhancement due to drainage flow from the tumor is seen in the portal or equilibrium phase with washout from the tumor (discussed later).308 This hemodynamic pattern is different from that in AP shunt.307,343 The area with AP shunt often shows focal sparing in a fatty liver and rarely focal iron deposition.108 AP shunt is often associated with HCC, cavernous hemangioma, and other tumors.

Pseudolesion Due to Portal Vein Obstruction.  Intrahepatic portal vein obstruction leads to various imaging findings in the obstructed segment that may result in its cause being obscured. Because of compensatory blood flow from the hepatic artery to the distal

1312

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

FIG 44-1  Arterioportal shunt. On dynamic CT scan (B) a wedge-shaped early enhancement is seen in the arterial phase, with visualization of the portal venule in the center of the lesion (arrow). However, there are no visible abnormal findings on precontrast (A) and equilibrium-phase (C) CT images.

A

B

C

FIG 44-2  Segmental staining caused by intrahepatic portal vein thrombosis. Although there are no attenuation differences in the liver on precontrast (A) and equilibrium-phase (C) CT scan, a wedge-shaped enhanced area is observed in segment VIII of the liver in the arterial phase (B) (arrow). A linear hypodense lesion that indicates a thrombus is observed in the portal vein (arrowhead).

1313

1314

A

PART II  CT and MR Imaging of the Whole Body

B

C

D FIG 44-3  Hypervascular pseudolesion in the posterior aspect of segment IV caused by an aberrant right gastric vein. A to C, Liver parenchyma of the posterior aspect of segment IV shows hyperdensity with a draining vein into this area on the portal phase of dynamic CT scan (arrow in B). This area is isodense on precontrast (A) and equilibrium-phase (C) images. D, Coronal maximum intensity projection during the portal phase indicates that the right gastric vein flows directly into this area and opacifies the parenchyma (arrow).

confirmation of the venous drainage by color Doppler is important.63 In addition, Gd-EOB-DTPA–enhanced MRI is also helpful for differentiation of these pseudolesions from HCC. On hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI, this hyperplastic area is not well visualized; on the other hand, the majority of early well-differentiated HCC shows hypointensity. Focal sparing of iron in hemosiderotic liver in this area was recently reported.116 The inferior veins of Sappey flow into the anterior aspect of segment IV from the chest wall through the falciform ligament;  they are also called paraumbilical veins. Focal fatty liver is often observed in this area136,137,259,338 (Fig. 44-8). This area also demonstrates early enhancement or hypodensity on dynamic CT and MRI.217 These changes are not seen in patients with portal hypertension because of reversed flow in this venous system.

Miscellaneous Types of Pseudolesion.  A hepatic segment with FIG 44-4  Hypervascular pseudolesion in the gallbladder fossa due to cholecystic venous drainage. Liver parenchyma around the gallbladder shows hyperdensity in the arterial phase of dynamic CT scan, owing to cystic venous drainage (arrows).

hepatic vein obstruction demonstrates the same imaging findings as  a hepatic segment in an area with portal vein obstruction.114,207 This phenomenon is often observed in living donor liver transplantation.127 The drainage flow from hypervascular HCC demonstrates corona enhancement surrounding the tumor, mimicking extratumoral 

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

1315

A

B

C

FIG 44-5  Focal sparing in the posterior aspect of segment IV caused by aberrant right gastric venous drainage. A, On precontrast CT the liver parenchyma in the posterior aspect of segment IV shows relative hyperdensity compared with the background fatty liver parenchyma (arrow). On T1-weighted in-phase (B) and opposed-phase (C) MRIs, the liver parenchyma—except for the posterior aspect of segment IV (arrow in C)—has decreased intensity in the opposed-phase image, which indicates fat deposition.

extension of the malignant neoplasm on dynamic CT or MRI (peritumoral pseudolesion).308 There may be a spared area in a fatty liver (peritumoral spared area)55,98 and rarely peritumoral focal fatty change.194

Nontumorous Hepatic Mass Lesions

Hepatic Cyst.  Solitary liver cysts are occasionally encountered. His-

FIG 44-6  Focal fatty liver in the posterior aspect of segment IV caused by aberrant right gastric venous drainage. There is a relatively hyperattenuated round lesion (arrow) in the liver parenchyma in the posterior aspect of segment IV on in-phase T1-weighted MRI. The lesion shows hypointensity on opposed-phase T1-weighted MRI (image not shown).

tologically a single layer of cuboidal epithelium is seen and the wall is composed of thin fibrous tissue. These cysts are frequently noted in older adults, implicating acquired factors in their pathogenesis. Although hepatic cysts are generally asymptomatic, compressive symptoms may develop if they become massive. Similar cysts may form diffusely throughout the liver in adult-type polycystic liver. On CT a hepatic cyst is depicted as a smooth-rimmed hypodense mass with CT values near zero. They appear hypointense on T1-weighted images and extremely hyperintense on T2-weighted images. They show no enhancement at all on either postcontrast CT or MRI (Fig. 44-9). When hemorrhage or inflammation occurs in the center of these cysts, the CT value increases and the signal intensity on MRI undergoes a variety of changes. Such lesions are referred to as complicated cysts (see Fig. 44-9).

1316

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 44-7  Hyperplastic change in the posterior aspect of segment IV in an alcoholic cirrhotic liver caused by aberrant right gastric venous drainage. The liver parenchyma in the posterior aspect of segment IV shows hyperintensity (arrow) on T1- (A) and hypointensity (arrow) on the T2-weighted image (B). On the SPIOenhanced T2-weighted MRI (C) the area shows uptake of SPIO (arrow) and is relatively hypointense compared with the background liver parenchyma.

A

B FIG 44-8  Focal fatty change in the drainage area of the inferior veins of Sappey. On precontrast (A) and arterial-phase (B) CT scans, the anterior aspect of segment IV adjacent to the falciform ligament shows hypoattenuation (arrow), which represents focal fat deposition.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

1317

B FIG 44-9  Hepatic cyst (polycystic liver). On T1-weighted (A) and T2-weighted (B) MRIs, there are multiple smooth-rimmed masses which demonstrate definite homogeneous hypointensity and hyperintensity, respectively, similar to cerebrospinal fluid. However, some of the cysts in the right lobe of the liver have a higher intensity than other cysts, which represents bleeding in the cyst (complicated cyst).

Polycystic Liver Disease.  Polycystic liver disease can be seen in the pediatric population; when complicated by congenital hepatic fibrosis, it may be associated with portal hypertension and esophageal varices. Renal cysts are also present in approximately 70% of cases. Innumerable cysts are observed within the liver, with calcification and complicated cysts frequently noted. The imaging features of these cysts are basically the same as those of hepatic cysts (see Fig. 44-9).

Ciliated Hepatic Foregut Cyst.  Ciliated hepatic foregut cyst (CHFC) is a unilobular and well-demarcated cyst whose inner surface is ciliated and covered with mucin-producing cylindrical or cubic epithelium. Histologically they are the same as bronchogenic cysts.53,109,292 These cysts tend to occur on the anterior surface of the liver beneath the hepatic capsule. They contain mucinous fluid with various concentrations of proteins and fats. Although CHFC is typically benign, malignant transformation to squamous cell carcinoma has been reported in few cases.53 The imaging findings of CHFC differ according to the concentration of mucin and the nature of the cyst’s contents.53,109 On nonenhanced CT, CHFC is often depicted as an isodense nodule relative to the surrounding liver; it is sometimes hyperdense or slightly hypodense. On dynamic CT the cyst does not enhance at all. On T2-weighted MRI, the findings are the same as those of hepatic cysts. On T1-weighted MRI, hyperintensity due to the presence of high concentrations of mucin is noted in some cases (Fig. 44-10).

Masslike Focal Parenchymal Changes  Focal sparing in fatty liver.  Focal sparing in fatty liver shows relative hyperdensity on nonenhanced CT (see Fig. 33-5), hypointensity on T1-weighted images, and relative hyperintensity on opposed-phase T1-weighted images (see Fig. 44-5). This finding is usually related to disturbances of localized portal flow, and its configuration is occasionally wedge shaped.13,180 Focal fatty change.  Focal fatty liver is also strongly related to localized portal flow disturbances.119 Common locations and imaging features are as described earlier under the heading “Pseudolesion Due to Portal Vein Obstruction” (see Figs. 44-6 and 44-8). Focal fatty liver can mimic dysplastic nodules or well-differentiated HCC with fatty change in cirrhotic livers (see “Hepatocellular Carcinoma,” later), and

the differentiation is often difficult when the change occurs in uncommon locations. Multiple focal nodular fatty infiltrations are rarely encountered and can mimic metastatic liver cancer, microabscesses, and biliary hamartomatosis (Fig. 44-11). Opposed-phase MRI can identify the fatty change and is useful in differentiating among these entities (see Fig. 44-6).148 Focal iron deposition.  Focal iron deposition can be seen in areas with disturbed portal perfusion and results in segmental hypointensity on T2-weighted images.108 Focal nodular hyperplasia.  Focal nodular hyperplasia (FNH) is defined as a tumorlike lesion characterized by a central fibrous scar with surrounding nodules of hyperplastic hepatocytes and small bile ductules in noncirrhotic livers. Most cases occur in women (the maleto-female ratio is 1 : 8).87 Because it is often associated with intrahepatic portal flow disturbance, FNH is now considered a reactive change  to abnormal blood circulation.143 Unlike hepatic adenoma, the asso­ ciation with oral contraceptive use is still controversial.67 Clinically FNH is usually asymptomatic, with epigastric pain or hepatomegaly infrequently developing. Unlike in hepatic adenoma, intraperitoneal rupture is extremely rare. Grossly FNH is a well-demarcated solitary mass without a capsule, often located beneath the surface of the liver. However, it can be multiple and pedunculated from the liver. Internal bleeding and necrosis are not usually seen. In the central scar, feeding arteries, draining veins connected to the hepatic vein, and bile duct proliferation are seen.56 The feeding arteries distribute to the peripheral portion of the tumor through the central scar in a spoke-wheel pattern. Variable amounts of Kupffer cells are present in FNH. On imaging, FNH usually has a slightly elliptical shape. On nonenhanced CT it appears as a homogeneous hypodense mass with a central scar showing more marked hypodensity. However, in small lesions, visualization of the central scar is not definite on CT. The entire nodule is well and homogenously enhanced in the arterial phase of CT and MRI. The central scar is seen as more hypodense,267 but in small lesions it is depicted infrequently.31 If vessels radiating from the central scar to the periphery of the tumor are visualized, a near-definitive diagnosis of FNH can be made. In this spoke-wheel pattern, the feeding vessels penetrate the fibrous septum from the central scar and distribute to the tumor (Figs. 44-12 and 44-13). This vasculature is most sensitively delineated by contrast-enhanced ultrasonography.

1318

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 44-10  Ciliated hepatic foregut cyst. A, Postcontrast CT scan shows a nonenhanced hypodense mass beneath the anterior surface of the liver (arrow). The internal density of the mass is higher than that of water. Both T1- (B) and T2-weighted (C) MRIs demonstrate extreme hyperintensity, reflecting a high intralesional concentration of protein (arrow).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

1319

hypervascular lesions show early enhancement in the arterial phase of dynamic CT and MRI, with variable signal intensity on T1- and T2-weighted images. On Gd-EOB-DTPA and SPIO-enhanced MRI, uptake of these contrast mediums can be seen. Recently this kind of lesion has attracted particular attention in patients with alcoholic liver cirrhosis. Furthermore, recent immunohistochemical analysis revealed that around two thirds of these nodules associated with alcoholic cirrhosis could be categorized into inflammatory-type hepatic adenoma (named serum amyloid A [SAA]-positive neoplasm because the existence of hepatic adenoma in cirrhotic livers are not well accepted), different from the other one third demonstrating the immunohistochemical features of FNH. Interestingly the former shows hypointensity but the latter iso- or hyperintensity on hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI.260

Macroregenerative Nodules.  Macroregenerative nodules are inciFIG 44-11  Multiple nodular fatty deposition in type C cirrhotic liver. On precontrast CT, multiple small hypodense nodules are observed in the liver. Multiple nodular fatty depositions were confirmed by MRI and follow-up studies.

FNH in most cases shows hypointensity on T1- and hyperintensity on T2-weighted images. However, because the tissue composition of FNH resembles that of the surrounding liver, it can exhibit isointensity on T1- or T2-weighted images. The central scar is depicted as hypointense on T1- and hyperintense on T2-weighted images (see Fig. 44-12). The finding of hyperintensity on the latter is attributable to the presence of a rich vasculature within the scar tissue (vascular scar).154 On dynamic MRI, homogeneous early enhancement is seen, with persistent staining. The central scar shows delayed enhancement.170 Both imaging and pathology studies have clarified that in FNH, the majority of the blood flow supplying the tumor from arteries refluxes to veins within the central scar and then enters the hepatic vein.56,193 The hemodynamics of FNH differs from that of hypervascular HCC (described later under “Hepatocellular Carcinoma”), and this difference is useful in the differential diagnosis. On dynamic CT and MRI of FNH, the margin of the stain is clear; hence washout of the contrast agent is not observed (see Figs. 44-12 and 44-13).193 In addition, direct draining from the tumor to the hepatic vein is occasionally depicted (see Fig. 44-12). Because Kupffer cells are present in FNH, the signal intensity decreases on SPIO-enhanced MRI (see Fig. 44-12). The degree of uptake is variable depending on differences in the distribution density of Kupffer cells.70,236 Hepatobiliary agents such as mangafodipir trisodium (Mn-DPDP) and Gd-EOB-DTPA are useful in the characterization of FNH. Because the hepatocytes in FNH take up these agents, FNH shows hyperintensity on T1-weighted images enhanced by them319,345 (Fig. 44-14). FNH often shows donut or ring-like peripheral hyperintensity on hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI (Fig. 44-15) because of lower expression of organic anion transporting polypeptide (OATP)1B3, which uptakes Gd-EOB-DTPA into hepatocytes in the hyperplastic hepatocyte surrounding the central scar.54,195,332 Recently, telangiectatic FNH, which was formerly categorized as one of the subtypes of FNH, was categorized into the subtypes of hepatic adenoma (inflammatory adenoma).22 FNH-like lesions.  FNH-like nodules are focal lesions occurring in cirrhotic livers that are morphologically similar to classic FNH  in otherwise normal livers.160,216,245,268,289 They are sometimes misdi­ agnosed as HCC on imaging because both lesions show definite enhancement in the arterial phase of dynamic CT or MRI.245 These

dentally detected in patients with cirrhosis and Budd-Chiari syndrome.26,168,317 They are larger than the usual regenerative nodules in some cases. Histologically, macroregenerative nodules resemble the regenerative nodules seen in cirrhosis. They typically exhibit hypointensity on T2-weighted images and various signal intensities on T1-weighted images. On dynamic CT and MRI they generally do not enhance, although in Budd-Chiari syndrome they may be multiple and show enhancement (Fig. 44-16).

Nodular Regenerative Hyperplasia.  Nodular regenerative hyperplasia consists of multiple monoacinar regenerative nodules in the absence of fibrous septa, with increased intranodular portal supply. It is frequently associated with various nonhepatic chronic diseases, including myeloproliferative disorders (leukemia, polycythemia vera), lymphoproliferative disorders (lymphoma, plasma cell dyscrasia), and collagen diseases (rheumatoid arthritis, polyarteritis nodosa, lupus erythematosus). The pathogenesis of nodular regenerative hyperplasia remains unclear, but it may be similar to that of partial nodular transformation; the differences between these two entities are related to the cause and the distribution of the portal vein obliteration.320 Imaging features of these nodules include hypointensity on T2-  weighted images; in contrast, true hepatic tumors usually show hyperintensity on T2-weighted images. On T1-weighted images they typically demonstrate hyperintensity.270 These MRI findings are similar to those of the hyperplastic changes caused by aberrant venous drainage in cirrhotic livers and by dysplastic nodules or well-differentiated HCC. Gd-EOB-DTPA may show donut-type hyperintensity similar to FNH on hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI (Fig. 44-17).

Peliosis Hepatis.  Peliosis hepatis consists of multiple small cystic lesions that communicate with sinusoids. Peliosis hepatis does not have endothelial cells in the cystic wall.321 On nonenhanced CT, peliosis hepatis is hypodense. It shows enhancement relative to the background liver in the arterial phase of dynamic CT and becomes progressively isodense. Peliosis hepatis usually demonstrates hyperintensity on T2-weighted MRIs; on T1-weighted images it is hypointense compared with the background liver. On T1-weighted images after contrast enhancement, peliosis hepatis may exhibit a centrifugal enhancement pattern.88

Biloma.  Biloma refers to a localized collection of bile that has escaped from the biliary tract, regardless of whether its location is intrahepatic or extrahepatic. Because the bile duct may rupture for a variety of reasons, including choledochal stones, bilomas may also form within the liver and subcapsular area.49,199 There has been a recent increase in the incidence of bilomas associated with bile duct injury

1320

PART II  CT and MR Imaging of the Whole Body

A

B

C

F

G

D

E

FIG 44-12  Focal nodular hyperplasia in a 44-year-old woman. A to C, On precontrast CT scan there is a homogeneous slightly hypodense mass with a central scar (slightly more hypodense) in the lateral segment of the liver (arrow in A). B, The lesion shows homogeneous enhancement with a less-enhanced central scar in the arterial phase (arrow). C, It is relatively hypodense compared with surrounding liver parenchyma, with a slightly enhanced central scar in the equilibrium phase (arrow). D, Direct draining from the tumor to the left hepatic vein (arrow) is observed on a coronal reconstruction of arterial-phase CT. The tumor shows isointensity (arrow) on T1- (E) and T2-weighted (F) MRIs. It demonstrates isointensity to the surrounding background liver parenchyma (arrow) on a SPIO-enhanced T2-weighted MRI (G), which represents the tumor’s uptake of SPIO.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

1321

FIG 44-13  Focal nodular hyperplasia in a 15-year-old girl. The arterial phase of dynamic CT shows an enhanced tumor (arrows) and the feeding arteries radiating from the central scar to the periphery (spokewheel pattern).

FIG 44-14  Retention of hepatobiliary contrast agent in a 34-year-old man with focal nodular hyperplasia. In the delayed-phase (hepatobiliaryphase) T1-weighted image obtained 20 minutes after administration of Gd-EOB-DTPA, the tumor shows prolonged enhancement (arrow) that represents uptake of the contrast agent by hepatocytes.

FIG 44-16  Hypervascular macroregenerative nodule in Budd-Chiari syndrome in a 28-year-old man. The nodule shows hyperdensity in the arterial phase of dynamic CT. The nodule demonstrated hyperintensity on T1- and isointensity on T2-weighted MRIs (not shown).

FIG 44-15  Donut or ringlike peripheral retention of hepatobiliary contrast agent in focal nodular hyperplasia. In the delayed-phase (hepatobiliary-phase) T1-weighted image obtained 20 minutes after administration of Gd-EOB-DTPA, the central scar portion shows lower expression of organic anion transporter polypeptide, and the tumor shows donut or ringlike peripheral retention of contrast agent.

1322

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 44-17  Nodular regenerative hyperplasia in idiopathic portal hypertension. A, On T1-weighted MRI, there is a hyperintense nodule with a central hypointense portion. B, On T2-weighted MRI, the nodule appears hypointense with a central hyperintense portion. C, On hepatobiliary-phase Gd-ROB-DTPA–enhanced MRI, the nodule shows hyperintensity that reflects uptake of EOB. D, On CTAP, the nodule appears hyperdense compared to the background liver, which suggests increase of portal flow compared to the surrounding liver.

after interventions such as transcatheter arterial embolization and radiofrequency ablation, as well as hepatic artery thrombosis following liver transplantation.129,138,261 The imaging findings of biloma resemble those of hepatic cyst, with proximal bile duct dilatation frequently noted. If hepatobiliary scintigraphy reveals accumulation, a definitive diagnosis can be made. Hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI may also help diagnose the presence of biloma.152

Biliary Hamartoma.  Biliary hamartoma, also known as von Meyenburg’s complex, is thought to be the result of a congenital defect in formation of the ductal plate. It is occasionally complicated by polycystic disease or peribiliary cysts.243 Biliary hamartoma is composed of bile ductule–like clusters surrounded by fibrous interstitium; it is not necessarily a rare condition, having been noted in 0.15% to 2.8% of cases in autopsy series. These nodules are usually small, in the range of 0.1 to 5 mm, but can occasionally measure up to 10 mm. In some cases they are limited to only one part of the liver; in other cases they are interspersed throughout. On nonenhanced CT, biliary hamartoma is depicted as tiny hypodense nodules; they do not show water attenuation because of a partial volume effect. These nodules may be difficult to differentiate

from hypovascular solid masses, particularly metastatic liver cancer, multiple microabscesses, and irregular fatty infiltration. On T2-  weighted MRIs and MR cholangiopancreatography (MRCP), they characteristically show marked hyperintensity, reflecting their cystic nature156,198,263 (Fig. 44-18).

Peribiliary Cysts.  Peribiliary cysts are considered to be retention cysts of the peribiliary glands. They are frequently associated with polycystic liver and cirrhosis.213,291 In liver cirrhosis they gradually increase in size and number with disease progression. On ultrasonography they are depicted as round or tubular anechoic lesions situated along the larger portal tracts.290 On CT and MRI they are depicted as tubular lesions along the portal vein or as beaded communicating cystic lesions.19,85,97,293 However, in some cases, the findings resemble a periportal collar on CT and periportal abnormal signal intensity on MRI.181 Characteristically, lesions are present on both sides of the proximal intrahepatic portal vein. Peribiliary cysts can exhibit the “central dot” sign on enhanced CT, like that seen in Caroli’s disease.1 MRCP is useful for obtaining the total picture of peribiliary cysts (Fig. 44-19). On endoscopic retrograde cholangiopancreatography (ERCP), contrast medium does not flow into peribiliary cysts, which is useful in differentiating them from Caroli’s disease.202

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

1323

B FIG 44-18  Multiple biliary hamartomatosis (von Meyenburg’s complex). A, On postcontrast CT, there are multiple minute hypodense nodules throughout the liver. B, On T2-weighted MRI at the same level, the nodules exhibit extreme hyperintensity. Most of the nodules are difficult to recognize on CT because of the partial volume effect.

B

A

C

FIG 44-19  Peribiliary cysts. A, Tubular hypointense lesions are apparent along the large portal vein, especially in the hepatic hilum area (arrow), on T1-weighted MRI. B, On T2-weighted MRI these lesions show hyperintensity (arrow). Actually, these lesions are not tubular but represent multiple small cysts around the large bile duct. C, On MRCP, multiple hyperintense cystic lesions are observed around large bile ducts in the hepatic hilum area (arrow).

Caroli’s Disease.  Caroli’s disease is a rare condition causing cystic dilation of the intrahepatic bile duct. It is somewhat more frequent in males and is detected in childhood or adolescence in most cases because of symptoms resulting from cholangitis or hepatic abscess. In the majority of cases, renal cystic disease, including renal tubular ectasia, is associated.214 On ultrasonography and CT, cystic dilatation of the bile duct is characteristic.157,190,199 The portal vein within a dilated bile duct is well visualized as a punctate or linear stain on enhanced CT, known as the central dot sign35 (Fig. 44-20). This portal vein also shows a flow void on MRI, called an intracystic central flow void sign.73

Infectious Liver Masses Pyogenic abscess.  Bacteria can reach the liver via the portal vein, hepatic artery, or bile duct, as well as by direct infection from adjacent structures.86 Many cases of pyogenic abscess are associated with ascending cholangitis due to bile duct disease; however, the cause is often

unclear, and in such cases the route of infection is presumed to be via the portal vein. Escherichia coli and Klebsiella species are the most common causative bacteria. Imaging findings reveal a variety of stages and degrees of severity, but basically an abscess cavity with a center filled with pus and inflammatory changes in the adjacent tissues are noted. Lesions are both unilocular and multilocular as well as multiple and solitary. Cases associated with cholangitis tend to be multiple. Cases due to bacteremia frequently show multiple microabscesses. The shape of the abscess cavity is generally irregular, while its margin tends to assume a clear, round shape in the healing stage. On nonenhanced CT, marked hypointensity due to the presence of pus in the center of the abscess is found. In addition, slight hypointensity, presumably due to the influence of inflammation on the adjacent tissues, produces a characteristic double-structured hypodense area.74 Although the shape is usually irregular, in a few cases it is virtually round, and the double structure is not clearly seen. Fusion of small

1324

PART II  CT and MR Imaging of the Whole Body

A

B FIG 44-20  Caroli’s disease. A, Precontrast CT scan shows multiple cystic lesions that contain dotlike structures. Some of these cystic lesions contain calcified stone (arrow). B, On postcontrast CT scan the dotlike structures are well enhanced and recognizable as portal veins.

A

B

C

FIG 44-21  Hepatic abscess caused by intrahepatic cholelithiasis. A, On precontrast CT scan there is a hypodense mass (arrow). B, On arterial-phase CT the mass has a target-like appearance (arrow)—a hypodense center surrounded by an enhanced hyperdense middle and a hypodense periphery. Transient wedge-shaped segmental enhancement is also observed surrounding the mass (arrowhead). C, On equilibrium-phase CT the middle and outer parts of the mass show a thick ringlike enhancement (arrow).

abscesses to form a large abscess is sometimes observed (cluster sign).106 During the arterial phase of dynamic CT, a double-target sign175 is seen: the area around the center, which is not enhanced, is stained (medial layer); the outermost layer shows hypodensity. In the portal to equilibrium phases, a thick ringlike stain due to staining of the medial and outermost layers is seen. Also, during the arterial phase, segmental or wedge-shaped enhancement is transiently found around the abscess (Fig. 44-21). This wedge-shaped enhancement disappears in the late phase. It may be attributed to narrowing or obstruction of the portal vein branches due to inflammatory changes of the portal tract around the abscess, which results in decreased portal flow and a compensatory increase in arterial flow.62 On MRI, liver abscesses show hyperintensity on T2-weighted images, with the abscess cavity showing even more marked hyperintensity.17,120,188 During the arterial phase of dynamic MRI, as on dynamic CT, transient wedge-shaped enhancement is shown around the abscess; in the same area on T2-weighted images, wedge-shaped hyperintensity

may be shown (Fig. 44-22). Because these findings are changeable after 1 to 2 weeks of antibiotic therapy, this helps differentiate abscesses from malignant tumors.62 Gas bubbles can be rarely found. Bile duct diseases are the most common cause of liver abscess. In particular, abscess frequently evolves from cholangitis, making early diagnosis of cholangitis of great importance. Clinically, cholangitis should be suspected in the presence of fever and elevated ductal enzymes. In the arterial phase of dynamic CT, acute cholangitis characteristically shows transient diffuse enhancement along the portal tracts within the liver12 (Fig. 44-23). Amebic abscess.  Amebic abscess, which results from Entamoeba histolytica infection, is a frequent type of hepatic abscess in tropical and subtropical regions. Because it is generally secondary to amebic infection of the colon, it reaches the liver via the portal vein.47 As the abscess progresses, its core undergoes liquefaction and cavitary changes and becomes surrounded by necrotic inflammatory cell infiltration and fibrous layers. In the liver, one large focus is usually found, although

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

1325

C

B

FIG 44-22  Hepatic abscess of unknown cause. A, On fat-suppressed T2-weighted MRI there is an inhomogeneous hyperintense mass (arrow) with a very hyperintense area, suggesting a collection of pus. B, On arterial-phase Gd-enhanced T1-weighted MRI the peripheral part of the mass shows wedge-shaped segmental enhancement (arrowhead). Septum-like enhancement in the mass is also observed (arrow). C, On the equilibrium-phase image a nonenhanced cystic mass with an enhanced irregular thick wall is visualized (arrow).

A

B FIG 44-23  Intrahepatic cholangitis with choledocholithiasis. A, In the portal phase of dynamic CT, there is transient diffuse enhancement along the portal tract within the liver; this disappears during the equilibrium phase (B).

multiple foci are sometimes seen. In a few cases the abscess ruptures, leading to abscess formation in the lung or peritoneal cavity.107,227 The imaging findings are essentially nonspecific and resemble those of pyogenic abscess.200,248,249 Postcontrast CT reveals an enhanced mural structure with a strong tendency to exhibit a hypodense area at its lateral side owing to the presence of edema. Segmental enhancement similar to that seen with bacterial abscess is noted in some cases.

Fungal abscess.  Fungal abscess is caused by an opportunistic infection seen in immunosuppressed patients such as those with acquired immunodeficiency syndrome (AIDS) or leukemia or those who have undergone bone marrow transplantation or treatment for malignant tumors.5,238 Candidiasis is the most common fungal infection encountered. In most cases it presents as multiple microabscesses. In many cases lesions develop in the spleen and liver simultaneously.

1326

PART II  CT and MR Imaging of the Whole Body

A

B FIG 44-24  Hepatic and splenic microabscesses (Candida) in a patient with acute myelogenous leukemia. A, In the arterial phase of dynamic CT there are multiple small nodules with faint peripheral ringlike enhancement in the liver and spleen (arrows). B, On equilibrium-phase CT they appear hypodense compared with the background liver (arrows).

A

B FIG 44-25  Hepatic microabscesses in a patient with acute lymphocytic leukemia. A, On arterial-phase Gd-enhanced T1-weighted MRI there are small hypointense nodules with faint peripheral ringlike enhancement (arrows). B, On fat-suppressed T2-weighted MRI these lesions appear hyperintense and additional smaller nodules are observed (arrows).

Microabscess shows faint ringlike enhancement in the arterial phase of dynamic CT and hypodensity in the equilibrium phase (Fig. 44-24). Calcifications are sometimes present in chronic microabscesses. On T1-weighted images they demonstrate hypointensity, and on T2-weighted images, marked hyperintensity. On dynamic MRI, as on CT, the rim of the lesion is preferentially enhanced8 (Fig. 44-25). Actinomycosis, an infection caused by Actinomyces israelii, usually assumes a chronic course.279 Most commonly, liver lesions develop via the portal vein from foci located in the intestines, particularly the ileocecal area. Occasionally pulmonary lesions penetrate the diaphragm and spread to the liver. Liver lesions characteristically form  a honeycomb abscess that contains 1- to 2-mm fungus balls. The

imaging findings resemble those of granuloma (described later under “Granuloma”), with slightly to moderately increased vascularity. In the arterial phase of dynamic CT, a solid mass that is isodense or hyperdense relative to the surrounding liver is noted. In the equilibrium phase, delayed enhancement is seen.326

Inflammatory Pseudotumor.  Inflammatory pseudotumor, which is also called plasma cell granuloma and postinflammatory tumor, is characterized histologically by connective tissue proliferation associated with plasma cell and fibroblast infiltration.262 Similar lesions are commonly found in the lungs, ocular fossa, and parotid glands; they are rare in the liver,220 although the number of reported cases has recently

CHAPTER 44  Liver: Focal Hepatic Mass Lesions increased. The pathogenesis of inflammatory pseudotumor is divided into two types. In the first, inflammation (e.g., liver abscess) is  surmised to be the cause (fibrohistiocytic type); in the other, autoimmune mechanisms are implicated (lymphoplasmacytic type).222,347 Hepatic inflammatory pseudotumor may be associated with other unusual inflammatory or autoimmune diseases such as sclerosing cholangitis, phlebitis, and retroperitoneal fibrosis.347 Inflammatory pseudotumor infrequently involves the porta hepatis or bile duct, which results in obstructive jaundice.347 Spontaneous regression has been reported.328 The imaging findings of pseudotumors caused by inflammation are the same as those of fresh granuloma, with the imaging findings differing according to the degree of fibrous change and inflammatory cell infiltration in the mass. When inflammatory cell infiltration is predominant, enhancement is seen in the arterial phase; when fibrous tissue predominates, delayed enhancement is seen on postcontrast CT, as in tumors with a rich fibrous interstitium58,333 (Fig. 44-26).

A

Autoimmune pseudotumor is associated with autoimmune pancreatitis in a high proportion of cases, either synchronously or nonsynchronously, and it is considered to be one type of immunoglobulin  (Ig)G4-related autoimmune disease.222,347 In this condition an inflammatory mass forms at a periductal site, resulting in dilation of peripheral bile ducts owing to intrahepatic bile duct stricture in some cases (Fig. 44-27). The imaging findings resemble those of hepatic hilar cholangiocarcinoma.

Granuloma.  Hepatic granuloma is associated with a wide range of underlying factors, including bacterial, viral, parasitic, and fungal infections; biliary tract diseases; systemic diseases such as sarcoidosis; and exposure to various pharmacologic and chemical substances.18,123,235,315 With the exception of cases associated with infection, most lesions are microscopic and thus not a suitable focus for diagnostic imaging. Among infection-related lesions, bacterial hepatic abscess frequently assumes a chronic course, as do some cases of 

C

B

D

1327

E FIG 44-26  Inflammatory pseudotumor. A to C, On precontrast CT a large isodense mass with a peripheral hypodense rim is observed (arrow in A). In the arterial phase the mass shows faint enhancement with a less-enhanced outer rimlike structure (arrow in B). In the equilibrium phase the central part demonstrates slight enhancement and the peripheral part demonstrates dense delayed enhancement (arrow in C). D, On T1-weighted MRI the central part of the mass shows isointensity and the peripheral part, hypointensity (arrow). E, On T2-weighted MRI it is visualized as an area of inhomogeneous hyperintensity (arrow).

1328

PART II  CT and MR Imaging of the Whole Body fungal and parasitic infection. Also, caseous necrosis develops in the core of some tuberculomas.118 Calcification is frequent in infectious granulomas.3,278 The imaging findings of granuloma differ according to stage. Granulomas with marked inflammatory cell infiltration show slight to moderate vascular proliferation and are enhanced in the arterial phase of dynamic CT and MRI. In the late phase they show isointensity or moderate hyperintensity relative to the surrounding liver, with this finding characteristically persisting long term. When an old fibrous component makes up the bulk of the lesion, it becomes hypovascular, and although it is not clearly enhanced on dynamic CT, it shows delayed enhancement relative to the surrounding liver on late-phase CT.65,337 On MRI it shows hypointensity and hyperintensity on T1- and T2-weighted images, respectively (Fig. 44-28), making it difficult to distinguish from tumor. Some lesions with prominent fibrosis exhibit hypointensity on T2-weighted images. These findings are similar to those of highly fibrous cholangiocarcinomas, cholangiolocellular carcinoma, and liver metastases of scirrhous cancer, making the differential diagnosis extremely difficult.

FIG 44-27  Autoimmune-related inflammatory pseudotumor. On postcontrast CT a slightly hypodense mass is observed along the left portal vein (arrow). This is an IgG4-related inflammatory mass in the area of the bile ducts.

A

B

D

C

E FIG 44-28  Hepatic granuloma (unknown cause) in a 39-year-old woman. A to C, On precontrast CT a hypodense nodule is observed (arrow in A). B, In the arterial phase it shows faint enhancement (arrow). C, In the equilibrium phase it demonstrates delayed enhancement and becomes isodense to the background liver (arrow). On T1- (D) and T2-weighted (E) MRIs, the nodule appears hypointense and hyperintense, respectively (arrow).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions Eosinophilic Abscess.  Hepatic eosinophilic abscess is commonly associated with peripheral eosinophilia. Activated Kupffer cells stimulate the eosinophils to release some humoral factors that induce  localized coagulopathy and thromboembolism. Histopathologically, eosinophilic abscess consists of coagulation necrosis and inflammatory cell infiltration, especially eosinophils. Eosinophilic abscess shows hypodensity on postcontrast CT, with indistinct margins and nonspherical contours344 (Fig. 44-29).

Parasitic Diseases.  The following parasitic diseases might show mass lesion on imaging: Echinococcus granulosus (hydatid) cyst,112,200 Echinococcus multilocularis,139,200 schistosomiasis (Schistosoma japonicum, Schistosoma haematobium, and Schistosoma mansoni),196,197,200 fascioliasis,76 clonorchiasis,33,161 and visceral larva migrans.32

Vascular Diseases Hepatic infarction.  The normal liver receives approximately 60% to 65% of its blood flow via the portal vein and 35% to 40% via the hepatic artery, making it less susceptible to infarction than most other organs. Most hepatic infarctions are due to hepatic artery occlusion

A

associated with systemic circulatory disturbance.228,303 Histologically, coagulation necrosis is found in the infarct center, with compensatory hyperemia seen in the periphery. An ill-defined wedge-shaped hypodense area is seen on CT. When bleeding is present, the center may appear hyperdense. On enhanced CT or MRI, the infarcted area does not stain. In acute hepatic necrosis, gas may be noted in the necrotic focus7,271 (Fig. 44-30). When portal flow is reduced for whatever reason, the number of hepatocytes decreases and the hepatic sinusoids undergo dilation and hyperemia. This is known as Zahn’s infarction. This entity is widely seen in cases of intrahepatic portal vein obstruction due to thrombosis and tumor growth (Fig. 44-31). Anoxic pseudolobular necrosis.  Anoxic pseudolobular necrosis (focal ischemic necrosis) is coagulative necrosis that involves one or more pseudolobules in cirrhotic livers. It usually follows massive gastrointestinal hemorrhage and shock. It is occasionally difficult to differentiate from HCC with massive internal coagulation necrosis.55,125,147 When ischemic necrosis of regenerative nodules is seen diffusely, it is called ischemic hepatitis. This mimics diffuse hypovascular HCC or metastasis111 (Fig. 44-32).

C

B

FIG 44-29  Eosinophilic abscesses in a 79-year-old man. On precontrast CT (A), multiple slightly hypodense masses of various sizes and shapes are seen. They are not enhanced on arterial-phase (B) and equilibriumphase (C) images.

A

1329

B FIG 44-30  Hepatic infarction. A, Precontrast CT scan shows wedge-shaped hypodense areas in the peripheral portion of the liver (arrows). B, They demonstrate no enhancement on equilibrium-phase CT scan (arrows).

1330

PART II  CT and MR Imaging of the Whole Body

A

B FIG 44-31  Zahn’s infarction. A, On precontrast CT the left and caudate lobes of the liver show slight hypodensity (arrows). B, In the arterial phase of dynamic CT these lobes demonstrate segmental staining (arrows).

FIG 44-32  Ischemic necrosis of regenerative nodules following esophageal variceal rupture in a cirrhotic patient. Postcontrast CT scan shows tiny nonenhanced nodules with an irregular and patchy configuration (arrows).

Confluent hepatic fibrosis.  Confluent hepatic fibrosis is observed in about 14% of patients with advanced cirrhosis who undergo liver transplantation. Histopathologically, approximation of remnant portal triads within the fibrotic areas are seen. On nonenhanced CT, confluent hepatic fibrosis shows wedge-shaped or geographic hypodensity. After contrast enhancement, the lesion becomes less obvious compared with nonenhanced CT.230 On MRI the lesions are hyperintense on T2-weighted images and hypointense on T1-weighted images compared with the background liver231 (Fig. 44-33).

Benign Hepatic Tumors

Cavernous Hemangioma.  Cavernous hemangioma is the most frequently encountered benign hepatic tumor in clinical practice, with an

incidence between 0.4% and 20%.241,265 It is seen in all age groups and has a female preponderance. It is detected in the absence of signs and symptoms in almost all cases. When the tumor exceeds 4 cm, abdominal pain or discomfort or a palpable mass may be present. Rupture occurs rarely.240,241 Rare complications of giant lesions include consumptive coagulopathy, thrombocytopenia, and hypofibrinogenemia (Kasabach-Merritt syndrome).6 Hemangiomas are usually solitary but are multiple in approximately 10% of cases. Their borders are clear, but they are not encapsulated. The cut surface is spongelike and typically dark reddish. Histologically it is made up of variously sized vascular spaces lined by flat endothelial cells and interspersed with fibrous tissue. As the hemangioma grows, various degenerative changes are seen in its center, including old and new thrombus formation, necrosis, scarring, hemorrhage, and calcification.316 When degeneration and fibrous changes become more prominent, the lesion is referred to as a sclerosed hemangioma.2,174 On nonenhanced CT, hemangioma is depicted as a well-demarcated hypodense area with the same density as the aorta. It is sometimes round but more often oval or irregular. Large lesions sometimes have a geographic (irregular) shape. In the arterial phase of dynamic CT, peripheral enhancement is seen first, followed by gradual filling toward the center (filling in) and prolonged enhancement on the equilibrium phase—a pattern characteristic of hemangioma. The density of hemangiomas reflects the vascular spaces, and on precontrast, arterial, and equilibrium-phase dynamic CT, the fact that the tumor’s density is similar to that of the aorta is useful diagnostic proof101,155,246 (Fig. 44-34). In the type of small hemangioma containing small sinusoids, dynamic CT may show the entire tumor to be enhanced from the early phase316 (Fig. 44-35). In hemangiomas exceeding 4 cm, a more irregular hypodense area in the center or a septum-like structure may be found because of thrombus, necrosis, and scar formation (Fig. 44-36). Calcification and cystic degeneration are also found in some cases.83,192 Rarely, fluid-fluid formation is seen.102 T1-weighted MRIs show homogeneous hypointensity, and T2-weighted images show homogeneous marked hyperintensity. The signal intensity on T2 is usually higher than that of the spleen (see Figs. 44-34 to 44-36).305 On dynamic MRI the enhancement pattern is the

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D FIG 44-33  Confluent hepatic fibrosis in hepatitis C–related cirrhotic liver. A, Precontrast CT shows a geographically hypodense area (arrow). B, The lesion demonstrates delayed enhancement on postcontrast CT (arrow). The lesion shows hypointensity (arrow) on T1- (C) and hyperintensity (arrow) on T2-weighted MRI (D).

1331

1332

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F

FIG 44-34  Cavernous hemangioma and HCC in a patient with hepatitis C–related chronic hepatitis. A to C, On precontrast CT a hypodense mass is observed (arrow in A). B, In the arterial phase it shows peripheral dense, spotty enhancement (arrow). C, In the equilibrium phase the attenuated area spreads throughout the mass; this is called filling in (arrow). The density of the lesion is almost the same as that of the aorta in all phases of dynamic CT. D, On T2-weighted MRI the mass demonstrates definite hyperintensity (arrow) similar to spinal fluid. It shows slight peripheral enhancement (arrow) on the arterial-phase Gd-enhanced T1-weighted image (E) and filling in (arrow) on the late-phase image (F). Arrowheads in B, D, and E indicate associated minute HCC.

B

A

C

D

FIG 44-35  Small hepatic hemangioma—early enhancing type. A, T1-weighted MRI shows a small hypointense mass (arrow). B, The mass demonstrates extreme hyperintensity (arrow) on the T2-weighted image. C, Arterial-phase Gd-enhanced T1-weighted MRI reveals definite enhancement of the entire tumor (arrow). D, Prolonged enhancement is shown in the late phase (arrow).

A

B

C

FIG 44-36  Giant cavernous hemangioma with central liquefaction. A, On the T1-weighted image the mass shows definite hypointensity, with even more hypointense central portions (arrow). B, On the T2-weighted image, the mass demonstrates definite hyperintensity, and the portions that appear very hypointense on T1 show more hyperintensity (arrow). C, The equilibrium-phase Gd-enhanced MRI reveals filling in and prolonged enhancement with central nonenhanced portions, indicating central liquefaction (arrow).

1334

PART II  CT and MR Imaging of the Whole Body

FIG 44-37  Sclerosed hemangioma. Postcontrast CT shows a slightly enhanced mass with internal heterogeneity (arrow).

same as that seen on CT. The larger the lesion becomes, the more inhomogeneous it appears on diagnostic imaging because of thrombus and scar formation in the center (see Fig. 44-36). Small AP shunts can complicate even small hemangiomas and are not rare. A hemangioma with an AP shunt usually shows enhancement of the entire lesion from the early phase, with less marked enhancement seen in the late phase compared with the early one.128 Sclerosed hemangiomas show hypodensity with delayed enhancement on dynamic CT and hypointensity in relation to cerebrospinal fluid on T2-weighted images2 (Fig. 44-37). Hemangioma occasionally appears as a so-called pseudo-washout sign in the transitional (late) phase of dynamic Gd-EOB-DTPA MRI and might mimic hypervascular HCC.43

Hepatic Adenoma.  Hepatic adenoma tends to occur in a younger group compared with FNH and overwhelmingly affects women,  most of whom have a history of oral contraceptive use. In addition the relationship between this tumor and glycogen storage disease is well recognized.122 In many cases hepatic adenoma is discovered because of an abdominal mass or recurrent abdominal pain, but it can also present as acute abdomen owing to tumor rupture. Hepatic adenoma develops in noncirrhotic livers and is usually solitary but may occur in multiple form (hepatic adenomatosis).72 The tumor border is clear and there is usually no capsule, although part of the tumor or its entire circumference may be covered by a fibrous capsule in some cases. In its core, bleeding, necrosis, and scar tissue are seen. Neither portal veins nor bile ducts are found, although bile ductules may develop bile plugs. Recently new genotype/phenotype classification of hepatic adenoma was released.22 According to it, hepatic adenoma is divided into the following four categories: HNF1αinactivated type, inflammatory type, β-catenin–activated type, and unclassified type. HNF1α-inactivated adenoma is female predominant and characterized by intratumoral fat deposition. Inflammatory adenoma is female predominant and has potential malignant transformation and steatosis in background liver. The lesions formerly

known as telangiectatic FNH are currently included in inflammatory adenoma.22 β-Catenin–activated adenoma is male predominant, has a higher potential of malignant transformation, and contains vague scar within the tumor. In most cases, nonenhanced CT depicts this tumor as a hypodense mass. Small tumors appear homogeneous, but in general the center shows inhomogeneous attenuation. Coagulum or thrombus may be noted in relatively early stages of ruptured hepatic adenoma as areas of hyperattenuation. In tumors with marked fatty metamorphosis of hepatocytes, a fat-related hypodense area may be seen. In addition, necrotic foci and scar tissue may appear as hypodense areas. Calcification is seen within the tumor in some cases. In the arterial phase of dynamic CT, moderate tumor enhancement is seen.72 HNF1α-inactivated adenoma shows diffuse signal dropout on T1-weighted chemical shift sequence due to steatosis, isosignal or slight hypersignal on T2-weighted images, and moderate enhancement in the arterial phase, with no persistent enhancement in the portal venous and delayed phases (Fig. 44-38). Inflammatory adenoma demonstrates an absence or only focal signal dropout on chemical shift sequences, marked hypersignal on T2-weighted sequences, with a stronger signal in the outer part of the lesions correlating with sinusoidal dilatation areas, and strong arterial enhancement with persistent enhancement in the portal venous and delayed phases151 (Fig. 44-39). In addition, on T2-weighted MRI, a hyperintense signal band in the periphery of the lesion with a center that is isointense to surrounding liver—called a characteristic atoll sign—can be seen.312 MRI findings of β-catenin– activated and unclassified adenoma are not yet well described. On hepatobiliary-phase Gd-EOB-MRI, more than 90% of hepatic adenomas demonstrated hypointensity relative to the surrounding liver distinctly different from FNH; this is valuable for differential diagnosis.71 However, β-catenin–activated adenoma commonly shows Gd-EOBDTPA uptake with iso- or hyperintensity on hepatobiliary-phase GdEOB-DTPA–enhanced MRI because of OATP1B3(8) overexpression in adenoma cells332 (Fig. 44-40).

Bile Duct Adenoma.  Bile duct adenoma is a rare benign epithelial tumor derived from bile duct cells. It is usually located on the surface of the liver. Pathologically it consists of a proliferation of ductules within a connective tissue stroma.4 In reported cases, ultrasonography showed it as a hyperechoic nodule with or without a hypoechoic rim. On dynamic CT or dynamic MRI it shows rapid enhancement in the arterial phase and delayed or prolonged enhancement owing to the fibrous component in the delayed phase.283

Angiomyolipoma.  Angiomyolipoma is composed of mature fat, blood vessels, and smooth muscle cells and is not encapsulated.223,304 Immunohistochemical staining of tumor cells reveals HMB-45 positivity.225 On imaging, if the fatty component predominates, it resembles lipoma,224,242 but in general it shows a mixture of a usual solid portion intermingled with a fatty component. Tuberous sclerosis has occasionally been reported to complicate this condition. On nonenhanced CT it appears as a solid mass containing markedly hypodense areas. On MRI, both T1- and T2-weighted images show the same hyperintensity as fatty tissue, with the intensity decreasing with fat suppression.242,329 To detect small volumes of fat, MRI chemical shift imaging (in-phase/opposed-phase) is useful.16 Also, when the vascular component is prominent it appears as a markedly enhanced area on dynamic CT and MRI. However, unlike renal angiomyolipoma, 50% of hepatic angiomyolipomas lack considerable fat content.304 Low-fat tumors may be difficult to differentiate from hypervascular hepatic tumors such as HCC.341 The drainage vein of angiomyolipoma is the hepatic vein, and identifying a perfusing vein communicating with the

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

D

E

C

FIG 44-38  Hepatic adenoma in a 37-year-old woman with a 14-year history of oral contraceptive use. There are two mass lesions in segment VII and segment I. On T1-weighted in-phase (A) and opposed-phase (B) MRIs, these tumors show decreased intensity in the opposed phase, which indicates fat deposition in the tumor. C, On T2-weighted MRI, these tumors show slightly hyperintense compared to the liver parenchyma. D, On arterial-phase Gd-EOB-DTPA–enhanced fat-suppressed T1-weighted MRI, these lesions appear isointense compared to the background liver, which indicates slightly early enhancement of these tumors. E, On hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI, these lesions appear hypointense compared to the background liver. The tumor in segment VII was proved to be HNF1α-inactivated adenoma by biopsy. (Courtesy L. Grazioli, University of Brescia, Brescia, Italy.)

1335

A

B

C

D

E

F FIG 44-39  Hepatic adenoma coexisting with focal nodular hyperplasia. There are two mass lesions on segment VII and segment I of the liver. On T1-weighted in-phase (A) and opposed-phase (B) MRIs, these lesions show no decreased intensity in the opposed phase, which indicates no fat deposition in the tumor. C, On T2-weighted MRI the lesion on segment VII appears hyperintense and the lesion on segment I appears isointense compared to the liver. On arterial-phase (D) and late-phase (E) Gd-EOB-DTPA–enhanced fatsuppressed T1-MRI, both lesions show marked enhancement on arterial phase, and the lesion on segment VII shows slightly decreased intensity with rimlike prolonged enhancement; in contrast, the lesion on segment I shows uniform prolonged enhancement. F, On hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI the lesion on segment VII appears hypointense and the lesion on segment I appears hyperintense compared to the background liver. The lesion on segment VII was proved to be inflammatory adenoma, and the lesion on segment I was proved to be focal nodular hyperplasia. (Courtesy L. Grazioli, University of Brescia, Brescia, Italy.)

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D

E

F FIG 44-40  Hepatic adenoma in a 15-year-old girl. There are multiple masses on the liver. On T1-weighted in-phase (A) and opposed-phase (B) MRIs, these lesions show no decreased intensity in the opposed phase, which indicates no fat deposition in the tumor. C, On T2-weighted MRI, these tumors appear slightly hyperintense compared to the liver parenchyma. On arterial-phase (D) and late-phase (E) Gd-EOB-DTPA–enhanced fat-suppressed T1-weighted MRI, these tumors show marked enhancement on arterial phase and prolonged enhancement on late phase. F, On hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI, these lesions appear hyperintense compared to the background liver. The largest tumor on segment IV was proved to be β-cateninactivated adenoma by biopsy. (Courtesy L. Grazioli, University of Brescia, Brescia, Italy.)

1337

1338

PART II  CT and MR Imaging of the Whole Body

B

A

C

E

D

FIG 44-41  Angiomyolipoma in a 28-year-old woman. A to C, Precontrast CT shows a large lobulated hypodense mass with very hypodense areas internally (arrows in A). B, In the arterial phase the mass is well attenuated, mainly in the peripheral part (arrows). C, In the equilibrium phase the mass is isodense to hypodense (arrows), with dense enhancement of the draining hepatic veins (arrowheads). The central part of the mass shows a decrease in signal in the opposed-phase T1-weighted image (E) (arrow) compared with the in-phase T1-weighted image (D), which indicates the presence of a fatty component in the central part of the mass.

hepatic vein from the tumor center can aid in differentiating angiomyolipoma from fat-containing HCC349 (Fig. 44-41).

Infantile Hemangioendothelioma.  This lesion resembles the capillary hemangioma seen in infantile skin and mucosa. Most cases become apparent within 6 months of birth and come to medical attention because of the presence of hepatomegaly and an abdominal mass and occasionally jaundice and congestive heart failure.38 Severe complications include heart failure, hepatic failure, and tumor rupture. If these complications can be avoided, the majority of these tumors regress spontaneously. Infantile hemangioendotheliomas are frequently multiple and range in size from barely discernible to 15 cm. They are well demarcated, and the tumor margin is made up of proliferating endothelial cells and new vessels. The center shows varying degrees of bleeding, fibrosis, necrosis, and calcification.

On nonenhanced CT, a relatively clear hypodense area is seen;  on dynamic CT, a staining pattern similar to that of cavernous  hemangiomas is noted. Dilated portal and hepatic veins are often visualized (Fig. 44-42). The MRI findings resemble those of hepatic hemangioma.201,251

Leiomyoma.  This tumor is extremely rare, and primary hepatic lesions and leiomyosarcoma must be differentiated.51,311 On dynamic CT it is depicted as a hypervascular tumor, and distinguishing leiomyoma from other hypervascular tumors such as HCC may be very difficult.

Mesenchymal Hamartoma.  Most mesenchymal hamartomas are seen in children younger than 2 years, and they account for 7% of all primary hepatic tumors in the pediatric population. In almost all cases

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

1339

C

FIG 44-42  Infantile hemangioendothelioma in a newborn (male). A, Precontrast CT shows a large hypodense mass (arrow). B, The arterial phase demonstrates inhomogeneous but marked enhancement (arrow). C, The equilibrium phase demonstrates dense prolonged enhancement with an irregular nonenhanced area (arrow). The density of the main structure of the mass is the same as that of the aorta, indicating large and abundant intratumoral blood spaces.

Pseudolymphoma.  Pseudolymphoma (reactive lymphoid hyperplasia) of the liver is a rare benign lesion. Microscopically it consists of hyperplastic lymphoid follicles with distinctive germinal centers and interfollicular areas containing mature lymphocytes and plasma cells. Histologically the lesion mimics Castleman’s lymphoma or low-grade malignant lymphoma of the liver.268 In reported cases it was seen as a hypervascular mass on dynamic CT and MRI186,233 (Fig. 44-44). Pseudolymphoma shows characteristic findings such as a homogeneous parenchymal pattern, vessel-penetrating sign in larger nodules, and a unique enhancement pattern known as perinodular enhancement, which reflects increased arterial supply in perinodular hepatic parenchyma due to portal venular stenosis and/or disappearance due to marked lymphoid cell infiltration in perinodular portal tracts.336 The lesions usually regress without medication.233

Miscellaneous Benign Hepatic Tumors.  Other uncommon FIG 44-43  Mesenchymal hamartoma in an 8-year-old girl. Postcontrast CT shows a well-demarcated enhanced solid mass with an internal cystic portion (arrow). (Courtesy B.I. Choi, Seoul National University, Seoul, Republic of Korea.)

they are discovered because of an abdominal mass and swelling. They occur more frequently in the right lobe, measure 2 to 29 cm, and are usually solitary and well demarcated, with a few showing pedunculated growth. Macroscopically they are fibrous masses containing small and large cysts; the larger cysts contain a septum.95,272,327 The cysts are filled with a transparent gelatin-like fluid. The fibrous portion is edematous, is made up of proliferating fibroblast-like mesenchymal cells, and includes bile ducts and liver parenchyma. Mesenchymal hamartoma with a predominant solid component has also been reported.172 On CT a well-demarcated cystic mass is seen; within or between the cysts, a septum-like structure is found. After administration of contrast medium, this structure within or between the cysts is enhanced (Fig. 44-43). T1-weighted MRIs show hypointensity, and T2-weighted images show marked hyperintensity, probably reflecting cystic changes in most cases.327,330

benign tumors include lymphangioma, fibroma, adrenal rest tumor, and pancreatic heterotopia.

Malignant Hepatic Tumors The most common malignant hepatic neoplasms are metastatic tumors. The majority of primary malignant neoplasms originate from hepatocytes and bile duct epithelium. Mesenchymal tumors such as epithelioid hemangioendothelioma, angiosarcoma, primary lymphoma, and other sarcomas are rare.

Hepatocellular Carcinoma Incidence.  HCC is one of the most frequent malignancies in the world; it ranks as the fifth most common cause of cancer in men and the seventh in women, with an estimated number of new cases of 520,000 and 230,000, respectively, for men and women in 2008.50 The highest age-adjusted incidence rates (>20 per 100,000) were reported from countries in Southeast Asia and sub-Saharan Africa that are endemic for hepatitis B virus (HBV) infection. Countries in Southern Europe have medium-high incidence rates, while low-incidence areas (<5 per 100,000) include South and Central America and the rest of Europe. Cirrhosis is present in about 80% to 90% of HCC patients and is thereby the largest single risk factor. Main risk factors include HBV, hepatitis C virus (HCV), aflatoxin, and possibly obesity and diabetes.

1340

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 44-44  Pseudolymphoma of the liver. A, Precontrast T1-weighted MRI shows a homogeneously hypointense mass (arrow). It demonstrates homogeneous enhancement (arrow) in the arterial phase (B) and washout (arrow) in the equilibrium phase (C) of dynamic contrast-enhanced T1-weighted MRI.

Together HBV and HCV account for 80% to 90% of all HCC worldwide. HBV continues to be the major HCC risk factor worldwide, although its importance will most likely decrease during the coming decades owing to widespread use of the HBV vaccine in newborns. HCV has been the dominant viral cause of HCC in North America, some Western countries, and Japan. Obesity and diabetes are increasing at a fast pace throughout the world, and if they are proven to be HCC risk factors, they would account for more HCC cases in the future.226 In endemic regions of HBV/HCV infection, it has become possible to detect small early-stage HCC in these high-risk patients. Periodic ultrasonography (every 3-6 months) allows early HCC detection, and dynamic CT or MRI permits an accurate assessment of the tumor in developed countries with a high incidence of HCC. Around 80% of newly diagnosed HCCs in Japan now measure less than 5 cm in diameter.90 Disease progression, classification, and presentation.  Two types of human hepatocarcinogenesis are now recognized: de novo carcinogenesis and multistep carcinogenesis in a liver with cirrhosis or chronic hepatitis. In the latter case, the progression is as follows: dysplastic nodule (DN), early HCC, well-differentiated HCC, moderately or poorly differentiated HCC.93,142,215,255 DNs are classified as low grade or high grade on the basis of cytologic and architectural atypia; highgrade DNs show varying degrees of one or both types of atypia. Early HCCs are indistinctly nodular and highly differentiated neoplasms. Although the pathologic diagnosis of high-grade DN and early HCC is controversial, the presence of tumor cell invasion of the intratumoral portal tracts (stromal invasion) is a helpful clue for differentiating the two entities.142,166,215,255 The gross features of HCC are classified into five major types: (1) small nodular with indistinct margin, (2) simple nodular, (3) simple nodular with extranodular growth, (4) confluent multinodular, and (5) infiltrative.166 The small nodular type with indistinct margin usually corresponds to early HCC. In advanced lesions, cancer cells often invade the portal veins through the blood drainage route from the tumor308 and then into the hepatic veins, forming tumor thrombi.

Microscopic tumor thrombi may be seen even in small HCCs, especially in the simple nodular type with extranodular growth and the confluent multinodular type.218 Patients with small localized tumors usually have no HCC-related symptoms. In the advanced terminal stage, hepatomegaly, upper abdominal mass, abdominal pain, general malaise, anorexia, abdominal fullness, weight loss, jaundice, ascites, edema, and gastrointestinal bleeding are commonly present. Intraperitoneal bleeding (rupture) is one of the most serious complications of advanced tumors, although it may occur in smaller lesions as well.34,115 Imaging of multistep hepatocarcinogenesis.  Ultrasound and CT features of DNs and early HCCs are similar, and a precise differential diagnosis is impossible. Therefore for the early diagnosis  of HCC, it is important to understand the concept of multistep  hepatocarcinogenesis and the sequential imaging changes that occur.48,80,176,179,284,286 The intranodular blood supply evaluated by CTAP and CT during hepatic arteriography (CTHA) changes with the progression from DN to overt HCC. The intranodular portal supply relative to the surrounding liver parenchyma observed by CTAP is decreased, whereas the intranodular arterial supply revealed by CTHA is initially decreased during the early stage of hepatocarcinogenesis and is then increased in parallel with increasing grade of malignancy of the nodules.80,179 Highgrade DN typically shows isoattenuation on CTAP (almost the same portal supply) and hypoattenuation on CTHA (decreased arterial supply) (Fig. 44-45); early HCC shows slight hypoattenuation on CTAP (decreased but not absent portal supply) and isoattenuation on CTHA (almost the same arterial supply) (Fig. 44-46); well-differentiated HCC shows partial definite hypoattenuation on CTAP (partially absent portal supply) and partial or slight hyperattenuation on CTHA (partially increased arterial supply) (Fig. 44-47); and moderately differentiated HCC shows definite hypoattenuation on CTAP (entirely absent portal supply) and hyperattenuation on CTHA (increased arterial supply) (see Fig. 44-45). Histologically the portal tracts, including the portal vein and hepatic artery, are decreased in accordance with highergrade malignancy and are virtually absent in moderately and poorly

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D

E

F

G FIG 44-45  High-grade dysplastic nodule and moderately differentiated HCC. A to C, Precontrast CT scan shows a high-grade dysplastic nodule (arrow in A) and a moderately differentiated HCC (arrowhead in A) as hypodense nodules. B, In the arterial phase the dysplastic nodule demonstrates no enhancement (arrow) and the HCC demonstrates definite enhancement (arrowhead). C, In the equilibrium phase, both appear hypodense. D, T1-weighted MRI shows the dysplastic nodule as markedly hyperintense (arrow) and the HCC as slightly hyperintense (arrowhead). E, On the T2-weighted image the nodule is hypointense (arrow) and the carcinoma is hyperintense (arrowhead). F, On the SPIO-enhanced T2-weighted image, only the HCC can be identified as a hyperintense nodule (arrowhead). G, The diffusion-weighted image visualizes the nodule as slightly hyperintense (arrow) and the carcinoma as definitely hyperintense (arrowhead). Continued

1341

1342

PART II  CT and MR Imaging of the Whole Body

I

H

FIG 44-45, cont’d  H, CT during hepatic arteriography demonstrates a definitely hyperdense HCC (arrowhead) with a markedly increased arterial supply and a hypodense dysplastic nodule (arrow) with a decreased arterial supply. I, On CTAP the nodule (arrow) is not well identified (with almost the same portal supply relative to the surrounding liver). In contrast the carcinoma (arrowhead) shows definite hypodensity (absence of portal supply).

A

B FIG 44-46  Early HCC. A, On CT during hepatic arteriography the nodule is not well identified (arrow), indicating almost the same intranodular arterial supply relative to the surrounding liver. B, CTAP demonstrates a slightly hypodense nodule (arrow), indicating a decreased but not absent intranodular portal supply.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D FIG 44-47  Well-differentiated HCC (borderline lesion with a definitely malignant focus). A, CT during hepatic arteriography shows a hypodense nodule with an internal hyperdense focus (arrow), indicating a partially increased intranodular arterial supply. B, CTAP demonstrates an isodense nodule with a definitely hypodense focus (arrow), indicating a partially absent intranodular portal supply. C, Part of the lesion demonstrates enhancement (arrow) in the arterial phase. D, Equilibrium phase of dynamic CT shows a hypodense nodule (arrow). Continued

1343

1344

E

PART II  CT and MR Imaging of the Whole Body

F

G

FIG 44-47, cont’d  E, T1-weighted MRI shows a hyperintense nodule with a relatively hypointense internal focus (arrow). F, On the T2-weighted image it is visualized as a hypointense nodule with a hyperintense internal focus (arrow). G, On the SPIO-enhanced image a hyperintense focus is seen (arrow).

differentiated HCC. In contrast, abnormal arteries due to tumor angiogenesis develop in high-grade DNs during the course of hepatocarcinogenesis and are markedly increased in number in moderately differentiated HCC211,217,310 (Fig. 44-48). Drainage vessels from the tumors also change during multistep hepatocarcinogenesis, namely from hepatic vein (DN, early HCC), hepatic sinusoids, and portal venules surrounding the tumor (hypervascular well/moderately differentiated HCC) to portal venules in a pseudocapsule (moderately differentiated HCC with pseudocapsule).134 Because of the drainage flow through these routes, hypervascular HCC appears as a peritumoral ring or coronal enhancement on late-phase dynamic CT/MRI.324 DNs or early HCCs usually show no definite staining in the arterial phase of dynamic CT or MRI (see Fig. 44-45).285 However, when a small focus of enhancement is demonstrated in a hypovascular nodule, it should be treated as overtly malignant because almost all such nodules progress to entirely hypervascular HCC (see Fig. 44-47).81 MRI signal intensity is also useful to estimate the grade of malignancy of hepatocellular nodules.178 DNs typically show hypointensity relative to the surrounding cirrhotic liver on T2-weighted images (see Fig. 44-45). Conversely, almost all moderately differentiated HCCs demonstrate hyperintensity (see Fig. 44-45), and early and welldifferentiated HCCs have a strong tendency to show isointensity (Fig. 44-49). In contrast, almost all DNs and early HCCs demonstrate hyperintensity on T1-weighted images (see Figs. 44-45, 44-47, and 44-49).21,44,48,159,176,178,314 It is unknown why borderline lesions or early HCCs show hypointensity on T2-weighted images, but it is probably due to a decrease in the number of intranodular sinusoids induced by hyperplastic changes of hepatocytes.269 Siderotic DNs may be one cause of hypointensity on T2-weighted images191,288; however, their frequency is not high. Malignant foci in hypointense borderline lesions can be seen as more hyperintense spots (see Fig. 44-47).69,178 Fat deposition in the tumor is one of the major causes of hyperintensity on T1-weighted images, and copper accumulation or decreased sinusoids may be another.45,110,159,178

Hepatic arterial supply Portal venous supply Abnormal arterial supply

High-grade DN

RN

Low-grade DN

Arterial supply

Moderately Well- differentiated Early differentiated HCC HCC HCC Portal supply

FIG 44-48  Stepwise changes of intranodular blood supply during multistep hepatocarcinogenesis. DN, dysplastic nodule; HCC, hepatocellular carcinoma; RN, regenerative nodule.

SPIO-enhanced MRI reflects the number of Kupffer cells in HCCs and DNs and is useful for estimating the histologic grade in HCCs, although uncertainties persist in differentiating DNs from welldifferentiated HCCs.91,117,163 SPIO typically accumulates in DNs and in early and some well-differentiated HCCs (see Figs. 44-45, 44-47, and 44-49). Conversely, SPIO is not taken up by moderately and poorly differentiated HCCs (Fig. 33-50). On hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI, the degree of EOB uptake in the hepatocellular nodules reflects cellular expression of OATP1B3, the membranous uptake transporter of EOB. OATP1B3 expression significantly decreases during multistep hepatocarcinogenesis (Fig. 44-51; see Fig. 44-50), so hepatobiliary-phase Gd-EOBDTPA–enhanced MRI is also useful for estimating the grade of malignancy of hepatocellular nodules in cirrhotic liver.135,132 Early hepatocellular carcinoma.  On gross pathologic examination, early HCC is distinguished with difficulty from the surrounding hepatic parenchyma. On microscopic examination, early HCC is a

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

D

B

E

C

F

FIG 44-49  Early HCC. A to C, Arterial phase of dynamic CT (B) demonstrates no definite enhancement of the lesion (arrow). On precontrast (A) and postcontrast (C) CT scans, it is shown as a nonspecific hypodense nodule (arrows). The margin of the tumor is indistinct. D to F, The lesion is visualized as a hyperintense nodule (arrow) on the T1-weighted image (D), as isointense (arrow) on the T2-weighted image (E), and as hypointense (arrow) on the SPIO-enhanced image (F).

1345

1346

PART II  CT and MR Imaging of the Whole Body

Well-differentiated HCC

Dysplastic nodule EOB uptake OATP1B3

EOB uptake OATP1B3

Moderately differentiated HCC EOB uptake OATP1B3

Poorly differentiated HCC EOB uptake OATP1B3

EOB uptake About 15%

OATP1B3

FIG 44-50  Relationship between multistep hepatocarcinogenesis and OATP1B3 expression/signal intensity of hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI of hepatocellular nodules.

well-differentiated neoplasm usually consisting of a relatively uniform population of small cells with nuclear atypia and a high nucleus-tocytoplasm ratio; these cells form irregular thin trabeculae and pseudoglandular structures.48,92,142,166 Imaging features of early HCC include an indistinct margin, hypoattenuation or isoattenuation on dynamic CT and MRI,  decreased but not absent intranodular portal supply on CTAP (see  Fig. 44-46), hyperintensity on T1- and isointensity or hypointensity  on T2-weighted images (see Fig. 44-49), occasional fatty changes,  and a nodule-in-nodule appearance (hypervascular focus on  dynamic CT or MRI and hyperintense focus on T2 images) when hypervascular less-differentiated HCC emerges within early HCC  (see Fig. 44-47). SPIO usually accumulates in early HCC (see  Fig. 44-49).48,80,91,176,178,205,269,285,286 Hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI can visualize the majority of early HCCs (probably ≈ 85%) as a hypointense nodule and drastically improved the detectability of hypovascular pathologic early HCC.132,258 Because around 30% of DNs may also show slight hypointensity,132 precise differentiation between early HCC and DN is still difficult, but when a hypovascular hypointense nodule is depicted in high-risk cirrhotic liver, the possibility of early HCC is extremely high.258 Classic hypervascular hepatocellular carcinoma.  Microscopically the majority of classic hypervascular HCCs are moderately differentiated. Macroscopically they are the simple nodular type (Fig. 44-52), simple nodular type with extranodular growth (Fig. 44-53), confluent multinodular type (Fig. 44-54), or infiltrative type166 (Fig. 44-55). The nodular type is a well-demarcated and often encapsulated nodule; its macroscopic features consist of a fibrous capsule and an internal mosaic architecture separated by fibrous septa (see Fig. 44-52). In larger tumors, internal necrosis, hemorrhage, and scarlike changes can be seen. In advanced lesions, cancer cells often invade the portal and hepatic veins, forming tumor thrombi. Histologic variants include fatty changes, prominent bile production, sarcomatous change, scirrhous change, intra–bile duct growth, and copper accumulation.

Nonenhanced CT usually reveals a hypodense area (see Figs. 44-52 and 44-55); a definitely hypodense area (total or partial) due to the presence of fatty changes (Fig. 44-56) may be seen. Calcification is rare. When copper accumulation or hemorrhage is associated, slight hyperdensity can be seen.131 In the arterial phase of dynamic CT, enhancement of the entire lesion is seen, and its intensity differs in accordance with its mosaic architecture; this represents a distinctive finding of HCC (see Fig. 44-52). However, when the lesions are small or predominantly well differentiated, they may appear hypoattenuating or isoattenuating even in the arterial phase. In this case, CTHA or contrast ultrasonography may be valuable for visualizing the increased arterial vascularity (see Fig. 44-47). Drainage flow from these HCCs into the surrounding hepatic sinusoids is through the connections between intratumoral blood sinusoids and hepatic sinusoids or tiny portal venules in the pseudocapsule308 (Fig. 44-57). This drainage flow results in “corona enhancement” surrounding the tumor in the arterial or portal phase of dynamic CT (see Fig. 44-52) and is a very useful imaging feature for differentiating benign hypervascular hepatic nodules such as cavernous hemangioma, AP shunting, FNH, angiomyolipoma, and peliosis hepatis, which seldom show this kind of corona enhancement.193,307,308 When poorly differentiated HCC is predominant in the nodule, it commonly shows relative hypovascularity with infiltrative growth (see Fig. 44-55). On T1-weighted images, classic HCC usually shows hypointensity. However, about one third of nodular-type HCCs demonstrate hyperintensity relative to the surrounding liver on T1 (Fig. 44-58; see Fig. 44-56). Almost all classic hypervascular HCCs show hyperintensity on T2-weighted images (see Figs. 44-45 and 44-56). This combination is unique among hepatic malignant tumors except for metastatic malignant melanoma; it is almost diagnostic for HCC, especially in high-risk patients. The histologic background of HCCs with hyperintensity on T1-weighted images is fatty metamorphosis in one third of cases, with copper accumulation, glycogen, and high-grade cancer cell differentiation other possible causes.45,110 The different signal intensities of this internal mosaic pattern and the various components are particularly

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D FIG 44-51  Hepatobiliary-phase images of hepatocellular nodules during hepatocarcinogenesis step. On T1-weighted images obtained 20 minutes after administration of Gd-EOB-DTPA, dysplastic nodule appears isointense compared to background liver parenchyma (A), early HCC appears slightly hypointense compared to background liver (B), moderately differentiated HCC usually appears hypointense compared to background liver (C), but about 15% of moderately differentiated HCC appears iso- or hyperintense compared to background liver (D).

1347

1348

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 44-52  Simple nodular (moderately differentiated) HCC. A, The arterial phase of dynamic CT shows a definitely enhanced tumor with a circular hypodense fibrous capsule and internal mosaic architecture separated by hypodense fibrous septa (arrow). B, In the portal phase the contour of the tumor becomes vague because of the drainage flow from the tumor itself (arrow). C, In the equilibrium phase the tumor is hypodense because of the washout of contrast medium; delayed enhancement of the capsule can be seen (arrow).

FIG 44-53  Simple nodular HCC with extranodular growth. The SPIOenhanced T2-weighted image shows nodular-type HCC with extranodular growth of part of the tumor (arrow).

A

B

FIG 44-54  Confluent multinodular HCC. The arterial phase of dynamic CT demonstrates a confluent multinodular type of HCC (arrow). Multiple intrahepatic metastases are also shown.

C

FIG 44-55  Infiltrative (poorly differentiated) HCC. Dynamic CT visualizes a faintly enhanced mass with unclear margins (arrow) in the arterial phase (B). Internal heterogeneity, including necrosis, is also shown on precontrast (A) and postcontrast (C) CT scans (arrow).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D

E

F

1349

FIG 44-56  HCC with marked fatty metamorphosis. A to C, Precontrast CT scans show a markedly hypodense mass (arrow in A). Arterial-phase (B) and equilibrium-phase (C) CT scans demonstrate enhancement with mosaic architecture (arrow). D, In-phase T1-weighted MRI demonstrates a definitely hyperintense mass (arrow). E, Opposed-phase T1-weighted MRI shows marked signal loss, indicating fat deposition in the mass (arrow). F, On the T2-weighted image the internal mosaic architecture and hyperintense capsule are well visualized (arrow).

well depicted on T2-weighted images and dynamic MRI (see Figs. 44-56 and 44-58). The fibrous capsule is demonstrated as a hypodense rim surrounding the tumor on nonenhanced CT; it is hyperdense on postcontrast CT, hypointense on T1-weighted images, and hyperintense on both T2-weighted and postcontrast MRIs. The internal mosaic architecture is visualized as components of variable density or intensity separated by fibrous septa, with imaging features similar to those of the fibrous capsule (see Figs. 44-52, 44-56, and 44-58). However, in tumors less than 2 cm in diameter, the capsule and mosaic architecture are seen in about 50% of cases, are poorly developed on macroscopic examina-

tion, and are visualized on CT or MRI in less than 20% of cases.110 Therefore in the diagnosis of small HCCs, various kinds of small hypervascular nodules should be differentiated. On MRI, SPIO and hepatobiliary agents such as Gd-EOB-DTPA do not accumulate in classic hypervascular HCCs (see Figs. 44-45 and 44-53). However, they may be taken up by some well-differentiated HCCs, and special care is required to detect these lesions and make the differential diagnosis.91 About 15% of classic hypervascular HCCs express membranous transporter OATP1B3 and uptake Gd-EOBDTPA.132 This type of HCC appears iso- to hyperintense on hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI (see Figs. 44-50

1350

PART II  CT and MR Imaging of the Whole Body

A

D

Pre

B

14s

E

5s

23s

9s

C

40s

F

FIG 44-57  Single-level dynamic during hepatic arteriography in a patient with hypervascular HCC (measuring 1.9 × 1.7 cm). A, Precontrast. B, At 5 seconds after injection of contrast material, the contour of the tumor is clearly visible. C, At 9 seconds, the adjacent liver, containing a bright portal venule (drainage area; arrows), starts to stain. The tumor is more hyperdense at this moment. D, At 14 seconds, both the tumor and the drainage area are stained and isodense compared with each other. At 23 seconds (E) and 40 seconds (F), contrast material is washed out from the tumor and the corona stain suggesting the drainage area appears. The corona stain has a thick area in the right upper part of the tumor (arrows in E). (From Ueda K et al: Hypervascular hepatocellular carcinoma: Evaluation of hemodynamics with dynamic CT during hepatic arteriography. Radiology 206:161–166, 1998.)

A

B

C

FIG 44-58  Moderately differentiated HCC. A, Precontrast T1-weighted image shows a slightly hyperintense nodule with a hypointense thin capsule surrounding the tumor (arrow). B, The tumor is well enhanced, with an internal mosaic pattern (arrow) in the arterial phase. C, In the equilibrium phase it demonstrates hypointensity owing to washout of contrast medium, with an enhanced hyperintense capsule (arrow).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

1351

B FIG 44-59  Portal vein tumor thrombus. A, The arterial phase of dynamic CT demonstrates multiple wellenhanced HCCs (arrowheads) and marked enhancement showing a thread- and streak-like pattern in the proximal right portal vein (arrows). B, Diffusely enhanced tumor thrombus is visualized in the equilibrium phase (arrows).

and 44-51). Hyperintense HCC on hepatobiliary-phase Gd-EOBDTPA–enhanced MRI tends to show less aggressive features than hypointense HCC.133 Tumor thrombus—whether in the main portal vein, the first and second branches of the intrahepatic portal vein, or the major branches of the hepatic vein—is visualized as an intravascular solid mass lesion. Differentiation from blood thrombus is usually possible by enhancement on contrast-enhanced CT or MRI. AP shunting with the thread and streak signs can be seen on dynamic CT (Figs. 44-59 and 44-60). As described earlier the area with obstructed portal vein shows segmental staining on dynamic CT and MRI.100,184 This is an important sign of peripheral portal vein invasion by HCC. On CTAP a wedgeshaped portal perfusion defect surrounding the tumor usually indicates the existence of portal tumor thrombus.177

Hepatocellular carcinoma with atypical histologic features.  Various types of HCCs with atypical histologic features have

atypical imaging findings. It is well known that HCC shows various degrees of fatty metamorphosis, especially in small well-differentiated HCCs. It is usually diffuse in small HCCs and focal in larger tumors. Fatty tissues are visualized as echogenic on ultrasonography, hypodense on CT, and hyperintense on T1-weighted images. When fatty metamorphosis is severe, CT shows an attenuation value of less than 10 Hounsfield units (HU) in the tumor, making a specific diagnosis of the fat component possible (see Fig. 44-56). For the specific detection of fatty deposition, chemical shift MRI is useful.173,242,339 HCCs with abundant fibrous stroma separating cords of tumor cells are called scirrhous type according to the World Health Organization (WHO) classification. This type of HCC demonstrates hypovascularity in the arterial phase and delayed enhancement in the delayed phase of dynamic CT and MRI. These imaging features are commonly seen in other mass lesions with abundant fibrous tissue (e.g., confluent hepatic fibrosis, granuloma, cholangiocarcinoma)65,149,337 (Fig. 44-61). Spontaneous massive necrosis can occur in small HCCs with a fibrous capsule. It is demonstrated as a nonenhanced cystic mass with an irregular wall on postcontrast CT or MRI. Intratumoral hemorrhage is common histologically in hypervascular HCCs. When it is relatively massive, intratumoral or peritumoral hyperdensity is shown on nonenhanced CT. A specific diagnosis of hemorrhage may be possible with MRI. On dynamic CT a pseudoan-

FIG 44-60  Hepatic vein tumor thrombus. Coronal postcontrast CT scan shows a giant HCC in the right lobe (arrowheads), with extension into the hepatic vein and inferior vena cava (arrow).

eurysm may be visualized. Intraperitoneal hemorrhage due to ruptured HCC is usually associated with a subcapsular HCC. Copper accumulation is seen in 20% to 40% of HCCs. A small HCC with marked copper accumulation may show hyperdensity (50-60 HU) on precontrast CT, which is probably attributable to the abundant copper-binding protein in cancer cells. On T1-weighted images this kind of HCC demonstrates hyperintensity.45,131 HCCs with sarcomatous changes have been reported sporadically. About 4% of HCC autopsy cases show varying degrees of sarcomatous change. They are usually hypodense or avascular in the arterial phase of dynamic CT or MRI84,331 (Fig. 44-62).

1352

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 44-61  Scirrhous-type HCC. A, The arterial phase of dynamic CT shows marked enhancement in the peripheral part of the tumor. B, Delayed central enhancement is observed in the equilibrium phase, which represents the presence of abundant fibrous stroma. C, T2-weighted MRI shows hyperintense mass with central hypointense area. D, On DWI the peripheral part of the tumor appears hyperintense and the central fibrous part is seen.

A

B FIG 44-62  HCC with sarcomatous change. A, The arterial phase of dynamic CT shows a markedly hypovascular mass with a peripheral enhanced rim (arrow). B, In the equilibrium phase the entire tumor demonstrates no detectable enhancement (arrow). Histologically, no definite necrosis was revealed.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions Cholangiocellular

FIG 44-63  HCC with intra–bile-duct growth. Coronal postcontrast CT reveals castlike tumor extension in the left hepatic duct, with dilated proximal intrahepatic bile ducts (arrow).

Tumor growth into the bile ducts is found in 6% of autopsied HCCs and can even occur with small HCCs. A castlike defect with a smooth margin differentiates it from bile duct cancer66 (Fig. 44-63).

Fibrolamellar Carcinoma.  Fibrolamellar carcinoma develops in normal livers in young adults, with no clear sex predilection. It is a slow-growing malignant tumor made up of eosinophilic neoplastic hepatocytes separated into cords by lamellar fibrous strands. It frequently presents with abdominal pain, which is nonspecific. Tumor markers are usually not elevated, although α-fetoprotein (AFP) may be slightly increased. The 5-year survival rate after surgery is approximately 60%. Most cases are discovered when the tumor becomes massive (>10 cm); it is solitary, well demarcated, and lobulated. Necrosis and hemorrhage are seen in the center. A central fibrous scar like that seen in FNH is sometimes present.253,301 On nonenhanced CT, a hypodense solitary mass is seen; in about 50% of cases, central calcification is present. On dynamic CT inhomogeneous enhancement is observed during the arterial phase. A poorly enhanced hypodense area generally reflects necrosis. In the equilibrium (or delayed) phase, delayed enhancement of the central scar is seen in about 25% of cases. On MRI the signal intensity is variable but is essentially hy­ pointense on T1- and hyperintense on T2-weighted images. The central scar characteristically shows hypointensity on both T1- and T2-weighted images. SPIO and hepatobiliary agents do not accumulate. The findings on dynamic MRI are the same as those on CT but are more sensitively depicted89,187,276,299 (Fig. 44-64). Differentiation from FNH can be based on these findings, although in some cases the findings are similar.75

1353

Carcinoma.  Cholangiocellular carcinoma (CCC) is a malignant tumor that develops from bile duct epithelium. It is classified as either intrahepatic cholangiocarcinoma or hilar cholangiocarcinoma and arises from the small intrahepatic bile ducts or hilar bile ducts, respectively. Here, intrahepatic cholangiocarcinoma is described. It has been reported that CCC occurs in those with hepatolithiasis, clonorchiasis or opisthorchiasis, sclerosing cholangitis, ulcerative colitis, hemochromatosis, α1-antitrypsin deficiency, extrahepatic biliary atresia, or Thorotrast exposure. Recently it has been reported that the occurrence of CCC is also associated with HCV infection or hepatitis C–related cirrhosis.212 CCC is often discovered as a huge tan or white tumor that is hard and contains dense fibrous tissue in its center. Histologically CCC consists of glandular adenocarcinoma with abundant sclerosis, with varying amounts of mucin secretion and necrosis noted. The tumors usually show infiltrative growth and a desmoplastic nature, and the contour of the adjacent liver is dented. Intravascular growth is not common, but the vessels and biliary tree are often encased.124,209,212,232 According to the Liver Cancer Study Group of Japan, CCC is divided into three types: mass forming, periductal infiltrating, and intraductal growing.166 The prognosis of the last is much better after surgical resection. On nonenhanced CT, mass-forming CCC is depicted as a hypodense mass. In the arterial phase of dynamic CT, only the periphery of the tumor where viable tumor cells exist is enhanced; the center of   the tumor where there is abundant fibrous stroma remains nonenhanced. In the equilibrium phase the center of the tumor may be enhanced (delayed enhancement). In the arterial phase the adjacent liver often shows wedge-shaped enhancement (Fig. 44-65) that reflects regional portal or hepatic venous obstruction. Dilatation of the peripheral intrahepatic bile ducts also reflects obstruction of the biliary  tree by the tumor (see Fig. 44-65). The desmoplasia in the tumor retracts the adjacent liver and depression of the hepatic contour is  seen15,126,150,162,252,348 (Fig. 44-66). On MRI the tumor is seen as hypointense on T1- and hyperintense on T2-weighted images, with an irregular contour (see Figs. 44-65 and 44-66). The center of the tumor is more hypointense than the periphery on T2, reflecting abundant fibrous stroma. Strongly hyperintense areas on T2-weighted images reflect areas with marked mucin secretion or necrosis.169,318 When the tumor infiltrates an intrahepatic bile duct branch, regional cholestasis results, and a hyperintense segment containing the tumor may be seen on T1-weighted images and even more clearly on fat-suppressed T1-weighted images (see Fig. 44-65).64 The findings on dynamic MRI are similar to those on dynamic CT, although more marked early peripheral and delayed central enhancement is seen on dynamic MRI (see Fig. 44-66). Hypervascularity on dynamic CT and MRI can also be demonstrated in some types of CCC with less internal sclerosis. Periductal-infiltrating type and intraductalgrowing type CCC, including intraductal papillary neoplasm of the bile duct (IPNB), are described in Chapter 42. Mucinous carcinoma is a variant of CCC. Histopathologically the tumor consists of myxomatous matrix in which tumor cells are scattered. On ultrasonography the tumor is hyperechoic. On nonenhanced CT the tumor is depicted as a definitely hypodense mass. The density of the tumor is slightly higher than that of water. During the arterial phase of dynamic CT, the periphery of the tumor demonstrates enhancement that gradually spreads toward the center during the portal and equilibrium phases. It shows marked hypointensity on T1and marked hyperintensity on T2-weighted images. Dynamic MRI findings are similar to those on dynamic CT but are more prominent; in the delayed phase when the tumor is relatively small, diffuse and definite enhancement of the entire tumor can be seen79,309 (Fig. 44-67).

1354

A

PART II  CT and MR Imaging of the Whole Body

B

C

E

D

FIG 44-64  Fibrolamellar carcinoma in a 24-year-old man. A, Precontrast CT demonstrates an inhomogeneous hypodense mass with punctate calcifications within a central scar (arrow). B, Arterial-phase CT shows definite enhancement, with dilated feeding arteries except for the central scar and necrotic portions (arrow). C, Nonnecrotic portions and the central scar exhibit delayed enhancement in the equilibrium phase (arrow). D, T1-weighted image demonstrates an inhomogeneous hypointense mass with a more hypointense central scar (arrow). E, On the T2-weighted image the mass shows inhomogeneous hyperintensity and the scar remains hypointense (arrow). Arrowheads in D and E indicate internal hemorrhagic necrosis. (Courtesy M. Satake, National Cancer Center East, Japan.)

These findings mimic those of cavernous hemangioma; the arterial phase of dynamic CT or MRI should be used for differentiation.309 Metastasis from mucinous carcinoma of the colon shows similar features.

Combined Hepatocellular and Cholangiocellular Carcinoma.  Combined hepatocellular and cholangiocellular carcinoma is defined as a tumor containing unequivocal intimately mixed elements of both HCC and CCC.297 It has areas of HCC with trabecular growth as well as areas of gland formation with fibrous stroma and mucin production. Combined HCC and CCC is considered a subtype of HCC.282 A hepatic stem cell origin of this tumor is now suspected because foci of intermediate morphology are seen at the interface of HCC and CCC in many cases.77,298 These with stem cell features are subdivided into typical, intermediate-cell, and cholangiolocellular subtypes (WHO classification).297

Both typical HCC areas and typical CCC areas are found within a single tumor. The HCC areas are hypervascular and show rapid runoff of contrast material. In contrast, the CCC areas are hypovascular and show delayed enhancement in the equivalent phase of postcontrast CT or MRI.10,46,57 When the HCC and CCC areas are clearly separated, the correct diagnosis can be made. However, the imaging findings of those with stem cell features are variable. Distinguishing poorly differentiated HCC, scirrhous-type HCC, CCC, and metastatic adenocarcinoma may be difficult.

Cholangiolocellular Carcinoma.  Cholangiolocellular carcinoma was first described by Steiner and defined as a tumor that originates from the Hering duct or cholangiole of the liver.277 The Hering duct is located at the connection of bile canaliculi, which is composed of hepatocytes, and the cholangiole, which is the first component of the intrahepatic bile duct. Thus the component cells of the Hering duct

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

B

A

D

C

E FIG 44-65  Mass-forming type of cholangiocellular carcinoma (intrahepatic cholangiocarcinoma) with cholestasis and Zhan’s infarction of the left lobe of the liver. A, Precontrast CT scans demonstrate a hypodense mass (arrow) with dilated intrahepatic bile ducts. B, The arterial phase reveals faint peripheral enhancement of the tumor (arrow) and segmental staining of the left lobe due to associated left portal vein obstruction. C, In the equilibrium phase the lesion shows hypodensity (arrow). Delayed enhancement caused by abundant fibrous tissue in the tumor is not prominent in this case. D, T1-weighted MRI demonstrates a hypointense nodule (arrow) and segmental hyperintensity in the left lobe due to cholestasis. E, T2-weighted MRI shows the nodule as hyperintense (arrow). The diffuse hypodensity on precontrast CT and hyperintensity on T2-weighted MRI in the left lobe may indicate Zhan’s infarction due to portal vein obstruction.

1355

1356

PART II  CT and MR Imaging of the Whole Body

B

A

C

D

FIG 44-66  Mass-forming type of cholangiocellular carcinoma. A, T1-weighted MRI demonstrates a hypointense mass with depression of the adjacent liver surface (arrow). B, T2-weighted MRI demonstrates a hyperintense mass with markedly hyperintense central necrosis (arrow). C and D, Dynamic contrast-enhanced T1-weighted MRI reveals peripheral enhancement in the arterial phase (C) (arrow) and internal delayed enhancement in the equilibrium phase (D) (arrow).

are thought to have intermediate features between hepatocyte and bile duct epithelium. Histopathologically, cholangiolocellular carcinoma tends to show a replacing growth pattern that is not common in typical CCC but is common in HCC. The concept of “bile ductular carcinoma” was proposed recently.146 In the new WHO classification, this is included in a subtype of combined HCC and CCC.297 Based on our experience, cholangiolocellular carcinoma is hypodense on precontrast CT, shows slight enhancement with a less-enhancing central part in the arterial phase of dynamic CT, and shows delayed enhancement in the central part of the tumor in the equilibrium phase (Fig. 44-68). Intratumoral portal tracts can often be seen.14,203

Biliary Cystadenoma and Cystadenocarcinoma.  Biliary cystadenoma and cystadenocarcinoma are rare solitary multilocular cystic masses that resemble mucinous cystic tumors of the pancreas. Cystadenoma is a large cystic neoplasm with internal septation and papillary projections. The locules are lined with columnar epithelium, and benignity and malignancy cannot be grossly differentiated. Papillary projections are more frequently present in malignant cases. Adenoma and adenocarcinoma may be intermingled in some cases, and

cystadenoma is considered a premalignant lesion. The locules contain turbid, purulent, or bloody mucin. Ovarian-like stroma is usually found in females.28,60,96,144,199,346 The locules of hepatobiliary cystadenoma and cystadenocarcinoma seldom communicate with the biliary tree. When such a communication exists, the distal and proximal biliary tree is dilated because of mucin. This type of tumor can be distinguished from the common cystadenoma or cystadenocarcinoma and is referred to as a mucin-hypersecreting bile duct tumor (or cholangiocarcinoma). It is considered the intrahepatic counterpart of intrapancreatic mucinous neoplasms.9,164,346 On imaging, biliary cystadenoma or adenocarcinoma appears as a unilocular or multilocular cystic mass with a thin wall and papillary projections. On nonenhanced CT the tumor is a uniformly hypodense mass. The density of the contents of the cyst is often greater than  that of simple hepatic or renal cysts. Focal calcification of the cyst wall may be seen. On postcontrast CT the wall and septa of the tumor,  as well as the papillary projections, are enhanced28,96,144,199 (Figs. 44-69 and 44-70). On T1-weighted images the tumor appears as a hypointense cystic mass. The intensity of the content of each locule may differ and is

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

A

B

C

D FIG 44-67  Mucinous carcinoma (a variant of cholangiocellular carcinoma). A to C, Precontrast CT shows a definitely hypodense mass (arrow in A). It demonstrates gradual enhancement (arrow) from the periphery of the tumor through the arterial phase (B) to the equilibrium phase (C). D, T2-weighted MRI demonstrates the lesion as a definitely hyperintense mass (arrow) because of abundant intratumoral mucin.

A

B

C

FIG 44-68  Cholangiolocellular carcinoma (bile ductular carcinoma). A, Precontrast CT shows a hypodense mass (arrow). B, Postcontrast CT demonstrates slight enhancement, more prominent in the periphery in the arterial phase (arrow), and delayed enhancement in the central part of the tumor in the equilibrium phase. Dilated bile ducts are observed within the tumor, which represents the presence of intratumoral portal tracts (C) (arrow).

1357

1358

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 44-69  Cystadenoma with ovarian stroma in a 52-year-old woman. A, Arterial phase of dynamic CT shows a large cystic mass (arrow) with a small internal daughter cyst (arrowhead) projecting from the wall of the main cyst. Cyst walls are well enhanced, and minute enhanced projections are seen in the daughter cyst. B, On T1-weighted MRI the main cyst demonstrates hyperintensity comparable to mucin with a high protein content (arrow). In contrast the daughter cyst shows hypointensity similar to cerebrospinal fluid (arrowhead). C, On T2-weighted MRI, both cysts demonstrate definite hyperintensity (arrow and arrowhead).

higher than that of cerebrospinal fluid. On T2-weighted images the intensity of the content is very high but varies depending on the nature of the fluid.60,158 For differentiation from cystic intraductal papillary neoplasm of the bile duct, visualization of the communication between the dilated bile duct and the surrounding liver, including proximal intrahepatic bile ducts, is valuable.9,164,346

Hepatoblastoma.  Hepatoblastoma is a malignant hepatocellular

FIG 44-70  Cystadenocarcinoma with ovarian stroma in a 65-year-old woman. Postcontrast CT shows a large cystic mass with enhanced papillary projections from the wall (arrows).

tumor of childhood. It seldom occurs after age 5 years and is the most common hepatic tumor in children younger than 5. Three quarters of patients are male. Like other tumors of childhood, genetic abnormalities are related to development of the disease. It is known that patients with hepatoblastoma also have Beckwith-Wiedemann syndrome, familial adenomatous polyposis coli, cardiovascular anomalies, diaphragmatic defects, malrotation of the intestine, anomalies of the urinary tract, and Down’s syndrome. Most of these tumors are detected after a parent notices abdominal enlargement. Serum AFP is elevated in most cases. Grossly, hepatoblastomas are often huge and single. The surrounding liver is not cirrhotic. Histopathologically, hepatoblastoma can be classified as either epithelial type or mixed epithelial and

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

1359

B

A

C

D FIG 44-71  Hepatoblastoma in a 3-year-old boy. A, Precontrast CT shows a well-demarcated heterogeneous hypodense mass (arrow). B, There is heterogeneous internal enhancement on postcontrast CT (arrow). The mass (arrow) demonstrates heterogeneous hypointensity on T1- (C) and hyperintensity on T2-weighted MRI (D).

mesenchymal type. The epithelial component is divided into two subtypes: embryonal and fetal.39,241 On nonenhanced CT, hepatoblastoma is a huge hypodense mass with occasional calcification. On dynamic CT it becomes heterogeneously hyperdense in the arterial phase and hypodense in the equilibrium phase. A typical lobulated appearance is more clearly seen after administration of contrast medium. On MRI these tumors show hypointensity on T1- and hyperin­ tensity on T2-weighted images. Hemosiderin deposition associated with intratumoral hemorrhage is sometimes present when hypointense parts are noted on both T2-weighted and long-TE (echo time) gradient echo images. Fibrous septa may be seen on CT or MRI25,39,241 (Fig. 44-71).

Epithelioid Hemangioendothelioma.  Epithelioid hemangioendothelioma (EHE) is a vascular tumor that occurs in the lung, bone, and soft tissue. Most cases are slowly progressive; thus the clinical behavior is considered low-grade malignant. Most cases occur in adults, two thirds of them in women. EHE tends to be located in the subcapsular area of the liver. The surface of the liver is often retracted by the tumor. Marked fibrous tissue and portal tract remnants are seen in the center of the tumor, with an increased number of cells (mainly epithelioid) found in the surrounding tissues. The invasive front of EHE infiltrates hepatic sinusoids and small vessels. In the center of the tumor there is dense hyalinized fibrous tissue. The tumor has abundant stromal

matrix and seems cartilaginous. Calcification is sometimes seen. The tumors tend to be multiple and often fuse in the subcapsular areas over the course of the disease, resulting in central hypertrophy. These changes are characteristic of EHE and are useful in making the imaging diagnosis.59,171,247,323 Nonenhanced CT depicts hypodense nodules in the subcapsular area of the liver. Some lesions show calcification. In the arterial phase of dynamic CT, no definite enhancement is apparent. On postcontrast CT the tumor shows slight enhancement and becomes somewhat indistinct. The contour of the liver is retracted by the tumor59,189,247 (Fig. 44-72). On MRI a variety of signal intensities are reported. On T1-weighted images EHE usually demonstrates hypointensity. On T2-weighted images moderate hyperintensity is shown, with a more hyperintense core (target sign; see Fig. 44-72). On dynamic MRI the tumor demonstrates a moderate peripheral enhancement and delayed central enhancement. In advanced cases a diffusely coalescent tumor in the subcapsular area and central hypertrophy of the liver are observed (see Fig. 44-72).167,313 In the early stage a solitary nodule or a few nodules constituting a nonspecific solid mass mimicking cholangiocarcinoma and metastatic liver cancer can be seen.

Angiosarcoma.  Angiosarcoma is a rare malignant hepatic tumor but is the most frequently encountered malignant mesenchymal tumor of the liver. It is derived from endothelial cells. Most patients are older

1360

A

PART II  CT and MR Imaging of the Whole Body

B

C

E

D

FIG 44-72  Epithelioid hemangioendothelioma. A, Precontrast CT scan depicts hypodense nodules in the subcapsular area of the liver (arrows). Coalescent masses are also shown. B, In the arterial phase, no definite enhancement is apparent. C, On postcontrast CT the tumor shows slight delayed enhancement and some lesions become somewhat indistinct. The contour of the liver is retracted by the tumor. Central hypertrophy of the liver is also evident. D, T1-weighted MRI demonstrates hypointense masses (arrows). E, On T2-weighted MRI they are visualized as hyperintense nodules with more hyperintense cores (arrows).

men who have a history of exposure to toxins such as vinyl chloride monomer, Thorotrast, arsenic, or anabolic steroids. Grossly, angiosarcoma consists of multiple masses spreading throughout the liver. They vary in size and can become massive. In the center of these tumors, necrosis, thrombosis, or hemorrhage may be found. Most cases are diagnosed in the advanced stage, with hepatic failure and intraperitoneal hemorrhage due to rupture seen in many cases. Intrahepatic tumors range from solitary to diffuse.30,145 On nonenhanced CT, angiosarcomas are seen as heterogeneous hypodense masses. Because of hemorrhage, the center may appear hyperdense. In cases with associated intraperitoneal bleeding, a hyperdense area due to acute clot adjacent to the tumor may be noted. On dynamic CT these tumors are well enhanced in the arterial phase, occasionally showing peripheral nodular enhancement and progressive and delayed filling enhancement on the delayed phase. Although the enhancement pattern of the mass is slightly different from the typical enhancement of hemangiomas, when the tumor is small and well demarcated it resembles cavernous hemangioma. However, in many cases these characteristics are absent, and instead, hypovascularity or ringlike enhancement is evident. T1-weighted images show a variety of hyperintense areas because of the presence of hemorrhage in basically hypointense masses. On T2-weighted images marked heterogeneous hyperintensity is seen 

(Fig. 44-73). Fluid-fluid levels and septation may be seen in some cases.82,103,145,239

Undifferentiated Embryonal Sarcoma.  Undifferentiated embryonal sarcoma is a rare malignant neoplasm that occurs in childhood and adolescence. It grows rapidly, and cystic degeneration, intratumoral hemorrhage, or necrosis frequently occurs. Serum AFP is not elevated. Grossly it is a huge mass composed of cystic, gelatinous, red hemorrhagic and yellow necrotic areas, with a pseudocapsule made up of fibrous tissue. Histologically, viable cells are located at the edge of the tumor; the remaining portion is necrotic and accounts for the majority of the tumor. The tumor cells are stellate or spindle shaped with myxoid or fibrous stroma. The resemblance to mesenchymal hamartoma suggests that undifferentiated embryonal sarcoma is its malignant counterpart.29,199 On nonenhanced CT the tumor is depicted as a huge wellcircumscribed mass with clustering of several central cystic and peripheral solid components. The contour of the tumor has clefts and the tumor is lobulated. Postcontrast CT clearly shows the demarcation of the central cystic and peripheral solid areas. On T2-weighted images the cystic components are very hyperintense. On T1-weighted images the tumor is a hypointense mass with hyperintense spots suggesting intratumoral hemorrhage29,199,244,334 (Fig. 44-74).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

B

A

1361

C

E

D

FIG 44-73  Angiosarcoma in a 67-year-old man. A, On precontrast CT, multiple nodules are seen as heterogeneous hypodense masses (arrows). Hyperdense areas are due to internal hemorrhage. Peripheral nodular enhancement in the arterial phase (B) and progressive and delayed filling enhancement in the equilibrium or delayed phase (C) are also visualized. D, On T1-weighted MRI, hyperintense areas due to hemorrhage are evident in basically hypointense masses (arrows). E, On T2-weighted MRI, marked heterogeneous hyperintensity is visualized (arrows).

Primary Hepatic Malignant Lymphoma.  Primary hepatic lymphoma is rare, although the liver is frequently involved by systemic malignant lymphoma. In most cases the cause is unknown, but associations with hepatitis C, hepatitis B, Epstein-Barr virus infection, and immunocompromising conditions have been described. The tumor is usually solitary, and a splenic mass is often associated. On CT it is depicted as a slightly enhanced, homogeneously hypodense nodule.256 Another characteristic feature is infiltrative growth showing hardly any mass effect. When imaging studies reveal piercing of the native portal tracts in the tumor, the possibility of lymphoma must be considered.11 The histologic findings are extremely varied, and in some tumors with extensive fibrous interstitium between tumor cells, postcontrast CT shows delayed enhancement.20 On MRI the tumor is well demarcated and shows homogeneous hypointensity on T1- and homogeneous hyperintensity on T2-weighted images in most cases.322,324 It may be seen along the porta hepatis.94

Metastatic Liver Tumor.  The most common liver metastasis is from the gastrointestinal tract. Adenocarcinoma with an abundant fibrous interstitium (scirrhous cancer) is characterized by the presence of abundant viable cancer cells in the peripheral portion of  the tumor and fibrous interstitium and necrosis in its center. In  most cases of medullary cancer metastases, the center also becomes

necrotic, although small lesions may lack necrotic foci. On both nonenhanced and postcontrast CT, the lesion is depicted as a nonspecific hypodense nodule. In the arterial phase of dynamic CT the peripheral hypercellular portion shows slight ringlike or donut-like enhancement. In the equilibrium phase it becomes hypodense, whereas fibrous necrotic tissue in the center of the tumor shows delayed enhancement65,206,295,296,337 (Fig. 44-75). In liver metastasis from medullary carcinoma, diffuse enhancement of the entire tumor can be seen. The most common hypervascular metastatic liver tumors are those derived from endocrine cancers (e.g., islet cell tumors, renal cell carcinoma, carcinoid).219,275 Metastatic liver cancer appears hypointense on T1- and hyperintense on T2-weighted images. Necrotic changes in the central portion are depicted as more hyperintense on T2-weighted images, and a bull’s-eye pattern may be seen, as on ultrasonography (Fig. 44-76). Contrast-enhanced MRI depicts the findings of dynamic CT more sensitively.65,264 In liver metastases from colon cancer, T2-weighted images may show the center to be hypointense (Fig. 44-77). The histologic background is considered to be fibrosis.234 Liver metastases from colon and stomach mucinous carcinoma often have imaging features similar to those of the mucinous carcinoma variant of CCC (see earlier under “Mucinous Carcinoma”) (Fig. 44-78). Because melanin is a paramagnetic substance, liver metastases from melanoma

1362

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 44-74  Undifferentiated embryonal sarcoma with diffuse myxoid change in a 15-year-old girl. A, Arterial phase of dynamic CT shows a large hypodense mass with internal inhomogeneous enhancement (arrow). B, T1-weighted MRI demonstrates an extremely hypodense mass with internal hyperintense areas due to hemorrhage (arrow). C, T2-weighted MRI visualizes a definitely hyperintense mass with internal septation (arrow). Cystic changes are not seen in this case. (Courtesy Y. Kodama, Hokkaido University, Japan.)

A

B

C

FIG 44-75  Liver metastasis from colon cancer. A, Precontrast CT demonstrates two hypodense nodules (arrows). B, Arterial phase shows peripheral enhancement (arrows). C, In the equilibrium phase, diffuse faint enhancement in the tumors is visualized (arrows).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

1363

typically appear hyperintense on T1- and hypointense on T2-weighted images121 (Fig. 44-79). Metastatic liver cancer forms nodules as a rule, and their distribution and size vary to some extent according to the type of underlying malignancy. For example, liver metastases from pancreatic cancer are

usually small and uniform and are distributed throughout the liver, whereas those from gallbladder cancer tend to cluster in the gallbladder fossa. In breast cancer, micronodules tend to be distributed diffusely. Liver metastases from smooth muscle sarcomas often show prominent liquefaction necrosis in their centers257 (Fig. 44-80). In the case of lymphogenic metastasis, tumor infiltration within the liver and along the portal tracts is sometimes seen; this is observed most frequently with bile duct and gallbladder cancers but also with gastric cancer, pancreatic cancer, and malignant lymphoma. In portal tract infiltration, T2-weighted images show a hyperintense area around the portal vein105,181,281 (Fig. 44-81). Peritoneal dissemination of cancer involves the subcapsular area of the liver, appearing as an intrahepatic mass in some cases. This is often seen in ovarian cancer.287

FIG 44-76  Liver metastasis from colon cancer. T2-weighted image

FIG 44-77  Liver metastasis from colon cancer. T2-weighted image

shows a hyperintense nodule with a more hyperintense central portion (bull’s-eye pattern).

shows a nodule with internal hypointensity and a hyperintense rim (arrow).

A

B FIG 44-78  Mucinous liver metastasis from colon cancer. A, Precontrast CT shows a definitely hypodense mass with a calcific deposit (arrow). B, T2-weighted MRI demonstrates an extremely hyperintense mass (arrow).

1364

PART II  CT and MR Imaging of the Whole Body

A

B FIG 44-79  Multiple liver metastases from malignant melanoma. A, T1-weighted image shows multiple hyperintense nodules (arrows). B, On the T2-weighted image they demonstrate hypointensity (arrows).

FIG 44-81  Pancreatic cancer invading the liver through the portal FIG 44-80  Cystic hepatic metastases from gastrointestinal stromal tumor of the small bowel. Postcontrast CT shows multiple nonenhanced masses with enhanced irregular rims (necrotic masses; arrows).

tracts. T2-weighted image shows periportal extension of the tumor as a hyperintense band along the portal tracts (arrows).

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

1365

B

A FIG 44-82  Moderately to poorly differentiated HCC. A, Postcontrast CT shows a moderately to poorly differentiated mass (arrow). B, PET CT shows a definite accumulation of FDG (arrow).

Miscellaneous Malignant Tumors.  Other primary malignant tumors include carcinoid tumors, other types of sarcomas such as leiomyosarcoma and liposarcoma, and malignant fibrous histiocytoma. These are all very rare.

IMAGING TECHNIQUES FDG PET in the Diagnosis of Liver Tumors There are two major goals when using fluorodeoxyglucose (FDG) positron emission tomography (PET) for liver imaging: to detect and characterize the liver tumor, and to monitor the effect of therapy. Malignant tumor cells generally show an overexpression of glucose transporters on cell membranes, resulting in increased uptake of glucose in tumor cells. For accumulation of FDG in tumor cells to occur, overexpression of hexokinase and lower expression of glucose6-phosphatase are necessary. However, the activity of glucose-6-phosphatase is high in both normal liver and well-differentiated HCC, and FDG PET is not sensitive for the diagnosis of HCC.266 However, Trojan et al. noted that in patients with moderately or poorly differentiated HCC, tumors larger than 5 cm, or markedly elevated AFP levels, FDG PET may contribute to effective noninvasive staging302 (Fig. 44-82). Although limitations include false-positive results in a minority of abscesses and false-negative results in well-differentiated HCCs, PET is useful in estimating the grade of malignancy, tumor staging, detecting recurrence, and monitoring response to therapy for most HCCs.40,165,325 FDG PET is also useful in detecting and staging both hilar and peripheral cholangiocarcinoma.130,250 For detection of colorectal liver metastases, FDG PET has a significantly higher sensitivity on a per-patient basis, but not on a per-lesion basis, than CT and MRI23 (Fig. 44-83). For detection of liver metastases from pancreatic tumors, the diagnostic accuracy of FDG PET is 90%, which is comparable to ultrasonography or CT.210 FDG PET can also be used to evaluate therapeutic interventions. Use of FDG PET in combination with CT at follow-up of radiofrequency ablation of primary and secondary liver tumors may lead to

A

B FIG 44-83  Metastasis from colon cancer. A, Postcontrast CT demonstrates a small metastasis (arrow) and associated hepatic cysts (arrowhead). B, PET CT clearly visualizes the lesion (arrow).

1366

PART II  CT and MR Imaging of the Whole Body

REFERENCES

FIG 44-84  Metastasis from colon cancer. DWI clearly demonstrates multiple liver metastases as hyperintense lesions.

earlier detection of tumor recurrence than use of contrast-enhanced CT alone.24 It is also valuable to evaluate the therapeutic effects of transarterial chemoembolization/radioembolization and sorafenib therapy for HCCs.153,254,258,273 FDG PET has a potential role in evaluating the antitumor effect of chemotherapy in patients with liver metastases from colorectal cancer.52

Diffusion-Weighted Imaging in the Diagnosis of Focal Liver Lesions Based on recent advances in MRI technique, diffusion-weighted imaging (DWI) has been applied to liver examinations.208,342 Detection of HCCs by DWI was around 50% in the analysis of explanted livers,78 because early HCC and well-differentiated HCC are not well visualized on DWI images. However, apparent diffusion coefficient (ADC) measurement with DWI may help estimate the grade of malignancy of hepatocellular nodules during hepatocarcinogenesis. ADCs tend to decrease as the grade of malignancy of the nodules rise from early to poorly differentiated HCC, but with significant overlaps.221 DWI is valuable to diagnose an early-stage hypovascular HCC when it shows nonspecific imaging features on hepatobiliary-phase Gd-EOB-DTPA–enhanced MRI (e.g., hyperintensity on DWI strongly indicates HCC).237 Metastatic liver cancers, including colorectal hepatic metastases, have higher ADCs compared with background liver140 and are well depicted as hyperintense nodules on DWI (Fig. 44-84). Gd-EOB-DTPA MRI combined with DWI is highly effective for detection and characterization of metastases from colorectal cancer.36 ADC measured by DWI is valuable for differential diagnosis. Hepatic cysts can be discriminated from other solid mass lesions with a cutoff of 2.5 × 10−3  mm2/sec at b values of 400 to 1000 sec/mm2. However, although ADCs tend to decrease in the order of cavernous hemangiomas, HCCs, and metastases, there are substantial overlaps among them.68 One major role of DWI in liver imaging is to detect and predict the effect of therapy on liver tumors (see Fig. 33-50). A higher pretreatment ADC value in a metastatic liver tumor is predictive of a poor response to chemotherapy, and a significant increase in ADC value is observed in metastatic lesions that respond to chemotherapy.141 Increase in the ADC value after yttrium-90 microsphere treatment of HCC may predict therapeutic response in the early posttreatment period.113

1. Ahmadi T, Itai Y, Minami M: Central dot sign in entities other than Caroli disease. Radiat Med 15:381–384, 1997. 2. Aibe H, Hondo H, Kuroiwa T, et al: Sclerosed hemangioma of the liver. Abdom Imaging 26:496–499, 2001. 3. Akimoto M, Yoshikawa J, Matsui O, et al: Hepatic granuloma with progressive calcification: CT appearance. Radiat Med 11:170–172, 1993. 4. Allaire GS, Rabin L, Ishak KG, et al: Bile duct adenoma: A study of 152 cases. Am J Surg Pathol 12:708–715, 1988. 5. Alvarez OA, Vanegas F, Maze GL, et al: Polymicrobial cholangitis and liver abscess in a patient with the acquired immunodeficiency syndrome. South Med J 93:232–234, 2000. 6. Alvarez-Mendoza A, Lourdes TS, Ridaura-Sanz C, et al: Histopathology of vascular lesions found in Kasabach-Merritt syndrome: Review based on 13 cases. Pediatr Dev Pathol 3:556–560, 2000. 7. Ametani F, Itoh K, Shibata T, et al: Spectrum of CT findings in pediatric patients after partial liver transplantation. Radiographics 21:53–63, 2001. 8. Anttila VJ, Lamminen AE, Bonestam S, et al: Magnetic resonance imaging is superior to computed tomography and ultrasonography in imaging infectious liver foci in acute leukemia. Eur J Haematol 56:82–87, 1996. 9. Aoki S, Okayama Y, Kitajima Y, et al: Intrahepatic biliary papilloma morphologically similar to biliary cystadenoma. J Gastroenterol Hepatol 20:321–324, 2005. 10. Aoki K, Takayasu K, Kawano T, et al: Combined hepatocellular carcinoma and cholangiocarcinoma: Clinical features and computed tomographic findings. Hepatology 18:1090–1095, 1993. 11. Apicella PL, Mirowitz SA, Weinreb JC: Extension of vessels through hepatic neoplasms: MR and CT findings. Radiology 191:135–136, 1994. 12. Arai K, Kawai K, Kohda W, et al: Dynamic CT of acute cholangitis: Early inhomogeneous enhancement of the liver. AJR Am J Roentgenol 181:115–118, 2003. 13. Arai K, Matsui O, Takashima T, et al: Focal spared areas in fatty liver caused by regional decreased portal flow. AJR Am J Roentgenol 151:300–302, 1988. 14. Asayama Y, Tajima T, Okamoto D, et al: Imaging of cholangiolocellular carcinoma of the liver. Eur J Radiol 75:e120–e125, 2010. 15. Asayama Y, Yoshimitsu K, Irie H, et al: Delayed-phase dynamic CT enhancement as a prognostic factor for mass-forming intrahepatic cholangiocarcinoma. Radiology 238:150–155, 2005. 16. Balci NC, Akinci A, Akun E, et al: Hepatic angiomyolipoma: Demonstration by out of phase MRI. Clin Imaging 26:418–420, 2002. 17. Balci NC, Semelka RC, Noone TC, et al: Pyogenic hepatic abscesses: MRI findings on T1- and T2-weighted and serial gadolinium-enhanced gradient-echo images. J Magn Reson Imaging 9:285–290, 1999. 18. Balci NC, Tunaci A, Akinci A, et al: Granulomatous hepatitis: MRI findings. Magn Reson Imaging 19:1107–1111, 2001. 19. Baron RL, Campbell WL, Dodd GD, III: Peribiliary cysts associated with severe liver disease: Imaging-pathologic correlation. AJR Am J Roentgenol 162:631–636, 1994. 20. Bechtold RE, Karstaedt N, Wolfman NT, et al: Prolonged hepatic enhancement on computed tomography in a case of hepatic lymphoma. J Comput Assist Tomogr 9:186–189, 1985. 21. Bhartia B, Ward J, Guthrie JA, et al: Hepatocellular carcinoma in cirrhotic livers: Double-contrast thin-section MR imaging with pathologic correlation of explanted tissue. AJR Am J Roentgenol 180:577–584, 2003. 22. Bioulac-Sage P, Rebouissou S, Thomas C, et al: Hepatocellular   adenoma subtype classification using molecular markers and immunohistochemistry. Hepatology 46:740–748, 2007. 23. Bipat S, van Leeuwen MS, Comans EF, et al: Colorectal liver metastases: CT, MR imaging, and PET for diagnosis—Meta-analysis. Radiology 237:123–131, 2005. 24. Blokhuis TJ, van der Schaaf MC, van den Tol MP, et al: Results of radiofrequency ablation of primary and secondary liver tumors: Long-term follow-up with computed tomography and positron

CHAPTER 44  Liver: Focal Hepatic Mass Lesions

25.

26.

27.

28.

29.

30.

31.

32.

33. 34. 35. 36.

37. 38.

39. 40.

41.

42.

43.

44.

45.

46.

emission tomography-18F-deoxyfluoroglucose scanning. Scand J Gastroenterol 241:93–97, 2004. Boechat MI, Kangarloo H, Ortega J, et al: Primary liver tumors in children: Comparison of CT and MR imaging. Radiology 169:727–732, 1988. Brancatelli G, Federle MP, Grazioli L, et al: Large regenerative nodules in Budd-Chiari syndrome and other vascular disorders of the liver: CT and MR imaging findings with clinicopathologic correlation. AJR Am J Roentgenol 178:877–883, 2002. Breen DJ, Rutherford EE, Stedman B, et al: Intrahepatic arterioportal shunting and anomalous venous drainage: Understanding the CT features in the liver. Eur Radiol 14:2249–2260, 2004. Buetow PC, Buck JL, Pantongrag-Brown L, et al: Biliary cystadenoma and cystadenoarcioma: Clinical-imaging-pathologic correlation with emphasis on the importance of ovarian stroma. Radiology 196:805–810, 1995. Buetow PC, Buck JL, Pantongrag-Brown L, et al: Undifferentiated (embryonal) sarcoma of the liver: Pathologic basis of imaging findings in 28 cases. Radiology 203:779–783, 1997. Buetow PC, Buck JL, Ros PR, et al: Malignant vascular tumors of the liver: Radiologic-pathologic correlation. Radiographics 14:153–166, 1994. Buetow PC, Pantongrag-Brown L, Buck JL, et al: Focal nodular hyperplasia of the liver: Radiologic-pathologic correlation. Radiographics 16:369–388, 1996. Chang S, Lim JH, Choi D, et al: Hepatic visceral larva migrans of Toxocara canis: CT and sonographic findings. AJR Am J Roentgenol 187:W622–W629, 2006. Choi BI, Kim HJ, Han MC, et al: CT finding of clonorchiasis. AJR Am J Roentgenol 152:281–284, 1989. Choi BG, Park SH, Byun JY, et al: The findings of ruptured hepatocellular carcinoma on helical CT. Br J Radiol 74:142–146, 2001. Choi BI, Yeon KM, Kim SH, et al: Caroli disease: Central dot sign in CT. Radiology 174:161–163, 1990. Chung WS, Kim MJ, Chung YE, et al: Comparison of gadoxetic acid-enhanced dynamic imaging and diffusion-weighted imaging for the preoperative evaluation of colorectal liver metastases. J Magn Reson Imaging 34:345–353, 2011. Couinaud C: The parabiliary venous system. Surg Radiol Anat 10:311–316, 1988. Dachman AH, Lichtenstein JE, Friedman AC, et al: Infantile hemangioendothelioma of the liver: A radiologic-pathologic-clinical correlation. AJR Am J Roentgenol 140:1091–1096, 1983. Dachman AH, Pakter RL, Ros PR, et al: Hepatoblastoma: Radiologicpathologic correlation in 50 cases. Radiology 164:15–20, 1987. Delbeke D, Martin WH, Sandler MP, et al: Evaluation of benign vs malignant hepatic lesions with positron emission tomography. Arch Surg 133:510–515, 1998. Demachi H, Matsui O, Kawamori Y, et al: The protective effect of portoarterial shunts after experimental hepatic artery embolization in rats with liver cirrhosis. Cardiovasc Intervent Radiol 18:97–101, 1995. Demachi H, Matsui O, Takashima T: Scanning electron microscopy of intrahepatic microvasculature casts following experimental hepatic artery embolization. Cardiovasc Intervent Radiol 14:158–162, 1991. Doo KW, Lee CH, Choi JW, et al: “Pseudo washout” sign in high-flow hepatic hemangioma on gadoxetic acid contrast-enhanced MRI mimicking hypervascular tumor. AJR Am J Roentgenol 193:W490– W496, 2009. Earls JP, Theise ND, Weinreb JC, et al: Dysplastic nodules and hepatocellular carcinoma: Thin-section MR imaging of explanted cirrhotic livers with pathologic correlation. Radiology 201:207–214, 1996. Ebara M, Watanabe S, Kita K, et al: MR imaging of small hepatocellular carcinoma: Effect of intratumoral copper content on signal intensity. Radiology 180:617–621, 1991. Ebied O, Federle MP, Blachar A, et al: Hepatocellularcholangiocarcinoma: Helical computed tomography findings in 30 patients. J Comput Assist Tomogr 27:117–124, 2003.

1367

47. Eckburg PB, Montoya JG: Hepatobiliary infections. In Wilson WR, Sande MA, editors: Diagnosis and treatment in infectious diseases: Lange current series, New York, 2001, McGraw-Hill, pp 269–286. 48. Efremidis SC, Hytiroglou P, Matsui O: Enhancement patterns and signal-intensity characteristics of small hepatocellular carcinoma in cirrhosis: Pathologic basis and diagnostic challenges. Eur Radiol 17:2969–2982, 2007. 49. Esensten M, Ralls PW, Coletti P, et al: Posttraumatic intrahepatic biloma: Sonographic diagnosis. AJR Am J Roentgenol 140:303–305, 1983. 50. Ferlay J, Shin HR, Bray F, et al: Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 127:2893–2917, 2010. 51. Ferrozzi F, Bova D, Zangrandi A, et al: Primary liver leiomyosarcoma: CT appearance. Abdom Imaging 21:157–160, 1996. 52. Findlay M, Young H, Cunningham D, et al: Noninvasive monitoring of tumor metabolism using fluorodeoxyglucose and positron emission tomography in colorectal cancer liver metastases: Correlation with tumor response to fluorouracil. J Clin Oncol 14:700–708, 1996. 53. Fujita AW, Steelman CK, Abramowsky CR, et al: Ciliated hepatic foregut cyst: Four case reports with a review of the literature. Pediatr Dev Pathol 14:418–421, 2011. 54. Fujiwara H, Sekine S, Onaya H, et al: Ring-like enhancement of focal nodular hyperplasia with hepatobiliary-phase Gd-EOB-DTPA-enhanced magnetic resonance imaging: Radiological-pathological correlation. Jpn J Radiol. 29:739–743, 2011. 55. Fukui N, Kitagawa K, Matsui O, et al: Focal ischemic necrosis of the liver associated with cirrhosis: Radiologic findings. AJR Am J Roentgenol 159:1021–1022, 1992. 56. Fukukura Y, Nakashima O, Kusaba A, et al: Angioarchitecture and blood circulation in focal nodular hyperplasia of the liver. J Hepatol 29:470–475, 1998. 57. Fukukura Y, Taguchi J, Nakashima O, et al: Combined hepatocellular and cholangiocarcinoma: Correlation between CT findings and clinicopathological features. J Comput Assist Tomogr 21:52–58, 1997. 58. Fukuya T, Honda H, Matsumata T, et al: Diagnosis of inflammatory pseudotumor of the liver: Value of CT. AJR Am J Roentgenol 163:1087– 1091, 1994. 59. Furui S, Itai Y, Ohtomo K, et al: Hepatic epithelioid hemangioendothelioma: Report of five cases. Radiology 171:63–68, 1989. 60. Gabata T, Kadoya M, Matsui O, et al: Biliary cystadenoma with mesenchymal stroma of the liver: Correlation between unusual MR appearance and pathologic findings. J Magn Reson Imaging 8:503–504, 1998. 61. Gabata T, Kadoya M, Matsui O, et al: Peritumoral spared area in fatty liver: Correlation between opposed-phase gradient-echo MR imaging and CT arteriography. Abdom Imaging 26:384–389, 2001. 62. Gabata T, Kodak M, Matsui O, et al: Dynamic CT of hepatic abscesses: Significance of transient segmental enhancement. AJR Am J Roentgenol 176:675–679, 2001. 63. Gabata T, Matsui O, Kadoya M, et al: Aberrant gastric venous drainage in a focal spared area of segment IV in fatty liver: Demonstration with color Doppler sonography. Radiology 203:461–463, 1997. 64. Gabata T, Matsui O, Kadoya M, et al: Segmental hyperintensity on T1-weighted MRI of the liver: Indication of segmental cholestasis.   J Magn Reson Imaging 7:855–857, 1997. 65. Gabata T, Matsui O, Kadoya M, et al: Delayed MR imaging of the liver: Correlation of delayed enhancement of hepatic tumors and pathologic appearance. Abdom Imaging 23:309–313, 1998. 66. Gabata T, Terayama N, Kobayashi S, et al: MR imaging of hepatocellular carcinomas with biliary tumor thrombi. Abdom Imaging 32:470–474, 2007. 67. Giannitrapani L, Soresi M, La Spada E, et al: Sex hormones and risk of liver tumor. Ann N Y Acad Sci 1089:228–236, 2006. 68. Goshima S, Kanematsu M, Kondo H, et al: Diffusion-weighted imaging of the liver: Optimizing b value for the detection and characterization of benign and malignant hepatic lesions. J Magn Reson Imaging 28:691–697, 2008.

1368

PART II  CT and MR Imaging of the Whole Body

69. Goshima S, Kanematsu M, Matsuo M, et al: Nodule-in-nodule appearance of hepatocellular carcinomas: Comparison of gadoliniumenhanced and ferumoxides-enhanced magnetic resonance imaging.   J Magn Reson Imaging 20:250–255, 2004. 70. Grandin C, Van Beers BE, Robert A, et al: Benign hepatocellular tumors: MRI after superparamagnetic iron oxide administration.   J Comput Assist Tomogr 19:412–418, 1995. 71. Grazioli L, Bondioni MP, Haradome H, et al: Hepatocellular adenoma and focal nodular hyperplasia: Value of gadoxetic acid-enhanced MR imaging in differential diagnosis. Radiology 262:520–529, 2012. 72. Grazioli L, Federle MP, Ichikawa T, et al: Liver adenomatosis: Clinical, histolopathologic, and imaging findings in 15 patients. Radiology 216:395–402, 2000. 73. Guy F, Cognet F, Dranssart M, et al: Caroli’s disease: Magnetic resonance imaging features. Eur Radiol 12:2730–2736, 2002. 74. Halvorsen RA, Korobkin M, Foster WL, et al: The variable CT appearance of hepatic abscesses. AJR Am J Roentgenol 141:941–946, 1984. 75. Hamrick-Turner JE, Shipkey FH, Cranston PE: Fibrolamellar hepatocellular carcinoma: MR appearance mimicking focal nodular hyperplasia. J Comput Assist Tomogr 18:301–304, 1994. 76. Han K, Choi BI, Cho JM, et al: Radiological findings of human fascioliasis. Abdom Imaging 18:261–264, 1993. 77. Harada K, Terada T, Nakanuma Y, et al: A case of small combined hepatocellular and cholangiocellular carcinoma arising in a nodule of atypical adenomatous hyperplasia of the liver. Am J Gastroenterol 88:1968–1969, 1993. 78. Hardie AD, Kizziah MK, Rissing MS: Can the patient with cirrhosis be imaged for hepatocellular carcinoma without gadolinium? Comparison of combined T2-weighted, T2*-weighted, and diffusion-weighted MRI with gadolinium-enhanced MRI using liver explantation standard.   J Comput Assist Tomogr 35:711–715, 2011. 79. Hayashi M, Matsui O, Ueda K, et al: Imaging findings of mucinous type of cholangiocellular carcinoma. J Comput Assist Tomogr 20:386– 389, 1996. 80. Hayashi M, Matsui O, Ueda K, et al: Correlation between the blood supply and grade of malignancy of hepatocellular nodules associated with liver cirrhosis: Evaluation by CT during intraarterial injection of contrast medium. AJR Am J Roentgenol 172:969–976, 1999. 81. Hayashi M, Matsui O, Ueda K, et al: Progression to hypervascular hepatocellular carcinoma: Correlation with intranodular blood supply evaluated with CT during intraarterial injection of contrast material. Radiology 225:143–149, 2002. 82. Heo SH, Jeong YY, Shin SS, et al: Solitary small hepatic angiosarcoma: Initial and follow-up imaging findings. Korean J Radiol 8:180–183, 2007. 83. Hihara T, Araki T, Katou K, et al: Cystic cavernous hemangioma of the liver. Gastrointest Radiol 15:112–114, 1990. 84. Honda H, Hayashi T, Yoshida K, et al: Hepatocellular carcinoma with sarcomatous change: Characteristic findings of two-phased incremental CT. Abdom Imaging 21:37–40, 1996. 85. Hoshiba K, Matsui O, Kadoya M, et al: Peribiliary cysts in cirrhotic liver: Observation on computed tomography. Abdom Imaging 21:228–232, 1996. 86. Huang CJ, Pitt HA, Lipsett PA, et al: Pyogenic hepatic abscess. Ann Surg 223:600–609, 1996. 87. Hussain SM, Terkivatan T, Zondervan PE, et al: Focal nodular hyperplasia: Findings at state-of-the-art MR imaging, US, CT, and pathologic analysis. Radiographics 24:3–17, 2004. 88. Iannaccone R, Federle MP, Brancatelli G, et al: Peliosis hepatis: Spectrum of imaging findings. AJR Am J Roentgenol 187:W43–W52, 2006. 89. Ichikawa T, Federle MP, Grazioli L, et al: Fibrolamellar hepatocellular carcinoma: Imaging and pathologic findings in 31 recent cases. Radiology 213:352–361, 1999. 90. Ikai I, Arii S, Okazaki M, et al: Report of the 17th Nationwide Follow-up Survey of Primary Liver Cancer in Japan. Hepatol Res 37:676–691, 2007.

91. Imai Y, Murakami T, Yoshida S, et al: Superparamagnetic iron oxide-enhanced magnetic resonance images of hepatocellular carcinoma: Correlation with histological grading. Hepatology 32:205–212, 2000. 92. International Consensus Group for Hepatocellular Neoplasia The International Consensus Group for Hepatocellular Neoplasia: Pathologic diagnosis of early hepatocellular carcinoma: A report of the international consensus group for hepatocellular neoplasia. Hepatology 49:658–664, 2009. 93. International Working Party: Terminology of nodular hepatocellular lesions. Hepatology 22:983–993, 1995. 94. Irie H, Honda H, Kaneko K, et al: Primary malignant lymphoma in the porta hepatis: A case report. Abdom Imaging 21:448–450, 1996. 95. Ishak K, Goodman ZD, Stocker JT: Tumors of the liver and intrahepatic bile ducts. In Atlas of tumor pathology, fasc 31, ser 2, Washington, DC, 2001, Armed Forces Institute of Pathology, pp 71–86. 96. Itai Y, Araki T, Furui S, et al: Computed tomography of primary intrahepatic biliary malignancy. Radiology 147:485–490, 1983. 97. Itai Y, Ebihara R, Tohno E, et al: Hepatic peribiliary cysts: Multiple tiny cysts within the larger portal tract, hepatic hilum, or both. Radiology 191:107–110, 1994. 98. Itai Y, Maeda M, Echigo J, et al: Hyperattenuating rim on noncontrast CT of the liver: Probable peritumoral sparing of fatty infiltration. Clin Radiol 51:406–410, 1996. 99. Itai Y, Matsui O: Blood flow and liver imaging. Radiology 202:306–314, 1997. 100. Itai Y, Moss AA, Goldberg HI: Transient hepatic attenuation difference of lobar or segmental distribution detected by dynamic computed tomography. Radiology 144:835–839, 1982. 101. Itai Y, Ohtomo K, Araki T, et al: Computed tomography and sonography of cavernous hemangioma of the liver. AJR Am J Roentgenol 141:315–320, 1983. 102. Itai Y, Ohtomo K, Kokubo T, et al: CT demonstration of fluid-fluid levels in nonenhancing hemangiomas of the liver. J Comput Assist Tomogr 11:763–765, 1987. 103. Itai Y, Teraoka T: Angiosarcoma of the liver mimicking cavernous hemangioma on dynamic CT. J Comput Assist Tomogr 13:910–912, 1989. 104. Itoh K, Awaya H, Mitchell DG, et al: Gallbladder disease: Appearance of associated transient increased attenuation in the liver at biphasic, contrast-enhanced dynamic CT. Radiology 204:723–728, 1997. 105. Itoh T, Kanaoka M, Obara A, et al: Lymphangiosis carcinomatosa of the liver. Acta Pathol Jpn 38:751–758, 1988. 106. Jeffrey RB, Tolentino CS, Chang FC, et al: CT of small pyogenic abscesses: The cluster sign. AJR Am J Roentgenol 144:463–467, 1985. 107. Juimo AG, Gervez F, Angwafo FF: Extraintestinal amebiasis. Radiology 182:181, 1992. 108. Kadoya M, Matsui O, Kitagawa K, et al: Segmental iron deposition in the liver due to decreased intrahepatic portal perfusion: Findings at MR imaging. Radiology 193:671–676, 1994. 109. Kadoya M, Matsui O, Nakanuma Y, et al: Ciliated hepatic foregut cyst: Radiologic features. Radiology 175:475–477, 1990. 110. Kadoya M, Matsui O, Takashima T, et al: Hepatocellular carcinoma: Correlation of MR imaging and histopathologic findings. Radiology 183:819–825, 1992. 111. Kahn JA, Haker KM, Petrovic LM, et al: Radiographic findings of ischemic hepatitis in a cirrhotic patient. J Comput Assist Tomogr 24:887–889, 2000. 112. Kalovidouris A, Gouliamos A, Vlachos L, et al: MRI of abdominal hydatid disease. Abdom Imaging 19:489–494, 1994. 113. Kamel IR, Reyes DK, Liapi E, et al: Functional MR imaging assessment of tumor response after 90Y microsphere treatment in patients with unresectable hepatocellular carcinoma. J Vasc Interv Radiol 18:49–56, 2007. 114. Kanazawa S, Wright KC, Kasi LP, et al: Preliminary experimental evaluation of temporary segmental hepatic venous occlusion: Angiographic, pathologic, and scintigraphic findings. J Vasc Interv Radiol 4:759–766, 1993.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions 115. Kanematsu M, Imaeda T, Yamawaki Y, et al: Rupture of hepatocellular carcinoma: Predictive value of CT findings. AJR Am J Roentgenol 158:1247–1250, 1992. 116. Karçaaltincaba M, Idilman I, Celik A: Focal sparing of iron and fat in liver tissue in patients with hemosiderosis: diagnosis with combination of R2* relaxometry and proton density fat fraction calculation by MRI. Diagn Interv Radiol 17:323–327, 2011. 117. Kawamori Y, Matsui O, Kadoya M, et al: Differentiation of hepatocellular carcinomas from hyperplastic nodules induced in rat liver with ferrite-enhanced MR imaging. Radiology 183:65–72, 1992. 118. Kawamori Y, Matsui O, Kitagawa K, et al: Macronodular tuberculoma of the liver: CT and MR findings. AJR Am J Roentgenol 158:311–313, 1992. 119. Kawamori Y, Matsui O, Takahashi S, et al: Focal hepatic fatty infiltration in the posterior edge of the medial segment associated with aberrant gastric venous drainage: CT, US and MR findings. J Comput Assist Tomogr 20:356–359, 1996. 120. Kawamoto S, Soyer PA, Fishman EK, et al: Nonneoplastic liver disease: Evaluation with CT and MR imaging. Radiographics 18:827–848, 1998. 121. Kelekis NL, Semelka RC, Woosley JT: Malignant lesions of the liver with high signal intensity on T1-weighted MR images. J Magn Reson Imaging 6:291–294, 1996. 122. Kerlin P, Davis GL, McGill DB, et al: Hepatic adenoma and focal nodular hyperplasia: Clinical, pathologic and radiologic features. Gastroenterology 84:994–1002, 1983. 123. Kessler A, Mitchell DG, Israel HL, et al: Hepatic and splenic sarcoidosis: Ultrasound and MR imaging. Abdom Imaging 18:159–163, 1993. 124. Khan SA, Thomas HC, Davidson BR, et al: Cholangiocarcinoma. Lancet 366:1303–1314, 2005. 125. Kim T, Baron RL, Nalesnik MA: Infarcted regenerative nodules in cirrhosis: CT and MR imaging findings with pathologic correlation. AJR Am J Roentgenol 175:1121–1125, 2000. 126. Kim TK, Choi BI, Han JK, et al: Peripheral cholangiocarcinoma of the liver: Two-phase spiral CT findings. Radiology 204:539, 1997. 127. Kim BS, Kim TK, Kim JS, et al: Hepatic venous congestion after living donor liver transplantation with right lobe graft: Two-phase CT findings. Radiology 232:173–180, 2004. 128. Kim KW, Kim AY, Kim TK, et al: Hepatic hemangioma with arterioportal shunt: Sonographic appearance with CT and MRI correlation. AJR Am J Roentgenol 187:406–414, 2006. 129. Kim SH, Lim HK, Choi D, et al: Changes in bile ducts after radiofrequency ablation of hepatocellular carcinoma: Frequency and clinical significance. AJR Am J Roentgenol 183:1611–1617, 2004. 130. Kim YJ, Yun M, Lee WJ, et al: Usefulness of 18F-FDG PET in intrahepatic cholangiocarcinoma. Eur J Nucl Med Mol Imaging 30:1467–1472, 2003. 131. Kitagawa K, Matsui O, Kadoya M, et al: Hepatocellular carcinomas   with excessive copper accumulation: CT and MR findings. Radiology 180:623–628, 1991. 132. Kitao A, Matsui O, Yoneda N, et al: The uptake transporter OATP8 expression decreases during multistep hepatocarcinogenesis: Correlation with gadoxetic acid enhanced MR imaging. Eur Radiol 21:2056–2066, 2011. 133. Kitao A, Matsui O, Yoneda N, et al: Hypervascular hepatocellular carcinoma: Correlation between biologic features and signal intensity on gadoxetic acid-enhanced MR images. Radiology 265:780–789, 2012. 134. Kitao A, Zen Y, Matsui O, et al: Multistep changes of drainage vessels at CT during arterial portography and hepatic arteriography—Radiologicpathologic correlation. Radiology 252:605–614, 2009. 135. Kitao A, Zen Y, Matsui O, et al: Hepatocellular carcinoma: Signal intensity at gadoxetic acid-enhanced MR imaging—Correlation with molecular transporters and histopathologic features. Radiology 256:817–826, 2010. 136. Kobayashi S, Matsui O, Gabata T: Pseudolesion in segment IV of the liver adjacent to the falciform ligament caused by drainage of the paraumbilical vein: Demonstration by power Doppler ultrasound. Br J Radiol 74:273–276, 2001.

1369

137. Kobayashi S, Matsui O, Kadoya M, et al: CT arteriographic confirmation of focal hepatic fatty infiltration adjacent to the falciform ligament associated with drainage of inferior vein of Sappey: A case report. Radiat Med 19:51–54, 2001. 138. Kobayashi S, Nakanuma Y, Terada T, et al: Postmortem survey of   bile duct necrosis and biloma in hepatocellular carcinoma after transcatheter arterial chemoembolization therapy: Relevance to microvascular damage of peribiliary capillary plexus. Am J Gastroenterol 88:1411–1415, 1993. 139. Kodama Y, Fujita N, Shimizu T, et al: Alveolar echinococcosis: MR findings in the liver. Radiology 228:172–177, 2003. 140. Koh DM, Scurr E, Collins DJ, et al: Colorectal hepatic metastases: Quantitative measurements using single-shot echo-planar diffusionweighted MR imaging. Eur Radiol 16:1898–1905, 2006. 141. Koh DM, Scurr E, Collins D, et al: Predicting response of colorectal hepatic metastasis: Value of pretreatment apparent diffusion coefficients. AJR Am J Roentgenol 188:1001–1008, 2007. 142. Kojiro M, Roskams T: Early hepatocellular carcinoma and dysplastic nodules. Semin Liver Dis 25:133–142, 2005. 143. Kondo F: Benign nodular hepatocellular lesions caused by abnormal hepatic circulation: Etiological analysis and introduction of a new concept. J Gastroenterol Hepatol 16:1319–1328, 2001. 144. Korobkin M, Stephens D, Lee J, et al: Biliary cystadenoma and cystadenocarcinoma: CT and sonographics findings. AJR Am J Roentgenol 153:507–511, 1989. 145. Koyama T, Fletcher JG, Johnson CD, et al: Primary hepatic angiosarcoma: Findings at CT and MR imaging. Radiology 222:667– 673, 2002. 146. Kozaka K, Sasaki M, Fujii T, et al: A subgroup of intrahepatic cholangiocarcinoma with an infiltrating replacement growth pattern and a resemblance to reactive proliferating bile ductules: “Bile ductular carcinoma”. Histopathology 51:390–400, 2007. 147. Kozakai I, Ichida T, Yamaguchi O, et al: Delayed anoxic pseudolobular necrosis (terminal hepatic necrosis) after chemo-lipiodolization for hepatocellular carcinoma. Am J Gastroenterol 91:1263–1265, 1996. 148. Kroncke TJ, Taupitz M, Kivelitz D, et al: Multifocal nodular fatty infiltration of the liver mimicking metastatic disease on CT: Imaging findings and diagnosis using MR imaging. Eur Radiol 10:1095–1100, 2000. 149. Kurogi M, Nakashima O, Miyaaki H, et al: Clinicopathological study of scirrhous hepatocellular carcinoma. J Gastroenterol Hepatol 21:1470– 1477, 2006. 150. Lacomis JM, Baron RL, Oliver JH, et al: Cholangiocarcinoma: Delayed CT contrast enhancement patterns. Radiology 203:98, 1997. 151. Laumonier H, Bioulac-Sage P, Laurent C, et al: Hepatocellular adenomas: Magnetic resonance imaging features as a function of molecular pathological classification. Hepatology 48:808–818, 2008. 152. Lee NK, Kim S, Lee JW, et al: Biliary MR imaging with Gd-EOB-DTPA and its clinical applications. Radiographics 29:1707–1724, 2009. 153. Lee JH, Park JY, Kim do Y, et al: Prognostic value of 18F-FDG PET for hepatocellular carcinoma patients treated with sorafenib. Liver Int 31:1144–1149, 2011. 154. Lee MJ, Saini S, Hamm B, et al: Focal nodular hyperplasia of the liver: MR findings in 35 proved cases. AJR Am J Roentgenol 156:317–320, 1991. 155. Leslie DF, Johnson CD, MacCarty RL, et al: Single-pass CT of hepatic tumors: Value of globular enhancement in distinguishing hemangiomas from hypervascular metastases. AJR Am J Roentgenol 165:1403–1406, 1995. 156. Lev-Toaff AS, Bach AM, Wechsler RJ, et al: The radiologic and pathologic spectrum of biliary hamartomas. AJR Am J Roentgenol 165:309–313, 1995. 157. Levy AD, Rohrmann CA, Jr, Murakata LA, et al: Caroli’s disease: Radiologic spectrum with pathologic correlation. AJR Am J Roentgenol 179:1053–1057, 2002. 158. Lewin M, Mourra N, Honigman I, et al: Assessment of MRI and MRCP in diagnosis of biliary cystadenoma and cystadenocarcinoma. Eur Radiol 16:407–413, 2006.

1370

PART II  CT and MR Imaging of the Whole Body

159. Li CS, Chen RC, Lii JM, et al: Magnetic resonance imaging appearance of well-differentiated hepatocellular carcinoma. J Comput Assist Tomogr 30:597–603, 2006. 160. Libbrecht L, Cassiman D, Verslype C, et al: Clinicopathological features of focal nodular hyperplasia-like nodules in 130 cirrhotic explant livers. Am J Gastroenterol 101:2341–2346, 2006. 161. Lim JH: Radiologic findings of clonorchiasis. AJR Am J Roentgenol 155:1001–1008, 1990. 162. Lim JH: Cholangiocarcinoma: Morphologic classification according to growth pattern and imaging findings. AJR Am J Roentgenol 181:819– 827, 2003. 163. Lim JH, Choi D, Cho SK, et al: Conspicuity of hepatocellular nodular lesions in cirrhotic livers at ferumoxides-enhanced MR imaging: Importance of Kupffer cell number. Radiology 220:669–676, 2001. 164. Lim JH, Jang KT, Rhim H, et al: Biliary cystic intraductal papillary mucinous tumor and cystadenoma/cystadenocarcinoma: Differentiation by CT. Abdom Imaging 32:644–651, 2007. 165. Lin CY, Chen JH, Liang JA, et al: 18F-FDG PET or PET/CT for detecting extrahepatic metastases or recurrent hepatocellular carcinoma: A systematic review and metaanalysis. Eur J Radiol 81:2417–2422, 2012. 166. Liver Cancer Study Group of Japan: General rules for the clinical and pathological study of primary liver cancer, Tokyo, 2003, Kanehara. 167. Lyburn ID, Torreggiani WC, Harris AC, et al: Hepatic epithelioid hemangioendothelioma: Sonographic, CT, and MR imaging appearances. AJR Am J Roentgenol 180:1359–1364, 2003. 168. Maetani Y, Itoh K, Egawa H, et al: Benign hepatic nodules in BuddChiari syndrome: Radiologic-pathologic correlation with emphasis on the central scar. AJR Am J Roentgenol 178:869–875, 2002. 169. Maetani Y, Itoh K, Watanabe C, et al: MR imaging of intrahepatic cholangiocarcinoma with pathologic correlation. AJR Am J Roentgenol 176:1499–1507, 2001. 170. Mahfouz AE, Hamm B, Taupitz M, et al: Hypervascular liver lesions: Differentiation of focal nodular hyperplasia from malignant tumors with dynamic gadolinium-enhanced MR imaging. Radiology 186:133– 138, 1993. 171. Makhlouf HR, Ishak KG, Goodman ZD: Epithelioid hemangioendothelioma of the liver: A clinicopatholgic study of 137 cases. Cancer 85:562–582, 1999. 172. Mansour A, Qandeel M, Sughayer M, et al: Solid (stromal predominance) hepatic mesenchymal hamartoma. Pediatr Radiol 35:349–350, 2005. 173. Martin J, Sentis M, Zidan A, et al: Fatty metamorphosis of hepatocellular carcinoma: Detection with chemical shift gradient-echo MR imaging. Radiology 195:125–130, 1995. 174. Mathieu R, Rahmouni A, Vasile N, et al: Sclerosed liver hemangioma mimicking malignant tumor at MR imaging: Pathologic correlation.   J Magn Reson Imaging 4:506–508, 1994. 175. Mathieu D, Vasile N, Fagniez P-L, et al: Dynamic CT features of hepatic abscesses. Radiology 154:749–752, 1985. 176. Matsui O: Imaging of multistep human hepatocarcinogenesis by CT during intra-arterial contrast injection. Intervirology 47:271–276, 2004. 177. Matsui O, Itai Y: Hepatocellular carcinoma: Japanese experience. In Freeny PC, Stevenson GW, editors: Margulis and Burhenne’s alimentary tract radiology, ed 5, St Louis, 1994, Mosby, p 1688. 178. Matsui O, Kadoya M, Kameyama T, et al: Adenomatous hyperplastic nodules in the cirrhotic liver: Differentiation from hepatocellular carcinoma with MR imaging. Radiology 173:123–126, 1989. 179. Matsui O, Kadoya M, Kameyama T, et al: Benign and malignant nodules in cirrhotic livers: Distinction based on blood supply. Radiology 178:493–497, 1991. 180. Matsui O, Kadoya M, Takahashi S, et al: Focal sparing of segment IV in fatty livers shown by sonography and CT: Correlation with aberrant gastric venous drainage. AJR Am J Roentgenol 164:1137–1140, 1995. 181. Matsui O, Kadoya M, Takashima T, et al: Intrahepatic periportal abnormal intensity on MR images: An indication of various hepatobiliary diseases. Radiology 171:335–338, 1989.

182. Matsui O, Kadoya M, Yoshikawa J, et al: Aberrant gastric venous drainage in cirrhotic livers: Imaging findings in focal areas of liver parenchyma. Radiology 197:345–349, 1995. 183. Matsui O, Takahashi S, Kadoya M, et al: Pseudolesions in segment IV of the liver at CT during arterial portography: Correlation with aberrant gastric venous drainage. Radiology 193:31–35, 1994. 184. Matsui O, Takashima T, Kadoya M, et al: Segmental staining on hepatic arteriography as a sign of intrahepatic portal vein obstruction. Radiology 152:601–606, 1984. 185. Matsui O, Takashima T, Kadoya M, et al: Staining in the liver surrounding gallbladder fossa on hepatic arteriography caused by increased cystic venous drainage. Gastrointest Radiol 12:307–312, 1987. 186. Matsumoto N, Ogawa M, Kawabata M, et al: Pseudolymphoma of the liver: Sonographic findings and review of the literature. J Clin Ultrasound 35:284–288, 2007. 187. McLarney JK, Rucker PT, Bender GN, et al: Fibrolamellar carcinoma of the liver: Radiologic-pathologic correlation. Radiographics 19:453–471, 1999. 188. Mendez RJ, Schiebler ML, Outwater EK, et al: Hepatic abscesses: MR imaging findings. Radiology 190:431–436, 1994. 189. Miller WJ, Dodd GD, 3rd, Federle MP, et al: Epithelioid hemangioendothelioma of the liver: Imaging findings with pathologic correlation. AJR Am J Roentgenol 159:53–57, 1992. 190. Miller WJ, Sechtin AG, Campbell WL, et al: Imaging findings in Caroli’s disease. AJR Am J Roentgenol 165:333–337, 1995. 191. Mitchell DG, Rubin R, Siegelman ES, et al: Hepatocellular carcinoma within siderotic regenerative nodules: Appearance as a nodule within a nodule on MR images. Radiology 178:101–103, 1991. 192. Mitsudo K, Watanabe Y, Saga T, et al: Nonenhanced hepatic cavernous hemangioma with multiple calcifications: CT and pathologic correlation. Abdom Imaging 20:459–461, 1995. 193. Miyayama S, Matsui O, Ueda K, et al: Hemodynamics of small hepatic focal nodular hyperplasia: Evaluation with single-level dynamic CT during hepatic arteriography. AJR Am J Roentgenol 174:1567–1569, 2000. 194. Miyayama S, Nishida H, Matsui O: Peritumoral fatty infiltration of the liver associated with venous drainage from metastatic liver tumor. AJR Am J Roentgenol 182:533–534, 2004. 195. Mohajer K, Frydrychowicz A, Robbins JB, et al: Characterization of hepatic adenoma and focal nodular hyperplasia with gadoxetic acid.   J Magn Reson Imaging 36:686–696, 2012. 196. Monzawa S, Ohtomo K, Oba H, et al: Septa in the liver of patients with chronic hepatic schistosomiasis japonica: MR appearance. AJR Am J Roentgenol 162:1347–1351, 1994. 197. Monzawa S, Uchiyama G, Ohtomo K, et al: Schistosomiasis japonica of the liver: Contrast-enhanced CT findings in 113 patients. AJR Am J Roentgenol 161:323–327, 1993. 198. Mortele B, Mortele K, Seynaeve P, et al: Hepatic bile duct hamartomas (von Meyenburg complexes): MR and MR cholangiography findings.   J Comput Assist Tomogr 26:438–443, 2002. 199. Mortele KJ, Ros PR: Cystic focal liver lesions in the adult: Differential CT and MR imaging features. Radiographics 21:895–910, 2001. 200. Mortele KJ, Segatto E, Ros PR: The infected liver: Radiologic-pathologic correlation. Radiographics 24:937–955, 2004. 201. Mortele KJ, Vanzieleghem B, Mortele B, et al: Solitary hepatic infantile hemangioendothelioma: Dynamic-enhanced MR imaging findings. Eur Radiol 12:862–865, 2002. 202. Motoo Y, Yamaguchi Y, Watanabe H, et al: Hepatic peribiliary cysts diagnosed by magnetic resonance cholangiography. J Gastroenterol 36:271–275, 2001. 203. Motosugi U, Ichikawa T, Nakajima H, et al: Cholangiolocellular carcinoma of the liver: Imaging findings. J Comput Assist Tomogr 33:682–688, 2009. 204. Motosugi U, Ichikawa T, Sou H, et al: Distinguishing hypervascular pseudolesions of the liver from hypervascular hepatocellular carcinomas with gadoxetic acid-enhanced MR imaging. Radiology 256(1):151–158, 2010.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions 205. Muramatsu Y, Nawano S, Takayasu K, et al: Early hepatocellular carcinoma: MR imaging. Radiology 181:209–221, 1991. 206. Muramatsu Y, Takayasu K, Moriyama N, et al: Peripheral low-density area of hepatic tumors: CT-pathologic correlation. Radiology 160:49– 52, 1986. 207. Murata S, Itai Y, Asato M, et al: Effect of temporary occlusion of the hepatic vein on dual blood in the liver: Evaluation with spiral CT. Radiology 197:351–356, 1995. 208. Naganawa S, Kawai H, Fukatsu H, et al: Diffusion-weighted imaging of the liver: Technical challenges and prospects for the future. Magn Reson Med Sci 4:175–186, 2005. 209. Nakajima T, Tajima Y, Sugano I, et al: Intrahepatic cholangiocarcinoma with sarcomatous change: Clinicopathologic and immunohistochemical evaluation of seven cases. Cancer 72:1872–1877, 1993. 210. Nakamoto Y, Higashi T, Sakahara H, et al: Contribution of PET in the detection of liver metastases from pancreatic tumours. Clin Radiol 54:248–252, 1999. 211. Nakamura K, Zen Y, Sato Y, et al: Vascular endothelial growth factor, its receptor Flk-1, and hypoxia inducible factor-1alpha are involved in malignant transformation in dysplastic nodules of the liver. Hum Pathol 38:1532–1546, 2007. 212. Nakanuma Y, Harada K, Ishikawa A, et al: Anatomic and molecular pathology of intrahepatic cholangiocarcinoma. J Hepatobiliary Pancreat Surg 10:265–281, 2003. 213. Nakanuma Y, Kurumaya H, Ohta G: Multiple cysts in the hepatic   hilum and their pathogenesis: A suggestion of periductal gland origin. Virchows Arch 404:341–350, 1984. 214. Nakanuma Y, Terada T, Ohta G, et al: Caroli’s disease in congenital hepatic fibrosis and infantile polycystic disease. Liver 2:346–354, 1982. 215. Nakanuma Y, Terada T, Terasaki S, et al: Atypical adenomatous hyperplasia in liver cirrhosis: Low-grade hepatocellular carcinoma or borderline lesion? Histopathology 17:27–35, 1990. 216. Nakashima O, Kurogi M, Yamaguchi R, et al: Unique hypervascular nodules in alcoholic liver cirrhosis: Identical to focal nodular hyperplasia-like nodules? J Hepatol 41:992–998, 2004. 217. Nakashima Y, Nakashima O, Hsia CC, et al: Vascularization of small hepatocellular carcinomas: Correlation with differentiation. Liver 19:12–18, 1999. 218. Nakashima Y, Nakashima O, Tanaka M, et al: Portal vein invasion and intrahepatic micrometastasis in small hepatocellular carcinoma by gross type. Hepatol Res 26:142–147, 2003. 219. Namasivayam S, Martin DR, Saini S: Imaging of liver metastases: MRI. Cancer Imaging 7:2–9, 2007. 220. Narla LD, Newman B, Spottswood SS, et al: Inflammatory pseudotumor. Radiographics 23:719–729, 2003. 221. Nasu K, Kuroki Y, Tsukamoto T, et al: Diffusion-weighted imaging of surgically resected hepatocellular carcinoma: Imaging characteristics and relationship among signal intensity, apparent diffusion coefficient, and histopathologic grade. Am J Roentgenol. 193:438–444, 2009. 222. Nonomura A, Minato H, Shimizu K, et al: Hepatic hilar inflammatory pseudotumor mimicking cholangiocarcinoma with cholangitis and phlebitis—A variant of primary sclerosing cholangitis? Pathol Res Pract 193:519–525, 1997. 223. Nonomura A, Mizukami Y, Kadoya M: Angiomyolipoma of the liver: A collective review. J Gastroenterol 29:95–105, 1994. 224. Nonomura A, Mizukami Y, Shimizu K, et al: Angiomyolipoma mimicking true lipoma of the liver: Report of two cases. Pathol Int 46:221–227, 1996. 225. Nonomura A, Mizukami Y, Takayanagi N, et al: Immunohistochemical study of hepatic angiomyolipoma. Pathol Int 46:24–32, 1996. 226. Nordenstedt H, White DL, El-Serag HB: The changing pattern of epidemiology in hepatocellular carcinoma. Dig Liver Dis 42(Suppl 3): S206–S214, 2010. 227. Nushijima Y, Ishida H, Watanabe Y, et al: Amebic liver abscess rupturing into the lesser omentum space. J Hepatobiliary Pancreat Surg 13:252–255, 2006. 228. O’Connor P, Buhac I, Balint JA: Spontaneous hepatic arterial thrombosis with infarction of the liver. Gastroenterology 70:599–601, 1976.

1371

229. Ohashi I, Ina H, Gomi N, et al: Hepatic pseudolesion in the left lobe around the falciform ligament at helical CT. Radiology 196:245–249, 1995. 230. Ohtomo K, Baron RL, Dodd GD, 3rd, et al: Confluent hepatic fibrosis in advanced cirrhosis: Appearance at CT. Radiology 188:31–35, 1993. 231. Ohtomo K, Baron RL, Dodd GD, 3rd, et al: Confluent hepatic fibrosis in advanced cirrhosis: Evaluation with MR imaging. Radiology 189:871–874, 1993. 232. Okuda K, Nakanuma Y, Miyazaki M: Cholangiocarcinoma: Recent progress. Part 1. Epidemiology and etiology. J Gastroenterol Hepatol 17:1049–1055, 2002. 233. Ota H, Isoda N, Sunada F, et al: A case of hepatic pseudolymphoma observed without surgical intervention. Hepatol Res 35:296–301, 2006. 234. Outwater E, Tomaszewski JE, Daly JM, et al: Hepatic colorectal metastases: Correlation of MR imaging and pathologic appearance. Radiology 180:327–332, 1991. 235. Ow C, Maldjian C, Shires T, et al: CT, US, and MR imaging of hepatic aspergilloma. J Comput Assist Tomogr 15:852–854, 1991. 236. Paley MR, Mergo PJ, Torres GM, et al: Characterization of focal hepatic lesions with ferumoxides-enhanced T2-weighted MR imaging. AJR Am J Roentgenol 175:159–163, 2000. 237. Park MJ, Kim YK, Lee MW, et al: Small hepatocellular carcinomas: Improved sensitivity by combining gadoxetic acid-enhanced and diffusion-weighted MR imaging patterns. Radiology 264:761–770, 2012. 238. Patterson JE: Epidermology of fungal infections in solid organ transplant patients. Transpl Infect Dis 1:229–236, 1999. 239. Peterson MS, Baron RL, Rankin SC: Hepatic angiosarcoma: Findings   on multiphasic contrast-enhanced helical CT do not mimic hepatic hemangioma. AJR Am J Roentgenol 175:165–170, 2000. 240. Popescu I, Ciurea S, Brasoveanu V, et al: Liver hemangioma   revisited: Current surgical indications, technical aspects, results. Hepatogastroenterology 48:770–776, 2001. 241. Powers C, Ros PR, Stoupis C, et al: Primary liver neoplasms: MR imaging with pathologic correlation. Radiographics 14:459–482, 1994. 242. Prasad SR, Wang H, Rosas H, et al: Fat-containing lesions of the liver: Radiologic-pathologic correlation. Radiographics 25:321–331, 2005. 243. Principe A, Lugaresi ML, Lords RC, et al: Bile duct hamartomas: Diagnostic problems and treatment. Hepatogastroenterology 44:994–997, 1997. 244. Psatha EA, Semelka RC, Fordham L, et al: Undifferentiated (embryonal) sarcoma of the liver (USL): MRI findings including dynamic gadolinium enhancement. Magn Reson Imaging 22:897–900, 2004. 245. Quaglia A, Tibballs J, Grasso A, et al: Focal nodular hyperplasia-like areas in cirrhosis. Histopathology 42:14–21, 2003. 246. Quinn SF, Benjamin GG: Hepatic cavernous hemangiomas: Simple diagnostic sign with dynamic bolus CT. Radiology 182:545–548, 1992. 247. Radin DR, Craig JR, Colletti PM, et al: Hepatic epithelioid hemangioendothelioma. Radiology 169:145–148, 1988. 248. Radin DR, Ralls PW, Colletti PM, et al: CT of amebic liver abscess. AJR Am J Roentgenol 150:1297–1301, 1988. 249. Ralls PW, Henley DS, Coletti PM, et al: Amebic liver abscess: MR imaging. Radiology 165:801–804, 1987. 250. Reinhardt MJ, Strunk H, Gerhardt T, et al: Detection of Klatskin’s tumor in extrahepatic bile duct strictures using delayed 18F-FDG PET/ CT: Preliminary results for 22 patient studies. J Nucl Med 46:1158– 1163, 2005. 251. Roos JE, Pfiffner R, Stallmach T, et al: Infantile hemangioendothelioma. Radiographics 23:1649–1655, 2003. 252. Ros PR, Buck JL, Goodman ZD, et al: Intrahepatic cholangiocarcinoma: Radiologic-pathologic correlation. Radiology 167:689, 1988. 253. Saab S, Yao F: Fibrolamellar hepatocellular carcinoma: Case reports and a review of the literature. Dig Dis Sci 41:1981–1985, 1996. 254. Sabet A, Ahmadzadehfar H, Bruhman J, et al: Survival in patients   with hepatocellular carcinoma treated with 90Y-microsphere radioembolization. Prediction by 18F-FDG PET. Nuklearmedizin 53:39–45, 2014.

1372

PART II  CT and MR Imaging of the Whole Body

255. Sakamoto M, Hirohashi S, Shimosato Y: Early stages of multistep hepatocarcinogenesis: Adenomatous hyperplasia and early hepatocellular carcinoma. Hum Pathol 22:172–178, 1991. 256. Sanders LM, Botet JF, Straus DJ, et al: CT of primary lymphoma of the liver. AJR Am J Roentgenol 152:973–976, 1989. 257. Sandrasegaran K, Rajesh A, Rushing DA, et al: Gastrointestinal stromal tumors: CT and MRI findings. Eur Radiol 15:1407–1414, 2005. 258. Sano K, Ichikawa T, Motosugi U, et al: Imaging study of early hepatocellular carcinoma: Usefulness of gadoxetic acid-enhanced MR imaging. Radiology 261:834–844, 2011. 259. Sappey C: Recherche sur queiques veines portes accessories. Comput rendus séances. Memoires Soc Biol 11:3–13, 1895. 260. Sasaki M, Yoneda N, Sawai Y, et al: Clinicopathological characteristics of serum amyloid A-positive hepatocellular neoplasms/nodules arising in alcoholic cirrhosis. Histopathology 2014. doi: 10.1111/his.12588. [Epub ahead of print]. 261. Schroeder T, Radtke A, Debatin JF, et al: Contrast-enhanced multidetector-CT cholangiography after living donor liver transplantation. Hepatogastroenterology 54:1176–1180, 2007. 262. Scott L, Blair G, Taylor G, et al: Inflammatory pseudotumors in children. J Pediatr Surg 23:755–758, 1988. 263. Semelka RC, Hussain SM, Marcos HB, et al: Biliary hamartomas: Solitary and multiple lesions shown on current MR techniques including gadolinium enhancement. J Magn Reson Imaging 10:196–201, 1999. 264. Semelka RC, Hussain SM, Marcos HB, et al: Perilesional enhancement of hepatic metastases: Correlation between MR imaging and histopathologic findings—Initial observations. Radiology 215:89–94, 2000. 265. Semelka RC, Sofka CM: Hepatic hemangiomas. Magn Reson Imaging Clin N Am 5:241–253, 1997. 266. Seo S, Hatano E, Higashi T, et al: Fluorine-18 fluorodeoxyglucose positron emission tomography predicts tumor differentiation, P-glycoprotein expression, and outcome after resection in hepatocellular carcinoma. Clin Cancer Res 13:427–433, 2007. 267. Shamsi K, De Schepper A, Degryse H, et al: Focal nodular   hyperplasia of the liver: Radiologic findings. Abdom Imaging 18:32–38, 1993. 268. Sharifi S, Murphy M, Loda M, et al: Nodular lymphoid lesion of the liver: An immune-mediated disorder mimicking low-grade malignant lymphoma. Am J Surg Pathol 23:302–308, 1999. 269. Shinmura R, Matsui O, Kobayashi S, et al: Cirrhotic nodules: Association between MR imaging signal intensity and intranodular blood supply. Radiology 237:512–519, 2005. 270. Siegelman ES, Outwater EK, Furth EE, et al: MR imaging of hepatic nodular regenerative hyperplasia. J Magn Reson Imaging 5:730–732, 1995. 271. Smith GS, Birnbaum BA, Jacobs JE: Hepatic infarction secondary to arterial insufficiency in native livers: CT findings in 10 patients. Radiology 208:223–229, 1998. 272. Smith SL, Ramli NM, Somers JM: Cystic mesenchymal hamartoma mimicking hepatic hydatid disease. Clin Radiol 56:599–601, 2001. 273. Song MJ, Bae SH, Yoo I, et al: Predictive value of 18F-  fluorodeoxyglucose PET/CT for transarterial chemolipiodolization   of hepatocellular carcinoma. World J Gastroenterol 18:3215–3222, 2012. 274. Soyer P, Devine N, Somveille E, et al: Hepatic pseudolesion around the falciform ligament: Prevalence on CT examination. Abdom Imaging 21:324–328, 1996. 275. Soyer P, Gueye C, Somveille E, et al: MR diagnosis of hepatic metastases from neuroendocrine tumors versus hemangiomas: Relative merits of dynamic gadolinium chelate-enhanced gradient-recalled echo and nonenhanced spin-echo images. AJR Am J Roentgenol 165:1407–1413, 1995. 276. Soyer P, Roche A, Levesque M, et al: CT of fibrolamellar hepatocellular carcinoma. J Comput Assist Tomogr 15:533–538, 1991. 277. Steiner PE, Higginson J: Cholangiolocellular carcinoma of the liver. Cancer 12:753–759, 1959.

278. Stoupis C, Taylor HM, Paley MR, et al: The rocky liver: Radiologicpathologic correlation of calcified masses. Radiographics 18:675–685, 1998. 279. Sugano S, Matsuda T, Suzuki T, et al: Hepatic actinomycosis: Case report and review of the literature in Japan. J Gastroenterol 32:672–676, 1997. 280. Sun HY, Lee JM, Shin CI, et al: Gadoxetic acid-enhanced magnetic resonance imaging for differentiating small hepatocellular carcinomas (< or =2 cm in diameter) from arterial enhancing pseudolesions: special emphasis on hepatobiliary phase imaging. Invest Radiol 45(2):96–103, 2010. 281. Tada H, Morimoto M, Shima T, et al: Progressive jaundice due to lymphangiosis carcinomatosa of the liver: CT appearance. J Comput Assist Tomogr 20:650–652, 1996. 282. Taguchi J, Nakashima O, Tanaka M, et al: A clinicopathological study on combined hepatocellular and cholangiocarcinoma. J Gastroenterol Hepatol 11:758–764, 1996. 283. Tajima T, Honda H, Kuroiwa T, et al: Radiologic features of intrahepatic bile duct adenoma: A look at the surface of the liver.   J Comput Assist Tomogr 23:690–695, 1999. 284. Tajima T, Honda H, Taguchi K, et al: Sequential hemodynamic change in hepatocellular carcinoma and dysplastic nodules: CT angiography and pathologic correlation. AJR Am J Roentgenol 178:885–897, 2002. 285. Takayasu K, Furukawa H, Wako F, et al: CT diagnosis of early hepatocellular carcinoma: Sensitivity, findings, and CT-pathologic correlation. AJR Am J Roentgenol 164:885–890, 1995. 286. Takayasu K, Muramatsu Y, Mizuguchi Y, et al: Imaging of early hepatocellular carcinoma and adenomatous hyperplasia (dysplastic nodules) with dynamic CT and a combination of CT and angiography: Experience with resected liver specimens. Intervirology 47:199–208, 2004. 287. Tang Y, Yamashita Y, Ogata I, et al: Metastatic liver tumor cystic ovarian carcinomas: CT and MRI appearance. Radiat Med 17:265–270, 1999. 288. Terada T, Kadoya M, Nakanuma Y, et al: Iron-accumulating adenomatous hyperplastic nodule with malignant foci in the cirrhotic liver: Histopathologic, quantitative iron, and magnetic resonance imaging in vitro studies. Cancer 65:1994–2000, 1990. 289. Terada T, Kitani S, Ueda K, et al: Adenomatous hyperplasia of the liver resembling focal nodular hyperplasia in patients with chronic liver disease. Virchows Arch A Pathol Anat Histopathol 422:247–252, 1993. 290. Terada T, Minato H, Nakanuma Y, et al: Ultrasound visualization of hepatic peribiliary cysts: A comparison with morphology. Am J Gastroenterol 87:1499–1502, 1992. 291. Terada T, Nakanuma Y: Pathological observations of intrahepatic peribiliary glands in 1000 consecutive autopsy livers. III. Survey of necroinflammation and cystic dilatation. Hepatology 12:1229–1233, 1990. 292. Terada T, Nakanuma Y, Kono N, et al: Ciliated hepatic foregut cyst: A mucus histochemical, immunohistochemical, and ultrastructural study in three cases in comparison with normal bronchi and intrahepatic bile ducts. Am J Surg Pathol 14:356–363, 1990. 293. Terayama N, Matsui O, Hoshiba K, et al: Peribiliary cysts in liver cirrhosis: US, CT, and MR findings. J Comput Assist Tomogr 19:419– 423, 1995. 294. Terayama N, Matsui O, Tatsu H, et al: Focal sparing of fatty liver in segment II associated with aberrant left gastric vein. Br J Radiol 77:150–152, 2004. 295. Terayama N, Matsui O, Ueda K, et al: Peritumoral rim enhancement of liver metastasis: Hemodynamics observed on single-level dynamic CT during hepatic arteriography and histopathologic correlation. J Comput Assist Tomogr 26:975–980, 2002. 296. Terayama N, Terada T, Nakanuma Y: A morphometric and immunohistochemical study on angiogenesis of human metastatic carcinomas of the liver. Hepatology 24:816–819, 1996. 297. Theise ND, Nakashima O, Park YN, et al: Combined hepatocellularcholangiocarcinoma. In Bosman FT, Camino F, Hruban RH, et al, editors: WHO classification of tumours of the digestive system, Lyon, 2010, IARC Press, pp 225–227.

CHAPTER 44  Liver: Focal Hepatic Mass Lesions 298. Theise ND, Yao JL, Harada K, et al: Hepatic “stem cell” malignancies in adults: Four cases. Histopathology 43:263–271, 2003. 299. Titelbaum DS, Hatabu H, Schiebler ML, et al: Fibrolamellar hepatocellular carcinoma: MR appearance. J Comput Assist Tomogr 12:588–591, 1988. 300. Tochio H, Kudo M, Okabe Y, et al: Association between a focal spared area in the fatty liver and intrahepatic efferent blood flow from the gallbladder wall: Evaluation with color Doppler sonography. AJR Am J Roentgenol 172:1249–1253, 1999. 301. Torbenson M: Review of the clinicopathologic features of fibrolamellar carcinoma. Adv Anat Pathol 14:217–223, 2007. 302. Trojan J, Schroeder O, Raedle J, et al: Fluorine-18 FDG positron emission tomography for imaging of hepatocellular carcinoma. Am J Gastroenterol 94:3314–3319, 1999. 303. Trojanowski JQ, Harrist TJ, Athanasoulis CA, et al: Hepatic and splenic infarctions: Complications of therapeutic transcatheter embolization. Am J Surg 139:272–277, 1980. 304. Tsui WM, Colombari R, Portmann BC, et al: Hepatic angiomyolipoma: A clinicopathologic study of 30 cases and delineation of unusual morphologic variants. Am J Surg Pathol 23:34–48, 1999. 305. Tung GA, Vaccaro JP, Cronan JJ, et al: Cavernous hemangioma of the liver: Pathologic correlation with high-field MR imaging. AJR Am J Roentgenol 162:1113–1117, 1994. 306. Ueda K, Matsui O, Kadoya M, et al: Pseudolesion in segment IV of the liver on MRI: Prevalence and morphology in 250 cirrhotic livers compared with 250 normal livers. J Comput Assist Tomogr 23:63–68, 1999. 307. Ueda K, Matsui O, Kawamori Y, et al: Differentiation of hypervascular hepatic pseudolesions from hepatocellular carcinoma: Value of single-level dynamic CT during hepatic arteriography. J Comput Assist Tomogr 22:703–708, 1998. 308. Ueda K, Matsui O, Kawamori Y, et al: Hypervascular hepatocellular carcinoma: Evaluation of hemodynamics with dynamic CT during hepatic arteriography. Radiology 206:161–166, 1998. 309. Ueda K, Matsui O, Nobata K, et al: Mucinous carcinoma of the liver mimicking cavernous hemangioma on pre- and postcontrast MR imaging. AJR Am J Roentgenol 166:468–469, 1996. 310. Ueda K, Terada T, Nakanuma Y, et al: Vascular supply in adenomatous hyperplasia of the liver and hepatocellular carcinoma: A morphometric study. Hum Pathol 23:619–626, 1992. 311. Urizono Y, Ko S, Kanehiro H, et al: Primary leiomyoma of the liver: Report of a case. Surg Today 36:629–632, 2006. 312. van Aalten SM, Thomeer MG, Terkivatan T, et al: Hepatocellular adenomas: Correlation of MR imaging findings with pathologic subtype classification. Radiology 261:172–181, 2011. 313. Van Beers B, Roche A, Mathieu D, et al: Epithelioid hemangioendothelioma of the liver: MR and CT findings. J Comput Assist Tomogr 16:420–424, 1992. 314. van den Bos IC, Hussain SM, Terkivatan T, et al: Stepwise carcinogenesis of hepatocellular carcinoma in the cirrhotic liver: Demonstration on serial MR imaging. J Magn Reson Imaging 24:1071–1080, 2006. 315. Vieco PT, Rochon L, Lisbona A: Multifocal cytomegalovirus-associated hepatic lesions simulating metastases in AIDS. Radiology 176:123–124, 1990. 316. Vilgrain V, Boulos L, Vullierme MP, et al: Imaging of atypical hemangioma of the liver with pathologic correlation. Radiographics 20:379–397, 2000. 317. Vilgrain V, Lewin M, Vons C, et al: Hepatic nodules in Budd-Chiari syndrome: Imaging features. Radiology 210:443–450, 1999. 318. Vilgrain V, Van Beers BE, Flejou JF, et al: Intrahepatic cholangiocarcinoma: MRI and pathologic correlation in 14 patients.   J Comput Assist Tomogr 21:59–65, 1997. 319. Vogl TJ, Kummel S, Hammerstingl R, et al: Liver tumors: Comparison of MR imaging with Gd-EOB-DTPA and Gd-DTPA. Radiology 200:59–67, 1996. 320. Wanless IR: Micronodular transformation (nodular regenerative hyperplasia) of the liver: A report of 64 cases among 2500 autopsies

321.

322. 323. 324. 325.

326. 327.

328.

329.

330.

331.

332.

333.

334. 335.

336.

337.

338.

339.

340.

341.

342.

1373

and a new classification of benign hepatocellular nodules. Hepatology 11:787–797, 1990. Wanless IR: Vascular disorders. In MacSween RNM, Burt AD, Portmann BC, et al, editors: Pathology of the liver, ed 4, Glasgow, 2002, Churchill Livingstone, pp 553–555. Weinreb JC, Brateman L, Maravilla KR: Magnetic resonance imaging of hepatic lymphoma. AJR Am J Roentgenol 143:1211–1214, 1984. Weiss SW, Enzinger FM: Epithelioid hemangioendothelioma: A vascular tumor often mistaken for a carcinoma. Cancer 50:970–981, 1982. Weissleder R, Stark DD, Elizondo G, et al: MRI of hepatic lymphoma. Magn Reson Imaging 6:675–681, 1988. Wolfort RM, Papillion PW, Turnage RH, et al: Role of FDG PET in the evaluation and staging of hepatocellular carcinoma with comparison of tumor size, AFP level, and histologic grade. Int Surg 95:67–75, 2010. Wong JJ, Kinney TB, Miller FJ, et al: Hepatic actinomycotic abscesses: Diagnosis and management. AJR Am J Roentgenol 186:174–176, 2006. Woodward PJ, Sohaey R, Kennedy A, et al: From the archives of AFIP: A comprehensive review of fetal tumors with pathologic correlation. Radiographics 25:215–242, 2005. Yamaguchi J, Sakamoto Y, Sano T, et al: Spontaneous regression of inflammatory pseudotumor of the liver: Report of three cases. Surg Today 37:525–529, 2007. Yan F, Zeng M, Zhou K, et al: Hepatic angiomyolipoma: Various appearances on two-phase contrast scanning of spiral CT. Eur J Radiol 41:12–18, 2002. Ye BB, Hu B, Wang LJ, et al: Mesenchymal hamartoma of liver: Magnetic resonance imaging and histopathologic correlation. World J Gastroenterol 11:5807–5810, 2005. Yokomizo J, Cho A, Yamamoto H, et al: Sarcomatous hepatocellular carcinoma without previous anticancer therapy. J Hepatobiliary Pancreat Surg 14:324–327, 2007. Yoneda N, Matsui O, Kitao A, et al: Beta-catenin-activated hepatocellular adenoma showing hyperintensity on hepatobiliary-phase gadoxetic-enhanced magnetic resonance imaging and overexpression of OATP8. Jpn J Radiol 30:777–782, 2012. Yoon KH, Ha HK, Seong J, et al: Inflammatory pseudotumor of the liver in patients with recurrent pyogenic cholangitis: CT-histopathologic correlation. Radiology 211:373–379, 1999. Yoon W, Kim JK, Kang HK: Hepatic undifferentiated embryonal sarcoma: MR findings. J Comput Assist Tomogr 21:100–102, 1997. Yoon KH, Matsui O, Kadoya M, et al: Pseudolesion in segments II and III of the liver on CT during arterial portography caused by aberrant right gastric venous drainage. J Comput Assist Tomogr 23:306–309, 1999. Yoshida K, Kobayashi S, Matsui O, et al: Hepatic pseudolymphoma: Imaging-pathologic correlation with special reference to hemodynamic analysis. Abdom Imaging 38:1277–1285, 2013. Yoshikawa J, Matsui O, Kadoya M, et al: Delayed enhancement of fibrotic areas in hepatic masses: CT-pathologic correlation. J Comput Assist Tomogr 16:206–211, 1992. Yoshikawa J, Matsui O, Takashima T, et al: Focal fatty change of the liver adjacent to the falciform ligament: CT and sonographic findings in five surgically confirmed cases. AJR Am J Roentgenol 149:491–494, 1987. Yoshikawa J, Matsui O, Takashima T, et al: Fatty metamorphosis in hepatocellular carcinoma: Radiologic features in 10 cases. AJR Am J Roentgenol 151:717–720, 1988. Yoshimitsu K, Honda H, Kaneko K, et al: Anatomy and clinical importance of cholecystic venous drainage: Helical CT observations during injection of contrast medium into the cholecystic artery. AJR Am J Roentgenol 169:505–510, 1997. Yoshimura H, Murakami T, Kim T, et al: Angiomyolipoma of the liver with least amount of fat component: Imaging features of CT, MR, and angiography. Abdom Imaging 27:184–187, 2002. Yu JS, Chung JJ, Kim JH, et al: Detection of small intrahepatic metastases of hepatocellular carcinomas using diffusion-weighted imaging: Comparison with conventional dynamic MRI. Magn Reson Imaging 29:985–992, 2011.

1374

PART II  CT and MR Imaging of the Whole Body

343. Yu JS, Kim KW, Sung KB, et al: Small arterial-portal venous shunts: A cause of pseudolesions at hepatic imaging. Radiology 203:737–742, 1997. 344. Yu JS, Yoon SW, Park MS, et al: Eosinophilic hepatic necrosis: Magnetic resonance imaging and computed tomography comparison. J Comput Assist Tomogr 29:765–771, 2005. 345. Zech CJ, Herrmann KA, Reiser MF, et al: MR imaging in patients with suspected liver metastases: Value of liver-specific contrast agent Gd-EOB-DTPA. Magn Reson Med Sci 6:43–52, 2007. 346. Zen Y, Fujii T, Itatsu K, et al: Biliary cystic tumors with bile duct communication: A cystic variant of intraductal papillary neoplasm of the bile duct. Mod Pathol 19:1243–1254, 2006.

347. Zen Y, Fujii T, Sato Y, et al: Pathological classification of hepatic inflammatory pseudotumor with respect to IgG4-related disease. Mod Pathol 20:884–894, 2007. 348. Zhang Y, Uchida M, Abe T, et al: Intrahepatic peripheral cholangiocarcinoma: Comparison of dynamic CT and dynamic MRI.   J Comput Assist Tomogr 23:670–677, 1999. 349. Zheng RQ, Kudo M: Hepatic angiomyolipoma of an efferent vessel to be hepatic vein by contrast-enhanced harmonic ultrasound. Br J Radiol 78:956–960, 2005.

45  Liver Transplantation Raj M. Paspulati

HISTORY Transplantation of the liver is the definitive treatment for irreversible acute and chronic liver disease. The first successful liver transplant was performed by Starlz in 1967 on a patient with hepatocellular carcinoma, who survived for more than 1 year.130 The development of cyclosporine in 1970, followed by other immunosuppressant agents and improved surgical techniques, has significantly improved the survival rates of liver transplant patients, leading to the acceptance of liver transplantation into clinical practice and the creation of several transplant centers in the United States and worldwide. The increased success of liver transplants led to a long waiting list of potential liver recipients owing to a shortage of cadaver livers. This shortage of cadaver livers led to the development of split cadaver transplantation and livingdonor liver transplantation (LDLT). According to data from the Organ Procurement and Transplantation Network (optn.org), from January 1988 to July 2015, 137,039 liver transplants were performed in the United States, 5519 of which were living-donor transplants.152 The United Network for Organ Sharing (UNOS) is a nonprofit scientific organization that acts as the only organ procurement network in the United States. Owing to the long waiting list, the allocation of donor livers is based on the principle that the sickest patient who has been waiting the longest is given priority. Disease severity was initially based on the Child-Turcotte-Pugh score, which had several limitations. The current severity assessment is based on the Model for End-Stage Liver Disease (MELD) scoring system, which includes three parameters: serum bilirubin, international normalized ratio, and serum creatinine. The MELD scoring system is a more accurate index of disease severity and a better assessment of mortality risk in patients with end-stage liver disease.16,34,158

INDICATIONS FOR LIVER TRANSPLANTATION The major indications for liver transplantation are hepatic decompensation secondary to chronic liver disease, acute liver failure, primary hepatic malignancy, and inborn errors of metabolism. Several other miscellaneous indications are listed in Box 45-1. Acute liver failure accounts for 5% to 6% of all liver transplants in the United States.94,144 Chronic liver disease due to hepatitis C infection is the leading indication for liver transplantation in the United States.113 Hepatocellular carcinoma (HCC) is a growing indication for liver transplantation. Chronic liver disease and cirrhosis are associated with a 2% to 8% annual incidence of HCC; patients with cirrhosis associated with HCC are given higher priority for liver transplantation.7,94,122 However, only patients with early HCC without vascular invasion or metastasis are suitable candidates. Early HCC, as defined by the Milan criteria,

includes a single HCC lesion less than 5 cm or no more than three lesions, all of which are less than 3 cm90 (Fig. 45-1). The 5-year survival rate after transplantation in patients with HCC fulfilling these selection criteria is reported to be 75%. There are conflicting results in patients with multifocal and diffuse infiltrative HCC (Fig. 45-2). Locoregional therapy using radiofrequency ablation, cryoablation,  or chemoembolization is being used to reduce the tumor burden  in these patients while they await liver transplantation9,35,49,83,126 (Fig. 45-3). Less common hepatic malignancies with promising survival rates following transplantation include fibrolamellar carcinoma, epithelioid hemangioendothelioma, and hepatoblastoma.4,60,78 Patients with hemangiosarcoma and hepatic metastases, with the exception of metastases from neuroendocrine tumors, have poor outcomes after transplantation. Unresectable cholangiocarcinoma is also treated by liver transplantation after downstaging with aggressive chemoradiation treatment.49,72,94,117,153

CONTRAINDICATIONS TO LIVER TRANSPLANTATION The absolute and relative contraindications for liver transplantation are outlined in Box 45-2. Brain death, extrahepatic malignancy or metastatic disease (Fig. 45-4), and poor compliance are the main absolute contraindications. Portal vein thrombosis is no longer an absolute contraindication to liver transplantation but remains a poor prognostic indicator for postoperative graft function (Fig. 45-5; see also Fig. 45-2). Cholangiocarcinoma without predisposing primary sclerosing cholangitis has a 5-year survival rate of 2% to 43% following curative surgical resection. Patients with unresectable cholangiocarcinoma or preexisting primary sclerosing cholangitis should be considered for liver transplantation.85 Inoperable hilar cholangiocarcinoma treated with neoadjuvant chemoradiation followed by liver transplantation has 1-, 3-, and 5-year survival rates of 92%, 82%, and 82%, respectively.47,100,110,132 Accurate staging of cholangiocarcinoma with crosssectional imaging is crucial for proper patient selection. HIV infection is no longer a contraindication to liver transplantation in patients with end-stage liver disease. Selection of patients is based on CD4 cell count and a viral load that is suppressible with antiretroviral treatment.92,94

EVALUATION OF THE DONOR Liver donors for LDLT are assessed for medical, psychosocial, and financial suitability. The medical evaluation includes a routine presurgical clinical assessment and laboratory and serologic tests. Preoperative imaging is mandatory for the exclusion of focal or diffuse liver lesions and identification of variations in biliary and vascular anatomy. Assessment of segmental and total liver volumes of both recipient and

1375

1376

PART II  CT and MR Imaging of the Whole Body

donor is necessary to ensure adequacy of the donor liver size. Contrastenhanced computed tomography (CT) and magnetic resonance imaging (MRI) with multiplanar reconstruction (MPR) are excellent for pretransplant evaluation of the liver in living-donor candidates. MRI is more specific for the characterization of focal liver lesions and

BOX 45-1  Indications for

Liver Transplantation Cirrhosis Chronic hepatitis B and C Alcoholic liver disease Autoimmune hepatitis Cryptogenic cirrhosis Primary biliary cirrhosis Primary sclerosing cholangitis

the diagnosis of hepatic steatosis. Evaluation of the donor liver for the presence and extent of steatosis is essential because fatty change carries a high risk of postoperative liver dysfunction in donors and graft nonfunction in recipients (Fig. 45-6). The acceptable upper limit of steatosis in the donor liver is 30%.54,71,125 The minimum graft volume required for adequate hepatic function in the recipient is reported to be 30% to 40% of the recipient weight. A donor remnant liver volume of greater than 30% of the original volume is necessary to maintain sufficient liver function in the donor.80,82,154 The reported incidence of variations in the portal vein, hepatic artery, and biliary anatomy is 14%, 29%, and 44%, respectively. Prior knowledge of these variations

BOX 45-2  Contraindications to

Liver Transplantation

Absolute Contraindications MELD score < 15 Advanced cardiopulmonary disease Brain death Extrahepatic malignancy Angiosarcoma of the liver Uncontrolled sepsis Persistent noncompliance Active alcohol and substance abuse Vascular variations incompatible with transplantation Anatomic abnormality that precludes liver transplantation

Acute Liver Failure Hepatitis A and B Drugs and toxins Metabolic Diseases α1-Antitrypsin deficiency Hemochromatosis Wilson’s disease Glycogen storage disorders Malignancy Hepatocellular carcinoma Metastases from neuroendocrine tumors Hepatoblastoma Hepatic epithelioid hemangioendothelioma Cholangiocarcinoma* Miscellaneous Nonalcoholic fatty liver disease Polycystic liver disease Budd-Chiari syndrome *Cholangiocarcinoma is no longer a contraindication for liver transplantation, provided that proper selection criteria are followed.

A

Relative Contraindications HIV infection* Portal vein thrombosis† Hepatopulmonary syndrome Advanced age Morbid obesity Lack of adequate social support system *HIV infection is no longer an absolute contraindication; patient selection is based on CD4 count and HIV-1 RNA viral load. † The extent of portal vein thrombosis should be evaluated by pretransplant imaging. Portal vein thrombosis alone is not a contraindication, but extensive thrombosis with involvement of the superior mesenteric vein is a contraindication.

B FIG 45-1  Hepatocellular carcinomas less than 5 cm in a liver before transplantation. A, Three hyperintense lesions (arrows) are visible on a pre-Gd T1-weighted image. B, Post-Gd T1-weighted image in the equilibrium phase shows central washout with peripheral enhancing capsules of the lesions (arrows).

CHAPTER 45  Liver Transplantation

A

B FIG 45-2  Diffuse infiltrating hepatocellular carcinoma in a liver before transplantation. Post-Gd axial (A) and coronal (B) MRI in the portal venous phase demonstrate a diffuse infiltrating mass (arrowheads) in the right lobe, with portal vein tumor thrombosis (arrows).

A

B

C

D FIG 45-3  Multifocal hepatocellular carcinoma in a liver before transplantation. Axial (A) and coronal (B) contrast-enhanced CT images demonstrate multiple focal enhancing lesions (arrows in A). Arterial-phase contrast-enhanced CT images before (C) and after (D) chemoembolization demonstrate right and left hepatic arteries (arrows) and embolized tumor (arrowheads in D).

1377

1378

PART II  CT and MR Imaging of the Whole Body

A

B FIG 45-4  Hepatocellular carcinoma with adrenal metastases in a liver before transplantation. A, Axial contrast-enhanced MRI in the arterial phase demonstrates a heterogeneously enhancing mass (arrowheads) in the right lobe and a new focal enhancing nodule (arrow) in the right adrenal gland. B, Axial contrastenhanced MRI in the portal venous phase shows an enhancing right adrenal nodule (arrow).

LIVER TRV MPV

A × ×

A

B

C

D FIG 45-5  Cirrhosis with portal vein thrombosis. A, Grayscale US demonstrates an enlarged main portal vein (arrows). B, Color-flow Doppler shows no flow in the main portal vein (white arrow) with periportal collateral veins (black arrow). Coronal (C) and axial (D) post-Gd MRI shows a thrombosed main portal vein (white arrows) and periportal collateral veins (black arrows).

CHAPTER 45  Liver Transplantation

A

1379

B FIG 45-6  Hepatic steatosis of a donor liver. In-phase (A) and opposed-phase (B) T1-weighted images demonstrate a diffuse drop in signal intensity of the liver on the opposed-phase image.

is useful to the operating surgeon in planning the vascular and biliary anastomosis because most posttransplantation vascular complications occur at the anastomosis.21,138 In adult-to-adult LDLT, the hepatectomy plane is to the right of the middle hepatic vein and involves preservation of the segment artery, left portal vein, and middle hepatic vein.

Hepatic Artery Anatomy and Variations The standard hepatic artery anatomy, with proper branching to right and left hepatic arteries, is seen in approximately 55% of patients. The hepatic artery variations were first classified by Michels in 1966.91 The most common variations described are the replaced or accessory right hepatic artery (RHA), arising from the superior mesenteric artery, and the replaced or accessory left hepatic artery (LHA), arising from the left gastric artery (Fig. 45-7A to D). A replaced RHA or LHA is not an absolute contraindication for donor selection, but additional measures must be taken to perform a safe anastomosis. An accessory RHA or LHA, in contrast, needs dual anastomoses in the recipient unless there is an intraparenchymal anastomosis between the two arteries (see Fig. 45-7E). A short RHA in the donor is significant because it makes performing the anastomosis in the recipient difficult and requires extensive reconstructive surgery. Origin of the RHA or LHA from the common hepatic artery proximal to the gastroduodenal artery can compromise perfusion of the donor stomach and duodenum after ligation of the common hepatic artery. Another significant variation in the donor is origin of the artery to segment 4 from the RHA (middle hepatic artery). Identification of this arterial branch to segment 4 is important in adult-to-adult LDLT because the hepatectomy line would cross the arterial supply to this segment, threatening the segment’s viability (see Fig. 45-7F). Documentation of acquired conditions that can compromise arterial inflow (e.g., celiac artery stenosis) can assist the surgeon in planning vascular reconstruction.44,45,75,93,94

Portal Vein Anatomy and Variations A normal portal vein bifurcates at the hilum into right and left main branches (Fig. 45-8A). The right portal vein (RPV) bifurcates into the anterior and posterior branches, which in turn bifurcate into ascending and descending branches, each supplying a segment of the right lobe. The left portal vein (LPV) divides into three branches, each supplying a segment of the left lobe. Variations in portal vein anatomy are seen in approximately 10% to 20% of patients. Variations in the RPV branching pattern are more common. An early bifurcation of the RPV and trifurcation of the main portal vein (MPV) into anterior and posterior divisions and the LPV are the most common variations (see Fig. 45-8B). The anterior division of the RPV may arise from the LPV,

and its origin may be intraparenchymal or extraparenchymal. An undivided MPV with all segmental branches arising directly from the MPV is a rare anatomic variation. Intraparenchymal origin of the anterior division of the RPV from the LPV and undivided MPV variants are absolute contraindications for right lobe donation. Trifurcation of the MPV and extraparenchymal branching of the anterior division of the RPV from the LPV may require surgical reconstruction of two venous anastomoses.20,23,45,66

Hepatic Vein Anatomy and Variations Normal hepatic vein anatomy consists of three major hepatic veins that drain into the inferior vena cava (IVC). The right hepatic vein (RHV) drains hepatic segments 5 to 8; the middle hepatic vein (MHV) drains segments 4, 5, and 8; and the left hepatic vein (LHV) drains segments 2 and 3. A common variation seen in nearly 60% of patients is a common opening of the MHV and LHV into the IVC (Fig. 45-9A). Variations in hepatic vein anatomy that are significant for both donor and recipient are seen in approximately 30% of patients. These variations challenge the operating surgeon to modify the hepatectomy plane to avoid excessive bleeding and preserve venous drainage in the donor as well as the recipient. The MHV is an important surgical landmark for the hepatectomy plane in both right lobe and left lateral segment grafts. Early bifurcation of the MHV and large branches draining into the MHV from adjacent segments of the right lobe may require alteration of the hepatectomy plane. Identification of accessory hepatic veins is essential because they may traverse the hepatectomy plane and cause significant hemorrhage and graft ischemia. All accessory hepatic veins larger than 5 mm have to be identified because they require reimplantation to avoid graft congestion and rejection. An accessory right inferior hepatic vein that drains the posterior segments (6 and 7) of the right lobe is an important variant that should be identified, and its distance from the main RHV should be measured (see Fig. 45-9B  and C). If these two veins are separated by more than 4 cm, it may be difficult to implant both veins with a single partially occluding clamp on the recipient’s IVC. A left lobe venous variant that may alter the surgical approach is an accessory LHV that drains directly into the IVC or MHV.20,45,58,89

Intrahepatic Biliary Duct Anatomy and Variations The right hepatic duct is formed by the confluence of the right posterior duct draining segments 6 and 7 and the right anterior duct draining segments 5 and 8. The left hepatic duct is formed by the segmental branches draining segments 2 to 4. The right hepatic duct, which is shorter, joins the left hepatic duct to form the common hepatic duct

1380

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F FIG 45-7  Arterial variations in a liver donor. A, Replaced right hepatic artery (RHA) (arrowhead) arising from the superior mesenteric artery (SMA) (arrow). B, Hepatic artery (white arrow), splenic artery (white arrowhead), and SMA (black arrowhead) arising from a common trunk (black arrow). C and D, Replaced left hepatic artery (LHA) (arrowhead) arising from the left gastric artery (curved arrow) and the RHA (arrow) arising from the celiac trunk. E, Accessory LHA (curved arrow) arising from the left gastric artery, with a normal LHA (arrowhead) and RHA (arrow). F, Artery to segment 4 (curved arrow) arising from the RHA (arrowhead); the LHA is shown (arrow).

CHAPTER 45  Liver Transplantation

1381

SURGICAL TECHNIQUES Cadaver Liver Transplantation

A

B FIG 45-8  Portal vein in the donor liver. A, Normal main portal vein (arrow) with right and left portal vein branches (arrowheads). B, Trifurcation of the main portal vein (arrow).

(CHD). The cystic duct joins the CHD distal to the confluence of right and left hepatic ducts. This normal biliary anatomy is present in about 58% of the population (Fig. 45-10A).77,109 The most common variant of biliary anatomy, seen in about 13% to 19% of the population, is drainage of the right posterior duct into the left hepatic duct (see Fig. 45-10B).38 Another common variant, seen in about 11% of the population, is the triple confluence of the right anterior duct, right posterior duct, and left hepatic duct to form the CHD.109 A less common variation, seen in about 5% of the population, is direct drainage of the right posterior duct into the CHD or common bile duct (CBD).38,109 Accessory hepatic ducts are present in approximately 2% of the population.24,109 Common cystic duct variations are low insertion of the cystic duct into the distal third of the extrahepatic bile duct in 9%, medial insertion of the cystic duct into the left side of the CHD in 10% to 17%, and a parallel course of the cystic duct and CHD.109,134 Uncommon cystic duct variations include a more proximal opening of the cystic duct into the CHD, aberrant opening of the cystic duct into either the left or right hepatic ducts, and drainage of an aberrant or accessory posterior duct into the cystic duct.50 The presence of these bile duct variations is not an absolute contraindication for transplantation, but preoperative knowledge of them helps the surgeon make appropriate alterations in the resection plane during graft retrieval and ductal anastomosis to avoid potential complications, including bile leaks and strictures.

The surgical technique for cadaver liver transplantation is well established, with minor individual and institutional differences. The basic surgical steps consist of the hepatectomy phase, anhepatic phase with engraftment (with or without venovenous bypass), reperfusion with arterialization, and biliary reconstruction.84,115,129 The vascular anastomosis proceeds in the following order: suprahepatic vena cava, infrahepatic vena cava, portal vein, hepatic artery. The hepatic artery is reconstructed by anastomosis between the donor celiac axis and the recipient hepatic artery bifurcation or the branch point of the gastroduodenal and proper hepatic arteries (Fig. 45-11A and B). The technique of arterial anastomosis depends on the presence of anatomic variations. If the recipient hepatic artery is inadequate, donor iliac artery graft is used to anastomose the hepatic artery to the aorta (see Fig. 45-11C). The donor and recipient portal veins are connected via an end-to-end anastomosis (Fig. 45-12). In recipients with an inadequate portal vein owing to either thrombosis or previous surgery, a venous jump graft from the donor portal vein or iliac vein is used. There are different techniques of IVC anastomosis. The conventional method includes resection of the retrohepatic segment of the recipient IVC and an end-to-end anastomosis of the donor IVC to the suprahepatic and infrahepatic ends of the recipient IVC (Fig. 45-13A to C). This technique may require a temporary venovenous bypass in the recipient between the iliac veins and superior vena cava through  the axillary vein. A new technique (piggyback) involves preserving the recipient retrohepatic IVC and creating an anastomosis between the donor IVC and a common stump of the recipient hepatic veins (see Fig. 45-13D). The piggyback technique requires greater surgical skill but has several advantages: avoidance of venovenous bypass, less bleeding, and less chance of adrenal or renal vein injury.111 After the vascular anastomosis, the biliary anastomosis is performed between the donor CBD and the recipient CHD after a cholecystectomy (Fig. 45-14A and B). The choledochocholedochostomy (CDC) can be performed in endto-end or side-to-side fashion. If the recipient duct is diseased or inadequate, a Roux-en-Y hepaticojejunostomy or choledochojejunostomy is performed (see Fig. 45-14C).10,84,115 The CDC technique is preferable because it provides physiologic bile flow to the duodenum with preservation of the sphincter of Oddi function, avoids intestinal contamination of the bile ducts, and allows endoscopic diagnostics and interventions.

Living-Donor Liver Transplantation LDLT is more technically challenging than whole cadaver liver transplantation. LDLT initially consisted of left lateral segment liver grafts from an adult donor to a pediatric recipient. The success of these procedures led to the use of right lobe grafts from living donors to adult recipients.133 There are four types of LDLT—left lateral segment graft, left lobe graft, right lobe graft, and extended right lobe graft— depending on the age and size of the recipient, the original disease and clinical status of the recipient, and the quality of the donor liver. The graft size is the major limiting factor in the outcome of LDLT. A minimum graft-to-recipient weight ratio of 0.8% is sufficient to meet metabolic demands after transplantation.8,100 The left lateral segment graft (segments 2 and 3) is used for small pediatric recipients, and a left lobe graft (segments 2 to 4) is used for larger children. The line of parenchymal transection is to the right of the falciform ligament in lateral segment grafts and 1 to 2 cm to the right of Cantlie’s line in left lobe grafts (Fig. 45-15A). Cantlie’s line runs from the gallbladder fossa to the IVC. Intraoperative ultrasonography (US) is useful to identify the MHV in the left lobe graft resection. A long length of the LHV is

1382

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 45-9  Hepatic vein variations in the donor liver. A, Contrast-enhanced axial MRI showing confluence (arrows) of the middle and left hepatic veins and IVC (arrowhead). B and C, Coronal images show the right hepatic vein (arrow) and accessory inferior hepatic vein (arrowhead).

preserved for anastomosis with a common stump of the LHV and MHV of the recipient. The graft portal vein is anastomosed with the LPV, portal vein bifurcation, or the MPV of the recipient. For adult recipients, either a right lobe (segments 5 to 8) or extended right lobe (segments 4 to 8) graft is used. The plane of resection is 1 cm to the right of the MHV for right lobe grafts, and the MHV is included in the graft for extended right lobe grafts (see Fig. 45-15B).43 Extended grafts are used when the graft liver is of limited quality owing to steatosis or elderly age of the donor. Assessment of the venous drainage of segment 4 and any variations in MHV anatomy is important in planning the resection plane for right lobe grafts.86 It is also important to assess preoperatively whether the donor will have adequate liver volume left after extended right lobe donation, because the volume of the left lobe lateral segments is variable. This can be assessed with CT volumetry, and 35% of liver volume is considered to be a safe cutoff point for donation.53 This percentage volume of residual liver is increased in aged donors and those with steatosis.76 Fatty livers are more vulnerable to injury by general anesthesia and ischemiareperfusion. Donor livers with steatosis of 20% or more, especially when the residual liver volume is less than 30%, are excluded from right lobe donor hepatectomy.31,69,88

POSTTRANSPLANT COMPLICATIONS Improved surgical techniques and early detection and correction of postoperative complications have significantly improved both graft and patient survival rates. According to the UNOS website (unos.org), the 1-, 5-, and 10-year graft survival rates of liver transplant recipients

who were younger than 65 years were 82.1%, 67.8%, and 52.6%, respectively; for recipients 65 years and older they were 77.5%, 59.7%, and 41.2%, respectively. Imaging plays an important role in the timely detection and treatment of complications. Imaging modalities for evaluation of the transplanted liver include US and cross-sectional imaging with either contrast-enhanced CT or MRI. Each of these modalities has advantages and limitations. US is the preferred imaging modality in the immediate postoperative period.27,37 The study can be performed at the patient’s bedside to detect perihepatic fluid collections, biliary dilatation, and vascular flow to the liver. Color-flow Doppler, with evaluation of the spectral waveform, provides baseline information about the perfusion of the transplanted liver. Although US is useful, it has several limitations. There is a limited imaging window available in the immediate postoperative period owing to extensive bowel gas, pneumoperitoneum, and perihepatic fluid collections. Doppler evaluation is also difficult owing to the patient’s inability to cooperate with breath holding. The hepatic arterial anastomosis site is technically difficult to evaluate, and the status of the arterial anastomosis is based on the spectral waveform and resistive index of the downstream arterial flow. Contrast-enhanced multidetector CT (MDCT) in the hepatic arterial phase and portal venous phase is ideal for evaluating the transplanted liver. In addition to arterial and venous anatomy, hepatic parenchymal enhancement and perihepatic hematomas and other fluid collections are well delineated.65 However, biliary anatomy and complications are not well evaluated by contrast-enhanced CT. Magnetic resonance cholangiopancreatography (MRCP) and gadolinium (Gd)-enhanced MRI can demonstrate vascular, biliary, and

CHAPTER 45  Liver Transplantation

1383

BOX 45-3  Posttransplant

Liver Complications

A

B FIG 45-10  Bile duct variations in the donor liver. A, MRCP demonstrates normal biliary anatomy, including the common bile duct (arrow), common hepatic duct (curved arrow), and cystic duct (arrowhead). B, Crossover anomaly of the right posterior hepatic duct (arrow) draining into the left hepatic duct (arrowhead).

parenchymal complications.101,119 Differentiation of perihepatic hematoma from other postsurgical fluid collections is also an advantage of MRI. Conventional angiography and cholangiography are used only for the management of vascular and biliary complications. Although CT and MRI are better modalities for imaging a transplanted liver, duplex sonography remains the primary imaging modality in the immediate postoperative period and during follow-up. CT and MRI are generally used when sonography findings are equivocal or there is a discrepancy between the clinical presentation and the sonographic findings. Post–liver transplant complications are broadly classified into vascular and nonvascular complications (Box 45-3).

Vascular Complications Vascular complications are more serious and are related to the hepatic artery, portal vein, and IVC reconstruction. Hepatic artery and portal vein thrombosis is the most important vascular complication, with a potential for graft failure and need for retransplantation.

Vascular Complications Hepatic artery Hepatic artery stenosis Hepatic artery thrombosis Hepatic artery pseudoaneurysm Portal vein Stenosis Thrombosis Inferior vena cava Stenosis Thrombosis Other Arterioportal fistula Splenic artery aneurysm and rupture Biliary Complications Bile leak Biliary obstruction Bile duct stricture Anastomotic stricture Nonanastomotic stricture Biliary calculi Cholangitis Extrinsic compression by fluid collections Recurrent cholangiocarcinoma Mucocele of the cystic duct remnant Sphincter of Oddi dysfunction Fluid Collections Hematoma Biloma Seroma Abscess Hepatic Parenchymal Complications Ischemia and infarction Abscess secondary to ischemic necrosis and opportunistic infections Adrenal Hemorrhage Malignancies Posttransplant lymphoproliferative disease Recurrent hepatocellular carcinoma Recurrent cholangiocarcinoma Skin cancer, Kaposi’s sarcoma Other malignancies (e.g., upper airway, lung, gastrointestinal, cervical cancer) Extrahepatic Opportunistic Infections Cytomegalovirus Tuberculosis Aspergillus Pneumocystis jiroveci infection

Hepatic Artery Complications.  Hepatic artery complications include thrombosis, stenosis, and pseudoaneurysm formation. Stenosis and thrombosis.  Stenosis and thrombosis are interrelated, with hepatic artery stenosis generally leading to thrombosis. Early hepatic artery thrombosis develops within the first 2 weeks after surgery. Delayed hepatic artery thrombosis develops several years after transplantation as a sequela of chronic rejection and sepsis. Hepatic artery thrombosis occurs more frequently in children (40%) than in adults (4%-12%).39,61 The clinical presentation of hepatic artery

A

B

FIG 45-11  Normal arterial anastomosis in a transplanted liver. Coronal (A) and coronal oblique (B) arterial-phase CT images demonstrate hepatic artery anastomosis between the recipient and donor common hepatic arteries (arrow) and right and left hepatic artery branches (arrowheads) of the transplanted liver. Coronal Gd-enhanced MR angiography (C) demonstrates hepatic artery anastomosis to the infrarenal aorta through an arterial conduit (arrow).

C

ANAS

A B

FIG 45-12  Normal portal vein anastomosis in a transplanted liver.

C

US with Doppler (A), coronal reconstructed CT (B), and contrastenhanced coronal MRI (C) demonstrate normal portal vein anastomosis (arrow).

CHAPTER 45  Liver Transplantation

ANAS

ANAS

B

A

C

D FIG 45-13  Normal IVC anastomosis in a transplanted liver. Color-flow Doppler (A), spectral Doppler (B), and coronal post-Gd MRI (C) demonstrate a conventional end-to-end IVC anastomosis (arrows). In a line diagram (D) showing the piggyback technique of IVC anastomosis, the preserved recipient retrohepatic IVC (straight black arrows), ligated donor IVC (curved black arrows), ligated recipient right hepatic vein (arrowhead), and vascular cuff of the recipient middle and left hepatic veins (red arrows) are anastomosed to the donor IVC.

1385

1386

PART II  CT and MR Imaging of the Whole Body

B A

C FIG 45-14  Bile duct anastomosis. Line diagram (A) and corresponding MRCP image (B) of a transplanted liver with choledochocholedochostomy (arrow in B). Line diagram (C) of choledochojejunostomy (black arrow). The portal vein anastomosis (curved arrow), hepatic artery anastomosis (red arrow), and T-tube (arrowhead) are illustrated in the line diagrams.

thrombosis can vary from catastrophic fulminant hepatic failure to a totally silent event. Other presentations include biliary obstruction, bile leak, or sepsis. Biliary complications result from ischemic necrosis of the bile ducts, which are entirely dependent on hepatic arterial supply, and manifest a few days to weeks after transplantation. Several anatomic or hemodynamic factors can increase the risk for thrombosis. Anatomic factors include significant size discrepancy between the donor and recipient arteries, internal flaps, prolonged clamping of the artery, kinking of a long artery, prolonged cold ischemia, ABO blood type incompatibility, and rejection.97 Hemodynamic factors include hypercoagulable state in the first postoperative week and graft edema from hepatic failure and celiac artery stenosis. Other

predisposing factors include interpositional conduit, technical errors, acute rejection, and cytomegalovirus infection. Because the nature of the arterial reconstruction has an effect on the prevalence of arterial thrombosis, an illustration of the type of vascular anastomosis should be provided to the radiologist by the operating surgeon.66,151 In all posttransplant patients with elevated hepatic enzymes and unexplained bacteremia, hepatic artery thrombosis should be excluded by imaging. Because thrombosis occurs at the anastomosis, US is insensitive in demonstrating the actual site of thrombosis, and the diagnosis is based on Doppler flow changes in the distal intrahepatic branches. Absence of flow in the proper hepatic artery at the porta hepatis and intrahepatic branches is diagnostic of complete hepatic

CHAPTER 45  Liver Transplantation

A

1387

B FIG 45-15  Lines of transection in living-donor liver transplantation. Coronal (A) and axial (B) contrastenhanced MRI demonstrates the lines of transection in a left lateral segment graft (yellow line) and a right lobe graft (red line).

artery thrombosis. A tardus parvus pattern of flow can be identified in the intrahepatic branches distal to partial thrombosis or significant stenosis (>50%) and in complete occlusion with collateral vessel formation. The tardus parvus flow is characterized by low systolic velocity with a resistive index of less than 0.5 and an increased acceleration time of greater than 0.08 second (Figs. 45-16A and 45-17A). A peak systolic velocity of more than 200 cm/sec at the anastomosis is diagnostic of significant stenosis. There is wide variation in the reported success rates of Doppler sonography in detecting hepatic artery stenosis. The reported sensitivity and specificity vary from 91% to 97% and 64% to 99%, respectively.10,28,56,107,140,150 A false-positive sonographic finding of hepatic artery thrombosis is seen with postsurgical edema, systemic hypotension, and high-grade hepatic artery stenosis. Development of periportal arterial collaterals in chronic hepatic artery thrombosis may result in false-negative Doppler findings. In all patients with equivocal sonographic findings, contrastenhanced CT or MRI is invaluable for the diagnosis of hepatic artery thrombosis (see Figs. 45-16B and 45-17B).14,66,67 As a diagnostic criterion for hepatic artery thrombosis, an abrupt cutoff of the hepatic artery at the anastomosis, with nonvisualization of the intrahepatic branches, has a low sensitivity (57%) and high specificity (99%). Using focal narrowing at the anastomosis as a diagnostic criterion, a sensitivity of 100% and specificity of 74% are possible. There is often overestimation of stenosis severity with MRI owing to postsurgical changes, susceptibility artifacts from surgical clips, and low-flow states.62 Lowflow states result from systemic hypotension or regional causes such as increased resistance to hepatic flow from edematous liver or decreased hepatic blood volume due to splenic artery steal phenomenon. These low-flow states cause decreased enhancement of intrahepatic arteries. Treatment of hepatic artery thrombosis depends on the timing after transplantation. Acute thrombectomy and revision of the anastomosis within the first 2 weeks have a success rate of about 50%, and most patients eventually require retransplantation. Management of hepatic artery thrombosis with interventional radiology includes intraarterial thrombolysis with urokinase.63,70,149,163 Hepatic artery stenosis of the anastomosis is managed with either balloon dilation or stent placement.68,114 These measures are palliative and provide a bridge for retransplantation. Hepatic artery pseudoaneurysm.  Hepatic artery pseudoaneurysm is an uncommon complication and can be extrahepatic or intra-

hepatic. Extrahepatic pseudoaneurysms develop at the anastomosis, and intrahepatic pseudoaneurysms are a sequela of prior intervention, such as biopsy or biliary intervention, and sepsis. These pseudoaneurysms may be asymptomatic, detected incidentally on imaging, but they are a potential source of rupture and fatal hemorrhage. Rupture of an intrahepatic pseudoaneurysm can lead to a fistulous communication with an adjacent portal vein or bile duct. Patients with arteriobiliary fistula present with hemobilia and upper gastrointestinal bleeding.81,87 The diagnosis can be made by contrast-enhanced CT or MRI, which shows an enhancing cystic lesion in relation to the artery10,14,64,97 (Fig. 45-18). US demonstrates a periportal or intrahepatic cystic structure with arterial flow on color and spectral Doppler analysis.142,156 All intrahepatic and perihepatic cystic structures should be evaluated with Doppler to exclude a pseudoaneurysm. The management options for extrahepatic pseudoaneurysm at the anastomosis are surgical excision and anastomosis, embolization, and exclusion by stent placement. Intrahepatic pseudoaneurysm can be successfully managed with endovascular coil embolization.5,30

Portal Vein Complications.  Portal vein thrombosis and stenosis at the anastomosis are uncommon complications reported in 1% to 3% of patients.62,97 These complications can develop in the immediate postoperative period or several months to years after transplantation and manifest clinically with symptoms of portal hypertension, massive ascites, edema, and hepatic failure.97 Factors predisposing to portal vein thrombosis are faulty surgical technique, vessel misalignment, differences in the caliber of donor and recipient veins, previous portal vein surgery, previous portal vein thrombosis in the recipient portal vein system, and hypercoagulable states.143,157 US demonstrates complete occlusion, filling defects, and narrowing of the portal vein (Fig. 45-19A). Mild narrowing at the anastomosis is natural; significant stenosis is diagnosed when the peak velocity exceeds 200 cm/sec and the mean anastomotic-to-preanastomotic velocity ratio is greater than 3 : 1 (see Fig. 45-19B).25,74,128 A decrease in the hepatic artery resistive index to less than 0.5 is a secondary sign of portal vein thrombosis.106 CT and MRI demonstrate portal vein thrombosis as filling defects with narrowing of the lumen, and the extent of thrombosis and stenosis is well delineated (Fig. 45-20A; also see Fig. 45-19C). More than 50% narrowing of the portal vein has 100% sensitivity and 84% specificity in the diagnosis of portal vein

1388

PART II  CT and MR Imaging of the Whole Body

A

A

B FIG 45-17  Hepatic artery thrombosis. A, Color-flow Doppler US shows

B FIG 45-16  Hepatic artery stenosis. A, Color-flow Doppler US of a transplanted liver shows low-resistance arterial flow in the liver, with a resistive index of less than 0.5. B, Coronal Gd-enhanced MRI demonstrates focal high-grade stenosis at the hepatic artery anastomosis (arrow), with poststenotic dilatation (arrowhead).

stenosis. The extent and severity of the portosystemic collateral vessels from portal hypertension due to significant portal vein stenosis are well delineated with both CT and MRI (see Fig. 45-20B).10,39,61 Percutaneous transhepatic portography can be performed when there is a discrepancy between noninvasive imaging findings and the clinical presentation. This allows measurement of a pressure gradient across the stenosis, with values higher than 5 mm being significant.105 Treatment options include surgical thrombectomy, creation of a venous jump graft, portosystemic shunt, and percutaneous transluminal angioplasty with or without stent placement. Patients refractory to these treatments require retransplantation.

Inferior Vena Cava Thrombosis and Stenosis.  IVC complications following liver transplantation are uncommon, with a prevalence of less than 1%.39 In a standard technique with suprahepatic and infrahepatic IVC anastomosis, stenosis can occur at either anastomosis. Suprahepatic IVC stenosis presents with clinical features of Budd-

low-resistance arterial flow in the liver, with a resistive index of 0.43. B, Coronal Gd-enhanced 3D MR angiogram demonstrates segmental narrowing (arrow) of the hepatic artery at the anastomosis.

Chiari syndrome, including hepatic venous congestion, ascites, decreased renal function, and peripheral edema. Infrahepatic stenosis presents with lower limb edema and decreased renal function without hepatomegaly or ascites. Secondary thrombosis of the IVC and hepatic veins can develop following suprahepatic IVC stenosis. In the piggyback technique, the recipient IVC is retained with an end-to-side anastomosis of the common cuff of the donor hepatic vein and the recipient IVC. Less commonly, stenosis can occur at this end-to-side anastomosis.95,102 Stenosis at the anastomosis can result from faulty surgical technique, swelling of the transplanted liver, kinking of the suprahepatic IVC from rotation of the liver, and discrepancy between the donor and recipient IVCs. In the case of IVC thrombosis, US may demonstrate IVC narrowing and echogenic intraluminal thrombus. Doppler evaluation at the anastomosis demonstrates a threefold increase in velocity compared with that in the prestenotic segment and monophasic flow in the IVC and hepatic veins proximal to the stenosis.25,112 The severity and extent of stenosis and thrombosis are well demonstrated on coronal and sagittal reconstructed CT and MRI10,39,61 (Figs. 45-21A and 45-22). IVC stenosis can be treated initially with percutaneous transluminal angioplasty and with stent placement if angioplasty is unsuccessful (see Fig. 45-21C).12,147,161

CHAPTER 45  Liver Transplantation

A

B

D

C

F

E FIG 45-18  Hepatic artery pseudoaneurysm and arteriobiliary fistula. This patient with a transplanted liver presented with upper gastrointestinal bleeding following a left hepatic lobe biopsy. Ultrasonography (A) and contrast-enhanced CT (B) show intrahepatic biliary dilatation (arrows). ERCP (C) demonstrates a dilated left hepatic duct with a filling defect (arrow). Selective hepatic angiogram in the arterial phase (D) demonstrates a small pseudoaneurysm (arrow) in relation to a peripheral branch of the left hepatic artery; a subsequent image (E) shows contrast filling of a dilated left hepatic bile duct (arrows) and the left hepatic artery (curved arrow), indicating an arteriobiliary fistula. The patient was successfully treated with coil embolization (arrow in F) of the left hepatic artery branch.

1389

1390

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 45-19  Portal vein stenosis. This liver transplant patient presented with recurrent portal hypertension. Color-flow (A) and spectral (B) Doppler US images demonstrate turbulent flow with high velocity at the portal vein anastomosis (arrow in A). Coronal reconstructed CT in the portal venous phase (C) demonstrates focal stenosis (arrow) of the main portal vein at the anastomosis.

Other Vascular Complications Splenic artery aneurysm.  The incidence of splenic artery aneurysm is higher in patients with cirrhosis and portal hypertension and varies from 7% to 10%. This increased incidence is presumably due to the hyperkinetic state of the spleen in portal hypertension. The risk of splenic artery aneurysm is persistently high after liver transplantation, with a reported prevalence of about 13%. There is also increased risk of rupture of these aneurysms after transplantation, especially if they are larger than 1.5 cm, owing to decreased portal venous flow and increased splenic artery flow. Therefore any splenic artery aneurysm detected during pretransplant evaluation of a recipient should be resected at the time of transplantation. Posttransplant splenic artery aneurysms tend to be multiple and occur more frequently in the distal third of the splenic artery.14,69,99 Intrahepatic arterioportal fistula.  Intrahepatic arterioportal fistula is a complication that develops in a transplanted liver following a surgical or percutaneous liver biopsy to exclude graft rejection. The reported incidence of these fistulas is 50% in the first week after the biopsy, and they tend to resolve spontaneously. These fistulas are small and are not usually detected by color-flow Doppler US. They have a characteristic appearance on contrast-enhanced CT and MRI; they are seen as peripheral portal branch enhancement in the arterial phase, without enhancement of the main portal vein and superior mesenteric vein. With very small arterioportal fistulas, there is increased focal, peripheral, wedge-shaped parenchymal enhancement in the arterial phase compared with the rest of the normal hepatic parenchyma; this is referred to as transient hepatic attenuation difference.22

Diffuse periportal hypodensity.  Diffuse periportal hypodensity on contrast-enhanced CT and hyperintensity on T2-weighted MRI in a transplanted liver are due to dilated lymphatics from impaired drainage into the extrahepatic lymphatic system (Fig. 45-23). This resolves with the development of alternative pathways of lymphatic drainage, which may take weeks to months. This is not a sign of graft rejection, as once believed, and should not be mistaken for dilated bile ducts.73 Splenic artery steal syndrome.  Splenic artery steal syndrome is characterized by arterial hypoperfusion of the transplant liver due to shift of blood flow from celiac trunk into splenic artery. It has a reported incidence of 3% to 8% after orthotopic liver transplant. It is thought to be due to hepatic arterial buffer response secondary to increased portal venous flow. Its onset may be as early as a few hours after the transplant to several weeks after transplant. Decreased hepatic arterial perfusion of the transplant liver may lead to ischemic biliary disease and graft loss. Doppler US demonstrates decreased peak systolic velocity of intrahepatic arteries with or without corresponding increase in resistivity index (RI). Celiac angiography demonstrates a patent hepatic artery but delayed and reduced contrast filling of the hepatic artery and its branches as compared to other branches of celiac trunk. Treatment is by selective arterial embolization of the proximal splenic artery.145

Biliary Complications Biliary complications have been a common cause of morbidity from the early days of orthotopic liver transplantation. The incidence of biliary complications has decreased owing to improved surgical

CHAPTER 45  Liver Transplantation

A

B FIG 45-20  Portal vein thrombosis. A, Coronal Gd-enhanced MRI in the portal venous phase shows thrombosis of the entire main portal vein (arrows), with dilated collaterals in the left upper quadrant (arrowheads). B, Delayed coronal MRI demonstrates dilated splenic collateral veins (arrowheads), with a splenorenal shunt and dilated left renal vein (arrow).

techniques, and the current reported incidence is 10% to 15%.46,104 However, with the wide use of split-liver, reduced-size, and livingdonor transplants, there has been a resurgence of biliary complications. The incidence of biliary complications is higher in LDLT, with multiple biliary reconstructions rather than a single biliary anastomosis. Irrespective of the type of biliary reconstruction, the two most common biliary complications are bile leak and bile stricture.40,52,128 Less common complications include bile duct obstruction from kinking, stone, or sludge; sphincter of Oddi dysfunction; and mucocele of the cystic duct remnant.159

Bile Leak.  The reported incidence of bile leak varies from approximately 6% to 10% following orthotopic liver transplantation. The clinical presentation of bile leak is variable and includes excess bile from intraabdominal drains, abdominal pain, elevated serum bilirubin and liver enzymes, and fever. Because the clinical presentation and laboratory findings are nonspecific, radiographic imaging and intervention play a crucial role in the early diagnosis and management of bile leaks. Types.  There are three types of bile leaks: leaks at the T-tube insertion, anastomotic leaks, and nonanastomotic leaks. Bile leak most commonly occurs at the T-tube site after its removal. It can be caused by errors in tube placement or displacement of the tube outside the duct. It is not associated with vascular complications

1391

and bile duct ischemia. The degree of bile leak varies from a small leak along the T-tube track to a significant leak with perihepatic biloma and bilious ascites. There is evidence of an increased incidence of bile leak with T-tube placement, so the new surgical strategy is to avoid T-tube placement after biliary reconstruction.96,98 Anastomotic leaks generally occur in the first month after transplantation. They are commonly the result of faulty surgical technique, with excessive tension on the biliary anastomosis and dissection around the duct, with ischemia and necrosis. This ischemia is most often due to local periductal dissection and less commonly to hepatic artery thrombosis. Anastomotic leaks can occur at both choledochocholedochostomy and choledochojejunostomy sites, but a much higher incidence is reported with the latter. This is believed to be due to a higher rate of infection and ischemia with choledochojejunostomy. Nonanastomotic leaks are a result of bile duct ischemia and necrosis as a sequela of hepatic artery thrombosis. This type of leak is often associated with infection and graft dysfunction and has a poor prognosis, with most patients requiring retransplantation. Hepatic artery thrombosis has to be excluded in all patients with nonanastomotic leak, because 90% of these leaks are a result of ischemia.10,96,98,121,145,151,160 Imaging techniques.  US demonstrates perihepatic fluid collections with or without biliary dilation. CT and MRI demonstrate the extent of fluid collections but cannot differentiate bile from nonhemorrhagic fluid collections (Fig. 45-24A). Because it is not possible to differentiate a bilious collection from other fluid collections with US, CT, or MRI, hepatobiliary scintigraphy should be obtained in all patients with suspected bile leak (see Fig. 45-24B). Alternatively, imaging-guided aspiration of the fluid collection may be performed to confirm a bile leak. Fluid collections at the anastomotic site should always be considered a possible bile leak, necessitating follow-up imaging. Conventional T2-weighted MRCP demonstrates the relationship of fluid collections to intrahepatic and extrahepatic bile ducts but cannot show direct evidence of a bile leak.10,51,59 Three MRI contrast agents that are taken up by functioning hepatocytes and excreted into the bile should be able to provide functional evidence of a bile leak. One such contrast medium is mangafodipir trisodium, not approved by the U.S. Food and Drug Administration (FDA) for regular use. Gadobenate dimeglumine (Gd-BOPTA)/Multihance and gadoxetic acid (Gd-EOB-DTPA)/Eovist are FDA approved for regular use. These are T1-shortening contrast agents, with hyperintense signal intensity of the bile on T1 imaging. The major limitation of Gd-BOPTA (Multihance) is that only 3% to 5% of the contrast agent is excreted into the biliary system and needs delayed imaging at 60 to 90 minutes after contrast administration. In contrast, 50% of gadoxetic acid (Eovist) contrast agent is excreted by the biliary system, and bile ducts are opacified by 20 minutes delay after contrast administration.3,15,63 Invasive imaging with endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTC) are undertaken if the patient is being considered for biliary stenting. Management.  The management of bile leaks depends on their location and origin. Small leaks at the T-tube insertion resolve spontaneously and are followed with imaging. Large bilomas and bilious ascites need imaging-guided percutaneous drainage because they are a potential source of infection in these immunocompromised patients. Patients with persistent bile leak and anastomotic stricture are initially managed with percutaneous and endoscopic stenting. Surgical correction should be considered only when nonoperative measures fail. Patients with nonanastomotic leaks from hepatic artery thrombosis eventually need retransplantation.52,57 Surgical management of nonanastomotic strictures is difficult because of their multiplicity and intrahepatic nature. Percutaneous nonoperative treatment with balloon dilation and stenting has a major

1392

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 45-21  IVC stenosis. A, Coronal Gd-enhanced MRI of the liver demonstrates stenosis (arrow) at the suprahepatic IVC anastomosis. B, Stenosis (arrow) is confirmed by an inferior venacavogram. C, Inferior venacavogram after balloon dilation of the stenosis (arrow).

role in the management of these strictures. The success rate of these nonoperative methods is variable, with better success in nonischemic strictures. In most cases, patients with nonanastomotic strictures caused by ischemia eventually need retransplantation, and the percutaneous methods are palliative until a suitable donor liver becomes available.118,128,139,151

Biliary Obstruction Biliary stricture.  Biliary obstruction due to strictures develops several months to years after transplantation. The actual incidence of biliary stricture varies among different centers, but the reported incidence ranges from 12% to 15%.26,32,121,135 Biliary strictures in a transplanted liver are classified as anastomotic or nonanastomotic, and distinguishing between the two is relevant because they differ in terms of morphology, pathogenesis, and prognosis. The reported incidence of anastomotic strictures in the literature is around 4% to 9%. The onset of these strictures depends upon the etiology. Early-onset strictures are related to improper surgical

technique, mismatch of the recipient and donor bile duct size, bile leak, and infection.128,151 Later-onset strictures are more related to fibrosis secondary to ischemia. The incidence is higher with right lobe livingdonor transplants than with whole-liver grafts, owing to blood supply at the anastomosis and multiple small-caliber donor ducts. Anastomotic strictures are seen with both choledochocho­ ledochostomy and choledochojejunostomy biliary reconstruction.26,40,42,46,52,96,98,104,121,128 MRCP with three-dimensional (3D) MPR is excellent for the diagnosis of anastomotic strictures. These strictures are seen as abrupt short-segment narrowing of the duct, with proximal biliary dilatation51 (Figs. 45-25 and 45-26). Anastomotic strictures may have clinical and biochemical evidence of cholestasis or can be incidental findings on routine follow-up imaging. Owing to the chronic and indolent nature of these strictures, intraductal calculi, sludge, and cholangitis may be associated. Because anastomotic strictures are generally not associated with other hepatic or vascular complications, they are managed successfully with percutaneous balloon dilation (see  Fig. 45-25C and D).26,32,135

CHAPTER 45  Liver Transplantation

1393

S

A

B

FIG 45-22  IVC thrombosis. A, Coronal MR venogram demonstrates thrombosis of the entire retrohepatic IVC of the transplanted liver (arrows). B, Sagittal MR venogram demonstrates IVC thrombosis (arrows) and dilated ascending lumbar and azygos collateral veins (arrowheads).

FIG 45-23  Periportal hypodensity. Contrast-enhanced CT of a transplanted liver 1 week after the surgery demonstrates diffuse periportal hypodensity (arrowheads) due to dilated lymphatics. This will eventually resolve with the development of new lymphatic pathways.

Nonanastomotic bile duct strictures are multiple and intrahepatic. They are often caused by bile duct ischemia secondary to hepatic artery thrombosis or stenosis. Other causative factors include prolonged cold ischemia, chronic rejection, and cholangitis from opportunistic infections such as cytomegalovirus.18,26,10,151,116,121 Nonanastomotic strictures may also be caused by recurrence of the primary disease in patients

who had liver transplantation for primary sclerosing cholangitis14,41 (Fig. 45-27). These strictures commonly involve the central bile ducts at the hilum, with variable involvement of the right and left intrahepatic ducts. Owing to ischemic necrosis, nonanastomotic strictures are associated with other complications such as leak and biloma with secondary infection and abscess formation. Nonanastomotic strictures present earlier than anastomotic strictures, and clinical presentation is variable and includes jaundice, cholangitis with fever and chills, and increased liver function tests. Imaging with US and CT may demonstrate biliary dilatation, intrahepatic biloma, abscess, and associated vascular thrombosis (Fig. 45-28A). However, MRCP with contrast-enhanced MRI is the best noninvasive imaging method to demonstrate short-segment anastomotic stricture and multiple intrahepatic bile duct nonanastomotic strictures, biliary dilatation, cholangitis (see Fig. 45-25), intrahepatic abscess, and vascular thrombosis (see Fig. 45-28B and C).18,51,62,121,128 ERCP and PTC are invasive and reserved for therapeutic intervention after identification of strictures by noninvasive imaging with MRCP. Endoscopic or percutaneous balloon dilation with or without stent placement is the first line of management of anastomotic strictures. Surgical revision and biliary reconstruction with hepaticojejunostomy is indicated when percutaneous or endoscopic treatment fails.10,26,121,128,155 Other causes of biliary obstruction.  Biliary cast syndrome is an uncommon complication caused by the diffuse formation of biliary casts, causing biliary obstruction. The incidence is not known, and the timing is variable after transplantation surgery. The exact cause is not known, but it is believed to be multifactorial, including acute cellular rejection, prolonged cold ischemic time, stricture, biliary drainage tubes, stasis, sludge formation, and infection. Ischemia and stricture formation are thought to be common predisposing factors for this syndrome. Noninvasive imaging demonstrates nonspecific biliary

1394

PART II  CT and MR Imaging of the Whole Body

A D

B

E

C

F FIG 45-24  Bile leak. A, Contrast-enhanced CT of a patient with a recent liver transplant demonstrates a perihepatic fluid collection (arrows). B, Hepatobiliary scan 3 hours after the tracer injection shows increased perihepatic activity (arrows) within the collection, confirming a bile leak. Axial (C) and coronal (D) T2-weighted MRI 48 hours after transplant demonstrates a large perihepatic fluid collection (arrow). Coronal (E) and axial (F) T1-weighted image 25 minutes after IV administration of hepatobiliary contrast (Eovist) demonstrates extravasation of biliary contrast (arrow) into the perihepatic collection, confirming bile leak.

CHAPTER 45  Liver Transplantation

A

B

C

D FIG 45-25  Anastomotic stricture of the bile duct. MRCP (A) and ERCP (B) images demonstrate a stricture (arrow) at the choledochocholedochostomy. Balloon dilation (C) and postdilation ERCP image (D) of the stricture (arrow).

1395

1396

PART II  CT and MR Imaging of the Whole Body

A A

B FIG 45-27  Nonanastomotic biliary strictures from recurrent primary

B

sclerosing cholangitis. A, Coronal MRCP of a transplanted liver shows multiple short segmental strictures of the intrahepatic bile ducts (arrows). B, Gd-enhanced MRI of the liver demonstrates dilated ducts with diffuse mural thickening of the bile ducts (arrows).

FIG 45-26  Anastomotic stricture of the bile duct. MRCP (A) and percutaneous transhepatic cholangiogram (B) of a transplanted liver demonstrate stenosis (arrow) at the biliary-enteric anastomosis.

dilatation, and a definitive diagnosis is made with ERCP and PTC. Initial management is by percutaneous or endoscopic cast clearance. Surgery is reserved for those refractory to nonoperative measures.103,120 Less common causes of biliary obstruction in transplanted livers are biliary calculi, sphincter of Oddi dysfunction, and recurrent cholangiocarcinoma. Biliary calculi are identified as hypointense dependent filling defects on MRCP, differentiating them from hypointense nondependent defects caused by air bubbles from pneumobilia and biliary enteric anastomosis (Fig. 45-29). In patients with sphincter of Oddi dysfunction, there is generalized ductal dilatation beyond the anastomosis up to the ampulla (Fig. 45-30). This papillary dysfunction is presumably due to denervation of the papilla during hepatectomy of the recipient liver.137 Recurrence of cholangiocarcinoma involves the ducts at the anastomosis or the intrahepatic ducts or involves the lymph nodes at the porta hepatis. Mucocele of the cystic duct remnant of an allograft can cause biliary obstruction by extrinsic compression of the CHD. US reveals intrahepatic biliary dilatation and a fluid-filled cystic structure at the level of the CHD. A definitive diagnosis can be made with MRCP or a cholangiogram.1,10,146

Fluid Collections Posttransplant perihepatic fluid collections are common and are due to hematoma, seroma, biloma, and abscess formation. Postoperative fluid collections and hematomas are common at the anastomotic areas in the hilum, adjacent to the IVC, and in the lesser sac and perihepatic and subphrenic spaces. Perihepatic fluid collections can be easily identified and quantified on postoperative surveillance US. Perihepatic hematomas have a variable sonographic appearance depending on their age. Acute and subacute hematomas have variable echogenicity on US and hyperattenuation on CT and can be readily distinguished from ascites (Fig. 45-31A and B).2,10,29,79 Large perihepatic hematomas in the immediate postoperative period should initiate a search for active hemorrhage and possibly timely surgical reexploration. Bilomas cannot be distinguished from other nonbilious fluid collections with any of the noninvasive imaging methods (see Fig. 45-24A). Perihepatic fluid collections that increase in size on follow-up imaging and may be caused by a bile leak should be aspirated under imaging guidance to confirm a biloma; the fluid should also be cultured because bilomas pose a high risk of infection and subsequent abscess formation. Once a biloma is confirmed, hepatobiliary scintigraphy is necessary to confirm an active bile leak (see Fig. 45-24B). Infected fluid collections

CHAPTER 45  Liver Transplantation

1397

A

C

B

FIG 45-28  Biliary strictures. A, US of a transplanted liver demonstrates intrahepatic bile duct dilatation (arrow). B, MRCP shows an anastomotic stricture (arrowhead) and nonanastomotic strictures (arrows). C, Intraoperative cholangiogram after contrast injection proximal to the anastomosis shows nonanastomotic strictures (arrows).

cannot be distinguished from other fluid collections by imaging, and all patients with unexplained fever should have this fluid aspirated and analyzed. Large perihepatic fluid collections can affect the vascular and biliary anastomosis and eventually survival of the graft. Percutaneous drainage of perihepatic fluid collections can be successfully performed with either US or CT guidance.2,17

Adrenal Hemorrhage Right adrenal hemorrhage is a known complication of liver transplantation; it is due to ligation or division of the adrenal vein at the time of hepatectomy of the recipient liver, resulting in adrenal hemorrhage or hemorrhagic infarction. It appears as a hyperechoic mass on US and a hyperdense mass on CT in the right adrenal region adjacent to the IVC (see Fig. 45-31C). MRI demonstrates a hyperintense mass on fatsaturated T1-weighted images.13,108,127 The hemorrhage resolves spontaneously without the development of adrenal insufficiency or other complications. The incidence of adrenal hemorrhage is reported to be lower following piggyback-type IVC reconstruction, because the recipient IVC is retained, making injury to the adrenal vein less likely.

Neoplasms of the Transplanted Liver Recurrence of the primary hepatic malignancy in the resected liver can develop within the transplanted liver. The reported recurrence of HCC in transplanted livers varies widely depending on the pretransplant selection criteria and the underlying primary hepatic disease.56 The original stage of HCC, histologic grade, and vascular invasion are all predictors of tumor recurrence and survival of the recipient. Assessment of the tumor burden by preoperative imaging, with attention to tumor size and multiplicity, is vital for proper patient selection.123,148 Adjuvant chemotherapy and preoperative chemoembolization can reduce the tumor burden while waiting for a donor liver.9,10,141 (Fig. 45-32) The most common site of recurrence is lung, followed by liver (Fig. 45-33) and regional and distant lymph nodes. Recurrence in the lung is due to the embolization of malignant cells before or during surgery and is augmented by immunosuppression.33 Liver transplantation after neoadjuvant chemoradiation is an accepted treatment option for selected patients with perihilar cholangiocarcinoma. Proper patient selection is important because there is a

1398

PART II  CT and MR Imaging of the Whole Body

FIG 45-30  Papillary dysfunction. MRCP of a transplanted liver demonstrates smooth narrowing of the common bile duct at the ampulla (arrow), with biliary dilatation proximal and distal to the anastomosis (arrowhead).

FIG 45-29  Choledocholithiasis. MRCP of a transplanted liver demonstrates a common bile duct calculus (arrow) distal to the anastomosis (arrowhead).

B A

C FIG 45-31  Perihepatic fluid collections. A, Immediate postoperative US of a transplanted liver shows an echogenic perihepatic collection (arrows). B, Corresponding CT shows a hyperdense perihepatic hematoma (arrows). C, Contrast-enhanced CT in another patient demonstrates a perihepatic hematoma (arrows) as well as right adrenal hemorrhage (arrowhead).

CHAPTER 45  Liver Transplantation

A

B

C

D

1399

FIG 45-32  Transarterial chemoembolization (TACE) of HCC in pretransplant liver to reduce tumor burden. Pretreatment post-Gd T1-weighted MRI in arterial (A) and portal venous (B) phases demonstrate enhancing mass (arrow) with washout. Post-TACE T1-weighted MRI in arterial (C) and portal venous (D) phases demonstrate no enhancement of the mass (arrow).

higher rate of recurrence in the transplanted liver in elderly patients with large primary tumors and high CA19-9 levels.47,48 Liver transplant recipients are also at a higher risk than the general population for malignancy unrelated to the original hepatic malignancy. Prolonged immunosuppression is the leading cause of this increased incidence of de novo malignancy, although other factors such as chronic viral infections, the patient’s age, alcohol use, and smoking play an important role.19,136 Skin cancer and lymphoproliferative disease are the most common malignancies in liver transplant recipients. Squamous cell carcinoma and Kaposi’s sarcoma are the most common types of skin cancer, though an increased incidence of basal cell and Merkel cell carcinomas has been reported.36 Posttransplant lymphoproliferative disease has an established association with the degree of immunosuppression and Epstein-Barr virus infection. B-cell lymphoma is the most common form; it has a predilection for extranodal involvement and develops within 1 year of transplantation

(Fig. 45-34). Extranodal sites include the liver, spleen, gastrointestinal tract, kidneys, and adrenals.82 Hepatic involvement can be intraparenchymal or hilar in location. The hilar type is more common, with hypodense masses encasing the hilar vessels and ducts. The hepatic parenchymal type presents as hypoechoic mass lesions on US and hypodense mass lesions on contrast-enhanced CT.37,97,131 Malignancies of the upper airways, lungs, and gastrointestinal tract are also reported to occur in liver transplant recipients.55 Patients with primary sclerosing cholangitis as the primary disease are at risk of developing colon cancer owing to its association with inflammatory bowel disease.11

Nonneoplastic Hepatic Parenchymal Complications Hepatic parenchymal ischemia and infarction are uncommon in a normal liver owing to hepatic arterial and portal vascularity and an abundant collateral supply. In a transplanted liver, infarction can occur following hepatic artery or portal vein thrombosis but is more frequent

1400

A

PART II  CT and MR Imaging of the Whole Body

C

B

FIG 45-33  Recurrent HCC in transplant liver. A, Axial T2-weighted MRI demonstrates a well-defined T2-hyperintense mass lesion in segment 8. Axial post-Gd T1-weighted MRI in arterial dominant phase (B) and portal venous phase (C) demonstrates enhancement in arterial dominant phase and washout in portal venous phase.

A

B

C FIG 45-34  Posttransplant lymphoma. Axial (A) and coronal (B and C) Gd-enhanced MRI of the abdomen and pelvis demonstrates large, enhancing, mesenteric mass lesions (arrows). CT-guided biopsy and histopathology of these lesions confirmed B-cell lymphoma.

CHAPTER 45  Liver Transplantation

1401

Epstein-Barr virus, Pneumocystis jiroveci, Aspergillus spp., and Nocardia spp. Prevention, early detection, and timely management of these infections are vital to decrease morbidity and mortality. Imaging is complementary to clinical and laboratory tests for early detection of these infections.6,124

CONCLUSION

1 2

1

2

A

Liver transplantation is an established technique for the management of patients with chronic liver disease. The addition of new indications for liver transplantation has further increased the number of patients waiting for liver transplants. As a result, new surgical techniques such as LDLT have been devised. This development has increased the responsibility of the radiologist from detecting postoperative complications to performing preoperative evaluations for the selection of both recipients and donors. Advances in imaging techniques and the feasibility of multimodality imaging of the liver have placed the radiologist at the forefront of liver transplantation. The radiologist must be familiar with new surgical techniques to enable a better understanding of the surgical anatomy and postoperative complications. The radiologist is now a key member of the transplant team.

REFERENCES

B FIG 45-35  Liver abscess. This patient with a liver transplant presented with fever and elevated white cell count. A, US demonstrates a large cystic mass in the left lobe of the liver. B, Contrast-enhanced CT demonstrates a cystic mass with an enhancing wall (arrow). The diagnosis was confirmed by CT-guided drainage of the abscess.

with the former. It is caused by disruption of the collateral vascular supply to the liver during transplantation surgery. Hepatic infarcts are seen as peripheral wedge-shaped low-attenuation areas on contrastenhanced CT or hypointense areas on MRI. The infarcted parenchyma undergoes liquefactive necrosis and is prone to infection and abscess formation10,29,37,39 (Fig. 45-35). Rejection of a liver transplant is diagnosed by biopsy and histopathology. The role of imaging in rejection is to exclude other vascular and nonvascular hepatic complications.97

Extrahepatic Opportunistic Infections Infection continues to be a major cause of morbidity and mortality in liver transplant patients. Apart from infections related to post­ surgical complications, these patients are prone to extrahepatic  infections owing to their immunosuppressed state. Extrahepatic sites of infection include the skin, upper airways, lungs, gastrointestinal tract, urinary tract, and central nervous system. These infections  can be caused by community-acquired or hospital-based infectious agents. Common opportunistic infectious agents are cytomegalovirus,

1. Abcarian PW, Emond JC, Ring EJ: Cystic duct remnant mucocele in a liver transplant recipient. J Vasc Interv Radiol 5:127–130, 1994. 2. Akin K, Ozturk A, Guvenc Z, et al: Localized fluid collections after liver transplantation. Transplant Proc 38:627–630, 2006. 3. Assaban M, Aube C, Lebigot J, et al: Mangafodipir trisodium-enhanced magnetic resonance cholangiography for detection of bile leaks.   J Radiol 87:41–47, 2006. 4. Avila LF, Luis AL, Hernandez F, et al: Liver transplantation for malignant tumours in children. Eur J Pediatr Surg 16:411–414, 2006. 5. Banga NR, Kessel DO, Patel JV, et al: Endovascular management of arterial conduit pseudoaneurysm after liver transplantation: A report of two cases. Transplantation 79:1763–1765, 2005. 6. Barkholt L, Ericzon BG, Tollemar J, et al: Infections in human liver recipients: Different patterns early and late after transplantation. Transpl Int 6:77–84, 1993. 7. Befeler AS, Hayashi PH, Di Bisceglie AM: Liver transplantation for hepatocellular carcinoma. Gastroenterology 128:1752–1764, 2005. 8. Ben-Haim M, Emre S, Fishbein TM, et al: Critical graft size in adult-to-adult living donor liver transplantation: Impact of the recipient’s disease. Liver Transpl 7:948–953, 2001. 9. Bharat A, Brown DB, Crippin JS, et al: Pre-liver transplantation locoregional adjuvant therapy for hepatocellular carcinoma as a strategy to improve longterm survival. J Am Coll Surg 203:411–420, 2006. 10. Bhargava P, Vaidya S, Dick AA, et al: Imaging of orthotopic liver transplantation: Review. AJR Am J Roentgenol 196(3 Suppl):WS15– WS25, 2011. 11. Bleday R, Lee E, Jessurun J, et al: Increased risk of early colorectal neoplasms after hepatic transplant in patients with inflammatory bowel disease. Dis Colon Rectum 36:908–912, 1993. 12. Borsa JJ, Daly CP, Fontaine AB, et al: Treatment of inferior vena cava anastomotic stenoses with the Wallstent endoprosthesis after orthotopic liver transplantation. J Vasc Interv Radiol 10:17–22, 1999. 13. Bowen AD, Keslar PJ, Newman B, et al: Adrenal hemorrhage after liver transplantation. Radiology 176:85–88, 1990. 14. Brancatelli G, Katyal S, Federle MP, et al: Three-dimensional multislice helical computed tomography with the volume rendering technique in the detection of vascular complications after liver transplantation. Transplantation 73(2):237–242, 2002. 15. Brandsaeter B, Schrumpf E, Bentdal O, et al: Recurrent primary sclerosing cholangitis after liver transplantation: A magnetic resonance

1402

16.

17.

18.

19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

29. 30.

31.

32.

33.

34. 35.

36.

PART II  CT and MR Imaging of the Whole Body

cholangiography study with analyses of predictive factors. Liver Transpl 11:1361–1369, 2005. Bridges MD, May GR, Harnois DM: Diagnosing biliary complications of orthotopic liver transplantation with mangafodipir trisodiumenhanced MR cholangiography: Comparison with conventional MR cholangiography. AJR Am J Roentgenol 182:1497–1504, 2004. Brown RS, Jr, Kumar KS, Russo MW, et al: Model for end-stage liver disease and Child-Turcotte-Pugh score as predictors of pretransplantation disease severity, posttransplantation outcome, and resource utilization in United Network for Organ Sharing status 2A patients. Liver Transpl 8:278–284, 2002. Campatelli A, Di Candio G, Morelli L, et al: Interventional ultrasound: Experience in 426 orthotopic liver transplantations. Transplant Proc 36:550–551, 2004. Campbell WL, Sheng R, Zajko AB, et al: Intrahepatic biliary strictures after liver transplantation. Radiology 191:735–740, 1994. Romero-Vargas ME, Flores-Cortes M, Valera Z, et al: Cancers of new appearance in liver transplant recipients: Incidence and evolution. Transplant Proc 38:2508–2510, 2006. Carr JC, Nemcek AA, Jr, Abecassis M, et al: Preoperative evaluation of the entire hepatic vasculature in living liver donors with use of contrast-enhanced MR angiography and true fast imaging with steady-state precession. J Vasc Interv Radiol 14:441–449, 2003. Chaib E, Ribeiro MA, Jr, Saad WA, et al: The main hepatic anatomic variations for the purpose of split-liver transplantation. Transplant Proc 37:1063–1066, 2005. Chen WP, Chen JH, Hwang JI, et al: Spectrum of transient hepatic attenuation differences in biphasic helical CT. AJR Am J Roentgenol 172:419–424, 1999. Cheng YF, Huang TL, Lee TY, et al: Variation of the intrahepatic portal vein; angiographic demonstration and application in living-related hepatic transplantation. Transplant Proc 28:1667–1668, 1996. Choi JW, Kim TK, Kim KW, et al: Anatomic variation in intrahepatic bile ducts: An analysis of intraoperative cholangiograms in 300 consecutive donors for living donor liver transplantation. Korean J Radiol 4:85–90, 2003. Chok KS1, Lo CM: Prevention and management of biliary anastomotic stricture in right-lobe living-donor liver transplantation. J Gastroenterol Hepatol 29(10):1756–1763, 2014. Chong WK, Beland JC, Weeks SM: Sonographic evaluation of venous obstruction in liver transplants. AJR Am J Roentgenol 188:W515–W521, 2007. Colonna JO, 2nd, Shaked A, Gomes AS, et al: Biliary strictures complicating liver transplantation: Incidence, pathogenesis, management, and outcome. Ann Surg 216:344–350, 1992. Crossin JD, Muradali D, Wilson SR: US of liver transplants: Normal and abnormal. Radiographics 23:1093–1114, 2003. Dodd GD, 3rd, Memel DS, Zajko AB, et al: Hepatic artery stenosis   and thrombosis in transplant recipients: Doppler diagnosis with resistive index and systolic acceleration time. Radiology 192:657–661, 1994. Dupuy D, Costello P, Lewis D, et al: Abdominal CT findings after liver transplantation in 66 patients. AJR Am J Roentgenol 156:1167–1170, 1991. Elias G, Rastellini C, Nsier H, et al: Successful long-term repair of hepatic artery pseudoaneurysm following liver transplantation with primary stent-grafting. Liver Transpl 13:1346–1348, 2007. Escartin A, Sapisochin G, Bilbao I, et al: Recurrence of hepatocellular carcinoma after liver transplantation. Transplant Proc 39:2308–2310, 2007. Fan ST, Lo CM, Liu CL, et al: Safety of donors in live donor liver transplantation using right lobe grafts. Arch Surg 135:336–340, 2000. Feller RB, Waugh RC, Selby WS, et al: Biliary strictures after liver transplantation: Clinical picture, correlates and outcomes.   J Gastroenterol Hepatol 11:21–25, 1996. Ferris JV, Baron RL, Marsh JW, Jr, et al: Recurrent hepatocellular carcinoma after liver transplantation: Spectrum of CT findings and recurrence patterns. Radiology 198:233–238, 1996.

37. Forman LM, Lucey MR: Predicting the prognosis of chronic liver disease: An evolution from child to MELD. Mayo end-stage liver disease. Hepatology 33:473–475, 2001. 38. Freeman RB, Jr: Transplantation for hepatocellular carcinoma: The Milan criteria and beyond. Liver Transpl 12(Suppl 2):S8–S13, 2006. 39. Fung JJ, Jain A, Kwak EJ, et al: De novo malignancies after liver transplantation: A major cause of late death. Liver Transpl 7(Suppl 1):S109–S118, 2001. 40. García-Criado A, Gilabert R, Bargalló X, et al: Radiology in liver transplantation. Semin Ultrasound CT MR 23:114–129, 2002. 41. Gazelle GS, Lee MJ, Mueller PR: Cholangiographic segmental anatomy of the liver. Radiographics 14:1005–1013, 1994.39. 42. Gondolesi GE, Varotti G, Florman SS, et al: Biliary complications in 96 consecutive right lobe living donor transplant recipients. Transplantation 77:1842–1848, 2004. 43. Graziadei IW: Recurrence of primary sclerosing cholangitis after liver transplantation. Liver Transpl 8:575–581, 2002. 44. Graziadei IW, Sandmueller H, Waldenberger P, et al: Chemoembolization followed by liver transplantation for hepatocellular carcinoma impedes tumor progression while on the waiting list and leads to excellent outcome. Liver Transpl 9:557–563, 2003. 45. Graziadei IW, Schwaighofer H, Koch R, et al: Long-term outcome of endoscopic treatment of biliary strictures after liver transplantation. Liver Transpl 12:718–725, 2006. 46. Greif F, Bronsther OL, Van Thiel DH, et al: The incidence, timing, and management of biliary tract complications after orthotopic liver transplantation. Ann Surg 219:40–45, 1994. 47. Grewal HP, Shokouh-Amiri MH, Vera S, et al: Surgical technique for right lobe adult living donor liver transplantation without venovenous bypass or portocaval shunting and with duct-to-duct biliary reconstruction. Ann Surg 233:502–508, 2001. 48. Gruttadauria S, Foglieni CS, Doria C, et al: The hepatic artery in liver transplantation and surgery: Vascular anomalies in 701 cases. Clin Transplant 15:359–363, 2001. 49. Guiney MJ, Kruskal JB, Sosna J, et al: Multi-detector row CT of relevant vascular anatomy of the surgical plane in split-liver transplantation. Radiology 229:401–407, 2003. 50. Hampe T, Dogan A, Encke J, et al: Biliary complications after liver transplantation. Clin Transplant 20(Suppl 17):93–96, 2006. 51. Heimbach JK, Gores GJ, Haddock MG, et al: Liver transplantation for unresectable perihilar cholangiocarcinoma. Semin Liver Dis 24:201–207, 2004. 52. Heimbach JK, Gores GJ, Haddock MG, et al: Predictors of disease recurrence following neoadjuvant chemoradiotherapy and liver transplantation for unresectable perihilar cholangiocarcinoma. Transplantation 82:1703–1707, 2006. 53. Hertl M, Cosimi AB: Liver transplantation for malignancy. Oncologist 10:269–281, 2005. 54. Hirao K, Miyazaki A, Fujimoto T, et al: Evaluation of aberrant   bile ducts before laparoscopic cholecystectomy: Helical CT cholangiography versus MR cholangiography. AJR Am J Roentgenol 175:713–720, 2000. 55. Hoeffel C, Azizi L, Lewin M, et al: Normal and pathologic features of the postoperative biliary tract at 3D MR cholangiopancreatography and MR imaging. Radiographics 26:1603–1620, 2006. 56. Horrow MM, Blumenthal BM, Reich DJ, et al: Sonographic   diagnosis and outcome of hepatic artery thrombosis after orthotopic liver transplantation in adults. AJR Am J Roentgenol 189(2):346–351, 2007. 57. Hwang S, Lee SG, Sung KB, et al: Long-term incidence, risk factors, and management of biliary complications after adult living donor liver transplantation. Liver Transpl 12:831–838, 2006. 58. Ito T, Kiuchi T, Egawa H, et al: Surgery-related morbidity in living donors of right-lobe liver graft: Lessons from the first 200 cases. Transplantation 76:158–163, 2003. 59. Iwasaki M, Takada Y, Hayashi M, et al: Noninvasive evaluation of graft steatosis in living donor liver transplantation. Transplantation 78:1501–1505, 2004.

CHAPTER 45  Liver Transplantation 60. Jimenez C, Marques E, Loinaz C, et al: Upper aerodigestive tract and lung tumors after liver transplantation. Transplant Proc 35:1900–1901, 2003. 61. Johnston TD, Gates R, Reddy KS, et al: Nonoperative management of bile leaks following liver transplantation. Clin Transplant 14:365–369, 2000. 62. Kamel IR, Lawler LP, Fishman EK: Variations in anatomy of the middle hepatic vein and their impact on formal right hepatectomy. Abdom Imaging 28:668–674, 2003. 63. Kantarcı M, Pirimoglu B, Karabulut N, et al: Non-invasive detection of biliary leaks using Gd-EOB-DTPA-enhanced MR cholangiography: Comparison with T2-weighted MR cholangiography. Eur Radiol 23(10):2713–2722, 2013. 64. Kapoor V, Baron RL, Peterson MS: Bile leaks after surgery. AJR Am J Roentgenol 182:451–458, 2004. 65. Kasahara M, Ueda M, Haga H, et al: Living-donor liver transplantation for hepatoblastoma. Am J Transplant 5:2229–2235, 2005. 66. Kayahan Ulu EM, Coskun M, Ozbek O, et al: Accuracy of multidetector computed tomographic angiography for detecting hepatic artery complications after liver transplantation. Transplant Proc 39(10):3239– 3244, 2007. 67. Kim BS, Kim TK, Jung DJ, et al: Vascular complications after living related liver transplantation: Evaluation with gadolinium-enhanced three-dimensional MR angiography. AJR Am J Roentgenol 181:467–474, 2003. 68. Kim HJ, Kim KW, Kim AY, et al: Hepatic artery pseudoaneurysms in adult living-donor liver transplantation: Efficacy of CT and Doppler sonography. AJR Am J Roentgenol 184:1549–1555, 2005. 69. Kim SY, Kim KW, Kim MJ, et al: Multidetector row CT of various hepatic artery complications after living donor liver transplantation. Abdom Imaging 32:635–643, 2007. 70. Kim BW, Won JH, Lee BM, et al: Intraarterial thrombolytic treatment for hepatic artery thrombosis immediately after living donor liver transplantation. Transplant Proc 38:3128–3131, 2006. 71. Kishi Y, Sugawara Y, Kaneko J, et al: Classification of portal vein anatomy for partial liver transplantation. Transplant Proc 36:3075–3076, 2004. 72. Kiuchi T, Kasahara M, Uryuhara K, et al: Impact of graft size mismatching on graft prognosis in liver transplantation from living donors. Transplantation 67:321–327, 1999. 73. Kobori L, van der Kolk MJ, de Jong KP, et al: Splenic artery aneurysms in liver transplant patients. Liver Transplant Group. J Hepatol 27:890–893, 1997. 74. Kodama Y, Sakuhara Y, Abo D, et al: Percutaneous transluminal angioplasty for hepatic artery stenosis after living donor liver transplantation. Liver Transpl 12:465–469, 2006. 75. Kooby DA, Fong Y, Suriawinata A, et al: Impact of steatosis on perioperative outcome following hepatic resection. J Gastrointest Surg 7:1034–1044, 2003. 76. Kwon CH, Joh JW, Lee KW, et al: Safety of donors with fatty liver in liver transplantation. Transplant Proc 38:2106–2107, 2006. 77. Lang H, Oldhafer KJ, Weimann A, et al: Liver transplantation for metastatic neuroendocrine tumors. Ann Surg 225:347–354, 1997. 78. Lang P, Schnarkowski P, Grampp S, et al: Liver transplantation: Significance of the periportal collar on MRI. J Comput Assist Tomogr 19:580–585, 1995. 79. Lee J, Ben-Ami T, Yousefzadeh D, et al: Extrahepatic portal vein stenosis in recipients of living-donor allografts: Doppler sonography. AJR Am J Roentgenol 167:85–90, 1996. 80. Lee SS, Kim TK, Byun JH, et al: Hepatic arteries in potential donors for living related liver transplantation: Evaluation with multi-detector row CT angiography. Radiology 227:391–399, 2003. 81. Lee VS, Krinsky GA, Nazzaro CA, et al: Defining intrahepatic biliary anatomy in living liver transplant donor candidates at mangafodipir trisodium-enhanced MR cholangiography versus conventional T2-weighted MR cholangiography. Radiology 233:659–666, 2004. 82. Lee SW, Park SH, Kim KW, et al: Unenhanced CT for assessment of macrovesicular hepatic steatosis in living liver donors: Comparison of

83.

84.

85.

86.

87.

88. 89. 90. 91.

92.

93.

94.

95.

96.

97. 98. 99. 100.

101. 102.

103. 104. 105. 106.

107.

1403

visual grading with liver attenuation index. Radiology 244:479–485, 2007. Lerut JP, Orlando G, Sempoux C, et al: Hepatic haemangioendothelioma in adults: Excellent outcome following liver transplantation. Transpl Int 17:202–207, 2004. Letourneau JG, Day DL, Ascher NL, et al: Abdominal sonography after hepatic transplantation: Results in 36 patients. AJR Am J Roentgenol 149:299–303, 1987. Liu B, Yan LN, Wang WT, et al: Clinical study on safety of adult-toadult living donor liver transplantation in both donors and recipients. World J Gastroenterol 13:955–959, 2007. Lo CM, Fan ST, Liu CL, et al: Adult-to-adult living donor liver transplantation using extended right lobe grafts. Ann Surg 226:261–269, 1997. Lowell JA, Coopersmith CM, Shenoy S, et al: Unusual presentations of nonmycotic hepatic artery pseudoaneurysms after liver transplantation. Liver Transpl Surg 5:200–203, 1999. Lubienski A: Hepatocellular carcinoma: Interventional bridging to liver transplantation. Transplantation 80(1 Suppl):S113–S119, 2005. Makowka L, Stieber AC, Sher L, et al: Surgical technique of orthotopic liver transplantation. Gastroenterol Clin North Am 17:33–51, 1988. Malhi H, Gores GJ: Review article: The modern diagnosis and therapy of cholangiocarcinoma. Aliment Pharmacol Ther 23:1287–1296, 2006. Marcos A, Ham JM, Fisher RA, et al: Surgical management of anatomical variations of the right lobe in living donor liver transplantation. Ann Surg 231:824–831, 2000. Marshall MM, Muiesan P, Srinivasan P, et al: Hepatic artery pseudoaneurysms following liver transplantation: Incidence, presenting features and management. Clin Radiol 56:579–587, 2001. Marsman WA, Wiesner RH, Rodriguez L, et al: Use of fatty donor liver is associated with diminished early patient and graft survival. Transplantation 62:1246–1251, 1996. Martin P, DiMartini A, Feng S, et al: Evaluation for liver transplantation in adults: 2013 practice guideline by the American Association for the Study of Liver Diseases and the American Society of Transplantation. Hepatology 59(3):1144–1165, 2014. Matsubara K, Cho A, Okazumi S, et al: Anatomy of the middle hepatic vein: Applications to living donor liver transplantation. Hepatogastroenterology 53:933–937, 2006. Mazzaferro V, Regalia E, Doci R, et al: Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 334:693–699, 1996. Michels NA: Newer anatomy of the liver and its variant blood supply and collateral circulation. Am J Surg 112:337–347, 1966. Miro JM, Aguero F, Laguno M, et al: Liver transplantation in HIV/ hepatitis co-infection. J HIV Ther 12:24–35, 2007. Molmenti EP, Pinto PA, Klein J, et al: Normal and variant arterial supply of the liver and gallbladder. Pediatr Transplant 7:80–82, 2003. Moon JI, Kwon CH, Joh JW, et al: Safety of small-for-size grafts in adult-to-adult living donor liver transplantation using the right lobe. Liver Transpl 16(7):864–869, 2010. Murray KF, Carithers RL, Jr: AASLD practice guidelines: Evaluation of the patient for liver transplantation. Hepatology 41:1407–1432, 2005. Navarro F, Le Moine MC, Fabre JM, et al: Specific vascular complications of orthotopic liver transplantation with preservation of the retrohepatic vena cava: Review of 1361 cases. Transplantation 68:646–650, 1999. Nemec P, Ondrasek J, Studenik P, et al: Biliary complications in liver transplantation. Ann Transplant 6:24–28, 2001. Nghiem HV: Imaging of hepatic transplantation. Radiol Clin North Am 36:429–443, 1998. O’Connor TP, Lewis WD, Jenkins RL: Biliary tract complications after liver transplantation. Arch Surg 130:312–317, 1995. Ohta M, Hashizume M, Tanoue K, et al: Splenic hyperkinetic state and splenic artery aneurysm in portal hypertension. Hepatogastroenterology 39:529–532, 1992. Pandey D, Lee KH, Tan KC: The role of liver transplantation for hilar cholangiocarcinoma. Hepatobiliary Pancreat Dis Int 6:248–253, 2007.

1404

PART II  CT and MR Imaging of the Whole Body

108. Pandharipande PV, Lee VS, Morgan GR, et al: Vascular and extravascular complications of liver transplantation: Comprehensive evaluation with three-dimensional contrast-enhanced volumetric MR imaging and MR cholangiopancreatography. AJR Am J Roentgenol 177:1101–1107, 2001. 109. Parrilla P, Sánchez-Bueno F, Figueras J, et al: Analysis of the complications of the piggy-back technique in 1112 liver transplants. Transplantation 67:1214–1217, 1999. 110. Parry SD, Muiesan P: Cholangiopathy and the biliary cast syndrome. Eur J Gastroenterol Hepatol 15:341–343, 2003. 111. Pascher A, Neuhaus P: Bile duct complications after liver transplantation. Transpl Int 18:627–642, 2005. 112. Pieters PC, Miller WJ, DeMeo JH: Evaluation of the portal venous system: Complementary roles of invasive and noninvasive imaging strategies. Radiographics 17:879–895, 1997. 113. Platt JF, Rubin JM, Ellis JH: Hepatic artery resistance changes in portal vein thrombosis. Radiology 196:95–98, 1995. 114. Platt JF, Yutzy GG, Bude RO, et al: Use of Doppler sonography for revealing hepatic artery stenosis in liver transplant recipients. AJR Am J Roentgenol 168:473–476, 1997. 115. Prokesch RW, Schima W, Berlakovich G, et al: Adrenal hemorrhage after orthotopic liver transplantation: MR appearance. Eur Radiol 11:2484–2487, 2001. 116. Puente SG, Bannura GC: Radiological anatomy of the biliary tract: Variations and congenital abnormalities. World J Surg 7:271–276, 1983. 117. Rea DJ, Heimbach JK, Rosen CB, et al: Liver transplantation with neoadjuvant chemoradiation is more effective than resection for hilar cholangiocarcinoma. Ann Surg 242:451–458, 2005. 118. Reddy KS, Johnston TD, Putnam LA, et al: Piggyback technique and selective use of veno-venous bypass in adult orthotopic liver transplantation. Clin Transplant 14:370–374, 2000. 119. Rossi AR, Pozniak MA, Zarvan NP: Upper inferior vena caval anastomotic stenosis in liver transplant recipients: Doppler US diagnosis. Radiology 187:387–389, 1993. 120. Rustgi VK: The epidemiology of hepatitis C infection in the United States. J Gastroenterol 42:513–521, 2007. 121. Ryu CH, Lee SK: Biliary strictures after liver transplantation. Gut Liver 5(2):133–142, 2011. 122. Saad WE, Davies MG, Sahler L, et al: Hepatic artery stenosis in liver transplant recipients: Primary treatment with percutaneous transluminal angioplasty. J Vasc Interv Radiol 16:795–805, 2005. 123. Saggi BH, Farmer DG, Yersiz H, et al: Surgical advances in liver and bowel transplantation. Anesthesiol Clin North America 22:713–740, 2004. 124. Sanchez-Urdazpal L, Gores GJ, Ward EM, et al: Diagnostic features and clinical outcome of ischemic-type biliary complications after liver transplantation. Hepatology 17:605–609, 1993. 125. Sarmiento JM, Que FG: Hepatic surgery for metastases from neuroendocrine tumors. Surg Oncol Clin N Am 12:231–242, 2003. 126. Scarborough JE, Desai DM: Treatment options for biliary complications after orthotopic liver transplantation. Curr Treat Options Gastroenterol 10:81–89, 2007. 127. Schroeder T, Malagó M, Debatin JF, et al: “All-in-one” imaging protocols for the evaluation of potential living liver donors: Comparison of magnetic resonance imaging and multidetector computed tomography. Liver Transpl 11:776–787, 2005. 128. Seehofer D, Eurich D, Veltzke-Schlieker W, et al: Biliary complications after liver transplantation: Old problems and new challenges. Am J Transplant 13(2):253–265, 2013. 129. Shah JN, Haigh WG, Lee SP, et al: Biliary casts after orthotopic liver transplantation: Clinical factors, treatment, biochemical analysis. Am J Gastroenterol 98:1861–1867, 2003. 130. Sheng R, Sammon JK, Zajko AB, et al: Bile leak after hepatic transplantation: Cholangiographic features, prevalence, and clinical outcome. Radiology 192:413–416, 1994. 131. Sherman M: Hepatocellular carcinoma: Epidemiology, risk factors, and screening. Semin Liver Dis 25:143–154, 2005.

132. Shimoda M, Ghobrial RM, Carmody IC, et al: Predictors of survival after liver transplantation for hepatocellular carcinoma associated with hepatitis C. Liver Transpl 10:1478–1486, 2004. 133. Singh N: Infectious diseases in the liver transplant recipient. Semin Gastrointest Dis 9:136–146, 1998. 134. Soejima Y, Shimada M, Suehiro T, et al: Use of steatotic graft in living-donor liver transplantation. Transplantation 76:344–348, 2003. 135. Soejima Y, Taketomi A, Yoshizumi T, et al: Extended indication for living donor liver transplantation in patients with hepatocellular carcinoma. Transplantation 83:893–899, 2007. 136. Solomon N, Sumkin J: Right adrenal gland hemorrhage as a complication of liver transplantation: CT appearance. J Comput Assist Tomogr 12:95–97, 1988. 137. Someda H, Moriyasu F, Fujimoto M, et al: Vascular complications in living related liver transplantation detected with intraoperative and postoperative Doppler US. J Hepatol 22:623–632, 1995. 138. Starzl TE, Iwatsuki S, Esquivel CO, et al: Refinements in the surgical technique of liver transplantation. Semin Liver Dis 5:349–356, 1985. 139. Starzl TE, Marchioro TL, Vonkaulla KN, et al: Homotransplantation of the liver in humans. Surg Gynecol Obstet 117:659–676, 1963. 140. Strouse PJ, Platt JF, Francis IR, et al: Tumorous intrahepatic lymphoproliferative disorder in transplanted livers. AJR Am J Roentgenol 167:1159–1162, 1996. 141. Sudan D, DeRoover A, Chinnakotla S, et al: Radiochemotherapy and transplantation allow long-term survival for nonresectable hilar cholangiocarcinoma. Am J Transplant 2:774–779, 2002. 142. Tanaka K, Uemoto S, Tokunaga Y, et al: Surgical techniques and innovations in living related liver transplantation. Ann Surg 217:82–91, 1993. 143. Taourel P, Bret PM, Reinhold C, et al: Anatomic variants of the biliary tree: Diagnosis with MR cholangiopancreatography. Radiology 199:521–527, 1996. 144. Theilmann L, Küppers B, Kadmon M, et al: Biliary tract strictures after orthotopic liver transplantation: Diagnosis and management. Endoscopy 26:517–522, 1994. 145. Uslu N, Aslan H, Tore HG, et al: Doppler ultrasonography findings of splenic arterial steal syndrome after liver transplant. Exp Clin Transplant 10(4):363–367, 2012. 146. Vallejo GH, Romero CJ, de Vicente JC: Incidence and risk factors for cancer after liver transplantation. Crit Rev Oncol Hematol 56:87–99, 2005. 147. Valls C, Alba E, Cruz M, et al: Biliary complications after liver transplantation: Diagnosis with MR cholangiopancreatography. AJR Am J Roentgenol 184:812–820, 2005. 148. Varotti G, Gondolesi GE, Goldman J, et al: Anatomic variations in right liver living donors. J Am Coll Surg 198:577–582, 2004. 149. Verdonk RC, Buis CI, van der Jagt EJ, et al: Nonanastomotic biliary strictures after liver transplantation. Part 2. Management, outcome, and risk factors for disease progression. Liver Transpl 13:725–732, 2007. 150. Vit A, De Candia A, Como G, et al: Doppler evaluation of arterial complications of adult orthotopic liver transplantation. J Clin Ultrasound 31:339–345, 2003. 151. Vivarelli M, Cucchetti A, La Barba G, et al: Ischemic arterial complications after liver transplantation in the adult: Multivariate analysis of risk factors. Arch Surg 139(10):1069–1074, 2004. 152. Waki K: UNOS Liver Registry: Ten year survivals. Clin Transpl 29–39, 2006. 153. Wall WJ: Liver transplantation for hepatic and biliary malignancy. Semin Liver Dis 20:425–436, 2000. 154. Wang F, Pan KT, Chu SY, et al: Preoperative estimation of the liver graft weight in adult right lobe living donor liver transplantation using maximal portal vein diameters. Liver Transpl 17(4):373–380, 2011. 155. Williams ED, Draganov PV: Endoscopic management of biliary strictures after liver transplantation. World J Gastroenterol 15(30):3725– 3733, 2009. 156. Worthy SA, Olliff JF, Olliff SP, et al: Color flow Doppler ultrasound diagnosis of a pseudoaneurysm of the hepatic artery following liver transplantation. J Clin Ultrasound 22:461–465, 1994.

CHAPTER 45  Liver Transplantation 157. Yerdel MA, Gunson B, Mirza D, et al: Portal vein thrombosis in adults undergoing liver transplantation: Risk factors, screening, management, and outcome. Transplantation 69:1873–1881, 2000. 158. Yu AS, Keeffe EB: Patient selection criteria for liver transplantation. Minerva Chir 58:635–648, 2003. 159. Zajko AB, Bennett MJ, Campbell WL, et al: Mucocele of the cystic duct remnant in eight liver transplant recipients: Findings at cholangiography, CT, and US. Radiology 177:691–693, 1990. 160. Zajko AB, Campbell WL, Logsdon GA, et al: Cholangiographic findings in hepatic artery occlusion after liver transplantation. AJR Am J Roentgenol 149:485–489, 1987.

1405

161. Zajko AB, Sheng R, Bron K, et al: Percutaneous transluminal angioplasty of venous anastomotic stenoses complicating liver transplantation: Intermediate-term results. J Vasc Interv Radiol 5:121–126, 1994. 162. Zavaglia C, De Carlis L, Alberti AB, et al: Predictors of long-term survival after liver transplantation for hepatocellular carcinoma. Am J Gastroenterol 100:2708–2716, 2005. 163. Zhou J, Fan J, Wang JH, et al: Continuous transcatheter   arterial thrombolysis for early hepatic artery thrombosis   after liver transplantation. Transplant Proc 37:4426–4429, 2005.

46  Pancreas Manuel Patino, Andrea Prochowski, Nisha Sainani, Onofrio Catalano, and Dushyant Sahani

“I’m tired of all this nonsense about beauty being only skin-deep. What do you want, an adorable pancreas?” —Jean Kerr Whether the pancreas is adorable or not is a debatable issue; however, advanced cross-sectional imaging has permitted generation of  excellent and exquisite images of the pancreas, which in turn has contributed immensely to the assessment of pancreatic diseases. Ultrasonography (US) has been used to evaluate abdominal pathologies, and its advantages include wide availability, low cost, and lack of radiation. Despite significant improvements and refinements in ultrasound technology, there are still inherent limitations when imaging deep in the abdomen, especially in large patients. Detailed visualization and definition of deeper and smaller structures and of subtle changes in density of the normal and abnormal pancreas are possible with images generated by multidetector computed tomography (MDCT) scanners. The scope of multiplanar reconstruction with MDCT scanners has added remarkably to the ability to visualize and understand complex anatomic structures and relationships. Dual-energy CT (DECT) technology, recently introduced, allows generation of material-density images founded on energy-based material separation, providing information about iodine distribution and tissue composition. In addition, processed energy-specific images can increase iodine enhancement, improving pancreatic lesion detection and characterization by increasing contrast between normal and abnormal tissue. Positron emission tomography (PET) CT allows the revelation of pathology at the molecular level. In addition, with the development of magnetic resonance hardware as well as new sequences and techniques, pancreatic parenchyma, ductal components, and relationships with surrounding structures can be defined with great detail. Even subtle lesions that do not alter the size and shape of the pancreas can be identified with highquality magnetic resonance imaging (MRI). Expertise in these modalities varies, although there is significant overlap in the information they can disclose.

ANATOMY The pancreas is an exocrine and endocrine organ approximately 15 cm long that is related to the stomach, duodenum, colon, and spleen. Considerable variations in pancreatic size, shape, and location are possible depending on body habitus and size and the position of contiguous organs (Figs. 46-1 to 46-9; Tables 46-1 to 46-3).

CT Anatomy The density of the nonenhanced pancreas is normally the same as that of soft tissue, between 30 and 50 Hounsfield units (HU); it increases to 100 to 150 HU after intravenous (IV) administration of iodine-

1406

based contrast agents.136,159 Rapid acquisition of images after a bolus injection allows visualization of normal enhancement during arterial, capillary, and venous phases. Homogeneous enhancement of the normal gland is a useful sign for excluding necrosis in pancreatitis. In the normal gland, variable normal fat tissue may be present, either along the margin of the gland or within the glandular portions. Fatty degeneration of the pancreas is common with aging; the entire pancreas may be replaced by fat, and the patient may have no clinical symptoms. Some pathologic conditions, such as chronic pancreatitis and cystic fibrosis, however, produce fatty infiltration and atrophy. The pancreas does not have a true fibrous capsule, and its contours may be smooth or lobular (Fig. 46-10). The pancreas is well visualized when peripancreatic retroperitoneal fat is abundant. The conspicuity of the gland can be improved in lean individuals and pediatric patients by the use of adequate oral contrast for opacifying the lumen of the stomach and bowel loops and by the use of IV contrast material for visualizing the vascular structures and pancreatic parenchyma.

MRI Anatomy The appearance of the pancreas on MRI depends on the specific pulse sequence. On T1-weighted images, the normal gland, owing to the aqueous protein content, reveals higher signal intensity than nonfatty tissue such as liver and muscle. On fat-suppressed T1-weighted sequences, the relatively high signal intensity of the pancreas increases further (Fig. 46-11B). On T2-weighted sequences, the normal pancreas is slightly hyperintense to muscle, whereas on fat-suppressed T2-weighted images, the contrast between the normal pancreas and surrounding suppressed fat is minimal (see Fig. 46-11A).229,269 The vessels are seen as flow voids on MRI (see Fig. 46-11C). On T1-weighted in-phase gradient echo (GRE) images, the pancreas appears hyperintense owing to the fat content, whereas it loses signal on T1-weighted opposed-phase images (see Fig. 46-11C and 46-11D). Being a very vascular organ, the pancreas shows intense contrast enhancement in the arterial phase, followed by a rapid washout (Fig. 46-12). This makes arterial-phase imaging crucial for accurate detection of most focal pancreatic lesions, which usually enhance less than the normal parenchyma and hence are rendered more conspicuous.117 The normal pancreatic duct measures 2 to 3 mm in caliber, demonstrates smooth margins, and increases slightly from the pancreatic tail to the head. The main pancreatic duct is routinely visualized on T2-weighted images and MR cholangiopancreatography (MRCP) as a high-signal-intensity structure (Fig. 46-13). Reviewing the source images and creating a maximum intensity projection (MIP) image can aid in visualizing and evaluating the entire duct (Fig. 46-14). The pancreatic duct receives 20 to 35 short tributaries or side branches that enter at right angles. These Text continued on p. 1415

CHAPTER 46  Pancreas

B CA

IVC A

Spl.art

A

B

N PV

B SMA

T

C

D

Bulb 2nd

E

H

F

FIG 46-1  Normal anatomy of the pancreas. A to I, Axial CT sections of the pancreas from superior to inferior. A and B, Superior portion of the body of the pancreas (B) with the splenic artery (Spl.art) posterior to the body. The splenic artery is a branch of the celiac artery (CA), which in turns arises from the aorta (A). IVC, inferior vena cava. C and D, The oblique orientation of the long axis of the pancreas is seen with the tail (T) extending from the hilum toward the midline of the body (B). The body continues as the neck (N) anterior to the portal vein (PV) and superior mesenteric artery (SMA), which in turn continues as the head. E and F, The head lies in the C loop of the duodenum, bound superiorly by the bulb and medially by the second part of the duodenum (2nd). The common bile duct is seen as a low-density structure posterior to the head (H) of the pancreas (arrow in F). Branches of the pancreaticoduodenal arcade are seen around the head of the pancreas. Continued

1407

1408

PART II  CT and MR Imaging of the Whole Body

SMV SMA U

IVC

G

H

I FIG 46-1, cont’d  G and H, The head lies anterior to the inferior vena cava (IVC) and is bound medially by the superior mesenteric vein (SMV). The head turns around and inferiorly to the triangular uncinate process (U), which lies posterior to the superior mesenteric artery (SMA) and vein. I, The inferiormost portion of the uncinate process is seen anterior to the superior mesenteric artery and vein.

FIG 46-2  Coronal reformatted MDCT of the pancreas reveals a normal

FIG 46-3  Axial MDCT shows a normal gradually tapering pancreas.

pancreatic groove. The pancreatic groove (arrow) is the region between the head of the pancreas and the C loop of the duodenum.

Most commonly the head is larger than either the body or the tail, with the tail being the narrowest.

CHAPTER 46  Pancreas

1409

A FIG 46-4  Axial MDCT of a dumbbell pancreas. In this variant the head and tail are larger, with a narrow neck.

B FIG 46-6  Curved axial (A) and coronal (B) reformatted MDCT images show the main pancreatic duct (duct of Wirsung) in its entirety.

A

B FIG 46-5  A, Bulbous and redundant body and tail of the pancreas (arrow) mimicking a mass lesion on axial MDCT. B, Axial T2-weighted MRI confirms the same finding (arrow) without any altered signal intensity.

1410

PART II  CT and MR Imaging of the Whole Body

Spl. Art

CA

CHA

SMA SMA

B A

CHA

PV

Spl. Art.

Spl. Vein

CA

C

FIG 46-7  High-quality MDCT angiographic images of the major vessels in relation to and supplying the pancreas.  A, Coronal section shows the common hepatic artery (CHA) and splenic artery (Spl. Art) originating from the celiac artery (not shown). The proximal portion of the superior mesenteric artery (SMA) originating from the aorta is visualized. B, Sagittal section in another patient reveals the origin of the celiac artery (CA) and superior mesenteric artery (SMA) from the aorta. Calcified plaques are seen along the aorta. C, Axial section reveals the celiac artery (CA) originating from the aorta, which in turn gives off its major branches, the common hepatic artery (CHA) and the splenic artery (Spl. Art.). The portal vein (PV) and splenic vein (Spl. Vein) are also visualized.

Spl. Art N H SMA Branch vessels

U SMA SMV

FIG 46-9  Coronal reformatted MDCT reveals the head (H), neck (N),

FIG 46-8  Coronal MDCT angiography reveals parenchymal feeding vessels in the region of the pancreatic head. The superior mesenteric artery (SMA) and splenic artery (Spl. Art) are visualized.

and uncinate process (U) of the pancreas. The superior mesenteric artery (SMA) arises from the aorta posterior to the neck of the pancreas and runs inferiorly, anterior to the uncinate process, along with the superior mesenteric vein (SMV).

CHAPTER 46  Pancreas TABLE 46-1  Anatomic Relationships of

the Pancreas

• Located in anterior pararenal space of retroperitoneum • Anterior to perirenal (Gerota’s) fascia and posterior to parietal peritoneum • Divided into head, uncinate process, neck, body, and tail from right to left (see Fig. 46-1) • Long axis of body and tail of pancreas is in an oblique orientation extending from hilum of spleen to midline of the body, where pancreas lies anterior to portal vein, which is the point of transition from the body to the neck (see Fig. 46-1C-F). • Region between head of pancreas and C loop of duodenum is known as the pancreatic groove (see Fig. 46-2). • In postnephrectomy cases or with agenesis of kidney or ectopic kidney, pancreas moves posteriorly to partially fill in the empty renal fossa; its soft tissue density should not be mistaken for recurrent tumor. Head • Located in C loop of duodenum • Bounded superiorly by the bulb • Laterally by second portion of duodenum • Inferiorly by third portion of duodenum • Medially by superior mesenteric vein • Anterior to inferior vena cava (see Fig. 46-1E-G) Uncinate process • Inferiormost portion of head of pancreas • Triangular or wedge shaped • Lies just posterior to superior mesenteric artery and vein (see Fig. 46-1G-I) Neck • Anterior to superior mesenteric artery • Vary greatly in thickness (see Fig. 46-1B-D) Tail • Anterior to splenic vein • Reaches splenic hilum (see Fig. 46-1C)

1411

TABLE 46-2  Pancreatic Ductal System • With curved reformats created using multislice CT scanners, entire pancreatic duct can be visualized in a single plane (see Fig. 46-6). • Course of pancreatic duct can vary considerably; most commonly it is descending (50% of cases). • At the point of embryologic fusion of ducts of Santorini and Wirsung in pancreatic neck, the duct may narrow slightly or demonstrate a loop that may be confused with a stricture. Main Pancreatic Duct (Duct of Wirsung)

Accessory Pancreatic Duct (Duct of Santorini)

• Fusion of dorsal duct in body and tail, with ventral duct in uncinate process • Joins common bile duct at ampulla of Vater • Empties into duodenum through major papilla • Major drainage route in 91% of individuals174

• Located in upper portion (head) of pancreas • More horizontal than duct of Wirsung • Drains proximal to major papilla, in minor papilla • Present in 44% of individuals • Major drainage route in 9% of individuals174

Size • Anteroposterior measurements in CT140: • Head: 23 ± 3 mm • Neck: 19 ± 2.5 mm • Body: 20 ± 3 mm • Tail: 15 ± 2.5 mm • Ultrasound measurements tend to be smaller. • Gland shrinks with age, but head-to-body ratio remains almost constant181 Shape • Pancreas can have several basic shapes. • Most common is gradually tapering pancreas: head larger than body and tail (see Fig. 46-3) • Dumbbell-shaped pancreas: large head and tail, narrow neck (see Fig. 46-4) • Both ends should be about equal in size. • If one end is larger than the other, pathologic process should be considered (see Fig. 46-5).

FIG 46-10  Axial curved reformatted MDCT of the entire pancreas reveals a lobulated contour (the pancreas lacks a true fibrous capsule).

1412

PART II  CT and MR Imaging of the Whole Body

TABLE 46-3  Pancreatic Vascular Anatomy • High-quality angiographic representation can be reconstructed using MDCT and angiography (see Fig. 46-7) in which not only major vessels but also tiny branches and parenchymal feeding vessels can be seen (see Fig. 46-8). • Calcifications commonly occur in splenic artery and should not be mistaken for pancreatic calcifications. Arterial Blood Supply

Veins of the Pancreas

• • • •

• Splenic vein runs along posterior and superior aspects of pancreas along with splenic artery • Splenic vein joins SMV behind neck of pancreas to form portal vein • Draining veins are parallel to arteries • Veins drain into pancreaticoduodenal, superior mesenteric, splenic, or portal veins • When pathologic occlusion of any of these vessels occurs, collateral vessels may enlarge and become visible.

From arteries originating from celiac and superior mesenteric arteries Celiac trunk is located superior to body of pancreas Splenic artery (branch of the celiac) courses along superior margin of pancreas Superior mesenteric artery (SMA) arises from aorta posterior to neck of pancreas and runs inferiorly, anterior to uncinate process, along with superior mesenteric vein (SMV) (see Fig. 46-9) • Main arteries are almost always present, but origin, size, and type of collateral supply may vary • Dorsal pancreatic • Great pancreatic • Transverse pancreatic • Pancreaticoduodenal arcade • Anterior and posterior superior pancreaticoduodenal branches of gastroduodenal artery anastomose with corresponding inferior pancreaticoduodenal branches of SMA to form an arterial arcade around head of pancreas

Lymphatic Drainage • Parallels venous drainage • Small lymph nodes around vessels and in various ligaments are commonly seen. However, if nodes are ≥10 mm, neoplasm should be suspected.

A

B

C

D FIG 46-11  A, Axial T2-weighted fat-suppressed image reveals the normal pancreas to be mildly hyperintense compared with liver and muscle. B, Axial T1-weighted fat-suppressed GRE image reveals a high-signalintensity pancreas compared with liver and muscle. C, The pancreas appears hyperintense on this axial T1-weighed GRE in-phase image owing to its fat content. Flow voids in the vessels (splenic artery and portal vein) are visualized (arrows). D, The pancreas loses its signal on the opposed-phase image.

CHAPTER 46  Pancreas

A

B FIG 46-12  Axial T1-weighted fat-suppressed noncontrast (A) and postcontrast (B) GRE images. The contrastenhanced image shows bright signal in the aorta and splenic artery (arrow) and intense contrast enhancement of the pancreas.

A

B FIG 46-13  Axial (A) and coronal (B) T2-weighted images reveal the normal pancreas and the pancreatic duct partially (arrow).

1413

1414

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F

FIG 46-14  Multisection coronal T2-weighted thin-slab single-shot FSE MRCP. A to F, Series of contiguous thin sections (generally 4 mm thick) acquired in the coronal plane with respiratory triggering; the images were acquired during expiration to allow motion-free imaging. When viewed sequentially, they reveal the pancreatic duct (arrow) and adjacent organ structures. A small pancreatic cyst (thick arrow in F) is seen inferior to the main pancreatic duct in the region of the uncinate process.

CHAPTER 46  Pancreas

A

1415

B

C FIG 46-15  Pancreas divisum type 1. A, Curved coronal reformatted MDCT image. B, Curved axial reformatted MDCT image. C, MRCP in a patient with pancreas divisum. The dorsal duct system (arrowheads in A, B, and C) drains through the duct of Santorini (thin arrow in A and C) into the minor papilla and above the opening of the duct of Wirsung (black arrow in C) into the major papilla. Note the confluence of the duct of Wirsung (black arrow) with the common bile duct (thick arrow in C).

side branches are not usually seen on MRI or MRCP unless they are dilated.84 Visualization of the pancreatic duct and its side branches on MRCP can be enhanced by administration of secretin.166

Embryology and Developmental Anomalies and Variants Congenital anomalies of the pancreas are usually discovered in childhood, although some may be detected later in life owing to the development of clinical symptoms related to the anomaly itself or symptoms mimicking acquired diseases, mainly pancreatic cancer (Figs. 46-15 to 46-21; Table 46-4).

IMAGING TECHNIQUES CT Protocols and Technique Biphasic pancreatic MDCT (pancreatic parenchymal phase [PP] and a portal venous phase [PVP] of contrast enhancement) is considered the most sensitive technique for both detection and staging of pancreatic adenocarcinoma (PDAC).76 It can also be applied to evaluation of other pathologies such as pancreatic cystic tumors. Both maximal pancreatic parenchymal enhancement and maximal tumor-to-  parenchyma attenuation differences usually occur in the PP, which is therefore the most sensitive phase to detect PDAC. Although tumorto-parenchyma attenuation differences tend to decrease during the PVP, the sensitivity for vascular invasion is highest in the PVP of dynamic imaging.76 Scan delays for PP and PVP can be fixed or established on an individual basis according to the patient’s personal

hemodynamics and injection rate. With the first approach a scan delay of 40 to 50 and 65 to 70 seconds is typically used for the PP and PVP, respectively. Nowadays, with the ability to scan the whole pancreas in less than 4 seconds and the possibility of scanning before consistent parenchymal enhancement has been achieved, such fixed delays may be inappropriate in many patients.228 The contrast material injection rate must also be taken into account when choosing the scanning protocol. It has been demonstrated that increasing the contrast injection rate shortens the aortic transit time, increases peak parenchymal pancreatic attenuation, does not change adenocarcinoma mean attenuation, and therefore increases the conspicuity of the tumor against the better-enhanced pancreatic parenchyma. It has also been demonstrated that individualized scan delays perform better than fixed ones.228 Usually 500 mL of water as a neutral contrast agent is given to the patient 20 to 30 minutes before the examination, and another 250 mL is given just before the start of the scan to achieve distention of the stomach and duodenum and improve detection of pancreatic abnormalities and infiltration or extension into surrounding anatomic structures, without masking common bile duct stones, pancreatic duct stones, or calcifications.253 A noncontrast examination of the abdomen is performed to detect calcifications and to clearly determine the location and extension of the pancreas in each patient. Subsequently, a PP acquisition extending from the diaphragm to below the third portion of the duodenum, and a PVP scan covering the whole abdomen are performed. The technical parameters used for each of these phases are detailed in Table 46-5. The raw data are subsequently

1416

PART II  CT and MR Imaging of the Whole Body processed to obtain multiplanar reconstructions, including MIPs and curved reconstructions along the course of the pancreatic duct. Using a structured report template helps ensure a complete description and reporting of imaging findings.31

Dual-Energy CT

A

B FIG 46-16  Pancreas divisum type 3. A, On MRCP the small filamentous communication between the ventral (arrow) and dorsal (arrowhead) pancreatic duct system is not visualized. B, On the corresponding ERCP the small communication (arrow) between the dorsal (two arrowheads) and ventral (single arrowhead) duct systems is displayed.

DECT technology has been recently introduced into clinical practice, allowing new applications including material decomposition on the basis of energy-dependent attenuation profiles of specific materials. This application provides additional qualitative and quantitative information of iodine distribution and tissue composition. For instance, material density–iodine (MD-I) images can improve conspicuity of isodense pancreatic masses including PDAC (Fig. 46-22), inflammatory lesions, and small neuroendocrine tumors170 (Fig. 46-23). Moreover, margins and extension of pancreatic lesions are more reliably defined along with a more accurate depiction of intralesional septa and nodules in cystic neoplasms (Fig. 46-24). MD-I images can contribute to the staging of pancreatic cancer because of the improved ability to assess vascular invasion, peripancreatic fat stranding, and distant metastasis. MD-I images can also improve characterization of surgical bed structures, allowing better differentiation between the postsurgical changes and local recurrence. Virtual unenhanced (VUE) images can be generated from contrast-enhanced acquisitions by a mathematical subtraction of the iodine. These images can substitute the true unenhanced images without losing depiction of small foci of calcium, fat, and hemorrhage, eliminating the need for unenhanced acquisition and reducing radiation dose1,93 (Fig. 46-25). DECT also offers the ability to generate virtual monochromatic (VMC) images. These processed images simulate CT acquisitions at different x-ray energies (keV), offering the ability to generate images at specific energies (40 keV to 140 keV). Images created with energies around 70 keV are similar to those commonly acquired with single-energy (SE)CT (120 kVp). Low-kVp images increase conspicuity of hypoattenuating lesions by improving contrast between normal pancreas and abnormal tissue.155,160 DECT is a growing technology currently under research for different pancreatic diseases. However, efforts have focused on pancreatic cancer, owing to the diagnostic difficulties with SECT and the nature of the disease. Research studies have demonstrated better depiction of pancreatic cancer at low-energy images (52 keV) rather than images acquired at 120 kVp or processed at 70 keV. Text continued on p. 1421

*

A

B FIG 46-17  Annular pancreas. A, Axial T1-weighted fat-suppressed MRI demonstrates pancreatic tissue (arrows) completely circumscribing the descending portion of the duodenum (asterisk). B, Corresponding MRCP shows the extrinsic compression around the second portion of the duodenum (arrow).

CHAPTER 46  Pancreas

1417

* * A

B FIG 46-18  Pancreatitis in annular pancreas. Coronal FIESTA (fast imaging employing steady-state acquisition) image (A) and axial T1-weighted fat-suppressed image (B) display annular pancreas (arrows in A) encircling the second portion of the duodenum (asterisk). Normal pancreatic tissue reveals high signal intensity (arrow) on the T1-weighted image (B), whereas the inflamed tissue is hypointense (arrowhead).

*

FIG 46-19  Annular pancreas. MDCT shows the annular pancreas partially encircling and narrowing the second portion of the duodenum (arrow). The annular pancreas is uniform in density. The pancreatic duct is indicated by the arrowhead.

* *

*

FIG 46-20  Axial contrast-enhanced MDCT reveals complete agenesis of the dorsal pancreas. Note the rounded head of the pancreas (arrow) and the absence of the neck and body. Also note polysplenia (asterisks) and the abnormal position of the bowel loops (arrowheads) behind the stomach.

FIG 46-21  Axial contrast-enhanced MDCT reveals contour lobulations (asterisk) of the pancreatic head, extending horizontally and laterally to the pancreatic-duodenal artery (arrow). The lobulations have welldefined margins, no signs of infiltration of the surrounding structures, and a homogeneous density similar to that of the remaining normal pancreas.

1418

PART II  CT and MR Imaging of the Whole Body

TABLE 46-4  Embryology and Developmental Anomalies and Variants Pancreas divisum

5% to 14% of population119 Failure of fusion of the dorsal and ventral ducts Separate drainage into duodenum Predominant drainage occurs through dorsal duct system (duct of Santorini). Classified into three types (see Figs. 46-15 to 46-16): • I (classic form): complete lack of fusion between dorsal and ventral duct systems • II: absent duct of Wirsung • III: small communicating branch connects dorsal and ventral duct systems Increased incidence in patients with idiopathic pancreatitis Acute recurrent pancreatitis has been demonstrated in 49% of those with symptomatic pancreas divisum. ERCP: considered the gold standard for diagnosis of this entity MRCP: 73% accurate in diagnosing complete pancreas divisum MRCP is less accurate in diagnosis of incomplete pancreas divisum. MDCT, with the aid of multiplanar reconstructions, may be helpful in diagnosing pancreas divisum by depicting the dominant dorsal pancreatic duct draining directly into the minor papilla. On cross-sectional imaging, pancreatic tissue is homogeneous without focal areas of architectural distortion or altered texture.130,212

Annular pancreas

Occurs in 1 in 20,000 births Defined as a portion of pancreatic tissue in continuity with the head that partially or completely circumscribes duodenum Usually encircles second portion of duodenum (74% of cases), less often first portion (21%), rarely third portion163 Usually associated with duodenal anomalies such as atresia, atrophy, or stenosis Up to 36% of patients with annular pancreas also have pancreas divisum.52,279 Usually discovered in childhood because of upper GI obstruction Classified into six types depending on site of drainage of annular duct277 Type I (most common type): annular duct drains into major papilla after joining duct of Wirsung from dorsal side ERCP discloses an aberrant pancreatic duct encircling duodenum. Both MRI and MDCT reveal pancreatic tissue encircling duodenum, which retains density and signal intensity of remaining normal pancreas. T1-weighted fat-suppressed axial images are particularly useful. MRCP may disclose associated aberrant duct encircling duodenum and clearly shows extrinsic obstruction around second portion of duodenum (see Figs. 46-17 to 46-19). In adults, must be differentiated from locally infiltrating pancreatic or duodenal cancer

Agenesis of dorsal pancreas

Complete agenesis of pancreas incompatible with life Agenesis of dorsal pancreas is more frequent than agenesis of ventral pancreas. Agenesis of dorsal pancreas can be complete or partial (more common). Complete agenesis of the dorsal pancreas: body and tail of pancreas and whole dorsal duct system, including minor papilla and accessory duct, are absent. CT/MRI: visible head of pancreas retains normal pancreatic lobulations; density and signal intensity are homogeneous both at baseline study and following contrast administration. MRCP: pancreatic duct in head of pancreas is short; dorsal pancreatic duct and accessory duct cannot be found (see Fig. 46-20). Partial agenesis of the dorsal pancreas: distal part of pancreatic body or at least a remnant of accessory duct and minor papilla are found.150,180,227 Dorsal pancreatic agenesis is usually associated with other congenital variants (polysplenia syndrome, lobulation of the liver, intestinal malrotation). Recurrent or chronic pancreatitis occurs mainly in patients with partial agenesis. Diabetes is found in 67% of the cases. Agenesis of dorsal pancreas must be distinguished from reduction in size of pancreas due to pancreatitis and pancreatic cancer with upstream atrophy. If a dorsal duct of normal length is documented at MRCP, partial or complete agenesis of dorsal pancreas is ruled out and other possibilities need to be entertained.83,180

Pancreatic head lobulations

Contour anomalies of head and neck of pancreas composed of normal pancreatic tissue May be found isolated or in association with other anomalies Lobulations of head of the pancreas >1 cm in maximal dimension have been found in 35% of general population. Clinical relevance: possible misinterpretation of lobulations with pancreatic neoplasms CT and MRI demonstrate normal density and signal intensity of lobulations, paralleling characteristics of normal pancreas. Margins are well defined; no signs of infiltration of surrounding structures, bile duct dilatation, or vascular encasement (see Fig. 46-21) Lobulations classified into three types according to their relation to gastroduodenal or pancreatic-duodenal artery: • Type 1: lobulation is directed anteriorly to artery. • Type 2: lobulation is directed posteriorly. • Type 3: lobulation is directed horizontally and laterally.217

CHAPTER 46  Pancreas

1419

TABLE 46-4  Embryology and Developmental Anomalies and Variants—cont’d Ectopic pancreas

Uneven pancreatic lipomatosis

May be found in different portions of GI tract Most commonly in proximal segments and Meckel’s diverticulum Tissue may be responsible for clinical symptoms related to secretion of pancreatic enzymes. Diagnosis not easy to make96 Not a congenital variant of pancreas Fatty infiltration of pancreas most common histologic change observed Age-related process that causes reduction in size of pancreas and increased evidence of its lobular structure Certain diseases (e.g., cystic fibrosis, diabetes) may be associated with increased fatty replacement of pancreas. When fatty replacement is focal or unevenly distributed through pancreas, condition is known as uneven pancreatic lipomatosis

TABLE 46-5  CT Scan Protocol for Pancreas Noncontrast (optional)

Parenchymal Phase

Portal Venous Phase

MDCT (16 and beyond) Thickness (mm) Interval (mm) FOV Tube Voltage (kV) Tube Current (mA) Delay Pitch Rotation Time Location

5 5 Large 120 80-160 — 1-1.5 0.5 Abdomen

1.25 0.625 Large 80-120 180-280 50 sec 1-1.5 0.5 Upper abdomen

5* 5* Large 80-120 180-280 70 sec 1-1.5 0.5 Abdomen

Coronal Reformats Thickness (mm) Interval (mm)

— —

1 0.6

3 3

*Retro-reconstructed to 1.25-mm thickness, 0.625-mm interval. FOV, field-of-view; MDCT, multidetector computed tomography.

1420

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 46-22  Pancreatic adenocarcinoma on DECT. Axial contrast-enhanced SECT image (A), DECT-VMC image at 65 keV (B), MD-I image (C), and color overlay iodine image (D). It is challenging to detect the lesion in the tail of the pancreas in SECT images because of the similar attenuation between the lesion and the adjacent pancreatic tissue. DECT images improve detection of pancreatic lesions. Note improved contrast between the hypoattenuating lesion (arrows) and the surrounding parenchyma on low monochromatic energies (65 keV) and iodine images.

A

B

C

D

FIG 46-23  Pancreatic neuroendocrine tumor on DECT. Axial contrast-enhanced SECT image (A), DECT-VMC image at 65 keV (B), MD-I image (C), and color overlay iodine image (D). Note the improved visualization of the lesion in the pancreatic tail, best seen on iodine images.

CHAPTER 46  Pancreas

A

B

C

D

1421

FIG 46-24  Axial contrast-enhanced SECT image (A), DECT-VMC image at 65 keV (B), MD-I image (C), and color overlay iodine image (D). Note the presence of a 3-cm hypoattenuating lesion in the body of the pancreas. MD-I images help to characterize this lesion by improving visualization of lesion morphology and depiction of cystic lesions (borders and septa) and excluding the presence of solid component or enhancing nodules. The lesion was diagnosed as a serous cyst adenoma.

A

B

C

FIG 46-25  DECT virtual noncontrast images. Axial true unenhanced SECT image (A), DECT virtual unenhanced image (B), and DECT contrast-enhanced image (C). The virtual unenhanced images obtained from a contrast-enhanced DECT acquisition show similar image quality and diagnostic performance compared to true noncontrast images. VNC images can potentially eliminate the need of a true noncontrast series, reducing radiation exposure.

Although there is no consensus on a pancreatic DECT protocol, Table 46-6 outlines the approach used at our department for pancreatic diseases.

MRI Protocols and Technique The pancreas is best imaged with a high-field-strength system (≥1.5 tesla [T]) and high-performance gradients that enable the use of fast MR sequences and phased-array torso coils.117,229 The protocol for MRI of the pancreas is detailed in Table 46-7.

T2-Weighted Images.  Breath-hold or respiratory-triggered multisection, thin-slice, T2-weighted images are obtained in the axial and coronal planes. With these sequences, fluid-filled lesions in or around the pancreas appear hyperintense; the common bile duct and the pancreatic duct are seen in cross sections and may be used to guide the acquisition of MRCP series. Fat suppression can increase the conspicuity of hyperintense lesions (see Fig. 46-11A).27

T1-Weighted Images.  These images can be acquired either with a fast-spin

echo

(FSE)

T1-weighted

sequence

using

multiple

1422

PART II  CT and MR Imaging of the Whole Body

TABLE 46-6  DECT Pancreas Protocol 1 2 3

Administration of negative oral contrast + scout Administration of IV contrast media from 80-120 mL + 40 mL saline at 4 mL/sec CT phase Scan parameters Pancreatic phase (PP) (only abdomen) after PP 140 kVp 50 sec from injection Thickness/interval 2.5 mm/2.5 mm

4

Portal-venous phase (PVP) (abdomen and pelvis) after 70 sec from injection

PVP 120 kVp Thickness/interval 5 mm/5 mm

Generated images PP MD iodine VUE from PP PP VMC 50 keV + PP VMC 65 keV PP VMC 65 keV thickness/interval 1.25 mm/0.625 mm MIP and curved MPR reconstruction PP coronal + sagittal 3 mm PVP coronal + sagittal 3 mm/3 mm MIP and curved MPR reconstruction

keV, kilo-electron voltage; kVp, kilo-voltage peak; MD, material-density; VMC, virtual monochromatic; VUE, virtual unenhanced.

TABLE 46-7  MRI Protocol for Pancreas MRI Localizer Precontrast

MRCP Presecretin Postsecretin* 3D MRCP

Sequence

Plane

Slice Thickness (mm)

Other Specifications

T1 (fast GRE) T2 (FSE) T1 (FSE or GRE) T1 in-phase/opposed-phase (GRE) T1 (GRE) T1 (GRE) delayed

Axial, coronal, sagittal Axial, coronal Axial Axial Axial, coronal Axial, coronal

4-6 4-6 4-6 4-6 4-6

± FS, BH/respiratory triggered ± FS, BH ± FS, BH Postcontrast FS, BH FS, BH

T2 2D MRCP (ssFSE—thin slab) T2 2D MRCP (ssFSE—thick slab) T2 2D MRCP (ssFSE—thick slab) T2 (ssFSE—thin slab)

Coronal, coronal oblique, axial Coronal, coronal oblique, axial Axial, coronal, coronal oblique Axial, coronal

4 20-50 20-50 1-1.5

BH BH BH, every 30 sec for 10-15 min BH

*Secretin is administered at a dose of 0.2 μg/kg intravenously (for >50 kg, 18.5 μg). BH, breath-hold; FS, fat-suppressed; FSE, fast-spin echo; GRE, gradient echo; MRCP, magnetic resonance cholangiopancreatography; ssFSE, single-shot fast-spin echo; 2D, two dimensional; 3D, three dimensional.

breath-hold acquisitions or with a single breath-hold GRE sequence with a water-stimulation prepulse, with or without fat suppression (see Fig. 46-11B). Fat suppression or selective water stimulation (1) improves delineation of the pancreas, which appears homogeneously bright compared with surrounding low-intensity fat; (2) is excellent for identifying a pancreatic mass or focal pancreatitis, which is less intense than the normal high-intensity pancreas; and (3) is suitable for contrast-enhanced studies. However, breath holding requires patient cooperation, and these sequences are less accurate in demonstrating focal disease in the presence of diffuse chronic pancreatitis because of the homogeneous low signal intensity of the fibrotic pancreas.27

Chemical Shift (In-Phase/Opposed-Phase) Imaging.  A hypoattenuating lesion in the pancreas on contrast-enhanced CT is nonspecific, and a non–border-deforming neoplasm cannot be ruled out. Although CT is accurate in revealing macroscopic fat (e.g., in pancreatic lipomas), it is less accurate in revealing microscopic lipid deposits. Fat-containing lesions in the pancreas have signal intensity similar to or higher than that of the remainder of the pancreas on in-phase T1-weighted GRE images (see Fig. 46-11C). This finding is in contradistinction to most pancreatic malignancies, which appear as areas of low signal intensity on T1-weighted sequences. Loss of signal intensity

on an opposed-phase T1-weighted GRE image establishes the lipid nature of the focal abnormalities (see Fig. 46-11D).110,122,269

Contrast-Enhanced MRI.  A high-resolution axial and coronal dynamic breath-hold gadolinium (Gd)-enhanced (0.1 mmol/kg body weight) T1-weighted three-dimensional (3D) fast spoiled GRE sequence is used to detect and characterize pancreatic mass lesions and assess inflammatory processes such as pancreatitis. This allows a combination of pancreatic parenchymal evaluation and MR angiography to assess possible vascular involvement (see Fig. 46-12).166,176,210 The time required for contrast to reach the abdominal aorta can be determined by a test bolus injection, MR fluoroscopy, or triggering software. Imaging at 15 seconds and 35 to 45 seconds after arrival of contrast in the abdominal aorta allows acquisition of high-quality images of pancreatic parenchyma and peripancreatic vessels.92 Alternatively a triplephase acquisition can be done by acquiring images in arterial (30-40 seconds), portal (70-80 seconds), and equilibrium (180 seconds) phases. Fat-suppressed high-resolution T1-weighted images acquired 10 minutes after administration of mangafodipir trisodium (5 µmol/ kg body weight) in a slow bolus over 1 to 2 minutes are useful for lesion detection in equivocal cases, for tumor detection, and for differentiating malignant from benign inflammatory masses.117 The basis for using mangafodipir is that normal pancreatic parenchyma enhances

CHAPTER 46  Pancreas following mangafodipir administration and becomes hyperintense on T1-weighted images, whereas tumors do not enhance.125

Magnetic Resonance Cholangiopancreatography MRCP has emerged as an accurate noninvasive means of evaluating the pancreatic duct. It uses heavily T2-weighted sequences that depict the fluid-containing pancreatic duct as a high-signal-intensity structure. MRCP can be acquired as two-dimensional (2D) or 3D breathingaveraged or breath-hold FSE sequences and represents a viable alternative to diagnostic ERCP.84,85,126,163,166 Two different and complementary approaches are generally used for MRCP: a 2D thick-slab single-shot fast (turbo)-spin echo (ssFSE) T2-weighted sequence and a 3D multisection thin-slab T2-weighted sequence.84,163 An additional approach is secretin-enhanced MRCP, which is a particular thick-slab MRCP acquired before and after administration of secretin (Figs. 46-26 to 46-28; Table 46-8).

1423

Through a combined qualitative and semiquantitative evaluation, FDG PET CT can help in diagnosing pancreatic cancer when other imaging modalities are equivocal, characterizing lymph nodes, preoperative staging, depicting metastases and tumor recurrence, and monitoring response to therapy. Moreover, PET CT can have a role in

PET CT Technique Fused PET CT is a recently developed technology that combines the functional information of PET with the anatomic details of CT, providing both PET and contrast-enhanced CT images of the entire body in one setting.139,220 On PET studies, malignant lesions generally demonstrate avid fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake, and most benign lesions are characterized by normal or minimal FDG uptake.25,199 However, false-positive and false-negative results may occur with FDG PET, and its inherent low spatial resolution may interfere with precise anatomic localization of the findings.215,262 Multidetector CT has the ability to image the entire abdomen and pelvis in a single breath hold with high spatial resolution, obtaining multiplanar reformats such as 3D and angiographic reconstructions, but it offers limited functional information. The limitations of CT for detection of pancreatic tumors are increasingly being recognized; a correlation between tumor size and sensitivity exists.248 Moreover about 10% of PDACs and pancreatic metastases are isoattenuating at contrast-enhanced CT and not depicted even when larger than 2 cm.209

A

FIG 46-26  Coronal T2-weighted thick-slab single-shot FSE MRCP provides an excellent display of the extrahepatic biliary tract and pancreatic duct. Residual fluid is seen in the duodenum (short arrows), which can be avoided by fasting or using a T2-negative oral contrast medium. A small pancreatic cyst (long arrow) is seen inferior to the main pancreatic duct; there is a subtle long thin communication.

B FIG 46-27  A, Coronal T2-weighted single-shot FSE MRCP reveals a pancreatic cyst (short arrow) overlapping the pancreatic duct (long arrows). B, 3D MRCP with rotation of image plane helps separate the cystic lesion from the pancreatic duct (thin arrows) and reveals the short communication (thick arrow).

1424

PART II  CT and MR Imaging of the Whole Body

A B

C FIG 46-28  A, Coronal MRCP reveals the pancreatic duct (arrowheads) and small hyperintense lesions (arrows) adjacent to the duct. These lesions probably represent side-branch intraductal papillary mucinous neoplasms (IPMNs); however, the communication with the pancreatic duct is not distinctly visualized. B and C, Postsecretin coronal projection in different planes improves the conspicuity of the pancreatic duct (arrowheads), side-branch IPMNs (arrows), and communication with the pancreatic duct (horizontal arrow in C).

diagnosing focal mass-forming autoimmune pancreatitis and guiding biopsy to sites that accurately represent the suspected pathology.220

PATHOLOGIC CONDITIONS Inflammatory Processes Acute Pancreatitis.  Acute pancreatitis is an acute inflammatory process that is followed by complete restoration of structural and functional normalcy after the attack subsides, provided no part of the pancreas has been destroyed by necrosis.235 The most common causes of acute pancreatitis are choledocholithiasis and ethanol abuse. Other causes include trauma, metabolic disorders (hyperlipidemia, hypercalcemia), ERCP-induced pancreatitis, medications (azathioprine, sulfonamides), tumors, and congenital anomalies such as pancreas divisum.149 Most patients present with nausea, vomiting, abdominal pain, and rebound tenderness. Fever, tachycardia, and leukocytosis are often present. The diagnosis is usually established by detection of elevated levels of pancreatic enzymes in blood, urine, or both.8 Once the clinical diagnosis of acute pancreatitis is made, treatment depends on the initial assessment of disease stage and severity. An increased frequency of death from acute pancreatitis is directly correlated with the development of multiorgan failure and the extent of pancreatic necrosis, which is a grave prognostic indicator.19 Whereas mortality rates of less than

1% are reported in patients with interstitial pancreatitis, mortality increases dramatically to 10% to 23% in patients with necrotizing pancreatitis.20,28 Virtually all life-threatening complications occur in patients with necrotizing pancreatitis. Secondary bacterial contamination occurs in 40% to 70% of patients with pancreatic necrosis, and it constitutes a major risk for death.214 Classification.  Acute pancreatitis was initially classified into mild acute (interstitial edematous) and severe acute (necrotizing) types.11,28 The classification was recently revised and updated, distinguishing between two phases of acute pancreatitis: an early phase (within the first week) and a late phase (after the first week).251 The early phase is characterized by pancreatic inflammation with variable degrees of peripancreatic edema and ischemia that can resolve or progress to necrosis; this phase is defined by clinical parameters such as organ failure. Patients with organ failure that resolves in 48 hours are considered to have mild pancreatitis without complications and a mortality rate of 0%. Patients are considered to have severe acute pancreatitis if organ failure lasts for more than 48 hours, which carries an increased risk of complications. The late phase is characterized by increasing necrosis, infection, and persistent multiorgan failure. This phase is defined by a combination of imaging findings and clinical parameters.251 Contrast-enhanced CT is able to depict and quantify pancreatic parenchymal injury, except

CHAPTER 46  Pancreas

1425

TABLE 46-8  Different Techniques of MRCP Thick-Slab ssFSE T2-Weighted Sequence (Projective MRCP)

Multisection Thin-Slab ssFSE T2-Weighted Sequence

Two-dimensional single thick (20-50 mm) section that can be obtained in any plane with a single short breath hold (<3 seconds). Final image resembles conventional image obtained during ERCP (see Fig. 46-26).85,162 Limitations include need for patient cooperation and degradation of ductal visibility owing to overlap with other fluid-containing organs (i.e., stomach, duodenum), ascites, or peripancreatic fluid. A T2-negative oral contrast agent can be administered in these cases (see Fig. 46-13). Could be useful adapting section thickness, orientation, and positioning of thick slab to patient’s duct anatomy, which requires on-site tailoring by radiologist during image acquisition (see Fig. 46-27). Adjacent solid organs are not visualized.163

Series of contiguous thin sections (generally 4 mm thick) acquired with a shorter echo time and echo train length; each single shot lasts <1 sec. Image quality is hence less dependent on patient cooperation and nearly unaffected by artifacts (chemical shift, magnetic susceptibility) or motion for the possibility to acquire each section with respiratory triggering. Also allows depiction of solid organs (see Fig. 46-14). Analysis of individual sections is mandatory for visualizing ductal abnormalities or filling defects (visible as areas of decreased signal intensity) that may be missed with thick-slab technique.85,162,163

for very mild forms, and detect pancreatic necrosis and local complications; it can also guide percutaneous fine-needle aspiration or catheter drainage.8,28 Mild acute (interstitial edematous) pancreatitis.  Mild acute pancreatitis, also called edematous or interstitial pancreatitis, is a selflimited disease with no or minimal organ dysfunction, no complications, a rapid response to appropriate conservative medical therapy, and prompt resolution of clinical manifestations and laboratory abnormalities, with an uneventful recovery.8 Pathologically there is predominant interstitial edema and microscopic acinar cell necrosis without evidence of macroscopic necrosis; intra- and peripancreatic adipose tissue necrosis is also commonly seen.8 No radiographic evaluation is necessary if mild acute pancreatitis is suspected and the clinical symptoms improve. However, imaging investigations are usually needed to detect causes of pancreatitis such as choledocholithiasis.115 US is a sensitive test for identifying gallstones, which may not be seen on CT. MRCP, though very sensitive, is expensive and may not be feasible in emergency situations.14 Failure to attain a clinical response after 48 to 72 hours of conservative treatment of acute pancreatitis indicates the need for CT to confirm the initial diagnosis, assess the severity of the attack, and depict any complications.8 CT reveals a normal or minimally enlarged gland with low or heterogeneous glandular attenuation due to interstitial edema and a shaggy contour. Peripancreatic fat may be normal or hazy due to inflammation. Diffuse involvement of the gland is common; however, segmental forms involving only the head of the pancreas are occasionally seen. After contrast administration, there is uniform enhancement of the gland due to an intact capillary network with vasodilation7-9 (Figs. 46-29 and 46-30). Heterogeneous low-attenuation areas in the peripancreatic space at the onset of an acute episode of pancreatitis represent a combination of fat necrosis, extravasated pancreatic fluid, nonspecific inflammation, and hemorrhage. This was formerly called phlegmon; however, the International Symposium on Acute Pancreatitis elected to abandon this term because it has been used ambiguously and erroneously.29 Extravasated fluid can be differentiated from fat necrosis by the presence of “fat islands”—intact uninvolved areas of fat within the edematous fluid.96

Secretin-Enhanced MRCP Thick-slab MRCP performed in coronal plane before and every 15-30 sec after IV administration of secretin (0.2 µg/kg body weight) for 10-15 minutes. Provides better delineation of full length of pancreatic duct (see Fig. 46-28), enhances detection of anatomic variants (e.g., pancreas divisum), reduces frequency of false-positive findings of duct strictures, assesses sphincter of Oddi function, and depicts progressive filling of duodenum with pancreatic fluid, indirectly assessing pancreatic exocrine reserve. Main limitation to its use is additional cost.61,89,143,158,163-165

Severe acute (necrotizing) pancreatitis.  Severe pancreatitis occurs in 20% to 30% of all patients with acute pancreatitis and is characterized by a protracted clinical course, multiorgan failure, and pancreatic necrosis.28 Pathologically there is severe inflammation with patchy or confluent areas of acinar cell necrosis and necrosis of vessel walls and ducts, with subsequent duct disruption and foci of hemorrhage. The necrosis can be diffuse or patchy, superficial or deep, and can affect any part of the pancreas.7 Based on the extension of necrosis, the revised Atlanta classification system distinguishes three types of acute necrotizing pancreatitis: pancreatic parenchymal necrosis alone, peripancreatic necrosis alone, and parenchymal and peripancreatic necrosis.251 Necrosis occurs early, within the first 24 to 48 hours; with few exceptions, it remains stable.19 An objective assessment of the severity of acute pancreatitis is based on numeric systems that combine clinical and laboratory parameters and contrast-enhanced CT findings. Numeric systems (APACHE II, Ranson) are commonly used to detect organ failure as indirect evidence of disease severity, with a sensitivity of about 70%.7 CT results are better prognostic indicators than numeric scoring systems owing to their greater sensitivity and specificity.16 When the clinical diagnosis of pancreatitis is in doubt, an initial noncontrast CT study is used to confirm the clinical suspicion or detect alternative acute abdominal conditions that can mimic acute pancreatitis. Noncontrast CT’s main drawback, however, is its inability to reliably depict pancreatic necrosis and consequently further define the risk of complications. Therefore more reliable results are obtained with bolus contrast-enhanced CT performed 48 to 72 hours after the onset of an acute attack of severe pancreatitis7 to visualize the pancreas, differentiate the gland from adjacent fluid collections and peripancreatic inflammatory tissue, and demonstrate pancreatic necrosis8 (Fig. 46-31). Focal or diffuse areas of decreased or no enhancement on dynamic contrast-enhanced CT represent areas of ischemia or necrosis. CT has an overall accuracy of 87%, with a sensitivity and specificity of 100% for detection of extended pancreatic necrosis and a sensitivity of 50% if only minor necrotic areas are present. CT scans obtained 3 days after the clinical onset of the attack are more accurate in depicting necrotizing pancreatitis and in discriminating normal variants or equivocal zones of ischemia from pancreatic necrosis.7

1426

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 46-29  Acute pancreatitis. A, Axial contrast-enhanced MDCT reveals a diffusely enlarged enhancing pancreas with an acute inflammatory fluid collection in the peripancreatic region (arrowheads). Fat islands (arrows) are seen in the extravasated inflammatory fluid, differentiating it from fat necrosis. B, Follow-up MDCT after conservative management shows improvement in the pancreatic swelling, with organization of the peripancreatic inflammation. C, Further follow-up reveals significant reduction of the pancreatic swelling, with the walled-off peripancreatic collection evolving into a pseudocyst (arrows). The peripancreatic inflammation adjacent to the tail of the pancreas has resolved (arrowhead).

The severity index (Table 46-9) is a scoring system that combines CT grading and the percentage of necrosis to obtain a number that correlates with the risk of developing serious local complications such as abscess or infected necrosis, which in turn correlates with increased morbidity and risk of death.7,254 In one study, patients with a severity index of 0 or 1 exhibited no mortality and no morbidity, patients with a severity index of 2 had no mortality and a 4% morbidity rate, and those with a severity index of 7 to 10 had a 17% mortality rate and a 92% complication rate.10 Also, it has been suggested that all heterogeneous peripancreatic collections should be considered areas of fat necrosis until proved otherwise. Infection cannot be excluded in these patchy areas of fat necrosis on the basis of imaging alone.7,53 Grades A to C (CT severity index 0-2) do not need follow-up unless there are changes in the patient’s clinical status suggesting a developing complication. Grades D or E (CT severity index 3-10) should be followed up 7 to 10 days after the initial scan to reevaluate CT findings and detect late complications such as arterial pseudoaneurysm and pseudocyst8 (Figs. 46-32 and 46-33). The modified CT severity index (Table 46-10) incorporates features reflecting organ failure and

extrapancreatic complications for predicting course. This index includes the presence or absence of acute fluid collections rather than the count of collections; it scores necrosis as absent, minimal (≤30%), or substantial (≥30%); and it takes into consideration extrapancreatic findings such as pleural fluid, ascites, extrapancreatic parenchymal abnormalities, vascular complications, or involvement of the gastrointestinal (GI) tract. The modified index is found to have a stronger prognostic correlation than the accepted index and could also predict the length of hospital stay and development of organ failure.

Clinical and CT presentation

Acute fluid collections.  Acute collections of enzyme-rich pancreatic juice occur in about 40% of patients early in the course of acute pancreatitis.138 These fluid collections may be around the gland or intrapancreatic. They lack a capsule and are confined only by the anatomic space within which they arise, most commonly the anterior pararenal space or lesser sac.234 They can dissect into other locations, including the mediastinum and the posterior pararenal space, and  can involve solid organs (liver, spleen, kidneys) or the wall of an adjacent bowel loop.8 Acute fluid collections appear hypodense on CT; they are poorly defined with no recognizable capsule or wall, which

CHAPTER 46  Pancreas

A

C

E

B

D

F

1427

FIG 46-30  Acute pancreatitis. Axial contrast-enhanced MDCT (A, C, and E) and corresponding coronal MDCT (B, D, and F). A, Inflammatory exudate is seen around the head and tail of the pancreas in the perirenal and perihepatic regions (arrows). B, Corresponding coronal image reveals the inflammation extending to the peripancreatic, perihepatic, lumbar, and iliac regions. C, Follow-up MDCT after conservative management shows improvement in the inflammation, which is well appreciated on the corresponding coronal image (D). Near-complete resolution of inflammation seen on the axial (E) and corresponding coronal (F) images.

FIG 46-31  Severe pancreatitis. Axial contrast-enhanced MDCT reveals ill-defined inflammation around the pancreas, especially in the region of the tail (arrow), with a fluid collection in the perihepatic region. The enhancing pancreas is partially visualized in the region of the head and body, with a focal nonenhancing area in the head of the pancreas (arrowhead), representing necrosis.

distinguishes them from pseudocysts.28 These collections resolve spontaneously in about 50% of cases (see Fig. 46-30). Those that do not resolve may evolve into pseudocysts (see Fig. 46-29) or may be associated with a variety of complications including pain, secondary infection, and hemorrhage.28 Fluid collections, whether encapsulated or not, are amenable to percutaneous catheter drainage if they do not resolve spontaneously or if they become infected.8 Pseudocyst.  Pseudocysts are encapsulated unilocular collections of pancreatic fluid and necrotic and proteinaceous material. They may evolve from acute fluid collections by the development of a nonepithelialized wall or capsule, or they may occur in patients with chronic pancreatitis during the course of their disease234,276 (Figs. 46-34 to 46-36). It takes about 4 weeks or longer for a pseudocyst to evolve from acute pancreatic fluid collections. Pseudocysts most often are peripancreatic but can be found throughout the abdomen as well as within the mediastinum and pelvis (Fig. 46-37). Symptoms associated with pseudocysts depend on the amount and severity of pancreatitis and the size and speed of their growth. Large pseudocysts can cause pain secondary to the inflammatory process, displacement of adjacent structures,  or pressure within the pseudocyst.96 About half of all pseudocysts that arise in patients with acute pancreatitis resolve spontaneously  by means of resorption of fluid or drainage in a bowel loop. The

TABLE 46-9  CT Severity Index CT Grade A: Normal pancreas B: Focal or diffuse pancreatic enlargement C: Inflammation—pancreas and/or peripancreatic fat D: Single ill-defined peripancreatic fluid collection E: Two or more ill-defined peripancreatic fluid collections

(Points)

Percentage Necrosis

(Points)

Severity Index = CT Grade + Percentage Necrosis (Points)

0 1 2 3 4

0 0 <30% 30-50% >50%

0 0 2 4 6

0 1 4 7 10

TABLE 46-10  Modified CT Severity Index

(Points)

Percentage Necrosis

(Points)

Normal pancreas

0

0

0

Inflammation—pancreas and/or peripancreatic fat Pancreatic or peripancreatic fluid collection or peripancreatic fat necrosis

2

≤30%

2

Pleural effusion, ascites, vascular complications, extrapancreatic parenchymal abnormalities, or GI tract involvement Moderate (4-6)

4

>30%

4

Severe (8-10)

CT Grade

Extrapancreatic Complications

(Points) 2

Modified Severity Index = CT Grade + Percentage Necrosis Extrapancreatic Complications (Points) Mild (0-2)

From Mortele KJ et al: A modified CT severity index for evaluating acute pancreatitis: Improved correlation with patient outcome. AJR Am J Roentgenol 183:1261–1265, 2004.

A

B

FIG 46-32  Severe pancreatitis. A, Axial contrast-enhanced

C

MDCT reveals a swollen pancreas that is replaced by inflammatory exudate in the region of the head and body (arrows). B, Axial image reveals a hypodense filling defect (arrow) in the enhancing portal vein. C, Corresponding coronal image showing this defect (arrow), representing portal vein thrombosis.

CHAPTER 46  Pancreas

D

A

B FIG 46-33  A, Axial contrast-enhanced CT reveals an enhancing focus in the pancreatic head (straight white arrow) located within the inflammatory mass (curved arrows). The pancreatic duct (D) is severely dilated. B, Conventional angiogram shows “pooling” of contrast material in a pseudoaneurysm (arrows).

A

B

C FIG 46-34  Evolution of pseudocyst in acute pancreatitis. A, Diffuse swelling of the pancreas (arrows) with a peripancreatic fluid collection in a case of acute pancreatitis visualized on axial contrast-enhanced MDCT. B, Follow-up MDCT reveals a hypoattenuating fluid collection in the region of the body of the pancreas, with loculation of the peripancreatic collection (arrows). C, There is a well-defined loculated hypoattenuating collection in the pancreatic body (arrows), with resolution of the peripancreatic collection. A nasogastric tube is seen in situ (arrowhead).

1429

1430

A

PART II  CT and MR Imaging of the Whole Body

B

C FIG 46-35  Evolution of pseudocyst in acute pancreatitis. A, Axial contrast-enhanced MDCT reveals focal acute pancreatitis in the region of the pancreatic tail, with swelling and a peripancreatic fluid collection and thickening of the perirenal fascia. B, Follow-up MDCT reveals an increase in peripancreatic inflammation. C, Further follow-up shows containment of the peripancreatic fluid, with development of a pseudocyst (arrows).

A

B FIG 46-36  Pseudocyst in chronic pancreatitis. A, Axial contrast-enhanced MDCT reveals a well-defined thick-walled low-attenuation pseudocyst in the region of the head of the pancreas (arrows). A fluid level is seen in the dependent position (horizontal arrow). B, Atrophy of the body and tail of the pancreas, with a prominent pancreatic duct due to chronic pancreatitis (arrows).

CHAPTER 46  Pancreas

A

1431

B FIG 46-37  A, Axial contrast-enhanced MDCT reveals a large pseudocyst (arrowheads) almost replacing the entire pancreas, with compression of the residual pancreatic parenchyma. B, Corresponding coronal image reveals a pseudocyst in the region of the pancreas (arrowheads) and in the mesentery in periumbilical, lumbar, and iliac regions (arrows), at sites distant from the pancreas.

remaining half may stabilize, partially resolve, or enlarge and cause complications, necessitating intervention.8 On CT, a pseudocyst appears as a round or oval fluid collection with a thin or a thick wall that shows contrast enhancement. MRI reveals a well-defined unilocular lesion that is hypointense on T1-weighted images and hyperintense on T2-weighted images (Fig. 46-38). Gas bubbles within the pseudocyst may be due to infection, fistula formation, or internal cystotomy, and percutaneous fine-needle aspiration is required for a specific diagnosis. Venous stenosis or occlusion with the formation of varices and arterial pseudoaneurysm may be seen. Acute hemorrhage due to enzymatic digestion of vessels or pseudocyst rupture is seen as high-attenuation material within the cyst (Fig. 46-39). Occasionally a pseudoaneurysm may be confused with a pseudocyst. A dynamic CT scan with IV contrast enhancement and careful scrutiny to identify a clot within the aneurysm are essential.26,233 Other complications such as biliary and GI tract obstruction, invasion into the spleen or liver, rupture into the peritoneum, and perforation into the GI tract can also be identified with CT.8,219 Pseudocysts smaller than 5 cm in diameter are likely to resolve spontaneously and should be monitored.276 Those larger than 5 cm, increasing in size, causing severe pain, or resulting in GI or biliary tract obstruction should be considered for drainage256 (Fig. 46-40). Fistula formation with the stomach or duodenum results in mechanical drainage, making surgery unnecessary. Fistula formation with colon, however, can result in bacterial colonization and infection and thus requires immediate débridement and bowel diversion26,96 (Fig. 46-41). Pancreatic abscess.  Pancreatic abscesses are circumscribed intraabdominal collections of pus located near the pancreas. They usually develop 4 weeks or more after the onset of acute pancreatitis and probably develop as a complication of limited necrosis with subsequent liquefaction and secondary infection.24 The source of infection can be hematogenous or lymphatic, due to GI fistula or perforation, or iatrogenic.101 The CT diagnosis of pancreatic abscess is based on the presence of a focal low-attenuation collection with a relatively thick

FIG 46-38  Axial T2-weighted fat-suppressed MRI shows a well-defined unilocular hyperintense lesion in the region of the body of the pancreas, representing a pseudocyst (arrows).

wall that often contains gas bubbles (Fig. 46-42). Gas bubbles are not specific for infection, and gas may not always be present. Hence the final diagnosis depends on correlation with the clinical condition  of the patient and confirmation with percutaneous fine-needle aspiration.69,213 Acute necrosis collection.  In necrotizing pancreatitis, a collection without an enhancing capsule within the first 4 weeks is called an acute necrotizing collection, and a collection with an enhancing capsule after 4 weeks is called walled-off necrosis. These necrotic pancreatic or peripancreatic collections can become infected.29,251 The process may be focal and minimal or widespread and severe. It can develop early or late and is often recognized on CT scans as bubbles of gas or air pockets within areas of pancreatic or peripancreatic necrosis (emphysematous

1432

PART II  CT and MR Imaging of the Whole Body

*

A

*

B FIG 46-39  A, Axial contrast-enhanced MDCT reveals a well-defined walled-off pseudocyst in the region of the pancreatic body. The pseudocyst is hypodense, with central hyperdense material representing hemorrhage with organization (asterisk). B, T1-weighted fat-suppressed contrast-enhanced MRI in another patient reveals a hypointense pseudocyst (arrows), with central focal hyperintensity representing organized hemorrhage (asterisk).

A

pancreatitis) (Fig. 46-43). CT is sensitive for detecting even the smallest amount of gas. However, when the pancreatic or peripancreatic necrotic collection does not contain gas, the infection cannot be diagnosed on CT unless percutaneous aspiration is performed.195 The outcome of pancreatitis depends not only on the severity of inflammation but also on the presence of infection.3 It is important to distinguish abscess from infected necrosis, because the mortality rate for the latter is nearly double that of the former and the specific therapy for each condition is different. On CT, an abscess is diagnosed when a normally enhancing pancreas is seen with an adjacent fluid collection composed of liquid pus (determined by needle aspiration). Infected necrosis is diagnosed when a zone of nonenhancing heterogeneous pancreas is seen and liquefied infected tissue is detected at needle aspiration. Needle aspiration is crucial because the CT appearance of a low-attenuation zone of infected necrosis may be similar to that of an abscess. Also, it is essential that the diagnostic aspiration be precise, with no contamination from intestinal material. Passage through the bowel may lead to a falsepositive result and introduce infection. CT can be a useful guide to precisely locate the areas of interest during aspiration. Also, it is important to perform contrast-enhanced CT to demonstrate any pseudo­ aneurysms. Abscesses can be treated effectively with percutaneous catheter drainage, whereas infected necrosis requires surgical necrosectomy and débridement (Fig. 46-44). Percutaneous drainage should  be attempted only when the necrotic tissue is totally liquefied3,8,24,96 (Fig. 46-45). Hemorrhage.  Hemorrhage in acute pancreatitis usually occurs as a late consequence due to either diffuse leakage from the inflamed granulation tissue or vascular injuries produced by the activated and extravasated pancreatic enzymes.78 The splenic artery and its branches or the pancreaticoduodenal arcade arteries are commonly affected. Bleeding is commonly preceded by development of an arterial pseudoaneurysm that ruptures, leading to sudden massive hemorrhage. CT usually shows high-attenuation fluid (blood) within the peritoneal cavity or retroperitoneum or within a preexisting fluid collection or pseudocyst. Emergency angiography with selective embolization of the bleeding vessel is the treatment of choice.263

B FIG 46-40  Cystogastrostomy for pseudocyst. A, Axial CT shows a fluid- and air-filled collection behind the stomach (arrow). B, CT after oral administration of contrast material reveals high-density contrast material in the stomach and within the collection behind the stomach (arrow) due to a surgically created communication between the two. Note the high-density suture line between the stomach and the cyst.

CHAPTER 46  Pancreas

1433

Pseudocyst

Larger than 5 cm

Smaller than 5 cm

Resolves spontaneously Should be monitored

Does not resolve

Observe for 6 wk • May resolve • Inflammation and lack of well-defined fibrous wall hinder adequate surgical intervention

Complications before 6 wk, mandating intervention

Cystotomy (6–13 wk) Beyond 13 wk, risk of spontaneous complications

Percutaneous drainage High rate of infection, fistula, and complications

FIG 46-41  Management approach to pseudocyst. (From Bodurtha AJ et al: Analysis of 29 cases of pancreatic pseudocyst treated surgically. Can J Surg 23:432–434, 1980.)

B

A

FIG 46-42  Pancreatic abscess. A, Axial contrast-enhanced MDCT reveals a thick-walled peripancreatic fluid collection around the body and tail, with gas bubbles (arrows). B, Follow-up MDCT after placing a drain in the infected collection (infection was confirmed on aspiration) reveals partial resolution.

Chronic Pancreatitis.  Chronic pancreatitis is a progressive fibroinflammatory disorder characterized by intermittent or continuous abdominal or back pain (or both) due to the persistence of structural damage after the primary cause has been eliminated. This damage results in loss of pancreatic parenchyma, functional insufficiency (endocrine and exocrine), and complications such as biliary stricture, pseudocyst, and pseudoaneurysm. Three principal forms of chronic pancreatitis are currently recognized39: 1. Calcifying chronic pancreatitis: characterized by acinar destruction and perilobular fibrosis with acute and chronic inflammatory cells. It presents with recurrent bouts of abdominal pain and eventual development of intraductal calculi in a large proportion of cases. Causative factors include alcohol and tobacco use. There are hereditary, tropical, idiopathic, and senile forms; the senile form is often painless.65,239 2. Obstructive chronic pancreatitis: persistent obstruction of the pancreatic duct due to tumor or postinflammatory ductal stricture leads to atrophy of the upstream pancreas. Though often

painless, it occasionally presents with clinically acute pancreatitis. Intraductal calculi are generally not seen.133 3. Autoimmune pancreatitis: chronic systemic lymphoplasmacytic inflammatory process involving the pancreas and other organs. Typically, chronic pancreatitis develops in patients with recurrent bouts of acute pancreatitis (e.g., alcoholic and hereditary forms of calcifying pancreatitis); however, it can also develop without clinical pancreatitis (e.g., obstructive and autoimmune forms). Thus in addition to repeated bouts of pancreatitis, other unidentified risk factors are required for development of fibrosis and formation of intraductal calculi.38 Irregular ductal dilatation and strictures, parenchymal atrophy, and pancreatic calcifications are typical CT manifestations of chronic pancreatitis. Pancreatic ductal dilatation, though a frequent manifestation of chronic pancreatitis, is not specific; it can also be seen with pancreatic and ampullary carcinomas.154 A smooth dilated duct with a duct width–to–total gland width ratio greater than 0.5 is suggestive of  carcinoma.120 In pancreatitis the dilated pancreatic duct is associated

1434

PART II  CT and MR Imaging of the Whole Body Gas in pancreatic or peripancreatic tissue on CT

P

Symptomatic

Asymptomatic

Early surgical or percutaneous drainage and antibiotics

Percutaneous aspiration with culture

Positive

Clinical deterioration

Negative

Observe, regular CT exam till gas disappears

FIG 46-45  Management approach to emphysematous pancreatitis.

FIG 46-43  Emphysematous pancreatitis (pancreatic necrosis). Axial CT scan shows a fluid collection around the pancreas (P), with air pockets (arrows) in the fluid collection and the gland.

A

B FIG 46-44  A, Axial contrast-enhanced MDCT shows a walled-off inflammatory peripancreatic collection (arrows) in a case of acute pancreatitis. Analysis of needle aspirate was suggestive of infection.  B, Follow-up imaging after surgical necrosectomy and débridement were performed.

with irregularity owing to dilated side branches (Figs. 46-46 to 46-48). Focal or diffuse pancreatic atrophy, a secondary manifestation of chronic pancreatitis, may be underestimated by CT because interlobular and periductal fibrous connective tissue and chronic inflammatory infiltrate can replace functioning parenchyma, resulting in little or no loss of total pancreatic mass185 (Fig. 46-49). Pancreatic calculi or calcifications are the most specific CT manifestations of chronic pancreatitis and are not found in association with neoplastic obstruction. Chronic inflammation results in local chemical changes, with deposition of calcium phosphate and carbonate. Pancreatic calcifications are almost always within the ductal system, although this intraductal location may not be evident on CT (Figs. 46-50 to 46-52). They develop late in the course of the disease and are often preceded by relatively radiolucent protein plugs that are detectable with ERCP but not with CT.154 Biliary ductal dilatation at the level of the pancreatic head is a nonspecific finding and can occur with distal common bile duct calculi, pancreatic carcinoma, cholangiocarcinoma, and ampullary carcinoma in addition to chronic pancreatitis. However, in chronic pancreatitis, common bile duct stenosis tends to be longer and more gradually tapered than in malignant obstructions.204 When present, mature pseudocysts with well-defined enhancing walls can be well appreciated on CT. The extensive lobular and periductal inflammation and fibrosis may result in the formation of benign inflammatory pancreatic masses. Although the clinical history, calcification within the mass, ductal calculi, and dilated ductal side branches can help differentiate benign from malignant mass lesions, this may not always be the case, because acute and chronic obstructive pancreatitis may antedate demonstrable carcinoma.68,87,177,185 In many cases the masses remain indeterminate despite the addition of US or ERCP to the diagnostic workup. Furthermore, percutaneous biopsy is helpful only when results are positive for malignancy. If percutaneous needle biopsy fails to demonstrate malignancy, the choice is between watchful waiting and a thorough biopsy.185 PET CT can help differentiate focal masslike pancreatitis from pancreatic cancer; cancer shows higher uptake of FDG.139 Bland pancreatic fibrosis without inflammation or parenchymal destruction can occur with aging, alcoholism, and smoking. It is clinically silent and should be distinguished from chronic pancreatitis.239 Autoimmune pancreatitis.  Autoimmune pancreatitis (AIP) is characterized by marked fibrosis and lymphoplasmacytic infiltration

CHAPTER 46  Pancreas

A

B

C

D

1435

FIG 46-46  Chronic pancreatitis. A, Axial contrast-enhanced MDCT reveals atrophied body and tail of the pancreas and a dilated pancreatic duct with intraductal (arrows) and parenchymal calcifications. B, ERCP reveals a prominent pancreatic duct with dilatation and irregularity of the side branches (arrow). Axial (C) and coronal (D) T2-weighted MRIs reveal the dilated and irregular pancreatic duct with prominent side branches (arrow), which was not appreciated on CT.

FIG 46-47  Axial curved reformatted MDCT reveals a low-attenuation nonenhancing mass lesion in the head and uncinate process of the pancreas (arrowheads), with upstream atrophy of the pancreas and smooth dilatation of the pancreatic duct, which terminates abruptly (arrow) at the mass lesion.

FIG 46-48  Axial curved reformatted MDCT reveals a high-density stent in the pancreatic duct. The stent was inserted (with ERCP guidance) to relieve a pancreatic duct stricture that developed as a result of chronic pancreatitis. Note the atrophy of the body and tail of the pancreas, dilatation of the pancreatic duct, and a small calcified focus. Air bubbles are seen in the pancreatic duct due to stent insertion.

1436

PART II  CT and MR Imaging of the Whole Body

FIG 46-49  Axial delayed contrast-enhanced MDCT reveals atrophy of the pancreas with mild prominence of the pancreatic duct (arrows) in the distal body and tail, secondary to chronic pancreatitis.

A

of the pancreas and increased serum immunoglobulin (Ig)G, especially the IgG4 subtype or positive autoantibodies.194 AIP is associated with other autoimmune disorders such as Sjögren’s syndrome, primary sclerosing cholangitis, primary biliary cirrhosis, ulcerative colitis, and systemic lupus erythematosus.194 AIP is classified into two types (1 and 2) with some overlap in clinical and histopathologic characteristics but also showing important differences54 (Table 46-11). Although differentiating the two types of AIP could be useful to predict the likelihood of disease recurrence, types 1 and 2 cannot be reliably distinguished by imaging.54,281 AIP can be diagnosed on the basis of criteria proposed by the Japan  Pancreas Society, which includes imaging, serologic, and histopathologic criteria73,118,167,262,280 (Table 46-12). The classic CT appearance of AIP is diffuse sausage-shaped enlargement of the pancreas, with homogeneous attenuation that after administration of contrast media shows reduced enhancement on the earlier phase and increased or prolonged enhancement in the delayed phase. Moreover, in contrast phases, AIP demonstrates a low-density rim surrounding the pancreas, also known as the halo sign.108

B

C FIG 46-50  Chronic calcifying pancreatitis. Axial contrast-enhanced MDCT through the pancreas reveals parenchymal calcification (arrow) in the region of the head (A), with intraductal calculi (arrow) in the region of the body (B) and tail (C). The pancreas is atrophied, with dilatation of the pancreatic duct.

CHAPTER 46  Pancreas

1437

Loss of lobularity is common; peripancreatic fat stranding is usually minimal (Fig. 46-53). As the disease progresses, involution or retraction of the pancreatic tail is evident. Diffuse or irregular attenuation of the pancreatic duct and compression or narrowing of the distal common bile duct due to pancreatic swelling are seen. Mild enlargement of the regional lymph nodes is also common. Focal involvement is encountered more often in the head of the pancreas than in the body and tail; it typically appears as a low-attenuation or isoattenuation mass (Fig. 46-54). Differentiating the focal form of AIP from pancreatic carcinoma can be difficult on the basis of CT alone; hence most of these patients are treated surgically. However, associated changes in the remainder of the gland and diffuse pancreatic ductal narrowing are highly diagnostic of AIP. Extrapancreatic manifestations include focal lesions in the lungs, kidneys, liver, or soft tissue around the aorta, described as inflammatory pseudotumors. Response to corticosteroids appears on CT as resolution of the pancreatic enlargement and improvement or resolution of the peripancreatic halo and stranding. Likewise, pancreatic and biliary narrowing may resolve partially or completely. Morphologic resolution becomes evident as early as 1 to 2 weeks after initiation of therapy.73,223 In some cases, however, narrowing of the pancreatic duct and common bile duct persists despite corticosteroid treatment. Chronic changes of

A

TABLE 46-11  Characteristics of AIP Types

1 and 2 Type 1

B Type 2

FIG 46-51  A, Axial contrast-enhanced MDCT reveals intraductal calculi (long arrow) with foci of parenchymal calcification (short arrow) in the region of the head of the pancreas in a case of chronic pancreatitis. B, Intraductal calculi (arrow) are seen on an axial T2-weighted fatsuppressed MRI; however, the parenchymal calcification is not well appreciated.

A

Serologic abnormalities (increased IgG4), extrapancreatic organ involvement, and disease relapse Clinically presents like nonobstructing jaundice in an elderly man Pathologically characterized by high cellularity infiltrate Confined to pancreas; no serologic abnormalities or systemic involvement Clinically presents with pain and symptoms similar to chronic pancreatitis Pathologically characterized by an evident ductal injury

B FIG 46-52  A, Axial contrast-enhanced MDCT reveals pancreatic atrophy, dilatation and irregularity of the pancreatic duct, and intraductal calculi (arrows) in a case of chronic pancreatitis. B, Corresponding ERCP reveals dilatation of the pancreatic duct with intraductal filling defects (arrows), representing calculi.

1438

PART II  CT and MR Imaging of the Whole Body

TABLE 46-12  Diagnostic Criteria for Autoimmune Pancreatitis* IMAGING FINDINGS (ONE REQUIRED) Cross-sectional Imaging

SEROLOGY AND HISTOLOGY FINDINGS (ONE REQUIRED) Pancreatic-Biliary Histologic Analysis

Non-GI Histologic Analysis

Elevated serum IgG4 level

Periductal lymphoplasmacytic infiltration or fibrosis

Focal pancreatic ductal narrowing

Elevated serum IgG or IgG4 level

Obliterative phlebitis

Diffuse pancreatic ductal narrowing

Presence of ALA, ACA II, ASMA, or ANA

IgG4-positive plasma cells in tissue†

Tubulointerstitial nephritis with immune deposits within tubular basement membranes Pulmonary interstitial lymphoplasmacytic infiltration with IgG4-positive plasma cells† Chronic sialoadenitis with IgG4-positive plasma cells†

ERCP or MRCP

Serologic Analysis

Diffusely enlarged pancreas

Segmental pancreatic ductal narrowing

Enhanced peripheral rim of hypoattenuation: “halo” Low-attenuation mass in head of pancreas

*Criteria modified from those of the Japan Pancreas Society.118 The presence of tissue IgG4-positive cells is not necessarily abnormal, but an increased number of infiltrating IgG4-positive plasma cells is abnormal. ACA II, anti–carbonic anhydrase II antibody; ALA, antilactoferrin antibody; ANA, antinuclear antibody; ASMA, anti–smooth muscle antibody; ERCP, endoscopic retrograde cholangiopancreatography; IgG, immunoglobulin G; MRCP, magnetic resonance cholangiopancreatography.

†

A

B

C

D FIG 46-53  Autoimmune pancreatitis. A, Axial contrast-enhanced MDCT reveals a diffusely swollen pancreas with a peripancreatic halo (arrows), minimal peripancreatic stranding, and attenuation of the pancreatic duct. B, Corresponding T2-weighted fat-suppressed MRI reveals mild hyperintensity of the pancreas, with minimal peripancreatic stranding. C, ERCP reveals multiple areas of irregularity in the pancreatic duct (arrows). D, Follow-up MDCT 3 months after corticosteroid treatment reveals resolution of the pancreatic swelling and peripancreatic halo, with normalization of the pancreatic parenchyma and pancreatic duct (arrow).

CHAPTER 46  Pancreas

A B

D C

E FIG 46-54  Autoimmune pancreatitis. A, Axial contrast-enhanced MDCT reveals a diffusely swollen pancreas (arrows) in a patient who presented with abdominal pain and jaundice. B, ERCP reveals dilatation of the common bile duct due to compression of the distal duct by the swelling in the pancreatic head. C, A stent was inserted under ERCP guidance to relieve the obstructive symptoms, and a course of corticosteroids was instituted. D, The diffuse swelling in the pancreas resolved with corticosteroid therapy; however, follow-up MDCT after 1 month reveals the development of a focal masslike lesion (arrows) in the head of the pancreas. The high-density stent is also seen in situ (horizontal arrow). The masslike lesion did not respond to continued corticosteroid therapy and was surgically removed to relieve the obstruction of the common bile duct; histopathology revealed chronic fibrosis and lymphoplasmacytic infiltration, characteristic of autoimmune pancreatitis. E, Delayed follow-up MDCT 4 months later reveals pancreatic atrophy.

1439

1440

PART II  CT and MR Imaging of the Whole Body

BOX 46-1  API Findings That Help

Differentiate from Pancreatic Cancer Delayed enhancement of affected pancreas Restricted diffusion and low ADC values on MRI Multiple MPD narrowings Long segment (>3 cm) MPD narrowing MPD dilatation not exceeding 4-5 mm Penetrating duct sign (dilated MPD penetrating the focal enlargement without disappearing) Halo sign Extrapancreatic findings Improvement of findings after corticosteroid therapy From Lee LK, Sahani DV: Autoimmune pancreatitis in the context of IgG4-related disease: Review of imaging findings. World J Gastroenterol 20:15177–15189, 2014. MPD, main pancreatic duct.

fibrosis and sclerosis are also responsible for pancreatic atrophy, which may be a long-term sequela in some cases (see Fig. 46-54). Findings on MR and MRCP in patients with typical AIP include a diffusely enlarged gland. As on CT, a peripheral rim of the hypointense halo can be seen on MRI. In addition, diffuse attenuation of the main pancreatic duct or a small strictured main pancreatic duct may be seen on MRCP.82 In patients with focal masslike lesions, fine-needle aspiration of the mass may be helpful. Because pathologic confirmation of AIP is difficult because of the patchy nature of the disease, corticosteroid therapy can be used as a diagnostic tool in patients whose clinical and laboratory findings are strongly suggestive of the disease. In these patients, short-interval imaging at 2 to 4 weeks after the initiation of therapy must be used to verify resolution of the mass145 (Box 46-1). However, because malignant masses may also respond to corticosteroids, complete resolution of the pancreatic mass is critical for the confirmation of AIP. If the clinical history and laboratory and pathologic findings do not support the diagnosis of AIP, or if corticosteroid therapy fails to resolve the pancreatic mass, surgical biopsy of the mass should be considered.73,86 Hereditary pancreatitis.  Hereditary pancreatitis is an autosomal dominant relapsing pancreatitis with an estimated 80% penetrance. It generally manifests during childhood, with a peak incidence at age 5 years. However, a second peak at age 17 may be attributable to the introduction of alcohol in the diet. In addition, factors such as hyperlipidemia and hypercalcemia may play a role. The symptoms, signs, and pathophysiology of this type of pancreatitis are no different from those of the other types of pancreatitis discussed. Parenchymal and intraductal calcifications occur in approximately 50% of patients. Intraductal calculi occur early in the course of the disease; they tend to be large and rounded and are arranged in a linear pattern in the dilated main pancreatic duct. Other features of pancreatitis and pseudocysts may be seen. Pancreatectomy is the treatment of choice.58,121,147

MRI in Pancreatitis.  MRI has a very limited role in acute settings and is typically used as a problem-solving tool to answer a specific clinical question. In acute pancreatitis, fat-suppressed T2-weighted images are helpful for defining subtle diffuse or focal parenchymal abnormalities. T2-weighted images can accurately depict fluid collections, pseudocysts, and areas of hemorrhage. MRI can assess the internal consistency and drainability of fluid collections, thus influencing the choice of treatment.179 Gd-enhanced T1-weighted GRE MRI can

depict pancreatic necrosis as areas of nonenhanced parenchyma.7 MRCP has the advantage of demonstrating possible choledocholithiasis, the presence or absence of ductal disruption or leakage, and the size, location, and possible communication of a pseudocyst with the pancreatic duct, thus making it useful for therapeutic planning.163 In patients with recurrent pancreatitis, MRCP can demonstrate causative factors such as pancreas divisum and annular pancreas, especially if performed with secretin stimulation, without the risk of post-ERCP pancreatitis.164 In early stages of chronic pancreatitis, increased signal intensity on T2-weighted MRI is due to fatty and fibrous replacement, inflammatory cell infiltration of the pancreatic parenchyma, and impaired pancreatic juice outflow.49 Pre- and post-Gd fat-suppressed T1-weighted images in chronic pancreatitis reveal decreased enhancement in the arterial phase and increased enhancement in the early venous phase. The signal intensity ratio (postcontrast divided by precontrast signal intensity) in the arterial phase is significantly lower (<1.7) than that in the early venous phase. These changes in enhancement in chronic pancreatitis are related to decreased pancreatic blood flow and delayed washout. Thus measurement of the pancreatic signal intensity on Gd-enhanced dynamic MRI may be helpful for the diagnosis of earlystage chronic pancreatitis, especially before the appearance of pancreatic morphologic changes.282 Morphologic features of advanced chronic pancreatitis on MRI include fibrosis, atrophy, focal enlargement of the gland, pancreatic calcification (which may be difficult to identify on MRI), and pseudocysts.198 MRCP offers several advantages over ERCP in the evaluation of chronic pancreatitis; it is noninvasive, avoids the need for contrast media and irradiation, provides projection images in several planes, and can be used to assess exocrine pancreatic function without the associated disadvantages of ERCP, such as the risk of complications and technical difficulties in postoperative patients (Billroth II operation or Roux-en-Y reconstruction).241,243 MRCP well demonstrates dilatation, stricture, and irregularity of the main pancreatic duct, as well as filling defects due to pancreatic stones and protein plugs; it visualizes the pancreatic ducts distal to sites of complete obstruction and pseudocysts, especially noncommunicating ones.241 Stones smaller than 2 mm and those lying within normal-caliber or minimally dilated side branches of the main pancreatic duct may be difficult to identify.166 MRCP is sometimes inadequate for evaluation of mild pancreatitis, which causes ectasia of the side branches, mild ductal dilatation, loss of the normal gentle tapering of the duct toward the tail, and mural irregularities.49,241,243 This may lead to false-negative results. Secretinenhanced MRCP improves delineation of ductal morphologic features and may be helpful in the diagnosis of mild pancreatitis.49,158,241 The benefit of secretin administration is even greater in the case of earlystage chronic pancreatitis. In one study the totality of the main pancreatic duct segments was visualized in 97% of cases by secretin-enhanced MRCP, versus 65% of cases by MRCP alone; dilated side branches were visualized in 63% of cases by secretin-enhanced MRCP, versus 4% by MRCP in patients with findings suggestive of chronic pancreatitis.158 Pancreatic exocrine function can be assessed by qualitative or semiquantitative analysis of duodenal filling with secretin-enhanced MRI. Another technique, secretin-enhanced diffusion-weighted imaging (DWI), evaluates regional or microscopic changes of water in the pancreatic parenchyma and can thus assess exocrine function. The apparent diffusion coefficient (ADC) time curve is created by the acquisition of dynamic MRI before and for 10 minutes after IV administration of secretin. In normal cases, in response to secretin stimulation, bicarbonate and fluid move from the pancreatic cells into the ductal lumen. This movement of water molecules causes molecular diffusion and

CHAPTER 46  Pancreas

1441

B

A

FIG 46-55  A, Normal pancreatic groove (arrow). B, Ill-defined area of low attenuation (white arrow) in the pancreatic groove on coronal reformatted MDCT, suggestive of groove pancreatitis. Note a pancreatic cyst (black arrow) in the uncinate process of the pancreas.

increases the ADC value, which peaks at 2 minutes and decreases thereafter. In patients with alcohol abuse and mild impairment of pancreatic exocrine function, fluid excretion into the pancreatic ducts is delayed, causing a delay in the ADC peak, although absolute values of the basal and peak ADC are normal. Severe pancreatitis shows marked functional impairment, which causes a low basal ADC value and no response to secretin. This technique is more sensitive than secretin-enhanced MRCP and may be useful for the diagnosis of mild or early chronic pancreatitis.49,64,241

Groove Pancreatitis.  This is a segmental form of pancreatitis with inflammation in the groove between the duodenum and the head of the pancreas; the rest of the pancreas enhances normally and there is normal ductal morphology238 (Fig. 46-55). Chronic inflammation and fibrosis may cause duodenal, common bile duct, and distal pancreatic duct stenosis. The clinical and imaging features may resemble periampullary or pancreatic carcinoma.18 Contrast-enhanced CT demonstrates a sheetlike lesion in the pancreaticoduodenal groove, effacing the groove and showing delayed enhancement. Cyst formation in the duodenal wall or pancreaticoduodenal groove has also been described.112 Tissue diagnosis is necessary when groove pancreatitis cannot be differentiated from periampullary or pancreatic carcinoma by imaging alone.274

Pancreatic Tuberculosis.  Tuberculosis of the pancreas is uncommon and usually occurs as a complication of miliary tuberculosis and immunodeficiency, with isolated involvement of the pancreas being exceedingly rare. The diagnosis usually is not suspected before laparotomy unless there is evidence of tuberculosis elsewhere or a relevant clinical history.278 In addition to the constitutional symptoms, pancreatic tuberculosis may present as acute or chronic pancreatitis, portal vein obstruction, pancreatic mass mimicking abscess, carcinoma, obstructive jaundice, GI bleeding, and peripancreatic abscess.47 Focal

involvement of the pancreas most frequently occurs in the pancreatic head, followed by the body and tail; diffuse pancreatic involvement is exceedingly rare.141 CT reveals a focal hypodense lesion often displaying internal densities. On contrast-enhanced CT the well-defined mass may show irregular margins with peripheral enhancement. Areas of central enhancement may give a multiloculated appearance. These features, however, are nonspecific and may resemble inflammatory or neoplastic cystic lesions of the pancreas. Rarely, diffuse enlargement of the pancreas along with hypodense areas may be seen. The latter morphologic abnormalities are also nonspecific and may be seen with pancreatitis and lymphoma.4,53 On MRI the focal lesion is hypointense compared with the normal pancreas on T1-weighted fat-suppressed images; the lesion appears heterogeneous on T2-weighted images. On Gd-enhanced T1-weighted fat-suppressed images, peripheral enhancement with areas of central enhancement may be seen, similar to CT. Diffuse involvement has signal intensities similar to those of focal lesions and may reveal heterogeneous enhancement after IV administration of gadolinium. Obstruction and dilatation of the common bile duct with dilatation of the intrahepatic ducts can rarely occur. A normal pancreatogram on ERCP despite a tuberculous mass in the pancreatic head has been reported to be typical in tuberculous pancreatic involvement. Displacement and stenosis of an otherwise normal main pancreatic duct may occur sporadically.47,66,74

Pancreatic Adenocarcinoma.  PDAC accounts for around 90% of the malignant tumors of the pancreas and is the eighth leading cause of cancer death in the world, with a poor overall 5-year survival rate of only 6%.271,275 The median age at diagnosis is 71 years, and it is rarely diagnosed in patients younger than 40.218 Because of its silent course, late clinical symptoms, and rapid growth pattern, it has been called  the “silent killer.”206 PDAC progresses from different precursor lesions

1442

PART II  CT and MR Imaging of the Whole Body

TABLE 46-13  Criteria for Resectability of Pancreatic Tumors258 Tumor Resectability

Venous Contact

Arterial Contact

Metastasis

Resectable tumors

Pancreatic head, body, and tail: • No tumor contact or solid tumor contact of ≤180 degrees with SMV or PV, without vein distortion SMV or portal vein encasement with contour irregularity of the vein and/or thrombosis but with suitable anatomy for SMV or portal vein reconstruction

Pancreatic head, body, and tail: • No arterial tumor contact (celiac axis [CA], SMA, or common hepatic artery [CHA])

No distant metastasis

Pancreatic head/uncinate process: • Abutment or encasement of CHA without extension to CA or hepatic artery bifurcation • Solid tumor contact with SMA of ≤180 degrees Pancreatic body/tail: • Encasement of CA of ≤180 degrees Pancreatic head/uncinate process: • Abutment of SMA >180 degrees • Any solid tumor contact with CA • Solid tumor invading aorta • Solid tumor contact with first SMA branch and most proximal draining vein into SMV Pancreatic body and tail: • Solid tumor contact of >180 degrees with SMA or CA • Solid tumor contact with aorta

No distant metastasis

Borderline tumors

Unresectable tumors

Pancreatic head/uncinate process: • Unreconstructible SMV or PV • SMV occlusion (can be due to tumor or bland thrombus) Pancreatic body and tail: • Unreconstructible SMV or PV • SMV occlusion (can be due to tumor or bland thrombus)

such as pancreatic intraepithelial neoplasm (PIN) and intraductal papillary mucinous neoplasm (IPMN). Although multiple oncogenes are activated through the progression pathway, KRAS activation is the most common oncogenic mutation in PDAC. KRAS mutation is associated with poor prognosis.192 Approximately 60% to 70% of PDACs occur in the head of the pancreas; the remainder are found in the body or tail or diffusely infiltrate the organ. PDAC is associated with an intense desmoplastic (stromal) reaction and tends to obstruct the pancreatic duct, with subsequent upstream duct dilatation and parenchymal atrophy; if it arises in the head, the common bile duct can be stenosed, with biliary tree dilatation. Pancreatic adenocarcinoma can extensively infiltrate the retroperitoneum and invade surrounding anatomic structures including the duodenum, stomach, mesenteric vessels, portal vein, neurovascular spaces, and lymphatic channels. It tends to metastasize to regional lymph nodes (at diagnosis, about 80% of cases of pancreatic head adenocarcinoma present with lymph node metastases), liver, and peritoneum.271 Because of the early occurrence of clinical symptoms, masses arising in the head of the pancreas tend to be smaller (2-3 cm in diameter) at diagnosis than those in the body and tail (5-7 cm in diameter).271 Clinical presentation varies according to the cancer’s site of origin and stage, but many patients recall a history of long-standing abdominal pain, asthenia, reduced appetite, and weight loss. New-onset diabetes mellitus is present in 10% of patients. Painless jaundice is present in 75% of patients at presentation, mostly in the cases of tumors arising from the head of the pancreas.36,123 Tumors in the body and tail tend to present with back pain related to tumor infiltration of the surrounding retroperitoneal structures and nerves.271 Late clinical presentation with advanced disease and the tumor’s biological aggressiveness result in a low rate of surgical intervention (10%20%) and high mortality.205 Complete surgical resection along with adjuvant therapy remains the best curative treatment for pancreatic adenocarcinoma. Currently, less than 20% of patients are candidates for surgical resection, and  the small tumors that are amenable to curative surgery are the 

Distant metastasis (including nonregional lymph node metastasis)

most difficult to detect.21,36,103 The treatment approach is based on whether the tumor is resectable or nonresectable at presentation56,77 (Table 46-13). For locally advanced pancreatic cancer (Figs. 46-56  to 46-58), chemotherapy, radiotherapy, or both are alternatives to surgery.153,265 Patients with distant metastases are considered ineligible for radiation treatment and undergo chemotherapy alone, although the traditional chemotherapy regimen for pancreatic adenocarcinoma, based on 5-fluorouracil (5-FU) or gentamicin, offers only a small survival advantage and only a slight improvement in the quality  of life.103,242 Despite the fact that tumor serum marker CA 19-9 is considered a sensitive but nonspecific marker for the diagnosis of adenocarcinoma of the pancreas, it is rarely positive in tumors less than 1 cm in diameter.203,271 Therefore diagnostic imaging plays a critical role in the evaluation of this disease. Multiphasic MDCT plays a major role in the assessment of pancreatic adenocarcinoma. Moreover, advanced CT technology such as DECT has become a commonly used imaging modality for evaluation of PDAC. Additional imaging techniques for diagnosis and staging include US, MRI, PET CT, ERCP, and endoscopic ultrasound (EUS). The diagnostic imaging approach to PDAC depends on the clinical indications: (1) patients who are clinically suspected of having pancreatic cancer and need diagnostic imaging to verify the presence or absence of disease, (2) patients who have already been diagnosed with PDAC and require staging, and (3) patients at high risk for this tumor who require screening. The TNM classification for pancreatic cancer staging is presented in Box 46-2. Role of CT in screening.  About 5% to 10% of pancreatic cancers are due to hereditary factors. For patients with a first-degree relative affected with pancreatic cancer, the relative risk for developing pancreatic cancer increases 2 to 6 times. Also, members of families with a history of hereditary cancer syndromes such as Peutz-Jeghers syndrome, hereditary pancreatitis, hereditary breast and ovarian cancer syndrome, familial atypical multiple mole melanoma syndrome, and hereditary nonpolyposis colorectal cancer are at increased risk of developing pancreatic cancer. In these cases, close follow-up might lead

CHAPTER 46  Pancreas to detection of tumors at an early stage and an improvement in overall survival.55 Although there is consensus in the need of screening in patients with a predisposition for pancreatic cancer, there is no consensus regarding the most effective modality for screening or the appropriate time interval between the screenings.56

FIG 46-56  Adenocarcinoma of the pancreas. Axial contrast-enhanced pancreatic-phase image shows a poorly defined area of hypodensity in the neck and body of the pancreas (white arrow) infiltrating the retroperitoneum. The superior mesenteric vein is severely reduced in caliber, is circumscribed by the lesion for more than 180 degrees, and is teardrop shaped (black arrow), suggesting infiltration.

Role of MDCT.  MDCT allows the pancreas to be imaged at high spatial and temporal resolution within a short breath hold.88,203 Contrast enhancement time is crucial in tumor detection. It has been demonstrated that both peak enhancement and time to peak enhancement of the pancreas are related to the contrast injection rate, and there is a strong positive correlation between pancreatic parenchymal

FIG 46-57  Adenocarcinoma of the pancreas. Axial contrast-enhanced PVP image shows a mass (arrows) in the neck and body of the pancreas, with poorly defined and infiltrating margins. The mass exhibits reduced and inhomogeneous contrast enhancement, and there is upstream dilatation of the pancreatic duct (single arrowhead). The mass surrounds the superior mesenteric artery (wavy arrow). Peritoneal fluid and thickening (two arrowheads) represent metastatic implants.

SMA

A

1443

B FIG 46-58  Adenocarcinoma of the pancreas. Coronal contrast-enhanced (A) and coronal MIP (B) MDCT images demonstrate a mass in the head and uncinate process of the pancreas (white arrows). It appears hypodense after contrast administration, exhibits poorly defined margins, infiltrates surrounding fat tissue, and reduces the caliber of the superior mesenteric artery (SMA) (arrowheads). Multiple small, rounded, hypodense metastatic foci are seen in the liver (black arrows in A).

1444

PART II  CT and MR Imaging of the Whole Body

enhancement and both the volume of contrast material and its rate  of injection. A dual-phase (PP and PVP) pancreatic protocol for MDCT is a sensitive technique for detecting and staging PDAC and for detecting metastases to the liver and peritoneum. Time delays can be predetermined or calculated according to the patient’s hemodynamic

BOX 46-2  AJCC TNM Staging System for

Pancreatic Cancer

Primary Tumor (T) Primary tumor cannot be assessed Tx T0 No evidence of primary tumor Tis Carcinoma in situ T1 Tumor limited to the pancreas, 2 cm or less in greatest dimension T2 Tumor limited to the pancreas, more than 2 cm in greatest dimension T3 Tumor extends beyond the pancreas but without involvement of the celiac axis or the superior mesenteric artery T4 Tumor involves the celiac axis or the superior mesenteric artery (unresectable primary tumor) Regional Lymph Nodes (N) Nx Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Regional lymph node metastasis Distant Metastasis (M) M0 No distant metastasis M1 Distant metastasis From Edge S et al (eds): AJCC cancer staging manual, ed 7, New York, 2010, Springer-Verlag.

parameters. In the case of a predetermined protocol, a scan delay of 40 to 50 seconds from the start of a bolus of IV contrast agent is used for the PP, and a scan delay of 65 to 70 seconds is used for the PVP.17,76,146 With 16-detector or higher MDCT scanners it is preferable to calculate delays based on the patient’s hemodynamics and injection rate. The PP allows optimal tumor detection and mapping of the regional vascular structures; in this phase the PDAC appears as a low-density lesion compared with the normal pancreatic parenchyma. In the PVP, the tumor conspicuity against the normal pancreas may be reduced owing to contrast diffusion into the interstitium of the tumor, but the detection of liver and peritoneal metastasis and the visualization of portal venous structures are improved.76,252 However, in about 10% of cases the mass is isoattenuating with contrast enhancement and cannot be directly observed. In these cases, useful indirect signs include stenosis of the distal common bile duct, stenosis of the pancreatic duct with upstream ductal dilatation, parenchymal atrophy, “double-duct” sign (stenosis of both the common bile duct and the pancreatic duct, with subsequent upstream dilatation), loss of lobulation of the pancreatic parenchyma, and deformity of the pancreatic contours.226 MDCT data can be postprocessed in various planes to maximize the diagnostic yield of the examination and improve visualization of the pancreatic vasculature and biliary structures. Among the different postprocessing techniques, MIP is particularly useful to display vascular structures; volume-rendered imaging is apt to depict both vessels and soft tissues. Multiplanar volume rendering and curved reconstructions, which can be constructed along the course of the pancreatic duct, provide MDCT pancreatograms17,75 (Figs. 46-59 to 46-61). MDCT pancreatograms allow one to follow the course of the dilated pancreatic duct, find the site of obstruction, analyze the characteristics of the stenosis, and visualize the surrounding parenchyma. Preoperative vascular evaluation can be accurately performed with MDCT

* A

B

C FIG 46-59  Adenocarcinoma of the pancreas. A, Coronal PVP MDCT demonstrates a mass in the head and uncinate process of the pancreas (short arrows) infiltrating the third portion of the duodenum (long arrow) and surrounding the superior mesenteric artery (arrowhead). Coronal (B) and axial (C) curved reformatted images of the same patient display the tumor (arrows) and dilatation of the main pancreatic duct (arrowheads), which abruptly terminates in the mass. A biliary stent is seen in situ (asterisk).

CHAPTER 46  Pancreas angiography and constitutes a crucial part of the MDCT examination. Its purposes are avoidance of intraoperative vascular complications and accurate staging. The negative predictive value of MDCT for overall tumor resectability was found to be high (87%) in a recent study.259 The accuracy of MDCT (86%) in the assessment of global locoregional extension is higher than that of EUS (65%) and MRI (75%), owing to the superior visualization of fat infiltration. However, EUS was most accurate in the

FIG 46-60  Adenocarcinoma of the pancreas. Curved reformatted MDCT shows an oval mass in the tail of the pancreas, with ill-defined margins and poor enhancement (arrow).

A

1445

evaluation of duodenal infiltration (EUS, 76%; CT, 74%; MRI, 67%). In some studies, EUS had the highest overall accuracy with respect to tumor involvement of regional lymph nodes. MDCT is the modality with the highest global accuracy in the assessment of vascular invasion.236 The likelihood of vascular infiltration by adenocarcinoma of the pancreas increases with greater tumor-to-vessel circumference contact. The likelihood of vessel infiltration is less than 3% when the tumor-to-vessel circumference contact is less than 90 degrees; it is between 29% and 57% for contact between 90 and 180 degrees, and it is more than 80% for contact greater than 180 degrees.153,191 Other useful criteria that indicate vascular invasion are the teardrop sign, which refers to the shape of the portal vein or the superior mesenteric vein (SMV); the lack of preservation of a fat plane around the vessels; and dilatation of the pancreaticoduodenal veins (see Figs. 46-56 and 46-58).123,148,272 A dual-phase dedicated MDCT scan of the pancreas must be performed in patients in whom there is a strong clinical suspicion of a pancreatic lesion. To reduce radiation-related risks to the patient, a relatively low-radiation examination consisting of a single PVP contrast-enhanced CT examination, with a 5- to 7-mm slice thickness, is generally sufficient for posttreatment follow-up. Potential role of DECT.  As previously discussed, a small percentage of PDACs are not directly visualized on conventional SECT scans—even with optimized protocols—because of their size or the slight difference in attenuation between the tumor and adjacent pancreatic tissue.170 Dual energy is an evolving technique with several potential clinical applications in the abdomen, including detection of challenging pancreatic lesions.93,178 Moreover, DECT has potential for contrast media reduction through selection of low-energy images and radiation dose reduction through elimination of the noncontrast acquisition. A dual-phase pancreatic protocol (PP and PVP) is used on DECT scanners, similar to that on SECT. However, the PP acquisition is acquired in the DECT mode (140/80 kVp), and the PVP is acquired with SECT mode (120 kVp). Different studies have demonstrated higher contrast of PDAC at low x-ray energies (50-60 keV) and increased tumor conspicuity, allowing identification of early-stage

B FIG 46-61  Adenocarcinoma of the pancreas. Axial contrast-enhanced pancreatic-phase MDCT (A) and curved reformatted MDCT (B) along the course of the main pancreatic duct display a hypodense mass in the body of the pancreas (arrows), with upstream dilatation of the main pancreatic duct (arrowheads).

1446

PART II  CT and MR Imaging of the Whole Body

tumors not visualized on conventional CT.170,202 MD-I images improve depiction and characterization of hypoattenuating malignancies such as PDAC based on the difference of tissue iodine concentration (see Fig. 46-22). Iodine maps also improve the morphologic and enhancing characteristics of a pancreatic mass and show better delineation of lesion margins and definition of the extension of the lesion. DECT can contribute to the staging of PDAC by improving evaluation of vascular compromise and distant metastasis. DECT allows more confident rule out of suspected pancreatic tumors. Role of MRI.  MRI is usually considered a second-line imaging modality in patients with a high clinical suspicion of pancreatic tumor or in those with a suspected pancreatic mass and indeterminate findings on high-quality MDCT.173 The factors that limit MRI’s use as a first-line tool for PDAC diagnosis and staging are related to motion and breathing artifacts in the case of poor patient cooperation, limited availability of the technique, and high costs. Currently MRI is mainly a problem-solving technique in patients with inconclusive MDCT studies. MRI can also be considered as an alternative in patients allergic to iodinated contrast agents, in patients with renal insufficiency, and in all cases in which reduced radiation exposure is required. In a recent study comparing state-of-the-art 16-detector MDCT and 1.5-T MRI, it was demonstrated that the sensitivity for detecting adenocarcinoma with MDCT alone, MDCT plus multiplanar reformations (MPR), and MRI alone was 83%, 96%, and 83%, respectively. In the same study, MRI had higher specificity than MDCT alone (98% vs. 89%), but MDCT plus MPR performed similarly to MRI (97%).22,172 Owing to its inherent soft tissue contrast and high resolution, MRI including DWI can improve detection of subtle pancreatic lesions and small non–organ-deforming adenocarcinomas.105,189 On MRI, PDAC appears hypointense on nonenhanced T1-weighted images and, depending on the amount of associated desmoplastic response, demonstrates variable signal intensity on T2-weighted images. The surrounding normal parenchyma shows effacement of its lobular architecture. On dynamic imaging following IV administration of gadolinium contrast agent, PDAC enhances relatively less than the background pancreatic parenchyma in the early phase. It shows progressive enhancement in subsequent phases23 (Figs. 46-62 to 46-64). As with MDCT, the regional vascular anatomy can be mapped  to assess for vascular involvement using 3D contrast-enhanced dynamic MR angiography. Moreover, 2D and 3D MRCP allows excellent visualization of the entire extrahepatic biliary tree and pancreatic duct. MRCP can demonstrate the typical double-duct pattern of obstruction, which is often seen with pancreatic or ampullary tumors.85 When MRCP is used in combination with a dynamic MRI examination of the pancreas, diagnostic preoperative ERCP can be avoided. MRI is particularly useful in detection and characterization of liver lesions and peritoneal implants, performing better than MDCT in this respect, especially in the setting of the small lesions.79,161 It has been demonstrated that peritoneal or hepatic metastases can be found at laparoscopy in 37% of patients with locoregional disease on CT.36 Secretin-enhanced MRCP can be performed as an integral part of the MRI examination. The increased duct distention improves MRCP quality and lesion conspicuity and allows more accurate assessment of pancreatic duct stenosis, helping to differentiate benign from malignant strictures. It can also enhance the diagnostic confidence.81,100,158,165,187 Because mass-forming chronic pancreatitis and pancreatic carcinoma share many morphologic characteristics on imaging studies, their differentiation is difficult with both CT and MRI. The “ductpenetrating” sign is considered highly specific (96%) for differentiating the two. It is characterized by visualization of the main pancreatic duct, without ductal wall irregularities, within the pancreatic mass.267 A

main pancreatic duct with an irregular wall running through the mass is not considered predictive of a benign cause. In this setting, secretinenhanced MRCP (owing to secretin-induced distention of the pancreatic duct) may be useful in making the correct diagnosis. Role of hybrid PET CT.  Although both MDCT and MRI provide excellent anatomic delineation of pathology of the pancreas, they are not always able to differentiate between malignant and benign disease.88,189 As a unique molecular and functional imaging technique, PET can complement structural modalities and overcome some of the deficiencies of structural imaging.5 PET performed with 18F FDG is increasingly being used for the diagnosis and staging of various cancers. The normal pancreas is not visualized on PET, so avid uptake in the pancreatic region is considered abnormal.137 An ongoing challenge in diagnostic radiology has been the accurate detection of pancreatic carcinoma and its differentiation from massforming pancreatitis with imaging methods.171 CT-guided fine-needle biopsy has been recommended for patients with ambiguous CT findings; however, even this technique has shortcomings owing to possible sampling errors.30 Accurate detection of lymph node metastases is also problematic because they are usually very small (<1 cm in diameter) and are frequently missed with CT or US; this may cause early tumor recurrence if extensive surgical clearance of lymphatic tissue is not performed.109 With the semiquantitative standardized uptake value (SUV) index cutoff of 1.53, the sensitivity and specificity for cancer detection with 18F FDG PET increase.98,237 Also, mass-forming chronic pancreatitis has relatively low levels of 18F FDG uptake in the pancreas, indicating the potential of 18F FDG PET to differentiate it from carcinoma.98 Lymphatic spread is very common in pancreatic cancer and can be present even with small tumors (<2 cm diameter) that are usually regarded as early cancer. Local and distant metastases are also correctly localized with PET with 81% sensitivity and 100% specificity, thus helping to stage the cancer.99 Because 18F FDG accumulation does not depend on tumor size, cancer can be detected even in normal-sized lymph nodes. This has clinical relevance in patients with small tumors that are potentially curable if complete resection (including local metastases) can be achieved.12 However, false-negative PET findings are observed in patients with insulin-dependent diabetes mellitus, which might be due to altered 18F FDG distribution in these patients; falsepositive results may occur in inflammatory or infectious benign lesions as well as macrophage-containing tumors after successful anticancer therapy.13 PET using agents such as C-11-l-dopa and C-11-5-HTP is more sensitive in detecting functional endocrine pancreatic tumors and relatively less sensitive in detecting nonfunctional tumors, thus offering the ability to characterize these tumors on the basis of their metabolic characteristics.2 PET CT has been useful in detecting malignancy in mucinous lesions, which tend to have a high malignant potential, thus adding to the sensitivity of CT and contributing to the management approach in these patients. Role of tissue sampling.  Not all pancreatic PDACs clearly present as hypodense lesions with biliary and pancreatic duct dilatation. Even when these features are present, they are not specific for PDAC; other neoplasms such as neuroendocrine tumors, acinar cell tumors, solid and papillary neoplasms, and even nonneoplastic pathologies such as focal chronic pancreatitis or autoimmune pancreatitis can share these features. In these cases, EUS-guided fine-needle aspiration cytology plays a role. The technique has high sensitivity (75%-90%) and specificity (94%-100%).91,197,268 However, because of the marked desmoplastic reaction secondary to PDAC, a benign pathology established on EUSguided fine-needle aspiration cytology does not completely exclude

CHAPTER 46  Pancreas

A

B

C

D

E FIG 46-62  Adenocarcinoma of the pancreas. Coronal oblique MRCP (A), axial T1-weighted image (B), axial T2-weighted image (C), axial contrast-enhanced T1-weighted image in the arterial phase (D), and axial pancreatic-phase MDCT (E) display the “double duct” sign, consisting of common bile duct and pancreatic duct dilatation (arrowheads in A). A biliary stent is in situ (thin arrow). An abrupt biliary tree and pancreatic duct obstruction (thick arrows in A) caused by the cancer (wavy arrow) is observed.

1447

1448

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E FIG 46-63  Adenocarcinoma of the pancreas. Coronal oblique MRCP (A), axial T1-weighted image (B), axial T2-weighted image (C), axial contrast-enhanced T1-weighted image (D), and axial MDCT pancreatic-phase image (E) demonstrate the “double duct” sign (arrowheads in A) and a head–uncinate process mass that appears ill-defined and hypointense on the T1-weighted image, hyperintense on the T2-weighted image, and poorly enhancing after the administration of gadolinium and iodinated contrast material (arrow in B to E).

CHAPTER 46  Pancreas

A

B

C

D

E

F

1449

FIG 46-64  Adenocarcinoma of the pancreas. Coronal single-shot FSE T2-weighted image (A), axial T1-weighted image (B), axial T2-weighted image (C), axial contrast-enhanced T1-weighted image in the pancreatic phase (D), delayed contrast-enhanced T1-weighted image (E), and axial contrast-enhanced MDCT (F) show dilatation and distal stenosis of the pancreatic duct (arrow in A), displacement and stenosis of the common hepatic duct (arrowhead in A), and a contour-deforming mass in the head of the pancreas (arrows in B to F) that appears poorly defined and hypointense on the T1-weighted image, hyperintense on the T2-weighted image, and poorly enhancing after contrast administration. It diffusely infiltrates the adjacent retroperitoneum (arrowheads in E and F). A necrotic metastatic node is indicated by the wavy arrow in E and F.

malignancy.257 For lesions smaller than 2.5 cm, EUS-guided biopsy is more accurate than CT-guided biopsy. CT-guided biopsy should be reserved for cases of advanced pancreatic carcinoma.104,114,261 Postoperative imaging.  Pancreaticoduodenectomy, also known as the Whipple procedure, is a potentially curative surgical technique for the treatment of PDAC. Although recognition of tumor recurrence and postoperative complications is not an easy task, diagnostic

imaging plays a critical role in postoperative assessment of patients undergoing the Whipple procedure. MDCT is considered the imaging modality of choice in the follow-up of postsurgical patients. It allows detection of both local and distant abdominal tumor recurrence  as well as metastatic lymphadenopathy. Local recurrence usually  presents as an irregular mass or soft tissue infiltration in the surgical bed, which may extend along the superior mesenteric vessels or the

1450

PART II  CT and MR Imaging of the Whole Body

portal vein. Differentiation of postoperative inflammation from recurrent tumor is the most important information provided. An increase in the size of the mass on serial follow-up imaging, loss of a distinct fat plane between mesenteric vessels, encasement of the vessels, and presence of adjacent bowel obstruction favor local recurrence rather than postsurgical inflammation or pseudotumor.129 Preliminary data from our institution have shown the capability of DECT-iodine images to improve readers’ evaluation of the surgical bed after a Whipple procedure. DECT allows improved visualization of surgical bed anatomy and better differentiation between normal postoperative changes and tumor recurrence. Local tumor recurrence shows more heterogeneous iodine distribution in the surgical bed  and higher iodine concentration compared to normal postsurgical changes. Additionally, DECT can improve assessment of lymph nodes and detection of liver metastasis. Local recurrence can be excluded with more confidence on iodine images and low-keV images from DECT. Patients with recurrent pathology on MDCT usually present with raised serial CA19-9 levels. PET CT performed 6 months after surgery is useful for detecting tumor recurrence and differentiating it from postoperative changes.216 Pitfalls and differential diagnosis.  Many anatomic variants can be confused with PDAC. Duodenal diverticula or a collapsed duodenum or small bowel can be confused with PDAC, but the presence of air-fluid levels filling with positive oral contrast material and characteristic enhancement of intestinal walls can lead to the correct diagnosis. It is more difficult to distinguish PDAC from mass-forming chronic pancreatitis; the most reliable finding is the duct-penetrating sign, as described earlier. Other neoplasms such as lymphoma, endocrine tumors, and acinar tumors must also be ruled out.

Endocrine Tumors.  Tumors arising from the islet cells of the pancreas account for about 2% of pancreatic neoplasms. These tumors have an incidence of 1 in 100,000 and are usually divided into functional and nonfunctional tumors on the basis of hormone overproduction and the associated clinical syndromes. Because they result in clinical symptoms, functional endocrine tumors of the pancreas are usually small at diagnosis. Nonfunctional endocrine tumors are usually discovered late, when they are large enough to cause symptoms from local infiltration or metastatic dissemination.177,190 On pathology, endocrine tumors of the pancreas are characterized by uniform nuclei, clear cytoplasm, and secretory granules. They stain positive for neuronspecific enolase and chromogranin A. Functional tumors are named by the predominant hormone they produce: insulinomas, gastrinomas, VIPomas, glucagonomas, and somatostatinomas. They can be associated with syndromes such as multiple endocrine neoplasia type 1 (MEN-1), von Hippel-Lindau disease, and neurofibromatosis.57,177 Both functional and nonfunctional tumors can be benign, malignant, or indeterminate on pathology. Insulinomas tend to be malignant in 10% of cases, gastrinomas in 60%, VIPomas in 50% to 75%, glucagonomas in 70%, and somatostatinomas in 50%. Ninety percent of nonfunctional endocrine tumors of the pancreas are malignant177,190 (Figs. 46-65 to 46-67; Table 46-14). Imaging of endocrine tumors.  When they are small, functional endocrine tumors are non–contour deforming, appear isodense at baseline CT, and enhance strongly during the arterial and portal phases of contrast administration. Sometimes they appear cystic and hypoenhancing and may contain discrete calcifications, especially if malignant. Some insulinomas may be hyperdense before contrast administration. Larger functional endocrine tumors may appear heterogeneous following contrast administration. Areas of central necrosis, as well as calcifications and retroperitoneal invasion, may be found in the case of malignancy.

Nonfunctioning endocrine tumors usually present as large, welldefined, enhancing masses, with the degree of heterogeneity proportional to their size. When small, they may present as strongly and homogeneously enhancing masses. Areas of central necrosis and calcifications may be found in larger lesions, usually malignant ones. Endocrine tumors of the pancreas tend to metastasize to regional lymph nodes and the liver. Their metastases are frequently hypervascular and therefore enhance during the arterial phase of contrast administration. On MRI, endocrine tumors appear hypointense on T1-weighted sequences and hyperintense on T2-weighted sequences, independent of their size. If smaller than 2 cm, homogeneous enhancement is common; ring enhancement predominates in larger tumors. Nonfunctional tumors tend to exhibit a higher incidence  of hemorrhage, necrosis, and cystic degeneration than functional tumors. The differential diagnosis includes pancreatic metastases from renal cancer (in the case of small tumors) and PDAC. Some imaging features of endocrine tumors that help differentiate them from adenocarcinoma are calcifications (present in 20% of cases), higher signal intensity on T2-weighted images, central necrosis or cystic degeneration, and lack of duct obstruction and vascular encasement. Moreover, glucagonomas and somatostatinomas may metastasize to the spleen. Lack of visualization of a renal primary tumor is useful to rule out pancreatic metastases from renal cancer.177,190,230

Lymphoma.  Primary pancreatic lymphoma is rare, constituting 0.5% of all pancreatic neoplasms. It presents with nonspecific symptoms such as abdominal pain, weight loss, and jaundice, mimicking the clinical picture of the more common PDAC. Sometimes CA 19-9 is elevated. Lymphomas tend to occur in those aged 35 to 75 years (mean age, 55) and predominate in men (7 : 1). Lymphomas usually occur in the head of the pancreas (80% of cases), and at diagnosis their mean diameter is 8 cm.184,224 At imaging, lymphomas may have one of two appearances: (1) a localized mass, which frequently extends to extrapancreatic regions, or (2) diffuse enlargement and replacement of the pancreas. Both forms of pancreatic lymphoma tend to show a diffuse invasive growth pattern that does not respect anatomic boundaries, infiltrating the retroperitoneal structures and GI tract. Mass-forming pancreatic lymphomas exhibit low signal intensity with a homogeneous appearance on T1-weighted sequences, and low to intermediate signal intensity with some heterogeneity on T2-weighted images. Diffusely infiltrating pancreatic lymphomas have low signal intensity on both T1- and T2-weighted sequences. Both forms usually enhance poorly but homogeneously (Figs. 46-68 to 46-70). Fine-needle aspiration biopsy is generally required for a definite histologic diagnosis. Primary pancreatic lymphomas must be differentiated from pancreatic carcinoma, endocrine tumors, and lymphoma with secondary involvement of the pancreas. Useful features to differentiate lymphoma from PDAC are the combination of a pancreatic head mass with a normal nonobstructed pancreatic duct, lymph node enlargement below the renal veins, and large size. High signal intensity on T2-weighted sequences, intense arterial contrast enhancement, and hypervascular metastases are relatively specific features commonly found in pancreatic endocrine tumors. Absence of calcifications and necrosis are other useful criteria to differentiate lymphomas from endocrine tumors. Lymph node enlargement limited to the peripancreatic region; absence of mediastinal, splenic, and hepatic involvement; no palpable superficial lymph nodes; and a normal leukocyte count favor primary pancreatic lymphoma rather than secondary involvement.94,184,211,224 Text continued on p. 1455

CHAPTER 46  Pancreas

A

B

C

D FIG 46-65  Insulinoma. Axial T2-weighted fat-suppressed image (A), axial arterial-phase T1-weighted fatsuppressed image (B), and axial PVP T1-weighted fat-suppressed image (C) demonstrate a small, rounded, well-demarcated, hyperintense lesion in the tail of the pancreas (arrow). The lesion exhibits avid contrast enhancement in the arterial phase of the dynamic image and retains contrast in the portal venous phase. D, Corresponding axial MDCT in the arterial phase of enhancement displays the same finding (arrow).

1451

A

B

C FIG 46-66  Gastrinoma. Coronal (A) and axial (B and C) contrast-enhanced MDCT images in the arterial phase display a rounded, well-circumscribed, homogeneously and intensely enhancing lesion (arrow) close to the junction of the third and fourth portions of the duodenum and the mesenteric vessels. Diffuse thickening and enhancement of the gastric folds (arrowheads in C) are seen in this patient with Zollinger-Ellison syndrome.

FIG 46-67  Nonfunctioning pancreatic endocrine tumor. Axial contrast-enhanced MDCT shows a large, welldefined, lobulated lesion in the tail of the pancreas. It appears inhomogeneous with enhancing solid areas (arrow) and has a poorly enhancing, partially necrotic component (arrowheads).

CHAPTER 46  Pancreas

1453

TABLE 46-14  Classification of Pancreatic Endocrine Tumors Functional Tumors Insulinoma

Gastrinoma

VIPoma

Glucagonoma

Somatostatinoma

Others

• • • • • • • • •

• • • • • • • • • • • • • • • • • •

50% of all endocrine tumors of pancreas Typical presentation in patients aged 30-60 years Similar distribution in male and female Can cause hypoglycemia and tend to be small at diagnosis Usually solitary, measure <2 cm in 90% of cases, and tend to be highly vascularized (see Fig. 46-65) 20% of pancreatic endocrine tumors Usually occur in fifth decade Predilection for males Arise predominantly in the so-called gastrinoma triangle: • Junction of neck and body of pancreas medially • Junction of second and third portions of duodenum inferiorly • Junction of cystic duct and common hepatic duct cranially 50% are located in duodenum, 14% in pancreas, 13% in lymph nodes; remaining gastrinomas are found in different locations Usually <1 cm, less vascularized than insulinomas, can be multiple Can induce Zollinger-Ellison syndrome About 30% are associated with MEN-1 (see Fig. 46-66) Median age at presentation is 47 years Female predilection (3 : 1) WDHA syndrome: watery diarrhea, hypokalemia, achlorhydria-hypochlorhydria 1% of all endocrine tumors of pancreas Usually occur in body or tail Main clinical manifestation: necrolytic migratory erythema Discovered when large (4-10 cm) Mild hyperglycemia may also occur Mean age of presentation: 51 years More often in women Originally pancreatic in 50% of patients and duodenal in remaining 50% Associated with neurofibromatosis type 1 in 50% of cases May cause diabetes, gallbladder disease, and steatorrhea Parathyroid hormone– and adrenocorticotropic hormone–secreting pancreatic tumors are extremely rare.

Nonfunctional Tumors • Usually occur in fourth or fifth decade • Usually arise in head of pancreas • Large size at presentation because of absence of symptoms of hormone overproduction: from 3-24 cm, with 30% exceeding 10 cm • Clinical manifestations when their size or infiltration causes abdominal pain, anorexia, and weight loss • 90% of nonfunctioning pancreatic tumors are malignant • Higher resectability rate and a better response to chemotherapy compared to pancreatic adenocarcinoma (see Fig. 46-67)

1454

A

PART II  CT and MR Imaging of the Whole Body

B

FIG 46-68  Pancreatic lymphoma. Axial T1-weighted opposed-

C

phase (A), axial T2-weighted fat-suppressed (B), and axial arterialphase T1-weighted fat-suppressed (C) MRIs display enlargement and rounding of the pancreatic head, with a well-demarcated area of low signal intensity within, corresponding to lymphoma (arrows). The low-signal-intensity area does not significantly enhance compared with the highly enhancing normal pancreatic parenchyma.

FIG 46-69  Pancreatic lymphoma. Axial contrast-enhanced MDCT shows a well-demarcated hypodense area in the head of the pancreas (arrow). The pancreatic duct is not dilated (arrowhead).

CHAPTER 46  Pancreas

A

B

C

D

1455

FIG 46-70  Pancreatic lymphoma. Axial T1-weighted opposed-phase (A), axial contrast-enhanced T1-weighted fat-suppressed (B), and axial T2-weighted fat-suppressed (C) images show enlargement and rounding of the head of the pancreas. An area of low signal intensity (arrow) is best seen on the contrast-enhanced image in the dorsal aspect (B). MRCP (D) displays smooth tapering of the distal common bile duct (large arrow) and of the duct of Wirsung (arrows), main pancreatic duct prominence (large arrowheads), and widening of the duodenal loop (small arrowheads).

Usually pancreatic lymphomas respond well to chemotherapy, radiotherapy, or both. Surgery is limited to cases in which a definite diagnosis cannot be achieved.94,224

Acinar Cell Carcinoma.  Acinar cell carcinoma (ACC) is a rare pancreatic tumor, accounting for about 1% of exocrine pancreatic neoplasms. It is characterized by the production of pancreatic enzyme by tumor cells. ACC occurs more often in women than in men, with a peak incidence in the seventh decade. ACC tends to arise as a single mass in the uncinate process and head of the pancreas (60% of cases). In about 80% of cases it exhibits exophytic growth (Fig. 46-71). The mean tumor size at presentation is 7 cm. A well-defined capsule circumscribes the tumor, which may present focal areas of discontinuity

and infiltration into the surrounding organs. At microscopy, tumor cells present an acinar arrangement. Electron microscopy and immunohistochemistry are usually required for a definite diagnosis.40,97,131,246 The clinical presentation is heterogeneous, ranging from lack of symptoms to jaundice, abdominal pain, vomiting, and weight loss. Some patients experience polyarthritis or subcutaneous fat necrosis induced by hyperamylasemia. CA 19-9 is elevated in 30% of patients. Acinar cell cancers tend to infiltrate locally and disseminate to lymph nodes and hepatic parenchyma. Such metastases are present in about 50% of cases at presentation. ACC recurs in 79% of patients after curative resection. Its 5-year survival rate is 5.9%.40,97,131,246 On imaging, about 50% of tumors harbor calcifications in the form of central punctate or stellate calcifications or peripheral punctuations

1456

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 46-71  Pancreatic acinar cancer. Axial pancreatic-phase (A) and PVP (B) MDCT images display an oval hypodense mass in the body and tail of the pancreas (white arrow), with an exophytic growth pattern, infiltrating the stomach (black arrows). Curved pancreatographic reformatting (C) demonstrates the relationship between the tumor (arrow) and the nondilated pancreatic duct (arrowheads), a useful criterion in distinguishing acinar cancer from adenocarcinoma.

or plaques. In about 60% of cases the tumor appears circumscribed by a well-defined and enhancing capsule, although areas of capsular discontinuity and infiltration can be detected. Central necrosis is a common finding, present in about 80% of cases; it is large in more than half of cases. Necrosis may account for the cystic or mixed cysticsolid appearance. Hemorrhage has not been reported in ACC. The tumor enhances mildly in the arterial and portal phases of dynamic contrast imaging.40,246 The differential diagnosis includes adenocarcinoma of the pancreas and endocrine tumors. The combination of calcifications and central necrosis is useful for this purpose. Calcifications are found in 50%  of ACCs, 22% of endocrine tumors, and rarely in adenocarcinomas. Central necrosis is rarely present in adenocarcinomas; it occurs in about 18% of functional and 54% of nonfunctional endocrine tumors. ACCs enhance more than adenocarcinomas and usually less than endocrine tumors. Moreover, ACCs are circumscribed by a  well-formed capsule, which is lacking in adenocarcinomas and endocrine tumors. Solid and pseudopapillary tumors may contain central

areas of necrosis, may be circumscribed by a capsule, and may enhance to the same degree as ACCs, but they present with macroscopic hemorrhage, which cannot be found in ACC, and they occur in young females.40,246

Intrapancreatic Metastases.  Intrapancreatic metastases are rare, accounting for 2% of pancreatic tumors. Renal, lung, thyroid, breast, and colorectal cancers and melanoma can spread to the pancreas, usually in the setting of diffuse disease, sometimes years after treatment of the primary cancer. They may be single or multiple, and the mean size at diagnosis is 4.6 cm. Most patients are asymptomatic, and pancreatic metastases are detected during routine oncology follow-up.67,90,144 At imaging, metastases tend to closely reflect the imaging appearance of the primary tumor. Therefore metastases from a hypervascular primary tumor such as renal cancer show intense enhancement during the arterial phase of dynamic contrast imaging (Figs. 46-72 and 46-73). If small, they appear homogeneous; if large, internal heterogeneity with necrotic areas is found. Metastases from the colon mimic PDACs;

CHAPTER 46  Pancreas they usually spread to the pancreas through the lymphatic network from right-sided colon cancers.37,90 The main differential diagnosis involves PDACs and endocrine tumors. Preoperative diagnosis is difficult and begins with clinical suspicion based on knowledge of the clinical oncology history. Comparing the appearance of the pancreatic mass and its enhancement characteristics with those of the primary tumor is mandatory. Usually biopsy is required.

Cystic Lesions.  Increased detection of cystic lesions of the pancreas at cross-sectional imaging has led to a rise in the number of surgical

FIG 46-72  Intrapancreatic metastases from renal cancer. Axial contrastenhanced MDCT in the arterial phase shows a rounded, relatively homogeneous, well-circumscribed area of enhancement in the uncinate process (arrow). This patient had undergone left nephrectomy for renal cancer.

A

1457

resections of the pancreas.70 However, many cystic lesions are benign, and a large number are detected incidentally; hence resection may  be unjustified in some cases.6,70 Therefore it is important to characterize cystic neoplasms and distinguish them from pseudocysts.222 Occasionally, solid tumors of the pancreas such as islet cell tumors and metastases have an associated cystic component or may undergo degeneration and can mimic a cystic neoplasm on imaging. Differ­ entiating cystic neoplasms from PDAC is important because the prognosis for malignant cystic neoplasms is better than that for ductal adenocarcinoma.157 Accurate preoperative characterization of the lesion aids in prognostication and guides therapeutic decision making (Fig. 46-74). Types of lesions.  Pseudocysts, serous cystadenomas, mucinous cystic neoplasms, and IPMNs are most often encountered in routine practice, accounting for more than 90% of cystic pancreatic lesions (Box 46-3; Figs. 46-75 to 46-90).32,222 MDCT that allows thin-section scanning of the pancreas is the preferred imaging modality for the initial assessment (detection and characterization) of pancreatic cysts.48 Unpublished data from our institution have shown improved assessment of the morphologic characteristics of the cystic lesions on DECT compared to conventional CT. Iodine images have shown more accurate depiction of mural nodules, septa, hemorrhage, and calcifications (see Fig. 46-24). MRI with MRCP also accurately depicts the morphologic features of the cyst and has the advantage of demonstrating the cyst’s relationship to the pancreatic duct.107 EUS and EUSguided aspiration and biopsy are useful in cases that are indeterminate on cross-sectional imaging owing to morphologic overlap or in cases that require observation.222 Cystic lesions may be asymptomatic or present with abdominal pain, jaundice, or recurrent pancreatitis, often indicating that the lesion is either in communication with the pancreatic ductal system or obstructing the pancreatic, or biliary duct.222,244 Patients with more advanced cystic malignancies may present with symptoms similar to those of pancreatic cancer.231 Table 46-15 summarizes the pancreatic cystic lesion characteristics. Text continued on p. 1467

B FIG 46-73  Intrapancreatic metastases from renal cancer. Coronal PVP (A) and axial arterial-phase (B) MDCT images demonstrate a rounded well-demarcated lesion in the tail of the pancreas (arrow). It exhibits heterogeneous contrast enhancement.

Surgery or EUS-guided cyst aspiration

Management depends on several factors*

Symptomatic/ large Surgery

Asymptomatic

Serial imaging (CT/MR) follow-up

Serous cystadenoma

Microcystic

Management depends on several factors†

Serial imaging (CT/MR) follow-up

Asymptomatic

Surgery

Symptomatic dilated MPD (combined IPMN)

SB IPMN

Surgery

MCN Rare (non Fn. NET, lymphangioma)

Macrocystic

Surgery

Malignant or malignant potential

True cystic lesions with solid component • MCN • IPMN Solid lesions with cystic component • Islet cell tumor • Adenocarcinoma • Metastasis • SPPT

Cyst with solid component

cystic neoplasm; MPD, main pancreatic duct; Non Fn. NET, nonfunctioning neuroendocrine tumor; S., serum; SB, side branch; SPPT, solid pseudopapillary tumor. (From Sahani DV et al: Cystic pancreatic lesions: A simple imaging-based classification system for guiding management. Radiographics 25:1471–1484, 2005.)

FIG 46-74  Imaging-based approach to cystic lesions of the pancreas. EUS, endoscopic ultrasonography; IPMN, intraductal papillary mucinous neoplasm; MCN, mucinous

*Multiple unilocular cysts include pseudocyst, von Hippel-Lindau disease, and IPMN. †Management decisions are based on several factors, including the patient’s age, surgical risk, and size and location of the cyst. Surgery should be considered for larger cysts and younger patients. Serial imaging follow-up (CT or MRI) is recommended for small asymptomatic cysts without any suspicious features and for older patients, including those who are at high risk for surgical complications.

Symptomatic

Alternative diagnosis • IPMN • Unilocular serous cystadenomas • Lymphoepithelial cysts

Pseudocyst

Asymptomatic

No pancreatitis Normal S. amylase

Pancreatitis ↑ S. amylase

Unilocular*

Cystic lesion of pancreas

1458 PART II  CT and MR Imaging of the Whole Body

CHAPTER 46  Pancreas

A

1459

B FIG 46-75  A, Axial contrast-enhanced MDCT reveals a finely lobulated lesion (arrow) with a microcystic appearance in the proximal body of the pancreas. B, Corresponding coronal T2-weighted MRI demonstrates the same findings (arrow) and also reveals a central scar (arrowhead) that is virtually pathognomonic for serous cystadenoma.

BOX 46-3  Pathologic Classification of

Cystic Pancreatic Tumors

Pseudocyst Common cystic pancreatic neoplasms Serous cystadenoma Mucinous cystic neoplasm Intraductal papillary mucinous neoplasm Rare cystic pancreatic neoplasms Solid pseudopapillary neoplasm Acinar cell cystadenocarcinoma Lymphangioma Hemangioma Paraganglioma Solid pancreatic lesions with cystic degeneration Cystic islet cell tumors (insulinoma, glucagonoma, gastrinoma) Pancreatic adenocarcinomas Metastasis Cystic teratoma Sarcoma True epithelial cysts*

FIG 46-76  Axial contrast-enhanced MDCT reveals a lobulated lesion with internal septations and a calcified central scar (arrow), a characteristic finding in serous cystadenoma.

*Associated with von Hippel-Lindau disease, autosomal dominant polycystic kidney disease, and cystic fibrosis.

FIG 46-77  Axial contrast-enhanced MDCT reveals a macrocystic or oligocystic variety of serous cystadenoma with fewer large (>2 cm) internal cysts (arrow). This variant may be difficult to differentiate from a mucinous cystic lesion.

1460

PART II  CT and MR Imaging of the Whole Body Mucinous cystic lesions of pancreas

DUCTAL Communication with pancreatic duct/ductal involvement/ductal dilatation

PERIPHERAL Lack of communication with pancreatic duct/ ductal involvement

Intraductal papillary mucinous neoplasm (IPMN) (lack of dense mesenchymal ovarian-like stroma)

Mucinous cystic neoplasm (dense mesenchymal ovarian-like stroma)

Main-duct or combined IPMN

Side-branch IPMN

Mucinous cystadenoma

Mucinous cystadenocarcinoma

FIG 46-78  Classification of mucinous cystic lesions.

A

FIG 46-79  Mucinous cystic neoplasm. Axial contrast-enhanced MDCT shows a large smooth cystic lesion with internal septations (arrow) and peripheral and septal calcification (arrowheads).

B FIG 46-80  Mucinous cystic neoplasm. A, Axial contrast-enhanced MDCT reveals a cystic lesion with an eccentric mural nodule (arrow). B, Corresponding coronal reformatted MDCT reveals the same findings (arrow). The presence of a mural nodule increases the likelihood of malignancy; however, it is sometimes difficult to distinguish a mural nodule from debris or inspissated mucin.

CHAPTER 46  Pancreas

A

1461

B FIG 46-81  Mucinous lesion. A, Axial contrast-enhanced MDCT image reveals a macrocystic lesion in the uncinate process (horizontal arrow in A and B). B, Corresponding axial T2-weighted fat-saturated MRI reveals a mural nodule (long arrow), which was not evident on CT, as an area of low signal intensity.

Intraductal papillary mucinous lesions

Benign (Adenoma)

Borderline (Dysplasia)

Malignant

Carcinoma in situ

R

L

Invasive

FIG 46-82  WHO classification of IPMN. (From Hruban RH et al: Intraductal neoplasms. In Hruban RH et al: Tumors of the pancreas. Washington, DC, 2007, The American Registry of Pathology, pp 71–110.)

A

B FIG 46-83  Main-duct intraductal papillary mucinous neoplasm. A, Coronal reformatted MDCT shows diffuse dilatation of the main pancreatic duct (arrows), which is filled with mucin. The ductal dilatation is disproportionate to the degree of parenchymal atrophy. B, ERCP reveals the diffuse ductal dilatation (arrows) without any evidence of side-branch irregularity.

1462

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 46-84  Main-duct intraductal papillary mucinous neoplasm. Axial contrast-enhanced (A) and coronal reformatted (B) MDCT images show the major papilla bulging into the duodenal lumen (arrow), which is considered a pathognomonic sign.

A

B FIG 46-85  Main-duct intraductal papillary mucinous neoplasm. Coronal reformatted MDCT (A) and coronal MRCP (B) reveal diffuse dilatation of the main pancreatic duct due to excessive mucin secretin (arrows in A). The distal duct (arrow in B) is not seen on the MRCP image, possibly owing to inspissated mucin.

CHAPTER 46  Pancreas

A

B

C

D FIG 46-86  Side-branch intraductal papillary mucinous neoplasm. A, Axial contrast-enhanced MDCT reveals a small cystic lesion (arrow) in the inferior body of the pancreas. B, Coronal reformatted image demonstrates the communication (arrow) of the lesion with the pancreatic duct. C and D, 2D (C) and 3D (D) MR pancreatograms demonstrate the lobulated cystic lesion and its communication (arrow) with the main pancreatic duct.

1463

1464

PART II  CT and MR Imaging of the Whole Body

FIG 46-88  Axial curved reformatted MDCT reveals multiple communicating side-branch intraductal papillary mucinous neoplasms (arrows) along the main pancreatic duct.

A

B FIG 46-87  Side-branch intraductal papillary mucinous neoplasm. A, Axial contrast-enhanced MDCT reveals a well-defined cystic lesion in the neck of the pancreas and its narrow communication (arrow) with the pancreatic duct. B, The cystic lesion and communication (arrow) are well demonstrated on a 2D MR pancreatogram acquired in the coronal oblique plane.

FIG 46-89  Combined or mixed intraductal papillary mucinous neoplasm. Axial curved reformatted MDCT reveals a loculated cystic lesion in the head, neck, and uncinate process of the pancreas (arrows), with diffuse dilatation of the main pancreatic duct (arrowheads). Histopathology revealed a mixed intraductal papillary mucinous neoplasm with papillary growth and mucin extending from the cystic lesion in the main pancreatic duct.

CHAPTER 46  Pancreas

R

L

A

R

B

L

R

C FIG 46-90  Main-duct intraductal papillary mucinous neoplasm. A, Axial contrast-enhanced MDCT reveals an isoattenuating lesion (arrow) in the terminal main pancreatic duct. B, 18F FDG PET shows a hot spot (arrow) in the same location. C, Findings are confirmed on a fused PET CT image (arrow). Histopathology revealed a high-grade malignancy.

TABLE 46-15  Characteristics of Pancreatic Cystic Lesions Pseudocyst Serous cystadenoma (SCA)

• • • • • • • • • • • • •

• • • •

Majority (30-90%) of cystic lesions of pancreas Unilocular cystic lesions in patients with clinical history of pancreatitis 30% of pancreatic cystic neoplasms Females, median age 65 years Benign lesion 70% have a polycystic or microcystic (honeycomb) pattern consisting of a collection of cysts (usually >6) that range from a few millimeters up to 2 cm in size.225 These lesions may appear solid on CT owing to the compact arrangement of cysts.208 Fine external lobulations are a common feature; enhancement of septa and cyst wall may be seen (Fig. 46-75). A fibrous central scar with or without a characteristic stellate calcification is seen in 30% of cases and considered virtually pathognomonic for serous cystadenoma (see Figs. 46-75B and 46-76).48 Hemorrhage may be seen. Patients with indeterminate CT findings: further characterization with MRI or EUS may be possible. MRI: microcysts are seen as numerous discrete foci with bright signal intensity on T2-weighted images.175 Macrocystic or oligocystic variant of SCA is uncommon (<10%) and can take the form of a single dominant macrocavity appearing as a unilocular cyst or may contain fewer large (>2 cm) cysts (Fig. 46-77). This variant may be difficult to differentiate from a mucinous cystic tumor.48 Imaging surveillance in asymptomatic patients264 Data obtained at our institution over a 3-yr follow-up period indicate average growth rate ≈ 4 mm/yr222 Large or symptomatic lesions can be resected conservatively with appropriate pancreatic segment.32,266 von Hippel-Lindau syndrome should be considered in rare cases of multiple SCAs.

1465

L

TABLE 46-15  Characteristics of Pancreatic Cystic Lesions—cont’d Mucinous cystic neoplasms (MCNs) (see Fig. 46-78)

• • • • • • • • • • • • • • • • •

Intraductal papillary mucinous neoplasms (IPMN)

• • • • • • • •

•

•

• •

•

•

•

44%-49% of pancreatic cystic lesions Women, mean age 47 years; also occurs in men Most MCNs are solitary and multilocular with a few large compartments (2-6 cm) and microcysts. Can be unilocular with a single compartment116,222 Most MCNs have a smooth contour, although a lobulated contour may be seen. Mean size > 5 cm (range, 3-20 cm); have a thick fibrotic wall that can calcify MCNs do not communicate with the pancreatic duct but can cause ductal obstruction.264 Asymptomatic (75%) or can exert mass effect225 Complex internal architecture of cyst, including septa and a solid component or mural nodule, is best appreciated on MRI and EUS (see Figs. 46-79 and 46-80).225 MRI with MRCP is considered superior to MDCT for detection of small mural nodules and communication with pancreatic duct.107,245 A mural nodule is seen as an area of low signal intensity on T2-weighted MRI; it enhances following injection of IV contrast material (see Fig. 46-81). Inspissated mucin or calcification in cyst may mimic a mural nodule on MRCP.107 High-resolution EUS is extremely sensitive for small mural nodule detection.222 MCN cysts occasionally contain debris or hemorrhage.222 Peripheral eggshell or septal calcifications on CT are specific for mucinous cystic lesions and predictive of malignancy.151 Often a single lesion may have multiple foci of gradations of histologic aggressiveness.182 Histologic heterogeneity in tumors has prompted classification of mucinous lesions into three categories: benign, borderline, and malignant (see Fig. 46-82).134 Main-duct IPMN Mucin-producing tumor of pancreas 21% to 33% of cystic pancreatic neoplasms • Does not present as a cystic lesion Affect both sexes, with a slight male • Composed of intraductal papillary growth of mucin-producing predominance, in sixth to seventh decades columnar epithelium From epithelial lining of main pancreatic duct or its • Predominantly in head of pancreas; occasionally in tail side branches, with variable ductal dilatation221 • Owing to mucin production, there is partial or diffuse dilatation of Prompt recognition and differentiation from other main duct, which is filled with mucin. Ductal dilatation is tumors are essential. disproportionate to degree of parenchymal atrophy (see Fig. Better prognosis than other pancreatic malignancies 46-83).207,244 Intermittent ductal obstruction occurs; recurrent • Excessive mucin secretion may result in bulging of major papilla abdominal pain is most frequent symptom. into duodenal lumen; this is considered a pathognomonic sign on Chronic obstruction may cause exocrine and ERCP and cross-sectional imaging and is seen more often with endocrine insufficiency leading to steatorrhea and malignant tumors (see Fig. 46-84).80 diabetes mellitus.15 Side-branch IPMN Anatomically, IPMNs can be classified as • Unilocular or multilocular cystic lesions that communicate with main-duct, branch-duct (side-branch), or mixed main pancreatic duct IPMNs, depending on morphology, site, and extent • Lack of visualization of communication on imaging does not of involvement.244 exclude a side-branch IPMN (see Figs. 46-46 and 46-86).107,222,244 Presence of mural nodules, thick septa, wall • One or more branches of pancreatic duct may be affected, which calcification, and segmental or diffuse dilatation of consequently show cystic dilatation (see Fig. 46-88).244 main pancreatic duct >10 mm in diameter indicates • Although cyst communication with pancreatic duct can also be increased likelihood of malignancy.107,221,273 seen with pseudocysts, IPMNs usually demonstrate a narrow neck Cysts >3 cm are also considered suspicious.240 at cyst-duct junction on imaging.222 Presence or absence of mural nodules or solid • Complications include obstructive pancreatitis or obstruction of components may be confirmed preoperatively by common bile duct with large lesions; malignant lesions in head EUS, which is superior to any other diagnostic may result in fistulas with duodenum or common bile duct, modality in this regard. whereas those in body or tail may result in stomach or colonic Enhanced CT and dynamic MRI may be able to fistulas.244 distinguish mural nodules or papillary neoplasms • Side-branch IPMNs have a low malignant potential; when such (≥3 mm) from mucin; nodules and neoplasms tumors can be clearly distinguished, treatment decisions should be enhance after contrast administration (see Fig. based on risk-benefit ratio, taking into account patient’s clinical 46-85). presentation, age, and surgical risk and size and morphologic IPMN can be invasive, including duodenal invasion, features of cyst.222,249 Asymptomatic side-branch IPMNs or those lymphadenopathy, peritoneal deposits, and liver without any suspicious features can be followed up without metastases.207 resection. If surgery is needed, a variety of function-preserving Malignancy is significantly higher in main-duct and minimal pancreatectomy techniques may be attempted.244 244,245 mixed IPMNs than in side-branch IPMNs. PET/ Mixed IPMN CT may be useful in increasing sensitivity of • Mixed-type IPMN is an advanced form of the side-branch type in detecting malignancy in mucinous lesions and can which the IPMN has spread to main pancreatic duct, or it is an complement anatomic details provided by CT (see ultimate form of main-duct type in which the IPMN involves Fig. 46-90). branch ducts as well (Fig. 46-89); it is not possible to establish the primary site of disease origin.207,244 • Biological behavior of mixed IPMN is similar to that of main-duct IPMN and should be dealt with surgically.

CHAPTER 46  Pancreas

*

1467

These malignant tumors are often older at presentation and have a male predilection; they are designated solid pseudopapillary carcinomas.134 The most common site for metastasis is the liver; metastases exhibit complex features similar to those of the primary tumor.42 Rarely lymph node metastasis, peritoneal spread, and multiplicity can occur.44 Invasion of the capsule and surrounding structures can occur.43,188 Benign SPT of the pancreas is treated with surgery, and complete resection is usually curative.124

Lymphangioma.  Pancreatic lymphangiomas constitute less than 1%

FIG 46-91  Axial contrast-enhanced MDCT shows a well-encapsulated lesion with varying solid and cystic components due to degeneration in a solid pseudopapillary neoplasm. The enhancing solid areas are typically noted peripherally (arrows), whereas cystic spaces are usually more centrally located (asterisk).

Solid Pseudopapillary Tumors.  Solid pseudopapillary tumor (SPT) is an uncommon benign exocrine pancreatic tumor commonly found in Asian or African American women between the second and third decades of life. These comprise about 9% of pancreatic cystic tumors. SPTs may be asymptomatic or they may present as a gradually enlarging nontender abdominal mass or with vague abdominal pain, discomfort, or obstructive symptoms.46 Pathologically these lesions are well encapsulated with varying amounts of necrosis, hemorrhage, and cystic changes. Both the capsule and intratumoral hemorrhage are important clues to the diagnosis because these features are rarely found in other pancreatic neoplasms.35 The classic CT features are a well-encapsulated lesion with varying solid and cystic components due to hemorrhagic degeneration.60 Following contrast administration, enhancing solid areas are typically noted peripherally, whereas cystic spaces are usually more centrally located (Fig. 46-91).120 MRI demonstrates a well-defined lesion with heterogeneous signal intensity on T1- and T2-weighted images, which reflects the complex nature of the mass. T2-weighted images show a thick fibrous capsule, which is seen as a rim of low signal intensity. SPTs can be differentiated from islet cell tumors by the presence of a hemorrhagic component that has high signal intensity on T1-weighted images and low or inhomogeneous signal intensity on T2-weighted images; the cystic component of islet cell tumors has moderately increased signal intensity on T1- and T2-weighted images. Moreover the peripheral portion of SPTs does not demonstrate the hypervascularity typically seen in islet cell tumors.124,193 Nonfunctional islet cell tumors are found more often in older adults, and unlike SPTs they do not have female predominance. However, SPTs with a minimal cystic components or no intratumoral hemorrhage are difficult to differentiate from islet cell tumors.43 Smaller lesions are less sharply circumscribed, often appear unencapsulated, are less likely to show prominent cystic change, and often appear as solid tumors.132 Some lesions may have internal septa or calcifications.124 Although the imaging features are informative, there is significant overlap with other lesions. Hence fine-needle aspiration biopsy, cytologic analysis, or excisional biopsy and histologic analysis are needed for definitive diagnosis.33,124 Although most SPTs exhibit benign behavior, malignant degeneration, invasion of adjacent structures, and metastases can occur.46,142

of all lymphangiomas.50 These slow-growing tumors are generally considered to be of pancreatic origin if they are within the parenchyma, adjacent to the pancreas, or connected to the organ by a pedicle.186 There is a female preponderance, and all regions of the pancreas are equally affected.196 Most patients are asymptomatic, and the lesion is discovered incidentally on imaging studies.196 Others may present with vague abdominal pain, nausea, vomiting, palpable mass, or symptoms related to compression of neighboring organs.34 Occasionally acute presentations may be related to torsion of the pedicle, rupture, or hemorrhage into the lymphangioma.102 CT or MRI reveals a septated fluid-density lesion and can assist in the preoperative evaluation by defining the exact anatomic location and content of the cyst196 (Fig. 46-92). Lymphangiomas are generally believed to be benign, so nonsurgical management may be reasonable if a definitive diagnosis can be made.196 However, because the clinical and radiographic appearance of lymphangiomas overlaps the findings for other pancreatic tumors (e.g., mucinous cystic neoplasms, cystadenomas, teratomas), exploratory laparotomy with complete surgical excision (rather than marsupialization) may be necessary.50

Hemangioma.  Pancreatic hemangioma may present as a palpable abdominal mass, GI bleeding, or a mass compressing adjacent structures such as the biliary tract or duodenum.37 The role of CT and MRI is similar to that for lymphangioma; however, MRI usually demonstrates more definitive features of hemangioma, including a lobulated mass with moderately high signal intensity on T2-weighted images  and marked enhancement after IV gadolinium administration.135,260 Enhancement may be delayed owing to sclerosis within the tumor.62 Histologically, hemangioma is easy to differentiate from lymphangioma, which has dilated lymphatic channels of varying sizes separated by thin septa.37

True Epithelial Cysts.  True cysts of the pancreas differ from pseudocysts and retention cysts in terms of origin, histologic appearance, and clinical significance. True cysts are believed to develop anomalously from remnants of the embryologic ductal systems. Histologically they have a lining of epithelial cells that may atrophy owing to pressure. In such cases the wall may consist of smooth fibrous tissue. Rarely, true cysts are complicated by infection or hemorrhage. They may be detected incidentally and are commonly associated with von HippelLindau disease. CT and MRI features reveal a well-defined unilocular cystic lesion with indistinct walls.96,232

von Hippel-Lindau Disease.  von Hippel-Lindau disease is a hereditary disease with a dominant trait. It is associated with a number of abnormalities, including angiomas of the central nervous system and eye and neoplasms and cysts of the pancreas, kidney, adrenal glands, and epididymis. Pancreatic involvement consists of cysts, serous cystadenomas, solid nonfunctional islet cell tumors,  and adenocarcinomas; calcifications are common (Figs. 46-93 and  46-94). Progression of the disease can result in steatorrhea and diabetes.44

1468

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 46-92  Axial contrast-enhanced MDCT (A) and the corresponding coronal reformatted image (B) reveal a well-defined irregular fluid-density lesion (arrow) in relation to the body and tail of the pancreas, with subtle internal septations. This lesion was histopathologically confirmed to be lymphangioma. Cross-sectional imaging helps by defining the exact anatomic location and the content of the lesion preoperatively.

B

A

FIG 46-93  von Hippel-Lindau disease. A and B, Axial T2-weighted MRIs reveal lobulated cystic lesions (arrow) in the head, body, and tail of the pancreas, with multiple cysts in the kidneys bilaterally (arrowheads).

TRAUMA Acutely, pancreatic injuries may lead to death due to associated vascular involvement (portal vein, splenic vein, inferior vena cava); this is twice as likely with injuries to the pancreatic head versus the tail.270 Delayed morbidity and mortality are due to complications such as abscess, pseudocyst, fistula, pancreatitis, hemorrhage, sepsis, and multiorgan failure. In addition, because surgical intervention can result in serious and prolonged postoperative complications, the decision to manage a pancreatic injury surgically must be weighed against the potential hazards of nonsurgical management.247 Isolated pancreatic injuries are rare; associated injuries, especially to the liver, stomach,

duodenum, and spleen, occur in more than 90% of cases. The pancreas is vulnerable to crush injury from blunt trauma due to its impact against the adjacent vertebral column. Two thirds of pancreatic injuries occur in the pancreatic body, and the remainder are equally distributed among the head, neck, and tail.156 In adults, more than 75% of blunt injuries to the pancreas are due to motor vehicle collisions; bicycle injuries in children and child abuse in infants may result in pancreatic injuries.106 CT is routinely used as a first-line imaging study in acute trauma patients and can be helpful in defining injuries to the pancreas and associated complications.95 Direct signs of pancreatic injury include focal enlargement, laceration, and transection, which is seen as

CHAPTER 46  Pancreas

1469

F

FIG 46-96  Axial CT shows transection (arrow) of the pancreas in the FIG 46-94  Axial contrast-enhanced MDCT detects multiple cystic lesions (arrows) in the pancreas and left kidney in this case of von Hippel-Lindau disease.

region of the neck, resulting in decreased vascular enhancement and edema of the body and tail, with a difference in density compared with the head. A large amount of free fluid (F) is also noted.

B

A

FIG 46-95  Pancreatic laceration. Axial (A) and coronal reformatted (B) MDCT images show focal enlargement and nonenhancing hypoattenuated areas (arrow) involving the head and uncinate process of the pancreas in this patient with abdominal injury.

hypoattenuating areas on CT; a fracture line may also be present (Figs. 46-95 and 46-96). Fluid collections such as hematomas, pseudocysts, and abscesses are often seen communicating with the pancreas at the site of fracture or transection (Figs. 46-97 and 46-98). Peripancreatic fat stranding, hemorrhage, and fluid between the splenic vein and pancreas are useful secondary signs. Although CT does not always directly demonstrate the pancreatic duct, injury to the duct can be suggested by the degree of parenchymal injury95,270 (Table 46-16). MDCT is 91% sensitive and 91% specific for pancreatic ductal injury.247

The prognosis of pancreatic injuries is related primarily to the integrity of the pancreatic duct; thus any patient with pancreatic injuries on CT and a high clinical index of suspicion for ductal injury should be considered for MRCP to allow direct imaging of the pancreatic duct.95 The associated parenchymal injuries can be evaluated using fat-suppressed T1- and T2-weighted images.95 If imaging of the pancreatic duct is not adequate or an injury is detected that might be amenable to  stent placement, ERCP should be performed promptly95 (Figs. 46-99 and 46-100).

1470

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 46-97  A and B, Axial contrast-enhanced MDCT (A) and axial T2-weighted MRI (B) show a small amount of fluid (arrows in B) adjacent to the pancreatic head and neck in this patient with abdominal trauma. The fracture through the neck of the pancreas is well appreciated on MDCT (arrow in A), while MRI reveals duct discontinuity. C and D, ERCP confirms duct disruption with extravasation (arrow in C) and insertion of a therapeutic stent (arrow in D).

R

A

L

R

B FIG 46-98  Axial (A) and coronal curved reformatted (B) images reveal a chronic well-contained fluid collection adjacent to the neck of the pancreas and communicating with the main pancreatic duct (arrow) in this patient with abdominal trauma.

L

CHAPTER 46  Pancreas

1471

TABLE 46-16  CT Grading Scheme of

Abdominal injury—suspected pancreatic trauma

Pancreatic Injury270 CT—first line of investigation

Grade

Injury

A

Pancreatitis or superficial laceration (<50% of pancreatic thickness) Deep laceration of pancreatic tail (>50% of pancreatic thickness) Transection of pancreatic tail Deep laceration of pancreatic head Transection of pancreatic head

B1 Parenchymal injury

Ductal injury

B2 C D

Further evaluation

MRCP

ERCP

Imaging not adequate/ amenable to stent placement

FIG 46-99  Diagnostic approach to pancreatic trauma.

Pancreatic lesion/disease suspected clinically or on US

MDCT • Evaluation of lesion • CT angiography • CT pancreatography • Preoperative planning/staging • PET CT—preop evaluation/staging

Further evaluation If contraindicated

MR/MR pancreatography • Evaluation of lesion • Preferred for characterization/ duct communication • MR angiography • MR pancreatography • Preoperative planning/staging

Is diagnosis made? No

Yes

• EUS/CT guided aspiration or biopsy • Secretin-enhanced MRCP/ERCP

• Conservative • Surgery • ERCP • Observation

FIG 46-100  Imaging approach to pancreatic lesions.

1472

PART II  CT and MR Imaging of the Whole Body

TABLE 46-17  Infiltrative, Metabolic, and Other Disorders Metabolic Disorder

Clinical/Pathologic

Imaging Appearance

Hemocromatosis

Results from excessive accumulation of iron in the body. The disease has a primary idiopathic form, resulting from excessive absorption of iron, and a secondary form, resulting from excessive accumulation of iron from multiple transfusions.63

CT shows an increased density in the pancreas and adjacent peripancreatic lymph nodes, although no correlation has been found between that density and pancreatic dysfunction or insufficiency.1152 Unopacified vessels may give the spurious impression of low-density masses in the pancreas, but this can be clarified by contrast studies. MRI reveals low signal intensity in the pancreas and other organs (liver, spleen) on T2-weighted GRE images, spin echo images being less sensitive.96

Cystic Fibrosis

Life-threatening autosomal recessive disease, results in exocrine pancreatic insufficiency in the majority of patients; only 10% to 15% of patients with cystic fibrosis retain sufficient pancreatic function to prevent steatorrhea.71 Although the course and prognosis of the disease are determined mainly by the progression of pulmonary disease, pancreatic insufficiency may result in earlier colonization of the airway by Pseudomonas aeruginosa.71 Exocrine pancreatic insufficiency may not become clinically apparent until 98% to 99% of the parenchyma undergoes destruction.250 Before total or near-total fatty replacement of the pancreas occurs, there is no correlation between histologic findings and the results of pancreatic function tests. The degree of fatty replacement of the pancreas and atrophy is related to the severity of expression of cystic fibrosis.71,200,250

CT is highly sensitive for depicting fatty replacement (Fig. 46-95). The pancreas has an inhomogeneous appearance on CT, with fatty changes, tiny calcification, and areas of lower attenuation representing cysts.51 Complete fatty replacement of the pancreas represents a late phenomenon and is often seen in older patients; despite diffuse fatty changes, the shape of the pancreas is maintained.51 MRI can show partial fatty replacement of the pancreas with focal areas of sparing; a lobulated, enlarged pancreas that appears hyperintense on T1- and T2-weighted images owing to nearly complete lipomatous replacement (lipomatous pseudohypertrophy); an atrophied fatty pancreas with variable degrees of fibrofatty replacement, seen as a small pancreas with inhomogeneous signal intensity; an atrophied pancreas without fatty replacement; and pancreatic cystosis, which reflects cyst formation due to inspissated secretions.71,72,169,183,250 In addition, the quantitative assessment of reduced T1 values secondary to fatty infiltration improves the overall sensitivity of MRI for the early diagnosis of fatty degeneration of the pancreas and hence allows the earlier detection of pancreatic compromise, before it is visually apparent on imaging.71 The main disadvantage of MRI is the inability to show calcification.250

Fatty

Pancreatic lipomatosis or fatty replacement of the pancreas is a rare benign disease. Fatty replacement may present as reticular patches or multiple tiny nodules throughout the pancreas; sometimes it presents as a focal fatty mass that may resemble a pancreatic neoplasm (Fig. 46-96).111,168 It is associated with a variety of diseases, including obesity, diabetes mellitus, chronic pancreatitis, hereditary pancreatitis, obstruction of the pancreatic duct by calculus or tumor, and cystic fibrosis.202

The involved gland appears hypodense on CT. A mild degree of focal fatty replacement may not be diagnosed with CT alone. Moreover, contrastenhanced CT cannot show the negative attenuation value of mild fatty replacement because normal pancreatic parenchyma trapped between fatty areas exhibits contrast enhancement. Also, a focal fatty mass may be mistaken for a true pancreatic neoplasm on CT.128 MRI may be helpful in confirming the presence of focal fatty replacement of the pancreas, especially opposed-phase chemical-shift MR images that reveal a reduction in signal intensity.113,128

Heterotopic

Heterotopic pancreas is defined as pancreatic tissue that lacks anatomic and vascular continuity with the main body of the pancreas.59 The heterotopic pancreatic tissue has all the normal pancreatic elements, including pancreatic acini, pancreatic ducts, and islets of Langerhans, at histologic examination. Most often, heterotopic pancreas is found in the stomach, duodenum, or upper part of the jejunum, although it may be found at several other locations.127

Heterotopic pancreas on upper gastrointestinal examination consist of a small, sharply marginated, round or oval, broad-based submucosal mass lesion that may extend into the muscularis and serosa, similar to any other gastrointestinal tumor; thus, it may be difficult to differentiate it from other submucosal lesions on imaging.41,127 Endoscopic biopsy should be performed before surgical resection, and a more conservative, nonsurgical approach should be considered in asymptomatic patients.41 Occasionally, complications such as pancreatitis, pseudocyst, cyst formation, insulinoma, adenoma, and malignant transformation can occur.45,59,127,255

CHAPTER 46  Pancreas

REFERENCES 1. Agrawal MD, Pinho DF, Kulkarni NM, et al: Oncologic applications of dual-energy CT in the abdomen. Radiographics 34(3):589–612, 2014. 2. Ahlstrom H, Eriksson B, Bergstrom M, et al: Pancreatic neuroendocrine tumors: Diagnosis with PET. Radiology 195(2):333–337, 1995. 3. Ahrenholz DH, Simmons RL: Fibrin in peritonitis. I. Beneficial and adverse effects of fibrin in experimental E. coli peritonitis. Surgery 88(1):41–47, 1980. 4. Akhan O, Pringot J: Imaging of abdominal tuberculosis. Eur Radiol 12(2):312–323, 2002. 5. Alavi A, Lakhani P, Mavi A, et al: PET: A revolution in medical imaging. Radiol Clin North Am 42(6):983–1001, vii, 2004. 6. Allen PJ, Jaques DP, D’Angelica M, et al: Cystic lesions of the pancreas: Selection criteria for operative and nonoperative management in 209 patients. J Gastrointest Surg 7(8):970–977, 2003. 7. Balthazar EJ: Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 223(3):603–613, 2002. 8. Balthazar EJ, Freeny PC, van Sonnenberg E: Imaging and intervention in acute pancreatitis. Radiology 193(2):297–306, 1994. 9. Balthazar EJ, Ranson JH, Naidich DP, et al: Acute pancreatitis: Prognostic value of CT. Radiology 156(3):767–772, 1985. 10. Balthazar EJ, Robinson DL, Megibow AJ, et al: Acute pancreatitis: Value of CT in establishing prognosis. Radiology 174(2):331–336, 1990. 11. Banks PA: A new classification system for acute pancreatitis. Am J Gastroenterol 89(2):151–152, 1994. 12. Bares R, Klever P, Hauptmann S, et al: F-18 fluorodeoxyglucose PET in vivo evaluation of pancreatic glucose metabolism for detection of pancreatic cancer. Radiology 192(1):79–86, 1994. 13. Bares R, Klever P, Hellwig D, et al: Pancreatic cancer detected by positron emission tomography with 18F-labelled deoxyglucose: Method and first results. Nucl Med Commun 14(7):596–601, 1993. 14. Barish MA, Soto JA: MR cholangiopancreatography: Techniques and clinical applications. AJR Am J Roentgenol 169(5):1295–1303, 1997. 15. Bassi C, Procacci C, Zamboni G, et al: Intraductal papillary mucinous tumors of the pancreas. Verona University Pancreatic Team. Int J Pancreatol 27(3):181–193, 2000. 16. Basterra G, Alvarez M, Marcaide A, et al: Acute pancreatitis: Evaluation of the prognostic criteria of the latest Balthazar tomographic classification. Rev Esp Enferm Dig 91(6):433–438, 1999. 17. Baum U, Lell M, Nomayr A, et al: [Multiplanar spiral CT in the diagnosis of pancreatic tumors]. Radiologe 39(11):958–964, 1999. 18. Becker V, Mischke U: Groove pancreatitis. Int J Pancreatol 10(3–4):173– 182, 1991. 19. Beger HG, Rau B, Mayer J, et al: Natural course of acute pancreatitis. World J Surg 21(2):130–135, 1997. 20. Beger HG, Schlosser W, Siech M, et al: The surgical management of chronic pancreatitis: Duodenum-preserving pancreatectomy. Adv Surg 32:87–104, 1999. 21. Biankin AV, Kench JG, Dijkman FP, et al: Molecular pathogenesis of precursor lesions of pancreatic ductal adenocarcinoma. Pathology 35(1):14–24, 2003. 22. Bipat S, Phoa SS, van Delden OM, et al: Ultrasonography, computed tomography and magnetic resonance imaging for diagnosis and determining resectability of pancreatic adenocarcinoma: A metaanalysis. J Comput Assist Tomogr 29(4):438–445, 2005. 23. Birchard KR, Semelka RC, Hyslop WB, et al: Suspected pancreatic cancer: Evaluation by dynamic gadolinium-enhanced 3D gradient-echo MRI. AJR Am J Roentgenol 185(3):700–703, 2005. 24. Bittner R, Block S, Buchler M, et al: Pancreatic abscess and infected pancreatic necrosis. Different local septic complications in acute pancreatitis. Dig Dis Sci 32(10):1082–1087, 1987. 25. Blodgett TM, Meltzer CC, Townsend DW: PET/CT: Form and function. Radiology 242(2):360–385, 2007. 26. Bodurtha AJ, Dajee H, You CK: Analysis of 29 cases of pancreatic pseudocyst treated surgically. Can J Surg 23(5):432–434, 1980. 27. Bosmans H, Van Hoe L, Gryspeerdt S, et al: Single-shot T2-weighted   MR imaging of the upper abdomen: Preliminary experience with

28. 29.

30. 31.

32. 33. 34. 35.

36.

37. 38.

39.

40.

41. 42.

43.

44.

45.

46. 47. 48.

49.

50. 51.

52.

1473

double-echo HASTE technique. AJR Am J Roentgenol 169(5):1291– 1293, 1997. Bradley EL, 3rd: A clinically based classification system for acute pancreatitis. Ann Chir 47(6):537–541, 1993. Bradley EL, 3rd: A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, GA, September 11 through 13, 1992. Arch Surg 128(5):586–590, 1993. Brandt KR, Charboneau JW, Stephens DH, et al: CT- and US-guided biopsy of the pancreas. Radiology 187(1):99–104, 1993. Brook OR, Brook A, Vollmer CM, et al: Structured reporting of multiphasic CT for pancreatic cancer: Potential effect on staging and surgical planning. Radiology 274(2):464–472, 2015. Brugge WR, Lauwers GY, Sahani D, et al: Cystic neoplasms of the pancreas. N Engl J Med 351(12):1218–1226, 2004. Buchino JJ: Fine-needle aspiration of solid and papillary cystic tumor of the pancreas. Pediatr Pathol Lab Med 16(2):235–242, 1996. Burkett JS, Pickleman J: The rationale for surgical treatment of mesenteric and retroperitoneal cysts. Am Surg 60(6):432–435, 1994. Cantisani V, Mortele KJ, Levy A, et al: MR imaging features of solid pseudopapillary tumor of the pancreas in adult and pediatric patients. AJR Am J Roentgenol 181(2):395–401, 2003. Cardenes HR, Chiorean EG, Dewitt J, et al: Locally advanced pancreatic cancer: Current therapeutic approach. Oncologist 11(6):612–623, 2006. Chang WT, Lee KT, Yang SF: Cavernous hemangioma of the pancreas: Report of a case. Pancreas 26(3):310–312, 2003. Chari ST: Chronic pancreatitis: classification, relationship to acute pancreatitis, and early diagnosis. J Gastroenterol 42(Suppl 17):58–59, 2007. Chari ST, Singer MV: The problem of classification and staging of chronic pancreatitis. Proposals based on current knowledge of its natural history. Scand J Gastroenterol 29(10):949–960, 1994. Chiou YY, Chiang JH, Hwang JI, et al: Acinar cell carcinoma of the pancreas: Clinical and computed tomography manifestations. J Comput Assist Tomogr 28(2):180–186, 2004. Cho JS, Shin KS, Kwon ST, et al: Heterotopic pancreas in the stomach: CT findings. Radiology 217(1):139–144, 2000. Choi BI, Kim KW, Han MC, et al: Solid and papillary epithelial neoplasms of the pancreas: CT findings. Radiology 166(2):413–416, 1988. Choi JY, Kim MJ, Kim JH, et al: Solid pseudopapillary tumor of the pancreas: Typical and atypical manifestations. AJR Am J Roentgenol 187(2):W178–W186, 2006. Choyke PL, Filling-Katz MR, Shawker TH, et al: von Hippel-Lindau disease: Radiologic screening for visceral manifestations. Radiology 174(3 Pt 1):815–820, 1990. Claudon M, Verain AL, Bigard MA, et al: Cyst formation in gastric heterotopic pancreas: Report of two cases. Radiology 169(3):659–660, 1988. Coleman KM, Doherty MC, Bigler SA: Solid-pseudopapillary tumor of the pancreas. Radiographics 23(6):1644–1648, 2003. Crowson MC, Perry M, Burden E: Tuberculosis of the pancreas: A rare cause of obstructive jaundice. Br J Surg 71(3):239, 1984. Curry CA, Eng J, Horton KM, et al: CT of primary cystic pancreatic neoplasms: Can CT be used for patient triage and treatment? AJR Am J Roentgenol 175(1):99–103, 2000. Czako L: Diagnosis of early-stage chronic pancreatitis by secretinenhanced magnetic resonance cholangiopancreatography. J Gastroenterol 42(Suppl 17):113–117, 2007. Daltrey IR, Johnson CD: Cystic lymphangioma of the pancreas. Postgrad Med J 72(851):564–566, 1996. Daneman A, Gaskin K, Martin DJ, et al: Pancreatic changes in cystic fibrosis: CT and sonographic appearances. AJR Am J Roentgenol 141(4):653–655, 1983. Davenport M, Howard ER: Acute pancreatitis associated with congenital anomalies. In Beger HG, Warshaw AL, Buchler MW, et al: The pancreas, Oxford, 1998, Blackwell Science, pp 343–353.

1474

PART II  CT and MR Imaging of the Whole Body

53. Dervenis C, Johnson CD, Bassi C, et al: Diagnosis, objective assessment of severity, and management of acute pancreatitis. Santorini consensus conference. Int J Pancreatol 25(3):195–210, 1999. 54. Deshpande V, Gupta R, Sainani N, et al: Subclassification of autoimmune pancreatitis: A histologic classification with clinical significance. Am J Surg Pathol 35(1):26–35, 2011. 55. de Vos tot Nederveen Cappel WH, Lagendijk MA, Lamers CB, et al: Surveillance for familial pancreatic cancer. Scand J Gastroenterol Suppl 239:94–99, 2003. 56. DeWitt J, Devereaux B, Chriswell M, et al: Comparison of endoscopic ultrasonography and multidetector computed tomography for   detecting and staging pancreatic cancer. Ann Intern Med 141(10):753– 763, 2004. 57. Dixon E, Pasieka JL: Functioning and nonfunctioning neuroendocrine tumors of the pancreas. Curr Opin Oncol 19(1):30–35, 2007. 58. Dodge JA: Paediatric and hereditary aspects of chronic pancreatitis. Digestion 59(Suppl 4):49–59, 1998. 59. Dolan RV, ReMine WH, Dockerty MB: The fate of heterotopic pancreatic tissue. A study of 212 cases. Arch Surg 109(6):762–765, 1974. 60. Dong PR, Lu DS, Degregario F, et al: Solid and papillary neoplasm of the pancreas: Radiological-pathological study of five cases and review of the literature. Clin Radiol 51(10):702–705, 1996. 61. Dreiling DA, Messer J: The secretin story: A saga in clinical medicine and gastrointestinal physiology. Am J Gastroenterol 70(5):455–479, 1978. 62. England RJ, Woodley H, Cullinane C, et al: Pediatric pancreatic hemangioma: A case report and literature review. JOP 7(5):496–501, 2006. 63. Erdman RA, Vandersall JH: Effect of rumen protein degradability on milk yield of dairy cows in early lactation. J Dairy Sci 66(9):1873–1880, 1983. 64. Erturk SM, Ichikawa T, Motosugi U, et al: Diffusion-weighted MR imaging in the evaluation of pancreatic exocrine function before and after secretin stimulation. Am J Gastroenterol 101(1):133–136, 2006. 65. Etemad B, Whitcomb DC: Chronic pancreatitis: Diagnosis, classification, and new genetic developments. Gastroenterology 120(3):682–707, 2001. 66. Evans JD, Hamanaka Y, Olliff SP, et al: Tuberculosis of the pancreas presenting as metastatic pancreatic carcinoma. A case report and review of the literature. Dig Surg 17(2):183–187, 2000. 67. Faure JP, Tuech JJ, Richer JP, et al: Pancreatic metastasis of renal cell carcinoma: Presentation, treatment and survival. J Urol 165(1):20–22, 2001. 68. Fawcitt RA, Forbes WS, Isherwood I, et al: Computed tomography in pancreatic disease. Br J Radiol 51(601):1–4, 1978. 69. Federle MP, Jeffrey RB, Crass RA, et al: Dalsem: Computed tomography of pancreatic abscesses. AJR Am J Roentgenol 136(5):879–882, 1981. 70. Fernandez-del Castillo C, Targarona J, Thayer SP, et al: Incidental pancreatic cysts: Clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg 138(4): 427–423,2003. 71. Ferrozzi F, Bova D, Campodonico F, et al: Cystic fibrosis: MR assessment of pancreatic damage. Radiology 198(3):875–879, 1996. 72. Fiel SB, Friedman AC, Caroline DF, et al: Magnetic resonance imaging in young adults with cystic fibrosis. Chest 91(2):181–184, 1987. 73. Finkelberg DL, Sahani D, Deshpande V, et al: Autoimmune pancreatitis. N Engl J Med 355(25):2670–2676, 2006. 74. Fischer G, Spengler U, Neubrand M, et al: Isolated tuberculosis of the pancreas masquerading as a pancreatic mass. Am J Gastroenterol 90(12):2227–2230, 1995. 75. Fishman EK, Horton KM, Urban BA: Multidetector CT angiography   in the evaluation of pancreatic carcinoma: Preliminary observations.   J Comput Assist Tomogr 24(6):849–853, 2000. 76. Fletcher JG, Wiersema MJ, Farrell MA, et al: Pancreatic malignancy: Value of arterial, pancreatic, and hepatic phase imaging with multidetector row CT. Radiology 229(1):81–90, 2003. 77. Freene FL, Page D, Felming ID, et al: AJCC cancer staging manual, ed 6, New York, 2002, Springer-Verlag.

78. Freeny PC: Radiology of the pancreas. Curr Opin Radiol 4(3):54–61, 1992. 79. Freeny PC: Pancreatic imaging. New modalities. Gastroenterol Clin North Am 28(3):723–746, 1999. 80. Fukukura Y, Fujiyoshi F, Sasaki M, et al: Intraductal papillary mucinous tumors of the pancreas: Thin-section helical CT findings. AJR Am J Roentgenol 174(2):441–447, 2000. 81. Fukukura Y, Fujiyoshi F, Sasaki M, et al: Pancreatic duct: Morphologic evaluation with MR cholangiopancreatography after secretin stimulation. Radiology 222(3):674–680, 2002. 82. Fukumori K, Shakado S, Miyahara T, et al: Atypical manifestations of pancreatitis with autoimmune phenomenon in an adolescent female. Intern Med 44(8):886–891, 2005. 83. Fukuoka K, Ajiki T, Yamamoto M, et al: Complete agenesis of the dorsal pancreas. J Hepatobiliary Pancreat Surg 6(1):94–97, 1999. 84. Fulcher AS, Turner MA: MR pancreatography: A useful tool for evaluating pancreatic disorders. Radiographics 19(1):5–24, discussion 41–44; quiz 148–149, 1999. 85. Fulcher AS, Turner MA, Capps GW, et al: Half-Fourier RARE MR cholangiopancreatography: Experience in 300 subjects. Radiology 207(1):21–32, 1998. 86. Furukawa N, Muranaka T, Yasumori K, et al: Autoimmune pancreatitis: Radiologic findings in three histologically proven cases. J Comput Assist Tomogr 22(6):880–883, 1998. 87. Gambill EE: Pancreatitis associated with pancreatic carcinoma: A study of 26 cases. Mayo Clin Proc 46(3):174–177, 1971. 88. Gangi S, Fletcher JG, Nathan MA, et al: Time interval between abnormalities seen on CT and the clinical diagnosis of pancreatic cancer: Retrospective review of CT scans obtained before diagnosis. AJR Am J Roentgenol 182(4):897–903, 2004. 89. Geenen JE, Hogan WJ, Dodds WJ, et al: Intraluminal pressure recording from the human sphincter of Oddi. Gastroenterology 78(2):317–324, 1980. 90. Ghavamian R, Klein KA, Stephens DH, et al: Renal cell carcinoma metastatic to the pancreas: Clinical and radiological features. Mayo Clin Proc 75(6):581–585, 2000. 91. Giovannini M, Seitz JF, Monges G, et al: Fine-needle aspiration cytology guided by endoscopic ultrasonography: Results in 141 patients. Endoscopy 27(2):171–177, 1995. 92. Gohde SC, Toth J, Krestin GP, et al: Dynamic contrast-enhanced FMPSPGR of the pancreas: Impact on diagnostic performance. AJR Am J Roentgenol 168(3):689–696, 1997. 93. Graser A, Johnson TR, Chandarana H, et al: Dual energy CT: Preliminary observations and potential clinical applications in the abdomen. Eur Radiol 19(1):13–23, 2009. 94. Grimison PS, Chin MT, Harrison ML, et al: Primary pancreatic lymphoma—Pancreatic tumours that are potentially curable without resection, a retrospective review of four cases. BMC Cancer 6:117, 2006. 95. Gupta A, Stuhlfaut JW, Fleming KW, et al: Blunt trauma of the pancreas and biliary tract: A multimodality imaging approach to diagnosis. Radiographics 24(5):1381–1395, 2004. 96. Haaga JR, Lanzieri CF, Gilkeson RC: The pancreas. In Haaga JR, editor: Computed tomography and magnetic resonance imaging of the whole body, ed 4, Philadelphia, 2003, Elsevier. 97. Hashimoto M, Miki K, Kokudo N, et al: A long-term survivor of metastatic acinar cell carcinoma. Pancreas 34(2):271–272, 2007. 98. Hawkins RA: Pancreatic tumors: Imaging with PET. Radiology 195(2):320–322, 1995. 99. Heinrich S, Goerres GW, Schafer M, et al: Positron emission tomography/computed tomography influences on the management of resectable pancreatic cancer and its cost-effectiveness. Ann Surg 242(2):235–243, 2005. 100. Hellerhoff KJ, Helmberger H, 3rd, Rosch T, et al: Dynamic MR pancreatography after secretin administration: Image quality and diagnostic accuracy. AJR Am J Roentgenol 179(1):121–129, 2002. 101. Henderson JM, Macdonald JA: Fistula formation complicating pancreatic abscess. Br J Surg 63(3):233–234, 1976.

CHAPTER 46  Pancreas 102. Henzel JH, Pories WJ, Burget DE, et al: Intra-abdominal lymphangiomata. Arch Surg 93(2):304–308, 1966. 103. Hines OJ, Reber HA: Pancreatic surgery. Curr Opin Gastroenterol 22(5):520–526, 2006. 104. Horwhat JD, Paulson EK, McGrath K, et al: A randomized comparison of EUS-guided FNA versus CT or US-guided FNA for the evaluation of pancreatic mass lesions. Gastrointest Endosc 63(7):966–975, 2006. 105. Ichikawa T, Peterson MS, Federle MP, et al: Islet cell tumor of the pancreas: Biphasic CT versus MR imaging in tumor detection. Radiology 216(1):163–171, 2000. 106. Ilahi O, Bochicchio GV, Scalea TM: Efficacy of computed tomography in the diagnosis of pancreatic injury in adult blunt trauma patients: A single-institutional study. Am Surg 68(8):704–707, discussion 707–708, 2002. 107. Irie H, Honda H, Aibe H, et al: MR cholangiopancreatographic differentiation of benign and malignant intraductal mucin-producing tumors of the pancreas. AJR Am J Roentgenol 174(5):1403–1408, 2000. 108. Irie H, Honda H, Baba S, et al: Autoimmune pancreatitis: CT and MR characteristics. AJR Am J Roentgenol 170(5):1323–1327, 1998. 109. Ishikawa O, Ohhigashi H, Sasaki Y, et al: Practical usefulness of lymphatic and connective tissue clearance for the carcinoma of the pancreas head. Ann Surg 208(2):215–220, 1988. 110. Isserow JA, Siegelman ES, Mammone J: Focal fatty infiltration of the pancreas: MR characterization with chemical shift imaging. AJR Am J Roentgenol 173(5):1263–1265, 1999. 111. Itai Y, Saida Y, Kurosaki Y, et al: Focal fatty masses of the pancreas. Acta Radiol 36(2):178–181, 1995. 112. Itoh S, Yamakawa K, Shimamoto K, et al: CT findings in groove pancreatitis: Correlation with histopathological findings. J Comput Assist Tomogr 18(6):911–915, 1994. 113. Jacobs JE, Coleman BG, Arger PH, et al: Pancreatic sparing of focal fatty infiltration. Radiology 190(2):437–439, 1994. 114. Jhala NC, Jhala D, Eltoum I, et al: Endoscopic ultrasound-guided fine-needle aspiration biopsy: A powerful tool to obtain samples from small lesions. Cancer 102(4):239–246, 2004. 115. Johnson CD: Timing of intervention in acute pancreatitis. Postgrad Med J 69(813):509–515, 1993. 116. Johnson CD, Stephens DH, Charboneau JW, et al: Cystic pancreatic tumors: CT and sonographic assessment. AJR Am J Roentgenol 151(6):1133–1138, 1988. 117. Kalra MK, Maher MM, Sahani DV, et al: Current status of imaging in pancreatic diseases. J Comput Assist Tomogr 26(5):661–675, 2002. 118. Kamisawa T, Okazaki K, Kawa S: Diagnostic criteria for autoimmune pancreatitis in Japan. World J Gastroenterol 14(32):4992–4994, 2008. 119. Kamisawa T, Tu Y, Egawa N, et al: MRCP of congenital pancreaticobiliary malformation. Abdom Imaging 32(1):129–133, 2007. 120. Karasawa E, Goldberg HI, Moss AA, et al: CT pancreatogram in carcinoma of the pancreas and chronic pancreatitis. Radiology 148(2):489–493, 1983. 121. Kattwinkel J, Fleming D, Cha CC, et al: A device for administration of continuous positive airway pressure by the nasal route. Pediatrics 52(1):131–134, 1973. 122. Katz DS, Nardi PM, Hines J, et al: Lipomas of the pancreas. AJR Am J Roentgenol 170(6):1485–1487, 1998. 123. Katz MH, Savides TJ, Moossa AR, et al: An evidence-based approach to the diagnosis and staging of pancreatic cancer. Pancreatology 5(6):576– 590, 2005. 124. Kehagias D, Smyrniotis V, Gouliamos A, et al: Cystic pancreatic neoplasms: Computed tomography and magnetic resonance imaging findings. Int J Pancreatol 28(3):223–230, 2000. 125. Kettritz U, Warshauer DM, Brown ED, et al: Enhancement of the normal pancreas: Comparison of manganese-DPDP and gadolinium chelate. Eur Radiol 6(1):14–18, 1996. 126. Khalid A, Peterson M, Slivka A: Secretin-stimulated magnetic resonance pancreaticogram to assess pancreatic duct outflow obstruction in evaluation of idiopathic acute recurrent pancreatitis: A pilot study. Dig Dis Sci 48(8):1475–1481, 2003.

1475

127. Kilman WJ, Berk RN: The spectrum of radiographic features of aberrant pancreatic rests involving the stomach. Radiology 123(2):291– 296, 1977. 128. Kim HJ, Byun JH, Park SH, et al: Focal fatty replacement of the pancreas: Usefulness of chemical shift MRI. AJR Am J Roentgenol 188(2):429–432, 2007. 129. Kim JK, Ha HK, Han DJ, et al: CT analysis of postoperative tumor recurrence patterns in periampullary cancer. Abdom Imaging 28(3):384– 391, 2003. 130. Klein SD, Affronti JP: Pancreas divisum, an evidence-based review: Part I, pathophysiology. Gastrointest Endosc 60(3):419–425, 2004. 131. Klimstra DS, Heffess CS, Oertel JE, et al: Acinar cell carcinoma of the pancreas. A clinicopathologic study of 28 cases. Am J Surg Pathol 16(9):815–837, 1992. 132. Klimstra DS, Wenig BM, Heffess CS: Solid-pseudopapillary tumor of the pancreas: A typically cystic carcinoma of low malignant potential. Semin Diagn Pathol 17(1):66–80, 2000. 133. Kloppel G, Detlefsen S, Feyerabend B: Fibrosis of the pancreas: The initial tissue damage and the resulting pattern. Virchows Arch 445(1):1–8, 2004. 134. Kloppel G, Luttges J: WHO-classification 2000: Exocrine pancreatic tumors. Verh Dtsch Ges Pathol 85:219–228, 2001. 135. Kobayashi H, Itoh T, Murata R, et al: Pancreatic cavernous hemangioma: CT, MRI, US, and angiography characteristics. Gastrointest Radiol 16(4):307–310, 1991. 136. Kolmannskog F, Swensen T, Vatn MH, et al: Computed tomography and ultrasound of the normal pancreas. Acta Radiol Diagn (Stockh) 23(5):443–451, 1982. 137. Kostakoglu L, Agress H, Jr, Goldsmith SJ: Clinical role of FDG PET in evaluation of cancer patients. Radiographics 23(2):315–340, quiz 533, 2003. 138. Kourtesis G, Wilson SE, Williams RA: The clinical significance of fluid collections in acute pancreatitis. Am Surg 56(12):796–799, 1990. 139. Koyama K, Okamura T, Kawabe J, et al: Diagnostic usefulness of FDG PET for pancreatic mass lesions. Ann Nucl Med 15(3):217–224, 2001. 140. Kreel L, Sandin B: Changes in pancreatic morphology associated with aging. Gut 14(12):962–970, 1973. 141. Ladas SD, Vaidakis E, Lariou C, et al: Pancreatic tuberculosis in non-immunocompromised patients: Reports of two cases, and a literature review. Eur J Gastroenterol Hepatol 10(11):973–976, 1998. 142. Lam KY, Lo CY, Fan ST: Pancreatic solid-cystic-papillary tumor: Clinicopathologic features in eight patients from Hong Kong and review of the literature. World J Surg 23(10):1045–1050, 1999. 143. Laugier R: Dynamic endoscopic manometry of the response to secretin in patients with chronic pancreatitis. Endoscopy 26(2):222–227, 1994. 144. Law CH, Wei AC, Hanna SS, et al: Pancreatic resection for metastatic renal cell carcinoma: Presentation, treatment, and outcome. Ann Surg Oncol 10(8):922–926, 2003. 145. Lee LK, Sahani DV: Autoimmune pancreatitis in the context of IgG4-related disease: Review of imaging findings. World J Gastroenterol 20(41):15177–15189, 2014. 146. Legmann P, Vignaux O, Dousset B, et al: Pancreatic tumors: Comparison of dual-phase helical CT and endoscopic sonography. AJR Am J Roentgenol 170(5):1315–1322, 1998. 147. Lesniak RJ, Hohenwalter MD, Taylor AJ: Spectrum of causes of pancreatic calcifications. AJR Am J Roentgenol 178(1):79–86, 2002. 148. Li H, Zeng MS, Zhou KR, et al: Pancreatic adenocarcinoma: The different CT criteria for peripancreatic major arterial and venous invasion. J Comput Assist Tomogr 29(2):170–175, 2005. 149. Lin A, Feller ER: Pancreatic carcinoma as a cause of unexplained pancreatitis: Report of ten cases. Ann Intern Med 113(2):166–167, 1990. 150. Lingareddy S, Duvvuru NR, Guduru VR, et al: Dorsal agenesis of pancreas: CT and ERCP. Gastrointest Endosc 65(1):157–158, discussion 158, 2007. 151. Loftus EV, Jr, Olivares-Pakzad BA, Batts KP, et al: Intraductal papillarymucinous tumors of the pancreas: Clinicopathologic features, outcome, and nomenclature. Members of the Pancreas Clinic, and Pancreatic Surgeons of Mayo Clinic. Gastroenterology 110(6):1909–1918, 1996.

1476

PART II  CT and MR Imaging of the Whole Body

152. Long JA, Jr, Doppman JL, Nienhus AW, et al: Computed tomographic analysis of beta-thalassemic syndromes with hemochromatosis: Pathologic findings with clinical and laboratory correlations. J Comput Assist Tomogr 4(2):159–165, 1980. 153. Lu DS, Reber HA, Krasny RM, et al: Local staging of pancreatic cancer: Criteria for unresectability of major vessels as revealed by pancreaticphase, thin-section helical CT. AJR Am J Roentgenol 168(6):1439–1443, 1997. 154. Luetmer PH, Stephens DH, Ward EM: Chronic pancreatitis: Reassessment with current CT. Radiology 171(2):353–357, 1989. 155. Macari M, Spieler B, Kim D, et al: Dual-source dual-energy MDCT of pancreatic adenocarcinoma: Initial observations with data generated at 80 kVp and at simulated weighted-average 120 kVp. AJR Am J Roentgenol 194(1):W27–W32, 2010. 156. Madiba TE, Mokoena TR: Favourable prognosis after surgical drainage of gunshot, stab or blunt trauma of the pancreas. Br J Surg 82(9):1236– 1239, 1995. 157. Madura JA, Wiebke EA, Howard TJ, et al: Mucin-hypersecreting intraductal neoplasms of the pancreas: A precursor to cystic pancreatic malignancies. Surgery 122(4):786–792, discussion 792–793, 1997. 158. Manfredi R, Costamagna G, Brizi MG, et al: Severe chronic pancreatitis versus suspected pancreatic disease: Dynamic MR cholangiopancreatography after secretin stimulation. Radiology 214(3):849–855, 2000. 159. Marchal G, Baert AL, Wilms G: Intravenous pancreaticography in computed tomography. J Comput Assist Tomogr 3(6):727–732, 1979. 160. Marin D, Nelson RC, Barnhart H, et al: Detection of pancreatic tumors, image quality, and radiation dose during the pancreatic parenchymal phase: Effect of a low-tube-voltage, high-tube-current CT technique– Preliminary results. Radiology 256(2):450–459, 2010. 161. Martin DR, Semelka RC: MR imaging of pancreatic masses. Magn Reson Imaging Clin N Am 8(4):787–812, 2000. 162. Masui T, Takehara Y, Ichijo K, et al: Evaluation of the pancreas: A comparison of single thick-slice MR cholangiopancreatography with multiple thin-slice volume reconstruction MR cholangiopancreatography. AJR Am J Roentgenol 173(6):1519–1526, 1999. 163. Matos C, Cappeliez O, Winant C, et al: MR imaging of the pancreas:   A pictorial tour. Radiographics 22(1):e2, 2002. 164. Matos C, Metens T, Deviere J, et al: Pancreas divisum: Evaluation with secretin-enhanced magnetic resonance cholangiopancreatography. Gastrointest Endosc 53(7):728–733, 2001. 165. Matos C, Metens T, Deviere J, et al: Pancreatic duct: Morphologic and functional evaluation with dynamic MR pancreatography after secretin stimulation. Radiology 203(2):435–441, 1997. 166. Matos C, Winant C, Deviere J: Magnetic resonance pancreatography. Abdom Imaging 26(3):243–253, 2001. 167. Matsubayashi H, Kakushima N, Takizawa K, et al: Diagnosis of autoimmune pancreatitis. World J Gastroenterol 20(44):16559–16569, 2014. 168. Matsumoto S, Mori H, Miyake H, et al: Uneven fatty replacement   of the pancreas: evaluation with CT. Radiology 194(2):453–458, 1995. 169. McHugo JM, McKeown C, Brown MT, et al: Ultrasound findings in children with cystic fibrosis. Br J Radiol 60(710):137–141, 1987. 170. McNamara MM, Little MD, Alexander LF, et al: Multireader   evaluation of lesion conspicuity in small pancreatic adenocarcinomas: Complimentary value of iodine material density and low keV simulated monoenergetic images using multiphasic rapid kVp-switching dual energy CT. Abdom Imaging 40(5):1230–1240, 2014. 171. Megibow AJ: Pancreatic adenocarcinoma: Designing the examination to evaluate the clinical questions. Radiology 183(2):297–303, 1992. 172. Mehmet Erturk S, Ichikawa T, Sou H, et al: Pancreatic adenocarcinoma: MDCT versus MRI in the detection and assessment of locoregional extension. J Comput Assist Tomogr 30(4):583–590, 2006. 173. Mertz HR, Sechopoulos P, Delbeke D, et al: EUS, PET, and CT scanning for evaluation of pancreatic adenocarcinoma. Gastrointest Endosc 52(3):367–371, 2000.

174. Millbourn E: On the excretory ducts of the pancreas in man, with special reference to their relations to each other, to the common   bile duct and to the duodenum. Acta Anat (Basel) 9(1–2):1–34, 1950. 175. Minami M, Itai Y, Ohtomo K, et al: Cystic neoplasms of the pancreas: Comparison of MR imaging with CT. Radiology 171(1):53–56, 1989. 176. Mitchell DG, Winston CB, Outwater EK, et al: Delineation of pancreas with MR imaging: Multiobserver comparison of five pulse sequences.   J Magn Reson Imaging 5(2):193–199, 1995. 177. Mittendorf EA, Shifrin AL, Inabnet WB, et al: Islet cell tumors. Curr Probl Surg 43(10):685–765, 2006. 178. Morgan DE: Dual-energy CT of the abdomen. Abdom Imaging 39(1):108–134, 2014. 179. Morgan DE, Baron TH, Smith JK, et al: Pancreatic fluid collections prior to intervention: Evaluation with MR imaging compared with CT and US. Radiology 203(3):773–778, 1997. 180. Mortele KJ, Rocha TC, Streeter JL, et al: Multimodality imaging of pancreatic and biliary congenital anomalies. Radiographics 26(3):715– 731, 2006. 181. Muranaka T, Teshima K, Honda H, et al: Computed tomography and histologic appearance of pancreatic metastases from distant sources. Acta Radiol 30(6):615–619, 1989. 182. Nagai E, Ueki T, Chijiiwa K, et al: Intraductal papillary mucinous neoplasms of the pancreas associated with so-called “mucinous ductal ectasia”. Histochemical and immunohistochemical analysis of 29 cases. Am J Surg Pathol 19(5):576–589, 1995. 183. Nakamura M, Katada N, Sakakibara A, et al: Huge lipomatous pseudohypertrophy of the pancreas. Am J Gastroenterol 72(2):171–174, 1979. 184. Nayer H, Weir EG, Sheth S, et al: Primary pancreatic lymphomas: A cytopathologic analysis of a rare malignancy. Cancer 102(5):315–321, 2004. 185. Neff CC, Simeone JF, Wittenberg J, et al: Inflammatory pancreatic masses. Problems in differentiating focal pancreatitis from carcinoma. Radiology 150(1):35–38, 1984. 186. Ng IO, Fan ST, Nicholls J: Lymphangioma of the pancreas.   J Gastroenterol Hepatol 4(3):299–302, 1989. 187. Nicaise N, Pellet O, Metens T, et al: Magnetic resonance cholangiopancreatography: Interest of IV secretin administration in the evaluation of pancreatic ducts. Eur Radiol 8(1):16–22, 1998. 188. Nishihara K, Nagoshi M, Tsuneyoshi M, et al: Papillary cystic tumors of the pancreas. Assessment of their malignant potential. Cancer 71(1):82–92, 1993. 189. Nishiharu T, Yamashita Y, Abe Y, et al: Local extension of pancreatic carcinoma: Assessment with thin-section helical CT versus with breath-hold fast MR imaging–ROC analysis. Radiology 212(2):445–452, 1999. 190. Noone TC, Hosey J, Firat Z, et al: Imaging and localization of islet-cell tumours of the pancreas on CT and MRI. Best Pract Res Clin Endocrinol Metab 19(2):195–211, 2005. 191. O’Malley ME, Boland GW, Wood BJ, et al: Adenocarcinoma of the head of the pancreas: Determination of surgical unresectability with thin-section pancreatic-phase helical CT. AJR Am J Roentgenol 173(6):1513–1518, 1999. 192. Ogura T, Yamao K, Hara K, et al: Prognostic value of K-ras mutation status and subtypes in endoscopic ultrasound-guided fine-needle aspiration specimens from patients with unresectable pancreatic cancer. J Gastroenterol 48(5):640–646, 2013. 193. Ohtomo K, Furui S, Onoue M, et al: Solid and papillary epithelial neoplasm of the pancreas: MR imaging and pathologic correlation. Radiology 184(2):567–570, 1992. 194. Okazaki K, Chiba T: Autoimmune related pancreatitis. Gut 51(1):1–4, 2002. 195. Banks P: The role of needle-aspiration bacteriology in the   management of necrotizing pancreatitis. In Bradley EL, editor:   Acute pancreatitis: Diagnosis and therapy, ed 3, New York, 1994, Raven, pp 99–103.

CHAPTER 46  Pancreas 196. Paal E, Thompson LD, Heffess CS: A clinicopathologic and immunohistochemical study of ten pancreatic lymphangiomas and a review of the literature. Cancer 82(11):2150–2158, 1998. 197. Palazzo L, Roseau G, Gayet B, et al: Endoscopic ultrasonography in the diagnosis and staging of pancreatic adenocarcinoma. Results of a prospective study with comparison to ultrasonography and CT scan. Endoscopy 25(2):143–150, 1993. 198. Pamuklar E, Semelka RC: MR imaging of the pancreas. Magn Reson Imaging Clin N Am 13(2):313–330, 2005. 199. Pandit-Taskar N, Schoder H, Gonen M, et al: Clinical significance   of unexplained abnormal focal FDG uptake in the abdomen   during whole-body PET. AJR Am J Roentgenol 183(4):1143–1147, 2004. 200. Park RW, Grand RJ: Gastrointestinal manifestations of cystic fibrosis: A review. Gastroenterology 81(6):1143–1161, 1981. 201. Patel S, Bellon EM, Haaga J, et al: Fat replacement of the exocrine pancreas. AJR Am J Roentgenol 135(4):843–845, 1980. 202. Patel BN, Thomas JV, Lockhart ME, et al: Single-source dual-energy spectral multidetector CT of pancreatic adenocarcinoma: Optimization of energy level viewing significantly increases lesion contrast. Clin Radiol 68(2):148–154, 2013. 203. Pavai S, Yap SF: The clinical significance of elevated levels of serum CA 19-9. Med J Malaysia 58(5):667–672, 2003. 204. Plumley TF, Rohrmann CA, Freeny PC, et al: Double duct sign: reassessed significance in ERCP. AJR Am J Roentgenol 138(1):31–35, 1982. 205. Poston GJ, Townsend CM, Jr, Rajaraman S, et al: Effect of somatostatin and tamoxifen on the growth of human pancreatic cancers in nude mice. Pancreas 5(2):151–157, 1990. 206. Pour PM: The silent killer. Int J Pancreatol 10(2):103–104, 1991. 207. Prasad SR, Sahani D, Nasser S, et al: Intraductal papillary mucinous tumors of the pancreas. Abdom Imaging 28(3):357–365, 2003. 208. Procacci C, Graziani R, Bicego E, et al: Serous cystadenoma of the pancreas: Report of 30 cases with emphasis on the imaging findings.   J Comput Assist Tomogr 21(3):373–382, 1997. 209. Prokesch RW, Chow LC, Beaulieu CF, et al: Isoattenuating pancreatic adenocarcinoma at multi-detector row CT: Secondary signs. Radiology 224(3):764–768, 2002. 210. Pruessmann KP, Weiger M, Scheidegger MB, et al: SENSE: Sensitivity encoding for fast MRI. Magn Reson Med 42(5):952–962, 1999. 211. Psatha EA, Hyslop WB, Woosley JT, et al: Immunoblastic large B-cell lymphoma of the peripancreatic head region: MR findings. Magn Reson Imaging 22(7):1053–1057, 2004. 212. Quest L, Lombard M: Pancreas divisum: Opinio divisa. Gut 47(3):317– 319, 2000. 213. Ranson JH, Balthazar E, Caccavale R, et al: Computed tomography and the prediction of pancreatic abscess in acute pancreatitis. Ann Surg 201(5):656–665, 1985. 214. Renner IG, Savage WT, 3rd, Pantoja JL, et al: Death due to acute pancreatitis. A retrospective analysis of 405 autopsy cases. Dig Dis Sci 30(10):1005–1018, 1985. 215. Rohren EM, Turkington TG, Coleman RE: Clinical applications of PET in oncology. Radiology 231(2):305–332, 2004. 216. Rose DM, Delbeke D, Beauchamp RD, et al: 18Fluorodeoxyglucosepositron emission tomography in the management of patients with suspected pancreatic cancer. Ann Surg 229(5):729–737, discussion 737–738, 1999. 217. Ross BA, Jeffrey RB, Jr, Mindelzun RE: Normal variations in the lateral contour of the head and neck of the pancreas mimicking neoplasm: Evaluation with dual-phase helical CT. AJR Am J Roentgenol 166(4):799–801, 1996. 218. Ryan DP, Hong TS, Bardeesy N: Pancreatic adenocarcinoma. N Engl J Med 371(22):2140–2141, 2014. 219. Safrit HD, Rice RP: Gastrointestinal complications of pancreatitis. Radiol Clin North Am 27(1):73–79, 1989. 220. Sahani DV, Bonaffini PA, Catalano OA, et al: State-of-the-art PET/CT of the pancreas: Current role and emerging indications. Radiographics 32(4):1133–1158, discussion 1158–1160, 2012.

1477

221. Sahani DV, Kadavigere R, Blake M, et al: Intraductal papillary mucinous neoplasm of pancreas: Multi-detector row CT with 2D curved reformations—Correlation with MRCP. Radiology 238(2):560– 569, 2006. 222. Sahani DV, Kadavigere R, Saokar A, et al: Cystic pancreatic lesions: A simple imaging-based classification system for guiding management. Radiographics 25(6):1471–1484, 2005. 223. Sahani DV, Kalva SP, Farrell J, et al: Autoimmune pancreatitis: Imaging features. Radiology 233(2):345–352, 2004. 224. Saif MW: Primary pancreatic lymphomas. JOP 7(3):262–273, 2006. 225. Sarr MG, Murr M, Smyrk TC, et al: Primary cystic neoplasms of the pancreas. Neoplastic disorders of emerging importance—Current state-of-the-art and unanswered questions. J Gastrointest Surg 7(3):417–428, 2003. 226. Schima W, Fugger R, Schober E, et al: Diagnosis and staging of pancreatic cancer: Comparison of mangafodipir trisodium-enhanced MR imaging and contrast-enhanced helical hydro-CT. AJR Am J Roentgenol 179(3):717–724, 2002. 227. Schnedl WJ, Reisinger EC, Schreiber F, et al: Complete and partial agenesis of the dorsal pancreas within one family. Gastrointest Endosc 42(5):485–487, 1995. 228. Schueller G, Schima W, Schueller-Weidekamm C, et al:   Multidetector CT of pancreas: Effects of contrast material   flow rate and individualized scan delay on enhancement   of pancreas and tumor contrast. Radiology 241(2):441–448, 2006. 229. Semelka RC, Ascher SM: MR imaging of the pancreas. Radiology 188(3):593–602, 1993. 230. Semelka RC, Custodio CM, Cem Balci N, et al: Neuroendocrine tumors of the pancreas: Spectrum of appearances on MRI. J Magn Reson Imaging 11(2):141–148, 2000. 231. Sheehan MK, Beck K, Pickleman J, et al: Spectrum of cystic neoplasms of the pancreas and their surgical management. Arch Surg 138(6):657– 660, discussion 660–662, 2003. 232. Shirkhoda A, Mittelstaedt CA: Demonstration of pancreatic cysts in adult polycystic disease by computed tomography and ultrasound. AJR Am J Roentgenol 131(6):1074–1076, 1978. 233. Shultz S, Druy EM, Friedman AC: Common hepatic artery aneurysm: Pseudopseudocyst of the pancreas. AJR Am J Roentgenol 144(6):1287– 1288, 1985. 234. Siegelman SS, Copeland BE, Saba GP, et al: CT of fluid collections associated with pancreatitis. AJR Am J Roentgenol 134(6):1121–1132, 1980. 235. Singer MV, Gyr K, Sarles H: Revised classification of pancreatitis. Report of the Second International Symposium on the Classification of Pancreatitis in Marseille, France, March 28-30, 1984. Gastroenterology 89(3):683–685, 1985. 236. Soriano A, Castells A, Ayuso C, et al: Preoperative staging and tumor resectability assessment of pancreatic cancer: Prospective study comparing endoscopic ultrasonography, helical computed tomography, magnetic resonance imaging, and angiography. Am J Gastroenterol 99(3):492–501, 2004. 237. Stollfuss JC, Glatting G, Friess H, et al: 2-(fluorine-18)-fluoro-2-deoxyD-glucose PET in detection of pancreatic cancer: Value of quantitative image interpretation. Radiology 195(2):339–344, 1995. 238. Stolte M, Weiss W, Volkholz H, et al: A special form of segmental pancreatitis: Groove pancreatitis. Hepatogastroenterology 29(5):198–208, 1982. 239. Suda K, Shiotsu H, Nakamura T, et al: Pancreatic fibrosis in patients with chronic alcohol abuse: Correlation with alcoholic pancreatitis. Am J Gastroenterol 89(11):2060–2062, 1994. 240. Sugiyama M, Atomi Y, Kuroda A: Two types of mucin-producing cystic tumors of the pancreas: Diagnosis and treatment. Surgery 122(3):617– 625, 1997. 241. Sugiyama M, Haradome H, Atomi Y: Magnetic resonance imaging for diagnosing chronic pancreatitis. J Gastroenterol 42(Suppl 17):108–112, 2007.

1478

PART II  CT and MR Imaging of the Whole Body

242. Tamm EP, Silverman PM, Charnsangavej C, et al: Diagnosis, staging, and surveillance of pancreatic cancer. AJR Am J Roentgenol 180(5):1311–1323, 2003. 243. Tamura R, Ishibashi T, Takahashi S: Chronic pancreatitis: MRCP versus ERCP for quantitative caliber measurement and qualitative evaluation. Radiology 238(3):920–928, 2006. 244. Tanaka M: Intraductal papillary mucinous neoplasm of the pancreas: Diagnosis and treatment. Pancreas 28(3):282–288, 2004. 245. Taouli B, Vilgrain V, Vullierme MP, et al: Intraductal papillary mucinous tumors of the pancreas: Helical CT with histopathologic correlation. Radiology 217(3):757–764, 2000. 246. Tatli S, Mortele KJ, Levy AD, et al: CT and MRI features of pure acinar cell carcinoma of the pancreas in adults. AJR Am J Roentgenol 184(2):511–519, 2005. 247. Teh SH, Sheppard BC, Mullins RJ, et al: Diagnosis and management   of blunt pancreatic ductal injury in the era of high-resolution computed axial tomography. Am J Surg 193(5):641–643, discussion 643, 2007. 248. Tempero MA, Arnoletti JP, Behrman S, et al: Pancreatic adenocarcinoma. J Natl Compr Canc Netw 8(9):972–1017, 2010. 249. Terris B, Ponsot P, Paye F, et al: Intraductal papillary mucinous tumors of the pancreas confined to secondary ducts show less aggressive pathologic features as compared with those involving the main pancreatic duct. Am J Surg Pathol 24(10):1372–1377, 2000. 250. Tham RT, Heyerman HG, Falke TH, et al: Cystic fibrosis: MR imaging of the pancreas. Radiology 179(1):183–186, 1991. 251. Thoeni RF: The revised Atlanta classification of acute pancreatitis: Its importance for the radiologist and its effect on treatment. Radiology 262(3):751–764, 2012. 252. Tublin ME, Tessler FN, Cheng SL, et al: Effect of injection rate of contrast medium on pancreatic and hepatic helical CT. Radiology 210(1):97–101, 1999. 253. Tunaci M: Multidetector row CT of the pancreas. Eur J Radiol 52(1):18–30, 2004. 254. Turner MA: The role of US and CT in pancreatitis. Gastrointest Endosc 56(6 Suppl):S241–S245, 2002. 255. Ura H, Denno R, Hirata K, et al: Carcinoma arising from ectopic pancreas in the stomach: Endosonographic detection of malignant change. J Clin Ultrasound 26(5):265–268, 1998. 256. vanSonnenberg E, Wittich GR, Casola G, et al: Percutaneous drainage of infected and noninfected pancreatic pseudocysts: Experience in 101 cases. Radiology 170(3 Pt 1):757–761, 1989. 257. Varadarajulu S, Wallace MB: Applications of endoscopic ultrasonography in pancreatic cancer. Cancer Control 11(1):15–22, 2004. 258. Varadhachary GR, Tamm EP, Abbruzzese JL, et al: Borderline resectable pancreatic cancer: Definitions, management, and role of preoperative therapy. Ann Surg Oncol 13(8):1035–1046, 2006. 259. Vargas R, Nino-Murcia M, Trueblood W, et al: MDCT in pancreatic adenocarcinoma: Prediction of vascular invasion and resectability using a multiphasic technique with curved planar reformations. AJR Am J Roentgenol 182(2):419–425, 2004. 260. Vogel AM, Alesbury JM, Fox VL, et al: Complex pancreatic vascular anomalies in children. J Pediatr Surg 41(3):473–478, 2006. 261. Volmar KE, Vollmer RT, Jowell PS, et al: Pancreatic FNA in 1000 cases: A comparison of imaging modalities. Gastrointest Endosc 61(7):854– 861, 2005. 262. von Schulthess GK, Steinert HC, Hany TF: Integrated PET/CT: Current applications and future directions. Radiology 238(2):405–422, 2006.

263. Vujic I: Vascular complications of pancreatitis. Radiol Clin North Am 27(1):81–91, 1989. 264. Warshaw AL, Compton CC, Lewandrowski K, et al: Cystic tumors of the pancreas. New clinical, radiologic, and pathologic observations in 67 patients. Ann Surg 212(4):432–443, discussion 444–445, 1990. 265. Warshaw AL, Fernandez-del Castillo C: Pancreatic carcinoma. N Engl J Med 326(7):455–465, 1992. 266. Warshaw AL, Rattner DW, Fernandez-del Castillo C, et al: Middle segment pancreatectomy: A novel technique for conserving pancreatic tissue. Arch Surg 133(3):327–331, 1998. 267. Wiersema MJ: Accuracy of endoscopic ultrasound in diagnosing and staging pancreatic carcinoma. Pancreatology 1(6):625–632, 2001. 268. Williams DB, Sahai AV, Aabakken L, et al: Endoscopic ultrasound guided fine needle aspiration biopsy: A large single centre experience. Gut 44(5):720–726, 1999. 269. Winston CB, Mitchell DG, Outwater EK, et al: Pancreatic signal intensity on T1-weighted fat saturation MR images: Clinical correlation. J Magn Reson Imaging 5(3):267–271, 1995. 270. Wong YC, Wang LJ, Lin BC, et al: CT grading of blunt pancreatic injuries: Prediction of ductal disruption and surgical correlation.   J Comput Assist Tomogr 21(2):246–250, 1997. 271. Yamada T, Alpers D, Kountz WB, et al: Textbook of gastroenterology, Philadelphia, 2003, Lippincott Williams & Wilkins. 272. Yamada Y, Mori H, Kiyosue H, et al: CT assessment of the inferior peripancreatic veins: Clinical significance. AJR Am J Roentgenol 174(3):677–684, 2000. 273. Yamaguchi K, Ogawa Y, Chijiiwa K, et al: Mucin-hypersecreting tumors of the pancreas: Assessing the grade of malignancy preoperatively. Am J Surg 171(4):427–431, 1996. 274. Yamaguchi K, Tanaka M: Groove pancreatitis masquerading as pancreatic carcinoma. Am J Surg 163(3):312–316, discussion 317–318, 1992. 275. Yeo TP: Demographics, epidemiology, and inheritance of pancreatic ductal adenocarcinoma. Semin Oncol 42(1):8–18, 2015. 276. Yeo CJ, Bastidas JA, Lynch-Nyhan A, et al: The natural history of pancreatic pseudocysts documented by computed tomography. Surg Gynecol Obstet 170(5):411–417, 1990. 277. Yogi Y, Shibue T, Hashimoto S: Annular pancreas detected in adults, diagnosed by endoscopic retrograde cholangiopancreatography: Report of four cases. Gastroenterol Jpn 22(1):92–99, 1987. 278. Yokoyama T, Miyagawa S, Noike T, et al: Isolated pancreatic tuberculosis. Hepatogastroenterology 46(27):2011–2014, 1999. 279. Yu J, Turner MA, Fulcher AS, et al: Congenital anomalies and normal variants of the pancreaticobiliary tract and the pancreas in adults: Part 2, Pancreatic duct and pancreas. AJR Am J Roentgenol 187(6):1544– 1553, 2006. 280. Zamboni G, Luttges J, Capelli P, et al: Histopathological features of diagnostic and clinical relevance in autoimmune pancreatitis: A study on 53 resection specimens and 9 biopsy specimens. Virchows Arch 445(6):552–563, 2004. 281. Zhang L, Chari S, Smyrk TC, et al: Autoimmune pancreatitis (AIP) type 1 and type 2: An international consensus study on histopathologic diagnostic criteria. Pancreas 40(8):1172–1179, 2011. 282. Zhang XM, Shi H, Parker L, et al: Suspected early or mild chronic pancreatitis: Enhancement patterns on gadolinium chelate dynamic MRI. Magnetic resonance imaging. J Magn Reson Imaging 17(1):86–94, 2003.

47  Mesentery Yong Ho Auh, George Shih, Jung Hoon Kim, and Stuart Bentley-Hibbert

EMBRYOLOGY

ANATOMY

At the end of the third week of gestation, the primitive gastrointestinal (GI) tract of an embryo consists of a straight tube suspended from the esophagus to the rectum by the dorsal mesentery, which contains the vascular supply. In the dorsal mesentery, the portion connected to  the stomach is known as the dorsal mesogastrium. As the embryo lengthens, the caudal portion of the septum transversum thins and becomes the ventral mesentery. It attaches the stomach and duodenum to the ventral wall of the abdominal cavity.97 Because there is no ventral mesentery below the foregut, the ventral mesentery is the same as the ventral mesogastrium. The sagittally oriented ventral and dorsal mesogastrium, along with the upper GI tract, divide the upper abdominal cavity into equally sized right and left peritoneal cavities. During the fourth gestational week, cords of tissue within the ventral mesogastrium begin to grow rapidly, forming the liver, bile ducts, and ventral pancreas within the ventral mesogastrium, while the spleen and dorsal pancreas arise within the dorsal mesogastrium. The remainder of the mesogastrium become the ligaments and omenta. The anterior part of the ventral mesogastrium between the anterior abdominal wall and the liver becomes the falciform ligament, and the posterior part between the liver and the stomach becomes the lesser omentum. The posterior portion of the dorsal mesogastrium between the spleen and posterior abdominal wall becomes the splenorenal ligament, while the anterior portion between the spleen and the stomach becomes the gastrosplenic ligament, which is a part of the greater omentum. The stomach initially develops as a fusiform bulge in the foregut. This bulge rapidly becomes asymmetric owing to the more rapid growth of its dorsal margin (greater curvature) than of its ventral margin (lesser curvature). The stomach then undergoes a 90-degree counterclockwise rotation along with the ventral and dorsal mesogastrium (Fig. 47-1). The dorsal mesentery inferior to the stomach evolves into small bowel mesentery that attaches the jejunum and ileum to the posterior abdominal wall and mesocolon in the transverse and sigmoid colon. The dorsal mesentery of the duodenum, along with the pancreas, is fused with the posterior abdominal wall and loses its mesentery. Rarely when they are not completely fused with the posterior abdominal wall, the duodenum and pancreas can become intraperitonealized (Fig. 47-2). The dorsal mesentery attached to ascending and descending colon and rectum is lost once these segments of colon are fused with the posterior abdominal and pelvic walls. Not infrequently the ascending and descending colon are incompletely fused with the posterior abdominal wall and become intraperitonealized (Fig. 47-3).3,4,21 Incomplete fusion with the posterior pelvic wall can cause the rectum to be intraperitonealized (Fig. 47-4).

In its broad meaning, mesentery may include any membranous derivatives of the embryonic dorsal and ventral mesentery (e.g., falciform ligament, lesser omentum, gastrosplenic and splenorenal ligaments).8 When it is defined as a membranous fold attached to various organs, its definition can be even broader and may include even the broad ligament of the uterus, mesosalpinx, gastropancreatic fold, and other structures. The usual definition of the mesentery is, however, limited to the derivatives of the embryonic dorsal mesentery below the foregut—the membranous fold connecting the small and large bowel to the posterior body wall. Mesenteries are double-layered folds of peritoneum within which lie continuations of the subperitoneal space,26,92,102 which contains various amounts of adipose tissue within which the arteries, veins, lymphatics, and nerves of the bowel course. On normal cross-sectional images, although the supplying and draining vessels can be traced to and from the major mesenteric vessels, the actual leaves of the mesentery are not discernible unless they are separated by intervening ascites or peritoneal thickening (Fig. 47-5). Mesenteries can act as barriers as well as conduits for the spread of intraperitoneal and extraperitoneal diseases.92

Small Bowel Mesentery The narrowest definition of the mesentery is the small bowel mesentery, also known as the mesentery proper or mesenterium. The small bowel mesentery is a broad fan-shaped fold of peritoneum connecting the jejunum and ileum to the root of the mesentery. The mesenteric root is approximately 15 cm long and is directed obliquely downward to the right from the duodenojejunal flexure to the ileocecal junction. The intestinal border of the mesentery is about 6 m long and is formed into numerous pleats. The mesentery consists of two layers of the peritoneum between which lie the jejunal and ileal branches of the superior mesenteric artery (SMA), with accompanying veins, nerve plexuses, lymphatics, lymph nodes, connective tissue, and fat (Fig. 47-6). The gastrocolic trunk is an anatomic landmark between the transverse mesocolon and root of the small bowel mesentery. On computed tomography (CT) and magnetic resonance imaging (MRI), the mesentery appears as a fat-containing structure inseparable from the other fat-containing peritoneal folds (e.g., greater omentum) or even from retroperitoneal fat (see Fig. 47-5). Normal mesenteric fat is similar in density to subcutaneous fat (−100 to −160 Hounsfield units [HU]) on CT125 and has the same signal intensity on MRI. The jejunal and ileal vessels can be identified as distinct round or linear densities within the mesenteric fat. Normal-sized lymph nodes within Text continued on p. 1484

1479

a Liver Liver

b

Stomach

c

Stomach

Spleen

Spleen

Pancreas Pancreas

d

B

A

FIG 47-1  Schematic drawing of early fetal development of the mesentery. A, During the fourth week of embryonal life, cords of tissue from the ventral and dorsal mesogastrium grow rapidly, forming the liver and ventral pancreas within the ventral mesogastrium, while the spleen and the dorsal pancreas develop within the dorsal mesogastrium. The primitive foregut is suspended within the abdomen by the ventral and dorsal mesogastrium along with cords of tissue that divide the cavity into symmetric right and left halves. B, During early fetal life the dorsal margin of the stomach bulges rapidly along with rapid growth of the dorsal mesogastrium, forming a large greater omentum. Meanwhile the dorsal pancreas fuses to the posterior abdominal wall (dashed lines). a, falciform ligament; b, gastrohepatic ligament; c, gastrosplenic ligament; d, lienorenal ligament.

*

A

B

*

*

C

D FIG 47-2  Incomplete fusion of the pancreas and duodenum to posterior abdominal wall. A to D, Four consecutive axial CT images of the abdomen at the level of the transverse portion of the duodenum (arrows) show anterior displacement of the transverse duodenum and uncinate process of the pancreas (black and white arrows) by a jejunal loop (asterisk).

CHAPTER 47  Mesentery

A

B

C FIG 47-3  Colocolic intussusception of descending colon. Two consecutive coronal reformatted CT images (A and B) and one axial CT image (C) show a bowel-within-bowel appearance (black arrows) in the descending colon. Descending mesocolon is seen as a curvilinear fat density (white arrow) as part of the intussusceptum.

1481

A

B FIG 47-4  Peritonealized rectum by mesorectum. A and B, Two consecutive axial CTs of the pelvis show ascites in bilateral pararectal fossae (white arrows) that extends behind the rectum, demarcating the fatcontaining mesorectum (black arrow).

A

B

C FIG 47-5  Small bowel mesentery in a normal subject (A) and a patient with ascites (B and C). A, Normally the mesentery is not identified as individual pleats. Mesenteric fat is inseparable from retroperitoneal and omental fat. It can be indirectly traced and identified by the undulating vessels (arrows). B and C, Once there is enough ascites, each pleat of mesentery can be separately identified. A long thin mesenteric attachment to the root of the mesentery (arrows in B) is well depicted. The mesentery can be easily traced and identified by the mesenteric vessels on coronal reformatted CT (arrows in C).

CHAPTER 47  Mesentery

A

B

C

D FIG 47-6  Peritoneal lymphomatosis. A to C, Axial CT of abdomen and pelvis and (D) coronal reformatted CT show omental caking (arrow) and masslike lymphoma implantation onto pleats of mesentery (arrowheads in D). A long thick mesenteric attachment to the root of the mesentery (arrows in D) is well depicted.

1483

1484

PART II  CT and MR Imaging of the Whole Body shaped like an inverted V, the apex of which is near the division of the left common iliac artery; the left line descends medially to the left psoas major muscle, and the right line descends into the pelvis and ends in the median plane at the level of the third sacral vertebra.138 Because of this attachment, a triangular-shaped fluid collection with a floating mesocolon is seen frequently in the presence of ascites. For the same reason, when a volvulus of the sigmoid colon occurs, the twisted colonic loop orients toward the right upper quadrant of the abdomen (Fig. 47-11). Between the layers of the sigmoid mesocolon are the sigmoidal and superior hemorrhoidal vessels. Both the ascending and descending colons are normally retroperitoneal organs but can be partially or totally intraperitoneal. The ascending colon possesses a mesocolon in up to 26% of cases, and the descending colon in up to 36% of cases49 (see Fig. 47-3). In such instances the ascending and descending colons can be more mobile. Ascites and seeded metastases can be found posterior to the ascending and descending colons, thereby mimicking a retroperitoneal collection or mass. Even the rectum is rarely peritonealized by having its own mesorectum (see Fig. 47-4). In this situation an intraperitoneal organ or mass can be mistaken for retroperitoneal structures.115,132

FIG 47-7  Normal anatomy of mesenteric vessels and lymph nodes. Contrast-enhanced fat-saturated coronal MRI shows enhanced branches of the SMV (large arrows) and SMA (small arrows). Arteries are smaller and run parallel to veins. Along with these vessels, normal mesenteric nodes (arrowheads) are scattered as round or ovoid structures.

the mesentery are routinely identified, particularly following contrast enhancement and in coronal images100 (Fig. 47-7).

Mesocolon The mesocolon is composed of two layers of peritoneum that connect the colon to the posterior abdominal wall; it contains its related blood vessels, lymphatics, nerves, and a variable amount of adipose tissue. The transverse colon and sigmoid colon have a well-formed mesocolon. The cecum is attached to the ileum by the ileocecal fold and has a complete peritoneal covering with no mesocolon in most cases.49 On CT and MRI, the mesocolon is not usually easily discernable, but it can be traced by identifying its blood vessels from the marginal vessels to the superior or inferior mesenteric vessel or vice versa18 (Fig. 47-8). The transverse mesocolon is a broad fold connecting the transverse colon to the posterior abdominal wall.88 The transverse mesocolon is typically absent close to the hepatic flexure, and therefore the colon is usually in direct posterior contact with the second portion of the duodenum as it crosses. The mesocolon is short over the first part of the head of the pancreas but is much longer where it is attached to the anterior border of the body of the pancreas.49,138 It ends at the splenic flexure but usually extends farther laterally as a peritoneal fold called the phrenicocolic ligament.87 Because of its intimate relationship with almost the entire length of the pancreas, pancreatitis easily extends to the transverse mesocolon135 (Fig. 47-9). The transverse mesocolon is formed by two layers of peritoneum; the upper layer is adherent to but separable from the greater omentum138 (Fig. 47-10). Between the two layers of the transverse mesocolon are the middle colic arteries and veins, lymphatics, and nerves of the transverse mesocolon. Because of the transverse mesocolon, the transverse colon has a great deal of mobility, and the position and orientation of vessels in the transverse mesocolon vary depending upon the position of the transverse colon.18 The sigmoid mesocolon is a fold of peritoneum that attaches the sigmoid colon to the posterior pelvic wall. Its line of attachment is

Omentum The omentum is a fold of peritoneum extending from the stomach to adjacent organs. Different from the small bowel mesentery and mesocolon, which connect the bowels to the posterior abdominal wall, the omentum connects two interperitoneal organs: the lesser omentum (between the stomach and liver) and the greater omentum (between the stomach and spleen or transverse colon). Like the small bowel mesentery and mesocolon, the greater omentum is formed by two layers of peritoneum within which lie vessels, lymphatics, lymph nodes, nerves, and varying amounts of adipose tissue127,143 (Fig. 47-12; see Fig. 47-10). The lesser omentum extends from the lesser curvature of the stomach and first portion of the duodenum to the liver at the porta hepatis into the fissure for the ligamentum venosum.3,92 The portion of the lesser omentum extending between the liver and stomach is called the gastrohepatic ligament, and that between the liver and duodenum, the hepatoduodenal ligament. The left gastric artery acts a marker for the gastrohepatic ligament, and the hepatic artery proper acts as an anatomic marker for the hepatoduodenal ligament. At its right free margin, the lesser omentum encloses the portal triad with its lymphatics and lymph nodes; this free margin is the anterior edge of the epiploic foramen.144 The lesser omentum contains the left and right gastric arteries and corresponding veins as well as lymphatics and lymph nodes.136 When there is sufficient ascites, it may appear as an undulating fat-containing plate in the fissure for the ligamentum venosum and serve as an anatomic landmark separating the lesser sac from the greater sac3 (Fig. 47-13). The left gastric vessels are easily identified at the lesser curvature aspect of the lesser omentum. Normal lymph nodes up to 8 mm in diameter are also frequently seen in the lesser omentum. The dilated coronary vein can easily be visualized in patients with portal hypertension.9 The greater omentum is the largest peritoneal fold, consisting of a double sheet folded on itself so that it is made of four layers in early development. The inner two layers are fused below the transverse colon and lose their mesothelial lining during fetal life (see Fig. 47-10). This fusion limits the inferior extent of the lesser sac. It stretches from the greater curvature of the stomach and first portion of the duodenum downward in front of the small intestine for a variable distance. It adheres to but is separable from the upper layer of the transverse mesocolon. The gastrocolic ligament is the part of the greater omentum Text continued on p. 1489

CHAPTER 47  Mesentery

A

B

C

D FIG 47-8  Sigmoid colon cancer. A to D, Coronal reformatted CT shows focal enhanced wall thickening in the upper rectum (arrowheads) with infiltration surrounding the sigmoid mesocolon. Small lymph node is noted on the sigmoid mesocolon (black arrow). The sigmoid mesocolon can be easily traced on coronal reformatted CT (arrows).

1485

1486

PART II  CT and MR Imaging of the Whole Body

*

A

B

*

*

C

D FIG 47-9  Acute pancreatitis. A, CT scout image shows sharply demarcated absence of colonic air (arrows) at midsegment of the transverse colon owing to involvement of pancreatitis through the transverse mesocolon. B to D, Three consecutive axial CT images show a large phlegmonous inflammatory mass (asterisk) encircling a segment of transverse colon (arrows in D), which extends from the pancreas.

CHAPTER 47  Mesentery

e

Lesser omentum Sto ma ch

a b c

d

A

Pancreas 3rd portion duodenum a

a

b c Transverse colon

B

FIG 47-10  Schematic drawing of posterior attachment of small bowel mesentery, transverse and sigmoid mesocolon, and greater omentum. Coronal (A) and sagittal (B) views show the attachment of greater omentum and transverse mesocolon running parallel to each other—the greater omentum being shorter and superior to the mesocolon. These two peritoneal folds are loosely attached but not fused. Note the inverted V–shaped attachment of the sigmoid mesocolon. a, greater omentum; b, transverse mesocolon; c, small bowel mesentery; d, sigmoid mesocolon; e, gastrosplenic ligament.

A

B

FIG 47-11  Sigmoid volvulus. A to D, Coronal reformatted CT shows distended sigmoid loop with inverted-U configuration and coffee bean sign (arrows). There is a whirlpool sign with torsion of the sigmoid vessels (arrowheads). A transition point with abrupt reduction in bowel caliber is seen. Continued

1487

1488

PART II  CT and MR Imaging of the Whole Body

C

D FIG 47-11, cont’d

a c b

d e

FIG 47-13  Lesser omentum. Axial CT shows the lesser omentum (black arrows) as a thin line in the fissure for the ligamentum venosum, extending from the lesser curvature of the stomach to the umbilical portion of the left portal vein (open arrow).

FIG 47-12  Schematic drawing of lesser omentum and greater omentum. Lesser omentum extends from lesser curvature of the stomach and superior border of the first portion of the duodenum to the liver into the fissure for the ligamentum venosum. Greater omentum drapes down from the greater curvature of the stomach and inferior border of the first portion of the duodenum, covering transverse colon and small bowel. Superior portion of the greater omentum extends to hilum of the spleen and becomes gastrosplenic ligament. a, gastrohepatic ligament; b, hepatoduodenal ligament; c, gastrosplenic ligament; d, gastrocolic ligament; e, duodenocolic ligament.

CHAPTER 47  Mesentery stretching from the stomach to the transverse colon, where as the duodenocolic ligament it extends from the first portion of the duodenum to the transverse colon (see Fig. 47-12).116 The gastrosplenic ligament connects the stomach to the spleen and is the part of the greater omentum. Contained within the greater omentum are the gastroepiploic arteries and veins. On cross-sectional images the greater omentum appears as a thin, broad, fat-containing area just beneath the anterior abdominal wall and ventral to the stomach, transverse colon, and small bowel loops.127 The greater omentum is frequently found wrapped about the organs in the upper part of the abdomen—only occasionally is it evenly dependent anterior to the intestines. Especially in the presence of a large amount of ascites, it is wrapped on itself and often misplaced at a particular area. It may limit the spread of infection by forming adhesions with areas of inflammation in the peritoneal cavity. It is frequently involved by peritoneal diseases, either infective or malignant.

RADIOLOGIC MANIFESTATIONS OF MESENTERIC DISEASE Mesenteric disease is difficult to separate from peritoneal or bowel disease, because the mesentery is intimately related to the peritoneum

A

and bowel anatomically and physiologically. Therefore primary peritoneal and bowel diseases often accompany changes of the mesentery and vice versa.89,91 The peritoneum is indeed a mesenteric component because it is the lining membrane of the mesentery. Peritonitis, particularly tuberculous peritonitis, and peritoneal carcinomatosis cause or are associated with mesenteric abnormality. By the same token, primary mesenteric disease processes cause and are associated with abnormal peritoneal findings like thickening and enhancement of peritoneal membrane or ascites or changes of bowel (Figs. 47-14 and 47-15). Inflammatory, infectious, and neoplastic diseases of bowel bring mesenteric fat changes and regional lymphadenopathy. Sclerosing mesenteritis induces, on the other hand, thickening and edema  of bowel wall and ascites secondary to vascular and/or lymphatic involvement (see Fig. 47-14). Because mesenteric space is continuous with extraperitoneal space as subperitoneal space, it becomes a conduit transmitting disease processes arising in the retroperitoneal space to GI organs and vice versa.102,130 Whether abnormal mesenteric changes are due to primary mesenteric causes or are secondary effects due to disease processes of  neighboring organs or structures, they manifest distinct and limited radiologic findings. These findings come together in different combinations or independently.

B

a

C

1489

D FIG 47-14  Sclerosing mesenteritis in a 70-year-old man. A and B, Noncontrast axial CT images show a well-circumscribed soft tissue mass (white arrows) in the small bowel mesentery, with coarse calcification (black arrow). C and D, On corresponding postcontrast images at the same level the mass encases the mesenteric vessels, causing vessel engorgement in the leaves of the mesentery and ascites (a). The superior mesenteric artery (black arrowhead in C) is encased by the soft tissue mass. (Courtesy Dr. Ha, Asan Medical Center, Seoul, Korea.)

1490

PART II  CT and MR Imaging of the Whole Body and mesenteric involvement of carcinoid tumor (see Fig. 47-20), misty mesentery may manifest as a stellate or spiculated mass due to extensive fibrosis; often calcifications are associated.59,121

Mass

A

Mesenteric mass, either cystic or solid, is quite uncommon. Other than conglomerated massive lymphadenopathy of lymphoma and mesenteric involvement of carcinoid tumor, a solid mesenteric mass is rare. Even though carcinoid tumor arises primarily from bowel— most frequently from ileum—it is often difficult to recognize the primary bowel mass. Instead the spiculated or stellate mesenteric mass is the dominant finding, which is a distinctive form of sclerosing mesenteritis due to the tissue reaction to released hormones by the tumor121 (see Fig. 47-20). Primary mesenteric GIST (GI stromal tumor)64 (Fig. 47-26), desmoid tumor associated with or without familial adenomatous polyposis73 (Fig. 47-27), and inflammatory pseudotumor92 (see Fig. 47-15) are known entities. Other benign and malignant primary mesenchymal tumors or metastatic tumors are extremely rare58 (Fig. 47-28). Also rare are cystic mesenteric lesions, commonly called mesenteric or omental cysts, which is rather descriptive terminology. A variety of different pathologic entities comprise such entities. Most common is lymphangioma (Figs. 47-29 and 47-30). Other types include enteric cyst, enteric duplication cyst, mesothelial cyst, nonpancreatic pseudocyst, cystic mesothelioma, and cystic teratoma.128

Lymphadenopathy B FIG 47-15  Inflammatory pseudotumor. Axial (A) and coronal (B) reformatted CT of the lower abdomen show spiculated soft tissue mass (arrows) in the mesentery. Mesenteric vessels (arrowhead in B) are encased and tethered, resulting in engorgement of vessels and thickening of the jejunal wall. (Courtesy Dr. Ha, Asan Medical Center, Seoul, Korea.)

Abnormal Density of Mesenteric Fat The most frequent radiologic manifestation of mesenteric abnormalities is alteration of fat density or intensity. The mesenteric fat becomes hazy due to lymphedema (see Fig. 47-24), edema (Fig. 47-16), hemorrhage (Fig. 47-17), cellular infiltration by inflammation or neoplasm (Figs. 47-18 to 47-20; see Fig. 47-15), or fibrosis (see Fig. 47-14). It is described as “misty mesentery.”93 Focal misty mesentery can be due to local etiologies such as trauma, strangulated bowel obstruction, infectious and inflammatory bowel disease (IBD), pancreatitis (see Fig. 47-16), and ischemic mesenteric disease36,61 (Fig. 47-21). It is diffuse when it is due to systemic causes like portal hypertension (Fig. 47-22), hypoalbuminemia, or heart failure. Whether localized or diffuse, it is usually ill-defined but is sometimes well demarcated, as in segmental misty mesentery120 (Figs. 47-23 to 47-25; see Fig. 47-16) and mesenteric panniculitis.52 It is sometimes demarcated by a so-called tumoral pseudocapsule (see Figs. 47-16 and 47-23), which is a peripheral band of soft tissue attenuation. Depending upon the severity and cause of disease, the misty mesentery may show variable densities. From cases of lymphedema to edema, to inflammatory and neoplastic cell infiltration, to hemorrhage, and to fibrosis with calcifications, the density of misty mesentery becomes higher. In cases of sclerosing mesenteritis (see Fig. 47-14)

Normal mesenteric lymph nodes are routinely seen at CT and MRI. They are more dramatically visualized and better appreciated on coronal images and following contrast enhancement (see Fig. 47-7). It is difficult to define the criteria of a normal mesenteric lymph node. As in other areas of the body, the size, shape, number, and distribution of nodes should be considered. Abnormality is generally defined as any mesenteric lymph node that measures larger than 6 mm in short axis or when a group of three or more lymph nodes is clustered in one anatomic location.19,81,82,83 A variety of diseases or conditions lead to mesenteric lymphadenopathy. Notably, lymphoma is the most common neoplastic cause (Figs. 47-31 and 47-32). Non-Hodgkin’s lymphoma in particular involves mesenteric nodes, usually along with paraaortic nodes. Hodgkin’s lymphoma usually does not involve mesenteric nodes. GI tract malignancy frequently metastasizes to regional mesenteric lymph nodes. Genitourinary tract malignancy and extraabdominal malignancy seldom metastasize to mesenteric nodes. Inflammatory or  infectious GI tract disease frequently cause regional mesenteric lymphadenopathy, but their size is usually less than 10 mm83 (Fig. 47-33). In the case of primary mesenteric lymphadenitis, inflamed lymphadenopathy is the sole finding, without accompanying abnormal bowel changes, although an underlying infectious terminal ileitis is thought to be the cause83 (Fig. 47-34). Other systemic diseases like systemic lupus erythematosus, systemic mastocytosis, and sarcoidosis may show mesenteric lymphadenopathy. Hypodense lymphadenopathy with Whipple’s disease has been overly emphasized; it is not a unique finding for Whipple’s disease. Hypodense lymphadenopathy is more commonly seen in other benign and malignant diseases, notably in abdominal tuberculous or atypical tuberculous disease (Figs. 47-35 and 47-36) and Kaposi’s sarcoma. Furthermore, Whipple’s disease is extremely rare.35 It is also worthwhile to recognize mesenteric cavitary lymph node syndrome as an entity characterized by cystic change in mesenteric lymph nodes and associated with celiac sprue.54 Text continued on p. 1503

CHAPTER 47  Mesentery

A

B

C

D

E

F FIG 47-16  Mimics of mesenteric panniculitis. A and B, Axial postcontrast CT images show hazy infiltration of the small bowel mesentery (white arrows) due to appendicitis (white arrow in B). C, A band of soft tissue—the tumoral pseudocapsule (black arrows)—demarcating a “misty mesentery” in a patient with Crohn’s disease. D, A hazy central small bowel mesentery (white arrows) with prominent lymph nodes (black arrows) in a patient who had been treated for testicular cancer. E, Misty mesentery (white arrow) with mesenteric vein engorgement (black arrow) in a patient previously treated for lymphoma. F, Hazy mesentery in a patient with pancreatitis. Fat ring sign (C to F) is well seen around vessels and lymph nodes.

1491

1492

PART II  CT and MR Imaging of the Whole Body

B

B

A

B

C

D FIG 47-17  Seatbelt injury. A and B, Axial CT images demonstrate a bowel wall and mesocolic hematoma in the cecum (white arrows) in a patient following blunt abdominal trauma sustained from a motor vehicle accident. The patient was wearing a seatbelt at the time of impact. B, urinary bladder. C and D, Images more superiorly show bruising in the abdominal wall in the right lower and left upper quadrants (white arrows) in the distribution of the seatbelt.

U

A

A

B FIG 47-18  Sigmoid diverticulitis in two patients. A, Coronal reformatted CT shows abundant inflammatory changes (black arrow) surrounding the sigmoid colon, which has multiple diverticula (arrowheads) and a small fluid collection within the sigmoid mesentery (white arrow). B, Axial CT in another patient with sigmoid diverticulitis demonstrating soft tissue stranding in the mesenteric fat adjacent to diverticula (black arrows). There is a large associated abscess (A). Urinary bladder is displaced anteriorly (U).

CHAPTER 47  Mesentery

A

B

F

C

D

E

F FIG 47-19  Active inflammatory bowel disease in two different patients. A to D, Serial axial CT images from a patient with Crohn’s disease show a focal segment of thickened distal and terminal ileum (white arrows). Prominence of the vasa recta (black arrowheads in B) indicates active inflammation. Soft tissue inflammatory changes are also evident in the RLQ mesentery (curved black arrow in A). There is abundant fatty proliferation of the mesenteric fat in the small bowel and sigmoid colon (F). E and F, CT images from another patient with active Crohn’s colitis show inflammatory changes in the fibrofatty proliferation of the sigmoid mesocolon (white arrows). Wall thickening and phlegmonous changes of the sigmoid colon are also present (arrowhead in F).

1493

1494

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 47-20  Small bowel carcinoid metastatic to mesentery and liver. A and B, Contrast-enhanced CT of the lower abdomen demonstrates a spiculated soft tissue mass in the mesentery (white arrows) with a “spoke-wheel” arrangement of the mesenteric vessels (black arrows). Adjacent small bowel thickening (curved arrows) represents bowel ischemia. C, CT image through the liver shows the typical enhancing metastatic lesions to the liver.

C

A

B

A A

C

FIG 47-21  Thrombosis of branches of the SMV in two different patients. A and B, Axial CT images in the lower abdomen, showing more distal thrombus (white arrows) in branches of the SMV, with resultant mesenteric haziness, reflecting edema (black arrow in B). C, A different patient with thrombus in small tributaries of the SMV (black arrows), marked bowel wall edema (white arrows), and ascites (A).

CHAPTER 47  Mesentery

A

B FIG 47-22  Liver cirrhosis with mesenteric edema and omental collateral. A, Nonenhanced CT at the mid abdomen shows omental collaterals (white arrows) mimicking mild omental caking. B, Lower-level CT slice shows hazy ill-defined mesenteric edema (black arrow). Note that engorged mesenteric vessels appear radiating. Small bowel wall is also thickened (arrowhead).

A

B

C

D FIG 47-23  Mesenteric panniculitis in a 70-year-old man. A to D, Serial axial CT images show segmental increased attenuation in the small bowel mesentery, separated from the adjacent fat by a tumoral pseudocapsule (black arrows in B). There is preservation of the fat around vessels (fat ring sign) that are surrounded by the increased density (white arrows in C) (Courtesy Dr. Ha, Asan Medical Center, Seoul, Korea.)

1495

1496

PART II  CT and MR Imaging of the Whole Body

*

A

B

C

D FIG 47-24  Idiopathic “misty mesentery.” A and B, Serial contrast-enhanced axial CT images show mild increased attenuation in the small bowel mesenteric fat (white asterisk). Additional linear branching soft tissue structures are within the mesentery, which are not identifiable as veins or arteries. These soft tissue structures are better appreciated on a coronal reformatted CT image (C) and presumably represent dilated lymphatics. D, A more anterior coronal reformatted CT image, redemonstrates the “misty mesentery” (arrow).

A

B FIG 47-25  Idiopathic “misty mesentery” in a patient with chronic abdominal pain. A, Axial enhanced CT image shows increased attenuation in the small bowel mesentery, with a tumoral pseudocapsule (arrows). B, Findings were relatively stable after 2 years.

CHAPTER 47  Mesentery

A

B

D

C FIG 47-26  Omental GIST. Axial CT (A and B) and coronal reformatted CT (C) show a heterogeneously enhancing mass in the greater omentum (arrows). This mass extends to the pelvic cavity (arrowheads). D, Cut section of specimen shows large mass with hemorrhagic necrosis.

A

B FIG 47-27  Mesenteric fibromatosis. A and B, Axial CT at two different levels of the abdomen show a soft tissue mass in the mesentery encasing mesenteric vessels (arrows) in a patient with total colectomy for Gardner’s syndrome.

1497

1498

PART II  CT and MR Imaging of the Whole Body

A

C

B

D

E FIG 47-28  Solitary fibrous tumor arising in the mesentery. Axial CT (A and B) and coronal reformatted CT (C and D) show a well-defined hypervascular mass in the mesentery (arrows). E, Cut section of specimen shows well-defined yellow to white mass with hemorrhagic necrosis.

CHAPTER 47  Mesentery

A

B

C

D FIG 47-29  Cystic lymphangioma in transverse mesocolon. A to D, Axial CT at multiple levels through the abdomen and pelvis show a low-density mass lesion in the transverse mesocolon (arrows). Transverse colon appears to be floating in the mass. Minimal mass effect on adjacent small bowel loops is seen (arrowhead).

1499

1500

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-30  Cystic lymphangioma in mesentery. A to C, Axial CT at multiple levels show low-density mass lesion in the mesentery (arrows). Minimal mass effect on adjacent small bowel loops is seen (arrowhead). D, Surgical specimen shows well-defined yellow mass in the mesentery.

CHAPTER 47  Mesentery

A

B

C

D FIG 47-31  Lymphoma. A to D, Four consecutive axial CTs of the abdomen and pelvis show conglomerate mesenteric lymph nodes (arrows). Mesenteric vessels (arrowheads) are surrounded and engulfed by the mass but not encased, distorted, or narrowed. Note a bulky mass in cecum (black arrows).

1501

1502

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-32  Lymphoma. Axial CT (A and B) and coronal reformatted CT (C and D) images of the abdomen and pelvis show omental caking (arrowheads) and a bulky masslike lymphoma implantation onto pleats of right side omentum (arrows).

CHAPTER 47  Mesentery

1503

mimic omental cake due to peritoneal carcinomatosis or peritonitis (Fig. 47-43; see Fig. 47-22). Mesenteric venous congestion or dilatation of veins is seen in high-grade bowel obstruction, segmental misty mesentery,120 or active IBD. Mesenteric venous air is an important sign of necrotizing enteritis.

PRIMARY MESENTERIC DISEASES Congenital

A

B FIG 47-33  Ileocolic lymphadenopathy with appendicitis. A, Lower abdominal CT during acute appendicitis shows several enlarged lymph nodes (arrows). B, Following antibiotic treatment these lymph nodes (arrows) dramatically decrease in size.

One commonly encounters cases of unexplainable mesenteric lymphadenopathy, most of them clinically insignificant and irrelevant. The best strategy to deal with concern is short-term follow-up study. If no changes arise in the short term, long-term follow-up can be done. If the lymphadenopathy still does not change, one can ignore it.

Abnormal Mesenteric Vessels Mesenteric disease may result from changes of the mesenteric vessels. Arterial thrombosis or embolism or venous thrombosis cause ischemic mesenteric (bowel) disease (Figs. 47-37 to 47-39; see Fig. 47-21). Mesenteric venous or arterial collaterals suggest current or past occlusion of mesenteric vessels or portal hypertension.50 Alteration of mesenteric vessels in position, course, and shape can be important clues of certain bowel-related diseases. When whirling of mesenteric vessels is associated with bowel obstruction, volvulus is the most likely cause.78 In congenital nonrotation or malrotation of bowel, the relationship of the SMA and superior mesenteric vein (SMV)  is reversed (Fig. 47-40). Anterior displacement of the inferior mesenteric vein (IMV) by clustered small bowel loops is the sign of left paraduodenal hernia (Fig. 47-41), whereas anterior displacement of the right colic vein by clustered small bowel loops is the sign of right paraduodenal hernia131 (Fig. 47-42). When there is portal hypertension, multiple mesenteric varices develop. Omental varices sometimes

Congenital defects within the mesentery cause internal hernias. Depending upon the different locations of the defect, different types of internal hernia occur. The most common internal hernias are right and left paraduodenal hernias, where the mesenteric defect occurs around the duodenojejunal junction (see Figs. 47-41 and 47-42). Much less commonly a defect in the small bowel mesentery or mesocolon can cause a transmesenteric hernia. When herniated bowel loops are incarcerated with or without volvulus, findings are those of closedloop obstruction. The mesenteric defect itself is not visualized, and it is difficult to distinguish from other causes of closed-loop obstruction. However, unique displacement of certain mesenteric vessels by clustered bowel is highly suggestive of certain forms of paraduodenal hernia131 (see Figs. 47-41 and 47-42). Malrotation is a congenital abnormal position of the bowel within the peritoneal cavity; it is due to faulty rotational or fixational evolution of the gut, usually involving both the small and large bowel. Abnormal bowel fixation by mesenteric bands (Ladd’s bands) or absence of fixation of portions of the bowel lead to increased risks of bowel obstruction with or without volvulus. Many subtypes of malrotations exist and involve various degrees of malrotation; the position of the duodenojejunal junction and colon depends on the developmental stage at which normal embryologic rotation failed. Some patients with malrotation show signs of bowel obstruction soon after birth (e.g., bilious vomiting), but some people affected by malrotation may not be diagnosed until long after infancy, and often it may be an incidental finding. In patients with malrotation the mesenteric attachment of the midgut, particularly the portion from the duodenojejunal junction to the cecum, is abnormally short. The bowel has an increased propensity to twist counterclockwise around the SMA and SMV, which is known as midgut volvulus. Midgut volvulus may present as intermittent abdominal pain and may cause bowel necrosis.2 Diagnosis is usually made on an upper GI series, which remains the imaging standard and shows an abnormal duodenojejunal junction, usually in the right abdomen or at least to the right of midline. On a CT the duodenum does not cross the midline, and no clear duodenojejunal junction is identified. Most small bowel loops are in the right side of the abdomen, and colonic loops are in the left. Additionally, mesenteric bands may lead to external obstruction, especially in the duodenum or ileum. Twisting can result in a whirl-like pattern on CT. The relationship between the SMA and SMV may also be reversed, with the SMV to the left of the SMA33 (see Fig. 47-40).

Neoplasm Primary neoplasm—either cystic or solid—of the mesentery is very rare; however, mesenteric fibromatosis is worth mentioning.108 Also known as intraabdominal fibromatosis or abdominal desmoid tumor, mesenteric fibromatosis is a locally aggressive but benign proliferative tumor that does not metastasize. It is part of the spectrum of deep fibromatoses, and the small bowel mesentery is the most common site of intraabdominal fibromatosis.76,124 Most cases are sporadic, but approximately 10% to 15% occur in patients with familial adenomatous polyposis (FAP) (i.e., Gardner’s

1504

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-34  Mesenteric adenitis. A to D, Axial CT of the lower abdomen and pelvis shows multiple enlarged RLQ mesenteric lymph nodes (white arrows in A and B) and mild thickening of the terminal ileum (white arrows in C and D) in a pediatric patient presenting with RLQ pain. The appendix is normal (black arrows in C and D).

CHAPTER 47  Mesentery

A

B

C

D FIG 47-35  Mesenteric adenopathy in a 41-year-old man with HIV, AIDS, and Mycobacterium avium complex (MAC). A to D, Serial axial contrast-enhanced CTs show conglomerate adenopathy (white arrows) in the mesentery. More discrete enlarged nodes are in the RLQ (black arrow). Nodes are of homogenous soft tissue attenuation. Note vascular engorgement and perivascular edema (black arrowheads).

1505

1506

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-36  Tuberculous lymphadenopathy. A to D, Axial CT of the abdomen and pelvis shows multiple enlarged lymph nodes in the mesentery (arrows). A hypodense center is noted (arrowhead).

CHAPTER 47  Mesentery

1507

B

A

FIG 47-37  SMA embolus. A, Curved planar reformatted image of the SMA shows a filling defect (arrow) in a branch of the SMA consistent with an embolus in a patient with known atrial fibrillation. B, Another curved planar reformatted image shows abnormal enhancement and wall thickening in a long segment of ileum (arrows) related to ischemia.

syndrome variant). About 83% of mesenteric fibromatosis with FAP patients have a history of abdominal surgery, most commonly total colectomy. Only about 10% of sporadic mesentery fibromatosis cases have had previous abdominal surgery. Patients with FAP may have fibromatosis in and around anastomotic or surgical incision sites. Abdominal desmoids can be solitary or multiple, and their CT manifestations are variable. CT findings include a focal mesenteric mass (see Fig. 47-27) that may either have a highly collagenous stroma demonstrating soft tissue density or a myxoid stroma that appears more hypodense. Lesions can have mixed collagenous and myxoid stroma and may appear striated or whorled because of their alternating composition. Infiltration into adjacent organs or growth into the abdominal wall musculature of the psoas muscle is occasionally noted. Variable contrast enhancement is seen because myxoid components do not enhance. CT is useful for planning surgical resection and predicting the patient’s prognosis. Large size, multiplicity, extensive infiltration, tethering, and encasement of small-bowel loops are poor prognostic CT findings. On MRI, there is low to intermediate T1 signal and intermediate T2 signal, with variable contrast enhancement after gadolinium injection.5 Interestingly, some lesions that do not enhance on CT may enhance on MRI with gadolinium. Treatment of mesenteric fibromatosis is controversial. Although surgery is an option, these lesions can recur and surgery predisposes these patients to complications of bowel obstruction and fistulas. Surgery involving cases of sporadic mesenteric fibromatosis are frequently curative. FAP mesenteric fibromatosis carries significantly higher rates of recurrence and disease is more difficult to control, so

surgery is often reserved to treat complications; nonsurgical methods are used initially.73 Extrapleural solitary fibrous tumor is a tumor of submesothelial origin with histology identical to that of solitary fibrous tumor of the pleura. Locations other than the pleura and mesentery include the lung, mediastinum, pericardium, peritoneum, extraperitoneal spaces, nose, and paranasal sinuses. The few case reports in the literature describe a well-defined mass with solid and cystic features on crosssectional modalities. The natural history is unknown, but because solitary fibrous tumor of the pleura may be aggressive, long-term follow-up is necessary.73 Cystic lymphangioma of the mesentery is a rare lesion but the most common cystic lesion involving mesentery. It can occur as a solitary cystic lesion or a more diffuse form (see Fig. 47-29). This tumor can occur in retroperitoneal locations alone or along with mesentery. Although benign in nature, it can cause symptoms if it attains large size and occurs in critical locations. This tumor should be differentiated from other types of cystic lesions of mesentery and omentum. These include enteric cyst, enteric duplication, mesothelial cyst, peritoneal inclusion cyst, and nonpancreatic pseudocyst.128 It is difficult to diagnose the different types of cystic lesions by imaging studies alone, particularly solitary forms.142 CT and MRI can provide important preoperative information regarding anatomic location, organ involvement, cyst size, and complications. A typical feature seen on a CT image is that multilocular masses often contain complex fluid. The presence of streaky low densities within a multicystic lesion is very suggestive of a cystic lymphangioma. MRI best displays the solid cystic

1508

PART II  CT and MR Imaging of the Whole Body

A

B

D

C FIG 47-38  SMA thrombosis and ileal ischemia. Axial CT (A and B) and coronal reformatted CT (C) images of the abdomen and pelvis show a filling defect (arrowhead) in SMA consistent with a thrombosis. Small bowel loops show poor enhancement and air in the wall (arrows). Surgical specimen shows transmural ischemic necrosis in the involved small bowel.

CHAPTER 47  Mesentery components, which aids in the preoperative diagnosis and surgical planning. Benign or malignant mesenchymal tumors of fat cell, muscle cell, nerve sheath cell, or connective tissue cell origin can develop in the mesentery very rarely (Fig. 47-44). GI stromal tumors can occur rarely as primary tumors of the mesentery, but it is often debatable whether they originated from the bowel wall or mesentery27,64,72,119,121 (see Fig. 47-26). Other common neoplasms originating from other sites that involve the mesentery include lymphoma, Kaposi’s sarcoma, and metastatic disease (e.g., carcinoid tumor; see later).71

Infectious and Inflammatory Disease Primary infectious and inflammatory diseases of the mesentery are rare. Most infectious and inflammatory diseases of the mesentery are

FIG 47-39  SMV thrombosis. Coronal reformatted image shows extensive occlusive thrombus (arrows) in the jejunal and ileal tributaries of the SMV in a patient who is hypercoagulable.

A

1509

secondary to infectious or inflammatory bowel disease or peritonitis and will be discussed further under “Secondary Mesenteric Diseases.” However, sclerosing mesenteritis, mesenteric lymphadenitis, and less importantly inflammatory pseudotumor are unique disease entities that occur as primary forms and deserve further mention.

Sclerosing Mesenteritis.  Sclerosing mesenteritis is a rare benign disorder characterized by nonspecific inflammation of mesenteric fat. It is distinguished histologically by varying amounts of fibrosis, chronic inflammation, and fat necrosis.30 This protean process has been known by a number of terms such as mesenteric panniculitis, mesenteric lipodystrophy, and retractile mesenteritis. Confusion surrounds these terms because they are often used synonymously in the literature and may describe histologic features or subtypes of different stages of the same disease. However, radiologically on CT there are two distinct forms: sclerosing mesenteritis and mesenteric panniculitis. Sclerosing mesenteritis usually involves the small bowel mesentery at its root but can sometimes affect the mesocolon and rarely the peripancreatic region, omentum, retroperitoneum, or pelvis.38,52 Its etiology remains unknown, although various causes have been suggested, including infection, ischemia, trauma, and vasculitis. There is a suggestion that it may link with immunoglobulin (Ig)G4-related sclerosing disease, because it has been associated with retroperitoneal fibrosis, Riedel’s thyroiditis, orbital pseudotumor, and sclerosing cholangitis. An association between prior abdominal surgery and sclerosing mesenteritis has been also reported. The average age of patients at presentation is 60 years, and the condition is more common in males than females. Patients may present with an abdominal mass, pain, bowel obstruction, bowel ischemia, or diarrhea. Not uncommonly, sclerosing mesenteritis may be an incidental finding in an asymptomatic patient. The CT appearance of sclerosing mesenteritis reflects the underlying histology of the lesion (see Fig. 47-14). Most commonly it presents as a soft tissue mass with variable enhancement (Fig. 47-45). The mass may show well-circumscribed or infiltrating margins, and lesions  may contain central calcification, possibly related to fat necrosis (Fig. 47-46). Larger masses may demonstrate cystic features, suggesting necrosis. Enlarged mesenteric or retroperitoneal nodes may be associated. Linear bands of fibrosis may radiate from the mass, affecting the small bowel by retraction and shortening of the mesentery rather than by direct invasion. The small bowel may then become kinked or fixed,

B FIG 47-40  Malrotation of the intestine. A and B, Two consecutive axial CTs show reversal of the SMA (arrow) and SMV (arrowhead) position, with SMV being to the left of the SMA.

1510

PART II  CT and MR Imaging of the Whole Body

A

B FIG 47-41  Left paraduodenal hernia. Axial (A) and coronal (B) CTs show contained clustered loops of small bowel in left upper abdomen. Note the anterior displacement of inferior mesenteric vein (arrow). (Courtesy Dr. Choi, Seoul National University Hospital, Seoul, Korea.)

A

B

C

D FIG 47-42  Right paraduodenal hernia in an asymptomatic patient. A to D, Serial contrast-enhanced axial CTs demonstrate mildly dilated small bowel loops herniating into Waldeyer’s fossa (curved arrow). The right colic vein (straight white arrows) is displaced anteriorly by the herniated bowel loops.

CHAPTER 47  Mesentery

A

1511

B FIG 47-43  Mesenteric and omental collateral vessels. A, Axial contrast-enhanced CT image shows a tangle of tortuous (long arrow) collateral vessels in the region of the pancreatic head in a patient with cavernous transformation of the portal vein from cryptogenic cirrhosis. Multiple small collaterals are also within the omentum (short arrow). B, More inferiorly numerous mesenteric collaterals are evident (long arrow), and small collaterals are also seen in the omental fat within an incisional hernia (short arrow).

resulting in obstruction; obstruction of lymphatics and veins may cause edematous thickening.52,67,73 On MRI, sclerosing mesenteritis has been reported to have signal intensity patterns consistent with fibrosis—low to intermediate signal intensity on T1- and low signal intensity on T2-weighted images.7,66 Sclerosing mesenteritis has also been reported to have high signal intensity on T2-weighted images,60,65 likely corresponding to the disease phase dominated by myxomatous change or active proliferation of fibrosis accompanied by hypervascularity.73 Lymphoma and small bowel carcinoid are the main differential diagnostic considerations for a soft tissue mass within the mesentery. Lymphoma, however, does not usually calcify unless previously treated and does not often cause bowel ischemia. Lymphoma is also more likely to demonstrate discrete enlarged nodes. Serotonin production from small bowel carcinoids can produce a fibrotic retraction that may cause desmoplastic kinking of the small bowel similar to the retraction observed at CT in cases of sclerosing mesenteritis. However, focal small bowel mural thickening or mass favors the diagnosis of carcinoid. Additionally, most small bowel carcinoid tumors are positive on somatostatin-receptor scintigraphy, and imaging with indium-111 pentetreotide can differentiate between the two diseases. Benign entities in the differential include atypical infectious etiologies such as Whipple’s disease, which may affect the mesentery and retroperitoneum with lipogranulomatous inflammation. A diagnosis of Whipple’s disease can be made by using a polymerase chain reaction to verify the causative bacillus. Actinomycosis can also cause an aggressive infiltrative soft tissue process at CT and may affect the mesentery (see Fig. 47-48). Additionally, Weber-Christian disease, a rare systemic inflammatory disorder of fat, may have multifocal areas of fat necrosis and inflammation in the mesentery that may be identical to sclerosing mesenteritis. Differentiation of Weber-Christian disease from sclerosing mesenteritis relies on identification of other features of the disease. Typically these patients also have lower extremity skin nodules, fevers, myalgias, arthritis, and arthralgia. Other benign fibrous lesions such as mesenteric fibromatosis, inflammatory pseudotumor, and extrapleural solitary fibrous tumor may also be considered within the differential. Definitive diagnosis of sclerosing mesenteritis requires histology, and surgical excisional biopsy is often needed for complete analysis.

Depending on the severity of symptoms, treatment may consist of simple observation, medical therapy with steroids, colchicine, orally administered progesterone, or immunosuppressive agents. Surgical resection may be indicated for those patients with complications such as bowel obstruction or perforation. Mesenteric panniculitis appears as subtle increased attenuation within the mesenteric fat, often at the root of the mesentery, with accompanying small nodes (Fig. 47-47; see Fig. 47-23). It may also exhibit a tumoral pseudocapsule consisting of a peripheral band of soft tissue attenuation that separates the normal mesentery from the inflammatory process118 and may show spared fat around vessels and lymph nodes—the fat ring sign.134 It does not progress or change for long periods and is mostly found incidentally on CT. It is self-limited and does not cause significant clinical issues. Although mesenteric panniculitis and sclerosing mesenteritis appear quite different on CT, it can be more difficult to distinguish them pathologically because both have inflammation, fat necrosis, and fibrosis. If fibrosis is the dominant feature, the diagnosis is sclerosing mesenteritis; if inflammation and fat necrosis are dominant features, the diagnosis is mesenteric panniculitis. These do not appear to be different stages of one disease spectrum, because no case of mesenteric panniculitis progressing to sclerosing mesenteritis has been reported. When one makes a diagnosis of mesenteric panniculitis, the rare possibility of mesenteric lymphoma, which has overlapping CT features with mesenteric panniculitis, should be excluded. Therefore it is essential to document stability of the lesion over a reasonable period of time to reach the diagnosis of mesenteric panniculitis by follow-up study or comparison with old studies if available. It is not known what causes mesenteric panniculitis or sclerosing mesenteritis. There are cases where CT findings of mesenteric panniculitis and possible causes  are conjoined: chronic pancreatitis, IBD, stigmas of prior abdominal surgery, and prior trauma. Misty mesentery is another confusing term. This is not a disease entity but rather CT descriptive terminology. Regardless of the etiology of the condition (e.g., edema, lymphedema, cellular infiltration by inflammation or neoplasm, fibrosis, bleeding in the mesentery), mesentery shows hazy increased density on CT24,93,120 (Fig. 47-48; see Figs. 47-16, 47-24, and 47-25). Most of these conditions are transient and

1512

PART II  CT and MR Imaging of the Whole Body

C

A

B

D

E FIG 47-44  Omental schwannoma. A, Sagittal T2-weighted MRI shows heterogeneously bright round mass (arrow) in the greater omentum. Postgadolinium early enhanced T1-weighted axial MRI (B) shows peripheral enhancing mass (arrow), and delayed enhanced coronal MRI (C) shows further filling in of peripheral enhancement, but central part does not get enhanced, representing necrosis. D, Axial postcontrast CT shows similar enhancement pattern (arrow). Note mobility of the mass from MRI to CT. A large fibroid in the uterine fundus (arrowheads) was incidentally noted on sagittal MRIs. E, Surgical specimen shows a well-defined yellow mass in the mesentery. (Courtesy Dr. Choi, Seoul National University Hospital, Seoul, Korea.)

CHAPTER 47  Mesentery

A

B

C

D

1513

FIG 47-45  Sclerosing mesenteritis. A to C, Axial CT images show a well-defined soft tissue mass (arrows) in the small bowel mesentery. There is mild mesenteric vessel engorgement (arrowhead). D, Surgical specimen shows yellow mass in the mesentery with speculated border (arrows).

temporary, and often the cause of the condition is readily apparent120 (see Figs. 47-16 and 47-17). Mesenteric panniculitis and sclerosing mesenteritis can be described as “misty mesentery” on CT initially, but chronicity of the condition and other unique findings will separate these two entities from other causes.

Inflammatory Pseudotumor.  Inflammatory pseudotumor (inflammatory myofibroblastic tumor, extrapulmonary inflammatory  pseudotumor, plasma cell granuloma, plasma cell pseudotumor, inflammatory fibrosarcoma) is a chronic inflammatory disorder of uncertain etiology.109 Although it most frequently occurs in lung and orbit, it can occur in many places including the mesentery (Fig. 47-49; see Fig. 47-15). Inflammatory pseudotumor may be due to the sequela of occult infection, minor trauma, or prior surgery and can be difficult to differentiate from malignant neoplasm or sclerosing mesenteritis. It most commonly occurs in the pediatric or young adult population. Symptoms include fever, malaise, weight loss, and abdominal pain. Surgical resection is usually curative.39,106

Mesenteric Adenitis.  Mesenteric adenitis has been characterized as benign inflammation of mesentery lymph nodes, usually in the right lower quadrant (RLQ). It has been subdivided into primary and secondary forms. Primary mesenteric adenitis has been defined as rightsided mesenteric adenopathy without identifiable concomitant acute inflammatory process, and with or without wall thickening of the terminal ileum. Although designated as primary, these cases are thought to be due to an infectious process (e.g., Yersinia enterocolitica and Yersinia pseudotuberculosis) causing a terminal ileitis that may not be apparent on imaging. Secondary mesenteric adenitis has been

described as mesenteric adenopathy (usually right-sided) associated with an identifiable intraabdominal inflammatory process such as Crohn’s disease, infectious colitis, ulcerative colitis, or diverticulitis. Although appendicitis can also cause enlarged right-sided mesenteric lymph nodes, mesenteric adenitis is often considered important in the differential diagnosis of appendicitis, especially in children. Therefore appendicitis is not usually classified as a form of secondary mesenteric adenitis even though enlarged right-sided lymph nodes are commonly identified with this process. Commonly the signs and symptoms of appendicitis and mesenteric adenitis overlap, including RLQ pain, nausea, and vomiting. Rao et al.111 found in a retrospective study that 50 of 651 (7.7%) patients who had an admission diagnosis of appendicitis had a discharge diagnosis of mesenteric adenitis, and that prospectively 18 of 100 (18%) patients had mesenteric adenitis with signs and symptoms of appendicitis. Distinguishing these two entities is important because mesenteric adenitis is treated conservatively, whereas appendicitis is a surgical disease. A CT scan may be the best way of distinguishing these entities. A normal appendix would effectively exclude appendicitis. Detection of a cluster of three or more enlarged RLQ lymph nodes (>5 mm) in the presence of a normal appendix and no other acute inflammatory process except mild (<5 mm) wall thickening of the terminal ileum would likely represent primary mesenteric adenitis, especially in the correct clinical setting (see Fig. 47-34). Finding enlarged mesenteric lymph nodes (>5 mm) with a concomitant intraabdominal inflammatory process (other than appendicitis) would represent secondary mesenteric adenitis. Confusion exists in the literature over what to do with lymph nodes incidentally found in the mesentery. Macari et al.83 in 2002 did not find

1514

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-46  Sclerosing mesenteritis with dense calcification. A, Plain film of the abdomen shows calcification (arrows) in the left side of the abdomen. B to D, Axial CT images show an ill-defined soft tissue mass (black arrows in B) in the small bowel mesentery, with dense calcification (arrowheads in C). There is dilated and thickened small bowel (white arrows in D).

A

B

FIG 47-47  A to C, Mesenteric panniculitis. Axial CT shows seg-

C

mental increased attenuation in the small bowel mesentery, separated from the adjacent fat by a tumoral pseudocapsule (arrows in C). There are several enlarged lymph nodes in the mesentery (arrowhead in A).

A

B

C

D FIG 47-48  A to D, Actinomycosis. Axial CT shows heterogeneously and irregularly enhanced mass in the mesentery, omentum, and abdominal wall (arrows in C and D). Hazy omentum is noted (arrowheads in A).

1516

PART II  CT and MR Imaging of the Whole Body

B

A

C

D FIG 47-49  Inflammatory pseudotumor. Axial (A and B) and coronal reformatted (C and D) CTs of the abdomen and pelvis show spiculated soft tissue mass (arrows) in the omentum. Hazy omentum (arrowhead) is noted.

RLQ mesenteric nodes in a group of 60 patients who had CT scans for abdominal trauma, and inferred that otherwise healthy patients did not have abnormal RLQ lymph nodes. The cutoff used for abnormal lymph node size was 5 mm or greater in short axis. Lucey et al.81 in 2004 found that 47 of 120 (39%) patients had mesenteric lymph nodes larger than 3 mm in short axis (mean size, 3.6 mm [range 3-6 mm]; mean size of largest lymph node, 4.6 mm [range 3-9 mm]), who had otherwise negative CT scans for blunt abdominal trauma. Five of these 47 patients had these lymph nodes in the RLQ. The discrepancies may partially be explained by the size criteria used (5 mm vs. 3 mm) and possibly be due to improving multidetector computed tomography (MDCT) and picture archiving and communication system (PACS) technologies that may facilitate detection of smaller lymph nodes. Given these findings, we think it would be wise to diagnose primary mesenteric adenitis only in the presence of clinical symptoms such as abdominal pain. It is likely that as imaging technology improves, more and more incidental mesenteric lymph nodes will be detected and may not be of clinical significance.

Trauma Injury to the mesentery from blunt abdominal trauma is relatively uncommon, with variable incidences of 1% to 13% reported in patients undergoing emergent laparotomy.85,99,139 Despite its infrequency, mesenteric injury from blunt abdominal trauma is a significant source of morbidity and mortality, often the result of a delay in diagnosis. CT has been shown to be accurate in detecting mesenteric injury and is the diagnostic test of choice in hemodynamically stable patients who have suffered blunt trauma.85 In recent years CT has been found in both surgical and imaging studies to have reasonably high accuracy for detecting and characterizing mesenteric injuries. Numerous CT findings of mesenteric injury from blunt abdominal trauma have been well described in the literature.12-15,28,34,43,46,62,98,99,114,129 Penetrating injury to the mesentery, on the other hand, typically coexists with other severe injury and is typically diagnosed by surgical exploration rather than by imaging, owing to the patient’s hemodynamic instability, often from catastrophic intraabdominal hemorrhage (Fig. 47-50).

CHAPTER 47  Mesentery

1517

A

C

B FIG 47-50  Mesenteric injuries and hematomas in a patient who had been in a motor vehicle collision. Axial precontrast (A), venous (B), and coronal reformatted (C) CTs of the abdomen and pelvis show a large hematoma on the left side (arrows). The high-attenuated contrast material (arrowhead) surrounded by a lowerattenuation hematoma is noted in the mesentery, suggesting mesenteric injuries with active bleeding.

Mesenteric injury may include contusion, laceration, and complete transection. Two mechanisms are thought to cause injury to the mesentery in blunt trauma: (1) crush or compressive force against the spinal column and (2) shear forces at the mesenteric attachments.53 The most commonly reported cause of mesenteric injury is a motor vehicle accident involving the use of seat belts—a situation where compressive and shear forces are both likely involved.53,117,139 The “seatbelt syndrome” is characterized by injuries in the plane of the seatbelt, often involving the mesenteries of the small bowel and colon. Patients may also have lumbar spine injuries (Chance fracture) and cutaneous bruising in the distribution of the seatbelt (“seatbelt sign”) (see Fig. 47-17). Additionally, injury to the small bowel mesentery is more frequent than injury to the colonic mesenteries,99 the most common locations involving the ligament of Treitz and the ileocecal region. These points of fixation of the mesentery are thought to be more susceptible to shearing forces.

Injuries to the mesentery result in an array of pathophysiologic processes. Laceration of the vessels may cause significant hemorrhage with subsequent hypovolemia and bowel ischemia. If the injury is unrecognized, bowel necrosis, peritonitis, and sepsis may ensue. Unrepaired small lacerations in the mesentery may allow for the formation of internal hernias and resultant bowel obstruction. Lastly, although rare, small tears or contusions of the mesentery may lead to partial-or full-thickness bowel ischemia with attendant mucosal ulceration, scarring, and bowel stenosis, eventually resulting in bowel obstruction. In the hemodynamically stable patient, MDCT plays a major role in the evaluation of mesenteric injury.28,64 CT findings that suggest mesenteric injury include active extravasation of intravenous (IV) contrast into the mesentery, mesenteric hematoma or infiltration associated with bowel wall thickening, isolated mesenteric hematoma or infiltration, and free intraperitoneal fluid (Figs. 47-51 and 47-52). Other morphologic vascular changes on MDCT, such as vascular irregularity or abrupt termination of vessels, have also been described.15 Of

1518

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-51  Mesenteric hemorrhage. A to D, Serial axial CTs demonstrate “misty mesentery,” increased attenuation in the small bowel mesentery (arrows) secondary to mesenteric bleeding in a patient following abdominal trauma from a motor vehicle accident.

H

H

A

B FIG 47-52  Iatrogenic mesenteric trauma. A and B, Axial nonenhanced CT in a patient who developed mesenteric hematoma (H) following bowel resection, with resultant obstruction (arrows) of more proximal small bowel loops.

CHAPTER 47  Mesentery these findings, active extravasation of IV contrast and bowel wall thickening associated with mesenteric hematoma/infiltration are the most specific signs for determining the need for surgical intervention. Active extravasation of IV contrast into the mesentery has a specificity of 100%15,28,62 for the diagnosis of significant injury and is usually an indication for urgent laparotomy. Despite its high specificity, this finding is not common. Bowel wall thickening associated with mesenteric infiltration or hematoma is highly suggestive of significant bowel injury requiring surgery.28,62 Again, although specific, these findings are not common. Mesenteric findings are more common when bowel injury is along the mesenteric border. The significance of an isolated mesenteric hematoma or isolated mesenteric infiltration is less clear with respect to surgical intervention. Although a hematoma is indicative of vascular laceration, this finding does not necessarily indicate a need for surgery. Furthermore, a number of studies have shown that patients with mesenteric infiltration or isolated hematoma who did not undergo surgery had resolution of findings on follow-up CT.95 Unexplained intraperitoneal free fluid (i.e., in the absence of solid organ injury, urinary bladder rupture, previous peritoneal lavage) is not an uncommon finding in patients with mesenteric injury requiring surgery.13,28,62 The presence of this finding in blunt trauma patients raises the possibility of surgical mesenteric injury, but it may not mandate laparotomy. A large prospective multiinstitutional trial showed a low (8%) rate of surgical bowel injury in the setting of isolated free fluid,77 and other retrospective trials have also noted that unexplained free fluid is not predictive of surgical bowel or mesenteric injury.28,62,85 Rather, these patients may benefit from clinical monitoring and follow-up CT.

Vascular Disease Mesenteric ischemia is characterized by diminished blood flow and an inadequate supply of oxygen to meet the metabolic demands of the bowel. Depending upon the severity of the insult, mesenteric ischemia may manifest variably from a transient change in bowel activity to frank bowel infarction and necrosis. Radiologic studies may detect the ischemic changes in the affected bowel loops and mesentery and determine the cause of ischemia. Common but nonspecific findings detectable on both CT and MRI in mesenteric ischemia include bowel distention, bowel wall thickening, mesenteric edema, and ascites. More specific findings may include lack of bowel wall enhancement and pneumatosis. In the past, catheter angiography was the gold standard for evaluating the mesenteric vasculature. However, with more recent developments in CT technology, MDCT angiography is probably the most frequently used technique for the diagnosis of mesenteric ischemia, with contrast-enhanced three-dimensional (3D) MR angiography (MRA) also being widely used.10,47,51,50,122,123 Mesenteric ischemia comprises a wide spectrum of disease processes that has traditionally been divided into acute and chronic causes given the stark contrast in clinical presentations.

Acute Mesenteric Ischemia.  Acute mesenteric ischemia (AMI) is a potentially fatal vascular emergency with overall mortality of 60% to 80%.101,140 Clinical presentation is nonspecific in most cases but has been classically characterized by sudden severe abdominal pain out of proportion to physical exam findings. AMI can be categorized into four specific types by cause: arterial embolism, arterial thrombosis, nonocclusive causes, and venous thrombosis. Emboli to the SMA are the most frequent cause of AMI, responsible for approximately 50% of cases.47,101 Most emboli are cardiac in origin

1519

and may dislodge from a left atrioventricular thrombus (in the setting of atrial fibrillation or myocardial infarction) or valvular thrombus. More than 95% of patients with SMA occlusion have documented histories of cardiac disease, and nearly a third have a history of a preceding embolic event. The onset of symptoms is usually sudden as a result of the poorly developed collateral circulation. Acute severe abdominal pain may be associated with diarrhea that may become bloody. Most emboli wedge at branch points in the mid- to distal SMA distal to the middle colic artery, allowing the inferior pancreaticoduodenal branches to be perfused and the proximal jejunum to be spared101,140 (see Fig. 47-37). The angiographic hallmark of an embolic occlusion is abrupt termination of the vessel (cutoff sign). Nonocclusive emboli are usually seen as filling defects in the vessel lumen. Typically no collateral vessels are noted. Both CT angiography (CTA) and contrast-enhanced MRA can depict these findings. Arterial thrombosis accounts for approximately 25% of AMI and occurs in the setting of severe atherosclerotic disease (see Fig. 47-38). The slowly progressive nature of atherosclerosis usually allows for development of a collateral circulation so that ischemia occurs only after the final artery supplying a bowel segment occludes. In up to 50% of cases, a history of intestinal angina is present. Additionally these patients tend to present with more extensive bowel ischemia/infarction than those with embolism. The perioperative mortality is high, ranging from 70% to 100% because of the delay in diagnosis, extent of bowel ischemia, and need for complex surgical revascularization. Unlike emboli, occlusion due to thrombosis usually involves the ostia and proximal 2 cm of the SMA. Typically no defined intraluminal filling defect or meniscus is noted on imaging, and a collateral circulation may be evident. Nonocclusive mesenteric ischemia accounts for 20% of AMI  and typically occurs in patients with cardiac dysfunction. The pathophysiology of nonocclusive mesenteric ischemia appears to involve splanchnic vasoconstriction in response to hypovolemia, hypotension, decreased cardiac output, or drugs such as digoxin, diuretics, or β-blockers. The frequency of nonocclusive mesenteric ischemia among cardiac surgery and hemodialysis patients has increased over the last few decades, with 9% to 20% of deaths of hemodialysis patients attributable to nonocclusive mesenteric ischemia.140 Other risk factors for nonocclusive mesenteric ischemia include cardiac (infarction, congestive failure, aortic insufficiency), renal, or hepatic disease. Bowel changes are similar to those of occlusive mesenteric ischemia; however, the mesenteric vasculature remains patent. Watershed areas such as the splenic flexure of the colon are more vulnerable to nonocclusive mesenteric ischemia. CTA findings are similar to findings evident on conventional catheter angiography but are often subtle and difficult to diagnose. The mesenteric vessels are patent but may appear diffusely small in caliber with fewer side branches visible. Reduced opacification of the bowel wall may also be evident. Acute mesenteric venous thrombosis is the least common cause, occurring in up to 10% of cases of small bowel ischemia. It is generally seen in younger patients, typically in the setting of abdominal surgery. Other predisposing factors include portal hypertension, abdominal or pelvic inflammation, trauma, renal or cardiac disease, oral contraceptive use, hematologic abnormalities, or a hypercoagulable state.10 The SMV is involved in 95% of cases (see Figs. 47-21 and 47-39). Involvement of the IMV and large bowel are uncommon. Mesenteric venous thrombosis often affects a focal segment of bowel with edema, hemorrhage, and mucosal sloughing. Occlusion of the proximal SMV rarely causes bowel necrosis, whereas occlusion of the distal SMV or its branches involving the vasa recta causes bowel necrosis by preventing development of collateral flow. Both CTA and contrast-enhanced MRA have been shown to be highly accurate for the evaluation of thrombosis.44,137

1520

PART II  CT and MR Imaging of the Whole Body

Chronic Mesenteric Ischemia.  Patients with chronic mesenteric ischemia (CMI) typically have a more benign and indolent presentation than those with AMI, typically presenting with mesenteric angina—postprandial abdominal pain, fear of food, and weight loss. Also in contrast to AMI, the majority of cases of CMI are due to atherosclerotic stenosis or occlusion of the proximal portions of the mesenteric arteries. Although atherosclerosis of the mesenteric vessels is common, CMI is relatively uncommon owing to the extensive collateral circulation within the mesentery. It is generally thought that at least two of the three main vessels must be affected before symptoms present, and at least 85% of cases do involve both the SMA and celiac axis.96 Involvement of only one vessel may occur but is uncommon. Unlike in patients with AMI, the small intestine usually appears normal in patients with CMI unless an acute thrombus is also present. In patients with CMI, mesenteric collateral vessels may form to preserve adequate blood flow to the intestines. These can also be identified at MDCT with 3D reformatting. Nonatherosclerotic causes of CMI are uncommon but include vasculitis due to connective tissue disorders and postradiation vasculitis. Fibromuscular dysplasia, abdominal coarctation, neurofibromatosis, and idiopathic fibrosis are also rare causes. Findings associated with vasculitides include stenosis, occlusion, aneurysm formation, vascular wall thickening and enhancement. The radiologic findings in various types of vasculitides may significantly overlap. Nevertheless, vasculitides should be considered whenever mesenteric ischemic changes are observed in young patients, coexist at unusual sites such as the stomach, duodenum, and rectum, involve both the small and large intestine, and are associated with genitourinary involvement or splenomegaly. Radiation enteritis is an infrequent sequela of abdominopelvic radiation therapy but a significant cause of morbidity. Radiation vasculitis and associated enteritis may develop as soon as 6 to 24 months after treatment but can present up to 20 years after the initial insult.63 The underlying pathologic process is an endarteritis obliterans— compromise of the microvascular circulation.17 The involved bowel segments depend on the radiation field, and this finding may point to the underlying cause. Diagnosis of chronic radiation enteritis is often problematic because of the latency period. CT findings are nonspecific but include thickening of the bowel wall and adjacent mesentery, increased density of the mesenteric fat, and confluent or adherent bowel loops. Barium studies may also demonstrate thickened valvulae conniventes, wall thickening, effacement of the mucosal fold pattern, ulceration, stenosis, and adhesions. Sinuses and fistulas may be demonstrated by either barium study or CT exams. Because of the variability and nonspecificity of findings in mesenteric ischemia, particularly in cases of nonocclusive and chronic ischemia, careful clinical correlation with CT findings is essential in evaluating patients with suspected mesenteric ischemia.

Appendices Epiploicae and Ischemic Infarction of Omentum

Epiploic appendagitis.  Appendices epiploicae are pedunculated fatty structures that arise from the serosal surface of the colon; they may measure 0.5 to 5 cm long and extend from the cecum to the rectosigmoid junction. Because of their pedunculated shape and mobility, appendices epiploicae may torse, leading to ischemic or hemorrhagic infarction. The resulting focal inflammatory process is termed epiploic appendagitis. Clinically, acute epiploic appendagitis manifests with acute onset of pain, typically in the lower quadrants, and is often clinically mistaken for appendicitis or diverticulitis. However, patients with

FIG 47-53  Epiploic appendagitis. Axial contrast-enhanced CT shows an ovoid fatty lesion adjacent to the cecum, with a hyperattenuating rim and surrounding inflammatory changes (short arrow). There is a vague hyperattenuating central area representing a thrombosed vessel or hemorrhage (long arrow).

epiploic appendagitis are usually afebrile with a normal white blood cell count.107,126 Epiploic appendagitis most commonly occurs in the sigmoid colon, followed by the descending colon and right colon. Typical CT findings include a paracolonic ovoid fatty mass (representing the infarcted or inflamed appendix epiploica) with surrounding inflammatory change, a well-defined hyperattenuating rim around the mass (reflecting the inflamed peritoneal lining), and on occasion, central high attenuation possibly representing thrombosed vessels or hemorrhage107,112,126 (Fig. 47-53). Mild local reactive thickening of the adjacent colonic wall may also be evident. Acute epiploic appendagitis is treated conservatively with oral antiinflammatory medication; antibiotics are not routinely indicated. The condition is self-limited and most patients recover with conservative management in less than 2 weeks. CT findings may persist longer but typically resolve by 6 months. Omental infarction.  The greater omentum, composed of a double layer of peritoneum, hangs from the greater curvature of the stomach and the proximal part of the duodenum covering the small bowel. At CT it appears as a band of fatty tissue of variable thickness just beneath the anterior abdominal wall and anterior to the stomach, transverse colon, and small bowel. Multiple small omental vessels are usually visible. Similar to epiploic appendagitis, torsion of the omentum or venous thrombosis of the omental vessels may lead to omental infarction, which usually occurs in the right lower or right upper quadrant, clinically mimicking appendicitis or cholecystitis. This bias for the right side of the abdomen is theorized to be related to variant embryologic vascular development predisposing to right-sided venous thrombosis.31 Other contributing factors include obesity, strenuous activity, congestive heart failure, digitalis administration, recent abdominal surgery, and abdominal trauma.126 Most cases occur in adults, but approximately 15% of cases affect the pediatric population41 (Figs. 47-54 and 47-55). Trauma, including rapid rotary movements, sustained blows to the abdomen, sudden change in body position, coughing, and straining, have been implicated as possible factors that may initiate primary torsion of the omentum. Secondary torsion is more common and can be induced by torsion caused by adhesion of the omentum and pathologic foci such as surgical scars, inflammation, cysts, tumors, and hernias and by thrombosis due to various causes such as hypercoagulopathy and vascular abnormalities. Regardless of the cause, similar patterns of progression from edema and congestion

CHAPTER 47  Mesentery

1521

B

A

C

D FIG 47-54  Omentum infarction. Axial (A to C) and coronal reformatted (D) CTs of the abdomen and pelvis show an oval-shaped streaky infiltration in the right side of the omentum (arrows). Coronal reformatted CT image clearly demonstrates streaky infiltration in the right side of omentum (arrows).

due to venous stasis and thrombosis to hemorrhagic necrosis and extravasation of serosanguineous peritoneal fluid are histopathologically observed. CT findings of omental infarct may vary from hazy soft tissue infiltration to a large heterogeneous mass. Most commonly the inflammatory mass is in the RLQ anteromedial to the ascending colon or anterior to the transverse colon.110 Although omental infarction may be difficult to distinguish from epiploic appendagitis, the inflammatory mass associated with omental infarction is typically larger. Like epiploic appendagitis, however, the condition is self-limited and only conservative management is required.107,126

Mesenteric Collateral Circulation.  Recognition of arterial collaterals in the mesentery is important in detecting occlusions and ste-

noses of the major mesenteric branches in the evaluation of suspected mesenteric ischemia. The marginal artery of Drummond is situated along the mesenteric border of the colon and formed by the anastomotic continuation between the right, middle, and left colic arteries. The marginal artery of Drummond serves as the connection between the distal branches of the SMA and inferior mesenteric artery (IMA). Riolan’s arch, the principal pathway between proximal branches of  the SMA and IMA, is located more centrally within the mesentery and is an inconstant anastomosis between the middle and left colic arteries. The meandering artery is also within the mesentery and is a very large tortuous vessel communicating between the SMA and IMA. It potentially represents an enlarged Riolan’s arch. Barkow’s arch is another collateral pathway between the SMA and the gastroepiploic artery.

1522

PART II  CT and MR Imaging of the Whole Body

A

B

D

C

FIG 47-55  Omental infarcts in two different patients. A and B, Two consecutive axial enhanced CTs in a 6-year-old girl with chronic abdominal pain for 2 months demonstrate irregular soft tissue in the omental fat (arrows) in the right upper quadrant anterior to the ascending colon. C and D, Images from a PET/CT study obtained for evaluation of a solitary pulmonary nodule in an adult patient with RLQ abdominal pain shortly following colectomy demonstrate a large area of soft tissue infiltration in the RLQ (C). The fused PET/CT image (D) demonstrates mild FDG uptake consistent with inflammation.

Collateral venous pathways in the mesentery typically develop from portal hypertension or occlusion of the portomesenteric vasculature secondary to inflammatory or neoplastic pancreatic disease. Mesenteric varices from portal hypertension are not common but have been described as dilated and tortuous branches of the SMV in the leaves of the mesentery23 (see Fig. 47-43). Mesenteric venous collateral pathways may arise from either the SMV or IMV and ultimately drain into the systemic circulation through pelvic or retroperitoneal networks. Occasionally there may be direct shunts from the mesenteric veins to the inferior vena cava or renal veins. GI bleeding from lower intestinal varices is reported to occur in up to 5% of patients with portal hypertension.1,23 Omental collaterals are also infrequently observed but when present can mimic the appearance of omental implants (see Figs. 47-22 and 47-43).

SECONDARY MESENTERIC DISEASES Bowel Origin Various infectious, inflammatory, and neoplastic processes originating in the bowel can involve the adjacent mesentery and show vascular engorgement, mesenteric haziness, and mesenteric lymphadenopathy.

Infection.  Infections of the GI tract commonly involve the mesentery with adenopathy. Yersinia enterocolitica, which typically affects the distal ileum and sometimes mimics Crohn’s disease, may show enlarged nodes in the RLQ.133 Mycobacterium avium complex (MAC) often infects the intestines in an HIV patient with a CD4 cell count of 50/ mL or less and may demonstrate massive mesenteric adenopathy.105 Approximately 80% of the nodes infected with MAC are of homogeneous soft tissue attenuation; 20% show low central attenuation (see Fig. 47-35). Whipple’s disease is a systemic infection of the grampositive bacillus Tropheryma whippelii, which typically infects the small intestine, causing malabsorption. Whipple’s disease can demonstrate enlarged nodes with central low density, thought to be caused by high fat content. These nodes have attenuation values between 10 and 20 HU.35 GI tuberculosis (TB) is another infection frequently associated with enlarged mesenteric nodes. Tuberculous nodes also generally have a lower attenuation value (closer to that of fluid or fat) and may show peripheral enhancement on CT, as opposed to malignant or inflammatory nodes, which tend to show homogeneous enhancement.82 Focal infectious processes such as appendicitis and diverticulitis may cause inflammatory changes in mesenteric fat with reactive mesenteric adenopathy75 (see Figs. 47-18 and 47-33).

CHAPTER 47  Mesentery Inflammatory Disease.  The mesentery is frequently involved in Crohn’s disease, and the inflammatory changes are well depicted on CT and MRI25,40,69,84 (see Fig. 47-19), including fibrofatty proliferation of the mesentery and complications such as abscess or phlegmon. Hypervascularity of the mesentery with vascular dilatation, tortuosity, and prominence of the vasa recta produce the comb sign, an indication of active inflammation. Mesenteric adenopathy is more commonly seen with Crohn’s disease than with ulcerative colitis. The adenopathy is often discrete rather than coalescent, as seen in lymphoma. Nodes can be located in the mesenteric root, mesenteric periphery, or RLQ and show homogeneous enhancement at CT. Inflammatory changes extending from the bowel into the mesentery in cases of Crohn’s disease are similar to those seen in appendicitis and diverticulitis—soft tissue stranding, fluid, abscess, and sinus tracts/fistulas. Celiac disease, a malabsorptive autoimmune disease of the small intestine, may rarely be complicated by cavitating mesenteric adenopathy.48 The adenopathy associated with the cavitating mesenteric lymph node syndrome of celiac disease also has low CT attenuation. The nodes may appear as fluid or fat attenuation similar to that described in Whipple’s disease, TB, or MAC infection. However, lymph nodes containing fat-fluid levels are thought to be specific for celiac disease. Histopathologic examination reveals lymph nodes with a central cavity containing chylous fluid and a thin peripheral rim of fibrous material and atrophic lymph node structures. The pathogenesis of this phenomenon is poorly understood but is theorized to result from chronic excessive antigenic exposure of the immune system to the damaged intestinal mucosa, resulting in depletion of cellular lymph node elements. Alternatively, cavitary nodal changes may reflect a disturbance of the usual mesenteric lymphatic drainage.

Neoplastic.  Tumors of bowel origin may involve the mesentery by multiple pathways: direct extension, lymphatic extension, nodal metastasis, and intraperitoneal seeding. GI carcinoid tumor is the most common malignant neoplasm of the small intestine, arising from neuroendocrine cells in the intestinal mucosa or submucosa, most often occurring in the distal ileum. Approximately 40% to 80% of GI carcinoids spread to the mesentery, either by direct extension or through local lymphatics.16,104 At CT the mesenteric mass is typically detected first, manifesting as an enhancing soft tissue mass with linear bands radiating in the mesenteric fat (see Fig. 47-20). Calcification is evident in the mesenteric mass in 70% of

A

1523

cases. The radiating soft tissue bands represent fibrosis and desmoplasia caused by serotonin and other hormones secreted by the primary bowel tumor. Fibrosis in the mesentery may create a “spoke-wheel” or “sunburst” arrangement of mesenteric vessels. The desmoplastic reaction may also cause angulation and kinking of the small bowel, sometimes termed a hairpin turn. Additionally there may be associated thickening of adjacent small bowel loops secondary to tumor infiltration or caused by ischemia due to sclerosis of mesenteric vessels  or lymphatic blockage.121 Extensive mesenteric disease may produce miliary implants, large masses, or mesenteric caking. The primary tumor, in contrast, is small (usually <3.5 cm) and is more often occult. The primary lesion on CT or MRI may appear as a well-defined hypervascular enhancing mass or regional intestinal wall thickening.6 Primary colonic adenocarcinoma may demonstrate direct extension into the mesentery, manifesting at CT as thickening and infiltration of the pericolic fat or an ill-defined soft tissue mass extending into the fat. When pericolic inflammatory change is evident, this appearance may mimic diverticulitis. Studies have found that pericolic lymph nodes adjacent to the focal area of colonic thickening are more  commonly seen in patients with colon cancer, whereas pericolic and mesenteric inflammatory changes are more commonly seen in diverticulitis22,103 (Fig. 47-56). Pathways of mesenteric nodal metastases can be reliably predicted based on the site of the primary tumor. For example, regional lymph node metastases from cancers located in the left colon will occur along the mesocolic, left colic, and IMA nodal chains.18,19 Mucinous adenocarcinoma of the colon can also cause widespread intraperitoneal metastases that can also affect the mesentery and calcify.

Retroperitoneal Origin The pancreas lies in the anterior pararenal space of the retroperitoneum, with the transverse mesocolon attached along its anterior surface and the root of the small bowel mesentery located immediately inferior. Because the pancreas lacks a capsule, carcinoma of the pancreatic head or body may directly invade the transverse mesocolon  and root of the small bowel mesentery.57 Metastatic spread of tumor to the mesentery may also occur via lymphatics and the adjacent neural plexus. Approximately 40% of patients with newly diagnosed adenocarcinoma of the pancreas have unresectable locally advanced disease with tumor extension along the root of the mesentery and encasement of the major mesenteric vessels86 (Fig. 47-57). Inflammatory

B FIG 47-56  Adenocarcinoma of the sigmoid colon. A and B, Contrast-enhanced CT showing circumferential thickening (white arrows) of a focal segment of sigmoid colon, with mild haziness in the adjacent fat. Enlarged pericolonic lymph nodes (black arrows) are evident.

1524

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 47-57  Pancreatic adenocarcinoma. A to C, Coronal reformatted CT of the abdomen and pelvis show an ill-defined hypodense tumor in the pancreatic uncinate process encasing the SMA and SMV (arrows).

complications of pancreatitis including fluid and phlegmonous changes also can spread to the small bowel mesentery and transverse mesocolon (Fig. 47-58; see Figs. 47-9 and 47-16),57 so pancreatic fluid can be seen in more remote sites adjacent to ileal loops, cecum, and transverse colon.70,94,100

Peritoneal Origin There are basically two different diseases in the peritoneum: peritonitis and primary and secondary neoplastic disease.29 Both involve mesentery secondarily. Ordinary bacterial peritonitis shows ascites with minimal peritoneal changes. It is often difficult to make a diagnosis of spontaneous bacterial peritonitis with image findings alone. Occasionally bacterial peritonitis secondary to other infections or as

complication of inflammatory bowel disease can spread to the mesentery (Figs. 47-59 and 47-60). Tuberculous peritonitis (Fig. 47-61) and peritoneal involvement by actinomycosis (see Fig. 47-48) are associated with profound peritoneal and/or mesenteric changes that often mimic peritoneal carcinomatosis. Tubercular involvement of the mesentery is less commonly seen today in developed countries but still may be seen in populations where TB is endemic and in immunocompromised patients. Typically TB involvement of the mesentery is the result of GI TB spread into the peritoneum. The pathophysiology of GI TB and subsequent spread is thought to start from swallowing a massive dose of TB bacilli that later pass into tubular glands of the bowel mucosa. The bacilli can then cause ulcerative or hypertrophic changes in the bowel mucosa.

CHAPTER 47  Mesentery

1525

A

A

A

A

C

B

FIG 47-58  Pancreatitis. A to C, Contrast-enhanced CT in a patient with pancreatitis shows fluid adjacent to the pancreatic body (arrow in A). Associated pancreatic fluid is within the transverse mesocolon (arrow in B). The small bowel mesentery is hazy (arrow in C). There is moderate ascites (A).

Eventually bacilli may be carried from the submucosa to mesenteric lymph nodes that drain those affected segments of bowel by hematogenous, lymphatic, or direct spread from the serosa. These infected lymph nodes often have a low-density or caseating appearance on CT. Rupture of caseating lymph nodes is thought to cause TB peritonitis. Three types of TB peritonitis have been described: (1) wet type, with ascites or pockets of loculated fluid, is the most common; (2) dry caseous nodules or adhesions creating a “plastic” abdomen; and (3) fixed fibrotic, characterized by large omental masses, matted and tethered bowel loops, and occasionally ascites.32,56 The appearance of peritoneal TB can be confused with peritoneal carcinomatosis (see Fig. 47-61). Mesenteric appearance with micronodular or macronodular (>5 mm) lesions—some of which are calcified—as well as thickening of mesenteric leaves and loss of normal mesenteric configuration is thought to be more commonly seen in patients with TB peritonitis than with carcinomatosis. Lymph node distribution is nonspecific, although calcification involving lymph nodes is thought to be more associated with TB. Ascites is commonly seen in both, but the amount of ascites is thought to be less with TB than carcinomatosis.42,79 Abdominal actinomycosis is a severe peritoneal infection that can also involve the mesentery. Presence of a long-standing intrauterine device is a known risk factor in young women. On CT, multiple solid

or encapsulated peritoneal and mesenteric masses that demonstrate enhancement may be seen (see Fig. 47-48). These masslike lesions represent abscesses that contain Actinomyces israelii. Resolution of these masses is seen over several weeks after antibiotic treatment.68 Peritoneal seeding of malignancy occurs most notoriously from ovarian cancer (Fig. 47-62).74 GI tract malignancy (Fig. 47-63) and pancreatic and biliary cancer commonly spread into the peritoneum and result in peritoneal carcinomatosis.11,90 Occasionally extraabdominal malignancy can metastasize to the peritoneum. A good example of this is breast cancer. Regardless of the primary tumor, the end result of peritoneal carcinomatosis has similar radiologic findings: ascites, serosal implant of tumor, omental caking, and marked enhancement of the thickened peritoneum, which is most strikingly seen on enhanced MRI. Some forms of lymphoma, notoriously effusion lymphoma, involve the peritoneum and show exactly the same radiologic findings as peritoneal carcinomatosis. This is called peritoneal lymphomatosis. Malignant peritoneal mesothelioma arises from the mesothelial cells lining the serosal surfaces of the peritoneal cavity. Radiologically this tumor is indistinguishable from typical peritoneal carcinomatosis (Fig. 47-64). Histologically it is also challenging to distinguish this tumor from peritoneal carcinomatosis due to adenocarcinoma. Often Text continued on p. 1530

A

B

FIG 47-59  Focal bacterial mesenteritis. A to C, Three consecu-

C

A

tive axial CTs of the abdomen show marked focal fat stranding in the mesentery (arrows) resembling epiploic appendagitis, following partial ileocolic resection for active Crohn’s disease. A right psoas abscess (arrowhead in C) is also complicated.

B

FIG 47-60  Bacterial peritonitis. A to C, Three consecutive axial

C

CTs through the abdomen show moderate fat stranding in the greater omentum (white arrows) and ascites as a complication of peritoneal dialysis. Note enhancement of the peritoneum (arrowheads) and peritoneal dialysis catheter (black arrow in C) in the RLQ.

CHAPTER 47  Mesentery

A

B

C

D

E

F FIG 47-61  Tuberculous peritonitis. Axial (A to D) and coronal reformatted (E and F) CTs of the abdomen and pelvis depict a large amount of ascites with infiltration in the omentum (arrows) and mesentery. Uniform thickening of the peritoneum is noted (arrowheads).

1527

1528

A

PART II  CT and MR Imaging of the Whole Body

B

C FIG 47-62  Peritoneal carcinomatosis. A to C, CTs at three consecutive levels of the abdomen and pelvis show omental caking (white arrow) and nodular tumor implantation onto pleats of mesentery (arrowheads). Moderate ascites and enhancement of parietal peritoneum (black arrows) are also noted.

CHAPTER 47  Mesentery

A

B

C

D FIG 47-63  Pseudomyxoma peritonei with appendiceal mucocele. Axial CT (A to C) and coronal reformatted CT (D) of the abdomen and pelvis show a small amount of ascites with infiltration in the mesentery (arrows). Appendix is enlarged with thin wall calcification (arrowheads).

1529

1530

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 47-64  Peritoneal malignant mesothelioma. A to D, Axial CTs of the abdomen show omental caking (arrows) with a small amount of ascites (arrowheads). Uniform thickening of the peritoneum is noted.

open biopsy is needed for a correct diagnosis. The majority of patients have had exposure to asbestos. Therefore it is often associated with pleural manifestations of asbestos, such as pleural calcifications, pleural thickening, and pleural effusions.121

Systemic Origin Systemic diseases affect the mesentery in many ways. Generalized anasarca due to any cause (e.g., congestive heart failure, septicemia, terminal cases of malignancy) causes edematous change of the mesentery. Mesenteric edema may also be associated with liver cirrhosis, nephrosis, vasculitis, or fluid overload.93 An increased bleeding tendency may cause mesenteric hematoma. Lymphoma is the most common malignant neoplasm to affect the mesentery. Involvement of the mesentery by lymphoma is seen commonly with non-Hodgkin’s lymphoma in more than 50% of patients at the time of presentation. Approximately 30% to 50% of cases of non-Hodgkin’s lymphoma spread to mesenteric lymph nodes, which commonly extend from retroperitoneal involvement.141 Enlarged nodes in lymphoma can present with three patterns of involvement: in the mesenteric root, scattered throughout the peripheral mesentery, or mixed root-peripheral pattern.141 The lymph nodes may be small and discrete in early lymphoma but may progressively coalesce to form a conglomerate soft tissue mass (see Fig. 47-31). Lymphoma is a soft tumor that tends to grow around and displace normal anatomic structures such as bowel or vessels. These lympho-

matous mesenteric masses may calcify after treatment of lymphoma. Calcification is rare without treatment (<1%). Mesenteric lymph node involvement is not usually associated with lymphomatous involvement of small or large bowel. Lymphoma involving the mesentery has similar attenuation to soft tissue and usually enhances homogeneously. Mesenteric lymphadenopathy in patients with lymphoma does not necessarily suggest active disease. There may be persistent lymphadenopathy with or without calcification even after treatment. Comparison with prior CT should carefully measure lymph node size to note any difference over time. Enlarged but stable lymph nodes are commonly seen in treated lymphoma, but increasing lymph node size is suspicious for recurrence.45,81,82,121 Kaposi’s sarcoma is a low-grade mesenchymal tumor usually seen in AIDS patients that involves the blood and lymphatic vessels, affecting primarily the skin and causing disseminated disease in a variety of organs. Involvement of the mesentery usually consists of homogeneously enhancing lymph nodes that can be significantly enlarged, measuring 3 to 5 cm in short-axis dimension, and may coalesce to form a masslike conglomeration. Lymphadenopathy involving other areas is also seen, including the porta hepatis, retroperitoneum, and pelvis. Involvement of the bowel, liver, and spleen is often seen in conjunction with lymphadenopathy.45,81,82,113 Connective tissue disorders such as systemic lupus erythematosus, systemic sclerosis, and rheumatoid arthritis may also result in mesenteric lymphadenopathy, although lymphadenopathy is usually seen in

CHAPTER 47  Mesentery other sites in the body. These lymph nodes measure up to about 2 cm and enhance homogeneously on postcontrast images. Other conditions that can cause mesenteric lymphadenopathy include primary biliary cirrhosis, sarcoidosis, amyloidosis, mastocytosis, and HenochSchönlein purpura.81,82

REFERENCES 1. Ahn J, Cooper JM, Silberzweig JE, et al: Venographic appearance   of portosystemic collateral pathways. Br J Radiol 70:1302–1306, 1997. 2. Applegate KE, Anderson JM, Klatte EC: Intestinal malrotation in children: A problem-solving approach to the upper gastrointestinal series. Radiographics 26:1485–1500, 2006. 3. Auh YH, Lim JH, Kim KW, et al: Loculated fluid collections in hepatic fissures and recesses: CT appearance and potential pitfalls. Radiographics 14:529–540, 1994. 4. Auh YH, Rubenstein WA, Markisz JA, et al: Intraperitoneal paravesical spaces: CT delineation with US correlation. Radiology 159:311–317, 1986. 5. Azizi L, Balu M, Belkacem A, et al: MRI features of mesenteric   desmoid tumors in familial adenomatous polyposis. AJR 184:1128– 1135, 2005. 6. Bader TR, Semelka RC, Chiu VC, et al: MRI of carcinoid tumors: Spectrum of appearances in the gastrointestinal tract and liver. J Magn Reson Imaging 14:261–269, 2001. 7. Badiola-Varela CM, Sussman SK, Glickstein MF: Mesenteric panniculitis: Findings on CT, MRI, and angiography—Case report.   Clin Imaging 15:265–267, 1991. 8. Balfe DM: Diseases of the peritoneum. In Freeny PC, Stevenson GW, editors: Margulis and Burhenne’s alimentary tract radiology, ed 5, St. Louis, 1994, Mosby, pp 1795–1825. 9. Balfe DM, Mauro MA, Koehler RE, et al: Gastrohepatic ligament: Normal and pathologic CT anatomy. Radiology 150:485–490, 1984. 10. Bradbury MS, Kavanagh PV, Bechtold RE, et al: Mesenteric venous thrombosis: Diagnosis and noninvasive imaging. Radiographics 22:527–541, 2002. 11. Bradley RF, Stewart JH, Russell GB, et al: Pseudomyxoma peritonei   of appendiceal origin: A clinicopathologic analysis of 101 patients uniformly treated at a single institution, with literature review. Am J Surg Pathol 30:551–559, 2006. 12. Brasel KJ, Olson CJ, Stafford RE, et al: Incidence and significance of free fluid on abdominal computed tomographic scan in blunt trauma.   J Trauma 44:889–892, 1998. 13. Breen DJ, Janzen DL, Zwirewich CV, et al: Blunt bowel and mesenteric injury: Diagnostic performance of CT signs. J Comput Assist Tomogr 21:706–712, 1997. 14. Brody JM, Leighton DB, Murphy BL, et al: CT of blunt trauma of bowel and mesenteric injury: Typical findings and pitfalls in diagnosis. Radiographics 20:1525–1536, 2000. 15. Brofman N, Atri M, Hanson JM, et al: Evaluation of bowel and mesenteric blunt trauma with multidetector CT. Radiographics 26:1119–1131, 2006. 16. Burke AP, Thomas RM, Elsayed AM, et al: Carcinoids of the jejunum and ileum: An immunohistochemical and clinicopathologic study of 167 cases. Cancer 79:1086–1093, 1997. 17. Carr ND, Pullen BR, Hasleton PS, et al: Microvascular studies in human radiation bowel disease. Gut 25:448–454, 1984. 18. Charnsangavej C, Dubrow RA, Varma DGK, et al: CT of the mesocolon: Part 1. Anatomic considerations. Radiographics 13:1035– 1045, 1993. 19. Charnsangavej C, Dubrow RA, Varma DGK, et al: CT of the mesocolon: Part 2, Pathologic considerations. Radiographics 13:1309– 1322, 1993. 20. Reference deleted in proofs.

1531

21. Chen A, Yang F, Shih S, et al: CT Diagnosis of volvulus of the descending colon with persistent mesocolon. AJR Am J Roentgenol 180:1003–1006, 2003. 22. Chintapalli KN, Esola CC, Chopra S, et al: Pericolic mesenteric lymph nodes: An aid in distinguishing diverticulitis from cancer of the colon. AJR 169:1253–1255, 1997. 23. Cho K, Patel Y, Wachsberg R, et al: Varices in portal hypertension: Evaluation with CT. Radiographics 15:609–622, 1995. 24. Daskalogiannaki M, Voloudaki A, Prassopoulos P, et al: CT evaluation of mesenteric panniculitis: Prevalence and associated diseases. AJR 174:427–431, 2000. 25. Del Campo L, Arribas I, Valbuena M, et al: Spiral CT findings in active and remission phases in patients with Crohn disease. J Comput Assist Tomogr 25:792–797, 2001. 26. DeMeo J, Fulcher A, Austin R, Jr: Anatomic CT demonstration of the peritoneal spaces, ligaments, and mesenteries: Normal and pathologic processes. Radiographics 15:755–770, 1995. 27. Dinauer PA, Brixey CJ, Moncur JT, et al: Pathologic and MR imaging features of benign fibrous soft-tissue tumors in adults. Radiographics 27:173–187, 2007. 28. Dowe M, Shanmuganathan K, Mirvis S, et al: CT findings of mesenteric injury after blunt trauma: Implications for surgical intervention. AJR 168:425–428, 1997. 29. Reference deleted in proofs. 30. Emory TS, Monihan JM, Carr NJ, et al: Sclerosing mesenteritis, mesenteric panniculitis and mesenteric lipodystrophy: A single entity? Am J Surg Pathol 21:392–398, 1997. 31. Epstein LI, Lempke RE: Primary idiopathic segmental infarction of the greater omentum: Case report and collective review of the literature. Ann Surg 167:437–443, 1968. 32. Epstein BM, Mann JH: CT of abdominal tuberculosis. AJR 139:861– 866, 1982. 33. Fisher JK: CT diagnosis of volvulus in intestinal malrotation. Radiology 140:145–146, 1981. 34. Frick EJJ, Pasquale MDFACS, Cipolle MDFACS: Small-bowel   and mesentery injuries in blunt trauma. J Trauma 46:920–926, 1999. 35. Friedman HD, Hadfield TL, Lamy Y, et al: Whipple’s disease presenting as chronic wastage and abdominal lymphadenopathy. Diagn Microbiol Infect Dis 23:111–113, 1995. 36. Garg AG, Singh AK: Inflammatory fatty masses of the abdomen. Semin Ultrasound CT MR 29:378–385, 2008. 37. Reference deleted in proofs. 38. Ghanem N, Pache G, Bley T, et al: MR findings in a rare case of sclerosing mesenteritis of the mesocolon. J Magn Reson Imaging 21:632–636, 2005. 39. Gleason BC, Hornick JL: Inflammatory myofibroblastic tumours: Where are we now? J Clin Pathol 61:428–437, 2008. 40. Goldberg HI, Gore RM, Margulis AR, et al: Computed tomography in the evaluation of Crohn disease. AJR 140:277–282, 1983. 41. Grattan-Smith JD, Blews DE, Brand T: Omental infarction in   pediatric patients: sonographic and CT findings. AJR 178:1537–1539, 2002. 42. Ha HK, Jung JI, Lee MS, et al: CT differentiation of tuberculous peritonitis and peritoneal carcinomatosis. AJR 167:743–748, 1996. 43. Hagiwara A, Akiyoshi Y, Tesuo S, et al: Early diagnosis of small intestine rupture from blunt abdominal trauma using computed tomography: Significance of the streaky density within the mesentery. J Trauma 38:630–633, 1995. 44. Hagspiel KD, Leung DA, Angle JF, et al: MR angiography of the mesenteric vasculature. Radiol Clin North Am 40:867–886, 2002. 45. Hamrick-Turner J, Chiechi M, Abbitt P, et al: Neoplastic and inflammatory processes of the peritoneum, omentum, and mesentery: Diagnosis with CT. Radiographics 12:1051–1068, 1992. 46. Hanks PW, Brody JM: Blunt injury to mesentery and small bowel: CT evaluation. Radiol Clin North Am 41:1171–1182, 2003. 47. Heiss SG, Li KC: Magnetic resonance angiography of mesenteric arteries. A review. Invest Radiol 33:670–681, 1998.

1532

PART II  CT and MR Imaging of the Whole Body

48. Hill CA, Harle TS, Mansell PW: The prodrome, Kaposi sarcoma, and infections associated with acquired immunodeficiency syndrome: Radiologic findings in 39 patients. Radiology 149:393–399, 1983. 49. Hollinshead WH: Anatomy for surgeons: Vol. 2. The thorax, abdomen and pelvis, ed 2, New York, 1971, Harper & Row. 50. Horton KM, Fishman EK: Multi-detector row CT of mesenteric ischemia: Can it be done? Radiographics 21:1463–1473, 2001. 51. Horton KM, Fishman EK: Volume-rendered 3D CT of the mesenteric vasculature: Normal anatomy, anatomic variants, and pathologic conditions. Radiographics 22:161–172, 2002. 52. Horton KM, Lawler LP, Fishman EK: CT findings in sclerosing mesenteritis (panniculitis): Spectrum of disease. Radiographics 23:1561–1567, 2003. 53. Hughes TM, Elton C: The pathophysiology and management of bowel and mesenteric injuries due to blunt trauma. Injury 33:295–302, 2002. 54. Huppert BJ, Farrell MA, Kawashima A, et al: Diagnosis of cavitating mesenteric lymph node syndrome in celiac disease using MRI. AJR 183:1375–1377, 2004. 55. Reference deleted in proofs. 56. Jadvar H, Mindelzun RE, Olcott EW, et al: Still the great mimicker: Abdominal tuberculosis. AJR 168:1455–1460, 1997. 57. Jeffrey RB, Federle MP, Laing FC: Computed tomography of mesenteric involvement in fulminant pancreatitis. Radiology 147:185–188, 1983. 58. Jeong YJ, Kim S, Kwak SW, et al: Neoplastic and nonneoplastic conditions of serosal membrane origin: CT findings. Radiographics 28:801–817, 2008. 59. Johnson PT, Horton KM, Fishman EK: Nonvascular mesenteric disease: Utility of multidetector CT with 3D volume rendering. Radiographics 29:721–740, 2009. 60. Kakitsubata Y, Umemura Y, Kakitsubata S, et al: CT and MRI manifestations of intraabdominal panniculitis. Clin Imaging 17:186– 188, 1993. 61. Kamaya A, Federle MP, Desser TS: Imaging manifestations of abdominal fat necrosis and its mimics. Radiographics 31:2021–2034, 2011. 62. Killeen KL, Shanmuganathan K, Poletti PA, et al: Helical computed tomography of bowel and mesenteric injuries. J Trauma 51:26–36, 2001. 63. Kim AY, Ha HK: Evaluation of suspected mesenteric ischemia. Radiol Clin North Am 41:327–342, 2003. 64. Kim H, Lee JM, Kim SH, et al: Primary gastrointestinal stromal tumors in the omentum and mesentery: CT findings and pathologic correlations. AJR 182:1463–1467, 2004. 65. Kobayashi S, Takeda K, Tanaka N, et al: Mesenteric panniculitis: MR findings. J Comput Assist Tomogr 17:500–502, 1993. 66. Kronthal A, Kang Y, Fishman E, et al: MR imaging in sclerosing mesenteritis. AJR 156:517–519, 1991. 67. Lawler LP, McCarthy DM, Fishman EK, et al: Sclerosing mesenteritis: depiction by multidetector CT and three-dimensional volume rendering. AJR 178:97–99, 2002. 68. Lee IJ, Ha HK, Park CM, et al: Abdominopelvic actinomycosis involving the gastrointestinal tract: CT features. Radiology 220:76–80, 2001. 69. Lee SS, Ha HK, Yang SK, et al: CT of prominent pericolic or perienteric vasculature in patients with Crohn’s disease correlation with clinical disease activity and findings on barium studies. AJR 179:1029–1036, 2002. 70. Lee SL, Ku YM, Rha SE: Comprehensive reviews of the interfascial plane of the retroperitoneum: Normal anatomy and pathologic entities. Emerg Radiol 17:3–11, 2010. 71. Levy AD, Arnáiz J, Shaw JC, et al: From the archives of the AFIP: Primary peritoneal tumors: Imaging features with pathologic correlation. Radiographics 28:583–607, 2008. 72. Levy AD, Remotti HE, Thompson WM, et al: From the archives of the AFIP: Gastrointestinal stromal tumors: Radiologic features with pathologic correlation. Radiographics 23:283–304, 2003. 73. Levy AD, Rimola J, Mehrotra AK, et al: From the archives of the AFIP: Benign fibrous tumors and tumorlike lesions of the mesentery: Radiologic-pathologic correlation. Radiographics 26:245–264, 2006. 74. Reference deleted in proofs.

75. Li Y, Yang ZG, Guo YK, et al: Distribution and characteristics of hematogenous disseminated tuberculosis within the abdomen on contrast-enhanced CT. Abdom Imaging 32:484–488, 2007. 76. Lips DJ, Barker N, Clevers H, et al: The role of APC and beta-catenin in the aetiology of aggressive fibromatosis (desmoid tumors). Eur J Surg Oncol 35:3–10, 2009. 77. Livingston DH, Lavery RF, Passannante MR, et al: Admission or observation is not necessary after a negative abdominal computed tomographic scan in patients with suspected blunt abdominal trauma: Results of a prospective, multi-institutional trial. J Trauma 44:273–282, 1998. 78. Lockhart ME, Tessler FN, Canon CL, et al: Internal hernia after gastric bypass: Sensitivity and specificity of seven CT signs with surgical correlation and controls. AJR 188(3):745–750, 2007. 79. Low RN: MR imaging of the peritoneal spread of malignancy. Abdom Imaging 32:267–283, 2007. 80. Reference deleted in proofs. 81. Lucey BC, Stuhlfaut JW, Soto JA: Mesenteric lymph nodes: Detection and significance on MDCT. AJR 184:41–44, 2005. 82. Lucey BC, Stuhlfaut JW, Soto JA: Mesenteric lymph nodes seen   at imaging: Causes and significance. Radiographics 25:351–365, 2005. 83. Macari M, Hines J, Balthazar E, et al: Mesenteric adenitis: CT diagnosis of primary versus secondary causes, incidence, and clinical significance in pediatric and adult patients. AJR 178:853–858, 2002. 84. Maccioni F, Viscido A, Broglia L, et al: Evaluation of Crohn disease activity with magnetic resonance imaging. Abdom Imaging 25:219–228, 2000. 85. Malhotra AK, Fabian TC, Katsis SB, et al: Blunt bowel and mesenteric injuries: The role of screening computed tomography. J Trauma 48:991–1000, 2000. 86. McMahon PM, Halpern EF, Fernandez-del Castillo C, et al: Pancreatic cancer: Cost-effectiveness of imaging technologies for assessing resectability. Radiology 221:93–106, 2001. 87. Meyers MA: Roentgen significance of the phrenicocolic ligament. Radiology 95:539–545, 1970. 88. Meyers MA: Distribution of intra-abdominal malignant seeding: Dependency on dynamics of flow of ascitic fluid. AJR 119:198–206, 1973. 89. Meyers MA: Metastatic seeding along the small bowel mesentery. Roentgen features. AJR 123:67–73, 1975. 90. Meyers MA: Intraperitoneal spread of malignancies and its effect on the bowel. Clin Radiol 32:129–146, 1981. 91. Meyers MA: Dynamic radiology of the abdomen: Normal and pathologic anatomy, ed 4, New York, 2000, Springer-Verlag. 92. Meyers MA, Oliphant M, Berne AS, et al: The peritoneal ligaments and mesenteries: Pathways of intraabdominal spread of disease. Radiology 163:593–604, 1987. 93. Mindelzun RE, Jeffrey RB, Jr, Lane MJ, et al: The misty mesentery on CT: Differential diagnosis. AJR 167:61–65, 1996. 94. Mirilas P, Skandalakis JE: Surgical anatomy of the retroperitoneal spaces: Part II—The architecture of the retroperitoneal space. Am Surg 76:33–42, 2010. 95. Mirvis S, Shanmuganathan K: Imaging in trauma and critical care, ed 2, Philadelphia, 1992, Elsevier Science. 96. Moawad J, Gewertz BL: Chronic mesenteric ischemia: Clinical presentation and diagnosis. Surg Clin North Am 77:357–369, 1997. 97. Moore KL: The developing human: Clinically oriented embryology, ed 4, Philadelphia, 1988, WB Saunders. 98. Nghiem H, Jeffrey R, Jr, Mindelzun R: CT of blunt trauma to the bowel and mesentery. AJR 160:53–58, 1993. 99. Nolan BW, Gabram SG, Schwartz RJ, et al: Mesenteric injury from blunt abdominal trauma. Am Surg 61:501–506, 1995. 100. Okino Y, Kiyosue H, Mori H, et al: Root of the small-bowel mesentery: Correlative anatomy and CT features of pathologic conditions. Radiographics 21:1475–1490, 2001. 101. Oldenburg WA, Lau LL, Rodenberg TJ, et al: Acute mesenteric ischemia: A clinical review. Arch Intern Med 164:1054–1062, 2004.

CHAPTER 47  Mesentery 102. Oliphant M, Berne AS, Meyers MA: Spread of disease via the subperitoneal space: Small bowel mesentery. Abdom Imaging 18:109– 116, 1993. 103. Padidar AM, Jeffrey RB, Jr, Mindelzun RE, et al: Differentiating sigmoid diverticulitis from carcinoma on CT scans: Mesenteric inflammation suggests diverticulitis. AJR 163:81–83, 1994. 104. Pantongrag-Brown L, Buetow PC, Carr NJ, et al: Calcification and fibrosis in mesenteric carcinoid tumor: CT findings and pathologic correlation. AJR Am J Roentgenol 164:387–391, 1995. 105. Pantongrag-Brown L, Nelson AM, Brown AE, et al: Gastrointestinal manifestations of acquired immunodeficiency syndrome: Radiologicpathologic correlation. Radiographics 15:1155–1178, 1995. 106. Patnana M, Sevrukov AB, Elsayes KM, et al: Inflammatory pseudotumor: The great mimicker. AJR Am J Roentgenol 198:W217– W227, 2012. 107. Pereira JM, Sirlin CB, Pinto PS, et al: CT and MR imaging of extrahepatic fatty masses of the abdomen and pelvis: Techniques, diagnosis, differential diagnosis, and pitfalls. Radiographics 25:69–85, 2005. 108. Pickhardt PJ, Bhalla S: Primary neoplasms of peritoneal and subperitoneal origin: CT findings. Radiographics 25:983–995, 2005. 109. Pickhardt PJ, Bhalla S: Unusual nonneoplastic peritoneal and subperitoneal conditions: CT findings. Radiographics 25:719–730, 2005. 110. Puylaert JB: Right-sided segmental infarction of the omentum: Clinical, US, and CT findings. Radiology 185:169–172, 1992. 111. Rao PM, Rhea JT, Novelline RA: CT diagnosis of mesenteric adenitis. Radiology 202:145–149, 1997. 112. Rao PM, Wittenberg J, Lawrason JN: Primary epiploic appendagitis: Evolutionary changes in CT appearance. Radiology 204:713–717, 1997. 113. Restrepo CS, Martinez S, Lemos JA, et al: Imaging manifestations of Kaposi sarcoma. Radiographics 26:1169–1185, 2006. 114. Rizzo M, Federle M, Griffiths B: Bowel and mesenteric injury following blunt abdominal trauma: Evaluation with CT. Radiology 173:143–148, 1989. 115. Rubenstein WA, Auh YH, Zirinsky K, et al: Posterior peritoneal recesses: Assessment using CT. Radiology 156:461–468, 1985. 116. Ruess L, Frazier A, Sivit C: CT of the mesentery, omentum, and peritoneum in children. Radiographics 15:89–104, 1995. 117. Rutledge RMD, Thomason MMD, Oller DMD, et al: The spectrum of abdominal injuries associated with the use of seat belts. J Trauma 31:820–826, 1991. 118. Sabate J, Torrubia S, Maideu J, et al: Sclerosing mesenteritis: Imaging findings in 17 patients. AJR 172:625–629, 1999. 119. Sandrasegaran K, Rajesh A, Rushing DA, et al: Gastrointestinal stromal tumors: CT and MRI findings. Eur Radiol 15:1407–1414, 2005. 120. Seo BK, Ha HK, Kim AY, et al: Segmental misty mesentery: Analysis of CT features and primary causes. Radiology 226:86–94, 2003. 121. Sheth S, Horton KM, Garland MR, et al: Mesenteric neoplasms: CT appearances of primary and secondary tumors and differential diagnosis. Radiographics 23:457–473, 2003.

1533

122. Shih MP, Angle JF, Leung DA, et al: CTA and MRA in mesenteric ischemia: Part 2, normal findings and complications after surgical and endovascular treatment. AJR 188:462–471, 2007. 123. Shih MP, Hagspiel KD: CTA and MRA in mesenteric ischemia:   Part 1, role in diagnosis and differential diagnosis. AJR 188:452–461, 2007. 124. Shinagare AB, Ramaiya NH, Jagannathan JP, et al: A to Z of desmoid tumors. AJR Am J Roentgenol 197:W1008–W1014, 2011. 125. Silverman PM, Baker ME, Cooper C, et al: CT appearance of diffuse mesenteric edema. J Comput Assist Tomogr 10:67–70, 1986. 126. Singh AK, Gervais DA, Hahn PF, et al: Acute epiploic appendagitis and its mimics. Radiographics 25:1521–1534, 2005. 127. Sompayrac SW, Mindelzun RE, Silverman PM, et al: The greater omentum. AJR 168:683–687, 1997. 128. Stoupis C, Ros P, Abbitt P, et al: Bubbles in the belly: Imaging of cystic mesenteric or omental masses. Radiographics 14:729–737, 1994. 129. Strouse PJ, Close BJ, Marshall KW, et al: CT of bowel and mesenteric trauma in children. Radiographics 19:1237–1250, 1999. 130. Stuart S, Booth TC, Cash CJ, et al: Complications of continuous ambulatory peritoneal dialysis. Radiographics 29:441–460, 2009. 131. Takeyama N, Gokan T, Ohgiya Y, et al: CT of internal hernias. Radiographics 25:997–1015, 2005. 132. Tirkes T, Sandrasegaran K, Patel AA, et al: Peritoneal and retroperitoneal anatomy and its relevance for cross-sectional imaging. Radiographics 32:437–451, 2012. 133. Trommer G, Bewer A, Kosling S: Mesenteric lymphadenopathy in Yersinia enterocolitica infection. Radiologe 38:37–40, 1998 [German]. 134. Valls C: Fat-ring sign in sclerosing mesenteritis. AJR 174:259–260, 2000. 135. Vikram R, Balachandran A, Bhosale PR, et al: Pancreas: Peritoneal reflections, ligamentous connections, and pathways of disease spread. Radiographics 29:e34, 2009. 136. Weinstein JB, Heiken JP, Lee JKT, et al: High resolution CT of the porta hepatis and hepatoduodenal ligament. Radiographics 6:55–74, 1986. 137. Wildermuth S, Leschka S, Alkadhi H, et al: Multislice CT in the pre- and postinterventional evaluation of mesenteric perfusion. Eur Radiol 15:1203–1210, 2005. 138. Williams PL, Dyson M, Bannister LH: Gray’s anatomy, ed 38, Edinburgh, 1995, Churchill Livingstone. 139. Xeropotamos NS, Nousias VE, Ioannou HV, et al: Mesenteric injury after blunt abdominal trauma. Eur J Surg 167:106–109, 2001. 140. Yasuhara H: Acute mesenteric ischemia: The challenge of gastroenterology. Surg Today 35:185–195, 2005. 141. Yenarkarn P, Thoeni RF, Hanks D: Lymphoma of the mesentery. Radiology 242:628–631, 2007. 142. Yoo E, Kim MJ, Kim KW, et al: A case of mesenteric cystic lymphangioma: Fat saturation and chemical shift MR imaging. J Magn Reson Imaging 23:77–80, 2006. 143. Yoo E, Kim JH, Kim MJ, et al: Greater and lesser omenta: Normal anatomy and pathologic processes. Radiographics 27:707–720, 2007. 144. Zirinsky K, Auh YH, Rubenstein WA, et al: The portacaval space: CT with MR correlation. Radiology 156:453–460, 1985.

48  Spleen Stephanie Soriano and Dean A. Nakamoto

The spleen is well evaluated with current cross-sectional imaging modalities. The primary imaging technique continues to be computed tomography (CT), although magnetic resonance imaging (MRI) has become important for lesion characterization. Ultrasonography is also used frequently, although there may be limited visualization of the spleen owing to a subcostal location or obscuration by adjacent bowel gas. 18Fluorodeoxyglucose (18FDG) positron emission tomography (PET) is becoming an important modality for oncologic imaging, particularly when fused with CT.

NORMAL ANATOMY AND VARIANTS Normal Spleen The spleen normally lies within the left upper quadrant. The adult spleen is typically 12 × 7 × 4 cm and weighs 100 to 200 g.37 The maximum craniocaudal length is 12 to 15 cm. The spleen is a network of white and red pulp. The white pulp consists of lymphocytes, plasma cells, and macrophages. The red pulp contains splenic cords, splenic sinuses, terminal branches of the central arteries, and pulp veins. The visceral surface of the spleen is adjacent to the stomach, left kidney, splenic flexure of the colon, and tail of the pancreas. The normal spleen may have rib notching and clefts that should not be confused with lacerations in patients who have experienced trauma. The spleen is usually a completely peritonealized organ. The splenorenal ligament, which connects the splenic hilum to the left kidney, contains the tail of the pancreas and the splenic artery and vein. The gastrosplenic ligament is formed by the union of the peritoneum of the lesser and greater sacs and connects the splenic hilum to the stomach. The phrenicocolic ligament connects the lower pole of the spleen to the splenic flexure of the colon and the diaphragm. On unenhanced CT scans, the normal splenic parenchyma is homogeneous; it measures 40 to 60 Hounsfield units (HU), usually 5 to 10 HU less than the normal liver. On T1-weighted MRI the normal spleen has a signal intensity equal to or less than that of normal liver, whereas on T2-weighted images, the spleen has uniformly high signal intensity. After intravenous (IV) contrast enhancement, the spleen can have a heterogeneous appearance on early arterial-phase images on both CT and MRI. This appearance is believed to be due to the differential enhancement of red and white pulp (Fig. 48-1). Early arterial-phase images may demonstrate a variety of patterns. Typically there is an arciform appearance. Other patterns include nodular enhancement, central enhancement, peripheral enhancement, and nonenhancing pseudomasses. Masses may be simulated, particularly during MRI, immediately after IV administration of a bolus of gadolinium. Therefore delayed images are important. All these patterns become homogeneous on delayed images.

1534

Normal Variants Splenic clefts are common. They can usually be easily recognized because of their sharp, smooth borders. They are typically located superiorly and medially in the spleen but may occur anywhere. In patients who have had trauma, one should avoid misinterpreting a congenital cleft as a splenic laceration. Splenic clefts are not associated with perisplenic edema (Fig. 48-2A), seen with splenic laceration. A retrorenal spleen occurs when the posterior border of the spleen lies posterior to the posterior border of the left kidney. In one series this finding was described in 15% to 20% of patients.46 It is more common in older adults. Accessory spleen represents normal splenic tissue in ectopic sites, arising from failure of fusion of some of the multiple buds of splenic tissue in the dorsal mesogastrium during embryologic life.21 Accessory spleens occur in up to 20% of the population. They are typically located near the splenic hilum but can be found anywhere in the peritoneal cavity (see Fig. 48-2E), usually measuring less than 4 cm in diameter. After surgical removal of the primary spleen, accessory spleens have been observed to hypertrophy. Accessory spleens have also been described in the pancreas; up to 16% of accessory spleens are located in or near the tail of the pancreas.9 Such accessory spleens may mimic an endocrine tumor of the pancreas; contrast-enhanced MRI with reticuloendothelial system–specific contrast media (e.g., superparamagnetic iron oxides) can help distinguish between these entities.9 Wandering or ectopic spleen is a rare entity whereby the spleen migrates from its normal site in the left upper quadrant, usually because of nonunion of the peritoneum of the lesser and greater sacs, creating a longer splenic mesentery and highly mobile spleen.21 This entity may be asymptomatic, but it can also cause acute abdominal pain due to torsion, and subsequently infarct37,93 (see Fig. 48-2D).

IMAGING TECHNIQUES Computed Tomography Current CT imaging uses multidetector scanning during a single breath hold and IV administration of a contrast agent. Using a mechanical injector and an angiocatheter (at least 20 gauge), a 90-mL bolus of nonionic contrast agent (ioversol 74% [Optiray 350]) is administered at a rate of 2 to 3 mL/second. Scanning is performed during the portal venous phase at 60 seconds after the bolus is given. This approach typically provides uniform enhancement of the spleen and enhancement of the liver during the portal venous phase. Routine images are acquired at 2.5-mm collimation and 5-mm thickness, with no interslice gap. In any patient being evaluated for trauma, delayed scans taken 2.5 to 3 minutes after the bolus can often exclude lacerations of the spleen or other abdominal organs.

CHAPTER 48  Spleen

A

B

C

D

1535

FIG 48-1  Various appearances of normal splenic enhancement. A, Arciform pattern. B, Peripheral enhancement. C, Normal enhancement in the portal venous phase. D, CT axial image showing a pseudomass in the arterial phase (arrow).

Magnetic Resonance Imaging Because of continuing advances in MR technology, we are continually refining our MRI protocols. The current protocol for imaging the spleen (also used to image the liver) is as follows: 1. Coronal T2-weighted half-Fourier acquisition single-shot turbo-spin echo (HASTE): repetition time/echo time (TR/TE) 1000/54 milliseconds, 5-mm slice thickness, 0 gap, flip angle (FA) 120 degrees, 256 × 180 matrix (Mx), bandwidth (BW) 320 Hz/pixel, approximate resolution 1.6 × 1.1 × 5 mm 2. Axial T2-weighted HASTE: TR/TE 650/54 milliseconds, 5-mm slice thickness, 0 gap, FA 120 degrees, 256 × 180 Mx, BW 319 Hz/ pixel, approximate resolution 1.6 × 1.1 × 5 mm 3. Axial T2-weighted fat-saturated HASTE: TR/TE 748/81  milliseconds, 5-mm slice thickness, 0 gap, FA 120 degrees, 

256 × 180 Mx, BW 319 Hz/pixel, approximate resolution 1.6 × 1.1 × 5 mm 4. Axial diffusion-weighted imaging (DWI) (b value = 0, 400, 600 sec/mm2): TR/TE 3800/71 milliseconds, 5-mm slice thickness, 0 gap, BW 1736 Hz/pixel, generalized autocalibrating partially parallel acquisitions (GRAPPA) (parallel imaging) factor 2, approximate resolution 2.0 × 1.8 × 5 mm 5. Axial in- and opposed-phase volumetric interpolated breath hold (VIBE): TR/TE 7.96/2.38 and 4.77 milliseconds, 3-mm slice thickness, 0 gap, FA 10 degrees, 256 × 180 Mx, BW 310 and 240 Hz/pixel, GRAPPA (parallel imaging) factor 2, approximate resolution 1.9 × 1.3 × 3 mm 6. Axial three-dimensional VIBE precontrast +20/70/180/300 seconds delayed postcontrast: TR/TE 4.74/2.38 milliseconds, 2.5- to 4-mm reconstructed slice thickness, FA 10 degrees, 

1536

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E FIG 48-2  Normal splenic variants. A, CT scan showing a splenic cleft (arrow). Polysplenia in the axial (B) and coronal planes (C). D, CT scan showing a wandering spleen within the pelvis (arrow). E, T2 HASTE fatsaturated axial image showing two splenules (arrows).

CHAPTER 48  Spleen 256 × 180 × 64 Mx, BW 400 Hz/pixel, GRAPPA (parallel imaging) factor 2, approximate resolution 1.7 × 1.2 × 3 mm 7. Coronal VIBE +120 seconds delayed postcontrast: TR/TE 2.97/1.1 milliseconds, 2.5 mm slice thickness, FA 10 degrees,  256 × 180 Mx, BW 600 Hz/pixel, GRAPPA (parallel imaging) factor 2, approximate resolution 1.9 × 1.4 × 2.5 mm The goal of this protocol is to image the organ as quickly as possible; examination time is about 25 minutes. The dynamic postgadolinium images are usually the most important ones. It can be helpful to obtain additional VIBE delayed postgadolinium images at 8 minutes after contrast administration to aid evaluation of atypical hemangiomas.

Positron Emission Tomography–Computed Tomography PET-CT is a powerful modality, particularly for oncologic imaging. In addition it is potentially useful to characterize incidentally discovered splenic lesions. Because splenic lesions often have complex and nonspecific imaging features, it may be difficult to differentiate between benign and malignant lesions. Although splenic biopsies can be performed, such procedures pose a high risk for bleeding. 18FDG PET-CT is a noninvasive modality that appears to have a high negative predictive value for malignancy.69,70 Benign lesions in the spleen are expected to have a typical appearance on PET-CT. The benign solid lesions (hemangiomas, hamartomas) should not have abnormal uptake of 18FDG. Splenic infarct, which can sometimes simulate a splenic mass, should not have uptake. However, infectious or inflammatory lesions, including splenic abscess, active granulomatous disease (either infectious or inflammatory), and inflammatory pseudotumor, may have some uptake of 18FDG. Splenic metastases are expected to have abnormal uptake if the primary malignancy is 18FDG avid (see Fig. 48-4F). However, some renal cell and thyroid carcinomas that can metastasize to the spleen may not be 18FDG avid. Splenic lymphoma, which may be primary or secondary, is perhaps most accurately diagnosed and staged by PET-CT.70,91 This is particularly true for Hodgkin’s disease and aggressive non-Hodgkin’s lymphoma (see Fig. 48-3). For some of the low-grade lymphomas, PET may not be as accurate.51

PATHOLOGIC CONDITIONS Lymphoma and Splenic Neoplasms Lymphoma.  Splenic lymphoma is the most common splenic malignancy.23,87,105 Primary splenic lymphoma is rare, representing 1% to 2.6% of all lymphomas; most primary splenic lymphomas are nonHodgkin’s lymphomas of B-cell origin.57,2,18 The definition of primary splenic lymphoma is controversial.2,34 In general a history of normal bone marrow findings, clinical detection of splenic hilar adenopathy, or both, combined with imaging data and surgical findings, are used to exclude generalized lymphoma.2,18,27,64,97 Because primary lymphoma is rare, this discussion focuses on the features of generalized lymphoma. The cross-sectional imaging appearance of lymphoma of the spleen depends on its pathologic appearance, which may consist of (1) homogeneous enlargement without a discrete mass, (2) a solitary mass,  (3) multifocal lesions, or (4) diffuse infiltration. Splenomegaly is the most common imaging finding, but it may be absent in up to one third of patients with lymphoma37,112,119 (Fig. 48-3). Conversely, in 30% of the enlarged spleens found in patients with lymphoma, the enlargement was benign in origin.34,37,127 There are typical splenic imaging manifestations associated with the lymphoma subtypes. Diffuse splenic infiltration is most common in Hodgkin’s lymphoma; diffuse large B-cell lymphoma presents with splenomegaly or large hypodense

1537

splenic mass with abdominal lymphadenopathy; and the indolent lymphomas present with marked splenomegaly and lymphadenopathy.97 Despite recent advances in PET, CT (even with its relatively low accuracy) is still the most widely used cross-sectional imaging modality for diagnosing lymphoma.34,87,119 Typically lymphoma lesions do not enhance with contrast agents during CT or MRI. In addition  the normal spleen enhances heterogeneously during dynamic bolus imaging. As previously noted, delayed scanning may be helpful, but occasionally the lesions may enhance to the same level as the remaining spleen. In general, MRI is less reliable than CT because the normal spleen and lymphomatous tissue may have similar T1 and T2 relaxation times and proton densities.87,116 Use of superparamagnetic iron oxide has been proposed to improve detection of lymphoma lesions on MRI.127 If lymphoma is present in the liver, the spleen is almost always involved as well.105 Although rare, calcifications can be seen with splenic lymphoma both before and after treatment. They are most likely dystrophic calcifications secondary to necrosis, hemorrhage, and subsequent fibrosis.87 Although imaging plays an important role in detecting abnormalities in patients with lymphoma, the ability to stage lymphoma with imaging is limited. Between 45% and 70% of splenic lymphoma lesions manifest as diffuse infiltration or tumor foci smaller than 1 cm.98,127 Occasionally tumor foci can be seen only microscopically; on gross visual inspection the spleen may appear normal. Such small lesions may be difficult or impossible to visualize even with state-of-the-art CT and MRI, making the staging of lymphoma difficult. However, in patients with known disease, imaging can be used to upgrade staging— that is, to detect new retroperitoneal or mediastinal lymph nodes. In recent years, FDG PET has become an important tool for evaluating lymphoma patients (see Fig. 48-4F and G).51,91 PET appears to be more accurate for staging and monitoring lymphoma than CT.91 The other role of cross-sectional imaging in patients with lymphoma is to differentiate between lymphoma and other disease  processes. For example, patients with lymphoma may undergo chemotherapy and become immunocompromised, making them susceptible to disseminated fungal infections such as candidiasis. Fungal infections may be difficult to diagnose, and fungal microabscesses should not be mistaken for foci of splenic lymphoma. MRI with gadolinium enhancement is an effective modality for differentiating between fungal abscesses and lymphoma lesions.101 Splenic fungal lesions on MRI appear as multiple small rounded lesions demonstrating hypointensity on T1- and hyperintensity on T2-weighted images, with rim enhancement on postcontrast images.23,97 Splenic abscess may demonstrate clinical and imaging features similar to those of lymphoma: fever, left upper quadrant pain, malaise, and cystic lesions within the spleen. In general, splenic abscesses rarely occur with lymphadenopathy; this fact can be helpful in differentiating these two entities.34,87 Even with current imaging technology, however, differentiating lymphoma from abscess can be difficult. For example, necrotic lymphoma lesions may appear cystic and resemble splenic abscess. Although necrosis of lymphoma lesions is rare,87 larger lesions can manifest as cystic-appearing masses or may have cystic components.119 Gas within a cystic-appearing splenic lesion suggests infection, although gas can also be seen with necrosis only. Splenic infarcts may occur in conjunction with lymphoma. Typically infarcts are peripheral wedge-shaped lesions. Splenic lymphoma may be associated with human immunodeficiency virus (HIV). Single or multiple low-attenuation foci may be present. Focal splenic lesions are more common in patients with  lymphoma related to acquired immunodeficiency syndrome (AIDS) than in those with lymphoma unrelated to AIDS.117 The diagnosis of

1538

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E FIG 48-3  Various appearances of splenic lymphoma in five patients. A, Patient with B-cell lymphoma and splenomegaly without evidence of a focal lesion. CT axial (B) and coronal (C) images demonstrating multifocal hypoattenuating lesions with the spleen in a patient with Hodgkin’s lymphoma (arrows). D and E, CT scans of a patient with diffuse large B-cell lymphoma demonstrating a large dominant, ill-defined, hypoattenuating central mass (arrow).

CHAPTER 48  Spleen

F

G FIG 48-3, cont’d  F, PET-CT of the same patient showing the large hypermetabolic tumor in the spleen. G, PET-CT in a patient with follicular lymphoma, demonstrating multiple hypermetabolic masses within the spleen.

1539

1540

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E FIG 48-4  Metastasis. A and B, Heterogeneously low-attenuating dominant metastatic mass in a patient with non–small cell lung cancer (NSCLC). C and D, Low-attenuating metastatic mass in a different patient with NSCLC, demonstrating perisplenic hematoma in the noncontrast image (C) and the portal venous phase (D) (arrows). E, Multiple low-attenuating metastatic lesions within the liver and spleen in a patient with metastatic melanoma (arrows).

CHAPTER 48  Spleen

1541

F FIG 48-4, cont’d  F, PET-CT of a different patient with NSCLC showing a hypermetabolic lesion within the spleen.

non-Hodgkin’s lymphoma in a patient with HIV infection fulfills the criteria for the diagnosis of AIDS.117 In general, AIDS-related lymphomas are aggressive and often manifest with multiple sites of abdominal involvement.117

Angiosarcoma.  Splenic angiosarcoma is a rare aggressive neoplasm that may be primary or metastatic. It has a poor prognosis and may be complicated by splenic rupture.14,26,114 Angiosarcomas can manifest as multiple nodules of various sizes or show permeative growth into the spleen; some may cause intrasplenic or perisplenic hemorrhage or frank hemoperitoneum.33,87 The spleen is almost always enlarged, and metastatic lesions are often found at presentation, primarily to the liver.1,57 Like hepatic hemangiomas, angiosarcomas may show varying degrees of enhancement.87 On CT and MRI the lesions are heterogeneous in enhancement, with areas of necrotic degeneration or hemorrhage. These areas of heterogeneous enhancement on MRI correspond to the solid portions of the tumor.1,56,57

Metastases.  Although more common than primary neoplasms, splenic metastases are infrequent despite the rich vascularity of the spleen.60 Splenic metastases tend to occur in patients with widely disseminated tumor. Primary tumors that commonly metastasize to the spleen include melanoma and lung, breast, and ovarian carcinomas (Fig. 48-4). Direct invasion may also occur from gastric, pancreatic, or colon carcinoma.

Cysts.  Various types of focal lesions of the spleen may appear cystic on cross-sectional imaging. These include congenital cysts, nonpancreatic pseudocysts (posttraumatic hematoma), cystic neoplasms (lymphangioma, hemangioma, lymphoma, metastasis), and inflammatory lesions (abscess, parasitic cyst).19,80,120 These cystic lesions can be further classified into true and false cysts. True cysts are differentiated by the presence of an epithelial lining, whereas false cysts do not have an epithelial lining.19 True cysts can then be subgrouped into parasitic (e.g., echinococcal cysts) and nonparasitic lesions (e.g., congenital cysts).92 Congenital cysts are believed to be developmental in origin. Possible explanations for the development of the epithelial lining include infolding of peritoneal mesothelium and collections of peritoneal mesothelial cells trapped within the splenic sulci.19,22,120 False cysts are presumed to be posttraumatic in origin; they may be hemorrhagic or serous and may be the end stage of intrasplenic hematoma.19,80,120 Posttraumatic false cysts probably account for 75% to 80% of all splenic cysts.80,120 Splenic cysts can also be found in women of childbearing age, which is thought to be due to congestion and infarction of the spleen during pregnancy, causing hemorrhage and secondary cyst formation.22 True and false cysts can appear identical on cross-sectional imaging. Typically both are well defined, homogeneous, and unilocular, though septations may be present. It may be impossible to differentiate between true and false cysts on imaging studies, but there are a few differential points to consider. False cysts, which tend to be smaller, have internal

1542

PART II  CT and MR Imaging of the Whole Body

debris and peripheral calcifications120; peripheral calcifications are less common in true cysts.19 The diagnosis of false cyst should also be favored when (1) there is a clear history of trauma, (2) the patient is older than the fourth decade, or (3) there is hematoma elsewhere in the spleen.19 On CT, both true and false cysts have an attenuation value near or equal that of water, walls that are either thin or imperceptible, and lack of enhancement after IV administration of a contrast agent (Fig. 48-5). On MRI, both types of cysts appear to contain fluid with a signal intensity similar to or slightly greater than that of water. After administration of gadolinium, the cysts show no central or rim enhancement.52 It may be difficult to differentiate the various cystic lesions of the spleen. The clinical and imaging features may have to be considered together to narrow the differential diagnosis. Diagnoses to consider include echinococcal cysts, large solitary abscess, hematoma, and cystic neoplasms of the spleen.19

Hemangioma.  Splenic hemangioma is the most common benign splenic neoplasm. It is most likely congenital in origin, composed of endothelial-lined blood-filled vascular channels of varying sizes.23,30,125 Its incidence is 0.03% to 14% in autopsy series.2 Hemangiomas of the spleen may be single or multiple and may occur as part of a generalized angiomatosis (Klippel-Trénaunay-Weber syndrome).89 The appearance of splenic hemangiomas depend on the tissue subtype, which includes cavernous, capillary, and mixed. The cavernous subtype predominates. Infarction, thrombosis, or fibrosis may occur in splenic hemangiomas, as in hemangiomas elsewhere in the body.89 Solitary hemangiomas are usually asymptomatic and are often discovered incidentally in adults. Hemangiomas may grow slowly and as they enlarge produce symptoms such as pain and fullness in the left upper quadrant. Up to 25% of splenic hemangiomas manifest acutely because of rupture.3,89 Anemia, thrombocytopenia, and coagulopathy (Kasabach-Merritt syndrome) have been reported with large hemangiomas, likely secondary to the sequestration of red cells and platelets and the consumption of clotting factors.94,125 The imaging appearance of the splenic hemangioma depends  on the proportion of cavernous and capillary components. On CT,  solitary splenic hemangiomas are usually small and solid; however, larger hemangiomas may be more complex and have both solid and cystic components. On nonenhanced scans, solid splenic capillary hemangiomas are typically isodense to hypodense relative to normal spleen; after IV administration of a contrast agent, they enhance from the periphery to the center like hepatic hemangiomas94 (Fig. 48-6). However, cavernous hemangiomas of the spleen tend to have mixed solid and cystic components and, unlike their classic appearance in the liver, demonstrate discrete mottled areas of heterogeneous density on delayed images after administration of a contrast agent.30,33 Rarely calcifications may be noted, with solid regions presenting with central punctate calcifications; curvilinear calcifications may be seen in the periphery of cystic portions.1,3,94 Splenic hemangiomas have low signal on T1-weighted MRI and higher signal (>120 milliseconds) on T2-weighted images, particularly with prolonged TE. Complex hemangiomas may have heterogeneous signal on T1- and T2-weighted images owing to the amount of   fibrosis and hemorrhage present and the variation in size of the blood-filled spaces. After IV administration of gadolinium, splenic hemangiomas demonstrate peripheral enhancement with progressive central filling.33,89 Ramani and associates found that 19 of 22 hemangiomas had peripheral enhancement with centripetal fill-in.89 Only two lesions in their series demonstrated a clearly defined peripheral nodular type of enhancement, typically seen in hepatic hemangiomas; the remaining lesions demonstrated a peripheral rim of variable

thickness with irregular inner margins. On delayed images, splenic hemangiomas also demonstrate retention of contrast material on delayed images, like hepatic hemangiomas. However, as described for CT, not all splenic hemangiomas have this classic pattern of enhancement on MRI. Differentiation of splenic hemangioma from other splenic neoplasms can be difficult, particularly in larger lesions complicated by hemorrhage or thrombosis.30 A positive result on a nuclear medicine– tagged red blood cell scan can establish the diagnosis of hemangioma of the spleen by demonstrating increased activity on delayed images.25,57,125

Lymphangioma and Lymphangiomatosis.  Lymphangiomas are rare, slow-growing, benign tumors composed of lymphatic vessels. No consensus has been reached regarding the pathogenesis of lymphangioma; however, some theories have posited that it could result from splenic lymphatic tissue agenesis or congenital malformation with obstruction.3,108 Splenic lymphangiomas are considered very rare, mostly occurring in children.1,33 These lesions may be solitary but more often appear as multiple thin-walled, well-marginated cysts that may be subcapsular in location.86,122 Splenic lymphangioma is typically an asymptomatic incidental finding; however, it may also present as a large mass with symptoms associated with mass effect on adjacent tissue.111 Diffuse lymphangiomas are called lymphangiomatosis. The prognosis of isolated splenic lymphangioma or lymphangiomatosis is believed to be good. In some cases, however, the presence of splenic lymphangiomatosis may signal multisystem involvement, known as cystic angiomatosis, consisting of lymphangiomatous or hemangiomatous involvement of bones, soft tissues, and viscera. This multisystem form may be progressive and have a poor prognosis.122 On CT and MRI, splenic lymphangiomas appear as multiloculated cystic lesions that do not enhance on arterial- or delayed-phase imaging52 (Fig, 48-7). Typically the lesions are in a subcapsular location where the lymphatics are normally concentrated.1 The lesions usually have water density on CT and follow water signal on MRI: low T1 signal intensity and high T2 signal intensity. However, if there is hemorrhage or proteinaceous debris, they can have high T1 signal intensity. Occasionally, peripheral calcifications can be present. Angiography demonstrates multiple well-defined avascular lesions of various sizes that give the spleen a “Swiss cheese” appearance.

Hamartoma.  Rare lesions, splenic hamartomas are usually solitary masses. They are classified based on the predominate histologic tissue—white or red pulp—though most are a mixture of the two cell types.89 They can be cystic or solid and may have punctate or stellate calcifications. Rarely these tumors are symptomatic, causing splenomegaly, pancytopenia, fatigue, and anorexia, likely due to sequestration of hematopoietic cells.17,115 Multiple splenic hamartomas may be associated with systemic hamartomas, such as with tuberous sclerosis or Wiskott-Aldrich syndrome. The imaging appearance on noncontrast CT images is a welldefined hypo- to isodense lesion that demonstrates uniform delayed enhancement that is hypo- to isodense to the spleen. The only finding may be a contour abnormality. On MRI the solid lesions may be isointense on T1- and hyperintense on T2-weighted images relative to the spleen. Unlike hemangiomas, however, hamartomas do not demonstrate an increase in signal intensity similar to that of cerebrospinal fluid when images with higher T2 weighting (TE = 120-150 milliseconds) are obtained.115 After IV administration of a contrast agent, hamartomas demonstrate diffuse heterogeneous enhancement on early postcontrast images, with more uniform enhancement on delayed postcontrast images.

CHAPTER 48  Spleen

A

B

C

D

E

F FIG 48-5  Cysts. A, CT noncontrast examination demonstrates a low-attenuation cyst with minimal peripheral calcification (arrow). B, Contrast-enhanced CT demonstrates multiple low-attenuating cysts within the dome of the liver (arrows). MRIs in a different patient show a lobulated cyst with increased signal on axial (C) and coronal (D) T2 HASTE images (arrows). E, T2 fat-saturated axial image in the same patient shows increased signal in the same lobulated cyst. F, A fat-saturated gradient echo T1-weighted image shows water signal within the cyst (arrow).

1543

1544

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E

F

G

H FIG 48-6  Hemangiomas. A, Axial CT image showing a low-attenuating hemangioma (arrow). B, Axial T2-weighted image showing increased signal within the same lesion (arrow). T1-weighted postcontrast axial images showing minimal enhancement at 70 seconds (C) with complete enhancement and retention of contrast at 8 minutes (D) (arrows). In a different patient, CT examination (E) demonstrates multiple small low-attenuating hemangiomas that are bright on T2 HASTE axial images (F). G and H, MR postgadolinium images at 20 seconds and 3 minutes demonstrate progressive peripheral enhancement.

CHAPTER 48  Spleen Hamartomas may be difficult to distinguish from malignant lesions. Splenic hamartomas may demonstrate radiotracer uptake on sulfur colloid scintigraphy, which may be helpful in differentiating these lesions from malignant neoplasms and hemangiomas.115

Littoral Cell Angioma.  Littoral cell angiomas were first described by Falk and colleagues in 1991 as splenic vascular tumors originating from the littoral cells of the splenic red pulp, typically appearing as multiple diffuse nodular masses within the spleen.28 These lesions are usually

benign.56 Patients present with splenomegaly and symptoms of hypersplenism. On noncontrast CT the nodular masses can appear iso- or hypodense. With contrast enhancement the masses can appear hypoattenuating in the early portal venous phase and homogeneously isodense in the late portal venous phase30,38,57 (Fig. 48-8). On MRI, littoral cell angiomas demonstrate low signal intensity on T1- and T2-weighted images because of cytoplasmic hemosiderin accumulation from the phagocytic action of the littoral cell.56

Infections and Diffuse Diseases Abscess

FIG 48-7  Axial CT image of a subcapsular low-attenuating lesion found to be a lymphangioma (arrow).

A

1545

Bacterial.  Splenic abscesses are uncommon, having been reported in 0.14% to 0.70% of large autopsy series.7,15,16 The observed rise in their incidence is believed to be due to an increase in the number of immunocompromised patients, such as oncology patients undergoing aggressive chemotherapeutic regimens, critically ill patients in intensive care units, and HIV-infected patients.24 Development of a splenic abscess may require both bacteremia and intrinsic splenic disease that alters the splenic architecture, such as infarct, hematoma, or sickle cell disease. Alternatively, splenic abscess can result from direct extension of infection from adjacent organs such as the pancreas61 or stomach.54 Aerobic organisms account for 57% of splenic abscesses, with the majority being caused by Staphylococcus species, Escherichia coli, and Salmonella species; 26% of abscesses are caused by fungi, predominantly Candida species, and 18% by anaerobic organisms.77 Mortality rates range from 15% in otherwise healthy patients with unilocular splenic abscess to 80% in immunocompromised patients with multiple abscesses.4 On cross-sectional imaging, splenic abscesses may be solitary or multiple (Fig. 48-9). CT is the study of choice. Solitary abscesses and multiloculated abscesses tend to be bacterial. Septations may be present, and there may be an enhancing capsule. The CT appearance of an abscess may overlap with that of an infarct, hematoma, or neoplasm. The lack of mass effect helps differentiate infarct from abscess or tumor. The presence of gas within the lesion is diagnostic of abscess; however, this finding is rare. Gas is more readily apparent on CT than

B FIG 48-8  Littoral cell angioma. Arterial phase (A) axial CT image shows a central hypodense lesion with peripheral enhancement, which becomes isodense on the late portal venous phase (B) (arrows).

1546

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

E FIG 48-9  Abscess. A and B, Axial and coronal CT images show multiple ill-defined hypodense abscesses with some locules of air. C, Axial CT of a dominant ill-defined hypodense abscess. D and E, Multiple small low-attenuating microabscesses within the spleen in patient found to be infected with Bartonella henselae.

CHAPTER 48  Spleen on MRI (see Fig. 48-9). Percutaneous splenic abscess drainage has been described as a safe and effective procedure.67,130 Zerem and Bergsland recommend that abscesses drained percutaneously be larger than 5 cm, be unilocular or at most bilocular, have a discrete wall, and preferably be in the periphery of the midpole or lower pole of the spleen; surgical treatment should be considered for multilocular abscesses or spleens with more than two collections.130 If the abscess is less than 5 cm, only aspiration is performed. If this is not successful, drainage is attempted. In their series, Zerem and Bergsland treated 36 patients with splenic abscess with percutaneous aspiration, percutaneous drainage, or both; they had a success rate of 89% (32 of 36).130 Four patients died of predisposing factors, not the abscess itself. Despite these encouraging results, percutaneous drainage of splenic abscess should be reserved for selected patients and performed with appropriate surgical backup. When performing splenic abscess drainage, we try to traverse as little normal splenic parenchyma as possible to avoid bleeding. Fungal.  Multiple small splenic abscesses (microabscesses) are typically nonbacterial in origin and are most commonly associated with fungal infections. The liver is usually involved as well. Fungal infections occur primarily in neutropenic patients, most commonly in those with acute leukemia or other lymphoproliferative disorders who are undergoing high-dose chemotherapy with or without bone marrow transplantation.95,101 An increase in the incidence of fungal abscesses of the spleen has resulted from widespread use of aggressive chemotherapy.77 Before the development of helical CT and current MRI techniques, it was difficult to establish the diagnosis of hepatosplenic candidiasis. The lesions were often undetectable because of their small size, the often nonspecific clinical signs and symptoms, and the commonly negative results of blood culture.65,95 Percutaneous needle biopsy may also yield a false-negative result.81 Typically the lesions of hepatosplenic candidiasis are smaller than 2 cm, usually measuring 5 to 10 mm in diameter; however, they may also be less than 5 mm (miliary). On CT they often have low attenuation, although there can be a focus of high attenuation or a “wheel-within-a-wheel” pattern. MRI is the modality of choice to evaluate for hepatosplenic candidiasis. Semelka and colleagues described the appearance of hepatosplenic candidiasis on MRI and distinguished the acute, subacute, and chronic phases of infection.101 They described acute microabscesses as small (<1 cm) round lesions best demonstrated on T2-weighted images as hyperintense foci. On T1-weighted images they are often poorly seen as low-signal foci and are only moderately demonstrated with gadolinium enhancement. Although MRI is accurate in the diagnosis of acute hepatosplenic fungal disease, the findings may be falsely negative if the patient is neutropenic and unable to mount an immune response to the infection. After treatment, acute microabscesses become necrotizing granulomas with a central abscess and peripheral fibrosis surrounded by a zone of lymphocytes and macrophages. In the setting of multiple blood transfusions, the macrophages contain iron,55 which produces a lowsignal ring around the lesions on all MRI sequences. These lesions  are still small (typically <1 cm). Subacute treated lesions have intermediate to mildly high signal intensity on T1- and T2-weighted MRI sequences and have a perilesional ring that is nearly void of signal intensity. If therapy is successful, subacute granulomas progress to chronic healed lesions composed of fibrous tissue of differing densities and degrees of vascularization. Chronic healed fungal abscesses have low signal intensity on T1-weighted images and are isointense to mildly hyperintense on T2-weighted images. They have an irregular polygonal shape and measure 1 to 3 cm in diameter. No perilesional dark ring is noted. These lesions are best noted on early postgadolinium images as

1547

moderately hypointense foci. They are minimally hypointense on delayed postgadolinium images.

Other Infectious Processes.  Other organisms, including Mycobacterium tuberculosis, Mycobacterium avium-intracellulare, and Pneumocystis carinii, may also infect the liver and spleen. Such infections tend to be miliary. M. tuberculosis infection can be seen in patients with immunodeficiency states associated with alcoholism, diabetes, steroid therapy, cancer, and HIV infection. Acutely, only splenomegaly may be present, without focal lesions. Subsequently, microabscesses may form and appear on CT as low-attenuation lesions (Fig. 48-10). They may then progress to scattered punctate calcifications, innumerable calcifications throughout the spleen, and finally almost complete splenic calcification.37 In patients with HIV who have disseminated M. tuberculosis, the spleen is more commonly involved than the liver.76 CT is preferred for evaluating disseminated processes such as mycobacterial infections, because other intraabdominal manifestations (e.g., lymphadenopathy, bowel wall thickening) are better appreciated on CT than on MRI. Splenic microabscesses are generally intermediate in signal intensity on T1-weighted MRI and measure from a few millimeters to 1 to 2 cm in diameter. They may show ring enhancement after administration of gadolinium. On T2-weighted images the microabscesses are high in signal intensity. As the lesions calcify, they become low in signal intensity on both T1- and T2-weighted images.116 M. avium-intracellulare and P. carinii infections occur more commonly in patients with HIV infection. M. avium-intracellulare infection occurs as a late manifestation of AIDS. Differentiating between M. tuberculosis and M. avium-intracellulare can be difficult in patients with HIV. Focal splenic or hepatic lesions are noted in about 30% of patients with HIV who are infected with M. tuberculosis but are rarely noted in those infected with M. avium- intracellulare. Marked splenomegaly is more common in M. avium-intracellulare than in M. tuberculosis infection.76,87,88 P. carinii infections occur very commonly in patients with AIDS, typically as pneumonia. However, with improvements in the therapies for P. carinii pneumonia, extrapulmonary P. carinii has become more common. When the spleen is involved, it may be enlarged with focal lesions that become progressively calcified either in a rimlike or punctate fashion over a few months.87,88 However, not all lesions demonstrate calcification, even at follow-up.

Peliosis and Bacillary Angiomatosis.  Peliosis is a rare disorder characterized by multiple blood-filled spaces that may involve the liver and spleen. It can lead to portal hypertension and organ rupture.68,104 Peliosis is primarily associated with HIV and AIDS. In patients with AIDS, visceral peliosis is due to infection by Bartonella henselae, a treatable manifestation of AIDS. Infection by B. henselae affects multiple organ systems, including the skin, airway, mucous membranes, lymphatics, visceral organs, bone, and brain.75 Kaposi’s sarcoma may have a similar presentation in patients with AIDS and if clinically suspected, the treatable infection, bacillary angiomatosis, should be considered in the differential diagnosis.75 Peliosis is also associated with other wasting diseases (e.g., tuberculosis, cancer) and with some drugs, including anabolic steroids and corticosteroids. In peliosis, multiple blood-filled cavities ranging in size from a few millimeters to 1 to 2 cm are noted in the liver and spleen. Such cavities may or may not have an endothelial lining and may communicate with dilated sinusoids.88 Some of the cavities may be thrombosed. The imaging findings depend on the pathologic appearance. On CT the liver lesions appear as multiple small low-attenuation foci that become isodense with IV administration of a contrast agent; the splenic lesions can be variable, ranging from nonenhancing hypodense nodules up

1548

PART II  CT and MR Imaging of the Whole Body

A

B FIG 48-10  A and B, Axial and coronal CT images showing small low-attenuating lesions in a patient with AIDS and tuberculosis.

to 1.5 cm in diameter to nodules with a target appearance and central enhancement. The thrombosed cavities remain low in attenuation with contrast enhancement.47,68,88 Sometimes, small peliotic lesions may demonstrate hyperattenuation on the arterial and portal venous phases of enhancement.47 On MRI, multiple foci may be seen, with increased signal intensity on T2-weighted images and variable signal intensity on T1-weighted images. The thrombosed cavities may show mixed signal on T2-weighted images because of the presence of deoxyhemoglobin and methemoglobin.68 The diagnosis can be confirmed histologically.74

Sarcoidosis.  Sarcoidosis is a systemic disease of unknown cause, characterized by multiple noncaseating granulomas in the affected organ. It involves primarily the thorax (mediastinal and hilar lymph nodes and lung parenchyma), skin, and eyes and less commonly the liver and spleen. Mesenteric and retroperitoneal lymph nodes can also be involved. The appearance may be similar to that of lymphoma, particularly non-Hodgkin’s lymphoma, and to metastases; however, involvement of retrocrural nodes by sarcoidosis is rare. This feature may help distinguish sarcoidosis from non-Hodgkin’s lymphoma, in which retrocrural node involvement is common. The most common splenic finding is nonspecific hepatosplenomegaly. Less commonly, sarcoidosis may also manifest as nodular splenic lesions, seen on CT in approximately 6% to 33%.13,35,100,123,124 The nodules demonstrate a diffuse distribution and range in size from 1 mm to 2 cm, with the larger lesions coalescing together.123,124,126 On CT the nodules of sarcoidosis are low in attenuation without contrast enhancement44,100 (Fig. 48-11). On MRI, nodular sarcoidosis has been described as having multiple low-signal-intensity foci on T1- and T2-weighted images and no substantial enhancement.116,126 In general the degree of splenic involvement confers a greater severity of disease. Warshauer and colleagues determined that a splenic diameter of greater than 14 cm was significantly associated with abdominal adenopathy, in comparison to patients with a splenic diameter less than 14 cm. There was also a positive correlation with splenic

FIG 48-11  Sarcoidosis. CT demonstrates diffuse involvement of the spleen by sarcoid. The lesions are nonenhancing and splenomegaly is present.

size and systemic symptoms. In addition, the presence of splenic nodules was also associated with a symptomatic presentation.123,124

Hydatid Disease.  Splenic involvement is uncommon in hydatid disease, occurring in less than 2% to 8% of all patients with echinococcal infection.36,82 The liver, lung, peritoneum, kidney, brain, mediastinum, and heart are more commonly affected than the spleen.48,84 Systemic dissemination and intraperitoneal spread from a ruptured

CHAPTER 48  Spleen liver cyst constitute the two main sources of splenic involvement.36 Hydatid cysts are seldom seen except in areas where Echinococcus species are endemic, such as Argentina, Greece, and Spain.120 The appearance of splenic hydatid cysts varies according to the age of the cyst and the associated complications, such as rupture and secondary infection. On CT, an uncomplicated cyst may have homogeneous fluid content with water attenuation; however, it may also contain debris, hydatid sand, or inflammatory cells that cause high attenuation.36 Peripheral calcifications may also be present. Classically, small daughter cysts, formed by invagination of the inner layer of the cyst, are noted at the periphery of the main cyst.120 Hydatid cysts are prone to rupture. Although rupture can be clinically silent, there are many potential complications, including anaphylaxis, infection, peritoneal seeding, perforation of hollow viscera, and hematogenous spread.82 Infected cysts may appear solid, have a mixed solid and cystic pattern, or contain air or air-fluid levels. CT and ultrasonography are the most useful modalities for diagnosis, with CT being preferred in cases of suspected infection.82

Splenomegaly Splenomegaly is a nonspecific finding with numerous etiologies. Although CT or MRI can accurately determine splenic volume, the cause of splenomegaly may not be apparent from imaging alone. Splenomegaly can be due to congestion (e.g., portal hypertension, splenic vein thrombus), infiltrative storage disease (e.g., Gaucher’s disease, sarcoidosis, amyloidosis), infection (e.g., infectious mononucleosis, tuberculosis), myeloproliferation (e.g., myelofibrosis), increased hematologic destruction (e.g., polycythemia vera, hemoglobinopathies), or neoplasms (e.g., leukemia, lymphoma)112 (Fig. 48-12).

Portal Hypertension.  Portal hypertension is a frequent congestive cause of splenomegaly. In addition to splenomegaly, there is enlargement of the splenic vein, formation of varices, and formation of Gamna-Gandy bodies, which are small remnants of intrasplenic hemorrhage composed of hemosiderin, fibrous tissue, and calcium.116 On MRI the Gamna-Gandy bodies have low signal intensity on T1- and

A

1549

T2-weighted images, appearing as multiple foci of signal void (Fig. 48-13). They are best visualized as low-signal-intensity foci on T1-weighted gradient echo images using in-phase and opposed-phase techniques. The lesions become more conspicuous on images with longer TE, because susceptibility artifacts from hemosiderin are increased with longer TE.

Gaucher’s Disease.  Gaucher’s disease is a metabolic disorder in which glucocerebroside accumulates abnormally in the macrophages of the reticuloendothelial system because of a deficiency of the enzyme glucocerebrosidase. These enlarged macrophages are called Gaucher cells. Affected patients have hepatosplenomegaly, thrombocytopenia, and anemia. In one study, focal lesions were described on MRI in 30% of 46 patients with Gaucher’s disease.43 Both red and white types of nodules were noted. The white foci of solid Gaucher cells in the spleen were isointense on T1-weighted images and of low signal intensity on T2-weighted images, but the red focal lesions contained sinusoids  and behaved more like hemangiomas, showing increased signal on T2-weighted images.23,43 Infarcts were also noted.

Hemochromatosis Primary hemochromatosis, an autosomal recessive disorder, causes organ damage such as cirrhosis and possibly hepatocellular carcinoma. Hemosiderosis and subsequent secondary hemochromatosis is caused by excess deposition of iron into the reticuloendothelial system (spleen, bone marrow, liver Kupffer cells), usually from blood transfusions or rhabdomyolysis. On MRI a decrease in signal intensity is seen on both T1- and T2-weighted images because the tissue iron is superparamagnetic (Fig. 48-14).

TRAUMA The spleen is the visceral organ most commonly injured by blunt abdominal trauma. The spleen is particularly susceptible given its thin capsule, mobile vascular pedicle, and exposed position on inspiration. The intraperitoneal spleen also has the potential for uncontained

B FIG 48-12  Splenomegaly. A, Splenomegaly in a patient with a myeloproliferative disorder. B, Splenomegaly in a patient with glucose-6-phosphate dehydrogenase deficiency.

1550

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 48-13  Patient with cirrhosis, splenomegaly, and Gamna-Gandy bodies in the spleen. A, T2-weighted MRI shows several low-signal-intensity Gamna-Gandy bodies. On T1-weighted in-phase (B) and opposedphase (C) images, the low-signal-intensity Gamna-Gandy bodies are more conspicuous, particularly on the in-phase image (B) compared with the opposed-phase image (C). The lesions’ loss of signal intensity is due to the longer TE of the in-phase image, which increases the susceptibility artifact. D, T1-weighted 3-minute postgadolinium image shows the Gamna-Gandy bodies as low signal intensity.

hemorrhage after trauma.10 A missed splenic injury is one of the most common causes of preventable death in patients with trauma.128 Splenic injuries can also be iatrogenic and may occur during surgery, percutaneous needle procedures, or colonoscopy.113 The spleen is a very vascular organ and may contain 500 to 600 mL of blood; thus splenic injury is potentially fatal. Over the past few decades, management of splenic injury has become more conservative, with the goal of salvaging the spleen whenever possible. In the emergency department, the initial physical evaluation of a trauma patient with blunt splenic injury may be limited. These patients may have multiple injuries or altered levels of consciousness due to central nervous system injury or intoxication. Triage of such patients may depend on local expertise and availability of resources. Patients who are hemodynamically unstable are typically evaluated for the presence of intraabdominal blood with a focused assessment with sonography for trauma (FAST) or diagnostic peritoneal lavage. A positive FAST in an unstable patient generally leads to laparotomy. If the patient is stable, CT is the imaging procedure of choice.

Focused Assessment with Sonography for Trauma Many trauma centers use abdominal ultrasonography (FAST) as the initial modality to evaluate hemodynamically unstable patients. It was first introduced in the early 1990s, and by 1999, 79% of level I trauma centers used FAST.11,118 The results of the FAST examination have been incorporated into decision algorithms for patients with blunt abdominal trauma.10,45 Initial reports of FAST demonstrated high sensitivity and accuracy; however, a recent systematic review and a retrospective chart review comparing the findings of the FAST examination to the CT evaluation have demonstrated that although a positive FAST examination proves the presence of abdominal injury, a negative examination fails to exclude intraabdominal traumatic injury.8,73,96,109 Therefore patients with certain injuries, such as pelvic ring fracture, spinal fracture, or hematuria from high-velocity impact, may benefit from CT even if the FAST is negative.6,96 In an unstable patient with a positive FAST, emergent laparotomy is generally performed. However, Farahmand and coworkers found that although 94% of hypotensive patients

CHAPTER 48  Spleen

A

B

C

D

1551

FIG 48-14  Hemosiderosis. A, Axial CT shows diffuse increased attenuation of the liver and spleen due to iron-deposition hemochromatosis from multiple blood transfusions. B, T2-weighted image demonstrates diffusely decreased signal intensity in the spleen and liver. T1-weighted in-phase (C) and opposed-phase (D) images demonstrate increased loss of signal intensity on the in-phase image owing to the longer TE, which increases susceptibility artifact from the excess iron in the spleen.

with blunt trauma and positive ultrasound findings had injuries, only 64% had injuries that required surgical intervention.29 Thus they recommended that even with positive ultrasound findings, hemodynamically unstable patients may benefit from diagnostic CT to identify injuries that might not necessarily require surgery if they can be stabilized well enough to undergo CT. The ability to stabilize a hypotensive trauma patient and thus allow CT depends on local institutional factors. The FAST examination is performed by evaluating four areas for the presence of blood: pericardial cavity, right upper quadrant, left upper quadrant, and pelvis. Ultrasonography is a very sensitive modality to detect hemoperitoneum. The advantages of FAST are that it can be performed rapidly in the emergency department and is less expensive than CT.12 FAST is also noninvasive, unlike diagnostic peritoneal lavage, and it is easily repeated. Disadvantages of FAST are the same as for any ultrasound examination; it may be of limited use in obese patients or in those with large amounts of bowel gas or subcutaneous emphysema. FAST is also operator dependent. In addition, some important injuries may not be detected by FAST via the presence of

hemoperitoneum, including intraperitoneal injuries, such as bowel and mesentery, and retroperitoneal organ injuries. Shanmuganathan and associates reported that 34% of trauma patients with abdominal visceral injuries did not have hemoperitoneum on CT or at surgery.102

Computed Tomography CT is the modality of choice to image hemodynamically stable trauma patients with blunt abdominal injuries. Unstable patients who become stabilized after initial resuscitation in the emergency department may also benefit from CT, particularly if a FAST examination is negative.63 The overall diagnostic accuracy of CT is approximately 95%.32,53,66,106 In the last decade there has been a move toward using only IV contrast for evaluation of splenic trauma, because there are safety risks with the use of oral contrast (e.g., aspiration) and the potential for delay in diagnosis. In addition, Stuhlfaut et al. have determined that for traumatic injuries requiring surgical repair, CT detection rates with use of IV contrast only are comparable to CT detection rates with use of oral and IV contrast.110 Meta-analyses performed by Lee and

1552

PART II  CT and MR Imaging of the Whole Body

A

B FIG 48-15  Trauma. A, Contrast-enhanced axial image demonstrates a linear hypodensity with a peripheral wedge-shaped hypodensity representing a splenic laceration with a peripheral infarct (arrow). B, Noncontrast axial CT image showing hyperdense blood in the area of the spleen with surrounding hemoperitoneum in a patient with splenic rupture.

colleagues reiterated that omission of oral contrast did not affect the accuracy of CT in the diagnosis of blunt abdominal injury.62 Inside the CT suite the patient receives the IV contrast agent, 90 mL of Optiray 350, at a rate of 2 to 3 mL/second via an automated mechanical injector. Images are obtained during the portal venous phase of imaging at approximately 80 seconds after initiation of the IV bolus of contrast agent. Images obtained too early show heterogeneous enhancement of the spleen because of differential enhancement of the red and white pulp. Splenic lacerations may not be visualized during this phase of enhancement. Delayed scans should then be performed approximately 5 minutes after initiation of the IV bolus. Uncooperative patients should be sedated to minimize motion artifacts. The entire abdomen and pelvis should be evaluated. Beam-hardening artifacts from the fracture board and from life-support monitor wires should not be misinterpreted as organ lacerations. These artifacts are usually recognized by using multiple window and level settings during image evaluation. Because trauma patients may have multisystem injuries, the images must be carefully evaluated for visceral organ injuries and hemoperitoneum. Attention should also be paid to subtle signs of bowel or mesentery injury, such as free fluid, mesenteric edema, and intramural hematoma. Splenic injuries include hematoma, laceration, active hemorrhage, posttraumatic infarct, and posttraumatic vascular injuries such as pseudoaneurysm and arteriovenous fistula. Splenic hematomas are typically subcapsular but can be intraparenchymal. On nonenhanced CT, a subcapsular hematoma is usually hyperdense compared with the spleen, but it may be isodense. After IV contrast administration, subcapsular hematomas appear as low-attenuation collections between the splenic capsule and the enhancing splenic parenchyma. They are usually crescent shaped. Most subcapsular hematomas have some mass effect on the adjacent splenic parenchyma, whereas perisplenic hematomas typically do not deform the spleen. Subcapsular hematomas generally resolve in 4 to 6 weeks. Intraparenchymal hematomas are typically round and can have a varied appearance; they may be heterogeneous and contain high-attenuation clot. Splenic lacerations are linear low-attenuation parenchymal defects. Lacerations are initially well defined and become less well defined over time. Splenic clefts may be mistaken for lacerations. Splenic clefts have smooth or even rounded margins and do not change over time. Most lacerations

have associated perisplenic edema, fluid, or clot; therefore the lack of such findings suggests a splenic cleft. A fracture is a laceration that traverses two opposed capsular surfaces. Multiple lacerations and fractures define a shattered spleen. Lack of enhancement of part or all of the spleen suggests injury or thrombosis of the splenic artery or segmental artery (Fig. 48-15A). In addition to the size of the splenic injury, the presence of active hemorrhage should be assessed. Active hemorrhage is defined as CT evidence of a focal intrasplenic or perisplenic collection that has an attenuation value similar to that of the aorta or a major adjacent artery and greater than the attenuation value of the spleen (see Fig. 48-15B).31,39,102,103 Delayed scans help differentiate active hemorrhage from contained vascular injuries. Anderson and colleagues found  that active hemorrhage increases in size or demonstrates attenuation at least 10 HU higher than the aorta on 5-minute delayed-phase images, whereas contained splenic vascular injuries (pseudoaneurysm, arteriovenous fistula) remain the same or decrease in size or have attenuation values within 10 HU of the aorta on delayed-phase images.5 Posttraumatic infarct is rarely seen in patients with blunt splenic injury. Infarcts appear as wedge-shaped areas of decreased attenuation on the portal venous and delayed phases of postcontrast scans.72 The mechanism of posttraumatic infarct is not known. Complications are rare and include new infarcts, abscess formation, and splenic rupture.72 Delayed rupture of the spleen is rare and can occur up to 10 days after injury. This complication is believed to be associated with subtle low-grade injury to the spleen that is not detectable on CT, ultrasonography, or diagnostic peritoneal lavage.58 The exact incidence of delayed rupture of the spleen is not known. Suboptimal scanning techniques may be responsible for the missed diagnosis of subtle splenic injuries in the emergency department.129

Splenic Injury Grading Scales The most commonly used splenic injury grading scale is that developed by the American Association for the Surgery of Trauma (AAST).74 This scale is based on the extent of splenic injury at the time of laparotomy. The grades are as follows: Grade I: subcapsular hematoma involving less than 10% of the splenic surface area or laceration less than 1 cm deep

CHAPTER 48  Spleen Grade II: subcapsular hematoma involving 10% to 50% of the splenic surface area, intraparenchymal hematoma less than 5 cm in diameter, or laceration 1 to 3 cm deep that does not involve a trabecular vessel Grade III: subcapsular hematoma involving more than 50% of   the splenic surface area, ruptured subcapsular or parenchymal hematoma, intraparenchymal hematoma greater than 5 cm in diameter, or laceration more than 3 cm deep or involving the trabecular vessels Grade IV: hilar laceration and devascularization of more than 25% of the splenic mass Grade V: shattered or early completely devascularized spleen Although such a scale is useful to categorize splenic injuries, it does not predict the need for surgical intervention. The decision to proceed to surgery is multifactorial and depends on clinical and imaging analysis.59 Treatment of splenic injury is aimed at conservative management, maximizing salvage therapy. Some of the CT prognostic factors associated with failure of nonoperative management include AAST injury grade, amount of hemoperitoneum, and evidence of ongoing hemorrhage.10,78,85,127

Angiography Angiography and splenic artery embolization have become important procedures in the nonoperative management of patients with blunt splenic injury.41,79 Meta-analysis performed by Requarth et al. demonstrated that the addition of angiography and splenic artery embolization to the overall nonoperative management provided significantly improved failure rates when compared to conservative management that did not include angiography and splenic artery embolization.90 Furthermore, infectious complication rates associated with splenic salvage and nonoperative management were found to be approximately 5% when compared to splenectomy at 32%.20 If a bleeding vessel or pseudoaneurysm is noted, embolization can be performed. Splenic artery embolization is performed either at the level of the main splenic artery or at the selected segmental level. Embolization of the main splenic artery eliminates all inflow to the spleen except for collateral flow through short gastric arteries. Additional collateral vessels may form over time. Selective embolization is technically more difficult to perform but preserves flow to more of the spleen.

A

In the appropriate setting, angiography is indicated for patients in whom contrast extravasation and active hemorrhage are demonstrated on the initial CT scan. Sclafani’s group documented a splenic salvage rate of 93% to 97%.98,99 Some providers have added higher-grade injury as an indication for angiography and subsequent embolization. Haan and associates reported a nonoperative salvage rate of 94%, with an 83% salvage rate for grade IV and V injuries; they also reported that the presence of an arteriovenous fistula was predictive of a 40% nonoperative failure rate.40 Miller et al. compared the failure rates between nonoperative management with and without splenic angiography and embolization for the treatment of grade III to V blunt splenic injuries. They determined significantly improved failure rates with splenic angiography and embolization, with a failure rate of 5% compared to 15% to 31%.71 Relative contraindications to splenic embolization include hemodynamic instability, peritonitis, and other indications for emergent surgery. One of the potential advantages of splenic salvage is retention of immune function. To indicate splenic function, Pirasteh and colleagues looked at the presence of Howell-Jolly bodies on peripheral smears obtained from blood samples in patients who had remote splenic embolization therapy. Howell-Jolly bodies are normally removed by the functioning spleen. They determined that none of the 32 patients studied were found to have Howell-Jolly bodies on peripheral smear, indicating retained splenic phagocytic function.83

SPLENIC INFARCT Splenic infarct occurs commonly and has various causes. The splenic arterial branches are end arteries with no intercommunication. Causes of splenic infarct include embolic disease (e.g., mitral valve disease, endocarditis, atheromatous plaques), arteritis, splenic artery aneurysm or occlusion (pancreatic disease), sickle cell disease (thrombosis), and mass lesions. Most cases are asymptomatic, although some patients experience acute onset of left upper quadrant pain. Infarcts are typically focal but can be diffuse. Infarcts in younger patients are more often due to thrombosis from a hematologic disorder; infarcts in older patients tend to be embolic. On CT and MRI, infarcts are usually wedge shaped but may be irregular in contour (Fig. 48-16). Although the infarct has no

B

FIG 48-16  Infarcts. Axial (A) and coronal (B) CT images demonstrate peripheral low-attenuating wedgeshaped infarcts.

1553

1554

PART II  CT and MR Imaging of the Whole Body

enhancement, there may be peripheral enhancement of the capsule. Infarcts can also be multiple. As described previously, differentiating between infarct and abscess or tumor can be difficult. The lack of mass effect and perisplenic changes suggests infarct. Over time, infarcts shrink. Any increase in size is suspicious for superimposed infection. On MRI the signal of the infarct can vary depending on its composition and age.52 Hemorrhagic infarcts may be of high signal intensity on T1- and T2-weighted images.52 Chronic infarcts have low signal intensity on T1 and high signal intensity on T2. Chronic splenic infarction, as seen in homozygous sickle cell anemia, leads to near-complete calcification and atrophy of the spleen (Fig. 48-17). Adults with heterozygous hemoglobinopathies may have enlarged spleens with subcapsular calcification.

MISCELLANEOUS DISORDERS OF THE SPLEEN Splenosis Splenosis is caused by autotransplantation of splenic tissue. It usually occurs after trauma. The implants are typically located throughout the peritoneum (Fig. 48-18), though thoracic implants have been described, with implantation depending upon the extent of traumatic injury. The diagnosis can be confirmed by nuclear medicine 99mTc sulfur colloid imaging.

Inflammatory Pseudotumor Inflammatory pseudotumor of the spleen is a rare benign lesion composed of localized areas of inflammatory cells.50 The pathogenesis is unclear, with some proposing the lesion results from an inflammatory reparative response.49,121 On CT, such areas have low attenuation on nonenhanced images and show progressive opacification after IV administration of contrast.87 Calcifications may be present. Inflammatory pseudotumors reportedly have low signal intensity on T2- and delayed enhancement on T1-weighted images after IV administration of gadolinium; however, because there are few reports, the imaging characteristics of these lesions have not been established.50

REFERENCES

FIG 48-17  CT shows a calcified atrophic spleen (arrow) in a patient with homozygous sickle cell anemia.

A

1. Abbott RM, Levy AD, Aguilera NS, et al: From the archives of the AFIP: Primary vascular neoplasms of the spleen: Radiologic-pathologic correlation. Radiographics 24(4):1137–1163, 2004. 2. Ahmann DL, Kiely JM, Harrison EG, Jr, et al: Malignant lymphoma of the spleen. A review of 49 cases in which the diagnosis was made at splenectomy. Cancer 19(4):461–469, 1966. 3. Ahmed S, Horton KM, Fishman EK: Splenic incidentalomas. Radiol Clin North Am 49(2):323–347, 2011. 4. Allen KB, Andrews G: Pediatric wandering spleen–The case for splenopexy: Review of 35 reported cases in the literature. J Pediatr Surg 24(5):432–435, 1989. 5. Anderson SW, Varghese JC, Lucey BC, et al: Blunt splenic trauma: Delayed-phase CT for differentiation of active hemorrhage from contained vascular injury in patients. Radiology 243(1):88–95, 2007.

B FIG 48-18  Splenosis. Axial (A) and coronal (B) postsplenectomy CT images showing small globular areas of splenic tissue (arrows).

CHAPTER 48  Spleen 6. Ballard RB, Rozycki GS, Newman PG, et al: An algorithm to reduce the incidence of false-negative FAST examinations in patients at high risk for occult injury. Focused assessment for the sonographic examination of the trauma patient. J Am Coll Surg 189(2):145–150, discussion 150–141, 1999. 7. Balthazar EJ, Hilton S, Naidich D, et al: CT of splenic and perisplenic abnormalities in septic patients. AJR Am J Roentgenol 144(1):53–56, 1985. 8. Bode PJ, Niezen RA, van Vugt AB, et al: Abdominal ultrasound as a reliable indicator for conclusive laparotomy in blunt abdominal trauma. J Trauma 34(1):27–31, 1993. 9. Boraschi P, Donati F, Volpi A, et al: On the AJR viewbox. Intrapancreatic accessory spleen: Diagnosis with RES-specific contrast-enhanced MRI. AJR Am J Roentgenol 184(5):1712–1713, 2005. 10. Boscak A, Shanmuganathan K: Splenic trauma: What is new? Radiol Clin North Am 50(1):105–122, 2012. 11. Boulanger BR, Kearney PA, Brenneman FD, et al: Utilization of FAST (focused assessment with sonography for trauma) in 1999: Results of a survey of North American trauma centers. Am Surg 66(11):1049–1055, 2000. 12. Boulanger BR, McLellan BA, Brenneman FD, et al: Prospective evidence of the superiority of a sonography-based algorithm in the assessment of blunt abdominal injury. J Trauma 47(4):632–637, 1999. 13. Britt AR, Francis IR, Glazer GM, et al: Sarcoidosis: Abdominal manifestations at CT. Radiology 178(1):91–94, 1991. 14. Chen KT, Bolles JC, Gilbert EF: Angiosarcoma of the spleen: A report of two cases and review of the literature. Arch Pathol Lab Med 103(3):122–124, 1979. 15. Chulay JD, Lankerani MR: Splenic abscess. Report of 10 cases and review of the literature. Am J Med 61(4):513–522, 1976. 16. Chun CH, Raff MJ, Contreras L, et al: Splenic abscess. Medicine (Baltimore) 59(1):50–65, 1980. 17. Conlon S, Royston D, Murphy P: Splenic hamartoma. Cytopathology 18(3):200–202, 2007. 18. Dachman AH, Buck JL, Krishnan J, et al: Primary non-Hodgkin’s splenic lymphoma. Clin Radiol 53(2):137–142, 1998. 19. Dachman AH, Ros PR, Murari PJ, et al: Nonparasitic splenic cysts: A report of 52 cases with radiologic-pathologic correlation. AJR Am J Roentgenol 147(3):537–542, 1986. 20. Demetriades D, Scalea TM, Degiannis E, et al: Blunt splenic trauma: Splenectomy increases early infectious complications: A prospective multicenter study. J Trauma Acute Care Surg 72(1):229–234, 2012. 21. Dodds WJ, Taylor AJ, Erickson SJ, et al: Radiologic imaging of splenic anomalies. AJR Am J Roentgenol 155(4):805–810, 1990. 22. Doolas A, Nolte M, McDonald OG, et al: Splenic cysts. J Surg Oncol 10(5):369–387, 1978. 23. Elsayes KM, Narra VR, Mukundan G, et al: MR imaging of the spleen: Spectrum of abnormalities. Radiographics 25(4):967–982, 2005. 24. Emery KH: Splenic emergencies. Radiol Clin North Am 35(4):831–843, 1997. 25. Eshaghi N, Ros PR: Imaging of splenic masses. Contemp Diagn Radiol 12:1–6, 1989. 26. Falk S, Krishnan J, Meis JM: Primary angiosarcoma of the spleen. A clinicopathologic study of 40 cases. Am J Surg Pathol 17(10):959–970, 1993. 27. Falk S, Stutte HJ: Primary malignant lymphomas of the spleen. A morphologic and immunohistochemical analysis of 17 cases. Cancer 66(12):2612–2619, 1990. 28. Falk S, Stutte HJ, Frizzera G: Littoral cell angioma. A novel splenic vascular lesion demonstrating histiocytic differentiation. Am J Surg Pathol 15(11):1023–1033, 1991. 29. Farahmand N, Sirlin CB, Brown MA, et al: Hypotensive patients with blunt abdominal trauma: Performance of screening US. Radiology 235(2):436–443, 2005. 30. Fasih N, Gulati A, Ryan J, et al: The mysterious organ. Spectrum of focal lesions within the splenic parenchyma: Cross-sectional imaging with emphasis on magnetic resonance imaging. Can Assoc Radiol J 65(1):19–28, 2014.

1555

31. Federle MP, Courcoulas AP, Powell M, et al: Blunt splenic injury in adults: Clinical and CT criteria for management, with emphasis on active extravasation. Radiology 206(1):137–142, 1998. 32. Federle MP, Goldberg HI, Kaiser JA, et al: Evaluation of abdominal trauma by computed tomography. Radiology 138(3):637–644, 1981. 33. Ferrozzi F, Bova D, Draghi F, et al: CT findings in primary vascular tumors of the spleen. AJR Am J Roentgenol 166(5):1097–1101, 1996. 34. Fishman EK, Kuhlman JE, Jones RJ: CT of lymphoma: spectrum of disease. Radiographics 11(4):647–669, 1991. 35. Folz SJ, Johnson CD, Swensen SJ: Abdominal manifestations of sarcoidosis in CT studies. J Comput Assist Tomogr 19(4):573–579, 1995. 36. Franquet T, Montes M, Lecumberri FJ, et al: Hydatid disease of the spleen: Imaging findings in nine patients. AJR Am J Roentgenol 154(3):525–528, 1990. 37. Freeman JL, Jafri SZ, Roberts JL, et al: CT of congenital and acquired abnormalities of the spleen. Radiographics 13(3):597–610, 1993. 38. Gaetke-Udager K, Wasnik AP, Kaza RK, et al: Multimodality imaging of splenic lesions and the role of non-vascular, image-guided intervention. Abdom Imaging 39(3):570–587, 2014. 39. Gavant ML, Schurr M, Flick PA, et al: Predicting clinical outcome of nonsurgical management of blunt splenic injury: using CT to reveal abnormalities of splenic vasculature. AJR Am J Roentgenol 168(1):207– 212, 1997. 40. Haan JM, Bochicchio GV, Kramer N, et al: Nonoperative management of blunt splenic injury: A 5-year experience. J Trauma 58(3):492–498, 2005. 41. Hagiwara A, Yukioka T, Ohta S, et al: Nonsurgical management of patients with blunt splenic injury: Efficacy of transcatheter arterial embolization. AJR Am J Roentgenol 167(1):159–166, 1996. 42. Reference deleted in proofs. 43. Hill SC, Damaska BM, Ling A, et al: Gaucher disease: Abdominal MR imaging findings in 46 patients. Radiology 184(2):561–566, 1992. 44. Hoeffel C, Bokemeyer C, Hoeffel JC, et al: CT hepatic and splenic appearances with sarcoidosis. Eur J Radiol 23(2):94–96, 1996. 45. Hoff WS, Holevar M, Nagy KK, et al: Practice management   guidelines for the evaluation of blunt abdominal trauma: The East Practice Management Guidelines Work Group. J Trauma 53(3):602– 615, 2002. 46. Hopper KD, Chantelois AE: The retrorenal spleen. Implications for percutaneous left renal invasive procedures. Invest Radiol 24(8):592– 595, 1989. 47. Iannaccone R, Federle MP, Brancatelli G, et al: Peliosis hepatis: Spectrum of imaging findings. AJR Am J Roentgenol 187(1):W43–W52, 2006. 48. Ilica AT, Kocaoglu M, Zeybek N, et al: Extrahepatic abdominal hydatid disease caused by Echinococcus granulosus: Imaging findings. AJR Am J Roentgenol 189(2):337–343, 2007. 49. Inada T, Yano T, Shima S, et al: Inflammatory pseudotumor of the spleen. Intern Med 31(7):941–945, 1992. 50. Irie H, Honda H, Kaneko K, et al: Inflammatory pseudotumors of the spleen: CT and MRI findings. J Comput Assist Tomogr 20(2):244–248, 1996. 51. Israel O, Keidar Z, Bar-Shalom R: Positron emission tomography in the evaluation of lymphoma. Semin Nucl Med 34(3):166–179, 2004. 52. Ito K, Mitchell DG, Honjo K, et al: MR imaging of acquired abnormalities of the spleen. AJR Am J Roentgenol 168(3):697–702, 1997. 53. Jeffrey RB, Laing FC, Federle MP, et al: Computed tomography of splenic trauma. Radiology 141(3):729–732, 1981. 54. Joffe N, Antonioli DA: Penetration into spleen by benign gastric ulcers. Clin Radiol 32(2):177–181, 1981. 55. Johnson TL, Barnett JL, Appelman HD, et al: Candida hepatitis. Histopathologic diagnosis. Am J Surg Pathol 12(9):716–720, 1988. 56. Kamaya A, Weinstein S, Desser TS: Multiple lesions of the spleen: Differential diagnosis of cystic and solid lesions. Semin Ultrasound CT MR 27(5):389–403, 2006. 57. Kaza RK, Azar S, Al-Hawary MM, et al: Primary and secondary neoplasms of the spleen. Cancer Imaging 10:173–182, 2010.

1556

PART II  CT and MR Imaging of the Whole Body

58. Kluger Y, Paul DB, Raves JJ, et al: Delayed rupture of the spleen–Myths, facts, and their importance: Case reports and literature review.   J Trauma 36(4):568–571, 1994. 59. Kohn JS, Clark DE, Isler RJ, et al: Is computed tomographic grading of splenic injury useful in the nonsurgical management of blunt trauma?   J Trauma 36(3):385–389, discussion 390, 1994. 60. Lam KY, Tang V: Metastatic tumors to the spleen: A 25-year clinicopathologic study. Arch Pathol Lab Med 124(4):526–530, 2000. 61. Lankisch PG: The spleen in inflammatory pancreatic disease. Gastroenterology 98(2):509–516, 1990. 62. Lee CH, Haaland B, Earnest A, et al: Use of positive oral contrast agents in abdominopelvic computed tomography for blunt abdominal injury: Meta-analysis and systematic review. Eur Radiol 23(9):2513–2521, 2013. 63. Lee BC, Ormsby EL, McGahan JP, et al: The utility of sonography for the triage of blunt abdominal trauma patients to exploratory laparotomy. AJR Am J Roentgenol 188(2):415–421, 2007. 64. Leite NP, Kased N, Hanna RF, et al: Cross-sectional imaging of extranodal involvement in abdominopelvic lymphoproliferative malignancies. Radiographics 27(6):1613–1634, 2007. 65. Lewis JH, Patel HR, Zimmerman HJ: The spectrum of hepatic candidiasis. Hepatology 2(4):479–487, 1982. 66. Livingston DH, Lavery RF, Passannante MR, et al: Admission or observation is not necessary after a negative abdominal computed tomographic scan in patients with suspected blunt abdominal trauma: Results of a prospective, multi-institutional trial. J Trauma 44(2):273– 280, discussion 280–272, 1998. 67. Lucey BC, Boland GW, Maher MM, et al: Percutaneous nonvascular splenic intervention: A 10-year review. AJR Am J Roentgenol 179(6):1591–1596, 2002. 68. Maves CK, Caron KH, Bisset GS, 3rd, et al: Splenic and hepatic peliosis: MR findings. AJR Am J Roentgenol 158(1):75–76, 1992. 69. Metser U, Even-Sapir E: The role of 18F-FDG PET/CT in the evaluation of solid splenic masses. Semin Ultrasound CT MR 27(5):420–425, 2006. 70. Metser U, Miller E, Kessler A, et al: Solid splenic masses: Evaluation with 18F-FDG PET/CT. J Nucl Med 46(1):52–59, 2005. 71. Miller PR, Chang MC, Hoth JJ, et al: Prospective trial of angiography and embolization for all grade III to V blunt splenic injuries: Nonoperative management success rate is significantly improved. J Am Coll Surg 218(4):644–648, 2014. 72. Miller LA, Mirvis SE, Shanmuganathan K, et al: CT diagnosis of splenic infarction in blunt trauma: Imaging features, clinical significance and complications. Clin Radiol 59(4):342–348, 2004. 73. Miller MT, Pasquale MD, Bromberg WJ, et al: Not so FAST. J Trauma 54(1):52–59, discussion 59–60, 2003. 74. Moore EE, Cogbill TH, Jurkovich GJ, et al: Organ injury scaling: Spleen and liver (1994 revision). J Trauma 38(3):323–324, 1995. 75. Moore EH, Russell LA, Klein JS, et al: Bacillary angiomatosis in patients with AIDS: Multiorgan imaging findings. Radiology 197(1):67–72, 1995. 76. Murray JG, Patel MD, Lee S, et al: Microabscesses of the liver and spleen in AIDS: Detection with 5-MHz sonography. Radiology 197(3):723–727, 1995. 77. Nelken N, Ignatius J, Skinner M, et al: Changing clinical spectrum of splenic abscess. A multicenter study and review of the literature. Am J Surg 154(1):27–34, 1987. 78. Olthof DC, Joosse P, van der Vlies CH, et al: Prognostic factors for failure of nonoperative management in adults with blunt splenic injury: A systematic review. J Trauma Acute Care Surg 74(2):546–557, 2013. 79. Pachter HL, Guth AA, Hofstetter SR, et al: Changing patterns in the management of splenic trauma: The impact of nonoperative management. Ann Surg 227(5):708–717, discussion 717–709, 1998. 80. Park JY, Song KT: Splenic cyst: A case report and review of literature. Am Surg 37(9):544–547, 1971. 81. Pastakia B, Shawker TH, Thaler M, et al: Hepatosplenic candidiasis: Wheels within wheels. Radiology 166(2):417–421, 1988. 82. Pedrosa I, Saiz A, Arrazola J, et al: Hydatid disease: Radiologic and pathologic features and complications. Radiographics 20(3):795–817, 2000.

83. Pirasteh A, Snyder LL, Lin R, et al: Temporal assessment of splenic function in patients who have undergone percutaneous image-guided splenic artery embolization in the setting of trauma. J Vasc Interv Radiol 23(1):80–82, 2012. 84. Polat P, Kantarci M, Alper F, et al: Hydatid disease from head to toe. Radiographics 23(2):475–494, quiz 536–477, 2003. 85. Reference deleted in proofs. 86. Pyatt RS, Williams ED, Clark M, et al: Case report. CT diagnosis of splenic cystic lymphangiomatosis. J Comput Assist Tomogr 5(3):446– 448, 1981. 87. Rabushka LS, Kawashima A, Fishman EK: Imaging of the spleen: CT with supplemental MR examination. Radiographics 14(2):307–332, 1994. 88. Radin DR: Intraabdominal Mycobacterium tuberculosis vs Mycobacterium avium-intracellulare infections in patients with AIDS: Distinction based on CT findings. AJR Am J Roentgenol 156(3):487– 491, 1991. 89. Ramani M, Reinhold C, Semelka RC, et al: Splenic hemangiomas and hamartomas: MR imaging characteristics of 28 lesions. Radiology 202(1):166–172, 1997. 90. Requarth JA, D’Agostino RB, Jr, Miller PR: Nonoperative management of adult blunt splenic injury with and without splenic artery embolotherapy: A meta-analysis. J Trauma 71(4):898–903, discussion 903, 2011. 91. Rini JN, Leonidas JC, Tomas MB, et al: 18F-FDG PET versus CT for evaluating the spleen during initial staging of lymphoma. J Nucl Med 44(7):1072–1074, 2003. 92. Robertson F, Leander P, Ekberg O: Radiology of the spleen. Eur Radiol 11(1):80–95, 2001. 93. Romero JR, Barksdale EM, Jr: Wandering spleen: A rare cause of abdominal pain. Pediatr Emerg Care 19(6):412–414, 2003. 94. Ros PR, Moser RP, Jr, Dachman AH, et al: Hemangioma of the spleen: Radiologic-pathologic correlation in ten cases. Radiology 162(1 Pt 1): 73–77, 1987. 95. Rossetti F, Brawner DL, Bowden R, et al: Fungal liver infection in marrow transplant recipients: Prevalence at autopsy, predisposing factors, and clinical features. Clin Infect Dis 20(4):801–811, 1995. 96. Rothlin MA, Naf R, Amgwerd M, et al: Ultrasound in blunt abdominal and thoracic trauma. J Trauma 34(4):488–495, 1993. 97. Saboo SS, Krajewski KM, O’Regan KN, et al: Spleen in haematological malignancies: Spectrum of imaging findings. Br J Radiol 85(1009): 81–92, 2012. 98. Sclafani SJ, Shaftan GW, Scalea TM, et al: Nonoperative salvage of computed tomography-diagnosed splenic injuries: utilization of angiography for triage and embolization for hemostasis. J Trauma 39(5):818–825, discussion 826–817, 1995. 99. Sclafani SJ, Weisberg A, Scalea TM, et al: Blunt splenic injuries: Nonsurgical treatment with CT, arteriography, and transcatheter arterial embolization of the splenic artery. Radiology 181(1):189–196, 1991. 100. Scott GC, Berman JM, Higgins JL, Jr: CT patterns of nodular hepatic and splenic sarcoidosis: A review of the literature. J Comput Assist Tomogr 21(3):369–372, 1997. 101. Semelka RC, Kelekis NL, Sallah S, et al: Hepatosplenic fungal disease: Diagnostic accuracy and spectrum of appearances on MR imaging. AJR Am J Roentgenol 169(5):1311–1316, 1997. 102. Shanmuganathan K, Mirvis SE, Sherbourne CD, et al: Hemoperitoneum as the sole indicator of abdominal visceral injuries: A potential limitation of screening abdominal US for trauma. Radiology 212(2):423–430, 1999. 103. Shanmuganathan K, Mirvis SE, Sover ER: Value of contrast-enhanced CT in detecting active hemorrhage in patients with blunt abdominal or pelvic trauma. AJR Am J Roentgenol 161(1):65–69, 1993. 104. Shimono T, Yamaoka T, Nishimura K, et al: Peliosis of the spleen: Splenic rupture with intraperitoneal hemorrhage. Abdom Imaging 23(2):201–202, 1998. 105. Shirkhoda A, Ros PR, Farah J, et al: Lymphoma of the solid abdominal viscera. Radiol Clin North Am 28(4):785–799, 1990.

CHAPTER 48  Spleen 106. Shuman WP: CT of blunt abdominal trauma in adults. Radiology 205(2):297–306, 1997. 107. Reference deleted in proofs. 108. Solomou EG, Patriarheas GV, Mpadra FA, et al: Asymptomatic adult cystic lymphangioma of the spleen: Case report and review of the literature. Magn Reson Imaging 21(1):81–84, 2003. 109. Stengel D, Bauwens K, Rademacher G, et al: Emergency ultrasoundbased algorithms for diagnosing blunt abdominal trauma. Cochrane Database Syst Rev (7):CD004446, 2013. 110. Stuhlfaut JW, Soto JA, Lucey BC, et al: Blunt abdominal trauma: Performance of CT without oral contrast material. Radiology 233(3):689–694, 2004. 111. Takayama A, Nakashima O, Kobayashi K, et al: Splenic lymphangioma with papillary endothelial proliferation: A case report and review of the literature. Pathol Int 53(7):483–488, 2003. 112. Taylor AJ, Dodds WJ, Erickson SJ, et al: CT of acquired abnormalities of the spleen. AJR Am J Roentgenol 157(6):1213–1219, 1991. 113. Taylor FC, Frankl HD, Riemer KD: Late presentation of splenic trauma after routine colonoscopy. Am J Gastroenterol 84(4):442–443, 1989. 114. Thompson WM, Levy AD, Aguilera NS, et al: Angiosarcoma of the spleen: Imaging characteristics in 12 patients. Radiology 235(1): 106–115, 2005. 115. Thompson SE, Walsh EA, Cramer BC, et al: Radiological features of a symptomatic splenic hamartoma. Pediatr Radiol 26(9):657–660, 1996. 116. Torres GM, Terry NL, Mergo PJ, et al: MR imaging of the spleen. Magn Reson Imaging Clin N Am 3(1):39–50, 1995. 117. Townsend RR: CT of AIDS-related lymphoma. AJR Am J Roentgenol 156(5):969–974, 1991. 118. Tso P, Rodriguez A, Cooper C, et al: Sonography in blunt abdominal trauma: A preliminary progress report. J Trauma 33(1):39–43, discussion 43–34, 1992.

1557

119. Urban BA, Fishman EK: Helical CT of the spleen. AJR Am J Roentgenol 170(4):997–1003, 1998. 120. Urrutia M, Mergo PJ, Ros LH, et al: Cystic masses of the spleen: Radiologic-pathologic correlation. Radiographics 16(1):107–129, 1996. 121. Kalaivani V, Vijayakumar HM, Grish KS, et al: Inflammatory pseudotumour of the spleen: A diagnostic dilemma. J Clin Diagn Res 7(7):1460–1462, 2013. 122. Wadsworth DT, Newman B, Abramson SJ, et al: Splenic lymphangiomatosis in children. Radiology 202(1):173–176, 1997. 123. Warshauer DM: Splenic sarcoidosis. Semin Ultrasound CT MR 28(1):21–27, 2007. 124. Warshauer DM, Dumbleton SA, Molina PL, et al: Abdominal CT findings in sarcoidosis: Radiologic and clinical correlation. Radiology 192(1):93–98, 1994. 125. Warshauer DM, Hall HL: Solitary splenic lesions. Semin Ultrasound CT MR 27(5):370–388, 2006. 126. Warshauer DM, Semelka RC, Ascher SM: Nodular sarcoidosis of the liver and spleen: Appearance on MR images. J Magn Reson Imaging 4(4):553–557, 1994. 127. Weissleder R, Elizondo G, Stark DD, et al: The diagnosis of splenic lymphoma by MR imaging: Value of superparamagnetic iron oxide. AJR Am J Roentgenol 152(1):175–180, 1989. 128. West JG, Trunkey DD, Lim RC: Systems of trauma care. A study of two counties. Arch Surg 114(4):455–460, 1979. 129. Wolfman NT, Bechtold RE, Scharling ES, et al: Blunt upper abdominal trauma: Evaluation by CT. AJR Am J Roentgenol 158(3):493–501, 1992. 130. Zerem E, Bergsland J: Ultrasound guided percutaneous treatment for splenic abscesses: The significance in treatment of critically ill patients. World J Gastroenterol 12(45):7341–7345, 2006.

49  Peritoneum John R. Haaga and David R. Di Lorenzo

Computed tomography (CT) and magnetic resonance imaging (MRI) have proven useful for the diagnosis of many peritoneal and mesenteric problems. Even in the absence of a primary peritoneal problem, many secondary signs that add valuable insights about other disorders may be observed. This chapter discusses the embryology, normal anatomy, and physiology of the peritoneum and retroperitoneum and a wide spectrum of relevant pathologic disorders. Because of the intimate relationship of these spaces to the various organs, a thorough knowledge of them improves one’s understanding of the various abdominal problems.

EMBRYOLOGY The gastrointestinal (GI) tract is formed when the endoderm develops in a tube within the coelomic cavity. It acquires a single-layer covering of mesodermal cells that evolve from the space between the ectoderm and endoderm (Fig. 49-1A). The dorsal or posterior mesoderm forms the dorsal mesentery, which eventually composes the visceral peritoneum of the abdomen (the pericardium and pleura in the chest; see Fig. 49-1A). The anterior splanchnic mesoderms fuse to form the ventral (anterior) mesentery and later completely disappear, leaving the falciform ligament of the liver as the only remnant (see Fig. 49-1B and C). Within the transverse septum (from which the diaphragm evolves), the liver forms and moves into the coelomic cavity, carrying an investment of peritoneum. The liver’s only attachments to the transverse septum are small portions of the peritoneum, which become the triangular and coronary ligaments (behind which are the “bare areas” of the liver). The blood vessels (also lymphatics) to the various developing viscera course through the posterior or dorsal mesentery. The anterior attachment of the mesentery between the liver and the stomach, esophagus, and duodenum becomes the lesser omentum, consisting of  two portions—the hepatogastric and hepatoduodenal ligaments (see Fig. 49-1B). One can best understand the relationship of the pancreas to the omental bursa by considering the rotation of the bowel and changes in the dorsal mesentery. The pancreas forms from endodermal buds of the duodenum, the (ventral) anterior bud being slightly caudal but close to the common bile duct and the (dorsal) posterior bud being directly opposite it. As the C loop of the duodenum rotates, the two pancreatic buds come together and fuse at the level of the head. The ventral segment remains as the uncinate and head portion. The spleen forms in the (dorsal) posterior mesentery and is carried laterally as the omental bursa forms. The dorsal mesentery containing the pancreas fuses with the posterior abdominal peritoneum (see Fig. 49-1C). The multiple layers of the embryologic mesentery as it relates to the retroperitoneum, pancreas, and colon can be seen in Figure 49-1C.

1558

The lesser omentum and omental bursa are formed by the rotation of the stomach and duodenum and the posterior movement of the pancreas (Fig. 49-1B to D). The anterior aspect of the bursa is limited by the lesser omentum, composed of the hepatoduodenal and hepatoduodenal ligaments, the posterior wall of the stomach, the gastrocolic ligament, and the mesocolon. The posterior wall is bounded by the retroperitoneum over the top of the pancreas. The lateral wall consists of the gastrocolic, gastrosplenic, and splenorenal ligaments (see Fig. 49-1C). The other boundaries of the omental bursa are described later under “Normal Peritoneal Anatomy.” As the bowel and colon rotate, the mesentery of the mesocolon fuses posteriorly with the covering of the retroperitoneum (see Fig. 49-1E). If the fusion is complete, the spaces remain intact and conform to the traditional anatomic concepts of the retroperitoneum and the peritoneum (i.e., the peritoneal space is the anterior space and has a number of different compartments).46 Traditionally the retroperitoneum is divided into three parts: the anterior pararenal, posterior pararenal, and perirenal spaces. After extensive CT experience, it has become apparent that the traditional concept cannot explain some observations. Dodds and coworkers reviewed the embryologic formation of the retroperitoneum and proposed a new concept that reconciles some of these inconsistent observations.12,13

Dodds’ Concepts of Retroperitoneum and Peritoneum With the extensive CT experience gained over the years, it has become apparent that the traditional descriptions of the retroperitoneum and posterior portions of the peritoneum do not fit some observations. The inconsistencies of the traditional approach involve primarily the relationship of the anterior pararenal space to the retroperitoneum and peritoneum.13,68 Authors have noted that the lateroconal fascia is not constant in its location and does not really function as a barrier to the spread of fluid or inflammation. Fluid from the pancreatic region does not always spread to the margins of the colon, which are supposedly in the same space. After reviewing the embryologic information, Dodds and colleagues came to several conclusions that differed from the traditional description.12,13 During embryologic development, the gut and its derivatives form in the intermesenteric space; they remain in this space regardless of any migration, folding, or fusion. Communication among the various components of the intermesenteric space follows the areolar connective tissue (the course of which can be followed by noting the vasculature; see Fig. 49-1B to D). Therefore communication between the retroperitoneum and intermesenteric space occurs only by means of spaces along the vessels, predominantly at the root of the mesentery. A number of rotations and fusions occur before the adult form results. With rotation of the bowel, the ventral and dorsal pancreata fuse and along with the duodenum lie against the embryologic

1559

CHAPTER 49  Peritoneum Intraembryonic coelomic cavity

Dorsal mesentery

Falciform lig. Liver

Somatic mesoderm layer

Gut

Lesser omentum Stomach

Pancreas

Spleen Omental bursa Dorsal mesogastrium Kidney

Surface ectoderm

A

B

Liver

Left leaf mesocolon Right leaf mesocolon Left leaf mesoduodenum Fused leaves of mesoduodenum Primary dorsal peritoneum Anterior renal fascia

Gastrolienal lig. Spleen

OB

Transversalis fascia

Peritoneum R. colon

L. colon

Pancreas

K L R

Transversalis fascia

C

Pancreas

Lienorenal lig.

D

Posterior renal fascia

Ascites

F

L

E Ascites

K

Ascites

K

L

K

K

K

L

L

FIG 49-1  Embryology of the GI tract. A, Embryologic formation of the gut as the endodermal canal within the layers of the mesoderm and ectoderm. The endoderm is attached posteriorly by the dorsal mesentery and anteriorly by the ventral mesentery (arrow). The coelomic cavity (which later becomes the peritoneal cavity) is lined by mesodermal cells (which later become the peritoneum). B, A later stage in the embryo shows the liver having formed in the anterior or ventral mesentery. This mesentery persists as the falciform ligament and the lesser omentum between the liver and stomach. The spleen and dorsal pancreas form in the posterior or dorsal mesentery. A folding of the posterior mesentery and mesogastrium forms the omental bursa. The kidney and adrenal gland are located in the retroperitoneum. (According to Dodds and associates,13 the space within the mesentery is and should continue to be called the intermesenteric space regardless of any changes in configuration.) C, This diagram approximates the adult configuration, with the liver having moved to the right. The rotation of the stomach, spleen, lienorenal ligament, and pancreas produces the margins of the omental bursa (OB). D, Adult retroperitoneal spaces at the level of the left kidney. The primary retroperitoneum is divided by the anterior renal fascia (dashed lines) and the posterior renal fascia (dotted lines) into perirenal and posterior pararenal spaces. Fusion of the anterior renal fascia to the primary dorsal peritoneum eliminates any anterior pararenal space. Ventral to the primary retroperitoneum, a secondary retroperitoneum forms, consisting of a pancreaticoduodenal space overlying a retroperitoneal colonic space. These secondary spaces are demarcated by folded laminated leaves of mesentery that fuse with one another or the primary dorsal peritoneum. E and F, Retroperitoneal colonic space. Right leaf of the mesocolon (long and short dashes), anterior perirenal fascia (long dashes), posterior perirenal fascia (dots), and lateral fascia (dots and dashes). E, Axial section through the pancreas of a 14-week human fetus. During colonic rotation (arrows), the left and right colons swing toward the flanks so that their dorsal mesentery, or mesocolon, lies flat against the posterior abdominal wall. At this stage the mesentery of the duodenum and pancreas has already fused with the posterior peritoneum to form the pancreaticoduodenal space. Thus bilateral retroperitoneal colonic spaces come to lie ventral to the pancreaticoduodenal space but extend more caudal.  F, Depending on complete or incomplete fusion of these spaces (curved arrow), the pericolic space may be very shallow, extend slightly posteriorly (vertical arrow), or extend medially into the fascia (horizontal arrow). These changes are appreciated only when ascites is present. K, kidney; L, left; R, right. (D and E, From Dodds WJ, et al: The retroperitoneal spaces revisited: A pictorial essay. AJR Am J Roentgenol 147:1155, 1986. ©American Roentgen Ray Society, 1986.)

1560

PART II  CT and MR Imaging of the Whole Body

peritoneum. Fusion of the left side of the dorsal mesentery and the retroperitoneal embryologic peritoneum occurs at a line that traditionalists called the anterior renal fascia (see Fig. 49-1D and E) but that Dodds and colleagues call the retropancreatic peritoneal recess.13 It was given this name because it represents a potential space that can be “split” by certain pathologic processes in the area.4 At about this time, the colon rotates and the mesocolon fuses to  the mesentery over the pancreaticoduodenal space and laterally to the posterior parietal peritoneum (see Fig. 49-1E). Most commonly  the fusion is complete and extends from the ascending colon and the cecum on the right to the descending colon by the sigmoid on the left. If fusion is complete, the right and left colons are “sandwiched” between the flattened pancreaticoduodenal space and the retroperitoneum. If the fusion is not complete, a variation in the spaces results. Without complete fusion, the colon has a mesentery (cecum or right colon in 10% to 20% of cases) and there is no colonic retroperitoneal space. When this occurs the splenic flexure is occasionally located medial to the spleen. In the adult the distribution of ascitic fluid in the pericolic gutter depends on the shape of the recess and the extent of fusion of the mesocolon with the retroperitoneum. Without fusion, fluid or bowel can be located posterior to the kidney by the psoas muscles. According to Dodds’ concept, the compartments of the retroperitoneal spaces differ. The anterior renal fascia fuses with the primordial peritoneum over the retroperitoneum (thus there is no anterior pararenal space). The posterior renal fascia fuses posteriorly with the fascia over the psoas and quadratus lumborum muscles. Laterally it fuses with the peritoneum, which extends anteriorly inside the properitoneal fat. The cephalic and caudal boundaries also differ. First, below the level of the renal fascia, there are two separate spaces (anterior and posterior) that are different areas of the intermesenteric space and do not communicate directly. The anterior space contains the vessels of the sigmoid and ascending colon. The posterior space contains the retroperitoneal structures. The cephalic relationships also change. The anterior renal fascia does not extend above the adrenal glands; the bare areas of the liver and spleen thus correspond to the posterior pararenal space. The new approach can be summarized as follows. The previously designated anterior pararenal space is called the intermesenteric space, which is divided into two compartments, one for the colon and one for the pancreas and duodenum. Communication between the different components, such as the pancreaticoduodenal and colonic intermesenteric spaces, is along the vessels. When fusion of the mesothelium is not complete, there may be separation of the potential spaces by pathologic processes. With pancreatitis, for example, the primordial space splits with inflammatory fluid (which looks like the splitting of what was previously called the anterior pararenal space).4,57 The depth of the paracolonic spaces can vary posteriorly, and the position of the colic mesentery can also vary, depending on fusion of the mesentery. The retroperitoneum above and below the cone of Gerota’s fascia varies from the traditional viewpoint because there is only a posterior pararenal space (no anterior spaces).

NORMAL PERITONEAL ANATOMY The peritoneal spaces are potential spaces that are not normally visualized unless they are distended with fluid or the fascia is thickened. Knowledge of these anatomic spaces is critical for diagnosis, staging, and treatment of fluid collections by either surgical or radiologic methods.70 Here, normal anatomy is shown in axial, coronal, and sagittal images provided by thin sections from a multislice scanner. Using

1-mm sections and subsecond scanning, these reconstructions show remarkably accurate anatomic relationships (Figs. 49-2 to 49-6).

Peritoneum The peritoneum is a serous membrane that is covered with a single layer of mesothelial cells, enveloping all portions of the abdomen. The total surface area is about 1.7 m2, but the functional area is about 1 m2.32 The cavity is closed except for the fallopian tubes and contains 50 to 75 mL of clear fluid, which has a specific gravity of 1.016 and fewer than 3000 cells/mm2. The cells are lymphocytes, macrophages, eosinophils, mast cells, and rare free mesothelial cells. The peritoneum is normally sterile and free of bacteria.

Supracolic and Infracolic Spaces.  The abdomen can be divided into the supracolic, infracolic, pericolic (pericolonic), and pelvic spaces. The supracolic and infracolic spaces are separated by the transverse colon and its mesentery (see Figs. 49-2B and C, 49-3D, and 49-4). Anteriorly the greater omentum is located between the peritoneal wall and the bowel (see Fig. 49-4D). The supracolic space includes the perihepatic spaces and the omental bursa. The infracolic space is divided into the right and left spaces by the root of the mesentery, which extends left to right from the ligament of Treitz to the ileocecal area. Laterally the pericolonic gutters communicate with the perihepatic spaces and the pelvic region.

Perihepatic Spaces.  The perihepatic spaces consist of the subphrenic (subdiaphragmatic) spaces and subhepatic spaces. There are right, left, anterior, and posterior spaces in each area. Subphrenic space.  The right subphrenic space is continuous with the right subhepatic space and the right pericolic space. The right subphrenic space lies adjacent to the lateral and superior portion of the liver between the diaphragm and the body wall (see Figs. 49-2A-C, 49-3A-C, and 49-4). On axial and coronal images, the posterior margin of the subphrenic space ends at the anterior edge of the coronary ligaments, which encompasses the bare area of the liver against the posterior surface of the abdominal cavity (see Figs. 49-2C and D, 49-3A,  and 49-4). The right subphrenic space is limited medially by the falciform ligament (see Fig. 49-3A) in the upper portion of the abdomen. Because of variability in size, the ligament may or may not serve as an effective barrier to the spread of material to the left side (see Figs. 49-2A and 49-3A). The left subphrenic space is beneath the left diaphragm and surrounds the fundus of the stomach anteriorly, the spleen, and the space between the liver and the stomach (see Figs. 49-3A-D and 49-4D). Fluid can also be localized to various areas of the space (e.g., anterior, posterior). The left subhepatic space is immediately continuous with the anterior part of the left subphrenic space (see Figs. 49-2B and 49-3A and B). The lateral perisplenic portion is immediately continuous with the left pericolic space (see Fig. 49-3D). Although it is traditionally taught that the phrenicocolic ligament restricts the flow of material from the left pericolic space to the left subphrenic space, in reality it is seldom an effective barrier (see Fig. 49-2A and B). CT scans in numerous patients have shown variability in size, shape, and location of the ligament, which explains its inconsistent performance as a barrier to infection. Subhepatic space.  The subhepatic space is divided into right and left parts (see Figs. 49-2A, 49-3B, and 49-4). The right subhepatic space is adjacent to the gallbladder and extends posteriorly to the anterior surface of the retroperitoneum over the right kidney (see Figs. 49-2A and C, 49-3D, and 49-4A). It is immediately adjacent to the duodenum Text continued on p. 1568

Ish

F

RSP

LSP

gh

PCL

P

RS

LC RSP

M

RC

LC

LP RP

A

F

P

RSP

LSP HD g

RSP RS RSP

L gh ST

D

P

RI

LC

RC

LP RP

B FIG 49-2  Peritoneal spaces in three dimensions. A, Coronal image of the anterior abdomen in a patient with dilute contrast material injected into the peritoneum. The various portions of the peritoneal cavity can be identified. The right subphrenic space (RSP) extends from the falciform ligament (F) laterally between the abdominal wall and liver. This space then communicates with the right pericolic space (RC) and the right subhepatic space (RS). The right pericolic space flows into the right side of the pelvis (RP). The left subphrenic space (LSP) communicates with the left subhepatic space (lsh). Flow from these spaces is limited caudad by the phrenicocolic ligament (PCL). The left pericolic space (LC) extends caudad into the left side of the pelvis (LP). In the region of the suprahepatic spaces, the right subhepatic space (RS) is below the right lobe of the liver and extends into the lesser sac (arrow) beneath the edge of the lesser omentum, which contains the portal vein, hepatic artery, and common bile duct. This is immediately adjacent to the pancreas (P). The upper margin of the lesser sac is bounded by the gastrohepatic ligament (gh) containing the gastric vessels. The mesentery (M) is noted in the midline, containing the mesenteric vessels. B, Coronal scan of a patient with dilute contrast in the peritoneum. The peritoneum is generally divided into supracolic and infracolic (RI) spaces by the mesocolon, which extends from the anterior margin of the pancreas (P). The supracolic space encompasses all of the subphrenic, perihepatic, and perisplenic spaces. The right subphrenic space (RSP) extends from the upper space below the diaphragm (arrow) and laterally around the liver. The falciform ligament is not seen, which demonstrates why it may not always serve as a complete barrier between the right and left sides. The subphrenic space communicates with the right pericolic space (RC), which extends into the right pelvis (RP). The right subphrenic space also communicates medially with the right subhepatic space (RS) (see Fig. 49-13B), which is adjacent to the gallbladder (g) and duodenum (D). This space extends into the lesser sac (L) under the hepatoduodenal ligament (HD), which contains the portal vein, hepatic artery, and common duct. The lesser sac continues behind the stomach (ST), which is better seen on the sagittal views that follow. The superior boundary is the gastrohepatic ligament (gh). The left subphrenic space (LSP) is located below the diaphragm, around the left lobe and the stomach. It is limited inferiorly by the phrenicocolic ligament. The left pericolic space (LC) extends into the pelvis (LP). Continued

1562

PART II  CT and MR Imaging of the Whole Body

S v

LSP

SP

A RSP RS RC

P

C

M

P

P

S

L

LK

RK

P

IP

P

IP

D FIG 49-2, cont’d  C, Coronal image more posterior than B. This level demonstrates the attachment of the coronary ligaments of the right lobe of the liver (straight arrows) to the more posterior portion of the diaphragm and abdominal cavity. Communication of the right subphrenic space (RSP) with the right pericolic space (RC) and right subhepatic space (RS) is well seen. The broad attachment of the mesentery located on the posterior abdominal wall extends from the ligament of Treitz (upward curved arrow) to the right lower quadrant (curved arrow). The recesses of the lesser sac, the superior recess (S) adjacent to the vena cava (v), and the splenic recess (SP) behind the stomach are seen. The aorta (A) is also noted. The left subphrenic space (LSP) and phrenicocolic ligament are again noted above the left pericolic space (see B). The mesentery of the small bowel (M) and the psoas muscles (P) are seen. Contrast material in the pelvis is also noted. D, Posterior coronal section shows the retroperitoneum and posterior portions of the peritoneum. The most posterior portions of the liver (L) and spleen (S) have medial and posterior “bare areas” containing the adrenal glands (curved arrows). Note the contrast material in the most posterior portions of the right and left subphrenic spaces (horizontal arrows). The retroperitoneum contains the right (RK) and left (LK) kidneys, which are surrounded by a cone-shaped space containing the perirenal fat bounded by the anterior fascia of Zuckerkandl and the posterior fascia of Gerota (see Fig. 49-6C). The highest extent of the pararenal spaces extends cephalad under the diaphragm, contains the adrenal glands (curved arrows), and merges with the bare areas. The psoas muscles (P) extend distally and merge with the iliopsoas muscles (IP).

CHAPTER 49  Peritoneum

1563

F rsp C

ST gs Sr Isp

cl

A

P

F

alsp

a

L

rsp

P

p

B

F

ST

w

SR

P

LL g

D

L P sma

RS

C

F

P

gs plsp

FIG 49-3  Axial images of the peritoneum. A, Axial scan through the upper abdomen shows the perihepatic spaces in the supracolic region. The right subphrenic space (rsp) extends from the falciform ligament (F), anteriorly around the liver, and posteriorly to the coronary ligaments (cl; arrowhead). Behind these ligaments is the “bare area” of the liver, which is the most cephalic portion of the retroperitoneum (see Fig. 49-2D). The left subphrenic space (lsp) extends posteriorly around the left lobe of the liver, stomach, and spleen. At this high level, the peritoneal space surrounds the spleen on all surfaces; the attached bare area of the spleen is attached posteriorly at the level of the splenic pedicle (see C). The left subhepatic space (curved arrow) is contiguous with the left subphrenic space (lsp). Also at this level, the superior recess (arrows) of the lesser sac is seen around the caudate lobe of the liver (C). The splenic recess (Sr) of the lesser sac is noted behind the stomach (ST). The gastrohepatic ligament and the gastrosplenic ligament (gs) are noted. B, This axial scan is at a slightly lower level over the lesser sac. The right subphrenic space (rsp) is noted lateral to the liver; inflammatory processes may sequester collections into anterior (a) and posterior (p) components. The left subphrenic space can be subdivided into the anterior (alsp) and posterior (plsp) spaces, and it ends posteriorly at the bare area of the spleen. The anatomic boundaries of the lesser sac are noted. The sac (L) is anterior  to the pancreas (P) and extends posteriorly behind the stomach (ST) as the splenic recess (SR). The margin of the foramen of Winslow (w) contains the portal vein, the hepatic artery, and the common bile duct. The classic boundary within the lesser sac is the left gastric artery, which resides in the gastrohepatic ligament (arrow) and the gastrosplenic ligament (gs). This peritoneal reflection over the left gastric artery divides the inferior and superior components of the lesser sac. At this level the spleen has a retroperitoneal component, the “bare area” of the spleen, which corresponds to the splenorenal ligament posteriorly (arrow). The retroperitoneum around the kidney consists of several planes (see Fig. 49-1). C, Axial scan at the level of the pancreatic head shows the perihepatic spaces, as noted in A and B, as well as additional peripancreatic anatomy. The inferior portion of the lesser sac (L) is anterior to the pancreas. The posterior right subhepatic space (RS) is behind the gallbladder (g), between the posterior right lobe of the liver and the kidney. The anterior boundary of the lesser sac consists of the gastrohepatic ligament (vertical arrow). The head of the pancreas (P), superior mesenteric artery (sma), left lobe of the liver (LL), and air-filled duodenum (D) are seen. Continued

1564

PART II  CT and MR Imaging of the Whole Body

P

rsp

M LIC Isp

RS K

D

K

S

P

F

FIG 49-3, cont’d D, Axial scan at a lower level than C. At this level, the right subphrenic space (rsp) merges with the posterior right subhepatic space (RS; also known as Morison’s pouch). The left subphrenic space (lsp) also begins to merge with the left pericolic space at this level. The left infracolic space (LIC) is seen to the left of the mesentery (M) and head of the pancreas (P). The stomach is anterior, and the greater omentum (arrows) lies anteriorly on either side. The kidneys (K) and spleen (S) are noted. E, Axial scan at a level below D. The right subphrenic space lateral to the liver converges with the right pericolic space (RC) adjacent to the cecum (C). The left pericolic space (LC) is on the left. The right infracolic space (RIC), adjacent to the mesentery (M), and the left pericolic space (LC) are noted. The fat-density greater omentum (arrows) is noted in the anterior peritoneum. F, Lower axial scan shows the upper pelvis. The right (RC) and the left (LC) spaces are noted; the right space is lateral to the cecum (CE). The psoas muscles (P) are noted posteriorly, medial to the retroperitoneum and below the renal fossa.

RIC

RC

M LC

C

E

P

F

RC

CE P

F

F

P

LC p

CHAPTER 49  Peritoneum

RS

G

F

H

F

P

P

FIG 49-3, cont’d G, Axial scan in the pelvis shows communication of the pericolic spaces. Perirectal spaces (arrows) are lateral to the rectosigmoid (RS). H, Scan taken through the lowest part of the pelvis, showing perirectal spaces.

1565

rsp

D

RK RS

HF IC

F

A

SC

F

P

arsp RL

pv F PH

RI

B

F

F

FIG 49-4  Sagittal images of the peritoneum. A, Sagittal scan to the right of midline. The anterior right subphrenic space (rsp) is seen anterior to the liver, and the right subhepatic space (RS) is below the right lobe. The posterior attachment of the phrenic space limits the space posteriorly (curved arrow). The duodenum (D) and the upper portion of the hepatoduodenal ligament (vertical arrow) are noted adjacent to the subhepatic space (RS). Also noted are the hepatic flexure (HF) of the colon, the right kidney (RK), the infracolic space (IC), and the sigmoid colon (SC). B, Sagittal scan slightly lateral to midline. The liver (RL), the anterior right subphrenic space (arsp), and the infracolic space (RI) are seen. The area of the foramen of Winslow (F) at the entrance of the lesser sac is noted. The lesser sac is limited posteriorly by the pancreatic head (PH), inferiorly by the transverse mesocolon (vertical arrow), and anteriorly by the hepatoduodenal ligament (arrow). The portal vein (pv) is in the margin of the ligament.

CHAPTER 49  Peritoneum

LL alsp C P

D

A

S

L M

C

F

P

alsp

LL

SP ST

S

av P

D FIG 49-4, cont’d C, Sagittal image close to midline. The liver (LL) and adjacent structures are seen, including the anterior left subphrenic space (alsp), the superior recess of the lesser sac (small angled arrow), and the “bare area” of the liver (horizontal arrow). The lesser sac (L) is bounded anteriorly by the edge of the gastrohepatic ligament and the hepatoduodenal ligament containing the duodenum (D), posteriorly by the pancreas (P), and inferiorly by the mesocolon (M). The aorta (A), with the celiac artery (C) and the superior mesenteric artery (S), is also seen. D, Sagittal section to the left of midline. The left lobe of the liver (LL) is attached by the triangular ligaments to the diaphragm (curved arrow). The anterior left subphrenic space (alsp) is in front of the liver; the left subhepatic space is between the liver and the stomach (ST). The splenic recess (SP) is behind the stomach, in front of the spleen (S) and the pancreas (P). The splenic artery and vein (av) are seen in the hilum. The mesocolon (vertical arrow), with the splenic flexure, is seen clearly separating the supracolic and infracolic spaces. Anteriorly the greater omentum (horizontal arrows) extends from the margin of the stomach.

1567

1568

PART II  CT and MR Imaging of the Whole Body are variable from one patient to another and may extend high beside the caudate lobe and vena cava and into the porta hepatis (see Figs. 49-3A and 49-4C).

Pericolic Spaces.  The right and left pericolic spaces are contiguous u

A

u

B

with the pelvic region and subphrenic spaces (see Figs. 49-2A to C and 49-3E to G). These spaces receive fluid displaced from around the small bowel and are bounded laterally by the peritoneum of the abdominal wall and medially by the peritoneum over the colon. The depth and configuration of these spaces may vary depending on the completeness of the fusion of mesentery in the embryologic state. The pericolic space may thus extend posteriorly, adjacent to the kidney.71 The posterior extension occurs most commonly on the right side, which accounts  for the observation that the right pericolic space is deeper than the  left. This common variation has prompted new theories about the retroperitoneal spaces (see “Dodds’ Concepts of Retroperitoneum and Peritoneum” earlier). If a portion of the colon (e.g., splenic flexure) is not in the normal location, the configuration of such spaces may be difficult to predict. Regardless of the anatomic variation, however, one can be confident that CT accurately displays the precise pathologic anatomy (even if there is some controversy about the nomenclature).

Infracolic Spaces.  The right and left infracolic spaces are separated B FIG 49-5  A, Axial scan shows the uterus (u), with a fluid collection posteriorly (arrow). B, Fluid collection (arrow) is present behind the uterus (u) and the bladder (B).

by the root of the mesentery, which extends from the ligament of Treitz in the left upper quadrant to the ileocecal region in the right lower quadrant (see Figs. 49-2B, 49-3D, and 49-4A and D). The left infracolic space communicates with the lower portions of the abdomen, but the right infracolic space is confined by the mesocolon, ascending colon, and posterior attachment of the mesentery.

Pelvis and the foramen of Winslow, whose anterior edge consists of the edge of the lesser omentum, which contains the hepatic artery, portal vein, common bile duct, and lymph nodes; virtually all nodes from the perihepatic area—except the lower portion of the colon—converge at this area (see Figs. 49-2B, 49-3B and C, and 49-4B and C). The left subhepatic space is immediately beneath the left lobe of the liver and is continuous with the anterior part of the left subphrenic space (see Figs. 49-2A, 49-3A, and 49-4D). It is bounded posteriorly by the lesser omentum and on the sides by the capsular surface of the left lobe of the liver and the stomach. The omental bursa (lesser sac)12 is behind the stomach and anterior to the pancreas (see Figs. 49-2A and B, 49-3B and C, and 49-4B and C). The left margin is the gastrolienal (gastrosplenic) ligament, and  the anterior margin is the gastrohepatic ligament; the right margin is the medial surface of the coronary ligament and the edge of the lesser omentum, which is the boundary of the superior recess. The caudal boundary of the omental bursa is the gastrocolic reflection and  the mesocolon. The greater omentum extends caudad from the anterior edge of the fundus of the stomach and the gastrocolic ligament. It contains a potential space in the fused membranes that may be  split into a true space by an active process (e.g., bleeding, infection). The omental bursa can be divided into inferior and superior recesses by the reflection over the left gastric artery (see Figs. 49-2A and B and 49-3A to C). The superior recess is a potential space on the right side of the omental bursa and is seldom seen in patients with a transudative effusion. In cases of pseudocyst, infection, or processes that incite a pressure phenomenon, it may be visualized. The margins of these spaces

The right pericolic, left pericolic, and left infracolic spaces have pathways into the pelvis. On the left side, the mesentery of the sigmoid provides a surface on which neoplastic or inflammatory processes may center (see Fig. 49-2A and B). In addition the broad ligaments of the uterus create a surface on which processes become involved. From an anatomic standpoint, the pelvis can be subdivided into an anterior portion, which has certain fossae common to both men and women, and a posterior portion, which has some differences between men and women. The anterior divisions are less consistently visualized, depending on individual anatomy (see Figs. 49-3H and 49-5). The anterior pelvis is divided into several smaller fossae by peritoneal ligaments over certain vestigial remnants. These include the supravesical fossa in the midline, bounded laterally by reflections over the closed umbilical arteries. The medial umbilical fold, which reflects over the vestigial remnant of the urachus, is in the midline of this space. Immediately lateral to these reflections, the medial inguinal fossae can be seen on the right and left; they are bounded laterally by the reflection over the inferior epigastric arteries. Lateral to these are the lateral inguinal fossae, which are lateral to the inferior epigastric arteries. These spaces have little clinical significance except to distinguish between direct inguinal hernias (which are lateral  to the epigastric artery) and indirect inguinal hernias (which are medial to it). The posterior pelvic space is divided into areas determined by gender. In women there is a space between the uterus and bladder called the vesicouterine space. The rectouterine space, also known as Douglas’s pouch or the pelvic cul-de-sac, is between the uterus and rectum. The spaces on each side of the rectum are called the right and left pararectal spaces. Some have also designated an ovarian fossa adjacent to the ovaries.

CHAPTER 49  Peritoneum

1569

Transversalis fascia Peritoneum Ant. renal fascia Ascending colon

Duodenum

K

A

Pancreas

K

Skin

Lateroconal fascia Descending colon

Liver Transversalis fascia

Post. renal fascia

Post. renal fascia

Transversalis fascia Peritoneum Perirenal space Ant. perirenal space Ant. renal fascia P M

B

K

K

Duodenum Post. peritoneum Ant. renal fascia

Cecum

Skin Descending colon

Post. renal fascia

C

Post. pararenal space

FIG 49-6  Retroperitoneal spaces based on the traditional concept of retroperitoneal anatomy. Anterior (dashed lines) and posterior (dotted lines) perirenal fascia join to form lateroconal fascia. These fasciae divide the retroperitoneum into (1) the anterior pararenal space, which contains the duodenum, pancreas, and right and left colon; (2) the perirenal space, which contains the kidney and adrenal gland; and (3) the posterior pararenal space, which contains fat. A, Axial section of the pancreas. B, Axial section of the left upper quadrant just caudal to the pancreatic tail. The anterior pararenal space is subtended ventrally by the posterior peritoneum and dorsolaterally by the anterior renal fascia and lateroconal fascia. The posterior pararenal space is demarcated ventrally by the posterior perirenal fascia and lateroconal fascia and dorsolaterally by transversalis fascia. C, Parasagittal section through the right kidney. The bare area of the liver is in continuity with the anterior pararenal space. The posterior pararenal space communicates with the anterior pararenal space in the iliac fossa, below the perirenal space. (See Fig. 49-1D to F for an explanation.) K, kidney; PM, psoas muscle. (B, From Dodds WJ, et al: The retroperitoneal spaces revisited: A pictorial essay. AJR Am J Roentgenol 147:1155, 1986. ©American Roentgen Ray Society, 1986.)

Men have fewer spaces in the pelvis: the rectovesicular space and the right and left pararectal spaces. There is no space associated with the prostate, which is below the bladder. In women the ovarian fossae and vesicouterine space vary according to the size, configuration, and angulation of the uterus. The right and left pararectal spaces may also vary in size. In some patients the right side may predominate, with no space on the left. In other patients the pararectal spaces may extend posteriorly around the rectum and may be extremely low, almost to the level of the coccyx.

Retroperitoneum Traditionally the retroperitoneum is divided into different compartments depending on the level. At the level of the kidney, the space behind the peritoneum can be divided into the anterior pararenal, perirenal, and posterior pararenal spaces (see Figs. 49-1D and E, 49-4, and 49-6). The division of this space is based on the perirenal fascia (Fig. 49-7). Gerota’s fascia is the posterior portion, and Zuckerkandl’s is the anterior one. The anterior pararenal space is bordered anteriorly by the peritoneum on retroperitoneum, posteriorly by the anterior perirenal fascia, and laterally by the lateroconal fascia. The perirenal space is the space around the kidney bordered anteriorly and posteri-

orly by the renal fascia. The posterior pararenal space is bordered anteriorly by the posterior renal fascia and posteriorly by the fascia over the psoas muscle, which is continuous with the transversalis fascia. Below the level of the kidneys, there is a single retroperitoneal space with direct contiguity between the anterior and posterior portions. The cephalic and caudal relationships of the spaces are also important. The perirenal space is shaped like an inverted cone (open at the bottom and closed at the top; see Figs. 49-2D, 49-4A, and 49-7A-C). The cone is attached superiorly to the diaphragm, whereas the open end of the cone is open to the pelvis. Both the anterior and posterior pararenal spaces extend into the pelvis and cephalic to the level of the diaphragm (see Fig. 49-7C and D). On the right side, the upper extent of this space is the “bare area” of the liver; on the left, it is the bare area of the spleen (see Figs. 49-2 to 49-4). With certain pathologic processes, the mediastinum can communicate with either of these areas.

Peritoneal Reflections or Ligaments The various “ligaments” of the peritoneum are actually peritoneal reflections over tissue containing areolar tissue, arteries, veins, lymphatics, vessels, and lymph nodes. The margins of the ligaments are

1570

PART II  CT and MR Imaging of the Whole Body

H

B

A

H H

C

D

FIG 49-7  Renal fascia after hemorrhage. A, The renal fascia can be well visualized when thickening occurs. This patient has a perinephric hematoma (H), with fluid in the anterior and posterior fasciae. The fluid (arrow) crosses the midline into the opposite space. B, The highest portion of the retroperitoneum at the level of the attachment to the undersurface of the diaphragm. The lower arrow shows the thickened diaphragm, likely due to engorged lymphatics (see Fig. 49-18). The upper arrow shows the thick upper fascia. C, Coronal image of the renal fascia, lateral and medial portions (arrows). H, hematoma. D, Sagittal appearance of the anterior and posterior fasciae (middle horizontal arrows), the upper attachment by the diaphragm (upper horizontal arrow), and the lower open end of the fascia (vertical arrow). H, hematoma.

not definable except when the edges of the peritoneum are seen because of the presence of fluid or contrast material. Understanding these reflections is important because they serve as boundaries to the peritoneal spaces and also as possible conduits for the contiguous spread of inflammation or neoplasm.

Coronary and Triangular Ligaments.  The coronary and triangular ligaments represent attachments between the liver and the parietal peritoneum over the abdominal wall. The coronary ligaments, inferior and superior, are contiguous with the capsule of the liver and attach the right lobe of the liver to the posterior abdominal wall (see

CHAPTER 49  Peritoneum Figs. 49-2D, 49-3A and B, and 49-4A and B). The right and left triangular ligaments (see Fig. 49-4D) are formed by reflections of the inferior and superior coronary ligaments and attach the right and left lobes, respectively, to the diaphragm. Although they are typically called suspensory ligaments, it is not clear what mechanical support they really provide. The triangular ligaments are virtually never seen on CT scans unless there is air in the peritoneum, which highlights their margins. Although typically one thinks of the peritoneum and its spaces as completely constant, congenital variations may occur. Authors have noted considerable variation in the size and position of the triangular ligaments. Both sets of ligaments have little involvement with intraabdominal processes except that they provide boundaries for the right subphrenic and subhepatic spaces. When a retroperitoneal process occurs on the right side, it may extend into the bare area of the liver behind the right coronary ligaments or cross from one side to the other.

Falciform Ligament.  The falciform ligament is a vestigial remnant that represents a peritoneal reflection (see Figs. 49-2A and 49-3A) over the closed umbilical vein, which remains as the ligamentum teres. In inflammatory disease of the pancreas, inflammation can extend along the fascial tissues around the porta hepatis, around the ligamentum teres, and to the abdominal wall. In patients with portal hypertension, the ligamentum teres may recanalize, resulting in a patent umbilical vein that courses to the abdominal wall. In the presence of tumors, such as those of the ovary, stomach, colon, or pancreas, the falciform ligament may be a site of peritoneal seeding or extension. The normal ligament is not seen on CT unless free air or fluid within the abdomen provides better visualization.

Gastrohepatic, Gastrosplenic (Gastrolienal), and Gastrocolic Ligaments.  These three ligaments are in direct contiguity with the margins of the stomach. They provide the peritoneal boundaries of the omental bursa in conjunction with the peritoneal surfaces of the hilum of the liver, the posterior wall of the stomach, and the parietal peritoneum over the retroperitoneum. Because they contain the arteries, veins, lymphatics, vessels, and lymph nodes, they are conduits for the contiguous spread of disease. A neoplasm beginning in the stomach can spread through any of these pathways into the lymph nodes along corresponding vessels (Fig. 49-8). Spread may occur along the left gastric artery and coronary vein in the gastrohepatic ligament, along the celiac trunk and base of the hepatic artery in the hepatoduodenal ligament, along the gastroepi-

1571

ploic and short gastric vessels in the gastrosplenic ligament, and along the gastroepiploic vessels in the gastrocolic ligament. Tumor involvement in the gastrohepatic ligament produces an impression or impingement on the lesser curvature of the stomach. Involvement of the gastroduodenal ligament produces impingement on the stomach antrum. Gastrosplenic involvement shows pressure on the greater curvature of the stomach, and gastrocolic involvement produces changes on the upper portion of the colon. Conversely, processes that begin in the colon can spread to the stomach by these pathways. Processes from the duodenum or pancreas can spread along the hepatoduodenal ligament. Extension within these various ligaments can also occur in some cases of pancreatitis or hemorrhage. The fluid may dissect within the ligaments and may even extend to the subcapsular areas.

Phrenicocolic Ligament and Mesocolon.  The mesocolon, which is attached to the anterior surface of the pancreas, has a lateral attachment on the left side called the phrenicocolic ligament. The mesocolon contains the various colic vessels and lymph nodes and can serve as a pathway for disease. Spread of inflammation or neoplasms from the colon to the pancreas is possible by either of these paths.

Splenorenal (Lienorenal) Ligament.  The splenorenal ligament is the connection between the spleen and kidney. It contains the tail of the pancreas, the splenic vessels, and small vessels to the retroperitoneum. They are of little significance except as a landmark to distinguish between peritoneal and pleural fluid. Tumor can spread and systemic portal collaterals can develop in this space.

Ileocolic Ligament.  Mesentery at the ileocolic region may serve as a site for tumor implantation, although its occurrence is infrequent (Fig. 49-9).

Sigmoid Mesentery.  The mesentery of the sigmoid colon in the left lower quadrant acts as a barrier or a margin along which processes can collect. At times it serves as a medial or caudal boundary for the spread of material down the left pericolic space (Fig. 49-10).

Broad Ligaments and Cul-de-sac.  The broad ligaments and culde-sac represent two low peritoneal reflections that border the lower portions of the cavity. Tumors within the peritoneal cavity can seed to either location; most commonly, ovary, colon, and gastric tumors involve these areas.

FIG 49-8  Scan of a patient with a single metastatic deposit in the

FIG 49-9  Patient with carcinoma of the pancreas has a metastatic

gastrosplenic ligament (arrow).

implant (arrow) adjacent to the ileocecal valve by the cecum.

1572

PART II  CT and MR Imaging of the Whole Body

FIG 49-10  Two metastatic implants (arrows) on the right and left sides of the sigmoid mesentery.

Mesentery The small bowel mesentery containing the superior mesenteric vessels extends from the left upper quadrant at the ligament of Treitz to the right lower quadrant at the ileocecal valve (see Figs. 49-2C, 49-3D, and 49-4C). It serves as a dividing point between the right and left infracolic spaces. Specific areas of the mesentery should be noted because they manifest changes with pathologic processes. These areas are the neurovascular bundles, the root of the mesentery, and the margin of the small bowel. Normally the neurovascular bundles are thin and barely perceptible as thin lines within the fat-density mesentery (see Figs. 49-2C and 49-3D). The mesentery as a whole exhibits fat density and normally extends high into the abdomen, permitting small bowel loops to appose themselves closely to the anterior abdominal peritoneum.74 The margins of the fat on the inner surface of the small bowel are typically poorly defined because the course of the bowel is irregular and overlapping. When pathologic processes occur, the bowel dilates, the wall thickens, and peristaltic motion lessens. Retroperitoneal, mesenteric, and peritoneal processes can produce changes along the mesentery.86

Greater Omentum The greater omentum originates along the margin of the greater curvature of the stomach and can cover a broad expanse of the anterior abdominal wall. The function and role of the greater omentum have not been defined, but removal does not produce any adverse effects in peritoneal function. The normal omentum is usually imperceptible on routine scans. It is visualized only when fluid is present (see Figs. 49-3D and E and 49-4D) or it is pathologically involved with metastatic disease (discussed later in this chapter). Two signs that can be helpful are the density of the omentum and the location of small bowel loops. The normal omentum exhibits fat density and is not seen; when it contains infectious processes or neoplasms, it increases in density and becomes thicker and can produce a mass effect on the small bowel loops.

PERITONEAL PHYSIOLOGY Peritoneal Function The function of the peritoneum can be separated into fluid dynamics and particle clearance. The dynamic movement of normal fluid in the peritoneum is illustrated in Figure 49-11. Fluid produced on the serosal surface of the bowel moves inferiorly toward the pelvis because of gravity (see Fig. 49-11A). The active flow of fluid toward the dia-

phragm is due to respiratory motion of the diaphragm, gravitational pull on the liver, and normal peristalsis of the bowel, which enhance the flow of fluid cephalad. Normal peristalsis displaces fluid to the lateral gutters, where upward flow occurs. Respiratory movement of the diaphragm creates a negative subphrenic pressure that enhances upward flow, and gravitational pull on the liver creates a negative pressure above the liver that also enhances upward flow. The sum of these influences is generally an upward movement of peritoneal fluid toward the diaphragm (Figs. 49-12 and 49-13). Anything that alters the normal pattern of respiratory motion or affects bowel motility inhibits this normal flow. An ileus created by laparotomy or some other problem delays the clearance of the peritoneum. Positive-pressure ventilation, increased intraabdominal pressure, or phrenic neurectomy alters the physiologic processes and movement and thereby impairs normal clearance. Gravity also plays an important role in the flow of fluid and particulate material.3 The in vivo effects of gravity on particle distribution were studied by analyzing the clearance of contrast material injected into peritoneal cavities after routine appendectomy and cholecystectomy. Material injected into the ileocecal region goes preferentially to the pelvis, right pericolic area, and right subhepatic area. A smaller amount passes to the left pericolic gutter and left subhepatic space. Material injected into the right subhepatic space next to the duodenum spreads to the right subphrenic space, left infrahepatic space, right pericolic gutter, and pelvis (see Figs. 49-12 and 49-13). The infracolic spaces were not involved in either case, even when contrast material was injected into the ileocolic region, probably because the net flow of fluid is from the bowel loop surfaces laterally. Fluid flow occurs by way of the peritoneal membrane, a semipermeable membrane that permits bidirectional diffusion of water and solutes. Fluid and solute exchange is related to membrane area and permeability. The permeability and transfer of material depend on blood flow, vasoactive compounds, and tonicity of the peritoneal fluid (or osmotic pressure of blood).47,90 Hyperosmotic fluid, inflammation, chemical irritants (e.g., bile, gastric acid, pancreatic fluid), or vascular agents can cause a net flow of water as high as 300 to 500 mL/hour into the cavity.33,44 Although passive exchange of fluids and solutes occurs over the entire surface, active movement of particulate material (e.g., bacteria, cellular metastases) occurs through the peritoneum or under the diaphragmatic surfaces. There are stomata or lacunae between the mesothelial cells of this specialized peritoneal surface through which particulate material and fluid can be absorbed, draining through specialized lymphatic vessels. Negrini and associates studied the distribution of lymphatics in the pleural and peritoneal surfaces of the tendinous and muscular portions of the diaphragm.61 Using microdissection techniques, they noted that the lacunae ran either parallel or perpendicular to the fibrous or muscular fibers; lacunae were more frequent in the tendinous peritoneal portion and seldom seen in the muscular areas. Measures of lacunar size showed them to be 108 to 137.9 µm. Stomata accommodate particles smaller than 20 µm (demonstrated using polystyrene beads); size varies slightly with the tension within the diaphragmatic muscles during respiratory movement (Fig. 49-14). The material is transported through the diaphragm and into the retrosternal and anterior mediastinal lymphatic vessels to the right thoracic duct; reverse flow is prevented by lymphatic valves (Fig. 49-15). Material that exceeds these dimensions (e.g., fibrin-coated tumor cells) adheres to the surface of the peritoneum (see Fig. 49-14). Work by Shinohara and colleagues showed that stomata exist on the peritoneal side but not the pleural side, explaining why patients with ascites may develop pleural effusions but not vice versa (see Fig. 49-15).73 Obliteration of the main lymphatic trunks in animal models

CHAPTER 49  Peritoneum

1573

Peritoneal injury

Mesothelial cells

Mesothelial cells

Release of histamine, serotonin, etc.

Thromboplastin

Mesothelial cells Blocking

Thrombin

Activators

Prothrombin

Plasmin

Plasminogen

Increased permeability of mesothelial venules

Exudation of proteinrich material, including fibrinogen

Fibrin-split products

Fibrin Fibrinous adhesions

A

Fibrous adhesions

B

Peritoneal contamination

Peritoneal cell injury

Diaphragmatic motion

Mesothelial injury

Mast cell degranulation

Histamine and other vasoactive substances

Absorption via diaphragmatic lymphatics

Systemic RES

PMN influx Opsonization

Fibrin Localization (abscess)

C

Chemotaxis

Vascular permeability Plasma exudation

Complement activation

Phagocytosis

Bacterial killing

FIG 49-11  Peritoneal dynamics and physiology. A, Directions of fluid flow and clearance within the peritoneum. The relative size of the arrows indicates the relative contribution from each area. B, Pathogenesis of peritoneal adhesions. C, Algorithm of bacterial interaction with the immune systems, and their clearance mechanisms. PMN, polymorphonuclear neutrophils; RES, reticuloendothelial system.

1574

PART II  CT and MR Imaging of the Whole Body

FA

C

A

A FA

S

B B

FA

FIG 49-12  Dynamic fluid flow in the peritoneum. A, Radiograph taken after the injection of contrast material through a T-tube tract (C), which was displaced from the common bile duct and into the subhepatic recess. The edge of the right lobe of the liver (large arrows) is seen. The falciform ligament (curved black arrow) is surrounded by contrast material. Contrast material flows to the left side into the anterior portion of the left subdiaphragmatic space (arrowheads) adjacent to the stomach. Some material has entered the omental bursa (small vertical arrows) through Winslow’s foramen. (The patient is rotated because of scoliosis.) B, In this later radiograph, contrast material is better seen in the subhepatic recess (Morison’s pouch) (arrowheads). It flowed cephalad into the right subdiaphragmatic space (large arrows). The impression of the kidney on the subhepatic recess is seen (curved arrow). (The space appears to be located more medially because of the patient’s rotation.) The left subdiaphragmatic space is also seen (open arrows).

C FIG 49-13  CT scans of barium leak show movement of fluid into perihepatic peritoneal spaces. A, Axial scan shows leakage of barium from a gastric ulcer (small arrow) into the peritoneum. Free air (FA) is visible in the anterior peritoneum. Free air at the edge of the anterior liver by the falciform is also noted (large arrow). High-density barium is “pulled into” the subphrenic space laterally (open arrow). Note the small amount of air between the gallbladder and liver. A small amount of barium is adjacent to air in the right subhepatic space. B, Higher scan shows free air (FA) anteriorly. A small amount of barium is noted adjacent to the falciform margin (large arrow). Barium is noted adjacent to the liver in the right subhepatic space (open arrow). There is also barium in the anterior right subhepatic space (S) by the gallbladder, as well as a small bubble of air (small arrow) in the posterior subhepatic space (Morison’s pouch) between the liver and the kidney. C, Highest scan shows free air (FA) anteriorly. Air is present in the hepatic portion of the lesser sac (omental bursa). Small arrow shows a small air bubble mixed in with barium at the lowest margin of the left subphrenic space. In this patient the posterior margin at the coronary ligaments is not as far posterior as usual.

CHAPTER 49  Peritoneum

FIG 49-14  Patient with early ovarian metastasis implanted on lacunae. Dynamic flow of fluid and particles occurs under the surface of the diaphragm. The negative pressure produced by the movement of the diaphragm up and down creates a small vacuum, pulling fluid upward (arrow shows fluid). Particulate material or tumor cells may enter the lymphatic system through the lacunae. When tumor cells are coated with fibrin and are larger than the lacunae, they adhere to the undersurface of the diaphragm (small arrow).

1575

permits ascites to enter the pleural space by passing through small collateral transmuscular lymphatics through the diaphragm; passage from the pleural space to the peritoneum does not occur. Lymph flow from the organs may converge to central lymph nodes (Figs. 49-16 and 49-17) and subsequently to the major draining thoracic duct or conversely to the retrosternal and posterior mediastinal trunks. Main trunks ultimately enter the great veins after being filtered through mediastinal lymph nodes (Fig. 49-18; also see Figs. 49-14 and 49-16). The original description by Gray in 1915 includes the sternal glands by the internal mammary vessels, intercostal glands, and diaphragmatic glands (see Fig. 49-18). The diaphragmatic lymph nodes receiving lymph through the diaphragm are located anteriorly behind the xiphoid, in the middle region by the pericardium adjacent to the vena cava, and posteriorly on the back of the crura, leading into the posterior mediastinum. Modern multidetector CT (MDCT) scans have sufficient resolution and reconstruction capabilities to demonstrate such nodes and channels. The primary significance of these structures is related to the spread of tumor from the peritoneum (see later and Fig. 49-16). Metabolic, infectious, or neoplastic processes in the abdominal peritoneum may produce lymph flow through the diaphragm to the supradiaphragmatic lymph nodes. These diaphragmatic lymph nodes represent “sentinel” nodes when enlarged. Ascites and tumor cells  may pass through the lymphatics and produce ipsilateral pleural effusion (see Fig. 49-18). Because of the unidirectional valves in lymph vessels, pleural effusion cannot pass into the peritoneum. These lymphatic pathways were graphically demonstrated in a patient who had chemoembolization of liver cancer (see Fig. 49-18). Careful scrutiny of these scans shows opacification in the subdiaphragmatic lymphatics, the parietal lymphatics in the lower pleura, and the posterior mediastinum.

Response to Infection

FIG 49-15  Patient with pancreatitis and sympathetic pleural effusion. The diaphragmatic lymphatics have one-way valves upward, so fluid passes cephalad through the diaphragm into the pleura (but not downward from the pleura to the peritoneum). Most of the lymphatics are in the muscular portion of the diaphragm, not in the central tendinous portion. Note the thickening of the left diaphragm (arrow) caused by lymphatic engorgement and pleural effusion from the movement of the fluid through the diaphragm.

There are two effective defense mechanisms against pyogenic infection in the peritoneum: particulate clearance through the lymph nodes47,57 (discussed earlier) and the typical immune response involving antibodies and white cells (see Fig. 49-11C). Soiling of the peritoneum with enteric material or bacteria causes the dynamic fluid flow to transport the material toward the diaphragm (see Figs. 49-12 and 49-13). Absorption of bacteria into the diaphragmatic lymphatic system permits their eradication by the systemic defense mechanisms of fixed tissue macrophages, reticuloendothelial cells, and polymorphonuclear leukocytes.28,29 Laboratory models have shown that bacteria begin to disappear from the peritoneum even before the influx of phagocytic cells. In such cases, bacteria have been isolated from mediastinal lymph channels within 6 minutes from the time of peritoneal injection. The attraction of phagocytes, which can lyse bacteria, into the peritoneum79 occurs by chemotaxis (see Fig. 49-11C). Chemotaxis occurs after the activation of various complement components.3 Killing of the bacteria with intracellular chemicals occurs after phagocytosis. Ingestion by phagocytes may occur unaided; however, opsonization by complement (C3) or immunoglobulin (Ig)G may be required with encapsulated organisms.82 Fibrin also serves as a local defense mechanism (see Fig. 49-11B), which is useful in one regard but detrimental in others. With peritoneal injury, histamine and other permeability factors are released, which may permit the leakage of a protein-rich material containing fibrinogen. This material, combined with tissue thromboplastin, causes the polymerization of fibrin. Normally fibrin can be split and removed as needed by the activation of plasminogen to plasmin; however, when injury to the mesothelium or peritonitis occurs, no fibrinolytic activity takes place.31,90

1576

PART II  CT and MR Imaging of the Whole Body

C

A

D B FIG 49-16  Calcified peritoneal metastases on the undersurface of diaphragm. A, This patient with ovarian cancer has small calcified implants on the undersurface of the diaphragm (arrows). Lesions were not found in the liver parenchyma. B, Coronal view of the same patient shows calcified implants on the undersurface of the diaphragm (arrows). Recall that the coronary ligaments attaching the liver to the abdominal wall are posterior (see normal anatomy in Figs. 49-2 to 49-4). Diagonal arrow shows a lymph node on the surface of the diaphragm adjacent to the pericardium. C, Axial scan of the same patient shows the lymph node filtering lymphatics that penetrate the diaphragm (arrow). Enlarged nodes are nonspecific but are an indicator of intraperitoneal disease, either inflammatory or neoplastic. As in other areas of the body, the probability of neoplasm is low unless the node measures 2 cm or greater. D, Dystrophic pleural calcifications should not be confused with peritoneal neoplasms on the undersurface of the diaphragm. This patient with previous pleurodesis shows calcification of the pleural surface (large arrow) posterior to the diaphragm, the top of which is marked (medium arrow). Note air in the lowest part of lung (small arrow) behind the calcification.

Although fibrin localizes the pathogens to a nearby area instead of permitting them to spread, it may impair the eradication of bacteria.31 Fibrin hinders the clearance of bacteria into the lymphatic system. Fibrin also protects bacteria from phagocytic activity and diffusion of antibiotics through the fibrin to the bacteria.29,31,58 With this knowledge, one can understand the impetus to develop fibrinolytic agents (e.g., urokinase, tissue plasminogen activator), as well as their effectiveness in improving abscess drainage. With the introduction of bacteria and deposition of fibrin, bacteria and phagocytes release enzymes that produce local tissue destruction. The outcome of this dynamic process may be sterilization of the site or formation of an abscess, depending on the effectiveness of the host’s

immune system and the growth and virulence of the bacteria. Once a large inoculum of bacteria is established, a low oxidation potential results29 and impairs phagocytosis and antibiotic activity. If the host immune response is effective, the site is sterilized; if it is not, an abscess will form, requiring medical treatment.

PATHOLOGIC CONDITIONS Ascites and Peritonitis Ascitic fluid occurs whenever there is active formation of peritoneal fluid or interference with its removal. It can be caused by anything that produces inflammation, venous obstruction, lymphatic obstruction, 

CHAPTER 49  Peritoneum

A

B FIG 49-17  A, Patient with an ovarian tumor has a normal left diaphragm with no evidence of thickening or pleural involvement. B, Same patient after tumor involvement of the diaphragm and pleura. Note implantation on the surface of the spleen, thickening of the diaphragm, and fluid-density tumor implants.

A

B

C FIG 49-18  MDCT of lymphatic pathways in and through the diaphragm in a patient with hepatoma embolized by selective angiography. These three images show the dynamic physiology of the diaphragmatic lymphatics and their communication with parietal lymphatics. A, High-density material is present in the embolized liver, with high-density lipiodol in the lymphatics of the diaphragm. The right side (large arrow) shows a wide area of lymphatics. The left side (small arrows) shows contrast material in the posterior mediastinal subpleural lymphatics and the lower pleural lymphatics. We postulate that owing to the edema of the infarcted liver, some of the contrast material passed through the posterior lymphatics of the liver to the diaphragm. B, This scan performed 2 weeks later shows “clearing” of contrast material from the lymphatics, documenting their nature. Note that the density of the liver site is less than in A, documenting some “clearance” of the material. The arrows show the low-density lymphatic no longer opacified. C, Sagittal scan shows contrast-filled lymphatic channels in the subpleural tissues posteriorly. The lower arrow shows the most inferior lymphatics of the parietal lymphatics, and the upper arrow shows lymphatics of the posterior mediastinum.

1577

1578

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 49-19  Contrast administration assists in detection of small amounts of ascites. A, Fluid-density ascites may be difficult to visualize against the margin of a fat-density liver (arrow). B, IV contrast material opacifies the liver sufficiently so that the liver edge (arrow) by the fluid can be better visualized.

or albumin deficiency, including liver failure, caval or portal vein obstruction, infection, and neoplasm. Ascitic fluid can be detected well with CT, and with certain diseases some specific ancillary findings  may suggest the origin of the ascitic fluid (Fig. 49-19). For example,  in patients with cirrhosis, there may be associated findings in the  liver. Similarly, with pancreatitis, there may be enlargement of the  gland and thickening of the perirenal fascia.4 If neoplasms of the liver, pancreas, lymph nodes, or ovaries are present, the tumor masses may be noted. Distribution and progression of ascitic fluid can be easily visualized with CT. In early ascites when the fluid is free moving and  small in amount, it localizes in the cul-de-sac or pericolic gutters  or the subhepatic space. Gravity causes the fluid to collect in the pelvis or the right subhepatic space. Negative pressure beneath the diaphragm “pulls” fluid beneath the diaphragm (Fig. 49-20), and  fluid flowing through the lymphatics produces a similar effect; as noted earlier, fluid can traverse the diaphragm and reach the pleural space. Signs indicating infection in fluid include the appearance of a pressure phenomenon, a thickened “wall,” or gas. Infected fluid loculates because of the presence of fibrin, which confines the margins of the fluid; as such, the contours of the fluid bulge, producing effacement of adjacent intestines or organs. A critical factor is the mobility of fluid (Fig. 49-21). If intraperitoneal fluid is localized to a single area, the patient can be placed in a decubitus position to see whether the fluid moves. Infected fluid does not move, whereas sterile fluid usually shifts (see Fig. 49-21B). However, in immunosuppressed patients, this is not a useful sign. Some authors

L

FIG 49-20  CT scan taken at the level of the subphrenic space shows the most posterior extent of the right subphrenic space (straight arrow) containing fluid. Its relationship to the adjacent bare area of the liver (L) containing the kidney (curved arrow) and the pleural space is well visualized.

CHAPTER 49  Peritoneum

1579

R

A

B

P

FIG 49-21  A, With the patient in supine position, fluid (arrow) adjacent to the small bowel suggests an abscess. B, When the patient is scanned in decubitus position, the fluid moves away, indicating that it is not loculated and is unlikely to be an abscess.

have attempted to characterize the density of ascitic fluid in disease processes,7,9,31 but these findings have been inconsistent. Although the finding of focal displacement or effacement of organs may suggest infection, this sign has limited utility with loculated ascites. Any active process—inflammatory, neoplastic, or other—can deposit fibrin at the margins of a space and thereby produce a fluid collection under pressure. Definitive confirmation depends on diagnostic aspiration. If the pyogenic peritonitis is produced by intestinal perforation or if there are gas-forming organisms, free gas can be seen within the peritoneum.34,49,79 The presence of free air in the other spaces suggests intestinal perforation (see Fig. 49-13).

Tuberculous and Granulomatous Peritonitis Tuberculous peritonitis reportedly occurs in 3.5% of patients with pulmonary tuberculosis.27,51 It is a rare disease that results from disseminated miliary tuberculosis of the bowel and peritoneum. It is thought to be caused by hematogenous or local spread from GI involvement (penetration of involved bowel or lymph nodes draining into the peritoneum).34,35 Classically it is divided into a “wet” form with large exudates and a “dry” form51,86,91 with large amounts of inflammation caused by fibrinous exudates and tubercles. The literature reports that soft tissue densities produce thickening of the mesentery, omentum, and bowel wall and enlarged lymph nodes in the mesentery and retroperitoneum (Fig 49-22). The findings are not specific, so unless tuberculosis is common, the entity cannot be diagnosed without a high index of clinical suspicion.

Sarcoidosis Sarcoidosis is a fairly common systemic disorder of noncaseating granulomas. Primary peritoneal involvement is rare but may occur secondarily from hepatic or cardiac involvement and appears as thickening of the involved peritoneum and lymph nodes (Fig. 49-23).

Complications of Peritoneal Dialysis Continuous ambulatory peritoneal dialysis (CAPD), in which dialysate is infused into the peritoneum via an indwelling catheter with the tip placed in the pelvis, has grown increasingly common in end-stage renal patients. With long-term use of CAPD, several complications can arise, from infectious entities such as bacterial and tuberculous peritonitis to noninfectious entities such as catheter entrapment (Fig. 49-24), dialysate leak (Fig. 49-25), hernias, and sclerosing encapsulating peritonitis (Fig. 49-26), the advanced stage of which can lead to a “cocoon”like encasement of small bowel80 (Fig 49-27). CT and MRI, including peritoneography, play a critical role in early detection and management of these conditions.

Hemorrhage Hemorrhage into the peritoneal spaces is characteristic because of its high density and distention of the peritoneal spaces. Typically when hemorrhage from an organ occurs, there is a high-density “sentinel clot” adjacent to the site of bleeding, thereby identifying its source (Figs. 49-28 and 49-29). This was initially reported by Orwig and Federle, who documented in numerous subsequent publications its usefulness as a sign to improve detection and localization of intraabdominal bleeding from all viscera.63 Blood released into the peritoneum is irritating, and its presence produces peritoneal fluid, so both fluid and clot may be present. In contrast to other areas of the body, where the breakdown of blood results in density changes over a period of days or weeks, density changes are much more rapid within the peritoneum. Because of the fibrinolytic activity of the peritoneum and the rapid influx  of fluid, density changes occur quickly over a matter of hours. As  lysis of the blood occurs, dual density levels may be observed,  which reflect the settling of the blood products dependently (see  Fig. 49-29).

1580

PART II  CT and MR Imaging of the Whole Body

A

FIG 49-23  Peritoneal sarcoidosis in a 33-year-old woman with abdominal distention, nausea, and vomiting. Contrast-enhanced axial CT scan shows ascites and mild but diffuse soft tissue thickening involving the mesentery, omentum (arrows), and parietal peritoneum. (From Pickhardt P, Bhalla S: Unusual non-neoplastic and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

A

B FIG 49-22  A, Tuberculous peritonitis in a 47-year-old man without evidence of thoracic disease. Contrast-enhanced axial CT scan shows the “dry” form of peritoneal tuberculosis, with diffuse fibrous thickening (arrows) of the peritoneal lining without ascites. B, Tuberculous peritonitis in a 39-year-old man with weight loss and fever. Contrastenhanced axial CT scan shows prominent low-attenuation mesenteric lymphadenopathy (arrows). (From Pickhardt P, Bhalla S: Unusual nonneoplastic and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

FIG 49-24  Axial T2-weighted MRI showing the low-signal-intensity CAPD catheter (arrow) with its tip entrapped in a small loculated collection of high-signal intensity dialysate (arrowhead). This finding is often associated with a history of peritonitis. (From Stuart S, et al: Complications of continuous ambulatory peritoneal dialysis. Radiographics 29:449, 2009).

CHAPTER 49  Peritoneum

FIG 49-25  Axial CT peritoneography image in a patient with end-stage renal failure secondary to autosomal dominant polycystic kidney disease, which demonstrates leakage of high-attenuating dialysate into the anterior abdominal wall. (From Stuart S, et al: Complications of continuous ambulatory peritoneal dialysis. Radiographics 29:450, 2009.)

1581

FIG 49-27  Axial CT image in a 47-year-old woman with end-stage renal failure, demonstrating multiple encased small bowel loops within a “cocoon,” a finding in advanced sclerosing encapsulating peritonitis as a complication of CAPD. This encasement may result in partial or complete bowel obstruction. (Courtesy Raj M. Paspulati, MD, Department of Radiology, University Hospitals Case Medical Center, Cleveland, Ohio.)

FIG 49-26  Axial CT image demonstrating advanced peritoneal calcifications in a 34-year-old woman with sclerosing encapsulating peritonitis as a complication of CAPD. (From Stuart S, et al: Complications of continuous ambulatory peritoneal dialysis. Radiographics 29:454, 2009.)

After hemorrhage, a localized collection occurs (Fig. 49-30A); this shrinks over time (see Fig. 49-30B) if it is sterile. In an acute situation, one cannot distinguish an infected hematoma from a sterile one except by needle aspiration for fluid culture.

FIG 49-28  Posttraumatic subphrenic blood in evolution. CT scan shows hematoma (arrow) around the margin of the spleen, resulting from traumatic injury. The high-density material indicates that the spleen is unstable, and angiographic embolization was performed to prevent catastrophic bleeding.

Abscess Intraabdominal abscesses are uniformly fatal if untreated and are usually curable if properly treated.1 CT has proved useful in the detection and treatment of this entity and has changed the therapeutic paradigm for abscesses; most abscesses are now treated by imageguided drainage rather than surgery. Before an abscess develops, inoculation with a bacterium or yeast occurs, and the body responds with antibodies and leukocytes to eliminate the infection. At this stage there is focal inflammation, phlegmon, or cellulitis. If the body defenses prevail and eliminate the organism,

the cellulitis clears30,31,75; if the organism prevails and the focus of infection continues, a collection of purulent material—an abscess—results (Fig. 49-31). Typically, removal of the fluid is required for healing to occur, because the hydrostatic pressure of an abscess and the low oxidation reduction potential of the contained enzymes prevent effective treatment by antibiotics alone.

Clinical Signs and Symptoms.  Signs and symptoms depend on the status of the patient’s immune system. Classically, patients with normal

1582

PART II  CT and MR Imaging of the Whole Body

A

FIG 49-29  Hematoma in the right subphrenic space shows separation between the low-density serum component anteriorly and the highdensity blood products (curved arrow) posteriorly. The posterior margin (vertical arrow) of the subphrenic space is noted adjacent to the bare area (horizontal arrow) of the liver.

B immune systems have a clinical presentation consisting of fever, tachycardia, leukocytosis, and symptoms localized to the area of the abscess. Localization of symptoms depends on the site of the abscess. Pain from the visceral peritoneum is carried by the splanchnic nerves and is poorly localized. When the parietal peritoneum is involved, the pain is more precisely localized. The typical imaging appearance of an abscess is that of a well-defined fluid collection producing effacement of any adjacent anatomy. Notably, our experience over the past 30 years differs from typical findings. In almost half of these patients undergoing percutaneous drainage, the classic signs of fever and leukocytosis have been absent. Because of the lack of correlation between clinical signs and imaging findings, we have a low threshold for recommending percutaneous needle aspiration of fluid collections for bacterial culture. Immunocompromised patients can be even more problematic because they lack the normal infection-fighting processes. Thus they are unlikely to exhibit fever, pain, leukocytosis, or discomfort on clinical examination. In addition to the classic causes of immunosuppression such as steroids, chemotherapy, and human immunodeficiency virus (HIV), less obvious causes include malnutrition, aging, and chronic illnesses. Garrison and colleagues found clinical signs in only 7% of 143 immunocompromised patients.15 Equally important, a poor immune response may alter the imaging appearance of an abscess; immunocompromised patients may not form a typical fluid collection because they cannot mount the normal immune response. In such cases the only practical way to assess the significance of a fluid collection is to sample it by needle aspiration.

FIG 49-30  A, Localized hematomas have a characteristic high density owing to internal hemoglobin product breakdown. A hematoma (arrow) such as this may or may not be infected, requiring needle sampling or follow-up scans. B, Uninfected hematomas such as this one (arrow) will shrink over time if sterile.

Diagnostic Imaging.  The diagnosis of focal inflammation or abscess on CT or MRI depends on the formation of a walled-off fluid collection, which is detectable on either modality. Positron emission tomography (PET) can also detect abscesses by virtue of the elevated glucose metabolism at the site of infection.

FIG 49-31  Large abscess (arrow) in the left lobe of the liver shows subtly decreased density, which represents a poorly liquefied inflammatory site. Administration of IV contrast material is critical to visualize such areas.

CHAPTER 49  Peritoneum Early focal infection prior to fluid formation can occasionally be visualized as focal low-density edema or the presence of air bubbles. In the kidney or liver an early focal infection produces a diffuse area of low attenuation. Focal interstitial nephritis may have the appearance of a subtle soft tissue mass, making it difficult to distinguish from tumor. If focal inflammation without fluid accumulation is detected in this early stage—even in serious disorders such as emphysematous pancreatitis—systemic antibiotics may eradicate the infection before an abscess forms (Fig. 49-32). One of the greatest advantages of CT is its ability to image all areas completely. Accurate localization is especially important in preoperative evaluation of an abscess to facilitate its drainage either surgically or percutaneously.17,21-24 With CT it is possible to determine the full extent of the inflammatory process and to detect multiple concurrent abscesses, which should always be a concern. The consensus in the literature is that CT is superior to other imaging methods for the accurate localization of abscesses before percutaneous drainage procedures and is thus the modality of choice.43 CT should be used as the initial diagnostic modality whenever percutaneous aspiration or drainage of an intraabdominal abscess is anticipated.

1583

On CT a fluid-density abscess has well-defined or irregular margins and may or may not contain septations in the cavity (Fig. 49-33). The density of the fluid may approximate that of water but can be higher, depending on the amount of debris, blood, or proteinaceous material. Other low-density fluid masses such as old hematomas, lobulated ascites, lymphoceles, necrotic tumors, mucinous tumors, and benign cysts may be mistaken for abscesses; aspiration may be required for diagnosis. Gas occurs in approximately one third of cases, appearing as a “black” density; it may be abundant, with air-fluid levels, or it may be scattered throughout as microbubbles. Bubbles usually indicate gasforming organisms (most commonly Escherichia coli, Enterobacter aerogenes, and Klebsiella and Clostridium spp.) and are common in infected aortic grafts, pyogenic phlebitis, emphysematous pyelonephritis, or emphysematous pancreatitis. An air-fluid level suggests a GI fistula (see Fig. 49-12). Once an abscess has been detected, a careful review of the anatomy is appropriate to seek a cause, and CT is often useful in this regard. If the patient has not undergone recent surgery, the most common causes are diverticulitis, Crohn’s disease, or perforated viscus. Entities such as appendicitis, diverticulitis, cholecystitis, and penetrating ulcers can

A

B

C

D FIG 49-32  Localized infection without abscess, such as emphysematous pancreatitis, may manifest by the presence of gas alone. This unusual case confirms that one cause of idiopathic emphysematous pancreatitis is pyogenic phlebitis of the inferior mesenteric vein that spreads cephalad into the pancreas and portal region. A, Air (arrow) in the inferior mesenteric vein is surrounded by inflammatory edema of the adjacent fat. B, Scan taken at a higher level shows “air” (lower arrow) at the convergence of the inferior mesenteric vein and the splenic vein. Note that the air and edema have spread to the adjacent inferior pancreas (upper arrow). C, Higher scan shows air in the pancreatic body (arrow). D, Clot and air are present in the portal vein (arrow). The infection completely resolved with antibiotics alone, and the patient did not develop an abscess.

1584

PART II  CT and MR Imaging of the Whole Body

A

B

A

D

C

FIG 49-33  Hepatic abscess eroding through the diaphragm and into the lung. A, The lowest portion of the abscess has two well-defined abscess cavities (arrows). B, Fluid-density abscess containing “air” (arrow). C, Axial scan through the lower lung shows an air-fluid level (middle arrow) within the enhanced lung. Upper arrow shows air in the top of the abscess; lower arrow shows the lung parenchyma below the abscess. D, Sagittal image of a fluid-density abscess (A) that extends through the diaphragm (straight arrows) and forms air-fluid levels (curved arrows) in the lung abscess (fluid levels are on the right side of the image, which was dependent when the patient was lying down).

sometimes be suspected or recognized on CT. With appendicitis the diagnosis is suggested by a fluid collection adjacent to the cecum or edema in the retroperitoneal fascia. With diverticulitis the air-filled diverticula adjacent to the abscess (which is of fluid density) can be recognized. Penetrating ulcers cannot be seen, but because of gas adjacent to the duodenum or in the omental bursa, the diagnosis should be suspected. In other cases, CT may not provide enough information and other radiologic procedures or investigations are needed. The MRI appearance of an abscess is that of a well-defined fluid collection with low signal intensity on T1 and high signal intensity on

T2. Diffuse inflammation by itself or adjacent to an abscess appears as diffuse edema with increased T2 signal intensity. The reported experience with MRI is limited. Numerous diagnostic caveats are appropriate. Both oral and intravenous (IV) contrast material should be considered when performing CT or MRI. Oral contrast may be helpful for subdiaphragmatic abscesses or interloop abscesses to distinguish bowel loops, but many patients are too ill to tolerate oral intake. IV contrast is critical in all cases to increase the density differential between normally vascularized organs and a fluid-density mass. In patients who cannot tolerate oral

CHAPTER 49  Peritoneum contrast, IV contrast is helpful because it produces intense enhancement of the bowel or intestinal wall, making the mucosal markings visible. Electronic cursor measurement of small gas bubbles may not show −1000 Hounsfield units (HU) because of the partial-volume effect; such bubbles may measure only −200 to −300 HU. The nature of such gas bubbles can easily be confirmed by adjusting the window center until the adjacent fat appears gray. Distinguishing a sterile collection from an infected one can be assisted by two approaches. Repeat scans in several days may show a reduction in size of the fluid collection, which suggests a sterile collection. Also, positional changes may cause sterile nonloculated fluid or gas to move (see Fig. 49-21). Careful scrutiny of the anatomy can distinguish small amounts of gas in the bowel, which may be confused with abscess (Fig. 49-34). There are no distinguishing features for the various pathogens such as bacteria, fungi, or parasites, but actinomycosis, tuberculosis, and echinococcosis have some unique features (see later sections in this chapter). On PET, abscess or inflammation appears as a site of increased fluorodeoxyglucose (18FDG) accumulation because of hypermetabolism (Fig. 49-35). A hypermetabolic inflammatory site has localized hyperemia and elevation of the glucose transporters (GLUT), which

1585

FIG 49-34  Small air collections in the bowel can be distinguished from abscess air by noting that the small air pockets (arrows) are trapped in the margins of the small bowel valvulae. Free air (curved arrow) is noted anteriorly in front of the liver.

A

B

C FIG 49-35  PET and CT scans show diverticulitis of the sigmoid colon. A, Nonenhanced CT scans in the axial, sagittal, and coronal views show stranding and edema adjacent to the colon (arrows). B, Corresponding FDG PET images in the axial, sagittal, and coronal views. The PET scans show increased uptake indistinguishable from neoplasm. 18FDG accumulates in both inflammation and neoplasm. C, Superimposed PET and CT scans in the same views.

1586

PART II  CT and MR Imaging of the Whole Body

d: 21.3 mm mean: 0.557 max: 0.781

d: 21.3 mm mean: 0.557 max: 0.781

d: 21.3 mm mean: 0.557 max: 0.781

A

B d: 21.3 mm mean: 0.525 max: 0.729

d: 21.3 mm mean: 0.525 max: 0.729

d: 21.3 mm mean: 0.525 max: 0.729

C FIG 49-36  This breast cancer patient had PET for suspected metastatic disease. A, PET scan shows high FDG uptake in the right upper lobe, suspected of being neoplasm. B, CT scan shows linear density in the right upper lobe. C, Repeat PET study after oral administration of nasal decongestant (a new method we devised to reduce the incidence of false-positive results) shows resolution of 18FDG uptake. 18

results in a “hot spot.” Although this is typically indistinguishable from a tumor, two approaches can be used to make the distinction. The more traditional method consists of delayed scans, which show a fading of the signal for infectious processes but not tumors. The second approach is oral administration of a decongestant, either pseudoephedrine or phenylephrine, at the same time as injection of 18FDG. Though not approved by the U.S. Food and Drug Administration (FDA), this approach was effective in distinguishing tumor from inflammation in four of nine false-positive cases (www.FDA.gov, The Belmont Report) (Fig. 49-36). The mechanism is likely similar to the decongestant effect in sinusitis. The maximally dilated vessels constrict, thereby reducing local inflammation and delivery of 18FDG to the site while having no such effect in tumor. Because of the remarkable accuracy of CT for detecting absces­ ses,2,8,16,23,40,41,45,87 data on other imaging methods are sparse, but some comparative information is available about ultrasonography and gallium-67 scanning. In an early report, CT correctly diagnosed 20 of 22 abscesses, gallium was correct in 6 of 9 abscesses, and ultrasonography was correct in 1 of 2 abscesses.23 Korobkin and colleagues reported that CT correctly diagnosed 9 of 9 abscesses, gallium 9 of 12, and ultrasonography 12 of 12.41 Levitt and coworkers found CT to be

accurate in identifying 6 of 6 abscesses, gallium in 9 of 9.45 Knochel and colleagues presented a larger number of cases comparing CT and ultrasonography, but they did not include gallium scanning.40 In their series, CT was the most accurate modality, with a sensitivity of 97.5% and a specificity of 95%; ultrasonography had a sensitivity of 82% and a specificity of 94.5%. An ongoing debate is whether the use of CT is justified when ultrasonography is available and only 10% to 13% less accurate. Proponents of ultrasonography argue that it is less expensive and does not use radiation. Advocates of CT point out that because abscesses are curable, the 10% increased accuracy is worth the additional expense and radiation exposure. Indium-tagged white cell studies may be useful on occasion (gallium has been abandoned) but only for questions of chronic infection. An acutely ill patient does not have the luxury of waiting hours or days for the results of an isotope test.52 One can formulate an approach to the diagnostic imaging of suspected abscesses on the basis of several principles. According to the medical decision-making guidelines set by McNeil’s group, falsenegative results must be minimized when dealing with curable diseases55; therefore significant costs and efforts are justifiable to find and treat such diseases. With this reasoning, one can rationalize using CT

CHAPTER 49  Peritoneum

FIG 49-37  Subphrenic abscess containing fluid and air. The posterior margin (arrow) of the space borders the bare area of the liver.

for all patients; however, ultrasonography is also a valuable tool for diagnosis or drainage. If access to CT is not immediately available, ultrasonography can be used. If an acutely ill patient has drains, wounds, or external appliances, CT is better suited immediately after surgery because it does not require a coupling agent on the skin.

Specific Anatomic Sites of Abscess Formation.  Abscesses in different anatomic areas have some unique clinical characteristics and appearances on CT images. In addition, each of the anatomic areas has the potential for false-positive findings associated with normal variants and other abnormalities, which should be sought to avoid an incorrect diagnosis of intraabdominal abscess. Subdiaphragmatic abscess.  Right subdiaphragmatic space abscesses can be caused by a variety of disorders because this area has open access from the right pericolic gutter and the right subhepatic recess. Virtually any inflammatory process in the abdomen can reach this area. The most common cause of an abscess, however, is appendicitis or a perforated viscus (e.g., gallbladder) or ulcer. The whole abscess may show fluid density, or it may contain gas (Figs. 49-37   and 49-38). The most important diagnostic point is that the peritoneal space ends at the coronary ligaments immediately lateral to the bare area of the liver. A pleural effusion should not be mistaken for a subdiaphragmatic abscess. If the fluid is behind the fat posterior to the bare area of the liver, it represents a pleural effusion and not a subdiaphragmatic abscess (see Figs. 49-2 and 49-3). Any fluid or gas within the fat of this bare area is in the superiormost portion of the retroperitoneal space. If both a pleural effusion and a subdiaphragmatic abscess exist, a decubitus view may permit the pleural fluid to move and confirm the diagnosis. In some unusual cases, a bolus injection of contrast material with dynamic scanning may be needed to visualize the diaphragm, which enhances because of its rich blood supply. Any fluid in the subdiaphragmatic space can resemble an abscess, whether it be bile, blood, necrotic tumor, or loculated ascites. Left subphrenic abscesses can be caused by a number of abnormalities (Fig. 49-39). The typical origin of the abscess is the stomach or colon secondary to a gastric ulcer, postoperative leaks, or a perforated neoplasm, left colon, or esophagus. In some cases of perforated esophagus, the inflammatory process involves the retroperitoneal space

1587

FIG 49-38  Three abscesses in different areas of the left subphrenic space. The most anterior portion of the space (curved arrow) is in front of the liver. An abscess is present in the gastrohepatic space (vertical arrow), sometimes known as the left subhepatic space. The posterior subphrenic space (horizontal arrow) is around the spleen.

rather than the subdiaphragmatic space; if a portion of the stomach is involved, however, the subdiaphragmatic space is also involved. Most abscesses can involve the entire diaphragmatic space, but they also may loculate in an anterior or posterior segment (see Fig. 49-38). Although it was traditionally thought that a process cannot extend from the left pericolic space or the right subphrenic space, experience has shown that depending on individual anatomic variations or the nature of the inflammatory process, the falciform ligament and phrenicocolic ligament may not serve as effective barriers to the spread of infection. To detect a small abscess in the left subdiaphragmatic space, one must administer oral contrast material to fill the stomach. At times a fluid-filled stomach may resemble an abscess and vice versa. The contrast material clearly defines the stomach and eliminates any question of the diagnosis. Left-sided pleural effusion should not be mistaken for a left subdiaphragmatic abscess. The region between the upper pole of the left kidney and the spleen, which contains the splenic pedicle and tail of the pancreas, represents the splenorenal ligament. This region, located behind the fat, is within the pleural space and not the subdiaphragmatic peritoneal space.25 Any fluid within this area is in the superiormost area of the retroperitoneum rather than in the left subdiaphragmatic space. Above the level of the splenorenal ligament, fluid can surround the spleen entirely and may become interposed between the diaphragm and the spleen (see pararenal space abscesses later). Any problems in distinguishing pleural effusions from subphrenic collections can be overcome by putting the patient in the decubitus position. Inverted diaphragms and pleural effusions can confound the diagnosis of abscess if one does not understand the anatomic relationships. When a pleural effusion is large and under pressure, it is easy to misdiagnose it as an abscess. It is often difficult to differentiate a subphrenic abscess from inversion of the diaphragm by a large pleural effusion. In the latter case, a large pleural effusion is seen adjacent to the diaphragm. On first glance, it is difficult to distinguish a pleural effusion from a subphrenic collection, but if one looks carefully, the fluid is behind the spleen and the bare area of the spleen. A decubitus view or sagittal reconstruction may clarify the nature of the inversion (Fig. 49-40).

1588

PART II  CT and MR Imaging of the Whole Body

S A

A

B

C

FIG 49-39  Left subphrenic abscess extension. A, The left subphrenic space around the spleen has variable appearances, depending on the level. An abscess (A) in the most superior portion of the space over the spleen is present against the diaphragm and displaces the spleen caudad. B, At this level the abscess fluid collection has extended along the medial and lateral portion of the spleen (S). C, At the midportion of the spleen, the posterior margin of the space is limited (arrow) by the lienorenal ligament (bare area of the spleen).

F

F

F

B

S

A FIG 49-40  Subphrenic abscess can be mimicked by an inverted diaphragm. A, Two fluid collections (F) below what appears to be the left diaphragm (arrow) but really represents inflammatory membranes. B, Sagittal image shows inversion of the diaphragm (curved arrows), which is depressing the spleen (S). The upper margin of the pleural collection has a curved surface (straight arrow).

CHAPTER 49  Peritoneum

FIG 49-41  Abscess in the posterior subhepatic space (Morison’s pouch) between the liver and kidney. On occasion, as here, the curved surfaces create a partial volume effect, which may make it difficult to distinguish the abscess from a renal or perirenal fluid collection.

Right subhepatic or hepatorenal abscess.  The right subhepatic space is located between the right lobe of the liver and the anterior surface of the kidney. It is divided into anterior and posterior (Morison’s pouch) compartments, a functional division based on causative factors (Fig. 49-41). The most common cause of abscess in the anterior portion is a perforated gallbladder, which borders immediately on  this area. The posterior portion of the space is a dependent area within the peritoneum that can collect fluid from the pericolic gutters or subphrenic space or by direct extension from the duodenum, liver, or bowel. Because this is such a common site for fluid collection, a distinction may have to be made between ascites and purulent material. In such cases it is best to place the patient in a decubitus position to let the fluid move out of the space. Ascites moves, whereas inflammatory collections (in immunocompetent patients) do not; this is a good criterion to mark the need for aspiration in some patients. Another uncommon problem is that the cephalic extent of this space can be high. When an axial scan is taken above the typical location of the space in such cases, it may give the spurious impression of an intrahepatic abscess. Left subhepatic abscess.  Abscesses in the left subhepatic space are caused by the same processes that cause subphrenic abscesses. They may spread from the right side, beneath the falciform ligament, or from other portions of the left subphrenic space (Fig. 49-42). Omental bursa abscess.  Abscesses in the omental bursa (lesser sac) may be secondary to a process extending through the foramen of Winslow, inflammation from the pancreas, penetrating ulcer, or tumor in the stomach or duodenum. The inferior recess of the omental bursa is anterior to the pancreas and is most commonly involved by penetrating ulcer or a pancreatic process. As noted previously the splenic recess of the omental bursa is located directly behind the fundus of the stomach (Fig. 49-43). It may be necessary to administer oral contrast material to see the posterior wall of the stomach and distinguish it from an abscess in the splenic recess of the omental bursa. A small amount of air may be the only sign of a perforated ulcer (Fig. 49-44). The medial superior recess extends cephalad adjacent to the caudate lobe and vena cava (Fig. 49-45).

1589

FIG 49-42  Abscess (arrow) in the left subhepatic (gastrohepatic) space behind the left lobe and adjacent to the stomach.

FIG 49-43  Abscess in a patient lying supine. CT scan shows a large abscess (arrow) in the splenic recess of the lesser sac, which is effacing the stomach anteriorly.

Air may remain in the omental bursa for several days after a surgical procedure and should not be mistaken for an abscess. Pericolic and interloop abscess.  Pericolic abscesses can result from virtually any inflammatory process within the abdomen (e.g., perforated colon or bowel, appendicitis, ruptured gallbladder) because the physiologic movement of fluid is into the pericolic space (see earlier discussion of peritoneal function). Such abscesses are usually associated with a subhepatic or subdiaphragmatic abscess. On CT scans a pericolic abscess appears as a mass of fluid density bordered on the lateral surface by the parietal peritoneum and medially by the serosa of the colon (Fig. 49-46). Such abscesses may spread into the right subhepatic or subdiaphragmatic space. Interloop abscesses can result from the same conditions causing any abscess but are more commonly associated with fistulas from the

1590

PART II  CT and MR Imaging of the Whole Body

FIG 49-44  CT scan shows a small collection of contrast medium and air in the lesser sac (arrow) from a small gastric perforation.

FIG 49-46  Pericolic abscesses (arrows) on the right and left sides of the peritoneum.

C

FIG 49-45  Fluid collection (arrow) in the superior recess of the lesser sac, around the caudate lobe (C) of the liver.

bowel. They can occur between isolated loops of bowel or within the infracolic spaces. In such cases, ample amounts of contrast material should be given to try to opacify the bowel; there must be sufficient contrast enhancement of the bowel wall and the contents to distinguish the abscess. When an abscess is present, bowel adjacent to the inflammation may be paralytic and may not carry the contrast. In such cases a fluidor gas-filled abscess can be distinguished from the normal bowel because the wall of the abscess is smooth and lacks the irregular contour of the bowel wall produced by typical haustra or valvulae. Moreover, the fluid in the bowel has a varied appearance with different densities, whereas an abscess is completely homogeneous. Pelvic abscess.  The pelvis, the lowest point in the abdomen and the area to which exudates gravitate, is one of the most common locations for intraabdominal abscesses. Abscesses in this area may result from a perforated viscus (Fig. 49-47), peritonitis, or trauma, 

but appendicitis and diverticulitis are the most common causes  (Fig. 49-48). On CT scans the abscess has a typical appearance; it is low in density, well marginated, and may or may not contain gas. It is important to administer IV contrast material, and on occasion it may be helpful to opacify the rectum and colon with rectal contrast. Posterior pararenal space abscess.  The posterior pararenal space is bounded posteriorly by the transverse fascia and anteriorly by the posterior renal fascia (Fig. 49-49A). This space extends caudad to the pelvis and cephalad to the diaphragm (see Fig. 49-49B). Although most clinicians are familiar with the posterior pararenal space in the area of the kidneys, the pararenal space in the area adjacent to the diaphragm is not as familiar. Inflammatory processes in the posterior pararenal space usually result from a ruptured aortic aneurysm, pancreatitis, osteomyelitis from the rib or spine, intestinal or esophageal perforation, or trauma. Abscesses may remain localized laterally within the pararenal space or become confluent with the psoas muscles. Processes from the aorta or spine may spread laterally. In addition to the abscess collection, there is concomitant thickening of the fascial planes. A psoas abscess produces enlargement of the psoas muscles.67 This abscess typically shows fluid density, but occasionally the density may be similar to that of muscle. In cases of bilateral psoas abscesses associated with destruction of the vertebral body, tuberculosis should be considered. Although little information is available in the literature, several cases indicate that the posterior pararenal space communicates with the posterior mediastinum. Esophageal perforations or other problems may extend caudad into the right or left posterior pararenal space or vice versa. Anterior pararenal space abscess.  The anterior pararenal space is bounded anteriorly by the posterior parietal peritoneum and posteriorly by the anterior renal fascia; this space is bounded laterally by the lateroconal fascia (which is the fused anterior and posterior perirenal fasciae). This compartment contains several alimentary organs, including the pancreas and ascending and descending colon. (This discussion is based on the traditional anatomic viewpoint; see “Dodds’ Concepts of Retroperitoneum and Peritoneum” earlier for a different view.)

CHAPTER 49  Peritoneum

A

1591

B FIG 49-47  A, CT scan of a child with vasculitis and local perforation of the bowel. A retroperitoneal abscess is seen in the right iliopsoas muscle (arrow). B, Abscess extends caudad through the psoas muscle and over the area of the lesser trochanter (arrow).

B

A

FIG 49-48  A, Large fluid- and air-filled abscess (arrow) adjacent to the sigmoid colon. B, Spontaneous resolution of the abscess (arrow), which drained into adjacent bowel, contains only air.

Most inflammatory processes in this area originate from the colon, appendix, pancreas (see Fig. 49-49A), or duodenum as a result of perforation from tumor, diverticulitis, inflammatory bowel disease, or trauma. Perforations of the bowel can produce gas, abscesses, or both, depending on the site of involvement. A localized perforation of the colon may occur in the retroperitoneal area and remain confined to the anterior pararenal space. If the process occurs below Gerota’s fascia, it can move superiorly into both the anterior and posterior pararenal spaces. The most cephalic extension of the pararenal spaces on the right is the retroperitoneum behind the bare area of the liver; on the left it is the retroperitoneum medial to the spleen. Another valuable sign in cases of inflammation is thickening of  the renal fascia. For example, in some cases it may be difficult to

distinguish a loop of bowel with the density of fluid from an abscess; with an abscess, the renal fascia is thickened.

Benign Peritoneal Abnormalities

Peritoneal Cyst.  Peritoneal cysts are unusual masses that may be attached to the mesentery or peritoneum. Their origin is poorly understood, but some believe that they are related to lymphatic abnormalities that produce sequestered secondary cysts. The cysts, which are typically lined with endothelium, may be unilocular or multilocular. Clinically they produce the same symptoms as an abdominal mass; surgical treatment is curative. On CT images these cysts appear as large low-density masses with clearly demarcated fibroconnective tissue walls (Fig. 49-50). No unique

1592

PART II  CT and MR Imaging of the Whole Body

AP

PP

A

B FIG 49-49  A, Retroperitoneal abscess in the anterior (AP) and posterior (PP) pararenal spaces. Fat (straight arrow) is seen within the renal fascia, bounded anteriorly by Gerota’s fascia and posteriorly by Zuckerkandl’s fascia. Slightly thickened fascia (curved arrow) can also be seen on the left. B, The superiormost extent of this retroperitoneal abscess (arrow) continues in the bare area behind the liver.

A

B FIG 49-50  Mesenteric cyst. A, CT scan shows a mass (arrow) that proved to be a mesenteric cyst. B, CT-guided procedure proved the true nature of the mass. The trajectory chosen avoids any loops of bowel, thereby preventing the possibility of contamination.

imaging features distinguish these cysts from other low-density masses such as mucinous metastases, abscesses, or hematomas. The clinical history is therefore important, and if indicated a percutaneous puncture for cytologic or bacteriologic study can be performed. The differential diagnosis of peritoneal cysts includes lymphan­ gioma, nonpancreatic pseudocysts, enteric cysts, duplication cysts,  and cystic mesotheliomas. According to Ros and coworkers, differentiation among these entities is not possible with current imaging techniques.69

Cystic Mesothelioma.  This rare benign tumor originates from the mesothelial cells and has a female predominance. Although it is not malignant, it may recur in 25% to 50% of cases. As one might anticipate, CT usually shows multiple cystic masses with thin walls that do not produce a mass effect on adjacent structures (Fig. 49-51). Early forms of pseudomyxoma peritonei may be difficult to differentiate except in cases with calcifications or mass effect.

Desmoid Fibromas (Aggressive Fibromatosis).  Desmoid fibromas are unencapsulated infiltrating fibromas occurring predominantly in women and in any part of the body, including the abdomen.48 There

is a high association between fibromas of the abdominal wall and peritoneum and familial polyposis syndrome,54 and fibromas are one of the classic signs of Gardner’s syndrome. In one series of 44 intraabdominal desmoid fibromas, 41 occurred within the mesentery.54 The main clinical problem is distinguishing these tumors from neoplastic masses resulting from colonic carcinoma. Pathologically, desmoid fibroma is benign but difficult to distinguish from low-grade fibrosarcoma; however, there have been no reported metastases. Occasionally the lesion may cavitate and enlarge quickly by imbibition of fluid. The treatment is surgical excision, but unless the resection is complete and meticulous, recurrence is the rule. On CT a desmoid fibroma appears as a mass of soft tissue density within the peritoneum, displacing other intraabdominal organs (Fig. 49-52). No unique imaging traits distinguish these fibromas from neoplastic masses such as lymphomas and metastatic disease or from other benign masses, hematomas, or high-density peritoneal cysts.

Inflammatory Pseudotumor.  Inflammatory pseudotumor or myofibroblastic tumor is a rare reactive lesion that may invade the  peritoneal cavity in early adulthood. The CT appearance may be infiltrative, with solid and necrotic areas (Fig. 49-53). Distinction from

CHAPTER 49  Peritoneum

1593

FIG 49-51  Cystic mesothelioma. CT scan in a young woman shows a large, peritoneum-based, multilocular cystic mass. Note the thin cyst walls (arrows) and lack of a prominent soft tissue component. (From Pickhardt P, Bhalla S: Unusual non-neoplastic and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

FIG 49-53  Inflammatory pseudotumor in a teenage boy. Contrastenhanced axial CT scan shows the infiltrative soft tissue appearance (arrows) of peritoneal inflammatory pseudotumor (myofibroblastic tumor). (From Pickhardt P, Bhalla S: Unusual non-neoplastic and subperitoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

FIG 49-52  Desmoid fibroma. Intraabdominal desmoid (fibromatosis) in a 36-year-old man with an increasing abdominal mass. Contrastenhanced axial CT scan shows a solitary mesenteric mass (arrow) with an irregular border. (From Pickhardt P, Bhalla S: Unusual non-neoplastic and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

fibrosarcoma is difficult on pathologic examination, but onset at a young age is typical.

Sclerosing Mesenteritis.  Sclerosing mesenteritis is nonspecific inflammation and fibrosis of mesenteric fatty tissue and can be categorized into retractile mesenteritis, mesenteric panniculitis, and mesenteric lipodystrophy based on the predominant process occurring in the tissue. Pathologically there is fibrosis, inflammation, and fatty infiltration. These entities typically present in the sixth to seventh decades of life, more commonly in men, with a predominantly self-limited course and favorable outcome.78 Retractile mesenteritis is a rare disorder in

which fibrosis predominates.11,39 The cause of the abnormality is unknown. The process originates in the root of the mesentery and can extend diffusely throughout the mesenteric fat. The disease process is confined to the mesentery, and there are no abnormalities of the intestines or vessels.39 On CT the mesentery is not of normal fat density but appears as an area of soft tissue density that can be masslike (Fig. 49-54A). The bowel is not in its normal anterior location but is retracted posteriorly. One group reported calcifications (see Fig. 49-54B) in some unusual cases of retractile mesenteritis and some resolution of the disease after treatment with steroids.18 In unusual cases, narrowing and rigidity of the colon with thumbprinting or displacement of the bowel by mesenteric masses may be seen.83 Han and coworkers reported a case that produced transmural thickening of the colon.26 Extensive metastatic disease to the peritoneum and mesentery may simulate the imaging findings of retractile mesenteritis (see Fig. 49-54C). In such cases, tumor implants in the mesentery produce increased density and may also restrict movement of the bowel, as seen with ascites. Mesenteric panniculitis, the acute form of mesenteric inflammation, can present clinically as acute abdomen. Imaging appearances vary, but a characteristic imaging finding—unlike tumors—is that vessels are spared, with a discrete fat margin known as the fat ring sign.78 Mesenteric lipodystrophy is characterized by fat necrosis as the predominant pathologic process.78

Omental Infarction and Epiploic Appendagitis.  In addition to mesenteric panniculitis, inflammatory changes of portions of the omentum and epiploic appendages are entities that may present

1594

PART II  CT and MR Imaging of the Whole Body

*

A

*

B

C FIG 49-54  A, A mass (curved arrows) with calcifications (large arrows) caused by retractile mesenteritis. B, Retractile mesenteritis in a 54-year-old man with a long history of abdominal pain. Contrast-enhanced axial CT scan shows a densely calcified spiculated mesenteric mass (asterisks). The CT appearance resembles that of metastatic carcinoid. C, A similar appearance can be produced by dystrophic calcifications in a treated lymphoma mass. The arrow points to thickening of the neurovascular bundle. (A, Courtesy Dr. Richard Palmer Gold, Columbia-Presbyterian Medical Center, New York. B, From Pickhardt P, Bhalla S: Unusual nonneoplastic and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

clinically as acute abdomen, and their proper diagnosis by imaging can prevent unnecessary surgery. Omental infarction can present with right lower quadrant pain, similar to appendicitis, and approximately 15% of cases occur in the pediatric population.20 Imaging characteristics include a large, solitary, well-circumscribed, nonenhancing, heterogeneous fatty omental mass, most commonly seen in the right lower quadrant between the abdominal wall and bowel (Fig. 49-55C).78 The nearby bowel wall is typically not thickened. Omental infarction is best differentiated from epiploic appendagitis by size (>5 cm in diameter), location in the right hemiabdomen adjacent to the cecum or ascending colon, and lack of enhancing rim.78 Normal fat-density colonic epiploic appendages are invisible among the other fat densities of the mesentery unless inflammation (most commonly due to torsion) is present. In such cases an edematous appendage can be easily seen on CT as an oval fat-density lesion along the colonic wall, measuring less than 5 cm in diameter, with an enhancing rim and surrounding edema (see Fig. 49-55A).77 The most common site is adjacent to the sigmoid colon, although they can be

found less frequently along the right hemicolon as well. The nearby bowel wall is typically not thickened. This condition most commonly occurs in middle-aged men and is typically self-limiting and treated conservatively.77

Lymphadenopathy.  A number of unusual disorders cause lymphadenopathy, including nonspecific adenitis, Whipple’s disease, celiac disease, amyloidosis, Castleman’s disease (angiofollicular lymph node hyperplasia, which is considered premalignant; Fig. 49-56), eosinophilic gastroenteritis, sarcoid, and others. These entities are discussed in the applicable chapters.

Benign Mesenchymal Tumors.  Benign mesenchymal tumors include hemangioma, lymphangioma, lipoma, nerve sheath tumor, gastrointestinal stromal tumor (GIST), and leiomyosarcomatosis peritonealis disseminata. Lymphangiomas may be congenital malformations or benign neoplasms of the lymphatic system. They may be large or small thin-walled cysts with visible vessels. They may be lower in density than water because of their chylous nature (Fig. 49-57).

CHAPTER 49  Peritoneum

A

1595

B

C FIG 49-55  A, Mesenteric panniculitis in a 28-year-old man with vague abdominal symptoms. Contrastenhanced CT scan shows mesenteric soft tissue infiltration (arrows) that is characteristically localized to a jejunal segment, spares the perivascular fat, and demonstrates a pseudocapsular appearance. B, Epiploic appendagitis in a 28-year-old woman with acute left lower quadrant pain. Contrast-enhanced axial CT scan shows soft tissue infiltration in and around a fatty appendage (arrow) extending off the sigmoid colon. C, Segmental omental infarction in a 42-year-old man with acute right lower quadrant pain. Axial contrastenhanced CT scan shows a localized masslike process (arrowhead), which represents an advanced example of right-sided omental infarction. (From Pickhardt P, Bhalla S: Unusual non-neoplastic and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

Hemangiomas may be cavernous, capillary, or venous and have a variable appearance. Having large blood-filled sinuses, they lack a welldefined border. The presence of phleboliths strongly suggests the diagnosis (Fig. 49-58). Lipomas arise from adipose tissue and exhibit a uniform lowdensity fat appearance; liposarcomas typically have higher-density soft tissue components except in unusual cases (Fig. 49-59). (See later under “Primary Mesenchymal Tumors.”) Leiomyomatosis peritonealis is a rare benign condition affecting women of reproductive age. It is characterized by subperitoneal nodules containing smooth muscle; ascites may be associated when lymphatic channels are infiltrated. Nerve sheath tumors are uncommon in the general population but occur in those with neurofibromatosis type 1 (NF1). Originating in

nerve sites, these tumors may bridge retroperitoneal and subperitoneal spaces, such as the celiac plexus area. Associated nerve root lesions typical of NF1 are helpful ancillary findings (Fig. 49-60).

Malignant Neoplasms

Peritoneal Mesothelioma.  Peritoneal mesothelioma is a rare neoplasm arising in the peritoneal lining of the abdomen. Thirty percent of all mesotheliomas are peritoneal, with the remainder being pleural. An association with asbestosis has been noted. Pathologically it occurs in several forms: sarcoma, polygonal cell type, and tubular papillary type. All types have a uniformly poor prognosis, and most patients die within 2 years of diagnosis. On CT examination (Fig. 49-61) the tumor appears as an area of soft tissue density within the abdomen.10 It may conform to the

1596

PART II  CT and MR Imaging of the Whole Body

FIG 49-58  Hemangioma. Contrast-enhanced CT scan shows a poorly circumscribed mesenteric soft tissue mass (arrow) with punctate calcifications that represent phleboliths. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983–995, 2005.)

FIG 49-56  Castleman’s disease in a 31-year-old man with a history of fever. Contrast-enhanced axial CT scan shows lymphadenopathy limited to the gastrosplenic region (arrow), which proved to represent Castleman’s disease. (From Pickhardt P, Bhalla S: Unusual non-neoplastic  and sub-peritoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

FIG 49-59  Lipoma. Contrast-enhanced CT scan shows a wellcircumscribed lesion composed of uniform fat attenuation (arrowheads). No soft tissue component is present. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983–995, 2005.)

contour of the liver or produce a mass effect on adjacent structures. No features clearly distinguish it from other types of benign or malignant tumors with a benign fibrotic reaction. A number of authors have recommended signs that suggest the diagnosis of mesothelioma.5,62,86,89 It may appear “dry,” with well-defined or confluent peritoneal masses. Calcification is uncommon.

Primary Serous Carcinoma.  Primary papillary tumor of the peri-

FIG 49-57  Mesenteric lymphangioma. Contrast-enhanced CT scan shows a large low-attenuation cystic lesion involving the small bowel mesentery. Note how vessels course through the lesion. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983–995, 2005.)

toneum is recognized as an unusual tumor of postmenopausal women. Initially it was thought to be related to either mesothelioma or ovarian tumors,37,66 but it is now generally thought to originate from mesothelial cells of the peritoneum. Because most patients present with extensive disease and symptoms, it is impossible to determine whether the origin is local or multifocal. Treatment is difficult, and the prognosis is poor.

CHAPTER 49  Peritoneum

1597

NF

FIG 49-62  Primary peritoneal papillary serous carcinoma. ContrastFIG 49-60  Multiple neurofibromas in NF1. Contrast-enhanced CT scan shows a prominent low-attenuation mass near the root of the mesentery that surrounds the superior mesenteric vessels. Although lymphadenopathy can have this appearance, note the presence of low-attenuation neurofibromas involving the paraspinal nerve roots (arrows). (From Pickhardt P, Bhalla S: Unusual non-neoplastic and subperitoneal conditions: CT findings. Radiographics 25:719–730, 2005.)

enhanced CT scan in a middle-aged woman shows a solid mass involving the right aspect of the omentum (arrow). Note the small amount of adjacent peritoneal fluid and prominent omental vessel. No ovarian mass was present. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983– 995, 2005.)

DSC

FIG 49-61  Malignant peritoneal mesothelioma. Contrast-enhanced CT scan shows relatively uniform and diffuse thickening of both the visceral and parietal (arrowheads) peritoneum associated with ascites. Omental infiltration is present (arrow), but no bulky masses or loculated fluid collections are seen. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983–995, 2005.)

CT findings of this disease are nonspecific. Scans may show a discrete tumor mass over the peritoneum or simply collections of loculated fluid (Fig. 49-62). Calcifications may occur, distinguishing it from mesothelioma.

Desmoplastic Small Round Cell Tumor.  These tumors differ from other peritoneal tumors in that they affect adolescents and young

FIG 49-63  Desmoplastic small cell tumor (DSC). Contrast-enhanced axial CT scan shows diffuse peritoneal thickening, including omental caking, which is an infrequent appearance of desmoplastic small round cell tumor. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983–995, 2005.)

adults. The appearance is similar to that of most peritoneal tumors, demonstrating peritoneal masses with variable enhancement and necrosis throughout, occasionally with a diffuse infiltrating appearance (Fig. 49-63).

Primary Mesenchymal Tumors Liposarcoma.  Liposarcoma has received considerable attention because of its striking appearance and the possibility of characterizing

1598

PART II  CT and MR Imaging of the Whole Body Other tumors.  Other mesenchymal tumors of the peritoneum and subperitoneum are diverse, being both benign and malignant; they have few distinguishing features. The malignant tumors include the malignant versions of liposarcoma, malignant fibrous histiocytoma, leiomyosarcoma (Fig. 49-65), nerve sheath tumor, GIST, fibrosarcoma, angiosarcoma, and synovial sarcoma.

the various cell types based on the histologic correlations reported in the literature. Kim and associates reported on a group that included well-differentiated, myxoid, pleomorphic, and round cell types.38 They found that the well-differentiated type had a similar appearance to lipoma and exhibited a low density similar to subcutaneous fat, with sclerosing components representing septae. The myxoid type had attenuation less than muscle on CT and signal intensity similar to water on MRI (hypointense on T1, hyperintense on T2). With contrast, the myxoid elements showed increased reticular enhancement. The pleomorphic and round cell types had density and attenuation similar to muscle on CT and MRI T1-weighted images and similar to fat on MRI T2-weighted images. These authors cautioned, however, that many tumors have mixed histologic cell types, and the classification is usually based on the most aggressive type. Kransdorf and colleagues reported on features that suggest malignancy to distinguish lipoma from well-differentiated liposarcoma; these include older patient age, large lesion size, presence of thick septa, and presence of nodular or nonadipose-like areas with a decreased percentage of fat composition.42 Einarsdottir and coworkers stated that in general, tumors that are predominantly fat are probably ordinary lipomas; lesions that are less fatty may be lipomas or liposarcomas.14 Suzuki and associates reported that PET showed significantly higher standardized uptake values for malignant lesions compared with benign ones.81 Welldifferentiated liposarcoma can show a fat density that is indistinguishable from normal fat, whereas poorly differentiated tumors are slightly higher in density (Fig. 49-64).

Metastatic Disease Metastatic involvement of the omentum and mesentery is common with neoplasms originating in the ovary, stomach, pancreas, and colon.6,36 Spread within the peritoneum can be along the surface of the peritoneum or within the ligaments (areolar tissue surrounded by peritoneum) that contain the lymph nodes draining a specific area. Lymphatic spread can occur along the neurovascular muscular bundles within the intermesenteric space. When fibrosis is associated with the lesions, there may be retraction in the mesentery. When metastatic lesions are on the peritoneal surface, several findings may be noted. First, if they are large enough, they may be seen as discrete nodular lesions (Fig. 49-66). If the tumor is of mucinous origin (e.g., ovary, colon) it may show soft tissue or fluid density. When the size of the lesion is below the resolving power of CT, collections of fluid indistinguishable from any other fluid may be seen. The metastatic sites of an ovarian tumor may be completely fluid in appearance.56 In such cases the tumors may resemble subcapsular collections or may implant on specific ligaments (Fig. 49-67) such as the falciform ligament (see Fig. 49-70A), gastrosplenic ligament (see Fig. 49-8),

S

A

S

B

C FIG 49-64  Liposarcoma varies in density, depending on the degree of differentiation. Mesenchymal tumors may have a variable appearance, but liposarcomas are somewhat typical, as can be seen in the following scans. A, Well-differentiated liposarcoma (S) is low in density, as is normal mesenteric fat. B, Well-defined higher-density tumor (S) has an intermediate differentiation. C, Poorly differentiated liposarcoma with myxoid changes; it is producing a pseudomyxomatous invasion of the peritoneum. The loops of the small bowel (curved arrow) are compressed centrally by the low-density tumor, which appears almost like fluid. The bowel does not float, as with normal ascites. The contrast-filled small bowel is compressed posteriorly on the left.

CHAPTER 49  Peritoneum reflection of the terminal ileum (see Fig. 49-9), broad ligaments, or sigmoid colon (see Fig. 49-10). Regardless of size, mucinous or other treated tumors can produce small calcifications throughout the peritoneum.64 Mitchell and colleagues reported a series of 15 patients with carcinoma of the ovary and found calcifications in six peritoneal sites of implantation, perihepatic calcifications in five sites, and lymph node calcifications in one.60 Making the diagnosis of peritoneal metastasis may be difficult if the mass is not big enough to cause displacement of the bowel and if total opacification of the bowel has not been achieved.84 The normal bowel, however, has a characteristic appearance that shows small amounts of fluid and gas in a diffuse pattern. Sometimes even after administration

1599

of contrast material, it may be difficult to distinguish an adynamic loop of bowel from a mass. In such cases, IV administration of contrast material may be helpful because the wall of the bowel enhances well. Also, placing the patient in the decubitus position permits moving “gas” to reveal the true nature of a loop of bowel.

Omental Changes.  The thickened omentum produced by “caking” appears as a large soft tissue mass with poorly defined edges (see  Fig. 49-66B). The fat plane between the anterior abdominal wall and the intestinal loops may be obscured. When the omentum (which exhibits fat density) is not thick, soft tissue masses within it or on it may be seen.

Sites of Implants.  The falciform ligament (see Fig. 49-67A), gastrohepatic ligament (see Fig. 49-8), reflection of the sigmoid colon (see Fig. 49-10), ileocolic region (see Fig. 49-9), posterior and dependent sites of the peritoneum (see Fig. 49-67C and D), and broad ligaments are common sites of malignant implantation.76 Data have shown that biopsy of the diaphragmatic areas is the best method for early detection of peritoneal seeding (at least with ovarian neoplasms).19 This is presumably a result of the circulation of peritoneal fluid to the undersurface of the diaphragm and the lacunae (see Fig. 49-12).

General Diagnostic Signs.  Whitley evaluated a group of 370

FIG 49-65  Leiomyosarcoma. Contrast-enhanced axial CT scan shows a large peritoneum-based soft tissue mass with heterogeneous enhancement (arrow). (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and subperitoneal origin: CT findings. Radiographics 25:983–995, 2005.)

A

patients with mesenteric tumors and suggested some diagnostic signs based on the lesion’s appearance: rounded, ill defined, cakelike, or stellate.86 Rounded masses (most common) were caused by non-Hodgkin’s disease in 33 of 41 cases, leukemia in 2 of 26 cases, and ovarian tumor in 3 of 38 cases. Whitley also noted a 30% occurrence of metastases in the nodes of the small bowel mesentery.86 Other authors have noted an overall occurrence of 50% if nodes at the root of the mesentery are included. Ill-defined masses, the next most common type, occurred in 27 patients. Of this group, non-Hodgkin’s and ovarian carcinomas were most common, followed by colon, pancreas, stomach, and other masses. Cakelike masses of the mesentery occurred with ovarian tumors, non-Hodgkin’s lymphoma, and leukemia; calcifications can occur primarily in ovarian tumors (Fig. 49-68) or secondarily after

B FIG 49-66  A, Small tumor implants (arrows) are noted under the anterior surface of the left diaphragm. Without fluid loculation, they are more difficult to appreciate. B, More extensive tumor implantation produces a thick soft tissue density (arrows), displacing the colon from the anterior peritoneum.

1600

PART II  CT and MR Imaging of the Whole Body

A B

C

D FIG 49-67  Patient with metastatic ovarian cancer foci in the upper abdomen. Because of their mucinous nature, the lesions appear as “fluid” density. A, Low-density mass adjacent to the falciform ligament (arrow). B, Low-density metastasis on the surface of the liver below the diaphragm (arrow). C, Mass in the hilum of the spleen (arrow). D, Low-density metastasis in Morison’s pouch, or the right subhepatic space between the right kidney and the liver (arrow).

treatment of other tumors (see Fig. 49-54C). Stellate infiltration of the neurovascular bundles of the mesentery occurred with all types of metastatic disease except lymphoma and leukemia. The stellate pattern is produced by infiltration along the inner margins of the mesentery. Goldhirsch and colleagues found that CT was not as accurate as second-look surgical exploration in detecting metastases.19 A positive CT scan is a useful guide for the surgeon, but a negative study does not obviate the need for a second-look surgical procedure.

Pseudomyxoma Peritonei.  Pseudomyxoma peritonei is a rare disease resulting from diffuse dissemination of mucinous material

derived from neoplasms.65 The histologic characteristics resemble mucinous adenocarcinoma or adenomucinosis. Typically the neoplasms producing the peritoneal seeding are mucinous cystadenocarcinomas of either the ovary or appendix, but tumors of the urachus, uterus, or omphalomesenteric duct are less common causes. The largest series was by Yan and colleagues, who reported 46 cases, with the great majority resulting from a missed appendiceal tumor.88 The disease is indolent, and massive deposition of material may occur before it is noticed. On CT there seem to be two possible appearances50,53,56,72,85: it may appear as discrete low-density cystic masses located in multiple anatomic spaces (Fig. 49-69), or the peritoneum

CHAPTER 49  Peritoneum

1601

may be completely filled and resemble ascitic fluid, sometimes producing mass effects on adjacent organs. Calcifications may occur following chemotherapy.59 Figure 49-69 shows a localized fluid density in the area of the lesser sac in the early stages; the unusual characteristic at that time was the “immobility” of the fluid density in variable body positions. Ultimately the fluid-density mass insinuated itself throughout the peritoneum.

PET IMAGING OF PERITONEAL PROCESSES

FIG 49-68  Psammomatous calcifications in metastatic ovarian implants. (From Pickhardt P, Bhalla S: Primary neoplasms of peritoneal and sub-peritoneal origin: CT findings. Radiographics 25:983–995, 2005.)

As one would expect, the use of FDG PET has not been extensively studied for peritoneal tumors and processes because of their infrequency. It is obvious that the hypermetabolism of cancer will be avid for 18FDG and therefore is likely to produce positive results. One of the more impressive uses of such scans has been to demonstrate diffuse metastatic disease in the peritoneum (Figs. 49-70 and 49-71), which shows excellent correlation with CT.

A B

C

D FIG 49-69  Pseudomyxoma peritonei. A, Scan taken early in the course of myxomatous deposits shows fluid-density material in the lesser sac by the foramen of Winslow (arrow). B, Lower scan at the same time shows fluid-density myxomatous material (arrow) in the right lower quadrant adjacent to the prior location of the appendix. The tumor origin was mucinous adenocarcinoma of the appendix, which had been resected. C, Late scan showing fluid-density myxomatous tumor in the right lower quadrant, surrounding the bowel. D, With the patient lying on the left side, one can see that the myxomatous material does not shift dependently as fluid would.

1602

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 49-70  CT PET imaging in combination has been helpful in the detection of some peritoneal tumors. A, CT scan shows a small round mass on the surface of the bowel, which is difficult to distinguish from bowel without contrast material. B, PET image shows “hot” 18FDG uptake over the lesion. C, CT PET image shows fusion of the “red” hot spot over the anatomy. Abnormal lymph nodes are also shown.

A

C

B

E

D

FIG 49-71  Five views from CT PET. A, Poorly defined streaking in the omentum anteriorly. B, Diffuse uptake in the anterior abdomen. C, Composite image shows a superimposed “red” hot spot on the CT image, representing omental caking. D, Hot areas anteriorly over the entire peritoneum. E, Diffuse uptake in the anterior view.

REFERENCES 1. Altemeier WA, et al: Intra-abdominal abscesses. Am J Surg 125:70–79, 1973. 2. Aronberg DJ, et al: Evaluation of abdominal abscess with computed tomography. J Comput Assist Tomogr 2:384–387, 1978. 3. Autio V: The spread of intraperitoneal infection: Studies with roentgen contrast medium. Acta Chir Scand Suppl 36:1–31, 1964. 4. Balthazar EJ, et al: Acute pancreatitis: Prognostic value of CT. Radiology 156:767–772, 1985. 5. Banner MP, Gohel VK: Peritoneal mesothelioma. Radiology 129:637–640, 1978. 6. Bernardino ME, Jing BS, Wallace S: Computed tomography diagnosis of mesenteric masses. AJR Am J Roentgenol 132:33–36, 1979. 7. Bydder GM, Kreel L: Attenuation values of fluid collections within the abdomen. J Comput Assist Tomogr 4:145–150, 1980.

8. Callen PW: Computed tomographic evaluation of abdominal and pelvic abscesses. Radiology 131:171–175, 1979. 9. Callen PW, Marks WM, Filly RA: Computed tomography and ultrasonography in the evaluation of the retroperitoneum in   patients with malignant ascites. J Comput Assist Tomogr 3:581–584, 1979. 10. Dach J, et al: Peritoneal mesothelioma: CT, sonography, and gallium-67 scan. AJR Am J Roentgenol 135:614–616, 1980. 11. Day DL, Sane S, Dehner LP: Inflammatory pseudotumor of the mesentery and small intestine. Pediatr Radiol 16:210–215, 1986. 12. Dodds WJ, et al: Anatomy and imaging of the lesser peritoneal sac. AJR Am J Roentgenol 144:567–575, 1985. 13. Dodds WJ, et al: The retroperitoneal spaces revisited. AJR Am J Roentgenol 147:1155–1161, 1986. 14. Einarsdottir H, et al: MR imaging of lipoma and liposarcoma. Acta Radiol 40:64–68, 1999.

CHAPTER 49  Peritoneum 15. Garrison RN, Fry DE, Spain D: Surgical infections. In Lawrence P, Bell RM, Dayton MT, et al, editors: Essentials of Surgery, Philadelphia, 2005, Lippincott Williams & Wilkins, pp 1162–1180. 16. Gerzof SG, Robbins AH, Birkett DH: Computed tomography in the diagnosis and management of abdominal abscesses. Gastrointest Radiol 3:287–294, 1978. 17. Gerzof SG, et al: Percutaneous catheter drainage of abdominal abscesses guided by ultrasound and computed tomography. AJR Am J Roentgenol 133:1–8, 1979. 18. Gold R: Personal communication, 1983. 19. Goldhirsch A, et al: Computed tomography prior to second-look operation in advanced ovarian cancer. Obstet Gynecol 62:630–634, 1983. 20. Grattan-Smith JD, et al: Omental infarction in pediatric patients: sonographic and CT findings. AJR Am J Roentgenol 178(6):1537–1539, 2002. 21. Haaga JR, Alfidi RJ: Precise biopsy localization by computer tomography. Radiology 118:603–607, 1976. 22. Haaga JR, Weinstein AJ: CT-guided percutaneous aspiration and drainage of abscesses. AJR Am J Roentgenol 135:1187–1194, 1980. 23. Haaga JR, et al: CT detection and aspiration of abdominal abscesses.   AJR Am J Roentgenol 128:465–474, 1977. 24. Haaga JR, et al: Interventional CT scanning. Radiol Clin North Am 15:449–456, 1977. 25. Halvorsen RA, et al: Ascites or pleural effusion? CT differentiation: Four useful criteria. Radiographics 6:135–149, 1986. 26. Han SY, et al: Retractile mesenteritis involving the colon: Pathologic and radiologic correlation (case report). AJR Am J Roentgenol 147:268–270, 1986. 27. Hanson RD, Hunter TB: Tuberculous peritonitis: CT appearance. AJR Am J Roentgenol 144:931–932, 1985. 28. Hau T, Ahrenholz DH, Simmons RL: Secondary bacterial peritonitis: The biologic basis of treatment. Curr Probl Surg 16:1–65, 1979. 29. Hau T, Hoffman R, Simmons RL: Mechanisms of the adjuvant effect of hemoglobin in experimental peritonitis. I. In vivo inhibition of peritoneal leukocytosis. Surgery 83:223–229, 1978. 30. Hau T, Nishikawa RA, Phuangsab A: The effect of bacterial trapping by fibrin on the efficacy of systemic antibiotics in experimental peritonitis. Surg Gynecol Obstet 157:252–256, 1983. 31. Hau T, Payne WD, Simmons RL: Fibrinolytic activity of the peritoneum during experimental peritonitis. Surg Gynecol Obstet 148:415–418, 1979. 32. Henderson LW: The problem of peritoneal membrane area and permeability. Kidney Int 3:409–410, 1973. 33. Henderson LW, Nolph KD: Altered permeability of the peritoneal membrane after using hypertonic peritoneal dialysis fluid. J Clin Invest 48:992–1001, 1969. 34. Hulnick DH, et al: Computed tomography in the evaluation of diverticulitis. Radiology 152:491–495, 1984. 35. Hulnick DH, et al: Abdominal tuberculosis: CT evaluation. Radiology 157:199–204, 1985. 36. Jeffrey RB, Jr: CT demonstration of peritoneal implants. AJR Am J Roentgenol 135:323–326, 1980. 37. Kannerstein M, et al: Papillary tumors of the peritoneum in women: Mesothelioma or papillary carcinoma. Am J Obstet Gynecol 127:306–314, 1977. 38. Kim T, et al: CT and MR imaging of abdominal liposarcoma. AJR Am J Roentgenol 166:829–833, 1996. 39. Kipfer RE, Moertel CG, Dahlin DC: Mesenteric lipodystrophy. Ann Intern Med 80:582–588, 1974. 40. Knochel JQ, et al: Diagnosis of abdominal abscesses with computed tomography, ultrasound, and 111In leukocyte scans. Radiology 137:425– 432, 1980. 41. Korobkin M, et al: Comparison of computed tomography, ultrasonography, and gallium-67 scanning in the evaluation of suspected abdominal abscess. Radiology 129:89–93, 1978. 42. Kransdorf MJ, et al: Imaging of fatty tumors: Distinction of lipoma and well-differentiated liposarcoma. Radiology 224:99–104, 2002. 43. Lang EK, et al: Abdominal abscess drainage under radiologic guidance: Causes of failure. Radiology 159:329–336, 1986.

1603

44. Leak L, Just EE: Permeability of peritoneal mesothelium: A TEM and SEM study. J Cell Biol 70:423, 1976. 45. Levitt RG, et al: Computed tomography and 67 Ga citrate radionuclide imaging for evaluating suspected abdominal abscess. AJR Am J Roentgenol 132:529–534, 1979. 46. Lieberman JM, Haaga JR: Duodenal malrotation. J Comput Assist Tomogr 6:1019–1020, 1982. 47. MacCallum W: On the mechanism of adsorption of granular materials from the peritoneum. Bull Johns Hopkins Hosp 14:105, 1903. 48. Magid D, et al: Desmoid tumors in Gardner syndrome: Use of computed tomography. AJR Am J Roentgenol 142:1141–1145, 1984. 49. Marks W: Computed tomographic demonstration of intraarterial air following hepatic artery ligation. Radiology 135:665, 1979. 50. Masaryk TJ, Chilcote WA: CT of pseudomyxoma peritonei: Case report. Comput Radiol 8:43–47, 1984. 51. Mathieu D, et al: Periportal tuberculous adenitis: CT features. Radiology 161:713–715, 1986. 52. Matthews AW, et al: The use of combined ultrasonic and isotope scanning in the diagnosis of amoebic liver disease. Gut 14:50–53, 1973. 53. Mayes B, Chuang V, Fisher R: CT of pseudomyxoma peritonei. AJR Am J Roentgenol 136:807, 1981. 54. McAdam WA, Goligher JC: The occurrence of desmoids in patients with familial polyposis coli. Br J Surg 57:618–631, 1970. 55. McNeil BJ, Keller E, Adelstein SJ: Primer on certain elements of medical decision making. N Engl J Med 293:211–215, 1975. 56. Megibow AJ, et al: Ovarian metastases: Computed tomographic appearances. Radiology 156:161–164, 1985. 57. Mendez G, Jr, Isikoff MB, Hill MC: Retroperitoneal processes involving the psoas demonstrated by computed tomography. J Comput Assist Tomogr 4:78–82, 1980. 58. Miles AA, Miles EM, Burke J: The value and duration of defence reactions of the skin to the primary lodgement of bacteria. Br J Exp Pathol 38:79–96, 1957. 59. Miller DL, Udelsman R, Sugarbaker PH: Calcification of pseudomyxoma peritonei following intraperitoneal chemotherapy: CT demonstration.   J Comput Assist Tomogr 9:1123–1124, 1985. 60. Mitchell DG, et al: Serous carcinoma of the ovary: CT identification of metastatic calcified implants. Radiology 158:649–652, 1986. 61. Negrini D, et al: Distribution of diaphragmatic lymphatic lacunae. J Appl Physiol 72:1166–1172, 1992. 62. O’Neil JD, et al: Cystic mesothelioma of the peritoneum. Radiology 170:333–337, 1989. 63. Orwig D, Federle MP: Localized clotted blood as evidence of visceral trauma on CT: The sentinel clot sign. AJR Am J Roentgenol 153:747–749, 1989. 64. Pandolfo I, et al: Calcified peritoneal metastases from papillary cystadenocarcinoma of the ovary: CT features. J Comput Assist Tomogr 10:545–546, 1986. 65. Pestieau SR, Esquivel J, Sugarbaker PH: Pleural extension of mucinous tumor in patients with pseudomyxoma peritonei syndrome. Ann Surg Oncol 7:199–203, 2000. 66. Pickhardt PJ, et al: Desmoplastic small round cell tumor of the abdomen: Radiologic-histopathologic correlation. Radiology 210:633– 638, 1999. 67. Ralls PW, et al: CT of inflammatory disease of the psoas muscle. AJR Am J Roentgenol 134:767–770, 1980. 68. Raptopoulos V, et al: Renal fascial pathway: Posterior extension of pancreatic effusions within the anterior pararenal space. Radiology 158:367–374, 1986. 69. Ros PR, et al: Mesenteric and omental cysts: Histologic classification with imaging correlation. Radiology 164:327–332, 1987. 70. Roub LW, et al: Computed tomographic positive contrast peritoneography. Radiology 131:699–704, 1979. 71. Rubenstein WA, et al: Posterior peritoneal recesses: Assessment using CT. Radiology 156:461–468, 1985. 72. Seshol M, Coulam C: Pseudomyxoma peritonei: Computed tomography and sonography. AJR Am J Roentgenol 136:803, 1981.

1604

PART II  CT and MR Imaging of the Whole Body

73. Shinohara H, et al: The distribution and morphology of lymphatic vessels on the peritoneal surface of the adult human diaphragm, as revealed by an ink-absorption method. Okajimas Folia Anat Jpn 79:175–183, 2003. 74. Silverman PM, et al: Computed tomography of the normal mesentery. AJR Am J Roentgenol 143:953–957, 1984. 75. Silverman PM, et al: CT appearance of diffuse mesenteric edema.   J Comput Assist Tomogr 10:67–70, 1986. 76. Silverman PM, et al: Computed tomography of mesenteric disease. Radiographics 7:309–320, 1987. 77. Singh AK, et al: CT of acute appendagitis. AJR Am J Roentgenol 183:1303–1307, 2004. 78. Singh AK, et al: Epiploic appendagitis and its mimics. Radiographics 25:1521–1534, 2005. 79. Steinberg B: Infections of the Peritoneum, New York, 1944, Paul Hoeber. 80. Stuart S, et al: Complications of continuous ambulatory peritoneal dialysis. Radiographics 29:441–460, 2009. 81. Suzuki R, et al: PET evaluation of fatty tumors in the extremity: Possibility of using the standardized uptake value (SUV) to differentiate benign tumors from liposarcoma. Ann Nucl Med 19:661–670, 2005. 82. Talley F: Determinants of virulence in anaerobic bacteria. Microbiology 34:219, 1976.

83. Thompson G, Fitzgerald ER, Somers SS: Retractile mesenteritis of the sigmoid colon. Radiology 157:565, 1985. 84. Warde P, et al: Computed tomography in advanced ovarian cancer: Inter- and intraobserver reliability. Invest Radiol 21:31–33, 1986. 85. Weigert F, Lindner P, Rohde U: Computed tomography and magnetic resonance of pseudomyxoma peritonei. J Comput Assist Tomogr 9:1120–1122, 1985. 86. Whitley N: Computed Tomography of the Gastrointestinal Tract: Including the Peritoneal Cavity and Mesentery, New York, 1986, Springer-Verlag. 87. Wolverson MK, et al: CT as a primary diagnostic method in evaluating intraabdominal abscess. AJR Am J Roentgenol 133:1089–1095, 1979. 88. Yan H, et al: Histopathologic analysis in 46 patients with pseudomyxoma peritonei syndrome: Failure versus success with a second-look operation. Mod Pathol 14:164–171, 2001. 89. Yeh HC, Chahinian AP: Ultrasonography and computed tomography of peritoneal mesothelioma. Radiology 135:705–712, 1980. 90. Zinsser HH, Pryde AW: Experimental study of physical factors, including fibrin formation, influencing the spread of fluids and small particles within and from the peritoneal cavity of the dog. Ann Surg 136:818–827, 1952. 91. Zirinsky K, et al: Computed tomography, sonography, and MR imaging of abdominal tuberculosis. J Comput Assist Tomogr 9:961–963, 1985.

50  Gastrointestinal Tract Hyun Kwon Ha, Seong Ho Park, Seung Soo Lee, and Ah Young Kim

ESOPHAGUS NORMAL ANATOMY The esophagus extends from the pharynx to the cardia of the stomach. It is 20 to 25 cm in length and distends up to 30 mm wide but is collapsed at rest.633 It may be divided into three parts: 1. The cervical segment is located near the midline and is posterior to the trachea. 2. The thoracic segment is the longest and lies behind the trachea until the trachea bifurcates. 3. The abdominal segment is the shortest segment; it is only a few centimeters in length, escapes the diaphragmatic hiatus, and runs to the cardia of the stomach.

IMAGING TECHNIQUES On computed tomography (CT) the esophagus appears as a welldelineated circle of soft tissue (Fig. 50-1). Although the wall thickness of the collapsed esophagus has no absolute normal size, a welldistended esophagus should not exceed 3 mm in wall thickness,718,772 and eccentric thickening of the wall should suggest a pathologic condition.303 An anteroposterior diameter of the esophagus larger than 16 mm and a lateral diameter larger than 24 mm are considered abnormal, although they may increase with advancing age (Fig. 50-2). A small amount of air in the esophagus is visible in approximately 65% of normal individuals, but an air-fluid level, fluid-filled lumen, or lumen caliber larger than 10 mm usually indicates obstruction or severe esophageal dysmotility.566,599 Progress in multidetector CT (MDCT) technology and postprocessing software has expanded clinical application of CT in the evaluation of various esophageal diseases. Virtual endoscopy technique has been recently introduced and tested for its potential usefulness in the esophagus.419,422,558 Adequate esophageal distention for MDCT is achieved by either oral effervescent agents or air insufflation via catheter inserted into the upper esophagus.419,422,558,642 Multiplanar reconstruction (MPR) of CT data may permit visualization of the anatomic relationship of the esophageal lesion and adjacent structure more clearly.642 Real-time endoluminal images are also readily available and permit clear visualization of small endoluminal lesions.558 Initial studies have demonstrated the potential value of this technique in evaluating esophageal neoplasms (Fig. 50-3) and varices.419,422,558,642 With magnetic resonance imaging (MRI) the normal esophagus is routinely well visualized in the upper thoracic and gastroesophageal regions (Fig. 50-4), whereas the middle portion of the esophagus is flattened behind the left atrium, sometimes making its visualization difficult.566 The MRI appearance of normal esophageal wall is a

low-signal-intensity structure, well defined by the high signal intensity of periesophageal fat on T1-weighted images. The normal esophagus also appears with low signal intensity contrasted with high signal intensity of internal contents on T2-weighting. After administration of intravenous (IV) gadolinium diethylenetriamine pentaacetic acid (GdDTPA), the muscle layer of the esophagus shows moderate enhancement.696,774 Application of phased-array surface coils can provide high-resolution MRI of the esophagus and visualize individual layers of esophageal wall. On high-resolution T2-weighted images, the esophagus is seen as intermediate-signal-intensity mucosa surrounded by high-signal-intensity submucosa and low-signal-intensity muscularis propria.732-734

PATHOLOGIC CONDITIONS Benign Esophageal Tumors Benign tumors of the esophagus are not a common condition, constituting about 20% of esophageal neoplasms.511 Depending on the site of origin, benign esophageal tumors comprise mucosal and submucosal lesions. Because small lesions cannot be demonstrated on CT or MRI, barium study appears to be superior to cross-sectional image in providing information about them.

Leiomyoma.  Leiomyoma is the most common benign esophageal tumor, accounting for more than 50% of all benign esophageal tumors.781 It is characterized histologically by proliferation of smooth muscle and is most frequently located in the distal third of the esophagus (60%), followed by the middle third (30%) and proximal third (10%).781 Most patients are asymptomatic, but some may present with dysphagia, depending on the size of the lesion and how much it encroaches on the lumen. Although most esophageal leiomyomas occur as a solitary lesion, multiple lesions (leiomyomatosis) are present in 3% to 4% of patients; the lesion is often associated with Alport’s syndrome in children.262,703 Occasionally leiomyoma arising near the gastroesophageal junction (GEJ) may involve not only the distal esophagus but also the gastric cardia and fundus.778 In this location these tumors may cause severe dysphagia due to obstruction of the distal esophagus, and it may be difficult to distinguish them from carcinoma of the cardia involving the distal esophagus. On CT a leiomyoma appears as an intraluminal mass or as eccentric or concentric esophageal wall thickening, which usually shows homogeneous soft tissue attenuation (Fig. 50-5) and in rare instances may contain irregular calcifications (Fig. 50-6). The peripheral margins of the lesion are usually smooth, and the surrounding mediastinal fat planes are intact.552,569 Sometimes CT appearances of leiomyoma are nonspecific and indistinguishable from those of a localized malignant

1605

1606

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-1  Normal esophagus. The cervical (A) and thoracic (B) segments (arrows) of normal esophagus appear as well-defined circular structures of soft tissue on this CT scan.

tumor. In leiomyomatosis, CT reveals marked circumferential or eccentric wall thickening of the distal esophageal wall, with soft tissue attenuation. In addition, CT may occasionally show a mass bulging into the gastric fundus or uncharacteristic achalasia associated with distal esophageal wall thickening.495,703

Fibrovascular Polyp.  Fibrovascular polyp of the esophagus is a rare

FIG 50-2  Normal esophagus of a 77-year-old man. The lumen of the esophagus (E) is somewhat dilated; however, in older people, mild airfilled dilatation of the esophagus is regarded as a normal condition.

benign intramural tumor composed of loose or dense fibrous tissue, adipose tissue, and a vascular structure. The polyp is covered by normal squamous epithelium. It usually develops from the upper third of  the esophagus near the level of the cricopharyngeus. This lesion probably originates from the loose submucosal tissue of the cervical esophagus, gradually elongating over a period of years as the polyps are dragged into the middle or distal third of the esophagus by esophageal peristalsis until the intraluminal portion of the lesion becomes massive.24,43,112,493,881 These polyps usually occur in older men. Common symptoms include long-standing dysphagia, vomiting, weight loss, and respiratory difficulty. Occasionally, fibrovascular polyps with long stalks can be regurgitated into the pharynx or mouth, resulting in sudden death from laryngeal occlusion.24,112 Small lesions can be removed endoscopically, but larger masses should be excised surgically because of the potential for hemorrhage. CT shows a wide spectrum of attenuation from fat to soft tissue, depending on the amount of adipose and fibrovascular tissue in the polyp.493 If adipose tissue content predominates, an intraluminal esophageal mass with fat attenuation is considered a typical CT sign

CHAPTER 50  Gastrointestinal Tract

A

C

B FIG 50-3  Small leiomyoma of the esophagus detected by CT esophagography. CT esophagography was performed with 16-detector CT scanner after air insufflation via catheter placed in the upper esophagus. A, Transparent rendering image shows a small smoothly marginated mass (arrow) in the upper esophagus. Catheter placed in the upper esophagus is seen as a tubular structure (arrowheads). B, Virtual endoscopic image shows a submucosal mass (arrow) and a catheter (arrowheads) in the upper esophagus. C, Coronal CT image shows a well-defined mass with homogeneous attenuation (arrow). (Courtesy Se Hyung Kim, Seoul National University Hospital, Seoul, Republic of Korea.)

1607

1608

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 50-4  MRI of normal esophagus. A, On the T1-weighted image, the signal intensity of normal esophageal wall (arrowheads) is low. B, On the T2-weighted image, the signal intensity of the esophageal wall (arrowheads) is also low, contrasted with internal contents of high signal intensity. C, After administration of Gd-DTPA, esophageal wall (arrowheads) shows moderate enhancement.

FIG 50-5  Leiomyoma of the esophagus. CT scan shows an eccentric mass (arrowheads) of the esophagus at the level of the carina. The lesion shows a smooth margin and homogeneous soft tissue attenuation.

of a fibrovascular polyp (Fig. 50-7).24 In addition, CT may help determine the intraluminal location of the lesion, particularly when oral contrast material surrounds a mass. MRI demonstrates a lipid-containing substance, mucoid secretion, and nonacute blood within the tumor; on T1, high signal intensity is present, and on T2, soft tissue and muscle within the tumor display low signal intensity. In addition to defining tissue composition, MRI is helpful in defining the spatial location of the mass with the use of MPR images.24,112

FIG 50-6  Calcified leiomyoma of the esophagus. CT scan demonstrates concentric wall thickening (arrowheads) of the GEJ. This lesion contains multifocal calcifications (arrows).

Hemangioma.  Hemangioma of the esophagus is a rare benign tumor that has been reported in fewer than 5% of all esophageal benign tumors. Although the pathogenesis of this neoplasm is still contro­ versial, most agree that hemangioma is a vascular hamartoma or tumorlike malformation.273 The most common symptoms include hemorrhage and dysphagia. Other minor symptoms are epigastric, chest, or retrosternal pain, weight loss, choking, and melena.273 On CT, hemangioma shows a well-defined homogeneous mass with soft tissue attenuation, occasionally containing a small calcification

CHAPTER 50  Gastrointestinal Tract

*

FIG 50-7  Fibrovascular polyp of the esophagus. CT scan shows a large intraluminal polypoid mass (arrows) expanding the lumen of the esophagus. The outer surface of the mass is well defined and contains central low attenuation (asterisk) resulting from adipose tissue component.

regarded as a phlebolith. Administration of a contrast agent produces significant enhancement, indicating high vascularity; in atypical cases, however, enhancement may be poor.640,845 With MRI, hemangioma of the esophagus appears as an isointense mass on T1-weighted images and as a hyperintense mass on T2-weighted images. Both fibrosis and reduction of blood flow in the hemangioma may raise the signal intensity on T1.845

Malignant Esophageal Tumors

Esophageal Carcinoma.  Most malignant neoplasms of the esophagus are squamous cell carcinoma or adenocarcinoma. More rare malignant tumors include small cell carcinoma, spindle cell carcinoma, malignant stromal tumor, Kaposi’s sarcoma, and malignant melanoma. Although squamous cell carcinoma contributes to most cases of esophageal cancer, adenocarcinoma has recently been found to  be the predominant type in patients with newly diagnosed esophageal carcinoma, and Barrett’s esophagus is seen in many of these patients.428,492,497 Various terms have been used to describe the status of esophageal carcinoma.543 For example: • Early esophageal cancer, limited to the mucosa or submucosa, without lymph node metastasis • Superficial spreading cancer, limited to mucosa or submucosa, with or without lymph node metastasis • Small esophageal cancer, in which tumor size is less than 3.5 cm regardless of the depth of invasion and presence or absence of lymph node metastasis CT and MRI have been used to stage esophageal carcinoma, because these techniques can reveal extraesophageal extension into adjacent mediastinal structures including the trachea, bronchi, aorta, and pericardium and can detect lymphadenopathy in the mediastinum. The main goal of staging is to identify nonresectable lesions and thereby avoid the morbidity of major surgery in patients who would not benefit from such intervention.306,599

1609

Squamous cell carcinoma.  Squamous cell carcinoma of the esophagus has been thought to comprise about 90% of cases of primary esophageal malignancy. The incidence of this tumor varies according to geographic location, with the highest incidences in China, Iran, and South Africa.210,620 In the United States, tobacco and alcohol are the two major risk factors; other predisposing factors include achalasia, lye strictures, head and neck tumors, celiac disease, Plummer-Vinson syndrome, and tylosis.897 Dysphagia and weight loss are the classic presenting symptoms of squamous cell carcinoma. By the time dysphagia has developed, the esophageal lumen has already been reduced by more than 50% of its normal caliber, and approximately two thirds of the circumference of the esophageal wall is invaded by the tumor.267,492,897 Ulceration in the tumor may produce odynophagia, and patients may have chest pain unrelated to swallowing if the tumor has invaded the mediastinum. Squamous cell carcinoma may directly invade adjacent structures, involving the trachea, bronchi, lungs, pericardium, aorta, and diaphragm, and may spread to lymph nodes in the mediastinum, neck, and upper abdomen. This tumor can also readily metastasize to the lungs, liver, adrenal glands, and other structures through the bloodstream.497 Adenocarcinoma and Barrett’s esophagus.  The incidence of esophageal adenocarcinoma in the United States and European countries has dramatically increased in the last 40 years, making a once rare tumor the predominant form of esophageal cancer.44,165,777,784 Barrett’s esophagus is the most important predisposing factor in development of adenocarcinoma of the esophagus and is reported to carry a 50- to 100-fold increased risk of adenocarcinoma compared with the general population.120,204,428,492,497,498,543,599,777 It is defined as replacement of the normal squamous epithelium of the distal esophagus by metaplastic columnar epithelium.227 Although it has generally been considered a possible complication of chronic reflux disease, Barrett’s esophagus was also detected in 5.6% to 25% of asymptomatic subjects.253,724,784 In patients with Barrett’s esophagus the risk of developing adenocarcinoma has been estimated at 0.5% per year,785 and it is now thought to follow a metaplasia-dysplasia-carcinoma sequence.777,784 Therefore many investigators advocate endoscopic surveillance of asymptomatic patients with known Barrett’s esophagus to detect the dysplastic changes before development of overt carcinoma.44,777,784 Unlike squamous cell carcinomas, which tend to occur in the upper midesophagus, two thirds of adenocarcinomas arising in Barrett’s esophagus are located in the distal esophagus and tend to invade  the gastric cardia and fundus.497 At present, 50% of tumors at the GEJ are thought to be esophageal adenocarcinoma that is invading the stomach.599 Patients with esophageal adenocarcinomas typically present with recent onset of dysphagia and weight loss. As with squamous cell carcinoma, tumors are usually advanced by the time dysphagia has occurred, so the prognosis is poor (5-year survival rate, ≈20%).784 Small cell carcinoma.  Small cell carcinoma of the esophagus is a rare but aggressive malignant tumor. It occurs in older adults and shows a slight male predominance.64 Clinical manifestations include rapidly progressive dysphagia, odynophagia, weight loss, and symptoms and signs related to hormone production from small cell–like carcinoma of the lung.64,618 It is notorious for its tendency for early metastasis and rapid fatal course and is indistinguishable from pulmonary small cell carcinoma.325,597 Only a few radiologic features of esophageal small cell carcinoma have been reported. Barium study reveals a smoothly marginated polypoid or fungating mass with a relatively flat central ulcer in the midesophagus.499 Usually the tumor expands rather than narrows the esophageal lumen. On CT the tumor shows irregular eccentric

1610

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-8  Small cell carcinoma of the esophagus. A, CT scan at the level of aortic arch demonstrates eccentric wall thickening with soft tissue attenuation in the proximal esophagus (arrowheads). The esophageal lumen (arrow) is filled with contrast medium. B, CT scan of the upper abdomen demonstrates multiple enlarged lymph nodes (arrowheads) in the perigastric area.

esophageal wall thickening and extensive lymph node metastasis to the mediastinum, supraclavicular region, and abdomen (Fig. 50-8).209,854 Because extensive and distant lymph node metastases appear to be a prominent feature of this tumor at presentation, CT examination should cover the abdomen as well as the entire thorax.209 Leiomyosarcoma.  Leiomyosarcoma of the esophagus is a rare lesion, accounting for fewer than 1% of all esophageal malignant neoplasms. It usually is found in middle-aged or older adults and usually occurs in the middle and distal third of the esophagus.891 It carries a better prognosis than squamous cell carcinoma, owing to its tendency for slow growth, low grade, and late metastasis.106,494,706 The growth pattern can be classified as polypoid or infiltrating. CT features include thickening of the esophageal wall and a soft tissue mass containing areas of low attenuation associated with necrosis, large exophytic components, and extraluminal gas or orally administered contrast material within the tumor (Fig. 50-9).494 In the polypoid-type lesion, even if the tumor is large, it does not usually cause obstruction. Therefore a bulky nonobstructing intraluminal mass with a large exophytic component should be considered a leiomyosarcoma. In contrast, an infiltrating tumor is indistinguishable from infiltrative esophageal cancer.491 On MRI this tumor may show a mass with similar signal intensity to skeletal muscle on T1-weighted images and higher signal intensity on T2-weighted images. There may be an area of signal void due to extraluminal gas within the tumor.494 Gastrointestinal stromal tumor.  Gastrointestinal stromal tumors (GISTs) are mesenchymal tumors that are thought to be derived from a precursor of the interstitial cells of Cajal, normally present in the myenteric plexus.617 They occur most frequently in the stomach (63%) and small bowel (30%) but can rarely occur in the esophagus (2%).415 According to Miettinen and colleagues,582 GISTs of the esophagus differ from leiomyomas in following clinicopathologic features: (1) GISTs occur in an older population than leiomyomas (mean age, 63 vs. 35 years); (2) GISTs show a higher overall cellularity than leiomyoma; and (3) on immunohistochemical analysis, GISTs are positive for c-kit (CD117) and CD34 and are negative for desmin, in contrast to leiomyomas, which have the opposite results. Size of the mass was not a

FIG 50-9  Leiomyosarcoma of the esophagus. CT scan shows a large mass (arrowheads) with heterogeneous enhancement. Despite the large size of the tumor, the esophageal lumen (arrow) is merely displaced, not collapsed.

reliable criterion for differentiation of GISTs from leiomyomas; 17% of leiomyomas were larger than 10 cm.582 GISTs of the esophagus have similar morphologic features to those that occur in other anatomic locations.764 On CT scan they appear as a well-defined, hypervascular, enhancing, extraluminal or intraluminal mass (Fig. 50-10). Small tumors tend to appear homogeneous, but larger tumors frequently have a central area of necrosis, hemorrhage, or cystic degeneration.330,337,764 About 50% of patients with GISTs present with metastasis. The most common sites of metastases are the liver and peritoneum.330 Malignant spindle cell tumor.  Malignant spindle cell tumor is an unusual malignant tumor of the esophagus that contains both

CHAPTER 50  Gastrointestinal Tract

U

*

FIG 50-10  GIST of the esophagus. CT scan shows a large well-defined exophytic mass (arrows) arising from the lower thoracic esophagus. The mass has central ulceration (U) that communicates with displaced esophageal lumen (asterisk).

carcinomatous and sarcomatous elements, with spindle cell metaplasia of the carcinomatous portion of the tumor (hence the name spindle cell carcinoma).8,49 On pathologic examination, these tumors consist of innumerable tumorlike spindle cells with islands or nests of carcinomatous cells interspersed throughout the lesion. Affected individuals usually present with recent onset of dysphagia and weight loss. Unfortunately they carry the same poor prognosis as squamous cell carcinoma of the esophagus.492 No reports have demonstrated CT features of malignant spindle cell tumor. In our experience, this disease appears as a large intraluminal mass without obstruction. The lesion shows homogeneous soft tissue attenuation similar to that of stromal tumors on CT scans and more intense enhancement than the adjacent esophageal wall on MRI (Fig. 50-11). On esophagography, malignant spindle cell tumor is shown as a giant polypoid intraluminal mass that characteristically expands the esophagus without obstruction.628 The differential diagnosis for these expansile intraluminal masses includes other rare malignant tumors of the esophagus such as leiomyosarcoma and malignant melanoma.492 Lymphoma.  The incidence of esophageal lymphoma is lowest among all GI tract tumors, accounting for only 1% to 2% of all cases of GI lymphoma.81,626,749 This tumor is usually secondary to generalized non-Hodgkin’s lymphoma, with direct invasion of the esophagus by lymphomatous nodes in the mediastinum, contiguous spread of lymphoma from the gastric fundus, or synchronous development of lymphoma in the wall of the esophagus.749 In rare instances, primary esophageal lymphoma may occur in patients without extraesophageal disease. Primary esophageal lymphomas in patients with acquired immunodeficiency syndrome (AIDS) have been reported.704,759 Most cases of esophageal lymphoma have been discovered at autopsy and in patients with widespread disease, because esophageal lymphoma does not usually produce symptoms. Occasionally esophageal lymphoma may cause dysphagia, bleeding, and fistulas.497,704,759 Radiographically, esophageal lymphoma shows irregular narrowing of the distal esophagus owing to contiguous spread of the tumor from the gastric fundus. However, this finding

1611

cannot be used to differentiate esophageal lymphoma from carcinoma extending from the gastric fundus to the esophagus.110,114,191,491 On CT, esophageal lymphoma appears as irregular thickening of the esophageal wall, with or without large ulceration.704,759 In addition, CT is useful in the evaluation of mediastinal involvement of the lymphoma. Although esophageal lymphoma is rare, the diagnosis should be considered in patients at risk for AIDS when radiologic findings are not typical for infectious esophagitis or Kaposi’s sarcoma. Occasionally these tumors can produce enlarged tortuous folds mimicking the appearance of varices. When multiple nodules are present in the esophagus, the differential diagnosis includes leukemic infiltrates, hematogenous metastases, Kaposi’s sarcoma, and multiple stromal tumors.704 Primary malignant melanoma.  Primary malignant melanoma of the esophagus is extremely rare. It originates from the small numbers of melanocytes and their precursors in the esophageal mucosa. As in skin melanoma, esophageal melanoma presumably develops because of malignant degeneration of these preexisting melanocytes.162 Primary malignant melanomas of the esophagus appear on barium studies as polypoid intraluminal masses with lobulations. Bulky masses may focally expand the esophagus but rarely cause considerable obstruction.263,905 On CT scans, malignant melanomas are shown as large intraluminal masses surrounded by a rim of intraluminal air or oral contrast, which may infrequently be associated with mediastinal lymphadenopathy.263,905 Primary melanoma of the esophagus should be differentiated from other polypoid tumors of the esophagus, including spindle cell carcinoma, leiomyosarcoma, fibrovascular polyp, and squamous cell carcinoma.

Tumor Staging Tumor-node-metastasis classification.  Adequate staging of esophageal cancer is important in the initial triage of patients to receive immediate surgery, neoadjuvant therapy, or palliative treatment. The tumor-node-metastasis (TNM) classification proposed by the American Joint Committee on Cancer (AJCC) is the most widely accepted staging system for esophageal cancer (Table 50-1).267,767 TNM staging applies to all histologic types of carcinomas. Regardless of pathologic types and longitudinal extents of tumors, only the depth of penetration in the esophageal wall, nodal status, and presence of metastasis are considered in staging. The definitions of regional lymph nodes are different according to the location of tumors. For intrathoracic tumors, paraesophageal and abdominal nodes cephalic to the celiac axis are considered regional lymph nodes, whereas cervical and celiac nodes are considered distant metastasis.198,602 On the other hand, celiac axis nodes for tumors in the GEJ and cervical nodes for those in cervical esophagus are regarded as regional lymph nodes.198 Primary tumor.  The depth of tumor infiltration into the esophageal wall—one of the most important prognostic factors—can best be evaluated by endoscopic ultrasonography (EUS), which can directly visualize all layers of the esophageal wall.329 Compared with EUS, CT is not a reliable tool for evaluation of the extent of the tumor in the esophageal wall because it cannot definitively demonstrate the individual layers of the wall. Therefore CT cannot differentiate T1 from T2 tumor.669 In a meta-analysis study,748 the pooled accuracy of EUS in T staging was 85%, but that of CT was only 45%. MRI also cannot reliably distinguish individual layers of the esophageal wall and has been shown to have accuracy similar to CT for predicting T stage.361 Recently, high-resolution MRI (HRMRI) using phased-array surface coils has been tested for its application to imaging of esophageal cancer.733,734 In these preliminary studies, HRMRI was technically feasible and could provide detailed imaging of the anatomic layers of the esophageal wall and tumor. Further study may be required to access the accuracy of HRMRI in staging esophageal cancer. 18Fluorodeoxyglucose positron

1612

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 50-11  Malignant spindle cell tumor of the esophagus. A, CT scan shows a giant polypoid intraluminal mass (arrowheads) with homogeneous soft tissue attenuation. This lesion is hypointense on T1-weighted image (B) and heterogeneously hyperintense on T2-weighted image (C). After administration of Gd-DTPA (D), the lesion (arrowheads) shows considerable heterogeneous enhancement. The lesion does not cause esophageal obstruction despite its large size.

emission tomography (FDG PET) also has a limited role in evaluating the depth of tumor invasion.196,361,602,731,894 Lowe and colleagues have reported that T staging was correctly evaluated by CT and FDG PET in only 42% of patients with esophageal cancer compared with 71% with EUS.518

On CT scans, esophageal carcinoma appears as a mass or wall thickening with soft tissue attenuation. Regarding optimal scan phase for depicting esophageal carcinoma, late arterial phase (35 seconds after attenuation of the descending aorta reached 200 Hounsfield units [HU]) has been shown to be better than early arterial and venous

CHAPTER 50  Gastrointestinal Tract

1613

TABLE 50-1  International Staging System

for Esophageal Carcinoma Stage

Description

Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of primary tumor Tis High-grade dysplasia T1 Tumor invades lamina propria, muscularis mucosae, or submucosa T1a Tumor invades lamina propria or muscularis mucosae T1b Tumor invades submucosa T2 Tumor invades muscularis propria T3 Tumor invades adventitia T4 Tumor invades adjacent structures T4a Resectable tumor invading pleura, pericardium, or diaphragm T4b Unresectable tumor invading other adjacent structures, such as aorta, vertebral body, trachea, etc. Regional Lymph Nodes (N)* NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis in 1-2 regional lymph nodes N2 Metastasis in 3-6 regional lymph nodes N3 Metastasis in 7 or more regional lymph nodes Distant Metastasis (M) M0 No distant metastasis M1 Distant metastasis† From Edge S et al (eds), American Joint Committee on Cancer: AJCC cancer staging manual, ed 7. New York, 2011, Springer-Verlag, p 109. *Regional lymph nodes include cervical, mediastinal, and perigastric nodes. † Celiac lymph node metastasis is regarded as N1 in adenocarcinoma; however, it is regarded as M1 in squamous cell carcinoma.

phases, because tumor enhancement and the difference in attenuation between tumor and normal esophageal wall were maximal at late arterial phase.850 With MRI, esophageal carcinoma is seen as an area of isointensity relative to residual esophageal wall on T1-weighted images and as an area of intermediate or high signal intensity on T2-weighted images.733,826 Macroscopic periesophageal fat invasion by the tumor can be demonstrated with high accuracy on CT and MRI, although microscopic invasion to periesophageal fat is not revealed.142,669 On CT scans, periesophageal fat invasion is shown as strandlike areas of soft tissue attenuation that generate a feathery appearance within the periesophageal fat plane (Fig. 50-12).599 MRI shows periesophageal fat invasion as having an ill-defined irregular margin, regional absence of a bright fat signal, and invasion of tumor mass into the adjacent mediastinal fat.329,899 Overall accuracy of periesophageal fat invasion of esophageal carcinoma on CT or MRI ranges from 88% to 100%.329 Although periesophageal fat infiltration adversely affects prognosis, it is not a contraindication to en bloc resection of tumor as a curative treatment.599 Accurate evaluation of tracheobronchial, aortic, and bone invasion is important because these are contraindications to surgical resection. The most frequent sites of tracheobronchial involvement are the distal trachea and left mainstem bronchus.149 CT or MRI appearances of tracheobronchial invasion are displacement or indentation of the

FIG 50-12  Esophageal cancer invading mediastinal fat. CT scan demonstrates circumferential wall thickening of the distal esophagus (arrow). The periesophageal fat is replaced by soft tissue attenuation (arrowheads).

posterior wall of the trachea or the bronchus by tumor mass, as well as evidence of direct extension of the tumor into the lumen of the airway (Fig. 50-13). Reported data indicate that the overall accuracy of tracheobronchial invasion of esophageal carcinoma ranges from 88% to 97%.329 Absence of an intervening fat plane alone, however, does not always suggest tumor invasion; this finding can be seen in normal patients.552 Moreover, an adjacent mass behind the trachea or mainstem bronchus can produce concavity of their posterior portions without invasion (Fig. 50-14), because the cartilage rings are incomplete in the posterior portion of the trachea and mainstem bronchi.249 According to Picus and coworkers, aortic invasion can be diagnosed by CT if the area of contact between the esophageal cancer and the aorta creates an arc of more than 90 degrees (Fig. 50-15).682 The accuracy of this 90-degree criterion is better than 90% when mediastinal fat planes are preserved above and below the tumor.51,682,842 If the arc is less than 45 degrees, aortic invasion is considered to be absent; an arc of 45 to 90 degrees is considered indeterminate.682 The overall accuracy of MRI in evaluating aortic invasion is similar to that of CT.249,669,826 Although the presence of the fat plane rules out invasion, absence of the fat plane between the aorta and tumor does not always indicate invasion, such as in tracheobronchial invasion. This major pitfall occurs in cachectic patients in whom the fat planes are absent throughout the mediastinum. Without clear evidence of displacement of the aorta and direct contact between the aorta and tumor, aortic invasion cannot be diagnosed.552 Bone invasion can be easily identified on CT or MRI with the findings of bone destruction and replacement of bone by soft tissue that are continuous with the main tumor mass.552,826 Pleural, pericardial, diaphragmatic, and gastric invasion are not contraindications to surgery, because the involved areas can be removed together with the

1614

PART II  CT and MR Imaging of the Whole Body

FIG 50-13  Esophageal cancer with tracheal invasion. CT scan shows circumferential wall thickening of the proximal esophagus (arrowheads), which shows irregular interface with the posterior wall of the trachea (arrows), indicating direct extension into the lumen.

FIG 50-14  Concavity of the posterior tracheal wall mimicking invasion. Although the posterior tracheal wall is concave (arrow), the surface of the posterior wall is smooth. Note that only the concavity of the trachea, without surface irregularity or apparent intraluminal protrusion, does not suggest the tumor invasion.

FIG 50-15  Esophageal cancer with aortic invasion. An arc (bent arrow) of the contact between the esophageal cancer (arrows) and the aorta (arrowheads) is more than 90 degrees, indicating aortic invasion.

tumor mass at the time of operation. Therefore the misdiagnosis with CT in identifying tumor involvement in these areas is less critical.51,842 Pericardial invasion can be readily diagnosed with high accuracy  (88% to 97%).697,841 CT or MRI appearances of pericardial invasion697,826 are as follows: • Pericardial thickening adjacent to the tumor • Pericardial effusion • Inward deformity of the heart with loss of the intervening fat plane at the level of the tumor, with intact fat above and below the tumor Although the patient may be cachectic or may have undergone radiation, which can obliterate normal fat planes, tumor invasion should be suggested when a localized obliteration of the fat plane is noted, with preserved fat planes above and below the lesion.307,599,844 The CT feature of gastric invasion is a soft tissue mass extending from the esophageal tumor into the proximal stomach under adequate distention.682 On MRI, gastric tumor extension is diagnosed when a T2-weighted image depicts a mass of high signal intensity to the gastric wall extending from the esophageal tumor into the gastric fundus.826 The overall accuracy of gastric invasion on CT and MRI ranges from 79% to 97%.329,826 Gastric invasion frequently occurs when the tumor locates in the distal esophagus, especially in adenocarcinoma. Pleural and diaphragmatic invasion cannot be clearly demonstrated on CT because adhesion to these structures by tumor can easily occur without invasion.682,697 Lymph node metastasis.  Metastatic spread of esophageal cancer occurs commonly in regional or distant lymph nodes because a rich network of lymphatic channels along the esophagus and lack of serosa allows early extraesophageal tumor invasion.767 Cross-sectional images, including EUS, CT, and MRI, have shown a wide range of accuracy in detecting lymph node metastasis.329,603,669 However, most studies concur that EUS shows the highest accuracy (73%-88%) among them.329,603 Use of EUS-guided fine-needle aspiration (FNA) can improve the accuracy of EUS for nodal staging, with reported accuracy ranging from 77% to 94%.194,547,861,866 Therefore EUS-guided FNA is currently regarded as the most reliable method to assess locoregional lymph node metastases.547,861

CHAPTER 50  Gastrointestinal Tract

A

1615

B FIG 50-16  Distant lymph node metastases of esophageal cancer detected by integrated CT PET. A, Integrated CT PET demonstrates paraaortic lymph node metastases showing increased FDG uptake (arrowheads). B, Corresponding CT image shows lymph nodes (arrowheads) measuring 5 to 8 mm in diameter. Based on size criteria, these lymph nodes may be considered benign on CT scan.

With regard to the accuracy of CT and MRI, one prospective series showed only 56% accuracy for mediastinal lymph nodes and 45% for abdominal lymph nodes.329 Although controversial, MRI may provide accuracy similar to that of CT in detecting a metastatic mediastinal lymph node.485 It has been widely accepted that lymph nodes with a short axis greater than 1 cm in diameter should be considered to be a metastasized node in the periesophageal region, other mediastinal regions, and the abdomen.606,892 However, this criterion still poses a problem because of the high frequency of false-positive and falsenegative results.841 In some instances, concomitant presence of inflammation associated with tumor may produce large lymph nodes, which limits accurate predictions of lymph node metastasis on CT. The size criteria for lymph node metastasis typically used for CT are not applicable to esophageal tumors because of microscopic invasion of normal-sized nodes. Application of different size criteria to some lymph node chains may improve the capability of cross-sectional imaging for determining metastatic nodes; according to Balfe and associates,31 the upper limit of normal lymph node in gastrohepatic ligament was 8 mm. The sensitivity of FDG PET in the detection of nodal metastases is poor overall and has been reported to be in the range of 22% to 57%.196,225,489,518,894 The poor sensitivity of FDG PET is probably due to its limited spatial resolution, which makes it difficult to differentiate FDG uptake in metastatic lymph nodes close to the primary tumor from that in the primary tumor.637 Distant metastasis.  The most common sites of distant metastases are nonregional lymph nodes, liver, lungs, bones, and adrenal glands. CT clearly demonstrates the metastatic lesions in solid upper abdominal organs and the lung. On CT scans, pulmonary metastasis is shown as one or multiple nodules, and pleural effusion may coexist. Often a single pulmonary nodule produces the dilemma in predicting metastasis. In this situation, biopsy of the solitary pulmonary nodule is usually recommended. Liver metastasis manifests as single or multiple nodules with variable size and ill-defined margins, which are not enhanced as much as normal parenchyma after administration of an IV contrast agent. Although the frequency is low, central necrosis in the metastatic nodule may mimic hepatic cyst.552 FDG PET has been shown to be potentially useful in the evaluation of distant metastases.95,225,858,859,870,874 FDG PET has been reported to

detect distant metastases in up to 20% of patients who are initially considered to have resectable disease based on conventional staging (Fig. 50-16).95,874 In addition it can detect distant metastases that are unusual in appearance and are in an unexpected location in 7% of patients.95 According to a recent meta-analysis,859 the sensitivity and specificity of FDG PET for detection of distant lymphatic and hematogenous metastases was 67% and 97%, respectively. Considering the high specificity of FDG PET in the detection of distant metastases, it may complement invasive methods such as biopsy for the evaluation of equivocal lesions detected on CT or MRI. Other staging methods.  Although many investigators have evaluated and compared the usefulness of CT and MRI in staging esophageal carcinoma, controversies remain about which imaging modality is superior. In clinical situations, CT is usually recommended as a routine examination. EUS has also taken an important role in the staging and is known to be more accurate for locoregional staging than CT or MRI.361,508,547,748,862,866 Because CT and MRI can detect distant metastases that cannot be evaluated by EUS, CT/MRI and EUS are considered complementary modalities for esophageal cancer staging.508 Therefore a combination of CT and EUS is usually performed for initial staging of patients with esophageal cancer. FDG PET is increasingly used for the staging of esophageal  cancer.95,196,390,391,435,436,445,518,520,573 Recent use of integrated CT PET imaging with coregistration of anatomic and functional imaging data improved the localization of regions with increased metabolic activity and the accuracy of staging in these patients.45,196 In the initial staging of esophageal cancer, the major advantage of FDG PET or integrated CT PET over anatomic imaging modalities is probably the ability to detect unexpected distant metastases, which can prevent noncurative surgery.95,196 Although several researchers claimed that FDG PET may be cost-effective in the prevention of noncurative surgery by detecting metastases not identified by conventional imaging modalities859,870 and should be included in the staging workup of patients with esophageal cancer,196,224,637,802 whether FDG PET or integrated CT PET has a direct impact on therapeutic decision making is still under debate.391,435,436,894

Tumor Resectability.  The goal of pretreatment staging in esophageal carcinoma is to evaluate tumor resectability. The major criteria for nonresectability include (1) aortic invasion, (2) tracheal invasion, and

1616

PART II  CT and MR Imaging of the Whole Body

(3) distant metastasis. According to Takashima and colleagues, sensitivity, specificity, and accuracy for resectability are 100%, 84%, and 87%, respectively, for MRI and 100%, 80%, and 84%, respectively,  for CT.826

Assessment of Therapeutic Response to Preoperative Neoadjuvant Therapy.  Primary surgery is considered the treatment of choice for patients with early-stage esophageal cancer, which is T1 to T2 cancer with limited or no lymph node involvement.488 Although there are ongoing debates on the optimal treatment method for resectable advanced esophageal cancer, preoperative neoadjuvant therapy aiming at downsizing or downstaging the disease has recently gained growing acceptance.217,488,866 Recent meta-analyses demonstrated that neoadjuvant chemoradiotherapy followed by surgery provides better survival than surgery alone.275,852 Accurate restaging of esophageal cancer after neoadjuvant chemoradiotherapy is important to select candidates for subsequent surgical resection. As in initial staging of esophageal cancer, CT and EUS are frequently used for restaging esophageal cancer. However, as opposed to their high accuracy in initial staging, CT and EUS have been shown to be inaccurate for restaging of the disease after adjuvant chemoradiotherapy.63,380,386,508,547,888 Accuracy in predicting T stage after neoadjuvant chemotherapy or radiotherapy of esophageal cancer has been reported as only 42% for CT380 and 27% to 82% for EUS.63,386,508 The main reason for poor accuracy of CT and EUS is their inability to distinguish inflammation and fibrosis associated with chemotherapy or radiotherapy from residual cancer.380,386,508,547 On CT scan, obliteration of the mediastinal fat plane that resulted from radiation-induced fibrosis cannot be differentiated from residual tumor infiltration.380 Although integrated CT PET may potentially distinguish residual tumor from inflammation or fibrosis, there is little evidence that integrated CT PET provides more accurate T staging. According to a prospective study to assess accuracy of CT, EUS-guided FNA, and integrated CT PET in restaging esophageal cancer after neoadjuvant therapy,117 accuracy of the three tests for distinguishing nonresectable T4 from resectable T1 to T3 disease was similar (≈80%), whereas FDG PET provided more accurate nodal staging than CT and EUS FNA. Response to neoadjuvant therapy is one of the most important factors determining patient prognosis.637 Patients who respond to neoadjuvant therapy benefit most from subsequent surgical resection, whereas “nonresponders” are likely to experience an unfavorable course.94 Furthermore, recent treatment strategy is moving toward individualization of therapeutic modalities by selecting a therapy that is expected to be most effective in each patient, which in turn requires a diagnostic test that can predict therapeutic response early in the course of neoadjuvant treatment and can aid in the decision about whether a treatment with considerable side effects should be continued or replaced with another therapeutic option. In this context, metabolic imaging with FDG PET or integrated CT PET rather than classic anatomic imaging modalities is a more suitable tool for monitoring therapeutic response.63,117,282,380,386,508,637,875,884 Compared with CT and EUS, FDG PET or integrated CT PET have been shown to more accurately reflect histologic response of the tumor after neoadjuvant therapy and to predict the response of the tumor earlier in the course of neoadjuvant therapy (Fig. 50-17).117,877 Change in tumor metabolic activity measured with FDG PET early in the course of neoadjuvant therapy can predict reduction in tumor size after the completion of therapy884 as well as differentiate responding and nonresponding tumors.875 Therefore FDG PET may potentially facilitate the use of neoadjuvant therapy by excluding nonresponders, who seem not to benefit from ineffective and potentially harmful treatment earlier.637,875

Postoperative Evaluation.  In evaluating patients with esophagogastrectomy, the radiologist must be aware of postoperative appearance. Normal postoperative CT features have been described by Heiken and coworkers.315 In patients who have undergone a high anastomosis with gastric conduit, both the esophagogastric junction and stomach are located in the right paravertebral area, and the proximal esophagus above the anastomosis can be dilated (Fig. 50-18). In the lower retrocardiac area, the stomach passes from right to left directly anterior to the descending aorta and then begins to narrow as it enters the diaphragmatic hiatus anterior to the aorta. In patients who have undergone low esophagogastric anastomosis, the esophagogastric junction and stomach lie in the left paravertebral region adjacent to the lower thoracic descending aorta. CT and MRI can play a role in postoperative evaluation because they aid diagnosis of early postoperative complications, including mediastinitis or abscess caused by anastomotic leak, lung parenchymal consolidation, pleural effusion or empyema, and subphrenic abscesses.283,315,753 Among these complications, anastomotic leak was known to be the most fatal complication. However, more recent data indicated that mortality rate dropped to 3.3%, probably owing to early diagnosis and adequate management.549 Although large leaks may require surgical intervention, most intrathoracic leaks can be managed with conservative treatment.423 In addition, anastomotic leak can cause fistula formation with adjacent structures, such as esophagobronchial, esophagopleural, or esophagocutaneous fistula. Serial postoperative CT plays a major role in detecting recurrent tumors.423 The overall accuracy of CT in detecting tumor recurrence is reported to be 87%.109 CT findings with postoperative recurrence include soft tissue mass, impression and displacement of normal structures, metastasis to distant organs, abdominal lymph node enlargement, pericardial effusion, and pleural effusion (Fig. 50-19).109 FDG PET or integrated CT PET may be a useful adjunct to conventional imaging studies because FDG PET may demonstrate increased metabolism of recurrent tumor before the development of any anatomic change at conventional imaging (see Fig. 50-19).423 Reported accuracy of FDG PET in the detection of recurrent esophageal cancer ranges from 82% to 87%.224,390

Other Diseases

Inflammation.  Most inflammations of the esophagus are caused by gastroesophageal reflux or infection. Gastroesophageal reflux disease (GERD) is the most common inflammatory disease involving the esophagus. Esophageal infection occurs frequently in immunosuppressed patients with malignancy, organ transplantation, and AIDS and includes Candida albicans, herpes simplex, and cytomegalovirus (CMV).552,620 Candidiasis is the most common cause of infectious esophagitis and is thought to result from downward spread of the fungus from the oropharynx to the esophagus. On CT scan, esophagitis appears as a long-segmental, uniform, circumferential esophageal wall thickening, often combined with a target sign (Fig. 50-20).61,148,718 According to Berkovich and associates, 55% of patients with esophagitis show an esophageal wall more than 5 mm thick, and 17% of patients exhibit the target sign.61 The thickening is caused by edema, infiltration with inflammatory cells, and muscular spasm. Compared with esophageal carcinoma, which usually shows asymmetric wall thickening and radial extension into the mediastinum, the mediastinal fat plane is preserved in esophagitis.552 Pearlberg and colleagues described the CT findings of esophageal intramural pseudodiverticulosis, which was associated with longstanding esophagitis and characterized by dilatation of the excretory ducts of esophageal mucous glands.662 CT features of this condition include intramural air collections at multiple levels within a

CHAPTER 50  Gastrointestinal Tract

A

B

C

D FIG 50-17  Assessment of treatment response of esophageal cancer to neoadjuvant chemoradiotherapy. A, Initial CT scan shows an encircling mass at the thoracic esophagus (arrowheads). B, Initial PET scan shows increased uptake of FDG in the esophageal cancer (arrowhead) and metastatic lymph nodes (arrows). C, Follow-up CT after completion of neoadjuvant chemoradiotherapy shows persistent wall thickening of the thoracic esophagus (arrowheads). Differentiation of radiation-induced fibrosis from residual tumor is not possible on CT. D, FDG PET scan after completion of neoadjuvant chemoradiotherapy shows complete metabolic response of the esophageal cancer as well as metastatic lymph nodes. Subsequent esophageal resection confirmed the absence of viable tumor.

1617

1618

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-18  Postoperative status of esophageal cancer (high esophagogastric anastomosis). A, Proximal anastomosis (esophagogastric junction) (arrows) is located posterior to the trachea and to the right of the vertebra. B, In the distal portion, the stomach (arrows) passes from right to left in front of the descending aorta.

A

B FIG 50-19  Recurrent esophageal cancer at the anastomotic site. A, Contrast-enhanced CT scan shows wall thickening in the esophagogastrostomy site (arrows). B, Integrated CT PET demonstrates increased FDG uptake of recurrent esophageal cancer at the esophagogastrostomy site (arrows).

thick-walled esophagus. Postinflammatory stricture may produce luminal narrowing without wall thickening. When circumferential soft tissue thickening is observed, it is due to fibrosis.552 The esophagus may be involved secondarily from adjacent mediastinal inflammatory processes. Histoplasmosis or tuberculosis (TB) may result in tracheoesophageal or bronchoesophageal fistula and subsequent aspiration pneumonia or lung abscess. The most common cause of esophageal TB is secondary involvement from adjacent

tuberculous lymphadenitis rather than primary esophageal TB. CT is useful in depicting esophageal TB and can demonstrate tuberculous mediastinal lymphadenopathy surrounding and displacing the esophagus and fistulous tract and extending from the esophagus into the nodal mass (Fig. 50-21).60,354,366,887 Even without a definite fistulous tract, amorphous gas collection at enlarged mediastinal lymph nodes should suggest the presence of an esophagonodal fistula in patients with TB.354

CHAPTER 50  Gastrointestinal Tract

1619

FIG 50-22  Tracheoesophageal fistula. CT scan shows an air-filled communication (arrow) between the esophagus (arrowhead) and the mainstem bronchus (curved arrows).

FIG 50-20  Esophagitis. CT scan shows circumferential esophageal wall thickening with the “target sign” (arrowheads). The esophageal fat is preserved in esophagitis, which is a point of differentiation from esophageal carcinoma.

Esophageal-airway fistulas can occur as complications of tumors or infections and can be caused by trauma.366 Berkmen and Auh also reported that CT could demonstrate the fistula that appeared as an air-filled communication between the esophagus and airway (Fig. 50-22).60 Thin section (2-4 mm) may be helpful in their detection. Even when the actual fistula is not demonstrated, associated findings, including infiltration of mediastinal fat due to inflammation and fibrosis, pulmonary consolidation, pulmonary cavitation, and pleural fluid, help suggest the diagnosis.

Achalasia.  Achalasia is an uncommon disease of the esophagus, with

FIG 50-21  Fistula between the esophagus and tuberculous lymphadenopathy. CT scan demonstrates an air-filled fistulous tract (arrows) between the esophagus (arrowheads) and the enlarged subcarinal lymph node (curved arrows).

an incidence of 0.2 to 0.6 per 100,000 per year.75,702 It is caused by aperistalsis of the lower esophageal segment and inadequate relaxation of the lower esophageal sphincter (LES). The pathogenesis of achalasia is obscure, and the histologic basis is unknown.702 There is no sex predilection. The condition usually occurs in middle-aged patients. The usual clinical symptom is dysphagia, which is usually worse with liquids than with solids, although sometimes only nonspecific symptoms such as heartburn may be present. The diagnosis of achalasia is usually made on the basis of esophagography, manometry, nuclear examination, and endoscopy. Radiographic findings of achalasia have been reported by many authors.189,522,635 On barium study the primary peristalsis is absent on all swallowing studies. The lower end of the esophagus shows a smooth “beaklike” tapering located at the level of the esophageal hiatus. This tapering reflects LES dysfunction and failure of the barium bolus to distend the tonically contracted LES.635 Recent preliminary studies have shown the potential value of MR fluoroscopy in the evaluation of achalasia and other esophageal motility disorders.447,643 The advantage of MR fluoroscopy over barium study is the fact that MR fluoroscopy can depict functional and morphologic alterations of the esophagus as well as abnormalities in esophageal wall and extraluminal structures at the same time. Ultrafast imaging sequences such as true fast imaging with steady-state precession (FISP) or T1-weighted turbo-fast low-angle shot (FLASH) were used for real-time evaluation of esophageal peristalsis. Initial studies have shown the feasibility of MR fluoroscopy in the evaluation of esophageal motility disorders. Further optimization

1620

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-23  Achalasia in a 47-year-old woman. A, CT shows dilated fluid-filled esophagus (arrow). B, In this patient the distal esophageal wall thickness (arrow) is normal, in contrast with pseudoachalasia resulting from esophageal cancer.

of imaging protocol and more clinical experiences may enable practical use of MR fluoroscopy in the future. CT features of achalasia have been described. Rabuschka and colleagues reported that a unique CT feature of achalasia is moderate dilatation of the esophagus, with a mean diameter of 1.4 cm at the carinal level, not combined with wall thickening in the distal segment of the esophagus702 (Fig. 50-23). They mentioned that normal wall thickness is the most important feature that distinguishes patients with achalasia from those with other esophageal diseases. The role of CT in evaluating patients with achalasia is not only to detect the disease and its complications but also to demonstrate atypical features that may indicate the presence of other diseases or superimposed benign or malignant processes. Other causes of LES narrowing that may mimic achalasia include intrinsic or extrinsic neoplasms, peptic stricture, and complications of scleroderma. Patients with tumor infiltration into the distal esophagus may report symptoms similar to achalasia. This entity has been named pseudoachalasia, and it is important to distinguish it from primary achalasia. Esophageal cancer occurs in 2% to 7% of patients with longstanding achalasia.621,702,896 In this situation, cancer develops in the midesophagus, and the etiologic mechanism has been presumed to be stasis and irritation of the esophagus.702 When patients with achalasia experience rapid worsening of symptoms, the possibility of superimposed esophageal cancer is suggested.

Esophageal Varices.  Esophageal varices may be classified as uphill or downhill; the former is more common than the latter. Uphill varices develop by portal hypertension, in which increased portal venous pressure leads to hepatofugal venous flow through the coronary vein into a plexus of dilated esophageal and periesophageal veins. Downhill

varices develop by obstruction of the superior vena cava and result in reversal of flow into the cervical and upper thoracic esophagus via the superior intercostal vein, bronchial vein, inferior thyroid vein, and other mediastinal collateral vessels.322 Esophageal varices are important because of the potentially catastrophic consequences of variceal rupture and hemorrhage. On CT and MRI, esophageal varices are recognized by the presence of a dilated vessel on the inner surface of the esophageal wall that is markedly enhanced after administration of IV contrast medium (Fig. 50-24).40,134,419,426,554,793 According to Balthazar and associates, CT is superior to barium esophagography in evaluating the presence and extent of periesophageal varices and portal hypertension, although the sensitivities in detection of the lesion are similar.40 Reported sensitivity of CT was more than 90% for detection of clinically significant large esophageal varices.426,793 Recently MDCT virtual endoscopy technique has been applied to the evaluation of esophageal varices.419 Kim and coworkers performed MDCT esophagography using air insufflation via a catheter inserted transorally into the upper esophagus in patients with liver cirrhosis and reported that CT esophagography can allow accurate grading of esophageal varices with better patient acceptance than endoscopy (see Fig. 50-24).419 Considering the promising results of initial studies,419,426,793 CT may play a role in selecting patients with high-risk varices who are potential candidates for preventive treatment. On CT scans, therapeutically sclerosed esophageal varices may sometimes mimic esophageal carcinoma.304 CT features of esophageal varices after uncomplicated endoscopic esophageal sclerotherapy include esophageal wall thickening, low attenuation within the wall (giving the esophagus a laminated appearance), a mediastinal effusion of predominantly low attenuation that is often masslike, obliteration

CHAPTER 50  Gastrointestinal Tract

1621

* * *

A

* *

B

* * *

C FIG 50-24  Esophageal varices. A, Contrast-enhanced CT scan that was performed after air insufflation via catheter placed in the upper esophagus shows a well-enhancing nodular lesion (arrows) protruding into the lumen of the esophagus. B, Virtual endoscopic image shows large esophageal varices seen as serpiginous protruding lesions (asterisks) in the lower esophagus. C, Endoscopic image also shows large esophageal varices (asterisks) in the lower esophagus. (Courtesy Se Hyung Kim, Seoul National University Hospital, Seoul, Republic of Korea.)

of mediastinal fat planes, thickening of the diaphragmatic crura, pleural effusions, and subsegmental atelectasis.556

TRAUMA Esophageal Hematoma Esophageal hematoma is related to variable conditions such as abnormal hemostasis, endoscopic instrumentation, variceal sclerotherapy, foreign body ingestion, remote trauma, and vomiting.167,374,918 Clinical manifestations include sudden onset of retrosternal chest pain, dys-

phagia, and hematemesis. Most esophageal hematomas resolve spontaneously with only conservative treatment and do not result in esophageal perforation.490 On barium studies, the hematoma appears as an ovoid or elongated submucosal mass in the esophagus. When a mucosal laceration is present (particularly after iatrogenic injury), contrast medium may dissect beneath the mucosa into the hematoma. This intramural dissection may produce a characteristic “double-barreled” appearance due to parallel collections of contrast medium in both true and  false lumens, separated by a thin radiolucent stripe.490,664 However,

1622

PART II  CT and MR Imaging of the Whole Body

intramural tracks can also be caused by Crohn’s disease, Candida esophagitis, or tuberculous esophagitis.664 On CT scans, esophageal hematoma appears as a nonenhancing submucosal lesion with high or low attenuation in the middle and lower esophagus. MRI reveals an intermediate signal intensity on T1-weighted images and a hyperintense signal on T2-weighted images.167,918

Esophageal Perforation Esophageal perforation is a catastrophic condition that may rapidly progress to fulminant mediastinitis and septic shock. Most cases are caused by endoscopy or other forms of esophageal instrumentation. In other cases, spontaneous perforation of the esophagus (Boerhaave’s syndrome) may occur as a result of a sudden rapid increase in intraluminal esophageal pressure. Initial symptoms and signs include sudden onset of severe retrosternal chest pain, vomiting, and subcutaneous emphysema with crepitus in the soft tissues of the anterior chest wall (mediastinal crunch). In

some cases, however, the initial signs and symptoms are nonspecific and may consist of hypotension, sepsis, or fever, falsely suggestive of myocardial infarction, acute aortic dissection, or intraabdominal abnormalities.368,378,560,658,879 The prognosis is directly related to the interval between perforation and initiation of therapy; within 24 hours the mortality rate for thoracic esophageal perforation exceeds 80%.25 In patients with spontaneous esophageal perforation, chest radiographs may reveal pneumomediastinum associated with left or (less commonly) right pleural effusion or bilateral effusions. Esophagography may demonstrate extravasation of contrast medium from the left lateral wall of the distal esophagus into the adjacent mediastinum. Often, however, the clinical features are atypical, and up to 10% of patients with esophageal perforation may have false-negative findings on contrast esophagography.76 CT features of esophageal perforation include esophageal wall thickening, periesophageal fluid, periesophageal air, pleural effusion, pericardial effusion, esophageal wall laceration, and extravasation of oral contrast material (Fig. 50-25). Pleural effusions are usually

P

*

*

A

B

C FIG 50-25  Esophageal perforation (Boerhaave’s syndrome) in a 32-year-old-man. After vigorous vomiting, the patient experienced sudden chest pain. A, CT scan shows a thickened esophageal wall (arrow), pneumomediastinum (asterisks), pleural effusion (arrowheads), and pneumothorax (P). B, Esophagography with water-soluble contrast agent shows leakage of oral contrast agent freely into the left pleural space (arrowheads). C, CT scan performed immediately after esophagography shows that high-density oral contrast agent in the esophagus (arrow) leaks into mediastinum and left pleural space (arrowheads).

CHAPTER 50  Gastrointestinal Tract bilateral. CT is ideally suited for defining the extent of extraluminal air-fluid collection and for detecting small amounts of extravasated contrast material not appreciated at esophagography. In addition, CT may also be useful in monitoring the clinical course of patients treated conservatively.26,164,195,203,378,671

1. Sliding hernia, in which the GEJ is above the esophageal hiatus of the diaphragm (Fig. 50-26) 2. Paraesophageal hernia, in which all or portions of the stomach herniate into the chest but the GEJ is below the diaphragm  (Fig. 50-27) Approximately 99% of hiatal hernias are sliding and the remaining 1% are paraesophageal.490 Although the incidence of paraesophageal hernia is much lower than that of sliding hernia, patients with a paraesophageal hernia have a higher risk for developing volvulus, incarceration, and strangulation of a herniated portion. Hiatal hernia can be readily detected on a single barium study, but CT and MRI also can demonstrate the size, content, and orientation

Foreign Body Impactions Esophageal foreign body impaction usually occurs in children, who tend to ingest coins, toys, or other foreign materials. In adults, foreign body impaction in the esophagus is usually caused by an inadequately chewed bolus of food or bones. Unchewed food tends to lodge in a lower esophageal ring or stricture, whereas bones tend to lodge in the pharynx near the cricopharyngeus.490 Most patients with food impaction present with sudden onset of chest pain, odynophagia, or dysphagia. In contrast, young children usually exhibit respiratory distress, drooling, or regurgitation.817 Esophageal perforation is rare, affecting fewer than 1% of all patients with foreign body impaction, but the risk of perforation increases significantly if the impaction persists longer than 24 hours and if the perforation of the cervical esophagus with a foreign body produces a retropharyngeal abscess.490 Plain radiography, especially lateral radiographs of the neck, may occasionally demonstrate bones or other radiopaque foreign bodies. Barium study is another option when a foreign body is suspected in the esophagus; barium can determine whether a foreign body is present and whether it is causing obstruction. The impacted foreign body typically appears as a polypoid filling defect with an irregular meniscus in the esophagus, owing to barium’s outlining the superior border of the impacted food bolus. CT is also useful in evaluating esophageal foreign body impaction. With its soft tissue and bone windows, CT may replace the barium swallow because of its superior detection of thin, small, minimally calcified foreign bodies that are often obscured by overlying tissues. CT can also visualize inflammatory changes in the adjacent structures.70,873

*

FIG 50-26  Sliding hernia. The stomach (asterisk) and esophagogastric

Hiatal Hernia

junction (arrow) are located in the mediastinum. Note that the esophagogastric junction is above the diaphragm, which is a crucial criterion in differentiating sliding hernia from paraesophageal hernia.

Hiatal hernia is caused by a weakness or tear of the phrenoesophageal membrane. It can be divided into two types:

St

A

1623

B FIG 50-27  Paraesophageal hernia. The gastric fundus (St) has migrated upward into the mediastinum (A), whereas the esophagogastric junction (arrow) is normally located below the diaphragm (B).

1624

PART II  CT and MR Imaging of the Whole Body

of the herniated stomach within the lower thoracic cavity, as well as the location of the GEJ.124,470,860 In some instances, only omental fat can be herniated into the chest while the stomach remains in the abdomen; this is referred to as a paraesophageal omental hernia. On CT and MRI this situation may mimic a mass containing fat in the mediastinum. Characteristic CT and MRI features of paraesophageal omental hernia include a bilobed retrocardiac fatty mass with branching blood vessels throughout, a midline septum, and extension through the diaphragm (Fig. 50-28).467,742

Diverticula Diverticula may be classified as pulsion and traction types. A pulsion diverticulum, the most common form, is caused by increased intraluminal pressure related to motility disorders; a traction diverticulum is caused by fibrosis in adjacent periesophageal tissues. Pulsion diverticula are further classified as a Zenker’s diverticulum in the proximal esophagus (Fig. 50-29) and as an epiphrenic diverticulum in the distal esophagus (Fig. 50-30). Traction diverticula are usually located in the midesophagus817 and can be readily diagnosed on esophagography. Pulsion diverticula show one or more rounded outpouching barium collections from the esophagus, whereas traction diverticula are manifested as a tented or triangular barium collection caused by scarring and retraction by granulomatous disease in adjacent subcarinal or hilar lymph nodes.817 Epiphrenic diverticula are typically demonstrated by CT as a thinwalled air-filled or air-fluid–filled structure communicating with the esophagus (see Fig. 50-30). However, those not associated with a distal esophageal obstruction (stricture, achalasia) may remain contracted in a resting state and thus may not be visible. Occasionally the diverticulum may have to be differentiated from mediastinal abscess or tumors and even hiatal hernia.402

STOMACH NORMAL ANATOMY FIG 50-28  Paraesophageal omental hernia. Omental fat (arrows) has herniated into the mediastinum, displacing the esophagus (curved arrow). Note the blood vessels within the lesion (arrowheads), which indicates that the fatty mass is the herniated omental fat.

Although the size and shape of the stomach varies, it can be conventionally divided into several areas based on its relationship with adjacent structures or the types of mucosal cells present. 1. The cardia is the area surrounding the GEJ. 2. The fundus is the portion of the stomach lying above a horizontal line drawn from the GEJ to the greater curvature.

B

A FIG 50-29  Zenker’s diverticulum. A, Esophagography demonstrates a barium-filled outpouching lesion from the esophagus (arrows). B, CT scan shows air-filled structures (arrows) anterior to the esophagus.

CHAPTER 50  Gastrointestinal Tract

A

1625

B

FIG 50-30  Epiphrenic diverticulum. A, CT scan demonstrates a thin-walled air-fluid–filled structure (arrows) from the laterally displaced distal esophagus (arrowheads). B, Esophagography shows a barium-filled large outpouching lesion (arrows) from the esophagus.

3. The body of the stomach is located between the fundus and a point on the lesser curvature designated as the incisura angularis. This area has abundant acid-secreting parietal cells. 4. The antrum begins where the gastric body ends and extends to the pylorus. This area contains parietal cells as well as G cells, which make gastrin, and D cells, which make somatostatin. 5. The pylorus is the transition point from the antrum to the duodenum. These designated areas of the stomach cannot always be well demonstrated on axial CT or MRI, so coronal or sagittal imaging planes are very helpful to understand the exact anatomic location of the lesion.

S

Gastric Wall In evaluating the gastric wall, the most commonly encountered difficulty is differentiation of pathologic thickening from spurious gastric wall thickening produced by redundant mucosa. In general the normal gastric wall is 2 to 5 mm thick, with 10 mm being the upper limit for normal266; in an air-filled stomach a 3-mm740 or 5-mm170 wall thickness is considered the upper limit of normal. In a suboptimally distended stomach, false-positive gastric wall thickening can frequently occur. In this circumstance, further studies with position change of the patient or additional administration of oral contrast media can be helpful to avoid misdiagnosis. When air techniques are used, the right decubitus position is selected to evaluate the proximal stomach and GEJ, and the left decubitus position is best to evaluate the distal stomach and duodenum; this is the reverse of the patient’s position when barium or water is used.421,843 Recent studies have reported that the left posterior oblique position is recommended for optimal distention of the gastric antrum, and prone position is necessary for distention of the gastric fundus.421 On CT or MRI, the normal gastric wall enhances homogeneously but shows a two- or three-layered structure (Figs. 50-31 and 50-32).126 The inner layer, corresponding to the mucosal layer, enhances markedly. The intermediate layer of low attenuation corresponds to the submucosal layer, and the occasionally seen outer layer of slightly higher attenuation corresponds to the muscular-serosal layer. This multilayered mural pattern, however, is difficult to appreciate when the wall measures less than 5 mm thick.331

FIG 50-31  Normal gastric wall of the stomach (S) with layered appearance.

S

S

FIG 50-32  Prone position with the water-filling method of the stomach (S) for evaluation of a patient with gastric cancer.

1626

PART II  CT and MR Imaging of the Whole Body

Perigastric Ligaments The perigastric ligaments contain major vascular structures and lymph nodes that are important pathways for lymphatic metastasis and direct extension of gastric pathologic processes. The gastrohepatic ligament is located in the most superior portion of the lesser omentum, containing the left gastric artery and coronary vein and lymph nodes. Any rounded structure larger than 8 mm within this ligament should suggest lymphadenopathy or collateral vessels. The gastrosplenic and gastrocolic ligaments contain the gastroepiploic vessels and lymph nodes that drain the greater curvature.

Vessels and Nodes The stomach is supplied by several vessels, including the celiac axis, hepatic artery, and splenic artery. The lymphatic vessels drain from the following major areas: 1. Cardia and most of the lesser curvature into left gastric nodes 2. Pylorus and the distal lesser curvature into the right gastric and hepatic nodes 3. Proximal portion of the greater curvature into the pancreaticosplenic nodes in the splenic hilum 4. Distal portion of the greater curvature into the right gastroepiploic nodes in the greater omentum and the pyloric nodes at the head of the pancreas Lymph nodes are classified on the basis of anatomic nodal location proposed by the Japanese Research Society for Gastric Cancer382,616 (Table 50-2).

CT GASTROGRAPHY With the advent of MDCT, new scanning protocols using thinner slices (<5 mm thick), position change of the patient, drug-induced hypotonia, and three-dimensional (3D) reconstruction as postprocessing are actively used in CT evaluation of gastric diseases.185,331,751 For virtual gastroscopic imaging and 3D volume-rendered images, gaseous gastric distention with effervescent granules is essential (Fig. 50-33).417 The patient should fast for at least 6 hours before CT gastrography. Adequate gastric distention is the most important prerequisite for successful CT gastrography; a collapsed stomach may result in mimicking,

TABLE 50-2  Numbering of Upper

obscuring, or missing small gastric pathologies. To distend the stomach, various oral contrast agents (e.g., diluted barium, water) have been used; recently, air has been most commonly used for this purpose. In general, 6 g of effervescent granules with a small amount of water may be sufficient. Drug-induced hypotonia of the stomach is also important to reduce the gastric peristaltic movement and maintain the gastric distention during CT scanning. A scanogram is obtained to ensure adequate gastric distention; if gastric distention is insufficient, additional effervescent granules are given. Thereafter the patient receives IV contrast materials, and CT images are obtained at about 72 seconds after contrast infusion, first with the patient in supine position. After completing CT scanning from diaphragm to pelvis, the patient is repositioned in prone position and additional scanogram and scanning is confined to the stomach; in most cases, additional administration of effervescent agents is not required at this positional change. Initial CT scanning with the patient in prone position is not recommended owing to a diminution of the amount of gastric air during positional change. CT images of the patient’s prone position are included at the end of the study because the prone position has the merit of adequate distention of the gastric fundus and high body,  but it is not appropriate for evaluating the gastric antrum; in this position, the gastric antrum is collapsed and filled with retained fluid or contents. Because of the popularity of gastric fiberoptic endoscopy, CT gastrography is not used as a screening tool for gastric cancer. CT gastrography seems to have nearly the same capability as upper GI series in depicting advanced and early gastric cancers (especially for depressed and protruded lesions). As with upper GI series, CT gastrography has trouble revealing both flat and slightly elevated or depressed lesions, especially against a background of mucosal surface nodularity. On occasion, some of the early gastric cancers that are very difficult or impossible to detect on upper GI series can be easily detectable on volume-rendering images. Therefore it seems likely that CT gastrography has a potential to replace upper GI series for detecting gastric pathologies. However, in several series the reported accuracy of CT gastrography in detecting gastric cancers may be problematic because the patient population that underwent CT gastrography was mostly restricted to those with already known gastric cancer.185,331,751 Further investigations are required for clarifying the advantages and disadvantages of CT gastrography in comparison with upper GI series and gastroscopy.

PATHOLOGIC CONDITIONS

Abdominal Lymph Node Stations

Benign Gastric Tumors

Station Number

Gastric Polyp and Adenoma.  Epithelial polyps of the stomach may Location

1/2 3/4 5 6 7 8 9 10 11 12 13 14 15 16

Right/left side of the cardia Along the lesser/greater curvature Suprapyloric, along the right gastric artery Infrapyloric Along the left gastric artery Along the common hepatic artery Around the celiac artery At the splenic hilus Along the splenic artery In the hepatoduodenal ligament Behind the pancreas head At the root of the mesentery (superior mesenteric artery) Along the middle colic artery Paraaortic

be hyperplastic or adenomatous. Hyperplastic polyps are not true neoplasms. They are believed to develop from excessive regeneration of superficial epithelium. Although most of these nonneoplastic polyps are often multiple and smaller than 1.5 cm in diameter, 9% are larger than 2 cm.588 In contrast, 80% of adenomatous polyps (adenoma) exceed that size.588 CT may detect epithelial polyps that are larger than 2 cm in diameter when the stomach is well distended with air or when oral contrast material has been given (Fig. 50-34). No specific CT features that can distinguish these polyps from gastric malignancy have been reported. Rarely, pedunculated polyps may prolapse into the duodenum or cause intussusception (Fig. 50-35).299

Lipoma.  Gastric lipoma is a rare submucosal tumor that usually occurs in the antrum. Large lipomas ulcerate and produce clinically significant hemorrhage.530 The CT finding of gastric mass having a CT

CHAPTER 50  Gastrointestinal Tract attenuation equivalent to fat is pathognomonic of a gastric lipoma (Fig. 50-36).

Hemangioma.  Gastric hemangiomas are rare neoplasms, constituting less than 2% of all benign tumors of the stomach. Pathologically, hemangiomas can be cavernous, capillary, or mixed. The gross appearance varies according to the size and density of the plexus of vessels forming the lesion, the degree of arteriovenous shunting, and the presence of thrombosis of fibrosis. Cavernous hemangiomas may be single or multiple, localized, or diffuse.398 Gastric hemangiomas are smooth submucosal masses that are radiographically indistinguishable from stromal tumors. Although rare, the presence of phleboliths, which are most often associated  with the cavernous types, help in establishing an accurate diagnosis  on CT.788

1627

Malignant Gastric Tumors

Adenocarcinoma.  Gastric carcinoma is one of the leading causes of cancer mortality worldwide. Its incidence varies widely, being higher in Asian than in Western countries. Many factors affect prognosis in patients with gastric cancer, including age, symptoms, tumor size, lymph node involvement, depth of tumor invasion, distant metastasis, histopathologic type, and surgery performed. These tumors may be detected either in an early stage (in which case the mucosa or the submucosa is involved) or in an advanced stage. The gross visual appearance of advanced carcinoma is polypoid, fungating, ulcerated, or infiltrative, whereas that of early carcinoma is protruded, flat, or depressed. Vascular structures in patients with gastric cancer have been examined with angiography, microangiography, or both.791 In these studies,

A

B

C

D FIG 50-33  CT gastrography of multiple advanced gastric cancers of the gastric body (A, C, E, and G) and antrum (B, D, F, and H). Axial CT images of the air-distended stomach show cancerous lesions (arrows) at the gastric body and antrum. Morphologic features of these tumors can be more clearly assessed by means of various postprocessing imaging reconstructions such as 2D reconstructions in the coronal (C) or oblique coronal (D) planes. Other postprocessing imaging includes 3D reconstructions such as surface-shaded mode (E), transparent volume rendering (F), and virtual endoscopy (G and H). Continued

1628

PART II  CT and MR Imaging of the Whole Body

E

F

G

H FIG 50-34, cont’d

M

FIG 50-34  Gastric adenoma with malignant transformation. CT scan shows a polypoid mass (M) with a short thick stalk.

CHAPTER 50  Gastrointestinal Tract

A

B

C FIG 50-35.  Gastroduodenal intussusception due to prolapsed gastric adenoma. A, Radiograph shows a large polypoid mass (arrowheads) in the antrum of the stomach. B and C, CT scans obtained 1 month later after an upper GI series show a large soft tissue mass (arrows in B) in the duodenum and beaklike narrowing (arrows in C) of the distal stomach. (From Ha HK, et al: Gastroduodenal intussusception due to a prolapsed gastric adenoma. AJR Am J Roentgenol 159:432, 1992.)

1629

1630

PART II  CT and MR Imaging of the Whole Body

most gastric cancers showed neovascularity in the arterial and capillary phases and tumor stain in the venous phase of angiography. Neoplastic vessels became more numerous, more tortuous, and wider than normal vessels. Tumor detection.  Complete resection of a gastric tumor and involved lymph nodes is the only potentially curative treatment. Therefore early detection and accurate tumor staging are critical in deciding the course of treatment and establishing the prognosis of gastric cancer patients. CT scan is the core of the staging process, although details of technique vary. In early gastric cancer, CT sensitivity in tumor detection depends on the examination technique as well as on the type and size of the lesion. On conventional CT scans, gastric wall thickening in early cancer appears minimal, usually smaller than 1 cm. Furthermore a flat or depressed type of early cancer may not be demonstrated by CT. Hence the reported detection rate of early gastric cancer on CT has been disappointing (≈50%).185,245,331 However, use of dynamic helical CT with water ingestion improves cancer detection. On dynamic CT scan, most gastric cancers show moderate or marked heterogeneous enhancement in the early phase and homogeneous enhancement in the equilibrium phase. Although it is still dif-

M

FIG 50-36  Gastric lipoma. A fatty mass (M) is seen in the antrum.

A

ficult to detect lesions less than 2 cm in diameter, most early cancers detected on CT show focal gastric wall thickening with strong enhancement in the most inner (mucosal) layer, especially in the arterialdominant phase.185 The presence of a low-attenuation stripe in the middle layer differentiates early from advanced tumors.476 In contrast with early gastric cancer, the sensitivity of CT in detecting advanced gastric cancer is high (92%-96%).126,331,587 Nevertheless, gastric cancer situated on the horizontally oriented portion of the gastric wall, such as the greater curvature and gastric angle, is difficult to evaluate because of the partial-volume averaging effect.126,331 This limitation of conventional axial CT scan can be overcome by using MPR.466 Common CT features of advanced gastric cancer include a polypoid mass with or without ulceration or an eccentric or circumferential gastric wall thickening with enhancement (Fig. 50-37). Because of marked desmoplastic response, mainly in the submucosal layer of the gastric wall, the thickened wall usually shows homogeneous and prolonged contrast enhancement, especially in the equilibrium phase.41 In contrast with the submucosal layer, the muscularis propria shows edematous thickening in the absence of fibrous connective tissues and thus appears to be hypoattenuated on CT, occasionally causing a target appearance in the thickened gastric wall. A less common variety is a scirrhous type of gastric carcinoma in which the tumor cells (most commonly, the signet ring cell type) spread predominantly in the submucosa, with no or minimal involvement of the mucosal layer (Fig. 50-38). In these cases, CT appears to be more diagnostic than endoscopy or barium study and shows diffuse circumferential thickening of the gastric wall, with a mean thickness of 1.8 cm (range, 1-3 cm).41 Other CT features diagnostic for scirrhous cancer include preservation of a thin rim of hypoattenuated mucosal line, intramural calcifications, and high-grade gastric outlet obstruction. The differential diagnosis of scirrhous carcinoma includes scirrhous metastases (lung, breast carcinoma), non-Hodgkin’s or Hodgkin’s lymphoma, and unusual chronic systemic or inflammatory conditions such as sarcoidosis, corrosive gastritis, amyloidosis, and Crohn’s disease.469 Calcified gastric carcinomas are rarely reported in the literature, and most of them are caused by mucinous adenocarcinoma (Fig. 50-39).348 These calcifications are nodular, miliary, punctate, or homogeneously distributed in the thickened gastric wall and are also noted in metastatic lesions, such as lymph nodes and liver.348 In certain instances of gastric carcinoma, the tumor grows exophytically, thereby simulating a stromal tumor (Fig. 50-40). The

B FIG 50-37  Gastric adenocarcinoma. Early phase (A) of a dynamic CT scan shows mild enhancement of thickened gastric wall (arrow). Scan shows good enhancement in the equilibrium phase (arrow) (B).

1631

CHAPTER 50  Gastrointestinal Tract

M

FIG 50-40  Gastric adenocarcinoma with exophytic tumor growth. CT FIG 50-38  Scirrhous adenocarcinoma of the stomach. CT scan shows considerable gastric wall thickening (arrows) with a markedly enhanced submucosal layer and a poorly enhanced thin rim of mucosal layer. Ascites is also seen.

scan shows a heterogeneous mass (M) in the left upper quadrant, simulating an exophytically growing malignant GIST. Enlarged perigastric lymph nodes (arrows) are seen along the lesser curvature of the stomach.

TABLE 50-3  International Staging System

for Adenocarcinoma Stage

FIG 50-39  Mucinous adenocarcinoma of the stomach. CT scan shows gastric wall thickening (arrows) with poor contrast enhancement and calcifications.

presence of gastric wall thickening adjacent to an exogastric mass or a gastric outlet obstruction favors gastric cancer, allowing it to be distinguished from gastric stromal tumors.469

Tumor staging

Tumor-node-metastasis classification.  As modern treatment for gastric cancer has become diverse, accurate staging becomes more essential in planning management. In general the TNM classification according to the AJCC (Table 50-3) is used.278 Primary tumor.  According to previous reports126,542,929 the overall accuracy of CT in preoperative tumor staging of gastric cancer ranges from 69% to 85%. The detectability of early cancer is very low, with a rate of 20% to 53%,245 such that in most studies the absence of abnormal findings on CT images has resulted in classification as T1. With the introduction of fast scanning by spiral CT, gastric water filling, prone position, and gastric hypotonia, the tumor detection rate has markedly increased to as high as 80% to 88%.163,410,479 Recently Kim et al.420 and

Description

Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of primary tumor Tis Carcinoma in situ: intraepithelial tumor without invasion of the lamina propria T1 Tumor invades lamina propria, muscularis mucosae, or submucosa T1a Tumor invades lamina propria or muscularis mucosae T1b Tumor invades submucosa T2 Tumor invades muscularis propria T3 Tumor penetrates subserosal connective tissue without invasion of visceral peritoneum or adjacent structures T4 Tumor invades serosa (visceral peritoneum) or adjacent structures T4a Tumor invades serosa (visceral peritoneum) T4b Tumor invades adjacent structures Regional Lymph Nodes (N) NX Regional lymph node(s) cannot be assessed N0 No regional lymph node metastasis N1 Metastasis in 1-2 regional lymph nodes N2 Metastasis in 3-6 regional lymph nodes N3 Metastasis in seven or more regional lymph nodes N3a Metastasis in 7-15 regional lymph nodes N3b Metastasis in 16 or more regional lymph nodes Distant Metastasis (M) M0 No distant metastasis M1 Distant metastasis From Edge S et al (eds), American Joint Committee on Cancer: AJCC cancer staging manual, ed 7. New York, 2011, Springer-Verlag, p 120.

1632

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 50-41  Early gastric cancer at gastric angle. A, Axial CT scan shows no definite pathologic lesion at gastric angle (arrow). B, Focal wall thickening with subtle enhancement of inner layer (arrow) is noted at gastric angle on coronal reconstructed image. This lesion (arrow) is well demonstrated on virtual endoscopic image (C) as early gastric cancer, type IIa.

Chen et al.121 reported highly improved tumor staging of gastric cancer with 3D CT using MPR and virtual gastroscopy techniques (detection rate of early cancer, 77%-78%; overall accuracy of tumor staging, 84%89%) (Fig. 50-41). However, the efficacy of CT in differentiation between T1 and T2 and between T3 and T4 is still controversial.7,163,685,868 These differentiations are important clinical decision points (T1 vs. T2 determines the candidates of endoscopic mucosal resection or laparoscopic surgery, while T3 vs. T4 decides the operability). Hence additional diagnostic tools such as EUS or diagnostic laparoscopy can  be used. The prognosis in surgically treated patients with gastric cancer is significantly affected by the presence or absence of intraperitoneal free cancer cells at the time of gastrectomy. According to a series by Kaibara and colleagues, who analyzed the relationship between the spatial extent of invasion of the gastric serosa and the 5-year survival rate, intraperitoneal free cancer cells were detected by lavage of the Douglas cavity in 135 (44%) of 309 patients with gross evidence of serosal invasion. Their study also showed that the 5-year survival rate worsened in patients with wider areas of serosal invasion.381 The accuracy of CT for

predicting serosal invasion ranges between 60% and 83%.126,185,245 In most series, understaging was more common than overstaging, probably as a result of CT’s limited ability to detect microscopic invasion. Lymph node metastasis.  Besides depth of invasion, extent of node metastases is one of the most important prognostic factors in gastric cancer without distant metastases.551 Under the new AJCC classification system, N staging is based on the number of positive nodes (see Table 50-3).278 This system differs from previous anatomic nodal classification following Japanese guidelines. Several studies have confirmed the superiority of using the number of positive nodes in estimating prognosis,441,627,743,906 but anatomic nodal location still remains a valuable criterion in description of the extent of lymphadenectomy.616 Common findings of metastatic nodes include larger size (>810 mm along the short axis), nearly round shape, central necrosis, and marked or heterogeneous enhancement.160,357,383 Until recently, CT diagnosis of lymph node metastasis had been based on size criteria. Most studies of GI tract tumors consider lymph nodes 10 mm or larger to have been infiltrated by metastases of solid tumors.126,244 According to this criterion, the sensitivity of CT in nodal staging of gastric cancer

CHAPTER 50  Gastrointestinal Tract has been reported to be variable (range, 48%-91%).80,126,823 Understaging is the main drawback of CT, because only lymph nodes with a diameter larger than 5 mm are visible, and metastases in smaller nodes cannot be detected. Furthermore, the results of two reports185,595 appear to be pessimistic with regard to the nodal staging of CT. In a study by Mönig and associates, who analyzed regional lymph nodes from 31 gastrectomy specimens of patients with gastric cancer, 55% of the metastatic lymph nodes were 5 mm or smaller in diameter, and 80% of tumor-free lymph nodes were smaller than 5 mm in diameter.595 In another study by Düx and colleagues, most metastatic lymph nodes were between 2 and 10 mm in diameter, and helical CT depicted 74% of histologic N0 stages, 45% of N1 stages, and only 32% of N2 stages.185 Therefore it may be true that lymph node size is not a reliable indicator for lymph node metastasis; accordingly, CT is not an effective modality for nodal staging of gastric cancer. Distant metastasis.  The major methods of spread of cancer are by local invasion through contiguity and tissue planes, by the lymphatic system, by the vascular system, and by implantation.135 In these ways gastric cancer spreads to the liver, lymph nodes, pancreas, duodenum, transverse mesocolon, peritoneal cavity, and ovary. In clinical practice, one of the major questions facing surgeons treating patients with gastric cancer who do not appear to have distant metastasis according to CT is whether there is pancreatic invasion, because its presence precludes curative surgery. On CT, pancreatic invasion is suggested when the low-attenuation band of fat between the tumor and the pancreas is absent (Fig. 50-42) or when the gastric tumor abuts the pancreas. When either of these criteria were applied, the sensitivity of CT in detecting pancreatic invasion was poor (27%), as seen in a series by Sussman’s group823; in this series, eight patients with intact fat planes had extension into the pancreas, and five of another eight patients who were presumed to have pancreatic invasion on CT were eventually shown to have no tumor invasion. A series by Cook and associates reported a sensitivity of 60%.137 Furthermore, coexisting acute or chronic pancreatitis or associated inflammatory response also causes obliteration of the peripancreatic fat plane.160 Therefore application of this finding alone seems to have a significant limitation in determining pancreatic invasion, and more strict criteria may be required.

The diagnostic value of the CT criterion of obliterated fat plane is increased when it is associated with prominent stranding in the perigastric fat plane as well as irregularity along the outer border of gastric wall involved by the tumor. Although additional studies with patients in a decubitus position are time consuming, they may be helpful in determining focal or diffuse fixation between the pancreas and stomach. In certain instances, gastric cancer spreads into the duodenum through the pyloric canal (Fig. 50-43). The incidence varies greatly in surgical and postmortem series (range, 9%-69%).383 However, the incidence of radiographically detectable transpyloric spread has been reported as only 5% to 25%.615 Such spread occurs by direct infiltration through the submucosa or subserosa or lymphatic transport through the submucosal layer.383 The 5-year survival rate in these patients is poor compared with patients whose tumor is confined to the antrum.383 Rarely, portal venous thrombosis occurs in approximately 0.95%  of patients with gastric cancer.357 The thrombosis develops as a

m

D

FIG 50-43  Intramural duodenal spread from gastric cancer. CT scan shows a large polypoid mass (m) in the stomach extending into the second portion of the duodenum (D).

n

A

1633

N

B FIG 50-42  Gastric cancer with pancreatic invasion. A, CT scan shows obliterated interface (arrow) between the tumor in the stomach and pancreas, along with linear strands. Enlarged perigastric nodes (n) are also seen. B, Axial T1-weighted MRI at a lower level shows obliteration of the fat plane between the enlarged perigastric node (N) and the uncinate process of the pancreas, along with diffuse stranding. (From Sohn KM, et al: Comparing MR imaging and CT in the staging of gastric carcinoma. AJR Am J Roentgenol 174:1551– 1557, 2000.)

1634

PART II  CT and MR Imaging of the Whole Body

M

P

*

M

B

A

FIG 50-44  Hepatoid type of gastric adenocarcinoma. A, CT scan shows concentric gastric wall thickening (M) involving the antrum of the stomach. B, CT scan at a higher level shows a portal vein thrombus (P) and a metastatic lesion (asterisk) in the caudate lobe of the liver.

consequence of splenic vein involvement by the gastric tumor, and a high incidence of liver metastasis is reported in these patients.357 Hepatoid adenocarcinoma, an adenocarcinoma originating primarily from the gastric mucosa and containing wide and distinctive foci of hepatocellular differentiation, tends to invade portal or hepatic veins and to metastasize to the liver; the prognosis is poor (Fig. 50-44).358 This tumor usually produces a large amount of serum α-fetoprotein (AFP), as high as the level released by hepatocellular cell carcinoma,358 although one report showed that not all AFP-producing adenocarcinomas have hepatoid features.606 MRI staging for gastric cancer.  Until recently, MRI had not been popular for staging gastric cancer. Its main drawbacks were its high cost and lack of effectiveness in comparison with CT. However, the recent development of fast MRI techniques allows an examination during breath holding for several seconds, thus providing high-quality images of the stomach (see Fig. 50-42). With the use of spoiled gradient-recalled acquisition in the steady state (GRASS) and the water-filling method, Matsushita and colleagues, who analyzed 48 patients with gastric cancer, reported an overall accuracy rate of 88% in predicting serosal invasion by observing the changes in a low-signalintensity band surrounding the gastric lesion.555 A later article by Sohn and associates, who compared CT with MRI for 30 patients with gastric cancer, showed that MRI was comparable with helical CT in T and N staging; the accuracies of MRI and CT were 73.3% and 66.7% for T staging and 55% and 58.6% for N staging, respectively.806 Although further investigation is necessary, MRI appears to be an alternative modality of choice for tumor staging in patients with gastric cancer. PET imaging of gastric adenocarcinoma.  FDG PET is a wellestablished method for detecting and staging a variety of malignancies, including lung, colon, and esophageal cancer.10,248,660 Despite the clinical importance and high frequency of gastric cancers, FDG PET has only rarely been used as a diagnostic modality for gastric cancer, and the results of these trials have not been satisfactory in the diagnosis of primary gastric cancers or metastatic nodes.168,445,489,559,593,812 In tumor detection, FDG PET shows variable uptake levels. Some gastric cancers such as mucinous, signet ring cell, and poorly differentiated adenocarcinomas tend to show significantly lower FDG uptake than do other histologic types of gastric cancer.812,913 Results of preoperative T staging by FDG PET were very disappointing.600 Detection rates of tumor less than 3 cm in size and early cancer were 16.8% and

25.9%, respectively. Therefore FDG PET is not helpful in T staging of gastric cancer. In N staging, FDG PET shows lower sensitivity (34%-40%) and higher specificity (95%-98%) than those of CT.119,414,593,600,919 Poor spatial resolution of FDG PET seems to prevent discrimination of nodes from the primary tumor (Fig. 50-45). In addition, FDG uptake of metastatic node is proportional to that of primary tumor. Contrary to lower sensitivity, higher specificity of FDG PET can detect more metastatic nodes less than 8 mm in size (not detectable at CT) in 10% of gastric cancers and can change the extent of lymphadenectomy or may preclude unnecessary surgery.919 There is no comparison study between PET and CT or whole-body bone scan in detection of distant metastasis to either lung or bone. Current evidence suggests that CT PET has a sensitivity of 98% to 100% and a specificity of up to 89% in detecting lung and liver metastasis in GI tumors,5,608,717,928 which are comparable results to CT’s sensitivity of 90% and specificity of 98%.5 If primary tumor shows poor FDG uptakes, however, its yield in detecting distant metastasis would be poor. In the assessment of peritoneal tumor seeding, FDG PET is also helpful in their detection (Fig. 50-46), but it seems to be less effective than CT owing to low spatial resolution.123,638 Several recent reports have demonstrated that response to preoperative chemotherapy can be predicted with FDG PET early in the course of therapy.636,875 Therefore it is expected that FDG PET may a play a valuable role in monitoring response of therapy in patients who undergo surgery or chemotherapy. Local recurrence of gastric adenocarcinoma.  Local recurrence of gastric carcinoma after surgery is defined as histologic evidence of a tumor in the surrounding tissues of the resected stomach (i.e., gastric remnant, esophageal anastomosis, pancreas, duodenal stump, hepatic pedicle, splenic hilum, celiac axis area, abdominal wound).646 According to autopsy studies, 80% of the patients who underwent gastrectomy for gastric cancer eventually experienced local recurrence with or without distant metastasis. However, only 19% of all patients with recurrence had resectable lesions on reexploration; almost 50% of stump recurrences were resectable, whereas only 2 of 34 patients with recurrence at an extragastric site had resectable lesions.646 Tumor recurrence at the gastric remnant may occur more commonly in patients with early-stage gastric cancers646,647 very late (>2 years) after surgery.360 Also, gastric cancers arising in the cardia were prone to anastomotic recurrence.647 Because barium studies are

CHAPTER 50  Gastrointestinal Tract

1635

N

A

B FIG 50-45  Advanced gastric adenocarcinoma. A, CT scan shows eccentric gastric mass (arrows) with extensive metastatic lymphadenopathy (N). B, FDG PET shows extensive extent of gastric tumor, without discrimination of metastatic lymph node from primary tumor.

B

A

FIG 50-46  Advanced gastric adenocarcinoma with peritoneal seeding. A, CT scan of the pelvis shows suspicious peritoneal seeding nodule at rectovesical pouch (arrows). B, This lesion (arrows) shows strong hot uptake at FDG PET scan, indicating the presence of definite malignant seeding nodule.

associated with difficulty in their capability to fully assess the areas mentioned earlier, CT plays an important role in determining the extent of local tumor recurrence or distal metastasis. On CT, recurrence at the gastric stump or anastomosis appears as a localized wall thickening or mass (Figs. 50-47 and 50-48). Afferent loop syndrome may be caused by development of the recurrent tumor near the orifice of the afferent loop (Fig. 50-49). However, potential sources of erroneous interpretation include improperly opacified bowel loop, food, surgical plication, gastritis, and gastritis cystica polyposa.291 The most frequently involved lymph nodes are those along the common hepatic artery, the celiac axis, and the hepatoduodenal ligament.291 The presence of lymphadenopathy along the hepatoduodenal ligament may cause moderate to severe biliary tract obstruction. When the pancreas is involved, CT features are indistinguishable from those of primary pancreatic tumor, and enlarged peripancreatic lymph nodes can mimic tumor recurrence in the pancreas. Most tumor recurrences at the abdominal incision are caused by iatrogenic dissemination of cancer cells during surgery (Fig. 50-50).137 However, postoperative fibrosis, hematoma, abscess, or granuloma

G

FIG 50-47  Local recurrence of gastric adenocarcinoma at the anastomotic site following subtotal gastrectomy. The bowel wall (arrows) at the surgical anastomosis is thickened and markedly enhanced along with marked distention of the remnant stomach (G).

1636

PART II  CT and MR Imaging of the Whole Body

M

FIG 50-50  Tumor recurrence at the incisional wound after surgery for FIG 50-48  Tumor recurrence at the surgical plication. CT scan shows prominent surgical plication (arrowheads). Surgery confirmed tumor infiltration at the plication. (From Ha HK, et al: Local recurrence after surgery for gastric carcinoma: CT findings. AJR Am J Roentgenol 161:975–977, 1993.)

gastric cancer. CT scan shows a lobulated well-enhancing mass (M) along the incisional wound, as well as a large amount of ascites and peritoneal thickening due to peritoneal carcinomatosis. (From Ha HK, et al: Local recurrence after surgery for gastric carcinoma: CT findings. AJR Am J Roentgenol 161:975–977, 1993.)

A A M

FIG 50-49  Afferent loop syndrome due to obstruction at the proximal

M

FIG 50-51  Gastric lymphoma. Gastric wall is markedly thickened (M)

afferent loop (A) by recurrent tumor (not shown).

with predominant involvement of the submucosa. The mucosal layer is well preserved.

cannot be differentiated from tumor recurrence at this site by imaging findings alone. Because tumor recurrences were most common within 2 years after surgery, follow-up CT scanning in a short interval is required at this period for early diagnosis.

Megibow’s group, gastric wall involvement can be divided into three patterns, as seen with CT570: 1. Diffuse infiltration involving more than 50% of the length of the stomach 2. Segmental infiltration 3. The localized polypoid form Gastric wall thickening is most commonly circumferential, with thickness greater than 4 cm. Thickening and distortion of the gastric rugal folds may be present, but outlet obstruction is rare (Fig. 50-51).100,570 Dynamic CT and MRI clearly depict submucosal spread of the tumor within the normally enhanced mucosal layer (Fig. 50-52).587 Lymphomatous tissues in the thickened gastric wall usually show poor and homogeneous contrast enhancement; this pattern may be due to lack of desmoplastic response in lymphoma, in contrast with gastric adenocarcinoma. Sometimes gastric lymphoma shows heterogeneous enhancement. In such cases the hypoattenuated areas may be caused by pathologic evidence of necrosis and hemorrhage.319 The localized polypoid form

Lymphoma.  The stomach is the most frequent site of lymphomatous involvement of the GI tract. Non-Hodgkin’s lymphoma accounts for approximately 80% of cases. Gastric involvement may be isolated but more commonly is part of a generalized disease process. Among various prognostic factors that include gender, age, size and location of tumor, histologic type, depth of invasion, and nodal status, depth of invasion and regional lymph node involvement play a decisive role in survival.88 Unlike carcinoma, lymphoma tends to spread laterally within the submucosal plane and spares the muscular coats until late in its course. With its ability to directly image the entire gastric wall and adjacent structures, CT has proved capable of determining the extent of spread of gastric lymphoma with a high degree of accuracy.100 According to

CHAPTER 50  Gastrointestinal Tract

1637

M M

A

B

M

M

C

D FIG 50-52  Gastric lymphoma. A, A mass (M) is seen in the gastric wall along the lesser curvature of the stomach on CT scan. B, T2-weighted MRI shows a mass (M) with intermediate signal intensity. C and D, Dynamic Gd-enhanced MRI shows poor contrast enhancement of the mass (M) in the early phase (C) and isointensity to that of the liver in the delayed phase (D).

is often associated with an ulcer. The perigastric fat plane is usually preserved. Most patients with gastric lymphoma have perigastric lymphadenopathy, and the presence of widespread retroperitoneal lymphadenopathy indicates systemic disease. Perforation is an important complication affecting therapy and prognosis. It may occur in 9% to 47% of patients as a result of the lack of desmoplastic response, rapid growth of the tumor, or a response to therapy. Tumor extension into the esophagus is present in about 10% of patients, and spread across the pylorus into the duodenal bulb has been noted in approximately one third of patients.339 Although not common, far-advanced cases of lymphoma can show diffuse peritoneal involvement with ascites, omental infiltration, and peritoneal implants, mimicking carcinomatosis (Fig. 50-53).403

Gastrointestinal Stromal Tumor.  GISTs are the most common nonepithelial neoplasms of the stomach and small bowel. Although GIST accounts for only 2.2% of malignant gastric tumors, most  GISTs are located at the stomach (60%-70%) and small intestine (20%-30%). GISTs are now thought to be derived from a precursor of the interstitial cells of Cajal and are clearly different from other mesenchymal tumors such as leiomyomas or leiomyosarcomas.617 Almost all GISTs express activating mutations in the KIT gene, which encodes transmembrane receptor tyrosine kinase in the interstitial cells of Cajal, and

result in promoting tumor growth. Therefore identification of KIT protein (CD117, stem cell factor receptor) is a key in confirming diagnosis of GISTs and improves clinical expectation of long-term survival by using imatinib (a targeted tyrosine kinase receptor blocker). Unlike leiomyoma, no GIST can truly be labeled benign.166 Even GISTs smaller than 2 cm, generally considered benign, show a very  low risk of recurrence. In addition to early detection of the tumor, therefore, monitoring for therapeutic response or tumor recurrence becomes more important. CT is the main imaging modality for these purposes. CT features of GISTs vary greatly depending on the size and aggressiveness of the tumor. Most primary GISTs are large, hypervascular, enhancing masses on contrast-enhanced CT; 90% are located in the fundus and body, and their growth is exogastric or dumbbell shaped.569 Dynamic CT may demonstrate their submucosal location by showing the intact mucosal layer enhancement (Fig. 50-54). Small localized GISTs are usually homogeneous, whereas larger tumors are heterogeneous on CT, with common occurrence of low-attenuated areas owing to internal hemorrhage, necrosis, or cystic changes (Fig. 50-55)133,707; the presence of hemorrhage produces a fluid-fluid level within the mass. Areas of necrosis can give the appearance of multiple thick septations within the tumor. Necrosis within the tumor mass may lead to excavation and communication with the GI tract or perforation into the peritoneal cavity.

1638

PART II  CT and MR Imaging of the Whole Body

FIG 50-53  Peritoneal lymphomatosis. CT scan shows omental cake

In one series,737 crescent-shaped necrosis due to necrosis in the periphery of the extraluminal component of the mass was reported to be suggestive of a stromal tumor on CT. Ulceration of the overlying mucosa can be demonstrated on CT as the presence of air or an oral contrast agent. Another cause of gas within the mass may be superinfection of the necrotic tumor (Fig. 50-56).768 Although not common, calcification is seen in stromal tumors (Fig. 50-57). Nearly 50% of patients with GISTs present with metastases615; the most common sites are the liver and peritoneum, followed by lung and bone (Fig. 50-58). Unlike in adenocarcinoma, lymphatic metastases or malignant ascites of GISTs are extremely rare. CT characteristics of metastatic lesions of GISTs are similar to those of primary GISTs. Histologic assessment of malignancy is essentially based on mitotic counts and size of the tumor. In general, tumors less than 5 cm and with fewer than 5 mitoses per 10 high-power fields (HPF) suggest lowgrade malignancy. However, tumors greater than 5 cm with higher than 5 mitoses/10 HPF are considered high-grade malignancy. CT

(arrows) and diffuse mesenteric infiltration.

m

FIG 50-56  Infected malignant GIST of the stomach. CT scan shows a FIG 50-54  Malignant GIST. CT scan shows a large heterogeneous mass (m) growing exophytically from the stomach (arrow).

large heterogeneous mass (arrows) with central cavitation and gas collection. The small orifice of the ulcer in the tumor was considered to be the cause of secondary infection.

M M

FIG 50-55  Malignant GIST. Huge exophytic growing mass (M) arising from the gastric high body is seen, with severe central necrosis. Note intact overlying mucosal layer enhancement (arrow).

FIG 50-57  Benign GIST of the stomach. A calcified mass (M) (arrow) is seen at the gastric high body.

CHAPTER 50  Gastrointestinal Tract

m

1639

m m

m

m

M

B

A

FIG 50-58  Malignant GIST of the stomach with distant metastases. CT scan shows hepatic metastasis (M) (A) and multiple peritoneal seeding masses (m) (B). Little lymphadenopathy is seen in the abdominal cavity.

A

B FIG 50-59  Metastasis to the stomach (linitis plastica) from breast cancer. A, Double-contrast upper GI series shows focal luminal narrowing (arrows), rigidity, and fold thickening in the proximal antrum and lower body of the stomach. B, CT scan shows involvement of the stomach by a scirrhous tumor (arrows).

features favoring malignancy of GISTs include tumor size larger than 5 cm, ulceration, heterogeneous enhancement with necrosis or hemorrhage, and distant metastases. During treatment with imatinib, therapeutic response is demonstrated by decreased tumor size, rapid transition from a heterogeneously hyperattenuating pattern to a homogeneously hypoattenuating pattern, resolution of the enhancing tumor nodules, and decreased intratumoral vessels. On the other hand, development of an enhancing nodule within the treated tumor suggests tumor recurrence regardless of changes in tumor size.20,128

Metastatic Tumor.  Hematogenous metastases may be discovered during abdominal CT performed for other reasons. Malignant melanoma and breast and lung carcinoma are the usual primary lesions. Metastasis from a lobular carcinoma of the breast produces a linitis plastica pattern of gastric wall thickening (Fig. 50-59).115 Squamous carcinoma of the esophagus reportedly involves the gastric cardia in up to 15% of patients and may simulate a primary mucosal or submucosal neoplasm.258 Carcinomas of the transverse colon and tail of the pancreas metastasize to the gastric wall by direct extension via the gastrocolic and splenorenal-gastrosplenic ligaments, respectively.

Liposarcoma.  Gastric liposarcoma originates in undifferentiated mesenchymal cells present within the submucosa and tunica muscularis of the gastric wall.459 Pathologically, four forms are recognized: (1) differentiated, (2) myxoid, (3) round cell, and (4) pleomorphic. Exuberant exophytic growth explains the lack of specific GI symptoms and delayed diagnosis. CT appearances correlate with the macroscopic appearance of the neoplasm. Fatty areas within the tumor are reported only in the differentiated form.516 Undifferentiated liposarcomas commonly contain areas of necrosis and hemorrhage along with prominent enhancement of the solid zones.213 Myxoid tumors present as a cystic form owing to the presence of gelatinous matrix.213

Plasmacytoma.  Malignant neoplasms of plasma cells (plasmacytoma) appear in two forms. The diffuse form includes multiple myeloma or plasma cell leukemia; the local form is a solitary myeloma of bone or an extramedullary plasmacytoma. Ten percent of cases of extramedullary plasmacytoma occur in the GI tract; the small bowel is most commonly involved, followed by the stomach, colon, and esophagus. Primary gastric plasmacytoma accounts for approximately 5% of all extramedullary plasmacytomas.

1640

PART II  CT and MR Imaging of the Whole Body

The diagnosis of primary gastric plasmacytoma is established by the absence of a serum M protein component; hence there is no Bence Jones proteinuria, no bone marrow infiltration, and no visceral involvement other than of the stomach.757 Pathologically, primary gastric plasmacytoma has been thought to originate from lymphoid follicles in the submucosa or from plasma cells in the lamina propria of the mucosa and can have various forms. Remingo and Klaum grossly classified gastric plasmacytoma into four types: (1) nodular, (2) infiltrative, (3) ulcerative, and (4) polypoid721; the nodular type was most commonly seen in their study. CT findings of gastric plasmacytoma simulate those of lymphoma or a mucinous type of gastric cancer owing to poor contrast enhancement of the gastric wall (Fig. 50-60).907

A

Deep endoscopic biopsy combined with immunohistochemical study of immunoglobulin is essential in diagnosis.780

Inflammation and Other Conditions

Peptic Ulcer Disease.  Although CT has no useful role in patients with uncomplicated peptic ulcer disease, it is effective in detecting its complications (e.g., acute free perforation, penetration). The ulcers are frequently identified on CT as round thin-walled collections of air or orally administered contrast medium projecting beyond the confines of the gastric or duodenal lumen. Although direct visualization of the ulcer may be difficult on CT in many instances, the involved gastric or duodenal wall is mostly thickened with considerable mural edema (Figs. 50-61 and 50-62). If free perforation is present, CT may show

B FIG 50-60  Gastric plasmacytoma. A, Upper GI radiograph shows large gastric mass (arrows) with a sharply marginated border. B, CT scan shows marked gastric wall thickening (arrows) with poor contrast enhancement, simulating lymphoma or adenocarcinoma. (From Yoon SE, et al: Upper gastrointestinal series and CT findings of primary gastric plasmacytoma: Report of two cases. AJR Am J Roentgenol 173:1266–1268, 1999.)

U

L U

A

B FIG 50-61  Benign gastric ulcer. A, A large penetrating ulcer (U) is seen in the posterior wall of the stomach, with markedly enhancing inner layer of the thickened gastric wall and edematous outer layer (arrows). Minimal perigastric infiltrate is also seen. A lipoma (L) is incidentally found in the duodenum. B, Doublecontrast upper GI series shows a large ulcer crater (U) with convergence of thickened gastric wall.

CHAPTER 50  Gastrointestinal Tract

1641

A P

FIG 50-62  Benign gastric ulcer. The swollen gastric wall surrounding the gastric ulcer simulates a polypoid mass (P). Gastric wall thickening (arrows) is also present.

free extravasation of oral contrast material or free intraperitoneal air, or both, along with evidence of diffuse peritonitis such as omental or mesenteric infiltration and intraperitoneal fluid collection. Although CT is not very accurate in detecting the perforating site,363,531 the use of thin-section spiral CT may improve detection by demonstrating interruption of the well-enhanced gastric wall.629 On occasion, peptic ulcers that extend beyond the serosa of the bowel wall may penetrate adjacent structures (confined perforation or penetration). The most common site of penetration is the pancreas, owing to an ulcer’s proximity to the distal stomach and duodenum and fixation of its position.531 Other sites include the gastrohepatic omentum, biliary tree, and liver, with the omentum being involved by gastric ulcers and the biliary tree by duodenal ulcers.313 On CT the diagnosis of ulcer penetration can be confirmed when a low-density linear band (sinus tract) containing perigastric contrast material is identified or when an ectopic pocket of gas is demonstrated in the perigastric space adjacent to the suspected ulcer crater (Figs. 50-63 and 50-64). In cases of pancreatic penetration, the fat plane between the ulcer and the pancreas is lost and the pancreas is focally enlarged with diminished CT attenuation, findings indistinguishable from those of acute pancreatitis. However, despite pancreatic enlargement, the remaining peripancreatic fat plane is commonly preserved, and only a small number of patients may show an increased level of serum amylase or lipase in such instances.363 CT can also depict penetration of a gastric ulcer crater or its surrounding inflammatory mass into the splenic parenchyma.257

Gastritis and Other Conditions Causing Gastric Wall Thickening.  Gastritis, recently associated with Helicobacter pylori infection, commonly involves the antrum and body of the stomach, with most cases involving the antrum.851 Enlargement of gastric folds and thickening of the gastric wall may be due to to a combination of mucosal inflammation, submucosal edema, and secondary hypertrophy and hyperplasia of the gastric mucosa. Gastritis appears as circumferential or focal gastric wall thickening on CT.851 Although CT is sensitive in detecting inflammatory processes in the stomach, it is not specific in differentiating gastritis from gastric malignancy. Absence of obliteration of the fat planes, perigastric infiltration, and lymphadenopathy may aid in diagnosis. Other infectious processes such as toxoplasmosis and cryptosporidiosis can also cause nonspecific gastric wall thickening.804,808

FIG 50-63  Confined peptic ulcer perforation. CT scan shows a perigastric cystic mass (A) with a thick peripheral wall in the perigastric region. The wall of the stomach and duodenum is considerably thickened, with heterogeneous contrast enhancement (arrows) and evidence of perigastric infiltrates.

D

FIG 50-64  Confined duodenal ulcer perforation. CT scan shows edematous bowel wall thickening of the duodenum (D) with an ulcer crater (arrow). A small amount of fluid is collected in the paraduodenal space.

In rare instances of acute infectious gastritis, an intramural gastric wall abscess develops, simulating a submucosal tumor.144 In patients with emphysematous gastritis, CT may show an irregular thickened gastric wall along with multiple intramural gas bubbles. When intramural gas is seen with noninfectious causes of gastric emphysema (e.g., secondary to gastric outlet obstruction), the gas is distributed as thin uniform linear streaks.548 Although rare, chronic granulomatous disease can involve the stomach, especially along the antrum, with diffuse nonspecific gastric wall thickening on CT. Ménétrier’s disease shows diffuse thickening of the gastric mucosal folds, predominantly along the body and fundus of the stomach (Fig. 50-65). In immunocompromised patients, graft-versus-host disease (GVHD) may demonstrate not only gastric wall thickening on CT but also a peculiar persistent coating of the gastric and small bowel mucosa with orally administered contrast material.377 CT also demonstrates gastric fold or wall thickening in Zollinger-Ellison syndrome (Fig. 50-66), amyloidosis, sarcoidosis, and eosinophilic gastritis. An extragastric inflammatory process (e.g., acute pancreatitis) may involve the gastric wall with focal thickening or cystic mass formation (Fig. 50-67).

1642

PART II  CT and MR Imaging of the Whole Body

Congenital Conditions

Diverticula.  Gastric diverticula generally arise from the posterior wall of the fundus. They appear as a thin-walled cystic-appearing mass on CT (Fig. 50-68). If the diverticulum is not filled with air or orally administered contrast material, it simulates a cystic inflammatory or neoplastic condition of adjacent structures.

Ectopic Pancreatic Rest.  The ectopic pancreas pathologically

FIG 50-65  Ménétrier’s disease of the stomach. CT scan shows marked thickening (arrows) of the gastric rugal folds along the greater curvature of gastric fundus and high body.

appears as a smooth nodule and occasionally as a mass with an irregular surface. It is usually situated in the submucosa and only occasionally expands into the muscularis. It most frequently occurs in the antrum within 3 to 6 cm of the pylorus and is on the greater curvature side of the posterior wall of the stomach; other sites include the second portion of the duodenum and the jejunum.62 On barium study, demonstration of umbilication, which represents filling of the rudimentary pancreatic duct, may suggest the proper diagnosis. When a nodule exceeds 2 cm, it appears as a mural nodule on CT (Fig. 50-69).

Gastric Duplication.  Most gastric duplications occur during the

FIG 50-66  Gastric fold thickening due to gastrinoma. CT scan shows marked gastric distention and fluid excess with prominent and thickened mucosal folds (arrowheads). A small mass (arrow), which was confirmed as gastrinoma at surgery, is noted in the perigastric region.

first year of life, but 28% of cases have been detected in patients older than 12 years.694 Approximately 82% are spherical cysts and do not communicate with the stomach; the remaining 18% are tubular and communicate with the gastric lumen.883 Although rare, a pedunculated type of duplication with a thin fibrous strand is also reported.448 The most common location is along the greater curvature, followed by the posterior wall, lesser curvature, anterior wall, and pylorus. Ectopic pancreatic tissue is found in the wall of up to 37% of gastric duplications328 and is associated with pancreatitis and elevated amylase levels. Detection of associated vertebral anomalies such as spina bifida, hemivertebra, and complete vertebral body cleft is a helpful clue for the diagnosis. Clinically, duplications occur twice as often in girls than in boys. Clinical symptoms and signs include a palpable mass, gastric obstruction, abdominal pain, melena, or hematemesis. Gastric duplication may undergo torsion and may perforate and cause a surgical emergency.779 CT and MRI may be used to determine the nature, size, and extent of the mass, revealing a cystic mass contiguous with the stomach and containing fluid, gas, or debris. Curvilinear calcification is rarely detected on the wall of the duplication.230 The differential diagnosis with other disease entities such as pancreatic cyst and pseudocyst, mesenteric cyst, and various intramural tumors of the stomach cannot be easily made by imaging findings alone. Technetium 99m-pertechnetate may show uptake of the radiopharmaceutical agent in the gastric mucosal lining.

Other Conditions

Varices.  Esophageal and gastric varices develop in patients with

P

FIG 50-67  Gastric wall involvement in a patient with pancreatitis. CT scan shows a cystic mass (P) in the gastric wall due to extension of pancreatic pseudocyst.

portal hypertension as a consequence of either intrahepatic obstruction (e.g., cirrhosis) or extrahepatic obstruction (e.g., portal or splenic vein occlusion). In contrast to the superficial location beneath the mucosa in esophageal varices, gastric varices tend to have a subserosal location and are thereby easily mistaken at endoscopy or radiographic studies for other submucosal masses.113 In the stomach the posteromedial border of the gastric fundus is the most common site, although any part of the stomach can be involved. CT is a sensitive tool for detecting gastric varices and determining their primary cause. Gastric varices appear as well-defined clusters of markedly enhancing rounded and tubular lesions within the scalloped gastric wall (Fig. 50-70).113 Intraabdominal collateral venous channels may also be seen along the distribution of the left gastric, gastroepiploic, or short gastric vein.

CHAPTER 50  Gastrointestinal Tract

F B C

A

B

C FIG 50-68  Gastric diverticulum. A, Coronal CT image shows a diverticulum (short arrow) arising from the gastric fundus (F) base, which is filled with oral contrast material. B and C, Axial CT images corresponding to the levels indicated by arrows B and C in part A show a contrast-filled outpouch (arrow) from the gastric fundus.

M

FIG 50-69  Ectopic pancreas in the stomach. CT scan shows a submucosal mass (M) in the prepyloric antrum of the stomach.

1643

1644

PART II  CT and MR Imaging of the Whole Body

FIG 50-71  Transient jejunojejunal intussusception after gastrectomy. CT scan shows a heterogeneous masslike lesion (arrow) in the jejunal loop caused by invagination of the regional mesentery into the bowel lumen. Small bowel follow-through study (not shown) did not demonstrate any mass.

A Although the diagnosis can be made with an upper GI series or gastroscopy, CT is also useful in diagnosis and early detection of possible strangulation in the intussuscepted jejunal loop. CT may show dilatation of the gastric remnant, with evidence of intragastric position of the jejunal loop and mesentery (Fig. 50-71).506,515

Gastric Herniation and Volvulus.  The diagnosis of gastric hernia-

B FIG 50-70  Varices of the gastric fundus in a patient with liver cirrhosis. A, Double-contrast barium study of the stomach shows nodular filling defects (arrows) at the gastric fundus, which reveals a smooth mucosal surface. B, On CT scan of portal venous phase, this lesion is matched with prolonged enhancing vascular structure (arrows) within the gastric fundic wall, indicating gastric varices.

Gastroduodenal Intussusception.  Gastroduodenal intussusception is an uncommon condition reported to be caused by benign pedunculated gastric tumors such as stromal tumor, adenoma, and lipoma or Ménétrier’s disease.173,299 CT findings are characteristic. The site of the invagination is seen as a target-like mass in the duodenal region. Both CT and barium study may also show foreshortening and beaklike narrowing of the distal stomach as well as gastric outlet obstruction (see Fig. 50-34).173,299

Jejunogastric Intussusception.  Jejunogastric intussusception is a rare complication of gastric surgery that may present either in acute form or as a chronic recurrent process. According to the type of invaginated loop, it can be classified as an afferent or efferent loop invagination or both. Efferent loop invagination is seen most commonly. Retrograde peristalsis is considered to be the main pathogenetic factor. Immediate laparotomy should be performed because an 80% mortality rate has been reported when treatment is delayed.136

tion through the lacerated diaphragm is often delayed or missed because its plain radiographic appearance resembles that of pneumothorax, hemothorax, pleural effusion, pulmonary consolidation or collapse, or raised hemidiaphragm. In addition to penetrating or blunt trauma to the diaphragm, laxity or absence of the tethering gastric ligaments, as well as conditions that increase the size of the esophageal hiatus, such as rapid changes in intraabdominal pressure (i.e., trauma) and degenerative changes (i.e., aging), may also predispose to herniation and volvulus of the stomach. Herniation may involve the fundus only or much of the body and antrum. In this condition a nasogastric tube may show a normal position at the esophagogastric junction but then pass up into the chest, although such findings can be seen in the presence of an elevated left diaphragm. A barium study may show the intrathoracic position of the herniated stomach as well as a waistlike constriction in the stomach where the stomach passes through the diaphragm. Although CT may be suboptimal in evaluating the diaphragm, several reports acknowledge its success (Fig. 50-72).314,846 Use of additional reformatted images in the coronal or sagittal plane helps facilitate detection of diaphragmatic abnormalities. MRI, especially with coronal or sagittal T1-weighted images, is useful not only for establishing the diagnosis but also for localizing the site of lacerated diaphragm (Fig. 50-73). To determine the presence or absence of strangulation of the herniated stomach that is mainly caused by volvulus, IV infusion of contrast material in both CT and MRI is mandatory.

Corrosive Gastric Injury.  Corrosive agents produce extensive damage to the GI tract, with perforation or death occurring in the acute phase. Alkaline caustic agents produce injury by liquefaction necrosis.261 Fat and proteins become saponified, and blood vessels are thrombosed, leading to further cellular necrosis and degeneration. This characteristic of an alkali enhances its penetration into tissues, resulting in complete dissolution of the upper GI tract and damage of

CHAPTER 50  Gastrointestinal Tract

1645

H S

S

A

B FIG 50-72  Delayed posttraumatic diaphragmatic hernia. A, CT scan shows herniation (H) of the proximal part of the stomach (S) into the thorax as well as evidence of disruption of the left crus of the diaphragm (arrows). B, Barium study shows gastric obstruction with markedly distended herniated intrathoracic stomach (S).

S

S

A

B FIG 50-73  Delayed posttraumatic diaphragmatic hernia. Gastric wasting (arrows) and intrathoracic position of the herniated stomach (S) are clearly depicted on coronal T1-weighted (A) and half-Fourier acquisition single-shot turbo-spin echo (HASTE) (B) MRI. Note the hyperintense thin diaphragm.

adjacent structures, even including in rare instances the transverse colon. In contrast to alkaline agents, concentrated acids produce a coagulative necrosis with a protective eschar that develops rapidly and may limit penetration to deeper muscle layers.285 The esophagus and stomach are the most common sites of involvement, but in severe cases, significant liquefaction necrosis can occur both extraluminally and much more distally in the GI tract, with the development of ischemia.622 Diagnostic evaluation of patients with corrosive injury should include examination of the entire esophagus, stomach, and duodenum by endoscopic or noninvasive methods. However, CT is useful in evaluating structures beyond the first site of injury, thus avoiding the risk of instrumental perforation associated with endoscopy (Figs. 50-74 and 50-75).

DUODENUM PATHOLOGIC CONDITIONS Malignant Neoplasms Primary Adenocarcinoma.  Most cases of primary duodenal carcinomas occur in the periampullary region and the third portion of the duodenum; duodenal bulb involvement is extremely rare.139 Duodenal adenocarcinoma may be associated with Gardner’s syndrome, neurofibromatosis, celiac sprue, Crohn’s disease, and Peutz-Jeghers syndrome. Lesions situated more distally usually have a napkin-ring appearance and produce partial intestinal obstruction; about 20% have

1646

PART II  CT and MR Imaging of the Whole Body

S

A FIG 50-74  Corrosive gastritis after acid ingestion, with nearly complete loss of posterior wall of the stomach (S), as seen on this CT scan.

S

FIG 50-75  Corrosive gastritis with pneumatosis in the gastric wall and multifocal hepatic infarction. CT scan shows intramural gas (arrowheads) along with pneumoperitoneum (white arrow) and ascites due to gastric perforation. Also noted are intrahepatic portal vein gas (black arrows) caused by hepatic infarction. (From Rha SE, et al: CT and MR imaging findings of bowel ischemia according to the various primary causes. Radiographics 20:29–42, 2000.)

B FIG 50-76  Ampulla of Vater cancer. A, CT scan shows an intraluminal protruded mass with cauliflower surface (arrows) in the region of the ampulla of Vater. B, This mass is well demonstrated on coronal CT image (arrows). Note the “double-duct sign” by the mass, which consists of marked dilatation of both the common bile duct (two arrows) and main pancreatic duct (thick arrow).

a predominantly polypoid or fungating appearance, probably caused by their origins in preexisting adenomatous polyps or villous adenomas (Fig. 50-76).85 Metastases are often present in the regional lymph nodes at the time of diagnosis, and thus the prognosis is poor. On CT, primary duodenal adenocarcinomas appear as an intrinsic mass with a short segment of bowel wall thickening (Fig. 50-77).206,394 The incidence of duodenal obstruction is common. Sometimes when tumors invade the pancreas and biliary ducts, it may be difficult  to distinguish these tumors from pancreatic and biliary tumors.  With primary adenocarcinomas, the duodenal wall is more circumferentially thickened, with a lesser degree of biliary and pancreatic duct dilation.

Lymphoma.  Duodenal lymphoma is rare because the distribution of lymphoma is proportionate to the amount of lymphoid tissue. Characteristic CT features include a long segment of a large annular wall thickening along with aneurysmal dilatation of the lumen and bulky regional or more widespread lymphadenopathy (Fig. 50-78). In  contrast to cases of primary adenocarcinoma, bowel obstruction is

FIG 50-77  Primary duodenal carcinoma. CT scan shows concentric bowel wall thickening (arrow), producing low-grade bowel obstruction. Note lymphadenopathy in the paraaortic and periduodenal regions.

CHAPTER 50  Gastrointestinal Tract

1647

S D

FIG 50-78  Duodenal lymphoma. CT scan shows irregular concentric

FIG 50-79  Duodenal obstruction caused by direct tumor invasion of

bowel wall thickening (arrows) in the duodenum. Note lymphadenopathy along the root of mesentery and lymphomatous nodules in the liver.

pancreatic cancer into the third portion of the duodenum. CT scan shows irregular soft tissue mass (arrows) obstructing the duodenum (D), with marked distention of the stomach (S). Pancreatic cancer is not shown.

uncommon. In rare instances, lymphoma involving the ampulla may produce biliary tract obstruction.

Metastatic Neoplasms.  Direct invasion from tumors of the adjacent organs is the most important route for metastatic neoplasm of  the duodenum. Primary tumor sites include the pancreas, stomach, gallbladder, liver, bile tract, right colon, and right kidney. In metastatic pancreatic or bile duct tumor, the medial aspect of the C-loop  is involved, with eccentric duodenal wall thickening. On the basis  of CT findings alone, however, it may be difficult to ascertain whether the engulfing mass is of duodenal, bile duct, or pancreatic origin  (Fig. 50-79). Metastases from left colon carcinoma occur via the lymphatic drainage to the central mesenteric nodes surrounding the fourth portion of the duodenum or across the short fascial plane of the lateral reflection of the transverse mesocolon206; it may cause duodenal or jejunal obstruction. Hematogenous and lymphatic metastasis to the duodenum also occurs in lung and breast cancer (Fig. 50-80). In patients receiving chemotherapy for lung cancer, bowel obstruction and perforation may develop as a result of tumor necrosis in a metastatic lesion.565 Gastric cancer may spread intramurally to the duodenum through submucosal lymphatic channels.211

Other Conditions

Trauma.  Because a large portion of the duodenum is within the retroperitoneum, most perforations occurring in these segments are clinically insidious. The horizontal segment of the duodenum crosses the vertebral column and is easily compressed against the spine by a direct blow. The distinction of duodenal perforation from an intramural hematoma without perforation is important because the latter condition does not require surgery.501 CT is becoming the most widely used modality for patients with abdominal trauma. Identification of extraluminal gas or extravasated oral contrast material or both in the right anterior pararenal space indicates the presence of duodenal perforation.453 Intramural duodenal hematomas are manifested on CT as tubular or ovoid, circumferential or eccentric, heterogeneous wall thickening (Fig. 50-81); a curvilinear area of high density (signet ring sign) can be demonstrated inside the margin of the hematoma.910 When the hematoma is resolved, the hematomas become cystic. Because of contiguity with the duodenum, it is important to search for an associated injury of the head of the pancreas.

M

FIG 50-80  Metastasis to the duodenum from squamous cell carcinoma of the lung. CT scan shows a polypoid mass (M) protruding into the duodenal lumen.

Pancreatitis.  Inflammation in pancreatitis may spread from the pancreas at the site of direct contact with the nonperitonealized posterior surface of the second part of the duodenum.55 Once the barrier is crossed, the inflammatory process may extend along the wall of duodenum. Duodenal involvement in pancreatitis can be identified on CT by nonspecific wall thickening (Fig. 50-82). In rare instances, intramural duodenal pseudocyst develops561; CT attenuation in this pseudocyst is related to the presence or absence of hemorrhage, cells, or proteinaceous material. In such cases, CT differentiation of duodenal pseudocyst from duodenal duplication cyst, choledochocele, or cystic dystrophy of the duodenal wall is often difficult.

Cystic Dystrophy.  Cystic dystrophy of the duodenal wall is a rare lesion characterized by the presence of cysts within the duodenal wall that originate from ectopic pancreatic tissue. This lesion can be induced

1648

PART II  CT and MR Imaging of the Whole Body

H D

FIG 50-81  Intramural duodenal hematoma after blunt abdominal trauma. CT scan shows a sausage-like cystic lesion (H) in the wall of the duodenum.

*

FIG 50-82  Pancreatitis with secondary duodenal involvement. CT scan shows concentric bowel wall thickening of the duodenum (asterisk) with heterogeneous contrast enhancement of the enlarged head of the pancreas.

by inflammation of ectopic pancreatic tissue within the duodenal wall secondary to obstruction of pancreatic excretory ducts. Since the first report in 1970 by Potet and Duclert,687 the radiologic appearances of this lesion have been described.693 The pancreas may appear to be normal or show evidence of chronic inflammation. In the cystic variant of cystic dystrophy, the cysts are easily demonstrated by both ultrasonography and CT between the duodenum and the pancreatic head. In contrast to a pseudocyst, the cystic variant is more frequently elongated or bilobate.693 The solid variant of cystic dystrophy is depicted as thickening of the duodenal wall along with microcysts embedded within it.693

Aortoduodenal Fistula.  Aortoduodenal fistula is a rare and usually fatal disease; it may be primary or secondary. Primary aortoduodenal fistula is caused mainly by atherosclerotic aneurysms (>70% of cases). Other causes of aortic aneurysms complicated by aortoduodenal fistula are trauma and infections (e.g., TB, syphilis, salmonellosis, mycosis).727 Although extremely rare, primary aortoduodenal fistula can also occur without an associated aortic aneurysm; causes include carcinomas, irradiation, peptic ulcers, foreign bodies, cholecystitis,

FIG 50-83  Primary aortoduodenal fistula in a patient with an aortic aneurysm. CT scan shows periaortic fibrosis (arrows) with subtle gas bubbles (arrowheads) in the periaortic region. The duodenal wall (D) adjacent to the aortic aneurysm is thickened, causing low-grade duodenal obstruction.

diverticulitis, and bacterial infections.440 Secondary aortoduodenal fistula is generally a complication of abdominal aortic aneurysm repair. The most common sites of involvement in all types of aortoduodenal fistula are the third and fourth parts of the duodenum. Clinical symptoms include abdominal or back pain, hematemesis, melena, and an abdominal pulsatile mass. Surgical therapy carries high mortality and morbidity rates, primarily because of the difficulty and delay in establishing the diagnosis. CT may offer a quick, inexpensive, and accurate method of diagnosis. On CT, an aortoduodenal fistula should be considered if extraluminal gas is present in the periaortic region (Fig. 50-83). Gas may also be noted within the calcified rim of the aneurysm in association with an adherent loop of intestine. Other findings include pseudoaneurysm formation and focal bowel wall thickening. In patients who have undergone aortic reconstructive surgery, complete resolution of hematoma should occur in 2 to 3 months, and no ectopic gas should be visualized at 3 to 4 weeks.700 After that time, any amount of perigraft soft tissue, fluid, or ectopic gas should be considered a sign of graft infection or aortoenteric fistula. However, the differentiation between aortoenteric fistula and perigraft infection is difficult, although the presence of ectopic gas favors the diagnosis of aortoenteric fistula.517

SMALL INTESTINE NORMAL ANATOMY On CT, normal small bowel wall thickness measures 2 to 3 mm; a wall thicker than 4 mm is considered abnormal except in the terminal ileum, where 5 mm is considered the upper limit of normal.740 The thickness of valvulae conniventes should not exceed 3 mm. When the intestine is collapsed or partially distended, the wall may appear to be 3 to 4 mm thick. If the wall seems to be concentrically and symmetrically thickened and is homogeneously enhanced, the clinical significance of a thickness greater than 4 mm should be interpreted cautiously and assessed in relation to the degree of distention of the small bowel loop.33 The fat in the mesentery has the same attenuation as fat elsewhere in the body. Major arteries and veins are identifiable as branching structures within mesenteric fat and do not exceed 3 mm in diameter.365 Mesenteric lymph nodes are occasionally observed within mesenteric fat and do not exceed 3 mm in diameter.

CHAPTER 50  Gastrointestinal Tract

1649

IMAGING TECHNIQUES Fluoroscopic examinations such as small bowel follow-through and barium enteroclysis used to be the most commonly performed radiologic tests to evaluate the small bowel. With recent advances in MDCT and fast MRI technologies, which have led to improved spatial and temporal resolution that allows exquisite high-resolution imaging of the small bowel, CT and MRI now have become the major radiologic modalities for small bowel evaluation. Compared to fluoroscopic studies, which give information mostly about inside the bowel, CT and MRI have the capability of mural and extraluminal evaluation. MPR or angiographic reconstruction of CT and MRI can also sometimes provide additional diagnostic information that is difficult to obtain otherwise. Optimal distention of the small bowel by administration of oral contrast is essential for accurate CT or MRI examination. Contrast administration can be performed perorally (i.e., CT enterography or MR enterography) or by infusion via nasoenteric intubation (i.e., CT enteroclysis or MR enteroclysis), similar to the procedure for barium enteroclysis. Although enteroclysis technique theoretically provides better small bowel distention, previous studies have not clearly shown the superiority of enteroclysis technique compared to enterography technique.611,776,890 Therefore it may not be necessary to use an invasive nasoenteric intubation for CT or MR examinations of the small bowel.

CT or MR Enterography For CT enterography, neutral oral contrast that produces attenuation similar to the attenuation of water is the choice in most examinations. Positive oral contrast, which was commonly used previously to distend and opacify the bowel loops, is helpful in delineation of luminal contour, detection of ulcer or fistula, and distinction between bowel loops and extraluminal structures such as mesenteric tumor or lymphadenopathy (Fig. 50-84). However, it makes evaluation of bowel wall enhancement difficult, which is a substantial drawback because bowel wall enhancement is crucial for detection and differentiation of many bowel diseases. Positive contrast also precludes 3D angiographic reconstruction of mesenteric vessels with volume rendering or maximum intensity projection (MIP) techniques because of the overlap between the contrast-filled bowel loops and contrast-enhanced vessels. Various types of neutral oral contrast used for small bowel imaging include water, water and methylcellulose, lactulose, polyethylene glycol, or a low-concentration (0.1% w/v) barium solution mixed with sorbitol (Fig. 50-85).659 Although water is inexpensive and well tolerated, it often is absorbed rapidly and provides suboptimal bowel distention, particularly in the distal small bowel.659,807 Methylcellulose has an unpleasant taste.215 Polyethylene glycol, despite its excellent bowel distention,455,458,537,807 is a cathartic, and the urge to evacuate can be great.215 The low-concentration (0.1% w/v) barium solution mixed with sorbitol is widely accepted; sorbitol minimizes small bowel absorption of water. The low concentration of barium produces approximately 20 HU, which is slightly higher than water attenuation. One study showed that this solution provided significantly better small bowel distention and visualization of mural features than water mixed with methylcellulose.567 The amount and timing of oral contrast for optimal distention of the entire small bowel is not yet standardized. Studies have used approximately 1 to 2 L administered in divided doses over time spans ranging from 30 minutes to 2 hours before the scan. Bowel distention for MRI examination can be achieved using methods similar to those for CT enterography. CT scan for enterography should be performed using a thin-slice thickness to allow high-quality coronal or sagittal images as well as 3D reconstruction. With the current scanners already in widespread use

FIG 50-84  Normal coronal CT enterography image after oral administration of Gastrografin as a positive oral contrast.

FIG 50-85  Normal coronal CT enterography image after oral administration of water as a neutral oral contrast.

1650

PART II  CT and MR Imaging of the Whole Body

(i.e., 16- or 64-detector MDCT), 1-mm slice thickness or less is easily obtained without motion artifact. IV contrast injection at a rate of 3 mL/sec or greater is recommended to obtain good bowel wall enhancement. Optimal scan delay depends on the indications. In most cases, particularly in the evaluation of Crohn’s disease and bowel obstruction, single-phase imaging with a scan delay of approximately 45 to 50 seconds, the so-called enteric phase wherein bowel wall enhancement is maximized, is most appropriate.659,771 In cases of suspicious GI bleeding or ischemia, multiphasic scans are necessary. In cases of suspicious GI bleeding, a nonenhanced scan may also be helpful in the detection of high-attenuating acute hematoma.

MRI MRI has the advantage of not using ionizing radiation and has better soft tissue contrast compared to CT, although it has inferior spatial and temporal resolution compared with CT. Appropriate sequences for small bowel examination are yet to be standardized. In general, both transverse and coronal images using fast T2-weighted sequences, such as half-Fourier acquisition single-shot turbo-spin echo (HASTE) technique and true-FISP, and using fast gradient echo T1-weighted sequences with and without IV contrast are often obtained. The appropriate scan delay for a contrast-enhanced study is yet to be determined. In contrast to CT, where the enteric phase is approximately 45 to 50 seconds after contrast injection,659,771 bowel wall enhancement on MRI was shown to be maximal during the portal phase (at 70 seconds).462 Given that MRI does not have a radiation hazard, multiple image acquisitions from 45 seconds into the portal phase may be a good strategy to image optimal bowel enhancement (Fig. 50-86).215

CT and MR Angiography CT angiography (CTA) and MR angiography (MRA) are highly effective methods for evaluating mesenteric vasculature and have replaced most conventional catheter-based diagnostic angiography techniques (Fig. 50-87). For high-quality imaging of mesenteric vessels using CT and MRI, careful attention to technical factors such as amount of IV contrast, injection rate, scan timing, slice collimation, reconstruction thickness, and reconstruction increment is required.

FIG 50-86  Normal coronal T2-weighted HASTE MR enterography image after oral administration of polyethylene glycol as a contrast agent.

Amount of IV Contrast and Injection Rate.  Regarding the amount of IV contrast, both fixed amount336,334 and body weight–adjusted amount (e.g., 2 mL/kg body weight for CTA)352 can be used. The body weight–adjusted method is based on the fact that the magnitude of vascular and parenchymal contrast enhancement of a given amount of contrast is inversely related to the body weight. With both methods, 120 to 150 mL of IV contrast agent is typically used for abdominal CTA of an adult patient. In terms of injection rate, higher rate is favored for two reasons. First, the magnitude of peak aortic enhancement increases with increase of injection rate.28 Second, faster injection rate shortens the time to peak arterial enhancement such that a greater temporal separation between the arterial and venous phases is obtained.28 For abdominal CTA, contrast injection at 3 to 5 mL/sec via 18- or 20-gauge angiocatheter in the antecubital fossa is typically used.

Scan Timing.  In terms of scan timing, fixed scan delay336,334 and customized scan delay using test-bolus28,352 or bolus-tracking28 methods can be used. Fixed scan delay is simpler to use. However, it does not reflect individual variations in circulation time and may result in suboptimal contrast enhancement in some patients. Customized scan delay considers the fact that optimal scan delay depends on several factors: (1) contrast medium injection duration, (2) contrast arrival time (Tarr), and (3) scan duration.27,28 Tarr varies among individuals according to their cardiac output and can be measured by using testbolus28,352 or bolus-tracking28 methods. The test-bolus method measures aortic attenuation by using a region of interest (ROI) placed over the lower portion of the descending thoracic aorta or the upper portion of the abdominal aorta after administration of a small amount of contrast agent as a test bolus (e.g., 15-20 mL for CTA). Then Tarr is defined as the time to maximal enhancement of the aorta after injection of a test bolus. Some have used Tarr measured with the test-bolus method as the scan start time for arterial-phase CTA103,352,755,789 or MRA,815 whereas others have added additional time delay to Tarr to calculate optimal scan start time, because the former method may result in earlier-than-optimal imaging acquisition.27,28 In the bolus-tracking method, the regular amount of contrast needed for angiographic examination is given to the patient, and ROI measurement of aortic attenuation is performed in a manner similar to that for the test-bolus method. In the case of CTA, Tarr is obtained by measuring the time needed for aortic attenuation to reach 50 to 100 HU above baseline attenuation after contrast injection.28 With the contrast arrival time (obtained as above) and injection duration, the time to peak aortic enhancement can be estimated as follows: injection duration + (0-10 seconds)29 and injection duration + (Tarr − 15) + (0-10 seconds)28 for CTA, and Tarr + injection duration/2187,478,691,692 for MRA. Optimal scan delay (i.e., optimal scan start time) for arterial-phase angiography can be calculated by subtracting half of the scan duration from the time to peak aortic enhancement so that peak aortic enhancement falls right in the middle of the scan duration or coincides with the center of k-space data acquisition.27,28 Optimal scan delay for arterial-phase angiography is therefore: injection duration + (0-10 seconds) − scan duration/229 and injection duration + (Tarr − 15) + (0-10 seconds) − scan duration/228 for CTA, and Tarr + injection duration/2 − scan duration/2187,478,691,692 for MRA. In contrast to the complexity of scan timing for arterial-phase angiographic examination, scan timing for the mesenteric or portal venous system is relatively straightforward. For most cases, a 60- to 70-second delay after contrast injection would be appropriate.28,336

Slice Collimation, Reconstruction Thickness, and Reconstruction Increment.  Slice collimation should be the thinnest possible to

CHAPTER 50  Gastrointestinal Tract

A

B

C FIG 50-87  Dynamic vascular phases on normal CT mesenteric angiography. A, Arterial phase (coronal MIP image). B, Portal phase (coronal MIP image). C, Coronal volume-rendering image.

1651

1652

PART II  CT and MR Imaging of the Whole Body

cover the entire scan range in a single breath hold in order to visualize fine details of mesenteric vessels: typically 0.75 mm on a 16-detector MDCT and 0.6 mm on a 64-detector MDCT. Overlapping image reconstruction should always be performed to improve 3D postprocessing.487 A reconstruction increment of 50% to 75% of the reconstruction thickness (e.g., 0.5- to 0.75-mm increment or 0.25- to 0.5-mm overlap with 1-mm reconstruction thickness) is generally appropriate.487 In addition to the traditional transverse images, coronal and sagittal images are very useful for evaluation of mesenteric vessels. Sagittal projection is particularly useful when evaluating the origin and proximal portions of the mesenteric arteries. Both volume-rendering and MIP are widely used postprocessing techniques for vascular evaluation. The smallest vascular branches are often best appreciated with MIP technique.

PATHOLOGIC CONDITIONS Benign Neoplasms Although the small intestine represents about 75% of the total length of the GI tract and more than 90% of its total mucosal surface, it is the site of fewer than 25% of all GI neoplasms and fewer than 2% of all malignant neoplasms.924 Five neoplasms making up 95% of benign small bowel tumors include benign stromal tumors, lipomas, adenomas, fibromas, and hemangiomas.151 Lipomas are slow-growing tumors composed of well-differentiated adipose tissue surrounded by a fibrous capsule. They are usually solitary, but small bowel lipomatosis has been reported.632 Some 90% to 95% of lipomas are submucosal and tend to prolapse into the bowel lumen. Lesions larger than 2 cm in diameter produce abdominal pain, diarrhea, constipation, or blood loss when the overlying mucosa is ulcerated or intestinal obstruction develops secondary to intussusception.833 On CT the diagnosis can be easily made by the tumor’s characteristic fatty CT attenuation (Fig. 50-88). MRI is also useful in diagnosis because lipomas appear hyperintense on T1-weighted images. Focal areas of soft tissue attenuation may be occasionally seen on CT in a lipomatous mass, especially in cases with intussusception. This finding may be caused by focal ischemic change and does not indicate liposarcoma. Multiple hemangiomas have been reported to involve the bowel and represent an important diagnostic consideration in patients with hemangiomatosis of the lower extremity. Phleboliths are frequently detected within hemangiomatous masses, providing a clue to their nature.605

Apart from these five benign neoplasms, inflammatory fibroid polyps can develop in the small intestine, especially at the ileum in 70% of cases (Fig. 50-89). They originate primarily in the submucosa and are composed of fibroblasts, blood vessels, and inflammatory cells within an edematous and collagenous stroma. Most lesions are solitary and large, often causing bowel obstruction due to intussusception. No distinctive radiologic features have been noted between these tumors and other mural or intraluminal lesions.

Malignant Neoplasms The most common histologic types of malignant tumors include adenocarcinoma, carcinoid tumor, lymphoma, and malignant stromal tumor. The nonspecific nature of symptoms contributes to the delayed diagnosis of small intestinal tumors, generally resulting in the poor prognosis of these patients. Even though barium studies such as enteroclysis and small bowel follow-through examinations have been used as primary diagnostic methods, CT and MRI not only have become primary tools in patients with a palpable mass, small bowel

M

FIG 50-89  Enterocolic intussusception due to inflammatory fibroid tumor of the ileum. CT scan shows a heterogeneous mass in the rightsided colon, with interposed mesenteric fat (arrowhead) and vessels (arrow). A round mass (M) is seen at the lead point.

B

A

FIG 50-88  Multiple lipomas in the small intestine. A and B, CT scan shows multiple lipomas (arrow) in the small bowel loops.

CHAPTER 50  Gastrointestinal Tract obstruction, or known lymphoma but also play a large role in preoperative staging and follow-up. The detection rate of the small bowel neoplasm was 80% in a series using single-detector CT,463 and 84.7% in another recent series that used CT enteroclysis.684 Dudiak and colleagues were able to suggest a specific tumor diagnosis in 60% of patients with adenocarcinomas, 58% of patients with lymphomas, and 33% of those with carcinoid tumors and malignant stromal tumors.184 CT is helpful in staging small bowel malignancy, but reported accuracy was not high (47%) in a series that included 15 patients with small bowel adenocarcinoma.97

Adenocarcinoma.  Primary adenocarcinoma of the small intestine is rare, accounting for approximately 40% of malignant tumors of the

small intestine. When this lesion occurs, it generally arises in the proximal jejunum. Risk factors for development of small bowel adenocarcinoma include a history of Crohn’s disease, sprue, Peutz-Jeghers syndrome, Lynch syndrome II, congenital bowel duplication, ileostomy, or duodenal or jejunal bypass surgery. The most frequent gross types are infiltrative, resulting in early bowel obstruction, and ulcerative, as represented by bleeding. Polypoid forms are uncommon. Adenocarcinomas tend to metastasize early to regional lymph nodes and invade the mesenteric vessels when they are located near the mesenteric root. On CT scans, the tumor appears as an eccentric focal mass or a circumferential bowel wall thickening in a short segment (Figs. 50-90 and 50-91). Narrowing of the lumen and dilatation of the proximal bowel are also common. However, tumors smaller than 2 cm

A

B FIG 50-90  Primary ileal adenocarcinoma. Axial (A) and coronal (B) CT scans show circumferential thickening (arrowheads) of the ileum, without obstruction.

A

1653

B FIG 50-91  Primary jejunal adenocarcinoma. A, Coronal true-FISP MRI shows a high-signal-intensity mass (arrowheads) obstructing the small bowel loop. B, Gd-enhanced MRI shows a high-signal-intensity mass (arrows) at the site of bowel obstruction.

1654

PART II  CT and MR Imaging of the Whole Body

in diameter may be missed by CT and are better detected by barium study.463

Carcinoid Tumor.  After the appendix, the small bowel is the second most frequent site of carcinoid tumors, representing up to 20% of GI carcinoids; 90% of these tumors are located in the ileum. Carcinoid tumors vary in appearance from a small submucosal lesion to a large intraluminal ulcerating lesion. The tumor tends to infiltrate the mesentery, causing characteristic desmoplastic mesenteric reactions such as angulation, kinking, rigidity, and separation of small bowel loops. Despite widespread metastases, the primary tumor itself may be difficult to detect on CT or even at laparotomy, owing to its small size. Therefore some researchers advocate the use of scintigraphy using meta-iodobenzylguanidine labeled with iodine-131 as a screening procedure in patients thought to have carcinoid tumors.6 The carcinoid syndrome, which includes diarrhea, flushing, and abdominal pain, occurs in 2% of cases of small bowel carcinoid, but it is much more likely to be present in cases of hepatic metastasis.820 CT may show virtually pathognomonic features, with a mesenteric mass associated with radiating linear or curvilinear soft tissue strands (Figs. 50-92 and 50-93). Calcifications are seen in approximately 70% of cases with mesenteric carcinoid tumor.644 The occurrence of retroperitoneal lymphadenopathy is not rare.683 Small bowel obstruction may develop as a result of desmoplastic reaction or serosal disease. CT features of small bowel carcinoid tumor may be similar in appearance to those of lymphoma, metastasis, Crohn’s disease, and mesothelioma. 18 FDG has not demonstrated significant uptake in well-differentiated neuroendocrine tumors such as carcinoid tumor and is only indicated in cases with less-differentiated tumors without somatostatin receptors and with high proliferative activity.79 For this reason, 18FDG positivity can be used as a prognostic parameter of aggressiveness.

Gastrointestinal Stromal Tumor.  GISTs are the most common mesenchymal malignancy of the GI tract despite the fact that they account for only approximately 1% to 3% of all malignant GI  tumors.252 Specific immunohistochemical properties of GISTs for KIT (CD117, a tyrosine kinase growth factor receptor) have allowed  GISTs to be distinguished from other mesenchymal tumors of the GI tract such as leiomyomas, leiomyosarcomas, schwannomas, and neurofibromas.583,765 Up to 94% of patients with GISTs express CD117 (KIT protein) and thus share similar markers with the interstitial cells

of Cajal.581 Mutations in the KIT oncogene of GIST resulted in constitutive activation of KIT receptor tyrosine kinase in GISTs. Thus KIT is not only the principal marker for diagnostic purposes but also a specific target for systemic therapy.324 Stromal tumors are usually solitary benign tumors that are categorized by their differentiation into smooth muscle or neural tissue. Tumors are classified as benign, borderline, or malignant. In contrast to benign stromal tumors (leiomyomas) that occur anywhere in the small intestine, malignant stromal tumors (leiomyosarcomas) arise more commonly in the ileum than in the jejunum. Symptoms depend on the growth pattern—submucosal, subserosal, or intraluminal. On CT and MRI, this tumor appears as a large heterogeneous mass with areas of low attenuation from hemorrhage, necrosis, or cyst formation502 and often grows in an exoenteric pattern with a large extraluminal component (Fig. 50-94). Severe hemorrhage within the mass, which is associated with hypervascular stromal tumors, produces a fluid-fluid level. Calcification is occasionally demonstrated.133 Even in the presence of large tumor, the incidence of bowel obstruction is rare. Some large cavitary tumors may show aneurysmal dilatation of the lumen, which is regarded as a fairly specific finding of lymphoma.764 Lymphadenopathy is not a common finding. Differentiation between benign and malignant stromal tumors is radiologically difficult, but CT findings favoring malignant stromal tumor include a mass larger than 5 cm, lobulated contour, heterogeneity with central necrosis or liquefaction, mesenteric infiltration, regional lymphadenopathy, and exophytic growth.133 Metastasis occurs by direct extension and the hematogenous route. However, intraperitoneal spread (peritoneal leiomyosarcomatosis) may occur by intraperitoneal seeding via ascitic fluid or by embolic metastases or rupture of the primary tumor,580,729 and it is more commonly seen in significantly larger or exophytically growing tumors.729 The most common CT manifestations of peritoneal tumor seeding include multiple well-defined peritoneal nodules or masses with or without a smudged or granular pattern of omental infiltration (Fig. 50-95).729 Lymphadenopathy is rare, and ascites may be seen but usually in small amount.863 These CT findings may simulate those of peritoneal carcinomatosis from GI or ovarian malignancies.729 However, multiple discrete peritoneal nodules are uncommon in

C

FIG 50-92  Carcinoid tumor. CT scan shows an intramural mass (C) in the bowel loop near the ileocecal region. Visualization of the primary carcinoid tumor is unusual. (From Megibow AJ, Balthazar EJ: CT of the gastrointestinal tract. St. Louis, 1986, CV Mosby.)

FIG 50-93  Mesenteric changes in a carcinoid tumor. Stellate, centripetally oriented, thickened mesenteric vessels are seen subtending a dilated “kinked” loop of distal ileum (arrows). The primary tumor is not visualized. (From Megibow AJ, Balthazar EJ: CT of the gastrointestinal tract. St. Louis, 1986, CV Mosby.)

CHAPTER 50  Gastrointestinal Tract

1655

B

A

FIG 50-94  Malignant GIST of the ileum. An ileal mass (arrows) with a lobulated surface can be seen on oblique coronal T2-weighted HASTE (A) and axial T2-weighted FISP (B) MRI. High signal intensity within the mass indicates tumoral necrosis.

*

*

*

FIG 50-95  Peritoneal tumor seeding in malignant GIST of the ileum. Axial CT scan shows multiple heterogeneously enhancing masses (asterisks) in the omentum and mesentery.

peritoneal carcinomatosis.290 Therefore CT features of peritoneal tumor seeding from GIST are more likely close to those of peritoneal tumor spread from hepatocellular carcinoma411 or other benign condition of leiomyomatosis peritonealis disseminata.648 Imatinib mesylate, a selective inhibitor of constitutive activity of KIT receptor tyrosine kinase in the cells of GIST, was introduced as a chemotherapeutic agent in the late 1990s and is now becoming promising for the management of patients with advanced GIST. In general, CT is regarded as the modality of choice in monitoring tumor response. The current gold standard for evaluating the response to imatinib treatment is monitoring the changes of tumor size on CT. However, use of tumor size criteria by measuring only the longest dimension is sometimes unreliable and may underestimate tumor response. Use of

the combination findings of tumor size, tumor attenuation, and the presence or absence of tumor nodules and vessels therefore seems reasonable.128,786 After chemotherapy with imatinib mesylate, in addition to size reduction, CT attenuation values become changed from hyperattenuating into hypoattenuating patterns (Fig. 50-96). Therefore liver metastasis responding to imatinib treatment appears to be cystic, sometimes mimicking simple hepatic cyst. In some tumors the size becomes paradoxically enlarged or tumor attenuation is increased despite significant clinical improvement, mostly owing to the development of intratumoral hemorrhage. In our experience in 21 patients in whom CT and pathologic correlation was possible, the histopathologic findings of the tumors contributing to the changes of tumor attenuation included hemorrhagic necrosis, cystic change, myxoid degeneration, hyalinization, or calcification. On FDG PET, metastasizing lesions show increased FDG uptake, and nonmetastasizing lesions do not show significant FDG uptake.318 In a study by Goerres and colleagues, who compared CT and CT PET in patients with GISTs after treatment with imatinib mesylate, CT detected more lesions than on PET, and FDG uptake was sometimes quite variable. However, CT PET gave more prognostic information than CT did; patients without FDG uptake after the start of treatment had a better prognosis than patients with residual activity.259

Lymphoma.  Types of lymphoma that may involve the small bowel include T-cell lymphoma, B-cell lymphoma, low-grade B-cell lymphoma (lymphoma of mucosa-associated lymphoid tissue [MALT]), Mediterranean lymphoma (immunoproliferative small intestinal disease), follicular lymphoma, Burkitt’s lymphoma, mantle cell lymphoma, and Hodgkin’s disease.747 Lymphoma accounts for approximately 20% of all small bowel tumors. The small intestine closely follows the stomach as the most frequent site of disease, and multicentric involvement occurs in 10% to 25% of cases.503 Almost all small bowel lymphomas are non-Hodgkin’s lymphoma; most of them are of

1656

PART II  CT and MR Imaging of the Whole Body

FIG 50-97  Lymphoma. Axial CT scan shows concentric bowel wall

A

thickening (arrows) in the ileum. Note enlarged lymph node (arrowhead) along the ileocolic vessels.

M

B FIG 50-96  Ileal malignant GIST after imatinib mesylate therapy. A, Axial CT scan shows a large heterogeneous enhancing ileal mass (arrowheads) with lobulated contour. B, CT obtained 12 months after starting chemotherapy shows marked shrinkage of tumor, with decreased CT attenuation value (arrows).

B-cell origin, but the cases complicating celiac disease are of T-cell origin. Predisposing conditions include AIDS, celiac disease, systemic lupus erythematosus (SLE), Crohn’s disease, and a history of chemotherapy.81 In children and young adults, lymphoma primarily affects the ileocecal region; in adults the distal ileum is the most frequent site of tumor.503 In lymphoma complicating celiac disease, however, the proximal jejunum is usually involved. Although the CT appearance of lymphomas is variable, these tumors can be classified as follows754: 1. Circumferential or infiltrative forms are characterized by concentric bowel wall thickening (Fig. 50-97). Aneurysmal dilatation of the lumen, regarded as one of the most characteristic features of lymphoma, occurs in approximately 50% of cases.184,463 Except for cases with developing intussusception, bowel obstruction is rare because the tumor weakens the muscularis propria of the bowel wall and does not elicit a desmoplastic response.754 2. Cavitary or endoexenteric forms are the second most commonly occurring pattern (Fig. 50-98). Cavitation of the tumor bulk results from ulceration of the mucosa with extension to the intramural portion of the tumor. Compared with stromal or epithelial tumors, the lymphomatous mass characteristically shows poorer contrast enhancement. Perforation of the tumor

A

n n

M

B FIG 50-98  Non-Hodgkin’s lymphoma of the jejunum—diffuse large B-cell type. A, CT shows a large cavitary mass (M) in the jejunum, with diffuse mesenteric lymphadenopathy (not shown). B, FDG PET scan clearly shows areas with abnormal FDG uptake corresponding to the lesions seen on CT scans. M, mass; n, nodes.

may occur secondary to a lack of desmoplastic response to tumor growth, as a result of rapid tumor growth, or as a response to therapy308 (Fig. 50-99). Fistulization to the skin, solid organs, or bladder develops in 5% to 15% of cases.96 3. Mesenteric forms are usually manifested as diffuse lymphadenopathy in the mesentery as well as in the retroperitoneum. A

CHAPTER 50  Gastrointestinal Tract

1657

n n

FIG 50-100  Non-Hodgkin’s lymphoma of the mesentery—diffuse large FIG 50-99  Perforated small bowel lymphoma after chemotherapy. On CT scan the bowel wall involved with lymphoma is focally perforated (arrow) along with evidence of a large amount of ascites and diffuse omental infiltration.

“sandwich” appearance is demonstrated when mesenteric masses surround mesenteric vessels but leave intact a thin rim of preserved perivascular fat (Fig. 50-100). The small bowel loops are displaced, focally invaded, or occasionally obstructed by mesenteric masses. Mesenteric lymphoma may also appear as an ill-defined confluent mass engulfing and encasing multiple loops of adjacent bowel or as a conglomerate mantle of mesenteric tissue. 4. Polypoid forms appear as single or multiple polypoid masses in the small intestine (Fig. 50-101). The incidence of intussusception is very high. Therefore differentiation of this form of lymphoma from intestinal polyposis syndrome may sometimes be difficult. However, the presence of lymphadenopathy in the mesentery or other sites may give the clue for establishing the diagnosis. 5. Nodular forms consist of submucosal nodules that are widely distributed in the small intestine. If these nodules do not reach a sufficient size, CT may not be helpful in their detection. The prevalence of lymphomas has been reported to be increased in patients with AIDS or in posttransplantation patients. In nearly all cases the pathologic type is non-Hodgkin’s lymphoma. The CT appearance of lymphoma in these immunocompromised hosts is indistinguishable from that of lymphoma in nonimmunocompromised patients. MRI appears comparable to CT in demonstrating the morphologic changes in the GI tract as well as in the peritoneal cavity in patients with small bowel lymphomas. Contrast-enhanced MRI had an advantage in depicting the submucosal tumor location by demonstrating a three-zone appearance: (1) a high-intensity mucosa, (2) an intermediate-intensity submucosal tumor infiltration, and (3) a lowintensity proper muscular layer.131 FDG PET has become a well-accepted method for diagnosis, staging, evaluating treatment response, and follow-up of patients with malignant lymphoma (see Figs. 50-98B and 50-101B). In addition, as the degree of FDG uptake is related to tumor grade, PET scanning can be used to estimate tumor grade prior to histologic classification.744 Although the use of only FDG PET may create difficulty in differentiating the high-grade non-Hodgkin’s lymphoma from other neoplastic or inflammatory disease, interpreting the results of a careful clinical

B-cell type. CT scan shows conglomerate lymphadenopathy (n) in the mesentery, surrounding mesenteric vessels with intact perivascular fat.

history, the extent of the disease, and corresponding CT images can contribute to the correct diagnosis. After treatment, evaluation of treatment response is an important aspect in the management of lymphomas. However, in the presence of residual mass, CT is often unable to discriminate between vital tumor and inactive fibrotic tissue. FDG PET is reported to be superior to CT in predicting the residual tumor or relapse after treatment, with higher specificity and positive predictive value than those of CT.452 In monitoring patients with high-grade non-Hodgkin’s lymphoma of the GI tract, the standard uptake value (SUV) plays a role; the SUV measurement is significantly decreased after therapy.674 Persistent FDG uptake after therapy may help predict treatment failure or a high risk of recurrence. False-positive scans can be caused by a variety of pathologic processes such as sarcoidosis, TB, histoplasmosis and other fungal infections, pyogenic abscess, spondylodiscitis, any process that results in an infiltration of metabolically active host cells,745 and inflammatory changes after treatment. Apparent accumulation of FDG PET in the GI tract may be a false-positive finding resulting from a combination of normal peristaltic activity, GI lymphoid tissue, and excreted radiotracer within the bowel lumen. False-negative scans can be seen in patients with minimal residual disease after treatment or in patients with malignant lymphoma that is not highly metabolically active. Thus low-grade lymphoma may be overlooked,373 and MALT lymphoma can be the cause of false-negative scan.326 Peripheral T-cell lymphoma.  Peripheral T-cell lymphoma (PTCL) is a rare disease in Western countries, affecting middle-aged patients; it represents lymphoma bearing a mature T-cell phenotype and excluding mycosis fungoides and cutaneous T-cell lymphoma.3 PTCL accounts for 5% to 30% of all non-Hodgkin’s lymphomas.4 Although the most common sites of extranodal involvement include bone marrow, skin, lung, and liver, the GI tract is rarely affected (<10% of cases).3,10,129 GI PTCL can be classified into several types, including PTCL unspecified, angioimmunoblastic T-cell lymphoma, angiocentric lymphoma, intestinal T-cell lymphoma, and adult T-cell lymphoma/ leukemia.311 Some investigators have also attempted to classify PTCLs of the intestine using histopathologic findings, including enteropathyassociated T-cell lymphoma (EATCL), EATCL-like lymphoma without enteropathy, and non-EATCL.130 The incidence of malabsorption and intestinal recurrence is high in EATCL.130 A history of autoimmune or

1658

PART II  CT and MR Imaging of the Whole Body

M

M

A

B FIG 50-101  Non-Hodgkin’s lymphoma of the terminal ileum—diffuse large B-cell type. A, CT scan shows an intramural polypoid mass (M) with homogeneous enhancement in the terminal ileum. Minimal lymphadenopathy is also noted along the ileocolic vessels. B, Coronal FDG PET scan shows areas with abnormal FDG uptake corresponding to the ileal mass (M) and enlarged lymph nodes seen on CT scan.

lymphoproliferative disorders or celiac disease has been reported in a minority of cases.130 Despite their rarity, PTCLs primarily involving the GI tract have some distinguishing features from those of B-cell lymphomas.130,176 First, in contrast to B-cell lymphoma, in which the stomach is the most frequent site and the distal ileum is a more common site in the small intestine, the preferential site of PTCL is the small intestine and the upper part of the jejunum.176 Second, PTCL has a high incidence of multifocal bowel involvement. Third, the incidence of bowel perforation is higher (41%-50%) in PTCLs than in B-cell lymphomas (<30%)130,176; angiocentric and angioinvasive infiltration of the vascular walls by lymphoid cells causes vascular occlusion with development of ischemic necrosis of both the tumor and normal tissues,311 subsequently developing bowel perforation. On CT, posttransplantation lymphoproliferative disorders (PTLDs) can show either bowel wall thickening or polypoid mass, but bowel wall thickening pattern is more commonly seen than polypoid mass (Fig. 50-102). In most patients with PTLD with bowel wall thickening, the degree of the thickening is not as severe as in B-cell lymphoma. It is attributed to the fact that the macroscopic features of PTCLs in the small intestine are predominantly those of thickened plaques, ulcers, or strictures.176 In addition, nonbulky lymphadenopathy is also considered strikingly different from B-cell lymphomas. Because of common prevalence of minimal bowel wall thickening, pneumoperitoneum, peritonitis, or minimal lymphadenopathy, PTLD is easily confused with acute inflammatory or ischemic enteritis. FDG PET appears to be a more sensitive method than CT in  detecting EATCL in patients with refractory celiac disease (see  Fig. 50-102B).300,327 Mucosa-associated lymphoid tissue lymphoma.  MALT lymphoma is one of the most common types of extranodal lymphomas.911 It can arise not only from an extranodal mucosal site such as the GI tract but also from a nonmucosal site such as bone marrow. Therefore

MALT lymphoma occasionally involves multiple organs or disseminates.311 In contrast to gastric MALT lymphoma confined to the stomach, in which Helicobacter pylori infection has a major role in development,893 H. pylori infection is not a causative factor in gastric MALT lymphoma involving multiple organs. Therefore MALT lymphomas involving multiple organs require combination chemotherapy or irradiation, contrary to MALT lymphoma confined to the stomach. In general, the stomach is known to be one of the most common organs involved, but MALT may occasionally disseminate preferentially into the GI tract.837 Multiple sites of GI tract involvement seem to be common among patients with multiorgan involvement of MALT lymphoma.838,912 CT findings of MALT lymphomas involving the GI tract other than the stomach do not differ from the subsets of non-Hodgkin’s lymphomas, showing homogeneous enhancing, circumferential wall thickening, or single or multiple polypoid mass in the small or large intestine, commonly in association with lymphadenopathy or hepatosplenomegaly (Fig. 50-103).

Metastases Metastatic disease in the small intestine may spread in the following ways: 1. Intraperitoneal seeding (ovary, stomach, appendix, colon) 2. Direct invasion from contiguous tumor (colon, pancreas, kidney, ovary, uterus) 3. Hematogenous dissemination (breast, lung, melanoma)

Intraperitoneal Seeding.  Intraperitoneal seeding produces ascites, and the ascitic fluid facilitates spread throughout the abdomen. Because of gravitational flow, the peritoneal implants tend to occur in four major areas: (1) the rectouterine or rectovesical pouch, (2) the lower small bowel mesentery near the ileocecal valve, (3) the sigmoid mesocolon, and (4) the right paracolic gutter. In the small intestine,

CHAPTER 50  Gastrointestinal Tract

A

1659

B FIG 50-102  Peripheral T-cell lymphoma of the jejunum. A, Coronal CT scan shows homogeneous enhancing wall thickening with aneurysmal dilatation (arrowheads) in the segmental jejunal loop. B, Coronal FDG PET scan shows areas with abnormal FDG uptake corresponding to the lesions seen on the CT scans.

Direct Invasion.  The small intestine can be directly invaded by the tumors of adjacent intraabdominal organs such as the pancreas, colon, kidney, ovary, or uterus. Common CT findings in the involved small intestine are focal or eccentric wall thickening or mass, along with displacement or mass effect on the small intestine. Small bowel obstruction commonly occurs. When primary tumor in other organs is not large, determination of the origin of primary tumors may be difficult.

Hematogenous Dissemination.  The incidence of small bowel

FIG 50-103  Mucosa-associated lymphoid tissue lymphomas (polypoid type) of the ileum. Coronal CT scan shows an intraluminal polypoid (black arrow) and a nodular lesion (white arrow) in the ileum. Another polypoid mass (not shown) in the ileum causes ileoileal intussusception (arrowheads). Also noted is regional mesenteric lymphadenopathy.

the tumor cells are implanted on the mesenteric border of the bowel, causing a direct bowel wall invasion. Once the bowel wall becomes involved, the tumors tend to spread circumferentially and longitudinally through the submucosa or subserosa in a short or long length, occasionally producing a linitis plastica pattern of bowel wall thickening. In such cases the target sign, a known sign of benign or inflammatory conditions, can be seen in the thickened bowel wall.289

metastasis from lung cancer is reported to be approximately 11% of cases.565 In such cases the presenting symptom may be bowel perforation, intestinal obstruction, malabsorption, or hemorrhage486; perforation is caused by tumor necrosis, obstruction by bulky tumor, hemorrhage by ulcerative lesions, and malabsorption by extensive involvement of mucosal surface. When metastatic tumor is found unexpectedly at laparotomy for perforated viscus in a heavy smoker after the fifth decade, lung cancer should be included as one of the possible sites of primary tumors. Although the GI tract had been considered a rare site for metastatic lung cancer, an autopsy series showed that metastasis to the GI tract has been found to exist in 14% in patients who died of lung cancer, and the small bowel was the most frequently involved organ by bloodborne metastases.19 Metastatic lung cancer to the GI tract appears as bowel wall thickening or polypoid mass on CT (Fig. 50-104).418 Multiple lesions are not rare with common incidence of bowel perforation and intussusception (Fig. 50-105).565 Lymphadenopathy adjacent to the involved bowel is also commonly seen. Of different histopathologic types of breast cancers, lobular carcinomas, which account for only a small fraction of breast carcinomas, metastasize significantly more often to the GI tract than do infiltrating ductal carcinomas. The most common and characteristic appearance of metastatic lobular carcinoma of the breast to the bowel is a linitis plastica–like appearance.274 Eighty percent of cases of melanoma metastatic to the GI tract involve the small bowel or mesentery. The site of hematogenously

1660

PART II  CT and MR Imaging of the Whole Body

P

A

* B FIG 50-104  Ileal metastasis from primary lung cancer (squamous cell type). A, Coronal CT scans show an encircling mass (arrowheads) without obstruction in the ileum. B, Coronal FDG PET scan shows an area with abnormal FDG uptake (asterisk) corresponding to the lesions seen on the CT scans. The FDG uptake is also seen in the primary lung cancer (P) of the middle lobe of the right lung.

seeded tumor cells influences their radiologic appearance. On CT scans, metastatic melanoma may manifest as mucosal or submucosal masses, serosal implants, or carcinomatosis.393 Mucosal mass may often lead to intussusception. Necrosis and ulceration may occur in large submucosal masses, but obstruction is less frequent. Lymphatic spread is less common, but when it occurs, lymphadenopathy may be marked.563 Uncommon malignancies that can metastasize to the GI tract may include renal cell carcinoma (RCC) and hepatocellular carcinoma (HCC). Although metastases may develop in more than 60% of patients with RCC, metastasis to the GI tract is uncommon. Batson’s vertebral venous plexus is considered a possible route in some cases. Typically, metastatic RCC to the GI tract presents as a solitary, bulky, and hypervascular lesion on CT.661 Metastasis to the GI tract from HCC is also an uncommon condition, and patients are usually asymptomatic. Hematogenous spread of HCC to the bowel may be caused by tumor thrombi via the portal system and progress through hepatofugal portal flow to the GI tract, but usually CT does not reveal any gross venous thrombi.653

Inflammatory Conditions Intestinal Tuberculosis

Incidence.  By the middle of the 20th century, intestinal TB had declined dramatically; since the mid 1980s, however, the incidence of TB has increased, especially in urban areas, owing to AIDS. Intestinal TB occurs at any age and is equally prevalent in men and women. The chest radiograph may show active disease in only about 15% to 20% of patients with intestinal TB. Intestinal TB may arise by several mechanisms546: • Swallowing of infected sputum in active pulmonary TB • Ingestion of contagious milk • Hematogenous spread • Direct extension from adjacent organs After the tubercle bacillus enters the GI tract, it traverses the mucosa to lodge in the submucosa. Lymphangitis, endarteritis, and fibrosis ensue, leading to mucosal ulceration, caseating necrosis, and narrowing of the intestinal tract. Intestinal TB manifests three gross pathologic types: (1) ulcerative, (2) hypertrophic, and (3) ulcerohypertrophic.

CHAPTER 50  Gastrointestinal Tract The highest incidence of abdominal TB has been noted in the GI tract and peritoneum, followed by the mesenteric lymph nodes.3 The affinity of the tubercle bacillus for lymphoid tissue and areas of physiologic stasis may be reasons why the ileocecum is the most common site of disease.546 CT and MRI findings.  There have been limited citations in the literature describing the CT features of intestinal TB. In our experience, however, CT is useful in determining the extent of disease, detecting its complications, and differentiating it from Crohn’s disease. The

A

B FIG 50-105  Ileal metastasis from primary lung cancer (squamous cell type). A and B, CT scans show a well-hyperattenuating polypoid mass (arrow in A) in the distal ileum, causing ileoileal intussusception (arrows in B).

common CT finding of tuberculous enteritis is bowel wall thickening, with a range of 1 to 2 cm in thickness (Fig. 50-106).293 The thickened bowel may show homogeneous attenuation on CT, but mural stratification may be rarely seen. Multiple sites of involvement with skipped areas are common. Bowel loop separation can be caused by mesenteric lymphadenopathy or lymphadenitis, intraperitoneal fluid collection or abscess, and rarely fibrofatty proliferation in the mesentery. Although not always the case, CT sometimes can give the clue in differentiating intestinal TB from Crohn’s disease.290,822,856 In intestinal TB, the thickened bowel is commonly less than 6 mm in thickness, asymmetric, and lacking mural stratification. It is also uncommon to see hypervascularity of mesenteric vessels (comb sign) in TB, unlike in active Crohn’s disease. Nodal involvement patterns are somewhat characteristic in intestinal TB: enlarged lymph nodes are commonly larger than 1 cm, may have low-attenuation centers caused by caseating necrosis in a third of patients, may contain calcification, and commonly involve the peripancreatic or paraaortic nodal chains (Fig. 50-107).290 However, lymph nodes with hypoattenuating centers can be seen in metastatic malignancy, lymphoma following radiotherapy, and inflammatory conditions such as pyogenic infection or Whipple’s disease.465 The presence of peritonitis on CT favors the diagnosis of TB rather than Crohn’s disease.290,538 CT findings of tuberculous peritonitis may mimic those of peritoneal carcinomatosis, including diffuse omental and mesenteric infiltration and nodules, peritoneal thickening, and ascites.290 In addition the incidence of splenic involvement is common in abdominal TB, which includes splenomegaly, hypoattenuating nodules, or calcifications.290 MRI features of abdominal tuberculous lymphadenopathy have been described in the literature.156,416 Although the signal intensity may differ depending on the stage of evolution, the involved lymph nodes most commonly show hypointensity on T1-weighted images, hyperintensity on T2-weighted images with perinodal hyperintensity, and predominant peripheral rimlike enhancement. Complications.  Although intestinal TB is a chronic disease, acute onset of abdominal symptoms may develop owing to complications. A broad spectrum of complications includes intestinal obstruction, bowel perforation (Fig. 50-108), fistulas, GI bleeding, enterolithiasis, venous thrombosis, and traction diverticula.65,546 Intestinal obstruction is the most common complication of tuberculous enteritis, with an incidence of 12% to 60%.546 Mechanisms may include inflammatory thickening of the bowel wall, especially in cases of hypertrophic or

I

A

1661

B FIG 50-106  Intestinal TB. A, Axial CT scan shows concentric bowel wall thickening of the ileum with diffuse perienteric infiltration (arrows). Note small calcified lymph node (arrowhead). B, Coronal CT scan shows patulous ileocecal valve (arrowheads) and focal stricture (arrow) at the proximal ascending colon. Also noted is diffuse bowel wall thickening of the ileum (I), cecum, and ascending colon.

1662

PART II  CT and MR Imaging of the Whole Body

*

A

B FIG 50-107  Intestinal TB. A, Coronal CT scan shows concentric bowel wall thickening (arrows) at the terminal ileum, cecum, and proximal ascending colon. The bowel wall is also thickened in the proximal ileum (asterisk). B, Coronal CT scan shows multiple enlarged lymph nodes (arrows) with hypoattenuating centers along the ileocolic vessels.

A

B

C FIG 50-108  Bowel perforation in a patient with intestinal TB during medical therapy. A, CT scan shows a stricture (arrowheads) at the site of low-grade small bowel obstruction. B, CT scan shows pneumoperitoneum (arrowheads) due to bowel perforation from the ulcer at the bowel loop proximal to the stricture. C, CT scan at a lower level shows an enterolith in the bowel (arrow) proximal to the stricture.

CHAPTER 50  Gastrointestinal Tract ulcerohypertrophic type with a long length of stricture or multiple areas of involvement and intraperitoneal adhesion.290 This complication commonly occurs during medical therapy. Healing by cicatrization in the course of antituberculous therapy increases the tendency  to result in obstruction, and the use of modern chemotherapeutic agents such as rifampicin also plays a partial role in producing cicatrization.14 Single or multiple bowel perforation may occur in patients with ulceration proximal to the obstruction,290 although it may be confined to a localized area when preexisting adhesive change is present. Fistula formation occurs less frequently than in Crohn’s disease but may result from bacterial invasion of a necrotic area in the bowel, a penetrating abscess, or sequelae of the confined perforation.657

Mycobacterium avium-intracellulare Complex Enteritis.  Mycobacterium avium-intracellulare complex (MAC) is the most common cause of systemic bacterial infection in patients with AIDS. It frequently affects the GI tract as well as the hepatobiliary system.916 Patients typically present with progressive weight loss, watery diarrhea, malabsorption, fever, and chills. This disease is often called pseudoWhipple’s disease because of the clinical, histologic, and radiologic similarities.686 The single most sensitive test for the diagnosis of disseminated MAC is peripheral blood culture, with a reported sensitivity of 86% to 98%.916 Common CT findings include diffuse jejunal wall thickening, intraabdominal lymphadenopathy, and hepatosplenomegaly (Fig. 50-109).443,645,705 The reported incidence of lymphadenopathy ranges between 42% and 100%. The appearance of abdominal lymph nodes appears to be useful in distinguishing Mycobacterium tuberculosis from MAC; lymph nodes in MAC are smaller and have a much lower frequency of hypoattenuating centers.

1663

ment than men. The most common sites of involvement include the distal small bowel and proximal colon; ileocecal area involvement occurs in approximately 50% of patients, the small bowel alone in 30% to 40%,624 and the colon alone in 20% to 27%.111 An increased risk of small bowel adenocarcinoma is likely to be present in patients with Crohn’s disease, and risk factors include Crohn’s disease with duration of 20 years or longer, intestinal bypass surgery, and chronic enteric fistula.281 An excessive rate of lymphomas has also been reported.280 Other extraintestinal complications include hepatitis, pericholangitis, hydronephrosis, renal calculi, and gallstones. Recurrence of small intestinal Crohn’s disease depends on the preoperative disease distribution and severity and on the extent of resection. The disease recurs in 40% to 80% of patients following ileocecal or ileocolic resection, almost always in the neoterminal ileum and within 2 years.921 CT findings.  CT yields accurate information regarding the extent of inflamed segment of the bowel, and it can detect potential complications. Accordingly, CT may sometimes play a critical role in the management of patients with Crohn’s disease; in one study of 80 patients with Crohn’s disease, CT demonstrated unsuspected abnormalities that altered medical or surgical management in 28% of cases.220 The distribution of bowel involvement on CT is commonly segmental with interposed loops of normal bowel. Although CT scans are usually normal when the disease is limited to the mucosa, the most common CT finding in Crohn’s disease is bowel wall thickening ranging from 1 to 2 cm (Fig. 50-110).270 Mural stratification, mucosal and mural hyperenhancement, edema in the perienteric mesenteric fat, and hypervascularity of the mesenteric vessels (comb sign) are well-known signs of active Crohn’s disease on CT (Fig. 50-111).78,260,268,474 Mural stratification in the thickened wall is composed of an inner ring of mucosal enhancement, an intermediate

Crohn’s Disease Incidence.  Crohn’s disease is a chronic transmural inflammation that may involve any part of the alimentary tract from the mouth to the anus. One fourth to one third of patients present before age 20 years, and about 15% of patients experience an onset after age 50. In the older age group, in contrast to younger people, the disease predominates in women, who show a higher frequency of colonic involve-

* *

*

* *

FIG 50-109  Mycobacterium avium-intracellulare enteritis. CT scan shows concentric bowel wall thickening (asterisks) of the small bowel loops with evidence of bowel dilatation and fold thickening. Mesenteric lymphadenopathy (arrows) is also seen. (Courtesy K.C. Cho, MD, Newark, N.J.)

FIG 50-110  Active Crohn’s disease. Coronal CT image shows focal bowel wall thickening (arrowheads) with mural stratification and mucosal enhancement in the distal jejunum.

1664

PART II  CT and MR Imaging of the Whole Body

* * *

*

* *

FIG 50-111  Comb sign in patient with active Crohn’s disease. Coronal

FIG 50-112  Enteroenteric fistula developed in patient with Crohn’s

MIP image shows prominent vasa recta (comb sign) adjacent to the bowel loops (black asterisks) involved by active Crohn’s disease. The bowel wall of the sigmoid colon (white asterisks) is also thickened with prominent pericolic vessels.

disease. Coronal CT scan shows enteroenteric fistula between bowel loops (arrows) involved by the Crohn’s disease. Also noted are mesenteric lymphadenopathy and diffuse small bowel dilatation due to multifocal ileal strictures.

hypoattenuating ring of edema in the submucosa and muscular layers, and an outer hyperattenuating ring of serosal enhancement260,268; with development of long-standing disease and transmural fibrosis, the thickened wall becomes homogeneous in attenuation. Mural hyperenhancement is depicted equally well during both arterial and portal phases,890 and quantitative measures of both mural enhancement and wall thickness are possible on CT enterography; quantitative measures are also reported to be highly correlated with ileoscopic and histologic findings of inflammatory Crohn’s disease.78 In many cases of clinically active diseases, CT scans may show hypervascularity of the involved mesentery, producing vascular tortuosity, dilatation, prominence, and wide spacing of the vasa recta in a comblike manner (comb sign) (see Fig. 50-111).474,579 Submucosal fat deposition and only mural thickening without enhancement or mural stratification correlate with fibrotic or quiescent disease. Fistula and sinus tracts are hallmarks of Crohn’s disease, occurring in about one third of patients.260 CT may have an advantage in demonstrating their full extent (Fig. 50-112). CT classification.  Some investigators attempted to classify the small bowel Crohn’s disease according to CT and/or MRI findings, insisting that such classification can be more objective than clinical classification of disease activity based on clinical and laboratory parameters.534 According to this classification, Crohn’s disease can be subdivided into four subtypes: (1) active inflammatory, (2) fibrostenotic, (3) fistulizing or perforating, and (4) reparative or regenerative. Newer imaging techniques such as MDCT, MRI, and CT or MR enteroclysis can provide information regarding the presence of ulceration,

edema, stricture, fistula formation, and inflammatory mass, which are required for classifying the subtypes. Causes of bowel loop separation.  A variety of conditions cause bowel loop separation on small bowel follow-through examination, and CT is useful in determining their primary causes, which are as follows: • Abscess • Phlegmon • Fibrofatty proliferation within the mesentery • Thickened bowel wall • Mesenteric lymphadenopathy Abscesses occur in about 15% to 20% of patients with Crohn’s disease owing to sinus tracts, fistulas, perforations, or postsurgical complication.434 On CT, they appear as a low-attenuation mass with a thick wall that may enhance with administration of IV contrast material, but the most specific finding may be the presence of gas bubbles within the mass (Fig. 50-113). CT is also useful for guiding percutaneous drainage of abscess collection, frequently obviating the need for surgical intervention. A phlegmon produces a loss of definition of surrounding organs and a smudgy or streaky appearance of the adjacent mesenteric or omental fat. Fibrofatty proliferation may present as an increase in the volume and attenuation of mesenteric fat on CT along with mild mesenteric lymphadenopathy and hypervascularity within the mesentery.579 Mesenteric lymphadenopathy is a common finding in patients with Crohn’s disease, but lymph node sizes usually range from 3 to 8 mm

CHAPTER 50  Gastrointestinal Tract

1665

B

A FIG 50-113  Peritoneal abscess complicating Crohn’s disease. A, Coronal T1-weighted FLASH MRI shows a heterogeneous mass (arrows) in the peritoneal cavity. B, Coronal Gd-enhanced MRI clearly depicts multiseptated abscess cavity (arrows) with well-enhanced thick peripheral wall and internal septations.

(Fig. 50-114). When the lymph nodes are larger than 1 cm, the presence of lymphoma or carcinoma as a complication of Crohn’s disease must be excluded.795 Newer imaging modalities.  In comparison with traditionally used barium small bowel follow-through examination and recently introduced wireless capsule endoscopy, which has unique advantages in direct visualization of the entire small bowel, the advantages and disadvantages of different types of newer imaging modalities such as noninvasive peroral CT enterography and CT or MR enteroclysis have also been reported in the literature. Compared to small bowel followthrough, CT enterography depicted more than twice as many cases of Crohn’s disease309 and was better in demonstrating the extraluminal findings and complications of Crohn’s disease. CT enterography may benefit patients with more advanced Crohn’s disease who cannot undergo capsule endoscopy. Several investigators have shown the utility of contrast-enhanced CT after placement of a nasojejunal tube and infusion of methylcellulose in the detection of active Crohn’s disease and its complications,539,690 but recent series890 show that noninvasive peroral CT enterography can be an effective alternative, with similar bowel distention and lesion detection as in CT enterography with placement of a nasojejunal tube. Because of its particular advantage of MPR imaging and excellent soft tissue resolution without radiation exposure, MRI can be an

alternative imaging modality in predicting disease activity in patients with Crohn’s disease by obtaining the information regarding mucosal, transmural, and extraintestinal imaging features. MRI findings that correlate well with active disease include significant bowel wall thickening, mural layer enhancement (mural stratification), hyperintensity in the thickened bowel wall on T2-weighted images, the degree of contrast uptake on contrast-enhanced MRI, detection of deep ulcer enhancing mesenteric lymph nodes, and T2 hyperintensity of fibrofatty proliferation on fat-suppressed images272,271,444,529 (Fig. 50-115; see Fig. 50-114). Until recently, PET or CT PET had either not been used or used sporadically in patients with inflammatory bowel diseases, despite widespread use of this modality for detection of certain primary malignant tumors as well as metastatic lesions. However, recent reports showed that PET activity correlated well with active inflammation  in Crohn’s disease.571 Several groups have also evaluated the usefulness of PET in inflammatory bowel disease. Although only seven patients were included in their study, Bicik and colleagues demonstrated that PET activity was high both in areas of macroscopic disease and in areas of active inflammation on biopsy in the absence of macroscopic disease.66 PET has sensitivity of 71% to 85% and specificity of 81% to 83% in identifying inflammation in patients with Crohn’s disease, and it appears to be more sensitive than MRI or granulocyte antibody scan.612,803

1666

PART II  CT and MR Imaging of the Whole Body

M

A

B FIG 50-114  Crohn’s disease. A, T1-weighted MRI shows mesenteric lymphadenopathy (arrows), especially along the jejunal branches of the mesenteric artery. Note the bowel wall thickening (arrowheads) of the ascending colon. B, Coronal FISP MRI shows fibrofatty proliferation (M) with hypervascularity (comb sign) of mesenteric vasculatures. Also note the focal bowel wall thickening (arrows) and pseudosacculation along the antimesenteric border of the adjacent bowel loop.

FIG 50-115  Active Crohn’s disease. Coronal contrast-enhanced MRI shows multifocal areas of well-enhancing bowel wall thickening (arrows) in the ileum.

Eosinophilic Enteritis.  Eosinophilic gastroenteritis is a rare inflammatory disorder characterized by focal or diffuse infiltration of the  GI tract by eosinophils. It usually occurs during the third decade of life. Although the etiologic mechanism is unknown, an allergic or immunologic disorder is the most commonly proposed cause for this

disease.864 Peripheral eosinophilia (present in 60%-96% of patients) is regarded as one of the most important clinical features for diagnosis.433,610 An allergic disorder such as bronchial asthma, food allergy, eczema, or atopic dermatitis occurs in approximately 50% of patients,433 and the clinical response to steroid treatment is usually dramatic. Patients usually present with recurrent episodes of abdominal pain, vomiting, diarrhea, or malabsorption, but the clinical features depend on the location and depth of tissue involvement. The Klein classification433 separates eosinophilic gastroenteritis into three groups according to the predominant tissue layer of involvement in the GI tract: (1) predominant mucosal, (2) muscular, and (3) subserosal disease. Mucosal involvement leads to protein-losing enter­ opathy, fecal blood loss, and malabsorption. Involvement of muscle layer (muscularis) often causes obstruction of the gastric outlet or small bowel, and subserosal involvement manifests as eosinophilic ascites. The stomach and small bowel are most frequently affected; the stomach alone is involved in approximately 40% of cases, both the stomach and small bowel in approximately 50%, and the small bowel alone in about 10%.610 The most common CT features include circumferential bowel wall thickening, predominantly in the jejunum (Fig. 50-116). The thickened bowel wall may show layered contrast enhancement (target or halo sign) or mucosal fold thickening.927 Polypoid mass is also reported in some cases on CT, rarely causing bowel obstruction.797 When subserosal involvement is predominant, eosinophilic ascites, eosinophilic pleural effusions, lymphadenopathy, and diffuse omental and mesenteric infiltration may also be seen.

Giardia lamblia Infection.  Giardiasis is a small intestinal infection with the protozoan parasite Giardia intestinalis. Isolates obtained from humans are commonly given the name Giardia lamblia. The major clinical impact of this infection is in young children, in whom it can produce persistent diarrhea, intestinal malabsorption, and retardation of growth and development. Outbreaks have been linked to contaminated water supplies and situations involving person-to-person contact. Immunocompromised patients are especially at risk. Most

CHAPTER 50  Gastrointestinal Tract

1667

* *

*

*

*

A

C

B

FIG 50-116  Eosinophilic gastroenteritis. A, CT scan shows diffuse thickening (asterisks) of the small intestine with ascites. B, Coronal T2-weighted HASTE MRI clearly depicts the discontinuous bowel wall involvement as well as bowel wall thickening (arrows) and ascites. C, Coronal Gd-enhanced MRI shows a target appearance of the thickened bowel wall with a markedly enhancing middle layer (arrows).

cases can be diagnosed by a single stool examination. CT findings show thickening of the wall of the duodenum and jejunum and mesenteric lymphadenopathy.630

Yersinia enterocolitica Infection.  Yersinioses are primarily diseases of animals, but human infections are recognized with increased frequency all over the world. In humans the alimentary tract is probably the portal of entry in most cases. Proposed modes of infection are (1) intake of contaminated food, (2) contact with infected animals, and (3) person-to-person transmission. Enteric infection causes mucosal ulcerations in the terminal ileum, necrotic lesions in Peyer’s patches, and enlargement of mesenteric lymph nodes.

Although most patients are younger than 5 years, clinical presentations vary with age, sex, and immunologic state of the host: acute enteritis (age < 5), mesenteric adenitis (age 5-15), acute terminal ileitis (age 10-20), gastroenteritis, erythema nodosum, and polyarthritis (adults).287 Most cases are self-limited, but complications have included appendicitis, diffuse ulceration and inflammation of the small intestine and colon, intestinal perforation, peritonitis, ileocolic intussusception,345 toxic megacolon, mesenteric venous thrombosis, and gangrene of the small bowel.143 A definitive diagnosis can be made by isolation of Y. enterocolitica together with demonstration of a rising antibody titer.857 The terminal ileum is most commonly affected in a length approximately 10

1668

PART II  CT and MR Imaging of the Whole Body

to 20 cm. The most typical lesions consist of shallow, small, round aphthoid ulcers to longitudinal ulcers of 2 cm or more, resembling those seen in Crohn’s disease. CT may help detect bowel wall thickening of the distal ileum as well as enlarged mesenteric lymph nodes  (Fig. 50-117).849

Salmonellosis.  Salmonellosis is one of the common causes of acute gastroenteritis. Salmonella species are gram-negative non–spore-forming bacilli, and in humans the disease is usually contracted by ingestion of contaminated foods, notably meat, dairy products, poultry, and eggs. Susceptibility to infection is heightened in patients with sickle cell anemia, hemolytic disease, and immunodeficiency. The terminal ileum may be affected mainly, although colonic changes are also seen. Salmonella organisms penetrate the wall of the small bowel, invading the lymphoid tissue of Peyer’s patches with development of small longitu-

dinally oriented ulcers. Ileal inflammation is usually superficial and involves the mucosa and lamina propria. Once the patient recovers, intestinal lesions heal with minimal fibrosis so that stricture formation is unusual. CT findings include circumferential and homogeneous thickening of the terminal ileum over a segment of 10 to 15 cm (Fig. 50-118).35 The colonic wall can also be mildly thickened. Small bowel perforation with peritonitis may occasionally occur.238

Whipple’s Disease.  Whipple’s disease is a chronic multisystemic condition resulting from systemic infection with Tropheryma whippelii. It frequently involves the jejunum and ileum but usually spares the duodenum. There is a characteristic infiltration of the small bowel mucosa and submucosa with foamy macrophages containing periodic acid–Schiff (PAS)-positive glycoprotein granules.228 The villi appear

B

A FIG 50-117  Yersiniosis. A, Small bowel follow-through scan shows multiple nodular defects (arrows) in the distal ileum. B, CT scan shows minimal mesenteric lymphadenopathy (arrows).

A

B FIG 50-118  Salmonellosis. A, Axial CT shows mild wall thickening in the small bowel (arrows) as well as in the colon (arrowhead). B, Diffuse edematous wall thickening (arrowheads) can be seen in the right-sided colon on CT scan.

1669

CHAPTER 50  Gastrointestinal Tract blunted and distorted by collections of foamy macrophages. Widespread fatty deposits and granulomas are present in the intestinal mucosa and intraabdominal lymph nodes, reflecting lymphatic obstruction with fatty extravasation. The most common clinical presentation is a generalized malabsorption syndrome. CT shows diffuse thickening of the small bowel folds. Characteristically the mesenteric or retroperitoneal nodes are of low CT attenuation owing to fatty deposits.735 Ancillary CT findings include hepatosplenomegaly and ascites.

Lymphangiectasia.  Intestinal lymphangiectasia, characterized by protein-losing enteropathy, hypoproteinemia, edema, malabsorption, and dilated lacteals, is a rare congenital obstructive defect of the  lymphatics affecting primarily children and young adults. Hypoplastic visceral lymphatics obstruct lymph flow, causing increased intestinal lymphatic pressure and thereby dilating lymphatic vessels through the small bowel and mesentery. Hypoproteinemia and steatorrhea are secondary to rupture of the dilated lymphatic vessels with discharge of lymph into the bowel lumen. The condition may also develop secondary to inflammatory or neoplastic involvement of the lymphatics.234 In both primary and secondary lymphangiectasia, CT shows diffuse small bowel and fold thickening (Fig. 50-119).695,816 A halo due to edema may be present within the wall. The presence of mesenteric infiltration and retroperitoneal or mesenteric lymphadenopathy may suggest the secondary form because adenopathy is not usually present in the primary form.816

A

Mucormycosis.  Mucormycosis, a relatively uncommon opportunistic fungal infection, is caused by fungi of the order Mucorales. The disease is known to occur especially in association with diabetes mellitus, leukemia, or lymphoma.104 When the GI tract is involved, the stomach is the most common site of involvement, and the intestinal form has a predilection for the terminal ileum and cecum.525 GI involvement usually complicates local disease processes such as intractable peptic ulceration, amebic colitis, persistent peritonitis, and malnutrition.17,104 However, a fulminant and rapidly fatal course can occur in certain instances.525 These fungi exhibit a remarkable tendency to infiltrate the walls of blood vessels, especially arteries. They grow profusely into the vessel lumen and initiate acute vasculitis and thrombosis of major blood vessels. As a result, ischemic infarction can occur in any organ.17 In some instances, venous involvement with thrombosis causes hemorrhagic necrosis. CT studies show diffuse circumferential wall thickening of the small bowel, intermingled with areas of both intense and poor contrast enhancement (Fig. 50-120).475 Pathologically, poorly enhanced areas coincide with the areas of necrosis and infarction, whereas intensely enhanced areas represent areas of edema and hemorrhage in the submucosa and muscle layers due to congestive changes. Therefore such CT findings can easily be confused with chemotherapy-induced necrotizing enteropathy if they have been seen in patients with lymphoma or leukemia who had received chemotherapeutic regimens.

B FIG 50-119  Lymphangiectasia. A, Axial CT scan shows diffuse thickening (arrowheads) of the bowel wall and mucosal folds in the small intestine. B, The image of double-balloon enterography shows thickening of the mucosal folds along with numerous nodular lesions diffusely scattered in the mucosal surfaces.

* * * *

Amyloidosis.  Amyloidosis is a group of heterogeneous disorders caused by the extracellular deposition of a protein substance or derivatives. Any part of the GI tract is involved, but the small intestine is most commonly affected. Use of a classification based on the histochemistry of the amyloid fibril is preferred825: amyloid A protein (AA), light chain protein (AL), prealbumin (AF), and β2-microglobulin (AH). Of these types, AL (primary) and AA (secondary) amyloidosis account for the overwhelming majority of cases. AA amyloidosis is the result of chronic

FIG 50-120  Intestinal mucormycosis. CT scan shows alternating areas of poor and prolonged contrast enhancement in the bowel wall of the small intestine (asterisks) along with regional mesenteric haziness. (From Lee JH, et al: CT and sonographically guided biopsy in a patient with intestinal mucormycosis. AJR Am J Roentgenol 175:129–131, 2000.)

1670

PART II  CT and MR Imaging of the Whole Body

inflammatory disease (e.g., Crohn’s disease, TB, adult or juvenile rheumatoid arthritis, SLE, ankylosing spondylitis). All four types of amyloid substance involve the GI tract, although each protein has its own specific mechanism of pathogenesis, radiologic findings, and clinical course. In the AL type, massive amyloid deposits commonly occur in the muscularis propria and submucosa of the intestinal wall.825 Wide granular deposits in the lamina propria mucosa are much more commonly seen in the AA type than in the AL type.254 Vascular deposits of amyloid in the submucosa are commonly seen in all types of amyloidosis. As a result, the vessel walls are thickened and the lumen is gradually reduced and eventually occluded, leading to ischemia, ulceration, infarction, and even perforation.825 When the GI tract is involved by amyloid deposit, many cases show normal CT findings without apparent bowel wall changes.21 When findings are present, the affected bowel shows focal or diffuse concentric and usually not severe wall thickening. Sometimes the thickened bowel demonstrates a target or double-halo appearance, so that differentiation of this disease from ischemic or other types of enteritis is difficult (Fig. 50-121).412 Bowel wall calcification reported in certain cases helps establish the diagnosis.332,384 Intestinal hypomotility results in delayed transit of the oral contrast agent.641 Adenopathy may be present but is not as bulky or as low attenuation as is seen in Whipple’s

*

* *

FIG 50-121  Amyloidosis. Axial CT scan shows minimal bowel wall thickening in the ileum (asterisks).

A

disease. Rarely, focal amyloidoma mimicking GI malignancy can cause bowel obstruction.760

Cytomegalovirus Enteritis.  CMV is a herpes virus that can be transmitted sexually or transmitted via infected organs, blood, or needles. CMV remains in a latent state in the host after the initial infection. With increasing impairment of cell-mediated immunity, particularly a CD4 lymphocyte count below 100 mm3, viral reactivation may occur.869 Clinically evident CMV infections most commonly affect immunosuppressed patients, including transplant recipients and patients with AIDS; however, they may occasionally occur in nonimmunosuppressed individuals, including those with diabetes mellitus, those who receive corticosteroids or radiation therapy, and those who are elderly.810 Clinical manifestations usually include diarrhea, hematochezia, abdominal pain, and constitutional symptoms. Although the colon is the organ most commonly affected by CMV infection, the incidence of small bowel involvement is not uncommon.604,835 Gross pathologic features include ulcers of various sizes and depth along with mucosal fold thickenings and luminal narrowing.835 The ulcers are usually numerous, round, or serpiginous, often reaching the muscular layer, with subsequently common occurrence of bowel perforation; CMV inclusions in the capillaries and venules of the mucosa and submucosa may produce focal ischemias that are responsible for mucosal ulcerations.265 These features simulate those of inflammatory bowel disease, bacterial or parasitic infection, ischemic bowel disease, lymphoma, and GVHD. CT may demonstrate bowel wall thickening in both the small and large intestines, lymphadenopathy, and ascites (Fig. 50-122).604

Gastrointestinal Disease in the Immunocompromised Host: Complications after Transplantation Human allogeneic bone marrow transplantation has been used with increasing frequency for treatment of lymphoma, leukemia, aplastic anemia, metabolic disorders of the hematopoietic system, and some metastatic disease. Preparatory regimens such as whole-body irradiation and high doses of chemotherapy, as well as immunosuppressive therapy, can cause profound immunosuppression after bone marrow transplantation. Apart from GVHD, a number of abdominal complications result from this immunosuppression, which include lymphoproliferative  disorders, enteritis from direct radiation or chemotherapy toxicity,

B FIG 50-122  CMV enterocolitis. A, Axial CT scan shows diffuse bowel wall thickening of both the ileum (arrows) and ascending colon (arrowheads). B, Coronal CT scan in the same patient demonstrates bowel wall thickening in both the ascending colon (arrowheads) and descending colon (long arrows). A large hemangioma (short arrow) is noted in the liver.

CHAPTER 50  Gastrointestinal Tract infective enteritis, neutropenic colitis, pseudomembranous colitis, pneumatosis intestinalis, abdominal abscess, and increased incidence of cancer.212

Graft-Versus-Host Disease.  GVHD is a significant complication after transplantation as a consequence of immunologic assault of donor lymphoid cells against host target organs.149 It develops when tissues containing immunocompetent cells (blood products, bone marrow, or solid organs) are transferred between persons. Acute GVHD.  Acute GVHD presents within the first 100 days of allogeneic bone marrow transplantation and is one of the major complications of this procedure. The incidence ranges from less than 10% to more than 80%.212 Major target organs in acute GVHD are the immune system, skin, liver, and intestine.212,840 The initial organ involved in almost all cases is the skin; hepatic and intestinal involvement usually appear several days after the rash.840 Therefore the hallmarks of the disease include rash, jaundice, and profuse secretory diarrhea.212,840 Pathologically the sine qua non of acute GVHD is selective epithelial damage of target organs.212 The destruction of intestinal crypts results in mucosal ulceration that may be either patchy or diffuse. The esophagus is less commonly involved. In acute GVHD, the most common CT findings include continuous or discontinuous bowel wall thickening of the small intestine, engorgement of the vasa recta adjacent to affected bowel, stranding of mesenteric fat, bowel dilatation, and ascites (Fig. 50-123).377,385 According to a report by Kalantari and colleagues, who analyzed CT findings of 22 patients with acute GVHD, all patients showed small bowel involvement, and 59% of patients also had large bowel wall thickening.385 The intestinal walls in involved segments are usually minimally thickened (5 mm),429 and the thickened small bowl wall commonly shows mucosal or serosal enhancement; a thin mucosal enhancement is caused by mucosal destruction and replacement by a thin layer of granulation tissue.177 CT findings associated with high-grade GVHD are thickening of the distal esophagus, ileum, or ascending colon, as well as increasing numbers of thickened bowel wall segments. It should be noted that gallbladder pathologies are not uncommonly combined with GI  tract abnormalities, which include gallbladder wall thickening, pericholecystic fluid collection, biliary sludge, and enhancement of the gallbladder wall.385 Unusual prolonged coating of the small intestine

and colon with contrast material has been reported in patients with severe mucosal disease; it seems likely that the barium may be adherent to mucosa, submucosa, or pseudomembranes or trapped in sloughed mucosa.333,379 The differential diagnosis includes GI infection, neutropenic enterocolitis, drug-induced enteropathy, and radiation enteritis. Of these diseases, the differentiation between acute GVHD and viral enteritis is important owing to different treatment strategy; immunosuppressive agents are required for acute GVHD but are contraindicated in infectious enterocolitis. Despite a difficulty in differentiation between these two conditions, the presence of mesenteric lymphadenopathy may favor the diagnosis of infectious enterocolitis, which is not reportedly identified in cases with acute GVHD.385 Colonic changes include luminal narrowing, wall thickening, ulceration, edema, and a haustral appearance.222,750 Chronic GVHD.  Although chronic GVHD was initially defined as a GVHD presenting more than 100 days after bone marrow transplantation, it can develop 40 or 50 days after transplantation.212 Its incidence ranges from 30% to 60% of patients with GVHD. As the number of long-term survivors increases, chronic GVHD has emerged as an increasingly frequent complication of allogeneic marrow transplantation. The clinicopathologic features resemble an overlap of several collagen-vascular diseases with frequent involvement of the skin, liver, eyes, mouth, upper respiratory tract, and esophagus and less frequent involvement of the serosal surfaces and lower GI tract.212,796 The distribution and histopathology of the lesions in the intestinal tract in chronic GVHD differ from those of the acute form, particularly with regard to the esophagus. Esophageal involvement causes a characteristic desquamation of upper esophageal mucosa.796 The small bowel and colonic mucosa seem to be spared in chronic GVHD.212,796

Posttransplantation Lymphoproliferative Disorders.  PTLDs are an uncommon but persistent complication of solid organ transplantation. When a primary infection with Epstein-Barr virus (EBV) occurs in an immunocompetent host, a transient lymphoproliferative disorder occurs. This proliferation of infected B lymphocytes is controlled by cytotoxic and suppressive T lymphocytes. When the infected host is immunodeficient, proliferation of B cells may not be controlled and EBV-induced PTLD can occur.609 It includes a spectrum of diseases

*

* A

1671

B FIG 50-123  GVHD in a patient who had a history of bone marrow transplantation. A, Axial CT shows intense contrast enhancement of mildly thickened bowel wall in the ileum (arrows). The bowel wall of the rectum and sigmoid colon (asterisks) are also minimally thickened. A small fluid collection is evident in the pelvic cavity. B, Coronal CT more clearly demonstrates that intense contrast enhancement occurs in the inner layer of the thickened bowel wall of the ileum (arrows).

1672

PART II  CT and MR Imaging of the Whole Body

ranging from polyclonal B-cell hyperplasia to monoclonal B-cell lymphoma. The reported incidence of PTLD varies from 12% in pancreas transplant recipients, to 9% in heart and lung transplant recipients, to 1% to 2% in renal and liver transplant recipients.521,708 Early diagnosis of PTLD is important because reduction or cessation of immunosuppression results in tumor regression. The anatomic region of the transplant organ was the most common site of involvement in lung and liver transplant recipients.681 According to a series that analyzed 51 patients with PTLD,681 CT manifestations of non-Hodgkin’s lymphoma occurring in the general population and PTLD in transplant recipients demonstrated important differences. The most striking distinction was the relatively high frequency of extranodal disease (81%) and lower frequency of nodular disease (22%) in patients with PTLD; in contrast, extranodal disease occurred in approximately 25% of nonimmunosuppressed patients with non-Hodgkin’s lymphoma.82 On CT, PTLD is similar in appearance to that of non-Hodgkin’s lymphoma when solid abdominal organs and hollow viscus are involved. The most common sites of involvement are the liver, small bowel, and kidney, but virtually any organ or space within the abdomen and pelvis can be affected (Fig. 50-124). Although the CT appearance of GI involvement in PTLD mimics that of lymphoma, PTLD has a propensity for ulceration and perforation of the alimentary tract.680

after transplantation. Contributing factors include mucosal damage from pretransplantation chemotherapy and radiation therapy, steroid therapy, acute GVHD, infectious colitis, small bowel obstruction, and severe humoral and cellular immunodeficiency.155 Day and colleagues reported on 18 patients with pneumatosis after bone marrow transplantation.155 The colon, predominantly the right side, was involved in 17 of 18 cases; 7 of these patients died within 2 weeks after detection of pneumatosis and 1 needed hemicolectomy, whereas the remaining 10 responded well to conservative treatment. Therefore in the clinical context of systemic infection or shock, the finding of pneumatosis may be a poor prognostic sign; if it is identified in asymptomatic patients, however, it is probably of little clinical significance.

Meckel’s diverticulum located on the antimesenteric aspect of the ileum 30 to 60 cm from the ileocecal valve. Acquired diverticula have only mucosal and submucosal layers that herniate through the muscular and serosal layers; they occur at the mesenteric border, in contrast with the true congenital Meckel’s diverticulum. Jejunal diverticula are found twice as frequently in men as in women and occur almost exclusively in persons older than 40 years.533 Diverticula tend to be larger and higher in number in the proximal jejunum and smaller and fewer as one progresses caudad.533 An exception is the terminal ileum, where the diverticula are often multiple. Acute complications of congenital or acquired diverticula include diverticulitis (with or without abscess or perforation), massive GI hemorrhage, intestinal obstruction, intussusception, and gangrene.157 Various types of neoplasm can occur in Meckel’s diverticulum,  owing to the presence of heterotopic tissues.904 Chronic symptoms include chronic abdominal pain, malabsorption, functional pseudoobstruction, and chronic low-grade GI hemorrhage. Small diverticula may not be recognized on CT scans because distinguishing between diverticula and bowel loops is usually impossible. CT findings that may be useful in identifying small bowel diverticula include (1) large size of the diverticulum (>3 cm), (2) different contents of the diverticulum and adjacent small bowel loops, and  (3) absence of valvulae conniventes.132 Intestinal obstruction may occur as a result of pressure on the intestinal wall from distended diverticula, an inflammatory mass associated with diverticulitis, stricture or adhesions from recent or past diverticulitis, intussusception at the site of the diverticulum, enteroliths developed within the diverticula, or volvulus of the diverticulacontaining segment.48,132 On CT, diverticulitis can be suggested when an inflammatory mass containing air or oral contrast material, with edema of the adjacent mesentery, is detected in an unusual location distant from the area of the ileocecal valve and from the sigmoid colon (Fig. 50-125). Contrast material within a diverticulum may remain for several days after CT has been completed.250 The presence of diverticulitis may mimic many other conditions such as abdominal abscess associated with appendicitis or inflammatory bowel disease, perforated colon cancer, and perforated right-sided colonic diverticulitis.56

Diverticular Diseases

Endometriosis

Small bowel diverticula are either congenital or acquired. A congenital diverticulum contains all three intestinal wall layers, usually as a

Endometriosis is a common disease characterized by the presence of ectopic endometrial tissue in an extramyometrial site. About 10% to

Pneumatosis Intestinalis.  Pneumatosis intestinalis can develop

m

* D

FIG 50-124  Posttransplantation lymphoproliferative disorder. Nonenhanced CT scan shows an ill-defined mass in the transplanted kidney (m), which appeared as a heterogeneous mass lesion on ultrasound examination. Another mass (asterisk) is noted along the ureteral course.

FIG 50-125  Jejunal diverticulitis. CT scan shows diverticular abscess (D) and diffuse perienteric haziness and infiltrates (arrows). (Courtesy K.C. Cho, Newark, N.J.)

CHAPTER 50  Gastrointestinal Tract

* *

FIG 50-126  Small bowel obstruction caused by endometrial implantation on the ileal loop. The small bowel loop is dilated (asterisks) with mild bowel wall thickening (arrows). Diffuse granular infiltrates are noted in the perienteric fat plane.

15% of women in their reproductive years are affected. GI tract lesions have been found in about 12% of patients with endometriosis; 85% of these lesions involved the rectosigmoid, 7% the small bowel, 3% the cecum, and 3% the appendix.526 Ileal endometriosis most often involves the terminal ileum within 10 cm of the ileocecal valve and can occur in the absence of colonic involvement.572 Endometriotic implants are typically located on the antimesenteric border of the bowel and have varied appearances, from short annular or plaquelike lesions619 or small pigmented nodules on the peritoneum to larger ones puckering the serosa. Small bowel endometriosis may present as cyclic episodes of abdominal pain coinciding with the menstrual cycle, nausea and vomiting, diarrhea, or constipation; however, classic symptoms are seen in only 18% to 40% of patients.589 Occasionally this disease may result in acute, chronic, or intermittent small bowel obstruction by kinking, stenosis, volvulus, or intussusception (Fig. 50-126).302 Conventional CT may not be considered as the primary imaging modality in detecting the tiny implanted lesions. However, when they become larger nodules with a diameter of approximately 2.5 cm, they can be detected on CT enteroclysis.69 Differentiation of endometriotic from metastatic nodules may be difficult, but metastatic nodules appear to be more intense than those of endometriosis.69

TRAUMA Incidence Bowel disruption due to blunt abdominal trauma is no longer rare and occurs in 5% to 25% of patients. The increasing frequency is attributed to automobile and industrial accidents. The mechanisms of bowel disruption have been thought to be at least one of the following different types of stresses: 1. Direct crushing impingement of the bowel against the prominent unyielding vertebral column631 2. A tearing or shearing force at areas of the bowel where either the bowel or mesentery is relatively fixed152 3. A sudden increase of intraluminal pressure in a closed loop of bowel, especially the duodenum It is generally agreed that the segments of the small intestine most commonly involved by blunt abdominal trauma are the duodenum near the ligament of Treitz, the proximal jejunum just beyond the ligament, and the ileum just proximal to the ileocecal valve, because these segments are relatively fixed. The relationship of intestinal injury to intraabdominal adhesions has been well documented,183 and the

1673

knowledge of a previous surgery should suggest possible intestinal damage. Similarly, fixation of the small intestine within an inguinal hernia has led to bowel perforation.152 Different types of injuries occur in blunt wounds of the small intestine, including transection, perforations, avulsions, intramural hematoma, and serosal tears. Delayed perforation up to 10 days after injury may occur. Clinically the classic triad of rigidity, absent or decreased bowel sounds, and tenderness occurs in approximately one third of patients.663 Diagnostic peritoneal lavage following blunt trauma markedly improves sensitivity (94%-100%) for detection of intraperitoneal injury but remains less sensitive for detection of retroperitoneal trauma to the duodenum, colon, and kidney.201,218 Because lavage findings are nonspecific for the type or extent of organ injury, positive lavage results may lead to nontherapeutic laparotomy rates of 6% to 25%, with associated morbidity.218 In addition, bowel perforation may be missed at lavage in up to 10% of cases in which the diagnostic procedure is performed soon after blunt abdominal trauma has occurred.205 As demonstrated by several researchers, blunt abdominal trauma is also one of the causes of segmental intestinal stenosis in association with acute ischemic insult to the intestine.323 Focal deprivation of the blood supply secondary to a tear in the mesenteric attachment is considered to be the main cause of the development of focal segmental ischemia. Although it depends on the site of the trauma, the stricture usually occurs near the proximal or distal extremes of the small intestine where mesenteric mobility commences.544 Intestinal stenosis is characterized by delayed onset of symptoms within a variable interval of months or years after the accident. Therefore if a patient’s history is overlooked, intestinal stenosis can be confused with other inflammatory or neoplastic GI diseases.

Imaging Findings CT is the diagnostic imaging method of choice in detecting bowel and mesenteric injuries caused by blunt abdominal trauma and may be useful in predicting the need for either surgical repair or conservative management, with sensitivities of 69% to 95% and specificities of 94% to 100%.84,182,590,738,801 Patients with a negative CT scan after blunt abdominal trauma do not benefit from hospital admission and prolonged observation.510 In performing CT scan in patients with blunt trauma, use of IV contrast material is necessary. Oral contrast material administration has been shown to be useful for demonstrating bowel injuries of the duodenum and proximal jejunum as well as mesenteric injuries.208 However, these benefits have been questioned by some researchers because of the risk of aspiration pneumonitis, potential delay in diagnosis, and lack of additional diagnostic information.811 In more recent series reported by Stuhlfaut and colleagues, MDCT without oral contrast material is adequate for diagnosing bowel and mesenteric injuries that require surgical repair.819 However, further investigation seems to be necessary for this controversy. When oral contrast material is administered, use of water-soluble contrast agent is recommended immediately before the study. In selected cases when colonic injury is suspected, rectal contrast material can be administered. Numerous CT signs of bowel and mesenteric injuries have been described in the literature.84,182,496,500,614,738 On CT, injury to the bowel may be manifested by bowel wall thickening, which is seen in 75%  of patients with transmural laceration.738 The most specific CT signs of bowel injury include pneumoperitoneum, frank extravasation of  oral contrast agent, and bowel wall discontinuity (Figs. 50-127 and 50-128).84 Unfortunately, both pneumoperitoneum and extravasation of contrast agent are insensitive, with free air seen in 32% to 56% and extravasated oral contrast seen in only 12%.496,738 The combination of

1674

PART II  CT and MR Imaging of the Whole Body

* *

* J

A

B FIG 50-127  Blunt trauma with bowel perforation and mesenteric injury. A, Bowel perforation (arrow) is noted in the transverse colon, with regional free gas collection (asterisks) and pericolonic infiltrates. The jejunal wall (J) is also thickened as a result of traumatic contusion. B, The mesenteric vessels (asterisk) are increased and prominent with their distorted courses, which resulted from mesenteric vascular injury. Intraperitoneal collection of free gas (arrows) and fluid is also seen.

*

*

*

H

FIG 50-128  Bowel perforation of the transverse colon and jejunum after blunt abdominal trauma. CT scan shows bowel wall thickening of both the transverse colon (asterisks) and the jejunal loop (arrows) along with diffuse pericolonic and perienteric infiltration. Also note pneumoperitoneum (arrowhead), intraperitoneal fluid collection, and right retroperitoneal hematoma (H).

pneumoperitoneum along with streaky mesenteric attenuation or mural thickening is suggestive of the presence of a significant bowel injury. Intraperitoneal gas may result from causes other than bowel perforation, such as pneumothorax, mechanical ventilation and pulmonary barotraumas, or diagnostic peritoneal lavage prior to CT; air entering via the female genital tract; or intraperitoneal laceration of the bladder secondary to cystography. Isolated, localized, unequivocal bowel wall thickening suggests the presence of bowel wall contusion but may not be associated with significant injury.540 In addition, diffuse bowel wall thickening can be sometimes encountered in trauma patients, but this finding is not typical for bowel contusion and may represent bowel edema secondary to systemic volume overload or to shock bowel (hypoperfusion complex). CT findings specific to significant mesenteric injuries include mesenteric extravasation of IV contrast agent, mesenteric vascular irregularity, and abrupt termination of mesenteric vessel.86 Mesenteric

vascular injury can cause vascular compromise, subsequently developing adjacent bowel wall thickening along with abnormal contrast enhancement. The presence of only localized mesenteric hematoma or focal bowel wall thickening indicates the need for close observation but not necessarily immediate surgical exploration without other definite indications.178 The finding of moderate to large amounts of intraperitoneal fluid with no visible solid organ injury is also a useful sign of bowel or mesenteric injury (see Figs. 50-127 and 50-128).84,500,738 Mesenteric fluid is extremely uncommon in solid organ injury, and its presence should be considered highly suggestive of bowel or mesenteric injury.496,614 Posttraumatic intestinal stenosis may cause diagnostic confusion when clinicians or radiologists do not have any information about a patient’s trauma history to the abdomen. This is because the radiologic features are similar to those of malignant or inflammatory strictures.535 On CT, posttraumatic intestinal stenosis shows nonspecific bowel wall thickening with or without bowel obstruction (Fig. 50-129). In such cases, barium studies of the colon or small intestine appear to be more useful than CT in demonstrating the characteristic findings of tubular or circumferential luminal narrowing in a short segment.

HERNIAS Abdominal hernias remain a common problem for both clinicians and radiologists. The main types of abdominal hernias are: • External, or abdominal wall, hernias • Internal hernias • Diaphragmatic hernias In the past the diagnosis of a hernia was made clinically either with plain radiographs or barium studies. Currently, specific diagnoses are confirmed more frequently by CT, especially in patients with nonspecific abdominal complaints. In addition to revealing the anatomic location and type of hernia, CT may provide information regarding the contents and complications of the hernia.

External Hernias Abdominal wall hernias account for most external hernias and include groin (inguinal and femoral), ventral (umbilical, paraumbilical,

CHAPTER 50  Gastrointestinal Tract

1675

S

* *

A

B FIG 50-129  Posttraumatic intestinal stenosis following blunt abdominal trauma that occurred 1 month before. A, Small bowel follow-through shows a short segment of tapered circumferential luminal narrowing (arrows) of the distal jejunum, along with thickened folds and ulcerations. B, CT shows marked contrast enhancement of the thickened bowel wall (S). Regional mesenteric vascular engorgement and haziness (asterisks) are also seen. (From Ha HK, et al: CT and MR diagnoses of intestinal ischemia. Semin Ultrasound CT MRI 21:40–55, 2000.)

epigastric, and hypogastric), lumbar, and incisional hernias. Less common hernias include (1) interparietal, Richter, Littre, and Spigelian hernias of the abdominal wall and (2) sciatic, obturator, and perineal hernias in the pelvis. All abdominal wall hernias consist of a peritoneal sac that protrudes through a weakness or defect in the muscular layers of the abdomen. Abdominal contents may include properitoneal fat, greater omentum, and viscera. Weakness in the transversalis fascia is the main source of predilection for abdominal wall hernia, especially in the groin. Trauma is a rare cause of these hernias. The most common complications of abdominal hernias include bowel obstruction, incarceration (irreducible hernia), and intestinal strangulation. CT helps in early detection of these complications. In most cases, bowel obstruction develops after incarceration and strangulation. The transition point in small bowel obstruction is at the level of the hernia, with CT appearances of dilated bowel proximal to the hernia and normal or collapsed intestinal loops distal to the obstruction. Although the diagnosis cannot be easily made by imaging findings alone, incarceration can be suspected when herniation occurs through a small defect and the hernia sac has a narrow neck.9 Their detection is important because of a high propensity of obstruction or strangulation. The known valuable CT criteria for diagnosing intestinal strangulation include free fluid collection within the hernia sac, bowel wall thickening, abnormal bowel wall enhancement (hypoattenuating or hyperattenuating), or proximal bowel dilatation292,723 (Fig. 50-130). However, it should be remembered that in contrast to those cases of small bowel obstruction developed in the abdominal cavity, the herniated bowel segment and involved mesentery are usually shorter in length, so there may be some difference in diagnostic value (sensitivity and specificity) of these CT criteria.347

Inguinal Hernias.  Inguinal hernias are either indirect or direct, depending on their relationship to the inferior epigastric vessels. In patients with an indirect hernia, the peritoneal sac passes down the course of the inguinal canal anterior to the spermatic cord and lateral to the inferior epigastric vessels. In women the hernia sac follows

S i i

i

FIG 50-130  Incarcerated spigelian hernia. CT scan shows bowel thickening with the “target sign” in the small bowel loop (S) herniated into the hernia sac, along with diffuse small bowel dilatation (i) caused by obstruction.

the course of the round ligament into the labium. The sac usually contains omentum and small bowel and occasionally ascites. A direct hernia is almost always acquired rather than congenital. The neck of the hernia is medial to the inferior epigastric vessels and generally does not traverse the inguinal canal. Bowel strangulation is seldom a complication because of the blunt and diffuse character of the protrusion. CT and MRI define the relationship of the inferior epigastric vessels to the hernia sac.

Femoral Hernias.  Femoral hernias are far less common than inguinal hernias. They are seldom seen in children and are more common in women. They occur below the inguinal ligament and lateral to the femoral vessels. Bowel strangulation is common.

1676

PART II  CT and MR Imaging of the Whole Body

S

FIG 50-131  Incisional external hernia with intestinal strangulation. CT scan shows marked enhancement of a small bowel loop (S) herniated in the hernia sac, along with a narrowed hernia neck (arrow). Fluid collection is present in the hernia sac.

The diagnosis is easily made by CT. A femoral hernia may be  distinguished from an inguinal hernia on CT because its sac lies  below and lateral to the pubic tubercle as it emerges from the femoral canal. Conversely, an inguinal hernia lies above and medial to the tubercle.876

Incisional Hernias.  Incisional hernias are a complication of laparotomy. Most of these hernias develop during the first 4 months after surgery but may remain clinically silent for up to 5 years until their detection.193 Vertical incisions are thought to be more likely to result in an incisional hernia than are transverse incisions. Incisional hernias can be seen with incisions as small as a puncture site for laparoscopic surgery.770 A common variation is the parastomal hernia, in which bowel and omental fat protrude through a fascial defect immediately adjacent to a colostomy or ileostomy stoma. This type has been reported to occur in as many as 36% of patients with an ileostomy.199 If the hernia is left untreated, bowel loops may be incorporated into the hernia and become incarcerated or strangulated. CT demonstrates the abdominal wall defects and hernial contents, as well as signs of bowel ischemia (Fig. 50-131).

Richter’s Hernias.  Richter’s hernia is an uncommon type of hernia in which only part of the bowel wall becomes involved. Because only a portion of the intestinal wall is incorporated into the hernia, the lumen remains patent and there is no obstruction. Incarceration is uncommon.

Littre Hernias.  Littre hernia is any hernia that contains a Meckel’s diverticulum. They are usually due to the presence of inflammatory adhesions that cause incarceration of the Meckel’s diverticulum within the hernia sac. Approximately 50% of Littre hernias develop in the inguinal region, 20% in the femoral, 20% in umbilical, and 10% in other sites.668

Spigelian Hernias.  Spigelian hernias occur at a weak point along the lateral border of the rectus muscle at a midpoint between the umbilicus and symphysis pubis. Although rare, they have a high frequency of bowel incarceration and strangulation.

FIG 50-132  Spigelian hernia. CT scan shows herniation of the mesentery and small bowel loop into the sac.

On CT a hernia sac protruding through the linear semilunaris and containing bowel or fat is identified. If it contains only peritoneal fat, it can be mistaken for an abdominal wall lipoma.876 However, careful inspection of the images reveals the muscular defect at the linear semilunaris (Fig. 50-132).

Obturator Hernias.  Obturator hernias protrude through the obturator foramen alongside the obturator vessels. They occur most often in thin older women. About one third of patients have neuralgia involving the thigh as a result of compression of the obturator nerve in the obturator canal.145 The diagnosis is made by the presence of herniated bowel protruding through the obturator canal on CT scans or MRI (Fig. 50-133).

Internal Hernias Internal hernias are the result of herniation of a viscus through a normal or abnormal aperture within the confines of the peritoneal cavity. Paraduodenal hernias account for approximately 50% of all internal hernias. Other internal hernias include transmesenteric, transomental, pericecal, intersigmoid, and retroanastomotic hernias; herniation through the foramen of Winslow or broad ligament; and supravesical and/or perivesical hernias. They are a rare cause of small bowel obstruction, and even small hernias are dangerous and may be fatal because of the risk of strangulation of the herniated contents. The clinical manifestations of most internal hernias are epigastric discomfort, periumbilical pain, and recurrent episodes of small bowel obstruction. CT may play an important role in establishing the diagnosis. The most important CT finding reported in the literature is a saclike mass of clustered small bowel loops with converging engorged mesenteric vessels toward its orifice; the converging point of the mesenteric vessels and/or intestinal loops may indicate the site of hernia defect.74,930 Depending on the types of internal hernia, other CT findings include displacement of the colon, cecum, or mesenteric vessels by a cluster of dilated small bowel loops and lack of overlying omental fat. The entrapped dilated loops within the hernia may be nonopacified and fluid filled. In addition to these findings, CT findings of intestinal strangulated obstruction, if developed, are combined (see “Strangulated Small Bowel Obstruction,” later).

Paraduodenal Hernias.  Paraduodenal hernias are the most common type of internal hernia and are usually left sided. They

CHAPTER 50  Gastrointestinal Tract

* *

A

*

1677

*

B FIG 50-133  Obturator hernia. A, Axial CT scan shows dilated small bowel loop (asterisk) in the space between the pectineus muscle (arrowhead) and the obturator muscle (arrow). B, Small bowel loops (asterisks) are dilated on axial CT owing to small bowel obstruction. Note the ovarian cyst in the left ovary (arrow).

frequently present with acute intestinal obstruction but may also cause chronic intermittent postprandial abdominal pain. A right paraduodenal hernia results from incomplete rotation of the duodenum during development. They usually contain only small bowel. The anterior wall of the sac comprises transverse colon and mesentery of the ascending colon. The entrance into the hernia sac is most commonly via the mesentericoparietal fossa of Waldeyer, which is in the first part of the jejunal mesentery immediately behind the superior mesenteric artery (SMA) and inferior to the third part of the duodenum.59,578 On CT, right paraduodenal hernias show a retroperitoneal mass comprising an encapsulated cluster of dilated small bowel loops and displacing the right ureter. The SMA and branches of the ileocolic artery are found in the anterior wall of the hernia. Left paraduodenal hernias also involve an anomaly of gut rotation. The bowel becomes entrapped behind the descending mesocolon in a hernia sac created by the fossa of Landzert. CT may show an abnormal cluster of normal-caliber or dilated bowel loops in an abnormal position either behind the body of the pancreas or between the stomach and the body of the pancreas (Fig. 50-134).153,656 There is usually a mass effect causing displacement of the posterior wall of the stomach, duodenojejunal flexure, and transverse colon. The inferior mesenteric vein and ascending left colic artery are located at the anteromedial border of a left paraduodenal hernia.

Transmesenteric Hernias.  Nearly 35% of transmesenteric hernias occur during the pediatric period and are probably caused by a congenital mechanism. Mesenteric defects are usually 2 to 5 cm in diameter and are located close to the ligament of Treitz or the ileocecal valve.255 The occurrence of this type of hernia has been increasing after patients undergo abdominal surgery requiring surgical construction of a Roux-en-Y loop, because such surgical procedures create iatrogenic defects in the mesentery.74 These hernias are usually not encapsulated and not enveloped in a sac, and so are not easy to detect if small bowel obstruction does not develop. On CT the herniated bowel is compressed against the abdominal wall, with usually no overlying omental fat, and is clustered and  lies outside the colon (Fig. 50-135). According to a study by Blachar and associates, valuable CT signs for the predictors of transmesenteric hernia include clustering of small bowel loops and mesenteric vessel abnormalities such as crowding, stretching, engorgement, and displacement of the main mesenteric trunk to the right.73 However, differentiation between these types of hernias from those caused 

FIG 50-134  Left paraduodenal hernia. The jejunal loops are enveloped by thin capsule of the mesentery (arrows), located at the subperitoneal space.

by prolapse of the intestine under adhesive bands may be very difficult.827

Herniation through Foramen of Winslow into Lesser Sac.  Herniation through the foramen of Winslow into the lesser sac may contain small bowel loops, cecum, ascending and transverse colon, gallbladder, and omentum. CT findings include gas-containing bowel loops within the lesser sac medial and posterior to the stomach (Fig. 50-136). The cecum and ascending colon may be absent from their usual location if they are part of the herniated viscera.

Hernia through Broad Ligament.  Broad ligament hernias can be classified as congenital or acquired. Congenital cases are the result of a developmental defect in this ligament. Acquired cases may result from prior surgery, pregnancy and birth trauma, or pelvic inflammatory disease.800 More than 85% of these hernias occur in parous middle-aged women. CT findings include a cluster of dilated small bowel loops in the pelvic cavity along with a point of convergence  of intestinal loops and/or mesenteric vessels in the pelvic cavity 

1678

PART II  CT and MR Imaging of the Whole Body

*

*

*

*

*

*

*

*

A

B FIG 50-135  Transmesenteric hernia with intestinal strangulation. A and B, CT scans show a cluster of fluidfilled, dilated small bowel loops (asterisks) in the left abdomen lateral and anterior to the descending colon (arrow). The involved bowel wall is mildly thickened, along with evidences of mesenteric haziness and poor opacification of regional mesenteric vasculatures. The mesenteric vessels are converged into the hernia orifice (arrowhead in B), with dilated regional mesenteric vein.

(Fig. 50-137). The dilated bowel loops may compress the rectosigmoid colon dorsolaterally and the uterus ventrally.827

INTUSSUSCEPTION Intussusception is the most common cause of intestinal tract obstruction in young children, but it is relatively uncommon in adults; 5% of all intussusceptions occur in adults, accounting for up to 5% of all cases of bowel obstruction in adults. In contrast with childhood intussusception, which is idiopathic in 90% of cases, almost half of adult cases may be idiopathic.872 As the intussusceptum enters into the intussuscipiens, the mesentery is carried forward and is trapped between the overlapping layers of bowel. Twisting or severe constriction of the mesenteric vessels may result in vascular compromise with subsequent edematous thickening of the involved bowel. Ischemic necrosis may develop if timely intervention is not undertaken.345 Intussusceptions may be enteroenteric, ileocolic, or colocolic types. In general, most lead points in the enteroenteric intussusceptions are benign, including benign neoplasms, Meckel’s diverticula, adhesions, lymphoid hyperplasia and adenitis, trauma, celiac disease, duplications, and inflammatory lesions; small bowel malignancy (either primary or metastatic) may account for less than one third of adult intussusceptions.872 From 8% to 20% of enteroenteric intussusceptions may be idiopathic.818 It should be noted that proximal small bowel intussusceptions are likely to be transient and nonobstructive and are unlikely to have a significant lead point.763 Colocolic intussusception is more likely to have a malignant etiology. Postoperative intussusception in adults has been also recognized as a distinct entity and may be related to a variety of predisposing factors, such as suture lines, oversewn ileum in jejunoileal bypass, previous jejunostomy site, adhesions, submucosal bowel edema, intestinal dysmotility, long intestinal tubes, and chronic bowel dilation.7 Intussusceptions in patients with AIDS are generally ileum based and may occur from lymphoma, atypical mycobacterial infection, or other unusual inflammatory process.108 Presenting signs and symptoms of intussusception include recurrent intestinal obstruction, intermittent abdominal cramps, and rectal bleeding. The diagnosis of intussusception is easily made by CT and MRI. The imaging appearance of a bowel-within-bowel configuration

M

*

*

FIG 50-136  Internal (lesser sac) hernia with strangulated small bowel obstruction. The dilated small bowel loops (asterisks) herniated in the lesser sac are poorly enhanced because of intestinal strangulation. Diffuse regional mesenteric haziness (M) is also present. (From Ha HK, et al: Differentiation of simple and strangulated small bowel obstructions: Usefulness of known CT criteria. Radiology 204:507–512, 1997.)

with or without contained fat and mesenteric vessels is pathognomonic (Fig. 50-138).872 Depending on severity and orientation to the scanning axis, intussusception may have three different CT appearances: a target lesion representing an intraluminal soft tissue mass with mesenteric fat, a sausage-shaped mass with alternating layers of low and high attenuation, and a reniform mass associated with focal ischemic changes.576 Although CT and MRI are sensitive in detecting the intussusception, morphologic analysis of the lead point may not provide the clue for determining their primary cause in many instances, except in cases of lipoma. However, the length of intussusception can be helpful in deciding clinical significance or need for surgery. Intussusceptions

CHAPTER 50  Gastrointestinal Tract

1679

* * B

A

FIG 50-137  Broad ligament hernia. A, Axial CT scan shows narrowing of the intestinal lumen at the site of hernia orifice (arrowhead) in the broad ligament. B, Axial CT scan shows a cluster of dilated bowel loops (asterisks) in the pelvic cavity between the uterus (arrowhead) and rectum (arrow).

FIG 50-138  Ileal lipoma with intussusception. CT scan shows a “bowel-within-bowel” appearance (arrowheads). A lipomatous mass (arrow) can be seen at the leading point.

shorter than 3.5 cm are almost invariably benign, and longer ones can be benign or have a significant cause.524,763 Vascular compromise is uncommonly seen in adult intussusception, requiring emergency surgery. It results from twisting or severe constriction of the mesenteric vessels trapped between the overlapping layers of bowel. Intestinal infarction may develop if timely surgical intervention is not undertaken. On CT the signs of vascular compromise in intussusception include loss of layered pattern or reniform appearance, along with evidence of vascular engorgement or twisting (whirl sign) of the mesenteric vessels, extraluminal fluid collection, or bowel perforation47,353,652 (Fig. 50-139). Intestinal ischemia or infarction may be confined to only the intussuscepted segment or can extend to a longer segment of the bowel. The presence of proximal small bowel obstruction does not always indicate vascular compromise.242

SMALL BOWEL OBSTRUCTION According to the degree, small bowel obstruction may be complete or high-grade partial obstruction or low-grade partial obstruction. The

diagnosis is usually made on the basis of clinical signs and patient history. However, because the clinical symptoms and signs are often nonspecific, radiologic studies (e.g., plain radiograph and barium studies) are considered essential for a diagnosis. Nevertheless, plain radiographic study is diagnostic in only 46% to 80% of cases, with common occurrence of false-positive and false-negative interpretations; up to 20% of patients may have no radiologic evidence of obstruction.87,237 Although barium studies are the standard of reference in diagnosing and characterizing intestinal obstruction, results of these studies are sometimes mixed. Furthermore, contrast studies are limited in the evaluation of changes in the peritoneal cavity and abdominal solid organs; may interfere with further diagnostic workup, such as CT or ultrasonography; and may be misleading, especially when a contrast agent passes into the colon through a high-grade partial obstruction. CT has been widely used for patients who are thought to have small bowel obstruction, because it is simple to perform and provides information about the site and cause of obstruction, the bowel wall and extraluminal changes, and the presence or absence of intestinal strangulation. CT diagnosis of bowel obstruction is based on detection of a transition zone, dilated fluid- or air-filled loops of small bowel proximal to the site of obstruction, and collapsed loops of small bowel or colon distal to the obstructed site. Bowel dilatation is indicated when the luminal diameter exceeds 2.5 cm243 and should be continuous.251 Additionally, small bowel feces sign, in which intraluminal particulate-like material is identified in the dilated small bowel, can be helpful in confirming small bowel obstruction (Fig. 50-140). This sign appears to be seen in approximately half of cases with small bowel obstruction, is present more frequently in moderate or high-grade obstruction, and is most conspicuous at the point of transition.464 Recent advent of MPR and 3D MDCT techniques allows easy demonstration of small bowel anatomy and the obstructive site with greater reliability. The addition of coronal reformations may add confidence to the diagnosis and exclusion of small bowel obstruction.364 Furthermore, because coronal reformations provide similar diagnostic accuracy as coronal images, and the coronal plane is well suited to display physiologic bowel anatomy, use of only coronal images may be sufficient for determining the presence/cause of small bowel obstruction.216,898 The speed of MDCT also optimizes the evaluation of patients with small bowel obstruction who have poor general condition;

1680

PART II  CT and MR Imaging of the Whole Body

* *

A

B FIG 50-139  Intussusception with vascular compromise in a patient who underwent gastrectomy 1 year before. A, CT scan shows a heterogeneous mass (arrows) caused by invagination of the mesentery. B, CT scan at another level shows diffuse haziness and mesenteric vascular prominence (arrows) along with suspected bowel wall thickening (asterisks) in the loops proximal to the obstruction. At surgery, intestinal strangulation was noted in the small bowel loops proximal to the intussuscepted site. Bowel adhesion was considered to be the primary cause of intussusception.

FIG 50-140  Small bowel feces sign. Nonenhanced axial CT scan shows the dilated ileal loop filled with small bowel feces (arrows). Also noted is a calcified enterolith (arrowheads) proximal to the feces. Surgery confirmed the high-grade benign stricture causing small bowel obstruction at the ileum.

artifacts caused by bowel motility, respiration, and patient motion are minimized. Furthermore, use of multiphase dynamic imaging gives information regarding hemodynamic changes in the intestinal wall. MRI is another alternative modality of choice in patients who are suspected of having small bowel obstruction,295 especially in those patients whose bowel is filled with residual barium suspension after barium study, who have hypersensitivity to the IV contrast material, or who are pregnant. The diagnostic capabilities of MRI for the  diagnosis of small bowel obstruction are similar to those of CT, providing rapid accurate identification of small bowel obstruction and  assisting in the determination of cause without exposing the patient to radiation.50,408 In most instances, coronal images are more useful than axial images in searching for an obstructed site, because they produce images comparable to those of small bowel follow-through (Fig. 50-141).

Degree of Obstruction

Complete or High-Grade Partial Obstruction.  Although CT is usually useful for evaluating high-grade obstruction, the transition

FIG 50-141  Simple intestinal obstruction caused by bowel adhesion. Coronal HASTE MRI shows diffuse small intestinal loop dilatation with tapered luminal narrowing at the obstructed site (arrow).

zone is sometimes difficult to detect. In these circumstances, use of MPR with coronal or sagittal images is an easy and fast way not only to determine the level of obstruction but also to reveal the cause. The presence or absence of colonic fluid or gas is also used to distinguish paralytic ileus from small bowel obstruction. In patients with highgrade small bowel obstruction, the sensitivity of CT in diagnosis ranges between 81% and 100%.237,532

Low-Grade Partial Obstruction.  The presence of low-grade partial small-bowel obstruction is considered to be present on CT when small bowel loops are mildly dilated, with an ill-defined transition zone, incomplete collapse of the distal bowel, and moderate colonic fluid and

1681

CHAPTER 50  Gastrointestinal Tract

FIG 50-142  Simple small bowel obstruction caused by adhesion and bands. CT scan shows diffuse dilatation of the small intestine, with a smooth beak (arrow) at the site of obstruction. (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

gas.237 The diagnosis can be easily misinterpreted as paralytic ileus. Furthermore, because the transition zone is difficult to detect, CT has limitations in determining the level and cause of obstruction. Therefore CT or MR enteroclysis or barium studies such as small bowel follow-through or enteroclysis may be necessary for further evaluating patients with suspected low-grade small bowel obstruction. In contrast to high-grade small bowel obstruction, the sensitivity of CT in diagnosis of low-grade obstruction is not high (48%).532 The occurrence of intestinal strangulation is rare in these patients.

FIG 50-143  Simple small bowel obstruction. CT scan shows diffuse dilatation of small intestinal loops with ascites. Accumulated fluid collection (arrow) between the mesenteric layers mimics mesenteric haziness. (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

S S

Types of Obstructions

Simple Small Bowel Obstruction.  Typical cases of simple obstruction may show diffuse bowel loop dilatation with a smooth transition zone or a smooth “beak” at the obstructed site on CT (Fig. 50-142). The bowel wall at the site of obstruction may be minimally thickened or of normal thickness. Mesenteric changes, such as vascular engorgement and haziness, are absent or minimal, and ascites is either absent or minimal (Fig. 50-143). However, the presence of such findings does not always indicate simple obstruction, because these changes can also be seen in some patients with strangulated obstruction.

Closed-Loop Obstruction.  Most cases of strangulated obstruction are related to closed-loop obstruction. Contrast radiography has no role in the acute clinical setting, but its value is in the further workup of partial obstruction in the subacute or chronic setting when CT cannot answer questions relevant to management.536 The most important CT indicators may include the whirl sign, convergence of mesenteric vessels toward the twisted site, and reversed position of the mesenteric artery and vein (Fig. 50-144)288; the whirl sign, however, is also seen in asymptomatic subjects or in patients who had undergone gastric surgery for gastric pathologies. Balthazar and colleagues described additional signs of U-shaped  or C-shaped configuration of dilated bowel loops and two adjacent collapsed, round, oval, or triangular loops with an arrowhead con­ figuration.34 The serrated beak sign (Fig. 50-145) described by Ha’s group297 also indicates the presence of a closed loop associated with regional mesenteric vascular engorgement, as well as bowel wall thickening at the obstructed bowel; this sign also suggests strangulated obstruction.

FIG 50-144  Closed-loop obstruction due to volvulus. CT scan shows twisting and engorgement of the mesenteric vessels (“whirl sign”) (arrows) adjacent to the obstructed site. Note diffuse dilatation of the small intestine (S) and intraperitoneal fluid collection. At surgery, strangulation was present in the involved segment. (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

Strangulated Small Bowel Obstruction.  Strangulation implies interference with the blood supply associated with an obstruction that may not necessarily be complete. In most instances it occurs as a complication of intussusception, torsion, volvulus, or any other form of closed-loop obstruction. Interference with the blood supply may occur either from twisting of the bowel on the mesentery or from distention of an obstructed closed loop. The incidence of strangulation is more than 40% in complete or high-grade obstruction766 but may be less than 10% in low-grade obstruction.71 Some researchers distinguish between long-loop, medium-loop (8-20 inches), and short-loop strangulated obstructions as follows200: 1. Long strangulated loops tend to cause death by sequestration of large quantities of plasma and blood from circulation. 2. Medium loops tend to cause shock and death by the formation of exotoxins of Clostridium welchii. 3. Short loops tend to perforate and cause peritonitis.

1682

PART II  CT and MR Imaging of the Whole Body

M

FIG 50-145  Strangulated small bowel obstruction due to adhesions. CT scan shows a serrated beak (arrows) in the bowel loop at the site of obstruction. The bowel wall (arrowheads) in the strangulated segment has lost its definition. A large amount of ascites is present. (From Ha HK: CT in the early detection of strangulation in intestinal obstruction. Semin Ultrasound CT MRI 16:141–150, 1995.)

With prompt surgical intervention, the strangulating element may be removed before gangrene of the intestine ensues. The sensitivity of CT in detecting the strangulated obstruction has been reported to range from 83% to 100%37,235,292,920; negative predictive value may be varied depending on the radiologist’s experience.406 The proposed well-known diagnostic CT criteria for strangulated obstruction are: • Portal or mesenteric venous gas, pneumatosis intestinalis • Abnormal bowel wall enhancement • Serrated beak sign • Unusual mesenteric vascular course • Diffuse mesenteric vascular engorgement and haziness • Bowel wall thickening • A large amount of ascites Although these CT criteria are useful in the diagnosis of intestinal strangulation, most of them, when used alone, are not sufficiently sensitive indicators of simple or strangulated obstruction. Furthermore, some of these findings are also noted in cases of simple intestinal obstruction. Therefore use of a combination of highly specific findings may increase the diagnostic accuracy of CT in detecting strangulation. In addition, the importance of clinical criteria (abdominal tenderness, tachycardia, fever, and leukocytosis) should not be overlooked. In equivocal cases demonstrated by CT, patients with none or one of these four clinical criteria had a higher chance of having simple obstruction, and patients with three or all clinical criteria had a higher chance of having strangulated obstruction. The presence of two criteria did not help in distinction between strangulated and simple obstruction.406

Portal or mesenteric venous gas or pneumatosis intestinalis.  Of the various findings reported, portal or mesenteric venous gas

and pneumatosis intestinalis (Fig. 50-146) are perhaps the most specific signs11,207 and are representative of bowel gangrene. However, the incidence of these signs is very low11,292; pneumatosis was noted in only one of the 19 patients in one series11 and was not found in any of the 41 patients in another series.292 Furthermore, some reports suggest that these findings are not always indicative of intestinal strangulation.235,828

FIG 50-146  Strangulated small bowel obstruction caused by bands and adhesion. Early phase of the spiral CT scan shows pneumatosis (arrows) and poor contrast enhancement in the strangulated segment of small intestinal loops. At the site just proximal to the obstruction, the superior mesenteric artery (arrowhead) is poorly opacified with contrast material as compared with the well-opacified superior mesenteric vein (curved arrow). Diffuse mesenteric haziness (M) is also noted. (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

Abnormal bowel wall enhancement.  Absent or poor contrast enhancement in the bowel wall is another specific CT finding for identifying vascular impairment in intestinal strangulation, with a sensitivity of 34% to 48% and a specificity of 100%292,920 (see Figs. 50-145 and 50-146). This finding in strangulated segments may be attributable mainly to a perfusion phenomenon (i.e., delayed return of venous blood, with subsequent slowing of arterial supply or arteriospasm). Delayed enhancement may be observed in strangulated segments, in which arterial or portal-phase CT shows poor contrast enhancement921 (Fig. 50-147). The most reliable method of determining whether abnormally enhancing bowel is present is to compare adjacent bowel loops. Because of the importance of analyzing the contrast enhancement pattern, it is better to avoid use of an oral contrast medium and to use dynamic CT study in patients thought to have small intestinal obstruction due to adhesions, bands, or hernia. Serrated beak sign.  This sign (see Fig. 50-145) as described in one series297 is highly specific in the diagnosis of strangulation; however, CT interpretation of this sign is problematic. The configuration of beaklike narrowing changes with different imaging planes, causing low sensitivity of this sign and yielding poor interobserver agreement.292 Unusual mesenteric vascular courses.  Most strangulated obstructions are associated with torsion or twisting of both bowel and mesenteric vessels. Therefore CT findings of an unusual mesenteric vascular course offer an important clue to the presence of closed-loop obstruction with intestinal strangulation. As described earlier, these findings include the whirl sign, converging of the mesenteric vessels toward one point, and reversed position of the mesenteric artery and vein (Fig. 50-148; see also Figs. 50-144 and 50-147). However, caution is needed when one is using CT to interpret these signs, because they can be seen in asymptomatic subjects (especially those patients who underwent gastrectomy) without closed-loop or strangulated obstruction (Fig. 50-149).

Diffuse mesenteric vascular engorgement and haziness. 

Although a mild degree of mesenteric changes (e.g., vascular

CHAPTER 50  Gastrointestinal Tract

A

1683

B FIG 50-147  Strangulated small bowel obstruction caused by bands and adhesion. A and B, CT scans show different degrees of contrast enhancement (arrows) in the thickened bowel wall of the same strangulated segment. Diffuse engorgement and haziness (arrowheads in A) are seen in the vasa recta of the regional mesentery. (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

FIG 50-149  Subtotal gastrectomy with gastrojejunostomy performed

FIG 50-148  Closed-loop obstruction caused by volvulus. CT scan shows convergence (arrow) and engorgement of the mesenteric vessels. Small intestinal loops are poorly enhanced compared with colonic loops. Intestinal strangulation was confirmed at surgery. (From Ha HK, et al: Differentiation of simple and strangulated small bowel obstructions: Usefulness of known CT criteria. Radiology 204:507–512, 1997.)

engorgement, haziness) is seen in simple obstruction,292,297 diffuse or severe mesenteric change is one of the highly sensitive and specific CT signs of intestinal strangulation (see Figs. 50-146 and 50-147).34,235,292 One should be cautious not to interpret the localized peritoneal fluid accumulated in the spaces between the two mesenteric layers as mesenteric haziness. Bowel wall thickening.  The likelihood of strangulated obstruction may increase when the bowel wall becomes thicker. To assess bowel wall thickness accurately, Maglinte and colleagues stressed the importance of not using oral contrast material in patients who showed definite or probable small intestinal obstruction on abdominal

for gastric cancer in an asymptomatic patient. The CT scan, obtained for routine follow-up study, shows twisting of the mesenteric vessels (“whirl sign”) (arrows). (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

radiography, because the presence of hyperattenuated intraluminal fluid made it difficult to determine wall thickening.532 When strict criteria (5 mm) were used, CT findings of bowel wall thickening showed high specificity (88%) with a sensitivity of 54% in distinguishing the simple from strangulated obstruction.292 However, the bowel wall is thinned or occasionally invisible when the bowel becomes gangrenous (see Fig. 50-142). The thickened bowel wall may sometimes show the target sign (Fig. 50-150) owing to submucosal edema or hemorrhage. Ascites.  In strangulated obstruction, there is rapid blood and plasma loss into the involved small bowel loops or peritoneal cavity. It is therefore not surprising that the presence of a large amount of ascites is an indicator of strangulation292 (see Fig. 50-150). However, the presence of a large amount of ascites in patients with small bowel obstruction is not always related to the obstruction and can be caused by many

1684

PART II  CT and MR Imaging of the Whole Body

A A

A

FIG 50-151  Intestinal obstruction caused by phytobezoar. CT scan FIG 50-150  Strangulated small bowel obstruction in an afferent loop in a 43-year-old man after subtotal gastrectomy with gastrojejunostomy for gastric cancer. CT scan shows bowel wall thickening of the afferent loop (A) with the “target sign.” A large amount of ascites is present in the peritoneal cavity. (From Ha HK, et al: CT diagnosis of strangulation in patients with small bowel obstruction: Current status and future direction. Emerg Radiol 2:47–55, 2000.)

other conditions. Therefore before this finding is used as a specific indicator in the diagnosis of strangulation, it is necessary to exclude other diseases. If results are questionable, paracentesis should be performed because peritoneal fluid is reddish brown or yields Escherichia coli with culture in patients with strangulated obstruction.484

Primary Causes of Obstruction Small bowel obstruction may be caused by three types of abnormalities: • Lesions extrinsic to the bowel (adhesions, hernia, peritoneal carcinomatosis) • Intrinsic bowel lesions (congenital atresia or stenosis, adenocarcinoma) • Obturation of the intestinal lumen (polypoid tumor, gallstone, bezoars) Among these lesions, the most common cause is adhesions, followed by malignancy and hernia.71 CT may determine the cause of obstruction in 73% to 85% of cases.237,568

Adhesions.  In general, adhesions and bands themselves are not visible on CT because they are thin fibrous connective tissues. However, CT can sometimes detect a linear impression on the bowel in the area of the transition zone (extraluminal band sign).670 As another CT finding of adhesions or bands, the obstructed site may show beaklike luminal narrowing or an abrupt change in bowel caliber without evidence of another cause of obstruction.107,297 Other intestinal loops distal to the obstructed site may be clustered together or collapsed with close attachment to the abdominal incisional wound.288 Inability to visualize an obstructive lesion such as a tumor, intussusception, or abscess may be diagnostic of obstruction caused by adhesions.36

Hernias.  A variety of external and internal abdominal hernias can cause bowel obstruction. Although CT is excellent for detecting and characterizing types of hernias and for demonstrating the bowel, mesentery, or omentum within a hernia sac, it is especially valuable in patients who have symptoms of small bowel obstruction possibly

shows an ovoid mass (arrows) containing mottled gas at the ileum, along with diffuse small intestinal dilatation due to obstruction.

resulting from hernia, but who have no hernia apparent on physical examination. Incarceration usually occurs as a result of a narrow hernia ring, which leads to bowel obstruction and may cause strangulation of the hernia sac and its contents (see Fig. 50-131). Small bowel obstruction caused by internal hernia can be predicted when CT shows abnormal location of bowel loops, crowding of small bowel loops, and an unusual course of mesenteric vessels (see Fig. 50-136). Because the hernia may not be recognizable if it is reduced spontaneously, it is better to perform CT during symptomatic periods. Defects in the small bowel mesentery, usually through a small hernia ring, may show a high frequency of strangulation and intestinal gangrene.

Bezoars.  A phytobezoar is a concretion of poorly digested fibers, fruit seeds, and pulpy fruits, especially oranges and persimmons, causing small intestinal obstruction usually after migration of a gastric phytobezoar. However, obstruction is also caused by a primary bezoar (enterolith) formed in the small intestine in association with underlying small bowel diseases such as diverticula, strictures, or tumors.239 Gastric surgery is the most predisposing factor for development of phytobezoar; other factors include poor mastication, autonomic neuropathy, and overconsumption of foods with a high fiber content.316 Early diagnosis is important because phytobezoars provoke decubitus ulcer, pressure necrosis, perforation, and even strangulation if the diagnosis is delayed.316 The most common CT findings of phytobezoars include a round or ovoid or a long sausage-shaped mass containing mottled gas at the obstructed site (Fig. 50-151).439 Phytobezoars are rarely manifested as a soft tissue mass without gas, resembling an intraluminal tumor or intussusception. In these cases, barium study may give the clue in suggesting the diagnosis, and the diagnosis should be suspected when intraluminal filling defects appear to be mobile or multiple. It should be remembered that in patients with small bowel obstruction due to phytobezoar, the stomach should be scrutinized before surgery; residual gastric phytobezoars can cause redevelopment of small bowel obstruction after surgery. In certain instances, phytobezoar can develop in the bowel proximal to underlying small bowel disease such as benign or malignant stricture. In contrast to cases of patients with underlying small bowel disease, the bowel wall thickening in cases of phytobezoar is seen on CT at the obstructed site and/or the bowel proximal to the obstruction.409 In the differential diagnosis, small bowel feces that develop proximal to the obstructed site in

CHAPTER 50  Gastrointestinal Tract patients with high-grade small bowel obstruction may mimic the signs of phytobezoars on CT. Some bezoars (enteroliths) appear as a calcified intraluminal mass with a laminated appearance.316 Phytobezoars appear as a hypointense mass on T1- and T2-weighted MRI, and the hypointensity may be due to their high fiber content and admixture of internal gas.

Gallstone Ileus.  Gallstone ileus refers to erosion of a gallstone into the GI tract, with subsequent development of bowel obstruction. This condition occurs in 0.3% to 0.5% of all cases of cholelithiasis.154 Although the duodenum is the most common site, cholecystoenteric or biliary-enteric fistula can occur in the colon or stomach.154 It is commonly agreed that a gallstone must be at least 2.5 cm to cause obstruction. In general, stones causing obstruction are single, large, and faceted.880 Bowel obstruction can occur preferentially at the narrow portion of the GI tract, such as the duodenum, the ligament of Treitz, the terminal ileum, or any area of stricture. CT findings include gas in the gallbladder or biliary tree, calcified gallstone in the dilated bowel, and intestinal obstruction284 (Fig. 50-152). The duodenal wall may appear thickened with inflammatory infiltration in the pericholecystic space and focal or diffuse loss of interface between the gallbladder and duodenum. MPR coronal or sagittal CT images using MDCT may provide more information regarding the exact number, size, location, and morphology of the stones than axial images.461

Bowel Obstruction after Abdominal Surgery for Malignancy.  Bowel obstruction in patients who have undergone abdominal surgery for malignancy remains a vexing problem in which the cause of obstruction may be benign postoperative adhesions, a focal malignant deposit, or peritoneal carcinomatosis. Approximately 25% of these cases are due to benign causes, a resectable or easily bypassed local recurrence, or a potentially curable new primary tumor.399,634 For this reason, patients with cancer who do not have proven recurrent tumor or peritoneal carcinomatosis should undergo aggressive treatment with early surgery in response to failure of conservative management.99 Benign obstruction is more likely if pelvic irradiation was used in the management of the primary tumor, whereas the chance of malignant obstruction is increased if the patient had known metastatic cancer  or if the primary cancer is in an advanced stage or of gynecologic origin.399,634

A

1685

When one is predicting the causes of intestinal obstruction, it is helpful to consider the time interval from the cancer surgery to the development of intestinal obstruction. The common causes of bowel obstruction during the early postoperative period (<3 months after surgery) are adhesions and bands, volvulus, or other benign conditions such as stricture due to postsurgical fibrosis.298,809 The site of intestinal obstruction is also another useful indicator in differentiating malignant from benign obstruction. The presence of colonic obstruction favors the diagnosis of malignant obstruction. According to a series by Megibow and colleagues, CT was successful in determining the cause of obstruction in 12 of 13 patients.568 A later report by Ha and associates showed that the diagnostic accuracy of CT in differentiating benign from malignant obstruction was 84%.298 The most diagnostic and specific CT findings to suggest malignant obstruction are a mass at the obstructed or prior surgical site and diffuse omental infiltration or cake (Fig. 50-153); however, if intestinal obstruction develops with implanted lesions smaller than 5 mm, they can be mistaken for a benign obstruction. Lymphadenopathy in the paraaortic space or along the upper major vessels of the aorta favors the diagnosis of malignant obstruction, whereas the presence of hematogenous metastases to other organs (e.g., liver, lung, bone) may not.298 In some instances, observation of bowel loop configuration at the site of obstruction on CT is also valuable. The transition zone at the site of obstruction is commonly abrupt in malignant obstruction and smooth in benign obstruction, whereas bowel wall thickening at obstructed sites is commonly irregular in malignant obstruction and smooth in benign obstruction. A large amount of ascites can be seen in both conditions.

MESENTERIC ISCHEMIA Mesenteric ischemia is a devastating disease process that frequently poses a challenge to clinicians. This disease can be classified as acute or chronic and has a venous or arterial origin. Bowel viability is not compromised in the chronic form because collateral blood flow will occur. In the acute form, infarction of bowel is more common. Ischemia of arterial origin occurs much more often than venous disease.83 The conditions most frequently leading to mesenteric ischemia include vascular occlusion due to arterial or venous disease and hypoperfusion associated with nonocclusive vascular disease. Many other diseases may also cause such vascular changes, including bowel

B FIG 50-152  Gallstone ileus. A, CT scan shows a high-attenuated stone (arrowheads) obstructing the small bowel loop. B, CT scan at higher level shows contracted and gas-filled gallbladder (arrow) with diffuse pericholecystic infiltration.

1686

PART II  CT and MR Imaging of the Whole Body

M

FIG 50-153  Malignant small bowel obstruction due to peritoneal tumor seeding after surgery for gastric cancer. CT scan shows irregular mass (M) obstructing the small bowel loop. (From Ha HK, et al: Usefulness of CT in patients with intestinal obstruction who have undergone abdominal surgery for malignancy. AJR Am J Roentgenol 171:1587– 1593, 1998.)

FIG 50-154  Mesenteric arterial occlusion due to thrombus. Coronal

obstruction, neoplasm, vasculitis, inflammatory conditions, trauma, and iatrogenic causes such as drug or radiation therapy. Irrespective of the cause of the ischemic insult, the end results are similar and range from transient alteration of bowel activity to transmural hemorrhagic necrosis. On pathologic examination, acute bowel ischemia has three stages.301,514 The first stage (reversible ischemic enteritis) is characterized by mucosal necrosis, mucosal erosions and ulceration, and hemorrhage. The second stage occurs when ischemic damage extends to submucosal and muscular layers. In contrast to stages I and II, which represent partial mural bowel ischemia, transmural bowel wall necrosis (bowel infarction) develops in stage III. However, the histopathologic manifestations of bowel ischemia can differ depending on their pathogenesis, the severity of the involvement, and the presence or absence of hemorrhage or superinfection. Although visceral angiography remains the most valuable method of diagnosis and the immediate therapy of acute arterial occlusion, CT has recently been regarded as a valuable noninvasive tool among a variety of diagnostic methods because it helps in detecting changes in the vessels and bowel and reveals other ancillary abdominal findings. CTA using MDCT, which is probably the most frequently used technique in patients suspected of having mesenteric ischemia, may replace angiography for diagnostic purpose. The most valuable advantages of using CTA include not only obtaining information about the presence or absence of thromboembolism and/or vascular narrowing in the mesenteric vessels but also demonstrating hemodynamic changes in the involved bowel segments (Fig. 50-154). Addition of 3D sagittal, coronal, and curved MPR images helps in easy and rapid identification of the site, level, and cause of mesenteric ischemia. In performing CTA, neutral or negative contrast agent should be used because positive oral contrast may obscure visualization and concomitant detection of diagnostically significant bowel wall enhancement patterns. Of the multiphase dynamic sequences in CTA, inclusion of nonenhanced CT in addition to both arterial- and portal-phase CT is recommended because nonenhanced CT has advantages not only in differentiating the hyperattenuating bowel wall caused by intramural hemorrhage from that caused by hyperemia885 but also in easy detec-

MIP CT image shows abrupt arterial occlusion (arrow) in the ileal branch of the mesenteric artery.

tion of atherosclerotic vascular wall calcification or blood clot occluding the mesenteric vessels. In addition, acquisition of arterial-phase CT scans should be included because one of the significant CT findings of bowel wall enhancement (i.e., hyperattenuation or hypoattenuation) in establishing the diagnosis of mesenteric ischemia and differentiating it from other nonischemic conditions (i.e., hypoattenuation, ischemia vs. hyperattenuation, infection) is best demonstrated in these images.

Mesenteric Ischemia of Arterial Origin In the radiologic literature, reports of sensitivity of CT in diagnosis of acute mesenteric ischemia have varied. In an early report,805 the sensitivity of CT was poor (only 39%); more updated series using highquality infusion techniques434,830 have shown much higher sensitivity (ranges, 64%-82%), and the most recent series using CTA MDCT has a sensitivity of 96%.430 This much-improved diagnostic accuracy results mainly from higher detection rates of specific signs of mesenteric ischemia, such as thromboembolism in the mesenteric vessels and lack of bowel wall enhancement (Fig. 50-155). If good contrast enhancement is achieved in the mesenteric vessels, most of the thromboemboli in the large branches of the superior mesenteric vessels are easily detected on CT. However, it is likely that significant limitation is encountered in detecting small thromboemboli in the peripheral arterial branches. In fact, angiography appears to be not much better than CT in cases of thromboembolism in the peripheral or terminal mesenteric branches. Occasionally the level of vascular occlusion in the mesenteric vessel in the presence of ischemic bowel changes is obscure even in goodquality CT images. In such instances, consideration of the length of involved bowel segment will help localize the thromboembolism in the mesenteric vessels; if ischemia develops at the proximal mesenteric artery, a larger small bowel segment may be involved, and in cases of peripheral branches, the involved bowel segment is relatively very short.

CHAPTER 50  Gastrointestinal Tract

1687

* *

* *

B

A

FIG 50-155  Focal intestinal infarction of the jejunum due to superior mesenteric artery (SMA) thrombosis in an 84-year-old woman. A and B, Coronal CT scans show multifocal areas of luminal narrowing (arrowheads in A) of the SMA. The infarcted jejunal loop (asterisks) shows lack of bowel wall enhancement, with evidence of intramural gas, perienteric haziness, and poor opacification of regional vasa recta.

B

A

FIG 50-156  Reversible intestinal ischemia due to embolus (lipiodol), which is mistakenly displaced into the superior mesenteric artery. A, Axial CT scan shows a hyperattenuating material in the ileocolic artery (arrow) 1 day after transarterial chemoembolization for hepatocellular carcinoma. B, Axial CT scan caudal to A shows nearly isoattenuating and concentric bowel wall thickening (arrows) in the ileum, without any perienteric changes. Wall thickening was normalized without surgery on follow-up CT (not shown).

Specific CT signs for establishing the diagnosis include throm­ boembolism in the mesenteric vessel, lack of bowel wall enhancement, intramural gas, portal venous gas, and ischemia of other  organs. Nonspecific signs are bowel dilatation, bowel wall thickening, bowel obstruction, mesenteric edema and vascular engorgement, and ascites.11,434,805,830

Abnormal Bowel Wall Enhancement.  In general it is considered that the bowel wall enhancement pattern in the involved bowel segment is the most important finding in establishing the diagnosis of mesenteric ischemia and also in differentiating reversible mesenteric ischemia (mucosal necrosis, pathologic grade of stage I) from infarction

(transmural bowel wall necrosis, pathologic grade of stage III). In addition, the bowel wall enhancement patterns are different according to which mesenteric vessels are involved (i.e., SMA or vein). In cases of mesenteric arterial occlusion, on arterial-phase CT, both reversible mesenteric ischemia (stage I) and mesenteric infarction (stage III) show hypoattenuation in the involved bowel segments, but on portal or enteral phase CT scans, the involved bowel shows persistence of hypoattenuation in mesenteric infarction (stage III) (see Fig. 50-155) and nearly isoattenuation in the reversible mesenteric ischemia (stage I) (Fig. 50-156). In stage II-type mesenteric ischemia, the involved bowel appears as hypoattenuation on arterial phase and slight hyperattenuation on portal phase.

1688

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-157  Intestinal infarction due to emboli in the superior mesenteric artery (SMA) in a patient with a history of atrial fibrillation. A, Axial CT scan shows lack of bowel wall enhancement of the small intestine, with nearly complete loss of bowel wall definition (arrowheads). Intramural gas can be seen in the involved bowel wall. A hypoattenuating lesion (arrow) occluding the lumen of the SMA represents embolus. B, Visceral mesenteric angiograph shows abrupt complete SMA occlusion (arrow). (From Kim JY, et al: Intestinal infarction secondary to mesenteric venous thrombosis: CT-pathologic correlation. J Comput Assist Tomogr 17:382–385, 1993.)

Bowel Wall Thickening.  Although bowel wall thickening is regarded as a nonspecific finding, paper-thin bowel or nearly invisible bowel wall can be a specific finding for the diagnosis of mesenteric infarction or gangrenous bowel (see Fig. 50-155).

Pneumatosis Intestinalis and Portomesenteric Venous Gas.  Pneumatosis intestinalis and portomesenteric venous gas are also considered highly specific and ominous signs of mesenteric infarction (Fig. 50-157), but these findings occur in only 6% to 28% and 3% to 14% of cases of mesenteric infarction, respectively.11,434,830 Furthermore, the presence of only one of these findings should be cautiously interpreted because they can be seen in many nonischemic conditions such as infectious, inflammatory, neoplastic, or iatrogenic mucosal injury due to increased intraluminal pressure and asthma.335,886 In one series, only nonspecific CT findings of mesenteric ischemia, such as diffuse bowel dilatation, bowel wall thickening, ascites, and mesenteric haziness and vascular engorgement, were observed in 36% of patients with proven mesenteric ischemia.830 Furthermore, in many instances, it is difficult to obtain high-quality images because of the generally poor condition of patients with mesenteric ischemia; thus specific CT findings can be easily missed. In such patients with nonspecific CT signs, it is difficult to differentiate mesenteric ischemia from other GI tract disease (e.g., inflammatory bowel disease, acute abdominal conditions). A high index of suspicion may be the key for early identification of acute mesenteric ischemia, especially in patients older than 50 years who have chronic heart disease and long-standing congestive heart failure and who show atheromatous vascular wall calcifications on CT studies.

Thromboembolism.  Occlusive intestinal ischemia is usually caused by either arterial or venous disease, although arterial occlusion accounts for most cases.83 More than 95% of patients with embolic occlusion of the SMA have documented histories of cardiac disease, with the emboli usually originating from a left atrial or ventricular mural thrombus. The embolus lodges at points of normal anatomic narrowing immediately distal to the origin of a major arterial branch. Most patients with SMA thrombosis also have a history of coronary, cerebrovascular,

or peripheral arterial insufficiency, and thrombosis occurs at points of severe atherosclerotic narrowing, most often at the origin of the SMA.83 Before symptoms are expected, the degree of arterial narrowing must be severe, with narrowing of the cross-sectional area by 50% to 80%.174 In addition to arterial luminal narrowing, systemic blood pressure and the degree of collateral circulation also play a role in determining the severity of a patient’s symptoms and extent of ischemic insult.46 If collateral circulation is adequate, bowel infarction may not develop even in cases with total occlusion of the main SMA or inferior mesenteric artery.

Mesenteric Ischemia of Venous Origin Acute mesenteric venous thrombosis (MVT) is an uncommon but often lethal form of bowel ischemia, accounting for 5% to 15% of all mesenteric ischemic events.451 Because of its rare occurrence and the nonspecificity of associated abdominal symptoms, delay in diagnosis is frequent and contributes to its high mortality rate, which ranges from 20% to 50%.2,451,730 Recent advances in imaging modalities have facilitated an early diagnosis, allowing more time for nonoperative management and leading to dramatic changes in management. Recently anticoagulation therapy has been instituted as the first-line treatment,312,451,596,926 and subsequent management decisions are largely determined by the severity of bowel ischemia.451,900 Important predisposing factors for developing superior mesenteric venous thrombosis include inherited or acquired conditions causing a hypercoagulable state, such as protein S or C deficiency or antiphospholipid antibody syndrome. Other causes include portal hypertension, abdominal inflammatory disease, previous surgery, trauma, and oral contraceptive use. Many cases are idiopathic. On CT the hallmark of mesenteric ischemia caused by superior mesenteric venous thrombosis is marked bowel wall thickening with severe mesenteric haziness or edema and a considerable degree of collateral vasculature in the mesentery as well as in the retroperitoneum (Fig. 50-158). The thickened bowel wall also shows a target appearance with a hyperattenuating inner layer, an isoattenuating or hypoattenuating middle layer, and a hyperattenuating outer layer. Pathologically the inner and middle layers correspond to alternating areas of

CHAPTER 50  Gastrointestinal Tract

1689

A

B FIG 50-158  Intestinal infarction due to superior mesenteric vein thrombosis. A, CT scan shows marked concentric bowel thickening (arrowheads) of the jejunum with alternating layers of high and low attenuation and dense enhancement of the peripheral wall. Also noted are evidence of regional mesenteric haziness and poor opacification of mesenteric vessels. B, Histologic section of the resected bowel shows marked hemorrhagic and edematous thickening of submucosa; the inner hyperattenuating layer on CT corresponds to the submucosal hemorrhage, and the middle hypoattenuating layer corresponds to the submucosal edema.

hemorrhage and edema in the submucosa of the intestinal wall, and the outer layer represents the change mainly in the serosal layer of the intestine.404 Because of these unique findings, the diagnosis is usually made easily by CT. However, great difficulty is encountered in differentiating medically treatable mesenteric ischemia from mesenteric infarction, which requires surgical intervention. According to our study, CT findings suggesting transmural infarction include homogeneous and/or lack of bowel wall enhancement, indistinct outer margins of the bowel wall, large amount of ascites, maximal lumen diameter of 20 mm or greater, and diffuse mesenteric haziness473 (Figs. 50-159 and 50-160). Although not included in this study, CTA findings of hypoattenuation in the involved bowel on arterial phase and hyperattenuation on portal phase favor the diagnosis of mesenteric infarction. In equivocal cases, in making a decision, follow-up CT after a short period (≈2-3 weeks) of anticoagulant therapy is strongly recommended. It also should be remembered that the incidence of a focal area of bowel infarction (not segmental) within the ischemic segment is not rare. Therefore considerable reduction of bowel wall thickening on CT following medical therapy does not always indicate complete recovery of the condition, and there is still a risk of bowel perforation–caused focal infarction; on CT, patients at such risk show hyperattenuation on portal phase (Fig. 50-161). MRI has been attempted in cases of suspected mesenteric ischemia. Nonenhanced images are comparable with CT for demonstrating changes in the bowel and adjacent mesentery, but a significant problem is posed in regard to detecting small thromboemboli in the peripheral mesenteric arterial branches (Fig. 50-162).295 As in CT, use of a contrast infusion study may be mandatory not only to assess the dynamic changes in the bowel wall but also to detect small thromboemboli. In

FIG 50-159  Intestinal infarction due to superior mesenteric vein thrombosis. Axial CT scan shows marked bowel wall thickening (white arrows) of the jejunum with poor contrast enhancement. There is evidence of perienteric haziness and small thrombi (black arrows) in the regional mesenteric veins. A small amount of ascites is also seen. Surgery confirmed transmural infarction.

addition, some researchers have attempted to use cine phase-contrast MRI to quantitate mesenteric vascular blood flow.507 In the study by Li and colleagues, the percentage of change in SMA blood flow 30 minutes after meal intake seems to be a promising criterion for differentiating healthy subjects from patients with chronic mesenteric ischemia.507 However, further study with a large number of patients is warranted.

1690

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-160  Superior mesenteric vein (SMV) thrombosis with edema and congestion of the intestine. A and B, CT scans show SMV thrombosis (arrow in A) with concentric and edematous wall thickening (arrowheads) in the jejunum. The involved jejunum shows a target appearance along with diffuse mesenteric haziness. A surgical specimen obtained after bowel resection revealed marked edema and congestion in the thickened bowel with ischemic changes limited to the intestinal mucosa (not shown).

A

B

C FIG 50-161  Intestinal infarction due to superior mesenteric vein (SMV) thrombosis. A, Axial CT shows diffuse concentric bowel wall thickening of the jejunum, with diffuse mesenteric haziness and ascites. Hypoattenuating thrombus (not shown) exists in the SMV. B, Axial CT obtained 1 week after anticoagulant therapy shows regression of both bowel wall thickening and mesenteric haziness. Collateral vessels are developed in the mesentery and retroperitoneum. C, Axial CT 4 weeks after anticoagulant therapy shows considerable regression of bowel wall thickening, but the involved bowel (arrows) shows hyperattenuation with diffuse perienteric infiltration. Surgery confirmed transmural intestinal necrosis with evidence of peritonitis due to bowel perforation.

CHAPTER 50  Gastrointestinal Tract

A

B

C

D

1691

FIG 50-162  Intestinal ischemia due to multiple arterial emboli. CT scan (A) shows a hypoattenuated occlusive lesion (arrow) in the root of the superior mesenteric artery (SMA), which is also well demonstrated on axial T2-weighted FISP MRI (B). CT scan (C) at a lower level shows a small embolus (arrow) in the branch of the SMA; however, the embolus is difficult to see on true-FISP MRI (D).

Nonocclusive Mesenteric Ischemia In patients with nonocclusive mesenteric ischemia, the mesenteric arteries and veins are patent, but flow through them is too slow to deliver enough oxygenated blood to the bowel. This condition is most commonly related to primary cardiac dysfunction, such as cardiac failure, arrhythmia, myocardial infarction, sepsis, trauma, and vascular insufficiency. Most cases of colonic ischemia belong to this group. Flow to the viscera is reduced by progressive vasoconstriction, and persistent vasoconstriction of the splanchnic vasculature results in intestinal hypoxia.83 Shock bowel, caused by prolonged hypoperfusion due to hypovolemic shock, is a subtype of nonocclusive mesenteric ischemia. It is transient and resolves with restoration of hypovolemia. CT scans may show diffuse bowel wall thickening and contrast enhancement exclusively in the small intestine (Fig. 50-163).591 Other reported CT manifestations include diminished caliber and increased enhancement of the inferior vena cava and aorta as well as intense contrast enhancement of the kidneys and mesentery.801 Shock bowel should be distinguished from diffuse edema of the small bowel that accompanies central venous hypertension due to aggressive volume resuscitation in a trauma setting. Such cases demonstrate other signs of elevated central venous pressure, including an enlarged inferior vena cava, periportal lymphedema, and accumulation

FIG 50-163  Shock bowel in a 26-year-old woman after cesarean section. CT scan shows diffuse bowel wall thickening in the ileum (arrows), with a large intraperitoneal fluid collection caused by hemoperitoneum. (From Rha SE, et al: CT and MR imaging findings of bowel ischemia according to the various primary causes. Radiographics 20:29– 42, 2000.)

1692

PART II  CT and MR Imaging of the Whole Body

of retroperitoneal fluid,787 and do not exhibit increased contrast enhancement of the small bowel wall.

VASCULITIS Vasculitis can affect blood vessels of all sizes, resulting in necrosis and inflammation. The extent and clinical course depend on the size and location of the vessels involved.775 In the 1990s the predominant type of affected vessel was generally accepted as the criterion for classification.371 According to this system, vasculitis can be classified into three types: 1. Large vessel vasculitis includes Takayasu’s arteritis and giant cell arteritis. 2. Medium-sized vessel vasculitis includes polyarteritis nodosa (see later) and Kawasaki’s disease. 3. Small vessel vasculitis comprises microscopic polyangiitis (hypersensitivity vasculitis), Wegener’s granulomatosis, ChurgStrauss syndrome, Henoch-Schönlein syndrome, SLE, rheumatoid vasculitis, and Behçet’s syndrome. Systemic vasculitis affecting the GI tract is uncommon. The frequency of GI involvement depends on the type of vasculitis, with the highest incidence (50%-70%) in cases of polyarteritis nodosa. With larger-vessel involvement, abdominal manifestations may be indistinguishable from those of mesenteric ischemia caused by emboli or thrombosis, except for associated evidence of systemic disease. Inflammation of medium-sized arteries may also lead to formation of aneurysms, as commonly occurs in polyarteritis nodosa, and their rupture may cause GI or intraabdominal hemorrhage. The vasa recta and intramural arteries and arterioles may be affected in virtually all disorders associated with systemic vasculitis. Ulceration and stricture formation are common, but perforation is less common with small vessel involvement. Until recently, the medical literature has not sufficiently dealt with the role of radiology in distinguishing ischemia caused by mesenteric vasculitides from ischemia due to other causes. If only radiologic findings are used, it may be difficult to determine the primary causes of GI manifestations. Knowledge of the systemic clinical manifestations in patients with mesenteric ischemia not only would suggest the pos-

R 1 7 1

sibility of vasculitis but also would be helpful in establishing the specific diagnosis. In our experience, CT helps suggest the diagnosis of vasculitis. The possibility of vasculitis should be considered in the following situations: • When multiple abdominal organs are involved • When mesenteric ischemic changes are suspected in young patients • When involvement of the stomach, duodenum, or rectum coexists with small bowel and large bowel changes • When bowel wall changes concomitantly involve both the small and large intestines371 Ancillary findings occurring in other organs may also suggest the diagnosis because a high incidence of genitourinary tract involvement (nephritis, cystitis, or hydronephrosis) or splenomegaly has also been reported.101

Polyarteritis Nodosa Polyarteritis nodosa (Fig. 50-164) is a fibrinoid necrotizing vasculitis that mainly involves the medium-sized muscular arteries. A characteristic finding is multiple aneurysm formation in these vessels, which are seen in 50% to 60% of cases; the individual lesions affect the branch points and bifurcation of arteries. The kidney is most commonly involved (80%-90%), followed by the GI tract (50%-70%), liver (50%-60%), spleen (45%), and pancreas (25%-35%).370,847 In the GI tract, the small intestine is most commonly affected, followed by the mesentery and colon. Approximately two thirds of patients with polyarteritis nodosa have abdominal pain, nausea, vomiting, or other GI symptoms in association with organ damage due to ischemia and infarction; GI hemorrhage occurs in roughly 6% of cases, bowel perforation in 5%, and bowel infarction in 1.4%.102,370 Therapy with corticosteroids and cyclophosphamide results in remission or cures in 90% of affected patients.775 The diagnosis can be suggested by the presence of aneurysms up to 1 cm in size within the renal, mesenteric, and hepatic vasculature on angiography. However, these visceral aneurysms are not always a pathognomonic finding; they are also seen in necrotizing angiitis associated with drug abuse305 and other vasculitides such as Wegener’s

B

R

A

B FIG 50-164  Polyarteritis nodosa. A, CT scan shows bowel wall thickening of the ileum and ascending and descending colon (arrows) with evidence of the target sign. Also noted are a large amount of ascites and bilateral ureteral dilatation (arrowheads). B, CT scan shows rectal wall thickening (arrows) with the target sign. The bladder (B) is also diffusely thickened. (From Ha HK, et al: Radiologic features of vasculitis involving the gastrointestinal tract. Radiographics 20:779–794, 2000.)

L 1 7 0

L

CHAPTER 50  Gastrointestinal Tract granulomatosis and SLE.847 Without angiography, the correct diagnosis is established by biopsy of subcutaneous nodules or skeletal muscle.557 CT may help in determining the extent of involvement in multiple abdominal organs. When the GI tract is involved, CT manifestations do not differ from intestinal ischemia resulting from other causes (see Fig. 50-164).369

Microscopic Polyangiitis Microscopic polyangiitis is characterized by pauci-immune necrotizing small vessel vasculitis without clinical or pathologic evidence of necrotizing granulomatous inflammation.371 The lesions are thought to represent a hypersensitivity reaction to drugs (e.g., penicillin), microorganisms, heterologous proteins, and tumor antigens,775 and they involve the skin, mucous membranes, lung, brain, heart, GI tract, kidneys, and muscle. In many instances the lesions are limited to the skin (cutaneous leukocytoclastic vasculitis). When the GI tract is involved, radiologic findings do not differ from findings of other types of vasculitis involving similar-size vessels, and bowel infarction or perforation is rarely reported (Fig. 50-165).404 In contrast to the findings of polyarteritis nodosa, visceral angiographic results are usually normal and do not reveal microaneurysms.

1693

predominate or precede skin lesions, the syndrome may mimic a number of acute abdominal diseases, resulting in unnecessary laparotomies. GI hemorrhage is confined mostly to the mucosa and submucosa, and full-thickness necrosis and perforation of bowel loop are rare. Therefore most GI manifestations are self-limited and without residua, and only 3% to 5% of patients develop bowel infarction, perforation, or irreducible intussusception.550 No specific findings are seen on radiologic examinations, although the CT finding of multifocal bowel wall thickening with skipped areas under the clinical backgrounds is important to establish a diagnosis (Fig. 50-166).372

* *

*

Henoch-Schönlein Syndrome Henoch-Schönlein syndrome is one of the subgroups of hypersensitivityrelated acute small vessel vasculitis.586 Although the cause is unknown, immunization, insect bites, medications, infections, and certain foods sometimes play a role in the development of this disease. The syndrome occurs most commonly in children between 3 and 10 years of age, but in some series, 30% of affected patients have been older than age 20. The diagnosis is based on characteristic clinical signs and symptoms: rash, arthritis involving the large joints, colicky abdominal pain, GI bleeding, and hematuria.586 The GI manifestations are thought to be related to edema and intramural hemorrhage. When GI symptoms

*

A

FIG 50-165  Microscopic polyangiitis (hypersensitivity vasculitis). CT scan shows bowel wall thickening of the ileum (asterisks) with the target sign. A small amount of ascites is also visible. (From Ha HK, et al: Radiologic features of vasculitis involving the gastrointestinal tract. Radiographics 20:779–794, 2000.)

*

B FIG 50-166  Henoch-Schönlein purpura with bowel infarct in a 6-year-old boy with symptoms of severe colicky abdominal pain, vomiting, and polyarthralgia. The patient had purpura on both buttocks on physical examination. The diagnosis was made on the basis of histopathologic results following surgery (segmental ileal resection). A, CT scan shows bowel wall thickening (asterisks) of the ileum, with the target sign. Ascites is also noted. B, Small bowel follow-through shows luminal narrowing and circular fold thickening in both the jejunum (arrows) and the ileum (arrowheads), with skipped areas. (From Ha HK, et al: Radiologic features of vasculitis involving the gastrointestinal tract. Radiographics 20:779–794, 2000.)

1694

PART II  CT and MR Imaging of the Whole Body

Systemic Lupus Erythematosus SLE is an autoimmune disorder affecting the musculoskeletal system, kidneys, GI tract, or skin.564 Local deposition of antigen-antibody complexes or antibodies inducing necrotizing vasculitis is the presumptive cause of this disease. Most patients range between 16 and 41 years of age, with the most common occurrence during childbearing years. A diagnosis of SLE can be made with 98% specificity and 97% sensitivity if at least 4 of 11 diagnostic criteria are present at any time during the course of disease828; these criteria include malar rash; discoid rash; photosensitivity; oral ulcers; arthritis; serositis; antinuclear antibodies; and renal, neurologic, hematologic, and immunologic disorders. SLE may involve any part of the GI tract from the esophagus to the colon, but the territory of the SMA is more preferentially affected. Other sites (e.g., stomach, duodenum, colon, rectum) may also be involved. Inflammation of the small blood vessels of the gut produces a spectrum of complications, including intestinal ischemia, hemorrhage, ileus, ulceration, infarction, and perforation.432 CT facilitates more accurate diagnosis of mesenteric arteritis. Common CT findings include dilated bowel, focal or diffuse bowel  wall thickening, engorged mesenteric vessels with a comblike arrangement (comb sign), ascites, and lymphadenopathy (Figs. 50-167 and 50-168).101 In most cases the segments of bowel wall thickening are multifocal and not confined to a single vascular territory, because mesenteric vasculitis may affect several vessels simultaneously.431 Rather than bowel wall changes, the ancillary findings occurring in

other organs may suggest the diagnosis because a high incidence of genitourinary tract involvement (lupus nephritis, cystitis, or hydronephrosis) has been reported.6 Hydronephrosis also occurs in other types of vasculitis (e.g., polyarteritis nodosa, Henoch-Schönlein syndrome) and seems to be caused either by detrusor muscle spasm with subsequent vesicoureteral reflux or by fibrosis of the ureterovesical junction.93,425

Behçet’s Syndrome Behçet’s syndrome is a well-recognized nonspecific necrotizing vasculitis that affects many organs; its well-known clinical manifestations include orogenital ulceration, uveitis, arthritis, and neurologic and GI involvement.18,553 Involvement of the GI tract has been reported in 10% to 40% of patients with Behçet’s syndrome792; the usual site is the ileocecal area, especially the terminal ileum. The hallmark of the syndrome is the presence of large, deeply  penetrating ulcerations of the submucosa, muscle layer, or entire intestinal wall. As a result, a high incidence of complications such as perforation, hemorrhage, fistula, and peritonitis has been reported.389 Appropriate treatment, therefore, is of great importance in avoiding these complications. Barium study shows a large ovoid or geographic ulcer with marked thickening of the surrounded intestinal wall at the sites of involvement, although in certain instances small, multiple, discrete, and punchedout ulcers can be seen. CT has proved effective not only for detecting primary lesions but also for excluding extraluminal complications.296 On CT scans the involved bowel shows concentric bowel wall

*

A

B

C FIG 50-167  Systemic lupus erythematosus. A, CT scan shows edematous gastric wall thickening (arrows). B and C, The duodenal (asterisk) and ileal (arrows) walls are also similarly thickened. (From Ha HK, et al: Radiologic features of vasculitis involving the gastrointestinal tract. Radiographics 20:779–794, 2000.)

CHAPTER 50  Gastrointestinal Tract

*

*

1695

R

1 120 200 75

120 40

B

A

0

FIG 50-168  Systemic lupus erythematosus in a 26-year-old woman. A, CT scan shows bowel wall thickening of the third and fourth portions of the duodenum (asterisks) with the target sign. Splenomegaly and a small amount of ascites are also noted. B, CT scan shows bowel wall thickening of the small intestine with the target sign. Mesenteric vascular engorgement (arrows) and a large amount of ascites are also noted.

7 .38 2.5 6

P

F 1.50

A

B FIG 50-169  Behçet’s syndrome. A, Double-contrast barium enema shows multiple discrete ulcers with thickened mucosal folds (arrows) at the terminal ileum. B, CT scan shows both bowel wall thickening (arrows) and a polypoid masslike lesion (P) at the ileocecal region. (From Ha HK, et al: Radiologic features of vasculitis involving the gastrointestinal tract. Radiographics 20:779–794, 2000.)

thickening or a polypoid mass (Figs. 50-169 and 50-170). The polypoid masslike lesions may reach a diameter of 8 to 10 cm in some cases, with central ulceration in one third of these cases. Most of the thickened bowel wall may show marked contrast enhancement, presumably owing to vasculitis and perivasculitis.296 Other CT findings include minimal lymphadenopathy of less than 1 cm in diameter and fibrofatty proliferation. In patients with no complications, perienteric or pericolonic infiltrations are usually absent or minimal. In contrast, the presence of severe perienteric or pericolonic infiltrations raises the possibility  of complications such as microperforation and localized peritonitis (Fig. 50-171).

DRUG-INDUCED ENTEROPATHY A variety of medications can cause injury to any part of the GI tract by different mechanisms. Although GI reactions may be minor in many instances, serious drug-induced complications such as bowel ischemia, perforation, peritonitis, or shock are not uncommon and sometimes fatal. Oral contraceptives and estrogens are the well-known drugs causing an increased incidence of thromboembolic complications in the superior and inferior mesenteric arteries, celiac artery, superior and inferior mesenteric veins, and hepatic vein.233 The duration of drug therapy with these agents before development of ischemic bowel disease ranges from 10 days to 11 years.233 Many instances of oral contraceptive

1696

PART II  CT and MR Imaging of the Whole Body

*

*

1 250 20

FIG 50-170  Recurrent Behçet’s syndrome at the surgical anastomosis. CT scan shows concentric bowel wall thickening (asterisks) at the anastomotic site after right hemicolectomy. Perienteric infiltration and regional mesenteric vascular engorgement are also noted. (From Chung SY, Ha HK, Kim JH, et al: Radiological findings of Behçet’s syndrome involving the gastrointestinal tract. Radiographics 21:911–924, 2001.)

Amromin and Solomon classified necrotic enteric lesions (necrotizing enteropathy) in 63 patients with treated leukemia or lymphoma into the following four types, according to the pathogenic mechanisms: (1) shock with a superficial variety of necrosis, (2) therapeutic necrosis of tumor or leukemic infiltrates of the intestinal mucosa, (3) hemorrhage with necrosis of GI mucosa, and (4) traumatic erosions.13 In their second group of patients, necrosis occurred at sites with the most abundant lymphoid tissue, and therefore the terminal ileum or appendix was commonly involved. Because little or no supporting tissue remained—either normal or neoplastic—perforations and peritonitis were common in this group of patients. Among various chemotherapeutic agents, 5-fluorouracil and its analogues are considered to be the most injurious.504 Ischemic enteritis or enterocolitis can be seen in patients with intraarterial infusion of floxuridine for the treatment of hepatic metastasis.395,688 Even though numerous reports regarding the clinical and surgicopathologic aspects of drug-induced enteropathy have been published, studies of imaging results have been limited. This is attributable to the fact that in patients with mild GI symptoms during or following drug therapy of any kind, CT is not usually performed and may be confined to patients with severe symptoms. Furthermore, as symptoms develop with periods of varying intervals after therapy, abnormal CT findings may be commonly confused with other conditions. In our experience, CT findings of drug-induced enteropathy appear to be nonspecific, including mild small bowel fold thickening, focal or diffuse bowel wall thickening, bowel obstruction, pneumatosis intestinalis, and signs of peritonitis (Figs. 50-172 and 50-173). Therefore drug-induced enteropathy may be difficult to distinguish from ischemic or inflammatory bowel diseases. Multifocal involvement or patchy distribution in the GI tract may be one of the important features for suggesting the diagnosis.

RADIATION ENTERITIS

FIG 50-171  Behçet’s syndrome with bowel perforation. CT scan shows intraperitoneal fluid collection and mesenteric infiltration (arrowheads) surrounding the involved bowel loop (arrow). (From Ha HK, et al: Radiologic features of vasculitis involving the gastrointestinal tract. Radiographics 20:779–794, 2000.)

steroid-induced mesenteric ischemia have mimicked Crohn’s disease but are reversible on discontinuation of the agent.834 Digoxin and certain cardiovascular agents504 can also cause bowel ischemia by diminishing splanchnic flow or by stimulating the reninangiotensin axis, leading to vasoconstriction. Other drugs causing bowel ischemia include various diuretics, antihypertensive medications, and ergot alkaloids. Chemotherapeutic agents also produce GI toxicity, ranging from mild clinical symptoms to acute surgical conditions due to peritonitis, perforation, or bowel necrosis. A high incidence of necrotizing lesions of the GI tract in patients with malignancy, lymphoma, or leukemia has been reported. Dosik and associates separated 26 patients with malignancy (most received antineoplastic and antimicrobial regimens and died of necrotizing colitis) into three histologically distinct categories: (1) those with pseudomembranous colitis (69%), (2) those with agranulocytic colitis (typhlitis) (19%), and (3) those with ischemic colitis (12%).180

Radiation enteritis is generally observed in patients who receive abdominal and pelvic radiation with doses of 4500 to 6000 rads (45-60  Gy).758 The latent period between radiation therapy and the development of radiation damage is usually 6 to 24 months, although this interval can be longer than 20 years. Predisposing factors include diabetes, hypertension, atherosclerosis, prior abdominal surgery with adhesive bowel loops, and peritonitis before the radiation therapy. The long-term effects of radiation on the vasculature may cause the involved vessel wall, especially of the small arterioles in the submucosa, to be obliterated with sufficient luminal narrowing, resulting in progressive occlusive vasculitis and diffuse collagen deposition with fibrosis.790 Fewer than 2% of patients who receive radiation therapy ultimately require surgery because of perforation, bleeding, fistula, or stricture formation.89 Although the involved bowel segments depend on the radiation mantle selected, the sigmoid colon, rectum, and terminal ileum are the most common sites affected by chronic radiation enteritis. The diagnosis of chronic radiation enteritis is difficult because patients often present many years after treatment and the disease often simulates recurrent cancer or adhesion. CT findings of radiation enteritis are relatively nonspecific. CT may demonstrate a masslike confluence of adherent bowel loops with thickening of the bowel wall, thickening of adjacent mesentery, or increased density of mesenteric fat (Figs. 50-174 and 50-175).221 The important CT findings that help differentiate chronic radiation enteritis from other conditions, such as recurrent malignant disease, include the distribution of changes within the radiation therapy port and the absence of a mass.

CHAPTER 50  Gastrointestinal Tract

1697

* *

A

*

B FIG 50-172  Chemotherapy-induced enteropathy with diffuse intestinal necrosis in a patient who received chemotherapy for central nervous system lymphoma. A, CT scan shows extensive intrahepatic portal venous gas. B, CT scan shows intestinal pneumatosis (asterisks) and gas in the mesenteric vascular branches. (From Rha SE, et al: CT and MR imaging findings of bowel ischemia according to the various primary causes. Radiographics 20:29–42, 2000.)

FIG 50-174  Acute radiation enteritis in a patient with a history of radia-

FIG 50-173  Chemotherapy-induced enteropathy with bowel necrosis and perforation in a patient receiving chemotherapy for leukemia. CT scan shows bowel wall thickening (arrows) of the ileum with free gas (curved arrow) in the pelvic cavity. Note the intraperitoneal fluid collection, which proved to be hemoperitoneum at surgery. (From Rha SE, et al: CT and MR imaging findings of bowel ischemia according to the various primary causes. Radiographics 20:29–42, 2000.)

OBSCURE GI BLEEDING Incidence Obscure GI bleeding is defined as recurrent or persistent bleeding for which no obvious etiology has been identified by standard endoscopic examinations (routine upper endoscopy and colonoscopy).931 Of the approximately 10% to 20% of patients with GI bleeding who do not have a bleeding source identified by standard endoscopic examinations, only half of them have recurrent bleeding. Therefore about 5% of patients with GI bleeding have recurrent bleeding of obscure origin,

tion therapy for periureteral metastases from rectal cancer. CT scan shows bowel wall thickening (arrows) of the ileal loops, with the target appearance confined to the radiation port. (From Rha SE, et al: CT and MR imaging findings of bowel ischemia according to the various primary causes. Radiographics 20:29–42, 2000.)

leading to extensive and repetitive testing.509,855 Although the source of obscure GI bleeding is located most commonly in the small bowel, lesions missed or underestimated during initial endoscopic examination comprise a considerable proportion of bleeding sources.509,739,855,931 In addition, pancreaticobiliary disease causing hemobilia can also be a cause of obscure GI bleeding. Therefore repeated endoscopy or investigation of the pancreaticobiliary system can be helpful to identify lesions responsible for obscure GI bleeding. Various small bowel lesions can cause obscure GI bleeding (Box 50-1). Among them, angiodysplasias are the most common, followed by small bowel tumor.855 However, frequencies of these lesions are variable depending on patients’ age: Meckel’s diverticula are the most common cause of small bowel bleeding in patients younger than 25 years. Tumors are common abnormalities in patients between ages 30 and 50. Angiodysplasias predominate in the elderly.509

1698

PART II  CT and MR Imaging of the Whole Body

FIG 50-175  Perforated radiation enteritis in a patient with a history of radiation therapy for cervical cancer. CT scan shows bowel wall thickening of the ileum with a large amount of ascites and free gas (not shown).

BOX 50-1  Causes of Small Bowel Bleeding Vascular Lesions • Vascular ectasia • Telangiectasia (hereditary hemorrhagic telangiectasia) • Hemangioma • Congenital vascular malformation • Ectopic varices Tumors Miscellaneous • Medication related • Infections • Crohn’s disease • Aortoenteric fistula • Meckel’s diverticulum • Jejunoileal diverticula • Ischemia • Small bowel polyposis

Diagnosis

Radiography.  Small bowel radiography has been frequently used in investigating the source of small bowel bleeding. However, the diagnostic yield is disappointing, ranging from 0% to 5.6%140,241,701 for small bowel follow-through and from 10% to 21%592,725 for small bowel enteroclysis.

Multidetector Computed Tomography.  The capability of rapid scanning with isotropic resolution provided by novel MDCT scanners enhanced the diagnostic efficacy of CT in the evaluation of obscure GI bleeding. For the investigation of GI bleeding, CT is performed without oral contrast agent in cases of hemodynamically significant bleeding908,909 or with neutral oral contrast agent for elective workup.343,519,585 Acquisition of an arterial-phase scan is essential to detect active extravasation of contrast material and enhancing vascular lesions.277,908,909 On CT scan the source of GI bleeding can be detected either by depicting extravasation of contrast material (Fig. 50-176) or by demonstrating lesions that are expected to be source of GI bleeding (i.e., vascular lesions, tumors, inflammatory bowel disease) (Fig. 50-177).

For the depiction of active GI bleeding, the diagnostic efficacy of MDCT appears to be comparable to that of conventional angiography. An animal study reported that helical CT can depict active colonic hemorrhage at rate of 0.5 mL/min or less, which is similar to the minimal bleeding rate that can be detected by conventional catheter angiography.450 Yoon and coworkers also found concordant results in their clinical study.909 They reported the sensitivity, specificity, and accuracy of arterial-phase MDCT in the localization of acute massive GI bleeding to be 90.9%, 99%, and 97.6%, respectively, compared with conventional angiography as a reference standard. Therefore MDCT appears to be a promising tool for the localization of active GI bleeding of obscure origin and is potentially helpful to guide further therapeutic approaches. However, the overall diagnostic efficacy of MDCT in evaluation of obscure GI bleeding is still uncertain. Although several studies reported a high diagnostic yield (75%-79%) of MDCT in depicting the source of obscure GI bleeding,197,585 these are retrospective studies and appear to have considerable selection bias resulting from the inclusion of patients with massive bleeding in a large proportion of their study population. Further well-controlled study is required to determine the diagnostic efficacy of MDCT as a routine elective examination tool for patients with obscure GI bleeding.

Capsule Endoscopy.  Capsule endoscopy was reported to be a promising noninvasive diagnostic method for patients with obscure GI bleeding; its yield in patients with obscure GI bleeding ranges from 45% to 66%.140,192,666 A meta-analysis of prospective trials comparing capsule endoscopy with other diagnostic modalities demonstrated superior diagnostic yields of capsule endoscopy compared to push enteroscopy (63% and 28%, respectively) as well as small bowel radiography (67% and 8%, respectively).848 The main advantage of capsule endoscopy over radiologic examinations is its capability to detect small mucosal lesions or angiodysplasia, which constitute most sources of small bowel bleeding but are hardly depicted by radiologic examinations. Therefore capsule endoscopy has gained acceptance as an initial investigation method for obscure GI bleeding.

COLON AND RECTUM NORMAL ANATOMY AND IMAGING TECHNIQUES The colon is visualized on CT by its anatomic location and its typical haustral morphology. It appears as a ringlike or tubular structure, depending on the orientation and position relative to the scanning plane. Recognition of the cecum is facilitated by visualization of the terminal ileum, ileocecal valve, or appendix. The appendix appears as a small ringlike or tubular structure. The presence of air or contrast material in the appendix along with normal-appearing surrounding fat may be indicative of absence of appendicitis. The ascending and descending colons are within the anterior pararenal space and usually are surrounded by homogeneous fatty tissue. The transverse colon is suspended within the peritoneal cavity by the transverse mesocolon. The rectum is about 12 to 15 cm in length. The peritoneum covers the anterior surface of the upper rectum, and the lower two thirds  of the rectum are enveloped by extraperitoneal connective and adipose tissue. The colon is commonly filled with fecal material that appears as gas-containing structures and often mimics a soft tissue mass. In certain instances, repeated CT with the patient in various positions may be required to differentiate fecal material from a polypoid colonic mass. The wall of the normal colon measures 3 mm or less in thickness when the colon is distended with oral contrast material. Walls measuring between 4 and 6 mm in thickness suggest pathology, and a 

CHAPTER 50  Gastrointestinal Tract

A

1699

B FIG 50-176  Active contrast extravasation in a patient with massive GI bleeding due to duodenal ulcer. A, Arterial-phase CT scan shows contrast extravasation (arrow) in the duodenum. B, Common hepatic arteriography also demonstrates contrast extravasation (arrow) from a small duodenal branch (arrowheads).

B

A

FIG 50-177  Small bowel ulceration of uncertain origin. A, Arterial-phase CT scan shows bowel wall thickening and hyperenhancement of inner layer in the ileum (arrows) suggesting the presence of small bowel ulceration. B, Enteroscopic image also demonstrates a broad ulceration (arrows) at the corresponding ileal segment.

thickness of more than 6 mm is definitely abnormal. When the aircontrast technique is used, however, measurements of 3 mm for the rectum and 2 mm for the colon are at the upper limit of normal.922 The luminal margin is usually smooth, being well outlined on the mucosal side by contrast or air. The outer colonic margin is sharply outlined by surrounding homogeneous pericolonic fat. Inflammation adjacent to the colonic wall obliterates the sharpness of the outer colonic margin by increasing the density of the pericolonic fat.

MRI, particularly high-resolution T2-weighted sequence, clearly depicts the details of the rectal wall and perirectal anatomy. On T2-weighted images, rectal wall can be recognized as three different layers: a thin inner line of low-signal intensity that represents the mucosal layer; a middle layer of high-signal intensity that represents the submucosa; and an outer layer of low-signal intensity that represents the muscularis propria (Fig. 50-178).91 The mesorectal fascia, which forms the boundary of the surgical excision plane in total 

1700

PART II  CT and MR Imaging of the Whole Body and expertise of those who perform the procedure,214,229,655 as demonstrated by large discrepancies in diagnostic performances of many studies performed with various degrees of technical completeness and examiner expertise.141,601,741

Bowel preparation

M SM PM

FIG 50-178  Normal rectal and perirectal anatomy on high-resolution T2-weighted MRI. Rectal mucosa (M), submucosa (SM), and muscularis propria (PM) are well discriminated. Mesorectal fascia appears as a thin, low-signal-intensity structure (arrowheads) and fuses with the remnant of urogenital septum making Denonvilliers fascia (arrows).

mesorectal excision, is identified as a thin low-signal-intensity structure on MRI that envelops the rectum and surrounding perirectal fat (see Fig. 50-178). Anteriorly the mesorectal fascia fuses with the remnant of urogenital septum, making the Denonvilliers fascia in the male (see Fig. 50-178) and the rectovaginal septum in the female.762

PATHOLOGIC CONDITIONS Colorectal Adenoma and Carcinoma CT Colonography and Screening for Colorectal Carcinoma.  Most colorectal carcinomas are considered to arise in adenomatous polyps through the adenoma-carcinoma sequence. The development of colorectal carcinomas, therefore, can be prevented by the detection and removal of adenomatous polyps. In addition, an earlier diagnosis of colorectal cancer increases the possibility of survival. Barium enema has played an important role in colorectal cancer screening, particularly in the pre-colonoscopy era. However, it is relatively insensitive for the detection of polyps and cancers compared with colonoscopy, detecting 21%, 42%, and 46% of polyps 5 mm or smaller, 6 to 10 mm, or greater than 10 mm, respectively,889 and 83% of cancers.726 Whereas conventional CT with thick cross-sectional images is insufficient to detect small mucosal polyps and small cancers, CT colonography (also known as virtual colonoscopy) (Fig. 50-179), which combines the recent technical developments in MDCT and 3D imaging, has shown a remarkable accuracy in polyp detection, detecting 85.7% and 92.2% of adenomatous polyps 6 mm or larger and 10 mm or larger, respectively.679 CT colonography is rapidly gaining acceptance as a viable option for colorectal cancer screening and is extending its role for the diagnosis of colonic diseases, promising a substantial role for radiology in the diagnosis of colonic diseases. However, a consistent successful performance of CT colonography is not always easy, requiring fulfillment of the following technical factors

Colonic purgation.  Polyethylene glycol, sodium phosphate, magnesium citrate, and bisacodyl can be used for colonic purgation before CT colonography. Sodium phosphate and magnesium citrate are advantageous in that they typically leave little fluid in the colon, whereas polyethylene glycol leaves a large amount of luminal fluid, which can impair visualization of colonic lesions.901 A smaller dose of sodium phosphate (i.e., typically 90 mL or 45 mL) or magnesium citrate (i.e., typically 250 mL) compared to a large quantity of polyethylene glycol (i.e., typically 236 g of polyethylene glycol mixed with 4 L of water) is another advantage. Sodium phosphate is a high-sodium preparation and in rare cases can cause significant electrolyte disturbances.190,867 It is therefore contraindicated in patients with known renal failure, preexisting electrolyte abnormalities, congestive heart failure, ascites, and ileus.901 Bisacodyl is often used in conjunction with sodium phosphate or magnesium citrate. It stimulates parasympathetic reflexes to induce evacuation of stool.901 Fecal and fluid tagging.  Although complete bowel cleansing using purgation is apparently helpful for a complete evaluation of the colon, it is not mandatory for the performance of CT colonography. CT colonography can be successfully performed with reduced bowel preparation by labeling residual feces and colonic fluid using radiopaque oral contrast materials and thus making them distinguishable from the soft tissue density of colonic polyps or masses, a procedure called fecal and fluid tagging. Fecal and fluid tagging is the labeling of fecal residue (Fig. 50-180) and fluid (Fig. 50-181) in the colon by using radiopaque contrast agent. The contrast agent is orally administered at each meal, typically the day before the CT colonography. It is then mixed with ingested food material leaving the stomach. As the mixture passes through the bowel, nutrients are digested and absorbed, and the oral contrast becomes more thoroughly mixed with undigested residue. Contrast material also mixes with residual fluid in the colon. Some of the tagged feces and fluid are evacuated after bowel preparation, and some remain in the colon. On CT, fecal matter that is mixed with the orally administered contrast material appears hyperdense or white, making it easily distinguishable from the homogeneous soft tissue density of colonic polyps and masses (see Figs. 50-180 and 50-181). There are two main advantages of performing fecal and fluid tagging. First, diagnostic performance may be improved owing to easier and more accurate differentiation of polyps from fecal residue.483,482 Second, rigorous dietary restriction the day prior to CT colonography to minimize colonic residue is not mandatory, so purgative bowel cleansing can be reduced.480 Even a fairly large amount of feces or fluid in the colon, if they are completely tagged and thus easily distinguished from true colonic lesions, would not compromise the interpretation of CT colonography. Even laxative-free CT colonography with the use of fecal tagging has been attempted105,351,481 and has shown a very high diagnostic accuracy, demonstrating 100% perpatient and per-lesion sensitivity for colonic polyps 10 mm or larger.351 Laxative-free CT colonography, although still investigational, has a great potential to substantially decrease patient discomfort and increase patient acceptance of colonic examination by eliminating the inconvenience of colonic purgation. Various oral contrast agents including various densities of barium sulfate, sodium amidotrizoate and meglumine amidotrizoate (Gastrografin), and their combinations have been used for tagging.351,477,482,679 At present, there is no consensus with regard to the most effective method (which agent, dose, administration protocol) of tagging.

CHAPTER 50  Gastrointestinal Tract

A

1701

B

C FIG 50-179  A 6-mm sessile tubular adenoma of the sigmoid colon seen on 3D endoluminal (A) and 2D (width, 1500 HU; level, −200 HU) transverse (B) views of CT colonography and colonoscopy (C).

Gastrografin is an iodinated agent that has a mild cathartic effect, which may provide some degree of fecal cleansing.677 It is generally safe; however, rare anaphylactoid reaction after oral administration of Gastrografin has been reported in a blunt trauma patient without bowel disease782 and in a patient with pseudomembranous colitis.584 A small amount (3%) of Gastrografin is absorbed from the GI tract after oral administration and eliminated mainly via the kidneys.613 Therefore it should not be used in those patients with hypersensitivity to iodine, manifest hyperthyroidism, or decreased renal function.613 Colonic distention.  The importance of proper colonic distention in the performance of CT colonography cannot be overstated. A suboptimally distended colon can obscure as well as mimic a lesion. Even a fairly large lesion can be missed if the colon is collapsed. Automated carbon dioxide (CO2) insufflation is increasingly used for colonic distention during CT colonography. Automated CO2 insufflation is

reported to improve colonic distention794 and to be safer than manual room air insufflations.678 Although CO2 has not been clearly shown to decrease patient discomfort during the examination, postprocedural discomfort has been found to be less with CO2 than with room air because of the rapid absorption of CO2 through the colonic mucosa.794 An antispasmodic agent such as hyoscine butylbromide (Buscopan) or glucagon may be administered before colonic insufflation to relieve colonic spasm and patient discomfort.256,831 IV administration of 20  mg of Buscopan immediately before gas insufflation has been shown to be associated with significantly improved colonic distention,831 whereas glucagon has not shown a definitive benefit in colonic distention.598 Buscopan should be used with some caution because glaucoma, cardiac ischemia, and urinary retention may all be precipitated, and self-limited dry mouth and blurred vision may occur.

1702

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-180  Fecal tagging using orally administered barium. A, A 3D endoluminal view of CT colonography shows an 8-mm polypoid structure (arrowhead) on a haustral fold of the sigmoid colon. B, A 2D transverse view (width, 1500 HU; level, −400 HU) of CT colonography clearly shows very high attenuation of the polypoid structure (white arrowhead), confirming a piece of tagged stool. Another piece of tagged stool is noted in the sigmoid colon (black arrowhead). (From Park SH, et al: Fundamental elements for successful performance of CT colonography (virtual colonoscopy). Korean J Radiol 8:264–275, 2007.)

A

B FIG 50-181  Fluid tagging using orally administered Gastrografin. A, An 8-mm pedunculated tubular adenoma (arrow) of the cecum is clearly visualized on 2D transverse view (width, 2000 HU; level, 0 HU) of CT colonography owing to the contrast between the lesion and the high attenuation of tagged fluid. B, Colonoscopy shows the same polyp (arrowheads).

Scanning procedure.  Both supine and prone scans, completely covering the colon and the rectum, are routinely obtained because the combination of supine and prone positions has shown a superior colonic distention and a higher sensitivity of polyp detection compared with those of either position alone.122,902 According to the American College of Radiology practice guidelines for performance of CT colonography in adults, a slice collimation of 3 mm or less, with a

reconstruction interval of 1.5 mm or less, is required.256 With modern scanners (i.e., 16- or 64-detector MDCT) that are already in widespread use, scanning with a thinner collimation is easily achieved. One should use the thinnest slice possible to cover the entire abdomen in a single breath hold. Because of the intrinsic high contrast between the colonic mucosa and luminal air, detailed images of the colonic luminal surface can be obtained even with use of a very low radiation dose. In

CHAPTER 50  Gastrointestinal Tract general, 120 kVp is used. It has been shown that radiation dose could be decreased to as low as 10 mAs without sacrificing diagnostic accuracy, with a total effective dose of 1.8 mSv in men and 2.4 mSv in women,350 which is less than half the radiation dose of barium enema.396 In practice, scanning with 50 mAs allows evaluation of colonic and extracolonic findings on CT colonography. IV contrast enhancement is generally not required for screening purposes. However, it is necessary in patients with known or suspicious colorectal cancer, those who are undergoing follow-up after curative surgery for colorectal cancer, and those with symptoms that suggest an increased prevalence of extracolonic abnormalities.231,256,457 When IV contrast enhancement is used for extracolonic evaluation, the standard radiation dose for body CT should be used.256 Interpretation of scanning results.  Areas that might contain lesions can be searched by using two-dimensional (2D) review (i.e., review by scrolling through serial transverse images in a stack mode using wide window setting, such as window level/width of 2000 HU/  0 HU, 1500 HU/−400 HU, or 1500 HU/−200 HU) or 3D review (i.e., review with 3D volume-rendered images of colonic lumen, typically using the colonoscopy-like endoluminal fly-through) (see Fig. 50-179). There is no one correct method for interpretation of CT colonography, and many cases will require a combination of both 3D and 2D review methods, but 3D review has some advantages. Polyp conspicuity and the opportunity for detection are greatly enhanced on 3D endoluminal view compared with 2D view.675-677 A 2D review consisting of rapid tracing of the colonic outline on each cross-sectional image to find small contour abnormalities generally requires a higher level of reader concentration and experience and is a tedious and fatiguing search pattern.655,676 A 3D review, however, also has some disadvantages. A 3D review using bidirectional (i.e., forward and backward) endoluminal fly-through is less time-efficient than a 2D review.146,528,564,832 A 3D review may not work well with the fecal tagging and reduced purgation protocol that leaves a large amount of fecal residue in the colon.655,677 Typical potential lesions have rounded luminal protrusions (see Fig. 50-179) or protruding masses. However, in contrast to what was known previously, recent studies have shown that approximately 40% of adenomatous lesions detected at colonoscopy were flat.344,720,761 Flat lesions appear as nodular mucosal surfaces, plaque-shaped mucosal elevations, and thickened haustral folds at CT colonography.654 Because flat lesions are more difficult to detect at CT colonography compared to sessile or pedunculated lesions, a careful interpretation not to miss flat lesions is particularly required.651,650,654 If a polyp candidate is found, its internal attenuation, morphology, and movability are carefully examined to distinguish feces from a true lesion. Internal air-density (Fig. 50-182), a higher density than soft tissue attenuation, and movability (i.e., positional change between supine and prone scans) are specific signs of residual feces. Some fecal matter may not show these findings and therefore may be impossible to distinguish from a true polyp without fecal tagging. Once a lesion is determined, the longest linear size is measured. Because polyp size is clinically the single most important feature of a colorectal polyp, serving as a rough surrogate of the risk of carcinoma, the size should be measured accurately and reliably. CT colonography is known to be more reliable and accurate than colonoscopy for polyp measurement.649 Polyps measuring 10 mm or larger should be reported with a recommendation for therapeutic colonoscopy.256,923 Reporting of polyps measuring 5 mm or less is not recommended, because they often represent false-positive results, are frequently nonneoplastic, or are associated with extremely low risk of malignancy.256,923 Reporting of polyps measuring between 6 and 9 mm varies, depending on the specific lesion size, certainty of the findings, patient’s age, and existing

1703

Spin: 0 Tilt: -90

FIG 50-182  A piece of polypoid residual feces with internal air-density (arrow).

BOX 50-2  Classification of Colonic

Findings and Suggested Follow-up

C0: Inadequate study/awaiting prior comparisons • Inadequate preparation: cannot exclude lesions ≥10 mm owing to presence of fluid/feces • Inadequate insufflation: one or more colonic segments collapsed on both views • Awaiting prior colon studies for comparison C1: Normal colon or benign lesion; continue routine screeninga • No visible abnormalities of the colon • No polyp ≥6 mm • Lipoma or inverted diverticulum • Nonneoplastic findings (e.g., colonic diverticula) C2: Intermediate polyp or indeterminate finding; surveillance or colonoscopy recommendedb • Intermediate polyp 6-9 mm, <3 in number • Indeterminate findings, cannot exclude polyp ≥6 mm in technically adequate exam C3: Polyp, possibly advanced adenoma; follow-up colonoscopy recommended • Polyp ≥10 mm • ≥3 polyps, each 6-9 mm C4: Colonic mass, likely malignant; surgical consultation recommendedc • Lesion compromises bowel lumen, demonstrates extracolonic invasion From Zalis ME, et al: CT colonography reporting and data system: A consensus proposal. Radiology 236:3–9, 2005. a Every 5-10 years. b Evidence suggests that surveillance can be delayed at least 3 years, subject to individual patient circumstance. c Communicate to referring physician as per accepted guidelines for communication, such as ACR Practice Guideline for Communication: Diagnostic Radiology. Subject to local practice, endoscopic biopsy may be indicated.

comorbid conditions.256 One guideline using the CT colonography reporting and data system (C-RADS) classification scheme (Box 50-2) suggests an immediate colonoscopy for patients with three or more 6- to 9-mm polyps and a follow-up colonoscopy in 3 years for patients with fewer than three 6- to 9-mm polyps.923

1704

PART II  CT and MR Imaging of the Whole Body

TABLE 50-4  International Staging System

for Colorectal Cancer Stage

Description

Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of primary tumor Tis Carcinoma in situ: intraepithelial or invasion of lamina propria T1 Tumor invades submucosa T2 Tumor invades muscularis propria T3 Tumor invades through the muscularis propria into pericolorectal tissues T4a Tumor penetrates to the surface of the visceral peritoneum T4b Tumor directly invades or is adherent to other organs or structures Regional Lymph Nodes (N) NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis in 1-3 regional lymph nodes N1a Metastasis in one regional lymph node N1b Metastasis in 2-3 regional lymph nodes N1c Tumor deposit(s) in the subserosa, mesentery, or nonperitonealized pericolic or perirectal tissues without regional nodal metastasis N2 Metastasis in four or more regional lymph nodes N2a Metastasis in 4-6 regional lymph nodes N2b Metastasis in seven or more regional lymph nodes Distant Metastasis (M) M0 No distant metastasis M1 Distant metastasis M1a Metastasis confined to one organ or site (e.g., liver, lung, ovary, nonregional node) M1b Metastases in more than one organ/site or the peritoneum From Edge S, et al (eds), American Joint Committee on Cancer: AJCC cancer staging manual, ed 7. New York, 2011, Springer-Verlag, p 155.

Diagnosis and Staging of Colorectal Carcinoma.  The ultimate goal of imaging in patients with colorectal carcinoma is to detect the lesion and determine its extent (i.e., staging) to help guide appropriate treatment. The most commonly used staging system for colorectal cancer is the AJCC/UICC TNM system (Table 50-4).279 In general, CT is most widely used for staging of colorectal cancer. In terms of staging rectal cancer, attention has recently been directed, particularly in Europe, to preoperative prediction of the exact depth of extramural tumor invasion and tumor involvement of the circumferential resection margin (CRM) or mesorectal fascia, which equals the circumferential boundary of the surgical excision plane in total mesorectal excision, using HRMRI (Fig. 50-183) in addition to traditional T staging. Total mesorectal excision, which involves resection of the rectum and the surrounding mesorectal fat and lymph nodes by sharp dissection along the mesorectal fascia, is currently considered a standard surgery with curative intent for rectal cancer. This surgical technique is known to be associated with a low recurrence rate (<10%) when used as a single modality therapy.388 In rectal cancer patients with extramural tumor invasion, preoperative radiation and/or chemotherapy is recommended to further decrease the risk of recurrence after the surgery. In many European

FIG 50-183  Rectal cancer with involvement of mesorectal fascia anterolaterally (arrowheads) shown on T2-weighted MRI.

centers, patients who have a small distance between tumor and CRM may get neoadjuvant chemotherapy in addition to a short course of preoperative radiation for a better local control,264 based on the fact that the depth of extramural tumor invasion and the presence or absence of tumor involvement of CRM are major factors that determine the risk of recurrence and prognosis after surgery.577,607,698,699,882 In such practice, it is incumbent for radiology to be able to accurately measure the extramural depth of tumor invasion preoperatively and thereby to identify those tumors that will have a small distance from CRM or a tumor-involved CRM with a high risk of postsurgical recurrence. On the other hand, exact depth of extramural tumor extension is considered as having less practical importance in the United States and elsewhere because all patients who are T3 to T4 or node-positive eventually get both preoperative chemotherapy and radiation.264 CT and MRI in diagnosis of primary tumor.  In general, appropriate colonic distention is crucial for clear visualization of colon cancer on CT or MRI. Even a fairly large advanced lesion can be missed if the colon is collapsed (Fig. 50-184). In staging rectal cancer, however, the role of colonic distention is controversial. Colonic distention will improve lesion conspicuity but may alter the distance between the outer tumor margin and mesorectal fascia, confounding the assessment of extramural depth of tumor invasion. On CT or MRI a colorectal cancer appears as a discrete mass (Fig. 50-185) or focal or short-segmental wall thickening. Asymmetric wall thickening of the colon, with or without an irregular surface, suggests a neoplastic process. Mucinous adenocarcinoma.  Histologic subtypes of colorectal cancer are generally indistinguishable on CT or MRI. Mucinous adenocarcinoma, however, may show a characteristic large area of low attenuation on CT or high signal intensity on T2-weighted MRI, with a peripheral or heterogeneous pattern of contrast enhancement as a result of abundant accumulation of extracellular mucin346,413,437 (Figs. 50-186 and 50-187). Intratumoral punctate or irregular calcifications (Fig. 50-188) are more frequent than nonmucinous types.437 According to World Health Organization criteria, the mucin composition should be more than 50% of the tumor area for the diagnosis of mucinous adenocarcinoma.914 Compared with nonmucinous tumors, mucinous

CHAPTER 50  Gastrointestinal Tract

1705

M

FIG 50-186  Mucinous adenocarcinoma of the rectum. CT scan shows

A

a large heterogeneous mass (M) with areas of cystic components. Note marked luminal narrowing of the rectum (arrow).

B FIG 50-184  Obscuration of an advanced cancer due to colonic collapse. A fairly large ulcerofungating cancer mass in rectosigmoid junction (A, colonoscopic image) is not apparent on CT when the colon is collapsed (arrowheads in B).

FIG 50-187  Mucinous adenocarcinoma of the rectum. T2-weighted MRI shows high signal intensity (arrowheads) of the cancer lesion in right anterolateral side of the rectal wall.

FIG 50-185  Polypoid rectal carcinoma. CT scan shows a large polypoid mass (arrows) with a lobulated contour protruding into the rectal lumen.

FIG 50-188  Development of mucinous colonic carcinoma in a patient with juvenile polyposis. A large mass (arrows) in the ascending colon contains multiple calcifications.

1706

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-189  Primary signet ring cell carcinoma of the rectum. A, CT scan shows concentric bowel wall thickening (arrows) with the target sign in the rectum. B, Double-contrast barium enema shows a long segment of concentric luminal narrowing (arrows) along the rectum with minimal irregularity of the mucosal surface.

tumors tend to be at a higher pathologic stage at diagnosis, have a greater tendency for metastasis and local recurrence, and carry an unfavorable prognosis.824 Primary signet ring cell carcinoma.  Primary signet ring cell carcinoma of the colon is a rare condition, constituting between 0.1% and 2.4% of all colorectal carcinomas.202 Signet ring cell carcinoma may show unique appearances at imaging. It usually produces minimal mucosal alterations but permeates the colonic wall extensively, thereby transforming the affected segment into a rigid and contracted structure, resulting in a so-called linitis plastica appearance. On CT scans, the most common feature of this tumor is a long segment of concentric bowel thickening405 (Fig. 50-189). Another important CT feature is target appearance in the thickened colorectal wall. The inner layer of a thickened bowel wall with a target sign is markedly enhanced owing to abundant fibrous tissues secondary to desmoplastic response. These features, however, are noted in only approximately 50% of cases with signet ring cell carcinoma.405 Because of these morphologic features, primary colorectal signet ring cell carcinoma may simulate other conditions; among those, metastatic linitis plastica from stomach, breast, gallbladder, bladder, or prostate211 may have exactly the same CT appearance.289 Therefore without knowledge of the clinical history of primary malignancy in other sites of the body, CT differentiation of primary signet ring cell carcinoma from metastases may not be possible. Other conditions that should be considered in the differential diagnosis are lymphoma and inflammatory and vascular disorders (e.g., Crohn’s disease, ulcerative colitis, and pseudomembranous  and ischemic colitis) and gynecologic diseases (e.g., endometriosis, actinomycosis). When the rectum is involved, extension to the level of anal verge is common. Compared with ordinary colorectal adenocarcinoma, signet ring cell carcinoma commonly occurs in younger patients and tends to spread to the pelvic lymph nodes, ovaries, and peritoneal surfaces, with a low incidence of hepatic metastasis.12

Moreover, its clinical course is aggressive, with a high incidence of local recurrence.12 Ischemic colitis.  Ischemic colitis may develop in a colonic segment proximal to the colon cancer in 1% to 7% of cases of colorectal carcinoma. An understanding of this condition is important because the concomitant presence of an ischemic segment may give a false impression regarding tumor length or depth of tumor invasion. According to one series,438 CT could distinguish ischemic segment from tumoral segment in 75% of patients by the differences of both bowel wall thickness and contrast enhancement patterns (Fig. 50-190). In contrast to irregular thickening in the tumoral segment, the ischemic segment shows smooth and regular thickening; the target sign with low attenuation in the central zone was the most specific sign of the ischemic segment.438 Maximum thickness also differed, with a mean thickness of 2 cm in the tumoral segment and 1 cm in the ischemic segment.438 In certain instances the ischemic segment may be present in the proximal colon, with intervening normal-appearing colonic mucosa (see Fig. 50-190).

Terminal ileum involvement by right-sided colon cancer.  Terminal ileum involvement by right-sided colon cancer (Fig. 50-191) was found in 10% of patients in one series.407 The presence of both colonic and distal ileal thickening might be misinterpreted as inflammatory bowel disease or lymphoma when CT readers are not aware of a patient’s history of colon cancer. The distal ileal involvement is caused by either direct tumor invasion or lymphatic spread and can be mimicked by nonneoplastic processes such as edema and congestion that are secondary to colon cancer. Perforation.  Perforation occurs in 2.5% to 8% of patients with colorectal cancer and is regarded as a poor prognostic sign. Because perforation is usually either localized or channeled into a fistulous tract, a significant percentage of patients with perforated colorectal cancer have no signs or symptoms of acute toxicity.340 Except for the

CHAPTER 50  Gastrointestinal Tract

M

FIG 50-190  Rectal cancer in the proximal colon in a patient with colitis. CT scan shows a mass (M) in the sigmoid colon along with proximal colonic dilatation. Bowel wall thickening with a target appearance (arrows), also shown in the proximal sigmoid colon, was confirmed as ischemic colitis on pathologic examination. Note the skipped area between tumoral and ischemic segments. (From Rha SE, et al: CT and MR imaging findings of bowel ischemia according to the various primary causes. Radiographics 20:29–42, 2000.)

1707

detection of enteroenteric fistulas, CT proved superior to barium studies in establishing the correct diagnosis. On CT scans, most of these patients showed a neoplastic mass with fluid-attenuation abscess or inflammatory changes in the pericolic or perirectal tissues340 (Fig. 50-192). Assessment of local tumor spread.  Colorectal carcinomas arise in the mucosa and grow to progressively invade the deeper layers of the bowel wall. When the tumor is confined to the bowel wall, the outer margins of the large bowel appear smooth on CT and MRI,  and the perirectal or pericolic fat tissue has uniform attenuation  on CT and uniform signal intensity on MRI, without soft tissue stranding. Extension of tumor beyond the bowel wall manifests as irregular outer borders of the bowel wall with soft tissue stranding  in the perirectal or pericolic fat tissue. Similar strandings can also be caused by desmoplastic reaction of the peritumoral tissue, inflam­ mation, or congestive changes, resulting in overstaging of a T2 lesion as T3 lesion.264,456 On the other hand, microscopic cancer invasion into pericolic fatty tissue may not be visualized on cross-sectional imaging, resulting in understaging of a T3 lesion as T2 lesion.264 A definitive diagnosis of tumor invasion beyond the bowel wall can be made with cross-sectional imaging if a tumor mass extends directly

I M

C

I

A

B

M

FIG 50-191  Adenocarcinoma of the ascending colon with extension

C

to the distal ileum. A, CT scan shows a heterogeneous mass (M) in the ascending colon with irregular thickening of the distal ileum (I). B, Coronal HASTE MRI shows irregular thickening of the bowel wall in the ascending colon (C) and distal ileum (I). C, Coronal Gd-enhanced MRI shows irregular bowel wall thickening of the distal ileum (arrows) with heterogeneous contrast enhancement. A mass (M) is visible in the ascending colon.

1708

PART II  CT and MR Imaging of the Whole Body

U M

FIG 50-194  Rectal cancer with uterine invasion. CT scan shows a large FIG 50-192  Perforated colon cancer. The colonic wall in the ascending colon is thickened and enhanced (arrow), representing colon cancer. Colonic wall in the tumor portion is focally disrupted, and a hypoattenuated abscess with internal air bubbles is formed adjacent to the perforation site (arrowheads).

heterogeneous rectal mass (M) with compression and direct invasion into the posterior wall of the uterus (U).

n

*

n

B

FIG 50-195  Ascending colon cancer (not shown) with duodenal inva-

FIG 50-193  Sigmoid colon cancer with bladder invasion. CT scan shows irregular bowel wall thickening of the sigmoid colon (arrows). The superior part of the bladder (B) is irregular because of direct tumor invasion.

into an adjacent muscle (levator ani, obturator internus, coccygeus, piriformis, gluteus maximus) or organs, obliterating the fat planes and enlarging the individual muscle or enveloping the neighboring structures (Figs. 50-193 to 50-195). Accuracy.  Accuracy of CT and MRI in local tumor staging has been studied mostly with rectal cancer. According to published reports, the accuracy of CT and MRI in the prediction of T stages is moderate. One meta-analysis showed T-stage sensitivity and specificity of 78% and 63%, respectively, for CT and 86% and 77%, respectively, for MRI, in comparison with 93% and 78%, respectively, for EUS.454 MRI with endorectal coil showed improved accuracy with T staging sensitivity and specificity of 89% and 79%, respectively.454 Owing to the rapid advances in imaging technology in recent years (particularly MDCT), however, many of the published results are already outdated and may not necessarily represent the diagnostic performance of modern scanners that are currently in widespread use.

Rectal HRMRI and assessment of extramural depth of tumor invasion.  In contrast to some limitations of CT and MRI in predict-

ing T stage of colorectal cancer, HRMRI of the rectum has recently

sion. CT scan shows heterogeneous duodenal wall thickening (asterisk) along with lymphadenopathy (n) in the retroperitoneum and peripancreatic regions.

been shown to be highly accurate and reproducible in assessing the absolute extramural depth of tumor invasion in rectal cancer. A phased-array surface coil is preferred for HRMRI of the rectum, because endorectal-coil MRI has a limited field of view despite its superior spatial resolution. Complete assessment of the perirectal structures is difficult with endorectal coils because portions of the mesorectal fascia, mesorectal fat, and lymph nodes lie outside the field of view.349 Similarly, EUS, although more accurate than CT or MRI in the evaluation of rectal wall and superficial rectal cancer, is limited in evaluation of the perirectum and mesorectal fascia, owing to limited tissue penetration.53,67 T2-weighted MRI sequences are the most suitable for depicting the rectal wall anatomy and perirectal cancer spread, and high-resolution T2-weighted imaging with a non–breath-hold turbo-spin echo sequence has been used in most studies54,90,349 (Fig. 50-196). Contrast-enhanced T1-weighted sequences, on the other hand, have not been shown to be effective.865 Identification and staging of rectal cancers by MRI is largely based on differences in T2 signal intensity between the tumor and the rectal wall and perirectal fat tissue. The tumor itself has an intermediate signal intensity between the high signal intensity of the submucosa or perirectal fat tissue and the low signal intensity of the muscular layer (see Fig. 50-196).91,349 The absolute depth of extramural invasion of

CHAPTER 50  Gastrointestinal Tract

FIG 50-196  Rectal cancer on T2-weighted MRI. An ulcerofungating cancer mass presents as a C-shaped lesion with intermediate signal intensity (arrowheads) on T2-weighted transverse MRI.

rectal cancer on MRI agrees well with the pathologic measurement,575 and the presence or absence of tumor involvement of CRM can be predicted accurately. In one study, a histologic tumor-free CRM with at least a 1- and 2-mm distance from the outermost tumor margin can be predicted with a high degree of certainty when the measured distance between the outermost part of the tumor and the adjacent mesorectal fascia on MRI is at least 5 mm (for a 1-mm distance) and 6 mm (for a 2-mm distance).54 In another study, when MRI criteria of 1 mm or less distance between the outermost tumor margin and mesorectal fascia was used to predict tumor involvement of the circumferential resection margin, the accuracy, specificity, and sensitivity of MRI were 88% (359/408), 92% (327/354), and 59% (32/54), respectively.574 Assessment of lymph node spread.  The patterns of lymph node spread of colorectal cancer are predictable to a certain degree on crosssectional images by the anatomy of lymphatic drainage of the colon. Colon cancer spreads first to epicolic and paracolic nodes, then follows the courses of corresponding colonic vessel branches. Cecal cancer spreads following the course of ileocolic vessel to the mesenteric root, and ascending colon cancer spreads following the course of right colic vessels to the root of the SMA.562 Tumors of the hepatic flexure and proximal transverse colon initially spread to the nodes of the gastrocolic trunk, whereas distal transverse colon and splenic flexure tumors spread to nodes in the mesocolon along middle colic vessels.562 Leftsided colonic tumors tend to spread first to nodes in the mesocolic artery and then to left colic or inferior mesenteric artery nodal groups.276 Rectal cancer has two main routes of lymphatic spread. For the upper portion of the rectum, the route is upward along the superior rectal vessels to the inferior mesenteric vessels.53 The lower portion of the rectum shows an additional lateral lymphatic route along the middle rectal vessels to the internal iliac vessels.53 Careful preoperative evaluation of internal iliac areas (Fig. 50-197) is important in rectal cancer patients, particularly in those with cancer in the lower portion of the rectum, because lymph nodes in those areas lie outside the resection margin of total mesorectal excision. Downward spread along the inferior rectal vessels to the groin is unusual except in very advanced cases and when the anal canal is involved.53

1709

FIG 50-197  An 11-mm metastatic lymph node in right internal iliac area. Heterogeneous, mottled, high signal intensity in the lymph node (arrow) is a characteristic of malignant lymphadenopathy noted on T2-weighted MRI.

Accuracy of CT and MRI in nodal staging has been studied mostly with rectal cancer. Despite the high spatial resolution of CT and MRI that allows identification of lymph nodes as small as 2 to 3 mm, reliable detection of nodal metastases is presently not possible. Radiologic assessment of nodal involvement generally relies on morphologic criteria such as the size and shape of the node. One meta-analysis showed 52% sensitivity and 78% specificity of CT for the detection of nodal metastasis and 65% sensitivity and 80% specificity for MRI.454 In another meta-analysis, the corresponding sensitivity and specificity were 55% (95% confidence interval: lower limit, 43; upper limit, 67) and 74% (95% confidence interval: 67, 80) for CT and 66% (95% confidence interval: 54, 76) and 76% (95% confidence interval: 59, 87) for MRI.67 The limitation with size criteria is that micrometastases within small lymph nodes are ignored, and with enlarged nodes it is difficult to distinguish between reactive and metastatic nodes. Rectal cancer is particularly known to have a high frequency of micrometastases in normal-sized nodes,16,72,186,594 with 45.4% to 78% of the involved nodes being 5 mm or smaller in size.16,317,446 Therefore a lower size criterion to predict malignant perirectal lymphadenopathy (e.g., ≥5 mm) compared to the criteria for the other intraabdominal nodal stations (e.g., ≥8 mm or ≥10 mm) is recommended in the interpretation of perirectal lymph nodes. Morphologic analysis of lymph nodes may help diagnose malignant lymph nodes to a certain degree. Spiculated or indistinct border of the nodes and heterogeneous pattern of signal intensity on MRI suggests malignant lymph nodes (see Fig. 50-197).92,400 Assessment of distant metastases.  The liver is the most common site of hematogenous spread in colorectal cancer, which typically appears as focal areas of lower attenuation on CT or higher signal intensity on T2-weighted MRI and lower signal intensity on T1-weighted MRI compared to hepatic parenchyma, with or without rim enhancement after contrast enhancement. Portal phase is an optimal scan timing for the evaluation of hepatic metastases, because lesion conspicuity is the highest and so is lesion detection. Sensitivity

1710

PART II  CT and MR Imaging of the Whole Body

of CT and MRI in the detection of hepatic metastases is moderate. One meta-analysis showed 72% (95% confidence interval: 63, 80) and 76% (95% confidence interval: 57, 91) sensitivity, respectively, for CT and MRI in the detection of hepatic metastases from various GI cancers.427 In another meta-analysis, CT and MRI showed 63.8% and 64.4% to 66.1% sensitivity, respectively, in the detection of colorectal liver metastases.68 Punctate or amorphous calcification can be seen in the hepatic metastasis from mucinous colonic adenocarcinoma. Sometimes a metastasis can appear identical to a simple cyst, an eosinophilic abscess, a granuloma, or a hemangioma. CT is highly sensitive in lung nodule detection and is the first choice to evaluate lung metastases of colorectal cancer but is not specific.264 CT is limited in the evaluation of peritoneal tumor seeding from colorectal cancer, showing per-patient sensitivity of 60% to 76%.158 Per-lesion sensitivity is even further limited, demonstrating only 9.1% to 24.3% for tumor implants smaller than 1 cm.158

FDG PET in Preoperative Assessment of Colorectal Carcinoma.  The use of FDG PET for routine preoperative staging of colorectal cancer is controversial,246,387 and its role in staging of colorectal cancer has yet to be established. FDG PET is insensitive in the diagnosis of nodal metastasis, showing disappointingly low sensitivity of 29% in one study.1 Therefore FDG PET is generally not recommended for the purpose of nodal staging. FDG PET is also not routinely recommended for the evaluation of metastatic disease in colorectal cancer patients. Several meta-analyses have shown that FDG PET was superior to CT or MRI in the detection of hepatic metastases from various cancers of the GI tract, with 90% sensitivity and 85% specificity,427 or similar to CT or MRI in the detection of hepatic metastases from colorectal cancer, with 75.9% sensitivity.68 However, FDG PET is not sensitive for small hepatic metastases, demonstrating 25% sensitivity for lesions smaller than 1 cm, in contrast to 85% sensitivity for lesions larger than 1 cm.232 On the other hand, FDG PET is helpful in the evaluation of extrahepatic metastases in a selected group of patients with potentially resectable hepatic metastases. FDG PET may detect unsuspected extrahepatic metastases in such patients and thereby dictate avoidance of inappropriate hepatic resection when such an operation is unlikely to result in long-term benefit owing to the presence of extrahepatic disease.232,756

Postsurgical Follow-up of Colorectal Carcinoma CT and MRI in postsurgical follow-up of colorectal carcinoma.  More than 90% of recurrences after surgery with curative intent occur by 5 years after the surgery, the vast majority within 3 years.172,672 More than half of recurrences after curative resection of colorectal cancer present as distant metastasis to the liver and lungs.541,722 Anastomotic or local tumor recurrences also present more frequently as extraluminal lesions than as intraluminal lesions.722 According to one meta-analysis, the rate of intraluminal anastomotic recurrence was only 3.2% (33 of 1028) of patients who underwent curative surgery for colorectal cancer, in contrast to 14.4% (193 of 1342) for all local recurrences.722 Therefore cross-sectional imaging modalities such as CT or MRI that typically evaluate extracolic structures rather than intraluminal colonic lesions play an important role in postsurgical surveillance of those who underwent curative surgery for colorectal cancer. The importance of CT in postsurgical surveillance was demonstrated by the study results that CT and frequent measurements of serum carcinoembryonic antigen (CEA) during postsurgical follow-up led to reduced cancer-related mortality by 9% to 13%.722 The importance of abdominal CT is also reflected in the recently revised surveillance guideline for postsurgical colorectal cancer patients proposed by

the American Society of Clinical Oncology.172 The guideline recommends annual CT of the abdomen for 3 years after primary therapy for patients who are at high risk of recurrence.172 Findings indicative of local tumor recurrence include a soft tissue mass increasing in size on follow-up scans, enlarged local lymph nodes, or invasion of surrounding structures. Another important sign that indicates recurrence is the presence of ischiorectal fossae invasion (Fig. 50-198).728 Both CT and MRI are useful in the diagnosis of pelvic recurrences.77,161,359,449,545,665 Posttherapy scarring and fibrosis can be confused with local recurrence. Single-time CT is generally limited in distinguishing recurrent tumor from posttherapy scarring and fibrosis (Fig. 50-199), and repeat examination after an interval of several weeks or months is required to assess size change. Some studies have shown that MRI was useful to a certain extent in the differential between recurrence and posttherapy fibrosis by distinguishing fibrosis with low signal intensity from the high signal intensity of recurrent tumor on T2-weighted images.359,449,545 However, other studies have shown that high signal intensity on T2-weighted images was not specific for recurrences but could be seen in nonneoplastic inflammatory changes or edema or in immature fibrosis (<1 year postoperatively).161,545 Low signal intensity in the mass could also be found in recurrent tumor with desmoplastic reaction.161,545 FDG PET in postsurgical follow-up of colorectal carcinoma.  In contrast to the rather nonspecific nature of CT and MRI findings of recurrences, FDG PET is highly effective in the diagnosis of suspected

M

FIG 50-198  Perineal recurrence of rectal cancer following abdominoperineal resection. A heterogeneous soft tissue mass (M) extends to the left ischiorectal fossae.

FIG 50-199  Postsurgical fibrosis following abdominoperineal resection for rectal cancer. A round well-enhancing soft tissue mass (arrows) is seen in the presacral space.

CHAPTER 50  Gastrointestinal Tract

A

1711

B FIG 50-200  Pericolic recurrence after low anterior resection of rectal cancer. A, CT obtained 25 months after surgery shows a moderately enhancing infiltrative soft tissue mass (arrowheads) located right posterolaterally to the anastomosis. B, 18FDG PET CT image at corresponding level shows hypermetabolism (arrowheads) of the lesion, confirming tumor recurrence. (From Choi YJ, et al: CT colonography for follow-up after curative surgery for colorectal cancer. AJR Am J Roentgenol 189:283–289, 2007. Reprinted with permission from the American Journal of Roentgenology.)

recurrent colorectal cancer23,52,171,223,341,355,813 and is particularly useful in distinguishing pelvic recurrence from posttherapy fibrosis in that residual or recurrent tumor shows uptake of the tracer (Fig. 50-200), whereas posttherapy scar tissue does not with high accuracy.359,397,625 FDG PET is also useful in the evaluation of postsurgical patients with CEA elevation without radiologic evidence of tumor recurrence or metastases.223,226,925 With the substantial improvement of cross-sectional imaging devices in the recent past, with the widespread implementation of MDCT and higher-field-strength MRI, however, such cases of radiologically occult CEA elevations are now rare.264

Other Benign Tumors

Lipoma and Teratoma.  Lipoma of the colon is a benign neoplasm that arises from the submucosal layer of the intestinal wall. Lesions may be asymptomatic but may cause symptoms when they are large (>2 cm), especially in cases associated with intussusception or ulceration of the overlying mucosa. Seventy percent of the lesions are located in the right colon and usually in the cecum.116 They are well demonstrated on CT scans because the masses present characteristic fatty attenuation (Fig. 50-201). However, small lesions may be missed because of artifacts and partial-volume averaging. Furthermore, intussuscepted lipomas occasionally show focal or diffuse areas of soft tissue attenuation within the lesion and may be mistaken for other neoplasms (Fig. 50-202). The changes in CT attenuation in such cases are directly related to the degree of infarction and fat necrosis within the lipoma.98 Although uncommon, teratoma can occur in the colon. CT may show a mass made up of fatty and soft tissue components.783 The presence of calcification may distinguish teratoma from lipoma.

Hemangioma.  Diffuse cavernous hemangioma of the colon is an uncommon disease that affects mainly young adults. The rectosigmoid is the most common site in the GI tract. Chronic rectal bleeding is the main complaint of these patients. Gross examination reveals an extensive network of vascular lakes involving the entire intestinal wall, with infiltration into the surrounding connective tissue. On microscopic examination, the disease is characterized by large thin-walled vessels with smooth muscle fibers and connective tissue stroma in a submucosal location.138 CT may show a transmural thickening of the involved segment, extramural lesions, intramural and extrarectal phleboliths, and vascu-

FIG 50-201  Colonic lipoma. A well-circumscribed polypoid mass with homogeneous fat attenuation (arrow) is noted in the ascending colon.

lar engorgement within the mesentery (Fig. 50-203).667 MRI shows a markedly thickened colonic wall of very high signal intensity on T2-weighted images, which may result from slow flow in the vascular malformation and is considered to be a highly specific finding. Perirectal fat also shows very high signal intensity on T2-weighted images and heterogeneity related to serpiginous vascular structures.523

Cystic Lymphangioma.  Cystic lymphangioma is a rare benign submucosal tumor of the colon. CT may show a sharply marginated, multilocular, or purely cystic mass (Fig. 50-204).915 Colitis cystica profunda may have similar appearances on CT.

Other Malignant Tumors

Metastatic Tumor.  As seen in the stomach and small bowel, metastatic disease can occur in the colon as a result of peritoneal seeding, direct invasion, and hematogenous or lymphatic spread. CT helps not only in discriminating primary from secondary neoplasms to the colon but also in identifying the primary tumor in other sites. The rectum may be affected by direct extension of tumor from the prostate or cervix. The anterior wall of the rectum is the most common site of

1712

PART II  CT and MR Imaging of the Whole Body

FIG 50-205  Metastatic linitis plastica involving the ascending colon in a patient who had undergone gastrectomy for gastric cancer. CT scan shows marked luminal narrowing of the ascending colon with concentric bowel wall thickening (arrows) in addition to the target sign markedly enhancing the inner zone.

FIG 50-202  Colonic lipoma with focal fat necrosis. There is evidence of colocolic-type intussusception with a lipoma at the lead point. A focal area of high attenuation (arrows) in the lipoma results from focal fat necrosis.

R

FIG 50-203  Hemangioma of the rectum. The bowel wall of the rectum (R) is concentrically thickened, with poor contrast enhancement. Small phleboliths (arrows) are noted in the involved rectal wall as well as in the perirectal space.

FIG 50-204  A 3.5-cm cystic lymphangioma of the cecum that shows homogeneously low attenuation (arrowheads) has moved to the transverse colon owing to colocolic-type intussusception.

peritoneal seeding; other locations include the sigmoid and transverse colons and along the paracolic gutters. Secondary involvement by extension along the mesenteric reflection is seen in the transverse colon, where neoplasms of either gastric or pancreatic origin extend to the colon. Although not common, hematogenous metastases, especially from breast, melanoma, or lung cancers, produce polypoid masses in the colon and even cause intussusception.798 In such instances, it may be difficult to differentiate them from mucosal polyps or primary colonic tumors. At other times, metastatic tumors may show diffuse involvement of the colonic wall, with an annular or linitis plastica appearance on CT, mimicking primary colonic carcinoma or inflammatory bowel disease (Fig. 50-205). This linitis plastica pattern of colonic involvement may result from intramural spread of tumor cells through the submucosa or subserosa in association with peristaltic bowel movement.211 Because an important pathologic characteristic of linitis plastica is the exuberant desmoplastic reaction in the submucosa and subserosa, a target sign in diffusely thickened wall is commonly seen on CT scans.289

Gastrointestinal Stromal Tumors.  GISTs occur rarely in the colorectum, with the rectum being the most common site. These tumors appear as a submucosal mass growing toward the lumen (endocolic), away from it (exocolic), or in a combined (dumbbell) form. In certain instances a constricted form has also been reported.169 In general, clinical symptoms correlate with the types of tumor growth. Intramural types cause pain, bowel habit changes, and intestinal obstruction; pedunculated tumors commonly develop intussusception or bleeding owing to degeneration; and exocolic tumors may commonly produce peritoneal metastasis.814 Seventy percent of rectal stromal tumors are located in the lower rectum, 20% are in the midrectum, and 10% are in the upper rectum. The remaining rectal stromal tumors are located equally in the ascending colon, transverse colon, and descending colon.513 Because of the high incidence of late recurrence, some investigators suggest that most colorectal stromal tumors should be regarded as potentially malignant neoplasms.871 Generally, stromal tumors have varying degrees of internal necrosis or cystic changes on CT, predominantly in cases of malignant forms (Fig. 50-206). The presence of internal dystrophic calcification within a tumor suggests the possibility of stromal tumor in the colon, although this finding can be seen in both benign and malignant forms 

CHAPTER 50  Gastrointestinal Tract

1713

R

FIG 50-206  Malignant rectal GIST. CT scan shows a heterogeneous mass (arrows) of rectal (R) origin growing exophytically and displacing the vagina anteriorly. (From Lee SH, et al: Radiological features of leiomyomatous tumors of the colon and rectum. J Comput Assist Tomogr 24:407–412, 2000.)

FIG 50-207  Malignant rectal GIST. CT scan shows a calcified mass (arrows), anterior to the rectum (arrowhead), simulating a prostatic mass. (From Lee SH, et al: Radiological features of leiomyomatous tumors of the colon and rectum. J Comput Assist Tomogr 24:407–412, 2000.)

FIG 50-208  Colonic lymphoma. Severe thickening of the cecal wall with mild enhancement (arrows) is noted.

3% of all extranodal lymphoma.175 Non-Hodgkin’s lymphoma is the most predominant type, and the cecum and rectum are the most common sites of involvement. Colonic lymphoma may be (1) infiltrative, (2) polypoid, (3) cavitary, or (4) diffuse. On CT the infiltrative form is characterized by a long segment of concentric luminal narrowing by diffuse tumor infiltration into the submucosa, simulating primary colonic carcinoma (Fig. 50-208). However, in patients with lymphoma the occurrence of aneurysmal dilatation is common, and the degree of contrast enhancement of the mass is usually poor owing to a lack of desmoplastic reaction. When a solitary mass is found in the colon, CT may not always distinguish lymphoma from primary colonic adenocarcinoma. How­ ever, lymphoma may be a possibility when cecal tumor extends to the terminal ileum (Fig. 50-209), when diffuse mesenteric lymphadenopathy and splenomegaly are associated, and when the tumor is fairly well demarcated without showing evidence of invasion or obstruction of neighboring viscera.895 In the cavitary form of lymphoma, large tumors may excavate into the mesentery, producing a giant cavitated mass. In the diffuse form of lymphoma characterized by multiple nodules of relatively small size, CT may not detect the nodules but helps in staging the tumor.

Inflammation (Fig. 50-207). Tumor growth patterns may be helpful in differentiating benign from malignant; endophytic tumor growth favors the diagnosis of the benign rather than the malignant form.133 Stromal tumors arising from the rectum may be difficult to diagnose preoperatively. Many possibilities can be included in the differential diagnosis, depending on the patient’s sex or the direction of tumor growth. Tumors growing in an anterorectal direction may be easily confused with prostatic and gynecologic tumors, such as uterine, cervical, and ovarian tumors or a mass arising from the rectovaginal septum. Stromal tumors growing in the retrorectal direction should be differentiated from other unusual sarcomas or lymphomas. In contrast with rectal cancer, rectal stromal tumors are usually  not associated with obstruction or they cause a mild degree of obstruction owing to the position of the main tumoral component in the extraluminal space.30 Except for its poor sensitivity for detection of internal tumoral calcification, MRI may be comparable to CT in tumor evaluation.471

Lymphoma.  Primary lymphoma of the colon is relatively rare, constituting approximately 0.05% of all colonic neoplasms and fewer than

Crohn’s Disease.  The most common CT finding of Crohn’s colitis is bowel wall thickening (mean thickness, 11 mm).673 It is often discontinuous and may have skip areas. Mural stratification is often noted after contrast enhancement on cross-sectional images, particularly in an acute inflammatory phase, and a double-halo or target appearance of the bowel wall may uncommonly be seen.236 In some patients, however, mural stratification may be lost, showing a homogeneously attenuated bowel wall thickening (Fig. 50-210), particularly in a chronic or cicatrizing phase.673 Marked contrast enhancement in the early phase of dynamic study and the presence of hypervascularity in the pericolic space and mesentery may indicate acute exacerbation of the disease. Perianal lesions occur in up to one third of patients and include anal fissures and ulceration, perirectal abscess, and fistulas. CT offers an excellent alternative to barium studies in fully defining the extent of perirectal disease, especially in patients with an exquisitely tender perineum or rectum in which the performance of barium enema and sinography is painful.716 In performing the CT scan for patients thought to have perianal disease, the scanning plane must continue caudad to include the entire perineum because of common occurrence of diseases beneath the level of the symphysis pubis.917

1714

PART II  CT and MR Imaging of the Whole Body

M M

B

A

FIG 50-209  Colonic lymphoma. A, A polypoid mass (M) is seen in the ascending colon. B, The mass (M) extends into the terminal ileum.

A

B FIG 50-210  Active Crohn’s disease. A, CT scan shows colonic wall thickening (arrowheads) in the distal sigmoid colon. The thickened bowel wall is homogeneously enhanced, and mural stratification is not apparent. A colovesical fistula (arrow) in the sigmoid colon with inflammatory changes in adjacent tissue is noted (urinary bladder not shown on this image). B, Single-contrast fluoroscopic study using Gastrografin enema shows a luminal narrowing with a cobblestone appearance of the luminal surface (arrowheads) in the distal sigmoid colon. A colovesical fistula is demonstrated as a thin line (arrow) that connects the rectosigmoid junction and the urinary bladder.

Because of MPR capability, MRI is superior to CT in demonstrating the relationship between the pelvic and abdominal sinus tracts or fistulas and the levator ani muscle.442 In the differentiation with ulcerative colitis, Crohn’s colitis may show even distribution between right-sided, left-sided, or bilateral disease and high frequency of both small bowel involvement and abscess.673 Rectal sparing is a characteristic of Crohn’s disease that helps distinguish it from ulcerative colitis. The presence of ascites is not common in Crohn’s disease and instead suggests pseudomembranous,

infectious, or ischemic colitis. Tuberculous colitis may have patterns of bowel wall changes similar to those of Crohn’s disease. In tuberculous colitis, however, the occurrence of rectal involvement is rare; lymph nodes are larger than 1 cm and may show central necrosis. Moreover, the presence of CT findings of peritonitis, such as diffuse omental and mesenteric infiltration or nodules and ascites, may indicate tuberculous origin. Patients with Crohn’s disease are at increased risk for colorectal cancer. According to one study, colorectal cancer occurred an average

CHAPTER 50  Gastrointestinal Tract

1715

A IC

I

FIG 50-211  Tuberculous colitis. CT scan shows concentric bowel wall thickening (arrows) of the ascending colon with pericolonic stranding and minimal lymphadenopathy (arrowhead).

of 10 years earlier than sporadic cancer and is more often of the mucinous type.223

FIG 50-212  Tuberculous colitis. The cecum and the proximal ascending colon (A), particularly those areas adjacent to the ileocecal valve (IC), are shrunken and shortened by TB. Deformity of the colon causes the dilated terminal ileum (I) to resemble the cecum or ascending colon. Bowel wall thickening is noted both in the ascending colon (A) and terminal ileum (I).

Tuberculous Colitis.  Tuberculous colitis may have several forms, most commonly segmental colitis, inflammatory strictures, and hypertrophic lesions resembling polyps and tumors.546 A rare form includes a diffuse type simulating ulcerative colitis.32 The ileocecal area is the most common site of tuberculous involvement in the GI tract.3 Multiple sites of involvement with skip segments are common.853 The length of stricture is usually shorter (<3 cm) than that of Crohn’s disease.829 Tuberculous lymphadenitis adjacent to the colon can result in colonic traction diverticula, bowel fixation, and sinus tract formation.546 Although not common, anal TB may cause ulcers and fissures, fistulas, and abscesses. On CT scans, the involved colon may show circumferential bowel wall thickening up to 3 cm. This thickening is usually homogeneous in attenuation (Fig. 50-211), and mural stratification (a double-halo or target sign) is rarely seen. Thickening of the ileocecal valve, the presence of inflammatory tissue (localized hypertrophic type), or segmental lesion in any sites of the colon may give the appearance of colonic carcinoma on CT. There are also no specific CT findings that can distinguish TB from Crohn’s disease in terms of the bowel wall involvement patterns, except for the distribution of disease; it is rare to see rectal involvement in tuberculous colitis, and the predominant left-sided colonic involvement may favor Crohn’s disease. Active tuberculous lesions may be suggested on CT in the following circumstances: • When there are diffuse inflammatory infiltrates or (rarely) hypervascularity in the pericolonic space • When the enlarged lymph nodes in the pericolonic space or other sites are diffusely enlarged or show central low attenuation • When there is evidence of diffuse peritonitis, such as ascites, infiltration, or nodules in the omentum or mesentery, and peritoneal thickening The latter two findings also favor the diagnosis of tuberculous colitis rather than Crohn’s disease. Although not common, fibrofatty proliferation can be seen in the pericolonic space. Healing of ileocecal lesions may result in shortening of the ascending colon. In such cases, the thickened terminal ileum may be easily misinterpreted as ascending

colon on CT (Fig. 50-212). In comparison with the rate of small bowel obstruction in tuberculous enteritis, the incidence of colonic obstruction is low because of a larger luminal diameter of the colon, but it rarely occurs in hypertrophic type of tuberculous colitis. Colonic perforation due to tuberculous colitis is also rare (Fig. 50-213).

Ulcerative Colitis.  Ulcerative colitis is characterized by a chronic mucosal disease with an inflammatory reaction that remains limited to the mucosa and the superficial portion of the submucosa. It usually starts in the rectum and then spreads more proximally, sometimes to involve the entire colon (pancolitis). In about one third of cases, the ileum is involved by backwash ileitis. In active stages of the process, the mucosa shows a diffuse, uniformly granular, and erythematous hemorrhagic appearance, and in the advanced stage, the entire bowel becomes fibrotic, narrowed, and shortened. When the disease is limited to the mucosa, the colon appears normal on CT scans; the presence of active ulceration does not significantly alter the CT appearance of the colonic wall. In patients with chronic disease, CT may show bowel wall thickening caused by hypertrophy of the muscularis mucosa and deposition of submucosal fat (Figs. 50-214 and 50-215).270 The mean bowel wall thickness is 7.8 mm in ulcerative colitis and 11 mm in Crohn’s disease.673 In about 70% of patients with ulcerative colitis, the thickened bowel wall may show heterogeneous attenuation, with a target appearance due primarily to submucosal fat. This target appearance is observed most commonly in the rectum but less frequently in the remainder of the colon.376 Submucosal fat deposition is seen only in the chronic or subacute stages of the disease673 and may result from a long-term use of corticosteroid therapy. An additional CT feature favoring the diagnosis of ulcerative colitis includes widening of the presacral space, which results from a combination of increased presacral fat deposition, luminal narrowing of the rectum, rectal wall thickening, and stranding of the perirectal fat.362 Although rarely seen, pneumatosis may develop in severe ulcerative colitis.673

1716

PART II  CT and MR Imaging of the Whole Body

A

B FIG 50-213  Perforated tuberculous colitis. A, CT scan shows masslike bowel wall thickening (arrowheads) with colonic obstruction. An ill-defined soft tissue–attenuated lesion is seen in the pericolic space. There is also evidence of loculated fluid collection (arrow). B, Barium study shows nearly complete colonic obstruction at the hepatic flexure of the colon (arrow).

FIG 50-214  Ulcerative colitis. The bowel wall of the rectum and sigmoid colon is minimally thickened, with a target appearance (arrowheads).

FIG 50-215  Ulcerative colitis. Heterogeneous bowel wall thickening

Local complications of ulcerative colitis include perforation with peritonitis and abscess (Fig. 50-216), toxic megacolon (Fig. 50-217), venous thrombosis, and carcinoma (Fig. 50-218). The presence of toxic megacolon should be suspected on CT when the colonic lumen is considerably dilated (>8 cm).269 Typically the bowel wall will be thinner rather than thicker. Because the risk of carcinoma is markedly increased in ulcerative colitis with a long-standing history (>10 years) or with pancolitis, CT findings of any focally exaggerated colonic wall thickening with pericolonic lymphadenopathy in these patients should suggest the possibility of cancer development.

(arrows) is noted in both the ascending and descending colon. Minimal pericolonic infiltration is noted.

Diverticulitis.  Colonic diverticulitis affects approximately 5% of the population, and acute inflammation occurs in 10% to 35% of these patients.118 Diverticulitis is primarily caused by obstruction of diverticular neck with fecal material, subsequently resulting in distention and inflammation of the peridiverticular tissues. In uncomplicated diverticulitis, the inflammatory process is confined to the colonic

CHAPTER 50  Gastrointestinal Tract

1717

FIG 50-216  Bowel perforation in a patient with ulcerative colitis. The bowel wall (arrows) of the hepatic flexure of the colon is thickened, with diffuse pericolonic infiltration and fluid collection. Free gas (arrowhead) is also noted.

FIG 50-218  Colonic carcinoma complicating long-standing ulcerative colitis. The splenic flexure of the colon shows irregular bowel wall thickening (arrow) with proximal colonic dilatation caused by obstruction.

M M

M

FIG 50-219  Acute diverticulitis. The inflamed diverticular wall (arrow)

transverse colon is markedly dilated (arrows), with nodular mucosal contour.

appears to be thickened, and there is diffuse mural edema (M) in the adjacent colonic wall. (From Jang HJ, et al: Acute diverticulitis of the cecum and ascending colon: Thin-section helical CT findings. AJR Am J Roentgenol 172:601–604, 1999.)

serosa. Microscopic perforations into the pericolic fat are usually walled off, forming focal abscesses; however, continued infection and inflammation can lead to peritonitis or fistulas. The inflammatory process can also dissect longitudinally within the wall of the colon, giving rise to intramural sinus tracts. Not commonly, diverticulitis also causes large- and small bowel obstruction,401 portal and mesenteric vein gas,878 venous thrombosis, and colovenous fistula to the inferior mesenteric vein.752 The classic clinical symptoms and signs, especially in sigmoid diverticulitis, include left lower quadrant pain, tenderness, fever, and leukocytosis. These signs may mimic many other acute abdominal conditions. Although barium studies can be used, CT has been advocated as the initial procedure for radiologic evaluation for suspected acute diverticulitis. It not only is more suitable for grading the severity of pericolic inflammation but also more accurately depicts the presence, location, and size of pericolic abscesses.127,689 In addition, the use of CT is more effective in determining the alternative diagnosis.

The sensitivity of CT in diagnosis of diverticulitis is reported to range between 79% and 97%, with a specificity of 77% to 100%.127,375,689,713 According to some reports, the use of helical CT with thin slice thickness367 can improve detection of diverticulitis. The most important CT findings of diverticulitis include diverticula, inflammatory bowel wall thickening, and pericolonic fat stranding (Fig. 50-219).127 More severe cases may depict extraluminal air bubbles, phlegmon, pericolonic abscess, colonic fistula and sinus tracts, colonic obstruction, and peritonitis. In some instances, contrast material may collect to form an arrowhead shape adjacent to focally inflamed colonic wall (arrowhead sign),709 and inflamed diverticulum appears as a thick, enhancing, ring shadow within the area of inflamed peridiverticular tissues (see Fig. 50-219).367 A few reports125,639 have indicated an overlap of image findings between diverticulitis and colon cancer in about 10% of cases of diverticulitis: colonic wall thickening of more than 1 cm, associated soft tissue mass, wall thickening with luminal narrowing, wall thickening

FIG 50-217  Toxic megacolon in a patient with ulcerative colitis. The

1718

PART II  CT and MR Imaging of the Whole Body

FIG 50-220  Colon cancer simulating diverticulitis. The sigmoid colon

FIG 50-221  Appendagitis in a patient with left lower abdominal pain

shows concentric bowel wall thickening (arrows) with an abrupt transition between normal and tumoral segments. Pericolonic stranding is also noted.

and tenderness, mild fever, and leukocytosis. CT scan shows a pericolonic oval fatty lesion (arrows) with an enhancing peripheral wall. Pericolonic fat strands also surround this lesion.

without pericolic inflammation, and short-segmental wall thickening.125 CT findings that favor the diagnosis of acute diverticulitis include fluid at the root of the mesentery and vascular engorgement.639 In contrast, an abrupt zone of transition with normal bowel, enlarged pericolic lymph nodes, and mural thickness greater than 1.5 cm favor colonic carcinoma (Fig. 50-220).125 Right-sided diverticulitis represents only about 5% of cases in Western populations but accounts for 17% in Asian patients.512 Although CT findings are similar to those of left-sided diverticulitis, it may be difficult to differentiate cecal diverticulitis from appendicitis unless a normal appendix is seen. The presence of an intramural abscess or cecal diverticula in association with an inflammatory process located cephalic to a normal-appearing cecal caput and the periappendicular region should suggest cecal diverticulitis.38 The offending diverticulum may contain gas, fluid, contrast material, or calcified material.367

lesion. Periappendageal fat stranding, a thickening of adjacent parietal peritoneum, and mass effect to adjacent colon or focal wall thickening in adjacent colon may also be noted. Omental infarction may produce similar CT findings, but it appears as a larger lesion without a welldefined rim of thickened visceral peritoneum.

Epiploic Appendagitis.  Epiploic appendices (appendices epiploicae) are sacs of visceral peritoneum that are filled with fat and vascular structures and protrude from the serosal surface of the colon. The average person has approximately 100 appendices epiploicae  that are distributed from the cecum to the rectosigmoid junction. Rarely, epiploic appendages may undergo torsion or spontaneous venous thrombosis with subsequent ischemic infarction and inflammation; this condition is termed primary epiploic appendagitis.839 The torsion has been reported to be related to heavy exercise or excessive stretching.736 Secondary epiploic appendagitis is an inflammation of the epiploic appendices that is caused by inflammation of adjacent organs (i.e., in diverticulitis, appendicitis, or cholecystitis). Patients with primary epiploic appendagitis characteristically present with an abrupt onset of focal abdominal pain, mild fever, and mild elevation of the white blood cell count, which is often mistaken for appendicitis or diverticulitis. Most cases can be managed conservatively with the use of analgesics only.714 The diagnosis can be made confidently on CT, and this disease can be differentiated from other, far more common conditions such as appendicitis and diverticulitis. The important CT findings include pericolonic oval lesions with fat attenuation and surrounding thickened peritoneal lining that represents an inflamed epiploic appendix (Fig. 50-221).150,736 A central high-attenuating dot may be seen in the

Pseudomembranous Colitis.  Pseudomembranous colitis is a potentially life-threatening acute infectious disorder associated with overgrowth of Clostridium difficile after antibiotic use. Although virtually all antibiotics may produce the disorder, clindamycin, ampicillin, and cephalosporins are agents that are most commonly associated. Most cases occur when the drug is given orally rather than parenterally. Pseudomembranous colitis may also occur in other clinical settings such as bowel ischemia, intestinal obstruction, uremia, abdominal surgery, or chemotherapy.746 The disease usually involves the colon but may include only the small bowel or both the colon and small bowel. In the colorectum, the rectum and sigmoid colon are typically involved. In approximately 30% to 40% of cases, however, pseudomembranous colitis can be limited to the right side of colon, with sparing of the rectosigmoid or even the entire left side of the colon.219 The classic clinical presentation includes abdominal pain, watery diarrhea, fever, and leukocytosis. CT is helpful not only in suggesting the diagnosis but also in evaluating the extent of disease. Common CT findings include colonic wall thickening, low-attenuation mural thickening corresponding to mucosal and submucosal edema, the accordion sign, the target sign, pericolonic stranding, and ascites (Figs. 50-222 and 50-223). The average colonic wall thickness (14.7 mm) is nearly twice that reported in ulcerative colitis and more than that reported in Crohn’s disease (7.8 and 13 mm, respectively).219 The accordion sign indicates alternating edematous haustral folds separated by transverse mucosal ridges filled with oral contrast material. Although this sign was once considered specific for the diagnosis of pseudomembranous colitis,219 it can be seen in other edematous and inflammatory conditions affecting the colon, such as portal hypertensive colopathy and ischemic colitis.527 Ascites is seen in approximately 35% of cases.219 Untreated cases of pseudomembranous colitis can result in toxic megacolon with subsequent perforation and peritonitis. However, CT does not help to distinguish patients requiring surgical intervention from patients requiring conservative management.392

CHAPTER 50  Gastrointestinal Tract

1719

*

FIG 50-222  Pseudomembranous colitis. CT scan shows marked bowel wall thickening of the sigmoid colon and rectum (asterisk), producing a target appearance. Diffuse pericolic and perirectal infiltration is seen.

FIG 50-224  Cytomegalovirus colitis with bowel perforation. The sigmoid colon shows a thickened wall with a target appearance (arrows). Note minimal pericolonic strands and a small amount of ascites.

FIG 50-223  Pseudomembranous colitis. The transverse colonic wall is diffusely thickened, resulting in the accordion sign (arrowheads).

Cytomegalovirus Colitis.  CMV is a herpesvirus that can be sexually transmitted or transmitted via infected organs, blood, or needles. CMV remains in a latent state in the host after initial infection. With increasing impairment of cell-mediated immunity, particularly a CD4 lymphocyte count lower than 100 mm3, reactivation may occur.869 Clinically evident CMV infections most commonly affect immunosuppressed patients, including transplant recipients and patients with AIDS. However, CMV infection may occasionally occur in nonimmunosuppressed individuals, including patients with diabetes mellitus, those who receive corticosteroids or radiation therapy, and older patients.810 Although the incidence of small bowel involvement is not rare, the colon is the organ most commonly affected by CMV infection. Before the AIDS epidemic, the cecum was the most common site of colonic involvement, but in approximately 50% of patient with AIDS the disease was confined to the descending colon, particularly to the rectosigmoid region. In severe forms, multiple ulcerations and perforations occur; vasculitis resulting from the endothelial cell involvement in the bowel wall by CMV leads to tissue edema and necrosis with ulceration, hemorrhage, and perforation.321 The clinical symptoms of CMV colitis are persistent diarrhea, fever, weight loss, abdominal pain, and hematochezia. The most prominent CT feature is colonic wall thickening with mural edema and deep ulceration (Fig. 50-224).604 Other findings include pericolonic strand-

FIG 50-225  Development of typhlitis during chemotherapy for leukemia. CT scan shows concentric bowel wall thickening (arrows) of the cecum, with poor contrast enhancement.

ing, lymphadenopathy, and ascites. Because of the nonspecificity of these findings, it is difficult to distinguish CMV infection from other colitides such as pseudomembranous colitis, although the relative high incidence of small bowel involvement is a suggestive feature of CMV infection.604

Neutropenic Colitis (Typhlitis).  Neutropenic colitis, a necrotizing enterocolitis, occurs as a complication of acute leukemia or other neutropenic states such as aplastic anemia, SLE, or cyclic neutropenia. The cecum is most commonly involved, but the remaining colon and distal ileum may also be affected. Various factors account for predominant cecal involvement by neutropenic colitis.4 The cecum represents an area of relative stasis of bowel contents and is easily distensible. Mucosal ulcerations create a mural port of entry for the resident colonic microflora and allow overgrowth of bacteria, viruses, or fungi, causing edema, thickening, and induration of the cecal wall.4,342 Typical clinical features are fever, watery diarrhea, abdominal pain, and occasionally a palpable mass. CT findings are nonspecific and include concentric homogeneous or heterogeneous thickening of the bowel wall with intramural edema and necrosis, pericolic fluid, and pneumatosis intestinalis (Fig. 50-225).240

1720

PART II  CT and MR Imaging of the Whole Body

Uncomplicated neutropenic colitis is managed conservatively; if left untreated, neutropenic colitis progresses to transmural necrosis and colonic perforation, requiring surgery.4

Actinomycosis.  Actinomycosis is a chronic, progressive, suppurative disease characterized by the formation of multiple abscesses, draining sinuses, abundant granulation, and dense fibrous tissue. This infection is considered to be caused by Actinomyces organisms (most commonly, Actinomyces israelii), which are gram-positive anaerobic bacteria; they are not regarded as virulent human pathogens and are best considered as opportunistic because they are normally present in healthy individuals, especially in the oral cavity, tonsillar crypts, and colon.58 Dental caries are also common reservoirs of Actinomyces species.57 A. israelii has a worldwide distribution and is present with equal frequency in urban and rural dwellers.58 Although that there is no reported discernible sex predilection, most patients (94%) are women.58,903 Human actinomycosis commonly occurs in three distinct forms. Most cases of clinical disease are cervicofacial (55%), with only 20% occurring in an abdominopelvic form and 15% in thoracopulmonic form.57,903 Abdominopelvic actinomycosis has been associated with abdominal surgery (e.g., appendectomy), bowel perforation, or trauma. A long-standing intrauterine device has been emphasized as a risk factor of the infection in young women.623 Various abdominal organs can be involved in abdominopelvic actinomycosis, including the GI tract, ovaries, liver, gallbladder, and pancreas.58 In the GI tract, the rectosigmoid colon and ileocecal region, including the appendix, are the most commonly involved.58,773 Clinical features depend on which organs are involved, but common symptoms and signs include fever and leukocytosis.58,294 The presumptive diagnosis is made when “sulfur granules” are seen in the Papanicolaou smears of pus in the abscess or discharged material from the sinus tract.286 Although histologic identification of actinomycotic granules or culture of Actinomyces58 is important to establish a definitive diagnosis, positive test results are seen in less than 50% of cases.57 Early diagnosis of actinomycosis is important to minimize morbidity and to avoid unnecessary surgery, because the disease usually responds favorably to high doses of penicillin.57,294 CT plays an important role in patients with the suspicion of abdominopelvic actinomycosis by suggesting the diagnosis and determining the anatomic location and extent of disease, as well as providing a monitoring of the treatment response and disease recurrence.294,310 Because of its aggressiveness and infiltrative nature, on CT scan actinomycosis can easily be mistaken for other GI disorders such as malignancy (including carcinoid tumor), Crohn’s disease, diverticulitis, intestinal TB, and other complicated GI infections. When the GI tract is involved, the main CT features include bowel wall thickening (mostly concentric) (Fig. 50-226). The mean wall thickness of 1.2 cm and the mean length of 8.3 cm overlap considerably with those of other inflammatory bowel diseases. One of the distinguishing CT features of actinomycosis may be the presence of a large mass adjacent to the involved bowel; we have noticed this finding in 17 of 18 patients with actinomycosis in the colon or the ileum.472 Such a mass appears predominantly cystic or solid. The presence of abundant granulation and dense fibrous tissues in the solid component of the mass may produce strong contrast enhancement. Actinomycosis shows an aggressive infiltrative pattern with a propensity for crossing the fascial planes or anatomic boundaries and involving multiple compartments. These tendencies are attributed to the proteolytic enzyme produced by A. israelii. Soft tissue strandings are also prominent, surrounding the involved bowel or mass. Because of the size of the bacterium, the

* *

FIG 50-226  Actinomycosis infection involving the GI tract. CT scan shows ill-defined soft tissue lesion (arrows) occupying the right lower abdomen, as well as evidence of bowel wall thickening of the distal ileum (asterisks) and diffuse infiltration in the regional fat plane. (From Lee IJ, et al: CT features of abdominopelvic actinomycosis involving the gastrointestinal tract. Radiology 220:76–80, 2001.)

FIG 50-227  Schistosoma japonica infection. CT scan shows diffuse colonic wall calcification (arrows) in the rectum and sigmoid colon. (Courtesy T. Araki, MD, Yamanashi University, Japan.)

organism usually does not spread via the lymphatics; therefore regional lymphadenopathy is uncommon or develops late.51,903

Schistosoma japonica Infection.  Schistosoma japonica infection is confined to eastern Asia. Patients are infected by contact with fresh water that is infested with larvae. The schistosomal larvae penetrate the skin and pass via the systemic circulation to the portal vein and its tributaries. The eggs are the main cause of tissue damage to the host. An intense granulomatous reaction is followed by irregular mucosal thickening, formation of granulomatous polyps, and eventually fibrosis. CT characteristically shows curvilinear or nodular calcification in the colon, appendix, or small bowel (Fig. 50-227).468 The calcified eggs are deposited more extensively in the submucosa and serosa.468

Vascular and Miscellaneous Conditions

Colonic Ischemia.  Colonic ischemia is one of the most common vascular disorders of the large bowel in older people. Several forms of ischemic colopathy are recognized719: • Reversible colopathy (submucosal or intramural hemorrhage) • Transient colitis

CHAPTER 50  Gastrointestinal Tract • Chronic ulcerating colitis • Stricture • Gangrene • Fulminant universal colitis Acute ischemia results from an acute reduction in blood flow through an intestinal segment caused by an acute occlusive event or decreased splanchnic blood flow. A number of predisposing factors have been suggested, including arteriosclerotic heart disease, hypotensive episodes, cardiac and aortic surgery, myocardial infarction, digitalis treatment, arrhythmias, vasculitis, colonic obstruction, colonic infection such as CMV, and hypercoagulable states. Although the common denominator producing ischemia in most of these circumstances is hypoxia, no specific cause of the ischemia is usually identified. Although any region of the colon is subject to focal ischemia,42 the segments commonly affected by colonic ischemia are the junction between the superior and inferior mesenteric arteries near the splenic flexure (Griffith’s point) and the anastomotic plexus between the inferior mesenteric artery and hypogastric vessels near the rectosigmoid region (critical point of Sudeck). According to one report, isolated cecal ischemia may occur because the ileocecal region is also a “watershed area.”799 The ischemic process tends to affect the mucosa initially; a severe and prolonged process causes involvement of only the muscularis propria. Mucosal damage is reversible, occurring as a self-limited condition, whereas necrosis of the muscle layer can lead to development of a fibrotic stricture or to necrosis with severe sepsis and perforation. The diagnostic accuracy of CT in detection of colonic ischemia has not been well described in the literature, but improved contrast enhancement and scanning techniques may increase its sensitivity. On CT images, colonic ischemia appears as circumferential bowel wall thickening with or without a target sign, poor bowel wall enhancement, focal or diffuse bowel dilatation, mesenteric vascular thrombus, mesenteric vascular engorgement, pneumatosis intestinalis, portal or mesenteric venous gas, and pneumoperitoneum11,207 (Figs. 50-228 and 50-229). Although similar CT findings can be seen in a variety of inflammatory conditions (e.g., ulcerative or granulomatous colitis, pseudomembranous colitis, CMV colitis), segmental colonic involvement is much more often seen in colonic ischemia.42 Differentiating reversible from irreversible ischemia may be important to establish the strategy of patient management. Except for the presence of portal air or intraperitoneal air, however, Balthazar and associates observed that a single CT examination was limited in

1721

predicting the development of complications (i.e., stricture, infarction, and perforation).42

Radiation Colitis.  Radiation colitis is generally observed in patients who receive abdominal and pelvic radiation with doses of 4500 to 6000 rads (45 to 60 Gy).758 The latent period between radiation therapy and the development of radiation damage is usually 6 to 24 months but can be longer than 30 years.159 Predisposing factors include diabetes, hypertension, atherosclerosis, prior abdominal surgery with adhesive bowel loops, and peritonitis before the radiation therapy. Although the small intestine is more radiosensitive than the rectum, radiation injury occurs more commonly in the rectosigmoid because the rectum is fixed in position and near many sites of malignancy, such as the cervix, uterus, ovary, bladder, and prostate. Depending on the area irradiated, other sites of the colon can be involved. In contrast to the acute radiation changes, in which viable regenerating cells within the crypts of the intestinal epithelium are damaged by the direct effects of the radiation, late radiation injury can cause bowel ischemia due to obliterative endarteritis in the fine intestinal vasculature of the submucosa. Fewer than 2% of patients who receive radiotherapy ultimately require surgery resulting from perforation, bleeding, fistula, or stricture formation.89 CT may show bowel wall thickening and luminal narrowing of the irradiated bowel segment (Figs. 50-230 and 50-231). The thickened bowel wall commonly shows heterogeneous attenuation with a target appearance. The perirectal changes include thickening of the perirectal fibrous tissue and increased perirectal fat, resulting in a widened presacral space.181 Symmetrically proliferated fat immediately adjacent to the rectum and fibrous tissue of a higher attenuation surrounding it create a halo effect.181 A fibrous connection between the sacrum and rectum may also be demonstrated. Because CT findings may closely simulate those of inflammatory bowel disease or colonic ischemia, correlation with the clinical history and radiation ports is necessary for a correct diagnosis. MRI can also be used to evaluate suspected radiation colitis. According to a series by Sugimura and colleagues, the frequency of MRI tissue changes increased with incremental radiation doses.821 In addition to the perirectal changes seen on CT scans, MRI may have advantages in revealing the muscle and bone marrow changes in the

P

P

FIG 50-229  Colonic infarction in the right-sided colon. CT scan shows FIG 50-228  Colonic ischemia. CT scan shows heterogeneous colonic wall thickening (arrows) resulting from diffuse intramural hemorrhage or edema, which is confined to the right-sided colonic segment.

pneumatosis (P) and poor enhancement (arrowheads) in the ascending and proximal transverse colon. Compare the attenuation of the infracted segments (arrowheads) to normal enhancement of the descending colon (arrow). The outer margin of the infracted colon is also indistinct.

1722

PART II  CT and MR Imaging of the Whole Body

region of the radiation port; the signal intensity of the affected muscle is increased on T2-weighted images, and that of the marrow becomes increased on T1-weighted images.821 MRI has a limited role in differentiating between radiation-induced changes and residual recurrent tumor, except in the presence of a bulky mass, which favors the diagnosis of recurrent tumor. In cases of infiltrative tumor growth in the rectum or bladder, differentiation is difficult.

Pneumatosis Intestinalis.  Pneumatosis intestinalis (intramural gas) is characterized by the presence of gas in the wall of the GI tract. It can occur without identifiable causes in 15% of cases (i.e., primary) or can be associated with various GI, pulmonary, and other conditions in 85% of cases (i.e., secondary).188 The common etiologies are various

GI diseases that cause bowel necrosis (see Fig. 50-229) or mucosal disruption, pulmonary diseases such as chronic obstructive pulmonary disease, and conditions that may increase mucosal permeability, such as immunotherapy or steroid therapy (Fig. 50-232). The prognosis varies widely from spontaneous resolution to a fatal outcome, depending largely on the etiology. CT enables precise localization and assessment of the extent of intramural gas accumulation (see Figs. 50-229 and 50-232).338 All wall layers may be affected, with the most common being the subserosal and submucosal layers. In addition, CT may allow for differentiation of primary and secondary forms and assessment of the prognosis.769

FIG 50-231  Development of radiation colitis after radiation therapy for FIG 50-230  Radiation colitis after radiation therapy for cervical cancer. CT scan shows bowel wall thickening (arrows) of the rectum and sigmoid colon, with the target sign.

cecal lymphoma. CT scan shows bowel wall thickening (arrows) of the distal ileum, cecum, and ascending colon, with marked and heterogeneous contrast enhancement. Pericolonic infiltration is minimal, without free gas. Histopathologic examination of the resected specimen showed transmural infarction with microperforation.

A

B FIG 50-232  Pneumatosis of the colon in a patient in a recovering state of GVHD after bone marrow transplantation. A, Abdominal CT (lung window setting) clearly shows extensive air collection in the colonic wall (arrowheads) and also in pericolic tissue (arrows). B, A plain abdominal radiograph also shows extensive pneumatosis (arrowheads) along the course of the entire colon.

CHAPTER 50  Gastrointestinal Tract

1723

FIG 50-233  Mucocele of the appendix. A bilobed cystic mass (arrow), which is a mucin-filled dilated appendix, is noted in the pericecal region. Pathologic examination revealed mucinous cystadenoma of the appendix.

Pneumatosis intestinalis may be occasionally overlooked as intraluminal gas; therefore the use of lung window settings (see Fig. 50-233) is necessary in equivocal cases.

Appendiceal Diseases Mucocele.  Mucocele of the appendix is a descriptive term for an abnormal mucus accumulation distending the appendiceal lumen, regardless of the underlying cause. Mucoceles usually result from obstruction of the appendiceal lumen. Both benign and malignant lesions can produce a mucocele, consisting of mucosal hyperplasia, mucinous cystadenoma, and mucinous cystadenocarcinoma.320 Preoperative diagnosis is important because of the possibility of rupture at surgery, with development of pseudomyxoma peritonei. On CT a mucocele appears as a low-attenuation, well-encapsulated mass with smooth regular walls in the right lower quadrant (Fig. 50-233). The degree of attenuation depends on the amount of mucin in the mass. No periappendiceal inflammation or abscess is the key differential point in excluding acute appendicitis. Calcification occasionally develops in the wall of the mucocele, owing to a dystrophic response to a chronic inflammatory process incited by the mucus; this can ultimately lead to a porcelain appendix.147 Although uncommon, intussusception may develop. Myxoglobulosis, a variant of mucocele, is seen in 0.35% to 8% of cases and is characterized by an appendix filled with “translucent globules,” or a cluster of “frog eggs.”147 When these spheres calcify, they become evident on CT and may shift in the appendix with changes in body position. The presence of a thickened wall suggests the possibility of a neoplastic mucocele, and enhancing nodules favor the diagnosis of mucinous cystadenocarcinoma.424 Rupture of a mucocele may result in pseudomyxoma peritonei, characterized by implants of mucinous epithelium on the peritoneal surfaces and mucus accumulation within the peritoneal cavity  (Fig. 50-234). Appendicitis.  Appendicitis is one of the most common causes of acute abdomen. Although the clinical diagnosis is usually made on the basis of history, physical examination, and laboratory data, approximately 20% to 33% of patients with suspected appendicitis present with atypical features.505 Up to 20% negative appendectomy rates have been considered to be acceptable.505 CT is now regarded as the most accurate test to diagnose appendicitis. According to two meta-analyses, sensitivity and specificity of CT for diagnosis of appendicitis were 94% and 95%, respectively, in both

FIG 50-234  Mucinous cystadenocarcinoma of the appendix with pseudomyxoma peritonei. CT scan shows nodular omental infiltration and a large amount of ascites. A cystic mass (arrow) is seen in the pericecal region.

adults and adolescents836; 94% and 95%, respectively, in children179; and 94% and 94%, respectively, in adults.179 A variety of CT techniques have been used for appendiceal imaging, including nonenhanced and contrast-enhanced CT, conventional and helical CT scanning, full and limited abdominopelvic scanning, and various combinations of contrast materials.39,460,711,712 Among various techniques, use of both nonenhanced and helical CT scans appear to have some merits.460,711,712 The advantages of a nonenhanced CT scan are that its use can eliminate the risk of an adverse reaction to IV contrast material, and it is less expensive. Helical CT scans may eliminate the risk of missing the appendix by misregistration; also, relatively thin section images can be easily obtained. The importance of optimal cecal opacification and distention with contrast material has been stressed.711,712 Contrast material can either be administered only through the colon or both by mouth and through the colon; neither technique shows a statistically significant difference in achieving high diagnostic accuracy.711,712 CT features in typical cases of acute appendicitis include an abnormal appendix, periappendiceal inflammation, and changes in the cecal apex. The abnormal appendix is distended and nonopacified, showing a diameter wider than 6 mm, wall thickening and enhancement after contrast infusion, and appendicoliths (Figs. 50-235 and 50-236). Periappendiceal inflammatory changes include fat stranding, phlegmon, fluid, free air bubbles, abscess, and adenopathy (Figs. 50-237 and 50238). Changes in the cecal apex include focal cecal apex thickening, an arrowhead sign, and a cecal bar; the arrowhead sign indicates a collection of contrast material between each side of the cecal apical thickening, and the cecal bar is a straight or slightly curved band of inflamed soft tissue that separates proximal calcified appendicolith from cecal lumen.715 The presence of appendicolith does not always indicate appendicitis and may be an incidental finding. Variants of appendiceal disease include recurrent, chronic, stump, and distal appendicitis. Stump appendicitis occurs after appendectomy with simple ligation, without appendiceal stump invagination; the residual stump acts as a small appendix or diverticulum (Fig. 50-239). Distal appendicitis is a potential cause of false-negative CT results, which may show a normal cecal apex and proximal appendix, a gradual or abrupt enlargement of the distal appendix, and periappendiceal fat stranding.710

1724

PART II  CT and MR Imaging of the Whole Body

* *

FIG 50-235  Acute appendicitis. Dilated appendix (asterisks) with thickened wall due to mural edema. A small appendicolith (arrow) is also seen.

FIG 50-238  Acute appendicitis with free perforation into the peritoneal cavity. A large amount of intraperitoneal fluid collection with free gas is present.

FIG 50-236  Periappendiceal abscess with small bowel obstruction. A thick-walled cystic mass (arrows) containing high-attenuated appendicolith (arrowhead) is seen at the right lower quadrant. Small bowel loops are diffusely dilated as a result of bowel obstruction.

FIG 50-239  Stump appendicitis. A small appendicolith (curved arrow) is noted within the appendiceal stump. The wall of the cecum (arrows) is considerably thickened as a result of mural edema. There is also evidence of inflammatory infiltrates in the pericecal fat plane.

MRI can also be used in the diagnosis of appendicitis.356 Intense contrast enhancement of the inflamed appendiceal wall indicates the presence of appendicitis. However, MRI has inherent limitations in detecting appendicoliths.

REFERENCES A

FIG 50-237  Perforated acute appendicitis with abscess. An abscess (A) with air-fluid level is noted in the root of the mesentery.

1. Abdel-Nabi H, Doerr RJ, Lamonica DM, et al: Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole-body PET: Correlation with histopathologic and CT findings. Radiology 206:755–760, 1998. 2. Abdu RA, Zakhour BJ, Dallis DJ: Mesenteric venous thrombosis—1911 to 1984. Surgery 101:383–388, 1987. 3. Abrams JS, Holden WD: Tuberculosis of the gastrointestinal tract. Arch Surg 89:282–293, 1964. 4. Abramson SJ, Berdon WE, Baker DH: Childhood typhlitis: Its increasing association with acute myelogenous leukemia. Report of five cases. Radiology 146:61–64, 1983. 5. Adachi Y, Sakino I, Matsumata T, et al: Preoperative assessment of advanced gastric carcinoma using computed tomography. Am J Gastroenterol 92:872–875, 1997.

CHAPTER 50  Gastrointestinal Tract 6. Adolph JM, Kimmig BN, Georgi P, et al: Carcinoid tumors: CT and I-131 meta-iodo-benzylguanidine scintigraphy. Radiology 164:199–203, 1987. 7. Agha FP: Intussusception in adults. AJR Am J Roentgenol 146:527–531, 1986. 8. Agha FP, Keren DF: Spindle-cell squamous carcinoma of the esophagus: A tumor with biphasic morphology. AJR Am J Roentgenol 145:541–545, 1985. 9. Aguirre DA, Santosa AC, Casola G, et al: Abdominal wall hernias: Imaging features, complications, and diagnostic pitfalls at multidetector row CT. Radiographics 25:1501–1520, 2005. 10. Ak I, Stokkel MP, Pauwels EK: Positron emission tomography with 2-[18F]fluoro-2-deoxy-D-glucose in oncology. Part II. The clinical value in detecting and staging primary tumours. J Cancer Res Clin Oncol 126:560–574, 2000. 11. Alpern MB, Glazer GM, Francis IR: Ischemic or infarcted bowel: CT findings. Radiology 166:149–152, 1988. 12. Amorn Y, Knight WA, Jr: Primary linitis plastica of the colon: Report of two cases and review of the literature. Cancer 41:2420–2425, 1978. 13. Amromin GD, Solomon RD: Necrotizing enteropathy: A complication of treated leukemia or lymphoma patients. JAMA 182:23–29, 1962. 14. Anand BS, Nanda R, Sachdev GK: Response of tuberculous stricture to antituberculous treatment. Gut 29:62–69, 1988. 15. Deleted in proofs. 16. Andreola S, Leo E, Belli F, et al: Manual dissection of adenocarcinoma of the lower third of the rectum specimens for detection of lymph node metastases smaller than 5 mm. Cancer 77:607–612, 1996. 17. Anonymous: Mucormycosis. Ann Intern Med 93:93–108, 1980. 18. Anonymous: Criteria for diagnosis of Behcet’s disease. International Study Group for Behcet’s Disease. Lancet 335:1078–1080, 1990. 19. Antler AS, Ough Y, Pitchumoni CS, et al: Gastrointestinal metastases from malignant tumors of the lung. Cancer 49:170–172, 1982. 20. Antoch G, Kanja J, Bauer S, et al: Comparison of PET, CT, and dual-modality PET/CT imaging for monitoring of imatinib (STI571) therapy in patients with gastrointestinal stromal tumors. J Nucl Med 45:357–365, 2004. 21. Araoz PA, Batts KP, MacCarty RL: Amyloidosis of the alimentary canal: Radiologic-pathologic correlation of CT findings. Abdom Imaging 25:38–44, 2000. 22. Deleted in proofs. 23. Arulampalam T, Costa D, Visvikis D, et al: The impact of FDG-PET on the management algorithm for recurrent colorectal cancer. Eur J Nucl Med 28:1758–1765, 2001. 24. Ascenti G, Racchiusa S, Mazziotti S, et al: Giant fibrovascular polyp   of the esophagus: CT and MR findings. Abdom Imaging 24:109–110, 1999. 25. Attar S, Hankins JR, Suter CM, et al: Esophageal perforation: A therapeutic challenge. Ann Thorac Surg 50:45–49, discussion 50-51, 1990. 26. Backer CL, LoCicero J, 3rd, Hartz RS, et al: Computed tomography in patients with esophageal perforation. Chest 98:1078–1080, 1990. 27. Bae KT: Peak contrast enhancement in CT and MR angiography: When does it occur and why? Pharmacokinetic Study in a Porcine Model. Radiology 227:809–816, 2003. 28. Bae KT: Principles of contrast medium delivery and scan timing in MDCT. In Saini S, Rubin GD, Kalra MK, editors: MDCT: A practical approach, New York, 2006, Springer, pp 10–24. 29. Bae KT, Heiken JP, Brink JA: Aortic and hepatic peak enhancement at CT: Effect of contrast medium injection rate—Pharmacokinetic analysis and experimental porcine model. Radiology 206:455–464, 1998. 30. Baek JH, Rhim HC, Koh BH, et al: Rectal leiomyosarcoma: Report of two cases. J Korean Radiologic Soc 31:511–514, 1994. 31. Balfe DM, Mauro MA, Koehler RE, et al: Gastrohepatic ligament: Normal and pathologic CT anatomy. Radiology 150:485–490, 1984. 32. Balikian JP, Uthman SM, Kabakian HA: Tuberculous colitis. Am J Proctol 28:75–79, 1977. 33. Balthazar EJ: CT of the gastrointestinal tract: Principles and interpretation. AJR Am J Roentgenol 156:23–32, 1991.

1725

34. Balthazar EJ, Birnbaum BA, Megibow AJ, et al: Closed-loop and strangulating intestinal obstruction: CT signs. Radiology 185:769–775, 1992. 35. Balthazar EJ, Charles HW, Megibow AJ: Salmonella- and Shigellainduced ileitis: CT findings in four patients. J Comput Assist Tomogr 20:375–378, 1996. 36. Balthazar EJ: George W. Holmes Lecture. CT of small-bowel obstruction. AJR Am J Roentgenol 162:255–261, 1994. 37. Balthazar EJ, Liebeskind ME, Macari M: Intestinal ischemia in patients in whom small bowel obstruction is suspected: Evaluation of accuracy, limitations, and clinical implications of CT in diagnosis. Radiology 205:519–522, 1997. 38. Balthazar EJ, Megibow AJ, Gordon RB, et al: Cecal diverticulitis: Evaluation with CT. Radiology 162:79–81, 1987. 39. Balthazar EJ, Megibow AJ, Siegel SE, et al: Appendicitis: Prospective evaluation with high-resolution CT. Radiology 180:21–24, 1991. 40. Balthazar EJ, Naidich DP, Megibow AJ, et al: CT evaluation of esophageal varices. AJR Am J Roentgenol 148:131–135, 1987. 41. Balthazar EJ, Siegel SE, Megibow AJ, et al: CT in patients with scirrhous carcinoma of the GI tract: Imaging findings and value for tumor detection and staging. AJR Am J Roentgenol 165:839–845, 1995. 42. Balthazar EJ, Yen BC, Gordon RB: Ischemic colitis: CT evaluation of 54 cases. Radiology 211:381–388, 1999. 43. Barki Y, Elias H, Tovi F, et al: A fibrovascular polyp of the oesophagus. Br J Radiol 54:142–144, 1981. 44. Barr H, Stone N, Rembacken B: Endoscopic therapy for Barrett’s oesophagus. Gut 54:875–884, 2005. 45. Bar-Shalom R, Guralnik L, Tsalic M, et al: The additional value of PET/ CT over PET in FDG imaging of oesophageal cancer. Eur J Nucl Med Mol Imaging 32:918–924, 2005. 46. Bartnicke BJ, Balfe DM: CT appearance of intestinal ischemia and intramural hemorrhage. Radiol Clin North Am 32:845–860, 1994. 47. Bar-Ziv J, Solomon A: Computed tomography in adult intussusception. Gastrointest Radiol 16:264–266, 1991. 48. Baskin RH, Jr, Mayo CW: Jejunal diverticulosis; a clinical study of 87 cases. Surg Clin North Am 32:1185–1196, 1952. 49. Battifora H: Spindle cell carcinoma: Ultrastructural evidence of squamous origin and collagen production by the tumor cells. Cancer 37:2275–2282, 1976. 50. Beall DP, Fortman BJ, Lawler BC, et al: Imaging bowel obstruction: A comparison between fast magnetic resonance imaging and helical computed tomography. Clin Radiol 57:719–724, 2002. 51. Becker CD, Barbier P, Porcellini B: CT evaluation of patients undergoing transhiatal esophagectomy for cancer. J Comput Assist Tomogr 10:607–611, 1986. 52. Beets G, Penninckx F, Schiepers C, et al: Clinical value of whole-body positron emission tomography with [18F]fluorodeoxyglucose in recurrent colorectal cancer. Br J Surg 81:1666–1670, 1994. 53. Beets-Tan RG, Beets GL: Rectal cancer: Review with emphasis on MR imaging. Radiology 232:335–346, 2004. 54. Beets-Tan RG, Beets GL, Vliegen RF, et al: Accuracy of magnetic resonance imaging in prediction of tumour-free resection margin in rectal cancer surgery. Lancet 357:497–504, 2001. 55. Bellon EM, George CR, Schreiber H, et al: Pancreatic pseudocysts of the duodenum. AJR Am J Roentgenol 133:827–831, 1979. 56. Bennett GL, Birnbaum BA, Balthazar EJ: CT of Meckel’s diverticulitis in 11 patients. AJR Am J Roentgenol 182:625–629, 2004. 57. Bennhoff DF: Actinomycosis: Diagnostic and therapeutic considerations and a review of 32 cases. Laryngoscope 94:1198–1217, 1984. 58. Berardi RS: Abdominal actinomycosis. Surg Gynecol Obstet 149:257– 266, 1979. 59. Berardi RS: Paraduodenal hernias. Surg Gynecol Obstet 152:99–110, 1981. 60. Berkmen YM, Auh YH: CT diagnosis of acquired tracheoesophageal fistula in adults. J Comput Assist Tomogr 9:302–304, 1985. 61. Berkovich GY, Levine MS, Miller WT, Jr: CT findings in patients with esophagitis. AJR Am J Roentgenol 175:1431–1434, 2000.

1726

PART II  CT and MR Imaging of the Whole Body

62. Besemann EF, Auerbach SH, Wolfe WW: The importance of roentgenologic diagnosis of aberrant pancreatic tissue in the gastrointestinal tract. Am J Roentgenol Radium Ther Nucl Med 107:71–76, 1969. 63. Beseth BD, Bedford R, Isacoff WH, et al: Endoscopic ultrasound does not accurately assess pathologic stage of esophageal cancer after neoadjuvant chemoradiotherapy. Am Surg 66:827–831, 2000. 64. Beyer KL, Marshall JB, Diaz-Arias AA, et al: Primary small-cell carcinoma of the esophagus. Report of 11 cases and review of the literature. J Clin Gastroenterol 13:135–141, 1991. 65. Bhansali SK: Abdominal tuberculosis. Experiences with 300 cases. Am J Gastroenterol 67:324–337, 1977. 66. Bicik I, Bauerfeind P, Breitbach T, et al: Inflammatory bowel disease activity measured by positron-emission tomography. Lancet 350:262, 1997. 67. Bipat S, Glas AS, Slors FJ, et al: Rectal cancer: Local staging and assessment of lymph node involvement with endoluminal US, CT, and MR imaging—A meta-analysis. Radiology 232:773–783, 2004. 68. Bipat S, van Leeuwen MS, Comans EF, et al: Colorectal liver metastases: CT, MR imaging, and PET for diagnosis—Meta-analysis. Radiology 237:123–131, 2005. 69. Biscaldi E, Ferrero S, Fulcheri E, et al: Multislice CT enteroclysis in the diagnosis of bowel endometriosis. Eur Radiol 17:211–219, 2007. 70. Bizakis JG, Segas J, Skoulakis H, et al: Retropharyngoesophageal   abscess associated with a swallowed bone. Am J Otolaryngol 14:354–357, 1993. 71. Bizer LS, Liebling RW, Delany HM, et al: Small bowel obstruction: The role of nonoperative treatment in simple intestinal obstruction and predictive criteria for strangulation obstruction. Surgery 89:407–413, 1981. 72. Bjelovic M, Kalezic V, Petrovic M, et al: Correlation of macroscopic and histological characteristics in the regional lymph nodes of patients with rectal and sigmoidal adenocarcinoma. Hepatogastroenterology 45:433–438, 1998. 73. Blachar A, Federle MP, Brancatelli G, et al: Radiologist performance in the diagnosis of internal hernia by using specific CT findings with emphasis on transmesenteric hernia. Radiology 221:422–428, 2001. 74. Blachar A, Federle MP, Dodson SF: Internal hernia: Clinical and imaging findings in 17 patients with emphasis on CT criteria. Radiology 218:68–74, 2001. 75. Blackwell JN, Castell DO: Oesophageal motility: Recent advances and implications. Clin Sci (Lond) 67:145–151, 1984. 76. Bladergroen MR, Lowe JE, Postlethwait RW: Diagnosis and recommended management of esophageal perforation and rupture. Ann Thorac Surg 42:235–239, 1986. 77. Blomqvist L, Holm T, Goranson H, et al: MR imaging, CT and CEA scintigraphy in the diagnosis of local recurrence of rectal carcinoma. Acta Radiol 37:779–784, 1996. 78. Bodily KD, Fletcher JG, Solem CA, et al: Crohn disease mural attenuation and thickness at contrast-enhanced CT enterography— Correlation with endoscopic and histologic findings of inflammation. Radiology 238:505–516, 2006. 79. Bombardieri E, Maccauro M, De Deckere E, et al: Nuclear medicine imaging of neuroendocrine tumours. Ann Oncol 12(Suppl 2):51–61, 2001. 80. Botet JF, Lightdale CJ, Zauber AG, et al: Preoperative staging of gastric cancer: Comparison of endoscopic US and dynamic CT. Radiology 181:426–432, 1991. 81. Brady LW, Asbell SO: Malignant lymphoma of the gastrointestinal tract. Erskine Memorial Lecture, 1979. Radiology 137:291–298, 1980. 82. Bragg DG, Colby TV, Ward JH: New concepts in the non-Hodgkin lymphomas: Radiologic implications. Radiology 159:291–304, 1986. 83. Brandt LJ, Boley SJ: Ischemic and vascular lesions of the bowel. In Sleisenger MH, Fordtran JS, editors: Disease: Pathophysiology, diagnosis, and management, ed 6, Philadelphia, 1993, WB Saunders, pp 7–1961. 84. Breen DJ, Janzen DL, Zwirewich CV, et al: Blunt bowel and mesenteric injury: Diagnostic performance of CT signs. J Comput Assist Tomogr 21:706–712, 1997.

85. Bridge MF, Perzin KH: Primary adenocarcinoma of the jejunum and ileum. A clinicopathologic study. Cancer 36:1876–1887, 1975. 86. Brofman N, Atri M, Hanson JM, et al: Evaluation of bowel and mesenteric blunt trauma with multidetector CT. Radiographics 26:1119–1131, 2006. 87. Brolin RE, Krasna MJ, Mast BA: Use of tubes and radiographs in   the management of small bowel obstruction. Ann Surg 206:126–133, 1987. 88. Brooks JJ, Enterline HT: Primary gastric lymphomas. A clinicopathologic study of 58 cases with long-term follow-up and literature review. Cancer 51:701–711, 1983. 89. Brown FA: Gastrointestinal complications associated with radiation therapy—Characteristic, confusing, or compounded. Am J Dig Dis 7:1006–1025, 1962. 90. Brown G, Daniels IR, Richardson C, et al: Techniques and troubleshooting in high spatial resolution thin slice MRI for rectal cancer. Br J Radiol 78:245–251, 2005. 91. Brown G, Kirkham A, Williams GT, et al: High-resolution MRI of the anatomy important in total mesorectal excision of the rectum. AJR Am J Roentgenol 182:431–439, 2004. 92. Brown G, Richards CJ, Bourne MW, et al: Morphologic predictors of lymph node status in rectal cancer with use of high-spatial-resolution MR imaging with histopathologic comparison. Radiology 227:371–377, 2003. 93. Bruce RG, Bishof NA, Jackson EC, et al: Bilateral ureteral obstruction associated with Henoch-Schoenlein purpura. Pediatr Nephrol 11:347–349, 1997. 94. Brucher BL, Stein HJ, Zimmermann F, et al: Responders benefit from neoadjuvant radiochemotherapy in esophageal squamous cell carcinoma: Results of a prospective phase-II trial. Eur J Surg Oncol 30:963–971, 2004. 95. Bruzzi JF, Truong MT, Macapinlac H, et al: Integrated CT-PET imaging of esophageal cancer: Unexpected and unusual distribution of distant organ metastases. Curr Probl Diagn Radiol 36:21–29, 2007. 96. Bubenik O, Lopez MJ, Greco AO, et al: Gastrosplenic fistula following successful chemotherapy for disseminated histiocytic lymphoma. Cancer 52:994–996, 1983. 97. Buckley JA, Siegelman SS, Jones B, et al: The accuracy of CT staging of small bowel adenocarcinoma: CT/pathologic correlation. J Comput Assist Tomogr 21:986–991, 1997. 98. Buetow PC, Buck JL, Carr NJ, et al: Intussuscepted colonic lipomas: Loss of fat attenuation on CT with pathologic correlation in 10 cases. Abdom Imaging 21:153–156, 1996. 99. Butler JA, Cameron BL, Morrow M, et al: Small bowel obstruction   in patients with a prior history of cancer. Am J Surg 162:624–628, 1991. 100. Buy JN, Moss AA: Computed tomography of gastric lymphoma. AJR Am J Roentgenol 138:859–865, 1982. 101. Byun JY, Ha HK, Yu SY, et al: CT features of systemic lupus erythematosus in patients with acute abdominal pain: Emphasis on ischemic bowel disease. Radiology 211:203–209, 1999. 102. Cabal E, Holtz S: Polyarteritis as a cause of intestinal hemorrhage. Gastroenterology 61:99–105, 1971. 103. Cademartiri F, Nieman K, van der Lugt A, et al: Intravenous contrast material administration at 16-detector row helical CT coronary angiography: Test bolus versus bolus-tracking technique. Radiology 233:817–823, 2004. 104. Calle S, Klatsky S: Intestinal phycomycosis (mucormycosis). Am J Clin Pathol 45:264–272, 1966. 105. Callstrom MR, Johnson CD, Fletcher JG, et al: CT colonography without cathartic preparation: Feasibility study. Radiology 219:693–698, 2001. 106. Camishion RC, Gibbon JH, Jr, Templeton JY, 3rd: Leiomyosarcoma of the esophagus: Review of the literature and report of two cases. Ann Surg 153:951–956, 1961. 107. Caoili EM, Paulson EK: CT of small-bowel obstruction: Another perspective using multiplanar reformations. AJR Am J Roentgenol 174:993–998, 2000.

CHAPTER 50  Gastrointestinal Tract 108. Cappell MS, Hassan T, Rosenthal S, et al: Gastrointestinal obstruction due to Mycobacterium avium intracellulare associated with the acquired immunodeficiency syndrome. Am J Gastroenterol 87:1823–1827, 1992. 109. Carlisle JG, Quint LE, Francis IR, et al: Recurrent esophageal carcinoma: CT evaluation after esophagectomy. Radiology 189:271–275, 1993. 110. Carnovale RL, Goldstein HM, Zornoza J, et al: Radiologic manifestations of esophageal lymphoma. AJR Am J Roentgenol 128:751–754, 1977. 111. Caroline DF, Friedman AC: The radiology of inflammatory bowel disease. Med Clin North Am 78:1353–1385, 1994. 112. Carter MM, Kulkarni MV: Giant fibrovascular polyp of the esophagus. Gastrointest Radiol 9:301–303, 1984. 113. Carucci LR, Levine MS, Rubesin SE, et al: Tumorous gastric varices: Radiographic findings in 10 patients. Radiology 212:861–865, 1999. 114. Caruso RD, Berk RN: Lymphoma of the esophagus. Radiology 95:381–382, 1970. 115. Caskey CI, Scatarige JC, Fishman EK: Distribution of metastases in breast carcinoma: CT evaluation of the abdomen. Clin Imaging 15:166–171, 1991. 116. Castro EB, Stearns MW: Lipoma of the large intestine: A review of 45 cases. Dis Colon Rectum 15:441–444, 1972. 117. Cerfolio RJ, Bryant AS, Ohja B, et al: The accuracy of endoscopic ultrasonography with fine-needle aspiration, integrated positron emission tomography with computed tomography, and computed tomography in restaging patients with esophageal cancer after neoadjuvant chemoradiotherapy. J Thorac Cardiovasc Surg 129:1232– 1241, 2005. 118. Chappuis CW, Cohn I, Jr: Acute colonic diverticulitis. Surg Clin North Am 68:301–313, 1988. 119. Chen J, Cheong JH, Yun MJ, et al: Improvement in preoperative staging of gastric adenocarcinoma with positron emission tomography. Cancer 103:2383–2390, 2005. 120. Chen MY, Frederick MG: Barrett esophagus and adenocarcinoma. Radiol Clin North Am 32:1167–1181, 1994. 121. Chen CY, Hsu JS, Wu DC, et al: Gastric cancer: Preoperative local staging with 3D multi-detector row CT—Correlation with surgical and histopathologic results. Radiology 242:472–482, 2007. 122. Chen SC, Lu DS, Hecht JR, et al: CT colonography: Value of scanning in both the supine and prone positions. AJR Am J Roentgenol 172:595–599, 1999. 123. Chen F, Ni YC, Zheng KE, et al: Spiral CT in gastric carcinoma: Comparison with barium study, fiberoptic gastroscopy and histopathology. World J Gastroenterol 9:1404–1408, 2003. 124. Chiechi MV, Hamrick-Turner J, Abbitt PL: Gastric herniation and volvulus: CT and MR appearance. Gastrointest Radiol 17:99–101, 1992. 125. Chintapalli KN, Esola CC, Chopra S, et al: Pericolic mesenteric lymph nodes: An aid in distinguishing diverticulitis from cancer of the colon. AJR Am J Roentgenol 169:1253–1255, 1997. 126. Cho JS, Kim JK, Rho SM, et al: Preoperative assessment of gastric carcinoma: Value of two-phase dynamic CT with mechanical IV injection of contrast material. AJR Am J Roentgenol 163:69–75, 1994. 127. Cho KC, Morehouse HT, Alterman DD, et al: Sigmoid diverticulitis: Diagnostic role of CT—Comparison with barium enema studies. Radiology 176:111–115, 1990. 128. Choi H, Charnsangavej C, de Castro Faria S, et al: CT evaluation of the response of gastrointestinal stromal tumors after imatinib mesylate treatment: A quantitative analysis correlated with FDG PET findings. AJR Am J Roentgenol 183:1619–1628, 2004. 129. Chott A, Augustin I, Wrba F, et al: Peripheral T-cell lymphomas: A clinicopathologic study of 75 cases. Hum Pathol 21:1117–1125, 1990. 130. Chott A, Dragosics B, Radaszkiewicz T: Peripheral T-cell lymphomas of the intestine. Am J Pathol 141:1361–1371, 1992. 131. Chou CK, Chen LT, Sheu RS, et al: MRI manifestations of gastrointestinal lymphoma. Abdom Imaging 19:495–500, 1994. 132. Chou CK, Mak CW, Hou CC, et al: CT of large small-bowel diverticulum. Abdom Imaging 23:132–134, 1998.

1727

133. Chun HJ, Byun JY, Chun KA, et al: Gastrointestinal leiomyoma and leiomyosarcoma: CT differentiation. J Comput Assist Tomogr 22:69–74, 1998. 134. Clark KE, Foley WD, Lawson TL, et al: CT evaluation of esophageal and upper abdominal varices. J Comput Assist Tomogr 4:510–515, 1980. 135. Cole WH: The mechanisms of spread of cancer. Surg Gynecol Obstet 137:853–871, 1973. 136. Conklin EF, Markowitz AM: Intussusception, a complication of gastric surgery. Surgery 57:480–488, 1965. 137. Cook AO, Levine BA, Sirinek KR, et al: Evaluation of gastric adenocarcinoma. Abdominal computed tomography does not replace celiotomy. Arch Surg 121:603–606, 1986. 138. Coppa GF, Eng K, Localio SA: Surgical management of diffuse cavernous hemangioma of the colon, rectum and anus. Surg Gynecol Obstet 159:17–22, 1984. 139. Cortese AF, Cornell GN: Carcinoma of the duodenum. Cancer 29:1010–1015, 1972. 140. Costamagna G, Shah SK, Riccioni ME, et al: A prospective trial comparing small bowel radiographs and video capsule endoscopy for suspected small bowel disease. Gastroenterology 123:999–1005, 2002. 141. Cotton PB, Durkalski VL, Pineau BC, et al: Computed tomographic colonography (virtual colonoscopy): A multicenter comparison with standard colonoscopy for detection of colorectal neoplasia. JAMA 291:1713–1719, 2004. 142. Coulomb M, Lebas JF, Sarrazin R, et al: [Oesophageal cancer extension. Diagnostic contribution and effects on therapy of computed tomography. Report on 40 cases (author’s transl)]. J Radiol 62:475–487, 1981. 143. Cover TL, Aber RC: Yersinia enterocolitica. N Engl J Med 321:16–24, 1989. 144. Cruz FO, Soffia PS, Del Rio PM, et al: Acute phlegmonous gastritis   with mural abscess: CT diagnosis. AJR Am J Roentgenol 159:767–768, 1992. 145. Cubillo E: Obturator hernia diagnosed by computed tomography. AJR Am J Roentgenol 140:735–736, 1983. 146. Dachman AH, Kuniyoshi JK, Boyle CM, et al: CT colonography with three-dimensional problem solving for detection of colonic polyps. AJR Am J Roentgenol 171:989–995, 1998. 147. Dachman AH, Lichtenstein JE, Friedman AC: Mucocele of the appendix and pseudomyxoma peritonei. AJR Am J Roentgenol 144:923–929, 1985. 148. Daffner RH: Computed tomography of the esophagus. Crit Rev Diagn Imaging 14:191–242, 1981. 149. Daffner RH, Halber MD, Postlethwait RW, et al: CT of the esophagus. II. Carcinoma. AJR Am J Roentgenol 133:1051–1055, 1979. 150. Danielson K, Chernin MM, Amberg JR, et al: Epiploic appendicitis: CT characteristics. J Comput Assist Tomogr 10:142–143, 1986. 151. Darling RC, Welch CE: Tumors of the small intestine. N Engl J Med 260:397–408, 1959. 152. Dauterive AH, Flancbaum L, Cox EF: Blunt intestinal trauma. A modern-day review. Ann Surg 201:198–203, 1985. 153. Day DL, Drake DG, Leonard AS, et al: CT findings in left paraduodenal herniae. Gastrointest Radiol 13:27–29, 1988. 154. Day EA, Marks C: Gallstone ileus. Review of the literature and presentation of thirty-four new cases. Am J Surg 129:552–558, 1975. 155. Day DL, Ramsay NK, Letourneau JG: Pneumatosis intestinalis after bone marrow transplantation. AJR Am J Roentgenol 151:85–87, 1988. 156. De Backer AI, Mortele KJ, Deeren D, et al: Abdominal tuberculous lymphadenopathy: MRI features. Eur Radiol 15:2104–2109, 2005. 157. de Bree E, Grammatikakis J, Christodoulakis M, et al: The clinical significance of acquired jejunoileal diverticula. Am J Gastroenterol 93:2523–2528, 1998. 158. de Bree E, Koops W, Kroger R, et al: Peritoneal carcinomatosis from colorectal or appendiceal origin: Correlation of preoperative CT with intraoperative findings and evaluation of interobserver agreement.   J Surg Oncol 86:64–73, 2004. 159. DeCosse JJ, Rhodes RS, Wentz WB, et al: The natural history and management of radiation induced injury of the gastrointestinal tract. Ann Surg 170:369–384, 1969.

1728

PART II  CT and MR Imaging of the Whole Body

160. Dehn TC, Reznek RH, Nockler IB, et al: The pre-operative assessment of advanced gastric cancer by computed tomography. Br J Surg 71:413–417, 1984. 161. de Lange EE, Fechner RE, Wanebo HJ: Suspected recurrent rectosigmoid carcinoma after abdominoperineal resection: MR imaging and histopathologic findings. Radiology 170:323–328, 1989. 162. De La Pava S, Nigogosyan G, Pickren JW, et al: Melanosis of the esophagus. Cancer 16:48–50, 1963. 163. D’Elia F, Zingarelli A, Palli D, et al: Hydro-dynamic CT preoperative staging of gastric cancer: Correlation with pathological findings. A prospective study of 107 cases. Eur Radiol 10:1877–1885, 2000. 164. de Lutio di Castelguidone E, Merola S, Pinto A, et al: Esophageal injuries: Spectrum of multidetector row CT findings. Eur J Radiol 59:344–348, 2006. 165. DeMeester SR: Adenocarcinoma of the esophagus and cardia: A review of the disease and its treatment. Ann Surg Oncol 13:12–30, 2006. 166. Demetri GD, Benjamin RS, Blanke CD, et al: NCCN Task Force report: Management of patients with gastrointestinal stromal tumor (GIST)— Update of the NCCN clinical practice guidelines. J Natl Compr Canc Netw 5(Suppl 2):1–29, quiz S30, 2007. 167. Demos TC, Okrent DH, Studlo JD, et al: Spontaneous esophageal hematoma diagnosed by computed tomography. J Comput Assist Tomogr 10:133–135, 1986. 168. De Potter T, Flamen P, Van Cutsem E, et al: Whole-body PET with FDG for the diagnosis of recurrent gastric cancer. Eur J Nucl Med Mol Imaging 29:525–529, 2002. 169. de Roo T, Vaas F: Leiomyosarcoma of the transverse and descending colon. Two case reports and review. Am J Gastroenterol 52:150–156, 1969. 170. Desai RK, Tagliabue JR, Wegryn SA, et al: CT evaluation of wall thickening in the alimentary tract. Radiographics 11:771–783, discussion 784, 1991. 171. Desai DC, Zervos EE, Arnold MW, et al: Positron emission tomography affects surgical management in recurrent colorectal cancer patients. Ann Surg Oncol 10:59–64, 2003. 172. Desch CE, Benson AB, 3rd, Somerfield MR, et al: Colorectal cancer surveillance: 2005 update of an American Society of Clinical Oncology practice guideline. J Clin Oncol 23:8512–8519, 2005. 173. Deutsch JP, Mariette D, Moukarbel N, et al: Gastroduodenal intussusception secondary to Menetrier’s disease. Abdom Imaging 19:207–209, 1994. 174. Dick AP, Graff R, Gregg DM, et al: An arteriographic study of mesenteric arterial disease. I. Large vessel changes. Gut 8:206–220, 1967. 175. Dodd GD: Lymphoma of the hollow abdominal viscera. Radiol Clin North Am 28:771–783, 1990. 176. Domizio P, Owen RA, Shepherd NA, et al: Primary lymphoma of the small intestine. A clinicopathological study of 119 cases. Am J Surg Pathol 17:429–442, 1993. 177. Donnelly LF, Morris CL: Acute graft-versus-host disease in children: Abdominal CT findings. Radiology 199:265–268, 1996. 178. Donohue JH, Federle MP, Griffiths BG, et al: Computed tomography   in the diagnosis of blunt intestinal and mesenteric injuries. J Trauma 27:11–17, 1987. 179. Doria AS, Moineddin R, Kellenberger CJ, et al: US or CT for diagnosis of appendicitis in children and adults? A meta-analysis. Radiology 241:83–94, 2006. 180. Dosik GM, Luna M, Valdivieso M, et al: Necrotizing colitis in patients with cancer. Am J Med 67:646–656, 1979. 181. Doubleday LC, Bernardino ME: CT findings in the perirectal area following radiation therapy. J Comput Assist Tomogr 4:634–638, 1980. 182. Dowe MF, Shanmuganathan K, Mirvis SE, et al: CT findings of mesenteric injury after blunt trauma: Implications for surgical intervention. AJR Am J Roentgenol 168:425–428, 1997. 183. Dreiling DA: Small bowel perforation in blunt trauma. Its relationship to previous laparotomy. Am J Gastroenterol 50:279–288, 1968. 184. Dudiak KM, Johnson CD, Stephens DH: Primary tumors of the small intestine: CT evaluation. AJR Am J Roentgenol 152:995–998, 1989.

185. Dux M, Richter GM, Hansmann J, et al: Helical hydro-CT for diagnosis and staging of gastric carcinoma. J Comput Assist Tomogr 23:913–922, 1999. 186. Dworak O: Number and size of lymph nodes and node metastases in rectal carcinomas. Surg Endosc 3:96–99, 1989. 187. Earls JP, Rofsky NM, DeCorato DR, et al: Breath-hold single-dose gadolinium-enhanced three-dimensional MR aortography: Usefulness of a timing examination and MR power injector. Radiology 201:705– 710, 1996. 188. Earnest DL, Hixson LJ: Other disease of the colon and rectum.   In Sleisenger M, Fordtran JS, editors: Gastrointestinal disease: Pathophysiology, diagnosis, and management, ed 5, Philadelphia, 1989, Saunders, pp 1537–1570. 189. Edwards DA: Achalasia: Symptoms and radiology. Postgrad Med J 50:209–210, 1974. 190. Ehrenpreis ED, Nogueras JJ, Botoman VA, et al: Serum electrolyte abnormalities secondary to Fleet’s Phospho-Soda colonoscopy prep. A review of three cases. Surg Endosc 10:1022–1024, 1996. 191. Ehrlich AN, Stalder G, Geller W, et al: Gastrointestinal manifestations of malignant lymphoma. Gastroenterology 54:1115–1121, 1968. 192. Ell C, Remke S, May A, et al: The first prospective controlled trial comparing wireless capsule endoscopy with push enteroscopy in chronic gastrointestinal bleeding. Endoscopy 34:685–689, 2002. 193. Ellis H, Gajraj H, George CD: Incisional hernias: When do they occur? Br J Surg 70:290–291, 1983. 194. Eloubeidi MA, Wallace MB, Reed CE, et al: The utility of EUS and EUS-guided fine needle aspiration in detecting celiac lymph node metastasis in patients with esophageal cancer: A single-center experience. Gastrointest Endosc 54:714–719, 2001. 195. Endicott JN, Molony TB, Campbell G, et al: Esophageal perforations: The role of computerized tomography in diagnosis and management decisions. Laryngoscope 96:751–757, 1986. 196. Erasmus JJ, Munden RF: The role of integrated computed tomography positron-emission tomography in esophageal cancer: Staging and assessment of therapeutic response. Semin Radiat Oncol 17:29–37, 2007. 197. Ernst O, Bulois P, Saint-Drenant S, et al: Helical CT in acute lower gastrointestinal bleeding. Eur Radiol 13:114–117, 2003. 198. Esophagus. In American Joint Committee on Cancer: AJCC Cancer Staging Manual, 6th ed. New York, 2002, pp 91-98. 199. Etherington RJ, Williams JG, Hayward MW, et al: Demonstration of para-ileostomy herniation using computed tomography. Clin Radiol 41:333–336, 1990. 200. Evans EI: The mechanism of shock in intestinal strangulation: An experimental study. Ann Surg 117:28–38, 1943. 201. Fabian TC, Mangiante EC, White TJ, et al: A prospective study of 91 patients undergoing both computed tomography and peritoneal lavage following blunt abdominal trauma. J Trauma 26:602–608, 1986. 202. Fahl JC, Dockerty MB, Judd ES: Scirrhous carcinoma of the colon and rectum. Surg Gynecol Obstet 111:759–766, 1960. 203. Faling LJ, Pugatch RD, Robbins AH: The diagnosis of unsuspected esophageal perforation by computed tomography. Am J Med Sci 281:31–34, 1981. 204. Falk GW, Richter JE: Why is esophageal adenocarcinoma increasing in incidence? Abdom Imaging 23:539–542, 1998. 205. Fang JF, Chen RJ, Lin BC: Cell count ratio: New criterion of diagnostic peritoneal lavage for detection of hollow organ perforation. J Trauma 45:540–544, 1998. 206. Farah MC, Jafri SZ, Schwab RE, et al: Duodenal neoplasms: Role of CT. Radiology 162:839–843, 1987. 207. Federle MP, Chun G, Jeffrey RB, et al: Computed tomographic findings in bowel infarction. AJR Am J Roentgenol 142:91–95, 1984. 208. Federle MP, Yagan N, Peitzman AB, et al: Abdominal trauma: Use of oral contrast material for CT is safe. Radiology 205:91–93, 1997. 209. Fenlon HM, O’Keane C, Carney DN, et al: Imaging of small cell carcinoma of the oesophagus. Clin Radiol 50:634–638, 1995. 210. Ferguson MK, Skinner DB: Carcinoma of the esophagus and cardia. In Zuidema GA, editor: Shackerford’s surgery of the alimentary tract, Philadelphia, 1995, WB Saunders, pp 305–332.

CHAPTER 50  Gastrointestinal Tract 211. Fernet P, Azar HA, Stout AP: Intramural (tubal) spread of linitis plastica along the alimentary tract. Gastroenterology 48:419–424, 1965. 212. Ferrara JL, Deeg HJ: Graft-versus-host disease. N Engl J Med 324:667– 674, 1991. 213. Ferrozzi F, Bova D, Garlaschi G: Gastric liposarcoma: CT appearance. Abdom Imaging 18:232–233, 1993. 214. Ferrucci JT: Colonoscopy: virtual and optical—Another look, another view. Radiology 235:13–16, 2005. 215. Fidler J: MR Imaging of the small bowel. Radiol Clin North Am 45:317–331, 2007. 216. Filippone A, Cianci R, Storto ML: Bowel obstruction: Comparison between multidetector-row CT axial and coronal planes. Abdom Imaging 32:310–316, 2007. 217. Fiorica F, Di Bona D, Schepis F, et al: Preoperative chemoradiotherapy for oesophageal cancer: A systematic review and meta-analysis. Gut 53:925–930, 2004. 218. Fischer RP, Beverlin BC, Engrav LH, et al: Diagnostic peritoneal   lavage: Fourteen years and 2,586 patients later. Am J Surg 136:701–704, 1978. 219. Fishman EK, Kavuru M, Jones B, et al: Pseudomembranous colitis: CT evaluation of 26 cases. Radiology 180:57–60, 1991. 220. Fishman EK, Wolf EJ, Jones B, et al: CT evaluation of Crohn’s disease: Effect on patient management. AJR Am J Roentgenol 148:537–540, 1987. 221. Fishman EK, Zinreich ES, Jones B, et al: Computed tomographic diagnosis of radiation ileitis. Gastrointest Radiol 9:149–152, 1984. 222. Fisk JD, Shulman HM, Greening RR, et al: Gastrointestinal radiographic features of human graft-vs.-host disease. AJR Am J Roentgenol 136:329–336, 1981. 223. Flamen P, Hoekstra OS, Homans F, et al: Unexplained rising carcinoembryonic antigen (CEA) in the postoperative surveillance of colorectal cancer: The utility of positron emission tomography (PET). Eur J Cancer 37:862–869, 2001. 224. Flamen P, Lerut A, Van Cutsem E, et al: The utility of positron emission tomography for the diagnosis and staging of recurrent esophageal cancer. J Thorac Cardiovasc Surg 120:1085–1092, 2000. 225. Flamen P, Lerut A, Van Cutsem E, et al: Utility of positron emission tomography for the staging of patients with potentially operable esophageal carcinoma. J Clin Oncol 18:3202–3210, 2000. 226. Flanagan FL, Dehdashti F, Ogunbiyi OA, et al: Utility of FDG-PET   for investigating unexplained plasma CEA elevation in patients with colorectal cancer. Ann Surg 227:319–323, 1998. 227. Flejou JF, Svrcek M: Barrett’s oesophagus—A pathologist’s view. Histopathology 50:3–14, 2007. 228. Fleming JL, Wiesner RH, Shorter RG: Whipple’s disease: Clinical, biochemical, and histopathologic features and assessment of treatment in 29 patients. Mayo Clin Proc 63:539–551, 1988. 229. Fletcher JG, Booya F, Johnson CD, et al: CT colonography: Unraveling the twists and turns. Curr Opin Gastroenterol 21:90–98, 2005. 230. Fletcher JD, Covell L, Shortsleeve MJ: Calcified gastric duplication. AJR Am J Roentgenol 163:994–995, 1994. 231. Fletcher JG, Johnson CD, Krueger WR, et al: Contrast-enhanced CT colonography in recurrent colorectal carcinoma: Feasibility of simultaneous evaluation for metastatic disease, local recurrence, and metachronous neoplasia in colorectal carcinoma. AJR Am J Roentgenol 178:283–290, 2002. 232. Fong Y, Saldinger PF, Akhurst T, et al: Utility of 18F-FDG positron emission tomography scanning on selection of patients for resection of hepatic colorectal metastases. Am J Surg 178:282–287, 1999. 233. Fortson WC, Tedesco FJ: Drug-induced colitis: A review. Am J Gastroenterol 79:878–883, 1984. 234. Fox U, Lucani G: Disorders of the intestinal mesenteric lymphatic system. Lymphology 26:61–66, 1993. 235. Frager D, Baer JW, Medwid SW, et al: Detection of intestinal ischemia in patients with acute small-bowel obstruction due to adhesions or hernia: Efficacy of CT. AJR Am J Roentgenol 166:67–71, 1996. 236. Frager DH, Goldman M, Beneventano TC: Computed tomography in Crohn disease. J Comput Assist Tomogr 7:819–824, 1983.

1729

237. Frager D, Medwid SW, Baer JW, et al: CT of small-bowel obstruction: Value in establishing the diagnosis and determining the degree and cause. AJR Am J Roentgenol 162:37–41, 1994. 238. Francis RS, Berk RN: Typhoid fever. Radiology 112:583–585, 1974. 239. Frazzini VI, Jr, English WJ, Bashist B, et al: Case report. Small bowel obstruction due to phytobezoar formation within Meckel diverticulum: CT findings. J Comput Assist Tomogr 20:390–392, 1996. 240. Frick MP, Maile CW, Crass JR, et al: Computed tomography of neutropenic colitis. AJR Am J Roentgenol 143:763–765, 1984. 241. Fried AM, Poulos A, Hatfield DR: The effectiveness of the incidental small-bowel series. Radiology 140:45–46, 1981. 242. Fujimoto T, Fukuda T, Uetani M, et al: Unenhanced CT findings of vascular compromise in association with intussusceptions in adults. AJR Am J Roentgenol 176:1167–1171, 2001. 243. Fukuya T, Hawes DR, Lu CC, et al: CT diagnosis of small-bowel obstruction: Efficacy in 60 patients. AJR Am J Roentgenol 158:765–769, discussion 771-772, 1992. 244. Fukuya T, Honda H, Hayashi T, et al: Lymph-node metastases: Efficacy for detection with helical CT in patients with gastric cancer. Radiology 197:705–711, 1995. 245. Fukuya T, Honda H, Kaneko K, et al: Efficacy of helical CT in   T-staging of gastric cancer. J Comput Assist Tomogr 21:73–81, 1997. 246. Furukawa H, Ikuma H, Seki A, et al: Positron emission tomography scanning is not superior to whole body multidetector helical computed tomography in the preoperative staging of colorectal cancer. Gut 55:1007–1011, 2006. 247. Deleted in proofs. 248. Gambhir SS, Czernin J, Schwimmer J, et al: A tabulated summary of the FDG PET literature. J Nucl Med 42:1S–93S, 2001. 249. Gamsu G, Webb WR: Computed tomography of the trachea: Normal and abnormal. AJR Am J Roentgenol 139:321–326, 1982. 250. Gayer G, Zissin R, Apter S, et al: Acute diverticulitis of the small bowel: CT findings. Abdom Imaging 24:452–455, 1999. 251. Gazelle GS, Goldberg MA, Wittenberg J, et al: Efficacy of CT in distinguishing small-bowel obstruction from other causes of smallbowel dilatation. AJR Am J Roentgenol 162:43–47, 1994. 252. George S, Desai J: Management of gastrointestinal stromal tumors in the era of tyrosine kinase inhibitors. Curr Treat Options Oncol 3:489–496, 2002. 253. Gerson LB, Shetler K, Triadafilopoulos G: Prevalence of Barrett’s esophagus in asymptomatic individuals. Gastroenterology 123:461–467, 2002. 254. Gertz MA: Secondary amyloidosis (AA). J Intern Med 232:517–518, 1992. 255. Ghahremani GG: Internal abdominal hernias. Surg Clin North Am 64:393–406, 1984. 256. Glick SG, Johnson CD, Macari M, et al: ACR practice guideline for   the performance of computed tomography (CT) colonography in adults. ACR Practice Guidelines and Technical Standards 2005-2006:  295-299. 257. Glick SN, Levine MS, Teplick SK, et al: Splenic penetration by benign gastric ulcer: Preoperative recognition with CT. Radiology 163:637–639, 1987. 258. Glick SN, Teplick SK, Levine MS, et al: Gastric cardia metastasis in esophageal carcinoma. Radiology 160:627–630, 1986. 259. Goerres GW, Stupp R, Barghouth G, et al: The value of PET, CT and in-line PET/CT in patients with gastrointestinal stromal tumours: Long-term outcome of treatment with imatinib mesylate. Eur J Nucl Med Mol Imaging 32:153–162, 2005. 260. Goldberg HI, Gore RM, Margulis AR, et al: Computed tomography in the evaluation of Crohn disease. AJR Am J Roentgenol 140:277–282, 1983. 261. Goldman LP, Weigert JM: Corrosive substance ingestion: A review. Am J Gastroenterol 79:85–90, 1984. 262. Goldstein HM, Zornoza J, Hopens T: Intrinsic diseases of the adult esophagus: Benign and malignant tumors. Semin Roentgenol 16:183– 197, 1981.

1730

PART II  CT and MR Imaging of the Whole Body

263. Gollub MJ, Prowda JC: Primary melanoma of the esophagus: Radiologic and clinical findings in six patients. Radiology 213:97–100, 1999. 264. Gollub MJ, Schwartz LH, Akhurst T: Update on colorectal cancer imaging. Radiol Clin North Am 45:85–118, 2007. 265. Goodman MD, Porter DD: Cytomegalovirus vasculitis with fatal colonic hemorrhage. Arch Pathol 96:281–284, 1973. 266. Gore RM: Cross-sectional imaging of inflammatory bowel disease. Radiol Clin North Am 25:115–131, 1987. 267. Gore RM: Esophageal cancer. Clinical and pathologic features. Radiol Clin North Am 35:243–263, 1997. 268. Gore RM, Balthazar EJ, Ghahremani GG, et al: CT features of   ulcerative colitis and Crohn’s disease. AJR Am J Roentgenol 167:3–15, 1996. 269. Gore RM, Laufer I: Ulcerative and granulomatous colitis. In Gore RM, Levine MS, Laufer I, editors: Textbook of gastrointestinal radiology, Philadelphia, 1994, Saunders, pp 1098–1141. 270. Gore RM, Marn CS, Kirby DF, et al: CT findings in ulcerative, granulomatous, and indeterminate colitis. AJR Am J Roentgenol 143:279–284, 1984. 271. Gourtsoyiannis NC, Grammatikakis J, Papamastorakis G, et al: Imaging of small intestinal Crohn’s disease: Comparison between MR enteroclysis and conventional enteroclysis. Eur Radiol 16:1915–1925, 2006. 272. Gourtsoyiannis N, Papanikolaou N, Grammatikakis J, et al: Assessment of Crohn’s disease activity in the small bowel with MR and conventional enteroclysis: Preliminary results. Eur Radiol 14:1017–1024, 2004. 273. Govoni AF: Hemangiomas of esophagus. Gastointest Radiol 7:113–117, 1982. 274. Graham WP, 3rd: Gastro-intestinal metastases from carcinoma of the breast. Ann Surg 159:477–480, 1964. 275. Graham AJ, Shrive FM, Ghali WA, et al: Defining the optimal treatment of locally advanced esophageal cancer: A systematic review and decision analysis. Ann Thorac Surg 83:1257–1264, 2007. 276. Granfield CA, Charnsangavej C, Dubrow RA, et al: Regional lymph node metastases in carcinoma of the left side of the colon and rectum: CT demonstration. AJR Am J Roentgenol 159:757–761, 1992. 277. Grassi R, di Mizio R, Romano S, et al: Multiple jejunal angiodysplasia detected by enema-helical CT. Clin Imaging 24:61–63, 2000. 278. Greene FL, Page DL, Fleming ID: AJCC Manual of Staging of Cancer, ed 6, New York, 2002, Spinger-Verlag. 279. Greene F, Page D, Fleming I, et al: AJCC Cancer Staging Manual, New York, 2002, Springer. 280. Greenstein AJ, Gennuso R, Sachar DB, et al: Extraintestinal cancers in inflammatory bowel disease. Cancer 56:2914–2921, 1985. 281. Greenstein AJ, Sachar D, Pucillo A, et al: Cancer in Crohn’s disease after diversionary surgery. A report of seven carcinomas occurring in excluded bowel. Am J Surg 135:86–90, 1978. 282. Griffith JF, Chan AC, Chow LT, et al: Assessing chemotherapy response of squamous cell oesophageal carcinoma with spiral CT. Br J Radiol 72:678–684, 1999. 283. Gross BH, Agha FP, Glazer GM, et al: Gastric interposition following transhiatal esophagectomy: CT evaluation. Radiology 155:177–179, 1985. 284. Grumbach K, Levine MS, Wexler JA: Gallstone ileus diagnosed by computed tomography. J Comput Assist Tomogr 10:146–148, 1986. 285. Gumaste VV, Dave PB: Ingestion of corrosive substances by adults. Am J Gastroenterol 87:1–5, 1992. 286. Gupta PK, Hollander DH, Frost JK: Actinomycetes in cervico-vaginal smears: An association with IUD usage. Acta Cytol 20:295–297, 1976. 287. Gurgui Ferrer M, Prats Pastor G, Mirelis Otero B: Yersinia enterocolitica: Intestinal features. Dig Dis 8:313–321, 1990. 288. Ha HK: CT in the early detection of strangulation in intestinal obstruction. Semin Ultrasound CT MR 16:141–150, 1995. 289. Ha HK, Jee KR, Yu E, et al: CT features of metastatic linitis plastica to the rectum in patients with peritoneal carcinomatosis. AJR Am J Roentgenol 174:463–466, 2000.

290. Ha HK, Jung JI, Lee MS, et al: CT differentiation of tuberculous peritonitis and peritoneal carcinomatosis. AJR Am J Roentgenol 167:743–748, 1996. 291. Ha HK, Kim HH, Kim HS, et al: Local recurrence after surgery for gastric carcinoma: CT findings. AJR Am J Roentgenol 161:975–977, 1993. 292. Ha HK, Kim JS, Lee MS, et al: Differentiation of simple and strangulated small-bowel obstructions: Usefulness of known CT criteria. Radiology 204:507–512, 1997. 293. Ha HK, Ko GY, Yu ES, et al: Intestinal tuberculosis with abdominal complications: Radiologic and pathologic features. Abdom Imaging 24:32–38, 1999. 294. Ha HK, Lee HJ, Kim H, et al: Abdominal actinomycosis: CT findings in 10 patients. AJR Am J Roentgenol 161:791–794, 1993. 295. Ha HK, Lee EH, Lim CH, et al: Application of MRI for small intestinal diseases. J Magn Reson Imaging 8:375–383, 1998. 296. Ha HK, Lee HJ, Yang SK, et al: Intestinal Behcet syndrome: CT features of patients with and patients without complications. Radiology 209:449–454, 1998. 297. Ha HK, Park CH, Kim SK, et al: CT analysis of intestinal obstruction due to adhesions: Early detection of strangulation. J Comput Assist Tomogr 17:386–389, 1993. 298. Ha HK, Shin BS, Lee SI, et al: Usefulness of CT in patients with intestinal obstruction who have undergone abdominal surgery for malignancy. AJR Am J Roentgenol 171:1587–1593, 1998. 299. Ha HK, Shinn KS, Kim IC, et al: Gastroduodenal intussusception due to a prolapsed gastric adenoma. AJR Am J Roentgenol 159:432, 1992. 300. Hadithi M, Mallant M, Oudejans J, et al: 18F-FDG PET versus CT for the detection of enteropathy-associated T-cell lym-phoma in refractory celiac disease. J Nucl Med 47:1622–1627, 2006. 301. Haglund U, Bergqvist D: Intestinal ischemia—The basics. Langenbecks Arch Surg 384:233–238, 1999. 302. Haines JD, Jr: Ileal endometriosis. A difficult preoperative diagnosis. Postgrad Med 85:145–146, 149, 1989. 303. Halber MD, Daffner RH, Thompson WM: CT of the esophagus: I. Normal appearance. AJR Am J Roentgenol 133:1047–1050, 1979. 304. Halden WJ, Harnsberger HR, Mancuso AA: Computed tomography of esophageal varices after sclerotherapy. AJR Am J Roentgenol 140:1195– 1196, 1983. 305. Halpern M, Citron BP: Necrotizing angiitis associated with drug abuse. Am J Roentgenol Radium Ther Nucl Med 111:663–671, 1971. 306. Halvorsen RA, Jr, Thompson WM: Computed tomographic staging of gastrointestinal tract malignancies. Part I. Esophagus and stomach. Invest Radiol 22:2–16, 1987. 307. Halvorsen RA, Thompson WM: CT and MRI of the esophagus. In Levine MS, editor: Radiology of the esophagus, Philadelphia, 1989, WB Saunders, pp 291–309. 308. Hande KR, Fisher RI, DeVita VT, et al: Diffuse histiocytic lymphoma involving the gastrointestinal tract. Cancer 41:1984–1989, 1978. 309. Hara AK, Leighton JA, Heigh RI, et al: Crohn disease of the small bowel: Preliminary comparison among CT enterography, capsule endoscopy, small-bowel follow-through, and ileoscopy. Radiology 238:128–134, 2006. 310. Harris LA, DeCosse JJ, Dannenberg A: Abdominal actinomycosis: Evaluation by computed tomography. Am J Gastroenterol 84:198–200, 1989. 311. Harris NL, Jaffe ES, Stein H, et al: A revised European-American classification of lymphoid neoplasms: A proposal from the International Lymphoma Study Group. Blood 84:1361–1392, 1994. 312. Hassan HA, Raufman JP: Mesenteric venous thrombosis. South Med J 92:558–562, 1999. 313. Haubrich WS, Pounder RE, Fraser AG: Peptic ulcer desease: Diagnosis, medical management, and complications. In Haubrich WS, Schaffner F, Berk JE, editors: Gastroenterology, Philadelphia, 1994, WB Saunders, pp 749–789. 314. Heiberg E, Wolverson MK, Hurd RN, et al: CT recognition of   traumatic rupture of the diaphragm. AJR Am J Roentgenol 135:369–372, 1980.

CHAPTER 50  Gastrointestinal Tract 315. Heiken JP, Balfe DM, Roper CL: CT evaluation after esophagogastrectomy. AJR Am J Roentgenol 143:555–560, 1984. 316. Herbetko J, Brunton FJ: Enteroliths of small bowel diverticula. Clin Radiol 43:311–313, 1991. 317. Herrera L, Villarreal JR: Incidence of metastases from rectal adenocarcinoma in small lymph nodes detected by a clearing technique. Dis Colon Rectum 35:783–788, 1992. 318. Hersh MR, Choi J, Garrett C, et al: Imaging gastrointestinal stromal tumors. Cancer Control 12:111–115, 2005. 319. Hertzer NR, Hoerr SO: An interpretive review of lymphoma of the stomach. Surg Gynecol Obstet 143:113–124, 1976. 320. Higa E, Rosai J, Pizzimbono CA, et al: Mucosal hyperplasia, mucinous cystadenoma, and mucinous cystadenocarcinoma of the appendix.   A re-evaluation of appendiceal “mucocele.”. Cancer 32:1525–1541, 1973. 321. Hinnant KL, Rotterdam HZ, Bell ET, et al: Cytomegalovirus infection of the alimentary tract: A clinicopathological correlation. Am J Gastroenterol 81:944–950, 1986. 322. Hirose J, Takashima T, Suzuki M, et al: “Downhill” esophageal varices demonstrated by dynamic computed tomography. J Comput Assist Tomogr 8:1007–1009, 1984. 323. Hirota C, Iida M, Aoyagi K, et al: Posttraumatic intestinal stenosis: Clinical and radiographic features in four patients. Radiology 194:813–815, 1995. 324. Hirota S, Isozaki K, Moriyama Y, et al: Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279:577–580, 1998. 325. Ho KJ, Herrera GA, Jones JM, et al: Small cell carcinoma of the esophagus: Evidence for unified histogenesis. Hum Pathol 15:460–468, 1984. 326. Hoffmann M, Kletter K, Diemling M, et al: Positron emission tomography with fluorine-18-2-fluoro-2-deoxy-D-glucose (F18-FDG) does not visualize extranodal B-cell lymphoma of the mucosaassociated lymphoid tissue (MALT)-type. Ann Oncol 10:1185–1189, 1999. 327. Hoffmann M, Vogelsang H, Kletter K, et al: 18F-fluorodeoxy-glucose positron emission tomography (18F-FDG-PET) for assessment of enteropathy-type T cell lymphoma. Gut 52:347–351, 2003. 328. Holcomb GW, 3rd, Gheissari A, O’Neill JA, Jr, et al: Surgical management of alimentary tract duplications. Ann Surg 209:167–174, 1989. 329. Holscher AH, Dittler HJ, Siewert JR: Staging of squamous esophageal cancer: Accuracy and value. World J Surg 18:312–320, 1994. 330. Hong X, Choi H, Loyer EM, et al: Gastrointestinal stromal tumor: Role of CT in diagnosis and in response evaluation and surveillance after treatment with imatinib. Radiographics 26:481–495, 2006. 331. Hori S, Tsuda K, Murayama S, et al: CT of gastric carcinoma: Preliminary results with a new scanning technique. Radiographics 12:257–268, 1992. 332. Horton KM, Corl FM, Fishman EK: CT of nonneoplastic diseases of the small bowel: Spectrum of disease. J Comput Assist Tomogr 23:417–428, 1999. 333. Horton KM, Corl FM, Fishman EK: CT evaluation of the colon: Inflammatory disease. Radiographics 20:399–418, 2000. 334. Horton KM, Fishman EK: Volume-rendered 3D CT of the mesenteric vasculature: Normal anatomy, anatomic variants, and pathologic conditions. Radiographics 22:161–172, 2002. 335. Horton KM, Fishman EK: The current status of multidetector row CT and three-dimensional imaging of the small bowel. Radiol Clin North Am 41:199–212, 2003. 336. Horton KM, Fishman EK: Multidetector CT angiography in the diagnosis of mesenteric ischemia. Radiol Clin North Am 45:275–288, 2007. 337. Horton KM, Juluru K, Montogomery E, et al: Computed tomography imaging of gastrointestinal stromal tumors with pathology correlation. J Comput Assist Tomogr 28:811–817, 2004. 338. Hosomi N, Yoshioka H, Kuroda C, et al: Pneumatosis cystoides intestinalis: CT findings. Abdom Imaging 19:137–139, 1994.

1731

339. Hricak H, Thoeni RF, Margulis AR, et al: Extension of gastric lymphoma into the esophagus and duodenum. Radiology 135:309–312, 1980. 340. Hulnick DH, Megibow AJ, Balthazar EJ, et al: Perforated colorectal neoplasms: Correlation of clinical, contrast enema, and CT examinations. Radiology 164:611–615, 1987. 341. Hung GU, Shiau YC, Tsai SC, et al: Value of 18F-fluoro-2-deoxyglucose positron emission tomography in the evaluation of recurrent colorectal cancer. Anticancer Res 21:1375–1378, 2001. 342. Hunter TB, Bjelland JC: Gastrointestinal complications of leukemia and its treatment. AJR Am J Roentgenol 142:513–518, 1984. 343. Huprich J, Fletcher J, Alexander J, et al: Obscure GI bleeding: 64 channel multi-phase, multiplanar CT enterography has a role (abstr). In Radiological Society of North America Scientific Assembly and Annual Meeting Program, Oak Brook, Ill, 2005, Radiological Society of North America. 344. Hurlstone DP, Cross SS, Adam I, et al: A prospective clinicopathological and endoscopic evaluation of flat and depressed colorectal lesions in the United Kingdom. Am J Gastroenterol 98:2543–2549, 2003. 345. Hurwitz LM, Gertler SL: Colonoscopic diagnosis of ileocolic intussusception. Gastrointest Endosc 32:217–218, 1986. 346. Hussain SM, Outwater EK, Siegelman ES: Mucinous versus nonmucinous rectal carcinomas: Differentiation with MR imaging. Radiology 213:79–85, 1999. 347. Huszty G, Mogami K, Sawada T, et al: Preoperative evaluation of irreversible bowel ischemia in obturator hernia. Hepatogastroenterology 54:775–779, 2007. 348. Hwang HY, Choi BI, Han JK, et al: Calcified gastric carcinoma: CT findings. Gastrointest Radiol 17:311–315, 1992. 349. Iafrate F, Laghi A, Paolantonio P, et al: Preoperative staging of rectal cancer with MR imaging: Correlation with surgical and histopathologic findings. Radiographics 26:701–714, 2006. 350. Iannaccone R, Laghi A, Catalano C, et al: Detection of colorectal lesions: Lower-dose multi-detector row helical CT colonography compared with conventional colonoscopy. Radiology 229:775–781, 2003. 351. Iannaccone R, Laghi A, Catalano C, et al: Computed tomographic colonography without cathartic preparation for the detection of colorectal polyps. Gastroenterology 127:1300–1311, 2004. 352. Iannaccone R, Laghi A, Passariello R: Multislice CT angiography of mesenteric vessels. Abdom Imaging 29:146–152, 2004. 353. Iko BO, Teal JS, Siram SM, et al: Computed tomography of adult colonic intussusception: Clinical and experimental studies. AJR Am J Roentgenol 143:769–772, 1984. 354. Im JG, Kim JH, Han MC, et al: Computed tomography of esophagomediastinal fistula in tuberculous mediastinal lymphadenitis.   J Comput Assist Tomogr 14:89–92, 1990. 355. Imdahl A, Reinhardt MJ, Nitzsche EU, et al: Impact of 18F-FDGpositron emission tomography for decision making in colorectal cancer recurrences. Langenbecks Arch Surg 385:129–134, 2000. 356. Incesu L, Coskun A, Selcuk MB, et al: Acute appendicitis: MR imaging and sonographic correlation. AJR Am J Roentgenol 168:669–674, 1997. 357. Ishida H, Konno K, Hamashima Y, et al: Portal tumor thrombus due to gastrointestinal cancer. Abdom Imaging 24:585–590, 1999. 358. Ishikura H, Kirimoto K, Shamoto M, et al: Hepatoid adenocarcinomas of the stomach. An analysis of seven cases. Cancer 58:119–126, 1986. 359. Ito K, Kato T, Tadokoro M, et al: Recurrent rectal cancer and scar: Differentiation with PET and MR imaging. Radiology 182:549–552, 1992. 360. Iwanaga T, Koyama H, Furukawa H, et al: Mechanisms of late recurrence after radical surgery for gastric carcinoma. Am J Surg 135:637–640, 1978. 361. Iyer RB, Silverman PM, Tamm EP, et al: Diagnosis, staging, and follow-up of esophageal cancer. AJR Am J Roentgenol 181:785–793, 2003. 362. Jacobs JE, Birnbaum BA: CT of inflammatory disease of the colon. Semin Ultrasound CT MR 16:91–101, 1995. 363. Jacobs JM, Hill MC, Steinberg WM: Peptic ulcer disease: CT evaluation. Radiology 178:745–748, 1991.

1732

PART II  CT and MR Imaging of the Whole Body

364. Jaffe TA, Martin LC, Thomas J, et al: Small-bowel obstruction: Coronal reformations from isotropic voxels at 16-section multi-detector row CT. Radiology 238:135–142, 2006. 365. James S, Balfe DM, Lee JK, et al: Small-bowel disease: Categorization by CT examination. AJR Am J Roentgenol 148:863–868, 1987. 366. Jang KM, Lee KS, Lee SJ, et al: The spectrum of benign esophageal lesions: Imaging findings. Korean J Radiol 3:199–210, 2002. 367. Jang HJ, Lim HK, Lee SJ, et al: Acute diverticulitis of the cecum and ascending colon: Thin-section helical CT findings. AJR Am J Roentgenol 172:601–604, 1999. 368. Jaworski A, Fischer R, Lippmann M: Boerhaave’s syndrome. Computed tomographic findings and diagnostic considerations. Arch Intern Med 148:223–224, 1988. 369. Jee KN, Ha HK, Lee IJ, et al: Radiologic findings of abdominal polyarteritis nodosa. AJR Am J Roentgenol 174:1675–1679, 2000. 370. Jennette JC, Falk RJ: Small-vessel vasculitis. N Engl J Med 337:1512– 1523, 1997. 371. Jennette JC, Falk RJ, Andrassy K, et al: Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 37:187–192, 1994. 372. Jeong YK, Ha HK, Yoon CH, et al: Gastrointestinal involvement in Henoch-Schonlein syndrome: CT findings. AJR Am J Roentgenol 168:965–968, 1997. 373. Jerusalem G, Beguin Y, Najjar F, et al: Positron emission tomography (PET) with 18F-fluorodeoxyglucose (18F-FDG) for the staging of low-grade non-Hodgkin’s lymphoma (NHL). Ann Oncol 12:825–830, 2001. 374. Joffe N, Millan VG: Postemetic dissecting intramural hematoma of the esophagus. Radiology 95:379–380, 1970. 375. Johnson CD, Baker ME, Rice RP, et al: Diagnosis of acute colonic diverticulitis: Comparison of barium enema and CT. AJR Am J Roentgenol 148:541–546, 1987. 376. Jones B, Fishman EK, Hamilton SR, et al: Submucosal accumulation of fat in inflammatory bowel disease: CT/pathologic correlation. J Comput Assist Tomogr 10:759–763, 1986. 377. Jones B, Fishman EK, Kramer SS, et al: Computed tomography of gastrointestinal inflammation after bone marrow transplantation. AJR Am J Roentgenol 146:691–695, 1986. 378. Jones WG, 2nd, Ginsberg RJ: Esophageal perforation: A continuing challenge. Ann Thorac Surg 53:534–543, 1992. 379. Jones B, Kramer SS, Saral R, et al: Gastrointestinal inflammation after bone marrow transplantation: Graft-versus-host disease or opportunistic infection? AJR Am J Roentgenol 150:277–281, 1988. 380. Jones DR, Parker LA, Jr, Detterbeck FC, et al: Inadequacy of computed tomography in assessing patients with esophageal carcinoma after induction chemoradiotherapy. Cancer 85:1026–1032, 1999. 381. Kaibara N, Iitsuka Y, Kimura A, et al: Relationship between area of serosal invasion and prognosis in patients with gastric carcinoma. Cancer 60:136–139, 1987. 382. Kajitani T: The general rules for the gastric cancer study in surgery   and pathology. Part I. Clinical classification. Jpn J Surg 11:127–139, 1981. 383. Kakeji Y, Tsujitani S, Baba H, et al: Clinicopathologic features and prognostic significance of duodenal invasion in patients with distal gastric carcinoma. Cancer 68:380–384, 1991. 384. Kala Z, Valek V, Kysela P: Amyloidosis of the small intestine. Eur J Radiol 63:105–109, 2007. 385. Kalantari BN, Mortele KJ, Cantisani V, et al: CT features with pathologic correlation of acute gastrointestinal graft-versus-host disease after bone marrow transplantation in adults. AJR Am J Roentgenol 181:1621–1625, 2003. 386. Kalha I, Kaw M, Fukami N, et al: The accuracy of endoscopic ultrasound for restaging esophageal carcinoma after chemoradiation therapy. Cancer 101:940–947, 2004. 387. Kantorova I, Lipska L, Belohlavek O, et al: Routine (18)F-FDG PET preoperative staging of colorectal cancer: Comparison with conventional staging and its impact on treatment decision making.   J Nucl Med 44:1784–1788, 2003.

388. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al: Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 345:638–646, 2001. 389. Kasahara Y, Tanaka S, Nishino M, et al: Intestinal involvement in Behcet’s disease: Review of 136 surgical cases in the Japanese literature. Dis Colon Rectum 24:103–106, 1981. 390. Kato H, Miyazaki T, Nakajima M, et al: Value of positron emission tomography in the diagnosis of recurrent oesophageal carcinoma. Br J Surg 91:1004–1009, 2004. 391. Kato H, Miyazaki T, Nakajima M, et al: The incremental effect of positron emission tomography on diagnostic accuracy in the initial staging of esophageal carcinoma. Cancer 103:148–156, 2005. 392. Kawamoto S, Horton KM, Fishman EK: Pseudomembranous colitis: Can CT predict which patients will need surgical intervention?   J Comput Assist Tomogr 23:79–85, 1999. 393. Kawashima A, Fishman EK, Kuhlman JE, et al: CT of malignant melanoma: Patterns of small bowel and mesenteric involvement.   J Comput Assist Tomogr 15:570–574, 1991. 394. Kazerooni EA, Quint LE, Francis IR: Duodenal neoplasms: Predictive value of CT for determining malignancy and tumor resectability. AJR Am J Roentgenol 159:303–309, 1992. 395. Kelvin FM, Gramm HF, Gluck WL, et al: Radiologic manifestations of small-bowel toxicity due to floxuridine therapy. AJR Am J Roentgenol 146:39–43, 1986. 396. Kemerink GJ, Borstlap AC, Frantzen MJ, et al: Patient and occupational dosimetry in double contrast barium enema examinations. Br J Radiol 74:420–428, 2001. 397. Keogan MT, Lowe VJ, Baker ME, et al: Local recurrence of rectal cancer: Evaluation with F-18 fluorodeoxyglucose PET imaging. Abdom Imaging 22:332–337, 1997. 398. Kerekes ES: Gastric hemangioma: A case report. Radiology 82:468–469, 1964. 399. Ketcham AS, Hoye RC, Pilch YH, et al: Delayed intestinal obstruction following treatment for cancer. Cancer 25:406–410, 1970. 400. Kim JH, Beets GL, Kim MJ, et al: High-resolution MR imaging for nodal staging in rectal cancer: Are there any criteria in addition to the size? Eur J Radiol 52:78–83, 2004. 401. Kim AY, Bennett GL, Bashist B, et al: Small-bowel obstruction associated with sigmoid diverticulitis: CT evaluation in 16 patients. AJR Am J Roentgenol 170:1311–1313, 1998. 402. Kim KW, Berkmen YM, Auh YH, et al: Diagnosis of epiphrenic esophageal diverticulum by computed tomography. J Comput Tomogr 12:25–28, 1988. 403. Kim Y, Cho O, Song S, et al: Peritoneal lymphomatosis: CT findings. Abdom Imaging 23:87–90, 1998. 404. Kim JY, Ha HK, Byun JY, et al: Intestinal infarction secondary to mesenteric venous thrombosis: CT-pathologic correlation. J Comput Assist Tomogr 17:382–385, 1993. 405. Kim HJ, Ha HK, Cho KS, et al: CT features of primary colorectal signet-ring cell carcinoma. J Comput Assist Tomogr 25:225–230, 2001. 406. Kim JH, Ha HK, Kim JK, et al: Usefulness of known computed tomography and clinical criteria for diagnosing strangulation in small-bowel obstruction: Analysis of true and false interpretation groups in computed tomography. World J Surg 28:63–68, 2004. 407. Kim AY, Ha HK, Seo BK, et al: CT of patients with right-sided colon cancer and distal ileal thickening. AJR Am J Roentgenol 175:1439–1444, 2000. 408. Kim JH, Ha HK, Sohn MJ, et al: Usefulness of MR imaging for diseases of the small intestine: Comparison with CT. Korean J Radiol 1:43–50, 2000. 409. Kim JH, Ha HK, Sohn MJ, et al: CT findings of phytobezoar associated with small bowel obstruction. Eur Radiol 13:299–304, 2003. 410. Kim HS, Han HY, Choi JA, et al: Preoperative evaluation of gastric cancer: Value of spiral CT during gastric arteriography (CTGA). Abdom Imaging 26:123–130, 2001. 411. Kim TK, Han JK, Chung JW, et al: Intraperitoneal drop metastases from hepatocellular carcinoma: CT and angiographic findings.   J Comput Assist Tomogr 20:638–642, 1996.

CHAPTER 50  Gastrointestinal Tract 412. Kim SH, Han JK, Lee KH, et al: Abdominal amyloidosis: Spectrum of radiological findings. Clin Radiol 58:610–620, 2003. 413. Kim MJ, Huh YM, Park YN, et al: Colorectal mucinous carcinoma: Findings on MRI. J Comput Assist Tomogr 23:291–296, 1999. 414. Kim SK, Kang KW, Lee JS, et al: Assessment of lymph node metastases using 18F-FDG PET in patients with advanced gastric cancer. Eur J Nucl Med Mol Imaging 33:148–155, 2006. 415. Kim KM, Kang DW, Moon WS, et al: Gastrointestinal stromal   tumors in Koreans: Its incidence and the clinical, pathologic and immunohistochemical findings. J Korean Med Sci 20:977–984, 2005. 416. Kim SY, Kim MJ, Chung JJ, et al: Abdominal tuberculous lymphadenopathy: MR imaging findings. Abdom Imaging 25:627–632, 2000. 417. Kim AY, Kim HJ, Ha HK: Gastric cancer by multidetector row CT: Preoperative staging. Abdom Imaging 30:465–472, 2005. 418. Kim SY, Kim KW, Kim AY, et al: Bloodborne metastatic tumors to the gastrointestinal tract: CT findings with clinicopathologic correlation. AJR Am J Roentgenol 186:1618–1626, 2006. 419. Kim SH, Kim YJ, Lee JM, et al: Esophageal varices in patients with cirrhosis: Multidetector CT esophagography—Comparison with endoscopy. Radiology 242:759–768, 2007. 420. Kim HJ, Kim AY, Oh ST, et al: Gastric cancer staging at multi-detector row CT gastrography: Comparison of transverse and volumetric CT scanning. Radiology 236:879–885, 2005. 421. Kim SH, Lee JM, Han JK, et al: Effect of adjusted positioning on gastric distention and fluid distribution during CT gastrography. AJR Am J Roentgenol 185:1180–1184, 2005. 422. Kim SH, Lee JM, Han JK, et al: Three-dimensional MDCT imaging and CT esophagography for evaluation of esophageal tumors: Preliminary study. Eur Radiol 16:2418–2426, 2006. 423. Kim TJ, Lee KH, Kim YH, et al: Postoperative imaging of esophageal cancer: What chest radiologists need to know. Radiographics 27:409– 429, 2007. 424. Kim SH, Lim HK, Lee WJ, et al: Mucocele of the appendix: Ultrasonographic and CT findings. Abdom Imaging 23:292–296, 1998. 425. Kim HJ, Park MH: Obstructive uropathy due to interstitial cystitis in a patient with systemic lupus erythematosus. Clin Nephrol 45:205–208, 1996. 426. Kim YJ, Raman SS, Yu NC, et al: Esophageal varices in cirrhotic patients: Evaluation with liver CT. AJR Am J Roentgenol 188:139–144, 2007. 427. Kinkel K, Lu Y, Both M, et al: Detection of hepatic metastases from cancers of the gastrointestinal tract by using noninvasive imaging methods (US, CT, MR imaging, PET): A meta-analysis. Radiology 224:748–756, 2002. 428. Kirby TJ, Rice TW: The epidemiology of esophageal carcinoma. The changing face of a disease. Chest Surg Clin N Am 4:217–225, 1994. 429. Kirkpatrick ID, Greenberg HM: Gastrointestinal complications in the neutropenic patient: Characterization and differentiation with abdominal CT. Radiology 226:668–674, 2003. 430. Kirkpatrick ID, Kroeker MA, Greenberg HM: Biphasic CT with mesenteric CT angiography in the evaluation of acute mesenteric ischemia: Initial experience. Radiology 229:91–98, 2003. 431. Kirshy DM, Gordon DH, Atweh NA: Abdominal computed tomography in lupus mesenteric arteritis. Comput Med Imaging Graph 15:369–372, 1991. 432. Kistin MG, Kaplan MM, Harrington JT: Diffuse ischemic colitis associated with systemic lupus erythematosus—Response to subtotal colectomy. Gastroenterology 75:1147–1151, 1978. 433. Klein NC, Hargrove RL, Sleisenger MH, et al: Eosinophilic gastroenteritis. Medicine (Baltimore) 49:299–319, 1970. 434. Klein HM, Lensing R, Klosterhalfen B, et al: Diagnostic imaging of mesenteric infarction. Radiology 197:79–82, 1995. 435. Kneist W, Schreckenberger M, Bartenstein P, et al: Positron emission tomography for staging esophageal cancer: Does it lead to a different therapeutic approach? World J Surg 27:1105–1112, 2003.

1733

436. Kneist W, Schreckenberger M, Bartenstein P, et al: Prospective evaluation of positron emission tomography in the preoperative staging of esophageal carcinoma. Arch Surg 139:1043–1049, 2004. 437. Ko EY, Ha HK, Kim AY, et al: CT differentiation of mucinous and nonmucinous colorectal carcinoma. AJR Am J Roentgenol 188:785–791, 2007. 438. Ko GY, Ha HK, Lee HJ, et al: Usefulness of CT in patients with ischemic colitis proximal to colonic cancer. AJR Am J Roentgenol 168:951–956, 1997. 439. Ko S, Lee T, Ng S: Small bowel obstruction due to phytobezoar: CT diagnosis. Abdom Imaging 22:471–473, 1997. 440. Kodaira Y, Shibuya T, Matsumoto K, et al: Primary aortoduodenal fistula caused by duodenal tuberculosis without an abdominal aortic aneurysm: Report of a case. Surg Today 27:745–748, 1997. 441. Kodera Y, Yamamura Y, Shimizu Y, et al: The number of metastatic lymph nodes: A promising prognostic determinant for gastric carcinoma in the latest edition of the TNM classification. J Am Coll Surg 187:597–603, 1998. 442. Koelbel G, Schmiedl U, Majer MC, et al: Diagnosis of fistulae and sinus tracts in patients with Crohn disease: Value of MR imaging. AJR Am J Roentgenol 152:999–1003, 1989. 443. Koh DM, Burn PR, Mathews G, et al: Abdominal computed tomographic findings of Mycobacterium tuberculosis and Mycobacterium avium intracellulare infection in HIV seropositive patients. Can Assoc Radiol J 54:45–50, 2003. 444. Koh DM, Miao Y, Chinn RJ, et al: MR imaging evaluation of the activity of Crohn’s disease. AJR Am J Roentgenol 177:1325–1332, 2001. 445. Kole AC, Plukker JT, Nieweg OE, et al: Positron emission tomography for staging of oesophageal and gastroesophageal malignancy. Br J Cancer 78:521–527, 1998. 446. Kotanagi H, Fukuoka T, Shibata Y, et al: The size of regional lymph nodes does not correlate with the presence or absence of metastasis in lymph nodes in rectal cancer. J Surg Oncol 54:252–254, 1993. 447. Koyama T, Umeoka S, Saga T, et al: Evaluation of esophageal peristalsis in patients with esophageal tumors: Initial experience with cine MR imaging. Magn Reson Med Sci 4:109–114, 2005. 448. Kremer RM, Lepoff RB, Izant RJ, Jr: Duplication of the stomach.   J Pediatr Surg 5:360–364, 1970. 449. Krestin GP, Steinbrich W, Friedmann G: Recurrent rectal cancer: Diagnosis with MR imaging versus CT. Radiology 168:307–311, 1988. 450. Kuhle WG, Sheiman RG: Detection of active colonic hemorrhage with use of helical CT: Findings in a swine model. Radiology 228:743–752, 2003. 451. Kumar S, Sarr MG, Kamath PS: Mesenteric venous thrombosis. N Engl J Med 345:1683–1688, 2001. 452. Kumar R, Xiu Y, Potenta S, et al: 18F-FDG PET for evaluation of the treatment response in patients with gastrointestinal tract lymphomas.   J Nucl Med 45:1796–1803, 2004. 453. Kunin JR, Korobkin M, Ellis JH, et al: Duodenal injuries caused by blunt abdominal trauma: Value of CT in differentiating perforation from hematoma. AJR Am J Roentgenol 160:1221–1223, 1993. 454. Kwok H, Bissett IP, Hill GL: Preoperative staging of rectal cancer. Int J Colorectal Dis 15:9–20, 2000. 455. Laghi A, Carbone I, Catalano C, et al: Polyethylene glycol solution as an oral contrast agent for MR imaging of the small bowel. AJR Am J Roentgenol 177:1333–1334, 2001. 456. Laghi A, Ferri M, Catalano C, et al: Local staging of rectal cancer with MRI using a phased array body coil. Abdom Imaging 27:425–431, 2002. 457. Laghi A, Iannaccone R, Bria E, et al: Contrast-enhanced computed tomographic colonography in the follow-up of colorectal cancer patients: A feasibility study. Eur Radiol 13:883–889, 2003. 458. Laghi A, Paolantonio P, Catalano C, et al: MR imaging of the small bowel using polyethylene glycol solution as an oral contrast agent in adults and children with celiac disease: Preliminary observations. AJR Am J Roentgenol 180:191–194, 2003. 459. Laky D, Stoica T: Gastric liposarcoma. A case report. Pathol Res Pract 181:112–117, 1986.

1734

PART II  CT and MR Imaging of the Whole Body

460. Lane MJ, Katz DS, Ross BA, et al: Unenhanced helical CT for suspected acute appendicitis. AJR Am J Roentgenol 168:405–409, 1997. 461. Lassandro F, Romano S, Ragozzino A, et al: Role of helical CT in diagnosis of gallstone ileus and related conditions. AJR Am J Roentgenol 185:1159–1165, 2005. 462. Lauenstein TC, Ajaj W, Narin B, et al: MR imaging of apparent small-bowel perfusion for diagnosing mesenteric ischemia: Feasibility study. Radiology 234:569–575, 2005. 463. Laurent F, Raynaud M, Biset JM, et al: Diagnosis and categorization of small bowel neoplasms: Role of computed tomography. Gastrointest Radiol 16:115–119, 1991. 464. Lazarus DE, Slywotsky C, Bennett GL, et al: Frequency and relevance   of the “small-bowel feces” sign on CT in patients with small-bowel obstruction. AJR Am J Roentgenol 183:1361–1366, 2004. 465. Leder RA, Low VH: Tuberculosis of the abdomen. Radiol Clin North Am 33:691–705, 1995. 466. Lee DH: Three-dimensional imaging of the stomach by spiral CT.   J Comput Assist Tomogr 22:52–58, 1998. 467. Lee MJ, Breathnach E: CT and MRI findings in paraoesophageal omental herniation. Clin Radiol 42:207–209, 1990. 468. Lee RC, Chiang JH, Chou YH, et al: Intestinal schistosomiasis japonica: CT-pathologic correlation. Radiology 193:539–542, 1994. 469. Lee DH, Choi BI, Lee MG, et al: Exophytic adenocarcinoma of the stomach: CT findings. AJR Am J Roentgenol 163:77–80, 1994. 470. Lee GH, Cohen AJ: CT imaging of abdominal hernias. AJR Am J Roentgenol 161:1209–1213, 1993. 471. Lee SH, Ha HK, Byun JY, et al: Radiological features of leiomyomatous tumors of the colon and rectum. J Comput Assist Tomogr 24:407–412, 2000. 472. Lee IJ, Ha HK, Park CM, et al: Abdominopelvic actinomycosis involving the gastrointestinal tract: CT features. Radiology 220:76–80, 2001. 473. Lee SS, Ha HK, Park SH, et al: Usefulness of CT in differentiating transmural infarction from non-transmural ischemia of the small intestine in patients with acute mesenteric venous thrombosis.   J Comput Assist Tomogr 32:730–737, 2008. 474. Lee SS, Ha HK, Yang SK, et al: CT of prominent pericolic or perienteric vasculature in patients with Crohn’s disease: Correlation with clinical disease activity and findings on barium studies. AJR Am J Roentgenol 179:1029–1036, 2002. 475. Lee JH, Ha HK, Yoo E, et al: CT and sonographically guided biopsy in a patient with intestinal mucormycosis. AJR Am J Roentgenol 175:129– 131, 2000. 476. Lee DH, Ko YT: The role of 3D spiral CT in early gastric carcinoma.   J Comput Assist Tomogr 22:709–713, 1998. 477. Lee SS, Park SH, Choi EK, et al: Colorectal polyps on portal phase contrast-enhanced CT colonography: Lesion attenuation and distinction from tagged feces. AJR Am J Roentgenol 189:35–40, 2007. 478. Lee VS, Rofsky NM, Krinsky GA, et al: Single dose breath-hold gadolinium-enhanced three-dimensional MR angiography of the renal arteries. Radiology 211:69–78, 1999. 479. Lee DH, Seo TS, Ko YT: Spiral CT of the gastric carcinoma: Staging and enhancement pattern. Clin Imaging 25:32–37, 2001. 480. Lefere P, Gryspeerdt S: The alternative: Fecal tagging. In Lefere P, Gryspeerdt S, editors: Virtual colonoscopy: A practical guide, Berlin, 2006, Springer, pp 35–50. 481. Lefere P, Gryspeerdt S, Baekelandt M, et al: Laxative-free CT colonography. AJR Am J Roentgenol 183:945–948, 2004. 482. Lefere PA, Gryspeerdt SS, Dewyspelaere J, et al: Dietary fecal tagging as a cleansing method before CT colonography: Initial results—Polyp detection and patient acceptance. Radiology 224:393–403, 2002. 483. Lefere P, Gryspeerdt S, Marrannes J, et al: CT colonography after fecal tagging with a reduced cathartic cleansing and a reduced volume of barium. AJR Am J Roentgenol 184:1836–1842, 2005. 484. Leffall LD, Jr, Quander J, Syphax B: Strangulation intestinal obstruction. Arch Surg 91:592–596, 1965. 485. Lehr L, Rupp N, Siewert JR: Assessment of resectability of esophageal cancer by computed tomography and magnetic resonance imaging. Surgery 103:344–350, 1988.

486. Leidich RB, Rudolf LE: Small bowel perforation secondary to metastatic lung carcinoma. Ann Surg 193:67–69, 1981. 487. Lell MM, Anders K, Uder M, et al: New techniques in CT angiography. Radiographics 26:S45–S62, 2006. 488. Lerut T, Coosemans W, Decker G, et al: Diagnosis and therapy in advanced cancer of the esophagus and the gastroesophageal junction. Curr Opin Gastroenterol 22:437–441, 2006. 489. Lerut T, Flamen P, Ectors N, et al: Histopathologic validation of lymph node staging with FDG-PET scan in cancer of the esophagus and gastroesophageal junction: A prospective study based on primary surgery with extensive lymphadenectomy. Ann Surg 232:743–752, 2000. 490. Levine MS: Miscellaneous abnormalities. In Gore RM, Levine MS, Laufer I, editors: Textbook of gastrointestinal radiology, Philadelphia, 1994, WB Saunders, pp 512–530. 491. Levine MS: Other malignant tumors. In Gore RM, Levine MS, Laufer I, editors: Textbook of gastrointestinal radiology, Philadelphia, 1994, WB Saunders, pp 479–498. 492. Levine MS: Esophageal cancer. Radiologic diagnosis. Radiol Clin North Am 35:265–279, 1997. 493. Levine MS, Buck JL, Pantongrag-Brown L, et al: Fibrovascular polyps   of the esophagus: Clinical, radiographic, and pathologic findings in 16 patients. AJR Am J Roentgenol 166:781–787, 1996. 494. Levine MS, Buck JL, Pantongrag-Brown L, et al: Leiomyosarcoma of the esophagus: Radiographic findings in 10 patients. AJR Am J Roentgenol 167:27–32, 1996. 495. Levine MS, Buck JL, Pantongrag-Brown L, et al: Esophageal leiomyomatosis. Radiology 199:533–536, 1996. 496. Levine CD, Gonzales RN, Wachsberg RH, et al: CT findings of bowel and mesenteric injury. J Comput Assist Tomogr 21:974–979, 1997. 497. Levine MS, Halvorsen RA: Esophageal carcinoma. In Gore RM,   Levine MS, Laufer I, editors: Textbook of gastrointestinal radiology, Philadelphia, 1994, WB Saunders, pp 446–478. 498. Levine MS, Herman JB, Furth EE: Barrett’s esophagus and esophageal adenocarcinoma: The scope of the problem. Abdom Imaging 20:291– 298, 1995. 499. Levine MS, Pantongrag-Brown L, Buck JL, et al: Small-cell carcinoma of the esophagus: Radiographic findings. Radiology 199:703–705, 1996. 500. Levine CD, Patel UJ, Wachsberg RH, et al: CT in patients with blunt abdominal trauma: Clinical significance of intraperitoneal fluid detected on a scan with otherwise normal findings. AJR Am J Roentgenol 164:1381–1385, 1995. 501. Levison MA, Petersen SR, Sheldon GF, et al: Duodenal trauma: Experience of a trauma center. J Trauma 24:475–480, 1984. 502. Levy AD, Remotti HE, Thompson WM, et al: Gastrointestinal stromal tumors: Radiologic features with pathologic correlation. Radiographics 23:283–304, 456; quiz 532, 2003. 503. Lewin KJ, Ranchod M, Dorfman RF: Lymphomas of the gastrointestinal tract: A study of 117 cases presenting with gastrointestinal disease. Cancer 42:693–707, 1978. 504. Lewis JH: Gastrointestinal injury due to medicinal agents. Am J Gastroenterol 81:819–834, 1986. 505. Lewis FR, Holcroft JW, Boey J, et al: Appendicitis. A critical review of diagnosis and treatment in 1,000 cases. Arch Surg 110:677–684, 1975. 506. Leyman P, Ponette E, Marchal G, et al: Computed tomography of acute jejunogastric intussusception. J Comput Assist Tomogr 13:531–533, 1989. 507. Li KC, Whitney WS, McDonnell CH, et al: Chronic mesenteric ischemia: Evaluation with phase-contrast cine MR imaging. Radiology 190:175–179, 1994. 508. Lightdale CJ, Kulkarni KG: Role of endoscopic ultrasonography in the staging and follow-up of esophageal cancer. J Clin Oncol 23:4483–4489, 2005. 509. Lin S, Rockey DC: Obscure gastrointestinal bleeding. Gastroenterol Clin North Am 34:679–698, 2005. 510. Livingston DH, Lavery RF, Passannante MR, et al: Admission or observation is not necessary after a negative abdominal computed tomographic scan in patients with suspected blunt abdominal trauma:

CHAPTER 50  Gastrointestinal Tract

511. 512. 513.

514. 515.

516. 517. 518.

519.

520.

521.

522.

523. 524.

525. 526.

527.

528.

529.

530.

531. 532.

533.

Results of a prospective, multi-institutional trial. J Trauma 44:273–280, discussion 280-272, 1998. Livstone EM, Skinner DB: Tumors of the esophagus. In Berk JE, editor: Gastroenterology, Philadelphia, 1995, WB Sanders, pp 818–850. Lo CY, Chu KW: Acute diverticulitis of the right colon. Am J Surg 171:244–246, 1996. Localio SA, Eng K, Coppa GF: Malignant tumors of the rectum. In Localio SA, Eng K, Coppa GF, editors: Anorectal, presacral and sacral tumors: Anatomy, physiology, pathogenesis, and management, Philadelphia, 1987, WB Saunders, pp 137–151. Longo WE, Ballantyne GH, Gusberg RJ: Ischemic colitis: Patterns and prognosis. Dis Colon Rectum 35:726–730, 1992. Lopez-Mut JV, Cubells M, Campos S, et al: Jejunogastric intussusception: A rare complication of gastric surgery. Abdom Imaging 23:558–559, 1998. Lopez-Negrete L, Luyando L, Sala J, et al: Liposarcoma of the stomach. Abdom Imaging 22:373–375, 1997. Low RN, Wall SD, Jeffrey RB, Jr, et al: Aortoenteric fistula and perigraft infection: Evaluation with CT. Radiology 175:157–162, 1990. Lowe VJ, Booya F, Fletcher JG, et al: Comparison of positron emission tomography, computed tomography, and endoscopic ultrasound in the initial staging of patients with esophageal cancer. Mol Imaging Biol 7:422–430, 2005. Lucey BC, Mishkin D, Ozonoff A, et al: Comparison of CT enterography using 64 row detector MDCT with wireless video capsule endoscopy in the investigation of small bowel disease (abst). In Radiological Society of North America Scientific Assembly and Annual Meeting Program, Oak Brook, Ill, 2006, Radiological Society of North America, p 346. Luketich JD, Friedman DM, Weigel TL, et al: Evaluation of distant metastases in esophageal cancer: 100 consecutive positron emission tomography scans. Ann Thorac Surg 68:1133–1137, 1999. Lumbreras C, Fernandez I, Velosa J, et al: Infectious complications following pancreatic transplantation: Incidence, microbiological and clinical characteristics, and outcome. Clin Infect Dis 20:514–520, 1995. Lundquist A, Olsson R, Ekberg O: Clinical and radiologic evaluation reveals high prevalence of abnormalities in young adults with dysphagia. Dysphagia 13:202–207, 1998. Lupetin AR: Diffuse cavernous hemangioma of the rectum: Evaluation and MRI. Gastrointest Radiol 15:343–345, 1990. Lvoff N, Breiman RS, Coakley FV, et al: Distinguishing features of self-limiting adult small-bowel intussusception identified at CT. Radiology 227:68–72, 2003. Lyon DT, Schubert TT, Mantia AG, et al: Phycomycosis of the gastrointestinal tract. Am J Gastroenterol 72:379–394, 1979. Macafee CH, Greer HL: Intestinal endometriosis. A report of 29 cases and a survey of the literature. J Obstet Gynaecol Br Emp 67:539–555, 1960. Macari M, Balthazar EJ, Megibow AJ: The accordion sign at CT: A nonspecific finding in patients with colonic edema. Radiology 211:743–746, 1999. Macari M, Milano A, Lavelle M, et al: Comparison of time-efficient CT colonography with two- and three-dimensional colonic evaluation for detecting colorectal polyps. AJR Am J Roentgenol 174:1543–1549, 2000. Maccioni F, Viscido A, Broglia L, et al: Evaluation of Crohn disease activity with magnetic resonance imaging. Abdom Imaging 25:219–228, 2000. Maderal F, Hunter F, Fuselier G, et al: Gastric lipomas—An update of clinical presentation, diagnosis, and treatment. Am J Gastroenterol 79:964–967, 1984. Madrazo BL, Halpert RD, Sandler MA, et al: Computed tomographic findings in penetrating peptic ulcer. Radiology 153:751–754, 1984. Maglinte DD, Balthazar EJ, Kelvin FM, et al: The role of radiology in the diagnosis of small-bowel obstruction. AJR Am J Roentgenol 168:1171–1180, 1997. Maglinte DD, Chernish SM, DeWeese R, et al: Acquired jejunoileal diverticular disease: Subject review. Radiology 158:577–580, 1986.

1735

534. Maglinte DD, Gourtsoyiannis N, Rex D, et al: Classification of small bowel Crohn’s subtypes based on multimodality imaging. Radiol Clin North Am 41:285–303, 2003. 535. Maglinte DD, Herlinger H: Vascular disorders of the small intestine. In Herlinger H, Maglinte DD, editors: Clinical radiology of the small intestine, Philadelphia, 1989, WB Saunders, pp 453–478. 536. Maglinte DD, Herlinger H, Nolan DJ: Radiologic features of closed loop obstruction: Analysis of 25 confirmed cases. Radiology 179:383– 387, 1991. 537. Magnano G, Granata C, Barabino A, et al: Polyethylene glycol and contrast-enhanced MRI of Crohn’s disease in children: Preliminary experience. Pediatr Radiol 33:385–391, 2003. 538. Makanjuola D: Is it Crohn’s disease or intestinal tuberculosis? CT analysis. Eur J Radiol 28:55–61, 1998. 539. Mako EK, Mester AR, Tarjan Z, et al: Enteroclysis and spiral CT examination in diagnosis and evaluation of small bowel Crohn’s disease. Eur J Radiol 35:168–175, 2000. 540. Malhotra AK, Fabian TC, Katsis SB, et al: Blunt bowel and mesenteric injuries: The role of screening computed tomography. J Trauma 48:991–1000, 2000. 541. Manfredi S, Bouvier AM, Lepage C, et al: Incidence and patterns of recurrence after resection for cure of colonic cancer in a well defined population. Br J Surg 93:1115–1122, 2006. 542. Mani NB, Suri S, Gupta S, et al: Two-phase dynamic contrast-enhanced computed tomography with water-filling method for staging of gastric carcinoma. Clin Imaging 25:38–43, 2001. 543. Mannell A: Carcinoma of the esophagus. Curr Probl Surg 19:553–647, 1982. 544. Marks CG, Nolan DJ, Piris J, et al: Small bowel strictures after blunt abdominal trauma. Br J Surg 66:663–664, 1979. 545. Markus J, Morrissey B, deGara C, et al: MRI of recurrent rectosigmoid carcinoma. Abdom Imaging 22:338–342, 1997. 546. Marshall JB: Tuberculosis of the gastrointestinal tract and peritoneum. Am J Gastroenterol 88:989–999, 1993. 547. Marsman WA, Fockens P: State of the art lecture: EUS for esophageal tumors. Endoscopy 38(Suppl 1):S17–S21, 2006. 548. Martin DF, Hartley G: Gastric emphysema demonstrated by computed tomography. Br J Radiol 59:505–507, 1986. 549. Martin LW, Swisher SG, Hofstetter W, et al: Intrathoracic leaks following esophagectomy are no longer associated with increased mortality. Ann Surg 242:392–399, discussion 399-402, 2005. 550. Martinez-Frontanilla LA, Haase GM, Ernster JA, et al: Surgical complications in Henoch-Schonlein Purpura. J Pediatr Surg 19:434– 436, 1984. 551. Maruyama K, Gunven P, Okabayashi K, et al: Lymph node metastases of gastric cancer. General pattern in 1931 patients. Ann Surg 210:596– 602, 1989. 552. Marx MV, Balfe DM: Computed tomography of the esophagus. Semin Ultrasound CT MR 8:316–348, 1987. 553. Mason RM, Barnes CG: Behcet’s syndrome with arthritis. Ann Rheum Dis 28:95–103, 1969. 554. Matsuo M, Kanematsu M, Kim T, et al: Esophageal varices: Diagnosis with gadolinium-enhanced MR imaging of the liver for patients with chronic liver damage. AJR Am J Roentgenol 180:461–466, 2003. 555. Matsushita M, Oi H, Murakami T, et al: Extraserosal invasion in advanced gastric cancer: Evaluation with MR imaging. Radiology 192:87–91, 1994. 556. Mauro MA, Jaques PF, Swantkowski TM, et al: CT after uncomplicated esophageal sclerotherapy. AJR Am J Roentgenol 147:57–60, 1986. 557. Maxeiner SR, Jr, McDonald JR, Kirklin JW: Muscle biopsy in the diagnosis of periarteritis nodosa: An evaluation. Surg Clin North Am 32:1225–1233, 1952. 558. Mazzeo S, Caramella D, Gennai A, et al: Multidetector CT and virtual endoscopy in the evaluation of the esophagus. Abdom Imaging 29:2–8, 2004. 559. McAteer D, Wallis F, Couper G, et al: Evaluation of 18F-FDG positron emission tomography in gastric and oesophageal carcinoma. Br J Radiol 72:525–529, 1999.

1736

PART II  CT and MR Imaging of the Whole Body

560. McCartney J, Dobrow J, Hendrix TR: Boerhaave syndrome. Johns Hopkins Med J 144:28–33, 1979. 561. McCowin MJ, Federle MP: Computed tomography of pancreatic pseudocysts of the duodenum. AJR Am J Roentgenol 145:1003–1007, 1985. 562. McDaniel KP, Charnsangavej C, DuBrow RA, et al: Pathways of nodal metastasis in carcinomas of the cecum, ascending colon, and transverse colon: CT demonstration. AJR Am J Roentgenol 161:61–64, 1993. 563. McDermott VG, Low VH, Keogan MT, et al: Malignant melanoma metastatic to the gastrointestinal tract. AJR Am J Roentgenol 166:809– 813, 1996. 564. McFarland EG, Brink JA, Pilgram TK, et al: Spiral CT colonography: Reader agreement and diagnostic performance with two- and three-dimensional image-display techniques. Radiology 218:375–383, 2001. 565. McNeill PM, Wagman LD, Neifeld JP: Small bowel metastases from primary carcinoma of the lung. Cancer 59:1486–1489, 1987. 566. Megibow AJ: The gastrointestinal tract. In Haaga JR, Alfidi RJ, editors: Computed tomography of the whole body, St. Louis, 1988, CV Mosby, pp 1390–1428. 567. Megibow AJ, Babb JS, Hecht EM, et al: Evaluation of bowel distention and bowel wall appearance by using neutral oral contrast agent for multi-detector row CT. Radiology 238:87–95, 2006. 568. Megibow AJ, Balthazar EJ, Cho KC, et al: Bowel obstruction: Evaluation with CT. Radiology 180:313–318, 1991. 569. Megibow AJ, Balthazar EJ, Hulnick DH, et al: CT evaluation of gastrointestinal leiomyomas and leiomyosarcomas. AJR Am J Roentgenol 144:727–731, 1985. 570. Megibow AJ, Balthazar EJ, Naidich DP, et al: Computed tomography   of gastrointestinal lymphoma. AJR Am J Roentgenol 141:541–547, 1983. 571. Meisner RS, Spier BJ, Einarsson S, et al: Pilot study using PET/CT as a novel, noninvasive assessment of disease activity in inflammatory bowel disease. Inflamm Bowel Dis 13:993–1000, 2007. 572. Melody GF: Endometriosis causing obstruction of the ileum. Obstet Gynecol 8:468–472, 1956. 573. Meltzer CC, Luketich JD, Friedman D, et al: Whole-body FDG positron emission tomographic imaging for staging esophageal cancer comparison with computed tomography. Clin Nucl Med 25:882–887, 2000. 574. MERCURY Study Group: Diagnostic accuracy of preoperative magnetic resonance imaging in predicting curative resection of rectal cancer: Prospective observational study. BMJ 2006. doi: 10.1136/ bmj.38937.646400.55. (published 25 September 2006). 575. MERCURY Study Group: Extramural depth of tumor invasion at thin-section MR in patients with rectal cancer: Results of the MERCURY study. Radiology 243:132–139, 2007. 576. Merine D, Fishman EK, Jones B, et al: Enteroenteric intussusception: CT findings in nine patients. AJR Am J Roentgenol 148:1129–1132, 1987. 577. Merkel S, Mansmann U, Siassi M, et al: The prognostic inhomogeneity in pT3 rectal carcinomas. Int J Colorectal Dis 16:298–304, 2001. 578. Meyers MA: Paraduodenal hernias. Radiologic and arteriographic diagnosis. Radiology 95:29–37, 1970. 579. Meyers MA, McGuire PV: Spiral CT demonstration of hypervascularity in Crohn disease: “Vascular jejunization of the ileum” or the “comb sign.”. Abdom Imaging 20:327–332, 1995. 580. Meyers MA, McSweeney J: Secondary neoplasms of the bowel. Radiology 105:1–11, 1972. 581. Miettinen M, Lasota J: Gastrointestinal stromal tumors—Definition, clinical, histological, immunohistochemical, and molecular genetic features and differential diagnosis. Virchows Arch 438:1–12, 2001. 582. Miettinen M, Sarlomo-Rikala M, Sobin LH, et al: Esophageal stromal tumors: A clinicopathologic, immunohistochemical, and molecular genetic study of 17 cases and comparison with esophageal leiomyomas and leiomyosarcomas. Am J Surg Pathol 24:211–222, 2000. 583. Miettinen M, Virolainen M, Maarit Sarlomo R: Gastrointestinal stromal tumors—Value of CD34 antigen in their identification and separation

584.

585.

586.

587.

588. 589. 590.

591.

592.

593.

594. 595.

596.

597. 598.

599.

600.

601. 602.

603. 604. 605.

606.

607.

from true leiomyomas and schwannomas. Am J Surg Pathol 19:207–216, 1995. Miller SH: Anaphylactoid reaction after oral administration of diatrizoate meglumine and diatrizoate sodium solution. AJR Am J Roentgenol 168:959–961, 1997. Miller FH, Hwang CM: An initial experience: Using helical CT imaging to detect obscure gastrointestinal bleeding. Clin Imaging 28:245–251, 2004. Mills JA, Michel BA, Bloch DA, et al: The American College of Rheumatology 1990 criteria for the classification of Henoch-Schonlein purpura. Arthritis Rheum 33:1114–1121, 1990. Minami M, Kawauchi N, Itai Y, et al: Gastric tumors: Radiologicpathologic correlation and accuracy of T staging with dynamic CT. Radiology 185:173–178, 1992. Ming SC, Goldman H: Gastric polyps; a histogenetic classification and its relation to carcinoma. Cancer 18:721–726, 1965. Minocha A, Davis MS, Wright RA: Small bowel endometriosis masquerading as regional enteritis. Dig Dis Sci 39:1126–1133, 1994. Mirvis SE, Gens DR, Shanmuganathan K: Rupture of the bowel after blunt abdominal trauma: Diagnosis with CT. AJR Am J Roentgenol 159:1217–1221, 1992. Mirvis SE, Shanmuganathan K, Erb R: Diffuse small-bowel ischemia in hypotensive adults after blunt trauma (shock bowel): CT findings and clinical significance. AJR Am J Roentgenol 163:1375–1379, 1994. Moch A, Herlinger H, Kochman ML, et al: Enteroclysis in the evaluation of obscure gastrointestinal bleeding. AJR Am J Roentgenol 163:1381–1384, 1994. Mochiki E, Kuwano H, Katoh H, et al: Evaluation of 18F-2-deoxy-2fluoro-D-glucose positron emission tomography for gastric cancer. World J Surg 28:247–253, 2004. Monig SP, Baldus SE, Zirbes TK, et al: Lymph node size and metastatic infiltration in colon cancer. Ann Surg Oncol 6:579–581, 1999. Monig SP, Zirbes TK, Schroder W, et al: Staging of gastric cancer: Correlation of lymph node size and metastatic infiltration. AJR Am J Roentgenol 173:365–367, 1999. Morasch MD, Ebaugh JL, Chiou AC, et al: Mesenteric venous thrombosis: A changing clinical entity. J Vasc Surg 34:680–684, 2001. Mori M, Matsukuma A, Adachi Y, et al: Small cell carcinoma of the esophagus. Cancer 63:564–573, 1989. Morrin MM, Farrell RJ, Keogan MT, et al: CT colonography: Colonic distention improved by dual positioning but not intravenous glucagon. Eur Radiol 12:525–530, 2002. Moss AA, Schnyder P, Thoeni RF, et al: Esophageal carcinoma: Pretherapy staging by computed tomography. AJR Am J Roentgenol 136:1051–1056, 1981. Mukai K, Ishida Y, Okajima K, et al: Usefulness of preoperative FDG-PET for detection of gastric cancer. Gastric Cancer 9:192–196, 2006. Mulhall BP, Veerappan GR, Jackson JL: Meta-analysis: Computed tomographic colonography. Ann Intern Med 142:635–650, 2005. Munden RF, Macapinlac HA, Erasmus JJ: Esophageal cancer: The role of integrated CT-PET in initial staging and response assessment after preoperative therapy. J Thorac Imaging 21:137–145, 2006. Murata Y, Suzuki S, Hashimoto H: Endoscopic ultrasonography of the upper gastrointestinal tract. Surg Endosc 2:180–183, 1988. Murray JG, Evans SJ, Jeffrey PB, et al: Cytomegalovirus colitis in AIDS: CT features. AJR Am J Roentgenol 165:67–71, 1995. Nader PR, Margolin F: Hemangioma causing gastrointestinal bleeding. Case report and review of the literature. Am J Dis Child 111:215–222, 1966. Nagai E, Ueyama T, Yao T, et al: Hepatoid adenocarcinoma of the stomach. A clinicopathologic and immunohistochemical analysis. Cancer 72:1827–1835, 1993. Nagtegaal ID, Marijnen CA, Kranenbarg EK, et al: Circumferential margin involvement is still an important predictor of local recurrence in rectal carcinoma: Not one millimeter but two millimeters is the limit. Am J Surg Pathol 26:350–357, 2002.

CHAPTER 50  Gastrointestinal Tract 608. Nakamoto Y, Higashi T, Sakahara H, et al: Contribution of PET in the detection of liver metastases from pancreatic tumours. Clin Radiol 54:248–252, 1999. 609. Nalesnik MA, Makowka L, Starzl TE: The diagnosis and treatment of posttransplant lymphoproliferative disorders. Curr Probl Surg 25:367–472, 1988. 610. Naylor AR, Pollet JE: Eosinophilic colitis. Dis Colon Rectum 28:615–618, 1985. 611. Negaard A, Paulsen V, Sandvik L, et al: A prospective randomized comparison between two MRI studies of the small bowel in Crohn’s disease, the oral contrast method and MR enteroclysis. Eur Radiol 17:2294–2301, 2007. 612. Neurath MF, Vehling D, Schunk K, et al: Noninvasive assessment of Crohn’s disease activity: A comparison of 18F-fluorodeoxyglucose positron emission tomography, hydromagnetic resonance imaging, and granulocyte scintigraphy with labeled antibodies. Am J Gastroenterol 97:1978–1985, 2002. 613. New Zealand Medicines and Medical Devices Safety Authority: Medicine data sheets. Available at . 614. Nghiem HV, Jeffrey RB, Jr, Mindelzun RE: CT of blunt trauma to the bowel and mesentery. AJR Am J Roentgenol 160:53–58, 1993. 615. Nilsson B, Bumming P, Meis-Kindblom JM, et al: Gastrointestinal stromal tumors: The incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate era—A population-based study in western Sweden. Cancer 103:821–829, 2005. 616. Nishi M, Omori Y, Miwa K: Japanese Research Society for Gastric Cancer (JRSGC) [English]. In Japanese classification of gastric carcinoma. Tokyo, Japan, 1995, pp 6-15. 617. Nishida T, Hirota S: Biological and clinical review of stromal tumors in the gastrointestinal tract. Histol Histopathol 15:1293–1301, 2000. 618. Nishimaki T, Suzuki T, Fukuda T, et al: Primary small cell carcinoma of the esophagus with ectopic gastrin production. Report of a case and review of the literature. Dig Dis Sci 38:767–771, 1993. 619. Nitsch B, Ho CS, Cullen J: Barium study of small bowel endometriosis. Gastrointest Radiol 13:361–363, 1988. 620. Noh HM, Fishman EK, Forastiere AA, et al: CT of the esophagus: Spectrum of disease with emphasis on esophageal carcinoma. Radiographics 15:1113–1134, 1995. 621. Norton GA, Postlethwait RW, Thompson WM: Esophageal carcinoma: A survey of populations at risk. South Med J 73:25–27, 1980. 622. Oakes DD: Reconsidering the diagnosis and treatment of patients following ingestion of liquid lye. J Clin Gastroenterol 21:85–86, 1995. 623. O’Connor KF, Bagg MN, Croley MR, et al: Pelvic actinomycosis associated with intrauterine devices. Radiology 170:559–560, 1989. 624. Ogorek CP, Fisher RS: Differentiation between Crohn’s disease and ulcerative colitis. Med Clin North Am 78:1249–1258, 1994. 625. Ogunbiyi OA, Flanagan FL, Dehdashti F, et al: Detection of recurrent and metastatic colorectal cancer: Comparison of positron emission tomography and computed tomography. Ann Surg Oncol 4:613–620, 1997. 626. Oguzkurt L, Karabulut N, Cakmakci E, et al: Primary non-Hodgkin’s lymphoma of the esophagus. Abdom Imaging 22:8–10, 1997. 627. Okamoto K, Sawai K, Minato H, et al: Number and anatomical extent of lymph node metastases in gastric cancer: Analysis using intra-lymph node injection of activated carbon particles (CH40). Jpn J Clin Oncol 29:74–77, 1999. 628. Olmsted WW, Lichtenstein JE, Hyams VJ: Polypoid epithelial malignancies of the esophagus. AJR Am J Roentgenol 140:921–925, 1983. 629. Ongolo-Zogo P, Borson O, Garcia P, et al: Acute gastroduodenal peptic ulcer perforation: Contrast-enhanced and thin-section spiral CT findings in 10 patients. Abdom Imaging 24:329–332, 1999. 630. Orchard JL: Petorak V: Abnormal abdominal CT findings in a patient with giardiasis. Resolution after treatment. Dig Dis Sci 40:346–348, 1995. 631. Orloff MJ, Charters AC: Injuries of the small bowel and mesentery and retroperitoneal hematoma. Surg Clin North Am 52:729–734, 1972.

1737

632. Ormson MJ, Stephens DH, Carlson HC: CT recognition of intestinal lipomatosis. AJR Am J Roentgenol 144:313–314, 1985. 633. Orringer MB: The esophagus: Historical aspects and anatomy. In Sabiston DC, editor: Textbook of surgery, Philadelphia, 1986, WB Saunders, pp 697–701. 634. Osteen RT, Guyton S, Steele G, Jr, et al: Malignant intestinal obstruction. Surgery 87:611–615, 1980. 635. Ott DJ: Motility disorders of the esophagus. Radiol Clin North Am 32:1117–1134, 1994. 636. Ott K, Fink U, Becker K, et al: Prediction of response to preoperative chemotherapy in gastric carcinoma by metabolic imaging: Results of a prospective trial. J Clin Oncol 21:4604–4610, 2003. 637. Ott K, Weber W, Siewert JR: The importance of PET in the diagnosis and response evaluation of esophageal cancer. Dis Esophagus 19:433– 442, 2006. 638. Ozmen MM, Zulfikaroglu B, Ozalp N, et al: Staging laparoscopy for gastric cancer. Surg Laparosc Endosc Percutan Tech 13:241–244, 2003. 639. Padidar AM, Jeffrey RB, Jr, Mindelzun RE, et al: Differentiating sigmoid diverticulitis from carcinoma on CT scans: Mesenteric inflammation suggests diverticulitis. AJR Am J Roentgenol 163:81–83, 1994. 640. Palchick BA, Alpert MA, Holmes RA, et al: Esophageal hemangioma: Diagnosis with computed tomography and radionuclide angiography. South Med J 76:1582–1584, 1983. 641. Pandarinath GS, Levine SM, Sorokin JJ, et al: Selective massive amyloidosis of the small intestine mimicking multiple tumors. Radiology 129:609–610, 1978. 642. Panebianco V, Grazhdani H, Iafrate F, et al: 3D CT protocol in the assessment of the esophageal neoplastic lesions: Can it improve TNM staging? Eur Radiol 16:414–421, 2006. 643. Panebianco V, Habib FI, Tomei E, et al: Initial experience with magnetic resonance fluoroscopy in the evaluation of oesophageal motility disorders. Comparison with manometry and barium fluoroscopy. Eur Radiol 16:1926–1933, 2006. 644. Pantongrag-Brown L, Buetow PC, Carr NJ, et al: Calcification and fibrosis in mesenteric carcinoid tumor: CT findings and pathologic correlation. AJR Am J Roentgenol 164:387–391, 1995. 645. Pantongrag-Brown L, Krebs TL, Daly BD, et al: Frequency of abdominal CT findings in AIDS patients with m. avium complex bacteraemia. Clin Radiol 53:816–819, 1998. 646. Papachristou DN, Fortner JG: Local recurrence of gastric adenocarcinomas after gastrectomy. J Surg Oncol 18:47–53, 1981. 647. Papachristou DN, Karas M, Fortner JG: Anastomotic recurrence in the oesophagus complicating gastrectomy for adenocarcinoma of the stomach. Br J Surg 66:609–612, 1979. 648. Papadatos D, Taourel P, Bret PM: CT of leiomyomatosis peritonealis disseminata mimicking peritoneal carcinomatosis. AJR Am J Roentgenol 167:475–476, 1996. 649. Park SH, Choi EK, Lee SS, et al: Polyp measurement reliability, accuracy, and discrepancy: Optical colonoscopy versus CT colonography with pig colonic specimens. Radiology 244:157–164, 2007. 650. Park SH, Ha HK, Kim MJ, et al: False-negative results at multi-detector row CT colonography: Multivariate analysis of causes for missed lesions. Radiology 235:495–502, 2005. 651. Park SH, Ha HK, Kim AY, et al: Flat polyps of the colon: Detection with 16-MDCT colonography—Preliminary results. AJR Am J Roentgenol 186:1611–1617, 2006. 652. Park SB, Ha HK, Kim AY, et al: The diagnostic role of abdominal CT imaging findings in adults intussusception: Focused on the vascular compromise. Eur J Radiol 62:406–415, 2007. 653. Park MS, Kim KW, Yu JS, et al: Radiologic findings of gastrointestinal tract involvement in hepatocellular carcinoma. J Comput Assist Tomogr 26:95–101, 2002. 654. Park SH, Lee SS, Choi EK, et al: Flat colorectal neoplasms: Definition, importance, and visualization on CT colonography. AJR Am J Roentgenol 188:953–959, 2007. 655. Park SH, Yee J, Kim SH, et al: Fundamental elements for successful performance of CT colonography (virtual colonoscopy). Korean J Radiol 8:264–275, 2007.

1738

PART II  CT and MR Imaging of the Whole Body

656. Passas V, Karavias D, Grilias D, et al: Computed tomography of left paraduodenal hernia. J Comput Assist Tomogr 10:542–543, 1986. 657. Patel MP, De I: Segmental tuberculosis of the colon with entero-colic fistula. Br J Radiol 45:150–152, 1972. 658. Patton AS, Lawson DW, Shannon JM, et al: Reevaluation of the Boerhaave syndrome. A review of fourteen cases. Am J Surg 137:560– 565, 1979. 659. Paulsen SR, Huprich JE, Hara AK: CT enterography: Noninvasive evaluation of Crohn’s disease and obscure gastrointestinal bleed. Radiol Clin North Am 45:303–315, 2007. 660. Pauwels EK, Sturm EJ, Bombardieri E, et al: Positron-emission tomography with [18F]fluorodeoxyglucose. Part I. Biochemical uptake mechanism and its implication for clinical studies. J Cancer Res Clin Oncol 126:549–559, 2000. 661. Pavlakis GM, Sakorafas GH, Anagnostopoulos GK: Intestinal metastases from renal cell carcinoma: A rare cause of intestinal obstruction and bleeding. Mt Sinai J Med 71:127–130, 2004. 662. Pearlberg JL, Sandler MA, Madrazo BL: Computed tomographic features of esophageal intramural pseudodiverticulosis. Radiology 147:189–190, 1983. 663. Peitzman AB, Makaroun MS, Slasky BS, et al: Prospective study of computed tomography in initial management of blunt abdominal trauma. J Trauma 26:585–592, 1986. 664. Pellicano A, Watier A, Gentile J: Spontaneous double-barrelled esophagus. Report of two cases and review of the literature. J Clin Gastroenterol 9:149–154, 1987. 665. Pema PJ, Bennett WF, Bova JG, et al: CT vs MRI in diagnosis of recurrent rectosigmoid carcinoma. J Comput Assist Tomogr 18:256–261, 1994. 666. Pennazio M, Santucci R, Rondonotti E, et al: Outcome of patients with obscure gastrointestinal bleeding after capsule endoscopy: Report of 100 consecutive cases. Gastroenterology 126:643–653, 2004. 667. Perez C, Andreu J, Llauger J, et al: Hemangioma of the rectum: CT appearance. Gastrointest Radiol 12:347–349, 1987. 668. Perlman JA, Hoover HC, Safer PK: Femoral hernia with strangulated Meckel’s diverticulum (Littre’s hernia). Am J Surg 139:286–289, 1980. 669. Petrillo R, Balzarini L, Bidoli P, et al: Esophageal squamous cell carcinoma: MRI evaluation of mediastinum. Gastrointest Radiol 15:275–278, 1990. 670. Petrovic B, Nikolaidis P, Hammond NA, et al: Identification of adhesions on CT in small-bowel obstruction. Emerg Radiol 12:88–93, 2006. 671. Pezzulli FA, Aronson D, Goldberg N: Computed tomography of mediastinal hematoma secondary to unusual esophageal laceration: A Boerhaave variant. J Comput Assist Tomogr 13:129–131, 1989. 672. Pfister DG, Benson AB, III, Somerfield MR: Surveillance strategies after curative treatment of colorectal cancer. N Engl J Med 350:2375–2382, 2004. 673. Philpotts LE, Heiken JP, Westcott MA, et al: Colitis: Use of CT findings in differential diagnosis. Radiology 190:445–449, 1994. 674. Phongkitkarun S, Varavithya V, Kazama T, et al: Lymphomatous involvement of gastrointestinal tract: Evaluation by positron emission tomography with (18)F-fluorodeoxyglucose. World J Gastroenterol 11:7284–7289, 2005. 675. Pickhardt PJ: Three-dimensional endoluminal CT colonography (virtual colonoscopy): Comparison of three commercially available systems. AJR Am J Roentgenol 181:1599–1606, 2003. 676. Pickhardt PJ: Differential diagnosis of polypoid lesions seen at CT colonography (virtual colonoscopy). Radiographics 24:1535–1556, 2004. 677. Pickhardt PJ: CT colonography without catharsis: The ultimate study or useful additional option? Gastroenterology 128:521–523, 2005. 678. Pickhardt PJ: Incidence of colonic perforation at CT colonography: Review of existing data and implications for screening of asymptomatic adults. Radiology 239:313–316, 2006. 679. Pickhardt PJ, Choi JR, Hwang I, et al: Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 349:2191–2200, 2003.

680. Pickhardt PJ, Siegel MJ: Abdominal manifestations of posttransplantation lymphoproliferative disorder. AJR Am J Roentgenol 171:1007–1013, 1998. 681. Pickhardt PJ, Siegel MJ: Posttransplantation lymphoproliferative disorder of the abdomen: CT evaluation in 51 patients. Radiology 213:73–78, 1999. 682. Picus D, Balfe DM, Koehler RE, et al: Computed tomography in the staging of esophageal carcinoma. Radiology 146:433–438, 1983. 683. Picus D, Glazer HS, Levitt RG, et al: Computed tomography of abdominal carcinoid tumors. AJR Am J Roentgenol 143:581–584, 1984. 684. Pilleul F, Penigaud M, Milot L, et al: Possible small-bowel neoplasms: Contrast-enhanced and water-enhanced multidetector CT enteroclysis. Radiology 241:796–801, 2006. 685. Polkowski M, Palucki J, Wronska E, et al: Endosonography versus helical computed tomography for locoregional staging of gastric cancer. Endoscopy 36:617–623, 2004. 686. Poorman JC, Katon RM: Small bowel involvement by Mycobacterium avium complex in a patient with AIDS: Endoscopic, histologic, and radiographic similarities to Whipple’s disease. Gastrointest Endosc 40:753–759, 1994. 687. Potet F, Duclert N: [Cystic dystrophy on aberrant pancreas of the duodenal wall]. Arch Fr Mal App Dig 59:223–238, 1970. 688. Pozniak MA, Babel SG, Trump DL: Complications of hepatic arterial infusion chemotherapy. Radiographics 11:67–79, 1991. 689. Pradel JA, Adell JF, Taourel P, et al: Acute colonic diverticulitis: Prospective comparative evaluation with US and CT. Radiology 205:503–512, 1997. 690. Prassopoulos P, Papanikolaou N, Grammatikakis J, et al: MR enteroclysis imaging of Crohn disease. Radiographics 21(Spec No):S161–S172, 2001. 691. Prince MR: Body MR angiography with gadolinium contrast agents. Magn Reson Imaging Clin N Am 4:11–24, 1996. 692. Prince MR, Narasimham DL, Stanley JC, et al: Breath-hold gadoliniumenhanced MR angiography of the abdominal aorta and its major branches. Radiology 197:785–792, 1995. 693. Procacci C, Graziani R, Zamboni G, et al: Cystic dystrophy of   the duodenal wall: Radiologic findings. Radiology 205:741–747, 1997. 694. Pruksapong C, Donovan RJ, Pinit A, et al: Gastric duplication. J Pediatr Surg 14:83–85, 1979. 695. Puri AS, Aggarwal R, Gupta RK, et al: Intestinal lymphangiectasia: Evaluation by CT and scintigraphy. Gastrointest Radiol 17:119–121, 1992. 696. Quint LE, Glazer GM, Orringer MB: Esophageal imaging by MR and CT: Study of normal anatomy and neoplasms. Radiology 156:727–731, 1985. 697. Quint LE, Glazer GM, Orringer MB, et al: Esophageal carcinoma: CT findings. Radiology 155:171–175, 1985. 698. Quirke P, Dixon MF: The prediction of local recurrence in rectal adenocarcinoma by histopathological examination. Int J Colorectal Dis 3:127–131, 1988. 699. Quirke P, Durdey P, Dixon MF, et al: Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumour spread and surgical excision. Lancet 2:996–999, 1986. 700. Qvarfordt PG, Reilly LM, Mark AS, et al: Computerized tomographic assessment of graft incorporation after aortic reconstruction. Am J Surg 150:227–231, 1985. 701. Rabe FE, Becker GJ, Besozzi MJ, et al: Efficacy study of the small-bowel examination. Radiology 140:47–50, 1981. 702. Rabushka LS, Fishman EK, Kuhlman JE: CT evaluation of achalasia.   J Comput Assist Tomogr 15:434–439, 1991. 703. Rabushka LS, Fishman EK, Kuhlman JE, et al: Diffuse esophageal leiomyomatosis in a patient with Alport syndrome: CT demonstration. Radiology 179:176–178, 1991. 704. Radin DR: Primary esophageal lymphoma in AIDS. Abdom Imaging 18:223–224, 1993.

CHAPTER 50  Gastrointestinal Tract 705. Radin R: HIV infection: Analysis in 259 consecutive patients with abnormal abdominal CT findings. Radiology 197:712–722, 1995. 706. Rainer WG, Brus R: Leiomyosarcoma of the esophagus; review of the literature and report of 3 cases. Surgery 58:343–350, 1965. 707. Ranchod M, Kempson RL: Smooth muscle tumors of the gastrointestinal tract and retroperitoneum: A pathologic analysis of 100 cases. Cancer 39:255–262, 1977. 708. Randhawa PS, Yousem SA, Paradis IL, et al: The clinical spectrum, pathology, and clonal analysis of Epstein-Barr virus-associated lymphoproliferative disorders in heart-lung transplant recipients. Am J Clin Pathol 92:177–185, 1989. 709. Rao PM, Rhea JT: Colonic diverticulitis: Evaluation of the arrowhead sign and the inflamed diverticulum for CT diagnosis. Radiology 209:775–779, 1998. 710. Rao PM, Rhea JT, Novelline RA: Distal appendicitis: CT appearance and diagnosis. Radiology 204:709–712, 1997. 711. Rao PM, Rhea JT, Novelline RA, et al: Helical CT technique for the diagnosis of appendicitis: Prospective evaluation of a focused appendix CT examination. Radiology 202:139–144, 1997. 712. Rao PM, Rhea JT, Novelline RA, et al: Helical CT combined with contrast material administered only through the colon for imaging of suspected appendicitis. AJR Am J Roentgenol 169:1275–1280, 1997. 713. Rao PM, Rhea JT, Novelline RA, et al: Helical CT with only colonic contrast material for diagnosing diverticulitis: Prospective evaluation of 150 patients. AJR Am J Roentgenol 170:1445–1449, 1998. 714. Rao PM, Wittenberg J, Lawrason JN: Primary epiploic appendagitis: Evolutionary changes in CT appearance. Radiology 204:713–717, 1997. 715. Rao PM, Wittenberg J, McDowell RK, et al: Appendicitis: Use of arrowhead sign for diagnosis at CT. Radiology 202:363–366, 1997. 716. Raptopoulos V, Schwartz RK, McNicholas MM, et al: Multiplanar helical CT enterography in patients with Crohn’s disease. AJR Am J Roentgenol 169:1545–1550, 1997. 717. Rasanen JV, Sihvo EI, Knuuti MJ, et al: Prospective analysis of accuracy of positron emission tomography, computed tomography, and endoscopic ultrasonography in staging of adenocarcinoma of the esophagus and the esophagogastric junction. Ann Surg Oncol 10:954– 960, 2003. 718. Reinig JW, Stanley JH, Schabel SI: CT evaluation of thickened esophageal walls. AJR Am J Roentgenol 140:931–934, 1983. 719. Reinus JF, Brandt LJ, Boley SJ: Ischemic diseases of the bowel. Gastroenterol Clin North Am 19:319–343, 1990. 720. Rembacken BJ, Fujii T, Cairns A, et al: Flat and depressed colonic neoplasms: A prospective study of 1000 colonoscopies in the UK. Lancet 355:1211–1214, 2000. 721. Remigio PA, Klaum A: Extramedullary plasmacytoma of stomach. Cancer 27:562–568, 1971. 722. Renehan AG, Egger M, Saunders MP, et al: Impact on survival of intensive follow up after curative resection for colorectal cancer: Systematic review and meta-analysis of randomised trials. BMJ 324:813–815, 2002. 723. Rettenbacher T, Hollerweger A, Macheiner P, et al: Abdominal wall hernias: Cross-sectional imaging signs of incarceration determined with sonography. AJR Am J Roentgenol 177:1061–1066, 2001. 724. Rex DK, Cummings OW, Shaw M, et al: Screening for Barrett’s esophagus in colonoscopy patients with and without heartburn. Gastroenterology 125:1670–1677, 2003. 725. Rex DK, Lappas JC, Maglinte DD, et al: Enteroclysis in the evaluation of suspected small intestinal bleeding. Gastroenterology 97:58–60, 1989. 726. Rex DK, Rahmani EY, Haseman JH, et al: Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology 112:17–23, 1997. 727. Reynolds JV, Galloway JM: Primary aortoduodenal fistula: A case report and review of the literature. Ir J Med Sci 160:381–384, 1991. 728. Reznek RH, White FE, Young JW, et al: The appearances on computed tomography after abdominoperineal resection for carcinoma of the rectum: A comparison between the normal appearances and those of recurrence. Br J Radiol 56:237–240, 1983.

1739

729. Rha SE, Ha HK, Kim AY, et al: Peritoneal leiomyosarcomatosis originating from gastrointestinal leiomyosarcomas: CT features. Radiology 227:385–390, 2003. 730. Rhee RY, Gloviczki P, Mendonca CT, et al: Mesenteric venous thrombosis: still a lethal disease in the 1990s. J Vasc Surg 20:688–697, 1994. 731. Rice TW: Clinical staging of esophageal carcinoma. CT, EUS, and PET. Chest Surg Clin N Am 10:471–485, 2000. 732. Riddell AM, Davies DC, Allum WH, et al: High-resolution MRI in evaluation of the surgical anatomy of the esophagus and posterior mediastinum. AJR Am J Roentgenol 188:W37–W43, 2007. 733. Riddell AM, Hillier J, Brown G, et al: Potential of surface-coil MRI for staging of esophageal cancer. AJR Am J Roentgenol 187:1280–1287, 2006. 734. Riddell AM, Richardson C, Scurr E, et al: The development and optimization of high spatial resolution MRI for imaging the oesophagus using an external surface coil. Br J Radiol 79:873–879, 2006. 735. Rijke AM, Falke TH, de Vries RR: Computed tomography in Whipple disease. J Comput Assist Tomogr 7:1101–1102, 1983. 736. Rioux M, Langis P: Primary epiploic appendagitis: Clinical, US, and CT findings in 14 cases. Radiology 191:523–526, 1994. 737. Rioux M, Mailloux C: Crescent-shaped necrosis: A new imaging sign suggestive of stromal tumor of the small bowel. Abdom Imaging 22:376–380, 1997. 738. Rizzo MJ, Federle MP, Griffiths BG: Bowel and mesenteric injury following blunt abdominal trauma: Evaluation with CT. Radiology 173:143–148, 1989. 739. Rockey DC: Gastrointestinal bleeding. In Feldman M, Friedman LS, Sleisenger MH, editors: Gastrointestinal and liver disease, ed 7, Philadelphia, 2002, Saunders, pp 211–243. 740. Rockey DC, Halvorsen RA, Jr, Higgins JL, et al: Prospective evaluation of patients with bowel wall thickening. Am J Gastroenterol 90:99–103, 1995. 741. Rockey DC, Paulson E, Niedzwiecki D, et al: Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: Prospective comparison. Lancet 365:305–311, 2005. 742. Rockoff SD, Aaron BL, Black C, et al: Diagnosis of paraesophageal omental hiatal hernia by magnetic resonance imaging. Chest 103:285– 287, 1993. 743. Roder JD, Bottcher K, Busch R, et al: Classification of regional lymph node metastasis from gastric carcinoma. German Gastric Cancer Study Group. Cancer 82:621–631, 1998. 744. Rodriguez M, Rehn S, Ahlstrom H, et al: Predicting malignancy grade with PET in non-Hodgkin’s lymphoma. J Nucl Med 36:1790–1796, 1995. 745. Rohren EM, Turkington TG, Coleman RE: Clinical applications of PET in oncology. Radiology 231:305–332, 2004. 746. Ros PR, Buetow PC, Pantograg-Brown L, et al: Pseudomembranous colitis. Radiology 198:1–9, 1996. 747. Rosai J: Gastrointestinal tract. In Rosai J, editor: Ackerman’s surgical pathology, ed 8, St Louis, 1995, Mosby-Year Book, pp 616–710. 748. Rosch T: Endosonographic staging of esophageal cancer: A review of literature results. Gastrointest Endosc Clin N Am 5:537–547, 1995. 749. Rosenberg SA, Diamond HD, Jaslowitz B, et al: Lymphosarcoma: A review of 1269 cases. Medicine (Baltimore) 40:31–84, 1961. 750. Rosenberg HK, Serota FT, Koch P, et al: Radiographic features of gastrointestinal graft-vs.-host disease. Radiology 138:371–374, 1981. 751. Rossi M, Broglia L, Graziano P, et al: Local invasion of gastric cancer: CT findings and pathologic correlation using 5-mm incremental scanning, hypotonia, and water filling. AJR Am J Roentgenol 172:383– 388, 1999. 752. Rossmann MD, Burr LJ, Thorpe PE: Colovenous fistula complicating diverticulitis: CT and radiographic findings. Abdom Imaging 22:513– 515, 1997. 753. Rubesin SE, Beatty SM: The postoperative esophagus. Semin Roentgenol 29:401–410, 1994. 754. Rubesin SE, Gilchrist AM, Bronner M, et al: Non-Hodgkin lymphoma of the small intestine. Radiographics 10:985–998, 1990.

1740

PART II  CT and MR Imaging of the Whole Body

755. Rubin GD, Alfrey EJ, Dake MD, et al: Assessment of living renal donors with spiral CT. Radiology 195:457–462, 1995. 756. Ruers TJ, Langenhoff BS, Neeleman N, et al: Value of positron   emission tomography with [F-18]fluorodeoxyglucose in patients with colorectal liver metastases: A prospective study. J Clin Oncol 20:388– 395, 2002. 757. Ruiz Montes F, Rene Espinet JM, Buenestado Garcia J, et al: Primary gastric plasmacytoma of rapid growth presenting with upper gastrointestinal bleeding. Am J Gastroenterol 90:1349–1350, 1995. 758. Russell JC, Welch JP: Operative management of radiation injuries of the intestinal tract. Am J Surg 137:433–442, 1979. 759. Sabate JM, Franquet T, Palmer J, et al: AIDS-related primary esophageal lymphoma. Abdom Imaging 22:11–13, 1997. 760. Saindane AM, Losada M, Macari M: Focal amyloidoma of the small bowel mimicking adenocarcinoma on CT. AJR Am J Roentgenol 185:1187–1189, 2005. 761. Saitoh Y, Waxman I, West AB, et al: Prevalence and distinctive biologic features of flat colorectal adenomas in a North American population. Gastroenterology 120:1657–1665, 2001. 762. Salerno G, Daniels IR, Brown G: Magnetic resonance imaging of the low rectum: Defining the radiological anatomy. Colorectal Dis 8(Suppl 3):10–13, 2006. 763. Sandrasegaran K, Kopecky KK, Rajesh A, et al: Proximal small bowel intussusceptions in adults: CT appearance and clinical significance. Abdom Imaging 29:653–657, 2004. 764. Sandrasegaran K, Rajesh A, Rushing DA, et al: Gastrointestinal stromal tumors: CT and MRI findings. Eur Radiol 15:1407–1414, 2005. 765. Sarlomo-Rikala M, Kovatich AJ, Barusevicius A, et al: CD117: A sensitive marker for gastrointestinal stromal tumors that is more specific than CD34. Mod Pathol 11:728–734, 1998. 766. Sarr MG, Bulkley GB, Zuidema GD: Preoperative recognition of intestinal strangulation obstruction. Prospective evaluation of diagnostic capability. Am J Surg 145:176–182, 1983. 767. Saunders HS, Wolfman NT, Ott DJ: Esophageal cancer. Radiologic staging. Radiol Clin North Am 35:281–294, 1997. 768. Scatarige JC, Fishman EK, Jones B, et al: Gastric leiomyosarcoma: CT observations. J Comput Assist Tomogr 9:320–327, 1985. 769. Scheidler J, Stabler A, Kleber G, et al: Computed tomography in pneumatosis intestinalis: Differential diagnosis and therapeutic consequences. Abdom Imaging 20:523–528, 1995. 770. Schiller VL, Joyce PW, Sarti DA: Small-bowel obstruction due to   hernia through the primary laparoscopic port: A complication of laparoscopic cholecystectomy. AJR Am J Roentgenol 163:480–481, 1994. 771. Schindera ST, Nelson RC, DeLong DM, et al: Multi-detector row CT of the small bowel: Peak enhancement temporal window—Initial experience. Radiology 243:438–444, 2007. 772. Schmalfuss IM, Mancuso AA, Tart RP: Postcricoid region and cervical esophagus: Normal appearance at CT and MR imaging. Radiology 214:237–246, 2000. 773. Schmidt P, Koltai JL, Weltzien A: Actinomycosis of the appendix in childhood. Pediatr Surg Int 15:63–65, 1999. 774. Schnall MD: Magnetic resonance of the hollow viscera. In Gore RM, Levine MS, Laufer I, editors: Textbook of gastrointestinal radiology, Philadelphia, 1994, WB Saunders, pp 113–126. 775. Schoen FJ, Cotran RS: Blood vessels. In Cotran RS, et al, editors: Robbins pathologic basis of disease, Philadelphia, 1999, WB Saunders, pp 493–542. 776. Schreyer AG, Geissler A, Albrich H, et al: Abdominal MRI after enteroclysis or with oral contrast in patients with suspected or proven Crohn’s disease. Clin Gastroenterol Hepatol 2:491–497, 2004. 777. Schuchert MJ, Luketich JD: Barrett’s esophagus-emerging concepts and controversies. J Surg Oncol 95:185–189, 2007. 778. Schung GE: Leiomyoma of the cardioesophageal junction. Arch Surg 65:342–346, 1952. 779. Schwartz DL, So HB, Becker JM, et al: An ectopic gastric duplication arising from the pancreas and presenting with a pneumoperitoneum.   J Pediatr Surg 14:187–188, 1979.

780. Scott FE, Dupont PA, Webb J: Plasmacytoma of the stomach: Diagnosis with the aid of immunoperoxidase technique. Cancer 41:675–681, 1978. 781. Seremetis MG, Lyons WS, deGuzman VC, et al: Leiomyomata of the esophagus. An analysis of 838 cases. Cancer 38:2166–2177, 1976. 782. Seymour CW, Pryor JP, Gupta R, et al: Anaphylactoid reaction to oral contrast for computed tomography. J Trauma 57:1105–1107, 2004. 783. Shah RS, Kaddu SJ, Kirtane JM: Benign mature teratoma of the large bowel: A case report. J Pediatr Surg 31:701–702, 1996. 784. Shaheen NJ: Advances in Barrett’s esophagus and esophageal adenocarcinoma. Gastroenterology 128:1554–1566, 2005. 785. Shaheen NJ, Crosby MA, Bozymski EM, et al: Is there publication bias in the reporting of cancer risk in Barrett’s esophagus? Gastroenterology 119:333–338, 2000. 786. Shankar S, vanSonnenberg E, Desai J, et al: Gastrointestinal stromal tumor: New nodule-within-a-mass pattern of recurrence after   partial response to imatinib mesylate. Radiology 235:892–898, 2005. 787. Shanmuganathan K, Mirvis SE, Amoroso M: Periportal low density on CT in patients with blunt trauma: Association with elevated venous pressure. AJR Am J Roentgenol 160:279–283, 1993. 788. Sharma S, Bandhu S, Gulati MS, et al: Gastric hemangioma associated with phleboliths: CT appearance. AJR Am J Roentgenol 173:859–860, 1999. 789. Sheiman RG, Raptopoulos V, Caruso P, et al: Comparison of tailored and empiric scan delays for CT angiography of the abdomen. AJR Am J Roentgenol 167:725–729, 1996. 790. Sher ME, Bauer J: Radiation-induced enteropathy. Am J Gastroenterol 85:121–128, 1990. 791. Shibata S, Iwasaki N: Angiographic findings in diseases of the stomach. Am J Roentgenol Radium Ther Nucl Med 110:322–331, 1970. 792. Shimizu T, Kagami T, Matsumoto T, et al: [Clinical Studies on Behcet’s syndrome—With special reference to epidemiology and attackpredisposing factors.]. Naika 12:526–536, 1963. 793. Shimizu T, Namba R, Matsuoka T, et al: Esophageal varices before and after endoscopic variceal ligation: Evaluation using helical CT. Eur Radiol 9:1546–1549, 1999. 794. Shinners TJ, Pickhardt PJ, Taylor AJ, et al: Patient-controlled room air insufflation versus automated carbon dioxide delivery for CT colonography. AJR Am J Roentgenol 186:1491–1496, 2006. 795. Shorter RG: Risks of intestinal cancer in Crohn’s disease. Dis Colon Rectum 26:686–689, 1983. 796. Shulman HM, Sullivan KM, Weiden PL, et al: Chronic graft-versus-host syndrome in man. A long-term clinicopathologic study of 20 Seattle patients. Am J Med 69:204–217, 1980. 797. Shweiki E, West JC, Klena JW, et al: Eosinophilic gastroenteritis presenting as an obstructing cecal mass—A case report and review of the literature. Am J Gastroenterol 94:3644–3645, 1999. 798. Silverman JM, Hamlin JA: Large melanoma metastases to the gastrointestinal tract. Gut 30:1783–1785, 1989. 799. Simon AM, Birnbaum BA, Jacobs JE: Isolated infarction of the cecum: CT findings in two patients. Radiology 214:513–516, 2000. 800. Simstein NL: Internal herniation through a defect in the broad ligament. Am Surg 53:258–259, 1987. 801. Sivit CJ, Taylor GA, Bulas DI, et al: Posttraumatic shock in children: CT findings associated with hemodynamic instability. Radiology 182:723– 726, 1992. 802. Skehan SJ, Brown AL, Thompson M, et al: Imaging features of primary and recurrent esophageal cancer at FDG PET. Radiographics 20:713– 723, 2000. 803. Skehan SJ, Issenman R, Mernagh J, et al: 18F-fluorodeoxyglucose positron tomography in diagnosis of paediatric inflammatory bowel disease. Lancet 354:836–837, 1999. 804. Smart PE, Weinfeld A, Thompson NE, et al: Toxoplasmosis of the stomach: A cause of antral narrowing. Radiology 174:369–370, 1990. 805. Smerud MJ, Johnson CD, Stephens DH: Diagnosis of bowel infarction: A comparison of plain films and CT scans in 23 cases. AJR Am J Roentgenol 154:99–103, 1990.

CHAPTER 50  Gastrointestinal Tract 806. Sohn KM, Lee JM, Lee SY, et al: Comparing MR imaging and CT in the staging of gastric carcinoma. AJR Am J Roentgenol 174:1551–1557, 2000. 807. Sood RR, Joubert I, Franklin H, et al: Small bowel MRI: Comparison   of a polyethylene glycol preparation and water as oral contrast media.   J Magn Reson Imaging 15:401–408, 2002. 808. Soulen MC, Fishman EK, Scatarige JC, et al: Cryptosporidiosis of the gastric antrum: Detection using CT. Radiology 159:705–706, 1986. 809. Spears H, Petrelli NJ, Herrera L, et al: Treatment of bowel obstruction after operation for colorectal carcinoma. Am J Surg 155:383–386, 1988. 810. Spiegel JS, Schwabe AD: Disseminated cytomegalovirus infection with gastrointestinal involvement. The role of altered immunity in the elderly. Am J Gastroenterol 73:37–44, 1980. 811. Stafford RE, McGonigal MD, Weigelt JA, et al: Oral contrast solution and computed tomography for blunt abdominal trauma: A randomized study. Arch Surg 134:622–627, 1999. 812. Stahl A, Ott K, Weber WA, et al: FDG PET imaging of locally advanced gastric carcinomas: Correlation with endoscopic and histopathological findings. Eur J Nucl Med Mol Imaging 30:288–295, 2003. 813. Staib L, Schirrmeister H, Reske SN, et al: Is (18)F-fluorodeoxyglucose positron emission tomography in recurrent colorectal cancer a contribution to surgical decision making? Am J Surg 180:1–5, 2000. 814. Stavorovsky M, Jaffa AJ, Papo J, et al: Leiomyosarcoma of the colon and rectum. Dis Colon Rectum 23:249–254, 1980. 815. Steffens JC, Link J, Grassner J, et al: Contrast-enhanced, K-spacecentered, breath-hold MR angiography of the renal arteries and the abdominal aorta. J Magn Reson Imaging 7:617–622, 1997. 816. Stevens RL, Jones B, Fishman EK: The CT halo sign: A new finding in intestinal lymphangiectasia. J Comput Assist Tomogr 21:1005–1007, 1997. 817. Stewart ET, Dodds WJ: Radiology of the esophagus. In Freeny PC, Stevenson GW, editors: Alimentary tract radiology, St. Louis, 1994, CV Mosby, pp 192–271. 818. Stubenbord WT, Thorbjarnarson B: Intussusception in adults. Ann Surg 172:306–310, 1970. 819. Stuhlfaut JW, Soto JA, Lucey BC, et al: Blunt abdominal trauma: Performance of CT without oral contrast material. Radiology 233:689–694, 2004. 820. Sugimoto E, Lorelius LE, Eriksson B, et al: Midgut carcinoid tumours. CT appearance. Acta Radiol 36:367–371, 1995. 821. Sugimura K, Carrington BM, Quivey JM, et al: Postirradiation changes in the pelvis: Assessment with MR imaging. Radiology 175:805–813, 1990. 822. Suri S, Gupta S, Suri R: Computed tomography in abdominal tuberculosis. Br J Radiol 72:92–98, 1999. 823. Sussman SK, Halvorsen RA, Jr, Illescas FF, et al: Gastric adenocarcinoma: CT versus surgical staging. Radiology 167:335–340, 1988. 824. Symonds DA, Vickery AL: Mucinous carcinoma of the colon and rectum. Cancer 37:1891–1900, 1976. 825. Tada S, Iida M, Yao T, et al: Gastrointestinal amyloidosis: Radiologic features by chemical types. Radiology 190:37–42, 1994. 826. Takashima S, Takeuchi N, Shiozaki H, et al: Carcinoma of the esophagus: CT vs MR imaging in determining resectability. AJR Am J Roentgenol 156:297–302, 1991. 827. Takeyama N, Gokan T, Ohgiya Y, et al: CT of internal hernias. Radiographics 25:997–1015, 2005. 828. Tan EM, Cohen AS, Fries JF, et al: The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 25:1271–1277, 1982. 829. Tandon HD, Prakash A: Pathology of intestinal tuberculosis and its distinction from Crohn’s disease. Gut 13:260–269, 1972. 830. Taourel PG, Deneuville M, Pradel JA, et al: Acute mesenteric ischemia: Diagnosis with contrast-enhanced CT. Radiology 199:632–636, 1996. 831. Taylor SA, Halligan S, Goh V, et al: Optimizing colonic distention for multi-detector row CT colonography: Effect of hyoscine butylbromide and rectal balloon catheter. Radiology 229:99–108, 2003.

1741

832. Taylor SA, Halligan S, Slater A, et al: Polyp detection with CT colonography: Primary 3D endoluminal analysis versus primary 2D transverse analysis with computer-assisted reader software. Radiology 239:759–767, 2006. 833. Taylor AJ, Stewart ET, Dodds WJ: Gastrointestinal lipomas: A radiologic and pathologic review. AJR Am J Roentgenol 155:1205–1210, 1990. 834. Tedesco FJ, Volpicelli NA, Moore FS: Estrogen- and progesteroneassociated colitis: A disorder with clinical and endoscopic features mimicking Crohn’s colitis. Gastrointest Endosc 28:247–249, 1982. 835. Teixidor HS, Honig CL, Norsoph E, et al: Cytomegalovirus infection of the alimentary canal: Radiologic findings with pathologic correlation. Radiology 163:317–323, 1987. 836. Terasawa T, Blackmore CC, Bent S, et al: Systematic review: Computed tomography and ultrasonography to detect acute appendicitis in adults and adolescents. Ann Intern Med 141:537–546, 2004. 837. Thieblemont C, Bastion Y, Berger F, et al: Mucosa-associated lymphoid tissue gastrointestinal and nongastrointestinal lymphoma behavior: analysis of 108 patients. J Clin Oncol 15:1624–1630, 1997. 838. Thieblemont C, Berger F, Dumontet C, et al: Mucosa-associated lymphoid tissue lymphoma is a disseminated disease in one third of 158 patients analyzed. Blood 95:802–806, 2000. 839. Thomas JH, Rosato FE, Patterson LT: Epiploic appendagitis. Surg Gynecol Obstet 138:23–25, 1974. 840. Thomas ED, Storb R, Clift RA, et al: Bone-marrow transplantation (second of two parts). N Engl J Med 292:895–902, 1975. 841. Thompson WM, Halvorsen RA, Jr: Staging esophageal carcinoma II: CT and MRI. Semin Oncol 21:447–452, 1994. 842. Thompson WM, Halvorsen RA, Foster WL, Jr, et al: Computed tomography for staging esophageal and gastroesophageal cancer: reevaluation. AJR Am J Roentgenol 141:951–958, 1983. 843. Thompson WM, Halvorsen RA, Foster W, Jr, et al: Computed tomography of the gastroesophageal junction: Value of the left lateral decubitus view. J Comput Assist Tomogr 8:346–349, 1984. 844. Tio TL, Cohen P, Coene PP, et al: Endosonography and computed tomography of esophageal carcinoma. Preoperative classification compared to the new (1987) TNM system. Gastroenterology 96:1478– 1486, 1989. 845. Tominaga K, Arakawa T, Ando K, et al: Oesophageal cavernous haemangioma diagnosed histologically, not by endoscopic procedures.   J Gastroenterol Hepatol 15:215–219, 2000. 846. Toombs BD, Sandler CM, Lester RG: Computed tomography of chest trauma. Radiology 140:733–738, 1981. 847. Travers RL, Allison DJ, Brettle RP, et al: Polyarteritis nodosa: A clinical and angiographic analysis of 17 cases. Semin Arthritis Rheum 8:184–199, 1979. 848. Triester SL, Leighton JA, Leontiadis GI, et al: A meta-analysis of the yield of capsule endoscopy compared to other diagnostic modalities in patients with obscure gastrointestinal bleeding. Am J Gastroenterol 100:2407–2418, 2005. 849. Trommer G, Bewer A, Kosling S: [Mesenteric lymphadenopathy in Yersinia enterocolitica infection]. Radiologe 38:37–40, 1998. 850. Umeoka S, Koyama T, Togashi K, et al: Esophageal cancer: Evaluation with triple-phase dynamic CT—Initial experience. Radiology 239:777– 783, 2006. 851. Urban BA, Fishman EK, Hruban RH: Helicobacter pylori gastritis mimicking gastric carcinoma at CT evaluation. Radiology 179:689–691, 1991. 852. Urschel JD, Vasan H: A meta-analysis of randomized controlled trials that compared neoadjuvant chemoradiation and surgery to surgery alone for resectable esophageal cancer. Am J Surg 185:538–543, 2003. 853. Vaidya MG, Sodhi JS: Gastrointestinal tract tuberculosis: A study   of 102 cases including 55 hemicolectomies. Clin Radiol 29:189–195, 1978. 854. Valls C, Sanchez-Laforga A, Benasco C: Oat-cell carcinoma of the esophagus. AJR Am J Roentgenol 163:1266–1267, 1994. 855. Van Gossum A: Obscure digestive bleeding. Best Pract Res Clin Gastroenterol 15:155–174, 2001.

1742

PART II  CT and MR Imaging of the Whole Body

856. Vanhoenacker FM, De Backer AI, Op de BB, et al: Imaging of gastrointestinal and abdominal tuberculosis. Eur Radiol 14(Suppl 3):E103–E115, 2004. 857. Vantrappen G, Geboes K, Ponette E: Yersinia enteritis. Med Clin North Am 66:639–653, 1982. 858. van Westreenen HL, Heeren PA, van Dullemen HM, et al: Positron emission tomography with F-18-fluorodeoxyglucose in a combined staging strategy of esophageal cancer prevents unnecessary surgical explorations. J Gastrointest Surg 9:54–61, 2005. 859. van Westreenen HL, Westerterp M, Bossuyt PM, et al: Systematic review of the staging performance of 18F-fluorodeoxyglucose positron emission tomography in esophageal cancer. J Clin Oncol 22:3805–3812, 2004. 860. Vas W, Malpani AR, Singer J, et al: Computed tomographic evaluation of paraesophageal hernia. Gastrointest Radiol 14:291–294, 1989. 861. Vazquez-Sequeiros E: Nodal staging: Number or site of nodes? How   to improve accuracy? Is FNA always necessary? Junctional tumors— What’s N and what’s M? Endoscopy 38(Suppl 1):4–8, 2006. 862. Vazquez-Sequeiros E, Wiersema MJ, Clain JE, et al: Impact of lymph node staging on therapy of esophageal carcinoma. Gastroenterology 125:1626–1635, 2003. 863. Villanueva A, Perez C, Sabate JM, et al: CT manifestations of peritoneal leiomyosarcomatosis. Eur J Radiol 17:166–169, 1993. 864. Vitellas KM, Bennett WF, Bova JG, et al: Radiographic manifestations of eosinophilic gastroenteritis. Abdom Imaging 20:406–413, 1995. 865. Vliegen RF, Beets GL, von Meyenfeldt MF, et al: Rectal cancer: MR imaging in local staging—Is gadolinium-based contrast material helpful? Radiology 234:179–188, 2005. 866. von Rahden BH, Stein HJ: Staging and treatment of advanced esophageal cancer. Curr Opin Gastroenterol 21:472–477, 2005. 867. Vukasin P, Weston LA, Beart RW: Oral Fleet Phospho-Soda laxativeinduced hyperphosphatemia and hypocalcemic tetany in an adult: Report of a case. Dis Colon Rectum 40:497–499, 1997. 868. Wakelin SJ, Deans C, Crofts TJ, et al: A comparison of computerised tomography, laparoscopic ultrasound and endoscopic ultrasound in the preoperative staging of oesophago-gastric carcinoma. Eur J Radiol 41:161–167, 2002. 869. Wall SD, Jones B: Gastrointestinal tract in the immunocompromised host: Opportunistic infections and other complications. Radiology 185:327–335, 1992. 870. Wallace MB, Nietert PJ, Earle C, et al: An analysis of multiple staging management strategies for carcinoma of the esophagus: Computed tomography, endoscopic ultrasound, positron emission tomography, and thoracoscopy/laparoscopy. Ann Thorac Surg 74:1026–1032, 2002. 871. Walsh TH, Mann CV: Smooth muscle neoplasms of the rectum and anal canal. Br J Surg 71:597–599, 1984. 872. Warshauer DM, Lee JK: Adult intussusception detected at CT or MR imaging: Clinical-imaging correlation. Radiology 212:853–860, 1999. 873. Watanabe K, Kikuchi T, Katori Y, et al: The usefulness of computed tomography in the diagnosis of impacted fish bones in the oesophagus. J Laryngol Otol 112:360–364, 1998. 874. Weber WA, Ott K: Imaging of esophageal and gastric cancer. Semin Oncol 31:530–541, 2004. 875. Weber WA, Ott K, Becker K, et al: Prediction of response to preoperative chemotherapy in adenocarcinomas of the esophagogastric junction by metabolic imaging. J Clin Oncol 19:3058–3065, 2001. 876. Wechsler RJ, Kurtz AB, Needleman L, et al: Cross-sectional imaging of abdominal wall hernias. AJR Am J Roentgenol 153:517–521, 1989. 877. Westerterp M, van Westreenen HL, Reitsma JB, et al: Esophageal cancer: CT, endoscopic US, and FDG PET for assessment of response to neoadjuvant therapy—Systematic review. Radiology 236:841–851, 2005. 878. White TB, Allen HA, 3rd, Ives CE: Portal and mesenteric vein gas in diverticulitis: CT findings. AJR Am J Roentgenol 171:525–526, 1998. 879. White CS, Templeton PA, Attar S: Esophageal perforation: CT findings. AJR Am J Roentgenol 160:767–770, 1993. 880. Whitesell FB, Jr: Gallstone ileus. Am Surg 36:317–322, 1970.

881. Whitman GJ, Borkowski GP: Giant fibrovascular polyp of the esophagus: CT and MR findings. AJR Am J Roentgenol 152:518–520, 1989. 882. Wibe A, Rendedal PR, Svensson E, et al: Prognostic significance of the circumferential resection margin following total mesorectal excision for rectal cancer. Br J Surg 89:327–334, 2002. 883. Wieczorek RL, Seidman I, Ranson JH, et al: Congenital duplication of the stomach: Case report and review of the English literature. Am J Gastroenterol 79:597–602, 1984. 884. Wieder HA, Beer AJ, Lordick F, et al: Comparison of changes in tumor metabolic activity and tumor size during chemotherapy of adenocarcinomas of the esophagogastric junction. J Nucl Med 46:2029–2034, 2005. 885. Wiesner W, Khurana B, Ji H, et al: CT of acute bowel ischemia. Radiology 226:635–650, 2003. 886. Wiesner W, Mortele KJ, Glickman JN, et al: Portal-venous gas unrelated to mesenteric ischemia. Eur Radiol 12:1432–1437, 2002. 887. Williford ME, Thompson WM, Hamilton JD, et al: Esophageal tuberculosis: Findings on barium swallow and computed tomography. Gastrointest Radiol 8:119–122, 1983. 888. Willis J, Cooper GS, Isenberg G, et al: Correlation of EUS measurement with pathologic assessment of neoadjuvant therapy response in esophageal carcinoma. Gastrointest Endosc 55:655–661, 2002. 889. Winawer SJ, Stewart ET, Zauber AG, et al: A comparison of colonoscopy and double-contrast barium enema for surveillance after polypectomy. National Polyp Study Work Group. N Engl J Med 342:1766–1772, 2000. 890. Wold PB, Fletcher JG, Johnson CD, et al: Assessment of small bowel Crohn disease: Noninvasive peroral CT enterography compared with other imaging methods and endoscopy—feasibility study. Radiology 229:275–281, 2003. 891. Wolfel DA: Leiomyosarcoma of the esophagus. Am J Roentgenol Radium Ther Nucl Med 89:127–131, 1963. 892. Wolfman NT, Scharling ES, Chen MY: Esophageal squamous carcinoma. Radiol Clin North Am 32:1183–1201, 1994. 893. Wotherspoon AC, Ortiz-Hidalgo C, Falzon MR, et al: Helicobacter pylori-associated gastritis and primary B-cell gastric lymphoma. Lancet 338:1175–1176, 1991. 894. Wren SM, Stijns P, Srinivas S: Positron emission tomography in the initial staging of esophageal cancer. Arch Surg 137:1001–1007, 2002. 895. Wyatt SH, Fishman EK, Jones B: Primary lymphoma of the colon and rectum: CT and barium enema correlation. Abdom Imaging 18:376– 380, 1993. 896. Wychulis AR, Woolam GL, Andersen HA, et al: Achalasia and carcinoma of the esophagus. JAMA 215:1638–1641, 1971. 897. Wynder EL, Mabuchi K: Cancer of the gastrointestinal tract. Etiological and environmental factors. JAMA 226:1546–1548, 1973. 898. Yaghmai V, Nikolaidis P, Hammond NA, et al: Multidetector-row computed tomography diagnosis of small bowel obstruction: Can coronal reformations replace axial images? Emerg Radiol 13:69–72, 2006. 899. Yamada I, Murata Y, Izumi Y, et al: Staging of esophageal carcinoma in vitro with 4.7-T MR imaging. Radiology 204:521–526, 1997. 900. Yasuhara H: Acute mesenteric ischemia: The challenge of gastroenterology. Surg Today 35:185–195, 2005. 901. Yee J: Patient preparation for CT colonography. In Lefere P, Gryspeerdt S, editors: Virtual colonoscopy: A practical guide, Berlin, 2006, Springer, pp 23–34. 902. Yee J, Kumar NN, Hung RK, et al: Comparison of supine and prone scanning separately and in combination at CT colonography. Radiology 226:653–661, 2003. 903. Yeguez JF, Martinez SA, Sands LR, et al: Pelvic actinomycosis presenting as malignant large bowel obstruction: A case report and a review of the literature. Am Surg 66:85–90, 2000. 904. Ymaguchi M, Takeuchi S, Awazu S: Meckel’s diverticulum. Investigation of 600 patients in Japanese literature. Am J Surg 136:247–249, 1978. 905. Yoo CC, Levine MS, McLarney JK, et al: Primary malignant melanoma of the esophagus: Radiographic findings in seven patients. Radiology 209:455–459, 1998.

CHAPTER 50  Gastrointestinal Tract 906. Yoo CH, Noh SH, Kim YI, et al: Comparison of prognostic significance of nodal staging between old (4th edition) and new (5th edition) UICC TNM classification for gastric carcinoma. International Union Against Cancer. World J Surg 23:492–498, 1999. 907. Yoon SE, Ha HK, Lee YS, et al: Upper gastrointestinal series and CT findings of primary gastric plasmacytoma: Report of two cases. AJR Am J Roentgenol 173:1266–1268, 1999. 908. Yoon W, Jeong YY, Kim JK: Acute gastrointestinal bleeding: Contrastenhanced MDCT. Abdom Imaging 31:1–8, 2006. 909. Yoon W, Jeong YY, Shin SS, et al: Acute massive gastrointestinal bleeding: Detection and localization with arterial phase multi-detector row helical CT. Radiology 239:160–167, 2006. 910. Yoshino MT: Duodenal hematoma: CT demonstration of the ring sign. Gastrointest Radiol 12:330–332, 1987. 911. Yoshino T, Akagi T: Gastric low-grade mucosa-associated lymphoid tissue lymphomas: Their histogenesis and high-grade transformation. Pathol Int 48:323–331, 1998. 912. Yoshino T, Ichimura K, Mannami T, et al: Multiple organ mucosaassociated lymphoid tissue lymphomas often involve the intestine. Cancer 91:346–353, 2001. 913. Yoshioka T, Yamaguchi K, Kubota K, et al: Evaluation of 18F-FDG PET in patients with a, metastatic, or recurrent gastric cancer. J Nucl Med 44:690–699, 2003. 914. Younes M, Katikaneni PR, Lechago J: The value of the preoperative mucosal biopsy in the diagnosis of colorectal mucinous adenocarcinoma. Cancer 72:3588–3592, 1993. 915. Young TH, Ho AS, Tang HS, et al: Cystic lymphangioma of the transverse colon: Report of a case and review of the literature. Abdom Imaging 21:415–417, 1996. 916. Young LS, Inderlied CB, Berlin OG, et al: Mycobacterial infections in AIDS patients, with an emphasis on the Mycobacterium avium complex. Rev Infect Dis 8:1024–1033, 1986. 917. Yousem DM, Fishman EK, Jones B: Crohn disease: Perirectal and perianal findings at CT. Radiology 167:331–334, 1988. 918. Yuen EH, Yang WT, Lam WW, et al: Spontaneous intramural haematoma of the oesophagus: CT and MRI appearances. Australas Radiol 42:139–142, 1998. 919. Yun M, Lim JS, Noh SH, et al: Lymph node staging of gastric cancer using (18)F-FDG PET: a comparison study with CT. J Nucl Med 46:1582–1588, 2005.

1743

920. Zalcman M, Sy M, Donckier V, et al: Helical CT signs in the diagnosis of intestinal ischemia in small-bowel obstruction. AJR Am J Roentgenol 175:1601–1607, 2000. 921. Zalcman M, Van Gansbeke D, Lalmand B, et al: Delayed enhancement of the bowel wall: A new CT sign of small bowel strangulation.   J Comput Assist Tomogr 20:379–381, 1996. 922. Zalev AH, Prokipchuk EJ, Jeejeebhoy KN, et al: Recurrent Crohn’s disease in the duodenum and jejunum following extensive   small bowel resection and jejunocolic anastomosis: Radiologic   findings in twenty-five patients. Abdom Imaging 24:538–543, 1999. 923. Zalis ME, Barish MA, Choi JR, et al: CT colonography reporting and data system: A consensus proposal. Radiology 236:3–9, 2005. 924. Zerhouni EA, Rutter C, Hamilton SR, et al: CT and MR imaging in the staging of colorectal carcinoma: Report of the Radiology Diagnostic Oncology Group II. Radiology 200:443–451, 1996. 925. Zervos EE, Badgwell BD, Burak WE, Jr, et al: Fluorodeoxyglucose positron emission tomography as an adjunct to carcinoembryonic antigen in the management of patients with presumed recurrent colorectal cancer and nondiagnostic radiologic workup. Surgery 130:636–643, 2001. 926. Zhang J, Duan ZQ, Song QB, et al: Acute mesenteric venous thrombosis: A better outcome achieved through improved imaging techniques and a changed policy of clinical management. Eur J Vasc Endovasc Surg 28:329–334, 2004. 927. Zheng X, Cheng J, Pan K, et al: Eosinophilic enteritis: CT features. Abdom Imaging 33:191–195, 2008. 928. Zhuang H, Sinha P, Pourdehnad M, et al: The role of positron   emission tomography with fluorine-18-deoxyglucose in identifying colorectal cancer metastases to liver. Nucl Med Commun 21:793–798, 2000. 929. Ziegler K, Sanft C, Zimmer T, et al: Comparison of computed tomography, endosonography, and intraoperative assessment in TN staging of gastric carcinoma. Gut 34:604–610, 1993. 930. Zissin R, Hertz M, Gayer G, et al: Congenital internal hernia as a cause of small bowel obstruction: CT findings in 11 adult patients. Br J Radiol 78:796–802, 2005. 931. Zuckerman GR, Prakash C, Askin MP, et al: AGA technical review on the evaluation and management of occult and obscure gastrointestinal bleeding. Gastroenterology 118:201–221, 2000.

51  Rectum Raj M. Paspulati

ANATOMY The rectum and anal canal form the terminal part of the intestine. The rectum varies in length from 10 to 15 cm and extends from the rectosigmoid junction at the level of the S3 sacral vertebra to the anorectal line. The surgical anorectal junction is at the upper edge of the anorectal muscle ring located about 4 cm from the anal verge. The anatomic anorectal junction called the dentate line marks the junction of the endodermal primitive gut with ectodermal proctodeum. It is located about 2 to 3 cm from the anal verge. The rectum is predominantly retroperitoneal, and only its proximal 1 to 2 cm has a peritoneal extension along its anterior and lateral walls. The depth of peritoneal reflection is variable between men and women. This peritoneal reflection in men (between the rectum and bladder) is called the rectovesical recess, and in women (between the rectum and uterus), the rectouterine recess. The depth of the rectovesical recess in men is deeper than the rectouterine recess in women. The rectum has three lateral curvatures or flexures with corresponding indentations on the opposite-side wall produced by crescent-like projections of mucosa, submucosa, and circular musculature of the rectal wall called rectal valves. The superior and inferior rectal valves are on the left side and located about 4 cm below the rectosigmoid junction and 2 to 3 cm above the dentate line, respectively. The middle rectal valve is on the right side about 6 to 7 cm above the dentate line. The rectal wall is composed of mucosa, submucosa, and muscularis propria comprising inner circular and outer longitudinal muscle layers. The rectum is surrounded by variable amounts of mesorectal fat limited peripherally by mesorectal fascia (MRF), which is a condensation of the visceral layer of the endopelvic fascia. The mesorectal fat has perirectal lymphatics, vessels, and lymph nodes. Posteriorly the MRF fuses with the presacral fascia, which separates it from the presacral space; anteriorly it fuses with the Denonvilliers fascia, separating it from the prostate and seminal vesicles in males and the vagina in females. The MRF forms an important local barrier for the spread of rectal carcinoma and an important plane for surgical resection. High-resolution T2-weighted MRI demonstrates the MRF as a thin hypointense layer surrounding the high-signal-intensity mesorectal fat (Fig. 51-1). The anal canal has an inner internal sphincter and an outer external sphincter. The internal sphincter is composed of obliquely oriented smooth muscle fibers that are contiguous with the inner circular muscle layer of the rectum. The external sphincter is composed of striated muscle fibers arranged in subcutaneous, superficial, and deep portions. The deep portion of the external sphincter blends with the puborectalis portion of the levator ani muscle. The rectum and upper anal canal are supplied by the inferior mesenteric artery through the superior rectal artery. The internal iliac artery, through its middle and inferior rectal arteries, supplies the lower

1744

anal canal. Lymphatic and venous drainage of the rectum and anal canal follows the corresponding arteries. The immediate lymphatic drainage is to regional perirectal lymph nodes and later follows the major lymph nodes along the superior rectal and inferior mesenteric vessels. There is also lateral lymphatic drainage along the middle rectal vessels to the internal iliac lymph nodes.

RECTAL CARCINOMA Introduction Colorectal cancer is the third most common cancer in the United States, and approximately 39,610 cases of rectal cancer were expected to be diagnosed in the year 2015.11 Adenocarcinoma is the predominant rectal malignancy. Other rare types include neuroendocrine, squamous cell, adenosquamous, spindle cell, and undifferentiated carcinomas.16 The epidemiology, etiology, and pathogenesis of rectal carcinoma are similar to colon cancer. Diagnosis is established by digital rectal examination followed by proctoscopy and tissue diagnosis. Advances in surgical technique and adjuvant preoperative chemoradiation treatment of rectal carcinoma have led to significant improvement in disease-free survival. Accurate staging is vital for treatment planning. Imaging is mainly used for staging and treatment planning. Table 51-1 outlines the American Joint Committee on Cancer’s (AJCC’s) tumor-node-metastasis (TNM) staging of rectal carcinoma.

Imaging Modalities

Endorectal Ultrasonography (EUS).  EUS has the ability to demonstrate different layers of the rectal wall and is accurate in assessing the depth of tumor invasion into the rectal wall. EUS has a sensitivity, specificity, and accuracy of 96% to 97% in detection of muscularis propria invasion in early-stage rectal carcinoma. Sensitivity, specificity, and accuracy of EUS in detection of positive regional perirectal lymph nodes in early-stage rectal carcinoma are 53%, 77%, and 72%, respectively. EUS has several limitations, including inability to image high rectal and stenosing carcinoma, a limited field of view (FOV) that makes it difficult to assess the extent of perirectal invasion, inability to identify the MRF, and limited evaluation of the lateral pelvic wall and surrounding visceral invasion.2,4

Magnetic Resonance Imaging (MRI).  High-resolution MRI has several advantages over EUS and computed tomography (CT) for accurate staging of rectal carcinoma. It can identify the MRF and can accurately identify involvement of the surgical resection margin for total mesorectal excision. Its large FOV allows accurate delineation of large locally invasive rectal carcinomas and extension to the lateral pelvic wall and invasion of surrounding viscera. MRI has sensitivities from 74% to 94% and specificities from 74% to 76% for staging T3

CHAPTER 51  Rectum

A

1745

B FIG 51-1  Normal mesorectal fascia. Axial (A) and coronal (B) T2-weighted images demonstrate well-defined thin hypointense layer (arrowheads) surrounding the hyperintense mesorectal fat.

TABLE 51-1  TNM Staging of

protocol in pretreatment staging of rectal carcinoma. An MRI protocol is given in Table 51-2.

Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of tumor Tis Carcinoma in situ: intraepithelial or invasion of lamina propria T1 Tumor invades submucosa T2 Tumor invades muscularis propria T3 Tumor invades through the muscularis propria into the perirectal fat T4a Tumor penetrates to the surface of visceral peritoneum T4b Tumor directly invades or is adherent to other organs or structures

CT.  Multidetector CT (MDCT) has low sensitivity and accuracy for T

Regional Lymph Nodes (N) NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis in 1-3 regional lymph nodes N1a Metastasis in 1 regional lymph node N1b Metastasis in 2-3 regional lymph nodes N1c Tumor deposit(s) in the subserosa, mesentery, or nonperitonealized pericolic or perirectal tissues without regional nodal metastasis N2 Metastasis in 4 or more regional lymph nodes N2a Metastasis in 4-6 regional lymph nodes N2b Metastasis in 7 or more regional lymph nodes

FDG-PET.  Hybrid imaging with fluorine-18 fluorodeoxyglucose

Rectal Carcinoma

Distant Metastasis (M) M0 No distant metastasis M1 Distant metastasis M1a Metastasis confined to one organ or site (e.g., liver, lung, ovary, nonregional node) M1b Metastases in more than one organ/site or the peritoneum From Edge S, et al (eds): AJCC cancer staging manual, 7th ed, American Joint Committee on Cancer, 2010, New York, Springer-Verlag.

and T4 rectal carcinomas.4,6,34,39 High-resolution MRI is a reliable and reproducible imaging technique with a high sensitivity, specificity, and accuracy in predicting a negative surgical or circumferential resection margin (CRM).3,18,28 High-resolution T2-weighted images obtained at three orthogonal planes to the tumor are the key images of the MRI

staging of rectal carcinoma and is not the imaging modality of choice for pretreatment staging of rectal carcinoma. A large meta-analysis by Kwok et al. has shown a sensitivity of 78% and an accuracy of 73% for detection of transmural extension, and sensitivity and specificity of 52% and 78%, respectively, for detection of mesorectal lymph node metastases.27,32,36 Similar to colon cancer, contrast-enhanced MDCT in portal venous phase is recommended by the American College of Radiology for identification of distant metastases.14

(18F FDG) positron emission tomography (PET) is not suitable for local T staging of rectal carcinoma owing to its poor spatial resolution. It is also not suitable for detection of regional mesorectal lymph node metastases because of the small size of these lymph nodes. FDG PET/CT is an important imaging modality for distant staging and has a higher detection rate of distant metastases as compared to MDCT.1,19 Though FDG PET is not suitable for T staging, it is found to be useful for prediction of treatment response to neoadjuvant chemoradiation and also in assessment of true pathologic response to treatment.12,30 Recent introduction of PET MRI into clinical practice has led to a potential role of PET MRI in initial TNM staging and follow-up of rectal carcinoma.33,38

Imaging in Rectal Carcinoma

T Staging.  Rectal carcinoma may present as a circumferential mass, with or without annular stenosis, or a polypoid mass protruding into the lumen (Fig. 51-2). A central indentation within the mass is suggestive of an ulcerating mass. Tumor mass commonly has minimal hyperintense signal intensity to the low-signal-intensity muscularis propria on T2-weighted images. They all demonstrate variable degrees of enhancement on postcontrast T1-weighted images. Mucinous tumors have very high signal intensity on T2-weighted images, with minimal or no enhancement on postcontrast T1-weighted images and no diffusion restriction (Fig. 51-3). Similar to mucinous adenocarcinomas of the colon, they have a poor prognosis.21,25 The management of T1 and T2 rectal carcinoma is by surgery, and management of T3 and T4 tumors is initially with neoadjuvant

1746

PART II  CT and MR Imaging of the Whole Body

TABLE 51-2  MRI PROTOCOL Patient preparation: Fleet enema in the morning on the day of examination to clear the rectum of stool, and 1 mg of intramuscular glucagon immediately prior to MRI to reduce bowel peristalsis. Rectal distension with about 60 mL of ultrasound gel will improve visualization of the primary tumor and accuracy of T staging. Hardware: 1.5T- or 3T-field-strength magnet with phased-array coils; 3T is preferable because it provides better signal-to-noise ratio. MR sequences: Sequence T2 HASTE / single shot T2 FSE T2 FSE T2 FSE DWI T1 3D GRE (VIBE / LAVA)

Imaging Plane Sagittal Axial Coronal Sagittal Axial Coronal Axial Sagittal Axial Coronal

TR/TE

FOV (cm)

Section Thickness

Matrix Size

ETL

NSA

2500-500/65-85

30-50 cm

4-5 mm

256 × 256

8-12

1-2

Provides location of the tumor for planning high-resolution sequences

4500-7000/65-80

16-25 cm

3 mm

320 × 256

8-12

2-4

4500-7000/65-80 4500-7000/65-80 5000-12000/80 3.5-4/1.4

16-25 cm 16-25 cm 25 cm 25 cm

3 mm 3 mm 4 mm 1.5-2 mm

320 × 256 320 × 256 256 × 192 256 × 320

8-12 8-12 — —

2-4 2-4 2-4 1

Craniocaudal extent of tumor and T staging Radial extension and T staging T staging Functional imaging Optional sequence not mandatory for T staging

Comment

DWI, diffusion-weighted imaging; ETL, echo train length; FOV, field of view; FSE, fast-spin echo; GRE, gradient recalled echo; HASTE, halfFourier acquisition single-shot turbo-spin echo; LAVA, liver acquisition volume acceleration; NSA, number of signal averages; TE, echo time; TR, repetition time; VIBE, volumetric interpolated breath-hold examination.

A

B FIG 51-2  Rectal cancer gross morphology. Sagittal T2-weighted images of the rectum demonstrate (A) a well-defined polypoid endoluminal mass (arrow) arising from the posterior wall and (B) an annular stenosing mass.

chemoradiation, followed by surgery. Tumors involving the mesorectal fat are treated with total mesorectal excision (TME) following initial neoadjuvant chemoradiation treatment. TME includes removal of the tumor and rectum en bloc with surrounding mesorectal fat. Axial T2-weighted MRI perpendicular to the tumor can differentiate T2 from T3 tumor by demonstrating breach of tumor through the T2-hypointense layer of muscularis propria into the hyperintense mesorectal fat (Figs. 51-4 and 51-5). MRI cannot differentiate

T1- from T2-stage tumors and is not important insofar as both stages are treated by surgery. Once the tumor is identified as stage T3, the extent of mesorectal fat extension and relation of the tumor margin to the MRF, surrounding viscera (prostate, seminal vesicles, vagina, cervix), lateral pelvic wall, and presacral fascia is identified. A T3 tumor with invasion of the mesorectal fat is subclassified based on its proximity to MRF into (1) those involving the MRF or within 1 mm from the MRF (circumferential resection margin) and (2) those not

CHAPTER 51  Rectum

A

B

FIG 51-3  Mucinous adenocarcinoma. A, Axial T2-weighted image of the rectum demonstrates lobulated hyperintense mass. B, Axial DWI ADC image of the rectum demonstrates hyperintense signal due to low diffusion restriction of mucinous tumor.

A

B FIG 51-4  Stage T2 rectal carcinoma. Sagittal (A) and axial (B) T2-weighted images reveal a plaquelike anterior rectal mass invading into the muscularis propria (arrows) but no transmural extension into the perirectal fat.

1747

1748

PART II  CT and MR Imaging of the Whole Body

extending to MRF (Fig. 51-6). MRI can identify the relation of the tumor margin to the MRF with high accuracy.9,22,35 This distinction is important because tumors involving or within 1 mm of MRF have a higher risk of local recurrence (Fig. 51-7). A T4 tumor is identified by involvement of surrounding viscera (prostate, seminal vesicles, uterus, vagina), anal sphincters, peritoneal reflection, presacral fascia, sacrum, lateral pelvic wall, and adjacent vasculature (Fig. 51-8). T2-weighted images in three planes can accurately identify involvement of the

surrounding viscera. Sagittal T2-weighted images are useful for involvement of peritoneal reflection in high rectal tumors, and tumor involvement of anal sphincters is best evaluated with coronal T2-weighted images (Fig. 51-9). Posterior tumor extension to the sacrum and sacral nerve roots above the S2 vertebra makes the tumor unresectable. MRI can identify extramural vascular invasion perirectal vessels, which is a poor prognostic sign and predictor of both local and distant treatment failure8,37 (see Fig. 51-9).

Nodal Staging.  The lymph nodal groups that can be involved by rectal carcinoma include mesorectal, superior rectal, inferior mesenteric, inguinal, internal, external and common iliac, and retroperitoneal lymph nodes. Superior soft tissue resolution of MRI facilitates better identification of regional mesorectal and pelvic wall lymph nodes. Though differentiation of benign from metastatic lymph nodes is still based on size criteria, additional morphologic data such as T2 signal intensity, margins, and enhancement pattern on postcontrast T1-weighted MRI are useful ancillary features. Irregular margins, heterogeneous signal intensity, and enhancement are indicators of a malignant lymph node (Fig. 51-10). A 5-mm cutoff value for differentiating a benign from a malignant lymph node has a sensitivity of 68% and specificity of 78%.4,9,13,24 Distant extrapelvic lymph nodal metastases are better identified by contrast-enhanced MDCT and 18F FDG PET12,13,15,19 (Fig. 51-11).

Distant Metastases.  Liver is the most common site of distant

FIG 51-5  Stage T3 rectal carcinoma. Axial T2-weighted image demonstrates a right rectal wall mass with transmural extension into the perirectal fat (arrow).

A

metastases from rectal carcinoma. Lung is the second most common site, with higher incidence in low rectal tumors owing to systemic venous drainage of this portion of the rectum and also with high-grade carcinoma. Low rectal tumors can present with pulmonary metastases without liver metastases. Adrenal glands, peritoneum, and omentum are other sites of metastases. Bone and brain are less common sites of metastases. A contrast-enhanced CT is the initial imaging modality for detection of distant metastases. FDG PET/CT and MRI are used for trouble-shooting, with indeterminate findings on MDCT (Fig. 51-12). Hepatic metastases are seen as hypodense lesions against an enhancing

B FIG 51-6  Stage T3 rectal carcinoma with extension into mesorectal fat. A, Axial T2-weighted image reveals a right lateral rectal wall mass extending into the mesorectal fat (dashed arrow) but no involvement of the mesorectal fascia (arrowheads) and a mesorectal fat lymph node (arrow). B, Axial T2-weighted image demonstrates a circumferential rectal mass with mesorectal fat invasion (arrow) and involvement of mesorectal fascia (arrowhead) on the left.

CHAPTER 51  Rectum hepatic parenchyma in a contrast-enhanced MDCT. The overall sensitivity of MDCT for liver metastases measuring larger than 1 cm is from 77% to 94%. Lesions smaller than 1 cm cannot be definitely characterized by MDCT, with sensitivity as low as 41.9%. MRI of the liver has higher specificity in characterizing these small (<1 cm) liver lesions. Diffusion weighted imaging (DWI) and use of liver-specific MRI

1749

contrast agents can further increase the sensitivity of MRI in detection of small hepatic metastases.5,19,26,29,31,40

Follow-up Imaging After Neoadjuvant Chemoradiation.  MRI is also useful in evaluation of tumor response following neoadjuvant chemoradiation treatment. Tumor shrinkage, necrosis, and fibrosis can be identified on T2-weighted images. Fibrosis is hypointense as compared to relatively hyperintense tumor on pretreatment MRI. It is important to have a baseline pretreatment MRI for comparison to minimize interpretation errors. Presence of viable tumor within the fibrosis cannot be distinguished by MRI. DWI is of added value in both predicting the response of rectal carcinoma to neoadjuvant treatment and identifying the degree of tumor response following treatment (Fig. 51-13). Comparison of apparent diffusion coefficient (ADC) values of tumor in the pre- and posttreatment MRI demonstrate increased ADC values within the responding tumor mass. There is negative correlation between the mean pretreatment ADC and tumor regression following neoadjuvant chemoradiation.7,10,15,17,20 18F FDG PET/CT-MRI is also found to be useful to assess tumor response following neoadjuvant chemoradiation treatment12,18,19,23,30 (Fig. 51-14).

Pitfalls in Local Staging of Rectal Carcinoma.  Most of the

FIG 51-7  Stage T3 rectal carcinoma with extension into mesorectal fat. Axial T2-weighted image demonstrates a circumferential rectal mass with mesorectal fat invasion (arrow) and more than 1 mm from the mesorectal fascia (arrowhead) on the left, indicating no involvement of the circumferential resection margin.

staging failure with MRI is seen in differentiation of T2 from early T3 stages. There is a tendency toward overstaging due to the presence of perirectal inflammatory and desmoplastic changes. Desmoplastic changes have a spiculated appearance as compared to the nodular growth pattern of tumor. Peritumoral inflammatory and desmoplastic changes can also overestimate MRF involvement in T3 tumors. Accurate staging of low rectal cancers is difficult because of the paucity of perirectal fat planes. T2-weighted high-resolution coronal images are more useful to assess tumor extension to puborectalis and levator muscle.

Standard Radiology Report.  Following is the format for a standardized radiology report for MRI staging of rectal carcinoma.

MASS

RECTUM

A

B FIG 51-8  Stage T4 rectal carcinoma. A, Sagittal T2-weighted image reveals an invasive low rectal and anal canal mass (arrows) invading into the prostate (arrowheads). B, Sagittal T2-weighted image reveals an invasive midrectal mass (arrows) invading the posterior wall of the cervix (arrowhead).

1750

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 51-9  Rectal carcinoma with levator muscle invasion. A, Coronal T2-weighted image demonstrates a low rectal and anal carcinoma (arrow) invading the left puborectalis muscle (arrowhead). B and C, Rectal carcinoma with perirectal vascular invasion. Axial T2-weighted image (B) and axial postcontrast T1-weighted image (C) reveal a T3 rectal carcinoma with perirectal vascular invasion (arrow).

A

B

C FIG 51-10  Rectal carcinoma with metastatic lymph nodes. A, Axial T2-weighted image reveals a right lateral wall tumor with perirectal fat invasion (dashed arrow) and an enlarged mesorectal lymph node (arrow). No invasion of the mesorectal fascia (arrowheads). B, Coronal T2-weighted image demonstrates a T3 right lateral wall tumor (dashed arrow) and an enlarged superior rectal lymph node (arrow) with heterogeneous signal intensity. C, Axial T2-weighted image reveals an enlarged right lateral wall pelvic lymph node (arrow) with irregular contour and heterogeneous signal intensity.

CHAPTER 51  Rectum

A

B FIG 51-11  Rectal carcinoma with distant lymph nodal metastasis. PET/MRI images. A, Axial postcontrast T1-weighted image demonstrates an enlarged distal left paraaortic lymph node (arrow). B, Hybrid fused axial 18 F FDG PET/MRI demonstrates increased FDG uptake within the enlarged distal left paraaortic lymph node (arrow).

A

B

FIG 51-12  Rectal carcinoma with liver metastasis. A, Coronal contrast-enhanced T1-weighted MRI demonstrates large hepatic mass lesions (arrows) with peripheral thick rim enhancement. B, Coronal hybrid 18F FDG PET/CT image demonstrates intense peripheral FDG uptake within the liver lesions (arrows).

1751

1752

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 51-13  Posttreatment response of T3 rectal carcinoma to neoadjuvant chemoradiation treatment. Pretreatment axial T2-weighted image (A) and DWI ADC image (B) demonstrate a semicircumferential low rectal mass with low ADC value due to diffusion restriction. Posttreatment axial T2-weighted image (C) and DWI ADC image (D) reveal decrease in size and T2 signal intensity of the mass but persistent low ADC signal on DWI image, indicating persistent viable tumor and incomplete tumor response to chemoradiation treatment.

CHAPTER 51  Rectum

A

B

C

G

D

E

F

H

1753

FIG 51-14  PET/MRI in assessment of posttreatment response of T3 rectal carcinoma to neoadjuvant chemoradiation treatment. Pretreatment sagittal (A) and axial (B) T2-weighted images with corresponding hybrid fused 18F FDG PET images in sagittal (B) and axial (G) planes demonstrate an invasive hypermetabolic high rectal tumor (arrowheads) invading the seminal vesicles and mesorectal fat (arrows). Posttreatment sagittal (D) and axial (F) T2-weighted images with corresponding hybrid fused 18F FDG PET images in sagittal (E) and axial (H) planes demonstrate marked decrease in size of the mass, without perirectal invasion, and also resolution of the hypermetabolic activity, indicating near-complete response to treatment.

1. MRI technique and image quality 2. Tumor location (from anal verge and anorectal junction ) Low: <5 cm from anal verge Mid: 5-10 cm from anal verge High: >10 cm from anal verge 3. Tumor characteristics Circumferential, stenosing, polypoid Mucinous 4. T staging 5. T3 tumor MRF involvement Distance from MRF Extramural vascular invasion 6. T4 tumor Organ involved (prostate, seminal vesicle, uterus, and bladder) Lateral pelvic wall extension Bone: sacrum invasion Vascular: iliac vessel invasion/encasement Levator muscle 7. Lymph nodes 8. Mesorectal lymph nodes 9. Number, size, morphology, and signal intensity 10. Additional comments Fluid collections, incidental findings

REFERENCES 1. Abdel-Nabi H, Doerr RJ, Lamonica DM, et al: Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole-body PET: Correlation with histopathologic and CT findings. Radiology 206:755–760, 1998.

2. Akasu T, Kondo H, Moriya Y, et al: Endorectal ultrasonography and treatment of early stage rectal cancer. WORLD 24(9):1061–1068, 2000. 3. Algebally AM, Mohey N, Szmigielski W, et al: The value of highresolution MRI technique in patients with rectal carcinoma: Preoperative assessment of mesorectal fascia involvement, circumferential resection margin and local staging. Pol J Radiol 80:115–121, 2015. 4. Al-Sukhni E, Milot L, Fruitman M, et al: Diagnostic accuracy   of MRI for assessment of T category, lymph node metastases, and circumferential resection margin involvement in patients with rectal cancer: A systematic review and meta-analysis. Ann Surg Oncol 19:2212–2223, 2012. 5. Berger-Kulemann V, Schima W, Baroud S, et al: Gadoxetic acid-enhanced 3.0 T MR imaging versus multidetector-row CT in the detection of colorectal metastases in fatty liver using intraoperative ultrasound and histopathology as a standard of reference. Eur J Surg Oncol 38:670–676, 2012. 6. Bipat S, Glas AS, Slors FJ, et al: Rectal cancer: Local staging and assessment of lymph node involvement with endoluminal US, CT, and MR imaging—A meta-analysis. Radiology 232(3):773–783, 2004. 7. Birlik B, Obuz F, Elibol FD, et al: Diffusion-weighted MRI and MR-volumetry—in the evaluation of tumor response after preoperative chemoradiotherapy in patients with locally advanced rectal cancer. Magn Reson Imaging 33(2):201–212, 2015. 8. Brown G, Radcliffe AG, Newcombe RG, et al: Preoperative assessment of prognostic factors in rectal cancer using high-resolution magnetic resonance imaging. Br J Surg 90:355–364, 2003. 9. Brown G, Richards CJ, Bourne MW, et al: Morphologic predictors of lymph node status in rectal cancer with use of high-spatial-resolution MR imaging with histopathologic comparison. Radiology 227(2):371– 377, 2003. 10. Cai G, Xu Y, Zhu J, et al: Diffusion-weighted imaging for predicting the response of rectal cancer to neoadjuvant concurrent chemoradiation. World J Gastroenterol 19(33):5520–5527, 2013. 11. Cancer facts and figures, Atlanta, 2015, American Cancer Society. .

1754

PART II  CT and MR Imaging of the Whole Body

12. de Geus-Oei LF, Vriens D, van Laarhoven HW, et al: Monitoring and predicting response to therapy with 18F-FDG PET in colorectal cancer: A systematic review. J Nucl Med 50(Suppl 1):43S–54S, 2009. 13. Dewhurst CE, Mortele KJ: Magnetic resonance imaging of rectal cancer. Radiol Clin North Am 51:121–131, 2013. 14. Dewhurst C, Rosen MP, Blake MA, et al: ACR appropriateness criteria pretreatment staging of colorectal cancer. J Am Coll Radiol 9:775–781, 2012. 15. Dzik-Jurasz A, Domenig C, George M, et al: Diffusion MRI for prediction of response of rectal cancer to chemoradiation. Lancet 360(9329):307–308, 2002. 16. Fleming M, Ravula S, Tatishchev SF, et al: Colorectal carcinoma: Pathologic aspects. J Gastrointest Oncol 3(3):153–173, 2012. 17. Genovesi D, Filippone A, Ausili Cèfaro G, et al: Diffusion-weighted magnetic resonance for prediction of response after neoadjuvant chemoradiation therapy for locally advanced rectal cancer: Preliminary results of a monoinstitutional prospective study. Eur J Surg Oncol 39(10):1071–1078, 2013. 18. Granero-Castro P, Muñoz E, Frasson M, et al: Evaluation of mesorectal fascia in mid and low anterior rectal cancer using endorectal ultrasound is feasible and reliable: A comparison with MRI findings. Dis Colon Rectum 57(6):709–714, 2014. 19. Grassetto G, Marzola MC, Minicozzi A, et al: F-18 FDG PET/CT in rectal carcinoma: Where are we now? Clin Nucl Med 36:884–888, 2011. 20. Hein PA, Kremser C, Judmaier W, et al: Diffusion weighted magnetic resonance imaging for monitoring diffusion changes in rectal carcinoma during combined, preoperative chemoradiation: Preliminary results of a prospective study. Eur J Radiol 45(3):214–222, 2003. 21. Hussain SM, Outwater EK, Siegelman ES: Mucinous versus nonmucinous differentiation with MR imaging. Radiology 213(1):79–85, 1999. 22. Iannicelli E, Di Renzo S, Ferri M, et al: Accuracy of high-resolution MRI with lumen distention in rectal cancer staging and circumferential margin involvement prediction. Korean J Radiol 15(1):37–44, 2014. 23. Joye I, Deroose CM, Vandecaveye V, et al: The role of diffusion-weighted MRI and (18) F-FDG PET/CT in the prediction of pathologic complete response after radiochemotherapy for rectal cancer: A systematic review. Radiother Oncol 113(2):158–165, 2014. 24. Kaur H, Choi H, You YN, et al: MR imaging for preoperative evaluation of primary rectal cancer: Practical considerations. Radiographics 32:389–409, 2012. 25. Kim MJ, Park JS, Park SI, et al: Accuracy in differentiation of mucinous and nonmucinous rectal carcinoma on MR imaging. J Comput Assist Tomogr 27:48–55, 2003. 26. Kirke R, Rajesh A, Verma R, et al: Rectal cancer: Incidence of pulmonary metastases on thoracic CT and correlation with T staging. J Comput Assist Tomogr 31:569–571, 2007. 27. Kwok H, Bissett IP, Hill GL: Preoperative staging of rectal cancer. Int J Colorectal Dis 15:9–20, 2000.

28. Lahaye MJ, Engelen SM, Nelemans PJ, et al: Imaging for predicting the risk factors—the circumferential resection margin and nodal disease—of local recurrence in rectal cancer: A meta-analysis. Semin Ultrasound CT MR 26:259–268, 2005. 29. Larsen LP, Rosenkilde M, Christensen H, et al: Can contrast-enhanced ultrasonography replace multidetector-computed tomography in the detection of liver metastases from colorectal cancer? Eur J Radiol 69:308–313, 2009. 30. Maffione AM, Marzola MC, Capirci C, et al: Value of (18) F-FDG PET for predicting response to neoadjuvant therapy in rectal cancer: Systematic review and meta-analysis. AJR Am J Roentgenol 204(6):1261– 1268, 2015. 31. Mainenti PP, Mancini M, Mainolfi C, et al: Detection of colo-rectal liver metastases: Prospective comparison of contrast enhanced US, multidetector CT, PET/CT, and 1.5 Tesla MR with extracellular and reticulo-endothelial cell specific contrast agents. Abdom Imaging 35:511–521, 2010. 32. Matsuoka H, Nakamura A, Masaki T, et al: Preoperative staging by multidetector-row computed tomography in patients with rectal carcinoma. Am J Surg 184:131–135, 2002. 33. Paspulati RM, Partovi S, Herrmann KA, et al: Comparison of hybrid FDG PET/MRI compared with PET/CT in colorectal cancer staging and restaging: A pilot study. Abdom Imaging 40(6):1415–1425, 2015. 34. Purkayastha S, Tekkis PP, Athanasiou T, et al: Diagnostic precision of magnetic resonance imaging for preoperative prediction of the circumferential margin involvement in patients with rectal cancer. Colorectal Dis 9:402–411, 2007. 35. Rao SX, Zeng MS, Xu JM, et al: Assessment of T staging and mesorectal fascia status using high-resolution MRI in rectal cancer with rectal distention. World J Gastroenterol 13(30):4141–4146, 2007. 36. Sinha R, Verma R, Rajesh A, et al: Diagnostic value of multidetector row CT in rectal cancer staging: Comparison of multiplanar and axial images with histopathology. Clin Radiol 61:924–931, 2006. 37. Smith NJ, Barbachano Y, Norman AR, et al: Prognostic significance of magnetic resonance imaging-detected extramural vascular invasion in rectal cancer. Br J Surg 95:229–236, 2008. 38. Teixera SR, Kohan AA, Paspulati RM, et al: Potential role of positron emission tomography/magnetic resonance imaging in gastrointestinal and abdominal malignancies: Preliminary experience. Semin Roentgenol 49(4):321–333, 2014. 39. Videhult P, Smedh K, Lundin P, et al: Magnetic resonance imaging for preoperative staging of rectal cancer in clinical practice: High accuracy in predicting circumferential margin with clinical benefit. Colorectal Dis 9:412–419, 2007. 40. Wicherts DA, de Haas RJ, van Kessel CS, et al: Incremental value of arterial and equilibrium phase compared to hepatic venous phase CT in the preoperative staging of colorectal liver metastases: An evaluation with different reference standards. Eur J Radiol 77:305–311, 2011.

52  Contrast Nephropathy and Its Management Sameh K. Morcos

Since the introduction of multidetector computed tomography (MDCT) technology, the number of patients undergoing contrastenhanced CT studies has steadily increased. In addition the patient population subjected to CT is becoming progressively older with more comorbid conditions. The benefits of contrast-enhanced CT are well recognized in the diagnosis of diseases, evaluation of trauma patients, and guidance of interventional and therapeutic procedures. However, adverse reactions to contrast administration may occur and remain a source of concern, particularly the development of contrast-induced nephropathy (CIN). This complication is defined as an acute deterioration of renal function—an increase of 25% or more or an absolute increase of 0.5 mg/dL (44.2 µmol/L) or more in serum creatinine (Cr) from the baseline value—that occurs within 3 days after administration of iodinated contrast media (CM) in the absence of other causes.41 CIN ranges in severity from an asymptomatic nonoliguric transient form  to oliguric severe acute kidney failure necessitating dialysis. Fortunately most episodes of CIN are self-limited and resolve within 1 to 2 weeks.31,34-36,41 Permanent kidney damage seems to be uncommon. Mild CIN may be underrecognized in clinical practice because serum Cr is not routinely measured after contrast administration, especially in outpatients. However, CIN may increase the risk of nonrenal complications and prolong hospital stays. In one study the mortality among hospital inpatients with CIN was 34% compared with 7% in a matched control group that did not develop CIN.27 Thus prevention of this complication is important to avoid substantial morbidity and mortality.20,27,35,36 In this chapter the risk factors, incidence, pathophysiology, and prevention of CIN are discussed.

RISK FACTORS FOR CONTRAST-INDUCED NEPHROPATHY Preexisting Renal Impairment Patients at highest risk for developing CIN are those with preexisting renal impairment, particularly when reduction in renal function is associated with diabetes mellitus.25,31,34-36,41,55,60 Renal impairment is defined as serum Cr greater than 1.5 mg/dL (130 µmol/L) or a calculated or estimated creatinine clearance (CrCl) less than 60 mL/ min/1.73 m2. To a large extent the degree of renal insufficiency present before administration of contrast material determines the severity of CIN. CrCl of 30 mL/min or less markedly increases the incidence and severity of CIN.31,34-36,41 The importance of diabetes mellitus without renal impairment as a risk factor for CIN remains to be clarified, but some studies have identified diabetes mellitus as an independent risk factor for this complication.31,65 Caution is warranted in diabetic patients because endothelial dysfunction may exist, predisposing the patient to CIN through alterations in the production of vasoactive mediators in the kidney, particularly the vasodilator nitric oxide.1,21,34,39,40

Type of Contrast Medium The type of CM is an important factor in determining the development and severity of CIN. In a meta-analysis of comparative trials, Barrett and Carlisle showed that the incidence of CIN in at-risk patients was significantly higher with high-osmolar contrast agents (>1500 mOsm/ kg) than with low-osmolar agents (<915 mOsm/kg).5 Thus highosmolar CM should be avoided in patients at increased risk of CIN.33,35,36,53 The question remains whether there is a difference in nephrotoxic potential between low-osmolar and iso-osmolar (290 mOsm/kg) agents. A total of nine low-osmolar contrast agents— eight nonionic monomers (iohexol, iomeprol, iopamidol, iopentol, ioxilan, iopromide, ioversol, iobitridol) and one ionic dimer (ioxaglate)—as well as one iso-osmolar nonionic dimer (iodixanol) are approved for intravascular use. In 2003 Aspelin and coworkers, in a study of 129 diabetic patients with preexisting renal impairment undergoing angiography (the NEPHRIC study), reported a 3% incidence of CIN with iodixanol and 26% with the low-osmolar nonionic monomer iohexol.3 In that study the two groups differed significantly with regard to the frequency of interventional procedures and the duration of diabetes but were otherwise comparable. Similarly in patients with baseline renal impairment, one third of whom had diabetes, Chalmers and colleagues observed less reduction in renal function with iodixanol than with iohexol.12 However, a low incidence of CIN with iodixanol was not observed in other studies.56 Stone and associates reported a 33.3% incidence of CIN with iodixanol and 25.3% with other types of low-osmolar agents, but the difference was not statistically significant.58 In two additional studies the incidence of CIN with iodixanol was 12% and 21%.4,9 Several new comparative studies have been carried out since the NEPHRIC study, with mixed results. Two recent reports24,30 were in concordance with the findings of the NEPHRIC study, but none of the others confirmed a lower incidence of CIN with iodixanol than with low-osmolar agents. Jo and coworkers reported a higher incidence of CIN after administration of ioxaglate (25%) in comparison to iodixanol (8.2%) in patients with renal impairment undergoing coronary angiography.24 In a meta-analysis of 2727 patients from 16 doubleblind randomized controlled studies, McCullough and colleagues reported that the incidence of CIN after intraarterial administration was 1.4% with iodixanol and 3.5% with low-osmolar CM.30 Analysis of a subgroup of patients with diabetes mellitus and chronic kidney disease showed that the incidence of CIN was 3.5% with iodixanol and 15.5% with low-osmolar media. The report concluded that renal failure appears to be significantly reduced by the use of iodixanol in patients undergoing angiography.30 However, the study had certain limitations: the dose of CM was significantly higher in the low-osmolar group (iodine 58.9 g/examination) than in the iodixanol group (iodine

1755

1756

PART II  CT and MR Imaging of the Whole Body

55.4 g/examination, P = .003), and no data on the state of hydration were available. In a retrospective analysis of 52,526 patients who received coronary angiography and intervention, the incidence of renal failure within 12 months was significantly higher with iodixanol (1.8%) in comparison to ioxaglate (1%).28 In a study of 285 patients undergoing cardiac angiography receiving either iopamidol or iodixanol-320, Jingwei and associates found no significant difference in the incidence of CIN between the tested agents: 19.4% with iopamidol and 11.7% with iodixanol.23 Barrett et al. reported a 2.6% rate of CIN with iodixanol and 0% with iopamidol in a randomized multicenter trial (IMPACT) of patients with renal impairment (glomerular filtration rate [GFR] ≤ 60 mL/min) undergoing MDCT who received intravenous (IV) equiiodine doses (40 g) of either iopamidol-370 or iodixanol-320 administered at 4 mL/sec.6 Results of the IMPACT study were similar to those of previous smaller studies by Carraro and coworkers11 and Kolehmainen and Soiva26 in patients with chronic kidney disease. Carraro’s group conducted a prospective randomized double-blind comparison of iodixanol and iopromide (nonionic monomer) in 64 patients with moderate to severe renal insufficiency undergoing excretory urography. Renal function was assessed before and 1, 6, 24, and 48 hours and 7 days after the contrast-enhanced examination. One nondiabetic patient in the iodixanol group developed CIN (serum Cr increasing from 2.5 to 5.4 mg/dL in 24 hours, returning to baseline by the 48-hour evaluation), and no cases of CIN were observed in the iopromide group.11 Using the same study design, Kolehmainen and Soiva compared iodixanol to the nonionic monomer iobitridol in 50 patients with severe kidney disease undergoing cranial or body CT procedures. Both groups received similar volumes of contrast (113.3 mL of iobitridol, 112.7 mL of iodixanol) and had similar baseline renal function (CrCl 28.7 mL/min in the iobitridol group, 27.5 mL/min in the iodixanol group). The incidence of CIN in both groups was 17%.26

Dose and Route of Administration The nephrotoxic effect of CM is dose dependent: the higher the dose, the greater the risk of CIN.31,71 In one study it was observed that 100 mL was the cutoff dose below which there was no CIN requiring dialysis.31 Multiple injections of CM within 72 hours in patients with renal impairment are also likely to increase the risk of CIN.36 The clearance of CM from the body is prolonged in patients with reduced renal function, and it is feasible that repeat injections within 72 hours could result in a cumulative renal insult.36,41 Route of administration is also important. Contrast media are less nephrotoxic when administered IV than when given intraarterially in the renal arteries or the aorta proximal to the origin of the renal blood vessels. The acute intrarenal concentration of CM is much higher after intraarterial injection than after IV administration.7,36

Other Risk Factors Other risk factors for CIN may include old age,16 dehydration, congestive cardiac failure, multiple myeloma, concurrent use of nephrotoxic drugs, hypertension, hyperuricemia, and proteinuria.31,34,36,41,55,60,67 Dehydration or congestive cardiac failure can lead to a reduction in renal perfusion, enhancing the ischemic insult of CM.36 Although multiple myeloma has been considered a risk factor for CIN, if dehydration is avoided, administration of low-osmolar CM rarely leads to CIN in patients with myeloma.36,41,55 Concurrent use of nephrotoxic drugs such as nonsteroidal antiinflammatory drugs (NSAIDs) and aminoglycosides is likely to exaggerate the nephrotoxic effects of CM. NSAID use, particularly in large doses or for prolonged periods, can lead to marked renal medullary ischemia, which theoretically can aggravate the ischemic and hypoxic

insult induced by CM in the kidney.21,36,41 Nephropathy is more common in patients with hypertension, hyperuricemia, or proteinuria than in patients without these factors, making them more vulnerable to the renal effects of CM.36

INCIDENCE CIN is unlikely to develop in patients with normal renal function receiving IV injections of CM. The incidence of CIN following IV administration in patients with renal impairment varies from 10% to 20% depending on the severity of the renal disease and the dose of CM.51,54 Following intraarterial injection, the incidence of CIN is approximately 5% to 10% in patients with mild to moderate kidney disease but increases to 50% or more in patients with advanced kidney disease, particularly if associated with diabetes mellitus.31,34,36,41,55,60,67

PATHOPHYSIOLOGY The renal effects of CM that precipitate development of CIN include an increase in renal vascular resistance, a decrease in GFR, natriuresis, and cytotoxic effects on tubular cells.31,34,36,38,39,41,50,55 The pathophysiology of these effects involves activation of the tubuloglomerular feedback (TGF) mechanism and modulation of intrarenal production of vasoactive mediators such as prostaglandins, nitric oxide, endothelin, and adenosine.1,21,34,37-41 The TGF response is osmolality dependent and accounts for about 50% of the acute functional effects of high-osmolar CM on the kidney.37 Reduced synthesis of the endogenous vasodilators nitric oxide and prostaglandins increases the nephrotoxicity of CM.1,21,40 Endothelin and adenosine play a crucial role in mediating the acute functional effects of CM.34,38,41,45-48 Vacuolization of the cells of the proximal tubules and apoptosis and necrosis of the cells of the medullary ascending limbs of the loops of Henle are the main structural effects of CM in the kidney.34,39,62

REDUCING THE RISK OF CONTRASTINDUCED NEPHROPATHY Identifying Patients at Risk The first step in reducing the incidence of CIN is identifying patients at high risk for this complication, particularly those with preexisting renal impairment. The serum Cr level is not a perfect marker for renal function because a greater than 50% reduction in GFR may occur before any rise is observed; however, it is adequate for identifying patients at risk for CIN.36,41,69 Patients with normal serum Cr (≤1.2 mg/ dL [114 µmol/L] for women and ≤1.4 mg/dL [123 µmol/L] for men) have almost no risk of developing CIN.36,41,69 Hence one recommendation is to measure serum Cr before intravascular injection of CM in all patients.35 A normal serum Cr up to 3 months before the examination in patients with no history of renal disease can be reassuring for hospital outpatients, but a recent serum Cr should be obtained for all inpatients.35 Alternatively, serum Cr measurement can be safely limited to those patients with a history of renal disease, proteinuria, prior kidney surgery, hypertension, gout, or diabetes.13 It has also been suggested that for a more accurate assessment of renal function, the measured serum Cr should be used to predict the GFR by means of the Cockcroft-Gault formula or the modification of diet in renal disease (MDRD) formula.14,36,69 Although the CockcroftGault formula provides more accurate information about renal function, it requires information about body weight, which can hinder its routine use in diagnostic imaging departments.36 The MDRD formula, which does not require body weight information, is more practical and is used regularly by many laboratories despite its potential inaccuracy

CHAPTER 52  Contrast Nephropathy and Its Management in patients with GFRs greater than 60 mL/min.36,69 Measuring serum cystatin C might be used in the future as a reliable assessment of renal function.17

Choice and Dose of Contrast Media Recent studies have shown no benefit in using iso-osmolar rather than low-osmolar IV CM in patients with renal impairment.6,11 The lowest possible dose should be used, however. Modern MDCT machines allow very fast scanning, potentially reducing the dose of CM required. MDCT protocols should be designed with the aim of minimizing  the contrast load on the patient. The use of gadolinium-based CM as an alternative to iodinated agents to reduce the risk of CIN is dangerous and should not be considered. Gadolinium agents are potentially more nephrotoxic than iodinated agents at equimolar concentrations.44,63,66 In addition they may induce the serious complication of nephrogenic systemic fibrosis in patients with marked reduction in renal function.64

Hydration (Volume Expansion) Volume expansion is generally accepted as an effective means of reducing the risk of CIN. It improves flow and dilutes the CM within the kidney. In addition it induces diuresis, which is associated with an increase in intrarenal production of prostacyclin, leading to vasodilation in the vulnerable region of the renal medulla.18,19,36,57,68 Hydration can be offered orally or IV. Normal saline given IV for 4 to 6 hours before and after CM exposure at a rate of 1 mL/kg/h is effective in reducing the risk of CIN.36,43,68 Isotonic bicarbonate starting 1 hour before contrast exposure at a rate of 3 mL/kg/h and continuing at a rate of 1 mL/kg/h for 6 hours after contrast exposure was shown in one study to reduce the incidence of CIN compared with isotonic saline hydration.32 Recent reports suggest that combining bicarbonate infusion with N-acetylcysteine (NAC) at the dose of 1200 mg twice a day for 48 hours, starting 24 hours before contrast administration, offers better protection against CIN than receiving 0.9% saline with or without ascorbic acid.10,52 The mechanism by which bicarbonate is protective is not fully understood, but it has been suggested that increasing the pH of the renal medulla and urine may reduce production of free radicals and protect the kidney from the oxidant injury that can be associated with CIN.32 Prolonged IV fluid therapy is a practical problem for outpatients and would require hospital admission overnight.59 Therefore oral hydration can be considered for outpatients; 1000 mL given at a rate of 100 mL/h before and after contrast exposure is a practical approach to outpatients with moderate renal impairment (GFR 45-60 mL/min) receiving 100 mL or less of CM.41 Patients receiving larger doses or with advanced chronic kidney disease (GFR < 45 mL/min) are better hydrated with IV saline.36 In summary, extracellular volume expansion seems to be the most effective measure to prevent CIN.68 The hydration regimen should start before and continue for several hours after CM exposure. Normal saline offers better protection than half-strength saline; however, isotonic bicarbonate may be the fluid of choice if recent findings are reproduced in several new studies.19,36,43,68 The optimal duration and intensity of fluid therapy have not been established. The main limitations of hydration prophylaxis are that it is not suitable for patients with congestive heart failure and that it is impractical in emergency situations when CM must be injected without delay.36,68

Pharmacologic Manipulation In recent years it has been claimed that various drugs may protect the kidney against CIN. However, no single drug has been proved efficacious beyond a doubt, and the results of clinical trials are not

1757

conclusive.35,36 The drug that has received the most attention is NAC, and it is discussed in more detail here. Other drugs, such as theophylline, fenoldopam, and calcium channel blockers, are covered in References 22 and 35. NAC might reduce the nephrotoxicity of CM through antioxidant and vasodilatory effects.36,61 Although the results of the initial trial were impressive, subsequent trials, including a meta-analysis of a large number of studies, showed mixed results.2,8-10,29,49,52,70,72 Recently it was suggested that increasing the dose of NAC from 600 to 1200 mg twice a day for 48 hours starting a day before contrast administration and adding IV hydration with saline could increase the protective effect of NAC.10,36,68 Despite all the published studies, a consistent protective effect of NAC against CIN has not been proved.36,68 In addition, the nephroprotective effect of NAC observed in some studies might be spurious because most studies used serum Cr as a surrogate marker of renal function. A recent study demonstrated that NAC can induce a decrease in the serum Cr level that is independent of alterations in GFR. It was suggested that NAC may enhance tubular secretion of Cr, precipitating a reduction in serum Cr level.36 The report concluded that use of Cr measurements alone to assess renal function in studies investigating the renoprotective effect of NAC is questionable, and direct measurement of GFR or measurement of another marker of renal function (e.g., cystatin C) should be considered.36 Thus although NAC has several attractive advantages, including cost, ease of administration, and limited side effects, further studies are required before its routine use can be endorsed.

BOX 52-1  Guidelines to Avoid Contrast

Medium Nephrotoxicity

Definition • Contrast medium nephrotoxicity is a condition in which an impairment in renal function (increase in serum Cr >25% or >44 µmol/L) occurs within 3 days after intravascular administration of contrast medium in the absence of an alternative cause. Risk Factors • Increased serum Cr levels, particularly secondary to diabetic nephropathy • Dehydration • Congestive heart failure • Age > 70 years • Concurrent administration of nephrotoxic drugs (e.g., NSAIDs) Do’s and Don’ts In Patients with Risk Factors, Do: • Make sure the patient is well hydrated: give at least 100 mL oral (e.g., soft drinks) or IV (normal saline) fluids (depending on the clinical situation) per hour, starting 4 hours before and continuing up to 24 hours after contrast administration; in hot climates, increase the fluid volume. • Use low-osmolar or iso-osmolar contrast media. • Stop administration of nephrotoxic drugs for at least 24 hours. • Consider alternative imaging techniques that do not require administration of iodinated contrast media. In Patients with Risk Factors, Do Not: • Give high-osmolar contrast media. • Administer large doses of contrast media. • Administer mannitol and diuretics (particularly loop diuretics). • Perform multiple studies with contrast media within a short period. Modified from Morcos SK et al: Contrast-media-induced nephrotoxicity: A consensus report. Eur Radiol 9:1602–1613, 1999.

1758

PART II  CT and MR Imaging of the Whole Body

SUMMARY Prevention of CIN relies on careful patient selection. Patients at high risk of CIN who require contrast-enhanced CT should receive the smallest possible dose of either nonionic iso-osmolar dimeric CM or nonionic low-osmolar monomeric CM. Oral hydration and IV infusion of 0.9% saline should be offered, provided there is no contraindication to volume expansion. Prophylactic hemodialysis offers no protection against CIN.15,36,42 The effectiveness of pharmacologic manipulation in preventing CIN remains controversial. Sodium bicarbonate infusion in combination with NAC shows promise and would be suitable for outpatients, because it can be given 1 hour before the procedure and continued for 6 hours after.8,52 However, further studies are required to confirm the effectiveness of this approach. The European Society of Urogenital Imaging guidelines for prevention of CIN are presented in Box 52-1.41,67

REFERENCES 1. Agmon Y, Peleg H, Greenfield Z, et al: Nitric oxide and prostanoids protect the renal outer medulla from radiocontrast toxicity in the rat. J Clin Invest 94:1069–1075, 1994. 2. Alonso A, Lau J, Jaber BL, et al: Prevention of radiocontrast nephropathy with N-acetylcysteine in patients with chronic kidney disease: A meta-analysis of randomized, controlled trials. Am J Kidney Dis 43:1–9, 2004. 3. Aspelin P, Aubry P, Fransson SG, et al: Nephrotoxic effects in high-risk patients undergoing angiography. N Engl J Med 348:491–499, 2003. 4. Baker CS, Wragg A, Kumar S, et al: A rapid protocol for the prevention of contrast-induced renal dysfunction: The RAPPID study. J Am Coll Cardiol 41:2114–2118, 2003. 5. Barrett BJ, Carlisle EJ: Metaanalysis of the relative nephrotoxicity of high- and low-osmolality iodinated contrast media. Radiology 188:171– 178, 1993. 6. Barrett BJ, Katzberg RW, Thomsen HS, et al: Contrast induced nephropathy in patients with chronic kidney disease undergoing computed tomography: A double blind comparison of iodixanol and iopamidol. Invest Radiol 41:815–821, 2006. 7. Berg KJ, Jakobsen JA: Nephrotoxicity related to x-ray contrast media. Adv X-ray Contrast 1:10–18, 1993. 8. Biondi-Zoccai GGL, Lotrionte M, Abbate A, et al: Compliance with QUOROM and quality of reporting over overlapping meta-analyses on the role of acetylcysteine in the prevention of contrast associated nephropathy: Case study. BMJ 332:202–209, 2006. 9. Boccalandro F, Amhad M, Smalling RW, et al: Oral acetylcysteine does not protect renal function from moderate to high doses of intravenous radiographic contrast. Catheter Cardiovasc Interv 58:336–341, 2003. 10. Briguori C, Airoldi F, Dandrea D, et al: Renal Insufficiency Following Contrast Media Administration Trial (REMEDIAL): A randomized comparison of 3 preventive strategies. Circulation 115:1211–1217, 2007. 11. Carraro M, Malalan F, Antonione R, et al: Effects of a dimeric vs a monomeric nonionic contrast medium on renal function in patients with mild to moderate renal insufficiency: A double-blind, randomized clinical trial. Eur Radiol 8:144–147, 1998. 12. Chalmers N, Jackson RW: Comparison of iodixanol and iohexol in renal impairment. Br J Radiol 72:701–703, 1999. 13. Choyke PL, Cady J, DePollar SL, et al: Determination of serum creatinine prior to iodinated contrast media: Is it necessary in all patients? Tech Urol 4:65–69, 1998. 14. Cockroft DW, Gault MH: Prediction of creatinine clearance from serum creatinine. Nephron 16:31–41, 1976. 15. Deray G: Dialysis and iodinated contrast media. Kidney Int 69:S25–S29, 2006. 16. Detrenis S, Meschi M, Bertolini L, et al: Contrast medium administration in the elderly patient: Is advancing age an independent

17.

18.

19. 20.

21.

22.

23.

24.

25.

26. 27. 28.

29.

30.

31.

32.

33.

34. 35. 36.

37.

38.

39.

risk factor for contrast nephropathy after angiographic procedures. J Vasc Interv Radiol 18:277–285, 2007. Drarnidharka VR, Kwon C, Stevens C: Serum cystatin C is superior to serum creatinine as a marker of kidney function: A meta-analysis. Am J Kidney Dis 40:221–226, 2002. Dussol B, Morange S, Loundoun A, et al: A randomized trial of saline hydration to prevent contrast nephropathy in chronic renal failure patients. Nephrol Dial Transplant 21:2120–2126, 2006. Erley CM: Does hydration prevent radiocontrast-induced acute renal failure? Nephrol Dial Transplant 14:1064–1066, 1999. Gruberg L, Mintz GS, Mehran R, et al: The prognostic implications of further renal function deterioration within 48 hours of interventional coronary procedures in patients with pre-existent chronic renal insufficiency. J Am Coll Cardiol 36:1542–1548, 2000. Heyman SN, Rosen S, Brezis M: Radiocontrast nephropathy: A paradigm for the synergism between toxic and hypoxic insults in the kidney. Exp Nephrol 2:153–157, 1994. Idee JM, Lancelot E, Pines E, et al: Prophylaxis of iodinated contrast media-induced nephropathy, a pharmacological point of view. Invest Radiol 39:155–170, 2004. Jingwei N, Ruiyan Z, Jainsheng Z, et al: Safety of isosmolar nonionic dimer during percutaneous coronary intervention. J Interv Radiol 15:327–329, 2006. Jo S-H, Youn T-J, Koo B-K, et al: Renal toxicity evaluation and comparison between Visipaque (iodixanol) and Hexabric (ioxaglate) in patients with renal insufficiency undergoing coronary angiography. The RECOVER study: A randomized controlled trial. J Am Coll Cardiol 48:924–930, 2006. Katzberg WR: Urography into the 21st century: New contrast media, renal handling, imaging characteristics, and nephrotoxicity. Radiology 204:297–312, 1997. Kolehmainen H, Soiva M: Comparison of Xenetix 300 and Visipaque 320 in patients with renal failure. Eur Radiol 13:B32–B33, 2003. Levy EM, Viscoli CM, Horwitz RI: The effect of acute renal failure on mortality, a cohort analysis. JAMA 257:1489–1494, 1996. Liss P, Presson PB, Hansell P, et al: Renal failure in 57,925 patients undergoing coronary procedures using iso-osmolar or low-osmolar contrast media. Kidney Int 70:1811–1817, 2006. Liu R, Nair D, Ix J, et al: N-acetylcysteine for the prevention of contrast-induced nephropathy: A systematic review and meta-analysis.   J Gen Intern Med 20:193–200, 2005. McCullough PA, Bertrand ME, Brinker JA, et al: A meta-analysis of the renal safety of isosmolar iodixanol compared with low osmolar contrast media. J Am Coll Cardiol 48:692–699, 2006. McCullough PA, Wolyn R, Rocher LL, et al: Acute renal failure after coronary intervention: Incidence, risk factors and relationship to mortality. Am J Med 103:368–375, 1997. Merten GJ, Burgess WP, Gray LV, et al: Prevention of contrast-induced nephropathy with sodium bicarbonate: A randomized controlled trial. JAMA 291:2328–2334, 2004. Moore RD, Steinberg EP, Powe NR, et al: Nephrotoxicity of high osmolality versus low osmolality contrast media: Randomized clinical trial. Radiology 182:649–655, 1992. Morcos SK: Contrast media-induced nephrotoxicit—Questions and answers. Br J Radiol 71:357–365, 1998. Morcos SK: Prevention of contrast media nephrotoxicit—The story so far. Clin Radiol 59:381–389, 2004. Morcos SK: Prevention of contrast media-induced nephrotoxicity   after angiographic procedures. J Vasc Interv Radiol 16:13–23, 2005. Morcos SK, Brown PWG, Oldroyd S, et al: Relationship between the diuretic effect of radiocontrast media and their ability to increase renal vascular resistance. Br J Radiol 68:850–853, 1995. Morcos SK, El Nahas AM: Advances in the understanding of the nephrotoxicity of radiocontrast media. Nephron 78:249–252, 1998. Morcos SK, Epstein FH, Haylor J, et al: Aspects of contrast media nephrotoxicity. Eur J Radiol 23:178–184, 1996.

CHAPTER 52  Contrast Nephropathy and Its Management 40. Morcos SK, Oldroyd S, Haylor J: Effect of radiographic contrast media on endothelium derived nitric oxide-dependent renal vasodilatation. Br J Radiol 70:154–159, 1997. 41. Morcos SK, Thomsen HS, Webb JA, et al: Contrast-media-induced nephrotoxicity: A consensus report. Eur Radiol 9:1602–1613, 1999. 42. Morcos SK, Thomsen HS, Webb JAW, et al: Dialysis and contrast media. Eur Radiol 12:3026–3030, 2002. 43. Mueller C, Buerkle G, Buettner HJ, et al: Prevention of contrast media-associated nephropathy: Randomized comparison of 2 hydration regimens in 1620 patients undergoing coronary angioplasty. Arch Intern Med 162:329–336, 2002. 44. Nyman U, Elmståhl B, Leander P, et al: Are gadolinium-based contrast media really safer than iodinated media for digital subtraction angiography in patients with azotemia? Radiology 223:311–318, 2002. 45. Oldroyd SD, Fang L, Haylor JL, et al: Effects of adenosine receptor antagonists on the responses to contrast media in the isolated rat kidney. Clin Sci 98:303–311, 2000. 46. Oldroyd SD, Haylor JL, Morcos SK: Bosentan, an orally active endothelin antagonist: Effect on renal response to contrast media. Radiology 196:661–665, 1995. 47. Oldroyd SD, Haylor JL, Morcos SK: The acute effect of ioversol on kidney function: Role of endothelin. Eur J Radiol 19:91–95, 1995. 48. Oldroyd SD, Slee SJ, Haylor J, et al: Role for endothelin in the renal response to radiocontrast media in the rat. Clin Sci 87:427–434, 1994. 49. Pannu N, Manns B, Lee H, et al: Systematic review of the impact of N-acetylcysteine on contrast nephropathy. Kidney Int 65:1366–1374, 2004. 50. Persson PB, Tepel M: Contrast medium-induced nephropathy: The pathophysiology. Kidney Int 69:S8–S10, 2006. 51. Reddan D, Fishman EK: Radiologists’ knowledge and perceptions of the impact of contrast-induced nephropathy and its risk factors when performing computed tomography examinations: A survey of European radiologists. Eur J Radiol 66:235–245, 2008. 52. Redio-Mayoral A, Chaparro M, Prado B, et al: The reno-protective effect of hydration with sodium bicarbonate plus N-acetylcysteine in patients undergoing emergency percutaneous coronary intervention: The RENO study. J Am Coll Cardiol 49:1283–1288, 2007. 53. Rudnick MR, Goldberg S, Wexler L, et al: Nephrotoxicity of ionic and non-ionic contrast media in 1196 patients: A randomized trial. Kidney Int 47:254–261, 1995. 54. Sandstede JJW, Roth A, Machann W, et al: Evaluation of nephrotoxicity of iodixanol in patients with predisposing factors to contrast medium induced nephropathy referred for contrast enhanced computed tomography. Eur J Radiol 63:120–123, 2007. 55. Solomon R: Contrast medium-induced acute renal failure. Kidney Int 53:230–242, 1998. 56. Solomon R, Natarajan M, Doucet S, et al: The Cardiac-Angiography in Renally Impaired Patients (CARE) study: A randomized, double blind

57.

58.

59.

60. 61.

62.

63. 64. 65.

66.

67.

68.

69.

70.

71.

72.

1759

trial of contrast-induced nephropathy (CIN) in high risk patients. Circulation 115:3189–3196, 2007. Solomon R, Werner C, Mann D, et al: Effects of saline, mannitol and furosemide on acute decrease in renal function induced by radiocontrast agents. N Engl J Med 331:1416–1420, 1994. Stone GW, McCullough PA, Tumlin JA, et al: Fenoldopam mesylate for the prevention of contrast-induced nephropathy: A randomized controlled trial. JAMA 290:2284–2291, 2003. Taylor AJ, Hotchkiss D, Morse RW, et al: PREPARED: Preparation for Angiography in Renal Dysfunction: A randomized trial of inpatient vs outpatient hydration protocols for cardiac catheterization in mild-tomoderate renal dysfunction. Chest 114:1570–1574, 1998. Tepel M, Aspelin P, Lameire N: Contrast-induced nephropathy: A clinical and evidence based approach. Circulation 113:1799–1806, 2006. Tepel M, van der Giet M, Schwarzfeld C, et al: Prevention of radiographic-contrast-agent-induced reductions in renal function by acetylcysteine. N Engl J Med 343:180–184, 2000. Tervahartiala P, Kivisaari L, Kivisaari R, et al: Contrast media-induced renal tubular vacuolisation: A light and electron microscopic study on rat kidneys. Invest Radiol 26:882–887, 1991. Thomsen HS: Gadolinium-based contrast media may be nephrotoxic even at approved doses. Eur Radiol 14:1654–1656, 2004. Thomsen HS: Nephrogenic systemic fibrosis: A serious late adverse reaction to gadodiamide. Eur Radiol 16:2619–2621, 2006. Thomsen HS: Non-insulin dependent diabetes and contrast media. In Thomsen HS, editor: Contrast media: Safety issues and ESUR guidelines, Heidelberg, 2006, Springer, pp 35–46. Thomsen HS, Almén T, Morcos SK, et al: Gadolinium-containing contrast media for radiographic examinations: A position paper. Eur Radiol 12:2600–2605, 2002. Thomsen HS, Morcos SK: Contrast media and the kidney: European Society of Urogenital Radiology (ESUR) guidelines. Br J Radiol 76:513–518, 2003. Thomsen HS, Morcos SK: Contrast-medium-induced nephropathy: Is there a new consensus? A review of published guidelines. Eur Radiol 16:1835–1840, 2006. Thomsen HS, Morcos SK, et al: In which patients should serumcreatinine be measured before contrast medium administration? Eur Radiol 15:749–756, 2005. Vaitkus PT, Brar C: N-acetylcysteine in the prevention of contrast induced nephropathy: Publication bias perpetuated by meta-analyses. Am Heart J 153:275–280, 2007. Wang Y-X, Jia Y-F, Chen K-M, et al: Radiographic contrast media induced nephropathy: Experimental observations and the protective effect of calcium channel blockers. Br J Radiol 74:1103–1108, 2001. Zagler A, Azadpour M, Mercado C, et al: N-acetylcysteine and contrastinduced nephropathy: A meta-analysis of 13 randomized trials. Am Heart J 151:140–145, 2006.

53  Adrenal Glands Lisa M. Ho

INTRODUCTION In 1564, Bartolomeo Eustachius (1520-1574) was the first anatomist to give a detailed description of the adrenal glands, which he depicted in a series of drawings on copper plates.68 However, the function of the adrenal glands remained a mystery for the next 3 centuries. In 1855, Thomas Addison (1793-1860) described the clinical syndrome of adrenal insufficiency in his monograph titled, “On the Constitutional and Local Effects of Disease of the Suprarenal Capsules.” In this work, Addison describes 11 patients who presented with “anæmia, general languor and debility, remarkable feebleness of the heart’s action, irritability of the stomach, and a peculiar change of colour in the skin, occurring in connexion with a diseased condition of the ‘supra-renal capsules’ ”.3 It was this monograph that spurred rigorous scientific investigation, which eventually led to discovery of the adrenal glands’ true function. Since the time of Addison, there has been extraordinary progress in understanding the structure and function of the adrenal glands. More recently, advances in imaging technology have led to significant improvements in the detection and diagnosis of adrenal pathology. Imaging now plays a critical role in the workup and clinical management of adrenal disease. The purpose of this chapter is to (1) describe the imaging appearance of common and uncommon adrenal diseases, (2) explain the current imaging techniques used to analyze, characterize, and diagnose adrenal diseases, and (3) provide a diagnostic imaging algorithm for the frequently encountered clinical dilemma of the “adrenal incidentaloma.”

NORMAL ANATOMY AND IMAGING APPEARANCE The adrenal glands are retroperitoneal structures that are located adjacent to the kidneys and enclosed within the perinephric fascia. Normal adrenal glands have a characteristic inverted Y, V, or T shape (Fig. 53-1). The right adrenal gland is typically superior to the right kidney, and the left adrenal gland is usually anterior to the superior pole of the left kidney. The adrenals are typically described as having a central body and two (medial and lateral) limbs. The body and limbs are typically smoothly shaped and measure less than 10 mm in thickness160; the limbs can measure up to 4 cm in length (see Fig. 53-1). In patients with congenital renal agenesis, the ipsilateral adrenal may have a disc shape with a parasagittal orientation81 (Fig. 53-2). The adrenal glands are composed of two layers. The outer layer, the adrenal cortex, is derived from the mesoderm. The inner layer, the adrenal medulla, originates from the neural crest. The adrenal cortex secretes cortisol, aldosterone, and androgens. The adrenal medulla secretes epinephrine and norepinephrine.102 The adrenal cortex and medulla cannot be differentiated with imaging.

1760

The adrenal glands are routinely visualized on cross-sectional imaging examinations. For adult patients, computed tomography (CT), magnetic resonance imaging (MRI), and fluorine-18 fluorodeoxyglucose (18F FDG) positron emission tomography (PET)/CT are superior to ultrasound (US) for identification of the adrenals.1 In children, US may be used for initial evaluation of the adrenal glands, but frequently CT or MRI is required for complete visualization. On unenhanced CT the adrenals appear similar in density to the liver. With contrast enhancement, the adrenals may enhance quickly in the early hepatic arterial phase and then fade slightly in the portal venous phase to appear similar in density to the liver. Intense contrast enhancement of the adrenal glands has been reported in adults and children in association with acute hypovolemic states such as shock associated with sepsis or acute traumatic blood loss37,115,157 (Fig. 53-3). On MRI the adrenals appear similar in signal to normal liver on T1-weighted images and mildly hyperintense to normal liver on fatsuppressed T2-weighted images. Following contrast administration using a gadolinium-based contrast medium, the adrenals demonstrate rapid increase in signal intensity on T1-weighted images, followed by relatively slower de-enhancement107,144 (see Fig. 53-1). Decreased signal on T2*-weighted gradient-recalled echo (GRE) images within the adrenals due to susceptibility artifact has been described in patients treated with ferumoxytol59 (Fig. 53-4). Ferumoxytol is an iron-containing medication used to treat iron deficiency anemia in patients with chronic renal insufficiency. It also has an off-label use as an alternative MRI contrast agent in selected patients with severe renal impairment.10 Susceptibility artifact can be seen within the adrenals at least 48 hours after ferumoxytol administration, and the effect could potentially persist for several months.59,149 The exact mechanism by which ferumoxytol is retained within the adrenals is not known. This phenomenon is not associated with any known specific clinical symptoms. With 18F FDG PET/CT, the normal adrenals may show activity ranging from the same as background in one third of patients to slightly greater than background in two thirds of patients.9 Rarely some normal adrenals may have activity slightly greater than liver activity.9

ADRENAL IMAGING TECHNIQUES Adrenal Protocol Computed Tomography Multidetector-row CT (MDCT) scanners are used for adrenal protocol CT imaging. Scanning with 64-row MDCT is usually performed using the following parameters: 120 or 140 kVp, 200 to 300 mAs; 0.5-second second gantry revolution time, 4-cm beam width, 1.2 to 1.5 pitch, and 0.25-cm reconstruction thickness with standard algorithm. Images are acquired through the bilateral adrenal glands. Unenhanced CT is performed first, and the radiologist reviews the images. If one or more

CHAPTER 53  Adrenal Glands

A

1761

C

B

FIG 53-1  Normal adrenal glands. Axial (A) and coronal (B) contrast-enhanced CT demonstrates typical location and appearance of the right and left adrenal glands. C, Axial contrast-enhanced T1-weighted MRI shows normal thickness of the adrenal body and limbs (<1 cm) and normal length (<4 cm).

A

B

FIG 53-2  Discoid adrenal gland. Axial (A) and coronal (B) contrast-enhanced CT demonstrates a flattened linear left adrenal gland (arrow) that has a parasagittal orientation and disc shape on sagittal view (not shown), which occurs as a result of ipsilateral (left) ectopic kidney located in the pelvis.

adrenal mass(es) is/are present, an elliptical region-of-interest measurement is drawn as large as possible within each adrenal nodule, carefully avoiding any calcification, and the mean attenuation value of the adrenal nodule is determined. If the attenuation measurement is +10 Hounsfield units (HU) or less, then no further imaging is performed. If the attenuation measurement is greater than +10 HU, then repeat imaging is performed with dynamic contrast enhancement, and also delayed images are performed. Contrast enhancement is performed with 100 mL of nonionic contrast material (300 mg iodine per milliliter [300 mg I/mL]) injected intravenously with a power injector at a rate of 3 mL/sec. MDCT images are acquired using identical technical parameters to the unenhanced phase, following a 65- to 70-second delay after initiating contrast medium injection. At 15 minutes after contrast medium injection, MDCT images are again acquired using identical technical parameters.

Dual-Energy CT Dual-energy CT (DECT) is a technique that allows nearly simultaneous acquisition of CT data sets at different tube voltages (commonly, 80 and 140 kVp) during a single breath-hold scan. Because the attenuation coefficients of substances in the human body vary with applied tube voltage, images acquired with DECT scanning can be analyzed to determine the presence of specific substances in human tissues, including fat, water, calcium, and iodine.60,100,142 DECT has the potential to improve the workup of adrenal masses by enabling characterization of adrenal nodules on the initial contrast-enhanced CT and thus obviating the need for additional imaging. One method subtracts iodine from contrast-enhanced DECT to generate “virtual” noncontrast (VNC) images that can be used to replace true unenhanced CT images. VNC data sets can then be used for characterization of adrenal nodules

1762

PART II  CT and MR Imaging of the Whole Body

with CT densitometry. Another method uses material decomposition analysis to generate material density basis pairs, which can quantify the proportions of materials such as lipid and water within adrenal nodules and enable differentiation of adenomas from metastases.

Adrenal Protocol Magnetic Resonance Imaging Chemical shift MRI is the key imaging technique for adrenal protocol MRI. Chemical shift imaging is performed with in-phase and opposedphase T1-weighted breath-hold GRE sequences. A phased-array body coil is centered on the upper abdomen. To reduce flow-related artifact, presaturation pulses can be applied above and below the imaging volume. Field of view is dependent on patient size. A typical adrenal MRI protocol employs the following sequences: three-plane localizer with large field of view, coronal breath-hold half-Fourier acquisition single-shot turbo-spin echo (HASTE), breath-hold axial and coronal T1-weighted in-phase and opposed-phase GRE using 256 × 192 matrix and 1 NEX, 4-mm slice thickness and 0-mm gap (axial) and 3-mm

slice thickness and 0-mm gap (coronal), axial breath-hold HASTE with fat saturation, and axial T1-weighted volumetric interpolated breathhold examination (VIBE) with fat saturation.

Nuclear Medicine Imaging There are two main types of nuclear medicine examination used with adrenal imaging. The traditional study uses iodine-123 (123I)-labeled metaiodobenzylguanidine (MIBG). MIBG is an analogue of norepinephrine and therefore is used to detect tumors originating from the adrenal medulla, most commonly adrenal pheochromocytomas. It can also be used to localize extraadrenal paragangliomas and metastatic disease from malignant pheochromocytomas. Another nuclear medicine imaging exam employs fluorine-18 fluorodeoxyglucose (18F FDG), a radiolabeled d-glucose analogue that is used with positron emission tomography (PET). FDG PET can detect tissue with high metabolic activity and is therefore sensitive at distinguishing malignant from benign disease. Simultaneous coregistration with CT imaging (FDG PET/CT) allows for fusion of functional and anatomic data.

Ultrasound US imaging of the adrenal glands is primarily limited to the pediatric population, owing to the lack of ionizing radiation; furthermore the small body habitus of children allows for better visualization of retroperitoneal structures. However, because of the relative nonspecific appearance of adrenal masses with US and the risk of missing an adrenal mass because of the operator-dependent nature of US imaging, most pediatric radiologists have a low threshold for referring children with suspected adrenal pathology for further imaging with CT or MRI.

Adrenal Venous Sampling

FIG 53-3  Shock adrenals. Coronal contrast-enhanced CT shows bilateral intensely hyperenhancing adrenal glands associated with septic shock due to numerous septic emboli in the liver, spleen, and kidneys.

A

Adrenal venous sampling is an invasive procedure where the right and left adrenal veins are cannulated via a transfemoral or transjugular approach under fluoroscopic guidance. Blood samples of the adrenal veins are obtained, usually to detect the presence of a hormonally active adrenal tumor, such as a hyperfunctioning adrenal adenoma. Additional blood samples are obtained from the inferior vena cava below the level of the adrenal veins to be used as controls. Most commonly this procedure is used to confirm the presence of an aldosteronesecreting adrenal adenoma initially detected on cross-sectional imaging prior to surgical resection.

B FIG 53-4  Ferumoxytol uptake in the normal adrenals. Axial in-phase T1-weighted GRE image (A) demonstrates profound signal loss within the bilateral adrenal glands (arrows) when compared with opposed-phase T1-weighted GRE image (B), indicative of magnetic susceptibility artifact related to ferumoxytol administration 3 weeks prior to the MRI examination. Note the similar effect of ferumoxytol uptake in the liver.

CHAPTER 53  Adrenal Glands

PATHOLOGY Adenomas The most common tumor of the adrenal gland is the adrenal adenoma. The overall prevalence of adenomas in the general population is estimated at approximately 4% to 6%, and the frequency increases with age.85,169 The majority of adrenal adenomas are nonhyperfunctioning, are clinically silent, and do not require further treatment.169 They are usually discovered as an incidental finding on imaging performed for indications other than evaluation for adrenal disease.169 On the basis of imaging criteria alone, no distinction can be made between nonhyperfunctioning and hyperfunctioning adenomas133,169; clinical evaluation and biochemical testing are needed to differentiate the two.133 Hyperfunctioning adrenal adenomas are treated with surgery or medical therapy depending on the patient’s clinical condition.169

Unenhanced CT of Adrenal Adenomas.  The adrenal adenoma usually appears as a small (<3 cm), smooth, and round or oval mass originating from the adrenal gland. Histologic examination reveals the presence of abundant intracellular cytoplasmic lipid within the typical adrenal adenoma.35 Korobkin et al.87 demonstrated that an inverse linear relationship exists between the percentage of lipid-rich cortical cells in the adrenal adenoma and the unenhanced CT attenuation number.87 Based on a meta-analysis of 10 studies that examined the unenhanced CT attenuation value of adrenal adenomas, Boland et al.29 found that using a threshold attenuation value of +10 HU or less on unenhanced CT resulted in a sensitivity of 71% and specificity of 98% for diagnosis of an adenoma. This threshold unenhanced CT attenuation value of +10 HU or less has become the reference standard for diagnosis of a lipid-rich adenoma (Fig. 53-5). However, approximately 30% of adrenal adenomas do not have abundant intracellular cytoplasmic lipid, and therefore these “lipid-poor” adenomas will measure greater than +10 HU on unenhanced CT imaging. Based

on attenuation characteristics alone at unenhanced CT, one usually cannot distinguish lipid-poor adenomas from other adrenal tumors such as a primary or secondary adrenal malignancy. Additional imaging tests or imaging follow-up to establish size stability is often needed to distinguish lipid-poor adenomas from other adrenal tumors. CT histogram analysis has been described as one method to improve the sensitivity of unenhanced CT for diagnosis of lipid-poor adenomas.8,62,70,77,131 Using a threshold of greater than 10% negative pixels on unenhanced CT, 46% to 52% of lipid-poor adenomas could be correctly diagnosed as benign adenomas on unenhanced CT.62,70 However, histogram analysis has been shown to become less sensitive when the unenhanced CT attenuation of the adrenal nodule measures greater than 20 HU.63,77

Enhanced CT of Adrenal Adenomas.  Adrenal adenomas demonstrate rapid enhancement and de-enhancement following intravenous (IV) contrast administration. This characteristic perfusion dynamic is independent of lipid content, and therefore this feature can be used to differentiate lipid-poor adenomas from other adrenal tumors (e.g., adrenal metastases, which usually have relatively slow enhancement and de-enhancement).86 The rate of de-enhancement of an adrenal adenoma can be ascertained by performing a dedicated adrenal protocol CT with IV contrast and absolute percentage washout (APW) calculation or relative percentage washout (RPW) calculation.34,123,136 APW is calculated by first determining the mean CT attenuation of the adrenal lesion on unenhanced CT. Next, the mean CT attenuation is measured on enhanced CT (65- to 70-second delay, portal-venous phase imaging) and delayed CT (15-minute delay, washout-phase imaging). APW is calculated by using the following formula: enhanced attenuation (E) minus delayed attenuation (D) divided by enhanced attenuation (E) minus unenhanced attenuation (U), or APW = E − D E − U

Aldosterone levels RT = 32378 ng/dL LT = 107 ng/dL

A

1763

B FIG 53-5  Conn’s syndrome in a 42-year-old woman who presented with hypertension and hypokalemia. A, Axial unenhanced CT shows a small (1.7 cm) smooth, oval-shaped, right adrenal mass of low density (−1.2 HU), consistent with a lipid-rich adrenal adenoma. B, Adrenal vein sampling was performed, which confirmed aldosterone-producing right adenoma with an elevated aldosterone level in the right adrenal vein, compared with normal aldosterone level in the left adrenal vein. This case also illustrates that there is no difference in CT appearance between hyperfunctioning and nonhyperfunctioning adrenal adenomas.

1764

PART II  CT and MR Imaging of the Whole Body

If no unenhanced-phase CT is performed, then RPW can be calculated using the following formula: enhanced attenuation (E) minus delayed attenuation (D) divided by enhanced attenuation, or RPW = E − D E Adenomas (lipid-rich and lipid-poor) will demonstrate greater than 60% APW and greater than 40% RPW when the delayed CT is performed at 15 minutes (Fig. 53-6). Although there have been some reports in the literature that propose using a 10-minute delayed CT to calculate washout,19,91 the sensitivity has been shown to be lower than when a 15-minute delayed CT is used.136 The sensitivity and specificity for APW using a 15-minute delayed CT is 86% to 89% and 92% to 95% for lipid-poor adenomas.33,34 Despite the high specificity of APW for diagnosis of lipid-poor adenomas, recent studies have found that certain types of nonadenomatous hypervascular tumors may also demonstrate greater than 60% APW and mimic the dynamic perfusional behavior of adenomas. Patel et al.121 reported that in their case series, 33% of pheochromocytomas met traditional washout criteria (>60% APW or >40% RPW) and could be misclassified as a lipid-poor adenoma. Similarly, Choi et al.38 found that 84% of hepatocellular

A

B

carcinoma (HCC) or renal cell carcinoma (RCC) adrenal metastases demonstrated greater than 60% APW, which suggests that these hypervascular malignant tumors have the potential to be misclassified as benign adenomas based on calculation of APW.

Dual-Energy CT.  DECT has the potential to improve the workup of incidental adrenal nodules compared with conventional single-energy MDCT.58,100 Several studies have shown that CT densitometry of adrenal nodules can be performed on VNC data sets reconstructed from contrast-enhanced DECT, thus replacing the need for true noncontrast imaging56-58,65,69 (Fig. 53-7). One study, however, found that VNC attenuation values of adrenal nodules were significantly higher than true noncontrast measurements.84 In that study, failure of the iodine-subtraction technique may be related to the enhancement phase chosen for reconstruction of VNC data sets.56 Another way DECT may be applied to characterization of adrenal nodules uses material decomposition analysis to identify the presence of lipid. Using this technique the presence of lipid within adenomas can be detected by generating material density basis pairs of fat-water, iodine-fat, or iodine-water.56,106 Both VNC images and material density analysis are

C

FIG 53-6  Lipid-poor adenoma diagnosed by contrast-enhanced CT with washout calculation. A, Axial unenhanced CT shows a left adrenal nodule that measures 20 HU. B, Axial contrast-enhanced CT shows the same left adrenal nodule, which measures 70 HU. C, Delayed (15 minutes after contrast injection) contrastenhanced CT shows that the same left adrenal nodule now measures 27 HU. Using the formula for calculation of absolute percentage washout, the calculated washout is 86%, which is greater than the threshold value of 60%, consistent with a lipid-poor adenoma.

A

B

C

FIG 53-7  Lipid-rich adenoma diagnosed by contrast-enhanced DECT. A, Axial contrast-enhanced DECT shows an incidentally discovered left adrenal nodule. B, Axial contrast-enhanced DECT after iodine subtraction. The “virtual” noncontrast image demonstrates the same left adrenal nodule (arrow) measures +4 HU, consistent with a diagnosis of lipid-rich adenoma. C, Axial true unenhanced CT shows the same left adrenal nodule, which measures −2 HU, suggesting that virtual noncontrast images created from contrast-enhanced DECT could be used to replace true unenhanced CT.

CHAPTER 53  Adrenal Glands

A

1765

B

FIG 53-8  Lipid-rich adenoma diagnosed by chemical shift MRI. A, Axial in-phase T1-weighted GRE demonstrates left adrenal nodule that is isointense to the spleen. B, Axial opposed-phase T1 GRE demonstrates the same left adrenal nodule, which demonstrates diffuse signal loss, indicating the presence of intracellular lipid, consistent with an adenoma.

new techniques that have the potential to obviate the need for additional imaging when an incidental adrenal nodule is detected on contrast-enhanced DECT.

MRI.  Chemical shift imaging (CSI) is the mainstay of MRI evaluation of solid adrenal masses. This imaging technique is performed with in-phase and opposed-phase T1-weighted breath-hold GRE. Because fat protons precess at a lower frequency than water protons, at a given time to echo (TE) their frequency signals will either be additive (in-phase) or cancel each other out (opposed-phase). Thus, owing to the presence of intracellular lipid, adenomas can be differentiated from nonadenomas (e.g., metastases) by demonstrating signal cancellation on opposed-phase images. The choice of TE to perform chemical shift imaging depends on the strength of the external magnetic field. At 1.5 teslas (T), a TE of 2.2 milliseconds corresponds to when fat and water are opposed-phase, and a TE of 4.4 milliseconds corresponds to when fat and water are in-phase. As with unenhanced CT imaging, there is also an inverse linear relationship between the percentage of lipid-rich cells and the relative change in MRI signal intensity of an adrenal adenoma on chemical shift imaging.87 Because of the abundant amounts of intracellular lipid within the typical adenoma, the signal loss on opposed-phase imaging is usually dramatic, and signal “dropout” is easily determined visually by comparing the in-phase signal images to the opposed-phase signal images (Fig. 53-8). However, if the adenoma does not contain large quantities of intracellular lipid, it may be necessary to quantify the degree of signal loss.103 There are two methods for quantitative analysis. One method is called the signal intensity index (SII). With this method, the signal value of the adrenal nodule is determined on the opposed-phase (O), and this value is subtracted from the in-phase signal value (I), which is then divided by the in-phase signal value (I): % SII = (I − O I) × 100 Fujiyoshi et al. found that an SII of greater than 16.5% indicates that the adrenal lesion is an adenoma.54 However, a range of SII thresholds are reported in the literature, depending on the strength of the external magnetic field (1.5 vs. 3 T) and the technical MRI parameters employed.101,109,128,138 An SII threshold of greater than 20% for diagnosis of an adenoma is commonly used in clinical practice.12,162 Another method for quantitative analysis is the adrenal-spleen ratio. With this method, the signal value of the adrenal nodule on the

opposed-phase (Ao) is divided by the signal value of the spleen on the opposed-phase (So), and then this result is divided by the result of dividing the signal value of the adrenal nodule on the in-phase (Ai) by the signal value of the spleen on the in-phase (Si): Adrenal-spleen = ( Ao So) ( Ai Si) An adrenal-spleen ratio of less than 0.71 indicates the adrenal lesion is an adenoma.117 The accuracy of qualitative (visual) versus quantitative methods has been shown to be not significantly different.103 The overall sensitivity and specificity of chemical shift imaging for differentiating adenomas from metastases has been reported to range from 81% to 100% and 94% to 100%, respectively.16,88,103,117,154 Recent reports indicate that the accuracy of chemical shift imaging is dependent on the lipid content of the adenoma, with lower sensitivity for lipid-poor adenomas. Studies have shown that adrenal adenomas that measure greater than 20 to 30 HU on unenhanced CT are less likely to demonstrate sufficient signal dropout on opposed-phase imaging.61,73,141 The sensitivity of chemical shift imaging for lipid-poor adenomas is reported to be 100% when the unenhanced CT attenuation is 10 to 20 HU61,141 and 75% when the unenhanced attenuation is 20 to 30 HU.61 In a recent study by Seo et al.141 the sensitivity of chemical shift imaging dropped to 64% for adenomas that measured greater than +20 HU on unenhanced CT. These results suggest that adrenal nodules that measure greater than +20 HU on unenhanced CT should be further characterized with a method that is independent of lipid content, such as contrast-enhanced CT with washout calculation. 18

F FDG PET/CT.  18F fluorodeoxyglucose is a radiolabeled d-glucose

analogue that is preferentially taken up by tissues with increased glucose metabolism. Because tumor cells often have high metabolic rates, FDG PET can be used for detection of metastatic disease. When FDG PET is coregistered with CT for anatomic data, combined FDG PET/CT is a powerful tool for cancer staging and follow-up. Adrenal metastases typically have substantial FDG uptake and can be detected with FDG PET/CT with high accuracy48,105 (Fig. 53-9). However, a small percentage (<5%) of adenomas can also demonstrate mild FDG uptake, and distinguishing between mildly FDG-avid adenomas and metastases can be challenging39 (Fig. 53-10). Studies have shown that the degree of FDG uptake in adenomas is either the same or less than the liver, or there is no uptake at all.9,28,31,48,105 By comparison, metastases usually demonstrate high FDG avidity and will have FDG uptake

1766

PART II  CT and MR Imaging of the Whole Body Conn’s Syndrome

FIG 53-9  Left adrenal metastasis from lung cancer on 18F FDG PET/ CT. Coronal fused 18F FDG PET/CT image demonstrates left adrenal nodule with high FDG activity, with SUVmax value of 10 HU. Diagnosis of metastasis was confirmed by biopsy.

FIG 53-10  Right adrenal adenoma by 18F FDG PET/CT. Axial fused 18F FDG PET/CT image of a right adrenal nodule shows mild FDG activity, with SUVmax value of 1.9. Diagnosis of adenoma was confirmed by biopsy.

higher than the liver. Quantitative studies have found that using a maximum standard uptake value (SUVmax) of greater than 3.1 or standard uptake ratio (SUR, which is the SUVmax of the adrenal nodule divided by the mean SUV of the liver) greater than 1 also separates adrenal metastases from adrenal adenomas.21,27,28,31,50,105 In a metaanalysis of 21 studies, Boland et al.28 found that FDG PET/CT has a mean sensitivity and specificity of 97% and 91% for diagnosis of adrenal metastatic disease, with no significant differences between qualitative (visual) analysis and quantitative analysis, which includes using SUVmax and SUR. Both quantitative and qualitative methods of analysis have been shown to have high interreader agreement.50 The CT portion of the FDG PET/CT can be performed with or without IV contrast,166 depending on the patient’s clinical situation. If the CT portion of the FDG PET/CT is performed without IV contrast, then adrenal nodules can be characterized using CT densitometry or histogram analysis (as described previously). When CT densitometry or histogram analysis is used in conjunction with FDG PET/CT, this combined analysis yields improved overall accuracy for distinguishing metastases from adenomas.31,124

Primary hyperaldosteronism, also known as Conn’s syndrome, is a disease characterized by excessive secretion of aldosterone and suppressed renin activity, leading to hypertension and usually hypokalemia. Primary hyperaldosteronism is caused by either a unilateral aldosterone-producing adenoma or by bilateral adrenal hyperplasia.122 Rarely it can be due to a hyperfunctioning adrenal cortical carcinoma. Because aldosterone-producing adenomas are treated surgically and adrenal hyperplasia is managed medically,122,169 distinguishing between these two entities requires imaging.97,98,122,168 The initial imaging test obtained is most often an adrenal protocol CT. Using a thin-slice  (2.0-2.5 mm) reconstruction thickness is important because most aldosterone-producing adenomas measure less than 2 cm, with a significant proportion less than 1 cm in size122 (see Fig. 53-5). Although MRI has been shown to be equally effective as CT at detecting aldosterone-producing adenomas,97 most centers prefer CT because of its generally higher spatial resolution. Detection of a unilateral adrenal adenoma on cross-sectional imaging is highly suggestive of the diagnosis of Conn’s syndrome due to an aldosterone-producing adenoma (see Fig. 53-5). Bilaterally diffusely enlarged adrenal limbs in the absence of a discrete adrenal nodule can be seen with primary hyperaldosteronism due to adrenal hyperplasia.98 Following identification of a unilateral adrenal adenoma, additional confirmatory testing with adrenal vein sampling is selectively recommended before proceeding to surgery, particularly in patients older than 40 years, because these patients have a higher propensity to develop an incidental nonhyperfunctioning adenoma.168 Adrenal vein sampling is performed by obtaining blood samples directly from the left and right adrenal veins and testing the samples for unilateral elevation of aldosterone levels that corresponds to the side of the adrenal adenoma detected on crosssectional imaging42,153 (see Fig. 53-5). Typically the left adrenal vein drains into the left renal vein and the right renal vein drains into the inferior vena cava. Accessing the right renal vein can be challenging and can lead to technical failure of the procedure if it cannot be successfully sampled.42,153

Cushing’s Syndrome Cushing’s syndrome is the clinical presentation of hypercortisolism, which is most frequently iatrogenic. Endogenous Cushing’s syndrome is relatively uncommon and may be adrenocorticotropic hormone (ACTH)-dependent in 80% to 85% of cases or ACTH-independent in 15% to 20% of cases.112 ACTH-dependent hypercortisolism is due to either an ACTHsecreting pituitary adenoma, also known as Cushing’s disease (75%85% of cases), or rarely due to an ectopic source of ACTH. Adrenal hyperplasia due to an ACTH-producing pituitary adenoma or ectopic source of ACTH can have variable appearance, but most frequently it appears as diffuse bilateral enlargement without focal mass, also called nodular hyperplasia76,146 (Fig. 53-11). ACTH-independent Cushing’s syndrome is almost always caused by primary adrenal disease, most commonly either a hyperfunctioning cortisol-producing adrenal adenoma or adrenal cortical carcinoma. The diagnosis of ACTH-independent Cushing’s syndrome is primarily based on clinical, biochemical, hormonal, and histopathologic criteria.133,169 Imaging is useful for identifying adrenal adenomas or adrenal cortical carcinomas and can be performed with CT or MRI.133 As stated previously, hyperfunctioning adrenal adenomas cannot be distinguished from nonhyperfunctioning adrenal adenomas by imaging alone. However, in some cases of cortisol-producing adrenal adenomas, atrophy of the nonadenomatous ipsilateral and contralateral adrenal tissue can be seen, and this reflects suppressed ACTH levels.40

CHAPTER 53  Adrenal Glands Adrenal cortical carcinomas are typically very large at presentation, may have central necrosis and calcification, and can invade the inferior vena cava.15 Further discussion of the imaging appearance of adrenal cortical carcinoma can be found later in this chapter. ACTHindependent macronodular hyperplasia is a rare disease that can have a dramatic appearance on cross-sectional imaging of the adrenals, with the appearance of massively enlarged bilateral adrenal glands with macronodules measuring up to 5 cm.133,161

Adrenal Metastases The adrenal gland is a frequent site of metastatic disease owing to its rich sinusoidal blood supply.93 In a study of 1000 consecutive autopsies,2 the adrenals were the site of metastatic disease in 27% of cases. Tumors that commonly metastasize to the adrenals include carcinomas of the lung, kidney, colon, breast, esophagus, pancreas, liver, and stomach.66,169 In patients with a known primary malignancy, up to 38% to 50% of detected adrenal masses will represent metastatic disease.39,169 Adrenal metastases are most often detected by imaging with CT, MRI, and FDG PET/CT.102

FIG 53-11  Bilateral adrenal nodular hyperplasia due to Cushing’s disease. Axial unenhanced CT demonstrates bilateral enlarged and nodular adrenal glands secondary to an ACTH-secreting pituitary microadenoma.

A

1767

Morphologic features of metastases that can be used to distinguish them from benign adrenal tumors such as adenomas include size greater than 4 cm, heterogeneous density and irregular shape or margins, and bilateral lesions. Although these morphologic features can be helpful, they are considered nonspecific and poorly sensitive for distinguishing benign from malignant adrenal lesions.102,119,148 An interval increase in size of an adrenal nodule over a 6-month interval is strongly suggestive of metastatic disease in patients with a known primary malignancy.23 Compared with adrenal adenomas, adrenal metastases are unlikely to contain significant amounts of intracellular lipid.95 Therefore adrenal metastases typically have Hounsfield attenuation greater than 10 HU on unenhanced CT and absence of signal loss on opposedphase MRI19,29,47,61,102 (Fig. 53-12). With contrast enhancement, adrenal metastases typically have slow contrast enhancement followed by delayed washout of contrast on contrast-enhanced CT and MRI.33,52,134 Using enhanced CT with APW and RPW calculations, most metastases can be distinguished from adenomas by demonstrating less than 60% APW and less than 40% RPW.33 However, a recent report found that hypervascular metastases, including those originating from HCC and RCC, may demonstrate perfusion dynamics similar to adenomas. It that study, 84% of HCC and RCC metastases demonstrated greater than 60% APW and greater than 40% RPW.38 FDG PET/CT has been shown to have high sensitivity and specificity for diagnosis of metastatic disease, with diagnostic performance percentages greater than 90%.17,28,45,120 Furthermore, absence of FDG uptake in an adrenal nodule has a high negative predictive value (93%97%) for the absence of metastatic disease.120,159 However, metastases that are FDG PET/CT negative have been reported. Causes of falsenegative FDG PET/CT for metastatic disease include previous treatment with chemotherapy, central necrosis, hemorrhage, small size of the lesion, and certain subtypes of primary malignancies.39,50,52 Primary malignancies that have been reported to have adrenal metastases that are falsely negative on FDG PET/CT include pulmonary carcinoid (neuroendocrine) tumor and pulmonary adenocarcinoma with predominant bronchioalveolar carcinoma component.39 Metastases of mucinous neoplasms have also been reported to be false negative with FDG PET/CT.13 False-positive FDG PET/CT results can also occur. As discussed previously, a small percentage (<5%) of adenomas can demonstrate mild FDG avidity but usually less than the activity of the

B

FIG 53-12  Left adrenal metastasis from breast cancer on chemical shift MRI. A, Axial in-phase T1-weighted GRE shows left adrenal nodule slightly hyperintense to spleen. B, Axial opposed-phase T1-weighted GRE shows left adrenal nodule without signal loss. Metastatic disease was confirmed by biopsy.

1768

PART II  CT and MR Imaging of the Whole Body

A

B FIG 53-13  Left adrenal collision tumor (adenoma and metastasis) in a patient with history of lung cancer. A, Axial unenhanced CT demonstrates lipid-rich left adrenal adenoma with focal oval high-density area (+23 HU) along the lateral aspect of the adrenal nodule (arrow). B, Axial 18F FDG PET/CT shows focal high FDG activity (arrow) corresponding to the high-density area within the adrenal nodule, consistent with metastasis coexisting with adenoma.

liver.39 Other tumors or conditions of the adrenals that can lead to a false-positive FDG PET/CT result include pheochromocytomas, adrenal cortical hyperplasia, and endothelial cysts, sarcoid, tuberculosis, and periadrenal abnormalities.17,39 A recent study suggests that in oncologic patients, the combination of FDG PET/CT and adrenal protocol CT with washout calculation yielded higher accuracy than either method alone.120 The term collision tumor refers to the coexistence of two adjacent but histologically distinct tumors without histologic admixture.140 In the adrenal gland, one of the tumors can represent a benign adenoma and the other a malignant tumor, such as a metastasis.22,140,155 A collision tumor may be suspected on unenhanced CT by observing that one portion of an adrenal nodule measures +10 HU or less and another area measures greater than +10 HU (Fig. 53-13). A collision tumor can be suspected on MRI by demonstrating the presence of opposed-phase signal loss in one area of an adrenal mass and the absence of opposedphase signal loss in another area of an adrenal mass.137,140 Collision tumors have also been detected with FDG PET/CT by demonstrating FDG avidity in only a small portion of an adrenal mass, suggesting the coexistence of an FDG-negative adenoma with an FDG-avid metastasis22,155 (see Fig. 53-13). In this situation, careful coregistration of PET/ CT data is important in establishing the presence of two different tumors with different FDG uptake within the same adrenal gland.

Adrenocortical Carcinoma Adrenocortical carcinoma (ACC) is a rare neoplasm that accounts for only 0.05% to 0.2% of all cancers, or 1 to 2 per 1 million population per year. There is a bimodal age distribution, with increased incidence in children younger than 5 years and a second peak in the fourth and fifth decades of life.15,32 This tumor is more commonly seen in female patients, and females are more likely than males to present with  a hyperfunctioning tumor. ACC can be associated with genetic  syndromes, particularly in children, and these include BeckwithWiedemann syndrome, Li-Fraumeni cancer syndrome, Carney’s complex, and familial adenomatous polyposis coli.4,15 This tumor is typically very large at presentation, with an average size reported to be approximately 9 cm (range, 2-25 cm). Nonhyperfunctioning tumors tend to be slightly larger than hyperfunctioning

tumors at presentation.15,32 Up to 60% of these tumors are hormonally active (85% in children and 15%-30% in adults), and they may secrete cortisol, androgens, estrogens, and aldosterone. In the adult population, hyperfunctioning ACC tumors present with Cushing’s syndrome (30%), virilization (20%), or a combination of the two (10%-20% of cases). Approximately 2% of cases of ACC will present with feminization or hyperaldosteronism.15,111 On CT, adrenocortical carcinomas appear as large tumors with heterogeneous enhancement and often with cystic areas, hemorrhagic changes, calcifications, and central necrosis15 (Fig. 53-14). Rarely, small foci of intracytoplasmic lipid has been described in these tumors and thought to be related to the presence of cortisol or other lipidcontaining hormonal precursors.51,139 These tumors demonstrate less than 60% APW and less than 40% RPW on 15-minute delayed enhanced CT.150 Metastatic disease occurs in liver (48%-96%), lung (45%-97%), lymph nodes (25%-46%), and bone (11%-33%). Local invasion most commonly involved the kidneys (26%) and inferior vena cava (9%-19%). Less common sites of metastases included pancreas, brain, diaphragm, small intestine, and thyroid.15,111 Liver metastases can be hypervascular.15 On MRI these tumors are generally T1 hypointense and T2 hyperintense but often heterogeneous in signal because of hemorrhage and necrosis139 (see Fig. 53-14). These lesions demonstrate heterogeneous enhancement and slow washout. Inferior vena cava invasion may be better detected on MRI than CT.15,71,145 Adrenocortical carcinoma has been shown to have high avidity on FDG PET/CT (Fig. 53-15). Sensitivity and specificity for detection of ACC with FDG PET/CT is reported to be 89% to 100% and 70% to 100%, respectively.43,151 FDG PET/CT has also been shown to be useful for staging because it can detect metastases and determine tumor response to chemotherapy.43,90,151

Myelolipoma Myelolipomas are benign tumors of the adrenals that are composed of fat and hematopoietic tissue and often resemble bone marrow. These tumors are usually small and round and may have a capsule. Symptomatology is related to size and potential compression of adjacent structures. They typically contain variable quantities of macroscopic

CHAPTER 53  Adrenal Glands

A

1769

C

B

FIG 53-14  Adrenocortical carcinoma in a 45-year-old woman who presented with hirsutism and elevated plasma androgens. A, Coronal unenhanced CT demonstrates a large right adrenal mass with heterogeneous attenuation and focal calcifications. B, Axial T2-weighted MRI shows right adrenal mass with heterogeneous T2 hyperintensity. C, Coronal contrast-enhanced T1-weighted MRI shows heterogeneously enhancing mass with central necrosis.

A

B

FIG 53-15  Adrenocortical carcinoma in a 50-year-old man with elevated cortisol levels. A, Axial contrastenhanced CT shows a large heterogeneously enhancing mass arising from the right adrenal. B, Axial 18F FDG PET/CT demonstrates high avidity within the right adrenal mass. Surgical pathology revealed adrenocortical carcinoma.

fat that is easily detected on CT and MRI (Fig. 53-16). The amount of recognizable macroscopic fat varies from almost 100% to none.26 Identification of macroscopic fat—even small foci—is usually diagnostic, and no further workup is needed. On CT the areas of fat attenuation measure between −150 HU and −30 HU.158 These tumors may also contain calcification in 20% of cases.26 Myelolipomas can hemorrhage when they get large (>5 cm), and as a result there may be areas of high and low attenuation depending on the age of the hemorrhage.41,82,130 On MRI the areas of fat within myelolipomas parallel the signal of retroperitoneal fat. On FDG PET/CT, myelolipomas typically show FDG uptake lower than that of liver background. Rarely, myelolipoma with dominant adenomatous and hematopoietic tissue can have high FDG avidity.45 As stated previously, small foci of macroscopic fat have been reported with adrenocortical carcinomas; however, ACC tumors are typically very large and malignant appearing at presentation  and therefore not likely to be mistaken for benign myelolipomas.51 Massive bilateral adrenal myelolipomas have been associated with long-standing or poorly managed congenital adrenal hyperplasia94 (see Fig. 53-16).

Pheochromocytoma Pheochromocytoma is a rare catecholamine-secreting tumor that originates from the adrenal medulla. This tumor often manifests clinically as a younger patient with hypertension. Other symptoms include palpitations, headaches, sweating, tremor, and anxiety.102 However, in 5% to 10% of cases, it can be detected by imaging as an incidental finding.26,169 In patients with suspected pheochromocytoma, the clinical diagnosis can be made by demonstrating elevated plasma catecholamine levels and elevated 24-hour urine vanillylmandelic acid and metanephrines. These tests have sensitivities ranging from 89% to 100%.102,169 If the diagnosis is suspected based on clinical grounds and biochemical testing, imaging can be performed to confirm the diagnosis and localize the tumor. The majority of pheochromocytomas are located in the adrenal gland. It is important to remember that 10% are extraadrenal, 10% are bilateral, and 10% are malignant. Approximately 10% are familial and associated with multiple endocrine neoplasia (MEN) types IIA and IIB,

1770

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 53-16  Myelolipoma. A, Axial contrast-enhanced CT shows right adrenal mass containing large areas of macroscopic fat (−50 to −100 HU), diagnostic of a benign myelolipoma. B, Massive bilateral myelolipomas associated with congenital adrenal hyperplasia.

A

B FIG 53-17  Malignant pheochromocytoma in a 61-year-old woman who presented with intermittent hypertension and tremors. A, Axial contrast-enhanced CT shows a large right adrenal mass that has heterogeneous enhancement and central necrosis. B, Whole-body planar image 123I-MIBG scan shows intense uptake of radionucleotide within the right adrenal mass compatible with a pheochromocytoma.

von Hippel-Lindau disease, neurofibromatosis, Sturge-Weber syndrome, Carney’s triad, and isolated familial pheochromocytoma. A familial association should be considered in cases of bilateral pheochromocytomas.18 Extraadrenal pheochromocytomas are more appropriately termed paragangliomas and can occur anywhere along the sympathetic nervous system, from the base of the brain to the urinary bladder. Common locations for paragangliomas include the level of the aortic bifurcation (also referred to as Zuckerkandl’s organ), retroperitoneum, bladder wall, heart, mediastinum, and carotid and glomus jugulare bodies.18 Adrenal pheochromocytomas and paragangliomas share similar imaging characteristics.126 Pheochromocytomas and paragangliomas can be imaged with a variety of modalities, with CT, MRI, and radionuclide imaging the most common imaging techniques employed. These tumors have both

characteristic and atypical imaging features owing to sometimes complex pathologic variations, and this can lead to challenges on diagnostic imaging.18 On CT these tumors usually appear as well-defined solid homogeneous masses when small but can appear as heterogeneous masses with foci of calcification and cystic areas, particularly when larger (Fig. 53-17). The average tumor size for sporadic cases is 5 cm, and familial cases tend to be smaller.53 The unenhanced CT attenuation is almost always greater than 10 HU. However, rare accounts of lipid degeneration within pheochromocytomas can result in an unenhanced attenuation of less than 10 HU.20,129 With contrast enhancement, most pheochromocytomas will demonstrate slow enhancement followed by delayed washout, with an APW calculation of less than 60% at 15 minutes or less than 40% for RPW.150 However, recent reports indicate

CHAPTER 53  Adrenal Glands

C

B

A

1771

FIG 53-18  Paraganglioma in a 30-year-old woman who presented with refractory hypertension. A, Axial T2-weighted MRI shows homogeneously T2-hyperintense left periaortic mass just above the level of the aortic bifurcation (Zuckerkandl’s organ), illustrating the “light bulb” T2-bright appearance of pheochromocytomas and paragangliomas. B, Axial contrast-enhanced T1-weighted MRI demonstrates heterogeneous enhancement within the mass. C, 123I-MIBG fused SPECT/CT axial image shows diffuse uptake within the tumor, compatible with paraganglioma.

that approximately 30% of pheochromocytomas will demonstrate greater than 60% APW or greater than 40% RPW and mimic the behavior of adrenal adenomas.64,79,121 With MRI, pheochromocytomas and paragangliomas are typically low signal on T1-weighted imaging. Although classically described as having a “light-bulb bright” signal on T2-weighted images, this sign has been shown to be neither sensitive nor specific for the diagnosis of pheochromocytoma89,156 (Fig. 53-18). More commonly these tumors demonstrate heterogeneous intermediate to high signal on T2-weighted imaging, and at least 30% showing moderate to low T2-weighted signal intensity.17 These tumors generally will exhibit substantial enhancement on MRI. A recent study reported maximal enhancement occurring in the interstitial or equilibrium phase of imaging.134 Radionuclide imaging can be performed with 123I-labeled MIBG (see Fig. 53-17). Whole-body imaging is particularly useful when a suspected pheochromocytoma cannot be localized or to exclude metastatic disease. A recent meta-analysis evaluating the diagnostic performance of 123I MIBG for diagnosis of pheochromocytoma demonstrated an overall sensitivity of 94% (range, 75%-100%) and specificity of 92% (range, 50%-100%).74 Because of its high specificity, 123I MIBG can be used to confirm that a mass is a pheochromocytoma and also identify patients who are candidates for therapeutic dosing of 131I MIBG.36 FDG PET imaging has been reported as an alternative nuclear imaging test, particularly in cases where there is a high clinical suspicion for pheochromocytoma but MIBG is negative. A previous study reported that most benign and malignant pheochromocytomas will demonstrate FDG avidity, with uptake found in a greater percentage of malignant than benign tumors.143

Adrenal Hemorrhage The adrenal gland has a complex vascular network that contributes to its susceptibility to hemorrhage.44 Etiologies for adrenal hemorrhage include traumatic and nontraumatic events or conditions. Blunt abdominal trauma accounts for 80% of cases of adrenal hemorrhage, and it is typically unilateral and on the right side.26,102 Approximately 2% of patients undergoing abdominal and pelvic CT for trauma have evidence of adrenal hemorrhage.94 Liver transplantation is associated with right adrenal hemorrhage and likely related to ligation and division of the right adrenal vein during recipient hepatectomy.30 Nontraumatic adrenal hemorrhage can be unilateral or bilateral; etiologies include systemic anticoagulation therapy, bleeding diatheses, severe stress, sepsis, underlying neoplasm, or rare conditions such as antiphospholipid syndrome.26,102,125 Bilateral adrenal hemorrhage occurs in 20% of cases and has rarely been associated with acute adrenal insufficiency

or adrenal crisis, a condition that can lead to cardiovascular collapse and death if not promptly diagnosed and treated.102,104,114 Acute adrenal hemorrhage or hematoma appears on CT as a round or oval hyperdense mass with smooth or irregular margins and a CT attenuation of 50 to 90 HU on unenhanced CT (Fig. 53-19). Adjacent retroperitoneal fat stranding is sometimes observed. Over the course of time, the size and density of the adrenal hemorrhage will decrease and calcifications may develop, usually after 1 year (see Fig. 53-19).  On MRI, acute (<7 days) hemorrhage (intracellular deoxyhemoglobin) will appear iso- or hypointense on T1-weighted images and markedly hypointense on T2-weighted images. Subacute (1 week to 2 months) hemorrhage (methemoglobin) will appear hyperintense on T1-weighted images. In the chronic stage (2-3 months) a T2-hypointense rim will appear along the periphery of the adrenal hemorrhage secondary to hemosiderin-laden macrophages.102

Adrenal Cyst Adrenal cysts are rare, with a reported incidence of less than 0.1% based on a combined autopsy series.110 Reported histologic types of adrenal cysts include endothelial cysts (45%), pseudocysts (40%),  epithelial cysts (10%), and parasitic cysts (5%)72,110,132 (Fig. 53-20). Endothelial cysts may be subclassified into lymphangiomatous or hemangiomatous. Pseudocysts are thought to arise from prior adrenal hemorrhage. Epithelial cysts are thought to develop from benign cystic neoplasms or a congenital origin. Parasitic cysts are rare but most commonly associated with Echinococcus granulosus infections132 (see Fig. 53-20). Adrenal cysts can occur at any age but have a peak incidence between the third and sixth decades of life. They are two times more common in women and usually unilateral. Most adrenal cysts are asymptomatic and discovered incidentally. However, some cysts can present with pain due to hemorrhage or cyst rupture. Rarely, adrenal cysts are associated with hypertension, postulated to be a consequence of compression of the renal artery or renal medulla.11,49 On CT, adrenal cysts typically appear as well-defined oval or round masses that have fluid-density attenuation (<20 HU) and no enhancement or mild wall enhancement (see Fig. 53-20). Not uncommonly, thin septations and smooth rim calcifications may be seen. Adrenal cysts are usually of small to medium size, with a reported average size of 5 cm, but can range from 1 to 20 cm.94 On MRI these lesions appear hypointense on T1-weighted images, hyperintense on T2-weighted images, and frequently demonstrate a T2-hypointense rim (see Fig. 53-20). Wall enhancement is seen more commonly on contrastenhanced MRI than CT.110,132

1772

PART II  CT and MR Imaging of the Whole Body

A

B 4 years later

C

D

E FIG 53-19  Adrenal hemorrhage in a 45-year-old man who was hit by lightning. A, Axial unenhanced CT shows a hyperdense (+60 HU) mass involving the left adrenal, with inflammatory stranding in the retroperitoneum consistent with acute adrenal hemorrhage. B, Axial unenhanced CT performed 2 months after initial CT (A) shows decreased size and density of left adrenal mass consistent with resolving adrenal hemorrhage. C, Axial unenhanced CT performed 4 years after initial CT (A) demonstrates continued decrease in size and density of left adrenal hemorrhage and development of focal calcification. D, Chronic left adrenal hemorrhage in a different patient demonstrates large areas of low or medium density and coarse calcifications. E, Axial 18 F FDG PET/CT image demonstrates no FDG avidity in the same mass (D). Owing to its large size and complexity it was surgically removed and found to be chronic adrenal hemorrhage.

CHAPTER 53  Adrenal Glands

A

B

1773

C

FIG 53-20  Left adrenal cyst. A, Axial contrast-enhanced CT shows a nonenhancing left adrenal mass that is water density (2 HU), compatible with a simple adrenal cyst. B, Axial T2-weighted MRI demonstrates small homogeneous T2-hyperintense right adrenal mass that does not enhance (not shown). C, Axial contrastenhanced CT demonstrates a cystic right adrenal mass with thick nodular rim, the result of Echinococcus granulosa infection. Note the cystic lesion in the head of the pancreas (arrow), also found to represent a sequela of E. granulosa infection.

Other adrenal tumors such as adenomas, cystic pheochromocytomas, and necrotic primary or secondary malignancies may appear similar to adrenal cysts on unenhanced CT or MRI. However, adrenal cysts can be differentiated from these tumors because most adenomas, cystic pheochromocytomas, or necrotic adrenal tumors will demonstrate significant solid or heterogeneous enhancement on contrastenhanced CT or MRI.132

Adrenal Infection Infection of the adrenal glands is most commonly due to granulomatous disease. Histoplasmosis and tuberculosis are the most commonly seen agents of adrenal infection. Less commonly, adrenal infection can occur with blastomycosis, Mycobacterium avium-intracellulare, and cytomegalovirus, particularly in patients with acquired immunodeficiency syndrome (AIDS).94 On CT, in the acute phase, bilateral diffuse or masslike adrenal enlargement with heterogeneous density and central low attenuation can be seen. In the chronic stage, calcifications, cysts, and gland atrophy may occur.26,55,108,163,164 Adrenal infection, like hemorrhage, can lead to adrenal insufficiency. Adrenal abscess, caused by Escherichia coli, group B Streptococcus, and Bacteroides, is rare and usually unilateral.94

Ganglioneuroma Ganglioneuroma is a benign tumor that arises from sympathetic ganglia. It can develop within the adrenal medulla (20%-40%) or extraadrenal retroperitoneum (60%-80%). These tumors are usually asymptomatic and are discovered incidentally. Most patients are younger than age 60, with a mean age of 49 years in one small series.78 These tumors are typically oval or lobulated with well-defined borders and are between 5 and 9 cm in size. On unenhanced CT these tumors appear homogeneous with attenuation less than muscle and sometimes may have calcification.127 With contrast enhancement, these tumors demonstrate mild or delayed enhancement, which can be centripetal78,116,127,135 (Fig. 53-21). On MRI these tumors are reported to have low signal intensity on T1-weighted images and heterogeneous high signal intensity on T2-weighted images.135 Markedly high T2-weighted signal has been described, and this finding has been attributed to the presence of myxoid matrix within the tumor, but others report mild T2-weighted signal and the absence of abundant myxoid material on histologic examination.116,127 In one small series, this tumor demonstrated low to moderate hypermetabolism on FDG PET scan, with a mean SUV value of 2.4 (range, 1.5-2.9).135

Adrenal Lymphoma Primary adrenal lymphoma is extremely rare, with only 70 cases reported in the literature5,92,96 (Fig. 53-22). Secondary involvement of the adrenal glands is more common and most often due to nonHodgkin lymphoma.118 Although autopsy studies have found histologic involvement of the adrenals in up to 25% of patients with lymphoma, only 4% of patients with non-Hodgkin’s lymphoma will demonstrate CT evidence for adrenal metastases.5,75,118 Primary adrenal lymphoma most commonly affects elderly men, and 50% of patients present with primary adrenal insufficiency. The most common type of lymphoma is diffuse large B cell. In a recent review, the majority of cases demonstrate bilateral involvement without nodal disease. CT demonstrates complex masses with variable density ranging from homogeneous density to a cystic appearance due to central necrosis. On MRI, primary adrenal lymphoma appears low signal intensity on T1-weighted images and high signal on T2-weighted images. Whole-body FDG PET imaging demonstrates intense FDG.92 Secondary adrenal lymphoma usually appears as diffuse enlargement of the adrenals or (less commonly) discrete masses. Frequently there is evidence for retroperitoneal lymphadenopathy. On CT, secondary adrenal lymphoma typically appears as homogeneous adrenal enlargement or masses with mild enhancement after contrast administration. Calcifications are sometimes seen after treatment. MRI appearance is nonspecific and typically low signal intensity on T1-weighted images and heterogeneous high signal on T2-weighted images.75,94,118 Secondary adrenal lymphoma demonstrates hypermetabolic activity with FDG PET/CT, and this imaging modality is helpful in distinguishing nonfunctioning adrenal neoplasm or hyperplasia from lymphomatous involvement. FDG PET/CT is also useful for following disease progression or regression. Only certain subtypes of lymphoma such as marginal zone, peripheral T-cell, and low-grade lymphomas may not demonstrate reliable FDG avidity.17,105

Adrenal Hemangioma Adrenal hemangioma is a rare benign tumor that is usually asymptomatic, discovered incidentally, and often large at presentation (mean size, 11 cm; range, 1-30 cm). In a recent review, the median age of patients at diagnosis was 62 years (range, 19-84 years). These tumors were more common in females compared with males.113 At unenhanced CT, adrenal hemangioma appears as a well-defined hypoattenuating or heterogeneously attenuating mass, often with focal or speckled

1774

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 53-21  Ganglioneuroma. A, Axial unenhanced CT shows a large heterogeneous low-density right adrenal mass with punctate calcification. B, Axial contrast-enhanced CT shows a low-attenuation mass with central area of increased enhancement. C, Axial delayed-enhanced CT demonstrates progressive increase in density or enhancement in periphery of the mass, which appeared as areas of nonenhancement during the dynamic phase of contrast enhancement (B).

FIG 53-22  Lymphoma. Axial contrast-enhanced CT shows bilateral large low-density adrenal masses that were diagnosed as primary large B-cell lymphoma in a 30-year-old man who was infected with human immunodeficiency virus.

calcifications (two thirds of cases). With contrast enhancement, peripheral discontinuous nodular enhancement may be seen, similar to that seen with cavernous hemangioma of the liver; however, complete fill-in is uncommon because of the presence of necrosis, fibrosis, and thrombosis in the central portion of the mass (Fig. 53-23). On MRI, low signal intensity on T1-weighted images and high signal intensity on T2-weighted images is usually seen; however, hemorrhage or central necrosis can result in increased T1-weighted signal intensity.80,89,94,113 The available literature on the FDG avidity of adrenal hemangiomas is sparse, but in one case report there was no demonstrable FDG uptake.113 Owing to their large size at presentation, biopsy or surgical removal often is needed to exclude malignancy.

ALGORITHM FOR WORKUP OF THE INCIDENTAL ADRENAL NODULE Adrenal incidentaloma is a term that generally refers to an adrenal mass measuring greater than 1 cm that is discovered during a radiologic examination that was performed for indications other than evaluation of adrenal disease.14,99,147,169,167 The prevalence of incidentally detected adrenal masses on imaging is approximately 5% (range, 3%-7%) in the adult population, which is similar to the frequency of incidental

adrenal nodules detected on autopsy series.14,147,169 The most common underlying pathology for an adrenal incidentaloma is the nonhyperfunctioning adrenal adenoma, which accounts for 75% of incidentalomas.147 The likelihood of this diagnosis increases with increasing age of the patient. The probability of detecting an incidental adrenal adenoma on CT is approximately 0.2% for patients aged 20 to 29 years, compared with approximately 7% in patients older than 70.169 A recent study by Song et al. demonstrated that for 1049 adrenal nodules in patients without a prior history of malignancy, none of the adrenal nodules turned out to be malignant.147 These data would suggest that adrenal incidentalomas in nononcologic patients could be left alone, with no further workup needed. However, most referring physicians and radiologists are too uncomfortable with the idea of no further follow-up, particularly because a small number of lesions may represent benign tumors, such as the clinically silent pheochromocytoma and subclinical hyperfunctioning adenoma, both of which can still cause morbidity or mortality if not removed surgically or treated medically.23-25,169 Furthermore, the possibility of an unknown primary or secondary adrenal malignancy, while remote, still exists. Therefore the workup of incidental adrenal nodules remains an important mission of the abdominal imager. The first part of the workup of an incidental adrenal nodule is to distinguish hormonally active adrenal tumors from those that are not (Fig. 53-24). This part of the workup requires a referral for hormonal testing that is performed by our clinical colleagues. If hormonal testing is positive, then clinical management is directed toward the potential need for surgical intervention.169 If hormonal testing is negative, then the workup is directed toward distinguishing benign from malignant adrenal tumors. The first and simplest step is to determine if the patient has any prior cross-sectional imaging studies to evaluate for size stability. The absence of growth  for at least 1 year or more can be used as an indication of benignity.14,25,26,46,102 Frequently, though, no prior imaging is available. It should be noted that adrenal masses measuring greater than 4 cm are generally considered by our clinical colleagues as an indication for surgical consult independent of imaging characteristics. This approach is based on a report involving 887 patients who had adrenal incidentalomas. It was shown that a diameter of greater than 4 cm correlated with a 90% sensitivity for diagnosis of adrenocortical carcinoma. However, this 4-cm threshold also demonstrated poor specificity, insofar as only 24% of adrenal lesions greater than 4 cm were malignant.6,14,99,167,169 It has been proposed that algorithms for the imaging workup of adrenal incidentalomas should be limited to adrenal masses measuring between 1 and 4 cm.14,148 Adrenal masses greater than 4 cm should be considered to be a higher risk for malignancy and therefore

CHAPTER 53  Adrenal Glands

A

C

B

FIG 53-23  Hemangioma. A, Frontal radiograph of the abdomen demonstrates a large soft tissue mass in the left upper quadrant of the abdomen, with stippled calcifications. B, Axial unenhanced CT shows large mixed-attenuation left adrenal mass with large calcifications. C, Axial contrast-enhanced CT shows peripheral nodular enhancement without centripetal fill-in due to central necrosis and fibrosis. This mass was removed surgically and found to be a large adrenal hemangioma.

INCIDENTAL ADRENAL NODULE

1-4 cm

>4 cm Surgical consult

Hormonal testing positive

Hormonal testing negative

10-20 HU Noncontrast CT

≤10 HU

MRI

>10 HU

Stop

<20% Signal drop

Benign

Indeterminate

>20 HU

>10% Neg pix Benign

>20% Signal drop

Enhanced CT

>60% Washout

<60% Washout

Benign

Indeterminate

Stop or F/U

BX, SURG, F/U, FDG PET

Stop

CT washout

BX, SURG, F/U, FDG PET

FIG 53-24  Flowchart shows imaging algorithm for workup of the incidental adrenal nodule. BX, biopsy; F/U, follow-up; SURG, surgical consult.

1775

1776

PART II  CT and MR Imaging of the Whole Body

should more often proceed toward a tissue diagnosis rather than imaging workup.14,148 Even so, many benign adrenal tumors may be larger than 4 cm, and these tumors include adenomas, myelolipomas, and adrenal cysts.67,83 The algorithm for imaging workup begins with the initial imaging test where the adrenal nodule was first discovered. If the adrenal mass contains evidence for macroscopic fat, the adrenal mass is most likely a benign myelolipoma, and no further workup is necessary.26 The presence of macroscopic fat is determined in two ways: on CT (unenhanced or contrast-enhanced CT) by identifying foci of low attenuation measuring between −150 and −30 HU, or on MRI by identifying foci of high signal on T1-weighted images, which suppress signal with fat saturation. If there is no macroscopic fat within the adrenal mass, then the workup is directed toward differentiation of benign nodules from indeterminate nodules, as outlined in the flowchart (see Fig. 53-24). If the initial imaging test is a noncontrast CT, then an attenuation value of the adrenal nodule is determined. If the attenuation value measures +10 HU or less, the adrenal nodule can be categorized as benign, most likely representing a lipid-rich adenoma but also possibly an adrenal cyst. Because both diagnoses are benign, no further imaging or workup is required. If the adrenal nodule measures greater than +10 HU on noncontrast CT, or if the initial imaging test was a contrast-enhanced CT, options are to either perform an adrenal protocol CT (with washout calculation) or perform MRI with chemical shift imaging. The option to perform MRI should only be followed if the adrenal nodule measures less than +20 HU on unenhanced CT, because previous studies have shown MRI has low sensitivity for detecting lipid within lipidpoor adenomas measuring greater than +20 HU on noncontrast CT.61,141 In addition, if the adrenal nodule measures greater than +10 HU, one may consider histogram analysis. Adrenal nodules that demonstrate greater than 10% negative pixels with histogram analysis can be considered benign, and no further workup is needed.70,77,131 The diagnosis of a lipid-poor adenoma can be made on adrenal protocol CT by calculating an APW of greater than 60% or RPW of greater than 40%, using 15-minute delayed imaging. As stated previously, a small percentage of pheochromocytomas and adrenal metastases from RCC or HCC uncommonly demonstrate washout patterns similar to adenomas.38,121 Careful attention to clinical history and overall imaging appearance may offset the risk of misclassification of these tumors as adenomas. On MRI the identification of signal loss on opposed-phase GRE within an adrenal nodule indicates the presence of intracellular lipid and is therefore diagnostic of an adrenal adenoma. Demonstration of signal loss on opposed-phase imaging can often be determined qualitatively or quantitatively using either SII or adrenal-spleen ratio. An SII value of greater than 20% or adrenal-spleen ratio of less than 0.71 is indicative of an adenoma. Rarely, though, intracytoplasmic lipid may occur in some adrenocortical carcinomas, metastases from HCC or RCC, and pheochromocytomas; these malignancies can mimic the appearance of adenomas on chemical shift MRI.137 Fortunately in most instances, the clinical history or overall imaging appearance  will suggest a malignant process and prompt further evaluation. If the APW or RPW on adrenal protocol CT is less than 60% or 40%, respectively, then the adrenal nodule remains indeterminate. At this point in the algorithm, the next possible options include follow-up to establish size stability, adrenal biopsy, surgical referral, or further imaging with FDG PET/CT. If there is insufficient signal loss by qualitative or quantitative analysis on chemical shift MRI, then the adrenal nodule remains indeterminate. In this case, the next-step options are the following: adrenal

protocol CT with washout, biopsy, surgical referral, FDG PET/CT, or follow-up to establish size stability. The role of FDG PET/CT in the workup of incidental adrenal nodules in nononcologic patients is mainly as an alternative noninvasive imaging test when adrenal protocol CT or chemical shift MRI is inconclusive or indeterminate. FDG PET/CT has a high negative predictive value for malignancy in nononcologic patients and can provide supportive data for observational imaging follow-up rather than biopsy or surgery if conservative management of an adrenal incidentaloma is favored.7,152,165 In the oncologic patient, FDG PET/CT is primarily used for cancer staging owing to its high sensitivity and specificity for identifying malignant disease. Because FDG PET/CT has high negative predictive value for malignancy (93%-97%), adrenal lesions that have low FDG avidity are likely to be benign.120,159 Adrenal nodules that have high FDG avidity (FDG uptake greater than liver) are likely to represent metastases.28 If the CT portion of the FDG PET/CT was performed as an unenhanced CT, then CT densitometry can be applied to distinguish benign adenomas from metastases.31 Even though FDG PET/CT is highly sensitive and specific for metastatic disease in the oncologic patient, false positives and false negatives have been reported as previously discussed. In cases where clinical management may be significantly affected or altered by a suspected adrenal metastasis on FDG PET/CT, further workup with adrenal biopsy should be considered when findings are borderline or equivocal.

Adrenal Biopsy When adrenal nodules cannot be characterized with CT, MRI, or FDG PET/CT, image-guided biopsy of adrenal masses can be performed. Adrenal biopsies are safe procedures with a complication rate of approximately 3%.102,169 The most common complications associated with this procedure include hemorrhage and pneumothorax.26 Other risks include pain, hematuria, pancreatitis, infection, and tumor tract seeding (an extremely rare complication).102,169 Biopsy of a pheochromocytoma can lead to precipitation of a life-threatening hypertensive crisis, and therefore the possibility of this diagnosis should be excluded with biochemical testing before any adrenal biopsy. The accuracy of adrenal biopsy has been reported to be 83% to 96%.26 CT guidance is the modality of choice for this procedure. Because of its relatively deep and retroperitoneal location, percutaneous adrenal biopsy can be technically challenging in some cases. A coaxial technique is often used to enable multiple biopsy passes through a single outer trocar needle. Changing the gantry angulation and placing the patient in decubitus position are two techniques to reduce the risk of causing a pneumothorax. Occasionally a transhepatic approach may be necessary to avoid traversing lung parenchyma when attempting biopsy of a rightsided adrenal nodule102 (Fig. 53-25). A posterior or posterolateral approach is preferred for left-sided adrenal nodule biopsy to avoid traversing the pancreas (see Fig. 53-25).

CONCLUSION The adrenal glands have been the subject of a tremendous amount of research for the past 150 years. Advances in diagnostic imaging and interventional radiology in the last 2 decades has led to the development of new techniques for detecting, diagnosing, and managing adrenal pathology. Although incidentally discovered adrenal nodules are most likely caused by adrenal adenomas, distinguishing this benign tumor from malignancy or other primary adrenal tumors can be challenging. The radiologist plays an important role in guiding our clinical colleagues toward the appropriate imaging workup of adrenal diseases.

CHAPTER 53  Adrenal Glands

A

1777

B

FIG 53-25  Adrenal biopsy. A, Axial unenhanced CT during biopsy of a right adrenal mass using a transhepatic approach. B, Axial unenhanced CT during biopsy of a left adrenal mass using a posterior paravertebral approach.

REFERENCES 1. Abrams HL, Siegelman SS, Adams DF, et al: Computed tomography versus ultrasound of the adrenal gland: A prospective study. Radiology 143(1):121–128, 1982. 2. Abrams HL, Spiro R, Goldstein N: Metastases in carcinoma: Analysis of 1000 autopsied cases. Cancer 3(1):74–85, 1950. 3. Addison T: On the constitutional and local effects of disease of the suprarenal capsules, London, 1855, Samual Highley. 4. Agrons GA, Lonergan GJ, Dickey GE, et al: Adrenocortical neoplasms in children: Radiologic-pathologic correlation. Radiographics 19(4):989–1008, 1999. 5. Alvarez-Castells A, Pedraza S, Tallada N, et al: CT of primary   bilateral adrenal lymphoma. J Comput Assist Tomogr 17(3):408–409, 1993. 6. Angeli A, Osella G, Ali A, et al: Adrenal incidentaloma: An overview of clinical and epidemiological data from the National Italian Study Group. Horm Res 47(4–6):279–283, 1997. 7. Ansquer C, Scigliano S, Mirallie E, et al: 18F-FDG PET/CT in the characterization and surgical decision concerning adrenal masses: A prospective multicentre evaluation. Eur J Nucl Med Mol Imaging 37(9):1669–1678, 2010. 8. Bae KT, Fuangtharnthip P, Prasad SR, et al: Adrenal masses: CT characterization with histogram analysis method. Radiology 228(3):735–742, 2003. 9. Bagheri B, Maurer AH, Cone L, et al: Characterization of the normal adrenal gland with 18F-FDG PET/CT. J Nucl Med 45(8):1340–1343, 2004. 10. Bashir MR, Bhatti L, Marin D, et al: Emerging applications for ferumoxytol as a contrast agent in MRI. J Magn Reson Imaging 2014. 11. Basiri A, Salim NS, Shadpour P: Hypertension secondary to an adrenal pseudocyst cured by laparoscopic partial adrenalectomy. J Endourol 16(9):689–691, 2002. 12. Bassignani MJ: Adrenal and retroperitoneal MR imaging. In Brant WE, de Lange EE, editors: Essentials of body MRI, New York, 2012, Oxford University Press, pp 194–214. 13. Berger KL, Nicholson SA, Dehdashti F, et al: FDG PET evaluation of mucinous neoplasms: Correlation of FDG uptake with histopathologic features. AJR Am J Roentgenol 174(4):1005–1008, 2000. 14. Berland LL, Silverman SG, Gore RM, et al: Managing incidental findings on abdominal CT: White paper of the ACR incidental findings committee. J Am Coll Radiol 7(10):754–773, 2010.

15. Bharwani N, Rockall AG, Sahdev A, et al: Adrenocortical carcinoma: The range of appearances on CT and MRI. AJR Am J Roentgenol 196(6):W706–W714, 2011. 16. Bilbey JH, McLoughlin RF, Kurkjian PS, et al: MR imaging of adrenal masses: Value of chemical-shift imaging for distinguishing adenomas from other tumors. AJR Am J Roentgenol 164(3):637–642, 1995. 17. Blake MA, Cronin CG, Boland GW: Adrenal imaging. AJR Am J Roentgenol 194(6):1450–1460, 2010. 18. Blake MA, Kalra MK, Maher MM, et al: Pheochromocytoma: An imaging chameleon. Radiographics 24(Suppl 1):S87–S99, 2004. 19. Blake MA, Kalra MK, Sweeney AT, et al: Distinguishing benign from malignant adrenal masses: Multi-detector row CT protocol with 10-minute delay. Radiology 238(2):578–585, 2006. 20. Blake MA, Krishnamoorthy SK, Boland GW, et al: Low-density pheochromocytoma on CT: A mimicker of adrenal adenoma. AJR Am J Roentgenol 181(6):1663–1668, 2003. 21. Blake MA, Slattery JM, Kalra MK, et al: Adrenal lesions: Characterization with fused PET/CT image in patients with proved or suspected malignancy—Initial experience. Radiology 238(3):970–977, 2006. 22. Blake MA, Sweeney AT, Kalra MK, et al: Collision adrenal tumors on PET/CT. AJR Am J Roentgenol 183(3):864–865, 2004. 23. Boland GW: Adrenal imaging: Why, when, what, and how? Part 1. Why and when to image? AJR Am J Roentgenol 195(6):W377–W381, 2010. 24. Boland GW: Adrenal imaging: Why, when, what, and how? Part 2. What technique? AJR Am J Roentgenol 196(1):W1–W5, 2011. 25. Boland GW: Adrenal imaging: Why, when, what, and how? Part 3. The algorithmic approach to definitive characterization of the adrenal incidentaloma. AJR Am J Roentgenol 196(2):W109–W111, 2011. 26. Boland GW, Blake MA, Hahn PF, et al: Incidental adrenal lesions: Principles, techniques, and algorithms for imaging characterization. Radiology 249(3):756–775, 2008. 27. Boland GW, Blake MA, Holalkere NS, et al: PET/CT for the characterization of adrenal masses in patients with cancer: Qualitative versus quantitative accuracy in 150 consecutive patients. AJR Am J Roentgenol 192(4):956–962, 2009. 28. Boland GW, Dwamena BA, Jagtiani Sangwaiya M, et al: Characterization of adrenal masses by using FDG PET: A systematic review and meta-analysis of diagnostic test performance. Radiology 259(1):117–126, 2011. 29. Boland GW, Lee MJ, Gazelle GS, et al: Characterization of adrenal masses using unenhanced CT: An analysis of the CT literature. AJR Am J Roentgenol 171(1):201–204, 1998.

1778

PART II  CT and MR Imaging of the Whole Body

30. Bowen AD, Keslar PJ, Newman B, et al: Adrenal hemorrhage after liver transplantation. Radiology 176(1):85–88, 1990. 31. Brady MJ, Thomas J, Wong TZ, et al: Adrenal nodules at FDG PET/CT in patients known to have or suspected of having lung cancer: A proposal for an efficient diagnostic algorithm. Radiology 250(2):523– 530, 2009. 32. Brennan MF: Adrenocortical carcinoma. CA Cancer J Clin 37(6):348– 365, 1987. 33. Caoili EM, Korobkin M, Francis IR, et al: Delayed enhanced CT of lipid-poor adrenal adenomas. AJR Am J Roentgenol 175(5):1411–1415, 2000. 34. Caoili EM, Korobkin M, Francis IR, et al: Adrenal masses: Characterization with combined unenhanced and delayed enhanced CT. Radiology 222(3):629–633, 2002. 35. Carney JA: Adrenal gland. In Sternberg SS, editor: Histology for pathologists, New York, 1992, Raven, pp 321–336. 36. Carrasquillo JA, Pandit-Taskar N, Chen CC: Radionuclide therapy of adrenal tumors. J Surg Oncol 106(5):632–642, 2012. 37. Cheung SC, Lee R, Tung HK, et al: Persistent adrenal enhancement may be the earliest CT sign of significant hypovolaemic shock. Clin Radiol 58(4):315–318, 2003. 38. Choi YA, Kim CK, Park BK, et al: Evaluation of adrenal metastases from renal cell carcinoma and hepatocellular carcinoma: Use   of delayed contrast-enhanced CT. Radiology 266(2):514–520, 2013. 39. Chong S, Lee KS, Kim HY, et al: Integrated PET-CT for the characterization of adrenal gland lesions in cancer patients: Diagnostic efficacy and interpretation pitfalls. Radiographics 26(6):1811–1824, discussion 1124–1126, 2006. 40. Choyke PL, Doppman JL: Case 18: Adrenocorticotropic hormonedependent Cushing syndrome. Radiology 214(1):195–198, 2000. 41. Cyran KM, Kenney PJ, Memel DS, et al: Adrenal myelolipoma. AJR Am J Roentgenol 166(2):395–400, 1996. 42. Daunt N: Adrenal vein sampling: How to make it quick, easy, and successful. Radiographics 25(Suppl 1):S143–S158, 2005. 43. Deandreis D, Leboulleux S, Caramella C, et al: FDG PET in the management of patients with adrenal masses and adrenocortical carcinoma. Horm Cancer 2(6):354–362, 2011. 44. Dobbie JW, Symington T: The human adrenal gland with special reference to the vasculature. J Endocrinol 34(4):479–489, 1966. 45. Dong A, Cui Y, Wang Y, et al: (18)F-FDG PET/CT of adrenal lesions. AJR Am J Roentgenol 203(2):245–252, 2014. 46. Dunnick NR, Korobkin M: Imaging of adrenal incidentalomas: Current status. AJR Am J Roentgenol 179(3):559–568, 2002. 47. Elsayes KM, Mukundan G, Narra VR, et al: Adrenal masses: MR imaging features with pathologic correlation. Radiographics 24(Suppl 1):S73–S86, 2004. 48. Erasmus JJ, Patz EF, Jr, McAdams HP, et al: Evaluation of adrenal masses in patients with bronchogenic carcinoma using 18 F-fluorodeoxyglucose positron emission tomography. AJR Am J Roentgenol 168(5):1357–1360, 1997. 49. Escudero MD, Sabater L, Calvete J, et al: Arterial hypertension due to primary adrenal hydatid cyst. Surgery 132(5):894–895, 2002. 50. Evans PD, Miller CM, Marin D, et al: FDG-PET/CT characterization   of adrenal nodules: Diagnostic accuracy and interreader agreement using quantitative and qualitative methods. Acad Radiol 20(8):923–929, 2013. 51. Ferrozzi F, Bova D: CT and MR demonstration of fat within an adrenal cortical carcinoma. Abdom Imaging 20(3):272–274, 1995. 52. Foti G, Faccioli N, Manfredi R, et al: Evaluation of relative wash-in ratio of adrenal lesions at early biphasic CT. AJR Am J Roentgenol 194(6):1484–1491, 2010. 53. Francis IR, Gross MD, Shapiro B, et al: Integrated imaging of adrenal disease. Radiology 184(1):1–13, 1992. 54. Fujiyoshi F, Nakajo M, Fukukura Y, et al: Characterization of adrenal tumors by chemical shift fast low-angle shot MR imaging: Comparison of four methods of quantitative evaluation. AJR Am J Roentgenol 180(6):1649–1657, 2003.

55. Glasgow BJ, Steinsapir KD, Anders K, et al: Adrenal pathology in the acquired immune deficiency syndrome. Am J Clin Pathol 84(5):594– 597, 1985. 56. Glazer DI, Maturen KE, Kaza RK, et al: Adrenal incidentaloma triage with single-source (fast-kilovoltage switch) dual-energy CT. AJR Am J Roentgenol 203(2):329–335, 2014. 57. Gnannt R, Fischer M, Goetti R, et al: Dual-energy CT for characterization of the incidental adrenal mass: Preliminary observations. AJR Am J Roentgenol 198(1):138–144, 2012. 58. Graser A, Johnson TR, Chandarana H, et al: Dual energy CT: Preliminary observations and potential clinical applications in the abdomen. Eur Radiol 19(1):13–23, 2009. 59. Gunn AJ, Seethamraju RT, Hedgire S, et al: Imaging behavior of the normal adrenal on ferumoxytol-enhanced MRI: Preliminary findings. AJR Am J Roentgenol 201(1):117–121, 2013. 60. Gupta RT, Ho LM, Marin D, et al: Dual-energy CT for characterization of adrenal nodules: Initial experience. AJR Am J Roentgenol 194(6):1479–1483, 2010. 61. Haider MA, Ghai S, Jhaveri K, et al: Chemical shift MR imaging of hyperattenuating (>10 HU) adrenal masses: Does it still have a role? Radiology 231(3):711–716, 2004. 62. Halefoglu AM, Bas N, Yasar A, et al: Differentiation of adrenal adenomas from nonadenomas using CT histogram analysis method: A prospective study. Eur J Radiol 73(3):643–651, 2010. 63. Halefoglu AM, Yasar A, Bas N, et al: Comparison of computed tomography histogram analysis and chemical-shift magnetic resonance imaging for adrenal mass characterization. Acta Radiol 50(9):1071– 1079, 2009. 64. Happel B, Heinz-Peer G: Enhancement characteristics of pheochromocytomas. Radiology 238(1):373, author reply 374, 2006. 65. Helck A, Hummel N, Meinel FG, et al: Can single-phase dual-energy CT reliably identify adrenal adenomas? Eur Radiol 24(7):1636–1642, 2014. 66. Hess KR, Varadhachary GR, Taylor SH, et al: Metastatic patterns in adenocarcinoma. Cancer 106(7):1624–1633, 2006. 67. Heylen S, Hubens G, Vaneerdeweg W: Giant adrenal myelolipoma: A case report. Acta Chir Belg 111(2):91–93, 2011. 68. Hiatt JR, Hiatt N: The conquest of Addison’s disease. Am J Surg 174(3):280–283, 1997. 69. Ho LM, Marin D, Neville AM, et al: Characterization of adrenal nodules with dual-energy CT: Can virtual unenhanced attenuation values replace true unenhanced attenuation values? AJR Am J Roentgenol 198(4):840–845, 2012. 70. Ho LM, Paulson EK, Brady MJ, et al: Lipid-poor adenomas on unenhanced CT: Does histogram analysis increase sensitivity compared with a mean attenuation threshold? AJR Am J Roentgenol 191(1):234– 238, 2008. 71. Hricak H, Amparo E, Fisher MR, et al: Abdominal venous system: Assessment using MR. Radiology 156(2):415–422, 1985. 72. Incze JS, Lui PS, Merriam JC, et al: Morphology and pathogenesis of adrenal cysts. Am J Pathol 95(2):423–432, 1979. 73. Israel GM, Korobkin M, Wang C, et al: Comparison of unenhanced CT and chemical shift MRI in evaluating lipid-rich adrenal adenomas. AJR Am J Roentgenol 183(1):215–219, 2004. 74. Jacobson AF, Deng H, Lombard J, et al: 123I-meta-iodobenzylguanidine scintigraphy for the detection of neuroblastoma and pheochromocytoma: Results of a meta-analysis. J Clin Endocrinol Metab 95(6):2596–2606, 2010. 75. Jafri SZ, Francis IR, Glazer GM, et al: CT detection of adrenal lymphoma. J Comput Assist Tomogr 7(2):254–256, 1983. 76. Jex RK, van Heerden JA, Carpenter PC, et al: Ectopic ACTH syndrome. Diagnostic and therapeutic aspects. Am J Surg 149(2):276–282, 1985. 77. Jhaveri KS, Wong F, Ghai S, et al: Comparison of CT histogram analysis and chemical shift MRI in the characterization of indeterminate adrenal nodules. AJR Am J Roentgenol 187(5):1303–1308, 2006. 78. Johnson GL, Hruban RH, Marshall FF, et al: Primary adrenal ganglioneuroma: CT findings in four patients. AJR Am J Roentgenol 169(1):169–171, 1997.

CHAPTER 53  Adrenal Glands 79. Kalra MK, Blake MA, Boland GW, et al: CT features of adrenal pheochromocytomas: Attenuation value and loss of contrast enhancement. Radiology 236(3):1112–1113, 2005. 80. Kawashima A, Sandler CM, Fishman EK, et al: Spectrum of CT findings in nonmalignant disease of the adrenal gland. Radiographics 18(2):393– 412, 1998. 81. Kenney PJ, Robbins GL, Ellis DA, et al: Adrenal glands in patients with congenital renal anomalies: CT appearance. Radiology 155(1):181–182, 1985. 82. Kenney PJ, Wagner BJ, Rao P, et al: Myelolipoma: CT and pathologic features. Radiology 208(1):87–95, 1998. 83. Khan MR, Ajmal S, Saleem T: Giant adrenal endothelial cyst associated with acute and chronic morbidity in a young female: A case report. Cases J 2:8841, 2009. 84. Kim YK, Park BK, Kim CK, et al: Adenoma characterization: Adrenal protocol with dual-energy CT. Radiology 267(1):155–163, 2013. 85. Kloos RT, Gross MD, Francis IR, et al: Incidentally discovered adrenal masses. Endocr Rev 16(4):460–484, 1995. 86. Korobkin M, Brodeur FJ, Francis IR, et al: CT time-attenuation washout curves of adrenal adenomas and nonadenomas. AJR Am J Roentgenol 170(3):747–752, 1998. 87. Korobkin M, Giordano TJ, Brodeur FJ, et al: Adrenal adenomas: Relationship between histologic lipid and CT and MR findings. Radiology 200(3):743–747, 1996. 88. Korobkin M, Lombardi TJ, Aisen AM, et al: Characterization of adrenal masses with chemical shift and gadolinium-enhanced MR imaging. Radiology 197(2):411–418, 1995. 89. Krebs TL, Wagner BJ: MR imaging of the adrenal gland: Radiologicpathologic correlation. Radiographics 18(6):1425–1440, 1998. 90. Kreissig R, Amthauer H, Krude H, et al: The use of FDG-PET and CT for the staging of adrenocortical carcinoma in children. Pediatr Radiol 30(5):306, 2000. 91. Kumagae Y, Fukukura Y, Takumi K, et al: Distinguishing adrenal adenomas from non-adenomas on dynamic enhanced CT: A comparison of 5 and 10 min delays after intravenous contrast medium injection. Clin Radiol 68(7):696–703, 2013. 92. Kumar R, Xiu Y, Mavi A, et al: FDG-PET imaging in primary bilateral adrenal lymphoma: A case report and review of the literature. Clin Nucl Med 30(4):222–230, 2005. 93. Lam KY, Lo CY: Metastatic tumours of the adrenal glands: A 30-year experience in a teaching hospital. Clin Endocrinol (Oxf) 56(1):95–101, 2002. 94. Lattin GE, Jr, Sturgill ED, Tujo CA, et al: From the radiologic pathology archives: Adrenal tumors and tumor-like conditions in the adult: Radiologic-pathologic correlation. Radiographics 34(3):805–829, 2014. 95. Leroy-Willig A, Bittoun J, Luton JP, et al: In vivo MR spectroscopic imaging of the adrenal glands: Distinction between adenomas and carcinomas larger than 15 mm based on lipid content. AJR Am J Roentgenol 153(4):771–773, 1989. 96. Li Y, Sun H, Gao S, et al: Primary bilateral adrenal lymphoma: 2 case reports. J Comput Assist Tomogr 30(5):791–793, 2006. 97. Lingam RK, Sohaib SA, Rockall AG, et al: Diagnostic performance of CT versus MR in detecting aldosterone-producing adenoma in primary hyperaldosteronism (Conn’s syndrome). Eur Radiol 14(10):1787–1792, 2004. 98. Lingam RK, Sohaib SA, Vlahos I, et al: CT of primary hyperaldosteronism (Conn’s syndrome): The value of measuring the adrenal gland. AJR Am J Roentgenol 181(3):843–849, 2003. 99. Mantero F, Arnaldi G: Management approaches to adrenal incidentalomas. A view from Ancona, Italy. Endocrinol Metab Clin North Am 29(1):107–125, 2000. 100. Marin D, Boll DT, Mileto A, et al: State of the art: Dual-energy CT of the abdomen. Radiology 271(2):327–342, 2014. 101. Marin D, Soher BJ, Dale BM, et al: Characterization of adrenal lesions: Comparison of 2D and 3D dual gradient-echo MR imaging at 3 T—Preliminary results. Radiology 254(1):179–187, 2010. 102. Mayo-Smith WW, Boland GW, Noto RB, et al: State-of-the-art adrenal imaging. Radiographics 21(4):995–1012, 2001.

1779

103. Mayo-Smith WW, Lee MJ, McNicholas MM, et al: Characterization of adrenal masses (<5 cm) by use of chemical shift MR imaging: Observer performance versus quantitative measures. AJR Am J Roentgenol 165(1):91–95, 1995. 104. McGowan-Smyth S: Bilateral adrenal haemorrhage leading to adrenal crisis. BMJ Case Rep 2014:2014. 105. Metser U, Miller E, Lerman H, et al: 18F-FDG PET/CT in the evaluation of adrenal masses. J Nucl Med 47(1):32–37, 2006. 106. Mileto A, Nelson RC, Marin D, et al: Dual-energy multidetector CT for the characterization of incidental adrenal nodules: Diagnostic performance of contrast-enhanced material density analysis. Radiology 274(2):445–454, 2015. 107. Mirowitz SA, Gutierrez E, Lee JK, et al: Normal abdominal enhancement patterns with dynamic gadolinium-enhanced MR imaging. Radiology 180(3):637–640, 1991. 108. Morgan HE, Austin JH, Follet DA: Bilateral adrenal enlargement in Addison’s disease caused by tuberculosis. Nephrotomographic demonstration. Radiology 115(2):357–358, 1975. 109. Nakamura S, Namimoto T, Morita K, et al: Characterization of adrenal lesions using chemical shift MRI: Comparison between 1.5 tesla and two echo time pair selection at 3.0 tesla MRI. J Magn Reson Imaging 35(1):95–102, 2012. 110. Neri LM, Nance FC: Management of adrenal cysts. Am Surg 65(2):151– 163, 1999. 111. Ng L, Libertino JM: Adrenocortical carcinoma: Diagnosis, evaluation and treatment. J Urol 169(1):5–11, 2003. 112. Nieman LKaC GB, Jr: Cushing’s syndrome. In DeGroot LJ, editor: Endocrinology, Philadelphia, 1995, Saunders, pp 1741–1770. 113. Noh JJ, Choi SH, Hwang HK, et al: Adrenal cavernous hemangioma:   A case report with review of the literature. JOP 15(3):254–257, 2014. 114. Ogino J, Toda J, Onitsuka S, et al: Idiopathic bilateral adrenal haemorrhage related to acute adrenal insufficiency. BMJ Case Rep 2013:2013. 115. O’Hara SM, Donnelly LF: Intense contrast enhancement of the adrenal glands: Another abdominal CT finding associated with hypoperfusion complex in children. AJR Am J Roentgenol 173(4):995–997, 1999. 116. Otal P, Escourrou G, Mazerolles C, et al: Imaging features of uncommon adrenal masses with histopathologic correlation. Radiographics 19(3):569–581, 1999. 117. Outwater EK, Siegelman ES, Huang AB, et al: Adrenal masses: Correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposed-phase sequences. Radiology 200(3):749–752, 1996. 118. Paling MR, Williamson BR: Adrenal involvement in non-Hodgkin lymphoma. AJR Am J Roentgenol 141(2):303–305, 1983. 119. Park SH, Kim MJ, Kim JH, et al: Differentiation of adrenal adenoma and nonadenoma in unenhanced CT: New optimal threshold value   and the usefulness of size criteria for differentiation. Korean J Radiol 8(4):328–335, 2007. 120. Park SY, Park BK, Kim CK: The value of adding (18)F-FDG PET/CT to adrenal protocol CT for characterizing adrenal metastasis (>/= 10 mm) in oncologic patients. AJR Am J Roentgenol 202(2):W153–W160, 2014. 121. Patel J, Davenport MS, Cohan RH, et al: Can established CT attenuation and washout criteria for adrenal adenoma accurately exclude pheochromocytoma? AJR Am J Roentgenol 201(1):122–127, 2013. 122. Patel SM, Lingam RK, Beaconsfield TI, et al: Role of radiology in the management of primary aldosteronism. Radiographics 27(4):1145–1157, 2007. 123. Pena CS, Boland GW, Hahn PF, et al: Characterization of indeterminate (lipid-poor) adrenal masses: Use of washout characteristics at contrast-enhanced CT. Radiology 217(3):798–802, 2000. 124. Perri M, Erba P, Volterrani D, et al: Adrenal masses in patients with cancer: PET/CT characterization with combined CT histogram and standardized uptake value PET analysis. AJR Am J Roentgenol 197(1):209–216, 2011.

1780

PART II  CT and MR Imaging of the Whole Body

125. Provenzale JM, Ortel TL, Nelson RC: Adrenal hemorrhage in patients with primary antiphospholipid syndrome: Imaging findings. AJR Am J Roentgenol 165(2):361–364, 1995. 126. Quint LE, Glazer GM, Francis IR, et al: Pheochromocytoma and paraganglioma: Comparison of MR imaging with CT and I-131 MIBG scintigraphy. Radiology 165(1):89–93, 1987. 127. Radin R, David CL, Goldfarb H, et al: Adrenal and extra-adrenal retroperitoneal ganglioneuroma: Imaging findings in 13 adults. Radiology 202(3):703–707, 1997. 128. Ramalho M, de Campos RO, Heredia V, et al: Characterization of adrenal lesions with 1.5-T MRI: Preliminary observations on comparison of three in-phase and out-of-phase gradient-echo techniques. AJR Am J Roentgenol 197(2):415–423, 2011. 129. Ramsay JA, Asa SL, van Nostrand AW, et al: Lipid degeneration in pheochromocytomas mimicking adrenal cortical tumors. Am J Surg Pathol 11(6):480–486, 1987. 130. Rao P, Kenney PJ, Wagner BJ, et al: Imaging and pathologic features of myelolipoma. Radiographics 17(6):1373–1385, 1997. 131. Remer EM, Motta-Ramirez GA, Shepardson LB, et al: CT histogram analysis in pathologically proven adrenal masses. AJR Am J Roentgenol 187(1):191–196, 2006. 132. Ricci Z, Chernyak V, Hsu K, et al: Adrenal cysts: Natural history by long-term imaging follow-up. AJR Am J Roentgenol 201(5):1009–1016, 2013. 133. Rockall AG, Babar SA, Sohaib SA, et al: CT and MR imaging of the adrenal glands in ACTH-independent Cushing syndrome. Radiographics 24(2):435–452, 2004. 134. Rodacki K, Ramalho M, Dale BM, et al: Combined chemical shift imaging with early dynamic serial gadolinium-enhanced MRI in the characterization of adrenal lesions. AJR Am J Roentgenol 203(1):99–106, 2014. 135. Rondeau G, Nolet S, Latour M, et al: Clinical and biochemical features of seven adult adrenal ganglioneuromas. J Clin Endocrinol Metab 95(7):3118–3125, 2010. 136. Sangwaiya MJ, Boland GW, Cronin CG, et al: Incidental adrenal lesions: Accuracy of characterization with contrast-enhanced washout multidetector CT–10-minute delayed imaging protocol revisited in a large patient cohort. Radiology 256(2):504–510, 2010. 137. Schieda N, Al Dandan O, Kielar AZ, et al: Pitfalls of adrenal imaging with chemical shift MRI. Clin Radiol 2014. 138. Schindera ST, Soher BJ, Delong DM, et al: Effect of echo time pair selection on quantitative analysis for adrenal tumor characterization with in-phase and opposed-phase MR imaging: Initial experience. Radiology 248(1):140–147, 2008. 139. Schlund JF, Kenney PJ, Brown ED, et al: Adrenocortical carcinoma: MR imaging appearance with current techniques. J Magn Reson Imaging 5(2):171–174, 1995. 140. Schwartz LH, Macari M, Huvos AG, et al: Collision tumors of the adrenal gland: Demonstration and characterization at MR imaging. Radiology 201(3):757–760, 1996. 141. Seo JM, Park BK, Park SY, et al: Characterization of lipid-poor adrenal adenoma: Chemical-shift MRI and washout CT. AJR Am J Roentgenol 202(5):1043–1050, 2014. 142. Shi JW, Dai HZ, Shen L, et al: Dual-energy CT: Clinical application in differentiating an adrenal adenoma from a metastasis. Acta Radiol 55(4):505–512, 2014. 143. Shulkin BL, Thompson NW, Shapiro B, et al: Pheochromocytomas: Imaging with 2-fluorine-18.fluoro-2-deoxy-D-glucose PET. Radiology 212(1):35–41, 1999. 144. Small WC, Bernardino ME: Gd-DTPA adrenal gland enhancement at 1.5 T. Magn Reson Imaging 9(3):309–312, 1991. 145. Smith SM, Patel SK, Turner DA, et al: Magnetic resonance imaging of adrenal cortical carcinoma. Urol Radiol 11(1):1–6, 1989. 146. Sohaib SA, Hanson JA, Newell-Price JD, et al: CT appearance   of the adrenal glands in adrenocorticotrophic hormone-dependent Cushing’s syndrome. AJR Am J Roentgenol 172(4):997–1002, 1999.

147. Song JH, Chaudhry FS, Mayo-Smith WW: The incidental adrenal mass on CT: Prevalence of adrenal disease in 1,049 consecutive adrenal masses in patients with no known malignancy. AJR Am J Roentgenol 190(5):1163–1168, 2008. 148. Song JH, Grand DJ, Beland MD, et al: Morphologic features of 211 adrenal masses at initial contrast-enhanced CT: Can we differentiate benign from malignant lesions using imaging features alone? AJR Am J Roentgenol 201(6):1248–1253, 2013. 149. Storey P, Lim RP, Chandarana H, et al: MRI assessment of hepatic iron clearance rates after USPIO administration in healthy adults. Invest Radiol 47(12):717–724, 2012. 150. Szolar DH, Korobkin M, Reittner P, et al: Adrenocortical carcinomas and adrenal pheochromocytomas: Mass and enhancement loss evaluation at delayed contrast-enhanced CT. Radiology 234(2):479–485, 2005. 151. Takeuchi S, Balachandran A, Habra MA, et al: Impact of (18)F-FDG PET/CT on the management of adrenocortical carcinoma: Analysis of 106 patients. Eur J Nucl Med Mol Imaging 41(11):2066–2073, 2014. 152. Tessonnier L, Sebag F, Palazzo FF, et al: Does 18F-FDG PET/CT add diagnostic accuracy in incidentally identified non-secreting adrenal tumours? Eur J Nucl Med Mol Imaging 35(11):2018–2025, 2008. 153. Trerotola SO, Asmar M, Yan Y, et al: Failure mode analysis in adrenal vein sampling: A single-center experience. J Vasc Interv Radiol 25(10):1611–1619, 2014. 154. Tsushima Y, Ishizaka H, Matsumoto M: Adrenal masses: Differentiation with chemical shift, fast low-angle shot MR imaging. Radiology 186(3):705–709, 1993. 155. Untch BR, Shia J, Downey RJ, et al: Imaging and management of a small cell lung cancer metastasis/adrenal adenoma collision tumor: A case report and review of the literature. World J Surg Oncol 12:45, 2014. 156. Varghese JC, Hahn PF, Papanicolaou N, et al: MR differentiation of phaeochromocytoma from other adrenal lesions based on qualitative analysis of T2 relaxation times. Clin Radiol 52(8):603–606, 1997. 157. Venkatanarasimha N, Roobottom C: Intense adrenal enhancement: A feature of hypoperfusion complex. AJR Am J Roentgenol 195(1):W82, author reply W3-4, 2010. 158. Vick CW, Zeman RK, Mannes E, et al: Adrenal myelolipoma: CT and ultrasound findings. Urol Radiol 6(1):7–13, 1984. 159. Vikram R, Yeung HD, Macapinlac HA, et al: Utility of PET/CT in differentiating benign from malignant adrenal nodules in patients with cancer. AJR Am J Roentgenol 191(5):1545–1551, 2008. 160. Vincent JM, Morrison ID, Armstrong P, et al: The size of normal adrenal glands on computed tomography. Clin Radiol 49(7):453–455, 1994. 161. Watson TD, Patel SJ, Nardi PM: Case 121: Familial adrenocorticotropin-  independent macronodular adrenal hyperplasia causing Cushing syndrome. Radiology 244(3):923–926, 2007. 162. Welch TJ, Sheedy SP, Sheedy PF, II: Adrenal glands. In Haaga JR, Boll DT, editors: CT and MRI of the whole body, ed 5, Philadelphia, 2008, Mosby, pp 1813–1862. 163. Wilms GE, Baert AL, Kint EJ, et al: Computed tomographic findings in bilateral adrenal tuberculosis. Radiology 146(3):729–730, 1983. 164. Wilson DA, Muchmore HG, Tisdal RG, et al: Histoplasmosis of the adrenal glands studied by CT. Radiology 150(3):779–783, 1984. 165. Wong KK, Arabi M, Bou-Assaly W, et al: Evaluation of incidentally discovered adrenal masses with PET and PET/CT. Eur J Radiol 81(3):441–450, 2012. 166. Wong TZ, Paulson EK, Nelson RC, et al: Practical approach to diagnostic CT combined with PET. AJR Am J Roentgenol 188(3):622– 629, 2007. 167. Young WF, Jr: Management approaches to adrenal incidentalomas. A view from Rochester, Minnesota. Endocrinol Metab Clin North Am 29(1):159–185, 2000. 168. Young WF, Jr: Minireview: Primary aldosteronism—Changing concepts in diagnosis and treatment. Endocrinology 144(6):2208–2213, 2003. 169. Young WF, Jr: Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 356(6):601–610, 2007.

54  Kidney Shweta Bhatt and Vikram S. Dogra

Computed tomography (CT) is a rapid, easily performed, and safe diagnostic imaging technique that provides valuable information about a wide spectrum of renal disorders. CT is highly accurate for determining the nature and extent of renal masses and plays a valuable role in assessing patients with renal cystic disease, renal trauma, renal infections, renal blood flow disturbances, and hydronephrosis of unknown cause. Multidetector (MD)CT (multislice CT) promises to provide even more rapid assessment of the kidneys and higher accuracy in the evaluation of renal masses and renal blood vessels than is provided by single-detector helical CT. The technology of magnetic resonance imaging (MRI) has advanced rapidly, and MRI is now almost equivalent to CT in detecting and characterizing renal masses. However, because CT is less expensive, quicker, and more generally available, renal MRI is mainly used for evaluating patients in whom CT findings are equivocal or in whom contrast-enhanced CT is contraindicated because of previous reactions to intravenous (IV) iodinated contrast medium or the presence of renal failure.

NORMAL CT ANATOMY The kidneys are surrounded by perinephric fat, which in turn is enveloped by a dense connective tissue sheath called the renal fascia (Fig. 54-1). The anterior renal fascia (Gerota’s fascia) covers the kidney anteriorly, whereas the posterior renal fascia (Zuckerkandl’s fascia) covers the kidney posteriorly (see Fig. 54-1). The renal fascial layers divide the general retroperitoneal space into three compartments extending from the diaphragm to the pelvic brim—the anterior pararenal space, the perinephric space, and the posterior pararenal space (see Fig. 54-1).217 Because it differentiates between fat and fascial tissue, CT usually shows the renal fascia and major extraperitoneal compartments (Fig. 54-2).

Perinephric Space The perinephric space contains the kidney, adrenal gland, inferior vena cava (IVC), lower aorta, renal pelvis, proximal ureter, renal blood vessels, renal capsular vessels, and perinephric fat (see Fig. 54-1). It is bounded by the anterior and posterior renal fascial layers and is demarcated by their sites of fusion. Above the adrenal glands the two fascial layers fuse and adhere firmly to the diaphragmatic fascia; laterally the layers fuse behind the ascending or descending colon to form the lateroconal fascia (see Fig. 54-1). Medially the anterior renal fascia blends into the connective tissue near the midline (see Fig. 54-1)88,217; the posterior renal fascia fuses with the psoas or quadratus lumborum fascia. The renal fascial cone extends inferior to the kidney, forming a caudal extension of the main perinephric space containing the proximal ureter and gonadal vessels (see Fig. 54-1).217 Right and left

perirenal spaces communicate with each other anterior to the IVC and below the level of the renal hilum.16,323 The perinephric space is divided into multiple compartments by fibrous lamellae, the bridging septa (Fig. 54-3).3,163 Some arise from the renal capsule and extend to the anterior and posterior renal fascia. Others are attached only to the renal capsule and are arranged nearly parallel to the renal surface (see Fig. 54-3). One of the more constant of these is the posterior renorenal bridging septum. It arises from the renal capsule at its posteromedial aspect and runs nearly parallel to the posterior surface of the kidney, inserting into the posterolateral aspect of the renal capsule. Still other bridging septa directly connect the anterior and posterior leaves of the renal fascia.163 The bridging septa determine the distribution of blood, pus, or urine collections in the perinephric space.3,163

Kidneys On CT the transverse contour of the kidney is smooth and oval, with an anteromedial break in the renal outline at the hilus where the vascular pedicle enters (Fig. 54-4). The renal sinus is a potential space contained by the renal parenchyma. The space is filled with fatty tissue and contains the renal arteries, veins, lymphatics, and pelvicalyces. The renal sinus fat is directly continuous with the perinephric fat through the renal hilus. On nonenhanced CT scans the normal renal parenchyma has an attenuation value of 30 to 50 Hounsfield units (HU), depending on patient hydration, and the cortex and medulla show no visible density differences. After a rapid IV injection of contrast medium, helical CT scanning shows a cortical nephrogram in which there is clear demarcation of the renal cortex and columns of Bertin from the renal medulla (Fig. 54-5). However, the cortical nephrogram phase is transient, and a homogeneous tubular nephrogram rapidly develops (see Fig. 54-5). The renal veins are usually densely opacified during dynamic CT scanning and are shown anterior to the renal pelvis as tubular structures joining the IVC. However, during dynamic scanning the IVC often shows significantly less contrast enhancement than the renal veins; this difference is due to mixing of opacified blood from the renal vein and unopacified blood from the lower extremities (Fig. 54-6). This appearance should not be mistaken for vena caval thrombosis or tumor extension. The longer left renal vein crosses the retroperitoneum between the aorta posteriorly and the superior mesenteric vessels anteriorly and joins the IVC at the level of the uncinate process of the pancreas. The right renal vein usually has a shorter more oblique course (see Figs. 54-4 and 54-6). The renal arteries are located posterior to the renal veins and are usually smaller. Normal renal arteries are seen by helical (spiral) CT performed during breath holding with the use of narrow scan collimation.274

1781

1782 AO

PART II  CT and MR Imaging of the Whole Body DS PP ARF

DC D

1

2 LRP PS QL

A

LC

3

LA

LK

LK 3

PRF Level A

AO UR PP ARF DC

Level B LC

1

2 LK

PS

FC

2

ARF PRF

Level C

QL

QL

B

PS

3 PRF

TF

AO UR ARF PP IL 1 DC 2 3 QL

PS

C

FIG 54-2  Normal CT anatomy. The posterior renal fascia (open arrows)

LC

PRF

D

separates the perinephric and posterior pararenal spaces. The anterior renal fascia (straight white arrow) separates the perinephric and anterior pararenal spaces. The lateroconal fascia (black arrow) extends lateral to the descending colon (curved arrow).

FIG 54-1  Transverse (A to C) and parasagittal (D) sections of the left flank. Sections A to C show the cross-sectional anatomy at the three levels indicated in D—the level of the renal hilus (A), the lower renal pole (B), and above the iliac crest (C). In D, the approximate positions of the psoas and quadratus lumborum muscles (stippled areas) are indicated by dashed lines. AO, aorta; ARF, anterior renal fascia; D, diaphragm; DC, descending colon; DS, deeper stratum of renal fascia; FC, fibrous capsule of kidney; IL, iliacus muscle; LA, left adrenal; LC, lateroconal fascia; LK, left kidney; LRP, left renal pelvis; PP, parietal peritoneum; PRF, posterior renal fascia; PS, psoas muscle; QL, quadratus lumborum muscle; TF, transversalis fascia; UR, ureter; 1, anterior pararenal space; 2, perinephric space; 3, posterior pararenal space. (From Feldberg MA, et al: Psoas compartment disease studied by computed tomography: Analysis of 50 cases and subject review. Radiology 148:505–512, 1983.)

Developmental renal anomalies and minor anatomic variants are commonly encountered and are usually evaluated by excretory urography. However, CT is sometimes necessary for further evaluation when urographic findings are confusing (Figs. 54-7 and 54-8). For example, on excretory urography, renal sinus lipomatosis may masquerade as a mass lesion in the renal sinus. CT reveals the benign nature of the process by showing that the renal sinus is occupied by tissue with a fat-attenuation value. Normal variants, such as a dromedary hump in the left kidney and prominent columns of Bertin, may simulate renal masses on excretory urography. CT easily recognizes these variants, which enhance to the same degree as normal renal parenchyma after IV administration of contrast medium.

IMAGING TECHNIQUES CT Technique Single-detector and multidetector spiral CT have dramatically refined the diagnostic evaluation of renal disorders by allowing rapid image acquisition through the entire kidney during various phases of contrast enhancement after administration of a single bolus of IV contrast material.160,314 The technique should be tailored according to the specific clinical problem involved. However, some aspects are common to

K

FIG 54-3  Bridging septa in perinephric space. Multiple delicate fibrous bridging septa traverse the upper part of the left perinephric space.  K, kidney.

all dedicated renal CT studies. Because unopacified bowel loops may simulate perinephric masses and retroperitoneal adenopathy, patients undergoing renal CT should receive oral contrast medium. CT protocol for evaluation of the kidneys consists of both nonenhanced and contrast-enhanced CT scans obtained in suspended respiration to overcome the motion artifact. To avoid artifactual differences in attenuation values, the same peak kilovoltage, milliampere-second setting, section thickness, and field of view should be used for both precontrast and postcontrast scans when small renal masses are being evaluated. The accuracy of attenuation values should also be tested by measuring the attenuation value of the gallbladder contents before and after IV administration of contrast medium. Apparent enhancement of the gallbladder contents suggests that mild enhancement of a renal mass may be spurious.34

CHAPTER 54  Kidney

1783

p v

a

FIG 54-4  Normal CT renal anatomy. The longer left renal vein (open arrow) passes between the aorta (a) posteriorly and the superior mesenteric artery (curved arrow) and vein (arrowhead) anteriorly. It joins the inferior vena cava (v) at the level of the uncinate process of the pancreas (p). The right renal vein (arrow) is shorter.

A

Images should be acquired with 1- to 3-mm collimation, subsecond scan time, and a pitch of up to 2 : 1 to allow coverage of the area of interest in a single breath hold. Multislice scanners allow acquisition of 1- to 2.5-mm slices through the region of interest in a single breath hold. Nonenhanced scans permit contrast enhancement of a renal lesion to be measured and also ensure that renal parenchymal calcifications, renal calculi, renal and perinephric hemorrhage and fat, and calcification in a renal mass will not be obscured by contrast medium. Nonenhanced scans also help differentiate hyperdense cyst from renal solid tumor.290

Corticomedullary Phase.  Administration of an IV contrast medium

B

is a fundamental requirement for CT evaluation of most renal lesions. Enhancement of a mass indicates that the lesion is vascular and is therefore possibly a neoplasm. Contrast medium should be administered rapidly with a mechanical injector through an antecubital vein as a 150-mL bolus containing 40 to 45 g of iodine at a rate of 2 to 4 mL/sec. The corticomedullary phase occurs between 25 and 70 seconds after the start of contrast administration. The corticomedullary phase begins as contrast material enters the cortical capillaries and peritubular spaces and filters into the proximal cortical tubules.278 Renal cortex can be differentiated from renal medulla at this stage because (1) the vascularity of the cortex is greater than that of the medulla, and (2) contrast material has not yet reached the distal aspect of the renal tubules (Fig. 54-9).82,278 The resultant CT nephrogram has been termed a cortical nephrogram. Corticomedullary-phase images should always be obtained when information about the renal vasculature is desired or when there is a possibility that a detected renal mass may represent an aneurysm or arteriovenous malformation or a fistula.355 Maximal opacification of the renal vein and arteries occurs during this phase, allowing confident diagnosis of tumor extension to the vein.

FIG 54-5  Normal CT anatomy. A, The corticomedullary phase demon-

Nephrographic Phase.  This begins as contrast material proceeds

CT urography is a relatively new technique that is progressively replacing conventional IV urography (IVU). It is being increasingly used for indications such as hematuria of unknown cause because of its ability to examine the entire genitourinary tract in a single study. However, it remains limited in evaluation of the urothelium compared with IVU because of its lower spatial resolution.

from the cortical vessels and extracellular-interstitial space and enters the loops of Henle and collecting tubules. A homogeneous or tubular nephrogram results in which corticomedullary differentiation is lost (Fig. 54-10).25 The nephrographic phase starts about 80 seconds and lasts up to 180 seconds after the start of injection, and it offers the best

strates dense enhancing cortex with minimal enhancement of the renal medulla. B, After a brief delay the parenchymal enhancement becomes uniform, resulting in the nephrographic phase; in the late nephrographic phase, calyces may be opacified.

opportunity for discrimination between the normal renal medulla and a renal mass.314,355 The nephrographic phase is the most valuable for detecting renal masses and characterizing indeterminate lesions.26,53,314

Excretory Phase.  Approximately 180 seconds after the start of contrast injection, the excretory phase begins. The contrast material is excreted into the collecting system, so the attenuation of the nephrogram progressively decreases (Fig. 54-11). This phase is occasionally helpful to better delineate the relationship of a centrally located mass with the collecting system.290 This phase is also useful for evaluating urothelial masses.

CT Urography

1784

PART II  CT and MR Imaging of the Whole Body

P

FIG 54-8  Agenesis of the right kidney and congenital left ureteropelvic FIG 54-6  Normal renal vasculature. In the corticomedullary phase, the renal veins (arrow) and arteries (arrowhead) have maximal opacification. The inferior vena cava (long arrow) appears less dense because of mixing of opacified blood from the renal vein and unopacified blood from the lower extremities. This appearance should not be confused with vena caval thrombus.

junction obstruction in an infant. The left renal pelvis (P) is markedly dilated, and there is also calyceal dilation (arrowheads).

FIG 54-9  Corticomedullary phase. The corticomedullary phase demonstrates dense enhancing cortex with minimal enhancement of the renal medulla.

A

B FIG 54-7  Horseshoe kidney. A, Nonenhanced CT scan shows staghorn calculi involving both elements of a horseshoe kidney. B, Contrastenhanced CT scan shows malrotation of both kidneys with the renal pelves directed anteriorly. The kidneys are joined by an isthmus (arrow) of functioning renal parenchyma.

FIG 54-10  Nephrographic phase. The renal parenchyma enhances uniformly, and corticomedullary differentiation is lost.

CHAPTER 54  Kidney

1785

FIG 54-11  Excretory phase. The attenuation of the nephrogram progressively decreases, and calyces demonstrate contrast. The left kidney shows renal carcinoma.

There are two major approaches for performing CT urography. Hybrid CT urography is a combination of CT and IVU that uses  projection radiographs along with acquisition of CT images after IV contrast injection. CT-only urography does not involve projection radiographs and uses CT alone to obtain two- and three-dimensional (2D and 3D) reformatted images to produce images similar to those obtained with IVU.

CT Hybrid Urography.  This technique combines the advantages of CT and IVU in imaging the urinary collecting system; it therefore does not require postprocessing of the CT images to produce coronal reformatted images.4 This technique may require imaging of the patient in two different locations. There are several problems associated with moving the patient between different procedure rooms, such as additional time and resources and possible insufficient pyelocalyceal distention at the time of the radiography.4,148 Therefore an alternative approach has been developed at the Mayo Clinic in an attempt to solve this problem. An MDCT scanner configured with a special tabletop apparatus is used to obtain both CT and radiographic images at the same location without moving the patient.339 This tabletop is capable of receiving a combined slip-on grid/standard film cassette. This system requires installation of a ceiling-mounted x-ray tube above the CT table and attachment of an auxiliary CT tabletop with a hollow bay under the patient surface in which to place the radiographic cassette. With this technique, an abdominal radiograph, a nonenhanced renal CT, and a multiphasic contrast-enhanced renal CT scan, followed by overhead excretory urographic and postvoid radiographs, are obtained. An alternative method of obtaining projection images without moving the patient from the CT table is the use of the CT scanned projection radiography (SPR) technique (scout view, topogram or scanogram). The spatial resolution of CT SPR is inferior to that of conventional radiography, whereas the contrast resolution of opacified structures is similar to that of conventional radiography.205 This CT urography approach, the combination of helical CT scans and enhanced CT SPR urographic images, is attractive because modification of the CT tabletop and installation of a ceiling-mounted x-ray tube is not required; therefore this technique can be performed on any MDCT equipment.

FIG 54-12  Normal CT urography. Coronal MIP image of the CT urogram in the excretory phase demonstrates the opacified proximal and middle ureters.

Preprocedure bowel preparation helps reduce gas and fecal material, enhancing visualization of the collecting system. IVU abdominal compression is applied after IV contrast medium injection for better opacification and distention of the intrarenal collecting system and ureter. It is important to use a low-osmolar iodinated contrast medium.81 Ten-minute decompressed film images help visualize almost the entire ureters. Twenty-minute and postvoiding films are useful for bladder evaluation.148

CT-Only CT Urography.  This technique is based on acquisition of nonenhanced and enhanced CT scans of the abdomen and pelvis, including the essential acquisition of thin-section helical CT scans of the urinary tract during the excretory phase of enhancement.148 No bowel preparation is necessary for this type of CT urography. Both 2D and 3D multiplanar reconstruction (MPR) images are produced from axial source images obtained during the excretory phase (Figs. 54-12 and 54-13). Optimal distention and opacification of the collecting system are the keys to detection of urothelial abnormalities. Abdominal compression improves opacification of the collecting system compared with CT scans without compression.41 Caoili and associates described four-phase MDCT urography as comprising a nonenhanced scan, a nephrographic-phase scan with abdominal compression, and two excretory-phase scans, one obtained 200 seconds after injection of contrast material with compression and the other at 300 seconds after release of compression.41 Renal, pelvic, ureteral, and bladder abnormalities are better visualized on compression-release excretory-phase images. The 3D images are particularly helpful for diagnosis of papillary abnormalities such as renal tubular ectasia and papillary necrosis (Fig. 54-14). Alternative techniques can be used to achieve optimal visualization of the collecting systems, such as normal saline chase and diuretic injection. CT urography with supplemental saline administration can

1786

PART II  CT and MR Imaging of the Whole Body significantly improve opacification of the distal ureters. Patient position (supine or prone) during CT examination does not affect ureteral visualization.206 The distal ureters are the most difficult segments to opacify and have a better chance to opacify if a saline infusion technique is used (250-mL saline chase after IV contrast administration). IV injection of low-dose diuretics (10 mg of furosemide) before IV contrast injection allows less dense, homogeneous opacification of the collecting system.225 Alternatively, oral water (1000 mL within 15-20 minutes) before the examination can cause sufficient opacification of the calyces and ureters in most instances.289 Evaluation of axial CT images (source images) with a wide window setting is important for accurate diagnosis.157 Coronal or oblique MPR images help define the location and extent of the lesions shown on axial images (Fig. 54-15). Maximum intensity projection (MIP) images resemble conventional IVU images and may be preferred by physicians. The protocol for CT urography used at the authors’ institution is summarized in Table 54-1.

MRI Techniques

FIG 54-13  Normal CT urography. 3D volume-rendered image of the excretory phase of the CT urogram demonstrates the kidneys, ureters, and bladder.

Gadolinium (Gd)-enhanced renal MRI is an attractive alternative to other techniques for evaluation of native and transplanted kidneys, because it can be used in patients with iodine allergies.268 Recent studies have shown a possible association of Gd-based MRI contrast agent administration in patients with renal insufficiency with development of a condition called nephrogenic systemic fibrosis (NSF). Clinical data from 2007-2009 showed less than 1,200 reports of NSF. It involved multiple Gadolinium-based contrast agents (GBCAs). There has been coincident decrease in new NSF cases with screening and avoidance of high-risk GBCAs. Use of Gd-based MRI contrast agent in patients with

A

B FIG 54-14  Papillary necrosis. Coronal (A) and sagittal (B) MIP images of the left kidney demonstrate a filling defect in the superiormost calyx (arrow) secondary to sloughing of the papilla into the calyx.

CHAPTER 54  Kidney

1787

A

B FIG 54-15  Ureteric injury. A, Axial CT scan of the pelvis in the excretory phase demonstrates extravasation of contrast (arrow) from the left distal ureter. B, Corresponding 3D volume-rendered image of the study shows the exact location of the ureteric injury (arrow). (From Kocakoc E, et al: Renal multidetector row CT. Radiol Clin North Am 43:1021–1047, 2005.)

TABLE 54-1  CT Urography Protocol Phase

Timing

Range

Slice Thickness

What to Detect

Nonenhanced Nephrogram* Excretory*†

Precontrast 100 sec 5-8 min

Lung bases to pubic symphysis Lung bases to pubic symphysis Lung bases to base of bladder

5 mm 3 mm 2 or 3 mm

Calculi, hemorrhage/hemorrhagic cysts Renal tumors, renal vein thrombosis Papillary necrosis, urothelial carcinoma

*125 mL IV contrast medium followed by 250 mL saline administration. † MPR of excretory-phase images using MIP and volume rendering.

renal insufficiency is not routinely recommended.320 Although this relationship is still unclear, it has significantly decreased the number of contrast-enhanced MRI studies being done in patients with renal failure who are otherwise unable to receive CT IV contrast because of increased risk of contrast-induced nephropathy.39,110,164,321 NSF, initially called nephrogenic fibrosing dermopathy because of its dominant cutaneous findings, is an acquired idiopathic systemic disorder that is restricted to patients with renal disease, particularly those undergoing dialysis. Patients with severe renal failure, whether acute or chronic, and those with acute renal failure of any severity who have hepatorenal syndrome or are in the perioperative period of liver transplantation are at greater risk of developing NSF. Characteristic cutaneous changes consist of indurated plaques and papules, most commonly on the extremities and trunk, with almost no involvement of the face.55

The condition is debilitating and potentially fatal. Diagnosis of NSF is made on the basis of histopathologic findings, which include thickened collagen bundles with surrounding clefts, mucin deposition, and a proliferation of fibroblasts and elastic fibers, with complete absence of signs of inflammation.55 Besides dermatologic changes, NSF also results in fibrosis of other organs such as skeletal muscle, myocardium, lungs, kidneys, and testes.101 Several studies have demonstrated an association of NSF with prior Gd administration, raising the possibility of Gd as the etiologic factor in this disorder.110 Radiologists should avoid using Gd-based contrast agents with high-risk patients unless the diagnostic information is essential and cannot be obtained any other way. The FDA recommends prompt dialysis for patients already receiving hemodialysis after administration of a Gd-based contrast agent. It is unknown whether

1788

PART II  CT and MR Imaging of the Whole Body

dialysis prevents NSF, but data indicate that it does speed up elimination of the agent from the body. There is no effective treatment for NSF; the condition leads to increased morbidity and mortality and is an absolute contraindication to contrast administration in severe renal insufficiency (glomerular filtration rate < 30 mL/min). Dynamic Gd-enhanced gradient echo imaging achieves coverage of the entire kidney during breath holding, thereby reducing respirationinduced phase artifacts.285 Gd-enhanced fat-suppressed T1-weighted spin echo technique images the kidneys with significantly fewer  artifacts than does conventional spin echo imaging.286 The ability of dedicated renal CT scanning and Gd-enhanced MRI to detect and characterize renal lesions greater than 1 cm is similar, although MRI is superior in detecting polar lesions because of its ability to image directly in nonaxial planes.267 Because CT is more generally available, quicker, and less expensive than MRI, it is usually preferred to MRI for evaluating patients with suspected renal masses. MRI is the preferred modality, however, for determining the cephalic extent of an intracaval tumor in a patient with renal cell carcinoma (RCC) when this finding is not adequately documented by CT.251 Other indications for renal MRI are differentiation between hemorrhagic renal cysts and renal neoplasms, characterization of small renal masses that are indeterminate on CT or ultrasonography (US),34 evaluation of renal donors,360 visualization of renal blood vessels (MR angiography [MRA]),279 and evaluation of transplanted kidneys.126 The MRI pulse sequences and the precise imaging parameters used depend on the nature of the available MRI equipment and on the clinical problem. Conventional spin echo sequences may be used for evaluating the kidneys; the development of newer and faster imaging techniques such as fast-spin echo and gradient echo imaging are preferred. T1-weighted gradient echo in-phase and opposed-phase images from the top of the liver to the inferior pole of the kidney are obtained to study the anatomy and also evaluate the liver for any metastatic disease from renal tumor. This technique also allows assessment of the presence of fat in the renal lesion. A T2-weighted fast-spin echo sequence with fat saturation is obtained in the axial plane corresponding to the T1-weighted gradient sequence. Coronal T2-weighted (very long echo time [TE], such as half-Fourier acquisition single-shot turbo-spin echo [HASTE] and single-shot fast-spin echo [SSFSE]) images are also acquired to identify renal cysts. T1- and T2-weighted sequences help identify hemorrhagic cysts (Fig. 54-16). Gradient echo images are helpful for evaluating renal masses.

Precontrast images of the kidneys are obtained first (Fig. 54-17). A 2D or 3D gradient echo breath-hold sequence with and without Gd enhancement is obtained in the arterial phase (corticomedullary phase; 20 seconds), venous phase (nephrographic phase; 70 seconds), and delayed phase (excretory phase; 180 seconds). Gadopentetate dimeglumine is injected intravenously by power injector in a dose of 0.1 to 0.2 mmol/kg over 5 seconds, with the patient positioned in the bore of the magnet. The coronal plane is often advantageous because it allows evaluation of both kidneys, renal vessels, and the IVC.251 Imaging is started during breath holding at 20 seconds after a rapid saline flush of the IV catheter has been performed. Images are also obtained at 70 seconds and 3 minutes after the start of injection. Corticomedullary-phase imaging demonstrates corticomedullary differentiation and also significant enhancement of the renal vein blood (see Fig. 54-17). This phase is optimal for assessing tumor extension to the veins. The nephrographic phase is the best sequence to detect a renal mass. Subtraction imaging is now being increasingly used in abdominal imaging, particularly in the kidneys. This technique involves the digital subtraction of a nonenhanced T1-weighted sequence from a contrast (Gd)-enhanced T1-weighted sequence, which cancels the native T1 signal and leaves behind only the enhancing portions as visible on the subtracted image.221 Differentiation of complex renal cysts from enhancing renal masses is often difficult on routine MRI sequences. A complex renal cyst or hemorrhagic cyst appears as high signal intensity on T1-weighted sequences because of the intracellular or extracellular methemoglobin, and evaluation for enhancement can thus be difficult on Gd-enhanced sequences. Subtraction images can prove extremely useful in identifying true enhancement in complex renal cysts (Fig. 54-18). However, this technique is prone to misregistration artifact and image degradation. For an accurate subtraction image, it is imperative that all parameters for nonenhanced and contrast-enhanced sequences remain the same, including patient position in the scanner. A slight change in any of the imaging parameters or patient position may result in the aforementioned artifacts.

Diffusion-Weighted Imaging (DWI) and Perfusion-Weighted Imaging (PWI) DWI and PWI are expanding the role of MRI in management of renal cell cancers. These techniques have been shown to be useful in renal lesion detection, characterization as benign or malignant, as well as

10cr

10

A

B FIG 54-16  Hemorrhagic cyst. A, Axial T1-weighted gradient image demonstrates a small right renal upper pole cyst with increased signal (arrow). B, Corresponding T2-weighted image demonstrates increased signal (arrow) compared with the renal cortex. Previously diagnosed fibrous dysplasia of the left ninth rib is an incidental finding.

SP

CHAPTER 54  Kidney

A

B

C FIG 54-17  Normal renal MRI. A T1-weighted (gradient opposed-phase) image (A) and a T2-weighted axial image (B) demonstrate the normal appearance of the kidneys. C, T1-weighted axial Gd-enhanced image reveals normal enhancing renal parenchyma. Renal vein (arrow) and renal artery (arrowhead) are also seen.

A

B

C FIG 54-18  Subtraction imaging in MRI. A, Axial T1-weighted image of the left kidney demonstrates a left renal lesion (arrow). B, T1-weighted axial Gd-enhanced image shows no discernible enhancement of the left renal lesion (arrow). C, Subtraction image (subtraction of A from B) clearly depicts enhancement of the renal lesion (arrow), suggestive of neoplasm. This was pathologically confirmed to be a papillary carcinoma.

1789

1790

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 54-19  Simple renal cyst. Nonenhanced (A) and contrast-enhanced (B) CT scans show that the mass is round with a smooth margin and distinct margination from adjoining renal parenchyma. It has a homogeneous water density, does not enhance (pre- and postcontrast attenuation value = 7 HU), and has no discernible wall thickness.

renal cancer subtype determination. Acute pyelonephritis can be easily identified by DWI compared to normal anatomic imaging without IV contrast administration. Pyonephrosis can be identified by demonstrating decreased apparent diffusion coefficient (ADC) values within the dilated collecting system, thus differentiating it from uncomplicated hydronephrosis. Chronic renal failure also demonstrated overall decreased ADC values compared to normally functioning kidneys. Any MRI sequence can be used to provide DWI by adding two equally large but opposite magnetic field gradients between an excitation pulse and readout. For renal DWI a combination of anterior and posterior phased-array body coils should ideally be used, which allows a high signal-to-noise ratio and imaging in any plane.319 Patient motion is detrimental to image quality in DWI, and hence efforts should be made to reduce motion artifacts to minimal. These include patient movement, respiratory motion, and bowel peristalsis. Since kidneys demonstrate a significant degree of motion during  respiration, it may potentially lead to volume averaging effects. To overcome or reduce these problems a breath-hold acquisition, a respiratory-triggering approach, or a navigator-controlled technique is recommended for DWI of kidneys. DWI of the kidneys should be ideally performed in the coronal plane. It gives a better anatomic view of the kidneys and is acquired in a shorter time, thus reducing motion artifacts. The choice of optimal b values for renal DWI remains difficult. There is as yet no real consensus about the optimum choice of b values for renal examinations, but a number of guidelines are provided in a recent consensus report on DWI as a cancer biomarker.230

PATHOLOGIC CONDITIONS Renal Cysts Renal cysts usually represent notably dilated nephrons or collecting ducts.97 A cystic kidney is a kidney with three or more cysts. The term renal cystic disease refers to any disorder that results from the presence of multiple renal cysts.

Simple Cysts.  Simple cysts are the most common renal masses. Most are clinically insignificant, being discovered incidentally at autopsy or on imaging studies. Their frequency increases with age, with only rare examples reported in children. Although the cause of renal cysts is

FIG 54-20  Simple renal cyst. Lesion (arrow) has an attenuation value of 9 HU and was detected during survey CT scan performed only with IV contrast enhancement. Even though a nonenhanced scan is unavailable, a cyst can be diagnosed with confidence because of the low attenuation value, smooth rounded shape, sharp margination and demarcation from surrounding renal parenchyma, and thin wall where the lesion projects beyond the renal outline.

unknown, their frequent occurrence in older patients suggests that they are acquired lesions.240 Although most simple cysts are asymptomatic, they occasionally cause pressure symptoms because of their size, and cyst hemorrhage or infection may cause pain and hematuria. Simple renal cysts may be solitary or multiple and are commonly bilateral. US is usually considered the most appropriate next investigation for further evaluation of a presumed renal cyst discovered on excretory urography. However, renal cysts are currently often found on abdominal CT scanning performed for nonrenal complaints. Such cysts may vary in size from less than 1 cm to 10 to 15 cm (Figs. 54-19 and 54-20). CT features.  The accuracy of CT diagnosis of a simple renal cyst approaches 100% if a renal mass strictly fulfills the following wellestablished criteria: (1) sharp margination and demarcation from surrounding renal parenchyma, (2) a smooth thin wall, (3) homogeneous

CHAPTER 54  Kidney water-density content with an attenuation value of 0 to 20 HU, and (4) no enhancement after IV administration of contrast medium (see Fig. 54-19).31,204 If a renal mass fulfills these CT criteria, no further evaluation is necessary. In routine CT performed for nonrenal complaints, usually only IV contrast-enhanced scanning is performed; therefore the presence or absence of contrast enhancement in renal masses often cannot be determined.31 A simple cyst can usually be diagnosed even without this information, however, if the other CT criteria listed previously are fulfilled (see Fig. 54-20).31 If CT findings are atypical or if the patient is referred because of symptoms such as hematuria, nonenhanced scans should also be obtained. Obtaining both sets of scans permits contrast enhancement to be measured and also precludes misinterpretation of nonenhancing high-density renal cysts as solid renal masses.31 Although CT is accurate in diagnosing simple renal cysts, interpretation of CT findings has some potential pitfalls. Small renal cysts may be volume averaged with normal renal tissue, causing spuriously high attenuation values,204 or falsely high attenuation values may be caused by streak artifacts.34 In these circumstances, US should be performed to determine whether the lesion is a cyst or a neoplasm. If US findings are indeterminate, dedicated renal CT with special attention to the lesion should be performed.31

A

1791

MRI features.  The use of Gd-enhanced MRI, particularly with gradient echo imaging techniques applied during breath hold, is helpful in evaluating renal cystic lesions in patients with contraindications to iodinated IV contrast medium (e.g., chronic renal failure, previous reactions to contrast media).268,286 Even a small simple renal cyst can be diagnosed on contrast-enhanced MRI if it fulfills the following criteria: (1) sharp margination and demarcation from surrounding renal parenchyma, (2) a smooth thin wall, (3) homogeneous contents with signal characteristics of water (such lesions have low signal intensity on T1-weighted spin echo images and high signal intensity on T2-weighted spin echo images), and (4) no enhancement after IV injection of Gd-DTPA (Fig. 54-21).286 Subsecond T2-weighted pulse sequences (SSFSE on GE machines and its equivalent on other machines) are better suited for detection of simple renal cysts, especially those smaller than 1 cm.218

Atypical Cysts and Cystic Neoplasms.  Many renal cystic lesions do not fulfill the criteria for simple cysts listed previously. Such lesions vary from minimally complicated simple cysts that usually do not require surgery to cystic renal neoplasms that generally need resection. Complicated cysts often result from bleeding in a simple cyst. Bosniak has suggested categorizing cystic renal masses according to CT criteria

B

C FIG 54-21  Simple renal cyst in renal transplant shown by MRI. A, T1-weighted (TR = 500 msec, TE = 15 msec) spin echo image shows a lesion (arrow) with lower signal intensity than renal parenchyma. B, On T2-weighted (TR = 2200 msec, TE = 90 msec) spin echo image, the cyst is hyperintense, indicating its fluid content. The cyst is well defined and has a smooth wall (arrow). C, Postcontrast gradient echo (TR = 168 msec, TE = 90 msec, flip angle = 75 degrees) image obtained after IV injection of Gd-DTPA shows that the lesion (arrow) does not enhance.

1792

PART II  CT and MR Imaging of the Whole Body in an attempt to separate lesions that require surgery from those that do not.31 Application of the Bosniak classification is most useful in evaluating and managing cystic renal lesions12; it takes into consideration their morphology and enhancement characteristics. This classification system has been updated recently, with introduction of a new category IIF (for follow-up) and revision of inclusion criteria for category III lesions.129,130 The revised Bosniak classification is described in Table 54-2. Category I lesions.  In the Bosniak classification, category I lesions are classic simple cysts as described previously (see Figs. 54-19 to 54-21). They require no further evaluation or management. Category II lesions.  Category II lesions are minimally complicated cysts that also usually do not require surgery. Occasional thin (<1 mm) smooth septa or small smooth plaques of fine linear calcification in the cyst wall or septa occur in these lesions (Fig. 54-22). Also included in category II are high-density cysts that are smaller than 3 cm (Fig. 54-23). These lesions have attenuation values ranging from 40 to 100 HU on nonenhanced CT scans and are usually caused

TABLE 54-2  Bosniak Renal Cyst

Classification System Category

Description

I

A benign simple cyst with a hairline-thin wall that does not contain septa, calcifications, or solid components. It measures water density and does not enhance. A benign cyst that may contain a few hairline-thin septa in which “perceived”* enhancement may be present. Fine calcification or a short segment of slightly thickened calcification may be present in the wall or septa. Uniformly high-attenuation lesions < 3 cm (so-called high-density cysts) that are well marginated and do not enhance are included in this group. Cysts in this category do not require further evaluation. Cysts that may contain multiple hairline-thin septa or minimal smooth thickening of their wall or septa. Perceived enhancement of their septa or wall may be present. Their wall or septa may contain calcification that may be thick and nodular, but no measurable contrast enhancement is present. These lesions are generally well marginated. Totally intrarenal, nonenhancing, highattenuation renal lesions > 3 cm are also included in this category. These lesions require follow-up studies to prove benignity. “Indeterminate” cystic masses that have thickened irregular or smooth walls or septa in which measurable enhancement is present. These are surgical lesions, and although some will prove to be benign (e.g., hemorrhagic cysts, chronic infected cysts, and multilocular cystic nephroma), some will be malignant, such as cystic renal cell carcinoma and multilocular cystic renal cell carcinoma. These are clearly malignant cystic masses that can have all the criteria of category III but also contain enhancing soft tissue components adjacent to, but independent of, the wall or septa. These lesions include cystic carcinomas and require surgical removal.

II

IIF (“F” for follow-up)

III

IV

FIG 54-22  Minimally calcified benign cyst (arrow), category II lesion.

From Israel GM, Bosniak MA: An update of the Bosniak renal cyst classification system. Urology 66:484–488, 2005, with permission. *Not measurable enhancement.

A

Contrast-enhanced CT scan reveals fine linear calcification in the cyst wall and septa. Lesion shows a thin wall and no focal nodules or areas of enhancement. Fluid in the cyst measured 12 HU.

B

FIG 54-23  Hyperdense benign renal cyst, category II lesion. Nonenhanced CT scan (A) shows a high-density (92 HU) mass (arrow) located anteriorly in the kidney. On contrast-enhanced CT scan (B), the lesion (arrow) did not enhance; it is homogeneous and reveals a smooth margin where it projects beyond the renal outline.

CHAPTER 54  Kidney

1793

B

A

C FIG 54-24  Hemorrhagic cyst in upper pole of left kidney. A, Contrast-enhanced CT scan shows a welldefined mass (arrow) in the upper pole of the left kidney. The lesion had a precontrast and postcontrast attenuation value of 45 HU. B, Gradient echo (TR = 168 msec, flip angle = 75 degrees) MRI shows fluid-iron level in the cyst, with high-intensity methemoglobin sediment located posteriorly (arrow) and lower intensity cyst fluid located anteriorly. C, T2-weighted (TR = 2200 msec, TE = 90 msec) spin echo image reveals that anteriorly located cyst fluid is hyperintense, whereas dependent sediment (arrow) has a lower signal intensity.

by high protein content or previous hemorrhage.144 Most high-density renal cysts are benign and can simply be monitored with serial imaging, provided the following CT criteria are fulfilled: (1) the lesion must  be perfectly smooth, round, sharply marginated, and homogeneous; (2) the lesion should not enhance with IV contrast medium; (3) at least one fourth of the lesion’s circumference should extend outside the kidney so that the smoothness of some of the wall can be evaluated; and (4) the lesion should be less than 3 cm in diameter.34 Although most lesions meeting these criteria are benign hemorrhagic cysts,33 cystic RCC may rarely show similar characteristics on CT.117 Highattenuation renal cysts and RCCs may have a similar appearance on nonenhanced CT images. Jonisch and colleagues demonstrated that a homogeneous renal mass with an attenuation of 70 HU or higher on nonenhanced CT has a greater than 99.9% chance of representing a high-attenuation renal cyst rather than an RCC.135 MRI may also help in characterizing high-density renal cysts that do not fulfill all the preceding CT criteria or that show internal echoes on US. Hemorrhagic cysts appear on MRI as sharply marginated, smooth, round, homogeneous lesions. They have high signal intensity on both T1- and T2-weighted sequences,179 and they do not enhance after IV administration of Gd-DTPA (see Fig. 54-21). In addition,

hemorrhagic cysts often show fluid-iron levels on MRI, probably because of dependent settling of methemoglobin-containing sediment (Fig. 54-24).179 Subtraction imaging can help in identifying the presence of enhancement in these high-signal lesions of the kidneys. This helps in differentiation of hemorrhagic cysts from hemorrhagic neoplasms. Category IIF lesions.  This category (“F” for follow-up) was introduced as a modification of the original Bosniak classification.131 It includes lesions that lie intermediately between categories II and III. They are slightly more complex than category II but do not fulfill the criteria for category III. These lesions are well defined and have multiple thin septa with minimal thickening of wall and septa (Fig. 54-25). They may also have thick nodular calcifications (Fig. 54-26). Intrarenal high-density lesions larger than 3 cm are also included in this category. These lesions are considered indeterminate and may advance to category III with time and therefore need follow-up at 6 months. Category III lesions.  Lesions in category III are indeterminate complicated cystic lesions. The presence of nodular and thick calcifications in walls or septa, which was considered a criterion for inclusion in category III according to the original Bosniak classification, is no longer a criterion. The modified version of the Bosniak system stresses

1794

PART II  CT and MR Imaging of the Whole Body studies.125 They may in rare cases cause hypertension, hematuria, or hydronephrosis or may become secondarily infected. Renal sinus cysts display the same CT features as simple renal parenchymal cysts.125 They have attenuation values in the water range (0-20 HU) and are difficult to distinguish from a dilated or extrarenal renal pelvis on nonenhanced CT scans. After IV administration of contrast medium the cysts remain of water density and cause displacement of the renal pelvis and calyces (see Fig. 54-30). Differentiation from hydronephrosis is thus readily made on contrast-enhanced CT.125 On CT the characteristic feature of a renal sinus cyst is a surrounding halo of renal sinus fat, indicating its extrarenal origin (see Fig. 54-29).57 CT easily distinguishes between a renal sinus cyst and renal sinus lipomatosis, which causes a similar deformity of the collecting system on excretory urography.125 In renal sinus lipomatosis, the attenuation value of the tissue is in the fat range. Solid masses in the renal sinus, such as lymphoma and invasive transitional cell carcinoma, are readily differentiated from renal sinus cysts because of their soft tissue attenuation values.

the presence of abnormal enhancement of the walls or septa for inclusion in category III (Fig. 54-27). Some category III lesions are benign (e.g., multilocular cystic nephromas); others are cystic RCCs. All such lesions should undergo surgical exploration unless the procedure is contraindicated because of the patient’s advanced age or poor general medical condition.31 Category IV lesions.  Category IV lesions are clearly malignant lesions with large cystic components; they should be resected.31 They may show marginal irregularity or solid vascular elements (Fig. 54-28).

Renal Sinus Cysts.  Renal sinus (parapelvic) cysts are benign extraparenchymal cysts located in the renal sinus.125 They are not true renal cysts but are probably lymphatic in origin.125 They may be unilocular or multilocular and are often bilateral (Figs. 54-29 and 54-30).125 They do not communicate with the renal collecting system. Most renal sinus cysts are asymptomatic and discovered incidentally on imaging

Renal Cystic Disease Renal cystic disease includes a wide variety of disorders. Although imaging studies help greatly in the evaluation of renal cystic disease, they are seldom diagnostic because many types of cystic kidneys show

FIG 54-27  Multilocular cystic renal cell carcinoma, category III lesion.

FIG 54-25  Category IIF lesion. A 4.2-cm cystic lesion at the lower pole

A contrast-enhanced CT scan reveals a well-defined, mainly fluidcontaining mass in the right kidney. Enhancing septa are seen throughout the lesion. A right nephrectomy was performed.

of the left kidney with enhancing (perceived enhancement, not measurable) minimally thick wall (arrow) and multiple septa (arrowhead).

A

B

FIG 54-26  Hemorrhagic benign renal cyst, category IIF lesion. Lesion (arrow) shows thick irregular mural calcification, has an attenuation value of 12 HU on a nonenhanced CT scan (A), and shows no enhancement after contrast administration (B). This lesion requires a follow-up.

CHAPTER 54  Kidney

1795

similar imaging findings. Therefore interpretation of imaging findings should always include knowledge of the patient’s age, family history, symptoms, clinical findings, and renal functional status.

Multicystic Dysplastic Kidney.  Multicystic dysplastic kidney

FIG 54-28  Cystic renal cell carcinoma, category IV lesion. Contrastenhanced CT scan shows a cystic mass (arrow) with irregular and enhancing solid elements located anteriorly.

FIG 54-29  Renal sinus (parapelvic) cyst. A contrast-enhanced CT scan reveals that the cyst (white arrow) is surrounded by a halo of renal sinus fat, indicating its extrarenal origin. The nondilated collecting system (curved arrow) and ureter (long arrow) are displaced by the cyst.

(MDK) is the most common form of cystic disease in infants, usually presenting as an asymptomatic flank mass. There is a male predominance (2 : 1 ratio), but female infants are twice as likely to have bilateral MDK disease and associated nonrenal abnormalities. The disorder is associated with intrauterine ureteral obstruction or atresia.277 The renal artery may also be hypoplastic or atretic.238 Contralateral renal anomalies, detected in 41% of patients with MDK, include obstruction of the ureteropelvic junction, renal agenesis, renal hypoplasia, vesicoureteral reflux, and bilateral MDK. Bilateral MDK is seen in 19% and contralateral agenesis is seen in 11% of patients with MDK.156 MDK can also be seen in patients with Meckel-Gruber syndrome, an autosomal recessive disorder characterized by the triad of occipital encephalocele, polydactyly, and cystic dysplastic kidneys. It is a rare syndrome that is incompatible with life.5 Other associated anomalies include cleft palate, ductal plate malformation of the liver, cardiac abnormalities, and incomplete development of external or internal genitalia. Two types of MDK may be shown by imaging, pelvoinfundibular atresia and the hydronephrotic type.310 In the classic type, pelvoin­ fundibular atresia, no discernible renal pelvis is demonstrated on imaging.276 The kidney may be small, normal in size, or enlarged and contains multiple variably-sized noncommunicating renal cysts (Fig. 54-31).238,276 No perfusion of the affected kidney occurs on renal scintigraphy, and contrast-enhanced CT shows no evidence of contrast excretion by the affected kidney.310 The hydronephrotic form of MDK is characterized by dilatation of the renal pelvis and calyces. Differentiation of this type of MDK from simple hydronephrosis is based on the imaging demonstration of parenchymal cysts that do not communicate with the collecting system.238,276 Currently most cases of unilateral MDK are managed nonsurgically and are followed with serial US.227,238 The affected kidney may remain unchanged, but it frequently undergoes spontaneous regression.238,337 MDK may occasionally be detected first in adults, in whom imaging studies show a small kidney with calcified cysts (Fig. 54-32) or an enlarged dysfunctional kidney.313 T2-weighted pulse sequences can be used to diagnose MDK, especially in utero.128

Polycystic Kidney Disease.  Hereditary polycystic kidney disease may be transmitted as an autosomal recessive or autosomal dominant trait. Autosomal recessive polycystic disease.  Autosomal recessive polycystic kidney disease is characterized by ectasia of the renal

C

C

FIG 54-30  Bilateral renal sinus cysts. The cysts cause displacement

FIG 54-31  Multicystic dysplastic kidney in a neonate. The right kidney

and attenuation of the opacified collecting systems. A simple renal cyst is also present anteriorly in the left kidney (arrow).

was replaced by cysts (C), one of which is larger than the other. The right kidney did not excrete contrast medium. The left kidney is normal.

1796

PART II  CT and MR Imaging of the Whole Body

S

FIG 54-32  Multicystic dysplasia of the left kidney in a 27-year-old man

a

with microscopic hematuria. The left kidney is small and composed of cysts with mural calcification (arrow). There is compensatory hypertrophy of the right kidney.

A

S

FIG 54-33  Autosomal recessive polycystic kidney disease in a 30-month-old child with normal renal function. A contrast-enhanced CT scan shows dilated collecting ducts in the renal medulla (striated nephrogram) bilaterally and relative preservation of the renal cortex. Occasional macroscopic cysts (arrowheads) are present in the left kidney.

collecting ducts, interstitial fibrosis, and variable degrees of portal hepatic fibrosis (congenital hepatic fibrosis), which often causes portal hypertension192; 10% to 90% of collecting ducts are involved. Severity of the disease depends on, and is inversely related to, the age at presentation. Patients with milder renal involvement may present at any age from infancy to early adulthood. The severity of renal involvement and extent of liver involvement in these patients are inversely related. On noncontrast CT examination, the kidneys are smooth, enlarged, and low in attenuation, likely a reflection of the large fluid volume in the dilated ducts. Administration of contrast medium reveals a striated nephrogram, representing accumulation of contrast in the dilated tubules. Tubular ectasia is confined mainly to the renal medulla, with only occasional macrocysts. The renal cortex is less severely affected (Fig. 54-33).192 CT may also show evidence of congenital hepatic fibrosis and portal hypertension.61 The liver contour is often irregular, and dilated intrahepatic bile ducts, ascites, splenomegaly, and enlarged portosystemic collateral veins may be seen (Fig. 54-34). Liver cysts are not usually evident. However, some patients have large liver cysts of biliary origin because of the presence of Caroli’s disease.67 Increased signal intensity of renal parenchyma is seen on T2-weighted MRI in autosomal recessive polycystic kidney disease.152,153,219 Autosomal dominant polycystic kidney disease.  Autosomal dominant polycystic kidney disease (ADPKD) is one of the most commonly inherited diseases in the United States. It affects nearly 500,000 Americans and accounts for 5% to 10% of patients with end-stage

B FIG 54-34  Autosomal recessive polycystic kidney disease in a 14-yearold girl with normal renal function. A, The liver contour is irregular as a result of congenital hepatic fibrosis. The spleen (S) is enlarged because of portal hypertension. Ascites (a) is present, and there is an enlarged left gastric vein (arrow) in the hepatogastric ligament. B, A more caudal scan shows mild nephromegaly and multiple medullary cysts with relative preservation of the renal cortex. (From Levine E: Computed tomography of renal masses. Crit Rev Diagn Imaging 24:91–200, 1985.)

renal disease (ESRD).18 ADPKD comprises at least three phenotypically indistinguishable but genetically distinct entities caused by mutations in three autosomal genes: PKD1 (chromosome 16p13.3) is present in about 85% of patients, PKD2 (chromosome 4q21-23) in 10%, and PKD3 (unknown chromosome) in a few families.287 The mean patient age at presentation with symptoms in ADPKD-1 is 44.8 years, versus 69.1 years for ADPKD-2.233 New mutations or nonfamilial development of ADPKD occurs in 1% to 2% of cases of ADPKD. Whites of European descent have the highest disease rates.43 ADPKD is associated with severely impaired renal function, but only 2% to 5% of the nephrons are directly affected by the cystic changes.44 Only 35% to 45% of patients with ADPKD experience ESRD.51 Cysts originate from renal tubules and collecting ducts. The ADPKD genotypes are characterized by bilateral kidney cysts, hypertension, hematuria, renal infection, stones, and renal insufficiency. ADPKD is a systemic disorder; cysts appear with decreasing order of frequency in the kidneys, liver, pancreas, brain, spleen, ovaries, and testes. Cardiac valvular disorders, abdominal and inguinal hernias, and aneurysms of cerebral and coronary arteries and the aorta are also associated with ADPKD. Colonic diverticula occur in about 80% of patients with the disease.90

CHAPTER 54  Kidney

A

1797

B FIG 54-35  Autosomal dominant polycystic kidney disease in a 39-year-old man with hypertension and mildly impaired renal function. Nonenhanced CT scan (A) shows bilateral nephromegaly. Contrast-enhanced CT scan (B) reveals large renal cysts with enhancing residual parenchyma between the cysts.

A

FIG 54-36  Autosomal dominant polycystic kidney disease in an asymptomatic 14-year-old boy whose father has the disease. A contrastenhanced CT scan shows multiple tiny bilateral renal cysts.

Although the disorder usually manifests in adults (Fig. 54-35), it may be diagnosed during infancy and childhood. Patients with fully developed ADPKD often show dramatic CT findings (see Fig. 54-35). CT is slightly more sensitive than US for detecting small renal cysts in the asymptomatic progeny of patients with ADPKD (Fig. 54-36).96 Diagnosis of ADPKD may be made based on US screening of at-risk individuals from unknown genotype using the following criteria: 1. Among individuals between 15 and 39 years of age, at least 3 unilateral or bilateral cysts 2. Among individuals 40 to 59 years of age, 2 cysts in each kidney 3. Among individuals 60 years or older, 4 cysts in each kidney These criteria were derived from a study of 948 individuals at risk for either PKD1 or PKD2 in which the performance of different US criteria was evaluated among at-risk individuals who subsequently underwent molecular genotyping.239 Flank pain and hematuria, the most common symptom and sign of ADPKD, may be due to cyst hemorrhage, calculi, or renal infection. Cyst hemorrhage is a common cause of pain in ADPKD and can be detected by CT in about 69% of patients.178 Hemorrhagic cysts have attenuation values of 40 to 100 HU on nonenhanced scans, do not enhance after IV administration of contrast medium, and are homogeneously hyperdense and well defined (Fig. 54-37). Pseudoen-

B FIG 54-37  Hemorrhagic renal cysts in autosomal dominant polycystic kidney disease. A, Nonenhanced CT scan shows bilateral subcapsular high-density cysts (arrows). B, After IV contrast enhancement, the highdensity cysts (arrows) are hypodense relative to enhanced residual renal parenchyma. Many fluid-density cysts are present. The highdensity cysts had an average attenuation value of 70 HU on both precontrast and postcontrast scans.

hancement of 10 HU can be seen in cysts smaller than 2 cm.54 They may rupture into the perinephric space, causing large hematomas (Fig. 54-38).179 Hemorrhagic cysts often develop mural calcification (Fig. 54-39). Renal calculi occur in 20% to 36% of patients with ADPKD and commonly cause flank pain.180 About 57% of calculi in ADPKD are composed predominantly of uric acid and are therefore radiolucent on conventional tomography.328 Such calculi are best evaluated by CT (Fig. 54-40).180 Urinary tract infections are more common in women with ADPKD and may cause cyst infection. Cyst infection is difficult to

1798

PART II  CT and MR Imaging of the Whole Body

FIG 54-38  Autosomal dominant polycystic kidney disease with perinephric hemorrhage. Nonenhanced CT scan reveals bilateral nephromegaly and multiple cysts. A posterior perinephric hematoma (short arrows) causes anterior displacement of the left kidney. A ruptured hemorrhagic cyst (open arrow) communicates (long arrow) with the hematoma. (From Levine E, Grantham JJ: Perinephric hemorrhage in autosomal dominant polycystic kidney disease: CT and MR findings.  J Comput Assist Tomogr 11:108–111, 1987.)

A

B FIG 54-40  Autosomal dominant polycystic kidney disease with a cal-

FIG 54-39  Autosomal dominant polycystic kidney disease with multiple curvilinear calcifications in cyst walls.

diagnose on CT; it is suggested by the finding of a cyst larger than surrounding cysts, with thickening and irregularity of its wall, an increase in the attenuation value of its contents, and localized thickening of the adjacent renal fascia (Fig. 54-41). Radionuclide scanning using gallium citrate 32– or indium III–labeled white blood cells may help confirm the diagnosis of cyst infection by showing increased tracer activity at the cyst periphery.281 Cysts of the liver are present in 70% to 75% of cases of ADPKD. The cysts are thought to arise from Meyenburg’s complexes. Hepatic cysts develop later than renal cysts, and women are more commonly affected than men.44 The CT appearance can vary from occasional small cysts (Fig. 54-42) to multiple large cysts that cause hepatomegaly (Fig. 54-43). Abnormal liver function test values suggest the development of complications such as common hepatic duct compression by cysts, cyst infection, and cholangiocarcinoma arising from the cyst lining (Fig. 54-44).177,317 The risk of renal cancer in patients with

culus in the lower pole of the left kidney. A, Nonenhanced CT scan shows a small rounded calcification (arrowhead). B, Contrast-enhanced CT scan reveals that the calcification is located in an opacified calyx (arrowhead). (From Levine E, Grantham JJ: Calcified renal stones and cyst calcifications in autosomal dominant polycystic kidney disease: Clinical and CT study in 85 patients. AJR Am J Roentgenol 159:77–81, 1992.)

ADPKD who are not undergoing dialysis is not increased; however, patients undergoing dialysis are at greater risk for development of renal malignancy relative to the general population.44 Living related kidney donors (with a family history of ADPKD) require screening for ADPKD, most commonly by US. US examination has a negative predictive value of 100% for donors older than 30 years, but of only 96% for donors 20 to 30 years of age. Heavily T2-weighted MRI in coronal and axial planes may be a more sensitive screening method for ADPKD in younger kidney donors (Fig. 54-45).360

Unilateral (Localized) Renal Cystic Disease.  In unilateral renal cystic disease, also known as localized renal cystic disease, most of one kidney is replaced by multiple cysts and the other kidney is normal. The affected patient usually has no family history of renal cysts, and

CHAPTER 54  Kidney

FIG 54-41  Acute pyelonephritis complicated by cyst infection in autosomal dominant polycystic kidney disease. A lower pole cyst (solid arrow) is larger than adjacent cysts and shows an irregular thickened wall and adjacent renal fascial thickening (open arrow). The patient demonstrated a response to antibiotic therapy, and a follow-up sonogram showed absence of the large cyst, suggesting rupture into the collecting system. (From Levine E, Grantham JJ: Calcified renal stones and cyst calcifications in autosomal dominant polycystic kidney disease: Clinical and CT study in 85 patients. AJR Am J Roentgenol 159:77–81, 1992.)

1799

FIG 54-43  Significant hepatomegaly caused by liver cysts in a 47-yearold woman with autosomal dominant polycystic kidney disease.

a

FIG 54-44  Autosomal dominant polycystic kidney disease complicated

FIG 54-42  Multiple small liver cysts in autosomal dominant polycystic kidney disease.

the condition is not related to ADPKD.58,159,183,297 Liver cysts and renal failure do not occur in this condition. Affected patients range in age from childhood to the sixth decade and may have hypertension, flank pain, or hematuria.183 The mechanism of hypertension in unilateral renal cystic disease is similar to that of ADPKD. Renin-containing cells are present in the walls of the cysts and in cells of the connective tissue surrounding the cysts. Renin can also be produced by the epithelial cells lining the cysts, and active renin is often present within the cyst fluid.302 On CT the cysts are seen to be separated by enhancing bands of normal renal parenchyma so that no distinct encapsulated renal mass is found (Fig. 54-46). The affected kidney may be enlarged and shows normal excretion of contrast medium. Although the cysts may predominate in one part of the affected kidney, careful assessment of the

by cholangiocarcinoma in a 69-year-old man complaining of abdominal pain and weight loss. Liver function values were notably abnormal. CT scan shows several cysts in the right lobe of the liver. A poorly defined low-density mass (arrowheads) involves the left lobe and part of the anterior segment of the right lobe. Ascites (a) is present. The neoplasm was not resectable, and the patient died of metastatic disease. (From Levine E, et al: Liver cysts in autosomal-dominant polycystic kidney disease: Clinical and computed tomographic study. AJR Am J Roentgenol 145:229–233, 1985.)

CT scan usually shows multiple smaller cysts elsewhere in the renal parenchyma.183 The absence of cysts in the other kidney on CT in adults usually excludes the diagnosis of ADPKD, but one must use caution in making a definitive diagnosis of unilateral renal cystic disease in children. In ADPKD, renal cysts develop slowly during childhood and may become apparent earlier in one kidney.247 Accordingly, in children, follow-up CT should be performed after an appropriate interval before a final diagnosis of unilateral renal cystic disease is made.

Acquired Cystic Kidney Disease.  Acquired cystic kidney disease (ACKD) is characterized by the development of multiple renal cysts in

1800

PART II  CT and MR Imaging of the Whole Body

patients with chronic renal failure but without a history of hereditary renal cystic disease.175 ACKD is seen in 90% of patients after 5 to 10 years of renal dialysis (hemodialysis or peritoneal dialysis); however, ACKD may be encountered in as many as 8% to 13% of patients with ESRD who are not undergoing dialysis.143,201 Nephrosclerosis, diabetic nephropathy, and chronic glomerulonephritis account for a large percentage of ACKD patients undergoing dialysis.168 The condition is

FIG 54-45  MRI of polycystic kidney. Coronal SSFSE images (TR =

67,500 msec, effective TE = 637 msec) demonstrate nephromegaly with multiple cysts that have variable signal intensities secondary to hemorrhage or high protein content.

A

usually asymptomatic but may become symptomatic when associated with a complication such as cyst or retroperitoneal hemorrhage, cyst infection, renal calculi, erythrocytosis, or renal neoplasm.45,175 The radiologic diagnosis of ACKD is best established with contrastenhanced CT, although US and MRI are useful for evaluating patients with ESRD who are not being treated with dialysis. The diagnosis is based on detection of at least three cysts in each kidney in a patient who has chronic renal failure not due to hereditary renal cystic disease.182 The affected kidneys are usually small (Fig. 54-47). However, ACKD is a progressive disorder, and nephromegaly may eventually develop (see Fig. 54-47).143 On MRI, renal cysts in ACKD are best shown on Gd-enhanced images (Fig. 54-48). Hemorrhagic cysts occur in about 50% of patients with ACKD and sometimes attain a large size.186 Hemorrhagic cysts manifest as welldefined masses with attenuation values ranging from 40 to 100 HU on nonenhanced CT scans (Fig. 54-49). Differentiation of these lesions from renal neoplasms is suggested by lack of contrast enhancement and a homogeneous appearance on contrast-enhanced CT scans (see Fig. 54-49).186 A nonhemorrhagic cyst that is more than 1 cm in diameter may show pseudoenhancement of less than 10 HU.13 Sonography may help in establishing the cystic nature of such a lesion by showing an anechoic center, posterior acoustic enhancement, and a smooth wall. On MRI, hemorrhagic cysts often have high signal intensity on both T1- and T2-weighted images and do not enhance after Gd administration (Fig. 54-50). Subcapsular and perinephric hematomas, which occur in as many as 13% of patients with ACKD, are usually due to rupture of hemorrhagic cysts (Fig. 54-51).186 The primary concern in patients with ACKD is the increased incidence (5%-19%) of RCC—about 12 to 18 times higher than that  in the general population—and the cancer may be asymptomatic 

B FIG 54-46  Unilateral renal cystic disease in a 15-year-old boy with intermittent gross hematuria starting at 5 years of age. Contrast-enhanced CT scans (A and B) show partial replacement of the left kidney by many cysts that are separated by enhancing bands of normal parenchyma. The right kidney is free of cysts, and a follow-up CT scan 7 years later showed no change.

A

B FIG 54-47  Acquired renal cystic disease in a 45-year-old man undergoing long-term hemodialysis. A, CT scan after 4 years of dialysis shows bilateral small cysts in shrunken kidneys. B, A follow-up CT scan obtained 4 years later shows enlargement of cysts and kidneys.

CHAPTER 54  Kidney

A

B FIG 54-48  Acquired renal cystic disease and left renal cell carcinoma in a 34-year-old man with azotemia not treated with dialysis. A, T1-weighted (TR = 500 msec, TE = 17 msec) spin echo MRI after Gd-DTPA administration shows several well-defined cysts (arrowheads) of low signal intensity in the left kidney. B, T2-weighted (TR = 2100 msec, TE = 90 msec) image without Gd administration reveals a heterogeneous 3.5-cm lower pole carcinoma (arrowheads).

A

B FIG 54-49  Acquired renal cystic disease with cyst hemorrhage in a 41-year-old man with gross hematuria. A, Nonenhanced CT scan shows a 3.8-cm high-density (44 HU) mass (arrow) in the upper pole of the left kidney. B, After contrast enhancement, the attenuation value of the mass (arrow) was 43 HU. Multiple bilateral fluid-density cysts are also present. The mass resolved completely on a follow-up CT scan obtained 3 months later, suggesting a hemorrhagic cyst that had ruptured into the collecting system. (From Levine E, et al: Natural history of acquired renal cystic disease in dialysis patients: A prospective longitudinal CT study. AJR Am J Roentgenol 156:501–506, 1991.)

A

B FIG 54-50  Hemorrhagic cysts in acquired renal cystic disease. A, Nonenhanced CT scan shows multiple bilateral fluid-density cysts with cyst wall calcifications. An area (arrow) located medially in the right kidney was considered suspect for a neoplasm. B, T1-weighted (TR = 500 msec, TE = 15 msec) spin echo MRI shows that most cysts have low signal intensity. However, two lesions (arrows) located medially in the right kidney have a high signal intensity, which persisted on T2-weighted images (not shown), findings consistent with cyst hemorrhage.

1801

1802

PART II  CT and MR Imaging of the Whole Body

(Fig. 54-52).143 RCC in patients undergoing dialysis occurs at an average age of 49 years, compared with 62 years in nonazotemic patients.248 The development of RCC in patients undergoing dialysis depends on the duration of dialysis (incidence rises after 3 years of dialysis), size of the kidneys, and male gender (seven times higher).45,175 Small renal neoplasms (<3 cm in diameter) are significantly more common than large carcinomas and occur in about 7% of patients undergoing dialysis (Fig. 54-53).175 Most such lesions remain unchanged with prolonged follow-up. However, some gradually enlarge and may become locally invasive or even metastasize (see Fig. 54-53).181 Because of their propensity to develop malignant renal neoplasms, the native kidneys of patients undergoing dialysis should be evaluated with CT if the patients experience flank pain or hematuria. Annual imaging of the native kidneys in asymptomatic patients undergoing dialysis should be restricted to those in good general medical condition and who have a good life expectancy.186

births. The disease, inherited as an autosomal dominant trait with high penetrance, is caused by a genetic defect in the short arm of chromosome 3.296 Manifestations of VHLD include retinal angiomas, central nervous system hemangioblastomas, pancreatic cysts, islet cell tumors, pheochromocytomas, endolymphatic sac tumors, cystadenomas of the epididymis, and broad ligament and renal cysts and tumors.44 The abdominal lesions are best detected on CT.46,47 VHLD that is not associated with pheochromocytoma is subclassified as VHLD1, and VHLD that is associated with pheochromocytoma as VHLD2. Multiple bilateral renal cysts and cystic and solid neoplasms are the most common manifestations of VHLD (Fig. 54-54). Most cysts remain stable with long-term CT follow-up, and transformation from typical simple cysts to solid lesions is rare.44,46 However, complex lesions with both cystic and solid elements may transform into solid lesions and therefore require careful imaging follow-up.46 RCC is reported to develop in 24% to 45% of patients with VHLD and is often multifocal and bilateral.197,198 VHLD-associated renal cysts are lined by hyperplastic epithelium that is commonly intermixed with foci of tumor.249 Once solid renal lesions develop, they grow at rates similar to those of sporadic RCCs (Fig. 54-55).46 Predominantly cystic lesions with solid components are characteristic of VHLD. The solid component usually represents RCC.249 The pancreas is commonly involved in VHLD, with pancreatic cysts occurring in about 30% of patients (Fig. 54-56). Solid pancreatic lesions, including nonfunctional islet cell tumor and ductal carcinoma, occur less commonly in VHLD. Pheochromocytomas, which occur in 10% of patients with VHLD, are often multiple and ectopic; approximately 50% to 80% are bilateral (Fig. 54-57).176,262 Screening in families affected by VHLD permits timely genetic counseling and early identification of potentially life-threatening lesions. Abdominal CT should first be performed late in the second decade in at-risk individuals (i.e., primary relatives of patients with the diagnosis). The frequency of imaging thereafter depends on the initial findings. Patients with suspect renal lesions may require annual CT evaluation.

Hereditary Syndromes

von Hippel-Lindau Disease.  It is estimated that approximately 7000 patients in the United States have von Hippel-Lindau disease (VHLD). The incidence of VHLD is approximately 1 in 40,000 live

Tuberous Sclerosis Complex.  Tuberous sclerosis complex (TSC) may be inherited as an autosomal dominant trait or may occur sporadically. Tuberous sclerosis is an autosomal dominant disorder often associated with a chromosome 9 (TSC1) and chromosome 16 (TSC2) abnormality, although up to 60% of cases occur spontaneously. The

FIG 54-51  Acquired renal cystic disease and a right perinephric hematoma (arrows) in a patient undergoing long-term dialysis. (From Levine E, et al: CT of acquired cystic kidney disease and renal tumors in longterm dialysis patients. AJR Am J Roentgenol 142:125–131, 1984.)

A

B FIG 54-52  Acquired renal cystic disease and metastatic renal cell carcinoma in a 46-year-old man treated with hemodialysis for 120 months. A, Contrast-enhanced CT scan shows multiple bilateral renal cysts and nephromegaly. The enlarged left renal hilar lymph node (arrow) is due to a metastasis. B, Caudal CT scan reveals two solid heterogeneous masses (arrows) with irregular contours and focal calcifications in the lower pole of the left kidney. The masses enhanced by 30 HU. (From Levine E, et al: Natural history of acquired renal cystic disease in dialysis patients: A prospective longitudinal CT study. AJR Am J Roentgenol 156:501– 506, 1991.)

CHAPTER 54  Kidney

A

1803

B FIG 54-53  Growth of renal cell carcinoma in a patient with acquired renal cystic disease who was undergoing hemodialysis. A, The first CT scan shows a small mass (arrow) in the upper pole of the left kidney. B, CT scan performed 18 months later during an episode of gross hematuria reveals that the left renal mass (arrows) has enlarged and become heterogeneous, suggesting renal cell carcinoma. On pathologic examination, the lesion had invaded the renal capsule. Linear strands related to the right kidney resulted from perinephric hemorrhage. (From Levine E, et al: Spontaneous subcapsular and perinephric hemorrhage in end-stage kidney disease: Clinical and CT findings. AJR Am J Roentgenol 148:755–758, 1987.)

A

B FIG 54-54  Multiple renal cysts in two different patients with von Hippel-Lindau disease. A, Contrastenhanced CT scan shows multiple bilateral renal cysts. B, T2-weighted (TR = 2200 msec, TE = 90 msec) spin echo MRI shows multiple cysts with high signal intensities involving both kidneys.

incidence of the disorder is between 1 in 10,000 and 100,000 live births, and it leads to multiple organ and skeletal abnormalities.44,84 The classic triad of epilepsy, mental retardation, and adenoma sebaceum defines the syndrome clinically.84 Renal lesions, seen in 50% of patients with TSC, include multiple cysts, angiomyolipomas (AMLs), tumors and perirenal cystic collections, or lymphangiomas.44,363 Other associated abnormalities are subependymal nodules, giant cell astrocytoma, peripheral tubers, retinal hamartoma, cardiac rhabdomyoma, lymphangioleiomyomatosis, shagreen patch, subungual fibroma, and bone cyst.44 Multiple renal AMLs are the sine qua non of TSC and are usually bilateral; they occur in about 15% of patients, mostly female.363 AMLs are benign renal tumors composed of fat, smooth muscle, and abnormal blood vessels. Abnormal vessels (aneurysmal dilations) predispose the lesions to hemorrhage. The rate of hemorrhage in TSCrelated AMLs is higher than that found in sporadic AMLs.44 CT scans usually demonstrate a large fatty mass intermixed with areas of tissue density that may represent nonfatty parts of the tumor or areas of hemorrhage (Fig. 54-58). The tumor commonly extends into the perinephric space and is often complicated by intratumoral or perinephric hemorrhage. CT is also useful in identifying perinephric extension and hemorrhage. The CT appearance of AMLs that are composed primarily of smooth muscle, or in which intratumoral hemorrhage has obscured the fatty part of the tumor, is indistinguishable from that of other solid renal tumors.288

Renal cystic disease in TSC may sometimes be the earliest and only clinical manifestation of the disorder in infancy and childhood.214 The kidneys are often significantly enlarged in such patients and are replaced by multiple cysts of varying size (see Fig. 54-58). The cysts involve the renal cortex and medulla, and the findings are similar to those of ADPKD.210 The cysts in TSC are not a diagnostic feature, occurring at an earlier age than in ADPKD. The combination of renal cysts and AMLs is strongly suggestive  of TSC.214 If no renal AMLs are found in a patient suspected of having TSC, cranial CT may help establish the diagnosis of TSC  by showing small, high-attenuation, periventricular subependymal nodules (Fig. 54-59).318 TSC is also associated with an increased incidence of renal cancer.42 TSC-associated tumors occur predominantly in female patients (81% female predominance in TSC vs. 70% male predominance for sporadic renal cancers), involve a younger age group (median age of 28 years vs. sixth and seventh decades for sporadic renal cancers), and are usually multiple and bilateral (in 43%).327,345 Clear cell renal carcinoma is the most common type of cancer found in association with TSC. Less common types of cancers include papillary and chromophobe carcinomas and oncocytomas.124,133,237

Renal Lymphangiectasia.  Renal lymphangiectasia is probably due to developmental obstruction of the larger lymphatics draining the

1804

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 54-55  Growth of a renal cell carcinoma in a patient with von Hippel-Lindau disease. The right kidney was previously resected because of renal cell carcinoma. The patient refused surgery for the left renal lesion until after the third CT scan. A, There is a 1.8-cm solid mass (arrow) in the lower pole of the left kidney. B, Follow-up CT scan obtained 2 years after the scan in A shows enlargement of the lesion (arrow) to 2.2 cm and mild heterogeneity. C, CT scan obtained 3 years after the scan in B shows that the lesion has now replaced most of the left kidney. A left nephrectomy was performed, and hemodialysis was started.

FIG 54-56  Multiple pancreatic cysts with mild pancreatic enlargement in von Hippel-Lindau disease.

FIG 54-57  Right adrenal pheochromocytoma (arrow) in von HippelLindau disease.

CHAPTER 54  Kidney kidney through the renal pedicle.28,189,242 The condition is usually bilateral and is characterized on imaging studies by large lymphatic cysts in the renal sinuses, perinephric tissues, and central retroperitoneum, as well as diffuse intrarenal lymphangiectasia (Fig. 54-60).216,256 Renal lymphangiectasia is usually discovered clinically because of abdominal masses caused by nephromegaly. The disorder may be complicated during pregnancy by large perinephric lymph collections and ascites

FIG 54-58  Tuberous sclerosis with renal cystic disease in a 47-year-old man. A, Contrast-enhanced CT scan shows enlarged kidneys with multiple fluid-density cysts. Also seen are bilateral fat-density lesions (arrows) consistent with angiomyolipomas and strongly suggestive of tuberous sclerosis. Liver also shows a fat-density lesion (arrowhead) consistent with a lipoma or angiomyolipoma.

1805

(Fig. 54-61).207 These findings may be secondary to lymphatic rupture resulting from increased renal lymph flow during pregnancy in the presence of lymphatic obstruction.207

Birt-Hogg-Dubé Syndrome.  Birt-Hogg-Dubé syndrome is primarily a dermatologic disorder characterized by tumors involving the  hair follicles, lung cysts, and renal tumors. This autosomal dominant hereditary syndrome initially included only the skin tumors, such  as fibrofolliculomas (tumors in hair follicles), trichodiscomas (early fibrofolliculomas), and acrochordons (skin tags consisting of collagen fibers and blood vessels). Later the association of renal cancers with these skin disorders was recognized and included in the syndrome. Renal cancers associated with Birt-Hogg-Dubé syndrome include most commonly chromophobe carcinomas (34%), oncocytomas (5%), hybrid chromophobe-oncocytomas (50%), clear cell carcinomas (9%), and papillary renal cancers (2%).222 Association with lung cysts and occasionally pneumothorax has also been confirmed.326 Diagnosis is based on the combined clinical, imaging, and histopathologic findings. On imaging, findings include lung cysts, commonly involving the lower lobes, with associated spontaneous pneumothorax in about 25% of patients.48 About 15% to 30% of patients with cutaneous findings also have renal cancers, commonly the chromophobe cancers. Fibrofolliculomas, which appear as domeshaped papules predominantly on the upper half of the body, are a characteristic finding in the Birt-Hogg-Dubé syndrome and help differentiate it from other hereditary syndromes such as TSC, which is also characterized by renal cancers and lung cysts, along with intracerebral lesions.

Lynch Type 2 Syndrome.  Lynch type 2 syndrome is a type of hereditary nonpolyposis colon cancer that along with colon cancer includes increased risk of other cancers such as carcinomas of the endometrium (second most common tumor), renal pelvis, urothelium, stomach, small bowel, pancreas, biliary system, and ovaries.1 Lynch type 1 syndrome does not carry a risk for noncolonic cancers.

A

B FIG 54-59  Tuberous sclerosis with renal cystic disease in a 14-year-old girl. A, Nonenhanced CT scan displays kidneys containing multiple cysts. No fat-containing angiomyolipomas were detected. B, Cranial CT scan reveals paraventricular calcifications (arrows), indicating the diagnosis of tuberous sclerosis and permitting differentiation from autosomal dominant polycystic kidney disease.

1806

PART II  CT and MR Imaging of the Whole Body On imaging, these renal tumors are usually hypovascular when small and become hypervascular as they enlarge.48

Familial Renal Oncocytoma.  Familial renal oncocytoma is another autosomal dominant condition characterized by the occurrence of multiple or bilateral oncocytomas (Fig. 54-63). Because of their similarity to RCCs on imaging, they are often treated in the same manner as RCCs.66

Renal Medullary Carcinoma.  Renal medullary carcinoma is a rare

FIG 54-60  Renal lymphangiectasia in a 10-month-old boy with abdominal masses on routine physical examination and normal renal function. Contrast-enhanced CT scan reveals nephromegaly and many renal sinus, perinephric, and central retroperitoneal lymphatic cysts. (From Blumhagen JD, et al: Sonographic evaluation of abdominal lymphangiomas in children. J Ultrasound Med 6:487–495, 1987.)

FIG 54-61  Renal lymphangiectasia in a 25-year-old woman who experienced severe abdominal distention during pregnancy. A nonenhanced CT scan shows large bilateral perinephric lymph collections and renal sinus lymphatic cysts. (From Meredith WT, et al: Exacerbation of familial renal lymphangiomatosis during pregnancy. AJR Am J Roentgenol 151:965–966, 1988.)

Hereditary Papillary Renal Cancer Syndrome.  This syndrome predisposes those affected to increased risk of type 1 papillary RCC.327 It was first recognized by Zbar and colleagues in 1994.361 These cancers are slow growing and therefore have a late clinical presentation. Detecting these tumors on imaging can be very challenging because of their hypovascular nature, which may result in their being mistaken for cysts if the subtle attenuation differences on postcontrast images are not identified.49 On US these tumors can be easily missed because of their isoechoic appearance.

Hereditary Leiomyoma Renal Cell Carcinoma.  Hereditary leiomyoma RCC syndrome is an autosomal dominant syndrome characterized by increased risk of uterine and skin leiomyomas and associated type 2 papillary RCC (in ≈ 20% of families).111,167,325 The papillary tumors are usually very aggressive (Fig. 54-62) and are solitary, unlike most of the hereditary tumors syndromes, which have multiple tumors.48

tumor of the kidney first described by Davis and colleagues in 1995.69 This tumor occurs exclusively in young black patients with sickle cell trait (hemoglobin [Hb] SA and HbSC) but not sickle cell anemia.40,69 Renal medullary carcinoma represents the seventh sickle cell nephropathy,69 the other six being hematuria, papillary necrosis, nephrotic syndrome, renal infarction, isosthenuria (inability to concentrate urine), and pyelonephritis. Affected patients range from 11 to 39 years of age.69 Flank pain and hematuria are the most common presenting complaints.344 The exact histologic origin of this tumor is controversial. It has been suggested that renal medullary carcinoma represents a particularly aggressive form of collecting duct carcinoma.304 Metastatic involvement of regional lymph nodes, liver, and lungs at presentation is common.154 Renal medullary carcinoma is thought to arise at the renal pelvis– mucosa interface. The tumor quickly grows to fill the renal pelvis and invade vascular and lymphatic structures. Intraparenchymal satellite nodules are frequently present.193 The radiologic appearance of renal medullary carcinoma is that of a prototypical infiltrative lesion. An ill-defined mass centered in the renal medulla with extension into the renal sinus and cortex is characteristic; caliectasis may be seen, and the reniform contour of the kidney is maintained (Fig. 54-64).244 The tumors are heterogeneous on US and contrast-enhanced CT, reflecting the characteristic tumor necrosis.64 The prognosis is extremely poor, with advanced disease at the time of diagnosis. The tumor responds poorly to chemotherapy or radiation therapy, with a mean survival of 15 weeks from diagnosis. The constellation of renal medullary mass, black race, sickle cell trait, and HbSC disease suggests the diagnosis.193

Malignant Renal Parenchymal Neoplasms Malignant renal parenchymal neoplasms may be either primary or secondary. RCC is the most common primary renal cancer, accounting for about 86% of all primary malignant renal parenchymal neoplasms. Of the remainder, 12% are Wilms’ tumors and 2% are renal sarcomas.173 Secondary neoplasms of the renal parenchyma include malignant lymphoma and metastases. The most common renal neoplasms overall are metastases.244

Primary Neoplasms Renal cell carcinoma.  RCC is a common urologic malignant lesion in adults and accounts for 2% of all cancers. It may occur at any age, although most patients are older than 40 years, and persons in the seventh and eighth decades of life are most commonly affected.251 Only 0.3% to 1.3% of all cases of RCC present in children older than 5 years.100 Known risk factors are cigarette smoking, exposure to petroleum products, obesity, ACKD, VHLD, TSC, and hereditary papillary renal cancer.215,251 Occurrence of sporadic clear cell RCC is associated with the existence of a tumor suppressor gene on chromosome 3p.9 Current imaging techniques have had a notable effect on tumor detection. Kidney cancers are being diagnosed at an earlier stage and with greater precision.200 Small early-stage carcinomas are now commonly identified during abdominal CT or US performed for nonrenal

CHAPTER 54  Kidney

1807

B A

C FIG 54-62  Papillary renal cell carcinoma type 2. Nonenhanced (A), nephrographic-phase (B), and excretoryphase (C) CT scans of the kidneys demonstrate a cystic lesion of the left kidney (arrow) that shows no appreciable enhancement. Incidentally seen is atrophic right kidney with dilated extrarenal pelvis. It was surgically resected and pathologically confirmed to be a papillary renal cell carcinoma type 2.

complaints (Fig. 54-65).184 Indeed, almost as many new RCCs are currently being detected incidentally as are being found in patients investigated because of hematuria or flank pain.

Imaging findings CT.  The scanning protocol for RCC consists of a combination of nonenhanced and contrast-enhanced CT in the corticomedullary and nephrographic phases. The corticomedullary phase (25-70 seconds from the start of injection of contrast material) is best for evaluating venous extension of RCC. The nephrographic phase (80-180 seconds after the start of injection) is best for evaluating the renal mass.290 Single-detector and multidetector spiral CT allow rapid image acquisition through the entire kidney in a single breath hold during various phases of contrast enhancement after administration of a single bolus of IV contrast material.290 CT is the primary modality for detection, diagnosis, and staging of RCC. The precontrast CT appearance of RCC varies considerably depending on the gross pathologic features. Neoplasms may be hypodense, isodense (see Fig. 54-65), or hyperdense (Fig. 54-66) compared with normal renal parenchyma on nonenhanced CT scans.359 Most RCCs are solid lesions with attenuation values of 20 HU or greater on nonenhanced CT.294 Tumor calcification occurs in as many as 31% of cases and may take the form of amorphous internal calcification (Fig. 54-67; see Fig. 54-66) or curvilinear calcification (Fig. 54-68), which may be peripheral or central.359

After administration of IV contrast medium, most RCCs enhance but usually to a lesser extent than normal renal parenchyma (see Figs. 54-65 and 54-66). This feature is best seen in the nephrographic phase,26,314 the most important phase for detecting renal masses.290 Enhancement is often heterogeneous because of tumor hemorrhage or necrosis (see Figs. 54-66 and 54-67).173 The mass often shows an indistinct interface with the surrounding parenchyma and frequently has a lobulated or irregular outer margin. Small RCCs, however, often have distinct smooth margins.287,359 MRI.  MRI assumes a primary role in tumor detection and staging in patients in whom contrast-enhanced CT scanning is contraindicated because of either previous major reactions to contrast material or renal failure.77,267 MRI may be used as a supplement to CT in patients who have renal carcinoma and venous tumor extension and in whom the status of the vena cava is not clearly established.77,83,106 The usual appearance of RCC on MRI is that of a lesion that  is hypointense to isointense to renal parenchyma on T1-weighted images, heterogeneously hyperintense on T2-weighted images, and demonstrates less enhancement than renal parenchyma after administration of Gd contrast. The appearance can vary, however, depending on tumor histology and the presence or absence of intratumoral  lipid, hemorrhage, and necrosis.251,291 RCC may be hyperintense, hypointense, or isointense to normal renal parenchyma on both T1and T2-weighted images. The use of IV Gd enhancement with fat

1808

PART II  CT and MR Imaging of the Whole Body

suppression on T1-weighted imaging makes tumors much more conspicuous.285,286 Solid RCCs often show contrast enhancement and may reveal irregular ill-defined margins and heterogeneity (Fig. 54-69). Tumor calcification is difficult to appreciate on MRI. Staging of renal cell carcinoma.  The prognosis of patients with RCC depends mainly on the histologic tumor grade and the anatomic extent or stage of the disease when first seen. RCC is now most commonly staged according to the tumor, node, and metastases (TNM) system (Table 54-3), which has replaced the almost 40-year-old Robson system.264 The TNM system has been revised several times to improve its accuracy, with the latest revision in 2002. In the Robson system, stage I RCC consists of tumor confined in the renal capsule; stage II of perinephric extension, including the ipsilateral adrenal gland; stage IIIA of venous invasion (renal vein, with potential extension into the IVC); stage IIIB of regional lymph node metastases; and stage IIIC of both venous tumor extension and regional lymph node metastases. Stage IVA neoplasms extend through Gerota’s fascia to

involve adjacent organs, and stage IVB neoplasms have distant metastases. Radiologic staging of RCC has an important influence on management. Small stage I neoplasms (<3 cm) are managed by partial nephrectomy in many institutions (Fig. 54-70).251 A radical nephrectomy is usually required for stage II tumors. The operation includes removal of the kidney, adrenal gland, perirenal fat, and Gerota’s fascia with or without a regional lymph node dissection.241 Lymphadenectomy is commonly used, but its effectiveness has not been definitively

A

B FIG 54-64  Medullary carcinoma. A, Contrast-enhanced CT scan of the FIG 54-63  Bilateral oncocytomas. Contrast-enhanced CT scan of the kidneys demonstrates bilateral solid enhancing masses (arrows) with a central scar.

A

kidneys shows a right renal mass (arrow) that enhances less than the normal renal parenchyma. B, Contrast-enhanced MRI shows the enhancing right renal mass (arrow) with slightly lower signal intensity than the normal renal parenchyma.

B

FIG 54-65  Renal cell carcinoma. A, The neoplasm (arrows) has a similar density to normal renal parenchyma on a nonenhanced CT scan. B, Contrast-enhanced CT scan shows a homogeneous 2.3-cm mass (white arrows) with a slightly irregular outline. The attenuation value increased from 37 to 87 HU. A small renal cyst (black arrow) is seen anteriorly.

CHAPTER 54  Kidney

A

1809

B

C FIG 54-66  Renal cell carcinoma. A, Small (2.9-cm) renal mass (arrow) has a higher density than normal renal parenchyma on a nonenhanced CT scan. B, A precontrast CT scan through the upper part of the lesion shows a small focus of amorphous calcification (arrow). C, Nephrographic-phase CT scan shows that the mass (arrow) enhances less than normal renal parenchyma and reveals low-density areas due to tumor necrosis. The mass has an indistinct interface with adjacent renal parenchyma.

FIG 54-68  Renal cell carcinoma. Left renal carcinoma (arrows) shows peripheral curvilinear calcification. The right kidney is small because of renal artery stenosis.

FIG 54-67  Renal cell carcinoma. Small foci of amorphous calcification (arrowheads) are present in a left renal carcinoma. The mass has an indistinct interface with adjacent renal parenchyma and is heterogeneous.

1810

PART II  CT and MR Imaging of the Whole Body

FIG 54-69  Gd-enhanced MRI of renal cell carcinoma. Coronal (T1-weighted and fat-suppressed sequence with Gd) image in the nephrographic phase demonstrates a mass (arrowhead) that enhances less than the adjoining renal parenchyma in the middle pole of the right kidney.

proved. Stage III tumors are also managed with radical nephrectomy, with extension of surgery to remove the entire renal vein and caval thrombus and a portion of the vena cava as necessary.118 Preoperative detection of venous tumor thrombus is important because the surgical approach changes according to the cephalad extension of the tumor into the IVC.229 Treatment of stage IV RCC differs from center to center. Almost all cases are incurable. Tumor embolization, external-beam irradiation, and nephrectomy can palliate symptoms caused by the primary tumor or from related ectopic hormone production. The main chemotherapeutic agents in use are interleukin-2 and alfa-interferon.228,347 CT is the single most effective modality for staging RCC, with an overall accuracy of about 90%.17,74 CT and MRI achieve approximately equal accuracy in the local staging of RCC.17 Perinephric invasion is suggested by a perinephric soft tissue mass at least 1 cm in diameter (Fig. 54-71), but lesser degrees of perinephric involvement are difficult to diagnose with CT (Fig. 54-72).134,160 Perinephric soft tissue stranding is an unreliable indicator of perinephric tumor invasion. Venous tumor invasion can be definitely diagnosed on CT only if there is identifiable thrombus in the renal vein (Figs. 54-73 and 54-74) or IVC (see Fig. 54-74) with or without venous enlargement.362 Ipsilateral

TABLE 54-3  Staging Systems for Renal Cell Carcinoma and Their CT Criteria Tumor Position

Robson Stage

TNM Class

CT Findings

CT Pitfalls

Confined within renal capsule

I

—

—

  Small (<7 cm diameter)   Large (≥7 cm diameter) Spread to perinephric fat

— — II

T1 T2 T3a

Soft tissue mass enhances less than normal renal parenchyma: central necrosis in large renal cell carcinoma — — Perinephric stranding; perinephric collateral vessels Soft tissue mass in perinephric space

Venous thrombus

IIIA

—

Filling defect within a distended vein; direct continuity of thrombus with primary mass; IV contrast enhancement indicates tumor thrombus; collateral veins

  Renal vein only   IVC infradiaphragmatic   IVC supradiaphragmatic Regional lymph node metastases

— — — IIIB

T3b T3c T4b N1-N3

— — — Lymph nodes ≥ 1 cm in diameter

Direct invasion of adjacent organs

IVA

T4a

Obliteration of normal soft tissue planes between tumor and adjacent organs

Distant metastases

IVB

M1a-d

—

N4

—

Metastases enhance with IV contrast material; hepatic metastases best seen in arterial phase —

— — Not reliable or specific; found in 50% of T1 and T2 tumors; false negative if spread is microscopic Specific, not sensitive in 45%-50% of cases False negative: right renal vein and IVC obscured False positive: venous enlargement due to increased flow; streaming of unopacified blood in IVC (perform delayed scanning) — — — False-negative rate: 4% False positive: enlarged inflammatory nodes False positive: partial volume averaging; tumor adherent but not directly invading Hypervascular metastases may be obscured in portal venous phase

From Sheth S, et al: Current concepts in the diagnosis and management of renal cell carcinoma: Role of multidetector CT and three-dimensional CT. Radiographics 21(Spec No):S237–S254, 2001, with permission. IV, intravenous; IVC, inferior vena cava.

CHAPTER 54  Kidney

A

1811

B FIG 54-70  Renal cell carcinoma. A, Corticomedullary phase demonstrates a small exophytic mass in the middle pole of the left kidney, isodense to the renal cortex (arrow). B, The nephrographic phase demonstrates that the mass (arrowhead) enhances less than the renal parenchyma. The patient underwent partial nephrectomy.

FIG 54-71  Renal cell carcinoma. A right-sided tumor has invaded the perinephric fat posteriorly (arrow) and medially (black arrowhead). The medial perinephric mass indents the inferior vena cava. The renal fascia (white arrowhead) is thickened.

renal vein enlargement on CT without identifiable tumor thrombus is not a reliable sign of venous tumor extension (Fig. 54-75). Although such enlargement may reflect the presence of tumor thrombus, it more frequently results from increased blood flow due to a hypervascular tumor (see Fig. 54-75).362 Errors in diagnosis of venous tumor extension on CT often occur when large right-sided tumors cause notable distortion of the ipsilateral renal vein and IVC.134,185 Tumor thrombus is distinguished from a bland thrombus by the former’s heterogeneous enhancement on the corticomedullary phase and continuity with the primary tumor (Fig. 54-76).290,356 MRI is superior in detecting venous tumor involvement and its extent and is a suitable alternative when CT findings are indeterminate (Fig. 54-77; see Fig. 54-75). Tumor thrombus in the renal vein and vena cava may be suspected on T1- and T2-weighted conventional imaging if the signal void of flowing blood is replaced by relatively high signal intensity, which is due to tumor thrombus (Fig. 54-78). Tumor thrombus has signal intensity similar to that of the primary neoplasm (see

FIG 54-72  Renal cell carcinoma with early perinephric extension. Nonenhanced fat-suppressed T1-weighted MRI (TR = 600 msec, TE = 15 msec) shows a large neoplasm (straight arrows) in the lower pole of the right kidney. There are early extensions into the perinephric fat (open arrow) and associated renal fascial thickening (curved arrows).

FIG 54-73  Left renal cell carcinoma with venous extension. The proximal part of the left renal vein is enlarged (arrowheads) and occupied by low-density tumor.

1812

PART II  CT and MR Imaging of the Whole Body

A

B FIG 54-74  Renal cell carcinoma with venous extension and regional lymph node metastases. A, Contrast-enhanced CT scan shows a right renal mass (black arrow). The right renal vein (white arrow) is enlarged and occupied by tumor, which has also extended into the inferior vena cava. There is a metastasis (arrowhead) in the right lobe of the liver. B, Multiple enlarged central retroperitoneal lymph nodes (arrows) were due to metastatic disease.

A

Fig. 54-78). Gradient echo images are also useful for detecting venous tumor extension.92,269 On such images, flowing blood has significantly increased signal intensity and appears white, whereas tumor thrombus has medium signal intensity and appears as an IV filling defect (Fig. 54-79).92,269 MRA 2D or 3D gradient echo sequences with Gd enhancement are most widely used. The 3D slab is centered over the renal vessels in the coronal plane.251 Sagittal or coronal MRI accurately evaluates the superior extent of vena caval thrombus relative to the diaphragm, hepatic veins, and right atrium.106,269 The sensitivity of MRI has been reported as 90% for detection of IVC thrombus, compared with 79% for CT (Fig. 54-80).139 Regional lymph nodes are considered to be involved by tumor if they are at least 1 cm in diameter in the short axis. However, lymph nodes may be 1 to 2 cm in diameter because of reactive hyperplasia.134 Enlargement of lymph nodes greater than 2 cm is usually due to metastases (see Fig. 54-74).134,185 Adjacent organ invasion is diagnosed only if an adjacent structure has enlargement, density change, or both (Fig. 54-81). The loss of a fat plane between the tumor and an adjacent organ is not reliable evidence of tumor invasion (see Fig. 54-81).134 RCC most commonly metastasizes to the lungs and mediastinum, bones, and liver (see Fig. 54-74). Less common sites are the adrenal glands, kidney, brain, pancreas, mesentery, and abdominal wall.74 Postnephrectomy evaluation.  Recurrence of neoplasm in the renal bed occurs in about 5% of patients after radical nephrectomy, as detected on follow-up CT. Patients with large invasive tumors have a higher risk of local recurrence than those with small early-stage tumors.104 Renal bed recurrence is most likely within 2 years of nephrectomy.349 The liver, ascending colon, second part of duodenum, pancreatic head, and small bowel migrate into the right renal fossa; the pancreatic tail, spleen, and large and small bowel migrate into the left renal fossa. Interpretational errors can be avoided in patients who have undergone nephrectomy by ensuring adequate bowel opacification with oral contrast medium. Postoperative scarring in the renal fossa is distinguished from neoplasm recurrence if the area shows no change on serial CT examinations.7 Local recurrence of neoplasms may also be suggested by enlargement or irregularity of the psoas muscle on the side of the nephrectomy (Fig. 54-82), although this finding may also be caused by

B FIG 54-75  A, Left renal cell carcinoma with renal vein enlargement (arrow) on contrast-enhanced CT scan. The right kidney is atrophic. B, On MRI (spin echo, TR = 2100 msec, TE = 30 msec) the left renal vein (arrows) is enlarged but shows a normal signal void, indicating that venous enlargement is probably due to increased blood flow rather than tumor extension. The inferior vena cava (open arrow) is also normal.

CHAPTER 54  Kidney

1813

B

A

C FIG 54-76  Renal cell carcinoma with inferior vena cava (IVC) extension. A, T1-weighted Gd-enhanced MRI of the abdomen shows a large right renal lesion. B, Image at slightly higher level shows extension of the tumor into the IVC (arrow). Also seen incidentally are multiple enhancing liver lesions (arrowheads) consistent with metastatic disease. C, Coronal reconstruction of contrast-enhanced image of the same patient again shows the renal cell carcinoma with direct extension into the IVC. Thrombus within the IVC shows heterogeneous enhancement and is in continuity with the primary tumor, suggesting a tumor thrombus.

postoperative scarring.21 In all patients undergoing radical nephrectomy or partial nephrectomy for RCC, a baseline postoperative CT scan of the abdomen should be obtained and CT scanning should be performed at regular intervals thereafter. Treatment.  Nephron-sparing surgery is the treatment of choice when radical nephrectomy would render the patient anephric with  a subsequent need for dialysis.290 Current accepted indications for nephron-sparing surgery are (1) RCC in a solitary functioning kidney, (2) RCC in a patient with compromised renal function, and (3) multiple bilateral tumors, which are common in VHLD and hereditary RCC.251,290 The most suitable lesion for nephron-sparing nephrectomy is smaller than 4 cm, polar, cortical, and far from the renal hilum and collecting system.290

Recently the R.E.N.A.L. nephrometry score has been introduced to describe the anatomic characteristics of the renal tumor, considering five specific descriptors of the mass. The scoring system is based on the five most reproducible features that characterize the anatomy of a solid renal mass: (R)adius (scores tumor size as maximal diameter), (E) xophytic/endophytic properties of the tumor, (N)earness of the deepest portion of the tumor to the collecting system or renal sinus, (A)nterior (a)/posterior (p) descriptor, and (L)ocation relative to the polar line. All components except for the (A) descriptor are scored on a 1, 2, or 3-point scale. This standard scoring system helps in surgery treatment decision making and potentially for predicting outcome.165 Collecting duct carcinoma.  Collecting duct carcinoma (Bellini duct carcinoma), originating from the collecting ducts of the kidney,

A

B

FIG 54-77  Significant retrocaval extension of renal cell

C

A

carcinoma with extrinsic vena caval invasion. The relationship of the vena cava to the tumor was not adequately determined by CT. A, Sagittal nonenhanced MRI (spin echo, TR = 500 msec, TE = 15 msec) shows retrocaval tumor extension with anterior caval displacement (open arrow). B, MRI (spin echo, TR = 500 msec, TE = 15 msec) performed after IV Gd-DTPA administration reveals marked tumor enhancement (arrow). The vena cava (open arrow) is displaced anteriorly and compressed. C, More cephalic MRI shows extrinsic caval tumor invasion (arrow) near the level of entry of the hepatic veins.

B

FIG 54-78  Inferior vena caval extension of right renal cell carcinoma. A, Nonenhanced MRI (spin echo, TR = 500 msec, TE = 17 msec) shows neoplasm extension into the right renal vein (black arrow) and inferior vena cava (white arrow). The venous tumor has a signal intensity similar to that of the primary neoplasm. B, More cephalic MRI shows intracaval tumor (arrow) at the level where the three hepatic veins join the vena cava. C, Sagittal MRI shows intracaval tumor (black arrows). The supradiaphragmatic part of the vena cava (white arrow) is not affected.

C

CHAPTER 54  Kidney

1815

FIG 54-79  Inferior vena caval extension of left renal cell carcinoma. A gradient echo MRI (TR = 60 msec, TE = 12 msec, flip angle = 60 degrees) shows tumor extension along the left renal vein (arrow) and into the inferior vena cava. Flowing blood in the vena cava appears white and the tumor is shown as a filling defect (arrowhead). (Courtesy Richard L. Ehman, MD, Rochester, Minn.)

FIG 54-80  MRI obtained after Gd enhancement in the same patient shown in Figure 54-62 demonstrates enlarged renal vein with enhancement (arrow). At surgery the enlargement was confirmed to be tumor extension into the renal vein.

p

A

B FIG 54-81  Left renal cell carcinoma with extension beyond the renal fascia. A, Tumor invades the abdominal wall in the region of the left quadratus lumborum muscle (arrow). Tumor has grown along a needle biopsy track in the subcutaneous tissues (arrowhead). Tumor invasion of the descending colon (open arrow) was confirmed at surgery. Although the neoplasm abuts the left psoas muscle, there was no actual muscle invasion. B, Contrast-enhanced MRI (spin echo, TR = 500 msec, TE = 15 msec) reveals that the neoplasm does not invade the psoas muscle (p). However, tumor has invaded the abdominal wall (long arrow) and colon (short arrow).

A

B FIG 54-82  Local recurrence of renal cell carcinoma 7 years after nephrectomy for right renal cell carcinoma. The patient presented with upper gastrointestinal bleeding. A, An enhancing mass (curved arrow) causes enlargement of the right psoas muscle. The third part of the duodenum (straight arrow) and the inferior vena cava (arrowhead) are adherent to the mass. B, MRI (spin echo, TR = 500 msec, TE = 15 msec) performed after IV injection of Gd-DTPA shows an enhancing mass (arrow) involving the right psoas muscle. Endoscopy with biopsy showed clear cell carcinoma invading the third part of the duodenum (open arrow).

1816

PART II  CT and MR Imaging of the Whole Body

FIG 54-83  Collecting duct carcinoma. Contrast-enhanced CT scan of the kidneys demonstrates a large, heterogeneous, infiltrating left renal lesion replacing the normal renal parenchyma. It was surgically resected and confirmed to be a collecting duct carcinoma.

accounts for only 1% of renal carcinomas.142 The age at presentation ranges from 16 to 62 years, with median age of 43 years.73 Hematuria is the most common presentation, and most patients have  metastases at presentation.73,203 Collecting duct tumors demonstrate aggressive behavior; a maximum survival of approximately 2 years is reported.73,149,203 CT demonstrates renal medullary involvement with an infiltrative appearance and renal sinus encroachment. The reniform contour  of the kidney is preserved except when an expansile component is present (Fig. 54-83). Collecting duct carcinomas are hypointense on T2-weighted images and hypovascular on angiography.243-245 Wilms’ tumor.  Wilms’ tumor accounts for 87% of pediatric renal neoplasms.138,263 Approximately 80% of children with this tumor present between 1 and 5 years of age, with the peak incidence between 3 and 4 years.38 Occasional cases are encountered during adult life.93 Most patients present because of an enlarging abdominal mass.121 Less common presenting symptoms include abdominal pain, fever, and microscopic or gross hematuria.100 The lung is the most frequent site of metastatic disease, and the liver is less frequently involved. With appropriate therapy about 95% of patients with Wilms’ tumor can be cured.350 Tumor histology and stage are the most useful predictors of patient outcome and help determine the nature of therapy. Although all Wilms’ tumors require surgical resection, preoperative imaging helps determine tumor stage and resectability, therefore influencing decisions about preoperative chemotherapy or radiation therapy.60,350 US is the most common method for initial diagnosis of Wilms’ tumor. However, CT and MRI have been found to be more accurate than US in tumor staging.60,292 Chest and abdominal CT may be performed during the same examination to determine the presence or absence of lung metastases.100 On CT a Wilms’ tumor usually appears as a large, spherical, intrarenal mass, very often with a well-defined rim of compressed renal parenchyma or pseudocapsule surrounding it (Fig. 54-84). Tumors that arise in the peripheral cortex may grow in an exophytic manner, with most of the tumor mass outside the kidney.93 The tumor is less dense than normal renal parenchyma on contrast-enhanced CT scans; areas of low attenuation coincide with tumor necrosis, fat deposition, or both (Fig. 54-85).100 About 13% of Wilms’ tumors contain calcifications as seen on nonenhanced CT scans.93,100 About 10% of patients show poor or absent function of the involved kidney because of venous

FIG 54-84  A large heterogeneous Wilms’ tumor is present in the left kidney. The residual renal parenchyma is displaced medially (arrows).

FIG 54-85  Wilms’ tumor of the left kidney with extensive necrosis. Contrast-enhanced CT scan shows that the tumor has a mainly fluid density. Solid and enhancing tumor elements are present medially (arrows).

tumor extension, compression of the collecting system, or extensive tumor infiltration throughout the kidney. Perinephric tumor extension thickens the renal fascia and obliterates the perinephric fat, but subtle perinephric extension may be missed on CT. Central retroperitoneal adenopathy may be detected on CT, although the distinction between nodes enlarged as a result of metastases and those enlarged because of reactive changes is usually not possible.60,292 Renal vein and IVC tumor extension is best shown after an IV bolus injection of contrast medium. The thrombus may be seen as a low-density intraluminal filling defect. CT may also identify bilateral Wilms’ tumors (Fig. 54-86), present in about 10% of patients,193 and liver and pulmonary metastases.60,259 On MRI, Wilms’ tumor appears as a large well-defined mass  with relatively distinct margins (Fig. 54-87). It has low signal intensity on T1-weighted images and high signal intensity on T2-weighted images.193 The tumor often appears heterogeneous on both T1- and T2-weighted images. Although MRI is sensitive for detecting regional lymph node enlargement, it cannot predict the cause of the enlargement.292 Currently CT and MRI appear equivalent for staging Wilms’ tumor. However, MRI shows venous extension better than CT (see  Fig. 54-87).292

CHAPTER 54  Kidney

1817

Nephroblastomatosis.  Nephrogenesis is normally complete by 34 to 35 weeks’ gestation. However, metanephric blastema may persist into infancy and childhood. Foci of persistent metanephric tissue are designated as nephrogenic “rests.” The presence of multiple nephrogenic rests is termed nephroblastomatosis. Nephrogenic rests are characterized as intralobar or perilobar. Intralobar rests are solitary and randomly distributed in the kidney, more centrally in the renal lobe. They are associated with two syndromes: WAGR syndrome (Wilms’ tumor, aniridia, genital anomalies, retardation) and Denys-Drash syndrome (DDS). Perilobar rests are multiple and peripherally located in the kidney. They are associated with Beckwith-Wiedemann syndrome and hemihyperplasia. There is wide acceptance of nephroblastomatosis as a precursor lesion to Wilms’ tumor; 30% to 34% of Wilms’ tumors arise from these rests. Approximately 35% of patients with diffuse nephroblastomatosis eventually have Wilms’ tumor.191 Hyperplastic nephrogenic rests are macroscopic and plaquelike (Fig. 54-88). Any macroscopic nephrogenic rest that is nodular and enlarges over time is considered neoplastic.348 Nephroblastomatosis occurs most often in neonates and is characterized by multiple, often bilateral, subcapsular masses.193 CT is the technique of choice for detection of this disorder. On contrast-enhanced CT, in addition to nephromegaly, the kidney may

show striate enhancement. The rests enhance less than adjacent normal parenchyma.244 CT is particularly useful for evaluating the other kidney in the child with Wilms’ tumor, for renal screening in children with syndromes associated with Wilms’ tumor, and for follow-up evaluation for neoplastic changes in patients with known nephroblastomatosis.89,348 The hallmark of neoplastic transformation of a benign nephrogenic rest is enlargement on serial CT scans.348 On MRI, nephroblastomatosis tends to be hypointense to renal cortex on T1-weighted images and isointense to cortex on T2-weighted images.112 Gd enhancement adds to the conspicuity of these lesions, although contrast-enhanced studies have not proved superior to nonenhanced studies for lesion detection.100 Renal sarcoma.  Primary renal sarcomas are rare (accounting for ≈ 1% of malignant renal parenchymal tumors) mesenchymal tumors that generally have a poor prognosis. Subtypes of renal sarcoma include leiomyosarcoma, angiosarcoma, hemangiopericytoma, liposarcoma, rhabdomyosarcoma, fibrosarcoma, and osteosarcoma. The biological behavior of the various types of sarcoma is highly variable.244,305 Leiomyosarcoma is the most common renal sarcoma, accounting for about 58% of all sarcomas.85 Hemangiopericytoma and liposarcoma each account for about 20% of renal sarcomas.85 Renal rhabdomyosarcoma and osteogenic sarcoma occur rarely. Capsular localization, a feature

FIG 54-86  Multicentric and bilateral Wilms’ tumors. The tumors have

FIG 54-88  A hyperplastic intralobar nephrogenic rest presents as a

a low density and compress residual enhancing renal tissue.

A

low-density subcapsular mass (arrow) on a contrast-enhanced CT scan.

B FIG 54-87  Wilms’ tumor of right kidney shown by MRI (spin echo, TR = 600 msec, TE = 15 msec). A, Axial

image reveals that the mass is well defined and displaces normal renal tissue (arrow) posteriorly. B, Coronal image shows that the mass indents the inferior vena cava (arrows) but does not extend to the vena cava.

1818

PART II  CT and MR Imaging of the Whole Body

of more than 50% of these tumors,85 should suggest the diagnosis on CT (Figs. 54-89 and 54-90). When these neoplasms arise in the renal parenchyma, however, they are indistinguishable from RCC on CT. Renal liposarcomas usually arise in the renal capsule or perinephric fat (see Fig. 54-89).85 On CT, liposarcomas cause compression without invasion of the renal parenchyma and show a variety of appearances correlating closely with their gross and histologic features.166 Tumors containing a large amount of mature fat show negative attenuation values. Myxoid liposarcomas contain little mature fat, and their predominantly fluid and connective tissue composition results in attenuation values nearer those of water.166 On CT, undifferentiated liposarcomas are indistinguishable from other types of sarcoma (see Fig. 54-89). The diagnosis of renal osteogenic sarcoma is suggested on CT by a peripherally located renal mass containing dense ossification (see Fig. 54-90).357

Secondary Neoplasms Renal lymphoma.  Renal involvement by lymphoma is usually secondary to systemic disease because the kidney does not contain lymphoid tissue.52,244,332 Secondary renal involvement occurs commonly either from generalized hematogenous dissemination of lymphoma or by direct extension of retroperitoneal disease.52 Bilateral renal involvement occurs more frequently, being present in 75% of all patients with lymphoma. Renal involvement occurs significantly more commonly in

R I G H T

FIG 54-89  Undifferentiated liposarcoma (arrows) arising in the perinephric fat. The neoplasm is confined in the renal fascia (arrowheads). The left kidney is displaced laterally, but the tumor has not invaded the renal parenchyma.

A

non-Hodgkin’s lymphoma than in Hodgkin’s disease.332 CT patterns of renal lymphoma are (1) multiple renal masses, (2) solitary mass, (3) renal invasion from contiguous retroperitoneal disease, (4) perirenal disease, and (5) diffuse renal infiltration.332 With extensive use of abdominal CT for lymphoma staging, renal lymphoma is now recognized on CT in about 8% of patients with lymphoma.52 Multiple renal nodules are the most common CT manifestation of renal lymphoma, occurring in about 59% of cases (Fig. 54-91).52,260 Nodules range from 1 to 3 cm, are less dense than normal renal parenchyma on contrast-enhanced CT scans, and typically show a homogeneous appearance. As multiple renal nodules grow, coalesce, and replace nephrons, the kidneys are almost diffusely replaced by lymphoma (Fig. 54-92). However, the multinodular nature of the disease process usually remains apparent. In most patients, the renal arteries and veins remain patent despite tumor encasement, a finding characteristic for lymphoma.332 Retroperitoneal lymph node enlargement is not visible on CT in a significant percentage of patients with renal lymphoma not due to direct spread of retroperitoneal tumor (see Fig. 54-92).52,260 Renal involvement from retroperitoneal lymphoma commonly occurs either by invasion through the renal capsule or by extension through the renal sinus (Fig. 54-93). This pattern accounts for about 28% of patients with renal lymphoma. Retroperitoneal adenopathy is invariably present, and many patients show obstructive hydronephrosis. Rarely a solitary mass in one kidney is the only CT manifestation of renal lymphoma (≈3% of cases; Fig. 54-94). Some patients (≈10%) show a predominance of perinephric disease without extensive parenchymal or retroperitoneal disease (Fig. 54-95). On MRI, renal lymphoma tends to be hypointense relative to the renal cortex on T1-weighted images and heterogeneously hypointense or isointense on T2-weighted images. Minimal enhancement is generally detected, although renal lymphoma enhances considerably less than normal renal parenchyma. MRI characteristics of treated lymphoma are variable.52,251 Metastases.  The frequency of metastases to the kidney in patients with cancer is 7% to 13% in large autopsy series.37,236 The kidney is the fifth most common site of metastases in the body, after lung, liver, bone, and adrenals; most metastases reach the kidney by the hematogenous route.211 The three neoplasms with the highest frequency of renal metastases are lung carcinoma, breast carcinoma, and carcinoma of the other kidney. Renal metastases are usually discovered incidentally during abdominal CT performed for tumor staging. However, some patients have hematuria and flank pain.211 The most common pattern on CT is that of multiple discrete bilateral lesions. Solitary exophytic metastases are

B FIG 54-90  A and B, Renal osteosarcoma. A densely ossified mass (arrows) arises peripherally in the lower pole of the right kidney. (Courtesy Clive Levine, MD, Columbia, Mo.)

CHAPTER 54  Kidney

FIG 54-91  Non-Hodgkin’s lymphoma with bilateral multiple renal masses. There is extensive retroperitoneal adenopathy (arrowheads).

1819

FIG 54-94  Hodgkin’s disease with solitary heterogeneous left renal mass.

FIG 54-92  Bilateral renal replacement and mild nephromegaly in nonHodgkin’s lymphoma. There is no enlargement of retroperitoneal lymph nodes.

FIG 54-95  Right perinephric mass (arrowheads) shown on nonenhanced CT scan in a patient with non-Hodgkin’s lymphoma. A large nodal mass (arrow) involves the small bowel mesentery.

primary neoplasm, such solitary renal lesions can usually be safely assumed to be metastases, and pathologic proof of their nature is unnecessary (Fig. 54-97).50,211 However, in the patient with no other metastases, it is difficult to determine whether the renal lesion is a synchronous RCC or a renal metastasis (Fig. 54-98). In these circumstances, further investigation is necessary, particularly if there has been an apparent remission of the primary neoplasm. Tissue sampling is usually performed in such cases.

Renal Pseudotumors

FIG 54-93  Renal involvement from retroperitoneal non-Hodgkin’s lymphoma. There is extensive central retroperitoneal adenopathy causing lateral displacement of the kidneys. The upper pole of the left kidney (arrow) is invaded through the renal capsule. Tumor extends into the right kidney through the renal sinus (arrowhead).

more common in patients with colon cancer, and perinephric tumor extension is typical of melanoma (Fig. 54-96).50 These lesions are best seen on contrast-enhanced CT scans. Although metastases are usually homogeneous on CT, larger lesions may show central necrosis.211 Occasionally patients with metastases have solitary renal masses.50 If the patient has other areas of metastatic disease and a current known

Certain renal masses that are composed of normal and benign renal tissue may mimic renal tumors. Such masses are termed renal pseudotumors.22 It is important to identify such pseudomasses because some of them are simply normal renal variants and do not require any further management or imaging. Some of the normal renal variants include prominent column (septum) of Bertin, dromedary hump,  persistent fetal lobulations, and congenital splenorenal fusion (Fig. 54-99).331 Other benign infectious conditions such as focal pyelonephritis, xanthogranulomatous pyelonephritis, and renal abscess may also look like renal tumors on imaging. These are benign conditions and may require only conservative management, unlike a true renal tumor, which requires a partial or radical nephrectomy. Some vascular lesions, such as arteriovenous malformation, renal extramedullary hematopoiesis (Fig. 54-100),22,258 and subepithelial pelvic hematomas (Antopol-Goldman lesions; Fig. 54-101),22,265 by virtue of their location and appearance, may mimic malignancy.

1820

PART II  CT and MR Imaging of the Whole Body A high index of suspicion for renal pseudotumor can avoid unnecessary imaging follow-up and surgery.

Benign Renal Neoplasms

Adenoma.  During the last decade there has been a significant increase

FIG 54-96  Renal metastases from colonic carcinoma. Axial contrastenhanced CT scan shows a low-attenuating mass with minimal enhancement in the inferior pole of right kidney (arrow). Multiple lowattenuating lesions are also seen in the liver, which are metastases from colonic carcinoma.

in detection of asymptomatic renal tumors 3 cm or less in diameter by sonography and CT (Fig. 54-102).184,340 Most such tumors are neoplasms of tubular origin. In the past, such lesions were often classified as renal adenomas because they rarely metastasize. However, serial imaging shows that they often gradually enlarge (Fig. 54-103)24 and that even small tumors may metastasize (Fig. 54-104).298 Indeed, such lesions are histologically indistinguishable from larger carcinomas more commonly encountered in the kidneys, and they often have histologic grades indicating a definite potential for malignant behavior. Accordingly, most investigators believe these lesions should be regarded as carcinomas in an early stage of evolution and should not be classified as adenomas.24,184 On CT these small renal tumors are usually well defined and homogeneous in appearance (see Fig. 54-102).184,298 Less commonly they show marginal irregularity, heterogeneity, or central calcification (see Fig. 54-66), findings that may suggest pathologically

FIG 54-98  Solitary renal metastasis from squamous cell carcinoma of FIG 54-97  Left renal metastasis (arrow) and retrocrural adenopathy (arrowhead) in a patient with small cell lung cancer. A follow-up scan after 3 weeks of chemotherapy showed complete resolution of the renal mass and adenopathy.

A

the lung. The patient underwent a left upper lobectomy. Two years later, he demonstrated left flank pain and hematuria. CT scan shows a heterogeneous left renal mass (arrow). No other lesions were found. A renal cell carcinoma was suspected, but percutaneous biopsy revealed metastatic squamous cell carcinoma.

B FIG 54-99  Splenorenal fusion. Coronal T2-weighted (A) and Gd-enhanced (B) MRIs of the kidneys demonstrate a well-defined mass (arrow) in the lower pole of left kidney, which shows low high signal on the T2-weighted image and intense enhancement after Gd administration. This mass was surgically resected and found to be consistent with splenic tissue.

CHAPTER 54  Kidney high-grade tumors.34 These lesions are best managed with partial nephrectomy,311,334 provided the other kidney is normal.

Oncocytoma.  The renal oncocytoma is a solid epithelial neoplasm with a generally benign course.70 It represents between 10% and 15% of small (<3 cm) solid renal neoplasms.267 They arise from proximal tubular epithelium (oncocyte) and range in size from 1.5 to 13 cm, with a mean of 6.3 cm.123 On CT, oncocytomas typically are well-defined masses with smooth rounded margins (Fig. 54-105).66 Tumor calcification occurs rarely, and oncocytomas are sometimes multiple and bilateral (see Fig. 54-63).59,127 Small oncocytomas are usually homogeneous in appearance on contrast-enhanced CT scans, although they are occasionally heterogeneous because of the presence of central scars (Fig. 54-106). On CT, small oncocytomas are usually indistinguishable from slowly growing small RCCs that lack hemorrhage or necrosis.66 A central sharply defined stellate scar (Fig. 54-107) is present in 25% to 33% of large oncocytomas and strongly suggests the diagnosis.255

1821

The classic angiographic findings for oncocytoma are a spokewheel pattern, a homogeneous nephrogram, and a sharp smooth rim. The finding of a homogeneous blush or a spoke-wheel pattern greatly increases the possibility of an oncocytoma, although RCC may have any or all of the classic findings described for oncocytoma.346 However, CT criteria are usually poor for distinguishing between oncocytoma and RCC, regardless of tumor size.66 Fine-needle aspiration cytology has been shown to be successful in discriminating between RCC and renal oncocytoma.190 If the diagnosis of oncocytoma is suspected before surgery, small tumors may be treated with partial nephrectomy because of the excellent prognosis.252 On MRI, oncocytoma is generally isointense to hypointense to normal parenchyma on T1-weighted images and has a variable appearance on T2-weighted images. The presence of a central scar suggests the diagnosis (Fig. 54-108). Oncocytomas do enhance with Gd administration, but less than renal parenchyma.

Neoplasm (Reninoma).223  Renin-producing tumors of juxtaglomerular cells (reninomas) are a rare78 but curable cause of hypertension. Two thirds of juxtaglomerular cell tumors of the kidney occur in young women in their reproductive years.121 Plasma renin and aldosterone values are usually elevated. None of the reported Juxtaglomerular

L SP

FIG 54-100  An 82-year-old woman with renal extramedullary hematopoiesis. Axial nonenhanced CT scan of the right kidney shows multiple perirenal masses (arrowheads) with severe splenomegaly (SP). The mass was sampled for biopsy on suspicion of malignancy, and microscopic findings were characterized by the presence of megakaryocytes and granulocyte precursors. Presence of these blood cell precursors suggests a hematopoietic process within these masses, consistent with renal extramedullary hematopoiesis.

A

FIG 54-102  Small (1.8-cm) renal cell carcinoma (arrow) detected incidentally by CT.

B FIG 54-101  Subepithelial pelvic hematomas (Antopol-Goldman lesions). A, Nonenhanced CT scan of the kidneys demonstrates a hyperdense lesion (arrow) in the left renal pelvis. Ureteroscopy was performed and did not show presence of any mass. Incidentally seen is a left renal cyst (arrowhead). B, Nonenhanced CT image obtained after 3 months revealed resolution of the hyperdense lesion. This is consistent with subepithelial pelvic hematoma, which resolves without treatment.

1822

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 54-103  Progression of renal cell carcinoma shown by serial CT in a 70-year-old man who underwent bladder carcinoma resection 6 years earlier. The patient initially refused surgery. A, Contrast-enhanced CT scan shows a 2.5-cm renal mass (arrow). B, Follow-up scan obtained 3 years later shows that the mass (arrow) has enlarged to 3.8 cm. C, Nonenhanced CT scan obtained 2 years after that in B shows further enlargement of the mass (arrow), which measures 6.5 cm.

A

B FIG 54-104  Small renal cell carcinoma with sacral metastasis. The patient was a 42-year-old man who complained of low back pain. A, A lytic metastasis (arrowhead) is present in the right sacral ala. B, CT performed to search for a primary neoplasm shows a small mass (arrow) with dense peripheral calcification in the left kidney. Histologic examination revealed a high-grade left renal cell carcinoma.

CHAPTER 54  Kidney

1823

neoplasms has been invasive or metastasized. Reninomas are usually well shown on contrast-enhanced CT scans, on which they show a smooth outline and sharp margination. Small foci of hemorrhage may cause a heterogeneous tumor appearance. Because of their benign nature, reninomas may be managed by partial nephrectomy.223 On MRI, reninomas are isointense to hypointense to normal cortex on T1-weighted images and hypointense or hyperintense to normal cortex on T2-weighted images. These tumors tend to enhance less than normal cortex on MRI because of their relative hypovascularity  (Fig. 54-109).2,141,251

Multilocular Cystic Renal Tumor.  Multilocular cystic renal tumor

FIG 54-105  Left renal oncocytoma. Contrast-enhanced CT scan shows a homogeneously enhancing well-circumscribed exophytic renal mass (arrow).

FIG 54-106  Left renal oncocytoma (arrow) with eccentric scar (arrowhead). Distinction from a small renal cell carcinoma is not possible on CT.

FIG 54-107  Large left renal oncocytoma (arrow). The centrally placed, sharply defined stellate area of low attenuation (arrowheads) suggests the diagnosis of oncocytoma.

(MCRT) refers to two types of generally benign tumors: cystic nephroma and cystic poorly differentiated nephroblastoma (CPDN).137 These are well-encapsulated, multilocular, noninfiltrating cystic lesions with no solid components. MCRT manifests in young children (usually boys) as a painless abdominal mass. Affected girls usually are older than 4 years, and the tumor demonstrates a biphasic age distribution, occurring in girls and women 4 to 20 years of age and in women in their fifth and sixth decades of life.196 The masses are often large, averaging 8 to 10 cm in diameter, single, and unilateral, and they involve only part of the kidney. Individual cysts may vary from microscopic to 2 to 4 cm.224 CPDN resembles cystic nephroma in every feature on histopathology and imaging examinations, except that the septa contain blastema cells with or without other embryonal stromal cells. On CT, MCRT presents as a well-defined encapsulated mass containing multiple cysts separated by septa (Fig. 54-110). The septa may be as thick as several millimeters, but they do not display significant nodularity. On contrast-enhanced CT, septa in the mass may enhance.196 Excretion of contrast by the affected kidney is usually normal. However, the lesion frequently herniates into the renal pelvis, causing a filling defect and sometimes obstructive calycectasis (Fig. 54-111).196 MRI does not add much information to that provided by CT (see Fig. 54-111). MRI shows an encapsulated multilocular mass with a capsule that has low signal on all pulse sequences, variable signal intensity of cyst contents on T1-weighted images, and high signal intensity of cyst contents on T2-weighted images.100 Cystic nephroma and CPDN are indistinguishable by imaging or by gross appearance. The treatment for MCRT is complete resection, with noninvasive imaging for follow-up if the tumor proves to be CPDN.100

Angiomyolipoma.  Angiomyolipomas are benign hamartomas composed of blood vessels, smooth muscle, and fatty tissue. AML is seen in two distinct clinical forms, a sporadic (isolated) form and in association with TSC. Although exact figures are difficult to obtain, the sporadic form accounts for approximately 80% to 90% of cases of AML.341 The sporadic form is typically seen in middle-aged patients (mean age, 43 years) and is more common in women by at least a 4 : 1 ratio. Up to 80% of all patients with TSC, including adults and children, have AMLs.100 When found in children, AMLs are almost exclusively associated with TSC.100 Masses less than 4 cm in diameter are usually asymptomatic, but those larger than 4 cm often cause flank pain, hematuria, and anemia and carry a significant risk of intratumoral or perinephric hemorrhage.308,335 AMLs are seen on CT as well-circumscribed renal masses. They vary in size from tiny renal nodules to large tumors (Fig. 54-112). Patients with TSC often have extensive bilateral renal involvement by AMLs (Fig. 54-113). The presence of intratumoral fat is almost diagnostic of AMLs,36 although in rare cases, fat has been reported in Wilms’ tumor in children235 and in renal oncocytomas.59 RCCs may also rarely show presence of fat cells.120,171,280,309 Retroperitoneal lymph node involvement may be discovered in patients with AML.316

1824

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 54-108  A 78-year-old woman with oncocytoma. A, Coronal T1-weighted MRI shows a mass in the inferior pole of the right kidney; the mass is slightly hyperintense to the renal parenchyma. B, T2-weighted fast-spin echo image shows the mass to be hyperintense to the surrounding renal parenchyma with a highsignal central scar. Precontrast T1-weighted axial fast multiplanar spoiled-gradient image (C) shows the mass to be slightly hyperintense to the renal parenchyma with low-signal central scar, which demonstrates enhancement on post-Gd enhancement sequences (D).

CHAPTER 54  Kidney

Long LT kidney inferior I

B

A

D

C FIG 54-109  Reninoma. A, Longitudinal sonogram of the left kidney shows an isoechoic well-defined mass (arrow) at the inferior pole of the kidney. On MRI this mass shows low signal intensity (arrow) on T1-weighted sequence (B) and high signal intensity (arrow) on T2-weighted sequence (C), with minimal enhancement (arrow) on Gd-enhanced sequence (D).

1825

1826

PART II  CT and MR Imaging of the Whole Body The presence of fat is best shown by CT. Thin-section (3-5 mm) technique is preferred to avoid volume averaging, which might prevent demonstration of low attenuation in the tumor. Use of IV contrast material is not necessary (see Fig. 54-112).341 Fatty tissue is considered to be present in a tumor if a region-of-interest attenuation measurement of −10 HU or lower is found within the tumor. Lesions without detectable fat cannot be distinguished from other renal neoplasms on CT (Fig. 54-114). Approximately 5% of AMLs do not show fat attenuation on CT scans and cannot be differentiated from RCC.341 Extensive intratumoral hemorrhage may also obscure the presence of fat. AMLs are always benign. However, the tumors may exhibit extrarenal extension. Large perinephric tumor components may be found (Fig. 54-115), and AMLs may extend into the renal vein and IVC.100 MRI may also detect the presence of fat in renal AMLs (Fig. 54-116). The MRI characteristics of AML depend on the relative amounts of fat and other tissue components. The presence of macroscopic fat in a lesion is very specific for AML, although rarely RCC can also contain fat.120 AML is identified on opposed-phase chemical-shift MRI by the presence of India-ink artifact (loss of signal at the interface of  fat and non–fat-containing areas) at the interface of mass and renal parenchyma, or within the renal mass.132 Non–fat-containing exophytic renal tumors demonstrate India-ink artifact only at the interface

L

FIG 54-110  Multilocular cystic nephroma (white arrow) in the right kidney in an infant. Multiple septa separate the fluid-containing loculi. The renal parenchyma is displaced medially and posteriorly (black arrows).

Front

A

B

C

D FIG 54-111  Multilocular cystic nephroma in the right kidney in an infant. A, CT scan shows fine septa in a well-defined cystic renal mass (arrow). Hydronephrosis (open arrow) was due to herniation of the tumor into the right renal pelvis. B, Coronal MRI (spin echo, TR = 350 msec, TE = 15 msec) obtained posteriorly through the right kidney shows dilated calyces (white arrows) and renal pelvis (black arrow). C, Coronal T1-weighted MRI (spin echo, TR = 350 msec, TE = 15 msec) shows that the tumor has a heterogeneous appearance with area of high signal intensity probably due to hemorrhage or high protein content in cyst fluid. D, Axial protondensity–weighted MRI (TR = 2000 msec, TE = 22 msec) shows that fluid loculi have significantly variable signal intensities.

CHAPTER 54  Kidney

A

1827

B

FIG 54-112  Small renal angiomyolipoma. A, Nonenhanced CT scan shows a low-density mass (arrow) in the lower pole of the left kidney. Attenuation values as low as −42 HU, indicating the presence of fat, were found in the lesion through the use of small regions of interest. B, On a contrast-enhanced CT scan, the attenuation values of the lesion (arrow) were in the soft tissue range.

Pelvicalyceal Neoplasms.  About 90% of pelvicalyceal cancers are

FIG 54-113  Tuberous sclerosis with multiple bilateral renal angiomyolipomas. The lesions contain low-density areas consistent with fat.

of the extrarenal component of the renal mass and perinephric fat and not within the kidney. If an AML is suspected on chemical-shift imaging, the diagnosis can be further confirmed by acquiring fatsaturated T1-weighted and water-saturated T1-weighted gradient echo sequences.251 Renal tumors that demonstrate focal (excluding Indiaink artifact) or diffuse signal suppression on opposed-phase gradient echo images but do not exhibit fat suppression on frequency-selective T1-weighted fat-suppressed techniques should raise the possibility of clear cell RCC, although they can also represent AML with microscopic fat. In such cases a biopsy of the mass is usually obtained to confirm the diagnosis.295 Surgery and biopsy are rarely needed in asymptomatic patients with typical imaging findings of AML. Patients with extensive intra­ tumoral or perinephric hemorrhage can usually be managed conservatively (Fig. 54-117). Partial nephrectomy or selective catheter embolization is advocated for masses larger than 4 cm.150,335 Generally, asymptomatic patients with an AML approaching 4 cm in diameter are followed annually with CT or US, although in most of these patients the lesions show little change over time. Patients with lesions larger than 4 cm can be evaluated at shorter intervals.341

Pelvicalyceal and Ureteral Neoplasms Primary neoplasms of the renal pelvis, calyces, and ureter are relatively uncommon. CT plays a limited but often important role in the diagnosis and staging of these tumors.

transitional cell carcinomas, about 9% are squamous cell carcinomas, and less than 1% are adenocarcinomas.108,168 Occupational exposure to industrial dyes, phenacetin abuse, cigarette smoking, prior cyclophosphamide therapy, and urinary stasis associated with horseshoe kidneys predispose an individual to the development of transitional cell carcinoma.168,174 Squamous cell carcinoma of the renal pelvis is commonly associated with chronic irritation secondary to renal calculi and chronic infection. Transitional cell carcinoma has a peak frequency in the sixth and seventh decades of life, and about 68% of affected patients are men. Patients present with painless hematuria.224 Transitional cell carcinoma of the renal pelvis is frequently multicentric, with either synchronous or metachronous tumor involvement of the ureter, bladder, and contralateral renal pelvis.224,352 Imaging findings.  Imaging findings in urothelial cancer consist of a calyceal or renal pelvic filling defect, hydronephrosis with a nonfunctioning kidney, and extensive renal parenchymal infiltration.168 The diagnosis of pelvicalyceal cancer is usually established by various combinations of excretory urography, retrograde pyelography, retrograde brush biopsy, ureteroscopy, and biopsy. CT helps in establishing the extent of renal parenchymal invasion, including the perinephric and periureteral extents of tumor. CT evaluation of pelvicalyceal filling defects of unknown nature shown by excretory urography or retrograde pyelography requires nonenhanced corticomedullary-, nephrographic-, and excretory-phase images through the abdomen and pelvis, using a slice thickness of 3 to 5 mm. Small pelvicalyceal neoplasms may be difficult to detect on precontrast scans, although occasional neoplasms show coarse punctate calcification.75 Contrast-enhanced CT demonstrates collecting system filling defects with smooth, lobulated, or irregular margins (Fig. 54-118).14 Precontrast attenuation values of pelvicalyceal cancers range from 20 to 46 HU.234 Although characteristically hypovascular on angiography, pelvicalyceal tumors often show mild enhancement on CT after IV administration of contrast medium. Postcontrast attenuation values range from 64 to 84 HU.234 Contrast enhancement differentiates pelvicalyceal neoplasms from other filling defects that occur in the collecting system. Radiolucent renal calculi are usually composed predominantly of uric acid and appear as densities with attenuation values greater than 200 HU on nonenhanced CT scans (Fig. 54-119).87,284 Attenuation

1828

PART II  CT and MR Imaging of the Whole Body

A

B FIG 54-114  Two left renal angiomyolipomas in a patient without tuberous sclerosis. A, A lesion (curved arrow) in the upper pole of the left kidney contains fat, indicating the diagnosis of angiomyolipoma. B, A mass (arrow) in the middle region of the kidney contained no detectable fat and was therefore indistinguishable from a renal cell carcinoma. Nephrectomy revealed that both lesions were angiomyolipomas, although the lesion shown in B had only a minimal fat content.

A

B FIG 54-115  Tuberous sclerosis with renal and liver involvement. A, Kidneys show multiple fat-containing lesions. On the right side there is extensive involvement of the perinephric space by an angiomyolipoma (arrows) containing fat and prominent blood vessels. The right kidney is displaced anteromedially. B, A fatcontaining mass (arrow) is present in the left lobe of the liver. (Courtesy Randi W. Hart, MD, Waukesha, Wis.)

A

B FIG 54-116  Multiple angiomyolipoma. A, In-phase MRI in the axial projection demonstrates multiple small lesions in the right kidney that have a slightly higher signal intensity compared with the surrounding renal parenchyma. B, Opposed-phase pulse sequence demonstrates a significant drop in the signal of these lesions, consistent with the presence of fat. Post-Gd enhancement images (not shown) did not reveal any enhancement.

CHAPTER 54  Kidney

A

1829

B

FIG 54-117  Extensive left perinephric hematoma due to hemorrhage from a renal angiomyolipoma. A, The hematoma (arrows) is partially contained by the posterior renorenal bridging septum. B, A more cephalic scan shows a fat-containing angiomyolipoma (arrows) closely related to the hematoma. The bleeding responded to conservative management.

FIG 54-118  Polypoid transitional cell carcinoma (arrow) of the right renal pelvis. The lesion invaded to but not beyond the muscularis. The lesion enhanced on this scan by 36 HU.

FIG 54-119  Urate calculus (arrow) in the left renal pelvis. Excretory urography showed a pelvic filling defect, but the lesion was radiolucent on conventional tomography. The stone has an attenuation value of  458 HU.

values of recent pelvicalyceal blood clots on nonenhanced CT scans range from 45 to 75 HU; these values are higher than those of urothelial cancer and lower than those of urate calculi.168 Moreover, blood clots do not show contrast enhancement, and follow-up studies show disappearance of blood clots with time.168 CT enables the clinician to recognize that an apparent intrinsic pelvic filling defect is produced by extrinsic compression of the renal pelvis by renal sinus lipomatosis, renal sinus cysts, RCCs, and lymphoma. The hydronephrotic form of urothelial cancer with a nonfunctioning kidney is due to ureteropelvic junction obstruction and may present a diagnostic problem on excretory urography.168 Although retrograde pyelography may suggest the diagnosis, CT sometimes provides a definitive diagnosis (Fig. 35-120).282 An enhancing soft tissue mass at the apex of a dilated renal pelvis or diffuse thickening of the renal pelvic wall on CT indicates a tumor causing ureteropelvic junction obstruction. The proximal ureter may be thickened and filled from invasion by the neoplasm (see Fig. 54-120).168 Most cases of ureteropelvic junction obstruction are congenital; on CT a dilated renal pelvis with a smooth wall and a smooth transition at the ureteropelvic junction to the collapsed proximal ureter is seen (Fig. 54-121).282 A calculus impacted at the ureteropelvic junction is easily shown by CT (Fig. 54-122). Extrinsic lesions causing ureteropelvic junction obstruction (e.g., enlarged lymph nodes or retroperitoneal neoplasms) are also easily diagnosed on CT.282 Another presentation of urothelial cancer is that of extensive parenchymal infiltration of the kidney.244 Urothelial cancer is centrally located, and its pattern of renal invasion differs from the eccentric origin and invasion of the renal sinus seen in RCC. Urothelial cancers do not usually distort the shape of the kidney. The demonstration of a central mass that is less dense than normal renal parenchyma and shows minimal tumor enhancement strongly suggests a urothelial cancer (Fig. 54-123).168 The renal sinus fat may be obliterated, and there is often a poorly defined margin between the tumor and the renal parenchyma.168 The calyces may be displaced, compressed, or dilated. Tumor staging and treatment.  Conservative therapy is considered an option in patients with a solitary functioning kidney, bilateral tumor, and renal insufficiency. However, nephroureterectomy with excision of a bladder cuff is considered the standard treatment and should be performed whenever possible.119 When conservative surgery is planned, CT is helpful for tumor staging.209 Patients with high-grade or high-volume pelvicalyceal cancers have a high frequency of

1830

PART II  CT and MR Imaging of the Whole Body

A

B FIG 54-120  Hydronephrosis due to transitional cell carcinoma of the right renal pelvis. A, A soft tissue mass (arrow) fills the right renal pelvis. Dilated calyces (arrowheads) are seen in the right kidney. An enlarged retrocaval lymph node (open arrow) was involved by metastatic disease. B, The tumor also involved the upper half of the right ureter (arrow).

A

B FIG 54-121  Congenital ureteropelvic junction obstruction. A, Longitudinal sonogram of left kidney shows marked hydronephrosis with medial tapering of the renal pelvis. B, Corresponding CT scan reveals the medial tapering of the renal pelvis and marked hydronephrosis. The ureter beyond the junction was collapsed. The patient underwent pyeloplasty.

metastases, primarily to regional lymph nodes, lung, bones, and liver, in decreasing order of frequency. Although CT cannot demonstrate the depth of tumor invasion in the renal pelvic wall (see Fig. 54-118), it can show whether the tumor has extended beyond the tunica muscularis into the renal sinus fat.14 Invasion of renal sinus fat is suggested by obliteration of the peripelvic fat stripe, which is the renal sinus fat separating the anterior and posterior cortical hilar lips from the renal pelvic adventitia (see Fig. 54-123).14 Invasive renal pelvic tumors may become attached to the aorta and IVC and may invade the psoas muscle (Fig. 54-124). In rare cases, tumor extension into the renal vein and IVC may be shown by CT.99 CT may also demonstrate regional lymph node metastases (Fig. 54-125), although the technique sometimes leads to “understaging” of tumors with metastases in normal-sized central retroperitoneal lymph nodes.14 MRI is not used frequently in the diagnosis and management of urothelial tumors. However, it may help determine local extent of tumor. FIG 54-122  Axial CT scan demonstrates a large right ureteropelvic junction calculus.

Ureteral Neoplasms.  Transitional cell carcinoma accounts for about 93% of ureteral neoplasms. Of the remainder, 5% are squamous

CHAPTER 54  Kidney

1831

B

A

FIG 54-123  Transitional cell carcinoma of the right renal pelvis in a 20-year-old man associated with cyclophosphamide therapy administered 6 years earlier. A, Dynamic CT scan obtained immediately after bolus injection of contrast material shows an enhancing mass (arrowheads; attenuation value = 71 HU) in the right renal pelvis. The precontrast attenuation value was 43 HU. The mass obliterates and invades the renal sinus fat. A poorly defined margin between the tumor and the renal parenchyma indicates parenchymal invasion. B, CT scan obtained 5 minutes after that in A shows displacement of the opacified collecting system. (From Levine E: Transitional cell carcinoma of the renal pelvis associated with cyclophosphamide therapy. AJR Am J Roentgenol 159:1027–1028, 1992.)

a T

FIG 54-124  Invasive transitional cell carcinoma of the left renal pelvis. The tumor (T) causes left hydronephrosis. It extends through the wall of the renal pelvis and is attached to the aorta (a) and the left psoas muscle (arrowhead). Complete resection was not possible.

Transitional cell carcinoma of the ureter is most commonly associated with the goblet sign on retrograde ureterography, a cup-shaped collection of contrast seen just distal to the intraluminal ureteric filling defect, which results from the slow growth of a ureteral neoplasm. However, CT may help by confirming that a ureteric filling defect represents a neoplasm and not a radiolucent calculus or blood clot. Ureteral neoplasms manifest on CT as soft tissue intraluminal filling defects (Fig. 54-126), often with ureteral widening or thickening of the ureteral wall. These findings can also occur in inflammatory conditions and ureteral metastases.168 CT also helps determine whether a ureteral neoplasm has extended outside the tunica muscularis to involve the retroperitoneal fat (see Fig. 54-126) and surrounding structures, or whether it has metastasized to regional lymph nodes. An important use of CT is to determine whether ureteral obstruction is due to an intrinsic neoplasm or to extrinsic disease when excretory urography and retrograde pyelography have failed to do so.35 CT is particularly useful in detecting such causes of extrinsic ureteric obstruction as idiopathic and malignant retroperitoneal fibrosis, inflammatory aortic aneurysm, and lymphomatous and metastatic retroperitoneal adenopathy (Fig. 54-127).35

Renal Infections

Acute Bacterial Renal Infection.  Acute bacterial renal infection in

FIG 54-125  Transitional cell carcinoma of the right renal pelvis with extensive parenchymal invasion and regional nodal metastases (arrows).

cell carcinomas and 2% are adenocarcinomas. The age and sex distribution, symptoms, and causes are similar to those for pelvicalyceal carcinoma. Tumors have a predilection for the middle and distal ureter. The diagnosis of ureteral carcinoma is usually readily established with a combination of excretory urography and retrograde pyelography.

adults frequently causes fever, chills, flank pain, leukocytosis, and pyuria. It occurs significantly more commonly in women. Vesicoureteric reflux is sometimes an important predisposing factor in children, but it is only rarely found in adults with acute pyelonephritis. Escherichia coli usually causes the typical ascending acute renal infection, whereas the hematogenous disease is often caused by Staphylococcus aureus.231 The diagnosis of acute bacterial renal infection is usually made clinically, and about 95% of patients with uncomplicated infections treated with appropriate antibiotics become afebrile in 48 hours, and nearly 100% do so in 72 hours.303 Such patients do not need renal imaging. However, imaging studies are valuable in patients with severe refractory renal infections. Imaging is needed to determine if there are complications requiring prolonged antibiotic therapy or surgical intervention (e.g., renal and perinephric abscess or pyonephrosis). Imaging is also useful for excluding abnormalities that predispose to refractory infections, such as nephrolithiasis or ureteric obstruction.

1832

PART II  CT and MR Imaging of the Whole Body

A

B

P

C FIG 54-126  Transitional cell carcinoma of the left ureter. The patient presented with hematuria, and on excretory urography the left kidney did not excrete contrast medium. A, Nonenhanced CT scan reveals left hydronephrosis. B, More caudal scan shows dilated left ureter (arrow) filled with fluid-density urine. C, Scan at the level of the aortic bifurcation shows that the ureter contains a filling defect (arrow) with soft tissue density. The tumor is separated from the left psoas muscle (P) and the left common iliac artery (open arrow) by intact fat planes.

B

A

FIG 54-127  Retroperitoneal metastasis obstructing the ureter. A, Contrast-enhanced CT scan of the kidneys demonstrates severe hydronephrosis (arrow) of the left kidney. B, Axial CT scan at the level of the middle ureter demonstrates an enhancing retroperitoneal mass (arrow) encasing the left ureter and causing obstruction. The patient underwent radical nephroureterectomy, and the lesion was pathologically proven to be a metastasis from squamous cell carcinoma of the vagina.

Acute bacterial pyelonephritis.  In patients with acute refractory bacterial renal infection, CT scans with contrast enhancement are acquired during the corticomedullary phase (30 seconds after initiation of injection) and during either the nephrographic phase (70-90 seconds after injection) or the excretory phase (5 minutes after injection).333 A striated nephrogram on contrast-enhanced CT that consists of discrete rays of alternating attenuation extending to the cortex is characteristic of pyelonephritis and is better demonstrated on CT than on excretory urography (Fig. 54-128).244 Striations result from stasis of

contrast material in edematous tubules that demonstrates increasing attenuation over time.333 Some patients with acute pyelonephritis show diffuse renal involvement on CT. This is manifested by poor renal contrast enhancement, delayed or absent excretion of contrast medium, and global renal enlargement (Fig. 54-129). Other patients show unifocal or multifocal renal abnormalities with extensive areas of apparently uninvolved renal parenchyma.105 These abnormalities are seen on contrast-enhanced CT scans as wedge-shaped zones of decreased attenuation (Fig. 54-130). They often manifest straight borders, radiate

CHAPTER 54  Kidney

1833

FIG 54-131  Acute pyelonephritis with focal mass formation. The kidney shows a rounded heterogeneous mass (arrow) with a poorly defined margin. Inflammatory changes in the adjacent perinephric fat and renal fascial thickening (arrowheads) are also present.

FIG 54-128  Striated nephrogram in acute pyelonephritis. Contrastenhanced CT of the kidneys demonstrates a striated pattern in the kidneys in the nephrographic phase.

FIG 54-129  Acute pyelonephritis with diffuse renal involvement. The affected left kidney shows diminished contrast enhancement, global enlargement, and delayed contrast excretion. The right kidney is normal.

FIG 54-130  Acute pyelonephritis. The left kidney is enlarged and shows several wedge-shaped zones (arrowhead) of decreased attenuation. The lesion exhibits straight borders, radiates from the collecting system to the renal capsule, and is widest at the kidney periphery.

from the collecting system to the renal capsule, and are widest at the periphery of the kidney.105,303,322 Severe tubulointerstitial inflammation may progress to form a hypodense mass or masses with rounded or irregular contours and bulging of the renal surface (Fig. 54-131). If treated too late or inadequately, such lesions may develop single or multiple small areas of liquefaction (i.e., small abscesses).105 These areas are often irregular and of near water attenuation (20-30 HU); they do not enhance after IV administration of contrast medium  (Fig. 54-132). Renal and perinephric abscesses.  If inadequately treated, acute pyelonephritis may progress to tissue necrosis, resulting in renal abscess. On contrast-enhanced CT scans, renal abscesses usually have attenuation values of about 30 HU; this feature distinguishes them from fluid-density renal cysts.303 Their contents do not enhance. They may have distinct rounded margins and thick enhancing walls (Figs. 54-133 and 54-134).172,303 Alternatively, abscesses may be ill-defined and surrounded by zones of decreased parenchymal enhancement, representing inflammation that has not yet progressed to necrosis. Gas is occasionally found in renal abscesses.213 Focal thickening of the adjacent renal fascia and stranding in the adjacent perinephric fat are common CT findings.172 A perinephric abscess may ultimately result if the infection extends through the renal capsule (Fig. 54-135). Although small renal abscesses may respond to antibiotic therapy alone, percutaneous drainage is usually the treatment of choice for larger renal abscesses. Perinephric abscesses always require drainage.213 Abscesses can be drained with the use of CT or US guidance. Emphysematous pyelonephritis.  Emphysematous pyelonephritis (EPN) is a relatively uncommon life-threatening bacterial infection characterized by gas formation within the renal parenchyma or perinephric space.231 Most patients with EPN are diabetic, with an average age of 55 years. E. coli is the organism found most commonly in patients with EPN; other bacteria found include Klebsiella, Pseudomonas, and Proteus species. Two types of EPN have been defined. Type I is characterized by destruction of more than one third of the renal parenchyma, with absence of fluid or fluid collection. A smaller gas collection may be present. Type II is characterized by (1) renal or perirenal fluid collection with bubbly or loculated gas or (2) gas in the collecting system and destruction of less than one third of the renal parenchyma. Type I EPN has a mortality rate of 69%, and type II a

1834

PART II  CT and MR Imaging of the Whole Body

R

L

FIG 54-134  Large right renal abscess. The lesion (arrow) has an attenuation value of 36 HU and shows a thick wall. High-density renal parenchyma on its anteromedial aspect (arrowhead) is probably due to retention of contrast medium in compressed and obstructed renal tubules on the periphery of the abscess.

FIG 54-132  Acute pyelonephritis with early abscess formation. The right kidney shows a small area of liquefaction with enhancing thick walls (arrowhead) along with scattered areas of low attenuation (arrow).

FIG 54-135  Nonenhanced CT scan shows a perinephric abscess (white arrow) complicating acute pyelonephritis in a diabetic patient. Gas collections (curved arrows) are present in the abscess, and the left kidney (arrowheads) is displaced anteriorly.

FIG 54-133  Right renal abscess (arrow) shows a thick wall and low density (30 HU). Inflammatory stranding is present in the perinephric fat.

rate of 18%.342 CT is the most sensitive method for establishing the diagnosis and can also distinguish among gas in the collecting system, parenchymal gas, and gas in the perinephric tissues (Fig. 54-136).263 Gas in the renal pelvis or calyces is a benign finding and may be related to emphysematous pyelitis, representing a gas-forming urinary tract infection that is associated with gas localized only in the renal collecting system (Fig. 54-137). Other causes of gas in the collecting system are ureteral instrumentation, a surgical or interventional radiologic procedure, and a fistulous connection with a hollow viscus.270 At present, nephrectomy remains the primary therapeutic interventional modality for EPN, although CT-guided drainage appears to have

good results in patients in whom surgery is contraindicated or who have a solitary kidney, bilateral disease, or disease with renal or perirenal fluid collections.113,266,358 Pyonephrosis.  Pyonephrosis is the accumulation of pus in an obstructed pelvicalyceal system. It may appear when urinary tract infection occurs in patients with ureteric obstruction due to calculi, strictures, developmental ureteric anomalies, or malignant disease.231 Pyonephrosis is often associated with severe urosepsis. It represents a true urologic emergency requiring either urgent percutaneous nephrostomy or passage of a ureteral stent to bypass the obstruction (Fig. 54-138). The preferred examination is contrast-enhanced CT. On CT, patients with pyonephrosis may have increased pelvic wall thickness, inflammatory changes in the perinephric fat, and in rare cases layering of intravenously injected contrast material anterior to the pus in the dilated renal pelvis.95,354 Pelvicalyceal system gas in the absence of a history of urinary tract instrumentation, although uncommon, is a strong diagnostic indicator of pyonephrosis (Fig. 54-139).95 However, CT usually cannot reliably distinguish between infected and uninfected hydronephrosis (see Fig. 54-138). The presence of clinical signs of both infection and hydronephrosis is a more sensitive indicator than CT

CHAPTER 54  Kidney

1835

findings of pyonephrosis. CT may show the site and cause of urinary tract obstruction (see Fig. 54-138).231

Chronic Renal Infections Xanthogranulomatous pyelonephritis.  Xanthogranulomatous pyelonephritis is a chronic renal inflammatory disease associated with indolent bacterial infection. The inflammatory process begins in the renal pelvis and later extends into the medulla and cortex, which are gradually destroyed and replaced by lipid-laden macrophages (xanthoma cells).107 The disease is usually diffuse but may be focal. Patients are typically middle-aged women. Systemic symptoms include malaise, fever, chills, and weight loss. Nonspecific urinary symptoms include flank pain, increased frequency of micturition, dysuria, and nocturia. E. coli or Proteus mirabilis is commonly cultured from the urine. CT is valuable in diagnosis and preoperative planning.107 Nephrectomy is usually required, particularly in the diffuse form of the disease. On CT, findings of diffuse xanthogranulomatous pyelonephritis are pathognomonic in most cases: diffuse reniform enlargement with illdefined central low attenuation, apparent cortical thinning, staghorn calculus, and unilateral decrease or (more commonly) absence of renal excretion of contrast material (Fig. 54-140).15,244 Less commonly, a small contracted kidney with replacement lipomatosis is found (Fig. 54-141).107 A staghorn calculus is present in up to 70% of cases (see

A

B FIG 54-136  Emphysematous pyelonephritis in a diabetic patient. A and B, There is extensive destruction of the right kidney, which is largely replaced by gas (white arrow). Gas is also present in the perinephric space (black arrows). A right nephrectomy was performed.

FIG 54-137  Emphysematous pyelitis in a diabetic patient. Gas is

FIG 54-139  Pyonephrosis complicating congenital ureteropelvic junc-

present in both collecting systems (arrows), but the right kidney is more significantly involved. The renal parenchyma is uninvolved, and the infection responded to antibiotic therapy.

tion obstruction in a diabetic patient. The significantly dilated right renal pelvis (arrows) shows a gas-fluid level, and gas is also present in dilated calyces (arrowheads).

A

B FIG 54-138  CT scans of pyonephrosis and severe urosepsis due to an obstructing left ureteric urate calculus not visible on plain films. A, There is decreased enhancement of the left renal parenchyma, and the left renal pelvis is dilated with delayed contrast excretion (arrow). B, An obstructing calculus (arrow) is present in the proximal ureter. Differentiation between simple hydronephrosis and pyonephrosis could not be made by CT. However, the diagnosis of pyonephrosis was suggested by clinical findings. The patient showed rapid response to percutaneous nephrostomy.

1836

PART II  CT and MR Imaging of the Whole Body

Fig. 54-141).324 Multiple fluid-density rounded areas almost completely replace the renal parenchyma (see Fig. 54-140). Pathologically these areas represent either dilated calyces or focal areas of parenchymal destruction filled with pus or debris. Histologically these cavities are lined by xanthoma cells.107 Rim enhancement of the low-density areas often occurs. Extrarenal extension of the inflammatory process is common and may involve the perinephric space, pararenal spaces, ipsilateral psoas muscle, flank muscles, diaphragm, and skin (see Fig. 54-141).107 Perinephric and psoas muscle abscesses often occur. In the focal form of xanthogranulomatous pyelonephritis, CT shows a poorly enhancing mass adjacent to a calyx or in one pole of a kidney. Associated calculi may be found. The focal form may be misdiagnosed as a renal neoplasm.151 Malacoplakia.  Malacoplakia is a rare inflammatory disease usually associated with chronic E. coli infection. The urinary collecting system (especially the bladder) is most frequently involved, but the renal pelvis, ureter, kidney, and other organs outside the genitourinary tract may be affected. Affected patients are usually middle-aged women with recurrent urinary tract infections.34 Malacoplakia is attributed to abnormal macrophage response. The histologic hallmark of the disorder is the presence of basophilic inclusions called Michaelis-Gutmann bodies in large eosinophilic macrophages.76 Renal parenchymal malacoplakia is usually unilateral but can be bilateral. Contrast-enhanced CT demonstrates multiple poorly defined hypodense cortical masses

and associated decreased excretion of the contrast material in cases with extensive parenchymal replacement.116,330 Nephromegaly and perinephric extension of disease may occur.151 CT findings are nonspecific, and renal parenchymal malacoplakia may be indistinguishable from a neoplasm (Fig. 54-142). Tuberculosis.  Renal tuberculosis usually results from hematogenous dissemination of a pulmonary infection. The diagnosis is usually best made on excretory urography, which shows typical calyceal and ureteric abnormalities. CT may show a variety of abnormalities, many of which are nonspecific. Obstruction of a single major calyx or of a group of minor calyces often occurs because of scarring. The obstructed calyces are dilated and show no excretion of contrast medium, and thinning of the overlying renal parenchyma occurs.151 Tuberculosis of the renal pelvis is manifested as either hydronephrosis secondary to ureteropelvic junction obstruction or diffuse pelvic contraction with fibrosis extending down a thickened ureter. The various patterns of hydronephrosis seen at CT depend on the site of the stricture and include focal caliectasis, caliectasis without pelvic dilation, and generalized hydronephrosis.343 Low-density parenchymal lesions, probably representing areas of caseous necrosis, may be seen.151 The calcifications in the renal parenchyma may be amorphous, granular, curvilinear, or lobar (“putty kidney”; Fig. 54-143).114 In a tuberculous putty kidney, large, round or oval, dense renal calcifications representing calcified caseating material in medullary cavities and dilated calyces are

FIG 54-140  Xanthogranulomatous pyelonephritis. The right kidney is globally enlarged, shows no contrast excretion, and is replaced by multiple rounded fluid-density areas. A calculus (arrow) is present in a nondilated renal pelvis with a thickened wall. (Courtesy David S. Hartman, MD, Hershey, Pa.)

A

FIG 54-142  Renal parenchymal malacoplakia. CT scan shows a nonspecific soft tissue mass (black arrow) extending from the left kidney and invading the perinephric space and the left flank (white arrow). (Courtesy David S. Hartman, MD, Hershey, Pa.)

B FIG 54-141  Xanthogranulomatous pyelonephritis with a contracted kidney. A, A staghorn calculus (arrow) is present in the right renal pelvis. The right kidney is small and replaced by oval fluid-density areas. B, A more cephalic CT scan shows extensive chronic inflammatory change (arrows) in the posterior pararenal space.

CHAPTER 54  Kidney

1837

FIG 54-143  Renal tuberculosis with diffuse parenchymal calcification involving most of the right kidney (arrow) as shown on a nonenhanced CT scan. (Courtesy David S. Hartman, MD, Hershey, Pa.)

FIG 54-145  Renal hydatid disease. Contrast-enhanced CT scan shows a multilocular multiseptate (arrows) cystic lesion in the right kidney. (From Vourganti S, et al: Ultrasonographic evaluation of renal infections. Radiol Clin North Am 44:763–775, 2006.)

FIG 54-144  Renal tuberculosis. The left kidney shows large, dense, oval calcifications (arrow) representing calcified caseating material in medullary cavities and dilated calyces (tuberculous autonephrectomy). Low-density areas in the right kidney probably represent foci of caseous necrosis.

found. This condition, also called tuberculous autonephrectomy, is due to the superimposition of ureteral obstruction on renal parenchymal tuberculosis (Fig. 54-144). Hydatid disease.  Hydatid disease is an endemic disease caused by the larval form of Echinococcus granulosus and affects almost every organ. Renal hydatidosis is extremely rare and is seen in less than 5% of patients with hydatid disease.10,11 Hydaturia is a pathognomonic sign of renal hydatidosis and results from the rupture of a hydatid cyst into the collecting system.338 Imaging findings in hydatid disease are characteristic. CT shows a unilocular or multilocular cyst with or without peripheral calcifications (Fig. 54-145). It can also determine the cyst’s communication with the urinary tract and its extrarenal extension.

RENAL TRAUMA Renal trauma may be caused by blunt or penetrating abdominal injuries. Blunt trauma is responsible for most renal injuries. Such injuries are usually mild and heal without specific therapy.202 Serious renal injury is often associated with damage to other structures, such as the liver, spleen, bowel, pancreas, or chest. Multiorgan involvement occurs in about 80% of patients with penetrating injuries and in about 75% of those with blunt trauma.79,275

Imaging of Renal Trauma Gross hematuria is the most reliable indication of potentially serious renal damage,146 but its absence does not preclude significant renal injury. Absence of hematuria has been reported in up to 24% of patients with renal artery thrombosis and in one third of cases of

FIG 54-146  Deep renal laceration (arrowheads) extending into the renal medulla without involvement of the collecting system. A small perinephric hematoma (arrow) is present.

ureteropelvic junction injury.29,306 Contrast-enhanced CT is the preferred modality for evaluation of patients with blunt or penetrating abdominal trauma, with scans obtained at 70 seconds and 3 minutes after the start of injection of contrast material.146 The three basic types of renal injury demonstrable by CT are contusions, lacerations, and infarcts, any of which may be further complicated by intrarenal or extrarenal hematomas or by urinary extravasation. The mildest form of renal injury is the contusion, characterized by an amorphous, interstitial extravasation of blood and edema. On nonenhanced CT scans the affected kidney zones may show focal swelling and irregular infiltrates of high-density fresh blood. On contrastenhanced CT, renal contusions appear as ill-defined round or ovoid areas of hypoattenuation. They may also appear as small interstitial accumulations of contrast material staining on delayed images (striated nephrogram).86,146,246 Superficial lacerations are limited to the renal cortex; deep lacerations extend into the medulla (Fig. 54-146), where they may enter the collecting system or transect the kidney.246 Contrast extravasation is often seen in the perinephric space (Fig. 54-147). In patients with multiple lacerations (shattered kidney), the fragments are separated and surrounded by blood clot (Fig. 54-148).246 Thrombosis or laceration of a segmental branch of the renal artery produces a focal area of renal infarction. Infarcts typically appear as peripherally based wedge-shaped areas of parenchyma that fail to

1838

PART II  CT and MR Imaging of the Whole Body

FIG 54-147  Category II renal trauma. A renal laceration (not shown) involved the collecting system, resulting in contrast extravasation into the right perinephric space (arrow). The lesion responded to conservative management.

FIG 54-149  Focal renal infarction after blunt trauma. The nephrogram

FIG 54-148  Category III renal trauma—shattered kidney. There are

FIG 54-150  Category III renal trauma—occlusion of the right main renal

multiple right renal lacerations, with the fragments separated by blood clot. An associated perinephric hematoma (arrow) is present. A right nephrectomy was performed.

enhance during both the corticomedullary and pyelographic phases of CT (Fig. 54-149).115 Posttraumatic bleeding is commonly associated with all injuries to the kidney. Hematomas may be intrarenal or subcapsular or may involve the perinephric or pararenal spaces. The most significant vascular injury after blunt trauma is thrombosis of the main renal artery.147 Main renal artery injuries can be diagnosed on CT based on the criterion of absence of renal contrast enhancement with or without cortical rim sign (Fig. 54-150).226 Other findings that suggest the diagnosis of pedicle injury include a hematoma surrounding the renal hilus (see Fig. 54-50) and abrupt cutoff of the contrast-filled renal artery.86 Acute renal vein occlusion may be suspected if the kidney is enlarged and shows thrombus in the renal vein. Venography may be performed if injuries to the renal vein and IVC are suspected, because CT may not reliably detect venous laceration.146 Ureteropelvic junction injuries are rare. In these patients, CT typically demonstrates excellent excretion of contrast material with an intact intrarenal collecting system, but with medial perinephric urinary extravasation rather than the lateral urinary extravasation in category II injuries with involvement of the collecting system. A circumrenal urinoma may be seen around the affected kidney, but typically there is no perinephric hematoma.146 Abnormal kidneys, including hydronephrotic kidneys (Fig. 54-151), ectopic kidneys, and horseshoe kidneys (Fig. 54-152), as well as kidneys containing neoplasms or cysts, may sustain injury second-

is absent anteriorly (arrow) in the right kidney. “Rim” cortical perfusion of the infarct is present.

artery. The right kidney shows absence of contrast enhancement except for a subcapsular rim of renal tissue perfused by capsular collateral vessels. Aortography showed occlusion of the main right renal artery. A hematoma (arrow) is present around the renal vascular pedicle.

FIG 54-151  Left perinephric urinoma (arrows) secondary to blunt trauma to a hydronephrotic left kidney. Hydronephrosis was due to congenital ureteropelvic junction obstruction.

ary to minor trauma.261 Gd-enhanced MRI may be used to assess suspected renal injury when the use of iodinated contrast material is contraindicated.170

Imaging Classification and Management Renal injuries can be classified into five grades according to the  American Association for the Surgery of Trauma, as summarized in Table 54-4.169

CHAPTER 54  Kidney

A

1839

FIG 54-153  Category I renal trauma. There is a small cortical laceration (arrowhead) involving the left kidney. An associated small perinephric hematoma is present (arrow).

TABLE 54-4  American Association for

the Surgery of Trauma Grading System for Renal Injury Grade*

Type of Injury

Description of Injury

I

Contusion

Microscopic or gross hematuria; urologic studies negative Subcapsular, nonexpanding without parenchymal laceration Nonexpanding perirenal hematoma confined to renal retroperitoneum <1 cm parenchymal depth of renal cortex without urinary extravasation >1 cm parenchymal depth of renal cortex without collecting system rupture or urinary extravasation Parenchymal laceration extending through renal cortex, medulla, and collecting system Main renal artery or vein injury with contained hemorrhage Completely shattered kidney Avulsion of renal hilum, which devascularizes kidney

Hematoma II

B

Laceration

FIG 54-152  Injury of horseshoe kidneys. A, Bilateral perinephric hematomas (arrows) are present. B, Caudal CT scan shows a hematoma (white arrow) involving the isthmus connecting the two kidneys. There is contrast extravasation (black arrow) from the left collecting system because of a laceration.

Hematoma

III

Laceration

IV

Laceration

Vascular

Categorization of renal injuries according to severity provides a helpful guide to patient management. Renal injuries are classified into four categories on the basis of imaging findings.86 Category I lesions (75%-85% of cases) are clinically insignificant; they consist of contusions and small corticomedullary lacerations that do not communicate with the collecting system (Fig. 54-153). They are managed nonoperatively.79,86 Category II lesions (10% of cases) are more serious and comprise major lacerations through the renal cortex extending to the medulla or collecting systems with or without urinary extravasation (see Fig. 54-147). Many of these lesions are managed nonoperatively. However, surgery is sometimes required in patients with extensive hemorrhage or urinary extravasation and a large amount of nonviable renal tissue.146 Category III lesions (5% of cases), which are catastrophic, consist of shattered kidney (multiple deep lacerations; see Fig. 54-148) and injury to the renal pedicle (see Figs. 54-146, 54-148, and 54-150). The rare entities of ureteropelvic junction avulsion and laceration of the renal pelvis are designated as category IV lesions. Category III and IV renal lesions usually require surgical intervention. Subcapsular and perinephric hematomas may occur in any of the categories of renal trauma, but when they occur in isolation, they are assigned to category I. Although an image-based classification of renal trauma is useful in guiding management, treatment is often ultimately determined by the patient’s clinical status.

V

Laceration Vascular

From Lee YJ, et al: Renal trauma. Radiol Clin North Am 45:581–592, 2007, with permission. *Advance one grade for bilateral injuries up to grade III.

RENAL BLOOD FLOW DISORDERS Renal Hemorrhage CT is the most valuable examination in the evaluation of patients with suspected acute renal hemorrhage because it accurately diagnoses the presence and location of such hemorrhage and often shows the underlying cause.20,283 Renal hemorrhage may be suburothelial, intraparenchymal, subcapsular, perinephric, or pararenal in location or may involve the renal sinus.68,94,162,232

Causes.  The most common cause of renal hemorrhage is trauma, either blunt or penetrating (Fig. 54-154; see Fig. 54-152). Extracorporeal shock-wave lithotripsy for nephrolithiasis is another form of renal trauma not infrequently associated with parenchymal and perinephric hemorrhage (Fig. 54-155).232 Spontaneous (nontraumatic) renal

1840

PART II  CT and MR Imaging of the Whole Body

hemorrhage may be caused by anticoagulation, blood dyscrasias (Fig. 54-156), renal infarction, polyarteritis nodosa (Fig. 54-157), renal aneurysms and arteriovenous malformations, RCC, AML (see Fig. 54-117), renal abscess, renal vein thrombosis, and rupture of hemorrhagic solitary cysts (Fig. 54-158) or of hemorrhagic cysts in renal cystic disease (see Figs. 54-38 and 54-51).20,179,181 Some cases are idiopathic.20 RCC is probably the most common cause of spontaneous subcapsular and perinephric hemorrhage.

CT Findings.  Recent renal hemorrhage is characterized by highattenuation blood (50-90 HU), which is best shown by nonenhanced CT scans (see Fig. 54-155). Postcontrast scans should also be obtained to facilitate identification of disorders such as small neoplasms causing spontaneous renal hemorrhage. Suburothelial hemorrhage is characterized on CT by thickening of the wall of the renal pelvis and upper ureter by blood that has a high attenuation value on nonenhanced scans.162 Spontaneous hemorrhage into the renal sinus is characterized by a high-density blood collection in the renal sinus, with displacement of the renal pelvis (Fig. 54-159).94 Renal sinus hematomas usually

A

B FIG 54-155  Left perinephric hematoma secondary to extracorporeal

FIG 54-154  Fat-suppressed T1-weighted MRI (TR = 600 msec, TE = 17 msec) shows a high-intensity left perinephric hematoma (arrow) due to a renal biopsy performed 2 weeks earlier. Loss of corticomedullary differentiation is due to chronic renal disease.

A

shockwave lithotripsy. A, Nonenhanced CT scan reveals anterior displacement of left kidney (black arrowheads) by high-density hematoma (arrow). There is a calculus (open arrow) in the collecting system. The hematoma is limited posteriorly by the posterior renorenal bridging septum, which separates it from the thickened posterior renal fascia (white arrowheads). B, The hematoma (white arrow) extends into the infrarenal fascial cone (arrowheads). Infrarenal extension confirms the perinephric location of the hematoma and distinguishes it from a subcapsular hematoma. The left ureter (curved arrow) contains a stent and is displaced anteriorly.

B FIG 54-156  Right subcapsular and perinephric hemorrhage associated with idiopathic thrombocytopenia. A, Nonenhanced CT scan shows indentation of the lateral border (arrowheads) of the right kidney by a denser subcapsular hematoma. A smaller perinephric hematoma (arrows) is also present. B, Contrast-enhanced CT scan shows that the hematoma is less dense than the enhanced renal parenchyma. (From Levine E, et al: Spontaneous subcapsular and perinephric hemorrhage in end-stage kidney disease: Clinical and CT findings. AJR Am J Roentgenol 148:755–758, 1987.)

CHAPTER 54  Kidney

A

1841

B

FIG 54-157  Multiple renal infarcts and perinephric hemorrhage complicating polyarteritis nodosa. A, A left perinephric hematoma (arrow) is contained by the posterior renorenal bridging septum. There are several wedge-shaped infarcts (arrowheads) involving the renal parenchyma. B, A left selective renal arteriogram shows multiple small aneurysms arising from interlobar arteries. Active contrast extravasation (arrows) in the lower pole indicates bleeding from a ruptured aneurysm. The patient showed response to selective embolization of the feeding artery. (Courtesy James Bergh, MD, Overland Park, Kan.)

FIG 54-158  Right perinephric hematoma (arrow) probably due to rupture of a simple renal cyst. The cyst (arrowheads) is flattened by the hematoma.

FIG 54-159  Acute hemorrhage (arrow) in the left renal sinus associated with anticoagulant therapy displaces the renal pelvis.

resolve spontaneously, as shown by follow-up CT. Both suburothelial and renal sinus hemorrhage occur most commonly in patients  undergoing anticoagulation therapy or with bleeding or coagulation disorders.94,162 On nonenhanced CT scan, a recent subcapsular hematoma is characterized by a mass with a higher attenuation value than that of adjacent renal parenchyma (see Fig. 54-156). Pressure on the underlying renal parenchyma characteristically causes flattening of the kidney, elevation of the renal capsule, and medial displacement of the collecting system. Sometimes subcapsular hematomas fail to resolve, causing chronic fluid collections that compress the renal parenchyma. This in turn may cause hypertension by reducing renal blood flow, thereby triggering the renin-angiotensin-aldosterone system. This clinical entity is called the Page kidney.146 The walls of chronic subcapsular hematomas may eventually calcify (Fig. 54-160). Perinephric hematomas involve the fat between the renal capsule and renal fascia. The configuration of such hematomas is determined largely by the bridging septa in the perinephric fat.163 When a hematoma is confined by the posterior renorenal bridging septum, it may compress and indent the renal surface, and because it is separated from the renal fascia, it may simulate a subcapsular hematoma (see Figs. 54-155 and 54-157).163 However, although subcapsular hematomas are confined to the kidney by the renal capsule, perinephric hematomas often extend caudad below the kidney into the cone of renal fascia (see Fig. 54-155). CT often shows the cause of spontaneous subcapsular and perinephric hematomas.283 RCCs or AMLs associated with such hematomas may be suspected on CT from the presence of a distinct mass clearly different in character from the surrounding hemorrhage.20 RCCs present as heterogeneous enhancing solid masses or as cystic masses, whereas an AML is suggested by the presence of tumor fat (see Fig. 54-117).20 CT may show that hemorrhage originated in either a solitary renal cyst (see Fig. 54-158)68 or a cyst associated with hereditary or acquired renal cystic disease (see Figs. 54-38 and 54-51).179,181 If CT shows no cause for hemorrhage, renal angiography may help with diagnosis by demonstrating underlying disorders such as angiitis, renal aneurysms, arteriovenous malformations, or polyarteritis nodosa (see Fig. 54-157).20

1842

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 54-160  Chronic calcified subcapsular hematoma in a hypertensive man. The low-density (17 HU) posterolateral mass with peripheral calcification (arrows) causes anteromedial displacement of the right kidney, as seen on nonenhanced (A) and contrast-enhanced (B) CT scans.

Treatment.  CT is valuable in the follow-up and management of renal hemorrhage, whether spontaneous or due to trauma. RCCs complicated by hemorrhage are managed with nephrectomy, whereas underlying hemorrhagic renal cysts, renal infarcts, or AMLs can usually be managed conservatively.20,181 When the bleeding is extensive and the patient is unstable, renal angiography with embolization (see Fig. 54-157) or occasionally surgery may be necessary to stop the bleeding.32,181 Patients with spontaneous renal hemorrhage who have no demonstrable cause for the bleeding or for whom the diagnosis of the cause of the hemorrhage is in doubt should also receive conservative management. To exclude an underlying RCC, however, serial CT should be used to monitor such patients until the hematoma completely resolves.20 This approach avoids unnecessary nephrectomy in patients who have spontaneous renal hemorrhage with benign renal disease or no underlying disease.20,32

Renal Infarction Renal infarction may be due to renal artery thrombosis or embolism, vasculitis as in polyarteritis nodosa, trauma, sickle cell disease, or aortic dissection. The most common cause is thromboembolism from cardiovascular disease.147 CT findings depend on both the extent and age of the infarction. The size and shape of a renal infarct are determined by the size of the occluded artery. If the main renal artery is occluded, global infarction of the kidney ensues. On contrast-enhanced CT scans, the affected kidney shows lack of enhancement, apart from a high-density cortical rim reflecting perfusion of the preserved outer rim of the cortex by collateral vessels (see Fig. 54-150).140 The renal collateral circulation is supplied by renal capsular vessels, peripelvic vessels, and periureteral vessels. If a major renal artery branch is occluded, an acute focal infarct results, which manifests as a wedge-shaped low-attenuation renal parenchymal lesion on a contrast-enhanced CT scan. The base of the wedge is contiguous with the renal capsule, and the apex is directed toward the renal hilus.115 There is usually sharp margination between infarcted tissue and the adjacent normal nephrogram (Fig. 54-161). A rim sign is often seen on the capsular margin of the infarct. Emboli and vasculitis cause multiple, often bilateral, focal renal infarcts (Fig. 54-162).351 Unilateral infarcts are often the result of renal trauma, which is also a common cause of global renal infarction.351 When smaller intrarenal arteries are occluded, the CT findings are less specific

FIG 54-161  Focal renal infarct associated with emboli from a prosthetic mitral valve. The lesion (arrow) manifests as a wedge-shaped zone of low attenuation. The base of the wedge is contiguous with the renal capsule, and the apex is directed toward the renal hilus. There is a straight sharp margin between the infarct and normal renal tissue. A rim of perfused tissue is seen on the capsular margin of the infarct.

and consist of multiple low-attenuation defects that are scattered throughout the nephrogram, although usually peripherally located. Vascular thrombosis in sickle cell disease causes multiple foci of “slitlike” areas of low attenuation.351 Both acute focal and global infarcts may be associated with subcapsular or perinephric blood or fluid collections and thickening of the renal fascia (see Fig. 54-162).351 Old infarcts manifest on CT as renal cortical scars or as a small, shrunken kidney (Fig. 54-163).147 The main CT differential diagnosis of acute renal infarction is  acute pyelonephritis; both conditions (1) often demonstrate wedgeshaped low-attenuation renal lesions on CT and (2) often manifest  as acute onset of flank pain and fever. A cortical rim sign should strongly suggest the diagnosis of renal infarction (see Fig. 54-162) because it is usually not seen in acute pyelonephritis.351 Small renal infarcts may also be confused with focal lymphomatous lesions or metastases on CT.

CHAPTER 54  Kidney

1843

A FIG 54-162  Left renal infarcts (arrowheads) associated with fibrodysplastic occlusions of branches of the renal artery. Both lesions show subcapsular rims of perfused tissue, and there is associated thickening of the anterior and posterior renal fascia (arrows).

B FIG 54-164  Right renal artery stenosis. A, Dynamic CT scan after a FIG 54-163  Bilateral renal scarring due to old infarcts in a hypertensive patient.

Renal Artery Stenosis Secondary hypertension due to diseases of the renal arteries (i.e., renovascular hypertension) accounts for about 1% to 5% of all cases of hypertension.212 Atherosclerosis and fibrodysplastic disease are the most common causes of significant renal artery narrowing. Although there are several tests for evaluating patients with suspected renovascular hypertension, the standard for the diagnosis of renal artery stenosis is angiography, with response to treatment the proof of its significance.257 CT angiography (CTA) and MRA provide noninvasive means to obtain anatomic information. CTA and Gd-enhanced 3D MRA seem to be preferred in patients referred for evaluation of renovascular hypertension.336

CT.  The renal arteries are not always well shown by conventional CT. However, significant renal artery disease may be suspected from conventional dynamic CT scans if the affected kidney shows prolonged corticomedullary differentiation compared with the normal side (Fig. 54-164). Global loss of renal parenchyma (if present) may be ancillary evidence of chronic arterial stenosis or occlusion (see Fig. 54-164).23 Helical (spiral) CT and now MDCT(multislice) with subsecond speeds allow volumetric acquisitions to be obtained during a single breath hold. Multislice CT systems allow simultaneous acquisition of multiple slices per gantry rotation. In combination with faster gantry rotation times of 0.5 second, the abdominal structures can be displayed in higher spatial and temporal resolution. The advantage of this

rapid IV injection of contrast agent shows prolonged corticomedullary differentiation in the right kidney compared with the left kidney, for which the nephrogram is almost homogeneously dense. The right kidney also shows global atrophy. B, A contiguous caudal image reveals a calcified atherosclerotic plaque (arrow) at the origin of the right renal artery. (From Birnbaum BA, et al: Asymmetry of the renal nephrogram on CT: Significance of the unilateral prolonged cortical nephrogram. Urol Radiol 12:173–177, 1991.)

technique is that it eliminates the respiratory misregistration that often degrades image quality.161 Sensitivity and specificity for significant renal artery stenosis with single-slice helical CT are 88% to 97% and 83% to 89%, respectively, compared with catheter angiography. Multislice CTA exceeds MRA in spatial resolution and is now able to display even small vascular side branches.254 Compared with catheter angiography, CTA has a sensitivity of 95% to 100% and a specificity of 99% to 100% for detection of accessory renal arteries (Fig. 54-165).91 Renal vascularity is evaluated during the vascular phase, which occurs approximately 15 to 25 seconds after the start of IV contrast material adminstration.147 Typical acquisition parameters for CTA of the renal artery include a collimation of 3 mm, a pitch of 1.5 to 2.0, a reconstruction interval of 1 to 2 mm, a scan delay of 25 seconds after administration of contrast material, and an injection rate of 3 mL/sec or greater.250,273,272 Images should be reconstructed with 50% overlap to improve detection of renal artery stenosis and 3D reconstruction. Commonly used 3D visualization techniques include surface rendering, MIP, and volume rendering. Volume rendering is superior to MIP and surface rendering because it uses the volume data and can provide lifelike images, including color. This permits MPR in any plane

1844

PART II  CT and MR Imaging of the Whole Body

IVC Aorta

Posterior view

FIG 54-165  3D MPR, posterior view. There is slight narrowing of the right renal artery in its midcourse, and both renal arteries are of normal caliber at their origin (arrows). IVC, inferior vena cava.

(Fig. 54-166). This technique is very useful in pretransplantation evaluation of renal donors and in presurgical planning.

MRA.  MRA with Gd enhancement has a sensitivity of 100% and a specificity in the range of 93% to 97% for detection of proximal renal artery stenosis (Fig. 54-167).71,253,307 In the past, both time-of-flight imaging and phase-contrast imaging were used; both methods have been virtually replaced by breath-hold 3D Gd-enhanced MRA. These sequences prevent image degradation from respiratory motion, reduce acquisition time, and overcome the problem of saturation of in-plane blood flow.279 The dose of Gd is usually 0.2 mmol/kg. Typical injection rates are 2 to 3 mL/sec. Commonly, four different degrees of stenosis are assessed, based on the reduction of vessel diameter: normal (0%), mild (<50%), moderate (50%-75%), or severe (75%; Fig. 54-168).279 In about 20% to 30% of patients, the kidney is supplied by more than one renal artery (Fig. 54-169).279 3D Gd-enhanced MRA allows an accurate diagnosis of proximal renal artery stenosis without the risks associated with nephrotoxic contrast agents, ionizing radiation, or arterial catheterization. Following is a general outline of renal artery stenosis protocol. A torso coil centered at midabdomen is used, with peripheral gating and respiratory bellows. An 18-gauge IV line is placed in the antecubital vein. The following sequences are obtained: 1. Breath-hold coronal large field of view of the upper abdomen 2. Non–breath-hold sagittal T1-weighted spin echo image over the aorta (6-8 slices), about 4 to 6 mm thick, giving a good target for SmartPrep tracker placement

FIG 54-166  3D MPR, posterior view, with overlying 12th rib. Arrows point to the renal arteries.

3. Breath-hold coronal SSFSE of the kidneys (for cysts) prescribe off series 2 4. Precontrast coronal 3D Gd-enhanced fast gradient-recalled echo sequence through the territory of interest (trial run), 3-mm thick (no gap because it is 3D). Cover from the celiac axis to groin. Keep scan time under 25 to 30 seconds for renal arteries. Maximize spatial resolution, increasing phase-encoding steps (can be 224 or 256 for some patients), using partial number of excitations. 5. Repeat step 4: coronal 3D Gd-enhanced fast gradient-recalled echo with tracker placed just above the origin of renal arteries, with the patient’s arms positioned up above the head to avoid the wraparound artifact. Run this sequence twice successively, allowing the patient to take three deep breaths before each run. 6. Optional: axial flow gradient from origin of the celiac axis to the iliac arteries in an elderly patient if an aneurysm is suspected; this improves visualization of the outer wall of the aneurysm.

Acute Cortical Necrosis Acute cortical necrosis, a rare form of acute renal failure, results from ischemic necrosis of the renal cortex with sparing of the medulla.147 The most common cause is hemorrhage in the third trimester of pregnancy, most often associated with placental abruption. Other causes are septic abortion, severe trauma with shock, transfusion reaction, severe dehydration, certain toxins including snake venom, the hemolytic-uremic syndrome, and acute aortic dissection.136 Some of these conditions appear to act by producing renal cortical vasoconstriction, causing necrosis of the renal cortex.65 The condition is usually bilateral but may occasionally be unilateral.136 Contrast-enhanced CT shows a distinctive nephrographic pattern characterized by a zone of nonenhanced cortex between a rim of enhancing subcapsular cortex and the enhancing juxtamedullary

CHAPTER 54  Kidney

1845

Right renal artery

A

B FIG 54-167  A and B, Gd-enhanced MRA shows stenosis of proximal right renal artery.

K

FIG 54-168  Gd-enhanced MRA shows severe stenosis of right renal artery at its origin (arrow). K, kidney.

cortex and medulla (Fig. 54-170).136,208 Global atrophy of the kidneys occurs within a few months, leading to small smooth kidneys. A dense linear “tramline” calcification can form in the renal cortex.80

Renal Artery Aneurysm and Arteriovenous Communications Renal artery aneurysms may be saccular, fusiform, or intrarenal; their most common cause is atherosclerosis.195 Other causes include medial

FIG 54-169  Gd-enhanced MRA (3D Gd-enhanced fast gradient-recalled echo) composite image shows two left renal arteries (arrows) and a single right renal artery (arrowhead).

fibrodysplasia, pregnancy, neurofibromatosis, and Ehlers-Danlos syndrome. Intrarenal aneurysms are usually associated with polyarteritis nodosa and Wegener’s granulomatosis.147 Most aneurysms are extrarenal (Fig. 54-171). Aneurysms of the main renal artery or its major branches may show rimlike calcification (Fig. 54-172) and significant enhancement on CT scans. Renal infarcts secondary to emboli from

1846

PART II  CT and MR Imaging of the Whole Body

mural thrombus in the aneurysm may be seen on contrast-enhanced CT.147 Demonstration of small intrarenal aneurysms requires selective renal arteriography. Renal arteriovenous communications that are developmental  are called arteriovenous malformations. Arteriovenous fistulas (AVFs) account for 70% to 80% of arteriovenous communications in the kidney.56 Penetrating trauma is the most common cause of AVF; other causes include surgery, tumors, and inflammation; AVFs may also be idiopathic.147 Small intrarenal AVFs or arteriovenous malformations are best diagnosed with selective renal arteriography. However, larger lesions may be calcified and may occur in the renal sinus, where they may be confused with other renal masses (Fig. 54-173). These vascular lesions are best assessed on contrast-enhanced CT. A nonthrombosed lesion enhances to the same extent as the adjacent aorta. A thrombosed lesion shows no contrast enhancement (see Fig. 54-173). Visualization

of an enlarged feeding renal artery and draining vein confirms the nature of an AVF (see Fig. 54-173). Nutcracker syndrome is a condition arising secondary to compression of the left renal vein between the aorta and superior mesenteric artery, with resultant hematuria, varicocele, and hypertension.155 This compression results in an enlarged left renal vein and may sometimes result in formation of collateral veins that can mimic an arteriovenous malformation. It has been seen in a relatively younger population, with patients presenting with gross or microscopic hematuria. Hematuria occurs secondary to left renal vein hypertension as a result of the compression, which may then result in rupture of thin-walled veins into the collecting system. Nutcracker “syndrome” (compression of left renal vein with hematuria) must be differentiated from nutcracker “phenomenon,” which is an asymptomatic compression of the left renal vein.187 Nutcracker phenomenon or syndrome can be diagnosed on a contrast-enhanced study such as CT or MRI. Diagnosis is usually made on the basis of an asymmetric enlargement of the left renal vein, with associated compression of the left renal vein between the aorta and superior mesenteric artery (Fig. 54-174). Occasionally collateral venous formation may be seen in the left renal hilum. The syndrome can also lead to formation of scrotal or vulvar varices.

FIG 54-170  Bilateral acute renal cortical necrosis complicating aortic dissection (arrow). There is no enhancement of the renal cortex apart from a thin subcapsular rim perfused by capsular collateral vessels. Enhancement of the renal medulla is preserved. (From Badiola-Varela CM: Acute renal cortical necrosis: Contrast-enhanced CT and pathologic correlation. Urol Radiol 14:159–160, 1992.)

A

FIG 54-172  Bilateral atherosclerotic renal artery aneurysms (arrows) with peripheral calcification.

B FIG 54-171  Left renal aneurysm. A, Contrast-enhanced axial CT shows an extrarenal large aneurysm with thrombus (arrow) and focal calcification in its wall posteriorly. The embolus from the aneurysm has resulted in segmental infarct (arrowhead). B, Selective angiogram of the left kidney confirms the extrarenal aneurysm of the renal artery.

CHAPTER 54  Kidney

1847

v a

B

A

C

D FIG 54-173  Bilateral renal artery pseudoaneurysms in a patient with a history of previous trauma. A, Contrast-enhanced CT scan shows an enhancing mass with peripheral calcification (arrow) in the lower part of the right renal hilus. There is also a nonenhancing mass with peripheral calcification (curved arrow) in the lower part of the left renal hilus. B, Cephalic scan reveals significant enlargement of the right renal artery (a) and vein (v), indicating an arteriovenous fistula. The left renal vein (arrow) is normal in size. C, Right selective renal arteriogram reveals enlargement of the right renal artery (arrow) and vein (curved arrow). The calcified pseudoaneurysm is opacified (open arrow). D, The left renal artery pseudoaneurysm (arrows) does not opacify because of thrombosis.

Renal Vein Occlusion

FIG 54-174  Nutcracker phenomenon. Contrast-enhanced CT scan shows decreased distance between the aorta and superior mesenteric artery, with compression of the left renal vein (arrow) resulting in dilatation of the left renal vein (arrowhead). Incidentally seen is right ureteropelvic junction obstruction.

Renal vein occlusions may be categorized into the following five groups: 1. Extrinsic occlusion of the renal vein by an adjacent neoplasm (e.g., pancreatic carcinoma or lymphoma)188 2. Direct renal vein extension of RCC or adrenal neoplasms134 3. Renal vein thrombosis associated with primary renal disease. About 20% of patients with the nephrotic syndrome resulting from causes such as membranous glomerulonephritis, lupus erythematosus, and amyloidosis have renal vein thrombosis, which may be acute or chronic.353 In adults the most common underlying abnormality is membranous glomerulonephritis.147 4. Secondary renal vein occlusion or thrombosis may occur when the IVC is thrombosed after caval extension of thrombus from pelvic or leg veins. 5. Renal vein thrombosis may occur as a primary phenomenon in infants with a history of maternal diabetes, sickle cell disease, hemoconcentration states such as diarrhea and sepsis, and hypoxia in cyanotic congenital heart disease.109 Adults with hypercoagulable states may also experience renal vein thrombosis.

1848

PART II  CT and MR Imaging of the Whole Body

A

FIG 54-176  Renal vein thrombosis. Dynamic CT scan shows a lowattenuation clot (arrow) in a partially thrombosed left renal vein. There is prolonged corticomedullary differentiation in the left kidney compared with the right kidney, which exhibits a homogeneous nephrogram. (From Birnbaum BA, et al: Asymmetry of the renal nephrogram on CT: Significance of the unilateral prolonged cortical nephrogram. Urol Radiol 12:173–177, 1991.)

B FIG 54-175  Renal vein thrombosis associated with the nephrotic syndrome. A, Low-density thrombus is present in a retroaortic left renal vein (arrowheads) and in the adjacent inferior vena cava (arrow). B, Spin echo MRI (TR = 650 msec, TE = 30 msec) shows high-signal thrombus in the retroaortic left renal vein (arrow) and in the adjacent inferior vena cava (black arrowhead). The aorta (white arrowhead) shows normal blood flow.

Clinical and imaging findings in renal vein thrombosis depend on the severity of the occlusion, rapidity of onset, location of the thrombus, and cause. Left renal vein thrombosis is more common. A classic acute presentation consists of flank pain, hematuria, and loss of renal function.147 Patients with chronic renal vein occlusion are usually asymptomatic. Thrombosis of the IVC at or near the renal vein orifices occurs in about 40% to 50% of patients with renal vein thrombosis (Fig. 54-175). Contrast-enhanced CT is an excellent method for noninvasive diagnosis of renal vein thrombosis, provided renal function is normal. CT permits differentiation between acute renal vein thrombosis and conditions that have similar clinical presentations, such as acute pyelonephritis, acute renal infarction, and acute renal obstruction.103 Renal vein thrombosis is generally unilateral. In acute and subacute cases, an enlarged swollen kidney is seen on CT. The nephrogram in the affected kidney is initially diminished because of impaired renal perfusion; however, once the nephrogram develops, it persists for a prolonged period and there is often prolonged enhancement of the renal cortex relative to the renal medulla (Fig. 54-176).147 Calyceal opacification is often delayed, diminished, or absent in the affected kidney. Stranding of the perinephric fat owing to edema and thickening of the renal fascia may occur. Enlarged perirenal collateral veins are often noted. Perinephric hemorrhage may occur. On CT the renal vein is commonly enlarged and may show a filling defect because of thrombus (see Fig. 54-176).102,293 Demonstration of venous thrombus is facilitated by scanning during the peak phase of vascular opacification after bolus injection of contrast medium and by obtaining 5-mm-thick sections (see Fig. 54-176). Infarction of the affected kidney may occur, and long-term

follow-up by CT may reveal severe atrophy and poor function in the affected kidney.102 Chronic renal vein thrombosis is often a silent complication of the nephrotic syndrome.98 CT frequently shows the presence of venous thrombus (see Fig. 54-175). In the chronic phase of renal vein thrombosis, the affected renal vein becomes attenuated because of retraction of the clot, and extensive collateral vessels develop along the proximal to middle ureter and around the kidney.147 MRI is widely used for detecting renal vein extension of RCC; it can also be used in the evaluation of nonneoplastic renal vein thrombosis. MRI is most useful for evaluating patients with significant renal functional impairment and symptoms suggesting renal vein thrombosis. Renal vein thrombus may be shown on T1-weighted spin echo pulse sequences when the signal void of flowing blood in the renal vein is replaced by high signal because of thrombus (see Fig. 54-175). Gradient echo technique shows thrombus as a filling defect of medium signal intensity that replaces the high signal of flowing blood (see Fig. 54-79). Coronal MRI helps determine the extent of vena caval involvement. MRI in patients with acute renal vein thrombosis may also show loss of corticomedullary differentiation on T1-weighted spin echo images, increased signal in the affected kidney on T2-weighted images, renal fascial thickening, and renal enlargement.158

URINARY TRACT STONES Plain film radiography, excretory urography, retrograde pyelography, and US have been used in various combinations for the diagnosis of urinary tract stones in patients presenting with flank pain. CT has virtually replaced all other imaging modalities for this purpose. Nonenhanced helical CT for the evaluation of patients with flank pain was first described by Smith and associates in 1995.300 Nonenhanced helical CT is superior to other imaging modalities in diagnosis of urinary tract calculi. Also, it can diagnose other causes of flank pain such as appendicitis, diverticulitis, pancreatitis, bowel obstruction, and intussusception (Table 54-5).194,299

Helical CT Technique CT examination for suspected ureterolithiasis is performed from the top of the kidneys to the base of the bladder, without the use of IV or

CHAPTER 54  Kidney

1849

TABLE 54-5  Composition of Urinary Tract Stones Calcium Containing

Uric Acid

60%-80% Ca oxalate and phosphate Uniform density and easily detectable on NECT and plain films

5%-10% Hyperuricosuria, gout, Lesch-Nyhan syndrome Radiolucent on plain films

Magnesium Ammonium Phosphate 5%-10% Triple phosphate (Ca, Mg, NH4) Form in association with UTI in alkaline urine 70% of staghorn calculi

Cystine <5% Autosomal recessive defect in tubular reabsorption

Protease Inhibitor Stones (Ritonavir or Indinavir) Rare Associated with protease inhibitor drugs used in HIV Radiolucent on plain films and CT

NECT, nonenhanced CT; UTI, urinary tract infection.

oral contrast material. A collimation of 5 mm with a pitch of 1 is recommended for single-detector spiral CT; 2.5-mm collimation with reconstruction at 2.5 mm and a pitch of 0.8 to 1 for MDCT is preferred. A distended urinary bladder at the time of imaging better defines the ureteral insertions.299 Using a pitch of 2 results in decreased radiation exposure without much sacrifice in the quality of scan.72 The helical CT data are acquired during breath hold, thus avoiding the respiratory misregistration artifact and allowing the retrospective reconstruction to delineate small calculi. The introduction of MDCT has further reduced the scan performance time, enabling the study to be completed in about 10 to 15 seconds.

K

CT Findings.  Virtually all stones (including uric acid stones) are of sufficient radiographic attenuation to be readily visible on CT,299 except for stones associated with protease inhibitors used for treating human immunodeficiency virus infection (e.g., indinavir).27,312 In the presence of flank pain, the primary CT finding in ureteral obstruction is identification of a stone in the lumen of the ureter (Fig. 54-177).299 Even calculi of submillimeter size can be easily identified. The most common sites of impaction are the ureteropelvic junction (35%; Fig. 54-178), midureter (7%), distal ureter (33%), and ureterovesical junction (18%; Fig. 54-179). In one study, 8% of the stones were already present in the bladder at the completion of the CT examination.63 The search for ureteric calculi should extend from the renal pelvis to the bladder. There is no substitute for establishing continuity of a calcification within the ureteral lumen on sequential images.62 The soft tissue rim sign (see Fig. 54-177) can be used to identify the ureter; it represents edema of the ureteral wall at the level where a stone has become impacted.122,300 Approximately 77% of ureteral stones have the soft tissue rim sign.62 Presence of this sign is specific for the diagnosis of ureterolithiasis, although absence of the sign does not preclude such a diagnosis.145 Differentiating between a phlebolith and a ureteric calculus is a significant problem, especially in patients with minimal retroperitoneal fat. Phleboliths, which are common in the pelvis, arise from thrombi in the pelvic veins and are not often seen in the middle and upper abdomen except when present in gonadal veins. The presence of the comet tail sign—an adjacent eccentric, tapering, soft tissue mass corresponding to the noncalcified portion of a pelvic vein—helps differentiate between ureteric calculus and phlebolith.19,30 CT is poor in identifying the central lucency present in a phlebolith on plain radiographs329 and is not helpful in the differentiation of phlebolith and calculus. When present, two important secondary CT signs of obstruction— ureteral dilatation and perinephric stranding (Fig. 54-180)—are helpful in the diagnosis of ureteral obstruction.299 The presence of ureteral dilation has a positive predictive value of 90% and a negative predictive value of 89% for diagnosis of ureteral obstruction.63 The dilated urine-filled ureter, a reliable guide to the point of obstruction,

R

FIG 54-177  Noncontrast renal CT protocol. Axial scan demonstrates a left ureteric calculus with surrounding ureteric wall edema—the so-called soft tissue rim sign (arrow). Part of the left kidney (K) is seen. There is some right perinephric edema (arrowhead) because the patient had bilateral ureteric calculi (not shown).

FIG 54-178  Noncontrast renal CT protocol. Axial scan demonstrates a 3- to 4-mm calculus (arrow) in the right ureteropelvic junction, with mild hydronephrosis (arrowhead).

1850

PART II  CT and MR Imaging of the Whole Body

is followed with sequential CT images until a change of caliber is seen (see Fig. 54-126). The mean diameter of the normal ureter on CT is 1 to 2 mm. The symptomatic side is compared with the normal side; it is important not to mistake a ureter for the gonadal vein. Stranding of the perinephric fat on the symptomatic side represents fluid in the bridging septa due to increased lymphatic pressure.62 This finding has a positive predictive value of 92% and a negative predictive value of 84% for diagnosis of ureteral obstruction.301 Perinephric stranding is most commonly seen in patients with acute or chronic ureteral obstruction; however, in the absence of obstruction, stranding can be seen in pyelonephritis, renal vein thrombosis, renal infarction, trauma, and renal tumor with hemorrhage.299 Periureteral stranding is less common and is rarely seen in the absence of perinephric stranding. Collecting system dilatation is best identified in the upper pole and lower poles. Dilated calyces and infundibula appear as rounded fluidfilled structures that partially obliterate the renal sinus fat compared with the contralateral side (Fig. 54-181).62 The presence of collecting system dilatation yields a positive predictive value of 93% for diagnosis of ureteral obstruction.301 After ureteral obstruction, renal enlargement may be seen; its presence has a positive predictive value of 86% for diagnosis of ureteral obstruction.301 Unilateral absence of the “white pyramid” has been described as an additional secondary sign of urinary tract obstruction (Fig. 54-182).271

Bilateral high-attenuation renal pyramids are an occasional incidental finding. Ureteral obstruction may result in tubular hydronephrosis, decreasing the attenuation of the medullary pyramid on the obstructed side. The finding of high-attenuation medullary pyramids in only one kidney suggests the presence of obstruction in the other kidney.62 Nephrolithiasis is found in 46% of patients with ureteral stones.301 Approximately 5% of patients have ureteral dilatation and perinephric stranding without any demonstrable ureteral calculus63; this constellation of findings should prompt a differential diagnosis of a passed stone, stone too small to visualize, pyelonephritis, or obstruction unrelated to stone disease. However, the constellation of findings in a patient who is undergoing treatment with a protease inhibitor such as ritonavir or indinavir should suggest the diagnosis of ritonavir stone disease (Fig. 54-183).62 Ritonavir stones are nonopaque on CT scans.27 If the patient shows no evidence of infection and has never been treated with protease inhibitors, CT scanning should be repeated with IV contrast material for further evaluation. The other common diagnoses detected on nonenhanced CT  in patients with flank pain that are unrelated to the urinary tract include ovarian masses (with torsion or hemorrhage), appendicitis,

FIG 54-179  Noncontrast renal CT protocol. Axial scan demonstrates a

FIG 54-181  Noncontrast renal CT protocol. Axial scan shows dilated

2- to 3-mm calculus in the ureterovesical junction (arrow) in a patient with left flank pain and hematuria.

calyx in the right renal upper pole (arrow) secondary to obstruction of the ureteropelvic junction (not shown).

A

B FIG 54-180  Noncontrast renal CT protocol. A, Axial scan in a patient with horseshoe kidney shows left ureteric dilatation (arrow); compare it with contralateral ureter (arrowhead). He had a small left ureterovesical calculus (not shown). B, Right perinephric stranding (arrow) secondary to distal ureteric calculus (not shown).

CHAPTER 54  Kidney

A

1851

B

C FIG 54-182  A to C, Noncontrast renal CT scans with “white pyramid sign.” Axial scans show hyperattenuating pyramids in the left kidney (arrows in A and B). The right kidney is edematous, with minimal perinephric stranding, and mild hydronephrosis with absence of hyperattenuating pyramids secondary to a small calculus in the right ureterovesical junction (arrow in C).

CT is particularly helpful in evaluating patients with ureteric obstruction secondary to retroperitoneal fibrosis. Retroperitoneal fibrosis may be due to a variety of causes. About 70% of cases are idiopathic (Fig. 54-184). Other causes are various drugs (especially methysergide), previous abdominal surgery or radiation therapy, inflammatory aortic aneurysm (Fig. 54-185), and inflammatory bowel diseases. Retroperitoneal fibrosis may also be caused by a desmoplastic reaction secondary to retroperitoneal spread of some neoplasms, notably lymphomas, as well as carcinomas of the breast, stomach, lung, colon, and bladder (Fig. 54-186).8

variety of conditions that cause bladder outlet obstruction.315 Urinomas may also occur in patients who have experienced penetrating abdominal trauma, renal surgery, or urinary tract instrumentation.315 On CT a urinoma is usually demonstrated to be confined to the perinephric space by the cone of the renal fascia and often to extend into the infrarenal part of the space. Most urinomas occur posterior to the kidney, which is displaced upward anteriorly and sometimes laterally (Figs. 54-187 and 54-188).315 They often have a thickened wall that may contain dystrophic calcification (see Fig. 54-188). Urinomas usually contain fluid of uniform water density. Hydronephrosis is usually present and is aggravated by ureteric compression due to the urinoma. The nature of a urinoma may be confirmed by delayed contrast-enhanced CT, which sometimes shows layering of contrast medium in the dependent part of the collection (see Fig. 54-187). Management of urinomas resulting from urinary obstruction consists of overcoming the obstruction and excision or drainage of the urinoma.315

URINOMA

LITHIUM NEPHROTOXICITY

Continued leakage of urine from the collecting system in the presence of urinary obstruction may lead to an encapsulated retroperitoneal urine collection called a urinoma. Urinomas may be associated with ureteropelvic junction obstruction, retroperitoneal fibrosis, retroperitoneal malignancy, cancer of the renal pelvis, ureter, or bladder, and a

Lithium-containing drugs are used for affective disorders. Chronic use of lithium can lead to nephrotoxicity or occasionally chronic renal insufficiency. Imaging findings characteristic for lithium use include distal tubular dilatation and microcysts, seen in about 33% to 62% of

diverticulitis, choledocholithiasis, Crohn’s disease, and pancreatitis. The two most common diagnoses detected on nonenhanced CT that are related to the urinary tract but are unrelated to stone disease are pyelonephritis and bladder outlet obstruction.63

RETROPERITONEAL FIBROSIS

1852

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 54-183  Protease inhibitor deposition in a 35-year-old man with human immunodeficiency virus disease who is undergoing indinavir therapy. A, Axial noncontrast CT scan reveals hydronephrosis and minimal perinephric stranding of the right kidney. B, A more caudal scan shows hydroureter (arrow). C, A more caudal scan demonstrates the persistence of hydroureter to the level of the ureterovesical junction (arrow). No calcification was identified. D, Retrograde ureterogram shows multiple filling defects in the distal right ureter. Stricture of the ureter distal to the indinavir fragments is likely caused by recent stone impaction near the ureterovesical junction. (From Dalrymple NC, et al: Pearls and pitfalls in the diagnosis of ureterolithiasis with nonenhanced helical CT. Radiographics 20:439–447, 2000.)

A

B FIG 54-184  Idiopathic retroperitoneal fibrosis. A, A rim of fibrous tissue (arrowheads) surrounds the aorta, which contains calcified atherosclerotic plaques. The fibrous tissue obscures the plane between the aorta and the vena cava (black arrow). Fibrous tissue strands (white arrows) extend laterally to surround the ureters, which contain stents. There is bilateral calyceal dilation. B, Proton-density–weighted MRI (spin echo, TR = 1800 msec, TE = 15 msec) obtained 6 months after bilateral ureterolysis reveals a periaortic rim of fibrous tissue (arrows) with a signal intensity higher than that of muscle. The inferior vena cava (arrowhead) is patent but is attached to the fibrous tissue.

CHAPTER 54  Kidney

1853

B

A

FIG 54-185  Retroperitoneal fibrosis and ureteric obstruction due to an inflammatory aortic aneurysm. A, There is bilateral hydronephrosis. B, A more caudal scan shows a calcified atherosclerotic aortic aneurysm (arrowheads) surrounded by thick fibrous tissue (arrows) that entraps the ureters (curved arrows).

a L

FIG 54-186  Malignant retroperitoneal fibrosis due to metastatic gastric carcinoma. There is bilateral hydronephrosis with no contrast excretion on the right. The contrast-enhanced aorta (a) is surrounded by an illdefined mantle of tissue (arrows) that displaces the vena cava (arrowhead) to the right. There is bilateral renal fascial thickening.

patients undergoing lithium therapy. They are typically tiny (≈1-2 mm), located in the cortex and medulla, and result from proliferation of distal tubular cells199 (Figure 54-189). MRI is superior to US and CT in detecting these cysts. T2-weighted imaging demonstrates tiny high-signal-intensity foci randomly distributed in the kidneys bilaterally. US has a characteristic appearance in lithium nephropathy, demonstrating multiple tiny echogenic foci in the renal parenchyma (Fig. 54-190). Since the cysts are too small for US resolution, the walls of the microcysts appear as tiny echogenic foci without posterior acoustic shadowing. Lithium nephropathy can be distinguished from ADPKD by the normal size of the kidneys and homogenous size of the cysts, ranging between 1 and 2 mm, as well as clinical correlation with lithium use.

RENAL TRANSPLANTS Radiologic evaluation for suspected surgical complications and kidney dysfunction in patients with renal transplants is best achieved by renal scintigraphy and US. Peritransplant fluid collections, including hematomas, lymphoceles, abscesses, and urinomas, are readily assessed on

A

L

B FIG 54-187  Right perinephric urinoma secondary to ureteric obstruction caused by lymph node metastases from a testicular carcinoma.  A, Perinephric fluid collection causes anterior displacement of the right kidney. Contrast material (thick arrow) is present posteriorly in the perinephric space. The right renal pelvis (curved arrow) is dilated. Enlarged retroperitoneal lymph nodes (L) are evident. B, A more caudal scan exhibits contrast material (black arrow) tracking posteriorly from a dilated and obstructed right ureter (white arrow). There is a bridging septum (open arrow) between the renal capsule and the posterior renal fascia.

1854

PART II  CT and MR Imaging of the Whole Body LOGIQ E9

TRANS RT KIDNEY MID

FIG 54-190  Lithium nephropathy. Sagittal grayscale US image of the

A

kidney demonstrates multiple tiny echogenic foci in the cortex, representing tiny microcysts.

MRI is a suitable alternative in the evaluation of the transplanted kidney and peritransplant region6,220; however, US and sonographicguided biopsy remain the primary imaging and interventional modalities.

REFERENCES

B FIG 54-188  Chronic asymptomatic urinoma secondary to renal surgery during childhood. A, A fluid-density mass with peripheral calcification (arrow) extends above the right kidney. B, A more caudal scan reveals that the mass (arrow) extends posterior to the kidney in the perinephric space.

FIG 54-189  Lithium nephropathy. Contrast-enhanced CT of the abdomen at the level of the kidneys demonstrates multiple bilateral tiny (1- to 2-mm) renal cysts.

US. Nonenhanced CT and MRI are reserved for cases in which US fails either because of lack of access due to a recent surgical incision or because the transplant area is obscured by intestinal gas. Contrastenhanced CT should be avoided because of the potential for nephrotoxicity.

1. Aarnio M, Mecklin JP, Aaltonen LA, et al: Life-time risk of different cancers in hereditary non-polyposis colorectal cancer (HNPCC) syndrome. Int J Cancer 64:430–433, 1995. 2. Agrawal R, Jafri SZ, Gibson DP, et al: Juxtaglomerular cell tumor: MR findings. J Comput Assist Tomogr 19:140–142, 1995. 3. Aizenstein RI, Owens C, Sabnis S, et al: The perinephric space and renal fascia: Review of normal anatomy, pathology, and pathways of disease spread. Crit Rev Diagn Imaging 38:325–367, 1997. 4. Akbar SA, Mortele KJ, Baeyens K, et al: Multidetector CT urography: Techniques, clinical applications, and pitfalls. Semin Ultrasound CT MR 25:41–54, 2004. 5. Alexiev BA: Meckel-Gruber syndrome: Pathologic manifestations, minimal diagnostic criteria, and differential diagnosis. Arch Pathol Lab Med 130:1236–1238, 2006. 6. Ali MG, Coakley FV, Hricak H, et al: Complex posttransplantation abnormalities of renal allografts: Evaluation with MR imaging. Radiology 211:95–100, 1999. 7. Alter AJ, Uehling DT, Zwiebel WJ: Computed tomography of the retroperitoneum following nephrectomy. Radiology 133:663–668, 1979. 8. Amis ES, Jr: Retroperitoneal fibrosis. AJR Am J Roentgenol 157:321–329, 1991. 9. Anglard P, Tory K, Brauch H, et al: Molecular analysis of genetic changes in the origin and development of renal cell carcinoma. Cancer Res 51:1071–1077, 1991. 10. Angulo JC, Sanchez-Chapado M, Diego A, et al: Renal echinococcosis: Clinical study of 34 cases. J Urol 157:787–794, 1997. 11. Aragona F, Di Candio G, Serretta V, et al: Renal hydatid disease: Report of 9 cases and discussion of urologic diagnostic procedures. Urol Radiol 6:182–186, 1984. 12. Aronson S, Frazier HA, Baluch JD, et al: Cystic renal masses: Usefulness of the Bosniak classification. Urol Radiol 13:83–90, 1991. 13. Bae KT, Heiken JP, Siegel CL, et al: Renal cysts: Is attenuation artifactually increased on contrast-enhanced CT images? Radiology 216:792–796, 2000. 14. Baron RL, McClennan BL, Lee JK, et al: Computed tomography of transitional-cell carcinoma of the renal pelvis and ureter. Radiology 144:125–130, 1982.

CHAPTER 54  Kidney 15. Beachley MC, Ranniger K, Roth FJ: Xanthogranulomatous pyelonephritis. Am J Roentgenol Radium Ther Nucl Med 121:500–507, 1974. 16. Bechtold RE, Dyer RB, Zagoria RJ, et al: The perirenal space: Relationship of pathologic processes to normal retroperitoneal anatomy. Radiographics 16:841–854, 1996. 17. Bechtold RE, Zagoria RJ: Imaging approach to staging of renal cell carcinoma. Urol Clin North Am 24:507–522, 1997. 18. Beebe DK: Autosomal dominant polycystic kidney disease. Am Fam Physician 53:925–931, 1996. 19. Bell TV, Fenlon HM, Davison BD, et al: Unenhanced helical CT criteria to differentiate distal ureteral calculi from pelvic phleboliths. Radiology 207:363–367, 1998. 20. Belville JS, Morgentaler A, Loughlin KR, et al: Spontaneous perinephric and subcapsular renal hemorrhage: Evaluation with CT, US, and angiography. Radiology 172:733–738, 1989. 21. Bernardino ME, deSantos LA, Johnson DE, et al: Computed tomography in the evaluation of post-nephrectomy patients. Radiology 130:183–187, 1979. 22. Bhatt S, MacLennan G, Dogra V: Renal pseudotumors. AJR Am J Roentgenol 188:1380–1387, 2007. 23. Birnbaum BA, Bosniak MA, Megibow AJ: Asymmetry of the renal nephrograms on CT: Significance of the unilateral prolonged cortical nephrogram. Urol Radiol 12:173–177, 1991. 24. Birnbaum BA, Bosniak MA, Megibow AJ, et al: Observations on the growth of renal neoplasms. Radiology 176:695–701, 1990. 25. Birnbaum BA, Jacobs JE, Langlotz CP, et al: Assessment of a bolustracking technique in helical renal CT to optimize nephrographic phase imaging. Radiology 211:87–94, 1999. 26. Birnbaum BA, Jacobs JE, Ramchandani P: Multiphasic renal CT: Comparison of renal mass enhancement during the corticomedullary and nephrographic phases. Radiology 20:753–758, 1996. 27. Blake SP, McNicholas MM, Raptopoulos V: Nonopaque crystal deposition causing ureteric obstruction in patients with HIV undergoing indinavir therapy. AJR Am J Roentgenol 171:717–720, 1998. 28. Blumhagen JD, Wood BJ, Rosenbaum DM: Sonographic evaluation of abdominal lymphangiomas in children. J Ultrasound Med 6:487–495, 1987. 29. Boone TB, Gilling PJ, Husmann DA: Ureteropelvic junction disruption following blunt abdominal trauma. J Urol 150:33–36, 1993. 30. Boridy IC, Nikolaidis P, Kawashima A, et al: Ureterolithiasis: Value of the tail sign in differentiating phleboliths from ureteral calculi at nonenhanced helical CT. Radiology 211:619–621, 1999. 31. Bosniak MA: The current radiological approach to renal cysts. Radiology 158:1–10, 1986. 32. Bosniak MA: Spontaneous subcapsular and perirenal hematomas. Radiology 172:601–602, 1989. 33. Bosniak MA: Difficulties in classifying cystic lesions of the kidney. Urol Radiol 13:91–93, 1991. 34. Bosniak MA: The small (≤3.0 cm) renal parenchymal tumor: Detection, diagnosis, and controversies. Radiology 179:307–317, 1991. 35. Bosniak MA, Megibow AJ, Ambos MA, et al: Computed tomography of ureteral obstruction. AJR Am J Roentgenol 138:1107–1113, 1982. 36. Bosniak MA, Megibow AJ, Hulnick DH, et al: CT diagnosis of renal angiomyolipoma: The importance of detecting small amounts of fat. AJR Am J Roentgenol 151:497–501, 1988. 37. Bracken RB, Chica G, Johnson DE, et al: Secondary renal neoplasms: An autopsy study. South Med J 72:806–807, 1979. 38. Breslow N, Beckwith JB, Ciol M, et al: Age distribution of Wilms’ tumor: Report from the National Wilms’ Tumor Study. Cancer Res 48:1653–1657, 1988. 39. Broome DR, Girguis MS, Baron PW, et al: Gadodiamide-associated nephrogenic systemic fibrosis: Why radiologists should be concerned. AJR Am J Roentgenol 188:586–592, 2007. 40. Bruno D, Wigfall DR, Zimmerman SA, et al: Genitourinary complications of sickle cell disease. J Urol 166:803–811, 2001. 41. Caoili EM, Cohan RH, Korobkin M, et al: Effectiveness of abdominal compression during helical renal CT. Acad Radiol 8:1100–1106, 2001.

1855

42. Chan HS, Daneman A, Gribbin M, et al: Renal cell carcinoma in the first two decades of life. Pediatr Radiol 13:324–328, 1983. 43. Chauveau D, Knebelmann B, Grunfeld JP: Inherited kidney diseases: Polycystic kidney disease and Alport syndrome. Adv Intern Med 40:303–339, 1995. 44. Choyke PL: Inherited cystic diseases of the kidney. Radiol Clin North Am 34:925–946, 1996. 45. Choyke PL: Acquired cystic kidney disease. Eur Radiol 10:1716–1721, 2000. 46. Choyke PL, Glenn GM, Walther MM, et al: The natural history of renal lesions in von Hippel-Lindau disease: A serial CT study in 28 patients. AJR Am J Roentgenol 159:1229–1234, 1992. 47. Choyke PL, Glenn GM, Walther MM, et al: von Hippel-Lindau disease: Genetic, clinical, and imaging features. Radiology 194:629–642, 1995. 48. Choyke PL, Glenn GM, Walther MM, et al: Hereditary renal cancers. Radiology 226:33–46, 2003. 49. Choyke PL, Walther MM, Glenn GM, et al: Imaging features of hereditary papillary renal cancers. J Comput Assist Tomogr 21:737–741, 1997. 50. Choyke PL, White EM, Zeman RK, et al: Renal metastases: Clinicopathologic and radiologic correlation. Radiology 162:359–363, 1987. 51. Churchill DN, Bear JC, Morgan J, et al: Prognosis of adult onset polycystic kidney disease re-evaluated. Kidney Int 26:190–193, 1984. 52. Cohan RH, Dunnick NR, Leder RA, et al: Computed tomography of renal lymphoma. J Comput Assist Tomogr 14:933–938, 1990. 53. Cohan RH, Sherman LS, Korobkin M, et al: Renal masses: Assessment of corticomedullary-phase and nephrographic-phase CT scans. Radiology 196:445–451, 1995. 54. Coulam CH, Sheafor DH, Leder RA, et al: Evaluation of pseudoenhancement of renal cysts during contrast-enhanced CT. AJR Am J Roentgenol 174:493–498, 2000. 55. Cowper SE, Su L, Robin H, et al: Nephrogenic fibrosing dermopathy. Am J Dermatopathol 23:383–393, 2001. 56. Crotty KL, Orihuela E, Warren MM: Recent advances in the diagnosis and treatment of renal arteriovenous malformations and fistulas. J Urol 150:1355–1359, 1993. 57. Crummy AB, Madsen PO: Parapelvic renal cyst: The peripheral fat sign. J Urol 96:436–438, 1966. 58. Curry NS, Chung CJ, Gordon B: Unilateral renal cystic disease in an adult. Abdom Imaging 19:366–368, 1994. 59. Curry NS, Schabel SI, Garvin AJ, et al: Intratumoral fat in a renal oncocytoma mimicking angiomyolipoma. AJR Am J Roentgenol 154:307–308, 1990. 60. Cushing B, Slovis TL: Imaging of Wilms’ tumor: What is important! Urol Radiol 14:241–251, 1992. 61. D’Agata ID, Jonas MM, Perez-Atayde AR, et al: Combined cystic disease of the liver and kidney. Semin Liver Dis 14:215–228, 1994. 62. Dalrymple NC, Casford B, Raiken DP, et al: Pearls and pitfalls in the diagnosis of ureterolithiasis with unenhanced helical CT. Radiographics 20:439–447, 2000. 63. Dalrymple NC, Verga M, Anderson KR, et al: The value of unenhanced helical computerized tomography in the management of acute flank pain. J Urol 159:735–740, 1998. 64. Davidson AJ, Choyke PL, Hartman DS, et al: Renal medullary carcinoma associated with sickle cell trait: Radiologic findings. Radiology 195:83–85, 1995. 65. Davidson AJ, Hartman DS, Choyke PL, et al: Davidson’s Radiology of the Kidney and Genitourinary Tract, Philadelphia, 1999, WB Saunders, pp 157–235. 66. Davidson AJ, Hayes WS, Hartman DS, et al: Renal oncocytoma and carcinoma: Failure of differentiation with CT. Radiology 186:693–696, 1993. 67. Davies CH, Stringer DA, Whyte H, et al: Congenital hepatic fibrosis with saccular dilatation of intrahepatic bile ducts and infantile polycystic kidneys. Pediatr Radiol 16:302–305, 1986. 68. Davis JM, 3rd, McLaughlin AP: Spontaneous renal hemorrhage due to cyst rupture: CT findings. AJR Am J Roentgenol 148:763–764, 1987.

1856

PART II  CT and MR Imaging of the Whole Body

69. Davis CJ, Jr, Mostofi FK, Sesterhenn IA: Renal medullary carcinoma: The seventh sickle cell nephropathy. Am J Surg Pathol 19:1–11, 1995. 70. De Carli P, Vidiri A, Lamanna L, et al: Renal oncocytoma: Image diagnostics and therapeutic aspects. J Exp Clin Cancer Res 19:287–290, 2000. 71. De Cobelli F, Vanzulli A, Sironi S, et al: Renal artery stenosis: Evaluation with breath-hold, three-dimensional, dynamic, gadolinium-enhanced versus three-dimensional, phase-contrast MR angiography. Radiology 205:689–695, 1997. 72. Diel J, Perlmutter S, Venkataramanan N, et al: Unenhanced helical CT using increased pitch for suspected renal colic: An effective technique for radiation dose reduction? J Comput Assist Tomogr 24:795–801, 2000. 73. Dimopoulos MA, Logothetis CJ, Markowitz A, et al: Collecting duct carcinoma of the kidney. Br J Urol 71:388–391, 1993. 74. Dinney CP, Awad SA, Gajewski JB, et al: Analysis of imaging modalities, staging systems, and prognostic indicators for renal cell carcinoma. Urology 39:122–129, 1992. 75. Dinsmore BJ, Pollack HM, Banner MP: Calcified transitional cell carcinoma of the renal pelvis. Radiology 167:401–404, 1988. 76. Dobyan DC, Truong LD, Eknoyan G: Renal malacoplakia reappraised. Am J Kidney Dis 22:243–252, 1993. 77. Dunnick NR: Renal lesions: Great strides in imaging. Radiology 182:305–306, 1992. 78. Dunnick NR, Hartman DS, Ford KK, et al: The radiology of juxtaglomerular tumors. Radiology 147:321–326, 1983. 79. Dunnick NR, Sandler CM, Amis ESJ, et al: Urinary Tract Trauma: Textbook of Uroradiology, Baltimore, 1997, Williams & Wilkins, pp 297–324. 80. Dyer RB, Chen MY, Zagoria RJ: Abnormal calcifications in the urinary tract. Radiographics 18:1405–1424, 1998. 81. Dyer RB, Chen MY, Zagoria RJ: Intravenous urography: Technique and interpretation. Radiographics 21:799–821, 2001. 82. Dyer RB, Munitz HA, Bechtold R, et al: The abnormal nephrogram. Radiographics 6:1039–1063, 1986. 83. Eilenberg SS, Lee JK, Brown J, et al: Renal masses: Evaluation with gradient-echo Gd-DTPA-enhanced dynamic MR imaging. Radiology 176:333–338, 1990. 84. Evans JC, Curtis J: The radiological appearances of tuberous sclerosis. Br J Radiol 73:91–98, 2000. 85. Farrow GM, Harrison EG, Jr, Utz DC, et al: Sarcomas and sarcomatoid and mixed malignant tumors of the kidney in adults: I. Cancer 22:545–550, 1968. 86. Federle MP: Evaluation of renal trauma. In Pollack HM, editor: Clinical Urography: An Atlas and Textbook of Urological Imaging, Philadelphia, 1990, WB Saunders. 87. Federle MP, McAninch JW, Kaiser JA, et al: Computed tomography of urinary calculi. AJR Am J Roentgenol 136:255–258, 1981. 88. Feldberg MA, Koehler PR, van Waes PF: Psoas compartment disease studied by computed tomography: Analysis of 50 cases and subject review. Radiology 148:505–512, 1983. 89. Fernbach SK, Feinstein KA, Donaldson JS, et al: Nephroblastomatosis: Comparison of CT with US and urography. Radiology 166:153–156, 1988. 90. Fick GM, Gabow PA: Natural history of autosomal dominant polycystic kidney disease. Annu Rev Med 45:23–29, 1994. 91. Fielding JR, Silverman SG, Rubin GD: Helical CT of the urinary tract. AJR Am J Roentgenol 172:199–206, 1999. 92. Finn JP, Goldmann A, Edelman RR: Magnetic resonance angiography in the body. Magn Reson Q 8:1–22, 1992. 93. Fishman EK, Hartman DS, Goldman SM, et al: The CT appearance of Wilms tumor. J Comput Assist Tomogr 7:659–665, 1983. 94. Fishman MC, Pollack HM, Arger PH, et al: Radiographic manifestations of spontaneous renal sinus hemorrhage. AJR Am J Roentgenol 142:1161–1164, 1984. 95. Fultz PJ, Hampton WR, Totterman SM: Computed tomography of pyonephrosis. Abdom Imaging 18:82–87, 1993. 96. Gabow PA: Autosomal dominant polycystic kidney disease-more than a renal disease. Am J Kidney Dis 16:403–413, 1990.

97. Gardner KD, Jr: Pathogenesis of human cystic renal disease. Annu Rev Med 39:185–191, 1988. 98. Gatewood OM, Fishman EK, Burrow CR, et al: Renal vein thrombosis in patients with nephrotic syndrome: CT diagnosis. Radiology 159:117–122, 1986. 99. Geiger J, Fong Q, Fay R: Transitional cell carcinoma of renal pelvis with invasion of renal vein and thrombosis of subhepatic inferior vena cava. Urology 28:52–54, 1986. 100. Geller E, Smergel EM, Lowry PA: Renal neoplasms of childhood. Radiol Clin North Am 35:1391–1413, 1997. 101. Gibson SE, Farver CF, Prayson RA: Multiorgan involvement in nephrogenic fibrosing dermopathy: An autopsy case and review of the literature. Arch Pathol Lab Med 130:209–212, 2006. 102. Glazer GM, Francis IR, Gross BH, et al: Computed tomography of renal vein thrombosis. J Comput Assist Tomogr 8:288–293, 1984. 103. Glazer GM, Gross BH, Orringer MB, et al: Computed tomography of pericardial masses: Further observations and comparison with echocardiography. J Comput Assist Tomogr 8:895–899, 1984. 104. Gofrit ON, Shapiro A, Kovalski N, et al: Renal cell carcinoma: Evaluation of the 1997 TNM system and recommendations for follow-up after surgery. Eur Urol 39:669–674, 2001. 105. Gold RP, McClennan BL, Rottenberg RR: CT appearance of acute inflammatory disease of the renal interstitium. AJR Am J Roentgenol 141:343–349, 1983. 106. Goldfarb DA, Novick AC, Lorig R, et al: Magnetic resonance imaging for assessment of vena caval tumor thrombi: A comparative study with venacavography and computerized tomography scanning. J Urol 144:1100–1103, 1990. 107. Goldman SM, Hartman DS, Fishman EK, et al: CT of xanthogranulomatous pyelonephritis: Radiologic-pathologic correlation. AJR Am J Roentgenol 142:963–969, 1984. 108. Grabstald H, Whitmore WF, Melamed MR: Renal pelvic tumors. JAMA 218:845–854, 1971. 109. Greene A, Cromie WJ, Goldman M: Computerized body   tomography in neonatal renal vein thrombosis. Urology 20:213–215, 1982. 110. Grobner T: Gadolinium: A specific trigger for the development of nephrogenic fibrosing dermopathy and nephrogenic systemic fibrosis? Nephrol Dial Transplant 21:1104–1108, 2006. 111. Grubb RL, Franks ME, Toro J, et al: Hereditary leiomyomatosis and renal cell cancer: A syndrome associated with an aggressive form of inherited renal cancer. J Urol 177:2074–2079, discussion 2079–2080, 2007. 112. Gylys-Morin V, Hoffer FA, Kozakewich H, et al: Wilms tumor and nephroblastomatosis: Imaging characteristics at gadolinium-enhanced MR imaging. Radiology 188:517–521, 1993. 113. Hall JR, Choa RG, Wells IP: Percutaneous drainage in emphysematous pyelonephritis: An alternative to major surgery. Clin Radiol 39:622–624, 1988. 114. Harisinghani MG, McLoud TC, Shepard JAO, et al: Tuberculosis from head to toe [CME available in print version and on RSNA Link]. Radiographics 20:449–470, 2000. 115. Harris AC, Zwirewich CV, Lyburn ID, et al: CT findings in blunt renal trauma. Radiographics 21(Spec No):S201–S214, 2001. 116. Hartman DS, Davis CJ, Jr, Lichtenstein JE, et al: Renal parenchymal malacoplakia. Radiology 136:33–42, 1980. 117. Hartman DS, Weatherby E, 3rd, Laskin WB, et al: Cystic renal cell carcinoma: CT findings simulating a benign hyperdense cyst. AJR Am J Roentgenol 159:1235–1237, 1992. 118. Hatcher PA, Anderson EE, Paulson DF, et al: Surgical management and prognosis of renal cell carcinoma invading the vena cava. J Urol 145:20–23, 1991. 119. Hauser S, Studer UE: Therapy of carcinoma of the kidney pelvis [in German]. Urologe A 40:452–455, 2001. 120. Helenon O, Chretien Y, Paraf F, et al: Renal cell carcinoma containing fat: Demonstration with CT. Radiology 188:429–430, 1993. 121. Henderson NL, Mason RC: Juxtaglomerular cell tumor in pregnancy. Obstet Gynecol 98:943–945, 2001.

CHAPTER 54  Kidney 122. Heneghan JP, Dalrymple NC, Verga M, et al: Soft-tissue “rim” sign in the diagnosis of ureteral calculi with use of unenhanced helical CT. Radiology 202:709–711, 1997. 123. Hes O, Michal M, Sulc M, et al: Oncocytoma of the kidney: Morphologic variation in 102 cases [in Czech]. Cesk Patol 37:51–56, 2001. 124. Hidai H, Chiba T, Takagi Y, et al: Bilateral chromophobe cell renal carcinoma in tuberous sclerosis complex. Int J Urol 4:86–89, 1997. 125. Hidalgo H, Dunnick NR, Rosenberg ER, et al: Parapelvic cysts: Appearance on CT and sonography. AJR Am J Roentgenol 138:667–671, 1982. 126. Hohenwalter MD, Skowlund CJ, Erickson SJ, et al: Renal transplant evaluation with MR angiography and MR imaging. Radiographics 21:1505–1517, 2001. 127. Honda H, Bonsib S, Barloon TJ, et al: Unusual renal oncocytomas: Pathologic and CT correlations. Urol Radiol 14:148–154, 1992. 128. Huppert BJ, Brandt KR, Ramin KD, et al: Single-shot fast spin-echo MR imaging of the fetus: A pictorial essay. Radiographics 19:215–227, 1999. 129. Israel GM, Bosniak MA: Calcification in cystic renal masses: Is it important in diagnosis? Radiology 226:47–52, 2003. 130. Israel GM, Bosniak MA: Follow-up CT studies for moderately complex cystic renal masses (Bosniak category IIF). AJR Am J Roentgenol 181:627–633, 2003. 131. Israel GM, Bosniak MA: An update of the Bosniak renal cyst classification system. Urology 66:484–488, 2005. 132. Israel GM, Hindman N, Hecht E, et al: The use of opposed-phase chemical shift MRI in the diagnosis of renal angiomyolipomas. AJR Am J Roentgenol 184:1868–1872, 2005. 133. Jimenez RE, Eble JN, Reuter VE, et al: Concurrent angiomyolipoma   and renal cell neoplasia: A study of 36 cases. Mod Pathol 14:157–163, 2001. 134. Johnson CD, Dunnick NR, Cohan RH, et al: Renal adenocarcinoma: CT staging of 100 tumors. AJR Am J Roentgenol 148:59–63, 1987. 135. Jonisch AI, Rubinowitz AN, Mutalik PG, et al: Can high-attenuation renal cysts be differentiated from renal cell carcinoma at unenhanced CT? Radiology 243:445–450, 2007. 136. Jordan J, Low R, Jeffrey RB, Jr: CT findings in acute renal cortical necrosis. J Comput Assist Tomogr 14:155–156, 1990. 137. Joshi VV, Beckwith JB: Multilocular cyst of the kidney (cystic nephroma) and cystic, partially differentiated nephroblastoma: Terminology and criteria for diagnosis. Cancer 64:466–479, 1989. 138. Julian JC, Merguerian PA, Shortliffe LM: Pediatric genitourinary tumors. Curr Opin Oncol 7:265–274, 1995. 139. Kallman DA, King BF, Hattery RR, et al: Renal vein and inferior vena cava tumor thrombus in renal cell carcinoma: CT, US, MRI and venacavography. J Comput Assist Tomogr 16:240–247, 1992. 140. Kamel IR, Berkowitz JF: Assessment of the cortical rim sign in posttraumatic renal infarction. J Comput Assist Tomogr 20:803–806, 1996. 141. Kashiwabara H, Inaba M, Itabashi A, et al: A case of renin-producing juxtaglomerular tumor: Effect of ACE inhibitor or angiotensin II receptor antagonist. Blood Press 6:147–153, 1997. 142. Kasper HU, Buhtz P, Kruger G, et al: Bellini duct carcinoma of the kidney: A case report. Gen Diagn Pathol 143:237–241, 1997. 143. Kawamura K, Ikeda R, Suzuki K: Acquired renal cystic disease [in Japanese]. Hinyokika Kiyo 46:651–656, 2000. 144. Kawashima A, Goldman SM, Sandler CM: The indeterminate renal mass. Radiol Clin North Am 34:997–1015, 1996. 145. Kawashima A, Sandler CM, Boridy IC, et al: Unenhanced helical CT of ureterolithiasis: Value of the tissue rim sign. AJR Am J Roentgenol 168:997–1000, 1997. 146. Kawashima A, Sandler CM, Corl FM, et al: Imaging of renal trauma: A comprehensive review. Radiographics 21:557–574, 2001. 147. Kawashima A, Sandler CM, Ernst RD, et al: CT evaluation of renovascular disease. Radiographics 20:1321–1340, 2000. 148. Kawashima A, Vrtiska TJ, LeRoy AJ, et al: CT urography. Radiographics 24(Suppl 1):S35–S54, 2004.

1857

149. Kennedy SM, Merino MJ, Linehan WM, et al: Collecting duct carcinoma of the kidney. Hum Pathol 21:449–456, 1990. 150. Kennelly MJ, Grossman HB, Cho KJ: Outcome analysis of 42 cases of renal angiomyolipoma. J Urol 152:1988–1991, 1994. 151. Kenney PJ: Imaging of chronic renal infections. AJR Am J Roentgenol 155:485–494, 1990. 152. Kern S, Zimmerhackl LB, Hildebrandt F, et al: RARE-MR-urography: A new diagnostic method in autosomal recessive polycystic kidney disease. Acta Radiol 40:543–544, 1999. 153. Kern S, Zimmerhackl LB, Hildebrandt F, et al: Appearance of autosomal recessive polycystic kidney disease in magnetic resonance imaging and RARE-MR-urography. Pediatr Radiol 30:156–160, 2000. 154. Khan A, Thomas N, Costello B, et al: Renal medullary carcinoma: Sonographic, computed tomography, magnetic resonance and angiographic findings. Eur J Radiol 35:1–7, 2000. 155. Kim SH, Cho SW, Kim HD, et al: Nutcracker syndrome: Diagnosis with Doppler US. Radiology 198:93–97, 1996. 156. Kleiner B, Filly RA, Mack L, et al: Multicystic dysplastic kidney: Observations of contralateral disease in the fetal population. Radiology 161:27–29, 1986. 157. Kocakoc E, Bhatt S, Dogra V: Renal multidetector row CT. Radiol Clin North Am 43:1021–1047, 2005. 158. Koch KJ, Cory DA: Simultaneous renal vein thrombosis and bilateral adrenal hemorrhage: MR demonstration. J Comput Assist Tomogr 10:681–683, 1986. 159. Kohno A, Yunoki M: Unilateral renal cystic disease. Radiat Med 17:423–426, 1999. 160. Kopka L, Fischer U, Zoeller G, et al: Dual-phase helical CT of the kidney: Value of the corticomedullary and nephrographic phase for evaluation of renal lesions and preoperative staging of renal cell carcinoma. AJR Am J Roentgenol 169:1573–1578, 1997. 161. Kopka L, Rogalla P, Hamm B: Multislice CT of the abdomen-current indications and future trends [in German]. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 174:273–282, 2002. 162. Kossol JM, Patel SK: Suburothelial hemorrhage: The value of preinfusion computed tomography. J Comput Assist Tomogr 10:157–158, 1986. 163. Kunin M: Bridging septa of the perinephric space: Anatomic, pathologic, and diagnostic considerations. Radiology 158:361–365, 1986. 164. Kuo PH, Kanal E, Abu-Alfa AK, et al: Gadolinium-based MR contrast agents and nephrogenic systemic fibrosis. Radiology 242:647–649, 2007. 165. Kutikov A, Uzzo RG: The R.E.N.A.L. nephrometry score: A comprehensive standardized system for quantitating renal tumor size, location and depth. J Urol 182:844–853, 2009. 166. Lane RH, Stephens DH, Reiman HM: Primary retroperitoneal neoplasms: CT findings in 90 cases with clinical and pathologic correlation. AJR Am J Roentgenol 152:83–89, 1989. 167. Launonen V, Vierimaa O, Kiuru M, et al: Inherited susceptibility to uterine leiomyomas and renal cell cancer. Proc Natl Acad Sci U S A 98:3387–3392, 2001. 168. Leder RA, Dunnick NR: Transitional cell carcinoma of the pelvicalices and ureter. AJR Am J Roentgenol 155:713–722, 1990. 169. Lee YJ, Oh SN, Rha SE, et al: Renal trauma. Radiol Clin North Am 45:581–592, 2007. 170. Leppaniemi A, Lamminen A, Tervahartiala P, et al: Comparison of high-field magnetic resonance imaging with computed tomography in the evaluation of blunt renal trauma. J Trauma 38:420–427, 1995. 171. Lesavre A, Correas J, Merran S, et al: CT of papillary renal cell carcinomas with cholesterol necrosis mimicking angiomyolipomas. AJR Am J Roentgenol 181:143–145, 2003. 172. Levine E: Computed tomography of renal abscesses complicating medullary sponge kidney. J Comput Assist Tomogr 13:440–442, 1989. 173. Levine E: Malignant renal parenchymal tumors in adults. In Pollack HM, editor: Clinical Urography: An Atlas and Textbook of Urological Imaging, Philadelphia, 1990, WB Saunders. 174. Levine E: Transitional cell carcinoma of the renal pelvis associated   with cyclophosphamide therapy. AJR Am J Roentgenol 159:1027–1028, 1992.

1858

PART II  CT and MR Imaging of the Whole Body

175. Levine E: Acquired cystic kidney disease. Radiol Clin North Am 34:947–964, 1996. 176. Levine E, Collins DL, Horton WA, et al: CT screening of the abdomen in von Hippel-Lindau disease. AJR Am J Roentgenol 139:505–510, 1982. 177. Levine E, Cook LT, Grantham JJ: Liver cysts in autosomal-dominant polycystic kidney disease: Clinical and computed tomographic study. Am J Roentgenol 145:229–233, 1985. 178. Levine E, Grantham JJ: High-density renal cysts in autosomal dominant polycystic kidney disease demonstrated by CT. Radiology 154:477–482, 1985. 179. Levine E, Grantham JJ: Perinephric hemorrhage in autosomal dominant polycystic kidney disease: CT and MR findings. J Comput Assist Tomogr 11:108–111, 1987. 180. Levine E, Grantham JJ: Calcified renal stones and cyst calcifications in autosomal dominant polycystic kidney disease: Clinical and CT study in 84 patients. AJR Am J Roentgenol 159:77–81, 1992. 181. Levine E, Grantham JJ, MacDougall ML: Spontaneous subcapsular and perinephric hemorrhage in end-stage kidney disease: Clinical and CT findings. AJR Am J Roentgenol 148:755–758, 1987. 182. Levine E, Grantham JJ, Slusher SL, et al: CT of acquired cystic kidney disease and renal tumors in long-term dialysis patients. AJR Am J Roentgenol 142:125–131, 1984. 183. Levine E, Huntrakoon M: Unilateral renal cystic disease: CT findings.   J Comput Assist Tomogr 13:273–276, 1989. 184. Levine E, Huntrakoon M, Wetzel LH: Small renal neoplasms: Clinical, pathologic, and imaging features. AJR Am J Roentgenol 153:69–73, 1989. 185. Levine E, Lee KR, Weigel J: Preoperative determination of abdominal extent of renal cell carcinoma by computed tomography. Radiology 132:395–398, 1979. 186. Levine E, Slusher SL, Grantham JJ, et al: Natural history of acquired renal cystic disease in dialysis patients: A prospective longitudinal CT study. AJR Am J Roentgenol 156:501–506, 1991. 187. Liebl R: Nutcracker phenomenon or nutcracker syndrome? Nephrol Dial Transplant 20:2009, author reply 2009, 2005. 188. Lien HH, Lund G: Collateral veins in inferior caval vein occlusion demonstrated via CT. Eur J Radiol 3:319–323, 1983. 189. Lindsey JR: Lymphangiectasia simulating polycystic disease. J Urol 104:658–662, 1970. 190. Liu J, Fanning CV: Can renal oncocytomas be distinguished from renal cell carcinoma on fine-needle aspiration specimens? A study of conventional smears in conjunction with ancillary studies. Cancer 93:390–397, 2001. 191. Lonergan GJ, Martinez-Leon MI, Agrons GA, et al: Nephrogenic rests, nephroblastomatosis, and associated lesions of the kidney. Radiographics 18:947–968, 1998. 192. Lonergan GJ, Rice RR, Suarez ES: Autosomal recessive polycystic kidney disease: Radiologic-pathologic correlation. Radiographics 20:837–855, 2000. 193. Lowe LH, Isuani BH, Heller RM, et al: Pediatric renal masses: Wilms tumor and beyond. Radiographics 20:1585–1603, 2000. 194. Lumerman J, Gershbaum MD, Hines J, et al: Unenhanced helical computed tomography for the evaluation of suspected renal colic in the adolescent population: A pilot study. Urology 57:342–346, 2001. 195. Lumsden AB, Salam TA, Walton KG: Renal artery aneurysm: A report of 28 cases. Cardiovasc Surg 4:185–189, 1996. 196. Madewell JE, Goldman SM, Davis CJ, Jr, et al: Multilocular cystic nephroma: A radiographic-pathologic correlation of 58 patients. Radiology 146:309–321, 1983. 197. Maher ER, Iselius L, Yates JR, et al: Von Hippel-Lindau disease: A genetic study. J Med Genet 28:443–447, 1991. 198. Maher ER, Yates JR, Harries R, et al: Clinical features and natural history of von Hippel-Lindau disease. Q J Med 77:1151–1163, 1990. 199. Markovitz GS, Radhakrishnan J, Kammbham N, et al: Lithium nephrotoxicity: A progressive combined glomerular and tubulointerstitial nephropathy. J Am Soc Nephrol 11:1439–1448, 2000.

200. Marshall FF, Stewart AK, Menck HR: The National Cancer Data Base: Report on kidney cancers. The American College of Surgeons Commission on Cancer and the American Cancer Society. Cancer 80:2167–2174, 1997. 201. Matson MA, Cohen EP: Acquired cystic kidney disease: Occurrence, prevalence, and renal cancers. Medicine (Baltimore) 69:217–226, 1990. 202. Matthews LA, Spirnak JP: The nonoperative approach to major blunt renal trauma. Semin Urol 13:77–82, 1995. 203. Matz LR, Latham BI, Fabian VA, et al: Collecting duct carcinoma of the kidney: A report of three cases and review of the literature. Pathology 29:354–359, 1997. 204. McClennan BL, Stanley RJ, Melson GL, et al: CT of the renal cyst: Is cyst aspiration necessary? AJR Am J Roentgenol 133:671–675, 1979. 205. McCollough CH, Bruesewitz MR, Vrtiska TJ, et al: Image quality and dose comparison among screen-film, computed, and CT scanned projection radiography: Applications to CT urography. Radiology 221:395–403, 2001. 206. McTavish JD, Jinzaki M, Zou KH, et al: Multi-detector row CT urography: Comparison of strategies for depicting the normal urinary collecting system. Radiology 225:783–790, 2002. 207. Meredith WT, Levine E, Ahlstrom NG, et al: Exacerbation of familial renal lymphangiomatosis during pregnancy. AJR Am J Roentgenol 151:965–966, 1988. 208. Mertens PR, Duque-Reina D, Ittel TH, et al: Contrast-enhanced computed tomography for demonstration of bilateral renal cortical necrosis. Clin Invest 72:499–501, 1994. 209. Millan-Rodriguez F, Palou J, de la Torre-Holguera P, et al: Conventional CT signs in staging transitional cell tumors of the upper urinary tract. Eur Urol 35:318–322, 1999. 210. Mitnick JS, Bosniak MA, Hilton S, et al: Cystic renal disease in tuberous sclerosis. Radiology 147:85–87, 1983. 211. Mitnick JS, Bosniak MA, Rothberg M, et al: Metastatic neoplasm to the kidney studied by computed tomography and sonography. J Comput Assist Tomogr 9:43–49, 1985. 212. Mitty HA, Shapiro RS, Parsons RB, et al: Renovascular hypertension. Radiol Clin North Am 34:1017–1036, 1996. 213. Morehouse HT, Weiner SN, Hoffman JC: Imaging in inflammatory disease of the kidney. AJR Am J Roentgenol 143:135–141, 1984. 214. Moss JG, Hendry GM: The natural history of renal cysts in an infant with tuberous sclerosis: Evaluation with ultrasound. Br J Radiol 61:1074–1076, 1988. 215. Motzer RJ, Bander NH, Nanus DM: Renal-cell carcinoma. N Engl J Med 335:865–875, 1996. 216. Murray KK, McLellan GL: Renal peripelvic lymphangiectasia: Appearance at CT. Radiology 180:455–456, 1991. 217. Myers MA: Dynamic Radiology of the Abdomen: Normal and Pathological Anatomy, New York, 1994, Springer-Verlag. 218. Nascimento AB, Mitchell DG, Zhang XM, et al: Rapid MR imaging detection of renal cysts: Age-based standards. Radiology 221:628–632, 2001. 219. Nasu K, Yoshimatsu J, Anai T, et al: Magnetic resonance imaging of fetal autosomal recessive polycystic kidney disease. J Obstet Gynaecol Res 24:33–36, 1998. 220. Neimatallah MA, Dong Q, Schoenberg SO, et al: Magnetic resonance imaging in renal transplantation. J Magn Reson Imaging 10:357–368, 1999. 221. Newatia A, Khatri G, Friedman B, et al: Subtraction imaging: Applications for nonvascular abdominal MRI. AJR Am J Roentgenol 188:1018–1025, 2007. 222. Nickerson ML, Warren MB, Toro JR, et al: Mutations in a novel gene lead to kidney tumors, lung wall defects and benign tumors of the hair follicle in patients with Birt-Hogg-Dube syndrome. Cancer Cell 2:157–164, 2002. 223. Niikura S, Komatsu K, Uchibayashi T, et al: Juxtaglomerular cell tumor of the kidney treated with nephron-sparing surgery. Urol Int 65:160– 162, 2000. 224. Nocks BN, Heney NM, Daly JJ, et al: Transitional cell carcinoma of renal pelvis. Urology 19:472–477, 1982.

CHAPTER 54  Kidney 225. Nolte-Ernsting CC, Wildberger JE, Borchers H, et al: Multi-slice CT urography after diuretic injection: Initial results. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 173:176–180, 2001. 226. Nunez D, Jr, Becerra JL, Fuentes D, et al: Traumatic occlusion of the renal artery: Helical CT diagnosis. AJR Am J Roentgenol 167:777–780, 1996. 227. Oliveira EA, Diniz JS, Vilasboas AS, et al: Multicystic dysplastic kidney detected by fetal sonography: Conservative management and follow-up. Pediatr Surg Int 17:54–57, 2001. 228. Oliver RT, Nethersell AB, Bottomley JM: Unexplained spontaneous regression and alpha-interferon as treatment for metastatic renal carcinoma. Br J Urol 63:128–131, 1989. 229. Oto A, Herts BR, Remer EM, et al: Inferior vena cava tumor thrombus in renal cell carcinoma: Staging by MR imaging and impact on surgical treatment. AJR Am J Roentgenol 171:1619–1624, 1998. 230. Padhani AR, Liu G, Koh DM, et al: Diffusion-weighted magnetic resonance imaging as a cancer biomarker: Consensus and recommendations. Neoplasia 11(2):102–125, 2009. 231. Papanicolaou N, Pfister RC: Acute renal infections. Radiol Clin North Am 34:965–995, 1996. 232. Papanicolaou N, Stafford SA, Pfister RC, et al: Significant renal hemorrhage following extracorporeal shock wave lithotripsy: Imaging and clinical features. Radiology 163:661–664, 1987. 233. Parfrey PS, Bear JC, Morgan J, et al: The diagnosis and prognosis of autosomal dominant polycystic kidney disease. N Engl J Med 323:1085– 1090, 1990. 234. Parienty RA, Ducellier R, Pradel J, et al: Diagnostic value of CT numbers in pelvicalyceal filling defects. Radiology 145:743–747, 1982. 235. Parvey LS, Warner RM, Callihan TR, et al: CT demonstration of fat tissue in malignant renal neoplasms: Atypical Wilms’ tumors. J Comput Assist Tomogr 5:851–854, 1981. 236. Pascal RR: Renal manifestations of extrarenal neoplasms. Hum Pathol 11:7–17, 1980. 237. Peccatori I, Pitingolo F, Battini G, et al: Pulmonary lymphangioleiomyomatosis and renal papillary cancer: Incomplete expression of tuberous sclerosis? Nephrol Dial Transplant 12:2740–2743, 1997. 238. Pedicelli G, Jequier S, Bowen AD, et al: Multicystic dysplastic kidneys: Spontaneous regression demonstrated with US. Radiology 161:23–26, 1986. 239. Pei Y, Obaji J, Dupuis A, et al: Unified criteria for ultrasonographic diagnosis of ADPKD. J Am Soc Nephrol 20(1):205, 2009. 240. Peterson RO: Urologic Pathology, Philadelphia, 1992, JB Lippincott. 241. Phillips E, Messing EM: Role of lymphadenectomy in the treatment of renal cell carcinoma. Urology 41:9–15, 1993. 242. Pickering SP, Fletcher BD, Bryan PJ, et al: Renal lymphangioma: A cause of neonatal nephromegaly. Pediatr Radiol 14:445–448, 1984. 243. Pickhardt PJ: Collecting duct carcinoma arising in a solitary kidney: Imaging findings. Clin Imaging 23:115–118, 1999. 244. Pickhardt PJ, Lonergan GJ, Davis CJ, et al: From the Archives of the AFIP: Infiltrative renal lesions: Radiologic-pathologic correlation. Radiographics 20:215–243, 2000. 245. Pickhardt PJ, Siegel CL, McLarney JK: Collecting duct carcinoma of the kidney: Are imaging findings suggestive of the diagnosis? AJR Am J Roentgenol 176:627–633, 2001. 246. Pollack HM, Wein AJ: Imaging of renal trauma. Radiology 172:297–308, 1989. 247. Porch P, Noe HN, Stapleton FB: Unilateral presentation of adult-type polycystic kidney disease in children. J Urol 135:744–746, 1986. 248. Port FK, Ragheb NE, Schwartz AG, et al: Neoplasms in dialysis patients: A population-based study. Am J Kidney Dis 14:119–123, 1989. 249. Poston CD, Jaffe GS, Lubensky IA, et al: Characterization of the renal pathology of a familial form of renal cell carcinoma associated with von Hippel-Lindau disease: Clinical and molecular genetic implications. J Urol 153:22–26, 1995. 250. Pozniak MA, Balison DJ, Lee FT, Jr, et al: CT angiography of potential renal transplant donors. Radiographics 18:565–587, 1998.

1859

251. Pretorius ES, Wickstrom ML, Siegelman ES: MR imaging of renal neoplasms. Magn Reson Imaging Clin N Am 8:813–836, 2000. 252. Pretorius ES, Siegelman ES, Ramchandani P, et al: Renal neoplasms amenable to partial nephrectomy: MR imaging. Radiology 212:28–34, 1999. 253. Prince MR, Narasimham DL, Stanley JC, et al: Breath-hold gadoliniumenhanced MR angiography of the abdominal aorta and its major branches. Radiology 197:785–792, 1995. 254. Prokop M: Multislice CT angiography. Eur J Radiol 36:86–96, 2000. 255. Quinn MJ, Hartman DS, Friedman AC, et al: Renal oncocytoma: New observations. Radiology 153:49–53, 1984. 256. Ramseyer LT: Case 34: Renal Lymphangiectasia. Radiology 219:442–444, 2001. 257. Rankin SC, Saunders AJ, Cook GJ, et al: Renovascular disease. Clin Radiol 55:1–12, 2000. 258. Rapezzi D, Racchi O, Ferraris AM: Perirenal extramedullary hematopoiesis in agnogenic myeloid metaplasia: MR imaging findings. AJR Am J Roentgenol 168:1388–1389, 1997. 259. Reiman TA, Siegel MJ, Shackelford GD: Wilms tumor in children: Abdominal CT and US evaluation. Radiology 160:501–505, 1986. 260. Reznek RH, Mootoosamy I, Webb JA, et al: CT in renal and perirenal lymphoma: A further look. Clin Radiol 42:233–238, 1990. 261. Rhyner P, Federle MP, Jeffrey RB: CT of trauma to the abnormal kidney. AJR Am J Roentgenol 142:747–750, 1984. 262. Richard S, Chauveau D, Chretien Y, et al: Renal lesions and pheochromocytoma in von Hippel-Lindau disease. Adv Nephrol Necker Hosp 23:1–27, 1994. 263. Ritchey ML, Azizkhan RG, Beckwith JB, et al: Neonatal Wilms tumor.   J Pediatr Surg 30:856–859, 1995. 264. Robson CJ, Churchill BM, Anderson W: The results of radical nephrectomy for renal cell carcinoma. J Urol 101:297–301, 1969. 265. Rodriguez Alonso A, Gonzalez Blanco A, Cespon Outeda E, et al: Subepithelial hematoma of kidney pelvis and ureter: Antopol-Goldman lesion. Actas Urol Esp 26:133–135, 2002 [in Spanish]. 266. Rodriguez-de-Velasquez A, Yoder IC, Velasquez PA, et al: Imaging the effects of diabetes on the genitourinary system. Radiographics 15:1051–1068, 1995. 267. Rofsky NM, Bosniak MA: MR imaging in the evaluation of small (<3.0 cm) renal masses. Magn Reson Imaging Clin N Am 5:67–81, 1997. 268. Rofsky NM, Weinreb JC, Bosniak MA, et al: Renal lesion characterization with gadolinium-enhanced MR imaging: Efficacy and safety in patients with renal insufficiency. Radiology 180:85–89, 1991. 269. Roubidoux MA, Dunnick NR, Sostman HD, et al: Renal carcinoma: Detection of venous extension with gradient-echo MR imaging. Radiology 182:269–272, 1992. 270. Roy C, Pfleger DD, Tuchmann CM, et al: Emphysematous pyelitis: Findings in five patients. Radiology 218:647–650, 2001. 271. Roychowdhary A, Makris J, Colby JM: Unilateral absence of the white pyramid sign on noncontrast CT: A sign of tubular hydronephrosis. Presented at the 98th meeting of the American Roentgen Ray Society, San Francisco, California, April 26–May 1, 1998. 272. Rubin GD, Dake MD, Napel S, et al: Spiral CT of renal artery stenosis: Comparison of three-dimensional rendering techniques. Radiology 190:181–189, 1994. 273. Rubin GD: Helical CT of potential living renal donors: Toward a greater understanding. Radiographics 18:601–604, 1998. 274. Rubin GD, Walker PJ, Dake MD, et al: Three-dimensional spiral computed tomographic angiography: An alternative imaging modality for the abdominal aorta and its branches. J Vasc Surg 18:656–664, 1993. 275. Sagalowsky AI, McConnell JD, Peters PC: Renal trauma requiring surgery: An analysis of 185 cases. J Trauma 23:128–131, 1983. 276. Sanders RC, Hartman DS: The sonographic distinction between neonatal multicystic kidney and hydronephrosis. Radiology 151:621– 625, 1984. 277. Sanders RC, Nussbaum AR, Solez K: Renal dysplasia: Sonographic findings. Radiology 167:623–626, 1988.

1860

PART II  CT and MR Imaging of the Whole Body

278. Saunders HS, Dyer RB, Shifrin RY, et al: The CT nephrogram: Implications for evaluation of urinary tract disease. Radiographics 15:1069–1085, 1995. 279. Schoenberg SO, Prince MR, Knopp MV, et al: Renal MR angiography. Magn Reson Imaging Clin N Am 6:351–370, 1998. 280. Schuster TG, Ferguson MR, Baker DE, et al: Papillary renal cell carcinoma containing fat without calcification mimicking angiomyolipoma on CT. AJR Am J Roentgenol 183:1402–1404, 2004. 281. Schwab SJ, Bander SJ, Klahr S: Renal infection in autosomal dominant polycystic kidney disease. Am J Med 82:714–718, 1987. 282. Schwartz JM, Bosniak MA, Hulnick DH, et al: The use of computed tomography in the diagnosis of carcinoma of the renal pelvis causing ureteropelvic junction obstruction. Urol Radiol 9:204–209, 1988. 283. Sebastia MC, Perez-Molina MO, Alvarez-Castells A, et al: CT evaluation of underlying cause in spontaneous subcapsular and perirenal hemorrhage. Eur Radiol 7:686–690, 1997. 284. Segal AJ, Spataro RF, Linke CA, et al: Diagnosis of nonopaque calculi by computed tomography. Radiology 129:447–450, 1978. 285. Semelka RC, Hricak H, Stevens SK, et al: Combined gadoliniumenhanced and fat-saturation MR imaging of renal masses. Radiology 178:803–809, 1991. 286. Semelka RC, Shoenut JP, Kroeker MA, et al: Renal lesions: Controlled comparison between CT and 1.5-T MR imaging with nonenhanced and gadolinium-enhanced fat-suppressed spin-echo and breath-hold FLASH techniques. Radiology 182:425–430, 1992. 287. Sessa A, Ghiggeri GM, Turco AE: Autosomal dominant polycystic kidney disease: Clinical and genetic aspects. J Nephrol 10:295–310, 1997. 288. Sherman JL, Hartman DS, Friedman AC, et al: Angiomyolipoma: Computed tomographic-pathologic correlation of 17 cases. AJR Am J Roentgenol 137:1221–1226, 1981. 289. Sheth S, Fishman EK: Multi-detector row CT of the kidneys and urinary tract: Techniques and applications in the diagnosis of benign diseases. Radiographics 24:e20, 2004. 290. Sheth S, Scatarige JC, Horton KM, et al: Current concepts in the diagnosis and management of renal cell carcinoma: Role of multidetector CT and three-dimensional CT. Radiographics 21(Spec No):S237–S254, 2001. 291. Shinmoto H, Yuasa Y, Tanimoto A, et al: Small renal cell carcinoma: MRI with pathologic correlation. J Magn Reson Imaging 8:690–694, 1998. 292. Siegel MJ: MR imaging of pediatric abdominal neoplasms. Magn Reson Imaging Clin N Am 8:837–851, 2000. 293. Si-Hoe CK, Thng CH, Chee SG, et al: Abdominal computed tomography in systemic lupus erythematosus. Clin Radiol 52:284–289, 1997. 294. Silverman SG, Lee BY, Seltzer SE, et al: Small (≤3 cm) renal masses: Correlation of spiral CT features and pathologic findings. AJR Am J Roentgenol 163:597–605, 1994. 295. Silverman SG, Mortele KJ, Tuncali K, et al: Hyperattenuating renal masses: Etiologies, pathogenesis, and imaging evaluation. Radiographics 27:1131–1143, 2007. 296. Singh AD, Shields CL, Shields JA: Von Hippel-Lindau disease. Surv Ophthalmol 46:117–142, 2001. 297. Slywotzky CM, Bosniak MA: Localized cystic disease of the kidney.   AJR Am J Roentgenol 176:843–849, 2001. 298. Smith SJ, Bosniak MA, Megibow AJ, et al: Renal cell carcinoma: Earlier discovery and increased detection. Radiology 170:699–703, 1989. 299. Smith RC, Levine J, Rosenfeld AT: Helical CT of urinary tract stones: Epidemiology, origin, pathophysiology, diagnosis, and management. Radiol Clin North Am 37:911–952, 1999. 300. Smith RC, Rosenfield AT, Choe KA, et al: Acute flank pain: Comparison of non-contrast-enhanced CT and intravenous urography. Radiology 194:789–794, 1995. 301. Smith RC, Verga M, Dalrymple N, et al: Acute ureteral obstruction: Value of secondary signs of helical unenhanced CT. AJR Am J Roentgenol 167:1109–1113, 1996.

302. Solak A, Gür MS, Genç B, et al: Localized cystic disease of the kidney: A rare cause of hypertension in a young adult. J Clin Imaging Sci 3:33, 2013. 303. Soulen MC, Fishman EK, Goldman SM, et al: Bacterial renal infection: Role of CT. Radiology 171:703–707, 1989. 304. Srigley JR, Eble JN: Collecting duct carcinoma of kidney. Semin Diagn Pathol 15:54–67, 1998. 305. Srinivas V, Sogani PC, Hajdu SI, et al: Sarcomas of the kidney. J Urol 132:13–16, 1984. 306. Stables DP, Fouche RF, de Villiers van Niekerk JP, et al: Traumatic renal artery occlusion: 21 cases. J Urol 115:229–233, 1976. 307. Steffens JC, Link J, Grassner J, et al: Contrast-enhanced, K-spacecentered, breath-hold MR angiography of the renal arteries and the abdominal aorta. J Magn Reson Imaging 7:617–622, 1997. 308. Steiner MS, Goldman SM, Fishman EK, et al: The natural history of renal angiomyolipoma. J Urol 150:1782–1786, 1993. 309. Strotzer M, Lehner KB, Becker K: Detection of fat in a renal cell carcinoma mimicking angiomyolipoma. Radiology 188:427–428, 1993. 310. Stuck KJ, Koff SA, Silver TM: Ultrasonic features of multicystic dysplastic kidney: Expanded diagnostic criteria. Radiology 143:217–221, 1982. 311. Sullivan M, Frydenberg M: Nephron-sparing surgery for small incidental renal cell carcinoma. Aust N Z J Surg 71:349–353, 2001. 312. Sundaram CP, Saltzman B: Urolithiasis associated with protease inhibitors. J Endourol 13:309–312, 1999. 313. Suzuki K, Kurokawa S, Muraishi O, et al: Segmental multicystic dysplastic kidney in an adult woman. Urol Int 66:51–54, 2001. 314. Szolar DH, Kammerhuber F, Altziebler S, et al: Multiphasic helical CT of the kidney: Increased conspicuity for detection and characterization of small (<3-cm) renal masses. Radiology 202:211–217, 1997. 315. Talner LB: Urinary obstruction. In Pollack HM, editor: Clinical Urography: An Atlas and Textbook of Urological Imaging, Philadelphia, 1990, WB Saunders. 316. Taylor RS, Joseph DB, Kohaut EC, et al: Renal angiomyolipoma associated with lymph node involvement and renal cell carcinoma in patients with tuberous sclerosis. J Urol 141:930–932, 1989. 317. Telenti A, Torres VE, Gross JB, Jr, et al: Hepatic cyst infection in autosomal dominant polycystic kidney disease. Mayo Clin Proc 65:933–942, 1990. 318. Thibaut H, Parizel PM, Van Goethem J, et al: Tuberous sclerosis: CT and MRI characteristics. Eur J Radiol 16:176–179, 1993. 319. Thoeny HC, De Keyzer F: Diffusion-weighted MR imaging of native and transplanted kidneys. Radiology 259(1):25–38, 2011. 320. Thomsen HS: European Society of Urogenital Radiology: European Society of Urogenital Radiology guidelines on contrast media application. Curr Opin Urol 17:70–76, 2007. 321. Thomsen HS, Marckmann P, Logager VB: Enhanced computed tomography or magnetic resonance imaging: A choice between contrast medium-induced nephropathy and nephrogenic systemic fibrosis? Acta Radiol 48:593–596, 2007. 322. Thornbury JR: Acute renal infections. Urol Radiol 12:209–213, 1991. 323. Thornton FJ, Kandiah SS, Monkhouse WS, et al: Helical CT evaluation of the perirenal space and its boundaries: A cadaveric study. Radiology 218:659–663, 2001. 324. Tolia BM, Iloreta A, Freed SZ, et al: Xanthogranulomatous pyelonephritis: Detailed analysis of 29 cases and a brief discussion of atypical presentations. J Urol 126:437–442, 1981. 325. Tomlinson IP, Alam NA, Rowan AJ, et al: Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 30:406–410, 2002. 326. Toro JR, Glenn G, Duray P, et al: Birt-Hogg-Dubé syndrome: A novel marker of kidney neoplasia. Arch Dermatol 135:1195–1202, 1999. 327. Torres VE, Bjornsson J, Zincke H: Inherited renal neoplasms. J Nephrol 11:229–238, 1998. 328. Torres VE, Wilson DM, Hattery RR, et al: Renal stone disease in autosomal dominant polycystic kidney disease. Am J Kidney Dis 22:513–519, 1993.

CHAPTER 54  Kidney 329. Traubici J, Neitlich JD, Smith RC: Distinguishing pelvic phleboliths from distal ureteral stones on routine unenhanced helical CT: Is there a radiolucent center? AJR Am J Roentgenol 172:13–17, 1999. 330. Trillo A, Lorentz WB, Whitley NO: Malakoplakia of kidney simulating renal neoplasm. Urology 10:472–477, 1977. 331. Tynski Z, MacLennan GT: Renal pseudotumors. J Urol 173:600, 2005. 332. Urban BA, Fishman EK: Renal lymphoma: CT patterns with emphasis on helical CT. Radiographics 20:197–212, 2000. 333. Urban BA, Fishman EK: Tailored helical CT evaluation of acute abdomen [CME available in print version and on RSNA Link]. Radiographics 20:725–749, 2000. 334. Uzzo RG, Novick AC: Nephron sparing surgery for renal tumors: Indications, techniques and outcomes. J Urol 166:6–18, 2001. 335. van Baal JG, Smits NJ, Keeman JN, et al: The evolution of renal angiomyolipomas in patients with tuberous sclerosis. J Urol 152:35–38, 1994. 336. Vasbinder GB, Nelemans PJ, Kessels AG, et al: Diagnostic tests for renal artery stenosis in patients suspected of having renovascular hypertension: A meta-analysis. Ann Intern Med 135:401–411, 2001. 337. Vinocur L, Slovis TL, Perlmutter AD, et al: Follow-up studies of multicystic dysplastic kidneys. Radiology 167:311–315, 1988. 338. Vourganti S, Agarwal PK, Bodner DR, et al: Ultrasonographic evaluation of renal infections. Radiol Clin North Am 44:763–775, 2006. 339. Vrtiska TJ, King BF, LeRoy AJ, et al: CT urography: Description of a novel technique using a modified multi-detector-row CT scanner [abstract]. Radiology 217(P):225, 2000. 340. Wagner BJ: The kidney: Radiologic-pathologic correlation. Magn Reson Imaging Clin N Am 5:13–28, 1997. 341. Wagner BJ, Wong-You-Cheong JJ, Davis CJ, Jr: Adult renal hamartomas. Radiographics 17:155–169, 1997. 342. Wan YL, Lee TY, Bullard MJ, et al: Acute gas-producing bacterial renal infection: Correlation between imaging findings and clinical outcome. Radiology 198:433–438, 1996. 343. Wang LJ, Wong YC, Chen CJ, et al: CT features of genitourinary tuberculosis. J Comput Assist Tomogr 21:254–258, 1997. 344. Warren KE, Gidvani-Diaz V, Duval-Arnould B: Renal medullary carcinoma in an adolescent with sickle cell trait. Pediatrics 103:E22, 1999. 345. Washecka R, Hanna M: Malignant renal tumors in tuberous sclerosis. Urology 37:340–343, 1991. 346. Weiner SN, Bernstein RG: Renal oncocytoma: Angiographic features of two cases. Radiology 125:633–635, 1977. 347. Weiss GR, Margolin KA, Aronson FR, et al: A randomized phase II trial of continuous infusion interleukin-2 or bolus injection interleukin-2

348. 349. 350. 351.

352.

353.

354. 355.

356.

357. 358.

359.

360.

361. 362.

363.

1861

plus lymphokine-activated killer cells for advanced renal cell carcinoma. J Clin Oncol 10:275–281, 1992. White KS, Kirks DR, Bove KE: Imaging of nephroblastomatosis: An overview. Radiology 182:1–5, 1992. Wiesner C, Jakse G, Rohde D: Therapy of local recurrence of renal cell carcinoma. Oncol Rep 9:189–192, 2002. Wilms’ tumor: Status report, 1990: By the National Wilms’ Tumor Study Committee. J Clin Oncol 9:877–887, 1991. Wong WS, Moss AA, Federle MP, et al: Renal infarction: CT diagnosis and correlation between CT findings and etiologies. Radiology 150:201–205, 1984. Wong-You-Cheong JJ, Wagner BJ, Davis CJ: Transitional cell carcinoma of the urinary tract: Radiologic-pathologic correlation. Radiographics 18:123–142, 1998. Wu ZL, Zhou KR, Liao L: Renal vein thrombosis and selective arterial or venous thrombolytic therapy. J Thromb Thrombolysis 3:67–70, 1996. Yoder IC, Pfister RC, Lindfors KK, et al: Pyonephrosis: Imaging and intervention. AJR Am J Roentgenol 141:735–740, 1983. Yuh BI, Cohan RH: Different phases of renal enhancement: Role in detecting and characterizing renal masses during helical CT. AJR Am J Roentgenol 173:747–755, 1999. Zagoria RJ, Bechtold RE, Dyer RB: Staging of renal adenocarcinoma: Role of various imaging procedures. AJR Am J Roentgenol 164:363–370, 1995. Zagoria RJ, Dyer RB: Computed tomography of primary renal osteosarcoma. J Comput Assist Tomogr 15:146–148, 1991. Zagoria RJ, Dyer RB, Harrison LH, et al: Percutaneous management of localized emphysematous pyelonephritis. J Vasc Interv Radiol 2:156–158, 1991. Zagoria RJ, Wolfman NT, Karstaedt N, et al: CT features of renal cell carcinoma with emphasis on relation to tumor size. Invest Radiol 25:261–266, 1990. Zand MS, Strang J, Dumlao M, et al: Screening a living kidney donor for polycystic kidney disease using heavily T2-weighted MRI. Am J Kidney Dis 37:612–619, 2001. Zbar B, Tory K, Merino M, et al: Hereditary papillary renal cell carcinoma. J Urol 151:561–566, 1994. Zeman RK, Cronan JJ, Rosenfield AT, et al: Renal cell carcinoma: Dynamic thin-section CT assessment of vascular invasion and tumor vascularity. Radiology 167:393–396, 1988. Zimmerhackl LB, Rehm M, Kaufmehl K, et al: Renal involvement in tuberous sclerosis complex: A retrospective survey. Pediatr Nephrol 8:451–457, 1994.

55  Retroperitoneum Drew A. Torigian and Parvati Ramchandani

INTRODUCTION The retroperitoneum is the compartmentalized space located external to and predominantly posterior to the posterior parietal peritoneum. About half a century ago, the retroperitoneum was described as a “veritable jungle of strange things” and as a “hinterland of straggling mesenchyme, with vascular and nervous plexuses, weird embryonic rests and shadowy fascial boundaries”.260,661 Since then, a better understanding of the anatomy and pathology of the retroperitoneum has taken place, related largely to cross-sectional imaging techniques. In this chapter, we discuss the use of computed tomography (CT) and magnetic resonance imaging (MRI) for evaluation of the retroperitoneum, the anatomy of the normal abdominal retroperitoneal and pelvic extraperitoneal spaces, and the imaging features of the pathologic processes that can occur in the retroperitoneum.

CT AND MRI OF THE RETROPERITONEUM CT and MRI both have advantages of high spatial resolution, multiplanar tomographic or volumetric image display, relatively good soft tissue contrast between normal structures and disease processes, a short examination time (generally <5-10 minutes for CT and <20-30 minutes for MRI), and the capability for whole-body imaging (related to multidetector row and parallel imaging technologies, respectively). As a result, they are the most widely used cross-sectional imaging modalities to assess the retroperitoneum.50,522,647,731,799 CT is frequently used for initial cross-sectional imaging evaluation of the retroperitoneum, particularly when the underlying etiology of a patient’s symptoms or signs is unclear. Although CT has higher spatial resolution than MRI, MRI has superior contrast resolution, does not use ionizing radiation, can be used when there are contraindications to obtaining an iodinated contrast-enhanced CT (e.g., renal insufficiency, balanced with concern for nephrogenic systemic fibrosis [NSF]) or contrast allergy), and allows for acquisition of multiple image sequences that may each be useful to depict different inherent characteristics of disease processes affecting the retroperitoneum.50,72,136,148,278,503 As such, if the diagnosis of a retroperitoneal malignancy is suspected, particularly when the pelvis is the primary site of disease, MRI may be used as the initial imaging test to potentially better delineate the nature and extent of disease.799,868 CT has superior sensitivity for the presence of calcification, ossification, or gas compared to that of MRI, is often more accessible than MRI at some institutions, and can be used when there are contraindications to MRI such as the presence of a transvenous pacemaker, intracranial ferromagnetic aneurysm clips, or orbital metallic foreign bodies.85,136,386,411,440,503,647,767 Despite these strengths, CT and MRI often cannot definitively characterize retroperitoneal lesions as benign or

1862

malignant based on morphologic, attenuation, signal intensity (SI), or enhancement features.146,431,616,647 However, both CT and MRI can be used to provide accurate guidance for percutaneous needle biopsy of retroperitoneal disease processes.275,647,799 CT examination of the retroperitoneum is typically performed using an oral contrast–enhanced (expediency now requires many studies to be performed without oral contrast, however) and intravenous (IV) contrast–enhanced abdominal and pelvic CT protocol with 5-mm axial sections extending from the lung bases to the symphysis pubis during the parenchymal/venous phase of enhancement. Precontrast images may be obtained to assess for the presence of calcification or ossification, macroscopic fat, hemorrhage, and cystic or necrotic change so as to serve as a baseline for assessment of enhancement after contrast administration and provide some diagnostic information if MRI or contrast-enhanced CT cannot be performed. Arterial-phase enhanced images may be obtained to characterize hypervascular retroperitoneal lesions such as paraganglioma, as well as disease processes involving arterial structures. Delayed- or excretory-phase enhanced images may at times be useful to assess for retroperitoneal disease processes that may involve or communicate with the urothelial tract.263,745,892 Acquired CT data sets may also be reconstructed retrospectively into sagittal, coronal, or oblique planes as needed, and three-dimensional (3D) volume-rendered images may also be created on a postprocessing workstation. Such images are useful for assessment of spatial relationships between lesions and surrounding anatomical structures, particularly if the masses are large or the organ of origin is unclear.213,615,737,957 Pulse sequences used for MRI examination of the retroperitoneum are similar to those of standard abdominal MRI and will only be  summarized here.193 T1-weighted images are useful for demonstrating high-SI-intensity fat or hemorrhage, lymphadenopathy, and vascular invasion by tumors. Fat-suppressed T2-weighted images are useful for demonstrating lymphadenopathy, muscle invasion by a disease process, cystic change or necrosis, fluid collections, bone marrow edema, and dilatation or obstruction of fluid-containing hollow structures such  as the biliary tree and gallbladder, bowel, or genitourinary (GU) system.502,799 Diffusion-weighted images (DWI) along with apparent diffusion coefficient (ADC) parametric maps are useful to detect lymphadenopathy and distant metastases.33,446,852 Precontrast, arterialphase, and parenchymal/venous-phase contrast-enhanced T1-weighted images may be obtained to provide information that is often useful to characterize retroperitoneal pathology.614 However, delayed-phase contrast-enhanced T1-weighted imaging is the most useful sequence for rapid screening of the extraparenchymal abdomen and pelvis because there is increased conspicuity of pathology. The solid or cystic/ necrotic nature of lesions, extent of disease, and presence and nature of vascular thrombosis or tumor encasement of abdominal vessels are demonstrated well on this sequence.148,193,502-504

CHAPTER 55  Retroperitoneum Abdominal Retroperitoneal Anatomy The abdominal retroperitoneum is bounded anteriorly by the posterior parietal peritoneum, posteriorly by the transversalis fascia, and extends craniocaudally from the diaphragm to the pelvic brim.553 It is divided by fascial planes into three compartments: the anterior pararenal space, the perirenal (or perinephric) space, and the posterior pararenal space (Figs. 55-1 to 55-4).425,556 The anterior pararenal space contains the retroperitoneal portions of the colon, duodenum, and pancreas, is continuous with the transverse mesocolon and root of the small bowel mesentery, and is confined by the posterior parietal peritoneum anteriorly, the anterior  renal fascia posteriorly, and the lateroconal fascia laterally (see Fig. 55-4).425,553,556,630 The junction of the anterior renal fascia, posterior renal fascia, and lateroconal fascia is most often located lateral to the kidney and tends to be located more posteriorly and medially at the craniocaudal level inferior to the renal hilum.425,698 Although the anterior pararenal space is potentially continuous across the midline, fluid collections are generally confined to their side of origin without crossing the midline, with exception of pancreatic fluid collections related to the presence of pancreatic enzymes.553 The perirenal space contains the kidney, renal vessels, adrenal gland, renal pelvis, proximal ureter, perirenal lymphatics, and perirenal fat and is confined by the anterior renal fascia and posterior renal fascia that comprise the renal fascia (also called Gerota’s fascia). It has the shape of a cone superiorly and an inverted cone inferiorly.118,425,553,556 Medially the anterior renal fascia blends into the dense connective tissue surrounding the great vessels in the root of the mesentery posterior to the pancreas and duodenum (which may leave a potential communication between the perirenal space and the central prevertebral space), and the posterior renal fascia fuses with the psoas or quadratus lumborum fascia.425,553,700 Laterally the anterior and posterior renal fascia fuse posterior to the ascending or descending colon to form the lateroconal fascia, which continues anteriorly around the flank to merge with the peritoneal reflection to form the paracolic gutter.553,698 Superiorly the posterior renal fascia fuses with the posterolateral hemidiaphragm, and the right anterior renal fascia merges with the right inferior coronary ligament. The anterior renal fascia also extends cranially to the diaphragm but may be deficient superiorly on the right side posterior to the bare area of the liver. As a result, lacerations of the liver involving the hepatic capsule adjacent to this bare area may be associated with hemorrhage that extends posteriorly into the right perirenal space. Furthermore, perirenal fluid collections may extend from the right perirenal space anteriorly to the bare area of the liver (Fig. 55-5).485,486,488,555,556,558 Similarly the splenorenal ligament may serve as a conduit for hemorrhage from the bare area of the spleen to the left anterior pararenal space in the setting of splenic trauma.43,781 The perirenal space is contiguous superiorly with the mediastinum through splanchnic foramina of the diaphragmatic crura and through small transdiaphragmatic perforations and lymphatic vessels, providing conduits of potential disease spread between the thorax and abdomen.57,853,922 The perirenal space is further divided into multiple compartments by fibrous lamellae, the bridging septa, which are of three types: those that arise from the renal capsule and extend to the renal fascia, those that are attached only to the renal capsule and are arranged more or less parallel to the renal surface, and those that connect the anterior renal fascia to the posterior renal fascia.442 These septa can exert an important influence on pathologic processes by either limiting the spread of disease or by serving as bidirectional conduits for spread of disease between the renal fascia, perirenal space, and central prevertebral space (Fig. 55-6; also see Fig. 55-2).262,442,700

1863

The posterior renal fascia can further be separated into two layers: the anterior and posterior laminae. The anterior lamina extends anteriorly and is continuous with the anterior renal fascia, the posterior lamina extends anterolaterally and is continuous with the lateroconal fascia, and multiple septa extend between the lateroconal and  renal fascia. As such, a potential space contiguous with the anterior pararenal space exists between these two laminae of the posterior renal fascia that can accumulate fluid collections, especially when related to pancreatitis.525,698 The posterior pararenal space almost always contains only fat, is confined by the posterior renal fascia anteriorly, by the transversalis fascia posteriorly, by the psoas muscle medially, and continues laterally external to the lateroconal fascia as the properitoneal fat of the abdominal wall. Inferiorly the posterior pararenal space is open to the pelvis, and superiorly it continues as a thin subdiaphragmatic layer of extraperitoneal fat.425,553,556 The retroperitoneal fascia are not composed of single membranes but are laminated owing to variable embryonic fusion of discrete layers of dorsal mesenteries. As a result, whereas most slowly accumulating fluid collections are confined to the retroperitoneal compartment in which they originate, these potentially expansile fascial planes may serve as routes of decompression for rapidly accumulating retroperitoneal fluid or gas collections, or as conduits for the spread of inflammatory or neoplastic disease to other abdominal or pelvic compartments (Figs. 55-7 and 55-8; also see Figs. 55-1 and 55-6).7,177,463,564,571 In addition, retroperitoneal fascial layers may not be intact in all patients owing to congenital variation, traumatic or iatrogenic disruption, or dissolution due to infected fluid or enzymatic digestion, allowing  for fluid collections and other pathologies to involve multiple retroperitoneal or peritoneal compartments. On CT and MRI, when a fluid collection abuts a fascial plane, the margin tends to be linear or curvilinear and sharply defined, whereas the interface between fluid and retroperitoneal fat usually has an uneven and feathery or spiculated appearance.425 The anterior interfascial retromesenteric plane is a laminated potentially expansile interfascial plane between the anterior pararenal space and the perirenal spaces. It is continuous across the midline and communicates with two other potentially expansile interfascial planes, the posterior interfascial retrorenal plane and the lateroconal plane, at the fascial trifurcation, providing potential routes for disease spread. At a level inferior to the origin of the superior mesenteric artery, retroperitoneal fluid collections can occasionally cross the midline anterior to the abdominal aorta and inferior vena cava (IVC) within the anterior interfascial retromesenteric plane.7,412,571 The posterior interfascial retrorenal plane is a complex potential compartment located between the perirenal space and the posterior pararenal space. Posterior interfascial fluid, peritoneal fluid, or posteriorly dissecting fluid in the anterior pararenal space may reside within this plane. The lateroconal interfascial plane is the potentially expansile space between layers of the lateroconal fascia.7,571 The normal thickness of the retroperitoneal fascial planes is 1 to 3 mm.425,553,807 On CT and MRI, retroperitoneal fascial planes are more frequently detected when there is an abundance of retroperitoneal fat. In general the posterior renal fascia is more often seen than the anterior renal fascia, and the anterior renal fascia is more commonly seen on the left side than on the right.425,553 Fascia that is focally thickened or greater than 3 mm in width is considered abnormal and can be caused by a large variety of pathologic conditions involving the retroperitoneal organs or nonparenchymal retroperitoneum.262,651,653 The anterior and posterior renal fascia fuse inferiorly in the iliac fossa to form a single multilaminar fascia called the combined interfascial plane.7,571 The combined interfascial plane continues in the pelvis

1864

PART II  CT and MR Imaging of the Whole Body Pancreas

Colon

LCP

RMP

Ascending ARF PP colon

*

Descending colon

Pancreas

APS

*

RMP

SFP II

II RRP

A

IVC

*

K

PM

*

I

IV LCF

QLM

PRS RRS

PRF

CP

PPS

TF

TF

DPS

B

A

Colon

LCP

SFP

Liver Pancreas

CIP

TF

RRP

RMP K

I

V IV

RRP

*

RMP

PM

II

TF V

CIP Ilium

CP III

C

Ilium

D

PV

PS

PM

FIG 55-1  Normal retroperitoneal anatomy in axial and sagittal planes. A, Line drawing at level of renal hila shows that perirenal and lateroconal fascia are laminated planes composed of apposed layers of embryonic mesentery. By dissecting layers, rapidly accumulating fluid collections or infiltrating diseases may spread within retroperitoneal fascial planes. Note that thickness of interfascial planes is exaggerated to illustrate potentially expansile nature. Also note that perirenal spaces are closed medially (open arrows). Anterior interfascial retromesenteric plane is continuous across midline (bidirectional arrows). Anterior interfascial retromesenteric plane, posterior interfascial retrorenal space, and lateroconal space communicate at fascial trifurcation (arrow). Dorsal pleural sinus may extend inferior to lie posterolateral to posterior pararenal space and transversalis fascia. A, aorta; APS, anterior pararenal space; ARF, anterior renal fascia; DPS, dorsal pleural sinus; IVC, inferior vena cava; LCF, lateroconal fascia;  PP, parietal peritoneum; PPS, posterior pararenal space; PRF, posterior renal fascia; PRS, perirenal space; RMP, retromesenteric plane; RRS, retrorenal space; TF, transversalis fascia; *, posterior peritoneal recess. (Reprinted and modified with permission from the American Journal  of Roentgenology.7) B to D, Line drawings through abdomen at level of renal hilum (B), upper pelvis (C), and right abdomen in sagittal plane (D) demonstrate retroperitoneal spaces, interfascial planes, and subfascial plane as well as associated potential routes of spread of retroperitoneal fluid or other pathology (gray arrows). Interfascial planes consist of anterior interfascial retromesenteric plane (RMP), posterior interfascial retrorenal plane (RRP), lateroconal interfascial plane (LCP), and combined interfascial plane (CIP). Subfascial plane (SFP) communicates with posterior interfascial retrorenal plane via narrow communicating passageway (CP) that may sometimes correspond to location of lumbar triangles. CIP continues into prevesical space (PV) and presacral space (PS) in pelvis, serving as bidirectional conduit for extension of fluid or other pathology between abdomen and pelvis. Note potential route of spread of fluid or other pathology along septa of Kunin (*) between perirenal space (stippled area), RMP, and RRP. K, kidney; PM, psoas muscle; QLM, quadratus lumborum muscle; TF, transversalis fascia. Horizontally-stippled area, anterior pararenal space; cross-hatched area, posterior pararenal space. (Modified from Ishikawa K, et al: Classification of acute pancreatitis based on retroperitoneal extension: Application of the concept of interfascial planes. Eur J Radiol 60:445-52, 2006.)

CHAPTER 55  Retroperitoneum

A

A D

D

PM QLM

B A

C

D FIG 55-2  Normal retroperitoneal anatomy in 59-year-old man with esophageal carcinoma and right abdominal pain due to infectious colitis. A and B, Axial contrast-enhanced CT (A) and axial heavily T2-weighted (B) images through abdomen show thin anterior and posterior renal fascia (short black arrows) that divide retroperitoneum into anterior pararenal space, which contains ascending (As) and descending (Ds) colon as well as duodenum; perirenal spaces, which contain kidneys; and posterior pararenal spaces, which contain fat. Edema within anterior interfascial retromesenteric plane (short white arrows) crosses midline posterior to superior mesenteric vessels. Note thickened septa of Kunin (long black arrows) due to high SI edema, and high-SI infiltration of fat in right anterior pararenal space surrounding ascending colon, with thickening of right anterior perirenal fascia due to infectious colitis. PM, psoas muscle; QLM, quadratus lumborum muscle. C and D, Coronal (C) and left sagittal (D) heavily T2-weighted images through abdomen demonstrate longitudinal configuration of perirenal spaces surrounded by thin renal fascia (short black arrows). Note V-shaped adrenal glands within perirenal spaces located superior to kidneys on coronal image and relationship of left perirenal space to bare area of spleen on sagittal image.

1865

1866

PART II  CT and MR Imaging of the Whole Body

A R

A P R

A

P

B FIG 55-3  Normal retroperitoneal anatomy in 47-year-old woman with nausea, vomiting, and abdominal pain due to acute pancreatitis. A and B, Axial contrast-enhanced CT (A) and axial out-of-phase T1-weighted gradient-recalled echo (B) images through abdomen show anterior renal fascia/anterior interfascial retromesenteric plane (short arrows), posterior renal fascia/posterior interfascial retrorenal plane (long arrow), and lateroconal fascia/lateroconal interfascial plane (arrowhead) with intersection at fascial trifurcation that divide retroperitoneum into anterior pararenal space (A), perirenal space (R), and posterior pararenal space (P). Thickening of left retroperitoneal fascial planes and infiltration of fat within anterior pararenal space surrounding pancreatic head are related to inflammation secondary to acute pancreatitis.

along the anterolateral margins of the psoas muscles, contiguous with the pelvic extraperitoneal perivesical and presacral spaces.262 There is controversy in the literature as to whether the inferior aspect of the perirenal space communicates openly with the pelvis; it is uncommon for perirenal fluid collections to extend into the pelvis or vice versa.412,529,564,623,699 However, the combined interfascial plane can serve as a conduit for disease extension from the abdominal retroperitoneum into the pelvis (see Figs. 55-4, 55-6, and 55-7).7,699,700 Fluid can also extend posteriorly from the posterior interfascial retrorenal plane through a cleft at the lumbar triangle between the medial aspect of the posterior pararenal space and the lateral aspect of the quadratus lumborum fat pad to the transversalis fascia, which accounts for the occurrence of flank discoloration (Grey Turner’s sign) in patients with acute pancreatitis.554 Furthermore, fluid within the posterior interfascial retrorenal plane can extend posteriorly through a narrow communicating passageway into the subfascial plane between the posterior pararenal space anteriorly and the transversalis fascia posteriorly at a level separate from that of the lumbar triangles, also potentially leading to Grey Turner’s sign (see Figs. 55-6 and 55-7).351,350 At the level of the iliac crest, inferior to the inverted cone of the renal fascia, the anterior and posterior pararenal spaces may communicate where the lateroconal fascia disappears as a distinct boundary. As a result, there is communication of these abdominal retroperitoneal spaces with the extraperitoneal pelvic spaces inferiorly and the properitoneal fat anterolaterally, serving as a multidirectional conduit for the spread of retroperitoneal pathology.56,550,553,623 In patients with large amounts of fluid in the infrarenal retroperitoneal space, extension into the pelvic extraperitoneal space is frequent. The most common pathways are posterior extension medial to the iliac vessels and/or a more medial extension into the prevesical space. Posterior extension lateral to the iliac vessels is less common.6,529,699

Pelvic Extraperitoneal Anatomy The pelvic extraperitoneal connective tissue is organized into groups of fascia that are fused together to demarcate the pelvic extraperitoneal spaces (Fig. 55-9).34,202 The parietal pelvic fascia (PPF) is the

variable dense fascial system that covers the structures limiting the pelvic cavity, including the levator ani, obturator, coccygeus, and piriformis muscles, the anterior surfaces of the sacrum and coccyx, and structures contiguous to the pelvic walls such as internal iliac vessels and sacral roots.202 The visceral pelvic fascia (VPF) is the fascial system derived from the visceral reflection of the PPF. It envelops the pelvic organs and attaches them to the pelvic walls.202 The VPF is composed of two symmetric layers oriented in the sagittal plane and courses along the lateral borders of the pelvic organs and medial borders of the vessels, thinning as it extends anteriorly. On CT and MRI it is best seen posteriorly where it is the thickest; the appearance is that of thin continuous lines enveloping the perirectal fat (Fig. 55-10). An additional group of fascia is oriented in the coronal plane that originates from the genital artery sheaths and from the inferolateral extensions of the peritoneal layer of the cul-de-sac. The anterior layer separates the urinary bladder and genital tract, and the posterior layer separates the genital tract from the rectum. These fascial planes are not well seen on CT or MRI owing to a paucity of surrounding fat.553 The extraserosal pelvic fascia is the variably dense connective tissue between the PPF and VPF in which the pelvic organs are embedded, which provides support to the vascular and neural structures of the pelvis and permits independent physiologic volume changes of the pelvic organs.202 The umbilicovesical fascia (UVF) is located anterior to the parietal peritoneum and posterior to the transversalis fascia, has a triangular configuration with its apex at the umbilicus, courses around the anterior aspect of the urachus and urinary bladder, and inferiorly blends with the pubovesical ligament medially and the VPF and PPF laterally (Fig. 55-10; also see Fig. 55-9).34,425 Its superior edges are occupied by the obliterated umbilical arteries that extend anteriorly from the internal iliac arteries, are covered by parietal peritoneum, and are visible as thin linear structures on CT and MRI in the location of the medial umbilical folds (see Fig. 55-10).34 The obliterated urachus is also often visible in the axial plane in the location of the median umbilical fold.34,553 Inferior to the peritoneal reflection the UVF extends Text continued on p. 1872

A

B

C

D

E FIG 55-4  Normal retroperitoneal anatomy in 34-year-old woman with acute diverticulitis of ascending colon. A and B, Axial contrast-enhanced CT images through lower abdomen (A) and upper pelvis (B) demonstrate infiltration of fat in right anterior pararenal space surrounding ascending colon, with thickening of anterior renal fascia (arrow), posterior renal fascia (long arrow), lateroconal fascia (arrowheads), and combined interfascial plane (short arrow). C and D, Sequential coronal contrast-enhanced CT images through abdomen again show infiltration within right anterior pararenal space. Note sparing of adjacent perirenal fat medial to thickened right anterior renal fascia (arrow) and sparing of adjacent posterior pararenal fat lateral to thickened lateroconal fascia (arrowheads). E, Sagittal contrast-enhanced CT image through right abdomen and upper pelvis reveals relationship between anterior renal fascia (arrow), posterior renal fascia (long arrow), and combined interfascial plane (short arrows). Note that level of fusion of anterior and posterior renal fascia to form combined interfascial plane is located superior to iliac crest in this particular patient. Again note thickening of anterior renal fascia and combined interfascial plane due to acute diverticulitis with sparing of perirenal fat and posterior pararenal fat.

1868

PART II  CT and MR Imaging of the Whole Body

FIG 55-5  Normal retroperitoneal anatomy in 82-year-old man status post motor vehicle crash. Axial contrast-enhanced CT image through upper abdomen demonstrates high-attenuation hemorrhage within superior right perirenal space surrounding adrenal gland and IVC extending to bare area of liver (short arrows) due to deficiency in superior aspect of right anterior renal fascia.

A

A

P

P

B

A

FIG 55-6  Normal retroperitoneal anatomy in 43-year-old woman

*

C

with drop in hematocrit level following percutaneous renal biopsy. A to C, Sequential axial unenhanced CT images through midabdomen (A), lower abdomen (B), and upper pelvis (C) show highattenuation hemorrhage within left renal subcapsular space and left perirenal space, with extension along septa of Kunin (arrowheads) into anterior and posterior renal fascia, as well as extension into adjacent anterior pararenal space and anterior interfascial retromesenteric plane (A) and posterior pararenal space and  posterior interfascial retrorenal plane (P). Note extension of  hemorrhage inferiorly to combined interfascial plane (short arrows) and posteriorly into communicating passageway (*) to subfascial plane inferior to location of inferior lumbar triangle.

CHAPTER 55  Retroperitoneum

S

*

A

B

C FIG 55-7  Normal retroperitoneal anatomy in 61-year-old man with retroperitoneal hemorrhage following mycotic thoracoabdominal aortic aneurysm repair. A to C, Sequential axial unenhanced CT images through lower abdomen (A) and upper pelvis (B and C) demonstrate high-attenuation hemorrhage within left posterior interfascial retrorenal plane (long arrows) extending posteriorly through communicating passageway (*) into subfascial plane (S) and inferiorly into combined interfascial plane (short arrows).

A

B FIG 55-8  Normal retroperitoneal anatomy in 80-year-old man with duodenal perforation. A and B, Axial unenhanced CT images through abdomen show complex fluid with very-high-attenuation foci of barium contrast material within anterior interfascial retromesenteric plane (arrows) and posterior interfascial retrorenal plane (long arrows) related to duodenal perforation.

1869

1870

PART II  CT and MR Imaging of the Whole Body

UVF S

C

B U R

A

UVF

Prevesical space R

U B

Perivesical space

B FIG 55-9  Normal extraperitoneal pelvic anatomy in axial and sagittal planes. A, Line drawing through pelvis in axial plane demonstrates large fluid collection (black area) distending and enlarging prevesical space producing characteristic “molar-tooth” configuration. B, Line drawing through pelvis in sagittal plane shows relationship of umbilicovesical fascia, prevesical space, and perivesical space. B, bladder; c, cecum; R, rectum; S, sigmoid colon; U, uterus, UVF, umbilicovesical fascia; dotted line, peritoneum. (Modified with permission from Auh YH, et al: Extraperitoneal paravesical spaces: CT delineation with US correlation. Radiology 159:319-328, 1986; and Korobkin M, et al: CT of the extraperitoneal space: Normal anatomy and fluid collections. AJR Am J Roentgenol 159:933-942, 1992.)

CHAPTER 55  Retroperitoneum

1871

B

*

* R

B

A

U B

R U

C

R

D

FIG 55-10  Normal extraperitoneal pelvic anatomy in 64-year-old U R B

E

man with melanoma (A), 75-year-old man with renal cell carcinoma (B), 33-year-old woman with renal anomaly (C), and 50-yearold woman status post trauma (D, E). A, Axial contrast-enhanced CT image through pelvis shows thin visceral pelvic fascia (short arrows) with normally increasing thickness posteriorly. B, Axial contrast-enhanced CT image demonstrates thin perirectal fascia (long arrows). C, Axial T2-weighted image shows obliterated umbilical arteries (arrows) delineating superior margin of umbilicovesical fascia as well as thin perirectal fascia (long arrows). D and E, Axial and sagittal contrast-enhanced maximum-intensity projection (MIP) CT images demonstrate thin umbilicovesical fascia (arrowheads) and obliterated umbilical artery (arrow) delineating its superior margin. B, bladder; R, rectum; U = uterus; *, seminal vesicle.

1872

PART II  CT and MR Imaging of the Whole Body

to the pelvic floor on each side of the urinary bladder but is not well visualized on CT or MRI.553 The extraperitoneal fat of the pelvis is divided by the UVF into the prevesical space and the perivesical space.34,425,553 The prevesical space lies anterior and lateral to the UVF, medial to the thin anterior portion of the VPF and posterior to the pubic bones where it is also called the retropubic space or space of Retzius (see Fig. 55-9).34,553 It is a large potential compartment that has multiple potential extensions and communications, allowing for bidirectional spread of fluid collections or other pathologies. It has posterior extensions on the lateral aspects of the urinary bladder and communicates with the presacral space. Superiorly it becomes thinner as it approaches the umbilicus and communicates with the abdominal retroperitoneal spaces; it is continuous with the rectus sheath (inferior to the semilunar line) and femoral sheaths and has extensions along the round ligaments, vasa deferentia, and spermatic cords.34,425,553 It may also communicate with the perivesical and perirectal spaces.623 Prevesical fluid collections typically have a “molar tooth” configuration in the axial plane on CT and MRI because of the orientation of the UVF (Fig. 55-11; also see Fig. 55-9). The “crown” of the tooth lies anterior to the urinary bladder between the UVF and transversalis fascia of the anterior abdominal wall and displaces the urinary bladder posteriorly. The “roots” extend posteriorly between the UVF and either

the peritoneum superiorly or the PPF inferiorly, displacing the urinary bladder medially if the roots are symmetric or laterally if the roots are asymmetric in size. As much as 2500 mL of fluid may occupy the prevesical space and its extensions without evidence of a palpable abdominal wall mass.34 Rectus sheath hematomas inferior to the arcuate line can also extend inferiorly into the prevesical space to displace and compress the pelvic viscera, and postpuncture hematomas in the femoral sheath can also expand rapidly to fill the prevesical space and subsequently spread to any communicating compartments.861 The perivesical space is a thin space bounded by the UVF and contains the urinary bladder, urachus, obliterated umbilical arteries, and a little fat (see Fig. 55-9). It is continuous with the seminal vesicles and prostate gland in males, and the supravaginal portion of the cervix and lower uterine segment in females.34 Perivesical fluid collections are typically small owing to compression by the UVF and may occasionally be confused with thickening of the bladder wall on CT or with fluid collections in the cul-de-sac when localized posteriorly in the perivesical space; in general, they are better characterized by MRI.553 The perirectal space, which contains the rectum, hemorrhoidal vessels, and fat, is limited by the posterior portion of the VPF, continues superiorly into the abdominal extraperitoneal space, and is continuous with the sigmoid mesocolon (Fig. 55-12).553 The Denonvilliers’ fascia comprises the anterior portion of the perirectal fascia and the posterior layer of the prostatic fascia.175,622,744

RETROPERITONEAL SARCOMA Soft tissue sarcomas are rare mesenchymal neoplasms, accounting for less than 1% of adult malignancies. About 15% of sarcomas originate within the retroperitoneum, 45% originate in the lower extremity, 15% in the upper extremity, 10% in the head and neck region, and the remainder in the abdominal wall and chest wall.256,765 Most retroperitoneal neoplasms are malignant, and one third of malignant retroperitoneal neoplasms are sarcomas, with an incidence rate of 1 to 2 cases per million per year.302,478 Retroperitoneal sarcomas may develop at any age, but most present in the sixth and seventh decades of life, occurring slightly more commonly in men. They are often of high histologic grade and have a mean

H

B V

FIG 55-11  Normal extraperitoneal pelvic anatomy in 51-year-old woman with abdominal pain while on anticoagulation therapy due to prevesical hematoma. Axial contrast-enhanced CT image through pelvis demonstrates large hematoma (H) in prevesical space with asymmetric “molar-tooth” configuration related to limitation by umbilicovesical fascia located just anterior to perivesical fat (arrowheads). Note hematocrit effect within hematoma due to dependent layering of cellular elements and nondependent layering of serum with focus of very high attenuation at boundary related to active extravasation of contrast material (long arrow), indicating active hemorrhage. Also note indistinct boundary and loss of intervening fat plane between hematoma and  right rectus muscle (arrow), indicating involvement of right rectus sheath. B, bladder; V, vagina.

R

FIG 55-12  Normal extraperitoneal pelvic anatomy in 80-year-old man with Clostridium difficile colitis. Axial unenhanced CT image through pelvis shows mild thickening of perirectal fascia, infiltration of fat  within perirectal space, and moderate thickening of rectum (R) due to C. difficile colitis.

CHAPTER 55  Retroperitoneum size of 17 cm at presentation.478,798 In a report by Gutierrez et al., 60.8% of retroperitoneal sarcomas were larger than 10 cm, and 21.2% were larger than 20 cm when diagnosed.272 As these reports indicate, retroperitoneal sarcomas are often very large before producing any symptoms or signs, leading to a delay in diagnosis and subsequent poor prognosis.55,272,543,668,798 Symptoms are related to compression of adjacent organs and can include early satiety, nausea, vomiting, constipation, urinary frequency, neurologic symptoms in the lower extremities, and pain in the flank, back, or radicular pain. Physical examination may demonstrate a nonspecific abdominal mass, increased abdominal girth, or leg edema.287,478,543 CT and MRI can define the extent of the primary tumor, evaluate for direct involvement of adjacent organs and vessels, and detect and delineate the extent of distant metastatic disease. Imaging features that suggest unresectability include extensive vascular involvement, peritoneal implantation, and distant metastatic disease.478 There are approximately 50 histologic subtypes of soft tissue sarcoma. In the retroperitoneum, they include, in descending order, liposarcoma (40%), leiomyosarcoma (30%), and undifferentiated pleomorphic sarcoma (UPS) (previously known as malignant fibrous histiocytoma) (15%).142,215,478,668 In contrast to its occurrence in the extremities, UPS is uncommonly found in the retroperitoneum.198,478 Sarcomas rarely develop from preexisting benign soft tissue tumors, with the exception of malignant peripheral nerve sheath tumors that can arise from neurofibromas, usually in patients with neurofibromatosis type 1 (NF1).256 CT and MRI may occasionally suggest a specific histologic diagnosis.616 For example, the presence of macroscopic fat within a large retroperitoneal mass favors the diagnosis of a well-differentiated liposarcoma, whereas caval involvement favors a leiomyosarcoma, particularly if cystic or necrotic intratumoral components and/or metastases are present.198,287,616 Similarly a large retroperitoneal mass that contains calcifications, extensive hemorrhage without fatty components, or central necrosis favors a diagnosis of UPS.198,616 Although crosssectional imaging cannot accurately predict the grade of a retroperitoneal sarcoma, visualization of tumor necrosis suggests the presence of a high-grade tumor and a poorer prognosis.270 Grading and staging of soft tissue sarcomas are essential for determining prognosis and planning therapy. Grading determines the degree of histologic malignancy of a sarcoma and is based on histologic parameters such as degree of cellularity, degree of necrosis, cellular pleomorphism, mitotic activity, and expansive or infiltrative and invasive growth, the first two of which are the most important. Molecular and cytogenetic analyses may also be used to differentiate various subtypes of retroperitoneal sarcoma, assess tumor biology, predict patient prognosis, and guide therapeutic interventions.817,863 Well-differentiated sarcomas are of low grade, and poorly differentiated sarcomas are of high grade. Well-differentiated low-grade  sarcomas rarely metastasize but can be locally aggressive, whereas higher-grade poorly differentiated sarcomas have a propensity to metastasize and exhibit locally aggressive behavior. However, there are also borderline lesions for which it is difficult to determine the malignant potential, and there are benign neoplastic and nonneoplastic lesions that morphologically appear malignant but follow a benign clinical course. Histologic examination can usually predict tumor aggressiveness, but follow-up imaging is often the ultimate arbiter of whether a resected sarcoma exhibits malignant behavior.256 Staging of retroperitoneal sarcomas is based on clinical, imaging, and histologic features. Retroperitoneal sarcomas are staged according to the GTNM classification (Table 55-1), where the G stage is based on the tumor grade, the T stage is based on lesion size and depth of the lesion relative to the fascia (“a” signifies that the lesion is entirely above the superficial fascia, whereas “b” indicates that there is invasion of the

1873

TABLE 55-1  AJCC GTNM Classification

and Stage Grouping of Soft Tissue Sarcomas189

Primary Tumor (T) TX Primary tumor cannot be assessed T0 No evidence of primary tumor Tumor 5 cm or less in greatest dimension* T1 T1a Superficial tumor T1b Deep tumor Tumor more than 5 cm in greatest dimension* T2 T2a Superficial tumor T2b Deep tumor *Note: Superficial tumor is located exclusively above the superficial fascia without invasion of the fascia; deep tumor is located either exclusively beneath the superficial fascia, superficial to the fascia with invasion of or through the fascia, or both superficial yet beneath the fascia. Regional Lymph Nodes (N) NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1* Regional lymph node metastasis *Note: Presence of positive nodes (N1) in M0 tumors is considered stage III. Distant Metastasis (M) M0 No distant metastasis M1 Distant metastasis

Stage IA Stage IB Stage IIA Stage IIB Stage III Stage IV

Primary Tumor

Regional Lymph Nodes

Distant Metastasis

Grade

T1a T1b T2a T2b T1a T1b T2a T2b T2a, T2b Any T Any T

N0 N0 N0 N0 N0 N0 N0 N0 N0 N1 Any N

M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M1

G1, GX G1, GX G1, GX G1, GX G2, G3 G2, G3 G2 G2 G3 Any G Any G

Histopathologic Grade (FNCLCC System Preferred) GX Grade cannot be assessed G1 Grade 1 G2 Grade 2 G3 Grade 3 From Edge SB, et al: Soft tissue sarcoma. In AJCC cancer staging manual, ed 7. New York, 2010, Springer, pp 291–298.

superficial fascia or that the lesion is entirely below the superficial fascia), the N stage is based on presence of regional lymph node metastases, and the M stage is dependent on presence of distant metastases. By definition, all intraabdominal sarcomas are “b” lesions. Overall, regional lymphadenopathy is uncommon in soft tissue sarcomas  (frequency < 4% at presentation), and less than one third of patients have metastases at presentation.189,256,817

1874

PART II  CT and MR Imaging of the Whole Body

Whole body 18F fluorodeoxyglucose (FDG) positron emission tomography (PET), often performed in combination with CT (and more recently in combination with MRI), is a powerful diagnostic tomographic imaging modality that provides metabolic information relevant to diagnosis, grading, staging, and therapeutic monitoring of patients with various cancers including retroperitoneal sarcoma.446-449,852,855 The sensitivity of FDG PET for primary sarcomas ranges from 74% to 100% and is greater for intermediate- and high-grade sarcomas than for low-grade sarcomas.507,613 Studies also suggest that FDG PET is reliable for separating high-grade from low-grade tumors and in determining the biological activity of a tumor. Furthermore, FDG PET allows for detection of malignant lymphadenopathy and distant metastatic disease during whole-body staging, as well as locally recurrent disease, although it does not allow for prediction of a specific histologic diagnosis.186,370,507,612,613,746 Complete surgical resection is the treatment of choice for primary and recurrent retroperitoneal sarcoma, but is often difficult to perform owing to frequent invasion of adjacent retroperitoneal structures.478,516,668,788,817,818 According to Herman et al., resectability rates for retroperitoneal sarcoma range between 38% and 100%.302 Concomitant resection of adjacent organs that are involved by tumor is performed to ensure clear surgical margins.55 Presence of satellite nodules around a primary retroperitoneal sarcoma seen on preoperative crosssectional imaging is associated with an increased risk of local recurrence and may alter the extent of resection to be performed.905 Chemotherapy for retroperitoneal soft tissue sarcoma is not generally effective, and adjuvant radiation therapy is limited by toxicity to adjacent intraabdominal structures.9,46,272,302,478,668,702,817,867 Radiation therapy may be useful to decrease the rate of onset of local recurrence of tumor following surgical removal.46,82,102,702,817,818,867,950 Lewis et al. reported on a study of 500 patients with retroperitoneal sarcoma. The median survival was 72 months for patients with primary disease, 28 months for those with local recurrence, and 10 months for those with metastasis; 5-year survival rates after complete resection ranged from 40% to 74%. In patients with either primary or locally recurrent tumor, survival was significantly decreased if there was unresectable disease, incomplete resection, and high histologic grade. The median survival was 103 months after complete resection as compared to a median survival of 18 months with incomplete resection. There was no significant difference in survival in patients with unresectable disease (e.g., distant metastases, peritoneal implants, extensive vascular involvement) compared to those with incompletely resected disease. These results suggest that patients with either primary disease or a local recurrence should have aggressive attempts at complete surgical resection, and that incomplete resection should be considered only for palliative symptom relief.478 Complete en bloc resection of retroperitoneal sarcoma including resection of involved adjacent structures as necessary remains the single most important independent prognostic factor for survival.120,272,478,668,798,817,884 Lewis et al. report a median survival after local recurrence of 60 months in resected patients as opposed to 20 months in unresected patients.478 Five-year survival rates after radical excision range from 55% to 92% for well-differentiated tumors, compared with 14% to 48% for less well-differentiated tumors.302,884 Most authors report high local recurrence rates for retroperitoneal sarcoma, and in a study of 204 patients with retroperitoneal sarcoma, Herman et al. demonstrated a local recurrence rate after complete resection of 40% after 2 years, 72% after 5 years, and 91% after 10 years.302 Early detection and aggressive surgical resection of recurrent tumors continue to form the basis of treatment of local recurrence, potentially resulting in long-term survival even with repeat recurrences.55,516 Patients generally undergo annual follow-up cross-sectional imaging because a delay in the appearance of some sarcoma

recurrences is well known, with some sarcomas of both low and high grades not recurring until more than 5 years after the initial surgery. Histologic grade is the next most important prognostic factor in determining local recurrence and overall survival.120,516,798,817,884 When sarcomas recur after surgery, patient survival is prolonged only in those for whom complete resection can be successfully performed or in those with low-grade tumors.269 Preoperative cross-sectional imaging in patients with recurrent abdominopelvic sarcoma is useful for treatment planning. Features that portend a poor prognosis include large (>5 cm) recurrent tumor involving the central pelvis, left lower quadrant of the pelvis, or distal jejunum, especially if the pelvic tumor also involves more than one segment of small bowel.662 The attenuation and SI features of recurrent sarcoma may differ from those of the original primary sarcoma.269 Distant metastases from retroperitoneal sarcoma occur mostly with high-grade tumors after a long time of primary tumor growth, with an incidence ranging from 15% to 34%.302,765 Resection of distant metastases with curative intent is the best manner of prolonging survival, especially in patients with a limited number of metastases, a long disease-free interval, slow clinical growth, and complete resection of the primary tumor.302,887

Liposarcoma Liposarcoma is the most common retroperitoneal sarcoma, accounting for approximately 40% of such cases, and is the second most common adult soft tissue sarcoma. UPS is the most common adult soft tissue sarcoma, accounting for 14% to 18% of all soft tissue sarcomas.198,402,478,813 Most liposarcomas occur in deep soft tissue, in contrast to lipomas, which occur more commonly in superficial soft tissue.256 Liposarcomas originate from primitive mesenchymal cells and not from adipocytes.813 The World Health Organization (WHO) divides liposarcomas into four subtypes: well-differentiated, myxoid, dedifferentiated, and pleomorphic.215 Liposarcomas can be conceptually divided into three subgroups. Well-differentiated and dedifferentiated liposarcomas comprise one subgroup, since over time a subset of welldifferentiated liposarcomas histologically can progress to dedifferentiated sarcomas, which have metastatic potential. A second subgroup is composed of the myxoid liposarcomas, with the continuum of lesions ranging from pure myxoid liposarcoma at one extreme to the more poorly differentiated round cell liposarcoma at the other. Finally, a third subgroup exhibits unusual features or combined patterns not accounted for in the other classifications.256 Retroperitoneal liposarcoma has a mean diameter of 20 cm and is associated with local recurrence, which is responsible for most patient morbidity and mortality.198,454,478,765 The rate of distant metastasis to both the liver and lung is less than 10%. In comparison, patients with nonliposarcoma retroperitoneal tumors have a fourfold higher risk of metastases compared with liposarcoma.798 In patients with unresectable retroperitoneal liposarcoma, particularly primary disease, incomplete or partial resection or debulking can provide improvement in survival and palliate symptoms, in contrast to patients with other types of retroperitoneal sarcoma. Partial resection increased survival as compared with exploration or biopsy alone (median survival, 26 months vs. 4 months), and successful palliation of symptoms was achieved in 75% with preoperative symptoms.765

Well-Differentiated Liposarcoma.  Well-differentiated liposarcoma is the most common subtype of retroperitoneal liposarcoma, reaching a peak incidence during the sixth and seventh decades of life, with men and women affected equally. Together the well-differentiated and dedifferentiated subtypes account for approximately 35% to 40% of liposarcomas. Grossly, well-differentiated liposarcomas are large

CHAPTER 55  Retroperitoneum

1875

* K

B A

D

C FIG 55-13  Retroperitoneal well-differentiated liposarcoma in 56-year-old man with bladder carcinoma. A, Axial contrast-enhanced CT image demonstrates large fat-attenuation mass involving right perirenal space causing anterior displacement of right kidney (K) and IVC (*). Note multiple soft tissue–attenuation mildly nodular septations within mass, as well as subtle soft tissue component laterally (arrow). B to D, Axial outof-phase T1-weighted GRE (B), axial fat-suppressed T2-weighted (C), and axial fat-suppressed contrastenhanced T1-weighted (D) images show loss of SI of fatty components of mass on fat-suppressed images and intermediate T1 and T2 SI of septations and soft tissue component (arrows) relative to skeletal muscle with enhancement.

multilobular lesions that could be mistaken for lipomas except for their extremely large size and their tendency to have more fibrous bands, gelatinous zones, or punctate hemorrhage.256 On CT, well-differentiated liposarcomas contain components that are similar in attenuation to macroscopic fat (Fig. 55-13).219,402,610 On T1- and T2-weighted images, well-differentiated liposarcomas are usually isointense to subcutaneous fat, with loss of SI on fat-suppressed imaging sequences.26,198,361,402,792,813 However, subtle differences in the attenuation or SI of normal abdominopelvic fat compared to areas of fat-containing liposarcoma may be seen, and in patients who have undergone surgical resection of liposarcoma, such findings must be viewed with suspicion for possible tumor recurrence.219,269,610 Streaky zones of fibrous or sclerotic components within a well-differentiated liposarcoma show SI similar to that of skeletal muscle (Fig. 55-14; also see Fig. 55-13).402 Well-differentiated liposarcoma demonstrates little or no contrast enhancement, which reflects

the paucity of intratumoral vessels.26 However, thickened irregular septa and minor nodular components within lipoma-like components may enhance, and sclerosing components may also enhance homogeneously.402 Areas of necrosis and calcification tend to be uncommon in all liposarcomas.198 Imaging features that favor well-differentiated liposarcoma over lipoma include large lesion size (>10 cm), presence of thick septa (>2 mm), nodular and/or globular areas, nonadipose masslike areas, and a decreased percentage of fat composition (<75% fat in mass). The presence of thick septa and associated nonadipose masslike areas increase the likelihood of a well-differentiated liposarcoma over  a lipoma by 9- and 32-fold, respectively.430 Large exophytic renal angiomyolipomas may mimic the appearance of perirenal welldifferentiated liposarcomas, but presence of a renal parenchymal defect, enlarged vessels within the lesion, and presence of other renal angiomyolipomas favor an angiomyolipoma.352

1876

PART II  CT and MR Imaging of the Whole Body

L

L

K

K

A

B

L

K

C

Well-differentiated liposarcomas are for the most part nonmetastasizing lesions that are of low-grade histology, but their rate of local recurrence in the retroperitoneum approaches 100%.919 With local recurrence, cachexia and intestinal obstruction are common.765 Approximately 10% of well-differentiated retroperitoneal liposarcomas can dedifferentiate after an average of 7 to 8 years.256

Dedifferentiated

Liposarcoma.  Dedifferentiated liposarcomas develop in approximately the same age group as well-differentiated liposarcomas, reaching a peak during the early seventh decade of life, with men and women affected equally.256,301 Histologically the lesions have areas of well-differentiated liposarcoma and a nonlipogenic (dedifferentiated) component that has an appearance of a high-grade fibrosarcoma or UPS.256 On CT and MRI, dedifferentiated liposarcomas have areas with attenuation and SI characteristic of fat-containing

FIG 55-14  Retroperitoneal well-differentiated liposarcoma in 62-year-old woman with palpable abdominal mass. A and B, Coronal in-phase (A) and out-of-phase (B) T1-weighted gradientrecalled echo images demonstrate large right retroperitoneal lesion causing mass effect upon lateral aspect of right lobe of liver (L) and right kidney (K), with associated medial displacement. Mass has intermediate SI relative to subcutaneous fat, with minimal loss of SI on out-of-phase image relative to in-phase image secondary to macroscopic fat, with more marked loss of SI on out-of-phase image in location of nonfatty septations (arrows). Latter phenomenon is due to phase cancellation of water and fat protons within voxels at interfaces of nonfatty septations with adjacent surrounding macroscopic fat. C, Coronal fat-suppressed T1-weighted image shows diffuse loss of SI of mass in keeping with macroscopic fat composition.

well-differentiated liposarcoma, but they have more masslike areas of nonfatty tissue with higher attenuation, as well as lower T1 and higher T2 SI (Fig. 55-15).319,432,792,833 Calcification or ossification may be seen in up to 30% of cases.833 The biological behavior of dedifferentiated liposarcomas is similar to that of other pleomorphic high-grade sarcomas in adults, with a 5-year survival rate of less than 25%.256,301,302 Up to 41% of patients may experience local recurrences, and 17% may experience metastases, with a 5-year survival rate as low as 16%.301,302 Overall, dedifferentiated retroperitoneal liposarcomas have the worst prognosis.256

Myxoid Liposarcoma.  Myxoid liposarcoma is the second most common subtype of retroperitoneal sarcoma.256,813 Unlike welldifferentiated liposarcomas, myxoid liposarcoma occurs in a younger age group, with a peak incidence during the fifth decade of life.256 The

CHAPTER 55  Retroperitoneum

1877

K

A

B

C

D FIG 55-15  Retroperitoneal mixed well-differentiated and dedifferentiated liposarcoma in 52-year-old woman with breast cancer. A, Axial contrast-enhanced CT image through abdomen shows large heterogeneously enhancing retroperitoneal mass with predominantly fatty attenuation well-differentiated component anteriorly and soft tissue attenuation dedifferentiated component posteriorly. Note soft tissue attenuation linear septations within fatty component anteriorly. Also note associated mass effect of lesion upon right lobe of liver and anterior displacement of right kidney (K). B, Axial out-of-phase T1-weighted gradient-recalled echo image demonstrates fatty component with SI similar to that of subcutaneous fat, and soft tissue component with SI similar to that of skeletal muscle. C, Axial fat-suppressed T2-weighted image reveals loss of SI of fatty component and heterogeneously increased SI of soft tissue component relative to skeletal muscle. D, Axial fat-suppressed contrast-enhanced T1-weighted image shows heterogeneous enhancement predominantly within soft tissue component of mass but also enhancement of linear soft tissue septations in fatty component.

myxoid subtype has an intermediate prognosis between those of   the well-differentiated and dedifferentiated subtypes. Histologically, myxoid liposarcomas are composed of a myxoid matrix containing mucopolysaccharides with small amounts of mature fat. Frequently the extracellular mucoid material forms large pools and often shows a lacelike pattern.26,813 Grossly the lesion is frequently lobulated, may be multiloculated or oval, and may be accompanied by surrounding soft tissue edema.813 A spectrum of features on CT and MRI may be present in myxoid liposarcomas, related to fat content, amount of myxoid material, degree of cellularity and vascularity, and presence of necrosis.813 On CT and MRI the myxoid dominant component of these tumors shows low attenuation, low SI on T1, and very high SI relative to muscle on T2 (Fig. 55-16).402,610,792,813 Foci of fat- and/or soft tissue–attenuation/ low-SI septa may be revealed within the predominately myxoid containing mass, and calcifications may also be seen.26,402,610,792,813 A potential pitfall is that a fatty component in the margin of a mass may simulate extraperitoneal fat and be missed.616

As many as 25% of myxoid liposarcomas may not exhibit foci of attenuation or SI typical of a fatty tumor, but may instead appear as completely cystic lesions on unenhanced images. The absence of foci with attenuation and SI similar to macroscopic fat makes a myxoid liposarcoma indistinguishable from most other soft tissue masses. In such cases the lesion can mimic a cyst, cystic degeneration, or necrosis on unenhanced CT (due to presence of water attenuation) as well as on T2-weighted images because of the presence of very high SI.616,813 However, the myxoid stroma will enhance to a variable degree depending on its degree of vascularity, and thus enhanced images can distinguish cystic and myxoid components of the neoplasm. Myxoid tissue may also be present within UPS, and in the absence of fat, the two lesions have a similar appearance.610,616

Pleomorphic Liposarcoma.  Pleomorphic liposarcoma is the least common subtype of retroperitoneal liposarcoma. Histologically there is a disorderly growth pattern with a marked degree of cellular pleomorphism, including bizarre giant cells.256 Pleomorphic liposarcomas

1878

PART II  CT and MR Imaging of the Whole Body

*

A

*

B

C FIG 55-16  Retroperitoneal myxoid liposarcoma in 76-year-old man with abdominal pain and palpable abdominal mass. A and B, Axial in-phase (A) and out-of-phase (B) T1-weighted gradient-recalled echo images show well-circumscribed heterogeneous left retroperitoneal mass that contains intermediate to slightly high SI components (*) relative to skeletal muscle that lose SI on opposed-phase image owing to microscopic lipid content within tumor. On out-of-phase image, low-SI etching artifact (arrows) present between tumor components indicates presence of fat-water interfaces within tumor. C, Axial fat-suppressed T2-weighted image demonstrates multiple septated high-SI foci relative to skeletal muscle within lesion representing myxoid tissue.

are heterogeneous nonfatty tumors with imaging characteristics indistinguishable from those of other malignant soft tissue masses.402 On CT they often contain little or fat, have predominant attenuation similar to muscle, may contain areas of low attenuation due to necrosis, and have variable enhancement.402,610 On MRI they appear heterogeneous on T1- and T2-weighted images and commonly reveal foci of intratumoral hemorrhage and necrosis.587,792 These tumors are highly malignant with a tendency toward local recurrence and metastasis.813

Leiomyosarcoma Leiomyosarcoma is the second most common retroperitoneal sarcoma (≈30% of retroperitoneal sarcomas) and accounts for about 8% of soft tissue sarcomas overall.198,429,478 Most leiomyosarcomas present in the fifth or sixth decade of life, with two thirds of all retroperitoneal leiomyosarcomas occurring in women.256,287 Approximately half of all soft tissue leiomyosarcomas develop in the retroperitoneum, making it the single most common soft tissue site. Retroperitoneal leiomyosarcomas are typically well circumscribed, have a mean diameter of 16 cm, and often contain intratumoral necrosis and hemorrhage. Histologically  a typical pattern of interlacing sweeping bundles of spindle-shaped cells with elongated blunt-ended nuclei is seen. Initially, adjacent organs are displaced without direct invasion but eventually do become

involved by direct extension. Two thirds of retroperitoneal leiomyosarcomas are located external to the lumen of the IVC, whereas approximately one third have both intraluminal and extraluminal components. Leiomyosarcoma is the most common intraluminal venous neoplasm and is the most common primary tumor of the IVC. The finding of a retroperitoneal mass that has both intraluminal and extraluminal components is very suggestive of a leiomyosarcoma.287 Purely intraluminal caval leiomyosarcomas account for 5% of leiomyosarcomas, occur primarily in women (80%-90% of patients), and present at a younger age (mean age, 50 years).287 On imaging, intracaval leiomyosarcomas are seen as polypoid or nodular masses that are firmly attached to the vessel wall. They are smaller than those that are entirely extravascular, are less likely to show intratumoral hemorrhage and necrosis, and are most frequently located between the diaphragm and renal veins. Leiomyosarcomas with an intraluminal component are more likely to produce early symptoms than those that are completely extraluminal. Patients with upper segment IVC involvement may develop symptoms and signs of Budd-Chiari syndrome. Patients with middle segment IVC involvement may develop right upper quadrant pain and tenderness, frequently mimicking biliary tract disease, and extension into the renal veins may cause a variable degree of renal insufficiency. Lower extremity edema occurs when there is lower

CHAPTER 55  Retroperitoneum

1879

A

B

C FIG 55-17  Retroperitoneal IVC leiomyosarcoma in 62-year-old woman with abdominal pain and elevated serum creatinine level. A, Axial fat-suppressed T2-weighted image shows expansion of intrahepatic IVC by high-SI tumor thrombus peripherally (long arrow) and low-intermediate SI bland thrombus centrally (arrow). B and C, Coronal (B) and axial (C) fat-suppressed contrast-enhanced T1-weighted image demonstrate cephalocaudal extent of thrombus within both intrahepatic and extrahepatic segments of IVC. Tumor thrombus enhances (long arrows) while bland thrombus does not enhance (arrow).

segment IVC involvement. Tumors that do not extend into or above the intrahepatic IVC are resectable, whereas those tumors that infiltrate the intrahepatic IVC, hepatic veins, right atrium, and beyond are often unresectable.148,287 On CT and MRI, leiomyosarcomas usually have attenuation similar to muscle, being low to intermediate SI on T1 and heterogeneous intermediate to high SI on T2 (Figs. 55-17 and 55-18).198,451,544 Central liquefactive necrosis is more common and extensive than in other sarcomas and is seen as foci with low attenuation, low T1 SI, and high T2 SI. Fat and calcification are not typically seen.198,287,544 Intratumoral hemorrhage typically appears as high attenuation and high T1-weighted SI relative to muscle. Smaller tumors may present as a solid nonnecrotic mass, and uncommonly, the tumor may present as a predominantly cystic mass with extensive necrosis. The enhancement of leiomyosarcomas is variable and depends on their muscular and fibrous components; it is usually delayed compared to enhancement of the surrounding skeletal muscles.287,451,544,610 Intraluminal bland thrombus and tumor thrombus can be differentiated with contrastenhanced CT and contrast-enhanced T1-weighted imaging; findings that favor tumor thrombus include enlargement of the vascular lumen and enhancement of the thrombus (see Fig. 55-17).147,198,287 As with other retroperitoneal sarcomas, complete surgical removal is the treatment of choice for retroperitoneal leiomyosarcoma and is

the most important factor affecting patient survival.287,515,568 Retroperitoneal leiomyosarcomas have a poor prognosis, with a 5-year survival rate of approximately 15%.565 Many patients have unresectable disease at the time of presentation secondary to infiltration of the intrahepatic IVC, hepatic veins, or right atrium, or metastatic disease, which is present in 40% of patients. Patients with resectable IVC leiomyosarcoma have the best prognosis, with two thirds of patients surviving beyond 5 years.308 Local recurrence after resection is seen in 40% to 77% of patients.287

Undifferentiated Pleomorphic Sarcoma UPS, previously known as malignant fibrous histiocytoma, is the most common soft tissue sarcoma in adults, with most patients presenting between the sixth and eighth decades of life. Two thirds of these tumors occur in men.50,256 UPS accounts for 25% of all soft tissue sarcomas, but only 16% arise in the retroperitoneum.256,429 Nevertheless it is the third most common retroperitoneal sarcoma after liposarcoma and leiomyosarcoma.198 On gross examination, retroperitoneal UPS appears as a solitary, multilobulated, large mass, often with hemorrhage and necrosis, sometimes with intratumoral calcification or ossification. Histologically there is typically a highly variable mixture of storiform and pleomorphic cellular elements.256

1880

PART II  CT and MR Imaging of the Whole Body

* M

M

B

A

*

L

M

K

C FIG 55-18  Retroperitoneal leiomyosarcoma in 51-year-old woman. A, Axial out-of-phase T1-weighted image shows homogeneous intermediate-SI retroperitoneal mass (M) located anterior to the right kidney that displaces and mildly compresses the IVC (*) anteriorly (which shows high-SI flow-related enhancement). Note posteriorly displaced right adrenal gland (arrow). B, Axial fat-suppressed T2-weighted image demonstrates that mass (M) has heterogeneous high SI relative to skeletal muscle and is well demarcated from adjacent right lobe of liver by thin rim of high SI. Note mildly prominent nonspecific periportal lymph nodes (long arrows) with slightly increased SI relative to skeletal muscle. C, Coronal fat-suppressed contrast-enhanced T1-weighted image shows that heterogeneously enhancing mass (M) is separate from liver (L) and right kidney (K), inferiorly displaces right kidney, and extrinsically compresses the IVC (*). At surgery, no hepatic, renal, or adrenal invasion was seen.

On CT and MRI, UPS appears as a large relatively well-circumscribed mass that spreads along fascial planes and between muscle fibers, with attenuation similar to muscle, low to intermediate SI on T1-weighted images, and heterogeneously increased SI on T2-weighted images relative to muscle (Fig. 55-19).198,454,610 Intratumoral fat is absent. Areas of low attenuation due to cystic degeneration or necrosis may be present.610 The “bowl of fruit” sign is a mosaic of mixed low, intermediate, and high SI on T2 that correlates to the presence of intratumoral solid components, cystic degeneration, hemorrhage, myxoid stroma, and fibrous tissue. Although the bowl of fruit sign is commonly revealed in UPS, it is not specific and has been described in other tumors such as synovial sarcoma and Ewing’s sarcoma; these latter subtypes only very rarely occur in the retroperitoneum, however. Extensive intratumoral hemorrhage frequently occurs in UPS, but most masses of UPS have nonhemorrhagic solid components, and thus most tumors can be differentiated from a bland benign hematoma. Intratumoral

calcifications may be seen in up to 20% on CT, often peripherally with a lumpy or ringlike configuration or sometimes more centrally with a speckled or amorphous configuration. Calcification is more difficult to detect prospectively on MRI.414,610,616

PRIMARY RETROPERITONEAL EXTRAGONADAL GERM CELL TUMOR Primary extragonadal germ cell tumors (EGCTs) represent between 1% and 3% of all germ cell tumors, occurring most commonly within the mediastinum and slightly less commonly within the retroperitoneum.77,123 Primary EGCTs are more common in men than women, with a peak occurrence in the fourth and fifth decades of life, which is a slightly older age group compared to patients who develop primary testicular germ cell neoplasms.77,123,198

CHAPTER 55  Retroperitoneum

1881

IP M

A

M

M

IP IP

C

B

FIG 55-19  Retroperitoneal malignant fibrous histiocytoma in 50-year-old woman with pain and palpable left lower quadrant mass. A, Axial T1-weighted image demonstrates infiltrative retroperitoneal mass (M) with intermediate SI relative to skeletal muscle that displaces left iliopsoas muscle (IP) and left common iliac artery and vein medially and extends into anterior left ischium and ilium (arrow). B and C, Coronal fatsuppressed T2-weighted (B) and fat-suppressed contrast-enhanced T1-weighted (C) images show high-T2-SI avidly enhancing myxoid components within superior and inferior portions of mass (M). Mass extends inferiorly through femoral canal into proximal thigh and extrinsically displaces left iliopsoas muscle (IP) medially. Again note direct invasion of left ischium and ilium (arrow) by tumor.

Primary retroperitoneal EGCTs are hypothesized to arise from primordial midline germ cell remnants of the genital ridge that fail to migrate properly.123,198,871 An alternative theory is that primary EGCTs are related synchronously or metachronously to an occult gonadal germ cell tumor (carcinoma in situ or “burned-out” tumor). Because the majority of retroperitoneal germ cell tumors are metastases from primary testicular tumors, careful clinical and imaging evaluation should be performed in affected men to exclude a coexistent primary testicular neoplasm.123,138,293 Seminomatous EGCTs account for 15% of EGCTs and are not associated with elevated tumor markers. Nonseminomatous EGCTs are associated with elevated tumor markers in 70% to 94% of cases. Levels of serum α-fetoprotein (AFP) may be elevated owing to the presence of yolk sac tumor or embryonal carcinoma, or human chorionic gonadotropin (hCG) levels may be elevated owing to the

presence of components of choriocarcinoma. Serial levels of tumor markers correlate with the clinical course and therapeutic response, and high levels are associated with poor survival.123 On gross examination, seminomas are usually homogeneous lobulated masses, whereas mixed and nonseminomatous tumors are heterogeneous with solid and cystic areas, necrosis, and hemorrhage.77,123,871 On CT and MRI, primary retroperitoneal EGCTs are typically large (mean size, 7-8 cm) midline-enhancing retroperitoneal masses of soft tissue attenuation that are of low to intermediate SI on T1 and intermediate to high SI on T2 relative to skeletal muscle. Areas of cystic change or necrosis are seen as very-high-SI foci on T2-weighted images.198,454 Seminomatous EGCTs tend to be homogeneous in attenuation and SI, whereas mixed and nonseminomatous EGCTs tend to be heterogeneous in attenuation and SI, with areas of cystic necrosis or hemorrhage.74,367,871 A midline location of a retroperitoneal mass is

1882

PART II  CT and MR Imaging of the Whole Body

probably the most helpful finding to suggest this diagnosis, whereas metastatic retroperitoneal lymphadenopathy from a primary testicular neoplasm tends not to be midline in location.123,198,871 The prognosis and treatment of primary retroperitoneal EGCT is equivalent to that of primary testicular neoplasms with retroperitoneal metastases.273 Nonseminomatous histology, presence of nonpulmonary visceral metastases, and elevated hCG are independent prognostic factors for shorter survival in patients with primary EGCT.77 The survival rate for patients with primary seminomatous retroperitoneal EGCT is 88% as compared to 63% for patients with primary nonseminomatous retroperitoneal EGCT, and the overall survival rate for patients with primary retroperitoneal EGCT is 65%.54,77

RETROPERITONEAL NEUROGENIC TUMORS Paraganglioma Paragangliomas, sometimes called extraadrenal pheochromocytomas, are rare neurogenic tumors that arise from highly vascularized specialized neural crest cells called paraganglia that are symmetrically distributed along the aortic axis in close association with the sympathetic chain in the neck, chest, abdomen, and pelvis. The largest collection of paraganglia includes the paired organs of Zuckerkandl that overlie the aorta at the level of the inferior mesenteric artery and have an uncertain physiologic role. They are prominent during early infancy and regress after 12 to 18 months.198,256 From 10% to 20% of pheochromocytomas are extraadrenal in location, and most often arise in the retroperitoneum from the organs of Zuckerkandl. Only a few tumors develop at other locations along the aorta or its branch vessels.156,198,294,790 On gross examination, paragangliomas are partially encapsulated solid and/or cystic brown masses that usually measure several centimeters in diameter and are commonly hemorrhagic. Histologically, paragangliomas are composed of small polygonal or slightly spindled cells with an amphophilic or eosinophilic cytoplasm, arranged in short irregular anastomosing sheets around a delicate vasculature.156 Welldefined cell nests (“Zellballen”) may also be seen, composed of chief cells with a peripheral thin layer of sustentacular cells separated by highly vascularized fibrous septa.464 Patients with paragangliomas present in the fourth and fifth decades of life, although malignant paragangliomas may sometimes arise in younger patients.132,256,294 Men and women are affected equally. Paragangliomas may be multicentric, particularly if there is a family history of paraganglioma, or may be associated with other tumors such as gastric malignant gastrointestinal stromal tumors (GISTs) and pulmonary chondromas as a component of Carney’s triad.162,790 Some 10% of paragangliomas are familial, occurring in conditions such as multiple endocrine neoplasia (MEN) types IIA and IIB, NF1, von Hippel-Lindau syndrome, and hereditary paraganglioma-pheochromocytoma syndrome (HPPS) related to genetic mutations in succinate dehydrogenase (SDH).511,953 Up to 40% of paragangliomas are malignant, as compared to 10% of adrenal pheochromocytomas. Malignant nature is recognized by metastatic spread or locally aggressive behavior; approximately 10% of patients initially present with metastatic disease.156,674 The malignant potential and biological behavior of a paraganglioma cannot be determined from its histologic appearance.325 Paragangliomas may spread both via the lymphatics and hematogenously, and the most common sites of metastatic disease are lymph nodes, bone, lung, and liver.325,634 Occasionally, spontaneous rupture of a retroperitoneal paraganglioma may occur with associated life-threatening retroperitoneal hemorrhage.706 Paragangliomas may be considered to be functional or nonfunctional depending on whether catecholamines are secreted or not.  Nonfunctional abdominal paragangliomas present with nonspecific

symptoms and signs and are rarely diagnosed as such prior to surgery. Paragangliomas may be functional in up to 60% of patients and may cause chronic or intermittent hypertension, headaches, or palpitations. In affected patients, detection of elevated urinary catecholamines is the most efficacious way of characterizing an abdominal mass as a paraganglioma.256,257,294 There is poor correlation between the functional activity of the tumor and the degree of malignancy, but functional paragangliomas are usually smaller than nonfunctional paragangliomas when detected (mean size, 7 cm vs. 12 cm, respectively).252,256,294,634 On CT, enhancing, well-circumscribed, lobular or round soft tissue–attenuation masses are seen that may be homogeneous in attenuation when small or heterogeneous when large. Central areas of low attenuation may be due to central necrosis or cystic change, which may be seen in up to 40% of cases.454,464,610,706,730 Punctate calcification may be seen in up to 15% of cases, and focal areas of high attenuation due to acute hemorrhage may also be seen in some tumors.294,610,706 Anecdotal reports in the literature report the induction of hyperadrenergic symptoms or hypertensive crises in patients with paraganglioma or pheochromocytoma after IV administration of older ionic contrast agents for CT. However, more recent reports in the literature suggest that IV administration of nonionic contrast material for CT is safe in patients with paraganglioma and pheochromocytoma, and that α-blocking medication is unnecessary in conjunction with contrast administration.67,240,584 On MRI, SI and enhancement characteristics of paragangliomas are similar to those of adrenal pheochromocytoma.193,198 The masses generally have low to intermediate SI on T1-weighted images, moderately  high SI on T2 relative to skeletal muscle, and are commonly heterogeneous secondary to foci of intratumoral necrosis or hemorrhage  (Fig. 55-20).256,464,538,685 Low-SI intratumoral septa and a low-SI peripheral capsule are also commonly seen on T2-weighted images.686 Progressive enhancement is present on parenchymal/venous and delayed phases of enhancement, but arterial-phase enhancement is variable.336 The size, attenuation and SI characteristics, and degree of heterogeneity cannot be used to differentiate between benign and malignant paragangliomas, the only definitive imaging criterion of malignancy being the presence of metastatic disease.256,294,538,685 CT and MRI also cannot differentiate between functional and nonfunctional paragangliomas.464 123 I or 131I-metaiodobenzylguanidine (MIBG) scintigraphy may be useful to detect clinically suspected tumor that cannot be localized, multiple primary tumors, tumors in unusual locations, or metastatic disease.63,68,687,888 The specificity of MIBG scintigraphy approaches 100% and is greater than that of CT and MRI, but its sensitivity is generally lower (range, ≈58% to 94%).68,92,268,345,726,888,898,926 Thus imaging with CT or MRI is recommended if there is a high index for suspicious for paraganglioma and the tumor is not found with scintigraphy.669,758 PET using other radiolabeled tracers such as 18F-fluorodopamine (FDOPA) or 11C-hydroxyephedrine is currently being studied and is promising in improving the diagnosis, staging, and therapeutic monitoring of such tumors.257,314,642,726,770,848,859,923 Complete surgical resection is the treatment of choice for abdominal paraganglioma; adjunctive therapies including radiotherapy,  131 I-MIBG, and chemotherapy are considered palliative.325,790 Surgical resection of localized metastatic disease may also be performed,  especially if solitary.325 If an abdominal paraganglioma is suspected clinically, assessment of the functional activity should be performed prior to surgery. Patients with functional tumors are premedicated with α-blockers to avoid intraoperative hypertensive crises during surgical manipulation of the tumor.205,256,669 Overall the 5-year survival rate may approach 82% for those with abdominal paraganglioma.674 The limitations of histopathologic criteria in predicting malignant behavior, the long natural history of the disease, and a

CHAPTER 55  Retroperitoneum

1883

L

L

M M

*

P

P

B

A

C FIG 55-20  Retroperitoneal paraganglioma in asymptomatic 35-year-old woman. A and B, Coronal T1-weighted (A) and T2-weighted (B) images show well-circumscribed retroperitoneal mass (M) that abuts right lobe of liver (L) and pancreatic head (P), with heterogeneous low to intermediate T1 SI and intermediate to high T2 SI relative to skeletal muscle. Very high T2 SI foci (*) represent intratumoral cysts or necrosis. C, Coronal arterial-phase fat-suppressed contrast-enhanced T1-weighted image demonstrates moderate heterogeneous enhancement of solid portions of mass without enhancement of intratumoral cystic or necrotic foci.

high propensity for subsequent metastasis make extended follow-up necessary.325

Ganglioneuroma Ganglioneuromas are uncommon benign neurogenic tumors that arise from sympathetic ganglia and represent 1% to 2 % of all primary retroperitoneal tumors, outnumbering neuroblastomas by about 3 to 1.256,638,961 They occur slightly more commonly in women than in men (1.5;1), most commonly in the first through fifth decades of life, with a mean age at diagnosis of 7 years. They are most often located in the posterior mediastinum (39%-43%), followed by the retroperitoneum (32%-52%), and the neck and pelvis (8%-9%).335,497,638,961 Although some ganglioneuromas result from maturation of neuroblastomas and ganglioneuroblastomas, the majority arise de novo.256,497 Most patients with ganglioneuromas are asymptomatic and have normal levels of urinary catecholamines.506 When symptomatic, abdominal pain and a palpable abdominal mass are the most frequent symptoms and signs.638 Patients with hormonally active ganglioneuro-

mas may clinically present with episodic hypertension, sweating, flushing, or diarrhea due to the excess catecholamine production.335 Secretion of androgenic hormones may lead to virilization with some tumors.691,706 On gross inspection, ganglioneuromas are well-circumscribed encapsulated benign neoplasms that have a mean size of 8 cm.497 Histologically they are composed of uniform bundles of longitudinal and transversely oriented mature Schwann cells that crisscross each other in an irregular fashion, along with scattered mature ganglion cells, variable amounts of myxoid stroma and collagen, and rarely stromal fat.256,497,961 Extremely rarely, ganglioneuromas may undergo malignant transformation.256 Retroperitoneal ganglioneuromas are typically well-defined longitudinally oriented masses that are lobulated or oval in shape, have a tendency to surround major blood vessels with little or no luminal narrowing, usually do not result in osseous changes, and only infrequently extend into the neural foramina.638,691,912,961 On CT and MRI, ganglioneuromas have relatively homogeneous attenuation similar to or lower than that of skeletal muscle, with low SI on T1-weighted

1884

PART II  CT and MR Imaging of the Whole Body

* *

B

A

* *

D

C FIG 55-21  Extraperitoneal pelvic ganglioneuroma in 22-year-old woman with abdominal pain. A to D, Axial contrast-enhanced CT (A), T1-weighted (B), T2-weighted (C), and fat-suppressed contrast-enhanced T1-weighted (D) images through pelvis demonstrate infiltrative lobulated pelvic extraperitoneal mass with low attenuation, intermediate to slightly high T1-weighted SI and heterogeneous high T2-weighted SI relative to skeletal muscle, scattered curvilinear foci of internal enhancement, and thin peripheral enhancement. Note that mass displaces and encases left external iliac artery and vein anteriorly (arrows) and extends into and expands left sacral foramen (*).

images and heterogeneous intermediate to high SI on T2-weighted images (Fig. 55-21).100,335,610,691,706 Intratumoral hemorrhage or fatty components may cause mixed intermediate to high SI on T1.691,961 SI on T2 is influenced by the proportion of myxoid stroma to cellular components and collagen fibers. Intermediate SI on T2-weighted images occurs when there is an abundance of cellular components and collagen fibers relative to myxoid stroma, whereas high SI on T2 occurs when there is a large amount of myxoid stroma with relatively few cellular components and collagen fibers.100,961 A whorled appearance may be present on T2-weighted images and sometimes on T1-weighted images and is due to interlacing bundles of Schwann cells and collagen fibers.616,961 In contrast to schwannomas, cystic degeneration is not present.291,961 On both T1 and T2 a low-SI peripheral rim may be visualized but is more often identified on contrast-enhanced images.961 Calcifications are seen in 42% to 60% of tumors on CT, tend to be discrete and punctuate, and appear as low-SI foci on T1- and T2-weighted images. In contrast the calcifications in neuroblastoma are amorphous and coarse in appearance.335,366,497,691 After contrast administration, ganglioneuromas show gradual progressive enhancement, although linear septa may enhance early.335,638,706,961 The CT and MRI imaging characteristics of ganglioneuroma overlap with those of ganglioneuroblastoma and neuroblastoma, and distinguishing between these lesions is not possible unless metastatic disease is present.497 Approximately 57% of ganglioneuromas may be functional,

produce increased amounts of catecholamines, and accumulate 123 I-MIBG.241,858 Management of retroperitoneal ganglioneuromas involves complete surgical resection when possible, especially when symptomatic due to large size or hormonal activity.256,497 After surgical resection, the prognosis is excellent and recurrence is rare, but periodic radiologic surveillance is advised.497,706

Ganglioneuroblastoma and Neuroblastoma Ganglioneuroblastoma most often develops in children aged 2 to 4 years. It is rare in adults and affects men and women equally. It occurs most often in the posterior mediastinum, retroperitoneum, and adrenal gland.203,934 Neuroblastoma most commonly occurs in the first decade of life, with 80% occurring in children younger than 5 years and 42% occurring in children younger than 1 year. It is slightly more common in boys than girls and rare in adults.84,271,497 Neuroblastoma is the third most common pediatric malignancy after leukemia and central nervous system tumors, accounting for close to 15% of childhood cancer fatalities.497 Two thirds of neuroblastomas are located in the abdomen, of which two thirds are located in the adrenal gland and the remainder in the extraadrenal retroperitoneum. The distribution of ganglioneuroblastomas and neuroblastomas follows the distribution of the sympathetic ganglia in a paramedian location, with the adrenal medulla and organ

CHAPTER 55  Retroperitoneum of Zuckerkandl being additional sites of involvement.256,497 Locally advanced tumors or distant metastases may be present at the time of diagnosis in 50% of patients with ganglioneuroblastoma and over 70% of patients with neuroblastoma, with involvement of bone, bone marrow, liver, lymph nodes, and skin.84,934 Serum or urinary catecholamines and their metabolites, vanillylmandelic acid (VMA) and homovanillic acid (HVA), may also be elevated in some cases.497,934 On gross examination, ganglioneuroblastomas may be partially or totally encapsulated, whereas neuroblastomas are unencapsulated with variable amounts of hemorrhage, calcification, and necrosis present. Histologically, ganglioneuroblastoma is composed of both mature ganglion cells and Schwann cells in addition to immature neuroblasts and has intermediate malignant potential.256,497,706,934 Neuroblastoma is composed of small neuroepithelial cells that may show glial or ganglionic differentiation, contains nests of primitive round cells with darkstaining nuclei and scant cytoplasm, and is frankly malignant.497,706 Homer Wright rosettes, circular or ovoid columns of tumor cells arranged around a central core of neuropil, if present, are characteristic of neuroblastoma but are not always present.497 There is remarkable heterogeneity observed in neuroblastoma phenotype, ranging from spontaneous regression to relentless progression, and there are dozens of clinical and biological markers that have been proposed as being predictive of disease outcome.322,524 Children who present with neuroblastoma earlier than age 12 months have a better prognosis than those who present after age 12 months.271 Even with metastatic disease, infants younger than 12 months can have a favorable outcome with treatment. The majority of children older than 12 months with advanced neuroblastoma at presentation will die from progressive disease.243,497,917 The most common staging system used for neuroblastoma is the International Neuroblastoma Staging System (INSS) (Box 55-1), which takes into account the clinical stage, imaging results, surgical findings, and bone marrow examination.93 Patients with stage 1 and 2 disease have a good prognosis, and there is a high rate of spontaneous remission of neuroblastoma in infants with stage 4S disease, as well as high survival rates in patients with stage 4S disease without MYCN

BOX 55-1  International Neuroblastoma

Staging System (INSS) of Neuroblastoma93,497 Stage/Prognostic Group 1 Localized tumor with complete gross excision, with or without microscopic residual disease; identifiable ipsilateral lymph nodes negative for tumor microscopically 2A Localized tumor with incomplete gross excision; identifiable ipsilateral nonadherent lymph nodes negative for tumor microscopically 2B Localized tumor with or without complete gross excision, with ipsilateral nonadherent lymph nodes positive for tumor; identifiable contralateral lymph nodes are negative for tumor microscopically 3 Unresectable unilateral tumor infiltrating across the midline (i.e., vertebral column) with or without regional lymph node involvement; or localized unilateral tumor with contralateral regional lymph node involvement; or midline tumor with bilateral extension by infiltration or regional lymph node involvement 4 Any primary tumor with dissemination to distant lymph nodes, bone, bone marrow, liver, skin, and/or other organs (except as defined for stage 4S) 4S Localized primary tumor (as defined for stage 1, 2A, or 2B) with dissemination limited to skin, liver, and/or bone marrow (<10% tumor) in infants younger than 12 months of age

1885

amplification.322 Patients older than age 18 months with stage 4 disease and unresectable disease have a worse outcome.536,917 On CT and MRI, ganglioneuroblastomas and neuroblastomas are less than 10 cm in size and have similar attenuation and SI characteristics to ganglioneuromas.497,706,912 They tend to have poorly defined margins with invasion of adjacent organs and encasement or occlusion of vessels. There may be cystic change or foci of hemorrhage.497,547,691,706,912,934,961 MRI is the preferred imaging modality for demonstrating intraspinal extension of tumor, which is seen with 10% of abdominal neuroblastomas, as well as for detecting metastatic involvement of bone marrow.144,497,771 Foci of amorphous and coarse calcification are commonly seen on CT in up to 90% of patients with neuroblastoma, in contrast to the discrete and punctate calcification seen in ganglioneuroma. These calcific foci are seen as very-low-SI areas on T1- and T2-weighted imaging.335,497 Heterogeneous arterialphase enhancement is also seen.335,834,934 Only 70% of ganglioneuroblastomas and neuroblastomas are MIBG positive, but FDG PET may be useful to evaluate patients with poorly differentiated tumors that are MIBG negative.322,497,753,769 Treatment of retroperitoneal ganglioneuroblastoma and neuroblastoma may include locally aggressive surgical excision, adjuvant or neoadjuvant chemotherapy, and external beam or targeted radiation therapy, depending on the age at diagnosis, stage of disease, tumor histopathology, MYCN amplification, tumor cell ploidy, and extent of involvement of vital anatomic structures by tumor.322,497,934 Patients with distant metastases have a poorer prognosis, while patients without distant metastases who undergo radical resection may have a favorable prognosis.322,497,934 The prognosis and response to therapy for patients with retroperitoneal ganglioneuroblastoma are better than those for patients with retroperitoneal neuroblastoma.395 The 5-year survival rate for adult patients (aged 21 or older) with neuroblastoma is about 36%, compared to 85% for infants.203

Schwannoma Schwannomas are benign tumors of nerve sheaths of peripheral nerves that account for up to 4% of all retroperitoneal tumors; they are most frequently found in the head and neck region and flexor surfaces of the extremities.48,256,590 They occur in the third through sixth decades of life, usually as solitary lesions that arise sporadically, and are twice as common in women than in men.256,291,491 Most patients are asymptomatic and present with a slowly growing painless soft tissue mass. Malignant transformation is rare.291,479,491,706,802 Schwannomas are usually less than 5 cm in diameter at presentation, but retroperitoneal schwannomas are typically larger than 8 cm at the time of presentation.479,706 Schwannomas are solitary fusiform masses derived from Schwann cells. They are surrounded by a fibrous capsule and are eccentrically located in relation to the parent nerve.256,491 Histologically the schwannoma is composed of a highly ordered cellular component of compact spindle cells in short bundles or interlacing fascicles (Antoni A area) and a less organized and less cellular loose myxoid component (Antoni B area).256,479 Larger retroperitoneal schwannomas are more likely to undergo degenerative changes, including cyst formation (in up to 66%), calcification, hemorrhage, and hyalinization.256,827 Ancient schwannomas refer to long-standing lesions with advanced degenerative changes such as cyst formation, hemorrhage, calcification, and hyalinization.347,491,495 On CT and MRI, schwannomas are sharply circumscribed, fusiform, round or oval masses, usually located in the paravertebral or presacral portions of the retroperitoneum. They are of low to intermediate SI on T1, high SI on T2, and have solid enhancing components. A low-SI capsule may sometimes be seen on T1- and T2-weighted

1886

PART II  CT and MR Imaging of the Whole Body

* *

B A

*

*

C

D FIG 55-22  Retroperitoneal schwannoma in 32-year-old woman with endometriosis and microhematuria. A to D, Axial contrast-enhanced CT (A), T1-weighted (B), T2-weighted (C), and fat-suppressed contrastenhanced T1-weighted (D) images through pelvis show round well-circumscribed retroperitoneal mass with predominantly low attenuation and scattered regions of soft tissue attenuation, slightly low to intermediate T1-weighted SI, high T2-weighted SI relative to skeletal muscle with enhancing regions of lower T2-weighted SI, and nonenhancing central fluid attenuation, low T1-weighted SI, and very high T2-weighted SI focus (*) corresponding to cystic or necrotic region.

images (Fig. 55-22).198,254,291,396,408,479,493,706,802,961 Heterogeneous attenuation and SI are much more common in schwannomas than in neurofibromas and may be due to the mixture of Antoni A and B areas, along with hemorrhage, cystic degeneration, or punctate, mottled, or curvilinear calcification that is often difficult to visualize on MRI prospectively.254,291,396,408,491,706 Myxoid components have low attenuation and high SI on T2 and show variable enhancement.590 A “target sign” may be present in both schwannomas and neurofibromas on T2-weighted imaging, which consists of a central area of low-to-intermediate-SI fibrous tissue surrounded by peripheral high-SI myxoid tissue.491,616 A target-like enhancement pattern with greater enhancement centrally than peripherally may be seen on contrast-enhanced CT or T1-weighted imaging.706 If the parent nerve of a schwannoma can be identified, the schwannoma can be seen to have an eccentric position in relation to the nerve.291,491 Schwannomas are treated by surgical excision or enucleation, particularly if large or symptomatic. The adjacent nerve can usually be spared because the schwannoma is separable from the underlying nerve fibers.254,479,491

Neurofibroma Neurofibromas are benign tumors of nerve sheaths of peripheral nerves that represent 5% of all benign soft tissue neoplasms.491 They

occur sporadically in the third through fifth decades of life, are more common in men, and may develop in up to 10% of patients with NF1 in a younger age range.256,491,706 Approximately one third of patients with a solitary neurofibroma have NF1, and almost every patient with multiple or plexiform neurofibromas has NF1.706 Neurofibromas commonly occur in deep anatomic locations in patients with NF1 (especially in retroperitoneal and paraspinal locations) and are commonly associated with neurologic symptoms.256,491 Other nonspecific symptoms and signs may also be seen with abdominal neurofibroma. Sporadic neurofibromas are usually smaller than 5 cm, whereas neurofibromas in those with NF1 tend to be multifocal and larger.491,593 Neurofibromas are not encapsulated and may occur as localized, plexiform, or diffuse types. The latter type is typically found in the subcutaneous tissues of patients with NF1 and rarely undergoes malignant transformation.256,491 Histologically, neurofibromas are composed of nerve sheath cells, thick wavy collagenous bundles, and variable amounts of myxoid degeneration.256 Often, longitudinal bundles of residual nerve fibers are centrally situated in the neurofibromas.930 Unlike schwannomas, neurofibromas are not encapsulated and lack a clear partition into Antoni A and B areas.706 Neurofibromas usually have a fusiform shape that is oriented longitudinally in the distribution of the nerve and has tapered ends that are centered upon and contiguous with the parent nerve. Associated

CHAPTER 55  Retroperitoneum muscle atrophy in a particular nerve distribution may also be seen. The “split-fat sign” may be seen when a rim of fat surrounds the tumor, originating from a nerve in an intermuscular location.491 A characteristic dumbbell shape may sometimes be seen with spinal nerve root neurofibromas that extend through and enlarge the neural foramen. This morphology may also be seen with other nerve and nerve sheath tumors involving the spinal nerve roots.330,491,610,616 When multiple or of the plexiform type, which is pathognomonic of NF1, large conglomerate infiltrative masses of innumerable neurofibromas diffusely thicken a parent nerve and extend into multiple nerve branches, resulting in a characteristic “bag of worms” appearance.491,706 Plexiform neurofibromas in the retroperitoneum are typically bilateral and symmetric in a parapsoas or presacral location and follow the distribution of the lumbosacral plexus. Parapsoas neurofibromas tend to be long and cylindrical in configuration, whereas confluent presacral neurofibromas often extend into the pelvis in a sheetlike fashion.53,217 Other findings of NF1 such as scoliosis, dural ectasia, or anterior and lateral meningoceles may be present.217 On CT, neurofibromas are of low soft tissue attenuation. On MRI, they are of low SI on T1 and high SI on T2 (Fig. 55-23).217,491,610 The SI on T2-weighted images may be homogeneous or may be heterogeneous with either a target sign, as described with schwannomas, or with a whorled appearance consisting of linear or curvilinear low-SI Schwann cell bundles and collagen fibers in a background of high SI.491,616 The whorled appearance may also be seen on CT, with high-attenuation linear or curvilinear foci seen in a background of low attenuation.53 After contrast administration, enhancement may be homogeneous or heterogeneous. Areas of fluid attenuation may be seen and are due to regions of myxoid degeneration; this may cause confusion with necrotic lymphadenopathy or other cystic lesions in the retroperitoneum.217,292,610 Target-like central enhancement may also be encountered on CT or MRI.53,706

Because neurofibromas are intimately associated with their underlying parent nerves, surgical resection requires removal of the nerves of origin.491 Therefore surgical treatment is performed only when neurofibromas are plexiform, symptomatic, or suspicious for malignant degeneration.

Malignant Peripheral Nerve Sheath Tumor Malignant peripheral nerve sheath tumors (MPNSTs) are malignant tumors that arise from or differentiate toward cells of the nerve sheaths of peripheral nerves. They represent 5% to 10% of soft tissue sarcomas, occur in 2% to 5% of patients with NF1, and arise from preexisting neurofibromas, usually after a latent period of at least 10 years.256,423,424,473,491,930 Conversely, about 50% of patients with MPNST have NF1. From 10% to 20% of MPNSTs are related to prior radiation exposure, occurring after a latent period of more than 15 years.256,473,491,930 Most MPNSTs occur during the third through sixth decades of life, which is earlier than other retroperitoneal sarcomas. MPNSTs that develop in association with NF1 tend to present within the third decade of life; they are of higher histologic grade, larger size, and greater aggressivity, and have a very poor prognosis compared to those that are sporadic.256,406,473,694,928 MPNSTs that arise from major nerves such as the sciatic nerve or sacral plexus may cause sensory and motor symptoms such as radiating pain, paresthesias, and weakness.256,473,491 Unexpected growth of a preexisting neurofibroma, especially a plexiform neurofibroma, or associated pain, particularly if spontaneous and unremitting, should raise clinical suspicion for MPNST, although benign neurofibromas may also grow and become painful.323,423,424,491 In patients with NF1, pain associated with a mass is the greatest risk factor for MPNST development, with a relative risk of 30-fold.406 On CT and MRI, benign and malignant neural tumors cannot be reliably differentiated; attenuation, SI, and enhancement characteristics

A

B

1887

C FIG 55-23  Extraperitoneal pelvic neurofibromas in 33-year-old woman with neurofibromatosis type 1, back pain, and right lower extremity pain. A to C, Axial T1-weighted (A), axial fat-suppressed T2-weighted (B), and coronal T2-weighted short tau inversion recovery (STIR) (C) images through pelvis demonstrate multiple well-circumscribed masses (arrows) in left ischiorectal fossa, right inguinal region, and right posterior pelvis. Masses have homogeneous low T1 SI and heterogeneous high T2 SI relative to skeletal muscle, with curvilinear areas of low T2 SI resulting in whorled appearance.

1888

PART II  CT and MR Imaging of the Whole Body

overlap. MPNSTs tend to be larger than 5 cm, may exhibit poorly defined margins suggesting infiltration of surrounding tissues, may have associated edema, and may expand neural foramina. These features may also be seen with benign neural tumors. Heterogeneity with central necrosis is common in MPNST, but benign tumors with cystic degeneration may also appear heterogeneous. Calcification is more commonly associated with MPNST but may also be seen in ancient schwannomas. The most important imaging finding that should raise suspicion of MPNST is rapid enlargement of a tumor mass, particularly if associated with pain (Fig. 55-24).53,323,423,491,706 Complete surgical resection is the treatment of choice for retroperitoneal MPNST, as with other retroperitoneal sarcomas, and is the  most important factor influencing patient survival.323,491,928 Local recurrence and distant metastatic disease are common complications of MPNST.473,491 The overall 5-year survival rate for patients with

MPNST is 34% to 52%, and the 5-year survival rate of MPNST in patients with NF1 is 15%.473,928

RETROPERITONEAL FIBROTIC LESIONS Retroperitoneal Fibrosis Retroperitoneal fibrosis (RPF) is a rare fibrotic reactive process considered a member of a family of disorders referred to as chronic periaortitis.16,876 Two thirds of all cases of RPF are considered idiopathic (also called Ormond’s disease), and approximately one third of cases develop in response to various medications, malignancies, or other etiologies.16,877 The exact etiology of idiopathic RPF is unclear, although chronic fibrosing periaortitis, possibly due to an immune reaction to a component of ruptured atherosclerotic plaque such as ceroid (a lipoproteic

* *

A

B

K

K

C

D FIG 55-24  Retroperitoneal malignant peripheral nerve sheath tumor (MPNST) in 31-year-old man with neurofibromatosis type 1 and palpable painful abdominal mass. A, Axial contrast-enhanced CT image shows large heterogeneous infiltrative right retroperitoneal mass with low-attenuation and soft tissue–attenuation components. B, Axial fat-suppressed T2-weighted image demonstrates high-SI components (*) relative to skeletal muscle within mass owing to cystic degeneration or necrosis. C and D, Coronal fat-suppressed T1-weighted image before (C) and after (D) contrast administration show high-T1-SI hemorrhagic foci (long arrows) within mass and mild heterogeneous central and peripheral enhancement of mass. There is associated superior displacement of right kidney (K) by mass. Note mildly enhancing well-circumscribed pedunculated cutaneous neurofibroma (arrows) along left lateral abdominal wall.

CHAPTER 55  Retroperitoneum polymer that results from low-density lipoprotein oxidation within plaque macrophages), may play a role.249,428,877 The observation that RPF tends to occur in areas where an arterial wall (usually of the abdominal aorta) has atherosclerotic plaque and attenuation of the media supports this theory.16 However, this theory has been challenged over the past decade, and an underlying systemic autoimmune process is believed to be responsible instead.876 In up to 15% of individuals with RPF, associated fibrotic processes outside the retroperitoneum may be present, including fibrosing mediastinitis, sclerosing mesenteritis, orbital pseudotumor, primary sclerosing cholangitis, Reidel’s thyroiditis, and immunoglobulin G4 (IgG4)-related disease.392,428,883,903 Other autoimmune or inflammatory disease processes such as systemic lupus erythematosus, rheumatoid arthritis, ankylosing spondylitis, small and medium-sized vessel vasculitides, and asbestos exposure also have an association with RPF.242,877 Up to 15% of abdominal aortic aneurysms (AAAs) may have an association with perianeurysmal fibrosis and are considered by some to be an early or mild form of RPF. These are discussed later under “Inflammatory Abdominal Aortic Aneurysm” (IAAA). One percent of patients who take methysergide, an ergot derivative used to treat migraine headaches, develop RPF, with symptomatic relief commonly occurring after drug withdrawal.16,428 This known side effect of methysergide (as well as pleural and valvular fibrosis) has led to decreased use of this drug to treat migraines and consequently fewer drug-related cases of RPF.774 Other ergot derivatives such as bromocriptine, used in the treatment of parkinsonism, may also be associated with RPF. Malignant RPF is an unusual subtype of RPF and is clinically difficult to distinguish from RPF due to benign or idiopathic causes. Malignant RPF occurs when small metastatic foci to the retroperitoneum (usually from lymphoma) elicit a desmoplastic response.16 Most commonly, Hodgkin’s disease and other lymphomas, retroperitoneal sarcomas, carcinoid tumors, and carcinomas from the breast, lung, thyroid, stomach, colon, kidney, bladder, prostate, and cervix are the culprits, although any primary malignancy may be involved.16,29,428 Malignant RPF is distinct from malignant retroperitoneal lymphadenopathy.428 Infections due to tuberculosis (TB), syphilis, actinomycosis, or fungi; nonspecific gastrointestinal (GI) inflammation including appendicitis, Crohn’s disease, or diverticulitis; and retroperitoneal hemorrhage, urine extravasation, or prior irradiation or surgery may also lead to RPF.16,428,877 RPF typically originates below the aortic bifurcation at the level of the sacral promontory or lower lumbar vertebrae and then extends superiorly along the anterior spinal surface in a periaortic and pericaval distribution toward the renal hila, where it may rarely surround the renal pelves. Typically one or both ureters, usually in the middle third, may be encased, often resulting in hydronephrosis. However, the fibrotic process may also spread inferiorly to involve the iliac and gonadal vessels, rectosigmoid colon, urinary bladder, and other pelvic organs, or anteriorly along the celiac and superior mesenteric arteries.16,428,742,883 Infrequently, involvement of the lungs, mediastinum, pericardium, pleura, spine, epidural space, spinal cord, gonadal vessels, pancreas, spleen, duodenum, colon, small bowel mesentery, urinary bladder, unilateral renal sinus or ureter, uterus, and cervix have also been reported.249,428,742,877 Symptoms and signs are related to entrapment and compression of retroperitoneal structures, including the ureters, IVC, abdominal aorta and its branches, and gonadal vessels.16 The ureters are the most frequently compressed structures, and oliguria, anuria, and eventual renal failure may occur.16,428,883 Compression of the IVC can result in lower extremity edema, whereas aortic and common iliac arterial compression may lead to lower extremity claudication. Renal artery compression may result in renovascular

1889

hypertension, and mesenteric artery compression may cause bowel ischemia.16 Rarely, pancreatic involvement, biliary obstruction, portal vein occlusion, GI obstruction, or spinal cord compression due to extension of RPF through the neural foramina may occur.16,169,235,428 Laboratory abnormalities may include abnormal renal function tests in those with ureteral obstruction, and there may be elevations in erythrocyte sedimentation rate (ESR) and C-reactive protein levels.249,877 Idiopathic RPF occurs more commonly in men by a ratio of 2-3 : 1, which has been attributed to the higher incidence of symptomatic atherosclerotic disease in men.16,249,742,883 However, RPF due to methysergide occurs twice as commonly in women as in men, and RPF associated with malignancy occurs equally in men and women.16 Most patients with RPF present during the fifth through seventh decades of life.249,428,742,883 On gross examination a dense grayish-white plaquelike mass surrounds the inferior aspect of the abdominal aorta and often encases or compresses the aortic branches and the ureters. Histologically, early RPF is characterized by immature fibrosis with numerous fibroblasts, inflammatory cells, capillary proliferation, and a loose network of collagen fibers with a high fluid content due to leaky vascular interendothelial junctions.16,29,460 Malignant RPF shows a similar histologic pattern, with the addition of scattered malignant cells in the stroma.16 As RPF matures, the cellular and inflammatory elements decrease and hyalinized collagen becomes more prominent.29,249,460 This maturation process progresses laterally from the midline such that the lateral edges of RPF tend to be inflammatory, whereas the central portion tends to be more fibrotic.16,249 The anterior margin of RPF is sharply marginated with the posterior peritoneum, whereas the posterior margin is poorly defined.16 The fibrotic plaque may be midline or eccentric, well circumscribed or poorly defined, localized or extensive, and may appear similar morphologically in both benign and malignant RPF.16,29 On CT, RPF has homogeneous soft tissue attenuation similar to or slightly greater than that of skeletal muscle.16,428,903 On MRI the T2-weighted SI of RPF is dependent on the activity of the disease. RPF has low to intermediate SI on T1-weighted images, and on T2, mature fibrotic plaque in benign RPF has low SI, whereas immature fibrotic plaque in benign RPF and malignant RPF has higher SI due to inflammatory edema or hypercellularity (Fig. 55-25).16,29,460,903 Chronic RPF tends to have a higher ADC value (i.e., less restricted diffusion) compared to that of active or malignant RPF.42,719 Enhancement of RPF on CT and MRI is variable and depends on the maturity of the fibrous process, with immature plaque enhancing to a greater degree owing to greater vascularity.16 After corticosteroid therapy and with maturation, there is a decrease in inflammatory reaction, with a subsequent decrease in T2-weighted SI as well as decreased enhancement on dynamic contrast-enhanced T1-weighted imaging.96,956 Imaging findings that suggest malignant RPF include the presence of other lymphadenopathy or metastatic disease, adjacent osseous destruction, a heterogeneous soft tissue mass with poorly defined margins, a longitudinal extent from above the renal arteries to below the aortic bifurcation, and associated high-SI changes of the adjacent psoas muscles on T2 (Fig. 55-26).29,198,566 Uptake of FDG on PET may be seen in RPF and is due to the increased metabolic activity of inflammatory cells; this may play a role in differentiating benign from malignant RPF, discriminating active from inactive disease, and monitoring disease activity following therapeutic intervention.5,65,357,437,578,735,822,860,878,883 Malignant retroperitoneal lymphadenopathy may mimic RPF, since it can occasionally become confluent and encase the great vessels. However, metastatic disease most commonly appears as lobulated paraaortic and paracaval masses that are due to enlarged lymph nodes.16 Lymphoma and other tumors that result in malignant

1890

PART II  CT and MR Imaging of the Whole Body

S

S

*

*

B

A

S

*

D

C

E FIG 55-25  Idiopathic retroperitoneal fibrosis (RPF) in 39-year-old man before and after medical treatment. A to C, Axial T1-weighted (A), fat-suppressed T2-weighted (B), and fat-suppressed contrast-enhanced T1-weighted (C) images demonstrate retroperitoneal soft tissue surrounding abdominal aorta (*) predominantly along its anterior and lateral aspects. Soft tissue has intermediate T1 SI and slightly high T2 SI relative to skeletal muscle, as well as avid enhancement, in keeping with immature fibrosis. Note lack of anterior displacement of aorta from lumbar spine. Also note subtle low T1 SI, high T2 SI, and mild enhancement in root of small bowel mesentery (S) in keeping with sclerosing mesenteritis, which is associated with RPF.  D and E, Axial fat-suppressed T2-weighted (D) and fat-suppressed contrast-enhanced T1-weighted (E) images obtained 6 months following medical therapy show interval decrease in size of retroperitoneal soft tissue as well as decreased T2 SI consistent with mature fibrosis.

lymphadenopathy typically displace the aorta anteriorly and the ureters laterally, whereas benign RPF usually does not cause significant anterior aortic displacement or lateral ureteral displacement.16,428 The goals of RPF treatment are to stop progression of the fibroinflammatory reaction, inhibit or relieve obstruction of the ureters or other retroperitoneal structures, switch off the acute-phase reaction and its systemic manifestations, and prevent disease recurrence or relapse.877 Although spontaneous regression of RPF has been reported rarely, most patients require some form of medical and/or surgical treatment. In patients with methysergide- or bromocriptine-related

RPF, discontinuation of the drug results in regression of both symptoms and fibrosis.428 In other patients with RPF, biopsies are obtained to exclude a malignant or infectious etiology before therapy is instituted.249,428 If RPF is secondary to infection, corticosteroids are contraindicated and specific antimicrobial therapy is instituted.877 Corticosteroids are used to decrease inflammation in early idiopathic disease, relieve the symptoms as well as shrink the RPF mantle on imaging, and relieve the ureteral obstruction within 1 to 2 weeks.16,225,389,482,877,881 Tamoxifen and immunosuppressants have also been used successfully to treat nonmalignant RPF.389,482,882 If medical

CHAPTER 55  Retroperitoneum

A

1891

B

*

*

C FIG 55-26  Malignant RPF in 54-year-old man with bladder carcinoma. A to C, Sequential axial contrastenhanced CT images through abdomen and pelvis demonstrate conglomerate soft tissue that surrounds abdominal aorta and proximal common iliac arteries. Note delayed nephrogram and moderate pelvicaliectasis of right kidney due to ureteral obstruction by RPF. Also note soft tissue masses (*) within posterior aspect of urinary bladder due to bladder carcinoma.

therapies are not effective in relieving the ureteral obstruction, ureteral stenting or ureterolysis (ureteral dissection from the surrounding fibrotic plaque) is successful.225,226,382,797 After ureterolysis, the ureters may be transposed laterally (after which retroperitoneal fat may be interposed between the ureters and the RPF), the ureters may be wrapped in omental fat, or the ureters may be transplanted into the peritoneal cavity, with the posterior peritoneal flap closed behind them to protect the ureters from future encasement and obstruction by the fibrotic process.16,428,877 The prognosis for patients with malignant RPF is poor, with a mean survival of 3 to 6 months after diagnosis.428,903 The prognosis for those with idiopathic RPF is generally favorable.428 Long-term success in prevention of recurrent urinary obstruction and maintaining renal function may be achieved with ureterolysis and medical therapy in over 90% of patients.903 However, because these patients usually have atherosclerotic disease with resultant complications such as myocardial infarctions and cerebrovascular accidents, the 10-year survival rate is usually reported as less than 70%.249

Inflammatory Abdominal Aortic Aneurysm From 5% to 15% of patients with AAAs have associated asymptomatic perianeurysmal fibrosis, which is similar morphologically and  histologically to RPF.249,650,841,908 IAAA is also considered a member of the family of disorders referred to as chronic periaortitis.876 The

pathogenesis of IAAA is thought to be similar to that of idiopathic RPF, with the perianeurysmal inflammatory response representing an immune reaction to antigens such as ceroid that leak from the aortic wall. Recent reports suggest that it may also be a local manifestation of a systemic autoimmune process, in particular IgG4-related disease.300,877,909 The aortic wall is thickened up to 3 cm with intense perianeurysmal and retroperitoneal fibrosis and extensive adhesions involving the surrounding organs.300,809 The anterior and lateral walls of IAAAs are covered with a thick white layer of fibrous tissue that fixes the duodenum and other adjacent structures to the aneurysm sac. In severe cases the perianeurysmal inflammatory/fibrous tissue may extend to involve the ureters, IVC, left renal vein, small bowel mesentery, or transverse mesocolon, with potential obstruction of the involved structures.701 Histologically, most of the aortic wall thickening results from expansion of the adventitia by a marked inflammatory reaction.300 Both atherosclerotic AAAs and IAAAs occur much more commonly in men compared to women, and in patients with IAAA, the mean age at presentation is in the seventh decade of life, 5 to 10 years younger than the mean age for atherosclerotic AAA.300,650,701 However, 75% of patients with IAAAs present with abdominal or back pain, compared to only 13.5% with noninflammatory AAA.300,739 Patients with IAAA are more likely to have a family history of aneurysms and tend to be current smokers.300 IAAAs tend to be larger than noninflammatory

1892

PART II  CT and MR Imaging of the Whole Body

AAAs at presentation (mean size, 8.0 cm vs. 6.4 cm, respectively), and the ESR is elevated in the majority of patients with IAAAs.78,618,739,810 The triad of chronic abdominal pain, weight loss, and elevated ESR in a patient with an AAA is highly suggestive of an IAAA.658 On CT, typical findings of an IAAA include a dilated abdominal aorta (diameter > 3 cm), thickening of the aortic wall (most marked anteriorly and laterally, with relative sparing posteriorly), and periaortic inflammation and fibrosis (which is of soft tissue attenuation with a variable amount of enhancement).28,300,341,809 The periaortic inflammation and fibrosis may be well defined or infiltrative in appearance.28 In a study of 355 patients with AAA, CT was 83.3% sensitive, 99.7% specific, and 93.4% accurate in diagnosis of IAAA.341 On MRI the typical appearance of an IAAA consists of a dilated aorta with concentric rings of high SI in the wall that are most prominent on T2 but are also present on T1. These mural concentric rings represent alternating layers of fibrosis and inflammation. IAAAs tend to have three or more concentric layers of high SI, whereas noninflammatory AAAs show at most two rings of high SI.841 On T1-weighted images, the surrounding rim of inflammatory tissue shows intermediate SI with poor differentiation from intraluminal thrombus and surrounding structures. On contrast-enhanced CT and T1-weighted imaging the inflammatory cuff usually enhances homogeneously, similar to that in patients with RPF who do not have an aneurysm. In contrast to retroperitoneal malignancy, IAAA rarely displaces the aorta anteriorly from the spine.832 Restricted diffusion may be seen on DWI and ADC map images.635 Uptake of FDG on PET may be seen in IAAA and is due to the increased metabolic activity of inflammatory cells.735,822 Preoperative treatment of IAAA patients with corticosteroids may be useful to control the inflammatory process and facilitate surgical repair of the IAAAs, particularly when the inflammatory process is severe or if there is associated involvement of adjacent retroperitoneal structures, and corticosteroids or other antiinflammatory immunosuppressive therapies may also be used after surgical repair if inflammation persists or is exacerbated.474,800,810,844 Aneurysm resection is the traditional treatment for IAAAs, owing to the risk of aortic rupture, especially when the maximal aortic diameter exceeds 5.5 cm. Over 75% of patients undergo spontaneous remission of perianeurysmal fibrosis following surgical repair. Perianeurysmal fibrosis is not protective against aneurysm rupture but often makes operative repair of such aneurysms more difficult and hazardous.78,474,649,841 Endovascular repair of IAAAs is also an option, reportedly with a high success  rate.140,277,300,307,609,649,704,805,877 However, the perianeurysmal fibrosis may not resolve but actually worsen over time in some patients, leading to or exacerbating ureteral obstruction.879 Endovascular repair is associated with a lower 1-year all-cause mortality compared to open surgical repair, although open surgical repair may be preferred in those patients who have hydronephrosis and are deemed low risk.649 The 5-year survival rates for IAAAs and noninflammatory AAAs are similar, ranging between 67% to 80%.78,739

Fibromatosis (Desmoid Tumor) The fibromatoses are divided into superficial (fascial) and deep (desmoid tumor) groups, and the deep group is further subdivided into extraabdominal, abdominal wall, and intraabdominal (mesenteric, mesocolic, omental, retroperitoneal) subgroups.256,439,452 In this chapter we will focus predominantly on the intraabdominal subgroup of desmoid tumors. Desmoid tumors are uncommon neoplastic lesions, comprising 0.1% of all tumors and 3.5% of fibrous tumors, occurring either sporadically or in association with familial adenomatous polyposis (FAP).439 FAP is inherited in an autosomal dominant fashion, and patients with a phenotypic variant of FAP known as Gardner’s

syndrome may develop desmoid tumors in addition to polyposis coli and colon carcinoma.416,707 The incidence of desmoid tumor in FAP ranges from 3.6% to 34%, and patients with FAP are at an approximately 1000-fold increased risk compared to the general population. Sporadic desmoid tumor occurs more commonly in women by a ratio of 2 to 5 : 1, whereas FAP-associated desmoid tumor occurs with equal gender frequency.439 Both types of desmoid tumor have a peak incidence in the fourth decade of life.439,523 Sporadic desmoid tumors are solitary in over 90% of patients; they are often present in the retroperitoneum, pelvis, and anterior abdominal wall and tend to be large, with a mean diameter of 13.8 cm. In contrast, FAP-associated desmoid tumors are multiple in 40% of patients, are more likely to involve the mesentery and the abdominal wall, and tend to be smaller, with a mean diameter of 4.8 cm.130,385,509,710 Retroperitoneal desmoid tumors are rare and principally originate from the connective tissue of the muscles and their overlying fascia or aponeuroses.393 They tend to be infiltrative, may be asymptomatic, or may have a variable clinical presentation due to ureteral obstruction or bowel invasion.393,434,523,633,660 Minor bone malformations such as exostoses, enostoses, and incomplete spinal segmentation may be seen on imaging in 80% of sporadic desmoid tumor patients compared to 5% of the general population.439,633 The exact etiology of desmoid tumors is not known. Most cases are idiopathic, but genetic abnormalities, trauma (including iatrogenic trauma), and estrogenic hormones are potential etiologic factors.256,439 Histologically, desmoids are composed of uniform, elongated, slender, highly differentiated spindle-shaped fibroblasts that are surrounded by abundant collagen in interlacing bundles.256,439,808 Cellularity is variable, with some portions of tumor showing complete replacement by dense fibrous tissue and other areas revealing stromal myxoid change.256 When large, desmoid tumors may have cystic degeneration or necrosis. No peripheral capsule is present, but instead an infiltrative margin is seen that extends several centimeters beyond the palpable edge of the lesion.439 No correlation between clinical behavior and histologic appearance has been observed.108 Interestingly the growth rate of desmoid tumors parallels the level of endogenous estrogen, which has more effect than progesterone, and between 25% and 75% of desmoid tumors have estrogen and antiestrogen binding sites.439 On CT and MRI, desmoid tumors are typically infiltrative and cross fascial boundaries, have homogeneous low to intermediate attenuation and low to intermediate SI on T1- and T2-weighted images relative to skeletal muscle, are uncommonly calcified, and tend to appear aggressive although they are not malignant (Fig. 55-27).297,467,616,710,886,965 However, because desmoid tumors may have a variable degree of cellularity, fibrosis, myxoid stroma, vascularity, and infiltration, they may be either well defined or poorly defined and may have variable SI.94,129,393,467,477 Slightly high SI components on T1-weighted imaging relative to skeletal muscle are frequently seen.467 The presence of high SI on T2 does not exclude a diagnosis of a desmoid tumor, because immature lesions with higher cellularity and less mature fibrosis can be of higher SI than skeletal muscle on T2. These immature high-SI desmoid tumors are associated with rapid growth on follow-up imaging. Occasionally a low-SI fibrous capsule may be partially or completely visualized on T1, and low-SI bands of collagen may be present on T1- and T2-weighted images.416,710 DWI may be useful to differentiate a desmoid tumor from a malignant soft tissue tumor; the former tends to have a higher mean ADC compared to the latter.627 Both mature and immature desmoid tumors tend to enhance, and the enhancement can be variable.297,467,710 Associated encasement of the bowel or mesenteric vessels or hydronephrosis may also be seen.297

CHAPTER 55  Retroperitoneum

A

1893

B

C FIG 55-27  Extraperitoneal pelvic fibromatosis (desmoid tumor) in 24-year-old man with palpable inguinal mass. A to C, Axial T1-weighted (A), T2-weighted (B), and fat-suppressed contrast-enhanced T1-weighted (C) images through pelvis show infiltrative left pelvic mass with intermediate T1 and intermediate to slightly high T2 SI relative to skeletal muscle, and marked enhancement. Mass infiltrates left bladder wall, left obturator internus muscle, left rectus abdominis muscle, and proximal left thigh musculature, with encasement of left common femoral artery and compression of left common femoral vein.

Imaging features of intraabdominal desmoid tumor that suggest a poor prognosis include a mesenteric mass greater than 10 cm, multiple mesenteric masses, extensive small bowel involvement, and bilateral hydronephrosis.94 Overall, the mean survival rate of patients with intraabdominal desmoid tumors is 5 years. The prognosis of patients with FAP-associated intraabdominal desmoid tumors is significantly worse than those without them.130,439,808 Surgical removal remains the mainstay of treatment of desmoid tumors, despite frequent local recurrence.439,628 Optimally, wide-field surgical resection is performed for small intraabdominal tumors so that tumor-free margins may be achieved.439,808 Large intraabdominal desmoid tumors cannot be completely removed without sacrificing vital structures such as the mesenteric vessels. Retroperitoneal desmoid tumors in particular are rarely completely resectable and often recur locally.393,434,523,633,660 Thus surgical removal of intraabdominal desmoids has a perioperative mortality rate and major morbidity rate of up to 60%, with recurrence in up to 90% of cases, and is not generally recommended unless there is GI or GU tract obstruction, fistula formation, or abscess formation.439,557,675,784 Postsurgical radiation therapy may help reduce the risk of recurrent disease.439,633 After surgery, recurrent desmoids have MRI characteristics similar to those of the original lesion, and the site of recurrence is frequently at the margin of the original lesion. MRI is also useful in evaluating the effectiveness of nonsurgical therapy; effective therapy decreases the size of the mass and increases the amount of low SI on T2-weighted imaging, reflecting increased intratumoral collagen.710 Noncytotoxic drugs, cytotoxic drugs, radiation therapy, and observation are nonsurgical treatment options for desmoid tumors, although

no single approach has been proven consistently effective.439,628,808 Noncytotoxic medications including nonsteroidal antiinflammatory drugs (NSAIDs) are considered by some as first-line treatment for desmoid tumors, particularly if they are small, asymptomatic, and unlikely to cause bowel obstruction.439,736 A trial of watchful waiting may also be implemented, especially in patients with FAP or with lesions that are intimately associated with the mesenteric vessels, because spontaneous regression may occur in 5% to 15% of cases.439,675,784 Cytotoxic medications may be used to induce variable degrees of remission in symptomatic, clinically aggressive, recurrent, nonresectable, or unresponsive desmoid tumors and may achieve a good initial response for inoperable desmoid tumors that have caused progressive bowel or ureteral obstruction.439,628,675,736

PRIMARY RETROPERITONEAL FAT-CONTAINING LESIONS Lipoma Lipomas are benign mesenchymal tumors composed of mature fat and represent the most common mesenchymal neoplasm. Even though retroperitoneal lipomas are rare, they are the most common benign tumor of the retroperitoneum.50,256,598 Variants of lipoma include angiolipoma, myolipoma, angiomyolipoma, myelolipoma, chondroid lipoma, spindle cell/pleomorphic lipoma, and lipoblastoma and are found much less commonly in the abdomen.256 Lipomatous tumors account for half of all soft tissue tumors in surgical series.430 Solitary lipomas occur in the fifth through seventh decades of   life at a significantly younger peak age range compared to that of

1894

PART II  CT and MR Imaging of the Whole Body

well-differentiated liposarcomas. They usually appear during periods of weight gain. Deep lipomas occur more commonly in men.256,430,592 Abdominal lipomas are usually not treated, although surgical removal may be performed if they are large or symptomatic.344 Lipomas are composed of mature adipocytes, similar to normal adult fat, although the cells are slightly larger and slightly more variable in size and shape. Admixed mesenchymal elements such as fibrous connective tissue may be present. Cartilaginous or osseous metaplasia may occur, particularly in large chronic lipomas. Lipomas do not undergo malignant transformation into liposarcomas, although they may occasionally grow to a large size.256,528,592 On CT and MRI, lipomas typically have homogeneous attenuation and SI identical to that of macroscopic fat on all pulse sequences, without enhancing components, in contrast to well-differentiated liposarcomas. In up to 49% of cases, however, a few thin (<2 mm) septae may be seen that have minimal to moderate enhancement, and mild enhancement of a thin fibrous capsule may also be seen.193,296,430,454,592,625,647 Many lipomas also have prominent nonadipose areas that may demonstrate enhancement, and the imaging appearance then overlaps with that of welldifferentiated liposarcomas. These nonadipose components may be seen in up to 31% of lipomas and are secondary to fat necrosis and associated calcification, fibrosis, inflammation, and areas of myxoid change. Lipomas tend to be smaller than well-differentiated liposarcomas, and most remain stable in size.181,256,430

Extrarenal Angiomyolipoma Angiomyolipomas (AMLs) are benign neoplasms composed of ectopic rests of normal tissue in an abnormal arrangement containing fat, abnormal blood vessels, and smooth muscle in varying relative proportions and are most often found in the kidneys. They are rarely encountered in the extrarenal retroperitoneum.187,223,299,324,865 Approximately 90% of AMLs occur sporadically in the kidney, most frequently in middle-aged women. Whereas 20% of patients with renal AMLs have tuberous sclerosis (TS), 80% of TS patients have AMLs, which tend to be multiple and bilateral. Up to 57% of patients with lymphangioleiomyomatosis (LAM) have AMLs.187,299,874 Although benign, some AMLs exhibit aggressive features that may mimic malignancy. These include vascular invasion, lymph node involvement, nonrenal organ involvement, and local recurrence after resection. Multicentric or extrarenal retroperitoneal AMLs are thought to be caused by the congenital  presence of cell precursors in multiple sites or by a form of benign metastases similar to that in benign metastasizing leiomyoma.75 Similar to renal AMLs, extrarenal AMLs have a propensity to undergo spontaneous hemorrhage, with a risk that correlates with the size of the lesion.471,594,695 On CT and MRI, extrarenal AMLs typically appear as masses that contain areas of macroscopic fat. Soft tissue components and vascular components are frequently present, and there is a variable amount of enhancement.41,462,594,604,763,865 The major differential diagnosis is retroperitoneal liposarcoma, but preoperative cross-sectional imaging combined with tissue sampling may permit a nonoperative diagnosis in lesions that pose a diagnostic problem. Surgical resection or arterial embolization may be performed for symptomatic lesions.604,865

Extraadrenal Myelolipoma Myelolipomas are uncommon benign well-circumscribed tumors that are composed of mature adipose cells and hematopoietic tissue. They are of uncertain etiology but usually arise in the adrenal gland. Extraadrenal retroperitoneal myelolipomas are rare, occurring most often in the presacral space. They are more common in women than in men by a ratio of 2 : 1, with a peak occurrence in the seventh decade of life.114,152,266,332,388,659,677,780

Extraadrenal myelolipomas can range in size from 2 to 26 cm, with a mean size of 8.2 cm. Smaller lesions are typically asymptomatic, whereas larger lesions are frequently symptomatic owing to mass effect or hemorrhage.114,266,378,388,677 Associated hemorrhage is more common with larger lesions than with smaller lesions.388 The natural history of extraadrenal myelolipoma is unknown, but adrenal myelolipomas can enlarge with time and become symptomatic.378,546 Surgical removal may be performed when they are symptomatic, large (increased risk of hemorrhage), or when they increase in size over time.677 Histologically, mature fat and focal collections of normal hematopoietic elements are seen in a well-encapsulated lesion.266 On CT and MRI, myelolipomas contain areas with attenuation and SI characteristics similar to macroscopic fat, often with enhancing soft tissue components due to mixed hematopoietic tissue, and less commonly may have foci of calcification. A thin peripheral rim of mild enhancement is often seen on contrast-enhanced CT and T1-weighted imaging. Associated retroperitoneal hemorrhage may occasionally be seen as well.98,378,388,659,696,840 Extraadrenal myelolipomas tend to have smaller amounts of macroscopic fat compared to adrenal myelolipomas.388 99mTc sulfur colloid scintigraphy may be useful for establishing the diagnosis by demonstrating the presence of radiotracer-avid erythropoietic elements within such masses.611

Primary Retroperitoneal Teratoma Primary retroperitoneal teratomas are rare lesions that represent 6% to 11% of primary retroperitoneal tumors.646,911 The majority occur in children. Less than 20% of lesions occur in adults, and they are 2 to 4 times more common in women than in men.156,236,426,646,911 Teratomas in adults have a greater chance of being malignant than those in children (14%-26% vs. 6%-7%, respectively). Teratomas are most common in the ovaries, testes, and anterior mediastinum.646,871 Primary retroperitoneal teratomas tend to be asymptomatic but may cause nonspecific symptoms, as seen with other retroperitoneal masses when large. AFP levels are normal in patients with benign teratomas but may be elevated with malignant teratomas.911 Primary retroperitoneal teratomas develop from totipotential germ cells that have failed to migrate to their normal gonadal locations.646 They may be cystic and/or solid, and frequently contain mature tissues including skin and dermal appendages, cartilage, bone, teeth, or fat.167 Cystic teratomas are benign well-defined lesions that contain multiple solid and cystic areas along with mature tissues, sebaceous material, and/or mucoid fluid.156,167,236,646,801 Solid teratomas are frequently malignant and contain immature embryonic tissue in addition to the mature components.156,646 Retroperitoneal teratomas tend to be located near the upper poles of the kidneys, with preponderance on the left side.197,911 From 60% to 83% of all primary retroperitoneal teratomas have calcifications that are visible on CT; up to 74% of benign teratomas may contain calcifications, and up to 25% of malignant teratomas may also contain calcifications. Therefore the presence of calcifications does not predict benignity.156 On CT and MRI, teratomas are usually heterogeneous well-defined solid and/or multiloculated cystic lesions that may contain fatty components, calcifications, ossifications, or teeth. Other tumoral components have less specific CT and MRI features.59,122,156,196,233,236,292,801,871,911 Occasionally a low-SI fibrous pseudocapsule may be visualized on T1and T2-weighted images, and a dermoid plug or Rokitansky nodule may also be seen along the inner surface of a cystic component, with variable attenuation and SI depending on components present within it.59,122,156,196,871 An almost pathognomonic finding of a mature cystic teratoma is an internal fat-fluid level, with the nondependent fat layer following the attenuation and SI of fat on all pulse sequences. This finding is usually seen in ovarian teratomas but may rarely be present

CHAPTER 55  Retroperitoneum in teratomas developing at other anatomic sites.196,444,646 Other fatty lesions such as a retroperitoneal liposarcoma may extremely rarely demonstrate an internal fat-fluid level.444 The demonstration of fatfluid levels in the peritoneum may be a sign of intraperitoneal rupture of an abdominal teratoma.210 Retroperitoneal teratomas are treated by surgical removal, regardless of whether benign or malignant, because even histologically benign lesions may result in significant morbidity from continued growth.236,911 Patients have an excellent prognosis after the resection of benign retroperitoneal teratomas.646,871

Fetus in Fetu Fetus in fetu (FIF) is a rare entity, infrequently occurring in adults, in which a vestigial vertebrate fetus is enclosed within the abdomen of a normally developing fetus and may mimic a teratoma on cross-sectional imaging. It is thought to arise from abnormal embryogenesis in a diamniotic monochorionic twin pregnancy, in which a malformed monozygotic twin lies within the body of its fellow twin owing to unequal division of totipotential cells of a blastocyst, although others consider it to represent a highly mature teratoma.37,113,313,567,655 Approximately 75% present in infancy, and men are affected up to two times more commonly than women. FIF may be asymptomatic or cause nonspecific symptoms due to mass effect or hemorrhage when large.313,655 In most cases, FIF is found in the upper retroperitoneum or in the nonretroperitoneal portion of the abdomen, often with a capsule and a vascular pedicle, whereas teratomas most commonly occur in the gonads or anterior mediastinum.37,313,646,845 The fetus is usually single and is always anencephalic, although up to 5 fetuses have been reported.313,405 Symptoms are mainly related to mass effects and include abdominal distention, feeding difficulty, emesis, jaundice, and dyspnea. Serum AFP levels may be normal or elevated.125 Malignant degeneration occurs very rarely in FIF, but surgical excision is still the recommended treatment.37,125,655 On CT and MRI a heterogeneous cystic mass containing mature fat and central osseous structures is typically seen, with imaging features similar to a teratoma, as described earlier.37,139,185,280,318,326,619,655,766,812,893 Identification of a fetal vertebral column or fetal long bones is essential to make the diagnosis, although 9% and 17.5% have no well-defined vertebral axis or limbs, respectively, on pathologic examination.37,313,619,655

Hibernoma Hibernomas are very rare benign, painless, slow-growing tumors consisting primarily of brown fatty tissue. They are usually seen in locations where normal brown adipose tissue is present, such as the thigh, shoulder, back, neck, chest, arm (in descending order of frequency), and rarely in the retroperitoneum.103,230,708,738 They occur most commonly during the third and fourth decades of life. A slight female predominance has been noted in the literature by some investigators, and a slight male predominance by others.230,592,708 Hibernomas are well-circumscribed, encapsulated, soft, lobulated masses that usually measure 5 to 10 cm in maximal diameter but can grow to larger sizes, particularly in the retroperitoneum. Histologically they are composed of multivacuolated fat cells with variably eosinophilic cytoplasm and small central nuclei, small granular eosinophilic cells with little or no lipid, and abundant capillaries. Univacuolated fat cells may be seen, often with abundant mitochondria.230,592,708,738 Morphologic variants of hibernoma (in descending order of frequency) are typical, myxoid, lipomalike, and spindle cell variants. The typical variant is composed primarily of brown fat cells and is the most common variant (82% of cases).230 Surgical resection is usually performed for symptomatic lesions, and local recurrence is unlikely with complete resection.592

1895

On CT, hibernomas have attenuation that varies between fat and skeletal muscle. They are well circumscribed and may have a peripheral soft tissue attenuation capsule. They enhance slightly and may contain prominent enhancing vessels.103,738,947 On MRI, hibernomas typically have high SI on T1-weighted images and heterogeneously high SI on T2-weighted images relative to skeletal muscle, may variably lose SI on fat-suppressed T1, and demonstrate variable enhancement, which can be marked.19,32 Prominent branching and serpentine enhancing vessels may also be seen on both CT and MRI.19,592 Retroperitoneal hibernomas are treated by surgical excision, particularly if large or symptomatic.103 Intense FDG uptake with PET has been reported as well.112

Fat Necrosis Retroperitoneal fat necrosis is another fatty process that is characterized by the coalescence of fat cells into fat cysts, sometimes bordered by foreign body giant cell granulomata, foam cells, chronic inflammatory infiltrates, and fibrosis. The most common cause is acute pancreatitis. Other etiologic factors that have been described include trauma, surgery, infection, sclerosing mesenteritis, and chronic hemorrhage or urine leaks. It may also be idiopathic.721,727 Fat necrosis can present as a palpable abdominal mass, mimic other abdominal masses including retroperitoneal well-differentiated liposarcoma, and rarely lead to ureteral obstruction when located in the retroperitoneum.17,295,721,826,847 On CT and MRI a lesion of predominantly fat attenuation or SI lesion is typically seen that contains internal foci of soft tissue attenuation or SI that may enhance. Foci of calcification may also be seen on CT. Fat necrosis tends to remain stable in appearance or decrease in size with time.17,500,826 Treatment is generally conservative. Percutaneous biopsy may be necessary to differentiate from retroperitoneal liposarcoma in an occasional case.17

Pelvic Lipomatosis Pelvic lipomatosis is a rare condition characterized by nonencapsulated fatty tissue hyperplastic overgrowth in the perirectal and perivesical regions of the pelvis, which results in compression of the lower urinary tract and rectosigmoid colon.851 The fatty growth is diffuse and composed entirely of mature fat, indistinguishable grossly and histologically from fatty tissue elsewhere in the body.256 The etiology of pelvic lipomatosis is unknown, but a genetic cause has been postulated.851 Pelvic lipomatosis is usually asymptomatic. African American men are affected in approximately two thirds of cases, with peak occurrence in the third and fourth decades of life.317,851 Associated proliferative cystitis, in particular cystitis glandularis, a premalignant bladder condition, is present in up to 80% of patients.261,304,530 Although the exact etiology of proliferative cystitis is still poorly understood, it is hypothesized to be secondary to lymphatic and venous stasis due to compression of the urinary bladder by the fatty tissue.530,851 The natural history of pelvic lipomatosis is variable. It tends to be relatively asymptomatic even though slowly progressive until severe proliferative cystitis or late complications of renal failure and bladder adenocarcinoma occur.759,851 Early in the disease, mild perineal pain, dysuria, and increased urinary frequency may be the only presenting symptoms and signs. Subsequently nausea, constipation, lower abdominal, flank, or back pain, hematuria, renal insufficiency, and lower extremity edema may present. Hypertension may also be found in up to one third of patients, and hydroureteronephrosis may be present as a result of lower ureteral obstruction by the pelvic fat and mucosal edema related to cystitis at the ureterovesical junctions.851,962 Rarely pelvic lipomatosis may lead to IVC obstruction, with recurrent deep venous thrombosis or even large bowel obstruction.410 On CT and MRI, pelvic lipomatosis appears as increased homogeneous nonencapsulated perivesical and perirectal fatty tissue that

1896

PART II  CT and MR Imaging of the Whole Body

B B

S

S

A

B FIG 55-28  Extraperitoneal pelvic lipomatosis in 54-year-old man with prostate carcinoma. A and B, Axial contrast-enhanced CT (A) and T1-weighted (B) images through pelvis demonstrate increased pelvic fat, with extrinsic compression of sigmoid colon (S) and anterior displacement of urinary bladder (B). Lobulated lesion is present in urinary bladder (not shown) in keeping with associated cystitis cystica/cystitis glandularis.

surrounds the pelvic organs symmetrically and has attenuation and SI similar to macroscopic fat on all pulse sequences (Fig. 55-28). Scattered intratumoral linear foci of fibrosis with soft tissue attenuation and low T1- and T2-weighted SI may also be present.13,39,222,246,816 Superior displacement of the bladder base, a pear-shaped bladder, elongation of the bladder neck and posterior urethra, elevation of the prostate gland, and superomedial displacement of the seminal vesicles may be visualized.165,222,851,962 Smooth extrinsic narrowing of the rectosigmoid colon as well as dilatation and displacement of one or both ureters may also be seen.13,39,851 No definitive treatment for pelvic lipomatosis has been established.304,410 Complete surgical removal of the fatty tissue is difficult because of indistinct surgical margins between the fat and adjacent normal structures and does not always help to improve clinical symptoms or signs.530 Other therapeutic measures including weight loss, antimicrobial medication, corticosteroid administration, radiation therapy, or chemotherapy are ineffective.304,410,530 Although the natural history of pelvic lipomatosis is not entirely clear, approximately 40% of patients will require urinary diversion at a mean of 5 years after diagnosis to prevent renal failure due to obstructive uropathy.410,851 Because patients with pelvic lipomatosis and proliferative cystitis have a higher incidence of bladder adenocarcinoma, periodic bladder cancer screening should be performed.530,794

On CT and MRI a well-circumscribed fusiform or branching mass with attenuation and SI similar to that of macroscopic fat is seen in the pelvic extraperitoneal space, usually along the expected course of a nerve. Internal cylindrical, linear, or serpentine soft tissue attenuation or SI nerve fascicles with a spaghetti-like appearance in longitudinal section and a coaxial cable–like appearance in cross section are also typically seen and are pathognomonic for this condition.526,854 Therapy is generally conservative.854

Pelvic Lipomatosis of Nerve

Overview

Lipomatosis of nerve, also known as fibrolipomatous hamartoma or fibrofatty proliferation of nerve, is a rare benign tumorlike lesion that most often affects the median nerve or its branches. Rarely, involvement of the nerves of the upper and lower extremities and the face has been reported.66,70,109,526,629,682,775 Even more rarely, involvement of the sciatic nerve or lumbosacral plexus may occur and may potentially be mistaken for a well-differentiated retroperitoneal liposarcoma.526,854 The affected nerves are diffusely enlarged by yellow fibrofatty tissue. Histologically, fatty infiltration separates the individual nerve fascicles. The nerve fibers are usually intrinsically normal, although atrophy can occur late in the disease.592

Nonparenchymal cysts and cystic lesions are rare, both in the abdominal retroperitoneum and the pelvic extraperitoneal space. Although controversy regarding the etiology and classification of most of these cystic lesions exists, these lesions can generally be placed into one of several categories (Box 55-2).62,167,296,715,801,938 The first step in evaluation of a cystic abdominal or pelvic lesion is determination of the organ of origin. If a cystic lesion does not arise from a solid organ but from the mesentery, omentum, mesocolon, or retroperitoneum, the differential diagnosis includes primary mesenteric, omental, mesocolic, or retroperitoneal cysts, as well as benign and malignant cystic neoplasms and congenital anomalies, among other

Retroperitoneal Lipomatosis Retroperitoneal lipomatosis is a condition in which excessive fat deposition in the retroperitoneum can exert a mass effect, with displacement of normal structures.234 The increase in retroperitoneal fat may be idiopathic in nature or associated with pelvic lipomatosis, obesity, corticosteroid administration, Cushing’s syndrome, or lipoblastic lymphadenopathy.222,234,647 On CT and MRI, homogeneous symmetrically distributed fat with attenuation and SI similar to that of macroscopic fat elsewhere in the body is demonstrated. Associated mass effect upon the kidneys, ureters, urinary bladder, and rectosigmoid colon may be present as well.222

RETROPERITONEAL NONPARENCHYMAL CYSTS AND CYSTIC LESIONS

CHAPTER 55  Retroperitoneum

1897

BOX 55-2  Differential Diagnosis of Nonparenchymal Retroperitoneal and Extraperitoneal

Cysts and Cystic Lesions62,167,296,715,801,938 Lymphatic origin Simple lymphatic cyst Lymphangioma Mesothelial origin Simple mesothelial cyst Benign cystic mesothelioma Enteric origin Enteric duplication cyst Enteric cyst Foregut origin Bronchogenic cyst Cystic pulmonary sequestration Hindgut origin Tailgut cyst Urogenital origin Müllerian urogenital cyst Pronephric urogenital cyst Mesonephric urogenital cyst Metanephric urogenital cyst Other developmental origin Epidermoid cyst Dermoid cyst Mature cystic teratoma Lymphangioleiomyomatosis (LAM) Neurogenous choristoma Nonpancreatic pseudocysts and loculated fluid collections Simple Proteinaceous Hemorrhagic Bilious Chylous

differential diagnostic possibilities. On CT and MRI the thickness of the cyst wall; the presence of internal septations, calcifications, or fat; and the attenuation, SI, and enhancement characteristics of the cyst contents are useful imaging features to help differentiate among the various cystic lesions. Because of overlap in the imaging features of different kinds of cysts and other cystic lesions, histologic analysis is usually required to establish a definitive diagnosis.167,715,801,938 The most common type of nonparenchymal abdominal “cyst” is the lymphangioma, followed by a nonpancreatic pseudocyst, an enteric duplication cyst, a simple mesothelial cyst, and an enteric cyst.801 All types of nonparenchymal abdominopelvic cysts may present either as large painful masses or be detected in asymptomatic individuals.715 Large nonparenchymal abdominopelvic cysts are generally treated with surgical removal to prevent complications such as rupture, hemorrhage, torsion, superinfection, or interval growth and exclude a cystic malignancy.167 Small asymptomatic nonparenchymal abdominopelvic cysts with no aggressive imaging features can be followed with imaging.

Lymphatic Cyst Simple lymphatic cysts are congenital in origin and usually small (1-5 cm) and unilocular. They tend to remain stable in size and are asymptomatic.79,167 Histologically they are lined by flat endothelial cells

Uriniferous Enteric Infectious Inflammatory noninfectious Malignant Neoplastic origin Metastatic disease Sarcomas Lymphoma Extragonadal germ cell tumor (EGCT) Neurogenic tumors Pseudomyxoma retroperitonei Benign cystic mesothelioma Mucinous cystadenoma/cystadenocarcinoma Serous cystadenoma/cystadenocarcinoma Solitary fibrous tumor (SFT) Extra-gastrointestinal stromal tumor (GIST) Extraarticular pigmented villonodular synovitis (PVNS) Cystic or necrotic lymphadenopathy Metastatic disease Lymphoma Mycobacterial infection Bacterial infection Cat-scratch disease Whipple’s disease Other bacterial infections Fungal infection Cavitary lymph node syndrome (celiac disease) Systemic lupus erythematosus Miscellaneous Retained surgical sponge (gossypiboma)

with a wall that contains smooth muscle fibers, lymphoid tissue, lymphatic space, and occasional foam cells containing lipoid material.168 On CT and MRI they are of simple fluid attenuation with minimal if any wall enhancement.952

Lymphangioma Lymphangiomas are uncommon cystic lesions that most often occur in the neck (75%) and axillary regions (20%). The remaining 5% can occur elsewhere in the body but are most common in the abdomen and very rare in the pelvis.79,413,641,891 The most common location of an abdominal lymphangioma is in the mesentery, followed by the omentum, mesocolon, and retroperitoneum.256,413,715 Rare multisystemic involvement and extensive intraabdominal cystic lymphangiomatosis may occur and has a poor prognosis.79 Up to 40% of retroperitoneal and other abdominal lymphangiomas are found in older children or adults, and the remaining 60% occur in children. Lymphangiomas of the neck and axillary regions are more frequently found in infants and children younger than 2 years.79,128,256,321 Lymphangiomas tend to occur slightly more commonly in men than in women.167,256,715 The etiology and pathogenesis of lymphangioma is not entirely clear. Some consider them to be acquired malformations resulting from obstruction of lymphatic vessels, whereas others consider them

1898

PART II  CT and MR Imaging of the Whole Body

to be congenital malformations of lymphangiectasia related to failure of communication with the lymphatic system.79 Histologically, lymphangiomas are large, thin-walled, usually multiloculated, cystic lesions that are lined by attenuated endothelium resembling that of normal lymphatics, with a stroma formed by a delicate meshwork of collagen punctuated by small lymphoid aggregates. Lymphangiomas are classified into capillary (simple), cavernous, and cystic (hygroma) types, depending on the size of the lymphatic spaces. Capillary lymphangiomas are composed of small thin-walled lymphatics, cavernous lymphangiomas are composed of larger lymphatic channels with adventitial coats, and cystic lymphangiomas are composed of large macroscopic lymphatic spaces with intervening collagen and smooth muscle.79,256,801 The fluid contents are predominantly chylous but may be serous or hemorrhagic.256,801 Retroperitoneal lymphangiomas are usually asymptomatic and  discovered either incidentally on cross-sectional imaging or during surgery for unrelated conditions.79,167,715 This is in contradistinction to lymphangiomas of the mesentery, omentum, or mesocolon, which tend to be symptomatic.167 On CT and MRI, lymphangiomas usually appear as large, cystic, unilocular or multilocular thin-walled lesions with multiple cystic components of various sizes that tend to invaginate between rather than displace structures, and may cause bowel dilatation due to obstruction (Fig. 55-29).155,167,292,413,586,610,801 Serous fluid contents have water attenuation and low SI on T1 and very high SI on T2 relative to skeletal muscle.193,292,586 Sometimes the fluid contents may be chylous with water or fat attenuation, increased SI on T1-weighted images, and high SI on T2-weighted images, or proteinaceous/hemorrhagic with high attenuation, increased SI on T1, and high SI on T2.155,222,801 Loss of SI of the internal fluid contents from in-phase to out-of-phase T1-weighted gradient echo images may sometimes be seen in chylous lymphangiomas of the mesentery.949 We have similarly observed this finding in chylous lymphangiomas of the retroperitoneum on MRI (unpublished data). Occasionally a small amount of fat may be detected within the septations of a lymphangioma, with high SI noted on T1 and loss of SI with both fat-suppressed and out-of-phase T1-weighted images.641,715,733,768 Rarely calcifications may be present, which appear as foci of very high attenuation and very low SI on both T1 and T2.155,290 Mild septal or wall enhancement may be present. With complications of infection or hemorrhage, the outer wall and internal septations tend to be thicker.891

Complete surgical excision of retroperitoneal lymphangioma may performed if large or symptomatic.79,641 However, in contrast to simple lymphatic cysts, which are generally easily enucleated, lymphangiomas may be adherent to vital intraabdominal structures, making complete excision difficult or impossible.167,801 Furthermore, recurrence after complete resection is rare but is common following incomplete resection.167,641 Alternatively, image-guided percutaneous drainage and sclerotherapy can be used for treatment.755

Mesothelial Cyst Simple mesothelial cysts are congenital lesions due to incomplete fusion of the mesothelial lined peritoneal surfaces.79,715,801 Most occur in young and middle-aged women and are usually located within the mesentery, although they can also occur in the omentum and very rarely in the retroperitoneum. Mesothelial cysts do not show interval growth and are usually asymptomatic.167,715,783,900 Histologically, mesothelial cysts are thin-walled unilocular cysts lined by flat, cuboidal, or columnar mesothelial cells and contain serous fluid.167,801 On CT and MRI, mesothelial cysts have no internal septations and typically have water attenuation, low T1-weighted SI, and very high T2-weighted SI fluid content. Mural calcification is rarely present, and minimal or no wall enhancement is generally seen.715,783,801 Surgical removal or percutaneous drainage with or without sclerotherapy may be performed for symptomatic lesions.8

Benign Cystic Mesothelioma Benign cystic mesothelioma is a rare lesion that originates from mesothelial cells lining the serosal cavities of the body. It almost always occurs in young to middle-aged women.256,421,749,801 It is usually found along the surfaces of the pelvic viscera but may less commonly occur in the retroperitoneum.480,595,631 The pathogenesis is controversial; some believe it is a benign neoplasm and others believe it is a reactive process.729,793 Unlike malignant mesothelioma, benign cystic mesothelioma is not related to prior asbestos exposure. Histologically a unilocular or multilocular thin-walled cystic lesion lined by mesothelial cells is seen, along with a delicate fibromuscular stroma and occasional inflammatory changes.383,722,801,918 On CT and MRI a nonspecific thin-walled multiloculated cystic lesion with water attenuation, low T1-weighted SI, and very high T2-weighted SI is usually seen.421,631,801,938 Occasionally calcification may be visualized.288 Surgery is the recommended therapy for benign

*

*

L

L

A

B FIG 55-29  Retroperitoneal lymphangioma in 90-year-old man with renal cell carcinoma. A and B, Axial T1-weighted (A) and fat-suppressed T2-weighted (B) images show well-circumscribed lobulated retroperitoneal lesion with low T1 SI and very high T2 SI relative to skeletal muscle, with several thin low-T2-SI internal septations. Note portion of lesion that invaginates between abdominal aorta (*) and lumbar spine (L), which is characteristic of lymphangioma.

CHAPTER 55  Retroperitoneum cystic mesothelioma, although recurrence is common even after aggressive surgery, particularly in women.480,631,641,793

Mucinous Cystadenoma and Cystadenocarcinoma Primary mucinous cystadenomas and cystadenocarcinomas are rare retroperitoneal neoplastic cystic lesions that most often occur in women who have normal ovaries, with a mean age of presentation in the fifth decade of life. Very rarely they can occur also in elderly men.227,265,676,723,846 The exact histogenesis of these lesions is unclear, but it is possible they arise from ectopic ovarian tissue, teratomas, remnants of the embryonic urogenital apparatus, or invagination of the peritoneal mesothelial layer that undergoes mucinous metaplasia with cyst formation.115,657,804,938 Primary mucinous tumors of the retroperitoneum can be classified into three clinicopathologic types: mucinous cystadenoma, mucinous cystic tumor of borderline malignancy, and mucinous cystadenocarcinoma. Histologically, mucin-producing or ciliated epithelial cells with benign, borderline malignant, or frankly malignant morphology are seen with mucinous fluid content.265,353,804,842,846,938 Early diagnosis is important because mucinous cystadenomas have malignant potential.265 On CT and MRI these lesions are nonspecific unilocular cystic lesions. Internal septations, papillary projections, or soft tissue solid enhancing components may sometimes be present.265,353,535,750,842,846,937,938 These lesions tend to be large, with a mean size of 16 cm (range, 3-26 cm).850 Surgical resection is the treatment of choice, although chemotherapy is also used.115,842,850

Serous Cystadenoma and Cystadenocarcinoma Primary serous cystadenoma and cystadenocarcinoma are extremely rare retroperitoneal lesions that have been reported a few times in the literature, most often in women.107,354,374,443,872 The origin of these epithelial neoplasms is unknown but may be similar to the histogenesis of primary retroperitoneal mucinous cystadenoma and cystadenocarcinoma described earlier. On cross-sectional imaging, a nonspecific retroperitoneal cystic mass is typically seen that may contain septations or enhancing solid mural nodules.107,354,374,443 In one report, retroperitoneal lymphadenopathy was the dominant finding.228 Surgical resection is generally performed, sometimes in conjunction with chemotherapy.228,374

Enteric Duplication Cyst Enteric duplication cysts are true diverticula of the bowel that form during gestation and contain an enteric mucosa, a muscle layer, and a nervous plexus.801 They are attached to the mesenteric side of normal bowel but occasionally migrate into the mesentery.715,801 Retroperitoneal enteric duplication cysts are extremely rare and usually detected during prenatal ultrasonography, and much less commonly during adulthood.494,873 They are thick-walled unilocular cysts with predominantly serous contents, although chylous or hemorrhagic content may be present.715,801 The wall of an enteric duplication cyst contains all normal enteric layers, including mucosa (usually gastric or pancreatic small bowel epithelium), circular and longitudinal muscle layers, and the mesenteric plexus.167,801 On CT and MRI a thick enhancing wall is usually visualized, and the contents are usually serous with water attenuation, low SI on T1, and very high SI on T2, rarely with calcification.715,801,829 Whereas nonretroperitoneal enteric duplication cysts are generally round in shape, retroperitoneal enteric duplication cysts are more frequently tubular in configuration.873

Enteric Cyst Enteric cysts result from migration of acquired bowel diverticula into the mesentery and are lined by enteric mucosa only, without reduplication of the bowel wall. Pathologically they are thin smooth-walled

1899

unilocular serous cysts lined with enteric epithelium and have a thin outer fibrous wall without muscle layers or a neural plexus.167,801 On CT and MRI, serous fluid contents are present with water attenuation, low T1-weighted SI, and very high T2-weighted SI, occasionally with septations but usually without a discernible wall.801

Bronchogenic Cyst Bronchogenic cysts are a subtype of congenital bronchopulmonary foregut malformation and are usually located in the pulmonary parenchyma or mediastinum but can very rarely occur in the retroperitoneum.298,401 A retroperitoneal bronchogenic cyst is formed when an abnormal bud of the tracheobronchial tree develops during fusion of the pleuroperitoneal membranes, with subsequent migration into the abdominal cavity.589 Pathologically, bronchogenic cysts are thinwalled, usually unilocular, cystic masses of nonfunctional pulmonary tissue composed of smooth muscle, cartilage, and glands, with a lining of ciliated columnar respiratory epithelium.95,540,589 The majority of retroperitoneal bronchogenic cysts are smaller than 5 cm and found within the upper retroperitoneum adjacent to the diaphragmatic crura, most frequently to the left of midline.95,483,589,913 Bronchogenic cysts are generally asymptomatic unless large, secondarily infected,  or ruptured.95 On CT and MRI a sharply marginated, thin-walled, round or lobulated cystic lesion is typically visualized.91,95,99,401,540 The contents are usually simple fluid, although fluid-fluid levels may be present if the contents are hemorrhagic, proteinaceous, or mucinous.95,401,540,589,602,938 Mural calcification or very-high-attenuation layering milk of calcium may sometimes be visualized as well.310,540 There is either minimal or no rim enhancement.95,540 Treatment for symptomatic lesions is surgical removal, and the prognosis is excellent.401

Extralobar Pulmonary Sequestration Pulmonary sequestration is another subtype of bronchopulmonary foregut malformation consisting of abnormal lung tissue that lacks a normal communication with the tracheobronchial tree and pulmonary arterial tree and receives blood supply from the systemic arteries. It is thought to form from anomalous dorsal budding of the primitive foregut.298,533,740 Pulmonary sequestrations can be divided into intralobar (75%) and extralobar (25%) types, where the former is invested in the visceral pleural covering of the surrounding lung tissue and the latter has its own visceral pleural covering.245 Extralobar pulmonary sequestration (EPS) usually presents in the first decade of life, most often within the first 6 months of life, although 10% are asymptomatic and may be detected incidentally on crosssectional imaging later in life.209,343,740 EPS is located in the left pleural cavity adjacent to the left lower lobe of the lung in up to 85% of cases, but subdiaphragmatic localization in the retroperitoneum can occur in up to 15% of cases.40,141,232,303,422,533,711,752,941 Pathologically, pulmonary sequestrations are composed of mature well-organized pulmonary and bronchial tissue and are covered by visceral pleura.298 On CT and MRI a nonspecific well-circumscribed low-attenuation mass with low T1-weighted and high T2-weighted SI is generally seen, sometimes with foci of cystic change or calcification. Increased attenuation or T1-weighted SI may occur if proteinaceous or mucinous fluid contents are present.40,232,303,422,533,711,752,941 Contrast enhancement is variable but is usually peripheral in distribution. Feeding systemic arterial branches to retroperitoneal extralobar pulmonary sequestrations are not usually well visualized on cross-sectional imaging, which makes a prospective diagnosis difficult.303 Occasionally, retroperitoneal sequestrations can be completely cystic, mimicking simple cysts on crosssectional imaging.27 Surgical resection may be performed if the lesions are symptomatic.232

1900

PART II  CT and MR Imaging of the Whole Body

Epidermoid and Dermoid Cysts Epidermoid and dermoid cysts are rare congenital lesions that result from closure defects of the ectodermal tube, with inclusions of skin as well as skin appendages such as sweat glands, hair follicles, or sebaceous glands in the case of dermoid cysts. They can occur anywhere in the body, most often in women, and rarely occur in the presacral space.358,870,944 The cysts can communicate with the skin surface, causing a characteristic postanal dimple.778 Histologically they have a wall composed of stratified squamous epithelium, and the internal contents consist of a mixture of desquamated debris, cholesterol, keratin, and water. In dermoid cysts, skin appendages are also seen.150,938 Although most lesions are benign, malignant degeneration to squamous cell carcinoma may rarely occur.939 On CT and MRI, epidermoid cysts generally appear as thin-walled unilocular cystic lesions.938,944 Multiple small foci of low SI on T2-weighted imaging may sometimes be seen as well and are due to aggregates of keratinous material.944 Dermoid cysts appear as thin-walled cystic lesions that may have increased attenuation and high T1-weighted SI of fluid contents as well as thin calcifications.150 Foci of malignant transformation may appear as enhancing irregular wall thickening or polypoid soft tissue.939 Surgical resection for epidermoid and dermoid cysts is the recommended treatment, especially if symptomatic or if malignant transformation is identified.150

Tailgut Cyst Tailgut cysts (also known as retrorectal cystic hamartomas) are rare congenital multicystic lesions that arise from vestiges of the embryonic hindgut; they are most common in middle-aged women, with a mean

age of 35 years. Most occur within the retrorectal space, although they can occur less frequently anterior to the rectum or in the subcutaneous tissue of the sacrococcygeal or anorectal locations.311,394 They are usually asymptomatic but may be associated with abdominal pain or perirectal symptoms due to mass effects.163,311,368,678 Tailgut cysts are multiloculated cystic lesions with thin walls and a glistening lining and are filled with mucin and debris.311,520,563 Histologically these cysts are typically lined by several different types of epithelium, predominantly mucin-secreting columnar epithelium, often with areas of squamous and transitional epithelium as well.311,368,520,678 Although most lesions are benign, malignant degeneration may rarely occur, most often to adenocarcinoma and less often to carcinoid or neuroendocrine malignancies.394,487,531,580,581,678 On CT, tailgut cysts are well-defined, thin-walled, multiloculated cystic lesions with variable attenuation that ranges from water to soft tissue attenuation and infrequently have thin calcifications. Wall thickening and surrounding inflammatory change may be present if superinfection is present. Invasion of contiguous structures and loss of discrete margins may indicate malignant degeneration.163,368,484,487,639,938 On MRI, tailgut cysts usually appear as well-defined multilocular cystic lesions that usually have low T1-weighted and very high T2-weighted SI fluid contents, although high T1 and variably decreased T2 SI  may occur if the fluid is mucinous, proteinaceous, or hemorrhagic  (Fig. 55-30).4,400,484,487,580,943 Sometimes a dominant cyst with a small peripheral cyst may be seen.943 Foci of malignant transformation or fibrous tissue in tailgut cysts may appear as enhancing irregular wall thickening or polypoid soft tissue.487,580 Surgical resection is the treatment of choice to prevent possible malignant transformation or alleviate associated symptoms.311,532,678,938

A

B FIG 55-30  Extraperitoneal pelvic tailgut cyst in 43-year-old woman. A and B, Axial fat-suppressed T1-weighted (A) and sagittal T2-weighted (B) images through pelvis demonstrate well-circumscribed multiloculated lesion (arrows) within presacral space with variable T1 and T2 SI within locules due to variable amounts of proteinaceous or hemorrhagic fluid content. Note incidentally detected uterine adenomyosis on sagittal image as manifested by thickened junctional zone that contains multiple cystic foci corresponding to trapped endometrial glands.

CHAPTER 55  Retroperitoneum Urogenital Cyst Urogenital cysts are rare cystic lesions that develop from vestigial remnants of the urogenital system and can be classified into pronephric, mesonephric, metanephric, and müllerian types based on their embryonic lines.466,542 Müllerian cysts are extremely rare benign lesions that are the most commonly documented subtype of retroperitoneal urogenital cyst.466 They usually occur in obese women in the second through fifth decades of life who are on hormonal therapy for menstrual irregularity and are believed to be related to growth of an aberrant müllerian duct remnant in response to the abnormal hormonal stimulation.466,796 Histologically the walls contain well-developed smooth muscle and are lined by cuboidal to columnar epithelial cells with cilia.420 On CT and MRI, nonspecific large unilocular or multilocular thinwalled cystic lesions are typically seen.466,938 Surgical removal is the usual treatment.938

Nonpancreatic Pseudocyst Nonpancreatic pseudocysts are unrelated to pancreatitis but are thought to develop from prior traumatic hematomas or abscesses that did not resorb.167,801 Pathologically these lesions are usually septated cystic masses with either hemorrhagic or purulent contents surrounded by a thick fibrous wall without an inner epithelial lining, histologically similar to pancreatic pseudocysts. Unlike pancreatic pseudocysts, they are not associated with high levels of amylase or lipase in the cystic fluid.715,801,938 On CT and MRI, one usually sees a thin- or thick-walled cystic lesion that may contain increased attenuation or high T1- and T2-weighted-SI hemorrhagic, proteinaceous, or purulent contents, sometimes with associated fluid-fluid levels and internal septations. Calcifications may sometimes be present.705,715,801,938 Often, peripheral enhancement is seen.

RETROPERITONEAL FLUID COLLECTIONS Overview Retroperitoneal fluid collections may result from neoplastic, infectious, inflammatory, and traumatic causes.262,425 Fluid collections may be classified as either transudative or exudative based on biochemical parameters such as the serum-fluid albumin gradient or total protein content. Cirrhosis, nephrotic syndrome, congestive heart failure, and low protein states are common causes of transudative fluid collections, whereas malignancy, infection, and pancreatitis are the common causes of exudative fluid collections.296 Retroperitoneal fluid collections may also be classified as simple, proteinaceous, hemorrhagic, bilious, chylous, uriniferous, enteric, infectious, inflammatory, or malignant, based on the composition of the fluid and the underlying etiology.62,296 Fascial planes or adhesions tend to confine retroperitoneal fluid collections to their compartment of origin.10,262 However, large or rapidly developing fluid collections may decompress along laminated, variably fused, retroperitoneal fascial planes.262,350,351 With severe inflammation, infection, or hemorrhage, transfascial spread between compartments as well as extension into the psoas muscle and abdominal wall musculature may occur, frequently with obliteration or displacement of adjacent structures or thickening of surrounding soft tissue planes.10,425 Nonloculated fluid collections flow to the most dependent position of the compartment they are located in, are shaped by the surrounding structures, and do not deform the normal organs.62 On CT and MRI, simple fluid collections and transudative fluid collections have water

1901

attenuation, low SI on T1-weighted images, and very high SI on T2-weighted images, whereas proteinaceous fluid collections and exudative (including infectious, inflammatory, or malignant) fluid collections tend to have greater attenuation than water, variably increased SI on T1, and variably decreased SI on T2.62,198,296,425,643

Hemorrhage/Hematoma Retroperitoneal hemorrhage may be secondary to rupture or leak of an aneurysm or vascular malformation, anticoagulant therapy, a  bleeding diathesis, trauma, arterial catheterization, underlying parenchymal organ pathology with rupture, or may be spontaneous.10,151,153,198,262,296,425,815 Spontaneous retroperitoneal hemorrhage classically originates in the posterior pararenal space and may extend into the properitoneal fat, pelvis, psoas muscle, or abdominal wall musculature.10 Traumatic retroperitoneal hemorrhage tends to be largely confined to the retroperitoneal interfascial planes and tends to be uncontrollable when extension into the subfascial plane is visualized.351 Extraperitoneal hemorrhage related to trauma is most commonly seen in the pelvis, second most commonly seen in the lateral aspects of the retroperitoneum, and least commonly seen in the central prevertebral aspect of the retroperitoneum.151 Retroperitoneal hematoma volume (≥1600 mL) is an independent predictor of decreased overall survival. Therefore surgical intervention may be warranted when a large retroperitoneal hematoma is present.499 Most AAAs tend to bleed posteriorly, with confinement either by the psoas space or extension into the left posterior interfascial retrorenal plane, whereas the IVC often bleeds directly into the right posterior interfascial retrorenal plane. In such cases, hemorrhage is often present in the perirenal spaces as well.672 Aortic hemorrhage can also extend into the pelvis via the combined interfascial plane and present as an inguinal mass, or into the anterior interfascial retromesenteric plane and cause obstructive jaundice or duodenal or colonic obstruction.262,671 On CT, acute clotted hemorrhage typically has high attenuation (45-70 Hounsfield units [HU]), whereas surrounding areas of acute nonclotted hemorrhage or more chronic hemorrhage have either lower attenuation (20-45 HU) or fluid attenuation (0-20 HU) (see Figs. 55-5 to 55-7 and 55-11).206,425,636,756,938 This is the basis for the sentinel clot sign, where areas of higher-attenuation acute clotted hemorrhage are more likely to indicate anatomic sites of hemorrhage.636 On MRI the appearance of hemorrhage is dependent upon on its age (Table 55-2).86,198 Hyperacute hemorrhage representing intracellular oxyhemoglobin that is minutes to hours old has low to intermediate SI on T1-weighted images and high SI on T2-weighted images.86,925 Acute hemorrhage representing intracellular deoxyhemoglobin that is hours to about 3 days old has low SI on T1 and T2.44,86,131 Subacute hemorrhage representing early intracellular methemoglobin that is about 3 to 7 days old has high SI on T1 and low SI on T2. Subacute hemorrhage representing late extracellular methemoglobin that is 1 week to several months old has high SI on T1- and T2-weighted images (Fig. 55-31).44,86,258 Chronic hemorrhage representing ferritin or hemosiderin that is months to years old has very low SI on T1 and T2.86,87 It is uncommon to perform an abdominal or pelvic MRI examination to evaluate for a hyperacute or acute hemorrhage. Thus most hematomas seen on abdominopelvic MRI examinations are either subacute or chronic. A subacute hematoma may show two outer characteristic layers of SI: a thin peripheral rim with low SI on all pulse sequences, corresponding to hemosiderin, and an inner peripheral high-SI zone on T1-weighted images due to methemoglobin. The central component of the hematoma has variable SI based on the amounts of seroma, deoxyhemoglobin, and methemoglobin present.198,276 This appearance of the periphery of a fluid collection on T1 is known as the concentric ring sign and is pathognomonic for a subacute hematoma.276,773

1902

PART II  CT and MR Imaging of the Whole Body

TABLE 55-2  MRI of Hemorrhage-Hematomas Age

Time Range

Principal

T1-Weighted SI

T2-Weighted SI COMPONENT

Hyperacute Acute Early subacute Late subacute Chronic

Minutes-hours Hours-3 days 3-7 days 1 week-months Months-years

IC oxyHb IC deoxyHb IC metHb EC metHb hemosiderin/ferritin

Low-int. Low High High Very low

High Low Low High Very low

EC, extracellular; Hb, hemoglobin; IC, intracellular; int., intermediate.

IP

A

C

B FIG 55-31  Extraperitoneal pelvic hematoma in 17-year-old male with lower extremity numbness for 2 weeks following trauma. A, Axial fat-suppressed T1-weighted image shows well-defined lesion anterior to right iliac bone that has central intermediate SI and peripheral high-SI rim (due to subacute methemoglobin) relative to skeletal muscle. Note anterior displacement of iliopsoas muscle (IP). B, Sagittal MIP image during arterialphase contrast-enhanced magnetic resonance angiography demonstrates enhancing pseudoaneurysm (arrow) of anterior division of right internal iliac artery. C, Axial fat-suppressed contrast-enhanced T1-weighted image reveals focal enhancement within medial portion of pseudoaneurysm (arrow).

A “hematocrit effect” with fluid-fluid levels due to layering of heavier cellular elements of blood located dependent to liquid supernatant may be seen on CT or MRI; this is most frequently seen in  the setting of anticoagulation therapy or coagulopathy (see Fig. 55-11).206,207,597 Active arterial hemorrhage is recognized as a focus or jet of very high attenuation (mean, ≈130 HU) or high SI on contrastenhanced CT and T1-weighted MRI, respectively, similar to the attenuation and T1-weighted SI of intraarterial contrast material, which does not conform to the lumen of a vessel. When present, this generally indicates a need for immediate supportive, angiographic, and/or surgical therapeutic intervention.296,359,588,756 When active arterial hemorrhage is present within a liquefied hematoma that has a hematocrit effect, a signal flare phenomenon may also be seen as a linear very-high-attenuation focus in the dependent layer, as well as very-high-attenuation foci at the

boundary of the two fluid-fluid levels. This is due to movement of veryhigh-attenuation extravasated contrast material through the dependent cellular layer to the boundary of the cellular and liquid layers, based on differences in specific gravity.334 As a hematoma matures and becomes smaller, acutely clotted hemorrhage lyses and decreases in attenuation, approaching the attenuation of water, and the SI of the core diminishes while the outer rim becomes very hypointense on both T1- and T2-weighted imaging secondary to hemosiderin deposition.64,198,206 Occasionally a hemorrhagic tumor may be difficult to differentiate from a nonneoplastic hematoma. However, hemorrhagic neoplasms typically have enhancing  soft tissue components, whereas hemorrhagic lesions with a large component of high attenuation or high T1 SI are usually benign hematomas.198

CHAPTER 55  Retroperitoneum Biloma Intraperitoneal bilious fluid collections (bilomas) may be caused by iatrogenic, traumatic, or spontaneous rupture of the biliary tree. Rarely, retroperitoneal bilomas may form following trauma, iatrogenic injury, or spontaneous rupture of the gallbladder or biliary tree.88,137,327,823 Some 70% of bilomas are localized to the right upper quadrant, whereas the remaining 30% develop in the left upper quadrant.896 Although usually asymptomatic, they may present with symptomatic bile peritonitis. On CT and MRI, bilious fluid demonstrates water attenuation, variable SI on T1-weighted images, and high SI on T2-weighted images, similar to the SI of gallbladder fluid.62,88,823,896 Both gadoliniumand manganese-based MRI contrast agents are available that are excreted through the biliary system. A delayed enhanced MRI examination using one of these agents may be useful to confirm that a localized fluid collection is composed of bile and to identify the site of bile leak.902 Most bilomas can be treated with percutaneous catheter drainage and repair of the underlying site of bile leak.896

Chyloretroperitoneum The very rare chyloretroperitoneum mainly occurs after retroperitoneal surgery and is due to collections of lymphatic fluid or chyle in  the retroperitoneum.15,161,296,606,620,803,869 However, any source of lymphatic vessel obstruction or lymphatic leakage, such as in the setting of trauma, can potentially lead to chyloretroperitoneum. Although  a small amount of lymphatic leakage due to injury to the extensive retroperitoneal lymphatic plexus usually occurs with every retroperitoneal operation, spontaneous closure usually occurs without complication.161,457,539,914 The loss of chyle can lead to increased morbidity and mortality secondary to serious nutritional deficits, large fluid volume losses, electrolyte depletion, and lymphocytopenia with an increased incidence of secondary infection.161,457 Chyloretroperitoneum is generally treated conservatively with watchful waiting and removal of underlying malignancy if present.161,803 Symptomatic chyloretroperitoneum may be treated with a low-fat medium-chain-triglyceride diet or  total parental nutrition to decrease lymphatic flow, drainage with or without sclerotherapy, or treatment of a causative primary tumor if

A

1903

present.15,161,457,469,606 Somatostatin analogues may also be effective in reducing chylous leakage and closing lymphatic fistulas.469,606 Profuse or persistent chronic lymphatic drainage may require surgical ligation or placement of a peritoneovenous shunt because untreated chronic chylous drainage can have up to a 50% mortality rate.73,161,457,869 On CT and MRI, chyle usually has water attenuation, low SI on T1-weighted images, and very high SI on T2-weighted images.425,539,606,869,914 A fat-fluid level due to the high fat content may potentially be identified in chyloretroperitoneum, as has been reported in chylous ascites, and is nearly pathognomonic.305,560,680 Chemical shift imaging may also potentially reveal the presence of microscopic lipid within the chylous fluid, although this has not previously been reported. Delayed enhancement of the fluid following administration of IV contrast material or following pedal lymphangiography may also be observed.459,901

Lymphocele Lymphoceles are fluid-filled cystic lesions that usually occur at least 3 to 4 weeks after pelvic or retroperitoneal lymphadenectomy for urologic or gynecologic malignancies or after renal transplantation.  They occur in up to 30% of patients who undergo radical lymphadenectomy and up to 18% who undergo renal transplantation, where they are due to surgical transection or inadequate ligation of lymphatic vessels.89,176,665,747 Small lymphoceles are often asymptomatic and spontaneously resorb.741 However, large retroperitoneal lymphoceles may be symptomatic and may lead to complications related to mass effect upon adjacent structures.665,741,890,920,968 Histologically the wall of a lymphocele consists of dense fibrotic tissue without a lining of epithelial cells, and the internal fluid is typically clear or yellow with biochemical characteristics similar to serum, although triglyceride levels may be higher than those in serum.381,398,665 Leukocytes with lymphocyte predominance and fat globules can also be seen.381 Treatment with percutaneous or surgical drainage is performed for symptomatic lymphoceles, sometimes in conjunction with sclerotherapy.111,342,398,476,577,968 On CT and MRI a water attenuation, low T1-weighted and very high T2-weighted-SI thin-walled cystic lesion is typically seen, usually adjacent to surgical clips in the setting of prior lymphadenectomy  (Fig. 55-32). The attenuation values of the internal fluid may sometimes be slightly negative owing to the lipid content.381,890 Internal

B FIG 55-32  Extraperitoneal pelvic lymphocele in 53-year-old man status post cystectomy, urinary diversion, and pelvic lymph node dissection for bladder carcinoma. A and B, Axial contrast-enhanced CT (A) and axial T2-weighted (B) images through pelvis demonstrate small unilocular lesion (arrows) medial to left external iliac vessels, with fluid attenuation and very high T2 SI relative to skeletal muscle. Note adjacent very-highattenuation surgical clip (short arrow) on CT image, with associated streak artifact.

1904

PART II  CT and MR Imaging of the Whole Body

septations may be present, and mural calcification may rarely be seen.381,398 Increased attenuation, variably increased T1-weighted SI, nonenhancing internal debris, and a thickened enhancing rim can be seen if the internal fluid is infected or proteinaceous.381,890

Uriniferous Fluid/Urinoma Uriniferous fluid collections (urinomas) are usually found in the retroperitoneum, most commonly within the perirenal spaces, sometimes with extension into the interfascial planes. They are most often due to urinary obstruction but may be due to abdominopelvic trauma, surgery, or diagnostic instrumentation.10,262,425,849 As urine extravasates into the retroperitoneum, it can cause lipolysis of the surrounding fat, with resultant encapsulation of urine, forming a urinoma.10,262,552 Small urinomas are usually treated conservatively because they tend to resorb spontaneously. Percutaneous drainage is often used for larger, persistent, or symptomatic urinomas, and treatment of the underlying cause of urine leakage is often necessary as well.455,849 On CT and MRI a urinoma shows water attenuation, low SI on T1, and very high SI on T2, similar to simple fluid elsewhere in the body. Urine leakage is usually directly demonstrated on contrast-enhanced CT or T1-weighted images, seen as enhancement of the fluid during the excretory phase of enhancement as a result of contrast extravasation from the lumen of the GU system (Fig. 55-33). Occasionally mural calcification may occur.62,238,296,425,521,849,938

Inflammatory Noninfectious and Infectious Fluid Collections Most inflammatory fluid collections originate in the anterior pararenal space of the retroperitoneum from extraperitoneal portions of the GI tract, including the pancreas, ascending and descending colon, duodenum, and retroperitoneal appendix. Inflammatory fluid collections are usually most prominent near their site of origin. Acute pancreatitis is one of the most common causes of an anterior pararenal fluid collection, and peripancreatic fluid often extends into the lesser peritoneal sac and subperitoneal spaces of the small bowel mesentery and  transverse mesocolon due to proteolytic enzymes.10,262,425 Sometimes

A

inflammatory fluid collections (in particular those related to acute pancreatitis) can access the posterior interfascial retrorenal plane of the retroperitoneum and extend into the transversalis fascia through a cleft between the medial posterior pararenal space and the lateral margin of the quadratus lumborum fat pad, also known as the lumbar triangle. Either by this lumbar triangle pathway or via a pathway from the posterior interfascial retrorenal plane to the subfascial plane at a level above or below the lumbar triangles, low flank discoloration (Grey Turner’s sign) associated with acute pancreatitis may be observed clinically, and other retroperitoneal fluid collections including abscesses may also follow this pathway to reach the abdominal wall.262,349-351 Inflammatory fluid may also traverse the midline via the anterior interfascial retromesenteric plane, spread inferiorly to the pelvic retroperitoneum via the combined interfascial plane, or extend superiorly along the diaphragm to enter the mediastinum.262 Inflammatory fluid collections confined to the posterior pararenal space of the retroperitoneum are rare, since no organs are present within this space, and most inflammatory fluid collections involving the posterior pararenal space tend to be secondary to severe infection in another space, such as osteomyelitis of the lumbosacral spine, ribs, or pelvis.10,198 Perirenal inflammatory fluid collections are most often secondary to renal infections.104 Infectious fluid collections in the retroperitoneum are typically subacute and are most often due to gram-negative bacilli. Clinical presentation is often nonspecific, with patients complaining of nausea, vomiting, weight loss, fever, chills, night sweats, anorexia, flank pain, or hip pain.10,104,134,198 As a result, the diagnosis of retroperitoneal infection may be overlooked in 25% to 50% of patients, with a subsequent increase in morbidity and mortality.10 Percutaneous abscess drainage is successful in the treatment of retroperitoneal abscesses, along with concomitant antimicrobial therapy. Multiple septations or fistulous communications decrease the success rate of percutaneous drainage.10,83,134,728 On CT and MRI, abscesses appear as localized complex fluid collections that show water or increased attenuation fluid, variable T1-weighted SI, intermediate to high T2-weighted SI, and thick

B FIG 55-33  Retroperitoneal urinoma and hemorrhage in 32-year-old woman status post fall from ladder 3 days earlier. A, Axial contrast-enhanced CT image through abdomen during nephrographic phase of enhancement shows hypoattenuating lacerations (long arrows) through lower pole of left kidney in association with mixed water attenuation and higher-attenuation fluid within left perirenal space, anterior interfascial retromesenteric plane, and posterior interfascial retrorenal plane. Note anterior displacement of left kidney. B, Axial contrast-enhanced CT MIP image during excretory phase of enhancement demonstrates leakage of contrast material (arrows) from medial left renal collecting system into perirenal space in keeping with perirenal urinoma.

CHAPTER 55  Retroperitoneum

1905

M

A

B

R R P S

*

*

S

C FIG 55-34  Retroperitoneal abscesses in 62-year-old man with acute pancreatitis. A to C, Sequential axial contrast-enhanced CT images through abdomen show loculated fluid and very-low-attenuation gas bubbles within fascial planes of retroperitoneum, with associated fascial thickening and enhancement. Note involvement of anterior interfascial retromesenteric planes (arrows), posterior interfascial retrorenal planes (long arrows), and root of small bowel mesentery (M) with bilateral extension via communicating passageways (*) to subfascial planes (S).

peripheral rim enhancement (Figs. 55-34 and 55-35).101,198,621,772 Layering debris may be present with increased attenuation and low SI on T1- and T2-weighted images, and gas in the form of bubbles or a gasfluid level may be present in half of cases, with very low attenuation and SI on T1- and T2-weighted images. The presence of gas increases the specificity for the diagnosis of an abscess.101,198,239,621 Most abscesses are round or oval in shape, but those adjacent to solid organs may have a lenticular or crescentic configuration.296 Associated displacement of adjacent structures with thickening, infiltration, or obliteration of adjacent fat planes is also frequently seen.62,296

Hydatid Cyst Hydatid cysts occur in patients of all ages and are most common in the liver, followed by the lungs. Retroperitoneal hydatid cysts are very rare.21,38,253,683,734,779 They are caused by the larval stage of the adult tapeworm Echinococcus granulosus and are endemic in Mediterranean countries, the Middle East, Asia, New Zealand, Australia, Africa, and South America. Dogs or other carnivores are the definitive hosts, whereas sheep or other ruminants are intermediate hosts. Humans

become secondarily infected by ingestion of food or water that has been contaminated by fecal material containing the parasitic eggs. Patients with retroperitoneal hydatid cysts most often present with flank pain and less commonly a palpable mass, and they often have positive test results upon serologic testing.21,38,656,670 Histologically, hydatid cysts have three layers: a rigid outer layer or pericyst composed of reactive host inflammatory and fibrotic cells, a middle acellular laminated membrane, and an inner germinative (or germinal) layer. Scolices, the infectious embryonic tapeworms, develop from outpouchings of the germinative layer known as the brood capsules (daughter vesicles). Freed scolices from ruptured brood capsules together with the brood capsules form hydatid sand, which settles in the dependent portion of the mother cyst.656,670 Treatment with total pericystectomy is typically performed, although partial removal may be performed following injection of scolocidal agent if located adjacent to vital structures.21,38 On CT and MRI, thick-walled cystic lesions with internal round peripheral daughter cysts and soft-tissue-attenuation, low-SI enhancing walls are characteristically seen. Typically the attenuation of the

1906

PART II  CT and MR Imaging of the Whole Body

K A

A A

*

A

B

FIG 55-35  Retroperitoneal abscesses in 46-year-old woman with immunosuppression. A and B, Sequential axial T2-weighted imaging through abdomen and pelvis demonstrate loculated fluid collections (A) within left posterior pararenal space and left subfascial plane, with extension medially into left psoas muscle, posteriorly into left paraspinal musculature (*), and inferiorly into left hemipelvis. Note anterior displacement of left kidney (K) and medial displacement of left external iliac vessels (arrows).

fluid within the mother cyst is greater than that of the daughter cysts, and the SI of the fluid within the mother cyst is greater than or similar to that of the daughter cysts on T1- and T2-weighted imaging. Detached germinative membranes, seen as linear, serpentine, or undulating filling defects within the mother cyst, as well as mural calcification may also be seen. Sometimes, unilocular cystic lesions may be encountered early on.38,670 On follow-up cross-sectional imaging, active hydatid cysts may demonstrate migration or herniation due to pressure differences between anatomic cavities or due to gravitational effects.417

Pseudomyxoma Retroperitonei The term pseudomyxoma peritonei, or “false mucinous tumor of the peritoneum,” has been used historically as a pathologic diagnostic term to describe any benign or malignant condition that results in intraperitoneal mucin accumulation.713,806 However, pseudomyxoma peritonei should be used strictly as a clinical descriptor for patients with mucinous ascites originating from a primary appendiceal mucinous adenoma with pathologic features of disseminated peritoneal adenomucinosis (DPAM).204,713,806 Pseudomyxoma peritonei (corresponding to DPAM pathologically) is a rare condition that occurs twice as often in men as in women, with a peak presentation in the sixth decade of life.204,713,762 Pseudomyxoma retroperitonei (or pseudomyxoma extraperitonei) occurs very rarely and is caused by retroperitoneal rupture of a primary appendiceal mucinous adenoma in a retrocecal appendix and fixation of the lesion to the posterior abdominal wall.234,576,713,760,789 It most often occurs in men in the fourth through ninth decades of life, with a peak occurrence in the sixth decade.789 Histologically the lesions are lined with glandular epithelium and filled with thick gelatinous mucinous material.150 On CT and MRI, multicystic lesions with low attenuation, low to intermediate SI on T1-weighted images, and high SI on T2-weighted images, as well as thickened walls or septations that displace and distort adjacent structures, are seen.150,534,575,760 Curvilinear or punctate mural calcifications may also be present.534 The primary appendiceal mucinous adenoma may be indistinguishable from adjacent mucinous fluid.309,806

Treatment of pseudomyxoma retroperitonei includes appendectomy and surgical débridement of retroperitoneal gelatinous deposits.789 It is important that patients undergo evaluation of the entire colon, since there is a strong association of appendiceal adenoma with synchronous or metachronous colorectal adenoma and carcinoma.157 The 5- and 10-year survival rates of patients treated for pseudomyxoma peritonei are up to 90% and 68%, respectively, with a mean survival of 112 months. The prognosis of patients with pseudomyxoma retroperitonei is probably better than pseudomyxoma peritonei, since vital abdominal structures are not involved in the adhesive processes.576,806

PSOAS MUSCLE (RETROFASCIAL) FLUID COLLECTIONS Fluid collections related to the psoas muscle are located in the retrofascial space rather than in the retroperitoneal space, because the psoas muscles (and quadratus lumborum muscles) are located posterior to the transversalis fascia, which is the posterior boundary of the retroperitoneum.777 Spread of infection from GI disease (e.g., appendicitis, diverticulitis, Crohn’s disease, or perforated colon carcinoma) is the most common source of a secondary psoas muscle fluid collection, and renal disease (e.g., due to perirenal abscess) is the second most common source. Extension from osteomyelitis of the lumbar spine is also common. Primary psoas muscle abscesses are rare in the Western world but may have an incidence as high as 20% in other parts of the world, such as in developing countries. They are more likely to occur in immunocompromised patients, with up to 90% due to Staphylococcus aureus infection. TB should be considered in the differential diagnosis of a psoas muscle fluid collection, because TB is an important cause of retroperitoneal abscess on a worldwide scale.597,643 Up to 5% of patients with spinal TB (Pott’s disease) develop a tuberculous psoas abscess, although large tuberculous psoas abscesses may occur without concomitant signs of bone involvement.198,492,643 Rarely, tuberculous psoas abscesses may be secondary to renal or colonic TB or due to the extension of retroperitoneal tuberculous lymphadenopathy to the psoas muscle.214

CHAPTER 55  Retroperitoneum Management of psoas muscle abscesses may include percutaneous drainage, surgical drainage, and/or antibiotic or antituberculous therapy. Successful treatment depends on differentiating primary from secondary psoas collections; those associated with renal or GI disease may require additional early surgery to correct the underlying disease process. However, percutaneous abscess drainage prior to definitive surgery may be useful to reduce morbidity.643,728 On CT and MRI an abscess of the psoas muscle has findings similar to an abscess in the retroperitoneum, as described previously. Secondary findings of psoas muscle enlargement, low attenuation and high  SI on T2-weighted images (due to muscle edema, bone destruction, and infiltration), and loss of surrounding fat planes may be present.208,403,425,643,857 Foci of gas, often considered a specific sign for an abscess, are not always present, may also be found in necrotic neoplasms, and are better demonstrated on CT than on MRI.425,597,643 Bacterial abscesses are usually larger in volume (mean volume, ≈500 mL) compared to tuberculous abscesses (mean volume, ≈80 mL) or compared to hematomas (mean volume, ≈50 mL).643 Psoas muscle hematomas can also occur and may be spontaneous or secondary to trauma, ruptured aneurysm, bleeding diathesis, anticoagulation, inflammatory disease, neoplastic disease, or recent surgery or biopsy.425,553 They often have high attenuation on CT and a peripheral rim of high SI on T1-weighted images secondary to methemoglobin (as part of the concentric rim sign described earlier). They may be associated with a hematocrit effect and are associated with psoas muscle enlargement.276,425,643,773,857 This finding coupled with a history of trauma or anticoagulation therapy is often diagnostic, and such a patient with this combination can be followed clinically. Malignancies with cystic components that involve the psoas muscle are usually large and aggressive and have solid enhancing tissue, so the distinction between abscess, hematoma, and tumor is not difficult on MRI.

PNEUMORETROPERITONEUM Retroperitoneal gas is most often the result of bowel perforation secondary to trauma, iatrogenic causes such as surgery or endoscopy, inflammatory bowel disease, peptic ulcer disease, foreign body, or gas-producing

infection. It can originate from the duodenum or ascending, descend­ ing, or rectosigmoid portions of the colon.18,76,154,194,316,328,551,617,664,681,709 Pneumoretroperitoneum can also be related to bowel infarction,  pneumatosis cystoids intestinalis, superinfected necrotizing pancre­ atitis, necrotizing fasciitis, abscess formation, percutaneous biopsy, epidural anesthesia, extracorporeal shockwave lithotripsy, acu­ puncture, hydrogen peroxide wound irrigation, as well as from  inferior extension of gas in the thorax, such as with mechanical ventilation.14,18,264,306,333,356,425,475,510,553,596,645,712,819 On CT and MRI, retroperitoneal gas has very low attenuation and very low SI on T1- and T2-weighted images and may either be punctate or confluent, sometimes with air-fluid levels (Fig. 55-36).18,328,664 Frequently, associated direct or indirect findings of GI or GU organ disease involvement are seen, which can vary in appearance depending on the pathologic process that is present.

RETROPERITONEAL AND PELVIC LYMPHADENOPATHY Overview There are approximately 400 to 500 lymph nodes in the human body, with approximately half located in the abdomen and pelvis. Retroperitoneal and pelvic lymph nodes are generally present in a perivascular distribution about the aorta, IVC, and iliac vessels.191 Both normal and malignant lymph nodes appear as small oval or flat structures that have soft tissue attenuation, intermediate SI on T1, intermediate to high SI on T2, high SI on DWI relative to skeletal muscle, and enhance. Fatty hila with attenuation and SI similar to that of subcutaneous fat may also be present. Conventional MRI has been shown equally effective or slightly superior to CT in detection of abdominopelvic lymph nodes and can more easily distinguish lymph nodes from vessels.198,461 Attenuation and SI characteristics of lymph nodes on conventional imaging sequences are not accurate in differentiating between benign and malignant lymph nodes.171,179,198,508 Calcifications may sometimes be present, appearing as foci of very high attenuation and very low  SI on T1- and T2-weighted images when visualized, and may be  secondary to neoplastic or inflammatory conditions.191 Hemorrhagic

K

A

1907

B FIG 55-36  Pneumoretroperitoneum in 27-year-old woman status post endoscopic retrograde cholangiopancreatography (ERCP) 3 days earlier. A and B, Sequential axial contrast-enhanced CT images in lung windows through abdomen show extensive very-low-attenuation gas throughout right retroperitoneal spaces. Note mild anterior displacement of right kidney (K).

1908

PART II  CT and MR Imaging of the Whole Body

lymphadenopathy or lymphadenopathy due to metastatic melanoma may demonstrate increased SI on T1.198 Fatty lymphadenopathy may be seen with lipoplastic lymphadenopathy, cavitary lymph node syndrome, and Whipple’s disease.222,537,667 Lipoplastic lymphadenopathy has a benign course and typically occurs in obese middle-aged women. The lymph nodes are enlarged and replaced by fat, with attenuation and SI similar to macroscopic fat on all pulse sequences, sometimes with a thin rim of nodal tissue remaining.222,248,513 In the pelvis, this can mimic the appearance of pelvic lipomatosis but can be differentiated at lymphangiography.222 Cavitary lymph node syndrome is sometimes found in association with celiac disease, along with splenic atrophy, and appears as low-attenuation lymphadenopathy that sometimes contains fat-fluid levels.220,537,703,724 Necrosis of lymph nodes suggests malignant lymphadenopathy, such as from metastatic squamous cell carcinoma or less commonly lymphoma or testicular carcinoma, but may also occur in benign inflammatory diseases including mycobacterial infections (Fig. 55-37), bacterial infection, Whipple’s disease, histoplasmosis infection, and systemic lupus erythematosus. On CT and MRI, necrotic lymph nodes have central low attenuation, central high SI on T2, and lack of central enhancement.191,274,282,360,399,409,573,688-690,885,940 Short-axis size estimation is the imaging criterion used most often to differentiate benign from malignant lymph nodes.198 In patients with a primary cancer, lymph nodes smaller than 10 mm in short-axis dimension are considered benign, whereas nodes 10 mm or larger are

A

suspicious for malignancy.191,508,901 However, overlap exists; malignant lymph nodes can sometimes be smaller than 10 mm, and benign reactive or hyperplastic lymph nodes can be 10 mm or larger.164,171 Lymph node location may also affect the criteria used to assess nodal size. For example, normal retrocrural and porta hepatis nodes should not exceed 6 mm, and gastrohepatic ligament nodes should not exceed 8 mm, whereas normal pelvic lymph nodes (particularly inguinal lymph nodes) may normally be up to 10 to 15 mm in size.45,180,191,512,663,901 The presence of multiple slightly smaller lymph nodes in the 8 to 10 mm range should be viewed with suspicion for underlying pathology (e.g., chronic lymphocytic leukemia).133,191 Most normal lymph nodes have an oval shape, whereas malignant lymph nodes are often round.508 Malignant neoplasms tend to spread initially to their regional nodal groups, although it is not unusual for lymphadenopathy to involve several contiguous or even widely separated nodal chains because there are complex intercommunications between regional groups of lymph nodes. Virtually any abdominal or pelvic neoplasm may lead to retroperitoneal lymphadenopathy, but the most common are lymphoma, renal cell carcinoma, testicular carcinoma, cervical carcinoma, and prostatic carcinoma. Carcinomas of the bladder, prostate, cervix, and uterus initially spread to the pelvic nodes, as may anorectal carcinoma after first involving nodes in the perirectal space. Testicular, ovarian, and fallopian tube malignancies spread first to the retroperitoneal nodes adjacent to or near the renal hila but may involve the pelvic

B

C FIG 55-37  Retroperitoneal and mesenteric lymphadenopathy and small bowel disease due to atypical mycobacterial infection in 43-year-old man with HIV infection. A and B, Axial contrast-enhanced CT images through abdomen and upper pelvis demonstrate extensive lymphadenopathy in retroperitoneum and mesentery, with central low attenuation seen in one mesenteric lymph node (arrow) in keeping with necrosis. Note associated infiltration of mesenteric fat. C, Axial contrast-enhanced CT image through pelvis shows associated mild fold thickening and wall thickening of multiple loops of small bowel.

CHAPTER 55  Retroperitoneum nodes via retrograde spread.191 Testicular carcinoma of the right testis tends to involve aortocaval and paracaval lymph nodes and can cross over to the left side of the aorta, whereas testicular carcinoma of  the left testis tends to involve left paraaortic lymph nodes.585,931 Anterior displacement of the aorta or IVC or lateral ureteral displacement may be seen when malignant retroperitoneal lymphadenopathy is present.16,428 Ultrasmall superparamagnetic iron oxide particles such as ferumoxtran-10 (not yet U.S. Food and Drug Administration [FDA] approved) may be used as a nontargeted MRI-negative lymph node contrast agent to differentiate between benign and malignant lymph nodes. Benign lymph nodal tissue accumulates iron particles within macrophages and has very low SI on T2*-weighted imaging, whereas foci of malignancy within lymph nodes remain intermediate to high SI on T2*-weighted imaging. Use of this approach increases the sensitivity and specificity for detecting nodal metastatic disease, compared to CT or conventional MRI, because micrometastases within normalsized lymph nodes can be indirectly detected. However, false-positive results due to focal nodal lipomatosis or a prominent fatty hilum can occur, as can false-positive results due to reactive lymphoid follicular hyperplasia or nodal necrosis. False-negative results can be due to limitations in spatial resolution.61,160,198,281,283,284,732,852 FDG PET and FDG PET/CT imaging can similarly be used to differentiate benign from malignant lymph nodes; sensitivity and specificity of detection of nodal metastatic disease is better than CT or conventional MRI. Furthermore, PET and PET/CT can simultaneously detect and quantify nonnodal malignancy and allow for quantitative monitoring of the response of nodal and nonnodal metastatic disease in the entire body to therapeutic intervention. False-positive results can be due to metabolically active inflammation in lymph nodes but can be decreased through use of dual time point PET imaging. Falsenegative results can be due to limitations in spatial resolution or malignancies with low metabolic activity. In general, PET/CT has improved diagnostic performance compared to CT or PET imaging alone, because the metabolic and anatomic data obtained are complementary to each other.23,446-449,786,852,897 PET/MRI may also be useful for this purpose.855 Lymphoma and metastatic disease (Figs. 55-38 to 55-40) are by far the most common causes of abdominopelvic lymphadenopathy. Other

less common causes of abdominopelvic lymphadenopathy include leukemia, Kaposi’s sarcoma, TB, atypical mycobacterial infection, bacterial, viral, or fungal infection, AIDS, mesenteric lymphadenitis, idiopathic RPF, sarcoidosis, cat-scratch disease, Whipple’s disease, amyloidosis, Castleman’s disease, mastocytosis, celiac sprue, and Crohn’s disease, among numerous other infectious or inflammatory etiologies.80,133,171 Several of these etiologies will be discussed.

Lymphoma Lymphoma is the most common retroperitoneal malignancy and accounts for one third of retroperitoneal malignant neoplasms.50 Patients of all ages may be affected, although the incidence of nonHodgkin’s lymphoma increases with age, whereas Hodgkin’s lymphoma has a bimodal age distribution in young adults and the elderly. The clinical course may range from indolent to highly aggressive, depending partly on the histology of the tumor.782,916 Lymphoma typically begins as local lymph node enlargement and then spreads via lymphatics to adjacent lymph nodes, commonly in the retroperitoneum, and sometimes becomes systemic.198,296 Conglomerate mesenteric and retroperitoneal masses may form, and they characteristically infiltrate the perinephric spaces of the retroperitoneum.198 Whereas abdominal Hodgkin’s lymphoma tends to be confined to the spleen and retroperitoneum, with spread of disease to contiguous lymph nodes, non-Hodgkin’s lymphoma more commonly involves a variety of nodal groups and extranodal sites.71,608,782,916 Approximately 50% of patients with non-Hodgkin’s lymphoma have mesenteric lymphadenopathy at presentation, which may become confluent with characteristic encasement of the mesenteric vessels. In contrast, enlarged lymph nodes due to other conditions tend to remain discrete, rarely forming a conglomerate mass.296 Abdominal pain or a palpable abdominal mass may be present, along with constitutional symptoms and signs such as fever, night sweats, weight loss, itching, and abnormal laboratory findings such as leukocytosis. The Ann Arbor Staging System for lymphoma (Box 55-3) is used to help predict prognosis and survival. This staging classification is less useful for non-Hodgkin’s lymphoma than for Hodgkin’s lymphoma, because non-Hodgkin’s lymphoma more frequently disseminates hematogenously. The prognosis worsens as the stage increases, and the prognosis within each stage also worsens when B

J A

M

Ds

PM

A

1909

B FIG 55-38  Retroperitoneal metastasis in 59-year-old woman with colon carcinoma. A and B, Sequential axial contrast-enhanced CT images through abdomen demonstrate left hydroureteronephrosis and delayed nephrogram of left kidney due to infiltrative soft tissue mass (M) within left retroperitoneum. Note that mass infiltrates adjacent left psoas muscle (PM), adjacent jejunum (J), descending colon (Ds), and wall of abdominal aorta (A).

1910

PART II  CT and MR Imaging of the Whole Body

M

M

B

A

M

C FIG 55-39  Retroperitoneal lymphadenopathy and hepatic metastasis in 55-year-old woman with metastatic adrenocortical carcinoma. A to C, Axial T1-weighted (A), fat-suppressed T2-weighted (B), and fat-suppressed contrast-enhanced T1-weighted (C) images through abdomen demonstrate mass (M) in left paraaortic region of retroperitoneum that displaces abdominal aorta anteriorly. Mass has intermediate T1 SI and heterogeneous slightly high T2 SI relative to skeletal muscle, as well as continuous rim enhancement in keeping with metastatic lymphadenopathy. Note coexistent hepatic lesion (arrow) with similar SI and enhancement characteristics in keeping with hepatic metastasis.

*

*

*

*

A

B FIG 55-40  Extraperitoneal pelvic lymphadenopathy in 44-year-old man with metastatic testicular carcinoma and palpable abdominopelvic mass. A and B, Axial T2-weighted (A) and fat-suppressed contrast-enhanced T1-weighted (B) images through pelvis show complex mass in right external iliac region that medially displaces right external iliac vessels. Note intermediate T2 SI enhancing components of mass corresponding to solid tumor components, as well as very high T2 SI nonenhancing components (*) corresponding to cystic or necrotic components. Also note associated medial displacement of right external iliac vessels (arrows) by mass.

CHAPTER 55  Retroperitoneum symptoms are present.106,190,574,718,782,916 The Revised European-American Classification of Lymphoid Neoplasms (REAL) and the World Health Organization (WHO) classification provide systems for classification of lymphomas. Studies have shown that these classification systems separate patients with clinically different types of lymphoma and

BOX 55-3  Ann Arbor Staging System for

Hodgkin and Non-Hodgkin’s Lymphoma106,190 Stage System Grouping Stage I Involvement of a single lymphatic site (i.e., nodal region, Waldeyer’s ring, thymus, or spleen) (I); or localized involvement of a single extralymphatic organ or site in the absence of any lymph node involvement (IE) Stage II Involvement of ≥ 2 lymph node regions on the same side of the diaphragm (II); or localized involvement of a single extralymphatic organ or site in association with regional lymph node involvement with or without involvement of other lymph node regions on the same side of the diaphragm (IIE) Stage III Involvement of lymph nodes on both sides of the diaphragm (III), which also may be accompanied by extralymphatic extension in association with adjacent lymph node involvement (IIIE), by involvement of the spleen (IIIS), or both (IIIE,S) Stage IV Diffuse or disseminated involvement of ≥ 1 extralymphatic organ, with or without associated lymph node involvement; or isolated extralymphatic organ involvement in the absence of adjacent regional lymph node involvement, but in conjunction with disease in distant site(s). This stage includes any involvement of liver or bone marrow, lungs (other than by direct extension from another site), or cerebrospinal fluid. The presence or absence of systemic symptoms should be noted with each stage designation: A, asymptomatic; B, presence of fever (>38°C), night sweats, or weight loss >10% of body weight in the 6 months prior to diagnosis.

stratify patients into different prognostic groups.1,110,285,286,505,820 The management approach to patients with lymphoma is dependent upon multiple factors and may either be conservative or involve chemotherapy, radiation therapy, immunotherapy, radioimmunotherapy, bone marrow transplantation, or surgical resection.782,916 On CT, nodal and extranodal lymphoma typically has homogeneous soft tissue attenuation even when large, because necrosis and calcification prior to therapy are uncommon.24,212,610,955 Confluent lymphadenopathy on both sides of the superior mesenteric vessels gives rise to an appearance described as the sandwich sign.583 On MRI, lymphoma is typically intermediate to slightly high in SI on T1, high in SI on T2 relative to muscle, high in SI on DWI, and low in SI on ADC parametric map images (Fig. 55-41).267,605,933 Associated involvement of the bone marrow, GI tract, and other parenchymal organs such as the liver or spleen may be present.212 Posttreatment assessment of the SI of lymphomatous nodal masses on T2-weighted images may be useful because low SI on T2 within treated lymph node masses represents nonviable tumor or fibrosis. However, most of the published literature on using T2-weighted SI to document response to treatment has focused on patients with mediastinal Hodgkin’s lymphoma. In addition, one should be cautious in diagnosing residual tumor when there is high SI on T2, because high SI due to edema, granulation tissue, hemorrhage, or immature fibrosis may persist up to 12 months following therapy.693 MRI and CT show equivalent accuracies for staging of abdominal lymphoma.372 However, MRI outperforms CT in detection and characterization of bone marrow involvement by lymphoma.312 DWI may also be useful to assess early therapeutic response of lymphoma to chemotherapy, where an increase in tumor SI on ADC parametric map images corresponds to a tumor response.933 FDG PET/CT is frequently used to stage nodal and extranodal lymphoma in the whole body, owing to its high sensitivity and  specificity, and is also used to quantitatively monitor the effects  of therapeutic intervention, prognosticate patient outcome, and restage disease.52,90,119,364,441,836,852,933,967 Interestingly, the utility of FDG PET for lymphoma may vary depending on the specific histologic subtype under study.195,380 Elstrom et al. showed that in 172 patients with lymphoma, FDG PET detected disease in 100% with large

P

I

A

1911

A

B FIG 55-41  Retroperitoneal lymphadenopathy in 56-year-old woman with non-Hodgkin’s lymphoma. A, Axial fat-suppressed T2-weighted image through abdomen demonstrates extensive retroperitoneal lymphadenopathy that anteriorly displaces pancreas (P), anteriorly displaces the IVC (I) from lumbar spine, and surrounds abdominal aorta (A). Lymphadenopathy has homogeneous slightly high T2 SI relative to skeletal muscle. Note incidental gallstone (arrow) with very low T2 SI. B, Axial fused FDG PET/CT image again shows extensive retroperitoneal soft tissue attenuation lymphadenopathy with multiple regions of avid FDG uptake (confluent white regions). Note normal urinary excretion of FDG within kidneys (long arrows).

1912

PART II  CT and MR Imaging of the Whole Body

B-cell lymphoma and mantle cell lymphoma and 98% with Hodgkin’s lymphoma and follicular lymphoma, whereas only 67% with marginal zone lymphoma and 40% with peripheral T-cell lymphoma were detected.195

Tuberculosis TB is endemic in most developing countries but is also globally increasing in incidence in both immunocompetent and immunocompromised patients, in part related to AIDS and multidrug-resistant TB.889 Although the thorax is the most frequently involved site, any of a number of organ systems may be involved, including the respiratory, cardiac, central nervous, musculoskeletal, GI, and GU systems. The abdomen is the most common site of extrapulmonary TB.97 Lymphadenopathy is the most common manifestation of abdominal TB and is the only abdominal finding in up to 55% of patients.282,399 Mesenteric, omental, peripancreatic, periportal, pericaval, and upper paraaortic lymph nodes are commonly involved because of lymphatic drainage from the small bowel and right colon, in addition to disease

spread along the hepatoduodenal ligament. More extensive lymphadenopathy may be present in patients with AIDS.399,942 TB may reach the abdominal lymph nodes either through infection of the GI tract or through hematogenous spread of disease.889 Pathologically, active tuberculous lymphadenopathy passes through  multiple stages: (1) lymphoid hyperplasia with formation of tubercles and granulomata without caseation necrosis, (2) caseation necrosis, (3) lymph node capsular destruction, adherence of multiple lymph nodes, and perilymphadenitis, (4) rupture into surrounding soft tissue forming a confluent abscess cavity, and (5) healing with subsequent fibrosis and calcification.573,889 On CT and MRI the lymph nodes are usually multiple, large (mean size, 2-3 cm), and of low attenuation, low to intermediate SI on T1, and central high SI on T2 relative to skeletal muscle in active disease, without associated urinary, biliary, or GI tract obstruction (Fig. 55-42).158,199,201,282,331,399,573,667,757 The degree of low attenuation and high SI on T2 within lymph nodes generally parallels the amount of central necrosis, and the degree of nodal necrosis

*

A

B

*

C

*

* FIG 55-42  Retroperitoneal lymphadenopathy due to Mycobacterium tuberculosis in 25-year-old woman with AIDS, weight loss, and abdominal pain. A and B, Sequential axial contrast-enhanced CT images demonstrate multiple enlarged retroperitoneal lymph nodes, many of which have central low attenuation (*) due to necrosis. Note diffuse infiltration of retroperitoneal fat due to direct spread of tuberculosis. C, Coronal heavily T2-weighted image again shows extensive retroperitoneal lymphadenopathy, some of which has central increased T2 SI (*) due to necrosis. Note very high T2 SI fluid within hepatorenal fossa of peritoneal cavity due to peritoneal spread of tuberculosis.

CHAPTER 55  Retroperitoneum correlates with clinical signs and symptoms. Low T2-weighted SI nodes may be present owing to inactive or fibrotic tissue in later stages of disease or owing to calcification.158,399,573 Homogeneous or mildly inhomogeneous enhancement of lymph nodes is seen if there is no or minimal central necrosis.158,573,673 However, peripheral intermediate SI on T1-weighted images and high SI on T2-weighted images with delayed peripheral enhancement is often present and is due to granulation tissue, inflammation, hypervascularity, and edema, along with central lack of enhancement due to caseation or liquefaction necrosis in up to 40% of cases.158,331,399,573,673,757,960 Nonenhancement of lymph nodes may be seen in AIDS, presumably because of a diminished inflammatory reaction association with this condition.889 Although the coexistence of findings of pulmonary TB may be seen, only 40% of patients with abdominal TB have findings of associated pleuropulmonary disease on chest radiography.519,889

Castleman’s Disease Castleman’s disease, also known as angiofollicular lymph node hyperplasia or benign giant lymph node hyperplasia, is a hyperplasticdysplastic-neoplastic lymphoproliferative process of uncertain etiology that is a rare cause of abdominopelvic lymph node enlargement.124,198,754 Individuals with Castleman’s disease can present clinically with either unicentric or localized lymphadenopathy that remains relatively stable, or with multicentric or disseminated disease characterized by generalized lymphadenopathy, constitutional symptoms, organomegaly, and a more aggressive clinical course, with 30% of patients developing malignant transformation.124,397,541,754,838 Patients with unicentric disease present at a significantly younger age (peak presentation in the third decade of life) than those with multicentric disease, who present in the sixth decade of life.124,754 Unicentric disease occurs equally in men and women, whereas multicentric disease occurs twice as commonly in men as in women.541 Patients with unicentric disease usually have intrathoracic disease most commonly involving mediastinal or hilar lymph nodes, whereas in patients with multicentric disease the mesentery, retroperitoneum, and pelvis are involved in 7% to 12% of patients.124,397,838 Multicentric Castleman’s disease is frequently associated with POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, M proteins, and skin changes), Kaposi’s sarcoma, and AIDS, and less commonly with amyloidosis and monoclonal gammopathy.397,838 The three basic histopathologic variants of Castleman’s disease include hyaline vascular, plasma cell, and mixed subtypes.124 The more common hyaline vascular subtype is characterized by small hyaline vascular follicles and interfollicular capillary proliferation, whereas the plasma cell subtype is characterized by large follicles with intervening sheets of plasma cells.397 Grossly, homogeneous, encapsulated, highly vascular masses are present, which are often adherent to surrounding structures.838 In patients with unicentric disease, the hyaline vascular subtype is most commonly found, whereas the majority of patients with multicentric disease (particularly if mesenteric and retroperitoneal lymph nodes are involved) are associated with the plasma cell or mixed subtypes.124,166,541 Most patients with the hyaline vascular subtype are asymptomatic, with fewer than 10% having associated systemic symptoms, whereas systemic symptoms are commonly associated with the plasma cell subtype.541,754 On CT and MRI the lesions of Castleman’s disease generally have soft tissue attenuation similar to muscle attenuation, low to intermediate SI on T1 and high SI on T2 relative to the skeletal muscle,  and mild to marked enhancement, particularly in the arterial phase, with slow washout.369,371,397,667,838,964 Lesions smaller than 5 cm tend to have homogeneous attenuation and enhancement, whereas larger lesions tend to have more heterogeneous attenuation and

1913

enhancement.545 A peripheral rim of even greater enhancement may sometimes be visualized as well, related to a greater quantity of small or capillary vessels in the periphery, and dilated feeding arteries due to lesional hypervascularity may be seen.348,963 Central irregular or stellate hypovascular areas of low attenuation and low SI on both T1and T2-weighted images, with a radial pattern, may be seen in the hyaline vascular subtype, suggesting fibrosis.397,838,964 Calcification may be present in one third of cases of abdominal Castleman’s disease, is better revealed on CT , and may be lumpy, flocculent, or irregular with an arborizing pattern that radiates outward from a central  location.397,545,811,838 Dilated feeding vessels and draining veins may also be visualized.811 In general, enhancement of lesions of the plasma cell subtype tends to be more variable but less than lesions of the hyaline vascular subtype.337,397 Unicentric disease tends to show a welldefined, homogeneous, single large intraabdominal mass or a single dominant mass with satellite nodules (mean size, 6 cm; range, 3-10 cm), whereas multicentric disease tends to show variably sized well-defined lymph nodes in multiple locations, sometimes also with associated findings of hepatosplenomegaly, ascites, and retroperitoneal fascial thickening.397,964 Surgical resection is the treatment for unicentric Castleman’s disease, usually with resolution of systemic symptoms and laboratory abnormalities within a few months, with a cure rate approaching 100%.116,124,541 In some cases, preoperative embolization of feeding vessels may be performed to reduce intraoperative hemorrhage.390 Alternatively, radiotherapy may be used for patients who are not good surgical candidates or who have undergone prior incomplete surgical resection.124 Combination chemotherapy and corticosteroids are used to treat multicentric Castleman’s disease, although interferon may be used as an alternative therapy.20,124 Multicentric Castleman’s disease still has a poor prognosis, with a median survival of 24 to 33 months and death most commonly due to sepsis and lymphoma.541,754,838

Cat-Scratch Disease Cat-scratch disease is an infectious disorder that is most often caused by the gram-negative bacterium Bartonella henselae and frequently occurs in children and adolescents. It is the leading cause of unilateral lymphadenopathy and chronic lymphadenopathy in this patient population and most commonly occurs in the fall and winter months. The clinical diagnosis is generally made by demonstrating a history of exposure to an animal scratch (usually by a cat), a dermal or ocular inoculation site that typically forms a pustule or papule, single or regional lymphadenopathy, or a positive cat-scratch disease serologic test. Other diagnostic criteria include histopathologic findings characteristic of the disease at skin or lymph node biopsy, demonstration of the offending bacillus, and failure to demonstrate other causative organisms.200,453,924 Although tender lymphadenopathy proximal to the site of inoculation is usually the presenting complaint, constitutional symptoms including generalized aches, fever, malaise, headache, sore throat, nausea, vomiting, and anorexia may occur. Cervical and axillary lymph nodes are the most common sites of involvement, and single lymph node involvement is seen in up to 85% of patients. However, multiple sites of lymph node involvement as well as multiorgan involvement may also be present, particularly in immunocompromised patients.201,924 Histologically the hallmark lesion is granulomatous inflammation with a central stellate microabscess.453 On CT and MRI, nonspecific retroperitoneal or pelvic lymphadenopathy may occasionally be seen, usually with low attenuation, low T1- and T2-weighted SI, nonenhancing centrally necrotic areas, along with associated high T2-weighted SI inflammation and edema in the surrounding tissues. Hepatosplenomegaly with granulomatous

1914

PART II  CT and MR Imaging of the Whole Body

changes and abscess formation may also be seen, as well as osseous lesions that are usually lytic.178,315,320,456 Cat-scratch disease is most often self-limited, persisting on the order of weeks or months, and the prognosis is generally excellent with few complications. Uncommonly, lymphadenopathy may persist for up to 2 years. Excision of involved lymph nodes is rarely performed, and needle aspiration and antibiotic therapy is implemented only for severe cases of suppurative lymphadenitis. Antibiotic therapy is also implemented when multiorgan involvement is present.200,315,501,924

Whipple’s Disease Whipple’s disease is an uncommon chronic multisystem disorder caused by infection with Tropheryma whippelii and characterized by lipogranulomatous inflammation of the small bowel, with systemic extension to virtually any organ system.221 The peak age for presentation is in the fifth decade of life. Caucasians are most often affected, and men are affected eight times more commonly than women.527 It is difficult to diagnose because the clinical features may be variable, but most often presents with arthropathy, weight loss, diarrhea, abdominal pain, and less commonly with fever, lymphadenopathy, or hepatosplenomegaly.221,527 A suspected diagnosis of Whipple’s disease can be confirmed by showing periodic acid–Schiff (PAS)-positive granular foamy macrophages from a sampled tissue such as the small bowel or a peripheral lymph node.527 Treatment of Whipple’s disease consists of prolonged antibiotic therapy.221 On CT and MRI, findings may resemble a malignancy such as lymphoma. Hepatosplenomegaly, small bowel wall and fold thickening, or abdominopelvic lymphadenopathy may be present.221,481,667 Low-attenuation lymphadenopathy may be seen, which may even have fat attenuation and SI due to infiltration of lipid-laden macrophages.667 Necrosis of lymph nodes may cause lack of central enhancement of the nodes.940 Ascites and sacroiliitis may also be seen.527

Rosai-Dorfman disease is nonspecific on CT and MRI, often requiring tissue sampling for a definitive diagnosis to be made.450,468,470,562,599,787

OTHER SELECTED RETROPERITONEAL NEOPLASMS Hemangioma Hemangiomas are one of the most common soft tissue tumors, accounting for approximately 7% of all benign tumors.127,256 Most hemangiomas occur superficially, with a predilection for the head and neck region, but may also occur deep in parenchymal organs such as the liver and spleen, or very rarely within the retroperitoneum in men and women of all ages.127,256,279,240,824 Retroperitoneal hemangiomas are almost always of the cavernous type, tend to be asymptomatic, especially in the early stages of their development, and symptoms, when present, are nonspecific owing to mass effect upon adjacent anatomic structures.3,340 Treatment is conservative, with surgical removal reserved for symptomatic lesions or to establish a tissue diagnosis.105 On CT and MRI a hemangioma is generally well defined, with round or lobular margins, low attenuation, low SI on T1, and high SI on  both T2 and heavily T2-weighted imaging relative to skeletal muscle (Fig. 55-43).279,340,454,496,801 Occasional heterogeneity in attenuation and SI due to the presence of adipocytes, hemorrhage, fibrosis, or thrombosis may be seen. Calcified phleboliths when visualized have very low SI on both T1 and T2, are better visualized on CT, and are strongly suggestive

Kikuchi’s Disease and Rosai-Dorfman Disease Kikuchi’s disease (also known as Kikuchi-Fujimoto disease and histiocytic necrotizing lymphadenitis) is an uncommon idiopathic condition characterized by benign self-limited lymphadenopathy that is most frequently seen in the cervical region but uncommonly the retroperitoneum, most often in young Asian women.438,518,562,725,946,958 It usually presents clinically with acute tender cervical lymphadenopathy and low-grade fever.81 Histologically there is necrotizing lymphadenitis with irregular paracortical areas of coagulative necrosis. Abundant karyorrhectic debris along with large numbers of histiocytes are typically seen at the margins of the necrotic areas. Thrombosed vessels  are also usually seen around areas of necrosis. Therapy is typically conservative.81 Rosai-Dorfman disease (also known as sinus histiocytosis) is a rare benign idiopathic disorder that commonly affects lymph nodes as well as extranodal tissues, occurs in young adults (mean age at presentation, 21 years), and occurs slightly more commonly in men than in women. It typically presents with bilateral cervical lymphadenopathy that is frequently massive, fever, leukocytosis, an increased ESR, and hypergammaglobulinemia. Infrequently, retroperitoneal and pelvic nodal and extranodal tissues can be involved, sometimes presenting as massively sized lesions.218,717,787,932,966 At histology, typical findings include large histiocytes with abundant pale cytoplasm, round vesicular nuclei, and lymphocytophagocytosis or emperiopolesis (i.e., active penetration of one intact cell by another).218 Treatment is typically conservative because the disease is usually benign and self-limited, although surgical resection may be used for symptomatic or widespread disease.470 The imaging appearance of retroperitoneal lymphadenopathy in Kikuchi’s disease and Rosai-Dorfman disease and of extranodal

M

A

M

B

*

FIG 55-43  Retroperitoneal hemangioma in 35-year-old man with palpable abdominal mass. A, Axial unenhanced CT image through abdomen demonstrates right retroperitoneal soft tissue–attenuation mass (M) adjacent to duodenum. Note multiple coarse focal calcifications due to phleboliths. B, Axial fat-suppressed T2-weighted image shows high SI of mass (M) relative to skeletal muscle, with lower SI relative to cerebrospinal fluid (*), characteristic of hemangioma.

CHAPTER 55  Retroperitoneum of the diagnosis.127,149,454,549,714,824,864 Nonparenchymal hemangiomas in the abdomen usually demonstrate progressive enhancement.340,454

Solitary Fibrous Tumor The solitary fibrous tumor (SFT) is a rare neoplasm of mesenchymal origin that comprises less than 2% of all soft tissue tumors.255 Less than 30 cases of retroperitoneal SFT have been reported in the English literature.373 The majority occur in an intrathoracic location, but up to one-third may occur in extrathoracic locations including the head and neck, extremities, abdominal parenchymal organs, retroperitoneum, peritoneum, and pelvic organs.192,229,255,259,373,407,436,906,954 Extrathoracic SFT occurs in men and women of all ages equally, and although typically asymptomatic, may present with symptoms and signs related to extrinsic compression of adjacent organs or occasionally be due to hypoglycemia.255 SFT most likely arises from submesothelial fibroblasts and is similar to a hemangiopericytoma.192,255 At histology, a collagenous matrix with arrays of spindle cells is generally seen.229,255,289,436 Areas of necrosis, cystic or myxoid change, calcification, hemorrhage, increased vascularity, atypia, or malignancy may also be seen.880 The majority are benign, although up to 20% may be malignant.255,289,906 A malignant SFT can arise de novo or by dedifferentiation or malignant degeneration within an otherwise histologically benign SFT. The clinical behavior of SFT is unpredictable by histopathologic features alone, emphasizing the need for long-term surveillance. A benign SFT smaller than 10 cm typically has a favorable outcome with surgical resection alone. An SFT greater than 10 cm that also harbors malignant foci and has positive surgical margins tends to have a poor prognosis despite surgical treatment. Optimal adjuvant therapy for this latter group is unknown, and close-interval follow-up is advised because there is an increased incidence of local recurrence.255,289,880 On CT a well-circumscribed, smooth, lobulated soft tissue mass is seen that may contain scattered calcifications. Smaller tumors tend to enhance homogeneously, whereas central tubular or rounded lowattenuation areas due to cystic or necrotic change may be seen in larger lesions.407,716,835,906 On MRI, benign SFT usually has relatively homogeneous low to intermediate SI relative to skeletal muscle on both T1and T2-weighted images that is due to fibrous tissue, as well as intense enhancement. Additionally there may be areas of subacute hemorrhage that have high T1-weighted SI, as well as nonenhancing cystic or necrotic foci that are more heterogeneous and higher in T2-weighted SI relative to the remainder of the tumor.126,373,407,716,835,906 However, when a central focus of heterogeneity and variable contrast enhancement is identified within an SFT at CT or MRI, malignant degeneration should be considered.906

Chordoma Chordoma is a low-grade malignant tumor arising from the remnants of the fetal notochord that usually occurs at either end of the neural axis, most often in the sacrococcygeal region (50%), followed by the sphenooccipital region (35%-50%) and other spinal segment regions (15%).150,647,814,821 It almost always occurs in a sagittal or parasagittal location in relation to the spine, most commonly occurs in the fourth through seventh decades of life (2 : 1 male-to-female ratio), and is the most common primary malignant sacral tumor.172 Grossly, chordomas are lobulated and surrounded by a pseudocapsule. At histology, typical chordomas have cords of characteristic highly vacuolated physaliphorous cells located in a background of abundant mucin. Foci of cartilaginous tissue may be seen in chondroid chordomas. Sarcomatous elements may be encountered in dedifferentiated chordoma.591 On CT a well-defined lobulated soft tissue attenuation mass in  the presacral space is seen, typically with osseous destruction of the

1915

sacrum and/or coccyx, commonly with amorphous internal calcifications.172,591,821 Sometimes a higher-attenuation fibrous pseudocapsule may be seen.591 On MRI a heterogeneous low to intermediate T1-weighted SI and high T2-weighted SI mass relative to skeletal muscle is typically seen. High-SI foci on T1-weighted images, as well as low-SI septa on T1- and T2-weighted images, may also be present. In general, chondroid chordomas have higher T1- and lower T2-weighted SI compared to typical chordomas.814,821 Enhancement on CT and MRI may be variable but is often moderate.172,814,821 Involvement of adjacent pelvic or spinal musculature as well as adjacent osseous structures may also occur.591,814 Major differential diagnostic considerations for other aggressive lesions in the sacrococcygeal  location that can extend into the presacral space include giant cell tumor, teratoma, primary sarcoma, plasmacytoma, lymphoma, and metastatic disease.172 Chordomas are usually treated with complete surgical resection, radiation therapy, and/or chemotherapy.31 However, there is a high local recurrence rate following surgery, for which palliative treatment using image-guided radiofrequency ablation may be used.22,839,948,951 Metastatic disease occurs in up to 40% of patients, and the estimated 5-year and 10-year overall survival rates are 87.8% and 48.9%, respectively.49 Overall, chondroid chordomas tend to have a better prognosis than typical chordomas.821

Parachordoma Parachordomas are rare neuroepithelial tumors with glial differentiation that are histologically similar to chordomas but do not occur in the axial skeleton.921,610 They usually occur in the deep soft tissues of the extremities or in the skin, are generally slow growing and nondisseminating, and sometimes have late local recurrence following resection.2,12 Occasionally they may occur within the retroperitoneum and have a nonspecific imaging appearance on CT and MRI. Calcifications may be visualized within the masses at CT.12

Aggressive Angiomyxoma Aggressive angiomyxoma is a rare, benign, locally infiltrative neoplasm that occurs most commonly in women in the second through fourth decades of life and has a predilection for the pelvic soft tissues, perineum, vagina, vulva, and less commonly affects the buttocks, retroperitoneum, and inguinal regions.58,517,776,785,795 Clinically it often presents as a palpable pelvic or perineal mass. On gross examination, a partially or completely encapsulated soft bulky mass with a homogeneous glistening cut surface is seen. Histologically, aggressive angiomyxoma is characterized by a predominantly myxoid stroma with fibroblasts, myofibroblasts, collagen fibers, and an abundance of thin- and thick-walled vascular channels along with scattered spindle cells. Nuclear atypia and mitosis are not present.384,795 Some lesions may express estrogen and progesterone receptors.514 On CT a well-defined mass with attenuation less than or equal to that of adjacent skeletal muscle is typically seen, often with swirled enhancing tissue internally (Fig. 55-44).363,640 On MRI, low to intermediate T1-weighted SI and very high T2-weighted SI relative to skeletal muscle are typically seen owing to the myxoid matrix and high water content, frequently with swirled or layered low-SI tissue internally that moderately enhances because of fibrovascular stroma.339,363,640 The mass typically grows around structures of the pelvic floor, causing displacement of adjacent organs without invasion.170,640 Treatment involves complete surgical resection, although local recurrences are common, usually because of incomplete resection.58,514,785 Hormonal therapy with gonadotropin-releasing hormone agonists may have potential utility for treatment of these lesions.514

1916

PART II  CT and MR Imaging of the Whole Body

M

U

V

B

B M

V

A U B

U

M

V

B

D M

C FIG 55-44  Extraperitoneal pelvic aggressive angiomyxoma in 45-year-old woman with intermittent low back pain. A, Coronal contrast-enhanced reconstructed CT image through pelvis demonstrates well-circumscribed soft tissue attenuation extraperitoneal right pelvic floor mass (M) that exerts mild mass effect upon the urinary bladder (B) and vagina (V). U, uterus. B to D, Axial T1-weighted (B), coronal fat-suppressed T2-weighted (C), and coronal fat-suppressed contrast-enhanced T1-weighted (D) images show homogeneous intermediate T1 SI and slightly high T2 SI of mass (M) relative to skeletal muscle with internal curvilinear low T2 SI focus (short arrow) due to fibromuscular stroma. Note homogeneous enhancement of mass on contrast-enhanced image. B, urinary bladder; U, uterus; V, vagina.

Extramedullary Plasmacytoma and Multiple Myeloma Extramedullary plasmacytoma (EMP) is an uncommon plasma cell tumor that can involve almost any location outside the bone marrow, although predominantly it affects the upper aerodigestive tract. It occurs most commonly during the fourth through seventh decades of life, more commonly in men than in women.11,433 Retroperitoneal involvement may occur but is extremely rare.117,391,415,862 EMP is a tumor composed of monoclonal plasma cells arranged in clusters or sheets. Progression to multiple myeloma (MM) varies from 10% to 30% in EMP.915 Diagnosis of EMP requires the exclusion of MM, and patients with EMP should have a negative bone marrow biopsy, normal protein

electrophoresis, and a normal radiographic skeletal survey.572 Presenting symptoms include ureteral obstruction or vascular occlusion, flank pain, lower abdominal pain, mechanical ileus, and even tumor thrombus in the renal vein.415,862 Hyperamylasemia may be an associated finding, and serum amylase levels may be used as a tumor marker for monitoring tumor progression and treatment response.862 Treatment with complete surgical resection, sometimes in combination with radiation therapy, is typically implemented.11 MM is a B-cell malignancy of antibody-secreting plasma cells that accounts for 1% of all malignancies, has a progressive increase in incidence with age and a peak occurrence in the sixth and seventh decades of

CHAPTER 55  Retroperitoneum life, and is more common in men than in women. It involves medullary sites, with predominant clinical manifestations related to bone marrow infiltration and osseous destruction, including anemia, immunosuppression, pathologic fractures, hypercalcemia, and renal failure. However, extramedullary involvement is seen in up to 70% of patients, most frequently involving the liver, spleen, kidneys, and lymph nodes, and it can involve the retroperitoneum.379,433,572,637,907 The diagnosis of MM is based on the presence of more than 10% plasma cells in the bone marrow and a biopsy-proven plasmacytoma, or the presence of one of these findings together with a paraprotein in the serum or urine, or focal lytic bone marrow lesions or diffuse osteopenia seen on plain film radiography, CT, or MRI.182,183 Treatment involves chemotherapy and transplantation of hematopoietic stem cells, sometimes with radiation therapy.51 Cross-sectional imaging of retroperitoneal EMP and retroperitoneal extramedullary MM may demonstrate nonspecific well-circumscribed or infiltrative homogeneous-enhancing retroperitoneal or pelvic soft  tissue masses or lymphadenopathy. There may be mass effect on other adjacent structures such as vessels, or invasion of adjacent organs. It may mimic the imaging appearance of other malignancies such as lymphoma.250,391,572,751,862,915 Active lesions demonstrate uptake of FDG on PET imaging.184 Bone marrow involvement and osseous destruction are also seen in patients with MM on CT, MRI, or FDG PET scanning.182,184,907

Granulocytic Sarcoma (Chloroma) Granulocytic sarcoma, also known as chloroma, is a rare solid tumor composed of primitive precursors of the granulocytic series of leukocytes that include myeloblasts, promyelocytes, and myelocytes. It represents a subset of extramedullary disease that is associated with acute myeloid leukemia as well as other myeloproliferative disorders. It has a predilection to involve soft tissue, bone, skin, lymph nodes, and periosteum and can involve the retroperitoneum. It occurs more commonly in children than in adults but equally among men and women.224,570,607,632,684 Treatment usually involves chemotherapy, sometimes along with radiation therapy.346,936 On CT and MRI, nonspecific focal or diffusely infiltrative lesions may be seen in the retroperitoneum or pelvis; these have soft tissue attenuation, intermediate T1-weighted SI and high T2-weighted SI relative to skeletal muscle, and variable enhancement, mimicking the appearance of other malignancies such as lymphoma.632,666,684

Extragastrointestinal Gastrointestinal Stromal Tumor GISTs are the most common mesenchymal tumors specific to the GI tract, are derived from Cajal cells or their precursors, and can be benign or malignant in behavior. They most commonly occur in patients older than the fifth decade of life within the stomach (60%), jejunum and ileum (30%), duodenum (4%), rectum (4%), colon and appendix (2%), and esophagus (<1%), and rarely as primary extra-GI tumors in the retroperitoneum.559,654,825,830 CD117 (c-kit proto-oncogene protein product) is expressed in most GISTs, including extra-GI GISTs, and plays a crucial role in the pathogenesis and treatment of this disease.600 Histologically, spindle cell types (70%), epithelioid types (20%) and mixed types (10%) may occur.559 On cross-sectional imaging, nonspecific enhancing heterogeneous large soft tissue masses are typically seen in the retroperitoneum, often with areas of cystic change or necrosis, sometimes with regions of hemorrhage or calcification.654,825,830 Metastatic lesions may also be seen in cases of malignant extra-GI GIST. Treatment is typically with surgical resection and use of imatinib mesylate (Gleevec).

Extraarticular Pigmented Villonodular Synovitis Pigmented villonodular synovitis (PVNS) describes a group of benign synovial-based locally aggressive neoplasms involving tendon sheaths,

1917

less commonly joints, and rarely bursae.355,791 Lesions are subtyped according to site (intraarticular vs. extraarticular) and growth pattern (localized vs. diffuse).875 PVNS most commonly refers to the diffuse intraarticular form that primarily affects joints of the lower extremity, most commonly the knee and less commonly the hip.251 The extraarticular form of diffuse PVNS, also known as diffuse-type giant cell tumor (D-TGCT), is characterized by an infiltrative extraarticular mass with or without involvement of an adjacent joint, which may mimic malignancy such as a sarcoma.256,791 Rarely, D-TGCT that involves the hip joint and ipsilateral iliopsoas bursa may extend superiorly along the iliopsoas muscle into the pelvis and present as a large retroperitoneal mass.376 D-TGCT affects patients of all ages and has a slight female predominance.791 At histology, characteristic findings of PVNS include a prominent mononuclear cell infiltrate, multinucleated giant cells, foamy macrophages, intracellular and extracellular hemosiderin deposition, and stromal fibrosis. However, the cell populations are not uniformly distributed, and in particular giant cells and foam cells may be difficult to identify.355 Furthermore, D-TGCT tends to have poorly circumscribed infiltrative margins and frequently demonstrates a more varied histologic appearance, with less frequent and less prominent hemosiderin deposition, making the differentiation from sarcoma more difficult.791 On CT and MRI, a large heterogeneous pelvic or retroperitoneal mass predominantly with soft tissue attenuation, intermediate SI on T1, and high SI on T2 relative to skeletal muscle is seen. Contiguous tissue in the region of the ipsilateral iliopsoas bursa or hip joint may also be present and is associated with a thickened enhancing hyperplastic hip joint synovium as well as high T2-weighted SI bone marrow edema subjacent to sites of osseous erosion.329,376,791 Foci of fat due to clusters of lipid-laden macrophages may be seen in the mass, with attenuation and SI characteristics similar to subcutaneous fat, and foci of encysted fluid may be present with water attenuation, low SI on T1, and very high SI on T2. Foci of hemosiderin deposition and fibrosis may also be seen as areas of low SI on T1- and T2-weighted images.329,376,489 Complete surgical resection of the mass is the treatment of choice, along with hip joint synovectomy and sometimes total hip arthroplasty if there is involvement of the hip joint.251,791,895 Wide excision margins for D-TGCT are typically used because of the high potential for disease recurrence.791

MISCELLANEOUS NONNEOPLASTIC RETROPERITONEAL CONDITIONS Amyloidosis Amyloidosis comprises a heterogeneous group of conditions caused by extracellular deposition of an insoluble fibrillar protein in organs and tissues, most often involving the kidneys and heart, less commonly involving the liver and spleen. Rarely there may be abdominal involvement manifested by either nonspecific mesenteric or retroperitoneal lymphadenopathy, which may demonstrate calcification, or diffuse mesenteric, omental, or retroperitoneal infiltration.145,435 The three common systemic amyloidoses include primary, familial, and secondary forms; primary and secondary amyloidosis account for the majority of cases.30,145 Amyloid light (AL) amyloidoses are associated with monoclonal immunoglobulin light chains and are related to primary systemic amyloidosis, multiple myeloma, or plasma cell dyscrasias. Amyloid A (AA) (secondary) amyloidoses are the result of systemic accumulation of serum amyloid A protein (SAA), an acutephase protein produced in response to chronic inflammatory disease such as rheumatoid arthritis, familial Mediterranean fever, and chronic

1918

PART II  CT and MR Imaging of the Whole Body

infection. Mutant transthyretin (ATTR) amyloidosis is the most common form of familial amyloidosis.145 The diagnosis of amyloidosis is generally confirmed on histologic examination using Congo red staining and optical birefringence under polarized light; amyloid deposits characteristically show apple-green birefringence.244 On CT and MRI, nonspecific retroperitoneal or pelvic masses, lymphadenopathy, or diffusely infiltrative soft tissue are seen, sometimes with very-high-attenuation coarse dystrophic calcification or avid enhancement.30,80,211,244,338,667,828 Treatment is generally targeted against the underlying plasma cell dyscrasia in AL amyloidosis using chemotherapy or hematopoietic stem cell transplantation, and against the underlying infectious or inflammatory disorder in AA amyloidosis.247 Systemic amyloidosis tends to be progressive and can be fatal, usually from renal or cardiac failure.

Malacoplakia Malacoplakia is an uncommon granulomatous disease that is most frequently found in the bladder or ureters but has been reported to occur in almost every other organ or structure, including the retroperitoneum, where it can present as a nonspecific soft tissue mass.692,831,843 Although the underlying etiology for malacoplakia is unknown, it most likely involves a defect in macrophage enzymatic function. Patients of all ages may be affected, with a peak incidence in the fifth to seventh decades of life and a 2 to 4 : 1 female-to-male ratio. The clinical presentation is variable and nonspecific, but most patients have active urinary tract infection (with Escherichia coli detected in 90%). A palpable abdominal mass, abdominal discomfort, and fever are the most frequent symptoms encountered.692 At histology, a chronic granulomatous reaction is seen, characterized by the accumulation of distinctive macrophages (von Hansemann cells) that contain pathognomonic basophilic staining inclusions (Michaelis-Gutmann bodies), which are thought to represent phagocytized fragments of bacteria.692,831 On cross-sectional imaging, a nonspecific retroperitoneal mass may be visualized. However, malacoplakic masses in the retroperitoneum are almost always due to extension from an adjacent organ such as the kidney, ureter, adrenal gland, or large bowel. Treatment with antibiotics is typically implemented, although surgical resection or radiation therapy may also be used if the disease does not respond to medical therapy. The mass is frequently unresectable, and recurrence and fistula formation may occur, producing significant morbidity.692

Lymphangioleiomyomatosis LAM, a forme fruste of tuberous sclerosis, is a rare idiopathic disorder that occurs in women of childbearing age and predominantly affects the lungs, less often the mediastinum and retroperitoneum.644,874 The disease is characterized by a hamartomatous proliferation of smooth muscle cells that involves lymph nodes and lymphatics, resulting in dilation of lymphatic spaces with cyst formation.143,837,927 On CT or MRI, cystic lung disease, chylous pleural effusions, pneumothoraces, and renal angiomyolipomas are most frequently encountered. Retroperitoneal or pelvic lymphangioleiomyomas are less common, occurring in up to 20% of patients. These appear as lowattenuation high-T2-SI cystic masses with enhancing walls, sometimes with delayed enhancement of the internal contents35,36,69,465,579,626,644,856 Diurnal variation in size of lymphangioleiomyomas (with an increase in lesion size during the day) is frequently encountered on crosssectional imaging and is thought to be related in part to an increase in abdominopelvic lymph flow during the day, related to an increase in chyle production after meals.35 Abdominopelvic lymphadenopathy may be seen in up to 40% of patients, some of which contains lowattenuation high-T2-SI areas that are due to chylous lymph collections,

or hamartomatous hyperattenuating areas that enhance.36,35,375,929 Chylous ascites may also be seen infrequently.644 Surgical resection of symptomatic retroperitoneal lymphangioleiomyomas may be performed, and additional therapeutic interventions for pulmonary, renal, or other organ involvement are also used as needed.

Xanthogranulomatosis/Erdheim-Chester Disease Xanthogranulomatosis, a rare multicentric masslike accumulation of non-Langerhans lipid-laden histiocytes, is an idiopathic process with a variable clinical course and a predisposition for the retroperitoneum. When xanthogranulomatosis affects multiple organs, it is called Erdheim-Chester disease, occurs slightly more commonly in men than in women, and predominantly affects middle-aged and elderly patients, with a peak incidence in the sixth decade of life.188,899 Clinical and laboratory findings are variable, as is clinical outcome; some patients have an indolent course, whereas others develop systemic disease that can be fatal. Therapeutic options include corticosteroid therapy, chemotherapy, radiation therapy, or immunotherapy, but all have variable efficacy.899 On CT and MRI, nonspecific mildly enhancing infiltrative soft tissue with attenuation similar to or less than that of skeletal muscle and intermediate T1- and T2-weighted SI relative to skeletal muscle are typically seen, sometimes with an admixture of macroscopic fat, often with infiltration of adjacent organs.159,173,216,387,743 Periaortic involvement by xanthogranulomatosis tends to be circumferential and is frequently associated with bilateral symmetric perirenal space involvement, whereas the IVC and pelvic ureters are typically spared, which are useful cross-sectional imaging findings for differentiation of retroperitoneal xanthogranulomatosis from RPF.173 Osteosclerosis is seen on CT. On MRI, symmetric bilateral replacement of the normal fatty bone marrow within the diaphyses and metaphyses of the long bones by heterogeneous low SI on T1 and intermediate to high SI on T2 is typically present in Erdheim-Chester disease, often in association with findings of periostitis, partial epiphyseal involvement, and medullary infarction.159,174

Extramedullary Hematopoiesis Extramedullary hematopoiesis (EMH) is usually of little clinical significance, is relatively limited in extent, typically occurs in a paraspinal location, and may occur as a compensatory mechanism of red cell production in the setting of myelofibrosis, chronic anemia, or hemoglobinopathy. In approximately 5% of cases, aggressive infiltrations and tumefactions of EMH may be noted, which may be clinically significant. EMH may cause urinary tract obstruction, bowel obstruction, portal hypertension with ascites, spinal cord compression, and occlusion of pulmonary and myocardial vessels by emboli of hematopoietic tissue.748 Large masses may rarely be seen in the retroperitoneum, abdominal lymph nodes, mesentery, omentum, or parenchymal organs.548,748 At histology, immature and mature cells, mainly from the erythroid and myeloid series, and dilated sinusoids containing precursors of red cells are seen early on, whereas fatty tissue, fibrosis, or massive iron deposits may be seen later in association with inactivity.866 On CT and MRI, multiple bilateral round or lobulated homogeneous masses with soft tissue attenuation, intermediate SI on T1-weighted images, intermediate to high SI on T2-weighted images relative to skeletal muscle, and variable enhancement may be seen in the retroperitoneum or extraperitoneal pelvis (Fig. 55-45).548,866,904 Areas with attenuation and SI similar to those of macroscopic fat may be present as well.548,935 Calcification and osseous destruction are not usually present. Similar appearing lesions are more commonly

CHAPTER 55  Retroperitoneum

*

1919

* *

M

FIG 55-46  Retroperitoneal varices in 78-year-old man with nonFIG 55-45  Extraperitoneal pelvic extramedullary hematopoiesis in 33-year-old gravid woman with abdominal pain and β-thalassemia. Axial contrast-enhanced CT image demonstrates homogeneous soft tissue mass (M) within presacral space of extraperitoneal pelvis and within sacral foramina, without osseous destruction. Note associated osteopenia and accentuated trabecular pattern of sacrum, as well as gravid uterus anteriorly.

Hodgkin’s lymphoma and chronically occluded IVC. Axial contrastenhanced CT image through abdomen demonstrates multiple tubular serpentine soft tissue attenuation retroperitoneal lesions (*) that communicate with each other (not shown), in keeping with varices. Note very-high-attenuation focus (arrow) associated with streak artifacts in the region of the obliterated IVC, in keeping with an IVC filter.

rare lesion that has been reported to occur in the retroperitoneum. On cross-sectional imaging a nonspecific, large, multicystic retroperitoneal mass is seen with enhancing septations.761,910 encountered in the posterior mediastinum, often in association with hepatosplenomegaly.748 Following therapy using blood transfusions, lesions may decrease in size, develop attenuation greater than that of skeletal muscle, as well as low SI on T1 and T2 due to massive iron deposition, and do not enhance.866 Skeletal changes due to chronic anemia or myelofibrosis may also be seen.498,548,748

Inflammatory Pseudotumor Inflammatory pseudotumor (inflammatory myofibroblastic tumor) is a rare benign idiopathic nonmetastasizing proliferation of myofibroblasts and inflammatory cells that mainly affects those in the first and second decades of life, most commonly affecting the lung and the orbit and less commonly occurring in the retroperitoneum.135,365,418,603 Clinical presentation may include constitutional symptoms, a palpable abnormality, and laboratory abnormalities such as an elevated ESR. At histology, it is characterized by myofibroblast proliferation, often associated with myxoid and inflammatory regions that may have myxoid, vascular, and inflammatory areas resembling nodular fasciitis, compact spindle cells with intermingled inflammatory cells (lymphocytes, plasma cells, and eosinophils) resembling fibrous histiocytoma, or dense platelike collagen resembling fibromatosis or scar.135 On CT or MRI a nonspecific mass with infiltrative or well-defined borders is seen, usually with soft tissue attenuation similar to or less than that of skeletal muscle, low T1- and high T2-weighted SI relative to skeletal muscle, and variable amounts of enhancement.365,418,603,667 Complete surgical resection is the treatment of choice, although high-dose corticosteroids, radiation therapy, and chemotherapy can also be implemented.603 Although inflammatory pseudotumor has the potential for local recurrence and persistent growth, distant metastases are rare and the prognosis is generally favorable.135

Neurogenous Choristoma A neurogenous choristoma is due to a collection of histologically normal mature glioneural tissue that is ectopically situated; it is a very

Ectopic Splenic Tissue Very rarely an ectopic spleen or accessory spleen may be found in the retroperitoneum as a normal variant.561,720,894

Ectopic Pancreatic Tissue Extremely rarely, ectopic pancreatic tissue may be encountered in the retroperitoneum, potentially mimicking a retroperitoneal mass.490

Varices Retroperitoneal varices may occur idiopathically or secondary to portal hypertension with formation of retroperitoneal portosystemic collaterals.601 Retroperitoneal hemorrhage or hemoperitoneum due to retroperitoneal varices can occur spontaneously and can be acutely life threatening.427,648 On CT and MRI, well-defined round, tubular, or serpentine smoothly marginated structures with enhancement similar to that of adjacent veins are typically seen (Fig. 55-46).121

Retained Surgical Sponge (Gossypiboma) Sponges are the most common postsurgical retained foreign body and are also referred to as gossypibomas and more recently as textilomas or cottonoids, because they are now made of synthetic materials. Retained surgical sponges are more frequent with emergency surgery, hemorrhagic procedures, time-consuming operations, operations in anatomic regions that are difficult to reach, an unexpected change in a surgical procedure, a change in operating theater personnel, and in patients with a higher body mass index.47,237,458,582 Surgical sponges are often left behind during operations in the lesser pelvis, explaining the high incidence of gossypibomas following gynecologic procedures, but have also been reported to occur in the retroperitoneum.60,458,569,945 Operations in the lower pelvis in combination with operations on the GI tract account for about 75% of all reported retained surgical sponges (digestive surgery 52%, gynecology 22%).458,472 Approximately 50% of retained surgical sponges are discovered 5 years or more after

1920

PART II  CT and MR Imaging of the Whole Body

surgery, and 40% are detected within the first year following surgery.697 An infrequently reported possible complication of a retained surgical sponge is development of an angiosarcoma.362 Variable reactions can occur in response to a retained surgical sponge, such as an exudative response with formation of abscess, fistula, bowel obstruction, or erosion into adjacent structures.624,959 Septic complications, including abscess formation, are more likely to occur in the early postoperative period.25,697,959 Alternatively, patients may have an aseptic inflammatory/fibrotic response with encapsulation, resulting in a granuloma, and can either present years later with symptoms resulting from mass effect upon other abdominal organs or may remain asymptomatic, with incidental detection of the sponge on imaging performed for an unrelated clinical indication.47,959 The cross-sectional imaging appearance of a retained sponge can vary with sponge composition, type of foreign body reaction, and time elapsed from surgery. On CT a well-circumscribed soft tissue attenuation mass with a whorled texture and very-low-attenuation foci related to gas trapped in the fiber meshwork is typically seen. Associated abscess formation may also be detected as a loculated fluid collection with thick rim enhancement.377,419,458,569,652,764 Peripheral calcification may sometimes be seen as well.458 Scout images may be useful in the detection of the radiopaque marker, which can potentially be obscured by beam-hardening artifact.624 On MRI a well-defined mass with variable T1-weighted and high T2-weighted SI central fluid is seen, often with wavy low-SI nonenhancing internal structures on T1- and T2-weighted images that have a whorled appearance due to the sponge material, as well as a low-SI enhancing peripheral capsule, frequently with a serrated inner contour.60,231,404,445,458,569,624 Adjacent surgical clips may also be seen, with high attenuation and very low SI on T1- and T2-weighted images. Consideration of the possibility that a retroperitoneal or pelvic mass may be the result of a retained sponge is probably the most important step in making the diagnosis.679 Treatment consists of thorough surgical exploration of the affected anatomic site, removal of the retained surgical sponge, drainage of purulent fluid, and treatment of accompanying lesions such as fistulas.458,945 When a retained surgical sponge is diagnosed and removed in the immediate postoperative period, morbidity and mortality are low. However, if there is a significant time delay, major surgical interventions may be needed and there is a higher risk of complications and mortality. Mortality due to a retained surgical sponge may be as high as 11% to 35%, and the morbidity rate is close to 50%.458

REFERENCES 1. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin’s lymphoma. The Non-Hodgkin’s Lymphoma Classification Project. Blood 89:3909–3918, 1997. 2. Abe S, Imamura T, Harasawa A, et al: Parachordoma with multiple metastases. J Comput Assist Tomogr 27:634–638, 2003. 3. Adam YG, Alberts W: Retroperitoneal hemangioma. Am Surg 56:374–376, 1990. 4. Aflalo-Hazan V, Rousset P, Mourra N, et al: Tailgut cysts: MRI findings. Eur Radiol 18:2586–2593, 2008. 5. Agrawal A, Nair N, Baghel N: F-18 FDG PET in Ormond disease in a patient with renal cell carcinoma. Clin Nucl Med 32:320–322, 2007. 6. Aikawa H, Tanoue S, Okino Y, et al: Pelvic extension of retroperitoneal fluid: Analysis in vivo. AJR Am J Roentgenol 171:671–677, 1998. 7. Aizenstein RI, Wilbur AC, O’Neil HK: Interfascial and perinephric pathways in the spread of retroperitoneal disease: Refined concepts based on CT observations. AJR Am J Roentgenol 168:639–643, 1997. 8. Akinci D, Akhan O, Ozmen M, et al: Diaphragmatic mesothelial   cysts in children: Radiologic findings and percutaneous ethanol sclerotherapy. AJR Am J Roentgenol 185:873–877, 2005.

9. Alektiar KM, Hu K, Anderson L, et al: High-dose-rate intraoperative radiation therapy (HDR-IORT) for retroperitoneal sarcomas.   Int J Radiat Oncol Biol Phys 47:157–163, 2000. 10. Alexander ES, Colley DP, Clark RA: Computed tomography of retroperitoneal fluid collections. Semin Roentgenol 16:268–276, 1981. 11. Alexiou C, Kau RJ, Dietzfelbinger H, et al: Extramedullary plasmacytoma: Tumor occurrence and therapeutic concepts. Cancer 85:2305–2314, 1999. 12. Alfieri S, Prete FP, Di Giorgio A, et al: Retroperitoneal parachordoma   in a patient with a history of recurrent pain. Lancet Oncol 6:350, 2005. 13. Allen FJ, De Kock ML: Pelvic lipomatosis: The nuclear magnetic resonance appearance and associated vesicoureteral reflux. J Urol 138:1228–1230, 1987. 14. Altman AR, Johnson TH: Pneumoperitoneum and pneumoretroperitoneum. Consequences of positive end-expiratory pressure therapy. Arch Surg 114:208–211, 1979. 15. Amin R: Chylous ascites from prostatic adenocarcinoma. Urology 59:773, 2002. 16. Amis ES, Jr: Retroperitoneal fibrosis. AJR Am J Roentgenol 157:321–329, 1991. 17. Andac N, Baltacioglu F, Cimsit NC, et al: Fat necrosis mimicking liposarcoma in a patient with pelvic lipomatosis. CT findings. Clin Imaging 27:109–111, 2003. 18. Anderson CM, Kerby JD, Perry WB, et al: Pneumoretroperitoneum in two patients with Clostridium perfringens necrotizing pancreatitis. Am Surg 70:268–271, 2004. 19. Anderson SE, Schwab C, Stauffer E, et al: Hibernoma: Imaging characteristics of a rare benign soft tissue tumor. Skeletal Radiol 30:590–595, 2001. 20. Andres E, Maloisel F: Interferon-alpha as first-line therapy for treatment of multicentric Castleman’s disease. Ann Oncol 11:1613–1614, 2000. 21. Angulo JC, Escribano J, Diego A, et al: Isolated retrovesical and extrarenal retroperitoneal hydatidosis: Clinical study of 10 cases and literature review. J Urol 159:76–82, 1998. 22. Anis N, Chawki N, Antoine K: Use of radio-frequency ablation for the palliative treatment of sacral chordoma. AJNR Am J Neuroradiol 25:1589–1591, 2004. 23. Antoch G, Vogt FM, Freudenberg LS, et al: Whole-body dual-modality PET/CT and whole-body MRI for tumor staging in oncology. JAMA 290:3199–3206, 2003. 24. Apter S, Avigdor A, Gayer G, et al: Calcification in lymphoma occurring before therapy: CT features and clinical correlation. AJR Am J Roentgenol 178:935–938, 2002. 25. Apter S, Hertz M, Rubinstein ZJ, et al: Gossypiboma in the early post-operative period: A diagnostic problem. Clin Radiol 42:128–129, 1990. 26. Arkun R, Memis A, Akalin T, et al: Liposarcoma of soft tissue: MRI findings with pathologic correlation. Skeletal Radiol 26:167–172, 1997. 27. Armatys SA, Cheng L, Gardner TA, et al: Pulmonary sequestration presenting as retroperitoneal cyst: Case report. J Endourol 19:997–999, 2005. 28. Arrive L, Correas JM, Leseche G, et al: Inflammatory aneurysms of the abdominal aorta: CT findings. AJR Am J Roentgenol 165:1481–1484, 1995. 29. Arrive L, Hricak H, Tavares NJ, et al: Malignant versus nonmalignant retroperitoneal fibrosis. Radiology 172:139–143, 1989. 30. Asaumi J, Yanagi Y, Hisatomi M, et al: CT and MR imaging of localized amyloidosis. Eur J Radiol 39:83–87, 2001. 31. Atalar H, Selek H, Yildiz Y, et al: Management of sacrococcygeal chordomas. Int Orthop 30:514–518, 2006. 32. Atilla S, Eilenberg SS, Brown JJ: Hibernoma: MRI appearance of a rare tumor. Magn Reson Imaging 13:335–337, 1995. 33. Attariwala R, Picker W: Whole body MRI: Improved lesion detection and characterization with diffusion weighted techniques. J Magn Reson Imaging 38:253–268, 2013.

CHAPTER 55  Retroperitoneum 34. Auh YH, Rubenstein WA, Schneider M, et al: Extraperitoneal paravesical spaces: CT delineation with US correlation. Radiology 159:319–328, 1986. 35. Avila NA, Bechtle J, Dwyer AJ, et al: Lymphangioleiomyomatosis: CT of diurnal variation of lymphangioleiomyomas. Radiology 221:415–421, 2001. 36. Avila NA, Kelly JA, Chu SC, et al: Lymphangioleiomyomatosis: abdominopelvic CT and US findings. Radiology 216:147–153, 2000. 37. Awasthi M, Narlawar R, Hira P, et al: Fetus in fetu. Rare cause of a lump in an adult’s abdomen. Australas Radiol 45:354–356, 2001. 38. Aydinli B, Ozturk G, Polat KY, et al: Extravisceral primary hydatid cyst of the retroperitoneum. ANZ J Surg 77:455–459, 2007. 39. Baath L, Nyman U, Aspelin P, et al: Computed tomography of pelvic lipomatosis. Report of a case. Acta Radiol Diagn (Stockh) 27:311–314, 1986. 40. Baker EL, Gore RM, Moss AA: Retroperitoneal pulmonary sequestration: Computed tomographic findings. AJR Am J Roentgenol 138:956–957, 1982. 41. Baker KS, Lee D, Huang M, et al: Presacral myelolipoma: a case report and review of imaging findings. J Radiol Case Rep 6:1–9, 2012. 42. Bakir B, Yilmaz F, Turkay R, et al: Role of diffusion-weighted MR imaging in the differentiation of benign retroperitoneal fibrosis from malignant neoplasm: Preliminary study. Radiology 272:438–445, 2014. 43. Balachandran S, Leonard MH, Jr, Kumar D, et al: Patterns of fluid accumulation in splenic trauma: Demonstration by CT. Abdom Imaging 19:515–520, 1994. 44. Balci NC, Semelka RC, Noone TC, et al: Acute and subacute liverrelated hemorrhage: MRI findings. Magn Reson Imaging 17:207–211, 1999. 45. Balfe DM, Mauro MA, Koehler RE, et al: Gastrohepatic ligament: normal and pathologic CT anatomy. Radiology 150:485–490, 1984. 46. Ballo MT, Zagars GK, Pollock RE, et al: Retroperitoneal soft tissue sarcoma: An analysis of radiation and surgical treatment. Int J Radiat Oncol Biol Phys 67:158–163, 2007. 47. Bani-Hani KE, Gharaibeh KA, Yaghan RJ: Retained surgical sponges (gossypiboma). Asian J Surg 28:109–115, 2005. 48. Bankier AA, Stanek C, Hubsch P: Case report: Benign solitary schwannoma of the greater omentum: A rare cause of acute intraperitoneal bleeding—Diagnosis by CT. Clin Radiol 51:517–518, 1996. 49. Baratti D, Gronchi A, Pennacchioli E, et al: Chordoma: Natural history and results in 28 patients treated at a single institution. Ann Surg Oncol 10:291–296, 2003. 50. Barker CD, Brown JJ: MR imaging of the retroperitoneum. Top Magn Reson Imaging 7:102–111, 1995. 51. Barlogie B, Shaughnessy J, Tricot G, et al: Treatment of multiple myeloma. Blood 103:20–32, 2004. 52. Barrington SF, Mikhaeel NG, Kostakoglu L, et al: Role of imaging in the staging and response assessment of lymphoma: Consensus of the International Conference on Malignant Lymphomas Imaging Working Group. J Clin Oncol 32:3048–3058, 2014. 53. Bass JC, Korobkin M, Francis IR, et al: Retroperitoneal plexiform neurofibromas: CT findings. AJR Am J Roentgenol 163:617–620, 1994. 54. Bassetto MA, Pasini F, Franceschi T, et al: Extragonadal germ cell tumor: A clinical study. Anticancer Res 15:2751–2754, 1995. 55. Bautista N, Su W, O’Connell TX: Retroperitoneal soft-tissue sarcomas: Prognosis and treatment of primary and recurrent disease. Am Surg 66:832–836, 2000. 56. Beaulieu CF, Mindelzun RE, Dolph J, et al: The infraconal compartment: A multidirectional pathway for spread of disease between the extraperitoneal abdomen and pelvis. J Comput Assist Tomogr 21:223–228, 1997. 57. Bechtold RE, Dyer RB, Zagoria RJ, et al: The perirenal space: Relationship of pathologic processes to normal retroperitoneal anatomy. Radiographics 16:841–854, 1996. 58. Begin LR, Clement PB, Kirk ME, et al: Aggressive angiomyxoma of pelvic soft parts: A clinicopathologic study of nine cases. Hum Pathol 16:621–628, 1985.

1921

59. Bellin MF, Duron JJ, Curet P, et al: Primary retroperitoneal teratoma in the adult: Correlation of MRI features with CT and pathology. Magn Reson Imaging 9:263–266, 1991. 60. Bellin M, Hornoy B, Richard F, et al: Perirenal textiloma: MR and serial CT appearance. Eur Radiol 8:57–59, 1998. 61. Bellin MF, Lebleu L, Meric JB: Evaluation of retroperitoneal and pelvic lymph node metastases with MRI and MR lymphangiography. Abdom Imaging 28:155–163, 2003. 62. Bennett HF, Balfe DM: MR imaging of the peritoneum and abdominal wall. Magn Reson Imaging Clin N Am 3:99–120, 1995. 63. Berglund AS, Hulthen UL, Manhem P, et al: Metaiodobenzylguanidine (MIBG) scintigraphy and computed tomography (CT) in clinical practice. Primary and secondary evaluation for localization of phaeochromocytomas. J Intern Med 249:247–251, 2001. 64. Bergstrom M, Ericson K, Levander B, et al: Variation with time of the attenuation values of intracranial hematomas. J Comput Assist Tomogr 1:57–63, 1977. 65. Bertagna F, Treglia G, Leccisotti L, et al: [(1)(8)F]FDG-PET/CT in patients affected by retroperitoneal fibrosis: A bicentric experience. Jpn J Radiol 30:415–421, 2012. 66. Berti E, Roncaroli F: Fibrolipomatous hamartoma of a cranial nerve. Histopathology 24:391–392, 1994. 67. Bessell-Browne R, O’Malley ME: CT of pheochromocytoma and paraganglioma: Risk of adverse events with i.v. administration of nonionic contrast material. AJR Am J Roentgenol 188:970–974, 2007. 68. Bhatia KS, Ismail MM, Sahdev A, et al: 123I-metaiodobenzylguanidine (MIBG) scintigraphy for the detection of adrenal and extra-adrenal phaeochromocytomas: CT and MRI correlation. Clin Endocrinol (Oxf) 69:181–188, 2008. 69. Bhattacharyya AK, Balogh K: Retroperitoneal lymphangioleiomyomatosis. A 36-year benign course in a postmenopausal woman. Cancer 56:1144–1146, 1985. 70. Bibbo C, Warren AM: Fibrolipomatous hamartoma of nerve. J Foot Ankle Surg 33:64–71, 1994. 71. Blackledge G, Best JJ, Crowther D, et al: Computed tomography (CT) in the staging of patients with Hodgkin’s Disease: A report on 136 patients. Clin Radiol 31:143–147, 1980. 72. Bland KI, McCoy DM, Kinard RE, et al: Application of magnetic resonance imaging and computerized tomography as an adjunct to the surgical management of soft tissue sarcomas. Ann Surg 205:473–481, 1987. 73. Blessios GA, Rokkas CK, Neulander O, et al: Laparoscopic management of large retroperitoneal lymphoceles complicating aortic surgery. Surg Endosc 16:218, 2002. 74. Blomlie V, Lien HH, Fossa SD, et al: CT in primary malignant germ cell tumors of the retroperitoneum. Acta Radiol 32:155–158, 1991. 75. Bloom DA, Scardino PT, Ehrlich RM, et al: The significance of lymph nodal involvement in renal angiomyolipoma. J Urol 128:1292–1295, 1982. 76. Boenicke L, Maier M, Merger M, et al: Retroperitoneal gas gangrene after colonoscopic polypectomy without bowel perforation in an otherwise healthy individual: Report of a case. Langenbecks Arch Surg 391:157–160, 2006. 77. Bokemeyer C, Droz JP, Horwich A, et al: Extragonadal seminoma: An international multicenter analysis of prognostic factors and long term treatment outcome. Cancer 91:1394–1401, 2001. 78. Bonamigo TP, Bianco C, Becker M, et al: Inflammatory aneurysms of infra-renal abdominal aorta. A case-control study. Minerva Cardioangiol 50:253–258, 2002. 79. Bonhomme A, Broeders A, Oyen RH, et al: Cystic lymphangioma of the retroperitoneum. Clin Radiol 56:156–158, 2001. 80. Borge MA, Parker LA, Mauro MA: Amyloidosis: CT appearance of calcified, enlarged periaortic lymph nodes. J Comput Assist Tomogr 15:855–857, 1991. 81. Bosch X, Guilabert A, Miquel R, et al: Enigmatic Kikuchi-Fujimoto disease: A comprehensive review. Am J Clin Pathol 122:141–152, 2004. 82. Bossi A, De Wever I, Van Limbergen E, et al: Intensity modulated radiation-therapy for preoperative posterior abdominal wall irradiation

1922

83.

84. 85.

86. 87. 88.

89.

90.

91. 92. 93.

94.

95.

96.

97. 98.

99. 100.

101. 102.

103.

104.

105.

106.

PART II  CT and MR Imaging of the Whole Body

of retroperitoneal liposarcomas. Int J Radiat Oncol Biol Phys 67:164– 170, 2007. Bouali K, Magotteaux P, Jadot A, et al: Percutaneous catheter drainage of abdominal abscess after abdominal surgery. Results in 121 cases.   J Belge Radiol 76:11–14, 1993. Bousvaros A, Kirks DR, Grossman H: Imaging of neuroblastoma: An overview. Pediatr Radiol 16:89–106, 1986. Boutin RD, Briggs JE, Williamson MR: Injuries associated with MR imaging: Survey of safety records and methods used to screen patients for metallic foreign bodies before imaging. AJR Am J Roentgenol 162:189–194, 1994. Bradley WG, Jr: MR appearance of hemorrhage in the brain. Radiology 189:15–26, 1993. Bradley WG, Jr: Hemorrhage and hemorrhagic infections in the brain. Neuroimaging Clin N Am 4:707–732, 1994. Brady RR, McAteer E, Weir CD: Biliscrotum and retroperitoneal biloma: Spontaneous rupture of the biliary system presenting as an incarcerated inguinal hernia. Ulster Med J 75:85–87, 2006. Braun WE, Banowsky LH, Straffon RA, et al: Lymphoceles   associated with renal transplantation: Report of fifteen cases   and review of the literature. Proc Clin Dial Transplant Forum 3:185–189, 1973. Brepoels L, Stroobants S, Verhoef G: PET and PET/CT for response evaluation in lymphoma: Current practice and developments. Leuk Lymphoma 48:270–282, 2007. Brient C, Muller C, Cassagneau P, et al: A retroperitoneal bronchogenic cyst. J Visc Surg 149:e361–e363, 2012. Brink I, Hoegerle S, Klisch J, et al: Imaging of pheochromocytoma and paraganglioma. Fam Cancer 4:61–68, 2005. Brodeur GM, Pritchard J, Berthold F, et al: Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment. J Clin Oncol 11:1466–1477, 1993. Brooks AP, Reznek RH, Nugent K, et al: CT appearances of desmoid tumours in familial adenomatous polyposis: Further observations.   Clin Radiol 49:601–607, 1994. Buckley JA, Siegelman ES, Birnbaum BA, et al: Bronchogenic cyst appearing as a retroperitoneal mass. AJR Am J Roentgenol 171:527–528, 1998. Burn PR, Singh S, Barbar S, et al: Role of gadolinium-enhanced magnetic resonance imaging in retroperitoneal fibrosis. Can Assoc Radiol J 53:168–170, 2002. Burrill J, Williams CJ, Bain G, et al: Tuberculosis: A radiologic review. Radiographics 27:1255–1273, 2007. Butori N, Guy F, Collin F, et al: Retroperitoneal extra-adrenal myelolipoma: Appearance in CT and MRI. Diagn Interv Imaging 93:e204–e207, 2012. Cai Y, Guo Z, Cai Q, et al: Bronchogenic cysts in retroperitoneal region. Abdom Imaging 38:211–214, 2013. Cai J, Zeng Y, Zheng H, et al: Retroperitoneal ganglioneuroma in children: CT and MRI features with histologic correlation. Eur J Radiol 75:315–320, 2010. Callen PW: Computed tomographic evaluation of abdominal and pelvic abscesses. Radiology 131:171–175, 1979. Canter RJ, Martinez SR, Tamurian RM, et al: Radiographic and histologic response to neoadjuvant radiotherapy in patients with soft tissue sarcoma. Ann Surg Oncol 17:2578–2584, 2010. Cantisani V, Mortele KJ, Glickman JN, et al: Large retroperitoneal hibernoma in an adult male: CT imaging findings with pathologic correlation. Abdom Imaging 28:721–724, 2003. Capitan Manjon C, Tejido Sanchez A, Piedra Lara JD, et al: Retroperitoneal abscesses—Analysis of a series of 66 cases. Scand J Urol Nephrol 37:139–144, 2003. Cappellani A, Zanghi A, Di Vita M, et al: Spontaneous rupture   of a giant hemangioma of the liver. Ann Ital Chir 71:379–383, 2000. Carbone PP, Kaplan HS, Musshoff K, et al: Report of the Committee on Hodgkin’s Disease Staging Classification. Cancer Res 31:1860–1861, 1971.

107. Caruncho M, Pombo F, Arnal-Monreal F: Primary retroperitoneal serous cystadenocarcinoma of ‘ovarian-type’: US and CT findings. Eur J Radiol 17:115–116, 1993. 108. Casillas J, Sais GJ, Greve JL, et al: Imaging of intra- and extraabdominal desmoid tumors. Radiographics 11:959–968, 1991. 109. Cavallaro MC, Taylor JA, Gorman JD, et al: Imaging findings in a patient with fibrolipomatous hamartoma of the median nerve. AJR Am J Roentgenol 161:837–838, 1993. 110. Chan JK: The new World Health Organization classification of lymphomas: The past, the present and the future. Hematol Oncol 19:129–150, 2001. 111. Charkviani L, Kekelidze N, Charkviani T: Management of lymphocysts after cervical carcinoma surgery. Eur J Gynaecol Oncol 21:487–490, 2000. 112. Chatterton BE, Mensforth D, Coventry BJ, et al: Hibernoma: Intense uptake seen on Tc-99m tetrofosmin and FDG positron emission tomographic scanning. Clin Nucl Med 27:369–370, 2002. 113. Chen CP, Chern SR, Liu FF, et al: Prenatal diagnosis, pathology, and genetic study of fetus in fetu. Prenat Diagn 17:13–21, 1997. 114. Chen KT, Felix EL, Flam MS: Extraadrenal myelolipoma. Am J Clin Pathol 78:386–389, 1982. 115. Chen JS, Lee WJ, Chang YJ, et al: Laparoscopic resection of a primary retroperitoneal mucinous cystadenoma: Report of a case. Surg Today 28:343–345, 1998. 116. Chen CH, Liu HC, Tung KY, et al: Surgical outcome of superficial and deep Castleman disease. ANZ J Surg 77:339–343, 2007. 117. Chen TC, Wu JH, Ng KF, et al: Solitary extramedullary plasmacytoma in the retroperitoneum. Am J Hematol 58:235–238, 1998. 118. Chesbrough RM, Burkhard TK, Martinez AJ, et al: Gerota versus Zuckerkandl: The renal fascia revisited. Radiology 173:845–846, 1989. 119. Cheson BD, Fisher RI, Barrington SF, et al: Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: The Lugano classification. J Clin Oncol 32:3059–3068, 2014. 120. Chiappa A, Zbar AP, Bertani E, et al: Primary and recurrent retroperitoneal soft tissue sarcoma: Prognostic factors affecting survival. J Surg Oncol 93:456–463, 2006. 121. Cho KC, Patel YD, Wachsberg RH, et al: Varices in portal hypertension: Evaluation with CT. Radiographics 15:609–622, 1995. 122. Choi BI, Chi JG, Kim SH, et al: MR imaging of retroperitoneal teratoma: Correlation with CT and pathology. J Comput Assist Tomogr 13:1083–1086, 1989. 123. Choyke PL, Hayes WS, Sesterhenn IA: Primary extragonadal germ cell tumors of the retroperitoneum: Differentiation of primary and secondary tumors. Radiographics 13:1365–1375, quiz 1377–8, 1993. 124. Chronowski GM, Ha CS, Wilder RB, et al: Treatment of unicentric and multicentric Castleman disease and the role of radiotherapy. Cancer 92:670–676, 2001. 125. Chua JH, Chui CH, Sai Prasad TR, et al: Fetus-in-fetu in the pelvis: Report of a case and literature review. Ann Acad Med Singapore 34:646–649, 2005. 126. Chun HJ, Byun JY, Jung SE, et al: Benign solitary fibrous tumour of the pre-sacral space: MRI findings. Br J Radiol 71:677–679, 1998. 127. Chung J, Kim M, Lee JT, et al: Cavernous hemangioma arising from the lesser omentum: MR findings. Abdom Imaging 25:542–544, 2000. 128. Chung JH, Suh YL, Park IA, et al: A pathologic study of abdominal lymphangiomas. J Korean Med Sci 14:257–262, 1999. 129. Ciftci E, Erden I, Koral K, et al: MR imaging in a desmoid tumor of the posterior mediastinum with extension into the abdominal cavity. A case report. Acta Radiol 39:301–303, 1998. 130. Clark SK, Neale KF, Landgrebe JC, et al: Desmoid tumours complicating familial adenomatous polyposis. Br J Surg 86:1185–1189, 1999. 131. Clark RA, Watanabe AT, Bradley WG, Jr, et al: Acute hematomas: Effects of deoxygenation, hematocrit, and fibrin-clot formation and retraction on T2 shortening. Radiology 175:201–206, 1990. 132. Clarke MR, Weyant RJ, Watson CG, et al: Prognostic markers in pheochromocytoma. Hum Pathol 29:522–526, 1998.

CHAPTER 55  Retroperitoneum 133. Coakley FV, Hricak H: Imaging of peritoneal and mesenteric disease: Key concepts for the clinical radiologist. Clin Radiol 54:563–574, 1999. 134. Coelho RF, Schneider-Monteiro ED, Mesquita JL, et al: Renal and perinephric abscesses: Analysis of 65 consecutive cases. World J Surg 31:431–436, 2007. 135. Coffin CM, Watterson J, Priest JR, et al: Extrapulmonary inflammatory myofibroblastic tumor (inflammatory pseudotumor). A clinicopathologic and immunohistochemical study of 84 cases. Am J Surg Pathol 19:859–872, 1995. 136. Cohen MD, Weetman RM, Provisor AJ, et al: Efficacy of magnetic resonance imaging in 139 children with tumors. Arch Surg 121:522–529, 1986. 137. Colovic R, Perisic-Savic M: Retroperitoneal biloma secondary to operative common bile duct injury. HPB Surg 3:193–197, 1991. 138. Comiter CV, Renshaw AA, Benson CB, et al: Burned-out primary testicular cancer: Sonographic and pathological characteristics. J Urol 156:85–88, 1996. 139. Coolen J, Bradshaw B, Bhargava R, et al: Fetus-in-fetu: Confirmation of prenatal diagnosis with MRI. Prenat Diagn 27:73–76, 2007. 140. Coppi G, Rametta F, Aiello S, et al: Inflammatory abdominal aortic aneurysm endovascular repair into the long-term follow-up. Ann Vasc Surg 24:1053–1059, 2010. 141. Corbett HJ, Humphrey GM: Pulmonary sequestration. Paediatr Respir Rev 5:59–68, 2004. 142. Cormier JN, Pollock RE: Soft tissue sarcomas. CA Cancer J Clin 54:94–109, 2004. 143. Corrin B, Liebow AA, Friedman PJ: Pulmonary lymphangiomyomatosis. A review. Am J Pathol 79:348–382, 1975. 144. Couanet D, Geoffray A, Hartmann O, et al: Bone marrow metastases in children’s neuroblastoma studied by magnetic resonance imaging. Prog Clin Biol Res 271:547–555, 1988. 145. Coumbaras M, Chopier J, Massiani MA, et al: Diffuse mesenteric and omental infiltration by amyloidosis with omental calcification mimicking abdominal carcinomatosis. Clin Radiol 56:674–676, 2001. 146. Crim JR, Seeger LL, Yao L, et al: Diagnosis of soft-tissue masses with MR imaging: Can benign masses be differentiated from malignant ones? Radiology 185:581–586, 1992. 147. Cuevas C, Raske M, Bush WH, et al: Imaging primary and secondary tumor thrombus of the inferior vena cava: Multi-detector computed tomography and magnetic resonance imaging. Curr Probl Diagn Radiol 35:90–101, 2006. 148. Cyran KM, Kenney PJ: Leiomyosarcoma of abdominal veins: Value of MRI with gadolinium DTPA. Abdom Imaging. 19:335–338, 1994. 149. Dachman AH, Ros PR, Shekitka KM, et al: Colorectal hemangioma: Radiologic findings. Radiology 167:31–34, 1988. 150. Dahan H, Arrive L, Wendum D, et al: Retrorectal developmental cysts in adults: Clinical and radiologic-histopathologic review, differential diagnosis, and treatment. Radiographics 21:575–584, 2001. 151. Daly KP, Ho CP, Persson DL, et al: Traumatic retroperitoneal injuries: Review of multidetector CT findings. Radiographics 28:1571–1590, 2008. 152. Dan D, Bahadursingh S, Hariharan S, et al: Extra-adrenal perirenal myelolipoma. A case report and review of literature. G Chir 33:62–65, 2012. 153. Danaci M, Kesici GE, Kesici H, et al: Coumadin-induced renal and retroperitoneal hemorrhage. Ren Fail 28:129–132, 2006. 154. Daniels IR, Sullivan T, Hale J: Retroperitoneal gas after colonoscopy.   J R Soc Med 92:21–22, 1999. 155. Davidson AJ, Hartman DS: Lymphangioma of the retroperitoneum: CT and sonographic characteristic. Radiology 175:507–510, 1990. 156. Davidson AJ, Hartman DS, Goldman SM: Mature teratoma of the retroperitoneum: Radiologic, pathologic, and clinical correlation. Radiology 172:421–425, 1989. 157. Deans GT, Spence RA: Neoplastic lesions of the appendix. Br J Surg 82:299–306, 1995. 158. De Backer AI, Mortele KJ, Van Den Heuvel E, et al: Tuberculous adenitis: Comparison of CT and MRI findings with histopathological features. Eur Radiol 17:1111–1117, 2007.

1923

159. De Filippo M, Ingegnoli A, Carloni A, et al: Erdheim-Chester disease: Clinical and radiological findings. Radiol Med 114:1319–1329, 2009. 160. De Gaetano AM, Vecchioli A, Minordi LM, et al: Role of diagnostic imaging in abdominal lymphadenopathy. Rays 25:463–484, 2000. 161. DeHart MM, Lauerman WC, Conely AH, et al: Management of retroperitoneal chylous leakage. Spine 19:716–718, 1994. 162. de Jong E, Mulder W, Nooitgedacht E, et al: Carney’s triad. Eur J Surg Oncol 24:147–149, 1998. 163. de Lange EE, Black WC, Mills SE: Radiologic features of retroperitoneal cystic hamartoma. Gastrointest Radiol 13:266–270, 1988. 164. del Olmo JA, Esteban JM, Maldonado L, et al: Clinical significance of abdominal lymphadenopathy in chronic liver disease. Ultrasound Med Biol 28:297–301, 2002. 165. Demas BE, Avallone A, Hricak H: Pelvic lipomatosis: Diagnosis and characterization by magnetic resonance imaging. Urol Radiol 10:198– 202, 1988. 166. Demirpolat G, Pourbagher A, Hekimgil M, et al: Mesenteric Castleman’s disease: Case report. Abdom Imaging 25:551–553, 2000. 167. de Perrot M, Brundler M, Totsch M, et al: Mesenteric cysts. Toward less confusion? Dig Surg 17:323–328, 2000. 168. de Perrot M, Rostan O, Morel P, et al: Abdominal lymphangioma in adults and children. Br J Surg 85:395–397, 1998. 169. de Sa J, Pimentel J, Carvalho M, et al: Spinal cord compression secondary to idiopathic retroperitoneal fibrosis. Neurosurgery 26:678–681, 1990. 170. Destian S, Ritchie WG: Aggressive angiomyxoma: CT appearance.   Am J Gastroenterol 81:711–713, 1986. 171. Deutch SJ, Sandler MA, Alpern MB: Abdominal lymphadenopathy in benign diseases: CT detection. Radiology 163:335–338, 1987. 172. Diel J, Ortiz O, Losada RA, et al: The sacrum: Pathologic spectrum, multimodality imaging, and subspecialty approach. Radiographics 21:83–104, 2001. 173. Dion E, Graef C, Haroche J, et al: Imaging of thoracoabdominal involvement in Erdheim-Chester disease. AJR Am J Roentgenol 183:1253–1260, 2004. 174. Dion E, Graef C, Miquel A, et al: Bone involvement in Erdheim-Chester disease: Imaging findings including periostitis and partial epiphyseal involvement. Radiology 238:632–639, 2006. 175. Diop M, Parratte B, Tatu L, et al: “Mesorectum”: The surgical value of an anatomical approach. Surg Radiol Anat 25:290–304, 2003. 176. Dodd GD, Rutledge F, Wallace S: Postoperative pelvic lymphocysts. Am J Roentgenol Radium Ther Nucl Med 108:312–323, 1970. 177. Dodds WJ, Darweesh RM, Lawson TL, et al: The retroperitoneal spaces revisited. AJR Am J Roentgenol 147:1155–1161, 1986. 178. Dong PR, Seeger LL, Yao L, et al: Uncomplicated cat-scratch disease: Findings at CT, MR imaging, and radiography. Radiology 195:837–839, 1995. 179. Dooms GC, Hricak H, Moseley ME, et al: Characterization of lymphadenopathy by magnetic resonance relaxation times: Preliminary results. Radiology 155:691–697, 1985. 180. Dorfman RE, Alpern MB, Gross BH, et al: Upper abdominal lymph nodes: Criteria for normal size determined with CT. Radiology 180:319–322, 1991. 181. Drop A, Czekajska-Chehab E, Maciejewski R: Giant retroperitoneal lipomas—Radiological case report. Ann Univ Mariae Curie Sklodowska [Med] 58:142–146, 2003. 182. Durie BG: The role of anatomic and functional staging in myeloma: Description of Durie/Salmon plus staging system. Eur J Cancer 42:1539–1543, 2006. 183. Durie BG, Salmon SE: A clinical staging system for multiple myeloma. Correlation of measured myeloma cell mass with presenting clinical features, response to treatment, and survival. Cancer 36:842–854, 1975. 184. Durie BG, Waxman AD, D’Agnolo A, et al: Whole-body (18)F-FDG PET identifies high-risk myeloma. J Nucl Med 43:1457–1463, 2002. 185. Dutta HK, Thomas JK, Sahewala NK, et al: Fetus in fetu in a neonate: Report of a case. Surg Today 43:547–549, 2013.

1924

PART II  CT and MR Imaging of the Whole Body

186. Eary JF, Conrad EU, Bruckner JD, et al: Quantitative [F-18] fluorodeoxyglucose positron emission tomography in pretreatment and grading of sarcoma. Clin Cancer Res 4:1215–1220, 1998. 187. Eble JN: Angiomyolipoma of kidney. Semin Diagn Pathol 15:21–40, 1998. 188. Eble JN, Rosenberg AE, Young RH: Retroperitoneal xanthogranuloma in a patient with Erdheim-Chester disease. Am J Surg Pathol 18:843– 848, 1994. 189. Edge SB, Byrd DR, Compton CC, et al: AJCC Cancer Staging   Manual, editor: Soft tissue sarcoma, ed 7, New York, 2010, Springer, pp 291–298. 190. Edge SB, Byrd DR, Compton CC, et al, editors: Hodgkin and nonHodgkin lymphomas, AJCC Cancer Staging Manual, ed 7, New York, 2010, Springer, pp 607–612. 191. Einstein DM, Singer AA, Chilcote WA, et al: Abdominal lymphadenopathy: Spectrum of CT findings. Radiographics 11:457–472, 1991. 192. el Naggar AK, Ro JY, Ayala AG, et al: Localized fibrous tumor of the serosal cavities. Immunohistochemical, electron-microscopic, and flow-cytometric DNA study. Am J Clin Pathol 92:561–565, 1989. 193. Elsayes KM, Staveteig PT, Narra VR, et al: Retroperitoneal masses: Magnetic resonance imaging findings with pathologic correlation. Curr Probl Diagn Radiol 36:97–106, 2007. 194. El Shallaly G, Raimes S: Cervical emphysema, pneumomediastinum, and retropneumoperitoneum following sigmoidoscopy and rectal biopsy. Int J Colorectal Dis 20:547–548, 2005. 195. Elstrom R, Guan L, Baker G, et al: Utility of FDG-PET scanning in lymphoma by WHO classification. Blood 101:3875–3876, 2003. [Epub 2003 Jan 16]. 196. Engel IA, Auh YH, Rubenstein WA, et al: Large posterior abdominal masses: Computed tomographic localization. Radiology 149:203–209, 1983. 197. Engel RM, Elkins RC, Fletcher BD: Retroperitoneal teratoma. Review of the literature and presentation of an unusual case. Cancer 22:1068– 1073, 1968. 198. Engelken JD, Ros PR: Retroperitoneal MR imaging. Magn Reson Imaging Clin N Am 5:165–178, 1997. 199. Engin G, Acunas B, Acunas G, et al: Imaging of extrapulmonary tuberculosis. Radiographics 20:471–488, quiz 529–30, 532, 2000. 200. English R: Cat-scratch disease. Pediatr Rev 27:123–128, quiz 128, 2006. 201. Epstein BM, Mann JH: CT of abdominal tuberculosis. AJR Am J Roentgenol 139:861–866, 1982. 202. Ercoli A, Delmas V, Fanfani F, et al: Terminologia Anatomica versus unofficial descriptions and nomenclature of the fasciae and ligaments of the female pelvis: A dissection-based comparative study. Am J Obstet Gynecol 193:1565–1573, 2005. 203. Esiashvili N, Goodman M, Ward K, et al: Neuroblastoma in adults: Incidence and survival analysis based on SEER data. Pediatr Blood Cancer 49:41–46, 2007. 204. Esquivel J, Sugarbaker PH: Clinical presentation of the pseudomyxoma peritonei syndrome. Br J Surg 87:1414–1418, 2000. 205. Favia G, Lumachi F, Polistina F, et al: Pheochromocytoma, a rare cause of hypertension: Long-term follow-up of 55 surgically treated patients. World J Surg 22:689–693, discussion 694, 1998. 206. Federle MP, Jeffrey RB, Jr: Hemoperitoneum studied by computed tomography. Radiology 148:187–192, 1983. 207. Federle MP, Pan KT, Pealer KM: CT criteria for differentiating abdominal hemorrhage: Anticoagulation or aortic aneurysm rupture? AJR Am J Roentgenol 188:1324–1330, 2007. 208. Feldberg MA, Koehler PR, van Waes PF: Psoas compartment disease studied by computed tomography. Analysis of 50 cases and subject review. Radiology 148:505–512, 1983. 209. Felker RE, Tonkin IL: Imaging of pulmonary sequestration. AJR Am J Roentgenol 154:241–249, 1990. 210. Ferrero A, Cespedes M, Cantarero JM, et al: Peritonitis due to rupture of retroperitoneal teratoma: Computed tomography diagnosis. Gastrointest Radiol 15:251–252, 1990.

211. Figler TJ, Keshavarzian A, Nand S, et al: Retroperitoneal amyloidosis, factor IX and X deficiency, and gastrointestinal bleeding. Abdom Imaging 21:266–268, 1996. 212. Fishman EK, Kuhlman JE, Jones RJ: CT of lymphoma: Spectrum of disease. Radiographics 11:647–669, 1991. 213. Fishman EK, Ney DR: Advanced computer applications in radiology: Clinical applications. Radiographics 13:463–475, 1993. 214. Fitoz S, Atasoy C, Yagmurlu A, et al: Psoas abscess secondary to tuberculous lymphadenopathy: Case report. Abdom Imaging. 26:323– 324, 2001. 215. Fletcher CDM, Bridge JA, Hogendoorn PCW, et al: WHO Classification of Tumours of Soft Tissue and Bone, ed 4, Lyon, France, 2013, World Health Organization, pp 1–427. 216. Fortman BJ, Beall DP: Erdheim-Chester disease of the retroperitoneum: A rare cause of ureteral obstruction. AJR Am J Roentgenol 176:1330– 1331, 2001. 217. Fortman BJ, Kuszyk BS, Urban BA, et al: Neurofibromatosis type 1: A diagnostic mimicker at CT. Radiographics 21:601–612, 2001. 218. Foucar E, Rosai J, Dorfman R: Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): Review of the entity. Semin Diagn Pathol 7:19–73, 1990. 219. Francis IR, Cohan RH, Varma DG, et al: Retroperitoneal sarcomas. Cancer Imaging 5:89–94, 2005. 220. Freeman HJ: Mesenteric lymph node cavitation syndrome. World J Gastroenterol 16:2991–2993, 2010. 221. Friedman HD, Hadfield TL, Lamy Y, et al: Whipple’s disease presenting as chronic wastage and abdominal lymphadenopathy. Diagn Microbiol Infect Dis 23:111–113, 1995. 222. Friedman AC, Hartman DS, Sherman J, et al: Computed tomography of abdominal fatty masses. Radiology 139:415–429, 1981. 223. Friis J, Hjortrup A: Extrarenal angiomyolipoma: Diagnosis and management. J Urol 127:528–529, 1982. 224. Fritz J, Vogel W, Bares R, et al: Radiologic spectrum of extramedullary relapse of myelogenous leukemia in adults. AJR Am J Roentgenol 189:209–218, 2007. 225. Fry AC, Singh S, Gunda SS, et al: Successful use of steroids and ureteric stents in 24 patients with idiopathic retroperitoneal fibrosis: A retrospective study. Nephron Clin Pract 108:c213–c220, 2008. 226. Fugita OE, Jarrett TW, Kavoussi P, et al: Laparoscopic treatment of retroperitoneal fibrosis. J Endourol 16:571–574, 2002. 227. Fujita N, Nishie A, Asayama Y, et al: A male case of primary retroperitoneal mucinous cystadenoma: A diagnostic dilemma. Jpn J Radiol 30:594–597, 2012. 228. Fujiwara K, Oda T, Suzuki S, et al: Primary serous adenocarcinoma of the retroperitoneum with a response to platinum-based chemotherapy: A case report. Int J Gynecol Cancer 9:170–172, 1999. 229. Fukunaga M, Naganuma H, Ushigome S, et al: Malignant solitary fibrous tumour of the peritoneum. Histopathology 28:463–466, 1996. 230. Furlong MA, Fanburg-Smith JC, Miettinen M: The morphologic spectrum of hibernoma: A clinicopathologic study of 170 cases. Am J Surg Pathol 25:809–814, 2001. 231. Furukawa H, Hara T, Taniguchi T: Two cases of retained foreign bodies after cholecystectomy: Diagnosis by sonography, CT, angiography, and MRI. Jpn J Surg 21:566–570, 1991. 232. Furuno T, Morita K, Kakizaki H, et al: Laparoscopic removal of a retroperitoneal extralobar pulmonary sequestration in an adult. Int J Urol 13:165–167, 2006. 233. Garg PK, Jaswal V, Jain BK: Retroperitoneal benign mature teratoma. Acta Med Indones 43:206–207, 2011. 234. Gass MA, Gould HR, Roggensack GF, et al: Lymphatic changes in retroperitoneal lipomatosis. J Comput Tomogr 7:229–232, 1983. 235. Gatanaga H, Ohnishi S, Miura H, et al: Retroperitoneal fibrosis leading to extrahepatic portal vein obstruction. Intern Med 33:346–350, 1994. 236. Gatcombe HG, Assikis V, Kooby D, et al: Primary retroperitoneal teratomas: A review of the literature. J Surg Oncol 86:107–113, 2004. 237. Gawande AA, Studdert DM, Orav EJ, et al: Risk factors for retained instruments and sponges after surgery. N Engl J Med 348:229–235, 2003.

CHAPTER 55  Retroperitoneum 238. Gayer G, Zissin R, Apter S, et al: Urinomas caused by ureteral injuries: CT appearance. Abdom Imaging 27:88–92, 2002. 239. Gazelle GS, Mueller PR: Abdominal abscess. Imaging and intervention. Radiol Clin North Am 32:913–932, 1994. 240. Geelhoed GW: CAT scans and catecholamines. Surgery 87:719–720, 1980. 241. Geoerger B, Hero B, Harms D, et al: Metabolic activity and clinical features of primary ganglioneuromas. Cancer 91:1905–1913, 2001. 242. Geoghegan T, Byrne AT, Benfayed W, et al: Imaging and intervention of retroperitoneal fibrosis. Australas Radiol 51:26–34, 2007. 243. George RE, London WB, Cohn SL, et al: Hyperdiploidy plus nonamplified MYCN confers a favorable prognosis in children 12 to 18 months old with disseminated neuroblastoma: A Pediatric Oncology Group study. J Clin Oncol 23:6466–6473, 2005. 244. Georgiades CS, Neyman EG, Barish MA, et al: Amyloidosis: Review and CT manifestations. Radiographics 24:405–416, 2004. 245. Gerle RD, Jaretzki A, 3rd, Ashley CA, et al: Congenital bronchopulmonary-foregut malformation. Pulmonary sequestration communicating with the gastrointestinal tract. N Engl J Med 278:1413– 1419, 1968. 246. Gerson ES, Gerzof SG, Robbins AH: CT confirmation of pelvic lipomatosis: Two cases. AJR Am J Roentgenol 129:338–340, 1977. 247. Gertz MA, Lacy MQ, Dispenzieri A, et al: Amyloidosis: Diagnosis and management. Clin Lymphoma Myeloma 6:208–219, 2005. 248. Gibbs MR, Pemberton LB, Litton NJ, et al: Lipolymph nodes of the mesentery. Am Surg 51:596–598, 1985. 249. Gilkeson GS, Allen NB: Retroperitoneal fibrosis. A true connective tissue disease. Rheum Dis Clin North Am 22:23–38, 1996. 250. Gionanlis LC, Bamichas GI, Smias CV, et al: Retroperitoneal plasmacytoma causing renal obstruction. Int Urol Nephrol 40:555–556, 2008. 251. Gitelis S, Heligman D, Morton T: The treatment of pigmented villonodular synovitis of the hip. A case report and literature review. Clin Orthop Relat Res 154–160, 1989. 252. Glodny B, Winde G, Herwig R, et al: Clinical differences between benign and malignant pheochromocytomas. Endocr J 48:151–159, 2001. 253. Goel A, Tiwari P, Sharma PK, et al: Extragenitourinary retroperitoneal primary hydatid cyst: A rare cause of bilateral lower ureteric obstruction and unilateral limb edema. Saudi J Kidney Dis Transpl 24:991–994, 2013. 254. Goh BK, Tan YM, Chung YF, et al: Retroperitoneal schwannoma. Am J Surg 192:14–18, 2006. 255. Gold JS, Antonescu CR, Hajdu C, et al: Clinicopathologic correlates of solitary fibrous tumors. Cancer 94:1057–1068, 2002. 256. Goldblum JR, Folpe AL, Weiss SW, editors: Enzinger and Weiss’s Soft Tissue Tumors, ed 6, Philadelphia, 2014, Elsevier Saunders. 257. Goldstein DS, Eisenhofer G, Flynn JA, et al: Diagnosis and localization of pheochromocytoma. Hypertension 43:907–910, 2004. 258. Gomori JM, Grossman RI, Hackney DB, et al: Variable appearances of subacute intracranial hematomas on high-field spin-echo MR. AJR Am J Roentgenol 150:171–178, 1988. 259. Goodlad JR, Fletcher CD: Solitary fibrous tumour arising at unusual sites: Analysis of a series. Histopathology 19:515–522, 1991. 260. Goodwin WE, Fonkalsrud EW, Goldman R, et al: Diagnostic problems in retroperitoneal disease. Ann Intern Med 65:160–184, 1966. 261. Gordon NS, Sinclair RA, Snow RM: Pelvic lipomatosis with cystitis cystica, cystitis glandularis and adenocarcinoma of the bladder: first reported case. Aust N Z J Surg 60:229–232, 1990. 262. Gore RM, Balfe DM, Aizenstein RI, et al: The great escape: Interfascial decompression planes of the retroperitoneum. AJR Am J Roentgenol 175:363–370, 2000. 263. Goshima S, Kanematsu M, Nishibori H, et al: Multi-detector row CT of the kidney: Optimizing scan delays for bolus tracking techniques of arterial, corticomedullary, and nephrographic phases. Eur J Radiol 2007. 264. Grassi R, Pinto A, Rossi G: Isolated pneumoretroperitoneum secondary to acute bowel infarction. Clin Radiol 55:321–323, 2000.

1925

265. Green JM, Bruner BC, Tang WW, et al: Retroperitoneal mucinous cystadenocarcinoma in a man: Case report and review of the literature. Urol Oncol 25:53–55, 2007. 266. Grignon DJ, Shkrum MJ, Smout MS: Extra-adrenal myelolipoma. Arch Pathol Lab Med 113:52–54, 1989. 267. Gu J, Chan T, Zhang J, et al: Whole-body diffusion-weighted imaging: The added value to whole-body MRI at initial diagnosis of lymphoma. AJR Am J Roentgenol 197:W384–W391, 2011. 268. Guller U, Turek J, Eubanks S, et al: Detecting pheochromocytoma: Defining the most sensitive test. Ann Surg 243:102–107, 2006. 269. Gupta AK, Cohan RH, Francis IR, et al: CT of recurrent retroperitoneal sarcomas. AJR Am J Roentgenol 174:1025–1030, 2000. 270. Gustafson P, Herrlin K, Biling L, et al: Necrosis observed on CT enhancement is of prognostic value in soft tissue sarcoma. Acta Radiol 33:474–476, 1992. 271. Gutierrez JC, Fischer AC, Sola JE, et al: Markedly improving survival of neuroblastoma: A 30-year analysis of 1,646 patients. Pediatr Surg Int 23:637–646, 2007. 272. Gutierrez JC, Perez EA, Franceschi D, et al: Outcomes for soft-tissue sarcoma in 8249 cases from a large state cancer registry. J Surg Res 141:105–114, 2007. 273. Gutierrez Delgado F, Tjulandin SA, Garin AM: Long term results of treatment in patients with extragonadal germ cell tumours. Eur J Cancer 29A:1002–1005, 1993. 274. Ha HK, Jung JI, Lee MS, et al: CT differentiation of tuberculous peritonitis and peritoneal carcinomatosis. AJR Am J Roentgenol 167:743–748, 1996. 275. Haaga JR, Reich NE, Havrilla TR, et al: Computed tomography-guided biopsy. I. Overview. J Comput Tomogr 2:25–30, 1978. 276. Hahn PF, Saini S, Stark DD, et al: Intraabdominal hematoma: The concentric-ring sign in MR imaging. AJR Am J Roentgenol 148:115–119, 1987. 277. Hallett JW, Jr: Management of abdominal aortic aneurysms. Mayo Clin Proc 75:395–399, 2000. 278. Hamberg LM, Rhea JT, Hunter GJ, et al: Multi-detector row CT: Radiation dose characteristics. Radiology 226:762–772, 2003. 279. Hanaoka M, Hashimoto M, Sasaki K, et al: Retroperitoneal   cavernous hemangioma resected by a pylorus preserving pancreaticoduodenectomy. World J Gastroenterol 19:4624–4629, 2013. 280. Hanquinet S, Damry N, Heimann P, et al: Association of a fetus in fetu and two teratomas: US and MRI. Pediatr Radiol 27:336–338, 1997. 281. Harisinghani MG, Barentsz J, Hahn PF, et al: Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 348:2491–2499, 2003. 282. Harisinghani MG, McLoud TC, Shepard JA, et al: Tuberculosis from head to toe. Radiographics 20:449–470, quiz 528–9, 532, 2000. 283. Harisinghani MG, Saksena MA, Hahn PF, et al: Ferumoxtran-10enhanced MR lymphangiography: Does contrast-enhanced imaging alone suffice for accurate lymph node characterization? AJR Am J Roentgenol 186:144–148, 2006. 284. Harisinghani MG, Saksena M, Ross RW, et al: A pilot study of lymphotrophic nanoparticle-enhanced magnetic resonance imaging technique in early stage testicular cancer: A new method for noninvasive lymph node evaluation. Urology 66:1066–1071, 2005. 285. Harris NL, Jaffe ES, Diebold J, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the Clinical Advisory Committee meeting-Airlie House, Virginia, November 1997. J Clin Oncol 17:3835–3849, 1999. 286. Harris NL, Jaffe ES, Stein H, et al: A revised European-American classification of lymphoid neoplasms: A proposal from the International Lymphoma Study Group. Blood 84:1361–1392, 1994. 287. Hartman DS, Hayes WS, Choyke PL, et al: From the archives of the AFIP. Leiomyosarcoma of the retroperitoneum and inferior vena cava: Radiologic-pathologic correlation. Radiographics 12:1203–1220, 1992. 288. Hasan AK, Sinclair DJ: Case report: Calcification in benign cystic peritoneal mesothelioma. Clin Radiol 48:66–67, 1993.

1926

PART II  CT and MR Imaging of the Whole Body

289. Hasegawa T, Matsuno Y, Shimoda T, et al: Extrathoracic solitary fibrous tumors: Their histological variability and potentially aggressive behavior. Hum Pathol 30:1464–1473, 1999. 290. Hatten MT, Hamrick-Turner JE, Smith DB: Mesenteric cystic lymphangioma: Radiologic appearance mimicking cystic teratoma. Pediatr Radiol 26:458–460, 1996. 291. Hayasaka K, Tanaka Y, Soeda S, et al: MR findings in primary retroperitoneal schwannoma. Acta Radiol 40:78–82, 1999. 292. Hayasaka K, Yamada T, Saitoh Y, et al: CT evaluation of primary benign retroperitoneal tumor. Radiat Med 12:115–120, 1994. 293. Hayashi T, Mine M, Kojima S, et al: Extragonadal germ cell tumor followed by metachronous testicular tumor. A case report. Urol Int 57:194–196, 1996. 294. Hayes WS, Davidson AJ, Grimley PM, et al: Extraadrenal retroperitoneal paraganglioma: Clinical, pathologic, and CT findings. AJR Am J Roentgenol 155:1247–1250, 1990. 295. Haynes JW, Brewer WH, Walsh JW: Focal fat necrosis presenting as a palpable abdominal mass: CT evaluation. J Comput Assist Tomogr 9:568–569, 1985. 296. Healy JC, Reznek RH: The peritoneum, mesenteries and omenta: Normal anatomy and pathological processes. Eur Radiol 8:886–900, 1998. 297. Healy JC, Reznek RH, Clark SK, et al: MR appearances of desmoid tumors in familial adenomatous polyposis. AJR Am J Roentgenol 169:465–472, 1997. 298. Heithoff KB, Sane SM, Williams HJ, et al: Bronchopulmonary foregut malformations. A unifying etiological concept. AJR Am J Roentgenol 126:46–55, 1976. 299. Helenon O, Merran S, Paraf F, et al: Unusual fat-containing   tumors of the kidney: A diagnostic dilemma. Radiographics 17:129–144, 1997. 300. Hellmann DB, Grand DJ, Freischlag JA: Inflammatory abdominal aortic aneurysm. JAMA 297:395–400, 2007. 301. Henricks WH, Chu YC, Goldblum JR, et al: Dedifferentiated liposarcoma: A clinicopathological analysis of 155 cases with a proposal for an expanded definition of dedifferentiation. Am J Surg Pathol 21:271–281, 1997. 302. Herman K, Kusy T: Retroperitoneal sarcoma—The continued challenge for surgery and oncology. Surg Oncol 7:77–81, 1998. 303. Hernanz-Schulman M, Johnson JE, Holcomb GW, 3rd, et al: Retroperitoneal pulmonary sequestration: Imaging findings, histopathologic correlation, and relationship to cystic adenomatoid malformation. AJR Am J Roentgenol 168:1277–1281, 1997. 304. Heyns CF, De Kock ML, Kirsten PH, et al: Pelvic lipomatosis associated with cystitis glandularis and adenocarcinoma of the bladder. J Urol 145:364–366, 1991. 305. Hibbeln JF, Wehmueller MD, Wilbur AC: Chylous ascites: CT and ultrasound appearance. Abdom Imaging. 20:138–140, 1995. 306. Hillman KM: Pneumoretroperitoneum. Anaesthesia 38:136–139, 1983. 307. Hinchliffe RJ, Macierewicz JA, Hopkinson BR: Endovascular repair of inflammatory abdominal aortic aneurysms. J Endovasc Ther 9:277–281, 2002. 308. Hines OJ, Nelson S, Quinones-Baldrich WJ, et al: Leiomyosarcoma of the inferior vena cava: Prognosis and comparison with leiomyosarcoma of other anatomic sites. Cancer 85:1077–1083, 1999. 309. Hinson FL, Ambrose NS: Pseudomyxoma peritonei. Br J Surg 85:1332–1339, 1998. 310. Hisatomi E, Miyajima K, Yasumori K, et al: Retroperitoneal bronchogenic cyst: A rare case showing the characteristic imaging feature of milk of calcium. Abdom Imaging 28:716–720, 2003. 311. Hjermstad BM, Helwig EB: Tailgut cysts. Report of 53 cases. Am J Clin Pathol 89:139–147, 1988. 312. Hoane BR, Shields AF, Porter BA, et al: Comparison of initial lymphoma staging using computed tomography (CT) and magnetic resonance (MR) imaging. Am J Hematol 47:100–105, 1994. 313. Hoeffel CC, Nguyen KQ, Phan HT, et al: Fetus in fetu: A case report and literature review. Pediatrics 105:1335–1344, 2000.

314. Hoegerle S, Altehoefer C, Ghanem N, et al: Whole-body 18F dopa PET for detection of gastrointestinal carcinoid tumors. Radiology 220:373– 380, 2001. 315. Holt PD, de Lang EE: Cat scratch disease: Magnetic resonance imaging findings. Skeletal Radiol 24:437–440, 1995. 316. Holton LH, Migaly J, Rolandelli RH: Pneumorrhachis, subcutaneous emphysema, pneumomediastinum, pneumopericardium, and pneumoretroperitoneum after proctocolectomy for ulcerative colitis: Report of a case. Dis Colon Rectum 45:567–570, 2002. 317. Honecke K, Butz M: Pelvic lipomatosis in a female: Diagnosis and initial therapy. Urol Int 46:93–95, 1991. 318. Hong SS, Goo HW, Jung MR, et al: Fetus in fetu: Three-dimensional imaging using multidetector CT. AJR Am J Roentgenol 179:1481– 1483, 2002. 319. Hong SH, Kim KA, Woo OH, et al: Dedifferentiated liposarcoma of retroperitoneum: Spectrum of imaging findings in 15 patients. Clin Imaging 34:203–210, 2010. 320. Hopkins KL, Simoneaux SF, Patrick LE, et al: Imaging manifestations of cat-scratch disease. AJR Am J Roentgenol 166:435–438, 1996. 321. Hovanessian LJ, Larsen DW, Raval JK, et al: Retroperitoneal cystic lymphangioma: MR findings. Magn Reson Imaging 8:91–93, 1990. 322. Howman-Giles R, Shaw PJ, Uren RF, et al: Neuroblastoma and other neuroendocrine tumors. Semin Nucl Med 37:286–302, 2007. 323. Hrehorovich PA, Franke HR, Maximin S, et al: Malignant peripheral nerve sheath tumor. Radiographics 23:790–794, 2003. 324. Hruban RH, Bhagavan BS, Epstein JI: Massive retroperitoneal angiomyolipoma. A lesion that may be confused with welldifferentiated liposarcoma. Am J Clin Pathol 92:805–808, 1989. 325. Hruby G, Lehman M, Barton M, et al: Malignant retroperitoneal paraganglioma: Case report and review of treatment options. Australas Radiol 44:478–482, 2000. 326. Hsiao YC, Chen CC, Shih TY, et al: [Fetus in fetu: report of a single case]. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 34:151–156, 1993. 327. Hsieh TJ, Chiang IC, Kuo KK, et al: Spontaneous retroperitoneal biloma: A case report. Kaohsiung J Med Sci 18:412–416, 2002. 328. Hsu SC, Huang JJ, Wang MC, et al: Gas-forming retroperitoneal abscess associated with crepitant myositis of right buttock and thigh. J Infect 40:295–297, 2000. 329. Hughes TH, Sartoris DJ, Schweitzer ME, et al: Pigmented villonodular synovitis: MRI characteristics. Skeletal Radiol 24:7–12, 1995. 330. Hughes MJ, Thomas JM, Fisher C, et al: Imaging features of retroperitoneal and pelvic schwannomas. Clin Radiol 60:886–893, 2005. 331. Hulnick DH, Megibow AJ, Naidich DP, et al: Abdominal tuberculosis: CT evaluation. Radiology 157:199–204, 1985. 332. Hunter SB, Schemankewitz EH, Patterson C, et al: Extraadrenal myelolipoma. A report of two cases. Am J Clin Pathol 97:402–404, 1992. 333. Hwang JK, Kim J, Lee BJ, et al: Pneumoretroperitoneum following acupuncture. J Altern Complement Med 14:1299–1301, 2008. 334. Ibukuro K, Oishi A, Tanaka R, et al: Signal flare phenomenon as active bleeding in retroperitoneal hematoma with hematocrit effect on dynamic CT scan: Three clinical cases and experimental study based on a specific gravity theory. J Comput Assist Tomogr 30:787–790, 2006. 335. Ichikawa T, Ohtomo K, Araki T, et al: Ganglioneuroma: Computed tomography and magnetic resonance features. Br J Radiol 69:114–121, 1996. 336. Ichikawa T, Ohtomo K, Uchiyama G, et al: Contrast-enhanced dynamic MRI of adrenal masses: Classification of characteristic enhancement patterns. Clin Radiol 50:295–300, 1995. 337. Ide C, De Coene B, Lawson G, et al: Castleman’s disease in the neck: MRI. Neuroradiology 39:520–522, 1997. 338. Idell S, Meltzer M, Knafelc E, et al: Primary systemic amyloidosis with a retroperitoneal mass. Clin Exp Rheumatol 2:181–183, 1984. 339. Idrees MT, Hoch BL, Wang BY, et al: Aggressive angiomyxoma of male genital region. Report of 4 cases with immunohistochemical evaluation including hormone receptor status. Ann Diagn Pathol 10:197–204, 2006. 340. Igarashi J, Hanazaki K: Retroperitoneal venous hemangioma. Am J Gastroenterol 93:2292–2293, 1998.

CHAPTER 55  Retroperitoneum 341. Iino M, Kuribayashi S, Imakita S, et al: Sensitivity and specificity of CT in the diagnosis of inflammatory abdominal aortic aneurysms.   J Comput Assist Tomogr 26:1006–1012, 2002. 342. Ikeda T, Miyauchi Y, Nishio S, et al: Primary retroperitoneal lymphocele in a pregnant woman: The first report. Int J Urol 13:445–446, 2006. 343. Ikezoe J, Murayama S, Godwin JD, et al: Bronchopulmonary sequestration: CT assessment. Radiology 176:375–379, 1990. 344. Ilhan H, Tokar B, Isiksoy S, et al: Giant mesenteric lipoma. J Pediatr Surg 34:639–640, 1999. 345. Ilias I, Shulkin B, Pacak K: New functional imaging modalities for chromaffin tumors, neuroblastomas and ganglioneuromas. Trends Endocrinol Metab 16:66–72, 2005. 346. Imrie KR, Kovacs MJ, Selby D, et al: Isolated chloroma: The effect of early antileukemic therapy. Ann Intern Med 123:351–353, 1995. 347. Inokuchi T, Takiuchi H, Moriwaki Y, et al: Retroperitoneal ancient schwannoma presenting as an adrenal incidentaloma: CT and MR findings. Magn Reson Imaging 24:1389–1393, 2006. 348. Inoue Y, Nakamura H, Yamazaki K, et al: Retroperitoneal Castleman’s tumors of hyaline vascular type: Imaging study. Case report. Clin Imaging 16:239–242, 1992. 349. Ishigami K, Khanna G, Samuel I, et al: Gas-forming abdominal wall abscess: Unusual manifestation of perforated retroperitoneal appendicitis extending through the superior lumbar triangle. Emerg Radiol 10:207–209, 2004. 350. Ishikawa K, Idoguchi K, Tanaka H, et al: Classification of acute pancreatitis based on retroperitoneal extension: Application of the concept of interfascial planes. Eur J Radiol 60:445–452, 2006. 351. Ishikawa K, Tohira H, Mizushima Y, et al: Traumatic retroperitoneal hematoma spreads through the interfascial planes. J Trauma 59:595– 607, discussion 607–8, 2005. 352. Israel GM, Bosniak MA, Slywotzky CM, et al: CT differentiation of large exophytic renal angiomyolipomas and perirenal liposarcomas. Am J Roentgenol 179:769–773, 2002. 353. Isse K, Harada K, Suzuki Y, et al: Retroperitoneal mucinous cystadenoma: Report of two cases and review of the literature. Pathol Int 54:132–138, 2004. 354. Iura A, Sasajima Y, Katsumata N, et al: Serous adenocarcinoma of the retroperitoneum, as a type of multifocal mullerian carcinoma. Int J Clin Oncol 14:254–257, 2009. 355. Jaffe HL, Lichtenstein L, Sutro CJ: Pigmented villonodular synovitis, bursitis and tenosynovitis. Arch Pathol 31:731–765, 1941. 356. Jang TN, Juang GD, Fung CP: Fulminating gas-forming psoas muscle abscess due to Klebsiella pneumoniae following a deep neck infection. J Formos Med Assoc 96:134–136, 1997. 357. Jansen I, Hendriksz TR, Han SH, et al: 18)F-fluorodeoxyglucose position emission tomography (FDG-PET) for monitoring disease activity and treatment response in idiopathic retroperitoneal fibrosis. Eur J Intern Med 21:216–221, 2010. 358. Jao SW, Beart RW, Jr, Spencer RJ, et al: Retrorectal tumors. Mayo Clinic experience, 1960-1979. Dis Colon Rectum 28:644–652, 1985. 359. Jeffrey RB, Jr, Cardoza JD, Olcott EW: Detection of active intraabdominal arterial hemorrhage: Value of dynamic contrastenhanced CT. AJR Am J Roentgenol 156:725–729, 1991. 360. Jeffrey RB, Jr, Nyberg DA, Bottles K, et al: Abdominal CT in acquired immunodeficiency syndrome. AJR Am J Roentgenol 146:7–13, 1986. 361. Jelinek JS, Kransdorf MJ, Shmookler BM, et al: Liposarcoma of the extremities: MR and CT findings in the histologic subtypes. Radiology 186:455–459, 1993. 362. Jennings TA, Peterson L, Axiotis CA, et al: Angiosarcoma associated with foreign body material. A report of three cases. Cancer 62:2436– 2444, 1988. 363. Jeyadevan NN, Sohaib SA, Thomas JM, et al: Imaging features of aggressive angiomyxoma. Clin Radiol 58:157–162, 2003. 364. Jhanwar YS, Straus DJ: The role of PET in lymphoma. J Nucl Med 47:1326–1334, 2006. 365. Jimenez JM, Poustchi-Amin M, Leonidas JC, et al: Extraperitoneal abdominopelvic inflammatory pseudotumor: Report of four cases. Pediatr Radiol 27:170–174, 1997.

1927

366. Johnson GL, Hruban RH, Marshall FF, et al: Primary adrenal ganglioneuroma: CT findings in four patients. AJR Am J Roentgenol 169:169–171, 1997. 367. Johnson J, Mattrey R, Phillipson J: Differentiation of seminomatous from nonseminomatous testicular tumors with MR imaging. Am J Roentgenol 154:539–543, 1990. 368. Johnson AR, Ros PR, Hjermstad BM: Tailgut cyst: Diagnosis with CT and sonography. AJR Am J Roentgenol 147:1309–1311, 1986. 369. Johnson WK, Ros PR, Powers C, et al: Castleman disease mimicking an aggressive retroperitoneal neoplasm. Abdom Imaging 19:342–344, 1994. 370. Johnson GR, Zhuang H, Khan J, et al: Roles of positron emission tomography with fluorine-18-deoxyglucose in the detection of local recurrent and distant metastatic sarcoma. Clin Nucl Med 28:815–820, 2003. 371. Joseph N, Vogelzang RL, Hidvegi D, et al: Computed tomography of retroperitoneal castleman disease (plasma cell type) with sonographic and angiographic correlation. J Comput Assist Tomogr 9:570–572, 1985. 372. Jung G, Heindel W, von Bergwelt-Baildon M, et al: Abdominal lymphoma staging: Is MR imaging with T2-weighted turbo-spin-echo sequence a diagnostic alternative to contrast-enhanced spiral CT?   J Comput Assist Tomogr 24:783–787, 2000. 373. Kakihara D, Yoshimitsu K, Eto M, et al: MRI of retroperitoneal solitary fibrous tumor in the suprarenal region. AJR Am J Roentgenol 188:W512–W514, 2007. 374. Kaku M, Ohara N, Seima Y, et al: A primary retroperitoneal serous cystadenocarcinoma with clinically aggressive behavior. Arch Gynecol Obstet 270:302–306, 2004. 375. Kalassian KG, Doyle R, Kao P, et al: Lymphangioleiomyomatosis: New insights. Am J Respir Crit Care Med 155:1183–1186, 1997. 376. Kallas KM, Vaughan L, Haghighi P, et al: Pigmented villonodular synovitis of the hip presenting as a retroperitoneal mass. Skeletal Radiol 30:469–474, 2001. 377. Kalovidouris A, Kehagias D, Moulopoulos L, et al: Abdominal retained surgical sponges: CT appearance. Eur Radiol 9:1407–1410, 1999. 378. Kammen BF, Elder DE, Fraker DL, et al: Extraadrenal myelolipoma: MR imaging findings. AJR Am J Roentgenol 171:721–723, 1998. 379. Kapadia SB: Multiple myeloma: A clinicopathologic study of 62 consecutively autopsied cases. Medicine (Baltimore) 59:380–392, 1980. 380. Karam M, Novak L, Cyriac J, et al: Role of fluorine-18 fluorodeoxyglucose positron emission tomography scan in the evaluation and follow-up of patients with low-grade lymphomas. Cancer 107:175–183, 2006. 381. Karcaaltincaba M, Akhan O: Radiologic imaging and percutaneous treatment of pelvic lymphocele. Eur J Radiol 55:340–354, 2005. 382. Kardar AH, Kattan S, Lindstedt E, et al: Steroid therapy for idiopathic retroperitoneal fibrosis: Dose and duration. J Urol 168:550–555, 2002. 383. Katsube Y, Mukai K, Silverberg SG: Cystic mesothelioma of the peritoneum: A report of five cases and review of the literature. Cancer 50:1615–1622, 1982. 384. Kaur A, Makhija PS, Vallikad E, et al: Multifocal aggressive angiomyxoma: A case report. J Clin Pathol 53:798–799, 2000. 385. Kawashima A, Goldman SM, Fishman EK, et al: CT of intraabdominal desmoid tumors: Is the tumor different in patients with Gardner’s disease? AJR Am J Roentgenol 162:339–342, 1994. 386. Kelly WM, Paglen PG, Pearson JA, et al: Ferromagnetism of intraocular foreign body causes unilateral blindness after MR study. AJNR Am J Neuroradiol 7:243–245, 1986. 387. Kenn W, Stabler A, Zachoval R, et al: Erdheim-Chester disease: A case report and literature overview. Eur Radiol 9:153–158, 1999. 388. Kenney PJ, Wagner BJ, Rao P, et al: Myelolipoma: CT and pathologic features. Radiology 208:87–95, 1998. 389. Kermani TA, Crowson CS, Achenbach SJ, et al: Idiopathic retroperitoneal fibrosis: a retrospective review of clinical presentation, treatment, and outcomes. Mayo Clin Proc 86:297–303, 2011. 390. Khan J, von Sinner W, Akhtar M, et al: Castleman’s disease of the chest. Magnetic resonance imaging features. Chest 105:1608–1610, 1994. 391. Khayyata S, Bentley G, Fregene TA, et al: Retroperitoneal extramedullary anaplastic plasmacytoma masquerading as sarcoma:

1928

392.

393. 394.

395. 396. 397.

398.

399.

400. 401. 402. 403. 404. 405. 406.

407. 408. 409.

410. 411.

412.

413.

414. 415.

416.

417. 418.

PART II  CT and MR Imaging of the Whole Body

Report of a case with an unusual presentation and imprint smears. Acta Cytol 51:434–436, 2007. Khosroshahi A, Carruthers MN, Stone JH, et al: Rethinking Ormond’s disease: “Idiopathic” retroperitoneal fibrosis in the era of IgG4-related disease. Medicine (Baltimore) 92:82–91, 2013. Kikkawa A, Kido A, Kumai T, et al: Extraabdominal fibromatosis in retroperitoneal space. World J Surg Oncol 2:33, 2004. Killingsworth C, Gadacz TR: Tailgut cyst (retrorectal cystic hamartoma): Report of a case and review of the literature. Am Surg 71:666–673, 2005. Kilton LJ, Aschenbrener C, Burns CP: Ganglioneuroblastoma in adults. Cancer 37:974–983, 1976. Kim SH, Choi BI, Han MC, et al: Retroperitoneal neurilemoma: CT and MR findings. AJR Am J Roentgenol 159:1023–1026, 1992. Kim TJ, Han JK, Kim YH, et al: Castleman disease of the abdomen: Imaging spectrum and clinicopathologic correlations. J Comput Assist Tomogr 25:207–214, 2001. Kim JK, Jeong YY, Kim YH, et al: Postoperative pelvic lymphocele: Treatment with simple percutaneous catheter drainage. Radiology 212:390–394, 1999. Kim SY, Kim MJ, Chung JJ, et al: Abdominal tuberculous lymphadenopathy: MR imaging findings. Abdom Imaging 25:627–632, 2000. Kim MJ, Kim WH, Kim NK, et al: Tailgut cyst: Multilocular cystic appearance on MRI. J Comput Assist Tomogr 21:731–732, 1997. Kim EY, Lee WJ, Jang KT: Retroperitoneal bronchogenic cyst mimicking a pancreatic cystic tumour. Clin Radiol 62:491–494, 2007. Kim T, Murakami T, Oi H, et al: CT and MR imaging of abdominal liposarcoma. AJR Am J Roentgenol 166:829–833, 1996. Kim JY, Park YH, Choi KH, et al: MRI of tuberculous pyomyositis.   J Comput Assist Tomogr 23:454–457, 1999. Kim CK, Park BK, Ha H: Gossypiboma in abdomen and pelvis: MRI findings in four patients. AJR Am J Roentgenol 189:814–817, 2007. Kimmel DL, Moyer EK, et al: A cerebral tumor containing five human fetuses; a case of fetus in fetu. Anat Rec 106:141–165, 1950. King AA, Debaun MR, Riccardi VM, et al: Malignant peripheral nerve sheath tumors in neurofibromatosis 1. Am J Med Genet 93:388–392, 2000. Kinoshita T, Ishii K, Higashiiwai H, et al: Malignant solitary fibrous tumour of the peritoneum. Clin Radiol 55:157–160, 2000. Kinoshita T, Naganuma H, Ishii K, et al: CT features of retroperitoneal neurilemmoma. Eur J Radiol 27:67–71, 1998. Kitsanou M, Andreopoulou E, Bai MK, et al: Extensive lymphadenopathy as the first clinical manifestation in systemic lupus erythematosus. Lupus 9:140–143, 2000. Klein FA, Smith MJ, Kasenetz I: Pelvic lipomatosis: 35-year experience. J Urol 139:998–1001, 1988. Klucznik RP, Carrier DA, Pyka R, et al: Placement of a ferromagnetic intracerebral aneurysm clip in a magnetic field with a fatal outcome. Radiology 187:855–856, 1993. Kneeland JB, Auh YH, Rubenstein WA, et al: Perirenal spaces: CT evidence for communication across the midline. Radiology 164:657– 664, 1987. Ko SF, Ng SH, Shieh CS, et al: Mesenteric cystic lymphangioma with myxoid degeneration: Unusual CT and MR manifestations. Pediatr Radiol 25:525–527, 1995. Ko SF, Wan YL, Lee TY, et al: CT features of calcifications in abdominal malignant fibrous histiocytoma. Clin Imaging 22:408–413, 1998. Kobayashi H, Itoh T, Murata R, et al: Primary retroperitoneal plasmacytoma with tumor thrombus within the renal vein. J Urol 147:452–454, 1992. Kobayashi H, Kotoura Y, Hosono M, et al: MRI and scintigraphic features of extraabdominal desmoid tumors. Clin Imaging 21:35–39, 1997. Koc Z, Ezer A: Migrating and herniating hydatid cysts. Eur J Radiol 65:120–124, 2008. Koirala R, Shakya VC, Agrawal CS, et al: Retroperitoneal inflammatory myofibroblastic tumor. Am J Surg 199:e17–e19, 2010.

419. Kokubo T, Itai Y, Ohtomo K, et al: Retained surgical sponges: CT and US appearance. Radiology 165:415–418, 1987. 420. Konishi E, Nakashima Y, Iwasaki T: Immunohistochemical analysis of retroperitoneal Mullerian cyst. Hum Pathol 34:194–198, 2003. 421. Koo PJ, Wills JS: Case 146: Benign multicystic mesothelioma. Radiology 251:944–946, 2009. 422. Kopecky KK, Bodnar A, Morphis JG, et al: Subdiaphragmatic pulmonary sequestrian simulating metastatic testicular cancer. Clin Radiol 55:794–796, 2000. 423. Korf BR: Plexiform neurofibromas. Am J Med Genet 89:31–37, 1999. 424. Korf BR: Malignancy in neurofibromatosis type 1. Oncologist 5:477–485, 2000. 425. Korobkin M, Silverman PM, Quint LE, et al: CT of the extraperitoneal space: Normal anatomy and fluid collections. AJR Am J Roentgenol 159:933–942, 1992. 426. Koroku M, Takagi Y, Suzuki N, et al: Primary retroperitoneal teratoma in an adult: A case report. Int J Urol 4:219–221, 1997. 427. Kosowsky JM, Gibler WB: Massive hemoperitoneum due to rupture of a retroperitoneal varix. J Emerg Med 19:347–349, 2000. 428. Kottra JJ, Dunnick NR: Retroperitoneal fibrosis. Radiol Clin North Am 34:1259–1275, 1996. 429. Kransdorf MJ: Malignant soft-tissue tumors in a large referral population: Distribution of diagnoses by age, sex, and location. AJR Am J Roentgenol 164:129–134, 1995. 430. Kransdorf MJ, Bancroft LW, Peterson JJ, et al: Imaging of fatty tumors: Distinction of lipoma and well-differentiated liposarcoma. Radiology 224:99–104, 2002. 431. Kransdorf MJ, Jelinek JS, Moser RP, Jr, et al: Soft-tissue masses: Diagnosis using MR imaging. AJR Am J Roentgenol 153:541–547, 1989. 432. Kransdorf MJ, Meis JM, Jelinek JS: Dedifferentiated liposarcoma of the extremities: Imaging findings in four patients. AJR Am J Roentgenol 161:127–130, 1993. 433. Kremer M, Ott G, Nathrath M, et al: Primary extramedullary plasmacytoma and multiple myeloma: Phenotypic differences revealed by immunohistochemical analysis. J Pathol 205:92–101, 2005. 434. Kreuzberg B, Koudelova J, Ferda J, et al: Diagnostic problems of abdominal desmoid tumors in various locations. Eur J Radiol 62:180–185, 2007. 435. Krishnan J, Chu WS, Elrod JP, et al: Tumoral presentation of amyloidosis (amyloidomas) in soft tissues. A report of 14 cases. Am J Clin Pathol 100:135–144, 1993. 436. Kubota Y, Kawai N, Tozawa K, et al: Solitary fibrous tumor of the peritoneum found in the prevesical space. Urol Int 65:53–56, 2000. 437. Kubota K, Yamada K, Yoshioka S, et al: Differential diagnosis of idiopathic fibrosis from malignant lymphadenopathy with PET and F-18 fluorodeoxyglucose. Clin Nucl Med 17:361–363, 1992. 438. Kucukardali Y, Solmazgul E, Kunter E, et al: Kikuchi-Fujimoto Disease: Analysis of 244 cases. Clin Rheumatol 26:50–54, 2007. 439. Kulaylat MN, Karakousis CP, Keaney CM, et al: Desmoid tumour: A pleomorphic lesion. Eur J Surg Oncol 25:487–497, 1999. 440. Kulshrestha M, Misson G: Magnetic resonance imaging and the dangers of orbital foreign bodies. Br J Ophthalmol 79:1149, 1995. 441. Kumar R, Maillard I, Schuster SJ, et al: Utility of fluorodeoxyglucosePET imaging in the management of patients with Hodgkin’s and non-Hodgkin’s lymphomas. Radiol Clin North Am 42:1083–1100, 2004. 442. Kunin M: Bridging septa of the perinephric space: Anatomic, pathologic, and diagnostic considerations. Radiology 158:361–365, 1986. 443. Kurosaki Y, Kuramoto K: Case report: Serous cystadenocarcinoma of the retroperitoneum: CT and sonographic appearance. Clin Radiol 53:916–918, 1998. 444. Kurosaki Y, Tanaka YO, Itai Y: Well-differentiated liposarcoma of the retroperitoneum with a fat-fluid level: US, CT, and MR appearance. Eur Radiol 8:474–475, 1998. 445. Kuwashima S, Yamato M, Fujioka M, et al: MR findings of surgically retained sponges and towels: Report of two cases. Radiat Med 11:98–101, 1993.

CHAPTER 55  Retroperitoneum 446. Kwee TC, Basu S, Saboury B, et al: Functional oncoimaging techniques with potential clinical applications. Front Biosci (Elite Ed) 4:1081–1096, 2012. 447. Kwee TC, Torigian DA, Alavi A: Overview of positron emission tomography, hybrid positron emission tomography instrumentation, and positron emission tomography quantification. J Thorac Imaging 28:4–10, 2013. 448. Kwee TC, Torigian DA, Alavi A: Oncological applications of positron emission tomography for evaluation of the thorax. J Thorac Imaging 28:11–24, 2013. 449. Kwee TC, Torigian DA, Alavi A: Nononcological applications of positron emission tomography for evaluation of the thorax. J Thorac Imaging 28:25–39, 2013. 450. Kwon SY, Kim TK, Kim YS, et al: CT findings in Kikuchi disease: Analysis of 96 cases. AJNR Am J Neuroradiol 25:1099–1102, 2004. 451. La Fianza A, Alberici E, Meloni G, et al: Extraperitoneal pelvic leiomyosarcoma. MR findings in a case. Clin Imaging 24:224–226, 2000. 452. Lai FM, Allen PW, Chan LW, et al: Aggressive fibromatosis of the spermatic cord. A typical lesion in a “new” location. Am J Clin Pathol 104:403–407, 1995. 453. Lamps LW, Scott MA: Cat-scratch disease: historic, clinical, and pathologic perspectives. Am J Clin Pathol 121(Suppl):S71–S80, 2004. 454. Lane RH, Stephens DH, Reiman HM: Primary retroperitoneal neoplasms: CT findings in 90 cases with clinical and pathologic correlation. AJR Am J Roentgenol 152:83–89, 1989. 455. Lang EK, Glorioso L, 3rd: Management of urinomas by percutaneous drainage procedures. Radiol Clin North Am 24:551–559, 1986. 456. Larsen CE, Patrick LE: Abdominal (liver, spleen) and bone manifestations of cat scratch disease. Pediatr Radiol 22:353–355, 1992. 457. Laterre PF, Dugernier T, Reynaert MS: Chylous ascites: Diagnosis, causes and treatment. Acta Gastroenterol Belg 63:260–263, 2000. 458. Lauwers PR, Van Hee RH: Intraperitoneal gossypibomas: The need to count sponges. World J Surg 24:521–527, 2000. 459. Lee KC, Cassar-Pullicino VN: Giant cisterna chyli: MRI depiction with gadolinium-DTPA enhancement. Clin Radiol 55:51–55, 2000. 460. Lee JK, Glazer HS: Controversy in the MR imaging appearance of fibrosis. Radiology 177:21–22, 1990. 461. Lee JK, Heiken JP, Ling D, et al: Magnetic resonance imaging of abdominal and pelvic lymphadenopathy. Radiology 153:181–188, 1984. 462. Lee YC, Huang SP, Liu CC, et al: Giant extrarenal retroperitoneal angiomyolipoma: A case report and literature review. Kaohsiung J Med Sci 19:579–582, 2003. 463. Lee SL, Ku YM, Rha SE: Comprehensive reviews of the interfascial plane of the retroperitoneum: Normal anatomy and pathologic entities. Emerg Radiol 17:3–11, 2010. 464. Lee KY, Oh YW, Noh HJ, et al: Extraadrenal paragangliomas of the body: Imaging features. AJR Am J Roentgenol 187:492–504, 2006. 465. Lee HJ, Park HS, Kim YJ, et al: Retroperitoneal lymphangioleiomyomatosis: Sonography, computed tomography, magnetic resonance imaging, and positron emission tomography with pathologic correlation. J Ultrasound Med 29:1837–1841, 2010. 466. Lee J, Song SY, Park CS, et al: Mullerian cysts of the mesentery and retroperitoneum: A case report and literature review. Pathol Int 48:902–906, 1998. 467. Lee JC, Thomas JM, Phillips S, et al: Aggressive fibromatosis: MRI features with pathologic correlation. AJR Am J Roentgenol 186:247–254, 2006. 468. Lee S, Yoo JH, Lee SW: Kikuchi disease: differentiation from tuberculous lymphadenitis based on patterns of nodal necrosis on CT. AJNR Am J Neuroradiol 33:135–140, 2012. 469. Leibovitch I, Mor Y, Golomb J, et al: Chylous ascites after radical nephrectomy and inferior vena cava thrombectomy. Successful conservative management with somatostatin analogue. Eur Urol 41:220–222, 2002. 470. Leiva BJ, Bofill RM, Sevin BU, et al: Extranodal Rosai-Dorfman disease as a pelvic mass. Int J Gynecol Cancer 16(Suppl 1):312–315, 2006. 471. Lemaitre L, Robert Y, Dubrulle F, et al: Renal angiomyolipoma: Growth followed up with CT and/or US. Radiology 197:598–602, 1995.

1929

472. Le Neel JC, De Cussac JB, Dupas B, et al: [Textiloma. Apropos of 25 cases and review of the literature]. Chirurgie 120:272–276, discussion 276–7, 1994. 473. Leroy K, Dumas V, Martin-Garcia N, et al: Malignant peripheral   nerve sheath tumors associated with neurofibromatosis type 1: A clinicopathologic and molecular study of 17 patients. Arch Dermatol 137:908–913, 2001. 474. Leseche G, Schaetz A, Arrive L, et al: Diagnosis and management of 17 consecutive patients with inflammatory abdominal aortic aneurysm. Am J Surg 164:39–44, 1992. 475. Levenson RB, Singh AK, Novelline RA: Fournier gangrene: Role of imaging. Radiographics 28:519–528, 2008. 476. Levi AD: Treatment of a retroperitoneal lymphocele after lumbar fusion surgery with intralesional povidone iodine: Technical case report. Neurosurgery 45:658–660, discussion 660–1, 1999. 477. Levy AD, Rimola J, Mehrotra AK, et al: From the archives of the AFIP: Benign fibrous tumors and tumorlike lesions of the mesentery: radiologic-pathologic correlation. Radiographics 26:245–264, 2006. 478. Lewis JJ, Leung D, Woodruff JM, et al: Retroperitoneal soft-tissue sarcoma: Analysis of 500 patients treated and followed at a single institution. Ann Surg 228:355–365, 1998. 479. Li Q, Gao C, Juzi JT, et al: Analysis of 82 cases of retroperitoneal schwannoma. ANZ J Surg 77:237–240, 2007. 480. Li YP, Guico R, Parikh S, et al: Cystic mesothelioma of the retroperitoneum. J Clin Ultrasound 20:65–68, 1992. 481. Li DK, Rennie CS: Abdominal computed tomography in Whipple’s disease. J Comput Assist Tomogr 5:249–252, 1981. 482. Li KP, Zhu J, Zhang JL, et al: Idiopathic retroperitoneal fibrosis (RPF): Clinical features of 61 cases and literature review. Clin Rheumatol 30:601–605, 2011. 483. Liang MK, Yee HT, Song JW, et al: Subdiaphragmatic bronchogenic cysts: A comprehensive review of the literature. Am Surg 71:1034–1041, 2005. 484. Liessi G, Cesari S, Pavanello M, et al: Tailgut cysts: CT and MR findings. Abdom Imaging 20:256–258, 1995. 485. Lim JH, Auh YH: Anatomic relation between the bare area of the liver and the right perirenal space. AJR Am J Roentgenol 162:733–734, 1994. 486. Lim JH, Auh YH, Suh SJ, et al: Right perirenal space. Computed tomography evidence of communication between the bare area of the liver. Clin Imaging 14:239–244, discussion 259, 1990. 487. Lim KE, Hsu WC, Wang CR: Tailgut cyst with malignancy: MR imaging findings. AJR Am J Roentgenol 170:1488–1490, 1998. 488. Lim JH, Kim B, Auh YH: Continuation of gas from the right perirenal space into the bare area of the liver. J Comput Assist Tomogr 21:667– 670, 1997. 489. Lin J, Jacobson JA, Jamadar DA, et al: Pigmented villonodular synovitis and related lesions: The spectrum of imaging findings. AJR Am J Roentgenol 172:191–197, 1999. 490. Lin LH, Ko SF, Huang CC, et al: Retroperitoneal ectopic pancreas: Imaging findings. Br J Radiol 82:e253–e255, 2009. 491. Lin J, Martel W: Cross-sectional imaging of peripheral nerve sheath tumors: Characteristic signs on CT, MR imaging, and sonography. AJR Am J Roentgenol 176:75–82, 2001. 492. Lindahl S, Nyman RS, Brismar J, et al: Imaging of tuberculosis. IV. Spinal manifestations in 63 patients. Acta Radiol 37:506–511, 1996. 493. Liu QY, Lin XF, Zhang WD, et al: Retroperitoneal schwannomas in the anterior pararenal space: Dynamic enhanced multi-slice CT and MR findings. Abdom Imaging 38:201–210, 2013. 494. Lo YS, Wang JS, Yu CC, et al: Retroperitoneal enteric duplication cyst.   J Chin Med Assoc 67:479–482, 2004. 495. Loke TK, Yuen NW, Lo KK, et al: Retroperitoneal ancient schwannoma: Review of clinico-radiological features. Australas Radiol 42:136–138, 1998. 496. Lombardo DM, Baker ME, Spritzer CE, et al: Hepatic hemangiomas vs metastases: MR differentiation at 1.5 T. AJR Am J Roentgenol 155:55–59, 1990. 497. Lonergan GJ, Schwab CM, Suarez ES, et al: Neuroblastoma, ganglioneuroblastoma, and ganglioneuroma: Radiologic-pathologic correlation. Radiographics 22:911–934, 2002.

1930

PART II  CT and MR Imaging of the Whole Body

498. Long JA, Jr, Doppman JL, Nienhuis AW: Computed tomographic studies of thoracic extramedullary hematopoiesis. J Comput Assist Tomogr 4:67–70, 1980. 499. Loor G, Bassiouny H, Valentin C, et al: Local and systemic consequences of large retroperitoneal clot burdens. World J Surg 33:1618–1625, 2009. 500. Lopez JA, Saez F, Alejandro Larena J, et al: MRI diagnosis and follow-up of subcutaneous fat necrosis. J Magn Reson Imaging 7:929–932, 1997. 501. Losanoff JE, Sauter ER, Rider KD: Cat scratch disease presenting with abdominal pain and retroperitoneal lymphadenopathy. J Clin Gastroenterol 38:300–301, 2004. 502. Low RN, Semelka RC, Worawattanakul S, et al: Extrahepatic abdominal imaging in patients with malignancy: Comparison of MR imaging and helical CT, with subsequent surgical correlation. Radiology 210:625–632, 1999. 503. Low RN, Semelka RC, Worawattanakul S, et al: Extrahepatic abdominal imaging in patients with malignancy: Comparison of MR imaging and helical CT in 164 patients. J Magn Reson Imaging 12:269–277, 2000. 504. Low RN, Sigeti JS: MR imaging of peritoneal disease: Comparison of contrast-enhanced fast multiplanar spoiled gradient-recalled and spin-echo imaging. AJR Am J Roentgenol 163:1131–1140, 1994. 505. Lu P: Staging and classification of lymphoma. Semin Nucl Med 35:160–164, 2005. 506. Lucas K, Gula MJ, Knisely AS, et al: Catecholamine metabolites in ganglioneuroma. Med Pediatr Oncol 22:240–243, 1994. 507. Lucas JD, O’Doherty MJ, Cronin BF, et al: Prospective evaluation of soft tissue masses and sarcomas using fluorodeoxyglucose positron emission tomography. Br J Surg 86:550–556, 1999. 508. Luciani A, Itti E, Rahmouni A, et al: Lymph node imaging: Basic principles. Eur J Radiol 58:338–344, 2006. 509. Lynch HT, Fitzgibbons R: Surgery, desmoid tumors, and familial adenomatous polyposis: Case report and literature review. Am J Gastroenterol 91:2598–2601, 1996. 510. Mack P: Pneumo-retroperitoneum associated with pneumatosis cystoides intestinals. Ann Acad Med Singapore 17:600–602, 1988. 511. Magliulo G, Zardo F, Varacalli S, et al: Multiple paragangliomas of the head and neck. An Otorrinolaringol Ibero Am 30:31–38, 2003. 512. Magnusson A: Size of normal retroperitoneal lymph nodes. Acta Radiol Diagn (Stockh) 24:315–318, 1983. 513. Magrina JF, Symmonds RE, Dahlin DC: Pelvic “lipolymph nodes”: A consideration in the differential diagnosis of pelvic masses. Am J Obstet Gynecol 136:727–731, 1980. 514. Magtibay PM, Salmon Z, Keeney GL, et al: Aggressive angiomyxoma of the female pelvis and perineum: A case series. Int J Gynecol Cancer 16:396–401, 2006. 515. Mahon DE, Carp NZ, Goldhahn RT, Jr, et al: Primary leiomyosarcoma of the greater omentum: Case report and review of the literature. Am Surg 59:160–163, 1993. 516. Makela J, Kiviniemi H, Laitinen S: Prognostic factors predicting survival in the treatment of retroperitoneal sarcoma. Eur J Surg Oncol 26:552–555, 2000. 517. Mallick S, Samantaray SP, Durgapal P, et al: Deep angiomyxoma of retroperitoneum presenting as a perineal lump in a young woman. BMJ Case Rep 2013:2013. 518. Mamoon N, Haroon A, Luqman M, et al: Kikuchi’s disease of lymph nodes. J Coll Physicians Surg Pak 13:138–142, 2003. 519. Manohar A, Simjee AE, Haffejee AA, et al: Symptoms and investigative findings in 145 patients with tuberculous peritonitis diagnosed by peritoneoscopy and biopsy over a five year period. Gut 31:1130–1132, 1990. 520. Marco V, Autonell J, Farre J, et al: Retrorectal cyst-hamartomas. Report of two cases with adenocarcinoma developing in one. Am J Surg Pathol 6:707–714, 1982. 521. Marcos HB, Noone TC, Semelka RC: MRI evaluation of acute renal trauma. J Magn Reson Imaging 8:989–990, 1998. 522. Margolis DJ, Bammer R, Chow LC: Parallel imaging of the abdomen. Top Magn Reson Imaging 15:197–206, 2004.

523. Mariani A, Nascimento AG, Webb MJ, et al: Surgical management of desmoid tumors of the female pelvis. J Am Coll Surg 191:175–183, 2000. 524. Maris JM: The biologic basis for neuroblastoma heterogeneity and risk stratification. Curr Opin Pediatr 17:7–13, 2005. 525. Marks SC, Jr, Raptopoulos V, Kleinman P, et al: The anatomical basis for retrorenal extensions of pancreatic effusions: The role of the renal fasciae. Surg Radiol Anat 8:89–97, 1986. 526. Marom EM, Helms CA: Fibrolipomatous hamartoma: Pathognomonic on MR imaging. Skeletal Radiol 28:260–264, 1999. 527. Marth T, Raoult D: Whipple’s disease. Lancet 361:239–246, 2003. 528. Martinez CA, Palma RT, Waisberg J: Giant retroperitoneal lipoma: A case report. Arq Gastroenterol 40:251–255, 2003. 529. Mastromatteo JF, Mindell HJ, Mastromatteo MF, et al: Communications of the pelvic extraperitoneal spaces and their relation to the abdominal extraperitoneal spaces: Helical CT cadaver study with pelvic extraperitoneal injections. Radiology 202:523–530, 1997. 530. Masumori N, Tsukamoto T: Pelvic lipomatosis associated with proliferative cystitis: Case report and review of the Japanese literature. Int J Urol 6:44–49, 1999. 531. Mathieu A, Chamlou R, Le Moine F, et al: Tailgut cyst associated with a carcinoid tumor: Case report and review of the literature. Histol Histopathol 20:1065–1069, 2005. 532. Mathis KL, Dozois EJ, Grewal MS, et al: Malignant risk and surgical outcomes of presacral tailgut cysts. Br J Surg 97:575–579, 2010. 533. Matsubayashi J, Ishida T, Ozawa T, et al: Subphrenic bronchopulmonary foregut malformation with pulmonarysequestration-like features. Pathol Int 53:313–316, 2003. 534. Matsuoka Y, Masumoto T, Suzuki K, et al: Pseudomyxoma retroperitonei. Eur Radiol 9:457–459, 1999. 535. Mattei J, Kim FJ, Phillips J, et al: Male primary retroperitoneal mucinous cystadenoma. Urology 82:e1–e2, 2013. 536. Matthay KK, Villablanca JG, Seeger RC, et al: Treatment of high-risk neuroblastoma with intensive chemotherapy, radiotherapy, autologous bone marrow transplantation, and 13-cis-retinoic acid. Children’s Cancer Group. N Engl J Med 341:1165–1173, 1999. 537. Matuchansky C, Colin R, Hemet J, et al: Cavitation of mesenteric lymph nodes, splenic atrophy, and a flat small intestinal mucosa. Report of six cases. Gastroenterology 87:606–614, 1984. 538. Maurea S, Cuocolo A, Reynolds JC, et al: [Role of magnetic resonance in the study of benign and malignant pheochromocytomas. Quantitative analysis of the intensity of the resonance signal]. Radiol Med 85:803–808, 1993. 539. Maurer CA, Wildi S, Muller MF, et al: Blunt abdominal trauma causing chyloretroperitoneum. J Trauma 43:696–697, 1997. 540. McAdams HP, Kirejczyk WM, Rosado-de-Christenson ML, et al: Bronchogenic cyst: Imaging features with clinical and histopathologic correlation. Radiology 217:441–446, 2000. 541. McCarty MJ, Vukelja SJ, Banks PM, et al: Angiofollicular lymph node hyperplasia (Castleman’s disease). Cancer Treat Rev 21:291–310, 1995. 542. McClellan DS, Brasch J, Rifkin H: Large retroperitoneal cysts in children and adolescents. J Urol 126:815–816, 1981. 543. McGinn CJ: The role of radiation therapy in resectable retroperitoneal sarcomas. Surg Oncol 9:61–65, 2000. 544. McLeod AJ, Zornoza J, Shirkhoda A: Leiomyosarcoma: Computed tomographic findings. Radiology 152:133–136, 1984. 545. Meador TL, McLarney JK: CT features of Castleman disease of the abdomen and pelvis. AJR Am J Roentgenol 175:115–118, 2000. 546. Meaglia JP, Schmidt JD: Natural history of an adrenal myelolipoma.   J Urol 147:1089–1090, 1992. 547. Mehta N, Tripathi RP, Popli MB, et al: Bilateral intraabdominal ganglioneuroblastoma in an adult. Br J Radiol 70:96–98, 1997. 548. Mesurolle B, Sayag E, Meingan P, et al: Retroperitoneal extramedullary hematopoiesis: Sonographic, CT, and MR imaging appearance. AJR Am J Roentgenol 167:1139–1140, 1996. 549. Meyer JS, Hoffer FA, Barnes PD, et al: Biological classification of soft-tissue vascular anomalies: MR correlation. AJR Am J Roentgenol 157:559–564, 1991.

CHAPTER 55  Retroperitoneum 550. Meyers MA: Acute extraperitoneal infection. Semin Roentgenol 8:445–464, 1973. 551. Meyers MA: Radiological features of the spread and localization of extraperitoneal gas and their relationship to its source. An anatomical approach. Radiology 111:17–26, 1974. 552. Meyers MA: Uriniferous perirenal pseudocyst: New observations. Radiology 117:539–545, 1975. 553. Meyers MA, Charnsangavej C, Oliphant M: Meyers’ Dynamic Radiology of the Abdomen: Normal and Pathologic Anatomy, ed 6, New York, NY, 2011, Springer, pp 1–419. 554. Meyers MA, Feldberg MA, Oliphant M: Grey Turner’s sign and Cullen’s sign in acute pancreatitis. Gastrointest Radiol 14:31–37, 1989. 555. Meyers MA, Oliphant M, Berne AS, et al: The peritoneal ligaments and mesenteries: Pathways of intraabdominal spread of disease. Radiology 163:593–604, 1987. 556. Meyers MA, Whalen JP, Peelle K, et al: Radiologic features of extraperitoneal effusions. An anatomic approach. Radiology 104:249– 257, 1972. 557. Middleton SB, Phillips RK: Surgery for large intra-abdominal desmoid tumors: Report of four cases. Dis Colon Rectum 43:1759–1762, discussion 1762–3, 2000. 558. Miele V, Andreoli C, De Cicco ML, et al: Hemoretroperitoneum associated with liver bare area injuries: CT evaluation. Eur Radiol 12:765–769, 2002. 559. Miettinen M, Lasota J: Gastrointestinal stromal tumors: Pathology and prognosis at different sites. Semin Diagn Pathol 23:70–83, 2006. 560. Miller FH, Keppke AL, Yaghmai V, et al: CT diagnosis of chyluria after partial nephrectomy. AJR Am J Roentgenol 188:W25–W28, 2007. 561. Miller EJ, Nowak E, Hair L, et al: Retroperitoneal accessory spleen. Am Surg 56:293–294, 1990. 562. Miller WT, Jr, Perez-Jaffe LA: Cross-sectional imaging of Kikuchi disease. J Comput Assist Tomogr 23:548–551, 1999. 563. Mills SE, Walker AN, Stallings RG, et al: Retrorectal cystic hamartoma. Report of three cases, including one with a perirenal component. Arch Pathol Lab Med 108:737–740, 1984. 564. Mindell HJ, Mastromatteo JF, Dickey KW, et al: Anatomic communications between the three retroperitoneal spaces: Determination by CT-guided injections of contrast material in cadavers. AJR Am J Roentgenol 164:1173–1178, 1995. 565. Mingoli A, Cavallaro A, Sapienza P, et al: International registry of inferior vena cava leiomyosarcoma: Analysis of a world series on 218 patients. Anticancer Res 16:3201–3205, 1996. 566. Mirault T, Lambert M, Puech P, et al: Malignant retroperitoneal fibrosis: MRI characteristics in 50 patients. Medicine (Baltimore) 91:242–250, 2012. 567. Miura S, Miura K, Yamamoto T, et al: Origin and mechanisms of formation of fetus-in-fetu: Two cases with genotype and methylation analyses. Am J Med Genet A 140:1737–1743, 2006. 568. Mizoe A, Takebe K, Kanematsu T: Primary leiomyosarcoma of the jejunal mesentery: Report of a case. Surg Today 28:87–90, 1998. 569. Mochizuki T, Takehara Y, Ichijo K, et al: Case report: MR appearance of a retained surgical sponge. Clin Radiol 46:66–67, 1992. 570. Mockli G: Granulocytic sarcoma presenting as a retroperitoneal lymph node mass. A case report with diagnosis by fine needle aspiration biopsy and flow cytometry. Acta Cytol 41:1525–1527, 1997. 571. Molmenti EP, Balfe DM, Kanterman RY, et al: Anatomy of the retroperitoneum: Observations of the distribution of pathologic fluid collections. Radiology 200:95–103, 1996. 572. Monill J, Pernas J, Montserrat E, et al: CT features of abdominal plasma cell neoplasms. Eur Radiol 15:1705–1712, 2005. 573. Moon WK, Im JG, Yu IK, et al: Mediastinal tuberculous lymphadenitis: MR imaging appearance with clinicopathologic correlation. AJR Am J Roentgenol 166:21–25, 1996. 574. Moormeier JA, Williams SF, Golomb HM: The staging of nonHodgkin’s lymphomas. Semin Oncol 17:43–50, 1990. 575. Mor Y, Leibovitch I, Garniek A, et al: Case report: Pseudomyxoma extraperitonei: An unusual presentation mimicking a huge hydronephrotic kidney. Clin Radiol 51:221–222, 1996.

1931

576. Moran CG, Morgan RH: Pseudomyxoma extraperitonei. J R Soc Med 81:668–669, 1988. 577. Morice P, Lassau N, Pautier P, et al: Retroperitoneal drainage after complete para-aortic lymphadenectomy for gynecologic cancer: A randomized trial. Obstet Gynecol 97:243–247, 2001. 578. Moroni G, Castellani M, Balzani A, et al: The value of (18)F-FDG PET/ CT in the assessment of active idiopathic retroperitoneal fibrosis. Eur J Nucl Med Mol Imaging 39:1635–1642, 2012. 579. Moss J, Avila NA, Barnes PM, et al: Prevalence and clinical characteristics of lymphangioleiomyomatosis (LAM) in patients with tuberous sclerosis complex. Am J Respir Crit Care Med 164:669–671, 2001. 580. Moulopoulos LA, Karvouni E, Kehagias D, et al: MR imaging of complex tail-gut cysts. Clin Radiol 54:118–122, 1999. 581. Mourra N, Caplin S, Parc R, et al: Presacral neuroendocrine carcinoma developed in a tailgut cyst: Report of a case. Dis Colon Rectum 46:411–413, 2003. 582. Moyle H, Hines OJ, McFadden DW: Gossypiboma of the abdomen. Arch Surg 131:566–568, 1996. 583. Mueller PR, Ferrucci JT, Jr, Harbin WP, et al: Appearance of lymphomatous involvement of the mesentery by ultrasonography and body computed tomography: The “sandwich sign”. Radiology 134:467– 473, 1980. 584. Mukherjee JJ, Peppercorn PD, Reznek RH, et al: Pheochromocytoma: Effect of nonionic contrast medium in CT on circulating catecholamine levels. Radiology 202:227–231, 1997. 585. Munechika H, Cohan RH, Dunnick NR: Non-seminomatous testicular tumors: Effect of lesion side on CT detection of lymph node metastasis. Comput Med Imaging Graph 12:343–348, 1988. 586. Munechika H, Honda M, Kushihashi T, et al: Computed tomography of retroperitoneal cystic lymphangiomas. J Comput Assist Tomogr 11:116–119, 1987. 587. Munk PL, Lee MJ, Janzen DL, et al: Lipoma and liposarcoma: Evaluation using CT and MR imaging. AJR Am J Roentgenol 169:589– 594, 1997. 588. Murakami AM, Anderson SW, Soto JA, et al: Active extravasation of the abdomen and pelvis in trauma using 64MDCT. Emerg Radiol 16:375–382, 2009. 589. Murakami R, Machida M, Kobayashi Y, et al: Retroperitoneal bronchogenic cyst: CT and MR imaging. Abdom Imaging 25:444–447, 2000. 590. Murakami R, Tajima H, Kobayashi Y, et al: Mesenteric schwannoma. Eur Radiol 8:277–279, 1998. 591. Murphey MD, Andrews CL, Flemming DJ, et al: From the archives of the AFIP. Primary tumors of the spine: Radiologic pathologic correlation. Radiographics 16:1131–1158, 1996. 592. Murphey MD, Carroll JF, Flemming DJ, et al: From the archives of the AFIP: Benign musculoskeletal lipomatous lesions. Radiographics 24:1433–1466, 2004. 593. Murphey MD, Smith WS, Smith SE, et al: From the archives of the AFIP. Imaging of musculoskeletal neurogenic tumors: Radiologicpathologic correlation. Radiographics 19:1253–1280, 1999. 594. Murphy DP, Glazier DB, Chenven ES, et al: Extrarenal retroperitoneal angiomyolipoma: Nonoperative management. J Urol 163:234–235, 2000. 595. Muscarella P, Cowgill S, Derenne LA, et al: Retroperitoneal benign cystic peritoneal mesothelioma. Surgery 135:228–231, 2004. 596. Mussack T, Wiedemann E, Szeimies U: Pneumoperitoneum, pneumoretroperitoneum, and pneumomediastinum caused by laryngeal fracture after multiple trauma. Am J Emerg Med 19:523–524, 2001. 597. Muttarak M, Peh WC: CT of unusual iliopsoas compartment lesions. Radiographics 20:S53–S66, 2000. 598. Myhre-Jensen O: A consecutive 7-year series of 1331 benign soft tissue tumours. Clinicopathologic data. Comparison with sarcomas. Acta Orthop Scand 52:287–293, 1981. 599. Na DG, Chung TS, Byun HS, et al: Kikuchi disease: CT and MR findings. AJNR Am J Neuroradiol 18:1729–1732, 1997.

1932

PART II  CT and MR Imaging of the Whole Body

600. Nakagawa M, Akasaka Y, Kanai T, et al: Clinicopathological and immunohistochemical features of extragastrointestinal stromal tumors: Report of two cases. Surg Today 35:336–340, 2005. 601. Nakamura T, Wakisaka M, Ohkubo H, et al: A case of solitary varices arising from the retroperitoneum. Urol Int 47:90–93, 1991. 602. Nakata H, Egashira K, Watanabe H, et al: MRI of bronchogenic cysts.   J Comput Assist Tomogr 17:267–270, 1993. 603. Narla LD, Newman B, Spottswood SS, et al: Inflammatory pseudotumor. Radiographics 23:719–729, 2003. 604. Neary P, Mathews R, Harries R, et al: Extrarenal retroperitoneal angiomyolipoma: Management of a rare benign tumour. Int J Colorectal Dis 18:526–528, 2003. 605. Negendank WG, al-Katib AM, Karanes C, et al: Lymphomas: MR imaging contrast characteristics with clinical-pathologic correlations. Radiology 177:209–216, 1990. 606. Negoro H, Oka H, Kawakita M: Two cases of chyloretroperitoneum following retroperitoneoscopic nephroureterectomy. Int J Urol 13:487–489, 2006. 607. Neiman RS, Barcos M, Berard C, et al: Granulocytic sarcoma: A clinicopathologic study of 61 biopsied cases. Cancer 48:1426–1437, 1981. 608. Neumann CH, Robert NJ, Canellos G, et al: Computed tomography of the abdomen and pelvis in non-Hodgkin lymphoma. J Comput Assist Tomogr 7:846–850, 1983. 609. Nevelsteen A, Lacroix H, Stockx L, et al: Inflammatory abdominal aortic aneurysm and bilateral complete ureteral obstruction: Treatment by endovascular graft and bilateral ureteric stenting. Ann Vasc Surg 13:222–224, 1999. 610. Neville A, Herts BR: CT characteristics of primary retroperitoneal neoplasms. Crit Rev Comput Tomogr 45:247–270, 2004. 611. Nguyen BD: Retroperitoneal extraadrenal myelolipoma: Technetium99m sulfur colloid scintigraphy and CT imaging. Clin Nucl Med 32:135–138, 2007. 612. Niccoli-Asabella A, Altini C, Notaristefano A, et al: A retrospective study comparing contrast-enhanced computed tomography with 18F-FDG-PET/CT in the early follow-up of patients with retroperitoneal sarcomas. Nucl Med Commun 34:32–39, 2013. 613. Nieweg OE, Pruim J, van Ginkel RJ, et al: Fluorine-18fluorodeoxyglucose PET imaging of soft-tissue sarcoma. J Nucl Med 37:257–261, 1996. 614. Nikken JJ, Krestin GP: MRI of the kidney-state of the art. Eur Radiol 17:2780–2793, 2007. 615. Nino-Murcia M, Tamm EP, Charnsangavej C, et al: Multidetector-row helical CT and advanced postprocessing techniques for the evaluation of pancreatic neoplasms. Abdom Imaging 28:366–377, 2003. 616. Nishimura H, Zhang Y, Ohkuma K, et al: MR imaging of soft-tissue masses of the extraperitoneal spaces. Radiographics 21:1141–1154, 2001. 617. Nishina M, Fujii C, Ogino R, et al: Pneumoperitoneum and pneumoretroperitoneum in blunt trauma patients. J Trauma 49:565– 566, 2000. 618. Nitecki SS, Hallett JW, Jr, Stanson AW, et al: Inflammatory abdominal aortic aneurysms: A case-control study. J Vasc Surg 23:860–868, discussion 868–9, 1996. 619. Nocera RM, Davis M, Hayden CK, Jr, et al: Fetus-in-fetu. AJR Am J Roentgenol 138:762–764, 1982. 620. Noel AA, Gloviczki P, Bender CE, et al: Treatment of symptomatic primary chylous disorders. J Vasc Surg 34:785–791, 2001. 621. Noone TC, Semelka RC, Worawattanakul S, et al: Intraperitoneal abscesses: Diagnostic accuracy of and appearances at MR imaging. Radiology 208:525–528, 1998. 622. Nunes LW, Schiebler MS, Rauschning W, et al: The normal prostate and periprostatic structures: Correlation between MR images made with an endorectal coil and cadaveric microtome sections. AJR Am J Roentgenol 164:923–927, 1995. 623. O’Connell AM, Duddy L, Lee C, et al: CT of pelvic extraperitoneal spaces: An anatomical study in cadavers. Clin Radiol 62:432–438, 2007.

624. O’Connor AR, Coakley FV, Meng MV, et al: Imaging of retained surgical sponges in the abdomen and pelvis. AJR Am J Roentgenol 180:481–489, 2003. 625. Ohguri T, Aoki T, Hisaoka M, et al: Differential diagnosis of benign peripheral lipoma from well-differentiated liposarcoma on MR imaging: Is comparison of margins and internal characteristics useful? AJR Am J Roentgenol 180:1689–1694, 2003. 626. Oikonomou A, Tzouvelekis A, Zezos P, et al: Pulmonary lymphangioleiomyomatosis associated with retroperitoneal lymphangioleiomyomas. BMJ Case Rep 2012:2012. 627. Oka K, Yakushiji T, Sato H, et al: Usefulness of diffusion-weighted imaging for differentiating between desmoid tumors and malignant soft tissue tumors. J Magn Reson Imaging 33:189–193, 2011. 628. Okuno S: The enigma of desmoid tumors. Curr Treat Options Oncol 7:438–443, 2006. 629. Oleaga L, Florencio MR, Ereno C, et al: Fibrolipomatous hamartoma of the radial nerve: MR imaging findings. Skeletal Radiol 24:559–561, 1995. 630. Oliphant M, Berne AS, Meyers MA: Spread of disease via the subperitoneal space: The small bowel mesentery. Abdom Imaging 18:109–116, 1993. 631. O’Neil JD, Ros PR, Storm BL, et al: Cystic mesothelioma of the peritoneum. Radiology 170:333–337, 1989. 632. Ooi GC, Chim CS, Khong PL, et al: Radiologic manifestations of granulocytic sarcoma in adult leukemia. AJR Am J Roentgenol 176:1427–1431, 2001. 633. Ooi BS, Lee CN, Ti TK, et al: Retroperitoneal fibromatosis presenting as acute duodenal obstruction. ANZ J Surg 71:74–76, 2001. 634. O’Riordain DS, Young WF, Jr, Grant CS, et al: Clinical spectrum and outcome of functional extraadrenal paraganglioma. World J Surg 20:916–921, discussion 922, 1996. 635. Orta Kilickesmez K, Kilickesmez O: Clear depiction of inflammatory abdominal aortic aneurysm with diffusion-weighted magnetic resonance imaging. Cardiovasc Intervent Radiol 33:379–382, 2010. 636. Orwig D, Federle MP: Localized clotted blood as evidence of visceral trauma on CT: The sentinel clot sign. AJR Am J Roentgenol 153:747– 749, 1989. 637. Oshima K, Kanda Y, Nannya Y, et al: Clinical and pathologic findings in 52 consecutively autopsied cases with multiple myeloma. Am J Hematol 67:1–5, 2001. 638. Otal P, Mezghani S, Hassissene S, et al: Imaging of retroperitoneal ganglioneuroma. Eur Radiol 11:940–945, 2001. 639. Ottery FD, Carlson RA, Gould H, et al: Retrorectal cyst-hamartomas: CT diagnosis. J Comput Assist Tomogr 10:260–263, 1986. 640. Outwater EK, Marchetto BE, Wagner BJ, et al: Aggressive angiomyxoma: Findings on CT and MR imaging. AJR Am J Roentgenol 172:435–438, 1999. 641. Ozgen A, Akata D, Akhan O, et al: Giant benign cystic peritoneal mesothelioma: US, CT, and MRI findings. Abdom Imaging 23:502–504, 1998. 642. Pacak K, Eisenhofer G, Carrasquillo JA, et al: Diagnostic localization of pheochromocytoma: The coming of age of positron emission tomography. Ann N Y Acad Sci 970:170–176, 2002. 643. Paley M, Sidhu PS, Evans RA, et al: Retroperitoneal collections— Aetiology and radiological implications. Clin Radiol 52:290–294, 1997. 644. Pallisa E, Sanz P, Roman A, et al: Lymphangioleiomyomatosis: Pulmonary and abdominal findings with pathologic correlation. Radiographics 22(Spec No):S185–S198, 2002. 645. Palmer BW, Norman RW: Retroperitoneal air, extracorporeal shock wave lithotripsy and epidural anesthesia. J Urol 141:844–845, 1989. 646. Panageas E: General diagnosis case of the day. Primary retroperitoneal teratoma. AJR Am J Roentgenol 156:1292–1294, 1991. 647. Papanicolaou N, Yoder IC, Lee MJ: Primary retroperitoneal neoplasms: How close can we come in making the correct diagnosis. Urol Radiol 14:221–228, 1992. 648. Paramesh AS, Meiners R, Fairchild RB, et al: Retroperitoneal varices presenting as an adrenal pseudotumor in a cirrhotic patient. WMJ 105:44–46, 2006.

CHAPTER 55  Retroperitoneum 649. Paravastu SC, Ghosh J, Murray D, et al: A systematic review of open versus endovascular repair of inflammatory abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 38:291–297, 2009. 650. Paravastu SC, Murray D, Ghosh J, et al: Inflammatory abdominal aortic aneurysms (IAAA): Past and present. Vasc Endovascular Surg 43:360– 363, 2009. 651. Parienty RA, Pradel J: Radiological evaluation of the peri- and pararenal spaces by computed tomography. Crit Rev Diagn Imaging 20:1–26, 1983. 652. Parienty RA, Pradel J, Lepreux JF, et al: Computed tomography of sponges retained after laparotomy. J Comput Assist Tomogr 5:187–189, 1981. 653. Parienty RA, Pradel J, Picard JD, et al: Visibility and thickening of the renal fascia on computed tomograms. Radiology 139:119–124, 1981. 654. Park SS, Min BW, Kim WB, et al: Malignant extragastrointestinal stromal tumor of retroperitoneum. Acta Oncol 44:497–499, 2005. 655. Patankar T, Fatterpekar GM, Prasad S, et al: Fetus in fetu: CT appearance—Report of two cases. Radiology 214:735–737, 2000. 656. Pedrosa I, Saiz A, Arrazola J, et al: Hydatid disease: Radiologic and pathologic features and complications. Radiographics 20:795–817, 2000. 657. Pennell TC, Gusdon JP, Jr: Retroperitoneal mucinous cystadenoma. Am J Obstet Gynecol 160:1229–1231, 1989. 658. Pennell RC, Hollier LH, Lie JT, et al: Inflammatory abdominal aortic aneurysms: A thirty-year review. J Vasc Surg 2:859–869, 1985. 659. Pereira JM, Sirlin CB, Pinto PS, et al: CT and MR imaging of extrahepatic fatty masses of the abdomen and pelvis: Techniques, diagnosis, differential diagnosis, and pitfalls. Radiographics 25:69–85, 2005. 660. Perera ND, Vithana VH: Giant retroperitoneal desmoid tumour. Ceylon Med J 51:124–125, 2006. 661. Periureteric fibrosis. Lancet 273:780–781, 1957. 662. Pestieau SR, Jelinek JS, Chang D, et al: CT in the selection of patients with abdominal or pelvic sarcoma for reoperative surgery. J Am Coll Surg 190:700–710, 2000. 663. Peters PE, Beyer K: [Normal lymph node cross sections in different anatomic regions and their significance for computed tomographic diagnosis]. Radiologe 25:193–198, 1985. 664. Petros JG, Thanikachalam M, Lopez MJ: Retroperitoneal and abdominal wall emphysema after transanal excision of a rectal carcinoma. Am Surg 62:759–761, 1996. 665. Petru E, Tamussino K, Lahousen M, et al: Pelvic and paraaortic lymphocysts after radical surgery because of cervical and ovarian cancer. Am J Obstet Gynecol 161:937–941, 1989. 666. Pickhardt PJ, Bhalla S: Primary neoplasms of peritoneal and subperitoneal origin: CT findings. Radiographics 25:983–995, 2005. 667. Pickhardt PJ, Bhalla S: Unusual nonneoplastic peritoneal and subperitoneal conditions: CT findings. Radiographics 25:719–730, 2005. 668. Pirayesh A, Chee Y, Helliwell TR, et al: The management of retroperitoneal soft tissue sarcoma: A single institution experience with a review of the literature. Eur J Surg Oncol 27:491–497, 2001. 669. Plouin PF, Gimenez-Roqueplo AP: Pheochromocytomas and secreting paragangliomas. Orphanet J Rare Dis 1:49, 2006. 670. Polat P, Kantarci M, Alper F, et al: Hydatid disease from head to toe. Radiographics 23:475–494, quiz 536–7, 2003. 671. Politoske EJ: Ruptured abdominal aortic aneurysm presenting as an obstruction of the left colon. Am J Gastroenterol 85:745–747, 1990. 672. Pollack HM: Imaging of the prostate gland. Eur Urol 20(Suppl 1): 50–58, 1991. 673. Pombo F, Rodriguez E, Mato J, et al: Patterns of contrast enhancement of tuberculous lymph nodes demonstrated by computed tomography. Clin Radiol 46:13–17, 1992. 674. Pommier RF, Vetto JT, Billingsly K, et al: Comparison of adrenal and extraadrenal pheochromocytomas. Surgery 114:1160–1165, discussion 1165–6, 1993. 675. Poritz LS, Blackstein M, Berk T, et al: Extended follow-up of patients treated with cytotoxic chemotherapy for intra-abdominal desmoid tumors. Dis Colon Rectum 44:1268–1273, 2001.

1933

676. Prabhuraj AR, Basu A, Sistla SC, et al: Primary retroperitoneal mucinous cystadenoma in a man. Am J Clin Oncol 31:519–520, 2008. 677. Prahlow JA, Loggie BW, Cappellari JO, et al: Extra-adrenal myelolipoma: Report of two cases. South Med J 88:639–643, 1995. 678. Prasad AR, Amin MB, Randolph TL, et al: Retrorectal cystic hamartoma: Report of 5 cases with malignancy arising in 2. Arch Pathol Lab Med 124:725–729, 2000. 679. Prasad S, Krishnan A, Limdi J, et al: Imaging features of gossypiboma: Report of two cases. J Postgrad Med 45:18–19, 1999. 680. Prasad S, Patankar T: Computed tomography demonstration of a fat-fluid level in tuberculous chylous ascites. Australas Radiol 43:542– 543, 1999. 681. Pretre R, Robert J, Mirescu D, et al: Pathophysiology, recognition and management of pneumoretroperitoneum. Br J Surg 80:1138–1140, 1993. 682. Price AJ, Compson JP, Calonje E: Fibrolipomatous hamartoma of nerve arising in the brachial plexus. J Hand Surg Br 20:16–18, 1995. 683. Prousalidis J, Tzardinoglou K, Sgouradis L, et al: Uncommon sites of hydatid disease. World J Surg 22:17–22, 1998. 684. Pui MH, Fletcher BD, Langston JW: Granulocytic sarcoma in childhood leukemia: imaging features. Radiology 190:698–702, 1994. 685. Pui MH, Liu MJ, Guo Y, et al: Computed tomography of retroperitoneal paragangliomas. Australas Radiol 43:303–306, 1999. 686. Qiao HS, Feng XL, Yong L, et al: The MRI of extraadrenal pheochromocytoma in the abdominal cavity. Eur J Radiol 62:335–341, 2007. 687. Quint LE, Glazer GM, Francis IR, et al: Pheochromocytoma and paraganglioma: Comparison of MR imaging with CT and I-131 MIBG scintigraphy. Radiology 165:89–93, 1987. 688. Radin DR: Disseminated histoplasmosis: Abdominal CT findings in 16 patients. AJR Am J Roentgenol 157:955–958, 1991. 689. Radin DR: Intraabdominal Mycobacterium tuberculosis vs Mycobacterium avium-intracellulare infections in patients with AIDS: Distinction based on CT findings. AJR Am J Roentgenol 156:487–491, 1991. 690. Radin R: Acute mesenteric and retroperitoneal lymphadenitis in systemic lupus erythematosus: Case report. Abdom Imaging 26:411–413, 2001. 691. Radin R, David CL, Goldfarb H, et al: Adrenal and extra-adrenal retroperitoneal ganglioneuroma: Imaging findings in 13 adults. Radiology 202:703–707, 1997. 692. Radin DR, Siskind B, Weiner S, et al: Retroperitoneal malacoplakia. Urol Radiol 6:218–220, 1984. 693. Rahmouni A, Tempany C, Jones R, et al: Lymphoma: Monitoring tumor size and signal intensity with MR imaging. Radiology 188:445– 451, 1993. 694. Ramanathan RC, Thomas JM: Malignant peripheral nerve sheath tumours associated with von Recklinghausen’s neurofibromatosis. Eur J Surg Oncol 25:190–193, 1999. 695. Randazzo RF, Neustein P, Koyle MA: Spontaneous perinephric hemorrhage from extrarenal angiomyolipoma. Urology 29:428–431, 1987. 696. Rao P, Kenney PJ, Wagner BJ, et al: Imaging and pathologic features of myelolipoma. Radiographics 17:1373–1385, 1997. 697. Rappaport W, Haynes K: The retained surgical sponge following intra-abdominal surgery. A continuing problem. Arch Surg 125:405– 407, 1990. 698. Raptopoulos V, Kleinman PK, Marks S, Jr, et al: Renal fascial pathway: Posterior extension of pancreatic effusions within the anterior pararenal space. Radiology 158:367–374, 1986. 699. Raptopoulos V, Lei QF, Touliopoulos P, et al: Why perirenal disease does not extend into the pelvis: The importance of closure of the cone of the renal fasciae. AJR Am J Roentgenol 164:1179–1184, 1995. 700. Raptopoulos V, Touliopoulos P, Lei QF, et al: Medial border of the perirenal space: CT and anatomic correlation. Radiology 205:777–784, 1997.

1934

PART II  CT and MR Imaging of the Whole Body

701. Rasmussen TE, Hallett JW, Jr: Inflammatory aortic aneurysms. A clinical review with new perspectives in pathogenesis. Ann Surg 225:155–164, 1997. 702. Raut CP, Pisters PW: Retroperitoneal sarcomas: Combined-modality treatment approaches. J Surg Oncol 94:81–87, 2006. 703. Reddy D, Salomon C, Demos TC, et al: Mesenteric lymph node cavitation in celiac disease. AJR Am J Roentgenol 178:247, 2002. 704. Rehring TF, Brewster DC, Kaufman JA, et al: Regression of perianeurysmal fibrosis and ureteral dilation following endovascular repair of inflammatory abdominal aortic aneurysm. Ann Vasc Surg 15:591–593, 2001. 705. Renzulli P, Candinas D: Symptomatic retroperitoneal cyst: A diagnostic challenge. Ann R Coll Surg Engl 91:W9–W11, 2009. 706. Rha SE, Byun JY, Jung SE, et al: Neurogenic tumors in the abdomen: Tumor types and imaging characteristics. Radiographics 23:29–43, 2003. 707. Richard HM, Thall EH, Mitty H, et al: Desmoid tumor-ureteral fistula in Gardner’s syndrome. Urology 49:135–138, 1997. 708. Rigor VU, Goldstone SE, Jones J, et al: Hibernoma. A case report and discussion of a rare tumor. Cancer 57:2207–2211, 1986. 709. Ripetti V, Caricato M: Arullani A: Rectal perforation, retropneumoperitoneum, and pneumomediastinum after stapling procedure for prolapsed hemorrhoids: Report of a case and subsequent considerations. Dis Colon Rectum 45:268–270, 2002. 710. Robbin MR, Murphey MD, Temple HT, et al: Imaging of musculoskeletal fibromatosis. Radiographics 21:585–600, 2001. 711. Roberts WW, Nelson JB, Fishman EK, et al: Diagnosis of retroperitoneal pulmonary sequestration using computerized tomography guided fine needle biopsy. J Urol 164:445, 2000. 712. Rohlfing BM, Webb WR, Schlobohm RM: Ventilator-related extraalveolar air in adults. Radiology 121:25–31, 1976. 713. Ronnett BM, Yan H, Kurman RJ, et al: Patients with pseudomyxoma peritonei associated with disseminated peritoneal adenomucinosis have a significantly more favorable prognosis than patients with peritoneal mucinous carcinomatosis. Cancer 92:85–91, 2001. 714. Ros PR, Lubbers PR, Olmsted WW, et al: Hemangioma of the liver: Heterogeneous appearance on T2-weighted images. AJR Am J Roentgenol 149:1167–1170, 1987. 715. Ros PR, Olmsted WW, Moser RP, Jr, et al: Mesenteric and omental cysts: Histologic classification with imaging correlation. Radiology 164:327–332, 1987. 716. Rosado-de-Christenson ML, Abbott GF, McAdams HP, et al: From the archives of the AFIP: Localized fibrous tumor of the pleura. Radiographics 23:759–783, 2003. 717. Rosai J, Dorfman RF: Sinus histiocytosis with massive lymphadenopathy. A newly recognized benign clinicopathological entity. Arch Pathol 87:63–70, 1969. 718. Rosenberg SA: Validity of the Ann Arbor staging classification for the non-Hodgkin’s lymphomas. Cancer Treat Rep 61:1023–1027, 1977. 719. Rosenkrantz AB, Spieler B, Seuss CR, et al: Utility of MRI features for differentiation of retroperitoneal fibrosis and lymphoma. AJR Am J Roentgenol 199:118–126, 2012. 720. Rosenthal CL, Bishop MC: Accessory spleen presenting as a retroperitoneal tumour. Eur Urol 7:314–316, 1981. 721. Ross JS, Prout GR, Jr: Retroperitoneal fat necrosis producing ureteral obstruction. J Urol 115:524–529, 1976. 722. Ross MJ, Welch WR, Scully RE: Multilocular peritoneal inclusion cysts (so-called cystic mesotheliomas). Cancer 64:1336–1346, 1989. 723. Rothacker D, Knolle J, Stiller D, et al: Primary retroperitoneal mucinous cystadenomas with gastric epithelial differentiation. Pathol Res Pract 189:1195–1204, 1993. 724. Rubesin SE, Herlinger H, Saul SH, et al: Adult celiac disease and its complications. Radiographics 9:1045–1066, 1989. 725. Rudniki C, Kessler E, Zarfati M, et al: Kikuchi’s necrotizing lymphadenitis: A cause of fever of unknown origin and splenomegaly. Acta Haematol 79:99–102, 1988. 726. Rufini V, Calcagni ML, Baum RP: Imaging of neuroendocrine tumors. Semin Nucl Med 36:228–247, 2006.

727. Sabate JM, Torrubia S, Maideu J, et al: Sclerosing mesenteritis: Imaging findings in 17 patients. AJR Am J Roentgenol 172:625–629, 1999. 728. Sacks D, Banner MP, Meranze SG, et al: Renal and related retroperitoneal abscesses: Percutaneous drainage. Radiology 167:447– 451, 1988. 729. Safioleas MC, Constantinos K, Michael S, et al: Benign multicystic peritoneal mesothelioma: A case report and review of the literature. World J Gastroenterol 12:5739–5742, 2006. 730. Sahdev A, Sohaib A, Monson JP, et al: CT and MR imaging of   unusual locations of extra-adrenal paragangliomas (pheochromocytomas). Eur Radiol 15:85–92, 2005. 731. Saini S: Multi-detector row CT: Principles and practice for abdominal applications. Radiology 233:323–327, 2004. 732. Saksena MA, Saokar A, Harisinghani MG: Lymphotropic nanoparticle enhanced MR imaging (LNMRI) technique for lymph node imaging. Eur J Radiol 58:367–374, 2006. 733. Salimi Z, Fishbein M, Wolverson MK, et al: Pancreatic lymphangioma: CT, MRI, and angiographic features. Gastrointest Radiol 16:248–250, 1991. 734. Sall I, Ali AA, El Kaoui H, et al: Primary hydatid cyst of the retroperitoneum. Am J Surg 199:e25–e26, 2010. 735. Salvarani C, Pipitone N, Versari A, et al: Positron emission tomography (PET): Evaluation of chronic periaortitis. Arthritis Rheum 53:298–303, 2005. 736. Samuels BL: Management of recurrent desmoid tumor after surgery and radiation: Role of cytotoxic and non-cytotoxic therapies. Surg Oncol 8:191–196, 1999. 737. Sandrasegaran K, Rydberg J, Tann M, et al: Benefits of routine use of coronal and sagittal reformations in multi-slice CT examination of the abdomen and pelvis. Clin Radiol 62:340–347, 2007. 738. Sansom HE, Blunt DM, Moskovic EC: Large retroperitoneal hibernoma—CT findings with pathological correlation. Clin Radiol 54:625–627, 1999. 739. Sasaki S, Yasuda K, Takigami K, et al: Inflammatory abdominal aortic aneurysms and atherosclerotic abdominal aortic aneurysms— Comparisons of clinical features and long-term results. Jpn Circ J 61:231–235, 1997. 740. Savic B, Birtel FJ, Tholen W, et al: Lung sequestration: Report of seven cases and review of 540 published cases. Thorax 34:96–101, 1979. 741. Sawhney R, D’Agostino HB, Zinck S, et al: Treatment of postoperative lymphoceles with percutaneous drainage and alcohol sclerotherapy.   J Vasc Interv Radiol 7:241–245, 1996. 742. Scheel PJ, Jr, Feeley N: Retroperitoneal fibrosis: The clinical, laboratory, and radiographic presentation. Medicine (Baltimore) 88:202–207, 2009. 743. Scheer M, Hon M, Fruauff AA, et al: Perinephric xanthogranulomatosis: CT diagnosis and confirmation by CT-guided percutaneous biopsy. Clin Imaging 24:64–67, 2000. 744. Schnall MD, Lenkinski RE, Pollack HM, et al: Prostate: MR imaging with an endorectal surface coil. Radiology 172:570–574, 1989. 745. Schreyer HH, Uggowitzer MM, Ruppert-Kohlmayr A: Helical CT of the urinary organs. Eur Radiol 12:575–591, 2002. 746. Schwarzbach MH, Hinz U, Dimitrakopoulou-Strauss A, et al: Prognostic significance of preoperative [18-F] fluorodeoxyglucose (FDG) positron emission tomography (PET) imaging in patients with resectable soft tissue sarcomas. Ann Surg 241:286–294, 2005. 747. Schweizer RT, Cho S, Koutz KS, et al: Lymphoceles following renal transplantation. Arch Surg 104:42–45, 1972. 748. Scott WW, Jr, Fishman EK: Extramedullary hematopoiesis mimicking the appearance of carcinomatosis or peritoneal mesothelioma: Computed tomography demonstration. Gastrointest Radiol 15:82–83, 1990. 749. Sebbag G, Yan H, Shmookler BM, et al: Results of treatment of 33 patients with peritoneal mesothelioma. Br J Surg 87:1587–1593, 2000. 750. Seki H, Shiina M, Nishihara M, et al: Primary retroperitoneal mucinous cystadenocarcinoma: Report of a case. Radiat Med 8:164–167, 1990. 751. Sered S, Nikolaidis P: CT findings of perirenal plasmacytoma. AJR Am J Roentgenol 181:888, 2003.

CHAPTER 55  Retroperitoneum 752. Shaffrey JK, Brinker DA, Horton KM, et al: Atypical extralobar sequestration: CT-pathological correlation. Clin Imaging 23:223–226, 1999. 753. Shah Syed GM, Naseer H, Usmani GN, et al: Role of iodine-131 MIBG scanning in the management of paediatric patients with neuroblastoma. Med Princ Pract 13:196–200, 2004. 754. Shahidi H, Myers JL, Kvale PA: Castleman’s disease. Mayo Clin Proc 70:969–977, 1995. 755. Shankar KR, Roche CJ, Carty HM, et al: Cystic retroperitoneal lymphangioma: Treatment by image-guided percutaneous catheter drainage and sclerotherapy. Eur Radiol 11:1021–1023, 2001. 756. Shanmuganathan K, Mirvis SE, Sover ER: Value of contrast-enhanced CT in detecting active hemorrhage in patients with blunt abdominal or pelvic trauma. AJR Am J Roentgenol 161:65–69, 1993. 757. Shao H, Yang ZG, Xu GH, et al: Tuberculosis in the abdominal lymph nodes: Evaluation with contrast-enhanced magnetic resonance imaging. Int J Tuberc Lung Dis 17:90–95, 2013. 758. Shapiro B, Sisson JC, Shulkin BL, et al: The current status of metaiodobenzylguanidine and related agents for the diagnosis of neuroendocrine tumors. Q J Nucl Med 39:3–8, 1995. 759. Sharma S, Nabi G, Seth A, et al: Pelvic lipomatosis presenting as uraemic encephalopathy. Int J Clin Pract 55:149–150, 2001. 760. Shelton MW, Morian JP, Radford DM: Pseudomyxoma retroperitonei associated with appendiceal cystadenoma. Am Surg 60:958–960, 1994. 761. Sherazi ZA, Dicks-Mireaux C: Suprarenal cystic masses: Unusual causes. Clin Radiol 52:953–955, 1997. 762. Sherer DM, Abulafia O, Eliakim R: Pseudomyxoma peritonei: A review of current literature. Gynecol Obstet Invest 51:73–80, 2001. 763. Sherman JL, Hartman DS, Friedman AC, et al: Angiomyolipoma: Computed tomographic-pathologic correlation of 17 cases. AJR Am J Roentgenol 137:1221–1226, 1981. 764. Sheward SE, Williams AG, Jr, Mettler FA, Jr, et al: CT appearance of a surgically retained towel (gossypiboma). J Comput Assist Tomogr 10:343–345, 1986. 765. Shibata D, Lewis JJ, Leung DH, et al: Is there a role for incomplete resection in the management of retroperitoneal liposarcomas? J Am Coll Surg 193:373–379, 2001. 766. Shin JH, Yoon CH, Cho KS, et al: Fetus-in-fetu in the scrotal sac of a newborn infant: Imaging, surgical and pathological findings. Eur Radiol 9:945–947, 1999. 767. Shinbane JS, Colletti PM, Shellock FG: MR in patients with pacemakers and ICDs: Defining the issues. J Cardiovasc Magn Reson 9:5–13, 2007. 768. Shinozaki K, Yoshimitsu K, Honda H, et al: Abdominal cystic tumors containing small amount of fat in the septa: Report of two cases. Abdom Imaging 26:315–318, 2001. 769. Shulkin BL, Hutchinson RJ, Castle VP, et al: Neuroblastoma: Positron emission tomography with 2-[fluorine-18]-fluoro-2-deoxy-D-glucose compared with metaiodobenzylguanidine scintigraphy. Radiology 199:743–750, 1996. 770. Shulkin BL, Thompson NW, Shapiro B, et al: Pheochromocytomas: imaging with 2-[fluorine-18]fluoro-2-deoxy-D-glucose PET. Radiology 212:35–41, 1999. 771. Siegel MJ, Jamroz GA, Glazer HS, et al: MR imaging of intraspinal extension of neuroblastoma. J Comput Assist Tomogr 10:593–595, 1986. 772. Siegelman SS, Copeland BE, Saba GP, et al: CT of fluid collections associated with pancreatitis. AJR Am J Roentgenol 134:1121–1132, 1980. 773. Siegelman ES, Outwater EK: The concentric-ring sign revisited. AJR Am J Roentgenol 166:1493, 1996. 774. Silberstein SD: Methysergide. Cephalalgia 18:421–435, 1998. 775. Silverman TA, Enzinger FM: Fibrolipomatous hamartoma of nerve. A clinicopathologic analysis of 26 cases. Am J Surg Pathol 9:7–14, 1985. 776. Simo M, Zapata C, Esquius J, et al: Aggressive angiomyxoma of the female pelvis and perineum. Report of two cases and review of the literature. Br J Obstet Gynaecol 99:925–927, 1992. 777. Simons GW, Sty JR, Starshak RJ: Retroperitoneal and retrofascial abscesses. A review. J Bone Joint Surg Am 65:1041–1058, 1983. 778. Singer MA, Cintron JR, Martz JE, et al: Retrorectal cyst: A rare tumor frequently misdiagnosed. J Am Coll Surg 196:880–886, 2003.

1935

779. Singh H, Walia DS, Samra NS, et al: Surgical management of   primary retroperitoneal hydatid cyst. J Indian Med Assoc 103:336–337, 2005. 780. Singla AK, Kechejian G, Lopez MJ: Giant presacral myelolipoma. Am Surg 69:334–338, 2003. 781. Sivit CJ, Frazier AA, Eichelberger MR: Prevalence and distribution of extraperitoneal hemorrhage associated with splenic injury in infants and children. AJR Am J Roentgenol 172:1015–1017, 1999. 782. Skarin AT, Dorfman DM: Non-Hodgkin’s lymphomas: Current classification and management. CA Cancer J Clin 47:351–372, 1997. 783. Smith VC, Edwards RA, Jorgensen JL, et al: Unilocular retroperitoneal cyst of mesothelial origin presenting as a renal mass. Arch Pathol Lab Med 124:766–769, 2000. 784. Smith AJ, Lewis JJ, Merchant NB, et al: Surgical management of intra-abdominal desmoid tumours. Br J Surg 87:608–613, 2000. 785. Smith HO, Worrell RV, Smith AY, et al: Aggressive angiomyxoma of the female pelvis and perineum: Review of the literature. Gynecol Oncol 42:79–85, 1991. 786. Sobic-Saranovic D, Artiko V, Obradovic V: FDG PET imaging in sarcoidosis. Semin Nucl Med 43:404–411, 2013. 787. Sodhi KS, Suri S, Nijhawan R, et al: Rosai-Dorfman disease: Unusual cause of diffuse and massive retroperitoneal lymphadenopathy. Br J Radiol 78:845–847, 2005. 788. Sogaard AS, Laurberg JM, Sorensen M, et al: Intraabdominal and retroperitoneal soft-tissue sarcomas—Outcome of surgical treatment in primary and recurrent tumors. World J Surg Oncol 8:81, 2010. 789. Solkar MH, Akhtar NM, Khan Z, et al: Pseudomyxoma extraperitonei occurring 35 years after appendicectomy: A case report and review of literature. World J Surg Oncol 2:19, 2004. 790. Somasundar P, Krouse R, Hostetter R, et al: Paragangliomas—A decade of clinical experience. J Surg Oncol Suppl 74:286–290, 2000. 791. Somerhausen NS, Fletcher CD: Diffuse-type giant cell tumor: Clinicopathologic and immunohistochemical analysis of 50 cases with extraarticular disease. Am J Surg Pathol 24:479–492, 2000. 792. Song T, Shen J, Liang BL, et al: Retroperitoneal liposarcoma: MR characteristics and pathological correlative analysis. Abdom Imaging 2007. 793. Soreide JA, Soreide K, Korner H, et al: Benign peritoneal cystic mesothelioma. World J Surg 30:560–566, 2006. 794. Sozen S, Gurocak S, Uzum N, et al: The importance of re-evaluation in patients with cystitis glandularis associated with pelvic lipomatosis: A case report. Urol Oncol 22:428–430, 2004. 795. Steeper TA, Rosai J: Aggressive angiomyxoma of the female pelvis and perineum. Report of nine cases of a distinctive type of gynecologic soft-tissue neoplasm. Am J Surg Pathol 7:463–475, 1983. 796. Steinberg L, Rothman D, Drey NW: Mullerian cyst of the retroperitoneum. Am J Obstet Gynecol 107:963–964, 1970. 797. Stifelman MD, Shah O, Mufarrij P, et al: Minimally invasive management of retroperitoneal fibrosis. Urology 71:201–204, 2008. 798. Stoeckle E, Coindre JM, Bonvalot S, et al: Prognostic factors in retroperitoneal sarcoma: A multivariate analysis of a series of 165 patients of the French Cancer Center Federation Sarcoma Group. Cancer 92:359–368, 2001. 799. Storm FK, Mahvi DM: Diagnosis and management of retroperitoneal soft-tissue sarcoma. Ann Surg 214:2–10, 1991. 800. Stotter AT, Grigg MJ, Mansfield AO: The response of peri-aneurysmal fibrosis–The “inflammatory” aneurysm—To surgery and steroid therapy. Eur J Vasc Surg 4:201–205, 1990. 801. Stoupis C, Ros PR, Abbitt PL, et al: Bubbles in the belly: Imaging of cystic mesenteric or omental masses. Radiographics 14:729–737, 1994. 802. Strauss DC, Qureshi YA, Hayes AJ, et al: Management of benign retroperitoneal schwannomas: A single-center experience. Am J Surg 202:194–198, 2011. 803. Su IC, Chen CM: Spontaneous healing of retroperitoneal chylous leakage following anterior lumbar spinal surgery: A case report and literature review. Eur Spine J 2007. 804. Subramony C, Habibpour S, Hashimoto LA: Retroperitoneal mucinous cystadenoma. Arch Pathol Lab Med 125:691–694, 2001.

1936

PART II  CT and MR Imaging of the Whole Body

805. Sueyoshi E, Sakamoto I, Uetani M: Endovascular repair of inflammatory abdominal aortic aneurysm: Serial changes of   periaortic fibrosis demonstrated by CT. Abdom Imaging 34:523–526, 2009. 806. Sugarbaker PH, Ronnett BM, Archer A, et al: Pseudomyxoma peritonei syndrome. Adv Surg 30:233–280, 1996. 807. Sugimura K, Carrington BM, Quivey JM, et al: Postirradiation changes in the pelvis: Assessment with MR imaging. Radiology 175:805–813, 1990. 808. Suit H, Spiro I: Radiation in the multidisciplinary management of desmoid tumors. Front Radiat Ther Oncol 35:107–119, 2001. 809. Sultan S, Duffy S, Madhavan P, et al: Fifteen-year experience of transperitoneal management of inflammatory abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 18:510–514, 1999. 810. Sumino H, Kanda T, Nakamura T, et al: Steroid therapy is effective in a young patient with an inflammatory abdominal aortic aneurysm. J Med 30:67–74, 1999. 811. Sun X, Liu C, Wang R, et al: The value of MDCT in diagnosis of hyaline-vascular Castleman’s disease. Eur J Radiol 81:2436–2439, 2012. 812. Sun J, VongPhet S, Zhang Z, et al: Fetus-in-fetu: Imaging and pathologic findings. Abdom Imaging 37:147–150, 2012. 813. Sung MS, Kang HS, Suh JS, et al: Myxoid liposarcoma: Appearance at MR imaging with histologic correlation. Radiographics 20:1007–1019, 2000. 814. Sung MS, Lee GK, Kang HS, et al: Sacrococcygeal chordoma: MR imaging in 30 patients. Skeletal Radiol 34:87–94, 2005. 815. Sunga KL, Bellolio MF, Gilmore RM, et al: Spontaneous retroperitoneal hematoma: Etiology, characteristics, management, and outcome.   J Emerg Med 43:e157–e161, 2012. 816. Susmano DE, Dolin EH: Computed tomography in diagnosis of pelvic lipomatosis. Urology 13:215–220, 1979. 817. Swallow CJ, Catton CN: Local management of adult soft tissue sarcomas. Semin Oncol 34:256–269, 2007. 818. Swallow CJ, Catton CN: Improving outcomes for retroperitoneal sarcomas: A work in progress. Surg Oncol Clin N Am 21:317–331, 2012. 819. Swayne LC, Ginsberg HN, Ginsburg A: Pneumoretroperitoneum secondary to hydrogen peroxide wound irrigations. AJR Am J Roentgenol 148:149–150, 1987. 820. Swerdlow S, Campo E, Harris NL, et al: WHO Classification of Tumours of Haematopoietic and Lymphoid Tissue (IARC WHO Classification of Tumours), ed 4, Lyon, France, 2008, World Health Organization, pp 1–441. 821. Sze G, Uichanco LS, 3rd, Brant-Zawadzki MN, et al: Chordomas: MR imaging. Radiology 166:187–191, 1988. 822. Takahashi M, Momose T, Kameyama M, et al: Abnormal accumulation of [18F]fluorodeoxyglucose in the aortic wall related to inflammatory changes: Three case reports. Ann Nucl Med 20:361–364, 2006. 823. Takahashi K, Okabe Y, Orino A, et al: Spontaneous rupture of a biliary diverticulum in the distal common bile duct, with formation of a retroperitoneal biloma. Gastrointest Endosc 61:783–787, 2005. 824. Takamura M, Murakami T, Kurachi H, et al: MR imaging of mesenteric hemangioma: A case report. Radiat Med 18:67–69, 2000. 825. Takao H, Yamahira K, Doi I, et al: Gastrointestinal stromal tumor of the retroperitoneum: CT and MR findings. Eur Radiol 14:1926–1929, 2004. 826. Takao H, Yamahira K, Watanabe T: Encapsulated fat necrosis mimicking abdominal liposarcoma: Computed tomography findings.   J Comput Assist Tomogr 28:193–194, 2004. 827. Takatera H, Takiuchi H, Namiki M, et al: Retroperitoneal schwannoma. Urology 28:529–531, 1986. 828. Takebayashi S, Ono Y, Sakai F, et al: Computed tomography of amyloidosis involving retroperitoneal lymph nodes mimicking lymphoma. J Comput Assist Tomogr 8:1025–1027, 1984. 829. Takiff H, Brems JJ, Elliott ML: Calcified retroperitoneal enteric duplication cyst. Am J Gastroenterol 88:470–471, 1993. 830. Takizawa I, Morishita H, Matsuki S, et al: Primary gastrointestinal stromal tumor in the retroperitoneum. Int J Urol 13:1245–1248, 2006.

831. Tanaka T, Sakuma H, Takahashi K, et al: Extravesical malacoplakia— Possibly originated from a superficial part of the renal cortex. Acta Pathol Jpn 31:323–334, 1981. 832. Tang T, Boyle JR, Dixon AK, et al: Inflammatory abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 29:353–362, 2005. 833. Tateishi U, Hasegawa T, Beppu Y, et al: Primary dedifferentiated liposarcoma of the retroperitoneum. Prognostic significance of computed tomography and magnetic resonance imaging features.   J Comput Assist Tomogr 27:799–804, 2003. 834. Tateishi U, Hasegawa T, Makimoto A, et al: Adult neuroblastoma: Radiologic and clinicopathologic features. J Comput Assist Tomogr 27:321–326, 2003. 835. Tateishi U, Nishihara H, Morikawa T, et al: Solitary fibrous tumor of the pleura: MR appearance and enhancement pattern. J Comput Assist Tomogr 26:174–179, 2002. 836. Tatsumi M, Cohade C, Nakamoto Y, et al: Direct comparison of FDG PET and CT findings in patients with lymphoma: Initial experience. Radiology 237:1038–1045, 2005. 837. Taylor JR, Ryu J, Colby TV, et al: Lymphangioleiomyomatosis. Clinical course in 32 patients. N Engl J Med 323:1254–1260, 1990. 838. Teh HS, Lin MB, Tan AS, et al: Retroperitoneal Castleman’s disease in the perinephric space—Imaging appearance: A case report and a review of the literature. Ann Acad Med Singapore 29:773–776, 2000. 839. Teichgraber V, Aube C, Schmidt D, et al: Percutaneous MR-guided radiofrequency ablation of recurrent sacrococcygeal chordomas. AJR Am J Roentgenol 187:571–574, 2006. 840. Temizoz O, Genchellac H, Demir MK, et al: Bilateral extra-adrenal perirenal myelolipomas: CT features. Br J Radiol 83:e198–e199, 2010. 841. Tennant WG, Hartnell GG, Baird RN, et al: Inflammatory aortic aneurysms: Characteristic appearance on magnetic resonance imaging. Eur J Vasc Surg 6:399–402, 1992. 842. Tenti P, Carnevali L, Tateo S, et al: Primary mucinous cystoadenocarcinoma of the retroperitoneum: Two cases. Gynecol Oncol 55:308–312, 1994. 843. Terner JY, Lattes R: Malakoplakia of colon and retroperitoneum. Report of a case with a histochemical study of the Michaelis-Gutman inclusion bodies. Am J Clin Pathol 44:20–31, 1965. 844. Testart J, Plissonnier D, Peillon C, et al: [Inflammatory abdominal aortic aneurysm. Role of corticosteroid therapy]. J Mal Vasc 25:201– 207, 2000. 845. Tezel E, Sare M, Edali N, et al: Retroperitoneal malignant teratoma. A case report. Mater Med Pol 27:123–125, 1995. 846. Thamboo TP, Sim R, Tan SY, et al: Primary retroperitoneal mucinous cystadenocarcinoma in a male patient. J Clin Pathol 59:655–657, 2006. 847. Thurnher MM, Schima W, Turetschek K, et al: Peripancreatic fat necrosis mimicking pancreatic cancer. Eur Radiol 11:922–925, 2001. 848. Timmers HJ, Kozupa A, Chen CC, et al: Superiority of fluorodeoxyglucose positron emission tomography to other functional imaging techniques in the evaluation of metastatic SDHB-associated pheochromocytoma and paraganglioma. J Clin Oncol 25:2262–2269, 2007. 849. Titton RL, Gervais DA, Hahn PF, et al: Urine leaks and urinomas: Diagnosis and imaging-guided intervention. Radiographics 23:1133– 1147, 2003. 850. Tjalma WA, Vaneerdeweg W: Primary retroperitoneal mucinous cystadenocarcinomas are a distinct entity. Int J Gynecol Cancer 2007. 851. Tong RS, Larner T, Finlay M, et al: Pelvic lipomatosis associated with proliferative cystitis occurring in two brothers. Urology 59:602, 2002. 852. Torigian DA, Huang SS, Houseni M, et al: Functional imaging of cancer with emphasis on molecular techniques. CA Cancer J Clin 57:206–224, 2007. 853. Torigian DA, Kaiser LR, Soma LA, et al: Symptomatic dysrhythmia caused by a posterior mediastinal angiomyolipoma. AJR Am J Roentgenol 178:93–96, 2002. 854. Torigian DA, Siegelman ES: CT findings of pelvic lipomatosis of nerve. AJR Am J Roentgenol 184:S94–S96, 2005. 855. Torigian DA, Zaidi H, Kwee TC, et al: PET/MR imaging: Technical aspects and potential clinical applications. Radiology 267:26–44, 2013.

CHAPTER 55  Retroperitoneum 856. Torres VE, Bjornsson J, King BF, et al: Extrapulmonary lymphangioleiomyomatosis and lymphangiomatous cysts in tuberous sclerosis complex. Mayo Clin Proc 70:641–648, 1995. 857. Torres GM, Cernigliaro JG, Abbitt PL, et al: Iliopsoas compartment: Normal anatomy and pathologic processes. Radiographics 15:1285– 1297, 1995. 858. Tosaka A, Ando M, Arisawa C, et al: Endocrinologically active retroperitoneal ganglioneuroma with positive iodine-131metaiodobenzylguanidine scintigraphy. Int J Urol 6:471–474, 1999. 859. Trampal C, Engler H, Juhlin C, et al: Pheochromocytomas: Detection with 11C hydroxyephedrine PET. Radiology 230:423–428, 2004. 860. Treglia G, Mattoli MV, Bertagna F, et al: Emerging role of Fluorine-18fluorodeoxyglucose positron emission tomography in patients with retroperitoneal fibrosis: A systematic review. Rheumatol Int 33:549–555, 2013. 861. Trerotola SO, Kuhlman JE, Fishman EK: Bleeding complications of femoral catheterization: CT evaluation. Radiology 174:37–40, 1990. 862. Tsai YS, Cheng HL, Lin JS, et al: Retroperitoneal plasmacytoma associated with hyperamylasemia. J Urol 162:1681–1682, 1999. 863. Tschoep K, Kohlmann A, Schlemmer M, et al: Gene expression profiling in sarcomas. Crit Rev Oncol Hematol 63:111–124, 2007. 864. Tseng TK, Lee RC, Chou YH, et al: Retroperitoneal venous hemangioma. J Formos Med Assoc 104:681–683, 2005. 865. Tseng CA, Pan YS, Su YC, et al: Extrarenal retroperitoneal angiomyolipoma: Case report and review of the literature. Abdom Imaging 29:721–723, 2004. 866. Tsitouridis J, Stamos S, Hassapopoulou E, et al: Extramedullary paraspinal hematopoiesis in thalassemia: CT and MRI evaluation. Eur J Radiol 30:33–38, 1999. 867. Tzeng CW, Fiveash JB, Heslin MJ: Radiation therapy for retroperitoneal sarcoma. Expert Rev Anticancer Ther 6:1251–1260, 2006. 868. Tzeng CW, Smith JK, Heslin MJ: Soft tissue sarcoma: Preoperative and postoperative imaging for staging. Surg Oncol Clin N Am 16:389–402, 2007. 869. Uchinami M, Morioka K, Doi K, et al: Retroperitoneal laparoscopic management of a lymphocele after abdominal aortic surgery: A case report. J Vasc Surg 42:552–555, 2005. 870. Ueda K, Tsunoda A, Nakamura A, et al: Presacral epidermoid cyst: Report of a case. Surg Today 28:665–668, 1998. 871. Ueno T, Tanaka YO, Nagata M, et al: Spectrum of germ cell tumors: From head to toe. Radiographics 24:387–404, 2004. 872. Ulbright TM, Morley DJ, Roth LM, et al: Papillary serous carcinoma of the retroperitoneum. Am J Clin Pathol 79:633–637, 1983. 873. Upadhyay N, Gomez D, Button MF, et al: Retroperitoneal enteric duplication cyst presenting as a pancreatic cystic lesion. A case report. JOP 7:492–495, 2006. 874. Urban T, Lazor R, Lacronique J, et al: Pulmonary lymphangioleiomyomatosis. A study of 69 patients. Groupe d’Etudes et de Recherche sur les Maladies “Orphelines” Pulmonaires (GERM“O”P). Medicine (Baltimore) 78:321–337, 1999. 875. Ushijima M, Hashimoto H, Tsuneyoshi M, et al: Giant cell tumor of the tendon sheath (nodular tenosynovitis). A study of 207 cases to compare the large joint group with the common digit group. Cancer 57:875–884, 1986. 876. Vaglio A, Buzio C: Chronic periaortitis: A spectrum of diseases. Curr Opin Rheumatol 17:34–40, 2005. 877. Vaglio A, Salvarani C, Buzio C: Retroperitoneal fibrosis. Lancet 367:241–251, 2006. 878. Vaglio A, Versari A, Fraternali A, et al: 18)F-fluorodeoxyglucose positron emission tomography in the diagnosis and followup of idiopathic retroperitoneal fibrosis. Arthritis Rheum 53:122–125, 2005. 879. Vallabhaneni SR, McWilliams RG, Anbarasu A, et al: Perianeurysmal fibrosis: A relative contra-indication to endovascular repair. Eur J Vasc Endovasc Surg 22:535–541, 2001. 880. Vallat-Decouvelaere AV, Dry SM, Fletcher CD: Atypical and malignant solitary fibrous tumors in extrathoracic locations: Evidence of their comparability to intra-thoracic tumors. Am J Surg Pathol 22:1501–1511, 1998.

1937

881. van Bommel EF: Retroperitoneal fibrosis. Neth J Med 60:231–242, 2002. 882. van Bommel EF, Hendriksz TR, Huiskes AW, et al: Brief communication: Tamoxifen therapy for nonmalignant retroperitoneal fibrosis. Ann Intern Med 144:101–106, 2006. 883. van Bommel EF, Jansen I, Hendriksz TR, et al: Idiopathic retroperitoneal fibrosis: Prospective evaluation of incidence and clinicoradiologic presentation. Medicine (Baltimore) 88:193–201, 2009. 884. van Dalen T, Plooij JM, van Coevorden F, et al: Long-term prognosis   of primary retroperitoneal soft tissue sarcoma. Eur J Surg Oncol 33:234–238, 2007. 885. van den Brekel MW, Castelijns JA, Stel HV, et al: Detection and characterization of metastatic cervical adenopathy by MR imaging: Comparison of different MR techniques. J Comput Assist Tomogr 14:581–589, 1990. 886. Vandevenne JE, De Schepper AM, De Beuckeleer L, et al: New concepts in understanding evolution of desmoid tumors: MR imaging of 30 lesions. Eur Radiol 7:1013–1019, 1997. 887. van Geel AN, Pastorino U, Jauch KW, et al: Surgical treatment of lung metastases: The European Organization for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group study of 255 patients. Cancer 77:675–682, 1996. 888. van Gils AP, Falke TH, van Erkel AR, et al: MR imaging and MIBG scintigraphy of pheochromocytomas and extraadrenal functioning paragangliomas. Radiographics 11:37–57, 1991. 889. Vanhoenacker FM, De Backer AI, de Op BB, et al: Imaging of gastrointestinal and abdominal tuberculosis. Eur Radiol 14(Suppl 3): E103–E115, 2004. 890. vanSonnenberg E, Wittich GR, Casola G, et al: Lymphoceles: Imaging characteristics and percutaneous management. Radiology 161:593–596, 1986. 891. Vargas-Serrano B, Alegre-Bernal N, Cortina-Moreno B, et al: Abdominal cystic lymphangiomas: US and CT findings. Eur J Radiol 19:183–187, 1995. 892. Vasanawala SS, Desser T, Jeffrey RB: Value of delayed imaging in MDCT of the abdomen and pelvis. AJR Am J Roentgenol 187:154–163, 2006. 893. Vasani PK, Soni HC, Murali Krishnan NN, et al: Fetus-in-fetu—A case report. Abdom Imaging 35:504–506, 2010. 894. Vassilopoulos PP, Apostolikas NG, Papajoglou I, et al: Ectopic spleen in the retroperitoneum. Case report. Acta Chir Scand 156:655–658, 1990. 895. Vastel L, Lambert P, De Pinieux G, et al: Surgical treatment of pigmented villonodular synovitis of the hip. J Bone Joint Surg Am 87:1019–1024, 2005. 896. Vazquez JL, Thorsen MK, Dodds WJ, et al: Evaluation and treatment of intraabdominal bilomas. AJR Am J Roentgenol 144:933–938, 1985. 897. Veit P, Ruehm S, Kuehl H, et al: Lymph node staging with dualmodality PET/CT: Enhancing the diagnostic accuracy in oncology. Eur J Radiol 58:383–389, 2006. 898. Velchik MG, Alavi A, Kressel HY, et al: Localization of pheochromocytoma: MIGB, CT, and MRI correlation. J Nucl Med 30:328–336, 1989. 899. Veyssier-Belot C, Cacoub P, Caparros-Lefebvre D, et al: ErdheimChester disease. Clinical and radiologic characteristics of 59 cases. Medicine (Baltimore) 75:157–169, 1996. 900. Vickery RE: Benign fibrous mesothelial cyst of the retroperitoneum. South Med J 64:762–763, 1971. 901. Vinnicombe SJ, Norman AR, Nicolson V, et al: Normal pelvic lymph nodes: Evaluation with CT after bipedal lymphangiography. Radiology 194:349–355, 1995. 902. Vitellas KM, El-Dieb A, Vaswani KK, et al: Using contrast-enhanced MR cholangiography with IV mangafodipir trisodium (Teslascan) to evaluate bile duct leaks after cholecystectomy: A prospective study of 11 patients. AJR Am J Roentgenol 179:409–416, 2002. 903. Vivas I, Nicolas AI, Velazquez P, et al: Retroperitoneal fibrosis: Typical and atypical manifestations. Br J Radiol 73:214–222, 2000. 904. Vlahos L, Trakadas S, Gouliamos A, et al: Retrocrural masses of extramedullary hemopoiesis in beta-thalassemia. Magn Reson Imaging 11:1227–1229, 1993.

1938

PART II  CT and MR Imaging of the Whole Body

905. Voros D, Theodorou D, Ventouri K, et al: Retroperitoneal tumors: Do the satellite tumors mean something? J Surg Oncol 68:30–33, 1998. 906. Vossough A, Torigian DA, Zhang PJ, et al: Extrathoracic solitary fibrous tumor of the pelvic peritoneum with central malignant degeneration on CT and MRI. J Magn Reson Imaging 2005. 907. Walker R, Barlogie B, Haessler J, et al: Magnetic resonance imaging in multiple myeloma: Diagnostic and clinical implications. J Clin Oncol 25:1121–1128, 2007. 908. Walker DI, Bloor K, Williams G, et al: Inflammatory aneurysms of the abdominal aorta. Br J Surg 59:609–614, 1972. 909. Wallis F, Roditi GH, Redpath TW, et al: Inflammatory abdominal aortic aneurysms: Diagnosis with gadolinium enhanced T1-weighted imaging. Clin Radiol 55:136–139, 2000. 910. Wan J, Ritchey ML, Muraszko K, et al: Retroperitoneal neurogenous choristoma. J Urol 148:1867–1868, 1992. 911. Wang RM, Chen CA: Primary retroperitoneal teratoma. Acta Obstet Gynecol Scand 79:707–708, 2000. 912. Wang YM, Li YW, Sheih CP, et al: Magnetic resonance imaging of neuroblastoma, ganglioneuroblastoma, and ganglioneuroma. ChungHua Min Kuo Hsiao Erh Ko i Hsueh Hui Tsa Chih 36:420–424, 1995. 913. Wang SE, Tsai YF, Su CH, et al: Retroperitoneal bronchogenic cyst mimicking pancreatic cystic lesion. J Chin Med Assoc 69:538–542, 2006. 914. Watanabe AT, Jeffrey RB, Jr: CT diagnosis of traumatic rupture of the cisterna chyli. J Comput Assist Tomogr 11:175–176, 1987. 915. Watanabe N, Morijiri M, Shimizu M, et al: A case of retroperitoneal extramedullary plasmacytoma with multiple metastases. Clin Imaging 24:365–367, 2000. 916. Weinshel EL, Peterson BA: Hodgkin’s disease. CA Cancer J Clin 43:327–346, 1993. 917. Weinstein JL, Katzenstein HM, Cohn SL: Advances in the diagnosis and treatment of neuroblastoma. Oncologist 8:278–292, 2003. 918. Weiss SW, Tavassoli FA: Multicystic mesothelioma. An analysis of pathologic findings and biologic behavior in 37 cases. Am J Surg Pathol 12:737–746, 1988. 919. Weiss SW, Rao VK: Well-differentiated liposarcoma (atypical lipoma) of deep soft tissue of the extremities, retroperitoneum, and miscellaneous sites. A follow-up study of 92 cases with analysis of the incidence of “dedifferentiation”. Am J Surg Pathol 16:1051–1058, 1992. 920. White M, Mueller PR, Ferrucci JT, Jr, et al: Percutaneous drainage of postoperative abdominal and pelvic lymphoceles. AJR Am J Roentgenol 145:1065–1069, 1985. 921. Wiebe BM, Jensen K, Laursen H: Parachordoma of the sacrococcygeal region—A neuroepithelial tumor. Clin Neuropathol 14:343–346, 1995. 922. Wilbur AC, Turk JN, Capek V: Perirenal metastases from lung cancer: CT diagnosis. J Comput Assist Tomogr 16:589–591, 1992. 923. Win Z, Al-Nahhas A, Towey D, et al: 68Ga-DOTATATE PET in neuroectodermal tumours: First experience. Nucl Med Commun 28:359–363, 2007. 924. Windsor JJ: Cat-scratch disease: Epidemiology, aetiology and treatment. Br J Biomed Sci 58:101–110, 2001. 925. Wintermark M, Maeder P, Reichhart M, et al: MR pattern of hyperacute cerebral hemorrhage. J Magn Reson Imaging 15:705–709, 2002. 926. Witteles RM, Kaplan EL, Roizen MF: Sensitivity of diagnostic and localization tests for pheochromocytoma in clinical practice. Arch Intern Med 160:2521–2524, 2000. 927. Wolff M: Lymphangiomyoma: Clinicopathologic study and ultrastructural confirmation of its histogenesis. Cancer 31:988–1007, 1973. 928. Wong WW, Hirose T, Scheithauer BW, et al: Malignant peripheral nerve sheath tumor: Analysis of treatment outcome. Int J Radiat Oncol Biol Phys 42:351–360, 1998. 929. Woodring JH, Howard RS, 2nd, Johnson MV: Massive low-attenuation mediastinal, retroperitoneal, and pelvic lymphadenopathy on CT   from lymphangioleiomyomatosis. Case report. Clin Imaging 18:7–11, 1994. 930. Woodruff JM: Pathology of tumors of the peripheral nerve sheath in type 1 neurofibromatosis. Am J Med Genet 89:23–30, 1999.

931. Woodward PJ, Sohaey R, O’Donoghue MJ, et al: From the archives of the AFIP: Tumors and tumorlike lesions of the testis: Radiologicpathologic correlation. Radiographics 22:189–216, 2002. 932. Wright DH, Richards DB: Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): Report of a case with widespread nodal and extra nodal dissemination. Histopathology 5:697–709, 1981. 933. Wu X, Kellokumpu-Lehtinen PL, Pertovaara H, et al: Diffusionweighted MRI in early chemotherapy response evaluation of patients with diffuse large B-cell lymphoma—A pilot study: comparison with 2-deoxy-2-fluoro- D-glucose-positron emission tomography/computed tomography. NMR Biomed 24:1181–1190, 2011. 934. Yamanaka M, Saitoh F, Saitoh H, et al: Primary retroperitoneal ganglioneuroblastoma in an adult. Int J Urol 8:130–132, 2001. 935. Yamato M, Fuhrman CR: Computed tomography of fatty replacement in extramedullary hematopoiesis. J Comput Assist Tomogr 11:541–542, 1987. 936. Yamauchi K, Yasuda M: Comparison in treatments of nonleukemic granulocytic sarcoma: Report of two cases and a review of 72 cases in the literature. Cancer 94:1739–1746, 2002. 937. Yan SL, Lin H, Kuo CL, et al: Primary retroperitoneal mucinous cystadenoma: Report of a case and review of the literature. World J Gastroenterol 14:5769–5772, 2008. 938. Yang DM, Jung DH, Kim H, et al: Retroperitoneal cystic masses: CT, clinical, and pathologic findings and literature review. Radiographics 24:1353–1365, 2004. 939. Yang DM, Kim HC, Lee HL, et al: Squamous cell carcinoma arising from a presacral epidermoid cyst: CT and MR findings. Abdom Imaging, 2007. 940. Yang WT, Lam WW, Yu MY, et al: Comparison of dynamic helical   CT and dynamic MR imaging in the evaluation of pelvic lymph   nodes in cervical carcinoma. AJR Am J Roentgenol 175:759–766, 2000. 941. Yang HJ, Lee SW, Lee HJ, et al: Extralobar pulmonary sequestration mimicking an adrenal tumor. JSLS 16:671–674, 2012. 942. Yang ZG, Min PQ, Sone S, et al: Tuberculosis versus lymphomas in the abdominal lymph nodes: Evaluation with contrast-enhanced CT. AJR Am J Roentgenol 172:619–623, 1999. 943. Yang DM, Park CH, Jin W, et al: Tailgut cyst: MRI evaluation. AJR Am J Roentgenol 184:1519–1523, 2005. 944. Yang DM, Yoon MH, Kim HS, et al: Presacral epidermoid cyst: Imaging findings with histopathologic correlation. Abdom Imaging 26:79–82, 2001. 945. Yildirim S, Tarim A, Nursal TZ, et al: Retained surgical sponge (gossypiboma) after intraabdominal or retroperitoneal surgery: 14 cases treated at a single center. Langenbecks Arch Surg 391:390–395, 2006. 946. Yilmaz M, Camci C, Sari I, et al: Histiocytic necrotizing lymphadenitis (Kikuchi-Fujimoto’s disease) mimicking systemic lupus erythematosus: A review of two cases. Lupus 15:384–387, 2006. 947. Yohannan J, Feng T, Allaf ME: Retroperitoneal hibernoma in a 28-year-old man. Urology 78:320–321, 2011. 948. Yonemoto T, Tatezaki S, Takenouchi T, et al: The surgical management of sacrococcygeal chordoma. Cancer 85:878–883, 1999. 949. Yoo E, Kim MJ, Kim KW, et al: A case of mesenteric cystic lymphangioma: Fat saturation and chemical shift MR imaging. J Magn Reson Imaging 23:77–80, 2006. 950. Yoon SS, Chen YL, Kirsch DG, et al: Proton-beam, intensity-modulated, and/or intraoperative electron radiation therapy combined with aggressive anterior surgical resection for retroperitoneal sarcomas. Ann Surg Oncol 17:1515–1529, 2010. 951. York JE, Kaczaraj A, Abi-Said D, et al: Sacral chordoma: 40-year experience at a major cancer center. Neurosurgery 44:74–79, discussion 79–80, 1999. 952. Younathan CM, Kaude JV: Renal peripelvic lymphatic cysts (lymphangiomas) associated with generalized lymphangiomatosis. Urol Radiol 14:161–164, 1992. 953. Young AL, Baysal BE, Deb A, et al: Familial malignant catecholaminesecreting paraganglioma with prolonged survival associated with mutation in the succinate dehydrogenase B gene. J Clin Endocrinol Metab 87:4101–4105, 2002.

CHAPTER 55  Retroperitoneum 954. Young RH, Clement PB, McCaughey WT: Solitary fibrous tumors (‘fibrous mesotheliomas’) of the peritoneum. A report of three cases and a review of the literature. Arch Pathol Lab Med 114:493–495, 1990. 955. Yu RS, Zhang WM, Liu YQ: CT diagnosis of 52 patients with lymphoma in abdominal lymph nodes. World J Gastroenterol 12:7869– 7873, 2006. 956. Yuh WT, Barloon TJ, Sickels WJ, et al: Magnetic resonance imaging in the diagnosis and followup of idiopathic retroperitoneal fibrosis. J Urol 141:602–605, 1989. 957. Zangos S, Steenburg SD, Phillips KD, et al: Acute abdomen: Added diagnostic value of coronal reformations with 64-slice multidetector row computed tomography. Acad Radiol 14:19–27, 2007. 958. Zar R, McClintock C: Kikuchi-Fujimoto disease: A case report and review of literature. Conn Med 70:491–494, 2006. 959. Zbar AP, Agrawal A, Saeed IT, et al: Gossypiboma revisited: A case report and review of the literature. J R Coll Surg Edinb 43:417–418, 1998. 960. Zhang M, Li M, Xu GP, et al: Neoplasm-like abdominal nonhematogenous disseminated tuberculous lymphadenopathy: CT evaluation of 12 cases and literature review. World J Gastroenterol 17:4038–4043, 2011.

1939

961. Zhang Y, Nishimura H, Kato S, et al: MRI of ganglioneuroma: Histologic correlation study. J Comput Assist Tomogr 25:617–623, 2001. 962. Zhang Y, Wu S, Xi Z, et al: Measuring diagnostic accuracy of imaging parameters in pelvic lipomatosis. Eur J Radiol 81:3107–3114, 2012. 963. Zheng X, Pan K, Cheng J, et al: Localized Castleman disease in retroperitoneum: Newly discovered features by multi-detector helical CT. Abdom Imaging 2007. 964. Zhou LP, Zhang B, Peng WJ, et al: Imaging findings of Castleman disease of the abdomen and pelvis. Abdom Imaging 33:482–488, 2008. 965. Zhu H, Chen H, Zhang S, et al: Intra-abdominal fibromatosis: Differentiation from gastrointestinal stromal tumour based on biphasic contrast-enhanced CT findings. Clin Radiol 68:1133–1139, 2013. 966. Zhu F, Zhang JT, Xing XW, et al: Rosai-Dorfman disease: A retrospective analysis of 13 cases. Am J Med Sci 345:200–210, 2013. 967. Zinzani PL, Musuraca G, Alinari L, et al: Predictive role of positron emission tomography in the outcome of patients with follicular lymphoma. Clin Lymphoma Myeloma 7:291–295, 2007. 968. Zuckerman DA, Yeager TD: Percutaneous ethanol sclerotherapy   of postoperative lymphoceles. AJR Am J Roentgenol 169:433–437, 1997.

56  Male Pelvis Joseph R. Grajo, Leslie K. Lee, and Mukesh G. Harisinghani

INTRODUCTION Evaluation of the male pelvis can be performed with several imaging modalities, including computed tomography (CT), magnetic resonance imaging (MRI), and ultrasonography (US), depending on the clinical indication and anatomic region of interest. US may be the first study of choice for selected indications because of its lack of radiation exposure, benefit of real-time examination, and ability of color Doppler to investigate blood flow. However, the limited resolution afforded by US in combination with its user dependence limits its use in many indications for imaging of the male pelvis. Since the development of multislice CT technology, CT has become the predominant method of evaluating numerous disease states, including staging pelvic malignancies, assessing soft tissue and bony injury, presurgical planning, and identifying postsurgical complications. However, MRI possesses several important advantages over conventional CT, including superior intrinsic soft tissue contrast, the ability to obtain direct multiplanar imaging, and lack of radiation exposure. These factors contribute to render MRI far superior to CT for evaluation of the male genitourinary (GU) system, particularly imaging of the prostate gland, seminal vesicles, testes, and penis. This chapter reviews the anatomy, imaging techniques, normal findings, benign conditions, and malignant disorders involving the organs of the male pelvis with reference to our standard protocols for common indications.

ANATOMY OF THE PELVIS The pelvis is demarcated superiorly by the iliac crest and inferiorly by the urogenital diaphragm. The area below the urogenital diaphragm is considered the perineum. The bony pelvis is composed of paired innominate bones (ilium, ischium, pubis), sacrum, and coccyx, which together facilitate ambulation by providing attachment sites for numerous muscles and fasciae (Fig. 56-1). In addition the pelvis provides support and protection for the visceral pelvic organs, including portions of the urinary and gastrointestinal tracts, reproductive organs, and neurovascular structures. The pelvis is classically divided into two regions—the false (major) pelvis and the true (minor) pelvis—with respect to the arcuate and iliopectineal lines running posterior to anterior from the sacral promontory to the pubic crest (see Fig. 56-1). The true pelvis contains the rectum, bladder, and pelvic ureters as well as the prostate and seminal vesicles in the male. The false pelvis is open anteriorly and bounded laterally by the iliac fossa; it includes the small bowel and portions of the colon. The pelvic muscles are composed internally of the psoas muscles and the laterally oriented iliacus along the inner brim within the iliac

1940

fossa. The external muscles, from deep to superficial, include the gluteus minimus, medius, and maximus. The floor of the pelvic cavity is also known as the pelvic diaphragm (or urogenital diaphragm) and is composed of the levator ani anteriorly and the coccygeus muscle posteriorly. The ischiorectal fossa is bordered medially by the levator ani and laterally by the obturator internus muscles. The obturator internus and the piriformis muscle form the majority of the lateral pelvic sidewall. Knowledge of these muscles is important because they may be involved in pelvic neoplasms, and their involvement may affect local therapy.

PROSTATE Anatomy The prostate is a tuboalveolar exocrine gland measuring approximately 4 cm in transverse dimension, located inferior to the bladder and  anterior to the rectum. The conical prostate gland surrounds the uppermost aspect of the urethra and is enveloped by an incomplete fibromuscular capsule. The gland contains a base superiorly, a midgland, and an apex inferiorly. The anatomic zonal architecture of the prostate gland is simpler than the histologic categorizations, and it is essentially divided into three main regions: peripheral zone, central gland, and anterior fibromuscular stroma (Fig. 56-2). Important neurovascular structures lie within the pericapsular fat both anterior to the apex, as the anterior periprostatic plexus, and posterolateral to it, as the neurovascular bundles. The neurovascular bundles supply the corpus cavernosum and are critical for normal erectile function. A well-defined fat plane exists between the prostate and the obturator internus and is delineated by Denonvilliers’ fascia, which acts as a physical barrier to the spread of disease.

Imaging Techniques MRI is the workhorse of prostate imaging because of its superior soft tissue resolution, allowing for visualization of the zonal anatomy. Newer software technology, including diffusion-weighted imaging (DWI), dynamic contrast-enhanced (DCE) imaging, and spectroscopy have dramatically increased characterization of prostate pathology because of the added functional data that complements the anatomic information provided by conventional MRI. See Appendix A for discussion of our prostate MRI protocol. Indications for CT imaging of the pelvis related to the prostate gland are most commonly packaged as part of a CT of the abdomen in the evaluation of metastases from prostate carcinoma.

Normal Findings.  The individual zones of the prostate gland are usually indistinct on CT examination (Fig. 56-3). T2-weighted MRI,

CHAPTER 56  Male Pelvis however, allows exquisite visualization of the zonal prostatic anatomy (see Fig. 56-3). The peripheral zone encompasses the most volume and almost completely envelops the central zone. On MRI the peripheral zone demonstrates high T2 signal intensity secondary to its increased glandular components. The central gland (encompassing the histologically separate but visually indistinct central and transitional zones) is decreased in T2 intensity compared with the peripheral zone (see Fig. 56-3). Owing to the stromal characteristics of the anterior fibromuscular stroma, this anteriorly located region is low in T2 signal intensity.

IL

S

A

Is

P

FIG 56-1  Bony pelvis in three-dimensional surface reconstruction. The iliac (IL), pubic (P), ischial (Is), and sacral (S) bones make up the bony pelvis. The acetabulum (A) accepts the femoral head in the hip joint. The green line demarcates the inferior true pelvis from the superior false pelvis.

Benign Conditions Benign prostatic hyperplasia.  Benign prostatic hyperplasia (BPH) is a benign increase in the volume of the prostate gland, a function that increases with age. Four different types of prostatic hyperplasia have been reported and the imaging characteristics described.75 Nodular prostatic enlargement, composed mainly of glandular tissue, is the most common type and can result in constriction of the urethra and subsequent bladder obstruction due to affliction of the central gland. On CT, the enlarged gland includes nodules that can have both high- and low-attenuation regions in addition to variable enhancement. Coarse calcifications are common in BPH. The MRI appearance of nodular prostatic hyperplasia is typically low to medium signal on T1-weighted images and moderate to high signal on T2-weighted images, consistent with glandular growth (Fig. 56-4). The interstitial forms of BPH are more heterogeneous in signal but typically T2 hypointense. Mixed glandular and stromal hyperplasia can have a heterogeneous appearance on MRI. BPH is indistinguishable from prostate carcinoma in most instances on CT. Initially MRI was also considered to be inadequate in the differentiation of benign and malignant entities when first introduced in the 1980s.57,62 However, more recent advances in MRI protocols and software packages have substantially increased the sensitivity and specificity of prostate MRI.9,14,15,46,69,81,87,97,105,114,122,130 Prostatitis.  Inflammation of the prostate is divided into clinical syndromes consisting of acute bacterial, chronic bacterial, and nonbacterial prostatitis, as well as prostatodynia. Although nonbacterial prostatitis is most common, the acute presentation of prostatitis is commonly due to bacterial causes; gram-negative anaerobes such as Escherichia coli are the most prevalent offenders. Prostatitis is a clinical diagnosis, and patients typically complain of dysuria and demonstrate a tender or “boggy” prostate on physical examination. CT is not routinely indicated; however, cross-sectional imaging may be useful if there is a concern for prostate abscess or malignancy. CT findings in prostatitis are nonspecific and include periprostatic stranding and an edematous appearance of the prostate (Fig. 56-5). Prostate abscess may appear as a focal region of hypodensity with surrounding enhancement on CT, with occasional internal septations7,11,71 (Fig. 56-6). MRI

SV

SV AFMS

CZ

CZ

CG PZ PZ

A

1941

B FIG 56-2  Prostate anatomy. A, The prostate gland is composed of the central gland (CG), which consists of the central zone and the transitional zone as well as the peripheral zone (PZ) and anterior fibromuscular stroma (AFMS). B, The paired seminal vesicles are perched posterolateral to the prostate gland. CZ, central zone; PZ, peripheral zone.

1942

PART II  CT and MR Imaging of the Whole Body

Ps

OI

G

A B B

P

FIG 56-3  Normal prostate. A, Contrast-enhanced CT scan at the

C

level of the prostate gland. The peripheral zone (yellow arrow) and central gland (red arrow) can occasionally be differentiated. The obturator internus (OI) forms the lateral border of the ischiorectal fossa. The psoas muscle (Ps) and gluteal muscles (G) can also be seen. B, Axial T2-weighted MRI shows better differentiation between the peripheral zone (yellow arrow) and the central gland (red arrow). C, Sagittal T2-weighted MRI shows the hyperintense seminal vesicle (yellow arrow) in relationship to the prostate (P) and bladder (B).

FIG 56-4  Benign prostatic hyperplasia. T1-weighted axial MRI shows an enlarged prostate with heterogeneous signal intensity in the central gland (red arrow). The peripheral zone (yellow arrow) is almost completely effaced by the enlarged central gland.

CHAPTER 56  Male Pelvis

A B

C FIG 56-5  Prostatitis and seminal vesicle abscess. A, Axial CT scan demonstrates an enlarged and edematous prostate (yellow arrow) with surrounding perirectal fat stranding (red arrow). B, Seminal vesicles demonstrate bilateral regions of hypodensity (white arrowheads) with adjacent fat stranding (red arrow), which resolved with antibiotic therapy. The black arrow indicates a small amount of free pelvic fluid. C, Bilateral seminal vesicle calcifications (yellow arrowheads) and vascular calcifications (white arrow) are consistent with a history of diabetes, which predisposes patients to infection.

1943

1944

PART II  CT and MR Imaging of the Whole Body

A B

C FIG 56-6  Prostate abscess. A, On CT scan, multiple rim-enhancing abscesses within the prostate gland (arrows) are associated with adjacent fat stranding. B, Axial T2-weighted image demonstrates corresponding hypointense lesions (arrows), consistent with pus. C, Coronal T1-weighted postgadolinium image shows rim enhancement of multiple abscesses (arrows).

CHAPTER 56  Male Pelvis

A

1945

B FIG 56-7  Granulomatous prostatitis. A, T2-weighted image shows hypointense nodules (arrow) within the peripheral zone. B, Nodules (arrow) appear hyperintense on the T1-weighted image. The prostate appears enlarged with superior contour change. Biopsy was consistent with granulomatous prostatitis.

findings of prostate abscess include low T1 signal within the abscess and enhancement of the peripheral rim with gadolinium (see Fig. 56-6). Granulomatous prostatitis is a form of benign postinfectious prostatitis associated with tuberculosis or fungal infections. It can also be seen following intravesical Bacillus Calmette-Guerin (BCG) therapy for bladder cancer. The CT appearance is nonspecific and includes enlargement of the gland, with associated calcifications. MRI can demonstrate T2 hyperintense nodularity (Fig. 56-7). Prostatic and periprostatic cysts.  Cysts in the region of the prostate gland are differentiated by intraprostatic or periprostatic locations. Although they can occasionally be seen on MRI, these lesions are typically not visualized on CT because they are isodense in signal. On MRI, prostatic cysts have signal characteristics that are hyperintense on fluid-sensitive sequences. Rarely complications such as hemorrhage or proteinaceous debris alter the signal characteristics. The key differentiating factor for diagnosis is anatomic localization, which is best evaluated with MRI. Intraprostatic cysts are most often composed of müllerian duct and utricle cysts. Müllerian duct cysts originate from müllerian duct remnants and do not communicate with the urethra. In addition they often arise slightly lateral to midline, with common extension above the prostate. Utricle cysts are a result of dilation of the prostatic utricle, with a midline location and urethral communication (Fig. 56-8). Other cystic structures in this location include ejaculatory duct cysts, prostatic retention cysts, malignancy, infection, and cystic degeneration of BPH. Periprostatic cystic lesions include seminal vesicle, vas deferens, and Cowper’s gland duct cysts.126

Malignant Diseases Prostate cancer.  Prostate adenocarcinoma is the most common malignancy in males. It typically affects men older than 50 years, and is currently the third leading cause of cancer deaths in men. Detection is commonly via digital rectal examination (DRE) or elevation of serum prostate-specific antigen (PSA) levels. Clinical evaluation of prostate cancer encompasses a serologic, anatomic, and pathologic assessment of disease. The combination of serum PSA levels, clinical tumor-node-metastasis (TNM) staging, and histopathologic Gleason grading dictates the course of treatment. The

measurement of PSA levels (though nonspecific owing to variability in glandular size, patient demographics, and concomitant inflammatory processes) as a screening tool has resulted in a significant decrease in tumor grade and frequency of metastases at initial presentation. PSA is also important for following disease progression or recurrence. TNM oncologic staging is the well-accepted standard for prostate adenocarcinoma. The Gleason score is a pathologic assessment of dysplastic degree and is calculated by combining the grade of each of the two most prevalent dysplastic tissue types for a total score between 1 and 10.80 The available treatments are numerous and include surveillance, surgical resection, hormonal therapy, radiation, or some combination of these therapies. The determination of which therapeutic option is appropriate for a patient is based on several established nomograms, taking into consideration clinical and pathologic data to generate a prognostic prediction for risk stratification.83,84,115,124 Generally patients with both PSA serum levels less than 10 ng/L and Gleason grades less than 7 are thought to be potentially curable, with options for treatment being prostatectomy, brachytherapy, and external radiation therapy. Patients believed to be incurable typically receive external radiation therapy in addition to hormonal therapy. Assessment of lymph node disease is performed if the clinical disease profile is of concern for metastatic spread. Nodal disease is classified as local lymph node involvement within the true pelvis, whereas metastatic lymph node disease is classified as lymph node involvement outside the true pelvis. Node-positive imaging findings are typically lymph nodes measuring greater than 10 mm in short-axis diameter or obturator/internal iliac nodes measuring greater than 8 mm in short axis.33,43,91 However, this has proven to be nonspecific at best.6 Prostate nodal disease is typically progressive in a stepwise fashion, with downstream retroperitoneal or mediastinal disease almost invariably accompanied by local pelvic lymphadenopathy. Spread usually occurs first to the internal iliac and obturator lymph nodes. Occasionally spread is seen in the external iliac lymph nodes. The poor reliability of standard cross-sectional imaging for lymph node detection is partially due to the fact that a significant portion of microscopic nodal metastases cannot be visualized on CT or MRI. Detection of suspicious lymph nodes with noninvasive imaging is also

1946

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 56-8  Prostatic utricle cyst. A, Coronal T2 sequence shows a round hyperintensity in the midline (arrow). B, T1-weighted coronal image confirms a well-demarcated hypointense lesion (arrow) without fat signal. C, Sagittal T2-weighted image identifies a posteriorly positioned hyperintensity in the location of the prostatic utricle (arrow).

helpful in directing a CT-guided biopsy to prove metastatic disease. This confirmation abrogates the need for more invasive pelvic lymphadenectomy, a procedure with associated morbidities such as incontinence and hemorrhage.3 Computed tomography of prostate cancer.  CT is not very reliable in determining tumor extension through the prostate capsule (T3 disease) because transgression is obvious only when there is gross extension (Fig. 56-9). Advantages of CT lie in the evaluation of abdominal and pelvic lymph nodes and bony anatomy. Bulky pelvic and retroperitoneal lymph nodes can be identified. However, studies demonstrate that lymph node involvement detectable by imaging—either CT or standard MRI—is a late manifestation of disease. In advanced disease with extracapsular extension and invasion of surrounding structures, a CT scan for baseline comparison is useful.

CT monitoring of metastatic bony lesions is not as sensitive as bone scan or MRI; however, CT allows direct visualization of the cortical bone and is superior to other modalities in differentiating nonmalignant lesions. CT is also critical for guidance in planning radiation therapy. Although MRI can also provide the necessary anatomic information, calculation of intrinsic electron densities is possible only with CT. Although not typically used to detect local recurrence, CT can be helpful in evaluating for new distant metastatic lymph node disease. Magnetic resonance imaging of prostate cancer.  MRI can be used in both preoperative and postoperative assessments of prostate cancer. MRI diagnosis of primary prostate carcinoma is often requested if there is continued clinical concern for malignancy despite negative prostate biopsy and negative transrectal US examination. More recently

CHAPTER 56  Male Pelvis

FIG 56-9  Prostate cancer with rectal invasion (T4 disease). Gross extension (arrow) into the anterior rectal wall is present.

MRI has also become a popular method for identifying any suspicious focal nodule(s) for the urologist to target with US-guided or US and MRI fusion–guided biopsy.30,65,98,109,116 The majority of prostate carcinomas arise from the peripheral zone, followed by the transitional zone. They arise much less commonly from the central zone, where tumor is usually more difficult to detect owing to the heterogeneity commonly seen throughout the gland in the setting of BPH. The most clinically significant cancers, because of their proximity to the capsule, arise from the peripheral zone; they  are also the easiest lesions to visualize on both US and MRI. On T1-weighted images, the homogeneity of the prostate can make tumor detection nearly impossible. Thus T2-weighted images are one of the key sequences in demonstrating prostate tumor. The normal peripheral zone T2 hyperintensity is replaced with low signal from the dense cellularity of the tumor (Fig. 56-10). Pitfalls for cancer detection in the peripheral zone involve hemorrhage from prior biopsy, prostatitis, or stromal fibrosis. High-intensity T1 and low-intensity T2 signal from hemorrhage can persist for weeks to months after biopsy. Transitional zone malignancies are more difficult to identify owing to the inherent relatively low T2 signal. However, studies using a combination of MRI and MR spectroscopy (MRS) can improve detection.93 Beyond abnormal T2 signal, two additional techniques are also critical in the evaluation of a potential prostate neoplasm: DWI and DCE imaging. These three sequences—T2 imaging, DWI, and DCE imaging—comprise the elements of so-called multiparametric MRI of the prostate. DWI is used to evaluate the prostate for focal areas of restricted diffusion, where the normal random brownian motion of the cells’ protons are inhibited owing to increased tissue cellularity, tightly packed cell membranes, and breakdown of interstitial spaces that allow for free water diffusion. Areas of restricted diffusion appear bright on DWI sequences and dark on corresponding apparent diffusion coefficient (ADC) maps. Hypointense signal on ADC is often easier to visually detect (Fig. 56-11). The ADC map is also not susceptible to T2

1947

shine-through artifacts like DWI sequences, therefore allowing for more sensitive detection.123 Some studies have shown that the combination of DWI and T2 sequences increases the accuracy of prostate cancer detection with 85% to 90% sensitivity and specificity compared to radical prostatectomy findings.111 Beyond its qualitative role, DWI can offer the ability to quantitate the amount of restricted diffusion with a nodule, therefore allowing analysis of tissue cellularity to attempt differentiation between benign and malignant tissue.123 Furthermore, quantitative ADC measurements are being used to predict tumor grade and aggressiveness.95 DWI is also being explored for its role in predicting extracapsular extension/seminal vesicle invasion, guiding targeted biopsy, and assessing response to treatment.123 However, improvements in spatial resolution must be made to allow for more confidence in these roles. Potential nodules can also be investigated for early arterial enhancement and washout on DCE images. The raw data from these dynamic sequences are now being used to generate perfusion maps in which the prostate can be evaluated for focal areas of altered wash-in/washout kinetics and capillary permeability50,104 (Fig. 56-12). Three types of perfusion versus time curves have been described to characterize prostate tissue.111 A type 1 curve demonstrates average enhancement and washout throughout the gland, typical of normal prostate tissue. A type 2 curve shows early and intense enhancement followed by slow washout, a pattern that can be seen in BPH, prostatitis, and multifocal low-grade tumor. A type 3 curve is characterized by early and intense contrast enhancement followed by rapid washout, a feature of highgrade tumors (Fig. 56-13). Abnormal perfusion in conjunction with focal nodular T2 hypointense signal and restricted diffusion makes the radiologist most comfortable in raising suspicion for potential tumor as well as an area to target on biopsy (Fig. 56-14).59,99 Studies have shown that interpretation of perfusion imaging with T2 signal abnormalities leads to sensitivity and specificity approaching 90% to 95% for detection of moderate to high-grade prostate cancer compared to radical prostatectomy specimens.111 Another technique used in prostate cancer is MRS, a unique application of MRI that allows for the relative quantification of tissue metabolites within a preselected voxel of interest. Choline (Cho), creatine (Cr), and citrate (Cit) are the main metabolites measured with proton spectroscopy in the prostate. An elevated Cho/Cr : Cit ratio greater than 0.75 (the cutoff value calculated to be 2 standard deviations above normal prostate ratios) is suggestive of malignancy94 (Fig. 56-15). Cho is a phospholipid cell membrane component that is increased in high turnover states, and Cit is a biochemical molecule produced by normal prostate tissue. Cr is involved in energy metabolism, but because its spectroscopic peak is relatively difficult to resolve from the adjacent creatine signal, the Cho and Cr peaks are added to get the Cho/Cr peak. The addition of MRS to MRI has not yielded a consistent increase in the accuracy of tumor detection in patients with elevated PSA levels and prior negative biopsy or no prior biopsy.29,39,53,86 MRS may be more promising in tumor staging or determining  tumor aggressiveness or for the evaluation of posttreatment recurrence.19,93,107,110 Studies also indicate that MRS may be helpful for distinguishing prostate adenocarcinoma from benign entities.113 Current MRS techniques are in part limited by resolution and finite voxel  size for tumor imaging; however, technologic advances may significantly improve the applicability of MRS in the detection of prostate malignancies. One of the main functions of MRI is to attempt to distinguish between T2 and T3 disease, specifically capsular transgression and neurovascular bundle invasion. Detailed investigation for capsular transgression includes high-resolution small field-of-view (FOV) images, most importantly in an oblique axial plane perpendicular to

A

B

FIG 56-10  Peripheral zone prostate cancer (T2 disease). A, Axial

C

T1-weighted MRI demonstrates no obvious prostate lesion. The periprostatic fat is well seen. B, Axial T2-weighted MRI shows a hypointense lesion (arrow) in the background of the normal bright peripheral zone signal. The prostate capsule appears to be intact. C, The tumor (arrow) is shown without capsule transgression on the coronal T2-weighted image.

FIG 56-11  Restricted diffusion in prostate cancer. ADC map demonstrates restricted diffusion within the peripheral zone of the right mid-gland in this patient with biopsy-proven Gleason 8 adenocarcinoma (arrow).

CHAPTER 56  Male Pelvis

FIG 56-12  Perfusion maps. Ktrans (magenta color map), Ve (yellow color map), Kep (green color map), and iAUGC (aqua color map) perfusion maps reflect increased capillary permeability/forward flux (Ktrans), rapid washout/reverse flux (Kep), and increased gadolinium concentration (iAUGC) within a biopsy proven Gleason 7 adenocarcinoma in the peripheral zone of the right mid-gland (arrows).

1949

Curve analysis ROI 7

160 150 140 130 120 110 100

Percent change

90 80 70 60 50 40 30 20 10 0 –10 0 0.0

1 n/a

2 n/a

3 n/a

4 0.0

5 12.0

6 24.0

7 36.0

8 48.0

9 60.0

10 72.0

11 84.0

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 96.0 108.0 120.0 132.0 144.0 156.0 168.0 180.0 192.0 204.0 216.0 228.0 240.0 252.0 264.0 276.0 288.0 Timepoints (post injuction time, sec) Average curve

FIG 56-13  Type 3 curve. A time curve generated from a region of interest placed on the same tumor as in Figure 56-12 demonstrates a type 3 curve, indicative of early arterial enhancement and rapid washout.

FIG 56-14  Multiparametric MRI. T2-weighted image, ADC map, DCE image, and perfusion color map all corresponding to tumor in the peripheral zone of the right mid-gland.

29 n/a

CHAPTER 56  Male Pelvis

A

B

–0.17 –0.48

1.50 0.24

2.74 0.30

3.03 0.33

0.99 1.04

2.31 0.34

1.73 1.24

–3.29 0.10

0.19 0.26

0.78 0.20

1.01 0.40

1.87 0.38

5.64 0.88

1.21 0.42

0.72 0.20

1.77 67.2

–0.01 –0.12

1.23 0.17

3.39 0.42

1.41 1.65

1.64 1.27

1.34 0.45

0.89 0.38

3.48 –32.1

3.70 0.20

0.98 0.33

–1.09 –0.03

1.27 0.64

0.91 0.51

1.97 0.36

1.50 1.02

7.25 3.90

0.46 0.73

0.75 0.50

0.43 0.22

0.23 0.57

1.28 1.24

2.72 0.27

1.99 1.31

5.45 2.66

1.30 0.27

0.58 0.31

0.21 0.12

3.03 0.24

4.01 1.03

5.94 1.18

*

4.75 4.45

C

1951

1.46 0.49

*

*

*

*

*

*

*

*

*

FIG 56-15  MRS of bilateral prostate cancer. A, Axial T2-weighted MRI shows bilateral hypointensities (arrows) in the peripheral zones. B, Voxel grid placement over the same axial image. C, Corresponding MRS readout. Abnormal choline-creatine peak (yellow arrows) and citrate peak (red asterisks) ratios are depicted by the lighter boxes, which correspond to the bilateral tumors seen in the peripheral prostate zones. (Courtesy Dr. Fergus Coakley, Department of Radiology, University of San Francisco, San Francisco, Calif.)

the long axis of the prostate gland. Many institutions attempt to further enhance signal to the prostate by using an endorectal coil rather than a standard phased array pelvic or body coil.76,121 Some argue that the endorectal coil improves sensitivity and specificity of prostate tumor detection and local staging.13,61,74 (See Appendix B for instructions on proper endorectal coil placement technique.) Other practices defer placement of the endorectal coil, especially when scanning at 3 teslas (T).18,47,67 Criteria for extracapsular spread include direct visualization of tumor extension, asymmetry or envelopment of the neurovascular bundle, angulated prostate contour, irregular or spiculated prostate margin, retracted capsule, and obliteration of the rectoprostate angle50 (Fig. 56-16). Seminal vesicle invasion can also be discerned on MRI. Findings consistent with seminal vesicle invasion include loss or disruption of the normal seminal vesicle architecture, replacement of the T2-hyperintense seminal vesicle with hypointense tumor, and obliteration of the normal prostate–seminal vesicle angle (see Fig. 56-16).32 Sensitivity for detecting extracapsular extension and seminal vesicle

invasion ranges from 13% to 95% and 23% to 80%, respectively; specificity ranges from 49% to 97% and 81% to 99%, respectively.50 Beyond evaluation for primary carcinoma within the prostate gland, staging is also dependent on detection of suspicious lymph nodes. Traditionally both MRI and CT use maximum short-axis diameter of a lymph node for distinguishing pathologic nodal disease. MRI has the benefit of superior soft tissue characterization of individual lymph nodes, but it suffers from lengthy examination times and an inability to assess all the lymph nodes in the abdomen and pelvis. However, the introduction of DWI and MR lymphangiography brings promise for an expanded role of MRI in the characterization of nodal disease. Although restricted diffusion is common and variable in normal lymph nodes because of the highly cellular structure of lymphoid tissue, quantitative ADC measurements may help distinguish metastatic nodal disease even in normal-sized lymph nodes.34 MR lymphangiography, an even newer technique, may help further characterize nodal disease based on the uptake of intravenously injected

1952

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 56-16  Prostate cancer with seminal vesicle invasion (T3B disease). A, Axial T2-weighted MRI identifies a low-intensity lesion in the right peripheral zone with irregular and expanded outer contours (arrow), consistent with extracapsular extension. B, On the T2-weighted image, the right seminal vesicle has been partially replaced with hypointense tumor (arrow). C, The left seminal vesicle is unaffected and demonstrates normal T2 hyperintensity (arrow).

iron nanoparticles. On T2 and T2* sequences, benign nodes demonstrate signal dropout due to phagocytic ingestion of the ferrous contrast material, whereas malignant nodes show signal retention. One study shows 90% accuracy in detecting malignant nodal disease when a combination of DWI and lymphangiography is used.56 Finally, prostate MRI is also performed in the postoperative setting, evaluating for recurrent disease following total or radical prostatectomy. If there is no known soft tissue mass in the prostatectomy bed, placement of an endorectal coil is unnecessary. However, studies have also indicated that endorectal coil MRI is useful in the detection of local recurrence in prostatectomy patients with biochemical PSA recurrence.2,70,72 In patients treated with radiation instead of surgery,

local recurrence is more difficult to assess secondary to posttreatment changes and fibrosis.

SEMINAL VESICLES Anatomy The seminal vesicles are perched posterior and superolateral atop the prostate gland. Superior soft tissue characterization makes fluidsensitive MRI sequences the modality of choice when evaluating the seminal vesicles, which are T2 hyperintense (see Fig. 56-3C). The secreted seminal vesicular fluid exits via the seminal ducts, which coalesce with the vas deferens along the posterior margin of the 

CHAPTER 56  Male Pelvis

1953

prostate gland and become the paired ejaculatory ducts. The ejaculatory ducts travel through the central gland transitional zone and empty into the prostatic urethra at the utricle. Coronal and sagittal imaging of the seminal vesicles is crucial because malignancies often arise from the glandular base or ejaculatory ducts.

Imaging Techniques MRI is the preferred modality in evaluating the seminal vesicles, whether for investigation of a primary seminal vesicle abnormality  or assessment of secondary invasion from prostate carcinoma. High soft tissue contrast and small FOV capabilities allow for exquisite examination of the seminal vesicles on MRI. However, seminal vesicle abnormalities are probably more often encountered on CT imaging performed for nonspecific pelvic pain or for another unrelated indication.

Normal Findings.  T2-weighted sequences demonstrate homogeneously hyperintense signal in the normal-appearing seminal vesicles. The T1 signal of the seminal vesicles is usually isointense to slightly hyperintense relative to skeletal muscle. A small percentage of patients demonstrate heterogeneous or increased T1 signal in the seminal vesicles, which is almost uniformly a benign finding and perhaps a function of aging.17 Some have postulated that this increased T1 signal in elderly men is related to amyloid deposition.38 On both CT and MRI, the seminal vesicles appear as elongated fluid-filled structures with fine internal septations. The vas deferens course posteriorly along the pelvic sidewalls, extending from the internal ring of the internal inguinal canal to the level of the ejaculatory duct.38

A

Benign Conditions Seminal vesicle cyst.  Congenital seminal vesicle cyst is a rare condition that is almost always associated with GU anomalies, the most common being ipsilateral renal agenesis or hypogenesis (Fig. 56-17). It is postulated that obstruction at the seminal vesicle– ejaculatory duct junction is responsible.66 Ectopic ureteral insertions into the seminal vesicles, ejaculatory duct, vas deferens, and prostatic urethra have been reported. Other ipsilateral genital anomalies can include vas deferens agenesis. Acquired seminal vesicle cysts can be secondary to any obstruction of the seminal vesicle or ejaculatory duct.66 Surgical procedures such as transurethral radical prostatectomy can cause fibrosis and scarring and lead to obstruction. Seminal vesiculitis and abscess.  Prostatitis causing subsequent edema may also obstruct the seminal vesicles. Complications of cystic disease in the seminal vesicles can result in infection because stasis with any incidental hemorrhage is an ideal culture medium for microbes. A rare complication of prostatitis or prostate abscess is seminal vesiculitis with possible seminal vesicle abscess formation. Seminal vesiculitis presents as diffuse seminal vesicle wall thickening with diffuse enhancement of the wall and septa on CT or MRI.38 Like abscesses in other locations, a seminal vesicle abscess will classically present as a thickwalled cyst on CT or MRI (see Fig. 56-5). In addition to prostatitis, diabetes, instrumentation, or surgery can be associated with seminal vesicle abscess formation.38 Benign tumors.  Tumors of the seminal vesicle are extremely rare. They can be detected clinically as a nonspecific retrovesical  mass on digital rectal examination. When a seminal vesicle tumor is detected on imaging, it is usually an incidental finding. Even large seminal vesicle tumors are most often asymptomatic. In most cases, seminal vesicle tumors are benign, although they can be difficult  to differentiate from malignant tumors even at histopathology. The majority of benign seminal vesicle tumors consist of adenomas, fibromas, and leiomyomas.16

B FIG 56-17  Congenital seminal vesicle cyst. A, Axial CT scan shows cystic structures in the left seminal vesicle (yellow arrow) and a normal right seminal vesicle (red arrow). B, Left renal aplasia is present, consistent with the urogenital anomalies associated with seminal vesicle cysts.

Malignant Diseases.  Characterization of any seminal vesicle mass should first involve exclusion of a separate primary malignancy within an adjacent organ in the pelvis (e.g., prostate, bladder, rectum), since metastasis to the seminal vesicles is much more common than a primary malignancy. Specifically the prostate should be carefully evaluated for possible carcinoma with subsequent invasion of the seminal vesicles. Findings consistent with seminal vesicle invasion include loss or disruption of the normal seminal vesicle architecture, replacement of the T2-hyperintense seminal vesicle with hypointense tumor, and obliteration of the normal prostate–seminal vesicle angle (see Fig. 56-16).32 Sensitivity for detecting seminal vesicle invasion ranges from 23% to 80% while specificity ranges from 81% to 99%.50 Primary malignant tumors of the seminal vesicles also occur but are even more rare than benign neoplasms. Adenocarcinoma is the most common primary malignancy of the seminal vesicles, although other histologies such as sarcomas, seminomas, and mesodermal tumors have also been reported.16 Primary malignancies of the seminal vesicle typically present as infiltrative enhancing soft tissue masses with imaging features similar to advanced prostate cancer.

1954

PART II  CT and MR Imaging of the Whole Body

TESTES, SCROTUM, AND SOFT TISSUES Anatomy The testes are the principal male reproductive organs and are located within the scrotal sac. The testes are surrounded by a thick layer of fibrous capsule called the tunica albuginea, which is fused at the bare area of the testes with the tunica vaginalis. This mesothelial layer is derived from its developmental precursor, which contributes to testicular descent and is known as the processus vaginalis. The testis itself is composed of seminiferous tubules responsible for production of sperm. The seminiferous tubules converge at the rete testis and subsequently merge into the efferent ductules, which become the posteriorly positioned epididymis. Leaving the epididymis is the vas deferens, which courses through the spermatic cord along with vascular and neural components to join the seminal ducts.

Imaging Techniques US is the initial modality of choice in the imaging evaluation of testicular pathology. The most important role of imaging in patient management is the determination of intratesticular or extratesticular location of a palpable mass, since most intratesticular lesions are malignant (>90%) and most extratesticular lesions are benign (97%).82 MRI can characterize testicular lesions but is not indicated for primary imaging of the testes. CT is usually reserved for staging of testicular cancer and screening for retroperitoneal adenopathy.

Normal Findings.  The normal testis is an ovoid structure demonstrating homogeneous signal on T1- and T2-weighted MRI (Fig. 56-18). It is intermediate in T1 signal and hyperintense in T2 signal

C

relative to skeletal muscle. The mediastinum testis is T1 isointense and T2 hypointense to the testicular parenchyma. Surrounding the testis is the tunica albuginea, which is hypointense on T1 and T2 imaging. The rete testis extends from the mediastinum testis to the level of the tunica. The epididymis is somewhat heterogeneous but isointense to the testis on T1 images and hypointense to the testis on T2 images. Postcontrast T1 images show homogeneous enhancement of the testis, with relative hyperintensity of the epididymis. The scrotal wall is hypointense on both T1 and T2.119

Benign Conditions Cryptorchidism.  Cryptorchidism refers to testes that are either atrophic or abnormally located along the route of descent during development. Undescended testes occur in about 3% of male infants, but the majority of these descend by 1 year of age without medical intervention. Approximately 25% of cases are bilateral.36 Undescended testes are associated with comorbidities that include a significantly increased risk of infertility and malignancy.20 Interestingly the risk of malignancy is not limited to the abnormally located testis.36 The majority of patients with cryptorchidism do not require imaging evaluation; the testes can be palpated within the inguinal canal in 80% of cases. However, when the testicular location remains indeterminate, imaging can be useful. US evaluation is rapid and easy  to perform; however, similar to physical examination the intraabdominal regions cannot be adequately assessed. Some reports indicate that CT or MRI may be useful in determining testicular location.55,103 CT is excellent for detecting testes in the inguinal canal (Fig. 56-19). On MRI the testes may appear hypoplastic with relatively low T2 signal and small size.55 Depending on the patient’s age, treatment is

C

S C C

* T

A

* T

B FIG 56-18  Normal penis and scrotum. A, T2-weighted MRI of the penis and scrotum. The bilateral dorsal corpora cavernosa (C) are surrounded by the tunica albuginea (yellow arrow). The urethra (white arrow) is embedded in the corpus spongiosum (S). The testes (T) and epididymis (asterisks) are seen in the scrotal sac. B, Penile MRI along the long axis, with the corpora cavernosa (C) delineated by the tunica albuginea (yellow arrows).

CHAPTER 56  Male Pelvis either orchiopexy or orchiectomy.20 Laparoscopy is the diagnostic procedure of choice because therapeutic surgery can be performed at the same time. Trauma.  Imaging of scrotal trauma typically begins with US. However, MRI can be used in equivocal cases. In the setting of trauma, heterogeneous low T2 signal in the background of normal hyperintense parenchymal signal should raise suspicion for injury. Testicular rupture can be excluded on MRI if the hypointense fibrous tunica albuginea is preserved.96 Vascular insults.  US is the test of choice to exclude testicular torsion and evaluate for infarct, given its real-time color Doppler capabilities. Occasionally, testicular infarct can be confused with a mass at US. In these equivocal cases, contrast-enhanced MRI can help differentiate the entities. Hyperintense T1 signal on unenhanced images suggest hemorrhage, which can often be seen with testicular infarcts.

FIG 56-19  Undescended testis. The right inguinal canal demonstrates a hypodensity, representing an undescended testis (arrow).

A

1955

Postcontrast T1 images of testicular infarcts will classically show nonenhancement of the avascular infarct with a peripheral rim of enhancement. Fournier’s disease.  Fournier’s disease is an uncommon polymicrobial necrotizing fasciitis that affects the perineal, perianal, and genital organs.44 Infection is initiated by a local wound that acts as a portal of entry and results in the synergistic growth of several different strains of microbes. This potentially devastating disease is a result of extracellular enzymatic destruction of fascial planes, which ultimately results in an obliterative endarteritis. Infection in the scrotum can spread to the penis through Buck’s and dartos fascial layers. Testicular involvement is uncommon because the blood supply of the testes differs from that of the scrotum. CT may show free gas within the affected region, indicating gangrenous tissue as well as soft tissue stranding and fascial thickening (Fig. 56-20). In addition to defining the extent of involvement, an offending infectious source such as a perirectal abscess may be identified.35 Intratesticular lesions.  Most intratesticular masses are malignant. However, a few benign intratesticular masses or masslike lesions are encountered at US and MRI. A commonly encountered entity that should not be mistaken for a mass is cystic dilatation (or tubular ectasia) of the rete testis, which presents as a nonenhancing fluid-filled T2-hyperintense structure emanating from the mediastinum testis. Testicular hematomas can occur in the setting of trauma, as alluded to above. Testicular abscess can also appear as an intratesticular masslike lesion, usually as a complication of bacterial epididymo-orchitis, or more rarely as a complication of trauma, infarction, or mumps.119 Lipomas can occur within the testicular parenchyma, demonstrating the classic well-circumscribed nonenhancing T1-hyperintense characteristics of lipomas in any other location. Adrenal rests can develop within the testis in the setting of increased circulating corticotropin (e.g., congenital adrenal hyperplasia, Cushing’s syndrome, Addison’s disease). Although nonspecific, MR features typically include isointense T1 signal, hypointense T2 signal, and diffuse enhancement  on postcontrast images.119 Sarcoidosis can also present as rounded enhancing T2-hypointense lesions or as an infiltrative process involving the testis and/or epididymis. The differential diagnosis for infiltrative processes involving the testis and/or epididymis also includes 

B

FIG 56-20  Fournier’s gangrene. A, Axial CT scan demonstrates gas (arrow) with significant surrounding soft tissue stranding in the perineum. B, Coronal CT image shows gas (arrow) in the perineum as well as significant edematous change within the scrotal sac (arrowhead). The patient was taken to the operating room, débrided, and given a diverting colostomy to facilitate healing.

1956

PART II  CT and MR Imaging of the Whole Body

epididymo-orchitis, lymphoma, leukemia, and tuberculosis.119 Epidermoid cysts are benign intratesticular lesions known for their classic “onionskin” appearance on US. MR features of epidermoid cysts have not been frequently described in the literature, but signal characteristics seem to be T1 hypointensity and T2 hyperintensity with a rim of low T1 and T2 signal.26,112 Extratesticular lesions.  An overwhelming majority of extratesticular lesions are benign. The most common benign extratesticular masses are lipoma and adenomatoid tumor. Lipomas are well-defined nonenhancing rounded masses showing homogeneous hyperintense T1 and T2 signal on MRI. Lipomas can be differentiated from other lesions showing high T1 signal, such as hemorrhagic or proteinaceous cysts, through the use of fat-suppression or chemical shift imaging. Adenomatoid tumors are also common benign extratesticular tumors that most commonly originate in the tail of the epididymis. Adenomatoid tumors are typically T2 hypointense relative to the testis and do not enhance to the same degree as the testicular parenchyma on MRI.119

Malignant Diseases Testicular cancer.  Testicular cancers are the most common malignancy of young men (aged 20-34 years) and though rare have been increasing in worldwide incidence.134 A validated risk factor for testicular cancer is cryptorchidism; other factors such as atrophy, infertility, and inguinal hernia have been suggested.1,125 A genetic predisposition for testicular malignancy has been explored, but no definitive molecular defect has been identified. Approximately 95% of primary testicular malignancies are germ cell tumors (GCTs), with the remaining 5% divided into lymphomas (Fig. 56-21) and sex cord stromal tumors. Based on histology, GCTs are categorized as seminomas and nonseminomas, with differing  treatment regimens.78 Seminoma, the more common tumor type, is composed of purely seminomatous histologic features. Any nonseminomatous histologic feature categorizes the tumor as a nonseminoma regardless of whether seminomatous elements are present. The  broad spectrum of nonseminomatous tumors includes embryonal  carcinoma, teratocarcinoma, teratoma, and choriocarcinoma. The clinical presentation of testicular cancer is usually a palpable painless

testicular abnormality. The TNM staging system for testicular cancer is complicated. Surgical orchiectomy is the treatment of choice for the early stages of either type of malignancy; however, considerations for adjuvant therapy differ and depend on tumor staging. Stage I seminomatous disease can be followed with surveillance for lower-risk patients. Higher-risk stage I seminoma patients traditionally receive adjuvant retroperitoneal lymph node irradiation, although recent data have demonstrated a virtually equal and effective chemotherapeutic regimen.100 Lymph node involvement in stage II and III disease is treated with either radiotherapy or chemotherapy.21,54,85,133 Stage I nonseminoma management includes surveillance or further surgical  retroperitoneal lymphadenectomy, with or without adjuvant chemotherapy.24 Treatment of stage II and III nonseminoma is based on different chemotherapeutic regimens.22,77,120 Computed tomography of testicular cancer.  The standard for testicular cancer staging is CT. It is typically performed following orchiectomy to evaluate the extent of disease, particularly lymph node involvement. Lymph node drainage of the testicles is important in determining the pattern of disease spread. Seminomatous tumors spread via lymphatic routes, whereas nonseminomatous tumors metastasize via either lymphatic drainage or vascular channels. Owing to the abdominal embryologic origin of the testes, normal lymphatic drainage from the testis exits through the inguinal ring to the abdominal aortocaval lymph node stations; therefore these lymph nodes are typically the initial regions affected by metastatic disease.63 Right-sided GCTs drain primarily to the caval side and involve the precaval, paracaval, and interaortocaval lymph nodes between the levels of L1 and L3. Left-sided malignancies affect the lymph nodes on the aortic side because drainage into the preaortic and paraaortic groups follows the left testicular artery (Fig. 56-22). Although these anatomic drainage patterns typically hold true, right-to-left crossover of metastatic disease can occur but only subsequent to ipsilateral lymphatic disease. Minor additional drainage pathways can exist for the testes. Though uncommon, metastatic drainage to the internal, external, and common iliac lymph nodes, which run adjacent to the epididymal and cremasteric vessels, is well documented. Inguinal lymph nodes are not regional drainage stations for the testes and hence are not involved in metastatic

GE

A

B FIG 56-21  Testicular lymphoma. A, Hypoechogenicity is present in the left testis (arrow). B, Axial CT scan demonstrates multiple bilateral enlarged lymph nodes (arrows) in the internal iliac lymphatic chain. Biopsy demonstrated lymphoma.

CHAPTER 56  Male Pelvis

1957

B

A

FIG 56-22  Metastatic seminoma. A, US of the left testis is normal. B, Enlarged rounded lymph node in the left paraaortic chain (arrow). Pathology demonstrated scar in the left testis, consistent with burned-out seminoma with metastatic lymphatic disease.

disease unless retrograde spread from affected iliac lymph nodes is present. In addition to regional lymph nodes, distant lymph nodes can be evaluated with CT. Mediastinal disease and pulmonary manifestations can be seen as sequential lymph node metastases in seminomas or as skip lesions in nonseminomas. Paratesticular malignancies can also be identified on pelvic CT imaging, and lesions within the spermatic cord can be easily identified (Fig. 56-23). Magnetic resonance imaging of testicular cancer.  Although US is the preferred modality for evaluation of a palpable testicular abnormality, testicular lesions can be encountered on MRI of the pelvis. Intratesticular masses are well visualized on T2 images because of their lower signal relative to the high signal of the surrounding testicular parenchyma. Generally, seminomas are homogeneous and nonseminomas are more heterogeneous on MRI. Seminomas are typically hypointense to the testis on T2 and isointense on T1 sequences, whereas nonseminomas are isointense to hyperintense on T1 and hypointense on T2 relative to testicular parenchyma.82 MRI can in fact characterize and locally stage testicular neoplasms in the preoperative state with high accuracy.12 Traditionally CT has been the preferred modality for screening of recurrent testicular cancer in the form of retroperitoneal adenopathy. However, MRI can also be used for this purpose. Specific MR protocols can be implemented to screen for retroperitoneal and pelvic adenopathy. In particular, DWI has proven to be a useful sequence for detecting lymph nodes. Discrimination of pathologically affected lymph nodes is restricted by size thresholds, the same limitations that affect CT. Studies comparing the accuracy of CT and MRI in identifying lymphadenopathy demonstrate comparable results.117 A study investigating the use of superoxide nanoparticles with MRI for the detection of lymph node metastases demonstrated a superior sensitivity and specificity compared with traditional MRI alone.132 These data have not been validated in randomized prospective trials. However, the results are promising.

FIG 56-23  Liposarcoma in the left spermatic cord. Axial CT scan demonstrates soft tissue density in the left inguinal canal (arrow). Pathology was consistent with liposarcoma.

Lymphoma.  Whereas primary testicular cancer is the most common malignancy in young men, lymphoma is the most common testicular neoplasm in elderly men. Lymphoma involving the testes may present as the primary manifestation of the disease but is much more commonly encountered in the setting of known extratesticular disease. Most instances of lymphoma involving the testis are of the

1958

PART II  CT and MR Imaging of the Whole Body

diffuse large B-cell non-Hodgkin’s variety. As with lymphoma in other locations, it typically demonstrates low T1 and T2 signal, with an infiltrative pattern and low-level enhancement on MRI. Lymphoma should be considered when both the testis and epididymis demonstrate this infiltrative pattern.119 Extratesticular malignancy.  Primary malignant extratesticular tumors predominantly involve the spermatic cord. The most common spermatic cord malignancy is a sarcoma, either rhabdomyosarcoma or liposarcoma. Rhabdomyosarcomas occur more commonly in children. Otherwise, spermatic cord sarcomas typically present in older men as large complex heterogeneously enhancing masses. On MRI, they commonly demonstrate hyperenhancement and heterogeneous signal due to a combination of hemorrhage and necrosis. Fat is seen in up to 80% of liposarcomas.119 Less than 10% of malignant extratesticular masses represent metastases, most commonly from prostate, renal, gastric, colon, ileal carcinoid, and pancreatic primary carcinomas.68

PENIS Anatomy The penis is composed of three cylindrical shafts—two corpora cavernosa and one corpus spongiosum. The corpora cavernosa, which lie along the dorsal penile surface and are invested with multiple chambers with sinusoidal spaces and anastomotic channels, are chiefly responsible for erection. The corpus spongiosum houses the urethra and becomes the glans penis distally and the penile bulb at its root. All three corpora are surrounded by a fibrous capsule of tunica albuginea and then two fascial layers (deep Buck’s fascia and superficial dartos fascia).

Imaging Techniques As in the case of the scrotum, US is the preferred imaging modality of the penis, owing to its tolerability, availability, and capability of Doppler imaging. However, the role for penile MRI is increasing in the setting of trauma and cancer staging because of its exquisite depiction of the individual fascial layers, which can help direct surgical repair.96 MRI should be performed with the penis in anatomic position (taped to the abdomen) with a small FOV and multiplanar DCE imaging.

A

Normal Findings.  On MRI the corpora demonstrate intermediate signal on T1-weighted images and high signal on T2-weighted images (see Fig. 56-18). The tunica albuginea and Buck’s fascia are hypointense on T1- and T2-weighted sequences, whereas dartos fascia is slightly T2 hyperintense relative to Buck’s fascia. Both the corpora cavernosa and the corpus spongiosum enhance with gadolinium administration. The corpus spongiosum enhances immediately, whereas the corpora cavernosa enhance in a centrifugal fashion owing to the location of the cavernosal artery.

Benign Conditions Trauma.  Penile injuries are often diagnosed clinically. The most common concern is urethral injury, which is best evaluated with retrograde urethrography under fluoroscopy. Penile fractures can result from traumatic rupture of the stiff tunica albuginea and corpora  cavernosa in an erect penis exposed to extreme lateral force or blunt trauma, often during sexual intercourse. The clinical diagnosis is usually straightforward; however, dorsal vein rupture may present similarly, and MRI can be useful. Regarding penile fractures, disruption of the involved corpora and tunica albuginea can be seen on MRI, which demonstrates high-intensity T1 signal, consistent with hemorrhage (Fig. 56-24). Management for penile fractures is surgical because conservative treatment has resulted in associated morbidities such as erectile dysfunction.73,90

Malignant Diseases Penile malignancy.  Penile cancer is a rare disease in developed nations but has a higher prevalence in developing countries. Histologically the majority (>95%) of penile cancers are squamous cell cancers. The remaining cases are melanoma, sarcoma, basal cell carcinoma, and lymphoma.28,102 Nearly half the cases of squamous cell carcinoma affect the glans penis. Other involved regions include the prepuce, coronal sulcus, and shaft.28 Metastatic disease to the penis is exceedingly rare but can occur; the primary tumor usually arises from the urogenital tract. TNM staging is used in penile cancer (Fig. 56-25). Penile cancer usually spreads through lymphatic channels. Buck’s fascia typically acts as a barrier and prevents local involvement of the corpora. Specific lymph node involvement depends on the location of the primary

B FIG 56-24  Penile fracture. T1-weighted penile long-axis (A) and axial (B) MRIs show a break within the left corpus cavernosum, with a hyperintense focus (arrows) extending beyond the tunica albuginea and representing hemorrhagic products.

CHAPTER 56  Male Pelvis

1959

BF

T1

DF

T3

T2

T4

Urethra

Tis

FIG 56-25  Penile cancer T staging. BF, Buck’s fascia; DF, dartos fascia.

tumor, owing to the normal lymph node drainage pathways. The superficial inguinal lymph nodes drain the skin of the penis and prepuce, the deep inguinal and external iliac lymph nodes drain  the glans penis, and the internal iliac lymph nodes drain the erectile tissue and urethra. Owing to crossover of the lymphatic channels, unilateral disease may result in bilaterally affected lymph nodes. In addition, pelvic lymphadenopathy is invariably preceded by inguinal lymphadenopathy. The most important prognostic and treatment-related factor is the presence and degree of nodal disease. The traditional clinical evaluation consists of physical examination with palpation for inguinal lymphadenopathy. However, because of its superior soft tissue resolution, MRI is the preferred method to evaluate the extent of local disease invasion and potentially the status of regional lymph node metastases. Local extent of disease is best seen on T2-weighted and T1-weighted postgadolinium images, which allow assessment of transgression into the corpora or the overlying fascial layers. Primary tumors enhance with gadolinium and are typically hypointense relative to the corpora on T1- and T2-weighted images (Fig. 56-26). Metastatic disease often appears as multiple discrete lesions within the corpora cavernosa and corpus spongiosum that are hypointense to normal surrounding tissue on T1- and T2-weighted sequences4 (Fig. 56-27). Determining lymph node involvement remains a difficult task despite the use of MRI, which allows visualization of all pelvic lymph nodes and is an improvement over dependence on physical examination alone. Gross appearance and size measurements are limited in sensitivity because occult metastases are not identified on crosssectional imaging. Because treatment with prophylactic inguinal lymphadenectomy has demonstrated an improvement in long-term survival, there are proponents for this surgical procedure, despite significant associated morbidity such as severe lymphedema and reconstruction complications. A new MRI technique using lymphotropic

FIG 56-26  Penile malignancy. Postgadolinium T1-weighted image of the penis demonstrates a hyperintense lesion (arrow) along the dorsal aspect of the penis arising from the corpus cavernosum.

nanoparticle technology, as described previously, may be more useful in identifying involved lymph nodes.

BLADDER AND URETERS Anatomy The urinary bladder is a primarily extraperitoneal distensible viscus composed mostly of a detrusor muscle wall and lined with loose connective tissue. The superior aspect of the bladder, or bladder dome, is covered by peritoneal lining. The bladder contains four distinct layers: the mucosa (thin urothelium), submucosa (lamina propria), muscularis propria (smooth muscle detrusor), and the outermost serosa. The distal ureters course from the abdomen into the pelvis anterior to the common iliac artery bifurcation. The bilateral ureters extend posteriorly before angling medially at the lateral aspect of the bladder to insert at the posterior base within the triangular trigone. The bladder is easily seen on T2-weighted images when filled with urine. Contrast-enhanced CT in the delayed phase is optimal in assessing the bladder and distal ureters.

Imaging Techniques The initial step in the imaging evaluation of a suspected bladder tumor or other urothelial lesion usually consists of a multiphase CT urogram. This protocol typically consists of a noncontrast acquisition (to evaluate for stones or hemorrhage), a nephrographic phase at about 90 seconds after IV contrast injection (to evaluate for solid renal mass), and a delayed phase at 10 to 15 minutes following contrast injection (to evaluate for urothelial masses, often presenting as filling defects within opacified collecting system/ureter/bladder). At our institution we reduce the number of acquisitions by using a splitbolus technique in which the nephrographic and delayed phases are obtained in a single acquisition following two injections of the normal

1960

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 56-27  Penile metastatic disease. A, Axial T2-weighted MRI shows multiple hypointense lesions in the corpora cavernosa (arrows). B, Sagittal image demonstrates hypointensities along the length of the corpora cavernosa (arrows). Metastatic disease was presumed to be from known prostate cancer.

contrast bolus (split in half). See Appendix C for our split-bolus CT urogram protocol. CT urography has demonstrated over 90% sensitivity and specificity in the detection of bladder cancer in patients with hematuria.127 MRI is also used for more detailed characterization of bladder cancer staging because of its high soft tissue contrast resolution, which allows for more accurate depiction of the individual bladder wall layers. In fact MR of the bladder has been shown to be more accurate in the staging of bladder carcinoma than CT.37 MRI has also demonstrated superiority in distinguishing intramural invasion and extravesical extension.45 The multiplanar functionality of MRI limits partial volume averaging while optimizing characterization of the bladder wall, regardless of tumor location. Coronal images can be particularly useful when assessing tumors along the lateral bladder wall; sagittal images are helpful in evaluating the anterior and posterior bladder wall as well as the bladder dome.45 DWI offers another tool for lesion detection and cancer staging on MRI. In both CT and MRI of the bladder, appropriate bladder distension is critical in the accurate characterization of bladder wall thickness. Underdistension causes a false appearance of bladder wall thickening that can lead to overcalling of potential pathology. Meanwhile, overdistension can cause patient discomfort, leading to patient motion with subsequent artifact (particularly on MRI). Flat or sessile tumors can also be overlooked when the bladder is overdistended. Optimal bladder distension is usually achieved by having the patient void 2 hours before the exam and then not again until the study is completed.

Normal Findings.  On T1-weighted MRI the urine within the bladder lumen is low in signal, the bladder wall is intermediate in signal, and the surrounding perivesical fat is high in signal. On T2-weighted MRI the urine is high in signal, the bladder wall is low in signal, and the surrounding fat is high in signal (on non–fat-saturated sequences). The detrusor muscle layer (muscularis propria) demonstrates low T2 signal. Loss of this normal hypointense T2 signal in the setting of bladder cancer defines T2 disease, and extension of a tumor through the detrusor into the perivesical fat denotes T3 disease.

Benign Conditions Trauma.  In the case of suspected bladder injury, a fluoroscopic static cystogram or CT cystogram can detect active contrast extravasation, indicating bladder rupture. An important determination is whether the rupture is intraperitoneal or extraperitoneal, because management is either surgical or conservative, respectively. Intraperitoneal rupture typically occurs at the dome of a urine-filled bladder secondary to increased intracystic pressure. Contrast medium is  characteristically seen within the peritoneum and mesenteric leaves, bathing the bowel loops (Fig. 56-28). Extraperitoneal bladder rupture is often a result of puncture or laceration from adjacent fractured bone, and contrast opacification of the retropubic space and fascial planes is typically present (Fig. 56-29). Urethral injuries can also occur in the setting of trauma, but retrograde urethrography rather than CT or MRI is the modality of choice for optimal evaluation. Stones.  Stones within the bladder or distal ureter are easily detected on CT. They present as dense rounded opacities on noncontrast images and as rounded filling defects on the delayed excretory phase. In very rare instances (e.g., indinavir stones), calculi can be radiolucent on CT. On MRI, stones will appear as rounded areas of dark T1 and T2 signal. Stones are much better characterized on CT. Masses.  Although the majority of bladder masses are malignant, benign masses are occasionally encountered within the bladder. These include papillomas, adenomas, leiomyomas, fibromas, hemangiomas, and paragangliomas.40 Leiomyomas are the most common benign bladder tumor but are predominantly seen in women. Fibroepithelial polyps are rare benign bladder lesions that are frequently mistaken for transitional cell carcinomas. They are most commonly seen in young adult males with painless hematuria and typically arise in the ureter. On CT or MRI, fibroepithelial polyps usually appear as wellcircumscribed sessile or frondlike masses measuring less than 5 cm.40 Paragangliomas in the bladder are extremely rare but can be seen as a manifestation of the 10% of pheochromocytomas that are extraadrenal in origin. On CT or MRI, bladder paragangliomas present as smooth intramural or submucosal masses demonstrating avid contrast enhancement. Peripheral calcification can be seen on CT, which is

CHAPTER 56  Male Pelvis

A

1961

B FIG 56-28  Intraperitoneal bladder rupture. A, Axial CT scan at the level of the kidneys demonstrates intraperitoneal bladder rupture, with contrast medium clearly seen in the peritoneal cavity and between the mesenteric leaves (red arrows). B, Axial CT scan at the level of the bladder (marked by the presence of a Foley catheter) demonstrates contrast extravasation within the pelvis (red arrow) and the likely site of perforation (yellow arrow).

FIG 56-29  Extraperitoneal bladder rupture. Axial CT scan shows contrast medium within the bladder as well as within the extraperitoneal space (arrows), consistent with bladder rupture.

highly suggestive of paraganglioma. MRI is even more sensitive and specific for characterization of a bladder paraganglioma, typically showing low T1 signal and intermediate to high T2 signal.40

Malignant Diseases Bladder cancer.  Bladder cancer is the fourth most common malignancy in men and is predominantly a disease of older men in the sixth and seventh decades of life. Cigarette smoke is a common causative factor, as well as urban living and certain occupations. The most common clinical presentation is painless hematuria that ultimately results in urinalysis, cystoscopy, and biopsy. Histologically the majority

of bladder cancers are transitional cell carcinomas, but a small percentage are squamous cell carcinomas, adenocarcinomas (arising in urachal remnants), and other much less common entities. Most bladder malignancies are thought to be secondary to chronic bladder irritation (e.g., chronic stones, occupational toxins) or infection. The TNM staging of bladder cancer is detailed in publications of the American Joint Committee on Cancer.27 T staging determines disease management (Fig. 56-30). Ta and T1 disease—in situ disease that has not spread beyond the subepithelial connective tissues— requires no further imaging after cystoscopy and biopsy. However, continued workup is recommended if the tumor extends into the bladder wall muscle (T2 disease). CT urography is typically the initial study of choice for bladder cancer staging. This allows for detection of stones, renal masses, and urothelial lesions in a patient presenting with hematuria. On CT, bladder lesions are seen either as enhancing masses in a background of hypodense urine or as filling defects in a background of the contrast-opacified bladder, often in the trigone region and posterolateral wall (Fig. 56-31). MRI findings are similar; the bladder mass is seen against a background of T2 bright urine (Fig. 56-32). A key anatomic consideration is whether there is overt transgression and macroscopic invasion through the bladder wall (stage T3b). On CT, soft tissue attenuation in the hypodense perivesical region is often the first sign of serosal penetration (Fig. 56-33). Involvement of and extension through the detrusor has been shown to be more accurately characterized by MRI.45 T1-weighted MRI is useful for evaluating intermediate-signal tumor that has spread into the highsignal perivesical fat. Traditional T2-weighted images demonstrate intermediate signal from the tumor in a background of high-signal urine (see Fig. 56-33). The morphology of bladder tumors can range from plaquelike to papillary in appearance (Fig. 56-34; also see Fig. 56-30). Calcifications in transitional cell carcinomas are relatively uncommon but have been reported in 5% of cases; they are relatively coarse and typically nodular or curved in appearance on CT89 (Fig. 56-35). As bladder malignancies become more advanced, they can spread circumferentially around the bladder wall (Fig. 56-36) and can eventually result in hydroureter if the ureteral insertion is affected. Because transitional cell carcinomas are commonly multifocal, synchronous lesions are often identified

1962

PART II  CT and MR Imaging of the Whole Body

T1

T2

T3a

T3b

T4

Prostate

FIG 56-30  Bladder cancer T staging.

A

B FIG 56-31  Polypoid bladder cancer. A, Axial enhanced CT scan in the early phase shows subtle density (arrow) along the right bladder wall. B, Presence of a polypoid mass (arrow) is confirmed by the delayed image during bladder contrast opacification. Pathology staged the tumor as T1.

CHAPTER 56  Male Pelvis

1963

B

A

FIG 56-32  Bladder cancer. Axial T2-weighted (A) and T1 fat-saturated (B) MRIs depict a hypointense lesion along the left trigone (arrow) without extracapsular extension into the perivesicular fat. Pathology staged the tumor as T2.

(Figs. 56-37 and 56-38). Bladder malignancy can also arise from a bladder diverticulum (Fig. 56-39), in which case the cancer is often more advanced at diagnosis, because extravesical spread is facilitated by thinning of the muscularis at the diverticulum. Evaluation of metastatic lymph node spread is limited by size criteria for pathologic lymph nodes. Lymphotropic nanoparticle MRI (see “Lymph Nodes”) has a high specificity and sensitivity in identifying diseased lymph nodes (Fig. 56-40).

LYMPH NODES Anatomy Lymph node drainage of the pelvis is important for evaluating metastatic disease and spread because it is factored into TNM staging of neoplasms. Understanding the normal lymphatic drainage pattern allows proper evaluation of metastatic disease. The pelvic nodal groups are classified in relation to the accompanying vessel and are composed of the paraaortic, common iliac, internal iliac, external iliac, deep inguinal, and superficial inguinal stations (Fig. 56-41). The obturator lymph nodes are part of the internal iliac chain and run along the medial border of the obturator internus muscle. The superficial inguinal lymph nodes drain the perineum but not the true pelvis. The generally accepted imaging criterion for pathologically involved lymph nodes is size greater than 1 cm in short-axis diameter. However, lymph nodes less than 1 cm are often reported if there is a history of malignancy. Identification of pelvic lymph nodes can be challenging on non– contrast-enhanced CT studies owing to their proximity to bowel and vascular structures. With the addition of contrast media, both oral and intravenous, lymph nodes are more apparent and can be distinguished from a background of enhancing structures. The unenhanced CT scan in Figure 56-42 demonstrates a soft tissue mass in the left pelvic  sidewall area that was thought to be an enlarged lymph node; it was suspicious for malignancy, and the patient was referred for biopsy. Subsequent MRI demonstrated flow voids and contrast opacification consistent with an arteriovenous malformation, a lesion with potentially catastrophic consequences if biopsied. When lymph nodes are

correctly identified, biopsy under CT guidance can be performed, usually with a posterior approach (Fig. 56-43).

Imaging Techniques Imaging of lymph nodes is considered separately because of the importance of detecting pathologic lymph nodes in the pelvis, especially in oncologic evaluation. Bipedal lymphangiography was traditionally used to evaluate for metastatic disease by assessing internal lymph node abnormality; however, it has been supplanted by cross-sectional imaging.64 CT diagnosis of lymphadenopathy is based mainly on size criteria. Any pelvic lymph node with a maximum short-axis diameter exceeding 10 mm or any obturator or internal iliac lymph node exceeding 8 mm in diameter is considered pathologic.43,91 Secondary findings that increase the suspicion of a metastatic lymph node are loss of a central fatty hilum and loss of the normal oval or round shape. Lymph node size is a fairly insensitive measure for determining the presence of metastatic disease; it is well established in the literature that “normal-sized” lymph nodes can contain micrometastases. The specificity of enlarged lymph nodes is compromised by increased lymph node size from reactivity secondary to infectious or inflammatory states.31 This tradeoff between sensitivity and specificity is well appreciated, and the aforementioned size thresholds are believed to best maximize the positive predictive value.118,129 MRI is just as inadequate as CT in assessing for metastatic lymph nodes when considering size criteria.8 On MRI the minimum shortaxis diameter of lymph nodes is slightly superior to the maximum short-axis diameter, with an accuracy of 93%, sensitivity of 62.2%, and specificity of 97.9%.58 The 10-mm size criterion for abnormal lymph nodes on CT has also been validated for MRI; in addition, 8-mm rounded lymph nodes are thought to be abnormal.33 The advantage of MRI over CT is the ability to assess the inherent internal signal characteristics of individual lymph nodes in addition to measuring size. Although MRI examination times are lengthy with the acquisition of pre- and postgadolinium T1-weighted sequences, T2-weighted sequences with and without fat saturation, and coronal and sagittal images, important additional information can be gained. The presence Text continued on p. 1968

1964

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 56-33  Bladder cancer with extracapsular extension. A and B, Axial CT scan during early (A) and delayed (B) phases demonstrates an enhancing lesion in the left bladder (arrow) with extension into the left perivesical fat, consistent with extracapsular transgression. C, Enlarged left iliac lymph node (arrow) seen on CT scan was positive for metastatic disease on biopsy. D, Perivesical fat involvement (arrow) is also seen on T2-weighted MRI.

CHAPTER 56  Male Pelvis

1965

FIG 56-34  Plaquelike bladder mass. Coronal contrast-enhanced image during the delayed phase demonstrates a plaquelike mass (arrow) along the left bladder, which was biopsy proven as malignant.

FIG 56-35  Calcification within bladder cancer. Axial CT scan shows calcification along the right bladder (arrow), adjacent to bladder wall thickening. Pathology confirmed the presence of transitional cell cancer, which can be associated with calcifications.

FIG 56-36  Circumferential bladder cancer. A bladder mass (arrows) can grow in a circumferential manner. Progressive encasement of the bladder wall, with involvement of the ureteral insertions, can result in obstruction and hydronephrosis.

1966

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 56-37  Multicentric bladder cancer. A, Axial contrast-enhanced CT scan identifies multiple masses (arrows) within the bladder. B, Additional lesions (arrows) are seen, with focal calcification (arrowhead) overlying a left bladder lesion.

R

L

FIG 56-38  Transitional cell cancer of the distal ureter. T1-weighted postgadolinium MRI shows an enhancing lesion (arrowhead) in the distal ureter.

CHAPTER 56  Male Pelvis

B

A

FIG 56-39  Development of bladder malignancy within a diverticulum. A, Initial axial CT scan shows a right lateral bladder wall diverticulum (arrow). B, Follow-up CT scan for symptomatic hematuria shows thickening and a mass (arrows) within the diverticulum.

1 1

2 2

A

B FIG 56-40  Metastatic bladder cancer. A, Precontrast T2*-weighted MRI shows left external iliac lymph nodes (arrows). B, Postlymphotropic nanoparticle contrast-enhanced T2*-weighted MRI demonstrates lack of normal signal loss, suggesting metastatic infiltration, which was confirmed by histopathology. Note normal darkening of the benign lymph node on the contralateral side (red arrow). Yellow arrows show the left external iliac lymph nodes.

1967

1968

PART II  CT and MR Imaging of the Whole Body

Aortic Aortic

Common iliac

Common iliac

Presacral

Presacral

External iliac

External iliac

Internal iliac

Internal iliac

Obturator

Obturator

Inguinal

Inguinal

A

B FIG 56-41  Lymph node stations of the pelvis in the anteroposterior (A) and lateral (B) positions.

of central necrosis is a predictive feature of metastatic lymph nodes.91 Furthermore, DCE sequences can demonstrate wash-in and washout characteristics of suspicious lymph nodes. Results have been mixed, but it has been suggested that the altered internal environment of metastatic lymph nodes changes several parameters of contrast flow dynamics.41,92 In addition, DWI has added a new tool for both nodal detection and characterization. Lymph nodes can be more easily detected on DWI than on other conventional pulse sequences because the highly cellular composition of lymphoid tissue causes hyperintense signal. Furthermore, quantitative measurements of ADC values has potential in identifying malignant nodal infiltration.95 An emerging MRI technique that has shown promise in the  evaluation of nodal metastatic disease is lymphotropic nanoparticle– enhanced MRI. It has been evaluated in numerous malignancies, including head and neck, breast, lung, and pelvic  cancers.5,10,23,25,42,48,49,52,60,106,108,131 As an imaging modality, lymphotropic nanoparticle MRI is more akin to positron emission tomography (PET) than traditional CT and MRI, in that it is a physiologic and functional evaluation of disease that provides additional anatomic information. Injected ferumoxtran-10, an ultrasmall superparamagnetic iron oxide nanoparticle coated with dextran, is actively phagocytosed by macrophages and causes a susceptibility change that results in a drop in T2 and T2* signal. These iron-containing macrophages circulate and populate the lymphatic system. In benign lymph nodes, these macrophages are deposited homogeneously; therefore signal dropout is due to the presence of ferrous contrast agent (Fig. 56-44). Lymph nodes with internal metastatic deposits prevent homogeneous distribution or completely exclude migrating macrophages, which results in lack of contrast agent and retention of either partial or highintensity T2 or T2* signal (Fig. 56-45). Gradient recalled echo (GRE)

T2* sequences are key for nodal characterization.128 In the pelvis, lymphotropic nanoparticle MRI has demonstrated excellent sensitivity and specificity—100% and 95.7%, respectively, for prostate cancer and 100% and 97%, respectively, for pelvic squamous cell carcinoma.25,108 One logistical downside is the need for pre- and postcontrast imaging on separate days. False-positive results have also been noted in the case of intranodal fibrosis, which can be caused by prior infection, radiation, or chemotherapy.

Normal Findings.  Normal lymph nodes are round or oval in shape and demonstrate a central fatty hilum. They are homogeneous and demonstrate intermediate signal on MRI. On T1-weighted images, normal lymph nodes are hyperintense to skeletal muscle and hypointense to fat. On T2-weighted images, they are hyperintense to skeletal muscle but only slightly hypointense to fat. In the pelvis, lymph nodes are often better visualized on T2-weighted sequences because they stand out compared to the signal of the adjacent musculature. Normal lymph nodes should be homogeneous in T2 signal. On both CT and MRI, lymph nodes also demonstrate homogeneous enhancement on postcontrast images. In addition, DWI can help identify lymph nodes because of the highly cellular composition of lymphoid tissue.

Benign Conditions.  Subcentimeter lymph nodes are frequently visualized in the pelvis, particularly in the inguinal nodal stations, in everyday nonpathologic situations. Lymph nodes begin to catch the attention of the interpreting radiologist once they approach a size of 10 mm in short-axis diameter. Conventionally a pelvic lymph node larger than 10 mm in short-axis diameter or an obturator or internal iliac lymph node larger than 8 mm in short-axis diameter is considered abnormal.43,91 Vinnicombe et al. proposed a more strict size threshold when

CHAPTER 56  Male Pelvis

A

B

C

D FIG 56-42  Imaging of a left pelvic mass. A, CT scan shows a left pelvic mass (arrow) suspicious for metastatic malignancy. The patient was referred for biopsy. B, Axial T2-weighted MRI shows hypointensity in the left pelvic region (arrow). C, Coronal T1-weighted postgadolinium image demonstrates an enhancing vascular structure (arrow) consistent with an arteriovascular malformation (AVM). D, Three-dimensional reconstruction of the AVM (arrow) showing its relationship to pelvic vessels. Scheduled biopsy was aborted.

1969

1970

PART II  CT and MR Imaging of the Whole Body

FIG 56-43  CT-guided lymph node biopsy. Enlarged retroperitoneal lymph nodes can be biopsied under CT guidance. The radiopaque needle (yellow arrow) targets a suspicious lymph node (red arrow).

A

B FIG 56-44  Lymphotropic nanoparticle MRI demonstrating a benign lymph node. A, Precontrast axial T2*weighted image shows a hyperintense right external iliac lymph node (arrow). B, The lymph node (arrow) loses signal on the postcontrast T2*-weighted image, signifying a lack of tumor infiltration. This was confirmed on biopsy.

1 1

A

B FIG 56-45  Lymphotropic nanoparticle MRI demonstrating a malignant lymph node from metastatic prostate cancer. A, Precontrast axial T2*-weighted image shows a hyperintense left internal iliac lymph node (arrow). B, The node (arrow) retains signal on the postcontrast T2*-weighted image, consistent with malignancy. This was confirmed on biopsy.

1971

CHAPTER 56  Male Pelvis evaluating lymph nodes in the pelvis. She ascertained that the maximum short-axis diameter of normal lymph nodes should not exceed 7 mm for internal iliac, 8 mm for obturator, 9 mm for common iliac, and 10 mm for external iliac nodal stations.126 Lymph nodes exceeding these size thresholds can be encountered in the pelvis in the setting of several benign conditions, including suppurative lymphadenitis, HIV infection, Castleman’s disease, tuberculosis, and inflammatory bowel disease. Suppurative lymphadenitis can present as lymph nodes increased in size and/or number in the right lower quadrant secondary to Yersinia enterocolitica infection or along the iliac or inguinal chains related to Staphylococcus aureus infection. Adenopathy is the most common presentation of both HIV and tuberculosis in the pelvis; it can be associated with both infectious and neoplastic processes. Castleman’s disease presents very rarely in the pelvis. Inflammatory bowel disease can demonstrate reactive adenopathy in the mesocolic or mesorectal nodal stations along the inferior mesenteric artery distribution.

Malignant Diseases Metastatic disease.  Malignant lymphadenopathy in the pelvis is most commonly encountered in the setting of metastatic disease. Most commonly, malignant nodal disease is related to metastatic prostate cancer, although it can be seen in men with less common neoplasms like penile or seminal vesicle cancer. In prostate cancer, metastatic adenopathy will occur in the usual route of lymphatic drainage of the prostate gland, typically involving the obturator and external iliac lymph nodes, the internal iliac and presacral lymph nodes, and ultimately extending superiorly into the retroperitoneum within the paraaortic and paracaval nodal stations.79 On both CT and MRI, characterization of lymph nodes based on maximum short-axis diameter is widely accepted, although it is well known that enlarged lymph nodes can be benign and that subcentimeter lymph nodes can harbor metastatic disease. Owing to its superior soft tissue resolution, MRI can further interrogate lymph node composition through variations in its inherent signal intensity and DCE patterns. Infiltration of a lymph node with metastatic disease can lead to inhomogeneity of its T2 signal. Development of necrosis within a node leads to a focal decrease in T1 signal and increase in T2 signal. T2 signal can also help detect the presence of recurrent disease in metastatic nodes, since treated nodes will typically exhibit decreased

T2 signal due to posttreatment fibrosis. It should be noted, however, that treated lymph nodes can remain T2 hyperintense for up to a year following successful treatment and that hyperintense signal could represent not only residual disease but also reactive inflammation.88 Pattern of lymph node enhancement can also be helpful in characterization. Although homogeneous enhancement is nonspecific, suspicion for a metastatic node can be raised when a node displaying peripheral rim enhancement and central necrosis is encountered. Additionally, metastatic nodes often show rapid contrast uptake and often mimic the enhancement pattern of the primary tumor.88 As with evaluation of primary tumors (e.g., prostate cancer), DWI can also be used to characterize lymph nodes on MRI. Although normal nodes show restricted diffusion because of their highly cellular composition, the degree of restriction in concert with nodal morphology may help in preoperative assessment of nodal staging, even for normal-sized nodes. One study reviewing such features in normal-sized nodes showed accuracies ranging from 64% to 79% and specificities from 79% to 85% in the detection of metastatic nodal disease.34 However, in addition to metastatic disease, it should also be noted that restricted diffusion can be seen in infectious or inflammatory processes.88 MR lymphangiography is a newer technique that might have  the most potential for distinguishing benign and malignant lymph nodes in the pelvis, because it is not dependent on nonspecific size criteria. As previously described, the injected ultrasmall superparamagnetic iron oxide nanoparticle (ferumoxtran-10) is phagocytosed by intranodal macrophages within benign lymph nodes, resulting in a drop in T2 and T2* signal.5,23,132 Metastatic nodes will completely exclude or inhomogeneously phagocytose the circulating nanoparticle, resulting in retention of either partial or high-intensity T2 or T2* signal (Fig. 56-45). Lymphoma.  Malignant adenopathy in the pelvis can also occur with lymphoma, most commonly the non-Hodgkin’s lymphoma subtypes. Affected lymph nodes are classically bulkier and more numerous compared to metastatic adenopathy in the pelvis and will appear in concert with widespread adenopathy elsewhere within the retroperitoneum, mesentery, and mediastinum.

APPENDIX A

Prostate MRI Protocol Name

Pulse Sequence

3-Plane localizer SSFSE axial SSFSE sagittal Asset cal BH T2 sagittal Diffusion axial T2 axial oblique T1 axial GRE T2 coronal oblique TEST DYNAMIC Dynacad

Gradient echo Spin echo Spin echo Gradient echo FSE-XL Spin echo FSE-XL Gradient echo FSE-XL SPGR SPGR

TE

TR

140 140

Minimum Minimum

130 Minimum 147 Minimum 130 Min full Min full

3384 3800 5084 170 4650

Frequency Direction Unswap A/P S/I A/P A/P R/L A/P A/P S/I R/L R/L

Phase Encode

Frequency Encode

128 192 192

256 384 384

192 128 192 160 192 128 128

384 128 384 384 384 128 128

1972

PART II  CT and MR Imaging of the Whole Body

APPENDIX B ENDORECTAL COIL PLACEMENT The endorectal coil should be placed with the blue line oriented anteriorly and fixed in position with balloon inflation. Scout localizer images are obtained in the axial and sagittal planes, which should be

reviewed by the radiologist to ensure optimal positioning of the coil. A, On axial localizer images, the “ears” should be symmetrically positioned at the 10:00 and 2:00 positions (arrows) such that signal from the coil is directly anteriorly. B, On sagittal localizer images, the coil should be positioned to cover the prostate base superiorly (including an enlarged median lobe) and apex inferiorly (arrows).

CHAPTER 56  Male Pelvis

APPENDIX C SPLIT BOLUS CT UROGRAM PROTOCOL HEMATURIA PROTOCOL Scan Overview Indications: IV contrast

assessment of stone, kidney, flank pain hematuria 370 mg/mL 120 mL 100 seconds

Contrast amount Scan delay

Series 1:

I- STONE

Topogram I- Stone I+ prone abd-pelvis

BILLING CODES: CTABPWWO

(supine)

Location

top of kidneys to S.P

Rotation Time

0.5

Care Dose kV

ON 4D 120

Pitch Direction

1.3 Craniocaudal

Slice Thickness Slice Increment

5mm 5mm

Kernel Window

B30F med smooth abdomen

32/24 x 1.2 160

FOV Recon Type

180

API Series Transf.

300 mm Axial Inspiration ON

ACQ Quality Ref. mAs

under 200 over 200

Inject 40 mL contrast, hand 250 mL bag of saline to gravity drip into the patient, once complete, place patient back on scan table prone, Topogram, then inject 80 mL of contrast @ 3 mL/sec

Series 2: kV mA

Series 3:

Prone Topogram 80 40 SCAN DELAY 100 seconds

I+ ABD-PELVIS

Recon 2: Thins

Location Care Dose

top of kidneys to S.P ON 4D

Kernel Window

kV

120

FOV

Slice Thickness Slice Increment

5mm 5mm

Recon Type Thickness

128/64 x 0.6 250 280

Interval send to PACS

Rotation Time

0.5

Pitch Direction Kernel Window FOV Recon Type API Series Transf.

0.75 Craniocaudal B30F med smooth abdomen 300 mm Axial Inspiration ON

MPR : Sagittal and Coronal Load into 3D

ACQ Quality Ref. mAs

Series 4:

under 200 over 200

B31F Abdomen skin to skin Axial 1 mm 0.6 mm no

select parallel ranges send to PACS Thickness Interval Kernel DFOVWindow

3 3 B31F skin to skin Abdomen

15 minute delays if ureters not visualized (same parameters)

1973

1974

PART II  CT and MR Imaging of the Whole Body

REFERENCES 1. Abolyosr A, et al: The management of penile fracture based on clinical and magnetic resonance imaging findings. BJU Int 96(3):373–377, 2005. 2. Akbar SA, et al: Multimodality imaging of paratesticular neoplasms and their rare mimics. Radiographics 23(6):1461–1476, 2003. 3. Akin O, et al: Incremental value of diffusion weighted and dynamic contrast enhanced MRI in the detection of locally recurrent prostate cancer after radiation treatment: Preliminary results. Eur Radiol 21(9):1970–1978, 2011. 4. Allaf ME, Partin AW, Carter HB: The importance of pelvic lymph node dissection in men with clinically localized prostate cancer. Rev Urol 8(3):112–119, 2006. 5. Amsellem-Ouazana D, et al: Negative prostatic biopsies in patients with a high risk of prostate cancer. Is the combination of endorectal MRI and magnetic resonance spectroscopy imaging (MRSI) a useful tool? A preliminary study. Eur Urol 47(5):582–586, 2005. 6. Anzai Y, et al: Dextran-coated superparamagnetic iron oxide, an MR contrast agent for assessing lymph nodes in the head and neck. AJNR Am J Neuroradiol 15(1):87–94, 1994. 7. Anzai Y, et al: Initial clinical experience with dextran-coated superparamagnetic iron oxide for detection of lymph node metastases in patients with head and neck cancer. Radiology 192(3):709–715, 1994. 8. Anzai Y, et al: Evaluation of neck and body metastases to nodes with ferumoxtran 10-enhanced MR imaging: Phase III safety and efficacy study. Radiology 228(3):777–788, 2003. 9. Asgari MA, et al: Penile fractures: Evaluation, therapeutic approaches and long-term results. J Urol 155(1):148–149, 1996. 10. Avery LL, Scheinfeld MH: Imaging of penile and scrotal emergencies. Radiographics 33(3):721–740, 2013. 11. Baco E, et al: Effect of targeted biopsy guided by elastic image fusion of MRI with 3D-TRUS on diagnosis of anterior prostate cancer. Urol Oncol 2014. 12. Bellin MF, Lebleu L, Meric JB: Evaluation of retroperitoneal and pelvic lymph node metastases with MRI and MR lymphangiography. Abdom Imaging 28(2):155–163, 2003. 13. Bellin MF, et al: Lymph node metastases: Safety and effectiveness of MR imaging with ultrasmall superparamagnetic iron oxide particles–Initial clinical experience. Radiology 207(3):799–808, 1998. 14. Beyersdorff D, et al: Patients with a history of elevated prostate-  specific antigen levels and negative transrectal US-guided quadrant or sextant biopsy results: Value of MR imaging. Radiology 224(3):701–706, 2002. 15. Bipat S, et al: Computed tomography and magnetic resonance imaging in staging of uterine cervical carcinoma: A systematic review. Gynecol Oncol 91(1):59–66, 2003. 16. Burgers JK, Badalament RA, Drago JR: Penile cancer. Clinical presentation, diagnosis, and staging. Urol Clin North Am 19(2):247–256, 1992. 17. Cassidy FH, et al: MR imaging of scrotal tumors and pseudotumors. Radiographics 30(3):665–683, 2010. 18. Caterino M, et al: Primary cancer of the urinary bladder: CT evaluation of the T parameter with different techniques. Abdom Imaging 26(4):433–438, 2001. 19. Cho JH, et al: Sonographic and MR imaging findings of testicular epidermoid cysts. AJR Am J Roentgenol 178(3):743–748, 2002. 20. Choi YJ, et al: Functional MR imaging of prostate cancer. Radiographics 27(1):63–75, discussion 75–7, 2007. 21. Chung PW, et al: Stage II testicular seminoma: Patterns of recurrence and outcome of treatment. Eur Urol 45(6):754–759, discussion 759–60, 2004. 22. Classen J, et al: Radiotherapy for stages IIA/B testicular seminoma: Final report of a prospective multicenter clinical trial. J Clin Oncol 21(6):1101–1106, 2003. 23. Claus FG, Hricak H, Hattery RR: Pretreatment evaluation of prostate cancer: Role of MR imaging and 1H MR spectroscopy. Radiographics 24(Suppl 1):S167–S180, 2004.

24. Coakley FV, et al: Brachytherapy for prostate cancer: Endorectal MR imaging of local treatment-related changes. Radiology 219(3):817–821, 2001. 25. Costouros NG, et al: Diagnosis of prostate cancer in patients with an elevated prostate-specific antigen level: Role of endorectal MRI and MR spectroscopic imaging. AJR Am J Roentgenol 188(3):812–816, 2007. 26. Curran S, et al: Endorectal MRI of prostatic and periprostatic cystic lesions and their mimics. AJR Am J Roentgenol 188(5):1373–1379, 2007. 27. Demuren OA, Koriech O: Isolated penile metastasis from bladder carcinoma. Eur Radiol 9(8):1596–1598, 1999. 28. Dennis MA, Donohue RE: Computed tomography of prostatic abscess. J Comput Assist Tomogr 9(1):201–202, 1985. 29. de Rooij M, et al: Accuracy of multiparametric MRI for prostate cancer detection: A meta-analysis. AJR Am J Roentgenol 202(2):343–351, 2014. 30. de Wit R, et al: Equivalence of three or four cycles of bleomycin, etoposide, and cisplatin chemotherapy and of a 3- or 5-day schedule in good-prognosis germ cell cancer: A randomized study of the European Organization for Research and Treatment of Cancer Genitourinary Tract Cancer Cooperative Group and the Medical Research Council.   J Clin Oncol 19(6):1629–1640, 2001. 31. Dighe MK, Bhargava P, Wright J: Urinary bladder masses: techniques, imaging spectrum, and staging. J Comput Assist Tomogr 35(4):411–424, 2011. 32. Dixon AK, Ellis M, Sikora K: Computed tomography of testicular tumours: Distribution of abdominal lymphadenopathy. Clin Radiol 37(6):519–523, 1986. 33. Engelbrecht MR, et al: Local staging of prostate cancer using magnetic resonance imaging: A meta-analysis. Eur Radiol 12(9):2294–2302, 2002. 34. Fischbein NJ, et al: Assessment of metastatic cervical adenopathy using dynamic contrast-enhanced MR imaging. AJNR Am J Neuroradiol 24(3):301–311, 2003. 35. Fleshner N, Warde P: Controversies in the management of testicular seminoma. Semin Urol Oncol 20(4):227–233, 2002. 36. Fritzsche PJ, et al: Undescended testis: Value of MR imaging. Radiology 164(1):169–173, 1987. 37. Fukuda H, Nakagawa T, Shibuya H: Metastases to pelvic lymph nodes from carcinoma in the pelvic cavity: Diagnosis using thin-section CT. Clin Radiol 54(4):237–242, 1999. 38. Futterer JJ, et al: Prostate cancer: comparison of local staging accuracy of pelvic phased-array coil alone versus integrated endorectal-pelvic phased-array coils. Local staging accuracy of prostate cancer using endorectal coil MR imaging. Eur Radiol 17(4):1055–1065, 2007. 39. Gill B, Kogan S: Cryptorchidism. Current concepts. Pediatr Clin North Am 44(5):1211–1227, 1997. 40. Grey A, et al: The diagnostic accuracy of MRI PI-RADS scoring in a transperineal prostate biopsy setting. BJU Int 2014. 41. Guermazi A, et al: Lymphography: An old technique retains its usefulness. Radiographics 23(6):1541–1558, discussion 1559–60, 2003. 42. Hara N, et al: Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a useful modality for the precise detection and staging of early prostate cancer. Prostate 62(2):140–147, 2005. 43. Harisinghani MG, et al: MR imaging of lymph nodes in patients with primary abdominal and pelvic malignancies using ultrasmall superparamagnetic iron oxide (Combidex). Acad Radiol 5(Suppl 1):S167–S169, discussion S183–4, 1998. 44. Harisinghani MG, et al: MR lymphangiography for detection of minimal nodal disease in patients with prostate cancer. Acad Radiol 9(Suppl 2):S312–S313, 2002. 45. Harisinghani MG, et al: Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 348(25):2491– 2499, 2003. 46. Harisinghani MG, et al: A pilot study of lymphotrophic nanoparticleenhanced magnetic resonance imaging technique in early stage testicular cancer: A new method for noninvasive lymph node evaluation. Urology 66(5):1066–1071, 2005. 47. Heijmink SW, et al: Prostate cancer: body-array versus endorectal coil MR imaging at 3 T–Comparison of image quality, localization, and staging performance. Radiology 244(1):184–195, 2007.

CHAPTER 56  Male Pelvis 48. Hilton S: Contemporary radiological imaging of testicular cancer. BJU Int 104(9 Pt B):1339–1345, 2009. 49. Horwich A, Shipley J, Huddart R: Testicular germ-cell cancer. Lancet 367(9512):754–765, 2006. 50. Humphrey PA: Diagnosis of adenocarcinoma in prostate needle biopsy tissue. J Clin Pathol 60(1):35–42, 2007. 51. Reference deleted in proofs. 52. Iwazawa J, et al: Prostate cancer detection with MRI: Is dynamic contrast-enhanced imaging necessary in addition to diffusion-weighted imaging? Diagn Interv Radiol 17(3):243–248, 2011. 53. Jager GJ, et al: Pelvic adenopathy in prostatic and urinary bladder carcinoma: MR imaging with a three-dimensional TI-weighted magnetization-prepared-rapid gradient-echo sequence. AJR Am J Roentgenol 167(6):1503–1507, 1996. 54. Kattan MW, et al: Counseling men with prostate cancer: A nomogram for predicting the presence of small, moderately differentiated, confined tumors. J Urol 170(5):1792–1797, 2003. 55. Kim CK, Park BK, Kim B: Diffusion-weighted MRI at 3 T for the evaluation of prostate cancer. AJR Am J Roentgenol 194(6):1461–1469, 2010. 56. Kim SH, et al: Uterine cervical carcinoma: Evaluation of pelvic   lymph node metastasis with MR imaging. Radiology 190(3):807–811, 1994. 57. Kim B, et al: Imaging of the seminal vesicle and vas deferens. Radiographics 29(4):1105–1121, 2009. 58. Kim BS, et al: Comparison of pelvic phased-array versus endorectal coil magnetic resonance imaging at 3 Tesla for local staging of prostate cancer. Yonsei Med J 53(3):550–556, 2012. 59. King BF, et al: Seminal vesicle imaging. Radiographics 9(4):653–676, 1989. 60. Kitajima K, et al: Prostate cancer detection with 3 T MRI: Comparison of diffusion-weighted imaging and dynamic contrast-enhanced MRI in combination with T2-weighted imaging. J Magn Reson Imaging 31(3):625–631, 2010. 61. Koh DM, Sohaib A: Diffusion-weighted imaging of the male pelvis. Radiol Clin North Am 50(6):1127–1144, 2012. 62. Kok HK, Leong S, Torreggiani WC: Is magnetic resonance imaging comparable with computed tomography in the diagnosis of retroperitoneal metastasis in patients with testicular cancer? Can Assoc Radiol J 65(3):196–198, 2014. 63. Lawrence EM, et al: Prostate cancer: Performance characteristics of combined T(2)W and DW-MRI scoring in the setting of template transperineal re-biopsy using MR-TRUS fusion. Eur Radiol 24(7):1497– 1505, 2014. 64. Le JD, et al: Multifocality and prostate cancer detection by multiparametric magnetic resonance imaging: Correlation with whole-mount histopathology. Eur Urol 2014. 65. Lee JK, et al: Utility of computed tomography in the localization of the undescended testis. Radiology 135(1):121–125, 1980. 66. Lee SH, et al: Is endorectal coil necessary for the staging of clinically localized prostate cancer? Comparison of non-endorectal versus endorectal MR imaging. World J Urol 28(6):667–672, 2010. 67. Linder BJ, et al: Early localization of recurrent prostate cancer after prostatectomy by endorectal coil magnetic resonance imaging. Can J Urol 21(3):7283–7289, 2014. 68. Ling D, et al: Prostatic carcinoma and benign prostatic hyperplasia: Inability of MR imaging to distinguish between the two diseases. Radiology 158(1):103–107, 1986. 69. Lista F, et al: Multiparametric magnetic resonance imaging predicts the presence of prostate cancer in patients with negative prostate biopsy. Actas Urol Esp 2014. 70. Lucia MS, Miller GJ: Histopathology of malignant lesions of the penis. Urol Clin North Am 19(2):227–246, 1992. 71. Mack MG, et al: Superparamagnetic iron oxide-enhanced MR imaging of head and neck lymph nodes. Radiology 222(1):239–244, 2002. 72. Maeda E, et al: Abnormal signal intensities of the seminal vesicles   in a screening population. J Magn Reson Imaging 39(6):1426–1430, 2014.

1975

73. Maio A, Rifkin MD: Magnetic resonance imaging of prostate cancer: Update. Top Magn Reson Imaging 7(1):54–68, 1995. 74. Males RG, et al: Clinical application of BASING and spectral/spatial water and lipid suppression pulses for prostate cancer staging and localization by in vivo 3D 1H magnetic resonance spectroscopic imaging. Magn Reson Med 43(1):17–22, 2000. 75. Martin DC: Malignancy in the cryptorchid testis. Urol Clin North Am 9(3):371–376, 1982. 76. Maxeiner A, et al: Real-time MRI/US fusion-guided biopsy improves detection rates of prostate cancer in pre-biopsied patients. Aktuelle Urol 45(3):197–203, 2014. 77. Michel SC, et al: Preoperative breast cancer staging: MR imaging of the axilla with ultrasmall superparamagnetic iron oxide enhancement. Radiology 225(2):527–536, 2002. 78. Moller H, Prener A, Skakkebaek NE: Testicular cancer, cryptorchidism, inguinal hernia, testicular atrophy, and genital malformations: Case-control studies in Denmark. Cancer Causes Control 7(2):264–274, 1996. 79. Moon WK, et al: Calcified bladder tumors. CT features. Acta Radiol 33(5):440–443, 1992. 80. Mueller-Lisse UG, et al: Localized prostate cancer: Effect of hormone deprivation therapy measured by using combined three-dimensional 1H MR spectroscopy and MR imaging: Clinicopathologic casecontrolled study. Radiology 221(2):380–390, 2001. 81. Nguyen BC, et al: Multicenter clinical trial of ultrasmall superparamagnetic iron oxide in the evaluation of mediastinal lymph nodes in patients with primary lung carcinoma. J Magn Reson Imaging 10(3):468–473, 1999. 82. Obek C, et al: Comparison of digital rectal examination and biopsy results with the radical prostatectomy specimen. J Urol 161(2):494–498, discussion 498–9, 1999. 83. Ohr J: Tumors of the testis, adnexa, spermatic cord, and scrotum. Arch Pathol Lab Med 124(12):1855, 2000. 84. Oliver T: One-dose carboplatin in seminoma. Lancet 366(9496):1526, 2005. 85. Partin AW, et al: Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer. A multi-institutional update. JAMA 277(18):1445–1451, 1997. 86. Partin AW, et al: Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium. Urology 58(6):843–848, 2001. 87. Patterson H, et al: Combination carboplatin and radiotherapy in the management of stage II testicular seminoma: Comparison with radiotherapy treatment alone. Radiother Oncol 59(1):5–11, 2001. 88. Peckham MJ, et al: The treatment of metastatic germ-cell testicular tumours with bleomycin, etoposide and cis-platin (BEP). Br J Cancer 47(5):613–619, 1983. 89. Rajan DK, Scharer KA: Radiology of Fournier’s gangrene. AJR Am J Roentgenol 170(1):163–168, 1998. 90. Riches SF, et al: MRI in the detection of prostate cancer: combined apparent diffusion coefficient, metabolite ratio, and vascular parameters. AJR Am J Roentgenol 193(6):1583–1591, 2009. 91. Rockall AG, et al: Diagnostic performance of nanoparticle-enhanced magnetic resonance imaging in the diagnosis of lymph node metastases in patients with endometrial and cervical cancer. J Clin Oncol 23(12):2813–2821, 2005. 92. Rothwax JT, et al: Multiparametric MRI in biopsy guidance for prostate cancer: Fusion-guided. Biomed Res Int 2014:439171, 2014. 93. Sahani D, Samir A: Lymph Node Imaging Techniques. In Abdominal Imaging, 2010, Saunders. 94. Saksena M, et al: Comparison of lymphotropic nanoparticle-enhanced MRI sequences in patients with various primary cancers. AJR Am J Roentgenol 187(6):W582–W588, 2006. 95. Sala E, et al: Endorectal MR imaging before salvage prostatectomy: Tumor localization and staging. Radiology 238(1):176–183, 2006. 96. Sankineni S, Osman M, Choyke PL: Functional MRI in prostate cancer detection. Biomed Res Int 2014:590638, 2014.

1976

PART II  CT and MR Imaging of the Whole Body

97. Schiebler ML, et al: Prostatic carcinoma and benign prostatic hyperplasia: Correlation of high-resolution MR and histopathologic findings. Radiology 172(1):131–137, 1989. 98. Schlemmer HP: Multiparametric MRI of the prostate: Method for early detection of prostate cancer? Rofo 182(12):1067–1075, 2010. 99. Sella T, et al: Suspected local recurrence after radical prostatectomy: Endorectal coil MR imaging. Radiology 231(2):379–385, 2004. 100. Sharpe RM, Skakkebaek NE: Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? Lancet 341(8857):1392–1395, 1993. 101. Reference deleted in proofs. 102. Silva L, et al: Seminal vesicle tumors. Appl Cancer Res 30(3):313–317, 2010. 103. Silverman JM, Krebs TL: MR imaging evaluation with a transrectal surface coil of local recurrence of prostatic cancer in men who have undergone radical prostatectomy. AJR Am J Roentgenol 168(2):379–385, 1997. 104. Sobin LH, Fleming ID: TNM Classification of Malignant Tumors, fifth edition (1997). Union Internationale Contre le Cancer and the American Joint Committee on Cancer. Cancer 80(9):1803–1804, 1997. 105. Sosna J, et al: MR imaging of the prostate at 3 Tesla: Comparison of an external phased-array coil to imaging with an endorectal coil at 1.5 Tesla. Acad Radiol 11(8):857–862, 2004. 106. Stattaus J, Forsting M: The relevance of magnetic resonance imaging (MRI) for the detection and exclusion of prostate cancer. Urologe A 49(3):351–355, 2010. 107. Stephenson AJ, et al: Preoperative nomogram predicting the 10-year probability of prostate cancer recurrence after radical prostatectomy. J Natl Cancer Inst 98(10):715–717, 2006. 108. Subak LL, et al: Cervical carcinoma: Computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol 86(1):43–50, 1995. 109. Sung YS, et al: Prostate cancer detection on dynamic contrast-enhanced MRI: Computer-aided diagnosis versus single perfusion parameter maps. AJR Am J Roentgenol 197(5):1122–1129, 2011. 110. Swanson GP, Hubbard JK: A better understanding of lymphatic drainage of the prostate with modern imaging and surgical techniques. Clin Genitourin Cancer 11(4):431–440, 2013. 111. Szabo BK, et al: Dynamic MR imaging of the breast. Analysis of kinetic and morphologic diagnostic criteria. Acta Radiol 44(4):379–386, 2003. 112. Tabatabaei S, Harisinghani M, McDougal WS: Regional lymph node staging using lymphotropic nanoparticle enhanced magnetic resonance imaging with ferumoxtran-10 in patients with penile cancer. J Urol 174(3):923–927, discussion 927, 2005. 113. Tan CH, et al: Diffusion-weighted MRI in the detection of prostate cancer: Meta-analysis. AJR Am J Roentgenol 199(4):822–829, 2012. 114. Tekes A, et al: Dynamic MRI of bladder cancer: Evaluation of staging accuracy. AJR Am J Roentgenol 184(1):121–127, 2005. 115. Tempany CM, et al: Staging of prostate cancer: Results of Radiology Diagnostic Oncology Group project comparison of three MR imaging techniques. Radiology 192(1):47–54, 1994.

116. Thoeny HC, et al: Combined ultrasmall superparamagnetic particles of iron oxide-enhanced and diffusion-weighted magnetic resonance imaging reliably detect pelvic lymph node metastases in normal-sized nodes of bladder and prostate cancer patients. Eur Urol 55(4):761–769, 2009. 117. Thoeny HC, et al: Metastases in normal-sized pelvic lymph nodes: Detection with diffusion-weighted MR imaging. Radiology 273(1):125– 135, 2014. 118. Thompson J, et al: The role of magnetic resonance imaging in the diagnosis and management of prostate cancer. BJU Int 112(Suppl 2):6–20, 2013. 119. Thornhill BA, et al: Prostatic abscess: CT and sonographic findings. AJR Am J Roentgenol 148(5):899–900, 1987. 120. Tiguert R, et al: Lymph node size does not correlate with the presence of prostate cancer metastasis. Urology 53(2):367–371, 1999. 121. Torabi M, Aquino SL, Harisinghani MG: Current concepts in lymph node imaging. J Nucl Med 45(9):1509–1518, 2004. 122. Tsili AC, et al: MRI in the characterization and local staging of testicular neoplasms. AJR Am J Roentgenol 194(3):682–689, 2010. 123. Turkbey B, et al: Comparison of endorectal coil and nonendorectal coil T2W and diffusion-weighted MRI at 3 Tesla for localizing prostate cancer: Correlation with whole-mount histopathology. J Magn Reson Imaging 39(6):1443–1448, 2014. 124. Verma S, et al: Urinary bladder cancer: Role of MR imaging. Radiographics 32(2):371–387, 2012. 125. Vick R, Carson CC, 3rd: Fournier’s disease. Urol Clin North Am 26(4):841–849, 1999. 126. Vinnicombe SJ, et al: Normal pelvic lymph nodes: Evaluation with CT after bipedal lymphangiography. Radiology 194(2):349–355, 1995. 127. Washecka R, Rumancik WM: Prostatic abscess evaluated by serial computed tomography. Urol Radiol 7(1):54–56, 1985. 128. Williams SD, et al: Treatment of disseminated germ-cell tumors with cisplatin, bleomycin, and either vinblastine or etoposide. N Engl J Med 316(23):1435–1440, 1987. 129. Yagci AB, et al: The value of diffusion-weighted MRI for prostate cancer detection and localization. Diagn Interv Radiol 17(2):130–134, 2011. 130. Yang WT, et al: Comparison of dynamic helical CT and dynamic MR imaging in the evaluation of pelvic lymph nodes in cervical carcinoma. AJR Am J Roentgenol 175(3):759–766, 2000. 131. Yoshida T: MRI of testicular epidermoid cyst. Radiat Med 22(5):354– 356, 2004. 132. Yu KK, et al: Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and three-dimensional proton MR spectroscopic imaging. Radiology 213(2):481–488, 1999. 133. Zakian KL, et al: Transition zone prostate cancer: Metabolic characteristics at 1H MR spectroscopic imaging–Initial results. Radiology 229(1):241–247, 2003. 134. Zakian KL, et al: Correlation of proton MR spectroscopic imaging with Gleason score based on step-section pathologic analysis after radical prostatectomy. Radiology 234(3):804–814, 2005.

57  Female Pelvis Sadhna B. Nandwana, Kelly L. Cox, and Sherif G. Nour

IMAGING CONSIDERATIONS OF THE FEMALE PELVIS Imaging of the female pelvis plays a critical role in the diagnosis and evaluation of various gynecologic diseases. Typically ultrasonography (US) is used as first-line evaluation of the female genital organs. The lack of ionizing radiation, relative low cost, and ability to evaluate normal and pathologic findings in a wide range of gynecologic diseases makes US an attractive choice in imaging evaluation of the female pelvis. However, there are several scenarios where magnetic resonance imaging (MRI) and computed tomography (CT) are preferred for the evaluation of both benign and malignant gynecologic diseases. These include staging of gynecologic cancers, monitoring the response of therapy and surveillance in treated malignancies, surgical planning, and assessment of patients with acute pelvic pain in an emergency setting.

Magnetic Resonance Imaging In general the patient should be imaged in the supine position with a phased-array pelvic coil on either a 1.5- or 3-tesla (T) magnet. The increased signal-to-noise ratio allows for acquisition of a small field of view and high-resolution images. At minimum, T2 axial and sagittal imaging along with T1-weighted imaging before and after contrast administration should be performed. T1-weighted images provide excellent differentiation between the pelvic structures and surrounding pelvic fat, allowing for detailed anatomic evaluation. In- and out-  of-phase imaging is helpful in the detection of macroscopic and  microscopic fat, oftentimes found in adnexal lesions. Fat suppression sequences will assist in identifying any proteinaceous or hemorrhagic material, and postcontrast imaging may be used to evaluate a variety of pathologic conditions. If needed, hyoscine butylbromide (e.g., Buscopan) or glucagon may be given as an intramuscular or intravenous (IV) injection before imaging to reduce movement artifact from peristalsing small bowel loops that may have fallen into the pelvis. In addition a partially filled bladder will also improve image quality by preventing motion artifacts and bladder contractions that may occur with a full bladder. Complete emptying of the bladder should be avoided because this may hinder evaluation of lesions in the anterior peritoneal pouch between the uterus and bladder. Sequences and patient preparation can be further modified for specific cases and indications as outlined in Table 57-1. For example, endovaginal or endorectal coils may be used for detailed presurgical planning, particularly in evaluation of the female urethra.33 Endovaginal ultrasound gel can be applied to aid in identifying vaginal anomalies, malignant involvement, or extension into the vagina and vaginal endometriosis. Similarly, ultrasound gel can be instilled into the rectum to aid in the diagnosis of rectal wall invasion of gynecologic

malignancies or rectal endometriosis. Obtaining a single large field of view, such as a T2-weighted coronal single-shot fast-spin echo (SSFSE), is useful in evaluating combined genitourinary malformations.202 Also, the addition of diffusion-weighted imaging (DWI) with apparent diffusion coefficient (ADC) maps aids in evaluation of gynecologic malignancies.

Computed Tomography In multidetector CT, multiplanar reformatted images in transaxial, coronal, and sagittal planes should be obtained using a 4- to 5-mm slice thickness at 3- to 4-mm intervals. However, given radiation exposure associated with CT imaging, care should be taken to deliver the least amount of radiation dose possible. This includes using techniques such as automatic tube modulation and automatic current settings. IV contrast is necessary, particularly for evaluation of adnexal lesions, staging malignancy, and evaluation of vascular or inflammatory disease. Ideally a postcontrast venous phase with a 70- to 90-second delay should be obtained. In addition, enteric contrast should be administered to differentiate fluid-filled bowel loops from cystic adnexal lesions. CT is ideal for staging pelvic cancers, because it allows for complete evaluation of the abdomen for the presence of metastatic disease in the same imaging session. In addition, evaluation of peritoneal carcinomatosis, defining the origin of large pelvic masses, and differentiating uterine from adnexal masses can be easily performed on CT. It is also useful as a first-line imaging test in evaluating a wide range of etiologies of pelvic pain that may be unrelated to gynecologic issues in patients presenting to the emergency department.

VAGINA Normal Anatomy The vagina demonstrates an H-shaped configuration on transaxial imaging and consists of a fibromuscular tube extending upward and backward from the vulva to the uterus.62 It measures approximately 7 to 9 cm in length and is divided into thirds.76 The upper third of the vagina corresponds to the lateral vaginal fornices and is supported by the levator ani, transverse cervical, pubocervical, and sacrocervical ligaments.62 The middle third corresponds to the vagina at the level of the bladder base (which is supported by the urogenital diaphragm), and the lower third is defined as vagina below the level of the bladder base at the level of the urethra (which is supported by the perineal body).62,76 Lymphatic drainage and blood supply to the vagina is somewhat site specific. The vagina receives its blood supply from the vaginal and uterine branches of the internal iliac arteries, with additional blood supply for the middle and lower thirds of the vagina arising from the

1977

1978

PART II  CT and MR Imaging of the Whole Body

TABLE 57-1  Protocol for MRI of the Female Pelvis According to Gynecologic Indications Indication

Technique

Sagittal T2

Axial T2

Coronal T2

T1 Native

T1 Fat-Suppressed

T1 ContrastEnhanced

Endometrial cancer

Strict

Parallel

Optional

Always

If T1 hyperintense

3D high resolution

Cervical cancer*

Strict

Parallel

Optional

Always

If T1 hyperintense

Always

Adnexal mass

Strict

Always

Strict

Always

If T1 hyperintense (in-phase, opposed-phase) If T1 hyperintense

Always

Tumor recurrence vs. scar Endometriosis†

Perpendicular to uterine cavity Perpendicular to endocervix Perpendicular to uterine cavity Strict

Strict

Parallel

Always

Always

Adenomyosis

Strict

Parallel

Always

If T1 hyperintense

Leiomyomas

Strict

Parallel

Always

If T1 hyperintense

If bladder, vaginal, or bowel wall thickening If associated with leiomyomas Always

Müllerian abnormality*

Strict

Perpendicular to uterine cavity Perpendicular to uterine cavity Perpendicular to uterine cavity Strict

If T1 hyperintense

Optional

Parallel Strict

Strict

Optional

Always

Always

*Vaginal opacification with 60 to 100 mL sonographic gel optional. † Rectal opacification with 150 to 200 mL sonographic gel.

middle rectal and internal pudendal arteries.62 The lymphatic drainage of the upper third of the vagina is predominantly to the external and internal iliac nodes, whereas drainage from the middle third of the vagina is to the internal iliac nodes, and drainage from the lower third of the vagina is to the superficial inguinal nodes.62 Venous plexuses exist in the adventitial layer surrounding the vagina and form the rectovaginal septum posteriorly, subsequently draining into the internal iliac veins.62 On MRI the normal vaginal mucosal layer is T2 hyperintense with corresponding T1 hypointensity and is lined by estrogen-sensitive stratified squamous epithelium and endoluminal mucus.62 The vaginal mucosa enhances after administration of IV gadolinium. The vaginal submucosa and muscularis layers demonstrate T1- and T2-hypointense signal and cannot be distinguished from each other on imaging.62 The outer adventitial layer may demonstrate serpentine T2-hyperintense structures due to slow flow within the vaginal venous plexus62,169 (Fig. 57-1). Signal characteristics of the vaginal wall and thickness of the mucosal layer parallels estrogen levels and will therefore vary depending on the menstrual cycle. As a result the mucosal layer is thicker and of higher T2-weighted signal intensity during the late proliferative and early to middle secretory phases of the menstrual cycle and during pregnancy. Before menarche and after menopause, the vaginal wall displays low T2 signal intensity and the mucosal layer is very thin.62,76,169

Developmental and Congenital Abnormalities MRI is the modality of choice to assess suspected developmental disorders of the vagina that may be isolated or associated with uterine abnormalities.190 MRI is superior to both CT and US in rendering relevant clinical information for diagnosis and treatment planning and evaluating concomitant abnormalities, owing to MRI’s superior soft tissue contrast and lack of ionizing radiation. Embryologic development of the vagina is derived from the paired müllerian ducts (upper two thirds) and the urogenital sinus.62 Vaginal malformations are commonly associated with other developmental abnormalities of the müllerian ducts, because these structures also give rise to the uterus, cervix, and fallopian tubes.26,62 The wolffian duct is also in close proximity, which may explain the association between

FIG 57-1  Normal anatomy of the vagina on MRI. Axial T2-weighted image demonstrates T2-hyperintense mucosa and endoluminal mucus centrally. T2-hypointense thin linear signal represents vaginal submucosa and muscularis layers (arrowhead) with the surrounding outer adventitial layer containing the venous plexus (arrow).

vaginal and renal anomalies in 30% to 80% of cases.109,177 The most common association is renal agenesis on the side of the müllerian duct defect.169

Vaginal Agenesis.  Partial or complete vaginal agenesis occurs in 1 in 5000 women and should be suspected with a clinical history of primary amenorrhea. This may be isolated or associated with agenesis or hypoplasia of the uterus.169 The most common cause of vaginal agenesis is Mayer-Rokitansky-Küster-Hauser syndrome, which is characterized by congenital aplasia of the uterus and upper two thirds of the vagina in women showing normal development of secondary sexual characteristics with a normal 46,XX karyotype.118 Other causes of vaginal agenesis include conditions associated with ambiguous genitalia, such as gonadal dysgenesis (Turner’s syndrome, 45,XO), male pseudohermaphroditism, and testicular feminization syndrome.62

CHAPTER 57  Female Pelvis

1979

B

B

A

FIG 57-2  Hematocolpos due to hymenal atresia in a 14-year-old girl. Coronal T1-weighted (A) and sagittal T2-weighted (B) images show dilatation of the vagina that extends to the level of L5 and displays high signal intensity on both images. Owing to obstruction, mild hematometra (arrow in B) is also seen. B, bladder.

Vaginal Septum.  Congenital vaginal septa may present as either a transverse or longitudinal septum and may be complete or incomplete. A transverse vaginal septum can occur anywhere along the vagina, although it is most commonly seen at the junction of the upper and middle thirds.62 It can appear as a thin T2-hypointense band that is composed of fibrous connective tissue with vascular and muscular components.62,190 It may present as primary amenorrhea or hematometrocolpos, with resulting complications of hematosalpinx and endometriosis.62 MRI is useful in defining the location and thickness of a vaginal septum for preoperative planning, as well as identifying other potential causes of hematocolpos such as hymenal atresia (Fig. 57-2). Longitudinal vaginal septa are usually asymptomatic and may arise either as a failure of lateral fusion of the müllerian ducts (resulting in duplication of the uterus, cervix, and vagina) or as a result of incomplete resorption of the vaginal septum.62 They are best visualized on T2-weighted imaging in either the coronal or axial plane and appear as thin low-signal-intensity structures.

Congenital Vaginal Cysts.  Congenital vaginal cysts are often incidentally found on pelvic MRI and include both müllerian (paramesonephric) and Gartner’s duct (mesonephric) cysts, which cannot be distinguished from each other on imaging.48 Both lesions typically appear as T1-hypointense, T2-hyperintense cystic structures within the vaginal anterolateral wall anywhere from the cervix to the introitus (Fig. 57-3). Lesions may contain intermediate to high T1 signal if protein-

aceous or hemorrhagic contents are present.62 The majority of lesions are small and asymptomatic and do not require intervention.62

Benign Pathologic Conditions

Noncongenital Cysts.  Noncongenital cystic lesions of the vagina include retention cysts of the vestibulum and epidermal inclusion cysts.48,62 Inclusion cysts are usually asymptomatic and often found within the lower posterior or lateral vaginal wall. They typically occur after trauma and are due to buried epithelial fragments, appearing as a complex cystic lesion on MRI. Although CT and US can be used to make the diagnosis, MRI is the modality of choice for assessment of vaginal cysts because MRI is often useful in excluding other etiologies in the differential diagnosis, such as vaginal prolapse, tumors, and ischiorectal fossa lesions.48

Vaginal Fistula.  Several different types of fistulas can develop, including vesicovaginal, colovaginal, or recto- or anovaginal fistulas. These typically develop after gynecologic surgery or radiation therapy, although other causes include Crohn’s disease, diverticulitis, malignancy, chronic infection, or trauma.62,168 Fistulous tracts appear as high-signal tracts extending from the vagina to adjacent structures, best visualized on T2-weighted sequencing and short tau inversion recovery (STIR) imaging (Fig. 57-4). Gadolinium-enhanced T1-  weighted images with fat suppression will often show parallel fistulous wall enhancement.81

1980

PART II  CT and MR Imaging of the Whole Body intense, best visualized on fat suppression images, with intermediate to high signal intensity on T2-weighted imaging.101 In addition, areas of fibrosis and adhesions due to a resulting inflammatory response may be seen, which is typically characterized by T2-hypointense fat stranding or spiculation. MRI has a sensitivity and specificity of 80% and 93%, respectively, for the diagnosis of vaginal endometriosis.11 Endometriosis is further discussed later under “Ovaries.”

Vaginal Leiomyomas.  These are rare solid, well-defined benign

FIG 57-3  Gartner’s duct cyst. Sagittal T2-weighted image demonstrates a T2-hyperintense lesion arising from the anterior wall of the vagina, consistent with a Gartner’s duct cyst (arrow). This lesion was T1 hypointense and did not enhance (not shown). (Courtesy Dr. Pardeep Mittal, Atlanta, Ga.)

tumors ranging from 1 to 5 cm that typically arise within the anterior midline vaginal wall and less likely the posterior or lateral walls.62,74 Leiomyomas arising from the anterior wall are associated with urinary tract symptoms (e.g., increased frequency, dysuria, urgency). Similar to uterine leiomyomas, vaginal leiomyomas may be intramural or pedunculated and demonstrate homogenous T1- and T2-hypointense signal with homogenous enhancement on postcontrast images. Hyaline degeneration results in low T2 signal intensity, whereas myxoid and cystic degeneration demonstrates high T2 signal intensity. Hemorrhagic degeneration demonstrates high signal intensity on both  T1- and T2-weighted imaging.62 Differential diagnosis includes a fibroepithelial polyp of the vagina, which is an uncommon benign asymptomatic lesion characterized by a squamous epithelial surface with a fibrovascular stalk.120

Malignant Pathologic Conditions

Vaginal Cancer.  Primary carcinoma of the vagina comprises

B

FIG 57-4  Anovaginal fistula on MRI. Sagittal T2-weighted fatsuppressed image demonstrates a T2-hyperintense linear tract extending from the posterior wall of the vagina to the anus (arrow). Patient had a history of Crohn’s disease. B, bladder.

Radiation Change.  The vagina may demonstrate significant changes after receiving radiation. Within the first 6 months post radiation, the vaginal wall shows T2-hyperintense signal primarily due to mucosal and intramuscular edema.174 Typically this is followed by vaginal atrophy, where the vaginal wall shows T2-hypointense signal. Additional findings can include narrowing, foreshortening, and/or stenosis of the vaginal canal. Severe radiation injury can result in vaginal ulceration, necrosis, and secondary fistula formation.168

Endometriosis.  Endometriosis is defined by the presence of functional endometrial glands and stroma outside the uterine cavity.197 The most common site of endometriosis is the ovary, but endometrial deposits may be found in any pelvic organ or other distant sites, including the vagina.62,197 Vaginal endometrial implants usually occur in the posterior aspect of the upper vagina and demonstrate intense enhancement on postcontrast images. Lesions are typically T1 hyper-

approximately 1% to 2% of all gynecologic malignancies, with 90% composed of squamous cell carcinomas, usually occurring in elderly women.12,62,152 Risk factors for development of vaginal cancer include chronic human papillomavirus (HPV) infection, multiple sexual partners, immunosuppression, prior radiation, and smoking.62 It commonly arises from the posterior wall of the proximal third of the vagina and tends to demonstrate early involvement of adjacent structures, including the parametrium, cervix, vulva, and rectovaginal and vesicovaginal septae.62 Lymph node involvement typically involves the iliac chain nodes from the upper third of the vagina and spreads to the inguinal lymph nodes from the lower two thirds of the vagina.62 MRI is helpful in defining the local extent of vaginal carcinoma and lymph node status. Tumor typically appears isointense on T1, and on T2 shows intermediate to high signal intensity relative to normal vaginal tissue, with enhancement on postcontrast images.31 Management depends upon the size, location, and tumor-node-metastasis (TNM) staging of the tumor. Stage I tumors are limited to the vagina, with the paravaginal fat retaining its normal high signal intensity on T1-weighted images. Stage II tumors extend into the paravaginal fat, disrupting the normal low signal intensity of the vaginal wall and replacing the paravaginal fat with abnormal low signal intensity on T1-weighted images. In stage III tumors, the carcinoma extends to the pelvic sidewall, with disruption of the normal low signal intensity of the pelvic sidewall muscles on T2-weighted imaging. Stage IV tumors extend beyond the true pelvis or invade the rectum or bladder, with resultant loss of the normal low-signal-intensity rectal or bladder wall on T2-weighted imaging.62 Additional vaginal cancers include vaginal adenocarcinoma, which is typically associated with in utero exposure to diethylstilbestrol (DES), and a T-shaped uterus.62 Very rarely, primary vaginal melanoma, lymphoma, and leiomyosarcoma can also occur.62

Secondary Vaginal Involvement.  Most tumors affecting the vagina (up to 90%) are due to direct extension from primary cervical, vulvar, bladder, or rectal neoplasms.62 The overall accuracy of MRI for all metastatic tumors involving the vagina has been reported as

CHAPTER 57  Female Pelvis 92%.31,76 Metastatic tumors demonstrate similar imaging characteristics to primary tumors, with low to intermediate signal intensity on T1-weighted imaging and high to intermediate signal intensity on T2-weighted imaging, with enhancement on postcontrast imaging sequences.62

VULVA Normal Anatomy The vulva is composed of the mons pubis, labia majora and minora, clitoris, vestibular bulb, vestibular glands, and the vestibule of the vagina.62 Arterial blood supply to the vulva is via branches of the external and internal pudendal arteries. Lymphatic drainage is into  the medial group of superficial inguinal nodes. The vulva demonstrates low to intermediate signal intensity on T1-weighted imaging and slightly higher signal intensity on T2-weighted imaging.26,62

1981

tion.48,62 Rarely, solid or nodular mural lesions can be indicative of the extremely rare Bartholin’s gland carcinoma.48

Solid Lesions.  Rarely, vulvar fibromas, lipomas, and leiomyomas can occur. Fibromas demonstrate areas of marked hypointensity consistent with fibrosis on T1- and T2-weighted MRI.79 Vulvar lipomas demonstrate signal intensity that follows fat sequences on MRI.62 Vulvar leiomyomas typically arise from the labia majora, are homogenously T1 and T2 hypointense, and may demonstrate changes of cystic degeneration.62

Vulvar Infection.  The vulva is susceptible to community-acquired infections such as community-acquired methicillin-resistant Staphylococcus aureus (MRSA) colonization, which can lead to vulvar abscesses, necrotizing fasciitis, or Fournier’s gangrene. Common risk factors include obesity and diabetes mellitus.74

Developmental and Congenital Abnormalities

Vulvar Thrombophlebitis.  Rarely, vulvar edema and varicose veins

Vascular malformations are considered developmental congenital defects and rarely may occur in the lower female genital tract.62 These malformations can be further subdivided into arterial, venous, capillary, lymphatic, or mixed types, with the most common type being venous malformations.62,74 Vulvar venous malformations are usually incidental and no treatment is required unless symptomatic. Typically, intermediate signal intensity is seen on T1-weighted imaging, with peripheral areas of high signal intensity due to intralesional fat. On T2-weighted imaging, venous malformations appear as high-signalintensity lobules, indicating slow blood flow, and fluid-fluid levels may be present in larger compartments of those slow-flow (venous) malformations. Areas of linear or punctate low T2 signal intensity may be seen, representing either fast flow within vessels or phleboliths.62,74

may develop during pregnancy as a result of compression from the gravid uterus.74 Thrombophlebitis can subsequently occur in these varicose veins either during pregnancy or in the postpartum period. Occasionally tamoxifen use for breast cancer has also been associated with superficial thrombophlebitis, presumably owing to the effect of a hypercoagulable state.74

Benign Pathologic Conditions

Bartholin’s Cysts.  Bartholin’s cysts are the most common vulvar cyst and are of urogenital sinus origin, arising from the Bartholin’s glands. They are commonly seen in women of reproductive age, range in size from 1 to 4 cm, and are typically located at the lateral introitus adjacent to the labia minora.48 Lesions are typically T2 hyperintense, with variable signal on T1 depending on the proteinaceous content (Fig. 57-5). Thickening and enhancement of the wall can be seen with inflamma-

A

Malignant Pathologic Conditions

Primary Vulvar Carcinoma.  Vulvar cancer is a rare malignancy that tends to affect elderly women, although a bimodal distribution exists.62,74 Approximately 90% of all vulvar carcinomas are squamous cell carcinomas. HPV-positive tumors (associated with HPV types 16, 18, and 33) occur in a younger age group, may be multifocal, and show an association with vulvar intraepithelial neoplasia. Other histologic types include sarcoma, basal cell carcinoma, adenocarcinomas, and extramammary Paget’s disease.62,74 Vulvar carcinoma usually involves the labia in two thirds of cases and demonstrates local invasion before lymph node involvement (Fig. 57-6). The most important prognostic factors for determining survival are tumor size, depth of invasion, and presence of inguinofemoral lymph node metastases.62,74 Imaging does not play a significant role in assessment of the primary tumor, inasmuch as this is readily assessed on clinical exam.

B FIG 57-5  Bartholin’s gland cysts. A, Bilateral lesions with T2-hyperintense signal and variable T1 signal, which depends on internal contents. B, Right-sided cyst contains proteinaceous or hemorrhagic contents (arrowhead), and left-sided cyst is simple (arrow).

1982

PART II  CT and MR Imaging of the Whole Body

A

B FIG 57-6  Vulvar cancer. Transaxial T1-weighted (A) and T2-weighted (B) images show diffuse irregular thickening of the right small labium (arrow), with an irregular surface and intermediate signal intensity on both sequences.

However, imaging does play an important role in assessment of any regional adenopathy and in surgical planning.62,74 Enlarged superficial inguinal lymph nodes measuring greater than 10 mm in short-axis diameter and deep inguinal nodes measuring greater than 8 mm in short-axis diameter have been shown to demonstrate a high specificity for malignant involvement (97%-100%).171 Additional vulvar malignancies may rarely include vulvar lymphoma, melanoma, leiomyosarcoma, and aggressive angiomyxoma.62,74

FEMALE URETHRA Normal Anatomy The normal female urethra is a tubular structure that measures approximately 3 to 4 cm in length. It extends from the internal urethral meatus at the bladder neck to the external urethral meatus at the vestibule, coursing through the urogenital diaphragm.33,42,176 The zonal anatomy of the urethra consists of mucosal, submucosal, and muscular layers, which can be distinguished on MRI, particularly on T2-weighted images, by alternating hyper- and hypointense rings.33,42,175,176 There are numerous small submucosal paraurethral and periurethral glands that secrete material into the lumen of the urethra.33,176 Transitional cell epithelium lines the proximal third of the urethra, and stratified squamous epithelium lines the distal two thirds of the urethra.33

Benign Pathologic Conditions

Urethral Diverticula.  Urethral diverticula are protrusions of the urethra into the periurethral fascia. The epithelial lining of a diverticulum is identical to the urethral mucosa, and communication with the urethral lumen is maintained. Some diverticula may be congenital,  but the vast majority are felt to arise from acquired causes. The most widely accepted theory involves rupture of a chronically obstructed and infected periurethral gland into the urethral lumen. Most patients present with nonspecific symptoms including urinary frequency, urgency, pelvic pain, urinary tract infections, incontinence, and urine retention. Urethral diverticula are classically most conspicuous on T2-weighted images, with the fluid-containing cystic cavity being T2 hyperintense relative to the surrounding soft tissues. They typically arise from the posterolateral wall of the midurethra at the level of the pubic symphysis and demonstrate various degrees of extension around

FIG 57-7  Urethral diverticulum. T2-weighted image demonstrates large urethral diverticulum that extends to involve nearly the entire circumference of the urethra (arrow). (Courtesy Dr. Aarti Sekhar, Atlanta, Ga.)

the circumference of the urethra (Fig. 57-7). Occasionally the diverticula can become infected with Escherichia coli and less often, gonococci and Chlamydia species. Inflamed urethral diverticula may demonstrate heterogeneous signal intensity on T1-weighted images and a possible fluid-fluid level. Additional complications include calculus formation, which occurs in approximately 10%. Rarely, neoplasm can develop within the diverticulum, with the majority being adenocarcinoma.33,74,176

Skene’s Gland Cyst.  Ductal obstruction secondary to infection or trauma leads to formation of Skene’s duct cysts, which are located at the floor of the distal urethra and are bilateral. Most of these cysts are clinically insignificant, but cysts larger than 2 cm often cause urinary symptoms such as dysuria or obstructive voiding symptoms.176 Differential diagnosis includes a urethral diverticulum, although MRI findings such as distal location, teardrop shape, and paired structures are more suggestive of Skene’s cysts.176

CHAPTER 57  Female Pelvis Urethral Leiomyoma.  Urethral leiomyomas are extremely rare benign smooth muscle tumors that may grow during pregnancy and result in dysuria.33 They frequently involve the proximal segment of the urethra and may be confined to the urethra, extend into the paraurethral region, or protrude through the urethral meatus.176 On MRI, lesions typically demonstrate intermediate T1 signal intensity, variable signal on T2-weighted images, and contrast enhancement.33,74,176

Urethral Caruncle.  Urethral caruncles are small benign lesions of the posterior margin of the external urethral meatus that are caused by distal urethral prolapse in hypoestrogenic postmenopausal women.33 Although most women are asymptomatic, urethral caruncles can occasionally cause pain and hematuria. On MRI a caruncle may appear as a T2-hyperintense mound of tissue surrounding the external urethral meatus.33

Urethral Fistulas.  Urethral fistulas are divided into urethrovaginal, rectourethral, and urethroperineal subtypes. Urethrovaginal fistulas are typically sequelae of Crohn’s disease, Behçet’s disease, or postsurgical complications.33 Rectourethral fistulas are typically congenital, whereas urethroperineal fistulas are thought to be acquired fistulas from chronic decubitus ulcers or abscesses in the perineal region.33 These fistulas appear as T2-hyperintense tracts extending from the urethra to the vagina, rectum, or perineum.33

Malignant Pathologic Conditions

1983

the uterus. The venous system parallels the arterial system and drains into the iliac veins. Lymphatic drainage from the lower and middle parts of the uterus are drained by the parametrial, paracervical, and obturator nodes, whereas lymphatic drainage from the upper corpus and uterine fundus is drained by the common iliac and paraaortic lymph nodes.112

Developmental and Congenital Abnormalities Congenital uterine abnormalities called müllerian duct abnormalities (MDAs) encompass a broad spectrum of findings that typically are due to interruptions in the development of müllerian ducts during the sixth week of pregnancy.202 The müllerian ducts normally give rise to the uterus, fallopian tubes, cervix, and upper two thirds of the vagina. When abnormalities occur they are often associated with renal/urinary tract and vertebral anomalies.202 Müllerian anomalies have been categorized by the American Fertility Society (AFS) into seven classes that are based primarily on morphologic features of the uterine corpus and are summarized in Fig. 57-8.3 Diagnostic imaging workup for MDAs may begin with hysterosalpingography and/or US. However, MRI is generally needed to classify the abnormality into a specific MDA category and is often preferred for its lack of invasiveness, lack of ionizing radiation, multiplanar imaging, and excellent soft tissue characterization.202

AFS Class I.  Agenesis or hypoplasia of the proximal two thirds of the

UTERUS

vagina, cervix, and/or uterus accounts for approximately 10% of MDAs.188 Absence or reduced development of the uterus and upper vaginal agenesis is the hallmark of Mayer-Rokitansky-Küster-Hauser syndrome, which comprises a vast majority of these MDA cases.188 Abnormalities may be isolated or associated with other nongynecologic abnormalities. The most common clinical manifestation is primary amenorrhea with normal development of secondary sexual characteristics.202 Typical MRI appearance includes bilateral uterine buds connected by a fibrous band that converge over the bladder dome, appearing as a triangular soft tissue (Fig. 57-9).202

Normal Anatomy

AFS Class II.  Unicornuate uterus results from normal development

The uterus is divided into three anatomic regions: fundus, body (uterine corpus), and cervix. The uterine body is formed of three layers: endometrium (inner lining), myometrium (muscular wall), and serosa (outer surface).68,173,184 On MRI the inner layer of high signal intensity on T2 corresponds to the endometrial lining, the endometrial cavity, and its contents. Normal thickness of the endometrium varies with the phase of menstrual cycle, with the greatest thickness during the midsecretory phase. In postmenopausal patients the endometrium should measure less than 5 mm but can measure up to 8 mm if the patient is undergoing hormone replacement therapy.61,64,106 Deep to the endometrium is a band of low T2 signal intensity known as the junctional zone, which is formed by the inner third of the myometrium.162 The tissue of the outer two thirds of the myometrium demonstrates intermediate signal intensity on T2 imaging.173 The cervix is separated from the uterine body by the internal os at the level of the entrance of uterine vessels. The endocervical canal contains small folds called plicae palmatae. The cervix itself is composed of dense fibrous stroma that is homogenously T2 hypointense. The outer cervical stroma is mildly T2 hyperintense and can be difficult to differentiate from the adjacent parametria. The uterus is supported by the round and utero-ovarian ligaments and is covered by peritoneum. The utero-ovarian ligaments blend  laterally to form the anterior and posterior leaves of the broad ligament. The vascular supply of the uterus is derived mainly from  the uterine arteries, which pass within the broad ligaments to enter 

of one müllerian duct and varying degrees of arrested development of the other side. Four subtypes exist: (1) unicornuate uterus with no rudimentary horn, (2) a rudimentary horn without a uterine cavity, (3) a rudimentary horn with a cavity communicating with the normal side, and (4) a rudimentary horn with a noncommunicating cavity. In patients with a noncommunicating uterine horn, endometriosis, hematometra, and dysmenorrhea may occur.202 The MRI of a banana- or cigar-shaped uterine cavity is pathognomonic of unicornuate uterus. It is typically small and located off the midline, with preserved zonal anatomy.188 The contralateral rudimentary horn shows variable findings. It may appear as a soft tissue lesion with the same signal intensity as myometrium, with or without the high signal intensity of normal endometrium on T2-weighted images. In obstructed rudimentary horns, central areas of high signal intensity on T1-weighted images and low signal intensity on T2-weighted images are indicative of hematometra.19

Urethral Carcinoma.  Primary urethral carcinoma is rare, although risk factors include HPV and urethral diverticula.141 The most common type of malignancy is squamous cell carcinoma with drainage to the inguinal nodes. However, the most common type of malignancy associated with a urethral diverticulum is clear cell adenocarcinoma.4 MRI findings typically include a T1-hypointense mass, with intermediate signal on T2 and heterogeneous enhancement.74

AFS Class III.  Uterus didelphys is caused by failure of the müllerian ducts to fuse, resulting in complete duplication of the uterine corpus and cervix. Imaging findings include two widely separated, almost normally developed uterine horns with maintained zonal anatomy and two separate cervices. Occasionally one side of the duplicated proximal vagina may be obstructed by a transverse hemivaginal septum.202 This can lead to dysmenorrhea, endometriosis, infections, and pelvic adhesions. When this constellation of findings is associated

1984

PART II  CT and MR Imaging of the Whole Body I Hypoplasia/agenesis

(a) Vaginal

(b) Cervical

II Unicornuate

(a) Communicating

III Didelphys

(b) Noncommunicating IV Bicornuate

(c) Fundal

(d) Tubal

(e) Combined

(d) No cavity

V Septate

(a) Complete

(d) No horn VI Arcuate

(a) Complete

(b) Partial

VII DES drug related

(b) Partial

FIG 57-8  Classification of müllerian duct abnormalities. (From American Fertility Society: The American Fertility Society classification of the adnexal adhesions, tubal occlusion secondary to tubal ligation, tubal pregnancies, müllerian anomalies, and intrauterine adhesions. Fertil Steril 49:944–955, 1988; modified by Troiano RN, McCarthy SM: Müllerian duct anomalies: Imaging and clinical issues. With permission from Radiology 23:19–34, 2004.)

with ipsilateral renal agenesis, it is referred to as Herlyn-WernerWunderlich syndrome.202

AFS Class IV.  Bicornuate uterus constitutes approximately 10% of MDAs and results from incomplete fusion of the müllerian ducts, with or without cervical duplication, referred to as bicornuate bicollis and bicornuate unicollis, respectively.202 In bicornuate uterus, two small uterine cavities with preserved zonal anatomy can be identified. Although the two uterine horns are not widely separated as in uterus didelphys, there is a wide fundal cleft of greater than 1 cm, and the distance between the two uterine cornua is usually greater than 5 cm.202 The central muscular septum displays signal intensity similar to peripheral myometrium and divides the uterine cavities. Occasionally the septum may demonstrate mixed muscular and fibrous components in its distal portion.188,190

AFS Class V.  Septate uterus results from incomplete resorption of the median septum. It is the most common type of MDA (55%) and is associated with the poorest reproductive outcome, although it is surgically correctable.202 The typical imaging feature is a uterus of normal size and normal fundal contour with duplication of the endometrial cavity by a median septum (Fig. 57-10). The septum arises from the midline fundus and can be composed of fibrous tissue, muscle, or both. On MRI a fibrous septum is hypointense to the myometrium and a muscular septum is isointense to the myometrium. The composition of the septum is important for surgical planning, with predominantly fibrous septae being treated with less invasive hysteroscopic techniques and muscular septae requiring a transabdominal surgical approach.202

AFS Class VI.  Arcuate uterus is a mild form of a resorption anomaly, with a normal-sized uterus, one endometrial cavity, and normal zonal anatomy with no known clinical significance.202 MRI findings include mild thickening of the fundal myometrium with a normal convex external fundal contour.202

AFS Class VII.  DES drug-related abnormalities of the reproductive system occur in approximately 70% of women exposed in utero.188 A small uterus with a T-shaped endometrial cavity is the most common abnormality, seen in one third of cases.190 Typical DES-related malformations are confidently diagnosed by hysterosalpingography.

Benign Pathologic Conditions

Uterine Leiomyoma.  Leiomyomas, also known as fibroids or myomas, are the most common benign neoplasms of the female genital tract. They are associated with younger age at menarche, obesity, and African American race.5 They tend to be multiple, of various sizes, and are hormone sensitive. Leiomyomas are composed of smooth muscle cells arranged in a whorl-like pattern, with various amounts of fibrous connective tissue, collagen, and extracellular matrix.5 The majority of leiomyomas involve the uterine corpus and can often be treated with uterine artery embolization, hysterectomy, or myomectomy if symptomatic. The type of treatment often depends on location, thickness of stalk (if pedunculated), and vascularity of fibroids.166 Degeneration of fibroids is related to a lack of blood supply and is more commonly found in postmenopausal women. Types of degeneration include hyaline or myxoid degeneration, calcification, cystic, and red (hemorrhagic) degeneration.122 Rarely, sarcomatous transformation can occur, but most leiomyosarcomas arise independently.122

CHAPTER 57  Female Pelvis

FIG 57-9  Mayer-Rokitansky-Küster-Hauser syndrome. T2 sagittal image demonstrates absence of uterus, with triangular soft tissue converging over the bladder dome, likely representing connected uterine buds (arrow). Free fluid is present within the pelvis. (Courtesy Dr. Aarti Sekhar, Atlanta Ga.)

A

1985

Leiomyomas are typically T1 isointense and T2 hypointense to the myometrium, although they may demonstrate increased T1 signal with the presence of hemorrhagic degeneration, fat content, or high cellular content.166 The appearance after IV contrast administration is variable, although most myomas enhance similar to or less than the surrounding myometrium on both contrast-enhanced MRI and CT (Fig. 57-11).5 Leiomyomas with hyaline degeneration or calcification remain of low signal intensity on T2-weighted images. Calcifications on CT appear as mottled, whorled, streaked, or ringlike.28 Increased T2 signal can occur with cystic degeneration, and these areas do not enhance.166 Uterine fibroids are typically classified according to their location within the uterus as submucosal, intramural, or subserosal types.122 Rarely, fibroids can also occur in the cervix or broad ligaments (Fig. 57-12). Submucosal leiomyomas are the least common but the most likely to cause symptoms such as abnormal uterine bleeding and infertility disorders. The most common leiomyomas are intramural fibroids, which may cause menorrhagia, infertility, or pelvic pain.122 Acute pain is usually a manifestation of acute degeneration or hemorrhage. However, rare cases of torsion, prolapse of a subserosal leiomyoma, rupture, or infection may also cause pain.5,28 Variants of leiomyomas.  Cellular leiomyomas are composed predominantly of smooth muscle cells, which are T2 homogenously hyperintense and demonstrate avid enhancement.201 Lipoleiomyoma is a fibroid that contains internal fat and demonstrates fat-attenuation values on CT or high signal intensity on T1-weighted images, with signal loss on fat-suppressed T1-weighted images.85 Parasitic leiomyoma is a pedunculated subserosal fibroid that develops a new blood supply from adjacent structures such as the omentum and becomes completely detached from the uterus.37 Disseminated peritoneal leiomyomatosis is a rare condition consisting of multiple peritoneal smooth muscle nodules mimicking diffuse peritoneal carcinomatosis or retroperitoneal masses resembling leiomyosarcoma.37 Benign metastasizing

B FIG 57-10  Septate uterus. A, Oblique axial T2-weighted image of the uterine body shows a hypointense septum (arrowhead) in the middle of the endometrial cavity. B, Oblique coronal T2-weighted image through the uterine cavity shows a hypointense septum (arrowhead) through the uterine cavity. A normal distance of 3.5 cm (bidirectional arrow) between the uterine cornua and a missing fundal cleft exclude the diagnosis of bicornuate uterus.

1986

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 57-11  MRIs of a 32-year-old patient in whom it was impossible to visualize the uterine cavity with sonography. A, Sagittal T2-weighted image allows adequate visualization of the uterine cavity, demonstrating large intramural fibroids and the absence of thickening of the junctional zone. The vaginal anatomy is normal, with a hyperintense mucosa (arrows) and hypointense vaginal wall. The left ovary (arrowheads) contains numerous tiny follicles. B, Axial T2-weighted image through the uterus demonstrates three hypointense fibroids and minimal peritoneal fluid between the ovaries. C, Contrast-enhanced T1-weighted image at the same level as in B shows strong enhancement of the normal myometrium and heterogeneous enhancement of the fibroids. The right ovary demonstrates typical features of a nonhemorrhagic corpus luteum (B and C).

leiomyomas are leiomyomas that are present in both the uterus and lung, manifested by single or multiple pulmonary nodules.37

Adenomyosis.  Adenomyosis is an idiopathic and common condition usually occurring in multiparous pre- or perimenopausal females with similar symptoms to leiomyomas, including abnormal uterine bleeding, dysmenorrhea, enlarged uterus, and pelvic pain.49,173 Pathologically there are endometrial glands and stroma ectopically located within the myometrium, with adjacent myometrial hypertrophy.49,173 Adenomyosis may be diffuse or focal. The diffuse form of adenomyosis can be readily identified on transvaginal US or MRI.45 On MRI there is ill-defined thickening of the low T2 signal band surrounding the endometrium, which represents the junctional zone. The junctional

zone is hormone dependent, with variable normal thickness depending on the phase of cycle; however, studies have shown that a junctional zone measurement of greater than 12 mm is highly specific for adenomyosis.145 Other studies have shown that the diagnosis may also be suggested with a junctional zone measuring between 8 and 12 mm.24,145 Adenomyosis can be excluded with a junctional zone of less than 8 mm.180 Additional imaging features that may support the diagnosis of adenomyosis include the presence of subendometrial cysts, which typically measure between 2 and 6 mm and appear as T2-hyperintense rounded foci in the myometrium (Fig. 57-13).145 These are present in approximately 50% of patients with adenomyosis and may occasionally show high T1 signal consistent with presence of blood products.145 In contrast, focal adenomyosis appears as a localized T2 low-signal

CHAPTER 57  Female Pelvis

A

1987

B FIG 57-12  Leiomyoma of the posterior cervix. A, Sagittal T2-weighted image demonstrates a large hypointense mass (arrow) protruding from the cervicoisthmic junction into the pouch of Douglas, suggesting a pedunculated leiomyoma. B, T1-weighted image with fat suppression shows moderate enhancement compared to the normal myometrium and a central stalk (arrowhead) arising from the cervical stroma.

mass (adenomyoma) within the myometrium, arising from the junctional zone with ill defined edges.173 Focal adenomyosis may occur as an isolated entity or with the diffuse form of adenomyosis.

Endometrial Polyps.  Endometrial polyps are of variable size, project into the endometrial cavity, and consist of dense stromal tissue, thickwalled vascular channels, and endometrial glands. Polyps typically demonstrate a central fibrous core, intralesional cysts, and areas of focal hemorrhage. On MRI, polyps appear T2 isointense to hypointense and demonstrate postcontrast enhancement.67 Differential diagnosis includes submucosal leiomyomas and endometrial cancer.

Endometrial Hyperplasia.  Endometrial hyperplasia refers to diffuse endometrial thickening due to estrogen stimulation caused by chronic anovulatory status, internal or external estrogen influence, obesity, or tamoxifen therapy. Tamoxifen therapy has also been associated with endometrial polyps, adenomyosis, and endometrial cancer.5 Imaging features of endometrial hyperplasia include irregular thickening of the endometrium, with T2 hyperintensity compared to normal myometrium and T1 hypointensity relative to myometrium on early contrastenhanced images.178 Endometrial thickening measuring 5 mm or greater in postmenopausal females with bleeding, regardless of hormonal therapy, should undergo further investigations including tissue sampling to exclude endometrial cancer.123

Benign Cervical Lesions.  Nabothian cysts are a common incidental finding that represent retention cysts of the cervical glands, often due to chronic inflammation with scarring. MRI imaging findings demonstrate well-circumscribed rounded cystic lesions with T2-hyperintense and variable T1 signal.15 Cervical polyps are primarily diagnosed on hysteroscopy, although they can be seen on MRI as a polypoid lesion with or without hyperintense T2 signal cystic components.173

Vascular Abnormalities.  Rarely, vascular malformations can occur within the uterus. These include arteriovenous malformations and venous malformations. Arteriovenous malformations can appear as a

focal uterine mass with serpiginous flow-related signal voids and prominent parametrial vessels.77 Venous malformations can be diagnosed on MRI, appearing as T2-hyperintense dilated veins with inhomogeneous enhancement due to thrombi or phleboliths.193 Venous malformations may be associated with pelvic venous congestion syndrome, which can be a cause of chronic pelvic pain. Other causes of pelvic congestion syndrome include any condition leading to dysfunctional venous valves, retrograde blood flow, venous hypertension, and dilatation of venous vessels.140 Imaging findings of dilated ovarian veins and tortuous pelvic and parametrial varices on CT and MRI are suggestive of pelvic venous congestion syndrome in the appropriate clinical setting.

Malignant Pathologic Conditions

Cancer of the Uterine Cervix.  Cervical cancer typically arises from the squamocolumnar junction (SCJ), with the majority of cancers being squamous cell carcinoma and only a small subset being adenocarcinoma or other rare subtypes (15%-20%).72 HPV infection is central to the development of cervical cancer, which can grow as an endophytic or exophytic mass. The normal cervix demonstrates a trilaminar pattern of signal intensity, with T2-hyperintense endocervical mucosal glands surrounded by T2-hypointense stroma and a rim  of intermediate-signal-intensity smooth muscle.54 Invasion of cervical cancer into the cervical stroma results in an enlarged barrel-shaped cervix with an abnormal intermediate- to high-intensity T2-signal mass that replaces the low-signal-intensity cervical stroma (Fig. 57-14). Tumors tend to enhance earlier than adjacent cervical stroma and may demonstrate restricted diffusion on DWI and ADC maps.34 Direct invasion into adjacent tissues and organs, lymphatic spread, and less commonly hematogenous spread occurs in cervical cancer.6 Regional lymph node metastases occur early in the course of the disease and may include any pelvic lymph nodes. Presence of lymphadenopathy is associated with a poor prognosis and alters treatment course but does not change the staging of cervical cancer.89 Staging of cervical cancer.  The International Federation of Gynecology and Obstetrics (FIGO) staging system is most commonly

1988

PART II  CT and MR Imaging of the Whole Body

A

B

* *

C FIG 57-13  MRIs from a 45-year-old patient with severe dysmenorrhea. A, Sagittal T2-weighted image of the uterus shows irregular thickening of the junctional zone (arrows), with several small subendometrial cysts suggesting adenomyosis. B, Axial T2-weighted image of the uterus confirms asymmetric thickening of the junctional zone, predominantly in the anterior myometrium. C, Axial T2-weighted image of the uterine isthmus shows two interstitial leiomyomas (asterisks) associated with adenomyosis, confirmed by subsequent hysterectomy. Arrowhead indicates an ovarian follicle cyst.

used and was recently revised in 2009 (Box 57-1). FIGO recommends a clinical staging system for cervical carcinoma that includes inspection, palpation, colposcopy, hysteroscopy, endocervical curettage, cystoscopy, proctoscopy, IV urography, and radiographic evaluation of lungs and skeleton. A major change in the new FIGO staging system is that the use of diagnostic imaging including CT and MRI is  now recommended if available, although not mandatory. In addition, examinations performed under anesthesia (including cystoscopy, proctoscopy) are no longer mandatory and are purely optional.54 Additional minor changes to stage groupings were also made and are discussed below. Role of imaging.  CT is of limited use in local staging, but it is able to depict extrauterine spread of disease, including enlarged lymph nodes, fistulation into the bladder or rectum, and distant metastases.

On CT, cervical cancers can have variable enhancement patterns from hypoattenuation to hyperattenuation.138 Identification of the primary cancer in a normal cervix is often difficult on CT because 50% of tumors appear isodense to cervical stroma.137 In contrast, MRI has excellent soft tissue contrast resolution and can clearly define the local extent of primary tumor and depict metastatic spread.125 Angled transaxial and sagittal T2-weighted images are useful for tumor extension and detection of the cervical cancer. Cervical cancer typically appears as a T2-hyperintense lesion in the T2-hypointense cervical stroma. The majority of lesions are isointense on T1-weighted images but may demonstrate areas of necrosis or blood products. Postcontrast imaging shows abnormal enhancement, which can aid in the detection of necrosis, visualization of small cancers not identified on T2-weighted images, and demonstration of

CHAPTER 57  Female Pelvis

1989

*

A

B FIG 57-14  Cervical cancer in a 47-year-old patient. A, T2-weighted image shows a mass larger than 4 cm in diameter (arrow) in the posterior aspect of the cervix, protruding into the vagina. The vagina shows high signal intensity (asterisk) due to opacification with ultrasound gel. The normal low signal intensity of the posterior cervix is replaced by the intermediate-signal-intensity mass. B, Axial T2-weighted image shows a small peripheral rim of normal cervical stroma of low signal intensity (arrows), ruling out parametrial invasion.

BOX 57-1  FIGO Staging of Cervical Carcinoma* Stage I: The carcinoma is strictly confined to the cervix (extension to the corpus would be disregarded). IA: Invasive carcinoma that can be diagnosed only by microscopy, with deepest invasion ≤5 mm and largest extension ≥7 mm IA1: Measured stromal invasion of ≤3.0 mm in depth and extension of ≤7.0 mm IA2: Measured stromal invasion of >3.0 mm and 5 mm, with an extension ≤7.0 mm IB: Clinically visible lesions limited to the cervix uteri or preclinical cancers greater than stage IA IB1: Clinically visible lesion ≤4.0 cm in greatest dimension IB2: Clinically visible lesion >4.0 cm in greatest dimension Stage II: Cervical carcinoma invades beyond the uterus but not to the pelvic wall or to the lower third of the vagina. IIA: Without parametrial invasion IIA1: Clinically visible lesion ≤4.0 cm in greatest dimension

IIA2: Clinically visible lesion >4.0 cm in greatest dimension IIB: With obvious parametrial invasion Stage III: The tumor extends to the pelvic wall and/or involves lower third of the vagina and/or causes hydronephrosis or nonfunctioning kidney. IIIA: Tumor involves lower third of the vagina, with no extension to the pelvic wall IIIB: Extension to the pelvic wall and/or hydronephrosis or nonfunctioning kidney Stage IV: The carcinoma has extended beyond the true pelvis or has involved (biopsy proven) the mucosa of the bladder or rectum. A bullous edema, as such, does not permit a case to be allotted to stage IV. IVA: Spread of the growth to adjacent organs IVB: Spread to distant organs

*Reprinted with permission and modified from Freeman SJ, et al: The revised FIGO staging system for uterine malignancies: Implications for MR imaging. Radiographics 32:1805–1827, 2012.

fistulas in advanced stages.125 Vaginal opacification with ultrasound gel and dynamic contrast imaging facilitate detection of small cervical cancers.191 The ability to accurately identify tumor size, parametrial and pelvic sidewall invasion, bladder or rectal invasion, and lymph node metastases markedly affects management and treatment decisions.54 MRI has been shown to be highly accurate in the evaluation of parametrial invasion, with an accuracy rate of 88% to 97% and a specificity of

93%.156,161 Parametrial invasion also can be reliably excluded on MRI, with a negative predictive value of 94% to 100%.153 MRI is able to appropriately identify the size of the primary tumor with an accuracy of 93%.172 Local invasion into the bladder and rectum can be excluded on MRI with a negative predictive value of 100%.150 Stage I.  In stage IA, tumors are confined to the cervix and defined as a microscopic tumor that cannot be visualized on MRI. Therefore there is no role of MRI in stage IA tumors. However, MRI is

1990

PART II  CT and MR Imaging of the Whole Body

B

B

A

FIG 57-15  Stage IVA cervical cancer. A, Bulky enlargement of the cervix by a large mass of intermediate signal intensity is seen on the sagittal T2-weighted image. It extends to the middle third of the anterior vagina. Disruption of the low signal intensity of the posterior deep bladder wall and the irregular surface of the bladder (B) wall over a distance of 2.5 cm is shown (arrow). B, Angulated transaxial T2-weighted image shows an enlarged left iliac lymph node suggestive of metastasis (arrow). Cystoscopy confirmed mucosal bladder wall invasion.

recommended in stage IB tumors or higher, owing to the need for accurate measurements of tumor size, identification of parametrial invasion, lower vaginal involvement, and lymph node metastasis. Stage IB tumors are subdivided by size, with stage IB1 tumors measuring  less than 4 cm and stage IB2 tumors 4 cm or larger. On T2-weighted images, tumors typically demonstrate intermediate to high signal intensity compared with the cervical stroma, which demonstrates low signal intensity. A rim of surrounding T2 low-intensity cervical stroma should remain intact at this stage to indicate tumor confinement within the cervix. Stage II.  In stage II, tumors extend beyond the uterus and involve the upper two thirds of the vagina but do not extend to the pelvic sidewall or the lower one third of the vagina. The absence or presence of parametrial invasion further subdivides stage II into stage IIA or stage IIB, respectively. Involvement of the upper two thirds of the vagina is characterized by a T2-hyperintense lesion disrupting the T2-hypointense vaginal wall.54 Differentiation between an invasive cervical cancer into the upper two thirds of the vagina and an exophytic polypoid cervical tumor, which may widen the vaginal fornix but not infiltrate it, is essential. The low-signal-intensity vaginal wall should remain intact in an exophytic polypoid cervical tumor.54 Parametrial invasion should be diagnosed when there is a disruption of the lowsignal-intensity cervical stromal ring, with nodular or irregular tumor extending into the parametrium.125 Additional features suggesting parametrial invasion include segmental disruption with a spiculated tumor-parametrium interface, soft tissue extension into the parametrium, and encasement of the periuterine vessels.89 Conversely, parametrial invasion can be confidently excluded if the T2-hypointense cervical stromal rim is thicker than 3 mm, with a specificity of 99%.204 Occasionally indistinct cervical stroma due the presence of peritumoral edema in some patients may overestimate parametrial extent of

disease.54 Other factors that may limit MRI evaluation of parametrial invasion are postbiopsy changes and cervicitis.99 Stage III and IV.  In stage IIIA, tumors extend to the lower one third of the vagina but not the pelvic sidewall. Stage IIIB is defined by extension to the pelvic sidewall or involvement of the ureters, which causes hydronephrosis. Stage IVA reflects local pelvic organ invasion, which is characterized by infiltration of the rectal mucosa or urinary bladder (Fig. 57-15). On T2-weighted images, rectal invasion is indicated by segmental disruption of the hypointense muscularis layer by the hyperintense tumor. T2-hyperintense thickening of the bladder mucosa is often due to bullous edema and should not be mistaken for direct bladder wall invasion. Contrast-enhanced MRI can help differentiate between bullous edema of the bladder and bladder wall invasion.192 In stage IVB, tumors spread beyond the pelvis to distant organs. Differential diagnosis.  Endometrial cancer with cervical invasion may have a similar appearance to primary cervical cancer, although usually the bulk of the tumor is centered in the endometrium rather than the cervix. Endometrial polyps or leiomyomas that protrude into the cervical canal can be differentiated by demonstration of a stalk arising from the endometrial cavity. Metastasis from breast or ovarian malignancy or melanoma may have a similar appearance and should be considered in the appropriate clinical context. Adenoma malignum is a rare subtype of mucinous adenocarcinoma of the cervix and should be considered when there is a predominantly complex multicystic mass with enhancing walls or solid components.44,127

Endometrial Cancer.  Endometrial cancer is the most common gynecologic malignancy in industrialized countries, including the United States. The most common clinical manifestation is abnormal uterine bleeding in a postmenopausal patient, which allows the majority of endometrial cancers to be diagnosed at stage I. Postmenopausal

CHAPTER 57  Female Pelvis BOX 57-2  FIGO Staging of Endometrial

Carcinoma*

Stage I: Tumor confined to the uterus IA: <50% invasion of the myometrium IB: ≥50% invasion of the myometrium Stage II: Tumor invades the cervical stroma but does not extend beyond the uterus Stage III: Local or regional spread of tumor IIIA: Serosal or adnexal invasion IIIB: Vaginal or parametrial involvement IIIC: Metastasis to pelvic or paraaortic lymph nodes IIIC1: Pelvic lymph node involvement IIIC2: Paraaortic lymph node involvement (with or without pelvic nodes) Stage IV: Extension into the pelvic wall, lower one third of the vagina, or hydronephrosis or nonfunctioning kidney IVA: Invasion of bladder or bowel mucosa IVB: Distant metastases, including abdominal, or involvement of inguinal lymph nodes *Reprinted with permission from Freeman SJ, et al: The revised FIGO staging system for uterine malignancies: Implications for MR imaging. Radiographics 32:1805–1827, 2012.

patients with vaginal bleeding should therefore undergo a pelvic US and tissue biopsy if the endometrial lining measures greater than 4 mm.18 There are two histologic subtypes: endometrioid adenocarcinoma, accounting for 90% of cases (type 1), and serous papillary or clear cell adenocarcinoma (type 2).54 Staging and role of imaging.  FIGO staging for endometrial cancer remains surgical because treatment remains predominantly surgical, including hysterectomy, bilateral salpingo-oophorectomy, node dissection, peritoneal washing, and an omental biopsy108 (Box 57-2). Therefore current National Comprehensive Cancer Network guidelines only recommend preoperative chest radiography and MRI if gross cervical invasion is suspected.124 However, these guidelines and FIGO staging do not take into account the vast amount of information that may be gained from preoperative MRI. Preoperative MRI in combination with results from endometrial biopsy can effectively stratify patients in differing risk categories, which may affect management and treatment decisions. Specifically, patients at low or intermediate risk may be treated with a minimally invasive laparoscopic hysterectomy and bilateral salpingo-oophorectomy, as compared to a more aggressive surgical staging including a lymphadenectomy. Minimally invasive laparoscopic hysterectomies have shown decreased morbidity and hospital stays when compared to more extensive surgical resections, while producing similar outcomes when comparing both types of treatment.144 Preoperative MRI adds valuable information such as identification of suspicious lymph nodes that may be otherwise unsuspected in  low- or intermediate-risk patients. In addition, MRI can confidently identify greater than 50% depth of myometrial invasion or cervical stromal invasion, which confers a higher risk for lymphovascular space invasion, resulting in lymph node metastases.104,151 Furthermore, MRI can provide a preoperative road map regarding resectability and the need for colorectal or urologic surgical intervention or adjuvant chemoradiotherapy.54 Owing to these advantages, several European guidelines now incorporate MRI for staging purposes, including current recommendations by the National Cancer Institute of France, the European Society of Radiology Guidelines, and the Royal College of Radiologists.78,100,143

1991

Imaging findings.  Endometrial carcinoma is T1 isointense and T2 hypointense relative to the normal endometrium, with relatively less enhancement compared to normal myometrium. Endometrial  carcinoma demonstrates restricted diffusion on DWI and ADC maps (Fig. 57-16).54 Stage I.  Stage I tumors are confined to the uterus and further subdivided by the amount of invasion into the myometrium (≤50% invasion) (Fig. 57-17). The sensitivity and specificity of MRI for depicting the depth of myometrial invasion vary between 69% to 94% and 64% to 100%, respectively.170 Confounding factors include the presence of fibroids and/or adenomyosis, although DWI and contrastenhanced sequences are helpful in differentiating between these entities.54 Stage II.  Stage II represents stromal invasion of the cervix, which is identified as a T2-hyperintense lesion within the T2-hypointense cervical stroma. Occasionally a stage I tumor that enters the endocervical canal may mimic a stage II cancer, although preservation of the normal T2-hypointense cervical stroma can help differentiate between the two entities.54 Stage III.  Stage III is defined by local or regional tumor spread beyond the uterus but not outside the true pelvis and is further subdivided by the extent of locoregional invasion and presence of lymphadenopathy. MRI can help identify tumors that have invaded through the serosa, which disrupts the contour of the outer myometrium or involves the adnexa (stage IIIA), and help identify tumor extension into the parametrium or show vaginal involvement (stage IIIB).54 Stage IIIC disease is characterized by lymph node involvement. Suspicious nodal features on MRI include a short-axis diameter greater than 10 mm, multiplicity, irregular contours, internal necrosis, or abnormal signal within the lymph node that is similar to the primary tumor.54 Positron emission tomography (PET) and CT are also helpful in identifying abnormal lymph nodes. Stage IV.  Stage IV disease is defined by direct invasion of the bladder or rectal mucosa or the presence of distant metastases. Extension of tumor into the normally T2-hyperintense bladder or rectal mucosa is indicative of endometrial tumor invasion (stage IVA).54 However, similar to cervical cancer, bullous edema—which appears  as thickening of the high-signal-intensity mucosal layer—should  not be confused with mucosal invasion. In stage IVB disease, distant metastasis includes peritoneal spread (omental caking, malignant ascites), paraaortic lymphadenopathy, and inguinal lymph node metastases.40 Rare Uterine Malignancies Uterine sarcomas.  Uterine sarcomas comprise a rare heterogeneous group of tumors of mesenchymal origin and represent approximately 8% of uterine malignancies.199 They may be classified as nonepithelial or mixed epithelial-nonepithelial tumors and include leiomyosarcoma, endometrial stromal sarcoma, undifferentiated endometrial sarcoma, and adenosarcoma.159 Typically these tumors are associated with a poorer prognosis due to their aggressive nature and commonly present as a rapidly growing pelvic mass that may be accompanied with pain or bleeding. Leiomyosarcoma is the most common histologic subtype of uterine sarcomas, and less than 0.2% of lesions arise from sarcomatous transformation of a benign leiomyoma.155 On MRI, leiomyosarcomas appear as an infiltrating heterogeneously enhancing uterine mass with T1 heterogeneous hypointense signal and irregular and ill-defined margins. On T2-weighted images, lesions are of intermediate to high signal intensity and may have areas of central necrosis, hemorrhage, or foci of calcifications.159 Differentiation of a leiomyosarcoma from a rapidly growing leiomyoma remains difficult because there is a large overlap in imaging features.

1992

PART II  CT and MR Imaging of the Whole Body

A B

C

D

E FIG 57-16  Endometrial cancer. A, T2-weighted image demonstrates a T2-hypointense mass filling the majority of the endometrial cavity, which demonstrates T1-isointense signal (B). C, Mild postcontrast enhancement of the endometrial lesion. D and E, Sagittal DWI and corresponding ADC map demonstrate restricted diffusion. (Courtesy Dr. Wes Carter, Jacksonville, Fla.)

CHAPTER 57  Female Pelvis

*

*

A

B

*

C FIG 57-17  MRIs from a 61-year-old patient with postmenopausal bleeding and positive endometrial biopsy for high-grade adenocarcinoma. A, Coronal T2-weighted image demonstrates an irregular interface between the enlarged uterine cavity (asterisk) and the myometrium. The right ovary shows a small round mass (arrow) of low signal intensity. B, Axial oblique 3D contrast-enhanced image 2 minutes after contrast injection at the same level confirms the irregular interface between the strongly enhancing normal myometrium and the less enhancing endometrial tumor (asterisk) invading the myometrium superficially. The right ovarian mass demonstrates very little enhancement (arrow). C, Sagittal 3D contrast-enhanced image 6 minutes after contrast injection shows a hypointense endometrial tumor (asterisk) invading the posterior and anterior myometrium in less than 50% of the entire myometrial thickness. Well-circumscribed, round, hypointense areas in the cervix and endocervical junction correspond to nabothian cysts. An enhancing double-layer line within the endometrial channel (arrow) represents a benign endocervical polyp. Pathology confirmed superficial myometrial invasion (stage IB), the coexistence of a benign endocervical polyp, and a small fibroma of the right ovary.

1993

1994

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 57-18  Uterine infiltration by metastatic breast cancer in a patient with progressive uterine enlargement on sonography. A, Diffuse enlargement of the uterus, broadening of the myometrium, and an irregular endomyometrial interface is demonstrated on CT. B, Sagittal T2-weighted image shows low signal intensity of the diffusely broadened myometrium and irregular bandlike areas of high signal intensity in the deep myometrium at the fundus. Histopathology revealed diffuse infiltration of the uterus by metastatic breast cancer.

Uterine carcinosarcoma.  Carcinosarcomas (also known as malignant mixed müllerian tumors) are rare highly aggressive mixed epithelial and mesenchymal malignancies that were previously classified as a subtype of sarcoma. They typically present as a polypoid bulky mass that fills the entire uterine cavity and demonstrates myometrial invasion and extension beyond the uterus.14 Occasionally the tumor may demonstrate a hemorrhagic and/or necrotic component.

Other Malignancies.  Rarely, primitive neuroectodermal tumor, lymphoma, and metastasis can be found in the uterus (Fig. 57-18).

OVARIES Normal Anatomy The ovaries are almond-shaped structures whose normal appearance and size varies depending on age, hormonal influence, and phase of the menstrual cycle.10,56,163 The adult ovary is approximately 2.5 to 5 cm long, 1.5 to 3 cm wide and 1 to 2 cm thick.66,195 Ovarian volume is preferred when assessing size and can be calculated with the formula for a prolate ellipse (0.523 × length × width × thickness).195 Normal ovaries are located in a peritoneal depression called the ovarian fossa, lateral to the uterus and between the internal and external iliac fossa.114,146 They are anchored in place by three structures: the mesoovarium, which anchors the ovary to the posterior surface of the  broad ligament; the utero-ovarian ligament (ovarian ligament), which anchors the ovary to the uterus; and the suspensory ligament, which anchors the ovary to the pelvic sidewall.43,146,195 The utero-ovarian ligament and suspensory ligament of the ovary have variable degrees of laxity, so ovarian location can vary from patient to patient.43,66 During pregnancy the ovaries are pushed into the abdomen by the enlarging

uterus, and after delivery the ovaries often do not return to their original position.154 In women of reproductive age, the ovaries exhibit a zonal anatomy on T2-weighted MRI. The cortex demonstrates lower signal intensity than the central medulla. The peripheral zone contains multiple cysts representing different stages of folliculogenesis, corpus luteum cysts, and surface inclusion cysts. Most cysts are thin walled with intermediate to low signal on T1-weighted images and high signal on T2-weighted images, reflecting simple fluid content.129,179 The dominant follicle enlarges to approximately 25 mm and after ovulation becomes the corpus luteum. Corpus luteum cysts demonstrate thicker, irregular, avidly enhancing walls and may contain more complex fluid signal intensity secondary to blood products (Fig. 57-19).129 On precontrast T1-weighted images the ovary is isointense to uterine myometrium but enhances less avidly than uterine myometrium following contrast administration. Alternatively, normal postmenopausal ovaries demonstrate lower signal on T2-weighted images without the presence of cysts.129

Benign Cystic Lesions Benign conditions of the ovary constitute the majority of pathologic and physiologic conditions. CT and MRI appearance of benign conditions vary significantly.

Physiologic Ovarian Cyst.  Functional cysts represent the majority of ovarian cysts and result from failure of an ovarian follicle to rupture or regress. Although the majority of follicular cysts are less than 5 cm in diameter, they can range in size from 3 cm to 8 cm.129 Physiologic cysts are extremely common in women of reproductive age, more so in the adolescent and early postmenopausal periods.

CHAPTER 57  Female Pelvis

FIG 57-19  Corpus luteum cyst in a 23-year-old woman with left lower quadrant pain demonstrates thick irregular wall with avidly enhancing walls. A small amount of surrounding free fluid suggests recent rupture.

They can also be seen in women on hormonal replacement therapy. Most cysts are asymptomatic and hormonally inactive. Complications of physiologic cysts include rupture, hemorrhage, and torsion. Follicular cysts in precocious puberty are a manifestation of McCuneAlbright syndrome.35 Physiologic cysts are unilocular and low density on CT. On MRI they demonstrate low signal on T1-weighted images and high signal on T2-weighted images. Physiologic cysts may demonstrate a thin enhancing wall on MRI, but no internal enhancement (Fig. 57-20).129 A physiologic cyst or corpus luteum cyst complicated by internal  hemorrhage is referred to as a hemorrhagic cyst. A hemorrhagic cyst demonstrates varying degrees of increased attenuation on CT and occasionally a fluid-fluid level depending on the stage of hemorrhage.50,70 On MRI a hemorrhagic cyst will often demonstrate high signal intensity on T1-weighted images and intermediate to high signal on T2-weighted images (Fig. 57-21).50 Uncommonly a physiologic or hemorrhagic cyst may rupture, resulting in severe abdominal pain, occasionally presenting in a catastrophic manner with shock secondary to massive intraperitoneal hemorrhage. CT of a ruptured cyst demonstrates high-attenuation

* A

1995

B

C FIG 57-20  Functional cyst in a 29-year-old woman. A 7-cm cystic lesion of the right ovary exhibits very high signal intensity on the T2-weighted image (A), intermediate signal intensity on the T1-weighted image (B), and a thin wall on T1 weighting with fat suppression after contrast medium administration (C) and on the T2 image (A). Minimal ascites (asterisk in A) is present in the cul-de-sac. Imaging could not differentiate a functional cyst from a cystadenoma; however, sonographic follow-up showed a considerable decrease in size within 3 months.

1996

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 57-21  Hemorrhagic cyst in a 30-year-old woman. A 3-cm cystic lesion of the left ovary demonstrates intermediate high signal on the T2-weighted image (A), high signal on the T1-weighted image (B), and no enhancement on T1 weighting with fat suppression after contrast administration (C). Follow-up US demonstrated resolution of the cystic lesion.

fluid in the pelvis, which can extend into the upper abdomen. The residual cyst may or may not be apparent depending on the amount of cyst decompression at the time of rupture. Active extravasation of IV contrast material can help identify the bleeding site.70

Peritoneal Inclusion Cyst.  Peritoneal inclusion cysts, also known as an entrapped ovary, pseudocyst, or multilocular inclusion cyst, are the result of active ovaries, pelvic adhesions, and impaired absorption of peritoneal fluid.82,119 They occur almost exclusively in premenopausal women, usually as a result of endometriosis, pelvic inflammatory disease (PID), or prior pelvic surgery.119 Pelvic inclusion cysts range from a few millimeters to large cystic lesions.82 Although pelvic inclusion cysts can cause pain and/or abdominal bloating, they are often found incidentally.119 At imaging, peritoneal inclusion cysts are unilocular or multilocular cystic adnexal masses that conform to the shape of the peritoneal cavity. Finding the ovary within and along the periphery of a cystic adnexal mass is a highly specific finding.179 Most peritoneal inclusion cysts contain hyperintense signal on T2-weighted images, hypointense signal on T1-weighted images, and are low density on CT. Occasionally, internal hemorrhage can cause high signal on T1-weighted images.82 Although there can be thin internal septations, there is no wall enhancement (Fig. 57-22).95 Peritoneal inclusion cysts may recur after surgery or drainage.107

*

FIG 57-22  Peritoneal inclusion cyst in a 32-year-old patient with a history of PID. CT scan demonstrates a cystic pelvic mass with fine septations and adhesions to the adjacent uterus (arrow). Lack of enhancing wall structures and demonstration of the centrally entrapped ovary (asterisk) are findings typical of a peritoneal pseudocyst. In this patient, sonographic follow-up showed reduction of size, and the lesion had resolved after 1 year.

CHAPTER 57  Female Pelvis Ovarian Hyperstimulation Syndrome.  Ovarian hyperstimulation syndrome, also known as hyperreactio luteinalis, is caused by disorders associated with increased human chorionic gonadotropin (hCG) levels, such as a hydatid mole, choriocarcinoma, multiple pregnancies, and assisted reproduction. As many as 65% of patients receiving gonadotropins for ovarian stimulation can develop ovarian hyperstimulation syndrome.35,179 Typical imaging findings are bilaterally enlarged ovaries (up to 20 cm) containing multiple variable-sized theca-lutein cysts. On CT and MRI the cysts demonstrate simple fluid density and signal without mural thickening or nodularity.58,179 Severe cases can result in acute Meigs’ syndrome with ascites and pleural effusions.35

Polycystic Ovary Syndrome.  Polycystic ovary syndrome (PCOS), also known as Stein-Leventhal syndrome, is an endocrine disorder characterized by hyperandrogenism and chronic anovulation.35 PCOS is the most common endocrine abnormality in women of reproductive age, affecting 4 to 5 million women in the United States.8,105 Severe insulin resistance and defects in insulin secretion are thought to be at the center of the syndrome.46 Besides infertility, women are at increased risk for hypertension, hyperlipidemia, type 2 diabetes mellitus, coronary artery disease, and cerebral vascular disease.105 Only half of patients with PCOS present with the classic triad of amenorrhea, hirsutism, and obesity.35 Because of multisystem involvement, imaging findings alone are insufficient for the diagnosis.105 The American College of Obstetricians and Gynecologists (ACOG) guidelines describes two criteria for polycystic ovaries: ovarian volume and number of follicles. Polycystic ovaries are present when one or both of the ovaries demonstrate 12 or more follicles measuring 2 to 9 mm in diameter or the ovarian volume exceeds 10 cm3.9 MRI can be used to complement US in the setting of PCOS. MRI findings of PCOS include enlarged ovaries and multiple 2- to 9-mm follicles in a subcapsular location.9,179 MRI can also be used to exclude a virilizing ovarian or adrenal tumor.156,181

Endometriosis.  Endometriosis occurs when endometrial tissue implants outside the uterine cavity and responds to cyclic hormonal stimulation.50,110 It affects as many as 10% of reproductive-aged women and is the most common cause for chronic pelvic pain in this population. The pathogenesis of endometriosis remains debated. Pain associated with endometriosis is initially treated with antiinflammatory agents and hormonal therapy.59 Surgical procedures may also be performed. It is estimated that 30% to 50% of patients with endometriosis are infertile; infertility is treated surgically.59 Pelvic endometriosis, especially in severe stages, is strongly associated with adenomyosis.49 Various forms of endometriosis exist, including peritoneal endometriosis and adhesions, deep infiltrating endometriosis, and tubal endometriosis. At MRI, adhesions can appear as spiculated lowsignal-intensity stranding on T1- and T2-weighted images. More often, only secondary signs are present, including angulation of bowel loops, elevation of the posterior vaginal fornix, or posterior displacement of the uterus and/or ovaries with loss of fat planes between the structures. Endometrial deposits can be located virtually anywhere but are commonly found in the anterior or posterior compartments. Common sites of involvement include the bladder, ureters, vagina, rectovaginal septum, uterosacral ligaments, rectum, and rectosigmoid junction.49 The most common manifestation of endometriosis is endometrial implants on the ovaries, also called endometriomas.167 Bilateral ovarian involvement occurs in one third to half of patients.36 Endometriomas present as unilateral or bilateral complex hemorrhagic lesions filled with blood products of different ages. The cyst wall is fibrotic, resulting

1997

in dense adhesions to the surrounding structures. The majority of endometriomas range in size from 3 to 6 cm and rarely exceed 15 cm.101,187 Classically, endometriomas demonstrate high signal intensity on T1-weighted images and low signal intensity on T2-weighted images; however, they can demonstrate high signal intensity on both T1- and T2-weighted images depending on the chronicity of blood products. The finding of low signal intensity on T2-weighted images is known as T2 shading.60 This results from T2 shortening secondary to chronic blood products and is specific to endometrioma, assisting in differentiating it from a hemorrhagic cyst.185 Endometriomas have a tendency to rupture. The cysts tend to subsequently reaccumulate blood and rupture again, resulting in multiplicity, a feature also specific for endometriosis (Fig. 57-23).50,167 Using the criteria of hyperintense lesions on T1-weighted images, shading on T2-weighted images, and multiple lesions, a sensitivity of 90% and specificity of 98% can be achieved in the diagnosis of endometriomas using MRI.185 Endometriomas can have a fibrous capsule, demonstrated by a lowsignal-intensity rim on T1- and T2-weighted images, with delayed capsular enhancement or enhancement of the adjacent structures.50 They may restrict diffusion on DWI, a nonspecific finding because it can be seen with both malignant and benign ovarian lesions.167 Endometriosis is a frequent cause of hydrosalpinx, and hematosalpinx is an ancillary finding suggestive of endometriosis.130 Contrast enhancement of mural nodules within an endometrioma is suggestive of malignant transformation, most commonly endometrioid and clear cell carcinoma.182,198 However, during pregnancy endometriosis can undergo decidualization due to hormonal changes, resulting in enhancing mural nodules that can mimic ovarian carcinoma.80,158

Neoplastic Conditions Ovarian neoplasms are classified by cell of origin (epithelial tumors, germ cell tumors, sex cord–stromal tumors, and metastatic tumors) rather than by malignant potential. Benign disease is more common than malignant disease; however, there is considerable overlap in the appearance of both benign and malignant neoplasms, making ovarian neoplasm a diagnostic challenge.39 Ovarian cancer is the fifth most common cause of cancer death in women in the United States, causing more deaths than any other gynecologic cancer. Most ovarian cancers present at a late stage after the disease has spread beyond the pelvis.2,83,111 Preoperative characterization of ovarian masses is of utmost im­ portance, enabling the surgeon to anticipate malignancy and plan appropriately.

Role of Imaging.  The role of diagnostic imaging in ovarian cancer is multifold: characterize the adnexal mass, determine the extent of preoperative disease, predict resectability, and evaluate treatment response.29,84,86 Whereas CT is more accurate in assessing extent of disease, MRI is more accurate in characterization of ovarian tumors, owing to its excellent contrast resolution.52,86,164 Thus CT is the imaging modality of choice for staging of ovarian cancer, and MRI is often reserved for mass characterization and problem solving or for patients who cannot undergo CT.51

Staging and Treatment.  Choice of treatment depends on accurate staging. Ovarian cancer is classified according to the FIGO and TNM staging systems.13 Stage I disease is limited to one or both ovaries. Stage II disease has spread to the surface of other pelvic organs. Stage III disease indicates spread to lymph nodes or abdominal peritoneal surfaces, and stage IV disease is distant metastases to solid organs outside the abdomen (Table 57-2).29,84 Cytoreductive surgery is the standard treatment for ovarian cancer. Imaging is used to define extent of

1998

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 57-23  Bilateral endometriomas with hemorrhagic content of various ages. A, Axial T2-weighted image demonstrates a slightly heterogeneous hyperintense cyst of the right ovary surrounded by numerous small follicles. The left ovary has a homogeneous hyperintense structure and a cyst with a heterogeneous fluidfluid level and shading (arrow). B, Corresponding T1-weighted image shows bilateral hyperintense cysts. C, T1-weighted image with fat suppression confirms the presence of blood and the likelihood of bilateral endometriomas.

disease, assess likelihood of optimal primary cytoreduction, and identify patients who may benefit from neoadjuvant chemotherapy.52,75,183

Recurrence.  Although initial treatment response is often good, recurrence is common, with an overall survival rate of 10% to 30% in patients with advanced cancer.25,136 The vaginal vault and cul-de-sac are the most common sites for recurrence, followed by abdominal peritoneal implants. Small and large bowel obstructions are a common complication in patients with recurrent disease and are the primary cause of mortality.54 Patients with ovarian cancer are followed with serial CA-125 measurements. A rising CA-125 in a patient in remission is highly predictive of recurrence and can precede imaging findings by 4 to 6 months.136 On CT and MRI, contrast enhancement is helpful in distinguishing peritoneal deposits from bowel loops. The reported

accuracy of MRI for detecting lesions greater than 2 cm is comparable to that of CT.52 However, fusion PET/CT has demonstrated higher accuracy for detecting recurrent disease when compared to CT and PET alone, with a sensitivity of 95% to 97% and specificity of 80% to 100%.121

Epithelial Tumors of the Ovaries.  Epithelial tumors are primarily cystic and account for 60% of all ovarian neoplasms and 85% of malignant neoplasms.39 Epithelial tumor prevalence increases with age, peaking in the sixth and seventh decades. Subtypes of epithelial tumors include serous, mucinous, endometrioid, clear cell, and Brenner’s tumors. Epithelial tumors are classified as benign (60%), malignant (35%), or borderline (5%) based on their histologic characteristics and clinical behavior.91

CHAPTER 57  Female Pelvis

1999

TABLE 57-2  FIGO Staging of

Ovarian Cancer Stage

Imaging Findings

I       II       III

Tumor limited to the ovaries Limited to one ovary, no ascites Limited to both ovaries, no ascites Stage IA or IB with ascites Tumor involving one or both ovaries, with pelvic extension Extension or metastases to the uterus or fallopian tubes Extension to other pelvic tissues Stage IIA or IIB with ascites Tumor involving one or both ovaries, peritoneal implants outside the pelvis, or implants in retroperitoneal or inguinal lymph nodes Tumor grossly limited to the true pelvis, large volume of ascites ≤2-cm implants in abdominal peritoneal surfaces >2-cm implants in abdominal peritoneal surface or retroperitoneal or inguinal lymph nodes Growth involving one or both ovaries, distant metastases, parenchymal liver metastases, pleural effusion with positive cytology

IA IB IC IIA IIB IIC

  IIIA   IIIB   IIIC IV

Modified from Prat J; FIGO Committee on Gynecologic Oncology: Staging classification for cancer of the ovary, fallopian tube, and peritoneum. Int J Gynaecol Obstet 124:1–5, 2014.

FIG 57-24  Serous cystadenoma in a 52-year-old patient. A 15-cm complex cystic lesion filling the pelvis demonstrates multiple thin septations with uniform signal on T2-weighted MRI. Lack of internal solid papillary projections suggests the neoplasm is benign. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

Imaging features most suggestive of benignity include entirely cystic internal contents, wall thickness less than 3 mm, lack of internal solid components, and absence of ascites.126 Multilocularity and wall thickening can occur in both benign and malignant masses and are  less reliable imaging findings.128 Imaging features more suggestive of malignancy include thick or irregular walls, thick septations, solid internal papillary projections, and large soft tissue components with necrosis.55,126 Secondary findings of pelvic organ invasion, peritoneal implants, ascites, and adenopathy are more specific for malignancy.84

Serous and Mucinous Tumor.  Serous and mucinous tumors are the most common epithelial tumors independent of malignancy.91 Although serous and mucinous tumors have different pathologic features, disease course, and prognosis, they can be difficult to differentiate based on gross inspection and imaging findings. However, imaging features can aid in differentiation as well as provide information about malignant potential.39,86 Serous tumors are the most common benign and malignant epithelial neoplasm, accounting for 60% to 80% of all epithelial neoplasms. Approximately 25% of serous tumors are malignant.194 On imaging, serous tumors are unilocular or multilocular, with homogeneous density on CT and homogeneous T1 and T2 signal intensity on MRI (Fig. 57-24).86 Psammoma bodies (microscopic calcifications) can be seen in up to 30% of malignant serous tumors, and calcifications on CT can be seen in up to 12% of tumors; however, calcifications  are nonspecific and can also be seen in benign serous tumors as well as a variety of other ovarian neoplasms.116 Thick or irregular walls, thick septations, solid internal papillary projections, and large soft tissue components with necrosis are more indicative of malignancy (Fig. 57-25).126 Mucinous tumors are less likely to be bilateral and tend to be larger at presentation.84,91 They are also more likely to be borderline or malignant.86,91 Mucinous tumors are classically multilocular with mucoid material, hemorrhage, or cellular debris within the cysts.91 These

FIG 57-25  Serous cystadenocarcinoma in a 50-year-old woman. T2-weighted images demonstrate a multilocular mass with uniform T2 hyperintensity within the locules. In contrast to a serous cystadenoma, the presence of a large solid papillary projection as well as thickened septations is indicative of malignancy. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

pathologic findings result in varying density on CT or signal intensity on MRI within the different locules, such that locules with thicker mucin demonstrate increased density or increased T1-weighted signal and decreased T2-weighted signal compared to locules with more watery mucin.86 This is often referred to as a “stained-glass” appearance (Fig. 57-26).39 Borderline and malignant mucinous tumors are one

2000

PART II  CT and MR Imaging of the Whole Body

FIG 57-27  Pseudomyxoma peritonei in a 48-year-old woman with bor-

FIG 57-26  “Stained glass” appearance of a mucinous cystadenoma. In contrast to a serous cystadenoma, this multilocular mass demonstrates varying signal intensity within different locules on T1-weighted images, corresponding with varying amounts of mucin. Thin septations and lack of internal solid papillary projections are consistent with a benign neoplasm. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

source of pseudomyxoma peritonei—implants of mucinous tumor on peritoneal surfaces. The MRI appearance of pseudomyxoma peritonei is similar to that of the primary mucinous ovarian neoplasm, with varying signal intensity depending on the thickness of mucin (Fig. 57-27).86 Pseudomyxoma peritonei may also be caused by benign and malignant mucin-secreting tumors of the appendix and large and small bowel cancer.196

Endometrioid Carcinoma.  Endometrioid carcinoma accounts for 10% to 15% of ovarian malignancies and is almost always malignant.135,194 From 30% to 50% of cases are bilateral, and up to a third are associated with synchronous endometrial hyperplasia or carcinoma.23,194 Endometrioid carcinoma is the most common malignancy arising within an endometrioma.182 Imaging characteristics of endometrioid carcinoma are nonspecific, but they tend to have varying degrees of cystic and solid components and are occasionally completely solid (Fig. 57-28).194

Clear Cell Carcinoma.  Clear cell carcinomas account for approximately 5% of ovarian carcinomas and are always malignant. The majority of clear cell carcinoma is stage I at the time of diagnosis and tends to have a better prognosis than other ovarian malignancies.194 Clear cell carcinoma is the second most common malignancy occurring in patients with endometriosis.86 The most common appearance of a clear cell carcinoma is a large smooth-walled unilocular cyst with solid protrusions. Internal signal intensity on T1-weighted images varies from high to low. Solid protrusions are usually small and few in number (Fig. 57-29).113 A serous tumor with low malignant potential should remain within the differential.86

Brenner’s Tumors.  Brenner’s tumors are rare lesions, representing 2% to 3% of all ovarian tumors, and are composed of transitional cells and dense stroma. Brenner’s tumors are rarely malignant, usually less than 2 cm, and typically discovered incidentally. In up to 30% of cases,

derline mucinous neoplasm. T2-hyperintense peritoneal implants are present throughout the abdomen, resulting in scalloping of the liver surface. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

Brenner’s tumors can be associated with other ovarian tumors.117,132 A Brenner’s tumor appears as either a multilocular cystic mass with a solid component or as a small predominately solid mass. The mass demonstrates mild or moderate enhancement at CT. Extensive calcification within the solid component is often present. On MRI the dense stroma demonstrates low T2-weighted signal similar to that of a fibroma117,132 (Fig. 57-30).

Germ Cell Tumors of the Ovaries.  Tumors of germ cell origin are the second most common type of ovarian neoplasms, accounting for 10% to 15%. The group includes mature teratoma, immature teratoma, dysgerminoma, endodermal sinus tumor, embryonal carcinoma, and choriocarcinoma.17,142 Mature teratoma is the most common of the group and the only benign neoplasm. The remaining tumors account for less than 5% of all malignant ovarian tumors. Malignant germ cell tumors are generally large and predominately solid. Elevated levels of serum α-fetoprotein (AFP) and hCG can be helpful in the diagnosis.17,38,131 Mature teratoma.  Mature teratomas are the most common benign ovarian neoplasm, accounting for 20% of adult tumors and 50% of pediatric tumors.139 Teratomas are composed of mature tissue from two or more embryonic germ cell layers. Gross pathologic examination reveals a unilocular mass, usually filled with sebaceous material and lined by squamous epithelium.38 Hair follicles, skin glands, muscle, and other tissue can be seen in the wall. A raised protuberance projecting into the cyst cavity (known as the Rokitansky nodule or dermoid plug) contains most of the hair, teeth, and fat.38,131 Mature teratomas demonstrate a wide range of imaging characteristics on both CT and MRI. On CT, fat attenuation with or without calcification with a cystic lesion is diagnostic of a mature teratoma22,63 (Fig. 57-31). On MRI the lipid-laden sebaceous component demonstrates increased signal on T1-weighted images and intermediate  signal on T2-weighted images.63,131 The use of in- and out-of-phase T1-weighted gradient echo images and fat-saturation images are essential for the diagnosis.39 The presence of intravoxel fat and water demonstrates loss of signal on the out-of-phase images, whereas the presence of pure fat demonstrates loss of signal on fat-saturated images, with India ink outline of the fat component on out-of-phase images39 (Fig. 57-32).

CHAPTER 57  Female Pelvis

A

2001

B

C FIG 57-28  Endometrioid carcinoma. T2-weighted images demonstrate a mixed cystic and solid mass in the left ovary (A). T2-hyperintense and T1-hyperintense endometrial thickening (A and B) filling the endometrial canal is consistent with coexisting endometrial carcinoma. Nodular enhancing soft tissue is also noted within the ovarian neoplasm (C). (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

FIG 57-29  Clear cell carcinoma. T2-weighted image demonstrates a unilocular mass with a single small solid papillary projection inferiorly. A serous neoplasm should remain in the differential. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

Malignant transformation of mature teratomas is rare, occurring in 1% to 2%.21,135 Findings associated with malignant transformation include age older than 45 years, tumor diameter greater than 9.9 cm, and elevated squamous carcinoma antigen level.53 The Rokitansky nodule is a common site of malignant transformation, and contrast enhancement suggests but is not conclusive for malignancy.139 Extracapsular growth is also a strong indication of malignant transformation.96 Rupture with leakage of sebaceous material into the peritoneum can occur in up to 4% of mature teratomas.135 Liquefied sebaceous material causes peritoneal irritation, resulting in both acute and chronic inflammation. Acute peritonitis can be caused by sudden rupture and may result in shock or hemorrhage. Sudden rupture is often associated with torsion, trauma, infection, or labor.139 Chronic leakage is more common, causing chronic granulomatous peritonitis or gliomatosis. Gliomatosis can result in extensive abdominal adhesions, which can lead to complications such as bowel obstruction.96 Immature teratoma.  Immature teratomas contain tissue from three germ cell layers; however, in contrary to mature teratomas, they contain immature or embryonic tissue. Immature teratomas are the only other ovarian neoplasm containing fat, but they can usually be differentiated from mature teratomas.139 In contrast to mature teratomas, immature teratomas occur in the first 2 decades of life and are usually very large (15-25 cm) at the time of diagnosis.27,131 They present

2002

PART II  CT and MR Imaging of the Whole Body

A B

C FIG 57-30  Brenner’s tumor in a 66-year-old patient. A, Solid 3-cm right adnexal tumor (arrow) is seen on noncontrast CT. B, Transaxial T2-weighted image demonstrates a well-delineated lesion (arrow) of low signal intensity. C, Contrast-enhanced T1-weighted image with fat suppression shows a slightly inhomogeneous architecture with small areas of low signal intensity surrounded by peripheral areas of intermediate signal intensity (arrow).

with large solid components, cystic areas, and intratumoral fat. The presence of a solid component is helpful for differentiating a mature teratoma from an immature teratoma.139 Dysgerminoma.  Dysgerminomas are the most malignant germ cell tumor but are rare, occurring predominately in children and young women, with 80% younger than age 30.86 Syncytiotrophoblastic giant cells are present in approximately 5% of dysgerminomas and can cause elevated serum hCG levels.93 On CT and MRI, dysgerminomas present as unilateral multilobulated solid masses with prominent fibrovascular septa.17,86 They demonstrate low signal on T1-weighted images and intermediate signal on T2-weighted images, with avidly enhancing fibrovascular septations and central areas of necrosis. CT may demonstrate a speckled pattern of calcification.93 Most patients are diagnosed with early-stage disease; however, in advanced disease nodal metastases are more common than peritoneal metastases.115 Endodermal sinus tumor.  Endodermal sinus tumor is a rare tumor occurring predominately in the second decade of life.17 On imaging, endodermal sinus tumors present as large cystic and solid masses with components extending into the abdominal cavity. Patients

have an elevated AFP level. These tumors grow rapidly and patients have a poor prognosis.200

Sex Cord–Stromal Tumors of the Ovaries.  Sex cord–stromal tumors account for 8% of all ovarian neoplasms and are derived from two cell types: stromal cells and primitive sex cords. Sex cord–stromal tumors account for most of the hormonally active ovarian neoplasms and are usually benign or a low-grade malignancy confined to the ovary at the time of diagnosis.134 Sex cord–stromal tumors are usually treated with surgical resection.87,181

Fibroma, Fibrothecoma, and Thecoma.  Fibroma, fibrothecoma, and thecoma are a spectrum of benign stromal tumors composed  of varying degrees of fibroblasts and spindle cells. They account  for 3% to 4% of all ovarian tumors and occur in pre- and postmenopausal patients.87 Lipid-rich thecomas have few fibroblasts, are hormonally active 60% of the time, and are associated with endometrial carcinoma in 20% of cases.86,133 Fibromas are the most common sex cord–stromal tumor.87 They have no thecal cells or estrogenic activity

CHAPTER 57  Female Pelvis and demonstrate whorled bundles of spindle-shaped fibroblasts and collagen at pathology. In 1% of cases, fibromas are associated with Meigs’ syndrome, which is characterized by an ovarian fibroma, ascites, and pleural effusions.86,134 On imaging, fibromas and fibrothecomas are typically unilateral well-circumscribed solid masses. CT demonstrates a homogeneous solid mass with delayed enhancement and sometimes

*

*

*

2003

dense calcifications.7 MRI demonstrates delayed enhancement with low signal on T1-weighted images and very low signal on T2-weighted images due to fibrous content. This is relatively diagnostic for a fibroma189 (Fig. 57-33). Foci of high intensity on T2-weighted images suggest edema or cystic degeneration.189

Granulosa Cell Tumors.  Granulosa cell tumors are the most common malignant sex cord–stromal tumor and the most common estrogen-producing ovarian tumor.86 Juvenile granulosa cell tumors often present with precocious puberty secondary to estrogenic activity.203 Adult granulosa cell tumors account for 95% of granulosa cell tumors, occur in women older than 30 years, and peak in the perimenopausal period.86 Increased estrogen levels can result in coexistent endometrial hyperplasia, polyps, and carcinoma in 3% to 25% of cases.102,134 The most common presenting symptom is abnormal vaginal bleeding. They have an excellent prognosis (10-year survival, >90%) but have a tendency to recur up to 10 to 20 years after diagnosis.134 CT and MRI findings vary widely, including a completely solid mass, a hemorrhagic mass with fibrotic changes, a multilocular cystic mass, or a completely cystic mass.39 The most common appearance of both juvenile and adult types is a spongelike cystic multilocular mass92 (Fig. 57-34). Size varies, but the average size at diagnosis is 12 cm.203 Granulosa cell tumors differ from epithelial tumors in that they do not have intratumoral papillary projections and are less likely to seed the peritoneum. Granulosa cell tumors may rupture, resulting in acute presentation with hemoperitoneum in as many as 6% of patients.86,203

Sclerosing Stromal Tumor.  Sclerosing stromal tumors are rare

FIG 57-31  Large 17-cm mass in an 18-year-old patient. Coronal CT shows a well-delineated tumor with areas of fluid, mural calcifications, and large focal areas of fat (asterisks). The latter are pathognomonic for benign teratoma. Fat in immature teratomas is usually punctate, minimal, or absent.

A

benign neoplasms, with 80% occurring in the second and third decades of life.86 The most common clinical presentation is menstrual irregularities.87 Imaging findings include a large mass with hyperintense cystic components and a heterogeneous solid component with intermediate to high signal intensity on T2-weighted images. On dynamic contrast-enhanced imaging, sclerosing stromal tumors have a similar enhancement pattern to liver hemangiomas: early peripheral enhancement with centripetal progression. Early arterial enhancement reflects the cellular area of the tumor with a rich peripheral vascular network,

B FIG 57-32  Non–fat-saturated T2 image (A) and fat-saturated T2-weighted image (B) demonstrate loss of signal within the cystic mass on the T2-weighted fat-saturation image, consistent with macroscopic fat and diagnostic of a mature teratoma in this 20-year-old patient. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

2004

PART II  CT and MR Imaging of the Whole Body

B

A

C FIG 57-33  A solid left ovarian mass on US in a 52-year-old postmenopausal patient on hormone replacement therapy. A, Coronal T2-weighted image shows a predominantly hypointense mass (arrow) of the left ovary below the sigmoid. B, Signal intensity of the ovarian mass in the corresponding T1-weighted image is isointense compared to myometrium. C, After contrast injection the upper part of the ovary enhances more than the ovarian mass, which exhibits little homogeneous contrast uptake, typical of ovarian fibroma.

whereas the prolonged central enhancement corresponds to the collagenous hypocellular central component. These findings can assist in differentiating this tumor from a delayed enhancing fibroma.86,186

intratumoral cysts.86 Low signal of the solid component on T2-weighted images can be seen, depending on the degree of fibrous stroma.87

Sertoli–Leydig Cell Tumors.  Sertoli–Leydig cell tumors are the

resent approximately 10% of ovarian neoplasms, with colon, gastric, and breast being the most common primary sites.20 Metastases are bilateral in up to 75% of cases. Krukenberg’s tumors account for 40% to 50% of ovarian metastases and contain mucin-secreting “signetring” cells classically originating from the gastrointestinal system.32,39 Krukenberg’s tumors have characteristic imaging features appearing as bilateral ovarian masses that tend to preserve the normal ovarian contour. Krukenberg’s tumors demonstrate hypointense solid components on T1-weighted images secondary to dense stromal reaction, and internal hyperintensity on T2-weighted images secondary to mucin

most common virilizing ovarian neoplasm but are extremely rare, accounting for less than 0.5% of ovarian neoplasms.87 They typically occur in young women (<30 years), and 40% to 50% present with androgenic activity.134,181 Most of these neoplasms behave in a benign fashion, but they tend to recur early, often within the first year after surgery.87,203 On imaging, Sertoli–Leydig cell tumors are almost always unilateral. They are typically solid enhancing masses, but they can also be mixed solid and cystic or completely cystic. A common imaging presentation is that of a solid enhancing mass with variable-sized

Metastatic Tumors of the Ovaries.  Metastatic ovarian tumors rep-

CHAPTER 57  Female Pelvis

FIG 57-34  Granulosa cell tumor. T2-weighted image demonstrates a mixed cystic and solid ovarian mass without internal papillary projections. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

2005

cystic ovarian masses, paraovarian cysts, and dermoids, especially those greater than 6 cm in size.165 Ovarian torsion is complete or partial twisting of an ovary on its ligamentous support, resulting in compromised blood supply; it is a gynecologic emergency.30 The ovary has a dual blood supply derived primarily from the ovarian artery, with secondary supply from a collateral uterine artery. As such the arterial flow is usually maintained until late; however, venous flow is compromised early, leading to venous congestion and hemorrhagic infarct.149 Ovarian torsion often presents with severe, sharp, localized pelvic pain with peritoneal signs. Nausea and vomiting is a common observation.30 US is the primary imaging modality for ovarian torsion. However, CT and MRI can be useful in equivocal cases. Furthermore, CT may be the initial diagnostic study in patients presenting to the emergency department with acute abdominal pain.30 CT and MRI findings include a midline adnexal mass rotated to the contralateral side of the pelvis, ipsilateral deviation of the uterus, and ascites.148 Multiple peripheral low-density or high T2-weighted signal follicles can be seen surrounding the edematous ovarian stroma.30 Lack of ovarian enhancement is indicative of torsion.71 In the setting of hemorrhagic infarct, the ovary may demonstrate hyperintense signal on T1-weighted images or be hyperattenuating on CT (Fig. 57-36).98 Twisting of the ovarian vascular pedicle is rarely seen.71

Ovarian Vein Thrombosis.  Ovarian vein thrombosis is typically seen in the puerperal state but has also been reported in PID, postoperatively, and in oncology patients.88 In puerperal patients, clinical symptoms typically develop in 1 to 5 days after delivery and are more common after cesarean section.47 Clinical manifestations are variable, ranging from asymptomatic to vague abdominal pain. The thrombus can extend into the inferior vena cava and result in pulmonary embolus and even death.88 Contrast-enhanced studies are the most reliable imaging test. Contrast-enhanced CT demonstrates a low-attenuation filling defect within a dilated ovarian vein (Fig. 57-37).88 On MRI the thrombus may show high signal intensity on T1- and T2-weighted images. Contrast-enhanced T1-weighted images allow for differentiation of thrombus from flow artifact.160

FALLOPIAN TUBES Normal Anatomy FIG 57-35  Metastatic gastric cancer in a 53-year-old patient. T2-weighted image demonstrates a solid ovarian mass with cystic areas of T2 hyperintensity, consistent with a mucin-secreting Krukenberg’s tumor. (©Cox KL, et al: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37:1–5, 2014.)

(Fig. 57-35).65 On CT and MRI they tend to be avidly enhancing. Imaging findings of non-Krukenberg’s ovarian metastases can be similar to primary ovarian malignancy, occasionally making differentiation impossible.20,32

Vascular Conditions

Ovarian Torsion.  Adnexal torsion is an inclusive term encompassing the ovary, fallopian tube, or both. Coexistent torsion of both the ovary and fallopian tube occur in up to two thirds of adnexal torsion cases.1,30 Ovarian torsion is bimodal, occurring in women in their first 3 decades of life and in postmenopausal women. Up to 20% occur during pregnancy, typically during the first and second trimesters.16,103 Ovarian torsion is a well-recognized complication of ovarian lesions such as

The fallopian tubes extend laterally from the uterine fundus to the ipsilateral ovary and lie in the mesosalpinx, a peritoneal fold along the superior margin of the broad ligament. A normal fallopian tube is approximately 9 to 11 cm long and 1 to 4 mm in diameter and is composed of four segments. The interstitial portion lies within the uterine myometrium and is approximately 1 cm long. The isthmus portion is the narrow midsection adjacent to the uterine wall, measuring 2 to 3 cm in length. The adjacent ampulla is slightly larger in diameter and accounts for more than half the length of the fallopian tube. Lastly, the infundibulum opens to the peritoneal cavity and contains approximately 25 fimbriae overhanging the ovary. The tube is composed of a mucosal lining, a muscular layer, and an outer serosa. The mucosal surface contains ciliated cells that assist in propelling the ovum into the uterus.97

Benign Pathologic Conditions

Hydrosalpinx.  Hydrosalpinx is caused by distal occlusion of the fallopian tube, which fills with fluid, pus, or blood, depending on  the cause of obstruction. Hydrosalpinx is most often a result of PID. Additional causes include endometriosis, fallopian tumors, ectopic pregnancy, and adhesions from prior surgery. Symptoms vary; many

2006

PART II  CT and MR Imaging of the Whole Body

A

B FIG 57-36  A 50-year-old patient with severe left pelvic pain, nausea, and vomiting. A, T2-weighted coronal image demonstrates a midline adnexal mass with ipsilateral deviation of the uterus and ascites. B, Lack of ovarian enhancement on T1-weighted fat-saturation images is indicative of ovarian torsion. (Courtesy Deborah Baumgarten, MD, MPH, Atlanta, Ga.)

A

B FIG 57-37  CT scans demonstrate right ovarian vein thrombosis at the level of the kidneys (A) and uterus (B). Central thrombus (arrows) is seen within the dilated ovarian vein, which shows mural enhancement and perivascular edema. An enlarged uterus and small amounts of air within the right abdominal wall are associated findings 2 days after cesarean section.

patients are asymptomatic, whereas others experience recurrent pelvic pain.97 Hydrosalpinx is often diagnosed by US. However, when a hydrosalpinx is large it can be difficult to differentiate from a multiloculated cystic mass, and MRI can be helpful. On imaging, hydrosalpinx appears as a fluid-filled tubular structure extending from the uterine fundus to the ipsilateral ovary, creating an S or C shape as it folds on itself. Incomplete longitudinal folds within the dilated tubular structure are a specific finding.97 The signal intensity of hydrosalpinx on MRI depends on the cause of obstruction. A tube dilated with simple fluid will have low signal on T1-weighted images and high

signal on T2-weighted images, whereas a tube with signal hyperintensity on T1-weighted images is consistent with a hematosalpinx, usually caused by endometriosis or ectopic pregnancy (Fig. 57-38). Enhancement of the wall suggests inflammation and can be seen in the setting of infection or ectopic pregnancy.97

Pelvic Inflammatory Disease.  PID is a spectrum of acute and chronic conditions that affects nearly 1 million women every year and is one of the most common reasons for emergency room visits.41,157 It is most often caused by an ascending sexually transmitted infection of

CHAPTER 57  Female Pelvis

2007

B

* A

B

*

C FIG 57-38  Hydrosalpinx in a perimenopausal woman with history of hysterectomy. Transaxial CT (A) and T2-weighted MRIs (B and C) show an adnexal lesion with thin walls and internal septations adjacent to the right ovary (asterisks). Its tubular character and the incomplete septations (arrow in C) are better appreciated on MRI. B, bladder.

the upper genital tract, including the endometrium, fallopian tubes, and ovaries. Less common causes include inflammation secondary to surgery, abortion-associated infections, and inflammatory conditions of the intestines.57 PID results in pyosalpinx, tubo-ovarian abscess (TOA), tubal damage/scarring, and infertility. Patients have a sixfold increased risk of ectopic pregnancy.157 Imaging findings of early PID are subtle and typically include pelvic edema resulting in thickening of the uterosacral ligaments  and haziness of the pelvic fat (Fig. 57-39). Inflammatory thickening of the fallopian tubes and enlarged, abnormally enhancing ovaries reflect mild oophoritis. Periovarian stranding and enhancement of the adjacent peritoneum are occasionally seen. Endometrial enhancement and simple endometrial fluid reflect endometritis. As PID advances, the fallopian tubes demonstrate more dramatic wall thickening and enhancement and fill with complex fluid, indicating pyosalpinx (Fig. 57-40).157

Formation of a TOA is usually caused by PID but can rarely be secondary to atypical infections such as tuberculosis, actinomycosis, and Pseudomonas.94 TOA is a late finding of PID, demonstrated by a thick-walled multiseptated fluid collection usually measuring between 5 and 10 cm (Fig. 57-41).157 On MRI, fluid within the abscess demonstrates variable signal intensity on T2-weighted images, often with a fluid-debris level.157 Internal gas is a specific but rare finding seen at CT; it is more common in secondary causes of TOA such as diverticulitis.57 Rupture of a TOA can cause life-threatening peritonitis.147 Abdominal and other pelvic organs can also be affected by PID. Dilated pelvic bowel loops can reflect adynamic ileus or mechanical obstruction secondary to adjacent inflammation. Similarly, functional or mechanical obstruction of the ureters may occur.157 Fitz-HughCurtis syndrome is another complication of PID and results from direct extension of bacteria along the right paracolic gutter or through the lymphatic system.157 Fluid and inflammatory stranding can be seen

2008

PART II  CT and MR Imaging of the Whole Body

FIG 57-39  This 23-year-old patient presented with fever and elevated white blood cell count. Early CT findings of mesenteric inflammatory changes and fat stranding are nonspecific. However, cultures were positive for a sexually transmitted disease, and with otherwise normal CT findings, such findings are consistent with early changes of PID. (Courtesy Deborah Baumgarten, MD, MPH, Atlanta, Ga.)

B

A

FIG 57-40  Pyosalpinx. A, MRI shows enhancing inflammatory thickening of the fallopian tube on a coronal T1-weighted fat-saturated contrast-enhanced image. B, In another patient with pyosalpinx there is layering fluid and pus within a dilated right fallopian tube on a T2-weighted image.

*

*

A

B FIG 57-41  Tubo-ovarian abscess in a 35-year-old patient with fever and elevated white blood cell count. A and B, CT scans show a multicystic left adnexal lesion. Within the mass, a tubular structure with homogeneous wall thickening consistent with pyosalpinx (asterisks) can be identified. The inflammatory nature of the lesion is underlined by stranding of the adjacent fat and thickening of the mesorectal fascia (arrow in A).

CHAPTER 57  Female Pelvis extending along the right paracolic gutter. Inflammation of Glisson’s capsule along the anterior liver reflects perihepatitis and occasionally results in adjacent hepatic perfusion variants. Gallbladder wall thickening and pericholecystic fluid may also be seen.147

Malignant Pathologic Conditions

Carcinoma of the Fallopian Tubes.  Primary fallopian tube carcinoma (PFTC) is the rarest of all gynecologic malignancies, accounting for 0.3% to 1.1%.69 Advanced cases of PFTC can be difficult to distinguish from epithelial ovarian carcinoma.69,147 Furthermore, the most common histologic type of PFTC is papillary serous carcinoma, which is identical to serous ovarian adenocarcinoma. A new theory based on histopathology findings in BRCA-positive women who  have undergone prophylactic salpingo-oophorectomy suggests that the fimbrial end of the fallopian tube may be the source of aggressive serous ovarian carcinoma. As a result, the prevalence of PFTC may be underestimated.147 PFTC usually originates in the ampulla and can be nodular, papillary, infiltrative, or masslike. Bilateral involvement is seen in up to 20% of women. The tumor produces copious amounts of fluid, leading to tubal distension. Some 15% of patients present with Latzko’s triad, which consists of profuse serosanguineous vaginal discharge, colicky lower abdominal and pelvic pain relieved by vaginal discharge, and an adnexal mass.147 Pain secondary to tubal distention often results in earlier detection of PFTC and thus a better prognosis than ovarian carcinoma.73,147 Imaging findings of PFTC depend on whether the solid component or the hydrosalpinx is the dominant component. If the hydrosalpinx is the dominant component, PFTC appears as a tubular fluid-filled structure with a solid nodular or papillary component. On MRI the solid component is iso- to hyperintense on T2-weighted images, hypointense on T1-weighted images, and enhances following contrast administration. The hydrosalpinx may appear as simple hyperintense fluid on T2-weighted images or contain blood products resulting in hyperintensity on T1-weighted images. It may also appear as a multilocular cystic mass with a spoke-wheel appearance attributed to folding of the dilated tube upon itself. In this case it may be difficult to distinguish a PFTC from an ovarian neoplasm. Intrauterine fluid collections and peritumoral ascites are common and attributed to decompression of tubal fluid either through the cornual end or fimbrial end, respectively.147 If the solid component dominates, PFTC presents as a sausageshaped enhancing adnexal mass. Peritoneal implants and lymph node metastases can also be seen.90

REFERENCES 1. Albayram F, Hamper UM: Ovarian and adnexal torsion: Spectrum of sonographic findings with pathologic correlation. J Ultrasound Med 20(10):1083–1089, 2001. 2. American Cancer Society: Cancer Facts & Figures 2011, Atlanta, 2011, American Cancer Society. 3. American Fertility Society: The American Fertility Society classification of the adnexal adhesions, tubal occlusion secondary to tubal ligation, tubal pregnancies, müllerian anomalies, and intrauterine adhesions. Fertil Steril 49:944–955, 1988. 4. Amin MB, Young RH: Primary carcinomas of the urethra. Semin Diagn Pathol 14(2):147–160, 1997. 5. Ascher SM, Jha RC, Reinhold C: Benign myometrial conditions: Leiomyomas and adenomyosis. Top Magn Reson Imaging 14(4):281– 304, 2003. 6. Ascher SM, Takahara J, Tha RC: Imaging of gynecologic malignancies. Top Magn Reson Imaging 12:105–129, 2001.

2009

7. Athey PA, Malone RS: Sonography of ovarian fibromas/thecomas.   J Ultrasound Med 6:431–436, 1987. 8. Azziz R, Woods KS, Reyna R, et al: The prevalence and features of the polycystic ovary syndrome in an unselected population. J Clin Endocrinol Metab 89(6):2745–2749, 2004. 9. Balen AH, Laven JS, Tan SL, et al: Ultrasound assessment of the polycystic ovary: International consensus definitions. Hum Reprod Update 9(6):505–514, 2003. 10. Bashist B, Freidman WN, Killackey MA: Surgical transposition of the ovary: Radiological appearance. Radiology 173:857–860, 1989. 11. Bazot M, Darai E, Hourani R, et al: Deep pelvic endometriosis: MR imaging for diagnosis and prediction of extension of disease. Radiology 232:379–389, 2004. 12. Beller U, Sideri M, Maisonneuve P, et al: Carcinoma of the vagina.   J Epidemiol Biostat 6:141–152, 2001. 13. Benedet JL, Bender H, Jones H, et al: FIGO classification and clinical practice guidelines in the management of gynaecologic cancers: FIGO Committee on Gynecologic Oncology. Int J Gynaecol Obstet 70:209– 262, 2000. 14. Berton-Rigaud D, Devouassoux-Shisheboran M, Ledermann JA, et al: Gynecologic Cancer InterGroup (GCIG) consensus review for uterine and ovarian carcinosarcoma. Int J Gynecol Cancer 24(9 Suppl 3):S55–S60, 2014. 15. Bin Park S, Lee JH, Lee YH, et al: Multilocular cystic lesions in the uterine cervix: Broad spectrum of imaging features and pathologic correlation. AJR Am J Roentgenol 195(2):517–523, 2010. 16. Bouguizane S, Bibi H, Farhat Y, et al: Adnexal torsion: A report of 135 cases. J Gynecol Obstet Biol Reprod 32(6):535–540, 2003. 17. Brammer HM, III, Buck JL, Hayes WS, et al: Malignant germ cell tumors of the ovary: Radiologic-pathologic correlation. Radiographics 10:715–724, 1990. 18. Breijer MC, Timmermans A, van Doorn HC, et al: Diagnostic   strategies for postmenopausal bleeding. Obstet Gynecol Int 2010. [Epub 2010]. 19. Brody JM, Koelliker SL, Frishman GN: Unicornuate uterus, imaging appearance, associated anomalies, and clinical applications. AJR Am J Roentgenol 171:1341–1347, 1998. 20. Brown DL, Zou KH, Tempany CM, et al: Primary versus secondary ovarian malignancy: Imaging findings of adnexal masses in the Radiology Diagnostic Oncology Group Study. Radiology 219:213–218, 2001. 21. Buist MR, Golding RP, Burger CW: Comparative evaluation of diagnostic method in ovarian carcinoma with emphasis on CT and MRI. Gynecol Oncol 52:191–198, 1994. 22. Buy JN, Ghossain MA, Moss AA, et al: Cystic teratoma of the ovary: CT detection. Radiology 171:697–701, 1989. 23. Buy JN, Ghossain MA, Sciot C, et al: Epithelial tumors of the ovary: CT findings and correlation with US. Radiology 178:811–818, 1991. 24. Byun JY, Kim SE, Choi BG, et al: Diffuse and focal adenomyosis: MR imaging findings. Radiographics 19:161–170, 1999. 25. Cannistra SA: Cancer of the ovary. N Engl J Med 351:2519–2529, 2004. 26. Carrington BM, Hricak H, Nuruddin RN, et al: Müllerian duct anomalies: MR imaging evaluation. Radiology 176:715–720, 1990. 27. Caruso PA, Marsh MR, Minkowitz S, et al: An intense clinicopathologic study of 305 teratomas of the ovary. Cancer 27:343–348, 1971. 28. Casillas J, Joseph RC, Guerra JJ, Jr: CT appearance of uterine leiomyomas. Radiographics 10(6):999–1007, 1990. 29. Chandrashekhara SH, Thulkar S, Srivastava DN, et al: Pre-operative evaluation of peritoneal deposits using multidetector computed tomography in ovarian cancer. Br J Radiol 84:38–43, 2011. 30. Chang HC, Bhatt S, Dogra VS: Pearls and pitfalls in diagnosis of ovarian torsion. Radiographics 28(5):1355–1368, 2008. 31. Chang YCF, Hricak H, Thurnher S, et al: Vagina: Evaluation with MR imaging. part II: Neoplasms. Radiology 169:175–179, 1998. 32. Chang WC, Meux MD, Yeh BM, et al: CT and MRI of adnexal masses in patients with primary nonovarian malignancy. AJR Am J Roentgenol 186:1039–1045, 2006.

2010

PART II  CT and MR Imaging of the Whole Body

33. Chaudhari VV, Patel MK, Douek M, et al: MR imaging and US of female urethral and periurethral disease. Radiographics 30(7):1857– 1874, 2010. 34. Chen J, Zhang Y, Liang B, et al: The utility of diffusion-weighted MR imaging in cervical cancer. Eur J Radiol 74(3):e101–e106, 2010. 35. Clement PB: Anatomy and histology of the ovary. In Kurman RJ, editor: Blaustein’s Pathology of the Female Genital Tract, New York, 2002, Springer, pp 649–674. 36. Clement PB: Diseases of the peritoneum. In Kurman RJ, editor: Blaustein’s Pathology of the Female Genital Tract, New York, 2002, Springer, pp 448–489. 37. Cohen DT, Oliva E, Hahn PF, et al: Uterine smooth-muscle tumors with unusual growth patterns: Imaging with pathologic correlation. AJR Am J Roentgenol 188(1):246–255, 2007. 38. Comerci JT, Jr, Licciardi F, Bergh PA, et al: Mature cystic teratoma: A clinicopathologic evaluation of 517 cases and review of the literature. Obstet Gynecol 84:22–28, 1994. 39. Cox KL, Baumgarten D, Mittal P: MR findings in cystic ovarian tumors. Contemp Diagn Radiol 37(6):1–6, 2014. 40. Creasman WT: New gynecologic cancer staging. Gynecol Oncol 58(2):157–158, 1995. 41. Curtis KM, Hillis SD, Kieke BA, et al: Visits to emergency departments for gynaecologic disorders in the USA, 1992-1994. Obstet Gynecol 91:1007–1012, 1998. 42. Del Gaizo A, Silva AC, Lam-Himlin DM, et al: Magnetic resonance imaging of solid urethral and peri-urethral lesions. Insights Imaging 4(4):461–469, 2013. 43. Doherty MG, Copeland LJ, Powell MA: Clinical anatomy of the pelvis. In Copeland LJ, Jarrell JF, editors: Textbook of Gynecology, ed 2, Philadelphia, 2000, Saunders, pp 17–57. 44. Doi T, Yamashita Y, Yasunaga T, et al: Adenoma malignum MR imaging and pathologic study. Radiology 204:39–42, 1997. 45. Dueholm M, Lundorf E: Transvaginal ultrasound or MRI for diagnosis of adenomyosis. Curr Opin Obstet Gynecol 19(6):505–512, 2007. 46. Dunaif A: Insulin resistance and the polycystic ovary syndrome: Mechanism and implications for pathogenesis. Endocr Rev 18(6):774– 800, 1997. 47. Dunnihoo DR, Gallaspy JW, Wise RB, et al: Post-partum ovarian vein thrombophlebitis: A review. Obstet Gynecol Surv 46:415–427, 1991. 48. Eilber KS, Raz S: Benign cystic lesions of the vagina: A literature review. J Urol 170:717–722, 2003. 49. Exacoustos C, Manganaro L, Zupi E: Imaging for the evaluation of endometriosis and adenomyosis. Best Pract Res Clin Obstet Gynaecol 28(5):655–681, 2014. 50. Fennessy FM: MRI of Benign Female Pelvis. Body MRI: 2013 ARRS Categorial Course. 223-230. . 51. Forstner R: Radiological staging of ovarian cancer: Imaging findings and contribution of CT and MRI. Eur Radiol 17:3223–3235, 2007. 52. Forstner R, Hricak H, Occhipinti KA, et al: Ovarian cancer: Staging with CT and MR imaging. Radiology 197:619–626, 1995. 53. Forstner R, Hricak H, Powell CB, et al: Ovarian cancer recurrence: value of MR imaging. Radiology 196:715–720, 1995. 54. Freeman SJ, Aly AM, Kataoka MY, et al: The revised FIGO staging system for uterine malignancies: implications for MR imaging. Radiographics 32(6):1805–1827, 2012. 55. Funt AS, Hricak H: Ovarian malignancies. Top Magn Reson Imaging 14:329–338, 2003. 56. Gardner E, Gray DJ, O’Rahilly R: Female genital organs. In Anatomy: A Regional Study of Human Structure, ed 3, Philadelphia, 1969, Saunders, pp 490–502. 57. Ghiatas AA: The spectrum of pelvic inflammatory disease. Eur Radiol 14:E184–E192, 2004. 58. Ghossain MA, Buy JN, Ruiz A, et al: Hyperreactio luteinalis in abnormal pregnancy: Sonographic and MRI findings. J Magn Reson Imaging 8:1203–1206, 1998. 59. Giudice LC: Clinical practice: Endometriosis. N Engl J Med 362(25):2389–2398, 2010.

60. Glastonbury CM: The shading sign. Radiology 224(1):199–201, 2002. 61. Goldstein SR, Nachtigall M, Snyder JR, et al: Endometrial assessment by vaginal ultrasonography before endometrial sampling in patients with postmenopausal bleeding. Am J Obstet Gynecol 163:119–123, 1990. 62. Grant LA, Sala E, Griffin N: Congenital and acquired conditions of the vulva and vagina on magnetic resonance imaging: A pictorial review. Semin Ultrasound CT MR 31(5):347–362, 2010. 63. Guinet C, Ghossain MA, Buy JN, et al: Mature cystic teratomas of the ovary: CT and MR findings. Eur J Radiol 20:137–143, 1995. 64. Gull B, Karlsson B, Milsom I, et al: Transvaginal sonography of the endometrium in a representative sample of postmenopausal women. Ultrasound Obstet Gynecol 7:322–327, 1996. 65. Ha HK, Baek SY, Kim SH, et al: Krukenberg’s tumor of the ovary: MR imaging features. AJR Am J Roentgenol 164:1435–1439, 1995. 66. Hall DA: Sonographic appearance of the normal ovary, of polycystic ovary disease, and of functional ovarian cysts. Semin Ultrasound CT MR 4:149–165, 1983. 67. Hase S, Mitsumori A, Inai R, et al: Endometrial polyps: MR imaging features. Acta Med Okayama 66(6):475–485, 2012. 68. Hauth EA, Jaeger HJ, Libera H, et al: MR imaging of the uterus and cervix in healthy women: Determination of normal values. Eur Radiol 17:734–742, 2007. 69. Hensley ML, Alektiar KM, Chi DS: Ovarian and fallopian-tube cancer. In Barakat RR, Bevers MW, Gershenson DM, et al, editors: Handbook of Gynecologic Oncology, London, 2001, Martin Dunitz, pp 243–263. 70. Hertzberg BS, Kliewer MA, Paulson EK: Ovarian cyst rupture causing hemoperitoneum: Imaging features and the potential for misdiagnosis. Abdom Imaging 24(3):304–308, 1999. 71. Hiller N, Appelbaum L, Simanovsky N, et al: CT features of adnexal torsion. AJR Am J Roentgenol 189:124–129, 2007. 72. Holschneider CH: Epidemiology, clinical features, and diagnosis of invasive cervical cancer. , 2007. 73. Hosokawa C, Tsubakimoto M, Inoue Y, et al: Bilateral primary fallopian tube carcinoma: Findings on sequential MRI. AJR Am J Roentgenol 186:1046–1050, 2006. 74. Hosseinzadeh K, Heller MT, Houshmand G: Imaging of the female perineum in adults. Radiographics 32(4):129–168, 2012. 75. Hou JY, Kelly MG, Yu H, et al: Neoadjuvant chemotherapy lessens surgical morbidity in advanced ovarian cancer and leads to improved survival in stage IV disease. Gynecol Oncol 105:211–217, 2007. 76. Hricak H, Chang Y, Thurnher S: Vagina: Evaluation with MR imaging: Part I. Normal anatomy and congenital anomalies. Radiology 169:169– 174, 1988. 77. Huang MW, Muradali D, Thurston WA, et al: Uterine arteriovenous malformations: Gray-scale and Doppler US features with MR imaging correlation. Radiology 206(1):115–123, 1998. 78. Husband JE, Padhani AR: Royal College of Radiologists: Recommendations for cross-sectional imaging in cancer management, 2006, Royal College of Radiologists. 79. Isoda H, Kurokawa H, Kuroda M, et al: Fibroma of the vulva. Comp Med Imaging Graphics 26:139–142, 2002. 80. Iwamoto H, Suzuki M, Watanabe N, et al: Case study of a pregnant woman with decidualized ovarian endometriosis whose preoperative findings suggested malignant transformation. Eur J Gynaecol Oncol 27(3):301–303, 2006. 81. Iyer RB, Jhingran A, Sawaf H, et al: Imaging findings after radiotherapy to the pelvis. AJR Am J Roentgenol 177:1083–1089, 2001. 82. Jain KA: Imaging of peritoneal inclusion cysts. AJR Am J Roentgenol 174:1559–1563, 2000. 83. Jemal A, Siegel R, Ward E, et al: Cancer statistics, 2009. CA Cancer J Clin 59(4):225–249, 2009. 84. Jeong YY, Outwater EK, Kang HK: Imaging evaluation of ovarian masses. Radiographics 20:1445–1470, 2000. 85. Johari B, Koshy M, Sidek S, et al: Lipoleiomyoma: A rare benign tumour of the uterus. BMJ Case Rep 2014. 86. Jung SE, Lee JM, Rha SE, et al: CT and MR imaging of ovarian tumors with emphasis on differential diagnosis. Radiographics 22:1305–1325, 2002.

CHAPTER 57  Female Pelvis 87. Jung SE, Rha SE, Lee JM, et al: CT and MRI findings of sex cordstromal tumor of the ovary. AJR Am J Roentgenol 185(1):207–215, 2005. 88. Karaosmanoglu D, Karcaaltincaba M, Karcaaltincaba D, et al: MDCT of the ovarian vein: Normal anatomy and pathology. AJR Am J Roentgenol 192(1):295–299, 2009. 89. Kaur H, Silverman PM, Iyer RB, et al: Diagnosis, staging, and surveillance of cervical carcinoma. AJR Am J Roentgenol 180(6):1621– 1631, 2003. 90. Kawakami S, Togashi K, Kimura I, et al: Primary malignant tumor of the fallopian tube: Appearance at CT and MR imaging. Radiology 186:503–508, 1993. 91. Kawamoto S, Urban BA, Fishman EK: CT of epithelial ovarian tumors. Radiographics 19:S85–S102, 1999. 92. Kim SH: Granulosa cell tumor of the ovary: Common findings and unusual appearances on CT and MR. J Comput Assist Tomogr 26:756–761, 2002. 93. Kim SH, Kang SB: Ovarian dysgerminoma: color Doppler ultrasonographic findings and comparison with CT and MR imaging findings. J Ultrasound Med 14:843–848, 1995. 94. Kim SH, Kim SH, Yang DM, et al: Unusual causes of tubo-ovarian abscess: CT and MR imaging findings. Radiographics 24:1575–1589, 2004. 95. Kim JS, Lee HJ, Woo SK, et al: Peritoneal inclusion cysts and their relationship to the ovaries: Evaluation with sonography. Radiology 204:481–484, 1997. 96. Kim KA, Park CM, Lee JH, et al: Benign ovarian tumors with solid and cystic components that mimic malignancy. AJR Am J Roentgenol 182(5):1259–1265, 2004. 97. Kim MY, Rha SE, Oh SN, et al: MR Imaging findings of hydrosalpinx: A comprehensive review. Radiographics 29(2):495–507, 2009. 98. Kimura I, Togashi K, Kawakami S, et al: Ovarian torsion: CT and MRI appearances. Radiology 190:337–341, 1994. 99. Kinkel K: Pitfalls in staging uterine neoplasm with imaging: A review. Abdom Imaging 31:164–173, 2006. 100. Kinkel K, Forstner R, Danza FM, et al: Staging of endometrial cancer with MRI: Guidelines of the European Society of Urogenital Imaging. Eur Radiol 19(7):1565–1574, 2009. 101. Kinkel K, Frei KA, Balleyguier C, et al: Diagnosis of endometriosis with imaging: A review. Eur Radiol 16:285–298, 2006. 102. Ko SF, Wan YL, Ng SH, et al: Adult ovarian granulosa cell tumors: Spectrum of sonographic and CT findings with pathologic correlation. AJR Am J Roentgenol 172:1227–1233, 1999. 103. Koonings PP, Grimes DA: Adnexal torsion in postmenopausal women. Obstet Gynecol 73:11–12, 1989. 104. Larson DM, Connor GP, Broste SK, et al: Prognostic significance of gross myometrial invasion with endometrial cancer. Obstet Gynecol 88(3):394–398, 1996. 105. Lee TT, Rausch ME: Polycystic ovarian syndrome: Role of imaging in diagnosis. Radiographics 32(6):1643–1657, 2012. 106. Levine D, Gosink BB, Johnson LA: Change in endometrial thickness in postmenopausal women undergoing hormone replacement therapy. Radiology 197:603–608, 1995. 107. Levy AD, Arnaiz J, Shaw JC, et al: Primary peritoneal tumors: Imaging features with pathologic correlation. Radiographics 28(2):583–607, 2008. 108. Lewin SN: Revised FIGO staging system for endometrial cancer. Clin Obstet Gynecol 54(2):215–218, 2011. 109. Lopez C, Balogun M, Ganesan R, et al: MRI of vaginal conditions. Clin Radiol 60:648–662, 2005. 110. Luciano DE, Luciano AA: Management of endometriosis-related pain: An update. Womens Health 7(5):585–590, 2007. 111. Lutz AM, Willmann JK, Drescher CW, et al: Early diagnosis of ovarian carcinoma: Is a solution in sight? Radiology 259:329–345, 2001. 112. Manfredi R, Gui B, Maresca G, et al: Endometrial cancer: Magnetic resonance imaging. Abdom Imaging 30(5):626–636, 2005. 113. Matsuoka Y, Ohtomo K, Araki T, et al: MR imaging of clear cell carcinoma of the ovary. Eur Radiol 11:946–951, 2001.

2011

114. McLean JM: Embryology and anatomy of the ovaries. In Fox H, Wells M, editors: Haines and Taylor Obstetrical and Gynaecological Pathology, ed 4, New York, 1995, Churchill Livingstone, pp 673–698. 115. Mironov S, Akin O, Pandit-Taskar N, et al: Ovarian cancer. Radiol Clin North Am 45:149–166, 2007. 116. Mitchell DG, Hill MC, Hill S, et al: Serous carcinoma of the ovary: CT identification of metastatic calcified implants. Radiology 158:649–652, 1986. 117. Moon WJ, Koh BH, Kim SK, et al: Brenner tumor of the ovary: CT and MR findings. J Comput Assist Tomogr 24:72–76, 2000. 118. Morcel K, Camborieux L, Guerrier D: Mayer-Rokitansky-Kuster-Hauser (MRKH) syndrome. Orphanet J Rare Dis 2:13, 2007. 119. Moyle PL, Kataoka MY, Nakai A, et al: Nonovarian cystic lesions of the pelvis. Radiographics 30(4):921–938, 2010. 120. Mucitelli DR, Charles EZ, Kraus FT: Vulvovaginal polyps: Histologic appearance, ultrastructure, immunocytochemical characteristics and clinicopathologic correlations. Int J Gynecol Pathol 9:20, 1990. 121. Murakami M, Miyamoto T, Iida T, et al: Whole-body positron emission tomography and tumor marker CA125 for detection of recurrence in epithelial ovarian cancer. Int J Gynecol Cancer 16(Suppl 1):99–107, 2006. 122. Murase E, Siegelman ES, Outwater EK, et al: Uterine leiomyomas: Histopathologic features, MR imaging findings, differential diagnosis, and treatment. Radiographics 19(5):1179–1197, 1999. 123. Nalaboff KM, Pellerito JS, Ben-Levi E: Imaging the endometrium: Disease and normal variants. Radiographics 21(6):1409–1424, 2001. 124. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Uterine neoplasms. Version 3. National Comprehensive Cancer Network. . Published March 20, 2012. Accessed Nov 17, 2014. 125. Nicolet V, Carignan L, Bourdon F, et al: MR imaging of cervical carcinoma: A practical staging approach. Radiographics 20(6):1539– 1549, 2000. 126. Occhipinti KA: Computed tomography and magnetic resonance imaging of the ovary. In Anderson JC, editor: Gynecologic Imaging, London, 1999, Churchill Livingstone, pp 345–359. 127. Okamoto Y, Tanaka YO, Nishida M, et al: MR imaging of the uterine cervix: Imaging-pathologic correlation. Radiographics 23(2):425–445, 2003. 128. Outwater EK, Dunton CJ: Imaging of the ovary and adnexa: Clinical issues and applications of MR imaging. Radiology 194:1–18, 1995. 129. Outwater EK, Mitchell DG: Normal ovaries and functional cysts: MR appearance. Radiology 198(2):397–402, 1996. 130. Outwater EK, Siegelman ES, Chiowanich P, et al: Dilated fallopian tubes: MR imaging characteristics. Radiology 208(2):463–469, 1998. 131. Outwater EK, Siegelman ES, Hunt JL: Ovarian teratomas: Tumor types and imaging characteristics. Radiographics 21:475–490, 2001. 132. Outwater EK, Siegelman ES, Kim B, et al: Ovarian Brenner tumors: MR imaging characteristics. Magn Reson Imaging 16:1147–1153, 1998. 133. Outwater EK, Siegelman ES, Talerman A, et al: Ovarian fibromas and cystadenofibromas: MRI features of the fibrous component. J Magn Reson Imaging 7:465–471, 1997. 134. Outwater EK, Wagner BJ, Mannion C, et al: Sex cord–stromal and steroid cell tumors of the ovary. Radiographics 18:1523–1546, 1998. 135. Ozols RF, Rubin SC, Thomas G, et al: Epithelial ovarian cancer. In Hoskins WJ, Perez CA, Young RC, editors: Principles and Practice of Gynecologic Oncology, ed 2, Philadelphia, 1997, Lippincott-Raven, pp 919–986. 136. Ozols RF, Schwartz PE, Eifel PJ: Ovarian cancer, fallopian tube carcinoma, and peritoneal carcinoma. In De Vita VT, Jr, Hellman S, Rosenberg SA, editors: Cancer: Principles and Practice of Oncology, ed 6, Philadelphia, 2001, Lippincott Williams & Wilkins, pp 1597–1632. 137. Özsarlak Ö, Tjalma W, Schepens E, et al: The correlation of preoperative CT, MR imaging, and clinical staging (FIGO) with histopathologic findings in primary cervical carcinoma. Eur Ra diol 13:2338–2345, 2003. 138. Pannu HK, Corl FM, Fishman EK: CT evaluation of cervical cancer: Spectrum of disease. Radiographics 21:1155–1168, 2001.

2012

PART II  CT and MR Imaging of the Whole Body

139. Park SB, Kim JK, Kim KR, et al: Imaging findings of complications and unusual manifestations of ovarian teratomas. Radiographics 28(4):969– 983, 2008. 140. Phillips D, Deipolyi AR, Hesketh RL, et al: Pelvic congestion syndrome: etiology of pain, diagnosis, and clinical management. J Vasc Interv Radiol 25(5):725–733, 2014. 141. Prasad SR, Menias CO, Narra VR, et al: Cross-sectional imaging of the female urethra: Technique and results. Radiographics 25(3):749–761, 2005. 142. Pretorius ES, Outwater EK, Hunt JL, et al: Magnetic resonance imaging of the ovary. Top Magn Reson Imaging 12:131–146, 2001. 143. Querleu D, Planchamp F, Narducci F, et al: Clinical practice guidelines for the management of patients with endometrial cancer in France: Recommendations of the Institut National du Cancer and the Société Française d’Oncologie Gynécologique. Int J Gynecol Cancer 21(5):945– 950, 2011. 144. Rauch GM, Kaur H, Choi H, et al: Optimization of MR imaging for pretreatment evaluation of patients with endometrial and cervical cancer. Radiographics 34(4):1082–1098, 2014. 145. Reinhold C, McCarthy S, Bret PM, et al: Diffuse adenomyosis: Comparison of endovaginal US and MR imaging with histopathologic correlation. Radiology 199:151–158, 1996. 146. Reproductive system. In Bannister LH, Dyson M, editors: Gray’s Anatomy: The Anatomical Basis of Medicine and Surgery, ed 38, Edinburgh, Scotland, 1995, Churchill Livingstone, pp 1861–1880. 147. Rezvani M, Shaaban AM: Fallopian tube disease in the nonpregnant patient. Radiographics 31(2):527–548, 2001. 148. Rha SE, Byun JY, Jung SE, et al: CT and MR imaging features of adnexal torsion. Radiographics 22(2):283–294, 2002. 149. Rha SE, Byun JY, Jung SE, et al: Atypical CT and MRI manifestations of mature ovarian cystic teratomas. AJR Am J Roentgenol 183:743–750, 2004. 150. Rockall AG, Ghosh S, Alexander-Sefre F, et al: Can MRI rule out bladder and rectal invasion in cervical cancer to help select patients for limited EUA? Gynecol Oncol 101(2):244–249, 2006. 151. Rockall AG, Meroni R, Sohaib SA, et al: Evaluation of endometrial carcinoma on magnetic resonance imaging. Int J Gynecol Cancer 17(1):188–196, 2007. 152. Rutledge F: Cancer of the vagina. Am J Obstet Gynecol 97:635–655, 1967. 153. Sahdev A, Sohaib SA, Wenaden AE, et al: The performance of magnetic resonance imaging in early cervical carcinoma: A long-term experience. Int J Gynecol Cancer 17(3):629–636, 2007. 154. Saksouk FA, Johnson SC: Recognition of the ovaries and ovarian   origin of pelvic masses with CT. Radiographics 24(1):S133–S146, 2004. 155. Sala E, Rockall AG, Freeman SJ, et al: The added role of MR imaging in treatment stratification of patients with gynecologic malignancies: What the radiologist needs to know. Radiology 266:717–740, 2013. 156. Sala E, Wakely S, Senior E, et al: MRI of malignant neoplasms of the uterine corpus and cervix. AJR Am J Roentgenol 188(6):1577–1587, 2007. 157. Sam JW, Jacobs JE, Birnbaum BA: Spectrum of CT findings in acute pyogenic pelvic inflammatory disease. Radiographics 22(6):1327–1334, 2002. 158. Sammour RN, Leibovitz Z, Shapiro I, et al: Decidualization of ovarian endometriosis during pregnancy mimicking malignancy. J Ultrasound Med 24(9):1289–1294, 2005. 159. Santos P, Cunha TM: Uterine sarcomas: Clinical presentation and MRI features. Diagn Interv Radiol 2014. [Epub ahead of print]. 160. Savader S, Otero RR, Savader BL: Puerperal ovarian vein thrombosis: Evaluation with CT, US, and MR imaging. Radiology 167:637–639, 1998. 161. Scheidler J, Heuck AF: Imaging of cancer of the cervix. Radiol Clin North Am 40(3):577–590, 2002. 162. Scoutt LM, Flynn SD, Luthringer DJ, et al: Junctional zone of the uterus: correlation of MR imaging and histologic examination of hysterectomy specimens. Radiology 179(2):403–407, 1991.

163. Seidman JD, Russell P, Kurman RJ: Surface epithelial tumors of the ovary. In Kurman RJ, editor: Blaustein’s Pathology of the Female Genital Tract, New York, 2002, Springer Verlag, pp 791–904. 164. Semelka RC, Lawrence PH, Shoenut JP, et al: Primary ovarian cancer: Prospective comparison of contrast-enhanced CT and pre- and postcontrast, fat-suppressed MR imaging, with histologic correlation.   J Magn Reson Imaging 3:99–106, 1993. 165. Sherard GB, Hodson CA, Williams HJ, et al: Adnexal masses and pregnancy: A 12-year experience. Am J Obstet Gynecol 189:358–362, 2003. 166. Siddiqui N, Nikolaidis P, Hammond N, et al: Uterine artery embolization: Pre- and post-procedural evaluation using magnetic resonance imaging. Abdom Imaging 38(5):1161–1177, 2013. 167. Siegelman ES, Oliver ER: MR imaging of endometriosis: Ten imaging pearls. Radiographics 32(6):1675–1691, 2012. 168. Siegelman ES, Outwater EK: Tissue characterization in the female pelvis by means of MR imaging. Radiology 212:5–18, 1999. 169. Siegelman ES, Outwater EK, Manner MP, et al: High resolution MR imaging of the vagina. Radiographics 17:1183–1203, 1997. 170. Sironi S, Colombo E, Villa G, et al: Myometrial invasion by endometrial carcinoma: Assessment with plain and gadolinium-enhanced MR imaging. Radiology 185(1):207–212, 1992. 171. Sohaib SA, Richards PS, Ind T, et al: MR imaging of carcinoma of the vulva. AJR Am J Roentgenol 178:373–377, 2002. 172. Subak LL, Hricak H, Powell CB, et al: Cervical carcinoma: Computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol 86(1):43–50, 1995. 173. Sudderuddin S, Helbren E, Telesca M, et al: MRI appearances of benign uterine disease. Clin Radiol 69(11):1095–1104, 2014. 174. Sugimura K, Carrington BM, Quivey JM, et al: Postirradiation changes in the pelvis: Assessment with MR imaging. Radiology 175:805–813, 1990. 175. Suh DD, Yang CC, Cao Y, et al: Magnetic resonance imaging anatomy of the female genitalia in premenopausal and postmenopausal women. J Urol 170:138–144, 2003. 176. Surabhi VR, Menias CO, George V, et al: Magnetic resonance imaging of female urethral and periurethral disorders. Radiol Clin North Am 51(6):941–953, 2013. 177. Takagi H, Matsunami K, Noda K, et al: Magnetic resonance   imaging in the evaluation of double uterus and associated urinary   tract anomalies: A report of five cases. J Obstet Gynaecol 23:525–527, 2003. 178. Takeuchi M, Matsuzali K, Uehara H, et al: Pathologies of the   uterine endometrial cavity: Usual and unusual manifestations and pitfalls on magnetic resonance imaging. Eur Radiol 15:2244–2255, 2005. 179. Tamai K, Koyama T, Saga T, et al: MR features of physiologic and benign conditions of the ovary. Eur Radiol 16:2700–2711, 2006. 180. Tamai K, Togashi K, Ito T: MR Imaging findings of adenomyosis: Correlation with histopathologic features and diagnostic pitfalls. Radiographics 25:21–40, 2005. 181. Tanaka YO, Tsunoda H, Kitagawa Y, et al: Functioning ovarian tumors: Direct and indirect findings at MR imaging. Radiographics 24(Suppl 1): S147–S166, 2004. 182. Tanaka YO, Yoshizako T, Nishida M, et al: Ovarian carcinoma in patients with endometriosis: MR imaging findings. AJR Am J Roentgenol 175:1423–1430, 2000. 183. Tangjitgamol S, Manusirivithaya S, Laopaiboon M, et al: Interval debulking surgery for advanced epithelial ovarian cancer: A Cochrane systematic review. Gynecol Oncol 112:257–264, 2009. 184. Togashi K, Nakai A, Sugimura K: Anatomy and physiology of the female pelvis: MR imaging revisited. J Magn Reson Imaging 13:842–849, 2001. 185. Togashi K, Nishimura K, Kimura I, et al: Endometrial cysts: Diagnosis with MR imaging. Radiology 180(1):73–78, 1991. 186. Torricelli P, Caruso Lombardi A, Boselli F, et al: Sclerosing stromal tumor of the ovary: US, CT and MR imaging findings. Abdom Imaging 27:588–591, 2002.

CHAPTER 57  Female Pelvis 187. Trattner M, Graf AH, Lax S, et al: Prognostic factors in surgically treated stage Ib-IIb cervical carcinomas with the special emphasis on the importance of volume. Gynecol Oncol 82:11–16, 2001. 188. Troiano RN: Magnetic resonance imaging of mullerian duct abnormalities of the uterus. Top Magn Reson Imaging 14:269–279, 2003. 189. Troiano RN, Lazzarini KM, Scoutt LM, et al: Fibroma and fibrothecoma of the ovary: MR imaging findings. Radiology 204:795–798, 1997. 190. Troiano RN, McCarthy SM: Müllerian duct anomalies: Imaging and clinical issues. Radiology 233:19–34, 2004. 191. Van Hoe L, Vanbeckevoort D, Oyen R, et al: Cervical carcinoma: Optimized local staging with intravaginal contrast-enhanced MR imaging preliminary results. Radiology 213:608–611, 1999. 192. Van Vierzen PB, Massuger LF, Ruys SH, et al: Fast dynamic contrast enhanced MR imaging of cervical carcinoma. Clin Radiol 53(3):183– 192, 1998. 193. Vijayakumar A, Srinivas A, Chandrashekar BM, et al: Uterine vascular lesions. Rev Obstet Gynecol 6(2):69–79, 2013. 194. Wagner BJ, Buck JL, Seidman JD, et al: Ovarian epithelial neoplasms: Radiologic-pathologic correlation. Radiographics 14:1351–1374, 1994. 195. Waldroup L, Liu JB: Sonographic anatomy of the female pelvis. In Berman MC, editor: Diagnostic Medical Sonography: Obstetrics and Gynecology, Philadelphia, 1991, Lippincott, pp 27–69. 196. Walensky RP, Venbrux AC, Presott CA, et al: Pseudomyxoma peritonei. AJR Am J Roentgenol 167:471–474, 1996.

2013

197. Woodward PJ, Sohaey R, Mezzetti TP: Endometriosis: Radiologic pathologic correlation. Radiographics 21:193–216, 2001. 198. Wu TT, Coakley FV, Quayyum A, et al: Magnetic resonance imaging of ovarian cancer arising in endometriomas. J Comput Assist Tomogr 28:836–838, 2004. 199. Wu TI, Yen TC, Lai CH: Clinical presentation and diagnosis of uterine sarcoma, including imaging. Best Pract Res Clin Obstet Gynaecol 25:681–689, 2011. 200. Yamaoka T, Togashi K, Koyama T, et al: Yolk sac tumor of the ovary: Radiologic-pathologic correlation in four cases. J Comput Assist Tomogr 24:605–609, 2000. 201. Yamashita Y, Torashima M, Takahashi M, et al: Hyperintense uterine leiomyoma at T2-weighted MR imaging: Differentiation with dynamic enhanced MR imaging and clinical implications. Radiology 189(3):721– 725, 1993. 202. Yoo RE, Cho JY, Kim SY, et al: A systematic approach to the magnetic resonance imaging-based differential diagnosis of congenital müllerian duct anomalies and their mimics. Abdom Imaging 2014 Jul 29. [Epub ahead of print]. 203. Young RH, Scully RE: Sex-cord-stromal, steroid cell, and other ovarian tumors. In Kurman RJ, editor: Blaustein’s Pathology of the Female Genital Tract, New York, 2002, Springer Verlag, pp 905–965. 204. Zand KR, Reinhold C, Abe H, et al: Magnetic resonance imaging of the cervix. Cancer Imaging 7:69–76, 2007.

59  Advanced Cardiovascular CT Imaging Matthias Renker, Carlo N. De Cecco, Julian Lukas Wichmann, Stephen R. Fuller, and U. Joseph Schoepf

CT MYOCARDIAL PERFUSION IMAGING According to the cascade model by Nesto et al.,24 myocardial ischemia is preceded by alterations in blood flow. Therefore stress myocardial perfusion imaging holds the potential to detect coronary artery disease at an early stage before adverse clinical events occur. Computed tomography myocardial perfusion imaging (CTMPI) provides functional information in addition to anatomic cardiovascular assessment via coronary CT angiography (cCTA) and may enhance diagnostic performance of cardiac CT for detection of hemodynamically significant coronary artery disease. CTMPI can be traced back to the 1980s.31 Until recently, however, major technical challenges precluded routine clinical use of this  technique. With emerging developments in CT technology, namely increased spatial and temporal resolution, the interest in CTMPI has been rekindled. To date, multiple studies have demonstrated the feasibility of CTMPI.1,15,19,37,38 For instance, the multicenter CORE320 study has recently provided evidence for an improvement of diagnostic accuracy for detection of hemodynamically relevant lesions by combining angiography and perfusion in a single CT protocol.29 For this purpose a combined protocol comprising both approaches has been shown to be feasible at a patient radiation dose level similar to exposure during nuclear perfusion imaging.29

Technical Prerequisites A priori, a high temporal resolution of the CT system is necessary to reduce motion artifacts during the cardiac cycle and ensure adequate partial volume correction. With the advent of 64-slice CT systems, temporal resolution (165 milliseconds) improved enough to allow for CTMPI in patients with low heart rates.30 Only with further technical advances, such as the introduction of dual-source CT (temporal resolution: first generation, 83 ms; second generation, 75 ms; third generation, 66 ms), has the scanning of individuals with higher heart rates become feasible.5 Development of 320-slice CT systems also allowed for cardiac image acquisition within a single heartbeat, but the moderate temporal resolution of 175 milliseconds impeded scanning at higher heart rates.29,39 Furthermore, high spatial resolution is essential to accurately assess and differentiate myocardial perfusion of the subepicardial and subendocardial layers.30 In addition, a motion correction algorithm is necessary to compensate for artifacts due to respiratory movements, since the required breath-hold periods are generally not realizable in clinical routine.

Measurement Principles Testing for myocardial ischemia by means of CTMPI benefits from the principle that contrast material distribution serves as an indicator of

myocardial blood flow. As opposed to normal myocardium, areas  of ischemia show delayed distribution and decreased peak contrast material enhancement, while the washout is normal. In the presence of infarcted myocardium, delayed contrast material distribution, decreased peak attenuation during first pass, and slow washout can be observed.30 Areas of myocardium with reduced contrast material attenuation represent a perfusion defect. Similar to cardiac stress testing in nuclear medicine, CTMPI protocols typically include acquisitions at rest and under stress conditions pharmacologically induced by vasodilative substances such as adenosine, regadenoson, dobutamine, dipyridamole, and adenosine triphosphate. The underlying reason is that reversible myocardial perfusion defects occur prior to fixed ones, and the addition of stress testing is expected to increase the sensitivity of CTMPI. Healthy coronary arteries respond to pharmacologic stress with increased perfusion and induction of hyperemic myocardium. Coronary arteries with obstructive disease have limited perfusion capacity because they are already operating within dilatory reserve to compensate for reduced perfusion. The chronological order of the rest and stress acquisitions is variable.30 Residual contrast material from the rest acquisition may limit the diagnostic conclusiveness of subsequent stress imaging for detection of ischemic myocardium. In theory, when the stress acquisition is performed first, the myocardium is not contaminated with previously administered contrast material. However, continuous elevation of the heart rate may impair the quality of a subsequent rest acquisition. Furthermore, if the rest acquisition is conducted first, a cCTA study with high negative predictive value (NPV) is effectively supplied. We advocate the strategy of an initial rest acquisition because the rest acquisition allows ruling out obstructive coronary artery disease. Therefore exposure of patients without obstructive coronary artery disease to a second dose of radiation can be avoided because these do not require an additional stress test. Meinel et al.19 recently investigated the contributions of rest, stress, and delayed enhancement at CTMPI and concluded that rest-stress acquisition should be the protocol of choice for an assessment of myocardial blood supply.

Scanning Technique Two techniques can be differentiated: static and dynamic CTMPI. A static image of myocardial attenuation at the time of arterial first-pass perfusion represents a snapshot of the distribution of iodinated contrast material.30 Dynamic myocardial perfusion imaging is obtained when myocardial attenuation is recorded at several consecutive time points. Scan parameters of single-energy static CTMPI do not vary from those for cCTA. The data sample displays the entire heart and can

2063

2064

PART II  CT and MR Imaging of the Whole Body

either be acquired over multiple heartbeats in a certain cardiac phase or within one or two heartbeats using wide-detector CT technology. Volumes of contrast medium generally range between 60 and 70 mL. A drawback of static CTMPI is the dependence of image acquisition on the timing of the contrast material bolus. The single data sample should be gathered during peak myocardial contrast material enhancement and therefore requires bolus tracking to ensure a proper CTMPI acquisition during the contrast enhancement peak. Furthermore, if the data are acquired during multiple heartbeats, myocardial contrast enhancement may vary along the myocardial wall.6 Static CTMPI can also be performed using a dual-energy acquisition technique.7 This technique takes advantage of the improved iodine discrimination under radiation from different tube voltages and makes quantification of the myocardial blood pool from iodine concentration possible.17 For this purpose the two tubes of a dual-source CT system may be operated simultaneously with different tube voltage settings (100 kV and 140 kV), resulting in a low- and high-photon energy level dataset that can be used to create an iodine map of the myocardium. A different approach to dual-energy static CTMPI is called kV switching and uses a dual-energy–capable single-source scanner that rapidly switches tube potential between a low and high tube voltage setting. Switching may be conducted within or between gantry rotations in a vendor-specific manner.12,14,33,34 Additionally, a “sandwich detector” can be used for dual-energy acquisition, with superimposition of two detectors in the same gantry, absorbing at different photon energy levels.4 Dynamic CTMPI allows for assessment of myocardial perfusion using sequential imaging for the evaluation of changes in contrast attenuation over time. Dynamic CTMPI can be performed either using a stationary table position or in shuttle mode. With the traditional stationary imaging technique, myocardial perfusion is recorded at a certain level in cine mode.11 By use of the shuttle technique, coverage of the whole heart during a certain cardiac phase is enabled through a backward-forward alternating position of the table.20 Favorably, dynamic CTMPI allows for absolute quantification of myocardial blood volume and perfusion, which provides information on myocardial blood volume and mean transit time for each vascular territory independently. However, these benefits come at the price of an increased applied radiation dose.

Patient Preparation and Image Acquisition Prior to CTMPI, patients should be instructed to avoid intake of caffeine and products containing theophylline, methylxanthine, and dipyridamole because the vasodilator capacity of pharmacologic stress agents may consequently be limited during stress imaging. Furthermore, the use of β-blockers before undergoing the stress acquisition is considered contraindicated by some because of the theoretical concern that myocardial ischemia may be masked. Nitrates are not recommended before CTMPI, as they may reduce the diagnostic capabilities for detecting myocardial perfusion defects.40 A second intravenous access is necessary because the infusion of contrast material and the pharmacologic stressor are simultaneously performed. In addition to 12-lead electrocardiographic monitoring, continuous blood pressure surveillance and supervision by a physician certified in Advanced Cardiovascular Life Support (ACLS) are indicated. Prior to image acquisition, emphasis should be placed on the tube voltage setting; it not only influences patient radiation dose exposure but also determines attenuation of contrast material. Because the atomic number of iodine is high, contrast material attenuation at lower tube voltages is more distinct.23 However, per societal guidelines it is strongly recommended to adjust the kilovoltage of CT scans to the patient’s body mass index, because image noise may increase with lowtube-voltage acquisition.10

Analysis of CTMPI Evaluation of myocardial perfusion is performed qualitatively by comparing the attenuation of an ischemic area to that of normal myocardium or by normalizing it with the blood pool of the left ventricular cavity. Using postprocessing software, an iodine distribution map can be displayed and normalized by placing a region of interest on an area of myocardium with high iodine content.32 If CTMPI reveals decreased myocardial attenuation during the stress phase only, or if the hypoattenuated area increases significantly, reversible ischemia can be suspected. Myocardial areas of hypoattenuation during both rest and stress acquisition correspond to irreversible ischemia. Myocardial ischemia could theoretically be missed in the presence of three-vessel coronary artery disease with balanced ischemia, wherein relative perfusion defects are not revealed owing to global ischemia—a limitation found similarly in single-photon emission computed tomography (SPECT). However, in static CTMPI the coronary arteries are also visualized anatomically, allowing for detection of diffuse disease. Compared with static CTMPI, the absolute quantification of blood volume via dynamic CTMPI allows an accurate assessment of myocardial perfusion without the need for visual comparison of myocardial segments. Furthermore, semiautomatic and fully automatic quantitative  analysis is available for dynamic CTMPI, using information about distribution and washout of contrast material over time (Figs. 59-1   and 59-2).

Future Implications Although initial results suggest a promising role for CTMPI, it is a technique that is still in its infancy. Refinement of protocols is necessary regarding contrast material injection, dataset acquisition, image reconstruction, and radiation exposure. For this purpose, implementation of iterative image reconstruction techniques may be beneficial.28 Further research is required to define, validate, and establish this novel technique in direct comparison with alternative hybrid imaging modalities such as SPECT/CT and positron emission tomography (PET)/CT. However, given its advantages of short examination time with comprehensive data acquisition, CTMPI may be a favorable option, particularly for evaluation of patients who present to the emergency department with acute chest pain and a low cardiovascular risk profile. Moreover, stent and bypass graft patency assessment may fall within the scope of this modality in the future.

CT-BASED FRACTIONAL FLOW RESERVE Because the dilative capacity of the coronary arteries is limited, the presence of severe luminal obstruction (>85%) leads to its exhaustion at rest conditions.25 When hyperemia is present—either induced by exercise or pharmacologic stress—an impairment of blood flow is observed for coronary artery stenosis grades of 50% and above.9 However, an erratic functional relevance has been demonstrated for intermediate grade lesions (50%-75% luminal stenosis): half are ischemia-causing and the remaining half are not.3 Evaluation of a lesion’s hemodynamic significance is important because it demonstrably provides guidance for patient care.27,36 A promising physiologic approach is assessment of coronary flow reserve, defined as an integrated measure of maximum flow through the major epicardial vessels as well as through the microvasculature; however, this does not indicate which specific component of the coronary arteries is affected.8 In parallel to this development, the concept of fractional flow reserve (FFR) was pursued to evaluate lesion-specific hemodynamic relevance. FFR represents the ratio between the maximum coronary

2065

CHAPTER 59  Advanced Cardiovascular CT Imaging

150

100

A 50

0 MBF (mL/100mL/min)

B

C FIG 59-1  Coronary CTA of a 67-year-old man with severe two-vessel disease involving the left anterior descending (A) and circumflex coronary artery (B). C, Dynamic CTMPI shows the presence of a multiterritory severe reduction of the myocardial blood flow. Red indicates peak values for myocardial blood flow, purple indicates lowest values.

A

B FIG 59-2  Virtual noncontrast color-coded reconstruction of images obtained during angiographic phase of a dual-energy triple-rule-out CT study in an 80-year-old man with acute myocardial infarction. A perfusion deficit in the anteroseptal segments on short-axis view (A) and horizontal long-axis view (B) can be demarcated, caused by an acute occlusion of the left anterior descending coronary artery in a patient with known multivessel disease.

2066

PART II  CT and MR Imaging of the Whole Body

flow in the presence of a stenosis and the hypothetical maximum coronary flow in its absence. In routine clinical practice, hyperemia is induced by pharmacologic stress agents (e.g., adenosine), and FFR is measured during invasive coronary angiography as the coronary pressure distal to the stenosis of interest divided by the proximal pressure within the ascending aorta. FFR values less than 0.80 (or <0.75 in another study) have been shown to relate to ischemia and indicate the necessity for revascularization.27,35 FFR was thus established as the reference standard for assessing the functional impact of a stenotic lesion on myocardial perfusion. Recent technical developments allow noninvasive CT-based FFR to enhance the diagnostic accuracy of cardiac CT for detecting hemodynamically relevant coronary artery stenoses. As a large body of evidence demonstrated, cCTA features a high NPV, allowing for a reliable exclusion of coronary artery disease.18,21 However, its moderate positive predictive value (PPV) may limit its reliability and accuracy in certain cases as a noninvasive diagnostic tool. A major limitation is the uncertainty regarding the functional relevance of anatomically identifiable stenotic lesions. In addition to myocardial perfusion imaging, CT-based FFR has shown high potential to mitigate this drawback.

Technical Aspects Unlike CTMPI, CT-based FFR has the potential to allow discrimination between flow-obstructive and nonobstructive lesions based on routine cCTA datasets, without the need for further image acquisitions entailing additional radiation dose exposure and contrast material usage. Additionally, the administration of pharmacologic stressors is not necessary for obtaining noninvasive FFR from cCTA. CT-based FFR is based on the principles of computational fluid dynamics, a branch of fluid mechanics that uses algorithms and numerical methods to model fluid flows; nonmedical applications of computational fluid dynamics include airplane design and a variety of onshore and offshore petroleum industry applications. To predict lesion-specific ischemia from coronary artery stenosis, CT-based FFR has been introduced as an offsite service that uses parallel supercomputers at core laboratories to solve complex Reynolds-averaged NavierStokes equations—a demonstrably accurate, elaborated method for hemodynamic assessment of indeterminate lesions of the major coronary vessels. Newly developed algorithms enable time-efficient in-hospital evaluation of FFR from cCTA data2 (Fig. 59-3). Operating on a regular postprocessing workstation, these algorithms integrate reduced- and full-order models for rapid computation of hemodynamic coronary artery tree models. This noninvasive detection of stenosis-specific ischemia at the point of care may improve patient management and guide rational therapeutic decision making. However, further studies with substantial cohort sizes to compare the diagnostic performance of this new CT-based FFR approach with anatomic interpretation of cCTA and coronary catheter angiography correlated with the invasive reference standard FFR are needed.

Diagnostic Performance Clinical trials have demonstrated that the CT-based FFR approach compares favorably to the reference standard of FFR obtained during invasive coronary angiography.16,22,26 Within the single-center

DISCOVER-FLOW (Diagnosis of Ischemia-Causing Stenoses Obtained Via Noninvasive Fractional Flow Reserve) trial,16 CT-based FFR in comparison with invasive FFR showed 81% accuracy, 93% sensitivity, 82% specificity, a PPV of 85%, and an NPV of 91% on a per-patient basis. Moreover, CT-based FFR resulted in higher diagnostic accuracy than anatomic cCTA for identification of ischemia-causing coronary artery lesions, because the area under the receiver-operator characteristics curve (AUC) was 0.90 for CT-based FFR and 0.75 for anatomic cCTA (P = 0.001); a 95% confidence interval (CI) was not reported.16 The international multicenter DeFACTO study (Determination of Fractional Flow Reserve by Anatomic Computed Tomographic Angiography)22 evaluated the diagnostic value of CT-based FFR and included a total of 252 patients. Based on a per-patient analysis, CT-based FFR yielded a higher diagnostic performance than anatomic cCTA for identification of functionally relevant stenoses (accuracy, 73% vs. 64%; sensitivity, 90% vs. 84%; specificity, 54% vs. 42%; PPV, 67% vs. 61%; NPV, 84% vs. 72%).22 Compared with obstructive coronary artery disease diagnosed by visual assessment of cCTA (AUC, 0.68; 95% CI, 0.62-0.74), CT-based FFR was shown to improve discriminatory power for the detection of hemodynamically significant lesions (AUC, 0.81; 95% CI, 0.75-0.86; P < 0.001).22 The sensitivity in patients with intermediate-grade stenosis (defined as 30%-70% luminal degree of obstruction) notably showed a more than twofold increase when comparing CT-based FFR to cCTA alone (82% vs. 37%), without any decrease in specificity (66% vs. 66%).22 More recently, Nørgaard et al.26 investigated the diagnostic performance of noninvasive FFR from cCTA data in patients with suspected coronary atherosclerotic disease for the identification of obstructive disease within the context of the NXT (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps) trial. This study was conducted on 484 vessels of 251 patients and demonstrated a superior diagnostic accuracy for CT-based FFR over mere anatomic cCTA on both a per-patient (AUC, 0.90 vs. 0.81) and a per-vessel basis (AUC, 0.93 vs. 0.79).26 Compared with cCTA alone, CT-based FFR correctly reclassified 68% of false-positive patient results and 67% of falsepositive vessel results as confirmed by invasive FFR.26 In addition to defining the hemodynamic relevance of lesions with greater than 50% stenosis on anatomic cCTA, the NXT trial also analyzed the performance of CT-based FFR against a cCTA threshold for ischemic lesions of greater than 70% stenosis, yielding a higher specificity for cCTA compared to CT-based FFR (84% vs. 79%) but at the cost of a substantially lower sensitivity (70% vs. 86%).26

Future Implications CT-based FFR is a promising new tool to assess stenosis-specific ischemia based on acquired standard cCTA datasets and could improve the accuracy of cardiac CT. However, further research is necessary to ascertain the value of FFR-CT in comparison to other cCTA-derived techniques for functional assessment of coronary artery stenoses and could be assessed in a routine clinical setting with the recently developed approach for onsite computation in real time. A novel application of this technique is virtual stent placement, allowing assessment and comparison of different stenting strategies and prediction of outcomes based on changes in the functional assessment of CT-based FFR.13

CHAPTER 59  Advanced Cardiovascular CT Imaging

2067

A

1

0.8

0.6

CT-based FFR = 0.94

Invasive FFR =0.93

B

C FIG 59-3  Coronary CTA MPR showing an intermediate-degree luminal stenosis of the middle portion of the right coronary artery (arrow in A). Coronary catheter angiography confirmed the lesion (arrow in B). Invasive fractional flow reserve was measured distal to the lesions as 0.93, indicating the absence of lesion-specific ischemia (arrowhead in B). CT-based FFR resulted in a value of 0.94, in good correlation with the invasive measurement (arrowhead in C).

REFERENCES 1. Bastarrika G, Ramos-Duran L, Rosenblum MA, et al: Adenosine-stress dynamic myocardial CT perfusion imaging: Initial clinical experience. Invest Radiol 45:306–313, 2010. 2. Baumann S, Wang R, Schoepf UJ, et al: Coronary CT angiographyderived fractional flow reserve correlated with invasive fractional flow reserve measurements – Initial experience with a novel physician-driven algorithm. Eur Radiol 25:1201–1207, 2015. 3. Bech GJ, De Bruyne B, Pijls NH, et al: Fractional flow reserve to determine the appropriateness of angioplasty in moderate coronary stenosis: A randomized trial. Circulation 103:2928–2934, 2001. 4. Boll DT, Merkle EM, Paulson EK, et al: Coronary stent patency: Dual-energy multidetector CT assessment in a pilot study with anthropomorphic phantom. Radiology 247:687–695, 2008. 5. De Cecco CN, Meinel FG, Chiaramida SA, et al: Coronary artery computed tomography scanning. Circulation 129:1341–1345, 2014.

6. de Feyter PJ: CT functional imaging using intracoronary gradient analysis: An indispensable boost for CT coronary angiography. Eur Heart J Cardiovasc Imaging 13:971–972, 2012. 7. Fornaro J, Leschka S, Hibbeln D, et al: Dual- and multi-energy CT: Approach to functional imaging. Insights Imaging 2:149–159, 2011. 8. Gould KL, Kirkeeide RL, Buchi M: Coronary flow reserve as a physiologic measure of stenosis severity. J Am Coll Cardiol 15:459–474, 1990. 9. Gould KL, Lipscomb K, Hamilton GW: Physiologic basis for assessing critical coronary stenosis. Instantaneous flow response and regional distribution during coronary hyperemia as measures of coronary flow reserve. Am J Cardiol 33:87–94, 1974. 10. Halliburton SS, Abbara S, Chen MY, et al: SCCT guidelines on radiation dose and dose-optimization strategies in cardiovascular CT. J Cardiovasc Comput Tomogr 5:198–224, 2011. 11. Huber AM, Leber V, Gramer BM, et al: Myocardium: Dynamic versus single-shot CT perfusion imaging. Radiology 269:378–386, 2013.

2068

PART II  CT and MR Imaging of the Whole Body

12. Karcaaltincaba M, Aktas A: Dual-energy CT revisited with multidetector CT: Review of principles and clinical applications. Diagn Interv Radiol 17:181–194, 2011. 13. Kim KH, Doh JH, Koo BK, et al: A novel noninvasive technology for treatment planning using virtual coronary stenting and computed tomography-derived computed fractional flow reserve. JACC Cardiovasc Interv 7:72–78, 2014. 14. Ko JP, Brandman S, Stember J, et al: Dual-energy computed tomography: Concepts, performance, and thoracic applications. J Thorac Imaging 27:7–22, 2012. 15. Ko BS, Cameron JD, Meredith IT, et al: Computed tomography stress myocardial perfusion imaging in patients considered for revascularization: A comparison with fractional flow reserve. Eur Heart J 33:67–77, 2012. 16. Koo BK, Erglis A, Doh JH, et al: Diagnosis of ischemia-causing coronary stenoses by noninvasive fractional flow reserve computed from coronary computed tomographic angiograms. Results from the prospective multicenter DISCOVER-FLOW (Diagnosis of Ischemia-Causing Stenoses Obtained Via Noninvasive Fractional Flow Reserve) study. J Am Coll Cardiol 58:1989–1997, 2011. 17. Koonce JD, Vliegenthart R, Schoepf UJ, et al: Accuracy of dual-energy computed tomography for the measurement of iodine concentration using cardiac CT protocols: Validation in a phantom model. Eur Radiol 24:512–518, 2014. 18. Meijboom WB, Meijs MF, Schuijf JD, et al: Diagnostic accuracy of 64-slice computed tomography coronary angiography: A prospective, multicenter, multivendor study. J Am Coll Cardiol 52:2135–2144, 2008. 19. Meinel FG, De Cecco CN, Schoepf UJ, et al: First-arterial-pass dual-energy CT for assessment of myocardial blood supply: Do we need rest, stress, and delayed acquisition? Comparison with SPECT. Radiology 270:708–716, 2014. 20. Meinel FG, Ebersberger U, Schoepf UJ, et al: Global quantification of left ventricular myocardial perfusion at dynamic CT: Feasibility in a multicenter patient population. AJR Am J Roentgenol 203:W174–W180, 2014. 21. Miller JM, Rochitte CE, Dewey M, et al: Diagnostic performance of coronary angiography by 64-row CT. N Engl J Med 359:2324–2336, 2008. 22. Min JK, Leipsic J, Pencina MJ, et al: Diagnostic accuracy of fractional flow reserve from anatomic CT angiography. JAMA 308:1237–1245, 2012. 23. Mongan J, Rathnayake S, Fu Y, et al: In vivo differentiation of complementary contrast media at dual-energy CT. Radiology 265: 267–272, 2012. 24. Nesto RW, Kowalchuk GJ: The ischemic cascade: Temporal sequence of hemodynamic, electrocardiographic and symptomatic expressions of ischemia. Am J Cardiol 59:23C–30C, 1987. 25. Nichols AB, Brown C, Han J, et al: Effect of coronary stenotic lesions on regional myocardial blood flow at rest. Circulation 74:746–757, 1986. 26. Nørgaard BL, Leipsic J, Gaur S, et al: Diagnostic performance of noninvasive fractional flow reserve derived from coronary computed

27.

28.

29.

30. 31.

32. 33.

34.

35.

36.

37.

38.

39.

40.

tomography angiography in suspected coronary artery disease: The NXT trial (Analysis of Coronary Blood Flow Using CT Angiography: Next Steps). J Am Coll Cardiol 63:1145–1155, 2014. Pijls NH, Fearon WF, Tonino PA, et al: Fractional flow reserve versus angiography for guiding percutaneous coronary intervention in patients with multivessel coronary artery disease: 2-year follow-up of the FAME (Fractional Flow Reserve Versus Angiography for Multivessel Evaluation) study. J Am Coll Cardiol 56:177–184, 2010. Renker M, Ramachandra A, Schoepf UJ, et al: Iterative image reconstruction techniques: Applications for cardiac CT. J Cardiovasc Comput Tomogr 5:225–230, 2011. Rochitte CE, George RT, Chen MY, et al: Computed tomography angiography and perfusion to assess coronary artery stenosis causing perfusion defects by single photon emission computed tomography:   The CORE320 study. Eur Heart J 35:1120–1130, 2014. Rossi A, Merkus D, Klotz E, et al: Stress myocardial perfusion: Imaging with multidetector CT. Radiology 270:25–46, 2014. Rumberger JA, Feiring AJ, Lipton MJ, et al: Use of ultrafast computed tomography to quantitate regional myocardial perfusion: A preliminary report. J Am Coll Cardiol 9:59–69, 1987. Schwarz F, Ruzsics B, Schoepf UJ, et al: Dual-energy CT of the heart–Principles and protocols. Eur J Radiol 68:423–433, 2008. Silva AC, Morse BG, Hara AK, et al: Dual-energy (spectral) CT: Applications in abdominal imaging. Radiographics 31:1031–1046, discussion 1047–1050, 2011. So A, Hsieh J, Imai Y, et al: Prospectively ECG-triggered rapid kV-switching dual-energy CT for quantitative imaging of myocardial perfusion. JACC Cardiovasc Imaging 5:829–836, 2012. Tonino PA, De Bruyne B, Pijls NH, et al: Fractional flow reserve versus angiography for guiding percutaneous coronary intervention. N Engl J Med 360:213–224, 2009. Toth G, De Bruyne B, Casselman F, et al: Fractional flow reserve-guided versus angiography-guided coronary artery bypass graft surgery. Circulation 128:1405–1411, 2013. Wang Y, Qin L, Shi X, et al: Adenosine-stress dynamic myocardial perfusion imaging with second-generation dual-source CT: Comparison with conventional catheter coronary angiography and SPECT nuclear myocardial perfusion imaging. AJR Am J Roentgenol 198:521–529, 2012. Weininger M, Schoepf UJ, Ramachandra A, et al: Adenosine-stress dynamic real-time myocardial perfusion CT and adenosine-stress first-pass dual-energy myocardial perfusion CT for the assessment of acute chest pain: Initial results. EurJ Radiol 81:3703–3710, 2012. Wong DT, Ko BS, Cameron JD, et al: Comparison of diagnostic   accuracy of combined assessment using adenosine stress computed tomography perfusion + computed tomography angiography with transluminal attenuation gradient + computed tomography angiography against invasive fractional flow reserve. J Am Coll Cardiol 63:1904–1912, 2014. Zoghbi GJ, Dorfman TA, Iskandrian AE: The effects of medications on myocardial perfusion. J Am Coll Cardiol 52:401–416, 2008.

58  Coronary Arteries, Heart, and Pericardium Charles S. White, Seth Kligerman, Gerdien Kramer, Minh Lu, and Christopher Brown

The coronary arteries are vessels that course through the epicardial fat to supply the myocardium with oxygenated blood. There is a great deal of variability in the anatomy of the coronary arteries from person to person, and the size and branching pattern of each coronary artery can vary significantly. Therefore it is important to remember that anatomic variants are quite common, and it is imperative to distinguish benign variants from congenital abnormalities that can lead to a compromise of blood flow with subsequent myocardial ischemia, infarction, or sudden cardiac death. In the first part of this chapter, basic anatomy and normal variants will be discussed.

CORONARY ARTERY ANATOMY The coronary arteries arise from the sinuses of Valsalva, which are three anatomic outpouchings in the ascending aorta. Although no coronary artery arises from the noncoronary sinus, the left main coronary artery and right coronary artery arise from the left and right sinuses, respectively (Fig. 58-1).

Left Main Coronary Artery The left main coronary artery arises from the left sinus of Valsalva and courses laterally before dividing. Although it is normally the vessel with the largest diameter compared to the other coronary arteries, it is of variable length.43 The average length of the left main coronary artery is nearly 1 cm, but this can vary significantly, with lengths ranging from a few millimeters to a few centimeters.3 In most patients, the left main coronary artery then bifurcates into two vessels, the left anterior descending coronary artery (LAD) and left circumflex coronary artery (LCx) (Fig. 58-2). However, in approximately 20% to 30% of the patients the left main coronary artery will trifurcate, with a ramus intermedius branch arising between the LAD and LCx14,27,105 (Fig. 58-3).

Left Anterior Descending Coronary Artery The LAD courses over the anterior surface of the left ventricle (LV) in the anterior interventricular sulcus or groove between the left and right ventricles (Fig. 58-4). Although its diameter can vary between patients, in most the LAD is a large vessel that courses along the anterior surface of the LV. Similar to other coronary arteries, the length of the LAD can vary significantly and can terminate prior to reaching the LV apex, terminate at the LV apex, or wrap around the LV apex to supply the inferior wall at the apical level.88 The LAD is divided into three portions (Fig. 58-5). The proximal LAD is defined as the portion of the LAD from its origin to the ostium of the first large septal branch or diagonal branch, whichever arises first. The definition of the mid-LAD is defined as the end of the proximal LAD to half the distance to the LV apex. The distal LAD is defined as the end of the mid-LAD to the termination of the LAD.158

The LAD gives rise to both septal branches and diagonal branches that are important in supplying oxygenated blood to the anteroseptal and anterolateral portions of the LV myocardium, respectively (Fig. 58-6). The septal branches arise from the inferomedial aspect of the LAD and dive into the anterior portion of the interventricular septum. They are usually smaller in size but more numerous than diagonal branches. The diagonal branches arise from the lateral aspect of the LAD and course over the anterolateral aspect of the LV. The number of diagonal branches varies, but in most patients there are between two and four diagonals.

Left Circumflex Coronary Artery With the bifurcation of the left main, the LAD courses anterior in  the anterior interventricular groove while the LCx courses posterolaterally between the LV and left atrium (see Fig. 58-2). The LCx gives rise to obtuse marginal (OM) vessels that supply oxygenated blood to the inferolateral aspect of the LV (Fig. 58-7). Similar to diagonal branches, the size and number of OM branches varies between patients. In most patients there are at least two visible OM vessels. Once the LCx reaches the inferolateral aspect of the left atrioventricular (AV) groove and begins to wrap around the inferior aspect of the LV, it is usually a diminutive vessel, since most people are right dominant. In patients who are left or codominant, the distal LCx will be larger in size (Fig. 58-8).

Ramus Intermedius In approximately 20% to 30% of patients the left main coronary artery will trifurcate into the LAD, LCx, and a middle branch called the ramus intermedius or intermediate branch (see Fig. 58-3).14,27,105 This branch can vary in size and distribution; in some patients it will course anterolaterally to supply the anterolateral portions of the LV (similar to diagonal branches), whereas in other cases it will course inferolaterally to supply the inferolateral portion of the LV (similar to OM branches).

Right Coronary Artery The right coronary artery (RCA) arises from the anterior-facing right sinus of Valsalva (Fig. 58-9). Since most patients are right dominant the RCA is a large vessel that courses anteriorly in the right AV groove. Similar to the LAD the RCA is divided into three territories (Fig. 58-10).158 The proximal RCA is defined as the ostium of the RCA to half the distance to the acute margin of the heart. The mid-RCA is defined as the end of the proximal RCA to the acute margin of the heart, and the distal RCA is defined as the end of the mid-RCA to the origin of the posterior descending artery. As the RCA courses in the right AV groove, branches arise from the RCA and course anteriorly and extend to the anterior surface of the right ventricle (RV) (see Fig. 58-9). Similar to branches of the LAD and LCX, these (acute) marginal arteries vary in size and number. In

2015

2016

PART II  CT and MR Imaging of the Whole Body

R L N

L

FIG 58-1  Normal coronary ostial anatomy. Transverse image through the aortic sinus from a coronary CTA shows the left (L), right (R), and noncoronary (N) sinuses. The left main coronary artery (white arrow) and right coronary artery (black arrow) arise normally from the left and right coronary sinuses, respectively.

FIG 58-2  Normal left main anatomy. Double oblique image from a coronary CTA shows the left main coronary artery (white arrowhead) arises from the left sinus of Valsalva (L). The left main bifurcates into the LAD (white arrow) and LCx (black arrow). The sinoatrial nodal branch arises from the LCx in this patient (black arrowhead).

* RV

LV

10.00mm

FIG 58-3  Ramus intermedius. The left main (black arrowhead) trifurcates into the LAD (white arrow), LCx (black arrow), and ramus intermedius branches (white star). A ramus intermedius is present in 20% to 30% of the population.

FIG 58-4  LAD anatomy. Volume-rendered image of the anterior heart from a coronary CTA shows the LAD (black arrows) coursing in the anterior interventricular sinus or groove between the right ventricle (RV) and left ventricle (LV). Diagonal branches (white arrows) arise from the lateral aspect of the LAD to supply the anterior lateral aspect of the LV.

2017

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

P

M LA L LV

D 10.00mm

FIG 58-5  LAD segmentation. Two-chamber MIP view of the LAD from a coronary CTA shows the vessel coursing over the anterior walls of the left ventricle (LV). The LAD is divided into three segments. The proximal LAD (P) is defined as the portion of the LAD from its origin to the ostium of the first large septal branch (white arrow) or diagonal branch, whichever arises first. The definition of the mid-LAD (M) is defined as the end of the proximal LAD to half the distance to the LV apex (white line). The distal LAD (D) is defined as the end of the mid-LAD to the termination of the LAD. LA, left atrium.

FIG 58-7  LCx anatomy. Volume-rendered image from a coronary CTA looking at the lateral aspect of the LV shows the left main coronary artery trifurcating into the LAD (white arrow), ramus intermedius (white arrowheads), and LCx (black arrows). The LCx courses in the groove between the left atrium and ventricle and gives rise to obtuse marginal braches (black arrowheads) that supply the inferolateral aspect of the LV. In right-dominant patients, the LCx becomes small in size near the inferolateral aspect of the LV distal to the ostia of the last obtuse marginal branch.

10.00mm/div

10.00mm/div

FIG 58-6  Septal and diagonal branches of the LAD. Double oblique axial MIP image from a coronary CTA shows the left main coronary artery (black arrowhead) arising from the left sinus of Valsalva (L). The left main bifurcates into the LAD (white arrowhead) and LCx (not shown in plane). The lateral branches of the LAD are diagonal vessels (white arrow) that supply the anterolateral aspect of the LV. The smaller septal arteries (black arrows) supply the anteroseptal portion of the LV.

FIG 58-8  Coronary artery codominance. Coronal oblique MIP image through the AV grooves from a coronary CTA shows the right coronary artery giving rise to the posterior descending coronary artery. The distal LCx (black arrow) is a large vessel and gives rise to the posterior ventricular branch (black arrowhead), consistent with a codominant system. In most patients who are right dominant, the distal LCx will be diminutive in size toward the inferior aspect of the LV.

2018

PART II  CT and MR Imaging of the Whole Body

RA RV

RV

R

LA

RA N

L

LV

FIG 58-11  Distal right coronary artery (RCA) anatomy. Volume-rendered

FIG 58-9  Right coronary artery anatomy. Axial oblique view transverse to the aortic sinuses from a coronary CTA shows the right coronary artery (RCA, black arrows) arising from the right sinus of Valsalva (R) and the left main coronary artery (white arrowhead) arising from the left sinus (L). No coronary artery arises from the noncoronary sinus (N). The RCA courses in the right AV groove between the right atrium (RA) and right ventricle (RV). Acute marginal branches (white arrow) of the right ventricle arise anteriorly and supply to the anterior free wall of the RV.

image of the inferior surface of the heart from a coronary CTA shows the distal RCA (white arrowheads) coursing in the inferior groove between the right atrium (RA) and right ventricle (RV). The RCA bifurcates into the posterior descending artery (PDA, white arrow), which courses in the inferior interventricular sulcus or groove, and the posterior ventricular branch (black arrow), which supplies the inferior aspect of the left ventricular base. A large acute marginal branch (black arrowhead) from the LCx supplies the inferolateral aspect of the left ventricle (LV), which is anterior to the left atrium (LA).

right dominant patients, the distal RCA will divide into two branches along the undersurface of the heart (Fig. 58-11; see Fig. 58-10). These branches are the posterior descending artery (PDA) and posterior lateral branch (PLB).

Posterior Descending Artery In approximately 75% to 88% of patients, the PDA arises from the RCA, and thus they are considered right dominant (see Figs. 58-10 and 58-11).27,102,105,197 In approximately 9% of patients, the PDA arises from the LCx and the patient is considered left dominant. In a smaller percentage of patients the PDA will arise from the RCA, and a large PLB branch will arise from the LCx; these patients are often considered balanced or codominant (see Fig. 58-8). Whichever vessel gives rise to the PDA, in some ways the PDA is  a mirror image of the LAD. While the LAD courses in the anterior interventricular sulcus, the PDA courses in the posterior interventricular sulcus along the inferior aspect of the LV. Similar to the LAD the PDA gives rise to septal branches that supply the inferolateral aspect of the LV septum. The PDA is of variable size and length. In patients with a smaller LAD the PDA tends to be a larger artery and vice versa  (Fig. 58-12).88

P

M

D

FIG 58-10.  Right coronary artery (RCA) anatomy. C-view of RCA from a coronary CTA shows its division into three territories. The proximal RCA (P) is defined as the ostium of the RCA to half the distance to the acute margin of the heart (white arrowhead). The mid-RCA (M) is defined as the end of the proximal RCA to the acute margin of the heart, and the distal RCA (D) is defined as the end of the mid-RCA to the origin of the posterior descending artery (PDA, white arrow). The posterior ventricular branch (black arrow) courses over the PDA in this patient with right dominance.

Posterior Lateral Branch While the PDA courses anteriorly in the posterior interventricular sulcus, the PLB (or posterior lateral ventricular branch [PLVB]) courses laterally and extends along the posterior AV groove between the inferior aspect of the left atrium and LV (see Figs. 58-10 and 58-11). It is of variable size and length but usually provides oxygenated blood to the inferolateral base of the LV.

Conus Branch The conus branch is usually the first branch of the RCA (Fig. 58-13), although it can have a separate origin from the right coronary sinus in

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2019

R

FIG 58-12  Variable size of the posterior descending coronary artery (PDA). Two-chamber view from a coronary CTA shows a large PDA (black arrows) extending to the left ventricular apex. The LAD (white arrows) is a smaller vessel that ends proximal to the apex.

FIG 58-14  Multiple ostia. C-view through the right coronary artery (white arrowhead) from a coronary CTA shows that the conus branch (black arrow) and sinoatrial nodal branch (white arrow) both have separate ostia from the right sinus of Valsalva (R).

RVOT SVC

FIG 58-13  Anatomy of the conus artery. Axial oblique image from a coronary CTA shows the conus branch (white arrows), which is usually the first vessel to arise from the right coronary artery unless it arises directly from the right aortic sinus. The vessel courses superiorly and anteriorly to supply the right ventricular outflow tract (RVOT).

17% to 50% of patients (Fig. 58-14).47 The conus branch extends anteriorly to supply blood to the right ventricular outflow tract (RVOT) or conus. In some instances the conus acts as a collateral pathway for blood flow to the LAD, and this circuit is often referred to as the arterial circle of Vieussens.37

Sinoatrial Nodal Branch The SA nodal branch is a small vessel that most often originates from the RCA (Fig. 58-15) but arises from the LCx in about one third of patients (see Fig. 58-2). Less commonly the SA branch can arise from the left main coronary artery, both the RCA and LCx, or directly from the aorta153 (see Fig. 58-14). Depending on its origin the SA nodal

LA

FIG 58-15  SA nodal branch anatomy. Axial oblique MIP image from a coronary CTA shows the SA branch (arrows) arising from the proximal right coronary artery (RCA) and coursing posteriorly to end between left atrium (LA) superior vena cava (SVC) just near its point of draining into the right atrium. The SA branch can arise from the LCx artery (see Fig. 58-2), both the RCA and LCx, or directly from the aorta.

branch courses posteriorly (RCA origin) or medially (LCx origin) and terminates in the region of the SA node, which is located along the posterior aspect of where the superior vena cava enters the right atrium.

Atrioventricular Nodal Branch In most patients the AV nodal branch arises from the very distal U-shaped aspect of the distal RCA as is courses superior to the PDA.153

2020

PART II  CT and MR Imaging of the Whole Body

L

FIG 58-16  AV nodal branch anatomy. Coronal oblique MIP image from a coronary CTA shows the AV nodal branch (white arrowhead) arising from the distal right coronary artery (black arrows) as it makes a U-shaped bend over the posterior descending coronary artery (PDA, white arrow). The AV nodal branch extends superiorly toward the inferior mitral anulus (black arrow).

FIG 58-17  Absent left main coronary artery. Axial oblique image from a coronary CTA shows complete absence of the left main coronary artery. The LAD (white arrow) and LCx (black arrow) arise directly from the left sinus of Valsalva (L).

It is a small vessel that courses superiorly toward the posterior anulus of the mitral valve (Fig. 58-16).

CORONARY ARTERY ANOMALIES Occurring in between 0.5% and 1.5% of the population, coronary artery anomalies are frequently encountered on cardiac computed tomography (CT). Although many of these anomalies have little clinical significance, anomalous coronary anatomy has been the cause of sudden cardiac death in 5% to 35% of young people.26,45 Some of these anomalies can lead to symptoms and thus be the reason for the study; however, they are often clinically silent and are frequently an incidental finding. It is also important to remember that although anomalous coronary anatomy is best visualized with electrocardiographically (ECG)-gated CT angiography (CTA), the improved temporal resolution of modern scanners often allows for basic assessment of coronary anatomy even on nongated thoracic CT scans. There are many ways to subdivide these anomalies. One of the more commonly used methods is to divide them based on anomalous origin, course, and termination. No matter how one chooses to classify these anomalies, it is important to recognize that although many of the anomalies are benign, others can lead to a reduction in coronary blood flow and major adverse cardiac events. Therefore it is imperative to recognize the various anomalies on imaging as well as their potential clinical ramifications.

Abnormalities in Origin Numerous congenital coronary artery anomalies are associated with an anomalous origin of a coronary artery. They may be an isolated finding but can be associated with multiple coronary anomalies, including the presence of a single coronary artery with various anomalous origins and courses. These anomalies include absence of the left main coronary artery, anomalous location of the coronary ostium within the aortic root but near the correct sinus, anomalous location of the coronary ostium outside the aortic root, and anomalous origin of the coronary ostium from the incorrect sinus. Within each of these

categories, some anomalies have few clinical repercussions while others can lead to adverse events.

Anomalies of Origin, Benign Absence of left main coronary artery.  Absence of the left main coronary artery is a benign variant that occurs in approximately 0.4% to 2% of the population.93,196 In this situation the LAD and LCx have independent origins from the left sinus of Valsalva (Fig. 58-17).

Anomalous origin of coronary arteries within the aortic root near the proper aortic sinus.  Coronary artery ostia usually arise from the center of their respective aortic sinus. However, in some patients the coronary artery may arise in a slightly eccentric location, although with the correct sinus. For instance a coronary may arise more laterally in the sinus (near one of the valvular commissures), slightly superiorly (near the sinotubular junction), or slightly inferiorly (closer to the valve). Except in rare circumstances, these are considered benign variants.

Anomalous origin of coronary arteries outside the aortic root.  In certain instances the coronary arteries may arise from a loca-

tion outside the sinuses. Although some of these anomalies may be potentially malignant, which will be discussed later, many are considered benign variants. For instance, one of the coronary arteries may have a high origin 1 cm or greater above the sinotubular junction and may be seen arising from the ascending aorta or even the aortic arch (Fig. 58-18). This can occur with either the left main or RCA and is a benign anomaly.100 However, it is important to recognize this anomaly because the vessel could be accidentally injured during ascending aortic surgical manipulation. In extremely rare instances a coronary artery can arise from one of the arch vessels (right brachiocephalic, carotid, or subclavian arteries), internal mammary artery, or even the descending thoracic aorta. Given the rarity of these anomalies, their clinical significance is unclear, although in most instances they should not lead to a compromise in blood flow.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2021

H

L

R

F

FIG 58-19  Retroaortic course. Coronal oblique image from a coronary

FIG 58-18  High origin of the RCA. Anterior volume-rendered image of the aorta and coronary ostia shows the RCA arising from the ascending aorta (white arrow), well above the sinotubular junction (white arrowhead). This is a benign anomaly.

CTA through the aortic root shows the LCx (black arrows) arising from the right coronary cusp and coursing posteriorly between the aortic root and atria. The LCx has an independent ostia from the RCA (white arrow). The LAD (white arrowhead) originates from the left coronary sinus. This is a benign anomaly.

Origin from anomalous sinus, benign

Retroaortic.  In a retroaortic course, the anomalous coronary artery arises from the opposite sinus and courses posteriorly and extends between the aorta and left atrium. This is most commonly seen with the LCx arising from the right sinus of Valsalva, although it can arise directly from the proximal RCA100 (Fig. 58-19). Less commonly the left main coronary artery can have a retroaortic course and be part of a single coronary anatomy. In rare instances the RCA can arise from the left sinus, with a retroaortic course. These are benign anomalies and require no intervention. Anterior to pulmonary outflow tract (prepulmonic or precardiac).  A prepulmonic course occurs when an anomalous coronary artery courses anterior to the RVOT (Fig. 58-20). Most commonly this abnormality involves the LAD or left main coronary artery, which often arises directly from the proximal RCA in the setting of a single coronary artery.206 If the LAD has a prepulmonic course, the LCx may have a retroaortic course or can arise directly from the left sinus. Less commonly the RCA may have a prepulmonic course in the setting of a single left coronary artery. Septal (intramyocardial) course.  A septal course usually involves the LAD arising from the right coronary sinus. It is important to recognize the inferomedial course of the vessel that dives into the proximal aspect of the LV septum (Fig. 58-21). This anomaly needs to be distinguished from the interarterial, potentially malignant course where the anomalous coronary extends medially between the two outflow tracts. Although it can rarely be associated with myocardial ischemia, in most instances a septal course does not lead to a compromise in coronary blood flow.206 Noncoronary sinus.  Origin from the noncoronary sinus is an extremely rare anomaly. This can occur with the RCA or left main.

R RVOT

FIG 58-20  Prepulmonic course. Axial oblique image from a coronary CTA shows LAD (black arrows) arising from the right sinus of Valsalva (R) and coursing anteriorly around the right ventricular outflow tract (RVOT), consistent with a prepulmonic or precardiac course. This is a benign anomaly.

2022

PART II  CT and MR Imaging of the Whole Body

Although there is little information on this anomaly given its rarity, it is often considered a benign anomaly.

artery or RVOT. Compression can worsen during exercise, owing to physiologic dilatation of the aorta and pulmonary artery at a time when myocardial oxygen demand increases. Compression is also due to the intramural course of the coronary where it courses in the wall of the aorta. The intramural course can be of varying length and can often be identified on imaging by soft tissue attenuation surrounding the proximal coronary artery, leading to luminal narrowing with an increased height/width ratio (Fig. 58-22).126 The ostium of the interarterial vessel may be narrowed and is often referred to as slitlike. The proximal vessel can also have a tangential course that leads to proximal kinking of the vessel. An interarterial course of the RCA (see Fig. 58-22) is more common than those involving the left system.106 In patients with an interarterial course of the RCA, a higher incidence of symptoms and major adverse cardiac events has been reported in those with a higher course of the aberrant vessel between the aorta and pulmonary artery as compared to those with a lower interarterial course between the aorta and RVOT.113 However, in most patients an interarterial course of the RCA is an asymptomatic finding and does not lead to ischemia. When an interarterial course of the RCA is identified, surgical correction is recommended if the patient has chest pain, ischemia, syncope, presyncope, or demonstrates LV dysfunction in the appropriate vascular territory.202 If the patient is asymptomatic and this course is an incidental finding, many recommend that the patient undergo a stress test with surgical correction in the setting of inducible ischemia in the affected vascular distribution. If ischemia is absent on stress testing, most currently favor a noninterventional approach.111 Compared to the RCA, an interarterial course of the LAD (Fig. 58-23) or left main has a higher association with myocardial ischemia and sudden cardiac death.126 Because of the high risk of sudden cardiac death with an interarterial course of the left main or LAD, surgical correction is recommended in most cases even if the finding is incidental.111 This can occur via various techniques including unroofing, reimplantation, or bypass.

Anomalies of Origin, Possibly Malignant Interarterial course.  In an interarterial course, a coronary artery arises from the opposite coronary sinus and courses medially between the two outflow tracts. This can involve the RCA, left main, or LAD. Unlike the retroaortic or septal course, an interarterial course, especially when involving the left main or LAD, can lead to myocardial ischemia, infarction, and sudden cardiac death.45 The cause of this is believed to be multifactorial. When the artery courses medially, it can experience extrinsic compression between the aorta and pulmonary

FIG 58-21  Septal course of the LAD. Sagittal oblique MIP image from a coronary CTA in a patient with a single right coronary artery (arrowhead) shows the LAD coursing inferiorly and medially (black arrows) before diving into the interventricular septum. This benign course needs to be delineated from the malignant interarterial course where the coronary artery crosses directly between the main pulmonary artery and aorta. The patient also had a prepulmonic LAD (white arrow) that supplied the anterior and anterolateral segments.

Anomalous origin of left main coronary artery from pulmonary artery (ALCAPA).  ALCAPA, or Bland-White-Garland syndrome, is a

rare congenital anomaly with an estimated incidence of 1 in 300,000 live births.207 In utero, admixture of blood and high pulmonary

PA

A

B

C

FIG 58-22  Incidental finding of an interarterial course of the right coronary artery (RCA) in a 56-year-old man. Axial oblique (A) and sagittal oblique (B) images from a coronary CTA show the RCA (black arrows), which arises above the left coronary sinus, coursing between the aorta and pulmonary artery (PA). The proximal vessel is surrounded by soft tissue density (A) and narrowed in width, with an increased height-to-width ratio. These findings are secondary to the intramural portion of the artery where it courses in the wall of the aorta. A few centimeters distal to the ostium, the RCA is normal in size (white arrow). C, Sagittal oblique image a few centimeters distal to the RCA ostium show a normal rounded appearance to the RCA (white arrow) where the height and width are equal.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

A

2023

B

FIG 58-23  Interarterial course of the left main coronary artery in a 13-year-old girl who presented to the emergency department in ventricular fibrillation. The girl had been playing lacrosse before her cardiac arrest. A, Axial oblique image from a coronary CTA shows the left main (black arrow), which arises above the right coronary sinus, coursing between the aorta and pulmonary artery. B, Sagittal oblique image through the proximal left main shows severe narrowing of the vessel (white arrow) as it courses between the aorta and pulmonary artery. A 7-mm intramural course of the proximal left main was confirmed at surgery when the patient underwent an unroofing procedure.

PA PA

A

B

FIG 58-24  ALCAPA in an 8-week-old infant with poor feeding, lethargy, and irritability. A, Axial oblique image from a CTA shows the left main coronary artery (black arrow) arising from the pulmonary artery (PA). B, Anterior image from catheter angiography confirms presence of ALCAPA, with the left main (white arrow) arising from the PA.

pressures allow for adequate perfusion of the left main coronary artery. In the neonatal period, pulmonary pressures remain high enough to allow for adequate perfusion of the left main coronary artery with pulmonary arterial blood. The age of presentation depends on the development of collaterals early in life. When pulmonary artery pressures decrease in the first few months of life, and if adequate collaterals do not develop, poor flow from the pulmonary artery to the left main coronary artery leads to myocardial ischemia, infarct, and cardiac death in 90% of infants in the first year of life (Fig. 58-24).36,154 However, if a good collateral system from the RCA to left coronary artery territory exists, the decrease in pulmonary pressure will allow for systemic arterial blood to perfuse the left main territory through collaterals from the RCA. This will create flow from the RCA to the left main and into the main pulmonary artery, leading to a physiologic state similar

to a fistula (Fig. 58-25). However, owing to preferential flow into  the pulmonary artery, capillary hypoperfusion may lead to chronic subendocardial ischemia, and patients often present in their third or fourth decade of life with subacute symptoms such as angina, dyspnea, palpitations, and fatigue. However, although patients with ALCAPA may be asymptomatically discovered in their eighth decade of life, ventricular arrhythmia and sudden death is still common in this population.154 Single coronary artery.  Single coronary anomalies encompass less than 1% of coronary anomalies. In most instances there is a single RCA, although a single left coronary artery can occur (see Fig. 58-21).120 In the setting of a single coronary artery, various anomalous courses can be present, including both benign (retroaortic, prepulmonic, septal, wraparound) and potentially malignant (interarterial) courses.

2024

PART II  CT and MR Imaging of the Whole Body

PA

PA Ao

A

*

B

FIG 58-25  ALCAPA in a 53-year-old woman with chest pain. A, Axial image from a CT pulmonary angiogram shows the left main coronary artery (white arrow) arising from the pulmonary artery (PA), consistent with ALCAPA. Note that attenuation in the left main coronary artery is very similar to that of the aorta (Ao) and much less dense than attenuation of the PA, suggesting flow from the aorta through collaterals into the left main. B, Coronary angiogram with a right coronary artery (RCA) injection shows a diffusely dilated RCA (black arrows) and acute marginal branch (black arrowheads), suggestive of a long-standing fistula. Flow from these dilated vessels fills the LCx (white arrowhead), LAD (white arrow), and left main (white star) via retrograde flow. The blood then drains into the PA. Owing to the presence of collateral circulation, the patient was able to survive into adulthood with this congenital abnormality.

It is important to distinguish a single coronary artery from the even rarer coronary ostial atresia. In most cases of ostial atresia, the proximal left main or RCA will be absent.174 However, distal to the focal atretic segment, normal coronary artery anatomy will be seen.

Abnormalities in Course

Myocardial Bridging.  An intramyocardial course of a coronary artery is a common incidental finding and has been reported to occur in up to 58% of patients undergoing coronary CTA98 and up to 86% of autopsies.86,156 Bridging most often involves the mid-LAD, where a band of myocardial tissue extends around the vessel (Fig. 58-26). The depth and length of the bridged segment can vary significantly from a few millimeters to a few centimeters. While the vessel is compressed during systole, this rarely leads to symptoms, because the coronary arteries fill during diastole.10 Although the vast majority of myocardial bridges are an incidental finding on coronary CTA, myocardial bridges can lead to angina and ischemia through various mechanisms including phasic systolic vessel compression, persistent diastolic lumen diameter reduction, increased blood flow velocities, retrograde systolic flow, and reduced coronary flow reserve.20 In addition there is an increased incidence of coronary artery atherosclerotic disease proximal to the bridge, although the bridged segment is typically free of disease.209 Although there are no definitive imaging findings that allow for distinction between incidental and symptomatic bridges, deep bridges are more likely to be symptomatic.133

Split (Double) Coronary Artery.  A split, or double, coronary artery is an extremely rare anomaly. In most instances there is one coronary artery arising from the sinus of Valsalva, which then divides in its proximal portion into two parallel coronary arteries that mirror their course (Fig. 58-27).177 This most commonly involves the RCA, although it has been described with the LAD. Since in most cases there is a single ostia, many have preferred to use the term split to describe this anomaly. In rarer instances there is a true “double” or duplicate coronary artery where each has an independent origin from the aortic sinus, with near

FIG 58-26  Myocardial bridging. Axial oblique image from a coronary CTA shows the mid-LAD diving into the left ventricular myocardium (white arrows), consistent with a myocardial bridge. This is a common incidental finding and of little clinical significance in the vast majority of patients.

parallel courses.30 However, there is continued debate regarding the terminology. In general this is a benign anomaly.

Abnormalities in Termination

Coronary Fistula.  Coronary artery fistulas may be acquired but are most often congenital. Fistulas can involve either the left or right

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2025

CS

FIG 58-28  Coronary artery fistula. Curved MPR image from a coronary CTA shows a diffusely enlarged LCx (white arrows) that drains into the coronary sinus (CS), consistent with a fistula. This was an incidental finding.

FIG 58-27  Split RCA in a 23-year-old man. C-view of the right coronary artery from a coronary CTA shows a single RCA proximally (white arrowhead). However, the RCA splits into two parallel vessels (black and white arrows), one of which terminates as the PDA (black arrows) and the other of which terminates as the proximal LV branch (white arrows). This is a very rare anomaly and is benign.

coronary systems, and the literature varies on which distribution is most common.78,117,171,210 No matter which coronary artery is involved, drainage is most commonly to the right side (from the coronary sinus to the pulmonary artery) and physiologically acts like a left-to-right shunt.117,210 In nearly all cases the involved coronary artery is markedly dilated and tortuous, and such a finding on CT should lead to suspicion of a fistula (Fig. 58-28). Although a fistula can be an incidental finding, patients may present with congestive heart failure due to longstanding shunt, ischemia due to a steal phenomenon (preferential flow of blood through lower-pressure fistula instead of through higherpressure capillary bed), or endocarditis.117

Coronary Veins.  Mapping of the coronary veins on cardiac CT is often valuable prior to biventricular pacer placement and transcoronary vein ablation.176 The coronary sinus provides drainage for most of the cardiac veins. It is located in the posterior AV groove and empties into the right atrium (Fig. 58-29). Its major feeding vessels are the great cardiac vein and middle cardiac vein. The larger coronary veins typically run parallel to the coronary arteries. The proximal portion of the great cardiac vein is adjacent to the LAD, where it is often denoted as the anterior interventricular vein. More distally the great cardiac vein curves into the posterior AV and extends inferiorly to drain into the coronary sinus176 (Fig. 58-30). The middle cardiac vein lies adjacent to the PDA in the posterior interventricular groove. It extends posteriorly and drains into the coronary sinus. The small cardiac vein runs parallel to the RCA in the anterior interventricular groove. It extends inferiorly to terminate in the coronary sinus.176 Small cardiac veins anterior to the RV drain directly into the right atrium. A small vein posterior to the left atrium, the oblique vein of Marshall, provides a connection between the coronary sinus and a left superior vena cava in patients with this abnormality.61

RA

FIG 58-29  Coronary sinus. The coronary sinus (arrow) provides drainage for most of the cardiac veins and drains into the right atrium (RA).

CT TECHNIQUE To optimize image quality, it is critical to properly prepare the patient and select the appropriate imaging protocol that takes into account the indication for the study. Prior to the scan, it is important to explain the procedure to the patient, determine that there are no contraindications to premedication or intravenous (IV) contrast, and coach the patient on breath holding. Appropriate contrast protocols and scanner acquisition techniques are essential.

Patient Preparation Multiple investigators have shown that the best images are obtained in the setting of a low heart rate, ideally below 65 beats per minute (bpm).79 Newer MDCT scanners that provide faster temporal resolution may permit adequate images at higher heart rates44; however, even

2026

PART II  CT and MR Imaging of the Whole Body left circulation is measured. After the TTP is determined, the timing of the diagnostic scan should be based on the TTP plus an additional 4 to 5 seconds’ delay.79 The contrast injection for coronary CTA is typically performed with 60 to 80 mL of a highly concentrated iodine solution (300400 mg/dL) at a flow rate of 5 to 6 mL/sec. For the latter part of the injection, iodine solution may be diluted with saline to decrease attenuation in the RV, which reduces artifact in the RCA system. A saline bolus is added at the end to facilitate transit of the contrast bolus and eliminate artifacts in the highly concentrated contrast in the superior vena cava. LA

FIG 58-30  Great cardiac vein. The great cardiac vein (long white arrows) courses alongside the LAD (black arrow) and drains into the coronary sinus. LA, left atrium.

with such scanners, premedication with a heart rate–decreasing agent may be valuable in individuals with heart rates above 70 bpm. Heart rate control can be achieved using one of several medications. Metoprolol is a β-blocking agent that is commonly employed. It can be administered orally or IV. When used orally a 30-mg dose is often given 45 to 60 minutes before the scan to ensure adequate absorption. For more rapid effect, IV metoprolol can be given a few minutes before the scan. A 5-mg dose is typically administered and can be repeated several times as needed. The main contraindications to use of metoprolol are significant bronchospastic disease and advanced heart block.38,96 An alternative medication is verapamil, a calcium channel blocker that is administered IV in 5-mg doses. Other β-blocking agents are occasionally employed.96 To achieve coronary vasodilatation, nitroglycerin is typically used.96 Prior to administration, it is important to determine that the blood pressure is adequate; nitroglycerin should be avoided in patients who are hypotensive. Nitroglycerin is given as a sublingual tablet or tablets or as a spray moments before the scan is initiated. Nitroglycerin should not be used in patients who are taking phosphodiesterase inhibitors (e.g., sildenafil, vardenafil) or who have hypersensitivity to organic nitrates.

Contrast Injection Because of the importance of rapid injection of IV contrast, it is desirable to insert a large-bore catheter into a vein, ideally 18 gauge or greater. Frontal and lateral low-energy scout images are first obtained to determine the position of the heart. Scan initiation can be performed using either a bolus-tracking or timing-bolus technique. In the former a region of interest is placed in the aorta or left heart and sequential axial images are obtained. The coronary CTA is initiated when the density within the region of interest exceeds a designated threshold value, usually set between 100 and 150 Hounsfield units (HU). For the timing-bolus technique, a preliminary scan using a small quantity of contrast, usually 10 to 20 mL, is injected followed by saline, and the transit time for the time for the contrast to peak (TTP) in the

CT Scanning Methods The initial use of CT technology to evaluate the coronary arteries primarily involved assessing coronary artery calcium and was gathered used electron beam CT (EBCT), a technology with stationary components that permitted a high-temporal-resolution solution needed to freeze physiologic coronary artery motion. With the advent and increasing acceptance of multidetector (MD)CT, EBCT gradually was abandoned. A wide variety of MDCT scanners are currently available. A minimum requirement for coronary CTA is a 64-slice scanner, and there are a number of scanners that meet or exceed this threshold. Scanner configurations range from 64 to 320 slices, consist of 1- or 2-tube configurations (i.e., single or dual source), permit spatial resolution in the 0.5-mm range for all 3 planes, and have temporal resolution as rapid as 66 milliseconds.44,170,191 There has also been substantial advancement in detector and other hardware technology as well as software. The focus on radiation dose reduction has led to meaningful innovations such as introduction of increasingly powerful iterative reconstruction techniques and other valuable automated or selectable parameters to decrease kilovoltage and milliamperage (kVp and mA) factors.65,112 Proper coronary CTA requires ECG gating to minimize or freeze coronary artery motion, and several gating options exist (Fig. 58-31). In retrospective ECG gating, a volume of data is obtained throughout the cardiac cycle using a helical mode. This technique produces the most complete set of images and permits assessment of myocardial wall motion and function in addition to coronary artery evaluation. However, it also produces the highest radiation exposure. One strategy to decrease such exposure is to use ECG-triggered dose modulation, which reduces tube output—and therefore dose—at a designated noncritical point in the cardiac cycle, generally systole. Prospective ECG gating or triggering can be performed in an axial or helical manner. In the axial mode, sequential scans are performed through the volume of interest using a step-and-shoot technique at a prespecified point in the cardiac cycle, usually mid- to late diastole. Tube output is turned off for the remainder of the cardiac cycle, leading to dose reduction on the order of 70% to 80% compared to standard retrospective ECG gating. In prospective helical gating (highpitched helical acquisition) an entire volume is scanned within one heart beat, beginning at a preselected time in the cardiac cycle.107 Major drawbacks of prospective ECG gating are the inability to evaluate cardiac function (because scanning only occurs during a limited part of the cardiac cycle) and difficulty with imaging at higher heart rates (>65 bpm).

Image Reconstruction and Postprocessing The volume of data acquired from scanning is typically reconstructed with slice thickness in the range of 0.5 to 0.8 mm using a 50% overlap. A field of view of 200 to 250 mm centered on the heart is generally reconstructed. A medium smooth reconstruction kernel is used.79

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2027

Tube output (x-ray) ON

A

Helical acquisition

Maximum tube output (100%)

B

Reduced tube output (20%)

Helical acquisition

Tube output ON

Tube output OFF

C FIG 58-31  A, In retrospective ECG gating a helical scan is obtained as the current remains at a constant level throughout acquisition. B, In retrospective ECG gating with dose modulation a helical scan is obtained, but the tube output is decreased during the noncritical parts of the cardiac cycle. C, In prospective ECG gating (triggering) a step-and-shoot axial scan is obtained, and the tube is turned on during critical points in the cardiac cycle, typically diastole.

Although axial images are the cornerstone of evaluation of the coronary arteries, several options exist for postprocessing of the image data, including multiplanar reformatted (MPR), maximum intensity projection (MIP), and volumetric images51 (Fig. 58-32). MPR images are critical for assessment of the coronary arteries. These are generated in the plane of the coronary vessel using a centerline technique. Interactive evaluation of MPR images using in-plane and through-plane visualization permits assessment of the presence and extent of coronary plaque and stenosis. MIP reformatted images may provide a supplemental approach to coronary artery analysis and can be viewed interactively in a sliding thin slab configuration. MIP images may be valuable when considerable image noise is present. The main disadvantage of MIP imaging is the risk of under- or overestimating stenosis due to partial volume averaging. Volumetric images provide a highlevel overview of the course and general status of the coronary arteries and may be used for demonstration or educational purposes. In general they should not be used for diagnosis.

Artifacts on Coronary CTA A host of artifacts may accompany coronary CTA.79 Because even the fastest CT scanners available today have temporal resolution inferior

to that required to freeze physiologic coronary motion—generally around 50 milliseconds—motion artifact remains problematic. Additional artifacts relate to cardiac gating, respiratory motion, blooming, and beam hardening. Cardiac gating artifacts occur in the presence of cardiac arrhythmias and can lead to poor image quality. In patients with premature contractions, software exists that in many cases allows exclusion of the abnormal beat or beats. In patients who undergo axial scanning with prospective ECG gating, there may be offset of the cardiac structures on adjacent axial acquisitions owing to misregistration. During the scan, respiratory motion can produce a stairstep artifact. Respiratory artifact is distinguishable from cardiac gating artifact because of involvement of nonmoving structures such as the sternum. Blooming artifact in the coronary arteries is caused by highattenuation structures, particularly calcified coronary plaque, that appear larger than they actually are because of partial volume averaging. Blooming artifact can be a significant factor in overestimation of coronary stenosis. Use of thinner slices, a sharper kernel, and viewing with wider windows may mitigate but not entirely eliminate the effects of blooming artifact. Beam-hardening artifact is due to absorption of low-energy photons by a high-density structure such as a calcified

2028

PART II  CT and MR Imaging of the Whole Body

LAD

A

C: 238 W: 1032

B

LAD

LAD

C: 487 W: 875

C ED volume : 93.8 ml ES volume : 40.1 ml Stroke volume : 53.7 ml Ejection fraction : 57 %

D

H A

L

Wall mass : 50.0 g Cardiac output : 4077.9 ml/min

LV volume

Vol(ml) 93 84 75 65 56 46 37 28 18 9

E

ED 0

ES 10

20

30 40

50 60

70 80

Phase 90 (%)

FIG 58-32  In addition to axial images, curved (A) and straightened MPR (B) images are typically obtained. C, MIP imaging may be valuable in certain instances, particularly as a problem-solving tool. D, Volumetric imaging may be helpful to show anatomic relationships. E, If retrospective gating is used a ventricular ejection fraction can be calculated.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium plaque or coronary stent, leading to a factitious low-density area in the plane of that structure that can mimic pathology such as a noncalcified plaque.

CORONARY ARTERY DISEASE Assessment of the coronary arteries by MDCT has gained widespread acceptance for multiple indications. A noncontrast coronary artery calcium scoring scan can provide an assessment of risk for major adverse coronary events. Coronary CTA allows evaluation of coronary artery stenosis and remodeling and characterization of coronary plaque. More recently, techniques aimed at determining the functional significance of coronary artery alterations, such as coronary CTA fractional flow reserve and CT-based perfusion imaging, have received considerable attention. These are described in a separate chapter.

Coronary Artery Calcification Coronary artery calcium scoring (CACS) has been well validated as a marker for cardiovascular risk, providing incremental value in some instances over information obtained from the population-based Framingham Risk Score (FRS).193 In particular, asymptomatic individuals with an intermediate FRS (10%-20%) as well as diabetics are viewed as appropriate to undergo CACS to assess the risk of coronary artery disease, according to the 2010 American College of Cardiology Foundation (ACCF) guidelines.193 A weaker recommendation has been given for those with a low to intermediate FRS (6%-10%). CACS is considered inappropriate in individuals at low risk (FRS > 6%).193 Higher CACS is indicative of a greater likelihood of cardiovascular death, with the highest scores associated with a relative risk of 10.8 as compared to individuals without coronary calcium.68 The capability of CT to detect coronary artery calcification on CT was initially described by Guthaner et al. in 1979 on EBCT, and this technique was shown to be more sensitive than chest radiography or fluoroscopy.72,83,188 In 1990, Agatston and his colleagues described the

2029

first practical scoring system for CACS.5 Using 3-mm collimation EBCT, they defined as a calcified focus any pixel in the coronary arteries with a threshold value of 130 HU or greater, using a minimum of 1-mm2 area to exclude noise artifact. They further stratified scoring based on the density measurement above the threshold, assigning a score of 1 for 130 to 199 HU, 2 for 200 to 299 HU, 3 for 300 to 399 HU, and 4 for 400 HU or greater. The Agatston score is still widely used and is the sum of the score (1-4 above) on each slice added together for each scan slice.5 In an effort to produce greater inter- and intrascan reproducibility, volume and mass CACS measurements were introduced subsequently.28,82 However, these approaches are less well validated than the Agatston score. Using the Agatston method, scores of 1 to 10 are considered to reflect minimal coronary artery calcification, 11 to 100 mild, 101 to 400 moderate, and greater than 400 severe calcification. An Agatston score of 400 or more indicates a strong possibility of hemodynamically significant coronary artery stenosis75,169 (Fig. 58-33). With current use of MDCT, CACS is typically obtained with ECG prospective triggering and a radiation exposure of approximately 1.5 mSv.15 A zero calcium score also has important clinical and prognostic value. Asymptomatic individuals with absent calcium on CACS are unlikely to develop a major adverse cardiac event. In a review of 25,903 asymptomatic subjects without coronary calcium, less than 1% had a cardiac event during a 51-month follow-up period.178 The association of a zero calcium score with a low risk for cardiovascular events remained true even in diabetic individuals.159 The timing of repeat CACS in the setting of a zero calcium score is uncertain. However, one study showed that only 14% of patients converted from a zero calcium score to a calcium score above zero after 4 years, indicating this interval might be an acceptable benchmark.131 Increase in the calcium score generally occurs over time, particularly in patients with a non-zero initial score. Slower progression of CACS may be associated with some forms of medical therapy. In a study of women randomized to treatment with estrogens or placebo,

FIG 58-33  A 61-year-old asymptomatic man at intermediate risk for coronary artery disease. Note that the automated program has detected areas of calcification in the LAD and descending coronary arteries. The total Agatston score (box) places this individual in the highest percentile or coronary risk (arrow).

2030

PART II  CT and MR Imaging of the Whole Body

the former group demonstrated a lower calcium score than the latter.121 However, therapy with statins, despite therapeutic effects, has not been shown to affect the calcium score, possibly because stabilized plaque associated with lipid-lowering therapy may also be prone to calcification.11 In contrast, unusually rapid progression of the calcium score is associated with increase in adverse cardiovascular events.23

for functional studies such as stress testing. The study can be performed as a dedicated coronary CTA or, in instances where more nonspecific chest pain is present, as a so-called triple rule-out (TRO) study.73 The TRO protocol consists of a comprehensive ECG-gated study through the entire thorax, enabling enhancement of both leftand right-sided circulations. Its main disadvantages as compared with

Coronary Stenosis Current capabilities of coronary CTA include anatomic assessment, plaque characterization, CT perfusion, and CT evaluation of fractional flow reserve (FFR).168 Advances in MDCT technology permit routine evaluation of the coronary artery lumen. The most recently published multisociety appropriateness guidelines for coronary CTA consider stable symptomatic patients at low to intermediate risk for coronary events according to the FRS suitable for this test.194 Coronary CTA in these patients often occurs after a nondiagnostic or equivocal stress test. A summary of indications for coronary CTA are listed in Box 58-1. As noted, evaluation of coronary stenosis is typically performed on thin-section axial and MPR images.51 Axial and MPR images can be supplemented by MIP or volume-rendered images (Fig. 58-34). A dedicated workstation can be used to perform semiautomated stenosis calculations with reference to normal proximal or distal coronary segments. Plaque morphology and quantification can also be assessed. Stenosis of 50% or greater is generally considered significant (Figs. 58-35 and 58-36). Beginning with four-slice MDCT, studies have shown a tendency toward increasing sensitivity and negative predictive value in assessing significant coronary artery stenosis in comparison to catheter-based angiography.4,57,163 More recent multicenter trials using 64-slice CT have demonstrated even greater efficacy for morphologic evaluation.22,123,127 A 2008 meta-analysis of 64-slice coronary CTA among 1296 patients in 28 studies showed a pooled sensitivity of 98%, specificity of 89%, positive predictive value of 93%, and negative predictive value of 100% in assessing coronary artery stenosis compared with catheter angiography.140 A later meta-analysis using dual-source CT technology demonstrated similarly excellent results even in the setting of higher heart rates.172 An additional focus of MDCT is to evaluate acute chest pain in the emergency department. Noninvasive imaging is useful for patients at low to intermediate risk for acute ischemia. MDCT provides an anatomic option to assess such patients and may complement or substitute

A

BOX 58-1  Summary of Indications for

Cardiac CTA

Detection of CAD • Symptomatic patients with low to intermediate pretest probability of coronary artery disease (CAD) • Suspicion of acute coronary syndrome (ACS) with normal ECG/biomarkers and low to intermediate pretest probability of CAD • Newly diagnosed heart failure with low to intermediate pretest probability of CAD • Preoperative evaluation for noncoronary cardiac surgery, with intermediate pretest probability of CAD • After ECG exercise testing with results indicating intermediate risk • After stress imaging with equivocal results • New or worsening symptoms in patients with previously normal stress imaging Postrevascularization • Evaluate graft patency in symptomatic patients after CABG • Localize grafts and other retrosternal anatomy prior to further cardiac surgery • Risk assessment in asymptomatic patients after stenting (stents > 3 mm) Evaluation of Cardiac Structure and Function • Vascular anomalies and variants • Coronary aneurysms and dissections • Assess for sequelae of known ischemic disease • Cardiac masses (tumor or thrombus) • Characterization of native and prosthetic valves • Cardiomyopathies • Vasculitis, collagen vascular diseases • Congenital heart disease Pericardial Diseases

B FIG 58-34  A 56-year-old woman with atypical chest pain. Multiplanar (A) and curved planar reformatted images (B) of the LAD demonstrate small areas of calcified and noncalcified plaque (arrow). No substantial coronary artery stenosis is evident.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

A

B

C

D FIG 58-35  A 59-year-old man with a chest pain syndrome. Initial coronary CTA demonstrated relatively mild coronary artery disease, as shown in MPR images of the LAD (A) and right coronary artery (B). Three years later, these vessels (LAD [C], RCA [D]) demonstrate substantial progression of coronary artery disease.

2031

2032

PART II  CT and MR Imaging of the Whole Body

A

LAD

B

LOSSY

+Occluded mid LAD

C FIG 58-36  A 57-year-old man with previous anterior myocardial infarction. A and B, Curved planar images demonstrate extensive coronary artery disease and chronic occlusion of the LAD (arrow). C, LAD occlusion is confirmed on catheter coronary angiography.

dedicated coronary CTA are a higher radiation exposure and contrast dose. Using dedicated coronary CTA, three recently published prospective trials have compared the use of coronary CTA to standard-of-care therapy in the emergency department setting.64,80,115 Each demonstrated the capability of coronary CTA to permit earlier discharge than the standard-of-care protocol, without a significant difference in adverse cardiac events. Additional efforts have been directed at assessing prognostic factors related to the use of coronary CTA. The CONFIRM (Coronary CT Angiography Evaluation for Clinical Outcomes: an International Multicenter) registry, an ongoing international multinational study of patients referred for coronary CTA, demonstrated a significantly lower all-cause mortality for patients without coronary artery disease, as compared to patients with nonobstructive and obstructive coronary artery disease.128 Annualized death rates in patients without coronary

artery disease followed for 4 or more years were extremely low (0.22%). In this group the authors suggested that repeat coronary imaging might not be necessary for 4 or more years. Patients with two- and three-vessel disease had significantly worse outcomes than those with nonobstructive and one-vessel coronary disease.

Coronary Plaque Coronary CTA provides a comprehensive assessment of the location, severity, and characteristics of coronary atherosclerotic plaque. A  vulnerable plaque is defined as a plaque at risk to cause an acute coronary event. Pathologically these plaques typically have a lipid-rich core, a thin fibrous cap, and loss of integrity of the endothelium, with platelet aggregation.205 Vulnerable plaques are often clinically silent until the fibrous cap ruptures and the necrotic core pours into the vascular lumen, leading to a thrombogenic reaction. This is because

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

A

RCA

B

CRX

C

2033

LAD

FIG 58-37  A 69-year-old man with severe coronary artery disease. Curved planar reformatted images of the right coronary (A) and LCx arteries (B) demonstrate extensive calcified plaque interspersed with noncalcified plaque. C, Curved planar reformatted image of the LAD demonstrates extensive mixed plaque and positive remodeling (arrows).

many vulnerable plaques tend to undergo positive vascular remodeling, with outward expansion of the vessel rather than narrowing (negative vascular remodeling).40,189,205 Therefore it is critical to assess for imaging findings of both coronary narrowing and vulnerable plaque (Fig. 58-37). Documented CT features of plaque associated with acute coronary events include positive vessel remodeling and low-attenuation plaques.137 In particular, plaques with attenuation values less than 30 HU correlate with low-attenuation atherosclerotic plaques as visualized on intravascular ultrasound.136 Motomaya et al. investigated plaque characteristics that were associated with subsequent development of an acute coronary event in 1059 patients. Of patients with both positive vessel remodeling and low-attenuation plaques, an acute event occurred in 22.5% of patients, compared with 3.7% of patients with one characteristic and 0.5% with neither positive remodeling nor low-attenuation plaque.138 An additional coronary CTA finding that may be indicative of a thin-cap atheromatous vulnerable plaque is termed the napkin-ring sign (Fig. 58-38). This is visualized as a rim of high attenuation, representing the inflamed fibrous cap, surrounding an area of low attenuation, representing the necrotic lipid core.143,147 In a study of 895 consecutive patients who underwent coronary CTA, Otsuka found that the napkinring sign was present in 41% of those who ultimately developed an acute coronary event and was an independent predictor of the acute coronary syndrome.151 Overall plaque volume as demonstrated on coronary CTA correlates with risk for a cardiac event. Using a semiautomated plaque quantification algorithm, Versteylen investigated plaque volumes among 1650 patients who underwent 64-slice CTA.198 They found that patients with the acute coronary syndrome had significantly higher total and noncalcified plaque volume than control patients. Moreover, the patients with acute coronary syndrome had higher per-plaque

maximal volume, higher percentage of noncalcified plaque, and higher plaque burden. Plaque quantification also provided incremental value over the conventional coronary CTA interpretation. Nevertheless, coronary CTA has constraints in assessing coronary plaque. Owing to limitations in spatial resolution it cannot reliably distinguish vulnerable thin-cap fibroatheroma from more stable thickcap atheroma. Moreover, contrast injection may alter CT densities, leading to difficulty in distinguishing lipid and fibrous plaque.27

Miscellaneous Coronary Artery Diseases Coronary artery aneurysm is defined as a segment of the coronary artery that measures more than 1.5 times the adjacent normal coronary artery.41 Coronary artery aneurysms can be classified by involvement of all coronary layers (true vs. false) and morphology (fusiform vs. saccular). There are myriad etiologies of coronary artery aneurysms, including atherosclerosis, Kawasaki’s disease and other vasculitides, inflammatory disorders, connective tissue disease, infection, trauma, fistulas and anomalies, and iatrogenic causes. In the developed world, atherosclerosis is the most common cause of coronary artery aneurysms in adults, whereas Kawasaki’s disease is the most common etiology in children. Coronary CTA accurately depicts the size and location of the aneurysm, as well as internal characteristics such as thrombus and calcification (Fig. 58-39). Diffuse narrowing of the coronary arteries with mural thickening may be caused by coronary vasculitis.

Additional Considerations In spite of a large body of evidence demonstrating the utility of coronary CTA, reimbursement for the procedure is not uniform. In the United States, two studies of patients at low to intermediate risk for major cardiac events have demonstrated a cost savings of 26% to 33% for coronary CTA in comparison with myocardial perfusion singlephoton emission computed tomography (SPECT).129,132 No additional

2034

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 58-38  Napkin-ring sign. A, Cross-sectional view of the middle segment of the right coronary artery shows atherosclerotic plaque with a thin rim of high attenuation (which represents the inflamed fibrous cap) surrounding an area of low attenuation (which represents the necrotic lipid core). B, Longitudinal view of the same plaque demonstrates stenosis with positive vascular remodeling. (Images courtesy Dr. Seung Min Yoo, MD, PhD. Department of Diagnostic Radiology, CHA University Bundang Medical Center, Seongnam, South Korea.)

adverse outcomes were observed when the coronary CTA algorithm was used. Nevertheless, substantially variability in reimbursement for coronary CTA persists. In asymptomatic patients at low to intermediate risk for coronary artery disease, CACS is considered an appropriate method of investigation, whereas coronary CTA is not.208 Nevertheless, screening coronary CTA is performed routinely in some parts of the world, particularly Asia, although it does not receive third-party reimbursement. Compared with CACS, disadvantages of screening coronary CTA include the requirement IV contrast, potential higher radiation exposure, and concern about an increased need for downstream testing. Moreover, there is an absence of literature supporting generalized use of screening coronary CTA in patients viewed as appropriate for CACS. Nevertheless, screening coronary CTA may have added value in certain subgroups. One study in asymptomatic diabetics showed that approximately one third of individuals with a calcium score of zero had subclinical atherosclerosis, and 10% had obstructive coronary stenosis.130 Whereas the negative predictive value of coronary CTA for assessment of coronary stenosis is excellent, the positive predictive value relating to the frequency of false-positive results is less optimal. Two functional techniques that may supplement coronary CTA and increase positive predictive value are coronary perfusion and CT-derived fractional flow reserve. These are described in a separate chapter.

CORONARY ARTERY BYPASS GRAFTS Since its introduction in 1962, coronary artery bypass grafting (CABG) has remained the definitive treatment for advanced coronary artery disease. Patency of the coronary grafts is critical for long-term survival and depends on the type of graft used. In the Veterans Administration Cooperative Study, internal mammary grafts had a patency rate of 85% after 10 years, as compared to 61% for saphenous vein grafts (SVG).63 Internal mammary artery grafts typically remain attached to their origin at the subclavian artery. They are dissected from the parasternal region and are often anastomosed to the LAD because of their higher patency rate. Saphenous vein grafts are harvested from the legs and

attached as a free graft to the ascending aorta and coronary artery distal to the site of obstruction.56 In patients with recurrent chest pain after CABG, the traditional approach to diagnosis has been catheter-based coronary angiography. The advent of MDCT provided a noninvasive alternative for assessment of CABG patency. Earlier generations of MDCT provided excellent sensitivity and specificity for detection of graft occlusion, and good specificity but only modest sensitivity for high-grade stenosis.145,165 Moreover, the number of patients with completely evaluable bypass grafts in whom stenosis could be excluded was limited.9 More recently, 64-slice or greater MDCT has resulted in further improvement in the assessment of bypass grafts (Fig. 58-40). Occluded grafts are readily detected, and there has been some improvement in evaluation of CABG stenosis119,164,204 (Fig. 58-41). Using 320-slice MDCT, de Graaf et al. found sensitivity, specificity, positive predictive value, and negative predictive value of 96%, 92%, 83%, and 98%, respectively.39 However, there remain challenges in evaluating nongrafted segments and areas distal to the bypass graft. Recent attention has focused on the prognostic value of coronary CTA in evaluating bypass graft patients. A number of studies have demonstrated that anatomic information on coronary CTA provides incremental value over clinical assessment in such individuals.33,185 Mushtaq et al. assessed 721 patients who underwent bypass grafting and whose grafts were evaluable with coronary CTA. They found that the number of nonbypassed coronary territories had significant prognostic value in predicting future myocardial infarction (MI) and death from cardiovascular disease.141

CORONARY STENTS Bare metal coronary stents were introduced for human use in 1986 and provided a more robust method to maintain coronary artery patency than previous techniques. The use of coronary stents was shown to result in a larger residual diameter and reduce the rate of in-stent restenosis to approximately 30% as compared to coronary angioplasty.53,180 More recently the introduction of drug-eluting stents has permitted further reduction in restenosis rates.81 Overall, in-stent

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

A

2035

B

FIG 58-39  A 67-year-old woman with coronary artery aneurysms. Axial (A) and volumetric images (B) reveal two coronary artery aneurysms (arrows), believed to be due to atherosclerosis.

A

C

B

A F

D

F

FIG 58-40  A 72-year-old man who had previously undergone coronary artery bypass grafting. Curved planar (A and C) and volumetrically reconstructed images (B and D) show patent left internal mammary artery– to-LAD graft (arrow in A and B) and saphenous vein–to–posterior descending artery (arrow in C and D) graft, respectively. Note the larger caliber of the saphenous vein graft.

restenosis occurs in 20% to 35% of patients after placement of bare metal stents and in 5% to 10% of patients with drug-eluting stents.53,180 However, there have been recent reports of late, sometimes fatal thrombosis in drug-eluting coronary stents, and longer-term antithrombotic therapy is now recommended.190 Coronary CTA can be used to evaluate stent occlusion and in-stent restenosis (Figs. 58-42 and 58-43). In general, image reconstruction

with a sharp kernel, a small field of view, and use of wide windows is recommended to optimize stent visualization obscured by blooming artifact from the metallic struts of the stent. Even with CT-scanning using 64 or more slices, there are limitations in imaging coronary stents. With current technology, studies demonstrate a high negative predictive value (usually 90%-100%) for significant stenosis and a rate of nonevaluable stents of 8%. In a

2036

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 58-41  Images in two patients showing a “ghost” sign (arrow in A) and a “nubbin” sign (arrow in B), indicating occlusion of coronary artery bypass grafts.

A

RCA

B

FIG 58-42  A 54-year-old man with coronary stent. Curved planar (A) and multiplanar (B) reformatted images of the RCA show a patent stent in the proximal vessel (arrows). Note contrast within the stent and excellent flow distally.

meta-analysis of in-stent restenosis in 14 studies using 64-slice or greater MDCT, Kumbhani et al. found overall sensitivity of 91%, specificity of 91%, positive predictive value of 68%, and negative predictive value of 98%. If nonevaluable stents are included, overall sensitivity, specificity, positive predictive value, and negative predictive value were 87%, 84%, 53%, and 97%, respectively.108 In general, stents 3 mm or more in diameter are more likely to be assessable. The 2010 multisociety expert consensus document on coronary CTA concluded that evaluation of coronary stents is appropriate for left main stents, but its utility is uncertain for other large stents (≥3 mm) placed at least 2 years earlier. CTA was deemed inappropriate for smaller or more recently placed stents.193 Limited data exist on the use of coronary CTA for evaluation of in-stent thrombosis or stent fracture. More recent technologic advances such as myocardial per­ fusion and iterative reconstruction may lead to improved stent visualization.118

wall motion and ejection fraction as well as infarction size. The lack of radiation exposure, excellent spatial and contrast resolution, reproducibility, and versatility are among the strengths of cardiac MRI.84

MRI IN CORONARY ISCHEMIA

with a steady-state free precession (SSFP) sequence that enables excellent contrast between the high-signal-intensity blood pool and medium-signal-intensity myocardium. When placed in a cine loop, SSFP imaging allows assessment of focal wall-motion abnormalities that may indicate ischemia or infarction, as well as evaluation of cardiac function.7 Although less frequently employed, dark-blood

Magnetic resonance imaging (MRI) has become a mainstay in the assessment of ischemic heart disease (Fig. 58-44). In particular it can distinguish myocardial ischemia from infarction and provide a measure of the extent of myocardial injury and prognosis by evaluating regional

MRI Technique The strength of the magnetic field of the scanner is measured in tesla (T) units. Most cardiac scanning is performed using a 1.5-T or 3-T magnet. A 3-T magnet has the advantage of an improved signal-tonoise ratio, but examinations can sometimes be limited by imaging artifacts due to magnetic field inhomogeneity.84 Important sequences of ischemia evaluation on MRI are bright-blood imaging and postcontrast evaluation of the myocardium using perfusion and delayed myocardial enhancement sequences. Phase-contrast imaging is valuable to assess flow through the mitral valve that may suggest incompetence due to MI. Stress MRI can also be performed.84

Bright-Blood Imaging.  Bright-blood imaging is often performed

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2037

FIG 58-43  A 61-year-old woman with coronary stent. Top, A 2D map from an MPR image reveals a stent in the LAD (arrows). Bottom, MPR image shows small foci of low density within the stent (arrows), consistent with mild endothelial ingrowth.

A

B

C

FIG 58-44  A 48-year-old man with recent myocardial infarction. A, Perfusion image in short axis shows a defect in the anterior wall of the LV (arrow). B, Late gadolinium-enhanced image demonstrates segmental high signal intensity in the same region (arrow). C, Catheter coronary angiography confirms occlusion of the LAD (arrow).

T2-weighted sequences can be used to demonstrate high signal intensity in the myocardium that indicates myocardial edema due to an acute ischemic event rather than chronic myocardial injury.2

Postcontrast Assessment.  Postcontrast assessment is typically obtained with a gadolinium chelate, a T1-shortening IV contrast medium that accumulates in the extracellular space following an MI. After infusion with gadolinium chelate, perfusion imaging permits a first-pass assessment of the integrity of myocardial flow. Segmental absence of flow involving the subendocardium is consistent with myocardial ischemia or infarction. Delayed images usually are acquired 10 to 15 minutes after gadolinium administration. Owing to expansion of the extracellular space in infarcted tissue, washout of gadolinium occurs slowly as compared to normal myocardium. This retained gadolinium leads to T1 shortening and high signal intensity on sequences with nulling of the normal myocardium, a phenomenon termed late gadolinium enhancement (LGE). LGE corresponding to a coronary artery segmental territory in

a subendocardial location is indicative of MI. The transmural extent of LGE is related to likelihood of successful revascularization with bypass grafting or percutaneous coronary intervention. Because the subendocardial myocardium is most vulnerable to ischemia, the extent of LGE in nontransmural MI may be limited to the subendocardial region. Areas of over 50% transmural extent of LGE, which are characteristic of a transmural MI, are much less likely to respond to revascularization than those with lesser amounts of transmural involvement.101 Stress MRI using SSFP and perfusion imaging can be used to distinguish ischemia from prior MI. Rest MRI followed by stress MRI is often performed using pharmacologic agents such as adenosine because of the challenges of introducing MRI-compatible equipment into the MRI suite to perform physiologic stress testing. Similar to other imaging-based stress techniques, segmental abnormalities that are present on both stress and rest imaging indicate infarction, whereas areas that are normal on rest imaging and abnormal on stress imaging suggest myocardial ischemia. Although not widely used, recent studies of stress MRI demonstrate sensitivity of 0.82 to 0.92 and specificity of

2038

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 58-45  A 20-year-old man who presented after a syncopal episode. A and B, Axial MRIs using a wholeheart technique demonstrate an anomalous left coronary artery coursing between the aorta and pulmonary artery (arrows), consistent with a malignant course.

0.75 to 0.94, suggesting that stress MRI is as good as conventional approaches of stress imaging.35

MRI for Direct Coronary Artery Visualization Coronary CTA remains the preferred noninvasive imaging technique for direct evaluation of the coronary arteries, but coronary MR angiography (MRA) has demonstrated value as a potential alternative technique, particularly in the assessment of anomalous coronary arteries (Fig. 58-45). Although there are multiple techniques used to assess the coronary arteries on MRI, coronary MRA is often performed with SSFP sequences using a free-breathing whole-heart technique that allows assessment of coronary artery position and disease. Contrast is not necessary, owing to the high intrinsic contrast of the blood pool on SSFP sequences. Using 50% or more coronary arterial stenosis on catheter coronary angiography as a standard, a multiinstitutional study demonstrated sensitivity of 88%, specificity of 72%, and accuracy of 79% for whole-heart MRA.95

CARDIAC CHAMBERS AND VALVES Chambers Right Atrium.  The primitive right atrium and sinus venosus give rise to the right atrium, which forms the right lateral margin of the heart and receives systemic and coronary venous return.134 The sinus venosus forms the smooth-walled portion of the right atrium, the sinus venarum, and receives blood from the superior vena cava, inferior vena cava, and coronary sinus. The primitive right atrium develops into the trabeculated component of the right atrial appendage (Fig. 58-46E  and F). The right atrial appendage has a pyramidal or triangular morphology with a wide opening at its base and rough trabeculation due to the pectinate muscles. These pectinate muscles should not be misinterpreted as a thrombus or mass.195 The crista terminalis is a smooth muscle ridge that extends in vertical orientation from the superior vena cava to the inferior vena cava (see Fig. 58-46G). It reflects the line of fusion between the primitive right atrium and the posterior smooth wall of the sinus venosus. The size and shape of the crista terminalis is variable, and it may be mistaken for an intracardiac thrombus or mass.134 The eustachian valve is found at the junction of the inferior vena cava and right atrium. This valve directs blood flow toward the foramen ovale in the embryo and normally regresses or disappears during embryonic development.60 When visible it is typically a thin, mobile, linear structure. The thebesian valve is found within the right atrium

at the orifice of the coronary sinus and is identifiable in a minority of patients on imaging. This valve prevents reflux of blood from the right atrium into the coronary sinus. The SA node is located in the right atrial wall at its junction with the superior vena cava, whereas the AV node is in the floor of the right atrium. Small vessels that supply blood to these nodal regions are often visible on imaging.74

Left Atrium.  The left atrium defines the posterosuperior aspect of the heart and receives pulmonary venous return (see Fig. 58-46B-H). The primitive left atrium connects to a common pulmonary vein that divides into four pulmonary veins, the proximal portion of which is incorporated into the posterior left atrium. This posterior component later becomes the smooth portion of the left atrium. The remaining trabeculated portion of the primitive left atrium becomes the left atrial appendage. The left atrial appendage is more tubular in shape, has a narrower base, and contains fewer trabeculated pectinate muscles than its counterpart on the right (see Fig. 58-46A). Pectinate muscles in the left atrial appendage have a linear or globular appearance that can be misinterpreted as an intracardiac mass or thrombus.195 The right and left superior pulmonary veins drain their respective upper lobes and the middle lobe and lingula, respectively (see Fig. 58-46A and B). The inferior pulmonary veins drain their respective lower lobes (see Fig. 58-46C-H). However, there is considerable variability in the size and configuration of the pulmonary veins.122 Common variants are a single left pulmonary vein and a separate rightsided vein draining the right middle lobe. The ostia of the superior pulmonary veins tend to be larger than the inferior pulmonary veins. The ligament of Marshall is a fold of the visceral pericardium containing muscle, vascular, and neural structures. It courses obliquely between the left atrial appendage and the left superior pulmonary vein bundle in the epicardial aspect of the left atrial fold, forming a focal area of smooth muscle also known as the coumadin or warfarin ridge, or Q-tip (see Fig. 58-46B). This structure should not be mistaken for a mass or thrombus.71 The interatrial septum is typically thin and may not be well visualized.201 During embryonic development of the atrial septum there is normally greatest thinning of the septum at the fossa ovalis, which can simulate an atrial septal defect on cross-sectional imaging. Lack of flow between the left atrium and right atrium on dynamic studies in the region of the thinned septum confirms the absence of a defect. There is typically fusion of a majority of the septum primum and septum secundum after birth, obliterating the embryonic foramen

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

MPA RAA AA

MPA

AA

RSPV SVC

LAA

SVC

RSPV

RPA LSPV LPA

DA

DA

A

B

PV

PV

AA SVC

SVC

LM LA

RSPV

LAD

AA

CX

RSPV

LA

LIPV

LIPV

DA

DA

C

D

RVOT

RVOT

RA

RA LAD

AA

AA

*

SVC

SVC CX

LA

LA

LIPV DA

E

F

DA

FIG 58-46  Normal cardiac anatomy. Axial contrast-enhanced ECG-gated CT images from the level of the pulmonary arteries (A) to the level of the inferior aspect of the heart (J) display normal cardiac anatomy. A, CT image obtained at the level of the right pulmonary artery (RPA). B, CT image at a slightly lower level shows the coumadin ridge (black arrow), a normal infolding of the wall of the left atrium posterior to the left atrial appendage. C, CT image at the level of the left atrium (LA) and pulmonary valves (PV). D, CT image at the level of the origin of the left main coronary artery (LM) giving rise to the left anterior descending (LAD) and circumflex (CX) coronary arteries. E, Left anterior descending (LAD) and circumflex (CX) coronary arteries. F, CT image at the level of the superior myocardium of the LV (asterisk). Continued

2039

2040

PART II  CT and MR Imaging of the Whole Body

RCA

RV

RV LV

RA

RA

LV

P

LA

LA RIPV

DA

G

DA

H

RV RV RA

LV

P

I

J

FIG 58-46, cont’d  G, CT image at the origin of the right coronary artery (RCA). The aortic valve leaflets are also visible (black arrows). H, More inferior CT image shows the moderator band extending from the interventricular septum to the anterior free wall of the RV (curved arrow). In the LV the chordae tendineae (arrowhead) can be seen extending from the papillary muscle (P) to the mitral valve leaflet (arrow). I, CT image at level of the coronary sinus (arrow) just before it enters the right atrium (RA). Note the papillary muscle (P) within the LV. J, CT image showing the posterior course of the RCA (arrows) in the AV groove. AA, ascending aorta; DA, descending aorta; LAA, left atrial appendage; LIPV, left inferior pulmonary vein; LM, left main coronary artery; LPA, left pulmonary artery; LSPV, left superior pulmonary vein; MPA, main pulmonary artery; RA, right atrium; RAA, right atrial appendage; RIPV, right inferior pulmonary vein; RSPV, right superior pulmonary vein; RV, right ventricle; RVOT, right ventricular outflow tract, also known as the infundibulum; SVC, superior vena cava.

ovale. If the foramen ovale is covered but the tissues fail to completely fuse, the result is a patent foramen ovale. A patent foramen ovale manifests as slitlike connection between the right and left atria, typically causing a left-to-right shunt.201 Although patients with patent foramen ovale are usually asymptomatic, they are at risk for paradoxical emboli.

Right Ventricle.  The RV is the most anterior chamber of the heart (see Fig. 58-46G-J). It has a triangular shape and receives blood from the right atrium via the tricuspid valve. It is more trabeculated, and its wall is thinner than the LV. In addition to heavy trabeculation, features of the RV that distinguish it from the LV are the presence of the moderator band and separation by contractile muscle of the AV and ventriculoarterial (semilunar) valves.195 The RV contains three components: (1) an inlet portion consisting of the tricuspid valve, chordae tendineae, and papillary muscles; (2) the trabeculated apical myocardium; and (3) the infundibulum, or

conus, corresponding to the smooth outflow tract. The right ventricular inlet and outlet are separated by the septomarginal trabeculation or the infundibuloventricular fold.175 Prominent muscular bands within the RV include the parietal band, septomarginal band, septal band, and moderator band. The most commonly visualized septomarginal trabeculation is the moderator band, which contains the right bundle branch (see Fig. 58-46H). The moderator band is a tissue ridge that extends across the RV apex from the anterior papillary muscle to the interventricular septum. A displaced moderator band can simulate a double-chambered RV, and a large moderator band in the setting of RV enlargement or hypertrophy can mimic a thrombus or tumor.195 The RV contains three papillary muscles: the anterior, posterior, and medial papillary muscles, respectively. Each muscle attaches to two cusps. On cross-sectional imaging, the most commonly visualized muscle is the anterior papillary muscle.60 If prominent, this structure can be mistaken for an intracardiac mass or thrombus.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2041

Nonspecific fat deposition is found in up to 17% of asymptomatic individuals in the RV and is most often encountered in the elderly. Most often it is found in the superior wall of the base, the middle portion of the RV, and the RVOT. Fat deposition can mimic findings of arrhythmogenic RV dysplasia or a prior MI involving the RV. Functional assessment can be used to differentiate nonspecific fat deposition from a diseased RV.97

Left Ventricle.  The LV consists of the inlet, apical, and outlet portions; it is separated from the RV by the interventricular septum (see Fig. 58-46H-J). It receives blood from the bicuspid mitral valve. The interventricular septum typically bulges toward the RV because of a relatively higher pressure.60 The LV contains two papillary muscles, the anterior and posterior papillary muscles, which are larger than the RV papillary muscles (see Fig. 58-46H and I). The papillary muscles are attached to the mitral valve leaflets by chordae tendineae.60 The posterior papillary muscle can have a globular shape that may lead to misinterpretation as a mass or thrombus. Two features of the LV that distinguish it from the RV are a smooth wall and a shared anulus between the mitral and aortic valves.195 The normal wall thickness of the LV myocardium is 6 to 11 mm. Focal wall thinning at the apex with normal cardiac function on crosssectional imaging may be present as a normal variant in the absence of symptoms or history of MI.74 The interventricular septum is composed of a thin membranous portion located immediately beneath the aortic valve and a larger muscular septum more basally. A thin membranous septum can simulate a ventricular septal defect.92 Lack of blood flow between the two ventricular chambers assures an intact ventricular septum.

Valves Cardiac valvular disease causes significant morbidity and mortality, particularly disease involving the aortic and mitral valves. It is estimated to represent 20% of cardiac surgical procedures performed in the United States.17 Echocardiography is generally considered the first line of investigation. However, both CT and MRI play important roles in the evaluation of valvular heart disease. CT scanning can precisely assess valve morphology, and if retrospective ECG gating is used, functional assessment can be performed, including direct assessment of valve morphology and the effects of valvular disease on cardiac dimensions, mass, and ejection fraction. MRI can also evaluate valve morphology using cine bright-blood SSFP sequences, and phase-contrast imaging is valuable to quantify flow across the valve. Stenosis and regurgitation is visible on SSFP sequences and is visualized as an area of signal dropout.84 However, qualitative assessment of flow on SSFP sequences often overestimates the degree of regurgitation or stenosis. Imaging of valvular vegetations or masses and postoperative valvular assessment can be achieved with CT or MRI.

Aortic Valve.  The normal aortic valve is composed of three cusps with semilunar attachments to the annular ring. Adjacent cusps join superiorly to form valve commissures. The area of the normal aortic valve is 2.5 to 3.5 cm2. The three sinuses of Valsalva are formed by shallow convexities on the aortic side of the valve and are designated right, left, and noncoronary sinuses based on the coronary arteries that arise from them.17 Congenital abnormalities.  Bicuspid aortic valve is common and occurs in 1% to 2% of individuals (Fig. 58-47). There are two functioning cusps, which may reflect two anatomic cusps or a fusion of two of three aortic cusps, most commonly the right and left cusps. Bicuspid

FIG 58-47  Bicuspid aortic valve. ECG-gated contrast-enhanced cardiac CT reformatted image of the aortic valve in a true axial plane (arrow). Note the false commissure, or raphe, in the anterior cusp (arrowhead).

aortic valve is most associated with aortic valvular stenosis but also can lead to valvular regurgitation, endocarditis, and aortic aneurysm and dissection, which occur as part of a proximal aortic aortopathy. These complications frequently occur at a younger age than in patients with tricuspid valves.184 Other rare congenital abnormalities include unicuspid aortic valve, which manifests with stenosis at an early age, and quadricuspid aortic valve, which is frequently accompanied by aortic regurgitation.17 CT and MRI effectively characterize congenital aortic valve abnormalities. Aortic stenosis.  The most common etiology of aortic stenosis is age-related degenerative change associated with atherosclerotic disease, which generally manifests in the sixth decade and beyond. This causes nodular calcification and fibrosis of the cusps. The same process occurs in bicuspid valves but at an earlier age.31 Rheumatic disease is the most common cause of aortic stenosis in developing countries and is typically also associated with aortic regurgitation. Aortic stenosis leads to an increase in afterload and initially LV hypertrophy. Ultimately LV dysfunction develops, associated with symptoms of angina, syncope, and heart failure. CT and MRI provide excellent definition of the thickened valve cusps, progressively decreasing valve area, poststenotic ascending aortic dilatation, and LV wall thickening. CT typically shows dense calcification of the valve leaflets (Fig. 58-48). MR SSFP and phase-contrast sequences demonstrate a stenotic jet, and phasecontrast imaging can be used to calculate the gradient across the valve.17 Surgery is performed in most patients with advanced aortic stenosis. In recent years, transcatheter aortic valve implantation (TAVI) is being used increasingly, particularly for patients at high risk for conventional surgery.186 CT is typically performed as part of the pre-TAVI work, and measurements of the aortic valve complex and aorta are performed to assess aortic root diameter, the status of the coronary arteries, and the extent of peripheral vascular disease. Aortic regurgitation.  Aortic regurgitation is due to a wide variety of conditions, including diseases of the valvular leaflets (i.e., bicuspid valve, rheumatic heart disease, infectious endocarditis), the aortic root (i.e., systemic hypertension, aortic dissection, cystic medial necrosis, Marfan’s or Ehlers-Danlos syndromes, syphilitic aortitis), or both.31 In

2042

PART II  CT and MR Imaging of the Whole Body

AA

LA

LV

FIG 58-48  Degenerative aortic stenosis in an 82-year-old man. Axial contrast-enhanced CT images show calcification of the aortic valve (arrowheads). LA, left atrium.

FIG 58-49  Aortic regurgitation. Coronal MPR image from spiral CT

the developed world, bicuspid aortic valve and aortic ectasia are the most common causes of aortic regurgitation, whereas rheumatic heart disease is most common in the developing world. In chronic aortic insufficiency, a gradual increase in LV volume and pressure lead to LV dilatation and hypertrophy, followed by remodeling and ultimately LV failure. Acute aortic insufficiency manifests as pulmonary edema due to sudden increase of left atrial and ventricular filling pressures.31 CT or MRI obtained during diastole shows lack of coaptation of the aortic valve leaflets and permits calculation of the size of the regurgitant orifice.17 The characteristics of the LV and ascending aorta can be accurately assessed, and the studies may provide an indication of the underlying etiology of aortic regurgitation (Fig. 58-49). On MRI a diastolic flow jet is visible emanating from the valve into the LV. Regurgitant volume and fraction can be calculated. Aortic valve replacement is performed in symptomatic patients and asymptomatic patients with substantial LV dilatation or dysfunction.17 Other aortic valve entities.  CT and MRI are useful to detect vegetations in patients with valvular endocarditis, particularly those that exceed 1 cm.17 Postoperative imaging of the aorta is sometimes performed to evaluate complications such as valvular dehiscence, pannus, and thrombus formation. MRI may be limited by artifacts from the prosthetic valve.

Mitral Valve.  The bicuspid mitral valve has a crescentic posterior cusp and a semicircular anterior cusp.135 It is connected by chordae tendineae to the papillary muscle. The anterior and posterior valve leaflets attach to the mitral anulus by a D-shaped ring in the anterior AV groove. The anterior leaflet covers more of the mitral orifice than the posterior leaflet and forms a portion of the left ventricular outflow tract (LVOT).135 Mitral stenosis.  The most common cause of mitral stenosis is rheumatic fever.135 Other etiologies include mitral annular calcification, congenital mitral stenosis with parachute deformity, and connec-

following administration of IV contrast medium. There is dilatation of the left ventricle (LV) and aneurysmal dilatation of the ascending aorta (AA), typical findings in patients with aortic regurgitation.

tive tissue disorders. In rheumatic mitral stenosis, fusion of the leaflet edges occurs along the commissures, accompanied by thickening of the chordae tendineae. Fibrosis and calcification associated with thickening and tethering of the mitral valve apparatus leads to elevation of atrial pressures, atrial fibrillation, and pulmonary venous hypertension. Mitral regurgitation is often present.31 On CT and MRI the valve leaflets are thickened and a characteristic fishmouth appearance is visible on the two-chamber view. Left atrial size can be quantified and the mitral orifice measured.125 CT may demonstrate calcification of the leaflets (Fig. 58-50). MRI shows a stenotic jet extending from the valve into the LV and permits quantification of the severity of mitral stenosis by measurement of the mean diastolic gradient across the valve during diastole.135 Mitral regurgitation.  Mitral regurgitation is the most common of all valvular lesions.135 Primary regurgitation directly affects the valve apparatus and includes rheumatic heart disease, mitral valve prolapse, and infectious endocarditis. Mitral valve prolapse is due to myxomatous degeneration of the valve, more often the posterior leaflet, and is associated with elongation and occasionally rupture of the chordae tendineae. Secondary etiologies of mitral regurgitation are due to changes in LV geometry or function of the papillary muscles and are frequently a result of ischemic or hypertrophic cardiomyopathy.31 Acute mitral regurgitation causes sudden volume overload and frequently results in pulmonary edema. It is infrequently evaluated with CT or MRI. In chronic mitral regurgitation there is compensatory dilatation of the LV and atrium until heart failure supervenes. CT or MRI is often performed as part of surgical planning.135 CT and MRI are useful to measure the regurgitant orifice and calculate the hemodynamic effects on the LV and atrium.70 A regurgitant volume and fraction can be calculated. In mitral valve prolapse,

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

RV

2043

RV RA

RA

LV

LA LA

DA

FIG 58-51  Mitral annular calcification. Axial contrast-enhanced CT FIG 58-50  Mitral stenosis. Marked thickening and calcification of the mitral valve (arrows), with enlargement of the left atrium (LA). RA, right atrium; RV, right ventricle.

prolapse of the valve more than 2 mm into the left atrium is considered diagnostic. Surgery for chronic mitral regurgitation is indicated for LV ejection fraction below 60% or LV end-systolic diameter greater than 40 mm.135 Other mitral conditions.  Mitral annular calcification is a degenerative process involving the anulus and adjacent myocardium and may undergo central necrosis. It is typically asymptomatic and is well visualized on CT as an area of calcification sometimes with internal areas of soft tissue attenuation89 (Fig. 58-51). On MRI the calcification may be recognized as low signal intensity on all sequences. If larger than 1 cm, valvular vegetations in the setting of infective endocarditis may be visualized on either CT or MRI. Other rare etiologies of mitral valve disease include tumors (e.g., papillary fibroelastoma) and congenital mitral windsock and cleft leaflet deformities.135

Pulmonary Valve.  The pulmonary, or pulmonic, valve is the right ventriculoarterial valve and consists of three cusps (anterior, right, left), similar to the aortic valve. The valve is more delicate than the aortic valve because of lower right-sided pressures, and it lacks coronary ostia. The normal area of the pulmonary valve is approximately 2 cm2 per square meter of body surface area.31 Diseases of the pulmonary valve, either stenosis or regurgitation, are uncommon compared to the left-sided valves. Pulmonary valve stenosis is nearly always congenital. The most common form is a dome-shaped valve consisting of two to four raphes with incomplete separation of the cusps. Less common is a dysplastic valve with thickened immobile cusps.13 A bicuspid or quadricuspid pulmonary valve is rare, occurring in 0.1% to 0.2% of individuals, and may cause stenosis or regurgitation.173 Congenital pulmonary stenosis should be suspected on CT or MRI if thickened valve leaflets are associated with enlargement of the main and left pulmonary artery, with a relatively normal-caliber right pulmonary artery (Fig. 58-52). This configuration is due to the flow jet that is directed posteriorly into the left pulmonary artery.173 Subvalvu-

image shows calcification of the mitral anulus (arrows). DA, descending aorta; RA, right atrium; LA, left atrium; LV, left ventricle; RV, right ventricle.

lar pulmonary stenosis is often present in patients with tetralogy of Fallot. MRI may be useful to distinguish valvular from subvalvular pulmonary stenosis and assess its hemodynamic consequences.173 Although mild pulmonary regurgitation can occur in conditions that dilate the pulmonary valve ring (e.g., pulmonary hypertension), substantial regurgitation is rare and is often found in patients with tetralogy of Fallot, particularly after surgery.31 MRI is widely used in this setting because of its ability to quantify pulmonary regurgitation and RV volumes and function as part of postsurgical follow-up.173 Other rare conditions that may cause pulmonary valve disease are carcinoid heart disease and infective endocarditis.

Tricuspid Valve.  The tricuspid valve consists of three cusps (septal, anterior, posterior), the anulus, chordae tendineae, and papillary muscles. The tricuspid valve is thinner than the mitral valve, presumably because of lower right-sided pressures.31 Most cases of tricuspid stenosis are due to rheumatic heart disease, which can also lead to tricuspid regurgitation.31 Rarer causes include carcinoid heart disease and congenital etiologies. CT and MRI demonstrate narrowing of the valve and tethering of the leaflets, although CT may be limited owing to mixing of contrast with unopacified blood in the right heart. Associated enlargement of the right atrium and venae cavae can be assessed.167 Flow and gradient measurements across the valve can be made using MRI. Tricuspid regurgitation is often due to secondary stretching of the tricuspid anulus, occurring in conditions with elevation of pulmonary arterial pressures due to severe left-sided heart disease, pulmonary vascular disease, or chronic pulmonary disease.31 Primary causes of tricuspid regurgitation include infective endocarditis, trauma, carcinoid heart disease, Marfan’s syndrome, and Ebstein’s anomaly, in which there is congenital displacement of the septal and posterior tricuspid leaflets into the RV. The excess volume in tricuspid regurgitation leads to dilatation of both right atrium and ventricle (Fig. 58-53). Chamber volumes and regurgitant volumes and fraction can be 

2044

PART II  CT and MR Imaging of the Whole Body

AA

MPA RPA LPA

A

B

FIG 58-52  Pulmonary stenosis. A, Axial enhanced CT imaging through the pulmonary valve level demonstrates calcification of the valve and thickening of the leaflets (arrow). B, More superiorly, the main (MPA) and left pulmonary (LPA) arteries are dilated with a normal-caliber right pulmonary artery (RPA).

measured most optimally with MRI.167 In patients with infective endocarditis, larger vegetations may be visible on the tricuspid valve.

Ventricular Aneurysm Cardiac aneurysms nearly always arise from the LV and are typically classified as true or false aneurysms. True aneurysms are composed of a continuous, thin, scarred myocardium, whereas false aneurysms reflect a ruptured wall with interruption of the myocardium.182 A true LV aneurysm is defined by the presence of systolic dyskinesia or paradoxical bulging. The large majority of true aneurysms are due to a prior transmural MI, particularly involving the LAD territory. Thus apicoanterior true aneurysms are more common than inferoposterior true aneurysms. Initial aneurysm formation is reported to occur between 2 days and 2 weeks after an MI.124 Less common etiologies of true LV aneurysms include trauma, Chagas’ disease, sarcoidosis, and congenital aneurysm. A false ventricular aneurysm is less common and typically occurs 5 to 10 days following MI.182 It represents a rupture through the myocardium, with containment of the false chamber by inflammatory tissue or pericardium in communication with the true cavity. An  inferior myocardial predominance has been reported for false LV aneurysms. Distinction between true and false ventricular aneurysm can be problematic. Findings that favor the latter are a narrow neck (<50% ratio) compared to the false cavity, absence of coronary arteries overlying the aneurysm, myocardial discontinuity at the neck of the aneurysm, and possibly inferior location.58,182 Echocardiography is often the first study to identify a ventricular aneurysm. CTA is quite effective at delineating the location, size, wall calcification, and associated thrombus (Fig. 58-54). MRI has incremental value in that LGE sequences can define the causative infarction and aid in distinguishing true from false aneurysms. Substantial pericardial enhancement adjacent to the aneurysmal segment on LGE sequences is highly suggestive of false aneurysm and is presumably due to pericardial inflammation.104 From a prognostic standpoint, false ventricular aneurysms have a high risk of lethal rupture and are therefore considered an indication for surgery, whereas true ventricular aneurysms have a low risk for rupture and are often managed medically in the early postinfarction period, followed by later consideration for surgery.182

Left Atrial Assessment for Radiofrequency Ablation Both CT and MRI provide excellent assessment of the left atrium prior to radiofrequency ablation (RFA). CT is generally performed with a technique similar to that outlined for coronary artery assessment, with focus on the left atrium. For MRI a contrast-enhanced MRA or a freebreathing whole-heart technique is generally used.42 Either technique provides an assessment of the size of the left atrium and the location, configuration, and size of the pulmonary veins. The pulmonary vein orifices are typically measured as part of the examination (Fig. 58-55). Variations in pulmonary anatomy are readily identified and can shorten the procedure, which involves identifying the ostia of the pulmonary veins, which are typically near areas of ablation.110 The presence of thrombus—a contraindication to RFA—can be evaluated. The left atrial appendage is often the site of thrombus. On CT a pseudothrombus caused by underfilling of the appendage may mimic a true thrombus. Delayed imaging is useful to show filling in of the left atrial appendage in this case.85 CT has also been used recently to define the size and shape of the left atrial appendage prior to percutaneous left atrial appendage closure.109 Cross-sectional imaging is valuable to identify complications of RFA such as pulmonary vein stenosis or occlusion, which occurs in 3% to 8% of patients undergoing the procedure.166 Either CT or MRI show the stenosis and allow quantification of its extent. Other complications that can be evaluated include pericardial effusion, cardiac tamponade, and atrioesophageal fistula and associated neurologic symptoms.

NONISCHEMIC CARDIOMYOPATHY AND MYOCARDITIS The classification of nonischemic cardiomyopathies has evolved considerably over the years. Current cardiomyopathy classification systems recognize that there is both a morphologic and genetic component.46 Common cardiomyopathies include hypertrophic cardiomyopathy, dilated cardiomyopathy, arrhythmogenic RV cardiomyopathy, various restrictive cardiomyopathies such as amyloidosis, sarcoidosis, and iron overload states, and unclassified cardiomyopathies. Cardiac MRI is most frequently used to image nonischemic cardiomyopathies. The general protocol is similar to that for ischemic  heart disease, with variation for certain cardiomyopathies. MRI can

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2045

RV

RA

A

B TRICUSPID VALVE REGURGITATION Systole

Diastole

25

80

20

60 40

10

20

5 0

0 0

100

200

300

400

500

600

–5

800 –20 –40

–10

–60

–15

–80

–20 –25

–100 Avg vel [cm/s]

C

700

Net flow [mL/sec]

Avg velocity [cm/sec]

15

Net flow [ml/s]

Time [msec]

FIG 58-53  Tricuspid regurgitation. A, Bright-blood image shows mild dilation of the right-sided cardiac chambers. B, Phase-contrast image shows a regurgitant jet in the region of the tricuspid valve. C, Reversal of flow seen during late systole is consistent with tricuspid regurgitation. The arrow denotes maximum reversal of flow and velocity. RA, right atrium; RV, right ventricle.

determine LV or RV ejection fraction and assess focal or global wall thickening, and LGE sequences can assess the extent of myocardial inflammation or fibrosis, which often has prognostic value. Cardiac CT with retrospective gating can also provide functional and morphologic assessment but has limited capability to detect fibrosis and involves radiation exposure and iodinated contrast material.

Hypertrophic Cardiomyopathy In hypertrophic cardiomyopathy (HCM), which has a strong genetic component, part or all of the myocardium becomes hypertrophied,

leading to diastolic dysfunction.34 Both obstructive and nonobstructive forms of HCM exist. In the most frequent obstructive form, subaortic valvular dynamic systolic obstruction occurs in the presence of systolic anterior motion (SAM) of the anterior leaflet of the mitral valve. Nonobstructive forms typically affect the myocardium globally or involve the apex. MRI or retrospectively gated cardiac CT can assess myocardial function, focal or global myocardial thickening, and identify SAM in the subaortic obstructive form of HCM.34 MRI has the additional advantages of allowing calculation of the subvalvular gradient and visualization of myocardial fibrosis on delayed enhancement

2046

PART II  CT and MR Imaging of the Whole Body

sequences (Fig. 58-56). There is increasing evidence that the extent of LGE correlates with prognosis.

Dilated Cardiomyopathy Nonischemic dilated cardiomyopathy (NICM) can be familial or due to diabetes, viral infections of the heart, toxins, or excessive alcohol use.59 MRI or retrospectively gated cardiac CT can assess LV ejection fraction, chamber morphology, and diastolic and systolic dysfunction. On MRI, linear midmyocardial-wall LGE is associated with ventricular arrhythmias and a worse prognosis. However, in most cases of NICM, delayed enhancement is absent.

Arrhythmogenic Right Ventricular Cardiomyopathy In ARVC, there is global or regional dilatation of the RV and occasionally the LV due to fibro-fatty replacement, resulting in ventricular dysfunction and tachycardia. A familial component is often present.160 The 2010 ARVC task force has defined a number of major and minor MRI criteria that contribute to the diagnosis. These include regional RV akinesia, dyskinesia, or dyssynchrony in addition to increased enddiastolic volume or reduced RV ejection fraction.160 In addition to RV dilatation, dyssynchrony, and areas of focal outpouching, fatty infiltration may be visualized in the RV free wall on MRI. Fat-suppression

LSPV

RSPV

LA

LV

RIPV LIPV H P

R Posterior view

FIG 58-55  Radiofrequency ablation. Preablation contrast-enhanced FIG 58-54  LV aneurysm. Vertical long-axis view from a cardiac CT demonstrates an inferior wall LV aneurysm (arrow). The wide mouth of the aneurysm compared to the aneurysm cavity is consistent with a true aneurysm. LA, left atrium.

A

ECG-gated CT exam. Volume-rendered CT image of the posterior wall of the left atrium shows the standard anatomy of the pulmonary veins. LIPV, left inferior pulmonary vein; LSPV, left superior pulmonary vein; RIPV, right inferior pulmonary vein; RSPV, right superior pulmonary vein.

B FIG 58-56  Hypertrophic cardiomyopathy. A, Three-chamber image from an SSFP MRI sequence demonstrates myocardial hypertrophy, particularly involving the basal septum (arrowhead), and a flow void (arrow) caused by subaortic stenosis. B, Short-axis LGE image shows multiple areas of enhancement in a nonvascular territory.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2047

sequences can be used to confirm its presence. Although LGE is typically absent, it may occasionally be found in the tricuspid region extending toward the RVOT.160

shorter T2* times. In a patient with a myocardial iron overload state, the finding of a T2* relaxation time less than 20 milliseconds is associated with LV failure and remodeling.

Restrictive Cardiomyopathy

Unclassified Cardiomyopathy

In restrictive cardiomyopathy (RCM), stiffening of the myocardium leads to poor diastolic filling and rapid increase in ventricular pressures with small volumes. Infiltrative myocardial processes such as amyloidosis, sarcoidosis, and iron overload states are well-established causes of RCM.46 MRI is the primary technique used to evaluate RCM. Cardiac amyloidosis involves deposition of extracellular glycoproteins and may be primary or occur in the setting of familial predilection or systemic disease. Morphologic alterations include biventricular thickening, left atrial enlargement, and thickening of the papillary muscles and valve leaflets.59 Because of the extent of involvement, global and transmural LGE is often present, although a global subendocardial pattern may occur (Fig. 58-57). Poor nulling of the myocardium in preparation for the LGE sequence is typical of cardiac amyloidosis.59 Cardiac sarcoidosis is characterized by infiltration of noncaseating granulomas into the myocardium. Cardiac involvement occurs in 25% to 50% of patients. It is symptomatic in only 5% but may lead to cardiac arrhythmias, heart failure, and sudden death.99 The disease progresses from an active inflammatory phase to a fibrotic phase.  On MRI, myocardial edema may be visualized on T2-weighted sequences in the inflammatory phase. Late gadolinium enhancement is most often found in the basal septal regions and free wall of the myocardium.199 Myocardial iron overload states may be due to hereditary hemochromatosis or multiple blood transfusions.59 Systolic dysfunction is present, and LV dilatation and heart failure may ultimately supervene. The presence of paramagnetic ferritin in iron overload states leads to changes in myocardial tissue characteristics. On MRI, SSFP sequences may show loss of signal in the myocardium in parallel to signal loss in the liver.59 Myocardial iron content is inversely proportional to T2* relaxation times, with greater amounts of iron being associated with

Unclassified cardiomyopathies include LV noncompaction (LVNC) and takotsubo cardiomyopathy. In LVNC there is a congenital failure of myocardium to undergo compaction, leading to persistence of the embryonic spongiform myocardium. Although it primarily involves the LV, the RV is affected in approximately 40% of patients. Systolic and diastolic dysfunction is common and may lead to heart failure.203 Both cardiac CT and bright-blood MRI demonstrate reduced ejection fraction and increased trabeculation corresponding to the noncompacted myocardium at the LV apex and extending along the lateral and inferior walls. Based on echocardiographic criteria, an MRI ratio of noncompacted to compacted left ventricular myocardium over 2.3 : 1 on diastolic short-axis images is consistent with LVNC.155 Takotsubo cardiomyopathy is characterized by apical ballooning of the LV myocardium in the absence of obstructive coronary artery disease. Its presentation can mimic an acute ischemic event. The condition is typically stress related and resolves in most instances over a period of days to weeks.46 Cardiac CT and MRI show apical akinesis and relative basal hyperkinesis that produce the characteristic ballooning found in this condition. A reduced ejection fraction is typical. MRI may also demonstrate myocardial edema on T2-weighted sequences  in a nonvascular distribution. Late gadolinium enhancement MRI sequences are typically normal.52

A

Myocarditis Myocarditis is a spectrum of acute and chronic inflammatory conditions that affect the myocardium. Most cases are due to viral infection. Other infectious etiologies include bacteria such as Corynebacterium diphtheriae and Streptococcus and parasites such as trypanosomes in Chagas’ disease. Cardiotoxic drugs are also responsible.187 Presenting symptoms may resemble acute myocardial ischemia, but obstructive coronary artery disease is absent. MRI is frequently used in this setting

B FIG 58-57  Amyloidosis. A, Four-chamber image from an SSFP MRI sequence demonstrates global thickening of the LV myocardium (arrows). B, Four-chamber LGE image demonstrates poor overall myocardial nulling with areas of septal and lateral enhancement (arrows).

2048

PART II  CT and MR Imaging of the Whole Body

LV

LV

RV

A

B FIG 58-58  Myocarditis in a 33-year-old man with acute chest pain, elevated troponin levels, and a normal catheter cardiac angiography. A, Short-axis image from an SSFP MRI sequence demonstrates normal myocardial thickness. No functional abnormality was noted. B, Short-axis LGE image shows patchy midmyocardial and subepicardial areas of enhancement. LV, left ventricle; RV, right ventricle.

to evaluate the myocardium. Typical features include myocardial edema on T2-weighted sequences and enhancement on LGE sequences that is typically midmyocardial or subepicardial and nonvascular in distribution187 (Fig. 58-58).

MASSES Cardiac masses can be classified as primary and secondary cardiac tumors and tumorlike conditions such as thrombus. Echocardiography is the most widely used imaging technique to evaluate cardiac masses; it is simple, noninvasive, widely available, and relatively inexpensive. CT and MRI are important adjunctive techniques in tumor assessment and can demonstrate the location and morphology of the mass.187 A standard cardiac CT protocol is used. Retrospective ECG gating may be valuable to demonstrate movement of the mass during the cardiac cycle but involves a higher radiation dose.94 On MRI, darkblood imaging with T1- and T2-weighted sequences can be added to the standard protocol, and tissue characterization of the mass using long inversion recovery times can distinguish thrombus from tumor. Fat-suppression imaging is valuable to identify tumors with fat content. Late gadolinium-enhancement sequences may be useful to distinguish thrombus and benign from malignant tumors, which typically demonstrate increasingly greater enhancement.21,148 Using CT or MRI the extent of tumor invasion into adjacent tissues can be shown. Cardiac tumors are divided into primary (benign and malignant) and secondary (metastatic) tumors. Metastatic tumors of the heart are more common than primary cardiac tumors, which are uncommon. The estimated prevalence of primary cardiac tumors is 0.001% to 0.06%. Malignant primary cardiac tumors are especially rare. Patients with cardiac tumors may be asymptomatic, but often the tumor exerts significant clinical effects by causing embolic, arrhythmogenic, or obstructive manifestations.

Benign Cardiac Tumors

Myxoma.  Myxoma is the most common primary cardiac tumor, accounting for about 50% of all benign cardiac tumors and 25% of all cardiac tumors. The most common anatomic location is the left atrium

(75%-80%), followed by the right atrium (10%-20%). Myxomas less commonly involve either ventricle (5%-10%). Involvement of the mitral valve has been rarely described.67 Constitutional symptoms of fever, malaise, weight loss, anemia, and elevated sedimentation rate are reported and may be related to an autoimmune reaction initiated by the tumor or due to superinfection of the myxoma. Approximately 7% of myxomas occur as part of Carney’s triad, which includes hyperpigmented skin lesions, extracardiac tumors such as pituitary adenomas, breast fibroadenomas, and melanotic schwannomas in addition to myxomas.67 Myxomas are well-defined, smooth, lobular, or oval and often pedunculated but can also be polypoid with a stalk. Villous or papillary myxomas have multiple fine, fragile, villous extensions and pose the greatest risk of embolization. The lesion has a gelatinous composition, histologically described as an acid-mucopolysaccharide–rich stroma, and often contains cysts and areas of necrosis, calcification, and hemorrhage.67 As visualized on CT or MRI, myxoma involving the left atrium nearly always arises from the atrial septum in the region of the fossa ovalis, and larger tumors may prolapse through the mitral valve during diastole.94 Myxomas involving the right atrium and ventricles are intracavitary and may arise from any surface. On CT, calcification is visible in approximately 10% of cases.94 On MRI, T1- and T2-weighted images show heterogeneous signal intensity, reflecting the variability of the lesion, which may contain myxomatous, fibrous, or hemorrhagic tissue or calcification.12 The tumor typically enhances following IV gadolinium administration148 (Fig. 58-59). Bright-blood SSFP imaging demonstrates a lesion that is hyperintense compared to myocardium but hypointense relative to the blood pool.187

Lipoma.  Lipoma consists of encapsulated fat cells and is the second most common primary tumor of the heart, constituting 8% to 12% of primary tumors.187 The majority arise in a subepicardial location, although they can also originate from the endocardium and pericardium. Most lipomas are incidental findings, but larger ones can cause obstruction to flow or ventricular compression. On CT the lesions are typically round and contain a homogeneous appearance, with HUs in

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2049

LV LV

LA

A

LA

B

FIG 58-59  Myxoma in a 67-year-old woman with a left atrial (LA) lesion on echocardiography. A, Fourchamber image from a dark-blood sequence shows a mass arising from the atrial septum in the region of the fossa ovalis (arrow). B, Four-chamber LGE image reveals speckled enhancement within the mass (arrows). LV, left ventricle.

the fatty range (≤50 HU).94 On MRI, lipomas have heterogeneous high signal intensity on T1-weighted images, somewhat lower signal intensity on T2, and low signal intensity on fat-suppression images. Postcontrast enhancement is absent.148 Lipomas are distinct from lipomatous hypertrophy of the atrial septum, a tumorlike condition that is discussed separately.

Papillary Fibroelastoma.  Papillary fibroelastoma occurs nearly as often as lipoma; 90% are associated with the cardiac valves, particularly the aortic and mitral valves. Some 75% of cardiac valvular tumors are papillary fibroelastomas.187 The lesion consists of dense connective tissue with an endothelial lining and is typically quite small, on the order of 1 to 1.5 cm. Most are asymptomatic, but their mobility may predispose to systemic embolization, and they are often removed surgically.66 Both CT and bright-blood MRI typically show a small, mobile, homogeneous mass attached to the valve, sometimes by a pedicle (Fig. 58-60). On MRI, T1-weighted images demonstrate intermediate signal intensity, whereas T2-weighted images show hyperintensity. Enhancement may be visible on LGE sequences.148 The main differential diagnostic consideration is a valvular vegetation, but a vegetation is more likely to be located at the free edge of the valve and associated with valvular destruction.187

Other Benign Cardiac Tumors.  Rhabdomyomas are the most common primary cardiac neoplasm in children. They are often diagnosed in the neonatal period or infancy. Although they may occur in isolation, multiple rhabdomyomas are associated with tuberous sclerosis and are present in approximately 50% of affected infants. The tumors tend to regress spontaneously, obviating surgery in many cases.16 Although uncommonly obtained, cardiac CT imaging shows a smooth lesion or lesions with attenuation similar to myocardium.94 MRI demonstrates a myocardial lesion that is typically hypointense relative to myocardium on T1-weighted imaging and mildly

hyperintense on T2-weighted imaging. There is little or no postgadolinium enhancement.148 Cardiac fibromas are the second most common primary tumor in children and may also occur in young adults. They may be asymptomatic or manifest with cardiac arrhythmias. They are intramyocardial tumors that consist of fibroblasts and may grow sufficiently to obstruct the myocardial cavity. Calcification within the lesion is common. CT demonstrates a frequently sizeable homogeneous low-attenuation mass arising from the myocardium that may contain a focus of central calcification.94 On MRI, fibromas are typically isointense or hypointense compared with myocardium on both T1- and T2-weighted imaging, in keeping with their fibrous composition. On postgadolinium sequences, variable appearances include absence of enhancement and enhancement with an isointense or hyperintense rim.187 Hemangiomas are benign vascular tumors that account for 5% to 10% of benign cardiac tumors.67 They may occur in any part of the heart and have a slight predilection for the ventricles. Patients with hemangiomas may present with dyspnea on exertion.148,187 The lesion appears heterogeneous on nonenhanced chest CT and usually demonstrates intense enhancement on postcontrast CT. On MRI, signal intensity resembles that of hepatic hemangiomas. They are isointense compared to myocardium on T1-weighted images and hyperintense on T2-weighted images. Hemangiomas enhance intensely on postgadolinium sequences in a pattern that can be inhomogeneous because of interspersed calcification and fibrous septa within the masses.94,187 Other rare benign tumors include teratoma and paraganglioma.

Malignant Cardiac Tumors

Metastatic Tumor.  Noncardiac malignancy may invade the heart from adjacent organs by transvenous extension or be due to metastatic spread via hematogenous or lymphatic routes.25 Lung carcinoma is the most common tumor to affect the heart, accounting for 30% to 40% of metastatic cardiac malignancies, followed by hematologic malignancies (10%-20%) and breast cancer (10%). Melanoma is the primary

2050

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 58-60  Papillary fibroelastoma in a 45-year-old man with a small valvular lesion on echocardiography. In-plane (A) and through-plane (B) views from cardiac CT demonstrate a small tumor attached to an aortic valve leaflet (arrows).

neoplasm most likely to metastasize to the heart. The most common site of metastasis is the pericardium, followed by the epicardium, myocardium, and endocardium.62 Cardiac CT offers precise anatomic localization of metastatic disease within the heart (Fig. 58-61). Myocardial thickening or multifocal nodularity is often present. Postcontrast imaging may demonstrate enhancement and is particularly useful for evaluating transvenous extension of tumor.94 On MRI, metastatic tumors generally have low signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging.187 The exception is metastatic melanoma, which may demonstrate hyperintensity on both T1- and T2-weighted sequences owing to the paramagnetic effects of melanin.139 Metastatic pericardial disease is discussed separately.

Angiosarcoma.  Angiosarcoma is the most common primary malignant cardiac tumor, comprising nearly 40% of cases. It consists of mesenchymal cells that form poorly defined vascular channels lined by atypical endothelial cells.25 The tumor nearly always involves the right side of the heart, with right atrial involvement in 80%. Encasement of the RCA and pericardial involvement may be present. The tumor is often metastatic at presentation and has a poor prognosis.90 Cardiac CT demonstrates a large, irregular, aggressive-appearing mass with a broad-based attachment to the right heart, particularly the right atrium. A connection between the tumor cavity and right atrium may be identified on early postcontrast imaging. Pericardial involvement is indicated by a sheetlike appearance as the tumor conforms to the pericardial space and provokes a pericardial effusion. On delayed CT imaging the lesion is heterogeneous and often grossly hemorrhagic.94 MRI demonstrates morphologic findings that are similar to CT (Fig. 58-62). Angiosarcoma demonstrates variable signal intensity on T1-weighted imaging, heterogeneously hyperintense signal intensity on T2-weighted imaging, and heterogeneous enhancement after contrast administration.148

Other Sarcomas.  Undifferentiated sarcoma is the second most common primary malignant tumor, accounting for 24% of cases.187 It is typically located in the left atrium and manifests as a discrete or

irregular mass. Primary cardiac leiomyosarcomas and osteosarcomas account for just fewer than 10% of primary malignant tumors each.94 Both most commonly occur in the left atrium and are invasive, potentially leading to left heart failure. Leiomyosarcoma may be multiple and may also originate from the inferior vena cava, leading to right heart failure. Osteosarcomas typically contain large amounts of calcification.148 Rhabdomyosarcoma is the most common primary malignant neoplasm in childhood and accounts for 4% to 7% of primary malignant cardiac tumors overall.187 The tumor manifests as an infiltrative mass arising from the myocardium and lacks a specific chamber pre­ dilection. Liposarcoma is extremely rare (1% of primary malignant cardiac tumors) and has a predilection for the left atrium. It contains little if any macroscopic fat.187 These sarcomas generally have a poor prognosis. Cardiac CT imaging of other sarcomas shows an irregular and often large mass and permits precise localization and definition of invasion. Osteosarcoma typically demonstrates extensive calcification. Although osteosarcoma can mimic myxoma because of its calcification and left atrial location, it typically is not related to the fossa ovalis, unlike myxoma. Liposarcoma may occasionally show areas of fat attenuation.94 On MRI, most sarcomas are isointense on T1-weighted imaging, hyperintense on T2-weighted imaging, and enhance after gadolinium administration.187

Lymphoma.  Primary cardiac lymphoma is quite rare, representing 1.3% of primary cardiac tumors, and typically is a non-Hodgkin’s B-cell lymphoma. Immunocompromised individuals are at higher risk, often related to infection with the Epstein-Barr virus. Primary cardiac lymphoma is most common in the right heart and pericardium. Prognosis is generally poor, although chemotherapy may occasionally be effective.91 On cardiac CT, lymphoma generally manifests as an ill-defined irregular mass affecting the right side of the heart or pericardium and often encasing adjacent structures such as the coronary arteries and aortic root.91 MRI reveals an ill-defined mass with nonspecific isointense signal intensity relative to myocardium on T1-weighted imaging,

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

A

2051

B FIG 58-61  Metastatic leiomyosarcoma in a 51-year-old woman with fibroid disease. Axial (A) and sagittal (B) images from cardiac CT show a large mass bridging in the right ventricular outflow tract extending to the main pulmonary artery. At surgery the lesion proved to be a metastatic leiomyosarcoma.

RV RV

RA

RA

B

A

FIG 58-62  Angiosarcoma in a 32-year-old woman with shortness of breath. Four-chamber dark-blood (A) and bright-blood (B) sequences demonstrate a mass infiltrating the anterolateral wall of the right atrium (RA). RV, right ventricle. (Case courtesy Elizabeth Weihe, MD.)

isointensity or hyperintensity on T2-weighted imaging, and variable enhancement after IV contrast administration.91

Tumorlike Lesions

Thrombus.  Thrombus is the most common cause of an intracardiac mass and is associated with atrial fibrillation, mitral valve disease, or a severely dysfunctional LV due to MI. A thrombus due to atrial fibrillation usually occurs in the left atrial appendage or posteriorly in the

left atrium, whereas a thrombus in the LV usually adheres to the akinetic wall or apex.94 The differentiation between thrombus and other types of intracardiac mass (particularly myxoma) is often impossible to make on echocardiography. Postcontrast cardiac CT provides excellent delineation of the location and morphology of cardiac thrombus (Fig. 58-63). Thrombus appears as a low-attenuation filling defect within the enhanced  blood pool. Chronic thrombus may occasionally demonstrate spotty

2052

PART II  CT and MR Imaging of the Whole Body

A

B FIG 58-63  Right atrial thrombus in a 64-year-old man with a central venous catheter. Axial (A) and coronal (B) images from a cardiac CT scan reveal a thrombus within the superior right atrium (arrow in A and lower white arrow in B). The coronal image demonstrates adherence to a central venous catheter (upper white arrow in B).

calcifications.94 However, thrombus may be difficult to distinguish from tumor on CT. The imaging characteristics of thrombus on MRI are more specific, taking into account that the imaging features differ throughout the evolution of the thrombus. Acute thrombus demonstrates high signal intensity on T1- and T2-weighted images, subacute thrombus shows high signal intensity on T1 and low signal intensity on T2, whereas chronic thrombus has low signal intensity on T1- and T2-weighted images.187 Using long inversion recovery times, thrombus generally demonstrates very low signal intensity compared with tumor, in which signal intensity is present.84 There is typically no contrast uptake in thrombus regardless of its stage.187

Lipomatous Hypertrophy of the Atrial Septum.  Lipomatous hypertrophy of the atrial septum (LHAS) occurs in up to 8% of the population and consists of unencapsulated fat with diameter greater than 2 cm that infiltrates into the atrial septum.77 It should be distinguished from cardiac lipoma, which is an encapsulated tumor. Risk factors include female gender, increasing age, and obesity. Most patients are asymptomatic, although cardiac arrhythmias may occur in patients with larger lesions.77 Cardiac CT demonstrates a fatty attenuation mass in the interatrial septum that has a dumbbell configuration, with the larger component generally located posterolaterally. The constricted area of the dumbbell corresponds to fatty sparing in the region of the fossa ovalis.94 MRI of LHAS is similar to that of CT. The lesion demonstrates the signal intensity characteristics of fat, with loss of signal intensity on fatsuppression sequences (Fig. 58-64).77

PERICARDIAL DISEASE Anatomy The pericardium has two distinct layers. The inner layer is a serous layer. It consists of two components, a visceral portion (or epicardium)

that is tightly adherent to the heart and proximal great vessels and a thicker parietal component that is firmly attached to the second outer fibrous layer. The fibrous layer attaches to the great vessels, manubrium, xiphoid process, central tendon of the diaphragm, and loosely to the descending aorta. The serous pericardium reflects superiorly approximately 3 cm from the origin of the great vessels. The pericardial space lies between the visceral and parietal serous layer. The configuration of the pericardial reflections is such that a portion of the left atrium is variably covered by pericardium. The pericardial space normally contains 15 to 35 mL of pericardial fluid.116 Sinuses are formed by reflections of the pericardium.32,69 The two major sinuses are the transverse and oblique sinuses. The transverse sinus lies between the arterial mesocardia, which surrounds the aorta and pulmonary artery, and the venous mesocardia, which encases the venae cavae and pulmonary veins. The transverse sinus lies superior to the left atrium and posterior to the aorta and main pulmonary artery, and it communicates with several recesses, including the superior and inferior aortic recesses and the bilateral pulmonary recesses. The superior pericardial recess is a component of the transverse sinus that extends between the ascending aorta and main pulmonary artery. It typically has a triangular shape but may be quite variable in appearance and extent. The bilateral pulmonary recesses lie inferior to their respective pulmonary arteries. The oblique sinus is located behind the left atrium and between the pulmonary veins and does not communicate with the transverse sinus.32,69 Its recesses include the left pulmonic, postcaval, and pulmonary vein recesses. On CT, portions of the pericardium are visible between the mediastinal and subepicardial fat. The normal measurement of the pericardium is less than 4 mm, and it typically measures 1 to 2 mm in thickness.24 The pericardium is best visualized anteriorly and inferiorly, but there may be areas where it is present but not identified (Fig. 58-65A and B). Focal thickening of the pericardium may be seen

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

2053

RA RA

LA

LA

A

B FIG 58-64  Lipomatous hypertrophy of the atrial septum. A, T1-weighted four-chamber MRI demonstrates fat signal intensity in the interatrial septum. The larger lateral component (large arrow) is separated from the smaller medial component (small arrow) by a low-signal-intensity area corresponding to the fossa ovalis. B, Fat-saturation sequence reveals signal dropout within the mass (arrow). LA, left atrium; RA, right atrium.

RV RA

A

B

AA

C

FIG 58-65  Normal pericardium. A and B, The thin membranes of the normal pericardium are visible anterior to the right heart chambers (arrows). C, The superior pericardial recess of the transverse sinus is demonstrated as a fluid-density ovoid structure (arrow) posterior to the ascending aorta (AA). RA, right atrium; RV, right ventricle.

anterior to the RV near the insertion into the diaphragm and should not be mistaken for a pathologic process. Recesses, particularly the superior pericardial recess, may be mistaken for a pathologically enlarged lymph node if their typical location and low-density composition are not recognized (see Fig. 58-65C). Fluid in the inferior pulmonary vein recess, particularly on the right, may mimic a medially located pulmonary nodule. On MRI, T1- and T2-weighted dark-blood imaging shows the pericardium as very low signal intensity between higher areas of mediastinal and subepicardial fat, similar in thickness to that on CT. Tissue characterization is possible in some instances, owing to alterations of normal signal intensity on dark-blood imaging. SSFP bright-blood sequences may show increased thickness due to chemical shift

artifacts.18 Spatial modulation of magnetization (tagging) sequences may be useful to assess for the presence of pericardial constriction. Cine loop imaging of SSFP sequences, which can be acquired during breath holding or free breathing, permits functional assessment of the pericardium throughout the cardiac cycle.152

Pericardial Cyst Pericardial cysts are congenital lesions that are lined by mesothelial cells. If a connection to the pericardium is present, they are termed pericardial diverticula. Pericardial cysts are uncommon lesions, occurring in 1 in 100,000 patients.87 Approximately 70% are found in the right cardiophrenic angle, followed in frequency by the left cardiophrenic angle. A minority occur in other parts of the mediastinum,

2054

PART II  CT and MR Imaging of the Whole Body

A

B FIG 58-66  Pericardial cyst. A, T1-weighted four-chamber image of the heart demonstrates an ovoid lowsignal-intensity lesion in the right cardiophrenic angle (arrow). B, On T2-weighted imaging the lesion shows high signal intensity (arrow) consistent with a pericardial cyst.

usually related to the pericardium.49 Most patients are asymptomatic, but up to one third may present with chest pain, dyspnea, or cough, particularly if the cyst compresses adjacent structures. Imaging features of pericardial cysts are often characteristic. On chest radiography the cyst typically manifests as a rounded density that contacts the hemidiaphragm and anterior chest wall.49 On CT a pericardial cyst typically appears as a thin-walled round or ovoid nonseptated lesion with fluid-density contents. If hemorrhage has occurred, the internal density may be higher.200 MRI demonstrates variable signal intensity on T1-weighted imaging depending on protein content. On T2-weighted images, pericardial cysts typically show high signal intensity200 (Fig. 58-66). Pericardial diverticula often cannot be distinguished from pericardial cysts but can be suspected if there is a direct connection of a cystlike structure to fluid in the pericardial space. Most pericardial cysts and diverticula do not undergo treatment, but symptomatic cysts may require aspiration or surgery.6,29,183

Pericardial Defect Congenital absence of the pericardium is an uncommon abnormality that may be partial or complete. It is estimated to occur in 1 in 7000 to 1 in 13,000 and is much more frequent on the left.144 One etiologic theory holds that the embryonic left common cardiac vein regresses earlier than normal, leading to a defect.48 A defect in the pleura usually accompanies the pericardial defect. Although most pericardial defects are asymptomatic, patients with partial defects—particularly on the left—may be symptomatic. Herniation through the partial defect may be associated with infarction of the left atrial appendage, causing syncope and sudden death.1 Although pericardial defects are usually isolated, a coexistent atrial septal defect, patent ductus arteriosus, bicuspid aortic valve, or pulmonary abnormalities may be present.179 Cross-sectional imaging of congenital pericardial defects relies on indirect findings, because the normal pericardium may not be visible. With left-sided lesions, there is typically leftward rotation of the heart into the left chest, with interposition of the lung between the aorta and

pulmonary artery. The appearance of congenital absence of the pericardium depends on whether the abnormality is partial or complete. If the defect is partial, the heart bulges leftward in the region of the left atrial appendage. If the left-sided defect is complete, the entire heart is typically displaced to the left and lung is interposed between the heart’s inferior margin and the diaphragm (Fig. 58-67). On functional (cine) CT or MRI images, excessive motion of the cardiac apex may be observed in individuals with complete absence of the pericardium.157

Pericardial Effusion and Pericarditis There are numerous etiologies of pericardial effusion, which can be defined as an abnormal accumulation of pericardial fluid. Pericardial effusion is frequent in congestive heart failure, renal insufficiency, and autoimmune diseases. Effusion can also be caused by infection, metastatic disease, and radiation therapy. Pericardial fluid collections can also occur after trauma or cardiothoracic surgery. Pericardial effusion can be an acute, chronic, or recurrent manifestation of disease. A pericardial collection may be characterized as transudative, exudative, hemopericardium, and pyomyocardium.8 Echocardiography is generally used as the initial technique to assess pericardial effusion but may be limited by poor acoustic windows or the location and characteristics of the pericardial collection. Both CT and MRI are valuable to quantify the pericardial effusion and may in some cases provide a degree of tissue characterization (Fig. 58-68). As a general rule, pericardial thickness of 4 mm or greater is considered abnormal on cross-sectional imaging. On CT an increase in density above that of simple fluid may indicate an exudative or hemorrhagic process, although there is overlap.149 Patients with inflammatory or malignant pericardial effusions may demonstrate enhancement after IV contrast. In patients with a history of trauma, hemopericardium may be evident. MRI is also valuable to demonstrate the characteristics of pericardial effusion. Whereas both transudative and exudative effusions  show high signal intensity on T2-weighted imaging, transudative effusions typically display low signal intensity on T1-weighted images, and

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

AA

2055

PA

A

B FIG 58-67  Congenital pericardial defect. A, Axial CT scan using lung windows demonstrates air (arrow) insinuating between the ascending aorta (AA) and main pulmonary artery (PA). B, Coronal CT scan using lung windows shows leftward shift of the heart, with air extending between the left heart border and diaphragm (arrows). The findings are consistent with complete absence of the left hemipericardium.

A

B FIG 58-68  Pericardial effusion. A, Axial enhanced CT scan demonstrates a large pericardial effusion (arrows). The low-density character of the effusion suggests that it is a transudate. B, Coronal enhanced CT scan shows the craniocaudal extent of the pericardial effusion (arrows).

exudative effusions are more likely to demonstrate higher signal intensity on T1-weighted images.19,55 In patients with myopericarditis, LGE sequences may show increased signal in both the myocardium and pericardium. Pericardial tamponade may complicate pericardial effusion and is a physiologic state in which the pericardial pressures exceed that of the RV. Tamponade can lead to hemodynamic compromise and rapid death. The quantity of pericardial effusion required to produce pericardial tamponade depends on the rapidity of accumulation and varies from 150 mL to 3 L.162,181 The diagnosis is usually made based on clinical criteria in combination with characteristic findings on echocardiography, notably RV free wall collapse.

CT and MRI play a secondary role in the diagnosis of pericardial tamponade. Typical findings include a large pericardial collection, compression with flattening or inversion of the right atrial and/or ventricular free wall, inversion of the septum, and right atrial enlargement with reflux of contrast into the inferior vena cava and azygos system.161 Dynamic imaging may reveal evidence of ventricular interdependence, a situation where elevated pericardial pressures negate the influence of respiration on the pressures within the cardiac chambers, leaving them dependent on transseptal pressures. This is manifested by a rocking motion or “bounce” of the interventricular septum, where the septum moves first toward the LV and later toward the RV during the cardiac cycle.54

2056

PART II  CT and MR Imaging of the Whole Body

Constrictive Pericarditis Constrictive pericarditis is a condition in which there is reduced compliance of the pericardial leading to elevated filling pressures. This entity is occasionally acute or subacute but typically is a sequela of long-standing pericardial disease, including infection (particularly purulent or tuberculous infection), uremia, connective tissue disease, metastatic disease, and surgery or radiation affecting the pericardium. In many instances the cause of pericardial constriction is unknown.146 The pathophysiology consists of equalization of pressures in all cardiac chambers associated with ventricular coupling, as described for cardiac tamponade. Constrictive pericarditis typically manifests with symptoms of low cardiac output, which particularly affects the right side of the heart.142 Echocardiography is often obtained initially and can demonstrate equalization of pressures during diastole but may not be able to distinguish constrictive pericarditis from restrictive cardiomyopathy. Such distinction is important because therapy differs, with pericardiectomy recommended for constrictive pericarditis. CT and MRI are critical techniques in the diagnosis of constrictive pericarditis. Thickening of the pericardium is well visualized with both techniques and may predominate anterior to the right heart. On CT, pericardial calcification is evident in up to 27% of patients with constrictive pericarditis.114,150 Conventionally a threshold pericardial thickness value of 4 mm has been used, but the limitations of such a sharp demarcation are evident. Up to 20% of patients with constrictive pericarditis demonstrate normal pericardial thickness.192 Moreover, patients with end-stage constrictive pericarditis are more likely than those with reversible constrictive pericarditis to have a normal diameter of the pericardium.50,211 Other morphologic features may be visible on CT or MRI. Poor compliance of the pericardium causes changes in the shape of the cardiac chambers. The ventricles may show a tubular appearance, and there may be enlargement of one or both atria (Fig. 58-69). Flattening of the RV free wall and evidence of elevated right-sided pressures are

also frequent. The venae cavae and azygous vein are often dilated, and ascites, pleural effusions, and peripheral edema are characteristic. On dynamic imaging by CT or MRI, equalization of pressures is manifested by signs of ventricular coupling, including a septal bounce. On MRI, free-breathing sequences may show increase in septal bounce during inspiration due to increased venous return to the RV. An MRI tagging sequence, which involves placing a grid of saturation lines over the pericardium and myocardium, can demonstrate adhesion between the two structures. Normally, grid lines from the myocardium and pericardium slip past each other during the cardiac cycle. Lack of such slippage indicates constrictive pericarditis.103 In pericardial constriction, the high atrial pressures lead to rapid early filling of the ventricles, followed by rapid decrease in the rate of filling during later diastole. Phase-contrast MRI, with its ability to measure flow, shows exaggeration of the early wave of diastolic filling (E wave) and a diminished or absent late diastolic wave (A wave), findings that may also be present in restrictive cardiomyopathy.

Pericardial Tumors and Tumorlike Conditions Focal pericardial lesions can be categorized as pseudomasses or primary and secondary tumors.19 Pseudomasses are not uncommon on echocardiography. CT and MRI often clarify the nature of a pseudomass and in some instances demonstrate its origin outside the pericardium. They may also permit characterization of a pericardial pseudomass within the pericardium as a focal collection of fluid or blood.19 Pericardial fat necrosis is another tumorlike condition of unknown etiology that may manifest with sudden chest pain. CT demonstrates a fatty lesion near the anterior pericardium, with adjacent infiltration of the fat.76 Primary pericardial tumors are quite rare and are more likely to be malignant. Examples of malignant tumors include primary pericardial mesothelioma, paraganglioma, lymphoma, angiosarcoma, and liposarcoma. Of these, pericardial mesothelioma is most common, accounting for approximately half of primary pericardial tumors, although

RA

LA

A

B FIG 58-69  Constrictive pericarditis. A, T1-weighted four-chamber image of the heart demonstrates marked thickening of the pericardium (arrows). The ventricles have assumed a spade-shaped appearance, with dilatation of the atrium. B, T1-weighted short-axis image reveals circumferential pericardial thickening at the ventricular level. RA, right atrium; LA, left atrium.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium

FIG 58-70  Lymphoma invading pericardium. Enhanced axial CT image demonstrates an anterior mediastinal mass showing pericardial invasion (arrow). At pathology a diagnosis of lymphoma was made.

pericardial mesotheliomas comprise less than 5% of all mesotheliomas. Rare benign pericardial tumors are teratoma, fibroma, and hemangioma.67 Secondary spread of tumors to the pericardium can occur by direct extension from adjacent structures or the veins, or by more distant spread via hematogenous or lymphatic extension.19 Common causes of direct extension include lung cancer, lymphoma, and renal cell carcinoma, whereas breast cancer and melanoma may affect the pericardium by distant metastasis. Metastatic disease to the pericardium is more commonly appreciated at autopsy than during life. Symptoms of pericardial involvement by tumor are related to the extent of disease. Nonspecific symptoms (chest pain, shortness of breath) are frequent.19 Diffuse involvement may lead to symptoms of pericardial constriction. CT and MRI may show the primary site of disease as well as the extent of pericardial involvement (Fig. 58-70). Typically, pericardial thickening is present, with a variable amount of pericardial nodularity. There may be an admixture of pericardial fluid and hemorrhage. Enhancement of portions of the pericardium may be present after contrast administration using either CT or MRI.

REFERENCES 1. Abbas AE, Appleton CP, Liu PT, et al: Congenital absence of the pericardium: Case presentation and review of literature. Int J Cardiol 98(1):21–25, 2005. 2. Abdel-Aty H, Zagrosek A, Schulz-Menger J, et al: Delayed enhancement and T2-weighted cardiovascular magnetic resonance imaging differentiate acute from chronic myocardial infarction. Circulation 109(20):2411–2416, 2004. 3. Abedin Z, Goldberg J: Origin and length of left main coronary artery: Its relation to height, weight, sex, age, pattern of coronary distribution, and presence or absence of coronary artery disease. Cathet Cardiovasc Diagn 4(3):335–340, 1978. 4. Achenbach S, Giesler T, Ropers D, et al: Detection of coronary artery stenoses by contrast-enhanced, retrospectively electrocardiographically gated, multislice spiral computed tomography. Circulation 103:2535– 2538, 2001. 5. Agatston AS, Janowitz WR, Hildner FJ, et al: Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 15(4):827–832, 1990.

2057

6. Akiba T, Marushima H, Masubuchi M, et al: Small symptomatic pericardial diverticula treated by video-assisted thoracic surgical resection. Ann Thorac Cardiovasc Surg 15(2):123–125, 2009. 7. Alfakih K, Reid S, Jones T, et al: Assessment of ventricular function   and mass by cardiac magnetic resonance imaging. Eur Radiol 14(10): 1813–1822, 2004. 8. Alter P, Figiel JH, Rupp TP, et al: MR, CT, and PET imaging in pericardial disease. Heart Fail Rev 18:289–306, 2013. 9. Anders K, Baum U, Schmid M, et al: Coronary artery bypass graft (CABG) patency: Assessment with high-resolution submillimeter 16-slice multidetector-row computed tomography (MDCT) versus coronary angiography. Eur J Radiol 57(3):336–344, 2006. 10. Angelini P: Coronary artery anomalies: An entity in search of an identity. Circulation 115(10):1296–1305, 2007. 11. Arad Y, Spadaro LA, Roth M, et al: Treatment of asymptomatic adults with elevated coronary calcium scores with atorvastatin, vitamin C, and vitamin E: The St. Francis Heart Study randomized clinical trial. J Am Coll Cardiol 46:166–172, 2005. 12. Araoz PA, Mulvagh SL, Tazelaar HD, et al: CT and MR imaging of benign primary cardiac neoplasms with echocardiographic correlation. Radiographics 20:1303–1319, 2000. 13. Bashore TM: Adult congenital heart disease: Right ventricular outflow tract lesions. Circulation 115(14):1933–1947, 2007. 14. Bazzocchi G, Romagnoli A, Sperandio M, et al: Evaluation with 64-slice CT of the prevalence of coronary artery variants and congenital anomalies: A retrospective study of 3,236 patients. Radiol Med 116(5):675–689, 2011. 15. Becker CR, Kleffel T, Crispin A, et al: Coronary artery calcium measurement: Agreement of multirow detector and electron beam CT. AJR Am J Roentgenol 176:1295–1298, 2001. 16. Beghetti M, Gow RM, Haney I, et al: Pediatric primary benign cardiac tumors: A 15-year review. Am Heart J 134:1107–1114, 1997. 17. Bennett CJ, Maleszewski JJ, Araoz PA: CT and MR imaging of the aortic valve: Radiologic-pathologic correlation. Radiographics 32(5):1399–1420, 2012. 18. Bogaert J, Francone M: Cardiovascular magnetic resonance in pericardial disease. J Cardiovasc Magn Reson 11:14, 2009. 19. Bogaert J, Francone M: Pericardial disease: Value of CT and MR imaging. Radiology 267(2):340–356, 2013. 20. Bourassa MG, Butnaru A, Lesperance J, et al: Symptomatic myocardial bridges: Overview of ischemic mechanisms and current diagnostic and treatment strategies. J Am Coll Cardiol 41(3):351–359, 2003. 21. Buckley O, Madan R, Kwong R, et al: Cardiac masses, part 1: Imaging strategies and technical considerations. AJR Am J Roentgenol 197(5):W837–W841, 2011. 22. Budoff MJ, Dowe D, Jollis JG, et al: Diagnostic performance of 64-multidetector row coronary computed tomographic angiography for evaluation of coronary artery stenosis in individuals without known coronary artery disease: Results from the prospective multicenter ACCURACY (Assessment by Coronary Computed Tomographic Angiography of Individuals Undergoing Invasive Coronary Angiography) trial. J Am Coll Cardiol 52(21):1724–1732, 2008. 23. Budoff MJ, Hokanson JE, Nasir K, et al: Progression of coronary artery calcium predicts all-cause mortality. JACC Cardiovascular Imag 3:1229–1236, 2010. 24. Bull RK, Edwards PD, Dixon AK: CT dimensions of the normal pericardium. British Journal of Radiology 71:923–925, 1998. 25. Burke A, Vermani R: Tumors of the heart and great vessels. Atlas of Tumor Pathology. 3rd Series, Fascicle 16, Washington, DC, 1996, Armed Forces Institute of Pathology. 26. Cademartiri F, La Grutta L, Malago R, et al: Prevalence of anatomical variants and coronary anomalies in 543 consecutive patients studied with 64-slice CT coronary angiography. Eur Radiol 18(4):781–791, 2008. 27. Cademartiri F, Mollet NR, Runza G, et al: Influence of intracoronary attenuation on coronary plaque measurements using multislice computed tomography: Observations in an ex vivo model of coronary computed tomography angiography. Eur Radiol 15(7):1426–1431, 2005.

2058

PART II  CT and MR Imaging of the Whole Body

28. Callister TQ, Cooil B, Raya SP, et al: Coronary artery disease: Improved reproducibility of calcium scoring with an electron-beam CT volumetric method. Radiology 208(3):807–814, 1998. 29. Caretta A, Negri G, Pansera M, et al: Thoracoscopic treatment of a pericardial diverticulum. Surg Endosc 17:158, 2003. 30. Chen YF, Chien TM, Chen CW, et al: Double right coronary artery or split right coronary artery? Int J Cardiol 154(3):243–245, 2012. 31. Chen JJ, Manning MA, Frazier AA, et al: CT angiography of the cardiac valves: Normal, diseased, and postoperative appearances. Radiographics 29(5):1393–1412, 2009. 32. Choe YH, Im JG, Park JH, et al: The anatomy of the pericardial space: A study in cadavers and patients. AJR Am J Roentgenol 149:693–697, 1987. 33. Chow BJ, Ahmed O, Small G, et al: Prognostic value of CT angiography in coronary bypass patients. JACC Cardiovasc Imaging 4(5):496–502, 2011. 34. Chun EJ, Choi SI, Jin KN, et al: Hypertrophic cardiomyopathy: Assessment with MR imaging and multidetector CT. Radiographics 30:1309–1328, 2010. 35. Coelho-Filho OR, Rickers C, Kwong RY, et al: MR myocardial perfusion imaging. Radiology 266(3):701–715, 2013. 36. Cowles RA, Berdon WE: Bland-White-Garland syndrome of anomalous left coronary artery arising from the pulmonary artery (ALCAPA): A historical review. Pediatr Radiol 37(9):890–895, 2007. 37. de Agustin JA, Marcos-Alberca P, Hernandez-Antolin R, et al: Collateral circulation from the conus coronary artery to the anterior descending coronary artery: Assessment using multislice coronary computed tomography. Rev Esp Cardiol 63(3):347–351, 2010. 38. Decramer I, Vanhoenacker PK, Sarno G, et al: Effects of sublingual nitroglycerin on coronary lumen diameter and number of visualized septal branches on 64-MDCT angiography. AJR Am J Roentgenol 190(1):219–225, 2008. 39. de Graaf FR, van Velzen JE, Witkowska AJ, et al: Diagnostic performance of 320-slice multidetector computed tomography coronary angiography in patients after coronary artery bypass grafting. Eur Radiol 21(11):2285–2296, 2011. 40. Dhawan SS, Avati Nanjundappa RP, Branch JR, et al: Shear stress and plaque development. Expert Rev Cardiovasc Ther 8(4):545–556, 2010. 41. Díaz-Zamudio M, Bacilio-Pérez U, Herrera-Zarza MC, et al: Coronary artery aneurysms and ectasia: Role of coronary CT angiography. Radiographics 29(7):1939–1954, 2009. 42. Dickfeld T, Calkins H, Zviman M, et al: Anatomic stereotactic catheter ablation on three-dimensional magnetic resonance images in real time. Circulation 108:2407–2413, 2003. 43. Dodge JT, Jr, Brown BG, Bolson EL, et al: Lumen diameter of normal human coronary arteries. Influence of age, sex, anatomic variation, and left ventricular hypertrophy or dilation. Circulation 86(1):232–246, 1992. 44. Donnino R, Jacobs JE, Doshi JV, et al: Dual-source versus single-source cardiac CT angiography: Comparison of diagnostic image quality. AJR Am J Roentgenol 192(4):1051–1056, 2009. 45. Eckart RE, Scoville SL, Campbell CL, et al: Sudden death in young adults: A 25-year review of autopsies in military recruits. Ann Intern Med 141(11):829–834, 2004. 46. Elliott P, Andersson B, Arbustini E, et al: Classification of the cardiomyopathies: A position statement from the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J 29:270–276, 2008. 47. Erol C, Seker M: The prevalence of coronary artery variations on coronary computed tomography angiography. Acta Radiol 53(3):278– 284, 2012. 48. Faridah Y, Julsrud PR: Congenital absence of pericardium revisited. Int J Cardiovasc Imag 18:67–73, 2002. 49. Feigin DS, Fenoglio JJ, McAllister HA, et al: Pericardial cysts. A radiologic-pathologic correlation and review. Radiology 125(1):15–20, 1977. 50. Feng DL, Glockner J, Kim K, et al: Cardiac magnetic resonance imaging pericardial late gadolinium enhancement and elevated inflammatory

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61. 62. 63.

64.

65.

66.

67.

68.

markers can predict the reversibility of constrictive pericarditis after anti-inflammatory medical therapy: A pilot study. Circulation 124(17):1830–1837, 2011. Ferencik M, Ropers D, Abbara S, et al: Diagnostic accuracy of image postprocessing methods for the detection of coronary artery stenoses by using multidetector CT. Radiology 243:696–702, 2007. Fernández-Pérez GC, Aguilar-Arjona JA, de la Fuente GT, et al: Takotsubo cardiomyopathy: Assessment with cardiac MRI. AJR Am J Roentgenol 195(2):W139–W145, 2010. Fischman DL, Leon MB, Baim DS, et al: Stent Restenosis Study Investigators: A randomized comparison of coronary-stent placement and balloon angioplasty in the treatment of coronary artery disease.   N Engl J Med 331:496–501, 1994. Francone M, Dymarkowski S, Kalantzi M, et al: Assessment of ventricular coupling with realtime cine MRI and its value to differentiate constrictive pericarditis from restrictive cardiomyopathy. Eur Radiol 16:944–951, 2006. Frank H, Globits S: Magnetic resonance imaging evaluation of myocardial and pericardial disease. J Magn Reson Imaging 10:617–626, 1999. Frazier AA, Qureshi F, Read KM, et al: Coronary artery bypass grafts: Assessment with multidetector CT in the early and late postoperative settings. Radiographics 25(4):881–896, 2005. Garcia MJ, Lessick J, Hoffmann MH: CATSCAN Study Investigators. Accuracy of 16-row multidetector computed tomography for the assessment of coronary artery stenosis. JAMA 296(4):403–411, 2006. Gatewood RP, Jr, Nanda NC: Differentiation of left ventricular pseudoaneurysm from true aneurysm with two dimensional echocardiography. Am J Cardiol 46:869–876, 1980. Giesbrandt KJ, Bolan CW, Shapiro BP, et al: Diffuse diseases of the myocardium: MRI-pathologic review of nondilated cardiomyopathies. AJR Am J Roentgenol 200(3):W266–W273, 2013. Ginat DT, Fong MW, Tuttle DJ, et al: Cardiac imaging: Part 1. MR pulse sequences, imaging planes, and basic anatomy. Am J Roentgenol 197:808–815, 2011. Godwin JD, Chen JT: Thoracic venous anatomy. AJR Am J Roentgenol 147:674–684, 1986. Goldberg AD, Blankstein R, Padera RF: Tumors metastatic to the heart. Circulation 128:1790–1794, 2013. Goldman S, Zadina K, Moritz T, et al: Long-term patency of saphenous vein and left internal mammary artery grafts after coronary artery bypass surgery: Results from a Department of Veterans Affairs Cooperative Study. J Am Coll Cardiol 44(11):2149–2156, 2004. Goldstein JA, Chinnaiyan KM, Abidov A, et al: The CT-STAT (Coronary Computed Tomographic Angiography for Systematic Triage of Acute Chest Pain Patients to Treatment) trial. J Am Coll Cardiol 58(14):1414–1422, 2011. Gosling O, Loader R, Venables P, et al: A comparison of radiation doses between state-of-the-art multislice CT coronary angiography with iterative reconstruction, multislice CT coronary angiography with standard filtered back-projection and invasive diagnostic coronary angiography. Heart 96:922–926, 2010. Gowda RM, Khan IA, Nair CK, et al: Cardiac papillary fibroelastoma:   A comprehensive analysis of 725 cases. Am Heart J 146:404–410, 2003. Grebenc ML, Rosado-de-Christenson ML, Green CE, et al: Cardiac myxoma: Imaging features in 83 patients. Radiographics 22:673–689, 2002. Greenland P, Bonow RO, Brundage BH, et al: ACCF/AHA 2007 clinical expert consensus document on coronary artery calcium scoring by computed tomography in global cardiovascular risk assessment and in evaluation of patients with chest pain: A report of the American College of Cardiology Foundation Clinical Expert Consensus Task Force (ACCF/AHA Writing Committee to Update the 2000 Expert Consensus Document on Electron Beam Computed Tomography) developed in collaboration with the Society of Atherosclerosis Imaging and Prevention and the Society of Cardiovascular Computed Tomography. J Am Coll Cardiol 49:378–402, 2007.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium 69. Groell R, Schaffler GJ, Reinmueller R: Pericardial sinuses and recesses: Findings at electrocardiographically triggered electron-beam CT. Radiology 212:69–73, 1999. 70. Guo YK, Yang ZG, Ning G, et al: Isolated mitral regurgitation: quantitative assessment with 64-section multidetector CT— Comparison with MR imaging and echocardiography. Radiology 252(2):369–376, 2009. 71. Gupta S, Plein S, Greenewood JP: The coumadin ridge: An important example of a left atrial pseudotumor demonstrated by cardiovascular magnetic resonance imaging. J Radiol Case Rep 3:1–5, 2009. 72. Guthaner DF, Wexler L, Harell G: CT demonstration of cardiac structures. AJR Am J Roentgenol 133(1):75–81, 1979. 73. Halpern EJ: Triple-rule-out CT angiography for evaluation of acute chest pain and possible acute coronary syndrome. Radiology 252(2):332–345, 2009. 74. Halpern EJ: Clinical cardiac CT anatomy and function. In Halpern AJ, Owen AN, editors: Cardiac Morphology and Function, ed 2, New York, 2011, Thieme, pp 174–229. 75. Hecht HS, Budoff MJ, Berman DS, et al: Coronary artery calcium scanning: Clinical paradigms for cardiac risk assessment and treatment. Am Heart J 151:1139–1146, 2006. 76. Hernandez D, Galimany J, Pernas JC, et al: Case 170: Pericardial fat necrosis. Radiology 259(3):919–922, 2011. 77. Heyer CM, Kagel T, Lemburg SP, et al: Lipomatous hypertrophy of the interatrial septum: A prospective study of incidence, imaging findings, and clinical symptoms. Chest 124:2068–2073, 2003. 78. Hobbs RE, Millit HD, Raghavan PV, et al: Coronary artery fistulae: A 10-year review. Cleve Clin Q 49(4):191–197, 1982. 79. Hoffmann U, Ferencik M, Cury RC, et al: angiography. J Nucl Med 47(5):797–806, 2006. 80. Hoffmann U, Truong QA, Schoenfeld DA, et al: Coronary CT angiography versus standard evaluation in acute chest pain. N Engl J Med 367(4):299–308, 2012. 81. Holmes DR, Jr, Leon MB, Moses JW, et al: Analysis of 1-year clinical outcomes in the SIRIUS trial: a randomized trial of a sirolimus-eluting stent versus a standard stent in patients at high risk for coronary restenosis. Circulation 109:634–640, 2004. 82. Hong C, Becker CR, Schoepf UJ, et al: Coronary artery calcium: Absolute quantification in nonenhanced and contrast-enhanced multi-detector row CT studies. Radiology 223(2):474–480, 2002. 83. Huang CW, Tsai YF, Hsiao CY: Different MRI signs in predicting   the treatment efficacy of epidural blood patch in spontaneous intracranial hypotension: A case report. Iran J Radiol 10(3):172–174, 2013. 84. Hundley WG, Bluemke DA, Finn JP, et al: ACCF/ACR/AHA/NASCI/ SCMR 2010 expert consensus document on cardiovascular magnetic resonance: A report of the American College of Cardiology Foundation Task Force on Expert Consensus Documents. Circulation 121(22):2462– 2508, 2010. 85. Hur J, Kim YH, Lee HJ, et al: Dual-enhanced cardiac CT for detection of left atrial appendage thrombus in patients with stroke: A prospective comparison study with transesophageal echocardiography. Stroke 42:2471–2477, 2011. 86. Hwang JH, Ko SM, Roh HG, et al: Myocardial bridging of the left anterior descending coronary artery: Depiction rate and morphologic features by dual-source CT coronary angiography. Korean J Radiol 11(5):514–521, 2010. 87. Hynes JK, Tajik AJ, Osborne MJ, et al: Two-dimensional echocardiographic diagnosis of pericardial cyst. Mayo Clin Proc 58:60–63, 1983. 88. Ilia R, Rosenshtein G, Weinstein J, et al: Left anterior descending artery length in left and right coronary artery dominance. Coron Artery Dis 12(1):77–78, 2001. 89. Izgi C, Cevik C, Basbayraktar F: Caseous calcification and liquefaction of the mitral annulus: A diagnostic confounder. Int J Cardiovasc Imaging 22(3–4):543–545, 2006. 90. Janigan DT, Husain A, Robinson NA: Cardiac angiosarcomas: A review and a case report. Cancer 57:852–859, 1986.

2059

91. Jeudy J, Kirsch J, Tavora F, et al: From the radiologic pathology archives: Cardiac lymphoma: Radiologic-pathologic correlation. Radiographics 32(5):1369–1380, 2012. 92. Johnson KM, Johnson HE, Dowe DA: Left ventricular apical thinning as normal anatomy. J Comput Assist Tomogr 33:334–337, 2009. 93. Joshi SD, Joshi SS, Athavale SA: Origins of the coronary arteries and their significance. Clinics (Sao Paulo) 65(1):79–84, 2010. 94. Kassop D, Donovan MS, Cheezum MK, et al: Cardiac masses on cardiac CT: A review. Curr Cardiovasc Imaging Rep 7:9281, 2014. 95. Kato S, Kitagawa K, Ishida N, et al: Assessment of coronary artery disease using magnetic resonance coronary angiography: A national multicenter trial. J Am Coll Cardiol 56(12):983–991, 2010. 96. Khan M, Cummings KW, Gutierrez FR, et al: Contraindications and side effects of commonly used medications in coronary CT angiography. Int J Cardiovasc Imaging 27(3):441–449, 2011. 97. Kim E, Choe YH, Han BK, et al: Right ventricular fat infiltration in asymptomatic subjects: Observations from ECG-gated 16 slice multidetector CT. J Comput Assist Tomogr 31:22–28, 2007. 98. Kim PJ, Hur G, Kim SY, et al: Frequency of myocardial bridges and dynamic compression of epicardial coronary arteries: A comparison between computed tomography and invasive coronary angiography. Circulation 119(10):1408–1416, 2009. 99. Kim JS, Judson MA, Donnino R, et al: Cardiac sarcoidosis. Am Heart J 157:9–21, 2009. 100. Kim SY, Seo JB, Do KH, et al: Coronary artery anomalies: Classification and ECG-gated multi-detector row CT findings with angiographic correlation. Radiographics 26(2):317–333, discussion 333–314, 2006. 101. Kim RJ, Wu E, Rafael A, et al: The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 343(20):1445–1453, 2000. 102. Knaapen M, Koch AH, Koch C, et al: Prevalence of left and balanced coronary arterial dominance decreases with increasing age of patients at autopsy. A postmortem coronary angiograms study. Cardiovasc Pathol 22(1):49–53, 2013. 103. Kojima S, Yamada N, Goto Y: Diagnosis of constrictive pericarditis by tagged cine magnetic resonance imaging. N Eng J Med 341:373, 1999. 104. Konen E, Merchant N, Gutierrez C, et al: True versus false left ventricular aneurysm: Differentiation with MR imaging initial experience. Radiology 236:65–75, 2005. 105. Kosar P, Ergun E, Ozturk C, et al: Anatomic variations and anomalies of the coronary arteries: 64-slice CT angiographic appearance. Diagn Interv Radiol 15(4):275–283, 2009. 106. Krasuski RA, Magyar D, Hart S, et al: Long-term outcome and impact of surgery on adults with coronary arteries originating from the opposite coronary cusp. Circulation 123(2):154–162, 2011. 107. Kröpil P, Rojas CA, Ghoshhajra B, et al: Prospectively ECG-triggered high-pitch spiral acquisition for cardiac CT angiography in routine clinical practice: initial results. J Thorac Imaging 27(3):194–201, 2012. 108. Kumbhani DJ, Ingelmo CP, Schoenhagen P, et al: Meta-analysis of diagnostic efficacy of 64-slice computed tomography in the evaluation of coronary in-stent restenosis. Am J Cardiol 103(12):1675–1681, 2009. 109. Kwong Y, Troupis J: Cardiac CT imaging in the context of left atrial appendage occlusion. J Cardiovasc Comput Tomogr 9(1):13–18, 2015. 110. Lacomis JM, Wigginton W, Fuhrman C, et al: Multi-detector row CT of the left atrium and pulmonary veins before radio-frequency catheter ablation for atrial fibrillation. Radiographics 23(Spec No):S35–S48, 2003. 111. Lee BY: Anomalous right coronary artery from the left coronary sinus with an interarterial course: is it really dangerous? Korean Circ J 39(5):175–179, 2009. 112. Lee TY, Chhem RK: Impact of new technologies on dose reduction in CT. Eur J Radiol 76:28–35, 2010. 113. Lee HJ, Hong YJ, Kim HY, et al: Anomalous origin of the right coronary artery from the left coronary sinus with an interarterial course: Subtypes and clinical importance. Radiology 262(1):101–108, 2012. 114. Ling LH, Oh JK, Breen JF, et al: Calcific constrictive pericarditis: Is it still with us? Ann Intern Med 132(6):444–450, 2000.

2060

PART II  CT and MR Imaging of the Whole Body

115. Litt HI, Gatsonis C, Snyder B, et al: CT angiography for safe discharge of patients with possible acute coronary syndromes. N Engl J Med 366(15):1393–1403, 94, 2012. 116. Little WC, Freeman GL: Pericardial disease. Circulation 113:1622–1632, 2006. 117. Lowe JE, Oldham HN, Jr, Sabiston DC, Jr: Surgical management   of congenital coronary artery fistulas. Ann Surg 194(4):373–380, 1981. 118. Mahnken AH: CT Imaging of coronary stents: Past, present, and future. ISRN Cardiol 2012:139823, 2012. [Epub 2012 Sep 11]. 119. Malagutti P, Nieman K, Meijboom WB, et al: Use of 64-slice CT in symptomatic patients after coronary bypass surgery: Evaluation of grafts and coronary arteries. Eur Heart J 28:1879–1885, 2007. 120. Mandal S, Tadros SS, Soni S, et al: Single coronary artery anomaly: Classification and evaluation using multidetector computed tomography and magnetic resonance angiography. Pediatr Cardiol 35(3):441–449, 2014. 121. Manson JE, Allison MA, Rossouw JE, et al: Estrogen therapy and coronary-artery calcification. N Engl J Med 356:2591–2602, 2007. 122. Marom EM, Herndon JE, Kim YH, et al: Variations in pulmonary venous drainage to the left atrium: Implications for radiofrequency ablation. Radiology 230:824–829, 2004. 123. Meijboom WB, Meijs MF, Schuijf JD, et al: Diagnostic accuracy of 64-slice computed tomography coronary angiography: A prospective, multicenter, multivendor study. J Am Coll Cardiol 52(25):2135–2144, 2008. 124. Meizlish JL, Berger HJ, Plankey M, et al: Functional left ventricular aneurysm formation after acute anterior transmural myocardial infarction: Incidence, natural history, and prognostic implications.   N Engl J Med 311:1001–1006, 1984. 125. Messika-Zeitoun D, Serfaty JM, Laissy JP, et al: Assessment of the mitral valve area in patients with mitral stenosis by multislice computed tomography. J Am Coll Cardiol 48(2):411–413, 2006. 126. Miller JA, Anavekar NS, El Yaman MM, et al: Computed tomographic angiography identification of intramural segments in anomalous coronary arteries with interarterial course. Int J Cardiovasc Imaging 28(6):1525–1532, 2012. 127. Miller JM, Rochitte CE, Dewey M, et al: Diagnostic performance of coronary angiography by 64-row CT. N Engl J Med 359(22):2324–2336, 2008. 128. Min JK, Dunning A, Lin FY, et al: Rationale and design of the CONFIRM (Coronary CT Angiography Evaluation for Clinical Outcomes: An International Multicenter) registry. J Cardiovasc Comput Tomogr 5:84e92, 2011. 129. Min JK, Kang N, Shaw LJ, et al: Costs and clinical outcomes after coronary multidetector CT angiography in patients without known coronary artery disease: Comparison to myocardial perfusion SPECT. Radiology 249(1):62–70, 2008. 130. Min JK, Labounty TM, Gomez MJ, et al: Incremental prognostic value of coronary computed tomographic angiography over coronary artery calcium score for risk prediction of major adverse cardiac events in asymptomatic diabetic individuals. Atherosclerosis 232(2):298–304, 2014. 131. Min JK, Lin FY, Gidseg DS, et al: Determinants of coronary calcium conversion among patients with a normal coronary calcium scan: What is the “warranty period” for remaining normal? J Am Coll Cardiol 55:1110–1117, 2010. 132. Min JK, Shaw LJ, Berman DS, et al: Costs and clinical outcomes in individuals without known coronary artery disease undergoing coronary computed tomographic angiography from an analysis of Medicare category III transaction codes. Am J Cardiol 102(6):672–678, 2008. 133. Mohlenkamp S, Hort W, Ge J, et al: Update on myocardial bridging. Circulation 106(20):2616–2622, 2002. 134. Moore KL, Persaud TVN: The cardiovascular system. In Moore KL, Persaud TVN, editors: The Developing Human Clinically Orientated Embryology, ed 8, Philadelphia, PA, 2008, Saunders Elsevier, pp 285–337.

135. Morris MF, Maleszewski JJ, Suri RM, et al: CT and MR imaging of the mitral valve: Radiologic-pathologic correlation. Radiographics 30(6):1603–1620, 2010. 136. Motoyama S, Kondo T, Anno H, et al: Atherosclerotic plaque characterization by 0.5-mm-slice multislice computed tomographic imaging. Circ J 71:363–366, 2007. 137. Motoyama S, Kondo T, Sarai M, et al: Multislice computed tomographic characteristics of coronary lesions in acute coronary syndromes. J Am Coll Cardiol 50:319–326, 2007. 138. Motoyama S, Sarai M, Harigaya H, et al: Computed tomographic angiography characteristics of atherosclerotic plaques subsequently resulting in acute coronary syndrome. J Am Coll Cardiol 54(1):49–57, 2009. 139. Mousseaux E, Meunier P, Azancott S, et al: Gaux JC. Cardiac metastatic melanoma investigated by magnetic resonance imaging. Magn Reson Imaging 16:91–95, 1998. 140. Mowatt G, Cook JA, Hillis GS, et al: 64-Slice computed tomography angiography in the diagnosis and assessment of coronary artery disease: Systematic review and meta-analysis. Heart 94(11):1386–1393, 2008. 141. Mushtaq S, Andreini D, Pontone G, et al: Prognostic value of coronary CTA in coronary bypass patients: A long-term follow-up study. JACC Cardiovasc Imaging 7(6):580–589, 2014. 142. Myers RB, Spodick DH: Constrictive pericarditis: Clinical and pathophysiologic characteristics. Am Heart J 138:219–232, 1999. 143. Narula J, Achenbach S: Napkin-ring necrotic cores: Defining circumferential extent of necrotic cores in unstable plaques. JACC Cardiovasc Imaging 2(12):1436–1438, 2009. 144. Nasser WK: Congenital diseases of the pericardium. Cardiovasc Clin 7:271–286, 1976. 145. Nieman K, Pattynama PM, Rensin BJ, et al: Evaluation of patients after coronary artery bypass surgery: CT angiographic assessment of grafts and coronary arteries. Radiology 229:749–756, 2003. 146. Nishimura RA: Constrictive pericarditis in the modern era: A diagnostic dilemma. Heart 86:619–623, 2001. 147. Nishio M, Ueda Y, Matsuo K, et al: Detection of disrupted plaques by coronary CT: Comparison with angioscopy. Heart 97(17):1397–1402, 2011. 148. O’Donnell DH, Abbara S, Chaithiraphan V, et al: Cardiac tumors: Optimal cardiac MR sequences and spectrum of imaging appearances. Am J Roentgenol 193:377–387, 2009. 149. O’Leary SM, Williams PL, Williams MP, et al: Imaging the pericardium: Appearances on ECG-gated 64-detector row cardiac computed tomography. Br J Radiol 83:194–205, 2010. 150. Oh KY, Shimizu M, Edwards WD, et al: Surgical pathology of the parietal pericardium: A study of 344 cases (1993–1999). Cardiovasc Pathol 10(4):157–168, 2001. 151. Otsuka K, Fukuda S, Tanaka A, et al: Napkin-ring sign on coronary CT angiography for the prediction of acute coronary syndrome. JACC Cardiovasc Imaging 6(4):448–457, 2013. 152. Peebles CR, Shambrook JS, Harden SP: Pericardial disease–Anatomy and function. Br J Radiol 84:S324–S337, 2011. 153. Pejkovic B, Krajnc I, Anderhuber F, et al: Anatomical aspects of the arterial blood supply to the sinoatrial and atrioventricular nodes of the human heart. J Int Med Res 36(4):691–698, 2008. 154. Pena E, Nguyen ET, Merchant N, et al: ALCAPA syndrome: Not just a pediatric disease. Radiographics 29(2):553–565, 2009. 155. Petersen SE, Selvanayagam JB, Wiesmann F, et al: Left ventricular non-compaction: Insights from cardiovascular magnetic resonance imaging. J Am Coll Cardiol 46(1):101–105, 2005. 156. Polacek P, Kralove H: Relation of myocardial bridges and loops on   the coronary arteries to coronary occlusions. Am Heart J 61:44–52, 1961. 157. Psychidis-Papakyritsis P, de Roos A, Kroft LJM: Functional MRI of congenital absence of the pericardium. AJR Am J Roentgenol 189:W312–W314, 2007. 158. Raff GL, Abidov A, Achenbach S, et al: SCCT guidelines for the interpretation and reporting of coronary computed tomographic angiography. J Cardiovasc Comput Tomogr 3(2):122–136, 2009.

CHAPTER 58  Coronary Arteries, Heart, and Pericardium 159. Raggi P, Shaw LJ, Berman DS, et al: Prognostic value of coronary artery calcium screening in subjects with and without diabetes. J Am Coll Cardiol 43:1663–1669, 2004. 160. Rastegar N, Burt JR, Corona-Villalobos CP, et al: Cardiac MR findings and potential diagnostic pitfalls in patients evaluated for arrhythmogenic right ventricular cardiomyopathy. Radiographics 34(6):1553–1570, 2014. 161. Restrepo CS, Lemos DF, Lemos JA, et al: Imaging findings in cardiac tamponade with emphasis on CT. Radiographics 27:1595–1610, 2007. 162. Rienmuller R, Groll R, Lipton MJ: CT and MR imaging of pericardial disease. Radiol Clin North Am 42:587–601, 2004. 163. Ropers D, Baum U, Pohle K, et al: Detection of coronary artery stenoses with thin-slice multi-detector row spiral computed tomography and multiplanar reconstruction. Circulation 107:664–666, 2003. 164. Ropers D, Pohle FK, Kuettner A, et al: Diagnostic accuracy of noninvasive coronary angiography in patients after bypass surgery using 64-slice spiral computed tomography with 330-ms gantry rotation. Circulation 114:2334–2341, 2006. 165. Ropers D, Ulzheimer S, Wenkle E, et al: Investigation of aortocoronary artery bypass grafts by multislice spiral computed tomography with electrocardiographic- gated image reconstruction. Am J Cardiol 88:792–795, 2001. 166. Rostamian A, Narayan SM, Thomson L, et al: The incidence, diagnosis, and management of pulmonary vein stenosis as a complication of atrial fibrillation ablation. J Interv Card Electrophysiol 40(1):63–74, 2014. 167. Rozenshtein A, Boxt LM: Computed tomography and magnetic resonance imaging of patients with valvular heart disease. J Thorac Imaging 15(4):252–264, 2000. 168. Rubin GD, Leipsic J, Joseph Schoepf U, et al: CT angiography after 20 years: A transformation in cardiovascular disease characterization continues to advance. Radiology 271(3):633–652, 2014. 169. Rumberger JA, Brundage BH, Rader DJ, et al: Electron beam computed tomographic coronary calcium scanning: A review and guidelines for use in asymptomatic persons. Mayo Clin Proc 74:243–252, 1999. 170. Rybicki F, Otero H, Steigner M, et al: Initial evaluation of coronary images from 320-detector row computed tomography. Int J Cardiovasc Imaging 24:535–546, 2008. 171. Said SA, van der Werf T: Dutch survey of coronary artery fistulas   in adults: Congenital solitary fistulas. Int J Cardiol 106(3):323–332, 2006. 172. Salavati A, Radmanesh F, Heidari K, et al: Dual-source computed tomography angiography for diagnosis and assessment of coronary artery disease: Systematic review and meta-analysis. J Cardiovasc Comput Tomogr 6(2):78–90, 2012. 173. Saremi F, Gera A, Ho SY, et al: CT and MR imaging of the pulmonary valve. Radiographics 34(1):51–71, 2014. 174. Saremi F, Goodman G, Wilcox A, et al: Coronary artery ostial atresia: Diagnosis of conotruncal anastomotic collateral rings using CT angiography. JACC Cardiovasc Imaging 4(12):1320–1323, 2011. 175. Saremi F, Ho SY, Sanchez-Quintana D: Morphological assessment of RVOT: CT and CMR imaging. JACC Cardiovasc Imaging 6(5):631–635, 2013. 176. Saremi F, Muresian H, Sánchez-Quintana D: Coronary veins: Comprehensive CT-anatomic classification and review of variants and clinical implications. Radiographics 32(1):E1–E32, 2012. 177. Sari I, Kizilkan N, Sucu M, et al: Double right coronary artery: Report of two cases and review of the literature. Int J Cardiol 130(2):e74–e77, 2008. 178. Sarwar A, Shaw LJ, Shapiro MD, et al: Diagnostic and prognostic value of absence of coronary artery calcification. JACC Cardiovasc Imag 2:675–688, 2009. 179. Scheuermann-Freestone M, Orchard E, Francis J, et al: Partial congenital absence of the pericardium. Circulation 116:e126–e129, 2007. 180. Serruys PW, de Jaegere P, Keimeneij G, et al: Benestent Study Group: A comparison of balloon-expandable-stent implantation with balloon angioplasty in patients with coronary artery disease. N Engl J Med 331:489–495, 1994.

2061

181. Shabetai R, Meaney E: Haemodynamics of cardiac restriction and tamponade. Br Heart J 37:780–785, 1975. 182. Shambrook JS, Chowdhury R, Brown IW, et al: Cross-sectional imaging appearances of cardiac aneurysms. Clin Radiol 65(5):349–357, 2010. 183. Sharma R, Harden S, Peebles C, et al: Percutaneous aspiration of a pericardial cyst: An acceptable treatment for a rare disorder. Heart 93:22, 2007. 184. Siu SC, Silversides CK: Bicuspid aortic valve disease. J Am Coll Cardiol 55(25):2789–2800, 2010. 185. Small GR, Yam Y, Chen L, et al: Prognostic assessment of coronary artery bypass patients with 64-slice computed tomography angiography: Anatomical information is incremental to clinical risk prediction. J Am Coll Cardiol 58(23):2389–2395, 2011. 186. Smith CR, Leon MB, Mack MJ, et al: Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med 364(23):2187–2198, 2011. 187. Sparrow PJ, Merchant N, Provost YL, et al: CT and MR imaging findings in patients with acquired heart disease at risk for sudden cardiac death. Radiographics 29(3):805–823, 2009. 188. Stanford W, Thompson BH, Weiss RM: Coronary artery calcification: Clinical significance and current methods of detection. AJR Am J Roentgenol 161(6):1139–1146, 1993. 189. Stone GW, Maehara A, Mintz GS: The reality of vulnerable plaque detection. JACC Cardiovasc Imaging 4(8):902–904, 2011. 190. Stone GW, Moses JW, Ellis SG, et al: Safety and efficacy of sirolimusand paclitaxel-eluting coronary stents. New Engl J Med 356:998–1008, 2007. 191. Sun Z, Choo GH, Ng KH, et al: angiography: Current status and continuing challenges. Br J Radiol 85(1013):495–510, 2012. 192. Talreja DR, Edwards WD, Danielson GK, et al: Constrictive pericarditis in 26 patients with histologically normal pericardial thickness. Circulation 108:1852–1857, 2003. 193. Taylor AJ, Cerqueira M, Hodgson JM, et al: ACCF/SCCT/ACR/ AHA/ ASE/ASNC/NASCI/SCAI/SCMR 2010 appropriate use criteria for cardiac computed tomography. A report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, the Society of Cardiovascular Computed Tomography, the American College of Radiology, the American Heart Association, the American Society of Echocardiography, the American Society of Nuclear Cardiology, the North American Society for Cardiovascular Imaging, the Society for Cardiovascular Angiography and Interventions, and the Society for Cardiovascular Magnetic Resonance. J Am Coll Cardiol 56:1864–1894, 2010. 194. Taylor AJ, Cerqueira M, Hodgson JM, et al: ACCF/SCCT/ACR/AHA/ ASE/ASNC/NASCI/SCAI/SCMR 2010 Appropriate Use Criteria for Cardiac Computed Tomography. A Report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, the Society of Cardiovascular Computed Tomography, the American College of Radiology, the American Heart Association, the American Society of Echocardiography, the American Society of Nuclear Cardiology, the North American Society for Cardiovascular Imaging, the Society for Cardiovascular Angiography and Interventions, and the Society for Cardiovascular Magnetic Resonance. Circulation 122(21):e525–e555, 2010. 195. Terpenning S, White CS: Imaging pitfalls, normal anatomy, and anatomical variants that can simulate disease on cardiac imaging as demonstrated on multidetector computed tomography. Acta Radiol Short Rep 4(1):2015 epub. 196. Topaz O, DiSciascio G, Cowley MJ, et al: Absent left main coronary artery: Angiographic findings in 83 patients with separate ostia of the left anterior descending and circumflex arteries at the left aortic sinus. Am Heart J 122(2):447–452, 1991. 197. Veltman CE, de Graaf FR, Schuijf JD, et al: Prognostic value of coronary vessel dominance in relation to significant coronary artery disease determined with non-invasive computed tomography coronary angiography. Eur Heart J 33(11):1367–1377, 2012. 198. Versteylen MO, Kietselaer BL, Dagnelie PC, et al: Additive value of semiautomated quantification of coronary artery disease using cardiac

2062

199. 200. 201. 202.

203. 204.

205.

PART II  CT and MR Imaging of the Whole Body

computed tomographic angiography to predict future acute coronary syndrome. J Am Coll Cardiol 61(22):2296–2305, 2013. Vignaux O: Cardiac sarcoidosis: Spectrum of MRI features. AJR Am J Roentgenol 184(1):249–254, 2005. Wang ZJ, Reddy GP, Gotway MB, et al: CT and MR imaging of pericardial disease. Radiographics 23(Spec No):S167–S180, 2003. Wang ZJ, Reddy GP, Gotway MB, et al: Cardiovascular shunts: MR imaging evaluation. Radiographics 23:S181–S194, 2003. Warnes CA, Williams RG, Bashore TM, et al: ACC/AHA 2008 Guidelines for the Management of Adults with Congenital Heart Disease: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to develop guidelines on the management of adults with congenital heart disease). Circulation 118(23):e714–e833, 2008. Weiford BC, Subbarao VD, Mulhern KM: Noncompaction of the ventricular myocardium. Circulation 109(24):2965–2971, 2004. Weustink AC, Nieman K, Pugliese F, et al: Diagnostic accuracy of computed tomography angiography in patients after bypass grafting: comparison with invasive coronary angiography. JACC Cardiovasc Imaging 2(7):816–824, 2009. Yamagishi M, Terashima M, Awano K, et al: Morphology of vulnerable coronary plaque: Insights from follow-up of patients examined by

206.

207.

208. 209. 210. 211.

intravascular ultrasound before an acute coronary syndrome. J Am Coll Cardiol 35(1):106–111, 2000. Yamanaka O, Hobbs RE: Coronary artery anomalies in 126,595 patients undergoing coronary arteriography. Cathet Cardiovasc Diagn 21(1): 28–40, 1990. Yau JM, Singh R, Halpern EJ, et al: Anomalous origin of the left coronary artery from the pulmonary artery in adults: A comprehensive review of 151 adult cases and a new diagnosis in a 53-year-old woman. Clin Cardiol 34(4):204–210, 2011. Yoo SM, Lee HY, White CS: Screening coronary CT angiography: Possibilities and pitfalls. Int J Cardiovasc Imaging 2014 Jul 16 epub. Zeina AR, Odeh M, Blinder J, et al: Myocardial bridge: Evaluation on MDCT. AJR Am J Roentgenol 188(4):1069–1073, 2007. Zenooz NA, Habibi R, Mammen L, et al: Coronary artery fistulas: CT findings. Radiographics 29(3):781–789, 2009. Zurick AO, Bolen MA, Kwon DH, et al: Pericardial delayed hyperenhancement with CMR imaging in patients with constrictive pericarditis undergoing surgical pericardiectomy: A case series with histopathological correlation. JACC Cardiovasc Imaging 4(11): 1180–1191, 2011.

60  Musculoskeletal Tumors Amilcare Gentili, Nicholas Bhojwani, and Mark R. Robbin

SKELETAL TUMORS Imaging Techniques Many imaging modalities are available for the evaluation of skeletal lesions. In most cases, standard radiographs are sufficient for detecting skeletal tumors.127 When evaluated together with clinical data (e.g., patient’s age and sex, anatomic location of the lesion, clinical presentation), radiographs are the best predictors of the histologic condition of the lesion. To fully characterize a tumor in an anatomically complex area such as the spine, scapula, or pelvis or in a periarticular location, however, magnetic resonance imaging (MRI) or computed tomography (CT) is often necessary. PET/CT has partially replaced bone scintigraphy as the modality  of choice to screen for skeletal metastases. PET/CT with sodium 18 F-fluoride (NaF) has higher sensitivity, specificity, positive predictive value, negative predictive value, and accuracy in detecting bone  metastases than conventional 99mTc-methylene diphosphonate (MDP) skeletal scintigraphy.17,82,117 Although MRI is more sensitive than scintigraphy in detecting metastases, it is not an efficient technique for evaluating the entire skeleton.9,41 However, with improvements in hardware, such as the introduction of multichannel MR scanners, a system of multiple phased-array coils covering the whole body, and acquisition acceleration techniques, high-resolution whole-body MRI has become clinically feasible.32,35,120,145,147 Whereas plain radiographs may be sufficient for diagnosis, MRI and CT are fundamental in staging skeletal tumors. Optimal treatment of primary skeletal neoplasms in the absence of distant metastases includes complete surgical resection of the tumor. In the past, amputation was the treatment of choice. With improved surgical technique and aggressive chemotherapy, however, limb-sparing operations can be performed, providing better functional results without affecting survival. In planning resection, accurate preoperative staging is essential. The initial CT or MRI procedures should be performed before biopsy; afterward it is often difficult to distinguish tumor from edema, hemorrhage, and granulation tissue. Numerous studies have compared the usefulness of CT and MRI in the staging of musculoskeletal tumors.8,11,12,14,49,51,84,108,114 MRI is superior for determining muscle compartment and vascular involvement because of the intrinsic contrast among tumor mass, muscle, and fat, without the need for enhancement with contrast medium. CT is superior for detecting subtle cortical invasion and periosteal and endosteal reaction and for depicting matrix calcification or ossification. Another advantage of CT is scan times of less than 10 seconds; this is particularly helpful when evaluating pediatric patients because it virtually eliminates the need for sedation and minimizes the dependence on patient cooperation.37 Both MRI and CT are useful in evaluating

2070

the response to chemotherapy and radiation therapy.57 Signs of response to treatment are decreased tumor size, improved delineation of the mass, reappearance of fat planes between muscle groups, and calcification or ossification of the tumor.

Computed Tomography With the advent of multidetector CT, images can be acquired with subsecond gantry rotation times at a slice thickness of 0.5 to 0.625 mm, which yields true isotropic high-resolution volume data sets. The isotropic volume data provide the ability to retrospectively reconstruct multiple high-resolution image sets from the original raw data, requiring only one acquisition for the production of three-dimensional CT images in numerous planes.25,37 For evaluation of bone lesions, images are reconstructed using bone algorithms, and the images are reviewed using both bone and soft tissue window settings. Intravenous administration of a contrast agent aids in visualizing major vascular bundles and assessing the vascularity of a lesion; it may also be helpful in distinguishing necrotic from viable tumor. CT angiography is useful in the assessment of the neurovascular bundle and in surgical planning.

Magnetic Resonance Imaging Numerous pulse sequences are available for the evaluation of musculoskeletal tumors, including spin echo and fast-spin echo, inversion recovery, gradient echo, chemical shift, and diffusion-weighted sequences. Our protocol for evaluating musculoskeletal masses includes images in at least two planes: usually a T1-weighted sequence in the sagittal or coronal plane, followed by either a double-echo (protondensity, T2-weighted) fast-spin echo sequence or a T2-weighted, fatsaturated, fast-spin echo sequence in the axial plane. A short tau inversion recovery (STIR) sequence is occasionally obtained to increase the conspicuity of a lesion.126 Whenever possible, imaging is performed with a surface coil for enhanced spatial resolution and an improved signal-to-noise ratio. Body coils are used when an entire bone or soft tissue region is being evaluated for the detection of skip lesions or for surgical planning. Use of contrast-enhanced MRI for characterization of musculoskeletal tumors has yielded controversial results.7,80,81,110,134 Contrast enhancement is useful in grading a lesion, assessing response to treatment, and distinguishing postoperative changes from recurrence.31,32,155

Benign Skeletal Tumors Enchondroma

Enchondroma is a common benign primary osseous tumor; only  nonossifying fibromas and exostosis are more common.4,72 There is no sex predilection, and approximately 50% of these tumors occur in the hands. They are asymptomatic unless they are undergoing malignant

CHAPTER 60  Musculoskeletal Tumors

A

C

2071

B

E

D

FIG 60-1  Enchondroma found incidentally on MRI in a 70-year-old woman. A and B, AP and lateral radiographs of the right knee demonstrate a lesion with ring-and-arc chondroid matrix mineralization in the distal femoral metadiaphysis. Coronal T1-weighted (C), T2-weighted (D), and T1-weighted fat-saturated postgadolinium (E) images demonstrate lobulated intramedullary lesion with heterogeneous areas of increased signal intensity on T2 corresponding to chondroid matrix, and low signal on all sequences corresponding to areas of matrix calcification. T1-weighted fat-saturated postgadolinium images demonstrate enhancement of cartilaginous tissue.

transformation or the bone is fractured, and they are often an incidental finding on radiographs. Enchondromas are typically round or oval lesions with well-defined lobulated borders. The unmineralized matrix exhibits soft tissue attenuation on CT and homogeneous high signal intensity on T2-weighted MRI. Calcification with a ring-and-arc appearance, which suggests a chondroid matrix,4,19 is best identified on plain radiography and CT. Calcifications are not consistently identified with MRI; when seen, however, they have low signal intensity on all sequences (Figs. 60-1   and 60-2). Benign and malignant chondroid lesions can have a similar appearance, making it difficult or impossible to differentiate an enchondroma from a low-grade chondrosarcoma. The presence of a soft tissue mass or cortical erosion or pain in the absence of fracture, however, increases the likelihood of malignancy.18,24,26,101,154

Chondroblastoma Chondroblastoma is a rare benign tumor of cartilaginous origin composed of chondroblasts. Its peak incidence is in individuals between 10 and 20 years of age. It is more common in males, with a male-to-female ratio of 2 : 1. Chondroblastoma involves the epiphysis, apophysis, or both; the tumor may extend into the metaphysis after destroying the growth plate.60 CT accurately demonstrates the extent of the lesion as well as matrix calcifications and subtle cortical infractions. Extension through the growth plate, however, is best detected with conventional tomography or MRI because the growth plate is usually best visualized in the sagittal or coronal plane (Fig. 60-3). Both T1- and T2-weighted MRIs demonstrate a peripheral rim of signal void that corresponds to the sclerotic margin of the lesion.65 Abnormal signal intensity in

C

B

A

FIG 60-2  Enchondroma in a 60-year-old man with hand pain. A, AP radiograph of left hand. Coronal STIR (B) and T1-weighted fat-saturated postgadolinium (C) images demonstrate an expansile lytic lesion with ringand-arc mineralization and endosteal scalloping in the distal second metacarpal. T1- and T2-weighted images demonstrate heterogeneous peripheral signal intensity and enhancement of the second metacarpal. Pathology post curettage showed enchondroma with no evidence of chondrosarcoma.

A

C

B

D FIG 60-3  Chondroblastoma in a 19-year-old male with shoulder pain. A, AP radiograph demonstrates a lytic lesion with sclerotic rim in the right humeral head epiphysis. Coronal proton density (B) and T2-weighted fat-saturated (C) images demonstrate T2 heterogeneous hyperintensity with low signal foci and surrounding bone marrow edema. D, Axial CT fluoroscopic image with core biopsy needle within the lesion.

CHAPTER 60  Musculoskeletal Tumors large areas of bone marrow surrounding the lesion has been seen and probably represents bone marrow edema.13,68,144,152 Fluid levels may be present in cystic components, corresponding to aneurysmal bone cysts.20,68,69,118,128

Osteochondroma Osteochondromas (exostoses) are the second most common benign tumor of bone after nonossifying fibromas. The male-to-female ratio is 1.5 : 1 to 2 : 1.136 Although these are congenital lesions, they are usually discovered in individuals between 10 and 20 years of age. Multiple exostoses are uncommon, occurring only one tenth as frequently  as solitary exostosis, but they manifest earlier, usually before 10 years of age.

A

C

2073

Ninety percent of osteochondromas originate from a long bone close to the metaphysis. The most common locations are around the knee (distal femur and proximal tibia) and the proximal humerus. The morphology of an osteochondroma is more important than the signal intensity in terms of diagnosis. The incidence of malignant transformation is approximately 1% for solitary osteochondromas and 5% to 25% for hereditary multiple exostoses.1,98,122 Both CT and MRI demonstrate the continuity of the cortex and medullary cavity of the osteochondroma with that of the parent bone116 (Fig. 60-4). The perichondrium is well seen on T2-weighted MRI as an area of low signal intensity surrounding the outer surface of the high-signal-intensity cartilage cap.77 MRI measurements of cartilage cap thickness are also accurate, whereas CT measurements of

B

D

FIG 60-4  Osteochondroma in a 23-year-old woman with knee pain. A, AP radiograph of the distal femur. On MRI, T1-weighted axial (B), T2-weighted fat-saturated axial (C), and T1-weighted fat-saturated coronal postgadolinium (D) images show a pedunculated lesion arising from the distal medial cortex of the femur, with cortical and medullary continuity and a thin overlying cartilage cap. Overlying the lobulated cartilaginous cap is a thin rim of high T2 signal of the formed bursa, with surrounding bursal enhancement from bursitis (arrows).

2074

PART II  CT and MR Imaging of the Whole Body

maximal cartilage thickness are often imprecise. The thickness of this structure is important in distinguishing benign osteochondroma from exostotic chondrosarcoma. According to most authors, the cartilage cap is usually thicker than 1.5 cm in chondrosarcoma.98 Complications of osteochondroma include nerve injury, vascular injury, deep venous thrombosis, and bursa formation.15,28,54,77,88,91,98 Formation of a bursa over an osteochondroma is common, and the bursa is usually asymptomatic; if it becomes inflamed and distended, it can be painful. On T2-weighted images, bursal fluid has high signal intensity similar to that of the cartilage cap, and it can be difficult to differentiate the two. On gradient echo sequences, cartilage has a lower signal intensity than that of fluid, and the diagnosis is easily made. A rarer complication is formation of a pseudoaneurysm adjacent to the exostosis.113

A

B

C

D

Chondromyxoid Fibroma Chondromyxoid fibroma is composed of three principal elements: chondroid, fibrous, and myxoid tissue. It is the least common of the benign cartilaginous bone neoplasms.136 There is a male predominance, with a male-to-female ratio of 1.5 : 1 to 2 : 1. The tumor tends to occur around the knee joint but can involve any bone. The tumor is usually solid but may contain cystic or hemorrhagic areas and rarely may be associated with an aneurysmal bone cyst.124 On MRI the signal intensities of this tumor vary with the proportions of chondroid, fibrous, and myxoid tissues present. CT may be better than MRI in delineating matrix calcification and sclerosis around the tumor149 (Fig. 60-5).

E

FIG 60-5  Chondromyxoid fibroma of the tibia in a 25-year-old woman with knee pain. AP (A) and lateral (B) radiographs, as well as sagittal T1-weighted (C), sagittal T2-weighted fat-saturated (D), and axial T2-weighted fat-saturated (E) images demonstrate eccentrically located lytic lesion at the tibial tubercle, with endosteal scalloping and sclerotic margins. MRI demonstrates hypointensity on T1 and hyperintensity on T2. No ring-and-arc matrix mineralization is seen on radiographs but can often be seen on CT.

CHAPTER 60  Musculoskeletal Tumors Osteoid Osteoma Osteoid osteomas are relatively common; only osteochondromas and fibrous cortical defects are more prevalent. The male-to-female ratio is 3 : 1. They typically occur in teenagers and young adults and are rare before age 5 and after age 30. These tumors occur most frequently in the femur, tibia, and humerus and involve the diaphysis and less commonly the metaphysis. When the spine is involved, the posterior elements are typically affected. The main clinical symptom is pain relieved by aspirin. CT is very accurate in detecting the nidus and is preferable to MRI for evaluating osteoid osteoma.45,78 Occasionally an osteoid osteoma can be confused with a stress fracture on MRI because of edema in the bone marrow. On CT the typical appearance of an osteoid osteoma is an area of sclerosis surrounding a small (<1 cm) radiolucent nidus (Fig. 60-6). Lesions larger than 1.5 cm are considered osteoblastomas. The nidus may contain a variable number of calcifications ranging from none to enough to cause almost complete calcification, with only a thin peripheral rim of low density. For full evaluation of the nidus,

A

C

2075

thin-section CT with less than 2-mm slices is needed because the nidus is often less than 1 cm in diameter. The nidus enhances on dynamic CT scans78; this enhancement helps in differentiating osteoid osteoma from Brodie’s abscess. Depending on the degree of calcification, the nidus has low signal intensity on T1-weighted images and variable intensity on T2-weighted images.43 The surrounding bone marrow has low signal intensity on all pulse sequences in the presence of reactive sclerosis or high signal intensity on T2-weighted images in the presence of edema (see Fig. 60-6).34 Osteoid osteomas in younger patients tend to be associated with more extensive peritumoral edema.34 When the osteoid osteoma is intraarticular, an associated joint effusion may be detected.119 CT is used to direct percutaneous ablation of osteoid osteomas.21,29,146

Osteoblastoma Osteoblastoma is a rare benign tumor. It is at least four times less common than osteoid osteoma.136 There is a male predominance, with a male-to-female ratio of 2.5 : 1. Its peak incidence is in the second

B

D FIG 60-6  Osteoid osteoma in a 12-year-old girl with hip pain, worse at night. Coronal CT reconstruction (A), coronal T1-weighted MRI (B), and coronal STIR MRI (C) through the right proximal femur show medial femoral neck sclerosis and eccentric lytic component with central sclerotic nidus (arrow in A). On MRI there is corresponding bone marrow edema with high signal on STIR and low signal on T1-weighted images. The central nidus is hypointense on both STIR and T1. D, Axial CT fluoroscopic image with ablation of the lesion via anterior approach.

2076

PART II  CT and MR Imaging of the Whole Body

B

A

C

D FIG 60-7  Osteoblastoma in a 26-year-old man with back pain. Axial CT in soft tissue (A) and bone windows (B), as well as sagittal T1- (C) and T2-weighted (D) MRI demonstrate a large, expansile, lytic lesion in the posterior elements, with internal osteoid matrix and surrounding dense sclerosis. On MRI the lesion is hypointense on T1 (arrow in C), corresponding to areas of mineralization, and hyperintense on T2 (arrow in D), with surrounding peritumoral edema.

decade, and it is rare before age 10 and after age 30. Its most common location is the spine, usually in the posterior elements. In the long bones it is most often located in the metadiaphyseal region; in rare circumstances it can extend into the epiphyses in adults.106 Osteoblastomas may be associated with aneurysmal bone cysts.139 On CT, osteoblastoma appears as an expansile lytic lesion, often with a mineralized matrix surrounded by a thin bony shell. Dense sclerosis and periosteal reaction may occasionally be present. Osteoblastoma typically has low to intermediate signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Edema in the surrounding soft tissues and in the bone marrow beyond the tumor margins is common22,76,123 (Fig. 60-7).

Intraosseous Lipoma Intraosseous lipoma is a rare benign tumor, accounting for less than 0.1% of primary bone tumors. There is a male predominance, with a male-to-female ratio of 1.6 : 1. Its peak incidence is in the fifth decade, but there is a wide age range from the second to eighth decades. Its most common location is in the metaphyses of long bones, especially the femur, tibia, and fibula, and in the calcaneus.87 The radiographic appearance of lipoma of bone is often pathognomonic; it is a well-defined lesion with sclerotic borders and central mineralization, and in most cases, no additional studies are necessary. CT and MRI are usually diagnostic because they demonstrate the fat

CHAPTER 60  Musculoskeletal Tumors

A

2077

C

B

D FIG 60-8  Intraosseous lipoma in a 52-year-old man. A and B, AP left hip radiographs. C, Coronal T1-weighted MRI. Images show a well-defined lucent lesion in the intertrochanteric region, with sclerotic border and central calcification. On T1 (C) the lesion is isointense, with central hypointensity corresponding to the region of calcification. D, Lateral foot radiograph in a different 55-year-old man. In the anterior calcaneus the central calcification of the lipoma is better visualized.

content of the lesion (Fig. 60-8). Central calcifications or ossifications are often present. With involution of the intraosseous lipoma, fibrous proliferation and cystic degeneration can be seen and may become a predominant feature on CT and MRI.72,102,151

Nonossifying Fibroma Nonossifying fibroma and fibrous cortical defects share the same tissue appearance and differ only in size, with fibrous cortical defects being smaller. They are sometimes called fibroxanthomas because they contain spindle-shaped fibroblasts and xanthoma (foam) cells. They are the most common benign lesion of bone and have been reported in 30% of normal children. Most of these nonneoplastic developmental aberrations are asymptomatic and disappear spontaneously. The male-to-female ratio is 1.5 : 1, and the peak incidence occurs around age 10. Nonossifying fibromas and fibrous cortical defects are usually localized to the metaphysis of a long bone and are most commonly found about the knee (distal femur and proximal tibia). They are uncommon

in the upper extremities. Their radiographic appearance is often pathognomonic, and usually no additional studies are necessary. They are frequently an incidental finding on MRI examinations of the knee and can have a varied MRI appearance (Fig. 60-9). In the early stages when the lesion is completely lytic on plain radiographs, it may have high signal intensity on T1-weighted images as a result of an abundance of xanthoma cells containing fat. Linear structures of low signal intensity are often present and represent fibrous septa or osseous pseudosepta. When the lesion starts healing, it has low signal intensity on both T1- and T2-weighted images because of increased fibrous collagen, mineralization, and possibly hemosiderin deposits.64 The lesions are always well defined and are never surrounded by bone marrow edema.73

Simple Bone Cyst The simple bone cyst is a benign lesion of unknown cause. It arises in the metaphysis of the bone, but with growth it tends to move away from the growth plate toward the diaphysis. It is a common bone

2078

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 60-9  Nonossifying fibroma incidentally found in a 15-year-old boy. AP radiograph of the tibia (A), coronal T1-weighted MRI (B), and sagittal STIR images (C) show a distal tibial metaphyseal lucent, lobulated, cortically based lesion with low to intermediate signal on T1, representing fibrous tissue, and heterogeneous high signal on T2. The surrounding hypointense rim represents the sclerotic border.

lesion, following only nonossifying fibromas and exostoses in prevalence. The male-to-female ratio is 2.5 : 1. It can be seen in any age group but is very rare before age 5 and after age 20. The most common site is the proximal humerus, followed by the proximal femur. All other locations are rare. Simple cysts are central, lytic, expansile lesions. Their radiographic appearance is usually diagnostic, and in most cases neither MRI nor CT is necessary. They are well defined and because of their fluid content have low signal intensity on T1-weighted images and high signal intensity on T2 (Fig. 60-10). Typically they are homogeneous on both T1- and T2-weighted images unless hemorrhage has occurred.89 Peripheral low-signal-intensity borders represent reactive sclerosis. No edema is present in the surrounding bone marrow or muscle. CT is superior to MRI in detecting subtle infractions of the lesion.

On both T1- and T2-weighted MRI, an aneurysmal bone cyst appears as a well-defined expansile mass with multiple internal septations surrounded by a well-defined, low-signal-intensity rim of variable thickness.85,150 The rim represents a bony shell. Fluid-fluid levels are often seen within well-defined cystic cavities.139 On T1-weighted images the signal intensity in the cyst may be higher than that of muscle because of intracystic hemorrhage. One cavity may have signal intensity characteristics that are markedly different from those of an adjacent cavity. The different signal characteristics probably reflect intracystic hemorrhage of different ages. Edema may be present in the surrounding muscles. Septations within the lesion enhance after administration of gadolinium, but the cystic component does not enhance.

Aneurysmal Bone Cyst

Although giant cell tumor is found relatively frequently, malignant transformation is rare, occurring in fewer than 10% of lesions. There is a slight female preponderance, and the peak incidence is between 20 and 49 years of age. It is rare before closure of the growth plate or after age 50. This tumor is localized to the metaphysis before closure of the growth plate. Ninety percent of cases involve the long bones, where the lesion is localized to the metaepiphyseal region. More than 50% of giant cell tumors involve the knee (metaepiphyseal region of the distal femur and proximal tibia). Other bones that can be affected include (in order of frequency) the distal radius, sacrum, distal tibia, proximal humerus, pelvis, and proximal femur. Other sites are rarely affected.3,16,23,137 On CT, giant cell tumors appear as lytic lesions with thinning and erosion of the cortex (Fig. 60-12). A sclerotic margin may be present between the tumor and the normal marrow cavity. On MRI, giant cell tumors are well defined, with a low-signal-intensity rim or halo surrounding the tumor. On T1-weighted images they have intermediate signal intensity and are usually homogeneous; on T2-weighted images

Aneurysmal bone cysts are expansile lytic lesions that contain thinwalled cystic cavities; their cause is unknown. They can be seen at any age, although 75% are seen in patients younger than 20 years; they are rare after age 30. There is a slight female predominance, with a maleto-female ratio of 1 : 1.5. Aneurysmal bone cysts can occur in any bone but arise predominantly in the long bones and spine. In the long bones they are usually metaphyseal or metadiaphyseal. In the spine they involve the posterior elements and may extend into the body, but involvement of the vertebral body alone is rare. On CT the cyst appears as an expansile lesion with a thin cortical shell (Fig. 60-11). Fluid-fluid levels may be seen and are thought to represent sedimentation of degraded blood cell products within the cystic cavity59; this is best visualized if the patient is immobilized for some time before scanning. Fluid-fluid levels are a nonspecific finding and have also been described in giant cell tumors, chondroblastoma, and telangiectatic osteosarcoma.118,124,128,133

Giant Cell Tumor

CHAPTER 60  Musculoskeletal Tumors

A

2079

C

B

FIG 60-10  Unicameral bone cyst in a 16-year-old boy with shoulder pain. A, AP radiograph of the right humerus shows expansile lytic lesion of the proximal humeral metadiaphysis, with pathologic fracture (arrows). On T1-weighted (B) and STIR (C) coronal MRI the lesion appears heterogeneously hypointense on T1 and hyperintense on STIR, corresponding to fluid signal, with high-signal areas representing subacute hemorrhage.

A

B

C

D

FIG 60-11  Aneurysmal bone cyst in a 12-year-old boy with wrist pain. A, AP right

E

distal forearm radiograph. Coronal T1-weighted (B), T2-weighted fat-saturated (C), T1-weighted fat-saturated postgadolinium (D), and axial T2-weighted fat-saturated (E) MRI through the distal right radius. The images demonstrate a well-defined expansile, lytic, metaphyseal intramedullary lesion with sclerotic rim. Multiple enhancing internal septations can be seen on postgadolinium image, with multiple hemorrhagic fluid levels of varying age seen on T2. No associated soft tissue mass or cortical destruction is present.

2080

PART II  CT and MR Imaging of the Whole Body

B

A

C

D

E

FIG 60-12  Giant cell tumor in a 27-year-old man with knee pain after trauma. A, AP left knee radiograph. B, Coronal CT. Coronal T1-weighted (C), T2-weighted fat-saturated (D), and T1-weighted fat-saturated postgadolinium (E) MRI. Images show a large, lytic, lateral femoral condylar lesion involving the metaphysis and epiphysis, with extension to the articular surface, cortical erosions, and pathologic fracture. There are intracystic components with hyperintensity on T1 and T2, with peripheral irregular hypointense areas representing hemosiderin and fibrous tissue (arrows in D and E) with surrounding edema in the metaepiphysis.

they have intermediate to high signal intensity and may be inhomogeneous.23,63,99 Local areas with increased signal intensity may be seen on both T1- and T2-weighted images and represent hemorrhage. Fluidfluid levels, often seen in aneurysmal bone cysts and telangiectatic osteosarcomas, are present only rarely in giant cell tumors.133 Both CT and MRI accurately demonstrate the intraosseous and  soft tissue extension of the tumor; CT is better at showing cortical destruction, whereas MRI is better at demonstrating intraarticular involvement.130

Malignant Skeletal Tumors Multiple Myeloma

Multiple myeloma is a multifocal malignant proliferation of plasma cells and is the most common primary malignant tumor of the skeleton.

There is a male predominance, with a male-to-female ratio of 1.5 : 1. It is common after age 50 and rare before age 30. Because this neoplasm derives from bone marrow elements, it involves bones containing red marrow: skull, ribs, sternum, pelvis, proximal humeral metaphysis, and proximal femoral metaphysis. At the time of diagnosis it is often disseminated throughout the red marrow of the axial skeleton. Infiltration of bone marrow has two forms: diffuse and focal. In the diffuse form, myeloma cells are mixed with hematopoietic cells. This form is difficult to image; the only MRI manifestation is an inhomogeneous bone marrow signal intensity that is subjective and difficult to quantitate.42,93,94 CT, plain radiographs, and scintigraphy may result in negative findings. In the focal form, normal bone marrow is displaced by nodules composed entirely of myeloma cells. Untreated myelomatous lesions

CHAPTER 60  Musculoskeletal Tumors

A

2081

B

C FIG 60-13  Plasmacytoma in a 72-year-old man with a history of multiple-week right lateral hip discomfort. AP right hip radiograph (A) and coronal T1-weighted (B), and T2-weighted fat-saturated MRIs (C) of the pelvis demonstrate a large, expansile, lytic mass originating from the right iliac wing, corresponding to the large soft tissue mass that is homogeneously hypointense on T1 and hyperintense on T2 (arrows).

have decreased signal intensity on T1-weighted images and increased signal intensity on T2-weighted images when compared to the surrounding bone marrow.42,93,94 CT demonstrates purely osteolytic lesions in the trabecular bone, which occasionally extend to involve the cortex (Fig. 60-13). Multidetector CT is able to demonstrate smaller lesions than MRI can.61 After irradiation, these lesions show low signal intensity on both T1- and T2-weighted images, and a sclerotic border develops, as seen on CT. Whole-body MRI is used for staging and monitoring response to treatment.33,50,71

Osteosarcoma Osteosarcoma is a malignant neoplasm characterized by production of osteoid by tumor cells. It is the second most common primary malignant tumor of bone after plasmacytoma. It is a rare tumor, representing only 0.2% of all malignant tumors. There is a male predominance, with a male-to-female ratio of 2 : 1. Its peak incidence is in the second decade; it is rare before age 10 and after age 30.

Osteosarcoma can occur in any bone, but it has a strong predilection for the distal femur, proximal tibia, and proximal humerus. Two thirds of osteosarcomas are localized to the knee or shoulder. Osteosarcomas can be divided into two categories: primary  osteosarcomas, which arise de novo, and secondary osteosarcomas, which develop in abnormal bones. Underlying bone abnormalities include Paget’s disease, complications of radiation therapy, multiple enchondromas, multiple osteochondromas, chronic osteomyelitis, fibrous dysplasia, or infarct. Secondary osteosarcomas usually affect older people. MRI is useful for staging the tumor and for follow-up after treatment.40,57,134,135 The osteoblastic component of the tumor has low signal intensity on all sequences. The nonmineralized component has low signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Bone marrow extension is best seen on T1, on which loss of the high signal intensity of normal bone marrow can be appreciated. Soft tissue extension is best seen on T2, whereas tumor

2082

PART II  CT and MR Imaging of the Whole Body

A

D

C

B

E

F

FIG 60-14  Osteosarcoma in an 11-year-old girl with a history of knee pain. A and B, AP and lateral left knee radiographs. C, CT fluoroscopic biopsy. Coronal T1-weighted (D), T2-weighted fat-saturated (E), and T1-weighted fat-saturated postgadolinium (F) MRI. Within the distal lateral femoral metaphysis there is an ill-defined osteoid matrix–producing lesion with adjacent destruction and sunburst periosteal reaction. There is decreased T1 and T2 signal, with associated peripheral nodular T2 increased signal of the soft tissue component, which enhances. Areas of decreased signal intensity correspond to the osteoid matrix.

and muscle may have the same signal intensity on T1. CT is better than MRI in demonstrating matrix mineralization but is less accurate in detecting skip lesions and bone marrow and soft tissue extension  (Fig. 60-14). Telangiectatic osteosarcomas contain large, cystic, blood-filled spaces with fluid-fluid levels that may mimic an aneurysmal bone cyst; however, these tumors are usually less well defined than aneurysmal bone cysts.46,133,139 Local staging of bone tumors must include evaluation for skip metastases, which are defined as synchronous smaller foci of tumor occurring in the same bone, anatomically separated from the primary lesion, or as synchronous smaller foci of tumor on the opposite side of a joint. MRI is the imaging modality of choice for detection of skip metastasis. In the last 2 decades the development of aggressive

chemotherapy has significantly improved survival of patients with osteosarcoma, and imaging studies are used to evaluate the tumor’s response to treatment. CT findings of a positive response to treatment include (1) marked decrease in size or complete disappearance of the soft tissue mass, (2) increased calcification of the mass, (3) improved delineation of the margins, and (4) formation of a peripheral rim of calcification.86,107,125 Decreased signal intensity of the nonmineralized mass on T2-weighted images is thought to represent fibrosis or sclerosis of the tumor. Persistent high signal intensity may be a result of nonresponding or necrotic tumor, reactive granulation tissue, or hemorrhage.40,55,58 Administration of gadolinium cannot help distinguish viable tumor from reactive inflammation, because both enhance; a lack of enhancement indicates tumor necrosis.38,121

CHAPTER 60  Musculoskeletal Tumors

A

B

D

E

2083

C

FIG 60-15  Chondrosarcoma in a 56-year-old woman with new left thigh pain. A, AP radiograph of proximal left femur. Coronal T1-weighted (B), T2-weighted fat-saturated (C), and T1-weighted fat-saturated postgadolinium (D) MRI. E, Axial CT during biopsy. Images show left femoral proximal metadiaphyseal intramedullary lesion with endosteal scalloping and chondroid matrix mineralization (arrow in A). The lesion has intermediate to low signal on T1, high signal on T2, and heterogeneously enhancing periosteum (arrows in D).

Chondrosarcoma Chondrosarcoma is a malignant chondroid tumor. It is the third most common primary malignant bone tumor after plasmacytoma and osteosarcoma. Its peak incidence is between 30 and 60 years of age. Based on their intraosseous location, chondrosarcomas can be divided into central and peripheral lesions. Most chondrosarcomas (≈75%) are primary and arise de novo, but the remaining 25% are secondary and develop from malignant transformation of a benign lesion such as enchondroma, osteochondroma,122 or rarely chondroblastoma. Central chondrosarcomas occur in both tubular bones (e.g., femur, proximal humerus, proximal tibia) and flat bones (e.g., pelvis). Peripheral chondrosarcomas most commonly arise in flat bones (e.g., pelvis, ribs) and the spine.

It is difficult to differentiate low-grade chondrosarcomas from benign chondroid lesions; however, the presence of pain in the absence of fracture and cortical destruction and the presence of a soft tissue mass are suggestive of malignancy.18,24,39,44,96 On MRI, chondrosarcoma has a characteristic multilobular configuration. The lobules of hyaline cartilage have intermediate signal intensity, similar to that of muscle, on T1-weighted images (Fig. 60-15); they have homogeneous high signal intensity on T2-weighted images.55,141 The fibrous septa have low signal intensity on both T1- and T2-weighted images, but they enhance after gadolinium administration.4,48 Calcifications are common in lowgrade chondrosarcoma and are best seen by conventional radiography and CT. Endosteal scalloping is also best depicted with CT and plain radiography.

2084

PART II  CT and MR Imaging of the Whole Body

Ewing’s Sarcoma Ewing’s sarcoma is a malignant neoplasm that is probably of neuroectodermal origin. Among primary malignant neoplasms of bone, it follows plasmacytoma, osteosarcoma, and chondrosarcoma in frequency. There is a male predominance, with a male-to-female ratio of 2 : 1. Its peak incidence is in the second decade, with 90% of cases occurring between 5 and 25 years of age.104 Ewing’s sarcoma may involve any bone but has a predilection for the long bones and pelvis. In the long bones it is localized to the diaphysis or metaphysis, and until the growth plate is open, it does not extend to the epiphysis. MRI facilitates staging of Ewing’s sarcoma and is better than CT for demonstrating bone marrow and soft tissue involvement.52 The MRI signal intensity characteristics of Ewing’s sarcoma are not specific and are similar to those of other malignant neoplasms.40 The tumor’s signal intensity is lower than or equal to that of muscle on T1-weighted images and is higher on T2-weighted images (Fig. 60-16).

A

D

MRI and CT are helpful in evaluating response to treatment. If the tumor is sensitive to treatment, it decreases in size, the periosteal reaction matures, and the bone becomes sclerotic.42,57,79 High signal intensity on T2-weighted images after treatment does not always indicate a poor response; it may represent necrosis, reactive granulation tissue, or hemorrhage. Enhancement after gadolinium administration does not aid in distinguishing reactive changes from residual tumor, but lack of enhancement indicates tumor necrosis.40,57,79

SOFT TISSUE TUMORS Imaging Techniques CT has been widely used for detection and staging of soft tissue tumors, but MRI is now the modality of choice. Because of its intrinsically high soft tissue contrast, MRI is more accurate and sensitive than CT in evaluating these lesions.1,13,28,55,92,99,105,113 On MRI, benign lesions tend to have well-defined margins and homogeneous signal intensity; they do not encase neurovascular

C

B

E FIG 60-16  Ewing’s sarcoma in a 12-year-old girl with a limp during increased activity. A and B, AP and lateral radiographs of the proximal right fibula. Coronal T1-weighted (C) and axial T2 fat-saturated (D) and T1-weighted fat-saturated postgadolinium (E) MRI of the proximal right fibula. Images demonstrate a large, expansile, lytic, and permeative lesion, cortical destruction of the proximal fibular metaphysis and epiphysis, with Codman’s triangle and permeative type of bone destruction. MRIs demonstrate circumferential solid mass extending into the epiphysis from the metaphysis, with high signal on T2, intermediate signal on T1, and avid enhancement.

CHAPTER 60  Musculoskeletal Tumors bundles and are not surrounded by peritumoral edema.10 Conversely, malignant lesions typically have poorly defined margins, heterogeneous signal intensity, neurovascular bundle encasement, and peritumoral edema.11,12 Unfortunately a large overlap exists between benign and malignant lesions, and MRI’s ability to differentiate benign from malignant soft tissue masses remains controversial.7,56,58

Benign Soft Tissue Tumors Lipoma

Lipoma is a benign tumor composed of mature adipose tissue. It is the most frequently occurring benign soft tissue tumor and is often asymptomatic.56 Superficial lipomas predominate in women, whereas deep lipomas are more common in men. The peak incidence of lipomas occurs between 40 and 60 years of age. Superficial lipomas are usually localized to the subcutaneous tissue of the trunk and proximal extremities.132 Deep lipomas are usually in the retroperitoneum, chest wall, and deep soft tissues of the hands  and feet. On CT, lipomas have a low attenuation coefficient, equal to that of subcutaneous fat, and they do not enhance after administration of contrast material. On MRI they have the same signal intensity characteristics as subcutaneous fat on all imaging sequences (Fig. 60-17). On both CT and MRI, lipomas have a homogeneous appearance, although occasional thin fibrous septa may be present.74,75,102

Vascular Anomalies Vascular anomalies include vascular tumors (infantile heman­ giomas, congenital hemangiomas) and vascular malformations.  Vascular malformations are divided into slow-flow vascular malformations (venous malformations, lymphatic malformations, capillary malformations), fast-flow vascular malformations (arteriovenous malformations/fistulas), and combined complex vascular malforma-

A

B

2085

tions (capillary-venous, capillary-arteriovenous, lymphaticovenous malformations).36,131 On MRI, hemangiomas have a signal intensity similar to that of muscle on T1-weighted images and high signal intensity on T2-weighted images. The signal is usually heterogeneous with T2 weighting and may be either homogeneous or inhomogeneous with T1 weighting; there may be variable amounts of fat and vessels within the mass.92 Phleboliths are seen as focal areas of low signal intensity on all imaging sequences, but they are better identified on plain radiographs or CT (Fig. 60-18). Serpiginous areas of low signal intensity are occasionally seen and represent flow voids in larger feeding arteries or draining veins. On MRI, arteriovenous malformations and arteriovenous fistulas have large tortuous vessels with flow voids due to rapidly flowing blood. Feeding and draining vessels can be identified as vascular structures, but veins and arteries cannot always be differentiated.62,92,112,140 After administration of contrast, hemangiomas and arterial malformations and fistulas enhance avidly, venous malformations have a variable enhancement, and lymphatic malformations do not enhance.36,131

Benign Nerve Sheath Tumors Benign nerve sheath tumors are divided into two groups: neurofibromas and schwannomas (neurilemomas). The peak incidence of benign nerve sheath tumors is between 20 and 30 years of age, with no sex predilection. On CT and MRI, visualization of a nerve entering or exiting the mass is strongly suggestive of a nerve sheath tumor. Another characteristic of nerve sheath tumors is the presence on T2-weighted images of the target sign, consisting of low to intermediate signal intensity centrally, with a ring of high signal intensity peripherally. Contrast enhancement on CT or MRI is variable and is commonly either heterogeneous and diffuse or peripheral. Lesions demonstrating the target

C

FIG 60-17  Intramuscular lipoma in an 80-year-old man with a palpable thigh mass. Coronal T1-weighted (A), sagittal T2-weighted fat-saturated (B), and coronal T1-weighted fat-saturated postgadolinium (C) MRI through the distal left femur, showing a well-defined ovoid mass within the vastus medialis. The mass has homogeneously high signal on T1, drops out on T1 fat-saturated images, and is hypointense on T2 fat-saturated images, with thin internal septations. No thickened or nodular areas of nonfatty septa are seen on postgadolinium images.

2086

PART II  CT and MR Imaging of the Whole Body

A

C

B GE L9 2

1

–2 cm/s

E D FIG 60-18  Venous malformation in a 10-year-old girl with a palpable soft tissue mass in her knee. A, AP knee radiograph. Coronal T1-weighted (B), STIR (C), and T1-weighted fat-saturated postgadolinium (D) MRI of the knee. Images show a soft tissue mass with phleboliths medial to the femoral metaphysis, corresponding to the ovoid lobulated mass, with intermediate signal on T1 and tiny areas of high signal, corresponding to fat. Heterogeneous high signal on T2 with areas of signal dropout correspond to the phleboliths (arrows). Multiple flow voids are also seen. E, Sonographic color Doppler image through the lesion demonstrates serpiginous areas of hypoechogenicity with scattered areas of flow.

sign on T2-weighted images usually enhance more prominently centrally, with a reversal of the target sign53,97,142,143 (Fig. 60-19).

Desmoid Tumor Desmoid tumor is composed of fibroblasts. Histologically the tumor is benign and does not metastasize, but it is locally aggressive, infiltrates contiguous structures, and tends to recur after resection. Desmoid tumors may occur at any age but are most often seen in the third and fourth decades. On cross-sectional imaging desmoids have well-defined borders in two thirds of cases; in the rest, the margins are infiltrative and poorly defined. CT scans obtained without contrast enhancement show  variable attenuation relative to muscle. After intravenous contrast

administration, desmoid tumors may or may not enhance. Most are hyperdense or isodense with muscle on contrast-enhanced scans. Desmoid tumors have a signal intensity lower than that of muscle on T1-weighted MRI and a variable signal intensity on T290,138 (Fig. 60-20). They may have an aggressive appearance and be confused with malignant tumors.10,22 The presence of areas of low signal intensity on both T1- and T2-weighted images, resulting from fibrous tissue, is a clue to the correct diagnosis.23,98

Intramuscular Myxoma Intramuscular myxoma is a rare benign tumor that contains myxoid tissue. It occurs in people between 40 and 70 years of age. Usually it is solitary and has a slow growth rate. Multiple myxomas have been seen

A

B

D

E

C

FIG 60-19  Neurofibromatosis type 1 in a 14-year-old girl. Coronal T1-weighted (A), T2-weighted fat-saturated (B), and T1-weighted fat-saturated postgadolinium (C) MRI of the pelvis demonstrates multiple paraspinal and pelvic ovoid round masses that are isointense on T1, peripherally high and centrally low signal on T2, and the reverse on postgadolinium images, representing the “target” sign (arrows in B and C). A solitary neurofibroma in the medial thigh is seen on coronal T1-weighted (D) and axial T2-weighted (E) MRI, demonstrating the “split fat” and “target” signs.

A

C

B

D

E

FIG 60-20  Aggressive fibromatosis in a 28-year-old woman. Axial T1-weighted (A), T2-weighted (B), T1-weighted fat-saturated postgadolinium (C), and sagittal T2-weighted fat-saturated (D and E) MRI through the posterior thigh. Images demonstrate elongated, infiltrative, dense, collagenous, musculoaponeurotic bands (arrows) of tissue extending from adjacent to the greater trochanter (cellular component [arrows]) and crossing the knee joint (arrows).

2088

PART II  CT and MR Imaging of the Whole Body

in patients with fibrous dysplasia. The tumor is usually localized in a large muscle of the thigh, shoulder, or hip. The CT appearance is a well-defined homogeneous mass with a density lower than that of muscle. The mass typically does not enhance after administration of contrast material. It has low signal intensity  on T1-weighted MRI and high signal intensity on T2; most lesions  have peripheral enhancement, and 55% have heterogeneous internal enhancement6,47,100,111,153 (Fig. 60-21).

Pigmented Villonodular Synovitis Pigmented villonodular synovitis is a benign proliferative process of the synovial lining of a joint, bursa, or tendon sheath. The cause of this lesion is unknown. Repeated trauma, repeated intraarticular hemorrhage, and inflammation have been suggested as causes.

A

There are two forms. The focal form (also called giant cell tumor of the tendon sheath) is more common and usually involves the tendon sheaths of the hands. The diffuse form involves large joints, especially the knee. There is no sex predilection. The peak incidence is between 20 and 40 years of age. MRI characteristics are typical and often lead to the correct  diagnosis. As a result of the deposition of hemosiderin within the tumor, the lesion has areas of low signal intensity on both T1- and T2-weighted images66,109 (Fig. 60-22). The presence of bone erosion on both sides of the joint also suggests the diagnosis. Joint effusion may be present. On CT, bone erosions have well-defined sclerotic borders. On noncontrast scans, the soft tissue component may contain areas of high attenuation; these correspond to deposits of hemosiderin.115

C

B

FIG 60-21  Intramuscular myxoma in a 49-year-old woman with a palpable left thigh lump. Axial T1-weighted (A), T2-weighted fat-saturated (B), and T1-weighted fat-saturated postgadolinium (C) MRI through the left thigh shows a well-defined soft tissue mass in the vastus medialis/lateralis. The lesion has homogeneously high signal on T2 and intermediate signal on T1, with a thin rim and nodular areas of contrast enhancement.

A

B FIG 60-22  Pigmented villonodular synovitis in a 29-year-old woman with knee pain. Sagittal T1-weighted (A) and T2-weighted fat-saturated (B) MRI of the knee demonstrates nodular areas of heterogeneous low signal intensity representing hemosiderin deposition in Hoffa’s fat pad, extending through the intercondylar notch into the posterior knee joint (arrows), also with erosion of the inferior patellar pole.

CHAPTER 60  Musculoskeletal Tumors Malignant Soft Tissue Tumors Liposarcoma

Liposarcoma is a malignant soft tissue neoplasm that contains cells with lipoblastic or lipocytic differentiation. It is the most common malignant soft tissue tumor. Its peak incidence is between 50 and 60 years of age; it is rare before age 20. Liposarcomas are most common in the extremities, especially the thigh, and in the retroperitoneum.95 Liposarcomas can be divided into four groups on the basis of their histologic characteristics: well differentiated, dedifferentiated, myxoid, and pleomorphic.27 The World Health Organization Committee for the Classification of Soft Tissue Tumors in 2002 recognized that atypical lipomatous tumor and well-differentiated liposarcoma are essentially synonymous. The CT or MRI appearance depends on the degree of differentiation.67,74 Portions of the tumor that contain fat demonstrate a low attenuation coefficient on CT scans and high signal intensity on T1-weighted MRI. Nonlipomatous portions of the tumor have a higher attenuation coefficient than fat on CT scans; their signal intensity is similar to that of muscle on T1 (Figs. 60-23 and 60-24) and is higher than that of fat on T2.95 More aggressive liposarcomas may contain no fat that is detectable by either CT or MRI. In these cases the tumor is indistinguishable from other malignant soft tissue tumors.

Undifferentiated Pleomorphic Sarcoma Undifferentiated pleomorphic sarcoma (UPS, previously known as malignant fibrous histiocytoma [MFH]) is a heterogeneous group of disorders. It is mainly composed of mesenchymal lineages that belong to established sarcoma subgroups such as liposarcoma, fibrosarcoma, and leiomyosarcoma. It was considered to be the most common malig-

A

B

2089

nant neoplasm of soft tissues, but with advances in immunohistochemistry and gene expression analysis, most of these tumors have been reclassified as myxofibrosarcoma, leiomyosarcoma, and liposarcoma.70 It is seen more often in men, with a peak incidence between 50 and 70 years of age. Fifty percent are localized in the lower extremities, 20% in the upper extremities, and 20% in the abdominal cavity and retroperitoneum. Other locations are rare. On MRI, undifferentiated pleomorphic sarcoma has low signal intensity with T1 weighting and heterogeneous high signal intensity with T2 weighting30,83 (Fig. 60-25). On CT its density is the same as that of muscle, and it frequently contains areas of lower attenuation that correspond to areas of necrosis.96 MRI is better than CT in determining the extent of tumor, but CT is superior in detecting bone involvement and calcifications within the tumor. Calcifications are seen in up to 20% of undifferentiated pleomorphic sarcoma.25

Synovial Sarcoma Synovial sarcoma is a malignant soft tissue tumor that develops from undifferentiated mesenchymal cells; despite its name, it does not arise from synovium. It is relatively common, following liposarcoma and rhabdomyosarcoma in prevalence. There is a slight male predilection and a peak incidence between 15 and 35 years of age. Less than 10% of synovial cell sarcomas are within a joint. Usually they are adherent to a joint capsule, bursa, fascia, or tendon sheath. They are most common in the lower extremity.5 Synovial sarcoma has low to intermediate signal intensity on T1-weighted and heterogeneous high signal intensity on T2-weighted MRI.103 This tumor is often multilocular with internal septation105 (Fig. 60-26), and fluid-fluid levels are occasionally seen.2,139 Calcifications are present in 30% of cases and are best seen on CT.5 The presence of extensive calcification indicates a better prognosis.129,148

C

FIG 60-23  Well-differentiated liposarcoma in a 70-year-old man with a right calf mass. Sagittal T1-weighted (A), T2-weighted fat-saturated (B), and T1-weighted fat-saturated postgadolinium (C) MRI through the left calf shows a large well-defined intramuscular mass that is hyperintense on T1 and drops out on fat-saturated images but contains nodular internal septations that enhance (arrows). The lesion meets criteria for a simple lipoma, but biopsy of non–fat-containing nodular areas demonstrated well-differentiated liposarcoma.

2090

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 60-24  Myxoid liposarcoma in a 42-year-old man with an anterior right leg lump. Sagittal T1-weighted (A), T2-weighted fat-saturated (B), and T1-weighted fat-saturated postgadolinium (C) MRI, as well as axial T2-weighted MRI through the mid right leg demonstrate a high-signal mass on T2-weighted imaging through the extensor hallucis and digitorum longus muscles. Precontrast T1 image has areas of high signal, representing fat (arrows in A). Postcontrast image (C) demonstrates solid irregular enhancement throughout the lesion.

CHAPTER 60  Musculoskeletal Tumors

A

E

C

B

D

F FIG 60-25  Undifferentiated pleomorphic sarcoma in a 58-year-old woman with left thigh pain. A, AP left femur radiograph. Coronal T1-weighted (B), T2-weighted fat-saturated (C), T1-weighted fat-saturated postgadolinium (D) MRI, as well as axial T2-weighted fat-saturated (E) and T1-weighted fat-saturated postgadolinium (F) MRI through the left thigh. Images show a large well-defined mass within the vastus intermedius/lateralis, extending into the inferior gluteus maximus muscle, which is heterogeneously hyperintense on T1 and T2, with heterogeneous enhancement and smaller dark cystic foci and septations. Multiple cystic structures with internal fluid-fluid levels representing hemorrhage can also be seen.

2091

2092

PART II  CT and MR Imaging of the Whole Body

B

A

C

D

E

F

G

FIG 60-26  Synovial sarcoma in a 16-year-old girl with an enlarging ankle mass. Images include an oblique ankle radiograph (A), axial CT images in soft tissue (B) and bone windows (C), axial T1-weighted (D) and T2-weighted fat-saturated (E) MRI, as well as sagittal T1-weighted (F) and T1-weighted fat-saturated post­ gadolinium (G) MRI through the distal tibia/fibula. A solid soft tissue mass can be seen anterior to the distal syndesmosis; the lesion is isointense on T1, heterogeneously hyperintense on T2, and demonstrates heterogeneous enhancement. Adjacent osseous erosions of the fibula and tibia can be seen on all three modalities.

CHAPTER 60  Musculoskeletal Tumors

REFERENCES 1. Altay M, et al: Secondary chondrosarcoma in cartilage bone tumors: Report of 32 patients. J Orthop Sci 12(5):415–423, 2007. 2. Alyas F, et al: Prevalence and diagnostic significance of fluid-fluid levels in soft-tissue neoplasms. Clin Radiol 62(8):769–774, discussion 75, 2007. 3. Amanatullah DF, et al: Giant cell tumor of bone. Orthopedics 37(2):112–120, 2014. 4. Aoki J, et al: MR of enchondroma and chondrosarcoma: Rings and arcs of Gd-DTPA enhancement. J Comput Assist Tomogr 15(6):1011–1016, 1991. 5. Bakri A, et al: Synovial sarcoma: Imaging features of common and uncommon primary sites, metastatic patterns, and treatment response. AJR Am J Roentgenol 199(2):W208–W215, 2012. 6. Bancroft LW, et al: Intramuscular myxoma: Characteristic MR imaging features. AJR Am J Roentgenol 178(5):1255–1259, 2002. 7. Barile A, et al: Musculoskeletal tumours: Preliminary experience with perfusion MRI. Radiol Med 112(4):550–561, 2007. 8. Beaman FD, Jelinek JS, Priebat DA: Current imaging and therapy of malignant soft tissue tumors and tumor-like lesions. Semin Musculoskelet Radiol 17(2):168–176, 2013. 9. Beatrous TE, Choyke PL, Frank JA: Diagnostic evaluation of cancer patients with pelvic pain: Comparison of scintigraphy, CT, and MR imaging. AJR Am J Roentgenol 155(1):85–88, 1990. 10. Berquist TH, et al: Value of MR imaging in differentiating benign from malignant soft-tissue masses: Study of 95 lesions. AJR Am J Roentgenol 155(6):1251–1255, 1990. 11. Bloem JL, et al: Radiologic staging of primary bone sarcoma: MR imaging, scintigraphy, angiography, and CT correlated with pathologic examination. Radiology 169(3):805–810, 1988. 12. Bohndorf K, et al: Magnetic resonance imaging of primary tumours and tumour-like lesions of bone. Skeletal Radiol 15(7):511–517, 1986. 13. Brower AC, Moser RP, Kransdorf MJ: The frequency and diagnostic significance of periostitis in chondroblastoma. AJR Am J Roentgenol 154(2):309–314, 1990. 14. Campanacci M, et al: The value of imaging in the diagnosis and treatment of bone tumors. Eur J Radiol 27(Suppl 1):S116–S122, 1998. 15. Ceberut K, et al: Bursa formation with scapular osteochondroma in hereditary multiple exostosis. J Coll Physicians Surg Pak 23(7):512–514, 2013. 16. Chakarun CJ, et al: Giant cell tumor of bone: Review, mimics,   and new developments in treatment. Radiographics 33(1):197–211, 2013. 17. Choi YY, Kim JY, Yang SO: PET/CT in benign and malignant musculoskeletal tumors and tumor-like conditions. Semin Musculoskelet Radiol 18(2):133–148, 2014. 18. Choi BB, et al: MR differentiation of low-grade chondrosarcoma from enchondroma. Clin Imaging 37(3):542–547, 2013. 19. Cohen EK, et al: Hyaline cartilage-origin bone and soft-tissue neoplasms: MR appearance and histologic correlation. Radiology 167(2):477–481, 1988. 20. Colangeli S, et al: Fluid-fluid levels detected on MRI and mimicking primary aneurismal bone cysts in a case of spinal bone metastasis.   Eur Rev Med Pharmacol Sci 18(1 Suppl):41–43, 2014. 21. Coupal TM, et al: CT-guided percutaneous cryoablation for osteoid osteoma: Initial experience in adults. AJR Am J Roentgenol 202(5):1136– 1139, 2014. 22. Crim JR, et al: Widespread inflammatory response to osteoblastoma: The flare phenomenon. Radiology 177(3):835–836, 1990. 23. Dahlin DC: Caldwell Lecture. Giant cell tumor of bone: Highlights of 407 cases. AJR Am J Roentgenol 144(5):955–960, 1985. 24. Dalinka MK, et al: The use of magnetic resonance imaging in the evaluation of bone and soft-tissue tumors. Radiol Clin North Am 28(2):461–470, 1990. 25. Dalrymple NC, et al: Price of isotropy in multidetector CT. Radiographics 27(1):49–62, 2007.

2093

26. De Coninck T, et al: Dynamic contrast-enhanced MR imaging for differentiation between enchondroma and chondrosarcoma. Eur Radiol 23(11):3140–3152, 2013. 27. Dei Tos AP: Liposarcomas: Diagnostic pitfalls and new insights. Histopathology 64(1):38–52, 2014. 28. Dhillon M, et al: Distal femoral osteochondroma masquerading as   deep vein thrombosis in an adolescent male. J Knee Surg 26(Suppl 1): S11–S15, 2013. 29. Donkol RH, Al-Nammi A, Moghazi K: Efficacy of percutaneous radiofrequency ablation of osteoid osteoma in children. Pediatr Radiol 38(2):180–185, 2008. 30. Dorfman HD, Bhagavan BS: Malignant fibrous histiocytoma of soft tissue with metaplastic bone and cartilage formation: A new radiologic sign. Skeletal Radiol 8(2):145–150, 1982. 31. Drape JL: Advances in magnetic resonance imaging of musculoskeletal tumours. Orthop Traumatol Surg Res 99(1 Suppl):S115–S123, 2013. 32. Dutoit JC, Vanderkerken MA, Verstraete KL: Value of whole body MRI and dynamic contrast enhanced MRI in the diagnosis, follow-up and evaluation of disease activity and extent in multiple myeloma. Eur J Radiol 82(9):1444–1452, 2013. 33. Duvauferrier R, et al: Current role of CT and whole body MRI in multiple myeloma. Diagn Interv Imaging 94(2):169–183, 2013. 34. Ehara S, et al: Peritumoral edema in osteoid osteoma on magnetic resonance imaging. Skeletal Radiol 28(5):265–270, 1999. 35. Eiber M, et al: Performance of whole-body integrated 18F-FDG PET/ MR in comparison to PET/CT for evaluation of malignant bone lesions. J Nucl Med 55(2):191–197, 2014. 36. El-Merhi F, et al: Peripheral vascular tumors and vascular malformations: Imaging (magnetic resonance imaging and conventional angiography), pathologic correlation and treatment options. Int J Cardiovasc Imaging 29(2):379–393, 2013. 37. Fayad LM, Johnson P, Fishman EK: Multidetector CT of musculoskeletal disease in the pediatric patient: Principles, techniques, and clinical applications. Radiographics 25(3):603–618, 2005. 38. Ferrari S, et al: Imaging in bone sarcomas. The chemotherapist’s point of view. Eur J Radiol 82(12):2076–2082, 2013. 39. Ferrer-Santacreu EM, et al: Enchondroma versus low-grade chondrosarcoma in appendicular skeleton: Clinical and radiological criteria. J Oncol 2012. doi: 10.1155/2012/437958. 40. Fletcher BD: Response of osteosarcoma and Ewing sarcoma to chemotherapy: Imaging evaluation. AJR Am J Roentgenol 157(4):825– 833, 1991. 41. Frank JA, et al: Detection of malignant bone tumors: MR imaging vs scintigraphy. AJR Am J Roentgenol 155(5):1043–1048, 1990. 42. Fruehwald FX, et al: Magnetic resonance imaging of the lower vertebral column in patients with multiple myeloma. Invest Radiol 23(3):193– 199, 1988. 43. Gaeta M, et al: Magnetic resonance imaging findings of osteoid osteoma of the proximal femur. Eur Radiol 14(9):1582–1589, 2004. 44. Gajewski DA, et al: Differentiating clinical and radiographic features of enchondroma and secondary chondrosarcoma in the foot. Foot Ankle Int 27(4):240–244, 2006. 45. Gamba JL, et al: Computed tomography of axial skeletal osteoid osteomas. AJR Am J Roentgenol 142(4):769–772, 1984. 46. Gao ZH, et al: Preoperative easily misdiagnosed telangiectatic osteosarcoma: Clinical-radiologic-pathologic correlations. Cancer Imaging 13(4):520–526, 2013. 47. Gavriilidis P, et al: Intramuscular myxoma of the soleus muscle: A rare tumor in an unusual location. Am J Case Rep 15:49–51, 2014. 48. Geirnaerdt MJ, et al: Cartilaginous tumors: Fast contrast-enhanced MR imaging. Radiology 214(2):539–546, 2000. 49. Ghanem N, et al: Diagnostic value of MRI in comparison to scintigraphy, PET, MS-CT and PET/CT for the detection of metastases of bone. Eur J Radiol 55(1):41–55, 2005. 50. Giles SL, et al: Whole-body diffusion-weighted MR imaging for assessment of treatment response in myeloma. Radiology 271(3):785– 794, 2014.

2094

PART II  CT and MR Imaging of the Whole Body

51. Gillespy T, 3rd, et al: Staging of intraosseous extent of osteosarcoma: Correlation of preoperative CT and MR imaging with pathologic macroslides. Radiology 167(3):765–767, 1988. 52. Ginaldi S, deSantos LA: Computed tomography in the evaluation of small round cell tumors of bone. Radiology 134(2):441–446, 1980. 53. Glasier CM, Williamson MR, Lange TA: MRI of peripheral neurofibromas in children. Orthopedics 12(2):269–272, 1989. 54. Gocmen S, et al: Peripheral nerve injuries due to osteochondromas: Analysis of 20 cases and review of the literature. J Neurosurg 120(5):1105–1112, 2014. 55. Golfieri R, et al: Primary bone tumors. MR morphologic appearance correlated with pathologic examinations. Acta Radiol 32(4):290–298, 1991. 56. Goodwin RW, O’Donnell P, Saifuddin A: MRI appearances of common benign soft-tissue tumours. Clin Radiol 62(9):843–853, 2007. 57. Holscher HC, et al: The value of MR imaging in monitoring the effect of chemotherapy on bone sarcomas. AJR Am J Roentgenol 154(4):763– 769, 1990. 58. Holscher HC, et al: Can MRI predict the histopathological response in patients with osteosarcoma after the first cycle of chemotherapy? Clin Radiol 50(6):384–390, 1995. 59. Hudson TM: Fluid levels in aneurysmal bone cysts: A CT feature. AJR Am J Roentgenol 142(5):1001–1004, 1984. 60. Hudson TM, Hawkins IF, Jr: Radiological evaluation of chondroblastoma. Radiology 139(1):1–10, 1981. 61. Hur J, et al: Efficacy of multidetector row computed tomography of the spine in patients with multiple myeloma: Comparison with magnetic resonance imaging and fluorodeoxyglucose-positron emission tomography. J Comput Assist Tomogr 31(3):342–347, 2007. 62. Hyodoh H, et al: Peripheral vascular malformations: Imaging, treatment approaches, and therapeutic issues. Radiographics 25(Suppl 1):S159–S171, 2005. 63. James SL, Davies AM: Giant-cell tumours of bone of the hand and wrist: A review of imaging findings and differential diagnoses. Eur Radiol 15(9):1855–1866, 2005. 64. Jee WH, et al: Nonossifying fibroma: Characteristics at MR imaging with pathologic correlation. Radiology 209(1):197–202, 1998. 65. Jee WH, et al: Chondroblastoma: MR characteristics with pathologic correlation. J Comput Assist Tomogr 23(5):721–726, 1999. 66. Jelinek JS, et al: Imaging of pigmented villonodular synovitis with emphasis on MR imaging. AJR Am J Roentgenol 152(2):337–342, 1989. 67. Jelinek JS, et al: Liposarcoma of the extremities: MR and CT findings in the histologic subtypes. Radiology 186(2):455–459, 1993. 68. Kaim AH, et al: Chondroblastoma and clear cell chondrosarcoma: Radiological and MRI characteristics with histopathological correlation. Skeletal Radiol 31(2):88–95, 2002. 69. Keenan S, Bui-Mansfield LT: Musculoskeletal lesions with fluid-fluid level: A pictorial essay. J Comput Assist Tomogr 30(3):517–524, 2006. 70. Kelleher FC, Viterbo A: Histologic and genetic advances in refining the diagnosis of “undifferentiated pleomorphic sarcoma”. Cancers (Basel) 5(1):218–233, 2013. 71. Kloth JK, et al: Appearance of monoclonal plasma cell diseases in whole-body magnetic resonance imaging and correlation with parameters of disease activity. Int J Cancer 2014. 72. Kransdorf MJ, Peterson JJ, Bancroft LW: MR imaging of the knee: Incidental osseous lesions. Radiol Clin North Am 45(6):943–954, v, 2007. 73. Kransdorf MJ, et al: MR appearance of fibroxanthoma. J Comput Assist Tomogr 12(4):612–615, 1988. 74. Kransdorf MJ, et al: Fat-containing soft-tissue masses of the extremities. Radiographics 11(1):81–106, 1991. 75. Kransdorf MJ, et al: Imaging of fatty tumors: Distinction of lipoma and well-differentiated liposarcoma. Radiology 224(1):99–104, 2002. 76. Kroon HM, Schurmans J: Osteoblastoma: Clinical and radiologic findings in 98 new cases. Radiology 175(3):783–790, 1990. 77. Lee KC, Davies AM, Cassar-Pullicino VN: Imaging the complications of osteochondromas. Clin Radiol 57(1):18–28, 2002.

78. Lee EH, Shafi M, Hui JH: Osteoid osteoma: A current review. J Pediatr Orthop 26(5):695–700, 2006. 79. Lemmi MA, et al: Use of MR imaging to assess results of chemotherapy for Ewing sarcoma. AJR Am J Roentgenol 155(2):343–346, 1990. 80. Leung JC, Dalinka MK: Magnetic resonance imaging in primary bone tumors. Semin Roentgenol 35(3):297–305, 2000. 81. Liu PT, et al: Imaging of osteoid osteoma with dynamic gadoliniumenhanced MR imaging. Radiology 227(3):691–700, 2003. 82. Liu NB, et al: Diagnostic value of 18F-FDG PET/CT in comparison to bone scintigraphy, CT and 18F-FDG PET for the detection of bone metastasis. Asian Pac J Cancer Prev 14(6):3647–3652, 2013. 83. Mahajan H, et al: Magnetic resonance imaging of malignant fibrous histiocytoma. Magn Reson Imaging 7(3):283–288, 1989. 84. Mahnken AH, et al: Multidetector CT of the spine in multiple myeloma: Comparison with MR imaging and radiography. AJR Am J Roentgenol 178(6):1429–1436, 2002. 85. Mahnken AH, et al: Aneurysmal bone cyst: Value of MR imaging and conventional radiography. Eur Radiol 13(5):1118–1124, 2003. 86. Mail JT, et al: Response of osteosarcoma to preoperative intravenous high-dose methotrexate chemotherapy: CT evaluation. AJR Am J Roentgenol 144(1):89–93, 1985. 87. Mannem RR, et al: AIRP best cases in radiologic-pathologic correlation: Intraosseous lipoma. Radiographics 32(5):1523–1528, 2012. 88. Maras Ozdemir Z, et al: Adventitious bursitis overlying an osteochondroma of the humerus facing the thoracic wall. Case Rep Radiol 2013:939372, 2013. 89. Margau R, et al: MR imaging of simple bone cysts in children: Not so simple. Pediatr Radiol 30(8):551–557, 2000. 90. McCarville MB, et al: MRI and biologic behavior of desmoid tumors in children. AJR Am J Roentgenol 189(3):633–640, 2007. 91. Mehta M, et al: MR imaging of symptomatic osteochondromas with pathological correlation. Skeletal Radiol 27(8):427–433, 1998. 92. Meyer JS, et al: Biological classification of soft-tissue vascular anomalies: MR correlation. AJR Am J Roentgenol 157(3):559–564, 1991. 93. Moulopoulos LA, et al: Prognostic significance of magnetic resonance imaging in patients with asymptomatic multiple myeloma. J Clin Oncol 13(1):251–256, 1995. 94. Moulopoulos LA, et al: Prognostic significance of magnetic resonance imaging of bone marrow in previously untreated patients with multiple myeloma. Ann Oncol 16(11):1824–1828, 2005. 95. Murphey MD, Arcara LK, Fanburg-Smith J: From the archives of the AFIP: Imaging of musculoskeletal liposarcoma with radiologicpathologic correlation. Radiographics 25(5):1371–1395, 2005. 96. Murphey MD, et al: Enchondroma versus chondrosarcoma in the appendicular skeleton: Differentiating features. Radiographics 18(5):1213–1237, quiz 44-5, 1998. 97. Murphey MD, et al: From the archives of the AFIP. Imaging of musculoskeletal neurogenic tumors: Radiologic-pathologic correlation. Radiographics 19(5):1253–1280, 1999. 98. Murphey MD, et al: Imaging of osteochondroma: Variants and complications with radiologic-pathologic correlation. Radiographics 20(5):1407–1434, 2000. 99. Murphey MD, et al: From the archives of AFIP. Imaging of giant cell tumor and giant cell reparative granuloma of bone: Radiologicpathologic correlation. Radiographics 21(5):1283–1309, 2001. 100. Murphey MD, et al: Imaging of soft-tissue myxoma with emphasis on CT and MR and comparison of radiologic and pathologic findings. Radiology 225(1):215–224, 2002. 101. Murphey MD, et al: From the archives of the AFIP: Imaging of primary chondrosarcoma: Radiologic-pathologic correlation. Radiographics 23(5):1245–1278, 2003. 102. Murphey MD, et al: From the archives of the AFIP: Benign musculoskeletal lipomatous lesions. Radiographics 24(5):1433–1466, 2004. 103. Murphey MD, et al: From the archives of the AFIP: Imaging of synovial sarcoma with radiologic-pathologic correlation. Radiographics 26(5):1543–1565, 2006.

CHAPTER 60  Musculoskeletal Tumors 104. Murphey MD, et al: From the radiologic pathology archives: Ewing sarcoma family of tumors: Radiologic-pathologic correlation. Radiographics 33(3):803–831, 2013. 105. Nakanishi H, et al: Cystic synovial sarcomas: Imaging features with clinical and histopathologic correlation. Skeletal Radiol 32(12):701–707, 2003. 106. Orguc S, Arkun R: Primary tumors of the spine. Semin Musculoskelet Radiol 18(3):280–299, 2014. 107. Pan G, et al: Osteosarcoma: MR imaging after preoperative chemotherapy. Radiology 174(2):517–526, 1990. 108. Panicek DM, et al: CT and MR imaging in the local staging of primary malignant musculoskeletal neoplasms: Report of the Radiology Diagnostic Oncology Group. Radiology 202(1):237–246, 1997. 109. Papp DF, et al: Magnetic resonance imaging of soft-tissue tumors: Determinate and indeterminate lesions. J Bone Joint Surg Am 89(Suppl 3):103–115, 2007. 110. Parkkola RK, et al: Dynamic contrast-enhanced MR imaging and MR-guided bone biopsy on a 0.23 T open imager. Skeletal Radiol 30(11):620–624, 2001. 111. Prado MA, et al: Intramuscular myxoma: Differential diagnosis and review of the literature. Orthopedics 25(11):1297–1299, 2002. 112. Rak KM, et al: MR imaging of symptomatic peripheral vascular malformations. AJR Am J Roentgenol 159(1):107–112, 1992. 113. Recht MP, et al: Popliteal artery pseudoaneurysm in a patient with hereditary multiple exostoses: MRI and MRA diagnosis. J Comput Assist Tomogr 17(2):300–302, 1993. 114. Richardson ML, et al: Magnetic resonance imaging of musculoskeletal neoplasms. Radiol Clin North Am 24(2):259–267, 1986. 115. Rosenthal DI, Aronow S, Murray WT: Iron content of pigmented villonodular synovitis detected by computed tomography. Radiology 133(2):409–411, 1979. 116. Rosenthal RE, Wozney P: Diagnostic value of gadopentetate dimeglumine for 1.5-T MR imaging of musculoskeletal masses: Comparison with unenhanced T1- and T2-weighted imaging. J Magn Reson Imaging 1(5):547–551, 1991. 117. Sahin E, et al: Is (99m)Tc-MDP whole body bone scintigraphy adjuvant to (18)F-FDG-PET for the detection of skeletal metastases? J BUON 19(1):291–296, 2014. 118. Sato E, et al: Chondroblastoma with secondary aneurysmal bone cyst of the capitate. Orthopedics 37(5):e500–e503, 2014. 119. Schlesinger AE, Hernandez RJ: Intracapsular osteoid osteoma of the proximal femur: Findings on plain film and CT. AJR Am J Roentgenol 154(6):1241–1244, 1990. 120. Schmidt GP, et al: Whole-body magnetic resonance imaging and positron emission tomography-computed tomography in oncology. Top Magn Reson Imaging 18(3):193–202, 2007. 121. Seeger LL, et al: Preoperative evaluation of osteosarcoma: Value of gadopentetate dimeglumine-enhanced MR imaging. AJR Am J Roentgenol 157(2):347–351, 1991. 122. Shah ZK, et al: Sarcomatous transformation in diaphyseal aclasis. Australas Radiol 51(2):110–119, 2007. 123. Shaikh MI, et al: Spinal osteoblastoma: CT and MR imaging with pathological correlation. Skeletal Radiol 28(1):33–40, 1999. 124. Shimada J, et al: Chondromyxoid fibroma with secondary aneurysmal bone cyst of the rib. Ann Thorac Cardiovasc Surg 19(3):250–252, 2013. 125. Shirkhoda A, et al: Computed tomography of osteosarcoma after intraarterial chemotherapy. AJR Am J Roentgenol 144(1):95–99, 1985. 126. Shuman WP, et al: Comparison of STIR and spin-echo MR imaging at 1.5 T in 45 suspected extremity tumors: Lesion conspicuity and extent. Radiology 179(1):247–252, 1991. 127. Sundaram M, McLeod RA: MR imaging of tumor and tumorlike lesions of bone and soft tissue. AJR Am J Roentgenol 155(4):817–824, 1990. 128. Tan H, et al: Chondroblastoma of the patella with aneurysmal bone cyst. Orthopedics 37(1):e87–e91, 2014. 129. Tateishi U, et al: Synovial sarcoma of the soft tissues: Prognostic significance of imaging features. J Comput Assist Tomogr 28(1):140–148, 2004.

2095

130. Tehranzadeh J, Murphy BJ, Mnaymneh W: Giant cell tumor of the proximal tibia: MR and CT appearance. J Comput Assist Tomogr 13(2):282–286, 1989. 131. Tekes A, et al: S.E. Mitchell Vascular Anomalies Flow Chart (SEMVAFC): A visual pathway combining clinical and imaging findings for classification of soft-tissue vascular anomalies. Clin Radiol 69(5): 443–457, 2014. 132. Toirkens J, et al: A comparison between histopathology and findings on magnetic resonance imaging of subcutaneous lipomatous soft-tissue tumors. Insights Imaging 2(5):599–607, 2011. 133. Tsai JC, et al: Fluid-fluid level: A nonspecific finding in tumors of bone and soft tissue. Radiology 175(3):779–782, 1990. 134. Uhl M, et al: Evaluation of tumour necrosis during chemotherapy with diffusion-weighted MR imaging: Preliminary results in osteosarcomas. Pediatr Radiol 36(12):1306–1311, 2006. 135. Uhl M, et al: Osteosarcoma: Preliminary results of in vivo assessment of tumor necrosis after chemotherapy with diffusion- and perfusionweighted magnetic resonance imaging. Invest Radiol 41(8):618–623, 2006. 136. van den Berg H, et al: Incidence of biopsy-proven bone tumors in children: A report based on the Dutch pathology registration “PALGA”.   J Pediatr Orthop 28(1):29–35, 2008. 137. van der Heijden L, et al: The clinical approach toward giant cell tumor of bone. Oncologist 19(5):550–561, 2014. 138. Vandevenne JE, et al: New concepts in understanding evolution of desmoid tumors: MR imaging of 30 lesions. Eur Radiol 7(7):1013–1019, 1997. 139. Van Dyck P, et al: Prevalence, extension and characteristics of fluid-fluid levels in bone and soft tissue tumors. Eur Radiol 16(12):2644–2651, 2006. 140. van Rijswijk CS, et al: Value of dynamic contrast-enhanced   MR imaging in diagnosing and classifying peripheral vascular malformations. AJR Am J Roentgenol 178(5):1181–1187, 2002. 141. Varma DG, et al: Chondrosarcoma: MR imaging with pathologic correlation. Radiographics 12(4):687–704, 1992. 142. Verstraete KL, et al: Nerve sheath tumors: Evaluation with CT and MR imaging. J Belge Radiol 75(4):311–320, 1992. 143. Wakoh M, et al: Two cases of schwannoma with marked cystic changes. Dentomaxillofac Radiol 34(1):44–50, 2005. 144. Weatherall PT, et al: Chondroblastoma: Classic and confusing appearance at MR imaging. Radiology 190(2):467–474, 1994. 145. Weininger M, et al: Whole-body MRI of multiple myeloma: Comparison of different MRI sequences in assessment of different growth patterns. Eur J Radiol 69(2):339–345, 2009. 146. Widmann G, et al: CT-guided stereotactic targeting accuracy of osteoid osteoma. Int J Med Robot 9(3):274–279, 2013. 147. Wilhelm T, Stieltjes B, Schlemmer HP: Whole-body-MR-diffusion weighted imaging in oncology. Rofo 185(10):950–958, 2013. 148. Wilkerson BW, et al: Characterization of synovial sarcoma calcification. AJR Am J Roentgenol 199(6):W730–W734, 2012. 149. Wilson AJ, Kyriakos M, Ackerman LV: Chondromyxoid fibroma: Radiographic appearance in 38 cases and in a review of the literature. Radiology 179(2):513–518, 1991. 150. Woertler K: Benign bone tumors and tumor-like lesions:   Value of cross-sectional imaging. Eur Radiol 13(8):1820–1835, 2003. 151. Woertler K: Soft tissue masses in the foot and ankle: Characteristics   on MR imaging. Semin Musculoskelet Radiol 9(3):227–242, 2005. 152. Yamamura S, et al: Inflammatory reaction in chondroblastoma. Skeletal Radiol 25(4):371–376, 1996. 153. Yao MS, et al: Magnetic resonance imaging of gluteal intramuscular myxoma. Clin Imaging 31(3):214–216, 2007. 154. Yildiz S, et al: Re: MR differentiation of low-grade chondrosarcoma from enchondroma. Clin Imaging 37(6):1149, 2013. 155. Yoo HJ, et al: Differentiating high-grade from low-grade chondrosarcoma with MR imaging. Eur Radiol 19(12):3008–3014, 2009.

61  Shoulder Kenneth Sheah and Miriam A. Bredella

NORMAL ANATOMY Gross Anatomy The glenohumeral joint is relatively shallow, with the humeral head large compared with the glenoid fossa. This configuration grants mobility at the expense of stability. The labrum is a meniscus-like fibrous or fibrocartilaginous structure that increases the depth of the glenoid, providing more contact area and thus more stability for the glenohumeral joint. A synovium-lined recess may be interposed between the labrum and origin of the long head of the biceps brachii muscle. Anterior to this muscle origin and extending to the midglenoid, the labrum may be absent or unattached, with a smooth foramen between itself and the glenoid rim. A sublabral recess or foramen may also be present in this area.36,108,150,185,213,229 Surrounding the glenohumeral joint is the joint capsule, which attaches to both the anatomic neck of the humerus and the scapula. Its scapular attachment is variable, particularly at the midpoint of the glenoid. In this area, the capsule may attach to the anterior labrum, to the glenoid just medial to the labrum, or medially along the glenoid.150,159 The glenohumeral joint capsule is relatively loose, with three normal extensions of the joint capsule where joint fluid may be seen at imaging in patients with an effusion. The first is a communication with the sheath of the long head of the biceps as it runs through the intertubercular groove. The second reaches under the coracoid process to communicate with the subscapular recess. The third is the axillary recess or axillary pouch, a loose fold of thickened capsule at the caudal extent of the joint that tightens when the arm is abducted. Normally, joint fluid is restricted to a thin film, with no distention of these recesses.147,203 Three glenohumeral ligaments are seen as thickened areas contiguous with the joint capsule: 1. The superior glenohumeral ligament arises from the superior glenoid within a centimeter anterior to the biceps origin and runs to the fovea capitis of the humerus. It is the most consistent of the three in its course and is often identified at cross-sectional imaging.37,56,172 2. The middle glenohumeral ligament is the most variable of the three in its course and thickness. It often arises in common with the superior glenohumeral ligament and runs anteriorly in an inferolateral direction to the lesser tuberosity of the humerus.37,56,172 3. The inferior glenohumeral ligament complex consists of distinct anterior and posterior bands separated by the axillary pouch. It runs from a broad-based origin along the middle to inferior labrum to the humeral neck. This complex is the most important ligamentous stabilizer, guarding against both anterior and posterior dislocations.166,172 A number of other ligaments also help stabilize the shoulder. The coracoacromial ligament passes deep to the deltoid and superficial to the

2096

subacromial-subdeltoid bursa. It supports the shoulder against upward pressure.147 The coracohumeral ligament runs from the lateral aspect of the base of the coracoid to insert on the humerus adjacent to the greater tuberosity, where its fibers blend at its insertion site with those of the joint capsule. It passively supports the humerus when the arm hangs at the side. It is also a major stabilizer of the biceps tendon in its groove, comprising a part of the biceps tendon pulley.147 The coracoclavicular ligament is made up of two parts, the trapezoid ligament and the conoid ligament. The former runs from the coracoid process to the undersurface of the distal clavicle, and the latter runs from the coracoid to the more medially located conoid tubercle of the clavicle.147 The acromioclavicular ligament strengthens the superior portion of the tight-fitting acromioclavicular joint capsule. Both it and the coracoclavicular ligament stabilize the acromioclavicular joint.147,199,218 Four muscles (supraspinatus, infraspinatus, teres minor, subscapularis), together with their tendons, form the rotator cuff. All originate on the scapula and insert on the greater tuberosity of the humerus except for the subscapularis, which attaches to the lesser tuberosity. They are responsible for abduction and internal and external rotation of the glenohumeral joint. Of these muscles, the supraspinatus is most prone to injury. It arches superiorly over the humerus to insert superolaterally on the greater tuberosity, gliding deep to the acromion. The infraspinatus and teres minor also insert on the greater tuberosity, each a little more inferiorly and posteriorly. The subscapularis has a broad origin on the anterior scapula and inserts with multiple tendon slips onto the lesser tuberosity. The large deltoid muscle extends from a very broad origin on the distal clavicle and acromion to insert into the deltoid tubercle of the lateral aspect of the proximal humeral diaphysis. Occasionally a proximal tendon slip of the deltoid arises from the undersurface of the clavicle or acromion, where it may mimic a spur. The deltoid muscle is external to the rotator cuff and is separated from it by the subacromial-subdeltoid bursa and its associated extrasynovial fat pad. The subacromial-subdeltoid bursa does not normally communicate with the glenohumeral joint.204 The peribursal fat pad has a variable appearance. On magnetic resonance imaging (MRI), part of this fat pad may not be visible, either in normal shoulders or in shoulders with rotator cuff pathology.43,96,143,160,219 The subcoracoid bursa lies caudal to the coracoid process and anterior to the subscapularis muscle. It communicates with the subacromial-subdeltoid bursa about half the time.202 This space is divided by the middle glenohumeral ligament into the foramina of Weitbrecht and Rouvière. Between the greater and lesser tuberosities lies the bicipital groove, also called the intertubercular sulcus. This depression cradles the tendon of the long head of the biceps brachii, which originates from the superior glenoid tubercle and superior labrum. Its proximal tendon passes anteriorly over the humeral head through the rotator interval,

CHAPTER 61  Shoulder an encapsulated space separating the supraspinatus and subscapularis muscles. It then passes through the bicipital groove to the muscle belly. The tendon is secured in its groove superiorly by the coracohumeral ligament, its most important stabilizer, and by the transverse humeral ligament, to which the subscapularis tendon may lend some fibers. More distally the tendon of the pectoralis major also contributes to stabilization of the biceps tendon in its groove.51 The coracoacromial ligament forms part of the coracoacromial arch, which consists of the coracoacromial ligament and the structures it connects—the coracoid process of the scapula and the anterior acromion. This relatively immobile complex not only provides support to the shoulder but may impinge on the tendon of the supraspinatus muscle, causing tendinitis and contributing to an eventual tear, especially where the coracoid process is impinging on the lesser tuberosity in flexion, adduction, and internal rotation (coracoid impingement syndrome). Three neurovascular structures are important to the imaging anatomy of the shoulder. Running on the posterior aspect of the scapula are the suprascapular nerve and artery. The nerve arises from the upper trunk of the brachial plexus (C5 and C6) and passes through the suprascapular notch. It then runs deep to the supraspinatus muscle. When it emerges from deep to the supraspinatus muscle, it passes through the spinoglenoid notch. It carries both sensory and motor fibers and supplies motor innervation to the supraspinatus and infraspinatus muscles.62,147 The anterolateral branch of the anterior circumflex humeral artery runs for a short distance next to the tendon of the long head of the biceps in the bicipital groove. This small vessel can be mistaken for a bicipital tendon disorder, particularly on gradient echo sequences in which arterial flow appears hyperintense.147

Cross-sectional Anatomy The following paragraphs describe the anatomy of the shoulder from the viewpoint of cross-sectional imaging in the four planes often obtained with MRI: • Axial • Coronal oblique • Sagittal oblique • Abduction, external rotation (ABER)

Axial Projection (Fig. 61-1).  At the level of the acromioclavicular joint, seen in Figure 61-1A, the acromion curves in on the lateral side of the shoulder to meet the distal end of the clavicle. Edema and cystic changes can be seen in the subarticular portions of the joint, as can the presence of an os acromiale. Figure 61-1B shows the level of the top of the humeral head. The supraspinatus muscle arches across the humeral head from posteromedial to anterolateral. A large principal tendon arises from the anterior fusiform portion of the muscle. Small individual tendon slips arise from the posterior supraspinatus muscle, which morphologically resembles a strap muscle. The course of the main tendon is at a slightly different angle from that of the muscle and curves anteriorly toward its attachment site on the greater tuberosity.236,237 The deltoid forms a thick semicircle of muscle external to the supraspinatus. The scapular spine can be noted posteriorly as it runs laterally and superiorly to form the acromion. In Figure 61-1C, at the level of the top of the coracoid process of the scapula, the humeral head normally has a round shape. The superior aspect of the glenoid is usually seen here. The supraspinatus muscle projects between the glenoid and the inferior extension of the scapular spine. The infraspinatus tendon emerges anteriorly to attach to the greater tuberosity. Running from posterior to anterior is the superior glenohumeral ligament.

2097

In Figure 61-1D, closer to the base of the coracoid, the superior labrum becomes identifiable. A Hill-Sachs deformity, if present, can be identified at this level or higher. Glenohumeral joint cartilage can be followed along the joint margins. In Figure 61-1E, the normal posterior labrum has a relatively constant blunt triangular shape. The anterior labrum may vary considerably in shape. It may appear either round or pointed and may exhibit longitudinal or transverse clefts.86,129,134,159 Both anteriorly and posteriorly, the articular cartilage comes to the edge of the glenoid, and this forms an area of increased signal intensity that undercuts the internal edge of the labrum and must not be mistaken for a tear.115 At this level the humeral head has an oblong shape with two indentations. Anteriorly the greater and lesser tuberosities surround the bicipital groove in which rests the tendon of the long head of the biceps. Posteriorly a normal small concavity or flattening of the humerus can be seen but should not be mistaken for a Hill-Sachs lesion, which is usually located superiorly.205 Other structures routinely visualized here include the anterior joint capsule, the middle glenohumeral ligament, and the subscapularis muscle and tendon. The tendon has  a broad insertion on the lesser tuberosity, and partial-thickness articular surface tears of the cranial aspect of the tendon are sometimes visualized. Biceps tendon subluxation is associated with tears of the  subscapularis tendon. The pectoralis major muscle tendon is also demonstrated and should be scrutinized for evidence of myotendinous junction strain or tear. In Figure 61-1F, the inferior labrum can be visualized, as well as the biceps tendon and the proximal humeral metaphysic. The glenoid neck is also seen.

Coronal Oblique Projection (Fig. 61-2).  In Figure 61-2A, the subscapularis muscle runs anterior to the scapular blade and inferior to the base of the coracoid to attach to the lesser tuberosity. Inferior to this are the flow voids of axillary vessels in the quadrilateral space. Superiorly the anteriormost extent of the clavicle is the origin of a portion of the deltoid. In Figure 61-2B, more posteriorly, the anterior humeral head  may be seen with the greater and lesser tuberosities and the  bicipital tendon in its groove. The coracohumeral ligament runs at  an angle with respect to this plane of imaging, but portions of it may be seen. In Figure 61-2C, at the midglenoid, the glenohumeral articulation is well seen. It is best visualized in the coronal projection at the point at which the glenoid is most prominent. Also well seen at this level is the tendon of the supraspinatus and the peribursal fat stripe. A portion of the inferior glenohumeral ligament is seen below the inferior labrum. In Figure 61-2D, far posterior in the shoulder, the infraspinatus muscle and tendon pass deep to the acromion to insert on the greater tuberosity. Depending on the degree of internal or external rotation, this attachment site may not always be visualized in the same plane as the muscle.

Sagittal Oblique Projection (Fig. 61-3).  In Figure 61-3A, the four rotator cuff tendons blend at their insertion sites into a fibrous band that appears on MRI as an arc of hypointense signal draped across the humeral head, forming the rotator crescent. The deltoid muscle projects superficial to the rotator cuff. The fused physeal plate is often visible in adults. In Figure 61-3B, at the level of the medial humeral head, the biceps tendon is seen close to the humeral head in the rotator interval between the subscapularis and the supraspinatus. It is bounded by the biceps tendon pulley. The coracoacromial ligament arches anteriorly and

2098

PART II  CT and MR Imaging of the Whole Body

23 5 8

1

3

7

6

2

10

24

8

8

7

13

23

14 12 4

9

9

C

B

A

19

21 20

8 24 16

7

8 22

24 20 19

21

24

23 19

D

21

22

7 18

15

20

25

17

11

11

E

F

FIG 61-1  A to F, MRI axial views of the shoulder (see text for details). 1, acromion; 2, clavicle; 3, acromioclavicular joint; 4, supraspinatus muscle; 5, anterior main tendon; 6, posterior tendon slip; 7, humeral head; 8, deltoid muscle; 9, scapular spine; 10, coracoid process; 11, glenoid; 12, infraspinatus muscle/tendon; 13, superior glenohumeral ligament; 14, superior labrum; 15, posterior capsule; 16, anterior labrum; 17, posterior labrum; 18, glenoid articular cartilage; 19, greater tuberosity; 20, lesser tuberosity; 21, bicipital groove; 22, middle glenohumeral ligament; 23, bicipital tendon; 24, subscapularis muscle and tendon; 25, inferior labrum.

inferiorly from the inferior margin of this slightly hooked acromion. It shows low signal intensity on all MRI sequences. In Figure 61-3C, slightly more medial, the lateral aspect of the coracoid process becomes visible. The coracoid is a reliable marker for the anterior part of the shoulder. In Figure 61-3D, medial to the humeral head, the anterior labrum is seen. The superior and middle glenohumeral ligaments are seen. The foramina of Weitbrecht and Rouvière are visualized, representing recesses of the subcoracoid bursa. In Figure 61-3E, the posterior and inferior labra are demonstrated. The biceps anchor is located at this level but is better visualized on coronal sections.

In Figure 61-3F, farther medially, the coracoid process arises from its base on the scapula. One of the coracoclavicular ligaments projects between the coracoid and the clavicle.

Abduction, External Rotation Positioning (Fig. 61-4).  In Figure 61-4A, a coronal scout image allows sections to be obtained  perpendicular to the glenohumeral joint and parallel to the  humeral shaft, running in this case from the inferior to the superior glenoid. In Figure 61-4B, at the level of the inferior labrum, the posterior joint capsule demonstrates laxity as the anterior capsule is placed under tension by this positioning technique.

CHAPTER 61  Shoulder

10

12 13 5

12

7

5 8

8

2

1 15

3

B

A

4

10 12 5 13 11

7

16

6 9

5 14 15 15 3 3 8

C

D FIG 61-2  A to D, MRI coronal oblique views of the shoulder (see text for details). 1, subscapularis muscle; 2, coracoid process; 3, axillary vessels; 4, clavicle; 5, deltoid; 6, supraglenoid fossa; 7, greater tuberosity; 8, bicipital tendon; 9, glenoid cartilage; 10, acromion; 11, humeral articular cartilage; 12, supraspinatus muscle and tendon; 13, superior labrum; 14, inferior labrum; 15, inferior glenohumeral ligament/axillary recess; 16, infraspinatus muscle/tendon.

2099

2100

PART II  CT and MR Imaging of the Whole Body

3 6

9

7 2

6

4 2

3

3

5

3

12

5

10

7

4

8

C

B

A

9 11

9 6 S

10

10 M

M 4

6

6

S

10

12

12

AL

7

7

PL

4

7

4

8

IL I

D

E

F

FIG 61-3  A to F, MRI sagittal oblique views of the shoulder (see text for details). 1, rotator cuff; 2, humeral head; 3, deltoid muscle; 4, subscapularis; 5, bicipital tendon; 6, supraspinatus; 7, infraspinatus; 8, teres minor; 9, acromion; 10, coracoid process; 11, clavicle; 12, scapular spine (root of acromion). AL, anterior labrum; I, inferior glenohumeral ligament; IL, inferior labrum; M, middle glenohumeral ligament; PL, posterior labrum. S, superior glenohumeral ligament.

In Figure 61-4C, as in other externally rotated sequences, the anterior labrum assumes a pointed triangular shape. The posterior labrum becomes rounded and globular. As the humerus rotates up into position, the medial aspect of the bone now faces anteriorly. The concave shape of the metaphysis is typical of the medial upper humerus. In Figure 61-4D, at approximately the midglenoid, the anterior band of the inferior glenohumeral ligament is visible and should appear taut at this level. In Figure 61-4E and F, at the level of the superior glenoid, the course of the tendon of the long head of the biceps brachii is demonstrated. Also, the undersurface of the supraspinatus is seen.

IMAGING TECHNIQUES Table 61-1 lists the protocols for MRI and computed tomography (CT) of the shoulder at our institution. It is important to stabilize the shoulder to reduce motion artifact. Motion arises from a combination of respiratory and small voluntary movements by the patient. When the coil allows it, this small amount of motion may be effectively suppressed by placing a sandbag on the shoulder. Some “sand” bags actually contain ferromagnetic metallic shot, and caution should be exercised before deploying them in the MRI scanning room.

2101

CHAPTER 61  Shoulder

1

1 6

5

2 10

8 5

A

6

2

9

9 7

7

3

8

3

C

B

1 1

9 6

2

8

10 7

11

6

11

2 2 3

4

3

12

10

3

5

D

E

F

FIG 61-4  A to F, MRI abduction, external rotation (ABER) oblique axial views of the shoulder. 1, humeral shaft; 2, humeral epiphysis; 3, glenoid; 4, redundant posterior capsule; 5, taut anterior capsule; 6, anterior humeral cortex; 7, anterior labrum; 8, posterior labrum; 9, posterior rotator cuff; 10, inferior glenohumeral ligament; 11, biceps tendon, long head; 12, coracoid process.

When performing gadolinium MRI shoulder arthrography, palpation, fluoroscopy,180 or ultrasonographic guidance238 can be used. The dilution of gadolinium chelate in a mixture of saline or mixed iodinated contrast with saline is 1 : 200.105 We use T1-weighted spin echo or gradient echo axial, coronal oblique, and sagittal oblique images with fat saturation. It is important to include at least one set of T2-weighted images, usually in the coronal oblique plane. This allows identification of intrasubstance rotator cuff disease and fluid collections not in communication with the glenohumeral joint. A fourth imaging plane, apart from the traditional oblique orthogonal planes, with MR arthrography is the oblique axial plane, with the patient’s shoulder in ABER.40,109,137,227 The patient places the hand under the head, with the palm up. A wraparound coil is required to

achieve adequate signal-to-noise ratio. A coronal localizing sequence then allows images to be obtained perpendicular to the glenohumeral joint and parallel to the shaft of the humerus. For CT of the shoulder with or without arthrography, both soft tissue and bone algorithms should be used.210

ROTATOR CUFF DISEASE, IMPINGEMENT, AND INSTABILITY Rotator Cable–Crescent Complex The rotator cable is a fibrous thickening of the articular surface of the distal supraspinatus tendon (running in a perpendicular direction to the supraspinatus tendon fiber orientation), located in the anterior 8

2102

PART II  CT and MR Imaging of the Whole Body

TABLE 61-1  Protocols for MRI and CT of the Shoulder Sequence

Slice Thickness

TR/TE (msec)

Field-of-View (mm)

Matrix Size (pixels)

Nonarthrographic MRI Axial gradient echo (flip angle 70 degrees) Oblique coronal proton density Oblique coronal T2 FSE with fat saturation Oblique coronal T2 FSE Oblique sagittal T1 Oblique sagittal T2 FSE with fat saturation Oblique sagittal T2 FSE with fat saturation

3 mm 3 mm 3 mm 3 mm 4 mm 4 mm 5 mm

717/17 2650/23 3017/50 3800/98 533/9 3000/45 5380/51

160 × 160 160 × 160 160 × 160 160 × 160 160 × 160 140 × 140 160 × 160

192 × 320 230 × 384 224 × 320 224 × 320 256 × 320 224 × 320 or 320 × 256

MR Arthrogram Axial T1 with fat saturation Oblique coronal T1 with fat saturation Oblique coronal T2 FSE with fat saturation Oblique sagittal T1 with fat saturation Oblique sagittal T1 FSE (without fat saturation) Oblique axial T1 ABER

3 mm 3 mm 3 mm 3 mm 3 mm 4 mm

565/10 600/12 3500/75 561/11 561/11 600/10

160 × 160 160 × 160 160 × 160 160 × 160 160 × 160 160 × 160

240 × 320 230 × 384 224 × 320 230 × 384 230 × 384 192 × 320

CT (with or without Arthrogram) Pitch

Kilovoltage and Milliamperage

Field of View

Slice Thickness (Reconstructed)

Table Speed

0.94 : 1

140, 198

220 mm

1.25 mm

9.37 mm/rotation

ABER, abduction, external rotation; FSE, fast-spin echo; TE, echo time; TR, repetition time.

to 12 mm of the distal tendon immediately posterior to the long head of biceps tendon. The rotator crescent represents the area of relative reduced vascularity in the distal supraspinatus commonly involved in rotator cuff tears. First described by Clark34 and characterized as the suspension bridge of the shoulder by Burkhart et al.,19 a retracted rotator cable is used as a proxy to identify undersurface tears of the supraspinatus tendon. Thickening and visualization of the rotator cable on MR arthrography has been described as a useful adjunct sign of partialthickness articular-sided supraspinatus tears.207 This finding further supports the role of the rotator cable as a stress shield for the vulnerable crescent. Meisha et al. recently described its biomechanical importance in a cadaveric model. They found that the rotator cable is the primary loadbearing structure within the supraspinatus for force transmission to the proximal part of the humerus. This stress shielding by the rotator cable protects the rotator crescent, the relatively avascular portion of the distal supraspinatus that is prone to tear.138 MRI detected the rotator cable in 74.3% of the MRI studies in both sagittal and coronal planes, with a mean (±SD) distance of the cable from the medial margin of the enthesis of 1.33 ± 0.27 cm, a mean width of the cable of 1.24 ± 0.31 cm, and a mean thickness of 0.19 ± 0.05 cm.76 Maldjian described thickening of the rotator cable in congenital absence of the long head of biceps tendon and a flat intertubercular groove, postulating that this is likely a related anatomic finding.122 This finding is useful in differentiating congenital biceps tendon absence from a completely torn biceps tendon.

Pathophysiology of Rotator Cuff Disease and Impingement Impingement syndrome is a clinical diagnosis, with or without supporting MRI findings. In 1991, Fu and coworkers drew on the work of

many authors to propose a unifying theory linking rotator cuff disease to impingement and glenohumeral instability.63 Intrinsic and extrinsic factors, acting together or separately, result in a common end point. Intrinsic factors lead to degenerative changes in the substance of the rotator cuff with increasing age.35,167,171 Degeneration may result from relative ischemia in the distal portion of the supraspinatus tendon  near its insertion site on the greater tuberosity (Fig. 61-5). This zone of relative ischemia called the critical zone corresponds to the most common site of rotator cuff disease, although its exact location and size vary from one patient to another.101,117,150,190 We now term this area the rotator crescent. Rotator cuff degeneration may incite secondary proliferative bony changes on the undersurface of the acromion. These changes in turn may lead to further cuff degeneration and ultimately to a cuff tear.171 Extrinsic causes of rotator cuff disease relate to mechanical impingement by surrounding structures.95 The impingement can lead to inflammatory and degenerative change in the underlying tendons. Primary impingement results from contact between the coracoacromial arch or acromioclavicular joint and the underlying tendons.10,63,99,157 The coracoacromial arch (consisting of the coracoid process, coracoacromial ligament, and undersurface of the anterior third of the acromion), acromioclavicular joint, rotator cuff tendons, and tendon of the long head of the biceps have a very close anatomic relationship. The supraspinatus tendon passes between the coracoacromial arch superiorly and the humeral head inferiorly to reach its insertion site on the greater tuberosity. The bicipital tendon passes through  the same area anteriorly. Mechanical compression of these tendons is particularly likely when the shoulder is flexed and internally  rotated, as during a tennis serve. Overuse in sports or work accelerates tendon degeneration resulting from either impingement or intrinsic degeneration.9,10,63,157,158,161 Neer and colleagues divide the impingement syndrome into three stages154,156,158:

CHAPTER 61  Shoulder

2103

B

A

FIG 61-5  Supraspinatus tendinopathy. A, PD-weighted coronal oblique image. The area of increased signal in the distal supraspinatus tendon shows thickening indicating an area of tendinopathy. B, T2-weighted coronal oblique image with fat saturation. The focus of increased signal is more pronounced, thus excluding the magic-angle effect. There is adjacent subchondral edema in the humeral head. Acromioclavicular joint degeneration is noted.

Stage I impingement syndrome is characterized by edema and hemorrhage in the distal supraspinatus tendon and is usually reversible with conservative treatment. Stage II impingement syndrome is associated with tendinitis and thickening and fibrosis of the subacromial-subdeltoid bursa. The bursal changes compound the problem by decreasing the subacromial space available to the tendons. Stage III impingement syndrome is characterized by a partial- or fullthickness tear of the rotator cuff or biceps tendon, along with osseous changes, including anterior acromial osteophyte formation and degenerative changes in the greater tuberosity. Stage III lesions usually occur in patients older than 40 years, whereas stages I and II lesions are found in younger adults. Longstanding rotator cuff tears may lead to erosions of the humerus and glenohumeral destruction, termed rotator cuff arthropathy.154,156,158 Whereas primary impingement syndrome is caused by the effects of the coracoacromial arch and acromioclavicular joint on the rotator cuff in a normally positioned and functioning glenohumeral joint, secondary impingement syndrome is related to glenohumeral instability. Secondary impingement usually occurs in young athletic individuals involved in sports that demand overhead arm movement. Repetitive external rotation in abduction, as during the cocking phase of a baseball pitch, may lead to laxity of the capsule and mild instability. This is initially compensated for by an increase in the tone of the rotator cuff, which pulls the humeral head into the glenoid and enhances stability. With continued vigorous activity, however, the rotator cuff loses its ability to compensate and the humeral head subluxates. The

subluxation results in secondary impingement.63,93,107 This atraumatic form of instability is often multidirectional and may damage different portions of the labrum or rotator cuff, depending on individual circumstances.195

Subcoracoid Impingement.  A distinct subtype of impingement occurring with the shoulder flexed, internally rotated, and adducted is associated with impingement of the coracoid process on the lesser tuberosity (Fig. 61-6). The normal interval between the coracoid process and the lesser tuberosity is greater than 11 mm in internal rotation on axial images.14 Subcoracoid impingement is deemed to be associated with interval measurements less than 11 mm.22 Friedman and colleagues found that in asymptomatic subjects the average coracohumeral interval measured 11 mm in maximum internal rotation, with none measuring less than 4 mm. Symptomatic patients had an average interval of 5.5 mm in maximum internal rotation; some measured 0 mm. Friedman’s group also noted that there is no redundancy of the subscapularis tendon in asymptomatic subjects with the arm in full internal rotation.61 Patients with subcoracoid impingement may demonstrate increased soft tissue in the subcoracoid space because of redundancy or folding of the subscapularis tendon and capsular tissues when the shoulder is in maximum internal rotation.57 This rare cause of anterior shoulder pain may be associated with an unusually long and medially oriented coracoid process. It may also follow a Bristow procedure, in which the tip of the coracoid process is moved further medially along the glenoid neck to correct anterior instability, or an osteotomy of the glenoid neck to correct posterior

2104

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 61-6  Subcoracoid impingement. A, Axial gradient echo image in a patient with clinical subcoracoid impingement demonstrating hyperintense thickening of the subscapularis tendon. B, Axial gradient echo image reveals a narrowed (6 mm) coracohumeral distance, as measured from the tip of the coracoid process (asterisk) to the humeral head. There is subcoracoid impingement on the subscapularis tendon, which itself shows a delaminating tear running between the remnant fibers (arrow).

instability. Patients without an iatrogenic risk factor often engage repeatedly in occupational or recreational activities requiring overhead motion in internal rotation.10,46,69

underlying subscapularis tendon reveals a partial-thickness intrasubstance tear.45

Noninsertional Tendinosis of Subscapularis.  Noninsertional ten-

Effect of Acromion Configuration on Rotator Cuff Tears and Impingement

dinosis of the subscapularis tendon is a term to describe intrasubstance delamination tears with no retraction of the subscapularis tendon footprint. Hence these are considered repairable tears with good functional outcome due to the intact footprint. Performance of MRI has not been formally evaluated in the assessment of noninsertional subscapularis tendon tears. This is an area that will undergo further study in future. Development of noninsertional subscapularis tendinosis is related to the finding that subcoracoid distance narrowing does not consistently reproduce impingement intraoperatively. Hence while the wringer-roller effect conveniently explains subscapularis pathology, its actual role in impingement remains a subject of discussion. A single case report raises the possibility of intrasubstance subscapularis tendon tears not involving the footprint as a cause of persistent anterior shoulder pain. MRI findings presented in the paper reflect irregularity in the distal subscapularis tendon, with increased intrasubstance T2-weighted signal and swelling. No tendon retraction is present in these cases. An interesting finding in noninsertional tears is the presence of an articular-sided nodular thickening (referred to as a Conrad lesion), which may cause triggering or catching. This nodule is located adjacent to the middle glenohumeral ligament, and when débrided the

Three different configurations of the acromion have been described by Bigliani and colleagues.9 Type I is flat inferiorly. Type II curves smoothly. In type III, the most anterior portion of the acromion has a hooked shape9 (Fig. 61-7). The acromion may also be convex along its inferior surface.55 The inferior margins of the acromion and distal clavicle usually lie at approximately the same level in the coronal plane. A type IV acromion with a convex inferior contour is described, but no convincing correlation between a type IV acromion and impingement syndrome exists.240 Bigliani et al. found an increasing incidence of rotator cuff tears with type II and particularly with type III acromia.9 The association of hooked acromia and rotator cuff tears has been confirmed by others.50,167 However, there is interobserver variation in acromial assessment,15,90 limiting its practical application. MRI and radiographic agreement on acromial morphology has also been somewhat variable, depending on factors such as radiographic angulation and MRI section,82,178 as well as the use of one versus two MR sections to form a composite image.127 A low-lying acromion has also been implicated as a cause of impingement. The acromion is considered to be low if its inferior border is lower than that of the clavicle at the acromioclavicular joint. Unfortunately there are no criteria for distinguishing a low-lying

CHAPTER 61  Shoulder

A

B

C FIG 61-7  Acromial variation. A, Type I is flat inferiorly. B, Type II curves smoothly. C, In type III, the most anterior portion of the acromion has a hooked shape.

2105

2106

PART II  CT and MR Imaging of the Whole Body

A

B FIG 61-8  Os acromiale. A, Axial CT at the level of the acromioclavicular joint demonstrating multiple os acromiale (arrows). B, T2-weighted axial oblique image with fat saturation in a different patient demonstrates a mesoacromion with periarticular edema (arrows).

acromion from a high-riding clavicle (such as may result from a previous acromioclavicular joint separation).97 An os acromiale occurs as a normal variant in approximately 8% of people. Since the 1930s, it has been considered to result from a failure of fusion of one of three ossification centers of the acromion. The os acromiale may be identified on axial and sagittal oblique images as a linear interruption of the marrow of the acromion (Fig. 61-8).176 A study of 270 scapulae in 1993 found no support for this theory and instead suggested a single site of occurrence, with the free fragment approximating one third the length of the acromion.48 The literature suggests an association between the presence of an os acromiale and development of impingement syndrome and rotator cuff tear.48,151,176 However, Ouellette and coworkers recently found no statistical difference in the incidence of supraspinatus and infraspinatus tears between patients with os acromiale and a control group. They did find that patients with a step-off deformity of the os acromiale had a greater incidence of rotator cuff tears than those without such deformity.169 Hence it is important to report on this step-off between the os acromiale and the clavicle if present. The acromial angle is not routinely measured.233 Degenerative changes at or near the acromioclavicular joint are variously considered either a cause or a result of impingement and rotator cuff disease. These include hypertrophic changes involving soft tissue in the joint capsule, as well as osteophytes that may project inferiorly from the acromion or clavicle and enthesophytes at the origin of the coracoacromial ligament. Spurs may visibly distort the supraspinatus muscle as it passes below them. At times it may be difficult to distinguish a spur from a tendinous or ligamentous attachment site, particularly because spurs may develop at such sites (Figs. 61-9 and 61-10). There is uncertainty whether spurs always contain marrow and thus match other marrow-containing bones in signal intensity.70,96 Thickening of the acromial origin of the coracoacromial ligament is associated with impingement syndrome even without enthesophyte formation.65 It may be identified on sagittal oblique MRI.

Presentation of Impingement Syndrome Patients with impingement syndrome present with a “painful arc” consisting of pain over the anterolateral humeral head exacerbated by flexion and abduction, particularly in the 70- to 120-degree portion of the arc.

FIG 61-9  Spur as shown on PD-weighted coronal oblique image. A spur containing normal marrow projects from the undersurface of the acromion (arrow).

At physical examination, a relocation maneuver may reveal the subtle instability found in these patients. With the patient supine and the shoulder abducted and externally rotated, posteriorly directed force is applied to the anterior proximal humerus. As the humerus is reduced posteriorly, the patient’s pain is alleviated.92,107 Differentiating primary from secondary impingement can be clinically difficult. Patients with secondary impingement are usually

CHAPTER 61  Shoulder

2107

younger than those with primary impingement, and they may relate the pain to a very specific activity. A baseball pitcher with pain restricted to the late cocking or early acceleration phase of throwing is a typical example. It is important to differentiate primary from secondary impingement; treatments differ significantly. Patients with primary impingement are usually treated with an anterior acromioplasty, with appropriate attention to the rotator cuff and biceps tendon. In patients with secondary impingement, treatment is aimed at the instability.93,94,125,126,157,158 In many patients with suspected impingement syndrome, a direct mechanical cause is not found. Hence intrinsic tendon degeneration may be the main cause of supraspinatus tendinopathy in this group of patients. There may also be a difference in pathogenesis of bursal versus articular surface partial-thickness tears of the supraspinatus. Ozaki and colleagues found that bursal-side and full-thickness rotator cuff tears correlated with degenerative changes of the coracoacromial ligament and anterior third of the inferior acromion.171 Articular-surface partialthickness tears of the supraspinatus, however, were associated with fewer pathologic changes in the acromion (Fig. 61-11).

Imaging Features of Rotator Cuff Disease

FIG 61-10  Spurious spur as shown on T2-weighted coronal oblique image with fat saturation. A slender hypointense structure projects from the undersurface of the lateral acromion (arrow). It represents a slip of deltoid tendon or a portion of the coracoacromial ligament (or both, because their fibers may fuse), but it may be mistaken for a spur. This can be distinguished from a true spur by the absence of marrow signal and the lack of radiographic confirmation of spur formation.

MRI can depict the continuum between rotator cuff degeneration or tendinopathy and partial- or full-thickness rotator cuff tears. Intrinsic cuff degeneration or tendinopathy is seen as increased signal intensity in the distal tendon on proton density (PD) images; this persists but does not intensify on T2-weighted images. Reported sensitivities for imaging of full-thickness rotator cuff tears have ranged from 80% to 100%, and specificities from 94% to 100%.53,103,104,260 Arriving at a diagnosis of partial-thickness tears on nonarthrogram MRI is more problematic, with sensitivities up to 92% and specificities of 83% to 100%. The use of 3-tesla (3T) MRI was reported by Magee and Williams, achieving a sensitivity of 98% and specificity of 96% (Table 61-2).121

TABLE 61-2  Performance of MRI in Diagnosis of Rotator Cuff Tears FULL-THICKNESS TEARS

PARTIAL-THICKNESS TEARS

No. of Patients

Sensitivity/ Specificity/ Accuracy (%)*

Sensitivity/ Specificity/Accuracy (%)*

Evancho et al., 1988 Burk et al., 198917a Flannigan et al., 199059a

31 38 23

80/94/89 92/100/94 —

— — —

Rafii et al., 1990189 Hodler et al., 199284a

Traughber et al., 1992231

80 36 (13 with partial tears, 4 with full tears) 28

Palmer et al., 1993171a

42

97/94/95 89/84/85 Combined full and partial tears a. MRI: 41/79/61 b. MR arthrography with gadolinium: 71/84/70 100/100/100 5 of 9 partial tears diagnosed with MRI 100/100/100 100/100/100

Quinn et al., 1995187a

100

85/99/96

Study 53a

82/99/97

Comment — — Improved performance of MR arthrography over MRI — Accuracy improved with MR arthrography

— Tested use of fat saturation; performance was lower without fat saturation (90%, 75%, and 84% for combined full and partial tears) Tested use of fat saturation Continued

2108

PART II  CT and MR Imaging of the Whole Body

TABLE 61-2  Performance of MRI in Diagnosis of Rotator Cuff Tears—cont’d FULL-THICKNESS TEARS

PARTIAL-THICKNESS TEARS

Sensitivity/ Specificity/ Accuracy (%)*

Sensitivity/ Specificity/Accuracy (%)* 10-20/83-93/—(spin echo without fat saturation) 24-45/93-97/—(spin echo with fat saturation) 19-57/85-93/— 92/—/—(with fat saturation) 67/—/—(without fat saturation) —

Study

No. of Patients

Reinus et al., 1995191a

80

Robertson et al., 1995194a Singson et al., 1996212a

82 43

80/87-92/—(spin echo without fat saturation) 100/77-87/—(spin echo with fat saturation) 81-100/89-98/— 100/—/—

Sonin et al., 1996215a

26

89/94/92

Sahin-Akyar et al., 1998198

39

83-100/—/—

Magee and Williams, 2006121

150

98/96/—

92/100/—

Comment Performance of MRI improved by using fat saturation

Compared four readers

Compared conventional and turbo-spin echo and found them equivalent Found that T2-weighted gradient echo and fast-spin echo with fat saturation both gave acceptable results Study performed with 3-tesla MRI sensitivity for detection of small tears (<1 cm) was 89.5% (26 of 28 small tears were seen), and specificity was 90%

*As given in the original article; not recalculated or derived from data in the article.

A

B FIG 61-11  A, T1-weighted fat-saturated coronal oblique MR arthrographic image shows an articular-side partial tear (arrow). B, T1-weighted sagittal oblique MR arthrographic image. On the articular side of the supraspinatus tendon is a defect filled with gadolinium-based contrast (arrow), with an intact bursal surface. No contrast is present in the subdeltoid bursa.

CHAPTER 61  Shoulder Partial-Thickness Tears.  Partial tears of the tendon may be intratendinous or may occur on either the bursal or the articular side of the tendon. They are most common along the articular surface of the cuff.235 On MRI, partial tears are depicted as focal areas of increased signal intensity on short echo time (TE) images; these areas increase in relative signal intensity on long-TE images but need not match the signal intensity of joint fluid (see Fig. 61-6). The abnormal focus does not extend through the entire thickness of the tendon. There may be increased fluid in the glenohumeral joint, particularly if the partial tear is on the articular side. Alternatively, fluid may be seen in the subacromial-subdeltoid bursa if the partial tear lies on the bursal side of the tendon. MRI findings in tendon degeneration and mild partial tears can overlap, compounded by the fact that the conditions tend to coexist. Careful attention to the bursal, intrasubstance, and articular surface morphology of the tendon, combined with distribution of abnormal signal intensity on PD- or T2-weighted sequences,212a is required. The presence of partial cuff fiber retraction is a sign of partial-thickness tear (Fig. 61-12). MR arthrograms are helpful in diagnosing articularsided partial-thickness tears by causing distention and capsular herniation into the tear defect. Snyder classified the severity of partial rotator cuff tears as follows221: • 0: Normal • I: Minimal superficial bursal or synovial irritation or mild capsular fraying in a small localized area (<1 cm) • II: Fraying and failure of some rotator cuff fibers with synovial bursal or capsular injury (<2 cm)

2109

• III: Fraying and fragmentation of tendon fibers usually involving the whole surface of a cuff tendon, most commonly the supraspinatus (<3 cm) • IV: Severe tear with tendon fraying, fragmentation, and a sizable flap tear involving more than a single tendon The depth of tendon fiber involvement is also used as a criterion for grading of partial tears259: • 1: Less than 3 mm deep • 2: 3 to 6 mm deep and less than 50% of the cuff thickness involved • 3: High-grade partial-thickness tear more than 6 mm deep with more than 50% of the rotator cuff thickness involved A subset of partial-thickness tears not well addressed includes  the articular rim-rent tear, which occurs at the greater tuberosity  (Fig. 61-13). Such small tears can extend along the tendon fibers, causing tendon delamination, which corresponds to Snyder’s III or IV classification. These resultant partial articular supraspinatus tendon avulsion (PASTA) lesions result in a tendon flap (Fig. 61-14). Involvement of the bursal surface can also occur (reverse PASTA). Surgical repair often involves suturing the delaminated tear, converting the tear to full thickness before repair.216

Full-Thickness Tears.  In a full-thickness rotator cuff tear, the tendon is interrupted by an abnormal focus of increased T2-weighted signal intensity extending across its complete superior-to-inferior extent on at least one image (Figs. 61-15 and 61-16). Fluid in the subacromial-subdeltoid bursa is a common association in fullthickness tears, although it can be present without a tear when the

FIG 61-12  Partial tear of the articular surface of supraspinatus as shown on T2-weighted coronal oblique image with fat saturation. There is a focal defect on the articular surface of the supraspinatus tendon, with adjacent bone edema. Medially there is thickening of the articular surface tendon, representing bunching of retracted torn articular surface fibers (arrow).

FIG 61-13  Rim-rent tear as shown on T2-weighted coronal oblique image with fat saturation. There is a focal defect on the articular surface of the supraspinatus tendon (arrow), extending slightly medially. There is no visible tendon retraction.

A

B FIG 61-14  Partial articular surface tendon avulsion (PASTA). A, T1-weighted sagittal oblique MR arthrographic image with fat saturation. There is a longitudinally oriented area of contrast signal in the distal supraspinatus tendon, implying connection with an articular-sided tear (arrow). This represents a delaminating tear within the tendon substance. There is an incidental superior labral tear. B, Oblique axial T1 fat-saturated MR arthrogram (abduction, external rotation [ABER]) demonstrates a large delaminating tear through the supraspinatus tendon. Note the extension of contrast between the supraspinatus fibers (arrows).

A

B FIG 61-15  Full-thickness tear. A, T2-weighted coronal oblique image with fat saturation. There is a fullthickness defect in the supraspinatus tendon at its attachment site on the greater tuberosity (arrow) that is filled with fluid. No intact tendon exists above or below the abnormal area. B, T2-weighted sagittal oblique image with fat saturation. There is a focus of increased T2 signal extending from the articular and bursal surfaces of the supraspinatus tendon, in keeping with a full-thickness tear (arrow).

CHAPTER 61  Shoulder

A

2111

B FIG 61-16  Large cuff tear. A, T2-weighted fat-saturated coronal oblique image. There is a large tear in the rotator cuff into the subacromial-subdeltoid bursa. The tendon is retracted (small arrow). A large amount of effusion is present in the glenohumeral joint, communicating with the biceps tendon sheath (large arrow). B, T2-weighted fat-saturated sagittal oblique image reveals a large defect in the anterior to midsubstance of the supraspinatus tendon.

patient has bursitis.54,85 Some full-thickness tears lack fluid-intensity signal because of a paucity of fluid in the joint or because the tendon gap is filled with scar.189 Other useful signs include tendon retraction and atrophy of the involved muscles with associated fatty infiltration. Tendon retraction becomes more obvious as the size of the tear increases, and it may be inapparent in small tears.189 Fatty infiltration of the muscle is expected in relatively chronic tears (Fig. 61-17) and is best appreciated on sagittal T1-weighted images. Large tears may be surprisingly subtle if there is little fluid in the joint to communicate with the tendon gap or subacromial bursa. Occasionally tendon retraction may bring the amputated stump to rest along the medial surface of the humeral head so that the humeral cortex resembles a smooth continuation of the tendon. In these cases, cephalad migration of the humeral head may be the most obvious sign of tearing. Another occasional accompanying sign is irregularity of the bony elements of the glenohumeral joint as a result of rotator cuff arthropathy (Fig. 61-18). An irregular greater tuberosity has been shown to correlate with rotator cuff tears at sonography.251 Cystlike lesions of the humeral head have a modest correlation with the presence of rotator cuff disease but are also common in asymptomatic individuals.87,152

S

I

Rotator Interval Tears.  The rotator interval capsule resists inferior and posterior glenohumeral translation through the direct contribution of the coracohumeral ligament and also helps maintain the negative pressure in the glenohumeral joint. The long head of the biceps maintains anterior stability, especially in ABER, by resisting torsional forces.149 Rotator cuff tears most commonly begin in the supraspinatus tendon, usually near its anterior edge. As the tear enlarges, it may extend completely across the anteroposterior dimension of the supraspinatus muscle and tendon to involve parts or all of the infraspinatus

FIG 61-17  Supraspinatus atrophy as shown on T1-weighted sagittal oblique image. The supraspinatus (S) and infraspinatus (I) muscles are markedly atrophied and infiltrated with fat. The supraspinatus belly lies inferior to a line connecting the tips of acromion and the coracoid process (white line).

2112

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 61-18  Cuff arthropathy. A, PD-weighted sagittal oblique image. There is also irregularity of the greater tuberosity cortex in keeping with degenerative change (arrowheads). B, PD-weighted coronal oblique image. Jagged, irregular humeral inferior osteophyte (arrow) is typical of cuff arthropathy.

and subscapularis. When the tear occurs at the critical zone of the supraspinatus muscle and extends into the subscapularis, it crosses and involves the intervening rotator interval capsule. Isolated tears of the rotator interval also occur, although uncommonly. They are usually  the result of an anterior dislocation. Synovium may herniate from  the glenohumeral joint into the rotator interval. With MRI it can  be difficult if not impossible to distinguish an isolated rotator interval tear from herniated synovium, especially without administration of intraarticular contrast medium.164,206 The relevance of rotator interval lesions lies in the pathologic processes affecting its constituents, namely the biceps tendon, superior glenohumeral ligament, anterior border of the supraspinatus and superior border of the subscapularis, and synovium. The intraarticular portion of the long head of the biceps tendon is supported by an interwoven arrangement between the coracohumeral, superior glenohumeral, and transverse humeral ligaments (Fig. 61-19). These structures prevent medial subluxation of the tendon. MRI depictions of biceps tendon pulley lesions include extravasation of contrast through the interval, irregularity of the superior border of subscapularis, and biceps tendon dislocation.8 An arthroscopic study of “hidden” rotator interval lesions in patients with supraspinatus tears showed a 27% incidence of subscapularis tears; 47% of these were associated with involvement of the superior glenohumeral ligament and medial fibers of the coracohumeral ligament.247 Anterosuperior internal impingement is another condition related to rotator interval lesions. Clinically the pain is provoked by elevation and internal rotation of the arm, but there is little evidence of instability.68 Impingement usually occurs between the long head of the biceps tendon and the pulley region with the superior labrum, or between  the tendinous insertion of the subscapularis and the anterior glenoid. The expected MR arthrographic findings include partial-thickness

subscapularis tendon tears (articular surface) with an associated pulley lesion (Fig. 61-20). Anterosuperior internal impingement can be demonstrated on ABER sequences.111 In adhesive capsulitis, Mengiardi and colleagues describe thickening of the coracohumeral ligament (>4 mm on sagittal oblique MRI), which can also be associated with thickening of the rotator interval capsule and synovial proliferation.136 Inferior instability of the glenohumeral joint can be due to a rotator interval lesion. This is important because the rotator interval itself is used as an arthroscopic portal, sometimes resulting in these lesions being missed surgically.

Imaging Procedure MR arthrography has a sensitivity, specificity, and accuracy of 96%, 99%, and 98%, respectively, for full-thickness tears.194a For partialthickness tears, it has a sensitivity, specificity, and accuracy of 80%, 97%, and 95%, respectively.89,243 Because nonarthrographic MRI of the shoulder already has a high degree of accuracy, particularly for fullthickness tears, any substantial gain in accuracy with MR arthrography is in the diagnosis of partial tears.84a Evaluation of the rotator cuff tendons should be performed in three orthogonal planes. The supraspinatus tendon is best imaged in coronal and sagittal planes. In degeneration, there is intermediate signal intensity on PD-weighted images, with no hyperintense signal on T2-weighted images. Use of fat-suppression sequences increases the sensitivity of MRI for tears/tendinopathy. Attention should be paid to cuff morphology, outline, and interstitial continuity through to the myotendinous junction unit. Fraying of the cuff can be appreciated as a contour irregularity on the bursal or articular surface. The presence of tendon retraction and muscle

CHAPTER 61  Shoulder

A

2113

B FIG 61-19  Biceps tendon pulley. A, Sagittal T1-weighted image of the rotator interval demonstrating the coracohumeral ligament (arrow) and the superior glenohumeral ligament fibers blending to envelop the intraarticular biceps tendon (arrowhead). B, Sagittal T1-weighted MR arthrogram with fat saturation showing an intact capsule, with the pulley surrounding the intraarticular biceps tendon (arrow).

atrophy should be noted. The following points should be specifically addressed when reporting on rotator cuff injuries: 1. Dimensions of a full-thickness tear: This has implications for selection of the surgical approach, prognosis, and probability of tear recurrence. For example, full-thickness tears greater than 1 cm2 have a poor outcome with conservative treatment and will likely benefit from surgery.3 In contrast, tears greater than 4 cm2 are not amenable to primary repair. Also, larger tears can be more easily repaired with an open rather than arthroscopic approach, depending on the operator. 2. Depth of partial-thickness tear: Articular surface tears are more common than bursal surface tears. There is potential for a partialthickness tear to convert to a full-thickness tear. The tear is propagated by the ever-increasing load placed on the remaining fibers. The normal rotator cuff tendons measure 10 to 12 mm in thickness, and tears extending 50% or beyond the tendon depth are deemed significant.49 In this situation, MRI has an advantage over arthroscopy because the depth of tear can be better measured, as opposed to the limited field of view afforded by arthroscopy.171a,187a 3. Tear shape: Viewed relative to the tendon surface, there are three configurations of rotator cuff tendon tears: U, L, and crescenteric.18,116 In crescentic tears, the tendon pulls away from the greater tuberosity but does not retract far medially and therefore can be reattached to bone with minimal tension. U-shaped tears are massive rotator cuff tears that may extend medially to the level of the glenoid fossa. L-shaped tears are massive tears with a longitudinal component along the orientation of the rotator cuff fibers and a transverse component along the cuff insertion.148

4. Tendon retraction: When the torn tendon ends are adjacent to each other, primary repair may be undertaken with the repaired tendon relatively free of tension. The ease of repair declines with greater degrees of retraction, being almost impossible when the tendon is retracted medial to the glenoid rim. 5. Tear extension: A supraspinatus tear may extend posteriorly through the conjoined tendon to involve the infraspinatus. It may also extend anterior into the rotator interval capsule and involve the coracohumeral ligament and the biceps tendon as well as subscapularis. Involvement of the rotator interval increases the rate of tear recurrence from 7% to 25%.67 Mellado and associates found that tendon re-tears were significantly more prevalent among patients with surgically demonstrated 2-tendon tears (78.9%) or 3-tendon tears (100%) and with preoperative fatty degeneration of the supraspinatus muscle greater than 1 (91.6%). Reruptures were also significantly larger in those subgroups.135

Pitfalls in Diagnosis On T1-weighted coronal oblique images, the supraspinatus tendon can present with an area of apparently increased signal intensity. This is located approximately 1 cm proximal to the insertion site of the tendon. It has been described by some authors as isointense to muscle.91,135 Others refer to it merely as increased in signal intensity, without providing a point of reference. Although most authors have noted it on short-TE technique, Kaplan and colleagues found it to be most apparent on gradient echo images,96 and Mirowitz also noted it in some T2-weighted images.142 It has been identified in asymptomatic individuals and does not necessarily represent a pathologic finding.

2114

PART II  CT and MR Imaging of the Whole Body Another source of discrepancy between surgical findings and MRI reports occurs when there is an interstitial tear of the tendon that does not extend to the surface. Because there is no surface communication, arthroscopy may not reveal a tear.220

Postsurgical Imaging of the Rotator Cuff

FIG 61-20  Sagittal T1-weighted MR arthrogram of the rotator interval demonstrating an ill-defined capsule with contrast extravasation through the rotator interval capsule (arrow). The biceps tendon is not displaced (arrowhead).

Attention to tendon morphology may help distinguish normal  from pathologic increases in signal intensity. A normal tendon  should taper smoothly from the musculotendinous junction to the insertion site. Several theories have been proposed to explain this area of increased signal intensity. It may represent subclinical degenerative change in the tendon.142,189,261 Erickson and coworkers noted increased signal intensity on T1-weighted images that they attributed to the “magic-angle effect,” in which an artifactual focus of increased signal intensity may occur on short-TE images of a tendon oriented 55 degrees to the constant magnetic field.52 They postulated that such increased signal intensity may also be expected on intermediate-weighted images but not on long-TE images (i.e., T2-weighted images; Fig. 61-21). Timins and colleagues bolstered this suggestion in 1995 by showing that the area of increased signal intensity changed position in the tendon with changes in patient position.226 One may reach the correct conclusion by following the intact tendons on oblique sagittal images.42,115,160,236 The supraspinatus tendon is divided into two structures—the anterior main tendon and smaller posterior slips. There are also slips of fatty connective tissue between the supraspinatus and infraspinatus muscles. The theory that these heterogeneous tissues can cause partial volume averaging has not been well accepted. Muscle/fat interposition has not been described near the cuff insertion where the magic-angle effect is seen, and true T1 shortening by fatty tissue persists independent of different imaging orientations.33,237 The reader should also be aware of previous injection therapy, resulting in increased T2 signal in the subacromial space (similar  to fluid).99

Commonly performed operations for rotator cuff repair are summarized in Table 61-3. MRI is useful in assessment of the rotator cuff after surgery. Ideally MR arthrograms should be performed on all postoperative shoulder joints to improve delineation of individual structures. Comparison with previous studies should also be routinely performed. Results, however, must be correlated with clinical findings. The presence of metallic artifacts in the subcutaneous tissues over the deltoid or deeper tissues is a good indicator of prior surgery.120 Suture anchor tracks can also be present in the glenoid rim. These artifacts can be exaggerated with gradient echo sequences (Fig. 61-22). Expected MRI findings after arthroscopic repair include changes in acromial morphology and widening of the acromioclavicular distance. Changes in acromial morphology arise from subacromial decompression and removal of a subacromial spur. The result is a flatter and tapered acromial configuration as opposed to a hooked appearance. There may be reduced T1- and T2-weighted signal intensity, reflecting sclerosis.145,193 Resection of the distal clavicle (Mumford procedure) results in widening of the acromioclavicular distance to 1 to 2 cm. There may be superior subluxation of the humeral head because of anterior capsular shift, related to anterior capsulorrhaphy. The presence of fluid in the subacromial-subdeltoid bursa is a nonspecific finding in a postoperative MR arthrogram owing to some rotator cuff repairs’ not being watertight and may be associated with a functional non-watertight repair, a recurrent tear in the rotator cuff, or an otherwise normal postoperative rotator cuff.218 On MR arthrography, the presence of contrast extending into the acromioclavicular joint from the subacromial bursa (known as the geyser sign) implies injury to the acromial undersurface. This may also be seen on non-MR arthrography as fluid in the acromioclavicular joint extending from the subacromial bursa (Fig. 61-23). In a majority of patients, acutely repaired rotator cuff tears exhibit increased signal intensity on T2-weighted images.170,217 Eventually the tendon returns to an intermediate to low signal intensity, reflecting granulation tissue and fibrosis, depending on the native tendon quality.259 Spielmann et al. studied the supraspinatus and infraspinatus tendons of a group of 15 asymptomatic patients after rotator cuff repair and found MRI evidence of rotator cuff tear in 11 (37%) of the 30 tendons. They also found subacromial-subdeltoid effusions, joint effusions, and bone marrow edema to be common in their patients.215 In addition to presence of infection, hematoma, and inadvertent soft tissue injuries, especially to nerves, rapid glenohumeral chondrolysis can occur after shoulder arthroscopy.113,182,200 Chondrolysis is a devastating condition in which widespread rapid chondrocyte death occurs after exposure of articular cartilage to chlorhexidine239 and methylmethacrylate.110 Patients usually present with progressive limitation of range of movement 3 to 6 months after arthroscopy and disproportional postoperative pain.

ABNORMALITIES OF THE BICEPS TENDON Bicipital Tendinopathy The biceps tendon is intimately associated with the rotator cuff. Most cases of bicipital tendinopathy are the result of mechanical impingement that occurs with rotator cuff disease in middle-aged or older patients. With flexion of the shoulder, the tendon of the long head of

CHAPTER 61  Shoulder

2115

B

A

FIG 61-21  Magic-angle artifact. A, Coronal PD-weighted image of the supraspinatus tendon reveals an area of apparent increased signal in the distal tendon, corresponding to a point where the tendon curves inferiorly. This is also the area often affected by degeneration and is a potential diagnostic pitfall. B, T2-weighted image of the same tendon demonstrating normal signal intensity, confirming the presence of magic-angle artifact.

TABLE 61-3  Commonly Performed Surgical Procedures of the Shoulder Indication

Procedure

Important Points

Impingement

Subacromial decompression

Impingement Rotator cuff tear (partial)

Full-thickness rotator cuff tear

Mumford procedure Débridement and subacromial decompression Débridement ± acromioplasty Open or mini-open cuff repair

Enlargement of the acromiohumeral distance Resection of coracoacromial ligament, anterior and posterior acromial resection, resection of acromioclavicular osteophytes Resection of distal clavicle in combination with subacromial decompression Partial-thickness bursal-sided tears

Glenohumeral instability

Bankart repair

Glenohumeral instability

Putti-Platt procedure Magnuson-Stack procedure Bristow-Helfet procedure

Rotator cuff tear

Glenohumeral instability Glenohumeral instability

Multidirectional instability

Capsular shift

Partial-thickness articular-sided tears (not more than two thirds tendon thickness) Deltoid muscle split Cuff débridement Side-to-side suturing of tendon ends and tendon reattachment to greater tuberosity Can be combined with subacromial decompression and Mumford procedure Suture anchors placed at the glenoid rim at 3-, 4-, 5-o’clock positions Anatomic repair Shortening procedure of the anterior capsule and subscapularis muscle Nonanatomic repair, limitation of external rotation Transfer of subscapularis tendon from the lesser to greater tuberosity Nonanatomic repair, limitation of external rotation Coracoid process and short head of biceps transfer to scapular neck Subscapularis split Nonanatomic bone block repair May result in nonunion and limitation of external rotation Inferior capsule shifted in a superior direction May result in overtightening and limitation of external rotation

2116

PART II  CT and MR Imaging of the Whole Body glenohumeral joint, where it may be mistaken for a detached anterior labrum. In rare circumstances when the subscapularis tendon remains intact yet the coracohumeral and transverse humeral ligaments are disrupted, the biceps tendon displaces extraarticularly and anterior to the subscapularis muscle. Occasionally the biceps tendon lodges in the fibers of the distal subscapularis tendon (see Fig. 61-25).24,25,181 Besides allowing identification of the displaced biceps tendon, both CT and MRI show an empty bicipital groove. On axial MRIs, particularly those using gradient echo sequences, tiny bony spicules projecting into the bicipital groove may mimic the biceps tendon. They may be distinguished from it, however, because they are probably not present on all the axial images where the biceps tendon would be expected to be seen. In addition, the far anterior coronal images also show an empty bicipital groove. Scar tissue in the bicipital groove may also resemble the tendon and mask a dislocation or tear.241 Surgery is usually performed in the young or middle-aged patient with dislocation of the biceps tendon. The tendon is replaced in the bicipital groove and the fibrous roof is reconstructed. Tenodesis is performed if the tendon is severely attenuated. Acromioplasty as well as rotator cuff arthroplasty can be performed. In older patients the treatment is usually directed to the rotator cuff disorder; the biceps tendon dislocation is considered secondary. If the biceps tendon is severely attenuated, tenotomy may be performed, which results in absence of the tendon in the bicipital groove.

Congenital Absence of the Biceps Tendon FIG 61-22  Axial gradient echo image of the shoulder. There is gross susceptibility artifact on axial MRI of the shoulder after biceps tenodesis and previously attempted subscapularis tendon repair. Note the presence of metallic artifact in the subcutaneous tissues anteriorly, demarcating the surgical approach. The subscapularis tendon has retorn and is likely retracted.

Congenital absence of the biceps tendon is a rare anatomic variant that should not be confused with complete tear. In case reports of congenital absence, a flat intertubercular groove is demonstrated. Thickening of the rotator cable (fibrous thickening on the undersurface of the distal supraspinatus tendon) is also seen and is likely due to increased stress on the structure.122

ANTERIOR GLENOHUMERAL INSTABILITY Pathophysiology

the biceps may become closely approximated to the coracoacromial ligament and the anterior acromion, particularly in the presence of a rotator cuff defect, and may be compressed between these structures and the humeral head.43,161,179 This impingement eventually results in tendinopathy, with flattening and thinning of the tendon that may lead to complete rupture.156,256 Distention of the biceps tendon sheath by fluid and synovial thickening is also seen. Bicipital tendinopathy can also occur in patients who have glenohumeral instability, particularly in the presence of pathologic conditions involving the superior labrum. Early bicipital tendinopathy is often undetected on imaging studies. Increased signal intensity in the abnormal tendon is rarely visualized on MRI (Fig. 61-24) and when present may be a result of the magicangle effect.55

Bicipital Tendon Dislocation Bicipital tendon dislocation usually occurs in conjunction with large long-standing tears of the rotator cuff that have extended to include the biceps tendon pulley and the subscapularis tendon (Fig. 61-25). The tendon dislocation may be overlooked clinically because the symptoms are masked by an associated cuff tear.177 Other lesions necessary for bicipital tendon dislocation to occur include disruption of both the strong coracohumeral ligament, an important stabilizer of the biceps tendon,124 and the weak transverse humeral ligament. In this setting, the tendon of the long head of the biceps dislocates medially, moving deep to the subscapularis tendon and muscle and into the

Glenohumeral joint stability depends on the supporting and restraining effect of the glenoid labrum, the joint capsule including the glenohumeral ligaments, and the rotator cuff. Instability can occur in all directions, but the most common is anterior, related to failure of the anterior labroligamentous complex.5 Subcoracoid dislocations with the shoulder in abduction and external rotation are the most common variant of anterior instability. Instability may present as subluxation or frank dislocation and can be chronic and recurrent, or it may present as an isolated episode.204 Shoulder dislocations, which most commonly occur in an anterior and subcoracoid location, are usually the result of acute trauma. During an initial episode of dislocation, the glenoid labrum and anterior capsule may become damaged, and an impaction fracture of the posterolateral humeral head, resulting in a Hill-Sachs deformity, may occur. These lesions predispose to recurrent dislocation, often after minimal or no trauma. Recurrent dislocations are common in young patients but become progressively less common with advancing age; they are relatively rare after age 40.2,43,126,175,230 Some patients can even dislocate the shoulder voluntarily.197 A general rule of thumb with the traumatic unidirectional instability is that it  is treated with Bankart surgery (traumatic unidirectional Bankart surgery [TUBS]). Atraumatic instability is less common than traumatic instability.126 It is more difficult to diagnose than the traumatic instabilities. Patients complain of the arm’s going numb or their shoulder sliding forward. Atraumatic multidirectional and bilateral instability (AMBRI) or  multidirectional instability (MDI) are treated with rehabilitation or

A

B FIG 61-23  Geyser sign. A, Coronal inversion recovery image through the acromioclavicular joint demonstrates free connection of subacromial fluid (arrow) extending into the acromioclavicular joint. B, Sagittal inversion recovery image confirming a full-thickness tear of the supraspinatus tendon and fluid in the subacromial space (arrow).

*

A

B FIG 61-24  Bicipital tendinopathy. A, Gradient echo axial image at the level of the inferior labrum shows distention of the biceps tendon sheath, with fluid and thickening of the tendon (asterisk), in keeping with tendinopathy. The tendon is intact. B, T2-weighted sagittal oblique image. Fluid within the biceps tendon sheath can extend along the tendon (arrowheads). It is important to assess the rest of the shoulder structures for concomitant rotator cuff pathology. In this case, increased signal in the substance of the infraspinatus is suggestive of muscle strain.

2118

PART II  CT and MR Imaging of the Whole Body

A

B FIG 61-25  Bicipital tendon dislocation. A, Gradient echo axial image. The bicipital tendon (arrow) is dislocated medially immediately adjacent to the lesser tuberosity and is embedded between the torn fibers of the subscapularis tendon. B, In a different patient, the biceps tendon is dislocated medially (arrow), deep to the subscapularis tendon. The bicipital groove is empty.

reconstruction of the rotator interval capsule–coracohumeral ligament mechanism with or without capsular shift. Patients can have a combination of TUBS and AMBRI. In MDI of the glenohumeral joint, inferior subluxation of the humeral head can be elicited clinically. This is thought to be related to ligamentous laxity. The patient is typically young and likely female, with or without a labral lesion. The ligaments are lax and the glenoid can be hypoplastic.

Imaging of Anterior Glenohumeral Instability MRI, CT arthrography, and MR arthrography can often demonstrate the anatomic residua of prior episodes of instability.5,188,209,211,212,250 Identification of the stigmata of instability, however, does not indicate whether the problem is ongoing, because (1) anatomic changes may occur at a single episode of dislocation, and (2) it is rare to identify actual displacement of the humeral head at the time of imaging. The imaging signs of instability include labral and glenoid rim abnormalities, capsular stripping, and deformities of the humeral head. Although practically any combination of injuries may occur in an individual patient, certain groupings of anatomic abnormalities occur repeatedly (Table 61-4).

Labral Tears Diagnosis.  Tears of the glenoid labrum are identified by complete absence of a portion of the labrum, displacement of the labrum from the glenoid rim, or abnormal linear areas of increased signal intensity in the labrum at MRI.66,75,98,112,126,176,260 Similar diagnostic criteria are used at CT or MR arthrography (Fig. 61-26). In addition, contrast

medium entering the labrum or tracking deep to the labrum to the external surface of the joint is consistent with a tear. At CT arthrography a characteristic rounded area of attenuation that is less than that of normal labral tissue, termed the cheerio sign,88 has been reported with displaced bucket-handle tear. The bucket-handle fragment may also be identified at MRI or MR arthrography as a low-signal-intensity intraarticular body.146 In addition, avulsions of the periosteal sleeve of the glenoid can occur in conjunction with an anteroinferior labral tear (anterior labroligamentous periosteal sleeve avulsion [ALPSA]), and this finding indicates a chronic lesion (Fig. 61-27). Good results in diagnosis of labral tears have been reported for MRI, CT arthrography, and MR arthrography, albeit somewhat less consistently for the first two than for MR arthrography (Table 61-5). In 1993, Chandnani and coworkers compared MRI, MR arthrography, and CT arthrography and found MR arthrography to be the most sensitive in detecting labral tears (96%); it also offered the best evaluation of the inferior labrum and inferior glenohumeral ligament.29 In 1999, Roger and colleagues compared conventional arthrography, CT arthrography, and MR arthrography in 17 throwing athletes and found that MR arthrography was both more sensitive and more specific than the other modalities, not only for labral abnormalities but also for rotator cuff lesions.195 Bankart lesions are classified into soft tissue (or cartilaginous) lesions or bony lesions. The cartilaginous Bankart lesion is due to avulsion of the anteroinferior labroligamentous complex without fracture of the bony glenoid rim. It is important to distinguish this from a bony Bankart lesion because glenoid reconstruction is indicated if there is a large bony defect. Oblique sagittal MRIs are used to assess accurately

CHAPTER 61  Shoulder

2119

TABLE 61-4  “Alphabet Soup”

of Glenohumeral Instability and Glenoid Lesions Lesion

Description

References

SLAP: superior labrum, anterior to posterior

Labral tear stretching across superior portion of labrum, crossing origin of long head of biceps brachii muscle Detachment of capsule and inferior glenohumeral ligament from their anterior humeral attachment site; uncommon lesion first described in 1942 Tear of anteroinferior glenoid labrum with associated injury to glenoid articular cartilage; lesion is caused not by instability but by an impaction injury Periosteum of the anteroinferior glenoid remains intact, whereas labrum is displaced medially and rotated inferiorly

24, 71, 80, 84a, 88, 112, 119, 146

HAGL: humeral avulsion of the glenohumeral ligament

GLAD: glenolabral articular disruption

ALPSA: anterior labroligamentous periosteal sleeve avulsion lesion

A

13a, 163a, 229a, 253

160a, 199a

160a FIG 61-26  Anterior labral damage (arrow) as shown on a CT arthrogram. The anterior labrum is diminutive and irregular. Focal indentation of the posterolateral humeral head is suggestive of a Hill-Sachs lesion. This should be confirmed with sections of the humeral head superiorly.

B FIG 61-27  Anterior labroligamentous periosteal sleeve avulsion (ALPSA). A, Anterior labral avulsion and periosteal stripping. T1-weighted fat-saturated axial oblique MR arthrogram. The anterior labrum (arrowhead) has been avulsed from the glenoid rim with a short segment of periosteum, allowing the intraarticular gadolinium solution to flow between these structures and the glenoid. B, CT arthrogram of a different patient reveals a torn anterior labrum as well as periosteal sleeve avulsion, in keeping with an ALPSA lesion. Note the periosteal reaction (arrow).

2120

PART II  CT and MR Imaging of the Whole Body

TABLE 61-5  Performance of CT Arthrography, MRI, and MR Arthrography in the Diagnosis of

Labral Tears Study

No. of Patients

Sensitivity/Specificity/ Accuracy (%)*

Deutsch et al., 198444 Callaghan et al., 198822a

44 30

Chandnani et al., 199329

28

96/—/86 66/100/96 (SLAP lesions) 100/100/100 (posterior labrum) 90/73/83 (anterior labrum) 73/—/—

MRI Garneau et al., 199166

24

44–77/67/—

Tuite et al., 1995233

24

36–50/90/55–67

Gusmer et al., 199675

103

89/97/95

MR Arthrography Chandnani et al., 199329 Palmer et al., 1994172 Palmer and Caslowitz, 1995173 Cvitanic et al., 199740

30 48 121 92

96/—/— 91/93/92 92/92/92 96/97/—

Willemsen et al., 1998249a

44

Bencardino et al., 20007

52

Comment

Compared MRI, MR arthrography, and CT arthrography and found MR arthrography to be the most sensitive modality Concluded that axial MRI was relatively inaccurate for labral evaluation Compared two readers’ performances with axial MRIs obtained in neutral and externally rotated positions Argued that technical advances in MRI have allowed better evaluation of the labrum than was possible previously; used both gradient echo and proton density–weighted, fast-spin echo axial sequences to obtain the reported level of accuracy

Tested ABER position and found better performance with combined ABER and axial images than with either alone (44% sensitivity, 92% specificity for axial images alone) Used intraarticular saline solution

93/80/89 (Bankart lesions) 89/89/89 (SLAP lesions) 89/91/90 (SLAP lesions)

Found 76% of SLAP lesions correctly graded by MR arthrography

*As given in the original article; not recalculated or derived from data in the article. ABER, abduction, external rotation; SLAP, superior labrum, anterior and posterior.

the anteroposterior dimension of the glenoid relative to the equator. CT is the preferred modality for assessment of the bony glenoid. ABER images, although not routinely used, are helpful in assessment of labral damage. With the shoulder in abduction and external rotation, the anterior capsule and the anterior band of the inferior glenohumeral ligament are taut. The Perthes lesion, a Bankart variant, is characterized by labroligamentous avulsion with intact scapular periosteum (Fig. 61-28). There is no medial displacement of the labral periosteal complex. Hence Perthes lesions are better diagnosed on the ABER position when there is stress on the anteroinferior structures. Similarly they can be missed arthroscopically until probed. Cvitanic and colleagues found that, read together, axial and ABER MR arthrograms had a sensitivity of 96% and specificity of 97% for anterior glenoid labral tears.40 Pitfalls and variants in diagnosis of labral tears.  The appearance of the labrum has many normal variations. The CT arthrographic configuration of the posterior labrum is rounded, and the MRI appearance has been described as usually triangular. The configuration of the anterior labrum may vary considerably from one individual to another, and in the same person labral shape changes with rotation of the humerus.150,176 With internal rotation, the anterior labrum assumes a large globular configuration; with external rotation it is attenuated and pointed. With gradient echo imaging, an area of smooth linear high signal intensity in the anterior labrum may occasionally be seen. This finding alone does not indicate a tear; however, its cause has not been ascertained.86,129,134,159

Some normal structures may be misinterpreted as labral tears. The hyaline articular cartilage of the glenoid extends under and undercuts the base of the labrum.115,205 At MRI, this produces a linear area of increased signal intensity that may resemble a tear. At the level of the subscapularis tendon, the middle glenohumeral ligament runs parallel to the anterior labrum and also shows low signal intensity. The space between the ligament and the labrum may mimic a tear along the outer border of the labrum, but its true nature is usually revealed when one follows the ligament along its normal course on sequential images  (Fig. 61-29). Superiorly the sublabral recess or foramen can mimic a labral tear, particularly when intraarticular contrast material enters the recess.108 A collection parallel to the shape of the bony glenoid and extending medially is probably a normal structure, whereas a collection that turns in a lateral direction into the substance of the labrum is most likely a tear (Figs. 61-30 and 61-31).7,38 The exact location of the fluid collection may also provide a hint because a normal recess or foramen should not extend posterior to the biceps origin, whereas a labral tear often (but not invariably) does.213,232 An absence of the anterosuperior glenoid labrum may be paired with a thick, cordlike middle glenohumeral ligament, an anatomic variant called the Buford complex (Fig. 61-32).229 Interpretation of labral tears may be particularly difficult in elderly patients because of the considerable overlap between pathologic labral changes and changes associated with normal aging. Intrasubstance degenerative changes are noted in individuals after age 30 and become

CHAPTER 61  Shoulder

A

2121

B FIG 61-28  Perthes lesion. A, T1-weighted fat-saturated axial oblique MR arthrogram. The anterior labrum (arrow) has been avulsed from the glenoid rim, but unlike the anterior labroligamentous periosteal sleeve avulsion (ALPSA), there is no medial labral displacement. B, T1-weighted fat-saturated axial oblique MR arthrogram. Abduction and external rotation (ABER) allow the anteroinferior structures of the glenohumeral joint to be inspected. With stress applied, there is elevation of the anterosuperior labrum, thus confirming the diagnosis of a Perthes lesion (arrow). A posterior labral tear is also noted.

increasingly evident with advancing age.43,185 On PD-weighted or gradient echo images, these may produce globular increased signal intensity in the labrum, reflecting degeneration.129,159 The labrum may also appear small in older adults. Hill-Sachs lesions arise in the setting of anterior shoulder dislocation, where the dislocated humeral head impacts against the anteroinferior glenoid rim. The resultant depression on the humeral head is called a Hill-Sachs lesion. This mechanism of injury is also responsible for Bankart tears. In 2008, Purchase et al. described the remplissage procedure to prevent subsequent reengagement of Hill-Sachs lesions.187 The remplissage procedure is arthroscopic capsulotenodesis of the posterior capsule and infraspinatus tendon to fill the Hill-Sachs lesion. This has the effect of simultaneously repairing the bony defect and isolating it into an extraarticular location. In a published self-audit of 2- to 10-year results in 45 patients, Wolf and Arianjam found a 4.4% incidence of anterior instability due to recurrent injury but had no significant complications or revision surgery.252 Cho et al. used CT with three-dimensional (3D) reconstruction to define Hill-Sachs lesions in 107 patients. They found that engaging lesions tended to be horizontally oriented to the humeral shaft and had an angle of 25.6 degrees ± 7.4 degrees in engaging lesions, which was significantly larger than 13.8 degrees ± 6.2 degrees in nonengaging lesions (P < .001).31

Other Structures in Stability.  The anterior joint capsule and glenohumeral ligaments are also important in maintaining shoulder  stability.230 In abduction and external rotation, the glenohumeral

ligaments restrain the humeral head and prevent anterior instability. Because of the capsular tension produced by fluid in the joint, capsular abnormalities may be better diagnosed with an arthrographic technique (either CT or MRI).249a The anterior capsule may arise from various positions on the neck or edge of the glenoid or from the labrum.44,150 The origin is easily demonstrated by these techniques, but a suspected association between a more medial capsular origin and glenohumeral instability has not been substantiated.37,173 Glenohumeral dislocation can result in bony abnormalities that are sometimes visible on plain films but are clearly identified on either CT or MRI. As the humeral head dislocates anteriorly, it encounters  the anterior glenoid rim, producing a bony Bankart lesion (anterior glenoid rim fracture) or a Hill-Sachs deformity. On axial images, HillSachs lesions are visualized in the posterolateral border of the humeral head at or superior to the base of the coracoid process (Fig. 61-33). They must not be confused with the normal flattening of the humeral head as it transitions into the humeral neck (see Fig. 61-33),  which is usually visualized more posteriorly and inferiorly but may be found at the level of the coracoid in a patient with a large rotator cuff defect and superior translation of the humerus.98,192,205,254 In patients who have had a posterior dislocation, a “reverse” Hill-Sachs (or “trough” lesion) may be identified on the anteromedial margin of the head of the humerus in association with a posterior glenoid or labral abnormality. Anterior glenoid bone loss can occur in the setting of anteroinferior dislocations as well as from a direct fracture. Glenoid bone loss reduces

FIG 61-29  Pitfall in diagnosis of labral tears. Gradient echo axial image. The anterior labrum and capsule proved to be normal at surgery. The labrum (small arrow) is separated from the glenoid rim by an unusually thick and easily seen layer of articular cartilage. The middle glenohumeral ligament (large arrow) runs parallel to the labrum, separated from it by a hyperintense curvilinear area of connective tissue that might be interpreted as a tear.

FIG 61-31  SLAP (superior labrum, anterior and posterior) lesion as shown on T2-weighted fat-saturated coronal oblique MR arthrogram. Fluid (arrow) fills the cleft in the labrum and turns laterally, away from the glenoid. There is also tendinopathy and partial-thickness tear of the bursal surface of the supraspinatus tendon (arrowheads).

FIG 61-30  Sublabral recess as shown on T1-weighted fat-saturated coronal oblique MR arthrogram. A smooth-bordered collection of gadolinium (arrow) parallels the cortex of the superior glenoid.

FIG 61-32  Buford complex. There is absence of the anterosuperior glenoid labrum (large arrow), with a cordlike, thickened middle glenohumeral ligament (small arrow), in keeping with a Buford complex. This can be distinguished from a cartilaginous Bankart lesion by following the thickened middle glenohumeral ligament toward the anterior capsule.

CHAPTER 61  Shoulder

*

A

*

B

C FIG 61-33  Hill-Sachs deformity. A, T1-weighted fat-saturated axial MR arthrogram. At the level of the coracoid process (asterisk), a notch-shaped defect in the posterolateral humeral head (arrow) is typical of a HillSachs defect. The patient had a history of recurrent anterior dislocation. Arrowheads indicate the superior glenohumeral ligament. B, Axial CT at the level of the coracoid process (asterisk) demonstrates a focal defect in the posterolateral humeral head (arrow), in keeping with a Hill-Sachs deformity. C, Axial T1-weighted fatsaturated MR arthrogram at a section inferior to A demonstrates normal contour flattening of the lateral aspect of the humerus (arrow), not to be mistaken for a Hill-Sachs defect.

2123

2124

PART II  CT and MR Imaging of the Whole Body

the articular contact area and increases the frequency of future dislocations. Dislocation frequency increases when bone loss involves 15% of surface area, but 20% to 25% is commonly used as the threshold to designate severe bone loss. Mild (<20%) bone loss can be treated with capsular tightening procedures, but severe bone loss (>20%-25%) may require bone augmentation.73,139,221 On MRI, glenoid bone loss is readily detected on sagittal images. Griffiths et al. defined the pear-shaped appearance of the normal glenoid on sagittal reconstruction CT scans.72 This is lost with glenoid bone loss, resulting in an anterior line at the anteroinferior glenoid. The usual maximal amount of bone loss is about one third of   total glenoid surface area, inasmuch as the middle third is buttressed by the glenoid neck.72,73 Fracture progression will result in a concave anterior edge. 3D CT of the shoulder (with or without arthrogram) has been used to quantify the amount of bone loss with great accuracy compared with arthroscopy.72,74 MRI confers the advantages of no ionizing radiation and one-stop bone and soft tissue imaging, but CT allows comparison of left and right sides. In a cadaver study (n = 14), it was found that CT and MRI measurements of actual glenoid defect size using the best-fit circle area method had a high degree of agreement between both modalities and surgery.186 A separate study also found high interobserver agreement when comparing the best-fit circle width method at multidetector CT and high-spatial-resolution 3D fatsuppressed MRI with volumetric-interpolated breath-hold examination in 41 subjects with recurrent shoulder dislocation in the detection of glenoid bone loss.225 3D isotropic MRI scans (using the 3D Dixon 3D T1-weighted FLASH [fast low-angle shot] sequence) had a 2.2% discrepancy rate relative to surgery, with interobserver agreement of about 90%.77

POSTERIOR GLENOHUMERAL INSTABILITY Pathophysiology Compared with anterior instability, less has been written about posterior glenohumeral instability in the orthopedic and radiologic literature. It is thought to account for 5% of shoulder instabilities.255 Affected patients are usually men between 20 and 30 years of age, playing overhead or contact sports. Patients usually present after trauma with pain. One of the major causes of posterior dislocation is epilepsy, and a neurologic consultation should always be considered before surgery. The etiology of posterior instability is thought to be multifactorial, with psychogenic,174,197 positional,174,197,222 muscular,64 and macrotraumatic or microtraumatic factors contributing. The jerk test (a maneuver to subluxate the humerus posteriorly) may elicit pain, but frank posterior dislocation is unusual. Posterior instability can occur as a component of MDI.

Imaging of Posterior Glenohumeral Instability The constellation of findings in posterior glenohumeral instability include a reverse Hill-Sachs notch in the anteromedial humeral head (the trough sign) (Fig. 61-34). This may be responsible for joint locking if there is engagement of the compression fracture on the glenoid. A posterior labral tear may be associated with fracture of the glenoid rim. This can be accompanied by a paralabral cyst (Fig. 61-35). A variant of the posterior labral tears includes the posterior labrocapsular periosteal sleeve avulsion, where the posterior scapular periosteum is stripped off, with medial migration of the labrum. A posterior capsular tear may also be demonstrated but may not be detected reliably, even with MR arthrography. A posterior SLAP (superior labrum, anterior and posterior) lesion can also be seen as part of the posterior peelback mechanism. The

FIG 61-34  Posterior labral tear as shown on T1-weighted fat-saturated axial MR arthrogram. There is a posterior labral tear, as indicated by dissection of gadolinium-based contrast between the labral substance and the underlying glenoid cartilage (arrow). An associated trough fracture of the anterior humeral head is present (arrowhead). The humeral head is posteriorly subluxated.

humeral head can be subluxated posteriorly on the glenoid as a result. End-stage findings of degenerative joint disease may also be seen. In addition, the humeral neck capsule may be stripped off with disruption of the posterior band of inferior glenohumeral ligament, resulting in a reversed humeral avulsion of the glenohumeral ligament (RHAGL) lesion (Fig. 61-36).32 There is discontinuity of the posterior capsule, correlating with an absent posterior capsule and direct visualization of the posterior rotator cuff arthroscopically. Arthrographically, fluid can be seen extending along the humerus, as in conventional humeral avulsion of the glenohumeral ligament (HAGL) lesions. It is important to recognize tears at both the humeral and glenoid attachments of the posteroinferior glenohumeral ligament (PIGHL). These are termed floating PIGHL lesions.183 When a floating PIGHL lesion coexists with a posterior labral tear, both lesions should be repaired to reduce the incidence of recurrent posterior instability. In fact, results from the Norwegian shoulder registry reveal that recurrent instability after primary repair occurs in 10% of anterior instability cases and 16% of posterior instability cases. The authors postulate that the difference in recurrence rate (recurrent posterior instability is higher than anterior) is related to posterior capsular tightening procedures being carried out less often than anterior tightening procedures13,255 (Fig. 61-37). Occasionally extraarticular posterosuperior ossification can be  seen associated with posterior labral injury and posterior articular surface rotator cuff damage. This should not be confused with

CHAPTER 61  Shoulder

2125

FIG 61-35  Posterior paralabral cyst. Axial gradient echo image obtained

FIG 61-36  Reverse humeral avulsion of the glenohumeral ligament as

through the posterior labrum reveals a posterior paralabral cyst with internal debris (arrow). The posterior labral substance is irregular, in keeping with a tear. There is also an anterior labral tear. The humeral head is posteriorly subluxated.

shown on T1-weighted fat-saturated coronal MR arthrogram slightly posterior to the glenoid midpoint. The tear of the glenohumeral ligament adjacent to the humerus is indicated (arrow), with a glenohumeral ligament remnant (arrowhead).

posterior dislocation (but can be associated with it); rather, it is a chronic posterior impingement syndrome. This is termed a Bennett lesion (Fig. 61-38) and can interfere with the act of throwing. Reactive glenoid sclerosis may be seen. This lesion is arthroscopically occult. In a selected population of patients with posterior labral tear, a tear of 15 mm or greater in craniocaudal length on MR arthrography had a sensitivity and specificity of 83% for posterior instability.234 Weishaupt and associates found that a posteroinferior defect in glenoid shape with a craniocaudal length of 12 mm or more predicted posterior instability with a sensitivity of 87% and a specificity of 83%.246 Glenoid hypoplasia and glenoid retroversion are associated with posterior instabilities, potentially requiring treatment with a bone block. On imaging, glenoid version is more accurately measured on CT because the bony edges of the glenoid are better depicted. Following the technique described by Mintzer and colleagues, a line is drawn along the midbody of the scapula, bisecting the glenoid fossa. Another line connecting the anterior and posterior tips of the glenoid is drawn at the level of the glenoid neck, intersecting the first line. The acute angle thus subscribed is the alpha angle, with retroversion indicated by negative values. Mean glenoid version in normal individuals can vary between −3 and 2 degrees relative to the true sagittal plane.141 In arthritis of the shoulder, the posterior glenoid margin is eroded, resulting in retroversion. This accounts for posterior glenohumeral instability and subluxation in arthritis. Bone reconstructive surgery or total shoulder arthroplasties are planned with information about posterior glenoid bone loss derived radiographically or with CT.

FIG 61-37  Posteroinferior glenohumeral ligament injury (PIGHL lesion). Indirect MR arthrogram of the shoulder performed to evaluate pain 3 months after low-velocity traffic accident. Axial intermediate-weighted fat-saturated scan reveals abnormal signal at the glenohumeral ligament attachments on the posterior glenoid rim (arrow) and also at the humeral insertion (arrowhead). This correlated with a positive posterior load shift test, consistent with a PIGHL lesion.

2126

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 61-38  Bennett lesion. A, Sagittal T1-weighted MR arthrogram with fat saturation at the level of the posterior glenoid rim demonstrates an area of bony proliferation and periosteal thickening (arrowheads), in keeping with a Bennett lesion. B, Axial T2-weighted image confirms the presence of a posterior labral tear (arrow), as well as thick bony proliferation at the posterior glenoid rim. There are also tears of the subscapularis tendon and anterior labrum. (Images courtesy Dr. Phillip Tirman.)

SUPERIOR LABRUM LESIONS Pathophysiology In 1990, Snyder and colleagues divided SLAP lesions into four types, depending on the nature of the labral injury and whether there is associated biceps tendon damage222: Type I lesions: The labrum is frayed, but it and the biceps tendon are not torn from the glenoid. Type II lesions: The labrum is stripped from the glenoid with the biceps tendon anchor. This is unstable, and attention should be paid to the cartilage-labrum interface. There is normally undercutting of the labrum by the cartilage, but in the presence of a SLAP II lesion, increased T2 signal is seen tracking between them. This is also associated with anterior glenohumeral instability and its constellation of MRI findings. SLAP II lesions involving the superoanterior labrum associated with undersurface tears of supraspinatus are now subclassified as SLAC (superior labrum–anterior cuff) lesions. Those involving the posterosuperior labrum are associated with the peelback mechanism in overhead-throwing athletes. Type III lesions: There is a bucket-handle tear of the labrum. The biceps anchor is stable. These tears can extend circumferentially to involve the anterior labrum (SLAP V). Type IV lesions: The bucket-handle tear extends into the biceps tendon. In 1995, Maffet and coworkers expanded the classification of Snyder’s group to account for additional injury patterns119: Type V lesions: The tear involves the labrum from the anteroinferior position to the posterosuperior position.

Type VI lesions: One finds an unstable flap tear of the labrum and separation of the biceps tendon from its anchor. Type VII lesions: The labral tear extends anteriorly beneath the middle glenohumeral ligament. Not all SLAP injuries are the result of instability. They may also be caused by sudden violent traction on the arm, as occurs when catching a heavy object or by falling on an outstretched arm.24,71,80,92,107,111

Superior Labrum–Anterior Cuff Lesions.  The SLAC lesion is a result of glenohumeral instability. The SLAP II component of this lesion involves the anterior portion of the biceps tendon. The resultant anterosuperior instability results in contact between the articular surface of the anterior supraspinatus tendon and the glenoid, resulting in a partial-thickness articular supraspinatus tear over time.201

Peelback Lesion and Glenohumeral Internal Rotation Deficit.  The peelback phenomenon has been noted in pitchers with a tight posteroinferior capsule and glenohumeral internal rotation deficit (GIRD; Fig. 61-39). The tight posteroinferior capsule and inferior glenohumeral ligament results from repetitive follow-through motions after throwing, and it shifts the glenohumeral contact point superoposteriorly during abduction and external rotation.20 It allows for external hyperrotation by increasing the clearance for the greater tuberosity, and it also avoids direct cam internal impingement of the posterosuperior labrum/rotator cuff. Over time, anterior capsular laxity and loss of glenohumeral internal rotation in abduction occur, resulting in a thrower’s “dead arm.”

CHAPTER 61  Shoulder

A

B

2127

C

FIG 61-39  Glenohumeral internal rotation deficit (GIRD) illustration. A, In GIRD, the fulcrum of rotation in the glenohumeral joint is shifted posterosuperiorly (black circle to red circle). B, With repeated abduction and external rotation, especially in the late cocking phase of overhead throwing, there is posterior peelback and torsion of the biceps tendon, which is transmitted to the anchor. C, With time, tears of the posterosuperior labrum result in a SLAP II (superior labrum, anterior and posterior) tear, extending into the biceps anchor (white jagged line).

In addition, there is a posterior shift of the biceps vector, with twisting of the biceps attachment in the late cocking phase of throwing. The increased torsional forces are transmitted back to the biceps anchor and the posterosuperior labrum, resulting in a SLAP II lesion  (Fig. 61-40). This cascade is potentiated by scapular malposition. Scapular malposition is a component of the SICK (scapular malposition, inferior medial border prominence, coracoid pain and malposition, and dyskinesis of scapular movement) syndrome. It is an overuse syndrome and presents in athletes as a dropped scapula. As the scapula tilts anteriorly in flexion, there is protraction, abduction, and riding up over the thorax.20 Impingement-like symptoms of traction and pain arise from this malpositioning, with subsequent anteroinferior and posterosuperior labral abnormalities and articular-sided rotator cuff tears.21

CALCIUM DEPOSITION DISEASES Hydroxyapatite deposition disease, including its manifestations as (1) calcific tendinitis or bursitis, (2) idiopathic destructive arthritis (“Milwaukee shoulder syndrome”), and (3) tumoral calcinosis, may occur in the shoulder.

Calcific Tendinitis Calcific tendinitis involves most frequently the supraspinatus, followed by infraspinatus, teres minor, and subscapularis tendons. Patients are typically in middle age, although the disease has been reported in a child as young as 3 years. Women are affected more commonly than men. The dominant shoulder is more likely to be affected. The cause is unknown, but calcific tendinitis is believed by some to be the result of calcification of metaplastic cartilage in the relatively ischemic distal supraspinatus tendon. The deposits may also extrude from the tendon into a bursa, usually the subacromial-subdeltoid bursa, producing calcific bursitis. The deposits may erode bone. Pain is the presenting complaint and seems to be correlated with cloudy ill-defined deposits

occurring with resorption, rather than with the well-defined deposits more commonly seen in asymptomatic patients.132,165,235,242 Microscopically, crystalline hydroxyapatite, inflammatory cells, macrophages, and multinuclear giant cells as well as fibroblasts are seen. From 3% to 20% of shoulders with radiographically demonstrable calcification are asymptomatic, and clinical correlation is required to make the diagnosis. Plain films usually are sufficient for the diagnosis. The calcium deposits are readily apparent on CT scans. If an MRI study is obtained, a small calcium deposit may be difficult to detect but typically demonstrates low signal intensity on all sequences, with apparent enlargement of the signal void on T2*-weighted sequences (Fig. 61-41A and B). Hyperintense perilesional edema can be demonstrated on T2-weighted, T2-weighted fat-saturated, or inversion-recovery images (see Fig. 61-41C and D). If ossification has occurred and marrow is present in the deposit, the signal intensity matches that of other marrow. Bursal deposits have the same appearance as those in the tendons but are often accompanied by bursal fluid, leading to a condition referred to as adhesive periarthritis.

Idiopathic Destructive Arthritis Idiopathic destructive arthritis of the shoulder, or Milwaukee shoulder syndrome (named for the city where it was first described), is another calcium crystal–related abnormality. It is most often seen in elderly women.248 It is characterized by large blood-tinged effusions containing calcium hydroxyapatite and calcium pyrophosphate dehydrate crystals, and by extensive destruction of cartilage and bone. It is often accompanied by derangement of the rotator cuff or the proximal tendon of the long head of the biceps brachii. The calcium-containing effusions distinguish it from simple rotator cuff arthropathy, as does a tendency toward bilateral involvement and involvement of other joints, particularly the hips and knees. The diagnosis of Milwaukee shoulder syndrome is generally made clinically. CT and MRI studies, if obtained, demonstrate the effusions

2128

PART II  CT and MR Imaging of the Whole Body

A

B FIG 61-40  Posterior impingement, peelback, and glenohumeral internal rotation deficit (GIRD). A, T1-weighted fat-saturated sagittal MR arthrogram reveals posterosuperior displacement of the humeral head relative to the glenoid (arrow). B, T1-weighted fat-saturated axial MR arthrogram confirms the presence of posterior labral tear (arrow) and periosteal thickening, in keeping with chronic posterior impingement/GIRD. There is also a tear of the anterior labrum. (Images courtesy Dr. Phillip Tirman.)

and destruction of articular surfaces. CT may also demonstrate calcium deposition on synovial surfaces.78,79,128,163

Periarticular Calcinosis and Tumoral Calcinosis There are multiple causes of periarticular calcinosis, spanning the gamut from neoplastic (e.g., synovial sarcomas), to trauma, to calcinosis of renal impairment. Tumoral calcinosis is a rare inherited condition caused by dysregulation of phosphate metabolism resulting in amorphous, often liquid, deposits of calcium forming near the joints, particularly the shoulder, hip, and elbow. It has been reported concurrently in siblings. The diagnosis is made on plain radiographs along with history, physical examination, and laboratory studies. Neither CT nor MRI is usually necessary. Occasionally, however, the plain film appearance of such masses may be confusing, and CT may help distinguish them from other causes of periarticular calcinosis.11,28,118,139,162 Although it is not the first line of investigation, MRI can be used to evaluate the integrity of soft tissue components in less intensely calcified masses.38,53

ADHESIVE CAPSULITIS This remains a clinical diagnosis. MRI is sometimes obtained to establish other concomitant pathologic processes rather than make the primary diagnosis. Mengiardi and coworkers found that in an arthroscopic capsulotomy population, MRI demonstrated thickening of the coracohumeral ligament at 4.1 mm versus 2.7 mm. The thickness of the rotator cuff interval capsule was also increased at 7.1 mm

versus 4.5 mm.136 Thickening of the inferior glenohumeral ligament, especially on coronal PD-weighted images,220 is often accompanied by synovial hypertrophy (Fig. 61-42). Carbone et al. proposed the use of subscapularis recess distension on MRI (without arthrogram) as an adjunct to diagnose adhesive capsulitis. In their retrospective review, the authors found that distension of the subscapularis recess was present it up to 90% of cases with freezing-phase (phase 1) adhesive capsulitis. The sign was positive in less than 6% of cases with superior rotator cuff tears and in 1.5% of control group patients. These results imply that even in situations where rotator cuff tears coexist with adhesive capsulitis (a common occurrence in clinical practice), subscapularis recess distension should raise suspicion of this pathology in patients presenting with painful arc syndrome.23

RHEUMATOID ARTHRITIS Rheumatoid arthritis (RA) is a common disease with a predilection for women of middle age. The glenohumeral joint, acromioclavicular joint, or both may be involved, but it is unusual for the disease to present in the shoulder before it has manifested elsewhere.39,133 In the glenohumeral joint, as elsewhere, RA is characterized by osteopenia, effusions, progressive loss of joint space, and osseous erosions. The effusions may cause outpouchings of the synovium; when very large they are termed giant synovial cysts. Acromioclavicular involvement has a similar appearance, except that the scapula and clavicle are held in their respective locations by surrounding structures

CHAPTER 61  Shoulder

A

2129

B

C

D FIG 61-41  Calcific tendinitis. A, Gradient echo axial image. A large area of signal void (arrow) is present near the attachment site of the supraspinatus tendon. B, T2-weighted coronal oblique image. Signal void is present in the relatively avascular area of the supraspinatus tendon (arrow). This area was also hypointense on T2-weighted images. C, T2-weighted fat-saturated coronal oblique image demonstrating a large focus of signal void in the distal infraspinatus tendon, with surrounding edema (arrow). D, Radiograph of the same structure as in C with humeral internal rotation, demonstrating calcification in the region of the infraspinatus tendon.

and do not migrate closer to one another with destruction of the joint surfaces. Thus the acromioclavicular joint space widens rather than narrows with rheumatic involvement.39,133 Inflammatory arthritides, especially RA, predispose the patient to tendon and ligament ruptures. In the shoulder, they may lead to rotator cuff tears. RA may be identified by clinical and serologic means. Radiographic techniques are less useful in diagnosis than in plotting progression of disease and response to therapy. MRI is more sensitive than plain

radiographs to the presence of effusions and small erosions and may directly depict pannus (Fig. 61-43). Use of intravenous gadolinium helps distinguish pannus from effusions. Both may appear low in signal intensity on nonenhanced T1-weighted images and high in signal intensity on T2-weighted images, but pannus enhances. Therefore MRI may have a role in evaluating therapeutic response. The hands and wrists are involved more often than the shoulder.6,60,196,249 At present there is uncertainty about the exact nature of “erosions” detected on MRI. Conventionally, erosions are defined as

2130

PART II  CT and MR Imaging of the Whole Body At this point, there is no exact definition of “early” or “late” RA. Large erosions can be seen in acute and late stages of disease, suggesting that their presence is not completely dependent on disease duration but rather on factors such as inflammatory activity.223

PIGMENTED VILLONODULAR SYNOVITIS Pigmented villonodular synovitis (PVNS) is an uncommon disorder of unknown cause. It usually affects young adults and is typically monoarticular, with most frequent involvement of the knee. It is  rare in the shoulder.47,59,207 Histologically, PVNS demonstrates fatty areas together with hemosiderin deposition. This results in a heterogeneous appearance on both CT and MRI. On MRI the hemosiderin creates areas of decreased signal intensity on intermediate weighting. The signal intensity decrease becomes more prominent on T2  weighting. On gradient echo imaging, the paramagnetic effect of hemosiderin causes even more dramatic signal loss. Effusions may be present.22,91,102,106,123,218

NEUROGENIC SHOULDER DYSFUNCTION Suprascapular Nerve

FIG 61-42  Adhesive capsulitis. T1-weighted fat-saturated coronal MR arthrogram demonstrates thickening of the inferior glenohumeral ligament (arrow). There is also lack of capsular distention with intraarticular contrast, in keeping with adhesive capsulitis. (Courtesy Dr. Phillip Tirman.)

surface lesions in bone and cartilage caused by inflammation in joints due to infiltration of lymphocytes, macrophages, and neutrophils. Thickened synovial tissue is formed as a result of proliferation of highly vascular connective tissue in the synovium. This synovium eventually invades bone to produce erosions, seen initially at the joint margins where cartilage is absent (bare areas). The inflammation progresses to destroy cartilage, bone, joint capsule, ligaments, and tendons. MRI is not as sensitive as radiographs (with or without magnification techniques) or CT,184 but it demonstrates subchondral signal change with greater sensitivity.28 The median time that MRI detection precedes radiographic detection of new erosions is currently thought to be around 2 years.60 These abnormal areas show focal loss of normal signal intensity of bone on T1-weighted images and hyperintense T2-weighted signal. However, pannus-filled and fluid-filled spaces in the bone can show similar T1- and T2-weighted signal on MRI. Hence contrast-enhanced T1-weighted fat-saturated sequences are required to differentiate the two. Pannus shows avid enhancement because of its inherent vascularity, whereas fluid-filled cysts show enhancement peripherally. If the joint is imaged too long after contrast injection (typically ≈ 15-20 minutes), contrast diffusion may occur—a pitfall seen in small joints (e.g., in the hand). Furthermore, larger intraosseous lesions on MRI should be differentiated from marrow space edema by the fact that the latter has no distinct margin and shows less intense enhancement.223

Pressure on the suprascapular nerve as it courses across the posterior scapula is an uncommon cause of shoulder pain. Any mass that arises in an appropriate location can produce suprascapular nerve entrapment; reported causes include neoplasms and hematoma. The most common cause, however, is a ganglion cyst. These cysts usually occur in young men. They typically arise from a labral tear in much the same fashion that a meniscal cyst arises from a torn meniscus.228,244 Ganglion cysts may occur at various sites about the shoulder, but those causing suprascapular nerve compression usually occur at the spinoglenoid notch (Fig. 61-44). They may extend superiorly into the suprascapular notch or inferiorly into the infraspinatus fossa. They may produce muscle paralysis and atrophy as well as pain. The infraspinatus muscle is most often affected, but the supraspinatus muscle may be involved as well. Ganglia may be detected as a water-density mass by CT. On MRI they are approximately isointense to muscle on short-TE images and hyperintense on T2-weighted images. If intravenous gadoliniumdiethylenetriamine pentaacetic acid (Gd-DTPA) is administered, the cysts may enhance about their rim. They often appear septate. Muscle atrophy, if it has ensued, may also be apparent on MRI; the earliest sign of denervation atrophy is increased signal intensity on T2-weighted images in the involved muscle. They can be aspirated under CT or ultrasonographic guidance. Some patients have successfully undergone surgery, and occasionally these cysts have resolved spontaneously.62,168 When the cyst arises from a labral tear, the tear must be addressed.228,244 Suprascapular nerve dysfunction may also be caused by direct injury to the nerve. One reported case was associated with a previous clavicular fracture. MRI assisted in the diagnosis by demonstrating abnormally high signal intensity on T2-weighted images in the belly of the supraspinatus and infraspinatus muscles.255

Acute Brachial Neuritis Acute brachial neuritis, or Parsonage-Turner syndrome, presents with a sudden onset of intense burning shoulder pain and muscle weakness, most commonly involving the supraspinatus and infraspinatus muscles, although other shoulder muscles may be involved as well. The patients are most typically young to middle-aged adults. There is a marked male predominance. This is a self-limited disease with resolution first of the pain. Muscle weakness may last years but usually

CHAPTER 61  Shoulder

B

A

FIG 61-43  Rheumatoid arthritis. A, T2-weighted fat-saturated coronal oblique image. There is uniform loss of articular cartilage. Multiple T2-weighted high-signal-intensity foci are noted in the subchondral bone, likely cysts. These findings may also be seen in late-stage osteoarthritis. B, Axial gradient echo image in a patient with RA confirming uniform loss of articular cartilage and subchondral cysts. There is a focal erosion on the posterior aspect of the humeral head.

* * A

B FIG 61-44  Spinoglenoid notch ganglion. A, Axial gradient echo image at the level of the superior labrum reveals a fluid-filled structure in the spinoglenoid notch ganglion (asterisk). B, Coronal T2-weighted image reveals that the cyst (asterisk) has occupied the spinoglenoid notch and displaced the belly of the adjacent supraspinatus superiorly. No associated muscle atrophy is present.

2131

2132

PART II  CT and MR Imaging of the Whole Body

TRAUMA

FIG 61-45  Parsonage-Turner syndrome. Sagittal T2-weighted fatsaturated image of the shoulder reveals edema in the supraspinatus and infraspinatus muscles, suggestive of brachial neuritis.

resolves in 2 to 3 months. MRI may reveal increased signal intensity in the involved muscles on T2 weighting (Fig. 61-45). Later in the course of the disease, muscle wasting without fatty infiltration may be evident. MRI is chiefly useful in excluding other causes of shoulder pain and dysfunction, particularly compressive masses in the spinoglenoid notch.83

Quadrilateral Space Syndrome Another cause of shoulder pain and dysfunction is quadrilateral space syndrome, in which the axillary nerve is compressed by fibrous bands as it traverses the quadrilateral space. This space is bounded by the teres minor and teres major muscles, the long head of the triceps muscle, and the surgical neck of the humerus. It manifests clinically with poorly localized shoulder pain, lateral shoulder and posterior upper arm paresthesias, and weakness and atrophy of the teres minor or deltoid muscles. MRI does not show the actual entrapment but may demonstrate denervation atrophy of the associated muscles and exclude other causes of shoulder pain.114 Angiography has been used to confirm the diagnosis on the basis of occlusion of the posterior humeral circumflex artery with the arm in abduction. In 1994, Mochizuki and coworkers cast doubt on the usefulness of this approach or, more specifically, of using MR angiography in place of conventional contrast angiography, when they found that MR angiography demonstrated occlusion of 80% of the posterior humeral circumflex arteries of six asymptomatic volunteers with abduction.144

Most fractures about the shoulder are adequately evaluated with plain radiography. In cases of complex fractures of the proximal humerus, CT may be helpful for pretreatment planning. Treatment of these fractures varies considerably depending on their position in the widely used Neer classification.153 This classification is based on the degree of separation or rotation of the four possible major fracture fragments (shaft, head, greater tuberosity, and lesser tuberosity). A fragment is considered displaced if it is distracted 1 cm or more from its neighbors or if it is angulated 45 degrees or more from its proper location.153 Fractures without displacement are commonly treated with early mobilization, whereas those involving displacement of one of the tuberosities often require open reduction. It is important, therefore, to determine accurately the position of all fragments before definitive treatment is carried out. This may be difficult with plain radiographs alone. Castagno’s group found that in 12 cases of acute fractures and 5 cases of persistent limitation of motion after treatment of fracture, CT provided useful information that was not evident on plain films.25 Even when the information is available on plain films, evaluation of fragment location may be difficult, as evidenced by reports of high interobserver variability in classifying humeral fractures according to Neer’s criteria. Furthermore, CT may be useful because it can provide a much more graphic depiction of fragments, particularly when 3D reconstruction is available. In addition to localizing major fracture fragments, CT can locate smaller ones; this is particularly useful for intraarticular fragments. Loose bodies of nontraumatic origin may also be located with CT, of course (Fig. 61-46).25,84,100,102,155 CT may be used to detect complex scapular fractures. These are rare injuries, usually seen in patients who have sustained multiple injuries as a result of direct blunt trauma, often an automobile accident.  These fractures may be missed on initial plain radiography, particularly on the portable chest films often obtained on patients who have  sustained severe trauma. When a suspected scapular fracture has not been revealed on such a film, it may be easier as well as more diagnostically rewarding to obtain a CT study rather than plain films of the shoulder.1,81,130,224 Treatment of scapular fractures is less well established than treatment of humeral fractures. Some surgeons believe that displaced fractures of the glenoid and scapular spine should receive internal fixation; others suggest that these injuries, like other scapular fractures, may be satisfactorily treated by immobilization followed by range-of-motion exercises shortly thereafter. In addition, the patient’s other injuries may preclude surgical intervention. When treatment hinges on the site of the fracture and the degree of displacement, however, CT may be more easily obtained and more revealing than plain films.1,41,131 MRI may reveal occult fractures, which appear as a line of abnormal signal intensity extending through cancellous bone to reach a cortical surface. Rotator cuff tears associated with greater tuberosity fractures can also be assessed.257 MRI can also demonstrate soft tissue injuries such as the glenoid labrum and cartilage disruption. Injuries to the muscles, such as the pectoralis major, can also be demonstrated. MRI findings on PD-weighted images include hyperintense edema in the clavicle, sternum, humerus, or ribs, reflecting injury to the attachment sites of the pectoralis major tendons. Edema and hemorrhage in the muscle, myotendinous junction, perifascial zone, or subcutaneous fat can also be demonstrated. The degree of tendon retraction can also be assessed. In addition, both CT and MRI are important to demonstrate incidental findings associated with trauma, such as a hemothorax or pneumothorax.

CHAPTER 61  Shoulder

2133

B C A

C

A

FIG 61-46  Humeral fracture. A, Radiograph. The humerus has been fractured. A crescent of bone typical of a detached greater tuberosity projects superolaterally to the humeral head. B, Axial CT scan demonstrates more graphically than the radiograph the amount of damage that has occurred to the region of the greater tuberosity. C, This 3D reconstruction shows the relationship of the detached greater tuberosity (white arrow) to the remainder of the humerus. A, acromion; C, clavicle.

IMAGING OF SHOULDER ARTHROPLASTY The end point of many conditions described in this chapter is severe functional immobility due to advanced osteoarthrosis. Irreparable rotator cuff tendon tears result from chronic tears, failed cuff surgery, arthritis, and humeral head fractures. Management of patients with cuff arthropathy can be challenging. When pain is severe and function is poor, surgical options include hemiarthroplasty, bipolar arthroplasty, extended head arthroplasty, arthroplasty with tendon transfer, reverse shoulder arthroplasty, and fusion.245 Reverse shoulder arthroplasty has been found to restore functional motion in patients with pseudoparalysis, failed fracture treatment, or failed hemiarthroplasty.245 The Norwegian shoulder registry records a 6% to 8% revision rate within 5 to 10 years for shoulder hemiarthroplasty surgery. For the reversed total shoulder arthroplasty, the results were 20% revision at 10 years. Glenoid loosening was the most common cause of revision.58 The components of a shoulder arthroplasty include a metallic humeral component, a lateralized polyethylene cup lining the articulating surface, a glenosphere, and a metaglene, which is anchored to the glenoid. Radiographically the humeral component and glenosphere as well as metaglene screw can be seen in alignment (Fig. 61-47).

Preoperative assessment for shoulder arthroplasty includes assessment of the axillary border of the scapula. A squared-off axillary border (vs. the normal sloped morphology) will require positioning the metaglene higher for DePuy or Encore prostheses. The reason is that these prostheses have a fixed angle for the inferior transfixing screw, which protrudes through bone in a squared-off glenoid neck. In other cases, the metaglene should be positioned as low as possible to avoid scapular notching by the humeral component, using additional bone purchase afforded by a gently sloping lateral scapular border. CT is used to preoperatively assess glenoid bone stock. The glenoid bone depth should be about 2 cm for placement of metaglene screws. MRI of the rotator cuff should include assessment of cuff tears, including retraction of the tendon end and the subjective degree of muscle atrophy. These have been discussed in the section on rotator cuff pathology. Specific attention should be paid to the integrity of the teres minor tendon and muscle. An intact teres minor results in better postoperative external rotation.15 The tendon insertion is resected in revision arthroplasty surgery, resulting in reduced external rotation in these cases. Radiographic assessment of shoulder arthroplasty requires examination and comment on the metaglene in relation to the glenoid

2134

PART II  CT and MR Imaging of the Whole Body

M

G

H

FIG 61-47  Reverse shoulder prosthesis. M represents the metaglene

FIG 61-48  Periprosthetic erosion. Coronal reconstructed CT of the

and its attached screws used to anchor the implant to the glenoid. It is important to assess the position of the screws with respect to bony margins, as well as implant integrity. G represents the glenoid component. H represents the humeral component. The humeral component may be posteriorly subluxed with respect to the glenoid component on axillary projections in a normal reverse shoulder arthroplasty.

shoulder reveals periprosthetic erosion adjacent to a Delta reverse shoulder prosthesis. Notching usually occurs on the inferior glenoid notch. Other sites of erosions include around the glenosphere, the screws, and polyethylene liner wear (usually inferior portion).

ACKNOWLEDGMENTS (specifically, and bony defects that may require grafting), placement of the inferior screw within the scapula (the other screws may be bicortical), the glenoid component alignment (should align with the humeral cup), and signs of radiolucency in the prosthesis-bone or cement-bone interface. There should be no hardware fracture. The inferior border of the scapula should be examined for erosions as well194 (Fig. 61-48).

CONCLUSION For any of these modalities to be maximally effective in a patient’s care, there must be careful correlation between clinical presentation and imaging results. In fact, preevaluation MRI (i.e., MRI obtained before clinical assessment) for patients with chronic shoulder pain rarely leads to a significantly different outcome than in patients who do not undergo MRI.16 In asymptomatic individuals, there is a 26% to 34% prevalence of MRI findings of rotator cuff tear.191,208 The purpose of imaging is to influence the management of patients and aid in presurgical planning. Blanchard and colleagues in 1996 found that MRI led to new diagnoses in 20 (34%) of 59 patients. For 25 of the 52 unchanged diagnoses (some patients had more than one diagnosis), MRI increased the referring physician’s confidence in the diagnosis.12 Bearcroft and associates examined the literature on the effectiveness of MRI for shoulder disorders and found that MRI alters the primary diagnosis in 23% to 68% of cases and led to changing the management plans in 15% to 61% of patients.4 This wide variation reflects the differences in practice across the surveyed institutions.

We acknowledge the contributions of Drs. Haygood, Oxner, Kneeland, and Dalinka pertaining to anatomy and rotator cuff disease.

REFERENCES 1. Ada JR, Miller ME: Scapular fractures: Analysis of 113 cases. Clin Orthop Relat Res 269:174–180, 1991. 2. Bankart AS: The pathology and treatment of recurrent dislocation of the shoulder joint. Br J Surg 26(101):23–29, 1938. 3. Bartolozzi A, Andreychik D, Ahmad S: Determinants of outcome in the treatment of rotator cuff disease. Clin Orthop Relat Res 308:90–97, 1994. 4. Bearcroft PWP, Blanchard TK, Dixon AK, et al: An assessment of the effectiveness of magnetic resonance imaging of the shoulder: Literature review. Skeletal Radiol 29(12):673–679, 2000. 5. Beltran J, Caudill JL, Herman LA, et al: Rheumatoid arthritis: MR imaging manifestations. Radiology 165(1):153–157, 1987. 6. Beltran J, Rosenberg ZS, Chandnani VP, et al: Glenohumeral instability: Evaluation with MR arthrography. Radiographics 17(3):657–673, 1997. 7. Bencardino JT, Beltran J, Rosenberg ZS, et al: Superior labrum anterior-posterior lesions: Diagnosis with MR arthrography of the shoulder. Radiology 214(1):267–271, 2000. 8. Bennett WF: Subscapularis, medial, and lateral head coracohumeral ligament insertion anatomy: Arthroscopic appearance and the incidence of” hidden” rotator interval lesions. Arthroscopy 17(2):173–180, 2001. 9. Bigliani LU: The morphology of the acromion and its relationship to rotator cuff tears. Orthop Trans 10:228, 1986. 10. Bigliani LU, Levine WN: Current concepts review–Subacromial impingement syndrome*. J Bone Joint Surg 79(12):1854–1868, 1997.

CHAPTER 61  Shoulder 11. Bishop AF, Destouet JM, Murphy WA, et al: Tumoral calcinosis: Case report and review. Skeletal Radiol 8(4):269–274, 1982. 12. Blanchard TK, Mackenzie R, Bearcroft PWP, et al: Magnetic resonance imaging of the shoulder: Assessment of effectiveness. Clin Radiol 52(5):363–368, 1997. 13. Blomquist J, Solheim E, Liavaag S, et al: Shoulder instability surgery in Norway: The first report from a multicenter register, with 1-year follow-up. Acta Orthop 83(2):165–170, 2012. 13a. Bokor DJ, Conboy VB, Olson C: Anterior instability of the glenohumeral joint with humeral avulsion of the glenohumeral ligament. J Bone Joint Surg Br 81:93–96, 1999. 14. Bonutti PM, Norfray JF, Friedman RJ, et al: Kinematic MRI of the shoulder. J Comput Assist Tomogr 17(4):666–669, 1993. 15. Boulahia A, Edwards TB, Walch G, et al: Early results of a reverse design prosthesis in the treatment of arthritis of the shoulder in elderly patients with a large rotator cuff tear. Orthopedics 25(2):129–133, 2002. 16. Bradley MP, Tung G, Green A: Overutilization of shoulder magnetic resonance imaging as a diagnostic screening tool in patients with chronic shoulder pain. J Shoulder Elbow Surg 14(3):233–237, 2005. 17. Bright AS, Torpey B, Magid D, et al: Reliability of radiographic evaluation for acromial morphology. Skeletal Radiol 26(12):718–721, 1997. 17a. Burk DL, Jr, Karasick D, Kurtz AB, et al: Rotator cuff tears: Prospective comparison of MR imaging with arthrography, sonography, and surgery. AJR Am J Roentgenol 153:87–92, 1989. 18. Burkhart SS, Danaceau SM, Pearce CE, Jr: Arthroscopic rotator cuff repair: Analysis of results by tear size and by repair technique-margin convergence versus direct tendon-to-bone repair. Arthroscopy 17(9):905–912, 2001. 19. Burkhart SS, Esch JC, Jolson RS: The rotator crescent and rotator cable: An anatomic description of the shoulder’s “suspension bridge”. Arthroscopy 9(6):611–616, 1993. 20. Burkhart SS, Morgan CD, Ben Kibler W: The disabled throwing shoulder: Spectrum of pathology. Part III: The SICK scapula, scapular dyskinesis, the kinetic chain, and rehabilitation. Arthroscopy 19:641– 661, 2003. 21. Burkhart SS, Morgan CD, Kibler WB: The disabled throwing shoulder: Spectrum of pathology. Part I: pathoanatomy and biomechanics. Arthroscopy 19:404–420, 2003. 22. Butt WP, Hardy G, Ostlere SJ: Pigmented villonodular synovitis of the knee: Computed tomographic appearances. Skeletal Radiol 19(3):191– 196, 1990. 22a. Callaghan JJ, McNiesh LM, DeHaven JP, et al: A prospective comparison study of double contrast computed tomography (CT) arthrography and arthroscopy of the shoulder. Am J Sports Med 16(1):13–20, 1988. 23. Carbone S, Napoli A, Gumina S: MRI of adhesive capsulitis of the shoulder: Distension of the bursa in the superior subscapularis recess is a suggestive sign of the pathology. Eur J Radiol 83(2):345–348, 2014. 24. Cartland JP, Crues JV, 3rd, Stauffer A, et al: MR imaging in the evaluation of SLAP injuries of the shoulder: Findings in 10 patients. AJR Am J Roentgenol 159(4):787–792, 1992. 25. Castagno AA, Shuman WP, Kilcoyne RF, et al: Complex fractures of the proximal humerus: Role of CT in treatment. Radiology 165(3):759–762, 1987. 26. Cervilla V, Schweitzer ME, Ho C, et al: Medial dislocation of the biceps brachii tendon: Appearance at MR imaging. Radiology 180(2):523–526, 1991. 27. Chan TW, Dalinka MK, Kneeland JB, et al: Biceps tendon dislocation: Evaluation with MR imaging. Radiology 179(3):649–652, 1991. 28. Chan WP, Lang P, Stevens MP, et al: Osteoarthritis of the knee: Comparison of radiography, CT, and MR imaging to assess extent and severity. AJR Am J Roentgenol 157(4):799–806, 1991. 29. Chandnani VP, Yeager TD, DeBerardino T, et al: Glenoid labral tears: Prospective evaluation with MRI imaging, MR arthrography, and CT arthrography. AJR Am J Roentgenol 161(6):1229–1235, 1993. 30. Chew FS, Crenshaw WB: Idiopathic tumoral calcinosis. AJR Am J Roentgenol 158(2):330, 1992.

2135

31. Cho SH, Cho NS, Rhee YG: Preoperative analysis of the Hill-Sachs lesion in anterior shoulder instability: How to predict engagement of the lesion. Am J Sports Med 39(11):2389–2395, 2011. 32. Chung CB, Sorenson S, Dwek JR, et al: Humeral avulsion of the posterior band of the inferior glenohumeral ligament: MR arthrography and clinical correlation in 17 patients. Am J Roentgenol 183(2):355–359, 2004. 33. Clark JM, Harryman DT: Tendons, ligaments, and capsule of the rotator cuff. Gross and microscopic anatomy. J Bone Joint Surg 74(5):713–725, 1992. 34. Clark J, Sidles JA, Matsen FA: The relationship of the glenohumeral joint capsule to the rotator cuff. Clin Orthop Relat Res 254:29–34, 1990. 35. Codman EA, Peltier LF: The classic: Rupture of the supraspinatus tendon. Clin Orthop Relat Res 254:3–26, 1990. 36. Cooper DE, Arnoczky SP, O’Brien SJ, et al: Anatomy, histology, and vascularity of the glenoid labrum. An anatomical study. J Bone Joint Surg 74(1):46–52, 1992. 37. Coumas JM, Waite RJ, Goss TP, et al: CT and MR evaluation of the labral capsular ligamentous complex of the shoulder. AJR Am J Roentgenol 158(3):591–597, 1992. 38. Crim JR, Seeger LL, Yao L, et al: Diagnosis of soft-tissue masses with MR imaging: Can benign masses be differentiated from malignant ones? Radiology 185(2):581–586, 1992. 39. Cruess RL: Rheumatoid arthritis of the shoulder. Orthop Clin North Am 11(2):333–342, 1980. 40. Cvitanic O, Tirman PF, Feller JF, et al: Using abduction and external rotation of the shoulder to increase the sensitivity of MR arthrography in revealing tears of the anterior glenoid labrum. AJR Am J Roentgenol 169(3):837–844, 1997. 41. Dalinka MK, Boorstein JM, Zlatkin MB: Computed tomography   of musculoskeletal trauma. Radiol Clin North Am 27(5):933–944, 1989. 42. Davis SJ, Teresi LM, Bradley WG, et al: Effect of arm rotation on MR imaging of the rotator cuff. Radiology 181(1):265–268, 1991. 43. De Palma AF: Surgery of the shoulder, Philadelphia, 1983, Lippincott. 44. Deutsch AL, Resnick D, Mink JH, et al: Computed and conventional arthrotomography of the glenohumeral joint: Normal anatomy and clinical experience. Radiology 153:603–609, 1984. 45. Dierckman BD, Shah NR, Larose CR, et al: Non-insertional tendinopathy of the subscapularis. Int J Shoulder Surg 7(3):83–90, 2013. 46. Dines DM, Warren RF, Inglis AE, et al: The coracoid impingement syndrome. J Bone Joint Surg Br 72(2):314–316, 1990. 47. Dorwart RH, Genant HK, Johnston WH, et al: Pigmented villonodular synovitis of the shoulder: radiologic-pathologic assessment. AJR Am J Roentgenol 143(4):886–888, 1984. 48. Edelson JG, Zuckerman J, Hershkovitz I: Os acromiale: Anatomy and surgical implications. J Bone Joint Surg Br 75(4):551–555, 1993. 49. Ellman H, Kay SP: Arthroscopic subacromial decompression for chronic impingement. Two- to five-year results. J Bone Joint Surg Br 73(3):395–398, 1991. 50. Epstein RE, Schweitzer ME, Frieman BG, et al: Hooked acromion: Prevalence on MR images of painful shoulders. Radiology 187(2):479– 481, 1993. 51. Erickson SJ, Cox IH, Hyde JS, et al: Estkowski LD. Effect of tendon orientation on MR imaging signal intensity: A manifestation of the “magic angle” phenomenon. Radiology 181(2):389–392, 1991. 52. Erickson SJ, Fitzgerald SW, Quinn SF, et al: Long bicipital tendon of the shoulder: Normal anatomy and pathologic findings on MR imaging. AJR Am J Roentgenol 158(5):1091–1096, 1992. 53. Erlemann R, Reiser MF, Peters PE, et al: Musculoskeletal neoplasms: Static and dynamic Gd-DTPA-enhanced MR imaging. Radiology 171(3):767–773, 1989. 53a. Evancho AM, Stiles RG, Fajman WA, et al: MR imaging diagnosis of rotator cuff tears. AJR Am J Roentgenol 151(4):751–754, 1988. 54. Farley TE, Neumann CH, Steinbach LS, et al: Full-thickness tears of the rotator cuff of the shoulder: Diagnosis with MR imaging. AJR Am J Roentgenol 158(2):347–351, 1992.

2136

PART II  CT and MR Imaging of the Whole Body

55. Farley TE, Neumann CH, Steinbach LS, et al: The coracoacromial arch: MR evaluation and correlation with rotator cuff pathology. Skeletal Radiol 23(8):641–645, 1994. 56. Ferrari DA: Capsular ligaments of the shoulder. Anatomical and functional study of the anterior superior capsule. Am J Sports Med 18(1):20–24, 1990. 57. Ferrick MR: Coracoid impingement. A case report and review of the literature. Am J Sports Med 28(1):117–119, 2000. 58. Fevang B-TS, Lie SA, Havelin LI, et al: Risk factors for revision after shoulder arthroplasty: 1,825 shoulder arthroplasties from the Norwegian Arthroplasty Register. Acta Orthop 80(1):83–91, 2009. 59. Flandry F, Norwood LA: Pigmented villonodular synovitis of the shoulder. Orthopedics 12(5):715–718, 1989. 59a. Flannigan B, Kursunoglu-Brahme S, Snyder S, et al: MR arthrography of the shoulder: Comparison with conventional MR imaging. AJR Am J Roentgenol 155:829–832, 1990. 60. Foley-Nolan D, Stack JP, Ryan M, et al: Magnetic resonance imaging in the assessment of rheumatoid arthritis–A comparison with plain film radiographs. Br J Rheumatol 30(2):101–106, 1991. 61. Friedman RJ, Bonutti PM, Genez B: Cine magnetic resonance imaging of the subcoracoid region. Orthopedics 21(5):545–548, 1998. 62. Fritz RC, Helms CA, Steinbach LS, et al: Suprascapular nerve entrapment: Evaluation with MR imaging. Radiology 182(2):437–444, 1992. 63. Fu FH, Harner CD, Klein AH: Shoulder impingement syndrome. A critical review. Clin Orthop Relat Res 269:162–173, 1991. 64. Fuchs B, Jost B, Gerber C: Posterior-inferior capsular shift for the treatment of recurrent, voluntary posterior subluxation of the shoulder. J Bone Joint Surg Am 82(1):16–25, 2000. 65. Gagey N, Ravaud E, Lassau JP: Anatomy of the acromial arch: correlation of anatomy and magnetic resonance imaging. Surg Radiol Anat 15(1):63–70, 1993. 66. Garneau RA, Renfrew DL, Moore TE, et al: Glenoid labrum: Evaluation with MR imaging. Radiology 179(2):519–522, 1991. 67. Gazielly DF, Gleyze P, Montagnon C, et al: [Functional and anatomical results after surgical treatment of ruptures of the rotator cuff. 1: Preoperative functional and anatomical evaluation of ruptures of the rotator cuff]. Rev Chir Orthop Reparatrice Appar Mot 81(1):8–16, 1995. 68. Gerber C, Sebesta A: Impingement of the deep surface of the subscapularis tendon and the reflection pulley on the anterosuperior glenoid rim: A preliminary report. J Shoulder Elbow Surg 9(6):483–490, 2000. 69. Gerber C, Terrier F, Ganz R: The role of the coracoid process in the chronic impingement syndrome. J Bone Joint Surg Br 67(5):703–708, 1985. 70. Gold RH, Seeger LL, Yao L: Imaging shoulder impingement. Skeletal Radiol 22(8):555–561, 1993. 71. Grauer JD, Paulos LE, Smutz WP: Biceps tendon and superior labral injuries. Arthroscopy 8(4):488–497, 1992. 72. Griffith JF, Antonio GE, Tong CW, et al: Anterior shoulder dislocation: Quantification of glenoid bone loss with CT. AJR Am J Roentgenol 180(5):1423–1430, 2003. 73. Griffith JF, Antonio GE, Yung PS, et al: Prevalence, pattern, and spectrum of glenoid bone loss in anterior shoulder dislocation: CT analysis of 218 patients. AJR Am J Roentgenol 190(5):1247–1254, 2008. 74. Griffith JF, Yung PS, Antonio GE, et al: CT compared with arthroscopy in quantifying glenoid bone loss. AJR Am J Roentgenol 189(6):1490– 1493, 2007. 75. Gusmer PB, Potter HG, Schatz JA, et al: Labral injuries: Accuracy of detection with unenhanced MR imaging of the shoulder. Radiology 200(2):519–524, 1996. 76. Gyftopoulos S, Beltran LS, Yemin A, et al: Use of 3D MR reconstructions in the evaluation of glenoid bone loss: A clinical study. Skeletal Radiol 43(2):213–218, 2014. 77. Gyftopoulos S, Bencardino J, Nevsky G, et al: Rotator cable: MRI study of its appearance in the intact rotator cuff with anatomic and histologic correlation. AJR Am J Roentgenol 200(5):1101–1105, 2013.

78. Halverson PB, Carrera GF, McCarty DJ: Milwaukee shoulder syndrome. Fifteen additional cases and a description of contributing factors. Arch Intern Med 150(3):677–682, 1990. 79. Halverson PB, McCarty DJ, Cheung HS, et al: Milwaukee Shoulder syndrome: Eleven additional cases with involvement of the knee in seven (basic calcium phosphate crystal deposition disease). Semin Arthritis Rheum 14(1):36–44, 1984. 80. Handelberg F, Willems S, Shahabpour M, et al: SLAP lesions: A retrospective multicenter study. Arthroscopy 14(8):856–862, 1998. 81. Harris RD, Harris JHJ: The prevalence and significance of missed scapular fractures in blunt chest trauma. AJR Am J Roentgenol 151(4):747–750, 1988. 82. Haygood TM, Langlotz CP, Kneeland JB, et al: Categorization of acromial shape: Interobserver variability with MR imaging and conventional radiography. AJR Am J Roentgenol 162(6):1377–1382, 1994. 83. Helms CA, Martinez S, Speer KP: Acute brachial neuritis (ParsonageTurner syndrome): MR imaging appearance–report of three cases. Radiology 207(1):255–259, 1998. 84. Heppenstall RB: Fractures of the proximal humerus. Orthop Clin North Am 6(2):467–475, 1975. 84a. Hodler J, Kursunoglu-Brahme S, Snyder SJ, et al: Rotator cuff disease: Assessment with MR arthrography versus standard MR Imaging in 36 patients with arthroscopic confirmation. Radiology 182:431–436, 1992. 85. Hollister MS, Mack LA, Patten RM, et al: Association of sonographically detected subacromial/subdeltoid bursal effusion and intraarticular fluid with rotator cuff tear. AJR Am J Roentgenol 165(3):605–608, 1995. 86. Holt RG, Helms CA, Steinbach L, et al: Magnetic resonance imaging of the shoulder: Rationale and current applications. Skeletal Radiol 19(1):5–14, 1990. 87. Huang LF, Rubin DA, Britton CA: Greater tuberosity changes as revealed by radiography: Lack of clinical usefulness in patients with rotator cuff disease. AJR Am J Roentgenol 172(5):1381–1388, 1999. 88. Hunter JC, Blatz DJ, Escobedo EM: SLAP lesions of the glenoid labrum: CT arthrographic and arthroscopic correlation. Radiology 184(2):513– 518, 1992. 89. Iannotti JP, Zlatkin MB, Esterhai JL, et al: Magnetic resonance imaging of the shoulder. Sensitivity, specificity, and predictive value. J Bone Joint Surg Am 73(1):17–29, 1991. 90. Jacobson SR, Speer KP, Moor JT, et al: Reliability of radiographic assessment of acromial morphology. J Shoulder Elbow Surg 4(6): 449–453, 1995. 91. Jelinek JS, Kransdorf MJ, Utz JA, et al: Imaging of pigmented villonodular synovitis with emphasis on MR imaging. AJR Am J Roentgenol 152(2):337–342, 1989. 92. Jobe CM: Superior glenoid impingement. Orthop Clin North Am 28(2):137–143, 1997. 93. Jobe FW, Bradley JP: Rotator cuff injuries in baseball. Prevention and rehabilitation. Sports Med 6(6):378–387, 1988. 94. Jobe FW, Giangarra CE, Kvitne RS, et al: Anterior capsulolabral reconstruction of the shoulder in athletes in overhand sports. Am J Sports Med 19(5):428–434, 1991. 95. Kaneko K, De Mouy EH, Brunet ME: Massive rotator cuff tears. Screening by routine radiographs. Clin Imaging 19(1):8–11, 1995. 96. Kaplan PA, Bryans KC, Davick JP, et al: MR imaging of the normal shoulder: Variants and pitfalls. Radiology 184(2):519–524, 1992. 97. Keats TE, Pope TLJ: The acromioclavicular joint: Normal variation and the diagnosis of dislocation. Skeletal Radiol 17(3):159–162, 1988. 98. Kieft GJ, Bloem JL, Rozing PM, et al: MR imaging of recurrent anterior dislocation of the shoulder: Comparison with CT arthrography. AJR Am J Roentgenol 150(5):1083–1087, 1988. 99. Kieft GJ, Bloem JL, Rozing PM, et al: Rotator cuff impingement syndrome: MR imaging. Radiology 166(1 Pt 1):211–214, 1988. 100. Kilcoyne RF, Shuman WP, Matsen FA, et al: The Neer classification of displaced proximal humeral fractures: Spectrum of findings on plain radiographs and CT scans. AJR Am J Roentgenol 154(5):1029–1033, 1990.

CHAPTER 61  Shoulder 101. Kjellin I, Ho CP, Cervilla V, et al: Alterations in the supraspinatus tendon at MR imaging: Correlation with histopathologic findings in cadavers. Radiology 181(3):837–841, 1991. 102. Klompmaker J, Veth RP, Robinson PH, et al: Pigmented villonodular synovitis. Arch Orthop Trauma Surg 109(4):205–210, 1990. 103. Kneeland JB, Carrera GF, Middleton WD, et al: Rotator cuff tears: Preliminary application of high-resolution MR imaging with counter rotating current loop-gap resonators. Radiology 160(3):695–699, 1986. 104. Kneeland JB, Middleton WD, Carrera GF, et al: MR imaging of the shoulder: Diagnosis of rotator cuff tears. AJR Am J Roentgenol 149(2):333–337, 1987. 105. Kopka L, Funke M, Fischer U, et al: MR arthrography of the shoulder with gadopentetate dimeglumine: Influence of concentration, iodinated contrast material, and time on signal intensity. AJR Am J Roentgenol 163(3):621–623, 1994. 106. Kottal RA, Vogler JB, Matamoros A, et al: Pigmented villonodular synovitis: A report of MR imaging in two cases. Radiology 163(2): 551–553, 1987. 107. Kvitne RS, Jobe FW, Jobe CM: Shoulder instability in the overhand or throwing athlete. Clin Sports Med 14(4):917–935, 1995. 108. Kwak SM, Brown RR, Resnick D, et al: Anatomy, anatomic variations, and pathology of the 11- to 3-o’clock position of the glenoid labrum: Findings on MR arthrography and anatomic sections. AJR Am J Roentgenol 171(1):235–238, 1998. 109. Kwak SM, Brown RR, Trudell D, et al: Glenohumeral joint: Comparison of shoulder positions at MR arthrography. Radiology 208(2):375–380, 1998. 110. Leclair A, Gangi A, Lacaze F, et al: Rapid chondrolysis after an intra-articular leak of bone cement in treatment of a benign acetabular subchondral cyst: An unusual complication of percutaneous injection of acrylic cement. Skeletal Radiol 29(5):275–278, 2000. 111. Lee JC, Guy S, Connell D, et al: MRI of the rotator interval of the shoulder. Clin Radiol 62(5):416–423, 2007. 112. Legan JM, Burkhard TK, Goff WB, et al: Tears of the glenoid labrum: MR imaging of 88 arthroscopically confirmed cases. Radiology 179(1):241–246, 1991. 113. Levine WN, Clark AMJ, D’Alessandro DF, et al: Chondrolysis following arthroscopic thermal capsulorrhaphy to treat shoulder instability. A report of two cases. J Bone Joint Surg Am 87(3):616–621, 2005. 114. Linker CS, Helms CA, Fritz RC: Quadrilateral space syndrome: Findings at MR imaging. Radiology 188(3):675–676, 1993. 115. Liou JT, Wilson AJ, Totty WG, et al: The normal shoulder: Common variations that simulate pathologic conditions at MR imaging. Radiology 186(2):435–441, 1993. 116. Lo IK, Burkhart SS: Arthroscopic repair of massive, contracted, immobile rotator cuff tears using single and double interval slides: Technique and preliminary results. Arthroscopy 20(1):22–33, 2004. 117. Lohr JF, Uhthoff HK: The microvascular pattern of the supraspinatus tendon. Clin Orthop Relat Res 254:35–38, 1990. 118. Longacre AM, Sheer AL: Tumoral calcinosis, case presentation and review of 55 cases in the literature. JFMA 61(3):221–225, 1974. 119. Maffet MW, Gartsman GM, Moseley B: Superior labrum-biceps tendon complex lesions of the shoulder. Am J Sports Med 23(1):93–98, 1995. 120. Magee TH, Gaenslen ES, Seitz R, et al: MR imaging of the shoulder after surgery. AJR Am J Roentgenol 168(4):925–928, 1997. 121. Magee T, Williams D: 3.0-T MRI of the supraspinatus tendon. AJR Am J Roentgenol 187(4):881–886, 2006. 122. Maldjian C, Borrero C, Adam R, et al: Nature abhors a vacuum: Bilateral prominent rotator cable in bilateral congenital absence of the long head of the biceps tendon. Skeletal Radiol 43(1):75–78, 2014. 123. Mandelbaum BR, Grant TT, Hartzman S, et al: The use of MRI to assist in diagnosis of pigmented villonodular synovitis of the knee joint. Clin Orthop Relat Res 231:135–139, 1988. 124. Matsen FA, Lippit SB, Bertelsen A, et al: Glenohumeral instability. In The shoulder, Philadelphia, 2009, WB Saunders, pp 617–700. 125. Matsen FA, Fehringer EV, Lippitt SB, et al: Rotator cuff. In The shoulder, Philadelphia, 2009, WB Saunders, pp 771–819.

2137

126. Matsen FA, Fehringer EV, Lippitt SB, et al: Rotator cuff. Relevant anatomy and mechanics. In The shoulder, Philadelphia, 2009, WB Saunders, pp 775–798. 127. Mayerhoefer ME, Breitenseher MJ, Roposch A, et al: Comparison of MRI and conventional radiography for assessment of acromial shape. AJR Am J Roentgenol 184(2):671–675, 2005. 128. McCarty DJ, Halverson PB, Carrera GF, et al: “Milwaukee shoulder”– Association of microspheroids containing hydroxyapatite crystals, active collagenase, and neutral protease with rotator cuff defects. I. Clinical aspects. Arthritis Rheum 24(3):464–473, 1981. 129. McCauley TR, Pope CF, Jokl P: Normal and abnormal glenoid labrum: Assessment with multiplanar gradient-echo MR imaging. Radiology 183(1):35–37, 1992. 130. McGahan JP, Rab GT, Dublin A: Fractures of the scapula. J Trauma 20:880–883, 1980. 131. McGinnis M, Denton JR: Fractures of the scapula: A retrospective study of 40 fractured scapulae. J Trauma 29(11):1488–1493, 1989. 132. McKendry RJ, Uhthoff HK, Sarkar K, et al: Calcifying tendinitis of the shoulder: Prognostic value of clinical, histologic, and radiologic features in 57 surgically treated cases. J Rheumatol 9(1):75–80, 1982. 133. McNair MM, Boyle JA, Buchanan WW, et al: A clinical and radiological study of rheumatoid arthritis with a note on the findings in osteoarthrosis. I. The shoulder joint. Clin Radiol 20(3):269–277, 1969. 134. McNiesh LM, Callaghan JJ: CT arthrography of the shoulder:   Variations of the glenoid labrum. AJR Am J Roentgenol 149(5):963–966, 1987. 135. Mellado JM, Calmet J, Olona M, et al: MR assessment of the repaired rotator cuff: Prevalence, size, location, and clinical relevance of tendon rerupture. Eur Radiol 16(10):2186–2196, 2006. 136. Mengiardi B, Pfirrmann CW, Gerber C, et al: Frozen shoulder: MR arthrographic findings. Radiology 233(2):486–492, 2004. 137. Merl T, Scholz M, Gerhardt P, et al: Results of a prospective multicenter study for evaluation of the diagnostic quality of an open whole-body low-field MRI unit. A comparison with high-field MRI measured by the applicable gold standard. Eur J Radiol 30:43–53, 1999. 138. Mesiha MM, Derwin KA, Sibole SC, et al: The biomechanical relevance of anterior rotator cuff cable tears in a cadaveric shoulder model.   J Bone Joint Surg Am 95(20):1817–1824, 2013. 139. Metzker A, Eisenstein B, Oren J, et al: Tumoral calcinosis revisited– Common and uncommon features. Report of ten cases and review.   Eur J Pediatr 147(2):128–132, 1988. 140. Millett PJ, Clavert P, Warner JJ: Open operative treatment for anterior shoulder instability: When and why? J Bone Joint Surg Am 87(2):419– 432, 2005. 141. Mintzer CM, Waters PM, Brown DJ: Glenoid version in children.   J Pediatr Orthop 16(5):563–566, 1996. 142. Mirowitz SA: Normal rotator cuff: MR imaging with conventional and fat-suppression techniques. Radiology 180(3):735–740, 1991. 143. Mitchell MJ, Causey G, Berthoty DP, et al: Peribursal fat plane of the shoulder: Anatomic study and clinical experience. Radiology 168(3):699–704, 1988. 144. Mochizuki T, Isoda H, Masui T, et al: Occlusion of the posterior humeral circumflex artery: Detection with MR angiography in healthy volunteers and in a patient with quadrilateral space syndrome. AJR Am J Roentgenol 163(3):625–627, 1994. 145. Mohana-Borges AV, Chung CB, Resnick D: MR imaging and MR arthrography of the postoperative shoulder: Spectrum of normal and abnormal findings. Radiographics 24(1):69–85, 2004. 146. Monu JU, Pope TLJ, Chabon SJ, et al: MR diagnosis of superior labral anterior posterior (SLAP) injuries of the glenoid labrum: Value of routine imaging without intraarticular injection of contrast material. AJR Am J Roentgenol 163(6):1425–1429, 1994. 147. Moore KL, Dalley AF, Agur AMR: Clinically Oriented Anatomy, 2013, Lippincott Williams & Wilkins. 148. Morag Y, Jacobson JA, Miller B, et al: MR imaging of rotator cuff injury: What the clinician needs to know. Radiographics 26(4):1045– 1065, 2006.

2138

PART II  CT and MR Imaging of the Whole Body

149. Morag Y, Jacobson JA, Shields G, et al: MR arthrography of rotator interval, long head of the biceps brachii, and biceps pulley of the shoulder. Radiology 235(1):21–30, 2005. 150. Moseley HF, Övergaard B: The anterior capsular mechanism in recurrent anterior dislocation of the shoulder. J Bone Joint Surg Br 44(4):913–927, 1962. 151. Mudge MK, Wood VE, Frykman G: Rotator cuff tears associated with os acromiale. J Bone Joint Surg 66(3):427–429, 1984. 152. Needell SD, Zlatkin MB, Sher JS, et al: MR imaging of the rotator cuff: Peritendinous and bone abnormalities in an asymptomatic population. AJR Am J Roentgenol 166(4):863–867, 1996. 153. Neer CS, 2nd.: Displaced proximal humeral fractures Part I. Classification and evaluation. J Bone Joint Surg 52(6):1077–1089, 1970. 154. Neer CS, 2nd.: Anterior acromioplasty for the chronic impingement syndrome in the shoulder. A preliminary report. J Bone Joint Surg 54(1):41–50, 1972. 155. Neer CS, Foster CR: Inferior capsular shift for involuntary inferior and multidirectional instability of the shoulder. A preliminary report.   J Bone Joint Surg 62(6):897–908, 1980. 156. Neer CS: Impingement lesions. Clin Orthop Relat Res 173:70–77, 1983. 157. Neer CS, Craig EV, Fukuda H: Cuff-tear arthropathy. J Bone Joint Surg 65(9):1232–1244, 1983. 158. Neer CS, 2nd, Welsh RP: The shoulder in sports. Orthop Clin North Am 8(3):583–591, 1977. 159. Neumann CH, Holt RG, Steinbach LS, et al: MR imaging of the shoulder: Appearance of the supraspinatus tendon in asymptomatic volunteers. AJR Am J Roentgenol 158(6):1281–1287, 1992. 160. Neumann CH, Petersen SA, Jahnke AH: MR imaging of the labralcapsular complex: Normal variations. AJR Am J Roentgenol 157(5):1015–1021, 1991. 160a. Neviaser TJ: The anterior labroligamentous periosteal sleeve avulsion lesion: A cause of anterior instability of the shoulder. Arthroscopy 9:17–21, 1993. 161. Neviaser TJ, Neviaser RJ, Neviaser JS, et al: The four-in-one arthroplasty for the painful arc syndrome. Clin Orthop Relat Res 163:107–112, 1982. 162. Nguyen VD, Nguyen KD: “Idiopathic destructive arthritis” of the shoulder: A still fascinating enigma. Comput Med Imaging Graph 14(4):249–255, 1990. 163. Nguyen VD, Nguyen KD, Kamath V: Unusual feature of soft-tissue calcification in chronic renal failure: Tumoral calcification. Comput Med Imaging Graph 15(6):397–402, 1991. 163a. Nicola T: Anterior dislocation of the shoulder: The role of the articular capsule. J Bone Joint Surg 24:614–616, 1942. 164. Nobuhara K, Ikeda H: Rotator interval lesion. Clin Orthop Relat Res 223:44–50, 1987. 165. Nutton RW, Stothard J: Acute calcific supraspinatus tendinitis in a three-year-old child. J Bone Joint Surg Br 69(1):148, 1987. 166. O’Brien SJ, Neves MC, Arnoczky SP, et al: The anatomy and histology of the inferior glenohumeral ligament complex of the shoulder. Am J Sports Med 18(5):449–456, 1990. 167. Ogata S, Uhthoff HK: Acromial enthesopathy and rotator cuff tear. A radiologic and histologic postmortem investigation of the coracoacromial arch. Clin Orthop Relat Res 254:39–48, 1990. 168. Ogino T, Minami A, Kato H, et al: Entrapment neuropathy of the suprascapular nerve by a ganglion. A report of three cases. J Bone Joint Surg Am 73(1):141–147, 1991. 169. Ouellette H, Thomas BJ, Kassarjian A, et al: Re-examining the association of os acromiale with supraspinatus and infraspinatus tears. Skeletal Radiol 36(9):835–839, 2007. 170. Owen RS, Iannotti JP, Kneeland JB, et al: Shoulder after surgery: MR imaging with surgical validation. Radiology 186(2):443–447, 1993. 171. Ozaki J, Fujimoto S, Nakagawa Y, et al: Tears of the rotator cuff of the shoulder associated with pathological changes in the acromion. A study in cadavera. J Bone Joint Surg Am 70(8):1224–1230, 1988. 171a. Palmer WE, Brown JH, Rosenthal DI: Rotator cuff: Evaluation with fat-suppressed MR arthrography. Radiology 188:683–687, 1993.

172. Palmer WE, Brown JH, Rosenthal DI: Labral-ligamentous complex of the shoulder: Evaluation with MR arthrography. Radiology 190(3): 645–651, 1994. 173. Palmer WE, Caslowitz PL: Anterior shoulder instability: Diagnostic criteria determined from prospective analysis of 121 MR arthrograms. Radiology 197(3):819–825, 1995. 174. Pande P, Hawkins R, Peat M: Electromyography in voluntary posterior instability of the shoulder. Am J Sports Med 17(5):644–648, 1989. 175. Pappas AM, Goss TP, Kleinman PK: Symptomatic shoulder instability due to lesions of the glenoid labrum. Am J Sports Med 11(5):279–288, 1983. 176. Park JG, Lee JK, Phelps CT: Os acromiale associated with rotator cuff impingement: MR imaging of the shoulder. Radiology 193(1):255–257, 1994. 177. Patten RM: Tears of the anterior portion of the rotator cuff (the subscapularis tendon): MR imaging findings. AJR Am J Roentgenol 162(2):351–354, 1994. 178. Peh WCG, Farmer THR, Totty WG: Acromial arch shape: Assessment with MR imaging. Radiology 195(2):501–505, 1995. 179. Penny JN, Welsh RP: Shoulder impingement syndromes in athletes and their surgical management*. Am J Sports Med 9(1):11–15, 1981. 180. Petersilge CA, Lewin JS, Duerk JL, et al: MR arthrography of the shoulder: Rethinking traditional imaging procedures to meet the technical requirements of MR imaging guidance. AJR Am J Roentgenol 169(5):1453–1457, 1997. 181. Petersson CJ: Spontaneous medial dislocation of the tendon of the long biceps brachii: An anatomic study of prevalence and pathomechanics. Clin Orthop Relat Res 211:224–227, 1986. 182. Petty DH, Jazrawi LM, Estrada LS, et al: Glenohumeral chondrolysis after shoulder arthroscopy: Case reports and review of the literature. Am J Sports Med 32(2):509–515, 2004. 183. Pokabla C, Hobgood ER, Field LD: Identification and management of “floating” posterior inferior glenohumeral ligament lesions. J Shoulder Elbow Surg 19(2):314–317, 2010. 184. Poleksic L, Zdravkovic D, Jablanovic D, et al: Magnetic resonance imaging of bone destruction in rheumatoid arthritis: Comparison with radiography. Skeletal Radiol 22(8):577–580, 1993. 185. Prodromos CC, Ferry JA, Schiller AL, et al: Histological studies of the glenoid labrum from fetal life to old age. J Bone Joint Surg 72(9): 1344–1348, 1990. 186. Provencher MT, Bhatia S, Ghodadra NS, et al: Recurrent shoulder instability: Current concepts for evaluation and management of glenoid bone loss. J Bone Joint Surg Am 92(Suppl 2):133–151, 2010. 187. Purchase RJ, Wolf EM, Hobgood ER, et al: Hill-Sachs “remplissage”: An arthroscopic solution for the engaging Hill-Sachs lesion. Arthroscopy 24(6):723–726, 2008. 187a. Quinn SF, Sheley RC, Demlow TA, et al: Rotator cuff tendon tears: Evaluation with fat-suppressed MR imaging with arthroscopic correlation in 100 patients. Radiology 195:497–501, 1995. 188. Rafii M, Firooznia H, Golimbu C, et al: CT arthrography of   capsular structures of the shoulder. Am J Roentgenol 146(2):361–367, 1986. 189. Rafii M, Firooznia H, Sherman O, et al: Rotator cuff lesions: Signal patterns at MR imaging. Radiology 177(3):817–823, 1990. 190. Rathbun JB, Macnab I: The microvascular pattern of the rotator cuff.   J Bone Joint Surg Br 52(3):540–553, 1970. 191. Reilly P, Macleod I, Macfarlane R, et al: Dead men and radiologists don’t lie: A review of cadaveric and radiological studies of rotator cuff tear prevalence. Ann R Coll Surg Engl 88(2):116–121, 2006. 191a. Reinus WR, Shady KL, Mirowitz SA, et al: MR diagnosis of rotator cuff tears of the shoulder: Value of using T2-weighted fat-saturated images. AJR Am J Roentgenol 164:1451–1455, 1995. 192. Richards RD, Sartoris DJ, Pathria MN, et al: Hill-Sachs lesion and normal humeral groove: MR imaging features allowing their differentiation. Radiology 190(3):665–668, 1994. 193. Richardson ML, Patten RM: Age-related changes in marrow distribution in the shoulder: MR imaging findings. Radiology 192(1):209–215, 1994.

CHAPTER 61  Shoulder 194. Roberts CC, Ekelund AL, Renfree KJ, et al: Radiologic assessment of reverse shoulder arthroplasty. Radiographics 27(1):223–235, 2007. 194a. Robertson PL, Schweitzer ME, Mitchell DG, et al: Rotator cuff disorders: Interobserver and intraobserver variation in diagnosis with MR imaging. Radiology 194:831–835, 1995. 195. Roger B, Skaf A, Hooper AW, et al: Imaging findings in the dominant shoulder of throwing athletes: Comparison of radiography, arthrography, CT arthrography, and MR arthrography with arthroscopic correlation. AJR Am J Roentgenol 172(5):1371–1380, 1999. 196. Rominger MB, Bernreuter WK, Kenney PJ, et al: MR imaging of the hands in early rheumatoid arthritis: Preliminary results. Radiographics 13(1):37–46, 1993. 197. Rowe CR, Pierce DS, Clark JG: Voluntary dislocation of the shoulder: A preliminary report on a clinical, electromyographic, and psychiatric study of twenty-six patients. J Bone Joint Surg 55(3):445–460, 1973. 198. Sahin-Akyar G, Miller TT, Staron RB, et al: Gradient-echo versus fat-suppressed fast spin-echo MR imaging of rotator cuff tears. AJR Am J Roentgenol 171(1):223–227, 1998. 199. Salter EGJ, Nasca RJ, Shelley BS: Anatomical observations on the acromioclavicular joint and supporting ligaments. Am J Sports Med 15(3):199–206, 1987. 199a. Sanders TG, Tirman PF, Linares R, et al: The glenolabral articular disruption lesion: MR arthrography with arthroscopic correlation. AJR Am J Roentgenol 172:171–175, 1999. 200. Sanders TG, Zlatkin MB, Paruchuri NB, et al: Chondrolysis of the glenohumeral joint after arthroscopy: Findings on radiography and low-field-strength MRI. AJR Am J Roentgenol 188(4):1094–1098, 2007. 201. Savoie FH, Field LD, Atchinson S: Anterior superior instability with rotator cuff tearing: SLAC lesion. Orthop Clin North Am 32(3):457–461, ix, 2001. 202. Schraner AB, Major NM: MR imaging of the subcoracoid bursa. AJR Am J Roentgenol 172(6):1567–1571, 1999. 203. Schweitzer ME, Magbalon MJ, Fenlin JM, et al: Effusion criteria and clinical importance of glenohumeral joint fluid: MR imaging evaluation. Radiology 194(3):821–824, 1995. 204. Seeger LL, Gold RH, Bassett LW: Shoulder instability: Evaluation with MR imaging. Radiology 168(3):695–697, 1988. 205. Seeger LL, Gold RH, Bassett LW, et al: Shoulder impingement syndrome: MR findings in 53 shoulders. AJR Am J Roentgenol 150(2):343–347, 1988. 206. Seeger LL, Lubowitz J, Thomas BJ: Case report 815: Tear of the rotator interval. Skeletal Radiol 22(8):615–617, 1993. 207. Sheah K, Bredella MA, Warner JJ, et al: Transverse thickening along the articular surface of the rotator cuff consistent with the rotator cable: Identification with MR arthrography and relevance in rotator cuff evaluation. AJR Am J Roentgenol 193(3):679–686, 2009. 208. Sher JS, Uribe JW, Posada A, et al: Abnormal findings on magnetic resonance images of asymptomatic shoulders. J Bone Joint Surg Am 77(1):10–15, 1995. 209. Shuman WP, Kilcoyne RF, Matsen FA, et al: Double-contrast computed tomography of the glenoid labrum. AJR Am J Roentgenol 141(3): 581–584, 1983. 210. Silverman PM: Helical (Spiral) Computed Tomography: A Practical Approach to Clinical Protocols, New York, 1998, Lippincott. 211. Singson RD, Feldman F, Bigliani L: CT arthrographic patterns in recurrent glenohumeral instability. AJR Am J Roentgenol 149(4): 749–753, 1987. 212. Singson RD, Feldman F, Bigliani LU, et al: Recurrent shoulder dislocation after surgical repair: Double-contrast CT arthrography. Work in progress. Radiology 164(2):425–428, 1987. 212a. Singson RD, Hoang T, Dan S, et al: MR evaluation of rotator cuff pathology using T2-weighted fast spin-echo technique with and without fat suppression. AJR Am J Roentgenol 166:1061–1065, 1996. 213. Smith DK, Chopp TM, Aufdemorte TB, et al: Sublabral recess of the superior glenoid labrum: Study of cadavers with conventional nonenhanced MR imaging, MR arthrography, anatomic dissection, and limited histologic examination. Radiology 201(1):251–256, 1996.

2139

214. Snyder SJ: Diagnostic arthroscopy of the shoulder: Normal anatomy and variations. 23–40, 1994. 215. Snyder SJ, Karzel RP, Del Pizzo W, et al: SLAP lesions of the shoulder. Arthroscopy 6(4):274–279, 1990. 215a. Sonin AH, Peduto AJ, Fitzgerald SW, et al: MR imaging of the rotator cuff mechanism: Comparison of spin-echo and turbo spin-echo sequences. AJR Am J Roentgenol 167:333–338, 1996. 216. Sperling JW, Dahm DL: Double suture technique to delineate PASTA lesions. Arthroscopy 22(6):681.e1–681.e2, 2006. 217. Spielmann AL, Forster BB, Kokan P, et al: Shoulder after rotator cuff repair: MR imaging findings in asymptomatic individuals–Initial experience. Radiology 213(3):705–708, 1999. 218. Steinbach LS, Neumann CH, Stoller DW, et al: MRI of the knee in diffuse pigmented villonodular synovitis. Clin Imaging 13(4):305–316, 1989. 219. Stiles RG, Otte MT: Imaging of the shoulder. Radiology 188(3):603–613, 1993. 220. Stoller DW, Wolf EM, Li AE, et al: The shoulder. In Magnetic Resonance Imaging in Orthopaedics and Sports Medicine, Philadelphia, 1997, Lippincott-Raven., p 665. 221. Sugaya H, Moriishi J, Dohi M, et al: Glenoid rim morphology in recurrent anterior glenohumeral instability. J Bone Joint Surg Am 85-A(5):878–884, 2003. 222. Takwale VJ, Calvert P, Rattue H: Involuntary positional instability of the shoulder in adolescents and young adults. Is there any benefit from treatment? J Bone Joint Surg Br 82(5):719–723, 2000. 223. Tehranzadeh J, Ashikyan O, Dascalos J, et al: MRI of large intraosseous lesions in patients with inflammatory arthritis. AJR Am J Roentgenol 183(5):1453–1463, 2004. 224. Thompson DA, Flynn TC, Miller PW, et al: The significance of scapular fractures. J Trauma 25(10):974–977, 1985. 225. Tian CY, Shang Y, Zheng ZZ: Glenoid bone lesions: Comparison between 3D VIBE images in MR arthrography and nonarthrographic MSCT. J Magn Reson Imaging 36(1):231–236, 2012. 226. Timins ME, Erickson SJ, Estkowski LD, et al: Increased signal in the normal supraspinatus tendon on MR imaging: Diagnostic pitfall caused by the magic-angle effect. AJR Am J Roentgenol 165(1):109–114, 1995. 227. Tirman PF, Bost FW, Steinbach LS, et al: MR arthrographic depiction of tears of the rotator cuff: Benefit of abduction and external rotation of the arm. Radiology 192(3):851–856, 1994. 228. Tirman PF, Feller JF, Janzen DL, et al: Association of glenoid labral cysts with labral tears and glenohumeral instability: Radiologic findings and clinical significance. Radiology 190(3):653–658, 1994. 229. Tirman PF, Feller JF, Palmer WE, et al: The Buford complex–A variation of normal shoulder anatomy: MR arthrographic imaging features. AJR Am J Roentgenol 166(4):869–873, 1996. 229a. Tirman PFJ, Steinbach LS, Feller JF, et al: Humeral avulsion of the anterior shoulder stabilizing structures after anterior shoulder dislocation: Demonstration by MRI and MR anthrography. Skeletal Radiol 25:743–748, 1996. 230. Townley CO: The capsular mechanism in recurrent dislocation of the shoulder. J Bone Joint Surg 32(2):370–380, 1950. 231. Traughber PD, Goodwin TE, Shoulder MRI: Arthroscopic correlation with emphasis on partial tears. J Comput Assist Tomogr 16(1):129–133, 1992. 232. Tuite MJ, Cirillo RL, De Smet AA, et al: Superior labrum anteriorposterior (SLAP) tears: Evaluation of three MR signs on T2-weighted images. Radiology 215(3):841–845, 2000. 233. Tuite MJ, Toivonen DA, Orwin JF, et al: Acromial angle on radiographs of the shoulder: Correlation with the impingement syndrome and rotator cuff tears. AJR Am J Roentgenol 165(3):609–613, 1995. 234. Tung GA, Hou DD: MR arthrography of the posterior labrocapsular complex: Relationship with glenohumeral joint alignment and clinical posterior instability. Am J Roentgenol 180(2):369–375, 2003. 235. Uhthoff HK, Sarkar K, Maynard J: Calcifying tendinitis: A new concept of its pathogenesis. Clin Orthop Relat Res 118:164–168, 1976. 236. Vahlensieck M, An Haack K, Schmidt HM: Two portions of the supraspinatus muscle: A new finding about the muscles macroscopy by

2140

PART II  CT and MR Imaging of the Whole Body

dissection and magnetic resonance imaging. Surg Radiol Anat 16(1):101–104, 1994. 237. Vahlensieck M, Pollack M, Lang P, et al: Two segments of the supraspinous muscle: Cause of high signal intensity at MR imaging? Radiology 186(2):449–454, 1993. 238. Valls R, Melloni P: Sonographic guidance of needle position for MR arthrography of the shoulder. AJR Am J Roentgenol 169(3):845–847, 1997. 239. Van Huyssteen AL, Bracey DJ: Chlorhexidine and chondrolysis in the knee. J Bone Joint Surg Br 81(6):995–996, 1999. 240. Vanarthos WJ, Monu JU: Type 4 acromion: A new classification. Contemp Orthop 30(3):227–229, 1995. 241. Van Leersum M, Schweitzer ME: Magnetic resonance imaging of the biceps complex. Magn Reson Imaging Clin N Am 1(1):77–86, 1993. 242. Vebostad A: Calcific tendinitis in the shoulder region: A review of 43 operated shoulders. Acta Orthop 46(2):205–210, 1975. 243. Waldt S, Bruegel M, Mueller D, et al: Rotator cuff tears: Assessment with MR arthrography in 275 patients with arthroscopic correlation. Eur Radiol 17(2):491–498, 2007. 244. Wang DH, Koehler SM: Isolated infraspinatus atrophy in a collegiate volleyball player. Clin J Sport Med 6(4):255–258, 1996. 245. Warner JJ, Shah A: Shoulder arthroplasty for the treatment of rotator cuff insufficiency. Instr Course Lect 60:113–121, 2011. 246. Weishaupt D, Zanetti M, Nyffeler RW, et al: Posterior glenoid rim deficiency in recurrent (atraumatic) posterior shoulder instability. Skeletal Radiol 29(4):204–210, 2000. 247. Weishaupt D, Zanetti M, Tanner A, et al: Lesions of the reflection pulley of the long biceps tendon. MR arthrographic findings. Invest Radiol 34(7):463–469, 1999. 248. Weiss JJ, Good A, Schumacher HR: Four cases of “Milwaukee shoulder,” with a description of clinical presentation and long-term treatment.   J Am Geriatr Soc 33(3):202–205, 1985. 249. Whitten CG, Moore TE, Yuh WT, et al: The use of intravenous gadopentetate dimeglumine in magnetic resonance imaging of synovial lesions. Skeletal Radiol 21(4):215–218, 1992. 249a. Willemsen UF, Wiedemann E, Brunner U, et al: Prospective evaluation of MR arthrography performed with high-volume intraarticular saline

250.

251.

252.

253.

254.

255. 256.

257.

258. 259. 260.

261.

enhancement in patients with recurrent anterior dislocations of the shoulder. AJR Am J Roentgenol 170:79–84, 1998. Wilson AJ, Totty WG, Murphy WA, et al: Shoulder joint: Arthrographic CT and long-term follow-up, with surgical correlation. Radiology 173(2):329–333, 1989. Wohlwend JR, van Holsbeeck M, Craig J, et al: The association between irregular greater tuberosities and rotator cuff tears: A sonographic study. AJR Am J Roentgenol 171(1):229–233, 1998. Wolf EM, Arianjam A: Hill-Sachs remplissage, an arthroscopic solution for the engaging Hill-Sachs lesion: 2- to 10-year follow-up and incidence of recurrence. J Shoulder Elbow Surg 2013. Wolf EM, Cheng JC, Dickson K: Humeral avulsion of glenohumeral ligaments as a cause of anterior shoulder instability. Arthroscopy 11(5):600–607, 1995. Workman TL, Burkhard TK, Resnick D, et al: Hill-Sachs lesion: Comparison of detection with MR imaging, radiography, and arthroscopy. Radiology 185(3):847–852, 1992. Yu JS, Fischer RA: Denervation atrophy caused by suprascapular nerve injury: MR findings. J Comput Assist Tomogr 21(2):302–303, 1997. Zanetti M, Weishaupt D, Gerber C, et al: Tendinopathy and rupture   of the tendon of the long head of the biceps brachii muscle: Evaluation with MR arthrography. AJR Am J Roentgenol 170(6):1557–1561, 1998. Zanetti M, Weishaupt D, Jost B, et al: MR imaging for traumatic tears of the rotator cuff: High prevalence of greater tuberosity fractures and subscapularis tendon tears. AJR Am J Roentgenol 172(2):463–467, 1999. Zlatkin MB: MRI of the postoperative shoulder. Skeletal Radiol 31(2):63–80, 2002. Zlatkin MB: MRI of the Shoulder, 2003, Lippincott Williams & Wilkins. Zlatkin MB, Iannotti JP, Roberts MC, et al: Rotator cuff tears: Diagnostic performance of MR imaging. Radiology 172(1):223–229, 1989. Zlatkin MB, Reicher MA, Kellerhouse LE, et al: The painful shoulder: MR imaging of the glenohumeral joint. J Comput Assist Tomogr 12(6):995–1001, 1988.

62  Hip and Pelvis Jason W. Stephenson and Donna G. Blankenbaker

INTRODUCTION The uses and applications of computed tomography (CT) and magnetic resonance imaging (MRI) to image the hip and pelvis continue to expand. These imaging techniques are effectively used to diagnose and characterize pathologic conditions including congenital and developmental abnormalities in their downstream sequelae, acquired intraarticular soft tissue and bony defects, infection, arthritis, vascular compromise, as well as tumors and tumorlike conditions. As with most musculoskeletal complaints, radiographs remain the primary first-line technique for imaging patients with problems of the hip and bony pelvis. Noncontrast CT is excellent for characterizing  the bony anatomy of the pelvis, evaluating heavily mineralized lesions of the bones and soft tissues, characterizing fracture patterns, and evaluating fracture healing. Contrast-enhanced CT may also be used in the emergent setting as a screen for abscess. Gas and radiopaque foreign body within the soft tissues may also be evident on CT. The excellent contrast resolution and cross-sectional capability of MRI make it an excellent tool for evaluating the muscles, tendons, ligaments, and nerves of the pelvis, hips, and proximal thighs. In addition, these characteristics make MRI an excellent choice for imaging the marrow and intraarticular compartments. MRI is the only technique that allows consistent visualization of articular cartilage.

NORMAL ANATOMY The bony pelvis is an osseous scaffolding composed of the ilium, ischium, and pubis, which collectively form the innominate bone. Right and left innominate bones and the sacrum create an osseous ring that connects the torso to the lower extremities. The bones and durable ligaments of the pelvis are critical in suspending the trunk and transmitting the loading of the trunk and upper extremities to the lower extremities during bipedal ambulation. In addition, the pelvic ring contains the pelvic viscera, transmits the gastrointestinal and genitourinary tracts, and forms the site of attachment for powerful lower extremity and body wall muscles and tendons. The hip joint is a deep ball-and-socket articulation between the acetabulum and femoral head that is surrounded by a durable collection of capsular ligaments. This arrangement confers stability for weightbearing while also accommodating a wide range of motion and numerous complex movements. The acetabular socket is angled inferiorly and anteriorly with a horseshoe-shaped hyaline cartilage–lined articular surface known as the lunate cartilage. The resulting nonarticular central and inferior depression is known as the cotyloid fossa and is filled with fibrofatty tissue known as the pulvinar (Fig. 62-1). The acetabular labrum is a fibrocartilaginous ring that surrounds the anterior, superior, and posterior portions of the acetabular rim. The

labrum is attached to the peripheral articular cartilage and osseous acetabular rim. The labrum, which is triangular in its cross-sectional dimension, serves to deepen and further stabilize the socket of the hip joint as well as distribute force during weightbearing and movement (Fig. 62-2). These functions also preserve the integrity of hip articular cartilage. The transverse ligament connects the anteroinferior and posteroinferior aspects of the acetabular labrum. The head of the femur is covered with hyaline cartilage except for the centrally positioned fovea capitis, which is the distal attachment for the ligamentum teres (Fig. 62-3). In the normal condition, the femoral head is spherical with a steep offset at its head-neck junction. The hip joint capsule extends from the acetabular rim adjacent to the base of the labrum to the intertrochanteric region of the proximal femur, resulting in an intracapsular position of the femoral head, headneck junction, and femoral neck (Fig. 62-4).

ANATOMIC VARIANTS Several well-characterized anatomic variants exist in the hip. These are frequently encountered on imaging and can be easily mistaken for pathology. Familiarity with the typical appearance of these variants can avoid overcalling hip pathology.

Os Acetabuli Visible on all imaging techniques, os acetabuli (Fig. 62-5) represent incompletely fused primary or accessory ossification centers that lie along the lateral margin of the acetabular rim in 2% to 3% of individuals.116 These foci may also be sequelae of prior acetabular rim fractures. Os acetabuli are most often asymptomatic, with intact underlying articular cartilage46,164 and acetabular labra, though prominent os acetabula have been associated with acetabular overcoverage and pincertype femoroacetabular impingement (FAI).108

Synovial Herniation Pit Also known as Pitt’s pit, synovial herniation pits are well-defined lucent defects, typically in the superolateral and anterosuperior head-neck junction, surrounded by a thin uniform sclerotic rim89,129 (Fig. 62-6). These defects, which are also visible on most imaging techniques,46,89 are thought to result from herniation of fluid and synovium into the bone after repetitive contact between the femur and acetabulum. The role of FAI in their origin is controversial.63,89

Sublabral Sulcus This somewhat controversial anatomic variant is a smooth linear groove along the articular surface of the base of the labrum, communicating directly with the hip joint.43,46,145 This defect can exist at any anatomic position along the labral circumference, is most common

2141

2142

PART II  CT and MR Imaging of the Whole Body

anteroinferiorly,46,145 and is most easily visualized on direct MR arthrography43,145 (Fig. 62-7).

typically in continuity with the cotyloid fossa. This variant, which is commonly only visible at arthroscopy, lies medial to the expected position of a supra-acetabular fossa and can be mistaken for a focal area of cartilage loss on CT and MRI.46,116

Supra-Acetabular Fossa The supra-acetabular fossa is a focal defect in the lunate cartilage at the 12 o’clock position41 (Fig. 62-8). The predictable location and morphology of this entity distinguishes it from an acquired area of cartilage loss.41 This defect, which can also be easily mistaken for an osteochondral lesion, is evident on radiographs, CT, and MRI.

Plicae and Pectinofoveal Fold Plicae are synovial folds traversing the hip joint compartment and represent remnants of embryonic septations. These flat or tubular synovial connections are often visible on MR and CT arthrography adjacent to the inferomedial acetabular labrum, anterior femoral headneck junction, and near the ligamentum teres.46,116 The pectinofoveal fold, or middle retinaculum of Weitbrecht, has an appearance similar to a plica, has variable attachment sites, is rarely symptomatic, and is characteristically located extending from the region of the lesser trochanter to the fovea capitis near the distal attachment of the ligamentum teres17,116 (Fig. 62-9).

Stellate Crease The stellate crease is a focal defect in the lunate cartilage that lies predictably in the superomedial aspect of the acetabular articular surface,

Fovea capitis

Ilium

Greater trochanter

Head

Sciatic notch

Pubis

Neck

Cotyloid fossa

Ischium Obdurator foramen

Lesser trochanter

FIG 62-1  Osseous pelvic landmarks.

FIG 62-3  Proximal femur anatomy.

Sciatic buttress

Posterior column

Anterior column

Anterior column

Posterior column

A

B FIG 62-2  Anatomy of pelvic and hip weightbearing (A) and acetabular (B) walls.

CHAPTER 62  Hip and Pelvis

2143

joint capsule and iliopsoas bursa has also been attributed to traumatic defects.

Iliopsoas Bursa The iliopsoas bursa lies posterior and medial to the iliopsoas tendon as it passes along the supra-acetabular ridge of the ilium and anterior to the hip joint (Fig. 62-10). In 15% of patients this bursa directly communicates with the hip joint. This can result in filling of the iliopsoas bursa during hip injection.46,116 Communication between the hip

ad

He

ck

Ne

CONGENITAL AND DEVELOPMENTAL ABNORMALITIES Developmental Dysplasia of the Hip Developmental dysplasia of the hip (DDH) results from abnormal laxity in the infantile hip. As hip and pelvic development proceeds, this abnormal laxity results in an underdeveloped acetabular socket and incompletely covered femoral head (Fig. 62-11). The hypoplastic acetabulum predisposes patients with DDH to hip subluxation, dislocation, labral injury, cartilage defects, and periarticular cysts. Initial imaging of patients at risk for DDH typically consists of ultrasound (US) and subsequently radiographs. Imaging with MRI and CT is often performed to assess for associated sequelae or for surgical planning. MRI in particular is useful in evaluating hip anatomy once femoral head and acetabular ossification centers have developed; these ossification centers hinder adequate visualization of acetabular structures on US by approximately 3 to 4 months of age.59,141 Furthermore, the multiplanar capability of MRI makes it an excellent tool for assessing acetabular and proximal femoral morphology and expected complications related to DDH, including labral tears, labral detachment, focal cartilaginous lesions, and ligamentum teres injury.18,141 In addition, patients with DDH are often found to have characteristic anatomic variations including labral hypertrophy, capsular thickening and laxity, and hypertrophy of the fibrofatty pulvinar.141 CT is useful for preoperative planning. Multiplanar reformatted (MPR) images are helpful in measuring the degree of morphologic abnormality,81 and three-dimensional (3D) volume-rendered images are useful for templating and further operative planning for periacetabular and proximal femoral osteotomies. In addition, CT is helpful in assessing the degree of osseous healing and the adequacy of acetabular coverage after surgical intervention.35

Proximal Femoral Focal Deficiency FIG 62-4  Intracapsular portions of the proximal femur (blue).

A

Proximal femoral focal deficiency (PFFD) is a congenital developmental defect affecting the proximal aspects of the femora. This disorder

B

FIG 62-5  Os acetabuli. A, Anteroposterior (AP) radiograph in a 42-year-old man demonstrates a large os acetabuli along the anterosuperior aspect of the acetabular rim (arrow). B, Axial T1-weighted MRI through the right hip of a different patient demonstrates an ovoid osseous fragment along the anterosuperior aspect of the right acetabulum (arrow), representing a large os acetabuli.

2144

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 62-6  Synovial herniation pit. AP radiograph of a 24-year-old man (A) and axial CT image of a 32-year-old woman (B) with synovial herniation pits (arrows). This appears as a well-corticated round focus located along the anterior or anterosuperior femoral head-neck junction.

anatomy for surgical planning, and identify the location and extent of pseudoarthroses. Owing to poor soft tissue contrast resolution, however, differentiation between cartilaginous or fibrous tufts and completely absent femoral segments can be difficult using CT. Thus MRI is a valuable tool in the assessment of PFFD patients and classification of disease severity.15,106 MRI is particularly helpful in identifying small ossific components, irregular cartilaginous or fibrous tufts, and completely absent segments.15 The presence or absence of marrow edema pattern adjacent to pseudoarthroses can also be assessed.

Rotation and Version Deformities

FIG 62-7  Sublabral sulcus. Sagittal T1-weighted fat-suppressed MR arthrogram of a 22-year-old woman with an arthroscopically proven anterosuperior acetabular labral tear (arrowhead) and an incidentally noted sublabral sulcus (arrow) in the posteroinferior aspect of the labrum.

results in limb shortening and varying degrees of hypoplasia.62 Severe cases are characterized by complete absence of one or both proximal femurs, with intermediate forms including absent segments of proximal femoral bone replaced by cartilaginous or fibrous tufts62 (Fig. 62-12). Mild forms include delayed proximal femoral ossification and pseudoarthroses in the subtrochanteric region. This spectrum of deformity typically produces varying degrees of varus hip deformity and acetabular dysplasia. Congenital hip dislocations may also result. Imaging is helpful in the classification of PFFD. CT is useful to assess the extent of proximal femoral ossification, characterize osseous

Rotational and angular deformities of the femur can cause pain and dysfunction throughout the lower extremity.148,153 Rotational deformities include the more common excessive femoral anteversion (in-toeing) and the relatively rare excessive retroversion (out-toeing). These deformities are often congenital, though they may also be acquired following trauma or be related to mechanical use or structural abnormalities in other parts of the lower extremity. Angular malalignment includes varus and valgus deformity and can occur from any of the above mechanisms in addition to developmental, neuromuscular, and metabolic disorders. Cross-sectional imaging with CT and MRI is valuable in quantifying the degree of geometric abnormality in patients with rotational deformities.83,111,148 This is accomplished by anatomic comparison of images through the hips and distal femora. Often these evaluations include the hip, knee, and ankle joints, and exclude the shafts of the intervening bones to reduce scan times and, in the case of CT, radiation exposure. CT is useful in evaluating patients following surgical correction of geometric deformity. This includes evaluation for residual deformity and assessing the extent of osseous healing across osteotomy defects.

PATHOLOGIC CONDITIONS FAI, Acetabular Labral Abnormalities, and Other Intraarticular Pathology Labral Tears.  Tears of the acetabular labrum are a common cause of chronic hip pain in the adult population. Patients with labral injuries often present with symptoms of clicking and locking with sharp anterior pain or pain localizing to or emanating from the groin region. Symptoms are often reproduced in flexion, abduction, and external rotation.3,150 Symptoms may also consist of dull aching pain that

CHAPTER 62  Hip and Pelvis

A

2145

B

FIG 62-8  Supra-acetabular fossa in a 16-year-old girl with hip pain. Coronal (A) and sagittal (B) T1-weighted fat-suppressed MRI shows a supra-acetabular fossa (arrows) within the superior acetabular roof. This normal variant should be located at the 12:00 position within the acetabular roof, have smooth margins, and have no associated bone marrow edema or adjacent cartilage defects.

*

FIG 62-9  Pectinofoveal fold in a 45-year-old man with hip pain. Coronal T1-weighted fat-suppressed MR arthrogram shows a normal pectinofoveal fold (arrows) extending from the medial proximal femoral neck onto the hip capsule. This structure should be recognized as a normal finding and not considered an abnormal or pathologic structure. Note the cam-type morphology of the femoral head-neck junction.

FIG 62-10  Normal iliopsoas bursa in a 19-year-old woman with rightsided hip pain. Axial T2-weighted fat-suppressed image of the asymptomatic left hip shows a small (physiologic) amount of fluid signal within the left iliopsoas bursa. This bursa, which is interposed between the iliopsoas tendon (arrow) and supra-acetabular ridge (asterisk), is often not visible if there is no fluid within it.

2146

PART II  CT and MR Imaging of the Whole Body

persists despite treatment measures. Identification of labral injuries is important to relieve pain but also to prevent development of osteoarthritis associated with the loss of stability resulting from an incompetent labrum. Radiographs are insensitive in the detection of labral pathology, though these may show signs of secondary osteoarthritis or evidence of a condition associated with labral tear, including FAI or acetabular

FIG 62-11  Developmental dysplasia of the hip (DDH). Coronal CT image of a 42-year-old man with right DDH shows a shallow right acetabulum, superolateral subluxation and increased uncovering of the right femoral head, and mild secondary right hip osteoarthritis. There is normal morphology of the left acetabulum and femoral head.

A

dysplasia. In addition, CT has a limited role in the evaluation of patients with a suspected labral injury. MRI and particularly MR arthrography are both highly sensitive and specific for detecting labral injury, with MR arthrography outperforming conventional noncontrast MRI.73,165 The normal labrum is triangular in cross section, with homogeneously low signal within its substance, sharp uninterrupted outer margins, and an uninterrupted attachment to all four quadrants of the acetabular rim (Fig. 62-13). Nearly 20% of patients will have a shallow sulcus along the articular margin of the acetabular side of the labrum, which should not penetrate greater than 50% of the thickness of the base of the labrum.43,97 Labral tears are indicated by linear areas of high signal within the labrum,165 absent segments of the labrum, and blunted and irregularly shaped segments. MR arthrography is particularly helpful in that tears will imbibe contrast (Fig. 62-14). Tears most commonly occur in the anterosuperior quadrant of the labrum, though such injuries often propagate into the superior aspect and posterosuperior quadrant. Tears of the anteroinferior labrum are rare. The labrum may also detach from the acetabular rim, which can accompany focal disruption in the adjacent acetabular articular cartilage.14,19 Degenerative fraying of the labrum is a pattern that typically explains a blunted appearance and is often accompanied by other signs of osteoarthritis. The presence of a paralabral cyst on MRI, CT, or US has been correlated with the presence of a labral tear. Identification of a labral tear is therapeutically important because the preferred treatment for such a tear is débridement and/or primary repair.14 MR arthrography has been shown to be useful in the identification of normal and abnormal postsurgical acetabular labra.5,73 In

B

FIG 62-12  Proximal focal femoral deficiency (PFFD). A, AP radiograph of the pelvis of a 3-year-old girl with PFFD shows absence of the left proximal femur. The remainder of the distal left femur and the entire right femur are normal in appearance. Also note that both acetabula are shallow. B, Coronal CT of the pelvis of a 2-year-old girl with bilateral PFFD shows absence of the majority of the femoral heads and necks and small osseous tufts in the expected regions of the femoral heads (arrowheads). There is a small fibrous or cartilaginous connection (arrow) between the left proximal femur and femoral head osseous tuft.

A

B

FIG 62-13  Normal acetabular labra. Sagittal T1-weighted fat-suppressed MR arthrogram of a 16-year-old girl with right hip pain (A) and oblique axial T1-weighted fat-suppressed MR arthrogram of a 19-year-old man with right hip pain (B) demonstrate the triangular shape in cross section, uniform internal low signal, and sharp outer margins of normal acetabular labra (arrows). Hip articular cartilage is completely intact in both of these patients.

A

B

C

D

FIG 62-14  Four patients with acetabular labral tears. A, Sagittal T1-weighted fat-suppressed MR arthrogram of a 43-year-old woman with left hip pain shows increased signal within the anterosuperior acetabular labrum (arrow). The labrum also demonstrates a ragged outer contour and a focally absent component along its articular margin. B, Sagittal T1-weighted fat-suppressed MR arthrogram of a 45-year-old woman with right hip pain shows a large linear focus of contrast imbibition along the articular aspect of the anterosuperior labrum. C, Sagittal T1-weighted fat-suppressed MR arthrogram in a 21-year-old male hockey player shows complete detachment of the anterosuperior labrum (arrow). Contrast fills the labral defect, and there is a ragged outer contour of the remainder of the labrum. D, Oblique axial T1-weighted fat-suppressed MR arthrogram of the painful left hip of a 32-year-old woman shows a blunted appearance of the anterior labrum (arrow). This represents a torn labrum.

2148

PART II  CT and MR Imaging of the Whole Body

A B

C

D

FIG 62-15  A, Cam impingement in a 22-year-old man with left hip pain and cam impingement. An oblique axial reformatted CT image shows a large anterior prominence centered at the anterior femoral head-neck junction. B, A 3D volume-rendered CT image in the same patient demonstrates the large osseous prominence (arrows) at the anterosuperior left femoral head-neck junction. C, Sagittal T1-weighted fat-suppressed MR arthrogram in a 44-year-old man with right hip pain and cam impingement shows maceration of the anterosuperior labrum (arrowhead) and a delaminating full-thickness anterosuperior acetabular cartilage defect (arrow). D, Coronal T2-weighted fat-suppressed MR arthrogram in 19-year-old man with left hip pain and cam femoroacetabular impingement shows a large superolateral osseous prominence at the left femoral head-neck junction, a subjacent synovial herniation pit (arrow), and an incidentally noted partial tear of the distal attachment of the ligamentum teres (arrowhead).

patients who are unable to undergo MRI, direct CT arthrography can aid in the detection of labral injury. MR and CT arthrography are also helpful in identifying the presence of intraarticular bodies and disruption of the ligamentum teres.

Femoroacetabular Impingement.  FAI consists of a group of structural acetabular and femoral abnormalities that result in abnormal contact between the anterior and anterosuperior femoral head-neck junction and acetabular rim during hip flexion and internal rotation. This contact can directly cause pain but also results in compromise of the underlying articular structures including the acetabular labrum and articular cartilage of the hip joint. There are two main types of FAI that may occur alone or more commonly in combination. Cam-type FAI results from excessive bony prominence of the anterior and anterosuperior femoral head-neck junction. During hip flexion and internal rotation this osseous “bump” contacts the anterosuperior acetabular rim, resulting in direct impac-

tion and shearing of the anterosuperior labrum and subsequently articular cartilage150 (Fig. 62-15). Accordingly, cam impingement results in focal labral tears, labral detachment, and complex degenerative tears of the labrum in this region. This also results in partial and/ or full-thickness cartilage wear and delaminating cartilage lesions along the anterosuperior peripheral aspects of the acetabular articular surface.14 The origin of the femoral head-neck junction prominences is not well understood, though congenital or developmental entities are possible. Acquired defects, including those from prior fracture or slipped capital femoral epiphysis, or coxa magna resulting from Legg-CalvéPerthes disease can also result in cam impingement. Pincer-type FAI results from overcoverage of the femoral head due to an excessively deep or malformed acetabulum. This can result from prior trauma, the presence of large os acetabuli, abnormal acetabular version, coxa profunda, or abnormal pelvic tilt.150 Such overcoverage causes contact between the acetabular rim and femoral head-neck

CHAPTER 62  Hip and Pelvis

A

2149

B FIG 62-16  Pincer impingement in a 48-year-old man with left hip pain and clinical symptoms of femoroacetabular impingement. A, Anteroposterior radiograph of the pelvis shows a lateral osseous projection from the superolateral acetabular rim. The anterior acetabular wall (white arrowheads) projects lateral to the posterior wall (black arrowheads), indicating acetabular retroversion. These findings correspond to the diagnosis of pincer impingement. B, An oblique axial T1-weighted fat-suppressed MR arthrogram in the same patient shows deep cartilage loss along the anterosuperior (arrows) and posteroinferior (arrowhead) aspects of the hip joint that involves both the femoral and acetabular surfaces. This pattern is typical in pincer impingement. Of note, the acetabular labrum was intact.

junction.14 In these cases the typical result is levering of the femoral neck, caudal shifting of the femoral head within the acetabulum, and cartilage wear of the posteroinferior and anterosuperior aspects of the hip joint (Fig. 62-16). Injury to the acetabular labrum occurs first in pincer FAI, followed by anterior cartilage injury. Both CT and MRI aid in the evaluation of patients with FAI. CT is useful in assessing the osseous anatomy of the acetabulum and proximal femur, especially in patients for whom surgery is being considered.128 MPR images allow excellent evaluation of osseous anatomy. This may include radial images of the proximal femur and acetabulum and oblique axial images through the femoral head and neck. In addition, 3D volume-rendered images of the pelvis and proximal femora are useful in localizing the position of osseous anatomic abnormalities for surgical planning purposes. These 3D images may also bring to light abnormalities that are not immediately apparent on 2D images.37,128 MR arthrography is an important diagnostic tool in patients with FAI; it allows for accurate identification and characterization of labral tears and hip articular cartilage defects. Surgical treatment for FAI must address both the structural osseous abnormalities and the underlying secondary soft tissue defects. Bony corrective procedures include femoral head-neck osteoplasty and acetabular osteoplasty. Soft tissue corrective procedures include labral repair and/or débridement, articular cartilage débridement, and potentially microfracture or other cartilage restorative procedure.14

Other Impingement

Snapping Tendon Syndromes and Other Symptomatic Tendon Abnormalities

Snapping iliopsoas.  Iliopsoas impingement is a common cause of snapping hip. Classically thought to result from catching of the iliopsoas tendon on the supra-acetabular ridge during hip flexion and extension, further investigation using dynamic US has determined that a majority of these cases result from sudden reversal of the mediolat-

eral relationship between the iliopsoas tendon and the adjacent overlying iliacus muscle belly.87 This arrangement results in exertional pain associated with a catching or popping sensation in the anterior hip and lateral groin during hip flexion and external rotation.53,87 MRI and CT do not demonstrate dynamic evidence of snapping iliopsoas tendon.87 MRI, however, can be helpful in demonstrating the secondary signs of iliopsoas tendon irritation and surrounding inflammation. If present, this usually includes excess fluid within iliopsoas bursa.87 Synovial thickening and debris within the iliopsoas bursa, surrounding soft tissue edema, and underlying bone marrow edema pattern may also be seen. Thickening and increased intrasubstance signal involving the iliopsoas tendon and/or edema within the overlying iliacus muscle belly could also be encountered. Typically there  is no role for CT in the evaluation of this abnormality, though if   abundant in sufficient quantities, iliopsoas bursal fluid is detectable (Fig. 62-17). Snapping iliotibial band/tensor fascia lata.  Another cause of snapping hip is catching of the iliotibial band (IT band) and/or tensor fascia lata muscle as it passes anterior and posterior to the greater trochanter during hip flexion and extension, respectively.33,79 This can lead to pain with movement as well as irritation and inflammation intervening periods of rest. MRI is useful in identifying the sequelae of adjacent inflammatory change, including adventitial bursa formation and surrounding soft tissue edema (Fig. 62-18). The typical MR examination will not show dynamic evidence of snapping. Hamstrings tendinopathy and snapping.  Tendinopathy and tearing of the proximal hamstrings tendons are a common cause of posterior thigh pain that is worsened with sitting and eccentric loading of the hamstrings muscles. In addition, strains of the proximal hamstring myotendinous unit are commonly encountered in athletic individuals.12 Findings of hamstring injury and/or degeneration are often invisible on radiographs unless there is an accompanying osseous avulsion fragment.

2150

PART II  CT and MR Imaging of the Whole Body

MRI is useful in identifying hamstring pathology.12,38 Research has established that normal proximal hamstring tendons can have small amounts of increased intrasubstance signal and increased fluid signal within the surrounding soft tissues.38 This includes increased tendon size and marrow edema within the ischial tuberosity. Nevertheless, extensive tendon thickening, focal discontinuity along the length of the

proximal hamstrings tendon as well as absent proximal tendon components, and bare ischial tuberosity cortex are evidence of clinically significant disruption12 (Fig. 62-19). The hamstrings may also catch on the ischial tuberosity during hip movement, resulting in a snapping sensation.12 Although the snapping phenomenon is not directly visible during a typical MR examination,

FIG 62-18  Iliotibial band impingement. Axial T2-weighted fatFIG 62-17  Iliopsoas bursitis in a 61-year-old woman with long-standing left hip and groin pain. This axial T2-weighted fat-suppressed MRI shows fluid signal within the distended left iliopsoas bursa, compatible with the diagnosis of iliopsoas bursitis.

A

suppressed MRI of the left hip in an 18-year-old woman with chronic left hip pain shows increased signal within thickened portions of the proximal left iliotibial (IT) band (arrows). MRI depicts increased fluid signal and adventitial bursa formation subjacent to the IT band. This constellation of findings can accompany the clinical diagnosis of IT band impingement on the greater trochanter.

B FIG 62-19  Hamstrings tendinopathy in a 44-year-old active woman who presented with buttock pain. Axial T2-weighted fat-suppressed (A) and coronal short tau inversion recovery (STIR) (B) MRI demonstrate increased signal and thickened left hamstring origin (arrows) that predominantly involves the conjoined tendon and represents tendinopathy. Also note subjacent marrow edema pattern in the ischial tuberosity and surrounding soft tissue edema.

CHAPTER 62  Hip and Pelvis

A

2151

B

C FIG 62-20  Ischiofemoral impingement in a 24-year-old woman with spastic quadriplegia. A, Coronal STIR MRI of the pelvis shows edema within the left medial femur and edema/fluid between the ischial tuberosity and lesser trochanter (arrow). B, Axial T2-weighted fat-suppressed image shows edema within the lesser trochanter and ischial tuberosity secondary to osseous impaction. Note the edema/fluid within the quadratus femoris muscle, which is compressed (arrow). C, Axial CT image demonstrates narrowing of both ischiofemoral spaces (arrows).

the sequelae of associated inflammation are often apparent, including tendinopathy, tendon tearing, pronounced surrounding soft tissue edema, and underlying bone marrow edema pattern.

Ischiofemoral Impingement.  Ischiofemoral impingement syndrome is a clinical diagnosis associated with pain related to entrapment of the quadratus femoris muscle belly between the lesser trochanter of the femur and the ischial tuberosity/hamstrings tendon unit.154 This entity is more frequently found in women154,155 and has been associated with ischiofemoral space narrowing of positional, acquired, or congenital2 origin. Narrowing is often bilateral and can represent an incidental finding in patients imaged to evaluate pain in other anatomic regions.2 Internal/external rotation, adduction/abduction, and flexion/ extension can provoke symptoms.149,154,155 Patients with ischiofemoral impingement present with chronic pain in the groin and/or buttock, without history of traumatic injury.13,118,126 Pain may radiate into the lower extremity, with thigh and/or knee pain,84,126,154 because the quadratus femoris muscle is in close proximity to the sciatic nerve, and pressure effects may produce nerve-related

symptoms. Patients may also suffer from snapping sensation, crepitation, or locking in the joint.2,154,155 Radiographs are typically normal, although cystic change and sclerosis may be seen within the lesser trochanter or ischial tuberosity.149 MRI demonstrates ischiofemoral or quadratus femoris space narrowing and quadratus femoris muscle abnormalities including edema and/ or fatty replacement. These changes are best seen at the site of maximal impingement, which is typically best visualized on axial images. Bone marrow edema of the ischium/lesser trochanter or cystic change may also be apparent (Fig. 62-20). Additional imaging features that can suggest impingement include bursa formation, edema within and tears of the hamstrings tendons, or edema surrounding the iliopsoas tendon insertion.154,155 Other causes of quadratus femoris muscle edema include muscle strains or tearing, which should be considered in cases without narrowing of the ischiofemoral or quadratus femoris space.

Greater Trochanteric Pain Syndrome Pain about the greater trochanter can result from a variety of different etiologies. The term greater trochanteric pain syndrome is meant to

2152

PART II  CT and MR Imaging of the Whole Body

describe a spectrum of disorders, including most notably gluteus medius and minimus tendinopathy, tendon tears, bursitis, enthesitis, and calcific tendinitis, among other more nonspecific causes of pain.91,144 Radiographs performed on these patients range from normal to subtly abnormal, including greater trochanteric enthesopathy, hydroxyapatite deposition in the region of the gluteal tendons or bursae, and periosteal reaction. In severe cases of bursitis there may be displacement of overlying subcutaneous fat. MRI has become the preferred imaging test to identify the source of greater trochanteric region pain and classification of the disease process for diagnostic and prognostic purposes.91 Three important bursae lie in the region of the greater trochanter. This includes the largest and most commonly symptomatic greater trochanteric bursa, which lies between the superficial surfaces of the gluteus medius tendon, gluteus minimus tendons, and linea aspera  of the proximal femur and the deep margin of the gluteus maximus muscle. The subgluteus minimus bursa lies anteriorly between the distal gluteus minimus tendon and lateral cortex of the greater trochanter, and the subgluteus medius bursa lies posterolaterally between the distal gluteus medius tendon and the posterolateral cortex of the greater trochanter.127 More than a trace physiologic amount of fluid within the bursae correlates with the diagnosis of bursitis. Increased fluid within the bursae may occur with or without associated internal debris and synovial thickening. There may also be surrounding soft tissue edema and bone marrow edema pattern within the sub­ jacent greater trochanter. It should be noted that the isolated presence of bursal fluid occurs commonly in asymptomatic patients.  Care should be taken to correlate imaging findings with the patient’s symptoms or the presence of supportive secondary findings when possible.127 Considered to be the rotator cuff tendons of the hip, the gluteus medius and minimus myotendinous units are important abductors of the hip and stabilizers of the pelvis during bipedal ambulation. Injuries of the tendons are common and range from tendinopathy to partial and full-thickness tearing. MRI is accurate in identifying and staging gluteal tendon injury. Coronal and axial fluid sensitive and T1-weighted images are useful in demonstrating abnormal thickening, focal discontinuity, and retraction of diseased gluteal tendons. This is often accompanied by surrounding bursitis and can also be associated with surrounding soft tissue edema and subjacent bone marrow edema pattern in the greater trochanter91 (Fig. 62-21). Gluteus medius and minimus muscle strains are also possible but are less common compared to the distal tendon disruption.

Athletic Pubalgia Previously known as sports hernia, athletic pubalgia refers to chronic or acute injury to the shared aponeurosis of the rectus abdominis and adductor longus muscles in the region of the pubic symphysis and parasymphyseal margins of the pubic bones.110,120 This entity commonly affects young athletic individuals, is more common in males, and results in focal groin pain that has at times been confused for inguinal hernia symptoms.1,167 The use of MRI to evaluate this condition has allowed further elucidation of the associated patterns of injury.167 The sheetlike inferior aponeurosis of the rectus abdominis muscle is continuous with the proximal aspect of the adductor longus aponeurosis. The resultant shared aponeurotic plate is fused to the anterior surface of the pubic tubercle and pubic body slightly off midline in the ipsilateral pubic region. The term athletic pubalgia refers to a spectrum of injuries that includes tendinopathy of the aponeurosis, separation of the aponeurotic plate from the pubic bone, frank tearing of the aponeurotic plate, myotendinous strain within the rectus abdominis

or abductor longus muscles, pubic tubercle apophysitis, and stress reaction or stress fracture in the parasymphyseal region of the pubic bones.120 Radiographs, though routinely performed in these patients as an initial diagnostic measure, rarely reveal abnormality.85,120 Further, CT has an extremely limited role in the evaluation of this condition.85,120 The majority of important prognostic and diagnostic information is obtained using MRI, where findings include thickening and increased signal within the aponeuroses or aponeurotic plates, secondary cleft formation between the aponeurotic plate and pubic bone, focal discontinuity at the myotendinous junction of either the rectus or adductor longus muscle bellies with surrounding muscle contusion and edema, or focal marrow edema pattern and/or visible fracture cleft in the adjacent pubic bone110,120 (Fig. 62-22). Massive tendinous avulsion is possible but is extremely uncommon. MRI is also helpful in the evaluation of patients post treatment, which usually consists of nonoperative techniques including decreased activity and directed rehabilitation.85 In the rare case that surgery is performed, MRI is helpful in evaluating the integrity of the surgical repair. Osteitis pubis is considered by some to be on the spectrum of athletic pubalgia,167 though this inflammatory degenerative condition of the pubic symphysis and adjacent pubic bones may also result from other inciting factors, including childbearing, pelvic surgery, trauma, abnormal lower extremity mechanics, rheumatologic and degenerative disease, as well as explosive running and jumping activities.57,71,123 Imaging findings of osteitis pubis include reactive sclerosis along the margins of the pubic symphysis, osseous remodeling and hypertrophy, subchondral cystic change, and occasionally erosion along the margins of the symphysis. On MRI these findings are often accompanied by adjacent marrow edema pattern, as well as increased fluid within the capsule of the pubic symphysis and degeneration of the articular disk and adjacent hyaline and fibrocartilage.105,123

Arthritis

Osteoarthritis.  Osteoarthritis of the hip and sacroiliac joints remains a diagnosis that is most commonly made and subsequently followed on radiographs.135,143 Characteristic findings of nonuniform superior joint space narrowing, subchondral sclerosis, and osteophyte formation are diagnostic.100,107,143 Earlier evidence of focal cartilage loss is often visible on cross-sectional imaging, particularly on MRI.30,100 Nevertheless, CT and MRI are generally of limited diagnostic use in typical patients with osteoarthritis, though they can be helpful in staging this disease process and assessing for associated complications.10,100,143 Because osteoarthritis or focal cartilage loss often accompanies or is secondary to other diseases of the pelvis and hips, these findings are often encountered incidentally or in combination with other pathology. CT and MRI are increasingly useful in preoperative planning for hip arthroplasty.135 Joint effusions in osteoarthritis are common. A subset of hip osteoarthritis known as rapidly progressive osteoarthritis can, in addition to rapid nonuniform loss of hip joint space over the course of a few months,22,39 also demonstrate exuberant inflammatory changes in the bones and surrounding soft tissues on MRI.22

Inflammatory Arthritis.  Inflammatory arthritides such as rheumatoid arthritis and seronegative spondyloarthropathy demonstrate characteristic findings of inflammation and marginal erosions on MRI and CT.109 CT is helpful to identify osseous manifestations of developing secondary osteoarthritis as well as any other associated acquired anatomic defects, including axial migration of the femoral head and protrusio acetabuli, which are often seen in patients with rheumatoid arthritis.68

CHAPTER 62  Hip and Pelvis

A

2153

B

C FIG 62-21  Greater trochanter pain syndrome. A and B, A 48-year-old man with left hip and proximal thigh pain. Axial T2-weighted fat-suppressed (A) and coronal STIR (B) MRI of the left hip show a linear band of fluid signal (arrows) in the region corresponding to the location of the left greater trochanteric bursa. This lies interposed between the distal portions of the gluteus maximus muscle belly and the greater trochanter of the proximal femur. C, Coronal STIR image of the right hip of a 64-year-old woman with right hip pain shows focal discontinuity involving a portion of the right gluteus medius tendon insertion. The tendinous defect is filled with fluid (short arrow). The retracted redundant tendon is also evident (long arrow).

MRI is helpful in demonstrating the presence of hypertrophic and enhancing synovium, which can be identified following administration of intravenous contrast. Active inflammatory arthritis typically also includes bone marrow edema pattern along joint margins, erosions, layering debris, synovitis, and capsular thickening. Bursitis and tenosynovitis often accompany these findings.72,109,121 MRI is a helpful tool in determining joint fluid versus synovitis (Fig. 62-23). T2-hyperintense material within the joint capsule on

precontrast images can represent either joint fluid or synovitis.121 Matching diffuse enhancement on postcontrast images indicates synovitis. Though these entities almost always coexist, attempted joint aspiration can be avoided in patients with extensive synovitis and scant amounts of intracapsular fluid.

Septic Arthritis.  Imaging findings in septic arthritis can appear similar to those in inflammatory arthritis, though involvement of

2154

PART II  CT and MR Imaging of the Whole Body

A

B FIG 62-22  Athletic pubalgia in a 20-year-old man with left groin pain. Sagittal (A) and oblique coronal (B) T2weighted fat-suppressed images through the left pubic region demonstrate a focal fluid-signal-intensity cleft (long arrows) beneath the pubic attachment of the left rectus/adductor aponeurotic plate. Thickening and increased signal are also apparent in the proximal attachment of the left adductor longus tendon (arrowhead). Also note marrow edema pattern within the subjacent aspect of the left pubic bone. This patient also had degenerative remodeling, subchondral cyst formation, and marrow edema pattern along the margins of the pubic symphysis (short arrow), with associated nonuniform loss of joint space representing osteitis pubis.

B

A

FIG 62-23  Joint fluid versus synovitis in an 81-year-old woman with rheumatoid arthritis and right hip pain. A, Axial T2-weighted fat-suppressed MRI of the right hip shows T2-hyperintense fluid signal material distending the right hip joint capsule (arrows). B, Axial T1-weighted fat-suppressed postcontrast MRI obtained at the same level confirms that large portions of the aforementioned T2-hyperintense material represents synovitis (arrowhead), indicated by intracapsular enhancement. Additional areas of intracapsular nonenhancement (arrow) represents excess reactive joint fluid surrounded by a rim of synovitis.

septic arthritis is almost always monoarticular. Findings of septic arthritis are often invisible on initial radiographs, except in subacute cases that have progressed to the point where associated periarticular osteomyelitis are evident.86 Clinical concern for septic arthritis exists for those with monoarticular arthritis and hot, painful, and swollen joints with or without fever. Urgent joint aspiration is warranted in these patients because bacterial infection can result in rapid loss of articular cartilage and subsequent development of secondary osteoarthritis. Findings of septic arthritis may be encountered unexpectedly in patients undergoing cross-sectional evaluation for nonspecific hip pain or for other clinical complaints.86,104 Contrast-enhanced MRI is

helpful in identifying the presence of a joint effusion for identifying complications of septic arthritis, including osteomyelitis, sinus tract formation, thrombophlebitis, and extension into adjacent bursae86,96,142 (Fig. 62-24).

Infiltrative Disorders

Pigmented Villonodular Synovitis.  Pigmented villonodular synovitis (PVNS) is a benign synovial proliferative disorder characterized by nodular and/or villous synovial proliferations that fill involved joints. These proliferations characteristically contain variable amounts of hemosiderin and are often accompanied by a robust inflammatory

CHAPTER 62  Hip and Pelvis

2155

* *

A

B

* *

* *

C

*

*

D FIG 62-24  Septic arthritis in a 58-year-old woman with progressive right hip pain and fever. A, AP radiograph of the right hip demonstrates large erosions in the femoral head and acetabulum (arrowheads), as well as complete loss of joint space and marked sclerosis along the margins of the joint. Coronal T1- (B) and T2-weighted fat-suppressed (C) MRI of the right hip demonstrate extensive hip joint destruction and erosion involving the acetabulum and femoral head, surrounded by extensive marrow edema pattern (asterisks). Fluid signal material and heterogeneous debris distend the right hip joint capsule (arrowheads). Extensive soft tissue edema in the surrounding musculature is also evident on the T2 image. D, Coronal T1-weighted fat-suppressed postcontrast MRI through the same region shows extensive enhancement (asterisks) within the femoral head and neck, as well as within adjacent portions of the acetabulum, with fluid and synovitis (arrowheads) within the joint capsule.

response. PVNS can result in destruction of articular cartilage, subchondral cyst formation, osseous erosion, and extensive reactive marrow edema pattern.23,25 The diffuse form of the disease is almost always monoarticular, and the hip is the second most commonly involved joint, with the knee accounting for approximately 80% of cases. Though the exact etiology of PVNS is unknown, most authors believe it results from an inflammatory process.122 Radiographic find-

ings of PVNS are more commonly seen surrounding the hip compared to the more capacious knee joint and include marginal pressure erosions in the head and neck of the proximal femora caused by hypertrophic synovial masses. Rarely these masses can calcify. Adjacent nonuniform joint space loss resulting from secondary osteoarthritis is also possible.25 Imaging findings on CT are similar, though the synovial masses are more easily identified in association with the joint space, given

2156

PART II  CT and MR Imaging of the Whole Body

the improved 3D resolution. CT often demonstrates capsular thickening, surrounding pericapsular fat stranding, and joint fluid.23 The synovial masses are higher in attenuation and therefore distinct from the bland reactive fluid within the joint capsule on noncontrast CT, although nonspecific. Masses are outlined by contrast on the right CT arthrography. MRI findings in diffuse PVNS are well characterized and in the absence of radiographic findings can be the path to initial diagnosis. Given that PVNS is found in a relatively young population, these patients are often imaged for nonspecific hip pain thought to be due to labral, capsular, or tendinous etiology. MRI findings include intraarticular masses that are intermediate signal on both T1- and T2-weighted images and classically contain scattered punctate areas of low signal corresponding to areas of hemosiderin deposition (Fig. 62-25). These low-signal areas characteristically bloom on gradient-recalled echo and T2* images. MRI also demonstrates joint fluid, capsular thickening and edema, and edema within the surrounding soft tissues.23,82 In addition, MRI provides additional characterization of the osseous changes adjacent to synovial proliferative masses, such as focal marrow edema pattern, pressure erosion, and loss of articular cartilage.23,82 Direct MR arthrography demonstrates the above findings, with PVNS masses appearing as well-defined globular filling defects.

A

B

C

FIG 62-25  Pigmented villonodular synovitis (PVNS). A, AP radiograph obtained following open synovectomy in a 13-year-old girl with left hip pain. This image demonstrates sclerosis in the medial and inferior aspects of the acetabulum (arrowheads), with associated erosions. Coronal T1-weighted (B) and fat-suppressed T2-weighted (C) images of the left hip show aggregations of hypointense material (arrows) that distends the left hip joint capsule. MRI confirms sclerosis and erosions in the medial acetabulum evident on radiograph.

Intravenous contrast-enhanced MRI will show enhancement of synovial masses and reactive otherwise normal synovium within the joint. The localized version of PVNS, also known as giant cell tumor of the tendon sheath, can involve the bursae and tendon sheaths about the hip. This entity tends to have fewer of the characteristic internal areas of low signal and blooming artifact, though other associated findings are similar.132 Treatment of diffuse PVNS can be operative or nonoperative. Operative treatment consists of synovectomy, and recurrence is fairly common (40%-50% of cases).122,132 External beam radiation has been used for recalcitrant cases, though rates of recurrence are similar to synovectomy.122 Intraarticular injection of yttrium 90 has also been shown to be effective in the treatment of PVNS in combination with surgical synovectomy.90 MRI and to a lesser extent CT are useful in evaluation of hip pain following treatment to assess disease recurrence, secondary osteoarthritis, and/or development of osteochondral lesions within the joint.23,25,82,117

Synovial Chondromatosis.  Synovial chondromatosis can occur in primary or secondary forms. The primary form consists of a synovial proliferative disorder characterized by accumulation of cartilaginous nodules in the subsynovium of large joints.138,166 As the nodules enlarge, they detach from the synovial lining and become loose within the joint. These nodules often undergo endochondral ossification, which would then be called synovial osteochondromatosis.138 In primary synovial chondromatosis, articular nodules tend to be small and relatively uniform in size. Like PVNS, synovial chondromatosis is most common in the knee and second most common in the hip.138,166 The cartilaginous nodules can be scattered throughout the joint or completely fill and distend the hip joint capsule. From 70% to 95% of radiographs demonstrate the typical appearance of intraarticular chondroid calcifications. Pressure erosions of the acetabulum or proximal femur are apparent in up to 50% of cases. Cross-sectional imaging with CT or MRI can be helpful if radiographic findings are nonspecific.54,138 On CT, calcified intraarticular nodules are evident within the joint in addition to varying amounts of joint fluid (Fig. 62-26). Pressure erosion is often clearly visible. The MRI appearance can vary depending on the ratio of cartilage and calcific or ossific components but typically includes low to intermediate T1 and high T2 signal nodules within the joint capsule that also contain punctate areas of low signal corresponding to areas of calcification.54,138 Ossific (as opposed to chondral) bodies have signal characteristics identical to normal bone marrow. In addition, MRI is useful in demonstrating joint effusion, erosion, and marrow edema pattern within the adjacent osseous structures. Secondary chondromatosis typically results from large dislodged cartilage fragments in the setting of osteoarthritis, following trauma, or following infection. In the secondary type, articular nodules are often large, vary widely in size, and frequently ossify. A lamellated appearance has been ascribed to the nodules of secondary chondromatosis, resulting from their gradual evolution within the joint space.

Amyloidosis.  Amyloidosis in the musculoskeletal system typically

results from β2-microglobulin deposition in patients undergoing chronic hemodialysis or light chain deposition in patients with multiple myeloma.34,50 Regardless of the source of amyloid, this disease results in deposition of amyloid within the synovial membranes of joints and tendon sheaths throughout the body.26 On MRI, amyloid deposits are evident as low to isointense T1 and iso- to hyperintense T2 aggregation along the synovial margins of the joint. As the disease progresses, these aggregations cause well-defined large erosions in the proximal femur and acetabulum and often extend

CHAPTER 62  Hip and Pelvis

2157

B

C

A

FIG 62-26  Synovial chondromatosis. A, AP radiograph of 42-year-old woman with chronic left hip pain demonstrates scattered rounded calcified radiopacities surrounding the left femoral head and neck. B and C, Axial CT images through the right hip of a 61-year-old man with right hip pain show similar rounded calcific densities clustered within the cotyloid fossa. Calcific joint bodies in both of these patients are small and relatively uniform, characteristic of primary synovial chondromatosis.

A

B

C

FIG 62-27  Articular amyloidosis. A, AP radiograph of the right hip demonstrates large erosions in the inferior aspect of the femoral head-neck junction (arrow) and increased soft tissue density about the right femoral neck (arrowheads). Axial T1- (B) and T2-weighted fat-suppressed (C) images through the right hip of a 59-yearold man on long-term hemodialysis for chronic renal failure show large amounts of isointense material (arrows in C) distending the right hip joint capsule and extending anteriorly into the iliopsoas bursa. The T1 image demonstrates extensive erosion (arrowhead in B) along the anterior aspect of the femoral head and neck. Though not depicted in these images, extensive acetabular erosions are also present in this patient.

into adjacent bursae, tendon sheaths, and surrounding soft tissues.34 In advanced cases, this creates a “lumpy-bumpy” appearance of the overlying soft tissue contours. Surrounding bony and soft tissue inflammation is also present34,61 (Fig. 62-27). In advanced cases, amyloid deposition adjacent to bone can result in marrow replacement simulating the appearance of disseminated metastatic disease. Patients with amyloidosis often undergo CT examination for evaluation of underlying chronic disease. As a result, articular manifesta-

tions are often visualized and include high-attenuation soft tissue aggregation along the synovial margin of the hip joint, with large articular erosion surrounded by a thin cortical rim and variable amounts of fluid attenuation within the joint capsule.61,157

Osteomyelitis Osteomyelitis of the pelvis and about the hip can result from direct seeding from adjacent skin ulceration, trauma, hematogenous seeding,

2158

PART II  CT and MR Imaging of the Whole Body

*

* A

*

B

C

FIG 62-28  Ischial decubitus ulcer and ischial osteomyelitis in a 41-year-old paraplegic man with long-standing bilateral ischial decubitus ulcers and recent onset of fever. A, Axial T1-weighted MRI demonstrates a low signal ulcer tract (arrow) that extends from the cutaneous margin to the cortex of the ischial tuberosity. Focal cortical erosions are evident along the posterior margin of the ischial tuberosity, with subjacent replacement of fatty marrow signal (asterisk). B, Axial T2-weighted fat-suppressed MRI through the same region demonstrates fluid signal and debris within the ulcer tract (arrow) and marrow edema pattern (asterisk) within and surrounding the corresponding aforementioned region of fatty marrow replacement. Also note surrounding soft tissue edema. C, Axial T1-weighted fat-suppressed postcontrast MRI shows medullary enhancement (asterisk) corresponding to the region of marrow edema pattern. This image also demonstrates nonenhancing debris within the ulcer tract, surrounded by reactive inflammatory enhancing soft tissue (arrow).

prior surgery/instrumentation,42,64,133 or spread of infection originating in an adjacent infected joint.86 CT is of limited utility in the identification of acute osteomyelitis insofar as bony information beyond what is already evident on radiographs is rarely revealed by additional CT images. In addition, radiographic and CT findings of osteomyelitis may take 7 to 10 days to become apparent. For these reasons, pre- and postcontrast MRI is the technique of choice to assess these patients.76,156 The presence of marrow replacement (low signal on T1-weighted images) is the most reliable indicator of osteomyelitis.69,76,156 This finding is usually accompanied by marrow edema pattern and enhancement within and surrounding the area of marrow replacement69,156 (Fig. 62-28). Edema pattern and enhancement within the marrow in the absence of replacement of marrow fat, though easy to identify, usually does not indicate active osteomyelitis. This pattern is instead thought to represent reactive inflammation in the adjacent bone, or reactive osteitis. Findings of true osteomyelitis are often accompanied by periostitis, soft tissue edema, and adjacent signs of soft tissue inflammation.11,156 MRI is helpful in identifying the presence of subperiosteal abscesses in addition to findings of chronic osteomyelitis, including soft tissue sinus tracts, osseous cloacae, bony sequestra, and Brodie’s abscesses.11,76 Surrounding signs of soft tissue infection and inflammation are also easily identified on MRI.11,156 CT changes indicative of osteomyelitis are cortical destruction, periostitis, and loss of medullary fatty attenuation (Fig. 62-29). CT  can also demonstrate signs of surrounding soft tissue inflammation, including fat stranding, muscle enlargement, skin thickening, and organized fluid collections.52 Patients with long-standing decubitus ulcers or known longstanding chronic osteomyelitis represent a subset of patients in whom the diagnosis of active osteomyelitis can be difficult on CT and MRI. This is because preexisting areas of chronic smoldering infection and fibrosis will appear as areas of replaced fatty marrow on MRI and dense

A

B FIG 62-29  Ischial decubitus ulcer and ischial osteomyelitis. A and B, Two CT images of a 55-year-old paraplegic man with a chronic left sacral decubitus ulcer show the base of a gas-filled ulcer tract that extends to the cortex of the ischial tuberosity (arrowhead). Both images show erosion of the adjacent cortex of the ischial tuberosity (arrows), with reactive sclerosis in the subjacent portions of the ischium. Also note lack of any detectable fat attenuation within the medullary space of the left ischial tuberosity compared to the noninfected right ischial tuberosity. CT images are not specific in distinguishing changes from chronic or prior osteomyelitis from active osteomyelitis.

CHAPTER 62  Hip and Pelvis sclerosis on CT in regions of prior osseous erosions and/or osseous débridement defects. In these cases, comparison to prior imaging is critical in assessing for new areas of marrow replacement or bone destruction.

Soft Tissue Infection Soft tissue infection about the hip can take the form of cellulitis, myositis, fasciitis, septic tenosynovitis, thrombophlebitis, and septic arthritis (previously discussed). These infections can range in severity from minimal to life threatening. Effective clinical triage is necessary to ensure that prudent imaging strategies are employed that guide therapeutic decision making, employ rapid efficient imaging assessment when life-threatening conditions are suspected, and avoid imaging in cases of minor superficial infections. As with other musculoskeletal pathology, radiographs are an important initial imaging technique to assess for the presence of retained radiopaque foreign body, soft tissue gas, and signs of underlying osseous involvement.52,76 In the appropriate clinical setting, organized or compartmentalized infected fluid collections or areas of infected devitalized soft tissue require targeted intervention for adequate treatment.76 Contrast-enhanced MRI and CT are helpful in identifying such foci.52 MRI and CT can also identify features of an organized fluid collection, such as internal debris, septations, thickened walls, and surrounding inflammatory changes that raise concern for infection if the initial diagnosis is uncertain. MRI is superior to CT in detecting fluid collections, though contrast-enhanced CT can improve the visibility of infected fluid collections by demonstrating the enhancing surrounding rim (Fig. 62-30). If necrotizing fasciitis is suspected, the diagnosis can often be made on a clinical basis. If imaging evaluation is necessary, MRI should be avoided because it could introduce an unreasonable delay in time  to surgical débridement, for which there could be life-threatening  consequences. The extremely short acquisition times and exceedingly rare contraindications of CT make it an acceptable choice for imaging in these patients, which will usually show extensive soft tissue gas  (Fig. 62-31).

*

A

Tumors and Tumorlike Conditions Osseous Neoplasms

Osteoid osteoma.  These benign cortically based tumors occur most frequently in patients aged 5 to 25, though occurrences in patients outside of this age range should still be expected.9,29,93 The clinical presentation most commonly involves deep bone pain that is worse at night and relieved by nonsteroidal antiinflammatory medications.9,29 Osteoid osteomas occur most frequently in the long bones and are particularly common in the femora.20 These tumors may rarely occur in the pelvis.20 The lesion consists of a central lytic nidus with an inner core composed of osteoblasts, osteoid, and woven bone surrounded by a rim of fibrovascular tissue. The inner core tends to have a varying amount of mineralization. The entire nidus is surrounded by thickened and reactive bone, generating a classic appearance on imaging examinations.93 Osteoid osteomas are frequently encountered on imaging examinations performed to evaluate hip or pelvic pain of unknown etiology. Although the radiographic appearance of osteoid osteomas is characteristic, they may be radiographically occult, especially if they occur in the pelvis. CT and to a lesser extent MRI can be useful in further characterization of a known lesion8; CT is the tool of choice for evaluating these lesions.8 On CT, osteoid osteomas appear as a spherical central lytic nidus that is less than 1.5 cm in diameter,  with a central focus of high-attenuating mineralization surrounded by thick dense bone9,93 (Fig. 62-32). An enlarged vascular channel is frequently seen leading into the nidus. On MRI the nidus is surrounded by a discreet sclerotic rim and contains soft tissue components of variable signal on both T1- and T2-weighted images. Central mineralized components will be low signal on all sequences.29,93 The nidus is surrounded by uniformly low-signal thickened, reactive bone. MRI also shows surrounding reactive periostitis, soft tissue edema, and bone marrow edema pattern. If an osteoid osteoma lies close to or within a joint, signs of exuberant joint inflammation will also be present, including synovitis, joint effusion, capsular thickening, and capsular edema.29,103 This appearance can mimic inflammatory or infectious arthritis.75

* *

* *

B

2159

*

C

FIG 62-30  Soft tissue abscess in a 72-year-old quadriplegic man with severe posttraumatic left hip osteoarthritis and development of erythema and swelling in the region of the left hip. Axial T1-weighted (A), T2-weighted fat-suppressed (B), and T1-weighted fat-suppressed postcontrast (C) MRI of the left hip show a lobulated organized intramuscular and subcutaneous fluid collection. The contents of this collection (asterisks) are complex, with scattered areas of proteinaceous or bloody nonenhancing fluid. The fluid is surrounded by a rim of enhancement, with edema and enhancement in the surrounding subcutaneous soft tissues.

2160

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 62-31  Necrotizing fasciitis. A and B, Two axial CT images through the pelvis of a 65-year-old diabetic man show scattered clusters of gas attenuation in the soft tissues of the left buttock, left posterior thigh, scrotum, and perirectal and presacral fat (arrows) in this patient with Fournier’s gangrene (perineal necrotizing fasciitis). There is also stranding of subcutaneous and intermuscular fat, skin thickening, and muscle enlargement in these regions.

A

B

C

FIG 62-32  Osteoid osteomas. Axial (A) and coronal (B) CT images in a 10-year-old boy with left hip pain show a cortically based lytic lesion in the medial cortex of the proximal femur (arrows). The lesion is surrounded by thickened reactive cortex. The lytic central nidus is fusiform with a central ovoid focus of relatively higher attenuation. C, An axial CT image of the pelvis in a 21-year-old man with right hip pain shows a similar-appearing lesion in the cotyloid fossa of the right acetabulum. Osteoid osteomas in the pelvis are rare compared to those in the proximal femur.

Chondroblastoma.  Chondroblastoma is a benign cartilaginous tumor that occurs in the epiphyses of long bones (and epiphyseal equivalents) in patients younger than 20 years.20 Patients typically present with focal pain, stiffness, and atrophy of surrounding musculature.20 On radiographs and CT, these appear as lytic lesions displaying characteristic ring-and-arc calcifications typical of chondroid matrix, along with smooth and/or lobulated outer contour. There may be varying amounts of surrounding reactive sclerosis, and in the hip an effusion may be apparent. CT is also helpful in demonstrating adjacent endosteal scalloping or if present breach of surrounding cortex. The MRI appearance of chondroblastomas includes the classic appearance of chondroid lesions, with T2-hyperintense and T1-  isointense signal within the lesion20 that is surrounded by a low-signal lobulated or smoothly rounded outer margin. Chondroid calcifications appear as punctate central areas of low signal. As with osteoid

osteomas, chondroblastomas are surrounded by signs of exuberant reactive inflammation, including periosteal edema and enhancement, soft tissue edema, bone marrow edema pattern as well as effusion, synovitis, and capsular inflammation involving the adjacent joint.20 Chondrosarcoma.  The pelvis and less commonly proximal femora are common sites for chondrosarcomas to occur.20,45 These tumors usually occur in patients in their 50s and 60s and arise from the medullary space or less commonly from a juxtacortical location or from a preexisting osteochondroma. Chondroid mineralization pattern is present.20,45,114 These lesions do, however, also demonstrate imaging findings of more aggressive osseous lesions, including a wide ill-defined zone of transition, periostitis, and cortical breakthrough. Endosteal scalloping and extension into surrounding soft tissues is common. The extent of osseous destruction is easily assessed on CT.102,114,140 MRI performed without and with contrast is much more useful in confirming T2-hyperintense and enhancing stroma, as well as in

CHAPTER 62  Hip and Pelvis

2161

A

B

C FIG 62-33  Pelvic chondrosarcoma. A, AP radiograph of the pelvis demonstrates a large soft tissue lesion centered just lateral to the left iliac wing. This lesion demonstrates central ring and arch calcifications (arrows) characteristic of chondroid matrix. The lesion obscures visualization of the majority of the iliac wing. Axial T1-weighted (B) and T2-weighted fat-suppressed (C) MRI of the left hip in a 61-year-old man with a painful slowly enlarging left buttock mass demonstrate a large lobulated lesion emanating from the lateral cortex of the left iliac bone (arrows). Lobulated contours and T2-hyperintense stroma with punctate foci of low signal calcification, some of which demonstrates a ring and arc configuration, indicate that this is a chondroid lesion. Scalloping of the lateral cortex of the posterior aspect of the iliac wing (arrows) is evident. These features combined with the location of the lesion and patient symptoms are characteristic of the diagnosis of chondrosarcoma.

determining the integrity of surrounding bone, extent of involvement of nearby muscular compartments, and the status of adjacent neurovascular structures102,114 (Fig. 62-33). Metastases and multiple myeloma.  Because of the preponderance of relatively vascular red marrow in the axial skeleton, both osseous metastases and multiple myeloma occur more frequently in this region.56 When present in the appendicular skeleton, for similar reasons, these lesions tend to occur more proximally.56 Accordingly the

pelvis and proximal femora are common sites of osseous metastases and myelomatous lesions. On CT, metastatic lesions tend to be focal spherical lytic and/or blastic lesions, depending on the tumor of origin.56,163 On MRI these lesions replace marrow fat, are enhancing, and have varying degrees of T2 hyperintensity depending on the sclerotic character, amount and age of internal blood products, and vascularity.139 Melanoma metastases may also demonstrate increased signal on T1-weighted images. CT

2162

PART II  CT and MR Imaging of the Whole Body

and MRI are also useful in assessing for the presence of cortical breakthrough and pathologic fracture.163 Multiple myeloma lesions are similar to metastases in their CT and MRI appearance. These lesions are most frequently lytic, though a rare blastic variant exists.6,74

Soft Tissue Masses.  There are numerous benign and malignant soft tissue masses that occur in the pelvis and about the hips.92,112 Common benign tumors include lipomas, fibromatosis, and nerve sheath tumors92,112,113 (Figs. 62-34 and 62-35). Soft tissue sarcomas and metastatic disease are the most common malignant etiologies.92,112 As with osseous neoplasms, MRI is a helpful tool for assessing the size and extent of lesions as well as evaluating for invasion of nearby

A

bones, muscular compartments, joints, and neurovascular structures. Such assessment is critical in the staging of malignant primary sarcoma lesions. The role of CT is much more limited in the evaluation of benign and malignant lesions of the soft tissues.92 Rarely CT can provide additional information regarding the mineralized or fatty components of a lesion or identify the presence of or further characterize the nature of any associated osseous destruction.92

Response to Therapy and Recurrence.  MRI is an extremely important tool in assessing treatment response and investigating the presence of recurrent or residual tumor following surgical resection, external beam radiation therapy, and systemic chemotherapy to treat

B FIG 62-34  Lipoma. Axial T1- (A) and T2-weighted fat-suppressed (B) images of the left hemipelvis show an ovoid lesion (arrows) arising between the left gluteus minimus and gluteus medius muscle bellies. The wellencapsulated appearance, lack of internal septations, and absence of soft tissue components are evidence of a lipoma in this 41-year-old patient.

A

B FIG 62-35  Nerve sheath tumor in a 29-year-old man. Axial STIR (A) and T1-weighted fat-suppressed postcontrast (B) images of the proximal right thigh show a round heterogeneously enhancing lesion (arrows) in the expected region of the right sciatic nerve that, following resection and pathologic analysis, was revealed to be a schwannoma.

CHAPTER 62  Hip and Pelvis malignant soft tissue lesions. In this setting, MRI without and with contrast is optimal in distinguishing viable tumor from surrounding treatment response.40,159 Findings of viable tumor include nodular or masslike areas of enhancement, notable or progressive mass effect, or expansion.40,60 Following radiation therapy, there is a characteristic evolution of findings in the soft tissues within the radiation port. This begins with subcutaneous edema, skin thickening, fascial edema, and feathery muscular edema in a well-defined region corresponding to the margins of the therapeutic radiation port. Edges of this region are often linear or angular, and there is associated diffuse enhancement of the involved subcutaneous fat, skin, and musculature. These changes initially pro­ gress before characteristically slowly receding over the subsequent months and years. Skin and fascia may remain thickened but with low

2163

signal on T1- and T2-weighted images. Fatty infiltration and/or atrophy of the muscles may also appear. Scarring and retraction of the soft tissues generates a characteristic linear or stellate pattern of enhancement. Though both will enhance, fascial tenting and retraction of chronic scarring appears in stark contrast to the nodular, rounded, expanded margins of tumor recurrence.60 Postsurgical changes follow a similar evolution and are centered around the surgical site. Surgical changes include edema and enhancement of the muscles, fascia, and subcutaneous fat about the operative bed60 (Fig. 62-36). Postoperative fluid collections (seroma, lymphocele, or hematoma) may also appear as nonenhancing aggregations, demonstrate variable internal signal characteristics on T1- and T2-weighted images, and are surrounded by a thin uniform enhancing rim (Fig. 62-37). Chemotherapy can cause necrosis of tumor and surrounding

A

B

C

D FIG 62-36  Evolution of posttreatment changes in a 51-year-old man with a previously resected anterior right thigh liposarcoma. Axial T2-weighted fat-suppressed (A) and T1-weighted fat-suppressed postcontrast (B) images obtained 6 months after surgery and external beam radiation therapy demonstrate edema and enhancement within the thickened subcutaneous fat and deep fascia (arrows), with adjacent edema within the bellies of the gluteal muscles but no areas of nodular enhancement or mass effect. Axial T2-weighted fat-suppressed (C) and T1-weighted fat-suppressed postcontrast images (D) in the same region obtained 2 years after treatment demonstrate decreased edema and enhancement within the fascia, soft tissues, and muscles. Once again there is no evidence of nodular enhancement or mass effect.

2164

PART II  CT and MR Imaging of the Whole Body

A

B

C

FIG 62-37  Postoperative seroma in a 47-year-old man 3 months following resection of a medial right thigh sarcoma. Axial T1-weighted (A), T2-weighted fat-suppressed (B), and T1-weighted postcontrast (C) MRI of the right thigh demonstrate a large fluid collection between the anterior and posterior compartment of the thigh in a region corresponding to the surgical bed. The fluid-filled seroma cavity (arrows) contains nonenhancing fluid and is surrounded by a thin rim of enhancement (arrowheads). There are no nodular areas of enhancement identified to suggest presence of recurrent tumor.

inflammatory change.60 Any nodular or enhancing internal elements should raise concern for the presence of viable tumor.60 Diffusionweighted MRI may be useful for characterizing lesions and assessing the treatment response of both osseous and soft tissue tumors.146

PROXIMAL FEMORAL BONE MARROW EDEMA Osteonecrosis This condition results from the devitalization of a fragment of bone that contacts the articular surface of the head of the proximal femur. There are numerous conditions that predispose to the development of osteonecrosis, including long-term or high-dose corticosteroid use, femoral head and neck fractures, hip dislocation, collagen vascular disease, hematologic disorders, radiation therapy, pancreatitis, and alcoholism among others. Most of the known effects of these conditions involve either a perturbation of blood flow to the proximal end of the femoral head or direct injury to femoral head marrow.21 A small subset of inciting conditions remains without a clear mechanistic link.21 When due to underlying systemic disease, osteonecrosis is most often bilateral. When due to prior trauma, it is most commonly unilateral. The initial insult results in a feed-forward cycle of inflammation, increased femoral head intramedullary pressure, further decrease in blood flow, additional necrosis of medullary bone, and further progression of inflammation.21 The normally vascularized bone surrounding the lesion becomes sclerotic in response to the adjacent inflammation. Once the acute process has subsided, the focus of subarticular bone remains devitalized and accordingly can no longer undergo the expected reparative mechanisms of normal bone. As a result, microscopic damage to the weakened subcortical trabeculae progresses, ultimately resulting in trabecular impaction. Loss of trabecular support in the subcortical medullary space leads to collapse and flattening of the articular surface and subsequently the development of secondary osteoarthritis. Radiographs are usually positive only in the latter stages of disease once loss of trabecular distinction and surrounding reactive sclerosis in the femoral head are apparent. The same is true for CT, though the improved 3D resolution capabilities allow for detection of these changes earlier than on radiographs.

MRI is the most sensitive and specific technique to identify osteonecrosis, especially during its initial stages.21,158 The earliest findings of osteonecrosis include patchy areas of marrow edema pattern within the femoral head158 and a characteristic serpentine rim of hypointense T1 and T2 signal surrounding the affected area of the femoral head, which usually involves at least the anterosuperior and superomedial aspects.21 In approximately 80% of patients, an additional serpentine rim of high-T2 signal lies within the low-signal rim. This appearance increases the specificity of diagnosis and has been termed the double line sign21,158 (Fig. 62-38). The marrow within the abnormal rim demonstrates a variable appearance on MRI depending on the extent of residual inflammatory change. Residual inflammation results in marrow edema pattern and low T1 signal, whereas burned-out quiescent defects demonstrate prominent fat signal but no edema.158 Inflammatory changes are also evident in the surrounding portions of the femoral head and neck, including marrow edema pattern or additional areas of bone infarct.158 Early pain associated with osteonecrosis results from the initial inflammation. Chronic pain symptoms result from additional inflammation related to subcortical trabecular failure and articular collapse, as well as from arthrosis related to the distorted geometry of the collapsed femoral head articular surface. On MRI these chronic changes are also readily identified.21,88,158 This includes subcortical cleft (crescent) formation in the regions of trabecular impaction that appear as linear areas of low signal beneath the subchondral cortex surrounded by varying amounts of marrow edema pattern. Flattening of the femoral head articular surface is also visible in cases of articular collapse. These patients typically have an associated joint effusion and varying amounts of synovial hypertrophy. Though not necessary for diagnosis, gadolinium-enhanced images will demonstrate nonenhancement of the central necrotic portions of the femoral head,158 enhancement in areas of surrounding reactive marrow edema pattern, and possibly enhancement of reactive synovium.

Transient Osteoporosis of the Hip Transient osteoporosis of the hip (TOH) is a rare entity that was initially thought to occur predominantly in the left proximal femora of young women in the latter stages of pregnancy or during the postpartum period. It is now known that this disorder can affect patients into middle age, more commonly occurs in men, and can involve either hip

CHAPTER 62  Hip and Pelvis

2165

*

B

A

*

C

*

D FIG 62-38  Osteonecrosis of the femoral head in a 28-year-old woman with sickle cell disease and right hip pain. A, AP radiograph of the right hip shows flattening of the medial aspect of the right femoral head articular surface (arrow), as well as focal sclerotic regions in the superolateral and inferomedial femoral head. Coronal T1-weighted (B), T2-weighted fat-suppressed (C), and T1-weighted fat-suppressed postcontrast (D) MRI of the right hip demonstrate a focus of osteonecrosis in the superior medial aspect of the right femoral head (asterisk). This contacts the articular surface of the femoral head and is surrounded by a characteristic serpentine rim of low signal intensity (arrowheads). The necrotic focus is surrounded by extensive marrow edema pattern and enhancement within the femoral head, extending into the femoral neck (arrows). Synovitis is also evident within the right hip joint capsule.

or (as is often the case in the postpartum period) both hips. The true etiology of this disorder remains unknown, though imaging findings are typically those associated with a reactive inflammatory process of the marrow in the head and neck of the proximal femur.160 Imaging findings in patients with TOH include asymmetrically decreased bone density in the involved proximal femur on radiographs and CT. Bilateral cases can be extremely difficult to detect radiographically. MRI is the study of choice for patients in whom the diagnosis of TOH is suspected.160 Typical MRI findings include bone marrow

edema pattern in the involved femoral head and neck that may spare the inferomedial portions. Marrow edema pattern typically corresponds in location to areas of low T1 signal160 (Fig. 62-39). Notably there is no visible fracture cleft in the femoral head or femoral neck.160 The presence of such should raise concern for an alternate diagnosis. The size of the region of edema is thought to have an inverse relationship to the likelihood of resolution following therapy, which usually includes a period of non–weightbearing and pain control. A subset of patient with TOH will progress to a condition known as

2166

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 62-39  Transient osteoporosis of the hip (TOH) in a 54-year-old man with left hip pain for 2 months and no history of trauma. A, AP radiograph of the pelvis shows extremely subtly decreased density in the head and neck of the left proximal femur compared to the right. B, Coronal STIR image of the left hip shows diffuse marrow edema within the femoral head and neck, compatible with TOH. No subchondral fracture visualized.

regional migratory osteoporosis in which patchy areas of marrow edema recurs in various locations in the affected lower extremity.27,152 A causal link between TOH and avascular necrosis has been  suggested by some, though the true relationship between TOH and either avascular necrosis or subchondral insufficiency fractures is unclear.88,131,158,160

Subchondral Insufficiency Fracture Recently subchondral insufficiency fracture of the femoral head (SIFFH) has been identified as an entity distinct from osteonecrosis and transient osteoporosis.77,78,162 This entity typically occurs in osteoporotic individuals who present with rapidly progressive severe hip pain.77,78,161,162 Radiographs in patients with SIFFH are either normal or, in cases without collapse, difficult to differentiate from osteonecrosis without articular collapse.77,80 In more advanced cases with articular collapse at the fracture site, radiographs may demonstrate a subtle area of flattening along the superior aspect of the femoral head articular surface and/ or a small linear sclerotic line in the subarticular marrow that parallels the articular surface.77,80 These findings are similar though slightly more apparent on CT. Owing to the overall increased risk of insufficiency fractures of the neck and intertrochanteric regions of the proximal femora in this patient population, if radiographs are negative, MRI is often performed to exclude a radiographically occult fracture. MRI in patients with decreased bone mineral density is both sensitive and specific in the diagnosis of SIFFH. On MRI the fracture line is typically evident and appears as low linear signal beneath the subchondral bone plate that parallels the articular surface of the femoral head.77,80,161 In the acute and subacute phases the fracture cleft is surrounded by T1 hypointensity and T2 hyperintensity corresponding to marrow edema pattern, medullary hemorrhage, and contusion (Fig. 62-40). Of note, there are no signs of devitalized bone in patients with SIFFH.77

TRAUMA Acetabular Fractures Acetabular fractures are the result of an axial impact to the femur or medially directed impact to the greater trochanter of the proximal

femur. Force is transmitted through the femur, driving the femoral head like the head of a hammer into the acetabulum. The result is one of several possible fracture patterns that involve not only the acetabular articular surfaces but may also disrupt the load-bearing properties of the bony pelvis (Fig. 62-41). High-energy fractures in patients involved in motor vehicle collisions or falls from great height make up the majority of these injuries; low-energy trauma in elderly patients with decreased bone mineral density make up a smaller but important subset.4,95 Associated injuries to pelvic viscera and pelvic vessels commonly accompany acetabular fractures resulting from high-energy mechanisms. Radiographs are typically obtained as the initial imaging evaluation. These usually identify the presence of a fracture, but they are often insufficient to adequately characterize the fracture pattern, and accordingly CT is performed. CT assessment is helpful in that MPR images are useful to map fracture extension and assess stability. In addition, CT is excellent in assessing the alignment of joints and osseous fragments.67,95 Despite limited contrast resolution, CT can also determine the location of associated hematoma and/or inflammation relative to the position of the sciatic and femoral nerves and nearby vessels.95 In the postoperative period, CT is helpful in establishing the extent of osseous union between fracture fragments. The Letournel classification system divides fractures based on the orientation and extension of the fracture line as well as the relative complexity of the fracture pattern. This system includes elementary patterns as well as complex patterns that represent combinations of elementary fractures and more extensive structural disruption98 (Fig. 62-42). The pattern of injury depends on the quality of the underlying bone, the position of the femur during trauma, and the force vector of the inciting event.98 A brief overview of this classification follows.

Elementary Fractures Posterior wall.  Most commonly occurring during a posterior hip dislocation event, posterior wall fractures involve fracture of some or all of the posterior rim of the acetabulum without involving the anterior or posterior columns. Head-on motor vehicle collision in which a patient’s knees forcefully contact the dashboard are a common inciting event resulting in posterior wall fractures.

CHAPTER 62  Hip and Pelvis

A

2167

C

B

FIG 62-40  Subchondral insufficiency fracture in a 64-year-old man with severe acute worsening of chronic right hip pain. A, AP radiograph of the right hip demonstrates focal sclerosis surrounding a flattened portion of the subarticular cortex of the superior weightbearing portion of the right femoral head (arrow). Coronal T1-weighted (B) and STIR (C) MRI of the right hip demonstrate flattening of the superior femoral head articular surface (arrows), with a subjacent linear hypointense fracture line (arrowhead) that parallels the articular surface of the femoral head, surrounded by marrow edema pattern.

lum co ior ter

Anterior wall

Posterior column

Posterior wall

lumn

An

A

Transverse

rior co

Poste

n

Anterior column

Posterior column + posterior wall

Transverse + posterior wall

Both column

FIG 62-41  Anterior and posterior columns of the acetabulum. Computer-generated illustration depicting the path of force transfer along the anterior and posterior columns during weightbearing.

B Posterior column.  A posterior column fracture results in a separation of the posterior column from the sciatic buttress such that it is no longer available for load bearing. Transverse.  Transverse fractures traverse both the anterior and posterior columns and typically result from a blow to the lateral aspect of the greater trochanter. Anterior column.  Anterior column fractures tend to occur in elderly patients and result in a separation of the anterior column from

Anterior column + hemitransverse

T-Shaped

FIG 62-42  Elementary (A) and complex (B) acetabular fracture patterns according to the Judet and Letournel classification system.

the sciatic buttress such that it is no longer available for load bearing. These can accompany an anterior dislocation or result from a medially directed blow to the posterolateral aspect of the greater trochanter. Anterior wall.  The anterior counterpart of posterior wall fractures, these fractures involve some or all of the anterior rim and

2168

PART II  CT and MR Imaging of the Whole Body

articular surface of the acetabulum. Unlike posterior wall fractures, anterior wall fractures are extremely rare.

Complex Fractures.  Complex acetabular fractures patterns include posterior column plus posterior wall, transverse plus posterior wall, both column, anterior column, and hemitransverse and T-shaped fractures. Of these injuries, transverse plus posterior wall is the most commonly occurring complex fracture pattern.

Proximal Femur Fractures Common fracture patterns to the proximal femur include femoral neck fractures, intertrochanteric fractures, and subtrochanteric fractures. These fractures tend to group in a bimodal distribution. The first of these two groups include elderly individuals with decreased bone mineral density.49,65,137 These patients, who are more frequently women, sustain low-energy traumatic mechanisms such as falls from standing. Because of decreased conspicuity of fracture defects in demineralized bones, these fractures are more likely to be invisible or extremely subtle on initial radiographs.28,124 In the elderly population, 10% of femoral neck fractures will be radiographically occult.124 MRI is the test of choice for additional evaluation in these patients because its sensitivity and specificity for fracture identification approaches 100%.58,65 On MRI fractures appear as linear areas of low signal surrounded by soft tissue edema, periosteal edema, marrow edema pattern, and osseous contusion.36,58 The second category includes young individuals involved in highenergy trauma, usually associated with motor vehicle collisions or falls from great heights. Though usually obvious on imaging, these injuries are often accompanied by additional injuries in the head, spine, torso, and extremities. CT can aid surgical planning in these patients, though cross-sectional imaging exclusively to evaluate the proximal femur fracture patterns is usually not necessary.

Stress Fractures Stress fractures can be grouped into two distinct categories: insufficiency (fragility) fractures that result from abnormal underlying bone that is exposed to normal levels of mechanical stress24 and fatigue fractures that result from normal underlying bone exposed to an abnormally high level of mechanical stress.101,151 Both types of stress fractures are frequently encountered about the hip. Fragility fractures of the femoral neck, intertrochanteric region, and proximal shaft occur most commonly in elderly osteopenic patients. These occur most commonly in the subcapital region of the femoral neck just distal to the head-neck junction. Fatigue-type fractures of the femoral neck and proximal femoral shaft are common in athletic individuals, especially distance runners and military recruits.48,151 These fractures tend to occur more frequently in the basicervical region. Subchondral fractures of the femoral head are discussed in a separate section. Insufficiency fractures are often occult on initial radiographs. This is thought to result from the poor conspicuity of these fracture patterns on a background of demineralized bone. CT is prone to similar insensitivity. As with other nondisplaced fractures of the proximal femur, MRI is both sensitive and specific in the diagnosis of stress fracture.32,44 Following initial radiographs, MRI is the diagnostic technique of choice in evaluating these patients44,151 (Fig. 62-43). Given the propensity of these fractures to progress with continued weightbearing and stress, authors do not distinguish between areas of stress reaction and stress fractures within the femoral neck. Stress fractures of the pelvis are also common and include fractures of the superior pubic ramus, ischiopubic ramus, supra-acetabular iliac bone, and sacrum.101

Atypical Bisphosphonate Fractures Recent investigation has revealed a link between long-term bisphosphonate use and insufficiency-type fractures of the proximal femoral shaft.94 These fractures tend to occur in patients who have been on bisphosphonate therapy for longer than 5 years.94,125 They are thought to result from decreased bone strength related to abnormal bone turnover that results from bisphosphonate-associated osteoclast inhibition.7,134 These fractures typically occur in the lateral cortex of the proximal to midfemoral diaphysis of the femur. Incomplete subtrochanteric and diaphyseal fractures may progress to complete fractures if treated conservatively, and therefore early detection of these incomplete fractures is critical to diagnose. Associated healing response  produces a characteristic lateral cortical thickening and a beaking appearance that may or may not include a visible transversely oriented fracture lucency on radiographs.7,31,130 These fractures are common, and imaging of the contralateral side should be performed because they are often asymptomatic. Compared to radiographs, CT is more helpful in identifying the fracture cleft and further visualizing adjacent periostitis.130 CT is also helpful in evaluating for fracture progression (Fig. 62-44). MRI aids in early detection of these fractures and is typically the first-line imaging technique after radiographs. MRI findings include focal cortical thickening and beaking of the lateral cortex of the proximal to midfemoral shaft, with periosteal edema, subjacent endosteal bone marrow edema, with or without fracture line (Fig. 62-45). Treatment usually includes internal fixation and a bisphosphonate and drug holiday.7,16,31

Avulsion Fractures Avulsion injuries about the hip are common in skeletally immature patients who overstress their extremities.51,147 Apophyseal avulsions of the anterior superior iliac spine, anterior inferior iliac spine, iliac crest, ischial tuberosity, and greater and lesser trochanters occur in patients nearing skeletal maturity.136,147 Forceful eccentric loading of the bonetendon-muscle unit can result in failure at the weakest link of this chain.51,147 In the adolescent, this weak point is the fusing physis as opposed to the adjacent tendon. If displaced or accompanied by apophyseal widening, these injuries are usually visible radiographically.70,136 Fractures that are nondisplaced, minimally displaced, or missed on radiographs are easily identified on MRI. These injuries demonstrate varying degrees of physeal widening, with exuberant surrounding soft tissue edema and marrow edema pattern within the apophysis and subjacent parent bone70 (Fig. 62-46). Laxity and redundancy of the tendon may also be present. These injuries are often encountered in the subacute healing phase, which includes callus formation and periosteal bone formation around the site of the injury.51 If encountered during an early stage of healing, immature periostitis and mineralized callus can mimic an aggressive osseous lesion.20,51 Location at the characteristic site of avulsion in a patient of appropriate age suggests the diagnosis. Short-interval follow-up imaging to confirm expected maturation is recommended in equivocal cases. In the adult, the myotendinous junction and to a lesser extent the tendinous attachment represents the weak point in the bone-tendonmuscle unit. Accordingly, myotendinous strain and tendon ruptures are more common in the adult population.119 Such injuries are well demonstrated on MRI and were discussed earlier (see “Snapping Tendon Syndromes and Other Symptomatic Tendon Abnormalities”). It is well known that avulsions in the skeletally mature patient without an appropriate history of trauma raise the specter of pathologic fracture. Avulsion fracture of the lesser trochanter in adults is

CHAPTER 62  Hip and Pelvis

A

B

C

D

2169

FIG 62-43  Femoral neck stress fracture in a 44-year-old woman runner with progressive right hip pain. A, Initial AP radiograph of the right hip demonstrates a subtle buckling of the medial cortex of the femoral neck (arrow), representing a stress fracture deformity. B, Follow-up AP radiographs performed 1 month later demonstrate the initial cortical contour defect (arrow) with an additional linear sclerotic band (arrowheads) that traverses the midcervical region of the femoral neck. Coronal T1-weighted (C) and STIR (D) images of the right hip obtained at the time of initial presentation show a low-signal linear fracture defect traversing the basicervical region of the right femoral neck (arrowheads), surrounded by marrow edema pattern (arrows).

highly correlated with the presence of an underlying neoplastic lesion even if no such lesion is identified on initial radiographs.47,55 Accordingly, radiographs showing this abnormality in a patient older than age 40 warrants additional imaging with MRI.

Soft Tissue Degloving Injuries Soft tissue degloving injuries are also known as Morel Lavallée lesions. These can occur in any region of the body following a shearing injury to the soft tissues but are most common about the hip and pelvis.66,99 This injury results in a separation of the subcutaneous fat from the deep fascia surrounding the muscular compartments.66 Initially, hemorrhage collects within the defect, resulting in a lenticular collection centered at the deep fascial margin. These collections, which are often chronic and resistant to treatment, evolve to contain varying degrees

of blood product, lymphatic fluid, and debris.66,115 Superinfection is possible. Soft tissue degloving lesions are well visualized on both CT and MRI regardless of their age and stage of evolution.66,115 Serial examination can be employed to assess for resolution following nonoperative or surgical intervention.

SUMMARY CT and MRI are important tools in the evaluation of patients with musculoskeletal hip and pelvis complaints. These techniques are useful in the diagnosis and classification of a variety of pathologic conditions. As is the case for almost all other parts of the musculoskeletal system, radiographs should be the initial (and in some cases only)

2170

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 62-44  Bisphosphonate proximal femur fracture in a 70-year-old woman on bisphosphonate therapy for osteoporosis who presented with progressive right thigh pain. A, AP radiograph of the right femur shows focal thickening of the lateral cortex of the proximal femoral shaft surrounding beaking along the periosteal (arrow) and endosteal surfaces of the lateral cortex. B, Coronal CT image through the same region demonstrates similar endosteal beaking and cortical thickening (arrow), as well as a transversely oriented cortical lucency that represents the atypical bisphosphonate-related fracture defect.

A

B

FIG 62-45  Coronal T1-weighted (A) and STIR (B) MRI of the right hip in a 63-year-old woman on long-term bisphosphonate therapy shows bilateral areas of lateral cortical thickening (arrows) in the proximal femoral shafts. There is subjacent marrow edema pattern (arrowheads) surrounding the symptomatic cortical defect.

CHAPTER 62  Hip and Pelvis

2171

A

B

C

FIG 62-46  Iliac apophyseal avulsion in a 14-year-old girl gymnast with left hip pain and point tenderness along the anterior aspect of the left iliac wing. A, AP radiograph of the pelvis demonstrates slight physeal contour irregularity adjacent to the anterior aspect of the left iliac apophysis (long arrow). Although left physeal width is symmetric compared to the right-sided counterpart, there is subtle callus formation more posteriorly (short arrow). Coronal T1-weighted (B) and STIR (C) images of the pelvis in the same patient show underlying avulsion fracture of the left iliac apophysis (arrows) surrounded by soft tissue edema, periosteal edema, and iliac marrow edema pattern. Fluid signal is evident within the fracture cleft.

imaging study to assess the patient with a painful pelvis or hip. In other cases, CT and/or MRI are the imaging techniques of choice to provide additional assessment and guide patient management.

REFERENCES 1. Ahumada LA, Ashruf S, Espinosa-De-Los-Monteros A, et al: Athletic pubalgia: Definition and surgical treatment. Ann Plast Surg 55:393–396, 2005. 2. Ali AM, Whitwell D, Ostlere SJ: Case report: Imaging and surgical treatment of a snapping hip due to ischiofemoral impingement. Skeletal Radiol 40:653–656, 2011. 3. Aliprandi A, Di Pietto F, Minafra P, et al: Femoro-acetabular impingement: What the general radiologist should know. La Radiol Med 119:103–112, 2014. 4. Amorosa LF, Kloen P, Helfet DL: High-energy pediatric pelvic and acetabular fractures. Orthop Clin North Am 45:483–500, 2014. 5. Anderson SE, Siebenrock KA, Tannast M: Femoroacetabular impingement. Eur J Radiol 81:3740–3744, 2012.

6. Angtuaco EJ, Fassas AB, Walker R, et al: Multiple myeloma: Clinical review and diagnostic imaging. Radiology 231:11–23, 2004. 7. Aspenberg P, Schilcher J: Atypical femoral fractures, bisphosphonates, and mechanical stress. Curr Osteoporos Rep 12:189–193, 2014. 8. Assoun J, Richardi G, Railhac JJ, et al: Osteoid osteoma: MR imaging versus CT. Radiology 191:217–223, 1994. 9. Atesok KI, Alman BA, Schemitsch EH, et al: Osteoid osteoma and osteoblastoma. J Am Acad Orthop Surg 19:678–689, 2011. 10. Atukorala I, Kwoh CK, Guermazi A, et al: Synovitis in knee osteoarthritis: A precursor of disease? Ann Rheum Dis 75:390–395, 2016. 11. Averill LW, Hernandez A, Gonzalez L, et al: Diagnosis of osteomyelitis in children: Utility of fat-suppressed contrast-enhanced MRI. AJR Am J Roentgenol 192:1232–1238, 2009. 12. Bancroft LW, Blankenbaker DG: Imaging of the tendons about the pelvis. AJR Am J Roentgenol 195:605–617, 2010. 13. Bano A, Karantanas A, Pasku D, et al: Persistent sciatica induced by quadratus femoris muscle tear and treated by surgical decompression: A case report. J Med Case Reports 4:236, 2010. 14. Bedi A, Kelly BT: Femoroacetabular impingement. J Bone Joint Surg Am 95:82–92, 2013.

2172

PART II  CT and MR Imaging of the Whole Body

15. Biko DM, Davidson R, Pena A, et al: Proximal focal femoral deficiency: Evaluation by MR imaging. Pediatr Radiol 42:50–56, 2012. 16. Black DM, Schwartz AV, Ensrud KE, et al: Effects of continuing or stopping alendronate after 5 years of treatment: The fracture intervention trial long-term extension (flex): A randomized trial. JAMA 296:2927–2938, 2006. 17. Blankenbaker DG, Davis KW, De Smet AA, et al: MRI appearance of the pectinofoveal fold. AJR Am J Roentgenol 192:93–95, 2009. 18. Blankenbaker DG, De Smet AA, Keene JS, et al: Imaging appearance of the normal and partially torn ligamentum teres on hip MR arthrography. AJR Am J Roentgenol 199:1093–1098, 2012. 19. Blankenbaker DG, Ullrick SR, Kijowski R, et al: MR arthrography of the hip: Comparison of ideal-SPGR volume sequence to standard MR sequences in the detection and grading of cartilage lesions. Radiology 261:863–871, 2011. 20. Bloem JL, Reidsma II: Bone and soft tissue tumors of hip and pelvis. Eur J Radiol 81:3793–3801, 2012. 21. Bluemke DA, Zerhouni EA: MRI of avascular necrosis of bone. Top Magn Reson Imaging 8:231–246, 1996. 22. Boutry N, Paul C, Leroy X, et al: Rapidly destructive osteoarthritis of the hip: MR imaging findings. AJR Am J Roentgenol 179:657–663, 2002. 23. Bravo SM, Winalski CS, Weissman BN: Pigmented villonodular synovitis. Radiol Clin North Am 34:311–326, 1996. 24. Burr DB, Forwood MR, Fyhrie DP, et al: Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J Bone Miner Res 12:6–15, 1997. 25. Byrd J, Jones KS, Maiers GP, II: Two to 10 years’ follow-up of arthroscopic management of pigmented villonodular synovitis in the hip: A case series. Arthroscopy 29:1783–1787, 2013. 26. Bywaters E, Dorling J: Amyloid deposits in articular cartilage. Ann Rheum Dis 29:294–306, 1970. 27. Cahir JG, Toms AP: Regional migratory osteoporosis. Eur J Radiol 67:2–10, 2008. 28. Cannon J, Silvestri S, Munro M: Imaging choices in occult hip fracture. J Emerg Med 37:144–152, 2009. 29. Chai JW, Hong SH, Choi JY, et al: Radiologic diagnosis of osteoid osteoma: From simple to challenging findings. Radiographics 30:737– 749, 2010. 30. Chan W, Lang P, Stevens M, et al: Osteoarthritis of the knee: Comparison of radiography, CT, and MR imaging to assess extent and severity. AJR Am J Roentgenol 157:799–806, 1991. 31. Chan SS, Rosenberg ZS, Chan K, et al: Subtrochanteric femoral fractures in patients receiving long-term alendronate therapy: Imaging features. AJR Am J Roentgenol 194:1581–1586, 2010. 32. Cheon S-H, Oh C-W, Lee J-Y, et al: Early diagnosis of impending femoral insufficiency fractures by use of MRI: Case report and review of the literature. J Orthop Sci 18:843–848, 2013. 33. Choi YS, Lee SM, Song BY, et al: Dynamic sonography of external snapping hip syndrome. J Ultrasound Med 21:753–758, 2002. 34. Cobby M, Adler R, Swartz R, et al: Dialysis-related amyloid arthropathy: MR findings in four patients. AJR Am J Roentgenol 157:1023–1027, 1991. 35. Cooper A, Evans O, Ali F, et al: A novel method for assessing postoperative femoral head reduction in developmental dysplasia of the hip. J Child Orthop 8:319–324, 2014. 36. Cosman F, De Beur S, Leboff M, et al: Clinician’s guide to prevention and treatment of osteoporosis. Osteoporos Int 25:2359–2381, 2014. 37. Dandachli W, Islam SU, Tippett R, et al: Analysis of acetabular version in the native hip: Comparison between 2D axial CT and 3D CT measurements. Skeletal Radiol 40:877–883, 2011. 38. De Smet AA, Blankenbaker DG, Alsheik NH, et al: MRI appearance of the proximal hamstring tendons in patients with and without symptomatic proximal hamstring tendinopathy. AJR Am J Roentgenol 198:418–422, 2012. 39. Della Torre P, Picuti G, Di Filippo P: Rapidly progressive osteoarthritis of the hip. Ital J Orthop Traumatol 13:187–200, 1987. 40. Diana Afonso P, Kosinski AS, Spritzer CE: Following unenhanced MRI assessment for local recurrence after surgical resection of mesenchymal

41.

42. 43.

44. 45. 46.

47. 48. 49. 50. 51. 52.

53.

54.

55.

56. 57. 58. 59.

60.

61. 62. 63.

64.

65.

soft tissue tumors, do additional gadolinium-enhanced images change reader confidence or diagnosis? Eur J Radiol 82:806–813, 2013. Dietrich TJ, Suter A, Pfirrmann CW, et al: Supraacetabular fossa (pseudodefect of acetabular cartilage): Frequency at MR arthrography and comparison of findings at MR arthrography and arthroscopy. Radiology 263:484–491, 2012. Dinan JJ: Two cases of primary osteomyelitis of the pubic bone. Can Med Assoc J 41:436–442, 1939. Dinauer PA, Murphy KP, Carroll JF: Sublabral sulcus at the posteroinferior acetabulum: A potential pitfall in MR arthrography diagnosis of acetabular labral tears. AJR Am J Roentgenol 183:1745– 1753, 2004. Dixon S, Newton J, Teh J: Stress fractures in the young athlete: A pictorial review. Curr Probl Diagn Radiol 40:29–44, 2011. Douis H, Saifuddin A: The imaging of cartilaginous bone tumours. II. Chondrosarcoma. Skeletal Radiol 42:611–626, 2013. Dubois DF, Omar IM: MR imaging of the hip: Normal anatomic variants and imaging pitfalls. Magn Reson Imaging Clin N Am 18:663–674, 2010. Duggleby L, Gupta R, Thomas D: A rare case of lesser trochanter fracture. Arch Dis Child 100:151, 2014. Ekstrand J, Torstveit M: Stress fractures in elite male football players. Scand J Med Sci Sports 22:341–346, 2012. Ellis AA, Trent RB: Hospitalized fall injuries and race in California. Inj Prev 7:316–320, 2001. Fautrel B, Fermand J-P, Sibilia J, et al: Amyloid arthropathy in the course of multiple myeloma. J Rheumatol 29:1473–1481, 2002. Fernbach S, Wilkinson R: Avulsion injuries of the pelvis and proximal femur. AJR Am J Roentgenol 137:581–584, 1981. Firooznia H, Rafii M, Golimbu C, et al: Computerized tomography in diagnosis of pelvic abscess in spinal-cord-injured patients. Comput Radiol 7:335–341, 1983. Flanum ME, Keene JS, Blankenbaker DG, et al: Arthroscopic treatment of the painful “internal” snapping hip results of a new endoscopic technique and imaging protocol. Am J Sports Med 35:770–779, 2007. Fox C, Hull R, Langdown A: 7. Synovial chondromatosis: A diagnosis to consider in persistent hip disease with normal plain x-rays. Rheumatology 53(Suppl 1):i59, 2014. doi: 10.193/rheumatology/ keu096.007. Fox TP, Lakkol S, Oliver G: Lesser trochanter fracture: The presenting feature of a more sinister pathology. BMJ Case Rep 2014:2014. doi: 10.1136/bcr-2013-202590. Freeman AK, Sumathi VP, Jeys L: Metastatic tumours of bone. Surgery (Oxford) 33:34–39, 2014. Fricker PA, Taunton JE, Ammann W: Osteitis pubis in athletes. Sports medicine 12:266–279, 1991. Frihagen F, Nordsletten L, Tariq R, et al: MRI diagnosis of occult hip fractures. Acta Orthop 76:524–530, 2005. Fukiage K, Yamasaki N, Suzuki S: Visualization of developmental dysplasia of the hip in infants by using three-dimensional magnetic resonance imaging. J Pediatr Orthop B 23:107–111, 2014. Garner HW, Kransdorf MJ, Peterson JJ: Posttherapy imaging of musculoskeletal neoplasms. Radiol Clin North Am 49:1307–1323, 2011. Georgiades CS, Neyman EG, Barish MA, et al: Amyloidosis: Review and CT manifestations. Radiographics 24:405–416, 2004. Goldman A, Schneider R, Wilson P, Jr: Proximal focal femoral deficiency. Can Assoc Radiol J 29:101–107, 1978. Goto T, Tamaki Y, Hamada D, et al: Computed tomography analysis of herniation pits in the general population in Japan: Association with morphological indicators of femoroacetabular impingement. Bone Joint J 95:2013. Graham CW, Dmochowski RR, Faerber GJ, et al: Pubic osteomyelitis following bladder neck surgery using bone anchors: A report of 9 cases. J Urol 168:2055–2057, 2002. Greenspan SL, Myers ER, Maitland LA, et al: Trochanteric bone mineral density is associated with type of hip fracture in the elderly. J Bone Miner Res 9:1889–1894, 1994.

CHAPTER 62  Hip and Pelvis 66. Hak DJ, Olson SA, Matta JM: Diagnosis and management of closed internal degloving injuries associated with pelvic and acetabular fractures: The Morel-Lavallee lesion. J Trauma Injury Infect Crit Care 42:1046–1051, 1997. 67. Harris JH, Jr, Coupe KJ, Lee JS, et al: Acetabular fractures revisited: Part 2, a new CT-based classification. AJR Am J Roentgenol 182:1367–1375, 2004. 68. Hastings D, Parker S: Protrusio acetabuli in rheumatoid arthritis. Clin Orthop Relat Res 108:76–83, 1975. 69. Hauptfleisch J, Meagher TM, Hughes RJ, et al: Interobserver agreement of magnetic resonance imaging signs of osteomyelitis in pelvic pressure ulcers in patients with spinal cord injury. Arch Phys Med Rehabil 94:1107–1111, 2013. 70. Hébert KJ, Laor T, Divine JG, et al: MRI appearance of chronic stress injury of the iliac crest apophysis in adolescent athletes. AJR Am J Roentgenol 190:1487–1491, 2008. 71. Hiti CJ, Stevens KJ, Jamati MK, et al: Athletic osteitis pubis. Sports Med 41:361–376, 2011. 72. Hodgson R, O’Connor P, Moots R: MRI of rheumatoid arthritis— Image quantitation for the assessment of disease activity, progression and response to therapy. Rheumatology 47:13–21, 2008. 73. Hodler J, Yu JS, Goodwin D, et al: MR arthrography of the hip: Improved imaging of the acetabular labrum with histologic correlation in cadavers. AJR Am J Roentgenol 165:887–891, 1995. 74. Horger M, Weisel K, Bares R, et al: Modern imaging techniques during therapy in patients with multiple myeloma. Acta Radiol 52:881–888, 2011. 75. Hosalkar HS, Garg S, Moroz L, et al: The diagnostic accuracy of MRI versus CT imaging for osteoid osteoma in children. Clin Orthop Relat Res 171–177, 2005. 76. Huang AB, Schweitzer ME, Hume E, et al: Osteomyelitis of the pelvis/ hips in paralyzed patients: Accuracy and clinical utility of MRI.   J Comput Assist Tomogr 22:437–443, 1998. 77. Ikemura S, Yamamoto T, Motomura G, et al: MRI evaluation   of collapsed femoral heads in patients 60 years old or older: Differentiation of subchondral insufficiency fracture from   osteonecrosis of the femoral head. AJR Am J Roentgenol 195:W63–W68, 2010. 78. Ikemura S, Yamamoto T, Motomura G, et al: The utility of clinical features for distinguishing subchondral insufficiency fracture from osteonecrosis of the femoral head. Arch Orthop Trauma Surg 133:1623– 1627, 2013. 79. Ilizaliturri VM, Jr, Camacho-Galindo J: Endoscopic treatment of snapping hips, iliotibial band, and iliopsoas tendon. Sports Med Arthrosc 18:2010. 80. Iwasaki K, Yamamoto T, Motomura G, et al: Subchondral insufficiency fracture of the femoral head in young adults. Clin Imaging 35:208–213, 2011. 81. Jacobsen S, Romer L, Soballe K: Degeneration in dysplastic hips. A computer tomography study. Skeletal Radiol 34:778–784, 2005. 82. Jelinek J, Kransdorf M, Utz J, et al: Imaging of pigmented villonodular synovitis with emphasis on MR imaging. AJR Am J Roentgenol 152:337–342, 1989. 83. Jia J, Li L, Zhang L, et al: Three dimensional-CT evaluation of femoral neck anteversion, acetabular anteversion and combined anteversion in unilateral DDH in an early walking age group. Int Orthop 36:119–124, 2012. 84. Johnson KA: Impingement of the lesser trochanter on the ischial ramus after total hip arthroplasty. Report of three cases. J Bone Joint Surg Am 59:268–269, 1977. 85. Kachingwe AF, Grech S: Proposed algorithm for the management of athletes with athletic pubalgia (sports hernia): A case series. J Orthop Sports Phys Ther 38:768–781, 2008. 86. Karchevsky M, Schweitzer ME, Morrison WB, et al: MRI findings of septic arthritis and associated osteomyelitis in adults. AJR Am J Roentgenol 182:119–122, 2004. 87. Keene JS: Coxa saltans: Iliopsoas snapping and tendonitis. Hip arthroscopy and hip joint preservation surgery. 2014.

2173

88. Kim Y-M, Oh H, Kim H: The pattern of bone marrow oedema on MRI in osteonecrosis of the femoral head. J Bone Joint Surg Br 82:837–841, 2000. 89. Kim JA, Park JS, Jin W, et al: Herniation pits in the femoral neck: A radiographic indicator of femoroacetabular impingement? Skeletal Radiol 40:167–172, 2011. 90. Kollender Y, Bender B, Nirkin A, et al: Intra-articular yttrium 90 injection as an adjuvant following resection of diffused pigmented villonodular synovitis. J Bone Joint Surg Br 87:2005. 91. Kong A, Van Der Vliet A, Zadow S: MRI and US of gluteal tendinopathy in greater trochanteric pain syndrome. Eur Radiol 17:1772–1783, 2007. 92. Kransdorf MJ, Murphey MD, Smith SE: Imaging of soft tissue neoplasms in the adult: Benign tumors. Semin Musculoskelet Radiol 3:21–38, 1999. 93. Kransdorf MJ, Stull MA, Gilkey FW, et al: Osteoid osteoma. Radiographics 11:671–696, 1991. 94. Lane JM: Bisphosphonate use for >/=5 years increased risk for subtrochanteric or femoral shaft fractures. J Bone Joint Surg Am 93:2011. 95. Lawrence DA, Menn K, Baumgaertner M, et al: Acetabular fractures: Anatomic and clinical considerations. AJR Am J Roentgenol 201:W425– W436, 2013. 96. Learch TJ, Farooki S: Magnetic resonance imaging of septic arthritis. Clin Imaging 24:236–242, 2000. 97. Lecouvet FE, Vande Berg BC, Malghem J, et al: MR imaging of the acetabular labrum: Variations in 200 asymptomatic hips. AJR Am J Roentgenol 167:1025–1028, 1996. 98. Letournel E: Acetabulum fractures: Classification and management. Clin Orthop Relat Res 151:81–106, 1980. 99. Letournel É, Judet R, Elson RA: Fractures of the acetabulum. 1993. 100. Li KC, Higgs J, Aisen AM, et al: MRI in osteoarthritis of the hip: Gradations of severity. Magn Reson Imaging 6:229–236, 1988. 101. Liong S, Whitehouse RW: Lower extremity and pelvic stress fractures in athletes. Br J Radiol 85:1148–1156, 2012. 102. Littrell LA, Wenger DE, Wold LE, et al: Radiographic, CT, and MR imaging features of dedifferentiated chondrosarcomas: A retrospective review of 174 de novo cases. Radiographics 24:1397–1409, 2004. 103. Liu PT, Chivers FS, Roberts CC, et al: Imaging of osteoid osteoma with dynamic gadolinium-enhanced MR imaging. Radiology 227:691–700, 2003. 104. Lopez M, Sauerbrei E: Septic arthritis of the hip joint: Sonographic and CT findings. Can Assoc Radiol J 36:322–324, 1985. 105. Lovell G, Galloway H, Hopkins W, et al: Osteitis pubis and assessment of bone marrow edema at the pubic symphysis with MRI in an elite junior male soccer squad. Clin J Sport Med 16:117–122, 2006. 106. Maldjian C, Patel T, Klein R, et al: Efficacy of MRI in classifying proximal focal femoral deficiency. Skeletal Radiol 36:215–220, 2007. 107. Mamisch TC, Zilkens C, Siebenrock KA, et al: MRI of hip osteoarthritis and implications for surgery. Radiol Clin North Am 47:713–722, 2009. 108. Martinez AE, Li SM, Ganz R, et al: Os acetabuli in femoro-acetabular impingement: Stress fracture or unfused secondary ossification centre of the acetabular rim? Hip Int 16:281–286, 2006. 109. McQueen F: Magnetic resonance imaging in early inflammatory arthritis: What is its role? Rheumatology 39:700–706, 2000. 110. Meyers WC, Yoo E, Devon ON, et al: Understanding “sports hernia”(athletic pubalgia): The anatomic and pathophysiologic basis   for abdominal and groin pain in athletes. Oper Tech Sports Med 20:2012. 111. Mootha AK, Saini R, Dhillon MS, et al: MRI evaluation of femoral   and acetabular anteversion in developmental dysplasia of the hip.   A study in an early walking age group. Acta Orthop Belg 76:174–180, 2010. 112. Murphey MD, Kransdorf MJ, Smith SE: Imaging of soft tissue neoplasms in the adult: Malignant tumors. Semin Musculoskelet Radiol 3:39–58, 1999. 113. Murphey MD, Smith WS, Smith SE, et al: From the archives of the AFIP. Imaging of musculoskeletal neurogenic tumors: Radiologicpathologic correlation. Radiographics 19:1253–1280, 1999.

2174

PART II  CT and MR Imaging of the Whole Body

114. Murphey MD, Walker EA, Wilson AJ, et al: From the archives of the AFIP: Imaging of primary chondrosarcoma: Radiologic-pathologic correlation. Radiographics 23:1245–1278, 2003. 115. Nair AV, Nazar P, Sekhar R, et al: Morel-lavallée lesion: A closed degloving injury that requires real attention. Indian J Radiol Imaging 24:288–290, 2014. 116. Nguyen MS, Kheyfits V, Giordano BD, et al: Hip anatomic variants that may mimic pathologic entities on MRI: Nonlabral variants. AJR Am J Roentgenol 201:W401–W408, 2013. 117. Nishida Y, Tsukushi S, Nakashima H, et al: Osteochondral destruction in pigmented villonodular synovitis during the clinical course. J Rheumatol 39:345–351, 2012. 118. O’Brien SD, Bui-Mansfield LT: MRI of quadratus femoris muscle tear: Another cause of hip pain. AJR Am J Roentgenol 189:1185–1189, 2007. 119. Ohishi T, Ito T, Suzuki D, et al: Occult hip and pelvic fractures and accompanying muscle injuries around the hip. Arch Orthop Trauma Surg 132:105–112, 2012. 120. Omar IM, Zoga AC, Kavanagh EC, et al: Athletic pubalgia and “sports hernia”: Optimal MR imaging technique and findings. Radiographics 28:1415–1438, 2008. 121. Østergaard M, Peterfy C, Conaghan P, et al: Omeract rheumatoid arthritis magnetic resonance imaging studies. Core set of MRI acquisitions, joint pathology definitions, and the omeract ra-MRI scoring system. J Rheumatol 30:1385–1386, 2003. 122. Ottaviani S, Ayral X, Dougados M, et al: Pigmented villonodular synovitis: A retrospective single-center study of 122 cases and review of the literature. Semin Arthritis Rheum 40:539–546, 2011. 123. Paajanen H, Hermunen H, Karonen J: Pubic magnetic resonance imaging findings in surgically and conservatively treated athletes with osteitis pubis compared to asymptomatic athletes during heavy training. Am J Sports Med 36:117–121, 2008. 124. Pandey R, McNally E, Ali A, et al: The role of MRI in the diagnosis of occult hip fractures. Injury 29:61–63, 1998. 125. Park-Wyllie LY, Mamdani MM, Juurlink DN, et al: Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. JAMA 305:783–789, 2011. 126. Patti JW, Ouellette H, Bredella MA, et al: Impingement of lesser trochanter on ischium as a potential cause for hip pain. Skeletal Radiol 37:939–941, 2008. 127. Pfirrmann CW, Chung CB, Theumann NH, et al: Greater trochanter of the hip: Attachment of the abductor mechanism and a complex of three bursae—MR imaging and MR bursography in cadavers and   MR imaging in asymptomatic volunteers. Radiology 221:469–477, 2001. 128. Philippon MJ, Wolff AB, Briggs KK, et al: Acetabular rim reduction for the treatment of femoroacetabular impingement correlates with preoperative and postoperative center-edge angle. Arthroscopy 26:757–761, 2010. 129. Pitt M, Graham A, Shipman J, et al: Herniation pit of the femoral neck. AJR Am J Roentgenol 138:1115–1121, 1982. 130. Porrino JA, Jr, Kohl CA, Taljanovic M, et al: Diagnosis of proximal femoral insufficiency fractures in patients receiving bisphosphonate therapy. AJR Am J Roentgenol 194:1061–1064, 2010. 131. Radke S, Kenn W, Eulert J: Transient bone marrow edema syndrome progressing to avascular necrosis of the hip–A case report and review of the literature. Clin Rheumatol 23:83–88, 2004. 132. Ravi V, Wang W-L, Lewis VO: Treatment of tenosynovial giant cell tumor and pigmented villonodular synovitis. Curr Opin Oncol 23:361–366, 2011. 133. Rennert R, Golinko M, Yan A, et al: Developing and evaluating outcomes of an evidence-based protocol for the treatment of osteomyelitis in stage iv pressure ulcers: A literature and wound electronic medical record database review. Ostomy Wound Manage 55:42–53, 2009. 134. Rodan GA, Fleisch H: Bisphosphonates. Mechanisms of action. Clin Invest 97:2692–2696, 1996. 135. Roemer FW, Guermazi A: Osteoarthritis year in review 2014: Imaging. Osteoarthritis Cartilage 22:2003–2012, 2014.

136. Rossi F, Dragoni S: Acute avulsion fractures of the pelvis in adolescent competitive athletes: Prevalence, location and sports distribution of 203 cases collected. Skeletal Radiol 30:127–131, 2001. 137. Samelson EJ, Zhang Y, Kiel DP, et al: Effect of birth cohort on risk of hip fracture: Age-specific incidence rates in the Framingham study. Am J Public Health 92:858–862, 2002. 138. Schagemann JC, Hunold P, Russlies M, et al: Synovial chondromatosis of the hip with atypical MRI morphology and mistakable clinical symptoms–A case report. Acta Orthop 82:246–249, 2011. 139. Sohaib S, Cook G, Allen S, et al: Comparison of whole-body MRI and bone scintigraphy in the detection of bone metastases in renal cancer. Br J Radiol 82:632–639, 2009. 140. Soldatos T, McCarthy EF, Attar S, et al: Imaging features of chondrosarcoma. J Comput Assist Tomogr 35:504–511, 2011. 141. Starr V, Ha BY: Imaging update on developmental dysplasia of the hip with the role of MRI. AJR Am J Roentgenol 203:1324–1335, 2014. 142. Steinbach L, Schneider R, Goldman A, et al: Bursae and abscess cavities communicating with the hip. Diagnosis using arthrography and CT. Radiology 156:303–307, 1985. 143. Stelzeneder D, Mamisch T, Kress I, et al: Patterns of joint damage seen on MRI in early hip osteoarthritis due to structural hip deformities. Osteoarthritis Cartilage 20:661–669, 2012. 144. Strauss EJ, Nho SJ, Kelly BT: Greater trochanteric pain syndrome. Sports Med Arthrosc 18:113–119, 2010. 145. Studler U, Kalberer F, Leunig M, et al: MR arthrography of the hip: Differentiation between an anterior sublabral recess as a normal variant and a labral tear. Radiology 249:2008. 146. Subhawong TK, Jacobs MA, Fayad LM: Insights into quantitative diffusion-weighted MRI for musculoskeletal tumor imaging. AJR Am J Roentgenol 203:560–572, 2014. 147. Sundar M, Carty H: Avulsion fractures of the pelvis in children: A report of 32 fractures and their outcome. Skeletal Radiol 23:85–90, 1994. 148. Sutter R, Pfirrmann CW: Atypical hip impingement. AJR Am J Roentgenol 201:W437–W442, 2013. 149. Taneja AK, Bredella MA, Torriani M: Ischiofemoral impingement. Magn Reson Imaging Clin N Am 21:65–73, 2013. 150. Tannast M, Siebenrock KA, Anderson SE: Femoroacetabular impingement: Radiographic diagnosis—What the radiologist should know. AJR Am J Roentgenol 188:1540–1552, 2007. 151. Thomas R, Wood A, Watson J, et al: Delay in diagnosis of neck of femur stress fracture in a female military recruit. J R Nav Med Serv 98:27–29, 2012. 152. Toms AP, Marshall TJ, Becker E, et al: Regional migratory osteoporosis: A review illustrated by five cases. Clin Radiol 60:425–438, 2005. 153. Tonnis D, Heinecke A: Acetabular and femoral anteversion: Relationship with osteoarthritis of the hip. J Bone Joint Surg Am 81:1747–1770, 1999. 154. Torriani M, Souto SC, Thomas BJ, et al: Ischiofemoral impingement syndrome: An entity with hip pain and abnormalities of the quadratus femoris muscle. AJR Am J Roentgenol 193:186–190, 2009. 155. Tosun O, Algin O, Yalcin N, et al: Ischiofemoral impingement: Evaluation with new MRI parameters and assessment of their reliability. Skeletal Radiol 41:575–587, 2012. 156. Unger E, Moldofsky P, Gatenby R, et al: Diagnosis of osteomyelitis by MR imaging. AJR Am J Roentgenol 150:605–610, 1988. 157. Urban BA, Fishman EK, Goldman S, et al: CT evaluation of amyloidosis: Spectrum of disease. Radiographics 13:1295–1308, 1993. 158. Vande Berg BE, Malghem JJ, Labaisse MA, et al: MR imaging of avascular necrosis and transient marrow edema of the femoral head. Radiographics 13:501–520, 1993. 159. Watts A, Teoh K, Beggs I, et al: Routine MRI surveillance following definitive treatment of bone and soft tissue sarcoma. J Bone Joint Surg Br 94:2012. 160. Wilson AJ, Murphy WA, Hardy DC, et al: Transient osteoporosis: Transient bone marrow edema? Radiology 167:757–760, 1988. 161. Yamamoto T: Subchondral insufficiency fractures of the femoral head. Clin Orthop Surg 4:173–180, 2012.

CHAPTER 62  Hip and Pelvis 162. Yamamoto T, Karasuyama K, Iwasaki K, et al: Subchondral insufficiency fracture of the femoral head in males. Arch Orthop Trauma Surg 134:1199–1203, 2014. 163. Yang H-L, Liu T, Wang X-M, et al: Diagnosis of bone metastases: A meta-analysis comparing 18FDG pet, CT, MRI and bone scintigraphy. Eur Radiol 21:2604–2617, 2011. 164. Zander G: “Os acetabuli” and other bone nuclei; periarticular calcifications at the hip-joint. Acta Radiol [Old Series] 24:1943. 165. Ziegert AJ, Blankenbaker DG, De Smet AA, et al: Comparison of standard hip MR arthrographic imaging planes and sequences for

2175

detection of arthroscopically proven labral tear. AJR Am J Roentgenol 192:1397–1400, 2009. 166. Zini R, Longo UG, De Benedetto M, et al: Arthroscopic management of primary synovial chondromatosis of the hip. Arthroscopy 29:420–426, 2013. 167. Zoga AC, Kavanagh EC, Omar IM, et al: Athletic pubalgia and the “sports hernia”: MR imaging findings. Radiology 247:797–807, 2008.

63  Knee Romulo Baltazar, Calvin T. Ma, Steven Shankman, Nicholas Bhojwani, and Javier Beltran

IMAGING TECHNIQUE Magnetic resonance imaging (MRI) is commonly used to evaluate important structures of the knee not adequately depicted on radiographs or computed tomography (CT). Radiography is the first-line imaging modality performed for most suspected abnormalities of the knee in the setting of trauma, degeneration, inflammatory arthritis, or suspected bone lesion. It will often suffice in diagnosing or excluding an abnormality. CT depicts to better advantage the detailed bony trabecular and cortical anatomy and can be helpful in the detection of nondisplaced fractures and in the characterization of bone lesions. MRI offers the advantage of inherent superior contrast resolution as compared to radiographs, and CT and is better suited for assessing soft tissue, cartilage, and marrow pathology. Common indications for performing MRI include but are not limited to the evaluation of meniscal tears, ligament sprains, cartilage lesions, marrow abnormalities, muscle strains, tendinopathy, neoplasms, and infections. MRI of the knee is typically performed with a dedicated extremity or knee coil to maximize the signal-to-noise ratio (SNR). The newer multichannel coils containing 8 or more coil elements will offer better SNR. Imaging is done with the knee placed in full extension in a neutral position. A 14- to 16-cm field of view (FOV) and a 3- to 4-mm slice thickness are generally recommended to allow adequate evaluation of the menisci and cartilage. An interslice gap of no more than 33% of slice thickness can help decrease signal loss due to crosstalk.73 Imaging matrix should be at least 196 steps in the phase-encoding direction and 256 steps in the frequency-encoding direction for twodimensional (2D) imaging. MRI is routinely obtained in the sagittal, coronal, and axial planes (Figs. 63-1 to 63-3). The sagittal images are oriented along the longitudinal axis of the lateral femoral condyle to obtain a good view of the anterior cruciate ligament (ACL). The images should cover an FOV extending from the suprapatellar region, including the distal portion of the quadriceps tendon, to the proximal tibia, including the patella tendon insertion on the tibial tuberosity. An extended FOV may be necessary to assess patients with suspected quadriceps tendon tears or tears involving the distal insertion of the medial collateral ligament. A wide variety of pulse sequences and imaging protocols are used in knee MRI, with variations often due to individual preference and available equipment. Most of the sequences are based on conventional spin echo (CSE) or fast-spin echo (FSE) techniques. Imaging of the menisci is most frequently done with sagittal proton density (PD)weighted images using either CSE or FSE techniques because PD is highly sensitive and accurate in the detection of meniscal tears. The sagittal plane is the primary workhorse sequence for the menisci, with coronal and axial planes providing additional views 

2176

of the menisci (Figs. 63-4 to 63-6). A second sagittal sequence, typically fat-saturated PD FSE or T2 FSE, is used to evaluate primarily the  cruciate ligaments and other structures of the knee such as tendons, bone marrow, joint fluid, and trochlear articular cartilage. Coronal T1-weighted images allow for evaluation of marrow pathology,  including subtle linear fracture lines that may be obscured by bone marrow edema on T2-weighted images. Coronal fat-saturated PD- or T2-weighted images allow for evaluation of the meniscal body and root attachments, collateral ligaments, and weightbearing surfaces of the articular cartilage of the medial and lateral knee compartments. Axial fat-saturated PD or T2 images provide secondary assessment of the menisci and tendons and primary assessment of the patellar cartilage. Oblique coronal and oblique sagittal images have been shown to improve accuracy in the diagnosis of ACL injury.75 The value of FSE pulse sequences compared with CSE in the detection of meniscal tears has been discussed extensively in the radiology literature.2,19,43,59,76,116,117,149 The sensitivity of FSE imaging described in many reports is approximately 80%, whereas CSE imaging has a reported sensitivity in the range of 90% to 95%. The decreased sensitivity of FSE sequences has been attributed to the blurring artifact inherent in FSE sequences, resulting from the use of short effective echo time (TE). However, with advances in pulse sequence technology, optimized FSE imaging may be equally effective and allow for shorter acquisition time compared with CSE.19,43,51,86,106 Different forms of fat suppression are frequently used for evaluation of musculoskeletal lesions.35 In the knee, the main value of fat suppression has been to study bone contusions and cartilage lesions77,142; for this purpose, fat suppression is often added to the array of pulse sequences commonly used for general evaluation of the knee. For assessment of meniscal lesions, fat-suppression techniques are used mostly in combination with direct or indirect MR arthrography to enhance the contrast between the injected gadolinium and the adjacent pericapsular fat.

MENISCI The menisci provide important structural and functional support to the knee. They provide joint stability by promoting congruity between femoral and tibial articular surfaces, absorbing shock to the articular cartilage during axial loading, lubricating joint structures, and limiting hyperextension and hyperflexion of the joint.107 The absence of normal menisci accelerates degenerative changes of the joint.

Normal Anatomy The menisci of the knee are two C-shaped fibrocartilaginous disks with concave superior surfaces that articulate with the convex femoral condyles and flat inferior surfaces that set on the peripheral margins of

CHAPTER 63  Knee

Vastus lateralis muscle

Vastus medialis muscle

Iliotibial band

Anterior cruciate ligament Lateral meniscus

Posterior cruciate ligament Medial collateral ligament Medial meniscus

A Vastus lateralis muscle

Lateral head of gastrocnemius muscle Biceps femoris muscle Lateral condyle of femur Posterior horn of lateral meniscus Lateral collateral ligament

Vastus medialis muscle

Medial head of gastrocnemius muscle

Medial condyle of femur Posterior cruciate ligament Posterior horn of medial meniscus

Popliteus muscle

B FIG 63-1  Normal knee anatomy, coronal views. A, Coronal PD MRI through the anterior knee joint. B, Coronal PD MRI through the posterior knee joint.

2177

2178

PART II  CT and MR Imaging of the Whole Body

Vastus medialis Semimembranosus muscle

Semitendinosus tendon Semimembranosus tendon Anterior horn of medial meniscus

Posterior horn of medial meniscus Medial head of gastrocnemius muscle

A

Quadriceps tendon

Popliteal vessels Tibial nerve

Anterior cruciate ligament Ligament of Wrisberg Transverse ligament Patellar tendon

Lateral head of gastrocnemius Posterior cruciate ligament

B FIG 63-2  Normal knee anatomy, sagittal views. A, PD fat-saturated MRI through the medial knee. B, PD MRI through the intercondylar notch.

CHAPTER 63  Knee

2179

Lateral condyle of femur Hyaline articular cartilage

Anterior horn of lateral meniscus

Posterior horn of lateral meniscus

Hoffa's fat pad Lateral tibial plateau Popliteus tendon

C FIG 63-2, cont’d  C, PD MRI through the lateral knee.

the tibial plateau (see Fig. 63-2). They are composed of collagen fibers that are primarily arranged circumferentially along the long axis of the meniscus, with fibers extending radially between the circumferential fibers. They are thick peripherally and thin centrally and have sharp inner free edges. Up to one third of the peripheral meniscus is vascularized and innervated (the “red zone”), whereas the remaining inner two thirds or more is strictly fibrocartilaginous (the “white zone”).58,107 The menisci are arbitrarily divided into thirds: anterior horn, body, and posterior horn. The medial meniscus is semilunar (like a half-circle) and is typically larger than the lateral meniscus. The width of the medial meniscus gradually tapers from posterior to anterior, resulting in a wider posterior horn that is almost twice the size of the anterior horn. The medial meniscus has a more open C-shaped configuration with wider central attachment sites compared with the lateral meniscus (see Fig. 63-6A). The anterior horn attaches to the tibial plateau anterior to the anterior horn of the lateral meniscus and anterior to the origin of the ACL. The posterior horn attaches to the tibial plateau immediately anterior to the posterior cruciate ligament (PCL) and just posterior to the posterior lateral meniscal attachment. The peripheral margin of the medial meniscus is firmly attached to the joint capsule and to the tibial plateau itself, the latter via the coronary ligament. The lateral meniscus is semicircular with a smaller radius and narrower central attachment sites (see Fig. 63-6B). Unlike the medial meniscus, it has the same width throughout, with the anterior and posterior horns almost equal in size. The anterior horn of the lateral meniscus attaches immediately lateral to the origin of the ACL.137 The posterior horn attaches anterior to the PCL and anterior to the root  of the posterior horn of the medial meniscus. The posterior horn is

also connected to the medial femoral condyle via meniscofemoral  ligaments (see Fig. 63-2C) and to the popliteus via popliteomeniscal fascicles. The lateral meniscus has a loose peripheral attachment to  the joint capsule and is normally more mobile than the medial meniscus. The menisci are connected by intermeniscal ligaments, most commonly the anterior transverse meniscal ligament, which links the anterior horns of the menisci (see Fig. 63-2B). This ligament arises at the most anterosuperior portion of the lateral meniscus, crosses in front of the ACL tibial attachment, and merges with the posterosuperior portion of the medial meniscus. Much less commonly seen are the oblique meniscomeniscal ligaments and the posterior transverse meniscal ligament. The oblique meniscomeniscal ligaments (medial and lateral) pass from the anterior horn of one meniscus to the  posterior horn of the opposite meniscus and course between the  ACL and PCL.119 The posterior transverse meniscal ligament con­ nects the posterior horns of the menisci and courses anterior to  the PCL.110 The meniscofemoral ligaments are inconsistently seen fibrous bands commonly arising from the posterior horn lateral meniscus and attaching to the lateral aspect of the medial femoral condyle. The meniscofemoral ligament that courses anterior to the PCL is known as the ligament of Humphrey, and that which courses posterior to the PCL is the ligament of Wrisberg (see Fig. 63-2B). At least one of the meniscofemoral ligaments can be seen on MRI more than two thirds of the time, and about 30% of the time both can be seen.54,78 Rarely anteromedial meniscofemoral ligaments have been reported as arising from the anterior horn of the medial meniscus and attaching to the posterolateral wall of the intercondylar fossa.132

2180

PART II  CT and MR Imaging of the Whole Body

Quadriceps tendon

Anterior cruciate ligament, posterolateral bundle Anterior cruciate ligament, anteromedial bundle Posterior cruciate ligament

Iliotibial band Biceps femoris muscle Popliteal artery and vein

A

Common peroneal nerve Lateral head of gastrocnemius muscle

Lateral articular facet of patella Hyaline articular cartilage

Medial collateral ligament Sartorius muscle Medial head of gastrocnemius muscle Gracilis tendon Semitendinosus tendon

Medial patellar retinaculum Medial articular facet of patella Trochlea

Iliotibial band

Biceps femoris muscle Popliteal artery and vein

B

Medial head of the gastrocnemius muscle Sartorius muscle Semimembranosus muscle Gracilis tendon Semitendinosus tendon

FIG 63-3  Normal knee anatomy, axial views. A, PD axial MRI through the level of the ACL origin. B, PD fat-saturated MRI through the level of the patella.

CHAPTER 63  Knee

A

B

C

D

E

F

2181

FIG 63-4  Normal sagittal sequence. The anterior and posterior horns of both menisci are best seen on sagittal views, as are the ACL and PCL. A-C, Note the “bowtie” configuration of the lateral meniscus owing to its smaller radius of curvature. The anterior and posterior horns of the lateral meniscus are similar in size in a normal meniscus. D-F, In the medial meniscus, note the posterior horn is larger than the anterior horn. The posterior horn of the medial meniscus should never be the same size or smaller than the anterior horn in a normal meniscus.

Imaging Anatomy The sagittal plane is the most important plane in assessing the menisci, with the coronal plane providing supportive rather than new information and the axial plane increasing the accuracy of the sagittal and coronal planes when combined.85,110,140 In general, sagittal images optimally show the anterior and posterior horns of the medial and lateral menisci, coronal images help in evaluating the meniscal bodies, and thin-section axial images provide an additional view of the menisci and their free edges (see Fig. 63-1A and B). On sagittal images, the anterior and posterior horns of the menisci appear as isosceles triangles. The posterior horn of the medial meniscus is almost twice the size of the anterior horn. The anterior and posterior horns of the lateral meniscus are more similar in size, with the posterior horn slightly larger than the anterior horn. The posterior horn of either meniscus should never appear smaller than the anterior horn. The more peripheral sagittal images show the bodies of the medial and lateral menisci, although not optimally. On both sides, the menisci appear as flat bands. On the lateral side, the more central slices take on a “bowtie” configuration because of the smaller radius of the curvature (see Fig. 63-2C). Coronal images at the midportion of the knee produce the best images of the bodies of both menisci (see Fig. 63-5D). They appear

triangular and slightly larger laterally than medially. The capsular attachment on the medial side is incorporated into the tibial or medial collateral ligament. A small amount of fat may be interposed between the body of the medial meniscus and the capsule. On posterior coronal cross-sections, the posterior horns appear as flat bands (see Fig. 63-1B). On sagittal cross-sections, the popliteal tendon courses upward and laterally at 45 degrees (Fig. 63-7; also see Fig. 63-2C). Synovium extends superiorly and inferiorly about the tendon through the opening in the capsule. It appears on both sagittal and coronal images as linear medium signal intensity on T1- and PD-weighted images and as high signal intensity on T2-weighted images. More anteriorly, coronal images show the anterior horn of the lateral meniscus as a bandlike structure. The anterior horn of the medial meniscus is small and extends more anteriorly than that of   the lateral meniscus. The medial and lateral menisci, transverse ligament, and meniscofemoral ligaments appear homogeneously dark on all pulse sequences. The vascular and avascular zones cannot be distinguished on MRI.58 Hyperintensity between the meniscus and capsule caused by fat and vascularized connective tissue should not be mistaken for the vascularized zone of the meniscus. The vascularized zone also does not demonstrate enhancement with intravenous gadolinium. With increasing

2182

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 63-5  Normal coronal sequence. A-D, Coronal images beginning from posterior knee. The bodies of the menisci are seen best on coronal views, as are the root attachments of the menisci.

A

B

FIG 63-6  Normal axial sequence. A and B, Axial images through the joint line provide good visualization of the free edges of the medial and lateral menisci.

CHAPTER 63  Knee

2183

presence does not increase the prevalence of meniscal tears. Because meniscal tears may also assume a wavy appearance, a flounce should be considered only in the absence of altered signal intensity or morphology indicative of a tear.152 On coronal images, the meniscal flounce can appear truncated and simulate a tear or degeneration. The meniscal ossicle is a rare lesion found in young individuals and is thought to be either developmental or posttraumatic.11,89,111 It is usually seen in the posterior horn of the medial meniscus near the tibial attachment and is associated with meniscal tears. It can be asymptomatic or associated with pain and a sensation of joint locking; it may clinically simulate a torn meniscus. Because they may be calcified or ossified, meniscal ossicles can be mistaken on radiographs for loose bodies. On MRI the meniscal ossicle follows the signal intensity of the bone marrow, with high signal intensity centrally on T1-weighted images surrounded by a low-signal-intensity margin representing the cortex.121

Trauma

Incidence.  Meniscal injuries can result from either excessive mechan-

FIG 63-7  Popliteus tendon, coronal view. Coronal fat-saturated T2-weighted image, far posteriorly, shows the popliteus tendon (arrow) crossing posterior to the posterior horn lateral meniscus.

patient age, degeneration leads to indistinct focal areas of medium to high signal intensity in the menisci and must not be confused with a meniscal tear.

Normal Variants The most frequent normal anatomic meniscal variant is the discoid meniscus, which refers to a meniscus that is not C shaped but disklike, covering most of the tibial plateau to varying degrees rather than just covering its periphery. The discoid meniscus is almost always seen in the lateral meniscus. Discoid meniscus may be incomplete (most common) or complete, which extends to the intercondylar notch (Fig. 63-8A). Discoid meniscus varies from mild hypertrophy to a bulky slab of fibrocartilage (see Fig. 63-8C and D). It is prone to tearing and usually seen in children and adolescents, in whom it may be asymptomatic and noted incidentally. The discoid meniscus is believed to be more likely developmental than congenital, because the fetal meniscus never assumes such a shape.107,110 Although much has been written about the increased number of sequential sagittal images that will depict the discoid meniscal body, it is simpler to judge for the presence of the enlarged meniscus on the coronal view (see Fig. 63-8A). A discoid meniscus is suggested when the meniscal body is wider than 15 mm or extends into the intercondylar notch.4,127,129,139 The Wrisberg variant is a subtype of the lateral discoid meniscus that lacks a posterior attachment to the tibia, resulting in a hypermobile meniscus.96,129 In these cases, the Wrisberg ligament becomes the only posterior stabilizer of the lateral meniscus. This variant is often symptomatic: the knee may snap during flexion and lock in extension. Another normal variant of the menisci is the so-called meniscal flounce.152 The meniscal flounce is a wavy or S-shaped fold along the free edge of the meniscus that occurs in the absence of a tear (Fig. 63-9). It may represent normal physiologic distortion related to knee positioning,102 or it may be observed in the setting of true ligamentous laxity in which femorotibial sliding results in buckling of the free edge of the meniscus.59 It has no known clinical significance, and its

ical load on the normal meniscus (traumatic tears) or normal mechanical load on the degenerating meniscus (degenerative tears). Acute traumatic tears are found in the young athletic population. In this situation, the lateral meniscus is most commonly torn, usually from twisting strains on the knee when it is either only slightly flexed or fully extended.40 Tears found in the older population generally occur at sites of meniscal degeneration and are often asymptomatic.13,74 In general, the medial meniscus is more commonly injured than the lateral meniscus, possibly because it is less mobile and because it transmits a greater load during most activities.40,47 The posterior horns of the menisci are more commonly injured than the anterior horns.40,136 Tears isolated to the anterior two thirds of the meniscus are uncommon.47 In one arthroscopic series of 1086 medial meniscal tears, the posterior horn was involved in 98% of torn medial meniscus.92 Lateral meniscal tears are more varied in location. In the posterior horn of the medial meniscus, tears are more frequently seen on the inferior surface because of increased stress and strain with femoral tibial rotations.136

Diagnostic Criteria.  The diagnostic criteria for meniscal tears seen on knee MRIs are (1) intrameniscal signal extending to the articular surface of the meniscus and (2) abnormal meniscal morphology. Intrameniscal signal.  It is important to differentiate accurately between an intrameniscal signal that does not extend to the articular surface and one that does. The former represents a degenerative change within the meniscus, and the latter represents a meniscal tear. A useful grading system first described by Stoller and colleagues recognizes three patterns of intrameniscal signal (Fig. 63-10).137 Globular signal intensity within the meniscus not contiguous with the articular surface is defined as a grade I lesion. A grade II designation is assigned to linear signal intensity within the meniscus that is not contiguous with the articular surface. Grade I and II signals represent myxoid or mucinous degenerative changes within the menisci that rarely correlate with arthroscopically demonstrable tears. They are most commonly seen in the posterior horn of the medial meniscus. Regions of intermediate signal intensity with extension to an articular surface constitute grade III lesions. Increased signal intensity representing a torn meniscus is best seen on T1- and PD-weighted images and might not be seen on T2-weighted images. T2-weighted images are more helpful when synovial fluid insinuates itself within the tear. Such signal changes represent a torn meniscus only when the signal extends to the superior or inferior articular surface. The periphery or capsular margin of the meniscus does not represent such a surface. The findings may be seen on only

2184

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 63-8  Discoid meniscus. A, Coronal fat-saturated T2-weighted image shows a complete discoid lateral meniscus extending to the intercondylar notch. B, Axial fat-saturated T2-weighted image shows a complete discoid lateral meniscus (arrow). Note absence of the free edge as present in the medial meniscus. C and D, Sagittal PD and coronal fat-saturated T2-weighted images demonstrate hypertrophied discoid meniscus with internal high signal. (A and B, Courtesy Sarah S. Chan, MD, Sutter Medical Center, Sacramento, Calif.; C and D, Courtesy Amin Matin, MD, Sutter Medical Center, Sacramento, Calif.)

CHAPTER 63  Knee

2185

When the posterior horn of the medial meniscus is equal in size to or smaller than its anterior horn, despite a normal triangular appearance and contour, a tear must be suspected because the posterior horn of a normal medial meniscus is normally almost twice the size of the anterior horn. In general, acute traumatic tears in young persons and athletes commonly demonstrate rounded or smooth margins, and chronic degenerative tears often have frayed, fibrillated, and irregular margins.

Classification of Meniscal Tears.  Meniscal tears can be horizontal

FIG 63-9  Meniscal flounce. Sagittal PD image shows an undulating portion of the body of the meniscus. (From Beltran J, Shankman S: The menisci. In Davies AM, Cassar-Pullicino VN, editors: Imaging of the knee. Heidelberg, Germany, 2002, Springer, p 139. With permission of Springer Science and Business Media.)

one image or in only one plane. When abnormal signal contacting the surface is seen on only one image, the diagnosis of a possible tear rather than a definite tear should be made.30,39 If abnormal signal is seen on more than one image, then a diagnosis of a meniscal tear is more definite, with a positive predictive value of greater than 90%.31 When it is unclear whether the signal extends to the articular surface, a true tear is probably not present. Altering the grayscale of an image should be done cautiously because false-positive images may result from exaggeration of intrameniscal signal. Abnormal meniscal morphology.  Morphologic changes of the meniscus associated with meniscal tears include blunting of the tip of the inner free edge of the meniscus, displacement of a portion of the meniscus, interrupted appearance of the meniscus, and abnormal size of a segment of the meniscus. When a portion of the free margin of the meniscus is detached (a bucket-handle tear), the remaining peripheral portion appears small and often truncated. The displaced portion usually lies within the intercondylar notch beneath the PCL. Occasionally it flips above the anterior horn. Coronal views are helpful in further identifying such displaced fragments, which should not be confused with the central attachment sites of posterior meniscal horns. Nondisplaced peripheral separations may be difficult to detect. The most specific finding is irregular, jagged, increased signal intensity at the periphery that brightens on T2-weighted images. An abrupt change in the contour of the meniscus (notch sign) is an important indicator of a meniscal tear. Although it can be simulated by the normal meniscal flounce, the presence of abnormal intrameniscal signal makes the notch sign a more definite indicator of a meniscal tear.24,110

(parallel to the tibial plateau), vertical (perpendicular to the tibial plateau), or complex (two or more tear configurations not easily categorized into a type of tear). Tears can be either full thickness or partial thickness and nondisplaced or displaced. Horizontal tears.  Horizontal tears are approximately parallel to the tibial plateau (Fig. 63-11). The pure horizontal cleavage type extends to the apex and divides the meniscus into upper and lower halves; sometimes they are referred to as fishmouth tears (see Fig. 63-11B).136 The noncleavage-type horizontal tear extends to the superior or inferior articular surface of the meniscus (see Fig. 63-11C). Horizontal tears are most common within the posterior horn of the medial meniscus. Linear horizontally oriented grade II signal intensity is sometimes mistaken for a complete horizontal tear. Meniscal cyst formation is commonly associated with horizontal tears. Vertical tears.  Vertical tears are perpendicular to the tibial plateau. A vertical longitudinal tear occurs between the circumferential collagen fibers parallel to the long axis of the meniscus (Fig. 63-12A). Longitudinal tears may be obliquely oriented, but they are not parallel to the tibial plateau. Vertical grade III signal intensity located within the peripheral third of the meniscus is considered a peripheral longitudinal tear (see Fig. 63-12B and C). If the vertical or oblique signal  is located in the inner third of the meniscus, it usually represents a  flap tear.136 A vertical radial tear occurs perpendicular to the circumferential collagen fibers and long axis of the meniscus. Radial tears are also known as free-edge tears because they involve the inner edge of the meniscus (Fig. 63-13). All radial tears are always vertical tear occurring at the free edge; this is the classic radial tear. MRI obtained perpendicular to the orientation of a radial tear easily demonstrates the tear as a small gap or cleft in the meniscus (cleft sign), which sometimes can be observed to move across the meniscus on successive images (marching cleft sign).57,59 If the orientation of the imaging plane is in the same direction as a full-thickness radial tear, a section through the same plane of the tear will demonstrate an absent meniscus or high signal intensity not representative of a meniscus because of volume averaging, which results in the so-called ghost sign. A partial-thickness radial tear will demonstrate blunting or abrupt truncation of the apex of the meniscus (truncated triangle sign). The term parrot-beak tear has been frequently used to describe a vertical radial tear at the inner edge of the meniscus that curves obliquely across the circumferential collagen fibers (Fig. 63-14). A secondary longitudinal component will be present more peripherally, resulting in the parrot-beak appearance.67 Some consider this a nondisplaced flap tear, with the flap formed by the free edge of the meniscus created by the curvilinear tear.136 The term parrot-beak tear should be reserved for arthroscopy reports and not used in MRI descriptions of meniscal tear patterns. Root tears.  Root tears are full-thickness radial tears at the central tibial attachment sites of the menisci and are called the root ligaments. These have been described posteriorly only (Fig. 63-15A). In the posterior horn of the medial meniscus, meniscal tissue should be present immediately medial to the tibial attachment of the PCL. In the lateral

2186

PART II  CT and MR Imaging of the Whole Body

A

C

Normal

Grade I

Grade II

Grade III

B

D FIG 63-10  Grading system for meniscal signal. A, Diagram of the grading system of signal intensity changes in the meniscus. B, Grade I signal—globular signal intensity within the meniscus not contiguous with the articular surface. C, Grade II signal—linear signal intensity within the meniscus not contiguous with the articular surface. D, Grade III signal—linear signal intensity within the meniscus extending to the articular surface. (A, from Beltran J, Shankman S: The menisci. In Davies AM, Cassar-Pullicino VN, editors: Imaging of the knee. Heidelberg, Germany, 2002, Springer, p 140. With permission of Springer Science and Business Media.)

CHAPTER 63  Knee

A

B

C FIG 63-11  Horizontal tear. A, Diagram of a horizontal tear. On cross section, a horizontal tear has a similar appearance to the grade 3 meniscal signal in Figure 63-9A and D. B, Horizontal tear may be cleavage type, which extends to the apex of the free edge of the meniscus, or C, noncleavage type extending to the inferior or superior articular surface.

2187

2188

A

PART II  CT and MR Imaging of the Whole Body

B

C

FIG 63-12  Longitudinal tear. A, Diagram of a longitudinal vertical tear running parallel to the long axis of the meniscus. B, Sagittal PD image shows vertically oriented high signal intensity at the periphery of the posterior horn medial meniscus extending to both superior and inferior articular surfaces, representing a peripheral longitudinal tear. C, Coronal T1-weighted image demonstrates an obliquely oriented high signal intensity at the periphery of the body of the medial meniscus extending to both superior and inferior articular surfaces, representing a peripheral longitudinal tear.

meniscus, the posterior horn should cover the most medial portion of the lateral tibial plateau. If this is not the case, a root tear is highly suspected. A secondary sign of root tear is the presence of subluxation of the meniscus in the absence of osteoarthritis.143 On posterior coronal MRI, abrupt loss or blunting of the normal meniscotibial attachment and foreshortening of the meniscus toward the posterior aspect of the intercondylar notch are seen (see Fig. 63-15B). On central sagittal images adjacent to the intercondylar notch, absent meniscal tissue or diffuse increased signal intensity in the expected location of the posterior root ligament is seen, resulting in a ghost appearance and the so-called ghost sign of radial tears (see Fig. 63-15D). Bucket-handle tears.  Bucket-handle tears are displaced longitudinal tears. The displaced inner fragment often has the appearance of a handle, and the remaining peripheral segment attached to the tibial plateau resembles a bucket—hence the name bucket-handle tear (Fig. 63-16A). A sensitive but nonspecific sign on MRI is the absent bowtie sign, in which fewer than two consecutive sagittal images of the bowtieshaped meniscal body is seen.59 A caveat of the absent bowtie sign is that it may be identified in the intact small meniscus of a child or small adult and in the partially resected meniscus. Thus confirmation of a bucket-handle tear is mandatory and can be made by identification  of a displaced fragment. In fact, the most reliable sign is the finding of a displaced fragment of meniscus in the intercondylar notch, which can be seen better on coronal images (see Fig. 63-16B and C).56,60,130,147 This sign is associated with a truncated or shortened meniscus on coronal images. On sagittal images, the displaced fragment can be seen lying anterior and parallel to the PCL. This is known as the double PCL sign (Fig. 63-17A).130,147 Bucket-handle tears may also result in the flipped meniscus sign, double delta sign, or double anterior horn sign, in which the displaced meniscal segment of the posterior horn is located anterior to, on top of, or posterior to the anterior horn (see Fig. 63-17B).56,118,136 Helpful clues in these cases include nonvisualization of the posterior horn of the meniscus, the appearance of a larger anterior horn than posterior horn, or an anterior horn taller than 6 mm. Flap tears.  Flap tears are commonly described as short-segment horizontal tears with fragments displaced into the notch or superior or inferior gutters (Fig. 63-18C).47,59,110,118 In general, flap tears can result whenever there is a change in direction of the tear.136 For instance,

a horizontal tear that extends superiorly or inferiorly resulting in a change in direction (into a nonhorizontal plane) is a flap tear (see Fig. 63-18A and B). A radial tear with a secondary longitudinal component, sometimes termed a parrot-beak tear in arthroscopy reports, is also considered by Stoller and colleagues to be a flap tear.136 Flap tears are the most common type of meniscal tears and most commonly occur in the posterior horn. They may develop following only minimal meniscal trauma of the degenerative meniscus from chronic shear stresses. They are frequently associated with oblique signal intensity on sagittal images and are sometimes referred to as oblique tears, although most surgeons would consider flap tear to be the proper terminology. On the other hand, this term should not be used to describe oblique signal intensity unless other MRI findings confirm the diagnosis of a flap tear.136 Both flap and longitudinal tears may have obliquely oriented signal intensity. One can differentiate between the two by noting that flap tears involve the inner one third to one half of the meniscus with a superior or inferior leaf extension that creates the flap. In contrast, longitudinal tears are more likely to involve the peripheral third of the meniscus. MRI findings of flap tears seen in the sagittal plane include (1) a vertical tear in the inner third of the meniscus, (2) deficiency of the inferior surface of the inner third of the meniscus, (3) blunted free edge of the meniscus with the displaced flap inferior to the periphery of the meniscus, (4) a change in the slope of the superior surface of the meniscus, reflecting a change in direction of the tear that creates the flap, and (5) folding and stacking of the displaced flap on the parent meniscus, producing a double layer of meniscus or double-decker appearance.136 On coronal MRI, extrusion of part of the meniscus into the coronary recess below the joint line is sometimes seen. Flap tears may also displace into the posterior intercondylar notch and mimic a bucket-handle tear. On coronal images, the displaced meniscal fragment is seen as an additional structure in the intercondylar notch, a similar finding with bucket-handle tear (see Fig 63-18D). On sagittal images, the meniscal fragment may displace beneath the posterior longitudinal ligament, simulating the double PCL sign of a bucket-handle tear (see Fig 63-18C). However, the important distinction between a flap tear and bucket-handle tear is that the meniscal Text continued on p. 2193

CHAPTER 63  Knee

A

B

C(1)

C(2)

C(3)

C(4)

C(5) FIG 63-13  Radial tear. A, Diagram of a radial tear. B, MRI obtained perpendicular to the orientation of a radial tear will show the tear as a small gap or cleft in the meniscus (cleft sign). C, The cleft in the meniscus may sometimes be observed to “march” across the meniscus on successive images (marching cleft sign). Continued

2189

2190

PART II  CT and MR Imaging of the Whole Body

E

D

FIG 63-13, cont’d  D, MRI obtained in the same direction and same plane as a full-thickness radial tear will demonstrate an absent meniscus (ghost sign). E, A partial-thickness radial tear will demonstrate abrupt truncation of the apex of the meniscus (truncated triangle sign). (A, From Beltran J, Shankman S: The menisci. In Davies AM, Cassar-Pullicino VN, editors: Imaging of the knee. Heidelberg, Germany, 2002, Springer, p 143. With permission of Springer Science and Business Media.)

A

B FIG 63-14  A, Diagram of a parrot-beak tear. B, Axial fat-saturated T2 image demonstrates a parrot-beak tear. (A, From Beltran J, Shankman S: The menisci. In Davies AM, Cassar-Pullicino VN, editors: Imaging of the knee. Heidelberg, Germany, 2002, Springer, p 144. With permission of Springer Science and Business Media.)

CHAPTER 63  Knee

B

A

C

D FIG 63-15  Root tear. A, Diagram of a posterior meniscal root ligament tear. B and C, Posterior coronal and axial fat-saturated T2-weighted images show abrupt loss of the normal meniscotibial attachment and foreshortening of the meniscus toward the posterior aspect of the intercondylar notch. D, On central sagittal image adjacent to the intercondylar notch, absence of meniscal tissue or diffuse increased signal intensity in the expected location of the posterior root ligament is seen, resulting in a ghost appearance and hence the ghost sign.

A

B

C

FIG 63-16  Bucket-handle tear. A, Diagram of a bucket-handle tear with the displaced fragment on top of the tibial spine. B, Coronal fat-saturated T2-weighted image shows displaced meniscal tissue from a buckethandle tear into the intercondylar notch. C, Far anterior coronal image shows displaced meniscal tissue in the intercondylar notch. Displaced meniscal tissue from a bucket-handle tear is present throughout the intercondylar notch from posterior to anterior, in contradistinction to a flap tear, which will only show displaced meniscal tissue in the intercondylar notch on posterior coronal images but not anterior coronal images. (A, From Beltran J, Shankman S: The menisci. In Davies AM, Cassar-Pullicino VN, editors: Imaging of the knee. Heidelberg, Germany, 2002, Springer, p 144. With permission of Springer Science and Business Media.)

2191

B

A

FIG 63-17  Bucket-handle tear. A, Sagittal PD image shows displaced meniscal tissue at the intercondylar notch, indicative of the bucket-handle tear. Meniscal tissue beneath the posterior cruciate ligament is known as the double PCL sign. B, Sagittal PD image shows a truncated posterior horn resulting from a bucket-handle tear. The displaced fragment has flipped anterior to the anterior horn of the lateral meniscus (arrow).

A

D

C

B

E FIG 63-18  Flap tear. A, Diagram of a horizontal flap tear of the medial meniscus. B, Coronal PD fat-saturated image demonstrating a displaced flap of the meniscus. C and D, Sagittal PD and coronal fat-saturated T2-weighted images show displaced flap tear into the posterior intercondylar notch. On the sagittal image, the meniscal fragment is located just beneath the PCL, similar to the PCL sign of bucket-handle tear. On coronal image, the additional structure within the intercondylar notch is a similar finding seen with buckethandle tear. However, the important distinction between a flap tear and bucket-handle tear is the meniscal fragment in a flap tear will only be present on posterior coronal images but not on anterior coronal images. Conversely, a bucket-handle tear will always demonstrate displaced meniscal fragment throughout the intercondylar notch from posterior to anterior coronal images. E, Axial fat-saturated T2-weighted image demonstrates the displaced meniscal fragment from a flap tear. Note that it does not extend completely anteriorly. (A, From Beltran J, Shankman S: The menisci. In Davies AM, Cassar-Pullicino VN, editors: Imaging of the knee. Heidelberg, Germany, 2002, Springer, pp 145–146. With permission of Springer Science and Business Media.)

CHAPTER 63  Knee fragment in a flap tear will only be present on posterior coronal images but not on anterior coronal images, because it most commonly occurs in the posterior horn of the meniscus. Conversely a bucket-handle tear will always demonstrate displaced meniscal fragment throughout the intercondylar notch from posterior to anterior coronal images. Clinically a flap tear causes intermittent locking because the flap tear can be displaced one moment and flipped back into position another moment, as opposed to a bucket-handle tear, which will always be displaced and cause persistent locking. Meniscocapsular separation.  Meniscocapsular separation (MCS) occurs when the peripheral portion of the meniscus detaches from its capsular attachment. This results in hypermobility of the meniscus, which can be a problem clinically.26 The tear occurs in the periphery of the meniscus, which is typically well vascularized and likely to heal. As a result, MCS is often treated conservatively without surgery in the absence of any other associated ligamentous injury. It should be noted that the absence of a distinct cleavage plane between the capsule and the meniscus might make it impossible to distinguish between a peripheral meniscal tear and MCS on MRI and arthroscopy. This distinction, however, is not clinically important; both conditions are treated conservatively. MCS cannot be reliably diagnosed on MRI because some of the signs used to diagnose MCS are nonspecific and often found in other conditions.115 In addition, many conditions can mimic MCS, such as perimeniscal fat, joint recess and fluid related to the MCL bursa, pes anserinus bursa, or meniscal cyst.26

2193

Lateral inferior geniculate artery. The lateral inferior geniculate artery is closely applied to the meniscus as it wraps around the knee and is found in the periarticular fat between the meniscus and the lateral collateral ligament. The space between the lateral inferior geniculate artery and the anterior horn of the lateral meniscus can produce the appearance of a tear on sagittal images.61 The artery can be traced on adjacent sagittal images, thereby allowing distinction from a tear. Popliteus tendon. The attachment of the posterior horn of the lateral meniscus to the joint capsule is interrupted by the popliteus tendon. The separation between the posterior horn of the lateral meniscus and the popliteus tendon can mimic a peripheral detachment or tear (Fig. 63-21).61 Meniscofemoral ligaments. The meniscal attachments of the meniscofemoral ligaments can simulate a tear. The posterior meniscofemoral ligament is usually seen at its origin at the superior margin of

Differential Diagnosis.  It is important to be familiar with the common normal anatomic structures, variants, and artifacts that can simulate a tear. In most cases, analysis of sequential MRI and knowledge of common variants and artifacts facilitate proper interpretation. Intermeniscal ligaments. The meniscal attachment sites of the anterior transverse meniscal ligament and rare posterior transverse meniscal ligament can simulate a tear at the point at which they separate from the menisci (Fig. 63-19). The rare medial and lateral oblique meniscomeniscal ligaments can be mistaken for displaced flap or bucket-handle tears.119 The course of these ligaments can be traced on sequential sagittal MRI of the central portion of the knee. Anterior horn of the lateral meniscus. The anterior horn of the lateral meniscus may demonstrate a speckled increased-signal-intensity pattern on PD-weighted images. This normal variation represents dense collagenous fibers of the ACL intertwined with the fibrocartilage of the meniscus itself and should not be mistaken for a tear (Fig. 63-20).123

A

B

FIG 63-20  Sagittal PD image shows a speckled or striated pattern of increased signal intensity at the anterior horn of the lateral meniscus near its central attachment site.

C

FIG 63-19  Transverse ligament. A-C, Sagittal PD images show the transverse ligament (arrow) joining the anterior horn lateral meniscus. The junction should not be confused with a tear.

2194

PART II  CT and MR Imaging of the Whole Body

FIG 63-21  Pseudotear, popliteus tendon sheath. Sagittal PD image

FIG 63-22  Pseudotear, posterior horn of the lateral meniscus. Sagittal

shows the normal high signal intensity of the popliteus tendon sheath between the posterior horn of the lateral meniscus and popliteus tendon (arrow).

T1-weighted image shows the takeoff of the ligament of Wrisberg at the posterior horn of the lateral meniscus. The signal between the ligament of Wrisberg and the lateral meniscus itself (arrow) should not be confused with a tear.

the posterior horn of the lateral meniscus. It sometimes appears as a round dot adjacent to the superior aspect of the posterior horn and should not be confused with a tear (Fig. 63-22). Rarely an anteromedial meniscofemoral ligament is seen arising from the anterior horn of the medial meniscus and attaching to the posterolateral wall of the intercondylar fossa, but this should not be mistaken for a tear in the anterior horn medial meniscus.132 Inspecting contiguous sections should help to avoid these errors. Bursa of the medial collateral ligament. Fluid within the medial collateral ligament bursa may be falsely mistaken for a peripheral meniscocapsular tear. Meniscal flounce. In meniscal flounce, the waviness or buckling of the free edge of the menisci seen on sagittal or coronal images simulates a tear (see Fig. 63-9). Additionally the truncated appearance of meniscal flounce on coronal images can be misinterpreted as a sign of tear.152 Intrameniscal signal intensity. Grade 2 linear intrameniscal signal intensity reflecting mucinous or myxoid degeneration may mimic a horizontal tear (see Fig. 63-10C). Volume averaging. Volume averaging of the capsule and the menisci on peripheral sagittal slices of the meniscal body is sometimes confused with a horizontal tear in the meniscus (Fig. 63-23).61 Coronal images will show no abnormality and can be used to exclude a tear. Truncation artifact. Truncation artifact is seen when parallel signal lines are produced in the phase-encoding direction along edges where there is a large abrupt transition in tissue signal intensity. An example is the periphery of the meniscal body (see Fig. 63-23B). This truncation artifact may be accentuated with meniscal windowing and with an acquisition matrix of 128 × 256. It is reduced or nearly eliminated with higher acquisition matrix of 256 × 256 and with the orientation of the phase-encoding gradient in the anteroposterior direction.144 Magic-angle phenomenon. The magic-angle phenomenon is observed on short-TE MRI and refers to the increased signal intensity that

occurs when the collagen fibers of the menisci are oriented at 55 degrees in relation to the main magnetic field along the long axis of the magnetic bore.114 The magic-angle phenomenon is most commonly observed in the posterior horn of the lateral meniscus as a result of the angular orientation of the meniscus at this particular location. It should not be interpreted as a tear (Fig. 63-24).103 Vacuum phenomenon. Magnetic susceptibility of intraarticular gas may produce a low-signal-intensity void or blooming artifact on gradient-recalled echo (GRE) images, which may be mistaken for a meniscal tear.126 This error can be avoided by correlating with radiographs.

Treatment.  Because of the importance of the menisci in stabilizing the knee and minimizing degenerative changes to the joint, management of meniscal tears has shifted away from total meniscectomy and toward meniscal preservation and repair.33,99,138 An accurate description of the meniscal injury by the interpreting radiologist is therefore important to aid the referring clinician in surgical planning. A description of the location of the tear should include whether it is in the anterior horn, body, or posterior horn and whether it is in the peripheral third of the meniscus (rough approximation of the vascularized zone), inner two thirds, or both. Location is important for the referring clinician because peripheral tears are more likely to heal than central tears.58 In addition, the identification of a small peripheral longitudinal tear may obviate arthroscopy.67 The specific type of tear is clinically relevant and should always  be described. Traumatic peripheral vertical longitudinal tears in the peripheral third and meniscocapsular injuries are reparable, whereas horizontal, large radial, flap, complex, and degenerative tears are not reparable and require partial meniscectomy.67 Short radial tears less than 5 mm long may not require resection. The extent of the tear, whether complete if it extends from one articular surface to another or partial if it does not, should likewise be described because it also

CHAPTER 63  Knee

A

2195

B FIG 63-23  Far peripheral sagittal pseudotear. A, Far sagittal PD image of the medial joint compartment shows normal linear high signal intensity representing volume averaging of the joint capsule, concave surface of the meniscus periphery, and normal fat in between (arrows). This may represent truncation artifact as well. B, Diagram of the medial meniscus in the coronal plane. The two lines indicate the sagittal images through the medial meniscus on which a pseudotumor may be encountered.

FIG 63-24  Magic-angle artifact (arrow). Coronal PD fat-saturated image shows the magic-angle phenomenon in the posterior horn of the lateral meniscus, where it is most commonly observed because of the angular orientation of the meniscus at this particular location. This artifact should not be interpreted as a tear.

factors in determining whether the tear is reparable. Identification of potentially reparable tears influences not only the decision for meniscal repair but also the timing of arthroscopy. Meniscal repairs have improved outcome if they are performed within 8 weeks of the injury.41 In addition, identification of reparable tears can influence an athlete’s decision about whether to finish the season, because continued sport activities could convert reparable tears into irreparable ones.67 The location of any displaced fragment should also be reported because some areas within the knee joint are difficult to see during arthroscopy. This information will guide the surgeon in inspecting regions that might otherwise be overlooked.

Postoperative Appearance.  The MRI appearance of the postsurgical meniscus depends on the nature of the surgical intervention. After partial meniscectomy, it can be difficult to evaluate tearing of the

meniscal remnant or incomplete meniscal resection. The MRI appearance of an arthroscopically normal meniscal stump varies. It may demonstrate a smooth contour with homogeneous signal intensity (the standard MRI criteria for meniscal tears are useful only in this situation), or it may demonstrate an irregular surface contour with inhomogeneous signal intensity. The latter postoperative appearance will in fact represent a normal meniscal remnant in most cases. For this reason, it may be very difficult to delineate true pathology (Fig. 63-25).131 Peripheral meniscal tears are often treated conservatively or with arthroscopic repair.5,32,53,55,83,93 In either situation, the increased signal on T1- or PD-weighted images seen on preoperative studies may persist even in patients who have become asymptomatic and presumably healed. The presence of such signal, therefore, should not be misinterpreted as meniscal re-tear. In this regard, some cases of falsepositive MRI findings may actually represent older injuries that have healed by the time arthroscopic examinations are performed.37 MR arthrography is recommended in patients with prior partial meniscectomy or meniscal repair. Both indirect and direct MR arthrography of the knee have proven highly accurate in detecting meniscal tears in the postoperative meniscus. Indirect MR arthrography involves acquiring MR images 10 to 15 minutes after intravenous injection of gadolinium. Owing to synovial excretion of contrast, a tear can be diagnosed if gadolinium is seen within the meniscus extending from the articular surface. However, false positives may occur because granulation or scar tissue may enhance.141 Direct MR arthrography offers several advantages over other methods.47,90,122,141 With direct MR arthrography, diluted gadolinium is injected intraarticularly and imaging performed after 10 to 15 minutes of moderate exercise. Contrast will insinuate itself into true tears and will not extend into areas of healing (Fig. 63-26). The lower viscosity of gadolinium compared with synovial fluid and distention of the joint by the contrast medium make the contrast medium more likely to extend into a tear. Additionally the T1-weighted sequences used to evaluate the gadolinium also provide an inherently higher SNR, which serves as an advantage of both indirect and direct MR arthrography over conventional MRI. If a joint effusion is present, conventional MRI with T2-weighted images of the knee may be diagnostic. Fluid, like contrast, may

2196

PART II  CT and MR Imaging of the Whole Body

A

C

B

FIG 63-25  Partial resection of the meniscus. A, Coronal fat-saturated PD image demonstrates a normal meniscal stump after partial resection. B and C, Coronal fat-saturated PD and sagittal PD images demonstrate intermediate linear signal within clinically “normal” postoperative meniscal stumps, which may be mistaken for meniscal re-tears. (Courtesy Amin Matin, MD, Sutter Medical Center, Sacramento, Calif.)

anterior and lateral to the anterior tibial spine. The fanlike tibial attachment is larger than the femoral attachment (Fig. 63-27A). The ligament measures approximately 4 × 1 cm. It consists of two distinct bundles separated by loose connective tissue and fat.  The smaller anteromedial bundle is taut in flexion, and the larger posterolateral bundle is taut in extension. Synovium covers the anterior surface of the ACL, making the entire structure intraarticular and extrasynovial.

Imaging Anatomy

FIG 63-26  Sagittal fat-saturated T1-weighted image of the knee following intraarticular injection of a gadolinium saline solution shows contrast material within the posterior horn of the medial meniscus after partial meniscectomy, indicative of a recurrent tear.

insinuate itself into a true tear.3 The specificity for diagnosing a tear in patients with prior meniscal surgery is increased if a displaced meniscal fragment is seen.141

ANTERIOR CRUCIATE LIGAMENT Normal Anatomy The ACL is principally a stabilizer against anterior translation of the tibia. Secondarily it resists internal tibial rotation. It extends from a semicircular attachment at the posteromedial aspect of the lateral femoral condyle to the anterior intercondylar region of the tibia just

The ACL is best seen on sagittal oblique images oriented parallel to the cortex of the lateral femoral condyle (see Fig. 63-27A). With the knee placed in extension for the examination, the ligament should appear taut and angled about 60 degrees to the tibial plateau. If the knee is flexed more than 5 degrees, it may appear lax. In general, the ACL may appear as a solid band or as several separate low-signal-intensity bundles. The tibial attachment is usually more clearly delineated than the femoral attachment on sagittal images because of partial-volume averaging with the lateral femoral condyle. The signal intensity on T1-weighted and spin-density images at the tibial insertion is usually increased, presumably the result of either interposed fat or the decreased density of the ligament itself. The anterior border of the taut ACL should nearly parallel the roof of the intercondylar notch.81,95 In patients suspected of ACL injury, MRI with the knee in mild (30-degree) flexion has been suggested by Lee and colleagues as a means to improve viewing of the proximal ACL. In this position, a reduction of the volume averaging may achieve a more accurate depiction of the ACL within the intercondylar notch.79 Coronal and axial images are useful in confirming the findings made on sagittal images. Coronal images show the ACL as a curvilinear fanlike structure adjacent to the horizontal segment of the PCL, near the medial surface of the lateral femoral condyle (see Fig. 63-27B). Axial images depict the proximal ACL as a low-signal band that is flattened against the medial surface of the lateral femoral condyle (see Fig. 63-27C). In the axial plane, the distal ligament is not as well seen. All imaging sequences demonstrate fat within the intercondylar notch adjacent to the ACL origin.

Trauma ACL tears are up to eight times more prevalent in women than  in men.45,65,81 Both partial and complete tears tend to occur at the

CHAPTER 63  Knee

A

B

C FIG 63-27  Normal ACL. A, Sagittal PD-weighted fat-saturated image shows a normal low-signal-intensity ACL (arrows). Slight increased signal intensity at the tibial attachment site is normal. B, Coronal PD-weighted image shows a normal low-signal-intensity ACL (arrows). The two discrete bundles (anteromedial and posterolateral) can occasionally be delineated. C, Contiguous axial PD-weighted images show the proximal ACL flattened against the lateral femoral condyle (arrows).

2197

2198

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 63-28  ACL tear. Sagittal (A) and coronal (B) PD fat-saturated images show discontinuity of the proximal fibers of the ACL (arrows) with reactive edema.

proximal to midportion. Avulsion fractures, usually seen in young patients, are more common at the tibial insertion.104

Mechanisms of Injury.  ACL tears may be the result of either contact or noncontact injury. Several mechanisms have been described. The pivot shift injury, the most commonly implicated noncontact mechanism, occurs in skiers and American football players as a result of rapid deceleration with simultaneous change in direction. The knee incurs a valgus load with internal rotation of the tibia. The clip injury, the most commonly implicated contact mechanism, occurs as a result of a highenergy blow to the lateral aspect of the knee. The acute valgus stress onto the partially flexed knee results in distractive forces at the medial joint compartment and compressive forces at the lateral joint compartment. Tearing of the ACL is thus commonly accompanied by impaction fracture at the lateral femoral condyle (the deep sulcus sign),112 bone bruise at the posterior aspect of the lateral tibial plateau, and MCL and medial meniscal tears (the O’Donoghue triad). A smaller contusion in the medial femoral condyle at the MCL attachment may also be present as a result of avulsive stress. Small contusion of the posterior lip of the medial tibial plateau has also been described and attributed to contrecoup impaction as the knee relocates following the ACL tear.69 Another mechanism of ACL tear involves hyperextension, which may be the result of an awkward landing or a direct blow onto the anterior tibia with the foot planted. Impaction occurs at the anterior femoral condyle and anterior tibia, and the ACL is torn along  with possible injury to the PCL, menisci, or posterolateral corner structures. A final mechanism of ACL injury involves varus stress and internal rotation of the tibia. The pathomechanism is essentially the reverse of the pivot shift injury, with distractive forces present laterally and compressive forces present medially. If there is avulsion fracture of the lateral tibial plateau (a Segond fracture) on radiography, an underlying ACL tear will be detected 75% to 100% of the time.17

Imaging Appearance.  The MRI appearance of an ACL tear varies according to the extent and chronicity of the injury. The most common complete interstitial tear appears as a fairly ill-defined mass of reactive edema and hemorrhage rather than as a discrete structure with a tear within. There will be complete discontinuity of the ACL fibers, often with increased T2 signal traversing buckled indistinct ligament fibers

(Fig. 63-28). The absence of these low-signal fibers at the expected site of attachment is occasionally the most definitive sign of proximal ACL tear in subtle cases (Fig. 63-29). In younger patients, avulsion fractures at the tibial eminence or (less commonly) the femoral attachment  may occur (Fig. 63-30). Partial tears demonstrate increased T2 signal with contour irregularity. In some cases, they may cause no contour abnormality but can lead to obscuration of the fasciculi within the ligament. Partial and complete tears of the ACL may heal over time, resulting in resolution of the increased T2 signal hyperintensity and eventual scar formation that restores its continuity. After 6 to 8 weeks, the ligament may demonstrate laxity, and it will be seen as bowed or dropped within the intercondylar notch. Occasionally a chronic tear with scar formation demonstrates a normal MRI appearance despite being insufficient on examination.145 Conversely the ligament may have an indistinct margin or might not be seen at all (Fig. 63-31). Indirect or secondary signs of ACL tear are also sometimes seen. Such signs include anterior translation of the tibia relative to the femur (the anterior tibial translocation sign),18,20,49 buckling of the PCL (see Fig. 63-31), posterior displacement of the lateral meniscus from the tibial plateau (the uncovered lateral meniscus sign), indentation of the anterior margin of the medial femoral condyle, avulsion fracture of the lateral tibial plateau (the Segond fracture), and fairly specific combinations of bone contusions that depend on the mechanism of trauma (Fig. 63-32). Injury to other supporting structures  of the knee are also commonly seen along with ACL injury. For instance, peripheral vertical tears of the posterior horn of the medial meniscus are characteristically encountered in the setting of clip injuries (Fig. 63-33). Mucinous (or mucoid) degeneration of the ACL represents a degenerative process and should not be confused with an acute tear (Fig. 63-34A). Pericruciate ganglion cyst formation may accompany this process (see Fig. 63-34B).

Treatment and Follow-up.  Most ACL repairs are intraarticular and consist of arthroscopic ACL reconstruction with patellar-bone centralthird patellar tendon tibial bone grafts. MRI demonstrates a welldefined, intact, low-signal-intensity autograft in clinically stable postoperative knees. Revascularization, which takes place 4 to 9 months postoperatively, may increase the signal intensity of the graft, leading to isointensity with the intraarticular space. This may account for

CHAPTER 63  Knee

A

2199

B FIG 63-29  ACL tear. A and B, Contiguous axial PD-weighted images with fat saturation show absence of low-signal ACL fibers at femoral attachment onto medial margin of the lateral femoral condyle (arrows).

A

B

FIG 63-30  ACL avulsion fracture. Sagittal (A) and coronal (B) images show avulsion fracture of the tibial eminence at the ACL tibial insertion (short arrows). The ACL fibers (long arrows) remain intact.

apparent discrepancies in graft size. In this regard, focal and diffuse signal changes may be physiologic. Therefore the object of postoperative evaluation in the first year following surgery is to identify the continuity of the graft itself rather than contour deformities or abnormal signal.105 Following the first year, graft impingement precedes graft disruption and is usually secondary to anterior placement of the tibial tunnel.

Localized anterior arthrofibrosis may result from such impingement or from debris raised by drilling the tibial tunnel. Because of its headlike appearance on arthroscopy and its characteristic focal reddish blue discoloration that resembles the eye of a Cyclops, it is often referred to as the Cyclops lesion. It usually appears as a low-signal-intensity mass surrounded by fluid.109 Other complications of ACL reconstruction include patellar fracture and dislodged hardware.8,108

2200

PART II  CT and MR Imaging of the Whole Body

POSTERIOR CRUCIATE LIGAMENT The clinical diagnosis of PCL injury is often more challenging than the diagnosis of ACL injury. The patient’s presentation may be confusing, and physical examination findings can be obscured by injury to other supporting structures of the knee. Arthroscopic assessment may likewise be misleading; the presence of an intact ACL can limit visualization of the PCL from the anterior approach,50,84,134 and the presence of an intact ligament of Humphrey can simulate an intact or injured PCL.34 With accuracies that approach 100% for complete tears,88 MRI is a reliable detector of PCL injury and a key contributor to clinical decision making.

Normal Anatomy The PCL is a central stabilizer of the knee and the primary restraint against posterior tibial translation. In addition, it restricts varus or valgus angulation and internal rotation of the tibia on the femur. The PCL originates from the lateral surface of the medial femoral condyle and inserts onto the posterior surface of the intercondylar region if the

FIG 63-32  ACL tear, bone contusion pattern. Sagittal PD-weighted FIG 63-31  Chronic ACL tear. Sagittal PD-weighted image shows absence of the ACL, which may be seen in the setting of chronic ACL tear with resorption of the torn fibers.

A

image with fat saturation shows bone contusion of the lateral condylopatellar sulcus (thin arrow) and impaction of posterior aspect of the lateral tibial plateau (short arrow). Note the cortical indentation of the posterior margin of the lateral tibial plateau.

B

FIG 63-33  ACL tear, associated meniscal injury. A, Sagittal PD-weighted image with fat saturation shows complex tear of the posterior horn of the medial meniscus (white arrow). B, Sagittal PD-weighted image shows peripheral vertical tear of the posterior horn of the lateral meniscus near the attachment of the ligament of Wrisberg (black arrow), which has been referred to as the Wrisberg rip.

CHAPTER 63  Knee

A

2201

B

FIG 63-34  Mucinous degeneration of the ACL. Sagittal PD-weighted images with fat saturation show A, mucinous degeneration of the ACL (arrow) and B, pericruciate ganglion cyst formation (arrow).

tibia below the level of the articular plateau. It is composed of an anterolateral bundle that becomes taut on flexion and a posteromedial bundle that becomes taut on extension. The PCL is wider, thicker, and stronger than the ACL.

Imaging Anatomy The PCL is best depicted on sagittal images, where it appears uniformly low in signal intensity, with a nearly horizontal takeoff from the femoral origin, followed by an abrupt descent at about 45 degrees to the tibia (Fig. 63-35). In up to two thirds of patients,54,78 the meniscofemoral ligaments of Humphrey and Wrisberg are seen on sagittal images as low-signal-intensity dots anterior and posterior to the PCL and should not be mistaken for displaced meniscal fragments or an intact PCL in the presence of PCL tear.

Trauma Tears of the PCL are relatively uncommon. About 50 ACL tears are imaged for each PCL tear. Most are incomplete and occur in the midportion of the ligament. Others involve the femoral and tibial attachments, where avulsion fracture may be present. In only a minority of patients, the PCL injury is isolated. A study of 47 patients with PCL injuries reported associated findings including joint effusions (64%) and injury to the ACL (17%), MCL (23%), lateral collateral ligament (LCL) (6%), medial meniscus (32%), lateral meniscus (30%), and bone marrow (36%).134

Mechanisms of Injury.  At least two common mechanisms of PCL injury have been identified. As in the case of ACL injuries, each pathomechanism may be accompanied by a predictable constellation of findings. A direct blow to the anterior tibia with the knee in flexion

FIG 63-35  Normal PCL. Sagittal PD-weighted image with fat saturation shows the normal course of the PCL (right arrows). The low-signalintensity “dot” anterior to the ligament represents Humphrey’s ligament in cross section (left arrow).

2202

PART II  CT and MR Imaging of the Whole Body

will forcefully displace the tibia posteriorly, causing a midsubstance tear of the PCL. These injuries may occur as a result of a fall onto a hyperflexed knee134,135 or in the context of a motor vehicle accident (the dashboard injury). Bone contusions may be seen at the site of contact in the anterolateral tibia and posterior lateral femoral condyle. Hyperextension of the knee may also result in PCL tear at the midsubstance or distal attachment. Continued extension may cause the ACL to rupture as well. Bone contusions are often seen at the site of contact in the anterior portion of the tibial articular surface and anterior aspect of the femoral condyles.

Imaging Appearance.  Injury to the PCL is reliably identified on sagittal MRI as focal discontinuity and signal hyperintensity (Figs. 63-36 and 63-37).88 Edema and hemorrhage from acute intrasubstance tears manifest as regions of increased signal intensity. Proximal or distal avulsion injuries will be evident with the presence of high-signalintensity ligamentous hemorrhage, subchondral marrow edema, intraarticular bone fragment with marrow signal intensity, and intraarticular hemorrhage in the intervening space. On conventional radiography or CT, PCL injury may be suggested by the recognition of secondary signs. These include avulsion fractures of the proximal fibula (arcuate sign) or periphery of the medial tibial plateau (reverse Segond fracture). Avulsion fracture of the head of the fibula most often occurs at the site of attachment of the arcuate complex. When this radiographic finding is encountered, there is an increased likelihood of concomitant injury of the PCL, and MRI should be strongly considered.62 Avulsion fracture of the periphery of the medial tibial plateau (reverse Segond fracture) may result  from avulsion of the deep fibers of the MCL. Similarly, MRI should be considered upon encountering this sign; there is an increased  likelihood of associated PCL disruption and peripheral medial meniscal tear.44

FIG 63-36  PCL tear. Sagittal PD-weighted image with fat saturation shows discontinuity of the distal PCL (arrow).

MEDIAL STABILIZING STRUCTURES OF THE KNEE Normal Anatomy The tibial or medial collateral ligament (MCL) is often conceptualized as a single ligamentous band extending from the medial femoral condyle to the medial tibial plateau. MRI and cadaver studies demonstrate that the medial knee anatomy is much more complicated, with the MCL actually blending with the adjacent structures in a complex configuration (Fig. 63-38). A three-layer approach has been suggested by Warren146 for conceptualizing the stabilizing structures of the medial knee and is supported by findings on both MRI and cadaver studies.27,82 The three layers, numbered from superficial to deep, make up the joint capsule and ligamentous structures, including the MCL. Together they function to restrain against valgus laxity during knee extension. The most superficial layer 1 consists of the deep crural fascia. It is continuous with the sartorius and superficial to the pes anserinus tendons (sartorius, semitendinosus, and gracilis) at its posterosuperior aspect. Anteriorly it fuses with layer 2 to form the medial patellar retinaculum. The intermediate layer 2 consists of the superficial layer of the MCL, a broad flat band 2 cm in width and 2 to 4 mm in thickness. These fibers extend from the medial epicondyle of the femur to the tibia, attaching about 2 cm below the joint line and then extending 3 to 4 cm more distally across the medial concavity of the proximal tibia, where fat and blood vessels are normally also present. Posteriorly, layer 2 fuses with layer 3 to form the posterior oblique ligament (POL). This ligament covers the posteromedial aspect of the femoral condyle, enveloping the semimembranosus tendon and attaching at three points—to the posterior portion of the medial meniscus via its prominent central (or tibial) arm, the posterior capsule via its superior (or capsular) arm, and the semimembranosus and tibia via its distal arm.82

FIG 63-37  PCL partial tear. Sagittal PD-weighted image with fat saturation shows intrasubstance increased signal (arrow), consistent with partial tear.

CHAPTER 63  Knee

A

2203

B

FIG 63-38  Normal medial collateral ligament complex. Contiguous coronal T2-weighted MRI demonstrates the A, superficial fibers (arrows) and B, deep fibers of the MCL complex. The deep layer of the MCL includes the meniscofemoral ligament (short white arrows) and the meniscotibial ligament (short black arrows).

The deep layer 3 consists of the joint capsule, the deep layer of the MCL including its meniscofemoral and meniscotibial (or coronary) extensions, and the patellomeniscal ligament. The deep fibers of the MCL attach loosely to the periphery of the body of the medial meniscus. Its thin bandlike superior extension constitutes the menisco­ femoral ligament. Its firmer shorter extension, attaching the medial meniscus to the tibia just below the joint line, constitutes the meniscotibial ligament. The MCL bursa and a small amount of fat are  interposed between the superficial and deep fibers of the MCL. The semimembranosus–tibial collateral ligament bursa is positioned between the semimembranosus tendon and the proximal MCL, and the pes anserine bursa is positioned between the pes anserinus and the distal MCL.

Trauma Given the degree of anatomic complexity, it is not surprising that injury to the medial knee rarely involves just one structure. A multitude of structures in the region—the menisci, medial retinaculum, POL, pes anserinus, and posteromedial joint capsule6—may also be injured.

Medial Collateral Ligament Injury.  Traumatic injuries of the MCL are common and range from intrasubstance tears (grade 1 sprain) to incomplete tears (grade 2 sprain) to complete disruption (grade 3 sprain). These are commonly seen in conjunction with tears of the medial meniscus and ACL as a result of valgus stress from clip injuries. Kissing bone contusions at the lateral femoral condyle and lateral tibial plateau may also be present. Other associated findings include lateral meniscal tear,124 joint effusions, and joint fluid extravasation. Imaging appearance.  The tibial collateral ligament is best seen on coronal images, where it appears as a homogeneously low-signal-

intensity structure on all pulse sequences. Moderately increased signal intensity may be seen between the superficial and deep fibers and below the superficial fibers at the distal tibial attachment site, where fat is normally interposed. A grade 1 sprain appears as increased signal intensity along the ligament secondary to edema. These changes are best seen on spin-density and T2-weighted images (Fig. 63-39). When there is focal irregularity, fraying, or discontinuity of only some of the fibers, a grade 2 sprain is diagnosed (Fig. 63-40). A grade 3 sprain is diagnosed when there is complete disruption of the low-signal-intensity band with a wavy or serpiginous contour of the disrupted fibers (Fig. 63-41). The ligament may be stripped from its femoral and tibial attachments. Two findings on conventional radiography are predictive of underlying MCL injury. Pellegrini-Stieda disease, or calcification within the soft tissue along the periphery of the medial femoral condyle, may be the consequence of chronic partial or complete tears of the MCL.91 The finding may alternatively be attributed to prior trauma to the ischiocondylar portion of the adductor magnus muscle. The reverse Segond fracture is an avulsion fracture of the periphery of the medial tibial plateau at the attachment of the deep fibers of the MCL. MCL avulsion fracture at this specific location is also associated with PCL disruption and peripheral medial meniscal tear.44 Either of these two radiographic findings should raise the possibility of MCL injury and prompt consideration of additional MRI to confirm the presence of MCL injury and check for the presence of concomitant injury.

Posterior Oblique Ligament Injury.  The POL is an important contributor to the static and dynamic stability of the knee.63 This fact as well as the propensity for the POL to become injured has been underappreciated.128 Its compromise is determined clinically by the detection of instability on knee extension. Because POL compromise usually

2204

PART II  CT and MR Imaging of the Whole Body occurs alongside injuries to the MCL, it is of key importance for the radiologist to recognize its presence on preoperative imaging. Addressing one without the other may leave the patient with persistent knee instability.6 Imaging appearance.  Evaluation of the POL is best performed on coronal and axial MRI. A coronal oblique plane along the superoinferior course of the POL has also been suggested for optimal imaging.6 Tearing is most commonly seen at the proximal aspect of the ligament.

LATERAL STABILIZING STRUCTURES OF THE KNEE Posterolateral Corner Normal Anatomy.  The posterolateral corner of the knee is composed of ligaments, muscles, and tendons that are static restraints to posterior translation, varus angulation, and external rotation of the knee.  Collectively referred to as the posterolateral (or arcuate) ligament complex, they include the fibular ligament or LCL, popliteus muscle and tendon, popliteofibular ligament, biceps femoris tendon, fabellofibular ligament, arcuate ligament, and lateral head of the gastrocnemius (Fig. 63-42). Independently the LCL restrains against varus angulation and external rotation. This extracapsular tubular cord arises at the lateral femoral epicondyle above the popliteal groove and just anterior to the origin of the lateral head of the gastrocnemius. It extends inferiorly and posteriorly where it joins with the biceps femoris insertion. These two tendons form the conjoined tendon and attach onto the lateral aspect of the fibular head. In contrast to the MCL, the LCL is separate from the lateral joint capsule, with the lateral geniculate veins and arteries interposed in between.

FIG 63-39  Medial collateral ligament grade 1 sprain. Coronal PDweighted image with fat saturation shows increased signal superficial to an otherwise intact MCL (arrows).

A

B

FIG 63-40  Medial collateral ligament grade 2 sprain. A, Coronal PD-weighted image with fat saturation shows increased signal superficial to the MCL fibers, which are somewhat indistinct proximally (white arrows). B, Axial image confirms discontinuity of the anterior margin of the ligament (black arrow) along the indistinct interval, with a normal appearance of the posterior fibers.

CHAPTER 63  Knee

A

2205

B

FIG 63-41  Medial collateral ligament grade 3 sprain. A, Coronal PD-weighted image with fat saturation shows complete discontinuity of the proximal fibers of the medial collateral ligament (white arrows). B, Axial image at the level of the discontinuity confirms absence of intact fibers (black arrow) at the level of the discontinuity.

The popliteus tendon is primarily a lateral stabilizer of the knee; secondarily it functions in internal rotation of the tibia. This intracapsular structure originates within the popliteal groove of the lateral femoral condyle just anterior and inferior to the proximal attachment of the LCL. It courses inferiorly and medially over its sulcus in the lateral femoral condyle and then meets its muscle belly posteromedially and deep to the arcuate ligament. It attaches to the posteromedial surface of the proximal tibia. Two important ligamentous structures emerge from the popliteus tendon. The popliteal fibular ligament extends from the popliteus tendon just proximal to the myotendinous junction to its distal attachment at the styloid process of the fibula. The site of attachment is medial to the attachment of the conjoined tendon. This ligament is one of the primary stabilizers of the knee and also functions to restrain against excessive posterior translation and varus angulation.80 The popliteal meniscal ligament extends from the posterolateral aspect of the popliteus tendon to the lateral meniscus. Upon knee extension, it restrains against forward meniscal displacement. The arcuate ligament is a Y-shaped thickening of the knee joint capsule and has medial and lateral limbs. It emanates superiorly from its attachment at the styloid process of the fibula to the level of the popliteus muscle. The medial limb is arcuate; it fans out over the popliteus muscle, blends with fibers of the oblique popliteal ligament, and attaches proximally to the posterior knee capsule. The lateral limb ascends along the deep portion of the capsule to its proximal attachment at the lateral femoral condyle. In the 20% of individuals who possess a fabella, the fabellofibular ligament extends from the fabella to the styloid process of the fibula. In the remainder, the fabellofibular ligament extends farther superiorly to attach to the posterolateral surface of the lateral femoral condyle.

Trauma.  Posterolateral corner injury, though less common than medial structure injury, may be more disabling. These injuries may occur as a result of hyperextension of an internally rotated knee or varus force onto an externally rotated knee.80 They are seen in the setting of sports activity, traffic accidents, or falls14 and manifest clinically as posterolateral rotatory instability, which may be subtle or underappreciated on physical examination. Disruption of posterolateral structures is rarely an isolated lesion.94 It usually occurs in association with injury to other supporting knee structures, such as the cruciate ligaments, the menisci, the MCL, and the anteromedial femoral and tibial condyles.10,22,113 Not surprisingly the broad spectrum and complicated nature of posterolateral knee injury can make its diagnosis challenging. If missed, these injuries can result in chronic knee instability, failure of subsequent cruciate ligament reconstruction, or osteoarthritis.23,46,64,101 Imaging appearance.  MRI can depict the structures of the posterolateral corner of the knee with varying degrees of success (see Fig. 63-42). LCL and popliteus tendon injuries are reliably seen on MRI. The other static support structures of the posterolateral corner of   the knee—including the arcuate, fabellofibular, and popliteofibular ligaments—are sometimes not well depicted on standard spin echo MRI in the sagittal, coronal, and axial planes.98 The use of coronal oblique images oriented along the plane of the popliteus tendon has been suggested for improved imaging of these structures.154 MRI of a posterolateral corner injury parallels the ligamentous and tendinous morphologic or signal-intensity changes seen elsewhere in the setting of injury. Indirect signs of abnormality such as focal edema can be relied on to suggest more subtle underlying injury. Their presence should prompt closer imaging inspection of the nearby tendons and ligaments for the presence of a tear. The common associated

2206

PART II  CT and MR Imaging of the Whole Body

A

B

C

D FIG 63-42  Posterolateral corner anatomy. A, Coronal T2-weighted image with fat saturation demonstrates a part of the popliteus tendon just proximal to the myotendinous junction coursing inferomedially (black arrow). The popliteofibular ligament is seen extending from the popliteus tendon toward its distal attachment at the fibular styloid process (white arrow). B, Sagittal PD-weighted image shows the fabellofibular ligament (white arrow) and the peroneal nerve (black arrows). C, T1-weighted coronal image shows the fabellofibular ligament (arrow). D, Sagittal PD-weighted image demonstrates the obliquely coursing fibers of the arcuate ligament (arrow).

structural injuries to the cruciate ligaments, menisci, or bone are usually well demonstrated. Findings on conventional radiography can be relied on as secondary signs of underlying posterolateral corner injury. The arcuate sign (avulsion fracture of the fibular head) is considered pathognomonic for posterolateral corner injury. This finding is attributed to one of two discrete foci of injury: (1) avulsion fracture at the distal attachment of the popliteal fibular, arcuate, and fabellofibular ligaments onto the fibular styloid process or (2) avulsion fracture at the distal attachment of conjoined tendon onto the lateral aspect of the fibular head.80 The presence of this lesion is also highly predictive of the presence of additional injury to the PCL. The Segond fracture, or avulsion of the lateral aspect of the lateral tibial plateau, is a classic radiographic entity that

is also highly suggestive of posterolateral corner injury as well as an underlying ACL tear.17,36,38,42,52,150 The presence of either the arcuate sign or the Segond fracture on radiography should prompt further investigation into the underlying injury using MRI. LCL injury.  Injuries to the LCL may occur in isolation or in association with other injuries of the posterolateral corner and the cruciate ligaments. They usually result from severe trauma involving varus force. Hemorrhage and edema with or without disruption of the ligament may be present. On coronal MRI, the LCL is well depicted as a thick, uniformly lowsignal-intensity cord on all pulse sequences (Fig. 63-43). As in the  case of MCL tears, the presence of edema along an otherwise intact ligament represents a low-grade sprain (Fig. 63-44). Midsubstance

CHAPTER 63  Knee

A

2207

B

FIG 63-43  Lateral and anterolateral corner normal anatomy. A, T1-weighted coronal image shows the lateral collateral ligament (white arrow) and biceps femoris tendon (black arrow) joining to form the conjoined tendon, which attaches onto the fibular head. B, T1-weighted coronal image shows the iliotibial band attaching onto Gerdy’s tubercle of the proximal tibia (white arrow).

signal change reflects edema and hemorrhage present in a partial  tear. Complete disruption is suggested by focal discontinuity and a serpiginous contour. The distal LCL and biceps femoris are commonly injured as a unit at the level of the conjoined tendon at the fibular head attachment. Isolated injury the distal biceps femoris is less common  (Fig. 63-45). Popliteus injury.  Popliteal tendon and muscle injuries are rarely isolated lesions148 and are usually associated with PCL tears, medial meniscus tears, bone bruises, or fractures.16,94 They occur in the setting of forced lateral rotation or varus stress. Most tears occur at the myotendinous junction,16 but intraarticular tearing of the popliteus tendon or avulsion at the femoral origin may also occur.14,16 Popliteus tendon injury is best imaged on sagittal and axial MRI. Focal enlargement of the tendon may be present as the result of hemorrhage or tearing.136 Partial tearing at the myotendinous junction will manifest as a feathery signal abnormality of the muscle belly, signifying edema and hemorrhage that extends into the tendon.94

Anterolateral Corner

Normal Anatomy.  The structures of the anterolateral corner of the knee include the iliotibial band (ITB), the anterior oblique band (AOB) of the LCL, and the capsular ligament (see Fig. 63-43). Together these structures contribute to anterior and posterolateral stability. The ITB is an extension of the tensor fasciae latae with distal attachments to the lateral femoral condyle, lateral patella, and Gerdy’s tubercle at the anterolateral aspect of the tibia. Just proximal to the attachment at Gerdy’s tubercle, the distal ITB fibers interdigitate with extensions  of the LCL. These extensions are referred to as the AOB.68 This

configuration allows forces that are applied to the anterolateral knee to become transmitted across both the ITB and the LCL.

Trauma.  Injury to anterolateral corner structures most commonly occurs in the setting of varus stress with internal rotation. These injuries are much more commonly encountered than those of the posterolateral corner. They include tears, disruption, and avulsion injury involving the ITB. Each of these entities is associated with ACL injury. Iliotibial band injury.  ITB injury is the most common injury to the anterolateral corner. Tearing most often involves the posterior fibers that interdigitate with the AOB. These are usually well demonstrated on coronal MRI. Avulsion fractures at the ITB distal attachments may also occur. Segond fracture.  The Segond fracture, or avulsion of the anterolateral aspect of the lateral tibial plateau, is a radiographic entity that signifies underlying trauma to the anterolateral corner structures of the knee. Although it was originally postulated to arise as a result of avulsion at the attachment site of the lateral capsular ligament, a recent MR study suggests that the Segond fracture actually occurs at the distal attachments of the iliotibial tract or AOB of the LCL.17 These lesions, like all anterolateral corner injuries, are highly suggestive of the presence of an underlying ACL tear (75%-100%), meniscal tear (66%75%), or posterolateral corner injury.17,36,38,42,52,150 Iliotibial band friction syndrome.  ITB friction syndrome (ITBFS) refers to lateral knee pain arising from the friction that occurs as the ITB passes over the lateral femoral epicondyle during repetitive flexion and extension of the knee. This injury is seen most often in individuals

2208

PART II  CT and MR Imaging of the Whole Body

who participate in long-distance running and cycling. On MRI, illdefined signal abnormality signifying soft tissue edema is present within the fat deep to the ITB (Fig. 63-46).97,100 Occasionally a more well-defined fluid collection, perhaps signifying the formation of an adventitious bursa, is seen medial to the iliotibial tract.

EXTENSOR MECHANISM Normal Anatomy The extensor mechanism of the knee includes the quadriceps muscle and tendon, the patellar tendon, and the components of the patellofemoral joint (Fig. 63-47). Damage to the patellofemoral articulation may arise from malalignment and tracking problems. Meanwhile, damage to the quadriceps tendon and patellar ligament may be caused by acute trauma or chronic overuse. In patients with systemic illness (e.g., diabetes, hyperparathyroidism, gout, collagen-vascular diseases), injury may be secondary to underlying structural weakness.9

Quadriceps Tendon.  The rectus femoris, vastus intermedius, vastus medialis, and vastus lateralis muscles converge to insert as the quadriceps tendon at the superior pole of the patella (see Fig. 63-47). The fasciae of these four individual muscles combine to varying degrees to form a multilayered structure. The overall thickness of the quadriceps tendon is 8 ± 2 mm. The transverse width of the tendon averages 35 ± 7 mm. Usually the vastus medialis and vastus lateralis merge, producing the central component of a three-layered quadriceps tendon, with the fascia from the rectus femoris seen anteriorly and the fascia from the vastus intermedius seen posteriorly.133,155 It is also common for the fasciae of the vastus medialis and lateralis muscles to combine with each other and with the fascia of either the rectus femoris or the vastus intermedius muscle, producing a twolayered tendon. This laminar structure is easily appreciated on axial and sagittal MRI scans.

Patellar Tendon.  The patellar tendon, alternatively referred to as the

FIG 63-44  Sprain of the lateral collateral ligament. Coronal PD-weighted image with fat saturation shows signal hyperintensity along the distal aspect of the lateral collateral ligament (arrow), which is otherwise intact.

A

patellar ligament, is technically a continuation of the fibers of the rectus femoris that course over the patella.153 It extends from the distal pole of the patella to the anterior tibial tubercle. Its sagittal dimension is  3.7 ± 1.2 mm proximally, 4.3 ± 1.1 mm at the midportion, and 5.6 ± 1 mm distally. The patellar tendon is a fairly uniformly low-signalintensity structure seen best on sagittal images (Fig. 63-48). Intermediate signal within the tendon may normally be present at the patellar attachment. Magic-angle artifact is not uncommonly seen as transverse

B

FIG 63-45  Coronal (A) and sagittal (B) PD-weighted images with fat saturation show tear of the distal biceps femoris tendon (arrows).

CHAPTER 63  Knee

2209

FIG 63-46  Iliotibial band friction syndrome. Coronal PD-weighted

FIG 63-48  Normal patellar tendon. Sagittal PD-weighted image with

image with fat saturation reveals ill-defined signal hyperintensity (white arrow) tracking deep to the intact iliotibial band (black arrow) at the level of the lateral femoral condyle.

fat saturation shows the taut patellar tendon (long arrow) with homogeneously low signal through nearly its entire course. Focal hyperintensity in a linear configuration (short arrow) at the patellar attachment is a normal finding.

intermediate signal bands within the patellar tendon on sagittal MRI with short TE (Fig. 63-49).

Patellofemoral Joint.  At 0 degrees of flexion, the patellofemoral joint is not congruent with the patella that is laterally displaced. With increasing flexion, the medial and lateral retinacula tighten and the patella descends toward the midline. Eventually, congruence of the patella and the trochlea is achieved. In a normal knee, this state of patellofemoral congruence is reached with a lesser degree of flexion than in an abnormal knee; there remains then a wide range of what is considered normal. Additional factors contributing to patellofemoral malalignment include structural abnormalities (e.g., misshapen patella), hypoplastic femoral trochlear sulcus, and abnormally high position of the patella itself.125,133 Patellar tracking abnormalities can be imaged with axial CT or MRI of the patellofemoral joints performed during incremental knee flexion. Both modalities show the relationship of the patella to the femoral trochlear sulcus during the initial 30 degrees of flexion, when most pathology is evident. MRI better defines the medial and lateral retinacula, although CT may image these structures as well. The patellar cartilage itself is better seen with MRI. CT arthrography may be helpful in evaluating gross cartilage defects.

Trauma

Quadriceps Tendon Tears.  Injury to the quadriceps tendon often FIG 63-47  Normal distal quadriceps tendon. Sagittal T1-weighted image shows the distal quadriceps tendon (arrow). Sagittal images obtained centrally within the tendon typically demonstrate this trilaminar appearance.

occurs as a result of either repetitive microtrauma or underlying illness. Acute injury may also occur in the setting of rapid deceleration from running with the foot planted. These injuries are often difficult to detect clinically. MRI is useful in grading tears of the distal

2210

A

PART II  CT and MR Imaging of the Whole Body

B

FIG 63-49  Normal patellar tendon, magic-angle artifact. A, Sagittal PD-weighted image with fat saturation shows discrete bands of intermediate signal (arrows) within the patellar tendon. The extensive signal abnormality superficial to the tendon may lead the novice interpreter to the wrong diagnosis. B, Sagittal T2-weighted image shows a normal patellar tendon.

quadriceps tendon, which usually occur distally at or near the patellar attachment (Fig. 63-50).9,133 Attention to the layers described previously is paramount because partial ruptures of the quadriceps tendon tend to involve one of these fascial groups. Full-thickness tears are more obvious (Fig. 63-51).155

Patellar Tendon Injury.  Patellar tendon injury ranges from tendinosis due to chronic overuse, to partial tearing, to complete disruption resulting from acute macrotrauma. A minority of patellar tendon injuries are complete tendon disruption resulting from acute trauma in the setting of sports-related activity. This most commonly occurs near the patellar attachment133 and is best seen on sagittal MRI as discontinuity of the patellar tendon with buckling of the tendon, retraction of the patella, and fluid signal in the intervening gap (Figs. 63-52 and 63-53). Partial-thickness tears are more common and are also most commonly seen at the patellar attachment (see Fig. 63-53).

Patellar Tendinosis (Jumper’s Knee).  Most patellar tendon injuries are the result of repetitive forced extension of the knee as seen in the context of basketball, volleyball, soccer, tennis, and track—hence the name jumper’s knee.70 This form of injury manifests as thickening and partial tearing of the proximal patellar tendon and is well seen on sagittal MRI as increased tendon thickness greater than 7 mm or as intrasubstance high signal intensity (Fig. 63-54).70,133 These same changes, however, can be seen in both symptomatic and asymptomatic individuals.87 In symptomatic patients, the findings may be accompanied by patellar tracking disorders in as many as 45% of patients.1

FIG 63-50  Quadriceps tendon rupture. Sagittal PD-weighted image with fat saturation shows complete discontinuity of the distal quadriceps tendon (short arrow) at the patellar attachment. Fluid extends from the knee joint space through the quadriceps defect and into the prepatellar soft tissues. The patellar tendon is buckled but intact (long arrow).

Hoffa’s fat pad impingement.  Patellar tendon–lateral femoral condyle friction syndrome21 refers to the constellation of MRI findings that likely corresponds to the clinical entity known as fat-pad impingement syndrome. Affected patients are both athletes and nonathletes who experience chronic anterior knee pain that localizes to the inferior patellar pole on physical examination. The etiology is unknown, but it is likely related to friction onto the fat interposed between the patellar ligament and the lateral femoral condyle due to abnormal patellar tracking or alignment. MRI findings include abnormal obliteration of the fat planes, abnormal signal, and enhancement within the lateral soft tissues of the inferior patellofemoral joint (Fig. 63-55A). Most patients demonstrate patellar alignment abnormality such as patella alta or lateral subluxation. Patients typically respond well to conservative treatment, which usually entails taping the superior pole of the patella. Lateral patellofemoral dislocation.  Transient lateral patellofemoral dislocation is often diagnosed on the basis of a characteristic constellation of findings: kissing bone contusions at the medial patella and lateral femoral condyle (as the patella reduces) (see Fig. 63-56), edema along the medial patellar retinaculum, and disruption of the medial patellofemoral ligament (Figs. 63-56 to 63-59). Osteochondral injury with displaced chondral or osteochondral fragments may also result, most frequently at the medial patellar facet (see Fig. 63-59).133

BONE AND BONE MARROW Imaging Anatomy On MRI the cortical bone of the distal femur, proximal tibia, and patella is of low signal intensity on all pulse sequences. The signal

CHAPTER 63  Knee

A

2211

B

FIG 63-51  Quadriceps tendon partial tear. Sagittal (A) and axial (B) PD-weighted MRI with fat saturation shows focal increased signal intensity within the distal quadriceps tendon, indicative of partial tearing (arrows).

*

FIG 63-52  Patellar tendon disruption. Sagittal PD-weighted fatsuppressed MRI shows complete disruption of the patellar tendon at its midportion (arrow). There is extensive reactive fluid in the prepatellar soft tissues, including within the prepatellar bursa (*).

FIG 63-53  Patellar tendon partial tear. Sagittal PD-weighted MRI with fat saturation shows focal signal hyperintensity of the patellar tendon at the patellar attachment (arrow). The reactive signal change in the adjacent portion of Hoffa’s fat pad seen here is often helpful in confirming the diagnosis in more subtle cases.

2212

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 63-55  Fat pad impingement. A, Patellar tendon–lateral femoral FIG 63-54  Patellar tendinosis. Sagittal PD-weighted MRI with fat saturation shows marked diffuse thickening of the patellar tendon (arrow).

condyle friction syndrome. Sagittal PD-weighted image with fat saturation demonstrates ill-defined signal hyperintensity (arrow) within the superolateral aspect of Hoffa’s fat pad. B, Quadriceps fat pad impingement. Sagittal PD-weighted image with fat saturation demonstrates diffuse signal hyperintensity and swelling of the quadriceps fat pad.

*

*

*

FIG 63-57  Patellar dislocation, medial patellofemoral ligament injury.

FIG 63-56  Patellar dislocation, bone contusion pattern. Axial PDweighted image with fat saturation shows osteochondral fracture of the patella involving the median ridge (black *) and one contusion of the anterior aspect of the lateral femoral condyle (white *). The medial patellofemoral ligament is also torn at the patellar attachment (arrow).

Axial PD-weighted image with fat saturation shows disruption of the medial patellofemoral ligament near the femoral attachment at the adductor tubercle (arrows).

CHAPTER 63  Knee

2213

B

A

FIG 63-58  Patellar dislocation, medial patellofemoral ligament injury. Axial PD-weighted images with fat saturation show injury of the medial patellofemoral ligament at the patellar attachment. A, The more cephalad image demonstrates disruption of the medial patellofemoral ligament at the patellar attachment (white arrows). B, The more caudal demonstrates osteochondral avulsion fracture of the patella (black arrow) at the MPFL attachment.

*

* B

A

FIG 63-59  Patellar dislocation, osteochondral injury of patella. Axial PD-weighted images with fat saturation in different patients show A, chondral defect extending from median ridge to lateral patellar facet (white arrows) and B, osteochondral fracture of medial patellar facet (black arrows). The superior margin of the femoral trochlea is noted to be shallow on this image (*). Bone contusion of the anterior aspect of the lateral femoral condyle is also noted.

intensity of the medullary canal, meanwhile, changes in reflection of the changing amounts of hematopoietic and fatty bone marrow with aging; over time, the fatty marrow increases in proportion. Older patients may have residual islands of hematopoietic marrow that should not be confused with infiltrative disease. The epiphyseal portions of the distal femur, proximal tibia, and epiphysioid patella always contain predominantly fatty marrow. The sensitivity of MRI to bone marrow alterations makes it ideal for detecting traumatic injury.

Trauma Essentially five types of bone injury occur about the knee: • Bone bruise or contusion • Stress fracture • Osteochondral fracture • Tibial plateau fracture • Classic avulsion fractures of the tibial plateau, femoral condyles, and patella • Bone infarction

2214

PART II  CT and MR Imaging of the Whole Body

Of these, only classic fractures are reliably diagnosed on plain film imaging. Even then, certain features (e.g., extension to the articular surface) might not be obvious without three-dimensional (3D) imaging. CT and MRI are therefore required (1) for diagnosis of subtle osseous lesions that are impossible to detect on conventional radiographic imaging and (2) to define more clearly the extent of osseous injury and the integrity of the additional supporting structures of  the knee when osseous abnormality has already been identified on plain film.

Bone Bruises.  A bone bruise represents an acute occult microfracture of subchondral trabecular bone, with subsequent hemorrhage, edema, and hyperemia. The cortex and articular cartilage are intact. These injuries may be secondary to either direct trauma or traction associated with twisting injuries and internal derangement of the knee. They usually resolve in 6 to 8 weeks. These injuries, which were not apparent before MRI became available, appear as ill-defined subchondral reticulated or stellate regions of low signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Five classic contusion patterns in the knee have been described120: 1. Hyperextension injury of the knee, as occurs with a direct blow to the anterior tibia onto the planted foot, results in “kissing contusions” of the anterior aspect of the proximal tibia and the distal femur. This finding on MRI (Fig. 63-60) is suggestive of additional trauma to the PCL, ACL, or menisci. 2. The dashboard injury, as the name suggests, occurs in passengers in automobiles who undergo traumatic impact of the anterior proximal tibia with the dashboard while their knee is in flexion. An identical injury pattern can result from a fall onto a bent knee. This injury will manifest on MRI as a bone contusion at the site of

impact. A significant association has been demonstrated with injury to the PCL. 3. The pivot shift injury occurs in skiers and American football players as a result of rapid deceleration with simultaneous change in direction. The tibia undergoes external rotation with the knee in flexion. On MRI, bone bruises are seen in the posterolateral tibial plateau and midportion of the lateral femoral condyle (see Fig. 63-32). This pattern is highly suggestive of an ACL tear, as detailed earlier. Osteochondral impaction of the lateral condylopatellar sulcus is also commonly seen in this setting (Fig. 63-61). 4. The clip injury results from pure valgus stress onto the partially flexed knee joint. Classically this injury is seen in American football players who experience a blow to the lateral knee with the foot planted. It may also occur in a pedestrian struck by a car. The direct impact to the lateral knee will result in bone contusion at the lateral femoral condyle and lateral tibial plateau. The distractive forces at the medial joint compartment may result in coexisting MCL and medial meniscal tears (the O’Donoghue triad). A smaller contusion in the medial femoral condyle at the MCL attachment may also be present as a result of the avulsive stress. Small contusion at the posterior lip of the medial tibial plateau has also been described and attributed to contrecoup impaction as the knee relocates following the ACL tear.69 5. Lateral patellar dislocation occurs in the setting of internal rotation of the femur with the knee in flexion. The resultant contraction of the quadriceps tendon in this configuration causes the patella to dislocate laterally. Bone contusions occur at the inferomedial patella and the anterior aspect of the lateral femoral condyle (see Fig. 63-57), presumably from the contact that occurs on relocation.

FIG 63-61  Osteochondral impaction. Sagittal PD-weighted image with FIG 63-60  Hyperextension injury. Sagittal PD-weighted image with fat saturation demonstrates bone contusion at the anterior aspect of the lateral tibial plateau.

fat saturation shows osteochondral impaction of the lateral condylopatellar sulcus (arrow) with deepening of the notch present normally. The pattern shown here is the prototypical bone marrow edema pattern seen in the setting of ACL tear due to pivot shift injury. This finding may even be evident on plain radiography in the lateral projection.

CHAPTER 63  Knee

A

2215

B

FIG 63-62  Subchondral insufficiency fractures. A, Coronal PD-weighted image with fat suppression shows extensive subchondral edema with discrete low-signal fracture line paralleling the subchondral bone plate. Note the high-grade thinning of the articular cartilage of the medial compartment, particularly the tibial plateau. B, Coronal PD-weighted image with fat suppression demonstrates identical findings in the medial tibial plateau in a younger patient.

Stress Fractures.  Stress fractures can result from either normal

Osteochondral Fractures.  Osteochondral fractures refer to articular

stress placed on deficient bone, as would occur in the osteoporotic elderly individual (insufficiency fracture), or excessive stress placed on normal bone, as would occur in the running, jumping, or hurdling athlete (fatigue fracture). Radiography is unreliable for detecting these lesions, particularly in the early stages. MRI, with a reported sensitivity and specificity as high as 88% and 100%, respectively, is the modality of choice.48 On MRI, stress fractures appear as linear low-signal-intensity bands away from the joint on T1- and T2-weighted images (Fig. 63-62). Increased signal intensity may be seen about the impacted fracture site on T2-weighted images, reflecting the associated edema. Although not nearly as sensitive as MRI, CT is also excellent in demonstrating cortical stress injuries. These appear as focal areas of cortical osteopenia, resorption cavities, or striation.48 Spontaneous osteonecrosis of the knee is an idiopathic disorder of middle-aged and older adults that usually affects the weightbearing portion of the medial femoral condyle. It less frequently affects the medial tibial plateau, lateral femoral condyle, and lateral tibial plateau. Associated osteoarthritis and degenerative tearing of the medial  meniscus characteristically precede the necrotic event.15 Radiography is normal for approximately 4 to 6 weeks. MRI, however, usually  demonstrates a poorly defined zone of decreased signal intensity on T1-weighted images and increased signal intensity on fluid-sensitive sequences. A poorly defined ring of moderately high signal intensity may be noted about the lesion (Fig. 63-63). Eventually the edema progresses to fracture and collapse of the subchondral bone, with fragmentation of the articular cartilage. Osteoarthritis then may be secondary to the necrosis or may exist coincidentally before the necrotic event.12 Current opinion is that this entity, which was originally described before the use of MRI, actually represents subchondral insufficiency fracture.151

surface fractures in which the fracture line parallels the articular surface. These must be distinguished from classic fractures in which the fracture line extends intraarticularly and thus perpendicular to an osteochondral fracture line. Acute osteochondral fractures may be displaced or impacted (see Fig. 63-62). Displaced osteochondral fragments may be elevated partially or completely dislodged from the parent bone into the joint cavity. They are best seen on T2-weighted images, especially in the presence of a joint effusion. These fractures can be seen at the medial patellar facet and the anterolateral femoral condyle secondary to patellar femoral dislocation. Impacted osteochondral lesions are commonly seen at the anterolateral femoral condyle. Decreased signal intensity is seen on T1-weighted images at the impaction site. Increased signal intensity is seen on T2-weighted images as a result of edema. The overlying cartilage may be intact. Chronic osteochondral fracture, or osteochondritis dissecans, classically occurs in the lateral aspect of the medial femoral condyle. It is well visualized on MRI, which is useful in both identifying and grading these lesions (see Fig. 63-63). Generally speaking, the presence of a break in the articular cartilage signifies an unstable lesion and calls for surgical intervention. Otherwise, conservative treatment usually suffices. A high-signal-intensity line at the interface between the osteochondral fragment and parent bone on T2-weighted images represents strong evidence for an unstable fragment.28 De Smet and colleagues identified additional MRI findings that suggest an unstable lesion. In decreasing order of frequency, they included a focal osteochondral defect, articular fracture, and a cyst deep to the lesion.29 Intraarticular injection of gadolinium contrast may improve diagnostic accuracy in grading osteochondral lesions. A bipartite patella is an occasionally encountered entity that may be misinterpreted as an unhealed chronic osteochondral fracture of the patella (Fig. 63-64). It results from failure of fusion of the patellar

2216

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 63-63  In situ osteochondritis dissecans. Coronal T1-weighted image (A) and coronal PD-weighted image with fat suppression (B) show fragmentation of the lateral aspect of the medial femoral condyle. There is irregularity of the overlying chondral surface and reactive marrow signal hyperintensity at the subchondral interface. However, no displacement of the osteochondral fragment is seen.

A

B

FIG 63-64  Bipartite patella. Coronal T1-weighted image (A) and axial PD-weighted image with fat suppression (B) show fragmentation of the superolateral aspect of the patella. This represents the most common type of bipartite patella. Note the intact overlying cartilage and the absence of edema or cystic change along the synchondrosis.

CHAPTER 63  Knee

2217

FIG 63-65  Nondisplaced tibial eminence avulsion fracture. Coronal PD-weighted image with fat suppression demonstrates nondisplaced avulsion fracture of the tibial eminence at the distal ACL insertion (arrows).

FIG 63-66  Bone infarction. Coronal T1-weighted image shows charac-

ossification centers. These may become symptomatic when there  is instability at the synchondrosis, which manifests on MRI as fluid  signal at the synchondrosis and cystic change at the apposing bony margins.

Tibial Plateau Fractures.  Tibial plateau fractures result from extreme valgus force at the knee and represent a common type of knee injury. Both CT and MRI can define the fracture, evaluate the fragment position, and grade the degree of depression. Because osseous detail is better seen with CT, tibial plateau fractures are often studied with this modality, complemented with 2D and 3D reconstructions. Fracture fragmentation, displacement, and depression may be seen more clearly. MRI, on the other hand, goes beyond the assessment of osseous injury and allows imaging of the menisci and the cruciate and collateral ligaments.

Avulsion Fractures.  Several types of avulsion fractures about the knee have been described: the Segond fracture, the reverse Segond fracture, the arcuate sign, ITB avulsion fractures, and fracture of the tibial eminence at the distal ACL insertion (Fig. 63-65). Each has a characteristic appearance on conventional radiography, and each should alert the imager to the presence of underlying structural abnormality. Predictable patterns of underlying tendinous, ligamentous, and osseous injury are often identified on subsequent MRI.

Bone Infarction.  Bone infarction may be seen anywhere in the knee in the setting of generalized osteonecrosis—for instance, in the setting of chronic steroid use, sickle cell anemia, or rheumatologic disorders. Although involvement of the distal femur and proximal tibia is commonly seen, involvement of the patella is relatively unusual and will almost exclusively involve the proximal pole. The serpiginous signal change, as seen elsewhere in the skeletal system, is pathognomonic (Fig. 63-66).

teristic serpiginous low signal throughout distal femur and proximal tibia, which is pathognomonic for bone infarction.

MUSCLES AND TENDONS Ten muscles and their respective tendons cross the knee joint. The anterior quadriceps represents the rectus femoris, vastus medialis, vastus intermedius, and vastus lateralis, which insert on the proximal pole of the patella. The anterolateral tensor fasciae latae and ITB insert at Gerdy’s tubercle at the anterolateral tibia. The posterolateral biceps femoris inserts with the fibular collateral ligament as the conjoined tendon at the fibular head. The posteromedial sartorius, gracilis, and semitendinosus insert at the posteromedial tibia, proximally at the pes anserinus. The posteromedial semimembranosus inserts separately, slightly more posteriorly. The medial and lateral heads of the gastrocnemius arise at the posterior supracondylar region of the femur. The popliteal muscle belly rests at the posterior proximal tibia, its tendon coursing intraarticularly through the posterior horn of the lateral meniscus toward the popliteal groove of the femur. The plantaris muscle originates at the posterolateral supracondylar femur, and its tendon, the longest in the body, courses toward the ankle.

KNEE BURSAE Bursae are synovium-lined potential spaces that are typically interposed between ligaments (or tendons) and the underlying bone to reduce friction. Numerous bursae are present about the knee.7 They are of significance to the imager because the mass effect, inflammation, hemorrhage, or rupture may account for a patient’s symptoms. Chronic friction, trauma, or internal derangement of the knee joint can result in bursitis, which fills the bursa with fluid and is complicated by infection, hemorrhage, or rupture.

2218

PART II  CT and MR Imaging of the Whole Body

FIG 63-67  Prepatellar bursitis. Sagittal PD-weighted image with fat

FIG 63-68  MCL bursa. Coronal PD-weighted image with fat suppres-

suppression demonstrates fluid-distended prepatellar bursa with a thick irregular capsule and extensive reactive edema in surrounding prepatellar soft tissues.

sion demonstrates fluid-filled bursa interposed between the periphery of a torn body of the medial meniscus and the superficial fibers of the MCL complex.

Normal Anatomy Four major bursae are located within the anterior aspect of the knee. The suprapatellar bursa is positioned between the quadriceps tendon and the distal femur. Fluid within this space is often seen in the setting of trauma or synovitis. The prepatellar bursa is positioned superficial to the patella within the subcutaneous soft tissue. Inflammation may result from friction in the setting of occupational kneeling—hence the term housemaid’s knee (Fig. 63-67). The pretibial bursa is positioned superficial to the tibial tubercle, where direct trauma or chronic overuse may result in inflammation. Pretibial bursitis is less common than prepatellar bursitis. The deep infrapatellar bursa is positioned deep to the patellar tendon’s distal attachment onto the tibial tubercle. Inflammation may result from repetitive knee extension in the context of running or jumping. Its presence can be distinguished from jumper’s knee on the basis of the absence of signal abnormality within the tendon itself. The main bursa within the posterior aspect of the knee emerges superficially from between the medial head of the gastrocnemius and the semimembranosus—the gastrocnemius-semimembranosus bursa. Fluid within this bursa, referred to as a popliteal cyst, most likely arises from continuity with the joint space. This entity may also be classified as a synovial cyst and is described in detail later in this chapter. Two bursae can be found within the lateral aspect of the knee. The ITB bursa is positioned between the distal ITB and the underlying tibia; the LCL–biceps femoris bursa is positioned between the distal biceps femoris and the FCL. ITB bursitis, usually seen in runners, sometimes mimics ITB friction syndrome or iliotibial tendonitis.

Three bursae are present within the medial aspect of the knee. The semimembranosus–tibial collateral ligament bursa is positioned between the semimembranosus tendon and the proximal MCL, the pes anserine bursa is positioned between the pes anserinus and the distal MCL, and the medial collateral ligament bursa is positioned between the superficial and deep fibers of the medial collateral ligament (Fig. 63-68).25 Pes anserine bursitis is seen most commonly in runners who experience medial knee pain. Because of its location, this condition must be distinguished from a meniscal cyst or a synovial cyst.

Imaging Bursae are not usually seen on MRI unless they are inflamed and fluid filled. Bursitis will manifest as a fluid collection within the soft tissues of the knee. The presence of internal heterogeneity or debris makes the presence of superimposed infection or hemorrhage more likely. Rupture is suggested when fluid extends beyond the expected bursa location and tracks between fascial layers.

CYSTIC LESIONS OF THE KNEE The common types and locations of the cystic lesions of the knee are important to recognize because they too can be the reason for a patient’s complaint. These lesions include synovial cysts, ganglion cysts, and meniscal cysts. Before any resection is done, the cystic nature of such a knee mass must be determined, the relationship to the joint cavity must be evaluated, and the presence of any associated intraarticular disease must be identified. MRI is best suited to study these issues. Generally, cystic lesions on MRI are well circumscribed with smooth walls.

CHAPTER 63  Knee

2219

* A

B FIG 63-69  Popliteal cyst. A, Axial PD-weighted image with fat suppression shows fluid within a posterior outpouching of the knee joint. Note its relationship to the semimembranosus tendon (arrow) and the medial head of the gastrocnemius tendon and muscle (*). B, Axial and coronal images demonstrate fluid (arrows) insinuating within fascial planes inferior to ruptured popliteal cyst, consistent with leak.

Synovial Cysts Synovial cysts are juxtaarticular fluid collections lined by synovial cells. They are best conceptualized as focal extensions of the joint capsule that may or may not demonstrate continuity with the joint space.  The most commonly encountered synovial cyst of the knee is the popliteal (or Baker’s) cyst, a fluid-filled cavity positioned within the posteromedial soft tissues of the knee. This cyst arises as an extension of knee joint fluid into the normally occurring gastrocnemiussemimembranosus bursa. MRI demonstrates a posteromedial fluid collection with a neck positioned between the medial head of the gastrocnemius and the semimembranosus tendon (Fig. 63-69A). Visible continuity with the joint space may or may not be identifiable. In nearly all cases, a joint effusion is present owing to the presence of underlying internal derangement or degeneration. Most likely the popliteal cyst arises as a sequela of primary knee joint pathology. Rupture is suggested when fluid extends beyond the bursal walls and tracks between fascial layers (see Fig. 63-69B).

Ganglion Cysts Ganglion cysts are benign cystic masses composed of viscous proteinaceous fluid that tend to arise adjacent to or within muscles and their tendon sheaths. Those occurring about the knee are usually found within the juxtaarticular soft tissue at the origin of the gastrocnemius66 or at the proximal tibiofibular joint or intraarticularly in close association with the tibial attachment of the cruciate ligaments (see Fig. 63-34). Ganglion cysts that arise at the proximal tibiofibular articulation are unique in that they can lead to compressive neuropathy of the common peroneal nerve. On MRI, features of ganglion cysts overlap those of synovial cysts, making this distinction sometimes difficult. The focus should instead be on the location and the relation of the cyst to the adjacent structures.7 Synovial cysts and ganglion cysts are easily distinguished from meniscal cysts by the absence of a direct communication with a meniscus.

Meniscal Cysts Meniscal cysts occur at the joint line, almost invariably in association with degenerative horizontal cleavage meniscal tears. Synovial fluid is forced through the tear and accumulates at the meniscal capsular junc-

tion. If the cyst is confined within the meniscus, it is termed intrameniscal. If the cyst extends into the adjacent soft tissue, it is termed perimeniscal. Medially, cysts are seen adjacent to the posterior horn. Extension of these cysts centrally to the tibial attachment of the PCL may mimic the appearance of a ganglion cyst. Extension of these cysts posteromedially may enlarge without restraint if they are posterior to the MCL and may be mistaken for popliteal cysts. Laterally, cysts are seen adjacent to the anterior horn. They appear to be contained by the LCL and the ITB and are therefore usually smaller. Meniscal cysts are easily distinguished from synovial or ganglion cysts on MRI because they will almost always communicate directly with a meniscal tear. Detection of the meniscal tear is essential because treatment of only the cyst itself will result in recurrence.

NONTRAUMATIC ABNORMALITY OF THE KNEE Nontraumatic conditions affecting the knee joint include proliferative disorders of the synovium, inflammatory arthritis, osteoarthritis, and hematopoietic and neoplastic diseases.

Pigmented Villonodular Synovitis Pigmented villonodular synovitis (PVNS) is a rare monoarticular proliferative disorder of the synovium that occurs in adults. The most common site of involvement is the knee. Although conventional radiographs may show associated pressure erosions of the underlying bone, most cases involving the knee demonstrate normal osseous structures at the time of clinical presentation. Soft tissue masses within the knee joint may be noted on plain films. MRI reveals a proliferative synovial mass throughout the confines of the joint capsule. The signal-intensity characteristics may vary as a result of the presence of abundant hemosiderin, inflamed synovium, and joint effusion. Certainly the presence of multiple areas of low signal intensity on all pulse sequences, especially at the periphery, is a clue to the abundant hemorrhage present (Fig. 63-70). The differential diagnosis includes other disorders that may be accompanied by hemarthrosis, such as hemophilia, synovial hemangioma, and inflammatory arthritis. PVNS may also present in the knee as a solitary lesion, which may demonstrate more masslike characteristics (Fig. 63-71). This entity is referred to as localized nodular synovitis.

2220

PART II  CT and MR Imaging of the Whole Body

B

A

D C

E

F

FIG 63-70  Pigmented villonodular synovitis. A, Lateral radiograph shows marked soft tissue prominence within the joint capsule, associated with osseous erosions. B, Axial CT scan shows mixed-attenuation fullness at the suprapatellar bursa. Sagittal PD-weighted (C and D) and T2-weighted (E and F) images show the extent of the mixed-signal-intensity mass within the joint. Persistent low-signal-intensity areas on both sequences, especially those at the capsular margins, indicate extensive hemosiderin deposition and a probable diagnosis of PVNS.

CHAPTER 63  Knee

2221

FIG 63-71  Localized nodular synovitis. Sagittal PD-weighted image with fat suppression and axial short tau inversion recovery image demonstrate rounded mass with heterogeneously low to intermediate signal interposed between Hoffa’s fat pad and the femoral trochlea. These imaging characteristics lead to a differential diagnosis of localized nodular synovitis or intraarticular chondroma.

Synovial Chondromatosis Synovial chondromatosis is the other major proliferative disorder of the synovium. Like PVNS, it is a rare monoarticular disease seen principally in adults. Cartilaginous metaplasia of the synovium leads to the production of multiple cartilaginous bodies within the confines of the joint. The knee is most commonly involved. As in PVNS, osseous erosions may be present. Although PVNS virtually never calcifies, cartilaginous bodies are prone to such calcification and ossification. The radiographic appearance of multiple calcified and ossified bodies of similar sizes distributed throughout the confines of the joint capsules is diagnostic of the disorder. When these bodies are not calcified, MRI detects them and is often diagnostic (Fig. 63-72).

Osteoarthritis Osteoarthritis may be found coincidentally during the search for other disorders. Osteophyte formation, subchondral sclerosis, subchondral cyst formation, and cartilage loss can all be seen. Degeneration of the menisci is independent of degeneration of the femorotibial compartments. It usually precedes the development of osteoarthritis, although it may accelerate the process. Therefore associated degenerative horizontal cleavage tears of the menisci may be seen. Such tears represent extension of internal meniscal disruption to the free edge, inferior articular surface, or capsular margin. With advancing osteoarthritis, what remains of the menisci may become adherent to the marginal osteophytes and eventually extrude beyond the margins of the joint. Lipoma arborescens is a somewhat rare disease of the synovium that is typically seen in the setting of severe degenerative joint disease. It is a benign condition characterized by diffuse villous hypertrophy of the synovium. Although cases have been described in other joints, involvement of the knee is the most common. Patients experience painful boggy swelling of the joint, usually in the region of the suprapatellar recess.72 The MRI appearance is diagnostic. Frondlike hypertrophy of the synovium mimicking fat signal is seen. This includes signal hyperintensity of T1-weighted images, with dropoff in signal on fatsuppressed sequences (Fig. 63-73).

FIG 63-72  Synovial chondromatosis. Sagittal PD-weighted image with fat suppression shows innumerable monomorphic intraarticular bodies throughout the joint.

Hematopoietic and Neoplastic Diseases As mentioned, MRI is exquisitely sensitive to changes in bone marrow, and hence myeloproliferative disorders, neoplasm, marrow reconversion, and marrow failure all can be imaged well. Processes affecting  the marrow itself can be imaged long before trabecular bony changes

2222

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 63-73  Lipoma arborescens. Sagittal T2-weighted (A) and T1-weighted (B) images show thick frondlike lipomatous proliferation of the synovium, with characteristic T1 hyperintensity.

T2-weighed images, with lobulated contours (Fig. 63-74) and enhancement on postcontrast images.

REFERENCES

FIG 63-74  Enchondroma. Coronal PD-weighted image with fat suppression demonstrates rounded hyperintense bone lesion with lobulated contours, which are characteristic.

are seen. Such changes are necessary before CT can be of value. CT, however, remains helpful in evaluating any lesion characterized by calcification or ossification. Enchondromas are the most commonly encountered benign neoplasm about the knee, usually seen in the distal femur. Radiographs will show a partially calcified well-circumscribed bone lesion without aggressive features. On MRI they appear heterogeneously bright on

1. Allen GM, Tauro PG, Ostlere SJ: Proximal patellar tendinosis and abnormalities of patellar tracking. Skeletal Radiol 28:220–223, 1999. 2. Anderson MW, Raghavan N, Seidenwurm DJ, et al: Evaluation of meniscal tears: Fast spin-echo versus conventional spin-echo magnetic resonance imaging. Acad Radiol 2:209–214, 1995. 3. Applegate GR, Flannigan BD, Tolin BS, et al: MR diagnosis of recurrent tears in the knee: Value of intraarticular contrast material. AJR Am J Roentgenol 161:821–825, 1993. 4. Araki Y, Yamamoto H, Nakamura H, et al: MR diagnosis of discoid lateral menisci of the knee. Eur J Radiol 18:92–95, 1994. 5. Arnoczky SP, Warren RF, Kaplan N: Meniscal remodeling following partial meniscectomy—An experimental study in the dog. Arthroscopy 1:247–252, 1985. 6. Beall DP, Googe JD, Moss JT, et al: Magnetic resonance imaging of the collateral ligaments and the anatomic quadrants of the knee. Magn Reson Imaging Clin N Am 15:53–72, 2007. 7. Beaman FD, Peterson JJ: MR Imaging of cysts, ganglia, and bursae about the knee. Magn Reson Imaging Clin N Am 15:39–52, 2007. 8. Beltran J: MRI musculoskeletal system, New York, 1990, Gower Medical. 9. Bencardino JT, Rosenberg ZS, Brown RR, et al: Traumatic musculotendinous injuries of the knee: Diagnosis with MR imaging. Radiographics 20 Spec No:S103–S120, 2000. 10. Bennett DL, George MJ, El-Khoury GY, et al: Anterior rim tibial plateau fractures and posterolateral corner knee injury. Emerg Radiol 10:76–83, 2003. 11. Berg EE: The meniscal ossicle: The consequence of a meniscal avulsion. Arthroscopy 7:241–243, 1991. 12. Bjorkengren AG, AlRowaih A, Lindstrand A, et al: Spontaneous osteonecrosis of the knee: Value of MR imaging in determining prognosis. AJR Am J Roentgenol 154:331–336, 1990.

CHAPTER 63  Knee 13. Boden SD, Davis DO, Dina TS, et al: A prospective and blinded investigation of magnetic resonance imaging of the knee: Abnormal findings in asymptomatic subjects. Clin Orthop Relat Res (282):177– 185, 1992. 14. Bolog N, Hodler J: MR imaging of the posterolateral corner of the knee. Skeletal Radiol 36(8):715–728, 2007. 15. Brahme SK, Fox JM, Ferkel RD, et al: Osteonecrosis of the knee after arthroscopic surgery: Diagnosis with MR imaging. Radiology 178:851– 853, 1991. 16. Brown TR, Quinn SF, Wensel JP, et al: Diagnosis of popliteus injuries with MR imaging. Skeletal Radiol 24:511–514, 1995. 17. Campos JC, Chung CB, Lektrakul N, et al: Pathogenesis of the Segond fracture: Anatomic and MR imaging evidence of an iliotibial tract or anterior oblique band avulsion. Radiology 219:381–386, 2001. 18. Chan WP, Peterfy C, Fritz RC, et al: MR diagnosis of complete tears of the anterior cruciate ligament of the knee: Importance of anterior subluxation of the tibia. AJR Am J Roentgenol 162:355–360, 1994. 19. Cheung LP, Li KC, Hollett MD, et al: Meniscal tears of the knee: Accuracy of detection with fast spin-echo MR imaging and arthroscopic correlation in 293 patients. Radiology 203:508–512, 1997. 20. Chiu SS: The anterior tibial translocation sign. Radiology 239:914–915, 2006. 21. Chung CB, Skaf A, Roger B, et al: Patellar tendon-lateral femoral condyle friction syndrome: MR imaging in 42 patients. Skeletal Radiol 30:694–697, 2001. 22. Cohen AP, King D, Gibbon AJ: Impingement fracture of the anteromedial tibial margin: A radiographic sign of combined posterolateral complex and posterior cruciate ligament disruption. Skeletal Radiol 30:114–116, 2001. 23. Covey DC: Injuries of the posterolateral corner of the knee. J Bone Joint Surg Am 83-A:106–118, 2001. 24. Davis SJ, Teresi LM, Bradley WG, et al: The “notch” sign: Meniscal contour deformities as indicators of tear in MR imaging of the knee.   J Comput Assist Tomogr 14:975–980, 1990. 25. De MM, Shahabpour M, Van RF, et al: MR imaging of the medial collateral ligament bursa: Findings in patients and anatomic data derived from cadavers. AJR Am J Roentgenol 177:911–917, 2001. 26. De MM, Shahabpour M, Vanderdood K, et al: Medial meniscocapsular separation: MR imaging criteria and diagnostic pitfalls. Eur J Radiol 41:242–252, 2002. 27. De MM, Van RF, Lenchik L, et al: Three layers of the medial capsular and supporting structures of the knee: MR imaging–anatomic correlation. Radiographics 20 Spec No:S83–S89, 2000. 28. De Smet AA, Fisher DR, Graf BK, et al: Osteochondritis dissecans of the knee: Value of MR imaging in determining lesion stability and the presence of articular cartilage defects. AJR Am J Roentgenol 155:549– 553, 1990. 29. De Smet AA, Ilahi OA, Graf BK: Reassessment of the MR criteria for stability of osteochondritis dissecans in the knee and ankle. Skeletal Radiol 25:159–163, 1996. 30. De Smet AA, Norris MA, Yandow DR, et al: MR diagnosis of meniscal tears of the knee: Importance of high signal in the meniscus that extends to the surface. AJR Am J Roentgenol 161:101–107, 1993. 31. De Smet AA, Tuite MJ: Use of the “two-slice-touch” rule for the MRI diagnosis of meniscal tears. AJR Am J Roentgenol 187:911–914, 2006. 32. Dehaven KE: Rationale for meniscus repair or excision. Clin Sports Med 4:267–273, 1985. 33. Dehaven KE: Meniscus repair. Am J Sports Med 27:242–250, 1999. 34. Dehaven KE: Arthroscopy in the cruciate-injured knee. In Feagin JA, editor: The cruciate ligament, New York, 2007, Churchill, p 334. 35. Delfaut EM, Beltran J, Johnson G, et al: Fat suppression in MR imaging: Techniques and pitfalls. Radiographics 19:373–382, 1999. 36. Delzell PB, Schils JP, Recht MP: Subtle fractures about the knee: Innocuous-appearing yet indicative of significant internal derangement. AJR Am J Roentgenol 167:699–703, 1996. 37. Deutsch AL, Mink JH, Fox JM, et al: Peripheral meniscal tears: MR findings after conservative treatment or arthroscopic repair. Radiology 176:485–488, 1990.

2223

38. Dietz GW, Wilcox DM, Montgomery JB: Segond tibial condyle fracture: Lateral capsular ligament avulsion. Radiology 159:467–469, 1986. 39. Dillon EH, Pope CF, Jokl P, et al: The clinical significance of stage 2 meniscal abnormalities on magnetic resonance knee images. Magn Reson Imaging 8:411–415, 1990. 40. Drosos GI, Pozo JL: The causes and mechanisms of meniscal injuries in the sporting and non-sporting environment in an unselected population. Knee 11:143–149, 2004. 41. Eggli S, Wegmuller H, Kosina J, et al: Long-term results of arthroscopic meniscal repair: An analysis of isolated tears. Am J Sports Med 23:715–720, 1995. 42. El-Khoury GY, Daniel WW, Kathol MH: Acute and chronic avulsive injuries. Radiol Clin North Am 35:747–766, 1997. 43. Escobedo EM, Hunter JC, Zink-Brody GC, et al: Usefulness of turbo spin-echo MR imaging in the evaluation of meniscal tears: Comparison with a conventional spin-echo sequence. AJR Am J Roentgenol 167:1223–1227, 1996. 44. Escobedo EM, Mills WJ, Hunter JC: The “reverse Segond” fracture: Association with a tear of the posterior cruciate ligament and medial meniscus. AJR Am J Roentgenol 178:979–983, 2002. 45. Fayad LM, Parellada JA, Parker L, et al: MR imaging of anterior cruciate ligament tears: Is there a gender gap? Skeletal Radiol 32:639–646, 2003. 46. Fleming RE, Jr, Blatz DJ, McCarroll JR: Posterior problems in the knee: Posterior cruciate insufficiency and posterolateral rotatory insufficiency. Am J Sports Med 9:107–113, 1981. 47. Fox MG: MR Imaging of the meniscus: Review, current trends, and clinical implications. Magn Reson Imaging Clin N Am 15:103–123, 2007. 48. Gaeta M, Minutoli F, Scribano E, et al: CT and MR imaging findings in athletes with early tibial stress injuries: Comparison with bone scintigraphy findings and emphasis on cortical abnormalities. Radiology 235:553–561, 2005. 49. Gentili A, Seeger LL, Yao L, et al: Anterior cruciate ligament tear: Indirect signs at MR imaging. Radiology 193:835–840, 1994. 50. Gillquist J, Hagberg G, Oretorp N: Arthroscopy in acute injuries of the knee joint. Acta Orthop Scand 48:190–196, 1977. 51. Gold GE, Busse RF, Beehler C, et al: Isotropic MRI of the knee with 3D fast spin-echo extended echo-train acquisition (XETA): Initial experience. Am J Roentgenol 188:1287–1293, 2007. 52. Goldman AB, Pavlov H, Rubenstein D: The Segond fracture of the proximal tibia: A small avulsion that reflects major ligamentous damage. AJR Am J Roentgenol 151:1163–1167, 1988. 53. Graf B, Docter T, Clancy W, Jr: Arthroscopic meniscal repair. Clin Sports Med 6:525–536, 1987. 54. Gupte CM, Bull AM, Thomas RD, et al: A review of the function and biomechanics of the meniscofemoral ligaments. Arthroscopy 19:161– 171, 2003. 55. Hamberg P, Gillquist J, Lysholm J: Suture of new and old peripheral meniscus tears. J Bone Joint Surg Am 65:193–197, 1983. 56. Haramati N, Staron RB, Rubin S, et al: The flipped meniscus sign. Skeletal Radiol 22:273–277, 1993. 57. Harper KW, Helms CA, Lambert HS, III, et al: Radial meniscal tears: Significance, incidence, and MR appearance. AJR Am J Roentgenol 185:1429–1434, 2005. 58. Hauger O, Frank LR, Boutin RD, et al: Characterization of the “red zone” of knee meniscus: MR imaging and histologic correlation. Radiology 217:193–200, 2000. 59. Helms CA: The meniscus: Recent advances in MR imaging of the knee. AJR Am J Roentgenol 179:1115–1122, 2002. 60. Helms CA, Laorr A, Cannon WD, Jr: The absent bow tie sign in bucket-handle tears of the menisci in the knee. AJR Am J Roentgenol 170:57–61, 1998. 61. Herman LJ, Beltran J: Pitfalls in MR imaging of the knee. Radiology 167:775–781, 1988. 62. Huang GS, Yu JS, Munshi M, et al: Avulsion fracture of the head of the fibula (the “arcuate” sign): MR imaging findings predictive of injuries to the posterolateral ligaments and posterior cruciate ligament. AJR Am J Roentgenol 180:381–387, 2003.

2224

PART II  CT and MR Imaging of the Whole Body

63. Hughston JC, Eilers AF: The role of the posterior oblique ligament in repairs of acute medial (collateral) ligament tears of the knee. J Bone Joint Surg Am 55:923–940, 1973. 64. Hughston JC, Jacobson KE: Chronic posterolateral rotatory instability of the knee. J Bone Joint Surg Am 67:351–359, 1985. 65. Huston LJ, Greenfield ML, Wojtys EM: Anterior cruciate ligament injuries in the female athlete: Potential risk factors. Clin Orthop Relat Res (372):50–63, 2000. 66. James SL, Connell DA, Bell J, et al: Ganglion cysts at the gastrocnemius origin: A series of ten cases. Skeletal Radiol 36:139–143, 2007. 67. Jee WH, McCauley TR, Kim JM, et al: Meniscal tear configurations: Categorization with MR imaging. AJR Am J Roentgenol 180:93–97, 2003. 68. Johnson LL: Lateral capsular ligament complex: Anatomical and surgical considerations. Am J Sports Med 7:156–160, 1979. 69. Kaplan PA, Gehl RH, Dussault RG, et al: Bone contusions of the posterior lip of the medial tibial plateau (contrecoup injury) and associated internal derangements of the knee at MR imaging. Radiology 211:747–753, 1999. 70. Khan KM, Bonar F, Desmond PM, et al: Patellar tendinosis (jumper’s knee): Findings at histopathologic examination, US, and MR imaging: Victorian Institute of Sport Tendon Study Group. Radiology 200:821– 827, 1996. 71. Reference deleted in proofs. 72. Kloen P, Keel SB, Chandler HP, et al: Lipoma arborescens of the knee. J Bone Joint Surg Br 80(2):298–301, 1998. 73. Kneeland JB, Shimakawa A, Wehrli FW: Effect of intersection spacing on MR image contrast and study time. Radiology 158:819–822, 1986. 74. Kornick J, Trefelner E, McCarthy S, et al: Meniscal abnormalities in   the asymptomatic population at MR imaging. Radiology 177:463–465, 1990. 75. Kosaka M, Nakase J, Toratani T, et al: Oblique coronal and oblique sagittal MRI for diagnosis of anterior cruciate ligament tears and evaluation of anterior cruciate ligament remnant tissue. Knee 21(1):54–57, 2014. 76. Kowalchuk RM, Kneeland JB, Dalinka MK, et al: MRI of the knee: Value of short echo time fast spin-echo using high performance gradients versus conventional spin-echo imaging for the detection of meniscal tears. Skeletal Radiol 29:520–524, 2000. 77. Lal NR, Jamadar DA, Doi K, et al: Evaluation of bone contusions with fat-saturated fast spin-echo proton-density magnetic resonance imaging. Can Assoc Radiol J 51:182–185, 2000. 78. Lee BY, Jee WH, Kim JM, et al: Incidence and significance of demonstrating the meniscofemoral ligament on MRI. Br J Radiol 73:271–274, 2000. 79. Lee SY, Matsui N, Yoshida K, et al: Magnetic resonance delineation of the anterior cruciate ligament of the knee: Flexed knee position within a surface coil. Clin Imaging 29:117–122, 2005. 80. Lee J, Papakonstantinou O, Brookenthal KR, et al: Arcuate sign of posterolateral knee injuries: Anatomic, radiographic, and MR   imaging data related to patterns of injury. Skeletal Radiol 32:619–627, 2003. 81. Lee K, Siegel MJ, Lau DM, et al: Anterior cruciate ligament tears: MR imaging-based diagnosis in a pediatric population. Radiology 213:697–704, 1999. 82. Loredo R, Hodler J, Pedowitz R, et al: Posteromedial corner of the knee: MR imaging with gross anatomic correlation. Skeletal Radiol 28:305– 311, 1999. 83. Lynch MA, Henning CE, Glick KR, Jr: Knee joint surface changes. Long-term follow-up meniscus tear treatment in stable anterior cruciate ligament reconstructions. Clin Orthop Relat Res (172):148–153, 1983. 84. Lysholm J, Gillquist J: Arthroscopic examination of the posterior cruciate ligament. J Bone Joint Surg Am 63:363–366, 1981. 85. Magee T, Williams D: Detection of meniscal tears and marrow lesions using coronal MRI. AJR Am J Roentgenol 183:1469–1473, 2004. 86. Magee T, Williams D: 3.0-T MRI of meniscal tears. Am J Roentgenol 187:371–375, 2006.

87. Major NM, Helms CA: MR imaging of the knee: Findings in asymptomatic collegiate basketball players. AJR Am J Roentgenol 179:641–644, 2002. 88. Margheritini F, Mariani PP: Diagnostic evaluation of posterior cruciate ligament injuries. Knee Surg Sports Traumatol Arthrosc 11:282–288, 2003. 89. Martinoli C, Bianchi S, Spadola L, et al: Multimodality imaging assessment of meniscal ossicle. Skeletal Radiol 29:481–484, 2000. 90. McCauley TR: MR imaging evaluation of the postoperative knee. Radiology 234:53–61, 2005. 91. Mendes LF, Pretterklieber ML, Cho JH, et al: Pellegrini-Stieda disease: A heterogeneous disorder not synonymous with ossification/calcification of the tibial collateral ligament—Anatomic and imaging investigation. Skeletal Radiol 35:916–922, 2006. 92. Metcalf MH, Barrett GR: Prospective evaluation of 1485 meniscal tear patterns in patients with stable knees. Am J Sports Med 32(3):675–680, 2004. 93. Miller DB, Jr: Arthroscopic meniscus repair. Am J Sports Med 16:315–320, 1988. 94. Miller TT, Gladden P, Staron RB, et al: Posterolateral stabilizers of the knee: Anatomy and injuries assessed with MR imaging. AJR Am J Roentgenol 169:1641–1647, 1997. 95. Moore SL: Imaging the anterior cruciate ligament. Orthop Clin North Am 33:663–674, 2002. 96. Moser MW, Dugas J, Hartzell J, et al: A hypermobile Wrisberg variant lateral discoid meniscus seen on MRI. Clin Orthop Relat Res 456:264– 267, 2007. 97. Muhle C, Ahn JM, Yeh L, et al: Iliotibial band friction syndrome: MR imaging findings in 16 patients and MR arthrographic study of six cadaveric knees. Radiology 212:103–110, 1999. 98. Munshi M, Pretterklieber ML, Kwak S, et al: MR imaging, MR arthrography, and specimen correlation of the posterolateral corner of the knee: An anatomic study. AJR Am J Roentgenol 180:1095–1101, 2003. 99. Newman AP, Daniels AU, Burks RT: Principles and decision making in meniscal surgery. Arthroscopy 9:33–51, 1993. 100. Nishimura G, Yamato M, Tamai K, et al: MR findings in iliotibial band syndrome. Skeletal Radiol 26:533–537, 1997. 101. O’Brien SJ, Warren RF, Pavlov H, et al: Reconstruction of the chronically insufficient anterior cruciate ligament with the central   third of the patellar ligament. J Bone Joint Surg Am 73:278–286, 1991. 102. Park JS, Ryu KN, Yoon KH: Meniscal flounce on knee MRI: Correlation with meniscal locations after positional changes. AJR Am J Roentgenol 187:364–370, 2006. 103. Peterfy CG, Janzen DL, Tirman PF, et al: “Magic-angle” phenomenon: A cause of increased signal in the normal lateral meniscus on short-TE MR images of the knee. AJR Am J Roentgenol 163:149–154, 1994. 104. Prince JS, Laor T, Bean JA: MRI of anterior cruciate ligament injuries and associated findings in the pediatric knee: Changes with skeletal maturation. AJR Am J Roentgenol 185:756–762, 2005. 105. Rak KM, Gillogly SD, Schaefer RA, et al: Anterior cruciate ligament reconstruction: Evaluation with MR imaging. Radiology 178:553–556, 1991. 106. Ramnath RR, Magee T, Wasudev N, et al: Accuracy of 3-T MRI using fast spin-echo technique to detect meniscal tears of the knee. Am J Roentgenol 187:221–225, 2006. 107. Rath E, Richmond JC: The menisci: Basic science and advances in treatment. Br J Sports Med 34:252–257, 2000. 108. Recht MP, Piraino DW, Applegate G, et al: Complications after anterior cruciate ligament reconstruction: Radio-graphic and MR findings. AJR Am J Roentgenol 167:705–710, 1996. 109. Recht MP, Piraino DW, Cohen MA, et al: Localized anterior arthrofibrosis (cyclops lesion) after reconstruction of the anterior cruciate ligament: MR imaging findings. AJR Am J Roentgenol 165:383–385, 1995. 110. Resnick D: Internal derangement of the joints, ed 4, Philadelphia, 2002, WB Saunders.

CHAPTER 63  Knee 111. Richmond J, Sarno RC: Posttrauma intracapsular bone fragments: Association with meniscal tears. AJR Am J Roentgenol 150:159–160, 1988. 112. Roberts CC, Towers JD, Spangehl MJ, et al: Advanced MR imaging of the cruciate ligaments. Magn Reson Imaging Clin N Am 15:73–86, 2007. 113. Ross G, Chapman AW, Newberg AR, et al: Magnetic resonance imaging for the evaluation of acute posterolateral complex injuries of the knee. Am J Sports Med 25:444–448, 1997. 114. Rubenstein JD, Kim JK, Morova-Protzner I, et al: Effects of collagen orientation on MR imaging characteristics of bovine articular cartilage. Radiology 188:219–226, 1993. 115. Rubin DA, Britton CA, Towers JD, et al: Are MR imaging signs of meniscocapsular separation valid? Radiology 201:829–836, 1996. 116. Rubin DA, Kneeland JB, Listerud J, et al: MR diagnosis of meniscal tears of the knee: Value of fast spin-echo vs conventional spin-echo pulse sequences. AJR Am J Roentgenol 162:1131–1135, 1994. 117. Rubin DA, Paletta GA, Jr: Current concepts and controversies in meniscal imaging. Magn Reson Imaging Clin N Am 8:243–270, 2000. 118. Ruff C, Weingardt JP, Russ PD, et al: MR imaging patterns of displaced meniscus injuries of the knee. AJR Am J Roentgenol 170:63–67, 1998. 119. Sanders TG, Linares RC, Lawhorn KW, et al: Oblique meniscomeniscal ligament: Another potential pitfall for a meniscal tear—Anatomic description and appearance at MR imaging in three cases. Radiology 213:213–216, 1999. 120. Sanders TG, Medynski MA, Feller JF, et al: Bone contusion patterns of the knee at MR imaging: Footprint of the mechanism of injury. Radiographics 20 Spec No:S135–S151, 2000. 121. Schnarkowski P, Tirman PF, Fuchigami KD, et al: Meniscal ossicle: Radiographic and MR imaging findings. Radiology 196:47–50, 1995. 122. Sciulli RL, Boutin RD, Brown RR, et al: Evaluation of the postoperative meniscus of the knee: A study comparing conventional arthrography, conventional MR imaging, MR arthrography with iodinated contrast material, and MR arthrography with gadolinium-based contrast material. Skeletal Radiol 28:508–514, 1999. 123. Shankman S, Beltran J, Melamed E, et al: Anterior horn of the lateral meniscus: Another potential pitfall in MR imaging of the knee. Radiology 204:181–184, 1997. 124. Shelbourne KD, Nitz PA: The O’Donoghue triad revisited: Combined knee injuries involving anterior cruciate and medial collateral ligament tears. Am J Sports Med 19:474–477, 1991. 125. Shellock FG, Mink JH, Deutsch AL, et al: Patellar tracking abnormalities: Clinical experience with kinematic MR imaging in 130 patients. Radiology 172:799–804, 1989. 126. Shogry ME, Pope TL, Jr: Vacuum phenomenon simulating meniscal or cartilaginous injury of the knee at MR imaging. Radiology 180:513–515, 1991. 127. Silverman JM, Mink JH, Deutsch AL: Discoid menisci of the knee: MR imaging appearance. Radiology 173:351–354, 1989. 128. Sims WF, Jacobson KE: The posteromedial corner of the knee: Medial-sided injury patterns revisited. Am J Sports Med 32:337–345, 2004. 129. Singh K, Helms CA, Jacobs MT, et al: MRI appearance of Wrisberg variant of discoid lateral meniscus. AJR Am J Roentgenol 187:384–387, 2006. 130. Singson RD, Feldman F, Staron R, et al: MR imaging of displaced bucket-handle tear of the medial meniscus. AJR Am J Roentgenol 156:121–124, 1991. 131. Smith DK, Totty WG: The knee after partial meniscectomy: MR imaging features. Radiology 176:141–144, 1990. 132. Soejima T, Murakami H, Tanaka N, et al: Anteromedial meniscofemoral ligament. Arthroscopy 19:90–95, 2003.

2225

133. Sonin AH, Fitzgerald SW, Bresler ME, et al: MR imaging appearance of the extensor mechanism of the knee: Functional anatomy and injury patterns. Radiographics 15:367–382, 1995. 134. Sonin AH, Fitzgerald SW, Friedman H, et al: Posterior cruciate ligament injury: MR imaging diagnosis and patterns of injury. Radiology 190:455–458, 1994. 135. Sonin AH, Fitzgerald SW, Hoff FL, et al: MR imaging of the posterior cruciate ligament: Normal, abnormal, and associated injury patterns. Radiographics 15:551–561, 1995. 136. Stoller DW: Magnetic resonance imaging in orthopaedics and sports medicine, ed 3, Baltimore, 2007, Lippincott Williams & Wilkins. 137. Stoller DW, Martin C, Crues JV, III, et al: Meniscal tears: Pathologic correlation with MR imaging. Radiology 163:731–735, 1987. 138. Stone KR: Current and future directions for meniscus repair and replacement. Clin Orthop Relat Res 367(Suppl):S273–S280, 1999. 139. Tachibana Y, Yamazaki Y, Ninomiya S: Discoid medial meniscus. Arthroscopy 19:E12–E18, 2003. 140. Tarhan NC, Chung CB, Mohana-Borges AV, et al: Meniscal tears: Role of axial MRI alone and in combination with other imaging planes. AJR Am J Roentgenol 183:9–15, 2004. 141. Toms AP, White LM, Marshall TJ, et al: Imaging the post-operative meniscus. Eur J Radiol 54:189–198, 2005. 142. Trattnig S, Huber M, Breitenseher MJ, et al: Imaging articular cartilage defects with 3D fat-suppressed echo planar imaging: Comparison with conventional 3D fat-suppressed gradient echo sequence and correlation with histology. J Comput Assist Tomogr 22:8–14, 1998. 143. Tuckman GA, Miller WJ, Remo JW, et al: Radial tears of the menisci: MR findings. AJR Am J Roentgenol 163:395–400, 1994. 144. Turner DA, Rapoport MI, Erwin WD, et al: Truncation artifact: A potential pitfall in MR imaging of the menisci of the knee. Radiology 179:629–633, 1991. 145. Vahey TN, Broome DR, Kayes KJ, et al: Acute and chronic tears of the anterior cruciate ligament: Differential features at MR imaging. Radiology 181:251–253, 1991. 146. Warren LF, Marshall JL: The supporting structures and layers on the medial side of the knee: An anatomical analysis. J Bone Joint Surg Am 61:56–62, 1979. 147. Weiss KL, Morehouse HT, Levy IM: Sagittal MR images of the knee:   A low-signal band parallel to the posterior cruciate ligament caused   by a displaced bucket-handle tear. AJR Am J Roentgenol 156:117–119, 1991. 148. Westrich GH, Hannafin JA, Potter HG: Isolated rupture and repair of the popliteus tendon. Arthroscopy 11:628–632, 1995. 149. White LM, Schweitzer ME, Johnson WJ, et al: The role of T2-weighted fast-spin-echo imaging in the diagnosis of meniscal tears. J Magn Reson Imaging 6:874–877, 1996. 150. Woods GW, Stanley RF, Tullos HS: Lateral capsular sign: X-ray clue to a significant knee instability. Am J Sports Med 7:27–33, 1979. 151. Yamamoto T, Bullough PG: Subchondral insufficiency fracture of the femoral head and medial femoral condyle. Skeletal Radiol 29:40–44, 2000. 152. Yu JS, Cosgarea AJ, Kaeding CC, et al: Meniscal flounce MR imaging. Radiology 203:513–515, 1997. 153. Yu JS, Petersilge C, Sartoris DJ, et al: MR imaging of injuries of the extensor mechanism of the knee. Radiographics 14:541–551, 1994. 154. Yu JS, Salonen DC, Hodler J, et al: Posterolateral aspect of the knee: Improved MR imaging with a coronal oblique technique. Radiology 198:199–204, 1996. 155. Zeiss J, Saddemi SR, Ebraheim NA: MR imaging of the quadriceps tendon: Normal layered configuration and its importance in cases of tendon rupture. AJR Am J Roentgenol 159:1031–1034, 1992.

64  Foot and Ankle Christos Kosmas, Ken L. Schreibman, and Mark R. Robbin

This chapter is intended to serve as a practical approach to imaging the ankle and foot using computed tomography (CT) and magnetic resonance imaging (MRI). We include a review of the anatomic structures and common pathologic processes as well as an overview of the CT and MRI techniques used. Finally we include references for those readers who wish to explore topics in more detail.

ANATOMY Tarsal Bones Gross Anatomy of the Tarsal Bones16

Talus.  Figures 64-1 through 64-4 are photographs of cadaveric bones arranged to illustrate the relationships of the major tarsal bones and joints. Figure 64-1 represents the bones of the ankle and foot. Central to all this is the talus, labeled Ta. (The label abbreviations in Fig. 64-1 will be consistent throughout all figures.) The word talus is Latin for “ankle,” indicating that early anatomists considered the talus the center of the ankle. Understanding the articulations between the talus and the surrounding bones is key to understanding the anatomy of the ankle and foot. Two views of the talus are shown in Figure 64-2. The dome is the broad curved articular surface on the top of the talus. (The specimen in Fig. 64-2 has an osteochondral lesion centrally in the medial edge of the talar dome. Osteochondral lesions are discussed later in the chapter.) The head is the rounded process at the anterior aspect of the talus, and it articulates with the navicular bone. The body of the talus comprises everything between the dome and head. The dome of the talus along with the distal ends of the tibia and fibula make up the ankle joint (Fig. 64-3). (Ankle joint is the preferred name of this joint in the radiology and orthopedic surgery literature, rather than “tibiotalar joint” or “crural joint.”) Mortise.  The flat talar dome articulates with the flat surface at the distal end of the tibia known as the plafond. Plafond is an architectural term meaning “a ceiling formed by the underside of a floor.” In essence the plafond is the ceiling of the ankle joint, formed by the floor of the tibia. The ankle joint is bounded on the sides by the inner articular surfaces of the medial and lateral malleoli. The plafond and malleoli together form a rectangular opening called the mortise into which the talar dome fits, analogous to a mortise-and-tenon joint in woodworking. The ankle mortise is a remarkably sturdy joint. Like the hip and knee joints, the ankle must bear our entire body weight with every step. But although it is common for primary osteoarthritis to affect the hips and knees of many of us as we age, it is uncommon to have primary osteoarthritis of the ankle.

2226

The joint between the distal tibia and fibula is called the syndesmosis. Syndesmosis is a Greek term meaning “to bind together,” and in general a syndesmotic joint is held together by thick connective ligaments. Most joints in the body, including the ankle and subtalar joints, are synovial joints enclosed by a synovium-lined capsule that creates synovial fluid. The distal fibula, just above the lateral malleolus, fits into a shallow groove in the adjacent tibia. Subtalar joint.  The talus articulates with the tibia superiorly and with the navicular anteriorly. Inferiorly it articulates with the calcaneus in the complex subtalar joint, a term that is preferred over “talocalcaneal joint.” Figure 64-4 illustrates the three facets that make up the subtalar joint. In Figure 64-4A to D, the talus and calcaneus were attached using colored modeling clay. In Figure 64-4E, the two bones have been disarticulated and the talus flipped over, displaying the talar and calcaneal articular surfaces of the posterior, middle, and anterior facets of the subtalar joint in red, blue, and green, respectively. The posterior facet is the largest and is the primary weightbearing portion of the subtalar joint. At the anterolateral corner of the posterior facet, the talus comes to an acutely angled corner, the lateral process of the talus. When the subtalar joint experiences an extreme axial load, such as when a person falls from a height or undergoes a deceleration injury in a motor vehicle collision, the pointy lateral process of the talus acts like a wedge, splitting and fracturing the calcaneus.13 Calcaneal fractures tend to extend into the posterior facet, and when imaging calcaneal fractures we obliquely angle our coronally reformatted CT slices to be perpendicular to the posterior facet. The middle facet is defined by the sustentaculum tali, a shelflike projection from the anteromedial portion of the calcaneus that supports the middle of the talus. Sustentaculum in Latin means “a supporting structure.” The flexor hallucis longus tendon passes under the sustentaculum tali. The middle facet of the subtalar joint is a completely separate articulation from the posterior facet. Across the middle facet of the subtalar joint is one of the two most common locations for tarsal coalitions to occur, the other being between the anterior process of the calcaneus and the lateral pole of the navicular. Unlike the posterior and middle facets, the anterior facet is not well defined and may even be absent. When present, the anterior facet is a smooth continuation of the middle facet, extending under the head of the talus. Directly lateral to the anterior and middle facet is the sinus tarsi, an area devoid of bone and filled primarily with fat.

Anatomic Divisions.  Figure 64-5 is a three-dimensionally (3D) reformatted CT image showing the anatomic divisions between the tarsals and metatarsals. The hindfoot consists of the talus and calcaneus

CHAPTER 64  Foot and Ankle

2227

Dome T OL

Ti

Fi OLT

Body

Ta ST

Ca Head

N

FIG 64-2  Gross anatomy of the talus as viewed from the top and

Cu 1 I

2 II

medial sides. The green arrows show an osteochondral lesion of the talus (OLT) in the medial edge of the dome.

3 III

IV V

FIG 64-1  Gross anatomy of the tarsals and surrounding bones. Ti, tibia; Fi; fibula; Ta, talus; Ca, calcaneus; ST, sustentaculum tali; N, navicular; Cu, cuboid; 1, 2, and 3 refer respectively to the first, second, and third cuneiforms (sometimes referred to as the medial, intermediate, and lateral cuneiforms, respectively); I, II, III, IV, and V refer to the first through fifth metatarsals, respectively.

Ti

Fi

Syndesmosis

Plafond MM

A

LM

Mortise

B

FIG 64-3  Gross anatomy of the ankle joint. A, The plafond (dotted line) is the transverse cortical articular surface at the distal end of the tibia. The mortise is the rectangular opening consisting of the plafond as well as the inner cortical articular surfaces (solid lines) of the medial malleolus (MM) and lateral malleolus (LM). B, The talar dome fits into the ankle mortise. The joint between the distal tibia and fibula is the syndesmosis (black bracket).

LPT

ST APC

A

C

ST

*

B

D

ANTERIOR

MIDDLE

ST

POSTERIOR

E FIG 64-4  Various views of the gross anatomy of the subtalar joint. A, Medial view. ST, sustentaculum tali. B, Inferior medial view. ST, sustentaculum tali. C, Lateral view. LPT, lateral process of talus; APC, anterior process of calcaneus. D, Anterior lateral view looking into the sinus tarsi (asterisk). E, The subtalar joint has been disarticulated: left, talus (flipped over); right, calcaneus. The articular surfaces of the three facets of the subtalar joint are coated with colored modeling clay: posterior (red), middle (blue), anterior (yellow).

CHAPTER 64  Foot and Ankle AIL UNIV OF M. METATARSALS

FOREFOOT

I

Lisfranc

II III IV

1 2 N

Chopart

Ta TARSALS

3 Cu Ca

2229

is a common site for fracture-dislocations to occur, particularly in diabetic patients with peripheral neuropathy. Figure 64-5 illustrates how the base of the second metatarsal (II) sticks down like a keystone, disrupting the otherwise relatively smooth tarsometatarsal joint. For this reason dislocations along the Lisfranc joint are typically accompanied by fractures across the base of the second metatarsal.

Cross-sectional Anatomy of the Tarsal Bones.  Figure 64-6 is a

V MIDFOOT HINDFOOT

Ti Fi

FIG 64-5  3D CT scan illustrating anatomic divisions of the foot. The Chopart joint separates the hindfoot (talus [Ta] and calcaneus [Ca]) from the midfoot (navicular [N], cuboid [Cu], and the three cuneiforms [1, 2, 3]). The Lisfranc joint separates the midfoot from the forefoot (metatarsals and phalanges).

and is separated from the midfoot by the Chopart* joint, a smooth continuation between the talonavicular and calcaneocuboid joints. The midfoot consists of the other five tarsal bones, the navicular, the cuboid, and the three cuneiforms. The forefoot consists of the metatarsals and phalanges and is separated from the midfoot by the tarsometatarsal joint, also known as the Lisfranc† joint. Along the Lisfranc joint

*François Chopart (1743-1795), a pioneer in urology, was known for the particular attention he gave to recording his numerous clinical observations. Thus it is somewhat surprising that he never wrote about the midtarsal amputation that bears his name almost 3 centuries later. He performed this surgery only once, on August 21, 1791, to resect a presumed liposarcoma of the foot. The approach was based on Chopart’s knowledge of the anatomy of the midfoot and was published by his student Laffiteau in 1792. † Jacques Lisfranc (1790-1865) was a very aggressive surgeon who wrote extensively and described many new procedures, including disarticulation of the shoulder, excision of the rectum, and amputation of the cervix. At age 23 he joined Napoleon’s army as a battlefront surgeon, a setting where amputations were the norm. Military surgeons (of the period) were not given the calm and unhurried atmosphere necessary for the task of laboriously picking out bone splinters and bits of clothing from gaping wounds. Locating the open ends of severed arteries and tying them off in the smoke of battle or by flickering candlelight was an enormous problem. Although some wounds did not themselves dictate amputation, it often had to be done because the patient could not otherwise survive the rigors of transport to the rear. The mind did not have time to reason. Experience and cold-bloodedness counted for more than talent. Everything had to be done with prompt and decisive action. In 1815, the final year of the war, Lisfranc wrote a 50-page paper describing his technique for performing a partial amputation of the foot at the tarsometatarsal joint, with the sole being preserved to make the flap. The technique was used to treat forefoot gangrene from frostbite. Lisfranc was widely known for his ability to amputate a foot in less than a minute, an important skill in that preanesthesia era.

series of straight axial images through the ankle and hindfoot from proximal (see Fig. 64-6A) to distal (see Fig. 64-6F). The straight axial plane is well suited to examine the syndesmosis (see Fig. 64-6B, arrow). The two joints that make up the Chopart joint, the talonavicular joint (see Fig. 64-6D) and the calcaneocuboid joint (see Fig. 64-6F), are also well profiled in the axial plane. However, the ankle and subtalar joints are not well profiled in the axial plane, and because examination of these two joints is usually the primary indication for requesting a CT of the ankle or hindfoot, other reformatted planes are required. Figure 64-7 is a series of straight sagittal images through the hindfoot from lateral (see Fig. 64-7A) to medial (see Fig. 64-7C). Nearly all of the joints are profiled in the sagittal plane, including the ankle joint, the calcaneocuboidal and talonavicular joints, and the posterior and middle facets of the subtalar joint. The only joint not well seen in the sagittal plane is the syndesmosis, but this is easily seen in the axial plane. Lateral sagittal images are also useful for visualizing the lateral process of the talus and the anterior process of the calcaneus (compare Fig. 64-7A with Fig. 64-4C). Figure 64-8 is a series of oblique coronal images through the hindfoot from posterior (see Fig. 64-8A) to anterior (see Fig. 64-8D). This plane best profiles the subtalar joint, and the broad posterior facet can be followed over several 3-mm slices (see Fig. 64-8A). As the posterior facet ends the middle facet begins, as defined by the sustentaculum tali (see Fig. 64-8B). When the oblique coronal slices are properly angled, the middle facet appears horizontally oriented (see Fig. 64-8C). The sinus tarsi is the cone of soft tissues directly lateral to the middle facet. Anterior to the subtalar joint, the round head of the talus is seen as a circle forming the talonavicular joint (see Fig. 64-8D). This demarcates the Chopart joint, the division between the hindfoot and midfoot.

Ankle Tendons There are 10 tendons that cross the ankle joint. For imaging purposes, these tendons can be clustered into four groups based on their anatomic locations, as illustrated by the colored curved lines drawn atop the 3D CT images in Figure 64-9. The anterior tendons are the anterior tibial, extensor hallucis longus, and extensor digitorum longus (see Fig. 64-9A). Posteriorly there are the Achilles and plantaris tendons (see Fig. 64-9B). Laterally the peroneus longus and peroneus brevis tendons pass under the lateral malleolus (see Fig. 64-9C). Medially the posterior tibial and flexor digitorum longus tendons pass under the medial malleolus, whereas the flexor hallucis longus passes under the sustentaculum tali (see Fig. 64-9D and E). On MRI, ankle tendons are best appreciated in cross section in the direct axial plane (Fig. 64-10). The oblique coronal plane (Fig. 64-11) is a good secondary plane to observe the medial and lateral tendons as they course under the malleoli. Normal tendons should appear uniformly black on all imaging sequences and have a sharply defined interface with adjacent fatty soft tissues. Any increased signal in a tendon on a T2-weighted image indicates the presence of pathology, typically an intrasubstance tear. In addition, more than a trace amount of fluid around an ankle tendon is abnormal, indicating inflammation or some other pathologic process. The exception to this is the flexor hallucis longus, which can normally contain some fluid in its tendon Text continued on p. 2236

2230

PART II  CT and MR Imaging of the Whole Body

MM

Ti

Ti

Fi

Ta

LM

Fi

A

C

B

I II III 2

1

1 3

3

N

N Cu

TNJ

Cu

Ta

CCJ

ST Ca

Ca

D

E

Ca

F

FIG 64-6  Straight axial images through the ankle and hindfoot, proximal (A) to distal (F). A, Proximal to syndesmosis. Fi, fibula; Ti, tibia. B, Through the syndesmosis (arrow). Fi, fibula; Ti, tibia. C, Through the top of the mortise. LM, lateral malleolus; MM, medial malleolus; Ta, talus. D, Through the sustentaculum tali (ST). Ca, calcaneus; N, navicular; Ta, talus; TNJ, talonavicular joint. E, Through the level where the calcaneus gets close to the navicular (arrowhead) but does not normally form a joint. If there were an articulation here or osseous bridging, that would be tarsal coalition. Ca, calcaneus; Cu, cuboid. Numerals indicate cuneiforms. F, Through the calcaneocuboid joint (CCJ). Ca, calcaneus; Cu, cuboid. Roman numerals indicate metatarsals.

CHAPTER 64  Foot and Ankle

Ti AJ

APC

Ta N

2

Ti

3

TNJ

P-STJ

AJ

Ta

P-STJ N

Ca

Ca M-STJ

Cu CCJ

A

TMJ

ls

sa

ar tat

LPT B

Me

Ti

Ta N 1 I

ST M-STU

C

FIG 64-7  Straight sagittal images. Ca, calcaneus; Cu, cuboid; N, navicular; Ta, talus; Ti, tibia. A, Through the lateral hindfoot, profiling the calcaneocuboid joint (CCJ), the ankle joint (AJ), and the posterior facet of the subtalar joint (P-STJ). The brown arrow points to the lateral process of the talus (LPT), and the red arrow points to the anterior process of the calcaneus (APC). Fractures through these pointed bony projections are often difficult to see on radiographs and are typically worked up with CT. B, Through the middle of the hindfoot, profiling the talonavicular joint (TNJ), the ankle joint (AJ), and the posterior facet of the subtalar joint (P-STJ). The middle facet of the subtalar joint (M-STJ) can now be seen. C, Through the medial hindfoot, now profiling the middle facet, above the sustentaculum tali (ST). Straight alignment should normally be present between the talus, navicular, medial cuneiform (1), and first metatarsal (I).

2231

2232

PART II  CT and MR Imaging of the Whole Body

Ti

Fi

Ta

Ta

Ta

*

ST

ST

Ca

Ca Ca

A

B

C

D

FIG 64-8  Oblique coronal images through the hindfoot, posterior (A) to anterior (D). Ca, calcaneus; Fi, fibula; ST, sustentaculum tali; Ta, talus; Ti, tibia. A, This plane best profiles the posterior facet of the subtalar joint (red arrow). The ankle mortise (yellow line) can be appreciated in the oblique coronal plane but would be better profiled in the mortise coronal plane. B, This oblique slice is just anterior to the ankle joint, where the posterior facet of the subtalar joint is ending (red arrow) and the middle facet is beginning (blue arrow). C, The oblique coronal slices are angled correctly if the middle facet of the subtalar joint (blue arrow) has a horizontal orientation. The cone of soft tissues lateral to the middle facet is the sinus tarsi (asterisk). D, The junction of the hindfoot and midfoot is at the round head of the talus at the talonavicular joint (circle).

CHAPTER 64  Foot and Ankle

Plantaris

N

ST Ach

AT

PL

A

EHL

EDL

C

B

PT FDL FHL

Tom Dick Harry

N

D

PB

ST

N ST

E

FIG 64-9  The 10 ankle tendons are illustrated as colored lines drawn over 3D CT images. A, Anterior view of the anterior tendons: anterior tibial (AT; red), extensor hallucis longus (EHL; green), and extensor digitorum longus (EDL; blue). B, Posterior view of the posterior tendons: Achilles (Ach; light blue) and plantaris (yellow). Also labeled are the medial pole of the navicular (N) and the sustentaculum tali (ST). C, Posterolateral view of the lateral tendons: peroneus brevis (PB; dark purple) inserting into the base of the fifth metatarsal, and peroneus longus (PL; light purple) wrapping under the cuboid. Medial (D) and posterior (E) views of the medial tendons, illustrating the “Tom, Dick, and Harry” mnemonic: the posterior tibial (PT; red) wraps under the medial malleolus and inserts on the medial pole of the navicular (N); the flexor digitorum longus (FDL; blue) runs behind the PT, under N, and out to the second to fifth toes; and the flexor hallucis longus (FHL; green) runs behind the talus, wraps under the sustentaculum tali (ST), crosses under the FDL at the master knot of Henry, and passes between the two great toe sesamoids (white arrows), inserting on the distal phalanx.

2233

2234

PART II  CT and MR Imaging of the Whole Body

AT

AT

EHL

EHL EDL

EDL

PT

PT

FDL

FDL

PB

A&N

PL

FHL

PB

A&N

PL FHL plantaris Ach

plantaris

B

A

Ach AT

AT EDL

EDL

N PT PT FDL

FDL

ST PB PL

PB

A&N

A&N

PL FHL

FHL

plantaris

C

Ach

D

Ach

FIG 64-10  MRI of normal ankle tendons in the straight axial plane. Ach, Achilles tendon; AT, anterior tibial tendon; EDL, extensor digitorum longus tendon; EHL, extensor hallucis longus tendon; FDL, flexor digitorum longus tendon; FHL, flexor hallucis longus tendon. PB, peroneus brevis tendon; PL, peroneus longus tendon; PT, posterior tibial tendon; A&N (artery and nerve) points to the dotted circle surrounding the neurovascular bundle that includes the posterior tibial artery and nerve. A, Just above the syndesmosis. B, Through the tip of the medial malleolus. C, One slice distal to B there is loss of the dark signal from the EHL tendon. D, Image through the talonavicular joint demonstrates the PT tendon inserting on the navicular (N), and the FHL tendon passing under the sustentaculum tali (ST). At this level, the EDL is dividing into separate tendon slips.

CHAPTER 64  Foot and Ankle

PT

PT

FDL

ST

A and N

FDL

FHL

PB

PB

PL

PL

FHL

A

B

N

PT PB FDL

PL FHL

C FIG 64-11  MRI of normal ankle tendons in the oblique coronal plane. FDL, flexor digitorum longus tendon; FHL, flexor hallucis longus tendon; PB, peroneus brevis tendon; PL, peroneus longus tendon; PT, posterior tibial tendon. A, Through the posterior facet of the subtalar joint. B, Through the middle facet of the subtalar joint. ST, sustentaculum tali. A and N (artery and nerve) points to the dotted circle surrounding the neurovascular bundle that includes the posterior tibial artery and nerve. C, Through the talonavicular joint. At this level the PT tendon has divided into separate slips. The white line with the round end points to the portion of the PT that inserts onto the medial pole of the navicular (N). The white line with the square end points to the portion of the PT that passes under the navicular. This patient has an os peroneum, which is why the PL tendon appears enlarged and gray at this level (dark gray arrow).

2235

2236

PART II  CT and MR Imaging of the Whole Body

sheath; it communicates with the ankle joint or the subtalar joint in up to 50% of cases.35

Anterior Tendons Normal anatomy.  The anterior tibial tendon is the largest tendon in axial cross section, except the Achilles tendon, and can be used as an internal standard for the other tendons.48 The anterior tendons extend uncrossed over the ankle joint and foot (see Fig. 64-9A). The anterior tibial is the most medial of the three anterior tendons. It extends along the medial aspect of the first tarsometatarsal joint to insert on the plantar aspect of the base of the first metatarsal and the adjacent medial cuneiform bone. The extensor hallucis longus is the middle of the three anterior tendons, proceeding straight to its insertion at the dorsal base of the great toe distal phalanx. The most lateral of the three anterior ankle tendons is the extensor digitorum longus. At the level of the midfoot, the extensor digitorum longus fans out into four separate tendon slips, which in turn proceed along the forefoot to insert at the dorsal bases of the second through fifth middle and distal phalanges.23,37 Whereas the anterior tibial and extensor digitorum longus tendons can be followed over a series of axial images (see Fig. 64-10), it is common to lose visualization of the extensor hallucis longus tendon as it curves anterior to the midfoot (see Fig. 64-10C). This is in part due to “magic-angle” effects.7 It is important not to misinterpret this lack of visualization as a rupture of the extensor hallucis longus tendon, a condition that is exceedingly rare. The lack of edematous signal along the course of the extensor hallucis longus on T2-weighted images should reassure the radiologist there is no pathologic process. Injury.  Tears of the anterior ankle tendons are rare, with the anterior tibial tendon being the most common to tear (Fig. 64-12). Ganglion cysts can arise from any synovium-lined structure, including the anterior ankle tendons. Figure 64-13 shows a synovial cyst arising from the anterior tibial tendon sheath.

A

Posterior Tendons Normal anatomy.  For anatomic purposes, the Achilles and plantaris tendons make up the posterior group. The Achilles tendon is the largest tendon in the body, originating in the midcalf at the junction of the two heads of the gastrocnemius muscle and the soleus muscle, and inserts onto the back of the calcaneal tuberosity. Unlike the anterior, medial, and lateral ankle tendons, all of which are surrounded by synovial sheaths, the Achilles is surrounded by thin layers of filmy fibrous tissue with fine internal blood vessels, called the paratenon or paratendon. This paratenon is analogous to synovium in that it provides nutrients for the tendon, but because the Achilles tendon does not change its axis of motion, there is no need for the lubrication function of synovium. Thus there should never be any fluid seen around a normal Achilles tendon. Directly anterior to the Achilles tendon is a triangular fat pad described radiographically by Kager in 1939.28 Kager’s fat pad is located in the retromalleolar region and is defined anteriorly by the posterior aspect of the tibia and posteriorly by the Achilles tendon, with the base being the proximal aspect of the calcaneus. The space contained within this triangle is filled with fatty tissue, producing a well-defined lucent triangle that can be seen on lateral radiographs of the ankle (Fig. 64-14A). On rupture of the Achilles tendon, this space becomes poorly demarcated and the normally lucent fatty tissue space becomes obscured (see Fig. 64-21A). The Achilles tendon is easily evaluated by physical examination as well as by MRI or ultrasonography (US).25 In the sagittal plane, the Achilles tendon should appear uniformly straight and black on T1-weighted images (see Fig. 64-14B) as well as on fluid-sensitive images (see Fig. 64-14C). There should be a sharp interface between the Achilles tendon and Kager’s fat pad directly ventral to it. A normal retrocalcaneal bursa may be present just in front of the Achilles tendon (white arrowhead in Fig. 64-14C). The normal retrocalcaneal bursa

B

FIG 64-12  Axial PD fat-suppressed (A) and sagittal T2 fat-suppressed (B) images in a 35-year-old man presenting with anterior ankle pain demonstrate increased signal within the anterior tibial tendon (arrow) consistent with a partial intrasubstance tear of the tendon. Surrounding fluid represents associated tenosynovitis (arrow).

2237

CHAPTER 64  Foot and Ankle

L

L

A

T2 FS

T1

T1 FS POST

P

P

T2 FS

L

T1

P

T1 FS POST

B FIG 64-13  Synovial cyst is identified arising from the anterior tibial tendon sheath in a 28-year-old woman. A, Axial T2 fat-suppressed, T1, and T1 fat-suppressed postcontrast images; and B, sagittal T2 fat-suppressed, T1, and T1 fat-suppressed postcontrast images demonstrate a cystic structure with high T2 signal, low T1 signal, and peripheral thin rim enhancement (arrow) consistent with a ganglion cyst. On the sagittal images a thin connection to the anterior tibial tendon is identified.

should measure less than 6 mm superior to inferior, 3 mm medial to lateral, and 2 mm anterior to posterior.47 Any fluid behind the Achilles tendon in a retro-Achilles bursa is abnormal. In the axial plane, the Achilles tendon should appear flattened in the anteroposterior direction. Distally the ventral margin of the tendon becomes concave with upturned corners, resembling a smile (see Fig. 64-10D). Injury.  Tears of the plantaris tendon tend to occur high in the calf at the myotendinous junction and have been called “tennis leg.” By MRI, plantaris tears present as fluid tracking along the length of the calf between the underlying soleus and more superficial gastrocnemius

muscles (Fig. 64-15). Figure 64-16 illustrates a chronically swollen and scarred posterior tibial tendon, with its cross-sectional area greater than that of the normal anterior tibial tendon. Ruptures of the Achilles tendon are usually diagnosed clinically, often by the patients themselves. The classic Achilles tendon rupture occurs with forced dorsiflexion of the planted foot, such as occurs in basketball or other jumping sports. The classic patient is a middle-aged “weekend warrior” who leads a sedentary life and attempts to participate in sports, perhaps with younger players, without an adequate warmup. Of all the tendons of the foot and ankle, the Achilles is the

2238

PART II  CT and MR Imaging of the Whole Body

A

T1

IR

Ach

Ach Kager’s fat pad

B

C FIG 64-14  Normal Achilles tendon in a 14-year-old with a calcaneal stress fracture. A, Lateral radiograph shows the normal sharp interface between the lucent Kager’s fat pad and the semiradiopaque Achilles tendon (white arrows). Sclerosis in the calcaneal tuberosity (black arrowheads) is more subtle radiographically. B, Midsagittal T1-weighted image shows the sharp interface between the normal bright Kager’s fat pad and the normal straight and uniformly dark Achilles tendon (Ach), which is uniform in thickness throughout its length. The dark line running perpendicular to the trabeculae in the calcaneal tuberosity is the stress fracture (black arrowheads). C, Midsagittal inversion recovery image reveals no abnormally increased signal in the uniformly dark Achilles tendon. A normal amount of fluid is present in the retrocalcaneal bursa (white arrowhead). There is bone marrow edema throughout the calcaneus as a response to the stress fracture in the tuberosity.

CHAPTER 64  Foot and Ankle

A

B

FIG 64-15  Plantaris tear in a 75-year-old who, while playing tennis, heard a “snap” and experienced sudden onset of posterior calf pain. A, Coronal T2-weighted fat-suppressed images and B, axial T2-weighted fatsuppressed images through the calf reveal soft tissue fluid tracking between the soleus and gastrocnemius muscles (arrow), consistent with a plantaris tear.

AT PT

A

T1

B

PD

C

T2

FIG 64-16  Stenosing tenosynovitis of the posterior tibial tendon (PT) in a 57-year-old with chronic medial ankle pain. These are straight axial images obtained at the same level with different sequences. A, T1-weighted image shows loss of the normal sharp fat-tendon interface around the PT (arrowhead). B, PD-weighted image shows that the chronically swollen and scarred PT is larger in axial cross section than the normal anterior tibial tendon (AT). C, T2-weighted image shows that the tissue surrounding the PT is not fluid but the chronic fibrotic scarring of stenosing tenosynovitis.

2239

2240

PART II  CT and MR Imaging of the Whole Body

T1

T2fs

A

B

T1

C FIG 64-17  Complete Achilles tendon rupture in a 41-year-old with history of renal transplantation and steroid use, who experienced acute posterior ankle pain 5 days earlier when bending over while gardening. Sagittal T1-weighted (A) and T2-weighted fat-suppressed (B) images reveal the complete Achilles tendon tear at the critical zone, 3 to 5 cm proximal to the calcaneal insertion. The arrows show the torn ends of the retracted fibers. C, Axial T1-weighted image through the tear reveals no intact Achilles tendon fibers (arrowhead). Incidentally noted is the intact plantaris tendon (arrow).

only one for which disorders have a male predominance. Complete ruptures of the Achilles tendon typically occur at one of two locations. One site is low, 3 to 5 cm just proximal to the calcaneal insertion (Fig. 64-17). This is a relatively hypovascular watershed region. The other site is relatively high, up at the myotendinous junction (Fig. 64-18). These more proximal tears may require that the imaging coil be repositioned around the lower calf rather than around the ankle to visualize the torn and retracted proximal end (Fig. 64-19). When it is clinically apparent to all that the Achilles tendon is completely ruptured, confirmation with MRI is usually unnecessary. However, imaging with MRI

or US is useful to measure the tendinous gap between the retracted ends of a complete tear prior to repair. Partial tears of the Achilles tendon are usually intrasubstance tears, and fluid-weighted images reveal increased signal in a swollen, abnormally rounded tendon (Fig. 64-20). Partial tears can also present  as nearly complete ruptures with only a few remaining fibers intact (Fig. 64-21). In these cases, abnormal fluid can be seen surrounding the intact fibers within the distended paratenon (see Fig. 64-21E). Imaging with MRI or US is used to assess the extent of partial  tears.

CHAPTER 64  Foot and Ankle

T1

T2fs

MTJ

A

2241

B

FIG 64-18  Complete Achilles tendon rupture in a 38-year-old who, while playing volleyball, felt a sudden “pop” and pain “like getting hit in the back of the leg.” A, Midsagittal T1-weighted image shows that the tear occurred at the musculotendinous junction (white arrow). B, Midsagittal T2-weighted fat-suppressed image shows the torn ends of the retracted fibers (arrows). This Achilles tendon tear required surgical repair.

m

A

B

FIG 64-19  Complete Achilles tendon rupture at the musculotendinous junction in a 52-year-old. A, An initial set of sagittal images was obtained with the extremity coil centered on the heel, which was too low to include the proximal end of the tear. B, The coil was repositioned proximal to the ankle joint to include the musculotendinous junction. The marker (m) is at the palpable defect.

An Achilles tendon that has undergone internal healing and scar formation from a prior intrasubstance tear tends to retain its thickened fusiform shape. However, unlike the acute intrasubstance tear, a healed Achilles tendon does not contain internal signal (Fig. 64-22).

Medial Tendons Normal anatomy.  The classic mnemonic “Tom, Dick, and Harry” is useful for remembering the order of the medial ankle tendons; T represents the posterior tibial tendon, D the flexor digitorum longus tendon, and H the flexor hallucis longus tendon. By changing the emphasis to “Tom, Dick, ANd Harry,” with the AN standing for the

posterior tibial artery and nerve, it is easier to remember the neurovascular bundle that runs between the flexor digitorum longus and flexor hallucis longus tendons (see Figs. 64-10 and 64-11). The posterior tibial tendon runs directly behind and under the medial malleolus, proceeds medial to the talus, and inserts on the medial pole of the navicular (see Fig. 64-10D). At this insertion site there is a focal osseous prominence called the navicular tubercle. The bulk of the posterior tibial tendon inserts on the navicular tubercle, although smaller slips of tendon pass under the navicular (see Fig. 64-11C) to insert on the plantar aspects of all three cuneiforms as well as the bases of the second through fourth metatarsals.

2242

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 64-20  Intrasubstance tear of the Achilles tendon in a 54-year-old with a history of rheumatoid arthritis and several months of persistent heel pain. T2-weighted fat-suppressed images in the sagittal (A) and axial (B) planes reveal that the distal Achilles tendon is abnormally swollen with increased intrasubstance signal (black arrow). An incidental finding is an abnormal amount of fluid in the posterior tibial tendon sheath (white arrow in B).

The flexor digitorum longus tendon runs directly behind the posterior tibial tendon, although these two tendons maintain individual tendon sheaths. The flexor digitorum longus tendon extends plantar to the bones of the midfoot, crossing superficially to the flexor hallucis longus tendon. This crossover point has been called the master knot of Henry,26 and the sheaths of these two flexor tendons may communicate at this point. Distally the flexor digitorum longus divides into separate tendon slips that insert on the plantar bases of the second through fifth distal phalanges. The flexor hallucis longus muscle is a posterior structure originating from the lower two thirds of the back of the fibula. The musculotendinous junction extends distally to the level of the ankle joint, and the proximal end of the tendon passes through a groove along the posterior talus. Whereas the posterior tibial and flexor digitorum longus tendons pass under the medial malleolus, the flexor hallucis longus tendon passes under the sustentaculum tali. The flexor hallucis longus then crosses deep to the flexor digitorum longus, extends under the first metatarsal, and passes between the two hallux sesamoids to insert on the plantar base of the distal phalanx (see Fig. 64-9D and E). Injury.  Of the three medial ankle tendons, the posterior tibial is the most prone to tear, characteristically along the portion that curves around the medial malleolus. The posterior tibial tendon is relatively hypovascular in this region.45 This region of the tendon is also susceptible to mechanical wear as the tendon rubs against the medial malleolus (Fig. 64-23). The posterior tibial tendon tends to tear with a longitudinal split rather than the transverse rupture seen in Achilles tendon tears. When imaged in the axial plane, a longitudinal split in the posterior tibial tendon resembles two individual tendons (Fig. 64-24). A tear of the posterior tibial tendon is associated with acquired pes planus deformity.15 Tenosynovitis refers to inflammation between the tendon and the surrounding synovial sheath, typically from chronic irritation. In the ankle, tenosynovitis most frequently occurs in the posterior tibial tendon and in the peroneal tendons. An abnormal amount of fluid in the tendon sheath indicates active tenosynovitis (Fig. 64-25). Dark fibrous tissue around the tendon suggests chronic scarring or stenosing

tenosynovitis (Fig. 64-26). Rheumatoid pannus can also be demonstrated by MRI (see Fig. 64-55) and enhances if intravenous (IV) contrast is administered.

Lateral Tendons Normal anatomy.  Laterally the peroneus brevis and longus tendons share a common sheath as they pass under the lateral malleolus. Distal to the lateral malleolus, the tendons are enveloped with individual sheaths. The peroneus brevis tendon extends along the lateral aspect of the midfoot and inserts on the tuberosity at the lateral base of the fifth metatarsal. The peroneus longus tendon passes through a groove in the plantar surface of the cuboid, crosses under the midfoot deep to the master knot of Henry, and extends medially to insert on the plantar aspect of the medial cuneiform and the base of the first metatarsal just lateral to the anterior tibial tendon insertion site. In their course the peroneus longus tendon runs posterior and inferior to the brevis (Fig. 64-27). Injury.  Owing to their long course spanning two joints, the peroneal tendons are prone to longitudinal split tear. This occurs more often in the peroneus brevis tendon, which is flatter as it passes around the lateral malleolus and lies between the fibula and the larger peroneus longus tendon. The presence of increased signal in the substance of the tendon, or the presence of fluid in the surrounding sheath, aids in the diagnosis of pathology of the peroneal tendons. Tenosynovitis, presenting with increased fluid in the peroneal tendon sheath, is also a common clinical problem (Fig. 64-28).30

Ankle Ligaments14 There are three sets of ligaments around the ankle joint. Laterally there are the syndesmotic ligaments and the lateral capsular ligaments. The syndesmotic ligaments consist of the thin anterior inferior tibiofibular ligament and the broader posterior inferior tibiofibular ligament. These ligaments are best seen in the axial plane (Fig. 64-29A), although they may be seen in the coronal plane if a single image serendipitously cuts through one (see Fig. 64-29C). It is these syndesmotic ligaments Text continued on p. 2249

A

T1

T2fs

C

B

T1

D

T2fs

E

FIG 64-21  Near-complete Achilles tendon rupture in a 54-year-old who, while playing racquetball, felt a pain “like being kicked in the back of the heel.” A, Lateral radiograph shows obscuration of the normally lucent Kager’s fat pad. B, Midsagittal T1-weighted image shows only a few remaining intact Achilles tendon fibers (arrow). C, Midsagittal T2-weighted fat-suppressed image shows the retracted ends of the torn fibers (black arrows). White arrow shows the few remaining fibers. D, Axial T1-weighted image through the level of the syndesmosis shows the markedly thinned intact Achilles tendon fibers (arrow). E, Axial T2-weighted fat-suppressed image at the same level shows bright abnormal fluid in the abnormally thickened and distended paratenon (arrowheads). White arrow shows the thinned intact Achilles tendon fibers. This Achilles tear ultimately required surgical repair.

2244

PART II  CT and MR Imaging of the Whole Body

T2

T1

A

T2

C

B

FIG 64-22  A 57-year-old with an Achilles tendon that has healed with chronic scarring. Axial T1-weighted (A), axial T2-weighted (B), and sagittal T2-weighted (C) images reveal that the distal Achilles tendon is too round and thick but contains no increased signal.

Medial malleolus

A

C

B

D

E

FIG 64-23  Illustration of posterior tibial tendon mechanical wear becoming a longitudinal tear. A, Medial view of the posterior tibial tendon (PT; red) as it wraps over the medial malleolus and under the flexor digitorum longus tendon (FDL; blue). The PT is susceptible to mechanical wear as it rubs back and forth (as indicated by the doubleheaded black arrow) between the underlying medial malleolus (gray lightning bolts) and the overlying FDL (white lightning bolts). B, A more anterior view of a partially torn PT as it might appear if it were laid flat. The tendon is thickened and butterflied open, with the gray region representing abnormal internal signal. (The dashed line represents the location of cross sections C to E). C to E, MRI cross sections of the PT only (now shown as a black ellipse), taken in the axial or oblique coronal plane through the longitudinal tear as it develops. In C, there is a gray wedge of abnormally increased signal along the inner aspect of the flattened PT (black ellipse). In D, tendinopathy (gray wedges) now involves the outer and inner surfaces of the PT. In E, the wedges of tendinopathy have progressed to a longitudinal tear, giving the appearance in cross section that the PT is two tendons.

CHAPTER 64  Foot and Ankle

1

2

MM

3 4

A

B

T1

C

PD

T2

FIG 64-24  Longitudinal split in the posterior tibial tendon (PT) in a 39-year-old. Shown are the same straight axial images obtained through the tip of the medial malleolus (MM). A, T1-weighted image well demonstrates the anatomy of the tendons as well as the neurovascular bundle (dotted oval). B, PD-weighted image shows what appears to be four medial tendons, the four-tendon sign, where 1 and 2 are the two halves of the split PT, and 3 and 4 are the normal flexor digitorum longus (FDL) and flexor hallucis longus (FHL) tendons. C, T2-weighted image demonstrates not bright fluid but gray scar (gray arrowhead) around the split PT, suggesting that this is chronic stenosing tenosynovitis. There is an abnormal amount of fluid in the FDL sheath (black arrowhead), suggesting active tenosynovitis here. The fluid in the FHL sheath (white arrowhead) is within normal limits for this tendon only.

A PD

B T2

FIG 64-25  Active posterior tibial tenosynovitis in a 46-year-old with chronic pain in the distribution of the posterior tibial tendon (PT). A, Straight axial PD-weighted image demonstrates that the PT is intact and contains no abnormal internal signal. The PT is slightly larger in cross section than the normal anterior tibial tendon, and there is loss of the normal fat signal around the tendon (arrowhead). B, Straight axial T2-weighted image at the same level reveals an abnormal amount of fluid in the posterior tibial tendon sheath (arrowhead), indicating active tenosynovitis.

2245

2246

PART II  CT and MR Imaging of the Whole Body

A

T1

B

C

PD

T2

FIG 64-26  Chronic posterior tibial stenosing tenosynovitis in a 57-year-old with chronic pain in the distribution of the posterior tibial tendon (PT). (This is the same patient as in Fig. 65-16; these straight axial images are two slices distal to those.) T1-weighted (A), PD-weighted (B), and T2-weighted (C) images all show abnormally dark signal (arrowheads) around the PT.

Lateral Malleolus

LM

PB PL

FIG 64-27  Coronal MRI (left) and graphic representation in the sagittal plane (right) demonstrate the relationship of the peroneal tendons to the lateral malleolus (LM); the peroneus brevis (PB) is closer to the distal fibula than is the peroneus longus (PL).

CHAPTER 64  Foot and Ankle

T1

LM

PD

T2

PB

PL

A

C

B PD

T2

PB

T1

LM

PL

D

E T1

PD

m

G

F T2

m

H

m

I

FIG 64-28  Longitudinal split of the peroneus brevis tendon (PB) in a 62-year-old. Straight axial T1-weighted (A), PD-weighted (B), and T2-weighted images through the syndesmosis. The PB (black arrow) is well seen on T1 and PD weighting, located between the lateral malleolus (LM) and the peroneus longus tendon (PL). At this level the PB has a normal flattened appearance. However, the T2-weighted image shows an abnormal amount of fluid in the common peroneal tendon sheath (white arrowhead). Straight axial T1-weighted (D), PD-weighted (E), and T2-weighted (F) images through the LM. At this level the PB is abnormally flattened to such an extent that it is draped over the PL, best seen on the PD image (E). Straight axial T1-weighted (G), PD-weighted (H), and T2-weighted (I) images distal to the LM. The marker (m) indicates the site of maximal tenderness. At this level the PB is split into two pieces (black arrows) separated by the intact PL (white arrow). Continued

2247

2248

PART II  CT and MR Imaging of the Whole Body

T1

PD

m

J

m

m

L

K

M

T2

N

FIG 64-28, cont’d  Oblique coronal T1-weighted (J), PD-weighted (K), and T2-weighted (L) images anterior to the LM through the pain marker (m). All three sequences show increased signal in the PB (black arrow) as opposed to the normal black signal in the PL (white arrow). Although the “magic angle” can artifactually increase the intratendinous signal on the short-TE sequences (T1 and PD), magic angle does not affect the long-TE T2-weighted sequence. Thus the bright signal in the PB on the T2-weighted image represents a true intrasubstance tear. The abnormal fluid in the common peroneal tendon sheath (white arrowhead) indicates active tenosynovitis. M and N, Far lateral sagittal inversion recovery images demonstrate the abnormal fluid in the common peroneal tendon sheath (white dotted rectangle) as well as the edema in the swollen PB (black dotted rectangle).

CHAPTER 64  Foot and Ankle

ATiFL

2249

ATaFL

PTiFL

PTaFL

B

A

PTiFL

PTaFL CFL

C

D

FIG 64-29  Normal lateral ankle ligaments. These are all T1-weighted images obtained using a high-resolution 512 acquisition matrix in the same normal volunteer as in Figure 64-10. A, Axial image through the bottom of the syndesmosis shows the two syndesmotic ligaments: the anterior tibiofibular ligament (ATiFL; white arrow) and the posterior tibiofibular ligament (PTiFL; black arrow). B, Axial image two slices distal to A through the top of the talar dome shows two of the three lateral capsular ligaments: the anterior talofibular ligament (ATaFL; white arrowhead) and the posterior talofibular ligament (PTaFL; black arrowhead). C, Coronal image through the back of the ankle joint shows the PTiFL (black arrow) running between the posterior malleolus of the talus and the fibula. D, Coronal image two slices anterior to C shows the PTaFL (black arrowhead) running between the back of the talus and the fibula. Also seen is a portion of the third of the three lateral capsular ligaments, the calcaneofibular ligament (CFL; gray arrowhead).

that are disrupted in a Weber type C ankle fracture (see Fig. 64-60C). Tear of the anterior inferior tibiofibular ligament represents a high ankle sprain and is much more common than a tear of the stronger posterior inferior tibiofibular ligament. The lateral capsular ligaments consist of the thin anterior talofibular ligament, the broader posterior talofibular ligament, both of which are transversely oriented and thus best seen in the axial plane (see Fig. 64-29B), and the longitudinally oriented calcaneofibular ligament, best seen in the coronal plane (see Fig. 64-29D). The anterior talofibular is more commonly injured, followed by the calcaneofibular; the posterior talofibular is rarely injured (Fig. 64-30). Medially the ankle joint is stabilized by a group of ligaments that fan out from the distal tip of the medial malleolus in a triangular

configuration and collectively are called the deltoid ligament. When viewed in the coronal plane (Fig. 64-31), the deltoid ligament can be seen to consist of deep fibers that insert on the medial process of the talus and superficial fibers that insert on the calcaneus at the sustentaculum tali. Injuries of the deltoid ligament tend to be sprains* rather than complete ruptures, although they may be accompanied by bone marrow edema (Fig. 64-32) or avulsion fractures.

*Sprains are defined as stretching or tearing of ligaments and are due to twisting injuries. Strains are defined as stretching or tearing of muscles, often at the musculotendinous junction, and are caused by sudden and powerful contractions or from overuse.

2250

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 64-30  Tear of the lateral ligaments of the ankle in a 38-year-old. A, Axial PD fat-suppressed images demonstrate tear of the anterior talofibular ligament (arrow). B, Coronal T2 fat-suppressed images demonstrate disruption of the calcaneofibular ligament as well (arrow).

MM

T

D

ST H

FIG 64-31  Normal medial ankle ligaments on a T1-weighted image obtained using a high-resolution 512 acquisition matrix. This coronal image is located just behind the middle facet of the subtalar joint. (This image is seven slices anterior to Fig. 65-29D). The magnified dashed box to the right shows the superficial and deep components of the deltoid ligament. The broader deep fibers (black arrow) run from the medial malleolus (MM) to the medial process of the talus. The more slender superficial fibers (white arrow) run from the MM to the calcaneus at the sustentaculum tali (ST). Also shown is the flexor retinaculum (open arrowheads), forming the roof of the tarsal tunnel atop the three medial tendons (T, posterior tibial; D, flexor digitorum longus; H, flexor hallucis longus) and the posterior tibial neurovascular bundle (dotted oval).

CHAPTER 64  Foot and Ankle

T1

A

2251

T2fs

B

FIG 64-32  Deltoid ligament sprain in an 18-year-old. Mortise coronal T1-weighted (A) and T2-weighted fatsuppressed (B) images show abnormally increased signal in the deep deltoid (black arrow). There is bone marrow edema at its insertion site on the medial talus (arrowhead). The superficial deltoid (white arrow) is intact.

Another important ligamentous structure in the ankle is the spring ligament complex, consisting of the superomedial calcaneonavicular, the medioplantar oblique calcaneonavicular and the inferoplantar  longitudinal calcaneonavicular ligaments. The spring ligament is an important stabilizer of the medial arch, providing support for the talar head (Fig. 64-33). Tear of the spring ligament is associated with posterior tibial tendon tear and pes planus deformity.32

Accessory Ossicles There are many accessory ossicles that can be present throughout the skeleton, and these are well documented in the normal variants textbook by Keats and Anderson.29 Many of these normal variants can be found in the feet. Three of the most commonly found accessory ossicles in the feet are the os trigonum posterior to the talus, the accessory navicular medial to the navicular bone, and the os peroneum plantar to the calcaneocuboid joint. Although in the vast majority of people these are nothing more than asymptomatic incidental findings, in rare circumstances they can cause pain and can be difficult to diagnose and treat.

Os Trigonum Syndrome.  The os trigonum is a common accessory ossicle located directly behind the talus at the posterior end of the subtalar joint, adjacent to where the flexor hallucis longus wraps around the back of the talus. The os trigonum develops as a separate ossification center. During growth this fuses to the talus in most people, but in 5% to 15% of normal feet it remains nonunited and is variable in size and shape. There are no radiographic findings in a patient with symptomatic os trigonum syndrome,50 although the diagnosis can be made with MRI by demonstrating marrow edema in the os trigonum and the adjacent talus (Fig. 64-34).

Accessory Navicular Syndrome.  The accessory navicular bone (os tibiale externum) is a common normal variant found adjacent to the medial pole of the navicular in approximately 10% of the population.

The medial pole of the navicular is the primary insertion site of the posterior tibial tendon (see Figs. 64-10D and 64-11C). Small accessory navicular bones are called type 1 and are simply sesamoid bones within the substance of the posterior tibial tendon (Fig. 64-35). The posterior tibial tendon still inserts normally on the navicular, and the type 1 accessory navicular bones are of no clinical significance. Larger accessory navicular bones are called type 2 (Fig. 64-36), and with these the posterior tibial tendon inserts onto the accessory navicular rather than on the navicular bone itself. Patients with this normal variation are typically asymptomatic unless they develop stress reaction or a tear through the normal fibrous union between the navicular and accessory navicular. A painful accessory navicular syndrome can be diagnosed on MRI by the presence of marrow edema in  the accessory navicular and the adjacent navicular bone (Fig. 64-37). Normally there should be a solid fibrous union between the type 2 accessory navicular and the navicular. The presence of a line of fluid between these bones is abnormal and indicates a pseudarthrosis, another MRI finding in accessory navicular syndrome (Fig. 64-38). A type 3 accessory navicular refers to an enlarged navicular tuberosity without synchondrosis.

Os Peroneum Syndrome.  The os peroneum is a common sesamoid bone located in the peroneus longus tendon as it passes under the cuboid. Painful os peroneum syndrome presents with marrow edema of the ossicle with surrounding soft tissue edema, best shown with  a fluid-sensitive MRI sequence targeted to the lesion (Figs. 64-39   and 64-40).

IMAGING PROTOCOL All imaging of the ankle and foot must begin with radiographs. Traumatic fractures are the most common cause of ankle and foot pain, Text continued on p. 2260

2252

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 64-33  Spring ligament complex. A, Axial PD image through the ankle, with arrows delineating the medioplantar oblique calcaneonavicular ligament. B, Coronal T2 fat-suppressed image, with arrows delineating the superomedial oblique calcaneonavicular ligament, which is the most commonly torn component of the spring ligament.

CHAPTER 64  Foot and Ankle

LEFT

RIGHT

A

B

C

T1

D

IR

PD

F

T2fs

T D & H

E

FIG 64-34  Os trigonum syndrome in an 11-year-old competitive Irish dancer. Lateral view of the symptomatic left ankle (A) shows a small os trigonum, a common normal variant. The asymptomatic right side (B) is shown for comparison. C, Midsagittal T1-weighted image shows the small os trigonum (arrow). D, Corresponding sagittal inversion recovery image shows bone marrow edema in the os trigonum (arrow) as well as in the adjacent talus (arrowhead). E, Straight axial PD-weighted image shows the irregular cleft (arrowhead) between the os trigonum and the talus. Well seen are the normal structures in the tarsal tunnel: the posterior tibial tendon (T), flexor digitorum longus tendon (D), neurovascular bundle (&), and flexor hallucis longus tendon (H). F, Corresponding axial T2-weighted fat-suppressed image shows bone marrow edema in the os trigonum (arrow) and in the adjacent talus (arrowhead).

2253

2254

PART II  CT and MR Imaging of the Whole Body

FIG 64-35  Normal type 1 (small) accessory navicular (arrow in dashed magnified box).

FIG 64-36  Normal type 2 (large) accessory navicular (white arrow in dashed magnified box). Incidentally seen is a normal os peroneum (black arrow).

CHAPTER 64  Foot and Ankle

LEFT

RIGHT

A

B

T1

IR

N

C

T A

D

FIG 64-37  Accessory navicular syndrome in a 20-year-old with focal pain directly over the left medial navicular. A, Standing anteroposterior radiograph of the asymptomatic right foot shows an elongated medial pole of an otherwise normal navicular (arrow). This has been referred to as a cornuate navicular and as a type 3 accessory navicular. B, Standing anteroposterior radiograph of the symptomatic left foot barely reveals the type 2 accessory navicular (arrow in the magnified dashed box). C, Far-medial sagittal T1-weighted image well shows the posterior tibial tendon (T) inserting onto the type 2 accessory navicular (A), as well as the fibrous union (arrowhead) between it and the navicular bone (N). D, Corresponding sagittal inversion recovery image shows subcortical edema (arrows) on both sides of this fibrous union. Continued

2255

2256

PART II  CT and MR Imaging of the Whole Body

Ta N m

m A

T1

T2fs

E

F

N m

G

A

T1

m

H

T2fs

FIG 64-37, cont’d  E, Oblique coronal T1-weighted image through the round head of the talus (Ta) shows the marker (m) indicating that the site of focal tenderness is directly over the abnormal articulation between the type 2 accessory navicular (A) and the navicular bone (N). F, Corresponding coronal T2-weighted fatsuppressed image shows subcortical edema (arrows) on both sides of this abnormal articulation. G, Oblique axial T1-weighted image shows that the marker (m) indicating the site of focal tenderness is directly over the abnormal articulation between the type 2 accessory navicular (A) and the navicular bone (N). H, Corresponding axial T2-weighted fat-suppressed image shows subcortical edema (arrows) on both sides of this abnormal articulation.

CHAPTER 64  Foot and Ankle

B

A

N A N

C

T A

D

FIG 64-38  Accessory navicular syndrome in a 38-year-old. A, Anteroposterior radiograph shows a type 2 accessory navicular (white arrow in magnified gray dashed box). B, On the lateral radiograph, the accessory navicular can faintly be seen through the anterior calcaneus (black arrow in magnified white dashed box). Axial (C) and far-medial sagittal (D) T2-weighted fat-suppressed images reveal a line of fluid (white arrowhead) indicating an abnormal joint where there should be a solid fibrous union between the navicular (N) and accessory navicular (A). The far-medial sagittal image shows the posterior tibial tendon (T) inserting on the type 2 accessory navicular.

2257

B

A

T1

T1 LM

LM

PB L P

Cu

5

C

Ca PL

D

IR LM

Cu

Ca

E FIG 64-39  Os peroneum syndrome in a 58-year-old who developed chronic lateral foot pain after ballroom dancing. (Case courtesy Edwin Rogers, MD.) Oblique (A) and lateral (B) radiographs reveal an os peroneum (white arrow) below the calcaneocuboid joint, a common normal variant. C, Far-lateral sagittal T1-weighted image shows the peroneus longus tendon (PB) wrapping around the lateral malleolus (LM) toward the base of the fifth metatarsal (5). Behind and below the PB is the peroneus longus tendon (PL). D, Sagittal T1-weighted image one slice medial to C. Here, the PL is passing under the calcaneus (Ca) and cuboid (Cu). Directly plantar to the calcaneocuboid joint is the os peroneum (black arrow), a sesamoid of the PL. (The os peroneum is difficult to see on this T1-weighted image because its edematous bone marrow is dark.) E, Corresponding inversion recovery image of slice at D demonstrates bone marrow edema throughout the os peroneum (arrow).

CHAPTER 64  Foot and Ankle

F

G

T1

T2

T1

H

2259

IR IR

J

I

K RT PLANTAR LT

L

FIG 64-39, cont’d F, Coronal T1-weighted slice through the os peroneum (arrow) shows its marrow to be darker than that of the other bones. G, T2-weighted image without fat suppression, corresponding to coronal slice at F. This sequence is not particularly edema sensitive, and the marrow signal of the os peroneum (arrow) is isointense to that of the other bones. H, Corresponding coronal inversion recovery image of slice at F and G. This sequence is so fluid sensitive it shows marrow edema not only in the os peroneum (arrow) but also in the adjacent cuboid (arrowhead). I, Axial T1-weighted image through the bottom of the foot shows the os peroneum (arrow). J, Axial inversion recovery image corresponding to slice at I shows the bone marrow edema in the os peroneum (arrow). K, Bone scan, both-feet-on-detector view, shows increased activity in the os peroneum (arrow) of the left foot. The normal right foot is included for comparison. To help with localization, also included is an axial scout MRI (L) showing the os peroneum (arrow).

2260

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 64-40  Axial (A) and sagittal (B) T2 fat-suppressed images demonstrate focal marrow edema in the os peroneum (arrow) and along the peroneus longus tendon (arrow).

and radiographs are the quickest and least expensive imaging modality for detection and follow-up of these fractures. However, the question often arises as to which imaging modality should be obtained next if radiographs do not sufficiently delineate the fracture or do not demonstrate the cause of symptoms. Figure 64-41 is a flow chart outlining which modalities we use to image various pathologic processes. CT is used when we specifically need to assess bone cortex. The ability to reformat CT data into two-dimensional (2D) cross-sectional images in any plane desired and to create 3D projections makes it the ideal modality to assess fractures, especially comminuted and intraarticular fractures. This is particularly helpful to orthopedic surgeons as part of their presurgical planning. CT is also useful for showing fractures that are difficult to see radiographically, such as the lateral process of the talus or the anterior process of the calcaneus. CT can be used to show arthritic narrowing of joints that are difficult to visualize radiographically, such as the subtalar joint. In patients who have undergone arthrodesis in an attempt to fuse a painful arthritic joint, CT can show the degree of bone fusion at the cortical level. CT is also the best modality to show the abnormal bone cortex in cases of tarsal coalition, both osseous and fibrous coalitions. MRI is essentially used for everything else. MRI is the best way to evaluate all of the ankle tendons at once for tears or tenosynovitis. It is the best way to evaluate masses arising from either the soft tissues or bones of the extremities. MRI is extremely sensitive for detection of bone marrow edema and soft tissue edema, and as such is it useful for the detection of conditions that may be radiographically occult, including stress fractures, osteochondral lesions, and infection. US can be used to perform a focused examination of the soft tissues of the ankle or foot. Unlike MRI, which images all of the ankle bones and tendons at once, US is usually targeted at one specific tendon. US is particularly useful for dynamic examinations and to evaluate small superficial structures that are sometimes difficult to see on MRI, such as Morton’s neuromas or the plantar plate. US can also be used to characterize soft tissue masses, particularly to determine if the mass is

Radiographs

CT

MRI

US

NM

Assess cortex • Fractures -Calcaneus -Distal tibia -Lateral process of talus • Arthritis • Fusions • Coalitions -Osseous -Nonosseous

Everything else • Tendons -Tears -Tenosynovitis • Masses -Soft tissue -Osseous • Bone pathology -Occult fracture -Osteochondral lesions -Infection

• Tendons -Achilles • Toes -Morton’s -Plantar plate • Masses -Vascularity -Cyst vs solid • Foreign bodies -Wood

• Screening -Sesamoid • Charcot

FIG 64-41  Flow chart for imaging the ankle and foot (see text).

cystic or solid. US is also extremely sensitive for the detection of subcutaneous foreign bodies in the extremities, particularly wooden splinters, which can be difficult to detect with radiographs, CT, or MRI. (A discussion of US of the ankle and foot is beyond the scope of this chapter.) Nuclear medicine studies, more specifically three-phase bone scan, can be used to detect stress fractures in the ankle and foot, although it is not specific. In neuropathic feet, a bone scan combined with a white blood cell scan can be as sensitive as and more specific for the detection of osteomyelitis than MRI. (A discussion of nuclear medicine is beyond the scope of this chapter.)

Radiography Because radiographs are a necessary first step in the workup of the ankle or foot, let us now briefly review how these should be obtained.

CHAPTER 64  Foot and Ankle Ankle Radiography.  Ankle radiographs can be either weight bearing or non–weight bearing, depending on the preference of the ordering clinician. A standard radiographic ankle series consists of three pro­ jectional views: anteroposterior (AP), mortise (oblique), and lateral  (Fig. 64-42). The mortise view is similar to the AP view, with the leg internally rotated 15 degrees to obtain a better profile of the ankle mortise. When obtaining radiographs of the ankle, it is important that the technologist include the base of the fifth metatarsal on at least one view. Patients with fractures of the base of the fifth metatarsal clinically present complaining of lateral ankle pain, and this can cause the clinician to request radiographs of the ankle rather than of the foot. Figure 64-42 is such a case, where a Jones* fracture can be seen at the edge of the lateral view.

Foot Radiography.  It is preferable to obtain radiographs of the foot with the patient standing to visualize the bones in their weightbearing alignment.5 AP and oblique views (Fig. 64-43A and B) can be obtained by placing the x-ray cassette on the floor and having the patient stand on the cassette while the x-ray beam is pointed downward. It is important to closely scrutinize the alignment of the tarsometatarsal joints on both of these views when assessing for a Lisfranc fracture-dislocation. Normally the first metatarsal should line up perfectly with the medial cuneiform, the second metatarsal with the middle cuneiform, the third metatarsal with the lateral cuneiform, and the fourth metatarsal with the cuboid. For obtaining the standing lateral view of the foot (see Fig. 64-43C) the patient stands on a raised set of wooden steps (Fig. 64-44). This elevates the feet to a level where the x-ray beam can be oriented horizontally while the cassette is held between the feet. Figure 64-45 is an example of differences that can be seen between standing and nonweightbearing views.

Computed Tomography

Overview.  Bone CT protocols have evolved as scanner technology has progressed. A CT gantry consists of a spinning ring of an x-ray tube and detectors. The tube emits a fan-shaped beam that reaches the array of detectors after traversing the patient or any other material in the gantry. With early generations of single-slice CT scanners the gantry would spin clockwise one rotation, then stop and spin counterclockwise one rotation to prevent tangling of the power cables supplying the x-ray tube. The patient table would be stationary during each of the scanning rotations while the gantry was spinning and the tube was emitting x-rays. The table would move between gantry rotations and stop at each slice position. Scans were relatively slow because time was lost stopping the gantry’s clockwise momentum to reverse its rotation.

*Sir Robert Jones (1857-1933) was the father of orthopedic surgery in England and revolutionized the care of wounded soldiers during World War I. An early proponent of x-rays, Jones imaged the transverse extraarticular fracture across the proximal diaphysis of the fifth metatarsal just a few months after Röntgen published “On a New Kind of Rays” (December 28, 1895). Jones first described this fracture after having sustained such an injury himself “whilst dancing.” (This was not ballroom dancing; rather it was “dancing in a circle round the tent pole” with his military colleagues. There was no mention as to whether alcohol was involved.) He subsequently identified this fracture on radiographs of two other patients and published his series of three in the Annals of Surgery in 1902, “Fracture of the Base of the Fifth Metatarsal Bone by Indirect Violence.”

2261

With the development of slip-ring technology, the problem of tangled cables was eliminated and the gantry could spin in one direction while the table moved continuously through it. Thus helical CT (also known as spiral CT), was developed. Now the data stream from the x-ray detectors represent a continuous volume of the patient. The raw data are then reconstructed into a series of axial slices, which we refer to as the source images. Because of its volumetric nature, helical data can be reconstructed at any slice width and with any interval spacing between slices. These axial source images can then be reformatted into 2D slices in any desired plane and of any desired width, or into 3D volume-rendered images. In the past decade, single-slice helical scanners have evolved into multislice scanners able to acquire larger volumes of patient data with each gantry rotation. This technology has largely been driven by the desire to scan the entire chest within a single breath hold and the coronary arteries in a single heartbeat. Although a multislice CT scanner is not absolutely required for bone CT, covering the desired volume faster minimizes artifacts related to patient motion as well as decreasing the amount of time the patient has to lie still on the scanner table. Thus the modern bone CT scan consists of the acquisition of three sets of imaging data. The raw data tend not to be archived. The source images are reconstructed from the raw data using a variety of filtered back-projection algorithms. With the increase in the computational capabilities of modern CT scanners, the use of iterative reconstruction algorithms has been used to improve image quality and reduce radiation dose. In musculoskeletal imaging this has been particularly useful in reducing metal artifact from prostheses and other orthopaedic hardware.34 The source images are typically oriented in a plane axial to the scanner gantry. These source images can be viewed by the radiologist as desired and can be sent to the picture archiving and communications system (PACS) for short- or long-term storage. However, the multiplanar 2D and 3D images reformatted from the source images are the ones primarily used for diagnostic and planning purposes and ultimately sent to the PACS for archiving. Achieving the highest-resolution 2D reformatted images requires the source images to be reconstructed into thin slices, often the width of a detector element. To minimize the stairstep artifact that can occur between axial slices, the source images should be reconstructed at intervals such that they overlap each other. We have found that a 50% overlap (interslice interval equals one-half slice width) works well. We use an edge-enhanced reconstruction algorithm (called a bone algorithm by some CT manufacturers) to yield 2D reformatted images with sharp cortical detail. Thin and overlapping source images also yield good 3D reformatted images. However, 3D images by their nature represent a smooth rendering of the volumetric data, and edge-enhanced source images can yield excessively noisy 3D images. Therefore a second set of source images using a smoothing reconstruction algorithm (called a standard or soft tissue algorithm by some manufacturers) is usually created for 3D reformatting.

Protocol for Foot, Ankle, and Tibia (Distal) Scanning technique.  For imaging of the foot and ankle the patient is positioned supine on the CT table, with legs straight, feet together in the center of the gantry, and toes pointing up at the ceiling (Fig. 64-46A). We typically scan through both ankles and feet simultaneously because this position is most comfortable for the patient and allows us to compare the injured side with the contralateral normal side when questions arise regarding subtle alignment abnormalities. Unless the contralateral foot contains metal, including it in the scanning field of view (FOV) does not cause excessive streak artifacts and

2262

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 64-42  Non-weightbearing radiographic ankle series in a 37-year-old with lateral ankle pain after an acute inversion injury. AP (A) and mortise (B) views demonstrate a normal appearance of the ankle joint. C, The lateral ankle view reveals no abnormalities of the hindfoot. (The region outlined by the dashed rectangle is magnified and displayed to the right.) Close inspection of the base of the fifth metatarsal on the lateral view of the ankle reveals a proximal diaphyseal fracture, a Jones fracture (arrow). Fractures of the base of the fifth metatarsal often present clinically as lateral ankle pain. The technologist must be careful always to include the base of the fifth metatarsal on at least one view of all ankle radiographic series.

CHAPTER 64  Foot and Ankle

2263

FIG 64-44  Method of obtaining weightbearing lateral view. In most

A

B

C FIG 64-43  Weightbearing radiographic foot series in an asymptomatic 41-year-old male volunteer. A, Anteroposterior view; B, oblique view; C, lateral view. The upward-pointing white arrow was placed by the technologist to indicate the patient was standing. The heel spur (black arrow in C) at the origin of the plantar fascia is of doubtful clinical significance in this normal volunteer who has never had heel pain. (This person is not standing on screws but on the wooden box in Fig. 64-44, held together by screws.)

does not increase the radiation exposure to organs in the torso. Securing the patient’s feet to a dedicated holder (see Fig. 64-46B) or to a box (see Fig. 64-46C) helps hold the feet in neutral position and prevent motion during the scan. Scout views are obtained in both the AP and lateral projections (Fig. 64-47). The scanning FOV should be set wide enough to include both the right and left lateral malleoli; for most patients this is less than 25 cm. The coverage should begin superior to both syndesmoses and extend below the calcanei to include the plantar soft tissues. In cases of pilon fractures, which are comminuted fractures involving the plafond, coverage is extended superiorly to include more of the distal tibia. We typically scan using 120 kVp at less than 200 mA. We can achieve the highest resolution on the reformatted images by reconstructing the source images at a width equal to the width of the narrowest detector. The reconstruction interval between source images is equal to half the detector width, and this allows for a 50%

radiology departments the x-ray tube cannot be lowered to the floor. Weightbearing lateral views can be obtained by having the patient stand on a wooden box. The central x-ray beam (dashed arrow) passes through the foot from lateral to medial, striking the x-ray cassette held upright between the feet.

overlap between slices. We reconstruct two sets of source images; one uses an edge-enhanced bone algorithm for the 2D multiplanar reformatted images, the other a smoothing standard algorithm for reformatting in 3D. Both sets of source images are sent to the PACS for archival storage and can be accessed any time in the future if additional 2D or 3D reformatted images are desired. Reformatting technique.  There are several different ways to create 2D reformatted images of the ankle and foot. The reformatting protocols are centered on the anatomic divisions illustrated in Figures 64-5 and 64-48. Ankle/distal tibia protocol.  For imaging of the ankle, scanning (Fig. 64-49A) is centered on the ankle joint. This protocol is used for scanning fractures of the distal tibia (e.g., pilon, malleoli, triplane, juvenile Tillaux) or of the talar dome (e.g., osteochondral lesions). Using a midsagittal reference image, straight axial images are created in a plane parallel to the bottom of the foot. Then, using an axial reference image through the top of the ankle mortise, mortise coronal and mortise sagittal images are created parallel and perpendicular to a line through the anterior cortex of the medial and lateral malleoli. For distal tibial fractures, we find that creating reformatted images that are 3 mm thick at 3-mm intervals (no gap or overlap between reformatted slices) yields crisp images that do not appear noisy. Thinner reformatted images can be performed however if required or to visualize smaller structures. Hindfoot/midfoot protocol.  Our hindfoot/midfoot protocol (see Fig. 64-49B) is centered on the Chopart joint and is used to evaluate hindfoot fractures (e.g., calcaneus, talar body) and the subtalar joint (e.g., tarsal coalitions). Using an axial reference image, straight sagittal images are reformatted along a plane parallel to the long axis of the foot. The other three planes are reformatted off a midsagittal reference image. Straight axial images are reformatted in a plane parallel to the bottom of the foot. Oblique coronal and oblique axial images are reformatted in planes both perpendicular and parallel to the posterior facet of the subtalar joint. Forefoot/midfoot protocol.  Our forefoot/midfoot protocol (see Fig. 64-49C) is primarily used to assess the alignment of the Lisfranc joint and the integrity of the adjacent bones. We find that it works best to create reformatted images in three planes relative to the first

2264

PART II  CT and MR Imaging of the Whole Body

PORTABLE

A

UPRIGHT

B

FIG 64-45  Importance of weightbearing view. A, Non-weightbearing lateral view, obtained portably. On this image, the long axis of the talus (dashed white line) is parallel to the long axis of the first metatarsal (dashed black line), suggesting normal alignment. B, Same patient as in A, obtained upright at a follow-up clinic visit 3 months later. On this weightbearing lateral view, the long axis of the talus (solid white line) is now angled downward relative to the first metatarsal (solid black line). This indicates that the patient has a nonrigid flatfoot deformity (pes planus), demonstrable only with weight bearing.

A

B

C FIG 64-46  A, The patient is positioned supine on the CT table with her legs straight, feet together, toes pointing to the ceiling. B, Example of a foot holder we built to help keep patients’ feet centered in the CT scanner in neutral position. C, In lieu of a dedicated foot holder, we have used a box.

CHAPTER 64  Foot and Ankle

A

2265

B

FIG 64-47  Anteroposterior (A) and lateral (B) scout CT views. The scanning field should cover both ankles and should extend from above the syndesmosis to below the calcaneus (white rectangles). In cases of pilon fractures, more of the distal tibia is covered (dashed rectangle).

these navicular fatigue fractures tend to be incomplete hairline cracks, we increase the resolution by creating thin (1-mm) reformatted images in a small (6-cm) FOV. Oblique coronal and oblique axial images are reformatted off a sagittal reference image, and oblique sagittal images are reformatted off an axial reference image.

Ti

Magnetic Resonance Imaging

AJ Ta

Hindfoot

t

par

Cho

ot dfo Mi nc fra Lis

t foo

Coils and Markers.  When imaging the ankle or foot with MRI, it is

P-STJ

Ca

re

Fo

FIG 64-48  Anatomic division and major joints shown on a sagittal CT. The joint between the tibia (Ti) and talus (Ta) is the ankle joint (AJ), shown with the yellow line. The joint between the talus and calcaneus (Ca) is the subtalar joint; the posterior facet (P-STJ) is shown with the red line. The Chopart joint, shown with the curved blue line, separates the hindfoot from midfoot. The Lisfranc joint, shown with the angled green line, separates the midfoot from forefoot.

metatarsal shaft. A sagittal reference image that best delineates the entire length of the first metatarsal is selected. Long-axis and short-axis planes are reformatted both parallel and perpendicular to the sagittal length of the first metatarsal. The third plane, sagittal to the first metatarsal, is best obtained off an axial reference image that has been obliqued to include the entire length of the first metatarsal. Navicular protocol.  A dedicated navicular protocol (see Fig. 64-49D) can be used to assess the healing of a known navicular fatigue fracture that has perhaps been previously diagnosed by MRI. Because

vital to tailor our protocol to answer the clinical question. The imaging planes, sequences, and even the selection of which coil to use will vary depending on the clinical circumstances. Whenever possible, an extremity coil should be used. We use a dedicated foot and ankle coil (Fig. 64-50) that incorporates a chimney-like extension so that the toes can be included in the FOV. This coil’s special design helps to hold the patient’s foot and ankle in a neutral position—that is, with the bottom of the foot perpendicular to the tibia as it would be if the patient were standing. Alternatively, knee coils or surface coils can be used. Some designs of knee coils have an open top that allows the toes to protrude (Fig. 64-51A). Custom cushioned inserts (see Fig. 64-51B) help to keep the heel immobilized and centered in the coil. Although the neutral positioning shown in Figure 64-51A is fine for imaging the ankle and hindfoot, it would not be appropriate for the phalanges. When it is necessary to image the toes and a foot coil as in Figure 64-51 is not available, the knee coil can be used with the patient’s foot held in plantar flexion. In these circumstances, to image the forefoot the patient can also be placed prone to maintain plantar flexion if this is more comfortable. Also, patients tend to feel less claustrophobic when prone. We encourage our technologists to place a vitamin E marker over the sight of maximal tenderness or near a nonhealing ulcer or a palpable mass. If the patient initially points to a wide area, he/she should be asked to point to the site of maximum pain with one finger. The marker should be placed in such a way as not to deform the tissues and alter their anatomic relationships. Optimally the skin markers should be placed at the borders of the site of interest. Markers can be helpful to draw the attention of both the technologist acquiring the images and the radiologist interpreting the images.

2266 A

PART II  CT and MR Imaging of the Whole Body Ankle/distal tibia (centered on ankle joint)

1. Straight axial (off a sagittal) Standard: 3 × 3 mm For OLT: 3 × 3 mm

B

2. Mortise coronal (off an axial) 3 × 3 mm 1 × 1 mm

Hindfoot/midfoot (centered on Chopart joint)

1. Straight sagittal (off an axial) 2. Oblique coronal (off a sagittal) Standard: 3 × 3 mm 3 × 3 mm

C

3. Mortise sagittal (off an axial) 3 × 3 mm 1 × 1 mm

3. Straight axial (off a sagittal) 3 × 3 mm

4. Oblique axial (off a sagittal) 3 × 3 mm

Forefoot/midfoot (centered on Lisfranc joint)

All planes are relative to 1st metatarsal 1. Axial (long axis) (off a sagittal) Standard: 3 × 3 mm

2. Short axis (off a sagittal) 3 × 3 mm

3. Sagittal (off an axial; may have to oblique reference image to see 1st metatarsal) 3 × 3 mm

D

Navicular (stress fracture) Reformat 6 cm FOV Reformat 1 × 1 mm 1. and 2. Coronal and axial (off a sagittal)

3. Sagittal (off an axial) 3

2

1

2 1

3

FIG 64-49  These are portions of the University of Wisconsin foot/ankle/distal tibia (F/A/T) protocol sheet. A, Ankle/distal tibia protocol. This protocol is appropriate for distal tibial fractures (pilon, malleoli, triplane, and juvenile Tillaux) and for talar dome fractures (osteochondral lesions of the talus [OLT], osteochondritis dissecans). B, Hindfoot/midfoot protocol. This protocol is appropriate for hindfoot fractures (calcaneus, talar body, and subtalar joint) and for tarsal coalitions. C, Forefoot/midfoot protocol. This protocol is appropriate for forefoot fractures (Lisfranc dislocation, metatarsals). D, Navicular protocol.

CHAPTER 64  Foot and Ankle Scanning Technique Imaging planes.  We use several different protocols for imaging the ankle and the foot, with the parameters depending on the area of interest and on the technical specifications of the scanner used. Use of a 3-tesla (T) scanner yields improved signal-to-noise and contrast-tonoise ratios and has proven particularly useful in imaging of small joints such as the ankle and foot. The straight sagittal plane (Fig. 64-52A) is our survey plane, and it is usually the first plane acquired in all of our ankle and foot MRI protocols. Most of the other imaging planes are acquired relative to a straight sagittal reference image. The straight sagittal slices are set up off an axial scout image and are oriented parallel to the long axis of the foot. Axial slices through the ankle (see Fig. 64-52B) are set up off a straight sagittal reference image, either a sagittal scout image or one of the midsagittal slices from the preceding acquisition. The axial slices should be roughly perpendicular to the long axis of the tibia, and if the ankle is held in the neutral position will be roughly parallel to the bottom of the foot. The slices should begin well proximal to the level of the malleoli and extend distal to the calcaneus. This is our primary plane for imaging the ankle tendons and is also useful for the ankle ligaments and syndesmosis.

FIG 64-50  This dedicated foot and ankle coil incorporates a chimneylike extension (arrow) so that the phalanges can be included in the FOV.

A

2267

In the imaging of the foot the axial slices (see Fig. 64-52C) are set up off the same straight sagittal reference image as previously but are oriented parallel to the long axis of the metatarsals. This is our primary plane for imaging the tarsal bones, and the slices should include all of the tarsals from the back of the calcaneus through the metatarsal bases. There are at least three ways to orient slices in the coronal plane. Least commonly used is the straight coronal plane. Coronal slices through the ankle are oriented perpendicular to the posterior facet of the subtalar joint (see Fig. 64-52D). Oblique coronal slices (see Fig. 64-52E) are used much more often than straight coronal slices and are set up off the same sagittal reference slice as previously. The oblique coronal slices are oriented perpendicular to the posterior facet of the subtalar joint. (See Fig. 64-7 to review the anatomy of the posterior facet of the subtalar joint.) This is a good secondary plane to evaluate the tendons and tarsals, and the slices should include all of the hindfoot and midfoot. Mortise coronal slices (see Fig. 64-52F) are set up off a straight axial reference image taken through the top of the talar dome. This axial reference image can be either an axial scout image or one of the straight axial slices from a preceding acquisition. The slices are aligned parallel to a line drawn between the medial and lateral malleoli. This is one of the primary planes used for imaging osteochondral defects of the talus (OCDs), and it is also good for looking at the malleoli and the ankle ligaments. Mortise sagittal slices (see Fig. 64-52G) are set up off the same straight axial reference image as previously, and the slices are oriented perpendicular to the mortise coronal slices. This, rather than the straight sagittal plane, is the survey plane we use for OCDs. With regard to the forefoot, we have found that there is some confusion among technologists as to the coronal and axial planes, and to avoid potential ambiguity we refer to the short-axis and long-axis planes relative to the metatarsals. Short-axis images are obtained off a straight sagittal reference image and are oriented perpendicular to the long axis of the metatarsals (see Fig. 64-52H). This yields a series of short-axis slices that cut transversely through the metatarsals and phalanges, an example of which is shown in Figure 64-52I. This is a good plane to evaluate for bone marrow edema in the forefoot. Long-axis images are obtained off a short-axis reference image through the metatarsals and are oriented to include all five, or at least four, metatarsals on individual slices (see Fig. 64-52I). This is the best way to obtain a side-by-side comparison of the metatarsals and is used when evaluating for stress fractures or osteomyelitis.

B FIG 64-51  A knee coil can be used to scan the ankle. A, The open top on this knee coil allows the toes to extend through the coil while keeping the foot in neutral position. B, A customized foam pad (white arrow) helps immobilize the hindfoot being scanned. A second pad next to the coil (black arrow) give the contralateral foot a place to rest.

2268

PART II  CT and MR Imaging of the Whole Body

Straight sagittal

Straight axial

A

Oblique axial

C

B

Straight coronal

Oblique coronal

E

D

m

F

Mortise coronal

Mortise sagittal

G

Short axis Long axis

H

I

CHAPTER 64  Foot and Ankle

2269

FIG 64-52  The standard imaging planes we use for MRI of the foot and ankle. The white lines represent the orientation but not the actual number or spacing of the slices. A, Straight sagittal slices are set up off an axial scout image and are oriented parallel to the long axis of the foot. (Using a cushioned foot holder, as in Figs. 65-48 and 65-50, helps keep the foot in place relative to the scanner.) This is our survey plane, and it is the first plane acquired in all of our ankle and foot MRI protocols. Most of the other imaging planes are acquired relative to a straight sagittal reference image. B, Straight axial slices are set up off a straight sagittal reference image, either a sagittal scout image or one of the midsagittal slices acquired in A. The straight axial slices should be roughly perpendicular to the long axis of the tibia and, if the ankle is held in the neutral position, will be roughly parallel to the bottom of the foot. The slices should begin well proximal to the level of the malleoli and extend distal to the calcaneus. This is our primary plane for imaging the ankle tendons. C, Oblique axial slices are set up off the same straight sagittal reference image as in B and are oriented parallel to the long axis of the metatarsals. This is our primary plane for imaging the tarsal bones, and the slices should include all of the tarsals from the back of the calcaneus through the metatarsal bases. (Although the FOV can be enlarged to include the metatarsals and phalanges in their entirety, we prefer to use the short-axis and long-axis planes delineated in H and I when the clinical question involves the forefoot.) D, Straight coronal slices set up off a sagittal reference image are used primarily when evaluating the plantar fascia and should be oriented roughly perpendicular to the plantar aponeurosis. E, Oblique coronal slices are used much more often then straight coronal slices and are set up off the same sagittal reference slice as in B. The oblique coronal slices are oriented perpendicular to the posterior facet of the subtalar joint. (Refer to Fig. 65-7 to review the anatomy of the posterior facet of the subtalar joint.) This is a good secondary plane to evaluate the tendons and tarsals, and the slices should include all of the hindfoot and midfoot. F, Mortise coronal slices are set up off a straight axial reference image taken through the top of the talar dome. This axial reference image can be either an axial scout image or one of the straight axial slices acquired in B. The slices are aligned parallel to a line drawn between the medial and lateral malleoli. This is one of the primary planes used for imaging osteochondral lesions of the talus (OLT), and it is also good for looking at the malleoli and the ankle ligaments. G, Mortise sagittal slices are set up off the same straight axial reference image as in F, and the slices are oriented perpendicular to the mortise coronal slices. This is the survey plane for OLT. (The marker [m] indicates the site of maximal tenderness.) With regard to the forefoot, we prefer to refer to the short-axis and long-axis planes (rather than coronal or axial) to avoid ambiguity. H, Short-axis images are obtained off a straight sagittal reference image and are oriented perpendicular to the long axis of the metatarsals. A short-axis image through the metatarsals is shown in I. I, Long-axis images are obtained off a short-axis reference image through the metatarsals and are oriented to try to include all five, or at least four, metatarsals on individual slices. This is the best way to obtain a side-by-side comparison of the metatarsals and is used when evaluating for stress fractures or osteomyelitis.

Protocols.  Our ankle/foot MRI protocols call for acquiring images in three of the standard planes and typically take 30 to 45 minutes. We start with straight sagittal images to survey the bones. For an ankle MRI, the FOV needs to include the distal tibia and fibula, all of the tarsal bones, and the bases of the metatarsals. For a forefoot MRI, the FOV needs to include all of the phalanges, all of the metatarsals, and the midfoot tarsal bones back to the Chopart joint. For an MRI of the hindfoot/ankle the FOV should include the midfoot articulations and cover the distal portion of the tibia to include the Achilles tendon myotendinous junction. Sequences.  We use fluid-weighted sequences, including a fatsuppressed fast-spin echo proton density (PD), fat-suppressed T2-weighted sequence or an inversion recovery in all the planes we acquire, usually a combination of these. The inversion recovery sequence is of particular use in the forefoot to counter heterogeneous fat suppression artifacts that are commonly encountered. The PD-weighted sequences are useful for evaluation of the articular cartilage, especially in the talar dome. They are also useful for evaluation of the tendons, having the best signal-to-noise ratio. T1-weighted images are optimal for evaluation of the bone marrow as well as the subcutaneous fat and the deeper fat between muscles and tendons. We use T1 weighting in at least one imaging plane, typically the sagittal. When the tendons are the site of clinical concern, we use PD-weighted images along with T2-weighted sequences in the straight axial and oblique coronal planes. The straight axial plane well images all 10 of the ankle tendons in cross section at and above the level of the ankle joint. The oblique coronal plane well images the medial and lateral ankle tendons in cross section as they curve under the malleoli. Tears in the substance of the ankle tendons are usually best seen with PD-weighted images (Fig. 64-53). However, because PD-weighted images are relatively insensitive for fluid, they should always be read side-by-side with

edema-sensitive images to look for abnormal amounts of fluid in the tendon sheaths, indicative of active tenosynovitis.

Use of Contrast.  Although IV administration of a gadolinium-based contrast agent (IVGd) is not needed for most musculoskeletal MRI, there are particular circumstances in which its use is invaluable.* Whenever possible, we use IVGd when there is a clinical concern for an inflammatory arthropathy or synovitis, such as in rheumatoid arthritis (Fig. 64-54). The IVGd causes T1 signal enhancement of the hypervascularized inflammatory synovium (pannus) but not of the adjacent synovial fluid, a distinction that may not otherwise be seen on noncontrast T1- or T2-weighted images. Our postcontrast images are fat suppressed to make the enhancing tissues more conspicuous. Likewise, we prefer to use IVGd whenever possible in cases of infection. IVGd can help distinguish an enhancing phlegmon from a centrally nonenhancing abscess. Also, IVGd can help distinguish cellulitis, which demonstrates contrast enhancement of the edematous skin and soft tissues, from nonspecific edema from other causes that usually does not enhance. We also use IVGd when there is a clinical question of a palpable mass, to detect if it is solid (Fig. 64-55) or cystic (see Fig. 64-13). Another indication for use of IVGd is the detection of Morton’s neuroma by MRI (Fig. 64-56).

*The U.S. Food and Drug Administration has issued warnings describing the risk for nephrogenic systemic fibrosis after exposure to gadolinium-containing contrast agents in patients with acute or chronic severe renal insufficiency. Information can be found at http:// www.fda.gov/cder/drug/infopage/gcca. Before any contrast agent is administered to any patient, that patient should be screened in accordance with your institution’s policies.

2270

PART II  CT and MR Imaging of the Whole Body

T1

A

B

PD

C

T2

FIG 64-53  Peroneus brevis tear in a 62-year-old. This is a side-by-side comparison of the same straight axial slice acquired using T1-weighted (A), PD-weighted (B), and T2-weighted (C) images. Although T1 shows the fat best and T2 shows the fluid best, PD shows the tendons best, particularly the abnormally increased signal in the split/abnormally flattened peroneus brevis tendon (white arrow).

A

T2fs

T1fs B IVGd

FIG 64-54  Comparison of T2 weighting with fat suppression (A) and T1 weighting with fat suppression after IVGd (B) in a 65-year-old with rheumatoid arthritis. The bright T2 signal in A in the posterior tibial (white arrow) and flexor digitorum longus (white arrowhead) tendon sheaths is shown to be enhancing pannus in B. In comparison, the bright T2 signal in A adjacent to the extensor digitorum longus (black arrow) and the anterolateral ankle joint (black arrowhead) is shown to be nonenhancing fluid surrounded by a thin rim of enhancing synovium in B.

CHAPTER 64  Foot and Ankle

T1

A

B

T1fs IVGd

IR

C

D

FIG 64-55  Schwannoma in a 67-year-old. A, Lateral radiograph shows a round soft tissue mass posterior to the talus. B, Sagittal T1-weighted image shows the mass to be homogeneously relatively dark, with no invasion into the overlying subcutaneous fat or the underlying talus. C, Sagittal inversion recovery image shows the mass to be heterogeneously bright. This is the classic appearance of a schwannoma. D, Sagittal T1-weighted fat-suppressed image after IVGd administration shows heterogeneous enhancement, confirming that this is a vascularized mass and not a cyst.

A

B

T1

m

T2fs

m

T1fs IVGd

m

C

FIG 64-56  Morton’s neuroma in a 44-year-old: short-axis images through the second metatarsal head. The marker (m) indicates the site of maximum tenderness. A, T1-weighted image demonstrates a small lobule of decreased signal (black arrow) between and below the heads of the second and third metatarsals. Morton’s neuromas are seldom more conspicuous than this on T1-weighted images. B, T2-weighted fatsuppressed images of Morton’s neuromas usually show little (open arrowhead) if any edema. C, Administration of IVGd makes the vascularized Morton’s neuroma (arrow) much more conspicuous on this T1-weighted fat-suppressed image.

2271

2272

PART II  CT and MR Imaging of the Whole Body

I O M

F i b u l a

Tibia Tib-Fib Lig

MM lt De g Li

Talus

LM ib -F lo g a i T L

FIG 64-57  Models of the ankle mortise. Left, Skeletal model. Right, Schematic. The intraosseous membrane (IOM) is shown in yellow. The syndesmotic ligaments, the anterior and posterior tibiofibular ligaments (TibFib Lig), are modeled in green. The anterior and posterior talofibular ligaments (Talo-Fib Lig) are modeled in purple. The deltoid ligament (Delt Lig) is shown in blue. LM, lateral malleolus; MM, medial malleolus.

ANKLE AND FOOT INJURIES Ankle Mortise Fractures Malleoli/Syndesmosis.12  Fractures of the medial and lateral malleoli are commonly the result of twisting injury of the talus in ankle mortise. Radiographs are usually sufficient for the management of what are typically simple fractures. CT axial images through both ankles are useful when the integrity of the syndesmosis is questioned. To understand the mechanism of syndesmotic injury, we find it is helpful to review the Weber* staging system for ankle fractures. Figure 64-57 shows two models of the ankle mortise. On the left is a skeleton model showing the relationship of the talus to the malleoli and syndesmosis. On the right is a schematic. The tibia is connected to the fibula by the intraosseous membrane (IOM), a sheet of connective tissue that runs along the length of the diaphyses. Where the distal fibula fits into a groove in the distal tibia is the syndesmosis. The syndesmotic ligaments, the anterior and posterior inferior tibiofibular ligaments, maintain the integrity of this syndesmotic joint. The integrity of the ankle joint is maintained laterally by the anterior and posterior talofibular and calcaneofibular ligaments, and medially by the deltoid ligament. Figure 64-58 illustrates how either inversion or eversion rotational injuries to the talus cause both avulsive and compressive forces on the

*Bernhard Georg Weber (1927-2002), a Swiss orthopedist, nearly gave up medicine and surgery to pursue his dream of becoming an architect. During his surgical training in Zurich, though, he recognized that orthopedics would satisfy his interest in medicine and technology and his need for artistic expression. Besides fracture treatment, he designed a new hip prosthesis and developed a tibial realignment osteotomy procedure to treat prematurely degenerated knees. In fact, he underwent this realignment procedure himself bilaterally to enable him to continue with two of his passions, skiing and tennis. His skill at skiing was such that he was certified as a championship instructor.

malleoli. Figure 64-58A illustrates a Weber type A injury, radiographically on the left and schematically on the right. As the talus undergoes an inversion rotational injury, it applies avulsive pulling forces on the lateral side of the mortise and compressive pushing forces on the medial side. The lateral avulsive forces may cause strain or tearing of the talofibular ligaments, or they may cause an avulsion fracture through the lateral malleolus, pulling it off the fibular shaft. Conversely the compressive forces on the medial side can fracture through the medial malleolus, pushing it away from the tibial plafond. Radiographically, avulsion fractures can be distinguished from compression fractures by the orientation of the fracture margins. Avulsion fractures are horizontally oriented in a direction roughly perpendicular to the lines of force. Compression fractures are more obliquely or vertically oriented in the same direction as the force. This principle is the key to understanding the Weber fractures. Figure 64-58B illustrates a Weber type B injury, radiographically on the left and schematically on the right. Here the talus is undergoing an eversion rotational injury, with the avulsive pulling forces on the medial malleolus and the compressive pushing forces on the lateral side. The medial avulsive forces may cause strain or tearing of the deltoid ligaments, or they may cause a horizontal avulsion fracture through the medial malleolus. The compressive forces on the lateral side cause a vertically oblique fracture through the fibula. If the fibular fracture is distal to the syndesmosis it is characterized as a Weber type B. The syndesmotic ligaments and IOM remain intact. Figure 64-58C illustrates a Weber type C injury, radiographically on the left and schematically on the right. This is the same mechanism as a Weber type B injury, except now the compressive lateral forces extend through the syndesmosis, tearing the tibiofibular ligament as well as the distal IOM. In this case the obliquely oriented fibula fracture will be higher up, above the level of the syndesmosis. Identifying this high fibula fracture is important to recognizing that the syndesmotic ligaments are disrupted, because radiographically the syndesmosis may not appear abnormally widened if not stressed. Indeed, sometimes the fibular fracture is so high that it occurs through the proximal fibula near the knee joint and is thus not imaged

CHAPTER 64  Foot and Ankle on ankle radiographs. This is referred to as a Maisonneuve* fracture and can be suspected when the ankle radiographs demonstrate an avulsion fracture through the medial malleolus without an accompanying fibula fracture. If you cannot tell from ankle radiographs whether you are looking at a Weber type B or C, this is a clue that you may be looking at a Maisonneuve, and radiographs that include the entire length of the fibula should be obtained. Determining the integrity of the syndesmosis is an important surgical consideration because syndesmotic injuries usually require screw fixation. When the integrity of the syndesmosis is unclear based on physical examination and radiographs, a CT scan can be helpful (Fig. 64-59). Scanning in the axial plane through both ankles simultaneously allows for side-by-side comparison of the widths of the injured and uninjured syndesmoses.

Fracture Through Tibial Plafond.  Intraarticular fractures through the tibial plafond often require surgical open reduction with internal fixation (ORIF) to restore the anatomic alignment of the articular surfaces, and multiplanar reformatted CT scans are often instrumental in presurgical planning. Three fractures in particular that typically come to CT are the pilon10 fracture in adults and the juvenile Tillaux42 and triplane44 fractures in adolescents. Pilon fracture.  Pilon fractures include any tibial fracture that involves the distal articular plafond and are typically the result of an axial loading force. Pilon is French for “pestle,” an instrument used for crushing or pounding, and was first used to describe this fracture in 1911 by Étienne Destot, the father of radiology in France. When they are the result of a high-energy injury, such as a fall from height or a high-speed motor vehicle front-end collision, pilon fractures can produce significant comminution with multiple displaced fracture fragments. Although these comminuted fractures invariably require internal fixation, they are typically not surgical emergencies. The patients with significantly displaced fractures may go to the operating room the day of the injury for traction reduction and external fixation to restore relative alignment to the mortise, and then wait several days for the swelling of the surrounding soft tissues to reduce before returning to the operating room for the more anatomic ORIF of the pilon fracture. This means that these patients typically receive their CT scans during this interim period after the external fixator is in place. However, as illustrated in Figure 64-60, such external fixation hardware is no impediment to obtaining the CT images the surgeon requires. To maintain alignment between the hindfoot and leg, the surgeon will percutaneously drill thick metal pins through the calcaneus (white arrows in Fig. 64-60A, B, and D) and through the tibia proximal to the fracture (this pin is not seen in Fig. 64-60). These pins are rigidly attached by metal clamps (white arrowheads in Fig. 64-60A and B) to nonmetallic connecting bars (gray arrows in Fig. 64-60A to C). It is these nonmetallic bars that span the length of the fracture and maintain the tibia length. Because these nonmetallic bars are made of materials (usually carbon fiber) that block very few x-rays from reaching the detectors, they are nearly radiolucent and cause no CT streak artifacts (see Fig. 64-60C). The metal pin-bar clamps block many x-rays from reaching detectors and thus will cause streak artifacts. However, because the clamps are always placed proximal and distal to the pilon fracture, they never cause any CT streak artifacts

*Jules Germain François Maisonneuve (1809-1897), a French surgeon and a student of Guillaume Dupuytren, was the first to describe external rotation as a contributing mechanism in the production of ankle fractures.

2273

across the reformatted fracture margins (see Fig. 64-60B and D). Using our standard bone CT scanning protocol of thin/overlapping slices, metallic streak artifacts are often not appreciable. Notice the good visualization of the calcaneus cortex in Figure 64-60B and D, which is only minimally affected by streaking caused by the metal pin-bar clamps. Juvenile tillaux fracture.42  Juvenile Tillaux fractures are SalterHarris type 3 fractures.† These fractures have a characteristic appearance, particularly on CT. The fracture is the result of an external rotation force pulling on the anterior tibiofibular ligament, causing avulsion of the anterolateral corner of the distal tibial epiphysis (Fig. 64-61A). These fractures always occur laterally because the distal tibial physis fuses from medial to lateral as a child matures (see Fig. 64-61B). As such, juvenile Tillaux fractures occur exclusively in adolescents in whom the medial growth plate has fused but the lateral remains open, usually between the ages of 12 and 15 years. Coronal and sagittal images are useful to demonstrate the degree of displacement, particularly at the articular surface (see Fig. 64-61B and C, white arrow). Minimally displaced juvenile Tillaux fractures are usually treated nonoperatively, but fractures displaced more than 2 mm should have orthopedic consultation and surgery to restore the congruity of the joint surface. Triplane fracture.  Triplane fractures are Salter-Harris type 4 fractures. Like the juvenile Tillaux fracture, triplane fractures occur in adolescents in whom the lateral growth plates have not yet fused. When minimally displaced, triplane fractures can be difficult to see radiographically, and frontal and lateral views are needed to appreciate their multiplanar nature (Fig. 64-62A to C): the epiphysis fracture running vertically in a sagittal orientation (plane 1), the physeal fracture running horizontally in the axial plane (plane 2), and the metaphyseal fracture running obliquely vertically in a coronal orientation (plane 3). Multiplanar CT scans are ideally suited to visualize these fractures in all planes (see Fig. 64-62D to F) and often reveal more deformity of the articular surface than would be anticipated from radiographs alone.

Talar Fractures3,47 Talar fractures can be either traumatic or insidious. Traumatic fractures are considered surgical emergencies because of the high risk of avascular necrosis, and patients typically undergo ORIF before obtaining a CT. Even with anatomic internal fixation, the risk of avascular necrosis persists, and CT or MR can be useful to confirm the presence of abnormal medullary sclerosis or abnormal marrow signal, suspected radiographically (Fig. 64-63).

Osteochondral Defects of the Talus.  Fractures of the talar dome are insidious. They typically occur at the medial edge or posterolateral corners of the talar dome and are thought to be the result of an impaction of the talar dome on the tibial plafond during an inversion or Text continued on p. 2279

†

The Salter-Harris system is applied to fractures that involve the growth plate (physis) at the ends of skeletally immature bones. Type 1 refers to simple transverse fractures that involve the physis only. Type 2, the most common, refers to fractures that involve the physis and the adjacent metaphysis. Type 3 fractures extend from the physis through the epiphysis at the end of the bone, typically disrupting the articular surface at a joint. Type 4 fractures involve the epiphysis, physis, and metaphysis. Type 5 fractures are rare and are crush injuries to the growth plate.

PART II  CT and MR Imaging of the Whole Body

Weber A I O M

F i b u l a

Tibia Tib-Fib Lig

Compression

Avulsion

MM

LM

Talus

VE R SI

O

IN

b Fi loTa Lig

N

lt De g Li

A Weber B I O M

F i b u l a

Tibia Avulsion

lt De g Li

Compression LM

s Talu

EV

b Fi loTa Lig

ER SIO

N

MM

Tib-Fib Lig

B Weber C I O M

F i b u l a

Tibia Tib-Fib Lig

MM lt De g Li

EV

C

LM

Talus

ER SIO

N

2274

b Fi loTa Lig

CHAPTER 64  Foot and Ankle

2275

FIG 64-58  Weber injuries. Structures are as identified in Figure 64-57. A, Weber type A. Left, Anteroposterior radiograph of the ankle showing medial displacement of the talus relative to the tibia, a horizontal avulsion fracture through the lateral malleolus, and a vertically oriented compression fracture through the medial malleolus. Right, Schematic showing the mechanism of a Weber type A ankle fracture. As the talus undergoes an inversion rotational injury, it applies avulsive pulling forces on the lateral side of the mortise and compressive pushing forces on the medial side. B, Weber type B. Left, Anteroposterior radiograph of the ankle showing lateral displacement of the talus relative to the tibia, a horizontal avulsion fracture through the medial malleolus, and an obliquely vertically oriented compression fracture through the distal fibula below the level of the syndesmosis. Right, Schematic showing the mechanism of a Weber type B ankle fracture. As the talus undergoes an eversion rotational injury, it applies avulsive pulling forces on the medial malleolus and compressive pushing forces on the fibula. C, Weber type C. Left, Anteroposterior radiograph of the ankle showing a horizontal avulsion fracture through the medial malleolus and an obliquely vertically oriented compression fracture through the distal fibula above the level of the syndesmosis. The syndesmosis is disrupted and abnormally widened, with no overlap between the tibia and fibula. Right, Schematic showing the mechanism of a Weber type C ankle fracture. This is the same as a Weber type B, except now the compressive forces extend through the syndesmosis, tearing the tibiofibular ligaments and the distal intraosseous membrane (IOM), with the oblique fracture higher up on the fibula. (If the compressive forces extend proximally up the length of the IOM, fracturing through the proximal fibula up near the knee, this is referred to as a Maisonneuve fracture [not illustrated].)

A

Right

Left

B

FIG 64-59  CT of Weber type C injury. A, Axial images through both ankles show the abnormally widened left syndesmosis (black arrows) compared with the width of the contralateral normal right syndesmosis (white arrows). B, Mortise coronal image shows widened syndesmosis (black arrows) and the high fibula fracture (white arrow), characteristic of a Weber type C injury.

2276

PART II  CT and MR Imaging of the Whole Body

L PORT

A

B

C

D FIG 64-60  Pilon fracture. A, Anteroposterior radiograph showing an external fixation device. The radiopaque hardware that could potentially cause streak artifacts on CT—the thick metal pin through the calcaneus (white arrows), the distal pin-bar clamps (white arrowheads), and the proximal pin-bar clamps (black arrowheads)— are below and above the pilon fracture and thus will not be in the axial CT scanning plane through the fracture. The longitudinal carbon fiber connecting bars are barely radiopaque, and as such they are barely discernible on this radiograph (black arrows). These will cause no CT streak artifacts. B, Coronal plane CT scan. The carbon fiber connecting bars (black arrows) cause no CT streak artifacts across the fractures. The CT streak artifacts from the metal percutaneous pin (white arrows) and pin-bar clamps (white arrowheads) are all distal to the pilon fracture and only minimally effect visualization of the calcaneus cortex. C, Axial plane CT scan through the level of the fractured plafond. The carbon fiber connecting bars (arrows) cause no CT streak artifacts across the fractures. D, Sagittal plane showing the talar dome impacted into a large cortical gap in the plafond. This is the type of visual information the surgeon needs to plan the open reduction and internal fixation. The arrow shows the percutaneous pin passing through the calcaneus.

CHAPTER 64  Foot and Ankle

A

B

RIGHT

LEFT

C FIG 64-61  Juvenile Tillaux fracture in a 13-year-old who reported hearing or feeling a snap when, during cheerleading, she landed very forcefully on the left foot with the ankle twisted. A Salter-Harris type 3 fracture was seen on outside radiographs (not shown). A CT series was requested to assess the degree of fracture displacement. A, Axial CT image through both distal tibial physes demonstrates the avulsion fracture of the left anterolateral quadrant (sad face). B, Coronal CT image shows the Salter-Harris type 3 fracture with a longitudinal component through the epiphysis (arrow) and a transverse component through the unfused lateral physis (white arrowheads). The fused medial physis is indicated by the black arrowheads. C, Sagittal CT image shows the Salter-Harris type 3 fracture with a longitudinal component through the epiphysis (arrow) and a transverse component through the unfused physis (arrowheads). Because CT showed that the fracture fragments were displaced more than 2 mm, open reduction and internal fixation was performed electively 1 week after the injury. Postoperatively the patient did well after being non–weight bearing in a cast for 6 weeks and in a weightbearing boot for 4 weeks.

2277

1

1 2

A

2

B

3 2

C

D

1

3

2 2

E

F

FIG 64-62  Triplane fracture in a 13-year-old who twisted an ankle in a sledding accident. Non-weightbearing anteroposterior (A) and mortise (B) radiographs. When minimally displaced, the fracture margins can be difficult to see on radiographs. The black arrow points to the epiphysis fracture running vertically in the sagittal plane (plane 1). The white arrow points to the physis fracture running horizontally in the axial plane (plane 2). C, Lateral non-weightbearing radiograph. The arrow points to the physis fracture running horizontally in the axial plane. The arrowheads point to the metaphysis fracture running obliquely vertically in the coronal plane (plane 3). D to F, CT scanning was performed after closed reduction and casting to assess the degree of fracture displacement. D, Axial CT scan. The avulsion fracture of the anterolateral quadrant (sad face) resembles the juvenile Tillaux fracture (see Fig. 65-61A). The surrounding plaster cast causes no streak artifacts and helps immobilize the patient’s ankle during scanning. E, Coronal CT scan. The black arrow points to the epiphysis fracture running vertically in the sagittal plane (plane 1). The white arrow points to the physis fracture running horizontally in the axial plane (plane 2). F, Sagittal CT scan. The arrow points to the physis fracture running horizontally in the axial plane (plane 2). The arrowheads point to the metaphysis fracture running obliquely vertically in the coronal plane (plane 3). These images clearly showed the surgeons that the closed reduction still had unacceptable displacement, and open reduction and internal fixation was performed the next day. After surgery, the patient was non–weight bearing in a cast for 4 weeks and was pain free after 1 week in a walking boot.

CHAPTER 64  Foot and Ankle

A

B

C

D

2279

FIG 64-63  Development of avascular necrosis (AVN) of the talus after trauma in a 25-year-old who was transferred to our emergency department with the ankle already in a cast. A, Our initial casted lateral radiograph revealed a vertical fracture (arrowhead) through the body of the talus. Because of the risk of AVN with talus fractures, the patient was immediately taken to the operating room. B, Intraoperative radiograph reveals anatomic reduction of the fracture with two screws. No sclerosis is present in the talus. C, On a lateral radiograph obtained 8 weeks later, the body of the talus appears more sclerotic than the surrounding bones. D, Midsagittal CT scan obtained 5 days after the CT scan in part C revealed a broad band of sclerosis in the talar dome and body, characteristic of AVN.

eversion twisting injury. Refer to the illustrations of Weber injuries  (see Fig. 64-58). Because they involve the cortical bone and the overlying articular hyaline cartilage, they are referred to as osteochondral fractures. A gross example is indicated by the green arrow in Figure 64-2. Osteochondral fractures notoriously occur on convex articular surfaces, including the femoral condyles of the knee and capitellum of the elbow. Several names have been traditionally used for these; we prefer the term osteochondral defects. The term osteochondral lesion is open to confusion as to the presence of an underlying mass, while the older term osteochondritis dissecans is misleading, for although the suffix “itis” by definition implies inflammation, histologically these lesions have not been shown to be inflammatory. Radiology and computed tomography.  Although the development of a symptomatic osteochondral defect of the talus (OCD) can often be traced to a specific injury, radiographs are usually read as normal early on. In part this is because many of these fractures are so nondisplaced that they can be difficult to see radiographically (Fig. 64-64). But sometimes, even in retrospect, the initial radiographs truly are negative, and it may take months for the OCD to become radiographically apparent (Fig. 64-65). Thin-slice (1-mm) CT can be used to assess for these defects. Coronal and sagittal reconstructed images are valuable for their detection. Magnetic resonance imaging and staging.  Although CT is good at showing a displaced fragment and the size of the talar dome defect, MRI is the study of choice because it is better at showing the integrity of the overlying articular hyaline cartilage and the underlying bone marrow and therefore can be used to stage the OCD and assess for healing potential or need for surgery.

Several staging systems have been proposed. In 1959, Berndt,8 an orthopedic surgeon from the Cleveland Clinic, working with Harty, an anatomist from the University of Pennsylvania, analyzed 24 cases of what they called “transchondral fractures of the talus.” In the process of tabulating their data, “an arbitrary classification was developed to aid understanding of the mechanism of the fracture and to help in determining the appropriate treatment.” This staging system was based solely on the radiographic appearance of the fracture: Stage I: a small compression fracture Stage II: incomplete avulsion fragment Stage III: complete avulsion without displacement Stage IV: avulsed fragment displaced within the joint Thirty years later, Anderson and colleagues2 from Australia modified this staging system based on the MRI appearance of the fracture. The following staging system was developed: Stage I: subchondral trabecular compression; radiographically occult but bone marrow edema on MRI Stage IIA: subchondral cyst formation Stage IIB: incomplete separation of the fragment Stage III: fluid undermining a nondisplaced nondisplaced fragment Stage IV: displaced fragment Anderson called stage I “subchondral trabecular compression” and defined it as radiographically negative, but with bone marrow edema on MRI (Fig. 64-66). He called stage II “incomplete separation of the fragment,” requiring demonstration of an intact attachment by either CT or MRI (Fig. 64-67). Anderson added a stage IIA, “formation of a subchondral cyst” (Fig. 64-68). Stage IIA cysts are thought to develop from stage I injuries with posttraumatic necrosis of bone and

2280

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 64-64  Acute talar dome fracture in a 23-year-old who fell from a ladder. A and B, Anteroposterior and mortise radiographs in which the cortical fracture of the lateral corner of the talar dome is so nondisplaced it is barely discernible (arrow). C and D, Mortise coronal CT images obtained the same day well demonstrate the cortical fragment (arrow) as well as the full extent of the fracture (arrowheads).

subsequent resorption of the necrotic trabeculae, leaving behind a subchondral cyst. Anderson stage III, “unattached, undisplaced fragment,” is the same as Berndt and Harty stage III. Anderson noted, “In the T2 weighted image, the presence of synovial fluid around a large fragment can help to differentiate between stages II and III” but went on to question the utility of MRI over CT in making this determination (Fig. 64-69). Anderson stage IV, “displaced fragment,” is the same as Berndt and Harty stage IV (Fig. 64-70). Evaluation of the OCD with MRI is useful to determine if an OCD is stable or unstable, with subsequent changes in management. Stable fragments are defined as being fixed firmly with fibrous tissue or fibrocartilage, and these patients typically do not require surgery (Fig. 64-71). Unstable fractures on the other hand require more aggressive treatment with surgery or prolonged immobilization. Several MRI findings suggest instability, and these include a detached fragment, fluid undermining the OCD, abnormalities of the overlying cartilage such as chondral thinning or disruption, and peripheral subchondral cysts in the bone surrounding the ODT (Fig. 64-72). On the contrary, signal within the fragment itself is not useful in determining stability. Once the diagnosis of OCD has been established, the decision as to whether to treat the patient conservatively or surgically often comes down to determining whether the fracture is stable and has a potential for continued healing, or unstable and at risk of dislocating.

Lateral Process of Talus.  The lateral process of the talus (LPT) is the pointed anterolateral corner of the posterior facet of the subtalar joint, indicated by the brown arrow on gross Figure 64-4C and on sagittal CT Figure 64-7A. LPT fractures are the result of trauma, often athletic trauma. Snowboarding in particular is so often cited that fractures of the LPT are also referred to as snowboarder’s ankle.53 The LPT fracture lines tend to be transversely oriented (Fig. 64-73), although vertically oriented LPT fractures can occur (Fig. 64-74). LPT fractures are often difficult to see radiographically (Fig. 64-75A) and are best imaged with CT. Because LPT fractures are typically transversely oriented in the axial plane, they are best visualized in the sagittal (see Fig. 64-75B) and oblique coronal planes (see Fig. 64-75C) to appreciate the

size of the fracture fragment as well as the extension of the fracture line into the subtalar joint. Like OLT, LPT fractures are often diagnosed months after injury, and reports in the orthopedic literature state that “40% are missed at initial presentation.” It is incumbent on anyone who looks at radiographs of the ankle to scrutinize the LPT on all views because these fractures can be subtle and sometimes are seen only on frontal views (Fig. 64-76).

Calcaneal Fractures6,19,40 The calcaneus is the most commonly fractured tarsal bone, with fractures typically occurring as the result of traumatic axial loading, such as can occur with a front-end automotive collision or falling from a height and striking the ground feet first. As a clinical note, patients who present to the emergency department with bilateral calcaneal fractures should also be evaluated for lumbar spine fractures at the time of the initial trauma workup; the same traumatic axial loading can result in a lumbar spine burst fracture with fragments retropulsed into the vertebral canal (Fig. 64-77). Calcaneal fractures can usually be recognized on the lateral radiograph by the presence of lucent fracture lines and displaced fragments (see Fig. 64-77A) or by a compression deformity of the calcaneus with flattening of “Böhler’s* angle” (see Fig. 64-77B).9 With traumatic axial loading, the wedge-shaped LPT is driven into the calcaneus at the angle of Gissane, fracturing and depressing  the calcaneus (Fig. 64-78; see Fig. 64-78B and J). This fracture invariably involves the calcaneal articular surface of the posterior facet of  the subtalar joint (see Figs. 64-77K and 64-78C). The fracture then propagates inferiorly and medially (see Fig. 64-78C), involving the Text continued on p. 2292

*Lorenz Böhler (1885-1973) is most notable as the creator of modern accident surgery. He was the head of the AUVA-Hospital in Vienna, Austria, that was later named for him. This hospital was an international model during his time as the leading surgeon there.

A

B

C

D

E

Ta

MM

MM

Ta LM

LM

F

RIGHT

LEFT

FIG 64-65  Osteochondral lesion (OLT) in a 9-year-old with right ankle pain without any specific trauma. A, Mortise radiograph, even in retrospect, is negative for a lesion in the medial talar dome. B, Nine months later, the same mortise view reveals a subtle lesion in the medial talar dome (arrows). C and D, CT scans obtained 1 month after the radiograph in B well demonstrate the medial osteochondral lesion of the talus (arrows). The difference between C and D is the way the images were reformatted: C was reformatted using our hindfoot protocol (3 × 3 mm in the oblique coronal plane), whereas D was reformatted using our specialized OLT protocol (1 × 1 mm in the mortise coronal plane). The thinner, 1-mm reformatted images yield edges with sharper margins. E, Mortise coronal reformatted CT well shows the extent of the OLT (arrows). F, Axial source images through both ankles demonstrate not only the symptomatic OLT in the posterior medial corner of the right talar dome (black arrows), but an asymptomatic OLT in the posterior medial corner of the left talar dome (white arrows). The patient was treated conservatively and was asymptomatic bilaterally. LM, lateral malleolus; MM, medial malleolus; Ta, talus.

2282

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 64-66  Anderson stage I osteochondral lesion of the talus—subchondral trabecular compression. Mortise coronal T1-weighted (A) and T2-weighted fat-suppressed (B) images show bone marrow edema (arrows) emanating from the lateral corner of the talar dome. The overlying cortex and cartilage are intact. (Courtesy Richard Kijowski, MD.)

A

B

C

D

FIG 64-67  Anderson stage II osteochondral lesion of the talus—incomplete separation of fragment. A, Mor­ tise radiograph; B, mortise coronal CT scan; C, mortise coronal T1-weighted MRI; D, mortise coronal T2-weighted fat-suppressed MRI. The black arrow points to the fragment and the black arrowheads to the donor site. The white arrow in B and D shows where the fragment is still attached to the donor site. (Courtesy Richard Kijowski, MD.)

A

B

C

D

FIG 64-68  Anderson stage IIA osteochondral lesion of the talus—formation of a subchondral cyst. A, Mortise radiograph; B, mortise coronal CT scan; C, mortise coronal T1-weighted MRI; D, mortise coronal T2-weighted fat-suppressed MRI. All images show the subcortical cyst of the medial talar dome with a thin sclerotic border (black arrows). The overlying cortex is intact except for a small focal irregularity (short white arrow). There is edema of the underlying bone marrow (long white arrow). (Courtesy Richard Kijowski, MD.)

A

B

C

D

FIG 64-69  Anderson stage III osteochondral lesion of the talus—unattached undisplaced fragment. A, Mortise radiograph; B, mortise coronal CT scan; C, mortise coronal T1-weighted MRI; D, mortise coronal T2-weighted fat-suppressed MRI. All images show the unattached nondisplaced fragment (short arrows). Arrowheads point to the donor site. Long arrow points to edema at the donor site. (Courtesy Richard Kijowski, MD.)

2284

PART II  CT and MR Imaging of the Whole Body

A

B

C

D

FIG 64-70  Anderson stage IV osteochondral lesion of the talus—displaced fragment. A, Mortise radiograph; B, mortise coronal CT scan; C, mortise coronal T1-weighted MRI; D, mortise coronal T2-weighted fatsuppressed MRI. All images show the displaced fragment (arrowheads) as well as the talar donor site (short arrows). Long arrows point to marrow edema at the donor site. (Courtesy Richard Kijowski, MD.)

CHAPTER 64  Foot and Ankle

A

T1

C

B

T2fs

D

FIG 64-71  Osteochondral lesion of the talus (OLT) in a 26-year-old with a remote history of an ankle strain, with diffuse ankle pain for the past year. Anteroposterior (A) and mortise (B) radiographs demonstrate the OLT of the medial talar dome (arrow). MRI was obtained 1 week later. C, Mortise coronal T1-weighted image demonstrates the OLT of the medial talar dome (arrow). D, The corresponding mortise coronal T2-weighted fat-suppressed image shows no bright signal around the OLT, indicating that it is stable.

2285

2286

PART II  CT and MR Imaging of the Whole Body

A T1

C

B T2fs

D

FIG 64-72  Osteochondral lesion of the talus (OLT) in a 14-year-old with ankle pain. Anteroposterior (A) and mortise (B) radiographs demonstrate a subtle OLT of the medial talar dome (arrow). MRI was obtained 2 months later. C, Coronal T1-weighted image demonstrates the OLT of the medial talar dome (arrow). D, The corresponding coronal T2-weighted fat-suppressed image shows a bright line of fluid (arrows) around the OLT, indicating it is unstable.

CHAPTER 64  Foot and Ankle

A

B FIG 64-73  Lateral process of the talus (LPT) fracture in a 17-year-old gymnast who landed awkwardly after a vault. A, Lateral radiograph demonstrates the slightly displaced transversely oriented LPT fracture (arrowheads). B, Sagittal CT confirms the LPT fracture (arrowheads) seen in A. Given the relatively small size of this fracture, the patient was treated nonoperatively with casting and then with physical therapy. (This scan was performed using an older protocol, with source images 1 mm thick at 1-mm intervals. This lack of overlap yields reformatted images with some stairstep artifacts that can be seen in the metatarsal shaft. This artifact can be avoided by reconstructing source images with a 50% overlap.)

FIG 64-74  Subtle lateral process of the talus (LPT) fractures in a 28-year-old who sustained multiple injuries from a motor vehicle collision. Lateral radiograph demonstrates a minimally displaced vertically oriented LPT fracture (arrowhead).

2287

2288

PART II  CT and MR Imaging of the Whole Body

os trigonum

bone island

A

B

C

FIG 64-75  Fracture of lateral process of the talus (LPT) in a 22-year-old who walked away from a motor vehicle collision and presented 1 day later with ankle pain. A, Lateral radiograph does not clearly demonstrate the fracture. Incidentally noted is an os trigonum (white arrow) and a bone island (black arrow), both of no clinical significance. B and C, CT scans obtained the same day as A, reformatted in the direct sagittal (B) and oblique coronal (C) planes. Both planes well demonstrate the transverse LPT fracture, with extension into the posterior facet of the subtalar joint (arrowheads). The patient did well after 6 weeks of non–weight bearing, and no surgery was required.

A

C

B

M-STJ

D

F

E

G

FIG 64-76  Fracture of lateral process of the talus (LPT) in a 45-year-old who was the driver in a front-end automobile collision. A, Lateral radiograph does not clearly demonstrate the LPT fracture. There is a small ossicle (arrow) just behind the subtalar joint that could be mistaken for an os trigonum but is in fact a small fragment off the posterior corner of the talus. B, Mortise radiograph shows a tiny ossicle (arrow) between the LPT and lateral malleolus, too small to characterize. C, Anteroposterior radiograph reveals the large LPT fragment (white arrow) as well as a medial fragment (black arrow). D and E, CT scans obtained the same day as the radiographs, reformatted in the oblique coronal plane, perpendicular to the subtalar joint. D, Image through the middle facet of the subtalar joint (M-STJ) shows the large LPT fragment (arrow). E, A more posterior image demonstrates the LPT fracture extending into the posterior facet (arrowhead), as well as the separate fragment off the medial talus (arrow). Surgery was performed 1 week later after the soft tissue swelling had diminished. Lateral (F) and mortise (G) radiographs were obtained after the LPT fracture was repaired with two screws. (There is also a Mitek suture anchor in the lateral malleolus.)

2290

PART II  CT and MR Imaging of the Whole Body

LEFT

RIGHT

B

A

C

D

L1 L1

E

F

L1

G

FIG 64-77  Bilateral calcaneal fractures in a 25-year-old who fell from a three-story parking garage, landing feet first. Lateral radiographs of the left (A) and right (B) ankles were obtained. In the left ankle, lucent fracture lines are clearly seen (arrowheads). In the right ankle, Böhler’s angle is flattened (compare with part N, after open reduction and internal fixation [ORIF]). As part of the trauma workup, a CT scan of the abdomen and pelvis was performed, hence the presence of oral contrast in the colon on the anteroposterior scout image (C). Because of the mechanism causing bilateral calcaneal fractures, it was necessary to evaluate the lumbar spine for fractures. The same raw data from the large-FOV scan of the abdomen and pelvis were reconstructed into thin overlapping small-FOV images centered on the lumbar spine as source images (D). The arrowheads point to a fracture through the anterosuperior end plate of L1. E, Sagittal reformatted CT image of the lumbar spine shows the thin fracture through the anterosuperior corner of L1 (arrowhead). F and G, MR sagittal T1- and PD-weighted images do not well demonstrate the L1 fracture.

L1 L1

I

H

LPT

J

M-STJ LPT

K

ST

P-STJ

L

FIG 64-77, cont’d  H and I, MR sagittal and coronal fat-suppressed T2-weighted images show bone marrow edema throughout the superior halves of the L1 and L2 vertebral bodies. The next day, a CT scan was performed through both ankles simultaneously and reformatted in multiple planes for each ankle individually using our hindfoot protocol. Displayed are images of the right calcaneus. J, Sagittal image through the lateral process of the talus (LPT). The calcaneus is fractured just below the LPT, where the wedge-shaped LPT was driven into the calcaneus at the angle of Gissane. The small back spots with the fractured calcaneus are air, indicating that this was an open fracture that was subsequently reduced. Coronal oblique images through the posterior (K) and middle (L) facets of the subtalar joint were obtained. These calcaneal fractures typically begin with impaction from the LPT, and there is extension into the posterior facet (white arrowhead). In this patient, the fracture also extends through the sustentaculum tali (ST) into the middle facet (black arrowhead). Continued

2292

PART II  CT and MR Imaging of the Whole Body

CCJ

Böhler’s angle

M

N

FIG 64-77, cont’d  M, Axial image through the calcaneocuboid joint (CCJ) shows that this joint is not involved in this patient. N, After ORIF of the calcaneal fracture, Böhler’s angle is restored (compare with B).

sustentaculum tali and the middle facet to varying degrees (see Figs. 64-77L and 64-78D). Assessment of the integrity of the middle facet of the subtalar joint is an important part of preoperative surgical planning. Surgeons prefer to operate on the calcaneus from the lateral side, meaning that they will not directly visualize the middle facet and sustentaculum tali. Thus they require the preoperative CT scan to evaluate these structures. For this reason, 2D and 3D reconstructions are crucial in the assessment of calcaneal fractures. Close attention should be made also to assess for extension into the calcaneocuboid joint (see Fig. 64-78E). Calcaneal fractures are not surgical emergencies and are typically treated surgically several days after the injury. Therefore treatment of more urgent injuries such as a lumbar spine fracture is usually warranted. As a result, the preoperative CT scan of the calcaneus does not need to be performed emergently when there may be other more serious injuries that need to be addressed.

Anterior Process of the Calcaneus.  The APC is the upper outer corner of the calcaneus where it articulates with the cuboid, indicated by the orange box on gross Figure 64-4C and the red arrow on sagittal CT Figure 64-7A. Like LPT fractures, APC fractures can be easily overlooked, and this structure should be carefully scrutinized on all lateral radiographs of the ankle and foot (Fig. 64-79). APC fractures are more common in women and are the result of an inversion injury while the foot is in plantar flexion, such as when wearing high-heeled shoes. Even when APC fractures are only minimally displaced they have a tendency for nonunion despite prolonged immobilization (Fig. 64-80). CT is useful for both detecting these fractures and following their progress. One potential pitfall in the diagnosis of an APC fracture is the os calcaneus secondarius, an accessory ossicle (Fig. 64-81) that resides

between the APC and the lateral pole of the navicular (see Fig. 64-39). The os calcaneus secondarius can be thought of as a forme fruste of tarsal coalition, and it should not articulate with the cuboid as the APC does. CT can be used to distinguish an acute APC fracture from the normal-variant accessory ossicle (Fig. 64-82).

Lisfranc Dislocation Dislocations along the tarsometatarsal joint are not uncommon. These can be the result of severe acute trauma, but the Lisfranc joint is also a common site for dislocation in diabetic patients with peripheral neuropathy. In a Lisfranc dislocation, the second to fifth metatarsals are dislocated laterally, or dorsolaterally, relative to the tarsal bones. The Lisfranc dislocations are subdivided into two categories based on  what happens to the first metatarsal relative to the other four. If the first metatarsal dislocates laterally along with the second to fifth metatarsals, it is called homolateral (Fig. 64-83). If the first metatarsal diverges from the other four metatarsals, remaining aligned with the medial cuneiform (Fig. 64-84), or if the first metatarsal dislocates medially (Fig. 64-85), it is called divergent. When a Lisfranc fracture is grossly displaced, a CT scan is not needed to confirm the diagnosis. However, because the exact location of dislocated metatarsals may be difficult to discern based solely on radiographs, a 3D reformatted CT scan may prove useful in presurgical planning (see Figs. 64-83H and 64-84F and G). These dislocations can be best appreciated in 3D volume rendered images. When only minimally displaced, Lisfranc dislocations can be difficult to discern radiographically, and close attention should be paid to the Lisfranc joint on all views of the foot. Normally there is perfect alignment between the first metatarsal base and first (or medial) Text continued on p. 2302

LPT NA PC

A

B

LM CCJ M-STJ ST

C

D

E

FIG 64-78  Calcaneal fracture in a 40-year-old who fell from a 4-foot ladder, landing on the heel. A, Lateral radiograph shows a calcaneal fracture. B, CT scan in the sagittal plane shows where the lateral process of the talus (LPT; black arrow) drove into the calcaneus, causing wide separation of the posterior facet of the subtalar joint (bidirectional arrow). Incidentally seen is a nonosseous tarsal coalition (arrowheads) between the anterior process of the calcaneus (APC) and the navicular (N). C, CT scan in the coronal oblique plane through the posterior facet of the subtalar joint demonstrates the typical inferomedial direction of the fracture (dashed arrow). There are often laterally displaced fragments (black arrow). Normally none of the calcaneus should be below the lateral malleolus (LM). D, CT scan in the coronal oblique plane through the middle facet (arrow) shows that in this patient the sustentaculum tali (ST) is in one large piece and that the fracture does not extend into the middle facet. E, Axial CT scan through the calcaneocuboid joint (CCJ) shows involvement of this joint (arrowhead).

APC

FIG 64-79  Anterior process of the calcaneus (APC) fracture in a 48-year-old who fell while standing on a picnic table, sustaining a twisting injury to the foot. This lateral radiograph was obtained in the emergency department the next day. The arrowheads in the magnified dashed box show the minimally displaced lucent fracture lines through the APC. The patient did well after nonoperative treatment with a non-weightbearing cast for 12 weeks.

2294

PART II  CT and MR Imaging of the Whole Body

A

RIGHT

C

B

LEFT

D

FIG 64-80  Anterior process of the calcaneus (APC) fracture in a 29-year-old who tripped down some steps. A, Oblique non-weightbearing foot radiograph shows a nondisplaced APC fracture (white arrowheads in magnified dashed box). B, Radiographs 6 months later show that the APC fracture is still unhealed. CT scans were also obtained on the same day as the initial radiographs. C, Source axial images through both hindfeet reveal the minimally displaced transverse fracture (white arrowheads in dotted magnified box) of the left APC. The contralateral right foot serves as a useful normal comparison when both feet are included in the small scanning FOV. D, Sagittal reformatted image shows the APC fracture disrupting the superior cortex (arrow). The acute fracture margins are not corticated. The patient was initially treated conservatively, including 4 months of non–weight bearing and 4 months with a bone stimulator. When the patient remained symptomatic 7 months later, a repeat CT scan was requested.

CHAPTER 64  Foot and Ankle

RIGHT

E

LEFT

F

G

H

I

J

FIG 64-80, cont’d E, The axial source images reveal that the transverse fracture remains nonunited (arrowheads in magnified dashed box). F, The sagittal image shows that the fracture margins are becoming sclerotic and corticated (arrow), a sign of nonunion. Because CT confirmed the clinical suspicion that the APC fracture was not healing, surgical intervention was warranted. G, Oblique radiograph obtained portably in the recovery room immediately after open reduction and internal fixation shows the lucent fracture line (arrowheads) bridged by a Whipple-type Herbert screw. H, Oblique radiograph obtained 9 months after surgery reveals that the fracture lines are essentially healed and barely discernible (arrowheads). I, Sagittal T2 fat-suppressed and J, T1 images through the anterior process of a talus on a different patient demonstrate a nondisplaced APC fracture with marrow edema.

2295

N

APC O C S

FIG 64-81  The os calcaneus secondarius (OCS) is an occasionally seen normal variant that resides between the anterior process of the calcaneus (APC) and the lateral pole of the navicular (N).

B

A

C FIG 64-82  Anterior process of the calcaneus (APC) fracture in a 34-year-old who presented to the urgent care clinic 1 day after a minor motor vehicle collision, complaining of pain along the lateral midfoot. A, Lateral radiograph shows the small APC fracture resembling an os calcaneus secondarius (arrowheads in magnified dashed box). CT scans were obtained the next day. B, Sagittal scan through the APC demonstrating sharp noncorticated margins (arrowheads in dotted magnified box) on the vertically oriented fracture. C, The axial source scan confirms that this is an acute fracture with sharp but nonsclerotic margins (arrowheads in dashed magnified box).

CHAPTER 64  Foot and Ankle

C I

I

II 1

1

2

II 2

III IV 3

V Cu

A

B

I 1

D

II 2

E

FIG 64-83  Example of progression from normal to neuropathic Lisfranc dislocation. The patient was 49 years of age at presentation and had diabetes mellitus. Anteroposterior (A) and oblique (B) radiographs reveals normal anatomic alignment between the first (1), second (2), and third (3) cuneiforms and the first (I), second (II), and third (III) metatarsals, as well as between the cuboid (Cu) and the fourth (IV) and fifth (V) metatarsals. C, Lateral radiograph shows normal alignment between the midfoot and forefoot. Arterial calcifications (arrow) are present, consistent with the patient’s history of diabetes. The patient returned 2.5 years later. He had been having episodes of passing out and falling, although he did not remember these episodes well. He did not remember injuring himself, and because of peripheral neuropathy he had no sensation in his foot. He first noticed swelling and blisters on his foot the morning the following radiographs were taken. He ambulated normally without the assistance of a walker or cane. Anteroposterior (D) and oblique (E) radiographs illustrate a homolateral Lisfranc dislocation, with lateral dislocations of all five metatarsals. The first metatarsal (I) is not articulating with the first cuneiform (1) but is instead articulating with the second cuneiform (2). The base of the second metatarsal (II; arrowhead) is not articulating with anything. Continued

2297

2298

PART II  CT and MR Imaging of the Whole Body

I

F

II 1

2

III

IV

3 Cu

I

1

II

2

III

IV

V

3

G

Cu

H FIG 64-83, cont’d F, Lateral radiograph shows the dorsal dislocation of the second metatarsal (arrowhead). A CT scan was obtained to understand better the extent of the dislocation. G, Axial oblique scan shows that none of the metatarsals are articulating with their appropriate tarsals. In cases of complex fracture-dislocations, 3D reformatted images can help in understanding the relative locations of the bones. H, This 3D image as viewed from above shows lateral displacement of all five metatarsals, and dorsal dislocation of the second through fourth metatarsals.

V

CHAPTER 64  Foot and Ankle

B

A

C

1 II 2 N

3

III

II

2 3

IV

III IV

Cu

V

E D FIG 64-84  Divergent Lisfranc dislocation in a 34-year-old who was the front passenger in a motor vehicle accident. Radiographs were obtained in the emergency department. Anteroposterior (A) and oblique (B) views of the foot reveal lateral dislocation of the second through fifth metatarsals. The white arrow points to a fragment fractured off the base of the second metatarsal. The black arrowhead points to the base of the fourth metatarsal, which is not articulating with anything. C, The Lisfranc dislocation is less obvious on the lateral view, although the base of the fourth metatarsal (black arrowhead) is not articulating with anything. A closed reduction was attempted at the bedside but was unsuccessful. D and E, CT scans were obtained to aid in surgical planning. Straight axial (long-axis) (D) and coronal oblique (long-axis) (E) images through the Lisfranc joint show that none of the metatarsals (II to V) is articulating properly with its respective tarsal bone. Cu, cuboid; N, navicular. Continued

2299

2300

PART II  CT and MR Imaging of the Whole Body

G

F

H

I

J FIG 64-84, cont’d F and G, 3D reformatted views help in understanding the multiplanar nature of the Lisfranc dislocation. F, Viewed from above, the second through fifth metatarsals can be seen to be dislocated dorsally and laterally. G, The view from below the foot (“Star Wars” view) shows a large fragment off the base of the second metatarsal (white arrow) and a smaller fragment off the third metatarsal (black arrow). Postoperative anteroposterior (H), oblique (I), and lateral (J) radiographs show that seven screws were required to restore the anatomic alignment of the Lisfranc joint.

CHAPTER 64  Foot and Ankle

A

B

D C

OR PORT

FIG 64-85  Divergent Lisfranc dislocation in a 22-year-old who was driving on the highway when struck by another car going the wrong way. The patient also sustained a fracture of the contralateral femoral shaft. Anteroposterior (A) and oblique (B) radiographs demonstrate medial dislocation of the first metatarsal and lateral dislocation of the second through fourth metatarsals. The patient was taken to the operating room for open reduction and internal fixation of the femoral fracture with an intramedullary nail. Once the patient was fully anesthetized, the surgeon was able to apply longitudinal traction on the first ray and achieve closed anatomic reduction of the Lisfranc joint. C, Oblique port intraoperative radiograph. D, The patient returned to the operating room 1 week later, after the swelling had diminished, for elective fixation of the Lisfranc joint.

2301

2302

PART II  CT and MR Imaging of the Whole Body

cuneiform, between the second metatarsal and the second (or middle) cuneiform, and between the third metatarsal and the third (or lateral) cuneiform. Also, the bases of the fourth and fifth metatarsals should be perfectly aligned with their individual facets on the cuboid (see Fig. 64-83A and B). One clue that a nondisplaced Lisfranc dislocation may be present is fracture fragments off the base of the second metatarsal. As shown on the 3D CT in Figure 64-5, the base of the second

metatarsal extends more proximally across the Lisfranc joint than do the other metatarsals. Thus when dislocations occur along the Lisfranc joint, it is common for the base of the second metatarsal to be avulsed. In some cases, axial oblique CT images can help to demonstrate subtle offsets at the tarsometatarsal joints, particularly when compared with the normal contralateral side (Fig. 64-86). In other cases, the dislocations along the Lisfranc joint may be manifest only when a lateral stress

LEFT

LEFT

A

B RIGHT

LEFT

V RIGHT

LEFT I

I III 3

C

II

1

1

2

Cu

Cu

III

II 2

V

3

D FIG 64-86  Subtle Lisfranc dislocation in a 39-year-old who fell backward down a 3-foot wall and injured the left foot. The patient bicycled home and continued to walk on this foot for 3 days before coming to the emergency department, concerned because the pain was not diminishing. Anteroposterior (A) and oblique (B) non-weightbearing radiographs were obtained. Close scrutiny of the Lisfranc joint on the oblique view (dashed box) reveals small fractures off the bases of the second and first metatarsals (arrowheads). The Lisfranc joint appears anatomically aligned. The presence of the fragments along the Lisfranc joint raised the concern that this may represent a Lisfranc dislocation. C and D, CT scans were performed to assess the integrity of the tarsometatarsal joint. C, Axial oblique scan through both medial Lisfranc joints. In the normal right foot there is anatomic alignment across the first, second, and third cuneiform-metatarsal joints. In the injured left foot there is lateral subluxation of the first (white arrow) and third (black arrow) metatarsals as well as small fragments off the second metatarsal (white arrowhead) and second cuneiform (black arrowhead). D, Axial oblique scan, slightly more plantar and more angled than C, through the lateral tarsometatarsal joint. On the normal right side the articular surfaces of the fifth metatarsal (V) and the cuboid (Cu) are aligned (white arrows). On the left, the lateral corner of the fifth metatarsal (black arrow) is laterally displaced relative to the cuboid’s impacted lateral corner (arrowhead). These findings confirmed that the patient had sustained a disruption of the Lisfranc joint, which was treated with a boot and non–weight bearing.

CHAPTER 64  Foot and Ankle

A

C

2303

B

D

FIG 64-87  Subtle Lisfranc dislocation in a 56-year-old nondiabetic patient who misstepped from a high curb and landed awkwardly, injuring the foot. A, Initial radiograph shows the fracture off the base of the second metatarsal (arrow). The alignment of the Lisfranc joint is relatively maintained. Because of the degree of soft tissue swelling, the patient was initially treated with a boot. B, Axial CT scan obtained 5 days later showed the second metatarsal fracture to be essentially nondisplaced and the first and second tarsometatarsal joints to be in anatomic alignment. When the soft tissue swelling subsided 4 days later, the boot was exchanged for a cast. C, Postcasting anteroposterior radiograph reveals that there is lateral displacement of the first (white arrow) and second (black arrow) metatarsals. Because this demonstrated that the Lisfranc joint was not stable, 3 days later the patient was taken to the operating room for open reduction and internal fixation (D).

is applied to the forefoot. In these patients, the nonstressed CT scan may fail to demonstrate the degree of potential displacement (Fig. 64-87). MRI can be particularly useful in assessing for Lisfranc dislocation because the Lisfranc ligament can be directly visualized and its integrity confirmed, while marrow edema within the bases of the metatarsals can confirm recent injury (Fig. 64-88).

Ankle Impingements21,46 Ankle impingement syndromes refer to painful restriction of movement in the ankle joint related to soft tissue and osseous abnormalities. Depending on their anatomic location, they are subdivided into anterolateral, anterior, anteromedial, posteromedial, and posterior impingement.

2304

PART II  CT and MR Imaging of the Whole Body

Gout.  Gout is uric acid crystal deposition arthropathy with a predilection for the foot, particularly the first metatarsophalangeal joint. The cortical erosions caused by gout form slowly and can take as long as a decade to manifest radiographically. These erosions are classically described as being well circumscribed with sharp overhanging edges (Fig. 64-91). The diagnosis of gout is confirmed when an aspirate of joint fluid reveals strongly negative birefringent crystals under a polarizing microscope. Recently the advent of dual-energy CT (DECT) or spectral CT has shown promise for the early detection of gout. DECT, through its inherent ability to differentiate between two compounds of different molecular weight, can assist in the early detection of urate crystal deposits and therefore in the subclinical diagnosis and early treatment of gout.39

Arthrodesis.  Surgical arthrodesis is usually performed in the ankle and foot for pain control, with the downside of losing function at that joint. However, typically patients with such a degree of pain have very limited motion in the affected joint anyway. After an arthrodesis, CT can be used to assess the degree of solid bony bridging. Although there may be several metal plates and screws within the scanning FOV, these tend to cause relatively little CT streak artifact, especially when the source images consist of thin overlapping slices (Fig. 64-92).

Tarsal Coalitions18,38

Anterolateral impingement represents the sequelae of an inversion injury, with disruption of the lateral ligamentous complex and the capsule and subsequent formation of synovial hypertrophy and scarring in the anterolateral gutter, with painful restriction of motion on dorsiflexion (Fig. 64-89). Anterior impingement refers to a predominantly osseous impingement from phytes arising from the anterior aspect of the tibia and the talus, with associated soft tissue edema  (see Fig. 64-89). Posterior impingement, or os trigonum syndrome, is related to a prominent os trigonum or enlarged posterior talar process that causes repetitive microtrauma during plantar flexion (see Fig. 64-34). Hallmarks of this on MRI are marrow edema along with posterior capsular thickening, soft tissue edema, and FHL tenosynovitis. Anteromedial and posteromedial impingement syndromes are less common and refer to capsular and ligamentous thickening with soft tissue edema.

Tarsal coalitions are abnormal unions between bones and are either osseous, with solid cortical bridging, or nonosseous, with a fibrous or cartilaginous union between the bones. In the hindfoot, tarsal coalitions most commonly occur across the middle facet of the subtalar joint and between the APC and the lateral pole of the navicular.52 An example of the latter was already seen as an incidental finding in Figure 64-78B. The subtalar joint complex consists of the subtalar joint itself and the talonavicular and calcaneocuboid joints. These joints function in unison during the gait cycle, and limitation of motion of any one of these joints limits the motion of the other joints.43 The clinical syndrome of tarsal coalition consists of pain and reduced or absent subtalar motion, as well as pes planus (flatfoot deformity) and peroneal muscle spasm (clonus on inversion stress).4,49 The exact cause of the peroneal spasm is uncertain; however, peroneal muscle tightness is the result of tarsal coalition, not the cause. Symptoms usually manifest between 12 and 16 years of age and worsen with increasing age. Conservative treatment options include antiinflammatory medication and a trial of reduced activity, cast immobilization, and molded orthoses. If conservative treatment fails, surgical options include resection of the coalition and arthrodesis if secondary osteoarthritis has developed. Tarsal coalitions are bilateral in the majority of cases (Fig. 64-93).

Arthritis51

Calcaneonavicular Coalition.  Of the two common locations for

The hallmarks of osteoarthritis—nonuniform joint space narrowing accompanied by the formation of osteophytes, subcortical sclerosis, and subcortical cysts—are typically well seen radiographically. But some joints, such as the ankle and subtalar joints, can be difficult to profile radiographically, and in such cases CT and MRI can be used to demonstrate all these osteoarthritic changes (Fig. 64-90).

tarsal coalitions, calcaneonavicular coalitions can often be seen radiographically on the oblique view (Fig. 64-94). Normally the calcaneus and navicular do not touch (Fig. 64-95A); it is abnormal any time they are close enough to each other to form a joint (see Fig. 64-95B), let alone a solid bony bridge. Another radiographic indication of a calcaneonavicular coalition is the presence of an elongated APC, sometimes called an ant-eater sign (see Fig. 64-95C and D). An elongated APC and an asymptomatic calcaneonavicular coalition may be seen as incidental findings on radiographs and CTs obtained for other reasons. And although an elongated APC may not cause a symptomatic coalition, it may be at increased risk of fracture (Fig. 64-96).

FIG 64-88  Axial T2 fat-suppressed image demonstrates thickening of the anterior talofibular ligament with soft tissue thickening and edema (arrow).

Rheumatoid Arthritis.  For rheumatoid arthritis, we prefer MRI to CT when cross-sectional imaging is required. MRI after the administration of IV contrast well demonstrates abnormally enhancing synovium and thickened pannus (see Fig. 64-54) as well as small cortical erosions before they become radiographically apparent.

Text continued on p. 2315

CHAPTER 64  Foot and Ankle

T2 FS

A

B

T1

C

T2 FS

D FIG 64-89  A, Lateral ankle radiograph of a 21-year-old man demonstrates anterior osteophytes in the distal tibia (arrow). B, Sagittal T2 fat-suppressed and C, T1 images show fluid and edema along with the productive changes at the anterior tibia (arrows). The extent of the associated synovitis and edema is seen on axial T2 fat-suppressed images, D (arrow).

2305

2306

PART II  CT and MR Imaging of the Whole Body

A

RIGHT

LEFT

B

RIGHT

LEFT

C FIG 64-90  CT scan of osteoarthritis in a 71-year-old patient. A, Axial scan through the tops of both ankle joints demonstrates nonuniform narrowing of the medial ankle mortise bilaterally (black arrows). Many small geodes with well-circumscribed, corticated margins (black arrowheads) are seen in both talar domes. B, Coronal scan demonstrates nonuniform narrowing medially in both mortises (black arrows), subcortical geodes (black arrowheads), and subcortical sclerosis (white arrowheads). C, Sagittal scan demonstrates nonuniform joint space narrowing (black arrows) as well as a small anterior osteophyte (white arrow).

CHAPTER 64  Foot and Ankle

LEFT

A

B

RIGHT

LEFT

FIG 64-91  A 34-year-old man came to the emergency department complaining of acute onset of nontraumatic pain of his left great toe. A, Anteroposterior radiograph of the left foot, with the area in the dashed box magnified to the right. A thin, white, curved line was observed just outside the lateral diaphyseal cortex of the first metatarsal (ellipse). Initially this was mistakenly thought to present a periosteal reaction from the first metatarsal rather than what it truly was: an eroded lateral sesamoid. There is also a marginal erosion of the medial first metatarsal head (white arrow). B, Axial CT scan through the sesamoids of the great toes bilaterally shows two normal sesamoids on the right, whereas on the left only a thin shell of the eroded lateral sesamoid remains (ellipse). Continued

2307

2308

PART II  CT and MR Imaging of the Whole Body

C

RIGHT

D

RIGHT

LEFT

LEFT

FIG 64-91, cont’d  C, Short-axis (coronal) CT scan confirms the erosion of the left lateral sesamoid (ellipse) as well as an erosion in the adjacent metatarsal head (arrow). D, Axial CT scan proximal to B reveals the marginal erosion seen radiographically in the left medial first metatarsal head (white arrow) and an erosion in the right second cuneiform (black arrow). Both erosions have well-defined, slightly sclerotic margins with sharp overhanging edges, characteristic of gout. Aspiration of the patient’s left great toe metatarsophalangeal joint yielded uric acid crystals.

CHAPTER 64  Foot and Ankle

A

C

B

D

E

FIG 64-92  Attempted arthrodesis. The patient is a 68-year-old farmer who injured his ankle 6 years earlier when he misstepped getting off his tractor. He underwent arthrodesis surgery 3 years ago, and this was revised 1 year ago because of failure of fusion. The patient is experiencing persistent pain. Anteroposterior (A), mortise (B), and lateral (C) radiographs reveal a plate and several screws across the ankle mortise and syndesmosis. Although the metal obscures visualization of portions of the mortise, no bony fusion is seen medially (black arrowheads). D and E, CT scans were requested to see if any fusion was present. Mortise coronal (D) and mortise sagittal (E) images clearly show no bony bridging throughout the ankle mortise (black arrowheads). Although there are some metallic streak artifacts, the resolution is high enough to visualize the widely spaced cancellous threads of the lag screw in the talus (white arrow) and the narrowly spaced cortical threads of the syndesmotic screw (white arrowhead). Continued

2309

2310

PART II  CT and MR Imaging of the Whole Body

H

G

F

Ti

Ti Fi Ta

I

Ta

J

FIG 64-92, cont’d One operation and 16 months later, anteroposterior (F), mortise (G), and lateral (H) radiographs no longer demonstrate residual lucency along the mortise. Mortise coronal (I) and mortise sagittal (J) CT scans now reveal solid bony fusion between the tibia (Ti), talus (Ta), and fibula (Fi).

A

RIGHT

N A P C

B

N

H

H APC

RIGHT

LEFT

N APC

OCS

N APC

C

RIGHT

LEFT

FIG 64-93  Calcaneonavicular coalition in an 11-year-old with right foot pain. A, Oblique radiograph shows the abnormal joint in this nonosseous coalition (arrowheads in magnified dashed box). B, Axial oblique CT scan through the bottoms of the talar heads (H) shows the symptomatic coalition on the right foot (black arrowhead) between the elongated anterior process of the calcaneus (APC) and the navicular (N). Incidentally seen is an asymptomatic coalition on the left (white arrowhead) between the abnormally broad APC and the navicular. C, Axial oblique CT scan slightly plantar to B. On the right foot is the symptomatic abnormal joint (arrowhead) between the broad APC and the navicular. On the left is an extra bone, an os calcaneus secondarius (OCS), articulating with both the APC and navicular. Continued

2312

PART II  CT and MR Imaging of the Whole Body

N

D

N

APC

RIGHT

E

APC O C S LEFT

Poin

ter

F

G

H

FIG 64-93, cont’d D, Sagittal CT scan of the right foot shows the nonosseous coalition (arrowhead) between the navicular and APC. E, Sagittal CT scan of the left foot shows the OCS between the navicular and APC. Surgery was elected. F and G, Fluoroscopic spot views were obtained at the beginning (F) and end (G) of the resection. The pointer in F is a metal instrument that the surgeon uses to localize the coalition fluoroscopically. The white rectangle in G outlines the resection site. H, Postoperative radiograph shows the calcaneus and navicular no longer in contact with each other (white rectangle).

CHAPTER 64  Foot and Ankle RIGHT

B

A

C

E

RIGHT

LEFT

D

RIGHT

F

LEFT

RIGHT

FIG 64-94  Tarsal coalition in a 29-year-old radiology resident with a long history of bilateral foot pain. The pain is worse on the right, is aggravated by athletics, and was improved with custom orthoses. A, Lateral radiograph of the more symptomatic right ankle reveals a talar beak (arrow). B, Lateral radiograph of the less symptomatic left ankle shows no enthesophyte arising from the dorsal head of the talus. Oblique radiographs of the right (C) and left (D) feet reveal no coalition between the calcaneus and navicular (bidirectional arrows). E, Sagittal CT scan of the right foot reveals the talar beak (arrow) as well as a portion of the solid osseous coalition across the subtalar joint (arrowhead). F, CT scan in the coronal oblique plane through the middle facets of both subtalar joints demonstrates solid osseous coalition across the right middle facet (white arrowheads). There is also a nonosseous coalition of the left middle facet (black arrowheads), as manifest by a joint that is abnormally broad with irregular, noncongruent articular surfaces.

LEFT

2313

2314

PART II  CT and MR Imaging of the Whole Body

N

N Ca

Ca

A

RIGHT

B

LEFT

C

RIGHT

D

LEFT

FIG 64-95  Calcaneonavicular coalition seen radiographically in this 21-year-old who has been complaining of left ankle pain for at least 7 years. On physical examination, the subtalar range of motion of the left foot is half that of the asymptomatic right. A, Oblique radiograph of the asymptomatic right midfoot shows the normal relationship between the calcaneus (Ca) and navicular (N), with no contact between them. B, Oblique radiograph of the symptomatic left foot shows the abnormal joint (arrowheads) between the calcaneus and navicular. This is a nonosseous coalition. Lateral radiographs of the right (C) and left (D) ankles reveal elongated anterior processes of the calcaneus bilaterally (arrows). These bilateral “ant-eater” signs suggest that the patient has an asymptomatic calcaneonavicular coalition on the right that was not radiographically apparent on the oblique view (A).

CHAPTER 64  Foot and Ankle

2315

ER

N APC

A

N

APC

C

B

FIG 64-96  Anterior process of the calcaneus (APC) fracture in a 51-year-old who was playing a jumping game. The patient, who was wearing sandals, landed awkwardly and felt a large “pop” after forcefully inverting the foot. A, Lateral radiograph obtained when the patient came into the emergency department the next day revealed the previously asymptomatic elongated APC (arrows). Careful scrutiny also reveals disruption of the cortex (arrowhead), indicating a nondisplaced APC fracture. B and C, CT scans were obtained the same day. Sagittal (B) and axial (C) scans revealed the APC fracture (black arrowheads) as well as the incidental finding of a calcaneonavicular nonosseous coalition (white arrowhead; N, navicular). The patient was treated nonoperatively with a non-weightbearing cast. When no healing was seen radiographically 9 weeks later, an ultrasonic bone stimulator was used. Five weeks after that, healing was evident radiographically. Twenty-three weeks later, the injury was essentially asymptomatic.

Talocalcaneal Coalition.  Talocalcaneal coalitions (see Fig. 64-94),

Tarsal Tunnel Syndrome

which occur across the middle facet of the subtalar joint, are difficult to demonstrate with conventional radiographs. An indirect radiographic sign is a talar beak, which represents an enthesophyte at the dorsal aspect of the talus related to altered biomechanics at the talonavicular joint (see Fig. 64-94A). Talar beaks occur less frequently with nonosseous coalition because some subtalar motion remains. Coronal CT images obliqued to be perpendicular to the subtalar joint are the key imaging plane (see Fig. 64-94F). With osseous coalitions, solid bony ankylosis is present across the middle facet (see Fig. 64-94F, white arrowheads). Nonosseous coalitions across the middle facet are not difficult to recognize by CT because they do not look like the flat uniform middle facet we typically see on coronal oblique CT. In nonosseous coalitions, the articular surfaces are not smooth or congruous and tend to have an overgrown appearance (see Fig. 64-94F, black arrowheads). An additional example of nonosseous middle facet coalition is shown in Figure 64-97.

The tarsal tunnel in the ankle is analogous to the carpal tunnel in the wrist. Both are spaces confined by the underlying bones and overlying fibrous ligaments through which pass tendons, blood vessels, and nerves. The roof of the tarsal tunnel is the flexor retinaculum, a broad fibrous band extending between the medial malleolus and the medial tubercle of the calcaneus (Fig. 64-98A). These retinacular fibers help prevent the medial tendons from becoming dislocated and can be identified on high-resolution images (see Fig. 64-98B and Fig. 64-31). Because the tarsal tunnel is a relatively tight space, otherwise innocuous volume-occupying lesions, such as synovial cysts (Fig. 64-99), small nerve sheath tumors, focal synovitis, or rarely even varicose veins, can potentially impinge on the posterior tibial nerve.24,36

Stress Injuries When the foot is subjected to new or excessive forces, such as a change in physical activity or an increased level of workout, certain bones may

2316

PART II  CT and MR Imaging of the Whole Body

B

A

FIG 64-97  Coronal (A) and sagittal (B) reconstructions of CT demonstrate nonosseous talocalcaneal coalition at the middle facet of the subtalar joint (arrows).

Tarsal tunnel

r xo um Fle acul tin re

A

B FIG 64-98  Location of the tarsal tunnel. A, Illustration of the location of the tarsal tunnel (arrow) deep to the flexor retinaculum. (Artist, M. Schenk, MS, CMI.) B, Axial high-resolution T1-weighted image shows the medial neurovascular bundle (dotted ellipse) deep to the flexor retinaculum (arrows).

CHAPTER 64  Foot and Ankle

A

C

2317

B

D

T2 FS

FIG 64-99  Tarsal tunnel containing a synovial cyst (arrow); axial (A) and sagittal (B) T2-weighted fatsuppressed images. Coronal (C) and axial (D) T2 fat-suppressed images in a different patient demonstrate a multilobulated synovial cyst within the distal aspect of the tarsal tunnel (arrows).

be subjected to a disproportionate amount of the increased stress and exhibit a stress response. The pattern of stress response depends on which bone is involved and how long it has been untreated.

Navicular Stress Fractures.41  Navicular stress fractures (Fig. 64-100A) tend to be the result of repetitive injuries rather than a specific traumatic event. Such fatigue injuries are commonly seen in college athletes. Often the athlete’s prognosis and the length of time needed to rest the fatigue injury depend on whether the cortex is broken. When MRI demonstrates just bone marrow edema without a breach in the cortex, these will be radiographically occult, and our sports medicine physicians prefer we use the term stress reaction. We

use the term stress fracture when a discrete line extending through the cortex is identified by MRI, CT, or plain radiography (Figs. 64-101 and 64-102). Although navicular fatigue fractures may be suspected clinically, initial radiographs are often negative, and MRI is the next imaging study ordered to confirm the diagnosis. As with most stress fractures, MRI is more sensitive than CT for detection of the bone marrow edema that develops before the cortex breaks (see Fig. 64-100). Whether seen by CT or MRI, navicular fatigue injuries begin at the dorsal, central, proximal navicular where it articulates with the head of the talus. This is illustrated by the black arrows pointing to the dark Text continued on p. 2322

T1

m

IR

m

A

m

m

B

C

T1

D

T2fs

E

T1

F

T2fs

FIG 64-100  Navicular stress reaction in a 36-year-old avid runner who had recently begun marathon training. The markers (m) indicate the proximal and distal extents of the patient’s pain. A, Sagittal T1-weighted MRI shows abnormally dark bone marrow in the dorsal half of the navicular (arrow). B, Sagittal inversion recovery (IR) image, being more sensitive for edema, shows abnormally bright signal throughout the navicular (arrows). C, Long-axis oblique axial T1-weighted image shows abnormally dark bone marrow in the central third of the navicular, emanating from the proximal articular surface adjacent to the head of the talus (arrow). D, Fatsuppressed T2-weighted image in the same plane, being more sensitive for edema, shows abnormally bright signal throughout the navicular (arrows). E, Short-axis oblique coronal T1-weighted image just distal to the talonavicular joint shows abnormally dark bone marrow in the dorsal central portion of the navicular (arrow). F, Fat-suppressed T2-weighted image at that same location, being more sensitive for edema, shows abnormally bright signal throughout the navicular (arrows). None of these MRIs demonstrates a discrete fracture line, and we call this a stress reaction rather than a stress fracture.

CHAPTER 64  Foot and Ankle

A

B

T1

C

T2fs

FIG 64-101  Navicular fatigue (stress) fracture in a 16-year-old who developed midfoot pain while crosscountry skiing. A, Anteroposterior radiograph of the foot reveals a subtle nondisplaced fracture in the middle third of the navicular (arrowhead in magnified dashed box). MRI was obtained 4 days later. Axial oblique T1- (B) and fat-suppressed T2-weighted (C) images reveal a discrete fracture in the middle third of the navicular (arrowhead) as well as diffuse bone marrow edema. Continued

2319

2320

PART II  CT and MR Imaging of the Whole Body

D

F

T1

E

T2fs

G G

FIG 64-101, cont’d  Coronal oblique T1- (D) and fat-suppressed T2-weighted (E) images show that this is an incomplete fracture, beginning from the dorsal cortex (arrowhead) and extending inferiorly in the sagittal plane, but not extending completely to the plantar cortex. A CT scan obtained 2 months later, reformatted using a 6-cm FOV in the oblique coronal (F) and oblique axial (G) planes, reveals that the fracture remains nonunited (arrowhead) and the bones are diffusely osteopenic from the patient’s being non–weight bearing.

CHAPTER 64  Foot and Ankle

A T1

B T2fs

C

D

E

F

FIG 64-102  Navicular stress fracture in a 20-year-old college decathlete complaining of lateral ankle pain not localized to the navicular. An MRI requested to evaluate the ankle joint found no abnormalities in or around the ankle but revealed abnormal bone marrow signal limited to the navicular. Coronal oblique T1(A) and fat-suppressed T2-weighted (B) images just distal to the talonavicular joint reveal the dark fracture line extending from the dorsal cortex (arrowhead) in a plantar-lateral direction. Serial CT scans using our navicular protocol were ordered to follow the progress of healing. Shown here is the same coronal oblique slice just distal to the talonavicular joint from scans taken over a period longer than 1 year. C, The first CT scan, obtained 2 months after the MRI, during which time the patient was non–weight bearing on this foot and using an ultrasonic bone stimulator. The fracture (arrowhead) is very narrow with indistinct noncorticated margins, suggesting that it is healing. D, The second CT scan, obtained 1 month after C, during which time the patient was weight bearing in a boot and using the bone stimulator. The fracture line (arrowhead) is much less distinct, consistent with continued interval healing. E, The third CT scan, obtained 3 months after D, during which time the patient resumed his training regimen. The fracture line (arrowhead) is now barely discernible. F, The final CT scan was obtained 9 months after E, when the patient’s symptoms returned. The fracture (arrowhead) has recurred along the original fracture lines, which are now wider and more distinct than on the first CT scan (C).

2321

2322

PART II  CT and MR Imaging of the Whole Body

T1

B

A

T2fs

C

IR

D FIG 64-103  Calcaneal stress fracture in a 62-year-old. A, Lateral radiograph shows a sclerotic band (black arrowheads) in the characteristic position of a calcaneal stress fracture, perpendicular to the trabeculae (white arrowheads). Incidentally seen is an os peroneum (arrow), a common normal variant. B, Midsagittal T1-weighted image shows the characteristic well-defined black line (arrowheads) of a calcaneal stress fracture. C, The corresponding midsagittal T2-weighted fat-suppressed (T2FS) image also shows the dark fracture line (arrowheads) as well as surrounding bone marrow edema. D, Sagittal inversion recovery (IR) image shown for comparison with the T2FS image (C). On the IR image, the normal fatty bone marrow is very dark, making the bone marrow edema more conspicuous. Arrowheads show the calcaneal stress fracture.

regions of bone marrow on the T1-weighted images in the stress reaction in Figure 64-100. More fluid-sensitive fat-suppressed T2-weighted or inversion recovery images show bone marrow edema emanating from this dorsal/central/proximal site. This is illustrated by the white arrows in Figure 64-100. When stress reactions progress to stress fractures, the cortical disruption starts at the dorsal/central/proximal site on the navicular and propagates in a plantar direction vertically in the sagittal plane (see Fig. 64-101) or in an oblique sagittal plane (see Fig. 64-102). Because of the primarily sagittal orientation of these fractures, they may be difficult to appreciate on sagittal CT images and are better seen on oblique coronal (see Fig. 64-101F) and oblique axial (see  Fig. 64-101G) images. Because they tend to be nondisplaced incomplete fractures, they are best seen on images that are reformatted into a small FOV with thin slices. Because these patients may undergo serial CT scans to follow the progress of fracture healing, it is useful to have a standard protocol (as in Fig. 64-63D) to help retain uniform reformatting parameters from one scan to the next (see Fig. 64-102C to F).

Calcaneal Stress Fractures.  Calcaneal stress fractures occur in a characteristic location, arising from the posterior third of the calcaneal tuberosity beginning at the superior cortex a few centimeters anterior

to the Achilles insertion, and extending inferiorly and slightly anteriorly, running perpendicular to the trabeculae. When radiographically apparent, these fractures are seen as a white sclerotic line on the lateral view (Fig. 64-103A). On MRI, calcaneal stress fractures are seen as a black line on sagittal T1-weighted images (see Fig. 64-103B) surrounded by bone marrow edema on fat-suppressed T2-weighted (see Fig. 64-103C) and inversion recovery (see Fig. 64-103D) images. Figure 64-104 is an example of a calcaneal stress fracture that was subtle on initial radiographs and was ultimately imaged using CT, a nuclear medicine bone scan, and MRI.

Plantar Fasciitis.  Plantar fasciitis is a common painful process of the plantar aponeurosis resulting from trauma and inflammation at its origin from the calcaneus, usually at the medial calcaneal tubercle. Causes include repetitive microtrauma in the origin of the plantar fascia causing traction periostitis and microtears, resulting in pain and inflammation. Plantar calcaneal spurs have been associated with plantar fasciitis. Plantar fasciitis is the most common cause of pain in the inferior aspect of the heel, and the diagnosis is typically made based on clinical symptoms and physical examination revealing tenderness along the medial calcaneal tuberosity. Most patients with

CHAPTER 64  Foot and Ankle

A

B

R

L

C FIG 64-104  Calcaneal stress fracture in a 14-year-old cross-country runner. A, Lateral radiograph shows a subtle sclerotic band (arrowheads in the magnified dashed box). CT scans in the axial (B) and sagittal (C) planes show the sclerotic line in the right calcaneus (arrowheads). The normal left calcaneus is included for comparison. Continued

2323

2324

PART II  CT and MR Imaging of the Whole Body

R

D

L

E

T1

F

T2fs

G

FIG 64-104, cont’d Bone scan images, both-feet-on-detector (D) and lateral (E) views, show increased activity in the right calcaneal tuberosity. Incidentally noted is normal activity in the distal tibial physis (arrowheads) in this skeletally immature patient. Midsagittal T1-weighted (F) and T2-weighted fat-suppressed (G) images show the characteristic well-defined black line (arrowheads) of a calcaneal stress fracture, as well as edema in the surrounding bone marrow.

plantar fasciitis respond to conservative treatments that include calf stretching, orthoses, nonsteroidal antiinflammatory medication, ultrasonic therapy, and occasionally casting. Patients with atypical clinical presentations or who fail conservative therapies may benefit from MRI to determine if their pain is indeed related to the plantar fascia or to some other etiology such as a tarsal stress fracture. An MRI of a patient with plantar fasciitis reveals calcaneal marrow edema at the origin of the aponeurosis, thickening of the aponeurosis and surrounding soft tissue edema with effacement of the subcutaneous fat. The plantar fascia may be partially torn or detached in severe cases (Figs. 64-105 and 64-106).

Metatarsal Stress Fractures.  Metatarsal stress fractures occur at such characteristic locations that some carry eponyms. The Jones fracture occurs at the proximal metadiaphysis of the fifth metatarsal and is seen radiographically as a transverse lucency (see Fig. 64-42C). Although Jones fractures can be caused by a single 

traumatic injury, they are commonly seen as the result of repetitive stress in athletes, or as in the case of Sir Robert Jones, “whilst dancing.”27 The Jones fracture is well recognized to have a high rate of nonunion or delayed union because of the relative hypovascularity of this portion of the fifth metatarsal, prompting orthopedic surgeons to recommend early screw fixation. MRI is useful in confirming the diagnosis when it is suspected in athletes with radiographically occult injuries (Fig. 64-107). The march fracture is found most commonly in the mid- to distal diaphysis of the second metatarsal and less often in the third. Unlike the Jones fracture, which occurs in high-performance athletes, stress fractures in the second and third metatarsals occur in individuals who have previously led relatively sedentary lifestyles then suddenly increase their level of activity. This fracture was first reported by Breithaupt in 1855,11 when he described foot pain and swelling in military recruits in the Prussian army who were forced to go on long marches—hence the name “march fracture.” Radiographically the second and third

CHAPTER 64  Foot and Ankle

LEFT

A

T1

med

B

IR

C

lat

T2fs

RIGHT

D

T1

lat

med

E

IR

F

T2fs

FIG 64-105  Plantar fasciitis in a 52-year-old with chronic bilateral heel pain. The left (A to C) and right (D to F) hindfeet were scanned individually. A, Sagittal T1-weighted image reveals thickening of the plantar fascia (white arrows). B, The corresponding sagittal inversion recovery (IR) image reveals edema (white arrowhead) deep to the plantar fascia (white arrow). Open arrow points to os trigonum. C, Coronal T2-weighted fatsuppressed image demonstrates a line of fluid (white arrowhead) as well as some focal bone marrow edema (black arrowhead) deep to the origin of the plantar fascia (white arrow). (“med” and “lat” represent the medial and lateral sides of the image, respectively.) Incidentally seen is a normal os trigonum (open arrow in A to C) with bone marrow signal isointense to the normal bone marrow in the other bones. D, Sagittal T1-weighted image of the contralateral right foot also demonstrates a thickened plantar fascia (arrows). The calcaneal bone marrow edema is less conspicuous than on more fluid-sensitive sequences. E, The corresponding IR images reveal extensive bone marrow edema along the plantar portion of the calcaneus (arrowheads). F, Coronal T2-weighted fat-suppressed image shows the bone marrow edema (white arrowhead) radiating from the medial-plantar surface of the calcaneus, at the origin of the plantar fascia.

2325

2326

PART II  CT and MR Imaging of the Whole Body

P

T2 FS

A

B

P

T1

C

FIG 64-106  Plantar fasciitis in a 68-year-old woman. A, Sagittal T2 fat-suppressed and T1 images demonstrate soft tissue edema surrounding a thickened plantar aponeurosis at its origin from the calcaneus (arrowhead). There is mild marrow edema in the calcaneus as well. B, Coronal T2 fat-suppressed and C, coronal T1 images again demonstrate the thickened medial bundle of the aponeurosis, with intrasubstance signal and surrounding soft tissue edema (arrowheads).

metatarsals respond to stress by forming a periosteal reaction, although this may be imperceptible or subtle early on (Fig. 64-108). Edemasensitive MRI reveals abnormally bright marrow signal in the diaphysis as well as bright periosteal reaction outside the cortex. A fracture line may eventually be seen, usually incomplete.

Sesamoid Stress Fractures.  Sesamoid stress fractures are notoriously difficult to diagnosis radiographically. However, when looking at the sesamoids of the first metatarsophalangeal joint, there are some helpful statistics to keep in mind. Although the other sesamoid bones

of the foot are present in less than 10% of people, the two sesamoids plantar to the head of the first metatarsal are present in 100% of the population.20 Absence of either the medial (tibial) or lateral (fibular) sesamoids is always abnormal, and a destructive process should be considered (see Fig. 64-91). And although a multipartite medial sesamoid bone is a common normal variant found in 13% to 30% of the population, a multipartite lateral sesamoid bone is an uncommon variant, found in only 1% of normal feet. When symptoms are referred to the lateral sesamoid and radiographs reveal it to be multipartite, this should be diagnosed as a

CHAPTER 64  Foot and Ankle

2327

IR 5th

B

4th

A

C

D

FIG 64-107  Early Jones fracture in a 21-year-old college athlete. A, On the initial anteroposterior radiograph of the base of the fifth metatarsal there is a very subtle periosteal reaction (white arrow). There is a questionable lucency (black arrowheads) extending transversely through the lateral cortex. B, Far lateral sagittal inversion recovery (IR) image reveals bone marrow edema throughout the fifth metatarsal. C, Sagittal IR image through the asymptomatic fourth metatarsal for comparison revealed normal dark marrow signal. Because of the high propensity for Jones fractures to have delayed union or nonunion, elective internal fixation with a lag screw was performed the next day. D, Follow-up anteroposterior radiograph 2 weeks later reveals bone resorption (arrowheads) along the questionable lucency in A, part of the normal early healing response. Subsequent radiographs (not shown) demonstrated solid bony bridging 1 month later.

fracture (Fig. 64-109). Additional diagnostic imaging should not be required, although MRI or a nuclear medicine bone scan could be obtained if confirmation is necessary. Fractures of the medial sesamoid are more difficult to diagnose radiographically because this sesamoid is not infrequently multipartite as a normal variant. Here radiographs are of limited value, and more sensitive imaging modalities are often required. Although MRI can demonstrate abnormal marrow signal in the sesamoids, owing  to the small size of these bones this may be present on only a single slice, and all imaging planes should be carefully scrutinized. Shortaxis images are particularly helpful in comparing the marrow signal from the medial and lateral sesamoids side-by-side (Fig. 64-110).  The imaging of sesamoiditis is one of the few instances when we recommend a nuclear medicine bone scan over an MRI. In particular, the both-feet-on-the-detector view is extremely effective for localizing abnormal radiotracer uptake to one of the sesamoids (see  Fig. 64-110C).

INFECTION Osteomyelitis is always a diagnostic dilemma. The term osteomyelitis comes from the Greek roots osteon meaning “bone,” myelos meaning “marrow,” and itis meaning “inflammation.” Thus, osteomyelitis literally means “inflammation of bone marrow,” and this is perhaps symbolic of the dilemma. MRI is extremely sensitive for the detection of marrow inflammation, but it is not specific for the inflammation

caused by infection. By MRI, the bone marrow edema caused by infection looks just like the bone marrow edema caused by a stress response as well as the edema caused by a nonhealing fracture or even a healing fracture. For this reason, an MRI for osteomyelitis should not be read in isolation. It should always be correlated with a clinical workup and history including prior surgical debridement.

Imaging Techniques

Radiography.  Radiographs are essential in the workup of osteomyelitis, and recent radiographs should be available before an MRI for infection is performed. First, radiographs are crucial to screen for the presence of metal, particularly in the feet of diabetic patients who may be insensate and thus unknowingly stepped on pins, not to mention the presence of orthopedic hardware or other postsurgical changes. Second, in diabetic feet it is necessary to screen for the jointcentered collapse that is typically seen with peripheral neuropathy, the Charcot joint. These radiographic findings have been described as “the six Ds”: destruction, increased density, dislocation, debris, distension, and disorganization. The bone marrow and soft tissue edema seen with MRI in patients with sterile neuropathic osseous changes may be indistinguishable from infection. The use of contrast and the presence of prior images may be helpful in these cases. Because neuropathic collapse can occur relatively quickly and go unnoticed by a patient with an insensate foot (Fig. 64-111), we require that the pre-MRI radiographs be recent, preferably within the last week. Text continued on p. 2333

2328

PART II  CT and MR Imaging of the Whole Body

A T1

T2fs

C

B T1

D

T2fs

E

FIG 64-108  Second metatarsal stress fracture in a 22-year-old who developed foot pain during a 1-week vacation in which the patient did a lot of walking in sandals. A, Oblique radiograph reveals a periosteal reaction along the medial cortex (white bracket in magnified dashed box). B, Long-axis T1-weighted image through the second metatarsal well illustrates the anatomy but not the pathology. C, Corresponding long-axis T2-weighted fat-suppressed image reveals bone marrow edema throughout the second metatarsal diaphysis, the thickened medial cortex/periosteum, and edema of the adjacent medial soft tissues.

CHAPTER 64  Foot and Ankle

T1

m

m

F

G

T2 FS

T1

H

J

IR

I

T2 FS

FIG 64-108, cont’d Short-axis T1-weighted (D) and T2-weighted fat-suppressed (E) images through the metatarsal shafts reveal edema in the second metatarsal bone marrow and in the adjacent medial soft tissues overlying the periosteal reaction. Sagittal T1-weighted (F) and inversion recovery (G) images through the marker (m) indicating the site of maximal tenderness show edema in and around the second metatarsal. H, Axial T2 fat-suppressed and I, axial T1 images in a different patient demonstrate fracture line and marrow edema at the distal fourth metatarsal consistent with a stress fracture.

2329

2330

PART II  CT and MR Imaging of the Whole Body

B

A

m

C

D

E

T1

IR

FIG 64-109  Fracture of the lateral sesamoid in a 34-year-old who complained of localized pain plantar and lateral to the first metatarsal head, made worse with weight bearing and extension of the great toe, for 1.5 years before the diagnosis was made. Anteroposterior (A), oblique (B), and sesamoid (C) radiographs all clearly show the transverse split across the lateral sesamoid of the great toe. A bipartite lateral sesamoid is an uncommon variant, present in only 1% of the population, and when symptomatic should be interpreted as a fracture. Short-axis T1-weighted (D) and inversion recovery (E) images through the marker (m) indicating the site of maximum pain show normal bone marrow signal in the medial sesamoid (white arrow) and bone marrow edema in the lateral sesamoid (black arrow).

CHAPTER 64  Foot and Ankle

A T1

B

T2fs

C

R

L

FIG 64-110  Fracture of the medial sesamoid in a 27-year-old with a several-month history of pain localized to the head of the first metatarsal. Short-axis T1-weighted (A) and T2-weighted fat-suppressed (B) images show normal bone marrow signal in the lateral sesamoid (white arrow) and bone marrow edema in the medial sesamoid (black arrow). C, Bone scan, both-feet-on-detector view, localizes the increased uptake to the medial sesamoid of the left foot. This patient failed to respond to conservative therapy and ultimately had the medial sesamoid resected.

2331

2332

PART II  CT and MR Imaging of the Whole Body

A

C

B

D

FIG 64-111  Rapid onset of severe Charcot neuropathic collapse in a diabetic patient. Anteroposterior (A) and oblique (B) radiographs show anatomic alignment along the Lisfranc joint. Anteroposterior (C) and oblique (D) radiographs only 3 months later reveal complete destruction of the Lisfranc joint.

CHAPTER 64  Foot and Ankle

T1 FS POST

A

C

2333

B

D

T2 FS

T1

FIG 64-112  Acute osteomyelitis in the great toe. A, AP radiograph of the toes reveals osseous destruction of the first distal phalanx with erosions in the proximal phalanx (arrows). B, Axial T2 fat-suppressed and C, sagittal T2 fat-suppressed and D, sagittal T1 images through the foot demonstrate marked marrow edema in the proximal phalanx of the great toe with osseous destruction of the distal phalanx (arrow). There is loss of the normal marrow signal on the T1 images (arrow).

Third, radiographs may reveal findings that, in the proper clinical setting, are diagnostic for osteomyelitis. New cortical erosions (Fig. 64-112) in a bone deep to a nonhealing ulcer or unresponsive cellulitis are as diagnostic as MRI for active osteomyelitis. Periosteal reactions can be useful as well. Gas in the soft tissues, such as from a gas-forming organism, is easily detected on radiographs yet may be hard to interpret on MRI because it can cause susceptibility artifacts similar to those caused by metal.

Magnetic Resonance Imaging.  Ultimately it is easier to rule out osteomyelitis by MRI than it is to confirm its presence. The absence of increased bone marrow signal on a fluid-weighted sequence effectively

rules out osteomyelitis. However, the presence of marrow edema on fluid-weighted sequence, although sensitive, is not specific for the diagnosis of osseous infection. Loss of T1 signal in the bone marrow is a more specific finding of osteomyelitis. Other findings that can aid in the diagnosis are the presence of a soft tissue abscess or a sinus tract between the infected bone and the skin. IVGd is useful for evaluation of these supporting signs. The presence of a soft tissue ulcer adjacent to the osseous changes is a significant supporting factor for the diagnosis of osteomyelitis as well. The diagnosis of osteomyelitis can be presumed when MRI shows not only marrow edema but also abscess in the adjacent soft tissues (Fig. 64-113) or a sinus tract communicating from the infected bone

A

C

IR

E

T1

D

T1fs IVGd

F

G

I

B

IR

T1

J

T1

T1fs IVGd

H

IR

K

T1fs IVGd

FIG 64-113  Developing calcaneal osteomyelitis in a 63-year-old diabetic patient. A, Lateral radiograph of the calcaneus shows intact cortex along the plantar surface (white arrowheads). Incidentally seen is mural calcification of the posterior tibial artery (black arrowheads). Such arterial calcifications are common in diabetic patients. B, Midsagittal T1-weighted image shows no bone destruction. C, Corresponding sagittal inversion recovery (IR) image shows little if any bone marrow edema. D, Corresponding sagittal post-IVGd contrast T1-weighted fat-suppressed image reveals diffuse enhancement of the plantar soft tissues, indicative of cellulitis, but no nonenhancing abscess pockets. When the patient’s symptoms did not respond to antibiotics, repeat imaging was obtained 2 weeks later. E, Lateral radiograph now demonstrates loss of cortex along the plantar surface of the calcaneus (arrowheads). F, Midsagittal T1-weighted image reveals infiltration of the fatty heel pad (arrows). G, Corresponding sagittal inversion recovery image reveals fluid bright signal (arrows) in the soft tissues adjacent to the calcaneus, as well as bone marrow edema in calcaneus (arrowheads). H, Corresponding sagittal post-IVGd contrast T1-weighted fat-suppressed image reveals a nonenhancing abscess pocket (arrows) as well as enhancing bone marrow (arrowheads). Coronal T1-weighted (I), inversion recovery (J), and post-IVGd contrast T1-weighted fat-suppressed (K) images through the abscess pocket confirm the findings seen in the sagittal plane: an abscess pocket (arrows) adjacent to the osteomyelitis (arrowhead) of the planter surface of the calcaneus.

CHAPTER 64  Foot and Ankle to the skin.1 IVGd-based contrast is extremely helpful in diagnosing the abscess, which exhibits thick irregular enhancement peripherally but not centrally (see Fig. 64-113H and K).

Brodie’s Abscess Brodie’s* abscess is a chronic intraosseous abscess resulting from incomplete resolution of acute osteomyelitis and isolation of the infection by surrounding bone. These abscess pockets are typically found in the metaphyses of skeletally immature children, and the usual pathogen is Staphylococcus aureus. However, the organisms tend to be of low virulence, and it is not uncommon for cultures of such aspirates to yield no growth. Clinical symptoms are often mild, generally presenting with persistent local pain. Radiographs show an intramedullary lucency with surrounding sclerosis, the density of which depends on the chronicity of the  abscess. A thick chronic periosteal reaction may also be present (Fig. 64-114B). MRI after the administration of IV contrast reveals an intraosseous abscess with peripheral but not central enhancement (see Fig. 64-114E).33

TUMORS Soft Tissue Masses Soft tissue tumors of the feet and ankle are common, and MRI is useful in determining the tissue type as well as demonstrating the relationship of the mass to the adjacent anatomic structures. Synovial cysts or ganglia are among the most common soft tissue “masses” found around the foot and ankle. These are uniformly bright on  fluid-sensitive images and exhibit minimal if any peripheral enhancement after administration of IVGd-based contrast (Fig. 64-115). Eventually the synovial lining of the ganglion cyst will excrete contrast, and therefore delayed postcontrast imaging may demonstrate “enhancement.” In comparison, nerve sheath tumors such as schwannomas  are heterogeneously bright on T2-weighted and inversion recovery images, and they demonstrate heterogeneous contrast enhancement (see Fig. 64-55). Plantar fibromas can have variable signal characteristics but are typically dark on all sequences (Fig. 64-116). These usually arise from the plantar aponeurosis, usually close to the calcaneus. Morton’s neuromas usually occur between the heads of the second and third or third and fourth metatarsals and are also usually dark on noncontrast images but demonstrate postcontrast enhancement (see Fig. 64-56). Giant cell tumor of the tendon sheath is a localized form of pigmented villonodular synovitis, the latter being a joint-centered synovial proliferative condition. Both diseases show areas of decreased signal on T1-, PD-, and T2-weighted images, secondary to hemosiderin

*Sir Benjamin Collins Brodie (1783-1862) was an English physiologist and surgeon who pioneered research into bone and joint disease. His most important work is widely acknowledged to be the 1818 treatise Pathological and Surgical Observations on the Diseases of the Joints, in which he attempts to trace the beginnings of disease in the different tissues that form a joint and to give an exact value to the symptom of pain as evidence of organic disease. This volume led to the adoption by surgeons of more conservative measures in the treatment of diseases of the joints, with consequent reduction in the number of amputations and the saving of many limbs and lives.

2335

deposition (Fig. 64-117A to C). The presence of hemosiderin can be detected on gradient echo and precontrast fat-suppressed T1-weighted images (see Fig. 64-117D), and the vascularized proliferative synovium should exhibit some contrast enhancement (see Fig. 64-117E).

Bone Tumors28 Osseous tumors are much less common than soft tissue tumors of the foot and ankle. Like all bone lesions, these tumors should be initially evaluated radiographically. MRI, however, is useful in localizing tumors and staging their extent. Because most tumors have nonspecific signal characteristics, MRI is often unable to render a specific preoperative diagnosis. Primary bone tumors of the feet are rare, accounting for only 4% of all bone tumors.22 Benign bone neoplasms are more common than malignant ones, although in the foot, most bone neoplasms are primary tumors because metastases to the foot are rare.

Benign Tumors.  The most common benign tumors of the foot are enchondromas and osteoid osteomas. An osteoid osteoma is a relatively common cause of bone pain in adolescents and young adults, accounting for approximately 10% of all benign bone tumors. The classic history is pain at night, relieved by aspirin. Osteoid osteomas are one of the few tumors that are better imaged by CT than MRI. Thin-slice CT well demonstrates the lucent nidus as well as the tiny sclerotic component that is characteristically associated with it (Fig. 64-118A). CT is also used by radiologists for the purpose of percutaneously ablating the nidus. On MRI, osteoid osteomas are seen as a nonspecific edema pattern emanating from a tiny dark nidus (see Fig. 64-118B to D). Chondroblastomas are rare benign cartilaginous neoplasms and one of the few tumors that arise from the epiphysis in a skeletally immature patient. Chondroblastomas can expand the cortex but should not cross the unfused growth plate (Fig. 64-119A). Radiographically, chondroblastomas can have either a lucent or chondroid matrix. By MRI, these lesions may exhibit considerable edema in the surrounding soft tissues, but the tumor itself should have a sharp, nonaggressive-appearing interface with the normal bone (see Fig. 64-119B to D). Intraosseous lipomas of the calcaneus are rare but have a characteristic radiographic appearance in that they are well circumscribed and nearly totally lucent except for a tiny central sclerotic focus (Fig. 64-120A). By MRI, the intraosseous lipoma is uniformly isointense to fat on all sequences except for a signal void corresponding to the sclerotic focus (see Fig. 64-120C and D).

Malignant Tumors.  The most common primary malignant tumor of the foot is chondrosarcoma, which has a propensity for the calcaneus (Fig. 64-121). High-grade chondrosarcomas have a calcified matrix that appears radiographically sclerotic (see Fig. 64-121A) and dark on T1- and T2-weighted sequences. Chondrosarcomas are not highly vascularized tumors and hence demonstrate poor contrast enhancement as well as a poor response to chemotherapy. Contrast enhancement can be seen at the margins of the tumor where it is aggressively invading the bone and in the surrounding soft tissues (see Fig. 64-121D). Low-grade chondrosarcomas are difficult to distinguish from benign enchondromas for both radiologists and pathologists, and the clinical symptom of pain is often the deciding factor. Although many enchondromas are asymptomatic and may be discovered incidentally when imaging the foot for other reasons, chondrosarcomas are painful. Text continued on p. 2343

A

B

C

T1

D

F

T1

G

T2fs

T2fs

E

T1fs IVGd

H

T1fs IVGd

FIG 64-114  Brodie’s abscess in a young child. A, Anteroposterior radiograph of the asymptomatic right leg. B, Anteroposterior radiograph of the swollen left leg reveals a lucency in the distal fibula metaphysis (arrow in magnified dashed box). This lucency has a well-defined and sclerotic margin, indicating chronicity. There are also thick, chronic periosteal reactions (arrowheads) extending up the diaphysis. C, Coronal T1-weighted image through the distal fibula confirms the radiographic findings of a thick chronic periosteal reaction (white arrowheads), as well as the wellcircumscribed dark line (open arrowheads) around the lesion corresponding to the sclerotic margin. D, The corresponding coronal T2-weighted fat-suppressed image shows that the well-circumscribed lesion (arrow) is as bright as fluid and thus probably cystic. E, The corresponding coronal T1-weighted fat-suppressed post-IVGd contrast image not only confirms that the lesion (arrow) is mostly nonenhancing and thus mostly cystic, but demonstrates peripheral enhancement, in some places thick (black arrowhead), characteristic of an abscess, in this case an intraosseous or Brodie’s abscess. (There is inadequate fat suppression of the heel pad [large white arrowhead] on both of the fat-suppressed sequences, D and E.) F to H, Axial images through the fibular abscess reveal it to be isointense to muscle on T1-weighted image (F, arrow) and fluid bright on T2-weighted fat-suppressed image (arrow in G), with peripheral but not central enhancement on fat-suppressed T1-weighted image after IVGd (arrow in H). There are edema and enhancement of the soft tissues surrounding the fibula, indicating an active inflammatory component to this chronic Brodie’s abscess.

CHAPTER 64  Foot and Ankle

A

B FIG 64-115  A, Coronal T2 fat-suppressed image through the forefoot demonstrates a low-signal irregular mass surrounding the second metatarsal (arrow), with solid enhancement after contrast administration (arrow) on B, coronal T1 fat-suppressed image. This lesion was biopsied and proven to be a giant cell tumor of the tendon sheath. This is the classic appearance of this lesion, which is histologically similar to pigmented villonodular synovitis

T1 FS POST

B

A

C

T2 FS

D

T1

T1 FS POST

FIG 64-116  Plantar fibroma in a 42-year-old woman. Coronal (short-axis) T2 fat-suppressed (A), T1 (B), and sagittal T2 fat-suppressed (C) images demonstrate a low- to isointense-signal lesion arising from the plantar aponeurosis (arrow). D, Sagittal T1 fat-suppressed image demonstrates enhancement of this lesion after contrast administration.

2337

2338

PART II  CT and MR Imaging of the Whole Body

N Ta

T D

ST

H

A T1

B

T1fs D (no Gd)

PD

C

T2

T1fs E (IV Gd)

FIG 64-117  Giant cell tumor of the tendon sheath in a 19-year-old with a palpable medial mass. A, Straight axial T1-weighted image through the mass (arrow) demonstrates that it lies within the soft tissues medial to the navicular (N), talus (Ta), and sustentaculum tali (ST) but does not invade the bones. B, The corresponding axial PD-weighted image shows that the mass has grown through a split tendon sheath (arrowheads). The tendons themselves—posterior tibial (T), flexor digitorum longus (D), and flexor hallucis longus (H)—are intact and normal in appearance with no evidence of tumor involvement. C, The corresponding axial T2-weighted image shows that the mass (arrow) has heterogeneous signal intensity, consistent with the presence of blood products of varying ages. D, Corresponding axial precontrast fat-suppressed T1-weighted image reveals that some signal in the mass is brighter than the surrounding suppressed fat, consistent with methemoglobin. E, Corresponding axial post-IVGd fat-suppressed T1-weighted image demonstrates heterogeneous enhancement, indicative of the vascularity of this synovial proliferation.

CHAPTER 64  Foot and Ankle

A

B

C

T1

T2fs

IR

D

E

F

G

FIG 64-118  Osteoid osteoma in a 19-year-old with symptoms clinically thought to be due to a tibial stress fracture. A, Axial CT scan through the level of the syndesmosis. The nidus is the small lucent lesion (black circle). B, Axial T1-weighted image through same level as A. The low-intensity nidus (white circle) is masked by the surrounding marrow edema. C, Corresponding T2-weighted fat-suppressed image. The low-intensity nidus (white circle) is unmasked by the bright signal of the surrounding marrow edema. D, Sagittal inversion recovery (IR) image reveals bone marrow edema around the small nidus (white circle). Osteoid osteoma in a 16-year-old girl. Coronal (E) and sagittal (F) reconstructed images of a CT of the foot demonstrate a lucent cortical based lesion in the plantar aspect of the second metatarsal with a central nidus (arrow). G, Radiofrequency ablation of this lesion is shown with a probe located within the lesion.

2339

2340

PART II  CT and MR Imaging of the Whole Body

T1

A

B

Cartilagesensitive sequence

T2

C

D

FIG 64-119  Chondroblastoma of the calcaneus. A, Lateral radiograph of the ankle demonstrates a lytic lesion in the body of the calcaneus. B, Axial T2 fat-suppressed and C, axial T1 images demonstrate a heterogeneous high T2 and low T1 signal corresponding to chondral elements with surrounding marrow edema, which is a common finding in chondroblastomas. No cortical break identified. There is enhancement after contrast administration on D, sagittal T1 fat-suppressed image.

CHAPTER 64  Foot and Ankle

RIGHT

B

A

C

LEFT

T1

D

T2fs

FIG 64-120  Intraosseous lipoma discovered as an incidental finding in a 43-year-old. A, Lateral radiograph of the right calcaneus shows the lucent lesion in the anterior half with a well-circumscribed border (black arrowheads). Centrally there is a small sclerotic focus (white arrowhead) characteristic of an intraosseous lipoma. B, Lateral radiograph of the contralateral left calcaneus is shown for comparison. A lipoma, intraosseous or otherwise, should be isointense to fat on all imaging sequences. C, Midsagittal T1-weighted image shows the signal intensity of the lipoma (black arrowheads) to be isointense to that of the surrounding fatty bone marrow. The central sclerotic focus (white arrowhead) is dark on all imaging sequences. D, On the corresponding T2-weighted fat-suppressed image, the fat in the lipoma suppresses to a degree similar to that of the surrounding fatty bone marrow so that it is nearly inconspicuous.

2341

2342

PART II  CT and MR Imaging of the Whole Body

T2 FS A

C

B

T1

T1 FS POST D

FIG 64-121  Chondrosarcoma arising from the calcaneal tuberosity 30-year-old. A, Lateral radiograph. This patient complained of heel pain for 10 months until the sclerosis in her calcaneus was recognized. B, Sagittal T1-weighted image shows diffusely decreased signal in the calcaneus corresponding to the radiographic sclerosis. C, Sagittal fat-suppressed T2-weighted image reveals marrow edema only at the periphery of the sclerotic region. There is, however, extensive bright signal in the soft tissues immediately plantar to the calcaneus, indicating that the tumor is extending out of the bone and into the soft tissues. D, Sagittal postIVGd contrast fat-suppressed T1-weighted image demonstrates enhancement only at the periphery of the osseous and soft tissue masses. This enhancement pattern is due to the relative hypovascularity of chondrosarcomas.

CHAPTER 64  Foot and Ankle

REFERENCES 1. Ahmadi ME, Morrison WB, Carrino JA, et al: Neuropathic arthropathy of the foot with and without superimposed osteomyelitis: MR imaging characteristics. Radiology 238:622–631, 2006. 2. Anderson IF, Crichton KJ, Grattan-Smith T, et al: Osteochondral fractures of the dome of the talus. J Bone Joint Surg Am 71:1143–1152, 1989. 3. Archdeacon M, Wilber R: Fractures of the talar neck. Orthop Clin North Am 33:247–262, 2002. 4. Badgley CE: Coalition of the calcaneus and the navicular. Arch Surg 15:75–88, 1927. 5. Ballinger PW, Frank ED: Merrill’s Atlas of Radiographic Positions and Radiologic Procedures, ed 9, St. Louis, 1999, Mosby. 6. Barei DP, Bellabarba C, Sangeorzan BJ, et al: Fractures of the calcaneus. Orthop Clin North Am 33:263–285, x, 2002. 7. Bencardino JT, Rosenberg ZS: Normal variants and pitfalls in MR imaging of the ankle and foot. Magn Reson Imaging Clin N Am 9:447–463, 2001. 8. Berndt AL, Harty M: Transchondral fractures (osteochondritis dissecans) of the talus. J Bone Joint Surg Am 41:988–1020, 1959. 9. Böhler L: Diagnosis, pathology, and treatment of fractures of the os calcis. J Bone Joint Surg 13:75–89, 1931. Available at: . 10. Borrelli J, Jr, Ellis E: Pilon fractures: Assessment and treatment. Orthop Clin North Am 33:231–245, x, 2002. 11. Breithaupt X: Zur pathologie des menschlichen fuses [The pathology of the human foot]. Medizinische Zeitung (Berlin) 24:169–175, 1855. 12. Bucholz R, Heckman J, Court-Brown C, et al, editors: Rockwood and Green’s Fractures in Adults, ed 6, Philadelphia, 2006, Lippincott Williams & Wilkins. 13. Cantrell MW, Tarquinio TA: Fractures of the lateral process of the talus. Orthopedics 23:55–58, 2000. 14. Cheung Y, Rosenberg ZS: MR imaging of ligamentous abnormalities of the ankle and foot. Magn Reson Imaging Clin N Am 9:507–531, 2001. 15. Chhabra A1, Soldatos T, Chalian M, et al: 3-Tesla magnetic resonance imaging evaluation of posterior tibial tendon dysfunction with relevance to clinical staging. J Foot Ankle Surg 50(3):320–328, 2011. doi: 10.1053/j. jfas.2011.02.004. [Epub 2011 Apr 2]. 16. Choplin RH, Buckwalter KA, Rydberg J, et al: CT with 3D rendering of the tendons of the foot and ankle: Technique, normal anatomy, and disease. Radiographics 24:343–356, 2004. 17. Reference deleted in proofs. 18. Crim JR, Kjeldsberg KM: Radiographic diagnosis of tarsal coalition. AJR Am J Roentgenol 182:323–328, 2004. 19. Daftary A, Haims AH, Baumgaertner MR: Fractures of the calcaneus: A review with emphasis on CT. Radiographics 25:1215–1226, 2005. 20. De Smet AA, Fisher DR, Burnstein MI, et al: Value of MR imaging in staging osteochondral lesions of the talus (osteochondritis dissecans): Results in 14 patients. AJR Am J Roentgenol 154:555–558, 1990. 21. Donovan A, Rosenberg ZS: MRI of ankle and lateral hindfoot impingement syndromes. AJR Am J Roentgenol 195(3):595–604, 2010. doi: 10.2214/AJR.09.4199. Review. 22. Enneking WF, Spanier SS, Goodman MA: A system for the surgical staging of musculoskeletal sarcoma. Clin Orthop Relat Res 153:106–120, 1980. 23. Farooki S, Sokoloff RM, Theodorou DJ, et al: MR imaging of the midfoot: Normal anatomy with cadaveric correlation. Magn Reson Imaging Clin N Am 9:419–434, 2001. 24. Finkel JE: Tarsal tunnel syndrome. Magn Reson Imaging Clin N Am 2:67–78, 1994. 25. Finlay K, Friedman L: Ultrasonography of the lower extremity. Orthop Clin North Am 37:245–275, 2006. 26. Henry AK: Extensile Exposure, ed 2, Edinburgh, 1957, E&S Livingstone, p 304. 27. Jones R: Fracture of the base of the fifth metatarsal bone by indirect violence. Ann Surg 35:697–700, 1902.

2343

28. Kager H: Zur Klinik und Diagnostik des Achillessehnenrisses. Chirurg 11:691–695, 1939. 29. Keats TE, Anderson MW: Atlas of Normal Roentgen Variants that May Simulate Disease, ed 8, St. Louis, 2006, Mosby. 30. Lee SJ1, Jacobson JA, Kim SM, et al: Ultrasound and MRI of the peroneal tendons and associated pathology. Skeletal Radiol 42(9):1191– 1200, 2013. doi: 10.1007/s00256-013-1631-6. [Epub 2013 May 18]. 31. Reference deleted in proofs. 32. Melão L, Canella C, Weber M, et al: Ligaments of the transverse tarsal joint complex: MRI-anatomic correlation in cadavers. AJR Am J Roentgenol 193(3):662–671, 2009. doi: 10.2214/AJR.08.2084. 33. Morrison WB, Ledermann HP, Schweitzer ME: MR imaging of inflammatory conditions of the ankle and foot. Magn Reson Imaging Clin N Am 9:615–637, xi-xii, 2001. 34. Morsbach F, Bickelhaupt S, Wanner GA, et al: Reduction of metal artifacts from hip prostheses on CT images of the pelvis: Value of iterative reconstructions. Radiology 268(1):237–244, 2013. doi: 10.1148/ radiol.13122089. [Epub 2013 Mar 19]. 35. Na JB, Bergman AG, Oloff LM, et al: The flexor hallucis longus: Tenographic technique and correlation of imaging findings with surgery in 39 ankles. Radiology 236(3):974–982, 2005. 36. Narvaez JA, Narvaez J, Ortega R, et al: Painful heel: MR imaging findings. Radiographics 20:333–352, 2000. 37. Netter FH: Bony attachments of muscles of leg. In Ciba Collection of Medical Illustration (vol 8). part 1. Summit, NJ, 1991, Ciba-Geigy, p 107. 38. Newman JS, Newberg AH: Congenital tarsal coalition: Multimodality evaluation with emphasis on CT and MR imaging. Radiographics 20:321–332, 2000. 39. Nicolaou S, Liang T, Murphy DT, et al: A promising new technique for assessment of the musculoskeletal system. AJR Am J Roentgenol 199(5 Suppl):S78–S86, 2012. doi: 10.2214/AJR.12.9117. 40. Petrover D, Schweitzer ME, Laredo JD: Anterior process calcaneal fractures: A systematic evaluation of associated conditions. Skeletal Radiol 36:627–632, 2007. 41. Pinney SJ, Sangeorzan BJ: Fractures of the tarsal bones. Orthop Clin North Am 32:21–33, 2001. 42. Protas JM, Kornblatt BA: Fractures of the lateral margin of the distal tibia: The Tillaux fracture. Radiology 138:55, 1981. 43. Resnick D: Talar ridges, osteophytes, and beaks: A radiologic commentary. Radiology 151:329–332, 1984. 44. Rogers LF, Poznanski AK: Imaging of epiphyseal injuries. Radiology 191:297, 1994. 45. Rosenberg ZS: Chronic rupture of the posterior tibial tendon. Magn Reson Imaging Clin N Am 2:79–87, 1994. 46. Russo A1, Zappia M, Reginelli A, et al: Ankle impingement: A review of multimodality imaging approach. Musculoskelet Surg 97(Suppl 2):S161–S168, 2013. doi: 10.1007/s12306-013-0286-8. [Epub 2013 Aug 15]. 47. Schweitzer ME, Karasick D: MR imaging of disorders of the Achilles tendon. AJR Am J Roentgenol 175:613–625, 2000. 48. Schweitzer ME, Karasick D: MR imaging of disorders of the posterior tibialis tendon. AJR Am J Roentgenol 175:627–635, 2000. 49. Slomann HC: On coalition calcaneo-navicularis. J Orthop Surg 3:586, 1921. 50. Sobel M, Pavlov H, Geppert MJ, et al: Painful os peroneum syndrome: A spectrum of conditions responsible for plantar lateral foot pain. Foot Ankle 15:112–124, 1994. 51. Sommer OJ, Kladosek A, Weiler V, et al: Rheumatoid arthritis: A practical guide to state-of-the-art imaging, image interpretation, and clinical implications. Radiographics 25:381–398, 2005. 52. Stormont DM, Peterson HA: The relative incidence of tarsal coalition. Clin Orthop Relat Res 181:28–36, 1983. 53. von Knoch F, Reckord U, von Knoch M, et al: Fracture of the lateral process of the talus in snowboarders. J Bone Joint Surg Br 89:772–777, 2007.

65  High-Resolution 3T Magnetic Resonance Neurography: Applications, Techniques, and Pitfalls Majid Chalian, Brian Cox, Bahar Mansoori, Mark R. Robbin, John A. Carrino, and Avneesh Chhabra

Dedicated magnetic resonance imaging (MRI) of peripheral nerves is referred to as magnetic resonance neurography (MRN). From the time Howe and Filler et al.48 originally described MRN, this technique has evolved with constant addition of newer and optimized MRI techniques, and it is being increasingly used in clinical practice.

PATHOPHYSIOLOGY OF NERVE DISEASES Basic nerve anatomy, pathophysiology, and injury grading should be understood to detect deviations from the normal morphology and accurately identify injury and entrapment related to nerve pathology. Peripheral nerves are covered by an outer epineurium, which contains multiple fascicles that are covered by perineurium. The fascicles contain endoneurium-covered axons. The peripheral nerves may be host to a large variety of diseases, which can be classified based on the functional involvement (motor, sensory, mixed, or autonomic) or the anatomic site of involvement (neuromuscular junction or nerve-cell body, axon, and myelin sheath). Nerve diseases can be categorized into three large groups: Group 1: systemic diseases such as vasculitis and ischemia; toxic, endocrine, and metabolic disorders such as diabetes mellitus, thyroid disorders, hyperlipidemia, and amyloidosis; acute and chronic demyelinating inflammatory neuropathies, hereditary neuropathies, and others. MRN is not expected to make these diagnoses, but it can be used to confirm a diagnosis by demonstrating the primary signs of neuropathy and secondary signs of regional muscle denervation changes while excluding any structural cause for mononeuropathy. Group 2: local conditions such as nerve injury, plexopathy, peripheral compressive lesions, and adhesive neuropathy in failed tunnel release cases, nerve sheath tumors, and so forth. MRN has a vital role in this group of diseases and provides supplementary information gained from clinical examination and electrodiagnostic tests by allowing direct visualization and interrogation of the nerves and muscles. Group 3: neuropathies related to functional anatomic changes, including habitual leg crossing, repeated typing, or repetitive exercise, which may cause traction as well as compressive neuropathy via functional compartment syndromes. MRN is rarely used in this group because the diagnosis is based on clinical examination and pressure catheter studies. However, MRN can play a role in complicated cases, with imaging before and after the instigating exercise/ effort in question to detect muscle signal alterations.30

2344

Nerve injuries can be related to blunt or penetrating trauma. Iatrogenic causes also play a role, such as nerve injuries after surgical plating, local surgery, or cryotherapy/radiofrequency ablation. Nerve injuries can also occur with fractures or fracture-dislocations, such as with hip dislocation (sciatic nerve), shoulder dislocation (axillary nerve), humeral fracture (radial nerve), and elbow fracture/dislocation (ulnar nerve). Nerve injuries are traditionally classified based on the Seddon and Sunderland classification system71,74 (Table 65-1). Seddon classification describes three different types of nerve injuries: neurapraxia, axonotmesis, and neurotmesis. Neurapraxia is the least severe type of nerve injury and is related to a temporary block of nerve conduction. Neurapraxia usually results from blunt trauma with predominant myelin sheath injury and usually resolves completely in less than 6 to 8 weeks. MRN may show abnormal increased signal on T2-weighted images without muscle denervation changes; no wallerian degenerative changes occur in this process, and electromyography (EMG) studies are normal. Nerve conduction velocity (NCV) studies may demonstrate temporary slowing or conduction block. In the authors’ experience, neural fascicular changes (disruption, effacement, or enlargement) are generally not seen in this type of mild nerve injury. Axonotmesis is a more severe nerve injury that results from severe nerve contusion or crush injury leading to axonal injury and wallerian degeneration distally. Axonotmesis results in loss of axonal continuity and disruption of the endoneurium with preservation of the epineurium and perineurium, and there is both sensory and motor function loss. EMG and NCV are both abnormal. On MRN there are muscle denervation changes with increased T2 signal intensity of the nerve, as well as neural fascicular enlargement, disruption, or effacement. Since the peripheral myelin sheath is intact, functional loss related to axonotmesis eventually recovers, albeit slowly. Neurotmesis is the most severe type of nerve injury, leading to axonal injury as well as disruption of the surrounding perineurium and epineurium. There is complete loss of motor, sensory, or autonomic function. EMG is usually unable to differentiate between axonotmesis and neurotmesis. On MRN there is complete transection of the nerve with end-bulb neuroma formation or sometimes neuroma in continuity. This type of injury does not have much potential for recovery. MRN plays a significant role in differentiating a reversible injury from irreversible injury by depicting nerve discontinuity and neuroma, and may therefore help guide treatment planning by identifying cases amenable to surgical intervention. Sunderland grading is  a further expanded injury grading system based on an inside-out anatomy model (i.e., injuries to endoneurium, perineurium, and

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

2345

TABLE 65-1  Combined Seddon and Sunderland Classification of Nerve Injuries Degree of Nerve Injury

Myelin

Axon

Endoneurium

Perineurium

Epineurium

MRN Findings

Recovery

I: Neurapraxia

+/−

No

No

No

No

Full

II: Axonotmesis

Yes

Yes

No

No

No

III IV

Yes Yes

Yes Yes

Yes Yes

No Yes

No No

V: Neurotmesis

Yes

Yes

Yes

Yes

Yes

VI: DellonMackinnon

Combination of above

Nerve: increased T2 SI Muscle: normal Nerve: increased T2 SI, abnormal fascicles; mildly enlarges nerves Muscle: denervation Same as II; worsening abnormalities Nerve: focal enlargement, effaced/disrupted fascicles, neuroma in continuity Muscle: denervation Nerve: complete discontinuity Muscle: denervation Nerve: mixed T1/T2 SI; surrounding soft tissue, ligament, and muscle injury changes

Full

Slow/incomplete Poor

None Poor to none

From Chhabra A, et al: Nerve anatomy, pathophysiology, nerve injury and entrapment. In Chhabra A, Andreisek G, editors: Magnetic resonance neurography. New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 1–9. MRN, magnetic resonance neurography; SI, signal intensity.

epineurium are encompassed in grades 3/4/5 in this system). The respective MRN findings are shown in Table 65-1. Dellon and Mackinnon added a sixth grade based on outside-in partial circumference injury to the nerve. The worst type of Sunderland (1-V) injury grade dictates the prognosis in such injury.34,40

COMMON APPLICATIONS OF MRN MRN provides information that may assist in diagnosis, therapeutic follow-up, and surgical planning for peripheral nerve disorders.30 The following are the most common indications for MRN based on the authors’ experience: 1. To confirm the clinical suspicion of peripheral neuropathy 2. To grade nerve injury 3. To assess the extent of the nerve abnormality 4. To perform preoperative mapping before neurolysis, as well as postoperative evaluation 5. To differentiate between postradiation changes and recurrent malignancy 6. To depict the intraneural or extraneural space-occupying lesions causing neuropathy 7. To help exclude the diagnosis of major neuropathy by depicting normal nerves and regional muscles 8. To provide imaging guidance for perineural and perimuscular injections

HIGH-RESOLUTION MRN TECHNIQUE MRN is a tissue-selective MRI technique. This technique may be broadly categorized as conventional and diffusion-based sequences. Conventional imaging encompassing T1- and T2-weighted images provides good anatomic and pathologic imaging. It is also widely available and provides consistent and reproducible imaging quality. Although still being optimized for widespread clinical application, diffusion-based imaging allows nerve signal quantification and the potential for functional imaging. The essentials of good conventional MRN are spin echo–type contrast, thin-section (2-4 mm), two-dimensional (2D),

T1-weighted and fat-suppressed T2-weighted MRI with in-plane resolution of 0.3 to 0.5 mm, uniform and excellent fat suppression, high echo times (TE > 60 milliseconds) to minimize magic-angle artifacts, and superior resolution (256 × 392 or higher) employing a combination of 2D nerve perpendicular and three-dimensional (3D) isotropic images. MRN can be performed using both 1.5 tesla (T) and higher-field 3T scanners. The advantages of higher-field scanners are higher signalto-noise ratio (SNR), better gradient performance, wider available bandwidth, and improved coil design. Increased SNR helps improve spatial resolution by allowing reduced slice thickness and voxel size. On non–fat-suppressed T2-weighted images, differentiation of abnormal high signal within the nerve from the normal hyperintense T2 signal of the adjacent fat and the fascicular morphology can be difficult to appreciate. Fat suppression can be performed in various ways, including frequency-selective fat saturation, short tau inversion recovery (STIR), spectral adiabatic inversion recovery (SPAIR), chemical shift imaging (CSI), and Dixon-type fat-suppression technique. Fat suppression can be challenging and often inhomogeneous along curvatures of the extremities, particularly on low-field scanners. STIR images frequently provide superior fat saturation, especially  on low-field scanners. However, STIR images have poorer spatial resolution, longer acquisition time, and less SNR than fast-spin echo (FSE) imaging overall, and they are prone to pulsation artifacts. 3D imaging is also degraded on low-field scanners; only gradient echo– type 3D sequences are feasible on low-field scanners to reduce the time penalty. Given the above-mentioned consequences, the authors recommend MRN on 3T scanners whenever possible, except when there is expected interference due to susceptibility artifacts, such as from metal prostheses. Susceptibility and chemical shift artifacts are aggravated on 3T images, although these effects can be partially mitigated using SPAIR imaging, employing parallel imaging, and increasing the bandwidth. To perform 3D imaging, the authors recommend isotropic (sub1.5 mm), predominantly spin echo–type sequences in most studies. These are single-slab turbo-spin echo (TSE)-type images with variable flip angle evolutions. The parameters can be adjusted, and output can

2346

PART II  CT and MR Imaging of the Whole Body

be set to provide T2-weighted, proton density (PD)-, SPAIR-, or STIRtype contrasts. 3D SPAIR TSE provides better images for extremity MRN as compared to STIR TSE, owing to higher SNR. 3D STIR TSE is ideal for brachial and lumbosacral (LS) plexus exams because of its better fat suppression. 3D SHINKEI (3D nerve-SHeath signal increased with INKed rest-tissue rarE Imaging [rapid acquisition with refocused echoes]) (Philips, Best, Netherlands) provides blood signal suppression to provide better nerve identification in the LS plexus. For these  3D images, the authors recommend a repetition time (TR) of 1500 to 2000 milliseconds, echo time (TE) of 70 to 80 effective, and 1.2 to 1.5 mm isotropic voxel. In the future, 3D Dixon sequences will be available with some diffusion weighting (b value, 60-80 sec/mm2) for higher contrast and spatial resolution. In general, these 3D sequences null the arterial signal and allow easier identification of nerves in multiple isotropic planes. In plexus cases, one should add axial and sagittal T2-weighted images of the spine or alternatively 3D non–fatsuppressed T2-weighted TSE sequences, since spine pathology is a major confounder in plexus pathology. These 3D images allow multiplanar isotropic reformats (MPR), curved planar reformats (CPR), and maximum intensity projection (MIP) to demonstrate the longitudinal extent of neural lesions. The use of intravenous contrast (gadolinium) provides limited value in peripheral neuropathies; in our practice, it is restricted to cases of suspected infections, acute-subacute inflammation, diffuse peripheral nerve lesions, and tumors. Intravenous contrast may better show enhancement of the denervation changes of muscles; however, these changes can already be apparent on fat-suppressed T2-weighted or STIR images. Diffusion-based MRN is technically demanding and remains in the experimental phase. It increases nerve conspicuity by suppressing the vascular signal, especially on higher b values. For nerve identification, one needs b values more than 60 to 80, whereas for tractography, one needs b values of approximately 600 to 1000 and 12 to 15 directions of interrogation. One may increase the specificity of MRN by parametric analysis of diffusion tensor imaging (DTI) data, including measurement of apparent diffusion coefficient (ADC) and fractional anisotropy (FA). DTI allows generation of quantitative values that can be used for follow-up assessment of nerve regeneration/degeneration. Diffusionbased 3D reversed fast imaging in steady-state free precession (FISP), or 3D DW-PSIF, is a hybrid-type sequence that can be used for effective vascular suppression and selective nerve distribution in extremities. Its parameters include TR, 12 milliseconds; TE, 2.5 milliseconds; b value, 60 to 80; and voxel, 0.8 mm isotropic. It has all the advantages of a 3D isotropic sequence, including MPR, CPR, and MIP.3,16,17,22,23,27 To perform extremity MRN, one could use a combination of axial T1 weighting, axial T2 SPAIR or T2Dixon, sagittal fat-suppressed proton density–weighting (fsPDW), coronal 3D SPAIR TSE, coronal 3D DW-PSIF, and axial DTI. To perform MRN of the lumbar plexus, one would use a combination of axial T1, axial T2 SPAIR or T2Dixon, coronal 3D STIR TSE or SHINKEI, sagittal 3D T2TSE for spine and axial DTI. Optimal brachial plexus MRN has similar requirements as LS plexus imaging, the principle difference being that axial T2 SPAIR of the LS plexus sequence is replaced by sagittal STIR on both sides.

nicely demonstrate fascicular anatomy and perineural fibrosis (strandlike T1 and T2 hypointensities). Atrophy and fatty replacement of the regional skeletal muscles are also better appreciated on T1-weighted images. T2-weighted images are better for detection of abnormal signal intensity of the nerve (approaching the signal intensity of the adjacent in-plane vessels). Normal nerves are isointense to regional muscles on T1 and T2 sequences and demonstrate a smooth course with uniformlysized fascicles, clean surrounding perineural fat planes, and caliber almost similar to adjacent arteries that gradually taper distally. Mild nonspecific T2 hyperintensity due to subclinical neuropathy or magicangle artifact can be seen. If the nerve is pathologic, one would see constant abnormal signal along the longer segment of the nerve beyond the expected location of magic-angle artifact (e.g., sciatic nerves show magic-angle artifact at the sciatic notch, whereas patients with sciatic neuropathy exhibit asymmetric increased signal of the nerve beyond the expected notch location). Normal nerves do not enhance, owing to a preserved blood-nerve barrier. Fascicle evaluation may demonstrate enlargement, effacement, or disruption due to distal obstruction, axonal degeneration, venous congestion and hyperemia, or direct injury. Postinjury neuroma can be differentiated from peripheral nerve sheath tumor (PNST) based on nonenhancement of the injury-related former lesion. Symmetric increased signal intensity is particularly common on 3D STIR images and is due to the high fluid sensitivity of these images and extremely good fat suppression. It is seen in all plexus nerves in the retroperitoneal course, whereas isointense signal is seen beyond the inguinal ligament. Therefore one should look for asymmetric signal alteration and other findings suggesting neuropathy, as described earlier. STIR sequences are also very useful for brachial and LS plexus abnormalities. Split variations of nerves (e.g., sciatic, femoral nerves) are common, and these findings by themselves are not pathologic. 3D images should be reconstructed as thick-slab (5-20 mm) MIP images to reduce image blurring and noise and improve conspicuity of the peripheral nerves. 3D images are great for looking at gross abnormal findings such as neuroma formation, focal or diffuse nerve swelling, nerve displacement or abnormal angulation, or nerve discontinuity. Secondary signs of neuropathy are very informative. For instance, muscle denervation changes can give a rough estimate of the acuity of the injury or lesion; acute-phase pathology demonstrates edema-like T2 signal intensity without fatty infiltration or atrophy, subacute phase pathology depicts muscle edema with  mild fatty infiltration with or without atrophy, and chronic phase pathology shows frank muscle atrophy and fatty infiltration. As a general rule, diffuse muscle involvement and lack of fascial edema/ enhancement can help in differentiating denervation changes from myositis and muscle strains. It is beyond the scope of this chapter to describe evaluation of peripheral nerves based on DWI sequences, given their technically demanding and experimental nature. It suffices to mention that neuropathic nerves are easily seen on tensor images and depict low FA and high ADC values. With nerve regeneration, the FA values increase and ADC values decrease. For nerve sheath tumors, low ADC values (<1.1 × 10−3 mm2/sec) suggest extremely high cellularity and likely malignancy.10,16,23,34,39,40,53

HOW TO INTERPRET MRN

MRN OF UPPER EXTREMITY NEUROPATHY

Accurate MRN interpretation requires detailed knowledge of peripheral nerve anatomy, its common variations, and nerve pathophysiology in various diseases. For 2D evaluation of peripheral nerves, T1- and T2-weighted images should be evaluated in tandem on the reading workstation. T1 images are useful for anatomic evaluation and differentiation of nerves from the adjacent vessels. T1-weighted images

Upper extremity nerves, especially when they pass through tunnels,  are vulnerable to intrinsic or extrinsic abnormalities that lead to compression or impingement syndromes. Tunnel syndromes may results  from various etiologies, including acromegaly, pregnancy, myxedema, rheumatoid arthritis, space-occupying lesion (e.g., hematoma), fracture deformity, tenosynovitis, ganglion cyst, anomalous muscle, 

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography varicosities, nerve sheath tumors, and other regional benign or malignant tumors. Traditionally EMG and NCV have been the diagnostic modalities of choice for evaluation of upper extremity nerve diseases. However, these studies provide limited information about the nature and severity of the lesion, existence of neuroma, or precise location of the pathology. High-resolution ultrasound (US) has the ability to provide more anatomic information but is very operator dependent, poor in the assessment of deep nerves, and gives limited or no information about regional muscle changes.26,55,65

Median Nerve Lesions and Associated Syndromes The median nerve arises from the medial roots of C8-T1 and lateral roots of C5-C7. It courses along the humerus between the biceps brachii and brachialis muscles and beneath the bicipital aponeurosis. It may travel beneath the ligament of Struthers that connects the supracondylar process to the medial epicondyle. The anterior interosseous nerve (AIN) arises from the radial aspect of the median nerve 5 to 8 cm distal to the medial epicondyle of the humerus. The median nerve may be injured or entrapped at different levels, producing different clinical symptoms.2,41,51

Ligament of Struthers and Supracondylar Spur.  Radiographs usually demonstrate a supracondylar spur when present. MRN shows high T2 intensity with or without nerve flattening, and sometimes the culprit ligament. Symptoms include paresthesia in the thumb, index and middle fingers, pain in the volar aspect of forearm, and weakness of forearm flexors. Treatment includes rehabilitation or surgical release, based on the severity of symptoms.

Pronator Syndrome.  Pronator syndrome may be due to a hypertrophied pronator muscle, thickened bicipital aponeurosis, thickened fascial edge of the flexor digitorum superficialis (FDS), aberrant median artery, or any kind of regional space-occupying lesion. Symptoms include pain in the volar forearm and proximal arm, paresthesia of the thumb, index, and middle fingers, and positive Tinel’s sign. EMG is positive in about two thirds of patients. MRN demonstrates moderate to severe T2 hyperintensity, with nerve flattening at the entrapment site. Proximal nerve enlargement may be present. Mild hyperintensity can be seen in asymptomatic cases. Treatment includes rehabilitation or surgical release, based on the severity of symptoms.

Anterior Interosseous Nerve Syndrome (Kiloh-Nevin Syndrome).  The AIN is predominantly a motor nerve, and it is the largest branch of the median nerve. It courses between the two heads of the pronator teres muscle and reaches the anterior aspect of the interosseous membrane to travel beside the anterior interosseous artery. The AIN provides motor innervation to the radial belly of the flexor digitorum profundus (FDP), flexor pollicis longus (FPL), and pronator quadratus (PQ) muscles. It also provides sensory innervations to the wrist. Compression of the nerve results in weakness of pinching, even though PQ weakness (pronation) may be compensated for by the concurrent action of the pronator teres. There are many causes of AIN palsy: fibrous bands from the deep (more common) or superficial head of the pronator teres to the brachialis fascia, fractures of the forearm bones or supracondylar fracture, local pressure such as that resulting from prolonged atypical sleeping position or a poorly applied cast, exercise and weightlifting, and viral neuritis. On MRI of AIN neuropathy, changes in muscle denervation are easily observed in a typical distribution. Although the PQ muscle is always the first one involved in AIN neuropathy, it should be noted that isolated PQ signal change or atrophy is also commonly seen as an asymptomatic finding. On high-resolution MRN, abnormal signal of

2347

the AIN itself can easily be differentiated from adjacent vessels, and changes in PQ size are appreciated in cases of neuropathy.26

Carpal Tunnel Syndrome.  Carpal tunnel syndrome (CTS) is the most common neuropathy of the upper extremity. The carpal tunnel is a fibroosseous canal formed by carpal bones and a rigid fibrous transverse carpal ligament. The median nerve branches into a recurrent motor branch that supplies the thenar muscles and multiple sensory branches that provide innervation to the three radial digits and the radial half of the fourth digit. The pathophysiology of CTS involves a combination of mechanical trauma, increased intratunnel pressure, and ischemic injury to the median nerve, leading to perineurial and epineurial edema and hyperemia, followed by demyelination and axonal loss. Clinically these patients complain of nocturnal pain, hand clumsiness, tingling, numbness, and paresthesia in the distribution of the median nerve. Sensory loss typically precedes the motor deficit. Patients with long-standing disease may have wasting of the soft tissues of the thenar eminence. Clinically there is a positive Phalen test (increased paresthesia after 1 minute of passive wrist flexion) and Tinel’s sign (paresthesia in nerve territory after gentle tapping over carpal tunnel). Although electrophysiology studies are frequently performed for wrist neuropathies, false-negative rates can be as high as 10% to 30% and positive predictive values can be as low as 33%, which clearly indicate a need for improved diagnosis. Isolated MRI signs of abnormal nerve T2 hyperintensity have low reported specificities of less than 40%. Other signs such as flattening of the median nerve in the carpal tunnel, proximal median nerve enlargement, volar bowing of the retinaculum, extension of T2 hyperintensity distally in the median nerve branches, and effacement of the deep fat pad of the carpal tunnel increase the specificity of the diagnosis of CTS. MR neurography also shows tenosynovitis, internal wrist derangement, and changes in regional muscle denervation in a noninvasive fashion. 3T MR neurography depicts the individual bundle abnormality of a bifid nerve, and nerve abnormalities can also be shown in the longitudinal plane of the nerve (Fig. 65-1). Failure to recognize the bifid components of the median nerve (Fig. 65-2) may lead to accidental injury to the median nerve during surgical intervention.11,50,51,58

Ulnar Nerve Lesions and Associated Syndromes The ulnar nerve arises from the medial cord of the brachial plexus (C8-T1 and sometimes C7) and courses along the medial border of the medial head of the triceps muscle. The nerve travels beneath the medial intermuscular septum and descends posteriorly through the cubital tunnel. The nerve courses distally between the two heads of flexor carpi ulnaris (FCU) muscle (Osborne band) and reaches the Guyon’s canal (ulnar tunnel). It provides a palmar cutaneous branch and a dorsal cutaneous branch to the medial 1.5 digits. The ulnar nerve gives off motor branches to the hypothenar, third palmar and fourth dorsal interossei, third and fourth lumbricals, adductor pollicis, and flexor pollicis brevis muscles. Like other major nerves, the ulnar nerve may be injured or entrapped at different levels, producing different clinical syndromes.

Cubital Tunnel Syndrome.  Also known as sleep palsy or cell phone neuritis, cubital tunnel syndrome is usually due to repetitive flexion of the elbow joint. Sometimes prolonged external pressure (driver’s elbow) can cause this syndrome. It can also be secondary to other intraneural or extraneural compressive lesions described earlier. Although clinical examination, EMG, NCV, and US can be diagnostic, MRN can clearly depict the underlying problem in most cases (Fig. 65-3). Mild increased T2 signal intensity can be seen in asymptomatic

2348

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 65-1  Carpal tunnel syndrome. A, Axial T1 image shows effacement of the deep fat within the carpal tunnel (arrow). B, Axial T2 SPAIR image shows enlarged median nerve with moderate increased signal within the carpal tunnel (arrow). C, Coronal T2 gradient recalled echo (GRE) PSIF MIP image shows mild to moderate enlargement with mild increased signal of the median nerve proximal as well as distal to the carpal tunnel, with flattening within the carpal tunnel.

individuals. The T2 signal intensity and its longitudinal extent increase with worsening neuropathy. Enlarged fascicles, nerve enlargement and flattening, angulation or displacement, and regional muscle changes can be visualized using MRN. Treatment includes rehabilitation or surgical release. Reentrapment of the anteriorly transposed nerve may occur secondarily to overzealous dissection and development of perineural fibrosis, leading to abnormal nerve flattening and angulation. MRN is quite useful in these cases as well.6,9,49,77

drome is in the detection of space-occupying lesions within the canal and/or signal intensity and size changes of the nerve itself. Secondary signs of neuropathy include denervation atrophy/edema of the hypothenar muscles, third/fourth lumbricals, and interossei muscles. Isolated deep-bundle neuropathy may also be diagnosed on high-resolution 3T MRN, which is a difficult diagnosis on electrodiagnostic testing. In the authors’ experience, neuropathy in Guyon’s canal without perineural fibrosis or mass lesion is very uncommon.26

Guyon’s Canal Syndrome.  The ulnar nerve can be affected at this

Radial Nerve Lesions and Associated Syndromes

level by external compression, space-occupying lesions, or chronic traction neuritis. Symptoms vary based on three zonal locations of involvement: Zone 1: proximal edge of the volar carpal ligament to bifurcation of the ulnar nerve, where patients present with combined motor and sensory deficits Zone 2: from the nerve bifurcation to just distal to the fibrous arch of the hypothenar muscles, where patients may present with pure motor deficit Zone 3: distal end of the canal containing the superficial sensory branch, where patients may present with pure sensory deficit Lesions in zones 1 and 3 are more common. Symptoms include sensory loss of the volar aspect of the medial fourth and fifth digits, with sparing of the dorsal aspect of the medial hand and digits and palmar aspect of the proximal hand because the respective nerves arise 5 cm above the wrist joint. The value of MRI in Guyon’s canal syn-

The radial nerve arises from the posterior cord of the brachial plexus (C6-T1). It courses posterior to the axillary and brachial arteries and anterior to the latissimus dorsi and teres major tendons, and it enters the spiral groove distally. It pierces the lateral intermuscular septum and arrives in the front of the lateral epicondyle. It enters the radial tunnel at the level of the capitellohumeral joint and divides into superficial and deep branches. The deep motor branch is also called the posterior interosseous nerve (PIN). The radial nerve innervates the triceps and anconeus above the spiral groove and the brachioradialis below the spiral groove. The PIN supplies all the other muscles of the dorsolateral aspect of the forearm. Anatomic sites where the radial nerve is prone to injury and entrapment are associated with several syndromes.

Spiral Groove Syndrome.  Spiral groove syndrome represents the most common type of radial nerve injury and mostly occurs after

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

2349

C

A

B FIG 65-2  Sunderland grade 5 median nerve injury. A, T2 GRE PSIF image shows a bifid median nerve (long arrows) with ulnar band gap (short arrow), indicating complete nerve discontinuity. B, Axial T2 SPAIR image shows fluid in the ulnar band gap (short arrow) and the normal radial band (long arrow). C, Coronal PSIF MIP image shows hyperintense proximal neuroma formation.

trauma/fracture. Less commonly it is due to a thickened lateral intermuscular septum or a prominent myotendinous edge of the lateral head of the triceps muscle, leading to “Saturday night palsy” due to improper positioning of the upper extremity during sleep, or it is secondary to other common pathologies affecting nerves in different locations as previously described. Symptoms include weak elbow flexion and supination, wrist extension, and finger extension. EMG is useful for diagnosis; however, MRN might be needed to demonstrate the underlying cause, such as thickened fascia or PNST. MRN shows abnormally increased T2 signal intensity at or above the spiral groove. Certain conditions, such as hereditary neuropathy with liability to pressure palsies, may be associated with multiple nerve entrapments (Fig. 65-4), including radial nerve entrapment in the spiral groove.26

Radial Tunnel Syndrome.  Radial tunnel syndrome is a distinct but controversial clinical entity, often occurring as the result of repetitive motion injuries to the elbow; it is distinguished by the general absence of motor weakness. Patients typically report pain in the region of the

proximal extensor and supinator muscles, exacerbated by forearm pronation, extension of the elbow, and flexion of the wrist. Although US is currently the modality of choice for evaluating the superficial sensory branch or for detection of an offending mass lesion such as a ganglion cyst, technological advances in high-resolution MRN allow demonstration of the abnormal nerve, associated regional joint derangement, and characterization of the soft tissue lesion. Accurate localization of nerve compression is a key factor in establishing the diagnosis and planning surgical treatment.37,61,70

PIN Syndrome (Supinator Syndrome).  Isolated motor neuropathy is the hallmark feature of PIN syndrome. Symptoms include loss of digit extension and accompanying radial wrist deviation related to weakness of extensor carpi ulnaris. Note should be made that many cases of refractory “tennis elbow” (i.e., lateral epicondylitis) may be related to PIN entrapment, since there is clinical overlap in their presentation. The only areas of the radial nerve and PIN prone to magicangle artifacts are in the axilla and intramuscular course through the

2350

PART II  CT and MR Imaging of the Whole Body

A

C

B FIG 65-3  Ulnar nerve reentrapment. A, Axial T2 SPAIR image shows a hyperintense enlarged ulnar nerve (arrow) at the level of the cubital tunnel. B, Axial T1 image shows scarring (arrows) from prior tunnel release surgery. C, Oblique sagittal PSIF MIP image shows the “triple B sign” of a bright enlarged ulnar nerve, black gap at the site of nerve entrapment (short arrow), and bright ulnar nerve distal to the entrapment site.

supinator muscle, respectively. At the latter site, however, the signal intensity of the normal PIN does not approach the adjacent in-plane veins. Patients with neuropathy show abnormal hyperintensity at the entrapment or injury site. Nerve enlargement may or may not occur. However, denervation edema in the associated muscles is the most common MRI manifestation of the syndrome and should not be overlooked.57

Wartenberg Syndrome (Cheiralgia Paresthetica).  This rare entity is a compression injury of the superficial radial nerve, often at the location of piercing of the deep fascia. It results from repetitive athletic activities such as rowing or throwing and is less likely from trauma or mass lesions. Symptoms include tingling, numbness, or pain at the posterior aspect of the hand near the anatomic snuffbox. MRN may show mass lesions such as a ganglion cyst or tumor compressing the nerve.

MRN OF LOWER EXTREMITY NEUROPATHY The etiologies, traditional diagnostic tools, and added value of MRN in evaluation of lower extremity neuropathies are similar to those affecting the upper extremity.

Sciatic Nerve Lesions and Associated Syndromes The sciatic nerve arises from the ventral rami of the L4-S3 nerve roots, which form three main nerve trunks: the tibial nerve (TN) medially, the common peroneal nerve (CPN) laterally, and the posterior cutaneous nerve of the thigh far posteriorly. The sciatic nerve exits the pelvis through the greater sciatic foramen and travels below the piriformis muscle. The TN and CPN are enclosed in a common sheath. The sciatic nerve then continues distally in the thigh between the adductor magnus and gluteus maximus muscles. The sciatic nerve divides into the TN and CPN in the distal third of the thigh. The TN

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

A

B

2351

C

FIG 65-4  Hereditary neuropathy with liability to pressure palsy. Axial T2 SPAIR image (A) and oblique coronal T2 GRE PSIF MIP image (B) show increased signal in the radial nerve at and distal to the spiral groove of the humerus (arrows). C, Oblique coronal DTI MIP image shows the hyperintense radial nerve and the resulting extensor compartment muscle edema (arrows) indicating denervation changes.

innervates the long head of the biceps femoris, semitendinosus, semimembranosus, and adductor magnus muscles, and the CPN innervates the short head of the biceps femoris. The long course of the sciatic nerve makes it prone to neuropathy at different levels, such as at the LS plexus level, sciatic notch, gluteal region of the hip, or the distal thigh region above the level of bifurcation. A myriad of systemic and local causes of sciatic neuropathy exist, similar to the other nerves (Figs. 65-5 and 65-6). One common tunnel syndrome is piriformis syndrome.8,18,35,42,54

Piriformis Syndrome.  Although spinal stenosis is the most common cause of sciatica, compression can occur anywhere along the course of the nerve. Compression of the sciatic nerve at the level of the piriformis muscle is called piriformis syndrome. This can result from various etiologies including (1) intramuscular course of the nerve, (2) posttraumatic changes including hematoma, fibroma, or myositis ossificans, (3) hypertrophied piriformis muscle in athletics, (4) accessory belly of the piriformis, and (5) intra- and extraperitoneal mass lesions including endometriosis. This syndrome presents with pain and paresthesia in the gluteal region, radiating downward posteriorly. The CPN component of the sciatic nerve is more vulnerable owing to its superficial and long course. Motor symptoms may manifest as difficulty in knee flexion, loss of foot dorsiflexion, and weak plantar flexion, inversion, or eversion. EMG has a limited role in the evaluation of sciatica owing to its deep location. However, MRN can directly visualize the affected nerve and extraneural soft tissues. A split sciatic nerve is a pertinent positive when evaluating for piriformis syndrome (Fig. 65-7). MRN usually shows asymmetric T2 hyperintensity or flattening of the nerve. High diagnostic accuracy is achieved when signal intensity approaches 90% signal intensity of the adjacent vessel. MRN also demonstrates fascicular effacement or enlargement in cases of nerve entrapment or infiltration by tumor, infection, or endometriosis. MRN can directly visualize PNSTs. Comparison with the contralateral side is always helpful for identification of the pathology. DTI images demonstrate the abnormal signal alteration quite well. 3D thick-slab MIP images from DTI can be obtained similar to conventional anatomic

3D images, leading to elegant depiction of nerve pathology. Treatment includes rehabilitation, steroid injection, Botox injection into the muscle, or surgical decompression. Identification of nerve pathology is helpful in directing CT- or MRI-guided perineural or intramuscular Botox injection into the piriformis muscle.8,18,19,35,38,42,54

Femoral Nerve Lesions The femoral nerve arises from the dorsal roots of L2-4, is a mixed sensory-motor nerve, and descends distally, coursing between the psoas major and iliacus muscles. It travels deep to the inguinal ligament as the most lateral structure of the femoral canal, and then divides into anterior and posterior divisions. The anterior division provides sensory branches (i.e., the medial and intermediate cutaneous nerves of the thigh) and gives off motor branches to the pectineus and sartorius muscles.20 The posterior division divides into the sensory saphenous nerve and muscular and articular branches. The saphenous nerve travels through Hunter’s canal along the medial aspect of distal thigh and gives off sartorial and infrapatellar cutaneous branches. Lesions affecting the femoral nerve are generally similar to the other extremity nerves. MRN plays an important role in visualization of entrapment neuropathy, mass lesions, and nerve injury. The most common site of entrapment is distal to inguinal ligament, which acts to protect the nerve from external impingement. In the authors’ experience, entrapment neuropathy without the presence of a mass lesion is rare. Femoral neuropathy can produce sensory or motor symptoms. Sensory symptoms include pain and tingling in the inguinal region, forcing the patient to hold the hip in antalgic position, as well as pain in the anterior thigh and anteromedial leg. Motor findings include poor hip flexion and knee extension, as well as reduced knee reflexes. EMG and NCV are helpful diagnostic tools if there are muscle denervation changes. MRN can depict asymmetric T2 hyperintensity and enlargement of contributing nerve roots or the femoral nerve itself.  In cases of injury, MRN can depict the degree of injury, neuroma in continuity, or end-bulb neuroma in cases of nerve discontinuity. Again, muscle denervation changes are helpful secondary signs of femoral neuropathy. Diabetic amyotrophy is not an uncommon cause of

2352

PART II  CT and MR Imaging of the Whole Body femoral neuropathy and can be unilateral or more commonly bilateral and asymmetric.18,67

Common Peroneal Nerve Lesions and Associated Syndromes

A

B

The CPN arises from the L4-S2 nerve roots. The only nerve that is innervated by the CPN division of the sciatic nerve above the knee is the short head of the biceps femoris. It gives rise to the lateral cutaneous nerve of the calf, while the medial cutaneous nerve of the calf arises from the TN. The lateral and medial cutaneous nerves of the calf combine more distally to form the sural nerve. The CPN then travels through the peroneal tunnel at the level of the fibular neck between the two heads of the peroneus muscle. The CPN provides sensory innervation to the anterolateral and dorsal aspect of the foot, as well as motor innervation to the anterior and lateral compartment muscles. The CPN divides into the superficial (SPN) and deep (DPN) peroneal nerves in the fibular tunnel. The SPN lies on the anterior intermuscular septum and innervates the peroneus longus and brevis muscles. Approximately 6 cm below the fibula, it divides into the medial and intermediate dorsal cutaneous nerves, where it can be compressed by the crural fascia (SPN syndrome). The DPN courses anteriorly at the level of fibular neck and pierces the intermuscular septum to innervate the anterior compartment muscles. Approximately 1.2 cm below the ankle mortis, the DPN divides into lateral and medial branches. Similar to other nerves, various lesions can affect the CPN. Also, it can be affected by extraneural compression from habitual leg crossing, anorexia, previous regional limb surgery with immobilization cast, ganglion or Baker’s cysts, and other regional space-occupying lesions. The most common location for CPN entrapment is at the level of peroneal tunnel (fibular tunnel syndrome), owing to the relatively fixed position of the nerve. At the level of the ankle, the DPN can be entrapped beneath the inferior extensor retinaculum as a result of bony productive changes, PNST, hypertrophied muscle, muscle herniation, or tight footwear (anterior tarsal tunnel syndrome). Clinical symptoms depend on the level of the nerve involvement. Although EMG and NCV can be helpful in diagnosis, MRN can visualize the underlying lesion. On MRN, CPN may show artifactual mild T2 hyperintensity at the level of the popliteal fossa due to magic-angle artifact owing to the CPN’s oblique course. Generally speaking, the size of the CPN should always be smaller than the ipsilateral TN. MRN shows abnormal nerve hyperintensity, flattening under the fascia, perineural fibrosis, and proximal enlargement (Fig. 65-8). Regional muscle denervation changes can be easily identified in the anterolateral compartment of the leg. MRN exhibits moderate to good diagnostic accuracy based on the authors’ published experience. Treatment includes rehabilitation or surgical decompression, based on the severity and nature of the underlying etiology.21,54,56,60,79

Tibial Nerve Lesions and Associated Syndromes C FIG 65-5  Mild stretch injury and neuropathy of the left sciatic nerve. A, Coronal 3D variable refocusing flip angle-TSE SHINKEI image shows mild hyperintensity and mild thickening of the left sciatic nerve. Coronal reconstructed DWI MIP (B) and axial T2 SPAIR (C) images confirm the finding.

The TN, the second major division of the sciatic nerve, arises from the L4-S3 nerve roots and is almost double the size of the CPN in diameter. The TN gives off the medial cutaneous nerve of the calf and motor branches to the posterior compartment muscles. At the level of the ankle, the TN is usually referred to as the posterior tibial nerve (PTN). The first branch of the PTN is the medial calcaneal nerve, which may originate proximal to the tarsal tunnel. The PTN travels beneath the flexor retinaculum and enters the tarsal tunnel, where it divides into the medial and lateral plantar nerves (MPN, LPN), which then travel within their own tunnels, named accordingly. The first branch of the LPN is named Baxter’s nerve (inferior calcaneal nerve) and courses along the medial inferior aspect of the calcaneus and heel spur, if present. The terminal branches of the MPN and LPN form the

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

2353

A

B FIG 65-7  Split right sciatic nerve on a coronal T1 image, a pertinent positive when evaluating for piriformis syndrome.

Proximal Tibial Neuropathy and Soleal Sling Syndrome.  Symp-

C FIG 65-6  Left sciatic and pudendal entrapment neuropathy due to posterior left hip/ischium melorheostosis. A, Axial T1 image shows sclerotic left ischium melorheostosis (long arrow) immediately adjacent to the left pudendal and sciatic nerves (medial and lateral short arrows, respectively). Axial T2 SPAIR (B) and coronal 3D SHINKEI SENSE (SENSitivity Encoding) MIP (C) images show hyperintense signal in the pudendal (short arrow) and sciatic (long arrow) nerves.

interdigital nerves. The most medial branch of the MPN is the medial plantar proper digital (MPPD) nerve, which gives sensation to the medial aspect of the great toe. On MRN, the TN can easily be seen in the posterior calf and can be followed to the tarsal tunnel. It is isointense to the adjacent muscles on T2-weighted images. The MPN and less commonly the LPN may show mild T2 hyperintensity in their tunnels in asymptomatic patients. A 3D DW-PSIF MRI sequence provides the most robust multiplanar depiction of the TN, MPN, LPN, and their major branches.

toms of proximal TN neuropathy include reduced plantar flexion, foot inversion, and sensation in the sole of the foot. Clinical symptoms and Tinel’s sign can be used to differentiate proximal TN entrapment from tarsal tunnel syndrome (TTS). The TN can be affected at this lesion by PNST, trauma, systematic disease, thickening and tightness of the fibromuscular sling of the origin of the soleus muscle (soleal sling syndrome), inflammatory processes, Baker’s cyst, ganglion cysts, arterial aneurysm, and other local space-occupying lesions.33,73 MRN is helpful to characterize the underlying pathology. PNST can be differentiated from neuroma in continuity, and intraneural or extraneural ganglion cysts by characteristic features of PNSTs, including target sign, tail sign, split-fat sign, and variable enhancement. MRN is also helpful in delineation of nerve abnormalities along with direct visualization of a thickened soleal sling in cases of soleal sling syndrome. Multifocal abnormal T2 hyperintensity or bilateral findings should raise suspicion of an underlying metabolic disorder.18,73

Tarsal Tunnel Syndrome.  TTS is defined as compression of the PTN or its branches in the tarsal tunnel, including the TN, MPN, LPN, and calcaneal nerve. Improperly fitted footwear, trauma, overuse injury, metabolic disorders, or any intraneural or extraneural spaceoccupying lesions can result in TTS. The most commonly reported symptoms are pain and tingling over the territory of the MPN. If more severe, it may radiate proximally as well (Valleix phenomenon). Tinel’s sign may be positive. EMG may or may not be useful in TTS and can be used to rule out radiculopathies and polyneuropathies.8,18 MRN can depict underlying causes of TTS such as tenosynovitis, venous malformation, PNSTs, muscle anomalies, and ganglion cysts. Mild T2 hyperintensity can be considered normal in the MPN and LPN; however, even mild hyperintensity should be viewed with suspicion in the PTN. Any nerve enlargement, flattening, or perineural scarring should be reported in the setting of suspected TTS. MRN is also useful for evaluation of nerve reentrapment or neuroma in

2354

PART II  CT and MR Imaging of the Whole Body

A

C

B FIG 65-8  Common peroneal nerve (CPN) entrapment neuropathy. A, Axial T1 image shows lateralization of the gastrocnemius (long arrow) displacing the CPN (short arrow) anteriorly in the fibular tunnel. B, Axial T2 SPAIR image shows hyperintensity of the CPN indicating moderate to severe entrapment. C, Oblique coronal T2 SPAIR MIP image shows hyperintensity of the CPN at the site of entrapment (lower arrow) relative to the proximal tibial nerve and CPN (upper arrows).

continuity formation. Readers should always combine regional skeletal muscle findings and clinical history with MRN findings. Denervation of plantar muscles is a helpful sign of neuropathy, especially in the abductor halluces or abductor digiti minimi.15,18,28

surgery, and tenosynovitis are other potential causes. On MRN, moderate to severe T2 hyperintensity with or without fascicular changes is diagnostic.18,59,68 Treatment includes rehabilitation or surgical release, based on severity and etiology.

Medial Plantar Neuropathy.  Medial plantar neuropathy is also

Lateral Plantar Neuropathy.  Lateral plantar neuropathy is also

known as medial plantar tunnel syndrome or lower tarsal tunnel syndrome. The most common locations for entrapment are between the abductor hallucis and quadratus plantae, beneath the talus or navicular bone, at the edge of the plantar fascia, or mass effect from a calcaneal spur or thickened abductor hallucis aponeurosis sling. Clinically it presents with a tingling and burning sensation over the territory of the MPN. Running with increased heel valgus and foot pronation is a common mechanism. Repetitive microtrauma from running can cause entrapment of the MPN between the abductor hallucis and a thickened master knot of Henry. This has been termed jogger’s foot.68 Trauma,

known as lateral plantar tunnel syndrome or lower tarsal tunnel syndrome. Baxter’s neuropathy is most common in athletes and results from entrapment of the first branch of the LPN, the inferior calcaneal nerve, between the abductor hallucis and quadratus plantae. Symptoms include sensory loss in the lateral third of the plantar aspect of the foot and weakness of the abductor digiti minimi muscles, which is hard to appreciate clinically; it may be misdiagnosed or associated with plantar fasciitis. On MRN, moderate to severe T2 hyperintensity with or without fascicular changes is diagnostic. Treatment includes rehabilitation or surgical release, based on severity and etiology.

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography Calcaneal Neuropathy.  This rare entity can arise from valgus hindfoot positioning or regional masses. MRN can visualize the nerve or associated fibrosis or mass on T1- and T2-weighted images.

Sural Neuropathy.  The long and superficial course of the sural nerve and its contributors predisposes this nerve to trauma or entrapment. Sural neuropathy manifests as pain, numbness, or paresthesia on the lateral aspect of ankle and foot, and it is often exacerbated at night and with exercise. The sural nerve is a preferable nerve for nerve biopsy, and hence postsurgical neuropathy may occur. On MRN, focal T2 hyperintensity, fascicular changes, neuroma formation, PNST, or nerve discontinuity may be found in the sural nerve or medial and lateral cutaneous nerves of the calf.18

Medial Plantar Proper Digital Neuropathy (Joplin’s Neuroma).  MPPD neuropathy can result from repetitive microtrauma, surgery, or ill-fitting footwear. Patients complain of pain, paresthesia, or numbness of the first medial aspect of the first toe on the plantar aspect. A palpable mass may be found on physical examination. MRN depicts fibrosis and entrapment of the MPPD nerve near the first metatarsal head and T2 hyperintensity of the nerve itself.72

Morton’s Neuroma.  Interdigital neuropathy, or Morton’s neuroma, is a fibrotic nodule arising from repetitive trauma, compression, or scarring of the nerve due to ischemia or entrapment between intermetatarsal ligaments. Therefore the term “neuroma” is a misnomer. Symptoms include pain and numbness within the interdigital web space, aggravated by activity. This entity is more common in middleaged women wearing high heels and usually occurs in the third or second web spaces. On MRN, short-axis (axial) T1-weighted images are important for the diagnosis. They can show the effacement of the intermetatarsal fat pad, with a teardrop/spindle-shaped mass on the plantar aspect of the foot. This can be associated with intermetatarsal bursitis, which projects along the dorsal aspect of the mass. MRN can confirm clinical suspicions and exclude other potential mimickers such as bursitis, gout, osteoarthritis, or rheumatoid arthritis. Symptomatic Morton’s neuromas are usually treated using image-guided injections or surgical resection.36,62,78

MRN OF THE BRACHIAL PLEXUS The brachial plexus (C5-T1) is a series of nerve ramifications that ultimately form the large terminal nerves innervating the upper extremity. These five terminal nerves include the median nerve, ulnar nerve, musculocutaneous nerve, axillary nerve, and radial nerve. Detailed description of brachial plexus anatomy is beyond the scope of this chapter. Clinical symptoms, history, EMG, and NCV studies traditionally make the diagnosis of brachial plexopathy, despite limitations due to the deep location of the plexus. However, MRN (Fig. 65-9) has been increasingly used for this purpose, given its wider availability and technical improvement to confirm the diagnosis of plexopathy and provide anatomic information for potential surgical planning.12,17,32,46,69 MRN is useful for patients with radiculopathy (Fig. 65-10), under consideration for surgery for thoracic outlet syndrome, PNST, subclavian aneurysm, trauma (Fig. 65-11), and symptomatic patients with history of radiation and suspected tumors or tumor-like lesions. Brachial plexus MRN is different from that for the peripheral nerves, owing to the small size of the nerves, complexity of the anatomy, and common anatomic variations. On STIR images, nerves show mild symmetric hyperintensity and size on either side of the midline. Any asymmetry or focal hyperintensity or enlargement/kinking should be considered suspicious in the proper clinical setting of suspected

2355

neuropathy. On sagittal images the roots and trunks are generally located posterior and superior to the flow void of the subclavian artery in the region of the interscalene triangle. The divisions are located superior to the subclavian artery. The cords are adjacent to the axillary artery in a clockwise order: lateral cord anterosuperior, posterior cord superoposterior, and medial cord in the axilla posteroinferior location. Like parts of the body, regional muscle denervation changes are helpful for localization of the abnormality.17,32

Brachial Plexus Lesions and Associated Syndromes

Thoracic Outlet Syndrome.  Thoracic outlet syndrome (TOS) is actually a misnomer insofar as symptoms are related to an abnormality of the thoracic inlet. TOC is by definition compression of a neurovascular bundle passing between the anterior and middle scalene muscles. The compression may be positional or static, such as that secondary to hypertrophied anterior scalene muscle, accessory muscle, fibrous bands, or cervical ribs. TOS can be classified as vascular (pain, swelling, discoloration), neurogenic (more commonly involves C8 root and lower trunk, produces pain and tingling of the medial forearm and hand, with atrophy of distal upper extremity muscles), combined, nonspecific/disputed (the most common form), or so-called droopy shoulder syndrome.17,32 On MRN, neurogenic TOS presents with abnormal T2 signal intensity and enlargement of C8 and T1 nerve roots, with the abnormality extending into the lower trunk. Treatment includes rehabilitation, physical therapy, corticosteroid perineural injection, Botox injection (in scalene), or surgical decompression.4,7

Brachial Plexitis.  Proposed etiologies for idiopathic brachial plexitis, a rare entity more common in young men, include inflammatory or immune responses. Brachial plexitis can be preceded by immunization, respiratory viral illness, recent surgery, or childbirth.5,44,75 Presentation includes acute onset of pain followed by severe atrophy and weakness of the shoulder muscles. Pain usually resolves in 2 to 3 weeks, but full recovery takes months or even years. Many nerves can be affected, but the ulnar nerve is usually spared. MRN can show diffuse enlargement and high T2 signal intensity, with denervation changes in the regional muscles.

Postradiation Brachial Plexopathy.  After radiation therapy for regional malignancies, we might see tumor recurrence or radiation neuropathy. Radiation neuropathy (unlikely at doses < 60 Gy) can occur within a few months to almost 20 years after radiation therapy and mostly involves the infraclavicular plexus, whereas tumor recurrence usually happens after 1 year and mostly involves C8-T1 and the lower trunk. Radiation neuropathy can manifest in three different ways: acute ischemic plexopathy related to vascular thrombosis and occlusion, transient plexopathy, and classic delayed plexopathy and fibrosis.45,66 EMG demonstrates classic “myokymic discharge” in about a third of cases. In cases of tumor recurrence, MRN shows focal nodular enlargement and enhancement. However, in radiation plexopathy, the abnormality is regional and geographic, diffuse, and without focal mass like lesion or enhancement. Postradiation fibrosis can be seen as T1- and T2-hypointense fat stranding with distortion of the nerve segments.32 Isolated involvement of the suprascapular nerve (C5-C6),14 axillary nerve (C5-C6),76 musculocutaneous nerve (C5-C7),24 and long thoracic nerve (C5-C7)60 has been reported in the literature. General etiologies are similar to peripheral nerves, albeit with few nerve-specific entrapment syndromes and mechanisms of injury. MRN would be helpful in localization of nerve lesions, regional compressions/ entrapments, or preoperative mapping.

2356

PART II  CT and MR Imaging of the Whole Body

MRN OF THE LUMBOSACRAL PLEXUS

FIG 65-9  Normal brachial plexus. Coronal 3D IR TSE MIP image shows normal symmetric nerve signal and size (arrows).

The LS plexus is a series of nerve ramifications that ultimately form terminal nerves innervating the lower extremity and pelvis. It comprises the lumbar plexus (L1-L4 with or without T12), sacral plexus (L4-S1, LS trunk, and S1-S4), and lower sacrococcygeal nerves (S2-S4 and C1). Anterior divisions of the lumbar plexus combine to form the iliohypogastric nerve (T12-L1), ilioinguinal nerve (L1), genitofemoral nerve (L1-L2), and obturator nerve (L2-L4). The posterior divisions of the lumbar plexus form the femoral nerve (L2-L4) and lateral femoral cutaneous nerve (L2-L3) within the psoas major muscle. Detailed description of the complicated LS plexus anatomy is beyond the scope of this chapter. The anterior divisions of the sacral plexus form the tibial part of the sciatic nerve (L4-S3), pudendal nerve (S1S4), and medial part of the posterior femoral cutaneous nerve (S1-S3). The posterior divisions of the sacral plexus form the common peroneal part of the sciatic nerve (L4-S2), superior (L4-S1) and inferior (L5-S2) gluteal nerves, lateral part of the posterior femoral cutaneous nerve (S1-S3), and piriformis muscle nerve (L5-S2).

A

C

B

D FIG 65-10  Left C7 radiculopathy. Coronal 3D STIR MIP image (A) shows asymmetric increased signal and enlargement of the left C7 nerve, which extends to the middle trunks, its divisions, and predominantly the posterior cord (B). C, Oblique axial 3D STIR MIP image shows asymmetric increased signal and enlargement of the left C7 nerve and dorsal nerve root ganglion. D, Axial T2 VISTA (volume isotropic turbo-spin echo acquisition) image shows moderate to severe left and severe right neuroforaminal narrowing.

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

A

B FIG 65-11  Left brachial plexus stretch injury. A, Coronal MPR STIR image shows an asymmetrically hyperintense enlarged left brachial plexus. B, Axial T1 image shows the resulting loss of left chest wall and shoulder girdle muscle bulk.

Common causes of LS plexopathy include nerve entrapment (although much less common than brachial plexopathy), ischemia, metabolic diseases, mass lesions, iatrogenic injury, fracture/dislocation, pregnancy, radiation, infection, vascular or muscular anomalies, and idiopathic etiologies. LS plexopathy commonly presents with asymmetric lower limb weakness, tingling, or paresthesia. Precise clinical findings depend on the location of the lesion and the territorial distribution and function of the affected nerves. EMG and NCV studies are more limited in the lower extremity because of the deep position of the nerves.25,47 General MRN findings in different disease entities like plexitis and postradiation plexopathy and metabolic diseases are basically similar to the findings mentioned in the brachial plexus section. MRN evaluation of the sciatic and femoral nerves has been described separately in the lower extremity section.

2357

FIG 65-12  Left pudendal neuropathy. Axial T2 SPAIR image shows mild increased signal of the left pudendal nerve (arrow) at the ischial spine.

ment at the level of the ischial spine and within the pudendal canal. Common causes of pudendal neuropathy include scarring from prior surgery, traction injury from pelvic instability, pelvic mass and hematoma, stretch injuries during parturition, motor vehicle accident, and bicycling. Symptoms present as sexual dysfunction and chronic perineal pain aggravated by sitting and relieved by standing or a recumbent position. On MRN the pudendal nerve can be identified at the level of the ischial spine (Fig. 65-12; also see Fig. 65-6) and pudendal canal on both T1- and T2-weighted images.25 MRN is helpful to exclude a lesion along the course of the nerve or for medication injection, by identifying the neural signal or size abnormality. Isolated involvement of the ilioinguinal nerve (L1),63 genitofemoral nerve (L1-L2),13 obturator nerve (L2-L4),49 lateral femoral cutaneous nerve (L2-L3)18 (Fig. 65-13), posterior femoral cutaneous nerve (S1S3),43 and superior and inferior gluteal nerves (L4-S1)64 has been reported in the literature. General etiologies are similar to peripheral nerves, with a few nerve-specific entrapment syndromes and mechanisms of injury. MRN can be helpful in localization of nerve lesions, regional compressions/entrapments, or preoperative mapping.

TUMORS AND TUMORLIKE LESIONS Peripheral nerve focal mass lesions include benign or malignant PNSTs, perineurioma, and primary nerve lymphoma. Besides these, peripheral nerves can be affected by tumorlike lesions such as Morton’s neuroma, posttraumatic neuroma, ganglion cysts (intraneural or extraneural), endometriosis, vascular tumors of the nerve, neuromuscular hamartoma (triton tumor), fibrolipomatous hamartoma, amyloidoma, neurothecoma, and neuritis ossificans.31 Diffuse polyneuropathies involve the nerves diffusely or at multiple different locations.

Pudendal Plexopathy

Focal Peripheral Nerve Mass Lesions

The pudendal plexus is formed from the anterior divisions of S2-S4 and C1 nerve roots and forms the pudendal nerve, perforating cutaneous nerve, anococcygeal nerve, and visceral and muscular nerves. The pudendal nerve (S2-S4) is the major branch with the most clinical importance. The pudendal nerve travels between the piriformis and coccygeus muscles and exits the pelvis through the greater sciatic foramen. It then reenters the pelvis through the lesser sciatic foramen and gives off the inferior hemorrhoidal nerve. It travels through the pudendal canal along the lateral wall of the ischiorectal fossa. Then it divides into the perineal nerve and dorsal nerve of the clitoris and penis. The pudendal nerve is vulnerable to compression and entrap-

These include benign and malignant PNST, neurolymphoma, perineurioma, lipoma, fibrolipoma, choristoma, and synovial sarcoma. Tumorlike conditions include ganglion cyst, endometrioma, and posttraumatic neuroma. The most common lesion is a nerve sheath tumor.

Peripheral Nerve Sheath Tumor.  PNSTs are most commonly benign tumors, with schwannoma and neurofibroma (Fig. 65-14) being the most common subtypes. These lesions can present with various symptoms based on the location of the involved nerve. Also, they can be part of neurocutaneous syndromes such as neurofibromatosis type 1 (NF1) or type 2 (NF2) (Fig. 65-15) and schwannomatosis.

2358

PART II  CT and MR Imaging of the Whole Body

FIG 65-13  Right lateral femoral cutaneous neuropathy. Axial T2 SPAIR (A) and DWI (B) images show mild asymmetrically increased signal in the right lateral femoral cutaneous nerve (arrows).

Schwannoma (neurinoma or neurilemoma), the most common benign PNST, is a slow-growing peripheral nerve tumor arising from the Schwann cells of the nerve sheaths.52 It usually presents as a solitary nerve tumor in individuals between 20 and 60 years old. Bilateral vestibular schwannomas are the hallmark of NF2. Plexiform schwannomas may also occur in NF2 or schwannomatosis. Schwannomas are well-circumscribed masses with nerve fascicles stretched over the lesion. Cystic formation, hemorrhage, or calcification can be seen in about 10% of cases. Neurofibroma is the second most common benign tumor of the peripheral nerve. Neurofibroma is 90% solitary and not by itself a criterion for the diagnosis of NF unless it is plexiform. They usually arise from small peripheral nerves and have superficial locations. These lesions usually manifest as painless slow-growing masses. Malignant transformation is rare in solitary forms and more common in NF1. Neurofibroma is composed of different cell types (Schwann cell, perineural cell, fibroblast, lymphocytes, and mast cells) and cannot be easily separated from the normal nerve fibers. Malignant PNSTs may be sporadic or arise from neurofibromas, especially following radiation therapy. Malignant PNSTs associated with NF1 are usually larger, present earlier, and have male predominance. Patients usually present with a firm mass associated with neurologic deficit and pain. NF1 with malignant PNST has the worst prognosis. MRN can be helpful to localize the lesion, determine whether it is intraneural or extraneural, and evaluate for nerve encasement, displacement, or impingement in cases of extrinsic masses. MRN is also helpful for differentiating nerve tumors from other possible soft tissue masses. It is challenging to differentiate schwannoma from neurofibroma based on imaging. However, neurofibroma tends to involve more fascicles than schwannoma. Other than secondary signs of nerve injury such as regional muscle denervation changes, neural tumors have some characteristic signs that can help differentiate them from extraneural masses1,29,78: Target sign: central fibrous/collagenous tissue of the tumor surrounded by peripheral myxoid stroma. On T1-weighted images, central hyperintensity with peripheral isointense lesion and also contrast enhancement are seen. On T2-weighted images, central hypointensity with peripheral homogeneous hyperintensity is seen. Split-fat sign: intramuscular fat capping the edges of the lesion Fascicular sign: prominent dots with peripheral hyperintensity on T2-weighted images and contrast enhancement on T1-weighted images Dumbbell sign: neoplasms of juxtaforaminal location adopt a dumbbelllike shape with a waist at the level of the foramen.

Smooth scalloping of the surrounding osseous structures due to the slow-growing nature of the tumor It is difficult by imaging to differentiate benign from malignant PNSTs. A diameter of more than 5 cm, peripheral gadolinium enhancement, peripheral edema-like ring, intralesional cyst, necrosis, or hemorrhage, and low ADC values on DWI favor a malignant nature of the lesion.1 MRN is also useful in postsurgical follow-up (Fig. 65-16). Treatment is typically sarcoma protocol surgical resection and oncologic treatment for malignant lesions, with attempts to preserve the nerve function as much as possible.

Diffuse Peripheral Nerve Lesions Diffuse peripheral nerve diseases have an extensive differential diagnosis and are almost always a diagnostic dilemma on MRI. Hence clinical findings, information regarding extension of disease, and (finally) nerve biopsy can be used as adjunctive information for more precise diagnosis. Whole-body MRN with contrast can be helpful for better diagnosis. The common etiologies include neurocutaneous syndromes, hereditary neuropathy, acute and chronic inflammatory demyelinating polyneuropathy (Fig. 65-17), hypertrophic motor and sensory neuropathies, diabetic neuropathy, radiation neuropathy, multifocal motor neuropathy (Fig. 65-18), amyloid neuropathy, and infections such as leprosy.

MRN TECHNICAL AND INTERPRETATIONAL PITFALLS Like other diagnostic modalities, MRN is also affected by some technical pitfalls and limitations. MRN is a technically demanding imaging study. Although sensitive, its specificity is questionable at times. MRI signal alterations cannot differentiate between mechanical, immune, or ischemic nerve injuries, and correlation with history and associated signs and symptoms is always recommended. Also, there is not a strong correlation between degree of T2 hyperintensity and degree of nerve injury; secondary signs of nerve injury (e.g., fascicular morphology, nerve enlargement, abnormal coarse, nerve discontinuity, neuroma formation, regional muscle changes) should be sought. Abnormal T2 signal in the setting of acute nerve injury can persist even after full functional recovery of the nerve; on the other hand, T2 signal intensity may become normal in chronically degenerated nerves despite persistent functional loss. Therefore T2-based MRN may not be able to differentiate regenerating nerves from chronically degenerating nerves based on signal intensity changes. DWI may play an

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

A

B

C

D FIG 65-14  Plexiform neurofibroma. Axial T2 SPAIR (A) and coronal T2 GRE PSIF (B) images show a hyperintense mass in the lateral ankle subcutaneous tissues. Axial T1 precontrast (C) and axial T1 fat-saturated postcontrast (D) images show the avid enhancement of the mass. Additional note is made of mild hyperintense signal in the sural nerve.

2359

2360

PART II  CT and MR Imaging of the Whole Body

A

B

C FIG 65-15  Neurofibromatosis with sciatic peripheral nerve sheath tumors. A and B, Coronal T2 GRE PSIF MIP images demonstrate innumerable neurofibromas, some involving the sciatic nerve (short arrows) and a dominant tumor in the vastus medialis (long arrow). C, Axial T2 SPAIR, T1, and ADC map images at the level of the dominant tumor (long arrows) show T2 hyperintensity with restricted diffusion. The ADC values in the tumor measured 1.2-2.0 × 10−3 mm2/sec, suggesting a cellular neurofibroma. Short arrows demonstrate the innumerable neurofibromas within the sciatic nerve at this level.

important role in evaluating the functional recovery of the nerves in the near future. When technical failure occurs, 3D sequences are usually more affected than 2D sequences. The common causes of technical failure include poor shimming, patient motion, field inhomogeneity, and susceptibility from nearby metal. In the case of field homogeneity, suboptimal fat saturation, or nearby metal susceptibility, a STIR sequence may provide better images. A 3D DW-PSIF sequence gives better images for the smaller peripheral nerves of the extremity, especially when high-resolution surface coils or multichannel joint-specific coils can be used. It usually provides suboptimal images in a large FOV (>15 cm), especially near the axial body, because of motion and breathing artifacts. DWI is vulnerable to ghosting and wraparound

artifacts, and any optimization should be performed with respect to the specificities of the individual MRI scanner. The nerve may be T2 hyperintense if traveling at 55 degrees to the main magnetic field, owing to magic-angle artifact. Therefore imaged anatomy should be scanned and placed in the field in a way to minimize this angulation. Also, confirmation in multiple planes, use of higher TE (>60 milliseconds), and attention to other findings such as nerve size, course, fascicular morphology, and regional skeletal muscle changes are helpful for more accurate interpretation. Intraneural prominent vessels may mimic enlarged fascicles. Assessment of both T1 and T2 images and branching or short-segment appearance of this intensity may help in differentiating vessels from fascicles. On 3D SPAIR SPACE and STIR images, nerves are usually

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography

2361

A

B

C FIG 65-16  Backpressure neuropathy and demyelination change following neurofibroma resection. A, Oblique axial and sagittal T2 fat-saturation images demonstrate a left paramedian path-proven neurofibroma intimately involving the left S1 and S2 neuroforamen. Coronal 3D SHINKEI SENSE MIP image (B) and axial T2 SPAIR image (C) obtained following resection show hyperintensity and prominence of the left S1 and S2 nerve roots extending from the sacral foramen to minimally involving the sciatic plexus (arrows).

hyperintense because of increased fluid sensitivity, good fat suppression, and increased dynamic range of contrast. Special attention to focal enlargement or flattening, comparison with contralateral or ipsilateral normal nerves, review of axial images, and correlation  with the patient’s symptomatology can be helpful for more precise evaluation. Nerves are usually hyperintense for a significant distance proximal and distal to the site of injury; this is due to proximal axoplasmic and venous flow congestion and distal wallerian degeneration, respectively. To guide the surgeon or treating physician to the appropriate site of injury, readers should keep in mind that the abnormality is most prominent at the site of entrapment or just proximally.16,22,30

CONCLUSION FIG 65-17  Inflammatory neuropathy, such as chronic inflammatory demyelinating polyneuropathy. Coronal TSE 3D STIR SPACE (Sampling Perfection with Application optimized Contrasts using different flip angle Evolutions) MIP image shows increased size and signal of bilateral C8 and T1 nerve roots extending into the brachial plexus trunks.

MRN is a helpful diagnostic tool for accurate diagnosis, follow-up, and preoperative mapping for focal and diffuse neuropathies. Knowledge of the normal and abnormal appearances of the peripheral nerves on MRN, familiarity with its technical pitfalls, and communication with the referring physician will lead to optimal imaging evaluation of the peripheral nerves.

2362

PART II  CT and MR Imaging of the Whole Body

A

B

FIG 65-18  Multifocal motor neuropathy. A, Axial T2 SPAIR image shows median and radial nerve increased signal (upper and lower arrows, respectively) at the level of the midhumerus. B, Oblique coronal PSIF MIP image shows the hyperintense median nerve (short arrow) and radial nerve (long arrow).

REFERENCES 1. Ahlawat S, Chhabra A, Blakely J: Magnetic resonance neurography of peripheral nerve tumors and tumorlike conditions. Neuroimaging Clin N Am 24(1):171–192, 2014. 2. Andreisek G, Crook DW, Burg D, et al: Peripheral neuropathies of the median, radial, and ulnar nerves: MR imaging features. Radiographics 26(5):1267–1287, 2006. 3. Andreisek G, White LM, Kassner A, et al: Diffusion tensor imaging and fiber tractography of the median nerve at 1.5T: Optimization of b value. Skeletal Radiol 38(1):51–59, 2009. 4. Aralasmak A, Karaali K, Cevikol C, et al: MR imaging findings in brachial plexopathy with thoracic outlet syndrome. AJNR Am J Neuroradiol 31(3):410–417, 2010. 5. Aydin S, Abuzayed B, Bozkus H, et al: Posttraumatic brachial plexitis.   J Trauma 71(6):E136, 2011. 6. Bäumer P, Dombert T, Staub F, et al: Ulnar neuropathy at the elbow: MR neurography-nerve T2 signal increase and caliber. Radiology 260(1):199– 206, 2011. 7. Baumer P, Kele H, Kretschmer T, et al: Thoracic outlet syndrome in 3T MR neurography-fibrous bands causing discernible lesions of the lower brachial plexus. Eur Radiol 24(3):756–761, 2014. 8. Beltran LS, Bencardino J, Ghazikhanian V, et al: Entrapment neuropathies III: Lower limb. Semin Musculoskelet Radiol 14(5):501–511, 2010. 9. Bencardino JT, Rosenberg ZS: Entrapment neuropathies of the   shoulder and elbow in the athlete. Clin Sports Med 25(3):465–487, 2006. 10. Bendszus M, Koltzenburg M, Wessig C, et al: Sequential MR imaging of denervated muscle: Experimental study. AJNR Am J Neuroradiol 23(8):1427–1431, 2002. 11. Campagna R, Pessis E, Feydy A, et al: MRI assessment of recurrent carpal tunnel syndrome after open surgical release of the median nerve. AJR Am J Roentgenol 193(3):644–650, 2009. 12. Caranci F, Briganti F, La Porta M, et al: Magnetic resonance imaging   in brachial plexus injury. Musculoskelet Surg 97(Suppl 2):S181–S190, 2013.

13. Cardosi RJ, Cox CS, Hoffman MS: Postoperative neuropathies after major pelvic surgery. Obstet Gynecol 100(2):240–244, 2002. 14. Chalian M, Faridian-Aragh N, Soldatos T, et al: High-resolution 3T MR neurography of suprascapular neuropathy. Acad Radiol 18(8):1049–1059, 2011. 15. Chalian M, Soldatos T, Faridian-Aragh N, et al: 3T magnetic resonance neurography of tibial nerve pathologies. J Neuroimaging 23(2):296–310, 2013. 16. Chhabra A, Batra K, Andreisek G: Magnetic resonance neurography interpretation. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 23–36. 17. Chhabra A, Chalian M, Andreisek G: Brachial plexus. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 133–160. 18. Chhabra A, Chalian M, Andreisek G: Magnetic resonance neurography of tunnels-Part II: Lower extremity nerves. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 73–111. 19. Chhabra A, Chalian M, Soldatos T, et al: 3-T high-resolution MR neurography of sciatic neuropathy. AJR Am J Roentgenol 198(4):W357– W364, 2012. 20. Chhabra A, Del Grande F, Soldatos T, et al: Meralgia paresthetica:   3-Tesla magnetic resonance neurography. Skeletal Radiol 42(6):803–808, 2013. 21. Chhabra A, Faridian-Aragh N, Chalian M, et al: High-resolution 3-T MR neurography of peroneal neuropathy. Skeletal Radiol 41(3):257–271, 2012. 22. Chhabra A, Flammang A, Andreisek G: Magnetic resonance neurography technique. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 10–22. 23. Chhabra A, Lee PP, Bizzell C, et al: 3 Tesla MR neurography-technique, interpretation, and pitfalls. Skeletal Radiol 40(10):1249–1260, 2011. 24. Chhabra A, Lee PP, Bizzell C, et al: High-resolution 3-tesla magnetic resonance neurography of musculocutaneous neuropathy. J Shoulder Elbow Surg 21(2):e1–e6, 2012.

CHAPTER 65  High-Resolution 3T Magnetic Resonance Neurography 25. Chhabra A, Soldatis T, Andreisek G: Lumbosacral plexus. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 161–181. 26. Chhabra A, Subhawong TK, Andreisek G: Magnetic resonance neurography of tunnels-Part I: Upper extremity nerves. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 37–72. 27. Chhabra A, Subhawong TK, et al: 3T MR neurography using threedimensional diffusion-weighted PSIF: Technical issues and advantages. Skeletal Radiol 40(10):1355–1360, 2011. 28. Chhabra A, Subhawong TK, Williams EH, et al: High-resolution MR neurography: Evaluation before repeat tarsal tunnel surgery. AJR Am J Roentgenol 197(1):175–183, 2011. 29. Chhabra A, Thakkar RS, Andreisek G, et al: Anatomic MR imaging   and functional diffusion tensor imaging of peripheral nerve tumors   and tumorlike conditions. AJNR Am J Neuroradiol 34(4):802–807, 2013. 30. Chhabra A, Thawait SK, Andreisek G: Nerve anatomy, pathophysiology, nerve injury and entrapment. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 1–9. 31. Chhabra A, Thawait GK, Andreisek G: Peripheral nerve tumor and tumor-like conditions. In Chhabra A, Andreisek G, editors: Magnetic Resonance Neurography, New Delhi, 2012, Jaypee Brothers Medical Publisher, pp 113–132. 32. Chhabra A, Thawait GK, Soldatos T, et al: High-resolution 3T MR neurography of the brachial plexus and its branches, with emphasis on 3D imaging. AJNR Am J Neuroradiol 34(3):486–497, 2013. 33. Chhabra A, Williams EH, Subhawong TK, et al: MR neurography findings of soleal sling entrapment. AJR Am J Roentgenol 196(3):W290– W297, 2011. 34. Chhabra A, Williams EH, Wang KC, et al: MR neurography of neuromas related to nerve injury and entrapment with surgical correlation. AJNR Am J Neuroradiol 31(8):1363–1368, 2010. 35. Donovan A, Rosenberg ZS, Cavalcanti CF: MR imaging of entrapment neuropathies of the lower extremity. Part 2. The knee, leg, ankle, and foot. Radiographics 30(4):1001–1019, 2010. 36. Espinosa N, Schmitt JW, Saupe N, et al: Morton neuroma: MR imaging after resection–Postoperative MR and histologic findings in asymptomatic and symptomatic intermetatarsal spaces. Radiology 255(3):850–856, 2010. 37. Faridian-Aragh N, Chalian M, Soldatos T, et al: High-resolution 3T   MR neurography of radial neuropathy. J Neuroradiol 38(5):265–274, 2011. 38. Filler AG, Haynes J, Jordan SE, et al: Sciatica of nondisc origin and piriformis syndrome: Diagnosis by magnetic resonance neurography and interventional magnetic resonance imaging with outcome study of resulting treatment. J Neurosurg Spine 2(2):99–115, 2005. 39. Filler AG, Kliot M, Howe FA, et al: Application of magnetic resonance neurography in the evaluation of patients with peripheral nerve pathology. J Neurosurg 85(2):299–309, 1996. 40. Filler AG, Maravilla KR, Tsuruda JS: MR neurography and muscle MR imaging for image diagnosis of disorders affecting the peripheral nerves and musculature. Neurol Clin 22(3):643–682, 2004. 41. Flak M, Durmala J, Czernicki K, et al: Double crush syndrome evaluation in the median nerve in clinical, radiological and electrophysiological examination. Stud Health Technol Inform 123:435– 441, 2006. 42. Flanigan RM, DiGiovanni BF: Peripheral nerve entrapments of the lower leg, ankle, and foot. Foot Ankle Clin 16(2):255–274, 2011. 43. Fritz J, Bizzell C, Kathuria S, et al: High-resolution magnetic resonanceguided posterior femoral cutaneous nerve blocks. Skeletal Radiol 42(4):579–586, 2013. 44. Griffin AC, Wood WG: Brachial plexitis: A rare and often misdiagnosed postoperative complication. Aesthetic Plast Surg 20(3):263–265, 1996. 45. Gu B, Yang Z, Huang S, et al: Radiation-induced brachial plexus injury after radiotherapy for nasopharyngeal carcinoma. Jpn J Clin Oncol 44(8):736–742, 2014.

2363

46. Guggenberger R, Nanz D, Bussmann L, et al: Diffusion tensor   imaging of the median nerve at 3.0 T using different MR scanners: Agreement of FA and ADC measurements. Eur J Radiol 82(10):590–596, 2013. 47. Höllinger P, Sturzenegger M: Chronic progressive primary lumbosacral plexus neuritis: MRI findings and response to immunoglobulin therapy. J Neurol 247(2):143–145, 2000. 48. Howe FA, Filler AG, Bell BA, et al: Magnetic resonance neurography. Magn Reson Med 28(2):328–338, 1992. 49. Husarik DB, Saupe N, Pfirrmann CW, et al: Elbow nerves: MR findings in 60 asymptomatic subjects–Normal anatomy, variants, and pitfalls. Radiology 252(1):148–156, 2009. 50. Jarvik JG, Comstock BA, Heagerty PJ, et al: Magnetic resonance imaging compared with electrodiagnostic studies in patients with suspected carpal tunnel syndrome: Predicting symptoms, function, and surgical benefit at 1 year. J Neurosurg 108(3):541–550, 2008. 51. Kabakci NT, Kovanlikaya A, Kovanlikaya I: Tractography of the median nerve. Semin Musculoskelet Radiol 13(1):18–23, 2009. 52. Kanta M, Petera J, Ehler E, et al: Malignant schwannoma of the obturator nerve. Bratisl Lek Listy 114(10):584–586, 2013. 53. Khalil C, Budzik JF, Kermarrec E, et al: Tractography of peripheral nerves and skeletal muscles. Eur J Radiol 76(3):391–397, 2010. 54. Kim S, Choi JY, Huh YM, et al: Role of magnetic resonance imaging   in entrapment and compressive neuropathy–What, where, and how to see the peripheral nerves on the musculoskeletal magnetic resonance image: Part 1. Overview and lower extremity. Eur Radiol 17(1):139–149, 2007. 55. Kim S, Choi JY, Huh YM, et al: Role of magnetic resonance imaging in entrapment and compressive neuropathy–What, where, and how to see the peripheral nerves on the musculoskeletal magnetic resonance image: Part 2. Upper extremity. Eur Radiol 17(2):509–522, 2007. 56. Kim JY, Ihn YK, Kim JS, et al: Non-traumatic peroneal nerve palsy: MRI findings. Clin Radiol 62(1):58–64, 2007. 57. Kinni V, Craig J, van Holsbeeck M, et al: Entrapment of the posterior interosseous nerve at the arcade of Frohse with sonographic, magnetic resonance imaging, and intraoperative confirmation. J Ultrasound Med 28(6):807–812, 2009. 58. Klauser AS, Faschingbauer R, Bauer T, et al: Entrapment neuropathies II: Carpal tunnel syndrome. Semin Musculoskelet Radiol 14(5):487–500, 2010. 59. Lareau CR, Sawyer GA, Wang JH, et al: Plantar and medial heel pain: Diagnosis and management. J Am Acad Orthop Surg 22(6):372–380, 2014. 60. Lee PP, Chalian M, Bizzell C, et al: Magnetic resonance neurography of common peroneal (fibular) neuropathy. J Comput Assist Tomogr 36(4):455–461, 2012. 61. Le Nail LR, Bacle G, Marteau E, et al: Isolated paralysis of the serratus anterior muscle: Surgical release of the distal segment of the long thoracic nerve in 52 patients. Orthop Traumatol Surg Res 100(4Suppl):S243–S248, 2014. 62. Linklater JM: Imaging of sports injuries in the foot. AJR Am J Roentgenol 199(3):500–508, 2012. 63. Lucini G, Mumenthaler M, Mumenthaler A, et al: Syndrome of the ilioinguinal nerve. Minerva Ortop 18(8):476–481, 1967. 64. Martinoli C, Miguel-Perez M, Padua L, et al: Imaging of neuropathies about the hip. Eur J Radiol 82(1):17–26, 2013. 65. Miller TT, Reinus WR: Nerve entrapment syndromes of the elbow, forearm, and wrist. AJR Am J Roentgenol 195(3):585–594, 2010. 66. Mullins GM, O’Sullivan SS, Neligan A, et al: Non-traumatic brachial plexopathies, clinical, radiological and neurophysiological findings from a tertiary centre. Clin Neurol Neurosurg 109(8):661–666, 2007. 67. Petchprapa CN, Rosenberg ZS, Sconfienza LM, et al: MR imaging of entrapment neuropathies of the lower extremity. Part 1. The pelvis and hip. Radiographics 30(4):983–1000, 2010. 68. Rask MR: Medial plantar neurapraxia (jogger’s foot): Report of 3 cases. Clin Orthop Relat Res 134:193–195, 1978. 69. Rehman I, Chokshi FH, Khosa F: MR Imaging of the brachial plexus. Clin Neuroradiol 24(3):207–216, 2014.

2364

PART II  CT and MR Imaging of the Whole Body

70. Rinkel WD, Schreuders TA, Koes BW, et al: Current evidence for effectiveness of interventions for cubital tunnel syndrome, radial tunnel syndrome, instability, or bursitis of the elbow: A systematic review. Clin J Pain 29(12):1087–1096, 2013. 71. Seddon HJ: A classification of nerve injuries. Br Med J 2(4260):237–239, 1942. 72. Still GP, Fowler MB: Joplin’s neuroma or compression neuropathy of the plantar proper digital nerve to the hallux: Clinicopathologic study of three cases. J Foot Ankle Surg 37(6):524–530, 1998. 73. Subhawong TK, Wang KC, Thawait SK, et al: High resolution imaging of tunnels by magnetic resonance neurography. Skeletal Radiol 41(1):15–31, 2012. 74. Sunderland S, Bradeley KC: The cross-sectional area of peripheral nerve trunks devoted to nerve fibers. Brain 72(3):428–449, 1949.

75. Sureka J, Cherian RA, Alexander M, et al: MRI of brachial plexopathies. Clin Radiol 64(2):208–218, 2009. 76. Tung GA, Entzian D, Stern JB, et al: MR imaging and MR arthrography of paraglenoid labral cysts. AJR Am J Roentgenol 174(6):1707–1715, 2000. 77. Vucic S, Cordato DJ, Yiannikas C, et al: Utility of magnetic resonance imaging in diagnosing ulnar neuropathy at the elbow. Clin Neurophysiol 117(3):590–595, 2006. 78. Wadhwa V, Lee PP, Strome GM, et al: Spectrum of superficial nerverelated tumor and tumor-like lesions: MRI features. Acta Radiol 55(3):345–358, 2014. 79. Weishaupt D, Andreisek G: Diagnostic imaging of nerve compression syndrome. Radiologe 47(3):231–239, 2007.

66  MRI-Guided Interventions Sherif G. Nour

INTRODUCTION The term interventional MRI is used to describe the use of magnetic resonance imaging (MRI) for rapid guidance and/or monitoring of a minimally invasive diagnosis or therapy where the entire procedure  is interactively performed within an interventional MRI suite in a manner that is equivalent to standard fluoroscopic or ultrasound (US)-guided procedures. The term interventional MRI itself is relatively new as a description of a full clinical service of MRI-guided interventions integrated within the daily practice of interventional radiology. The idea of adapting MRI for interventional procedure guidance, however, dates back to the mid-1980s shortly after the introduction of MRI as a diagnostic imaging modality. Continued growth of the field of interventional MRI (iMRI) is marked by the increasing number of academic institutions with added iMRI capabilities, the growing diversity of iMRI applications adopted at these institutions, and the expanding technologic innovations that back this growth.

iMRI SUITE SETUP An iMRI suite is an interventional procedure room equivalent to other medical procedure rooms where the main interventional equipment may consist of an x-ray fluoroscopy unit, a computed tomography (CT) scanner, a US scanner, an endoscopy unit, and so forth. In this case the room activity revolves around an MRI scanner. This arrangement is a hybrid design that has to fulfill the requirements of both a regular diagnostic MRI scanner and an interventional radiology suite. Specifically the fundamental requirements for an iMRI suite include: (1) Adequate access to the patient within the scanner to perform an interventional procedure. Access to the patient is restricted within the regular narrow, long bores of standard scanners. Interventional solutions include use of an open-configuration scanner or adopting other workflow solutions as described in this chapter. (2) Availability of MRI-compatible anesthesia equipment, monitoring tools, and interventional devices. These are currently available from several vendors on a commercial basis. (3) Ability to achieve proper draping and to ensure a completely sterile field during interventions. This is a more challenging job compared to the customary draping of flat surfaces during traditional x-ray fluoroscopy, CT, or US-guided interventions. (4) Ability to implement rapid pulse sequences with reasonable signalto-noise ratio to ensure adequate visualization of the interventional device and timely updates of its location to achieve safe device placement. (5) Ability to operate the scanner and to review updated images at the patient’s bedside without having to remove the operator’s hand at any time from the interventional setting.

2366

iMRI SYSTEM DESIGN AND WORKFLOW SOLUTIONS Designing an MRI scanner for interventional purposes is a technically and logistically complex endeavor. On the technical side, the environment suitable for adequate access to the patient and comfortable application of anesthesia equipment and monitoring devices is basically the opposite of that “closed magnet” environment suitable for producing high-quality diagnostic MRI scans. Logistically there will probably be a need—at least for the foreseeable future—to use an iMRI scanner for routine diagnostic imaging to maximize the revenue of the scanner and support the associated interventional programs. Several interventional solutions have been implemented over the past years that included both magnet design and workflow/combined-suite solutions (Fig. 66-1).

Magnet Design Solutions The one recognized limitation the leaders of the field of interventional MRI agree on is that the ultimate interventional MRI scanner is yet to be built. The General Electric Medical Systems magnet with the doubledoughnut design (Signa SP system, Milwaukee, Wisconsin, USA) was the world’s first MRI scanner designed and produced for the sole purpose of MRI-guided interventions.79,85,86 This design is based on a 0.5-tesla (T) superconducting magnet where the central segment has been removed from the cylindrical system, leaving two “doughnuts,” one on each side, allowing access to the patient from the sides and top at the isocenter of the imaging system.28 The system was first installed at the Brigham and Women’s Hospital in Boston in the mid-1990s and promised a prosperous future for this field. Unfortunately that era of development was not the prime time for a dedicated iMRI scanner, and a sufficient clinical practice was never generated to support commercialization of this scanner. In fact the last one of these scanners operating today is currently installed at Shiga University of Medical Science, Otsu, Shiga, Japan, awaiting the formal label of a “historic item.” The rise and fall of the double-doughnut system has managed to keep GE furthest away from iMRI applications, with other vendors stepping in to offer some interventional solutions, albeit modestly. With all lessons learned, these solutions have been centered on constructing scanners with more open configurations to allow better patient access for interventions while maintaining a platform that is fully operational for standard diagnostic imaging. In-room radiofrequency-shielded liquid crystal display (LCD) monitors have also been made available to immediately review reconstructed images at the scanner side and operate the scanner from within the room. Vendors also have provided some software solutions for device guidance (e.g., the Integrated Front End [IFE] program from Siemens) and treatment monitoring. Vendors are, however, still reluctant to reengage in designing a full-scale interventional scanner or to respond to the need for more involved

CHAPTER 66  MRI-Guided Interventions

GE Signa SP 0.5 T

Magnetom Open 0.2 T

Philips Panorama 1.0 T

Siemens Espree 1.5T

Hybrid MRI / Fluorosuites

IMRIS

2367

AMIGO Suite

FIG 66-1  Various magnet design and workflow/combined-suite solutions for iMRI.

hardware modifications geared toward satisfying interventional needs. In the current iMRI practice, magnet design solutions are centered on the use of open-configuration cylindrical superconducting systems and biplanar lower-field-strength open systems designed to accommodate claustrophobic and obese patients and to allow MRI-guided interventions.

Workflow/Combined-Suite Solutions This group represents a growing sector of iMRI solutions and is more suitable when the intervention is not intended to be solely conducted within the MRI environment but rather when MRI is needed as a part of a procedure, such as to obtain updated information on organ shift, confirm extent of a resection, or monitor effect of therapy or drug delivery. The primary concept in these solutions is related to meeting space requirements and installation specifications and cost. Interactive MRI-guided device placement is usually not feasible with this setup. Examples of these workflow/combined-suite solutions follow.

Hybrid MRI/Fluoroscopy Suites.  In this arrangement a conventional x-ray fluoroscopy unit is installed within the same MRI suite. Patients are positioned on the same procedure table throughout the entire intervention while the table is moved between the two systems on a track or wheels. The value of these suites is highlighted during cardiovascular catheter intervention procedures where vascular access and catheter navigation is still more straightforward under standard fluoroscopic guidance, whereas MRI provides state-of-the-art additions to these procedures, such as plaque burden imaging and characterization within the vessel wall, imaging perivascular soft tissues for procedure complications (e.g., infracted territory, hemorrhage, etc.), or accurately locating catheter tips in relation to cardiac chambers and assessing the effect of radiofrequency ablation (RFA) during electrophysiology procedures.

IMRIS System.  This system is primarily geared toward facilitating intraoperative imaging and has become a popular component of modern neurosurgery suites in recent years. Surgeons operate using their standard operating table and operative techniques. When imaging is needed during the course of surgery to quantify the extent of brain shift following dural incision or to evaluate the extent of resection, the MRI unit is moved on giant ceiling-mounted tracks to cover the patient’s head.

MedTrack System.  Instead of designing special scanners or scanning tables, the MedTrack system consists of a large track mounted to the ceiling across several lined rooms (e.g., operating room, MRI scanner, and a positron emission tomography [PET] scanner). Attached to this ceiling track is a table with room for both the patient and anesthesia equipment to facilitate moving both as one unit, thereby eliminating the usual complexity related to handling general anesthesia in these situations.

Advanced Multimodality Image-Guided Operating (AMIGO) Suite.  This is the most recent addition to workflow solutions. The suite is a three-room design aiming at offering a state-of-the-art operating room with access to advanced imaging technology to meet any intraoperative imaging need. The center of the AMIGO suite houses an operating room that features an electronically controlled operating table  and advanced mobile imaging devices, including surgical navigation systems, C-arm x-ray, three dimensional (3D) US, and a near-infrared imaging system. The second room features a ceiling-mounted MRI  unit that can be moved into the operating room on demand (IMRIS system). The third room is a PET-CT scanner room that is contiguous with the operating room on the other side. The patient is transferred during surgery to the PET-CT room to obtain metabolic information as needed to optimize surgery (e.g., during tumor resections).

2368

PART III  Image-Guided Procedures

APPLICATIONS OF iMRI MRI may be used to guide procedures for a solely diagnostic purpose, such as performing a percutaneous biopsy or aspiration procedure, and may be used for more complex therapeutic tasks where MRI is used to guide the placement of an interventional device and to monitor the deployment of therapy or the progress of surgery.

MRI-Guided Biopsies and Aspirations The traditional use of US and CT to guide biopsies is expected to remain the primary approach for standard interventions, owing to wide availability, ease of use, and lower cost. Significant advances in diagnostic MRI techniques in recent years has resulted in markedly increased use of MRI as a first-line diagnostic modality for many indications, such as evaluation for liver metastases in the setting of tumor workup. This has led to an increase in the number of indeterminate lesions that cannot otherwise be seen on other imaging modalities, creating a need for similarly improved visualization during intervention to allow successful tissue sampling.

Head, Neck, and Spine Biopsies.  Much of the early literature reporting the use of MRI to guide percutaneous biopsies comprised reports on the feasibility and safety of head and neck biopsies within low-field biplanar46 and midfield double-donut20 MRI scanners. The use of MRI for guidance of biopsies and aspirations in the head and neck was found to be most advantageous for sampling suprahyoid46 and retropharyngeal39,48 neck lesions and other masses within the deep head and neck spaces.16,25,53,77,99 The use of rapid gradient echo T2-weighted sequences was found most useful for biopsies and aspiration procedures within the spinal canal, because they can readily distinguish cerebrospinal fluid or cyst contents and the adjacent spinal cord, thereby ensuring safe placement of biopsy devices.20,45,46

A

B

Interactive near-real-time MRI guidance of intracranial fine-needle aspirations and core biopsies have been reported using openconfiguration MRI systems, including direct placement into brain tumors3,18,35,40,42,43,82 and transnasal and transsphenoidal MRI-guided tissue sampling of petroclival lesions.84 The feasibility, safety, and efficacy of intracranial cyst and abscess aspirations were also reported as viable applications of iMRI.5,34 The interactive approach of MRI guidance (as opposed to intermittent intraprocedural imaging) is favored during those intracranial procedures to facilitate instantaneous online monitoring of the extent of cyst or abscess decompression and of any potential associated complications such as hemorrhage or brain edema. A significantly favorable neurologic outcome was reported following MRI-guided evacuation of hypertensive hematomas in the basal ganglia.4

Abdominal and Pelvic Biopsies.  With the exception of prostate biopsies, abdominal and pelvic biopsies are typically not a primary field for MRI-guided biopsies. There is a long history of radiologic expertise using x-ray fluoroscopy, US, and CT to guide interventions within the abdomen and pelvis. These imaging modalities are clearly inferior to MRI when imaging the central nervous system and the head and neck, a factor that explains the early interest in these areas for MRI-guided interventions. The same logic has led to the late application of iMRI for abdominal and pelvic interventions. The literature does, however, report use of MRI guidance for abdominal and pelvic procedures. These are mainly biopsy procedures performed to sample lesions that are not visualized on, or difficult to access under (Fig. 66-2), US or CT scans as an alternative to surgical biopsy.45,46 The significant improvement in diagnostic MRI techniques in the abdomen in recent years has led to a sharp spike in use of abdominal MRI as a first-line diagnostic modality in several instances, including assessment for hepatic metastases in the setting of tumor workup. This has led  to an increase in the number of indeterminate lesions that would

C

FIG 66-2  MRI-guided biopsy in a 50-year-old man with retroperitoneal fibrosis. A retroperitoneal soft tissue mantle (arrowheads) is seen encasing the aorta (black arrow) and inferior vena cava (IVC) (white arrow) and entrapping the dilated right ureter (curved arrow) at the level of the L4 vertebral body. Renal function was impaired, precluding intravenous administration of iodinated contrast material. MRI-guided biopsy was performed using true-FISP image guidance. The flow-enhancement phenomenon inherent in this GRE sequence facilitated a clear delineation of the aorta and IVC throughout the procedure, without the need for contrast administration. Additionally the T2 weighting provided by this sequence helped visualization of the dilated ureter as a rather hyperintense structure. A, The biopsy needle is introduced through a transpsoas approach. B, The needle is safely advanced between the IVC and the dilated right ureter. C, The cutting notch is deployed further into the retroperitoneal abnormal tissues anterolateral to the abdominal aorta. (From Nour SG, Lewin JS: Percutaneous biopsy from blinded to MR guided: An update on current techniques and applications. Magn Reson Imaging Clin N Am 13:441-464, 2005, with permission.)

CHAPTER 66  MRI-Guided Interventions

A

B

2369

C I : Interroll

Cho I : 4.73

CI I : 3.02

0.2

Cr I : 0.133

? 50 pixels

0.0

–0.2 ppm

D

E

4.0

F

3.5

3.0

2.5

2.0

1.5

FIG 66-3  Multiparametric prostate MRI of a cancer-suspicious area was proven to be adenocarcinoma after MRI-guided biopsy. Axial (A), coronal (B), and sagittal (C) high-resolution T2-weighted images. Axial diffusion weighted, b = 2000 (D) and ADC map (E). 1H-spectroscopy demonstrating elevated choline/creatine-to-citrate ratio (F). (From Nguyen ML et al; International Geriatric Radiotherapy Group. The potential role of magnetic resonance spectroscopy in image-guided radiotherapy. Front Oncol 4:91, 2014, with permission.)

otherwise not be seen on other imaging modalities. The wide range of different image contrast mechanisms obtainable by MRI offers iMRI a unique capacity to target these lesions for percutaneous biopsy and/or locoregional treatment.9,37,102 MRI guidance has also proven helpful to guide biopsies of transiently enhancing lesions and lesions within the dome of the liver,80 in the subphrenic regions,1 in the pancreas,32 and in the adrenal glands.19,36 The use of MRI to guide prostate gland biopsies is currently a hot topic in the application of body iMRI. Transrectal,6 transgluteal,101 and transperineal38,52,89,90 approaches have been described for MRI-guided prostate biopsies. The current standard of care to establish the diagnosis in patients with suspected prostate cancer owing to an elevated prostate-specific antigen (PSA) level and/or the presence of a palpable nodule is to obtain a systematic US-guided biopsy that randomly samples various parts of the gland. The initial transrectal US (TRUS) biopsy result is typically negative in 65% to 75% of cases of increased PSA, and this negative rate for cancer detection is only increased at  the second TRUS biopsy attempt to 81% to 83%.73,88 In patients with repeated negative TRUS biopsies and continued rise of PSA level, a saturation biopsy may be pursued; however, cancer detection rates of saturation biopsy protocols including 20 to 38 cores have been reported to be as low as 14% to 41% without significantly increasing the likelihood of detecting clinically significant cancers.78 Much of the missed cancer diagnoses during TRUS biopsies are related to anterior gland and apical foci of cancer, particularly in the presence of a large gland.

A more important scenario than missing the cancerous focus on a TRUS biopsy is to sample a low-grade cancer and miss a higher-grade focus, thereby placing the patient on the path for failed treatment secondary to erroneous risk stratification. The recent advent of targeted prostate biopsies under MRI is currently changing the landscape of prostate cancer care and is the fastest growing sector of the field of iMRI. MRI-guided biopsies of cancersuspicious areas detected on multiparametric prostate MRI (Fig. 66-3) have been reported to detect cancer in 37% to 59% of patients with elevated PSA level and one or more previous negative TRUS biopsy sessions.30 More importantly, the implementation of MRI-guided biopsy of cancer-suspicious areas has been associated with detection of predominantly (87%-93%) clinically significant prostate cancer,23,30 compared to an estimated 56% of clinically significant cancers for repeat TRUS biopsy,74 thereby providing a relevant impact on risk stratification49 and subsequent management. In contrast to the freehand technique usually used for most MRIguided body biopsies,75,97 MRI-guided prostate biopsies are typically obtained using some sort of needle holder, commonly via the transrectal or transperineal approach. With the transperineal approach, the patient is placed in lithotomy position on the MRI table. A needle guide grid is placed against the perineum and an MRI scan is performed to localize the grid relative to the prostate target(s) in a manner similar to the earlier concept of MRI-guided breast biopsies.69,100 We typically perform MRI-guided biopsies on a 0.3T MRI scanner

2370

A

PART III  Image-Guided Procedures

B FIG 66-4  Transrectal approach for MRI-guided prostate biopsy using interventional planning hardware (A) and computer-aided detection software (B) (DynaTRIM, DynaCAD, Invivo, Pewaukee, WI). (From Nour SG: MRI Guided Focal Laser Ablation for Prostate Cancer. http://www.diagnosticimaging.com/authors/sherifg-nour-md#sthash.37ew1lmZ.dpuf; with permission.)

(Magnetom Trio, Siemens, Germany). The patient is placed in the prone position on the MRI table. The rectal piece of the Dyna-TRIM system (Invivo, USA) is inserted following lubrication with lidocaine gel. The transrectal piece is then connected to the rest of the system. Preliminary images are obtained to identify the transrectal fiducial line and the target lesion(s) and to calculate the trajectory angles in three planes (Fig. 66-4). Subsequently, an 18-gauge core biopsy needle is introduced into the target lesion to collect an average of three cores per lesion.21

Breast Biopsies.  Although the use of breast MRI has improved the sensitivity for detecting occult breast lesions,24 the reported specificity of MRI techniques, including dynamic contrast-enhanced MRI, in characterizing breast cancer remains low to moderate, ranging between 59.4% and 85%.27,81,83,92 The concept of using MRI to guide biopsies of lesions seen on MRI but occult on mammography and US is one of the early applications of iMRI. Although some groups have described the use of an interactive freehand approach for breast biopsy and wire localization under MRI17,70,83,87 similar to that used for other body biopsy applications, mainstream practice has always revolved around using a technique of “dead-reckoning navigation” that matches the skill set of mammographers performing stereotactic breast biopsies and avoids the higher rate of manipulations and wire displacement reported with the freehand technique.68 The technology for MRI-guided breast biopsy has been commercially available for more than a decade and is currently an integral component of numerous modern breast imaging and intervention centers. The MRI coil is modified to provide both immobilization and stereotaxy, most commonly in the prone position.26,27,67,68,98 The goal is to achieve moderate immobilization with compression plates that is sufficient for accurate targeting without altering blood flow and enhancement characteristics of the target lesion.93 The coil/compression plate is designed as a fenestrated grid equipped with MR fiducial markers that facilitate calculation of the stereotactic access trajectory.

Musculoskeletal Biopsies.  The use of MRI to guide biopsies of the musculoskeletal system is complementary to the current modes of guidance under CT and US. The major role of MRI in musculoskeletal tissue sampling is to identify and accurately target small (Fig. 66-5)

and ill-defined (Fig. 66-6) lesions that cannot be accessed using the traditional approaches under CT or US guidance.65 Device guidance is typically performed using the freehand technique, although other modes of MRI guidance are technically applicable.65 Obviously large soft tissue lesions and osseous tumors breaching the cortex can be readily sampled with CT or US-guided biopsies, with reported successful sampling in up to 98% of cases for each modality.47,76 MRI guidance may, however, have an added value in sampling large deeply seated lesions in complex locations when a double oblique approach or precise vascular localization is considered necessary for a safer procedure.65 Furthermore, in cases of inconclusive previous biopsies of large musculoskeletal lesions, MRI may be used to guide the biopsy needle into nonnecrotic parts of a tumor through the use of T2 contrast weighting.65 The fast T2-weighted gradient recalled echo (GRE) sequences usually used to guide interactive navigation during MRI-guided interventions are typically sufficient to identify necrotic areas; however, these fast guidance sequences may need to be reviewed in conjunction with more reliable planning turbo-spin echo (TSE) T2-weighted33 scans to accurately delineate areas of tumor necrosis if needed.

MRI-Guided Thermal Ablation

Concepts for Image Guidance and Monitoring.  Performing percutaneous image guided thermal ablation entails guiding the ablative device to the target, confirming placement relative to the 3D geometry of the target, and deploying the ablative energy with the goal of achieving complete tissue necrosis of the tumor along with a rim of adjacent normal tissue. The vast majority of percutaneous tumor ablations are currently performed under CT or US guidance. These modalities usually suffice for guiding the ablative device and in most cases confirming accurate placement. Some target tumors are, however, difficult to access because of complex trajectory, target invisibility, or proximity to vital structures. In these cases the excellent soft tissue contrast, high spatial resolution, multiplanar capabilities, and vascular conspicuity14,15,79 are all features specific to MRI that facilitate greater suitability for safe navigation and accurate placement of the ablative device. Treatment monitoring and quantifying the deposited ablative energy is where the existing standard-of-care technology is lacking and results in ablative treatments being delivered based on predetermined

CHAPTER 66  MRI-Guided Interventions

2371

A

CORE 1

B

CORE 2

C

CORE 3

D

CORE 4

E

FIG 66-5  A 54-year-old woman with left calf pain. A, Axial True-FISP (repetition time [TR]/echo time[TE]/ fractional anisotropy [FA] = 1454.2/1.89/60 degrees) from a series of dynamic triorthogonal plane scans demonstrate a focus of high signal abnormality (arrows) within the soleus muscle. B-E, sequential interactive true-FISP scans obtained during deployment of the cutting blade of an 18-gauge MRI-compatible semiautomatic biopsy gun. Images highlight the ability to clearly visualize the boundaries of the target lesion, avoid injury to vascular structures, and plan the sample acquisition in such a way as to represent all areas of lesion from the anterior-most (B) to the posterior-most (E) aspect. Pathology result revealed regenerating and necrotic myofibroblasts and focal hemosiderin with fibrosis, indicating injury. (From Nour SG, Monson DK: MRI-guided musculoskeletal soft tissue interventions. Top Magn Reson Imaging 22:197-205, 2011, with permission.)

ablation parameters. Therefore the major contribution of MRI to thermal ablation technology is its ability to monitor the zone of thermal coagulation during the procedure, thereby facilitating online feedback on the progress of energy deposition and its effect on the target tissues. Through MRI monitoring, the size and configuration of the thermal ablation zone can be directly controlled by the interventionist and adjusted during the procedure to compensate for deviations from preoperative predictions and define the treatment end point without moving the patient from the interventional suite (Fig. 66-7). This individually tailored response-based treatment represents a significant departure from the currently practiced standard “recipe” treatment strategies. This monitoring ability is achieved either by direct anatomic imaging of the zone of ablation with T2-weighted imaging techniques or by creating a thermal map allowing real-time updates of temperature changes to be superimposed on the anatomic image of the ablation zone. A modification of the latter entails creating a total damage estimate map highlighting the areas that have been exposed to

lethal temperatures for a certain amount of time, rather than the mere reflection of temperature distribution at the time of sampling. These features are exclusive to MRI and could not be reliably duplicated by any other currently used imaging modality. The sensitivity of MRI to the immediate changes occurring at the ablation zone and surrounding reactive tissue applies to all modes of thermal ablation but is easier to monitor during laser or cryoablation than RFA. MRI can simply be simultaneously performed during laser or cryoablation but is more complicated during RFA, owing to interference with the scanner. This limitation is usually circumvented by implementing intermittent MR scanning between the ablation cycles, although an RF switching circuit may be employed to allow simultaneous RFA and MRI if desired.103 CT scans are insensitive to the immediate temperature-mediated tissue changes, except in cryoablation, where the iceball can be temporarily detected as a hypodensity before thawing is performed. If intravenous iodinated contrast is administered after RFA or laser ablation, a nonenhancing area appears at the

2372

PART III  Image-Guided Procedures HAR

A

HRP

B

C

FIG 66-6  A 66-year-old man with history of leg sarcoma, status post resection and radiation therapy. A, Sagittal fat-saturated proton density scan (TR/TE/FA = 3000/49/90 degrees) obtained 2 years after surgery demonstrate a subtle focus of increased signal (arrow) within the proximal aspect of the right soleus muscle. B, A similar scan obtained 8 months later demonstrates interval increase in the size of signal abnormality (arrow) with associated fluid smearing the soleus-gastrocnemius fascial plane. C, A similar plane from an interactive triorthogonal plane true-FISP scan (TR/TE/FA = 975.7/1.89/60 degrees) obtained during MRIguided biopsy of this signal abnormality, demonstrating the biopsy needle in cross section (arrow) within the lesion. Pathology result revealed recurrent myxoid sarcoma. (From Nour SG, Monson DK: MRI-guided musculoskeletal soft tissue interventions. Top Magn Reson Imaging 22:197-205, 2011, with permission.)

site of ablation that is typically interpreted as the extent of the ablation zone (Fig. 66-8), although direct correlation with the exact extent of cell death has not been investigated. During US monitoring of RFA procedures, gas microbubbles begin to accumulate in the tissues, thus creating progressively increasing echogenic areas that cast posterior acoustic shadowing, which masks the exact margins of the target tumor, precluding accurate determination of the treatment end point29 (Fig. 66-9).

Technical Considerations.  Early reports on MRI-guided thermal ablations described the use of a biplanar low-field (0.2T) magnet design that allows ample room for ablation device handling and manipulations.44 Subsequent trends in iMRI have favored a shift to higher field interventions in order to take advantage of the higher spatial and temporal resolutions offered by the high-field MRI systems. The widebore, open-configuration cylindrical 1.5T interventional scanners do not, however, offer the free side access allowed by the design of biplanar magnets. The more restricted procedure environment within the bores of these cylindrical scanners dictated different access strategies than those originally described and rendered the adoption of flexible laser fiber71,96 optics a more appealing choice for in-bore thermal ablation compared to the cumbersome handling of device handles and cables during RFA, cryoablation, and microwave ablation.66 At our institution, we have therefore adopted laser energy in the recent several years as the primary mode of ablation within the MRI environment. We currently use a 980-nm diode laser system (Visualase, Inc.) to deliver direct ablative energy to primary and metastatic lesions in various body organs under direct MRI guidance and monitoring (Fig. 66-10). The laser fiber is housed within a flexible applicator that

is inserted through an MRI-compatible 14.5-cm-long, 14-gauge introducing needle. Advancing the MRI-compatible introducing needle under interactive MRI guidance is simple despite the limited room available within the MRI gantry, owing to the needle’s short shaft and lack of bulky handles and connecting cables. MRI monitoring is feasible during deployment of laser energy because of its lack of interference with the scanner, as is the case during RFA. This facilitates continuous acquisition of temperature-sensitive sequences during ablation, thereby allowing online monitoring of the progress of tissue damage. This process can be achieved with temperature mapping without or with the application of motion correction algorithms, depending on the target organ and possible effects of respiratory or cardiac motion on the accuracy of monitoring of heat deposition. MRI-compatible versions of several other ablative devices are currently available on a commercial basis. For example, MRI-compatible RF electrodes are offered as titanium electrodes of the Cool Tip RF system (Covidien) and nitinol electrodes of the RITA system (AngioDynamics). The Cool Tip RF system electrodes operate with continuous cooling via circulating iced water inside the electrode shaft to reduce charring at the electrode-tissue interface, thereby maximizing the achievable ablation zone size. The RITA system offers electrodes with multiple active tines that can be extended within the target tumor to produce larger ablation zones. A flexible-shaft RF electrode (StarBurst Semi-Flex) is provided by RITA systems. This offers similar expandable tines, but the shaft can also be bent to improve the capability for electrode navigation in the limited in-bore space available during iMRI procedures. In our experience, only mild bending of these electrodes is practical. The torque associated with excessive electrode bending within the MR gantry may result in inaccurate tip placement

CHAPTER 66  MRI-Guided Interventions

k

k

A

C

B

k

k

*

*

D

E

F

FIG 66-7  MRI scans obtained intermittently during RFA of an exophytic anterior lower pole right-sided clearcell carcinoma in a 75-year-old man. A, Transverse fast-spin echo (FSE) T2-weighted image (TR/TE: 3465/105 milliseconds, number of signal averages [NSA] = 4, echo train length [ETL] = 17) after MR-guided insertion of electrode (arrowheads) but before ablation demonstrates intermediate-signal anterior exophytic tumor (arrows) involving the lower pole of the right kidney (k). B, Transverse FSE T2-weighted imaging scan (TR/ TE: 3465/105 milliseconds, NSA = 4, ETL = 17) after two ablation cycles lasting 15 and 6 minutes at the same electrode location demonstrates complete thermal damage of the tumor at this transverse level as indicated by the development of uniformly low signal (arrows) around the needle electrode (arrowheads). C, When evaluated on coronal FSE T2-weighted imaging (TR/TE: 1856/105 milliseconds, NSA = 4, ETL = 17), most of the tumor is treated, as indicated by the development of hypointensity (straight arrow) around the needle electrode (arrowhead), but a residual untreated portion of the tumor is detected by the intermediate signal crescent (curved arrow) capping the ablation zone near the junction of the tumor with the kidney (k). D, Coronal FISP image (TR/TE: 17.8/8.1 milliseconds, NSA = 3, flip angle of 90) demonstrates the RF electrode position (arrowhead) after interactive repositioning superoposteriorly into residual tumor tissue (straight arrows). Guidance into residual tumor was based on localization from T2-weighted images (C) and was confirmed with additional FSE T2-weighted images before the next ablation cycle, because the treated tumor (curved arrow) does not appear hypointense on FISP images. A third ablation cycle was performed at this location for 12 additional minutes. E, Coronal FSE T2-weighted image (TR/TE: 1898/105 milliseconds, NSA = 4, ETL = 17) after the third RF application reveals complete replacement of the tumor by hypointense thermally damaged tissue (*) surrounded by a faint hyperintense rim of reactive tissue changes (arrowheads). F, Transverse FSE T2-weighted image (TR/TE: 1898/105 milliseconds, NSA = 4, ETL = 17) after the third RF application shows a hypointense ablation zone completely replacing the superior portion of the tumor (*) and documents adequate treatment of a margin of normal kidney (arrowheads). (From Lewin JS et al: Phase II clinical trial of interactive MR-guided interstitial radiofrequency thermal ablation of primary kidney tumors: Initial experience. Radiology 232:839, 2004, with permission.)

2373

2374

A

PART III  Image-Guided Procedures

C

B

FIG 66-8  A, Axial postgadolinium volumetric interpolated breath-hold examination (VIBE) (TR/TE/FA = 3.81/2.11/10 degrees) obtained in the venous phase (70 seconds following contrast administration), demonstrating a posterior lower pole intraparenchymal hypervascular renal neoplasm (arrowheads). B, Axial confirmation CT scan following RF electrode (arrows) insertion during CT-guided RFA. Distinction of tumor margins from the surrounding renal parenchyma is difficult, and the RF electrode insertion in this procedure was targeting the approximate area of the tumor as shown on the MRI scan (A) that was kept on display on a side monitor to be used as a “road map” for CT-guided insertion. C, Contrast-enhanced CT scan obtained through the ablation zone immediately after the procedure. The nonenhancing area (arrowheads) is presumed to represent the area of necrosis, but direct correlation with extent of cell death has not been investigated. In contrast to MRI monitoring, intraprocedural assessment of the extent of treatment is not possible. Treatment planning is therefore based on a “recipe” that assumes the required parameters of ablation rather than a tailored plan based on individual tumor response. This approach typically results in a larger ablation zone to ensure inclusion of a suboptimally targeted tumor in a treatment area that is difficult to monitor during the procedure. (From Nour SG, Lewin JS: MRI-guided RF ablation in the kidney. In Kahn T, Busse H [eds]: Interventional Magnetic Resonance Imaging, New York, Springer, 2012, pp 319-339, with permission.)

A

B

RT KIDNEY

FIG 66-9  A, Grayscale US image prior to RFA. Renal cell carcinoma is identified as a rounded mass extending off the posterior kidney (arrow). B, Grayscale US image during RFA. The renal cell carcinoma is now obscured by the gas bubbles that form as the tissue water boils during RFA (arrows). (From Hinshaw JL, Lee FT Jr: Image-guided ablation of renal cell carcinoma. Magn Reson Imaging Clin N Am 12:429-447, 2004, with permission.)

and potential tissue injury. Cryohit (Galil Medical) is an example of an MRI-compatible argon-based cryotherapy system. Microtaze (Alfresa) is a microwave system that has long been used for liver ablation. They provide an MRI-compatible version of a microwave needle that is made of brass coated with silver and gold. Regardless of the source of ablative energy used, the ablation generator is typically placed outside the MRI room, with the standard laser fibers or special extension cables of other ablative devices extended through a waveguide to the patient inside the magnet bore.

At our institution, we currently perform MRI-guided thermal ablations under general anesthesia. It is of utmost importance that the anesthesia and nursing teams as well as the ancillary staff be familiar with the requirements and limitations of working in the MRI suite. MRI-compatible anesthesia equipment and monitoring devices are integral parts of the iMRI suite that are typically already available in most diagnostic MRI suites and are therefore not associated with additional cost. We perform most of our iMRI procedures on a cylindrical high-field, short-bore iMRI scanner (Magnetom Espree, Siemens

CHAPTER 66  MRI-Guided Interventions

A

B

C

D

E

F

2375

FIG 66-10  A 51-year-old man with history of prostate cancer. A, Axial gadolinium-enhanced VIBE scan (TR/ TE/FA = 3.81/2.11/10 degrees) demonstrates an enhancing left inferior sacral lesion with early soft tissue extension (arrows). MRI-guided biopsy (not shown) revealed metastatic adenocarcinoma from a prostatic primary. PSA level = 33.47 ng/mL. B, Intraprocedural axial TSE T1-weighted image (TR/TE/FA = 500/4.4/150 degrees) obtained to confirm the locations of 2 laser fibers (arrows) (Visualase, Inc.) within the lesion for MRI-guided laser ablation. C, Intraprocedural monitoring with cumulative damage estimate map showing the extent of predicted thermal ablation zone (arrows) superimposed on the magnitude image. D, Immediate intraprocedural postablation TSE T1-weighted image following intravenous gadolinium administration, demonstrating the extent of tissue necrosis as an iso-/hypointense area surrounded by a thin enhancing rim (arrows). E, Three-month follow-up with axial gadolinium-enhanced VIBE scan demonstrates the ablation zone with slight interval involution in size (arrows) and complete lack of the internal enhancement seen on the pretreatment scan. PSA level has dropped to 0.09 ng/mL. F, Nine-month follow-up with axial gadoliniumenhanced VIBE scan demonstrates further involution of the ablation zone (arrows). PSA level has further dropped to less than 0.05 ng/mL. (From Nour SG, Monson DK: MRI-guided musculoskeletal soft tissue interventions. Top Magn Reson Imaging 22:197-205, 2011, with permission.)

Medical Solutions, Erlangen, Germany) while accessing the patient from the backside of the scanner. When using general anesthesia, the patient lies on the scanner in feet-first position, allowing the anesthesia team to have access to the patient’s head and use the entire space around the front of the gantry.

The approach for MRI-guided focal prostate laser ablation is different. High-field 3T MRI guidance is preferred in these procedures because of improved delineation of the target lesion, leading to more accurate targeting. These procedures, like MRI-guided diagnostic  prostate biopsies, can be easily performed within the bore of a 

2376

PART III  Image-Guided Procedures = 117.7°C = 97.5°C = 77.3°C = 57.2°C = 37.0°C

FIG 66-11  MRI-guided and monitored focal laser ablation of a 1.7-cm left central midgland localized prostate cancer (Gleason 3+3=6). Intraprocedural real-time temperature and cumulative damage maps demonstrating interactive online assessment of the extent of laser ablation that allows accurate response-based determination of the treatment end point. (From Nour SG: MRI Guided Focal Laser Ablation for Prostate Cancer. http:// www.diagnosticimaging.com/authors/sherif-g-nour-md#sthash.37ew1lmZ.dpuf. With permission.)

regular (long, narrow) diagnostic MRI scanner, with the prostate target accessed transperineally via a perineal template device or transrectally via a fiducial rectal needle holder (Fig. 66-11) as described earlier. The simple setup for these procedures render them applicable in virtually any practice with standard MRI capabilities. This ease of incorporation into existing workflow paradigms is well matched with a currently realized need for a minimally invasive option to manage prostate cancer and to bridge the existing gap between surveillance and wholegland treatment protocols offered to these patients. One of the current issues with prostate cancer care is the concern that patients with lowrisk prostate cancer are being overtreated. Miller et al. studied the various treatments offered to 24,405 patients with low-risk prostate cancer and reported that 55% received whole-gland treatment in the form of radiation therapy (45%) or radical prostatectomy (10%), compared to only 45% who received expectant management.56 We envision that minimally invasive MRI-guided focal ablation will be established as a viable treatment option for a growing number of these patients over the next few years.

MRI Monitoring of the Thermal Ablation Process.  Using MRI to directly monitor the effect of thermal ablation allows detection of any focus of the target tumor that is either undertreated or properly treated but more resistant to thermal damage (e.g., owing to heat sink by adjacent vasculature). The ability to identify these areas and use MRI to accurately redirect additional treatment represents a significant improvement of thermal ablation technology and is a primary guard against potential treatment failures. This capability also allows a confident cessation of therapy once appropriate coverage is believed to be achieved and therefore minimizes the risk of complications from improperly aggressive therapy. Providing just the adequate amount of treatment is essential to preserving parenchymal function, such as in cases of liver metastases following partial hepatectomy and renal tumors in patients with prior nephrectomies. Temperature-sensitive MRI sequences have been developed to enable direct online monitoring of heat deposition.10,13,95 MRI can also detect the lethal effect of hyperthermia on viable tissues, rather than measuring the temperature change itself, through detecting the changes in the tissue relaxation parameters that accompany the phase

transition from the viable to the necrotic state.7,50 The ability of MRI to reliably define the extent of thermally induced tissue necrosis using several various sources of energy has been repeatedly demonstrated and correlated with histopathologic analysis.2,11,12,51,94 The development of tissue necrosis associated with thermal ablation involves several processes at the cellular level, including22,54 denaturation, shrinkage, aggregation of cytoplasmic proteins, and increased hydrophobic interactions resulting in the extrusion of water. The latter effect, along with binding between the denatured proteins and any residual free water, most likely represents the underlying cause for shortening of the T2 relaxation time after thermal ablation, which ultimately leads to the uniform hypointense appearance of ablation zones seen on the T2-weighted and short tau inversion recovery (STIR) images (Fig. 66-12A) acquired intermittently during the procedure. This feature allows direct observation of the size and configuration of the developing thermal ablation zone and permits identification of any foci of residual viable tumor, which appear as relatively hyperintense areas (see Fig. 66-12B) compared with the hypointense effectively ablated zones.

MRI Follow-up after Thermal Ablation.  On precontrast T1weighted MRI the thermal ablation zone may demonstrate a hypointense, isointense, slightly hyperintense, or markedly hyperintense54 signal intensity (Fig. 66-13). Inevitable hemorrhage occurs within the target tissues during the ablation process. The amount of ablationrelated parenchymal hemorrhage appears to correlate with the degree of target tissue vascularity22 and is believed to be responsible for the reduction of T1 relaxation time following thermal ablation procedures. Additionally, blood itself exhibits an abrupt decrease in T1 and T2 relaxation times at temperatures greater than 60°C, resulting in a hyperintense appearance on T1-weighted imaging.22 These T1-weighted signal changes are usually observed as early as the intraprocedural or immediate postprocedure scans and typically involute gradually over several months following the ablation procedure. On T2-weighted and STIR images, the uniform hypointense appearance of the ablation zone noted during the ablation procedure continues to be observed throughout the follow-up scans and gradually decreases in size with involution of the ablation zone. The ablative

CHAPTER 66  MRI-Guided Interventions

A

2377

B FIG 66-12  Careful evaluation of the margins of thermal ablation zones on follow-up MRI scans is essential to detect any new edge irregularity. Only hypointense signals on T2-weighted or STIR images are equivalent to coagulation necrosis. A, Axial T2-weighted image obtained 15 months after RFA of renal cell carcinoma. The regularly ovoid and diffusely hypointense area (arrowheads) seen along the anterior margin of the interpolar segment of the left kidney is consistent with a completely necrotic ablation zone. B, Another T2-weighted image obtained 1 year later demonstrates the previously seen hypointense necrotic zone (arrowheads) with interval development of an isointense nodule (arrow) along the posterior margin of the ablation zone. This appearance is typical for local tumor recurrence. The length of tumor-free period in this patient (27 months) should stress the value of carefully scrutinizing the margins of ablation zones on each follow-up scan. Comparison with prior scans, especially the first postablation (baseline) scan, is indispensable. (From Nour SG, Lewin JS: Radiofrequency thermal ablation: The role of MR imaging in guiding and monitoring tumor therapy. Magn Reson Imaging Clin N Am 13:561-581, 2005, with permission.)

A

B

C

D FIG 66-13  T1-weighted in-phase GRE images of four different ablated renal cell carcinomas (arrowheads, A-D) demonstrate the spectrum of possible MRI signal characteristics on high-field (1.5T) scans. Thermal ablation zones may appear hypointense (A), isointense (B), or hyperintense (C) on high-field GRE in-phase imaging. (From Merkle EM et al: MR imaging follow-up after percutaneous radiofrequency ablation of renal cell carcinoma: Findings in eighteen patients during first 6 months. Radiology 235:1068, with permission.)

device track is typically seen as a linear high-T2-weighted signal within the ablation zone, particularly on early postablation follow-up scans. Thermal ablation zones are typically associated with acute inflammatory reaction that encircles the ablation site and comprises edema, hyperemia, and foci of hemorrhage on histopathologic analysis.8,12,55,64,72

This inflammatory response presents as a bright rim marginating the area of necrosis on T2-weighted scans and demonstrates enhancement on postgadolinium T1-weighted scans. The inner margin of this rim is typically sharp, whereas its outer margin usually fades gradually into the adjacent intact tissues.44 Although we have demonstrated that the

2378

PART III  Image-Guided Procedures

extent of actual cell death does eventually extend to the outer margin of the inflammatory rim,12 we use the sharp inner margin as an unequivocal indicator of the definite extent of tissue necrosis on our intraprocedural T2-weighted scans, as well as on the confirmatory postablation imaging that includes the gadolinium-enhanced T1-  weighted series to ensure complete treatment of the targeted tumor. Careful observation of the extent of this inflammatory signal change is critical when ablation is performed close to a vital structure such as the colon, small bowel, or gallbladder. As the thermal ablation zone matures, the surrounding reactive tissue inflammation subsides and starts to be gradually invaded by granulation tissue laid concentrically around the necrotic ablated center. This granulation tissue ring subsequently undergoes active organization, resulting in progressive maturation into fibrous tissue in a concentric fashion proceeding from the periphery of the ablation zone toward the center (Fig. 66-14).64 These tissue changes are reflected on follow-up MRI scans as a gradual resolution of the hyperintense rim seen on T2-weighted scans, until it becomes barely detectable after approximately 3 months in most cases.44 The rim enhancement around the thermal ablation zone is explained during the early postablation period by the enhancing granulation tissue and later by the enhancing mature fibrotic tissue and is seen for longer periods during follow-up. Ultimately the chronic thermal ablation zone involutes into an area of featureless ‘‘mummified’’ coagulated tissue encased by a variable amount of fibrous tissue. The latter may further progress to replace the entire ablation zone, thus rendering it into a small contracted scar.64 Evaluation of postablation follow-up MRI scans warrants careful attention to the margins of the thermal ablation zone, where early tumor recurrence typically occurs. Marginal irregularity along the contour of the ablation zone should raise a concern for recurrence unless the configuration of the ablation zone has been known to be related to overlapping ablations at multiple device insertion sites. The presence of focal enhancement or focal loss of the uniform hypointense T2-weighted signal indicates a true recurrent neoplasm (Fig. 66-15). The ability to examine the ablation zone in multiple planes, apply different tissue contrast weightings, and implement modern functional imaging techniques such as diffusion-weighted imaging render MRI a superior modality for detecting early postablation recurrences, as compared to standard CT scans. Even with the use of MRI, there are factors that may complicate the evaluation for tumor recurrence. Ablation zones may demonstrate irregular outlines rather than the expected perfectly ovoid appearance even when no overlapping ablations have been performed. This occurs when the energy distribution is modified during the ablation procedure by the presence of an adjacent sizable blood vessel exerting a focal heat sink effect at one side of the ablation zone. Techniques used to augment the size of ablation zones, such as saline perfusion either directly or through perfused electrodes, have been described. These are intended to increase the ionic concentration in the tissues to enhance RF current flow and thereby the ablation zone size. These may also result in irregular zones of ablation; the final geography of tissue necrosis is related to the direction of saline flow in the tissues.59 This unpredictability of the final outcome of the ablation procedure has prevented widespread adoption of these techniques and is an important factor to consider when using MRI for procedure guidance and monitoring to provide a more controlled outcome whenever salineaugmented ablations are considered.59 A similar effect may occur when the housing applicator encasing the laser fiber is damaged by excessive heating during MRI-guided laser ablation procedures, resulting in leakage of the saline used to cool down the fiber tip during ablation. Postablation complications such as infection, hematoma, biloma, or urinoma occurring adjacent to the site of ablation are another factor

that can further complicate interpretation of follow-up MRI scans. Review of prior scans, particularly the first postablation baseline scans, should therefore be a routine practice that can help monitor interval development of new iso- or hyperintense areas within the confines of the hypointense ablation zone on T2-weighted images or the development of new focal enhancement along a uniform margin on postgadolinium scans. There is currently no consensus recommendation regarding the timing of acquisition of the first postablation baseline scans. Some groups do not acquire these scans until 3 months following the ablation procedure to allow time for the initial inflammatory response to settle down to simplify interpretation of these scans. Other groups believe in acquiring these scans as early as 2 or 3 weeks following the procedure to detect and re-treat any residual tumor without delay. Performing the actual intraprocedural monitoring under MRI provides an additional time point for comparison that facilitates a more accurate interpretation of early-acquired baseline scans, despite the presence of inflammatory tissue response.

MRI-Guided Fiducial Marker Placement and Wire Localization This is a relatively uncommon application of iMRI technology that applies a concept similar to the long-practiced breast fiducial marker placement and wire localization procedures. MRI can be used to mark the locations of soft tissue neoplasms in various body parts that are difficult to delineate for subsequent radiation treatment or surgical resection. Some soft tissue neoplasms, such as sarcomas, are characterized by their infiltrative nature precluding reliable delineation of their margins on CT or US for safe and effective delivery of radiation treatment or surgical resection. MRI is therefore considered the imaging modality of choice for tumor margin identification in these cases. MRIcompatible fiducial markers (ClipLoc, Invivo, USA) are currently commercially available for percutaneous placement under MRI guidance. The goal of placing these markers should be to bracket the margins of the target tumor for identification during image-guided radiation treatments. Each marker is supplied preloaded in a needle introducer that is inserted toward the target location under MRI. The deployed fiducial is a 7-mm-diameter C-shaped marker. These markers are not only MRI compatible but are also visible under x-ray, CT, and US to offer a wide flexibility of usage. Planning the placement of these fiducial markers around the margins of a tumor should be discussed with the radiation oncologist prior to the procedure to ensure useful placement during stereotactic radiation treatment. As a rule, the markers should ideally be placed in different planes along the x, y, and z axes of the patient.31 The same principle of bracketing infiltrative tumors by fiducial markers applies to subsequent surgical resection to ensure a tumor-free margin of the resected tissue. Similar to bracketing for radiation treatment, the placement of these fiducial markers before surgery should ideally be discussed with the surgical oncologist on a case-by-case basis to understand the intended surgical approach and therefore maximize usefulness of marker placement. In addition to infiltrative neoplasms, small-sized soft tissue tumors represent a surgical challenge even when they are quite superficial, owing to the inability to reliably palpate them at the time of surgery. Because these tumors are typically well visualized on MRI, we apply iMRI technology for interactive wire localization before surgical resection (Fig. 66-16). These procedures are typically performed on the same day as surgery. We use an MRI-compatible wire localization system (Invivo, USA) that features a hooking wire preloaded within a needle introducer. The latter is advanced into the target tumor, which may be a small deep tumor or a subcutaneous lesion/nodule, under MRI guidance using a freehand technique similar to that used during

CHAPTER 66  MRI-Guided Interventions

4 3 2 1

A

B 1

3

C

D

4

FIG 66-14  A, Gross pathologic specimen demonstrates bisected zones of coagulation necrosis after RFA within a porcine kidney. The pale coagulated tissues are surrounded by darker rims of reactive tissue response consisting of hyperemia, hemorrhage, and edema. RF electrode tracts are visible in the centers of the ablated zones and usually demonstrate some hemorrhage and edema as well. B, On histologic section, the central zone (zone 1) is an empty cavity around the needle track, which is surrounded by an area characterized by shrunken nuclei, collapsed tubules, and disorganized tissue architecture (zone 2) corresponding to the hypointense area seen on MRI. Outermost region (zone 3) shows shrunken cells, high density, and size of areas of vacuolization in histologic images, representing hyperemic edematous rim. Note welldemarcated border of lesion to normal kidney tissue (zone 4). Low-power (C) and high-power (D) trichromestained histologic specimens obtained 1 month following experimental RFA in porcine tongue muscle, demonstrating a closer detailed insight on the process of edge maturation of the ablation zone. Four distinct layers of histologic and pathologic findings encase the area of coagulation necrosis from outside inward as follows: (1) normal tissue, (2) mature fibrous tissue, (3) active granulation tissue, and (4) coagulated (mummified) muscle tissue. (A and B from Merkle EM et al: MR-guided RF thermal ablation of the kidney in a porcine model. AJR Am J Roentgenol 173:649, 1999, with permission. C and D from Nour SG et al. Percutaneous MR imaging–guided radiofrequency interstitial thermal ablation of tongue base in porcine models: Implications for obstructive sleep apnea syndrome. Radiology 230:367, 2004, with permission.)

1 mm

2

2379

2380

PART III  Image-Guided Procedures

A

B

C FIG 66-15  Two-week follow-up high-field MRI of the first of two debulking thermal ablation procedures performed to palliate this patient with a large centrally located renal cell carcinoma of the right kidney and global impairment of renal function. The ablated zone (arrows, A-C) demonstrates hypointense signal on FSE T2-weighted image (A), hyperintense blood signal on a T1-weighted GRE in-phase image (B), and no enhancement on a postgadolinium scan (C). The area of residual tumor tissue (white arrowheads, A-C) can be readily identified on all three pulse sequences as hyperintense tissue capping the anterior aspect of the ablated zone on FSE T2-weighted images (A) and as an isointense area on the in-phase image (B) that enhances on the postgadolinium scan (C). Note the left-sided nephrostomy tube (black arrowheads, A, C). (From Merkle EM et al: MR imaging follow-up after percutaneous radiofrequency ablation of renal cell carcinoma: Findings in eighteen patients during first 6 months. Radiology 235:1068, 2005, with permission.)

MRI-guided percutaneous biopsies. When the introducer is well seated within the target tumor (or immediately beyond it in the case of tiny lesions), the wire localizer is deployed to release two fishhook-like blades that anchor the tissues. The rest of the wire extends out through the skin surface, where it is taped and secured while the patient is transported to the operating room.

MRI-Guided Percutaneous Sclerotherapy iMRI technology has been applied to treat slow-flow (venouslymphatic) vascular malformations. These are congenital anomalies that usually become symptomatic during late childhood or early

adulthood. Symptoms are typically related to pain and/or cosmetic disfigurement. The current standard treatment for these lesions is percutaneous sclerotherapy under x-ray fluoroscopy. Because of the inability to visualize the malformation and its surrounding anatomy under conventional fluoroscopy, skin ulceration, scarring, and incomplete treatment are not infrequent problems with this approach. Lewin et al.41 introduced the concept of using MRI for guiding sclerotherapy on low-field interventional scanners. We believe that the current widespread shift to higher-field interventions provides an opportunity to revive this approach and improve it for targeting large, complex, and deep malformations, aiming at full therapeutic results60 through exploiting the fast scan times and improved image resolution and signal-to-noise ratio. We use triorthogonal image plane MR guidance to interactively observe and redirect a small (typically 22-gauge) MR-compatible needle into the targeted part of the malformation under MR fluoroscopic guidance using continuously updated sets of true fast imaging with steady-state precession (FISP) images. Once the needle tip is confirmed in place on MRI scans on three planes, further confirmation of appropriate location is performed by checking blood return through the introducing needle. Subsequently the sclerosing material is injected, mixed with dilute gadolinium, and monitored on a triorthogonal fast low-angle shot (FLASH) sequence (Fig. 66-17). We use 5% ethanolamine oleate (Ethamolin) mixed with 6% dilute gadolinium (1.0 mL:0.15 mL vol). Injection end points are tailored for individual lesions based on lesion size and adequacy of filling under real-time monitoring. A characteristic feature of these vascular malformations is the presence of multiple vascular compartments that do not have a predictable pattern of communication, and the presence of completely secluded compartments is not an uncommon finding. This attribute highlights the value of interactive MRI monitoring of the distribution of the sclerosing material during injection in relation to the anatomy of each individual vascular malformation. This technique facilitates identification of undertreated or completely secluded areas that require dedicated targeting by needle redirection or addition punctures. Additionally the use of MRI guidance facilitates selective targeting of deeper or complex areas of the malformation, thereby eliminating potential niduses of recurrence that could otherwise be left behind. The current standard-of-care treatment under x-ray fluoroscopy cannot offer this level of selective targeting or detailed information on the extent of filling in relation to the overall volume of the malformation. Other than contributing to the efficacy of treatment, the use of MRI to guide and monitor sclerotherapy procedures adds to procedure safety by rapidly identifying sclerosing material extravasation toward the skin and other critical structures while monitoring injections on planes. The flexibility of independently aligning the three planes of guidance/ monitoring also allows continuous monitoring of sizeable draining veins as desired. Finally, using MRI as a radiation-free tool to perform these procedures adds another safety dimension compared to standard x-ray fluoroscopy, given the often required repeated treatment session in this predominantly younger population of patients.

CONCLUSION The use of MRI to guide interventions and interactively monitor local therapies is feasible and gaining increasing popularity as a recognized mode of minimally invasive intervention. With its unequivocally superior soft tissue contrast, multiplanar capabilities, and inherent vascular conspicuity, MRI can positively influence interventional procedures by adding safety during device navigation, by facilitating interventional

CHAPTER 66  MRI-Guided Interventions

A

B

C

D

E

FIG 66-16  A 61-year-old man with metastatic renal cell carcinoma. A, Axial gadolinium-enhanced VIBE scan (TR/TE/FA = 3.81/2.11/10 degrees) demonstrates a 1.5-cm-diameter cutaneous/subcutaneous enhancing nodule (0.9 cm in thickness) (arrow). B, Axial FLASH (TR/TE/FA = 1220/1.85/15 degrees) image from interactive triorthogonal plane guidance series obtained during insertion of an MRI-compatible wire localization system (arrow) to facilitate subsequent identification at during surgical resection. Axial (C), sagittal (D), and coronal (E) gadolinium-enhanced VIBE scans obtained immediately after wire (arrows) placement, confirming its central location. Subsequent pathologic evaluation of the surgical specimen revealed a neurofibroma. (From Nour SG, Monson DK: MRI-guided musculoskeletal soft tissue interventions. Top Magn Reson Imaging 22:197-205, 2011, with permission.)

2381

2382

PART III  Image-Guided Procedures

A

B

C

D

FIG 66-17  A 14-year-old boy with elbow pain and swelling precluding him from practicing football. A, Axial fat-saturated TSE T2-weighted image demonstrating a large venous malformation (arrows) underneath the triceps tendon. B, Selected axial FLASH (TR/TE, 2484/5.4) image from a series of interactive triorthogonal plane scans obtained during injection of sclerosing material (ethanolamine oleate) demonstrating the MRIcompatible 22-gauge needle (arrow) with the tip placed deep at the farthest end of the malformation. The early phase of injection is seen here, with a small amount of the sclerosing material (mixed with dilute gadolinium) starting to accumulate within the malformation. C, A similar scan from the same continuous FLASH series during monitoring of the injection process, demonstrating a later phase of the injection. The needle (arrow) has been pulled out gradually to fill the proximal aspect of the malformation. D, Follow-up axial fatsaturated TSE T2-weighted image obtained after 27 months. The patient has had a total of 6 similar MRIguided sclerotherapy sessions. The follow-up scan demonstrates almost complete resolution of the vascular malformation except for the tiny remnant indicated by the arrow. The patient became symptom free and was able to resume his sports activities. (From Nour SG, Monson DK: MRI-guided musculoskeletal soft tissue interventions. Top Magn Reson Imaging 22:197-205, 2011, with permission.)

device manipulation in areas that were until recently considered strictly surgical zones, and by introducing new procedures to the field of minimally invasive interventional radiology. It is to be stated that the development of the field of iMRI is not meant to replace the already established earlier methods of fluoroscopic-, US-, and CT-guided interventions. Rather it is meant to expand the options of minimally invasive diagnosis and therapy to a subset of patients who would otherwise revert to more conventional invasive forms of management. Finally, it is essential to realize that iMRI does not only involve the actual MRI-guided interventional event but should be perceived as a whole foundation of technology, development, setup, and conceptual training and as an institutional culture that realizes the opportunities offered by MRI-guided interventions and understands the challenges and limitations. Support staff including anesthesia  personnel, MRI technologists, nursing staff, and patient transportation personnel are integral to the success of these procedures and should be similarly trained and oriented in the various aspects of iMRI, particularly those pertaining to patient and personnel safety within the MRI environment.

REFERENCES 1. Adam G, Bucker A, Nolte-Ernsting C, et al: imaging: Percutaneous abdominal and skeletal biopsies and drainages of the abdomen. Eur Radiol 9(8):1471–1478, 1999. Review. 2. Anzai Y, Lufkin RB, Hirschowitz S, et al: MR imaging-histopathologic correlation of thermal injuries induced with interstitial Nd:YAG laser irradiation in the chronic model. J Magn Reson Imaging 2(6):671–678, 1992. 3. Bernays RL, Kollias SS, Khan N, et al: Histological yield, complications, and technological considerations in 114 consecutive frameless stereotactic biopsy procedures aided by open intraoperative magnetic resonance imaging. J Neurosurg 97(2):354–362, 2002. 4. Bernays RL, Kollias SS, Romanowski B, et al: Near-real-time guidance using intraoperative magnetic resonance imaging for radical evacuation

5.

6.

7.

8.

9.

10.

11.

12.

13.

14. 15. 16.

of hypertensive hematomas in the basal ganglia. Neurosurgery 47(5):1081–1089, discussion 1089-90, 2000. Bernays RL, Kollias SS, Yonekawa Y: Dynamic changes during evacuation of a left temporal abscess in open MRI: Technical case report. Neuroradiology 44(5):438–442, 2002. [Epub 2002 Feb 08]. Beyersdorff D, Winkel A, Hamm B, et al: MR Imaging-guided prostate biopsy with a closed MR unit at 1.5 T: Initial results. Radiology 234(2):576–581, 2005. Bleier AR, Jolesz FA, Cohen MS, et al: Real-time magnetic resonance imaging of laser heat deposition in tissue. Magn Reson Med 21(1):132– 137, 1991. Boaz TL, Lewin JS, Chung YC, et al: MR monitoring of MR-guided radiofrequency thermal ablation of normal liver in an animal model.   J Magn Reson Imaging 8(1):64–69, 1998. Boll DT, Lewin JS, Duerk JL, et al: Comparison of MR imaging sequences for liver and head and neck interventions: Is there a   single optimal sequence for all purposes? Acad Radiol 11(5):506–515, 2004. Botnar RM, Steiner P, Dubno B, et al: Temperature quantification using the proton frequency shift technique: In vitro and in vivo validation in an open 0.5 Tesla interventional MR scanner during RF ablation. J Magn Reson Imaging 13(3):437–444, 2001. Breen MS, Lancaster TL, Lazebnik RS, et al: Three-dimensional method for comparing in vivo interventional MR images of thermally ablated tissue with tissue response. J Magn Reson Imaging 18(1):90–102, 2003. Breen MS, Lazebnik RS, Fitzmaurice M, et al: Radiofrequency thermal ablation: Correlation of hyperacute MR ablation zone images with tissue response. J Magn Reson Imaging 20(3):475–486, 2004. Chung YC, Duerk JL, Shankaranarayanan A, et al: Temperature measurement using echo-shifted FLASH at low field for interventional MRI. J Magn Reson Imaging 9(1):138–145, 1999. Cline HE, Hynynen K, Watkins RD, et al: Focused US system for MR imaging-guided tumor ablation. Radiology 194(3):731–737, 1995. Cline HE, Schenck JF, Watkins RD, et al: Magnetic resonance-guided thermal surgery. Magn Reson Med 30(1):98–106, 1993. Davis SP, Anand VK, Dhillon G: Magnetic resonance navigation for head and neck lesions. Laryngoscope 109(6):862–867, 1999.

CHAPTER 66  MRI-Guided Interventions 17. deSouza NM, Kormos DW, Krausz T, et al: MR-guided biopsy of the breast after lumpectomy and radiation therapy using two methods of immobilization in the lateral decubitus position. J Magn Reson Imaging 5(5):525–528, 1995. 18. Fenchel S, Boll DT, Lewin JS: Intraoperative MR imaging. Magn Reson Imaging Clin N Am 11(3):431–447, 2003. Review. 19. Frahm C, Gehl HB, Weiss HD, et al: Technique of MRT-guided core biopsy in the abdomen using an open low-field scanner: Feasibility and initial clinical results. Rofo 164(1):62–67, 1996. German. 20. Fried MP, Hsu L, Jolesz FA: Interactive magnetic resonance imagingguided biopsy in the head and neck: Initial patient experience. Laryngoscope 108(4 Pt 1):488–493, 1998. 21. Geller RL, Nour SG, Osunkoya AO: Pathologic Findings in Patients with Targeted Magnetic Resonance Imaging-Guided Prostate Needle Core Biopsies. Int J Clin Exp Pathol 8(9):9790–9795, 2015. 22. Graham SJ, Stanisz GJ, Kecojevic A, et al: Analysis of changes in MR properties of tissues after heat treatment. Magn Reson Med 42(6):1061– 1071, 1999. 23. Hambrock T, Somford DM, Hoeks C, et al: Magnetic resonance imaging guided prostate biopsy in men with repeat negative biopsies and increased prostate specific antigen. J Urol 183(2):520–527, 2010. 24. Harms SE, Flamig DP, Hesley KL, et al: MR imaging of the breast with rotating delivery of excitation off resonance: Clinical experience with pathologic correlation. Radiology 187(2):493–501, 1993. 25. He Y, Zhang Z, Tian Z, et al: The application of magnetic resonance imaging-guided fine-needle aspiration cytology in the diagnosis of deep lesions in the head and neck. J Oral Maxillofac Surg 62(8):953–958, 2004. 26. Heywang-Kobrunner SH, Heinig A, Schaumloffel U, et al: MR-guided percutaneous excisional and incisional biopsy of breast lesions. Eur Radiol 9(8):1656–1665, 1999. 27. Heywang-Kobrunner SH, Huynh AT, Viehweg P, et al: Requardt H, Paprosch I. Prototype breast coil for MR-guided needle localization.   J Comput Assist Tomogr 18(6):876–881, 1994. 28. Hinks RS, Bronskill MJ, Kucharczyk W, et al: MR systems for image-guided therapy. J Magn Reson Imaging 8:19–25, 1998. 29. Hinshaw JL, Lee FT, Jr: Image-guided ablation of renal cell carcinoma. Magn Reson Imaging Clin N Am 12(3):429–447, 2004. 30. Hoeks CM, Schouten MG, Bomers JG, et al: Three-tesla magnetic resonance-guided prostate biopsy in men with increased prostatespecific antigen and repeated, negative, random, systematic, transrectal ultrasound biopsies: detection of clinically significant prostate cancers. Eur Urol 62(5):902–909, 2012. 31. . 32. Kariniemi J, Blanco Sequeiros R, Ojala R, et al: MRI-guided abdominal biopsy in a 0.23-T open-configuration MRI system. Eur Radiol 15(6):1256–1262, 2005. 33. Koenig CW, Duda SH, Truebenbach J, et al: MR-guided biopsy of musculoskeletal lesions in a low-field system. J Magn Reson Imaging 13:761–768, 2001. 34. Kollias SS, Bernays RL: Interactive magnetic resonance imaging-guided management of intracranial cystic lesions by using an open magnetic resonance imaging system. J Neurosurg 95(1):15–23, 2001. 35. Kollias SS, Bernays R, Marugg RA, et al: Target definition and trajectory optimization for interactive MR-guided biopsies of brain tumors in an open configuration MRI system. J Magn Reson Imaging 8(1):143–159, 1998. 36. Konig CW, Pereira PL, Trubenbach J, et al: MR imaging-guided adrenal biopsy using an open low-field-strength scanner and MR fluoroscopy. AJR Am J Roentgenol 180(6):1567–1570, 2003. 37. Konig CW, Trubenbach J, Fritz J, et al: Contrast enhanced MR-guided biopsy of hepatocellular carcinoma. Abdom Imaging 29(1):71–76, 2004. 38. Kooy HM, Cormack RA, Mathiowitz G, et al: A software system for interventional magnetic resonance image-guided prostate brachytherapy. Comput Aided Surg 5(6):401–413, 2000.

2383

39. Lai A, Maghami E, Borges A, et al: MRI-guided access to the retropharynx. J Magn Reson Imaging 17(3):317–322, 2003. 40. Lewin JS: Interventional MR imaging: Concepts, systems, and applications in neuroradiology. AJNR Am J Neuroradiol 20(5):735–748, 1999. Review. 41. Lewin JS, Merkle EM, Duerk JL, et al: Low-flow vascular malformations in the head and neck: Safety and feasibility of MR imaging-guided percutaneous sclerotherapy: Preliminary experience with 14 procedures in three patients. Radiology 211:566–570, 1999. 42. Lewin JS, Metzger AK: Intraoperative MR systems. Low-field approaches. Neuroimaging Clin N Am 11(4):611–628, 2001. Review. 43. Lewin JS, Metzger A, Selman WR: Intraoperative magnetic resonance image guidance in neurosurgery. J Magn Reson Imaging 12(4):512–524, 2000. Review. 44. Lewin JS, Nour SG, Connell CF, et al: Phase II clinical trial of interactive MR imaging guided interstitial radiofrequency thermal ablation of primary kidney tumors: Initial experience. Radiology 232:835–845, 2004. 45. Lewin JS, Nour SG, Duerk JL: Magnetic resonance image-guided   biopsy and aspiration. Top Magn Reson Imaging 11(3):173–183, 2000. Review. 46. Lewin JS, Petersilge CA, Hatem SF, et al: Interactive MR imagingguided biopsy and aspiration with a modified clinical C-arm system. AJR Am J Roentgenol 170(6):1593–1601, 1998. 47. Logan PM, Connell DG, O’Connell JX, et al: Image-guided percutaneous biopsy of musculoskeletal tumors: An algorithm for selection of specific biopsy techniques. AJR Am J Roentgenol 166:137– 141, 1996. 48. Maghami EG, Bonyadlou S, Larian B, et al: Magnetic resonance imaging–guided fine-needle aspiration biopsies of retropharyngeal lesions. Laryngoscope 111(12):2218–2224, 2001. 49. Marcus DM, Rossi PJ, Nour SG, et al: The impact of multiparametric pelvic magnetic resonance imaging on risk stratification in patients with localized prostate cancer. Urology 84(1):132–137, 2014. 50. Matsumoto R, Oshio K, Jolesz FA: Monitoring of laser and freezinginduced ablation in the liver with T1-weighted MR imaging. J Magn Reson Imaging 2(5):555–562, 1992. 51. Matsumoto R, Selig AM, Colucci VM, et al: MR monitoring during cryotherapy in the liver: Predictability of histologic outcome. J Magn Reson Imaging 3(5):770–776, 1993. 52. Menard C, Susil RC, Choyke P, et al: MRI-guided HDR prostate brachytherapy in standard 1.5T scanner. Int J Radiat Oncol Biol Phys 59(5):1414–1423, 2004. 53. Merkle EM, Lewin JS, Aschoff AJ, et al: Percutaneous magnetic resonance image-guided biopsy and aspiration in the head and neck. Laryngoscope 110(3 Pt 1):382–385, 2000. 54. Merkle EM, Nour SG, Lewin JS: MR imaging follow-up after percutaneous radiofrequency ablation of renal cell carcinoma: Findings in eighteen patients during first 6 months. Radiology 235(3):1065–1071, 2005. 55. Merkle EM, Shonk JR, Duerk JL, et al: MR-guided RF thermal ablation of the kidney in a porcine model. AJR Am J Roentgenol 173(3):645–651, 1999. 56. Miller DC, Gruber SB, Hollenbeck BK, et al: Incidence of initial local therapy among men with lower-risk prostate cancer in the United States. J Natl Cancer Inst 98(16):1134–1141, 2006. 57. Reference deleted in proofs. 58. Reference deleted in proofs. 59. Nour SG, Goldberg SN, Wacker FK, et al: MR monitoring of NaClenhanced radiofrequency ablations: observations on low- and high-field-strength MR images with pathologic correlation. Radiology 254(2):449–459, 2010. 60. Nour SG, Hsu DP, Derakhshan JJ, et al: MRI-guided sclerotherapy of veno-lymphatic vascular malformations: evaluation of procedure safety and long-term efficacy. Proceedings of the International Society for Magnetic Resonance in Medicine (ISMRM) 19th Scientific Meeting, May 2011. 61. Reference deleted in proofs.

2384

PART III  Image-Guided Procedures

62. Reference deleted in proofs. 63. Reference deleted in proofs. 64. Nour SG, Lewin JS, Gutman M, et al: Percutaneous MR imagingguided radiofrequency interstitial thermal ablation of tongue base in porcine models: Implications for obstructive sleep apnea syndrome. Radiology 230(2):359–368, 2004. 65. Nour SG, Monson DK: MRI-guided musculoskeletal soft tissue interventions. Top Magn Reson Imaging 22(4):197–205, 2011. Review. 66. Nour SG, Nicholson AD, Powell TE, et al: MRI-guided and monitored laser ablation for renal malignancy: A step toward moving interventional MRI technology to mainstream usage? Milan, Italy: Proceedings of the Joint Annual Meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) and European Society for Magnetic Resonance in Medicine and Biology (ESMRMB). May 2014. 67. Orel SG, Schnall MD: MR imaging of the breast for the detection, diagnosis, and staging of breast cancer. Radiology 220(1):13–30, 2001. 68. Orel SG, Schnall MD, Newman RW, et al: MR imaging-guided localization and biopsy of breast lesions: Initial experience. Radiology 193(1):97–102, 1994. 69. Penzkofer T, Tuncali K, Fedorov A, et al: Transperineal in-bore 3-T MR imaging-guided prostate biopsy: A prospective clinical observational study. Radiology 274(1):170–180, 2015. 70. Perlet C, Schneider P, Amaya B, et al: MR-Guided vacuum biopsy of 206 contrast-enhancing breast lesions. Rofo 174(1):88–95, 2002. 71. Reither K, Wacker F, Ritz JP, et al: [Laser-induced thermotherapy (LITT) for liver metastasis in an open 0.2T MRI]. Rofo 172(2):175–178, 2000. German. 72. Ringe KI, Wacker F, Raatschen HJ: Is there a need for MRI within 24 hours after CT-guided percutaneous thermoablation of the liver? Acta Radiol 56(1):10–17, 2015. doi: 10.1177/0284185114520858. [Epub 2014 Jan 20]. 73. Roehl KA, Antenor JA, Catalona WJ: Serial biopsy results in prostate cancer screening study. J Urol 167(6):2435–2439, 2002. 74. Roemeling S, Roobol MJ, Kattan MW, et al: Nomogram use for the prediction of indolent prostate cancer. Cancer 110:2218–2221, 2007. 75. Rothgang E, Gilson WD, Wacker F, et al: Rapid freehand MR-guided percutaneous needle interventions: An image-based approach to improve workflow and feasibility. J Magn Reson Imaging 37(5):1202– 1212, 2013. doi: 10.1002/jmri.23894. [Epub 2013 Jan 18]. 76. Rubens DJ, Fultz PJ, Gottlieb RH, et al: Effective ultrasonographically guided intervention for diagnosis of musculoskeletal lesions. J Ultrasound Med 16:831–842, 1997. 77. Sack MJ, Weber RS, Weinstein GS, et al: Image-guided fine-needle aspiration of the head and neck: Five years’ experience. Arch Otolaryngol Head Neck Surg 124(10):1155–1161, 1998. 78. Scattoni V, Zlotta A, Montironi R, et al: Extended and saturation prostatic biopsy in the diagnosis and characterisation of prostate cancer: A critical analysis of the literature. Eur Urol 52(5):1309–1322, 2007. [Epub 2007 Aug 17]. 79. Schenck JF, Jolesz FA, Roemer PB, et al: Superconducting openconfiguration MR imaging system for image-guided therapy. Radiology 195:805–814, 1995. 80. Schmidt AJ, Kee ST, Sze DY, et al: Diagnostic yield of MR-guided liver biopsies compared with CT- and US-guided liver biopsies. J Vasc Interv Radiol 10(10):1323–1329, 1999. 81. Schnall MD, Orel SG, Connick TJ: MR guided biopsy of the breast. Magn Reson Imaging Clin N Am 2(4):585–589, 1994. 82. Schneider JP, Dietrich J, Lieberenz S, et al: Preliminary experience with interactive guided brain biopsies using a vertically opened 0.5-T MR system. Eur Radiol 9(2):230–236, 1999.

83. Schneider JP, Schulz T, Horn LC, et al: MR-guided percutaneous core biopsy of small breast lesions: First experience with a vertically open 0.5T scanner. J Magn Reson Imaging 15(4):374–385, 2002. 84. Schulz T, Schneider JP, Bootz F, et al: Transnasal and transsphenoidal MRI-guided biopsies of petroclival tumors. J Magn Reson Imaging 13(1):3–11, 2001. 85. Silverman SG, Collick BD, Figueira MR, et al: Interactive MR-guided biopsy in an open-configuration MR imaging system. Radiology 197:175–181, 1995. 86. Silverman SG, Jolesz FA, Newman RW, et al: Design and implementation of an interventional MR imaging suite. AJR Am J Roentgenol 168:1465–1471, 1997. 87. Sittek H, Perlet C, Herrmann K, et al: MR mammography. Preoperative marking of non-palpable breast lesions with the Magnetom open at 0.2 T. Radiologe 37(9):685–691, 1997. German. 88. Smith DS, Humphrey PA, Catalona WJ: The early detection of prostate carcinoma with prostate specific antigen: The Washington University experience. Cancer 80(9):1852–1856, 1997. 89. Susil RC, Camphausen K, Choyke P, et al: System for prostate brachytherapy and biopsy in a standard 1.5 T MRI scanner. Magn Reson Med 52(3):683–687, 2004. 90. Susil RC, Krieger A, Derbyshire JA, et al: System for MR image-guided prostate interventions: Canine study. Radiology 228(3):886–894, 2003. 91. Reference deleted in proofs. 92. Teifke A, Hlawatsch A, Beier T, et al: Undetected malignancies of the breast: Dynamic contrast-enhanced MR imaging at 1.0 T. Radiology 224(3):881–888, 2002. 93. Thiele J, Schneider JP, Franke P, et al: New method of MR-guided mammary biopsy. Rofo 168(4):374–379, 1998. German. 94. Tracz RA, Wyman DR, Little PB, et al: Comparison of magnetic resonance images and the histopathological findings of ablation zones induced by interstitial laser photocoagulation in the brain. Lasers Surg Med 13(1):45–54, 1993. 95. Vogl TJ, Muller PK, Hammerstingl R, et al: Malignant liver tumors treated with MR imaging-guided laser-induced thermotherapy: Technique and prospective results. Radiology 196(1):257–265, 1995. 96. Wacker FK, Reither K, Ritz JP, et al: MR-guided interstitial laserinduced thermotherapy of hepatic metastasis combined with arterial blood flow reduction: Technique and first clinical results in an open MR system. J Magn Reson Imaging 13(1):31–36, 2001. 97. Wacker FK, Vogt S, Khamene A, et al: An augmented reality system for MR image-guided needle biopsy: Initial results in a swine model. Radiology 238(2):497–504, 2006. 98. Wald DS, Weinreb JC, Newstead G, et al: MR-guided fine needle aspiration of breast lesions: Initial experience. J Comput Assist Tomogr 20(1):1–8, 1996. 99. Wang SJ, Sercarz JA, Lufkin RB, et al: MRI-guided needle localization in the head and neck using contemporaneous imaging in an open configuration system. Head Neck 22(4):355–359, 2000. 100. Woodrum DA, Gorny KR, Mynderse LA, et al: Feasibility of 3.0T magnetic resonance imaging-guided laser ablation of a cadaveric prostate. Urology 75(6):1514, e1-6, 2010. doi: 10.1016/j. urology.2010.01.059. 101. Zangos S, Eichler K, Engelmann K, et al: MR-guided transgluteal biopsies with an open low-field system in patients with clinically suspected prostate cancer: technique and preliminary results. Eur Radiol 15(1):174–182, 2005. [Epub 2004 Sep 04]. 102. Zangos S, Kiefl D, Eichler K, et al: [MR-guided biopsies of undetermined liver lesions: Technique and results.]. Rofo 175(5):688– 694, 2003. German. 103. Zhang Q, Chung YC, Lewin JS, et al: A method for simultaneous RF ablation and MRI. J Magn Reson Imaging 8(1):110–114, 1998.

67  Image-Guided Aspirations and Biopsies John R. Haaga, Timothy L. Haaga, and Hanping Wu

The use of image-guided interventional procedures has become ubiquitous in every hospital in the world. Although diagnostic imaging has improved the diagnosis of many disease processes and facilitated the implementation of patient treatment, interventional procedures have truly revolutionized the treatment paradigm. The treatment process can be initiated only after a definitive diagnosis is confirmed and the biopsy procedures provide the obligatory histopathologic diagnosis. With reported experience in the field, the specific roles of fine-needle aspiration (FNA) and core biopsies have been defined to permit optimal diagnosis of pathology and anomalies with very high levels of safety. Image-guided interventions have evolved into definitive treatment procedures such as fluid drainages and tumor treatment and ablation. Nonvascular image-guided interventions have continued to increase in recognition, frequency, complexity, and success rates and have finally gained their appropriate role in radiology practice. Although other subspecialties (vascular surgery, cardiology) have captured major portions of the vascular field, nonvascular image-guided interventions can remain part of the radiology field if radiologists incorporate them into their practice as an adjunct to diagnostic imaging; superior diagnostic skills are obligatory for superior image-guided interventions. In our opinion, image-guided interventions are the key for preservation of the radiology field as a subspecialty.180 Imaging advances such as multidetector computed tomography (MDCT), three-dimensional (3D) reconstructions, and computer-assisted diagnosis facilitate visualization of normal anatomy and enhance the diagnosis of disease for radiologists and nonradiologists alike, but radiologists are familiar with the subtleties of image interpretation required for high-quality interventions. The future of image-guided interventions is bright, but excellence must be fostered to protect the field of radiology. Although for the foreseeable future procedures will be performed manually, it is clear from initial robotic experiences that some type of robotic guidance will be forthcoming (Fig. 67-1). Magnetic resonance imaging (MRI) and ultrasound (US) devices continue to be refined and are likely to become more efficient and accurate for image-guided interventions. Future hybrid machines will combine digital images based on topographic images and 3D reconstructions to facilitate placement and confirmation of instrument location (Fig. 67-2). Radiologists must be in the forefront of such developments and applications.

GENERAL PERSPECTIVE In earlier editions of this book, it was challenging to find data and references to support our perspective on procedures, but now the challenge is to be selective among the many reports that are available in the field. Vast amounts of new information have been generated and

published on all aspects of image-guided interventional procedures. We are pleased that the accumulated knowledge validates many of the principles we and others have espoused over the years. Although many of the techniques have not changed, their clinical applications have expanded so that now, instead of using the histologic samples simply to diagnose a problem, chemical assays and biomarkers are being measured. These biomarkers permit better prognostication, predetermine the sensitivity of selective chemotherapeutic regimens (e.g., estrogen markers, human epidermal growth factor receptor markers, tyrosine kinase growth factor receptor sites), and permit DNA clonality when appropriate. The National Institutes of Health (NIH) is facilitating further application of biomarkers measured from biopsy samples by encouraging the routine use of image-guided large-needle biopsies  for biomarker assay by many new initiatives. It is hoped that special stains and tissue biomarkers will open the field of personalized patient care, which is contingent on earlier specific molecular or genetic diagnosis. Equipment and devices have become more refined and userfriendly, facilitating the performance and positive results of most procedures. To simplify the acquisition of such samples and lessen the inconsistency of manual needle techniques, a wide array of automated needles are available in the marketplace. Relative to cutting needles in general, a survey by Mayoral and Lewis showed that 95% of liver biopsies were performed by cutting needle “guns.”335

Use of Scan Time: Diagnostic Imaging Versus Procedures Some radiologists have looked at image-guided interventional procedures as time-consuming and not an important part of their practice. Financial considerations have been the driving principle for the bias not to use “scanner time” for such procedures. From a financial perspective, it is true that multiple diagnostic scans can be performed during the same time as several procedures, but this need not be the case if patient flow is organized. It is likely that interventional procedures will provide a mainstay for a radiologic practice. Time efficiency for performance of procedures can be improved by many factors, including experience, designated interventional individuals, support personnel, and defined and refined processes. Clearly the more experience one gains, the quicker the entire process occurs. If only a few individuals are designated as interventionalists, their judgment, skill, and operational efficiency increase. Selecting and training the appropriate support personnel, such as technicians, nurses, and physician assistants, improves efficiency remarkably. With appropriate assistance, these support personnel can assume responsibility for patient selection, performing the patient interview, obtaining informed consent, performing laboratory studies, obtaining a pertinent medical and drug history, and writing preprocedure and postprocedure orders.

2385

2386

PART III  Image-Guided Procedures y x

z

z

x

A

y

B

C FIG 67-1  A, Placement of a biopsy needle within an anatomic location. The relationship of the needle to the plane of the z axis and the plane of the x-y axis is illustrated. B and C, Current approaches have evolved far beyond earlier methods and even initiated developmental work with robotic assistance. B, CT scanner with robot mounted on the side of the table, placing the instrument in a swine model. C, CT scan taken after insertion of the needle by the robot to a site adjacent to the aorta. This corresponds to the celiac plexus. With the positive results of these tests, we are convinced that robotics will be used in future devices. See “CT Fluoroscopy, Guidance Devices, and Robotics.” (A, From Haaga JR et al: CT-guided biopsy. Cleve Clin Q 44:27–33, 1977.)

CHAPTER 67  Image-Guided Aspirations and Biopsies

A UP

B

D

C FIG 67-2  A, First CT radiographic image ever produced or reported. One of its original purposes was to aid with interventional procedures (see discussion under “Insertion of Needle and Catheter”). This primitive scan shows the lower chest and upper abdomen. This scan preceded all of the current commercial images, including the Topogram, Scoutview, and Synerview. B, With insertion of a wire, the wire moves posteriorly and cephalad. C, With a digital radiograph, the cephalic positioning of the wire at the highest level of the abscess (arrow) below the diaphragm can be appreciated. There is no need to traverse the pleura to drain these very high collections. D, Reconstructed 3D image with transparency and color treatment shows the remarkable visual effects possible. (D, Courtesy Dr. Damian Dupuy, Brown University, Providence, RI.)

2387

2388

PART III  Image-Guided Procedures

When a well-qualified assistant is selected and properly trained, such an individual can even begin a procedure and save considerable time by selecting the patient position on the table, selecting and preparing the skin site, and administering the initial local anesthesia. Although performance of interventional procedures may not increase referrals at all sites, these procedures can assist in maintaining referral bases. Most referring physicians appreciate how diagnostic biopsies facilitate patient care by providing a definitive diagnosis that permits initiation of treatment or an alternative therapy option in certain disease processes. Assisting in patient care improves the general reputation of radiologists among referring physicians and patients. The lack of such procedures in such a practice may erode the current referral pattern and result in referrals to other radiology groups or other competitive clinical groups. With computer-assisted diagnosis on the horizon, skillful image-guided interventional procedures require a higher level of interpretation than can be provided by a computer or other clinical groups, which in the future may entertain the possibility of image interpretation. Radiologists who are competent in image-guided interventions add unique value to the patient care process.

our extensive experience to radiologists who wish or need to be more proficient with CT procedures. The information and insights we provide are based on extensive long-term research and decades of clinical experience. We originated181 and have helped refine many of today’s CT-guided procedures181 and drainages (see Fig. 67-3).177,179,182,187 We are flattered that the official history of radiology by the American College of Radiology acknowledges that we innovated CT-guided biopsies, drainages, and treatments, and that such experts as former Surgeon-in-Chief at Massachusetts General Hospital Claude Welch noted our origination of CT-guided abscess drainages in the New England Journal of Medicine.509 Other contributions have included image-guided large-needle biopsies,189 radiation dose reduction for diagnostic and image-guided interventions,178 CT-guided radioactive seed placement,193 gas injection to move structures,184 percutaneous cecostomy,185 CT longitudinal scan (e.g., topogram, scanogram), CT-guided nerve blocks,188 and fibrinolytics to facilitate abscess drainage.192,278,371,372,399

History of Percutaneous Procedures Reviewing the historical perspective of interventional procedures can be confusing if one takes a narrow viewpoint and assumes that their evolution in the past 50 years has been along a single pathway. If one takes an objective global viewpoint, however, one can see that the development of percutaneous procedures has been varied and has involved four separate disciplines: (1) image guidance methods, (2) instrument development, (3) clinical procedure techniques, and (4) pathologic and cytopathologic techniques. Incremental advances in these four areas have occurred concurrently and at times separately. For individuals who have a limited personal experience or take a cursory review of the literature, it is difficult to appreciate the overall perspective. Many of the controversies related to needle sizes and types have been resolved in the intervening years since we first discussed them. The historical context we previously used to “defend” or encourage use of the biopsy instruments and drainage techniques is no longer necessary. Some discussion is still appropriate, however, because those who do not know history are condemned to repeat it, or some academicians may resurrect all concepts to build their careers until the concepts are disproved again.

Cost-Efficiency of Image-Guided Interventions Although the cost is greater than for other examinations, the cost savings of a CT biopsy compared with alternative costs of surgery, hospitalization, or both are remarkable. Mitty and colleagues calculated a savings for CT-guided pancreatic biopsies of more than $3000 per case.354 Similarly, Silverman and coworkers found a savings of $3149 per case.449 Considering that more than half of all such biopsies are performed on an outpatient basis, the savings is actually more than that amount. By comparison, US is much cheaper, and MRI is much more expensive.

Authors’ Experience and Perspective The senior author has performed CT procedures for more than 40 years (Fig. 67-3), encompassing about 30,000+ procedures, and has observed many changes in the field. With the refinement and expansion of the imaging field, all modalities, including CT, US, MRI, and traditional fluoroscopy, are used for interventional procedures. The roles of the various modalities have changed and will change in the future as new developments and refinements occur. Although it is true that the choice of image-guided procedures typically follows the radiologist’s own interest, expertise, and psychomotor skills sets, imagers should be familiar with the various modalities to enable the use of the best guidance modality for procedure success. This chapter imparts

A

Brief Historical Summary of Image-Guided Biopsies.  When biopsies were first performed in the 1930s through the 1960s, they  were performed at the bedside by clinicians without imaging. At that  time, the selected organ, usually liver or lung, was located by physical

B

FIG 67-3  A, First CT-guided biopsy was of a retroperitoneal mass (by the author). Although the results of this biopsy were negative, the merits of CT guidance were obvious. B, First CT-guided procedure of the lung in 1975 (by the author).

CHAPTER 67  Image-Guided Aspirations and Biopsies examination. Entrance sites were chosen using palpation and auscultation with a stethoscope. The skin was prepared with antiseptic, but local anesthesia was not always used. The biopsy was performed quickly during a single breath hold, the theory being that if the needle was inserted and withdrawn quickly enough the patient would not feel it or would not move. Laboratory measurements of hematocrit and coagulation studies were not used frequently. Complications were frequent. It is almost amusing in retrospect to reflect on two of the most important historical advances for unguided liver biopsies.343 First, it was discovered that the lateral intercostal approach was found safer than the anterior subcostal approach. With the clear knowledge of anatomy today, one reflects why they did not realize that the other contiguous organs, such as the stomach, gallbladder, colon, duodenum, and pancreas, could potentially be damaged by this path. Imaging guidance precludes this possibility. Second, the liver biopsy was proved safer when performed while the patient suspended respiration and motion. One eminent physician of the time, Menghini, gained great notoriety because he had the patient suspend respiration, and the needle was thrust in and out during 1 second.343 This practice prevented motion of the diaphragm and reduced the number of complications. His other great innovation was having a single physician apply suction with a syringe while inserting the needle; previously, one person applied suction by means of an attached tubing while the other inserted the needle. It is very strange to note that some physicians today are proposing the “old Menghini” two-person technique of the 1940s where two people perform a biopsy. One person does the needle work and the second provides vacuum with an attached plastic tube. (To each his own!) Relative to guided procedures, Lundquist was the first to show the usefulness of “guided” biopsies using US and FNA; the sample was cytologic.318 The US for guidance was the B mode (before real time), which was used to mark the skin surface; no direct visualization of the needle was possible. Lundquist’s reported experience showed conclusively that this method was superior to “blind” procedures.319 He also reported the use of the 18-gauge Menghini needle (end-hole cutting needle compared with the side cutting needles of today) for selection of the needle entrance site.319 In a series of 1000 biopsies with the Menghini needle and 2611 procedures with the “skinny” needle, there was a very low complication rate; one hematoma occurred with the skinny needle. For perspective, one should realize that in the 1970s, the imaging technology was such that only extremely large and numerous lesions could be visualized and even considered for biopsy.

FNA Versus Core Cutting Needles.  From today’s perspective, it seems inconceivable that any controversy about the effectiveness or specific benefit of FNA versus core cutting needles existed, but at the major meetings of the 1980s and 1990s, debates about skinny needles versus large cutting needles were extremely controversial and charged with emotion. Proponents of skinny needles used references from the unguided biopsies of the antiquated days of the 1950s and 1960s to counter the benefits of larger needle samples. With the unguided biopsies, a variety of core biopsy needles were designed and evaluated; a brief summary is important because invariably vendors reintroduce different permutations of these needles periodically using a different name. A prospective series in 1978 by Perrault and associates showed no complication difference among the VimSilverman, Menghini, and Tru-Cut needles.404 The comparison by Bateson and colleagues in 1980 of the Tru-Cut needle and the Menghini needle (end cutting needle) showed no diagnostic or other complications, although the physical integrity of the sample was more

2389

consistently intact with the Tru-Cut needle, especially with cirrhotic livers.28 Large cutting needles were not used routinely in imaging until the 1980s, when our group published numerous articles about the basic characteristics of needles,17 the diagnostic yield of larger needles,197 and the comparison of outcomes with skinny and cutting needles.329 In the 1980s, we were accused of “malpractice” for using such needles because of the “reputed” dangerous complications they posed for patients,238 and yet today we believe some physicians may overuse cutting needles and do not follow the guidelines similar to ours, which ensures a low complication rate. It is ironic, looking back at the early days of interventional procedures, that it took 2 intense years of reviews and negotiations to get our first paper on lung biopsies and cutting needles by Goralnik and coworkers160 published, and now such procedures are commonplace and perhaps overused (see “Lungs and Mediastinum” under the heading “Selection of Cutting Needle Type and Caliber,” later). We initially proposed the routine use of such needles in 1979. Since the 1990s, the superiority of cutting needles has been repeatedly documented and confirmed. Such large needles are used preferentially in all organ systems, with excellent results and acceptable complication rates. Having championed their usage, we remain enthusiastic, but we believe that careful choice of patients, lesions, anatomic approaches, and use of contrast material is important to balance the benefit and risk for each patient. With proper techniques, the diagnostic yield approaches 95% and the complication rate is less than 1%. Some physicians use cutting needles without appropriate patient selection and precautions (see later sections).

IMAGING PREFERENCES There was a period of about 25 years (1950-1975) when traditional x-ray fluoroscopy was used for aspiration biopsies and fluid aspiration procedures, but these procedures have for the most part fallen into disuse. In the 1970s, US29,168,226 and CT were introduced as guidance modalities. US was introduced as a biopsy guidance system by Holm and colleagues in 1975.224,225 CT guidance was introduced by Haaga and Alfidi in 1976.181 MRI procedures were pioneered by numerous authors, most notably Lewin of our group, who reported many animal studies and human trials (Fig. 67-4).296-298 This chapter provides insights and details for CT procedures. The reader is referred to other texts and articles for more details regarding US or MRI procedures. The experienced imager should be familiar with all major imaging systems and should have at least passing familiarity of the various techniques and preferably hands-on experience for procedures. One should strive to become expert with any modality used for procedures. The major general tenets for choice of guidance modality are as follows: 1. Act on the behalf of the patient and use the modality best suited to achieve favorable results. 2. Choose the modality for the right reason—improved visualization of the lesion or equipment—not because it is the “easiest” or most convenient for other reasons. 3. Minimize risk in all circumstances. 4. Never let ego determine the course of the procedure; never hesitate to terminate a procedure if things are not going well. 5. If the patient has significant comorbidities, do not push the limits unless it is critical for the patient’s survival; adverse events may occur from a comorbidity, and if it occurs, the procedure might be inappropriately blamed.528 6. If general anesthesia is needed or the lesion is unsuitable, cancellation is better than a complication.

2390

PART III  Image-Guided Procedures

A

B

C

D

FIG 67-4  A, Turbo-spin echo T2-weighted axial image shows a 1-cm metastasis near the dome of the liver. B, Turbo-spin echo T2-weighted coronal image shows the RF electrode, placed under MR guidance, with its tip centered within the tumor nodule (arrow). C and D, Axial (C) and coronal (D) contrast-enhanced T1-weighted images at the end of the ablation session show a hypointense zone corresponding to the treated metastasis. The lesion was treated to ensure complete ablation of the tumor plus a thin surrounding margin of normal tissue.

It is reasonable to combine the advantages of several imaging systems for procedures, as has occurred with diagnostic equipment. If an operator is more comfortable with US than CT or vice versa, he or she may wish to combine the two as appropriate. US can show realtime placement of an instrument conveniently, whereas CT is unparalleled in documenting precise needle placement. Combinations with MRI and US and CT and positron emission tomography (PET) are likely for procedures.

CT

Advantages.  CT uses a beam of radiation to penetrate virtually all natural and artificial materials within the human body. Information concerning attenuation of these substances can be measured and an image can be reconstructed. Because of the wide latitude of imaging capabilities, an extremely low-density material (e.g., gas) can be imaged, as can high-density materials such as metal, bone, or synthetic catheters (Fig. 67-5). In addition, CT is sensitive enough to detect subtle abnormalities such as fluid densities, cysts, or gas-containing small pockets or fistulas (Fig. 67-6).195,196 The ability to image high-density material accurately offers wide versatility so that any type of instrument (i.e., needle, drainage tube, or other device) may be used in a procedure. Contrast material can be

FIG 67-5  RFA of kidney tumor shows the high-density probe with individual tines clearly visible. Small gas bubbles are present in the tumor site from the heating effects. Such bubbles are not visible on MRI or US (see text).

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2391

B

FIG 67-6  Patient with Crohn’s disease and abscess. A, Scan shows a needle being inserted in the abdominal wall abscess and a subtle air-filled fistula between the colon and the abscess (arrow). Visualization permitted placement of the catheter adjacent to the fistula to improve drainage and resolution. B, Repeat scan shows a catheter close to the fistula into the bowel.

injected when vascularity, an anatomic space, or an abnormal cavity needs further delineation. An abscess can have unpredictable serpiginous tracts or communicating channels that can be difficult to perceive unless contrast material is used (Fig. 67-7). Because the CT unit uses an air interface, it consistently provides detailed and complete image even in patients with skin issues. In patients who have draining sinuses, wounds, ileostomies, scars, or other unusual surface problems, the examination can be performed without difficulty. Another advantage is that the longitudinal scan (e.g., Topogram, Scoutview) that we originated183 is now available on all CT scanners, making it possible to obtain conventional radiographs of the anatomy and to visualize instruments or catheters during insertion or manipulation. Finally, real-time CT fluoroscopy can be used for cases of very small mobile masses or abnormalities. With the evolution of patient referral, lesions sent for biopsy have become smaller and smaller, testing the skills of even the most experienced radiologist. Accuracy of localization.  The potential accuracy of the CT-guided method can best be appreciated by examining a schematic drawing of a cross-sectional image (see Fig. 67-1). If the instrument is placed within an anatomic area and the CT scan shows the needle in that area, the needle must lie within the 10-mm section of the abdomen corresponding to the CT section. Location of an instrument in this plane is limited by the section thickness, which is 10 mm; the resolution in the z axis is +5 mm. If one uses narrower slices, such as 5 mm, the accuracy in the y-z plane is +2.5 mm. Resolution of the CT system in the x-y plane is less than 1 mm with modern scanners. The precision required to determine the position of such an instrument is less limited by the resolution of imaging equipment than it is by the degree of the operator’s accuracy, as determined by manual dexterity and by the degree of accuracy of the guidance system, rather than by the imaging equipment. Because of the increased detection of smaller lesions, incidental findings must be managed, and a definitive biopsy would be the best alternative; robotic assistance that positions the device during realtime fluoroscopy would be required. The ability to localize the needle tip consistently and accurately in an axial plane is a distinct advantage of CT over fluoroscopy or US. Although fluoroscopy can provide multiplanar visualization of a needle or instrument, it is incapable of directly visualizing soft tissue abnormalities except in the chest. Fluoroscopic localization of instruments depends on indirect information obtained from displacement of contrast-filled structures (e.g., bowel, biliary).

A

B

FIG 67-7  A, Patient with presacral fluid collection after surgery and radiation. The cavity has a subtle channel adjacent to the rectum, suggesting communication. Diagnostic needle aspiration was performed for culture and delineation of fistula tract by contrast injection. B, Patient was rolled slightly to enhance the flow of contrast material into any existing path. Note the contrast material entering the margin of the rectum.

The special advantage of CT guidance is that visualization of the needle is a foregone conclusion, and special technical approaches or concerted efforts are not required to visualize the needle. CT is not as sensitive for subtle lesions, however, and occasionally the combination of MRI and CT guidance can prove worthwhile (Fig. 67-8). Versatility.  Versatility is still one of the most valuable advantages of CT-guided procedures. There is significant versatility in selecting the entry point for the biopsies because of the completeness of the CT image: a full 360-degree view of the patient is provided. All organs, including the bowel, are visualized so that a variety of needle pathways can be chosen. It is possible to choose appropriate pathways to avoid specific organs or the bowel if clinically indicated. Changing the body

2392

PART III  Image-Guided Procedures

A

C

B FIG 67-8  A, MRI with contrast material shows mass in medial right lobe segment 7 (arrow), with slightly increased signal and rim of edema suspicious for tumor. B, CT scan shows a very subtle lesion difficult to see clearly. Using landmarks, the site was localized. C, Biopsy of the subtle tumor mass (arrow) yielded hepatocellular carcinoma.

position of the patient may be advantageous for opening windows for instruments because of organ shifting or patient comfort (changing positions creates no adverse effect on image quality). Finally, a wide variety of instruments can be used, ranging in size from a 23-gauge Chiba needle to a 14-gauge Tru-Cut needle. Of all of the advantages of CT procedures, versatility is the one that is least recognized. Varying the patient position is useful for patient comfort or for planning and execution of the procedure (Figs. 67-9 to 67-11). As noted in the subsequent clinical sections, three major benefits are

derived. First, if patients are going to cooperate during a procedure, it is crucial that they are comfortable on the table. Patients with back pain or localized discomfort in the flank can be induced to remain still during a procedure if they can become comfortable by shifting their position. Permitting the patient to lie in a prone, lateral, or oblique position can go a long way toward improving patient comfort and cooperation and success of the procedure. The second benefit of moving patients into different positions is that the internal or external anatomy shifts accordingly and permits

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2393

varying the patient position slightly, it may be possible to produce movement of a loop of bowel, permitting an anterior biopsy without traversing bowel (Fig. 67-12). In other cases, one can approach retroperitoneal lesions posteriorly through the flank instead of traversing bowel or other structures by using an anterior approach. Although traversing bowel is seldom a problem, in selected cases, secondary infection may develop from a microcontamination (Fig. 67-13). Occasionally it is beneficial with simple and complex procedures to perform 3D imaging with axial, coronal, or variable planes. For unusual biopsy procedures or ablations, it may be beneficial to determine the exact location of a cutting needle before acquiring the tissue sample (Fig. 67-14). In complex drainages, such as rectal drainages (Fig. 67-15), 3D imaging is well suited because it is difficult to ascertain the precise trajectory of the needle when the approach is made through the rectum. Also, the precise placement of the catheter in the abscess cavity can be confirmed. For the new ablation treatments, 3D rendering is especially important because accurate planning and execution of a treatment plan  is crucial for consistently successful procedures. Accurate placement  is necessary to ensure that all portions of the tumor are treated and there is adequate overlap of the treatment areas. It is of further benefit to confirm that no adjacent uninvolved anatomy is being damaged (Fig. 67-16).

Disadvantages.  CT guidance has three disadvantages that must be

B FIG 67-9  A, CT scan shows a small mass posteriorly in the right upper lobe. With the patient supine, there is no pathway clear of intervening structures such as the ribs and scapula. An approach through the axilla is not possible because of the axillary artery and vein, which are anterior to the small metal suture. B, The patient was repositioned in an anterior oblique position, which permits needle insertion through a clear pathway posteriorly.

better and safer trajectory planning and execution. For chest procedures, such movement may permit displacement of the scapula, making a posterolateral approach easier (see Fig. 67-9). Within the chest, lesions in the anterior mediastinum can be sampled more safely because positioning patients on their side may shift the anterior junction line sufficiently to permit a pathway avoiding the lung parenchyma (see Fig. 67-10). In the abdomen, bowel and organs shift sufficiently to permit a safer sampling procedure. Third, in some cases when a patient has difficulty taking reproducible breaths and the lesion is mobile, placing the patient in a lateral or oblique position can help in limiting the movement of the lesion. When the ipsilateral side of the patient is dependent, the weight of the viscera partially immobilizes the diaphragm and limits movement of the lesion (see Fig. 67-11) even when it is in various precarious sites, such as adjacent to the spleen. It is perplexing that some radiologists do not change the patient position or trajectory of the needle to avoid uninvolved structures. By

considered when planning image-guided procedures: (1) use of radiation, (2) less sensitivity for imaging primary liver tumors, and (3) less sensitivity for imaging musculoskeletal tumors. The CT scan uses radiation for generation of the scans, but the dose delivered to the patient is much less than with other specialized studies, such as angiography. With modern MDCT scanning devices, the detector systems are remarkably efficient, using almost 100% of the dose. Although excessive doses should be avoided in all circumstances, a single scan has a sufficiently low level that in cases of acute illness that may jeopardize a patient (e.g., appendicitis) a CT scan and procedure may be justified. In the more recent generation of scanning devices, as a result of increased awareness about radiation safety, vendors are producing “photo-timed” CT scanners. The detector systems are being used in conjunction with the computers to measure the dosage “on the fly” so that optimization of dosage occurs even within the context of a specific scan. I cannot avoid the opportunity to say, “I told you so and it’s about time,” noting that this was proposed many years ago in our second edition of the textbook, after our first article on CT dose reduction in 1981190 and subsequent recommendations to use the lowest possible dosage for image-guided procedures.188 There is a definite lower detectability rate of tumors in the liver and muscle tissues. Because primary tumors of the liver are sometimes difficult to see, one should perform procedures in conjunction with real-time US or reference an available MRI scan (see Fig. 67-8).

Ultrasound Advantages

Nonionizing radiation.  One significant advantage of US is that the sound waves used are not dangerous to living tissues in the amount and wavelength used. Repetitive examinations are not an issue. The overuse of CT scanners for diagnosis has raised radiation dose issues for radiologists, who must be cognizant of the radiation issue related to repetitive scans, 3D reconstructions, and the tendency to overuse MDCT scanners. Instrument localization.  US has gone through several periods of great enthusiasm as technology has improved. In the early years of interventions, the metal instruments were difficult to visualize because

2394

PART III  Image-Guided Procedures

A

B FIG 67-10  A, CT scan shows a subtle mass in the anterior mediastinum (arrow), thought to be thymus, but the clinical situation dictated that a biopsy be performed. The anterior junction line is concealed behind the sternum. The only possible biopsy approach would be a very lateral one traversing much lung parenchyma. B, With the patient lying on the right side, the mediastinum shifted so that a clear access to the mass without crossing the lung was provided (arrows).

the needle completely reflected the US beam, preventing sufficient return of echoes to permit visualization. The sensitivity of new US equipment is so great that it is easy to visualize the tip and pathway of needles in the range of 20 to 18 gauge (Figs. 67-17 and 67-18). The very slight motion of the needle during insertion is well visualized, permitting the “trail” of the needle to be seen clearly. Alterations in the “surface” characteristics of the needle sides and the tip show the end of the needle well. Any tiny gas bubble that is adherent to the end is displayed well.

Disadvantages.  There still remains significant difficulty in special circumstances when bone or gas is present either before or during a procedure. Although US may be used for single or small instrument placement for radiofrequency ablation (RFA) or cryoablation, it is limited when numerous probes must be inserted and localized. Assessing treatment in such cases is difficult because the iceball or gas generated may not be completely assessed or imaged.

MRI

Advantages.  MRI advantages relate to its great sensitivity for displaying subtle changes in tissue characteristics, permitting instrumentation. This sensitivity enables visualization of subtle abnormalities not seen as clearly on CT or US (see Fig. 67-8); this is frequent with primary liver mass lesions and some cystic tumors of the pancreas. MRI can survey such nodules and determine which are most suspicious for change to neoplasm. Differentiation of dysplastic nodules compared with hepatocellular carcinomas usually can be made by quantitating T2 differences (see Chapter 43). Also, when ablative procedures are performed, the “real-time” damage or change to tumor tissue can be determined by change in the signal. Another example of superior diagnostic sensitivity is in pancreatic tumors (see Chapter 46).

MRI is remarkably superior to CT for the detection of subtle islet cell tumors and the details of some cystic tumors of the pancreas. Because of these qualities, MRI is well suited in selected cases to follow the progress of the treatment procedures. It has been well documented that MRI is the most accurate modality for displaying the tissue changes from RF treatment, as noted by Lewin and coworkers297 and Merkle and colleagues344-346 in numerous reports (Fig. 67-19). Experimentally, MRI has been able to display characteristics correlating with absolute tissue temperature. With the current emphasis on treatment, it also may be possible with higher-field magnets having broadband capabilities to visualize drugs in situ as they circulate within the tissues and the target molecules. It is not illogical to assume that in the future, these qualities will be exploited by interventional strategies, most likely therapeutic approaches, if issues related to accurate and convenient instrument placement can be resolved. Another advantage of MRI is its ability to image blood vessels and to assess vascularity as it relates to general blood flow. MRI during standard pulsing sequences shows vessels even without contrast material. Moderate-sized vessels are clearly seen as a flow void. Although the anatomic spatial resolution of MRI is less than that of CT, it can also display specific blood flow characteristics, as one might see in hemangiomas, using gadolinium-enhanced dynamic contrast enhancement (DCE). An indirect advantage is that gadolinium has a much lower idiosyncratic rate than iodinated contrast material, so that patients who are allergic to traditional contrast material used with CT can be imaged and evaluated by MRI flow studies. Because of the proven efficacy of MRI-guided procedures, we have included a complete chapter on the topic in this sixth edition. Chapter 66—“MRI-Guided Interventions” by Dr. Sherif Nour—provides an excellent comprehensive review, and the subject will not be discussed further in this chapter.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2395

A

A

B FIG 67-11  A, CT scan shows a patient lying in the left decubitus position to partially immobilize the small renal mass, proven to be an oncocytoma, adjacent to the spleen. B, Scan shows placement of a small cutting needle in the small mass.

B FIG 67-13  A, Patient with a mesenteric mucoid mass that was biopsied using a transenteric approach. CT scan shows mass in peritoneum (arrow), which was biopsied with the needle inadvertently, traversing the bowel loop. B, After several days the patient developed a severe fever and elevated white blood cell count. CT scan shows air in the mass indicating secondary infection.

FIG 67-12  CT scan shows the choice of a pathway that avoids adjacent loops of bowel, permitting cutting needle biopsy of the mesenteric mass.

2396

PART III  Image-Guided Procedures

B

A

D

C

E FIG 67-14  A, Reconstructions can be worthwhile for assessing the location of masses and determining the angulation required. This axial scan shows a central lesion very high in the liver. B, Sagittal image shows the lesion close to the diaphragm. C, Coronal image shows the very high location of the lesion and most importantly the insertion of the diaphragm. By using an entrance site in the midaxillary line, the pleural space can be avoided. Although some authors have used the transpleural approach for high liver lesions, there is a higher incidence of pneumothorax and possible tumor seeding of the pleura. The literature regarding traversing of the diaphragm is comforting, but the best approach is to avoid such issues. D, CT scan shows the lowest insertion of the needle angled cephalad. E, Scan shows the tip of the needle at the margin of the lesion.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2397

FIG 67-15  Sagittal view during rectal drainage of abscess shows the needle directed into the pelvic abscess.

A

B FIG 67-16  A, Renal ablation shows two probes within a tumor mass. Because of the four probes placed, it is difficult to give a total perspective of probe placement in the mass. B, Reconstructed image shows the four probes properly positioned within the mass, confirming correct placement of the treatment probes for cryoablation overlap.

FIG 67-18  US showing placement of wire and catheter (arrow) into the pelvic abscess.

FIG 67-17  Real-time US showing needle puncture of the gallbladder (arrow) during cholecystotomy.

2398

PART III  Image-Guided Procedures

Readout gradient Readout gradient

B

A FIG 67-19  These two sets of images show the importance of properly orienting the readout gradient to the alignment of the needles. A, This image shows various-sized needles within an apple with the gradient parallel to the needles. Note how visualization of the needles is lessened by the apparent thin diameters of the needles. B, This set of images of the same needles in the same apple shows the diameter of the needles as being very large, enhancing their visualization. Localization of the needle tip is more accurate with the gradient perpendicular to the orientation of the needles than with the gradient parallel. This was well shown by Lewin and colleagues.297

PET Scanning Because of the ability of PET to image actively growing tumor, it is extremely valuable for targeting viable tumor for biopsy. Generally, any tumor that is “avid” for glucose appears hot and can be localized on associated CT images. The avidity of glucose is due to the hypermetabolism and increased level of glucose transport molecules. PET is of great benefit for visualizing virtually all viable tumor masses, and with the ubiquitous nature of PET, it is likely that most patients coming for biopsy will have such a study. The exceptional value of PET is threefold: (1) to provide overall assessment of disease extent to identify suitable targets for biopsy (Fig. 67-20A), (2) to differentiate viable versus nonviable tumor and central necrosis, which does not contain diagnostic tissue and should be avoided (see Fig. 67-20B), and (3) to perform postablation assessment of tumors. PET has proven to be the modality of choice for follow-up of lung ablation lesions379,388 and to be of great importance for liver ablations (see later under “General Techniques for Tumor Ablation”).502 Several organ sites and categories may be problematic, related to the nature of the primary tumor, tumor metabolism, and inflammatory sites. Primary tumors of the liver and kidney are variable and may not consistently be avid for fluorodeoxyglucose (FDG). A problem with PET is that it is occasionally difficult to differentiate inflammatory sites from tumor unless special techniques are used. The literature indicates that inflammatory tissue is avid for glucose in a similar fashion as cancer because of the enhanced blood flow due to dilated inflamed vessels and elevation of the glucose transporter molecules. There are two methods to differentiate inflammatory sites from neoplastic sites; the delayed scanning technique is described in Chapter 53, and the second is a new method that we have devised but not yet published. The first method reported consistently in the literature is to take delayed image acquisitions 2 to 3 hours after the injection of FDG.

Tumors typically retain their increased high uptake, whereas inflammatory sites reduce their uptake. The second method is a new method to temporarily reduce the inflammatory vascularity by administering an over-the-counter (OTC) U.S. Food and Drug Administration (FDA)-approved oral decongestant before injecting FDG. Although this is not an FDA usage of an approved drug, this technique is legal using the guidelines under the Belmont report, which appears on the FDA official website. We devised this new technique to reduce the incidence of false-positive PET scans sent for biopsy. In a group of nine patients with positive PET studies who presented for biopsy, we altered the appearance of the CT scan as follows. Recognizing that inflammatory sites respond to oral decongestants, we administered a single oral dose of an OTC FDA-approved decongestant 30 minutes before injection of the FDG. The medication we used was either pseudoephedrine or phenylephrine in the recommended OTC dose. Uptake of the FDG was sufficiently reduced to dispel the concern about malignancy, and the procedures were deferred (Figs. 67-21 and 67-22). We found that this technique worked in four of nine cases, eliminating the false positives from inflammation without changing the positivity of the cancers scanned. Further studies are required, but this seems to be a worthwhile technique for the future. Although this technique is not approved by the FDA as yet, the patents have been filed for the method, and we are hopeful that a company will partner with us to bring this new technique to the marketplace and general usage.

Future of Combined-Modality Imaging In recent years, devices that combine PET and CT within the same piece of hardware have been built. As one might expect, these devices offer great potential for enhancement of interventional procedures. The functional aspects of PET permit better visualization of viable tumor within mass lesions to permit more accurate and reproducible biopsies and treatments.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2399

A

B FIG 67-20  A, PET study shows multiple locations of lymphoma (arrow), providing a geographic distribution for possible target selection. B, PET study shows a necrotic center (arrow) indicating nonviability. Any biopsy should avoid the necrotic center.

Health care will become more personalized with the standardization of biomarkers, made possible from the development of new ligands related to reporter genes, proteomics, enzymatic metabolic pathways, and cellular processes. Specific traces will permit assessment and treatment of disease processes, eliminating the need for some diagnostic biopsies, but it is likely that sometime during the course of such evaluations and treatments, biopsy to recurrent target tissues will still be needed. It also is very likely that image-guided procedures will be used for delivery of treatment agents (i.e., chemicals, antibodies, or modulators).

INDICATIONS FOR VARIOUS PROCEDURES Indications for CT-Guided Procedures The indications for CT-guided procedures vary from institution to institution, depending on the expertise of the radiologists and the available imaging modalities. Generally the important factors to consider are the visibility of the lesion on CT compared with other modalities, the type of pathologic sample required, and location of the lesion. Considering that the prime function of the guidance modality is to direct the needle to the lesion, visibility of the lesion is the most obvious factor. If one cannot see the lesion, one cannot biopsy the lesion. If a lesion is seen only on CT or much better on CT, CT is the

modality of choice. Conversely, if a lesion is seen only on US or MRI, those modalities would be the preferred guidance methods. When an aspiration biopsy is clinically indicated, various factors should be considered. The simplest modality that satisfactorily shows the abnormality, the pathway of the instrument, and adjacent anatomic information should be used. In most instances, anatomic areas such as the porta hepatis, pancreas, retroperitoneum, and pelvis are best suited for CT guidance. Specific considerations of each anatomic area are discussed later. When lesions are small (3-5 cm) and located in critical areas, CT is the preferred modality (Figs. 67-23 to 67-25). CT puncture is best performed to assess abnormalities in critical areas such as adjacent to major vessels, the hilum of the spleen, and the mediastinum. Visualization of such small lesions may be attempted first using other modalities, but most authors agree that successful punctures on the first attempt are more likely with CT guidance. Any procedure that fails with other modalities should be attempted using CT. CT should be used when avoidance of intestinal loops or vascular structures is important (see Figs. 67-24 and 67-25). We believe such avoidance is advisable in most patients but especially in several situations. In immunologically compromised patients, CT procedures should be used to avoid penetration of the bowel and possible contamination (even with the skinny needle). Even in immunocompetent

2400

PART III  Image-Guided Procedures

A B

C

D FIG 67-21  Male patient with recurrent squamous cell carcinoma in left lung. A, PET scan with FDG with standard technique shows a large area of intense uptake in the left chest. Note eccentric uptake lateral to a large mass, corresponding to the needle location in B. B, CT scan of the left lung shows a large mass (arrow) adjacent to the left pleura. It appears to show a mass with adjacent infiltrate medially, but the lateral portion appears to be solid tumor. The biopsy needle is seen at the edge of the mass, corresponding to the edge of intense area on PET scan. Biopsy was negative for cancer. C, Repeat FDG PET following oral pseudoephedrine. Note the reduction in size of the overall mass, with resolution of the small focal hot spot noted in A. From this scan, it was decided that the tumor was more medial in location, which was used to plan second biopsy. D, Repeat CT scan taken during the performance of the repeat biopsy. Note that the needle tip was positioned more medially in the mass. Pathology was positive for squamous cell cancer.

patients, if the abnormality to be punctured is a fluid density, penetration of the bowel should be avoided if possible, to prevent pyogenic contamination. Although the incidence of clinical infection after microcontamination occurs is rare, it does occur (see Fig. 67-13). Three approaches can be used to move bowel: variation of patient position; manual compression of the abdomen; and injection of carbon dioxide, air, or fluid such as saline (see later under “Intentional Organ Movement or Tissue Retraction”). When a large cutting needle biopsy is clinically indicated, CT and US are the modalities of choice because of the clarity of needle visualization and the accuracy of real-time monitoring with US. State-ofthe-art CT permits optimal visualization of the needle, assessment of vascularity with dynamic bolus scanning, and real-time CT fluoroscopy when needed. Virtually all authors accept dynamic bolus scanning as being highly accurate for assessment of vascularity, whether in

the liver or other areas. These bolus scans can detect excessively vascular lesions, which may have a propensity to bleed if biopsied. Occasionally, remarkable unexpected vascular anomalies are detected, which can avoid catastrophic accidents (Fig. 67-26). CT also may be the modality of choice when some type of contrast agent, such as urographic contrast material or air, is to be injected. Such contrast can provide additional information about cysts, abscesses, pseudocysts, or dissemination of medication.

Indications for US-Guided Procedures The main indications for US are the clear visualization of the abnormality on US and the need for real-time guidance of the procedure. Although US is not as accurate for visualization of the needle, it is effective for real-time monitoring of needle placement. This is a judgment call by the radiologist, who must consider the proximity of

CHAPTER 67  Image-Guided Aspirations and Biopsies

B A

D

C

FIG 67-22  Female patient with history of hepatitis C. A, MRI scan shows a large solid mass (arrow) in the porta hepatis suspected to be a metastatic lymph node. Because of the location, percutaneous biopsy was not possible. B, PET image obtained using standard technique shows a large area of intense uptake (arrow) in the porta hepatis. C, PET study performed 5 days later with oral ingestion of pseudoephedrine 60 mg (oral phenylephrine 10 mg may also be used) before FDG injection. Scan shows an absent area of uptake. D, CT scan 6 weeks later shows resolution of lymph node mass, supporting the premise that the mass was inflammatory. The patient declined surgery.

M

A

B

FIG 67-23  A, Scans adjacent to a mass (M) in the tail of the pancreas shows the close proximity of varices (arrow) in the gastrosplenic ligament. B, With the patient in an oblique position and an angled approach, aspiration needle biopsy of the pancreatic mass avoiding the colon, spleen, and varices was accomplished. In contrast to other authors who recommend traversing uninvolved organs, we recommend that one at least attempt to avoid them. In most cases, penetration can be avoided. Over a large series of procedures, the absolute number of complications would be less because at least a relatively smaller number of inadvertent penetrations would occur.

2401

2402

PART III  Image-Guided Procedures

A

B

FIG 67-24  A, Small lesions in deep locations can be sampled using only CT. Scan shows a tiny mass in the root of the mesentery. This first pass shows the needle deflecting past the mass. The patient was positioned obliquely, which permitted the bowel to float away slightly, clearing a pathway for the needle. B, Repositioning of the needle and using the bevel of the needle to prevent deflection, the mass was sampled adequately. See discussion under “Intentional Organ Movement or Tissue Retraction.”

benefits all parties involved, including patients, physicians, and administrators.

POTENTIAL COMPLICATIONS Some complications are specific to the various organ systems, but some general concepts are relevant to all areas. The use of CT-guided procedures has significantly reduced the possibility of complications. Assuming that the correct CT technique has been selected, the major causes of complications are limited to poor technique, lack of patient cooperation, undetected coagulopathy, or undiagnosed obstructive or diffuse interstitial processes in the lung. FIG 67-25  Lymph nodes adjacent to aorta are suited for CT guidance only because of the accuracy required and difficulty with visualization of such areas with MRI or US.

crucial structures, the need for precise visualization of the needle, and vascular enhancement issues. Knowing that the skill level at different institutions varies, when there is a high level of experience, US may be used almost interchangeably with CT when the aforementioned factors are not critical.

Impact of Greater Experience Greater personal experience translates into better results regardless of the imaging modality or technique used. The outcome and complication rate are better for an experienced operator than for an inexperienced operator, regardless of whether biopsies or ablations are performed. This is true for different ablation methods, needles, and organ systems. Also, if one looks at the reports in the literature, one usually finds an improvement in the authors’ results as their reported experience accumulates. The improvement of diagnostic accuracy with experience was documented by Evander,117 Zornoza,533 and Hopper227 and their colleagues. Similarly, with ablations, Poon and coworkers409 documented that after experience has been gained, the ablation success rate increases and complication rate decreases. Because of the importance of experience, it is prudent for departments to designate a limited number of individuals who perform procedures so that greater expertise can be developed. This practice

Needle Complications Before discussing in detail how one can avoid certain problems, several comments are warranted concerning complications as they relate to the two major categories of instruments: the aspirating, or skinny, needles and the large cutting needles. Contrary to the opinions of some, severe complications may occur with any needle (including skinny needles, manual cutting needles, and automated biopsy guns) unless one employs caution, common sense, and meticulous technique.

Skinny Needle Complications.  There is no question that the skinny needle is the safest instrument available, but even this needle can produce complications when proper patient selection and techniques are not used. In a careful review of reports of skinny needle biopsies of the pancreas, one can find instances of pancreatitis, pancreatic fistula, cholangitis, needle-tract seeding, significant hemorrhage, abscess formation, and fatal pancreatitis. Such complications were usually the result of incorrect patient selection or choice of anatomic pathway and an excessive number of needle passes. Each organ system has specific issues, as noted and explored subsequently under each of the organ system sections below, but general factors can be mentioned. Improper patient selection and evaluation of anatomy and vascular properties have produced numerous hemorrhagic complications and deaths. Reported fatalities from liver and adrenal hemorrhage occurred with vascular hemangiomas, hemangiosarcoma,218 severe coagulopathy,112 and pheochromocytoma.339 One can see in reviewing the literature that most complications have occurred in cases with four or five needle passes.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2403

v M

BV

A

B L

C FIG 67-26  A, Noncontrast scan of a veteran sent to our institution for a “lymph node” (M) biopsy. The large mass to the right of vena cava was thought to be a lymphoma mass; a 14-gauge cutting needle was ordered. B, As dictated by our protocol, a bolus dynamic scan was performed that showed a large enhancing varix (BV) fed by smaller varices (v) in the right lower quadrant. Biopsy was not performed. C, Digital subtraction image shows large tubular varix (arrows) flowing toward the liver.

The impact of numerous needle passes with skinny needles on complications was reported by Smith of the University of Massachusetts.459 Smith reported a survey sent to 100 university hospitals and every fifth hospital in the United States, which solicited information about complications and risks from FNA biopsies. In this report, he noted the occurrence of 4 deaths, 27 cases of hemorrhage, 3 occurrences of needle-tract seeding, and 16 cases of generalized infections (nonfatal). Smith stated, “It is logical to assume that the number of complications should be related to the number of needle passes made during the biopsy procedure. However, serious and even fatal complications, although rare, can and do occur and it is important to be aware of the possibility and take all the appropriate precautions in order to reduce their incidence.”459 Even with extensive experience and caution, hemorrhage may occur with FNA aspirations (Fig. 67-27). In most instances, such events are rare if proper methods are used.

Cutting Needle Complications.  One has a greater probability of incurring possible complications with large needles because a large “hole” is being produced. With proper patient selection, proper target and trajectory choice, and meticulous technique, however, the potential risks of these needles can be offset, and very low complication rates can be achieved. Considering the previous controversies in the field about using large needles, we have some concern that such needles are used too frequently by some groups. One can use very

large needles safely with good results, but one must pay careful attention to meticulous technique, or serious and fatal complications can result. The new automated cutting devices are no safer than the standard “manual” Tru-Cut needle that has been used for many years. Some authors have asserted that the rapid speed of the cutting action may lessen local damage, but this concept is nonsensical. In the first report on the biopsy gun, the authors described a serious hemorrhagic complication after a liver biopsy. Potential problems are minimized by CT guidance and a careful, meticulous technique such as I have described. These measures can eliminate many potential causes of complications (e.g., hemorrhage or inadvertent penetration of an uninvolved organ) and can result in excellent accuracy with few problems. If one uses a large-core biopsy needle with an automated mechanism, it is best to avoid the single-action devices and use instead double-action devices. Double-action devices permit confirmation of the cutting gap in the anatomy before tissue cutting occurs. Critical biopsies can be performed safely because the device cuts only where the “gap” is seen on the scan. If a patient wiggles or moves slightly, the location of the needle and the cutting action are not altered. Singleaction devices “cut” the tissue with a single action by the operator; if a patient moves, the anatomy may move a vital structure into the cutting site, and catastrophe may result (see “Lung and Mediastinum” for an example of a patient in another institution who died because the patient moved slightly in such a case).

2404

PART III  Image-Guided Procedures

A

A

A

B

C FIG 67-27  A, FNA biopsy of small retroperitoneal nodule. B, After biopsy, minimal stranding is noted but no significant hemorrhage. C, After several minutes the patient complained of pain, and a repeat scan was performed. The scan shows substantial hematoma (arrow) that is high density from clot formation. The bleeding was self-limited, and no transfusion was required.

Avoidance of Uninvolved Structures A major cause of complications is penetration of uninvolved structures. With the spectacular imaging characteristic of CT and US, intentional penetration of adjacent structures is inexcusable if other clear pathways are readily available. We differ from other authors who proselytize routinely using a trajectory through an adjacent uninvolved structure such as the pancreas or liver during an adrenal biopsy, the bowel during pancreas biopsy, the aorta or vena cava during pancreas biopsy, the spleen during pancreas biopsy, and the pleura during abdominal biopsy. Such approaches are likely to result in higher complication rates if used routinely. Complications are statistically more likely to occur if a less optimal approach or technique is used frequently. The fewer times one penetrates an avoidable uninvolved structure, the lower the probability of a related complication. If one chooses to penetrate the aorta or venous structure during a celiac nerve block, there is a finite rate of complications even if it has yet to be defined. The fewer times one uses the suboptimal approach, the less likely it is that one would have such a bad outcome. At least if one attempts to avoid and not target such structures with a suboptimal trajectory, it would take many more occasions for complications to occur.

Complications Will Occur, or “Complications of No Biopsy” Complications may occur even if no error was made (Fig. 67-28). This was the subject of the article published by Yohann and our group called

“Complications of No Biopsy.”528 In this report, we noted the spontaneous occurrence of three events that would have been interpreted as complications even though there was no association with the procedure. The most remarkable case was a life-threatening spontaneous bleed, which occurred even though the scheduled procedure was canceled and never performed. We emphasize this principle for two reasons. First, any physician who has had a complication despite his or her best efforts and good techniques can be consoled. Second, if the reader is ever involved in contentious case or lawsuit, he or she can use this as a reference to support the unintended complication.

Avoidance of Hemorrhage Although the complication rate is typically low and acceptable with proper needle selection and technique, numerous individuals did early work to minimize or prevent any issues related to hemostasis. Among investigators who have contributed in this area are Chuang and associates73 (injection of absorbable gelatin sponge [Gelfoam] with coaxial system), Allison and Adam13 (angiographic coil through coaxial system), Gazelle and Haaga139 (occlusive protein sheath), and Haaga and Nakamoto191 (local injection of blood products). The most likely potential risk during any biopsy procedure is  the occurrence of bleeding. The vascularity of the pathologic lesion and the anatomy, occurrence rate, and risk of bleeding are specific  for each organ system; these factors are discussed in relation to each organ system. Highly vascular lesions of the liver (e.g., hepatomas, sarcomas, hemangiomas) potentially have a high incidence of serious

CHAPTER 67  Image-Guided Aspirations and Biopsies

2405

Anterior 15

5 C M

0

A

B

FIG 67-28  A, Scan shows thickening of the gastric pull through the segment, believed to represent tumor. Based on this study, I scheduled the patient for a biopsy procedure. B, Repeat scan on the day of the procedure shows a normal-appearing stomach, so the procedure was canceled. There was no skin preparation or local anesthetic. The patient was admitted to a local hospital the next day with life-threatening hemoptysis. She ultimately died as a result of bleeding.

hemorrhage unless meticulous methods are used. Some lesions should be avoided if they are too vascular on dynamic bolus scan unless one is prepared to perform a hemostatic method (see Fig. 67-32). Assessment for a coagulopathy is essential to avoid hemorrhagic complications. In contrast to some radiologists who do not require preprocedure laboratory studies, we believe that measurement of laboratory values is important to ensure that no significant deviation from the norm is present, and we make appropriate arrangements if coagulation studies are abnormal. Although the literature is unclear about what level of laboratory values is required, there is little question that bleeding is directly related to abnormal states. Several animal studies by our group evaluated the effect of needle caliber, type, prothrombin time, and platelet counts.138-141 In these studies, platelet function seemed to be an even more important factor than prothrombin time. With either abnormal platelets or prothrombin time, studies showed that there was a considerable difference between needles smaller than 16 gauge and needles larger than 16 gauge. We prefer that the international normalized ratio (INR) be less than 1.3, with the partial prothrombin time within 25% of normal, and the platelet count be greater than 50,000/µL. In some critical clinical circumstances, exceptions can be made, but with such variances the potential risk is greater, and one is obliged to consider using one of the hemostatic methods described subsequently.

Pyogenic Contamination Contamination of sterile areas is possible from either a cutaneous source or, if a loop of bowel is penetrated, an enteric source. Cutaneous contamination is improbable if povidone-iodine (Betadine) is used for preparation of the skin. Contamination from bowel or normal flora of the liver or kidney can occur, depending on the immune status of the patient, nature of the abnormality, size of the pyogenic inoculum, and virulence of the pathogens. Numerous secondary infections have occurred and been reported in the literature. One reported case of peritonitis occurred in an elderly woman after penetration of the colon by a skinny needle during a lymph node biopsy. In his survey, Smith accumulated 16 cases of generalized bacteremia occurring from aspiration biopsy that were treated by antibiotics.459 Although traversing a loop of bowel during the course of a procedure uncommonly produces a clinical infection, statistically the possibility exists; we prefer to avoid such events if possible. Although an immunocompetent patient can tolerate spillage of a small amount of bacteria within the peritoneum, the clearance of

bacteria is impaired in certain circumstances.206-209 If bacteria are introduced into a fluid space, experimental data show that the effectiveness of bacterial “killing” is impaired because (1) opsonization of pyogens by antibodies and complement is less efficient in a fluid space, and (2) phagocyte-clearing activity is impaired in a fluid space based on physical spatial considerations and the probabilities of random particle encounters. An immunologically compromised patient is at greater risk from pyogenic organisms because of the reduced number or function of immunologic cells. Special precautions should be taken as noted subsequently. If a patient has a normal immune system, we attempt to avoid a trajectory crossing bowel, but if no other pathway is feasible, we accept a trajectory crossing the stomach or small bowel (not the colon). In normal patients, any small inoculum of bacteria is easily eradicated by the normal immune system (practically speaking, it is difficult to penetrate bowel unless one tries, because it moves away from the needle). Even in normal patients, if the lesion being sampled is cystic, we insist on avoidance of bowel for the reasons stated earlier (see Fig. 67-13) because the phagocytes cannot clear bacteria as easily.207,208 Passage through bowel also would be counterproductive for the culture of cystic masses because of possible spurious results from contamination. If the patient is known to be immunocompromised, we avoid traversing the colon or small bowel in all cases. The potential risk to the patient exceeds the benefit, and it is advisable to cancel the procedure rather than put the patient at risk. Avoidance of the bowel in either scenario usually can be achieved by changing the patient position on the table, which typically changes the location of the bowel. Although one cannot be completely sure how the bowel will shift, the air in the bowel will rise and the “floating” action will move it. If a chosen position does not open up the potential pathway, changing the patient position in an opposite fashion (usually one lateral or the other) may open clear pathways for needle insertion. In other cases, injection of saline, air, or carbon dioxide may be used to move bowel loops (see later under “Intentional Organ Movement or Tissue Retraction”).

Vasovagal Response–Initiated Hypotension and Bradycardia The vasovagal reaction is a sudden discharge of the parasympathetic vagal nerve, resulting in bradycardia, vasodilation, and hypotension.

2406

PART III  Image-Guided Procedures

Vasovagal responses must be dealt with properly and expeditiously to prevent severe sequelae or inappropriate iatrogenic intervention. Patients appear pale and ashen gray and may be partially responsive or unresponsive. The appearance is similar to that of a patient who  has had a cardiopulmonary arrest. Close inspection of the patient shows the presence of bradycardia, diaphoresis, dilated pupils, and hypotension. Treatment of this event should be tailored to the severity of the situation. The patient first should be put in Trendelenburg position (raising the patient’s legs while he/she lies flat) to enhance venous return to the heart and cerebral blood flow. The patient should not be seated upright—this worsens the hypotension. If the patient improves with Trendelenburg positioning within 1 to 2 minutes, no other action needs to be taken. If the patient does not improve quickly, oxygen should be administered by nasal cannula to preserve oxygenation, and an appropriate dose of atropine, either intramuscularly or intravenously, should be administered. Before administration of the drug, a quick review of the patient’s clinical notes is appropriate; patients with cardiac arrhythmias may be sensitive to the drug. A single intravenous (IV) or intramuscular (IM) dose of 0.5 or 1 mg usually restores the normal heart rate. If the patient’s condition persists (i.e., if atropine does not improve the situation), urgent consultation with a resuscitation team is appropriate. Such patients, especially elderly patients, can develop a complete heart block, vascular collapse, and cardiac arrest if not properly treated.

Hysterical Reaction The other event, which is more spectacular but less dangerous, is a hysterical reaction. These patients appear agitated—hyperventilating, tremulous, and possibly crying. Close inspection of the patient may reveal tachycardia and systolic hypertension. The pupils are of normal size or smaller. Patients may be breathing fast and complain of subjective shortness of breath. If hyperventilation seems to be occurring, the patient complains of tingling of the lips and being lightheaded. After auscultation of the chest confirms good air exchange, the best treatment is to have patients breathe into their hands or into a bag, if available, to prevent compensatory changes in the cerebral blood flow and subsequent fainting. The next step is to give the patient gentle but firm reassurance; if there is no response, the patient should be removed from the biopsy suite. Before any repeat attempt, the patient should receive an appropriate sedative. Careful assessment of all biopsy patients during the informed consent period usually indicates those who need sedation, preventing any such hysterical reactions.

Tumor Recurrence in Needle Pathway We have seen two cases of tumor recurrence in the needle pathway, one referred from another hospital (Fig. 67-29) and one from a liver biopsy. Although the exact reason why tumor implants occur is unknown, there are two main factors affecting the outcome: type of tumor and the implant site milieu. Mechanisms and pathophysiology of tumor seeding have been studied. Many reports have been performed in animal models assessing the seeding of tumor cells in tissues during and after needle biopsy. Ryd and colleagues showed that 10,000 to 100,000 cells may be deposited in the tissues.428 Most such cells do not survive or produce local tumor growth. Why a selected number of sites develop tumor is unknown, but according to Mahalingam and coworkers, platelet aggregation and fibrin are important for tumor growth to occur.324 The best human information on this topic is related to breast cancer. Diaz and colleagues101 and Stolier and associates478 have reported on the frequency of tumor cell displacement occurring in

1.5

0 cm

L

FIG 67-29  Enlarged view of the chest wall shows a tumor implant (arrow) to the left of the sternum from a previous lung biopsy from another hospital. (Courtesy Dr. Julie Clayman, Cleveland, Ohio.)

tissues after large-needle biopsies and have provided significant insights into tumor seeding. Diaz and colleagues studied a series of 352 cases and looked at the occurrence of tumor cell displacement in specimens that were resected.101 These authors found that a greater number of tumor cells were present in a short period after the biopsies than at a later time. This finding suggests that tumor cells do not survive displacement in usual circumstances in the breast. In the group that underwent resection in less than 15 days, 42% had displaced tumor cells present. In the group that underwent resection more than 28 days, only 15% had displaced tumor cells. The reason a few residual cells may survive to cause low seeding or recurrence likely depends on the type of tumor, general differentiation, immune system of the individual, antigenicity of the tumor, and local milieu (e.g., vascular characteristics). We once speculated that with the expansion of image-guided  biopsies, it was likely that more cases of tumor implantation in the needle track would be observed and reported, and this has occurred. Tumor implants at needle sites have been increasingly reported. We discuss some of the more common sites of needle-associated tumor implants with pertinent comments. When studied, the residual tumor implants from surgical procedures are many times greater than needle procedures. The incidence of implants in the lung is well defined, but most investigators express the opinion that their use is crucial for optimal, expeditious clinical care. Considering the thousands of such procedures performed, the incidence, as noted by Wolinsky and Lischner,516 is very low. More recent data from a surgically resected group, however, indicate that although needle track recurrence is low, patients who have had a needle biopsy before surgical resection have a higher incidence of recurrence in the pleura than do patients without needle biopsy. Matsuguma and coworkers reported that pleural recurrence or needle track implantation was observed for 8.6% of the patients who underwent a needle biopsy, whereas the rate was 0.9% for patients who were examined using other diagnostic modalities (P = .0009).331 There are no specific details about the technique used, so there is no way to determine if that represented a factor related to the increased occurrence. Pleural mesotheliomas are recognized as having a very high incidence of tumor seeding in the needle track (see “Lungs and Mediastinum” and Chapter 39). According to Agarwal and associates, although the incidence of seeding at the needle site is 4% for image-guided procedures, the incidence after surgical biopsy is 22%.5

CHAPTER 67  Image-Guided Aspirations and Biopsies Concerning the liver, there have been numerous reports in the literature of tumor recurrence after needle biopsy and ablation. In recent years, numerous incidental cases of tumor seeding have occurred in patients with hepatocellular carcinoma. Studies led by Soyer,468 Llovet,315 Sato,435 and Kim259 have reported tumor seeding in relation to needle biopsy, tumor ablation, biliary drainage, and laparoscopic procedures. The increase in reports is undoubtedly due to the increased frequency of image-guided microsurgery and minimally invasive surgery and the nature of hepatocellular carcinoma. Needle tracking of renal tumors is infrequent with aspiration biopsies, but it may occur rarely with cutting-type needles. Dr. David Wood at the annual meeting of the American Urologic Society in Orlando, Florida, in June 2008 reported the incidence of tumor seeding at less than 0.01%. Von Schreeb and associates reported a controlled series of about 150 cases of proven renal cell carcinoma used to study the impact of needle aspiration.504 Half of these cases were aspirated and half were not. There were no instances of needle track seeding. The pancreas has had a frequent rate of tumor seeding compared with other organ systems.125,245,349,414 Johnson and coworkers studied the effect of skinny needle aspiration in 41 patients who went to surgery for curative intent.245 There were no instances of tumor seeding. Peritoneal washings were positive in 18% of patients who had needle biopsy, and 20% of a small group without needle biopsy had positive peritoneal washings. Tumor seeding in most patients with pancreatic cancer does not seem to be a significant issue20 because few patients are surgical candidates, so a minimally invasive method of diagnosing the disease is crucial. For patient care, the literature is replete with reports about the benefit of percutaneous procedures for early diagnosis, confirmation of recurrence, and image-guided treatments. There are no proponents for reducing the use of these procedures, but rather recommendations that techniques be refined to minimize the occurrence of such events. It seems that the issues related to tumor seeding in the needle track can be easily avoided by using a coaxial system. The two recommendations we previously made (i.e., using a pathway through normal liver and using a coaxial system) have been validated. Several authors have reported that tumor tracking can be avoided if such techniques are used. The factors that seem to prevent tumor seeding are an approach through an adjacent loop of normal tissue408 and a coaxial system.332

PREPROCEDURE HISTORY, REVIEW OF IMAGES, PLANNING, AND INFORMED CONSENT Patient History and Informed Consent Before the procedure and administration of any medication, it is important to review the pertinent medical and drug history and to discuss the procedure in depth with the patient. From the history, it is important to ensure that the proper procedure has been chosen and that necessary preparatory laboratory studies and arrangements have been made. With the fast pace of medical care today, it is common for the referring physician to have delegated the order and coordination of the procedure to a nurse, receptionist, or physician assistant. Frequently the patient’s history provided by the paperwork is incomplete or incorrect. Depending on the purpose of the procedure, one has to ensure that the proper pathologic analysis is being ordered (i.e., flow cytometry, electron microscopy, biomarkers). If the patient is chronically ill or fragile, special arrangements such as anesthesia or hospital admission might be appropriate. During this time, one should inquire about any medication that may alter coagulation or platelet function. The latter is especially important because platelet inhibitors are widely used and not described to patients as “blood thinners.” Routine coagulation studies do not detect the

2407

presence or activity of platelet inhibitors. Laboratory work performed by our group has confirmed the serious alteration and risk of the coagulation status in patients taking aspirin-like compounds.141 Two of the most troublesome situations are patients who are being treated with low-molecular-weight heparin and patients with systemic lupus erythematosus. Use of low-molecular-weight heparin is widespread because it has a much lower incidence of overanticoagulation than traditional heparin. The problem with this agent from the perspective of interventional procedures is that the methods for assessing the degree of anticoagulation are not readily available at many hospitals. Few clinicians are aware of the required test, which is an assessment of factor X inhibitor. Within our community, several catastrophic events have occurred in association with this new agent. With systemic lupus erythematosus, there is a high incidence of patients who have spuriously elevated coagulation tests owing to unknown circulating antibodies. The dilemma that arises is that when such tests are positive, one cannot be sure if the patient would clot normally. Frequently we are asked to perform biopsy in such patients, and clinical physicians typically assert that there are few risks. Both of these situations are fraught with difficulty because the attending radiologist, not the referring clinician, bears the brunt of the legal consequences if bleeding occurs. The best approach in these situations or others that might arise is to rely on a more clinical coagulation study—the bleeding test—when available. This test evaluates the time required for a standard puncture site to stop bleeding; it takes into account all factors, including the intrinsic, extrinsic, and capillary properties. A substitute for this test for the interventionalist is to observe carefully a small stab wound made at the entrance site for  the procedure, as noted later under “Bedside Assessment of Bleeding Diathesis.” During the interview process and informed consent, all details, including the rationale, medical alternatives, method of the procedure, sensations to be felt by the patient, and postprocedure routine, should be mentioned. After this discussion, the patient may sign the informed consent form and help confer about the need for any sedation.

Patient Cooperation and Sedation One cause of complications common to all procedures is lack of patient cooperation. If a patient moves or cannot suspend respiration at the appropriate time, difficulties may arise. The most critical factor to enhance patient cooperation is adequate local anesthesia in the area of the biopsy, the goal being “pain-free anesthesia” as noted subsequently. If local anesthesia is not administered correctly throughout the pathway, including the serosal surface of the organ (e.g., pleura, capsule), the patient may jerk at the time of instrument insertion and produce a tear, bleed, or pneumothorax. The other factor that has been alluded to and is critical is patient comfort on the table. The patient should be positioned in whatever position is the most comfortable within the context of the anatomic trajectory. This position must be specific to the patient and may include pillows under the knees or shoulders. If patients must lie immobile, they may fidget if they are uncomfortable. For patients who are emotionally upset, administration of an appropriate hypnotic or sedative can be helpful. Because of the powerful but effective sedative and analgesic agents now available, one must be conversant with basic life-support resuscitation and the sedation protocols of the home institution. Most institutions now require some type of basic proficiency for sedation administration, so the operators must ensure that they comply with institutional, state, and federal guidelines and certification. Two types of medication may be given to patients: one for sedation and anxiety and another for analgesia. For anxiety, a mild sedative such

2408

PART III  Image-Guided Procedures

as midazolam (Versed) should be given. In most average-sized adult patients, 1 mg intravenously can be given at the onset and then titrated to the needs of the patient during the procedure. It is probably better to err on the side of giving medication, but if there is no apparent need by the history or from clinical observations of the patient, one should withhold the drugs.

Laboratory Studies For routine aspiration and cutting needle procedures, we request a complete blood cell count (CBC), partial thromboplastin time, INR, and platelet count. In view of the greater context of the litigious society and multiple layers of administrative oversight, such studies are of minimal cost and effort to prevent even a rare catastrophic event. The CBC is important to ensure that the patient has sufficient hematocrit reserve to cope with a hemorrhage. The hematocrit should be at least adequate; the exact value is relative, but I prefer a level greater than 33% for routine cases. If there is any question about the patient’s reserve capacity, a type and crossmatch of blood products should be made available if needed. We prefer the platelet count to be greater than 50,000/µL and the INR less than 1.3. Although it is difficult to find a definitive recommendation for the INR value in the literature, we derived this value from an NIH study formulated by the Transfusion Medicine/Hemostasis Clinical Trials Network, a subsidiary of the New England Research Institutes. In this proposal formulated for the Study of Hemostasis and Invasive Procedures, a consensus opinion by the participants recommended an INR level of 1.3 or greater for replacement therapy.111 In cases in which patients have an altered coagulation status, the laboratory values, clinical factors, remedial corrective steps, attendant risks, and associated precautions should be understood by all involved (patient and physicians), and appropriate contingencies should be prepared for.

Bedside Assessment of Bleeding Diathesis When laboratory values are unavailable for an emergency procedure or other reason, we have developed great confidence in the use of a simple bedside method for evaluating bleeding diathesis. The “bleeding time” is a very old laboratory method of evaluating the entire coagulation cascade by making a small stab wound in the arm. The test is performed by observing how long it takes for a bleeding wound to stop bleeding. This test is superior because it assesses in vivo the function of all components in the coagulation pathway, including platelets, blood proteins, capillary fragility, and circulating inhibitors or antibodies. While we do not order this test, we use a different permutation at the bedside that we believe is helpful for avoiding possible catastrophe. After the skin is prepared, we always make a small stab wound at the entrance site of the needle. Observation of this wound provides a “bleeding test.” Although we do not time the event, we pay careful attention to the clotting of blood from this site. If bleeding continues to “stream” from the site despite several minutes and “blotting” with gauze, we become concerned about an inaccurate history of anticoagulative medication (i.e., aspirin) or an unknown bleeding diathesis. During the course of our practice, we have identified patients who have had incorrect laboratory values, inappropriate administration of anticoagulants, or other issues. In questionable cases we would rather postpone a procedure than cause a complication. In such cases, the laboratory tests and history are reviewed before the procedure is rescheduled.

Needle Preferences in Coagulopathic Patients For patients with a normal coagulation state, there is little clinical indication that needle caliber is the prime determinant of hemorrhagic

TABLE 67-1  Separation of Means of

Biopsy Blood Loss Data by Needle Gauges Target Organ Liver   No warfarin   Warfarin Kidney   No warfarin   Warfarin

Groupings of Needle Gauges

P Value

(14) (16) (18, 20, 22) (14) (16) (18, 20, 22)

<.01 <.01

(18, 20, 22) (18, 20) (22)

NS <.05

Note: Groupings within a single set of parentheses are equal versus those grouped in other parentheses. From Gazelle GS et al: Effect of needle gauge, level of anticoagulation and target organ on bleeding associated with aspiration biopsy. Radiology 183:509–513, 1992. NS, not significant.

complications, but common sense dictates that smaller needles be used if such needles can provide sufficient tissue or sample for the intended histopathologic evaluation. The impact of anticoagulation was elucidated by two well-designed animal models from our institution. Studies led by Nakamoto (Nakamoto D: personal communication, 2006) and Gazelle140 reported on the effect of anticoagulation on blood loss in two organs: liver and kidney. Using aspiration and cutting needles, the investigators found that there were absolute differences in local bleeding with various calibers of needles (Table 67-1). We also showed that by using an occlusive hemostatic device, the amount of bleeding was reduced to essentially normal levels. If a coagulopathic patient has laboratory values completely restored to normal by the methods described subsequently, needles of the caliber typically used are suitable. If coagulation values remain slightly abnormal, one should use one of the described hemostatic methods.

Coagulopathy Treatment

Abnormal INR.  For an abnormal INR, there are two options depending on the time issues involved. If the patient is taking a warfarin-like drug, it should be stopped. The patient can be given vitamin K, which enhances recovery of the prothrombin level to normal over several days, but the laboratory values must be rechecked until at a normal range. If correction of vitamin K is ineffective, fresh frozen plasma (FFP) may be administered intravenously before the procedure. Repeat INR values should be drawn until a normal level is reached. Some authors recommend administration of several units before performance of the procedure, without repeat measurement of the laboratory values, but we believe a repeat coagulation screen is important to ensure that factors have returned to normal. For patients whose coagulopathy cannot be corrected by the aforementioned means, or in emergency circumstances that require early performance of the procedure, we have developed a new method to prevent hemorrhage, which involves the local injection of blood products as described later under “Local Infusion of Blood Elements for Hemostasis.” This is not an FDA-approved technique, but it is legal under current statutes; we have used it for several years with excellent results and no bleeding complications to date in more than 50 cases.

Low Count or Impaired Platelets.  With platelet abnormalities— either absolute count or dysfunction—several corrective approaches are possible. If the platelet impairment is simply due to drug administration, cessation of the drug and waiting an appropriate period is

CHAPTER 67  Image-Guided Aspirations and Biopsies appropriate. If drug cessation does not work, a transfusion of platelets can be given. If the absolute count is depleted or an untreatable dysfunction exists, IV platelets should be given. For an impaired platelet function or low platelet count, three possible approaches can be taken: IV systemic platelet replacement, infusion of desmopressin, or local platelet injection in the pathway. For impaired platelet function, especially for elevated blood urea nitrogen (BUN), desmopressin may be administered. The dosage and regimen for infusions should be obtained from the local pharmacy. Although the exact mechanism of action is unknown, it has been shown over many years that administration of this medication improves the adhesiveness of platelets and minimizes the probability of hemorrhage. The medication is typically most effective on the first administration and is less effective when later doses are administered. In many patients, platelet impairment is produced by medications such as aspirin, other analgesics, or antiplatelet agents given for cardiovascular reasons. Administration of platelet inhibitors or aspirin has become established as the mainstay of cardiovascular treatment. The significance of this platelet impairment was well described in a laboratory model by Gazelle and colleagues,141 who studied the effects of aspirin. Patients with impaired platelet function on the basis of either medication such as aspirin or other inhibitors are best treated by stopping the medication and waiting the appropriate time for the effect to resolve. Although the effects of aspirin are long-lasting, we consider a 7-day period without medication as the optimal time interval before a procedure should be performed. Nonaspirin pain relievers are usually shorter acting; a waiting period of 1 day after cessation is usually adequate, but the pharmacy should be consulted for specifics of any medication.

Traditional Infusion Replacement.  The traditional approach for correction of coagulopathy is to administer IV platelets or FFP over the course of many hours to obtain a systemic improvement of these factors. If replacement platelets are used, the infusion should be continued until the platelet count reaches greater than 50,000/µL. Restoration of the INR to acceptable levels is accomplished by IV administration of FFP. Numerous units are given until the INR reaches 1.3, as noted earlier. Fresh plasma permits this restoration of the INR because it replenishes numerous factors such as prothrombin and factors VII, IX, and X, which are produced by the liver and are deficient when liver function is impaired. There are several difficulties with the traditional method. The most difficult issues are related to obtaining the blood products, administering the products, rechecking the tests, and transporting the patient to the imaging site to perform the procedure. Regardless of the institution, anywhere in the world, there are always logistical problems. If patients have cardiovascular compromise, the multiple units infused may cause cardiac overload and decompensation. In some patients, there may be antibodies or unknown reasons why such products do not correct deficiencies. If patients show a lack of response to such methods or the clinical scenario is so dire that the procedure must be performed emergently, one can opt for the new method of local infusion of platelets, as described next.

Local Infusion of Blood Elements for Hemostasis.  We continue to refine and study our technique for treating bleeding, the local infusion of blood products. This is not an FDA-approved use of blood products, but under current legal guidelines, a physician can act on the best behalf of a patient in unusual circumstances. If a product is approved for human usage, a licensed physician can use that product for other uses when the benefit to the patient seems to outweigh  the risks. Under these guidelines, a physician can use the technique

2409

described later in this section, but he or she takes on any associated ethical or legal risks. The innovation of this technique evolved from the treatment of a radiologist colleague who had refractory leukemia unresponsive to systemic platelet infusion. Having received many platelet infusions over the course of his treatment, he developed antibodies to platelets and hypersplenism. This condition was so severe that if he was given 10 U of platelets, his body would consume the platelets and his platelet count would never increase above “0-10.” After developing a severe spontaneous hemothorax causing compression of his heart and diaphragm, he was refused treatment by thoracic surgery and was sent  to us for percutaneous treatment. We recognized the possibility of rebleeding but consented to perform the procedure for temporary relief. At the time, we reasoned that one might be able to use platelets with local administration as an alternative method. The rationale was that blood products injected into the pathway with a small-caliber needle before the biopsy or drainage instrument would create a “seroma” of the missing product (e.g., platelets, FFP). The concentration of the blood product would be “astronomical” in the local site. If the instrument encountered a vessel in the path, the local product would form a local clot. Because there was no other alternative for our radiologist colleague, we tried the method and it worked (Fig. 67-30A). With drainage of hemothorax, the patient improved symptomatically and lived for 3 months afterward. We have since used local injection of platelets and FFP in patients who have an abnormal prothrombin time or low platelet count as a partial solution to the problem. An even more extraordinary case involved a teenager with hypersplenism whose low platelet count was unresponsive to replacement. To exclude the possibility of portal hypertension secondary to portal vein stenosis, a splenoportogram was ordered by the surgeons (see Fig. 67-30B). Platelets were injected locally into the pathway during insertion (and removal) of an 18-gauge needle, and the study was performed without incident. The many clinical circumstances under which we feel use of this technique is justified include: (1) the patient is too ill to wait for infusion therapy and in a life-threatening situation; (2) an existing condition of patient (e.g., pulmonary embolus, failed vena cava filter, recent coronary stent placement, cerebral occlusion) requires maintenance of anticoagulation; and (3) failure of the patient to respond or recent history of failure to respond to replacement therapy (i.e., antibodies, hypersplenism). If these factors are present, we perform the local injection of blood elements as follows. After preparation of the site with povidone-iodine, the site and pathway are prepared in two steps. The syringe containing lidocaine is mixed 50% by volume with either FFP or platelets. The pathway is anesthetized (which includes blood product) in the routine manner, ensuring that the pathway is treated from skin to the organ surface and edge. Next we make a repeat injection of approximately 5 to 10 mL of blood product into the expected trajectory of the needle, including the skin, subcutaneous tissue, and surface of the organ and target (see Fig. 67-30). It is crucial that the entire pathway be injected, not simply the subcutaneous tissues or the capsule of the organ; bleeding can potentially occur from any site along the pathway and can push out any localized plug of clot. When a biopsy rather than a drainage is being performed, it is prudent to use a coaxial system that includes a guidance cannula and the biopsy instrument. The appropriate samples can be taken and the needle removed, leaving the cannula. After the biopsy, the blood product is injected in the pathway through the coaxial cannula as it is removed (Fig. 67-31). If bleeding occurs, one can inject more blood product into the site to provide hemostasis (see Fig. 67-31) (this has not happened to date in our

2410

PART III  Image-Guided Procedures

19

0

A

FIG 67-31  Patient with an INR of 1.6 required a needle biopsy of the liver. CT scan shows the needle within the coaxial cannula during biopsy of the liver (see text).

B FIG 67-30  A, CT scan of a former colleague (an interventional radiologist) showing a large hematoma in the right chest. Because the patient had refractory leukemia and antibodies to platelets, the platelet level could never be elevated beyond a total value of 0 to 5. Other clinical services refused to attempt drainage to permit expansion of the lung. A small needle is used along the pathway to infiltrate platelets. (This is not an FDA-approved technique, but it is legal under FDA guidelines.) B, A 16-year-old girl had hypersplenism and low platelet count of 16,000/µL owing to consumption by spleen. She was not responsive to platelet infusions. Surgeons required a splenoportogram to delineate the portal vein to determine if occlusion was present; the operating room was kept open in case bleeding occurred. Local injection of platelets in the pathway was used during the procedure, which was performed safely without bleeding.

experience), or one can use the coaxial/coil/thrombin method if bleeding continues (see subsequent technique and Fig. 67-32). If a drainage is being performed, no coaxial system is used. After this step, the drainage procedure is performed in the routine fashion. Our results over the past 5 years have been very favorable. The rationale for this technique is that the local injection creates a local “seroma” of the missing product so that if any adjacent vessel is

inadvertently punctured, the local elevated concentrate of the blood product provides material for an effective blood clot. With FFP the seroma concept is appropriate. With platelets, the platelets probably undergo immediate activation, which provides overabundant amounts of thromboplastin to initiate the thrombosis cascade. From a physiologic perspective, investigators have shown that a native clot that forms at the site is superior to an injected “clot.” A clot formed after restoration of local or systemic coagulation factors is of better quality56,123,161,467 than one formed when an imbalance of coagulation factors exists.27 A “balanced” clot is more effective because studies have shown that the strength of a clot is due half to the platelet agglutination and half to the serum proteins.

Occlusive Hemostatic Methods for Cutting Needles.  The most consistent and practical method to prevent localized bleeding after a biopsy is placement of more solid occlusive material through a coaxial system after the biopsy has been taken and the needle has been removed. Four such materials have been used: Gelfoam, occlusive liquid “glues,” angiographic coils, and angiographic coils with thrombin. In regards to injectable “glues,” numerous products have been evaluated in animal models400,460 and approved for clinical use, but in our experience their application is difficult. These products consist pri­ marily of fiber material such as cellulose/Gelfoam460 or liquid protein material such as fibrin glue,120,400 which solidifies in the pathway after injection. A major problem of these products is that the fixation  time of these materials is not predictable. When these are injected through a needle pathway, one cannot always be confident of the placement of the protein material. As the injection cannula is withdrawn, the protein plug may be pulled out. From the research data in the

A

B

C

D

F

E FIG 67-32  A, Biopsy needle with coaxial system for biopsy. The cannula with stylet is inserted into the target site, and then the biopsy needle is inserted through the cannula. The biopsy is completed and the needle is withdrawn, leaving the cannula in place for insertion of the hemostatic coil with thrombin. B, Equipment needed for assembly before insertion of coils are thrombin, dilators, coils, and angiographic wire. C, The coils are pushed to the end of the dilator using an angiographic wire. D, To prevent ejection of the coil, a finger is held firmly over the tip until the coil end is at the dilator tip. Finger pressure prevents ejection. E, Radiograph shows an empty dilator on the left and a dilator with coil at the tip (arrows). F, Thrombin solution mixed according to the label is injected slowly through the dilator until it drips through the end; the coil is held in place by the fingertip during the injection. This small amount of thrombin in the dilator and on the coil produces clot formation; no thrombin is injected directly into the patient. Continued

2412

G

PART III  Image-Guided Procedures

H

FIG 67-32, cont’d  G, Radiograph shows simulated biopsy with the stylet of the cutting needle extending beyond the cannula (arrows). H, After the biopsy needle is removed, the coil is pushed beyond the end, depositing the coil at the precise site of the biopsy (arrow). Because of the coil, mechanical pressure does not permit displacement of the coil. Thrombin on the coil immediately produces clot formation.

previous section, one appreciates that such artificial clots are not as good as natural clots. Injection of Gelfoam through a coaxial system was reported by Chuang and colleagues.73 This method does not always work well because the Gelfoam is difficult to position properly and may be pulled out with the cannula, similar to the injectable materials. Coaxial sheath with angiographic coils was reported by Allison and Adam.13 Although this technique has been reported as effective, we have altered it to use thrombin because angiographic coils have been changed from their original design. Until about 15 years ago, the coils had fabric on the end intended to serve as foci or anchors for clot formation. Those early coils had “wool” fibers on the coils, which was very thrombogenic, but the material was replaced with Dacron, which is not as thrombogenic. After being frustrated with the use of such coils because clot did not form quickly, we devised the coaxial/coil/thrombin method that we now prefer, which is the use of a coaxial system, angiographic coils, and thrombin. We modified our technique to add thrombin after numerous studies. In a study of many coagulopathic animal experiments, we found that the coil with thrombin has three distinct advantages: (1) the coil is positioned precisely at the site of the biopsy “cut,” (2) the spring coil stays in place and does not “pull out” with the coaxial cannula, and (3) there is instant clot formation for the thrombin. After one becomes familiar with this technique, one easily prepares the items required for hemostasis either before a procedure or during the procedure when bleeding from the guidance cannula is noted. The coaxial cannula must always be left in place for access to insert the material. Preparation for this technique can be made before the procedure if the bolus injection shows increased vascularity or large collateral vessels. Required equipment includes two vascular dilators 6F or larger, angiographic coils, and thrombin (Fig. 67-32A and B). Before the coaxial/coil/thrombin technique can be executed, the equipment must be assembled such that the dilators are loaded with the coils and thrombin so that they can be inserted as needed and noted in Figure 67-33. The coils are loaded into the dilators and pushed carefully to the end of the dilator (see Fig. 67-32C) and not beyond; holding the finger over the end as they are inserted prevents inadvertent deployment (see Fig. 67-32D). The thrombin is mixed according to the instructions and injected into the dilator slowly while holding a finger loosely over the end (see Fig. 67-32F). The idea is to permit the thrombin fluid to flow through, but not eject the coils if the injection is too forceful. Figure 67-32G shows a radiograph of a simulated biopsy, with the inner stylet of the needle inserted. Figure 67-32H shows a radiograph of the coaxial cannula after removal of the biopsy needle, and the coil with thrombin is pushed out the end of the cannula. In cases in which an enhancing lesion is to be biopsied (see Fig. 67-33A), we prepare the equipment as noted previously and inform the attending physician and the patient of the increased risk of bleeding and the method to be used to prevent bleeding. It is the prerogative of the patient and the physician to decline the procedure, but if it is decided to proceed, the routine steps are taken as noted in Figure 67-33B through D. The biopsy using the coaxial guidance system is performed in the routine fashion. The biopsy is performed through the cannula, and the tissue sample is removed with the cannula left in place (see Fig. 67-33B and C). Through the cannula (see Fig. 67-33D), an angiographic wire is used to push the coil through the dilator into the cannula and out the end of the cannula (Fig. 67-33E). After the procedure, the patient should lie quietly for 30 minutes to let a clot secure the occlusive material in place. If patients get up too quickly, the device may dislodge into the peritoneum if they happen to cough or exert themselves; there is no direct harm to the peritoneum except

CHAPTER 67  Image-Guided Aspirations and Biopsies

2413

B

A

D

C

E FIG 67-33  A, Preplanned placement of hemostatic coil should be performed if intense vascular enhancement is shown on a dynamic contrast bolus scan. The left kidney shows increased enhancement of renal tumor and large collateral vessels in the retroperitoneum. B, A coaxial cannula is positioned adjacent to the target in the anterior margin of kidney. C, CT scan shows the stylet extending beyond the end of the cannula (arrows). D, After biopsy, the biopsy needle is removed, leaving the cannula in place for the hemostatic procedure. E, The angiographic coil soaked with thrombin is pushed into the site by the angiographic wire (thrombin is not injected but rather is on the coil; see Fig. 67-32). The coil remains precisely at the site of the biopsy (arrow).

for possible bleeding from the site. This method may be used in the liver or other sites as needed. This coaxial method can be used with either a guided procedure of a mass or a diffuse sample of the liver. Occasionally a unanticipated potential problem with bleeding may occur (Fig. 67-34). In such cases, if excessive bleeding is observed from the hub of the cannula after biopsy of a lesion (see Fig. 67-34A and B), one could decide to do nothing, but we prefer to proceed with this hemostatic method. If bleeding occurs, a stylet is inserted to delay the

bleeding temporarily, and the equipment is prepared as noted earlier. If one has the equipment available, preparation takes only several minutes. As noted earlier, the coils are inserted through the cannula to provide hemostasis (see Fig. 67-34C through E). When first learning the technique, it is helpful to have fluoroscopy available because it is easier to follow the insertion of the thrombin/ coil in “real time.” One can use the topogram or rapid repeat CT scans, however, to monitor the coil insertion. Such procedures can be

2414

PART III  Image-Guided Procedures

B

A

D C

E FIG 67-34  A, Emergent insertion of hemostatic coil. CT scan shows a soft tissue mass in the posterior abdominal wall. B, Coaxial needle biopsy of the mass. Note the stylet extending beyond the cannula (arrows). C, Note the cannula remaining at the site after the needle removal. Rapid bleeding occurred backward through the cannula. The stylet was replaced. Preparations for hemostatic coil were made as noted in Figure 67-32. D, Thrombin-soaked coil is pushed through the cannula, depositing it in the biopsy site (arrows). E, Two coils were inserted beyond the tip of the cannula. Hemostasis was accomplished, and the cannula was withdrawn.

performed at the bedside with CT or US, but this requires extensive experience because with this approach the procedure is essentially “blind.”

GENERAL PROCEDURE STEPS There are many consistent steps involved for performing all procedures. This section discusses in general the approach to be used, but for each organ system there are some variations, which are discussed in the corresponding organ sections.

Review of Imaging Studies and Selection of Target, Pathway, and Patient Positioning For every procedure, the diagnostic imaging studies performed should be reviewed before scanning the patient, and the appropriate lesion target, the safest possible pathway, and the optimal patient position should be chosen to avoid uninvolved organs, vessels, bowel, and possible nerve sites. Common sense dictates that potential problem areas such as arteries and veins should be avoided. Muscles can be a nuisance because

CHAPTER 67  Image-Guided Aspirations and Biopsies they are highly vascular and can bleed locally. Also, when penetrated, muscles can be tightened purposely or unintentionally by the patient and cause difficulty with needle placement. The trajectory of a needle can be changed by a muscle contraction. If a powerful muscle group such as the spinal erector or gluteal muscles contracts, a needle can be physically bent. The most notable areas where this can be a problem are pathways through the sciatic notch and the retroperitoneum. Patient position is decided to optimize the chosen anatomic approach. Patients may be in the supine, prone, oblique, or lateral position. Patient positioning is important for three reasons: (1) patient comfort, (2) anatomic approach for trajectory planning, and (3) producing immobility of a mass in cases where the target may be partially mobile or in patients who cannot completely cooperate. Having the patient in a comfortable position is important because patient discomfort may result in a movement at an inopportune time or increase the patient’s anxiety and ability to cooperate. Increased patient anxiety generally affects the patient’s ability to control his or her respirations. Another important aspect of patient positioning is using patient position to shift internal anatomy to facilitate the biopsy; this has been discussed earlier, but it is worth repeating because of its great benefit. By having the patient lie in different positions, the anatomic access may be easier if internal anatomy shifts. Notable examples are seen in Figures 67-10 and 67-11. As noted earlier, if a lesion being sampled seems to be mobile owing to respiratory variation, placing the ipsilateral side down can be helpful to limit the motion of the mass (Fig. 67-35). In such cases, the weight of the viscera partially immobilizes the ipsilateral diaphragm and limits motion during quiet respiration (see Fig. 67-11).

Site Preparation for Cleansing and “Painless” Local Anesthesia When the entrance site has been chosen, the skin is prepared with povidone-iodine and a local anesthetic agent such as lidocaine is administered. At the time of lidocaine administration, it is important to palpate the tissue carefully to ascertain the location of the ribs, cartilages, and processes so they can be avoided. This is true especially next to ribs, the sacrum, the scapula, or other bony structures that have a tapered edge or traverse the scan obliquely. Such structures produce a partial volume error in the CT scan that results in a discrepancy

A

B

2415

between their apparent location on the scan and their actual anatomic location. Administration of local anesthesia can be painless if it is performed properly using newer techniques instead of the traditional method. We believe the traditional method typically taught to medical students is antiquated; it was based on the use of procainamide, an older agent that had significant cardiovascular effects. The traditional teaching was that one should make multiple insertions and injections. According to the old method, one inserts the needle a small distance, pulls back on the syringe plunger to see if blood is aspirated, and then injects the anesthetic. That process was repeated numerous times until the entire pathway was infiltrated with medication. After one uses this method, one quickly realizes that every insertion of the needle is very painful because the tip extends beyond into unanesthetized tissues. The new optimal method is as follows. The first step to permit painless local anesthesia is to modify lidocaine for injection to prevent the “burning” during injection and change the injection technique of the medication. We learned from plastic surgeons that the burning sensation produced during lidocaine injection can be eliminated by neutralizing the slight acid pH of lidocaine. To do this, one injects 5 mL of sterile injectable sodium bicarbonate into a 50-mL bottle of lidocaine. This preparation can be used in the standard manner as lidocaine. It can be used only during the 8-hour period following, however, because the sodium bicarbonate degrades the lidocaine slowly; it must be discarded, and a new bottle remixed after that time. To lessen the discomfort during the needle insertions for the local injection, the technique is modified. The first important point is the method of injecting the lidocaine into the projected needle pathway. The skin wheal is made in the skin and subcutaneous tissues first, as always done before. During the subsequent injection of the lidocaine, the lidocaine is injected very slowly and the needle advanced slowly behind the injection. In this manner, a fringe of anesthesia precedes the needle tip, and the patient typically feels nothing but pressure—no needle sticks or burning from the medication. In effect, a small boundary of anesthesia precedes the tip of the needle as the pathway is extended. At first consideration, this method seems to violate the concern about the possible injection of anesthetic agent into a vessel. The possibilities of this event are extremely remote; consider how difficult it is

C

FIG 67-35  Small lesions may be difficult if the patient has any variation in respiration. These three images show the method for immobilizing the lesion and a method of providing a pain-free procedure for the patient. A, This scan shows the patient in a decubitus position with the lesion side down. In such cases, one must be sure access is possible anteriorly or posteriorly. On this scan, the needle is on the margin of the rib, where lidocaine is infiltrated. B, This scan shows infiltration with a small amount of lidocaine in the pleura (arrow), not the raised pleura from the injection. C, This scan documents the placement of the needle in the margin of the lesion.

2416

PART III  Image-Guided Procedures

to perform a vascular puncture even when it is being intentionally done. Second, the “vascular/arrhythmia” concern arose in the early 1950s when procainamide was used as the anesthetic agent. If one reviews the literature, one cannot find any actual report about such an event. It is well recognized that lidocaine even when used as a cardiac antiarrhythmic agent is very safe, and very large doses are given; even if a microscopic injection occurred, it would be of no consequence. We have used this technique for 20 years and have never observed an adverse event. Finally, when infiltrating local anesthetic, it is important to infiltrate the tissues down to the level of the pleura or capsule of the organ being sampled. One should completely infiltrate the pathway in two steps. In the first step, the skin, subcutaneous tissue, and superficial muscles are infiltrated. After that point is reached, the patient suspends respiration, and the local anesthetic is infiltrated down to and including the pleura, peritoneum, or the capsule of the organ (e.g., kidney or liver) (see Fig. 67-35). Such infiltrations should be made as the patient suspends respiration. If the capsule is not adequately anesthetized, the patient may jerk the diaphragm during the procedure. In our opinion, inadequate administration of local anesthesia is the most common cause of poor patient cooperation. If a needle is inserted without anesthesia of the capsule, a “spinal level reflex” occurs; this moves the organ and produces a potential lacerating effect. Adequate anesthesia also is most important with anxious patients because they will not cooperate well if they feel pain.

Indications for and Selection of FNA and Cutting Needles

Comparison of Aspiration and Cutting Needles.  The superior sample quality from a core cutting needle is no longer disputed. Numerous authors have reconfirmed the advantage of diagnostic  accuracy with little difference in complications in most organ systems.160,176,265,329 There is no longer any controversy, and virtually all workers concur that cutting needles are superior to aspiration needles. Aspiration needles (FNA).  An aspiration needle is selected for four different reasons: (1) clinical cancer history, (2) increased vascularity of the lesion, (3) anatomic location, and (4) safety factors. If the clinical history includes a known malignancy and there is no suspicion of a second primary, an aspiration needle consistently provides confirmation of the disease. Comparison of the cytologic sample with previously obtained tissue samples provides definitive information. Although highly vascular lesions may bleed with any type of needle and procedure, FNA does provide a lower incidence of bleeding. Certain anatomic sites are predisposed to complications or are close  to critical structures. Thyroid, lung, pancreas, and spleen are sites  with higher possibilities of complication, so an FNA sample would suffice. Anatomy may dictate an aspiration needle because of close proximity to structures such as vena cava, aorta, bowel, or nerves. With extensive experience, many such difficult lesions can be biopsied  safely with small-core cutting needles because of the precision with CT placement. Safety is the overriding factor that sways most operators to use the skinny needle. Reasons to stress safety and to defer to an aspiration needle include an uncooperative patient, a bleeding diathesis, or other unique factors in the patient’s history. Cutting needles.  A cutting type of needle is preferred whenever possible because it provides such a superior tissue sample. If the patient presents with unknown disease, possible change of cell type, previously failed aspiration, or suspected benign disease, a cutting needle biopsy should be performed. A large needle can be used if clear access is available to the organ of the abnormality (i.e., no intervening uninvolved structure or close

proximity of the aorta, vena cava, portal vein, or bowel). As experience is gained, skill increases, and one’s safety tolerance relative to anatomy improves. Although a novice might be fearful of performing a biopsy of a node several centimeters from the aorta, an experienced individual would be comfortable biopsying some lesions that abut the aorta. One factor we still consider important is to assess the vascularity by observing the enhancement during a stationary dynamic bolus scan over the site of the lesion. Vascularity is an important issue for prevention of bleeding complications. The importance of the basic vascularity on the incidence of biopsy complications can be appreciated by noting the incidence of complications in the liver when bolus scans are used to biopsy the liver and the series of cases where no such assessment occurs. Bret and coworkers45,46 reported a series of biopsies on hepatoma patients using 22-gauge aspiration needles and had a hemorrhagic complication rate of almost 5%. In our 32-year experience of liver biopsies using 14-gauge needles, we reported a hemorrhagic complication rate of 0.5% in a group of 200 patients (see “Liver” section).106 Some slightly vascular masses may be biopsied using the hemostatic occlusive method discussed earlier.

Selection of Aspiration Needle Type and Caliber.  The selection of aspiration needles is straightforward because there is clear laboratory and clinical information that indicates the best type and caliber of FNA needle to use. These points were shown by historical articles and have been reconfirmed over the years without controversy. In the earliest laboratory study by Andriole and associates, our group showed that the best samples with the least amount of tissue artifact were obtained by the small-angle bevel needle tip (25- to 30-degree angle) and by the largest caliber needle (Fig. 67-36).17 This was later confirmed by Dahnert and colleagues.89 In the direct clinical comparisons, these results were reconfirmed. Pagani showed with aspiration biopsies of the liver that an 18-gauge aspiration needle was better than a 22-gauge needle.394 In the lung, there has been no definite benefit shown with large aspiration needles, and there was a larger incidence of hemoptysis shown by Khouri and colleagues.257 In the pancreas, Dickey and associates of our group showed the diagnostic return for a 20-gauge needle compared with a 22-gauge needle, with no difference in complications.102 Reporting on the prostate, more recently Norberg and coworkers showed increased diagnostic yield of the 0.9-mm needle (20 gauge) over the 0.8-mm (22 gauge) or 0.7-mm aspiration needles.378 From our experience and the numerous reports as noted earlier, we have consolidated the needle choices in our institution to use the 20-gauge FNA needle in all organs. It is suitable for lung and solid organs.

Selection of Cutting Needle Type and Caliber.  Many variations of cutting needles exist, so numerous factors should be considered in the decision making. The first question is whether to use an end cutting needle or a side cutting needle for soft tissue or bone. There seems to be little controversy from either the scientific perspective or the clinical perspective about the merits and shortcomings of these needles. Numerous authors, including Bateson and colleagues28 and Andriole and coworkers,17 have shown that the side cutting needle provides the best sample without crush artifact or fragmentation of the core. For bone, the appropriate needle is the end cutting needle unless the target is associated with soft tissue. The cylindrical end cutting needle is the strongest and does not have the risk of bending. The side cutting needle with the thin waist of the stylet can bend even if within the cannula. The first decision to be made with cutting needles is whether to use a manual or automated type; most operators choose an automated

CHAPTER 67  Image-Guided Aspirations and Biopsies

2417

A

D

B C

FIG 67-36  The type and relative size of specimens obtained with four different needles. Specimen A was obtained with three passes of the 22-gauge skinny core needle using 30 mL of suction and a 2-cm insertion of the needle. Specimen B was obtained in one pass of an 18-gauge bevel-end needle using 30 mL of suction and a 2-cm insertion of the needle. Specimen C was obtained with an 18-gauge Menghini needle using 30 mL of suction and a 2-cm insertion of the needle. Specimen D was obtained with the Tru-Cut needle and no suction.

variety. A survey by Mayoral and Lewis showed that more than 90% of radiologists doing liver biopsies use these devices; the only variables are the vendor and caliber of the needle.335 Probably the most important decision to be made is whether to use a single- or double-action needle. The terminology does not refer to the action of the needle, but rather the requirement of the operator. One can understand the two-step mechanism by reviewing the images of the manual type of needle shown in Figure 67-37. A single-action device means that the operator needs to do only one thing: “push” the button or “pull” the trigger. The double-action device requires that one insert the inner stylet with the tissue receptacle and actuate the cutting action of the outer cannula as a secondary step. When the stylet is introduced, it defines the site of the biopsy. The double-action type  is safer than the single-action device when there is critical anatomy close by. In our opinion and experience, single-action devices should be avoided (Fig. 67-38) for reasons described subsequently. When a small lesion in a precarious location is being sampled, one has to be absolutely certain about where the cutting action occurs. An experienced interventionalist knows that if a patient shifts position slightly, the internal anatomy can change by numerous millimeters or centimeters. This change can happen very subtly, and if undetected, catastrophe can occur if one is using a single-action device (see Fig. 67-38). In a case in Ohio that was litigated, a patient died because the main pulmonary artery was biopsied inadvertently. Because a single-action needle was used, the final position of the stylet was not seen because it fired instantly before the cutting action. With the double-action needle, the stylet is inserted separately, and the exact location can be noted. The cutting action occurs precisely at that site. Hemostasis is more effective with use of a hemostatic occlusive method such as the coaxial/coil/ thrombin technique.

The final decision to be made is the caliber of the needle. This decision depends on the amount of tissue required for the histopathologic studies to be performed and the organ sampled. In the new era of tissue biomarkers, routine and possibly repetitive biopsies are expected to become the norm. Workers from our group have provided some insights relative to needle caliber and its relationship to bleeding and the amount of tissue acquired.140,407 Gazelle and colleagues showed that increased bleeding occurs with certain calibers of needles in the normal and coagulopathic states (see Table 67-1).140 From the work by Plecha and associates, one can determine the amount of tissue obtained with each size of needle (Table 67-2).407 Plecha’s group studied the outcome in an animal model using 20-gauge, 18-gauge, and 14-gauge needles.407 These authors evaluated the amount of diagnostic information compared with the amount of bleeding by weighing the amount of DNA (nuclear material) and blood in the tissue sample. They noted that the larger the amount of tissue required for the diagnostic information, the more efficient it was to use a large-caliber needle. Equally important as obtaining adequate samples is reducing the possibility of bleeding. Recognizing that bleeding is the result of multiple factors, including the natural vascularity of the organ (spleen > kidney > liver), the specific vascularity of the tumor, and the random event of striking a vein or artery, one can choose the optimal caliber. Smaller needles require more passes, producing greater statistical chance of striking a vessel. We select the caliber of the needle according to the type of organ being biopsied, the specific type of histopathologic evaluation to be performed, and the characteristics of the organ being sampled (see discussions under each organ system). Assessing vascularity before sampling an abnormality is especially important with a cutting-type needle. In most instances, the vascular character of the lesion is apparent from a prior contrast-enhanced study. If not, one should perform a study with a bolus of contrast

2418

PART III  Image-Guided Procedures

material and repetitive scans at the site of the biopsy to evaluate the vascularity. Not only do cancers have increased vascularity, but unusual benign vascular abnormalities (e.g., arteriovenous malformations [AVMs], aneurysms, cavernous transformation of the veins) occur in virtually all organs; detection of such lesions by a dynamic scan prevents a bleeding complication. From our experience and review of the literature, the vascular nature of an abnormality is a major determinant of bleeding complications. High bleeding rates occur even with 22-gauge needles when vascular tumors such as hepatomas are biopsied.46 In our latest series of 200 liver biopsies, the complication rate with a 14-gauge needle was

0.5% because we assessed tumor vascularity with a dynamic bolus scan over the site of the selected target.106 The lesions that were too vascular or close to vessels were deferred; this study was before our refinement of the coaxial/coil/thrombin technique. It is likely that we would not have deferred any biopsy because of vascularity had we had the technique during this period. As noted earlier, with coagulopathic patients, the smallest caliber appropriate should be used depending on the “contingency” hemostatic method used—systemic blood products, local injection of blood products, or the coaxial/coil/thrombin technique (see “Coagulopathic Treatment”). The systemic replacement to normal status permits any

A

B FIG 67-37  Core biopsy of an apple using the Tru-Cut needle. A, Insert the needle in the closed position to the margin of the lesion. B, Insert the inner stylet with the “gap” (arrows) briskly through the area of the lesion. (If the needle is improperly positioned, it can be withdrawn in this position; a specimen would not be cut unless the outer needle is advanced over the gap.)

CHAPTER 67  Image-Guided Aspirations and Biopsies

2419

C

D FIG 67-37, cont’d  C, With the inner stylet (arrow) held stationary, advance the outer needle rapidly over the gap. (The outer needle cuts the specimen into the gap.) D, A hard material such as bone (in this case, an apple seed) can deflect the thin portion of the needle. This has never been a problem with soft tissue tumors, but the Tru-Cut needle is unsuitable for lesions in or against bone for this reason.

TABLE 67-2  Blood Loss and Amount of

DNA After Biopsy

BLOOD LOSS Needle Size (gauge) 14 18 20

No. Samples

Unadjusted Blood Loss (g)* 1.69 (2.22) 0.74 (0.89) 0.32 (0.50)

64 63 63 AMOUNT OF DNA

Needle Size (gauge) 14 18 20

No. Samples

Amount of DNA per Sample (µg)*

17 16 16

40.38 (10.29) 12.18 (3.84) 5.86 (1.58)

*Data are from men. Numbers in parentheses are ±1 SD.

size of needle, as does the local injection method. The coaxial/coil/ thrombin technique requires a coaxial needle system of 18 gauge or larger; we defer to the 18 gauge. Numerous clinical articles have studied the histologic information obtained from tissue samples for kidney and other organs. Hopper and colleagues showed that sufficient diagnostic tissue could be obtained from the kidney with an 18-gauge needle in pediatric patients.228 Haaga and associates found in a comparison of 18-gauge and 14-gauge needles that the diagnosis was equivalent for both devices.194 A survey of gastroenterologists and radiologists showed that most physicians use an automated side cutting needle in the range of 18 gauge.335 Although the use of the 18-gauge needle is the norm in the United States, there is good reason for radiologists to begin using large-caliber needles in selected cases. In recent years, new methods developed at our institution in conjunction with the National Cancer Institute portend to enhance the development of new cancer agents106,472,473,514 and drugs by using samples from 14-gauge needles in selected cases. As reported by Dowlati and coworkers, we had a 0.5% bleeding rate

2420

PART III  Image-Guided Procedures 10

0 cm

R

L

A 10

0 cm

R

L

importance of judgment when selecting them to achieve the greatest tissue yield, using proper patient selection, approach, and techniques to minimize the potential complications. We have not changed our opinion on the superiority of the tissue samples, but rather we are concerned about the cavalier attitude some operators have for minimizing the possibility for complications. We have noted that some physicians use the cutting needle with impunity and put patients at great risk and themselves in great legal peril. Lest one misconstrue these statements as indicating that we have a high complication rate, our complication rate (according to the National Cancer Institute, which monitored all of the NIH studies and reviewed the charts) was 0.5% in a group of 200 patients (see “Liver” section).106 Figure 67-39 presents a case that shows the exact opposite of all our recommendations. This case was performed by a community radiologist in a nonuniversity hospital and resulted in two significant lifethreatening complications. The radiologist used an 18-gauge cutting needle for a deep lung lesion, which predisposed to severe hemoptysis (a 20-gauge cutting needle should be used for such deep lesions). The cutting needle was deployed with the inner stylet outside of the mass and adjacent to a vessel (see Fig. 67-39). The patient had a large pneumothorax and large hemorrhage in the right lung; the pathologic diagnosis was inconclusive (not a surprise, seeing that the needle was not properly positioned in the mass). The reader is urged to take our recommendations and suggestions because we have performed these procedures for 33 years, including approximately 20,000 personal experiences. Everyone must develop their own guidelines, but it is always better to err on the safe side.

Needle Techniques B FIG 67-38  Patient at another institution who had biopsy by singleaction needle and died as a result of biopsy of the main pulmonary artery and subsequent bleeding (we testified in the lawsuit). A, Preliminary localization with a single-action cutting needle shows proper positioning of the needle in front of the mass. B, CT scan after the biopsy shows hematoma in front of the main pulmonary artery (arrow) after the biopsy. Because the patient moved slightly during the procedure, the internal anatomy shifted slightly so that when the single-action needle was fired, it cut the main pulmonary artery, resulting in hemopericardium and death. With the single-action needle, the site of the biopsy is never seen; only the targeting is seen. Double-action needles permit confirmation of the needle stylet in the target zone before firing.

with 14-gauge cutting needles in the liver in a series of 200 patients.106 In these trials, large needles were used to recover histologic samples, suitable for microchemical assays of molecules that were the specific target of cytolytic or modulating agents. These trials permitted more rapid drug development at lower costs. Improvements in determination of pharmacokinetics and dosage regimens occurred to the extent that a new term, the BMD (biochemical modulatory dose), was coined. In some cases, this term was substituted for the MTD (maximum tolerable dose before unacceptable complications occur).106 Possible overuse of cutting needles.  Having innovated cutting needle biopsies using CT guidance and advocated their use for many years, it is ironic that we are now concerned that cutting needles occasionally are used too frequently by some operators. We have always been strong advocates of their use but continue to emphasize the

Since the prior edition, significant data have been published that confirm a higher incidence of needle track seeding than previously reported. In view of this, the single-needle technique probably should be reserved for lung lesions (hemoptysis and pneumothorax are more likely to occur with larger needles). Even mediastinal biopsies are  well suited for the coaxial system because when the cannula has been carefully placed past the major vessels, repeat biopsies can be taken if needed. The single-needle technique is simple because it uses the primary instrument as the device primarily localized. This technique is typically used when one anticipates taking only a single sample. The usual clinical scenarios differ with the aspiration needle and the cutting needle. With a Chiba needle and a lesion that is large and accessible, it is easy to perform several procedures as needed without concern about repeat positioning of the needle. With large cutting needles, a 14-gauge needle usually is used to obtain a sample; seldom are multiple samples needed in such a circumstance. Coaxial methods offer significant advantages. In the prior editions, we made a distinction between a short cannula versus a long cannula, but in view of needle track recurrences reported, the long cannula method is the only one that should be used. There are three benefits to the coaxial cannula method: 1. With the cannula in place, one can take repetitive biopsies of skinny needles and larger cutting needles without requiring relocalization of the needle. 2. If one encounters a localized bleeding problem or question, it is simple to occlude the pathway with an angiographic coil/thrombin even if the bleeding was not anticipated. 3. The coaxial system prevents tumor seeding in the needle track. Maturen showed that the coaxial method reduces the occurrence of needle track seeding because the guidance cannula prevents shedding of tumor cells in the pathway.332

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2421

B

C FIG 67-39  A, Patient from another hospital had a cutting needle biopsy improperly performed from two perspectives: (1) the mass (arrow) is smaller than the cutting gap of the needle, and (2) the needle is adjacent to the mass and not within. The cutting of the needle occurred in the lung and not the mass. Hemorrhage has occurred posteriorly from the initial placement. B, The biopsy was taken, which produced more bleeding posteriorly that was aspirated into the posterior bronchus. C, The lung subsequently collapsed with more bleeding around the mass (arrow). This patient survived after emergency resuscitation.

BASIC SAMPLING METHOD FOR ASPIRATION AND CUTTING BIOPSY Aspiration Technique The standard aspiration procedure is simple and can be broken down into four individual steps: (1) localization of the needle tip in the lesion, (2) application of suction, (3) movement of the needle, and (4) delivery of the sample to the appropriate medium for evaluation. Repeat needle passes are made as appropriate. Localization of the needle tip in the site of the mass is crucial and likely the most important step. Having the tip close to the target does not suffice. If small changes are necessary for correct placement, the needle deflection methods described subsequently may be helpful. Application of suction is important because it enhances recovery of adequate cellular material. The amount of suction to apply has been evaluated by our group and others.119,233,272 In a laboratory model, we found that displacing the syringe barrel by 5 mL produced about 84% of obtainable sample, and that displacement of the barrel by 10 mL produced 94% of the obtainable sample (Fig. 67-40). Kreula and associates had similar findings; there was an “apparent” slight discrepancy

between our data and Kreula’s data,272 which can be easily reconciled. Kreula’s group used absolute pressures, whereas we used a clinical analogy of displacement volume of syringe during application of suction. Our curve is asymptomatic, reaching a limitation after a linear rise. Kreula and associates reported a straight-line relationship between the sample and the vacuum because they measured only the effects corresponding from 0 to 5 mL of syringe displacement.272 Suction can be applied by two different methods. A two-person technique can be used with one person applying suction and the second holding and moving the needle. The usual approach entails plastic connecting tubing between the needle and the syringe. This is more convenient from a personal perspective, but it actually may provide less suction because the plastic tubing is pliable: when suction is applied, it partially collapses, reducing suction. This method was first applied by Menghini in the 1950s but was later abandoned for the one-person method. The one-person method has the radiologist applying suction while at the same time stabilizing and moving the needle. One important point is to maximize the vacuum during the movement of the barrel; it is important to ensure that all of the air has been expressed from the syringe before hooking it up to the hub. Even a small amount of air dilutes the vacuum when the barrel is moved.

2422

PART III  Image-Guided Procedures 20-GAUGE NEEDLE 18-GAUGE NEEDLE

6

25 5 Sample size (mg)

Sample size (mg)

20 15 10 5

4 3 2 1

5

10

A

20

30

5

B

Suction (cc)

80

20

Average sample size (mg)

Maximal pressure change (%)

100

60

40 – 20-gauge – 18-gauge

5

10

20 Suction (cc)

30

VARIATIONS IN SUCTION APPLICATION 25

1

C

20 Suction (cc)

THEORETICAL PRESSURE CHANGE WITH CHANGE IN SUCTION

20

10

30

20 cc c release 20 cc s release 20 cc c pause

15

10

5

1

D

2

3

4

5

Specimen

FIG 67-40  A, Sample size versus suction, 18-gauge needle. Each point equals an average of five samples. B, Sample size versus suction, 20-gauge needle. Each point equals an average of five samples. C, Theoretical pressure change as calculated by Poiseuille’s law for noncompressible fluids (Q = τ8(Pa4/l)n; Q = volume of fluid through cylinder, P = pressure difference at ends of cylinder, a = radius of cylinder, l = length, n = viscosity of fluid) and Boyle’s law for ideal gases (PV = PV at constant temperature; P = initial pressure, V = initial volume, P = final pressure, V = final volume). D, Specimens 1 to 4, average of 5 human samples; specimen 5, average of 10 bovine samples; c = with, s = without. (D, From Hueftle MG, Haaga JR: Effect of suction on biopsy sample size. AJR Am J Roentgenol 147:1014–1016, 1986.)

Historically, some groups advocated the use of no suction during the sample acquisition. This “capillary” method works with very necrotic tissue such as squamous cell tumor, but not with most tumors. Some authors have asserted that the application of suction pulls too much extraneous material, especially blood, into the needle, but a more recent study by Ha and coworkers showed that such degradation occurred in only 6% of cases176; this slight tradeoff is minimal compared with the improvement of the sample. Depending on the fixative chosen, the red blood cells can be lysed by certain additives to the fixative material, eliminating the problem. Movement of the needle should be made to free tumor cells for the sample, with the needle tip confined to the tumor site. The movement is best performed by very small reciprocating motions of 2 to 3 mm,

which enhance displacement of cells into the needle barrel. Movement of several centimeters is excessive because the needle tip moves in and out of the mass into the adjacent normal tissue. This movement introduces blood into the sample and injures the adjacent tissue. A common error by inexperienced operators is making these large movements until blood is aspirated into the syringe or plastic tubing (usually occurs with the two-person technique). During the removal of the needle, it is important to maintain the suction to retain any fragments in the lumen of the cannula. The in vitro study by Hueftle and Haaga of our group documented the definite benefit of this movement (see Fig. 67-40).233 The diagnostic sample should be delivered on the slide and into a separate solution for washing the needle and syringe. A “perfect”

CHAPTER 67  Image-Guided Aspirations and Biopsies sample is orange, owing to a small amount of blood and a large number of white tumor cells. After a sample has been taken, it is crucial that it be handled properly. The smear is performed quickly near the fixative to prevent drying, which can occur quickly. The syringe and needle are always flushed with fixative, and the fluid is sent for tissue cell block or cytocentrifuge, depending on the routine approach at an institution. This practice can increase the diagnostic yield by several percentage points. Kasugai and colleagues reported that heparinizing the needle and syringe is helpful,252 but no other authors have confirmed this. Other authors have measured the carcinoembryonic antigen in the cytologic samples and found this test useful.32 Repeat needle passes (four to five passes) have been recommended by some authors238 to enhance diagnostic accuracy. From our experience, equivalent diagnostic accuracy can be achieved with two wellpositioned, meticulously performed aspirations or a small-caliber cutting needle; this has been well documented by Smith and associates460 and Sargent colleagues433 (see “Brief Historical Summary of Image-Guided Biopsies”). The emphasis on taking many needle passes has evolved incorrectly from misinterpretation of an article by Ferrucci and colleagues.124,126 These authors took a sample of 20 cases, proven positive by cytologic results, and looked at the incidence of positivity with each needle pass. At first glance, one would believe that 100% accuracy could be achieved with many needle passes, but this is not true; 100% of this group had positive results because they were selected as positive at the onset. In the same study, the total group with five needle passes had only an 82% accuracy, not the 100% of the selected group.335

2423

recovery. To prevent push-back of the needle and a smaller sample, the device should be held firmly. Kellermeyer found that holding the needle firmly would improve the yield by reducing the recoil (personal communication, 1993). In his unpublished data, there was a 14% improvement of the yield. Another anecdotal factor is that one should hold the device exactly as explained in the instructions. If an operator holds such a device casually without the fingers in the correct location, the movement of the actions may be impeded and the sample less than optimal. Several additional important comments about side cutting needles should be made relating to the stylet for acquisition and safety. First, when the cutting needle is removed from the packaging, one should inspect the stylet with the cutting gap. Occasionally it may be bent slightly forward (Fig. 67-41A) and should be straightened or bent

Side Cutting Needles As noted previously, the best cutting needle for soft tissue tumor is the single-action side cutting needle, and the best cutting needle for bone or hard tissues is the end cutting needle. The side action cutting needle is based on the original manual Tru-Cut design. This manual device (see Fig. 67-37) is the prototype of virtually all cutting needles used today. We have retained this figure even though it has been used in numerous editions of this book because it clearly shows the mechanism used in all manual or automated side cutting needles. By understanding the basic mechanism, the various devices can be better understood, especially the significance of the single- and double-action needles. Side cutting needle advantages are severalfold. As affirmed by numerous in vitro tests conducted by ourselves and Hopper’s  group,227-229 this design provides the most complete and preserved samples of all needle types. Also, because the location of the biopsy is confined to the “gap” of the needle, one can carefully localize the site of the biopsy. The action of the needle is so precisely predicted that it can be used near other structures. Another advantage of this needle is that if a double-action needle is used, the tissue gap of the stylet can be moved or adjusted as needed without any adverse effects on the tissue. The cutting action occurs only when the outer cannula is advanced forward manually or by the power device of the automated devices. Catastrophic events as noted in Figure 67-38 can be avoided. Automated needle use can be optimized by several approaches. These approaches improve the diagnostic yield and minimize the possibility of complication. To improve the diagnostic yield, several additional factors and steps can be taken. The gap of the stylet defines the precise site of the needle biopsy; although the sliding cannula cuts the tissue, it moves over the stylet. One can precisely determine the location of the biopsy, which optimizes tissue acquisition and minimizes collateral damage. When automated devices push the stylet forward, there is a backward recoil of the device, which may affect tissue

A

B FIG 67-41  A, Before using a side cutting needle, it is always prudent to inspect the end of the stylet for any manufacturing problem. One frequently observed finding is that the stylet of the needle may be slightly bent forward; this impairs the cutting action when the cannula slides forward. We have found that if the tip is bent back slightly, the sample acquisition is improved. B, Sample container showing two tissue cores, one obtained when the stylet of a side cutting needle was used with the stylet forward as it came out of the box and the second large piece after the stylet was bent back before sample was obtained. The larger sample is almost twice as large.

2424

PART III  Image-Guided Procedures

slightly back. If the needle is bent forward, a smaller sample would be obtained than if the stylet were bent to the proper shape (see Fig. 67-41B). Second, there is one word of caution about these needles relative to bending: These needles should not be used for very hard tumors or bone. The thick waist of these needles can bend if too much force is applied, even if the needle is closed (Fig. 67-42).

Some new scanners have a cluster scan feature consisting of three or more scans, which is well suited for this purpose. When the end of the needle is found, it appears to have a square end and may have an artifact from the end. Occasionally the end of the needle is not square but is easily appreciated by noting when the end is no longer seen on sequential scans.

Needle Localization and Positioning

Unintended Needle Deflection.  With a bevel needle, the needle

mality is the simplest approach if possible. In these cases, with the needle in the superficial tissues, the angle of the needle is adjusted with repetitive scans until angled correctly. When the correct angle is achieved, the device is inserted to the desired depth either with incremental adjustments and repeat scans or to a premeasured distance. Subtle adjustments of the needle can be made using the deflection principles, or cannula manipulations noted in the following sections may be used meticulously to optimize the needle position.

either moves straight or deflects depending on the hardness of the mass, angle of the needle, and speed of insertion. A hard mass is likely to deflect the needle if the trajectory pathway is not perpendicular to the tangent of the entrance point. If a beveled needle is pushed slowly, it is more likely to move the point away from the flatter part of the needle. If a needle is pushed quickly, it tends to go straighter. For very hard masses, deflection may not be avoided unless one continuously rotates the needle during insertion or a stiff guidance cannula greater than 18 gauge is used. In moderate-density material, controlled deflection is possible.

Axial Slice.  Positioning a needle in the same xy plane of an abnor-

Localization of the Needle Tip.  After insertion of the needle, one

Intentional Needle Deflection.  With the intentional needle deflec-

occasionally may have some difficulty in localizing the tip. Several methods can be used to solve this problem. The simplest method is to take a longitudinal scan and take an axial slice at the end; this is virtually foolproof and is the most expeditious. Another method that can be used if the angle of the needle is slight involves taking multiple scans in one direction until the end of the needle is reached (Fig. 67-43).

A

tion method, one can change direction of a needle when it is located in a patient below the surface of the skin. Purposeful deflection can change the angle of the needle at the latest stage of adjustment before striking the mass. This deflection is best accomplished by rotating the needle so that the flat side of the bevel is facing the area one wishes to

B

FIG 67-42  A, CT scan shows a bend in the middle of the gap (arrow) that occurred while it was in a patient. The fibrous material was so hard it bent the needle. B, CT scan shows the bent needle after it was extracted.

A

B

FIG 67-43  Localization of the needle tip may be problematic at times. A, When the needle is tapered, it means that a variable number of pixels is being occupied along the length of the needle, and the end may not be seen. B, When the end is square, the tip is being accurately localized.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2425

B FIG 67-44  A, Diagram shows a schematic where the needle is not accurately positioned adjacent to the mass. B, Diagram shows how, by using the bevel of the needle, it can be redirected toward the mass by pulling the skin away from the point and pushing the needle slowly.

avoid, and the point is directed more toward the target direction. The skin adjacent to the needle is pulled away from the point. This changes the fulcrum of the needle so that it is more angled toward the target (Fig. 67-44). The needle is then pushed slowly. By using this method, one can change the direction of the needle by 10 to 15 degrees, which can amount to a 10-mm change; this can prevent the need for a major readjustment or reinsertion of the needle (Fig. 67-45). This technique works only with a needle with a single bevel at the tip and not with needles with two bevels or a central point, such as a pencil or trocar point. This technique also can be used with the thin stylet of a cutting needle (Fig. 67-46), which permits repositioning without removing the entire needle. To achieve this repositioning with the double-action technique, the needle is positioned adjacent to the target, slight pressure is used to deflect the shaft of the needle away from the bevel, and the stylet with the bevel is pushed slowly into the target.

Coaxial Cannula Manipulation.  Because the coaxial approach has such benefits as described earlier, one should consider the possibilities with the guidance cannula. Because the guidance cannula is stiffer than the basic needle, it can be used to torque the needle. This is done by simply angling the cannula in the desired direction. One must be aware that such devices are stiff, and moving such a device potentially can produce shearing. The intended deflection noted earlier can occur with a coaxial cannula. Another technique that can be used to move a needle into a difficult area combines the benefits of a Chiba needle with a cannula. For lack of a better term, we have called this a cannula slide maneuver. If one wishes to use a cutting needle beyond a vital structure, one can advance the cannula in the following fashion. Through the cannula, one pushes a bevel Chiba needle beyond the vessel. Then one advances the cannula forward over the small-caliber needle. When the cannula is distal to the vital structure, one can use the cutting needle safely. We have used this technique in the mediastinum and the pelvis (see Figs. 67-77 and 67-131) in lung sections.

Rotation of Cutting Needle to Optimize Sampling.  In some cases when a biopsy is being attempted with a side cutting needle, the placement is not completely optimal because the “needle gap” misses the mass slightly. In such cases, one can increase the probability of getting an adequate sample by rotating the needle so that the gap is oriented toward the center of the mass (Figs. 67-47 and 67-48).

Angled or Out-of-Slice Targeting When a needle must be inserted into a lesion located at a different z axis position, different approaches must be taken, 3 of which are: (1) freehand, (2) angle estimated, or (3) device assistance (see “CT Fluoroscopy, Guidance Devices, and Robotics”).

Freehand Method.  With the freehand approach, it is best to consider the approach to the lesion as a 3D system. The final vector or angle of the needle to the lesion would be a combined angle consisting of the xy angle and the yz angle. When doing this, it is best to establish the xy angle in the superficial tissues of the lowest slice before attempting the z angulation. When the xy angle is proper, one can consciously hold that angle constant while moving the needle into the z axis (Fig. 67-49). When the needle is close to the target site, numerous scans are made over the path of the needle to judge the position in 3D. If a minor adjustment needs to be made, one should use the tip deflection method described earlier under “Intentional Needle Deflection.” Using this technique after the first major angulation has been accomplished saves time and makes success more likely.

Angle-Estimate Method.  The angle-estimate method is a permutation of the freehand method, but more care is given to trying to calculate or simulate the angle more precisely. Most simply, this is done by using a calculation of the Pythagorean angle from the scans to arrive at a numerical angle in the different planes (Fig. 67-50). The execution of the instrument insertion is performed by the freehand method or by using a manual angulation device.

Intentional Organ Movement or Tissue Retraction

Patient Positional Changes.  Placing the patient in lateral or oblique positions produces shift of internal anatomy so that loops of bowel or organs may move sufficiently to open a procedure trajectory pathway. In most instances, bowel shifts to “float,” and solid organs “gravitate” or “roll” depending on capsular attachments. Although one might predict such movements, trial and error may be needed. If one positional change does not work, rotation to the oblique or the opposite side might produce the desired effect.

Injection of Gas or Fluid.  Although one can penetrate certain loops of bowel and uninvolved organs in some circumstances, it may  be advisable to avoid certain structures or to protect them from

2426

PART III  Image-Guided Procedures

A FIG 67-46  Diagram shows the same result of deflecting the needle tip with a side cutting needle. By pushing the needle at an angle and then slowly pushing out the stylet, the bevel of the stylet can direct the needle into a mass when there is a slight trajectory error.

A B

B FIG 67-47  In some circumstances, the “gap” of a cutting needle is close to the edge (A), and one cannot be completely certain that the cutting side would obtain a suitable sample. In such cases, one can optimize this possibility by rotating the needle 180 degrees so the gap faces the tumor (B).

C FIG 67-45  Three radiographs show the position of a needle tip (arrowheads) that is moved by the deflection method (see text). A, The needle tip is adjacent to the lower end plate of the L4 body. B, The needle has been retracted back slightly before the deflection maneuver. The skin was retracted and the bevel rotated cephalad, and the needle was pushed forward. C, The needle tip (arrow) was deflected caudad almost 10 mm to the top of the L5 end plate.

unintended damage during ablation procedures. Two methods have been used for dealing with this problem: injection of carbon dioxide and injection of saline. We reported in 1986 that carbon dioxide insufflation or room air can be used to move structures short distances.184 Other authors have proposed the use of injected saline to increase the size of pathways through which one can insert a needle.167 This technique has been used predominantly in the chest for widening the paraspinal space, but potentially it could be used in other locations (see later under “Lung and Mediastinum”) (Fig. 67-51). Langen and colleagues studied the efficiency of carbon dioxide and saline in an animal model to move structures from a needle trajectory pathway.286 With equivalent amounts of gas or saline, they found that the saline produced substantially more movement of organs than carbon dioxide. Neither fluid nor gas moved solid organs, such as kidneys.

More recently, many authors have reported using such materials to prevent inadvertent damage to adjacent structures during ablation procedures. Authors have reported the use of saline or water to provide a conductive layer between the thermal ablation site and any adjacent normal structures. Laeseke and coworkers reported the use of dextrose water and saline to prevent thermal damage to adjacent uninvolved structures.277 In their animal model, dextrose prevented damage better than saline, permitting smaller sites of damage (perhaps because of conductivity of saline). Kariya and associates reported the use of injected carbon dioxide to produce “insulation” between the thermal site and normal structures.251 The maximum distance was slightly greater than 1 cm. Caveats for saline injection include patient discomfort, decrease in target lesion conspicuity, and adverse effects of absorbed fluid in patients with kidney or heart disease. Both of these methods have a similar problem. Depending on the location of connective tissue, either the gas or the fluid may dissect cephalad or caudad and not produce substantial enlargement in the axial plane. Nevertheless, both of the methods can work in selected cases. Another caveat is the effect of severe inflammation or tumor infiltration. If either of these is present, movement or displacement may be impaired. The procedure is performed easily. After the trajectory is planned, the operator decides from the scans the proper location of the gas or

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

2427

A

B

FIG 67-48  A, CT scan through the liver shows the gap of the needle faintly visible in the lesion. B, By rotating the needle, the gap is turned so that the cutting edge is toward the mass. There is a greater probability that an adequate sample will be obtained.

liquid. The operator inserts a needle and injects by hand 50 to 100 mL of the material (Fig. 67-52). If the intended loop of structure does not move, the operator attempts to reposition the needle or repeats the injection. Depending on the length of the procedure, repetitive injections might be necessary because the material disperses gradually along fascial planes, and the displacement decreases accordingly.

External Manipulation of Bowel Loops.  In some patients, it may be possible to move an intervening loop of bowel by manual palpation through the skin. Anyone who has ever performed fluoroscopy is familiar with how mobile loops of bowel may be. Tuncali and colleagues used manual displacement of bowel to provide greater clearance between the iceball of cryoablation and adjacent bowel loops to prevent collateral damage.493 In a group of 14 patients, mean distance between tumor margin and closest adjacent bowel increased from 0.8 cm (range 0-2 cm) before external manual compression to 2.6 cm (range 1.6-4.1 cm) with manual displacement (P < .01). Occasionally we have used this technique for biopsies, and it has been beneficial.

CT FLUOROSCOPY, GUIDANCE DEVICES, AND ROBOTICS CT Fluoroscopy Real-time monitoring of instrument insertion by rapid CT acquisition is now possible. First conceived by Katada of Fujita Health  University, Japan, the technique was implemented by Toshiba and subsequently by other vendors. This technique has been used only occasionally by radiologists; some estimate that only about 30% of

C FIG 67-49  A, An entrance site below the mass is chosen so that the needle is introduced below the diaphragmatic insertions. The XY angle is approximated, and the needle is inserted partially. B, Higher CT scan at the level of the cyst (curved arrow) shows that the trajectory of the needle (straight arrow) is slightly too medial for puncture. The needle was withdrawn slightly and the bevel rotated, as shown in later scans, so that the point is directed laterally and the needle is pushed slowly to facilitate the subtle bending. C, Final CT scan shows the needle centrally in the cyst. Fluid confirmed the simple nature of the cyst.

radiologists use it. The reason for the low usage is unclear, but it is likely related to inexperience and the potential higher radiation dose to the patient and the operator. Logically, if radiologists are capable of high-quality effective procedures without direct exposure to themselves, they would not use CT fluoroscopy. With the growing interest in screening CT and earlier diagnosis of tumors, however, many more patients are being triaged for percutaneous needle biopsy. The issue of respiratory variation becomes much more significant with small lesions, so we believe that this tool will be used more and more for biopsy procedures.

2428

PART III  Image-Guided Procedures A

C

B

FIG 67-50  Calculation of the distance to the lesion and angle of entry when the entrance point to the skin and the target lesion is not in the same CT section. When the entrance point C and distances A and B are known, the distance to the lesion and the angle of insertion can be calculated. Although theoretically this method has merit, pragmatically its use is limited because of the difficulty in duplicating the angle freehand. We prefer to estimate the angle with such lesions and take a series of sequential scans or use the longitudinal scan method.

Guidance Assistance Devices Numerous devices have been conceived, developed, and commercialized for guidance assistance, but these have not become a part of the mainstream practice of most radiologists. More sophisticated ancillary guidance devices are being designed; these are expected to simplify the performance of CT-guided, US-guided, and MRI-guided procedures. One device introduced by Magnusson and Akerfeldt323 seemed to be effective, but for unknown reasons, attempts to market the device were unsuccessful. The likely reason that these devices have not succeeded is that interventional radiologists have unique skills and are well trained. Most develop a skill set that permits them to execute such procedures almost freehand or using the simple methods described earlier.

Robotics With high-resolution MDCT scanners, 3D reconstructions, and the progress in robotic systems, there is no question that in the remote future robot-assisted procedures will become available. The robot will truly facilitate and not replace the physicians. A caring physician/ radiologist will always be needed on site to use the robotic assistances, monitor the procedure, and manage the patient, but targeting lesions will be much easier. In our opinion, such procedures are not likely to be done remotely from across the country or world because caregivers must be present to manage the psychological and physical needs of the patient. We have performed some initial testing using a standard industrial robot and have found it to be promising (see Fig. 67-1). The optimal device has yet to be designed, but there is little question that such a device will become available. In a swine model, we tested a robot for the execution of a celiac nerve block. The execution of the procedure was faster and more precise than we ever imagined. In contrast to manually performed procedures, during which needle adjustments are made below the skin surface to account for small errors during the insertion, the robot executed the insertion quickly without error (see Fig. 67-1). The final configuration and type of robot have yet to be determined, but its time will surely come. As already stated, robots will never replace the interventionalist because patients require the physical presence of a sympathetic and skilled physician. It is likely that the final device will be an “assisting device” controlled by a physician in the room to direct the robotic hand into the radiation beam during CT fluoroscopy. With a physician

nearby in the room but away from the x-ray source, the physician can serve as a failsafe in case of instrument failure.

LUNGS AND MEDIASTINUM Image-Guided Background The first experience using skinny needles with image intensification guidance in the lung was reported by Dohlgren104 and Nordenstrom380 in the 1960s. Sargent and colleagues433 also performed biopsies without image intensification, with good results. When image intensification became standard, many authors, including Lauby and coworkers,288 Lalli and associates,279 Westcott,511,512,513 and Sinner,454,455,456 reported their series. Many important observations were made about different factors that alter diagnostic accuracy and complications, which are discussed in the following section on lung procedures. Although the early studies were typically small, were not designed well, and did not assess the proper variables, many different calibers of needles and techniques were used with similar results. During this time, there was an obsession with the assumption that the most important factor was the needle caliber. Regardless of the caliber of the needle used by the authors (e.g., Lalli and associates279 used an 18-gauge needle, and Zornoza and colleagues535 used a 23-gauge needle [called “skinny needles”]), the diagnostic accuracy of all reports was generally good, and the complication rate was acceptable. Further progress for sampling the lung and mediastinum occurred with our introduction of CT lung procedures in 1976 (see Fig. 67-3B).177 Although traditional fluoroscopy was previously used for well-defined peripheral lung nodules, it has receded into obscurity. US has become more frequently used for sampling some mediastinal and pleural lesions when there is no lung intervening in the path that obscures imaging. CT has become the mainstay, probably because of its inherent advantages for accurate needle placement and avoidance of complications. Precise visualization of the needle tip permits precise localization of the needle into selected portions of a lung mass, increasing the probability of a positive histopathologic interpretive result. Trajectory selection is easier, and one can choose the shortest pathway and certain anatomic features of specific lesions to avoid complications.

Indications The clinical situations that justify the performance of a biopsy include sampling of nodules for neoplasm or sampling of infiltrate for

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

C

D

FIG 67-51  A, CT scan taken during a biopsy of a middle

E

mediastinal mass in the area of the aorticopulmonary window. The trajectory chosen to reach the mass is through the anterior junction line but is difficult because of the narrow pathway. B, Injection of saline at the site moves the pleura 7 to 8 mm laterally. This widened pathway permits passage of a needle. C, Chiba aspiration biopsy needle was passed through the parasternal path into the mediastinal mass. The lung parenchyma was totally avoided. D, The same pathway was used for passage of a cutting needle. The needle is seen before insertion of the cutting stylet. E, The stylet is inserted with the bevel turned toward the mass so that the stylet deflects laterally away from the margin of the aorta.

2429

2430

PART III  Image-Guided Procedures

T

A

C

B

FIG 67-52  A, CT scan shows a tumor (T) in the right lower quadrant. The needle (arrow) is inserted adjacent to the bowel, which is to be moved. B, Air (10-15 cc) is injected quickly to push the bowel away. The gas with the needle is noted adjacent to the bowel (arrow). The metal needle shaft is not well seen because of the partial volume of air with the metal needle. C, With the bowel moved away, the needle could be inserted into the mass.

infectious agents. One can justify performing a biopsy of a suspected primary nodule for two reasons. First, aspiration biopsy can distinguish small cell tumors from non–small cell tumors.456 Many oncologists believe that small cell tumors should be treated medically rather than surgically, or at least with combination therapy. Second, Sinner stated that percutaneous biopsy can expedite surgical treatment.454 For metastatic or recurrent tumors, confirmation of malignancy provides tissue for palliative therapy or closure if hospice care is chosen. In planning patient management, more recent data indicate that there is a higher incidence of tumor recurrence in the pleura of the biopsy side, but coaxial needle methods have been shown to eliminate this problem. In cases of pulmonary infiltrate when the causative organism is unknown, diagnostic aspiration can obtain material for culture. Although the results of these procedures are less accurate than for tumor, such procedures can distinguish among pyogenic, fungal, or opportunistic organisms.

Imaging Preferences

CT.  CT is the modality of choice for the biopsy of most lung lesions regardless of their locations. Because the shape and angle of the ribs varies from the front to the back of the ribs, approaches for biopsying different lesions must be varied (Fig. 67-53), as noted in later sections. Lesions that were problematic for fluoroscopy, such as lesions in the apices (Fig. 67-54; see Fig. 67-35), costophrenic angles (Fig. 67-55), costovertebral angles (Fig. 67-56), pleurae (Fig. 67-57), and hila, and lesions obscured by fluid and mediastinum, are well suited for CT. Parenchymal lesions with ill-defined edges are well seen, and one can distinguish necrotic areas or contiguous areas of infiltrate adjacent to the solid portions of a mass. Avoidance of complications is improved because visualization of anatomy is excellent, and one can exploit anatomic detail to avoid problems in patients with parenchymal disease or vascular anomalies or blebs. In such cases, if one takes thin scan sections, one can sometimes find a pleural attachment that permits a “risk-free” access, avoiding penetration of the lung (Fig. 67-58). In patients with an unknown or unexpected vascular anomaly, bolus dynamic CT scan is an important adjunct for the mediastinum and hila because major vessels can be opacified and delineated, making separation from masses in these areas easy.

PET Scanning.  PET scanning has become a valuable tool for the diagnosis and characterization of lung masses and is commonly a

F

A

FIG 67-53  Sagittal CT scan close to the central chest shows the relationship of the ribs. Anteriorly there is fusion of several costal cartilages. The ribs are much closer posteriorly than anteriorly. In addition, the narrowing of the ribs as they approach the apex shows how the interspaces become more narrow owing to the close proximity of the ribs.

FIG 67-54  Lesion in anterior apex adjacent to the mediastinum. A needle can be noted anteriorly entering the mass without traversing any intervening lung. In such cases, one must be careful to avoid the subclavian vein, which was at a level slightly higher.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2431

B

FIG 67-55  A, CT scan shows a mass adjacent to the costophrenic angle of the right lung, slightly above the diaphragm, making it very mobile with respiration. Note the many small cysts behind the mass. B, The patient is in the oblique position to immobilize the diaphragm and mass partially. The needle was inserted laterally to avoid the cysts, which might cause a pneumothorax.

FIG 67-57  CT scan showing biopsy of a pleural-based mass. Rather FIG 67-56  Mass in the posterior paraspinal region can be approached without traversing the lung parenchyma.

prime indicator for biopsy. Cancer is hypermetabolic and consumes larger amounts of radioactive glucose than benign or noncancerous masses (Fig. 67-59). In most cases, PET is unparalleled for characterizing the cancerous nature of pulmonary nodules. Its use  has revolutionized the practice of medicine relative to pulmonary  nodules. Occasionally an inflammatory site or nodules in the lung may  be problematic because of associated inflammation. Although  delayed scans have been used in the PET protocol to reduce the uptake and differentiate cancer from inflammation, this has not always been effective. After encountering numerous cases intended for percutaneous biopsy, we used a new technique to reduce the inflammatory reaction to eliminate the question of tumor in such cases. In nine patients referred for biopsy, we used a new technique that is not yet FDA

than a straight approach, an oblique longer path was chosen, which increases the chance of positive results (see text). Arrow shows the “gap” of cutting needle.

approved. For problematic cases, this new technique can reduce the inflammatory vascularity and FDG uptake in inflammatory foci sufficiently to eliminate some false-positive findings for inflammatory pulmonary nodules (Fig. 67-60).

Contraindications Refinement of techniques has reduced the number and significance of contraindications to image-guided biopsy of the lungs and mediastinum. Standard contraindications include very severe lung disease, congestive heart failure, severe anemia, general debilitation, or a very uncooperative patient. A patient with severe lung disease may have such pulmonary compromise from decreased compliance and air exchange that chest tube placement would not successfully re-expand the lung, making resuscitation impossible.

2432

PART III  Image-Guided Procedures

Patients with severe congestive heart failure are excluded for two reasons. First, having their cardiopulmonary status compromised by output failure, these patients are unlikely to be able to tolerate a significant pneumothorax. Second, because of the increased pulmonary vascularity owing to poor output function, the vessels are larger and more edematous, making severe hemoptysis more likely. Severe anemia reduces the oxygen carrying capacity of the blood so that a pneumothorax can induce severe anoxia. Such anoxia could lead to serious metabolic issues, making resuscitation difficult. Very uncooperative patients cannot follow instructions during such a procedure, such as staying immobile and holding their breath. Inability to comply with such instructions can increase the chance of a severe pneumothorax or hemoptysis, making resuscitation very difficult. Absolute contraindications include uncorrectable coagulopathy, severe pulmonary hypertension, uncontrollable coughing, possible Echinococcus infection, and an uncooperative patient insufficiently controlled by anesthetics.

Laboratory Values The levels of the coagulopathy that we find acceptable are more restrictive than for other organ sites. We insist on an INR of 1.0 and a platelet count of at least 50,000/µL. Unlike in other solid organs that are not

air filled and can provide local tissue tamponade, normal values are important because the air-filled lung does not provide local tamponade for bleeding. This situation can result in serious bleeding and even intrabronchial bleeding and death. Uncontrollable coughing and a patient with an uncooperative nature are important because movement of the lung during a procedure may cause a tear of the lung by the needle, which might cause air embolism, pneumothorax, or bleeding (see “Complications,” later).

General Technique for Lung Biopsy Review of patient history and diagnostic scans are performed, and informed consent is obtained as noted in the earlier general section. Review of patient history must include smoking, chronic lung disease, and oxygen therapy because it has been clearly shown that these are associated with increased pneumothorax rate. Any history of recent infection may suggest the possibility that the mass could be a focal inflammation. It is important to know whether the patient has been on intermittent oxygen, which gives an indication of how difficult it may be to resuscitate the patient if a pneumothorax or bleeding occurs. Medication history is important to screen patients with coagulopathy-causing drugs. Patients frequently may self-medicate with aspirin or other analgesics, which might affect blood clotting. If such medication has been taken, it is best to defer the procedure because the lung has a greater propensity to bleed than other organs. Review of prior scans helps preplan the procedure relative to possible approaches and patient positions; it saves time to pre-position the patient if a position other than supine is to be used. Any enhanced CT scan may show vascularity and other anatomic features. PET scan can show the metabolically active areas or necrotic areas for target planning (see Fig. 67-59). Laboratory values for blood count and coagulation factors should be suitable.

Target and Trajectory Planning.  A diagnostic series of scans is taken in anticipation of the procedure. From the current scans, the mass referred for biopsy is reevaluated. If the lesion is stable or larger, the procedure should proceed. If the lesion is significantly smaller, one should consider deferring the procedure until a later date after follow-up scans are performed. One selects the tentative target for biopsy and the appropriate pathway for the trajectory back to the skin surface. If the mass is large, one selects a portion of the mass that is not necrotic and a portion that is definitely solid if there is associated infiltrate or collapse. When there

FIG 67-58  CT scan shows the approach of a mass in the right apex. The needle is positioned to traverse the “fibrous band” (arrow) while traversing to the pleura. This approach minimizes the possibility of a pneumothorax.

A

B FIG 67-59  A, PET scan shows a large mass laterally in the left lung. Note the central area of low uptake indicating necrosis. B, CT scan shows biopsy of the mass with the patient in the decubitus position, partially immobilizing the mass. Because of central necrosis on PET, the biopsy was taken peripherally in the mass.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2433

B

A

C

D

FIG 67-60  A, CT scan of a patient with a history of breast cancer shows a linear density in the right middle lobe anteriorly, which was a new finding suggestive of recurrence. B, PET scan shows increased uptake in the mass, suggestive of tumor. SUV was higher than the mediastinum. C, Repeat PET scan after an oral dose of OTC pseudoephedrine shows absence of uptake. (Oral phenylephrine 10 mg may also be used.) Our theory is that the inflammatory vessels constricted and changed the uptake of FDG. D, Repeat CT scan on same day as the second PET scan shows that the lesion looks unchanged, confirming that the changes in the PET were not due to resolution of the lesion.

are questions about viability of the mass, correlation with a PET scan may be helpful. The trajectory pathway to be used (which influences the entrance point selection) should avoid bullae or blebs, bronchi, and vessels. Penetration of bullae or blebs may cause pneumothorax. Puncture of the bronchi causes coughing, which can cause pneumothorax or possible air embolism to the vessels. If possible, one should choose the shortest pathway to the lesion, realizing that the distance through aerated lung is directly related to the pneumothorax rate (see “Lung and Mediastinum,” later).152,253 If there is an infiltrate or pneumonia, one should plan to traverse those areas if possible because this lessens the chance of pneumothorax. If there is a fibrous attachment or pleural thickening nearby, the trajectory should traverse them; it is likely that the fibrotic tissue anchors the visceral pleura to the parietal pleura, lessening the chance of a pneumothorax.

Bolus dynamic contrast scan over the chosen target site should be performed for numerous reasons. First, if the patient has any history of bruit on auscultation of the chest or if the scan appears to show a prominent vessel adjacent to or entering the mass, a bolus scan probably should be performed (Figs. 67-61 and 67-62). The other circumstance would be in instances when there is no clear history of smoking or other risk factors for a tumor. Most importantly, bolus scanning is absolutely indicated if any consideration of a cutting needle is being made. The short time required for such a study is a small price to pay for avoiding a life-threatening catastrophe. Patient position may be adjusted depending on the target site and the specific anatomy of the trajectory. In most instances, it is simplest to have the patient lie in the prone or supine position, but other positions also might be considered (Fig. 67-63; see Fig. 67-35). In some cases, changing the patient position may “clear” the pathway for the

2434

PART III  Image-Guided Procedures biopsy. Specifically, the position of the scapula and the relationship of the ribs to the mass may change with an oblique position, converting an unlikely approach to a suitable one (see Fig. 67-63). If the lesion appears mobile, which is common close to the diaphragm, the patient can be positioned with the ipsilateral side down, which may lessen motion of the diaphragm (see Fig. 67-35). If the lesion is small and moves with respiratory motion, it may be wise to put the patient in an ipsilateral decubitus position and insert the needle through a direct anterior or posterior approach. This position minimizes the motion of the ipsilateral diaphragm because of the weight of the abdominal organs against the diaphragm. This restriction of motion minimizes movement of the lesion and simplifies the performance of the procedure. Kinoshita and colleagues263 and Rozenblit and associates427 reported that if a biopsy can be done with the ipsilateral side dependent, the pneumothorax rate is reduced.

R

P

FIG 67-61  Because a cutting needle biopsy was being considered, this bolus injection was performed. Note the remarkable vascularity in the mass, which later proved to be an angiosarcoma. If a biopsy had been performed, it surely would have been catastrophic.

Procedure Evaluation and Planning.  During the procedure, vital signs, neurologic status, and pulse oximetry should be monitored.

12

13

0

0

A

B FIG 67-62  A, This small lesion was being considered for a cutting needle technique because of the close proximity to the pleura. B, Bolus scan shows enhancement of an AVM.

A

B

FIG 67-63  A, A patient with a small posterior lateral mass is positioned so that the rib and scapula block a potential pathway to the lesion (arrow). B, When the patient was placed in the left posterior oblique position, the anatomy shifted so that biopsy could be performed (arrow). If a more anterior or posterior approach were taken, a pneumothorax would have been more likely because of the length of lung traversed.

CHAPTER 67  Image-Guided Aspirations and Biopsies Analgesics and sedatives can be administered as needed (see earlier under “Patient Cooperation and Sedation”). Instrument choice and techniques.  Our needles of choice for lung and mediastinal biopsies, considering the diagnostic outcome and complication rates as noted in those sections, are (1) the 20-gauge FNA needle (spinal needle with 30-degree bevel is equivalent) for aspiration biopsy, (2) 20-gauge cutting needle for suitable parenchymal or pleural masses, and (3) 18-gauge or larger cutting needles for larger pleuralbased masses. A single-needle method or a coaxial method can be used for a needle aspiration of the lung. With the single needle, the needle is started into the superficial tissues and a scan is obtained to verify its angulation. When the correct angle is determined, the needle is inserted to the correct depth. The coaxial method should be used for deep or small lesions. The guidance cannula is inserted into the chest wall and directed immediately adjacent to the lesion. The needle is inserted through the cannula into the mass. When a mass is accessible and in the appropriate location, a cutting needle may be used judiciously. The most critical factor with any cutting needle procedure is that a bolus injection be performed before the procedure. Although infrequent, vascular lesions, such as AVMs or vascular sarcomas, may occur, and catastrophe can be avoided by aborting the procedure (see Figs. 67-61 and 67-62). If a lesion within the parenchyma is accessible and of sufficient size, a 20-gauge cutting needle may be considered because the risk is no different than an aspiration needle (see Fig. 67-35). The size and location of the mass must be adequate to permit complete placement of the cutting area within the lesion (Fig. 67-64). In such cases, the intervening lung has no additional trauma than an aspiration needle because there is no caliber difference or exposure of the traumatic tip to the lung. One must be prepared to compensate for deflection and bending of the needle when the stylet is inserted. Benign lesions tend to be very hard, and the surrounding lung does not provide much resistance. Such lesions may bend the stylet or deflect it to the side (Fig. 67-65). Choice and palpation of entrance site.  When the tentative trajectory and entrance site are chosen, a marker is placed on the skin and a repeat scan is performed to see the marker relative to the pathway. The marker may be a purchased device or a hypodermic needle in the package. After one has selected a tentative entrance point and before the needle insertion, one should inspect the chest carefully at the entrance site for several problems. First, if there are any obvious superficial veins noted, the site should be moved accordingly. Second, one should

2435

palpate deeply for the intercostal space between the margins of the ribs. The interspace between the ribs is greater the more anterior the location and less posteriorly (see Fig. 67-53). The ribs become smaller as they approach the apex, making posterior approaches the most difficult in that area because of the narrow spaces and angle. Palpation of the ribs is necessary because the location of the intercostal space as seen on the CT scan is unreliable. The tapered margin of the rib and the oblique orientation of the rib as it traverses the imaging section produce considerable partial volume effect and discrepancy regarding the actual location of the intercostal space on the scan. The problem of identifying the location of the ribs is especially relevant to the anterior chest wall where the ribs are mostly cartilaginous and fuse together in an unpredictable pattern. Although one can use a needle or a guidance cannula to “drill” through a cartilaginous rib or sternum, we do not recommend this with CT guidance because if one drills through the cartilage, the pathway of the needle is fixed, preventing corrective angulation. Also, the fixed needle cannot move with respiration, making a small tear more likely if the patient moves the lung while breathing. In the lung, as in other areas, the needle should be permitted to “swing” freely with respiration. Any time the needle is adjusted, the patient should be asked to suspend respiration temporarily during the manipulation of the needle or removal of the stylet to prevent tearing of the parenchyma or a pneumothorax. Another structure that should be carefully noted is the scapula. When approaching a lesion from the posterior chest wall, at any location, one should carefully palpate for the scapula. Although on first consideration one might think that the scapula would not typically interfere with a procedure, the tip of the scapula is tapered in two directions such that the location of the tip of the scapula is “averaged out,” so that what is seen on the scan does not correspond to the palpable tip on the patient. In some cases, the needle is deflected without the operator’s being aware of the problem. Also, in some cases, the patient moves the scapula slightly into the path of the needle and deflects it sufficiently to preclude a successful needle placement. Site preparation.  Careful antiseptic preparation of the skin is important. Proper administration of lidocaine to produce a “painless” procedure, as noted earlier, is important. When the area is anesthetized, it is important to infiltrate anesthetic through the subcutaneous tissues and down to and including the pleura. Adequate anesthesia of the pleura makes the procedure virtually pain free and permits insertion of the needle without the patient’s moving because of pain. So important is this step of anesthetizing the pleura that as we perform the actual insertion of the needle, we stop the needle tip at the pleura to inject a small amount of anesthetic at the pleura and to check the final angle of needle insertion before entering the lung (see Fig. 67-35). If a

FIG 67-64  Biopsy of pleural-based masses can be easily and safely accomplished with cutting needles if the gap of the needle is confined to the mass (arrows). Note the thin fat line of retracted pleura adjacent to the mass, indicating adherence.

FIG 67-65  Scan of a cutting needle biopsy of a granuloma. Note the bend (arrow) in the needle produced by the hard character of the mass.

2436

PART III  Image-Guided Procedures

rib is inadvertently hit during the initial anesthetic administration, we give a small bolus of lidocaine on the rib; this prevents any recurrence of pain if that area is struck again during the procedure. Needle selection and insertion.  The appropriate needle, either FNA or cutting needle, is selected using the factors discussed previously. The needle is inserted with incremental steps and adjusted until it is in place in the target site. The patient is asked to hold the breath during each movement of the needle. With the aspiration needle, localization of the tip precisely in the chosen site is important if adequate cellular material is to be obtained (see Fig. 67-35). If a cutting needle is being used, one must be confident that the cutting gap of the needle would fit entirely within the mass and adjacent normal lung would not be traumatized (see Fig. 67-63). If this is not the case, a severe hemorrhage or pneumothorax or both may occur (see Fig. 67-39). During needle insertion and sample acquisition, whenever the needle is moved the patient should be asked to suspend respiration. Not doing so may cause shearing of the pleura and cause a complication. Sample acquisition.  The sample should be acquired as noted previously in the general section on technique. When the stylet of the needle is being removed to permit attachment of the syringe, patients should be asked to hold their breath so that they do not inhale and pull atmospheric air into the chest. We routinely take only a single needle pass unless there has been an error in technique or a poor sample obtained on visual inspection. We prefer to defer a second needle pass until after the results of the first pass are known unless there has been a break in the technique. If a technical error has been made or the final position of the needle placement is unsatisfactory, a second needle pass can be made at once. If the cytologist is available for a rapid stain interpretation, the patient remains in the department until the results are known. If this service is unavailable, I still prefer to wait for the results instead of taking a second pass, because the 24-hour delay is a small tradeoff considering the high success rate.

Postprocedure Care.  Postprocedure care is simple and consists of observing patients for severe hemoptysis and pneumothorax. Immediately after completion of the procedure, I prefer to have the patient lie still for 5 to 10 minutes to let the puncture site seal. Moore and coworkers have shown that the incidence of pneumothorax is reduced if the patient is positioned with the puncture site down after the procedure.358 The pneumothorax rate for patients placed in this position was 17.9%, whereas the rate was 33.6% for patients not so positioned. Chest tube placement rate was 9.8% for patients who were positioned with the procedure side down compared with 0.4% for patients who were not. Before removing the patient from the table, we take a repeat scan to evaluate for postprocedure problems. Although Miller and associates reported that CT and plain radiography were equivalent for pneumothorax detection,352 Bungay and colleagues reported that CT is more sensitive for detection of pneumothorax (radiography detected only 22 of the 35 pneumothoraces detected by CT).48 A small pneumothorax is of little significance, as noted later. Some bleeding in the parenchyma in the area of the needle track occurs frequently, but it is of no significance. If there is any question of the patient’s stability, oxygen should be administered, and appropriate treatment should be given. Postprocedure chest films should be taken 1 hour, 4 hours, and 8 hours after the procedure. It makes no sense to use valuable CT time for follow-up of pneumothorax; chest tube insertion is required only with larger pneumothoraces when patients are symptomatic. Perlmutt and coworkers showed that pneumothoraces could be detected in the

following percentages at the noted timed delays: 89% immediately after procedure, 9% after 1 hour, and only 2% after 4 hours.402

Accuracy of Diagnostic Tissue Recovery.  The success rate of lung biopsies has varied widely among the many reports of fluoroscopic and CT-guided procedures and with the use of aspiration and cutting needles. Before cutting needle usage, favorable results were reported for aspiration cytologic results. With increased use of cutting needles, the literature shows improved recovery of diagnostic tissues for all malignant and benign lesions.

Aspiration Biopsy Success rates of aspiration biopsy have ranged from 60% to 95%. Several reports by Sinner,455,456 Thornbury and colleagues,487 and Greene and colleagues163 have documented that the results of aspiration biopsies correlate well with histologic results. Greene and colleagues found the correlation to be 92.3% for adenocarcinomas, 100% for squamous cell carcinomas, 88.9% for large cell undifferentiated carcinomas, and 100% for small cell undifferentiated carcinomas.163 Taft and coworkers found an interobserver error to be only 6%, with 94% overall consistency.483 These results are exclusive of lymphoma and mesothelioma cases, which are difficult to diagnose with aspiration biopsies. Weisbrod and associates noted that in about 67% of lung cases the diagnosis of lymphoma can be suggested, but not the definitive subtype.508

Cutting Needle Biopsy Goralnik and coworkers, who reported the first cutting needle biopsy in the thorax, found that the specific subtype could be diagnosed using a large cutting needle for suspected lymphoma lesions.160 Initial reports by numerous authors, including Kushihashi and colleagues,275 Laurent and colleagues,289 Rotte,426 Greif and colleagues,164-166 and Klein and Stewart,265 have verified the usefulness of the cutting needles but confirm the role of aspiration needles for diagnosis of malignant tissues. Klein and Stewart found little difference in the diagnosis of malignant lesions but great superiority for benign disorders.265 The correct diagnosis of benign lesions was 44% with aspiration needles and 100% with cutting needles. With peripheral lung nodules, Greif and colleagues reported an accuracy of 81% for benign lesions compared with 16% for aspiration needles.165 More recent articles have confirmed the superior results obtained with cutting needles for unusual tumors and benign abnormalities. Bungay and associates reported 86% success rate for diagnosis of lymphoma and noncarcinomatous tumors.48 Lopez and coworkers reported 93% diagnosis in benign lesions.316 Carruthers and associates reported the usefulness of cutting needles for diagnosing Wegener’s granulomatosis.57 Greif and colleagues reported that in the mediastinum, cutting needles could diagnose lymphoma, thymoma, and bronchogenic carcinoma accurately.166 Zinzani and colleagues reported on a series of 83 patients, in which the accurate diagnosis of lymphoma was achieved in 81%.532 Savage and coworkers reported on a total series of 836 cases; of the 95 benign diagnoses by needle biopsy, a “specific benign” diagnosis was a true negative for malignancy.436 Savage and coworkers reported on a series of cutting needle biopsy of benign masses; they noted that when a specific diagnosis is made, the diagnosis can be considered definitive.436 Even the problematic diagnosis of mesothelioma can be made. Scott and colleagues reported diagnosing 13 of 14 cases of mesothelioma.441 The tumor is problematic because even with a large tissue sample, it is not always possible to distinguish a benign tumor from a malignant one. Based on our experience with about eight such cases, the diagnosis of mesothelioma can be made, but the differentiation of

CHAPTER 67  Image-Guided Aspirations and Biopsies benign versus malignant can be determined only from its biological activity.

Improving Diagnostic Yield Precise localization of the needle into the viable portion of the tumor is important. One should avoid areas of necrosis or infiltrate. If the nodule being sampled contains a small amount of calcification, the calcified area should be avoided and the adjacent soft tissue density should be targeted. If the nodule is cavitated, appears necrotic, or is large, the central portion should be avoided in preference to the peripheral area (Fig. 67-66; see Fig. 67-59). The importance of accurate localization is emphasized by the numerous reports that note that the diagnostic yield is lower and the complication rate higher with smaller lesions. The size and location of a mass affect the probability of a positive diagnostic aspiration. Bergquist and coworkers noted that the diagnostic accuracy was higher for larger lesions and for more peripheral lesions.34 A 1-cm nodule was successfully biopsied in 58% of cases when it was central; it was successfully biopsied in 85% of cases when it was peripheral. More recently, Li and colleagues299,300 and Miller and associates352 have confirmed the improved diagnostic recovery from large lesions compared with smaller ones. Localization in such lesions is more difficult, usually requiring several repositioning attempts in the parenchyma. If using an aspiration needle, multiple needle passes increase diagnostic yield but at a cost of increasing pneumothorax rate. Early reports by Sargent and colleagues433 and Stevens and Jackman477 described the effects of increasing the number of needle passes. Stevens and Jackman477 reported that a second needle pass increased diagnostic yield by 10%, and a third needle pass increased yield by 1.5%. Sargent and colleagues433 reported that a second needle pass increased the diagnostic yield by an additional 5%, and that the third needle pass

FIG 67-66  CT scan shows a necrotic mass in the anterior mediastinum. A biopsy is taken of the wall and may include some of the necrotic material.

2437

increased the diagnostic yield by another 3%. Multiple passes increase the pneumothorax rate, however. Appropriate use of the cutting needle can improve diagnostic yield in specific cases. As noted in the discussion of the cutting needle technique, the needle caliber should be 20 gauge if the mass is not pleural based, and the cutting action of the needle should be confined to the mass, without extension into the adjacent lung parenchyma. Use of the cutting needle is justified if there has been a failed aspiration procedure or suspicion of lymphoma, mesothelioma, or thymoma.

Complications The main complications reported for lung biopsies are pneumothorax, hemoptysis, air embolism, and intrabronchial bleeding. Pneumothorax and hemoptysis are the most common complications. The incidence of pneumothorax varies widely among the numerous clinical reports. The asymptomatic rate ranges from 17.5% to 72%. The symptomatic rate requiring chest tube is 6% to 18%. The hemoptysis rate varies for different needle sizes from 5% to 21%. The treatments for these complications are discussed at the end of this section.

Pneumothoraces—Factors Affecting Incidence.  Effective administration of local anesthesia through the trajectory site to the pleura helps reduce pneumothoraces. If the patient does not have the pleura adequately anesthetized, even the most cooperative patient may jerk at the instant a needle traverses the pleura, potentially tearing a small hole and producing an air leak. Theoretically, placing a postbiopsy patient with the ipsilateral side dependent minimizes the incidence of a pneumothorax. This theory is controversial and not yet confirmed because there are conflicting reports on the topic. In a series of 262 patients, Moore and colleagues reported that the pneumothorax rate was reduced in a statistically significant manner.358 Collings and associates reported in a series of 423 patients that no benefit occurred.76 Although early authors attributed the high pneumothorax rate to the prolonged length of the procedure, a more recent report by Ko and coworkers refutes this claim.267 These authors found no difference between patients with procedures with different times. Two major determinants of pneumothorax rate are the status of the lung and the number of needle passes. If a patient’s lung has severe chronic obstructive pulmonary disease (COPD), more blebs are present. Decreased lung compliance in patients with interstitial disease favors collapse of the lung when a small air leak occurs. In prior editions, we noted from our experience numerous parameters important for pathway selection, which have since been validated by reports in the literature. One can minimize the possibility of a complication by looking carefully for a safe pathway through the lung to the nodule being biopsied. In such cases, one might locate the attachment to the pleura of an area of fibrosis or thickened pleura that can be traversed by the needle. By planning the trajectory through such an area, only a minimal amount of parenchyma is traversed or the fibrosis can serve to keep the layers of the pleura attached and avoid a pneumothorax (Fig. 67-67). Ko and coworkers reported that pleural thickening associated with prior surgery decreased the pneumothorax rate (P < .05).267 By looking for a small amount of “fat” that may be tented at the pleura, one usually can distinguish between a mass that is simply adjacent to the pleura and one that has a fibrous extension. This type of approach is especially important in high-risk patients with pulmonary hypertension, severe restrictive disease, or other serious underlying pulmonary problems. Minimizing the pathway of lung crossed (Fig. 67-68) and using a pathway through nonaerated lung can minimize the incidence of pneumothorax. The significance of minimizing the needle pathway

2438

PART III  Image-Guided Procedures

A

B

FIG 67-67  A, Small lateral mass close to the pleura with a linear band connecting to the pleura. The patient had platelet inhibitor therapy, which could not be stopped, so a low infusion of platelets was performed. We use this method if a mass is adjacent to the pleura but are reluctant to do so if the mass is deeper. B, Biopsy was performed laterally through the fibrous band.

B

A

FIG 67-68  Difference in approach for two lesions that are almost identical. A, This CT scan shows a direct lateral approach, with the distance of lung being crossed almost half of that in B. B, This biopsy approach traverses almost twice the distance of lung and increases the potential risk by a factor of almost 3 according to new data (see text).

length was first evaluated by one of our colleagues, Locke (Locke J, Haaga JR: Personal communication, unpublished data), who studied a series of 235 cases from our center and looked at the incidence of pneumothorax compared with the distance of the lung traversed. The distances crossed and the corresponding incidence of symptomatic pneumothorax were 1 cm, 3.7%; 1 to 2 cm, 10.7%; 3 to 4 cm, 19%; and 5 to 6 cm, 44.4%. Because these measurements were taken from

CT scans, they were accurate without errors in magnification; other authors who have not supported this relationship made their measurements from radiographs, which are more subject to error. In recent years, other authors have confirmed that lesion depth was the prime determinant of pneumothorax for CT procedures.134,253,289 That traversing nonaerated lung, as we initially published in prior editions, does diminish pneumothorax has been documented.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2439

B FIG 67-69  A, A large mass in the anterior right middle lobe. Note the collapsed lung peripherally and the bulging mass centrally, similar to the “golden S sign” described for plain radiographs. B, Biopsy through the collapsed lung was performed. Because there is no aerated lung, the change of pneumothorax is minimal.

Haramati and Austin showed convincingly that if aerated lung was avoided, the pneumothorax rate decreased significantly.204 These investigators found that the pneumothorax incidence through aerated lung was 46% compared with 0% through nonaerated lung. Cox and associates reported 15% pneumothorax for nonaerated lung and 50% for aerated lung.80 Stated differently, if there is an area of infiltrate adjacent to a nodule, one should intentionally traverse the consolidated area (Fig. 67-69). Status of the lung was evaluated by Fish and colleagues127 and Miller and associates.352 Topal and Ediz491 and Cox and colleagues79 showed a definite relationship of pneumothorax to lung disease. Fish and colleagues127 evaluated chest films and spirometry to determine the presence of COPD, and they found that patients without COPD had a pneumothorax rate lower than patients with COPD. With spirometry measurements, the ratio for pneumothorax was 45% for COPD patients compared with 25% for patients without COPD. Using the chest x-ray as the standard, the rate was 42% for COPD patients compared with 25% for patients without COPD (see earlier under “Pneumothoraces—Factors Affecting Incidence”). Cox and colleagues reported that patients with emphysema were three times more likely to have a pneumothorax.79 Avoiding cystic abnormalities reduces the chance of a pneumothorax (see Fig. 67-55B). Several authors have reported a blood patch technique to prevent pneumothorax.42,217,338 With this method, a coaxial needle method is used, and a blood clot is injected through the long cannula during its removal from the lung. The results of this method vary; Bourgouin and associates reported no difference in outcome when it was evaluated in a randomized prospective fashion.42 Lang and coworkers studied a group using a blood patch method; in patients with emphysema, pneumothorax occurred in 3 of the 20 patients (15%) who received autologous blood clot and 10 of the 14 (71%) patients who did not (P < .001).285 In our opinion, the difference in results is probably based on the exact method of instilling the blood clot. If clot is simply deposited at the pleural margin, it is unlikely to seal the leak of air. If blood is injected throughout the pathway of the needle within the lung and the pleural margin, a seal is more likely to succeed.

Hemoptysis—Factors Affecting Incidence.  Hemoptysis is expectoration of blood after biopsy; it does not refer to the imaging appearance of a small amount of bleeding, Commonly, minimal localized hemorrhage in the parenchyma may occur after a biopsy, but it is of no significance (Fig. 67-70). The factors affecting hemoptysis are the needle caliber, depth of lesions, coagulation status, presence of congestive failure, and an improper trajectory that passes through a moderatesized pulmonary vessel. Excessive damage to the lung parenchyma with a cutting needle also can occur if the needle is not positioned properly (see Fig. 67-39). These circumstances should be avoided if possible. Although this is an unsettling event for the patient and the physician, hemoptysis was common in the early days of fluoroscopic biopsy. Needle caliber directly affects hemoptysis; this was confirmed by Khouri and coworkers, who noted in their series that the hemoptysis rate was 21% with an 18-gauge needle but only 5% with a 20-gauge needle.257 Deep lesions and hilar lesions are at greatest risk because of the long pathway and proximity of blood vessels. Yeow and colleagues reported depth and size of lesion as the most important factors for hemoptysis.527 The cases reported in the literature were associated with illdefined masses adjacent to the hila. Also, one must remember the importance of bolus studies in selected patients (see Figs. 67-61 and 67-62). Vascular lesions, aneurysms, and AVMs are very vascular  and may bleed. Bolus dynamic scan is crucial to avoid biopsy of such lesions. Patients with coagulopathies, pulmonary hypertension, or congestive heart failure are at greatest risk for hemoptysis and intrabronchial bleeding (as noted later under “Intrabronchial Bleeding”) and should not undergo biopsy without correction of the basic problem. Biopsy in a patient with congestive heart failure should be avoided because the pulmonary vessels have elevated pressure, are edematous, and are more fragile, being predisposed to tearing. Anticoagulated patients should never undergo biopsy, even if the anticoagulation consists only of aspirin. The lung parenchyma has a very low tissue pressure because of the alveoli, and there is no compressive effect to help retard bleeding.

2440

PART III  Image-Guided Procedures

0

0

A

B

0

C FIG 67-70  These three scans show the optimal approach for performing a biopsy. A, CT scan shows the needle positioned adjacent to the pleura, not the bulging of the pleura produced by the injection of lidocaine. B, The needle was advanced into the lesion for the biopsy. C, CT scan after removal of the needle shows a small infiltrate in the lung, representing hemorrhage. Small amounts of bleeding such as this are common and are of no significance.

Fatal Events.  We have located nine reported deaths in the literature from percutaneous lung biopsy, five from endobronchial bleeding and four from air embolism.64,268 Having read those reports, some valid observations can be made. Air embolism.  Air embolism is caused by introduction of air into the arterial system that results in occlusion of the blood supply to organs. Although most of the cases occurred with percutaneous biopsy, a case of fatal air embolism has been reported with bronchoscopic biopsy.444 In the early reported cases of air embolism to the head the patients died, but there have been three reported cases of patients who have survived, two with hyperbaric oxygen therapy and one spontaneously246,287,385; very few centers have hyperbaric facilities. After reading the original reports on these cases and other research papers on the pressures in the tracheobronchial system, one can hypothesize the cause of air embolism and steps to use to avoid such catastrophes. In all of the cases, the clinical discussion is very similar. The authors related that unexpectedly the involved patients began an uncontrollable coughing episode, which could not be ameliorated by consoling the patient and supportive action. In each case, the needle was left in place during the episode. Each case notes that the patient stopped coughing suddenly and became unresponsive and died immediately. One more recent case was associated with positive-pressure ventilation. Several authors hypothesized that air embolism is a result of the elevated intrabronchial pressure that occurs with coughing; the highpressure air follows along the needle track into the blood vessels.489,510 The literature notes that during a cough when the epiglottis closes, the pressure can reach more than 100 pounds of pressure. We theorize that with the needle in place, the high-pressure air passes along the shaft of the needle into the pulmonary veins and subsequently into the arterial system.

We believe the way to prevent any such event is to withdraw the needle immediately whenever a patient starts coughing during a needle biopsy. If one is careful to avoid visible bronchi during the trajectory planning, coughing is a rare event, and this policy does not defer many patients. Patients with a propensity to cough should be treated with codeine before attempting a biopsy, but if coughing ensues the procedure should be canceled. Intrabronchial bleeding.  Intrabronchial bleeding is a very rare complication, reported only several times in the literature with percutaneous and transbronchial biopsy, with numerous deaths.334,381 Most instances of intrabronchial bleeding have occurred with cutting needles before guided methods, but anecdotal cases with guided aspiration and cutting biopsies have occurred. If one reads the original reports, the patients have the same general story. The patient experiences hemoptysis, and this continues until large amounts of blood are coughed up and the patient quickly dies. We have not seen such a case, but from the reports and images provided, it seems likely that the patients had ongoing congestive heart failure. We hypothesize that because the pulmonary vessels were edematous and friable, a tear occurred in a pulmonary vessel and the patients bled into the bronchial tree.

Specific Sites in the Lung

Hilar Lesions.  Hilar lesions are best sampled by bronchoscopy, and percutaneous biopsy should be employed only as a second choice. With percutaneous procedures, the length of the lung to be crossed and the proximity of vessels make the incidence of pneumothorax and hemoptysis more likely. The success rate for bronchoscopy is better and the complication rate lower than for percutaneous procedures. If a bronchoscopy has been attempted unsuccessfully, a percutaneous procedure is performed with the following guidelines.

CHAPTER 67  Image-Guided Aspirations and Biopsies Before the procedure, intense contrast enhancement of the vasculature should be performed to highlight the location of the anterior, lateral, or posterior approach to be taken. With such lesions, a coaxial short cannula method should be used. This method is especially helpful because, as noted earlier, before the biopsy needle is inserted, the guidance cannula predetermines the angle. With a long distance from the pleura to the hilum, a few degrees in error would result in a large localization error centrally.

Apical Lesions.  Careful planning for apical lesions is important because of the unique anatomy of the upper chest. Anteriorly the brachiocephalic vessels, mediastinal vessels, clavicle, and axillary vessels make the approach more difficult. Posteriorly the scapula covers a substantial portion of the posterior chest. Avoidance is best accomplished by having the patient fully abduct the shoulder so the scapula moves laterally. Careful palpation usually reveals the margin of the scapula. Another anatomic factor making the approach more difficult is that the ribs become smaller as one progresses from the midchest cephalad to the apex. It is usually easier to plan the entrance site so that the angle also is slightly caudal, which makes penetration of the intercostal space easier (see Figs. 67-35, 67-53, and 67-54). Several final points should be made about unusual masses in the apex. Neurogenic tumors and lateral meningocele may occur in the posterior mediastinum adjacent to the neural foramina (Fig. 67-71). Neurogenic tumors are hard and painful in biopsy77; a diagnostic sample seldom can be obtained. Puncture of a lateral meningocele yields clear fluid consistent with spinal fluid.

Pleural Lesions.  Pleural lesion biopsies are difficult to perform fluoroscopically because as the lesion is turned perpendicularly to the fluoroscope, it becomes en face and more difficult to see. With CT, sampling of such lesions is easy and provides considerable flexibility for needle selection. An aspiration needle can be used at any time; in cases in which it is indicated, a cutting needle can be used for a more adequate

2441

tissue sample. Goralnik and colleagues reported that these cutting needles could be used safely for pleural lesions.160 These investigators recommended their use when either lymphomas or unusual tumors such as mesotheliomas were suspected. An automated biopsy device also can be used if the lesion is large enough to accommodate the length of the needle action. Scott and coworkers noted that “by combining findings at biopsy and at CT (presence or absence of pleural thickness >1 cm, mediastinal-circumferential involvement, irregular contour), sensitivity and negative predictive values reached 100% and 1.0.”441 Maskell and associates reported that guided biopsy of the pleura is superior to unguided pleural biopsy.330 Adams and Gleeson reported on pleural biopsy of patients with suspected malignant effusions.3 In their series, accuracy was 91% overall and 93% for mesothelioma. Several authors have reported on the use of long versus short pathways for biopsy of small subpleural nodules. For small subpleural nodules about 1 cm, Gupta and colleagues compared the usefulness of using a direct short pathway with a long pathway (entrance site and trajectory slightly away from the nodule).174 They reported better accuracy for a diagnostic result but a higher incidence of pneumothorax for the long pathway (Fig. 67-72). The diagnostic yield in the short pathway group was 71% compared with 91% for the long pathway. Although the frequency of all pneumothoraces for both groups was the same, the chest tube rate was higher for the long pathway (38% vs. 17%). Ko and coworkers reported that a shallow angle of insertion was more likely to produce a pneumothorax.267

Cavitated Nodules.  Cavitated nodules are slightly problematic because two major issues arise. To obtain viable diagnostic tissue, one must sample the wall carefully (see Fig. 67-72). If the needle enters the cavity, a pneumothorax is likely and there may be spillage of material into the pleural space; spillage of neoplastic or infectious material into the space should be avoided.

Mediastinal Lesions.  Before the use of CT, only a few authors ventured to perform biopsy of the mediastinum using fluoroscopic guidance. With the refinement of CT techniques, however, CT is now the modality of choice and is commonly used for this area.4,273 Several limited studies by Lauby and colleagues,288 Jereb and Us-Krasovec,241 and Rosenberger and Adler421 showed the feasibility of

FIG 67-71  Rounded masses such as this by the neural foramina are

FIG 67-72  CT scan shows biopsy of a large cavitated nodule. The

likely to be of neurogenic origin. Lateral meningoceles yield cerebrospinal fluid on aspiration. Schwannomas such as this one yield diagnostic tissue when sampled by cutting needle, as noted here, but are painful during biopsy because they are attached to the nerve roots.

periphery of the mass is targeted for best results. Some authors have suggested that in such cases a higher incidence of pneumothorax and hemorrhage exists. A pathway through the thickened pleura was chosen to minimize pneumothorax.

2442

PART III  Image-Guided Procedures

mediastinal biopsies with fine needles guided by fluoroscopy. Weisbrod and colleagues508 and Westcott513 confirmed the benefits of fluoroscopic procedures but also noted some of the disadvantages and potential hazards. The most extensive series, consisting of 100 patients, was reported by Westcott.513 Although the results showed 96% accuracy, there were 10 patients with vascular aneurysms that created some difficulty. Five of the patients were excluded from a biopsy because of a difficulty in evaluating the mediastinum. In another five cases, inadvertent puncture of aortic aneurysms occurred; in one patient, a hemomediastinum developed. Weisbrod and colleagues also reported several important observations.508 In their series they noted that although aspiration needles were satisfactory for metastatic disease, the needles were suboptimal for lymphomas. They also reported one case of cardiac tamponade that was successfully treated. Mediastinal anatomy and approaches.  The mediastinum is an exciting area for the application of procedures. Before the advent of CT, when most biopsies were performed with fluoroscopy, this area was considered off limits because there was no reliable way to image the anatomy, either diagnostically or for intervention. With the evolution of CT imaging and its corresponding procedures, radiologists and clinicians have become confident of mediastinal anatomy and pathology. The reliability of anatomic display and refinement of the interventional techniques are such that procedures of the mediastinum are now commonplace. These procedures require considerable skill, however, and should be performed only by the most skillful radiologists who are well versed in anatomy and procedures. As one would expect, performing procedures in the mediastinum is contingent on comprehensive and complete knowledge of the mediastinal anatomy. One must be familiar with virtually all variations so that inadvertent puncture of a vascular structure does not occur. Aberrant subclavian vessels, left-sided vena cava, and prominent azygos veins are the most likely variations to occur. Appropriate use of bolus dynamic scanning is crucial to ensure that all vascular structures are well defined. Although it is seldom that tumors are vascular and represent a bleeding risk, it is common for vascular anomalies, aneurysms, or pseudoaneurysms to exist. Any rational interventionalist would use bolus dynamic scanning in the mediastinum to visualize normal anatomy and exclude a vascular anomaly (Fig. 67-73); a more complete discussion of variant anatomy is provided in Chapter 38. Procedures of the different portions of the mediastinum (superior, anterior, middle, or posterior) require uniquely different approaches. Successful and safe procedures in these areas require careful attention to the subtle techniques used to guide and manipulate instruments in these areas. Superior mediastinum.  The superior mediastinum consists of the space above the aortic arch and contains the various brachiocephalic vessels (venous and arterial). This is the most common area for congenital variations because arteries emanating from the arch and venous variations are common. This space requires an anterior insertion in virtually all cases and usually results in an approach through the suprasternal notch or the parasternal region (Fig. 67-74). In all cases a good vascular study with bolus injection is crucial to clearly define the carotids, jugulars, and small vessels. In most cases the lower portion of the thyroid gland and the trachea must be dealt with to approach such masses. It is always crucial to clearly identify the lobes of the thyroid because it is common for substernal or low-lying thyroid to be mistaken for a mass in the upper mediastinum. If the ipsilateral lobe of the gland is not identified, one should always consider the possibility that the observed mass might be the lobe of the thyroid.

A

D

B

A

C FIG 67-73  Common vascular variations are noted in these three images, which should not be mistaken for tumors; distinction is easy with contrast material. A, CT scan shows a very prominent azygos vein. B, CT scan at the level of the aortic arch shows a small aneurysm at the ductus region (D). C, CT scan of the upper mediastinum shows an enhancing mass behind the trachea, which represents an aberrant subclavian artery.

CHAPTER 67  Image-Guided Aspirations and Biopsies Methods for choosing the correct entrance site and pathway are essentially the same as in other areas. As the needle is inserted, the pathway and tip of the needle are monitored carefully with incremental adjustments. Use of the bevel to slide past the different structures is important to provide clearance around the vessels. As noted in the earlier technical section, the bevel is turned toward the structure to be avoided, and the point is rotated so that it is directed toward the intended target. Also as noted in earlier sections, successful maneuvering of the needle around a vessel or structure is best accomplished by pushing the instrument slowly, which facilitates the deflection of the needle. Anterior mediastinum.  Lesions in the anterior mediastinum are in the anterior space behind the sternum or anterior to the aortic arch. In most cases, one can access these lesions by using an insertion site just lateral to either side of the sternum. One must be cautious not to sample the internal mammary artery or vein. With these lesions, one can use either an aspiration or cutting needle as in other portions of the mediastinum. Careful localization of the needle tip or cutting needle gap is crucial in such cases (Figs. 67-75 and 67-76) because cutting of the major vessels would be catastrophic. As noted in the earlier section, placing a patient in a lateral or oblique position can shift the anterior junction line to make such procedures simpler. One can inject saline into the mediastinum to increase the potential pathway for needle insertion and to avoid traversing the lung (see Fig. 67-51). Gupta and colleagues have proposed using a transsternal approach with a coaxial system; these authors advocate using a coaxial cannula to penetrate the sternum and use a needle through the cannula.175 We have not found the need for this approach and have preferred to use the methods discussed here. Middle mediastinum.  The middle mediastinum is a large space below the arch and encompasses the azygos node (Fig. 67-77), peritracheal region (Fig. 67-78), aorticopulmonary window (Figs. 67-79 and

A

2443

67-80), subcarinal region (Fig. 67-81), and periesophageal region (Fig. 67-82). The peritracheal regions and the aorticopulmonary windows are typically best approached anteriorly through the anterior junction line, avoiding the lung parenchyma and other structures. The internal mammary vessels should be identified before any procedure so that the entrance and pathway can be planned appropriately to avoid these structures. The anterior approach may be performed if the patient has a wide anterior junction region, but in many cases it is helpful to place the patient in a lateral position, either on the right or on the left side (see Fig. 67-76). By doing this, the anatomy may shift so that a clear pathway through the anterior fat space may be possible, preventing any possible pneumothorax that could occur if the lung were traversed (see Fig. 67-79). In such cases, injection of saline or gas may be helpful to enlarge slightly the potential pathway (see Fig. 67-51). Widening of the pathway is typically small because these materials disperse along  the long axis of tissues and the cross-sectional areas, so the change in the anatomy is usually minimal. Maneuvering the instruments past corresponding vessels is a crucial part of the procedure. Paratracheal or azygos lymph nodes can be approached by traversing the anterior mediastinum and the space between the superior vena cava and the ascending aorta. The level of the entrance is below the origins of the brachiocephalic vessels and the subclavian veins. When using this approach, it is crucial to choose the correct entrance site and trajectory to ensure uneventful passage of the needle past the vessels. Traversing the space between the vena cava and aorta is simpler if the process has slightly widened the space. In cases in which the vessels have not been separated in this fashion, successful placement can be accomplished by using a coaxial system as follows. With the newly designed devices, blunt or rounded stylets are available so that the walls can be pushed slightly and penetration can be avoided. In other cases, one might choose to use a small-caliber bevel needle to slide between

B FIG 67-74  A, The most cephalic portion of the superior mediastinum includes the brachiocephalic vessels. This shows a mass (long arrow) adjacent to the subclavian artery. A needle (horizontal arrow) is inserted through the suprasternal notch and carefully advanced into the mass. In such cases it is best to keep the bevel directed toward the vessel so that it skims over the surface of the vessel. B, A different patient with small mass posteriorly, adjacent to the esophagus and trachea. A needle is noted passing by the trachea. The bevel of the needle was directed toward the trachea.

2444

PART III  Image-Guided Procedures

A

B FIG 67-75  A, CT scan shows an ill-defined mass in the anterior mediastinum, believed to be a subtle mass in the thymus. With the patient in the supine position, access to the mass cannot be obtained without crossing the lung. B, When the patient is positioned on the side, the mediastinal structures shift, permitting access for needle placement into the mediastinum.

A

B

C

FIG 67-76  A, An experienced interventionalist can use a pathway through the fat of the anterior junction space, from the right to the left side, if careful to avoid the internal mammary vessels. Note the needle in the proper angle of insertion, passing medial to the vessels. The correct needle angle must be attained before entering the space because there is no room for adjustment. B, The needle passes anterior to the vessels and past the aorta. The standard bevel needle is best; the flat side of the needle is rotated to be against the innominate vein so that the point does not penetrate it. C, Final placement of the needle in the mass permits successful biopsy.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

C

D FIG 67-77  A, A mass located in the azygos region shows tumor posterior to the cava and aorta. Initial positioning of needle required slight medial direction. B, A Chiba needle within the coaxial cannula is pushed past the vessels, with the bevel directed toward the vena cava. C, Through the cannula, Chiba needle aspiration was performed. D, After removal of Chiba needle, a cutting needle (20-gauge) procedure was performed. We have called this the coaxial slide method.

M

A

B

C

FIG 67-78  A, Mass (M) is located lateral to the trachea. Note the marker on the skin surface (arrow) used to choose the entrance site. B, The pathway for the needle (arrow) is most appropriate when it traverses the juncture point between the transverse process and the rib. C, By using ample local anesthetic along the ribs and paraspinal path, the needle (arrow) can slide along the vertebral body and enter the mass.

2445

2446

PART III  Image-Guided Procedures

A

M

A

B

C FIG 67-79  A, CT scan during bolus injection shows enhancement of the aortic arch (A). The mass (M) in the aorticopulmonary window of the middle mediastinum is clearly seen. B, Because the anterior juncture is wide, insertion of a Chiba needle is simple. C, With precise visualization of the margins of a cutting needle, accurate placement of the gap and sampling of the mass are easily accomplished. Because the action of the needle is confined to the gap, there is no risk of damage to the aorta.

the two structures, and the coaxial cannula can be advanced over the aspiration needle (see Fig. 67-77). The subcarinal area is best approached through a posterior paraspinal approach from either the right or left side, depending on the location of the mass. The left approach between the aorta and the vertebral body is usually the best approach because there is more space compared with the paraspinal region on the right side (see Figs. 67-81 and 67-82). As one reviews the figures, the approach seems almost

unbelievable, but the truth is that we have consistently performed this approach for 15 years with no complications. It would be gratifying to believe we were the only ones capable of doing this, but the facts are that the approach and methodology are valid, and most experienced and motivated interventional radiologists can learn to perform this procedure. Choice of the correct pathway and entrance site is crucial to the success of the procedure. The most suitable trajectory from which an entrance site must be chosen is a pathway that includes the margin of the transverse process/rib junction, the margin of the vertebral body, and the mass itself. The trajectory path should not include any vessel. A small amount of lung parenchyma in the path is acceptable but not preferable. Selection of any other pathway more medial than this would be unsuccessful because the contiguous overlap of the transverse process and the rib articulation provides an almost impenetrable “shield” that is painful and traps the needle so that no angulation or adjustment is possible. When this technique is executed, ample use of local anesthetic is important not only in the skin but also at the site of the transverse process and vertebral body. Unless adequately anesthetized, these two sites are painful because of the periosteum. The steps of the procedure are as follows. The site is prepared and the needle is advanced just to the margin of the transverse process. At this level, local anesthesia is infiltrated on the periosteum. The needle is rotated so that the bevel is turned against the body, and the needle is pushed slowly. The tip of the needle is located close to the margin of the vertebral body, and local anesthesia is injected. The needle is adjusted if needed so that the bevel of the needle is against the edge of the vertebral body and pushed slowly several millimeters. The bevel is rotated so that it is toward the aorta and pushed forward 4 to 5 mm to pass the aorta. If the hemiazygos vein or esophagus is close, the bevel also may be adjusted to pass by those structures. When positioned in the mass, the sample is acquired as appropriate with either the aspiration or the cutting needle. The sequence of events in this procedure typically proceeds predictably, but the procedure can be aborted any time at the judgment of the operator. As one begins to learn this technique, each succeeding attempt provides additional confidence that enhances eventual success and comfort with the procedure. One almost insurmountable difficulty is the presence of a large osteophyte that may occur at the edge of the vertebral body. When a large spur is present, one might not even choose to attempt the procedure. If the procedure is being performed and an obstruction occurs at the edge of the vertebral body, minor adjustments of the needle can be made to determine if passage is possible. Persisting beyond several attempts is seldom beneficial. Posterior mediastinum.  Sampling abnormalities in the posterior mediastinum is much easier than in either the anterior or middle mediastinum. The posterior mediastinum represents the space adjacent to the vertebral body and the pleural region. In most cases, there is clear access to the mass without crossing the lung parenchyma. In such circumstances, it is usually most expeditious to perform a quick sample with an aspiration needle and follow it with a cutting needle. Performing the aspiration first permits one to ascertain the effectiveness of the local anesthesia and the cooperation of the patient before using the cutting needle.

Pneumonias and Infiltrates Recovery of samples for culture and identification of infectious processes has become more popular in recent years. The results have been variable for routine pyogenic infections, so we have seldom been asked to sample such problems. The opportunistic infections in immunocompromised patients have become a more significant diagnostic

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

FIG 67-80  A, Access for this biopsy was not present when the patient was lying supine (see Fig. 67-10). The mediastinum shifted, opening up access through the anterior fat (bidirectional arrows). The path of the needle was not appropriate for sampling a small lymph node (arrow). B, The needle was redirected using the deflection method described and shown earlier. The node was sampled without removal of the needle and redirection.

M

A

B

C

D

E

F

FIG 67-81  A, Middle mediastinum shows a mass (M) clearly distinguished from the opacified vessels (arrow). The method shown here is effective for consistently sampling such masses. B, The needle is inserted at a point where the trajectory crosses the junction between the transverse process and the rib. Saline is injected into the paraspinal tissues, moving the pleura more laterally, facilitating easy passage of the needle. C, CT scan shows the next step in the procedure, which consists of positioning the needle against the anterolateral margin of the vertebral body (arrow). The bevel of the needle is directed toward the body; the needle is pushed forward close to the aorta. D, When the needle is against the aorta, the bevel is turned toward the aorta and pushed forward past the aorta (arrow). E, At this point the bevel of the needle is turned toward the azygos vein and pushed forward (arrow). F, Final positioning of the needle is within the mass (arrow).

2447

2448

PART III  Image-Guided Procedures

M

A

B

C

D FIG 67-82  A, A patient with a history of hepatoma shows a mass (M) in the middle mediastinum lateral to the esophagus. The lidocaine needle is used to inject saline in the paraspinal space, enlarging the pathway; compare with the opposite paraspinal space (arrow). B, Aspiration needle biopsy is performed. Performing aspiration needle biopsy before cutting needle biopsy permits assessment of pathway clearance and cooperation of the patient. C, The cutting needle has been inserted after aspiration biopsy, and the stylet has been inserted. The gap is medial, and it was considered too close to the esophagus (arrow). The stylet was retracted and the needle rotated 180 degrees. D, Using the deflection of the bevel of the stylet through the cannula (arrow), the gap is more laterally positioned so that the sample could be taken without a major readjustment of the needle. Note the position of the gap compared with that in C.

dilemma, and the occasion may arise in which sampling is more critical. In our experience, the diagnosis of fungal infection with cavitation or fungus ball by needle aspiration is reliable and easy to perform. In such cases, the procedure is performed in a standard fashion except that the sample is sent for appropriate stains and culture. When a more difficult pathogen such as Pneumocystis jiroveci is suspected, simple aspiration may not suffice. Although some authors have implied that recovery of diagnostic tissue is easy, we have found that it is not. In

many cases, one may require a small cutting sample that can provide more ample material for evaluation. When the infiltrate abuts the pleural membrane, we have employed a small cutting 20-gauge needle with considerable success and no morbidity. Our experience is limited, but we are optimistic for this technique.

Treatment of Complications

Pneumothorax.  The most frequent complication for chest procedures is pneumothorax, which can be treated easily with a small-caliber

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2449

B FIG 67-83  A, CT scan shows almost complete collapse of the left lung. The thin plastic needle sheath is noted in the air space just anterior and contiguous to the lung. B, After aspiration of the air, the lung is completely expanded. The patient had repeat chest x-rays over 4 hours documenting that the lung remained expanded. The patient was discharged home without incident.

tube and a one-way valve. The decision to insert a chest tube depends on patient symptoms, patient attitude, attending service, and patient reliability. In ideal circumstances, the chest tube is inserted only if the patient is symptomatic. Even a 50% pneumothorax does not require treatment unless the patient is symptomatic. If there is any question of the patient’s symptoms, reliability, or housing status, the chest tube should be inserted. If a patient is healthy, one can attempt to defer chest tube insertion by performing a simple aspiration of the air and observing for a recurrence of a pneumothorax over several hours. Simple aspiration of air from a pneumothorax may prevent the necessity of chest tube insertion (Fig. 67-83). Yamagami and associates reported the effectiveness of simple aspiration in two different groups of patients.522 In their report on CT-guided biopsies, 283 consecutive patients were treated with either simple aspiration or chest tubes. Of a group of 52 patients with pneumothoraces, manual aspiration of the air was performed immediately after the biopsy; 91% had total resolution of the pneumothorax. If the amount of air aspirated was more than 543 cc, a chest tube was required. In their second report, Yamagami’s group used the technique on pneumothoraces after RFA.522 In 129 patients treated with RF, there was a pneumothorax rate of 29.5%. Fourteen of these cases were treated by manual aspiration, and 24 were simply observed. We use this approach in some patients when the pneumothorax is small and the patient has few symptoms. In such a case, if the repeat chest x-ray shows no recurrence, we discharge the patient without a chest tube as long as the patient is reasonably healthy, lives close to a healthcare facility, and does not live alone. Any recurrence or progression of the pneumothorax from the first chest x-ray (after air aspiration) to the second would prompt us to insert a chest tube. Chest tube insertion is required if a patient becomes short of breath or the follow-up imaging shows a progressively enlarging pneumothorax, or both. It is best for an inexperienced operator to insert the chest tube under fluoroscopy, but when experience has been gained, it is important to be able to insert the tube quickly at the bedside. The

placement of the chest tube is simple and can be accomplished in the following fashion. One should use an entrance site in the anterior second or third intercostal space on the side of the pneumothorax. After preparation of the site with antiseptic solution, one should thoroughly anesthetize the site down to and including the pleura. One can estimate the depth required for the tube insertion by using the anesthetic needle to aspirate air at the site. After administration of the local anesthetic, a nick in the skin is made with a scalpel to provide ample space for the tube. As one pushes the tube through the chest wall, a “give” is noted; this indicates passage through the pleura. The tube is kept perpendicular relative to the axial plane until it passes into the pleural space. It is then angled slightly cephalad so that the tube lies in the apical lateral region (Fig. 67-84). This location is optimal because in this position it would evacuate “rising” air when the patient is either lying down or standing up. After placement of the tube, one should always reconfirm the proper location of the tube even if the placement was optimal. This can be done by checking the tube position with oblique or lateral films. Another method is to place the end of the valve close to one’s ear and listen for the escape of air through the valve. After the placement of the tube, patients usually experience some pain as the lung reexpands against the tube. A single small dose of a narcotic agent relieves the discomfort, and it seldom recurs. The tube is securely attached to the chest wall by suture and tape to prevent displacement. A complete explanation should be provided to the patient and nursing staff to prevent inappropriate or premature closure of the tube. If the lung is reinflated for 10 to 12 hours, the tube can be removed as follows. The tube should be clamped for several hours and a repeat film taken. If the lung remains expanded, the tube can be removed. If not, the valve should be reopened for 24 hours and the procedure should be repeated later. If severe primary lung disease is present, expansion could take 2 to 3 days.

2450

PART III  Image-Guided Procedures

A

B FIG 67-84  A, Disposable chest tube trays such as this are available. These are complete and provide all necessary equipment, including the instruments, tube, and Heimlich valve. B, This chest film shows the chest tube in the left upper lobe. Whenever the tube is positioned, one should clinically check for the egress of air to ensure proper function.

Hemoptysis and Intrabronchial Bleeding.  Coughing up a small amount of blood is of no great concern. In earlier times when only fluoroscopy was available and approaches were limited to direct anterior or posterior pathways, there was a higher incidence of hemoptysis. With that method, it was the accepted convention that a patient could cough up one third cup of blood without consequence. Seeing this amount of bleeding is very disconcerting. The best management is to place the patient on oxygen and calmly reassure him or her. Repeat scans might be appropriate to ensure bleeding is not progressing. If bleeding persists, a resuscitation should be called for intubation and other supportive measures. Treatment for intrabronchial bleeding is temporizing until a formal resuscitation can be performed with the anesthesia service. If a patient has severe hemoptysis, he or she should be placed in the ipsilateralside-down position. This position lets blood pool in the impaired lung and protects the uninvolved lung, which is elevated. The patient should be given intranasal oxygen so that oxygenation is enhanced. An emergency page for anesthesia is appropriate so an endotracheal tube can be placed into the functioning lung. Support of the patient can be via ventilating the unimpaired lung.

Intrabronchial Bleeding and Air Embolism.  Intrabronchial bleeding is a serious problem with potentially fatal consequences, so immediate action is required. The appropriate action for the radiologist is to place the patient on oxygen, prepare for cardiopulmonary resuscitation (CPR), and call an immediate resuscitation with anesthesia support. The anesthesia team should intubate the patient, placing the tube in the nonbleeding side to tamponade the bleeding in the affected side. The complication of air embolism was uniformly fatal until more recently. Two cases reported by Lattin and colleagues287 and Ohashi and coworkers385 noted the recovery of patients who were treated with hyperbaric oxygen. Based on this limited experience, if air embolism occurs, one should administer 100% oxygen immediately and use hyperbaric oxygen if available.

LIVER Liver biopsy is a frequent procedure because the liver is commonly involved with malignant processes and because benign disorders mimic malignancy. Because specific differentiation of these various processes is impossible by imaging methods, definitive diagnosis by biopsy is necessary for planning proper patient management and avoiding unnecessary surgical procedures.

Image-Guided Background Although Lundquist318 showed the usefulness of blind skinny needle biopsies, guided biopsy of the liver with the skinny needle was infrequent until the introduction of real-time Doppler US and CT scanning. Rasmussen and colleagues415 performed the first US biopsy of the liver. Their study noted the increased diagnostic yield of US-guided skinny needle biopsy of liver masses compared with blind Menghini biopsy. The improved yield was secondary to the accurate targeting provided by US. This article is sometimes mistakenly quoted as showing the superiority of the US-guided skinny needle over the larger cutting needle guided by CT. However, this article does not even include CT-guided procedures, so any comparison or inference is not valid. Many other authors later confirmed the high accuracy of the skinny needle procedure of the liver and the US guidance. Since the first CT biopsy of the liver by our group, a large body of experience with the skinny needle and larger cutting needles has accumulated. Haaga and colleagues,177,182,183 Ferrucci and associates,126 and Iselin and colleagues,237 among others, confirmed that CT-guided skinny needle procedures are effective for the diagnosis of many primary and metastatic malignancies. During direct comparisons of the skinny needles and cutting needles, Haaga and colleagues189 and Ha and coworkers176 confirmed the usefulness of the skinny needle and noted additional advantages with the larger cutting needles. Currently, US and CT are used for the performance of biopsies using large cutting needles. US guidance is adequate for easily visualized lesions, but for more difficult or complicated cases, CT is the modality of choice

CHAPTER 67  Image-Guided Aspirations and Biopsies

2451

because of advantages associated with guidance and diagnostic scanning (see later). Pagani investigated the significance of needle caliber on the diagnostic yield and complication rate of aspiration biopsy of liver masses.394 With CT techniques, he found that there was an increased diagnostic yield with the larger needle and no difference in complication rate (see “Diagnostic Accuracy,” later).

New Cancer Initiatives The topic of new cancer initiatives is discussed in this section because of the common occurrence of cancer in the liver and the feasibility of obtaining substantial core biopsies for tissue assay. In recent years with the development of new sophisticated treatments for cancer, a new approach for cancer drug development and treatment has evolved based on acquisition of biopsy tissue. The success of this new approach is contingent on obtaining sufficient tissue for assays. These initiatives have focused on the ability to measure molecules and microchemistries of tissue samples. From the tissue samples, microassays are performed to determine the presence and effectiveness of the administered drug. Using these methods, remarkable findings have revolutionized the outcome of such studies.106,472,473 Probably the greatest testimony to this trend is the change in the vernacular for cancer therapy. Previously researchers would talk about the MTD (maximum tolerable dose),472,473,514 whereas they now talk about the OMD (optimal modulatory dose)106,514 of the chosen drug or modulating agent. Because the success of these methods depends on acquisition of adequate tissue, we encourage physicians to become familiar with core biopsy needles of various sizes, including 18-gauge and 14-gauge needles. In the future, cancer patients are likely to undergo repetitive biopsies over the course of their treatment to assay drug, change drug therapy, and modulate treatment based on such samples. The approach would be similar to that of sampling and treatment of infectious agents that occur in abscesses or blood cultures. Although these methods are currently used only in study protocols, it is likely they will enter the general clinical community in coming years.

FIG 67-85  Diagnostic CT scan of a patient with colon cancer shows a highly vascular lesion in the left lobe of the liver (arrow). This lesion is so vascular that no biopsy should be performed with standard methods (aspiration or cutting).

Indications The indications are broad, considering the diagnostic and guidance accuracy of US and CT. Any patient with a focal hepatic mass or unusual anatomic configuration should have a guided rather than a blind procedure. Diffuse processes of the liver are probably best performed by US or a transjugular approach if the coagulation factors are abnormal. Most clinicians have abandoned the unguided bedside method.

Contraindications Contraindications are few and consist of uncorrectable coagulopathy, numerous varices or collateral vessels obscuring the lesion, unusual anatomic location within the porta hepatis or adjacent to venae cavae, hypervascularity of the lesion, and inability of the patient to cooperate. Although hemostasis is now possible using the technique described previously, it is unlikely hemostasis would be effective if a large vessel were cut. Some masses are too vascular to biopsy safely and should be triaged for laparoscopic biopsy (Fig. 67-85). Lesions that may be too vascular include focal nodular hyperplasia, hepatomas, hemangiomas, sarcomas, and metastases (see Chapter 44). The issue of poor cooperation can be remedied by deep or general anesthesia. The presence of ascites is not a contraindication for biopsy. Murphy and colleagues reported a group of patients who had liver biopsies, some with ascites and others without.370 The complication rate in the patients who had ascites (32%) was less than that in the patients who

FIG 67-86  CT-guided biopsy of the liver with the presence of ascites (arrow). The presence of ascites is not a contraindication to biopsy (see text). In the series by Murphy and colleagues, patients with ascites had lower bleeding rates as long as the coagulation laboratory values were normal.370

did not have ascites (43%). This was not statistically significant, but the converse must be true; there were no increased complications associated with the presence of ascites (Fig. 67-86).

General Technique for Liver Biopsy

Preprocedure History, Review of Images, Planning, and Informed Consent.  Patient history and diagnostic scans are reviewed

2452

PART III  Image-Guided Procedures

and informed consent obtained as noted in the earlier general section. Review of patient history, including liver failure, bleeding issues, and current medications, is important. Laboratory studies should be assessed to evaluate for coagulopathy; if issues exist, they can be addressed as noted earlier under “Coagulopathic Treatment.” Abnormal liver function should alert the radiologist to be observant for any capsular varices or typical varices that might be subject to bleeding if inadvertently damaged. Laboratory values for CBC and coagulation factors should be suitable, as discussed earlier. Most liver biopsies can be performed as an outpatient procedure, assuming inpatient facilities are available as needed. After studying a series of 1000 patients, Perrault and coworkers advocated that virtually all liver biopsies can be performed as outpatient procedures if there are no complicating factors.404 Appropriate consent should be obtained and documented. Review of prior scans helps preplan the procedure relative to possible approaches and patient positions. It saves time to pre-position the patient if a position other than supine is to be used. Any enhanced CT scan may show vascularity and other anatomic features that are important. PET scan can show metabolically active areas (Fig. 67-87A and B) or necrotic areas for target planning, but prior surgery or ablation may produce findings suggesting neoplasm. Metastatic lesions are typically positive, but hepatocellular carcinoma may have a high false-positive rate. In some unusual cases, tumors may alter their metabolism so as not to consume glucose and may be PET negative (see Fig. 67-87C through E) even if the biopsy is positive; any rapidly growing mass should be biopsied (see Fig. 67-87F). Hepatocellular carcinoma is less amenable to evaluation by PET using FDG. Many authors report an accuracy of detection ranging from 50% to 75%,256,492,520 and others have reported that FDG uptake

A

D

correlates well with poorly differentiated tumors and tumors with fibrosis. In postoperative patients with increasing α-fetoprotein (AFP), Chen and associates found that 71% could be detected with FDG PET.65 In addition to FDG tracers for PET, other authors have reported the use of acetate222 and choline derivatives484 with good results. Additional studies are indicated before the role of these new agents can be defined.

Imaging Preferences.  Choice of modality between US, MRI, or CT must be made by the individual physician, but CT generally is still considered the gold standard. Difficult anatomic location, poor visualization under US, or concern about increased vascularity is a clear indication for CT guidance. With increasing expertise, modern US refinements, and increased demand on CT scanning time, routine mass biopsies most often should be performed with US guidance. Cases that are difficult to visualize and access because of location or large body size are well suited to CT. Lesions in precarious locations, such as adjacent to the portal vein, gallbladder, diaphragm, or other site, may be triaged for CT guidance. Biopsy of patients with abnormal coagulopathy should be performed with CT or combined with fluoroscopy so that a hemostatic method can be used should bleeding occur. The changes produced by bleeding and coil insertion make US less useful for performance of the hemostatic coaxial/coil/thrombin technique.

Procedure Steps.  During the procedure, vital signs, neurologic status, and pulse oximetry should be monitored. Analgesics and sedative medications can be administered as needed (see earlier under “Patient Cooperation and Sedation”). Target and trajectory determination.  A diagnostic series of scans is taken in anticipation of the procedure. From the current scans, the

C

B

E

F

FIG 67-87  This patient with breast cancer had coexisting masses in the liver, some of which were FDG positive and one that was not. A, PET study shows an FDG hot spot in the anterior right lobe of the liver. B, CT scan of the same area shows a low-density mass with tumor in the anterior right lobe (arrow). C, CT scan shows a small lesion in the posterior right lobe, which grew and was biopsied. D, Repeat CT scan shows that the lesion (arrow) has grown. E, CT-guided biopsy of site is positive for cancer. New information in the cancer literature indicates that tumors can change their metabolism to self-consumption of proteins, called autophagy, and not burn glucose, being negative on FDG PET. F, PET scan performed before biopsy shows negative uptake in the area of the liver that was biopsied.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2453

FIG 67-89  Avoidance of the diaphragm is the most reasonable FIG 67-88  Large mass in the right lobe bordering against the capsule. As shown, the needle trajectory should include a cuff of normal liver and avoid direct biopsy of the capsular area; this prevents bleeding and tumor seeding in the needle path. The tissue core should be obtained in the transition area between normal tissue and tumor, avoiding the necrotic center (see text).

mass referred for biopsy is reevaluated. If the lesion is stable or larger, the procedure should proceed. If the lesion is significantly smaller, one should consider deferring the procedure until a later date after follow-up scans are performed. Triage to MRI or US likely should be performed if the mass cannot be well visualized on CT. One selects the tentative target for biopsy and the appropriate pathway for the trajectory. If the mass is large, one selects a portion of the mass that is not necrotic and attempts to target a site at the interface of the tumor and normal tissue. In many cases, especially with primary liver tumors, the transitional area may provide more diagnostic information to differentiate the various cirrhotic nodules associated with transition from dysplastic tissue to neoplasm (Fig. 67-88). The trajectory pathway should be planned as follows. If the lesion appears mobile, which is common close to the diaphragm, the patient can be positioned with the ipsilateral side down, which may lessen motion of the diaphragm. The pathway should provide a small cuff of normal tissue (which tamponades any possible bleeding and prevents tumor seeding in needle track) and yet minimize the pathway through the liver. Capsular lesions should not be approached directly but have an angled approach, including a cuff of normal liver (see Fig. 67-88); this has been shown to prevent tumor seeding and reduce the probability of bleeding (see “Complications and Factors Affecting Incidence,” later). The diaphragm and pleural space should be avoided if possible; it is best never to have a pathway more posterior than the midaxillary line or to use patient position to better display the pleural space (Fig. 67-89). Although some operators have used the transdiaphragmatic approach, we do not recommend it.

approach. This patient was positioned in the left posterior oblique position, which exaggerated the visibility of the pleural space. No pleural space was noted, but the needle path was close to the midaxillary line, which minimizes the probability of traversing the pleura based on the fact that the diaphragm inserts higher anteriorly than posteriorly (see Fig. 67-53).

Bolus dynamic contrast scan over the chosen target site should  be performed to visualize and avoid any significant-sized vessels  (Figs. 67-90 and 67-91) and exclude excessive vascularity, as noted  in the contraindications section. Lesions such as hemangiomas (Fig. 67-92) can be detected and the procedure terminated. Intensely  vascular lesions may bleed even with FNA (Fig. 67-93) and cause complications. Instrument choice and techniques.  The choice of needles is based on the clinical history of malignancy, accessibility of the lesion, vascularity of the lesion, and type of tissue sample required. The choices are as noted in the earlier section on FNA for cytology or a cutting needle for larger amounts of tissue (accuracy of tissue recovery and steps to improve outcome are discussed later). If the patient has had a proven malignancy and there is need only for confirmation of recurrence, FNA is adequate. If lesions are hypervascular, FNA is preferred over cutting needles, but even then issues may arise (see earlier). If the patient has an undiagnosed tumor or additional tissue is being obtained for further characterization and analysis to select chemotherapy or other course of treatment, a cutting needle is appropriate. The caliber of the cutting needle can be chosen according to the amount of tissue needed for analysis, experience of the interventionalist, and concern for complications. The relative amount of tissue acquired and the associated bleeding are presented in Tables 67-3 and 67-4. Of the needle techniques described earlier, the coaxial system should be used regardless of the needle type chosen. As noted earlier, with greater dissemination of liver lesions, numerous cases of tumor seeding have been reported. The choice of a cuff of normal liver and the coaxial method have been shown to prevent tumor seeding (see “Complications and Factors Affecting Incidence,” later).

2454

PART III  Image-Guided Procedures

A

B FIG 67-90  A, Bolus CT scan of the liver before biopsy. Note enhancement of the vessels and visualization of the medium-sized portal vein anteriorly (arrow). B, The needle trajectory avoided the site of the vessel, preventing possible damage.

FIG 67-92  Significant hemorrhage can have numerous causes. This scan shows a large hematoma outlining the capsule of the liver in the lower portion of the abdomen, caused by a single pass of a 20-gauge Chiba needle (performed by the author). The patient has diffuse metastatic renal cell carcinoma to the liver that was very vascular.

FIG 67-91  Needle biopsy of a lesion in the left lobe is noted. The level of the needle pathway was chosen at this level above the left portal vein to avoid penetration of the vessel.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

P

B

P

2455

FIG 67-93  Hemangioma is a common benign problem that has the potential for catastrophic bleeding. A, This CT scan was taken during the contrast phase and shows the characteristics of feeding vessels in the wall of the lesion (arrows). B, Late-phase CT scan shows the mass to be completely homogeneous in the delayed phase. These qualities are pathognomonic, and no procedure should be performed.

TABLE 67-3  Statistical Results for Each Outcome Measure

Comparison

Blood Loss (g)

Blood Loss (g) with Amount of DNA per Sample (µg)

Blood Loss (g) with Needle Chamber Volume (mm3)

Blood Loss (g) with Amount of DNA per Sample (µg) and with Needle Chamber Volume (mm3)

14 vs. 18 gauge 18 vs. 20 gauge 14 vs. 20 gauge

.025* .059 <.001†

.24 .76 .47

.14 .42 .07

.002† .025* .007†

Note: Data are P values. *Difference was statistically significant (P < .05). † Difference was statistically significant (P < .001).

TABLE 67-4  Factors Affecting Hemorrhagic Complication Rates After Renal Biopsy Factor Hemodialysis   No   Yes Patient sex   Male   Female Side of biopsy   Right   Left Pretreatment with desmopressin acetate   No   Yes

Hemorrhagic Complication Rate (%)*

Increased Risk of Hemorrhagic Complications

5.9 (13/220) 24 (5/21)

— ×4.1

<.025

4.5 (7/156) 13 (11/85)

— ×2.9

<.01

5.2 (8/155) 14 (12/86)

— ×2.7

<.01

6.4 (15/236) 60 (3/5)

— ×9.4

<.001

P Value†

From Sateriale M et al: A 5-year experience with 307 CT-guided renal biopsies: Results and complications. J Vasc Interv Radiol 2:401–407, 1991. *Numbers in parentheses are actual number of patients with hemorrhagic complications per total number of patients in each subgroup. † P values indicate the significance of the difference in complication rates for each set of paired values.

2456

PART III  Image-Guided Procedures

Choice and palpation of the entrance site.  When the tentative trajectory and entrance site are chosen, a marker is placed on the skin and a repeat scan is performed to see the marker relative to the pathway. The marker may be a purchased device or a hypodermic needle in the package. After one has selected a tentative entrance point and before needle insertion, one should carefully inspect the chest at the entrance site for several problems. If the patient has cirrhosis, one should inspect the site for any collateral veins, which may form secondary to portal systemic collaterals, such as a caput medusae. The ribs should be palpated to find the intercostal space, and the entrance site should be moved if necessary to avoid ribs. One should not attempt to anesthetize a site that is not over an intercostal space. First, pricking the periosteum of the rib is very painful to the patient, and his or her confidence in the radiologist and personal cooperation may be lost by this painful experience. Second, if one does not move the site, a successful procedure is unlikely because attempts to “angle” around a rib or leverage against it seldom work well, and the procedure would be prolonged and likely unsuccessful. Site preparation.  Careful antiseptic preparation of the skin is important. Proper administration of lidocaine to produce a “painless” procedure, as noted earlier in the general section, is important. When the area is anesthetized, it is important to infiltrate anesthetic through the subcutaneous tissues and down to and including the capsule of the liver. Adequate anesthesia of the capsule makes the procedure virtually pain free and permits insertion of the needle without the patient’s moving because of pain. If a rib is inadvertently hit during the initial anesthetic administration, we give a small bolus of lidocaine on the rib; this prevents any recurrence of pain if that area is struck again during the procedure. Needle selection and insertion.  The appropriate needle, either FNA or a cutting needle, is selected using the factors discussed earlier. The needle is inserted with incremental steps and adjusted until placement of the needle in the target site is accomplished. The patient is asked to hold his or her breath during each movement of the needle. If the lesion and needle are in the same axis, the procedure is usually straightforward, but if not, the angled methods noted earlier are appropriate. With the aspiration needle, localization of the tip precisely in the chosen site is important if adequate cellular material is to be obtained. If a cutting needle is being used, one should adjust the position of the needle and stylet so the chosen area is included in the gap (see Fig. 67-88). If the stylet is not properly positioned, one can try the maneuvers noted in the early section to position the stylet better; intentional deflection, coaxial manipulation, or rotation of the needle may be performed (see Figs. 67-44 through 67-47). During needle insertion and sample acquisition, whenever the needle is moved, the patient should be asked to suspend respiration. Not doing so may cause shearing of the capsule and cause a complication. Sample acquisition.  The sample should be acquired as noted previously in the general section technique. We routinely take only a single needle pass unless there has been an error in the technique or a poor sample obtained on visual inspection. If a proper sample is not obtained with FNA, we move up to a 20-gauge cutting needle at the same session. If a special sample has been requested for flow cytometry, DNA analysis, or microchemistries, larger needles such as 15 gauge or 14 gauge may be used and as many samples taken as needed (see Tables 67-3 and 67-4). The coaxial system is left in place until the multiple samples are obtained. If the site seems to be bleeding, we suggest using the coaxial/ coil/thrombin hemostatic method described earlier (see Fig. 67-32). If local injection of blood products is being used, another aliquot of FFP

or platelets should be injected into the site as the coaxial system is withdrawn (see Fig. 67-31). The coaxial cannula is withdrawn and a repeat scan is performed to assess for bleeding.

Postprocedure Care.  Before removing the patient from the table, we take a repeat scan to evaluate for postprocedure problems. If any hematoma is noted, the clinical team is contacted and a type and crossmatch of blood products may be ordered if there is a concern that the bleeding is significant. The postprocedure care of these patients is simple and consists of observing them for abdominal discomfort and monitoring vital signs for possible bleeding. The patient is kept on bed rest for 4 hours. Vital signs are checked every 15 minutes for 1 hour, every 30 minutes for 1 hour, and every hour for the next 2 hours. After 4 hours, if no issues arise, the outpatient can be discharged home with appropriate instructions. Patients should be instructed to rest quietly at home for 1 day with no strenuous activity. If any pain or other symptoms develop, the patient should seek medical attention.

Diagnostic Accuracy The success of percutaneous liver biopsy by CT guidance depends on the various important factors noted earlier. In our experience, the success rate for malignancy and benign disease has varied according to the type of needle used and the abnormality being sought. Looking at the literature, the average success rate for sampling all malignant disease with aspiration needles is 77.6%, and for cutting needles it is 90.2%. The success rate for benign disease is 39.4% with aspiration needles and 88.5% with cutting needles. With individual techniques and as experience has been gained, the diagnostic accuracy has increased. Historically, guided biopsy procedures are superior to blind or unguided biopsy procedures. According to Conn and Yesner,78 blind biopsy positivity depended on the large amount of disease present, so such a random biopsy had a greater probability of being positive if extensive disease was present. Lundquist318 did his first biopsies without guidance by palpating the masses through the skin surface for localization. Rasmussen and colleagues reported the first series on US biopsy using a 22-gauge needle with multiple passes.415 They recovered positive tissue in 70% of cases with US guidance compared with an anticipated rate of 23% for blind biopsies (extrapolating Conn’s78 classification of his cases as described in the report). Subsequent reports showed similar results, with success rates ranging from 66% and 94%. All of the comparison articles concerning diagnostic yield from aspiration and cutting needles (single passes) have shown that cutting needles are superior to aspiration needles. The two comparison articles using a single pass were reported by our group. Martino and associates showed that the diagnostic yield of Menghini and Tru-Cut biopsy needles was 88.1% compared with 60.9% for 22-gauge aspiration needles.329 Our more recent experience reported by Ha and colleagues showed the 14-gauge Tru-Cut needle to be superior to the 20-gauge aspiration needle, with yields of 90.2% compared with 77.6%.176 A comparison article by Jacobsen and coworkers showed a slightly greater accuracy with a skinny needle239; however, their comparison was flawed by the fact that they took three passes with the skinny needle and only one with the cutting needle. For benign disease, all reporting authors have shown a clear advantage for the cutting types of needles relative to determining specific disease processes. In our more recent experience reported by Ha and colleagues, the cutting needles provided a specific diagnosis in 88.5%, compared with 39.4% for aspiration needles; nonspecific diagnosis was provided in 11.5% by the cutting needle and 48.8% by the aspiration needle.176

CHAPTER 67  Image-Guided Aspirations and Biopsies The combination of slide smears and histology can provide an added increment of positivity for the results. Numerous authors, including ourselves, attest to the benefit of combining aspirate samples with cell block histology. In the study by Ha and colleagues, the addition of cell block histology increased diagnostic yield from 54.8% to 77.4%.176

Factors Affecting Accuracy.  Many factors affect the diagnostic results from liver biopsies. In our own experience with FNA needles, we limit our needle passes to two in virtually all cases. The expected yield should be about 75% to 85%. These factors include (1) extent of disease in liver and size of lesions, (2) type and size of needle used, (3) proper technique, (4) number of needle passes, (5) experience of the interventionalist, and (6) choice of modality. The extent of disease has a great impact on biopsy results. If mass lesions are large, numerous, and accessible, the chances of a positive biopsy are greater. Conn and Yesner studied this factor with a series of blind biopsies performed on cadaver livers at autopsy.78 They graded the degree of involvement of the liver by the number of lesions and found the following recovery accuracy: grade I, 1 to 3 lesions, had a 16% success rate; grade II, 4 to 25 lesions, had a 27% success rate; grade III, 25 to 50 lesions, had a 45% success rate; and grade IV, with greater than 50% involvement by weight, had a 86% success rate. This finding is important to consider whenever a new method, technique, or type of needle is first being reported in the literature. Proper technique includes accurate placement of the needle’s sampling portion within the lesion, application of vacuum if needed, use of the appropriate needle device, and proper recovery of the tissue sample (see earlier under “General Procedure Steps”). There are some variations of different aspects, and technique can and should be modified according to the preferences of the operator, but a consistent technique is crucial. The type of needle affects outcome, with cutting needles providing greater diagnostic tissue than cytologic samples. All authors reporting the use of cutting needles reported improved diagnostic accuracy and more specific information based on sophisticated histopathologic evaluation or assays. Automated cutting needles are better than manual types for most interventionalists. Before their introduction, cutting needles were seldom used for biopsies because only a few radiologists had the dexterity to use the standard manual devices, such as the Tru-Cut. In a survey, more than 90% of radiologists performing liver biopsies used the automated devices. Hopper and associates, in the evaluation of numerous automated devices, noted that the gold standard device with which all devices should be compared and the one as yet unsurpassed is the 14-gauge manual Tru-Cut device228; such manual devices have a long learning curve, so automated devices are preferred. Meticulous placement of the needle is critical regardless of the guidance modality used. With a standard aspiration needle, the needle end must be in the lesion and not adjacent to it. With a side cutting needle, the tissue receptacle of the stylet should be positioned at the interface of the focal lesion, with only a small portion traversing the normal rim of tissue. Better quality pathologic interpretation is likely if an interface between normal tissue and neoplastic tissue is provided for review and staining. Also, areas of definite necrosis, more noticeable with contrast material, should be avoided. Larger-caliber needles provide more ample tissue, enhancing pathologic interpretation. Larger samples permit recutting of samples so that different modern biomarkers can be assessed. Such biomarkers are crucial for modern oncologic treatment, which targets specific enzymes, peptides, or membrane markers. The effect of needle size on diagnostic success has been well documented for aspiration needles in

2457

in vitro and in vivo studies. Andriole and colleagues studied the yield in liver biopsy of various aspiration needles and found that in vitro there was a definite increase in the diagnostic material with the larger caliber needles.17 In a series of liver biopsies, Pagani reported the results of a small clinical group using 22-gauge and 18-gauge needles.395 Pagani found a significant improved diagnostic accuracy, 85% versus 98%, without any difference in the complication rate. In our own institution, the diagnostic yield and complication rates of various aspiration needles have been determined and compared with cutting needles; these rates were reported by Martino and associates329 and Ha and colleagues.176 With a single pass with aspiration needles and providing samples for slide smears and cell block material, the diagnostic yield for malignant disease with a 22-gauge needle was 60.9% compared with 77.6% with a 20-gauge needle. Pagani’s data were based on the performance of multiple passes, and our data involved single passes, which explains their higher yields. A laboratory study by Plecha and coworkers compared the amount of diagnostic tissue obtained with the amount of local bleeding with various calibers of Tru-Cut needles (see Tables 67-2 and 67-3).407 They found that per unit of tissue, there was slightly more bleeding with the larger 14-gauge needle than with the 20-gauge needle. The data showed further, however, that if one looks at the relative tissue yield (reflected by the quantity of DNA) compared with the blood loss, there is relatively no difference among the various needle calibers (20 gauge, 18 gauge, and 14 gauge). Theoretically if one was required to obtain a finite amount of tissue, the efficiency and risk of using a 14-gauge needle may be less. As noted earlier, the amount of bleeding per unit of diagnostic tissue is no different among the needles. Also, significant bleeding is usually the result of inadvertently striking a small artery. This is essentially a random event, so that if multiple passes with a small-caliber needle are required, there is a statistically greater chance of bleeding. Three or four passes with a 20-gauge needle may be less safe than one pass of a 14-gauge needle. It has not been clinically determined whether multiple smallcaliber Tru-Cut needles are safer to use than a single pass with a large needle. In our practice, we have had a major complication with the use of a 20-gauge Tru-Cut, which resulted in a major hemorrhage and required a surgical procedure. Multiple needle passes increases accuracy. From these data and the data of Ferrucci and colleagues,124,126 it is clear that the more needle passes taken, the higher the diagnostic accuracy. In their series, Ferrucci’s group selected 20 cases that were positive and looked at the diagnostic yield for each needle pass.124,126 The yield was 75% with one pass and 90% with two passes. (One can never expect 100% yield even with five passes, because the data were selective; in the same article, Ferrucci’s overall yield was 82%.) Even with blind biopsies, the diagnostic yield increases appreciably with multiple passes. Experience of the operator directly affects the outcome of interventional procedures. Poon and colleagues reported on procedures of the liver, including biopsy and treatment, and found lower complication rate occurs with experience.409

Complications and Factors Affecting Incidence As experience with guided biopsy procedures has accumulated, the same types and spectrum of complications have occurred with guided biopsies as have occurred with unguided procedures. The most common complication is bleeding; others include biliary peritonitis, arteriovenous fistula,411 tumor seeding, hepatic infarction, pneumothorax, inadvertent puncture of bowel, and abscess. Before discussing the factors that affect complications, we note several issues that have been shown to be irrelevant and note issues contributing to complications. The presence of ascites is not a

2458

PART III  Image-Guided Procedures

contraindication for biopsy of the liver. A review of our cases and several reports by other authors confirm this. Murphy and coworkers370 and Little and colleagues305 showed that ascites had no impact on complications. From a common sense perspective, one can understand this easily. If one assumes that local pressure on a biopsy site might tamponade bleeding, the presence of intraperitoneal fluid increases intraabdominal pressure and surely would not have an adverse effect on hemostasis. If there is any relationship between ascites and complications noted in the earlier literature, it is likely that those physicians had trouble percussing the liver because of the ascites, and the coagulation factors were likely abnormal because of liver compromise associated with ascites. If coagulation studies are normal, ascites should be irrelevant. The more obvious factors that affect complications include coagulation status, lesion characteristics, characteristics of the needle, the number of needle passes, and the technique used. Coagulation status affects the incidence of bleeding complication. For biopsy using either an aspiration or a cutting needle, we prefer to have the platelet count greater than 50,000/µL and the INR less than 1.3. If either value is not within the desired range, corrective action should be taken as needed. See under “Coagulopathy Treatment” for specific details of evaluation and treatment. Anatomic factors such as lesion location, characteristics, and vascularity (see Fig. 67-92) affect the potential for complications. The location of the lesion affects complications in obvious circumstances. Lesions in the porta hepatis or adjacent to the gallbladder add the potential risk of biliary damage. Traversing or biopsying large vessels introduces a greater chance for bleeding; we do not accept the cavalier attitude of some who assert that this is of no significance. Masses close to the diaphragm increase the risk of pneumothorax and require special consideration of trajectory to avoid penetration of the diaphragm (see Figs. 67-88 and 67-89). Most importantly, lesions on the capsular surface are difficult to access, and direct biopsy on the capsular surface amplifies the chance of bleeding or tearing of the capsule. Historically, in the early period of liver biopsies, metastatic disease was considered a contraindication for biopsy because of the high incidence of severe life-threatening hemorrhage associated with blind biopsy. Blind biopsy inadvertently targeted such capsular lesions because manual palpation was used as a means to locate masses. The vascular nature of the abnormality is important. We believe that performing a stationary bolus dynamic scan on every patient having a cutting needle biopsy is important. Many primary and metastatic cancers, benign masses, and conditions can produce large or numerous vessels that could produce catastrophic bleeding if inadvertently damaged. The most notable lesions are high-flow hemangiomas or cancers such as hepatoma, angiosarcomas, or metastatic tumors. Bret and colleagues reported a small series of hepatomas biopsied using a 22-gauge needle for aspiration with associated serious hemorrhagic complications.45,46 One patient required transfusion, two patients required corrective surgery, and one patient died. Considering the small caliber of the needle and the experience level of the authors, the only explanation for the bleeding was the vascular nature of the tumors. There also is a case report of a bleeding death from a 22-gauge needle biopsy of a cavernous hemangioma that was not evaluated by a dynamic study. Smith also reported that one cannot prevent complications by using a fine needle (see Fig. 67-92).459 From a mail survey, Smith determined that in a series of 63,000 reported cases, there were 27 cases of significant hemorrhage requiring transfusion and 3 deaths.459 Cirrhosis also presents a special problem because cirrhotic patients may have networks of small varices in or adjacent to the capsule or

umbilical vessels of the falciform ligament, which are the result of portal hypertension. Such patients typically have poor coagulation status and a “stiff liver.” The needle characteristics have an impact on complications, but considering the favorable results reported by many, it is likely that any traits that tend to increase the potential can be minimized by proper techniques. Relative to aspiration needles, the caliber of needle has minimal impact, but experience has shown that careful proper techniques can minimize any adverse affect. Our own group documented an improvement in complication rates over years of experience and refinement of our technique. In our first report, Martino and coworkers reported a complication rate of 0.83% for serious complications.329 In an interim study by Ha and coworkers, we reported a complication rate of 0%.176 In these two sequential series, the only objective difference was that we switched from a 22-gauge to a 20-gauge needle. One could postulate that the very flexible 22-gauge needle required multiple reinsertions because of needle deflection, but experience also is the likely cause for improvement. Pagani reported no difference in the complication rate between 18-gauge and 22-gauge needles.394 Relative to cutting needles, variable results are reported in both the earlier and the more recent literature. Although many authors have reported different results, it is insightful to review our experience at University Hospitals of Cleveland, Case Western Reserve University, which has had little change in the patient referral pattern, selection of patients, and manual technique over the past 30 years. Three series from our group over time reported improved complication rates with cutting needles. In 1984, Martino and coworkers reported a complication rate of 1.44%,329 and in 1991, Ha and coworkers reported a complication rate of 1.1%.176 In 2001, Dowlati and associates reported that the complication rate for cutting needles was 0.5% in 200 cases.106 The last series reported on exclusively 14-gauge side cutting needles, whereas the other studies had two needle calibers, 18 gauge and 14 gauge. The first study included end cutting and side cutting needles. In a laboratory study on animals by our group (see Table 67-1), there were only minimal differences among any of the needles less than 18 gauge, but there was significant difference between these and the larger gauge needle. Multiple needle passes increase the diagnostic outcome but introduce a higher incidence of complication with all needles, including small aspiration types. If the procedure is not a successful one, a repeat two-pass biopsy is warranted, after which one should choose either a cutting needle or a surgical procedure. The adverse effect of numerous needle passes was best documented in a survey by Smith.459 We concur with his opinion that the number of needle passes even with a small needle should be limited to as few as possible. As is seen in reports in the literature, serious complications occurred in cases involving four or five needle passes, and none occurred in cases involving one or two passes.

Special Pathologic Entities

Hemangiomas.  Cavernous hemangioma is the most common hepatic tumor; it produces few symptoms until bleeding occurs, which is rare. The imaging appearance is typical, showing a low-density  mass on nonenhanced scans but a characteristic appearance with IV contrast material (Figs. 67-94 and 67-95; see Fig. 67-93). With contrast administration, there is a margin of irregular enhancement at the edge of the lesion. The mass becomes almost completely isodense to the adjacent normal liver in most cases, although very slight differences may be noted within owing to variation in histologic traits, such as fibrosis. Correct diagnosis is important to prevent unnecessary surgery and errors in cancer staging. If inappropriate imaging studies are

CHAPTER 67  Image-Guided Aspirations and Biopsies

2459

Hyperplastic Nodules and Cirrhosis.  Successful sampling of nodules of the cirrhotic liver is a complicated process because of the imaging appearance of dysplastic nodules. With cirrhotic livers the appearance of the liver can be very complex, with focal scars, irregular contours, and the presence of numerous nodules. MRI is superior to CT for characterization of the most suspicious nodules, so although one might choose to sample such nodules under US or CT, careful correlation with MRI is essential. In such cases, MRI scans should be immediately available to help determine the optimal target and pathway. It is likely in the future that fusion of the two studies will be used to facilitate the process. The other issues surrounding the biopsy of such cases is that many have coagulopathies related to liver function (see “Coagulopathy Treatment,” earlier).

Hepatoma.  Primary tumors warrant a separate discussion because

FIG 67-94  Hemangioma showing increased vascular enhancement in the anterior margin (arrow). Biopsy performed in such a case would be through a capsular approach. Localized bleeding would be likely.

performed or incorrectly interpreted, a patient may present for a biopsy procedure. In most instances, typical hemangiomas are correctly diagnosed on imaging, but occasionally an atypical one is presented for biopsy. When typical vascular hemangiomas are excluded, there is little risk of a significant complication from the atypical variety. The reason for this is that the pathologic features that make the imaging appearance atypical also preclude the potential for bleeding complications. Stated differently, the increased vascularity of hemangiomas is what gives the typical appearance on such imaging modalities as CT enhancement and isotope pooling. When the increased vascularity is not present because of thrombosis, fibrosis, or hyalinization, the lesion is unlikely to bleed with a biopsy (see Fig. 67-95). Clinically the literature supports the concept that inadvertent or even intentional biopsy of hemangioma is unlikely to cause bleeding if performed in a meticulous manner.83,329 Such lesions reported were generally small, not bordering on the capsule, and without extreme hypervascularity. We have inadvertently biopsied numerous atypical hemangiomas with 14-gauge cutting needles under the assumption that they were lesions other than hemangiomas. Other authors, including Spamer and associates,469 Cronan and associates,82,83 Solbiati and colleagues,464 and Heilo and Stenwig,215 have reported the safe biopsy of atypical hemangiomas (without hypervascularity). These authors also have noted the difficulty in diagnosing hemangiomas with small aspiration needles; they note that hemangiomas can be inconsistently diagnosed with special-purpose needles and consistently diagnosed with the standard cutting needles.

they have some unique traits from several perspectives. Clinically the specific nature can usually be suspected by an elevation of the serum AFP level. Differentiation from other entities may be difficult from an imaging perspective; more recent data suggest, however, that MRI can differentiate between regenerating nodules and hepatomas. Hepatomas tend to have a higher T2 intensity than do regenerating nodules. If such a lesion were visible only on MRI, it would be appropriate to perform an MRI-guided procedure. Results of biopsy of hepatomas vary depending on the technique used. Successful diagnosis may be difficult with small needles because differentiating normal hepatocytes from well-differentiated tumors can be difficult with small aspiration samples. This problem can be overcome by taking numerous aspiration samples or by using a larger needle. The tradeoff is in the form of a potentially higher complication rate. It is well documented that hepatomas have an increased vascularity and predispose to a higher complication rate. Bret and coworkers reported successful diagnosis in 63 of 69 hepatomas with a 22-gauge needle and multiple passes, but numerous complications resulted.46 In their series, they had four major complications in the 69 patients: one death and three hemorrhages, two treated by surgery and one by transfusion alone.

Cystic Abnormalities.  With cystic abnormalities, one important point should be made. The fluid obtained from such lesions may not be diagnostic, and one should always endeavor to sample the wall of such lesions. If one is using an aspiration needle, one should send two samples as follows. The first sample of fluid should be aspirated and sent for cytology. A second cellular sample should be obtained from the opposite wall of the cyst by “reciprocating” the needle against the opposite wall.

Endocrine Tumors.  Biopsy of nonfunctional metastatic endocrine tumors to the liver is straightforward.509 When functional tumors are to be biopsied, one should be cautious and prepared to treat any symptoms if a hormone release occurs. A death has been reported after biopsy of metastatic carcinoid in the liver.36 Generally, biopsy of functional tumors is discouraged. Echinococcus Infection.  Several series of percutaneous treatment of Echinococcus infection are reported in the literature. Although anaphylaxis is not as likely as was previously thought (see “Echinococcal Drainage,” later), percutaneous puncture for diagnosis is discouraged because the imaging appearance of daughter cysts is characteristic, and a serum test can accurately confirm the diagnosis. The major concern in such patients is to prevent spillage of daughter cysts into the peritoneum or other spaces. If therapy is considered, special precautions should be taken; these are discussed under “Echinococcal Drainage.”

2460

PART III  Image-Guided Procedures

M

A

B

M

M

C

D

E

F FIG 67-95  The reliability of dynamic bolus scanning to determine the degree of increased or decreased vascularity can be shown by this unusual case. This series of scans shows images of an atypical hemangioma that does not show the typical qualities. Such lesions are partially thrombosed and have very small capillaries, so no significant increased blood flow is perceptible. A, This CT scan shows a low-density mass (M) in the right lobe of the liver. B, Later CT scan did not show typical enhancement but did show homogeneous appearance in the late phase. C, T2-weighted image shows a large mass (M) with inhomogeneous intensity, not completely characteristic of hemangioma. D, Because there was still concern about hemangioma, an angiogram was performed. This early-phase image shows hepatic vasculature but no increased vascularity. M, mass. E, Late-phase angiogram shows no evidence of “puddling,” said to be characteristic of hemangioma. F, A 14-gauge biopsy of the lesion shows a gap centrally in the lesion. Pathology reports showed hemangioma. The point is not that biopsy should not be performed in hemangiomas, but rather that the bolus study is so sensitive to blood flow that if a lesion is avascular with this method, one can perform a biopsy of an unknown mass with confidence that there would be no bleeding. Rarely such avascular masses ultimately prove to be atypical hemangiomas.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2461

PANCREAS Percutaneous procedures of the pancreas have become widely accepted as the method of choice for the diagnosis of various neoplastic and benign pancreatic problems. Percutaneous aspiration and drainage of pseudocysts and abscess of the pancreas are accepted by surgeons as an important adjunct to management. The major issues of pancreatic procedures are which cases should be triaged to endoscopic physicians and which should be performed by image guidance.

Image-Guided Background Percutaneous FNA biopsy of the pancreas developed shortly after FNA was performed at open laparotomy. Under direct visualization at surgery, Christopherson72 used the skinny needle in the surgical suite to show that the diagnosis of malignancy could be made with the skinny needle. Forsgren and Orell130 used percutaneous skinny needle aspiration for diagnosis of palpable pancreatic tumors. Because it was easy to use FNA percutaneously guided by imaging, the technique quickly developed acceptance as an effective means of eliminating the need for surgery in the predominantly fatal disease so that other palliative therapies could be initiated more quickly. The first guided percutaneous pancreatic biopsy was performed in 1972 by Oscarson and colleagues,390 who used fluoroscopy in conjunction with angiography to localize the needle placement. Although other authors, such as Tylen and associates,495 used fluoroscopy with contrast studies, widespread acceptance occurred after the introduction of US and CT guidance. Holm224-226 first pioneered US-guided pancreas biopsy in 1975. CT-guided biopsy of the pancreas was introduced in 1976 by Haaga and Alfidi.181 US and CT remain the main tools for providing a definitive diagnosis without surgical exploration in nonsurgical candidates. Percutaneous sampling and drainage of pseudocysts, and puncture of dilated pancreatic duct and abscesses began during the 1980s with contributions by numerous authors.126,237,328

Indications The indications for using percutaneous procedures for the pancreas are broad and relate to diagnosis of tumors, pseudocysts, or infectious processes. Relative to the diagnosis of tumors, the threshold for biopsying any solid or cystic mass should be low because of the effectiveness of these methods and their associated low morbidity and mortality. The prospects for early definitive curable surgery or palliative treatment for relief of the severe pain associated with tumor invasion are important. Pancreatic neoplastic processes are lethal, so if a patient has a solid tumor that seems to be operable, a quick diagnosis can permit earlier surgical intervention. If a tumor is obviously inoperable because of local extension or metastatic disease, initiation of palliation to relieve symptoms (e.g., celiac nerve block, gastric bypass) is compelling. Cystic masses can be a simple pseudocyst, necrotic squamous cell tumor, benign cystic tumor, premalignant cystic tumor, or cystadenocarcinoma; diagnosis and distinction of these entities should be made at the earliest time.

Solid Tumors.  The most frequently biopsied pancreatic tumors are adenocarcinomas (Figs. 67-96 to 67-98) derived from the ductal cells, but other cell lines of the gland, stroma, islet cells, metastases, or lymph nodes can produce solid masses in the pancreas. There are some characteristic features of the various tumors, but there are no real patho­ gnomonic signs (see Chapter 46). As with all tumors, initiation of therapy can begin only after a definitive diagnosis is made, and a biopsy by percutaneous or surgical means is necessary for the histopathologic diagnosis.

FIG 67-96  Biopsy of the pancreas performed from an anterior approach. The needle is positioned in the low-density area (arrows) of the tumor. The trajectory was planned to avoid the superior mesenteric vein and smaller veins (curved arrows).

Patients frequently present to a tertiary institution with the presumed diagnosis of inoperable and fatal pancreatic cancer, but biopsy reveals a treatable lymphoma instead (Fig. 67-99). Percutaneous biopsy of pancreatic cancer probably should not be performed unless the cancer is clearly inoperable; most cancers are inoperable at the time of presentation. Reserving percutaneous biopsies for inoperable patients is an especially important principle because new published data indicate that percutaneous biopsies introduce tumor cells into adjacent tissues unless a coaxial method is used. In support of this premise, Balen and colleagues reported that biopsy does not adversely affect the patient survival time for inoperable patients, which includes most patients because of late presentation.20

Cystic Masses or Tumors.  Low-density cystic masses of the pancreas can represent a wide variety of pathologies, including true cysts (commonly occurs in generalized polycystic disease), lymphangioma, pseudocysts, and cystic tumor variants. In the early reported experience, there was little evidence that percutaneous procedures could obtain adequate material to diagnose and differentiate the various entities, but more recent data indicate that such distinctions can be made using some specific biomarkers. Cystic masses from pseudocysts and abscesses are simple to diagnose from the fluid sample. Fluid from abscesses are opaque and purulent. Pseudocyst fluid may have a variable appearance—clear, turbid, or hemorrhagic—but laboratory analysis invariably yields an elevated amylase content (Figs. 67-100 and 67-101). Fluid obtained from cystic tumors (Fig. 67-102) can be analyzed from biomarkers to diagnose cystic tumors. Measurement of carcinoembryonic antigen (CEA) and CA 19-9471 has been effective in diagnosing malignant cystic tumors (see “Diagnostic Accuracy,” later). For these reasons, one can justify percutaneous biopsy of cystic pancreatic masses to make a strongly suggestive diagnosis, at least sufficient to suggest the need for surgery or other intervention.

PET Scanning The role of PET scanning has been studied extensively; apparently there is very little role for PET scanning of the pancreas for selecting for biopsy or improving the biopsy technique. The method is not applicable because the primary diagnostic accuracy for tumor is unfavorable, and differentiation of inflammatory and neoplastic masses is not consistently possible.

2462

PART III  Image-Guided Procedures

T

A

B

C

D

FIG 67-97  A, Dynamic bolus scan shows a low-density mass in the pancreas. The superior mesenteric artery (horizontal arrow) and superior mesenteric vein (vertical arrow) are noted. The patient was positioned in the lateral position to move the bowel out of the way. B, Initial approach shows a loop of small bowel (arrow) intervening in the pathway; also, the direction is such that the needle might strike the superior mesenteric artery (curved arrow). Tumor (T) is noted. C, The needle is readjusted so that it is directed toward the mass and partially compressing the loop of small bowel. By turning the needle so that the bevel is caudal toward the bowel and pushing slowly, the needle can be pushed by. Small bowel typically is pushed aside by peristalsis. D, Final image shows placement of the needle in the pancreatic mass.

20

M

0

FIG 67-99  A large mass (M) in the pancreas was believed to be lymphoma. A cutting needle sample revealed lymphoma, which was effectively treated with chemotherapy.

FIG 67-98  A large mass in the body of the pancreas extends into the submucosa of the stomach. A transgastric approach was used to sample the mass.

CHAPTER 67  Image-Guided Aspirations and Biopsies

FIG 67-100  Guided aspirations of pancreatic pseudocysts are crucial to their differentiation from cystic tumors. This scan shows aspiration of a small cystic mass adjacent to but behind the pancreas. Because of the high amylase content, the concern about a cystic tumor was eliminated.

A

2463

FIG 67-101  Posterior approach to cystic mass, which proved to be a pseudocyst. Although a posterior approach is uncommon, if one has developed sufficient skills and all of the retroperitoneal structures are well seen, a careful approach even between the vena cava and aorta may be used.

C

B FIG 67-102  Mucinous adenocarcinoma. A, Large mass in the head of pancreas (arrow) has solid and cystic components. With contrast material, enhancement of the splenic vein, superior mesenteric artery, and renal vein is noted. No collateral vessels are seen anterior through the potential trajectory. B, CT scan shows placement of a small cutting needle with 5-mm gap in the solid component of the mass. C, FNA placement in the cystic component yields clear mucinous material, which was sent for amylase, CEA, and CA 19-9 analysis to attempt to characterize for cystadenocarcinoma.

2464

PART III  Image-Guided Procedures

Many authors reporting small series did not come to any definite conclusion or recommendation,21,236 but several reports of larger series do not support the use of PET for pancreatic cancer. Lytras and colleagues reported on a series of 112 patients and stated: “PET had a similar accuracy to that of CT for imaging pancreatic cancer but it did not provide any additional information in patients with equivocal CT findings and currently would seem of little benefit for the staging of pancreatic cancer.”321 Sendler and coworkers concluded that PET imaging provides only fair diagnostic accuracy (69%) for characterizing enlarged pancreatic masses.442 PET does not allow exclusion of malignant tumors. In doubtful cases, the method must be combined with other imaging modalities. Sperti and associates reported an uncommon use of PET that has not yet been validated by other studies.470 In a group of 50 patients, these investigators reported that PET was more accurate than CT imaging for evaluation for cystic masses (94% compared with 64%).

Contraindications Entities causing bleeding include severe coagulopathy, large aneurysms or pseudoaneurysms (Fig. 67-103), and large varices, which may be associated with the secondary effects of local tumor invasion. In the individual case, one may or may not choose to perform pancreatic procedures because at least local hemorrhage is likely if a varix is penetrated (Fig. 67-104). As noted in the cases discussed here, although such problems may not be evident on prior diagnostic scans, their true nature can be shown by a dynamic bolus scan. Endocrine-secreting tumors are a second important category for exclusion; this includes patients whose history indicates that they are likely to have an endocrine-secreting tumor, such as insulinomas or other islet cell or neuroendocrine-secreting tumors. Although there are no reports about such complications, it is well known that endocrinesecreting adrenal tumors are prone to catastrophic crises from bleeding or hormone release. It is only rational to assume that pancreatic endocrine tumors have a similar risk. Also, some nonfunctioning islet cell tumors and islet cell tumors may be very vascular, and they should not be biopsied if they appear so (see Chapter 46).

17

0

A

17

A

0

B

General Technique for Pancreas Biopsy

Preprocedure History, Review of Images, Planning, and Informed Consent.  Patient history and diagnostic scans are reviewed and informed consent is obtained as noted in the earlier general section. Review of patient history should include weight loss, thrombophlebitis (occurs with pancreatic adenocarcinoma), jaundice, dark urine, abdominal pain, bleeding issues, and current medications. Other pertinent inquiries related to possible islet cell tumors include hypoglycemia (insulinoma), ulcers (gastrinomas), diarrhea (vasoactive intestinal polypeptide-secreting tumors), or periarticular symptoms (acinar cell). Standard laboratory tests should include CBC, INR, and amylase; test results should be normal, as discussed earlier. Pancreatic biopsies can be performed as outpatient procedures unless the patient seems frail and has other significant issues such as vomiting or pain control. Depending on the outcome of the procedure, the patient may be sent home or admitted as needed. Review of prior scans helps plan the procedure relative to possible approaches and patient positions; it saves time to pre-position the patient if a position other than supine is to be used. Any enhanced CT scan may show vascularity and other important anatomic features. Special note should be given to collateral veins, which might occur with splenic vein occlusion, associated pseudocysts, or dilated pancreatic ducts; avoidance of these structures is appropriate. PET scan has not been helpful for the diagnosis of pancreatic cancer, as noted previously. After all of the steps are accomplished, the final discussion with the

S A

C FIG 67-103  Images show the importance of performing a dynamic bolus injection study for every biopsy of the pancreas. A, Image shows what appears to be a large mass in the body of the pancreas, with no indication of any prior history. Biopsy was planned. There was a prior history of contrast allergy, so no contrast study was available. B, At the time the patient was on the table, a bolus injection was performed. Note the very large enhancing pseudoaneurysm (A). C, Angiogram was performed in the course of embolizing the aneurysm (A). Note that the origin is from the splenic artery (S).

CHAPTER 67  Image-Guided Aspirations and Biopsies

2465

considered the gold standard. With the increasing expertise, modern US refinements, and increased demand on CT scanning time, biopsy of large masses can be performed with US guidance. Cases that are difficult to visualize or have collateral veins, dilated loops of bowel, or large body size are well suited to CT. Lesions in precarious locations, such as adjacent to the portal vein, gallbladder, or splenic hilum, are triaged for CT guidance.

Procedure Steps.  During the procedure, vital signs, neurologic

A

B FIG 67-104  A, Enhanced CT scan before biopsy shows many enhancing vessels representing varices anterior to the pancreatic mass (arrows). Such varices are usually formed secondary to splenic vein thrombosis. B, Despite these vessels, a junior faculty member attempted a biopsy of the mass. Scan taken during the biopsy shows hematoma forming around the varices adjacent to the mass (arrows).

patient should be performed and informed consent should be obtained, according to the local standards.

Imaging Preferences.  There are two modalities from which to choose: endoscopic ultrasound (EUS) FNA or percutaneous imageguided procedures. The choice of EUS versus image guidance is a complicated issue that can be addressed only on a local level; expertise varies among institutions. If one were to accept the literature reported in the gastroenterologic journals, EUS is the method of choice. Percutaneous methods are effective, however, and it is likely that choices will be made on a local level without a national consensus (see “Diagnostic Accuracy,” later). For percutaneous procedures, a choice between US, MRI, or CT must be made by the individual physician, but generally CT is still

status, and pulse oximetry should be monitored. Analgesics and sedatives can be administered as needed (see “General Procedure Steps,” earlier). Target and trajectory determination.  A diagnostic series of scans is taken before the procedure. From the current scans, the mass referred for biopsy is reevaluated. A dynamic bolus contrast scan over the chosen target site should be performed to better visualize the mass with its contours and its relationship to the normal glandular tissue. This scan also visualizes any significant sized vessels and excludes excessive vascularity of the tumor if it is not an adenocarcinoma, as noted earlier under “Contraindications.” Because the tumor mass is typically avascular compared with the normal gland, the margins of the normal gland can be seen. The appropriate pathway to the target is the one that avoids the normal gland and other structures if possible. Although it is unnecessary in an immunocompetent patient, we attempt to avoid loops of bowel if possible by using an angulation of the needle or changing the position of the patient to permit the bowel to shift, clearing the trajectory path (Fig. 67-105). A posterior approach (see Fig. 67-101) is suitable,  but one should use this approach only if uninvolved structures are avoided. We do not agree with other authors who have used posterior approaches and penetrated structures such as the vena cava; those trajectory choices have a higher associated complication rate. Rarely we have traversed liver for a pancreatic head biopsy (preferable to crossing the cava posteriorly), but this should be done only with a small-caliber needle and with avoidance of any visible vessels in the liver. If bowel must be crossed, we prefer the stomach (see Fig. 67-98) over the small bowel, but the bacterial contents of these loops seldom cause a problem. Penetration of the small bowel seldom occurs, because it “slithers” out of the way when the needle comes close to it. Colon should be avoided because of the many pathogens present. It is best to avoid the dilated biliary system or pancreatic duct. Penetration of these structures may produce peritoneal soilage and chemical peritonitis. Instrument choice and techniques.  The choice of needles is based on the clinical history of malignancy, accessibility of the lesion, vascularity of the lesion, and type of tissue sample required. The choices are as noted earlier in “Indications for and Selection of FNA and Cutting Needles”: FNA for cytology or a cutting needle for larger amounts of tissue (accuracy of tissue recovery and steps to improve outcome are discussed subsequently). A coaxial system should be used with any needle biopsy. This is because of reported contamination of the peritoneum with tumor cells after a percutaneous biopsy. In most cases, FNA is appropriate because if a cutting needle is used there is the possibility of some associated pancreatitis (if a portion of the normal gland is injured). With FNA, we use a 20-gauge needle and take two passes. If a cutting needle is being used, we use either a 20-gauge or 18-gauge needle, but only an 18-gauge needle if the mass is large and easily accessible. As noted numerous times earlier, if a cutting needle is to be used, a dynamic bolus scan is critical; the peripancreatic vessels are nearby and are in jeopardy if they are not visualized with contrast material.

2466

PART III  Image-Guided Procedures

A

B

FIG 67-105  Avoidance of the colon is important if possible because of the high bacterial count in the colon. A, The approach used was to insert the needle from a lower origin to traverse below the transverse colon. At this level the needle is pushing by the lower edge of the colon and is directed cephalad (arrow). B, Slightly higher scan shows the needle entering the pancreatic mass. Such attempts are worthwhile, and occasionally one may penetrate the bowel. At least over a long experience, the probability of a complication is less likely the fewer penetrations are made.

Choice and palpation of the entrance site.  When the tentative trajectory and entrance site are chosen, a marker is placed on the skin and a repeat scan is performed to see the marker relative to the pathway. The marker may be a purchased device or a hypodermic needle in the package. After one has selected a tentative entrance point and before needle insertion, one should always look for subcutaneous vascular collaterals. These can occur if portal hypertension develops secondary to tumor occlusion of splenic or portal veins. Site preparation.  Careful antiseptic preparation of the skin is important. Proper administration of lidocaine to produce a “painless” procedure, as noted earlier, is important. When the area is anesthetized, it is important to infiltrate anesthetic through the subcutaneous tissues and down to and including the surface of the pancreas. Adequate anesthesia makes the procedure virtually pain free and permits insertion of the needle without the patient moving because of pain. Needle selection and insertion.  The appropriate needle, either the aspiration needle or the cutting needle, is selected using the factors discussed previously. The needle is inserted with incremental steps and then adjusted until it is placed in the target site. The patient is asked to hold his or her breath during each movement of the needle. If the lesion and needle are in the same axis, the procedure is usually straightforward, but if not, the angled methods noted earlier are appropriate. With the aspiration needle, localization of the tip precisely in the chosen site is important if adequate cellular material is to be obtained. Care should be taken to avoid trauma to the normal gland because it has been shown that inadvertent injury to the glandular tissue can cause severe pancreatitis and even death (see “Complications,” later). If a cutting needle is being used, one should adjust the position of the needle and stylet so that the chosen area and not normal glandular tissue is included in the gap (see Figs. 67-99 and 67-102). If the stylet is not positioned properly, one can try the maneuvers noted earlier to position the stylet better; intentional deflection, coaxial manipulation, or rotation of the needle may be performed (see Figs. 67-44 through 67-47). During needle insertion and sample acquisition, whenever  the needle is moved, the patient should be asked to suspend respiration. Not doing so may cause shearing of the capsule and cause a complication.

Sample acquisition.  The sample should be acquired as noted earlier under “General Technique for Pancreas Biopsy.” We routinely take two FNA samples (see “Factors Affecting Accuracy,” later) and one or several cutting needle samples depending on the quality of the sample as noted on visual inspection or if a special sample has been requested for flow cytometry, DNA analysis, or microchemistries. Cutting needles larger than 20 gauge should not be used; a 20-gauge needle is used routinely, but if a mass is large and a lymphoma is suspected, an 18-gauge needle is appropriate if the anatomy is suitable (see Fig. 67-99). If a mucinous cystadenoma is suspected or is being excluded, multiple samples are taken as follows. The solid portion of the mass is sampled with either an aspiration needle or a small cutting needle (see Fig. 67-102). A second sample of the fluid component is taken and sent for CEA and CA 19-9; authors have reported that these markers can consistently differentiate benign cystadenomas from cystadenocarcinomas (see “Diagnostic Accuracy”). The coaxial system is left in place until the multiple samples are obtained. If the site seems to be bleeding, we suggest using the coaxial/ coil/thrombin hemostatic method described earlier. If the local injection of blood products is being used, another aliquot of FFP or platelets should be injected into the site as the coaxial system is withdrawn. The coaxial cannula is withdrawn and a repeat scan performed to assess for bleeding.

Postprocedure Care.  Before removing the patient from the table, we take a repeat scan to evaluate for postprocedure problems. If any hematoma is noted, the clinical team is contacted, and a type and crossmatch of blood products may be ordered if there is a concern the bleeding is significant. The postprocedure care is simple and consists of observing the patient for abdominal discomfort and monitoring vital signs for possible bleeding. The patient is kept on bed rest for 4 hours. Vital signs are checked every 15 minutes for 1 hour, every 30 minutes for the next hour, and every hour for the next 2 hours. After 4 hours, the outpatient can be discharged home with appropriate instructions. We do not routinely order an amylase level unless the patient develops abdominal discomfort after the procedure.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2467

Diagnostic Accuracy

Complications

Diagnostic accuracy is a direct result of obtaining an adequate and appropriate sample for cytohistologic and chemical analysis. For solid tumors, the FNA can recover diagnostic cells to determine whether a tumor is malignant and the specific cell type. If a more definitive diagnosis is desired, a small-caliber cutting needle can be used, which diagnoses pancreatic cell types and lymphoma; greater caution should be used with the cutting needle, as has been emphasized. For cystic masses, recovery of fluid for cytologic and chemical analysis is important for characterization of the abnormality. In the earliest report, Dickey and associates showed the impact of FNA needle caliber.102 They noted that with a single pass of a needle, the diagnostic recovery was better with a 20-gauge FNA needle than with a 22-gauge FNA needle. In a randomized prospective singleinstitution study, Horwhat and colleagues reported that EUS FNA is numerically (although not quite statistically) superior to CT or US FNA for the diagnosis of pancreatic malignancy.232 Erturk and coworkers reported no difference in the outcome of EUS-guided or CT-guided procedures.116 From these data, we conclude that FNA biopsy guided by imaging likely performs equivalently to EUS and US, with any difference reported being secondary to individual expertise, institutional bias, and variable patient selection. Information also is available for nonadenocarcinomatous tumors. The older literature reports that FNA can be used to diagnose the less common types of tumors. Iselin and colleagues reported a small series in which cytology distinguished mucinous and serous tumors.237 Gupta and alAnsari reported success in diagnosing four cases of mucinous cystadenoma of the pancreas.173 Acinar tumors diagnosed by aspiration were reported by Samuel and Frierson.431 Suarez and coworkers479 and Mandrekar and colleagues325 diagnosed papillary tumor. Islet cell tumors can be diagnosed by cytology according to Collins and Cramer77 and Jhala and associates.242 Data are insufficient to determine whether any of the modalities is a better method, but it is likely that samples from CT, US, or EUS would be suitable. Numerous authors have reported on an approach for diagnosis of mucinous adenocarcinoma of the pancreas and differentiation from pseudocysts that seems to provide improved results for differentiating premalignant and malignant pancreatic cystic tumors from seroustype tumors.295,304,392,471,518 These authors concluded that chemical analysis of the fluid aspirated from cystic tumors, either percutaneously or by EUS, can diagnose malignant forms. CEA has been found to be present in the macrocystic form of the mucinous tumor, which is thought to be premalignant or malignant. Wu and colleagues noted that CEA and CA 19-9 permit early diagnosis of cystadenocarcinoma.518 According to O’Toole and colleagues392 and van der Waaij and colleagues,496 pseudocysts have much higher amylase levels in aspirated fluid.

The incidence of complications varies among the various reports in the literature. Brandt and coworkers reported a major complication rate of 1.1%.44 Mueller and colleagues reported the highest complication rate for FNA of the pancreas, which was 3%.365 The mild complications reported by Smith,458 Goldstein and associates,159 and McLoughlin and colleagues342 included transient rectal or colonic bleeding, probably because of trauma to the transverse colon. Other mild complications included four cases of mild pancreatitis that resolved.118 Vasovagal reaction has been reported in two cases by Elvin and colleagues, who used a cutting needle with any pancreatitis.113 Such reactions should be treated as noted earlier under “Vasovagal Reaction.”

Factors Affecting Accuracy.  Factors favoring improved accuracy are, as one might expect, straightforward. Proper use of the needle technique as described earlier under the heading “General Technique for Pancreas Biopsy” are applicable. Careful targeting of the tumor mass and meticulous technique with FNA or cutting needle are important. High-quality enhancement of the pancreas most clearly delineates the margins of the tumor in the pancreas and porta hepatis. Shin and associates reported a series on EUS and CT and found that the results were good with either method, and results were not influenced by lesion characteristics, aspiration site, and tumor type.448 For biopsy of pancreas transplant, the literature indicates that CT or US is useful.8,19 Because this is such an esoteric indication, the reader is referred to the literature.

Serious Complications.  Serious complications have been related to tumor sampling (pseudocysts are reported later) and have included hemorrhage, fatal pancreatitis, sepsis, cholangitis, fistula, and needle track seeding. Fatal pancreatitis has been the most common. Evans and colleagues118 and Levin and Bret293 reported cases of fatal pancreatitis after US-guided biopsies. In the study by Mueller and colleagues of 184 biopsies, 3% (5 patients) developed severe pancreatitis.365 One patient ultimately died of sepsis. The patients with severe pancreatitis were different from most patients. Mueller and colleagues noted that 60% had passage of the needle through bowel compared with 21% of the total.365 The masses in the patients were 3 cm or smaller. Four of five patients had 20-gauge needle biopsy. Although the authors do not attribute the complication to the number of needle passes and do not mention the number of needle passes, the four passes of the needles used by these authors are more than the number of needle passes used by most groups. Finally, three of the five cases with severe pancreatitis had normal tissue and not tumor. Serious bleeding also has been a serious complication associated with a posterior approach.122,461 Fataar and associates reported two cases of hemorrhage; one required CPR and 2 U of blood, and one resulted in a 6.5-cm hematoma in the mesocolon.122 Sofocleous and coworkers reported bleeding in 14% of patients with biopsies using a posterior approach in a series of 56 patients.461 No transfusions were required. Four inadvertent punctures of the right renal artery occurred; there was no long-term follow-up for these patients reported to date. Tumor recurrence or seeding has been noted. Rashleigh-Belcher and colleagues414 and Ferrucci and associates125 reported separate cases of needle track seeding from pancreatic carcinoma. The occurrence of peritoneal carcinomatosis was reported as higher for CT procedures by Micames and colleagues.349 One patient in the EUS FNA group developed peritoneal carcinomatosis, compared with seven patients in the percutaneous FNA group (2.2% versus 16.3%; P < .025). This is in contradistinction to the data of Johnson and associates, who found no statistical difference in the presence of tumor cells in peritoneal washings in patients with and without percutaneous biopsy.245 Phillips and colleagues reported one case of cholangitis and pancreatic fistula and severe vasovagal reaction.406 Ferrucci and associates had a case of cystadenoma of the pancreas that resulted in generalized sepsis.125

Factors Affecting Complications.  Careful targeting of the tumor site and avoiding traumatization of the normal glandular tissue are important. In contrast to other organs, the pancreas is the one organ in which the traumatized normal gland can develop severe inflammation, which may be difficult to manage and can result in death. In a series by Mueller and colleagues, three of five cases of pancreatitis occurred when normal gland was inappropriately sampled and trauma was induced.365 FNA needles should be carefully targeted into the

2468

PART III  Image-Guided Procedures

tumor, avoiding trauma to the normal gland. Cutting needles are used only when an approach permits insertion of the needle clearly into the suspected tumor and the “cutting gap” is limited to the tumor; if a portion of the normal gland is cut, pancreatitis can ensue. In practical terms, one avoids the most expeditious pathways that cross other structures, but rather seeks alternative pathways that might avoid vascular structures, intestinal loop, bile duct, pancreatic duct, or other uninvolved structures. Limiting the number of needle passes and careful targeting of abnormalities prevent damage to normal gland. We limit our needle passes with FNA to two, whereas others have used as many as five.238,365 One should avoid the posterior approach for pancreas biopsy unless extreme care is taken to avoid vessels and uninvolved structures (see Fig. 67-104). Although Fataar and coworkers,122 Sofocleous and associates,461 and Gupta and coworkers175 have reported intentional passage of the biopsy needle through the vena cava, we believe this is not a rational approach. In all such circumstances, such structures should be avoided, even with the injection of saline or gas, in attempting an unusual approach. Although complications are more likely to occur with such approaches, one can take consolation that one can “get lucky” and a complication may not happen.

such a biopsy is inappropriate. Biopsy of nonfunctional tumors is not an issue relative to hormone release, but nonfunctional tumors can be hypervascular on dynamic bolus scan. If such a lesion were too vascular, we would defer such a biopsy. It is not unusual for a nonspecificappearing mass to be biopsied without knowledge of its endocrine origin, with its true identity being revealed by the cytohistologic report; because the tumor is nonfunctional, there are no adverse consequences. Chapter 46 provides a complete review of nonfunctional endocrine tumors. The definition of “functioning” is made on a clinical basis. During the patient interview and informed consent process, inquiry into clinical symptoms that might suggest one of these clinical syndromes is appropriate. Briefly, insulinomas from beta cells produce the Whipple triad of (1) spontaneous hypoglycemia with central nervous system and vasomotor symptoms, (2) blood glucose less than 50 mg/mL,  and (3) relief of symptoms with administration of glucose. Gastrinomas produce the classic Zollinger-Ellison syndrome with hyperacidity and ulcer disease. Glucagonomas are associated with a rash, diabetes mellitus, and weight loss. Vasoactive intestinal polypeptide-secreting tumors and somatostatinomas are very rare and do not have a clearly defined syndrome, although both may be associated with diarrhea.

Implications for Surgery

Cystic Tumors.  The indications for sampling cystic tumors are dif-

Surgeons have become more aggressive in approaching the resection of pancreatic cancer, because new imaging methods can better define anatomy and screen patients for surgery. Several recurring questions can be answered by reviewing certain reports. First, the accuracy of percutaneous image-guided procedures was shown by Savarino and colleagues to be equivalent to biopsies performed during laparotomy.437 Another common question raised is whether percutaneous biopsy “seeds” the peritoneum and affects the ultimate outcome. Johnson and associates confirmed two important principles.245 These authors studied the presence of tumor cells in the peritoneum at the time of laparotomy in two groups of patients. They found viable tumor cells in both groups, biopsied and nonbiopsied, with no statistical difference between the two groups. Second, they found that the ultimate survival was no different between the groups. Balen and coworkers also confirmed that biopsy produced no adverse effect on the survival of patients.20 Micames and associates reported that the occurrence of peritoneal carcinomatosis was higher for patients having CT-guided procedures.349 One patient in the EUS FNA group had developed peritoneal carcinomatosis, compared with seven in the percutaneous FNA group (2.2% versus 16.3%; P < .025). As one can tell from the conflicting data in the surgical literature about percutaneous biopsies, there are varying opinions about the role of such procedures. Local expertise and experience are likely to dictate the role of such procedures at individual institutions.

ferent than the indications for other problems. With cystic masses, there are two objectives. The first one is to differentiate cystic tumors from pseudocysts of the pancreas. By imaging alone, it is virtually impossible to distinguish between a pseudocyst (see Fig. 67-101) and a cystic tumor (see Fig. 67-102). In such cases, the goal is to aspirate fluid from a fluid-density mass so that the fluid can be sent for amylase content and biomarkers CEA and CA 19-9, which are characteristic of cystic tumors. Surgery typically is the definitive treatment for such cases.295,304,392,471,518 As noted in Chapter 46, cystic tumors of the pancreas are unusual entities that include a benign and a premalignant variety. Correlation with demographics can be helpful because of the greater incidence in different age groups. Differentiation of these tumors by imaging and histocytologic methods has been reported occasionally. In the early literature, several authors173,237 reported success in differentiating the mucinous/serous cystic tumors and diagnosing cystadenocarcinoma, but this experience has not been confirmed by other authors. In recent years, numerous reports have confirmed that assay of aspirated fluid or protein for the presence of CEA and CA 19-9 can be diagnostic (see earlier under “Diagnostic Accuracy”).

Specific Tumors

Acinar Cell Carcinoma.  Acinar cell carcinoma is a rare tumor, so it is doubtful there will ever be accumulation of a significant number of cases. As predicted, it has now been documented that diagnosis can be made by aspiration biopsy; several cases were reported by Samuel and Frierson.431

Endocrine Tumors—Functional and Nonfunctional.  Endocrine tumors of the pancreas originate from the islet cells and can be functional or nonfunctional. We do not recommend biopsy of endocrine tumors that are functional because of concern about a crisis caused by rapid release of the involved hormone product. Although there is a case report of a biopsy of a pancreatic insulinoma, we believe that planning

Pseudocyst Aspiration Aspiration of pseudocyst fluid or inflammatory fluid is important for the exclusion of cystic tumors (as noted earlier) or the diagnosis of an abscess. Laboratory analysis of fluid from a pseudocyst shows a high amylase content (see Figs. 67-100 and 67-101). Aspiration of any suspicious fluid collection in patients with pancreatitis should be performed to exclude pyogenic infection. When such procedures are being performed, one should ensure that bowel loops are not being traversed to exclude the possibility of contaminating a fluid collection or recovering spurious culture results. Good evidence for this cautionary approach was provided by Smith, who collected 16 cases of generalized sepsis from a physician survey.459

CAVERNOUS TRANSFORMATION OF PORTAL VEIN In some cases, cavernous transformation of the portal vein may produce a mass that appears avascular, which would be recommended for a biopsy. In usual cases when the vascular channels are large, the

CHAPTER 67  Image-Guided Aspirations and Biopsies

v a

A

2469

of solid renal tumors, follow-up in the form of observation over time, percutaneous biopsy, or surgical investigation or treatment is required. The sophistication of image-guided interventions now also provides treatment for hydronephrosis, abscesses, and even solid tumors. With the recognized benefit of less invasive treatment of tumors by percutaneous methods, definitive diagnosis of incidental masses has become a topic of great interest. Although the standard of care espoused by urologists is to treat any solid mass as a cancer, newer data and approaches advocate definitive diagnosis by biopsy before treatment. Three types of CT biopsies are performed on the kidney: (1) biopsy of indeterminate partially cystic masses, (2) biopsy of solid masses, and (3) biopsy of diffuse parenchymal disease.

Indications

B FIG 67-106  A, Mass in the head of the pancreas shows increased enhancement (v) with contrast material, a finding consistent with a cavernous transformation of the portal vein. Note aortic (a) enhancement. B, Enhancing mass in the head of the pancreas, representing AVM of the head (arrows).

portal cavernomas can be recognized with a dynamic bolus scan (Fig. 67-106). If a cavernous transformation of the portal vein is partially thrombosed and has small capillaries without significant vascular channels providing blood flow, such masses may not enhance with a dynamic bolus scan (Fig. 67-107). In such cases, if a biopsy is inadvertently performed in a patient, a complication is unlikely because the vascular channels are very small. Unaware of their pathologic identity, I have biopsied three such cases without difficulty. Careful correlation with clinical history and scrutiny of sequential diagnostic scans can be diagnostic (see Fig. 67-107).

KIDNEY Imaged-Guided Background Image-guided procedures on the kidney have been performed for many years, but their role and purpose have changed. In the early years of intervention before accurate special imaging, percutaneous puncture and aspiration of cysts was common and was performed after urograms, with fluoroscopic guidance. With the refinement of US and CT, the diagnosis of simple cysts is reliable, and a cyst aspiration or biopsy is seldom required. Biopsy of solid masses was performed incidentally when the mass was determined not to be cystic. Renal mass biopsy of solid masses has taken on new importance with the ubiquitous availability and use of imaging. Masses are commonly discovered incidentally, and if they are suspicious or suggestive

Virtually all authors agree that there is no indication for aspiration or biopsy of a simple renal cyst that is unequivocal on CT, MRI, or US. The reader is referred to Chapter 54 for specific criteria for simple cysts, but a summary is provided here. Indeterminate cystic masses with high density, thickened or nodular walls, calcifications, or enhancing areas should be considered for diagnostic FNA or core tissue biopsy. In most cases, only an aspiration needle is indicated. In some cases, when the cytologic sample does not provide adequate histologic data (assuming the dynamic bolus scan shows the mass to be avascular), a cutting needle should be used. Indeterminate masses were studied by Harisinghani and colleagues, who proved the importance of biopsying Bosniak III masses.205 In their series, they precluded the need for surgical treatment in 39% of patients. Solid renal masses should be considered malignant until a biopsy proves otherwise. We differ from the standard urologic approach because biopsies are easy and reliable to perform. In our opinion, it is not logical to perform a therapeutic procedure that has some potential complications or side effects without a diagnosis. New data from several authors are mentioned in support of this position. With the wide accessibility of scanners, more patients are having scans and such lesions are detected. The resolution of such incidental findings without incurring the occasional serious complications of “immediate” surgery on such lesions is to biopsy them (Fig. 67-108). The remarkably accurate MDCT scanners commonly detect additional abnormalities during the workup of a primary tumor (Fig. 67-109). In such cases, one might assume that two lesions indicate metastatic spread or two malignancies, but only a diagnostic biopsy can determine the true state of affairs. We have encountered numerous patients whose management was significantly affected when the nature of unusual inflammatory lesions was confirmed by biopsy (Fig. 67-110). Solid masses may be unusual inflammatory lesions, such as granulomatous or fungal masses (see Figs. 67-109 and 67-110); nephrectomy or ablation in such cases is not rational unless there are extraordinary circumstances. The accuracy of such biopsies and the need for preablation diagnosis was emphasized by Tuncali and coworkers494 and Beland and associates.30 Angiomyolipomas are the exception among solid masses and do not require biopsy, because the presence of fat is virtually pathognomonic of this hamartoma. The only possible differential diagnosis would be a liposarcoma, which is so rare it is not worth considering. The second important reason not to biopsy such masses is that they have many pseudoaneurysms and increased vascularity, which are prone to bleeding should a biopsy be performed. Biopsying such a mass might put the kidney in jeopardy.

General Technique for Renal Biopsy

Preprocedure History, Review of Images, Planning, and Informed Consent.  Patient history, laboratory studies, and diagnostic

2470

PART III  Image-Guided Procedures

172.0 mm

A

B

C

D FIG 67-107  A, Biopsy of cavernous vein was performed inadvertently by a junior faculty member. A bolus CT scan was performed to assess vascularity in a patient with a large tumor mass, edema around the mass (arrows), and an enlarged spleen. Many collaterals (curved arrows) were noted in the peripancreatic area, so a biopsy through the lower right liver lobe was performed. No enhancement of the mass was noted. B, Biopsy of the mass through the right lobe. Note the cutting gaps of the small core biopsy needle in the mass. Results were not diagnostic of tumor. Clinically the patient was thought to have a lymphoma, but his clinical status improved. Review of the prior US revealed thrombosis of the portal vein causing the edematous pancreas and collateral vessels. C, Repeat CT scan shows a smaller mass (arrows), less edema, a smaller spleen, and fewer collaterals after the portal vein had recanalized. The mass resolved and the patient became asymptomatic, indicating that the findings were the result of portal vein thrombosis and an edematous pancreatic head.

scans are reviewed and informed consent obtained as noted earlier. Review of patient history includes history of hematuria, history of renal disease, prior surgery, history of any other malignancies, unusual infections, and status of the immune system. Also important are bleeding issues and current medications related to cardiac disease, hypertension, and anticoagulation drugs (prescribed or OTC). Laboratory values for CBC and coagulation factors should be suitable, as discussed earlier. The venue of the procedure is decided based on the nature of the biopsy. Renal mass biopsies can be performed as an outpatient procedure unless the patient is frail and has other significant issues. Depending on the outcome of the procedure, the patient may be sent home or admitted. Renal parenchymal biopsy patients should be kept overnight because of the high incidence of bleeding that occurs with parenchymal biopsy (see Table 67-4).

Review of prior scans helps plan the procedure relative to possible approaches and patient positions; it saves time to pre-position the patient if a position other than supine is to be used. For tumors, enhanced CT scan may show tumor vascularity, collateral vessels, and adjacent anatomic features that are important for success or avoidance of complications. Also, the location of the mass may suggest initial plans for patient position and needle trajectory, as discussed subsequently. Visualization of the renal pedicle and segmental arteries, splenic pedicle, gonadal vein, inferior mesenteric vein, and ureter may be of interest depending on the location of the mass. PET scanning has not been as helpful as one might anticipate because many renal tumors are not avid for FDG (see Chapter 53). Informed consent is discussed and obtained with the patient before the procedure is initiated.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2471

local expertise and preference, either US or MRI can be used instead of CT.

Procedure Steps.  During the procedure, vital signs, neurologic

A

B FIG 67-108  A, Elderly woman with an incidental solid mass identified on routine scan. Because a thermal ablation was contemplated, a biopsy was performed. Because of the small size of the mass, the patient was scanned in the left decubitus position, which facilitates immobilization of the mass because of the weight of the abdominal viscera on the dependent diaphragm. This diagnostic scan shows the mass (arrow), which does not display excessive vascularity, permitting a safe core biopsy. B, A small cutting needle localized in mass, which proved to be renal cell carcinoma.

Imaging Preferences.  Choice of modality for guidance is CT, US, or MRI. Because the retroperitoneum has some characteristics that make it difficult to obtain a good window, CT typically is chosen. The ability to accurately see the adjacent anatomy, renal pedicle, and collecting system may be helpful. The use of the dynamic contrast scan is very helpful to assess vascularity and collaterals to plan the possible need for a hemostatic approach, as discussed earlier. Depending on the

status, and pulse oximetry should be monitored. Analgesics and sedatives can be administered as needed (see earlier under “Patient Cooperation and Sedation”). Target and trajectory determination.  A diagnostic series of scans is taken in anticipation of the procedure. From the current scans, the mass referred for biopsy is reevaluated. A dynamic bolus contrast scan over the chosen target site should be performed to better visualize the mass with its contours and its relationship to the normal structures. The collecting system should be avoided. This scan also visualizes any significant sized vessels and excludes excessive vascularity of the tumor if it is not an adenocarcinoma (see Fig. 67-33). Different issues exist with different sizes of lesions. Small lesions are more difficult to target accurately. Larger tumors tend to have increased vascularity but can have areas of necrotic tissue. Targeting the nonnecrotic area provides the most suitable diagnostic sample  (Fig. 67-111). In most cases, the patient is not in the supine position but in the prone, lateral, or oblique position. Other anatomic considerations include the colon and the muscles of the posterior abdomen. If a lateral approach is used, the location of the colon is important so that it can be avoided; if the colon needs to be moved, local injection of gas or fluid might be considered. The patient may be placed in any position to facilitate the approach, including the supine, prone, oblique, or lateral position (see Fig. 67-11). Such movement can move loops of bowel out of the way if needed for an oblique or lateral position. Another benefit of having a patient lie in the lateral decubitus position is that the weight of the viscera reduces the motion of the ipsilateral diaphragm. The decreased movement of the diaphragm assists in immobilizing the kidney with a mass (Fig. 67-112; see Fig. 67-11). This may be crucial for small lesions located anteriorly adjacent to spleen or colon or both. The choice of the pathway should avoid the major muscles posteriorly and traverse the thinnest area of the muscle if possible. Most hematomas in our cases have been from the posterior abdominal wall rather than the organs themselves (see Fig. 67-27). The bleeding may be from muscular vessels or lumbar vessels. For parenchymal biopsies, the cortex of the kidney is targeted to provide adequate tissue for analysis of the glomeruli. Traditionally the lower pole of either the right or left kidney was targeted to ensure a good sample of the cortex without damage to the medullary region with the segmental arteries. Although this site can be used, with CT one can target the tangential cortex at different sites and accomplish the same thing (Fig. 67-113). Use of the coaxial system is desirable to permit hemostasis as needed; biopsies of parenchyma have a high bleeding rate even when done properly; the highly vascular cortex is “arterial” and bleeds readily (see “Complications,” later). Instrument choice and techniques.  The choice of needles for tumors is based on the clinical history of malignancy, accessibility of the lesion, vascularity of the lesion, and type of tissue sample required. The choices are as noted earlier: FNA needle for cytology or a cutting needle for larger amounts of tissue (accuracy of tissue recovery and steps to improve outcome are discussed subsequently). Use of a coaxial system for any needle biopsy is prudent because of the possible vascularity of tumors. The use of a coaxial system permits the use of a hemostatic method if any bleeding occurs. With FNA, we use a 20-gauge needle and take two passes (see Fig. 67-112). If a cutting needle is being used, we use either a 20-gauge or an 18-gauge needle, but only the 18-gauge needle if the mass is

2472

PART III  Image-Guided Procedures

A

B FIG 67-109  A, A patient with a lung mass proven to be lung cancer by biopsy also had a renal mass. The presumption was that it was a metastatic or a second primary lesion, until a biopsy was performed. B, CTguided biopsy of renal mass with an 18-gauge cutting needle yielded granulomatous inflammation (arrow). Because of the reluctance of clinical services to accept the diagnosis, three separate biopsies with an 18-gauge core biopsy needle were performed; the histologic result was identical on every biopsy. The patient refused to have a nephrectomy, contrary to the clinical advice, and went to another facility for lung surgery without a nephrectomy.

FIG 67-110  A similar case occurred in a patient who was status post

FIG 67-111  Large renal cell tumor with a necrotic center. Core biopsy

liver transplant for a hepatocellular carcinoma. The solid mass (arrow) was assumed to be a metastatic lesion, but at our insistence a biopsy was performed. Biopsy showed aspergilloma of kidney, not renal cancer.

taken of the periphery with viable tissue. The edge of the necrotic tissue (arrow) can be seen centrally.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2473

C

A

B FIG 67-112  A, A patient with a small cystic mass on the left kidney (arrow) was scanned with the left side down. The weight of the abdominal viscera partially impairs movement of the diaphragm, making such masses more immobile. During this initial approach, the needle deflected from the cyst. B, CT scan shows repositioning with needle poised adjacent to the mass. The bevel was turned downward to deflect the needle up. C, Puncture of cyst was successful, and clear fluid was extracted.

sufficiently large and accessible (see Fig. 67-111). As noted numerous times earlier, if a cutting needle is to be used, a dynamic bolus scan is critical before the biopsy to avoid possible bleeding. Choice of the needles for parenchymal biopsies varies according to the local nephrology service and the requirements of pathology. At least an 18-gauge needle is required, with two samples. We prefer a 15-gauge or 14-gauge needle and prefer a single biopsy if possible (see Fig. 67-113). Song and Cronan reported equivalency for glomerular specimens with 18-gauge and 14-gauge needles.466 Mostbeck and colleagues compared 18-gauge and 16-gauge needles and found the 16-gauge needle to be more effective.361 As noted subsequently, we are not biased toward any of these needle sizes and use whichever is suggested by the clinical nephrology service. Choice and palpation of entrance site.  When the tentative trajectory and entrance site is chosen, a marker is placed on the skin and a repeat scan is performed to see the marker relative to the pathway. The marker may be a purchased device or a hypodermic needle in the package. We prefer a lateral entrance site to avoid the spinal erector

muscle, which has a propensity to bleed, is uncomfortable to the patient, and can alter the pathway or bend the needle if a patient becomes uncooperative (see Fig 67-113). Site preparation.  Careful antiseptic preparation of the skin is important. Proper administration of lidocaine to produce a “painless” procedure, as noted earlier, is important. When the area is anesthetized, it is important to infiltrate anesthetic through the subcutaneous tissues and down to and including the capsule of the kidney. Adequate anesthesia of the capsule makes the procedure virtually pain free and permits insertion of the needle without the patient’s moving because of pain. If a rib is inadvertently hit during the initial anesthetic administration, I give a small bolus of lidocaine on the rib; this prevents any recurrence of pain if that area is struck again during the procedure. Needle insertion.  The appropriate needle, either FNA needle or cutting needle, is selected using the factors discussed earlier. The needle is inserted with incremental steps and adjusted until it has been placed in the target site. The patient is asked to hold his or her breath during each movement of the needle. If the lesion and needle are in the same

2474

PART III  Image-Guided Procedures

24

is delivered to the slides and fixative. With the coaxial cannula in place, a fluid sample can be withdrawn for cytology and culture if it appears purulent. The coaxial system is left in place until the required samples are obtained. If the site seems to be bleeding, we suggest using the coaxial/ coil/thrombin hemostatic method described earlier under “Coagulopathic Treatment.” If local injection of blood products is being used, another aliquot of FFP or platelets should be injected into the site as the coaxial system is withdrawn. The coaxial cannula is withdrawn and a repeat scan performed to assess for bleeding.

Postprocedure Care.  Before removing the patient from the table, 0 80

FIG 67-113  Because the goal of parenchymal biopsy is to recover glomeruli and not produce bleeding, one tries to sample the most peripheral cortex. CT scan shows the gap of the needle in the most peripheral cortex for the sample. The central area should be avoided to prevent inadvertent injury to segmental arteries.

we take a repeat scan to evaluate for postprocedure problems. If any hematoma is noted, the clinical team is contacted, and a type and crossmatch of blood products may be ordered if there is a concern that the bleeding is significant. Postprocedure care is simple and consists of observing patients for flank discomfort and monitoring vital signs for possible bleeding. The patient is kept on bed rest for 4 hours. Vital signs are checked every 15 minutes for 1 hour, every 30 minutes the next hour, and every hour for the next 2 hours. After 4 hours the outpatient can be sent home with appropriate instructions.

Diagnostic Accuracy axis, the procedure is usually straightforward, but if not, the angled methods noted earlier are appropriate (see Fig. 67-49). With the FNA needle, localization of the tip precisely in the chosen site is important to obtain adequate cellular material. Care should be taken to avoid trauma to the normal gland; it has been shown that inadvertent injury to the glandular tissue can cause severe pancreatitis and even death (see “Complications,” later). If a cutting needle is being used, one should adjust the position of the needle and stylet so that the chosen area is included in the gap (see Figs. 67-109 and 67-111). If the stylet is not positioned properly, one can try the maneuvers noted earlier to better position the stylet; intentional deflection, coaxial manipulation, or rotation of the needle may be performed (see Figs. 67-44 through 67-47). During needle insertion and sample acquisition, whenever the needle is moved, the patient should be asked to suspend respiration. Not doing so may cause shearing of the capsule and cause a complication. Several technical points may be helpful for small masses. First, when approaching a small mass, the procedure should be performed in two steps. The first step is to position the needle carefully immediately adjacent to the area being sampled. Several initial approaches may be needed in this way because the kidney is mobile, and with variable respirations a small mass may move in and out of the slice. After the needle position has been confirmed adjacent to the mass, the needle is “popped” into the mass. If the needle is pushed slowly toward the mass, it may deflect because the bevel of the needle may slide over its curved surface. Sample acquisition.  The sample should be acquired as noted in the general section on technique. We routinely take two FNA samples and one or several cutting needle samples depending on the quality of the sample as noted on visual inspection. If a special sample has been requested for flow cytometry, DNA analysis, or microchemistries, larger needles (18 gauge through 14 gauge) may be used, and as many samples taken as needed (see Tables 67-3 and 67-4). If the mass proves to be cystic and a fluid sample is obtained, one should attempt to get two samples—one for the inner wall of the cyst and a fluid sample. With a coaxial system in place, this is straightforward. The needle is advanced to the opposite side of the wall, and a reciprocating motion is used to get cells for the inner wall. The sample

In our unpublished study of needle biopsy of renal masses (J. Isariyanwongse and L. Ponsky: personal communication, 1967), we confirmed the favorable qualities of the FNA and core biopsy procedures, reconfirming the improved diagnostic outcome from cutting needles. Sixty patients had a renal biopsy with CT guidance for an enhancing renal mass; 49 (82%) had a positive biopsy for malignancy, and 11 (18%) biopsies were negative for malignancy. Of the positive biopsies, 41 (84%) tumors were primary renal malignancies, and 8 (16%) were metastatic or lymphomatous infiltration of the kidney. Ten patients ultimately underwent a radical or partial nephrectomy, and all surgical specimens showed concordance with initial pathology on biopsy. Seven of 19 patients (37%) with a known extrarenal primary neoplasm had biopsies positive for metastatic disease, prompting nonsurgical management in each case. The sensitivity of the percutaneous renal biopsy was 91%, and overall accuracy of the procedure was 92%. Although some authors, such as Heilbrun and colleagues, recommend FNA aspiration as the first appropriate step in the algorithm for diagnosis before ablation,214 we do not agree. Heilbrun’s group reported a 15% nondiagnostic rate for FNA procedures of the kidney.214 Tuncali and associates emphasized the importance of definitive diagnosis of preablative treatment, and their results confirm the limitations of smaller-needle biopsies.494 Tuncali’s group of 27 patients had 10 (37%) with a benign diagnosis.494 This group also contained 10 patients with the diagnosis of cancer by biopsy and another 7 that were not definitively diagnosed before treatment. Use of a cutting needle is warranted at the onset for expeditious diagnosis and management to prevent a delay in treatment. In our experience, it is difficult to convince the urologist even to perform a biopsy before an ablation, and the urologist is unlikely to accept the incomplete results of FNA in a high percentage of cases. Cutting biopsies are more definitive for malignant and benign lesions, and complication rates are sufficiently low. Additional preventive steps can be taken to reduce the risk potential of bleeding by selection of a smallercaliber needle or use of a hemostatic method. In addition to our data, the merits of core biopsies can be appreciated by reviewing the results of Silverman,450 Maturen,333 and Beland30 and their colleagues. Maturen and colleagues reported a favorable

CHAPTER 67  Image-Guided Aspirations and Biopsies success rate in a series of 152 renal mass biopsies performed using an 18-gauge core coaxial system.333 Malignant tumors were found in 56%; benign findings occurred in 40%. There was no tumor seeding. The sensitivity was 97% and the specificity 100%. In 62% there was a great impact on patient management. Beland and colleagues reported a 90% diagnostic yield by biopsy for tumor.30 Of the 52 biopsies, 69% were renal cell carcinomas and 27% were benign lesions. There was one case of lymphoma and one metastatic lesion. There was one false-negative result for renal cell cancer. Harisinghani and coworkers studied Bosniak III indeterminate masses relative to the incidence of cancer and the success rate of biopsy of Bosniak III cystic masses.205 Of the 28 biopsied category III lesions, 17 (60.7%) were malignant (16 renal cell carcinomas and 1 lymphoma) and 11 (39.3%) were benign (6 hemorrhagic cysts, 3 inflammatory cysts, 1 metanephric adenoma, and 1 cystic oncocytoma). The investigators concluded that surgical (also ablative) treatment could be avoided in 39% of patients.

Factors Affecting Accuracy.  Use of a cutting needle rather than an aspiration needle is the greatest factor improving the diagnostic results for solid masses. The preferred size of the cutting needle differs among various authors. For parenchymal biopsies, Song and Cronan compared 18-gauge and 14-gauge needles and found no difference in the suitability of the sample or complication rates.466 This provides an excellent example of how the nature of the target, not the needle caliber, is a prime determinant of the outcome and complications.

Complications Complications for renal tumor biopsy vary among published reports. In our experience of biopsying renal masses, we have encountered numerous complications related to bleeding. One hematoma occurred during the biopsy of a renal mass close to the hilum, suspected to be a focal lymphoma. Because of inadvertent damage to a segmental artery, a hematoma that required neither surgical intervention nor the administration of blood developed. From the major series reported by Maturen and colleagues, there were 3 complications in a series of 152 cases.333 These included two postprocedure hematomas, one requiring blood transfusion. One case of delayed pseudoaneurysm was noted. Beland and colleagues reported no complications, but there was one false-negative result, which was later proven by nephrectomy.30 A potential complication mentioned by clinicians is the possible occurrence of needle track seeding, for which we can find only one reported case in the literature. A carefully controlled study by von Schreeb and associates compared the outcome and complications in two groups of patients with proven renal carcinoma.504 There was no difference in complication rates, morbidity rates, or mortality rates between the groups (and no needle track seeding in 75 cases of carcinoma that were sampled). Reported complications of parenchymal biopsies are greater than for solid renal masses. The presence of the many arteries in the parenchyma account for this increased incidence. The bleeding complication rate from parenchymal biopsies is high, varying in some reports from 5.2% to 8.8%. Sateriale and coworkers reported numerous risk factors related to renal biopsies, including hemodialysis, sex, site of biopsy, and treatment with desmopressin.434 Risk of hemorrhagic complications had no correlation with patient age (when > 16 years old), creatinine level, hematocrit, pathologic features of resultant biopsy specimen, or whether the patient was admitted solely for the biopsy. Arteriovenous fistula formation has been reported after biopsy, and the problem usually can be resolved by percutaneous embolization. deSouza and colleagues reported a series of seven arteriovenous fistulas

2475

occurring after biopsy of renal transplants.99 All patients were successfully embolized without loss of renal function as determined by creatinine level.

Factors Affecting Complications.  The factors affecting complications are similar to other organ sites. The most important factors are related to careful targeting of the abnormality. Minimizing the number of needle passes is important; the more passes, the greater the chance of bleeding. Inadvertently biopsying adjacent normal structures, such as calyces or vessels, causes bad results.

ADRENAL GLAND Image-Guided Background Adrenal biopsy is best accomplished by CT guidance rather than US or MRI because the gland is very posterior, adjacent to many vessels, and adjacent to the undersurface of the diaphragm. With experience, successful biopsy can be consistently accomplished. Either an aspiration needle or cutting needle biopsy can be performed using different approaches, as described. Many authors have reported on CT-guided biopsies of adrenal masses.

Indications The major indication for biopsy is the presence of a mass that is suspected to be either a malignant or an infectious process when determination of the process would change the management of the patient. The major conundrum is to minimize unnecessary biopsy of commonly occurring adrenal adenomas, found incidentally and in oncologic patients. The classic CT findings of a lipid-laden adenoma are small size (<5 cm), well-circumscribed margins, and low-density center. More sophisticated diagnostic approaches are available to confirm the diagnosis (see Chapter 53). Non–lipid-laden adenoma is a more difficult problem; it has been studied extensively using a contrast time washout curve.271 Korobkin and coworkers have stated that if an adenoma would characteristically wash out 50% at 5 minutes and 70% at 15 minutes, tumors would have washout of 8% and 20% at the same time intervals.271 If there are inhomogeneous enhancements, a biopsy is indicated according to Blake and Schwartz37,440 and their colleagues,37,440 who have reported coexisting tumor and adenomatous tissue. More recent literature indicates that PET scanning has assumed a new role in the diagnosis of adrenal masses in oncology patients.274,348 Caoili and associates determined that the combination of the nonenhanced CT scan correlated with the PET scan was the most accurate (Fig. 67-114).55 The standardized uptake value (SUV) of the adrenal mass was 4.2, essentially equivalent to the adjacent liver. This article is of great significance because this group originated the concept of the adrenal washout in the literature, and it implies that they consider this the most accurate study. Finally, controversy has emerged because some authors, such as Park and colleagues, have reported CT washout as more accurate.398 Any atypicality of a non–lipid-laden adrenal mass that is not smooth and homogeneous should be biopsied. When atypical traits are present, making the diagnosis of adenoma less confident, a biopsy is indicated. These principles have been confirmed by numerous authors. In a patient with even a small adrenal mass and a lung cancer, biopsy should be performed (Fig. 67-115; see Fig. 67-114).

Contraindications General contraindications are similar to those in other areas relative to coagulopathies and lack of appropriate anatomy or hypervascularity. Specific contraindications exist related to endocrine function.

2476

PART III  Image-Guided Procedures

d: 39.7 mm mean: –15.7 max: 185.0 d: 39.7 mm mean: 1. max: 3.6

A

C

d: 39.7 mm mean: 1.284 max: 3.693

B FIG 67-114  A, CT scan shows a mass in the adrenal area suggestive of adenoma. B, PET study showed increased uptake, which is common with nonfunctioning adenomas. C, Because a small lung cancer (arrow) is present and has increased uptake, biopsy of an adrenal mass even if suspected as adenoma is mandatory.

A specific contraindication is the presence of an endocrineproducing adrenal mass or a history of hypertension suggestive of such a mass. It is well documented that manipulation of functioning tumors, such as pheochromocytomas, at surgery can produce discharge of hormone, which produces severe physiologic effects. Percutaneous instrumentation can have the same result. Several reports in the literature indicate the great risk associated with biopsying such lesions. McCorkell and Niles reported two cases of FNA of pheochromocytomas with catastrophic results339; both patients had severe hemorrhage, one of whom later died. Casola and coworkers reported four cases of unsuspected pheochromocytoma58; one patient had severe hypertensive instability during the biopsy. As an additional precaution for an unanticipated pheochromocytoma, it is prudent to have an α-blocking medication available in the department for IV use should such a lesion be inadvertently sampled and a hypertensive crisis develop.

General Technique for Adrenal Biopsy

Preprocedure History, Review of Images, Planning, and Informed Consent.  Patient history and diagnostic scans are reviewed

and informed consent obtained as noted earlier. Review of patient history includes hypertension, adrenal history, bleeding propensity, pulmonary disease (risk of pneumothorax), and malignancies that might metastasize to the gland. Also important are bleeding issues and current medications related to cardiac disease, hypertension, and anticoagulation drugs (prescribed or OTC). Laboratory values for CBC and coagulation factors should be suitable, as discussed earlier. If there is any irregularity in the clinical history, such as severe headaches or hypertension, evaluation of serum adrenal hormones is appropriate. Deferral of a procedure until these values are available is appropriate. Adrenal mass biopsies can be performed as outpatient procedures unless the patient is frail or has other significant issues. Depending on the outcome of the procedure, the patient may be sent home or admitted. There is no reason to keep the patient overnight unless other issues arise. Planning of the approach is made from a review of prior CT, MRI, and PET scans (see Fig. 67-114) to note the location of the mass, which may suggest initial plans for patient position and needle trajectory help plan the procedure relative to possible approaches and patient

CHAPTER 67  Image-Guided Aspirations and Biopsies

FIG 67-115  Oblique biopsy of the adrenal gland can be performed if the mass is large enough and the anatomic structure is suitable. This scan shows a mass in the right adrenal gland, which can be approached from a retroperitoneal direction without traversing the liver or kidney.

positions; it saves time to pre-position the patient if a position other than supine is to be used (Figs. 67-116 and 67-117; see Figs. 67-114 and 67-115). Visualization of the adjacent pancreas, spleen, splenic pedicle (see Fig. 67-115), liver, vena cava, renal pedicle (if the adrenal is lower than usual), and segmental arteries is important to avoid collateral damage.

Imaging Preferences.  Choice of modality for guidance is CT because the adrenal area is difficult to image owing to its high and deep location near the spleen, liver, and vessels. Use of the dynamic contrast scan is helpful to assess lesion vascularity, necrosis, and vascular anatomy.

Procedure Steps.  During the procedure, vital signs, neurologic status, and pulse oximetry should be monitored. Analgesics and sedatives can be administered as needed. Target and trajectory determination.  A diagnostic series of scans is taken in anticipation of the procedure, which includes a dynamic bolus scan when the site is chosen. After the evaluation, one should consider the four different anatomic approaches as follows, recognizing that the complication rate varies with each: an oblique-angled approach, a direct posterior paraspinal approach, a transrenal route, and a transhepatic approach. Depending on the selected approach, the patient should be positioned appropriately. Dynamic bolus contrast scan over the chosen target site should be performed with the patient in the chosen position. This provides better visualization of the mass with its contours and its relationship to adjacent structures, such as the vena cava and the left side of the splenic vessels (Fig. 67-118; see Figs. 67-116 and 67-117). Targeting the most viable portion of an adrenal mass is appropriate. If necrosis or fluid is present, these areas should be avoided (see Fig. 67-116). Necrosis can be better appreciated (see Fig. 67-118). The first and most desirable approach uses an oblique angled pathway that avoids all intervening structures (see Fig. 67-114). This approach is ideally suited for cutting and aspiration needles because all structures are avoided and it can be used on the right or the left side. This approach is technically difficult, however, and requires an

2477

entrance site below the level of the lesion. On the right side, the angled pathway is between the right lobe of the liver and the kidney; on the left side, the pathway traverses the retroperitoneum behind the tail of the pancreas, medial to the spleen and anterior to the kidney. Sometimes, having the patient in an oblique position expedites this approach. This approach and the next one are suited to aspiration and cutting needle procedures. The second most desirable approach, which is probably the easiest to execute, is a posterior paraspinal route that traverses the crura. This is a straightforward method that can be used by almost all interventionists because it does not require an angled approach. If a paraspinal approach is taken, an aspiration or a cutting needle can be used (see Figs. 67-116 to 67-118). Karampekio and colleagues described the injection of saline into the paraspinal space to widen the potential anatomic pathway.250 The next most desirable approach, which is limited to an aspiration needle, is the transrenal approach (Fig. 67-119). Although this approach is contrary to our general preference for not penetrating uninvolved structures, historically, punctures of the kidney have been safe. Literally thousands of aspiration procedures were performed on renal cysts under standard fluoroscopy before modern imaging without significant problems. This approach is less appealing than the prior two but is preferable to the transhepatic approach, which has the highest complication rate. If a complication occurs, it is confined to the retroperitoneum, making diagnosis and treatment easier than if a pathway crossing the peritoneum and retroperitoneum is used. The last approach for the right adrenal gland is a transhepatic approach, which has been proposed by several authors.33,355 We have not used this approach because of its high complication rate (see “Complications,” later). This approach can be used with the patient in a supine or oblique position. If the transhepatic approach is chosen, the large portal and hepatic veins should be avoided. Two approaches reported in the literature that we do not recommend because of serious potential complications are the transpancreatic and transsplenic approaches. Anecdotal experience is reported in the literature and at meetings, but the risk is too high for routine use (see “Complications”). Instrument choice and techniques.  The choice of needles for tumors is based on the clinical history of malignancy, accessibility of the lesion, vascularity of the lesion, and type of tissue sample required. A coaxial system should be used for either the FNA or the cutting core needle. In either case, a 20-gauge needle is appropriate, but if a mass is very large, an 18-gauge needle can be used (Fig. 67-120). Use of a coaxial system is helpful for the acquisition of several samples. If two FNA samples are taken, the coaxial cannula can be angulated slightly so that the second sample can be at a slightly different location. In most instances, if the FNA procedure goes well,  the second sample can be a cutting needle; caliber choice depends on the size of the lesion, vascularity, and accessibility (see Fig. 67-116). Even with small-caliber needles, bleeding can occur, so if there is  any question about vascularity, a 20-gauge needle should be used (see Fig. 67-120). Choice, palpation, and preparation of the entrance site.  When the tentative trajectory and entrance site are chosen, a marker is placed on the skin and a repeat scan is performed to see the marker relative to the pathway. The marker may be a purchased device or a hypodermic needle in the package. Careful antiseptic preparation of the skin is important. Proper administration of lidocaine to produce a “painless” procedure, as noted earlier, is important. When the area is anesthetized, it is important to infiltrate anesthetic through the subcutaneous tissues and down to and including the capsule of the kidney. Adequate anesthesia of the capsule makes the procedure virtually pain free and permits insertion of the needle without the patient’s

T

A

B

C FIG 67-116  A, The patient was positioned in the decubitus position for a paraspinal approach and bolus dynamic scan. Note enhancement of the splenic vein anterior to the adrenal mass; obviously this should be avoided. Because of the location of the upper pole of the left kidney and the low position of the costophrenic angle of the lung, a lower entrance point for the needle was chosen. This required cephalad angulation to avoid the pleura. B, By using a coaxial guidance cannula, a single placement of the cannula permits repetitive biopsies. This shows a Chiba needle placement into the mass. C, After completion of the aspiration sampling, a small cutting needle was inserted through the guidance. This shows placement of the cutting gap of the 20-gauge cutting needle in mass (arrow). The sample was taken without difficulty.

A

B

C

FIG 67-117  A, Sampling of adrenal masses can be accomplished best by placing the patient in an ipsilateralside-down position to immobilize the mass. The entrance site should be at the junction of the transverse process and the rib. Local anesthetic should be amply applied. B, At the level of the diaphragmatic slip, more local anesthetic should be given because movement would move the needle. C, Final placement of the needle tip into the gland. Final adjustments can be made by turning the bevel as previously described.

CHAPTER 67  Image-Guided Aspirations and Biopsies

C A

B

D FIG 67-118  A, Bolus dynamic scan clearly shows the necrotic central core of a mass, indicating that the needle should be localized to the margin. B, The tumor has a necrotic center (arrow) medial to the viable wall sampled in C and D. C, FNA biopsy of the wall through the cannula. Because the biopsy went well, cutting needle biopsy was performed. D, Cutting needle biopsy of the wall; note the gap of the stylet localized in the viable periphery.

2479

2480

PART III  Image-Guided Procedures

A

larger needles (e.g., 18 gauge) may be used, but we do not normally use larger sizes. If the mass proves to be cystic and a fluid sample is obtained, one should attempt to get two samples—one for the inner wall of the cyst and a fluid sample. With a coaxial system in place, this is straightforward. The needle is advanced to the opposite side of the wall and a reciprocating motion is used to get cells from the inner wall. The sample is delivered to the slides and fixative. With the coaxial cannula in place, a fluid sample can be withdrawn for cytology and culture if it appears purulent. The coaxial system is left in place until the required samples are obtained. If the site seems to be bleeding, we suggest using the coaxial/ coil/thrombin hemostatic method described earlier. If the local injection of blood products is being used, another aliquot of FFP or platelets should be injected into the site as the coaxial system is withdrawn. The coaxial cannula is withdrawn, and a repeat scan is performed to assess for bleeding.

Postprocedure Care.  Before removing the patient from the table, we take a repeat scan to evaluate for postprocedure problems. If any hematoma is noted, the clinical team is contacted, and a type and crossmatch of blood products may be ordered if there is a concern that the bleeding is significant. The postprocedure care is simple and consists of observing patients for flank discomfort and monitoring vital signs for possible bleeding. The patient is kept on bed rest for 4 hours. Vital signs are checked every 15 minutes for 1 hour, every 30 minutes the next hour, and every hour for the next 2 hours. After 4 hours, the outpatient can be sent home with appropriate instructions.

Diagnostic Accuracy

B FIG 67-119  A, Mass in the left adrenal gland in front of the upper pole of the left kidney. Other trajectories were not suitable, so coaxial cannula with a Chiba needle was inserted through the upper pole of the kidney. B, Final needle position in the adrenal mass. Transrenal approach is a good alternative route to such a mass; trajectory through pancreas or spleen should not be used because of frequent complications (see text).

moving because of pain. If a rib is inadvertently hit during the initial anesthetic administration, we give a small bolus of lidocaine on the rib; this prevents any recurrence of pain if that area is struck again during the procedure. Needle insertion.  The appropriate needle, either aspiration or cutting needle, is selected using the factors discussed earlier. The needle is inserted with incremental steps and adjusted until localization in the target site is accomplished. The patient is asked to hold his or her breath during each movement of the needle. If the lesion and needle are in the same axis, the procedure is usually straightforward, but if not, the angled methods noted earlier are appropriate (see Fig. 67-49). Sample acquisition.  The sample should be acquired as noted in the general section on technique. We routinely take two FNA samples and one or several cutting needle samples depending on the quality of the sample as noted on visual inspection. If a special sample has been requested for flow cytometry, DNA analysis, or microchemistries,

The accuracy of CT sampling of adrenal masses is high. Heaston reported a series of 14 cases and had correct results in 18 patients without complications.212 Bernardino and coworkers reported a series of 50 cases with an accuracy rate of 83%.33 They had an 11% complication rate overall, with serious complications requiring blood transfusions in 3.5% of cases. Silverman and colleagues reported a positive predictive accuracy of 96% and a negative predictive accuracy of 91%.451 Obtaining benign tissue was a high indicator of benignity. Lumachi and colleagues reported definitive diagnosis in 97%.317 The authors discounted the risk of sampling pheochromocytomas, to which we strongly object (see “Contraindications,” earlier). The feasibility of sampling even the most subtle mass in a patient with a malignancy is best emphasized by the experience of Pagani.395 In his series of patients with small cell lung tumors, he sampled normal-appearing glands and recovered malignant cells in a significant number. With an adenoma or prominent normal gland, our experience and Pagani’s395 has shown that a cytologic sample shows the presence of normal adrenal cells in nonfunctioning adenomas or hyperplasia, excluding the possibility of a malignant or inflammatory disease. If the cytologic interpretation raises any question about the benign nature of the adrenal cells, further exploration is needed.

Complications Numerous problems have been reported. Bernardino and colleagues, using a transhepatic approach in a reported series of 50 patients, had an overall complication rate of 11%, with a serious complication rate of 3.5%.33 Two patients required blood transfusions. Pagani reported a case of pneumothorax that required chest tube placement.396 Heaston212 and our group, using an indirect approach that attempted to avoid other structures, had no significant complications. There have been two deaths reported (see “Contraindications,” earlier).

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2481

B FIG 67-120  A, A 20-gauge FNA needle in margin of mass. B, Small limited hemorrhage occurred, so cutting needle biopsy was not performed.

More recently, other authors have reported their experience with complications. Kane and coworkers reported a 6% incidence of severe pancreatitis when the left adrenal gland was sampled with aspiration needles using an anterior transpancreatic approach.247 Mesurolle and colleagues reported 6 cases of pneumothorax in a series of 44 patients.347 Mody and associates described an 8.4% complication rate in a series of 44 patients using various approaches.355 Complications have occurred with biopsy of pheochromocytomas. The transhepatic approach had a 12% complication rate, and the posterior approach had an 8% rate. Two pneumothoraces and one hematoma occurred with the posterior approach, and four hematomas occurred with the transhepatic approach. These authors reported seven cases of complications (pneumothoraces in two, severe pain in two, hematomas in three, and needle track metastasis in one). It also was noted that five of the seven complications occurred with a 22-gauge needle. McCorkell and Niles reported two cases of FNA of pheochromocytomas with catastrophic results339; both patients had severe hemorrhage, one of whom later died. Casola and associates reported four cases of unsuspected pheochromocytoma58; one patient had severe hypertensive instability during the biopsy. In view of the numerous reported complications, we believe our conservative approach is warranted. To date, we have had only two pneumothoraces over the past 30 years. Since the adoption of the transcrural paravertebral approach, no pneumothoraces have occurred.

RETROPERITONEUM AND LYMPHOMA The retroperitoneum is an anatomic space and not an organ, so the relevant discussion relates to malignant, infectious, and nonspecific inflammatory lymph node pathology. Lymphoma is the most common primary tumor of the retroperitoneum and can be difficult to diagnose without adequate tissue biopsy. Because lymphoma is the one tumor that has predictably favorable prognosis with medical therapy, early reliable diagnosis is critical. Metastatic nodes are common and occur from the spread of various organ malignancies, including pancreas, kidney, colon, ovary, prostate, and others. Infectious processes (e.g.,

M

FIG 67-121  For all retroperitoneal biopsies, it is prudent to perform a dynamic bolus CT scan. This scan shows a mass in the hilum (arrow) of the right kidney; note the numerous small vessels around the mass (M). A small-needle aspiration biopsy might be possible, but in cases such as this when lymphoma is more likely, such samples are not helpful. Although the risk of a problem is small, the probability of success is low.

with Mycobacterium spp.) or granulomatous processes also produce extensive lymph node disease, which can be problematic to diagnose and differentiate from malignancies.

Image-Guided Background Because of the retroperitoneal anatomy and the need for large cutting needle samples to differentiate among the various lymphomas, CT procedures are uniquely advantageous. CT is the preferred guidance modality, especially with a dynamic bolus scan, because of the many vessels and subtle anatomy of the area, the clear visualization of the complicated anatomy and instrumentation that CT provides, and its unparalleled unlimited display of anatomy (Fig. 67-121).

2482

PART III  Image-Guided Procedures

In prior editions, our discussions were based predominantly on our own experience because very few groups had reported significant experience. In recent years, CT procedures have been widely adopted, and numerous authors have reported their experiences, affirming the important role of CT procedures for diagnosing lymphomas.

Indications The indications for biopsy of retroperitoneal structures are twofold: (1) clinical suspicion of malignancy and severe idiopathic inflammatory processes, and (2) abnormal appearance of lymph nodes or appropriate findings in the area suggesting a high probability of cancer or unusual benign disease. Because of the credibility and reliability of CT detection of disease, the mere observation of appropriate lymph node or mass findings is usually sufficient indication to pursue a biopsy. For more detailed discussion, the reader is referred to Chapter 55, but the criteria we use are straightforward. Lymph node size and number are the primary determinants of whether a node should be biopsied, because the appearance of lymph nodes is not specific. Lymph nodes measuring 2 cm or greater have a high probability of being pathologic or cancerous and are appropriate for biopsy. Such enlargement results from cancer, but significant inflammatory processes, such as sarcoidosis, tuberculosis, and others, likewise can produce enlargement. Biopsy should be considered for lymph nodes that have increased in number over many months. Nodes that are less than 1 cm are unlikely to be neoplastic, but in unusual circumstances, biopsy may be appropriate. The procedure is more difficult because of the small size of the target.

Contraindications The proximity of vessels and other structures can prevent safe biopsy and is a contraindication. Arteries traverse all portions of the retroperitoneum because the primary blood vessels to and from the various organs traverse the retroperitoneum throughout the axial planes (e.g., renal artery, superior mesenteric arteries, adrenals); these blood vessels usually have a predictable location and pathway, but aberrant or collateral vessels may occur, embryologically or as a result of pathologic processes. Anatomic variations of various organs can make trajectory planning difficult or impossible depending on the location of the nodes.  In the upper abdomen, periaortic and pericaval vessels are central, and the overlying organs and intervening structures may preclude biopsy. The most notable example of such a situation are nodes in the upper abdomen, where the pancreas, splenic pedicles, porta hepatis, and diaphragms can make planning and execution of such procedures difficult (see Fig. 67-121). In the pelvis, the combination of the pelvic bones, the anterior overlying bowel, the iliac vessels, and the femoral/sciatic nerves sometimes makes trajectory planning problematic. Although some authors have suggested traversing such uninvolved structures as appropriate, common sense suggests that complications would ensue. Uncorrectable coagulopathy precludes biopsy because of the risk of bleeding complications, although the hemostatic methods have been described earlier for managing coagulation disorders.

Retroperitoneal Biopsy

Preprocedure History, Review of Images, Planning, and Informed Consent.  Patient history and diagnostic scans are reviewed and informed consent obtained as noted earlier. Review of patient history is generally related to illnesses, symptoms, and medications. Also important are bleeding issues and current medications related to cardiac disease, hypertension, and anticoagulation drugs (prescribed or OTC).

Review of prior scans to note the location of the mass may suggest initial plans for patient position and needle trajectory to help plan the procedure relative to possible approaches and patient positions; it saves time to pre-position the patient if a position other than supine is to be used. PET scanning has been used a great deal for staging the overall extent of lymphoproliferative disease, so it may provide a road map for choosing the largest lymph node suitable for biopsy. Laboratory values for CBC and coagulation factors should be suitable, as discussed earlier.

Imaging Preferences.  Choice of modality for guidance is CT because neither US nor MRI is as consistent imaging this area. CT is unparalleled in its ability to see accurately the adjacent anatomy, renal pedicle, and collecting system, which occasionally may be helpful. Use of the dynamic contrast scan is helpful to visualize vessels, making it possible to biopsy paraaortic and paracaval lymph node masses.

Procedure Steps.  During the procedure, vital signs, neurologic status, and pulse oximetry should be monitored. Analgesics and sedative medications can be administered as needed. Target and trajectory determination.  A diagnostic series of scans is taken in anticipation of the procedure. From the current scans, the mass referred for biopsy is reevaluated. Dynamic bolus contrast scan over the chosen target site should be performed to better visualize the mass with its contours and its relationship to adjacent structures. Depending on the part of the retroperitoneum being sampled, it is crucial to visualize the aorta, vena cava, renal pedicle, iliac vessels, mesenteric vessels, gonadal vein, inferior mesenteric vein, and ureters (Figs. 67-122 and 67-123). If the vena cava cannot be identified, an injection of the foot vein can be performed to get better opacification. Unusual vascular abnormalities may occur that could produce catastrophe if an inappropriate biopsy is performed (Fig. 67-124; see Fig. 67-26). Because of the varying anatomy in this area, the trajectory differs at the levels of the retroperitoneum, as follows. In the upper abdomen, the approach may be posterior or lateral depending on the location of the abnormality and uninvolved organs. In most cases a suitable trajectory just lateral to the margin of the vertebral body at an angle that essentially “blocks” any path toward the aorta (see Fig. 67-123) is usually most suitable. Avoidance of the psoas muscle is best because the psoas can be flexed inadvertently by the patient and slightly deviate the needle. Also, a path through the psoas is more uncomfortable to a patient. If the crura of the diaphragm are present, traversing through the crura is better than traversing adjacent to them, to avoid any medial portion of the pleural space. In the midabdomen, a posterior approach that passes just lateral to the psoas muscle is best. The posterior approach for such procedures is easier with the patient lying in a lateral, anterior oblique, or prone position (Fig. 67-125). If an aspiration needle is used, an anterior or posterior approach can be taken, but if a cutting needle is used, a posterior approach is necessary. The areas that are difficult to biopsy safely are the retropancreatic region and the area of the aortic bifurcations. In these areas, multiple vessels are branching and converging that make exact delineation difficult. In these areas, I usually perform only an aspiration biopsy to avoid the possibility of injuring one of the vessels. Instrument choice and techniques.  The choice of needles for tumors is based on the clinical history of malignancy, accessibility of the lesion, vascularity of the lesion, and type of tissue sample required. If the purpose of the procedure is to confirm a know malignancy, only FNA is required (Fig. 67-126). If the disease has not yet been diagnosed, a cutting needle is important because of the common occurrence of lymphoma in this area.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

FIG 67-122  A, CT scan shows a poorly defined mass around the aorta, enhanced by contrast medium. B, CT scan during biopsy shows a mass adjacent to the aorta, which proved to be retroperitoneal fibrosis. Tissue samples with a 14-gauge needle are ample to diagnose unusual processes.

m

A

B

FIG 67-123  A, Patient with mass (m) adjacent to the kidney. This is more posterior than noted in the case in Figure 67-121, and the vessels are located anteriorly. B, Cutting needle biopsy provided definitive information for lymphoma.

A

B FIG 67-124  A, CT scan of a patient referred for biopsy shows a large mass to the left of the spine. B, CT scan with patient on the right side was performed with bolus. The mass proved to be a large hematoma in a contained rupture of the aorta. No biopsy was done.

2483

2484

PART III  Image-Guided Procedures

A

B

FIG 67-125  A, CT scan shows a poorly defined mass around the aorta, enhanced by contrast medium. B, CT scan during biopsy shows the mass adjacent to the aorta, which proved to be retroperitoneal fibrosis. Tissue samples with a 14-gauge needle are ample to diagnose unusual processes.

C

A B FIG 67-126  A, CT scan shows a small mass (arrow) between the vena cava and aorta. After bolus injection, the needle was inserted between the vessels for cytologic sampling. B, The accuracy of needle placement is so precise, one can even “thread” a needle between the vessels and sample structures (C) slightly anterior. When passing needles by such vessels, it is always important to have the bevel directed toward the vessel.

A coaxial system should be used for either the aspiration or the cutting core needle. A 20-gauge needle is our choice if only an FNA is to be obtained. As noted earlier, it provides a consistent sample with few complications. Use of a coaxial system is helpful for acquisition of several samples. If two FNA samples are taken, the coaxial cannula can be angulated slightly so the second sample can be at a slightly different location. In many instances, even if a cutting needle is to be used, we perform an FNA first because it provides a “test” of the feasibility of the chosen trajectory and target and assesses how cooperative the patient will be. A coaxial system also is used for a cutting needle biopsy. In contrast to other organs, the most common neoplasm encountered is lymphoma, so it is best to use the largest possible cutting needle with which the physician is comfortable using. Side cutting needles are superior to end cutting needles because they provide the most intact samples with the least crush artifacts. We believe that only double-action needles

should be used because of safety issues (see “Indications for and Selection of FNA and Cutting Needles,” earlier). Choice between the two types of side cutting needles is important in this area to prevent serious complications (see Fig. 67-39). As discussed earlier, there is a distinction between single-action and doubleaction needles of the automated varieties. It is critical to ensure safe tissue acquisition in the retroperitoneum because many of the lymph nodes sampled are near the vena cava, organ vessels, or aorta. The single-action needle activates both moving components of the side cutting needle—the inner stylet (with the tissue gap) and the outer cutting cannula. The action of the needle is two sequential steps. The stylet pushes into the tissue, then the outer cannula advances over the stylet, cutting the tissue into the gap. The location of the stylet determines where the cut occurs. With the single-action needle (called single because the operator has a single action, pushing the button), both components move in sequence after the tip is positioned adjacent to

CHAPTER 67  Image-Guided Aspirations and Biopsies the target. Because the cutting occurs quickly without the operator’s seeing the stylet, the precise location of the cut can vary if the patient moves or the operator changes the angle slightly. Although this is not an issue with large masses, it may be an issue with small masses. Theoretically and practically, the target can be missed and an uninvolved structure can be damaged. The double-action needle, requiring two actions by the operator, is the best device to accurately target the abnormality and avoid other structures. With this device, the stylet is pushed in first by the operator, and a scan can be obtained; the cutting gap can be seen clearly, and this determines whether the cutting action occurs, because the cutting cannula follows the stylet. If the gap is positioned completely in the mass, even if very close to the aorta or vena cava, the biopsy occurs safely because “what you see is what you get.” If the needle is held stationary, there is only a remote chance of any complication. Choice and palpation of the entrance site.  When the tentative trajectory and entrance site are chosen, a marker is placed on the skin and a repeat scan is performed to see the marker relative to the pathway. The marker may be a purchased device or a hypodermic needle in the package. Site preparation.  Careful antiseptic preparation of the skin is important. Proper administration of lidocaine to produce a “painless” procedure, as noted earlier, is important. When the area is anesthetized, it is important to infiltrate anesthetic through the subcutaneous tissues and down to the surface of the retroperitoneal mass; one can do this through an aspiration needle during the course of the FNA. We have evaluated the effect of lidocaine on cytologic samples and interpretation and have found no ill effects on the sample. If, during the course of the local anesthesia, the side of the vertebral body or the transverse processes are stuck with the needle, we give a small bolus at that site so that pain does not occur again at a later time. Needle selection and insertion.  The appropriate needle, either FNA or cutting needle, is selected using the factors discussed previously. For critically located masses, such as between the aorta and vena cava, only FNA should be performed (see Fig. 67-126). The needle is inserted with incremental steps and adjusted until placement in the target site is accomplished. Needle insertion should occur within the slice axis of the scan. An angled approach is not recommended, because there are so many structures in the retroperitoneum that should be avoided. Vascular pedicles would be in jeopardy with an angled approach. Sample acquisition.  The sample should be acquired as noted in the general section on technique. We routinely take two FNA samples and one or several cutting needle samples depending on the quality of the sample as noted on visual inspection. If a special sample has been requested for flow cytometry, DNA analysis, or microchemistries, larger needles (e.g., 18 gauge) may be used, but we do not normally use larger sizes. As many samples are taken as needed (see Tables 67-3 and 67-4). The coaxial system is left in place until the required samples are obtained. If bleeding from the site seems to be occurring, we suggest using the coaxial/coil/thrombin hemostatic method described earlier. If the local injection of blood products is being used, another aliquot of FFP or platelets should be injected into the site as the coaxial system is withdrawn. The coaxial cannula is withdrawn and a repeat scan is performed to assess for bleeding.

Postprocedure Care.  Before removing the patient from the table, we take a repeat scan to evaluate for postprocedure problems. If any hematoma is noted, the clinical team is contacted, and a type and crossmatch of blood products may be ordered if there is a concern that bleeding is significant.

2485

The postprocedure care is simple and consists of observing patients for flank discomfort and monitoring vital signs for possible bleeding. The patient is kept on bed rest for 4 hours. Vital signs are checked every 15 minutes for 1 hour, every 30 minutes for the next hour, and every hour for the next 2 hours. After 4 hours, the outpatient can be discharged home with appropriate instructions.

Diagnostic Accuracy A review of the history of biopsy for lymphoma and retroperitoneal tumors reveals remarkable improvement in the diagnosis after the introduction of the core biopsy needle. The first specific report on retroperitoneal biopsies was by Knelson and colleagues of our group, who reported accuracy for diagnosis of lymphoma and other retroperitoneal disease.266 In this seminal article, we compared the 20-gauge FNA needle and the 14-gauge side cutting needle and reported that for lymphoma, solid tumors, and benign disorders, the cutting needle was far superior. The aspiration needle diagnosed metastatic disease in 20 of 22 cases and suggested the diagnosis in 10 of 15 lymphoma cases. With the larger cutting needle, the diagnosis for metastases was made in 13 of 13 cases, and the specific subtype diagnosis of lymphoma was made in 10 of 11 cases. Affirmation of the concepts introduced by Knelson and colleagues266 was provided by Quinn and associates412 and Hunter and associates.234 Hunter and associates234 noted that for the retroperitoneum, CT guidance was the preferred modality in 9 of 11 cases, and definitive diagnosis was made in 11 of 11 cases. The core biopsy correctly diagnosed the subtype of lymphoma in all cases and diagnosed the unusual variant of lymphomatoid granulomatosis. Cytology was successful in only 7 of 11 cases. Flow cytometry confirmed these diagnoses in 8 of the 11 cases. Several large series have defined the global benefit of core biopsy needles with regard to the impact of core biopsy needles on the diagnosis, treatment, and management of lymphomas. Pappa and colleagues reported on a group of 106 patients with lymphoma.397 The authors achieved an accurate diagnosis in 88 of 106 patients and noted that the small size of the sample was the impediment for better numerical results; improvement could be predicted if multiple needles or a large needle size was used. The technique was equally as successful in the diagnosis of Hodgkin’s disease and high-grade and low-grade nonHodgkin’s lymphoma as in nonlymphoproliferative disorders. The procedure was equally successful at initial diagnosis as at suspected recurrence of progression. Change in the histologic appearance was observed in 33 of 80 cases and influenced treatment. Transformation of follicular non-Hodgkin’s lymphoma was diagnosed in 31 of 33 cases, allowing appropriate adjustment of treatment. de Kerviler and colleagues reported two large series.97,98 In 2000, they reported a group of 212 biopsies on lymphoma patients. The group included 103 patients for original diagnosis and 109 patients for recurrence or progression. The group used needle calibers ranging from 20 gauge to 14 gauge. Immunohistochemistry studies were performed in all lymphoma cases. A diagnosis of lymphoma with subtyping was observed in 88% of all cases. Therapy was initiated on the basis of core biopsy in 93% of all cases, in 91% at initial diagnosis, and in 94% at follow-up. Benign complications occurred in 7.5% of cases and did not require specific treatment. The more recent study in 2007 by de Kerviler and colleagues provides additional confirmation of CT-guided core biopsies of lymphoma in a group of 180 patients with peripheral lymph node disease97; 123 image-guided node biopsies were performed at first diagnosis, and 57 large cutting needle core biopsies were performed to diagnose progression. A diagnosis of lymphoma with sufficient information such that a therapeutic decision could be made was obtained in 146 of the 152 patients with lymphoproliferative disorders (96%).

2486

PART III  Image-Guided Procedures

Ravinsky and Morales detailed the additional diagnostic benefit of core needles compared with FNA.417 The key benefits for providing information about subtype included observation of “nodularity,” improved evaluation of grading, and use of specific immunohistochemical prognostic markers Ki-67 and p53. Goldschmidt and colleagues reported similar results for the common types of lymphoma, but they also emphasized the usefulness for more unusual tumors.158 The total group of 130 patients with lymphoproliferative disorders included 10 patients with chronic lymphocytic leukemia, 6 with combined malignancies, and 2 with Castleman’s disease who underwent CT-guided core needle biopsy. Their series is unique in the wider range of lymphoproliferative diseases diagnosed successfully by CT core biopsies. In retrospect it is difficult for contemporary physicians to realize, but during the introduction of this new technique there was remarkable controversy and political turmoil against using these larger needles. In subsequent years, virtually every author who has reported on the topic has reconfirmed the advantages of the core biopsy needle; the only point of contention has been the caliber of the instrument required.

Factors Affecting Accuracy.  Experience, larger needle size, and multiple needle passes improve the accuracy. The diagnostic accuracy for de Kerviler and colleagues improved from 88% to 96% with the use of larger needles and multiple needle passes.97,98 They noted that the results of the later series did not depend on the biopsy site, lymph nodes size, or needle type. Larger needles provide larger samples, which are superior for more complete analysis. Hesselmann and coworkers, reporting on lymphomas, concluded, “The most important parameter for diagnostic success is the size of the specimen.”219 From de Kerviler’s reports, it is insightful to note that their earlier report,98 in which they used 20-gauge and 14-gauge needles, had a lower success rate than their later report, in which they used 18-gauge and 14-gauge needles.97 The same argument can be derived from Agid and colleagues, who reported their series using 20-gauge and 18-gauge side cutting needles.6 Their total accuracy of 76% and 82% was less favorable compared with de Kerviler’s group, which used 20-gauge and 14-gauge needles. Anatomic location of lesion site undoubtedly has an indirect effect on the accuracy rate. When anatomic approach is more difficult because of organ proximity or proximity of vessels or other vital structures, most authors are appropriately inclined toward smaller needles to ensure safety. Sklair-Levy and associates, reporting on mediastinal lymphoma, reported a success rate of 71.5% owing to choice of a smaller caliber needle.457 Results reported by de Kerviler’s group varied between their two series on deep lymphoma and peripheral node disease.97,98 Because peripheral nodes were less critically positioned, the authors opted to use larger needles, and they had better results (96%) for 14-gauge cutting needles used for peripheral nodes compared with results (88%) for 18-gauge needles used for more central tumors. Finally, a combination of all of these factors was reported by Hesselman and colleagues, who stated, “Diagnostic yield was correlated with features such as size of specimen, location and depth of the target lesion, and experience of the investigator.”219

Complications Considering the benefit of the cutting needles and the flexibility of using larger gauges, and the history of acceptably low complication rates, the risk-to-benefit ratio of this technique is favorable. In the first clinical comparison of large and small needles by our group,266 only one complication occurred. In one of the largest series of 212 biopsies on lymphoma patients by de Kerviler and colleagues, minor

complications occurred in 7.5% of cases and did not require specific treatment.97,98 Pappa and coworkers reported no complications in a group of 106 patients.397

Factors Affecting Complications.  Dynamic bolus scans for vascularity are probably the most critical step for planning the trajectory for retroperitoneal biopsy. The myriad vessels are potential sources of bleeding if any are accidentally biopsied. The stationary dynamic bolus scan during the trajectory phase planning is the most important for avoiding arterial structures. Venous structures are more difficult to see, but they are less risky relative to serious bleeding events. At our institution, one patient who had a musculoskeletal biopsy died of bleeding that might have been prevented if a dynamic bolus scan had been performed to visualize the vessels, permitting avoidance before the procedure. There is little question that type and size of needles and the number of needle passes affect potential complication rates. The reported experience with retroperitoneal biopsies theoretically would have the highest chance of complications, but from the reports in the literature, very few complications have been reported. Paradoxically, the low rate is likely due to a heightened state of awareness and the use of meticulous technique relative to selection of needle size, anatomic pathway, and avoidance of collateral damage. With the experience of biopsy in other organ systems, which usually precedes these biopsies, it is likely that interventionalists had sufficient experience and expertise to expand this technique into the retroperitoneal area without significant problems. Experience and technique outweigh the significance of needle geometry or size. Movement of bowel by positional changes or injection of gas or fluid as described earlier is important in the lower portion of the retroperitoneum and the pelvis. Because of the bony ring of the pelvis, access is restricted to only the anterior abdominal wall without the latitude available in the other portions of the abdomen. Comprehensive anatomic knowledge of the retroperitoneal region is crucial to prevent collateral damage to normal structures. The reader is referred to Chapter 55 for an understanding of the complex structures present. Whether or not one can see the unopacified ureter, gonadal veins, inferior mesenteric vessels, or lumbar vessels, one should always anticipate their location and avoid the areas where they typically lie if possible. In the lower pelvis, knowledge of the location and path of the femoral, iliac, and obturator nerves is important to avoid damage. The site of peritoneal reflections over the pelvic organs is important if one wishes to avoid spreading tumor cells from an ovarian mass if a biopsy is performed. Recognition of unusual variants also is important in all aspects of intervention. One unusual variant that may be seen is an anomaly of the cisterna chyli. Primary cystic masses are rare in the retroperitoneum, but there is an unusual variant of normal adjacent to the crura of the diaphragm that might be misconstrued as an abscess (Fig. 67-127). This usually represents an ectatic cisterna chyli, and puncture is unnecessary unless there are clinical factors that would suggest a pathologic condition. If the structure is punctured, the fluid recovered is very white and “milky” in appearance because it contains chyle accumulated by the lymphatic system of the bowel (see Fig. 67-127B).

PERITONEUM AND MESENTERY Biopsy of peritoneal lesions is uncommonly performed, so discussion of this topic is brief. There are many similarities to other sites relative to the need for preprocedure steps, laboratory studies, and choice of needles. Lesions can be primary or metastatic, so one chooses an aspiration needle to confirm malignancy or a cutting needle to do a

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2487

B FIG 67-127  A, CT scan at the level of the diaphragm shows a cystic mass behind the right crura (arrow), which represents an ectatic thoracic duct or cisterna chyli. B, White fluid containing chyle, which was aspirated from such a lesion.

definitive biopsy. Whenever a cutting needle is used, it is prudent to perform a bolus scan to visualize any mesenteric vessels, splanchnic aneurysms, or vascularity of lesions. As in other areas, if there is a possibility of penetrating a loop of bowel, a small aspirating needle (20 gauge) is used. If the mass can be approached without penetrating a bowel loop and the other criteria are met, a larger cutting needle (e.g., 18-or 14-gauge) can be used. These procedures are slightly more difficult than procedures in other sites because of the interspersed loops of bowel, the vessels in the mesentery, and the mobility of the structure (Fig. 67-128). Administration of oral contrast material is usually helpful for small masses to avoid inadvertent puncture of an unopacified loop of bowel. When a cutting needle is used, it is important to ensure that the cutting gap  is completely within the mass (Figs. 67-129 and 67-130). If a part  of the cutting needle were to overlap, some of the abundant small vascular network of the mesentery or omentum may be cut, causing a hemorrhage.

SPLEEN The spleen is a problematic area because it is an organ that is prone to serious injury or complications from trauma and surgery owing to its natural high-flow blood supply. Tumorous lesions to the spleen are rare, so large clinical experience has not been accumulated. Although the literature contains observational studies about biopsying the spleen, no extensive experience is available to review or study. Although we have performed image-guided interventions for 30 years, we have not pursued cases involving the spleen because of our bias. We are not convinced that the benefit sufficiently outweighs the potential risks for widespread application. This is one of the few areas we think should be left to academic centers until more information is obtained. Several reports are available. Splenic biopsy was reported by Keogan and colleagues.254 They performed 20 biopsies in their series of 18 patients. They used a variety of needles and achieved an accuracy of 88.9%. Lieberman and associates reported on the use of FNA for the diagnosis of lymphoma and recommended such procedures.301 Although these authors conclude that this is a safe procedure, we believe the experience is not sufficiently large to recommend it as a routine procedure because of the potential risk.

FIG 67-128  CT scan showing a large deep mesenteric mass with a cutting needle in place.

FIG 67-129  Biopsy of small mobile masses in the peritoneum is possible, but one must adjust the bevel of the needle as required and insert it quickly to prevent movement of the mass.

2488

PART III  Image-Guided Procedures

FIG 67-130  Biopsy of a thickened omentum (arrows) caused by peritoneal metastases.

PELVIS The pelvis is a frequent site for biopsy with CT guidance because of the common occurrence of gynecologic, urologic, and colonic pathosis.49,530

Indications Indications for biopsy in the pelvis are varied; some are well defined and others are nebulous. Clearly defined entities that should be sampled include metastatic tumors from the colon, bladder, prostate, or other organs. General principles for these procedures are the  same as the retroperitoneum, including site selection, preparation, local anesthesia, needle selection, dynamic bolus scan, and sample acquisition.

Contraindications Biopsy of an ovarian mass for clinical reasons is relatively contraindicated. Because ovarian tumors have a propensity to spread to the peritoneum, we do not perform biopsy of an ovarian mass unless a gynecologic oncologist has rendered a consultation. If the oncologist believes it is necessary, we perform an ovarian biopsy, but we use  care to try to approach the mass from an extraperitoneal approach, either from a very low entrance site at the level of the bladder or posteriorly.

General Technique for Pelvic Biopsy The indications, preprocedure steps, trajectory planning, use of dynamic bolus scanning, and choice of needles is similar to that of the retroperitoneum. Anatomic and trajectory approaches are the anterior, lateral, or posterior approach through the sciatic notch. The major normal structures to identify in this region are the iliopsoas muscles with the femoral nerve within and the associated iliac vessels. When abnormalities are in the upper portion of the pelvis, an anterior or anterolateral approach may be used. At this level it is

impossible to pass a needle posteriorly because the bony structures are intact posteriorly. When a mass is large enough to push anteriorly, access is easy because the bowel has been pushed laterally. When masses are obscured anteriorly by bowel, access can be gained by several methods. The patient can lie on either side, which may cause the bowel to move, permitting access for a needle placement. If the bowel seems mobile, one can push the needle in slowly, and the loop of bowel can be pushed aside. If neither of these methods works, injection of air or carbon dioxide may be used to move the cecum or other bowel. When such procedures are performed, the needle is placed in the small space between the loop of bowel being moved and the lateral structures (see earlier section “Intentional Organ Movement or Tissue Retraction”) (see Fig. 67-52). Lateral masses can be approached with a pathway parallel to the iliac crest, with the needle directed either medially or laterally to the iliac vessels and femoral nerve. The vessels can be identified by injecting contrast material for opacification. The femoral nerve lies in the small “notch” at the margin of the psoas and iliac muscles. When significant expertise is developed, one can sample masses or nodes even closer to the vessels. The bevel of a needle can be used to pass by a vessel. In some circumstances, one can even pass a needle between the artery and vein to reach a subtle mass (Fig. 67-131). More inferior masses can be approached via anterior, lateral, or posterior trajectories. The bladder can be avoided by ensuring that the bladder is evacuated or contains contrast material. One can avoid crossing the pelvic peritoneal surface by choosing an entrance site just above the pubic ramus and directing the needle cephalad. The posterior approach through either sciatic notch is suitable for sampling any presacral mass (Fig. 67-132). Choice of entrance point is important to avoid impingement against the sacrum because there is a discrepancy between the apparent location of the sacrum as seen on the CT image and its real location. Because the anterior, posterior, and lateral margins of the sacrum are tapered, there is a significant partial volume error created on the CT image of the sacrum. The sciatic notch should be palpated carefully before any needle insertion because the margin of the bone is more lateral and caudal than would appear on the CT image. With the information from palpation of the notch, the location of the notch can be anticipated better and the discomfort associated with the procedure minimized. One should avoid the gluteal and hemorrhoidal vessels, which are numerous and near the notch.

Accuracy There are several articles from two perspectives in the literature about the accuracy of CT-guided procedures in this area. The first article on this topic from my group530 examined the accuracy of percutaneous procedures relative to surgical biopsies. We found that percutaneous procedures were more successful than the surgical approach (either perineal or peritoneal). Butch and colleagues reported a series of CT-guided procedures evaluating the overall diagnostic accuracy, correctly diagnosing 15 of 19 cases of neoplasm.49

Complications There are few reports specifically about pelvic biopsies, so documentation of complications is difficult. The probability of bleeding is likely similar to that of the retroperitoneum.

Therapeutic Procedures CT can be used readily for performing many aspirations and drainages in different portions of the body. The greatest number of drainages that are performed are those on abscesses, but occasionally cysts, lymphoceles, pseudocysts, bilomas, or other fluid collections may be drained.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

C

D

2489

FIG 67-131  A, A mass is located posterior to an artery and vein, making the approach difficult. B, Cannula of the coaxial system and Chiba needle are noted anterior to the space between the iliac artery and the psoas muscle. The bevel was turned toward the artery to permit safe passage of both instruments past the artery. C, The Chiba needle was inserted beyond the cannula, and a sample was taken. D, The Chiba needle was replaced with a small-caliber cutting needle (Tenmo). This could be performed safely because the cutting gap was beyond the vessels, with the Chiba needle and cutting needle inserted.

LYMPHOCELES Lymphoceles are collections of lymph fluid, usually caused by the local disruption of lymphatic vessels after surgery. The appearance of these lymphoceles is indistinguishable from abscesses, appearing as welldelineated fluid collections (Fig. 67-133A). Commonly these fluid collections resolve spontaneously without any type of therapeutic intervention when new lymphatic channels open (see Fig. 67-133B). Although it is not well defined in the literature, one should wait several months to give the fluid time to resorb in the new lymphatics.

Indications There are only two valid indications for intervening when such a fluid collection is found. If a clinical suspicion of an infection exists, a diagnostic aspiration is indicated (Fig. 67-134). There is no rationalization for catheter insertion in such cases unless purulent material

is present. Inserting a catheter may cause secondary infection, but in most cases the cavity closes after the infection has resolved. The second indication is when the lymphocele is obstructing the urinary tract (Fig. 67-135) or gastrointestinal (GI) tract. In such cases, needle puncture and catheter insertion are the same as for the other fluid collections. A lymphocele persists in draining fluid until the lymphatics open new channels, whether a catheter is present or not. Removal of the tube results in reaccumulation of lymph fluid if such channels have not formed. When a lymphocele has been drained for one of these reasons, some radiologists have instilled sclerosing material, such as tetracycline, to speed resolution. In recent years, numerous authors have reported the use of various agents that sclerose lymphoceles, including doxycycline,51 bleomycin,255 ethanol,537 Betadine,147 and mitomycin C.356 The results of such sclerosing procedures have varied, and in many cases, surgeons default to internal drainage. Laparoscopic procedures have

2490

PART III  Image-Guided Procedures

been reported by Gruessner and coworkers170 and Hamilton and Winfield.202

LARGE CYSTS—ASPIRATION AND SCLEROSIS True cysts with endothelial linings can occur in many organs, but they occur most commonly in the liver and kidney. They have characteristic appearances showing a very low density centrally with a clearly demarcated margin.

Diagnostic Aspiration Diagnostic aspiration is indicated only if there is suspicion of a diagnosis other than simple cyst, such as infection or malignancy. Using

M

FIG 67-132  Biopsy of a presacral mass (M). The position of the lateral margin of the sacrum (straight arrow) can be deceptive. Because of the tapered edges, the margin is more lateral than one might expect. Careful palpation of the site should be performed to prevent any undue discomfort. The sciatic nerve (curved arrow) should always be avoided if possible.

A

the techniques described in the liver and renal sections, an aspiration needle can be used for fluid recovery. The one important issue with cystic masses is that one should be careful not to spill fluid into the peritoneal space to reduce the possibility of local spread of infection or tumor. It is always wise to approach such cysts through a cuff of normal tissue to reduce the chance of spillage.

Cyst Sclerosis

Indications.  The mere presence of a cyst in an organ is no reason for sclerosis, but in some cases cyst size may be so great as to impair respiration, produce pain, obstruct the portal structures, or obstruct the renal collecting system. The decision to sclerose a large cyst in the kidney (Fig. 67-136) or liver (Fig. 67-137) is not an easy one for several reasons. Any procedure—no matter how benign—can have adverse side effects, and the use of these sclerosing agents is not without potential risk. Also, it is unusual when a cyst produces physical discomfort, so we are very selective of our patients for sclerosis. To ensure that a patient with a cyst is experiencing symptoms and not having psychological overlay from “knowing” a cyst is there, we use a two-step approach. On the patient’s first presentation, we aspirate the cyst without sclerosing agent. In many cases, a simple aspiration suffices, and a cyst does not recur. If the patient remains symptom free and the cyst does not reaccumulate, no sclerosis is needed. If the patient comes back because of symptoms and the repeat scan shows no recurrence of the cyst, the symptoms are not caused by the cyst and there is no reason to sclerose. Finally, if the patient comes back because the cyst has reaccumulated, this proves the symptoms are from the cyst, and sclerosis is needed.

General Technique for Cyst Sclerosis

Preprocedure History, Review of Images, Planning, and Informed Consent.  Preprocedure steps, as noted earlier, include physical examination, clinical and medication history, and review of

B FIG 67-133  A, CT scan shows a postoperative lymphocele (arrow), which was symptomatic, but the patient was advised that the fluid not be drained in order to permit lymphatics to develop spontaneously. B, The patient became irate because nothing was being done, so she returned, insisting on percutaneous drainage. The fluid has now reduced spontaneously (arrow), indicating that the procedure was unnecessary.

2491

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

C FIG 67-134  A, CT scan showing postoperative lymphocele in the retroperitoneum adjacent to the spleen. B, Needle aspiration of fluid collection. C, Appearance of fluid from a postoperative lymphocele from a prior surgery. More acute lymphocele fluid may appear hemorrhagic.

LEFT U

17 B

L R

0

A

B

C

FIG 67-135  A, A large lymphocele (L) is noted in the left side of pelvis after kidney transplant. Access to the collection is impaired because of the proximity of bladder (B) and the clips from surgery (arrow). B, Positioning of the patient with the right side down increases the space for the needle trajectory (arrow), so the cyst can be punctured without penetration of the bowel or bladder, which potentially could introduce infectious agents. C, Radiograph shows the catheter in the lymphocele (arrow) and the ureter (U).

2492

PART III  Image-Guided Procedures general anesthesia may lessen the depth of general anesthesia required, which is beneficial to the patient.

Procedure Steps.  Patient vital signs and pulse oximeter monitoring

A

B

C FIG 67-136  A, This large renal cyst was symptomatic, so drainage and sclerosis was performed. CT scan shows the needle traversing the entire cyst, with a cytology sample taken from the inner wall. The coaxial system was used for puncture (see “General Technique for Cyst Sclerosis”). B, A catheter was inserted for slow and complete drainage. Note the redundant catheter in the cyst and reduced size of cyst. After fluid is removed, alcohol is injected. C, Fluid recovered is yellow, resembling urine, but the sample was sent for cytologic evaluation.

all imaging and clinical data. Medication history is especially important to determine if any anticoagulative drugs are taken (either prescribed or OTC). Laboratory studies, including CBC and coagulation, are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam (Versed) or fentanyl (Sublimaze) or both as needed. Local anesthesia is administered as described earlier; either deep sedation or general anesthesia can be used. Use of sedation with

with automated devices and local anesthetic and sedation administration are the same as used for other sites. The general method for selection of target, entrance site, and trajectory is used as described earlier. Because a sclerosing agent is to be injected, it is wise to avoid traversing a blood vessel because during injection of the sclerosing agent, some fluid could “track” along the needle into the cyst. This fluid could damage the organ and produce local infarction. The other critical point for selection of the trajectory is to choose a pathway that includes a 2-cm cuff of normal liver. During and after the procedure, this cuff of tissue reduces the chance of leakage of the cyst contents or sclerosing agent into the adjacent peritoneum during or after the procedure. Instruments used are a plastic sheath needle, small-caliber angiographic catheter, and angiographic wire. Choice of sclerosing agent is left to the operator, but we prefer ethyl alcohol because it has been widely used for many purposes, is easily metabolized if absorbed, and has low systemic toxicity. The volume we use is about 45 mL as a flush (see “Insertion of Needle and Catheter and Alcohol Treatment”). We do not use the large volumes described by others. Our approach differs from that of other authors. Other authors advocate evacuation of a portion of the fluid and injecting alcohol equivalent to 25% of the volume. If a single treatment does not work, the procedure is repeated on multiple occasions as needed. We object to this method for two reasons. First, the alcohol is diluted significantly; this limits its local effect of destroying the endothelial lining. Second, because large volumes of alcohol may be needed in large cysts, there is the possibility of administering a large systemic dose. We prefer to insert a small-caliber catheter to remove most of the fluid and inject smaller amounts of alcohol, as noted subsequently. Insertion of needle and catheter and alcohol treatment.  Our approach begins with insertion of a small 5F pigtail catheter as a conduit (see Fig. 67-137). Before any further manipulation of the catheter in the cavity, a contrast injection is performed to ensure that there is no communication with the biliary or urinary system. If there is, the procedure must be aborted because injection of alcohol would sclerose the system. We evacuate under suction as much of the fluid as possible and leave the catheter to gravity drainage for several hours while the patient remains in the department. After as much fluid as possible has been removed, the walls of the cyst are apposed or in close proximity so that a smaller volume of alcohol may be used to sclerose the lining of the cyst wall. Small amounts of alcohol are injected as follows. We inject 15 mL of alcohol and leave it in place for 10 minutes. During this time, the patient is moved to several positions to permit the alcohol to gravitate to the sides and anterior surface of the cyst. After 10 minutes, attempts are made to aspirate the alcohol. A second instillation of 15 mL of alcohol is made, and the process is repeated. After 10 minutes, we attempt to recover the alcohol, and then inject the last 15 mL of alcohol. If removal of alcohol is possible, a total of 45 mL is used but we never leave in more than 30 mL. The catheter is removed and the patient is left to recover for 1 to 2 hours. During this time the patient’s vital signs and general well-being are monitored. The alcohol that is injected shows a remarkable change. The returned alcohol becomes very cloudy, turbid, and pink owing to the sloughing of cells from the cyst wall into the fluid.

Postprocedure Care.  Postprocedure follow-up is straightforward. There is no need for repeat studies if the patient remains

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2493

B FIG 67-137  Sclerosis of a hepatic cyst. A, CT scan shows two large cysts that are separated by a septum (arrow). The needle was inserted to permit puncture of both cysts. The path was chosen to provide a cuff of normal liver to prevent the spillage of cyst fluid or alcohol when it is injected at a later time. B, Radiograph shows insertion of a wire and catheter into the cyst cavity. The cyst is left to drain for several hours before instillation of alcohol (see text).

asymptomatic, but repeat scans are typically reassuring to the patient and clinical physicians (Fig. 67-138). Cysts that have been sclerosed shrink gradually over weeks and months. If the cyst returns, the patient returns with symptoms, and the procedure can be repeated with slightly larger amounts of alcohol if desired. We believe this conservative method is better than the large-volume methods used by other authors. Such large volumes invariably result in some intoxication of the patient and can put other portions of the liver at risk because of the local toxicity of the alcohol. When readers go to a conference, they will be amazed at how cases being presented using the high-volume method show destruction of other liver lobes; lecturers seldom make note of that in their presentations. Another approach has been offered by Cellier and colleagues.61 In a group of five patients, the authors performed the following technique. The cyst was punctured and half of the cyst contents were aspirated. A single dose of 100 to 500 mg of minocycline hydrochloride was injected and left in place. At follow-up, four of the five cysts had resolved.

PANCREATIC PSEUDOCYST Pseudocysts of the pancreas are collections of necrotic debris, inflammatory material, and digestive enzymes that require treatment because they are associated with numerous complications. Although surgical treatment has been the standard of the past, percutaneous methods developed in the past decade have a major role in management.

Indications The presence of a cyst in the pancreas warrants at least a diagnostic aspiration to confirm the diagnosis of pseudocyst and exclude the possibility of a cystic tumor. Evaluation of the aspirated fluid confirms a pseudocyst by virtue of the elevated amylase within the fluid. Cystic tumors can be diagnosed by biomarkers in the fluid. Treatment of a pseudocyst should be considered to prevent the occurrence of recognized associated complications such as bleeding, infection, or rupture of the pseudocyst. The decision to treat a patient’s

pseudocyst must be made in conjunction with the medical and surgical team to ensure a collaborative approach to treatment. Numerous selection criteria that have been defined in the surgical literature in recent years seem to be good indicators of success (see “Success Rates for Percutaneous and Surgical Methods”). Another facet of patient selection depends on an appropriate anatomic appearance permitting proper access and instrumentation. One must choose to perform aspiration or catheter drainage.

General Technique for Pancreatic Pseudocyst Drainage

Preprocedure History, Review of Images, Planning, and Informed Consent.  Preprocedure steps include physical examination, clinical and medication history, and review of all imaging and clinical data. Medication history is especially important to determine if any anticoagulative drugs are taken (either prescribed or OTC). Laboratory studies, including CBC and coagulation, are reviewed. The patient’s history is reviewed relative to any history of alcohol abuse, pancreatitis, or surgical interventions related to the pancreas. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam or fentanyl or both as needed. Local anesthesia is administered as described earlier; deep sedation or general anesthesia is used. Use of sedation with general anesthesia may lessen the depth of general anesthesia required, which is beneficial to the patient. Planning the procedure from the diagnostic scans taken before the procedure is important. A thorough understanding of the nature of the anatomy is important for planning. In contrast to biopsies, where the procedure is straightforward, drainage may be complicated because of the complexity of the inflammatory process and involvement of numerous organs and anatomic spaces (see Chapter 49). The modality chosen is usually CT because of the remarkable complexity of pancreatitis and associated pseudocysts requiring clear delineation of all anatomic detail. CT is the best modality for visualization

2494

PART III  Image-Guided Procedures

C

A

B

C FIG 67-138  A to C, Three images show gradual reduction in size of a cyst (C) that has been sclerosed with alcohol. The reduction in the cyst occurs over many weeks.

of the anatomy and instruments so that uninvolved structures can be avoided. Avoidance of small bowel and colon is important because secondary contamination can occur. With CT dynamic contrast enhancement scanning, pseudoaneurysms can be diagnosed accurately and avoided. Life-threatening hemorrhage has occurred when these structures were inappropriately penetrated. Morgan and associates advocated the use of MRI to evaluate fluid collections to assess the presence of large pieces of necrotic debris within fluid collections to determine if percutaneous methods are suitable.359 Such collections were judged to be not drainable if the material measured more than 1 cm.

Procedure Steps.  Patient vital signs and pulse oximeter monitoring with automated devices and local anesthetic and sedation are the same as used for other sites. Determination of entrance site, target, and trajectory are critical factors to ensure success and to avoid complications. Common sense dictates that if external drainage is being used, structures such as the spleen, vessels (splenic artery, splenic vein, superior mesenteric vessels, varices), colon, liver, and others should be avoided. Although some authors have advocated that penetration of such structures is not relevant, these techniques have higher risks. If a transgastric approach is being used, one should ensure that the pathway is clear and there are no gastric varices on the wall. Dynamic bolus study with IV contrast material is mandatory. Pancreatitis and pseudocysts are known to be associated with pseudoaneurysms (see Fig. 67-103). We have had patients referred for cyst puncture who actually had large aneurysms; damage to these could be

catastrophic. Varices can be variable, with different sizes of vessels and unusual locations. Preparation of the entrance site is routine. The skin site should always be examined and palpated to look for possible varices in the subcutaneous tissues. Patients with pancreatitis are likely to have problems with cirrhosis and likely to have varices. Penetration of a large skin vessel could be dangerous. Introduction of the needle is routine. If only an aspiration is performed, only a sample is obtained. If the intention is to try a palliative aspiration, the entire contents of the pseudocyst are aspirated. If drainage is being performed, the angiographic wire and catheter are the same as described later under “Abscess Aspirations and Drainages.” Catheter choice depends on whether an external drainage or transgastric method is being used. For external drainage, the catheter size is chosen based on the thickness of the aspirated fluid. The largestcaliber pigtail catheter should be used if the patient can tolerate the insertion and if the fluid space would easily accommodate the catheter. If the material is thick, it may be necessary to insert multiple tubes and institute periodic lavage through the catheters. Placement of the catheter should be throughout the length of the cavity if an external drainage is being used (Fig. 67-139A and B). This action prevents formation of an internal loculation caused by internal septation. If a transgastric approach is being used, the selected catheter must be less than the internal diameter of the gastrostomy tube. A sample of fluid should always be obtained (see Fig. 67-139C) and sent to the laboratory for evaluation. In cases of acute inflammatory pseudocyst, the fluid is hemorrhage, but if the pseudocyst is chronic,

CHAPTER 67  Image-Guided Aspirations and Biopsies

2495

A B

C FIG 67-139  A, CT-guided external drainage of a pseudocyst. CT scan shows a fluid collection in an anterior pancreatic inflammatory mass. B, CT scan shows catheter coiled within the pseudocyst. C, Fluid obtained from a pseudocyst typically appears hemorrhagic.

the fluid may appear clear as with cysts or lymphoceles (see Figs. 67-134C and 67-136C). Determination of amylase evaluation, biomarkers CEA and CA 19-9, and bacterial cultures is done if appropriate. Pseudocysts have elevated amylase levels. Tumors have elevated biomarkers.

Postprocedure Care.  Postprocedure care is routine. If only an aspiration has been performed, routine observation of the patient over a 4-hour period is appropriate. If the aspirated material appeared purulent, appropriate antibiotics should be administered after consultation with the clinical services. If a catheter has been inserted, it should be attached to a drainage bag and the drainage amounts monitored and observed for any changes. Change to a purulent nature necessitates administration of systemic antibiotics to be determined by consultation with infectious disease specialists. If the drainage turns very bloody, evaluations for aneurysms or pseudoaneurysms should be initiated by means of CT angiography, MRI, or standard angiography if embolization is to be considered. The duration of drainage varies, ranging from days to many weeks. With external drainage, fistula formation is always possible, and if it occurs the course may be prolonged. To lessen pancreatic secretion and flow during percutaneous drainage, the hormone octreotide has been used by some authors. D’Agostino and associates used octreotide in recommended doses, and it seemed to be beneficial in four of eight cases.87 The catheter may be removed whenever the drainage ceases and a repeat CT scan shows resolution of the pseudocyst. If the drainage does not cease after several weeks, a sinogram should be performed to evaluate for ductal or enteric communication or fistula. These can be treated with surgical or percutaneous methods as needed.

Factors for Success

Selection Criteria for Treatment.  Selection criteria for different treatment methods have been formulated based on anatomic factors that have been defined by surgeons. A classification of pseudocysts  into groups has been proposed by D’Egidio and Schein96 and supported in concept and studies by other authors, including Nealon and Walser374-376 and Zhang and Zheng.531 Group I consists of acute postnecrotic pseudocysts with normal pancreatic duct anatomy and rarely duct communication. This group can be cured in most cases by percutaneous drainage. Group II consists of postnecrotic pancreatitis with chronic pancreatitis. Percutaneous drainage may be effective, but the treatment must be prolonged. Group III consists of chronic “retention” pseudocysts associated with a diseased pancreatic duct, ductal strictures, and duct-pseudocyst communication. Operative treatment of the ductal disease resolves the pseudocyst. Timing of treatment is important. Approximately 6 weeks should elapse after detection of pseudocyst before any procedure, because a high percentage (≥25%) of pseudocysts may disappear spontaneously, being absorbed or spontaneously drained internally. Sanfey and colleagues reported that in their series of 97 cases of inflammatory pseudocysts, 32% resolved spontaneously, not requiring intervention.432 It is reasonable to follow patients with pseudocysts to see if they resolve spontaneously, but if not, intervention should be considered. If pseudocysts go untreated longer than 13 weeks, some authors have reported a high incidence of infection, rupture, or bleeding, all of which can be catastrophic. For these reasons, the optimal time to drain these is after 6 weeks and before 13 weeks to prevent such complications. Warshaw and Rattner recommend a similar timetable.507 Poorly defined inflammatory fluid collections representing phlegmons or edema collections should be avoided.169 If drainage of these

2496

PART III  Image-Guided Procedures

poorly walled off collections is attempted, drainage is typically unsuccessful and secondary infection is likely. Proper patient selection using different parameters can predict likely success. Fluid and ductal traits predict success. In the early period of technique development, treatment of almost every pseudocyst was attempted, but over the years, predictive factors have evolved to select patients for the various treatments. Early authors reported that the amylase content and communication with the pancreatic duct could predict success. Duvnjak and colleagues reported that pseudocysts with a high amylase content were unlikely to respond to percutaneous methods.110 These findings may be an indirect observation based on communication with the pancreatic duct; pseudocysts having direct communication with the duct are more likely to have high levels of amylase. Barthet and coworkers believed that percutaneous methods were less likely to be successful if there was communication with  the pancreatic duct,25 but other authors have not documented their opinion.

Success Rates for Percutaneous and Surgical Methods.  The success rates for the various approaches vary and likely depend on the selection of patients. The success of external drainage, percutaneous drainage, and internal drainage varies accordingly. Early reports on the various methods are inconsistent, and it is likely that results depend on selection of patients and individual idiosyncrasies of different institutions and practices. One of the largest series reported in the surgical literature was by Heider and associates.213 In their retrospective study of 173 patients, percutaneous methods were inferior to surgical treatment. They reported successful treatment in only 42% for percutaneous treatment and 88% for surgical treatment. The best example of the confusion surrounding this controversy can be appreciated by noting two articles from the same institution in the same year by Adams and Anderson2 and Criado and associates.81 The two groups from the same institution seem to contradict one another in their support of surgical or radiologic methods. Zhang and Zheng,531 using the classification of D’Egidio and Schein,96 reported the following results, which generally indicated expected cure rates. Group I, consisting of acute postnecrotic pseudocysts without chronic duct changes, would have percutaneous drainage success in 82%, whereas external drainage or internal drainage would succeed in 50%. Group II, consisting of postnecrotic pancreatitis with chronic pancreatitis, would have percutaneous or external drainage success in only 44% and internal drainage success in 92%. Group III, consisting of chronic pancreatitis and diseased ducts, would have internal drainage or resection cure rates of 80%.

Complications Numerous complications have been reported, including hemorrhage, infection, fistula, death, and inadvertent damage to adjacent structures. These complications vary according to the institutions, technique used, anatomic approach, and catheter size. The factors are so variable it  is difficult to ascertain individual factors, but meticulous technique with adherence to technical principles is important. Transvascular approaches as recommended by Gupta and colleagues172 should be avoided to minimize bleeding. The most common complications are bleeding, sepsis, or abscess formation. The literature has many articles recounting the complications associated with these procedures,121 but for the sake of brevity, we use the data from a more recent article in the surgical literature by Nealon and Walser.374 Recognizing the authors’ apparent bias against nonsurgical treatment, their data would represent the worst-case scenario. The authors provide short-term and long-term follow-up. A defined episode of clinical sepsis was identified in 72 of 79 patients (91%) and was the most common complication. Hemorrhage requiring transfusion was

identified in 16 of the 79 patients (20%), clinical shock in 51 of the 79 (65%), renal failure in 16 of the 79 (20%), and ventilator support for longer than 24 hours in 19 of the 79 (24%). A persistent pancreatic fistula occurred in 66 of the 79 patients (84%); mean duration was 61.4 ± 9.6 days. Among the 100 operated patients, 69 complications occurred in 6 patients (6%). A second large study by Heider and coworkers included 173 patients and reported that percutaneous methods were inferior to surgical treatment.213 Mortality for percutaneous procedures was stated as 16%, whereas surgical mortality was 0%. Complication rates were 64% for percutaneous treatment and 27% for surgical treatment. Pancreatic fistula may result after tube drainage of a pseudocyst. Of the several approaches that might be used, only one has been reported. Some authors have used the hormone somatostatin to decrease the exocrine output of the gland.87 The results have varied, however, and the medication is expensive. We have used “fibrin glue” successfully to seal such a pathway from a pseudocyst or the pancreatic duct; we have not published the data or technique, but an experienced interventionalist might consider this approach and improvise it. Singh reported biliary and pancreatic duct fistulas.453 The success with biliary communication was 93% (100% of cystic duct fistulas), and pancreatic communication was 67% in a group of 57 patients. The biliary communications were cystic duct, common duct, and right and left ducts. The three failures in the biliary group were associated with common duct communication. In the pancreatic duct group, the site of communication was the main duct in 14 of 15 and the side duct in 1 of 15. The presence of a fistula increases the catheter drainage time required for healing. Freeny and colleagues reported that the drainage time from collections without and with percutaneous drainage was  29 days and 96 days, respectively.131,132 Singh reported that the drainage time of pancreatic and biliary fistulas was 25 days and 20 days, respectively.453 In our opinion, the merits of various methods depend on the individual physicians at each site and the selection criteria for the methods. Each institution tends to have its own idiosyncrasies, and the skill levels of each group probably differ from one center to the next. Depending on the local talent and experience, a closely working team of physicians can establish which patients are appropriate for surgery and which for radiology.

Different Methods of Treatment The above-described pseudocyst technique can be modified depending on the chosen method. Four methods of treatment have been reported, and after many years of development, the advantages and disadvantages of each are well known. These methods in chronology of development are aspiration, external drainage, and internal drainage. None of the reports in the radiology literature reveals the deficiencies of the methods. An impression is given that each method has solved the problem. In our extensive experience with the different treatments of pseudocysts, we have formed personal opinions of their various roles, and we have found each method to have some merit but also some shortcomings. With each succeeding method the results have been better, but a subset of patients remains who are not “cured” by percutaneous methods and who require surgical treatment. Each of the methods is described, and the benefits and disadvantages are noted from our perspective.

Aspiration.  Simple aspiration of a pseudocyst seems to work in slightly less than 25% of cases, but it is advantageous because of its simplicity. This procedure can be performed with the standard methods described previously and with little or no risk (assuming that one avoids bowel loops and vascular structures). The procedure is so simple

CHAPTER 67  Image-Guided Aspirations and Biopsies that most can be performed on an outpatient basis, as noted earlier. Some workers have recommended that several aspirations may be required over weeks or months. Pseudocysts most likely to be cured by this method are those occurring after a localized and minor leakage from the pancreas after a surgical procedure.

External Drainage.  External drainage is analogous to the external drainage performed for abscesses. With this method, various structures are avoided and a catheter that drains to the external environment is put in place. Some authors have reported good results using this method, but in our experience a fairly high percentage persists or recurs. The risk of a permanent fistula or secondary infection also exists. Although the literature on this problem does not indicate significant recurrences, we have found the recurrence rate to be similar to the aspiration method. Two subsets of pseudocysts have responded well to this method or the aspiration method of therapy. Well-defined pseudocysts that are secondarily injected are well suited for percutaneous external drainage.499 In patients who have a pseudocyst secondary to external or surgical or traumatic injury,26 the pseudocyst can be treated easily by aspiration or external drainage. In most of the cases we have treated and in the cases reported in the literature, such drainages are uniformly successful without any sequela or recurrence.

Transgastric Drainage.  The most promising percutaneous method for draining pseudocysts is a transgastric approach as described by Nunez and colleagues,383 Bilbao and coworkers,35 and Ho and Taylor.221 With this method, the trajectory chosen includes the walls of the stomach and the pseudocyst in the lesser sac. The catheter provides internal and external drainage until such time as the pseudocyst resolves. In the largest series by Nunez and colleagues, 8 pseudocysts in 7 patients were successfully drained.383 The length of drainage was 3 weeks in 6 cases and 6 weeks in 2 cases. No recurrences were noted, but the length of follow-up was not mentioned.

Modified Drainage or Cystogastrostomy.  Instead of the simple transgastric approach, we have developed a modification using access through an established percutaneous endoscopic gastrostomy (PEG) tube (Figs. 67-140 and 67-141). The PEG lessens the concerns about possible gastric spillage into the peritoneum. Our modification of this method eliminates this potential problem and has been well received

A

2497

by clinical physicians. Since we have published this method in our prior editions, other authors16,80,90 have independently developed a similar technique described as cystogastrostomy. With this method, a combined procedure is performed in collaboration with the endoscopist, preferably a surgeon. The procedure is contingent on the endoscopist’s inserting a PEG. The tip of the PEG should be cut off to permit passage of the catheters used more easily. The PEG secures the anterior stomach wall to the abdominal wall and minimizes the possibility of accidental gastric content leakage. This tube serves as a conduit into the stomach and farther into the pseudocysts (see Fig. 67-141). A plastic sheathed needle is inserted via the PEG through the posterior stomach wall into the pseudocyst. Execution of the chosen trajectory into the pseudocyst is easy because the gastrostomy site in the skin serves as a fulcrum, making angulation of the needle through  the posterior stomach wall adjacent to the pseudocyst simple. When the pseudocyst has been successfully punctured, a catheter is inserted using Seldinger technique. The pathway is dilated and the chosen catheter inserted. Because an 18F or 20F gastrostomy tube is typically inserted during the PEG procedures, angiographic catheters 9F or 10F can be used to penetrate the pseudocyst. The outcome of the procedure is a tube traversing the abdominal wall, gastric wall, and pseudocyst wall. Patients can begin eating the following day and require no special follow-up. It has been our experience that leaving the tube in for about 6 to 8 weeks is sufficient. Our experience and the surgical experience with the same procedure are good. In contrast to endoscopic cystostomies, in which double-J tubes are difficult to replace should occlusion occur, exchanging the catheter from the external site is easy. The transgastric method has three advantages. First, residual or leaking pancreatic juice can exit into the stomach, preventing a fistula. Second, replacement of drainage tubes is easy. Third, the collaboration with the endoscopists keeps political issues in hand.

ABSCESS ASPIRATIONS AND DRAINAGES Percutaneous aspiration and percutaneous drainages guided by CT represent the most significant advance in the care of acutely ill patients since the development of surgical drainage and the discovery of antibiotics. Since the first CT-guided aspiration and drainage procedure (Fig. 67-142), which was performed in 1975 and reported in 1976,177

B FIG 67-140  Percutaneous transgastric drainage of pseudocyst is a viable option for drainage. A, CT scan shows the needle trajectory through the stomach into a large retrogastric pseudocyst. B, Later CT scan with a catheter in place shows remarkable resolution of the pseudocyst.

2498

PART III  Image-Guided Procedures

C P

A

B S

C FIG 67-141  A, Scan shows pseudocyst (C) anterior to the pancreas (P). Note the percutaneous endoscopic gastrostomy (arrow) anteriorly in the stomach. B, CT scan shows the needle (curved arrow) penetrating the end of the tube into the cyst, and the wire (straight arrow) looped in the pseudocyst. C, Radiograph shows a large tube traversing the stomach and located in the pseudocyst; contrast medium defines the stomach (S).

B

A

FIG 67-142  A, First CT-guided drainage of an abscess (arrows) performed in 1975 by the author. B, Plain radiograph of the abdomen shows a drainage tube in place (arrow). The patient recovered uneventfully. (B, From Haaga JR: Definitive treatment of a large pyogenic liver abscess with CT guidance. Cleve Clin Q 43:85–88, 1976.)

use of this technique has expanded rapidly. This expansion has occurred because of the increased number of scanners available, the excellent accuracy of CT for the detection of any fluid collection, the high accuracy for guidance, and the popularization of the technique by enthusiasts such as Haaga,177,198,199 Gerzof,145,146 Mueller,364 and Van Sonnenberg498-500 and their colleagues.

Diagnostic Aspirations Because patients with infectious fluid collections may or may not manifest the typical signs of infection, we believe that if there is even the slightest clinical concern about an abscess, a diagnostic aspiration should be performed. This aggressive diagnostic approach is justified

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

C

B

D

2499

FIG 67-143  Infected fluid samples can have different appearances. The visual appearance cannot be used to determine whether a sample is infected. The serum-colored infected sample (A) is turbid and cloudy. The dark gray sample (B) is an indolent chronic abscess. The hemorrhagic sample (C) represents infected blood. The turbid red sample (D) is hemorrhagic purulent material. Except for cultures and stains, on examination of a sample at the time of the procedure, the malodorous smell of a sample is actually the most accurate predictor of infection.

because with careful technique, including avoidance of bowel, a sample can be successfully recovered in almost 100% of cases with little or no risk (Fig. 67-143). In our experience with innumerable cases of CT guidance over the years, to our knowledge we have not had any case of secondary infection. To avoid local contamination with bacteria, the diagnostic sampling of fluid collections requires the same planning as that of all of the other lesions we have discussed, but several factors must be kept in mind. The material being obtained may contain viable pathogenic material and can disseminate infection if it is inappropriately spilled into or contaminates sterile spaces. Even though modern potent antibiotics can eradicate any minor spillage of purulent material, great care should be taken not to contaminate sterile areas. Patients may develop serious peritonitis or bacteremia if the inoculum is larger or the organism virulent (or if the patient is immunocompromised by any cause). A question occasionally arises concerning the interpretation of culture results—that is, how culture results should be interpreted when only a few colonies grow. In our opinion, any number of colonies involving organisms other than Staphylococcus species should be considered an infection at least warranting antibiotic therapy based on the culture sensitivities. Depending on the number of bacteria within a fluid collection or the presence of antibiotics, there is the possibility

for a sampling error that can result in a negative culture or possibly growth of a few pathogens. Finally, avoidance of GI loops is important when fluid is aspirated for culture. First, a sample taken through bowel may provide spurious culture results. Some workers claim that the presence of mixed flora and absence of a “pure” culture of a single organism can distinguish an abscess from GI colonization. As has been well documented in the radiology literature, however, many abscesses are caused by a GI fistula that contains many types of organisms. Second, the function of humoral cellular defenses is lessened in a fluid space so that clearing of bacteria by phagocytes is impaired, and a secondary infection might occur if a small inoculum is introduced into a fluid collection.

Mycotic Aneurysms or Infected Vascular Prostheses Because of the difficult treatment required for infected vessels or synthetic aortic prostheses, any patient with such a suspected infection should have a diagnostic aspiration as early as possible. In the infected prostheses and their primary mycotic infections that we have encountered, the cultures have been positive and reliable. In each of these cases, the culture results were crucial and helped expedite treatment of the problem. With prompt treatment, most of these patients survived the surgical therapy. In another group of 10 patients with uninfected

2500

PART III  Image-Guided Procedures Percutaneous Abscess Drainage Before CT, percutaneous abscess drainage (PAD) was occasionally performed when abscesses were so large they could be palpated, were visible on fluoroscopy by displacement of bowel, or were superficially located and visible by US. With the advent of CT assistance, PAD has become a routine procedure. CT has been especially useful because it has a high accuracy for detecting abscesses and allows accurate determination of the anatomic location and extent of the lesion. Because CT is not impeded by gas or bone, any involvement of peritoneal spaces or anatomic areas can be delineated clearly. This information is necessary if one is to determine whether an abscess can be drained by a percutaneous method. Although there was some controversy in the early literature concerning the merits of different guidance systems for drainage, there now seems to be a consensus that CT guidance is the best method available.

Perspectives on Abscess Drainages

FIG 67-144  CT aspiration of periaortic fluid collection (arrow). A posterior approach avoids bowel loops and other intervening structures. Clear delineation of the needle and aorta makes this procedure a viable alternative to surgical intervention.

grafts, the results were negative. No complications have been associated with the needle procedures (Figs. 67-144 and 67-145). Such cases should be approached from the flank, and a pathway should be chosen to avoid the kidney and major vessels. In most instances a clearly defined fluid collection is seen either adjacent to the graft or between the wall of the aneurysm and the graft. As with other procedures performed from this approach, we try to avoid the psoas muscle, the spinal erector muscles, and the spinal elements to prevent patient discomfort. If there is any question about the location of the vessel or prosthetic wall, a dynamic bolus CT scan should be performed (see Fig. 67-144). Dynamic CT is highly effective to show a vascular problem, even better than Doppler US and angiography. In one patient with a clotted mycotic aneurysm, only CT correctly showed the vascular nature of the lesion. Some operators have raised concerns about possible injury to an infected vessel, but in our experience this is not a problem. In the more than 20 suspected cases we have aspirated, there have been no complications. The need for aggressive procedures to expedite patient care is clear. The mortality rate of untreated prosthesis infection is almost 100%. Corrective treatment and surgery can be effective only if initiated early. In some instances, percutaneous drainage of infected prostheses may be effective if the infection involves only the midportion and not the anastomotic ends. Other authors, such as Rossi and colleagues424 and Lambiase and coworkers,280 also have proposed an aggressive approach to diagnosis and management of these difficult infections. These authors also have advocated percutaneous aspiration for diagnosis of infectious agents and palliating the infection with preoperative catheter drainage. We have effectively cured a patient with graft infection using local catheter drainage (see Fig. 67-145).

With a broad perspective toward abscess drainage, the intent of interventional percutaneous procedures is at least to duplicate the excellent treatment results established by surgeons. The benefit of percutaneous procedures lies in their simplicity, lower complication rates, and presumably lower overall expense. Publications in the surgical literature have illuminated some of the factors that relate to the outcome of percutaneous and surgical procedures. The benefits of improved detection and localization of abscess have been documented by Deveney and coworkers, who showed conclusively that the improved localization and intensive care methods have resulted in a greater percentage of drainage success and decreased mortality resulting from organ failure.100 These authors studied two 5-year periods (1973-1978 and 1981-1986) in a veteran population. The population included patients treated either surgically or percutaneously. The mortality of the two groups was 39% from 1973 to 1978 and 21% from 1981 to 1986. The successful initial drainage was 55% in the earlier group compared with 74% in the later group. Organ failure rate decreased from 52% to 23%. The surgical success rate was 76% compared with the percutaneous success rate of 72%. The authors concluded that accurate localization and early drainage are the most important factors and that there is no difference in outcome between the surgical and percutaneous methods. Overall patient status predicts the outcome of PAD, as has been documented by Levison and Zeigler.294 These authors used the APACHE (Acute Physiology, Age, and Chronic Health Evaluation) II system to grade patients and follow their outcome. The final results were that mortality was directly related to the APACHE score. Of patients with a score less than 15, the mortality was 1.7%; when the score was greater than 15, mortality was 78%. There was no significant statistical difference in outcome between percutaneous and surgical drainages. (One is cautioned not to quote this finding out of context because the authors subjectively maintain that surgical drainage was better by absolute numbers. The finding was not statistically significant, however.) The status of the patient’s immune system clearly affects the effectiveness of PAD. Lambiase and colleagues documented how crucial this factor is as it relates to abscess drainage.280 These authors considered the following states as immunocompromised: absolute neutropenia, infection with human immunodeficiency virus, cancer with distant spread, chemotherapy, radiation therapy, lymphoproliferative disorders, diabetes requiring treatment, long-term renal dialysis, splenectomy, and severe alcoholism. They found that with PAD the overall cure rate for immunocompromised patients was 53.1% compared with the normal rate of 72.6% for normal immunocompetent patients.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2501

B

C FIG 67-145  A, Percutaneous drainage of a fluid collection around the iliac limb of a prosthetic graft. Note the fluid collection (arrow) confined by the aneurysm. B, CT scan shows a catheter placed within the fluid collection. Note the opaque contrast medium around the patent graft (arrow) and the air cavity anteriorly within the cavity. C, Follow-up CT scan shows complete resolution of the fluid collection. A small residual amount of gas is left after removal of the catheter. Four months after drainage, the patient continues to be completely well. Urokinase irrigations were used to enhance drainage (see text).

Patient Selection In recent years, the indication for percutaneous methods has expanded significantly. In contrast to most new techniques and procedures, in which there is an initial period of enthusiasm followed by either a moderation or a restriction of the method, percutaneous drainage guided by CT has continued to expand its role. The results of percutaneous procedures have been so positive and so widely accepted that the indications and applications have been defined. To maintain a proper clinical perspective, it is important that the appropriate patients be selected for curative or palliative PAD procedures and for surgical procedures. Simplistically, one could try PAD for palliation in every patient, but it is not logical to expand its use for patients with certain surgical problems. On a practical level, although the role and the indications for percutaneous procedures have expanded greatly, there are still clear indications for surgical drainages. The imaging appearance of the abscess indicates how easily an abscess might be drained. Abscesses that are of low density and clearly defined without septations or not involving numerous peritoneal spaces are easily drained. The results with septated abscess have varied;

some drain well and others do not. The most plausible explanation is that fibrin septa in such cavities may be complete or incomplete. When they are complete, drainage is poor, and when they communicate and are incomplete, curative drainage can occur. We have studied the use of fibrinolytic agents in the former cases and have found them effective for improving drainage (see “Fibrinolytic Agents,” later). Also, the multiplicity of abscesses is no longer considered a contraindication to percutaneous drainage, but multiple catheters in each site may be required for optimal drainage. For the anatomic pathways, we still prefer a clear pathway into the abscess cavity without traversing an uninvolved space or organ. I believe that a routine penetration through an uninvolved organ is inappropriate because of possible seeding of pyogenic organisms into the organ. Although with the potent antibiotics of today most such procedures seem feasible, I believe that when larger series are collected, the success and complication rate will favor avoidance of uninvolved organs and spaces. Viscosity or thickness of the purulent material in the abscess affects drainage. If the entire cavity can be evacuated easily after aspiration, percutaneous drainage probably would work. If the material appears

2502

PART III  Image-Guided Procedures

to be viscous, techniques should be adjusted appropriately (i.e., special attention given to catheter size, catheter placement, and use of an appropriate irrigant) (see “Clinical Management of Catheters”). Using PAD for palliation is most appropriate. Adequate palliative drainage reduces the “toxic load” resulting from drainage of purulent material and usually produces considerable improvement in such patients. Some patients are cured, but even patients who are not would benefit from an improved physiologic state that makes them better surgical candidates. The appropriateness of percutaneous drainages for abscess associated with entities requiring surgical treatment is an area of significant controversy. One must weigh the palliative benefits of PAD against the final curative benefits of surgery. Similar views are shared by Welch, who noted in an editorial509: Unfortunately, abscesses vary so greatly in many ways that studies with proper, randomized controls would be very difficult to establish. It is far more likely that further experience will lead to greater concurrence between radiologists and surgeons. Thus, the letter by Haaga and colleagues198 outlines conclusions that appear to be closely in accord with our own studies in the Massachusetts General Hospital. Since Haaga and his co-workers originated the radiologic method, their opinion must be given great respect.

Indications for Surgery Some patients are not good candidates for percutaneous procedures and are better suited for surgical drainage. This group consists of patients with fungal infections, infected hematomas, echinococcal disease, surgical disease requiring correction, most pancreatic abscesses, and failed percutaneous drainages. Patients with fungal abscesses require surgical drainage because there is extensive tissue invasion, necrosis, and mycotic plaque formation with the wall of the cavity. Our experience with several such cases and reports in the literature187,281 show that surgical drainage or resection is best for adequate débridement. Infected hematomas do not drain well because of the extensive amount of fibrin deposition and the protective effects of fibrin on bacteria. As shown from our early experience and from reports in the literature, such hematomas have uniformly failed by percutaneous drainage. In our more recent experience, fibrinolytic agents used as a catheter irrigant can improve the drainage of infected hematomas. Echinococcal disease requires a special approach to prevent spread of the daughter cysts into the peritoneum or other areas. The issues related to anaphylaxis are essentially minimal with appropriate technique, and many authors have now reported on the technique of percutaneous drainage (see “Echinococcal Drainage,” later). Unless the interventionalist practices in an endemic area, however, it is unlikely that he or she would develop sufficient expertise to manage echinococcal disease well. In nonendemic areas, echinococcal disease should be treated by the surgical method. The specific approach for percutaneous treatment is discussed at the end of the section “Echinococcal Drainage.” Finally, for patients who have had a failed percutaneous procedure, surgical drainage is appropriate. Because of the difference in philosophy among surgeons, radiologists, and institutions, the definition of a failed procedure can be determined only at the local level and not in this text.

Treatment of Abscesses by Aspiration and Antibiotics Alone In recent years a new approach for abscesses has been introduced that is avant-garde and may have a role in the future. These techniques have

FIG 67-146  Transplant patient with a small abscess in right lobe, which is too small for catheter insertion. Note the needle inserted into small abscess with an air-fluid level.

been based on single or multiple aspiration procedures on abscesses in lieu of drainage tube placement (Figs. 67-146 and 67-147). Because of the proven record of percutaneous catheter drainage, we have been slow to adopt this method as a standard approach, but we do use it when two factors exist: (1) lack of safe access to permit catheter insertion and (2) small size preventing catheter placement. Techniques have been published by McFadzien and Wong,340 Giorgio and colleagues,149 Saluja and associates,430 Ch Yu and coworkers,62 and Wroblicka and Kuligowska.517 The specific details of each author differed slightly, but most reported single or multiple aspirations of the abscess cavity while systemic antibiotics were given. Wroblicka and Kuligowska reported irrigating the cavities with saline.517 McFadzien and Wong340 and Giorgio and colleagues149 also injected intracavitary antibiotics. All authors claimed high success rates. The authors also reported numerous failures and complications, including transient sepsis, death by sepsis, failure because of enteric fistula, and peritonitis. This technique was heralded in an editorial comment by Miller and associates, who stated that this might be the biggest advance for abscess treatment in 80 years.350 This is probably overstated, but this method clearly has a role in the modern treatment of abscess. We believe the greatest advance was surely the use of CT and drainage methods, which made percutaneous treatment a routine daily method for abscess treatment. When the first such diagnoses and procedures were performed in 1975, patients were dying from undiagnosed and ineffectively treated abscesses. When CT methods were developed, abscesses became a radiologic and not a surgical disease.

General Technique for Aspiration Treatment Selection of patients for aspiration treatment is based on the characteristics of the abscess. We use this technique when an abscess is too small or inaccessible for catheter drainage. Preprocedure steps are identical to other procedures relative to review of history, medications, imaging, and patient consent. Laboratory values and sedation are as noted in prior sections. Needle aspiration of the abscess is performed using similar guidelines as described for FNA of lesions, but one must use the 18-gauge needle. Purulent material may be thick, and a larger needle is more reliable for recovering a sample. On recovery of the purulent material, we perform (1) aspiration of purulent contents, (2) lavage with saline, and (3) injection of a small

CHAPTER 67  Image-Guided Aspirations and Biopsies

A A

C

2503

B

D FIG 67-147  Sagittal reconstruction taken through the lateral portion of the chest, about 3 cm lateral to the midclavicular line. A, Scan clearly shows the widened space between the anterior ribs (straight arrow) compared with the posterior ribs (curved arrow). One can appreciate how the structure of the ribs enhances the cephalad direction of the needle or catheter—there is an upward slope of the ribs. This permits remarkable flexibility in angling the device when inserted through the ribs. B, Sagittal reconstruction through the left side of the chest. The anterior diaphragm is much higher than the posterior portion, and the muscle slips are much higher anteriorly. C, Relationships of the diaphragm also can be appreciated by noting the position of the diaphragm in a patient with an inverted diaphragm. This scan shows what would apparently be a left subdiaphragmatic collection. Examination of contiguous slices suggests its true nature. D, Sagittal CT scan shows how a large pleural fluid collection pushes the posterior portion of the diaphragm down and forward. The anterior portion is virtually stationary because it is higher and less mobile; more anterior approaches are less variable and less prone to motion during the procedure.

dose of antibiotic. The antibiotic chosen should be the one selected from cultures and sensitivities (if available from fluid or blood cultures). The amount of antibiotic we inject is equivalent to a nondiluted single IM or IV dose. That amount of drug is deducted from the daily limit stipulated by the manufacturer and approved by the FDA. It also is important to pay attention to the volume required, because one must not inject volumes too large for the cavity. For typical pyogens, the volume is not an issue when routine broad-spectrum antibiotics are used. With fungal abscesses, the volume of the dilute drug might be too high, so the dosage would need to be reduced.

General Technique and Steps for Catheter Drainage.  The technique of abscess drainage can be divided into the following steps: (1)

abscess detection, (2) preprocedure steps, (3) choice of the guidance modality, (4) trajectory planning, (5) diagnostic aspiration, (6) catheter selection and insertion, (7) catheter management and irrigation/ fibrinolytics, (8) sinogram and catheter adjustment, (9) follow-up diagnostic evaluation, and (10) catheter removal. Abscess detection.  If an abscess has been detected by plain radiography, radioisotope-tagged white blood cells, or US, a CT scan should be performed to determine the complete extent of the abscess. The geographic location and degree of involvement are essential factors in choosing the trajectory, determining the number of catheters required, and anticipating possible difficulties. An abscess can be detected in selected cases by any modality, but CT plays the most critical role. CT is the preferred examination because

2504

PART III  Image-Guided Procedures

it is the most accurate of all imaging systems for abscess detection. Not only can CT confirm or exclude the presence of drainable fluid, but it can also diagnose the causative or associated problem. CT can provide valuable information about diverticulitis, tumors, appendicitis, fistulas, infarcts, cellulitis, phlegmons, inflammatory bowel disease, bowel obstructions, or other problems.

Preprocedure history, review of images, planning, and informed consent.  Preprocedure steps include physical examination, clinical

and medication history, and review of all imaging and clinical data. Medication history is especially important to determine if any anticoagulative drugs are taken (either prescribed or OTC). Laboratory studies including CBC and coagulation values are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained. Imaging preferences.  Virtually all authors now agree that CT is the modality of choice for performing drainage procedures. CT provides excellent visualization of the abscess cavity and its extensions. Air is accurately imaged and localized. Adjacent anatomy and virtually all instruments can be visualized without difficulty. The “longitudinal scan” (originated by our group183), or digital radiograph, is now available on all scanners, and the 3D capabilities are of great advantage in certain cases for contrast injection and catheter insertion. Procedure steps.  Patient vital signs and pulse oximeter monitoring with automated devices and local anesthetic and sedation administration are the same as used for other sites. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam and fentanyl or both as needed. Local anesthesia is administered as described earlier; deep sedation or general anesthesia is used. Sedation used with general anesthesia may lessen the depth of general anesthesia required, which is beneficial to the patient. Target and trajectory determination.  Trajectory planning is best accomplished by CT, regardless of the imaging modality chosen for guidance. Although, as most authors agree, it is best to choose the shortest pathway, it also is optimal to avoid uninvolved organs and noncontiguous peritoneal spaces. The most important areas for drainages are the supracolic spaces above the mesocolon and below the diaphragm (see Chapter 49) because these are the most common site for abscess formation. The most common spaces are the right subphrenic, left subphrenic, and lesser bursa (sac). To plan the trajectory in these areas, it is crucial to understand and appreciate the spaces as they relate to the diaphragm and rib cage. These relationships can be understood best by noting the interspaces between the ribs (see Fig. 67-147), which are larger anteriorly and laterally than posteriorly. Also, the diaphragm inserts higher anteriorly than posteriorly, making it more likely to avoid the pleural space the more anterior the trajectory. Improved delineation of anatomy and access to these spaces can be optimized by having the patients in variable positions. The subphrenic collections typically can be punctured and drained without traversing the pleura if one uses an angled approach and chooses a trajectory as anterior and lateral as possible. One should never choose an entrance pathway that traverses the pleura if avoidable, although operators who have done so intentionally claim no adverse outcomes. For either a right or a left subphrenic abscess, it is best to choose one of the lowest levels of the abscess and to choose a point in the midaxillary line (Figs. 67-148 to 67-150). The instrument is angled cephalad, the puncture is made, and the wire is advanced cephalad. Several points should be made concerning the configuration and nature of the abscess collection. If an abscess collection is long, the trajectory of the puncture should be planned so that when the catheter is inserted, it lies throughout the length of the cavity. Two approaches

A

B FIG 67-148  A, The patient has two left subphrenic abscesses, one anteriorly (curved arrow) and one posteriorly adjacent to the spleen. Note that the needle trajectory is planned at the lowest possible level of the margin of the abscess and through the midaxillary line (straight arrow). By angling cephalad, the needle, wire, and catheter remain below the diaphragm. B, Puncture and drainage of the anterior collection was made at a higher level even more anterior than the posterior cavity. Because the diaphragm is higher anteriorly, a straight in-axis approach is made.

may be used. First, one can choose an entrance site at the upper or lower end of the cavity so that the catheter traverses the entire length or width (Fig. 67-151). The alternative is to make two punctures at the center point, one directed cephalad and one directed caudad, to permit positioning of a catheter at the top and bottom of the cavity (see “Catheter Selection and Insertion Method,” later) (Fig. 67-152). If such an abscess is elongated in a transverse direction, catheters should be similarly placed in that transverse plane. If multiple septations are seen within a cavity, it is reasonable to plan the trajectory so that the maximum number of cavities would be traversed. Although best surgical results are obtained if an extraperitoneal approach is used, this is not true with percutaneous procedures. Neither we nor other authors have found any indications that a peritoneal approach differs in outcome from an extraperitoneal approach with percutaneous drainage. With percutaneous drainage of peritoneal abscesses, the high-pressure abscess evacuates through the catheter and

CHAPTER 67  Image-Guided Aspirations and Biopsies

C

B

A

FIG 67-149  A, Anterior subphrenic collection is adjacent to the anterior margin of the liver (arrow). An anterior approach is made so that no angulation is required to remain below the diaphragm. B, Seldinger method with wire is best used because metal clips can be fibrotic and painful unless careful dilation is used. C, Radiograph showing catheter in place.

A

B

C FIG 67-150  The pathway for needle puncture and catheter insertion should avoid the pleural space if possible. Because the pleural space is deeper posteriorly, a more anterior approach facilitates the avoidance of the pleural space. A, Axial CT scan shows the needle (arrow) being inserted anteriorly, adjacent to the anterior slip of diaphragm. B, Higher CT scan shows the needle tip in the abscess cavity. Note how the pleural space has been avoided. C, Topogram shows a radiographic type of image showing the catheter (arrow) in the right subphrenic space.

2505

2506

PART III  Image-Guided Procedures

not into the peritoneum because the margins of the peritoneum around the abscess are likely fused from inflammation. Drainage of the liver is different and requires a cuff of normal liver (see “Approaches to Specific Abscess Sites,” later). Diagnostic aspiration.  Before catheter or trocar placement, a diagnostic aspiration is performed to confirm the presence of an abscess. We use an 18-gauge needle to perform diagnostic aspirations for two reasons. First, a larger needle permits aspiration of even the thickest purulent material. Second, if at the time of aspiration the fluid appears purulent, the sheath permits insertion of an angiographic wire and catheter without a second puncture. The actual puncture and placement of the needle are performed in the same manner as any other needle procedure. We take two steps to minimize the risk of contamination along the pathway during the subsequent steps of catheter insertion. First, when draining abscesses in a solid organ, we choose the path to include a “cuff ” of normal parenchyma to lessen the chance of tearing the thin wall of an abscess. Second, at the time the puncture is made, we aspirate a sufficient amount of material to lessen the “back pressure” to minimize the chance of local spillage (Fig. 67-153). There is one reported case of spillage into the peritoneum that produced peritonitis and a case of dissemination into the subphrenic area reported by Johnson and coworkers.244 One must always strive for meticulous technique, but if spillage of purulent material occurs, placement of a second catheter at the site would prevent any problem (Fig. 67-154). If the aspirated material is not obviously purulent, we aspirate only a small amount of fluid and send it for culture. In the event that a repeat puncture and drainage might be required, we try not to aspirate the fluid collection “dry”; if such a cavity is aspirated dry, there would be no well-defined lesion to target should a drainage be required later. We do not insert a catheter when the fluid is not obviously infected, because if such a collection is sterile an indwelling catheter might

introduce pyogens and cause a secondary infection (Fig. 67-155). In patients with tumors in the area, a cytologic sample should be sent to determine if there is a necrotic neoplasm or secondarily infected tumor giving toxic symptoms similar to abscesses. Catheter selection and insertion method.  Two types of catheters are available for drainage: drainage catheters inserted by the  Seldinger method and single-step catheters. Either may be used, depending on the specific nature of each case. The Seldinger method is well suited to all abscesses regardless of the size or location; even the most difficult deep abscess can be drained with this method (Fig. 67-156). The single-step catheters are used only for superficial abscesses (Fig. 67-157). The size of the catheter and the size of the pigtail shape to be used should be matched to the size of the abscess cavity so that the entire curved end is in the cavity. It is important to use the largest catheter possible from the beginning; if the procedure becomes technically difficult, one can settle for almost any size of catheter and upgrade it later. Larger catheters permit faster drainage of material. It is important not to have holes outside the cavity. If holes are outside, leakage into a sterile space or organ can occur. Conversely, if a cavity is large, it may be important to cut additional side holes in the catheter if needed. Seldinger method.  After the needle has entered the cavity and a sample has been aspirated for culture and to relieve the pressure, we inject a small amount of contrast material to visualize the cavity and any leakage of material. Contrast material in the cavity facilitates the use of fluoroscopy or topograms if they are employed for catheter insertion. Also, if the wire or the needle inadvertently becomes displaced, the contrast material permits a clear target for a puncture under fluoroscopy. An angiographic wire is then inserted. Choice of the wire is individual preference; we prefer a slightly stiffer wire, such as a Rosen or an Amplatz, so that dilation of the tract and catheter insertion are facilitated. Although some interventionalists

A B

C

D FIG 67-151  A to D, Four scans taken before the drainage procedures show a long abscess in a child who had perforation of the duodenum secondary to a vasculitis. The abscess designated by the arrow is seen in the iliacus in A, the iliopsoas in B, the psoas in C, and the distal psoas in D. Others had attempted percutaneous drainage by draining only the fluid in D.

CHAPTER 67  Image-Guided Aspirations and Biopsies

E

F

H

G

I FIG 67-151, cont’d  E to I, Five images show the proper procedure with positioning of the catheter. In E, the needle puncture through the widest extent of the abscess was made. In F, the CT catheter (vertical arrow) is seen as it is used to perform a sonogram of the cavity that flows down through the psoas over the femoral head. The horizontal arrow shows the surgical catheter previously placed, which did not drain the abscess. In G, the wire and catheter (arrow) were manipulated to follow the entire length of the cavity. In H and I, the axial scans show the upper and lower extent of the catheter in all portions of the abscess.

2507

2508

PART III  Image-Guided Procedures

A

B FIG 67-152  A, Coronal reconstruction shows a long abscess (arrows) extending through the retroperitoneum. B, Plain film shows placement of two catheters entering from the middle of the cavity (curved arrow), extending to the top and bottom of the cavity (straight arrow).

A

B FIG 67-153  A, Accidental spillage of purulent material occurred because fluid was not removed to relieve the internal pressure of an abscess. CT scan shows a large abscess under pressure behind the stomach. Note that the needle is inserted in the margin. Apparently, insufficient material was aspirated to relieve the pressure before attempts to insert wire and catheter. B, The patient complained of severe abdominal pain after catheter insertion. The small amount of fluid surrounding the spleen represents spilled purulent material.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B FIG 67-154  A, After insertion of the large catheter, it was noted that a leakage of purulent material occurred into the subphrenic space (arrow). B, Puncture of the small subphrenic fluid was performed. When such punctures are attempted, they should be planned obliquely and so that the needle penetrates beyond the space. The instrument is withdrawn while suction is applied; the location of the tip in the space is heralded by aspiration of fluid. The wire and catheter are then inserted.

A

B FIG 67-155  A, CT scan shows a small sterile subphrenic fluid collection (arrow), which had a catheter placed within. We recommended that the catheter be removed in 24 hours if culture was not positive. B, The clinical team lost track of the patient and did not remove the catheter when the culture results were negative. The patient subsequently developed an abscess at the site from secondary infection (arrow).

2509

2510

PART III  Image-Guided Procedures

A

B FIG 67-156  A, The wire is inserted posteriorly and cephalad. B, With a digital radiograph, the cephalic positioning of the wire at the highest level of the abscess (arrow) below the diaphragm can be appreciated. There is no need to traverse the pleura to drain these very high collections.

A

B

FIG 67-157  A, Single-stick insertion in anterior abscess. One should use incremental steps inserting such catheters to ensure that the trajectory is correct, regardless of the size of the collection. Note the femoral vessels located to the left of the needle. B, Final placement of a singlestick catheter in the cavity.

are reluctant to use stiff wires, the stiffness prevents kinking of the wire or catheter during the dilation or insertion process. If the abscess is secondary to a local infarction or necrosis of the organ, a flexible wire should be used. Availability of a portable fluoroscopy device or an attached fluoroscope is helpful. With experience, one can insert the catheter on the CT table, using the topogram as the method for

confirming the catheter location (see Fig. 67-156B) or blindly using a sense of “feel.” Single-stick or trocar catheters.  Trocar catheters commercially available from many different companies can be used. I have found that these trocar catheters are useful for large superficial abscesses (see Fig. 67-157), but they are unsuitable for most small or critically located abscesses (see discussion in “Drainage Errors,” later). The large trocar associated with the catheters makes the instrument heavy, making incremental adjustment difficult, and the thickness of the trocar can create large artifacts on the scan. These two factors limit their use in deep smaller lesions that might be close to critical structures. Final positioning of the catheter is important. It should be placed into the deepest part of the abscess and not superficially to lessen the chance of dislodgment (see discussion in “Drainage Errors,” later). If the cavity is long, the catheter is positioned through the length of the cavity, or two catheters are inserted (see Figs. 67-151 and 67-152) to prevent regional loculations and enhance complete drainage. We make sure that none of the holes in the catheter are outside the cavity. After the catheter is positioned, we gently evacuate most of the purulent material from the cavity and irrigate it with saline. If the fluid becomes blood tinged, we cease irrigation. When loculations are present, we do not overirrigate or overmanipulate the catheter because this may cause bacteremia. Some authors have recommended use of a wire at this point to disrupt any septations or adhesions, but this is imprudent. Using a wire may traumatize the wall and cause significant bacteremia or hemorrhage from the granulation tissue. A review of the incidence of drainage complications shows that bacteremia is the most common. It is better to use fibrinolytic irrigations to treat such cases (see “Catheter Irrigation and Fibrinolytics,” later). Clinical management of catheters.  When the catheter is in place, a few guidelines must be followed to ensure a successful drainage. In most cases, catheters should be left only to gravity drainage or low suction. High suction may pull debris and tissue into the catheter holes and impair drainage. The one circumstance in which active

CHAPTER 67  Image-Guided Aspirations and Biopsies

2511

A

B FIG 67-158  A, CT scan shows two abscess cavities in the right side: one in the psoas muscle and one in the gluteus muscle behind the iliac crest. The high-density objects are cross sections of two different catheters that were used to drain the abscesses. B, Sinogram taken later in the course of the patient’s recovery shows communication of the long cavity in the psoas muscle with the small bowel. Total resolution without surgery followed (see text).

suction should be used is when there is free communication with either an enteric or a biliary fistula (Fig. 67-158). Patients should be followed on a daily basis to ensure that the patient’s response follows the expected course. When a drainage procedure has gone well, the patient responds quickly and shows dramatic clinical improvement. In such patients, the fever abates, the white blood cell count decreases, and sweating episodes stop. The progress of the drainage also can be assessed by following the sequential changes in the appearance of the fluid. Initially the material is cloudy and turbid; it then changes to a serosanguineous appearance, and finally it changes to a clear serous material. Cultures of the fluid also change sequentially so that when the abscess has resolved, drainage fluid is sterile. After this clinical response and the change in character of the material, the catheter can be removed. If the patient does not respond as expected, the fluid drainage is scant, purulent fluid does not evolve as expected, or the drainage begins to show bile or enteric material, repeat imaging is indicated. Fistulas usually are not apparent on initial postprocedure scans because the small fistulas are usually sealed by inflammatory edema in the area, which temporarily occludes the fistula. As healing begins and edema reduces, the fistula opens up and changes the appearance of the drainage material or the appearance on CT. A repeat CT scan should be performed to evaluate the abscess to determine if fistula is present (shows large amount of air) (see Fig. 67-158), if portions of the cavity are incompletely drained (Fig. 67-159), or if additional cavities have developed. Evaluation of the catheter status is important to detect any displacement or kinking (Figs. 67-160 and 67-161). Repeat puncture and catheter insertion or repositioning of the catheter may be necessary. The catheter also should be inspected to ensure nothing “crazy” has happened, such as someone tying a knot in the catheter (Fig. 67-162). Catheter irrigation and fibrinolytics.  Catheter irrigation should be performed routinely with saline, but either of two fibrino-

lytic agents may be used when the material is thick or septations are present. If the material being drained is thin and no definite septations are noted on imaging studies, we inject saline. Saline in the amount of 10 to 15 mL should be injected every 6 to 8 hours to keep the catheter lumen clear. If the material being drained is thick or there are septations in the abscess, urokinase or tissue plasminogen activator (tPA) should be injected into the abscess cavity every 8 to 12 hours for 3 to 4 days. Both of these fibrinolytic agents change the substrate plasminogen to the active fibrinolytic agent plasmin. There is no reason to suspect that its activity should be any different than urokinase. If one reviews the table of information from the urokinase study (Table 67-5), it can be seen that urokinase improved the clinical factors of the patient and shortened the drainage time and hospital stay.192 The amount of urokinase used depends on the size of the cavity as follows: 0 to 3 cm, 12,500 U; 3 to 5 cm, 25,000 U; 5 to 10 cm, 50,000 U; and greater than 10 cm, 100,000 U. After the urokinase has been injected, a small amount of saline should be used to clear the catheter of the urokinase. The catheter should be clamped for 5 minutes and then left to gravity drainage. If tPA is chosen, the corresponding amount is administered; one should check with the pharmacy or package insert for the amount relative to urokinase. This use of tPA  is not approved by the FDA but is legal according to off-label regulations. The effectiveness and safety of fibrinolytics has been documented by our group in the literature.192,278,399 Park and associates showed that in vitro, urokinase decreased the viscosity of purulent material more than saline did (Fig. 67-163).399 The safety of urokinase in vivo was proved by Lahorra and colleagues, who showed that  intracavitary urokinase produced no systemic effects.278 A later article by Haaga and colleagues showed considerable benefit for using urokinase compared with saline (Fig. 67-164).192 This article showed that its use improved treatment of abscesses. Its use produced earlier

2512

PART III  Image-Guided Procedures TABLE 67-5  Results of Student’s Test for

Length of Stay, Treatment Cost, Febrile Course, Days of Elevated WBC Count, and Days of Drainage Test for

Group

Mean

Saline 29.00 ± 17.2 13.11 ± 9.5 Urokinase Treatment cost ($) Saline 62,102.11 ± 42,844 Urokinase 24,178.00 ± 17,871 Febrile course (days) Saline 4.00 ± 6.25 1.45 ± 1.73 Urokinase WBC elevation (days) Saline 9.12 ± 8.79 5.62 ± 5.04 Urokinase Drainage (days) Saline 14.88 ± 20.5 6.93 ± 4.2 Urokinase Length of stay (days)

t

df

3.56 24

P* .0025

3.49 22.3 .0021 1.63 18.1 .1202 1.41 25.7 .1710 1.57 17.5 .1361

From Haaga JR et al: Intracavitary urokinase for enhancement of percutaneous abscess drainage: Phase II trial. AJR Am J Roentgenol 174:1681–1685, 2000. *Separate variance t-test used. df, degrees of freedom; WBC, white blood cell count.

A

B FIG 67-159  A, CT scan shows a large abscess with a catheter positioned anteriorly with its tip close to the wall. B, The catheter became displaced either because of retraction of the abscess wall away from the catheter or because of movement of the abscess cavity during patient motion. A more laterally positioned catheter was inserted with a deeper location.

improvement of clinical parameters and shortened hospital stay (Table 67-6; see Table 67-5). Continuous irrigation.  In some cases, when especially thick material is present or there is injury to the vascular supply, placement of several catheters for irrigation may be beneficial. One catheter is used for inflow of saline, and one or more catheters may be used for the evacuation of the fluid. In such cases, one should ensure that fluid is moving freely between the catheters so that there is no buildup of fluid. Sinograms.  Some authors have performed sinograms to follow abscess cavities. A CT sinogram can show if cavities are undrained and if additional catheters are necessary (see Fig. 67-159). During the sinogram, one evaluates whether the catheter placed is still in the correct position and is of adequate size for the evacuation of any residual purulent material. If there is considerable purulent material left, one

may choose to increase the size of the catheter. At this time, sequential dilation may be performed safely without bacteremia because the pathogens are fewer as a result of the treatment. If the drainage material is thick or loculations are present, one might institute catheter irrigation with fibrinolytic agents at this time, if it was not used initially. With the cavity partially evacuated, the contour may change and shift the location of the catheter, requiring adjustment of the position. Repositioning, replacement, or adjustment should be performed as appropriate (see Fig. 67-160). Fluoroscopic sinograms reveal any communication with bowel (see Fig. 67-158B) or biliary system that should be treated with suction or additional catheters. When performing such sinograms, one should be careful not to inject the catheter and cavity with too large a volume. Excessive volume may increase the intracavitary pressure sufficiently to cause bacteremia or to rupture a weakened wall of a cavity and produce spillage. To prevent political conflict with clinical services, it is always helpful to remind clinicians that in many cases a preexisting fistula may be the primary cause of the abscess. The late appearance of a fistula during the drainage process is due to the fact that inflammatory edema in the wall of the abscess occludes a fistula temporarily. At a later time, when the inflammation partially resolves, the fistula opens and manifests itself. One should not accept such fistulas as caused by the PAD unless the bowel was clearly crossed during the procedure. Catheter removal.  Before final removal of the catheter in a routine case, a final CT scan is performed to ensure that all areas of the abscess are drained. An abscess that is completely resolved should show virtually complete obliteration of the cavity (but one may see a small amount of residual fluid and air) or decreased density in the liver. If a considerable amount of fluid remains, it is best to wait until the cavity has more completely resolved and has been reduced in size by granulation tissue. If a patient is doing well clinically, he or she may be discharged and managed as an outpatient during the time it takes for the cavity to granulate closed. A liver abscess may be confusing on follow-up scans. An abscess may be completely drained, but a low, ill-defined density area at the site slightly higher in density than fluid may remain. This represents local fatty change or edema or both, which resolves spontaneously.

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2513

B FIG 67-160  A, CT scan taken during a contrast sinogram shows a catheter at the bottom of a cavity with contrast material not filling an anterior portion of the cavity. B, A second catheter was inserted from an anterior approach.

CATHETER SIZE VERSUS TIME TO DRAIN 10 ML OF FLUID 900 Water 800

Pseudocyst fluid Blood

700

Abscess with urokinase Abscess without urokinase

FIG 67-161  CT scan over the pelvis shows that the catheter tip has become severely kinked (arrow), impairing the drainage flow.

Time (sec)

600 500 400 300 200 100 0 0

0.1

0.2

0.3

0.4

0.5

Inner diameter (cm)

FIG 67-163  An in vitro experiment was performed evaluating different parameters including urokinase treatment. Graph shows the drainage time of various fluid types through various sizes of catheters. The upper two lines show how urokinase improves drainage of purulent material. (Data from Park JK et al: Fluid flow during percutaneous drainage procedures: An in vitro study of the effects of fluid viscosity, catheter size, and adjunctive urokinase. AJR Am J Roentgenol 160:165–169, 1993.)

FIG 67-162  On close inspection of the draining apparatus, someone had tied a knot in the drainage tubing.

2514

PART III  Image-Guided Procedures

A

A

B FIG 67-164  A, CT scan of patient lying on the left side during the planning phase of abscess drainage. The patient has had a proctectomy with an ileal anastomosis; the curved arrow indicates the dense suture line. The abscess (A) located in front of the sacrum is too far posterior to be drained by the transrectal route (the preferred route). Note the air-filled cannula (arrow) in the rectal region demonstrating that the abscess cannot be reached through the rectum. For this reason it was decided to use transsciatic route. B, CT scan shows the catheter positioned in the abscess cavity through the transsciatic route. Transcatheter fibrinolytic was used to facilitate drainage. The abscess completely resolved.

TABLE 67-6  Effect of Loculation on Length of Stay and Hospital Costs Loculation

Test for

Group

No

Length of stay (days)

No

Treatment cost ($)

Yes

Length of stay (days)

Yes

Treatment cost ($)

Saline Urokinase Saline Urokinase Saline Urokinase Saline Urokinase

Mean

t

df

P*

27.09 ± 17.8 11.04 ± 4.72 58,023.33 ± 44,075 21,567.40 ± 7797 32.50 ± 17 13.20 ± 11.81 70,259.66 ± 42,981 17,315.60 ± 6760

2.81

2.81

.0168

2.81

11.8

.0169

2.13

9

.0167

2.97

5.3

.031

Note: χ2 = 1.57, df = 17.5, P = .8933. From Haaga JR et al: Intracavitary urokinase for enhancement of percutaneous abscess drainage: Phase II trial. AJR Am J Roentgenol 174:1681–1685, 2000. *Separate variance t-test used. df, degrees of freedom.

The decision to remove the catheter is easy if the course of the drainage (i.e., clinical response, evolution of fluid changes, and appearance of the sinograms) is as expected. Before removal of a catheter, a final scan should be obtained to ensure that the cavity has closed completely. If this has occurred and there is no continued drainage, the catheter can be removed. The patient’s antibiotics should be continued for several days after removal of the catheter.

Fistulas.  GI or biliary fistulas commonly occur with intraabdominal abscesses; some fistulas probably cause abscesses and some are sequelae.363 Management depends on the amount of flow, whether it is of low or high output.280 Low-output fistulas (320 mL/day) usually close spontaneously without additional therapy (see Fig. 67-158). High-output fistulas require treatment. High suction on the catheter

should be maintained, and the bowel should be put at rest. This usually means insertion of a nasogastric tube. Hyperalimentation also may be needed for several weeks in some severe cases. Failure of these methods indicates a need for surgery.

Success Rates Since the introduction and development of PADs, there has been controversy among various groups about the achievable success rate, which overall has varied between 62% and 90%. The differences in success rates depend on numerous factors, including selection criteria, drainage techniques and catheters, and final criteria for outcome determination (i.e., definition of cure, palliation). Numerous authors have studied the factors that affect the outcome of the drainage procedures. Several major determinants have been found, including the status of

CHAPTER 67  Image-Guided Aspirations and Biopsies the immune system, the overall severity of the patient’s illness, and the accuracy of abscess localization. Lambiase and colleagues reported a large series of patients treated with PAD and noted a significant difference in the outcomes of their patients.281 They found that immunocompetent patients had a successful outcome of 72.6%, compared with immunocompromised patients with a success rate of 53.4%. Levison and Zeigler correlated the APACHE II score with drainage techniques and outcomes.294 These authors found that patients with scores less than 15 had a mortality rate of 1.7%, and patients with scores greater than 15 had a mortality rate of 78%. There was no benefit of one method—surgical or percutaneous drainage—over the other. Deveney and colleagues studied the outcome of abscess drainage in two periods: the first in the 1970s and the second in the 1980s.100 The mortality in the two periods was 39% compared with 21%. There was no difference in outcome between patients treated with surgical methods and patients treated with percutaneous methods. The authors concluded that the improvement in localization had a major impact on the treatment of abscesses. There are many reasons for the failures reported, including premature withdrawal of the catheter, presence of a GI fistula, pleural contamination, hemorrhage, and the errors of selecting cases, as noted earlier. Lang reported a series of 136 patients with a failure of 29 of 119 attempted curative procedures.283 With infected pancreatic processes, Lang and colleagues were successful in 11 of 22 cases. In their series, they had 17 patients with identified GI fistulas; 9 resolved spontaneously, and 8 required surgical management. Failure to document the complete obliteration of abscesses by CT resulted in failure in seven cases. They had two cases of hemorrhage associated with inappropriate placement of catheters in areas with phlegmons. The most common misdiagnosis was failure to identify loculated or residual cavities. Premature withdrawal was made in many cases because of the cessation of drainage fluid. The authors recommended that a repeat CT scan be performed before the removal of the catheter.

Success with Unusual Clinical Circumstances.  Many unusual clinical circumstances are encountered in daily practice that are worthwhile to note specifically, including patients with Crohn’s disease,  critically ill patients in intensive care, and infections caused by antibiotic-resistant organisms. Relative to Crohn’s disease, Casola and associates59 and Garcia and colleagues137 reported on the outcome of medical/percutaneous drainage versus surgical treatment. Casola and associates reported success in 15 of 15 patients without any fistulas.59 In the series of 51 patients by Garcia and colleagues, fewer recurrent abscesses developed in postoperative patients than in patients treated by the more conservative method (12% vs. 56%).137 Schurawitzki and coworkers noted the superiority of CT guidance over US in the intensive care setting.439 With US there were numerous equivocal studies that prevented adequate evaluation or performance of a drainage procedure, or both. Antibiotic-resistant organisms are becoming more prevalent in the environment, so the question of drainage effectiveness is raised. Catalano’s group reported a series of 21 patients with vancomycin-resistant enterococcal abscesses60; 71% were drained successfully using standard guidance methods and systemic antibiotics.

Drainage Errors Errors associated with abscess drainages have been documented by numerous authors. Perhaps the most common error made is the improper choice of trajectory, which can cause either a portion of the pleura to be crossed when draining a subphrenic abscess or a solid

2515

A

B FIG 67-165  A, CT scan shows a large abscess (arrow) anterior to the spleen, with the needle entering the cavity. The anterior vertical arrow indicates the falciform ligament. The radiologist did not aspirate enough fluid before insertion of the wire and the catheter. The high pressure caused leak of purulent material and the resulting bacteremia. B, CT scan performed after catheter insertion shows purulent material anteriorly (curved arrow) and around the spleen (arrow).

organ to be traversed. Although with today’s potent antibiotics dissemination of an infection is infrequent, it can occur and produce significant difficulty. Removal of some purulent material is important after the initial puncture to relieve the tension of a large abscess. If this is not done, there is a chance that leakage of infected material may spill into the peritoneum during the insertion and manipulation of the wire and catheter (Fig. 67-165). Relative to placement of the catheter, it is important that the largest catheter possible be inserted, not only to permit rapid drainage of the fluid but also to provide ample diameter should some purulent material become inspissated on the inner wall of the catheter. This does not mean that a massive catheter is required, but one at least 10F to 12F should be used. Also, the catheter should be placed in the most remote site of the cavity so that loculations do not occur distal to the location of the catheter. When a very large abscess is present, the catheter should be inserted deep into the cavity; if it is not, problems can develop. If a large cavity were to drain well or a patient to shift position, a superficially placed catheter may be displaced from the cavity, requiring a repeat puncture (see Fig. 67-160). Inadvertent or early removal of the catheter results in recurrence of the abscess. Failure either to identify or to correct the cause results

2516

PART III  Image-Guided Procedures

in a recurrence, such as fistula or ulcer. If a final CT scan is not obtained, a recurrent or undrained abscess can be unrevealed. Alternatively, if a catheter has been inserted into a sterile fluid collection and it is not removed promptly, a secondary infection of the fluid space can occur. If an abscess is very large, one can be overly confident and attempt to hasten a procedure by “skipping” some steps, such as using CT localization. In one such patient, one of our experienced physicians used a one-stick catheter without guidance and penetrated the femoral vein (Fig. 67-166). The patient did well after insertion of a vena cava filter to prevent emboli, insertion of a proper drainage catheter, and removal of the offending catheter from the vein. Finally, if an abscess does not resolve with drainage or enlarges, one should suspect a fungal infection or spontaneous bleeding. It is frequently observed that the infectious agent in an abscess may change based on elimination of sensitive organisms and selection of resistant organisms. Candida commonly becomes the dominant organism in a patient with multiorganism infection who is treated with broad-spectrum antibiotics selected for bacterial pyogens. In several cases, we have encountered an abscess that appeared to increase in size and developed an elevated density. In such cases, localized bleeding may occur because of the invasion of the local blood vessels; effective percutaneous drainage is typically impossible, and surgery is preferred.

Complications The incidence of complications varies widely depending on technique, types of abscess being drained, and various authors’ experience. Reported rates vary between 5% and 9.8%. In regard to possible abscess locations, the complication rate ranges from 1.4% to 30%. The lowest complication rate is associated with the liver, the highest with the spleen. The most common complication in all series is that of sepsis immediately after catheter insertion. According to our review of the literature, this occurred in about 6% of all cases. As seen in the review article by Lambiase and coworkers, it occurred in about 4.2% of cases.281 Other less frequent complications included spontaneous hemorrhage, spillage of infected material, and fistula formation.

Approaches to Specific Anatomic Sites The various anatomic areas involved by abscesses have unique characteristics as discussed in Chapter 49. Information unique to the technical approach and outcomes in each area are discussed in the following sections.

Subphrenic Abscesses.  With a subphrenic abscess, one must always look at the contiguous anatomic areas to see if there are any extensions of the abscess into these areas. Contiguous spaces that may require drainage include the right subhepatic and right pericolonic spaces with right subphrenic abscesses. With left subphrenic abscesses, extension into the right subphrenic, left subhepatic, and left pericolonic spaces also may occur. Traditionally such extension is not expected, but because of anatomic variability, some peritoneal margins do not serve as effective barriers to the spread of infection. When planning the trajectory and performing such a procedure, the operator should always visualize and avoid the pleural space, gallbladder, and hepatic flexure of the colon. If there is any problem with localizing a subphrenic abscess or separating hepatic flexure of the colon from the liver, the patient should be placed in the lateral or decubitus position. The key elements of the trajectory approach were described earlier in the general technique of catheter drainage (see Figs. 67-148 and 67-149).

The rate of cure for subphrenic collections is 79% to 81%. Complications include empyema and sepsis, the incidence of which is 2%.145,191,198,281,366

Right Subhepatic Abscess.  An abscess in the right subhepatic space can be approached easily either laterally or posterolaterally (Fig. 67-167). Penetration of an adjacent structure, such as the colon, gallbladder, or duodenum, can be easily avoided. In some cases, a left decubitus position also can be helpful in moving the hepatic flexure of the colon to simplify a puncture. In cases in which clear access is unavailable, it is reasonable to traverse a portion of the liver. Because the liver in such cases is clearly exposed to the pyogens owing to close proximity, there is likely no difference in risk between the two approaches.

Left Subhepatic Abscess.  A left subhepatic abscess, located in the anterior subphrenic space below the left lobe of the liver, appears between the liver and the stomach on axial CT scan. The administration of oral contrast material is important to confirm the stomach location. The optimal approach for this area is a directly anterior approach that avoids liver and stomach (Fig. 67-168).

Omental Bursa.  Drainage of the omental bursa can be approached from several directions. Fluid collections in the inferior recess of the lesser sac can be approached anteriorly beneath the lower margin of the stomach if a fluid collection is large enough. If the fluid collection is in the splenic portion of the inferior recess, the best approach is through the gastrosplenic ligament. This approach is best accomplished when the patient is in a right decubitus position to permit the stomach to shift anteriorly. In such cases, one must be careful to avoid the splenic flexure of the colon, looking carefully for varices in the gastrosplenic ligament to avoid penetration of any such vessels (Figs. 67-169 and 67-170). If the approach through the gastrosplenic ligament is unavailable, two other approaches are possible. For a second choice, I would use a transgastric approach, ideally through a PEG as described earlier for pseudocysts. Another approach is a transhepatic approach through the left lobe of the liver (other authors recommend this as a primary approach).364 We dislike this approach because of possible injury or pyogenic dissemination into the left biliary or portal system.

Inferior Recess of Lesser Bursa and Porta Region.  With the mastery of these techniques, it is now possible to approach drainage of deep areas, which can provide remarkable benefit. These deep areas are difficult to drain by inexperienced operators, but with experience and knowledge of the techniques described in this chapter, drainage of very deep areas is possible. To accomplish these procedures successfully, one must take advantage of patient positioning and needle deflection as described in earlier sections (Fig. 67-171).

Pericolonic Spaces.  The pericolonic spaces are easily approached laterally. One should always look for extension of an abscess into the pelvis or either subphrenic space. Contrary to traditional teaching, an abscess in the left pericolonic space may extend into the left subphrenic space unimpeded by the phrenicocolic ligament. When draining an abscess in these spaces, always make sure the catheter traverses the entire length of the cavity. To do this, it is always best to enter either the upper or the lower end of the cavity. Another option is to insert two catheters at the midportion of the cavity, with one extending cephalad and one extending caudad.

Empyema and Pleural Effusions.  Drainage of empyemas is straightforward, and the same methods are used as for other areas. The

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

A

C

D

FIG 67-166  A, CT scan shows a large abscess in pelvis (arrow). The physician thought that the collection was so large that guidance was not necessary. A one-stick catheter was inserted at the tableside. B, Fluoroscopic image of contrast injection of catheter shows filling of right iliac vein and vena cava. C, Repeat scan shows abscess (A) near the catheter (curved arrow) in vein. Needle (horizontal arrow) is seen anteriorly entering the abscess cavity (A).

2517

2518

PART III  Image-Guided Procedures

A FIG 67-168  Collections in the left subhepatic space are best approached through the gastrohepatic space. The needle (arrow) usually avoids the liver and stomach; if inadvertent puncture occurs occasionally, it is better than intentionally penetrating these structures in every case.

B FIG 67-167  A, Direct lateral approach to right subhepatic abscess. B, When the patient is positioned on the left side, the fluid remains in the same location and does not shift. Note the drainage needle in the right flank.

most important point relative to empyemas is to use a more lateral or anterior approach in preference to a posterior approach. If a catheter is inserted posteriorly, it is uncomfortable for the patient, and the tube is likely to kink and limit the drainage. Because such abscesses have a significant amount of inflammatory reaction, vacuum on the catheter is required, not to keep the lung reexpanded, but rather to improve removal of the material. Empyemas do not require a water-seal vacuum device, because the cavity is a closed space owing to fibrous reaction.

Liver Abscesses.  Abscesses in the liver should be approached by the shortest trajectory available, with an entrance point on the anterior or lateral surface. When a lateral entrance site is used, the operator should not go more posterior than the midaxillary line to avoid traversing the pleura. When puncturing and draining hepatic abscess, the operator should avoid any visible hepatic vessel, the gallbladder, and pleura. There should always be a cuff of normal liver to prevent spillage of material into the peritoneum. Many hepatic abscesses may appear to be multiseptated,105,244 but experience has shown that many clear completely with one drainage

tube. Various explanations have been proposed, including that (1) the abscesses intercommunicate, (2) the abscesses drain spontaneously into the drained cavity, and (3) antibiotics clear the different cavities. A repeat scan is important to ensure that all such cavities eventually clear, because separate cavities that persist occasionally require drainage (Fig. 67-172). The success rate for hepatic abscesses varies between 69% and 89%.223,244 The complication rate is 1.7% to 7.3%. Complications include sepsis, contamination of peritoneum, and production of an empyema.

Pancreatic Abscesses.  Drainage of pancreatic abscess by PAD is a complex issue that has not yet been resolved, but certain comments can be made. The issue is complex because there is a discrepancy in the literature about the objective results, recommendations, and the experience of various authors. The overall success rate is 57%, with a fairly high complication rate. Complications have been short term and long term. Short-term complications have been related to hemorrhage, either for pseudoaneurysms or for inappropriately attempted drainage of poorly defined phlegmons. Initial attempts to drain pancreatic abscesses were nonselective and the methods were not clearly defined. Most workers differentiate necrotic ill-defined pancreatitis with superimposed infection from well-delineated low-density-fluid collections, which more closely resemble infected pseudocysts. The welldefined fluid-density collections are best suited for percutaneous drainage. The standard methods using the Seldinger technique or trocar methods provide good results. Freeny and coworkers reported a high success rate with such methods.131,132 In their group of 21 patients, the success rate was 85%. Van Sonnenberg also claimed a high success rate of 69%.499 This author encountered 13 fistulas, 10 of which closed spontaneously. Lambiase and colleagues reported successful drainage of two infected pseudocysts compared with one of seven cures in a group of patients with abscesses exclusive of pseudocysts.281

CHAPTER 67  Image-Guided Aspirations and Biopsies

2519

S A

B

A

Perisplenic

FIG 67-169  A, Radiograph shows a catheter in the splenic recess of the lesser sac. B, CT scan shows the standard approach for drainage of the superior recess abscess (A) of lesser sac behind the stomach (S). The space is wide enough between the stomach and spleen to permit easy access.

S S A

C P

A

B

FIG 67-170  A, Drainage of the splenic recess abscess (A) of the lesser bursa, behind the stomach (S), was impossible with patient in the supine position. Note colon (C) and spleen (P). B, With patient lying in the right decubitus position, the stomach (S) moved anteriorly, providing a wide pathway between the spleen and stomach (arrows) for the needle (curved arrow).

Regardless of the primary approach taken for drainage of pancreatic abscesses, successful resolution of the problem depends on diligent repetitive clinical and imaging evaluation of the patient. Close cooperation between the surgeon and radiologist is essential because of the recurring and unpredictable nature of the disease. A radiologic procedure often is needed after an operative drainage or vice versa to produce complete resolution of the problem. In our experience, the technique and approach used depend on the appearance and location of the pancreatic fluid collections. For welldefined collections, we are willing to perform percutaneous drainage because of the high success rate. Collections representing infected necrotic tissue or poorly defined multiple collections are best drained surgically. If a patient has had a surgical drainage and develops a collection postoperatively, we attempt percutaneous drainage regardless of the appearance, knowing that the infection is most likely well confined and that a second surgical procedure would be difficult.

We use either an anatomic approach that avoids all adjacent  organs or a transgastric approach. If one is attempting to avoid bowel or solid organs, the approach often must be retroperitoneal, anterior to the kidney or directly anterior. We use such a direct anterior approach only if there is a “window” between the various loops of bowel. If there is any problem with access, we choose a transgastric approach using the same method we have described for pseudocyst drainage, with a gastrostomy as a conduit for entry (see “Pancreatic Pseudocyst”). One must be prepared for multiple catheter insertions because pancreatic abscesses may be difficult to drain effectively (see Fig. 67-160).

Renal Abscesses.  Renal abscesses are best approached posteriorly through a path avoiding spinal erector muscles, colon, liver, and spleen or from a posterior axillary line location (Fig. 67-173). Puncture and drainage of most such abscesses can be performed with US or CT

2520

PART III  Image-Guided Procedures

C

A

C

B

FIG 67-171  A, Patient with a small abscess (vertical arrow) beside a metallic stent in the biliary system (horizontal arrow) anterior to the vena cava (C). An anterior approach would be impossible, so a posterior approach was chosen. For the needle to clear the vena cava, the bevel was rotated toward the left (vena cava) and pushed slowly to permit deflection. B, Placement of the needle precisely in the abscess was accomplished. C, Final placement of the catheter (vertical arrow) next to the stent (horizontal arrows) was accomplished. The patient defervesced and survived, without surgical drainage.

A

B FIG 67-172  A, CT scan shows multiple large abscess cavities in the liver. Because of the large number, only the two largest cavities were drained. B, Follow-up scan shows the reduced size of two drained cavities and the other smaller cavities.

guidance. When the upper pole is involved, CT is indicated to avoid inadvertent trauma to the spleen, splenic pedicle, or pancreas.

Splenic Abscess.  Drainage of splenic abscess is controversial for several reasons. Some early investigators reported favorably on drainage of splenic abscesses, but if one carefully reviews their reports, there was a high incidence of emergent splenectomy. Having never been an advocate of such procedures on a routine basis, we have waited for additional experience to be accumulated before recommending this approach. The more recent information has remained inconsistent. Liu and coworkers306 reported a 90% success rate, whereas Green162 reported a 25% success rate. Thanos and Diliana486 reported favorable results with no complications. These reports come from different areas of the world, and there may be differences in selection of patients and geographic and ethnic differences among patients.

Because the incidence of splenic abscesses worldwide is increasing owing to the increasing number of immunocompromised patients, we believe the procedure should be used but only in selected patients. Because the spleen is so vascular, one probably should drain only abscesses that are very close to the surface of the spleen. Realizing that this procedure has not proven to be safe in controlled studies, one must always have close communication with surgeons to provide “expectant observation” and make sure an emergent splenectomy can be performed expeditiously if needed (Fig. 67-174).

Perirenal and Pararenal Spaces.  The anterior and posterior pararenal spaces are frequent sites of abscess and can be approached safely284; however, certain facts must be remembered. These spaces extend from the pelvis to the diaphragm and are long. Unless a catheter or catheters are positioned throughout the entire length of the cavity, drainage may be unsuccessful. It is especially important to use CT for

CHAPTER 67  Image-Guided Aspirations and Biopsies

2521

A

FIG 67-173  Puncture of an air-containing abscess in the kidney through the posterolateral trajectory.

follow-up of such cases because the abscesses may spread toward the diaphragm or pelvis. Examination of these areas may be difficult with US. The success rate for these abscesses ranges from 68% to 81%. The complication rate is 3% to 6%, with sepsis and hemorrhage being the most common.281,284 S

Pelvic Abscesses.  The procedure for drainage of pelvic abscesses is similar to that of other areas. The most significant controversy involves the choice of drainage pathways. In our opinion, the best pathways are an anterolateral transrectal one (Figs. 67-175 and 67-176) or a transsciatic one. When abscess cavities are close to the rectum, a transrectal approach is indicated (Fig. 67-177; see Fig. 67-176).103 To perform this procedure, a series of baseline scans are taken over the pelvis to show the collection. A short plastic tube (with povidone-iodine ointment) is inserted into the rectum. A plastic sheath needle is inserted into the tube, and a series of repeat scans are taken over the pelvis. When the needle is aligned properly, puncture into the fluid collection is made. The needle is removed, but the plastic tube is left in place to serve as a conduit for catheter insertion. The tube prevents buckling of the wire or catheter in the rectum during the insertion. The second choice for draining pelvic abscesses is through the sciatic notch (Fig. 67-178). This pathway is satisfactory for biopsy procedures, but long-term drainage catheters are not recommended for several reasons. The use of this pathway can produce a lot of pain, and there have been reports of abscess tracking into the gluteal muscles, of sciatic neuritis, and of pseudoaneurysm formation in the iliac vessels. Because of the proximity of the sacral margin and passage through the muscles, catheter kinking is common. For prostatic abscesses, the transsciatic approach is the only route available for CT. This abscess has a high mortality rate and must be drained expeditiously (Fig. 67-179). Transrectal US is perhaps a better approach for prostatic abscess.303

B FIG 67-174  A, CT scan shows air contained within splenic hematoma (arrow), representing an abscess. Note surgical clips on the splenic vessel, placed during laparoscopic surgery. B, CT placement of catheter was performed. Under expectant observation, the patient had splenic hemorrhage (s) requiring splenectomy.

Controversial Areas

Antibiotics.  We believe all patients who have a percutaneous procedure should have systemic antibiotics before and during the course of abscess drainage. With antibiotics, patients have quicker resolution of the abscess and fewer local complications. Some patients not receiving antibiotics may develop a localized cellulitis around the site of catheter insertion; administration of antibiotics resolves or prevents this. If antibiotics are not given, septicemia is more likely to occur after or during instrument manipulation.

2522

PART III  Image-Guided Procedures

A

B FIG 67-175  A, CT puncture of appendiceal abscess; note an appendicolith (curved arrow). B, Radiograph showing catheter positioning.

In our opinion, the only valid controversy is whether antibiotics should be started before or after the diagnostic aspiration. Starting antibiotics when a patient initially develops septic symptoms can be justified to prevent further deterioration. It can be argued, however, that a valid culture cannot be obtained if the patient is receiving antibiotics. Either philosophy seems sound; we usually accept the preference expressed by clinical colleagues attending the patient.

Treatment by Aspiration Instead of Catheter Drainage.  Several authors have reported on the concept of treating abscesses with aspiration, saline lavage of an abscess cavity, and intracavitary antibiotics. Multiple series have been reported with good success despite variations in technique. The numerous reports were discussed in the earlier section “Treatment of Abscesses by Aspiration and Antibiotics Alone.” These were not well-designed studies because they did not have welldefined selection criteria, were nonconsecutive, and had no comparison with standard drainage methods. If better data can be obtained, this may be an important, more simplified approach for abscess treatment.

Catheters.  Abscess drainage has become so ubiquitous that there are innumerable drainage catheters available on the market. Each has its own unique quality advertised by the various vendors. The list is too exhaustive to recommend any specific catheters; readers can decide which products meet their own needs. Controversy exists about whether gravity (siphon) drainage, sump drainage, or irrigation drainage (multiple catheters) is the most effective. A report by Park and coworkers showed that there was little difference in drainage success rates regardless of whether small or large catheters were used.399 This is probably the result of selection of catheters by operators (i.e., small catheters for thin material and large catheters for thick material).

Mucolytic Agents.  There have been several suggestions in the literature about using materials to try to enhance drainage from abscesses. Van Waes and colleagues proposed the use of acetylcysteine (Mucomyst) to decrease the viscosity of purulent material501; most workers

do not accept this technique as being effective, and no additional reports have appeared.

Fibrinolytic Agents.  Extensive work has been performed by our group192,278,399 using urokinase initially, and now other fibrinolytics have been studied by other groups. Fibrinolytic agents all work by the same pathway, converting plasminogen to plasmin. Gervais, presenting data on the use of tPA to facilitate abscess drainage at the 2006 meeting of the Radiology Society of North America (Chicago, Illinois, November, 2006), reported favorable results with no issues related to bleeding. Based on these reports, fibrinolytics are no longer considered controversial. Virtually any fibrinolytic can be used, including streptokinase, urokinase, and tPA, as noted earlier. Other fibrinolytics are now being evaluated as irrigants for catheters.

Drainage of Splenic Abscesses.  Percutaneous drainage of splenic abscess is a questionable procedure because of surgical experience and the reported data in the literature. The spleen is a large vascular organ, and any manipulation presents a potential risk. Even with the current surgical trends favoring preservation of the spleen in traumatic injury, most surgeons perform a splenectomy rather than attempt a catheter drainage. In the accumulated literature about PAD of the spleen, a high failure rate and high complication and mortality rates are reported. Surgery is the most acceptable treatment for a splenic abscess except in extraordinary cases. In such cases, if the patient and family and all of the physicians are aware of such issues, a drainage in the splenic area can be performed with “expectant” observation, with the knowledge that a hemorrhage can occur (see Fig. 67-174).

Diverticular Abscesses.  Percutaneous drainage of diverticular abscesses is a technically simple procedure for which the indications are limited because in most patients with diverticular abscesses, a colonic diverticular fistula results after successful drainage of the abscess. We concur with the many authors who reported that the primary purpose in performing this procedure is to change a potential two-stage procedure into a one-stage procedure.377,429,474 Specifically,

CHAPTER 67  Image-Guided Aspirations and Biopsies

2523

A

A

B

C FIG 67-176  A, CT scan shows a patient lying in a lateral position, with a large fluid collection (A) in the midpelvis. B, Repeat scan shows an axial slice of the cavity with a needle (arrow) centrally in the cavity. C, Sagittal reconstruction shows the needle position (arrow) and trajectory approaching abscess.

most cases of small diverticular abscesses can be treated with a single surgical procedure consisting of an en bloc resection of a segment of the sigmoid colon and a confined abscess. The entire problem is resolved with a single procedure. In cases in which the abscess is so large that this cannot be accomplished, surgical treatment would require two procedures: one to drain the abscess and divert the bowel, and a second to reanastomose the bowel. In such cases, percutaneous drainage can remove the infectious material, and a single surgical procedure can be performed to remove the offending segment of sigmoid colon (Fig. 67-180). We try to approach such lesions from an angle not traversing the bowel; we believe including the bowel within the pathway increases the possibility of fistulas. Neff and colleagues reported a series of 16 patients with diverticular abscesses.377 According to the authors, all responded to drainage. Eleven patients required sigmoid resection after the drainage, and one

patient had a three-stage procedure. Ten fistulas resulted; two required surgical treatment.

Infected Neoplasms.  Although experience has shown that percutaneous drainage of infected neoplasms is not curative,368 there has been some more recent interest in using percutaneous drainage as a palliative procedure. Our experience indicates that this is not a worthwhile procedure; such abscesses never heal because antibiotics are not delivered to the site through the tumor. A chronic drainage fistula typically develops, making skin care difficult.

Echinococcal Drainage.  Numerous authors have reported on the percutaneous drainage of echinococcal cysts.41,150,258,362,389,438 The method is simple and consists of typical percutaneous puncture and aspiration using US or CT. With the Seldinger technique, catheters

2524

PART III  Image-Guided Procedures

A

B FIG 67-177  A, Various devices used for a transrectal approach include the plastic sheath needle, plastic cannula, angiographic wire, and catheter. B, Simulated procedure using the bony pelvis of a skeleton and the equipment as noted in A. This shows the plastic sheath along the side of the finger. The needle is within the sheath (protecting the rectum and the operator’s finger), and povidone-iodine (Betadine) ointment occludes the end of the plastic sheath.

A

B FIG 67-178  A, CT scan of the pelvis shows a large abscess posterior to the bladder and anterior to the rectum. B, Follow-up CT scan shows catheter drainage through the sciatic notch. The cavity is much smaller, but there is also an abscess that has tracked posteriorly into the gluteal muscle.

between 12F and 20F are inserted. The cavity is irrigated with alcohol and povidone-iodine and left in the site for 30 minutes. The catheter is left to drain until fluid ceases. Concurrent treatment with albendazole also is important in the appropriate doses and duration. Anaphylactoid reactions occurred in several cases but were not fatal. The most definitive paper by Khuroo and colleagues reported 50 patients with hydatidosis; 25 were treated with surgery and 25 were treated by catheter drainage, and both groups were treated with albendazole.258 Percutaneous treatment was effective in 88%, compared with 72% by surgery. Complications occurred in 32% of the percutaneous group and 84% of the surgical group.

Amebic Abscess.  Typically, amebic abscesses respond to systemic treatment with amebicidal therapy; however, in some cases, the abscesses are resistant to drug therapy. Akgun and associates,10 Tandon and colleagues,485 and Hanna and coworkers203 reported on the use of aspiration and drainage in conjunction with amebicidal therapy. It was the consensus that percutaneous aspiration and drainage may expedite therapy and provide a cure when there is resistance to medical therapy.

Percutaneous Nephrostomy The same advantages for CT procedures are applicable to nephrostomy placements, but several different points should be emphasized. CT has been successful in a high percentage of cases; in one early report, we completed 12 of 13 attempts.182 Barbaric and associates reported a series of 148 procedures using CT and fluoroscopy.22 Their success rate was 97%, and there were no major complications. Authors such as Hagspiel and colleagues201 and Merkle and colleagues345 have used MRI for placement of nephrostomy tubes. The indication for CT guidance is limited, but I believe that most nephrostomies can be performed easily using either fluoroscopic or US guidance. The cases for which we use CT guidance are renal transplants, unilateral kidney, high-risk patients, unusual anatomic variation, or a kidney with a urinoma associated with the hydronephrosis (Fig. 67-181). Renal transplants are well suited for CT nephrostomies. Patients with transplants are high-risk patients who are immunosuppressed, and a higher margin of safety is needed. The transplants are mobile,

CHAPTER 67  Image-Guided Aspirations and Biopsies

R

FIG 67-179  Transsciatic drainage through sacroiliac notch for drainage of prostatic abscess, located anterior to the rectum (R). Note the needle (upper arrow) and contrast in the cavity (lower arrow).

2525

and the axis is unpredictable. CT may be needed to ensure proper placement of the catheter through the parenchyma. Nephrostomies of unilateral kidneys should be performed with CT because these cases present a high-risk situation. Potential damage or complications in such cases must be minimized. When a collecting system with hydronephrosis has ruptured and produced an adjacent urinoma, CT is better than other methods for several reasons. First, if the urinoma is displacing the axis of the kidney, the orientation can be appreciated. Second, if there is active excretion of contrast material, the perinephric space may fill with contrast  material, making it difficult to visualize the collection system under fluoroscopy. The technique is similar to that described for abscess punctures. The skin is prepared with the standard antiseptic preparation, and a local anesthetic is administered. An 18-gauge plastic-sheathed needle is used to puncture the renal pelvis. An angiographic wire is inserted, the plastic sheath is removed, and the chosen nephrostomy catheter is inserted. The longitudinal type of scan or fluoroscopy may be used after the puncture to insert the catheter. When the nephrostomy is performed under CT, several technical maneuvers are important. After the collecting system has been penetrated, a small sample of fluid should be aspirated for culture, then a small amount of contrast material should be injected immediately. This injected contrast material permits direct visualization under fluoroscopy when the catheter is being inserted. In addition, I recommend the 18-gauge Teflon-sheathed needle so that with successful puncture the angiographic wire can be inserted into the collecting system

A

A

C

B FIG 67-180  A, CT scan shows an approach to an abscess (A) posterior to the bladder and adjacent to the sigmoid colon (arrow) from diverticular abscess. B, CT scan shows insertion of a wire throughout the length of the cavity. C, Radiograph shows the injection of contrast material, showing a fistula to the diverticulum of the colon.

2526

PART III  Image-Guided Procedures

C C

FIG 67-181  CT-guided nephrostomy with patient in a right lateral position. Needle puncture of the collecting system (arrow) was performed between two cysts (C). The catheter was inserted using the Seldinger technique.

without additional manipulation (see Fig. 67-181). The patient can be transported to the fluoroscopy area without concern about displacement of the sheath or wire. In some cases, the final positioning of the catheter should be made under fluoroscopy, but in many cases adjustments can be made with localizing scans. If one has access to a fluoroscope attached to a CT scanner, this step is simplified.

Percutaneous Cholecystostomy Percutaneous cholecystostomy is widely accepted as a viable palliative alternative to surgical cholecystostomy in cases of acute sepsis from cholecystitis. Either US or CT may be used for performing the procedure.

Indications.  An acutely inflamed gallbladder producing sepsis is the sole indication for this procedure. Abnormal gallbladders have the expected finding of a very thick wall and may have associated inflammatory stranding in the adjacent tissues.

Contraindications.  There are two contraindications to the procedure. Rarely the gallbladder appears so inflamed that one might be concerned about rupture and spillage of bile into the peritoneum. The interventionalist must consider this possibility, although it is very unlikely because of the adherence of tissues from inflammation around the gallbladder. Unmanageable coagulopathy may preclude the procedure.

General Technique for Percutaneous Cholecystostomy.  The chosen modality can be either US or CT, depending on the anatomic approach desired. One can choose a transhepatic route or a peritoneal pathway. Anatomic pathway.  Either a transhepatic or a peritoneal approach may be used. During the early development of this technique, most authors preferred a transhepatic approach through the liver. The premise was that penetrating the gallbladder from the hepatic side would prevent spillage of bile into the peritoneum (Fig. 67-182). As the technique matured, it was determined that when necessary a transperitoneal approach could be used without adverse effects.

The pathway chosen should avoid uninvolved structures and may be either anterior or lateral. The target should be the fundal area of the gallbladder and not the neck region, because the neck is too small to easily accommodate the instruments used for this procedure. Preprocedure preparation.  The key element in preprocedure preparation is to ensure that the patient is receiving appropriate broadspectrum antibiotics. Just as with abscesses, bacteremia can occur during instrumentation of an infected gallbladder. Instrument choice.  This procedure is essentially a modification of abscess drainage, so one can choose to use a one-stick instrument or Seldinger technique. If using the Seldinger technique, we prefer an 18-gauge plastic-sheathed needle and a stiff wire; the stiff wire permits insertion of the catheter more easily. The size of catheter used varies from 5F to 7F depending on the size of the gallbladder lumen. When a single-stick catheter with stylet is used, the operator should choose one with a small enough coil to fit easily within the gallbladder lumen; in most cases, 6F is appropriate. Insertion of needle and catheter.  With the Seldinger technique, the central portion of the gallbladder lumen is targeted with the 18-gauge needle. The fluid recovered may be thin or thick; sometimes it is purulent, but in most cases, the fluid is dark green and thick. It is seldom foul smelling. After a return of fluid is obtained, injection of a small amount of contrast material is helpful to define the exact site and determine whether spillage of material occurs. If a small amount spills into the peritoneum or into the subhepatic space, it is seldom an issue. In contrast to other catheter insertions, we prefer a “floppy” wire that can be coiled easily within the gallbladder. Because the wall of the gallbladder may be fragile, we seldom use a stiff wire. The standard dilation and insertion of the catheter is performed; usually a 6F is used. With the one-stick device, incremental adjustments are made to position the catheter. Because of the weight of the device, it is best to support the weight of the catheter with folded towels during the movement of the patient in and out of the CT scanner. Postprocedure care.  Postprocedure care is straightforward. The catheter is left to drainage, and the patient is given the proper course of antibiotics. When the patient has improved sufficiently, an elective cholecystectomy can be performed. Patients who are unsuitable for surgery, such as patients in hospice, can have the catheter in place  long term.

Success and Complications.  Among the authors reporting, there is consistent agreement that these procedures are effective and provide effective palliation for acute cholecystitis.115,206,496 Complications vary and include bile peritonitis, bleeding, vagal reactions, hypotension, and acute respiratory distress. Rates are 4% to 10%.

Percutaneous Enteric Stoma A logical extension of the percutaneous puncture and catheter insertions is the development of percutaneous methods for insertion of catheters into bowel loops. Gastric tube placement can be performed for decompression of gastric bypass remnants or as a guide for surgical placement of a gastrostomy tube in difficult cases. An entrance site is selected according to the access available and the lack of intervening or overlapping structures. The catheter is inserted by means of a Seldinger technique with standard wire, dilators, and catheters. To increase the probability of a successful tube placement, there are two critical technical points to consider. First, during puncture of the wall, dilation of the tract, and insertion of the catheter, one should ensure that the bowel loop is well distended with gas so the wall is not flaccid but taut. This makes manipulation of the instruments easier. Second, the stiffest angiographic wire available should be used to enhance the exchange of dilators and the insertion of the catheter.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2527

A

B

C

FIG 67-182  A, CT scan shows a needle puncture of the gallbladder through a cuff of the liver. The needle was intentionally directed toward the gallstones (arrow). This was done in case there was a need for percutaneous removal of the stones later. B, Radiograph shows the needle in place adjacent to the gallstones (arrow). Theoretically one could consider performing this study under fluoroscopy, but the margin of the liver could not be visualized in this manner. C, Radiograph shows the insertion of the catheter in the gallbladder with contrast injection.

Colonic decompression by catheter insertion into the colon can be performed; CT and fluoroscopy are required for optimal placement into the chosen site. We reported the first image-guided cecostomy in a patient with toxic megacolon to facilitate direct instillation of antibiotics.185 CT is well suited to ensure that the entrance site is on the retroperitoneal side of the colon so that peritoneal spillage does not occur. Lynch and coworkers reported a series of percutaneous endoscopically assisted cecostomies in eight patients for pseudo-obstruction and severe chronic constipation; they were successful in seven of eight cases.320

they converge to form the right and left splanchnic nerves, which traverse the retroperitoneum lateral to the aorta. The splanchnic nerves penetrate the diaphragm and lie on both sides of the posterior mediastinum until they join higher plexuses and enter the central nervous system. Interruption of these sensory fibers prevents pain stimuli from being carried back to the brain, eliminating pain. Numerous methods have been used to interrupt these pathways, including surgical ablation, unguided percutaneous neurolysis, and guided percutaneous neurolysis.

NERVE BLOCKS

ment physicians used either an unguided method or fluoroscopic guidance to provide general localization of the treatment needle insertion relative to the vertebral bodies (soft tissues such as the aorta and plexus were invisible with standard fluoroscopy). These physicians used the L1 vertebral body to approximate the location of the aorta and the level of the celiac plexus. Kappis described localizing the L1 vertebral body with the needle positioned posteriorly behind the crura.249 Because of variation in individual patient anatomy and physician expertise, the results and complications were inconsistent among the reporting groups and institutions. Some centers have excellent results, others have poor success with a high incidence of complications.

Celiac Nerve Block One of the most exciting and gratifying areas of interventional radiology is CT-guided nerve block. The same types of nerve blocks that were previously performed blindly or by fluoroscopy can be performed better with CT guidance. It is important to review the rationale behind celiac nerve blocks to gain a better understanding of the process, side effects, and possible complications. The sensory nerve fibers transmitting pain sensation from the bowel, mesentery, capsule of the liver, and some portions of the retroperitoneum around the major vessels pass through a major neural plexus adjacent to the celiac artery called the celiac plexus. This large plexus contains ganglia from the sympathetic, parasympathetic, and sensory nerves. When the sensory fibers pass through this area,

Image-Guided Background.  Before CT guidance, pain manage-

Choice of Approach.  Historically, five anatomic approaches for celiac plexus block have been used for guided methods (unguided methods were posterior with only approximations of location relative

2528

PART III  Image-Guided Procedures

to the crura): (1) posterior approach including the retrocrural area, (2) posterior approach transcrural, (3) posterior approach subcrural, (4) anterior approach,291 and (5) transaortic approach. Our preferred approach is a posterior abdominal approach, which is subcrural. The approach is below the diaphragm and is performed at the level of the celiac plexus itself. Several decades ago, we began with a posterior transcrural double-needle approach, but with experience and refinement the current method evolved. Although our approach requires greater experience, we have been very pleased with the success and lack of any significant complications. The use of an anterior approach to destroy the plexus has been associated with more procedure pain and risk of pancreatitis.419 Although not reported, we have concerns about potential damage to the portal vein, hepatic artery, or pancreas.488 Fitzgibbon and coworkers reported a case of superior mesenteric vein occlusion associated with injection of alcohol from an anterior approach.128 The retrocrural approach, which is done above the insertion of the diaphragm, attempts to interrupt the pathway after the splanchnic nerves have formed and as the nerves course in the posterior mediastinum. This is safe but less effective because the pathway of the splanchnic nerves above the plexus varies, and anticipating their location for blocking is difficult and requires large volumes of ablative solution. Lieberman and Waldman reported a transaortic approach in which the authors planned a trajectory to traverse the aorta.302 After both walls of the aorta were penetrated, the alcohol was injected. Although it was reported as safe, we believe this method is inherently dangerous, and one should avoid the aorta and target the plexus instead. This report confirms that complications are unlikely if the aorta is inadvertently punctured when the plexus is targeted. Over many years, we have tried most of the methods described (except transaortic) and learned many things. As with most CT procedures, our experience is extensive and dates back to the first such CT procedure in 1977 (Fig. 67-183).196 We prefer the subcrural posterior abdominal approach, which interrupts the pathway at the plexus before formation of the nerves. The plexus is penetrated directly using a posterior angled trajectory so that the entire needle pathway is beneath the diaphragm. This approach lessens possible complications by avoiding penetration of uninvolved structures and tracking of medication into the chest or abdomen. Chemical ablation is likely to be used for the foreseeable future, but ultimately thermal ablation is expected to become prevalent.

FIG 67-183  This CT procedure, which was done in 1976, shows the placement of two skinny needles for performance of a celiac nerve block (see text).

Indications and Patient Selection.  The indications for celiac nerve blocks are straightforward. All patients who have severe pain in the upper abdomen supplied by the splanchnic nerves are potential candidates. From a CT anatomic perspective, pathologies in the pancreatic, peripancreatic, retroperitoneal, and renal hila are appropriate if the tumor is not so extensive that it has invaded the plexus or outside the innervation of the plexus. The procedure usually is performed in concert with the clinical pain service. With this approach, the patient gets the best care and political issues are minimized. The patient is evaluated by the pain physician, who refers the case to the radiologist. The radiologist places the instrument, and the pain physician injects the medication and follows the patient in the postprocedure period. Selecting candidates for this procedure has been difficult because clinical physicians tend to defer celiac blocks until the disease is too extensive for their performance. Commonly such patients are managed with increasing doses of opiates until there is no symptomatic relief. When that point of the disease is reached, the tumor has usually invaded the plexus itself or outside the innervation, making the procedure impossible. Although cancer is the typical indication, some patients with pancreatitis may be suitable. The only issue with such patients is that because their disease is benign, they outlive the effects of the procedure because alternative pain pathways develop. When pain recurs, the level of pain is typically less than the original intensity.

General Technique for Celiac Nerve Block Preprocedure history, review of images, planning, and informed consent.  The patient’s history and imaging studies are reviewed. The most important part of the interview is to ensure that the pain symptoms correspond to the celiac distribution, because that is the only area improved by this nerve ablation. The outside imaging studies are reviewed. If tumor is noted to have invaded the plexus directly, the procedure should not be performed; it is unlikely to produce relief and is more likely to result in complications (see “Complications,” later). As with other procedures, the patient’s laboratory studies should be normal. The patient should not be on any anticoagulant medication. The procedure and side effects (see “Side Effect [Not Complication] of Orthostatic Hypotension,” later) should be explained to the patient. The consent form should be signed by the patient and a family member because most patients are on long-term narcotic drugs, which alters their mental competency from a legal perspective. Another important point relative to pain control is that such patients should have their pain medication stopped or reduced before the procedure so the interventionalist can judge the effectiveness of the test block. To establish as a reference point the level of preprocedure pain so we can assess the effectiveness of the preliminary anesthetic block, as described subsequently, we ask the patient to grade the pain by rating the degree of pain on a scale of 0 to 10, 0 being no pain or normal and 10 being unbearable. This scale is used for a reference value after the test block to help decide if the alcohol should be injected. To facilitate this, we ask that the patient not have pain medication the morning of the procedure. Finally, if the patient has new symptoms suggestive of an active process that differs from the long-term symptoms, the procedure should be deferred until the cause of the symptoms can be determined. One would not want to mask symptoms of any ongoing pathologic process, which might jeopardize the patient in the short term. Procedure steps.  During the procedure, the patient’s vital signs are actively monitored and maintained as appropriate. The patient is positioned in the right lateral decubitus position with the left side up. During a routine diagnostic scan, we look for any signs of an active

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

2529

C

FIG 67-184  Appropriate positioning for a celiac nerve block. A, This scan shows introduction of the needle into the retroperitoneum on the left. B, With cephalad angulation, the needle is passed behind the renal pedicle and then passed beside the aorta. C, CT scan shows the final placement of the needle tip beside the superior mesenteric artery behind the pancreas. Even if inadvertent penetration of the vessels should occur, any adverse developments are unlikely (see text under other methods as proposed by Lieberman and Waldman).302

process that should be treated. During the course of our practice, we have diagnosed and deferred patients with emphysematous pancreatitis, small bowel obstruction, and gallbladder disease that required treatment rather than a neurolytic block for pain, or disease so extensive that a block would not be beneficial. We exclude patients with severe local tumor extension into the musculoskeletal portions of the retroperitoneum. A celiac block would not significantly ablate the pain in such patients. Selection of entrance site, trajectory, and target.  A diagnostic scan is performed to assess the status of the tumor process to ensure that progression has not occurred, which might preclude the procedure. Because internal structures can shift with positional changes, the findings on a diagnostic supine scan cannot be used to extrapolate an approach, so we perform the diagnostic scan in the position in which the procedure would be performed. The patient is positioned on the right side with the left side up, to provide visualization of the anatomy and pathology to be used for the procedure. A dynamic bolus scan with IV contrast material over the areas highlights vascular structures. This position provides easy access to the left side of the celiac plexus (Fig. 67-184), which is the side treated when a unilateral block is performed. The entrance site chosen should avoid the spinal erector muscle if possible because if a patient moves or rotates slightly, the needle trajectory may be altered. The trajectory chosen is a posterior angled abdominal approach below the insertion of the crura. We do not penetrate the kidney, bowel, or other organs because we believe that the complications previously reported by the blind method were the result of inadvertent injection of a neurolytic agent into such structures. The skin entrance point is at the level of the renal vein, with slight trajectory angulation cephalad toward the area above the superior mesenteric artery or adjacent to its base by the aorta (Figs. 67-185 and 67-186; see Fig. 67-184). The needle is angled slightly medially and cephalad so that the final placement would be adjacent to the side of the aorta at a level between the celiac artery and the mesenteric artery. With this precise placement, we can interrupt the nerve fibers in the plexus itself; because of the precise placement, a smaller volume of alcohol (≈15 mL) can be injected (older techniques use 50-100 mL of alcohol). Needle selection.  Instead of the 22-gauge or 20-gauge needles used during the early development of this method,50,196 we use the

FIG 67-185  CT scan of a patient in the right decubitus position after injection of air into the retroperitoneum. The air (arrow) crosses from left to right in the plexus area.

larger-caliber 18-gauge needle with plastic sheath for three reasons. First, the plastic sheath is left in situ after correct placement of the needle, and if the patient moves during the time required for the procedure, structures are not displaced or damaged. Second, if a vessel has been entered (e.g., the spinal artery of Adamkiewicz or lumbar arteries), brisk back-bleeding occurs, heralding the event; the needle can be repositioned or removed as needed. If the ablative solution is not injected intravascularly, organ damage or spinal cord injury may occur. Third, the flow of medication during injection is easy; if there is resistance to the injection of the solution flow during injection, the needle is likely present in a fascia or even the wall of the aorta. The needle should be repositioned because it could be in the wall of a vessel. Catastrophic damage to the aorta has occurred at some institutions. Smallcaliber needles have a high resistance to injection, so this sign is not valid with those devices. Insertion and placement of needles.  Insertion and positioning of the needles depends on whether a unilateral or bilateral block

2530

PART III  Image-Guided Procedures

A

B

C

D FIG 67-186  A, CT scan with a patient in the right decubitus position shows a needle adjacent to the superior mesenteric artery. Note the small amount of tumor encasing the anterior portion of the vessel. B, Injection of air shows the gas localized by the artery. It does not cross the midline, indicating that coverage of the right side will not occur and right-sided treatment is needed. C, Placement of a right-sided needle is performed. The tip is between the crus of the diaphragm and the vena cava. D, Injection of air shows gas distributing only on the right side.

is being performed. We have found that in most properly selected cases, a unilateral block from the left side would be effective. We prefer the left side over the right because in this position the aorta is seen and serves as an excellent landmark to avoid; navigating a needle between the vena cava and aorta with a right-sided approach is difficult. The determination of a unilateral or bilateral block is made after the test injection of bupivacaine (Marcaine) and air, as noted subsequently. The needle is incrementally inserted and readjusted as needed to enter the plexus target zone. The celiac plexus is located on the anterior surface of the aorta between the celiac and superior mesenteric arteries and drapes slightly to the sides of the aorta. It is optimal to have the needle tip precisely at this site (see Figs. 67-185 and 67-186). If an appropriate placement of a single needle is made, the next steps of injection are taken.

Injection of test dose of local anesthetic and “air” marker.  After successful placement of a left-side needle has been

made, a test injection of local anesthetic (bupivacaine) is performed for several reasons. There can be congenital variation in the location of the plexus, and confirmation of pain relief confirms that this nerve plexus carries the pain sensation. If the pain is not relieved, it does not make sense to inject the alcohol, because any liquid medication would not reach the corresponding nerves. Lack of pain relief can be due to congenital variation of the location or infiltration of the plexus by tumor, which prevents flow of the liquid medication into the site. Whereas injection of alcohol or other sclerosing agent is unlikely to produce pain relief, it is likely to disperse into inappropriate areas and cause complications (see “Complications,” later). We have replaced lidocaine in the test dose with bupivacaine. Not only does bupivacaine produce rapid local anesthesia, but its effects persist so that the subsequent injection of the alcohol is not as painful. We inject 15 mL of 2% bupivacaine and 5 cc of room air into the site. The “air” marker indicates where the flow of alcohol will occur, and

CHAPTER 67  Image-Guided Aspirations and Biopsies one can adjust or remove the needles as indicated. A repeat scan is performed over the area of the needles to assess the distribution of the air in the plexus area of the retroperitoneum. If the air crosses from left to right, covering the entire plexus area around the superior mesenteric artery and the anterior region adjacent to the aorta, a unilateral block, rather than a bilateral block, can be performed because the entire plexus area can be treated (see Figs. 67-185 and 67-186). If the air does not cross, a bilateral block is performed. If the air tracks extensively throughout the retroperitoneum away from the plexus site into the psoas or mediastinal area, the sclerosing agent is not injected. To do so might cause unintended damage to other structures (Fig. 67-187). We let the medication permeate for 10 minutes and then ask the patient to grade the pain. If there is good pain relief, the patient reports a remarkable improvement in the grading scale (i.e., a reduction from 8 or 9 to 0 or 3), and if the air has not tracked into an inappropriate area, the final block is performed. The alcohol for the final block should

2531

not be injected if there is no relief of pain or if there is an inappropriate tracking of the air. In some cases, when the patient has very severe pain from cancer and the patient and clinician insist, we inject the alcohol despite a poor response (but this is uncommon). If a bilateral block is required, we typically finish the left-sided procedure completely and then insert the needle on the right side. Injection of alcohol.  If there is a return of blood from the needles, we do not inject the alcohol, because injection of a vessel with alcohol results in destruction of the end organ (e.g., kidney, spinal cord). At this time the alcohol is injected very slowly into the unilateral or bilateral needles. Because of the effective outcome of our procedures, we have reduced the volume of alcohol injected from 25 mL to 15 mL of alcohol. If a unilateral block is performed, a single dose of 15 mL is injected. If a bilateral block is performed, 15 mL is injected on each side. (Plastic syringes must be used because with glass syringes the evaporation of alcohol is so rapid that the barrels “lock up.”) Uncommonly the unilateral block provides only left-sided relief, so the right side must be blocked to ensure adequate pain relief (see Fig. 67-186). The patient may be rolled more obliquely or prone if needed to provide access for the second needle insertion and treatment. The right-side celiac block is performed in a similar fashion with the approach passing through the space between the vena cava and aorta (a difficult procedure).

Side Effect (Not Complication) of Orthostatic Hypotension.  M

A

When the sensory, sympathetic, and parasympathetic nerves are effectively destroyed, one expects to see temporary side effects; two are commonly observed. First, short-term orthostatic hypotension is produced by splanchnic pooling of blood caused by interruption of the sympathetic fibers. The dilatation produces pooling of blood; if patients attempt to stand after a completed block, they lose consciousness. Second, alteration of bowel function occurs; although one would expect diarrhea, we have had patients complain of constipation and diarrhea. Both side effects must be treated appropriately, the orthostatic hypotension immediately after the procedure and the bowel changes in the following weeks.

Postprocedure Care.  For orthostatic hypotension, the patient must

B

lie flat for 12 hours and gradually have the posture elevated from flat position to upright over an 8-hour period. Adherence to these guidelines is recommended because the orthostatic hypotension can be severe. It is better to avoid testing for orthostatic hypotension, because unnecessary discomfort and distress for the patient may result. One patient in our institution was made unconscious three times by three different resident physicians who wanted to test her orthostatic hypotension by taking her blood pressure in an upright position. She had a very bad headache after these “tests” but eventually did fine after the appropriate bed rest.

Successful Outcomes.  Three large international series have been C FIG 67-187  A, Diagnostic CT scan shows a large mass in the head of pancreas (M). There is suggestion of infiltration (arrow), but it was decided to proceed with the celiac block. B, Needle placement in the space anterior and adjacent to the superior mesenteric artery was accomplished. Note the air in the celiac area (straight arrow). Also, there is some air posteriorly (curved arrow), which has tracked back because of infiltrated tumor. C, Higher scans were obtained to evaluate for diversion of local anesthetic and air. The CT scan shows that air did not enter the celiac area properly and was diverted into the retroperitoneum under the diaphragm (arrow). Alcohol was not injected because this would have given a poor result.

reported by Prasanna,410 Marra and colleagues,327 and Gimenez and coworkers.148 Prasanna reported on 200 cases with good results for reduction of pain using a posterior unilateral left-sided block.410 Marra and colleagues327 and Gimenez and coworkers148 used an identical anterior approach employing a 22-gauge needle with injection of alcohol and reported significant difference in the outcome. Marra and colleagues reported that 21% were pain free at 15 days,327 and Gimenez and coworkers reported that 61% were pain free at 6 months.148

Factors Affecting Success and Complications.  Extent of tumor invasion strongly affects the outcome of celiac nerve blocks. Akhan and colleagues showed that when tumor has invaded the plexus, alcohol

2532

PART III  Image-Guided Procedures

FIG 67-188  Radiograph of a patient with fluoroscopic localization for celiac nerve block shows injection of contrast material (arrow). Permanent paraplegia developed after the procedure, despite the interventionist’s considerable fluoroscopic experience and meticulous fluoroscopic technique.

does not distribute and may not ablate the nerve tissues.11 To improve the results of plexus blocks further, we have introduced a new technique that we believe offers great promise for the future of pain relief and perhaps even treatment of pancreatic cancer. Relative to patient cooperation during the procedure, pain control during the alcohol injection is key. Local anesthesia, as described earlier for a test block, helps relieve pain. Also, slow rather than fast injection of the alcohol reduces the intense pain. Finally, use of the transcrural method and meticulous technique described previously maximizes benefit and reduces complications. Potential problems are reduced with careful placement of needles in target areas and avoidance of uninvolved anatomy. Some severe complications can be avoided by not injecting the alcohol if there is a blood return. Using the unguided methods, individuals have had catastrophic outcomes (Fig. 67-188).

Complications.  In the past 30 years, we have had only two transient complications. One patient had hematuria for 2 days, and one patient developed a small subcapsular renal hematoma. Complications from blind or other guidance methods are many and severe, including death, paraplegia, loss of kidneys or liver, hemorrhage, and mediastinitis (see Fig. 67-188).

Cryoablation of Celiac Plexus Cryoablation of the celiac plexus is the preferred method when tumor invasion exists. Because of the intense clinical competition fostered by the new emphasis of the “business” approach to medicine it is likely that radiologists will only do a small number of these procedures. This is because pain centers and anesthesiology departments have used nonguided methods for many years and play a major role in this field. However, a comprehensive radiology interventional approach requires that interventional radiologists maintain their skills in this field. The need for thermal ablation for selected patients is obvious. If one reviews the process of chemical ablation, one can appreciate that ablation of tumor or tissues depends on uniform and effective distribution of the ablative liquid agent into the target area. When a liquid is injected into tissue, however, the liquid follows pathways of least resistance and does not always distribute evenly or to the intended site

if nonpermeable tissue intervenes. With celiac nerve blocks, if tumor blocks the pathway for the alcohol, it distributes inefficiently and does not produce the expected results. This theoretical problem with alcohol distribution was reported by Akhan and colleagues.11 They noted that when tumor had invaded the area of the plexus, it affected the outcome of the block and increased complications. Patients with high-grade invasion did not have an effective block, and there was increased incidence of complications because the alcohol dispersed into inappropriate areas. The solution to this problem is to use thermal ablation, which can treat through solid tumor and still destroy the nerves. We approached this problem a number of years ago and have evaluated the feasibility of using radiofrequency or cryoablation to ablate the celiac nerve plexus. In the last edition we briefly discussed the use of RFA (heat), which we have now discarded. Although such ablations were effective in a number of cases, the inability to see the heat damage was stressful for the operator. This stress was caused by the inability to see in real time the effects or damage to the plexus, adjacent structures, and vessels. We have possible role of radiofrequency but use heat ablation using radiofrequency but have found that cryoablation is a better methodology. Cryoablation is a better methodology because the location of the celiac plexus is very deep and surrounded by many vessels and other structures. It is reassuring to any operator performing this procedure to be able to visualize the size and extent of the cryoablation “iceball” (Fig. 67-189; see Fig. 67-190). When an inexperienced person contemplates this procedure it is quite stressful because of all of the contiguous vessels (i.e., portal vein, superior mesenteric artery, etc.). However, this modality is perfectly suited because cryoablation does not affect major vessels because of the flow, which provides a “heat source” to the site. It is virtually impossible to damage the vessels compared to the possible damage for the chemical approach discussed above. On the other hand, cryoablation can penetrate into the tumor and provide a better result, which avoids the issues reported by Akhan et al. This technique is quite straightforward and similar to the other methods. The anatomic approach is from the back through the retroperitoneum. Efforts are made to avoid intervening anatomic structures. Although there have been few complications reported in the literature, we do not accept the technique others have reported, going “through” the aorta. The benefit of CT is being able to avoid such structures.

TUMOR ABLATION Local treatment of tumors using percutaneous methods has evolved naturally from the biopsy and drainage techniques described previously. After imagers developed and refined these techniques for placement of instruments, the stage was set for the ablative methods and technology to evolve. The enthusiasm for local treatment originated from the seminal work of several surgeons who performed meticulous resection of multiple nodules throughout the liver. These authors, most notably Minton from Ohio State University, showed increased survival of these patients. Although this approach was considered remarkably controversial at the time, it represents the underlying basis for the current philosophy. As a result of the general acceptance of the surgical principles, the methods appropriately evolved to simpler means of surgical treatment (cryoablation and heat ablation at open laparotomy) and ultimately to percutaneous application of the same methods. The radiology and the surgical literature43,85,86 confirm that the RF cumulative survival outcomes parallel the surgical results for primary and metastatic diseases. Treatment of primary hepatocellular 

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2533

B FIG 67-189  A, CT scan with patient lying right side down during IV contrast material injection. Note the soft tissue density surrounding the aortic, celiac, and splenic arteries. Such tumor invasion precludes the use of alcohol for celiac block because it prevents alcohol from flowing into the nerve plexus. Also note the small needle medial to the left kidney, which was inserted as a marker to help localize the anatomy during cryoprobe insertion. B, The long cryoprobe is noted adjacent to the aorta, and the smaller needle serving as a guidance marker is seen posteriorly. This was needed because the margin of the aorta and plexus were obscured by tumor invasion. In the typical case, the fat planes around the vessels and the plexus are visible because invasion is not so dense. The tube-shaped iceball is seen on the lateral side of the aorta. There is no danger of damaging the aorta because blood flow provides protective heat that prevents damage.

carcinoma is either surgical or percutaneous depending on the character of the tumor and status of the patient. When patients have lesions smaller than 4 cm, the percutaneous method suffices, but larger lesions are better treated by surgery. Patients who are not surgical candidates because of the size of lesions, proximity to vital structures, compromised status of the liver (cannot endure loss of functioning liver parenchyma associated with lobar liver resection), or generally poor health also are candidates for the percutaneous method. With this remarkable change in tumor therapy supported by oncologic surgeons, the local ablative therapies have been developed, refined, and widely introduced into standard clinical practice. The literature is replete with reports by many authors who have devoted considerable effort and time to develop, introduce clinically, and refine these methods. The most notable of these are Goldberg,153,154 Solbiati,462,463 Lewin,296 Livraghi,311,312,314 Curley,84,85 and others. Publications have been so numerous and expansive, it is difficult to reference properly all who have made contributions, so in this chapter we do our best to provide due credit, believing those not mentioned would be gracious enough to understand any errors of omission.

Standardized Terminology and Criteria With the remarkable growth and dissemination of ablative methods, a vast clinical and research experience has been accumulated, but understanding, comparison, and validation of the different and diverse applications have been impaired by lack of standard nomenclature.

The purpose of standardization is to facilitate refinement and development of this important new technique by permitting better precision and communication in this field. The international working group convened by the Society of Interventional Radiology Technology (SIRT) assessment committee has defined and refined this topic. The primary source can be reviewed for more complete information if needed (Goldberg and colleagues154), but the essence of the report has been paraphrased and quoted here. Tumor ablation is defined as the direct application of chemical or thermal therapies to a focal site for eradication or substantial destruction of the tumor. With chemical ablation, such as ethanol or acetic acid, the term alcohol ablation should replace the term percutaneous ethanol instillation. When scientific results are reported, the route of administration (e.g., intravascular, interstitial) should be noted with the specific chemical, rate of delivery, and volume. Thermal ablation uses either heat (RF, laser) or cold (cryoablation). The energy is applied; terms such as “irradiation” should be avoided. Applicators are used for all thermal ablations. Applicators for RF are electrodes, those for microwaves are antennas, those for lasers are fibers, and those for freezing are cryoprobes. RF electrodes are multitined, expandable tines, internally cooled electrodes, or perfusion electrodes. Cluster electrodes describe three or more closely spaced electrodes. Whenever possible, the algorithm of energy deposition should be noted, such as ramped energy deposition or impedance regulated. The model of the generator should be noted. Additional

2534

PART III  Image-Guided Procedures

parameters such as monopolar, bipolar, power in watts, and amperage in milliamperes should be provided in publications. Laser ablation should replace the terms laser coagulation therapy and laser interstitial photocoagulation. The term direct or interstitial indicates that fibers are inserted into the tissue. The laser source (e.g., YAG, holmium) with wavelength should be specified. The fiber type, including tip characteristics (e.g., diffuser tip, scattering dome), should be noted. The number of applications and technical parameters, including watts, duration of energy application, amount of energy applied, and number of applications, should be noted. Microwave ablation is used for all electromagnetic methods of inducing tumor destruction with frequencies from 30 MHz to 30 GHz. Type of applicator and frequency should be noted. US ablation can be performed extracorporeally or with a direct percutaneous applicator. The term high intensity is not relevant. Cryoablation describes all methods of destroying tissue by means of low-temperature “freezing.” The temperature of the cryoablation should be specified. For publication, the system type, gases used, probe dimensions, length, and number of freeze/thaw cycles should be specified. Blood flow effects should be called heat sink effect, referring to vessels larger than 1 mm, or perfusion-mediated tissue cooling, which encompasses larger heat sink vessels or capillary level perfusion.

Role of Imaging Imaging has five separate roles for tumor ablation: (1) planning, (2) targeting, (3) monitoring, (4) controlling, and (5) assessing treatment response. These terms are self-explanatory, with the exception of controlling, which means using image-based information during monitoring to modify the procedure. This includes repositioning the applicator with thermal feedback or image findings and automatically terminating the ablation at a critical point. Pathologic and imaging findings may not always coincide. Specifically the imaging traits may not coincide with the extent of tumor as visualized on the pretreatment scan or the induced imaging changes produced by the applied treatment. The “zone of death” is referred to as coagulation. Thermal sites have a central “white zone” corresponding to coagulated tissue and a “red zone” corresponding to hyperemia (not visible in ex vivo specimens). Both zones should be stated in scientific publications.

Postprocedure Imaging After ablation, two imaging findings can be observed: (1) zones of perfusion and (2) changes in modality units (i.e., attenuation units on CT, signal intensity on MRI, and echogenicity with US). In all organs, but especially the liver and kidney, areas of minimal enhancement may later prove to be dead tissue. The cause is thought to be “pseudoenhancement” from technical reasons or leaking capillaries at the treatment margin (see “Long-Term Follow-up”). Ablative margin is used to describe the margin of normal tissue ablated to ensure complete treatment; this margin is appropriate in selected organs and cases. Transient hyperechoic zone is the increased echogenicity around a site after ablation, which clears after 30 to 90 minutes. Benign periablational enhancement can be seen in pathologic examinations and imaging, which reflects reactive hyperemia. This persists for 6 months and typically is a thin uniform rim, which differs from the irregular enhancement typical of residual tumor. Irregular peripheral enhancement indicative of tumor differs in that it is typically scattered, irregular, and nodular. Involution of coagulation describes the process by which the zone of induced coagulation is eliminated by the body. The terms “shrinkage” and “regression” should be avoided.

Reporting of Tumor and Sizes Index tumor describes the initially identifiable tumor before ablation. Tumors are classified by size. Small tumors are 3 cm or less, intermediate tumors are 3 to 5 cm, and large tumors are 5 to 10 cm. Comparison of zones of coagulation should be made from a volumetric measurement of three dimensions. Although the International Working Group Committee convened by SIRI is silent about using different modalities for follow-up and correlating the results, we believe that once a modality has been used for the first index size, it should be used for all subsequent comparisons.

Standardization of Follow-up Terms Technical success addresses whether the tumor was treated according to the selected protocol with regard to complete coverage by the thermal treatment. Documentation of the effects can be determined from MRI or a contrast-enhanced CT scan. Technique effectiveness should refer to a defined time point after the ablation (it may vary but should be defined; e.g., immediate postprocedure, 1 week) when complete ablation was confirmed by the chosen imaging modality. For percutaneous thermal ablation, a time window and number of treatments (should not exceed four) should be prospectively determined; if complete ablation cannot be achieved, the tumor should be classified as unsuccessfully treated. Primary effectiveness is the percentage of tumors eradicated after initial treatment. Secondary or assisted effectiveness rate includes tumors that have had repeated ablations. Complete ablation is an accepted term, but “partial ablation” has no defined meaning, so its use is not condoned. Tumor palliation is an accepted term if the treatment was used to relieve clinical symptoms and complete ablation was not intended; this is most applicable for carcinoid symptoms or paraneoplastic symptoms. Debulking should be used to describe reduction of tumor burden. Treatment failure is associated with incomplete therapy or tumor progression, new foci of disease in the target organ, or distant metastases. These entities should be distinguished and reported as such. When the distinction between residual tumor or new tumor growth cannot be made or is not relevant for treatment planning, the term local tumor progression should be used. Complications should be reported using the Society of Interventional Radiology standard table. The definition of death is self-evident, and this event should be reported within the 30-day period after the treatment ends. Causality related to ablation should be determined. A major complication is an event that leads to substantial morbidity and disability, increasing the level of care, or results in a hospital stay or extension of stay. Immediate complications occur within 6 to 24 hours, periprocedure complications occur within 30 days, and delayed complications occur after 30 days. Side effects are undesired complications that occur frequently but rarely result in substantial morbidity. Side effects include pain, asymptomatic pleural effusions, blood collections, and thermal damage to adjacent structures with negative sequelae. Postablation syndrome is a transient self-limited symptom or sign complex of low-grade fever and general malaise. It is dependent on the volume of necrosis produced. The syndrome may persist for 2 to 3 weeks.

Ablative Methods Three major methods have been developed and disseminated widely throughout the medical subspecialties: (1) injection of chemical liquid agents (e.g., alcohol, acetic acid, hypertonic saline, hot saline); (2) thermal ablation by heat generated by RF, microwave, or laser; and (3)

CHAPTER 67  Image-Guided Aspirations and Biopsies thermal ablation by cryoablation, which uses freezing to kill tumor cells. Other novel methods are being investigated, such as DNA vectors, polypeptides, hormones, and chemotherapeutic agents, but these have yet to be refined and tested in clinical trials.

Alcohol and Other Chemical Agents.  Before RF refinement, chemical ablation—typically with alcohol—was the most common minimally invasive method for treating hepatocellular carcinoma and metastatic lesions because of its elegant simplicity and ease of use. It is low tech, economical, and available to any center in the world. Although in recent years chemical ablation for small hepatocellular carcinoma and metastatic disease has been predominantly replaced by thermal ablation, in the absence of the equipment or technical expertise needed for thermal ablation, chemical ablation provides an effective method for the local treatment of cancer. Alcohol ablation produces cytoplasmic dehydration, which causes coagulation necrosis, endothelial damage, platelet aggregation, local tissue ischemia, and ultimately fibrous reaction. The advantages of alcohol are that it is only locally destructive, it can be metabolized readily, and it does not typically cross major anatomic boundaries. It is effective for hepatocellular carcinoma because this cancer typically has well-defined borders that confine and restrict the spread and localize the effects of alcohol. Metastatic lesions are less well suited because the margins are less well defined, and ethanol spreads in a more unpredictable fashion. Two methods for alcohol ablation have been reported, best described by Shiina and associates447 and Livraghi and colleagues.308,309 The methods are similar in their approach for injection of the material but differ in the quantity and intervals of injection. Both methods use a small-caliber needle that is inserted into the most distal portion of the lesion and then reinserted for injection of the alcohol to the lateral, posterior, caudal, cranial, and anterior walls of the mass. Shiina and associates calculated the amount of alcohol to be used by using a volumetric calculation of a sphere447: Volume = 4 3 π(R + 0.5)3 The volume is in cubic centimeters, and R is the radius of the lesion. An additional 0.5 cm is added to account for spread into the margin of the lesion with the normal liver, with anticipation that this may produce a chemical margin of the lesion. The injection is made over 10 to 20 minutes, with some modification based on the patient’s compliance. Typically a second treatment is performed several months later. Livraghi’s group used US to monitor the injection of alcohol, with no precalculated volume specified.308 Their procedures were performed under general anesthesia to permit better pain control and management of ventilation. With their method the lesion was monitored with US and the injection was continued as long as the “flow” from the end of the needle continued without undue injection pressure. They carefully infiltrated all walls and the central portion of the lesion. Repetitive injections were performed over several months if repeat imaging studies showed evidence of nonresponse or recurrence. Postprocedure side effects and follow-up were similar with both methods. There was invariably a transient increase in the liver enzyme profiles, which returned to normal within 10 days. Blood alcohol levels were elevated after the procedure, and Shiina and associates447 noted mild intoxication in 25% of patients. Successful treatment of confined or encapsulated tumors showed significant reduction in AFP levels. Complications were frequent. The most common adverse effect was localized pain during and after the procedure. Pain persisted for several days, with upper abdominal discomfort enhanced by palpation, but this was easily controlled by analgesics. Patients may have spread of

2535

alcohol to uninvolved areas of the liver, which produces necrosis of normal tissue. Both studies reported intraperitoneal bleeding when a cuff of normal liver did not exist below the capsule. Barnett and Curley24 and Livraghi and coworkers314 summarized the general results of alcohol ablation. For hepatocellular carcinoma, success depends on the size of the lesions. With lesions less than 5 cm, ethanol injection was successful in 75% of lesions that underwent subsequent resection. With lesions 5 to 8 cm, the ablation rate was 60%. Livraghi and coworkers reported a 1-year survival rate of 90% and a 3-year survival rate of 63%.314 Both of these authors and their groups now favor RFA over alcohol injection. Ethanol injection has much less effect for metastatic lesions and is seldom used, in favor of RF treatment.462,465 Most authors recommend using alcohol injection when lesions are in an inopportune site for RF or when lesions are larger than 3 to 4 cm. Although other chemical agents have been injected, such as acetic acid386 and gel cisplatin,114 they failed to garner general acceptance because of the variable results and complications owing to irregular distribution of the liquid agents, need for repetitive treatments, and nonuniform response. A change from use of these chemical agents to thermal ablation occurred with the completion of studies directly comparing alcohol ablations with thermal ablation by diverse authors and even by the proponents of alcohol ablation. Multiple authors performed comparison studies of alcohol injection and RFA. Livraghi and colleagues reported better results with hepatocellular carcinoma tumors smaller than 3 cm for RF than alcohol injection309; however, there were higher complication rates with RF. Ikeda and coworkers studied small hepatocellular carcinoma lesions and found almost identical results of success and complications but favored RF because it required fewer treatments.235 Lencioni and associates292 compared RFA and alcohol injection in a series of 102 patients. They concluded that RF treatment was confirmed as an independent prognostic factor for local recurrencefree survival rates with multivariate analysis (adjusted relative risk 0.20; 95% confidence interval 0.05, 0.73; P = .015).

Thermal Ablation by Radiofrequency.  Application of heat effectively kills tumor cells in two methods, depending on the temperature. At 42°C, tumor cells may die in preference to normal cells because of the induction of hypermetabolism from increased temperature. Because tumor cells lack the homeostatic controls of normal cells, they typically die a metabolic death over many minutes. With temperatures at 45° to 60°C, intracellular proteins denature and bilipid membranes melt, resulting in cell death. With temperatures beyond this level (80°110°C), heat coagulation necrosis occurs for all tissues. The temperature levels and effects reported by different authors vary slightly, but this is probably due to minor differences in the basic models studied and differences in cell types. A seminal paper by Goldberg and coworkers reported the effects of three parameters of tissue ablation size: probe gauge, duration of treatment, and temperature on the lesion volume.153 In an experimental model, they noted that diameter of the ablated site increased with greater probe gauge and duration of heat application. Increased temperature had little effect on the treatment size. One critical factor is the influence of blood flow microscopically and macroscopically on the effect on local tissues and the survivability of normal and tumorous cells. Flowing blood acts as a heat sink or radiator effect so that heat is moved away from the immediate site of the treatment region. This result is helpful because one wishes to preserve normal vessels near the tumor, but it has a negative effect because tumor cells may survive in immediate proximity to the vessels. The residual tumor cells within treated tumors or especially at the margins of the treated areas are the cause of local recurrences.85,463,465

2536

PART III  Image-Guided Procedures

A

B

C FIG 67-190  A, RF Radionics cool-tip probe is available in single or cluster configuration. The single shaft permits precise placement in a small lesion. The cluster probe permits treatment of larger lesions. B, RF Spoke/Tine configuration permits adjustment to provide appropriate coverage to different-sized lesions. C, Cryoprobes come in various sizes. Multiple probes can be placed in various positions so that they can be frozen to produce larger treatment areas (see Figs. 67-208 and 67-210). (From Rose SC et al: Lung cancer and radiofrequency ablation. J Vasc Interv Radiol 17:927–951, 2006.)

Among the various application devices, RF heat coagulation remains the preeminent method for heat coagulation. Cryoablation represents the second most commonly used device and seems to be growing in popularity. Laser, microwave, and US heating have been attempted and are being refined, but there is little indication they would provide heat deposition superior to that of RF. Radiofrequency devices.  Numerous RF devices are used for these treatments. General configuration of the probe is the most recognizable difference. One type is a single straight probe and the other is a spoked device shaped like an umbrella (Fig. 67-190). A flexible shaft is available with either configuration. As noted earlier, we use these probes for smaller lesions and difficult sites to target. De Baere and colleagues compared the two types of RF probes and found that although the cooled-tip straight probe produced larger lesions, the lesions produced by the expandable device were more reproducible, uniform, and spherical.93 Shibata and associates compared the watercooled straight probe with the expandable probe in a series of 74 patients (83 nodules, ≤3 cm)446; there was no significant difference in the primary technique effectiveness or rates of major complications.

Shibata and associates also preferred the straight probe for smaller lesions because it is easier to position in difficult anatomic sites.446 Other devices for heat thermal ablation are available, including laser, microwave devices, and bipolar devices, but we have no experience with them. Shibata and associates compared RF thermal ablation with microwave and found equivalent effectiveness for hepatocellular carcinoma, but microwave required more treatment episodes.445 In our opinion, any of the devices work if the operator is familiar with the technology and method and is meticulous with the planning and execution of the procedure as reported in the literature.

Thermal Ablation by Cryoablation.  Cryoablation has been available for many years, but the instruments were so large that the methodologies were confined to surgical methods at open laparotomy. This method depends on repetitive freezing that produces death of tumor cells by creating ice crystals within the cells. Two freezing cycles were required to ensure reliable death of tumor cells. The advantage of  this approach is that large tumors can be treated. The disadvantages were that there is a likelihood that large cracks in the liver can occur

CHAPTER 67  Image-Guided Aspirations and Biopsies and require surgical repair. Another adverse outcome was a systemic  reduction in platelets, which would be problematic. Smaller percutaneous probes have become available more recently, and the experience being gained is favorable. There are three significant advantages of cryoablation over RFA. First, the freezing process can be followed more easily on the favored guidance modality of CT. The iceball produced is easily seen, and complete treatment is verified more easily. Second, because of this property, one is better able to treat large lesions effectively, since the overlap of contiguous treatment sites can be better appreciated, increasing the chances of an effective ablation. Third, Allaf and colleagues have shown that the pain produced by cryoablation is much less than RFA.12 This reduced pain lessens the need for narcotics or general anesthesia or both. Several cryoablation devices are commercially available; the reader is referred to trade shows or meetings for specifics of each brand. Distinction of different properties of the various vendors is not completely apparent, and there has not yet been research reporting the advantages and disadvantages of each (see Fig. 67-190).

Microwave Ablation.  Microwave ablation is another method for performing thermal ablations that is gaining acceptance. It greatest advantage is that the generation of heat is extremely rapid. Rather than requiring 30 to 40 minutes per heating cycle and sometimes needling multiple applications as with other methods, microwave delivers the “killing” thermal dose within a matter of minutes. It is so efficient, it seldom requires repeat treatment except when the target area has been missed or incompletely covered by the ablation The disadvantage of microwave ablation is that the treatment zone is not well visualized on routine scans. A repeat scan during the IV injection of contrast material may be required to clearly document the treatment site. The reader is referred to the current literature for comparison studies about efficacy. For technical details about microwave equipment and technique, the appropriate manufacturer should be consulted.

General Techniques for Tumor Ablation We discuss here numerous facets of thermal applicators and techniques using our own experience and the data available in the literature. Similar to many groups, our preferred primary imaging modality is CT scanning for multiple reasons, with US having a limited role. Occasionally we combine real-time US with CT, which can be an effective synergist. The real-time US is used for initial targeting of a small site, and CT follows with final assessment of location, treatment effects, and imaging for any complications, such as hematoma. The best overall assessment of the target tumor during the procedure is provided by CT, whether one uses a single straight probe, cluster probe, spoke probe, or cryoablation probe. Because of the ability to image metallic objects clearly, every part and component of the device can be seen and accurately localized into such lesions. When microbubbles of gas are generated by RFA, they can be well seen and characterized by CT (Fig. 67-191) (they are more difficult to see with US or MRI). Some authors believe the detection of this gas to be important because it indicates during real-time monitoring of ablative procedures that effective coagulation of tissues is occurring, especially in the lung. Excessive gas might impair the heat deposition in the tumor and the effectiveness of the procedure. Gas may be produced in adjacent structures and has been associated rarely with complications, so accurate imaging is important. Gas emboli to the brain have occurred, and animal models have shown that it is related to high-energy input, so adjustment of the treatment might be considered if excessive gas were noted.

2537

With RF thermal ablation, one can well observe the margins of tissue damage in the target lesion only if repetitive IV injections of contrast material are given (see Fig. 67-191). With cryoablation, the treatment area encompassed by the iceball can be visualized well and accurately because the site becomes less dense with the expansion of the frozen tissue. Different types of RF and cryoablation probes have clearly defined advantages and disadvantages. The choice of applicator type and specific probe design is based on the size, visibility, shape, and location of the lesion. The anatomic pathway available to approach the lesion may dictate the use of a thinner more flexible probe if it is located high under the diaphragm. For small (1-2 cm) target areas or lesions in critical areas where the anatomy is very close and difficult, a straight RF probe device is preferred. The most notable examples of this are in the porta hepatis, hilum of kidney, peripancreatic areas (e.g., celiac nerve block), or osteoid osteomas. For small (2-4 cm) lesions in a difficult area that requires precise placement and angulation of the probe, we prefer the spoke-shaped RF probe (Fig. 67-192). Because the shaft of these probes is smaller than the cryoprobes, angulation and positioning are performed more easily. Probes with a flexible shaft permit some bending during movement of the patient into the scan diameter, minimizing any issues related to patient breathing and contact with the scan gantry. Finally, the ability to insert the tines quickly for securing the tip placement is timeefficient, not requiring the delay needed for the partial freeze localization of the cryoprobe. Repetitive adjustments and placement can be made quickly. For larger lesions with unencumbered access, the cryoprobes are best for several reasons. Large lesions are treated more easily because multiple probes can be combined to cover the lesion completely at one time, simplifying the treatment. The iceball is easily seen and is more consistent in its production of a large treatment area because of the ease of overlap. Finally, the discomfort created for patients during ablations varies between the two forms of ablation. Allaf and colleagues12 reported that lower doses of narcotics and hypnotics are needed for cryoablation than for RFA.

Liver Thermal Ablation

Indications.  The general indications for liver ablation are the presence of a mass amenable to percutaneous treatment. The treatment is intended to destroy the cancer or reduce associated symptoms. The best expectation is to prolong life or improve morbidity. For neuroendocrine neoplasms metastatic to the liver, systemic symptoms produced by hormone release can be reduced by destruction of the tissues using RFA according to Henn and coworkers.216

Surgery Versus Percutaneous Ablation for Liver Lesions.  As one might expect, the outcomes of properly selected and planned ablative procedures compared with meticulous surgical resections are virtually identical. Although initially some controversy existed in the literature between the treatment methods, a consensus has developed for patient selection and triage to achieve such favorable results. Surgical outcomes.  Advocates of surgical treatment, such as Marco,326 have noted that the overall and disease-free survival were significantly higher in the surgical group than the percutaneous RF group (P = .002 and P = .001, respectively). The advantage of surgery was more evident for Child-Pugh class A patients and for single tumors of more than 3 cm in diameter. Results were similar in two groups for Child-Pugh class B patients.

Different minimally invasive techniques for ablative method.  Numerous authors have reported at meetings that when meticulous technique is used, there is little difference in the outcome

2538

A

D

PART III  Image-Guided Procedures

B

E

C

F

FIG 67-191  A, CT scan shows a subtle low-density metastasis (arrow) from breast cancer. In most cases, it is best to target the central part of the lesion so that progressive extension of the tines would treat the entire sphere of the tumor more completely. B, CT scan shows the device anchored by slight extension of the tines. C, Repeat CT scan shows slight extension of the tines during application of RF heat. Note the low-density area surrounding the tips, indicating edema from heating. D, To ensure treatment of a 1-cm margin as recommended by most authors, the tines can be extended farther out into the adjacent normal liver, and ablation can be repeated. E, Lower portion of same mass as in D was treated in the area of the vessel (not RF probe). F, Repeat CT scan during administration of a bolus of contrast material shows complete ablation of the site with a 10-mm margin (oblique arrow). Also note the small gas bubble in the upper region (vertical arrow). Note the wedge-shaped peripheral area indicating damage to the corresponding vessel (a good thing).

CHAPTER 67  Image-Guided Aspirations and Biopsies

G

H

I

J

2539

FIG 67-191, cont’d  G, Repeat postablation diagnostic scan shows a new metastatic site in the lower liver margin (arrow), which was not seen on the preprocedure scan. Contrast-enhanced diagnostic study was not performed before the procedure in an effort to reserve the contrast material for perfusion studies during the ablation procedure. H, Preprocedure scan performed 6 weeks prior shows no evidence of lesion. Some authors have noted that although individual breast cancer lesions respond to RF, it is common for new sites to grow quickly.

for percutaneous, surgical, or laparoscopic methods, but only limited documentation has appeared in peer-reviewed journals. Kuvshinoff and Ota reported that the success of RFA performed at open surgery or by laparoscopy is directly related to size in a similar fashion as the percutaneous method; lesions less than 4 cm have a high success rate.276 Percutaneous outcomes.  Experienced groups report favorable outcomes with percutaneous methods. Machi and associates,322 Lencioni and coworkers,292 and Raut and associates416 reported favorable outcomes with hepatocellular carcinoma (see Fig. 67-192). Lencioni and coworkers292 used RFA as first-line treatment in a group of 206 patients with cirrhosis Child class A or B. The nodules were less than 5 cm; 9% were excluded because of inappropriate location. Overall survival rates were 97% at 1 year, 67% at 3 years, and 41% at 5 years. Survival depended on Child classification (P = .0006) and multiplicity of tumors (P = .013). The 1-, 3-, and 5-year recurrence rates were 14%,

49%, and 81%, respectively, for the emergence of new tumors and 4%, 10%, and 10%, respectively, for local tumor progression. As an adjunct to surgery, percutaneous ablative methods show good results for palliation before liver transplantation and for recurrence after surgical treatment. Brillet and coworkers47 and Mazzaferro coworkers337 showed that ablative methods could be used for temporary treatment of patients waiting for liver transplantation. They studied the possible adverse effect of treating hepatocellular carcinoma before transplant on the recurrence rate and survivorship of transplant candidates, and they reported no adverse effects. With regard to recurrent hepatocellular carcinoma after hepatectomy, Yang and colleagues reported that RFA was effective and safe for late recurrences after surgery.525 The survival of the early recurrence group was less than the control group or the late recurrence group (late recurrence survival was 45.9 months compared with control group survival of 16.4 months).

2540

PART III  Image-Guided Procedures

D A

E

B

F

C FIG 67-192  A, CT scan from an elderly woman awaiting liver transplant. A small hepatocellular carcinoma appears as a subtle mass in the right lobe of liver (arrow) with contrast material. B, A small probe in place within the lesion after ablation. The lesion is difficult to see without contrast material. C, IV contrast material may be required to visualize ablation effects best. After contrast administration, a low-density ablation area can be seen (arrow). D, Three months after RF, the eccentric focal enhancement is noted in the upper margin of a low-density fluid mass (arrow). E, Eight months after RF procedure, CT scan shows resolution of focal enhancement. F, Fifteen months after RF, the cystic mass is reduced in size (arrow), confirming resolution of the tumor and RF site.

CHAPTER 67  Image-Guided Aspirations and Biopsies With metastatic disease, especially colorectal disease, there are conflicting data, some suggesting that surgery is better than ablative therapy, and some suggesting that ablative therapy is essentially equivalent. Kornprat and associates,270 Pawlik and colleagues,401 and Aloia and coworkers14 presented data indicating that for solitary and multilesion metastases from colorectal cancer, surgery is the treatment of choice. Other authors, such as Abitabile and associates, propose RFA as the treatment of choice.1 Authors reporting on the ablation of breast cancer are equivocal in their results (see “Factors Affecting Success for Liver Ablations,” later).

Contraindications.  General contraindications include uncorrectable coagulopathy and severe comorbidities affecting complications and recovery. It has been suggested that if lesions are too close to the gallbladder, hilum, or bowel wall, injury may occur (although there are some contrary data—see “Factors Affecting Success for Liver Ablations,” later). A distance of 1 cm from these structures is considered adequate to prevent severe damage. Likewise, close proximity to major vessels is considered a relative contraindication because such vessels reduce the efficiency of heating and reduce the local coagulation benefit. The presence of multifocal disease in widespread sites is a relative contraindication. Although most cancer patients are extremely motivated, one should not embark on treatment of numerous lesions unless all members of the cancer team and the patient have come to a consensus about prognosis, chemotherapy, surgery, image-guided treatment, or even hospice considerations. Technical feasibility should not take precedence over appropriate humane patient care.

General Technique for Liver Thermal Ablation Preprocedure history, review of images, planning, and informed consent.  Preprocedure steps are taken as noted earlier, including physical examination, clinical and medication history, and review of all imaging and clinical data. Medication history is especially important to determine if any anticoagulative drugs are taken (either prescribed or OTC). Laboratory studies, including CBC and coagulation, are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectation, an appropriate witnessed informed consent is obtained. A diagnostic scan with imaging assessment of lesions before the procedure is important for selection of the probe, appreciation of vascular structures of anatomy, and assessment of anatomy. We believe imaging studies should be performed within 2 to 3 weeks prior, to lessen the probability that multiple new lesions may arise, changing the clinical context of the patient (see Fig. 67-191). A preliminary plan, including probe choice and number, anatomic approach, trajectory, probes to be used, and technique, is helpful. This plan enables the technical staff to ensure that the proper type and number of probes are available. The probes are too expensive to have large numbers on the shelf waiting for use and becoming outdated. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam or fentanyl or both as needed. Local anesthesia is administered as described earlier; either deep sedation or general anesthesia is used. Use of sedation with general anesthesia may lessen the depth of general anesthesia required, which is beneficial to the patient. During the procedure, vital signs and pulse oximetry should be monitored continuously with automated devices. Immediately before the procedure, the patient should receive broad-spectrum antibiotics, such as ceftriaxone, in appropriate doses. The specific antibiotic and dose should be chosen based on the patient history (e.g., allergy, renal failure). Although infections are unlikely,

2541

this is a small step to minimize the occurrence of an infection caused by a random hematologic contamination.465 Although we have not seen any infectious issues with liver procedures, we have had a patient who developed a localized abscess after a kidney ablation when an unrelated focus of diverticulitis “seeded” the site. Procedure steps.  If RF is being used, the grounding pads must be appropriately attached to the skin as recommended by the manufacturer to enable good conduction of the current from the electrode to the ground. The skin should be shaved to permit good contact. The pads should be attached with the long side of the pads perpendicular to the long axis of the body, taking care that there is good contact providing good “grounding” and avoiding local skin burns. One plans appropriately depending on the ablation type and type of probe used. The most important factor for effective treatment is providing effective coverage of the target site with either the single probe or the multiple probes, as selected. For small lesions it may be possible to use a single probe, but large lesions require multiple probes, cluster probes, or multitine devices, which can provide adequate coverage. It is better to use multiple probes rather than reposition probes because repositioning is too time-consuming. Careful image-guided placement of the treatment probes is crucial for a successful outcome and avoidance of unanticipated complications. The probe or probes are positioned and repositioned as needed to target the lesion correctly and avoid adjacent uninvolved structures. If a normal structure is in jeopardy, fluid or air can be injected either to conduct thermal or dry effects away or to insulate the tissues of concern. The intent of monitoring the lesions under imaging is twofold: (1) to document complete ablation of all viable tumor and (2) to ensure that a tumor-free cuff has been created in the adjacent liver. Both of these factors have been shown to be major determinants of success (see “Factors Affecting Success for Liver Ablations,” later). CT, US, and MRI have been used to monitor ablations63; however, most authors prefer CT because of its consistent diverse traits. CT permits accurate monitoring of metal instruments, vessels, gas formation, edema, hemorrhage, and perfusion changes with IV contrast material. At our site, real-time US has been beneficial for initial placement of a probe if the lesion is single or small, but otherwise CT is the “workhorse.” Minami and colleagues353 reported that standard realtime US was not effective clinically, but contrast-enhanced harmonic imaging was useful for ablation of liver lesions. MRI guidance also has been studied and refined (Fig. 67-193), but it has not assumed a major role for ablations. Lewin and colleagues,297 Aschoff and colleagues,18 and Merkle and colleagues346 have reported that MRI produces changes in T1 and T2 that are well visualized during real-time MRI, and that IV gadolinium affords assessment of vascular changes and tissue destruction. After the treatment probes have been placed according to the plan, the ablative steps are taken either to heat or to freeze the probes and tissue according to the recommendations of the vendor and the literature. The reader is referred to the literature or refresher courses for the most current methods relative to number of treatments, temperature, repetitive cycles, and adjustments of probes as appropriate. Vendor representatives are a good source of current information because of the FDA oversight and their vested financial interest for ensuring proper use of the equipment. Such vendors have literature and contact information available for clinical experts using their devices. Earlier devices required some manual adjustments of the equipment during the process, but most are virtually turnkey with computer feedback mechanisms. With thermal ablation, changes in the tissue are difficult to visualize unless IV contrast material is given to enhance viable tissue and show

2542

PART III  Image-Guided Procedures

A

B

FIG 67-193  A and B, Turbo-spin echo T2-weighted oblique axial (A) and coronal (B) images obtained during MRI-guided RFA. During the ablation session, T2-weighted and STIR images allow assessment of the developing zone of tissue damage, seen as a hypointense region surrounded by a bright ring of edema at the tip of the RF electrode. These images are intermittently acquired during the ablation session to ensure ablation of the entire tumor and a thin rim of surrounding normal tissue.

the loss of enhancement in thermally treated areas. If scans taken during this time do not show lack of enhancement and tissue destruction in the entire site, modification of the probe placement and retreatment should be performed (see Figs. 67-191 and 67-192). For thermal ablation in the liver, we prefer the multitine devices, which provide flexibility relative to positioning the device and adjusting the location of the tines to enlarge the size of the ablative zone (see Fig. 67-191). With each placement of the device, the position can be secured to prevent movement of the device during the scan by extending the tines a short distance; this creates the effect of an anchor (see Fig. 67-191B). When the probe has been positioned properly, incremental areas can be ablated by insertion of the tines (see Fig. 67-191C and D). Repositioning is necessary if the lesion is elongated or not spherical to ensure that the entire mass is treated (see Fig. 67-191E). Final assessment of the ablation treatment should be made with a contrast study (see Fig. 67-191F through H). With cryoablation, observation of the treated tissue is much easier. The iceball produced during the treatment appears as low density because the attenuation of the tissue decreases as the iceball expands within the tissues. After all probes have been positioned and the treatment has been performed, one can assess whether complete coverage of the tumor target has been accomplished. Postprocedure care.  Short-term postprocedure care is standard, with bed rest for 4 to 6 hours, assessment of vital signs, and laboratory tests (e.g., hematocrit, liver enzymes, BUN, creatinine). Depending on symptoms and laboratory values, the patient may be discharged within 24 hours after the procedure. Patients are treated supportively with IV fluids as needed and advancement to oral intake as tolerated. Any deviation from expected improvement over the first 24 hours warrants repeat scanning to evaluate for any postprocedure complications, such as hematoma or inadvertent damage to other tissues. A postablation flulike syndrome develops within the 10 days of the procedure in approximately one third of patients treated by RFA. Dodd and associates reported that 36% of patients develop symptoms, which

consist of fever, malaise, chills, delayed pain, and nausea lasting for 3 to 5 days.103 The severity depended on the volume of tissue ablated. Only conservative treatment was required. Wah and coworkers also reported a postablation syndrome in approximately one third of all patients.506

Imaging Follow-up.  Postprocedure imaging follow-up should be performed 24 hours after the procedure to establish a baseline study and detect any unexpected complications. Imaging should be repeated at 3-month intervals for at least 18 months and at yearly intervals for several years until the mass has undergone expected evolution and disappearance. One can use either CT or MRI for follow-up in the liver. If the patient has any issues with renal function or contrast material, MRI has proven to be equivalent or better for assessing any subtle changes. Long-term follow-up.  For long-term follow-up, CT or MRI, but not US, is suitable for assessment of treatment. Solbiati and colleagues showed that acute appearance of US changes cannot predict ultimate tumor necrosis.463-465 PET has been used, but with some limitation, as noted subsequently. The imaging appearance of the treated tumor evolves over time, showing a predictable sequence of changes over months (Figs. 67-194 to 67-196).69 Immediate postprocedure findings on imaging typically show some effects of edema, inflammation (Fig. 67-197), and a small amount of hemorrhage, as one might expect. According to Choi and colleagues, the size of the necrotic defect immediately after treatment should be larger than that of the pretreatment tumor, reflecting ideally a 5- to 10-mm margin clear of tumor.69 The margins should be sharp, delineating the edge of normal liver from the necrotic tissue (see Fig. 67-196); the delineation becomes more distinct as time passes. A confined hematoma in the site of treatment is common (see Figs. 67-191 and 67-194), appearing as high density on CT or increased signal on T2 MRI scans. Another common finding is the presence of gas in the lesion or the portal veins (see Fig. 67-195). Oei and associates reported the presence

CHAPTER 67  Image-Guided Aspirations and Biopsies

2543

1 month

FIG 67-194  Early changes after complete RFA in a 32-year-old man. A CT scan obtained 1 month after ablation shows gas in the defect and diffuse peripheral enhancement (arrows) indicating hyperemia. (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

3 months

FIG 67-195  Same patient as in Figure 67-194. Nonenhanced CT scan obtained 3 months after ablation shows resolution of the gas. It shows high-attenuation hemorrhage within the low-attenuation defects in segments III (arrowheads) and IV and VIII (arrow). (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

1 month

Pre RFA

A

B FIG 67-196  Complete ablation in a 65-year-old woman with metastatic colon carcinoma. A, CT scan obtained before ablation shows a metastatic lesion in segment V (arrow). Because the duodenum abuts the lesion, the percutaneous approach would bear the risk of bowel injury. B, CT scan obtained 1 month after ablation shows a low-attenuation defect (arrow) larger than the pretreatment metastasis and with a smooth margin. (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

2544

PART III  Image-Guided Procedures

3 weeks

FIG 67-197  Benign peripheral enhancement in a 61-year-old man. CT scan obtained 3 weeks after RFA shows a smooth, thin, regular rim of enhancement (arrow) along the ablation defect that is consistent with granulation tissue. (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

of intraportal gas after RFA and RFA/ethanol technique.384 In 25 procedures (74%), gas was found in portal vein branches; in 30 procedures (88%), gas also was found in tumoral and peritumoral tissues. Also, there was no significant difference in frequency of portal venous gas between the ablation methods. Recurrence of tumor at the ablation site has a typical appearance on CT or MRI. The margin of the treated site should be smooth and clearly defined from the adjacent normal liver. The clear delineation is amplified by the adjacent normal liver (see Fig. 67-191) with administration of IV contrast material. The lesion has rounded margins with some undulating contours produced by the close proximity of a large vessel (see Fig. 67-196). Within the first 6 months, uniform concentric enhancement of the wall may occur from inflammatory enhancement produced by the healing process (see Fig. 67-197). Irregular enhancement typically indicates tumor recurrence (Fig. 67-198), but occasionally eccentric benign enhancement occurs and resolves (see Fig. 67-192). Chopra and colleagues noted that the local intrahepatic recurrences appeared in three patterns of abnormal enhancement: nodular, halo, or gross enlargement.71 In the 6-month period after ablation, the inflammatory healing response may produce a false-positive appearance, suggesting tumor recurrence. Curley reported that the presence of inflammatory vessels in the treatment site may be asymmetric and indistinguishable from tumor.84 The enhancement reduces on subsequent scans and resolves after 6 months, clarifying the status of this spurious finding. Recently Wu and colleagues519 have reported that vasoactive constrictive drugs can cause sufficient vasoconstriction to eliminate this false positive finding. MRI is a very sensitive tool for detecting and confirming tumor recurrence even when CT may be subtle (Fig. 67-199A). Typically,

nonviable tumor shows low signal on T2 images but increased signal with tumor recurrence (see Fig. 67-199B). Dromain and coworkers noted that MRI may be slightly more sensitive for residual or tumor recurrence than CT.107 We concur, but during the 6-month postprocedure period, the greater sensitivity may result in more false-positive results. PET scanning can confirm the presence of tumor recurrence with typical FDG uptake (see Fig. 67-199C). PET scanning shows uptake of FDG in recurrent tumor and inflammatory changes from the ablation procedure. Differentiation of the two is possible by configuration and the appearance of specialized technique. Authors have found that delayed scans with PET may distinguish inflammatory areas for tumor because inflammation SUV values reduce at 1 to 2 hours. Curley noted that in a small study of 11 patients in which CT was equivocal in 6 patients 6 months postprocedure after resolution of inflammatory changes, PET was able to diagnose recurrence.84 The problem of using PET for postablation detection of tumor was discussed by Okuma and coworkers388 and Norbert,379 who noted the importance of a significant time delay to permit the resolution of periablation inflammatory response. Veit and associates502 reported a small series of 16 patients in whom RFA for colon tumor had been performed. They noted that CT PET was superior to CT alone, with an accuracy of 68% compared with 47%.

Factors Affecting Success for Liver Ablations.  Experience derived from a learning curve and meticulous technique are crucial factors. Poon and colleagues documented that morbidity and diseasefree outcome directly depend on experience.409 It is likely that the differences of opinion expressed by authors about the effect of tumor cell type are related to experience and technique. Meticulous technique for imaging assessment of probe placement and treatment results is crucial409; imaging should be used as an indicator for adjustment of the probe site and thermal application. Permpongkosol and coworkers reported that similarly with cryoablation, lesion size varied significantly in different organs (liver, kidney, lung) despite identical technical parameters, undoubtedly dependent on tissue characteristics such as blood flow and air content.403 If the expected destructive lesion is not created, repositioning and retreatment should occur. Montgomery and associates emphasized the need to use imaging to follow the treatment effect because of the variability in size and volumes of ablation that occurred despite the use of identical technical parameters.357 Treatment margins are a well-recognized parameter that affects the success of tumor ablation. Nakazawa and coworkers studied the impact of multiple factors that might have an impact on the recurrence after a single small episode of RF treatment.373 Of the factors statistically analyzed (which included clinical and biochemical features, tumor characteristics, tumor location ≤5 mm from intrahepatic vessels, needle biopsy before treatment, and presence of ablative margin of ≥5 mm), it was shown that location of a vessel within 5 mm (P = .0292) and an ablated margin of less than 5 mm (P = .019) correlated with tumor recurrence. Size of liver lesions is a primary determinant for successful treatment. Smaller metastatic lesions are more likely to be treated successfully by percutaneous methods than large ones. Solbiati and colleagues465 and Livraghi and colleagues311,314 showed that lesions smaller than 3 cm are likely to have a good outcome. In a multiinstitutional study involving metastases, Livraghi and colleagues314 showed that lesions less than 2 cm were cured, lesions 2 to 3 cm had a cure rate of 70% (recurrence rate of 30%), and lesions greater than 3 cm had a cure rate of 58% (recurrence rate of 42%). The success and recurrence rates with hepatocellular carcinoma are similar. Hepatocellular carcinoma lesions less

CHAPTER 67  Image-Guided Aspirations and Biopsies

6 weeks

1 week

A

2545

B FIG 67-198  Incomplete ablation and local recurrence in a 62-year-old woman with metastatic colon carcinoma. A, CT scan obtained 1 week after RFA shows an ablation defect in segment III with an irregular rim of enhancement (arrowhead). More diffuse ill-defined enhancement (arrow) is around the defect in segments IV and VIII. These findings are indeterminate. B, CT scan obtained 6 weeks after ablation shows complete ablation in segment III (arrowhead), with resolution of the enhancing rim. The margin of the low-attenuation defect is smooth. In segments IV and VIII, recurrence or incomplete ablation (arrow) is seen as a lowattenuation margin along the border of the defect. (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

than 2.5 cm are effectively destroyed, whereas large lesions are prone to recur.43,85,230,269,521 Livraghi also studied the effectiveness of treating medium (3.1-5 cm) and large (5.1-9.5 cm) hepatocellular carcinomas; complete necrosis was produced in only 47.6%.307 Tumor location related to the capsule for liver tumors was studied by several authors, but the results are inconclusive. Cho and associates,67 Poon and colleagues,409 and Chung and coworkers74 reported that location of tumor nodules, either subcapsular or not subcapsular, had no effect on the outcome of ablations. Hori and colleagues230 and Komorizono and coworkers269 found that tumors in the subcapsular location were the most likely to recur in a group of 99 hepatomas; we theorize that differences in probe placement techniques may be the differential factor. Cell type of tumor likely has an impact on the probability of success. Colorectal lesions have responded well according to multiple authors. Breast cancer ablative therapy has had mixed reviews as reported in the literature. For breast cancer, Gunabushanam and colleagues reported a high success rate for technical success, but the recurrence rate with new metastases remote from the treatment site was 50% (see Fig. 67-191).171 These results suggest that the ablative treatment itself is effective, but it is likely there were small metastatic sites that probably were invisible on imaging. From our experience, this seems likely because often breast cancer lesions are poorly visualized, since the lesions are infiltrative and difficult to detect until they reveal themselves with a definite focal mass (see Fig. 67-191G and H). Ovarian lesions as reported by Gervais and associates143 have responded well,

and cytoreductive approaches are useful. A differing view expressed by Montgomery and coworkers was that cell type had no effect.357 Differentiation of liver tumors greatly affects outcome as reported by Kim and associates.262 They reported that the therapeutic results of hepatocellular carcinoma greatly depended on histologic grades of tumors. Using the Edmondson-Steiner grade system of I, II, and III (III the most poorly differentiated), the survivorship varied accordingly. The 5-year cumulative survival rates for groups I, II, and III were 71%, 44%, and 43%, respectively. The 4-year cancer-free survival rates for groups I, II, and III were 39%, 10%, and 0%, respectively. The presence or absence of cirrhosis and tissue temperature during treatment had no relationship to ablation size according to Montgomery and coworkers.357

Side Effects.  Approximately one third of patients develop a flulike illness in the several days following ablation, as reported by Dodd and coworkers103 and Wah and colleagues506 (see “Postprocedure Care,” earlier). Kim and coworkers reported that patients with decompensated cirrhosis manifested changes in liver function as shown by transient elevation of bilirubin and aminotransferase levels.261 They reported on the appearance of the biliary system after hepatic RFA. In a series of 389 patients with 521 hepatocellular carcinomas (571 sessions of RFA), these authors reported mild upstream dilation of the bile ducts in 82% of patients, bilomas at the site of ablation in 5.8%, and both in 11.6%. Only one patient required later treatment for cholangitis (Figs. 67-200 and 67-201).

2546

PART III  Image-Guided Procedures

23 months

A

24 months

B

24 months

C

FIG 67-199  Local recurrence in a 56-year-old woman with metastatic colon carcinoma. A and B, CT scans obtained 23 months (A) and 24 months (B) after RFA show an apparent increase in the size of the ablation defect in segment VIII, which indicates recurrence. The area of low attenuation extends behind the right hepatic vein (arrowhead). The small lymph node near the esophagus (arrow) has enlarged. C, Positive PET study showing recurrent tumor (arrowhead). (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

Complications.  Extensive review of the literature reveals numerous complications reported in large series. Although complications occur, the rate of complications compares favorably with surgical and laparoscopic techniques. Mulier and colleagues reviewed all cases of RF treatment of liver tumors up to December 2001 and collated 3670 patients reported in 82 independent reports.369 Complications occurred in 8.9%: abdominal bleeding in 1.6%, abdominal infection in 1.1%,

biliary tract damage in 1%, liver failure in 0.8%, pulmonary complications in 0.8%, dispersive pad skin burn in 0.6%, hepatic vascular damage in 0.6% (Fig. 67-202), visceral damage in 0.5%, cardiac complications in 0.4%, myoglobinemia or myoglobinuria in 0.2%, renal failure in 0.1%, tumor seeding in 0.2%, coagulopathy in 0.2%, and hormonal complications in 0.1%. Complication rates were 7.2%, 9.5%, 9.9%, and 31.8% after percutaneous, laparoscopic, simple open, and

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

2 months

2 years

FIG 67-200  Biliary stricture and parenchymal atrophy in a 71-year-old man with hepatocellular carcinoma. A, CT scan obtained 2 months after ablation shows complete ablation (arrowheads) in the lateral segment of the left lobe. A subtle dilation of the biliary ducts is seen centrally (arrow). B, CT scan obtained 2 years after ablation shows progression of the biliary ductal dilation with atrophy of the lateral segment of the left lobe (arrow). (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

combined open approaches. Mortality rates were 0.5%, 0%, 0%, and 4.5%, respectively. Livraghi and colleagues reported complications in a large multicenter group consisting of 2320 patients with 3554 lesions.314 The lesions included 2320 hepatocellular carcinomas with liver disease, 693 metastatic sites, mostly colon cancer, and 17 cholangiocarcinomas. The main determinants of complications were multiple probe placements with no impact on lesion size or probe type. Six deaths (0.3%) were noted, including two caused by multiorgan failure after intestinal perforation, one case each of septic shock after Staphylococcus aureus– caused peritonitis, massive hemorrhage after tumor rupture, liver failure after stenosis of the right bile duct, and one case of sudden death

2547

of unknown cause 3 days after the procedure. Additional major complications occurred in 50 (2.2%) patients. The most frequent of these complications were peritoneal hemorrhage, neoplastic seeding, intrahepatic abscesses, and intestinal perforation. A third large series by Curley and associates studied early and late complications in a series of 608 patients who had either open or percutaneous RFA of hepatic malignancies.86 The treatment-related mortality rate was 0.5%. Early complications developed in 43 patients (7.1%). Early complications were more likely to occur in patients treated with open RFA (33 [8.6%] of 382 patients) compared with percutaneous RFA (10 [4.4%] 226 patients; P < .01), and in patients with cirrhosis (25 [12.9%] complications in 194 patients) compared with noncirrhotic patients (31 [7.5%] complications in 414 patients; P < .05). Late complications arose in 15 patients (2.4%), with no difference in incidence between open and percutaneous RFA treatment. The combined overall early and late complication rate was 9.5% Tumor seeding in the probe pathway has been extensively described and may occur in the abdomen, abdominal wall, or pleural region. This has been confirmed by Llovet and associates315 (12.5%), Livraghi and colleagues314 (0.9%), and Jaskolka and colleagues240 (4%). Factors affecting liver complications.  Avoidance of damage to collateral organs is important for minimizing complications. Some authors, such as Choi and coworkers, have reported RFA to be safe if a small amount of the target tumor abuts the colon or stomach.68 They noted no complications when 5 mm of the tumor abutted these organs. Authors have reported the use of fluid and gas to provide a conductive layer between the thermal ablation site and any adjacent normal structures. Laeseke and coworkers reported the use of dextrose water and saline to prevent thermal damage to adjacent uninvolved structures, such as the diaphragm.277 In their animal model, dextrose prevented damage better than saline did, permitting smaller sites of damage (perhaps because of conductivity of saline). Kariya and associates reported the use of injected carbon dioxide to produce “insulation” between the thermal site and normal structures.251 We have chosen to use “air” in a similar fashion and have used it for selected cases of cryoablation (see Fig. 67-197). Unintentional biliary damage has only rarely been observed. Raman and colleagues reported in a swine model that the circulation of chilled saline remarkably reduced attempted injury to the ductal system.413 Hepatoma has a high incidence for tumor seeding in the needle track. Incidence reported by Llovet and colleagues315 is 12%, by Livraghi and associates,313 0.9%, and by Jaskolka and associates,240 4%. Livraghi and associates313 believe a subcapsular location has a statistical correlation, so approach through an oblique path of normal liver may be helpful to avoid this occurrence. Numerous authors have noted an increased incidence for tumor seeding and bleeding for subcapsular lesions. Cho and colleagues evaluated lesions that were subcapsular and lesions that were not subcapsular and found no difference in complications.67 Livraghi and associates showed an association of bleeding with subcapsular location.311 Proximity of the gallbladder to an ablation site has been reported to be an issue for increasing complications with RFA, but Chopra and coworkers70 reported different results. In 6 patients with 14 lesions close to the gallbladder, they reported no significant complications after RFA. They theorized that the patients developed a mild iatrogenic cholecystitis that was self-limited, but they stated that with proper care and probe placement such lesions could be treated safely. We support this concept but point out that care should be taken not to ablate the wall of the gallbladder intentionally, especially if it abuts the peritoneal cavity. Head and colleagues studied the complications associated with ablation of lesions adjacent to the diaphragm.211 In 29 patients with

2548

PART III  Image-Guided Procedures

A

3 months

3 months

B

FIG 67-201  Biliary fistula in a 62-year-old man with metastatic colon carcinoma. A, CT scan obtained 3 months after ablation shows an ablation defect (arrowheads) in segments V and VI. Dilated bile ducts (arrow) abut the defect at a slightly lower level than in B. B, Endoscopic retrograde cholangiopancreatogram shows opacified portal vein branches (two arrows) at the ablation site and leakage of contrast medium into the ablation defect (single arrow). (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

A

1 month

3 months

B

FIG 67-202  Arteriovenous fistula in a 42-year-old man with metastatic colon carcinoma. A, One month after RFA, arterial-dominant–phase CT scan shows parenchymal enhancement consistent with arteriovenous fistula. Within the ablation defect, early opacification of a portal branch (arrow) is adjacent to a small artery (arrowhead). B, CT scan obtained 3 months after ablation shows that the parenchymal enhancement has not changed, but the sizes of the artery and portal vein have increased with development of a pseudoaneurysm (arrowhead). (From Choi H et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001.)

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

2549

B FIG 67-203  A, CT of multitined RF probe centrally positioned with a tumor. B, After application of heat, small gas bubbles (arrow) are noted in the mass.

lesions contiguous to the liver, 17% of patients developed diaphragmatic thickening thought to be indicative of injury. All these patients had referred shoulder pain immediately after the procedure. Hinshaw and coworkers reported that the preprocedure intraperitoneal infusion of 5% dextrose in water reduced the postprocedure pain that occurred with treated surface lesions, lessening the need for narcotic administration.220

Kidney Ablations Patient selection criteria vary among authors, but most select patients because of comorbidities and size of the tumor. The success and low complication rate of renal tumor ablation have been so favorable that there is a strong tendency to approach most renal masses measuring 3 cm or less with ablation. With the current structure of health care, the radiologist cannot act without interaction with other clinical experts. Each individual case should be considered by the attending clinician, urologist, and interventional radiologist to decide the most appropriate treatment method.

Indications.  The indications for renal ablation are straightforward; the only controversy is the specific techniques or the need for preablation biopsy. Generally, any mass lesion that is believed to be an enhancing tumor mass on imaging should be considered a cancer and treated by surgical or ablative techniques. The standard of care in urology is to treat all enhancing lesions as tumors, either surgically or percutaneously without tissue confirmation by biopsy. Selection criteria for ablation include the size, location, and accessibility of the lesion and the medical considerations of the patient. Tumors measuring less than 3 to 4 cm are excellent candidates for ablation, unless the mass is inaccessible or too close to adjacent structures. Experience has shown that small lesions (≤3 cm) are likely to be cured by ablation, whereas larger ones are likely to recur (Figs. 67-203 and 67-204) (see “Outcomes and Factors Affecting Success,” later). Lesions located centrally within the kidney adjacent to vessels are more difficult to approach because of the possibility of damaging the renal vessels or collecting system. Lesions that are small and exophytic may

be very close to adjacent structures, such as the spleen, pancreas, or bowel, and may not be approachable unless the radiologist is comfortable with complicated instrument placement and the use of techniques to lessen potential damage to these structures, such as the injection of conductive fluids or air (Fig. 67-205). Tumors larger than 4 cm are not favorable candidates because of the recurrence rate, but even such lesions can be considered if the comorbidities are severe, such as morbid obesity. There is nationwide controversy about the need for preablation or presurgical biopsy, with the standard of urologic care being that any enhancing lesion should be treated as if it were cancer. We believe biopsies of renal masses should be performed to prove the presence of cancer before ablation, but even in our institution, this is controversial. The controversy is based on the success rate of such procedures and the possibility of bleeding and tumor seeding of the needle track. At our institution, the outcome is not yet settled, but studies are being conducted to set better guidelines. Anecdotally, we argue that it is irrational to treat masses if they are proven to be inflammatory or benign lesions, unless there are extenuating circumstances. The controversy about the accuracy of needle biopsy results probably occurred because of the paucity of available data on image-guided core needle biopsy until recently. Early reports about FNA results were inconsistent, but reports are now emerging on core biopsy documenting the merits of core needle biopsies. Maturen reported a favorable success rate in a series of 152 renal mass biopsies performed using an 18-gauge core coaxial system.332 Malignant tumors were found in 56%, and benign findings occurred in 40%. There was no tumor seeding. The sensitivity was 97%, and the specificity was 100%. In 62%, there was a great impact on patient management. Complications included two postprocedure hematomas, one requiring blood transfusion. One case of delayed pseudoaneurysm was noted. Other likely contributing factors are the high incidence of postbiopsy bleeding and the occasional occurrence of tumor seeding in the needle track. There is little doubt that hypervascular tumors are prone to serious bleeding, but modern techniques to provide preemptive hemostasis are well developed and easily executed by experienced

2550

PART III  Image-Guided Procedures

E A

F B

G

C

16.88 mm

H D FIG 67-204  A, RFA was performed under MRI guidance with straight Radionics probe. Note the probe in the mass (lower arrow). The mass shows decreasing signal with necrosis of tissue (upper arrow). B, Follow-up scan at 4 months shows low signal intensity on T1, with some reaction in fat around tumor (arrow) creating a halo. C, T2 scan shows decreased signal intensity in the mass (arrow), with no evidence of recurrence (see Fig. 67-210). D, Four-month follow-up T1 image after gadolinium injection shows no enhancement of the mass (arrow). E, Four-month follow-up CT scan shows a small residual mass (arrow) with an inflammatory ring. F, Twelve-month follow-up image with gadolinium enhancement shows no signal in the mass (arrow), indicating no residual tumor. Note that the mass is reducing in size. G, Twelve-month follow-up CT scan shows similar findings with no enhancement and reduction in the size of the mass (arrow). H, Two-year follow-up CT scan shows that the mass (arrow) has almost totally resolved.

CHAPTER 67  Image-Guided Aspirations and Biopsies

2551

A C

B

D FIG 67-205  A, Cryoablation of a small renal mass located centrally. Note the low-density iceball. B, Axial MRI with gadolinium shows no enhancement within the mass (arrow). C, Coronal T2 MRI shows increased signal centrally, suggestive of hemorrhage (arrow). D, Follow-up T2 MRI several months later shows resolution of hemorrhage and lack of signal in this completely treated mass.

radiologists. The potential bleeding risk of an avascular lesion can be anticipated by studying the vascular characteristics of a tumor with a dynamic bolus scan by CT, microbubble flow by real-time US, or gadolinium injection with MRI. If a lesion is vascular, three options exist: (1) choose a site in the lesion that is less vascular, (2) use a smaller cutting needle (e.g., 20 gauge), or (3) employ a coaxial cannula system, which permits insertion of a hemostatic occlusive device if needed (see “Coagulopathy Treatment”). Finally, tumor seeding in the needle track can be prevented by use of a coaxial guidance system, which prevents deposition of tumor cells. With no cells being left, tumor deposits cannot grow.

The importance of a preprocedure biopsy can be appreciated by noting the two patients in Figures 67-109 and 67-110. Both patients had benign renal masses that would have been inappropriately treated if biopsies had not been performed. One had an aspergilloma, and the second had “granulomatous inflammation.”

Contraindications.  Contraindications include the presence of severe coagulopathy or anatomic issues related to proximity of adjacent structures. Lesions that are too close to the bowel or spleen may not be candidates (Fig. 67-206). A distance of 1 cm from these structures is considered adequate to prevent severe damage. Occasionally injury to

2552

PART III  Image-Guided Procedures

FIG 67-206  Small mass on the lateral side of the left kidney adjacent to the spleen was judged to be too close to the spleen for ablation.

adjacent structures can be prevented by injection of fluid, which conducts heat (liquid) or insulates the structure (air) (Fig. 67-207).

General Technique for Kidney Ablation Preprocedure history, review of images, planning, and informed consent.  Preprocedure steps are taken as noted earlier, including physical examination, clinical and medication history, and review of all imaging and clinical data. Medication history is especially important to determine if any anticoagulative drugs are taken (either prescribed or OTC). Laboratory studies including CBC and coagulation are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained. Imaging assessment of lesions before the procedure is important for selection of the probe, appreciation of vascular structures of anatomy, and assessment of anatomy. Ablations are technically challenging and difficult if not planned correctly, so it is unwise to defer review of data until the patient is on the table. A preliminary plan, including probe choice and number, anatomic approach, trajectory, probes to be used, and technique, is helpful. This plan enables the technical staff to ensure that the proper type and number of probes are available. Probes are too expensive to have large numbers on the shelf waiting for use and becoming outdated. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam, fentanyl, or both as needed. Local anesthesia is administered as described earlier; either deep sedation or general anesthesia is used. Use of sedation with general anesthesia may lessen the depth of general anesthesia required, which is beneficial to the patient. During the procedure, vital signs and pulse oximetry are monitored continuously with automated devices. Immediately before the procedure, the patient should receive broad-spectrum antibiotics, such as ceftriaxone, in appropriate doses. The specific antibiotic and dose should be chosen based on the patient history (e.g., allergy, renal failure). Although infections are unlikely, administering antibiotics is a small step to minimize the occurrence of an infection caused by a random hematologic contamination. Procedure steps.  If RF is being used, the grounding pads must be appropriately attached to the skin as recommended by the manufac-

turer to enable good conduction of the current from the electrode to ground. The skin should be shaved to permit good contact. The pads should be attached with the long side of the pads perpendicular to the long axis of the body. Lack of good grounding can result in local burns. One plans appropriately depending on the ablation type and type of probe used (see Figs. 67-203 to 67-205). Most RF devices have a computer-controlled application of the heat relative to wattage, temperature, and duration of the treatment. One can try to consult the most recent literature, but it is most efficient to communicate with the vendor representatives, who can provide the best source of recent literature experience. The recommendations of the manufacturer should be consulted if questions of most appropriate techniques exist. Because these vendors are so competitive and responsible to the FDA, they are knowledgeable about the current recommendations by experienced clinical investigators and the FDA. These devices modulate the heat application based on impedance feedback information to prevent “charring,” which might compromise the treatment. Information about cryoablation techniques relative to iceball issues and treatment times likewise are available from vendors. One of the most important factors for effective treatment is providing effective coverage of the target site with either single or multiple probes as selected. When a probe is placed correctly, the treatment is carried out according to the above-recommended techniques. Multiple insertions require repetitive heating periods for each geographic site. The size and degree of the thermal lesion depend on the temperature achieved, the duration of the heat application, and the adjacent anatomic structures such as vessels. In our opinion, it is important to take vessel location into consideration and to place the treatment probe in a slightly asymmetric location to enhance thermal injury to the area adjacent to the vessels. This means placing the probe slightly closer to the vessels side. We have found that this helps to better treat the site close to the vascular heat sink. The therapeutic response can be assessed in two ways: tissue impedance measurements and direct observation of tissue changes with imaging. Following impedance changes can be done with older RF devices, but this is no longer needed now that RF devices have computerized feedback. Observation of treatment effects to modify probe placement or treatment parameters is important. As noted earlier, cryoablation is superior to RFA because it permits clear visualization of the treatment area by displaying the iceball and a low-density site (Fig. 67-208; see Fig. 67-207). When multiple probes are placed and the freezing cycle is performed, the position of the probes relative to the entire iceball in the treatment area can be evaluated (see Fig. 67-208B and C). With MDCT, a variety of planes can be viewed to assess the treatment and any involvement of adjacent structures (see Fig. 67-208E). Typically two freeze cycles are performed for each treatment site. The second freeze cycle creates a larger iceball than the first one (see Fig. 67-208D). Real-time monitoring of the probe and the ablation effects is critical for a successful outcome and treatment. The location of the treatment area can be assessed, and the probe position can be changed. Real or potential damage to adjacent normal structures can be assessed, and modifications can be made as necessary. If the probe needs to be inserted closer to a site or farther away from a site, it can be executed. If the treatment seems to be putting a normal structure in jeopardy, fluid or air can be injected to insulate the tissues (see Fig. 67-207). The intent of monitoring the lesions under imaging is to produce necrosis of a cuff of tumor-free area by necrosing 10 mm of adjacent normal liver, which has been shown to be a major determinant of success. Because the kidney is smaller than the liver, the organ can move if multiple probes are placed, which requires careful reassessment of

SB

A

B

D

C

E

F

FIG 67-207  A, Cryoablation was performed on the small tumor noted anteriorly on the lower pole of the left kidney (straight arrow). Note the close proximity of the small bowel loop (SB) and the inferior mesenteric vein adjacent to the lesion (curved arrow). B, CT scan taken during cryoablation shows the large cryoprobe in the tumor mass (curved arrow) and a small needle (straight arrow) inserted between the mass and the structures noted above. A small amount of air has been injected, which outlines the gonadal and inferior mesenteric veins (arrowheads), insulating them from freezing. C, The real-time monitoring of the freeze process was used to limit the size of the iceball (arrow) so that it did not extend too far. Because the size of the iceball needed to be limited, four cycles of freezing for 5 minutes were used. D, CT Scan taken slightly below the ablation site shows air surrounding the bowel loop and the gonadal and inferior mesenteric veins (arrowheads). E, Cryoablation site filled with air (arrow) after probe removal shows air in the retroperitoneum and in the pathway of the probe. Air (curved arrow) is between iceball and small bowel loop. F, Follow-up scan at 1 month shows a necrotic site in the lower pole (arrow), which provided an adequate margin for treatment. Continued

2554

PART III  Image-Guided Procedures dure complications such as hematoma or inadvertent damage to other tissues. Patients may experience some general malaise and fever for days, but postablation symptoms are not as frequent as described by Dodd and associates103 for liver lesions.

Posttreatment Appearances.  Postprocedure imaging follow-up

G FIG 67-207, cont’d  G, Follow-up scan 6 months later shows a decreased size and partial resolution of the treatment site, without evidence of recurrence.

anatomy during the procedure. If multiple cryoablation probes are being used, 3D reconstructions can be of great benefit in assessing the completeness of the treatment (see Fig. 67-208). This is critical with cryoablation because the probes are larger than the RF probes, and the weight and configuration of the probes are such that the kidney rotates and moves away from the probes. As a result, the planned approach may need to be modified to accomplish successful treatment. Any of three imaging devices can be used to visually monitor the change in tissue appearance that indirectly reflects tissue coagulation, but in our experience CT seems to be superior for the procedure. US shows a change of increased echogenicity when tissue coagulates, but the associated shadowing that is produced restricts evaluation of the posterior margins of the tumor. For this reason, we use US only to provide the initial placement of the probe. Although MRI can also be used and shows change in tissue signal with certain pulse sequences during coagulation of protein, changing the T2 appearance with IV contrast material (see Fig. 67-204), as noted by Aschoff and colleagues,18 Merkle and colleagues,346 and others, is needed to confirm the loss of perfusion and treatment. CT shows imaging changes from ablation with injection of contrast material. Cha and coworkers noted in an animal model that CT is superior to US for monitoring of an ablation site.63 From clinical experience, CT has the broad imaging traits necessary for accurate monitoring, being able to image instruments, gas, edema, and perfusion changes with IV contrast material.

Postprocedure Care.  Short-term postprocedure care is standard, with bed rest for 4 to 6 hours, monitoring vital signs, and laboratory tests, including hematocrit, liver enzymes, BUN, and creatinine. Depending on symptoms and laboratory values, the patient may be discharged within 24 hours after the procedure. Patients are treated supportively with fluids as needed and advanced to oral intake as tolerated. Any deviation from expected improvement over the first 24 hours warrants rescanning to evaluate for postproce-

should be performed 24 hours after the procedure to determine the short-term completeness of the treatment and at 3-month intervals for at least 18 months. After this, yearly scans should be performed until the mass has undergone expected evolution and disappearance. Either CT or MRI can be used, depending on the expertise and the preference of an experienced imager. If CT is used, triple-phase study with nonenhanced baseline, arterial, and capillary phase should be performed so that evaluation during the different enhancement phases can be performed. A completely necrotic tumor shows no enhancement with contrast material and may appear almost fluidlike except for focal areas of hemorrhage. Over time, the lesion shrinks, resulting in a small focal scar (see Figs. 67-204 and 67-205). When MRI is used, T1, T2, and gadolinium scans should be used, permitting assessment of the paramagnetic effects of any hematoma and intensity values reflecting vascularity. Adequately treated tumors appear dark on all pulsing sequences but may show periablational rim enhancement from inflammatory reaction. Any local hemorrhage shows increased T2 from the paramagnetic effects of the hematoma. Adequately treated tumor does not show enhancement during gadolinium injection. With either modality, it is common to see a slight increase in the mass 2 to 4 weeks after the ablation, but subsequently the mass involutes gradually over months.

Tumor Recurrence.  Tumor recurrence in the kidney is detectable based on three major findings: (1) appearance or continued evidence of viability, such as abnormal signal or density, (2) continued enhancement in the site, and (3) failure of the mass to reduce in size over time (Fig. 67-209). Although CT and MRI are typically used, Doppler US, especially with microbubble contrast material, also shows tumor recurrence (Fig. 67-210). When recurrence has been found, retreatment of the area is mandatory, with adjustments in technique as needed. In one of our cases with recurrence, our appreciation of treatment effects from RF was compromised because we were unable to give the patient contrast material owing to renal failure. Our treatment was changed to cryoablation because of the ability to better appreciate the iceball relative to the tumor mass.

Outcomes and Factors Affecting Success.  Many different factors have been evaluated to seek improvement of the outcome, as discussed subsequently. Considering the wide variation among authors and apparent direct contradiction of results, it seems that the three major determinants of outcome are (1) the experience of the operator, (2) the size of the lesion, and (2) meticulous technique used by a talented interventional physician. Renal tumor size is a strong determinant of renal ablation, so it must be considered as a criterion for indication. Concerning renal tumor ablations, numerous reports define the important determinants of success. In an early report by Zagoria and colleagues that was derived from a group of 22 patients (27 treatments), size was the major determinant for achieving tumor eradication with a single session of ablation, with all 11 tumors 3 cm or smaller being completely ablated after one session.529 Mayo-Smith and coworkers also emphasized the significance of size for effective treatment with RF.336 In their small series, the larger masses required retreatment compared with the smaller tumors (3.5 cm compared with 2.4 cm). In a series of 100 renal tumors by

B

A

D

F

C

E

F

FIG 67-208  A, Cryoablation of a large left renal carcinoma. Two probes are shown in the tumor mass before freezing. B, MDCT permits reconstruction, which shows the four probes required to produce a 6-cm iceball, including the 4-cm tumor and 10-mm margin. No iceball is seen yet because freezing was not begun until confirmation of probe placement was made. C, After freezing, the iceball in the mass can be well seen (straight arrows), with some extension into the adjacent abdominal muscle (curved arrows). D, A repeat freezing cycle typically produces a larger iceball because of the change in the tissue consistency and breakdown of proteins. Note that the iceball now extends farther into the abdominal musculature (arrows), which is in itself not an issue because adequate treatment margins are more important than some localized muscle damage. E, MDCT reconstruction after freezing and removal of the probes shows the extent of treatment. Note also the air left in the partially frozen pathways (arrows). F, Several weeks after treatment, high-density hemorrhage is noted in the mass (arrow). Fluid collection (F) is noted laterally. Continued

2556

PART III  Image-Guided Procedures = .015). We concur with Gervais and coworkers that location and proximity to vessels is important, with more peripheral lesions being easier to heat and more prone to necrosis. For this reason, we prefer to place the treatment probe so that it is closer to the interface with the normal kidney first. We believe that this placement produces some reduction of blood flow through the feeding vessels as a result of ablation or even edema produced at the feeding vessel level. The ablative process is more successful because of the reduction of the heat sink effect. Even though this has not been proven as effective, we believe this is still a worthwhile approach, especially if aggressive treatment at that location is performed. Individual results (i.e., equipment, operators) vary, but this approach at least potentially improves margins. Histology and the presence of coexisting renal disease were reported by Zagoria and colleagues as not correlating with treatment success.529 Anecdotally, we believe that cryoablation may have a slight increased incidence of bleeding compared with RFA. The bleeding occurs from the muscular abdominal wall or the lumbar region. It is unclear if it is related to the freezing effect on tissues or the size of the probes; as more experience is reported, the validity of this belief will be clarified.

G

Side Effects.  Syvanthong and coworkers reported on the effect of RFA on renal function in patients with unilateral kidneys.482 They noted tumor eradication to be effective. There was an associated 16% increase in serum creatinine concentration and a 13% decrease in creatinine clearance. C

H FIG 67-208, cont’d  G, Later scan shows development of air (arrows) in the tumor and the fluid collection, indicating abscess. Because the operator was confident that the bowel was not damaged, a rationale for the infection was sought. H, Review of patient’s previous CT scan shows that patient previously had a case of diverticulitis (see the thick colon [C] and edema [arrow]), which was treated medically. We hypothesize that the bacterial contamination occurred by hematogenous seeding from the colon even though diverticulitis was not present at the time of the cryoablation.

Gervais and coworkers, for a single ablation episode, small size (P < .0001) and noncentral location (P = .0049) were independent predictors of complete necrosis.144 For any number of treatment episodes, location was a significant predictor of complete necrosis (P = .015). Tumor location for kidney masses was studied by Gervais and coworkers143 and Zagoria and colleagues.529 Zagoria and colleagues529 did not find location significant, but Gervais and coworkers143 did. Gervais and coworkers143 found that for any number of treatment episodes, location was a significant predictor of complete necrosis (P

Complications.  Attention to anatomic detail and avoidance of certain structures are important. Avoidance of the central region of the kidney prevents the possibility of doing unnecessary damage to the remaining normal kidney because the probability of damaging major vessels is minimized. Use of various maneuvers to minimize damage to adjacent loops of bowel and vessels lessens the chance of collateral damage. When renal tumors lie anteriorly, they are near the colon, small bowel, and retroperitoneal vessels. Using variable patient positions on the table permits movement of the structures so that the uninvolved tissues are not near the tumor treatment area. Injection of air or fluid can provide a partial barrier to the thermal damage, preventing damage to adjacent structures. Gas and liquid (e.g., saline, dextrose water) dissect or diffuse along the longitudinal plane, lessening the effect over time. Such materials must be replenished periodically to maintain the volume against the tissues to insulate them. Boll and colleagues reported in an animal model that surgical clips do not represent an impediment or hazard for RFA.40 Gervais and associates142 and others also noted relative to the collecting system that contiguity of the tumor or zone of ablation to the system did not increase the complication rate, but that obscuration of the calyces by a central tumor was a predictor of collecting system hemorrhage necessitating treatment. Urine leaks were rare and related only to downstream obstruction. McGahan and coworkers reported on a small series using a transhepatic approach for ablation of a right renal masses.341 Although they encountered no complications, one must assume that extensive use of such an approach would result in increased bleeding from the liver; this approach with adrenal biopsies resulted in a high hemorrhage rate. The method should be considered within the context of this history and used as appropriate with each patient.

Lung Ablations With the acceptance and success of liver and kidney ablative treatments, the technique has been extended to the lung. Intuitively one might have expected that ablations of the lung would be associated

CHAPTER 67  Image-Guided Aspirations and Biopsies

2557

A D

B E

C FIG 67-209  A, Tumor recurrence after RF treatment. A small tumor in the anterior left kidney was treated with RFA. B, Follow-up scan showed lack of enhancement in the tumor but raised the question of eccentric enhancement at the margin adjacent to the cortex (arrow). C, MRI with gadolinium enhancement in the mass area (arrow) indicated recurrence and the need for retreatment. D, The patient had refused retreatment until this repeat scan showed further growth of the tumor and enhancement (arrow).

with too many complications, but animal studies and preliminary human trials indicate that lung ablations have some role. Currently lung ablations are used as an alternative treatment or palliation in patients with a complicated medical status, but some experience suggests that it may become a primary treatment in the future. Because our experience with lung ablations is limited, we rely heavily on other sources for a more comprehensive discussion.108,387,420,452,475,524 Dupuy and colleagues were the first group to perform percutaneous RFA on small lung lesions, and many other authors followed with supportive data and information.109 Feasibility of lung ablations was proven on animal models by numerous authors. Wacker and associates, using MRI, noted the ability to create ablative lesions in the lung and the detection of an associated pneumothorax.505 Morrison and colleagues studied CT and MRI of ablative lesions in the normal rabbit lung parenchyma.360 On histologic

examination of the lung tissue after ablation, a generally spherical lesion was observed; irregularity of shape was not as consistent as had been noted on lesion description of solid organs. The central zone was characterized by coagulation necrosis, and the outer zone was characterized by hyperemia and edema. On CT, the authors noted a heterogeneous inner zone surrounded by a high-density outer zone. On MRI, the inner zone was typically hyperintense on T1-weighted fast-spin echo imaging, and the outer zone was hyperintense on T2-weighted fast-spin echo imaging. Oshima and coworkers reported in swine models that occlusion of the bronchus resulted in greater tissue temperature around the probe and a large ablation diameter (P = .03).391

Indications.  The defining indications for lung ablation are more difficult than for other organs. With tumors of the liver and kidneys, it had been proven that local surgical treatment provided excellent results

2558

PART III  Image-Guided Procedures

C

A

DG:32/ PRF:2.2k

B FIG 67-210  A, Case showing the advantage of cryoablation over RF relative to visualization of ablative effects. This patient had initial treatment with RF and recurrent tumor treated with cryoablation. CT scan shows a RF probe in a tumor mass with gas forming (arrow), indicating tumor necrosis. Because the patient had renal failure, contrast material could not be used, and it was difficult to assess the ablative effects accurately. B, Follow-up Doppler US showed increased blood flow in the mass indicating the presence of viable tumor. C, Repeat ablation was performed with cryoablation, which permitted better assessment of ablation effects because of clear visualization of the iceball (arrows).

and improvement in patient outcomes, setting the background for formulation and refinement of percutaneous methods. With lung cancer, it was shown in the surgical literature that local treatment did not afford results as favorable as lobectomy and lymph node dissection because of the propensity for regional lymph node involvement. As a result, lung ablation must be considered a palliative procedure for prolongation of life and relief of symptoms rather than potentially curable. Diagnostic biopsy of the lesion should be performed before the decision to perform ablation to help decide the role of other adjunctive treatments.

In light of these differences of lung cancer compared with other organ cancers, proper indications include (1) prolongation of survival, (2) symptomatic relief, and (3) potential cure.420,452 Palliative treatments are reasonable if large tumors in the lung are slow growing and producing symptoms related to size, such as dyspnea, hemoptysis, cough, or localized pain. With these criteria for palliation, even tumors with extrathoracic spread can be candidates. Reviewing the literature (see “Assessment of Success by Patient Survivorship,” later), cure is possible only if lesions are smaller than 3 cm, there is no remote or local regional lymph node spread, optimal treatment with proper

CHAPTER 67  Image-Guided Aspirations and Biopsies margins can be achieved, and comorbidities are not so significant that serious complications would result. Ambrogi and colleagues summarized the role succinctly by concluding15: Efficacy of RF ablation in the mid-term seems to settle at a promising level, with better results for metastatic lesions and, above all, for lesions smaller than 3 cm. Notwithstanding these encouraging results, RF ablation remains an alternative local therapy only when surgery cannot be performed, especially in NSCLC [non– small cell lung carcinoma]. Although Ambrogi and colleagues spoke of thermal ablation, cryoablation would be similar.

Contraindications.  Because of the basic nature of the lung and the cardiopulmonary compromises that are typically associated with lung cancer patients, numerous contraindications exist. Most patients with lung cancer have underlying pathology of the lung reducing its compliance, making it more susceptible to pneumothoraces and bleeding. The large size of the instruments used makes mechanical injury to the lung more likely. Patients with bleeding diathesis, anticoagulation, severe cardiopulmonary compromise, lung infection, and the use of bronchodilators should be excluded. Patients with cardiac pacemakers can be considered on an individual basis. RF can interfere with pacemaker function, although the theoretical heat coagulation of the myocardium has not proven to be valid. Authors have postulated that the lack of tissue burning results from the massive heat sink effect of the blood flow of the cardiac output. Pacemaker patients can be treated if patients can tolerate temporary cessation of pacing, with backup provided by external or other pacing should cardiac rhythm anomalies develop.66,490 Extracardiac RFA interferes with pacemaker function but does not damage the pacemaker.490

General Technique for Lung Ablation Preprocedure history, review of images, planning, and informed consent.  Preprocedure steps are taken as noted earlier, including physical examination, clinical and medication history, and review of all imaging and clinical data. Although one might assume that clinical physicians are aware of the complete medical history, communication deficiencies are common, especially during these times of healthcare cost reductions, so a careful clinical history is important. Clinical history should be detailed to assess for any recent unexpected changes in cardiopulmonary function or significant comorbidities. Patients with coexisting lung compromise must be evaluated for pulmonary function and suitability for general anesthesia. Medication history is important to determine if any anticoagulative drugs, bronchodilators, oxygen therapy, cardiac drugs, or antibiotics are being taken that would be indicative of coexisting problems that might exclude patients. Patients with pulmonary infections should be excluded to prevent further spread of the infection. Patients with pacemakers require special consideration, as noted previously. Laboratory studies including CBC and coagulation are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained. Consultation for and arrangements must be made for general anesthesia. In contrast to other organs for which general anesthesia is an option, it is required for lung ablations. Intubation and general anesthesia are mandatory because of the need for immobilization, coexisting lung disease, discomfort to the patient, oxygenation requirement,

2559

and the possibility of required assisted ventilation. Finally, because such patients are prone to severe pneumothorax or endobronchial bleeding, acute complications in such patients are best managed with intubation and general anesthesia. The only guidance modality used has been CT scan because of the advantages of rapid image acquisition and clear delineation of the lung anatomy, instruments, and pneumothorax. The patient is placed in a position that is optimal for access to the lesion to minimize the pathway length and to avoid structures not intended to be punctured. From review of prior imaging studies, a tentative target, entrance site, and trajectory pathway are chosen. Choice of the thermal probe by reporting authors varies. All of the various available probes, including the straight cool tip, cluster probes, the multitine RF probes, and cryoablation, have been reported (see Fig. 67-190). Direct comparisons of effectiveness and complications of the different probes have not been well documented, but from the numerous reports, there seem to be few differences among the various methods or devices when meticulous planning and technique are used. The RFA parameters, including probe type, number of probes, and number and duration of treatments, are preplanned. Target and trajectory planning should include a volumetric assessment of the mass with evaluation of adjacent structures, such as blebs or vessels, which should be avoided. Entrance site and trajectory planning can lessen the chance of pneumothorax or bleeding by using selected pathways. Paths traversing areas abutting pleura or regions with numerous fibrous strands are less likely to lead to bleeding, pneumothorax, or lung collapse. Procedure steps.  During the procedure the patient is monitored by the anesthesia team for vital signs and adequate ventilation. The lung is more critical than other ablation sites because numerous patients have compromised cardiopulmonary status owing to lung impairment. Before the procedure a dose of broad-spectrum antibiotics should be given to cover most common pathogens. Although different operators disagree on this point, if one considers the potential benefit of this simple step and the issues related to an albeit rare but potentially serious infection from hematogenous seeding, antibiotics should be given. We are more conservative in this regard because we have had an infectious complication caused by hematogenous seeding, but we believe this is the appropriate approach. Grounding pads are positioned on the opposite chest wall or thighs to enhance proper ground function. Multiple pads are used; the device manufacturer should be consulted to ensure that the recommended method of grounding for thoracic lesions is followed. Simon and colleagues452 applied pads to the opposite chest wall, whereas Belfiore and colleagues31 and others applied pads to the thighs in the traditional manner. Patients with cardiac pacemakers should be managed similarly, but consultations with cardiology and anesthesia colleagues are appropriate to assess the need for cessation of pacing or planning remedial action should arrhythmias occur during the procedure. The entrance site is selected using the plan devised during the preprocedure phase. The specific site is examined and palpated to ensure that there are no intervening local collateral vessels and no intervening obstacles at the surface, such as ribs or scapula. Depending on the probe used, proper placement to optimize the probability of successful complete ablation is important. With the straight probe, it is best to position the device as close to the center as possible (Figs. 67-211 and 67-212). The cluster probe provides more extensive treatment of large lesions (Fig. 67-213). With the multitined probe, positioning of the tines around the lesion is more likely to produce complete ablation (Fig. 67-214; see Fig. 67-212F). With cryoprobes, the exact placement is less critical because of the large

2560

PART III  Image-Guided Procedures 38.0cm

512

R 1 9 0

E A 38.0cm

512

[R]

[L]

R 1 9 0

F B

[R]

[L]

[R]

[L]

C G

[R]

[L]

D FIG 67-211  A, Cluster probe ablation of non–small cell lung cancer. A posterior approach was chosen to minimize the length of the pathway through the lung and to avoid crossing the major fissure. B, Effective treatment of the mass with 10-mm margin is documented by the creation of the hazy ground-glass appearance in parenchyma around the lesion (arrow). C, One-month follow-up CT scan shows a large necrotic area with necrosis centrally and a thick rim of reactive tissue (arrows). D, Coronal view of the treated mass gives a different perspective of the treated site. A large amount of central necrosis with air is seen. E, PET with fusion on axial CT at the level of the cavity shows minimal uptake in the margin of the treated area, which is equivalent to the mediastinal signal. This indicates that the central masslike area is not a viable tumor. F, CT scan taken at 2 months after treatment. Necrosis of site has occurred, with cavitation (arrow) visible in the mass. The overall size of mass is reducing, indicating further resolution. G, Coronal view providing different perspective, with reduction in size of mass. (A and G, Courtesy Damian Dupuy, Brown University, Providence, RI.)

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

C

D FIG 67-212  A, RFA in the lung and value of multiplanar reformatted image display in an 80-year-old woman with a history of smoking and earlier right upper lobe non–small cell lung carcinoma (NSCLC) retreated with RFA 2 years earlier, now with a 1.7-cm speculated mass in the left lower lobe. Initial CT section of lung windows with the patient prone. Arrow denotes the targeted left lung tumor; arrowheads indicate the metallic wires contained on the adhesive grid applied to the skin over the anticipated skin entry site. Incidentally noted is a residual scar from the previously treated NSCLC in the contralateral lung. B, CT section of lung windows after advancement of a 22-gauge, 15-cm-long “finder” needle (arrow) into the tumor. C, CT section of bone windows after advancement of a 14-gauge cannula (arrow) into the tumor with the use of tandem trocar technique with the finder needle acting as a guidance landmark. Aspiration biopsy with 22-gauge needles deployed coaxially through the 14-gauge cannula showed malignant cells consistent with NSCLC. D, CT section with bone windows after deployment of the 3.5-cm-diameter Le Veen multiple-tine RF electrode (arrow) coaxially through the 14-gauge cannula. Before RFA the finder needle was removed. Continued

2561

2562

PART III  Image-Guided Procedures

R A

L P

E

F FIG 67-212, cont’d  E, Multiplanar reformatted image display of the lung window perpendicular to the axis of the RF electrode shaft. Although the electrode shaft has penetrated the tumor eccentrically, the punctate radiodense tines (arrows) are shown to encircle the tumor completely. Complete ablation of the tumor and a rim of at least 1 cm of nontumorous lung would be anticipated. RFA was performed along a cylinder, with the tines being sequentially deployed at the deep aspect, the epicenter, then the superficial aspect of the tumor. F, CT section of lung windows obtained immediately after the RF electrode was removed shows that perilesional opacities are present (arrow), indicating probable complete tumor ablation. The small pneumothorax (curved arrow) was not treated and remained asymptomatic. (From Rose SC et al: Lung cancer and radiofrequency ablation. J Vasc Interv Radiol 17:927–951, 2006.)

destructive area of the iceball (Fig. 67-215). Generally, because of the insulating effect of the air in the lung parenchyma, the application of thermal energy requires less time than with other organs because the exchange of heat or cold to adjacent tissues is less effective, and more tumor damage is caused with less time interval. The site is treated according to the recommendations of the manufacturer and methods as described in the literature referenced in this chapter or new sources available through Internet search engines. As is true with biopsies, repositioning likely increases the incidence of pneumothorax and complications. Despite the best technique, however, pneumothoraces commonly occur simply because of the usual coexisting lung disease (see Figs. 67-212F and G and 67-214). As with lung biopsies, minimizing the number of insertions and readjustments is important to minimize complications. One general approach with larger lesions is to insert the probes to the distal side of the lesion for the first treatment. A subsequent more superficial treatment can be accomplished by withdrawing the probe, reimaging, and retreating if properly positioned. The specific technique used by several authors is as follows. Simon and colleagues used a cooled RF straight electrode, either a single for small lesions or a cluster for larger lesions.452 Using the impedance feedback mechanism, they applied RF for the time it took the generator to shut off three times in 1 minute in impedance control mode. Jin and associates used the cool tip straight probe.243 Tissue impedance was continuously monitored using the feedback system of the computerized generator. At the end of each treatment, the saline perfusion was stopped and the maximal temperature recorded. If the temperature exceeded 60°C, the electrode was withdrawn slowly in increments of 1 cm up to the length of the active tip. Simultaneously the intratumoral temperature was measured. If after the first treatment, the maximal intratumoral temperature did not exceed 60°C, an additional treat-

ment was performed at the same site. These authors chose to apply RF for 12 minutes during the initial ablation episode and for 6 to 12 minutes during subsequent episodes for a maximum peak current of 1000 to 2000 mA and 80 to 150 W. After the procedure, the electrode was withdrawn without cauterizing the probe tract. Belfiore and colleagues reported a multitine probe technique.31 The probe was used for a single application or multiple applications depending on the size of the tumor. The device was positioned and then energized for 9 to 12 minutes when the target temperature of 90° to 95°C had been reached. Grounding pads were applied to the thighs and not the chest by these authors. Steinke and coworkers preferred the multitine device also because when deployed, the tines prevent dislocation secondary to motion or respiration.475 They used a similar protocol with temperature, automated impedance feedback, and time of 9 of 12 minutes. The treatment temperature and time are modified according to the effects noted during the image monitoring of the ablation. The ability to acquire 3D imaging is helpful when multiple probe placement is required (Fig. 67-216), and MDCT scanners have proven worthwhile (Fig. 67-217A and B; see Fig. 67-216). MDCT scanners provide the ability to see the multiple probes and assess the location relative to tumor and adjust the treatment accordingly. Specific treatment decisions are modified by the physician at the time of the procedure. For small lesions, several authors recommended gradual increase in the power level and temperature owing to the insulating effect of the surrounding lung. They expressed the concern that with rapid heating, charring might occur and impair homogeneous deposition of heat. Virtually all authors agree that achieving a 10-mm margin around the tumor is important. The critical sign used by all is the occurrence of a hazy ground-glass change in the region of 10 mm in the lung Text continued on p. 2567

2563

CHAPTER 67  Image-Guided Aspirations and Biopsies

R

R 2 0 0

A

2 0 0

B [A]

[R]

C

[A]

[L] [R]

[P]

D

[L]

[P]

E FIG 67-213  A, Cluster probe ablation of non–small cell lung carcinoma in the left lower lobe. B, Effective treatment documented by hazy halo (arrow) in the lung around the mass. C, CT follow-up at 2 months shows a soft tissue mass with a pleural base corresponding to the treated area. D, CT at 14-month follow-up shows reduction in the size of the mass. E, PET scan with fusion to CT at 17 months shows almost complete resolution with no increased uptake in the mass compared with other tissues and mediastinum. (Courtesy Damian Dupuy, Brown University, Providence, RI.)

2564

A

PART III  Image-Guided Procedures

C

B

D FIG 67-214  A, Cryoablation of a right upper lobe mass from metastatic breast cancer. Very early iceball formation is noted around the probe. B, Complete formed iceball (arrow) after application of the second freezing cycle. C, One-month follow-up scan shows complete necrosis of the treatment site with air bubbles noted in the mass (arrow). D, Coronal reconstruction of a follow-up CT scan shows the relationship of the treated mass (arrow) containing gas bubbles to the bronchi and mediastinum. (Courtesy Damian Dupuy, Brown University, Providence, RI.)

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

C

B FIG 67-215  A, Placement of an RF probe in non–small cell lung carcinoma in the right upper lobe. CT scan during the time of treatment shows air in the mass (arrow) and halo edema indicating effective treatment. B, Chest radiograph shows that pneumothorax with air in pigtail chest tube (arrow) is in place in the chest wall, which may occur with larger probes and chronic pulmonary disease. C, Follow-up chest radiograph shows necrosis of the site with thin-wall residual cavity and necrotic material in the base of the cavity (arrow). (Courtesy Damian Dupuy, Brown University, Providence, RI.)

A

B FIG 67-216  A, Multiplanar reformatted (MPR) image of a CT scan oriented perpendicular to the shaft of the electrodes. Two electrodes were selected to treat this oblong tumor and achieve adequate coverage of the larger inferior tumor in one power application. B, MPR image of a CT scan with the same orientation but through a smaller superior tumor. (From Rose SC et al: Lung cancer and radiofrequency ablation. J Vasc Interv Radiol 17:927–951, 2006.)

2565

2566

PART III  Image-Guided Procedures

[R]

[L]

A

R 2 0 0

B

C SL: 35.9mm

[R]

D

Study time:11:35:36 AM MRN: DOB: Sex:

[L]

E

FIG 67-217  A, Pretreatment contrast-enhanced CT scan shows a metastatic destructive lesion in the posterior left rib and lung field (arrow). B, After radiation treatment, PET scan with overlay on reconstructed coronal CT was performed. There is increased uptake in lower portion of mass (arrow), indicating residual viable tumor after therapy. C, With the patient lying prone, a posterior approach for RFA was used. CT scan shows localization of multiple RF probes into lung and rib portions of the tumor. D, After RF treatment, CT scan with coronal reconstruction was performed. A necrotic tumor site containing gas (arrow) is noted. E, Repeat PET scan overlaid on a coronal CT scan shows absence of uptake at the site of the tumor mass. Some posttreatment inflammatory changes in the rib are noted. Compare with B. (Courtesy Damian Dupuy, Brown University, Providence, RI.)

CHAPTER 67  Image-Guided Aspirations and Biopsies

A

B

2567

C

FIG 67-218  A, Ablation of testicular metastasis led to empyema. B, Follow-up CT scan shows a multiloculated cavity in the left pleural space (arrow). C, Complete resolution of empyema occurred at 1 month. (Courtesy Damian Dupuy, Brown University, Providence, RI.)

parenchyma around the tumor mass (see Figs. 67-211, 67-212, and 67-213). The other signs used to assess complete treatment are the achievement of the correct temperature levels and impedance changes at the treatment site. Belfiori and colleagues reported that in most cases, contrastenhanced CT scans obtained after RFA showed a nonenhancing central area with decreased density (ablation zone), with intralesional bubbles and the ground-glass opacity surrounding the tumor (see Fig. 67-214).31 This observation has not been made by other authors.

Postprocedure Care.  After the procedure, a delayed CT scan is taken to check for pneumothorax. If one occurs, a chest tube is inserted by the radiologist on site as needed. Patients with chronic obstructive pulmonary disease or other lung problems have poor lung compliance, and their lungs may be difficult to reexpand. Small catheters and Heimlich valves do not always expand such stiff lungs. In such cases, a thoracic surgeon may need to be consulted for placement of a large chest tube, water seal, and active suction (see Fig. 67-214C). Patients are observed in the radiology area for at least 1 hour and then may be moved to a recovery area in the hospital. Patients should be kept overnight to provide adequate observation and care should a later pneumothorax, hemothorax, infection, or other problems develop. For the foreseeable future, lung ablations are unlikely to be performed as outpatient procedures because of the many factors affecting the procedure and the increased risk of complications compared with other ablation sites. Patients may have moderate postprocedure pain, which should be treated with appropriate narcotic or nonnarcotic analgesics as needed. If the patient experiences dyspnea, administration of oxygen as needed is appropriate. Any sudden change in symptoms or discomfort warrants evaluation with imaging to determine if a pneumothorax or hemothorax may have developed. Low-grade fever also may persist for several days.420 If a significant complication arises, hospitalization is appropriate. Rose and associates in a review article noted that postprocedure hospitalization ranged from 1 to 35 days (mean, 1.8-6 days).420

Side Effects.  Postprocedure side effects have been noted similar to general symptoms noted for other organs, but specific to the lung. Gadaleta and colleagues reported transient fever, cough and hemoptysis, pain, and dyspnea in 46%, 29%, 38%, and 13% of patients, respectively.136

Postprocedure Findings.  Immediate technical success has been defined by virtually all authors as the presence of a ground-glass

appearance or soft tissue density in the lung around the pulmonary nodule. Suh and colleagues described the enhancement of treated nodules as decreasing significantly from 46 to 10 Hounsfield units (HU) when measured at 1 to 2 months.480 Bojarski and colleagues described postprocedure findings of lung ablations in a series of 26 patients with 32 neoplasms (14 primary and 18 metastatic) with a mean size of 3.1 cm (range, 1-7.1 cm).39 The most common postprocedure finding was a peripheral ground-glass opacity, seen in 84% of patients, which rapidly resolved in all but one patient in 1 month. Cavitation occurred in 31% and was most frequent in the inner two thirds of the lung (Fig. 67-218; see Fig. 67-214). Patients not developing cavitation developed “bubble” lucencies at the site.

Long-Term Follow-up.  Chronic imaging changes have been noted by most authors. Bojarski and colleagues noted imaging findings as follows.39 Scarring appeared as pleural thickening in 55% of parenchymal neoplasms and as linear densities between the lesion and pleura in 64%. Pleural effusions were seen in 15%. Assessment of response and detection of tumor recurrence.  CT and PET have been used to diagnose the recurrence of tumor after therapy by other methods and postablation (see Figs. 67-211, 67-213, and 67-217). PET has the greatest apparent success, although the reported series have been too small to prove statistically. Generally the success of ablative therapy is difficult to judge by CT criteria because of the different methods to judge adequacy of ablation. Authors have used the size of lesions and contrast enhancement by CT for other organs, and these do not seem to be consistent observations. Using the size change of treated lesions has not been as consistent as for other organs. Bojarski and colleagues noted that the size of the mass enlarged from pretreatment scans at 1 month but remained stable at 3 to 6 months.39 Jin and associates noted a more consistent decrease in the size of the lesions.243 The mean percentage of completely ablated lesions shrank (5.7% at 6 months, 9% at 9 months, 14.3% at 12 months, and 40% at 15 months) compared with the lesion size on the immediate postprocedure CT scan. Jin and associates also noted in the partially ablated group (n = 12 patients) that the ablated lesions gradually increased after the 6-month follow-up CT examination.243 Steinke,475 Akeboshi,9 Gadaleta,135,136 and Simon452 and their colleagues, and others have shown little correlation with response and size of the lesions after ablation. Lee and coworkers used contrast-enhanced CT to diagnose tumor recurrence.290 They reported that an increase of at least 10 HU within a region of tumor between the initial noncontrast CT image and the subsequent contrast-enhanced CT image was held to represent viable

2568

PART III  Image-Guided Procedures

tumor. Belfiore and colleagues reported using contrast-enhanced CT to assess the necrotic site in follow-up, but they made no specific reference to the degree or extent of enhancement or to direct correlation with viability of tumor.31 Gadaleta and associates noted using contrast enhancement for evaluation of viable tumor, but they did not report any observations about its utility for assessing viable tumor; lack of comment implies that the method was not worthwhile.136 PET scanning has been considered to be worthwhile for assessing successful tumor ablation, especially better than CT (see Figs. 67-211, 67-213, and 67-217). Suzawa and coworkers studied the efficacy of using FDG PET to study the outcome of RF-ablated pleural lung lesions in 24 lung tumors.481 These tumors were PET positive before ablation and had a mean size of 1.9 cm. This prospective study was undertaken to determine the clinical utility of CT FDG PET in evaluating early therapeutic response to lung RFA used for the treatment of lung tumors. Defining involution of the ablative zone by 30% on CT as anatomic tumor response and reduction of the SUV by 30% on PET, the anatomic change occurred in 42% as studied by CT and 96% as studied by PET. As one might anticipate, the authors concluded that the metabolic change occurred before the anatomic changes. Kang and associates evaluated the conditions of patients 1 week after RFA with the use of PET imaging.248 They reported that all tumors smaller than 3.5 cm in diameter were negative for residual tumor, whereas all tumors larger than 3.5 cm in diameter had increased activity that suggested viable residual tumor in the outer rim of the tumor. Akeboshi and colleagues studied 31 patients with lung ablations by enhanced CT and PET scanning.9 For a standard of comparison, the authors performed biopsies in many cases and showed that PET was more sensitive and accurate than contrast-enhanced CT for detection of residual tumor. Assessment of success by patient survivorship.  Because of the remarkable variability among the methods used to assess tumors and the variable observations, the best substantiation of lung ablation effectiveness is patient survivorship. Numerous reports have documented this. Simon and colleagues reported 153 patients with inoperable lung cancers under CT guidance.452 The overall 1-, 2-, 3-, 4-, and 5-year survival rates for stage I non–small cell lung cancer were 78%, 57%, 36%, 27%, and 27%, respectively; rates for colorectal pulmonary metastasis were 87%, 78%, 57%, 57%, and 57%, respectively. The 1-, 2-, 3-, 4-, and 5-year local tumor progression-free rates were 83%, 64%, 57%, 47%, and 47%, respectively, for tumors 3 cm or smaller and 45%, 25%, 25%, 25%, and 25%, respectively, for tumors larger than 3 cm. The difference between the survival curves associated with large (>3 cm) and small (≤3 cm) tumors was significant (P < .002). Lee and coworkers described RF ablative results in 26 patients with 27 non–small cell lung carcinomas and four patients with five lung metastases, with cooled-tip electrodes under CT guidance.290 Patients were not candidates for surgery because of advanced-stage disease (n = 20), comorbid processes (n = 4), or refusal to undergo surgery (n = 6). The procedure was performed with the intent to cure in 10 (33%) patients with stage I tumors and as palliative therapy in 20 (67%) patients. Using IV contrast enhancement, their results showed that complete necrosis was achieved in 38% of 32 lesions; partial necrosis was achieved in 62% of the remaining 20 lesions. Tumor size was a major determinant of complete necrosis. Complete necrosis resulted in all 6 tumors smaller than 3 cm but in only 6 of 26 larger tumors (P < .05). Mean survival of patients with complete necrosis (19.7 months) was significantly better than that of patients with partial necrosis (8.7 months, P < .01). In a series of 60 patients, de Baere and colleagues noted several indicators of probable success.94 An ablation area at least four times larger than the initial tumor was predictive of complete ablation

treatment (P = .02). There was a trend toward better efficacy for tumors smaller than 2 cm in diameter (P = .066). Overall survival and lung disease–free survival at 18 months were 71% and 34%, respectively. As one might anticipate, the large size of ablative sites indicated a more aggressive approach, which had an associated higher pneumothorax rate. Significant data also are available from authors related to the success of ablative therapy for metastatic colon lesions. Yamakado,524 Steinke,475 and Simon452 and their colleagues showed a clear improvement of survivorship. The factors affecting morbidity were size of the lesions and extrapulmonary lesions.

Complications.  Lung procedures have significantly more complications than solid organ ablations. Most patients chosen for ablation have compromised lung status. Although a specific study has not examined this association, such factors are recognized as a major factor for lung biopsies and likely translate to lung ablations, which use larger instruments. Simon and colleagues reported complications for a mixed group of patients, including primary and metastatic lesions.452 The overall pneumothorax rate was 28.4% (52 of 183 ablation sessions), with a 9.8% (18 of 183 ablation sessions) chest tube insertion rate. The overall 30-day mortality rate was 3.9% (6 of 153 patients), with a 2.6% (4 of 153 patients) procedure-specific 30-day mortality rate. In the series by Lee and associates, 10% of patients had major complications that included acute respiratory distress syndrome and two pneumothoraces that required thoracostomy.290 De Baere reported that pneumothorax occurred in 54% of procedures, but a chest tube was required in only 9% of the procedures.91 Relative to lung function, de Baere reported no change in pulmonary function as measured by spirometry.91 Yamagami and coworkers reported a group of 129 patients with a pneumothorax rate of 29.5%.522 Fourteen of these cases were treated by manual aspiration, and 24 were simply observed. In five cases (3.9%), chest tube placement was required as therapy for pneumothorax. The risk of pneumothorax was significantly increased in patients with pulmonary emphysema. Rossi and associates reported complication rates similar to those of other authors but also noted the occurrence of pneumonia in 4 of 36 patients, which was successfully treated by antibiotics.425 A survey of seven large centers was reported by Steinke and associates.476 These centers reported 493 percutaneous procedures in lung tumors. Two deaths have been reported. Complications were subdivided into major and minor complications. Pneumothorax occurred in 30% of interventions, with less than 10% requiring intercostal drainage. Pleural effusion requiring aspiration occurred in less than 10% of cases. Anecdotal cases reported include fistula (cavitary cutaneous and bronchopleural fistulas)420 and pseudoaneurysm malformations (Figs. 67-219 and 67-220; see Fig. 67-218).452 Finally, several cases of air embolism to the brain have occurred, so any sign of neurologic change during this procedure warrants immediate evaluation. Air embolism has been associated with high fatality, but several authors have used hyperbaric oxygen successfully to effect recovery. Ahrar and colleagues studied the generation of microbubbles causing such cerebral microemboli in a dog model.7 Microbubbles in the carotid arteries can be detected in humans and animals during such procedures. Although Ahrar and colleagues asserted that these microbubbles are too small to create microinfarctions of the brain, this model may not be valid because such models likely do not have the comorbidities that occur in humans, such as pulmonary and cardiovascular predispositions. Frich and associates showed in an animal model that microbubbles were created during ablation of the liver and could be detected in the

CHAPTER 67  Image-Guided Aspirations and Biopsies

2569

R 2 0 0

C

A

B FIG 67-219  A, CT scan of metastatic testicular carcinoma in posterior left lower lobe during cryoablation with single probe. B, Short-term follow-up shows loculated empyema, which was drained effectively. C, Delayed follow-up shows minimal pleural thickening with focal scarring (arrow).

pulmonary arteries.133 These authors noted that the bubbles were rarely observed when lesions were small, the treatment was brief, and the RF power was low. With the widespread use of liver ablations and complete lack of any cerebral symptoms in the vast reported experience, the presence of comorbidities combined with microbubbles is the likely cause of such cerebral events. Further work is needed to help determine the nature and severity of comorbidities that cause such problems so that these patients can be avoided during the selection process.

Factors for Successful Outcome.  The single consistent factor for successful outcome is the small size of a lesion. Virtually all authors report that lesions less than 3 to 3.5 cm have a high probability of successful outcome (see “Assessment of Response and Detection of Tumor Recurrence”). Likewise, the successful treatment of a 10-mm margin free of tumor is important; data to support this are not yet available. The choice of thermal ablation device has not been considered a factor, but no data have been collected to assess this factor.

Factors Affecting Side Effects and Complications.  The most significant factors contributing to the occurrence of complications are

the pulmonary status of the patient, the anatomic characteristics of the site being treated, the large size of the probes used, and the skill and meticulous technique of the operator. It is likely that even with these factors being known, little can be done to minimize the complications except carefully selecting patients, using meticulous technique for choosing the pathway and executing the procedure, minimizing the number of probe placements, and having a dedicated cadre of interventionalists who become skillful. One specific postprocedure finding, cavitation, has been studied by Okuma and colleagues.387 The frequency of cavitation was significantly higher in patients with lung cancer as the primary lesion (see Figs. 67-215 and 67-220), patients who underwent RFA for lesions located within 1 cm of the chest wall, and patients with pulmonary emphysema (P < .05). Age, sex, maximum power output of the RF generator, total ablation time, minimum impedance, and tumor diameter were not significant contributors to cavity formation. Information about the influence of ablation site on complications is insufficient. Okuma and colleagues avoided centrally located lesions with the belief that they are more likely to cause severe hemorrhage or bronchopulmonary fistula.387,388 The relevance of these factors has not been confirmed by other authors.

2570

PART III  Image-Guided Procedures

B

A

FIG 67-220  Fatal pulmonary abscess of bronchocavitary and cavitary-cutaneous fistulas in a 44-year-old woman with stage IIIB non–small cell lung carcinoma of the right upper lobe previously treated with chemotherapy and external-beam irradiation. A, RFA was performed for relief of chest and shoulder pain and possible prolongation of life by cytoreduction of the tumor. CT scan of soft tissue windows with an RF electrode located in the deep portion of the tumor shows the proximity of the electrode tines to the anterior segment of the right upper lobe bronchus (arrow). B, Three weeks after RFA, the patient experienced progressive dyspnea over several days. CT scan of lung windows shows an air-containing necrotic abscess in the region of RFA with fistulous connection to the anterior segmental bronchus and skin via the previous RF electrode tract (arrows). The patient died 7 weeks after RFA as a result of progressive respiratory distress and air space disease despite aggressive antibiotic therapy, attempts to close the bronchial fistula with an endoscopically placed covered stent, and repeated transcavitary catheter-directed administration of biological adhesive (fibrin glue followed by bovine albumin) and catheter drainage of the abscess cavity. (From Rose SC et al: Lung cancer and radiofrequency ablation. J Vasc Interv Radiol 17:927–951, 2006.)

Musculoskeletal Ablations Musculoskeletal ablations have gained great popularity for benign and malignant lesions and have become the standard of care for many abnormalities. Ablation of osseous lesions has proven effective for pain relief and reduction in size of local tumor growth. For osteoid osteomas and localized painful metastatic lesions, RFA has become the treatment of choice.

Indications.  The two major indications for bone ablations are osteoid osteomas and painful metastatic bone lesions. Ablations for these indications have evolved into effective treatments for severe problems. RFA has revolutionized the treatment of osteoid osteoma, establishing a new standard of care. Because of the success and experience with RFA techniques in many other organs and the failure of other minimally invasive methods for treatment of osteoid osteomas, RFA has become the treatment of choice. Numerous authors have confirmed good results with RFA.23,95,423,515 Painful metastatic bone lesions are problematic for cancer patients because typically these are refractory to standard chemotherapy, radiopharmaceuticals, or hormonal therapy, and analgesics may not provide adequate relief of symptoms. RF treatment is always possible, but within the context of advanced disease or poor functional status,  RF treatment may not be a rational option except if fracture is anticipated.

General Technique for Musculoskeletal Ablation Preprocedure history, review of images, planning, and informed consent.  Preprocedure steps include physical examination, clinical and medication history, and review of all imaging and clinical data. Medication history is especially important to determine if any anticoagulative drugs are taken (either prescribed or OTC). Laboratory

studies, including CBC and coagulation, are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam, fentanyl, or both as needed. Local anesthesia is administered as described earlier; either deep sedation or general anesthesia is used. Use of sedation with general anesthesia may lessen the depth of general anesthesia required, which is beneficial to the patient. Procedure steps.  Patient vital sign and pulse oximeter monitoring with automated devices, and local anesthetic and sedation are the same as used for other sites. Choice of target, pathway, and entrance site.  Choice of target, pathway, and entrance site is similar as for other sites in that uninvolved tissues should be avoided; avoidance of uninvolved nerves, vessels, ligaments, and joint spaces is particularly important. Relative to osteoid osteoma, Rosenthal and colleagues advised that most spinal lesions should not be treated because of proximity to nerves within the spinal canal, the lateral recesses, or the neural foramina.422 Although cortical bone may offer some protection from heat transmission, these authors were not convinced that it is a reliable insulator, and they stated that a probe should be at least 1 cm from major nerves. Similar considerations apply to rare tumors of the hand. Generally the shortest distance through the bone is selected for access. If such an approach is unsafe (because of neurovascular or other anatomic structures) or technically difficult (because of a steeply oblique approach to the bone surface that would risk the needle’s skidding off the cortex), the lesion may be approached by drilling from the opposite side of the bone. For metastatic lesions, the technique is less complicated. Because an osteolytic destructive site is being targeted, one chooses a trajectory

CHAPTER 67  Image-Guided Aspirations and Biopsies that does not involve nerves, joint space or cartilage, weight-bearing sites, uninvolved structures, vessels, or other organs. Imaging preference.  The guidance modality of choice is CT because of its ability to visualize instruments and bone well and accurately. RF probe selection, placement, and treatment.  A biopsy of the lesion may be performed before the ablation. Various instruments can be used based on the operator’s preference. RFA and cryoablation may be performed, but there are two definite advantages for cryoablation. RFA is effective but is associated with increased pain in the treatment and posttreatment periods. Also, the treatment effects of RFA are not well visualized with CT. Cryoablation has advantages and one disadvantage. The treatment is less painful. The iceball produced during the treatment is easily visualized as a low-density area on CT scan and thus can be well monitored during treatment. The disadvantage of cryoablation is the prolonged time required because of the repetitive freeze/thaw cycles required. Typically the protocol is two separate freeze cycles followed by two thaw cycles; each is about 8 to 10 minutes long. A guidance cannula may be used in such instances. Treatment of osteoid osteoma differs from metastatic treatments. RFA is more commonly used for osteoid osteomas, whereas metastatic lesions can be treated by either RFA or cryoablation. Advantages among the various RF probes have not been apparent from the literature. Cantwell and colleagues found no difference between saline cooled straight probes and multitine probes.54 Rosenthal and coworkers described osteoid osteoma treatment as follows.422 A direct approach to the lesion may be used; if drilling is necessary, the lesion may be approached by drilling from the opposite side of the bone. Such an approach also may be used to obviate turning the patient into a prone position, which can be difficult to do and hard to monitor, especially in large individuals. In most cases, a Bonopty drill (Radi Medical Systems, Uppsala, Sweden) was used. The drill is valuable when it is necessary to penetrate dense bone. Alternatively, if a cannula was used to biopsy the site, it may be used for probe insertion. An RF electrode with a 5-mm exposed tip is introduced, and ablation is performed at 90°C for 6 minutes (Fig. 67-221). If complete ablation or improper positioning occurred, the RF episode can be repeated. Treatment of metastatic lesions with either RF or cryoablation is more straightforward than treatment of osteoid osteomas. The treatment itself involves accurate placement of RF or cryoablation probes into the site, with sufficient overlap to treat the entire area. Cryoablation provides better visualization of the treatment site because of the iceball.

Postprocedure Care.  Patients recover in the routine fashion for several hours. Clinical signs of unexpected pain, injury, or bleeding are noted, and any short-term problems are addressed immediately. For osteoid osteoma, Rosenthal and coworkers recommend that when weight-bearing bones are treated, strenuous training activities such as distance running should be avoided.422 For metastatic lesions, the same precaution is important because treated lesions may be of significant size.

Successful Outcomes.  The largest experience to date with osteoid osteomas was reported by Rosenthal and coworkers.422 In their series of 263 patients performed with CT guidance, there was technical success in all cases. Preablation biopsies were successful in 73%. For ablation procedures performed as the initial treatment, the success rate was 91% (107 of 117 procedures). Procedures for recurrent lesions had a significantly lower success rate of 60% (6 of 10 procedures; P < .001).

2571

FIG 67-221  RFA of osteoid osteoma. Scan shows the RF device carefully positioned in the nidus site (arrow). Success rate for ablation of osteoid osteoma is very high (see text). (Courtesy Mark Robbins, MD, University Hospitals of Cleveland, Case Western Reserve University.)

Peyser and colleagues reported a series of 51 patients with good success, which included lesions from the femur, tibia, calcaneus, talus, metatarsus, humerus, scapula, olecranon, patella, sacrum, and vertebral bodies.405 Rimondi and colleagues reported a series of 97 nonspinal osteoid osteomas with most occurring in the metaphysis of the femur; they achieved primary success in 85% of patients.418 Favorable results have been reported by numerous other authors. Clinical success was achieved in 94% of patients (44 of 47).515 Successful treatment resulted in early pain relief and return to normal activity in 30 of 38 patients.75 Metastatic tumor ablations have been favorable for RFA and cryoablation sites. Using RF, Goetz and associates reported that 95% (41 of 43 patients) experienced a decrease in pain that was considered clinically significant.151 Opioid usage significantly decreased at weeks 8 and 12. Cryoablation can result in substantial pain reduction, as shown with a 43% mean reduction in worst pain at 4 weeks, which is considered to be clinically important.52,53 Patients also reported pain relief that ranged from 50% to 100% 4 weeks after cryoablation, which compares favorably with the reported radiation therapy response.

Complications.  For osteoid osteomas, reported complications include local skin burns,75 local infection,405 cellulitis,422 and sympathetic dystrophy.422 Rosenthal and coworkers reported two deaths from aspiration during general anesthesia.422 For metastatic ablations,

2572

PART III  Image-Guided Procedures

reported by Goetz and associates, adverse events were seen in three patients and included (1) a second-degree skin burn at the grounding pad site, (2) transient bowel and bladder incontinence after treatment of a metastasis involving the sacrum, and (3) a fracture of the acetabulum after RFA of an acetabular lesion.151

Side Effects.  An incidental observation made by Rosenthal and associates is important for clinical radiologists to know.423 In a group of 10 patients, the cardiac rate and respiratory rate while under general anesthesia changed dramatically during needle placement into the site. The mean cardiac rate increased an average of 26 beats/min (40%) to 91 beats/min (range, 62-114 beats/min; P < .001), and the mean respiratory rate increased an average of 12 breaths/min (50%) to a mean of 37 breaths/min (range, 25-52 breaths/min; P < .001). These changes occurred within seconds of tumor puncture. In our clinical experience, the change in vital signs is common.

Factors for Success.  Relative to osteoid osteoma treated by thermal ablation, de Berg and associates reported that the outcome was not affected by patient age or sex, or lesion size or location.95 According to Venbrux and coworkers, entering the nidus and adequate ablation time are critical factors.503

Factors Affecting Complications.  Factors affecting complications include selection of patients for proper lesions, avoidance of certain anatomic structures, and meticulous technique execution.

New Methods to Improve Ablation Efficiency Numerous methods have been studied in animal models and some humans to increase ablation efficiency and increase the size of ablation sites. The ability to create larger lesions potentially can facilitate adequate treatment margins and reduce the incidence of local tumor recurrence. Most of these methods have been shown in animal models, but according to our reading of the literature, they have not made an impact on clinical trials. New thermal methods, including laser devices and microwave devices, have been introduced, but they have not become established in the mainstream. RF has continued to dominate the field. Cryoprobes of varying sizes and vendor brands have been introduced and have established a niche in the clinical market and offer great promise for the future. Because of the ability to visualize the ablated areas clearly and to treat larger areas effectively, these may become a major contributor to ablative therapy. Combining ablation with other treatments such as alcohol, embolization, or temporary compression of vessels has produced some benefit. Few of these treatments have entered clinical practice as a practical approach. Shankar and colleagues reported a series of 30 patients, including 15 with combined alcohol/RFA and 15 with RFA only.443 They reported significantly larger ablated sites with the combined therapy. The measured impedance in the lesion was greater for the combined treatment group. Yamagami and colleagues reported favorable results with metastatic disease using RF after using intraarterial chemotherapy.523 Local injection of saline, which increases heat production by the ion effect, has been reported by several authors.38,310 These authors reported injecting saline in different volumes and concentration and found that the resulting coagulation area was increased; however, there have been no comparison studies that indicate this technique is of clinical benefit. Combined chemotherapy and local chemotherapy have been reported. Goldberg and colleagues reported animal models using systemic doxorubicin, which increases the size of the treated ablation

site.156 Goldberg and colleagues studied the benefits of combining RFA with ethanol injection and doxorubicin injection into the tumor site.153,156 With ethanol injection, they found increased areas of coagulation when the ethanol was injected before the RF heating. The size of the lesion increased from 6.7 mm from ablation alone to 10.1 mm with the ethanol combination. There was no improvement when ethanol was injected after RFA. Increase in the tissue impedance was found after the alcohol injection. In another animal model, Goldberg’s group injected the chemotherapeutic agent doxorubicin into the tumor before and after RFA. The authors found that doxorubicin improved tumor coagulation when it followed the ablation but not before.157 The increase in the ablation site changed from 6.7 mm in the ablation site alone to 11.1 mm with the combination therapy. Haaga and associates186 reported the combination used with local deposition of chemotherapeutic-laden biodegradable polymer, which increased the size of the ablation. Goldberg and colleagues reported on a small pilot group of 10 patients who received systemic doxorubicin before percutaneous ablation.155 They noted an increased ablation site 2 to 4 weeks after the ablation in the group with combined therapy compared with the RF group alone. Since this pilot group was reported in 2002, no human studies have yet been reported. Transarterial chemoembolization and embolic therapy have been combined with RFA. Kitamoto and associates reported on a series of 21 patients who had doxorubicin chemoembolization and percutaneous RFA or RFA alone for small nodules (<3 cm).264 The size of the induced necrosis was greater for the combined group than the RFAalone group. Mechanical reduction of organ blood flow has been used. De Baere and colleagues reported in a preliminary study that temporary occlusion of a hepatic vein by catheter occlusion facilitated ablation of larger liver lesions.92 In a small series, 5-cm lesions were more completely ablated with venous occlusion than a matched control group (P < .0003). Numerous surgeons performed temporary occlusion of the portal vein using the Pringle maneuver, which does provide improved treatment effect. Reduction of blood flow using vasoactive drugs to reduce flow by pharmacologic means has been reported. In animals, direct arterial injection of epinephrine has been performed to reduce general blood flow, but this has not been routinely performed clinically. Some investigators have used powerful systemic agents affecting the general vascular system, with variable results. Horkan and coworkers231 and Goldberg and colleagues156 used powerful systemic drugs such as epinephrine, halothane, and arsenic trioxide, which did show improvement in the size of ablation sites. The problem with this approach is that the systemic effects necessitate that general anesthesia is used to permit monitoring, modulation, and treatment of any adverse vascular effects. Our group has studied physiologic vasoactive drugs, which react specifically with organ-specific receptor sites so that there is minimal systemic effect. The term physiologic drugs is used because we have concentrated on agents that do not affect the general vascular system, but have greater effects in target organs with receptor sites. The rationale of this approach is that if blood flow to the target site can be reduced, the heat sink effect is minimized and the deposition of heat is more efficient. Two approaches can be used. If the vessel to the target organ can be constricted, flow is reduced. An alternative approach is to exploit a “steal” effect. If the normal smooth muscle–laden vessels of the normal organ containing the cancer can be dilated by specific agents, the blood flow would be shunted from the cancerous area, stealing the blood and reducing the heat sink effect. Han Ping Wu of our group (personal communication, 2008) has reported this steal

CHAPTER 67  Image-Guided Aspirations and Biopsies method in an animal model and showed a 100% increase in the size of the ablation site. Haaga (personal communication, 2008) treated a patient with a hypervascular renal tumor according to the Belmont Report of FDA guidelines for nonapproved usage of acetazolamide and improved RFA treatment.

Future Clinical Applications for Radiofrequency One area of great possibility for future clinical applications is treatment of pancreatic cancer. Two factors that make this a promising area are the high mortality of the disease and the heat sink effect of blood vessels. Pancreatic cancer is well recognized as a hypovascular or avascular process; thermal ablation might be effective in clearing the perivascular tumor because of the heat sink effect. Also, animal models by Merkle and associates have shown that RFA can kill pancreatic tissue without inciting a continuing inflammatory process.344 The thyroid, breast, and tongue also may become sites for ablation. Thyroid ablations have been reported by Pacella and coworkers.393 There is not sufficient experience or information to determine if this will become a standard treatment. Nour and associates reported the use of RFA for destruction of the base of the tongue to improve sleep apnea.382 Primary breast cancer lesions have been ablated with RFA129 and cryoablation.

REFERENCES 1. Abitabile P, Hartl U, Lange J, et al: Radiofrequency ablation permits an effective treatment for colorectal liver metastasis. Eur J Surg Oncol 33:67–71, 2007. 2. Adams DB, Anderson MC: Percutaneous catheter drainage compared with internal drainage in the management of pancreatic pseudocyst. Ann Surg 215:571–576, discussion 576–578, 1992. 3. Adams RF, Gleeson FV: Percutaneous image-guided cutting-needle biopsy of the pleura in the presence of a suspected malignant effusion. Radiology 219:510–514, 2001. 4. Adler O, Rosenbort A, Peleg H: Fine needle aspiration biopsy of mediastinal masses: Evaluation of 136 experiences. AJR Am J Roentgenol 140:893–896, 1983. 5. Agarwal AK, Seely JM, Matzinger FR, et al: Pleural mesothelioma: Sensitivity and incidence of needle track seeding after image-guided biopsy versus surgical biopsy. Radiology 241:589–594, 2006. 6. Agid R, Sklair-Levy M, Bloom AI, et al: CT-guided biopsy with cutting-edge needle for the diagnosis of malignant lymphoma: Experience of 267 biopsies. Clin Radiol 58:147–157, 2003. 7. Ahrar K, Price RE, Wallace MR, et al: Percutaneous radiofrequency ablation of lung tumors in a large animal model. J Vasc Interv Radiol 14:1037–1043, 2003. 8. Aideyan OA, Schmidt AJ, Trenkner SW, et al: CT-guided percutaneous biopsy of pancreas transplants. Radiology 201:825–828, 1996. 9. Akeboshi M, Yamakado K, Nakatsuka A, et al: Percutaneous radiofrequency ablation of lung neoplasms: Initial therapeutic response. J Vasc Interv Radiol 15:463–470, 2004. 10. Akgun Y, Tacyildiz IH, Celik Y: Amebic liver abscess: Changing trends over 20 years. World J Surg 23:102–106, 1999. 11. Akhan O, Altinok D, Ozmen MN, et al: Correlation between the grade of tumoral invasion and pain relief in patients with celiac ganglia block. AJR Am J Roentgenol 168:1565–1567, 1997. 12. Allaf ME, Varkarakis IM, Bhayani SB, et al: Pain control requirements for percutaneous ablation of renal tumors: Cryoablation versus radiofrequency ablation—Initial observations. Radiology 237:366–370, 2005. 13. Allison DJ, Adam A: Percutaneous liver biopsy and track embolization with steel coils. Radiology 169:261–263, 1988. 14. Aloia TA, Vauthey JN, Loyer EM, et al: Solitary colorectal liver metastasis: Resection determines outcome. Arch Surg 141:460–466, discussion 466–467, 2006.

2573

15. Ambrogi MC, Lucchi M, Dini P, et al: Percutaneous radiofrequency ablation of lung tumours: Results in the mid-term. Eur J Cardiothorac Surg 30:177–183, 2006. 16. Andersson R, Cwikiel W: Percutaneous cystogastrostomy in patients with pancreatic pseudocysts. Eur J Surg 168:345–348, 2002. 17. Andriole JG, Haaga JR, Adams RB, et al: Biopsy needle characteristics assessed in the laboratory. Radiology 148:659–662, 1983. 18. Aschoff AJ, Rafie N, Jesberger JA, et al: Thermal lesion conspicuity following interstitial radiofrequency thermal tumor ablation in humans: A comparison of STIR, turbo-spin echo T2-weighted, and contrast-enhanced T1-weighted MR images at 0.2 T. J Magn Reson Imaging 12:584–589, 2000. 19. Atwell TD, Gorman B, Larson TS, et al: Pancreas transplants: Experience with 232 percutaneous US-guided biopsy procedures in 88 patients. Radiology 231:845–849, 2004. 20. Balen FG, Little A, Smith AC, et al: Biopsy of inoperable pancreatic tumors does not adversely influence patient survival time. Radiology 193:753–755, 1994. 21. Bang S, Chung HW, Park SW, et al: The clinical usefulness of 18-fluorodeoxyglucose positron emission tomography in the   differential diagnosis, staging, and response evaluation after concurrent chemoradiotherapy for pancreatic cancer. J Clin Gastroenterol 40:923–929, 2006. 22. Barbaric ZL, Hall T, Cochran ST, et al: Percutaneous nephrostomy: Placement under CT and fluoroscopy guidance. AJR Am J Roentgenol 169:151–155, 1997. 23. Barei DP, Moreau G, Scarborough MT, et al: Percutaneous radiofrequency ablation of osteoid osteoma. Clin Orthop 373:115–124, 2000. 24. Barnett CC, Jr, Curley SA: Ablative techniques for hepatocellular carcinoma. Semin Oncol 28:487–496, 2001. 25. Barthet M, Bugallo M, Moreira LS, et al: Management of cysts and pseudocysts complicating chronic pancreatitis: A retrospective study of 143 patients. Gastroenterol Clin Biol 17:270–276, 1993. 26. Bass J, Di Lorenzo M, Desjardins JG, et al: Blunt pancreatic   injuries in children: The role of percutaneous external drainage   in the treatment of pancreatic pseudocysts. J Pediatr Surg 23:721–724, 1988. 27. Basu S, Marini CP, Bauman FG, et al: Comparative study of biological glues: Cryoprecipitate glue, two-component fibrin sealant, and French glue. Ann Thorac Surg 62:317–318, 1996. 28. Bateson MC, Hopwood D, Duguid HL, et al: A comparative trial of liver biopsy needles. J Clin Pathol 33:131–133, 1980. 29. Becker JA, Schneider M, Staiano S, et al: Needle aspiration and B mode scanning. Invest Radiol 10:173–175, 1975. 30. Beland MD, Mayo-Smith WW, Dupuy DE, et al: Diagnostic yield of 58 consecutive imaging-guided biopsies of solid renal masses: Should we biopsy all that are indeterminate? AJR Am J Roentgenol 188:792–797, 2007. 31. Belfiore G, Moggio G, Tedeschi E, et al: CT-guided radiofrequency ablation: A potential complementary therapy for patients with unresectable primary lung cancer—A preliminary report of 33 patients. AJR Am J Roentgenol 183:1003–1011, 2004. 32. Berger H, Permanetter W, Steiner W: Percutaneous biopsy technic: Aspiration fine needle and cutting biopsy cannula in an experimental comparison. Ultraschall Med 9:128–131, 1988. 33. Bernardino ME, Walther MM, Phillips VM, et al: CT-guided adrenal biopsy: Accuracy, safety, and indications. AJR Am J Roentgenol 144:67–69, 1985. 34. Berquist TH, Bailey PB, Cortese DA, et al: Transthoracic needle biopsy: Accuracy and complications in relation to location and type of lesion. Mayo Clin Proc 55:475–481, 1980. 35. Bilbao JI, Alejandre PL, Longo JM, et al: Percutaneous transgastric cystoduodenostomy in the treatment of a pancreatic pseudocyst: A new approach. Cardiovasc Intervent Radiol 18:422–425, 1995. 36. Bissonnette RT, Gibney RG, Berry BR, et al: Fatal carcinoid crisis after percutaneous fine-needle biopsy of hepatic metastasis: Case report and literature review. Radiology 174:751–752, 1990.

2574

PART III  Image-Guided Procedures

37. Blake MA, Sweeney AT, Kalra MK, et al: Collision adrenal tumors on PET/CT. AJR Am J Roentgenol 183:864–865, 2004. 38. Bohm T, Hilger I, Muller W, et al: Saline-enhanced radiofrequency ablation of breast tissue: An in vitro feasibility study. Invest Radiol 35:149–157, 2000. 39. Bojarski JD, Dupuy DE, Mayo-Smith WW: CT imaging findings of pulmonary neoplasms after treatment with radiofrequency ablation: Results in 32 tumors. AJR Am J Roentgenol 185:466–471, 2005. 40. Boll DT, Lewin JS, Duerk JL, et al: Do surgical clips interfere with radiofrequency thermal ablation? AJR Am J Roentgenol 180:1557–1560, 2003. 41. Bosanac ZB, Lisanin L: Percutaneous drainage of hydatid cyst in the liver as a primary treatment: Review of 52 consecutive cases with long-term follow-up. Clin Radiol 55:839–848, 2000. 42. Bourgouin PM, Shepard JA, McLoud TC, et al: Transthoracic needle aspiration biopsy: Evaluation of the blood patch technique. Radiology 166:93–95, 1988. 43. Bowles BJ, Machi J, Limm WM, et al: Safety and efficacy of radiofrequency thermal ablation in advanced liver tumors. Arch Surg 136:864–869, 2001. 44. Brandt KR, Charboneau JW, Stephens DH, et al: CT- and US-guided biopsy of the pancreas. Radiology 187:99–104, 1993. 45. Bret PM, Fond A, Casola G, et al: Abdominal lesions: A prospective study of clinical efficacy of percutaneous fine-needle biopsy. Radiology 159:345–346, 1986. 46. Bret PM, Sente JM, Bretagnolle M, et al: Ultrasonically guided fine-needle biopsy in focal intrahepatic lesions: Six years’ experience. Can Assoc Radiol J 37:5–8, 1986. 47. Brillet PY, Paradis V, Brancatelli G, et al: Percutaneous radiofrequency ablation for hepatocellular carcinoma before liver transplantation: A prospective study with histopathologic comparison. AJR Am J Roentgenol 186:S296–S305, 2006. 48. Bungay HK, Adams RF, Morris CM, et al: Cutting needle biopsy in the diagnosis of clinically suspected non-carcinomatous disease of the lung. Br J Radiol 73:349–355, 2000. 49. Butch RJ, Wittenberg J, Mueller PR, et al: Presacral masses after abdominoperineal resection for colorectal carcinoma: The need for needle biopsy. AJR Am J Roentgenol 144:309–312, 1985. 50. Buy JN, Moss AA, Singler RC: CT guided celiac plexus and splanchnic nerve neurolysis. J Comput Assist Tomogr 6:315–319, 1982. 51. Caliendo MV, Lee DE, Queiroz R, et al: Sclerotherapy with use of doxycycline after percutaneous drainage of postoperative lymphoceles. J Vasc Interv Radiol 12:73–77, 2001. 52. Callstrom MR, Atwell TD, Charboneau JW, et al: Painful metastases involving bone: Percutaneous image-guided cryoablation—prospective trial interim analysis. Radiology 241:572–580, 2006. 53. Callstrom MR, Charboneau JW: Percutaneous ablation: Safe, effective treatment of bone tumors. Oncology 19:22–26, 2005. 54. Cantwell CP, O’Byrne J, Eustace S: Radiofrequency ablation of osteoid osteoma with cooled probes and impedance-control energy delivery. AJR Am J Roentgenol 186:S244–S248, 2006. 55. Caoili EM, Korobkin M, Brown RK, et al: Differentiating adrenal adenomas from nonadenomas using (18)F-FDG PET/CT: Quantitative and qualitative evaluation. Acad Radiol 14:468–475, 2007. 56. Carr ME, Jr, Martin EJ, Carr SL: Delayed, reduced, or inhibited thrombin production reduces platelet contractile force and results in weaker clot formation. Blood Coagul Fibrinolysis 13:193–197, 2002. 57. Carruthers DM, Connor S, Howie AJ, et al: Percutaneous image-guided biopsy of lung nodules in the assessment of disease activity in Wegener’s granulomatosis. Rheumatology (Oxf) 39:776–782, 2000. 58. Casola G, Nicolet V, Van Sonnenberg E, et al: Unsuspected pheochromocytoma: Risk of blood-pressure alterations during percutaneous adrenal biopsy. Radiology 159:733–735, 1986. 59. Casola G, van Sonnenberg E, Neff CC, et al: Abscesses in Crohn disease: Percutaneous drainage. Radiology 163:19–22, 1987. 60. Catalano OA, Hahn PF, Hooper DC, et al: Efficacy of percutaneous abscess drainage in patients with vancomycin-resistant enterococci. AJR Am J Roentgenol 175:533–536, 2000.

61. Cellier C, Cuenod CA, Deslandes P, et al: Symptomatic hepatic cysts: Treatment with single-shot injection of minocycline hydrochloride. Radiology 206:205–209, 1998. 62. Ch Yu S, Hg Lo R, Kan PS, et al: Pyogenic liver abscess: Treatment with needle aspiration. Clin Radiol 52:912–916, 1997. 63. Cha CH, Lee FT, Jr, Gurney JM, et al: CT versus sonography for monitoring radiofrequency ablation in a porcine liver. AJR Am J Roentgenol 175:705–711, 2000. 64. Chakravarti R, Singh V, Isaac R, et al: Fatal paradoxical pulmonary air embolism complicating percutaneous computed tomography-guided needle biopsy of the lung. Australas Radiol 48:204–206, 2004. 65. Chen YK, Hsieh DS, Liao CS, et al: Utility of FDG-PET for investigating unexplained serum AFP elevation in patients with suspected hepatocellular carcinoma recurrence. Anticancer Res 25:4719–4725, 2005. 66. Chin J: The effect of radiofrequency catheter ablation on permanent pacemakers: An experimental study. Pacing Clin Electrophysiol 12:23–29, 1990. 67. Cho YK, Rhim H, Ahn YS, et al: Percutaneous radiofrequency ablation therapy of hepatocellular carcinoma using multitined expandable electrodes: Comparison of subcapsular and nonsubcapsular tumors. AJR Am J Roentgenol 186:S269–S274, 2006. 68. Choi D, Lim HK, Kim MJ, et al: Therapeutic efficacy and safety of percutaneous radiofrequency ablation of hepatocellular carcinoma abutting the gastrointestinal tract. AJR Am J Roentgenol 183:1417–1424, 2004. 69. Choi H, Loyer EM, DuBrow RA, et al: Radio-frequency ablation of liver tumors: Assessment of therapeutic response and complications. Radiographics 21:S41–S54, 2001. 70. Chopra S, Dodd GD, 3rd, Chanin MP, et al: Radiofrequency ablation of hepatic tumors adjacent to the gallbladder: Feasibility and safety. AJR Am J Roentgenol 180:697–701, 2003. 71. Chopra S, Dodd GD, 3rd, Chintapalli KN, et al: Tumor recurrence after radiofrequency thermal ablation of hepatic tumors: Spectrum of findings on dual-phase contrast-enhanced CT. AJR Am J Roentgenol 177:381–387, 2001. 72. Christopherson P: Preoperative pancreas aspiration biopsies. Acta Pathol Microbiol Scand 112:28–33, 1970. 73. Chuang VP, Alspaugh JP: Sheath needle for liver biopsy in high-risk patients. Radiology 166:261–262, 1988. 74. Chung IK, Park MJ, Kwon KT, et al: The factors related to the prognosis of solitary hepatocellular carcinoma after radiofrequency ablation. Korean J Hepatol 11:371–380, 2005. 75. Cioni R, Armillotta N, Bargellini I, et al: CT-guided radiofrequency ablation of osteoid osteoma: Long-term results. Eur Radiol 14:1203– 1208, 2004. 76. Collings CL, Westcott JL, Banson NL, et al: Pneumothorax and dependent versus nondependent patient position after needle biopsy of the lung. Radiology 210:59–64, 1999. 77. Collins BT, Cramer HM: Fine-needle aspiration cytology of islet cell tumors. Diagn Cytopathol 15:37–45, 1996. 78. Conn HO, Yesner R: A re-evaluation of needle biopsy in the diagnosis of metastatic cancer of the liver. Ann Intern Med 59:53–61, 1963. 79. Cox JE, Chiles C, McManus CM, et al: Transthoracic needle aspiration biopsy: Variables that affect risk of pneumothorax. Radiology 212:165–168, 1999. 80. Cox MR, Davies RP, Bowyer RC, et al: Percutaneous cystogastrostomy for treatment of pancreatic pseudocysts. Aust N Z J Surg 63:693–698, 1993. 81. Criado E, De Stefano AA, Weiner TM, et al: Long term results of percutaneous catheter drainage of pancreatic pseudocysts. Surg Gynecol Obstet 175:293–298, 1992. 82. Cronan J: Cavernous hemangioma of the liver: Role of percutaneous biopsy. Radiology 116:135–138, 1998. 83. Cronan JJ, Esparza AR, Dorfman GS, et al: Cavernous hemangioma of the liver: Role of percutaneous biopsy. Radiology 166:135–138, 1988. 84. Curley SA: Radiofrequency ablation of malignant liver tumors. Ann Surg Oncol 10:338–347, 2003.

CHAPTER 67  Image-Guided Aspirations and Biopsies 85. Curley SA, Izzo F, Ellis LM, et al: Radiofrequency ablation of hepatocellular cancer in 110 patients with cirrhosis. Ann Surg 232:381–391, 2000. 86. Curley SA, Marra P, Beaty K, et al: Early and late complications after radiofrequency ablation of malignant liver tumors in 608 patients. Ann Surg 239:450–458, 2004. 87. D’Agostino HB, van Sonnenberg E, Sanchez RB, et al: Treatment of pancreatic pseudocysts with percutaneous drainage and octreotide: Work in progress. Radiology 187:685–688, 1993. 88. Reference deleted in proofs. 89. Dahnert WF, Hoagland MH, Hamper UM, et al: Fine-needle aspiration biopsy of abdominal lesions: Diagnostic yield for different needle tip configurations. Radiology 185:263–268, 1992. 90. Davies RP, Cox MR, Wilson TG, et al: Percutaneous cystogastrostomy with a new catheter for drainage of pancreatic pseudocysts and fluid collections. Cardiovasc Intervent Radiol 19:128–131, 1996. 91. de Baere T: Radiofrequency ablation of lung and bone tumors. Cancer Radiother 10:430–436, 2006. 92. de Baere T, Besoud B, Dromain C, et al: Percutaneous radiofrequency ablation of hepatic tumors during temporary venous occlusion. AJR Am J Roentgenol 178:53–59, 2002. 93. de Baere T, Denys A, Wood BJ, et al: Radiofrequency liver ablation: Experimental comparative study of water-cooled versus expandable systems. AJR Am J Roentgenol 176:187–192, 2001. 94. de Baere T, Palussiere J, Auperin A, et al: Midterm local efficacy and survival after radiofrequency ablation of lung tumors with minimum follow-up of 1 year: Prospective evaluation. Radiology 240:587–596, 2006. 95. de Berg JC, Pattynama PM, Obermann WR, et al: Percutaneous computed-tomography-guided thermocoagulation for osteoid osteoma. Lancet 346:350–351, 1995. 96. D’Egidio A, Schein M: Pancreatic pseudocysts: A proposed classification and its management implications. Br J Surg 78:981–984, 1991. 97. de Kerviler E, de Bazelaire C, Mounier N, et al: Image-guided core-needle biopsy of peripheral lymph nodes allows the diagnosis of lymphomas. Eur Radiol 17:843–849, 2007. 98. de Kerviler E, Guermazi A, Zagdanski AM, et al: Image-guided core-needle biopsy in patients with suspected or recurrent lymphomas. Cancer 89:647–652, 2000. 99. deSouza NM, Reidy JF, Koffman CG: Arteriovenous fistulas complicating biopsy of renal allografts: Treatment of bleeding with superselective embolization. AJR Am J Roentgenol 156:507–510, 1991. 100. Deveney CW, Lurie K, Deveney KE: Improved treatment of intraabdominal abscess: A result of improved localization, drainage, and patient care, not technique. Arch Surg 123:1126–1130, 1988. 101. Diaz LK, Wiley EL, Venta LA: Are malignant cells displaced by large-gauge needle core biopsy of the breast? AJR Am J Roentgenol 173:1303–1313, 1999. 102. Dickey JE, Haaga JR, Stellato TA, et al: Evaluation of computed tomography guided percutaneous biopsy of the pancreas. Surg Gynecol Obstet 163:497–503, 1986. 103. Dodd GD, 3rd, Napier D, Schoolfield JD, et al: Percutaneous radiofrequency ablation of hepatic tumors: Postablation syndrome. AJR Am J Roentgenol 185:51–57, 2005. 104. Dohlgren SE: Transthoracic Needle Biopsy, Stockholm, 1966, Almquist & Wiksell/Bebers. 105. Dondelinger RF, Kurdziel JC, Gathy C: Percutaneous treatment of pyogenic liver abscess: A critical analysis of results. Cardiovasc Intervent Radiol 13:174–182, 1990. 106. Dowlati A, Haaga J, Remick SC, et al: Sequential tumor biopsies in early phase clinical trials of anticancer agents for pharmacodynamic evaluation. Clin Cancer Res 7:2971–2976, 2001. 107. Dromain C, de Baere T, Elias D, et al: Hepatic tumors treated with percutaneous radiofrequency ablation: CT and MR imaging follow-up. Radiology 223:255–262, 2002. 108. Dupuy DE, DiPetrillo T, Gandhi S, et al: Radiofrequency ablation followed by conventional radiotherapy for medically inoperable stage I non-small cell lung cancer. Chest 129:738–745, 2006.

2575

109. Dupuy DE, Zagoria RJ, Akerley W, et al: Percutaneous radiofrequency ablation of malignancies in the lung. AJR Am J Roentgenol 174:57–59, 2000. 110. Duvnjak M, Vucelic B, Rotkvic I, et al: Assessment of value of pancreatic pseudocyst amylase concentration in the treatment of pancreatic pseudocysts by percutaneous evacuation. J Clin Ultrasound 20:183–186, 1992. 111. Dzik S: The use of blood components prior to invasive bedside procedures: A critical appraisal. In Transfusion Therapy: Clinical Principles and Practice, Bethesda, MD, 2005, AABB Press, pp 151–194. 112. Edoute Y, Tibon-Fischer O, Ben-Haim SA, et al: Imaging-guided and nonimaging-guided fine needle aspiration of liver lesions: Experience with 406 patients. J Surg Oncol 48:246–251, 2006. 113. Elvin A, Andersson T, Scheibenpflug L, et al: Biopsy of the pancreas with a biopsy gun. Radiology 176:677–679, 1990. 114. Engelmann K, Mack MG, Strub R, et al: CT-guided percutaneous intratumoral chemotherapy with a novel cisplatin/epinephrine injectable gel for the treatment of inoperable malignant liver tumors. Rofo 172:1020–1027, 2000. 115. England RE, McDermott VG, Smith TP, et al: Percutaneous cholecystostomy: Who responds? AJR Am J Roentgenol 168:1247–1251, 1997. 116. Erturk SM, Mortele KJ, Tuncali K, et al: Fine-needle aspiration biopsy of solid pancreatic masses: Comparison of CT and endoscopic sonography guidance. AJR Am J Roentgenol 187:1531–1535, 2006. 117. Evander A, Ihse I, Lunderquist A, et al: Percutaneous cytodiagnosis of carcinoma of the pancreas and bile duct. Ann Surg 188:90–92, 1978. 118. Evans WK, Ho CS, McLoughlin MJ, et al: Fatal necrotizing pancreatitis following fine-needle aspiration biopsy of the pancreas. Radiology 141:61–62, 1981. 119. Fagelman D, Chess Q: Nonaspiration fine-needle cytology of the liver: A new technique for obtaining diagnostic samples. AJR Am J Roentgenol 155:1217–1219, 1990. 120. Falstrom JK, Moore MM, Caldwell SH, et al: Use of fibrin sealant to reduce bleeding after needle liver biopsy in an anticoagulated canine model: Work in progress. J Vasc Interv Radiol 10:457–462, 1999. 121. Fataar S: Percutaneous drainage of pancreatic pseudocysts: Technique and problems. Australas Radiol 34:334–338, 1982. 122. Fataar S, Harries-Jones EP, Dalrymple R, et al: Fine-needle aspiration biopsy of malignant tumours of the abdomen. S Afr Med J 62:637–641, 1982. 123. Fenger-Eriksen C, Anker-Moller E, Hoslop J, et al: Thromboelastographic whole blood clot formation after ex vivo addition of plasma substitutes: Improvements of the induced coagulopathy with fibrinogen concentrate. Br J Anaesth 94:324–329, 2005. 124. Ferrucci JT, Jr, Wittenberg J: CT biopsy of abdominal tumors: Aids for lesion localization. Radiology 129:739–744, 1978. 125. Ferrucci JT, Wittenberg J, Margolies MN, et al: Malignant seeding of the tract after thin-needle aspiration biopsy. Radiology 130:345–346, 1979. 126. Ferrucci JT, Jr, Wittenberg J, Muller PR, et al: Diagnosis of abdominal malignancy by radiologic fine-needle aspiration biopsy. AJR Am J Roentgenol 134:323–330, 1980. 127. Fish GD, Stanley JH, Miller KS, et al: Postbiopsy pneumothorax: Estimating the risk by chest radiography and pulmonary function tests. AJR Am J Roentgenol 150:71–74, 1988. 128. Fitzgibbon DR, Schmiedl UP, Sinanan MN: Computed tomographyguided neurolytic celiac plexus block with alcohol complicated by superior mesenteric venous thrombosis. Pain 92:307–310, 2001. 129. Fornage BD, Sneig N, Ross MI, et al: Small (< or = 2-cm) breast cancer treated with US-guided radiofrequency ablation: Feasibility study. Radiology 231:215–224, 2004. 130. Forsgren L, Orell S: Aspiration cytology in carcinoma of the pancreas. Surgery 73:38–42, 1973. 131. Freeny PC, Hauptmann E, Althaus SJ, et al: Percutaneous CT-guided catheter drainage of infected acute necrotizing pancreatitis: Techniques and results. AJR Am J Roentgenol 170:969–975, 1998.

2576

PART III  Image-Guided Procedures

132. Freeny PC, Lewis GP, Traverso LW, et al: Infected pancreatic fluid collections: Percutaneous catheter drainage. Radiology 167:435–441, 1988. 133. Frich L, Halvorsen PS, Skulstad H, et al: Microbubbles in the pulmonary artery generated during experimental hepatic radiofrequency ablation is correlated with increased pulmonary arterial pressure. J Vasc Interv Radiol 18:437–442, 2007. 134. Fukushima A, Ashizawa K, Aso N, et al: CT-guided needle biopsy of the lung: Factors affecting risk of complications. Nippon Igaku Hoshasen Gakkai Zasshi 61:96–99, 2001. 135. Gadaleta C, Catino A, Ranieri G, et al: Radiofrequency thermal ablation of 69 lung neoplasms. J Chemother 16(Suppl 5):86–89, 2004. 136. Gadaleta C, Mattioli V, Colucci G, et al: Radiofrequency ablation of 40 lung neoplasms: Preliminary results. AJR Am J Roentgenol 183:361–368, 2004. 137. Garcia JC, Persky SE, Bonis PA, et al: Abscesses in Crohn’s disease: Outcome of medical versus surgical treatment. J Clin Gastroenterol 32:409–412, 2001. 138. Gazelle GS, Haaga JR: Biopsy needle characteristics. Cardiovasc Intervent Radiol 14:13–16, 1991. 139. Gazelle GS, Haaga JR, Neuhauser D: Hemostatic protein-polymer sheath: New method to enhance hemostasis at percutaneous biopsy. Radiology 175:671–674, 1990. 140. Gazelle GS, Haaga JR, Rowland DY: Effect of needle gauge, level of anticoagulation, and target organ on bleeding associated with aspiration biopsy: Work in progress. Radiology 183:509–513, 1992. 141. Gazelle GS, Haaga JR, Rowland DY: Bleeding due to needle biopsy: Effect of venopirin in an animal model and implications for humans. J Vasc Interv Radiol 4:305–310, 1993. 142. Gervais DA, Arellano RS, McGovern FJ, et al: Radiofrequency ablation of renal cell carcinoma, Part 1: Indications, results, and role in patient management over a 6-year period and ablation of 100 tumors. AJR Am J Roentgenol 185:64–71, 2005. 143. Gervais DA, Arellano RS, Mueller PR: Percutaneous radiofrequency ablation of ovarian cancer metastasis to the liver: Indications, outcomes, and role in patient management. AJR Am J Roentgenol 187:746–750, 2006. 144. Gervais DA, McGovern FJ, Arellano RS, et al: Radiofrequency ablation of renal cell carcinoma, Part 2: Lessons learned with ablation of 100 tumors. AJR Am J Roentgenol 185:72–80, 2005. 145. Gerzof SF, Johnson WC, Robbins AH, et al: Expanded criteria for percutaneous abscess drainage. Arch Surg 120:227–232, 1985. 146. Gerzof SG, Robbins AH, Johnson WC, et al: Percutaneous catheter drainage of abdominal abscesses: A five-year experience. N Engl J Med 305:653–657, 1981. 147. Gilliland JD, Spies JB, Brown SB, et al: Lymphoceles: Percutaneous treatment with povidone-iodine sclerosis. Radiology 171:227–229, 1989. 148. Gimenez A, Martinez-Noguera A, Donoso L, et al: Percutaneous neurolysis of the celiac plexus via the anterior approach with sonographic guidance. AJR Am J Roentgenol 161:1061–1063, 1993. 149. Giorgio A, de Stefano G, Di Sarno A, et al: Percutaneous needle aspiration of multiple pyogenic abscesses of the liver: 13-year single-center experience. AJR Am J Roentgenol 187:1585–1590, 2006. 150. Giorgio A, Tarantiono L, de Stefano G, et al: Hydatid liver cyst: An 11-year experience of treatment with percutaneous aspiration and ethanol injection. J Ultrasound Med 20:729–738, 2001. 151. Goetz MP, Callstrom MR, Charboneau JW, et al: Percutaneous image-guided radiofrequency ablation of painful metastases involving bone: A multicenter study. J Clin Oncol 22:300–306, 2004. 152. Gohari A, Haramati LB: Complications of CT scan-guided lung biopsy: Lesion size and depth matter. Chest 126:666–668, 2004. 153. Goldberg SN, Gazelle GS, Dawson SL, et al: Tissue ablation with radiofrequency: Effect of probe size, gauge, duration, and temperature on lesion volume. Acad Radiol 2:399–404, 1995. 154. Goldberg SN, Grassi CJ, Cardella JF, et al: Image-guided tumor ablation: Standardization of terminology and reporting criteria. Radiology 235:728–739, 2005.

155. Goldberg SN, Kamel IR, Kruskal JB, et al: Radiofrequency ablation of hepatic tumors: Increased tumor destruction with adjuvant liposomal doxorubicin therapy. AJR Am J Roentgenol 179:93–101, 2002. 156. Goldberg SN, Kruskal JB, Oliver BS, et al: Percutaneous tumor ablation: Increased coagulation by combining radio-frequency ablation and ethanol instillation in a rat breast tumor model. Radiology 217:827–831, 2000. 157. Goldberg SN, Saldinger PF, Gazelle GS, et al: Percutaneous tumor ablation: Increased necrosis with combined radiofrequency ablation and intratumoral doxyrubicin injection in a rat breast tumor model. Radiology 220:420–427, 2001. 158. Goldschmidt N, Libson E, Bloom A, et al: Clinical utility of computed tomography-guided core needle biopsy in the diagnostic re-evaluation of patients with lymphoproliferative disorders and suspected disease progression. Ann Oncol 14:1438–1441, 2003. 159. Goldstein HM, Zornoza J, Wallace S, et al: Percutaneous fine needle aspiration biopsy of pancreatic and other abdominal masses. Radiology 123:319–322, 1977. 160. Goralnik CH, O’Connel DM, el Yousef SJ, et al: CT-guided cuttingneedle biopsies of selected chest lesions. AJR Am J Roentgenol 151:903–907, 1988. 161. Gottumukkala VN, Sharma SK, Philip J: Assessing platelet and fibrinogen contribution to clot strength using modified thromboelastography in pregnant women. Anesth Analg 89:1453, 1999. 162. Green BT: Splenic abscess: Report of six cases and review of the literature. Am Surg 67:80–85, 2001. 163. Greene R, Szyfelbein WM, Isler RJ, et al: Supplementary tissue-core histology from fine-needle transthoracic aspiration biopsy. AJR Am J Roentgenol 144:787–792, 1985. 164. Greif J, Marmor S, Schwarz Y, et al: Percutaneous core needle biopsy vs. fine needle aspiration in diagnosing benign lung lesions. Acta Cytol 43:756–760, 1999. 165. Greif J, Marmur S, Schwarz Y, et al: Percutaneous core cutting needle biopsy compared with fine-needle aspiration in the diagnosis of peripheral lung malignant lesions: Results in 156 patients. Cancer 84:144–147, 1998. 166. Greif J, Staroselsky AN, Gernjac M, et al: Percutaneous core needle biopsy in the diagnosis of mediastinal tumors. Lung Cancer 25:169–173, 1999. 167. Gronemeyer DH, Schirp S, Gevargez A: Image-guided radiofrequency ablation of spinal tumors: Preliminary experience with an expandable array electrode. Cancer J 8:33–39, 2002. 168. Gronvall J, Gronvall S, Hegedus V: Ultrasound-guided drainage of fluid-containing masses using angiographic catheterization techniques. AJR Am J Roentgenol 129:997–1002, 1977. 169. Grossman M: Percutaneous treatment (including pseudocystogastrostomy) of 74 pancreatic pseudocysts. Radiology 173:493–497, 1989. 170. Gruessner RW, Fasola C, Benedetti E, et al: Laparoscopic drainage of lymphoceles after kidney transplantation: Indications and limitations. Surgery 117:288–295, 1995. 171. Gunabushanam G, Sharma S, Thulkar S, et al: Radiofrequency ablation of liver metastases from breast cancer: Results in 14 patients. J Vasc Interv Radiol 18:67–72, 2007. 172. Gupta S, Ahrar K, Morello FA, Jr, et al: Masses in or around the pancreatic head: CT-guided coaxial fine-needle aspiration biopsy with a posterior transcaval approach. Radiology 222:63–69, 2002. 173. Gupta RK, alAnsari AG: Needle aspiration cytology in the diagnosis of mucinous cystadenocarcinoma of pancreas: A study of five cases with an emphasis on utility and differential diagnosis. Int J Pancreatol 15:149–153, 1994. 174. Gupta S, Krishnamurthy S, Broemeling LD, et al: Small (
CHAPTER 67  Image-Guided Aspirations and Biopsies 176. Ha HK, Sachs PB, Haaga JR, et al: CT-guided liver biopsy: An update. Clin Imaging 15:99–104, 1991. 177. Haaga J: Definitive treatment of a large pyogenic liver abscess with CT guidance. Cleve Clin Q 43:85–88, 1976. 178. Haaga JR: New techniques for CT-guided biopsies. AJR Am J Roentgenol 133:633–641, 1979. 179. Haaga JR: Imaging intraabdominal abscesses and nonoperative drainage procedures. World J Surg 14:204–209, 1990. 180. Haaga JR: Imaging-guided microsurgery and the future of radiology. AJR Am J Roentgenol 175:303–304, 2000. 181. Haaga JR, Alfidi RJ: Precise biopsy localization by computer tomography. Radiology 118:603–607, 1976. 182. Haaga JR, Alfidi RJ, Havrilla TR, et al: CT detection and aspiration of abdominal abscesses. AJR Am J Roentgenol 128:465–474, 1977. 183. Haaga JR, Alfidi RJ, MacIntyre WJ, et al: CT longitudinal scan. AJR Am J Roentgenol 127:1059–1060, 1976. 184. Haaga JR, Beale SM: Use of CO2 to move structures as an aid to percutaneous procedures. Radiology 161:829–830, 1986. 185. Haaga JR, Bick RJ, Zollinger RM, Jr: CT-guided percutaneous catheter cecostomy. Gastrointest Radiol 12:166–168, 1987. 186. Haaga JR, Exner AA, Wang Y, et al: Combined tumor therapy by using radiofrequency ablation and 5-FU-laden polymer implants: Evaluation in rats and rabbits. Radiology 237:911–918, 2005. 187. Haaga JR, George C, Weinstein AJ, et al: New interventional techniques in the diagnosis and management of inflammatory disease within the abdomen. Radiol Clin North Am 17:485–513, 1979. 188. Haaga JR, Kori SH, Eastwood DW, et al: Improved technique for CT-guided celiac ganglia block. AJR Am J Roentgenol 142:1201–1204, 1984. 189. Haaga JR, LiPuma JP, Bryan PJ, et al: Clinical comparison of small-and large-caliber cutting needles for biopsy. Radiology 146:665–667, 1983. 190. Haaga JR, Miraldi F, MacIntyre W, et al: The effect of MAS variation upon computed tomography image quality as evaluated by in vivo and in vitro studies. Radiology 138:449–454, 1981. 191. Haaga JR, Nakamoto D: Computed tomography-guided drainage of intra-abdominal infections. Curr Infect Dis Rep 6:105–114, 2004. 192. Haaga JR, Nakamoto D, Stellato T, et al: Intracavitary urokinase for enhancement of percutaneous abscess drainage: Phase II trial. AJR Am J Roentgenol 174:1681–1685, 2000. 193. Haaga JR, Ovens DB, Kellermeyer RW, et al: CT guided interstitial therapy of pancreatic carcinoma. J Comput Assist Tomogr 11:1077–1078, 1987. 194. Haaga P, Piroth W, Staatz G, et al: CT-guided percutaneous biopsies for the classification of focal liver lesions: A comparison between 14 G and 18 G puncture biopsy needles. Rofo 171:44–48, 1999. 195. Haaga JR, Reich NE, Havrilla TR, et al: Interventional CT scanning. Radiol Clin North Am 15:449–456, 1977. 196. Haaga JR, Reich NE, Havrilla TR, et al: Computed tomography-guided biopsy, I: Overview. J Comput Tomogr 2:25–30, 1978. 197. Haaga JR, Vanek J: Computed tomographic guided liver biopsy using the Menghini needle. Radiology 133:405–408, 1979. 198. Haaga JR, Weinstein AJ: CT-guided percutaneous aspiration and drainage of abscesses. AJR Am J Roentgenol 135:1187–1194, 1980. 199. Haaga JR, Weinstein AR: Percutaneous catheter drainage of abdominal abscesses. Letter to the. N Engl J Med 306:106–108, 1982. 200. Reference deleted in proofs. 201. Hagspiel KD, Kandarpa K, Silverman SG: Interactive MR-guided percutaneous nephrostomy. J Magn Reson Imaging 8:1319–1322, 1998. 202. Hamilton BD, Winfield HN: Laparoscopic marsupialization of pelvic lymphoceles. Tech Urol 2:220–224, 1996. 203. Hanna RM, Dahniya MH, Badr SS, et al: Percutaneous catheter drainage in drug-resistant amoebic liver abscess. Trop Med Int Health 5:578–581, 2000. 204. Haramati LB, Austin JH: Complications after CT-guided needle biopsy through aerated versus nonaerated lung. Radiology 181:778, 1991. 205. Harisinghani MG, Maher MM, Hahn PF, et al: Predictive value of benign percutaneous adrenal biopsies in oncology patients. Clin Radiol 57:898–901, 2002.

2577

206. Hatjidakis AA, Karampekios S, Prassopoulos P, et al: Maturation of the tract after percutaneous cholecystostomy with regard to the access route. Cardiovasc Intervent Radiol 21:36–40, 1998. 207. Hau T, Ahrenholz DH, Simmons RL: Secondary bacterial peritonitis: The biologic basis of treatment. Curr Probl Surg 16:1–65, 1979. 208. Hau T, Hoffman R, Simmons RL: Mechanisms of the adjuvant effect of hemoglobin in experimental peritonitis, I: In vivo inhibition of peritoneal leukocytosis. Surgery 83:223–229, 1978. 209. Hau T, Nishikawa RA, Phuangsab A: The effect of bacterial trapping by fibrin on the efficacy of systemic antibiotics in experimental peritonitis. Surg Gynecol Obstet 157:252–256, 1983. 210. Reference deleted in proofs. 211. Head HW, Dodd GD, III, Dalrymple NC, et al: Percutaneous radiofrequency ablation of hepatic tumors against the diaphragm: Frequency of diaphragmatic injury. Radiology 243:877–884, 2007. 212. Heaston DK: Narrow-gauge needle aspiration of solid adrenal masses. AJR Am J Roentgenol 141:169, 1983. 213. Heider R, Meyer AA, Galanko JA, et al: Percutaneous drainage of pancreatic pseudocysts is associated with a higher failure rate than surgical treatment in unselected patients. Ann Surg 229:781–787, discussion 787–789, 1999. 214. Heilbrun ME, Zagoria RJ, Garvin AJ, et al: CT-guided biopsy for the diagnosis of renal tumors before treatment with percutaneous ablation. AJR Am J Roentgenol 188:1500–1505, 2007. 215. Heilo A, Stenwig AE: Liver hemangioma: US-guided 18-gauge core-needle biopsy. Radiology 204:719–722, 1997. 216. Henn AR, Levine EA, McNulty W, et al: Percutaneous radiofrequency ablation of hepatic metastases for symptomatic relief of neuroendocrine syndromes. AJR Am J Roentgenol 181:1005–1010, 2003. 217. Herman SJ, Weisbrod GL: Usefulness of the blood patch technique   after transthoracic needle aspiration biopsy. Radiology 176:395–397, 1990. 218. Hertznu J, Peiser H, Zirkin A: Massive bleeding after fine needle aspiration of liver angiosarcoma. Abdominal Imaging 15:43–46, 1990. 219. Hesselmann V, Zahringer M, Krug B, et al: Computed-tomographyguided percutaneous core needle biopsies of suspected malignant lymphomas: Impact of biopsy, lesion, and patient parameters on diagnostic yield. Acta Radiol 45:641–645, 2004. 220. Hinshaw JL, Lubner MG, Lee FT: Percutaneous tumor ablation tools: Microwave, radiofrequency, or cryoablation–What should you use and why? Radiographics 35:1344–1362, 2014. 221. Ho CS, Taylor B: Percutaneous transgastric drainage for pancreatic pseudocyst. AJR Am J Roentgenol 143:623–625, 1984. 222. Ho CL, Yu SC, Yeung DW: 11C-acetate PET imaging in hepatocellular carcinoma and other liver masses. J Nucl Med 44:213–221, 2003. 223. Hochbergs P, Forsberg L, Hederstrom E, et al: Diagnosis and percutaneous treatment of pyogenic hepatic abscesses. Acta Radiol 31:351–353, 1990. 224. Holm HH: Editorial: The future of ultrasonic diagnosis. Ugeskr Laeger 136:44, 1973. 225. Holm HH, Pedersen JF, Kristensen JK, et al: Ultrasonically guided percutaneous puncture. Radiol Clin North Am 13:493–503, 1975. 226. Holm HH, Rasmussen SN, Kristensen JR: Ultrasonically guided percutaneous puncture technique. J Clin Ultrasound 1:27–31, 1973. 227. Hopper KD, Abendroth CS, Strutz KW, et al: Fine-needle aspiration biopsy for cytopathologic analysis: Utility of syringe handles, automated guns, and the nonsuction method. Radiology 185:819–824, 1992. 228. Hopper KD, Abendroth CS, Sturtz KW, et al: Automated biopsy devices: A blinded evaluation. Radiology 187:653–660, 1993. 229. Hopper KD, Baird DE, Reddy VV, et al: Efficacy of automated biopsy guns versus conventional biopsy needles in the pygmy pig. Radiology 176:671–676, 1990. 230. Hori T, Nagata K, Kasuike S, et al: Risk factors for the local recurrence of hepatocellular carcinoma after a single session of percutaneous radiofrequency ablation. J Gastroenterol 38:977–981, 2003. 231. Horkan C, Ahmed M, Liu Z, et al: Radiofrequency ablation: Effect of pharmacologic modulation of hepatic and renal blood flow on

2578

232.

233. 234.

235.

236.

237.

238.

239.

240.

241. 242.

243.

244. 245.

246.

247.

248.

249.

250.

251.

252.

PART III  Image-Guided Procedures

coagulation diameter in a VX2 tumor model. J Vasc Interv Radiol 15:269–274, 2004. Horwhat JD, Paulson EK, McGrath K, et al: A randomized comparison of EUS-guided FNA versus CT or US-guided FNA for the evaluation of pancreatic mass lesions. Gastrointest Endosc 63:966–975, 2006. Hueftle MG, Haaga JR: Effect of suction on biopsy sample size. AJR Am J Roentgenol 147:1014–1016, 1986. Hunter S, Samir A, Eisner B, et al: Diagnosis of renal lymphoma by percutaneous image guided biopsy: Experience with 11 cases. J Urol 176:1952–1956, discussion 1956, 2006. Ikeda M, Okada S, Ueno H, et al: Radiofrequency ablation and percutaneous ethanol injection in patients with small hepatocellular carcinoma: A comparative study. Jpn J Clin Oncol 31:322–326, 2001. Imdahl A, Nitzsche E, Krautmann F, et al: Evaluation of positron emission tomography with 2-[18F]fluoro-2-deoxy-D-glucose for the differentiation of chronic pancreatitis and pancreatic cancer. Br J Surg 86:194–199, 1999. Iselin CE, Mayer P, Hauser H, et al: Computed tomography and fine-needle aspiration cytology for preoperative evaluation of cystic tumours of the pancreas. Br J Surg 80:1166–1169, 1993. Isler RJ, Ferrucci JT, Jr, Wittenberg J, et al: Tissue core biopsy of abdominal tumors with a 22 gauge cutting needle. AJR Am J Roentgenol 136:725–728, 1981. Jacobsen GK, Gammelgaard J, Fuglo M: Coarse needle biopsy versus fine needle aspiration biopsy in the diagnosis of focal lesions of the liver: Ultrasonically guided needle biopsy in suspected hepatic malignancy. Acta Cytol 27:152–156, 1983. Jaskolka JD, Asch MR, Kachura JR, et al: Needle tract seeding after radiofrequency ablation of hepatic tumors. J Vasc Interv Radiol 16:485–491, 2005. Jereb M, Us-Krasovec M: Transthoracic needle biopsy of mediastinal and hilar lesions. Cancer J 40:1354–1357, 1977. Jhala D, Eloubeidi M, Chhieng DC, et al: Fine needle aspiration biopsy of the islet cell tumor of pancreas: A comparison between computerized axial tomography and endoscopic ultrasound-guided fine needle aspiration biopsy. Ann Diagn Pathol 6:106–112, 2002. Jin GY, Lee JM, Lee YC, et al: Primary and secondary lung malignancies treated with percutaneous radiofrequency ablation: Evaluation with follow-up helical CT. AJR Am J Roentgenol 183:1013–1020, 2004. Johnson RD, Mueller PR, Ferrucci JT, Jr, et al: Percutaneous drainage of pyogenic liver abscesses. AJR Am J Roentgenol 144:463–467, 1985. Johnson DE, Pendurthi TK, Balshem AM, et al: Implications of fine-needle aspiration in patients with resectable pancreatic cancer. Am Surg 63:675–679, discussion 679–680, 1997. Kamiyoshihara M, Sakata K, Ishikawa S, et al: Cerebral arterial air embolism following CT-guided lung needle marking: Report of a case.   J Cardiovasc Surg (Torino) 42:699–700, 2001. Kane NM, Korobkin M, Francis IR, et al: Percutaneous biopsy of left adrenal masses: Prevalence of pancreatitis after anterior approach. AJR Am J Roentgenol 157:777–780, 1991. Kang S, Luo R, Liao W, et al: Single group study to evaluate the feasibility and complications of radiofrequency ablation and usefulness of post treatment positron emission tomography in lung tumours. World J Surg Oncol 2:30, 2004. Kappis M: Sensibilitat und Lokale anestheie im chirurgischen Geblet der bauchkokle mit besonderer berucksichtigung der Splanchniscus anesthesia. Beitr Klin Chir 151:161–175, 1919. Karampekio S, Hatjidakis AA, Jannis D, et al: Artificial paravertebral widening for percutaneous CT-guided adrenal biopsy. J Comput Assist Tomogr 22:308–310, 1998. Kariya S, Tanigawa N, Kojima H, et al: Radiofrequency ablation combined with CO2 injection for treatment of retroperitoneal tumor: Protecting surrounding organs against thermal injury. AJR Am J Roentgenol 185:890–893, 2005. Kasugai H, Yamamoto R, Tatauta M, et al: Value of heparinized fine-needle aspiration biopsy in liver malignancy. AJR Am J Roentgenol 144:243–244, 1985.

253. Kazerooni EA, Lim FT, Mikhail A, et al: Risk of pneumothorax in CT-guided transthoracic needle aspiration biopsy of the lung. Radiology 198:371–375, 1996. 254. Keogan MT, Freed KS, Paulson EK, et al: Imaging-guided percutaneous biopsy of focal splenic lesions: Update on safety and effectiveness. AJR Am J Roentgenol 172:933–937, 1999. 255. Kerlan RK, Jr, LaBerge JM, Wilson MW, et al: Bleomycin sclerosis of pelvic lymphoceles. J Vasc Interv Radiol 8:885–887, 1997. 256. Khan MA, Combs CS, Brunt EM, et al: Positron emission tomography scanning in the evaluation of hepatocellular carcinoma. J Hepatol 32:792–797, 2000. 257. Khouri NF, Stitik FP, Erozan YS, et al: Transthoracic needle aspiration biopsy of benign and malignant lung lesions. AJR Am J Roentgenol 144:281–288, 1985. 258. Khuroo MS, Wani NA, Javid G, et al: Percutaneous drainage compared with surgery for hepatic hydatid cysts. N Engl J Med 337:881–887, 1997. 259. Kim S: Needle-tract implantation in hepatocellular carcinoma: Frequency and CT findings after biopsy with a 19.5-gauge automated biopsy gun. Abdom Imaging 25:246–250, 2000. 260. Reference deleted in proofs 261. Kim YK, Kim CS, Chung GH, et al: Radiofrequency ablation of hepatocellular carcinoma in patients with decompensated cirrhosis: Evaluation of therapeutic efficacy and safety. AJR Am J Roentgenol 186:S261–S268, 2006. 262. Kim SH, Lim HK, Choi D, et al: Percutaneous radiofrequency ablation of hepatocellular carcinoma: Effect of histologic grade on therapeutic results. AJR Am J Roentgenol 186:S327–S333, 2006. 263. Kinoshita F, Kato T, Sugiura K, et al: CT-guided transthoracic needle biopsy using a puncture site-down positioning technique. AJR Am J Roentgenol 187:926–932, 2006. 264. Kitamoto M, Imagawa M, Yamada H, et al: Radiofrequency ablation in the treatment of small hepatocellular carcinomas: Comparison of the radiofrequency effect with and without chemoembolization. AJR Am J Roentgenol 181:997–1003, 2003. 265. Klein JS, Stewart EA: Transthoracic needle biopsy with a coaxially placed 20-gauge automated cutting needle: Results in 122 patients. Radiology 198:715–720, 1996. 266. Knelson M, Haaga J, Lazarus H, et al: Computed tomography-guided retroperitoneal biopsies. J Clin Oncol 7:1169–1173, 1989. 267. Ko JP, Shepard JO, Drucker EA, et al: Factors influencing pneumothorax rate at lung biopsy: Are dwell time and angle of pleural puncture contributing factors? Radiology 218:491–496, 2001. 268. Kodama F, Ogawa T, Hashimoto M, et al: Fatal air embolism as a complication of CT-guided needle biopsy of the lung. J Comput Assist Tomogr 23:949–951, 1999. 269. Komorizono Y, Okentani M, Sako K, et al: Risk factors for local recurrence of small hepatocellular carcinoma tumors after a single session, single application of percutaneous radiofrequency ablation. Cancer 97:1253–1262, 2003. 270. Kornprat P, Jarnagin WR, Gonen M, et al: Outcome after hepatectomy for multiple (four or more) colorectal metastases in the era of effective chemotherapy. Ann Surg Oncol 14:1151–1160, 2007. 271. Korobkin M, Brodeur FJ, Francis IR, et al: CT time-attenuation washout curves of adrenal adenomas and nonadenomas. AJR Am J Roentgenol 170:747–752, 1998. 272. Kreula J, Virkkunen P, Bondestam S: Effect of suction on specimen size in fine-needle aspiration biopsy. Invest Radiol 25:1175–1181, 1990. 273. Kulkarni S, Kulkarni A, Roy D, et al: Percutaneous computed tomography-guided core biopsy for the diagnosis of mediastinal masses. Ann Thorac Med 3:13–17, 2008. 274. Kumar R, Xiu Y, Yu JQ, et al: 18F-FDG PET in evaluation of   adrenal lesions in patients with lung cancer. J Nucl Med 45:2058–2062, 2004. 275. Kushihashi T, Munechika H, Otsuki N, et al: CT-guided percutaneous lung biopsy using a cutting needle and an automated biopsy gun: Comparison with lung biopsy using a manual aspiration needle. Nippon Igaku Hoshasen Gakkai Zasshi 52:191–198, 1992.

CHAPTER 67  Image-Guided Aspirations and Biopsies 276. Kuvshinoff BW, Ota DM: Radiofrequency ablation of liver tumors: Influence of technique and tumor size. Surgery 132:605–611, discussion 611–612, 2002. 277. Laeseke PF, Sampson LA, Brace CL, et al: Unintended thermal injuries from radiofrequency ablation: Protection with 5% dextrose in water. AJR Am J Roentgenol 186:S249–S254, 2006. 278. Lahorra JM, Haaga JR, Stellato T, et al: Safety of intracavitary urokinase with percutaneous abscess drainage. AJR Am J Roentgenol 160:171–174, 1993. 279. Lalli AF, McCormack LJ, Zelch M, et al: Aspiration biopsies of chest lesions. Radiology 127:35–40, 1978. 280. Lambiase RE, Cronan JJ, Dorfman GS, et al: Postoperative abscesses with enteric communication: Percutaneous treatment. Radiology 171:497–500, 1989. 281. Lambiase RE, Deyoe L, Cronan JJ, et al: Percutaneous drainage of 335 consecutive abscesses: Results of primary drainage with 1-year follow-up. Radiology 184:167–179, 1992. 282. Reference deleted in proofs. 283. Lang EK: Abdominal abscess drainage under radiologic guidance: Causes of failure. Radiology 174:329–336, 1986. 284. Lang EK: Renal, perirenal, and pararenal abscesses: Percutaneous drainage. Radiology 174:109–113, 1990. 285. Lang EK, Ghavami R, Schreiner VC, et al: Autologous blood clot seal to prevent pneumothorax at CT-guided lung biopsy. Radiology 216:93–96, 2000. 286. Langen HJ, Jochims M, Gunther RW: Artificial displacement of kidneys, spleen, and colon by injection of physiologic saline and CO2 as an aid to percutaneous procedures: Experimental results. J Vasc Interv Radiol 6:411–416, 1995. 287. Lattin G, Jr, O’Brien W, Sr, McCarry B, et al: Massive systemic air embolism treated with hyperbaric oxygen therapy following CT-guided transthoracic needle biopsy of a pulmonary nodule. J Vasc Interv Radiol 17:1355–1358, 2006. 288. Lauby VW, Burnett WE, Rosemond GP, et al: Value and risk of biopsy of pulmonary lesions by needle aspiration: Twenty-one years’ experience. J Thorac Cardiovasc Surg 1:159–172, 1965. 289. Laurent F, Latrabe V, Vergier B, et al: Percutaneous CT-guided biopsy of the lung: Comparison between aspiration and automated cutting needles using a coaxial technique. Cardiovasc Intervent Radiol 23:266–272, 2000. 290. Lee JM, Jin GY, Goldberg SN, et al: Percutaneous radiofrequency ablation for inoperable non-small cell lung cancer and metastases: Preliminary report. Radiology 230:125–134, 2004. 291. Lee MJ, Mueller PR, van Sonnenberg E, et al: CT-guided celiac ganglion block with alcohol. AJR Am J Roentgenol 161:633–636, 1993. 292. Lencioni RA, Allgaier HP, Cioni D, et al: Small hepatocellular carcinoma in cirrhosis: Randomized comparison of radio-frequency thermal ablation versus percutaneous ethanol injection. Radiology 228:235–240, 2003. 293. Levin DP, Bret PM: Percutaneous fine-needle aspiration biopsy of the pancreas resulting in death. Gastrointest Radiol 16:67–69, 1991. 294. Levison MA, Zeigler D: Correlation of APACHE II score, drainage technique and outcome in postoperative intra-abdominal abscess. Surg Gynecol Obstet 172:89–94, 1991. 295. Lewandrowski KB, Southern JF, Pins MR, et al: Cyst fluid analysis in the differential diagnosis of pancreatic cysts: A comparison of pseudocysts, serous cystadenomas, mucinous cystic neoplasms, and mucinous cystadenocarcinoma. Ann Surg 217:41–47, 1993. 296. Lewin JS, Connell CF, Duerk JL, et al: Interactive MRI-guided radiofrequency interstitial thermal ablation of abdominal tumors: Clinical trial for evaluation of safety and feasibility. J Magn Reson Imaging 8:40–47, 1998. 297. Lewin JS, Duerk JL, Jain VR, et al: Needle localization in MR-guided biopsy and aspiration: Effects of field strength, sequence design, and magnetic field orientation. AJR Am J Roentgenol 166:1337–1345, 1996. 298. Lewin JS, Petersilge CA, Hatem SF, et al: Interactive MR imagingguided biopsy and aspiration with a modified clinical C-arm system. AJR Am J Roentgenol 170:1593–1601, 1998.

2579

299. Li H, Boiselle PM, Shepard JO, et al: Diagnostic accuracy and safety of CT-guided percutaneous needle aspiration biopsy of the lung: Comparison of small and large pulmonary nodules. AJR Am J Roentgenol 167:105–109, 1996. 300. Li L, Wu QL, Liu LZ, et al: Value of CT-guided core-needle biopsy in diagnosis and classification of malignant lymphomas using automated biopsy gun. World J Gastroenterol 11:4843–4847, 2005. 301. Lieberman S, Libson E, Maly B, et al: Imaging-guided percutaneous splenic biopsy using a 20- or 22-gauge cutting-edge core biopsy needle for the diagnosis of malignant lymphoma. AJR Am J Roentgenol 181:1025–1027, 2003. 302. Lieberman RP, Waldman SD: Celiac plexus neurolysis with the modified transaortic approach. Radiology 175:274–276, 1990. 303. Lim JW, Ko YT, Lee DH, et al: Treatment of prostatic abscess: Value of transrectal ultrasonographically guided needle aspiration. J Ultrasound Med 19:609–617, 2000. 304. Linder JD, Geenen JE, Catalano MF: Cyst fluid analysis obtained by EUS-guided FNA in the evaluation of discrete cystic neoplasms of the pancreas: A prospective single-center experience. Gastrointest Endosc 64:697–702, 2006. 305. Little AF, Ferris JV, Dodd GD, 3rd, et al: Image-guided percutaneous hepatic biopsy: Effect of ascites on the complication rate. Radiology 199:79–83, 1996. 306. Liu KY, Shyr YM, Su CH, et al: Splenic abscess—A changing trend in treatment. S Afr J Surg 38:55–57, 2000. 307. Livraghi T: Guidelines for treatment of liver cancer. Eur J Ultrasound 13:167–176, 2001. 308. Livraghi T, Bolondi L, Lazzaroni S, et al: Percutaneous ethanol injection in the treatment of hepatocellular carcinoma in cirrhosis: A study on 207 patients. Cancer 69:925–929, 1992. 309. Livraghi T, Goldberg SN, Lazzaroni S, et al: Small hepatocellular carcinoma: Treatment with radio-frequency ablation versus ethanol injection. Radiology 210:655–661, 1999. 310. Livraghi T, Goldberg SN, Monti F, et al: Saline-enhanced radiofrequency tissue ablation in the treatment of liver metastases. Radiology 202:205–210, 1997. 311. Livraghi T, Goldberg SN, Solbiati L, et al: Percutaneous radio-frequency ablation of liver metastases from breast cancer: Initial experience in 24 patients. Radiology 220:145–149, 2001. 312. Livraghi T, Lazzaroni S, Meloni F: Radiofrequency thermal ablation of hepatocellular carcinoma. Eur J Ultrasound 13:159–166, 2001. 313. Livraghi T, Lazzaroni S, Meloni F, et al: Risk of tumour seeding after percutaneous radiofrequency ablation for hepatocellular carcinoma. Br J Surg 92:856–858, 2005. 314. Livraghi T, Solbiati L, Meloni MF, et al: Treatment of focal liver tumors with percutaneous radio-frequency ablation: Complications encountered in a multicenter study. Radiology 226:441–451, 2003. 315. Llovet JM, Vilana R, Bru C, et al: Increased risk of tumor seeding after percutaneous radiofrequency ablation for single hepatocellular carcinoma. Hepatology 33:1124–1129, 2001. 316. Lopez Hanninen E, Vogl TJ, Ricke J, et al: CT-guided percutaneous core biopsies of pulmonary lesions: Diagnostic accuracy, complications and therapeutic impact. Acta Radiol 42:151–155, 2001. 317. Lumachi F, Borsato S, Brandes AA, et al: Fine-needle aspiration cytology of adrenal masses in noncancer patients: Clinicoradiologic and histologic correlations in functioning and nonfunctioning tumors. Cancer 93:323–329, 2001. 318. Lundquist A: Fine-needle aspiration biopsy for cytodiagnosis of malignant tumour in the liver. Acta Med Scand 188:465–470, 1970. 319. Lundquist A: Fine-needle aspiration biopsy of the liver: Applications in clinical diagnosis and investigation. Acta Med Scand Suppl 520:1–28, 1971. 320. Lynch CR, Jones RG, Hilden K, et al: Percutaneous endoscopic cecostomy in adults: A case series. Gastrointest Endosc 64:279–282, 2006. 321. Lytras D, Connor S, Bosonnet L, et al: Positron emission tomography does not add to computed tomography for the diagnosis and staging of pancreatic cancer. Dig Surg 22:55–61, discussion 62, 2005.

2580

PART III  Image-Guided Procedures

322. Machi J, Bueno RS, Wong LL: Long-term follow-up outcome of patients undergoing radiofrequency ablation for unresectable hepatocellular carcinoma. World J Surg 29:1364–1373, 2005. 323. Magnusson A, Akerfeldt D: CT-guided core biopsy using a new guidance device. Acta Radiol 32:83–85, 1991. 324. Mahalingam M, Ugen KE, Kao KJ, et al: Functional role of platelets in experimental metastasis studied with cloned murine fibrosarcoma cell variants. Cancer Res 48:1460–1464, 1988. 325. Mandrekar SR, Pinto RG, Patil R, et al: Solid and papillary epithelial neoplasm of the pancreas: Report of a case with diagnosis by fine needle aspiration cytology. Acta Cytol 41:1823–1827, 1997. 326. Marco V: Surgical resection versus percutaneous radiofrequency ablation in the treatment of hepatocellular carcinoma on cirrhotic liver. Ann Surg 240:102–107, 2004. 327. Marra V, Debernardi F, Frigerio A, et al: Neurolytic block of the celiac plexus and splanchnic nerves with computed tomography: The experience in 150 cases and an optimization of the technic. Radiol Med (Torino) 98:183–188, 1999. 328. Martin H: The advantages and limitations of aspiration biopsy. AJR Am J Roentgenol 35:54, 1936. 329. Martino CR, Haaga JR, Bryan PJ, et al: CT-guided liver biopsies:   Eight years’ experience: Work in progress. Radiology 152:755–757, 1984. 330. Maskell NA, Gleeson FV, Davies RJ: Standard pleural biopsy versus CT-guided cutting-needle biopsy for diagnosis of malignant disease in pleural effusions: A randomised controlled trial. Lancet 361:1326–1330, 2003. 331. Matsuguma H, Nakahara R, Kondo T, et al: Risk of pleural recurrence after needle biopsy in patients with resected early stage lung cancer. Ann Thorac Surg 80:2026–2031, 2005. 332. Maturen KE: Lack of tumor seeding of hepatocellular carcinoma after percutaneous needle biopsy using coaxial cutting needle technique. AJR Am J Roentgenol 187:1184–1187, 2006. 333. Maturen KE, Nghiem HV, Caoili EM, et al: Renal mass core biopsy: Accuracy and impact on clinical management. AJR Am J Roentgenol 188:563–570, 2007. 334. Maxeiner H: Lethal hemoptysis caused by biopsy injury of an abnormal bronchial artery. Chest 119:1612–1615, 2001. 335. Mayoral W, Lewis JH: Percutaneous liver biopsy: What is the current approach? Results of a questionnaire survey. Dig Dis Sci 46:118–127, 2001. 336. Mayo-Smith WW, Dupuy DE, Parikh PM, et al: Imaging-guided percutaneous radiofrequency ablation of solid renal masses: Techniques and outcomes of 38 treatment sessions in 32 consecutive patients. AJR Am J Roentgenol 180:1503–1508, 2003. 337. Mazzaferro V, Battiston C, Perrone S, et al: Radiofrequency ablation of small hepatocellular carcinoma in cirrhotic patients awaiting liver transplantation: A prospective study. Ann Surg 240:900–909, 2004. 338. McCartney R, Tait D, Stilson M, et al: A technique for the prevention of pneumothorax in pulmonary aspiration biopsy. Am J Roentgenol Radium Ther Nucl Med 120:872–875, 1974. 339. McCorkell SJ, Niles NL: Fine-needle aspiration of catecholamineproducing adrenal masses: A possibly fatal mistake. AJR Am J Roentgenol 145:113–114, 1985. 340. McFadzien A, Wong CK: Solitary pyogenic abscess treated by closed aspiration and antibiotics: Fourteen consecutive cases with recovery.   Br J Surg 41:141–152, 1953. 341. McGahan JP, Ro KM, Evans CP, et al: Efficacy of transhepatic radiofrequency ablation of renal cell carcinoma. AJR Am J Roentgenol 186:S311–S315, 2006. 342. McLoughlin MJ, Ho CS, Langer B, et al: Fine needle aspiration biopsy of malignant lesions in and around the pancreas. Cancer 41:2413–2419, 1978. 343. Menghini G: One-second biopsy of the liver—Problems of its clinical application. N Engl J Med 283:582–585, 1970. 344. Merkle EM, Haaga JR, Duerk JL, et al: MR imaging-guided radiofrequency thermal ablation in the pancreas in a porcine model with a modified clinical C-arm system. Radiology 213:461–467, 1999.

345. Merkle EM, Hashim M, Wendt M, et al: MR-guided percutaneous nephrostomy of the nondilated upper urinary tract in a porcine model. AJR Am J Roentgenol 172:1221–1225, 1999. 346. Merkle EM, Shonk JR, Duerk JL, et al: MR-guided RF thermal ablation of the kidney in a porcine model. AJR Am J Roentgenol 173:645–651, 1999. 347. Mesurolle B, Ariche-Cohen M, Tardivon A, et al: Retrospective analysis of 44 adrenal puncture biopsies under x-ray computed tomographic guidance. J Radiol 77:17–21, 1996. 348. Metser U, Miller E, Lerman H, et al: 18F-FDG PET/CT in the evaluation of adrenal masses. J Nucl Med 47:32–37, 2006. 349. Micames C, Jowell PS, White R, et al: Lower frequency of peritoneal carcinomatosis in patients with pancreatic cancer diagnosed by EUS-guided FNA vs. percutaneous FNA. Gastrointest Endosc 58:690– 695, 2003. 350. Miller FJ, Ahola DT, Bretzman PA, et al: Percutaneous management of hepatic abscess: A perspective by interventional radiologists. J Vasc Interv Radiol 8:241–247, 1997. 351. Reference deleted in proofs. 352. Miller JA, Pramanik BK, Lavenhar MA: Predicting the rates of success and complications of computed tomography-guided percutaneous core-needle biopsies of the thorax from the findings of the preprocedure chest computed tomography scan. J Thorac Imaging 13:7–13, 1998. 353. Minami Y, Masatoshi K, Kawasaki Y: Percutaneous radiofrequency ablation guided by contrast-enhanced harmonic sonography with artificial pleural effusion for hepatocellular carcinoma in the hepatic dome. AJR Am J Roentgenol 182:1224–1226, 2004. 354. Mitty HA, Efremidis SC, Yeh HC: Impact of fine-needle biopsy on management of patients with carcinoma of the pancreas. AJR Am J Roentgenol 137:1119–1121, 1981. 355. Mody MK, Kazerooni EA, Korobkin M: Percutaneous CT-guided biopsy of adrenal masses: Immediate and delayed complications. J Comput Assist Tomogr 19:434–439, 1995. 356. Montalvo BM, Yrizarry JM, Casillas VJ, et al: Percutaneous sclerotherapy of lymphoceles related to renal transplantation. J Vasc Interv Radiol 7:117–123, 1996. 357. Montgomery RS, Rahal A, Dodd GD, et al: Radiofrequency ablation of hepatic tumors: Variability of lesion size using a single ablation device. AJR Am J Roentgenol 182:657–661, 2004. 358. Moore EH, LeBlanc J, Montesi SA, et al: Effect of patient positioning after needle aspiration lung biopsy. Radiology 181:385–387, 1991. 359. Morgan DE, Baron TH, Smith JK, et al: Pancreatic fluid collections prior to intervention: Evaluation with MR imaging compared with CT and US. Radiology 203:773–778, 1997. 360. Morrison PR, van Sonnenberg E, Shankar S, et al: Radiofrequency ablation of thoracic lesions, Part 1: Experiments in the normal porcine thorax. AJR Am J Roentgenol 184:375–380, 2005. 361. Mostbeck GH, Wittich GR, Derfler K, et al: Optimal needle size for renal biopsy: In vitro and in vivo evaluation. Radiology 173:819–822, 1989. 362. Mueller PR, Dawson SL, Ferrucci JT, Jr, et al: Hepatic echinococcal cyst: Successful percutaneous drainage. Radiology 155:627–628, 1985. 363. Mueller PR, Ferrucci JT, Jr, Butch RJ, et al: Inadvertent percutaneous catheter gastroenterostomy during abscess drainage: Significance and management. AJR Am J Roentgenol 145:387–391, 1985. 364. Mueller PR, Ferrucci JT, Jr, Simeone JF, et al: Lesser sac abscesses and fluid collections: Drainage by transhepatic approach. Radiology 155:615–618, 1985. 365. Mueller PR, Miketic LM, Simeone JF, et al: Severe acute pancreatitis after percutaneous biopsy of the pancreas. AJR Am J Roentgenol 151:493–494, 1988. 366. Mueller PR, Simeone JF, Butch RJ, et al: Percutaneous drainage of subphrenic abscess: A review of 62 patients. AJR Am J Roentgenol 147:1237–1240, 1986. 367. Reference deleted in proofs. 368. Mueller PR, White EM, Glass-Royal M, et al: Infected abdominal tumors: Percutaneous catheter drainage. Radiology 173:627–629, 1989.

CHAPTER 67  Image-Guided Aspirations and Biopsies 369. Mulier S, Mulier P, Ni Y, et al: Complications of radiofrequency coagulation of liver tumours. Br J Surg 89:1206–1222, 2002. 370. Murphy FB, Barefield KP, Steinberg HV, et al: CT- or sonographyguided biopsy of the liver in the presence of ascites: Frequency of complications. AJR Am J Roentgenol 151:485–486, 1988. 371. Nakamoto DA, Haaga JR, Bove P, et al: Use of fibrinolytic agents to coat wire implants to decrease infection: An animal model. Invest Radiol 30:341–344, 1995. 372. Nakamoto DA, Rosenfield ML, Haaga JR, et al: Young Investigator Award. In vivo treatment of infected prosthetic graft material with urokinase: An animal model. J Vasc Interv Radiol 5:549–552, 1994. 373. Nakazawa T, Kokubu S, Shibuya A, et al: Radiofrequency ablation of hepatocellular carcinoma: Correlation between local tumor progression after ablation and ablative margin. AJR Am J Roentgenol 188:480–488, 2007. 374. Nealon WH, Walser E: Main pancreatic ductal anatomy can direct choice of modality for treating pancreatic pseudocysts (surgery versus percutaneous drainage). Ann Surg 235:751–758, 2002. 375. Nealon WH, Walser E: Surgical management of complications associated with percutaneous and/or endoscopic management of pseudocyst of the pancreas. Ann Surg 241:948–957, discussion 957–960, 2005. 376. Nealon WH, Walser E: Surgical management of complications associated with percutaneous and/or endoscopic management of pseudocyst of the pancreas. Ann Surg 241:948–957, discussion 957–960, 2006. 377. Neff CC, van Sonnenberg E, Casola G, et al: Diverticular abscesses: Percutaneous drainage. Radiology 163:15–18, 1987. 378. Norberg M, Busch C, Stavinoha J, et al: Transrectal ultrasound-guided core biopsies of the prostate: A comparison between the standard 1.2-mm needle and three thinner needles. Acta Radiol 35:463–467, 1994. 379. Norbert A: 18F-FDG PET after radiofrequency ablation: Is timing everything? J Nucl Med 47:1235–1237, 2006. 380. Nordenstrom B: A new technique for transthoracic biopsy of lung changes. Br J Radiol 38:550–553, 1965. 381. Norenberg R: Percutaneous needle biopsy of the lung: Report of two fatal complications. Chest 66:216–218, 1974. 382. Nour SG, Lewin JS, Gutman M, et al: Percutaneous MR imagingguided radiofrequency interstitial thermal ablation of tongue base in porcine models: Implications for obstructive sleep apnea syndrome. Radiology 230:359–368, 2004. 383. Nunez D, Jr, Yrizarry JM, Russell E, et al: Transgastric drainage of pancreatic fluid collections. AJR Am J Roentgenol 145:815–818, 1985. 384. Oei T, van Sonnenberg E, Shankar S, et al: Radiofrequency ablation of liver tumors: A new cause of benign portal venous gas. Radiology 237:709–717, 2005. 385. Ohashi S, Endoh H, Honda T, et al: Cerebral air embolism   complicating percutaneous thin-needle biopsy of the lung: Complete neurological recovery after hyperbaric oxygen therapy. J Anesth 15:233–236, 2001. 386. Ohnishi K: Comparison of percutaneous acetic acid injection and percutaneous ethanol injection for small hepatocellular carcinoma. Hepatogastroenterology 45(Suppl 3):1254–1258, 1998. 387. Okuma T, Matsuoka T, Yamamoto A, et al: Factors contributing to cavitation after CT-guided percutaneous radiofrequency ablation for lung tumors. J Vasc Interv Radiol 18:399–404, 2007. 388. Okuma T, Okamura T, Matsuoka T, et al: Fluorine-18fluorodeoxyglucose positron emission tomography for assessment of patients with unresectable recurrent or metastatic lung cancers after CT-guided radiofrequency ablation: Preliminary results. Ann Nucl Med 20:115–121, 2006. 389. Ormeci N, Soykan I, Bektas A, et al: A new percutaneous approach for the treatment of hydatid cysts of the liver. Am J Gastroenterol 96:2225–2230, 2001. 390. Oscarson J, Stormby N, Sundgren R: Selective angiography in fine-needle aspiration cytodiagnosis of gastric and pancreatic tumours. Acta Radiol Diagn (Stockh) 12:737–750, 1972.

2581

391. Oshima F, Yamakado K, Akeboshi M, et al: Lung radiofrequency ablation with and without bronchial occlusion: Experimental study in porcine lungs. J Vasc Interv Radiol 15:1451–1456, 2004. 392. O’Toole D, Palazzo L, Hammel P, et al: Macrocystic pancreatic cystadenoma: The role of EUS and cyst fluid analysis in distinguishing mucinous and serous lesions. Gastrointest Endosc 59:823–829, 2004. 393. Pacella CM, Bizzarri G, Francica G, et al: Percutaneous laser ablation in the treatment of hepatocellular carcinoma with small tumors: Analysis of factors affecting the achievement of tumor necrosis. J Vasc Interv Radiol 16:1447–1457, 2005. 394. Pagani JJ: Biopsy of focal hepatic lesions: Comparison of 18 and 22 gauge needles. Radiology 147:673–675, 1983. 395. Pagani JJ: Normal adrenal glands in small cell lung carcinoma: CT-guided biopsy. AJR Am J Roentgenol 140:949–951, 1983. 396. Pagani JJ: Non-small cell lung carcinoma adrenal metastases: Computed tomography and percutaneous needle biopsy in their diagnosis. Cancer 53:1058–1060, 1984. 397. Pappa VI, Hussain HK, Reznek RH, et al: Role of image-guided core-needle biopsy in the management of patients with lymphoma.   J Clin Oncol 14:2427–2430, 1996. 398. Park BK, Kim CK, Kim B, et al: Comparison of delayed enhanced CT and 18F-FDG PET/CT in the evaluation of adrenal masses in oncology patients. J Comput Assist Tomogr 31:550–556, 2007. 399. Park JK, Kraus FC, Haaga JR: Fluid flow during percutaneous drainage procedures: An in vitro study of the effects of fluid viscosity, catheter size, and adjunctive urokinase. AJR Am J Roentgenol 160:165–169, 1993. 400. Paulson EK, Stephenson GR, Neal MC, et al: Use of fibrin sealant as a hemostatic agent after liver biopsy in swine. J Vasc Interv Radiol 11:905–911, 2000. 401. Pawlik TM, Abdalla EK, Ellis LM, et al: Debunking dogma: Surgery for four or more colorectal liver metastases is justified. J Gastrointest Surg 10:240–248, 2006. 402. Perlmutt LM, Braun SD, Newman GE, et al: Timing of chest film follow-up after transthoracic needle aspiration. AJR Am J Roentgenol 146:1049–1050, 1986. 403. Permpongkosol S, Nicol TL, Link RE, et al: Differences in ablation size in porcine kidney, liver, and lung after cryoablation using the same ablation protocol. AJR Am J Roentgenol 188:1028–1032, 2007. 404. Perrault J, McGill DB, Ott BJ, et al: Liver biopsy: Complications in 1000 inpatients and outpatients. Gastroenterology 74:103–106, 1978. 405. Peyser A, Applbaum Y, Khoury A, et al: Osteoid osteoma: CT-guided radiofrequency ablation using a water-cooled probe. Ann Surg Oncol 14:591–596, 2007. 406. Phillips VM, Hersh T, Erwin BC, et al: Percutaneous biopsy of pancreatic masses. J Clin Gastroenterol 7:506–510, 1985. 407. Plecha DM, Goodwin DW, Rowland DY, et al: Liver biopsy: Effects of biopsy needle caliber on bleeding and tissue recovery. Radiology 204:101–104, 1997. 408. Poon RT, Ng KK, Lam CM, et al: Radiofrequency ablation for subcapsular hepatocellular carcinoma. Ann Surg Oncol 11:281–289, 2004. 409. Poon RT, Ng KK, Lam CM, et al: Learning curve for radiofrequency ablation of liver tumors: Prospective analysis of initial 100 patients in a tertiary institution. Ann Surg 239:441–449, 2004. 410. Prasanna A: Unilateral celiac plexus block. J Pain Symptom Manage 11:154–157, 1996. 411. Preger L: Hepatic arteriovenous fistula after percutaneous liver biopsy. Am J Roentgenol Radium Ther Nucl Med 101:619–620, 1967. 412. Quinn SF, Sheley RC, Nelson HA, et al: The role of percutaneous needle biopsies in the original diagnosis of lymphoma: A prospective evaluation. J Vasc Interv Radiol 6:947–952, 1995. 413. Raman SS, Lu DS, Vodopich DJ, et al: Minimizing diaphragmatic injury during radio-frequency ablation: Efficacy of subphrenic peritoneal saline injection in a porcine model. Radiology 222:819–823, 2002. 414. Rashleigh-Belcher HJ, Russell RC, Lees WR: Cutaneous seeding of pancreatic carcinoma by fine-needle aspiration biopsy. Br J Radiol 59:182–183, 1986.

2582

PART III  Image-Guided Procedures

415. Rasmussen SN, Holm HH, Kristensen JK, et al: Ultrasonically-guided liver biopsy. BMJ 2:500–502, 1972. 416. Raut CP, Izzo F, Marra P, et al: Significant long-term survival after radiofrequency ablation of unresectable hepatocellular carcinoma in patients with cirrhosis. Ann Surg Oncol 12:616–628, 2005. 417. Ravinsky E, Morales C: Diagnosis of lymphoma by image-guided needle biopsies: Fine needle aspiration biopsy, core biopsy or both? Acta Cytol 49:51–57, 2005. 418. Rimondi E, Bianchi G, Malaguti MC, et al: Radiofrequency thermoablation of primary non-spinal osteoid osteoma: Optimization of the procedure. Eur Radiol 15:1393–1399, 2005. 419. Romanelli DF, Beckmann CF, Heiss FW: Celiac plexus block: Efficacy and safety of the anterior approach. AJR Am J Roentgenol 160:497–500, 1993. 420. Rose SC, Thistlethwaite PA, Sewell PE, et al: Lung cancer and radiofrequency ablation. J Vasc Interv Radiol 17:927–951, 2006. 421. Rosenberger A, Adler O: Fine needle biopsy in the diagnosis of mediastinal lesions. AJR Am J Roentgenol 131:239–242, 1978. 422. Rosenthal DI, Hornicek FJ, Torriani M, et al: Osteoid osteoma: Percutaneous treatment with radiofrequency energy. Radiology 229:171–175, 2003. 423. Rosenthal DI, Marota JJ, Hornicek FJ: Osteoid osteoma: Elevation of cardiac and respiratory rates at biopsy needle entry into tumor in 10 patients. Radiology 226:125–128, 2003. 424. Rossi P, Arata FM, Salvatori FM, et al: Prosthetic graft infection: Diagnostic and therapeutic role of interventional radiology. J Vasc Interv Radiol 8:271–277, 1997. 425. Rossi S, Dore R, Cascina A, et al: Percutaneous computed tomographyguided radiofrequency thermal ablation of small unresectable lung tumours. Eur Respir J 27:556–563, 2006. 426. Rotte KH: CT-assisted needle biopsy in pulmonary and mediastinal space-occupying lesions: The effect of spiral CT and various biopsy needles on results. Aktuelle Radiol 5:136–169, 1995. 427. Rozenblit AM, Tuvia J, Rozenblit GN, et al: CT-guided transthoracic needle biopsy using an ipsilateral dependent position. AJR Am J Roentgenol 174:1759–1764, 2000. 428. Ryd W, Hagmar B, Eriksson O: Local tumour cell seeding by fineneedle aspiration biopsy: A semiquantitative study. Acta Pathol Microbiol Immunol Scand [A] 91:17–21, 1983. 429. Saini S, Mueller PR, Wittenberg J, et al: Percutaneous drainage of diverticular abscess: An adjunct to surgical therapy. Arch Surg 121:475–478, 1986. 430. Saluja S, Fields JM, Schwartz DS, et al: Percutaneous needle aspiration of small interloop abscesses in children. Pediatr Surg Int 13:528–530, 1998. 431. Samuel LH, Frierson HF, Jr: Fine needle aspiration cytology of acinar cell carcinoma of the pancreas: A report of two cases. Acta Cytol 40:585–591, 1996. 432. Sanfey H, Aguilar M, Jones RS: Pseudocysts of the pancreas, a review of 97 cases. Am Surg 60:661–668, 1994. 433. Sargent EN, Turner AF, Gordonson J, et al: Percutaneous pulmonary needle biopsy: Report of 350 patients. Am J Roentgenol Radium Ther Nucl Med 122:758–768, 1974. 434. Sateriale M, Cronan JJ, Savadler LD: A five year experience with 307 CT guided renal biopsies: Results and complications. J Vasc Interv Radiol 2:401–407, 1991. 435. Sato M, Tokui K, Watanabe Y, et al: Generalized intraperitoneal seeding of hepatocellular carcinoma after microwave coagulation therapy: A case report. Hepatogastroenterology 46:2561–2564, 1999. 436. Savage C, Walser EM, Schnadig V, et al: Transthoracic image-guided biopsy of lung nodules: When is benign really benign? J Vasc Interv Radiol 15:161–164, 2004. 437. Savarino V, Ceppa P, Biggi E, et al: Comparative study of percutaneous and preoperative fine-needle aspirations in the diagnosis of pancreatic cancer. Hepatogastroenterology 33:75–78, 1986. 438. Sayek I, Onat D: Diagnosis and treatment of uncomplicated hydatid cyst of the liver. World J Surg 25:21–27, 2001. 439. Schurawitzki H, Karnel F, Stiglbauer R, et al: CT guided percutaneous drainage and fluid aspiration in intensive care patients. Acta Radiol 33:131–136, 1992.

440. Schwartz LH, Macari M, Huvos AG, et al: Collision tumors of the adrenal gland: Demonstration and characterization at MR imaging. Radiology 201:757–760, 1996. 441. Scott EM, Marshall TJ, Flower CD, et al: Diffuse pleural thickening: Percutaneous CT-guided cutting needle biopsy. Radiology 194:867–870, 1995. 442. Sendler A, Avril N, Helmberger H, et al: Preoperative evaluation of pancreatic masses with positron emission tomography using 18F-fluorodeoxyglucose: Diagnostic limitations. World J Surg 24:1121–1129, 2000. 443. Shankar S, vanSonnenberg E, Morrison PR, et al: Combined radiofrequency and alcohol injection for percutaneous hepatic tumor ablation. AJR Am J Roentgenol 183:1425–1429, 2004. 444. Shetty PG, Fatterpekar GM, Manohar S, et al: Fatal cerebral air embolism as a complication of transbronchoscopic lung biopsy: A case report. Australas Radiol 45:215–217, 2001. 445. Shibata T, Iimuro Y, Yamamoto Y, et al: Small hepatocellular carcinoma: Comparison of radio-frequency ablation and percutaneous microwave coagulation therapy. Radiology 223:331–337, 2002. 446. Shibata T, Shibata T, Maetani Y, et al: Radiofrequency ablation for small hepatocellular carcinoma: Prospective comparison of internally cooled electrode and expandable electrode. Radiology 238:346–353, 2006. 447. Shiina S, Tagawa K, Niwa Y, et al: Percutaneous ethanol injection therapy for hepatocellular carcinoma: Results in 146 patients. AJR Am J Roentgenol 160:1023–1028, 1993. 448. Shin HJ, Lahoti S, Sneige N: Endoscopic ultrasound-guided fine-needle aspiration in 179 cases: The M.D. Anderson Cancer Center experience. Cancer 96:174–180, 2002. 449. Silverman S, Deuson TF, Kane N: Percutaneous abdominal biopsy: Cost-identification analysis. Radiology 206:429–435, 1998. 450. Silverman SG, Gan YU, Mortele KJ, et al: Renal masses in the adult patient: The role of percutaneous biopsy. Radiology 240:6–22, 2006. 451. Silverman SG, Mueller PR, Pinkney LP, et al: Predictive value of image-guided adrenal biopsy: Analysis of results of 101 biopsies. Radiology 187:715–718, 1993. 452. Simon CJ, Dupuy DE, DiPetrillo TA, et al: Pulmonary radiofrequency ablation: Long-term safety and efficacy in 153 patients. Radiology 243:268–275, 2007. 453. Singh S: Surgical management of complications associated with percutaneous and/or endoscopic management of pseudocyst of the pancreas. Ann Surg 244:630, 2006. 454. Sinner WN: Transthoracic needle biopsy of small peripheral malignant lung lesions. Invest Radiol 8:305–314, 1973. 455. Sinner WN: Fine-needle biopsy of hamartomas of the lung. AJR Am J Roentgenol 138:65–69, 1982. 456. Sinner WN, Sandstedt B: Small-cell carcinoma of the lung: Cytological, roentgenologic, and clinical findings in a consecutive series diagnosed by fine-needle aspiration biopsy. Radiology 121:269–274, 1976. 457. Sklair-Levy M, Polliak A, Shaham D, et al: CT-guided core-needle biopsy in the diagnosis of mediastinal lymphoma. Eur Radiol 10:714–718, 2000. 458. Smith EH: Fine-Needle Aspiration Biopsy: Are There Any Risks?, New York, 1985, Thieme. 459. Smith EH: Complications of percutaneous abdominal fine-needle biopsy. Radiology 178:253–258, 1986. 460. Smith TP, McDermott VG, Ayoub DM, et al: Percutaneous transhepatic liver biopsy with tract embolization. Radiology 198:769–774, 1996. 461. Sofocleous CT, Schubert J, Brown KT, et al: CT-guided transvenous or transcaval needle biopsy of pancreatic and peripancreatic lesions. J Vasc Interv Radiol 15:1099–1104, 2004. 462. Solbiati L, Goldberg SN, Lerace T, et al: Hepatic metastases: Percutaneous radio-frequency ablation with cooled-tip electrodes. Radiology 205:367–373, 1997. 463. Solbiati L, Lerace T, Goldberg SN, et al: Percutaneous US-guided radio-frequency tissue ablation of liver metastases: Treatment and follow-up in 16 patients. Radiology 202:195–203, 1997. 464. Solbiati L, Livraghi T, De Pra L, et al: Fine-needle biopsy of hepatic hemangioma with sonographic guidance. AJR Am J Roentgenol 144:471–474, 1985.

CHAPTER 67  Image-Guided Aspirations and Biopsies 465. Solbiati L, Livraghi T, Goldberg SN, et al: Percutaneous radio-frequency ablation of hepatic metastases from colorectal cancer: Long-term results in 117 patients. Radiology 221:159–166, 2001. 466. Song J, Cronan J: Percutaneous biopsy in diffuse renal disease: Comparison of 18 and 14 gauge automated biopsy devices. J Vasc Interv Radiol 9:651–655, 1998. 467. Sorensen B, Johansen P, Christiansen K, et al: Whole blood coagulation thromboelastographic profiles employing minimal tissue factor activation. J Thromb Haemost 1:551, 2003. 468. Soyer P, Pelage JP, Dufresne AC, et al: CT of abdominal wall implantation metastases after abdominal percutaneous procedures. J Comput Assist Tomogr 22:889–893, 1998. 469. Spamer C, Brambs HJ, Koch HK, et al: Benign circumscribed lesions of the liver diagnosed by ultrasonically guided fine-needle biopsy. J Clin Ultrasound 14:83–88, 1986. 470. Sperti C, Pasquali C, Decet G, et al: F-18-fluorodeoxyglucose positron emission tomography in differentiating malignant from benign pancreatic cysts: A prospective study. J Gastrointest Surg 9:22–28, discussion 28–29, 2005. 471. Sperti C, Pasquali C, Pedrazzoli S, et al: Expression of mucin-like carcinoma-associated antigen in the cyst fluid differentiates mucinous from nonmucinous pancreatic cysts. Am J Gastroenterol 92:672–675, 1997. 472. Spiro TP, Gerson SL, Liu L, et al: O6-benzylguanine: A clinical trial establishing the biochemical modulatory dose in tumor tissue for alkyltransferase-directed DNA repair. Cancer Res 59:2402–2410, 1999. 473. Spiro TP, Liu L, Majka S, et al: Temozolomide: The effect of once- and twice-a-day dosing on tumor tissue levels of the DNA repair protein O-alkylguanine-DNA-alkyltransferase. Clin Cancer Res 7:2309–2317, 2001. 474. Stabile BE, Puccio E, van Sonnenberg E, et al: Preoperative percutaneous drainage of diverticular abscesses. Am J Surg 159:99–104, 1990. 475. Steinke K, King J, Glenn DW, et al: Radiologic appearance and complications of percutaneous computed tomography-guided radiofrequency-ablated pulmonary metastases from colorectal carcinoma. J Comput Assist Tomogr 27:750–757, 2003. 476. Steinke K, King J, Glenn DW, et al: Percutaneous radiofrequency ablation of lung tumors with expandable needle electrodes: Tips from preliminary experience. AJR Am J Roentgenol 183:605–611, 2004. 477. Stevens GM, Jackman RJ: Outpatient needle biopsy of the lung: Its safety and utility. Radiology 151:301–304, 1984. 478. Stolier A, Skinner J, Levine EA: A prospective study of seeding of the skin after core biopsy of the breast. Am J Surg 180:104–107, 2000. 479. Suarez JF, De Teresa J, Gonzalez A, et al: Papillary tumor of the pancreas diagnosed by fine-needle aspiration. Rev Esp Enferm Dig 88:53–55, 1996. 480. Suh RD, Wallace AB, Sheehan RE, et al: Unresectable pulmonary malignancies: CT-guided percutaneous radiofrequency ablation– Preliminary results. Radiology 229:821–829, 2003. 481. Suzawa N, Yamakado K, Takeda K: Therapeutic response to lung radiofrequency ablation for lung neoplasms using F-18 deoxyglucose (FDG)-PET/CT: A prospective study. J Nucl Med 48:85, 2007. 482. Syvanthong C, Wile GE, Zagoria RJ: Effect of radiofrequency ablation of renal tumors on renal function in patients with a solitary kidney. AJR Am J Roentgenol 188:1619–1621, 2007. 483. Taft PD, Szyfelbein WM, Greene R: A study of variability in cytologic diagnoses based on pulmonary aspiration specimens. Am J Clin Pathol 73:36–40, 1980. 484. Talbot JN, Gutman F, Fartoux L, et al: PET/CT in patients with hepatocellular carcinoma using [(18)F]fluorocholine: Preliminary comparison with [(18)F]FDG PET CT. Eur J Nucl Med Mol Imaging 33:1285–1289, 2006. 485. Tandon A, Jain AK, Dixit VK, et al: Needle aspiration in large amoebic liver abscess. Trop Gastroenterol 18:19–21, 1997. 486. Thanos I, Diliana T: Percutaneous CT guided drainage of splenic abscess. AJR Am J Roentgenol 179:629–632, 2002.

2583

487. Thornbury JR, Burke DP, Naylor B: Transthoracic needle aspiration biopsy: Accuracy of cytologic typing of malignant neoplasms. AJR Am J Roentgenol 136:719–724, 1981. 488. Titton RL, Lucey B, Gervais DA, et al: Celiac plexus block: A palliative tool underused by radiologists. AJR Am J Roentgenol 179:633–636, 2002. 489. Tolly TL, Feldmeier JE, Czarnecki D: Air embolism complicating percutaneous lung biopsy. AJR Am J Roentgenol 150:555–556, 1988. 490. Tong NY, Ru HJ, Ling HY, et al: Extracardiac radiofrequency ablation interferes with pacemaker function but does not damage the device. Anesthesiology 100:1041, 2004. 491. Topal U, Ediz B: Transthoracic needle biopsy: Factors effecting risk of pneumothorax. Eur J Radiol 48:263–267, 2003. 492. Trojan J, Schroeder O, Raedle J, et al: Fluorine-18 FDG positron emission tomography for imaging of hepatocellular carcinoma. Am J Gastroenterol 94:3314–3319, 1999. 493. Tuncali K, Morrison PR, Tatli S, et al: MRI guided percutaneous cryoablation of renal tumors: Use of external manual displacement of adjacent bowel loops. Eur J Radiol 59:198–202, 2006. 494. Tuncali K, van Sonnenberg E, Shankar S, et al: Evaluation of patients referred for percutaneous ablation of renal tumors: Importance   of a preprocedural diagnosis. AJR Am J Roentgenol 183:575–582, 2004. 495. Tylen U, Arnesjo B, Lindberg LG, et al: Percutaneous biopsy of carcinoma of the pancreas guided by angiography. Surg Gynecol Obstet 142:737–739, 1976. 496. van der Waaij LA, van Dullemen HM, Porte RJ: Cyst fluid analysis in the differential diagnosis of pancreatic cystic lesions: A pooled analysis. Gastrointest Endosc 62:383–389, 2005. 497. Van Sonnenberg E: Percutaneous drainage of abscesses and fluid collection: Technique, results, and applications. Radiology 142:1–10, 1982. 498. Van Sonnenberg E: Percutaneous drainage of infected and noninfected pancreatic pseudocysts: Experience in 101 cases. Radiology 170:757–761, 1989. 499. Van Sonnenberg E: Percutaneous abscess drainage. Curr Concepts Radiol 181:617–626, 1991. 500. Van Sonnenberg E, Goodacre BW: Percutaneous gallbladder puncture and cholecystostomy: Results, complications, and caveats for safety. Radiology 183:167–170, 1992. 501. Van Waes PF, Feldberg MA, Mali WP, et al: Management of loculated abscesses that are difficult to drain: A new approach. Radiology 147:57–63, 1983. 502. Veit P, Antoch G, Stergar H, et al: Detection of residual tumor after radiofrequency ablation of liver metastasis with dual-modality PET/CT: Initial results. Eur Radiol 16:80–87, 2006. 503. Venbrux AC, Montague BJ, Murphy KP, et al: Image-guided percutaneous radiofrequency ablation for osteoid osteomas. J Vasc Interv Radiol 14:375–380, 2003. 504. von Schreeb T, Arner O, Skovsted G, et al: Renal adenocarcinoma: Is there a risk of spreading tumour cells in diagnostic puncture? Scand J Urol Nephrol 1:270–276, 1967. 505. Wacker FK, Nour SG, Eisenberg R, et al: MRI-guided radiofrequency thermal ablation of normal lung tissue: In vivo study in a rabbit model. AJR Am J Roentgenol 183:599–603, 2004. 506. Wah TM, Arellano RS, Gervais DA, et al: Image-guided percutaneous radiofrequency ablation and incidence of post-radiofrequency ablation syndrome: Prospective survey. Radiology 237:1097–1102, 2005. 507. Warshaw AL, Rattner DW: Timing of surgical drainage for pancreatic pseudocyst: Clinical and chemical criteria. Ann Surg 202:720–724, 1985. 508. Weisbrod GL, Lyons DJ, Tao LC, et al: Percutaneous fine-needle aspiration biopsy of mediastinal lesions. AJR Am J Roentgenol 143:525–529, 1984. 509. Welch CE: Percutaneous catheter drainage of abdominal abscesses. Letter to the. N Engl J Med 306:108, 1982. 510. Westcott JL: Air embolism complicating percutaneous needle biopsy of the lung. Chest 63:108–110, 1973.

2584

PART III  Image-Guided Procedures

511. Westcott JL: Direct percutaneous needle aspiration of localized pulmonary lesions: Result in 422 patients. Radiology 137:31–35, 1980. 512. Westcott JL: Percutaneous needle aspiration of hilar and mediastinal masses. Radiology 141:323–329, 1981. 513. Westcott JL: Percutaneous transthoracic needle biopsy. Radiology 169:593–601, 1988. 514. Willson JK, Haaga JR, Trey JE, et al: Modulation of O6-alkylguanine alkyltransferase-directed DNA repair in metastatic colon cancers. J Clin Oncol 13:2301–2308, 1995. 515. Woertler K, Vestring T, Boettner F, et al: Osteoid osteoma: CT-guided percutaneous radiofrequency ablation and follow-up in 47 patients.   J Vasc Interv Radiol 12:717–722, 2001. 516. Wolinsky H, Lischner MW: Needle track implantation of tumor after percutaneous lung biopsy. Ann Intern Med 71:359–362, 1969. 517. Wroblicka JT, Kuligowska E: One-step needle aspiration and lavage for the treatment of abdominal and pelvic abscesses. AJR Am J Roentgenol 170:1197–1203, 1998. 518. Wu H, Cheng NS, Zhang YG, et al: Improved early diagnosis of cystadenocarcinoma of the pancreas. Hepatobiliary Pancreat Dis Int 6:87–91, 2007. 519. Wu H, Haaga J, Exner A: Radiofrequency ablation postablation assessment using CT perfusion with pharmacologic modulation in a rat subcutaneous tumor model. Acad Radiol 16:321–331, 2009. 520. Wudel LJ, Jr, Delbeke D, Morris D, et al: The role of [18F] fluorodeoxyglucose positron emission tomography imaging in the evaluation of hepatocellular carcinoma. Am Surg 69:117–124, discussion 124–126, 2003. 521. Xia JL, Ye SL, Zou JH, et al: Prognosis and risk factors for recurrence of small liver cancer after a single session of percutaneous radiofrequency ablation. Ai Zheng 23:977–980, 2004. 522. Yamagami T, Kato T, Hirota T, et al: Pneumothorax as a complication of percutaneous radiofrequency ablation for lung neoplasms. J Vasc Interv Radiol 17:1625–1629, 2006. 523. Yamagami T, Kato T, Tanaka O, et al: Radiofrequency ablation therapy of remnant colorectal liver metastases after a course of hepatic arterial infusion chemotherapy. J Vasc Interv Radiol 16:549–554, 2005.

524. Yamakado K, Hase S, Matsuoka T, et al: Radiofrequency ablation for the treatment of unresectable lung metastases in patients with colorectal cancer: A multicenter study in Japan. J Vasc Interv Radiol 18:393–398, 2007. 525. Yang W, Chen MH, Yin SS, et al: Radiofrequency ablation of recurrent hepatocellular carcinoma after hepatectomy: Therapeutic efficacy on early- and late-phase recurrence. AJR Am J Roentgenol 186:S275–S283, 2006. 526. Reference deleted in proofs. 527. Yeow KM, See LC, Lui KW, et al: Risk factors for pneumothorax and bleeding after CT-guided percutaneous coaxial needle biopsy of lung lesions. J Vasc Interv Radiol 12:1305–1312, 2001. 528. Yohann MM, Haaga JR: Complications of no biopsy. AJR Am J Roentgenol 172:1475–1476, 1999. 529. Zagoria RJ, Hawkins AD, Clark PE, et al: Percutaneous CT-guided radiofrequency ablation of renal neoplasms: Factors influencing success. AJR Am J Roentgenol 183:201–207, 2004. 530. Zelas P, Haaga JR, Fazio VW: The diagnosis by percutaneous biopsy with computed tomography of recurrence of carcinoma of the rectum in the pelvis. Surg Gynecol Obstet 151:525–527, 1980. 531. Zhang AB, Zheng SS: Treatment of pancreatic pseudocysts in line with D’Egidio’s classification. World J Gastroenterol 11:729–732, 2005. 532. Zinzani PL, Corneli G, Cancellieri A, et al: Core needle biopsy is effective in the initial diagnosis of mediastinal lymphoma. Haematologica 84:600–603, 1999. 533. Zornoza J, Cabanillas FF, Altoff TM, et al: Percutaneous needle biopsy in abdominal lymphoma. AJR Am J Roentgenol 136:97–103, 1981. 534. Reference deleted in proofs. 535. Zornoza J, Snow J, Jr, Lukeman JM, et al: Aspiration biopsy of discrete pulmonary lesions using a new thin needle: Results in the first 100 cases. Radiology 123:519–520, 1977. 536. Reference deleted in proofs. 537. Zuckerman DA, Yeager TD: Percutaneous ethanol sclerotherapy of postoperative lymphoceles. AJR Am J Roentgenol 169:433–437, 1997.

68  Computed Tomography-Guided Drainage Dean A. Nakamoto and Ji Y. Buethe

With developments in cross-sectional imaging, there have been major advances in technical materials enabling interventional radiologists to perform procedures previously considered impossible. Before computed tomography (CT), percutaneous abscess drainage (PAD) was occasionally performed when abscesses were so large they could be palpated, visible on fluoroscopy by displacement of bowel, or superficially located and visible by ultrasound (US). Since the first CT-guided percutaneous aspiration and drainage procedure, which was performed in 1975 and reported in 197688 by Haaga, use of this technique has expanded rapidly. CT has been especially useful because it has a high accuracy for detecting abscesses and allows accurate determination of the anatomic location and extent of the lesion. Because CT is not impeded by gas or bone, any involvement of peritoneal spaces or anatomic areas can be delineated clearly. This information is necessary if one is to determine whether an abscess can be drained by a percutaneous method. Although there was some controversy in the early literature concerning the merits of different guidance systems for drainage, there now seems to be a consensus that CT guidance is the best method available. Primary and postoperative abscesses and fluid collections in nearly every organ system have been successfully treated by percutaneous means. PAD has become the treatment of choice for a wide variety of fluid collections, and this attests to its advantages over the alternative of surgical drainage by offering precise noninvasive localization of fluid collections, minimally invasive therapeutic techniques, and avoidance of general anesthesia in most cases.76,89,180,264 This has resulted in reduced morbidity and mortality and has helped reduce length of hospital stay and hospital costs compared with open surgical drainage. There has been a transition from open surgical drainage to PAD over the last 30 years, except in the most difficult or inaccessible cases. Results vary according to the complexity of the abscess. Although complications are relatively uncommon, they may be serious, even life threatening. Close communication and strategy between the surgeon and interventional radiologist provide synergy that ultimately benefits the patient.

PERSPECTIVES ON ABSCESS DRAINAGES In the past 3 decades, percutaneous drainage of primary and postoperative abscesses and fluid collections has had a profound effect on the management of the critically ill patient population. The benefits of improved detection and localization of abscess have been documented by Deveney and coworkers, who showed conclusively that the improved localization and intensive care methods have resulted in a greater percentage of drainage success and decreased mortality resulting from organ failure.52 These authors studied two 5-year periods (1973-1978

and 1981-1986) in a veteran population. The population included patients treated either surgically or percutaneously. The mortality of the two groups was 39% from 1973 to 1978 and 21% from 1981 to 1986. The successful initial drainage was 55% in the earlier group compared with 74% in the latter group. Organ failure rate decreased from 52% to 23%. The surgical success rate was 76% compared with the percutaneous success rate of 72%. The authors concluded that accurate localization and early drainage are the most important factors and that there is no difference in outcome between the surgical and percutaneous methods. Despite the lack of randomized studies comparing percutaneous to surgical drainage, many studies have been conducted to retrospectively compare the results of percutaneous drainage with the results of surgical drainage in patients with intraabdominal abscesses. Their findings suggest that percutaneous drainage of an abscess can be performed as safely as surgical drainage.96,98 With a broad perspective toward drainage, the intent of interventional percutaneous procedures is at least to duplicate the treatment result established by surgeons. The benefit of percutaneous procedures lies in their simplicity, lower complication rates, and presumably lower overall expense. The general view is increasing that PAD is a safe and effective alternative to surgical drainage.82,160,271 Percutaneous drainage can be effective in the management of complex abscesses, including multiple abscesses, ones associated with fistula, and splenic abscesses.

PATIENT SELECTION In recent years the indication for percutaneous methods has expanded significantly. Unlike most new techniques and procedures in which there is an initial period of enthusiasm followed by either moderation or a restriction of the method, percutaneous drainage guided by CT has continued to expand its role. Accumulated experience and refined techniques have expanded the indications for image-guided abscess drainage to permit drainage of complicated abscesses. Examples of such collections are gallbladder infection, pancreatic pseudocysts and abscesses, lymphoceles, and mediastinal and lung abscesses. The prerequisite for percutaneous drainage is an abnormal fluid collection and one of the following: suspicion that the fluid is infected, need for fluid characterization, or suspicion that the collection is producing symptoms to warrant drainage.11 To maintain a proper clinical perspective, it is important that the appropriate patients be selected for curative or palliative PAD procedures and for surgical procedures. Simplistically one could try PAD for palliation in every patient, but it is not logical to expand its use for patients with certain surgical problems. On a practical level, although the role and the indications for percutaneous procedures have expanded greatly, there are still clear indications for surgical drainages.

2585

2586

PART III  Image-Guided Procedures

Although there are few contraindications to most imaging-guided fluid aspirations and drainages, one important contraindication is an uncooperative patient. Relative contraindications include patients with abnormal clotting parameters: an INR (international normalized ratio) of greater than 2, an elevated prothrombin time 3 seconds or more above control, a partial thromboplastin time 6 seconds or more above control, or a platelet count less than 50,000/mm3. These abnormalities can usually be corrected with appropriate blood products. If a patient is febrile and an infected collection is suspected, we prefer that the patient have at least one dose of intravenous (IV) broadspectrum antibiotics prior to the procedure. One dose of antibiotics usually does not alter the results of the specimen culture. The imaging appearance of the abscess indicates how easily an abscess might be drained. Abscesses that are of low density and clearly defined, without septations or not involving numerous peritoneal spaces, are easily drained. The results with septated abscess have varied; some drain well, and others do not. The most plausible explanation is that fibrin septa in such cavities may be complete or incomplete. When they are complete, drainage is poor, and when they communicate and are incomplete, curative drainage can occur. We have studied the use of fibrinolytic agents in the former cases and have found them effective for improving drainage (see “Fibrinolytic Agents,” later). Also, the multiplicity of abscess is no longer considered a contraindication to percutaneous drainage, but multiple catheters in each site may be required for optimal drainage. For the anatomic pathways, we still prefer a clear pathway into the abscess cavity without traversing an uninvolved space or organ. A routine penetration through an uninvolved organ is inappropriate because of possible seeding of pyogenic organisms into the organ, as well as possible injury to the otherwise uninvolved organ. Although with the potent antibiotics of today most such procedures seem feasible, the success and complication rate will favor avoidance of uninvolved organs and spaces.72,108 There are several techniques that can be used to create a safe window for percutaneous access, including changing patient position and instillation of air or saline through the access needle (Fig. 68-1). Viscosity or thickness of the purulent material in the abscess affects drainage success. If the entire cavity can be evacuated easily after aspiration, percutaneous drainage probably would work. If the material appears to be viscous, techniques should be adjusted appropriately (i.e., special attention given to catheter size, catheter placement, and use of an appropriate irrigant; see “Clinical Management of Catheters,” later). The appropriateness of percutaneous drainages for abscess associated with entities requiring surgical treatment is an area of significant controversy. One must weigh the palliative benefits of PAD against the final curative benefits of surgery. On the other hand, the use of PAD for complex abscesses may offer significant therapeutic benefits and palliation by draining most of the infected fluid, allowing the patient to undergo surgery in a more stable condition.82,271 This approach for gaining time has been applied successfully in patients with enteric abscesses, appendicitis,110 Crohn’s disease,35 diverticulitis,217 and cholecystitis. PAD has converted emergency surgery into elective surgery and has increased the possibility of a successful one-stage surgical procedure in recent years.62,220 The success of conservative diverticular disease management including PAD has led some investigators to argue against obligatory elective surgery.286 Patient preparation requires evaluation and normalization of coagulation status, administration of periprocedural, culture-specific, or broad-spectrum antibiotics, and evaluation of the patient for tolerance of conscious sedation. Procedures are performed with IV conscious sedation and local anesthesia. Conscious sedation is critical to reduce

patient pain and anxiety, increase patient ease and comfort, and thereby decrease the risk of complications. In patients with an inability to tolerate conscious sedation owing to pain, cardiopulmonary instability, or allergies to local anesthetic agents, general anesthesia may be required. For some patients, the procedure may be performed with only local anesthetic because the risk of sedation or anesthesia may be greater than the risk of the procedure. There is a spectrum of complexity for intraabdominal abscess. Examples of more complex situations include multiple abscesses, abscess due to Crohn’s disease, pancreatic abscess, drainage route that traverses bowel or pleura, infected hematoma, or infected tumor.52,236 Although some authors have reported curative or partially successful PAD in patients with these complex pathologies, one should expect that these cases are associated with lower success rates, higher rates of complications, technical difficulties, longer periods of time for drainage, and increased chance of recurrence.108,135 Decision regarding percutaneous versus surgical drainage of complex collections should be made in concert with other services involved in the patient’s care.

INDICATIONS FOR SURGERY Some abscesses are difficult to percutaneously drain and may be better suited for surgical drainage. This group consists of fungal abscesses, infected hematomas, echinococcal disease, most pancreatic abscesses, particularly infected necrotic pancreatitis, and failed percutaneous drainage. Patients with fungal abscesses often require surgical drainage because there is extensive tissue invasion, necrosis, and mycotic plaque formation within the wall of the cavity. Our experience shows that surgical drainage or resection is best for adequate débridement in these patients.91,129 Infected hematomas do not drain well because of the extensive amount of fibrin deposition and the protective effects of fibrin  on bacteria. As shown from our early experience and from reports  in the literature, such hematomas have uniformly failed by percu­ taneous drainage. In our more recent experience, fibrinolytic agents used as a catheter irrigant can improve the drainage of infected hematomas.93 Echinococcal disease requires a special approach to prevent spread of the daughter cysts into the peritoneum or other areas. The issues related to anaphylaxis are essentially minimal with appropriate technique, and many authors have now reported on the technique of percutaneous drainage (see “Echinococcal Drainage,” later). Unless the interventionalist practices in an endemic area, however, it is unlikely that he or she would develop sufficient expertise to manage echinococcal disease well. In nonendemic areas, echinococcal disease may best be treated by the surgical method. Finally, for patients who have had a failed percutaneous procedure, surgical drainage is appropriate. Because of the difference in philosophy among surgeons, radiologists, and institutions, the definition of a failed procedure can be determined only at the local level and not in this text.

IMAGING GUIDANCE AND ACCESS The most important step in percutaneous drainage procedures is selection of the access route and imaging modality for guidance of needle and catheter insertion. Whenever possible, guidance with US should be considered because it allows real-time monitoring of manipulations and no radiation to the patient. Seemingly inaccessible pelvic fluid collections may be approached with US using a transvaginal or trans­ rectal route. These techniques will be discussed in the pelvic abscess section.

CHAPTER 68  Computed Tomography-Guided Drainage

A

B

C

D

E FIG 68-1  Technique to increase pleural space for safe access to middle mediastinum using saline injection. A, This 68-year-old with a history of cervical carcinoma and enlarging subcarinal lymph node presented for CT-guided fiducial placement. B, Soft tissue and bone window images demonstrate narrow window of pleural space, with a 15-gauge coaxial needle being advanced via transcostal approach. C, Saline (5 mL) was instilled to increase pleural space. D, This technique created a safe window for further advancement of the needle to the middle mediastinum through the pleural space without traversing lung parenchyma. E, Final CT images following fiducial placement. No evidence of hematoma or pneumothorax is identified.

2587

2588

PART III  Image-Guided Procedures

In general the shortest distance between the skin and the collection without interposing organ or vessel is chosen for needle entry and catheter insertion. Selection of the skin entry site is based on previous diagnostic studies and anatomic considerations. The preselected pathway is confirmed by immediate limited preprocedure CT study. A wide area is sterilized and local anesthesia is applied to the skin and underlying soft tissues where the needle and catheter will be inserted. There is still a debate about which is better, aspiration alone with and without intermittent aspiration or continuous catheter drainage. Some authors believe that aspiration and lavage of the abscess is as effective as more conventional treatment of prolonged catheter drainage.289 They believe that more manipulation of instruments in the abscess cavity may increase the risk of septicemia.136 Other advantages of this treatment method include less invasiveness, lower cost, and no need for follow-up catheter care.10,78 Rae and colleagues performed percutaneous aspiration alone on 25 patients with hepatic abscesses smaller than 5 cm and had a 100% cure rate.203 Zarem and Hadzic had 100% success rate for US-guided aspiration of small liver abscesses.293 They also concluded that continuous catheter drainage was more efficient than intermittent aspiration for liver abscess bigger than 5 cm. Giorgio et al. reported a 98% cure rate of 115 patients who suffered from liver abscess and were treated by US-guided needle aspiration alone.78 On the other hand, many authors believe that continuous percutaneous catheter drainage is more efficient than intermittent percutaneous needle aspiration. Catheter drainage is also reported to be more efficient for large abscesses (diameter > 5 cm) and multiloculated abscesses.205,294

GENERAL TECHNIQUE FOR ASPIRATION TREATMENT Selection of patients for aspiration treatment is based on the characteristics of the abscess. We use this technique when an abscess is too small or inaccessible for catheter drainage. Preprocedure steps are identical to other procedures relative to review of history, medications, imaging, and patient consent. Needle aspiration of the abscess is performed using similar guidelines as described later under “General Technique and Steps for Catheter Drainage.” We typically use one of the commercially available trocar-based catheter needles to access the collection, such as a Yueh Centesis Catheter Needle (Cook Medical) or a One-Step Centesis Catheter (Merit Medical), which has a 5-French (F) catheter over a 19-gauge needle. Such access needles are preferred because purulent material may be thick and the multiholed 5F catheters are more reliable for recovering a sample than a simple hollow needle. In addition, a standard 0.035-inch guidewire can then be placed through the 5F catheter into the abscess for subsequent catheter insertion using the Seldinger technique. On recovery of the purulent material, we perform aspiration of purulent contents and sometimes injection of a small dose of antibiotics. The antibiotic chosen should be the one selected from cultures and sensitivities (if available from fluid or blood cultures). The amount of antibiotics injected is equivalent to a nondiluted single intramuscular or IV dose. That amount of drug is deducted from the daily limit stipulated by the manufacturer and approved by the U.S. Food and Drug Administration (FDA). It also is important to pay attention to the volume required, because one must not inject volumes too large for the cavity. For typical pyogens, the volume is not an issue when routine broad-spectrum antibiotics are used. With fungal abscesses,

the volume of the dilute drug might be too high, so the dosage needs to be reduced.

GENERAL TECHNIQUE AND STEPS FOR CATHETER DRAINAGE The technique of abscess drainage can be divided into the following steps: (1) abscess detection, (2) preprocedure patient workup, (3) choice of the guidance modality, (4) trajectory planning, (5) diagnostic aspiration, (6) catheter selection and insertion.

Abscess Detection If an abscess has been detected by magnetic resonance imaging (MRI), radioisotope-tagged white blood cells (WBCs), or US, a CT scan should be performed to determine the complete extent of the abscess. The geographic location and degree of involvement are essential factors in choosing the trajectory, determining the number of catheters required, and anticipating possible difficulties. An abscess can be detected in selected cases by any modality, but CT plays the most critical role. CT is the preferred examination because it is the most accurate of all imaging systems for abscess detection. Not only can CT confirm or exclude the presence of drainable fluid, it can also diagnose the causative or associated problem. CT can provide valuable information about diverticulitis, tumors, appendicitis, fistulas, infarcts, cellulitis, phlegmons, inflammatory bowel disease, bowel obstructions, or other problems.

Preprocedure History, Review of Images, Planning, and Informed Consent Preprocedure steps include physical examination, clinical and medication history, and review of all imaging and clinical data. Medication history is especially important to determine if any anticoagulants are taken (either prescribed or over the counter [OTC]). Laboratory studies including complete blood counts and coagulation values are reviewed. After complete patient interview and discussion of the indications, procedure techniques, expected outcomes, possible complications, and follow-up expectations, an appropriate witnessed informed consent is obtained.

Imaging Preferences An appropriate choice of imaging modality for guidance is a critical step because it provides information regarding feasibility, appropriate technique, and associated risks for the purposes of informed consent and procedural planning. US is becoming increasingly popular as interventionalists become more experienced with this modality. However, CT is the preferred modality to identify the extent of abscesses and plan the route of access; distinction of the fluid collection from the adjacent normal structures can be exquisitely enhanced by the administration of oral and IV contrast agents. Air is accurately imaged and localized. Adjacent anatomy and virtually all instruments can be visualized without difficulty. CT is also advantageous for guiding drainage, particularly in cases in which collections are small and deep, in close proximity to vital structures, or located in regions that are difficult to assess. In addition, CT fluoroscopy promises to increase accuracy and decrease procedure time.

Target and Trajectory Determination Trajectory planning is best accomplished by CT, regardless of the imaging modality chosen for guidance. Although as most authors agree, it is best to choose the shortest pathway, it is also optimal to avoid uninvolved organs and noncontiguous peritoneal spaces. For

CHAPTER 68  Computed Tomography-Guided Drainage example, common spaces where abscesses can form in the abdomen are the right subphrenic, left subphrenic, and lesser sac. To plan the trajectory in these areas, it is crucial to understand and appreciate the spaces as they relate to the diaphragm and rib cage (Fig. 68-2). Several points should be made concerning the configuration and nature of the abscess collection. If an abscess collection is long, the trajectory of the puncture should be planned so that the catheter lies throughout the length of the cavity when it is inserted. Two approaches may be used. First, one can choose an entrance site at the upper or lower end of the cavity so that the catheter traverses the entire length or width. The alternative is to make two punctures at the center point, one directed cephalad and one directed caudad, to permit positioning of a catheter at the top and bottom of the cavity (see “Catheter Selection and Insertion Method”). If such an abscess is elongated in a traverse direction, catheters should be similarly placed in that transverse plane. If multiple septations are seen within a cavity, it is reasonable to plan the trajectory so that the maximum number of cavities would be traversed. Although the best surgical results are obtained if an extraperitoneal approach is used, this is not true with percutaneous procedures. Neither we nor other authors have found any indications that a peritoneal approach differs in outcome from an extraperitoneal approach with percutaneous drainage. With percutaneous drainage of peritoneal abscesses, the high-pressure abscess evacuates through the catheter and not into the peritoneum, because the margins of the peritoneum around the abscess are likely fused from inflammation. Drainage of the liver is different and requires a cuff of normal liver (see “Approaches to Specific Abscess Sites”).

Diagnostic Aspiration Patient vital signs and electrocardiographic (ECG) monitoring with automated devices and local anesthetic and sedation administration are the same as used for other sites. Sedation is administered as needed. Most authors use short-acting drugs such as midazolam and fentanyl (or both) as needed. Local anesthesia is administered. Before catheter or trocar placement, a diagnostic aspiration is performed to confirm the presence of an abscess. We typically use a 5F Yueh needle to perform diagnostic aspiration for two reasons. First, a larger needle permits aspiration of even the thickest purulent material. Second, if at the time of aspiration the fluid appears purulent, the sheath permits insertion of an angiographic wire and catheter without a second puncture. The actual puncture and placement of the needle are performed in the same manner as any other needle procedure. We take two steps to minimize the risk of contamination along the pathway during the subsequent steps of catheter insertion. First, when draining abscesses in the liver, we choose the path to include a “cuff ” of normal parenchyma to lessen the chance of tearing the thin wall of an abscess. We try to avoid traversing normal spleen or kidney while draining splenic or renal abscesses. Second, at the time the puncture is made, we aspirate a sufficient amount of material to lessen the “back pressure” to minimize the chance of local spillage. There is a reported case of spillage into the peritoneum that produced peritonitis and a case of dissemination into the subphrenic area reported by Johnson and coworkers.112 One must always strive for meticulous technique, but if spillage of purulent material occurs, placement of a second catheter at the site would prevent further problems. If the aspirated material is not obviously purulent, then one can decide among three different management options: (1) place a catheter, (2) aspirate the entire collection, or (3) aspirate a small diagnostic sample and send it to the laboratory for evaluation. There are

2589

advantages to each of these options. We often perform the first option of placing a catheter, particularly if the collection is in a difficult location to access. If the cultures are negative, the catheter can be removed in a day or two. With current antibiotics, the chances of a sterile collection getting a secondary infection from the catheter are low. If the collection is small and simple, the entire collection can be aspirated and thus completely resolve. If the collection is septated and easy to access, then the third option of a small diagnostic aspirate can be performed. In the event that a repeat puncture and drainage might be required, we try not to aspirate the fluid collection “dry”; if such a collection is aspirated dry, there would be no well-defined lesion to target should a drainage be required later. This third option is very conservative and may be considered in such collections in patients who are immunocompromised. Extracted fluid may be characteristic for collections other than abscess. Creatinine is elevated in urinomas; the presence of lymphocytes and fat globules signifies lymphoceles; amylase characterizes pancreatic pseudocysts; and bilirubin is diagnostic for bile collections. Seromas typically yield clear yellowish fluid. In cancer patients with a possible necrotic tumor vs. abscess, a cytologic sample should be sent to determine if there is a necrotic neoplasm or secondarily infected tumor giving toxic symptoms similar to abscesses.

Catheter Selection and Insertion Method Assessment of the type of fluid to be drained (e.g., pus, bile, lymph, urine, blood) and its characteristics (e.g., clear, particulate, viscous, clot filled) are used to select the drainage catheter size and type. Generally a 10F to 14F catheter with multiple side holes can effectively drain thick, purulent material; similar but smaller catheters (6F or 8F) may be used to drain collections of nonviscous fluid such as that in cysts or seromas and uninfected effusions. The size of the catheters and the size of the pigtail shape to be used should be matched to the size of the abscess cavity so that the entire curved end is in the cavity. It is important to use the largest catheter possible from the beginning; if the procedure becomes technically difficult, one can settle for almost any size of catheter and upgrade it later. It is important not to have holes outside the cavity. If holes are outside, leakage into a sterile space or organ can occur. The catheter is secured to the skin with an adhesive device or suture and sterilely dressed. In patients at high risk for accidental catheter removal, the catheter should be secured directly to the skin with suture. Catheter position should be noted and marked if necessary to detect accidental removal. The catheter should be drained to a suction bulb. Dressings can be changed during rounds or by the nursing staff. Two types of techniques can be used to place drainage catheters: Seldinger method and single-step catheters. Either may be used depending on the specific nature of each case. The Seldinger method is well suited to all abscesses regardless of the size or location; even the most difficult deep abscess can be drained with this method. We use the single-step catheters with larger superficial abscesses.

Seldinger Method.  Seldinger technique is the preferred method for placement of catheters in most collections, including small collections and those that are difficult to assess. Seldinger technique carries the advantage of verification of successful access into the collection and successful avoidance of adjacent structures prior to the creation of a large-bore tract. The principal disadvantage of this technique is potential risk of loss of access and cross-contamination while exchanging dilators, during which abscess contents have the potential to spread to the bloodstream or adjacent spaces. Access is gained using a 5F Yueh needle (or similar catheter-based needle). After the needle has entered the cavity and a sample has been

2590

PART III  Image-Guided Procedures

A

B

C

D

E FIG 68-2  This 37-year-old patient presented with worsening leukocytosis and epigastric pain status post pancreatitis. A, CT through the upper abdomen demonstrates a peripancreatic abscess within the left subphrenic region. Abscess is centered at the level of the diaphragm. B, Axial CT image obtained at a level 1 cm caudal to the first image demonstrates adjacent loops of colon in the region of the potential percutaneous access route. A decision was made to target the level between these two images to avoid diaphragm/pleura and adjacent bowel loops. C, A 5F Yueh needle was placed under CT guidance. D, Axial CT image demonstrating Yueh needle within the abscess. E, Final CT image demonstrating satisfactory position of the catheter.

CHAPTER 68  Computed Tomography-Guided Drainage aspirated for culture and to relieve the pressure. A 0.035-inch guidewire is advanced into the cavity. Choice of the guidewire is an individual preference; we prefer a slightly stiffer wire such as a Rosen or an Amplatz so that dilation of the tract and catheter insertion is facilitated. Although some interventionalists are reluctant to use stiff wires, the stiffness prevents kinking of the wire or catheter during the dilation or insertion process. If the abscess is secondary to a local infarction or necrosis of the organ, a softer, more flexible wire may be used. Availability of a portable fluoroscopy device or an attached fluoroscopy unit is helpful to confirm guidewire location. With experience, one can insert the catheter on the CT table, using the topogram as the method for confirming the catheter location or blindly using a sense of “feel.” Under direct visualization, serial fascial dilation over the guidewire is performed to accommodate the size of the drainage catheter. A lockingloop pigtail catheter with a metal or plastic stiffener is advanced over the wire into the collection. The stiffener and wire are removed and the catheter is locked.

Single-Stick or Trocar Technique.  Some practitioners use the single-step trocar technique routinely, but one should be cautious when using this technique for small or critically located abscesses (see “Drainage Errors”). We use this technique with US-guided transvaginal, transrectal, and selected cholecystostomy catheter placements and CT-guided drainage of larger more superficial collections in our institution. This technique facilitates rapid drainage of the collection and minimal potential for spreading the infection owing to the absence of serial fascial dilation. However, the long trocar associated with the catheters may not fit into the CT gantry, particularly with larger patients. The trocar-based catheter does not penetrate the soft tissues as easily as a 5F Yueh needle. These two factors limit their use in deep smaller lesions that might be close to critical structures. Another principal disadvantage is the direct advancement of a large-bore catheter and sharp stylet, which can have serious consequences in cases of inadvertent nontarget access.175 In this technique, direct CT guidance is used to access the fluid collection by advancing the combination of a drainage catheter loaded with a metal stiffening cannula and a sharp inner stylet. The distance between the skin and the collection (measured by US or CT) is marked on the catheter shaft. The trocar-based catheter is then advanced into the collection. Catheter advancement can be monitored by CT, US, or fluoroscopy. The stylet is removed and fluid is aspirated to confirm catheter position in the collection. Final positioning of the catheter is important. To lessen the chance of dislodgment, it should be placed into the most dependent part of the abscess and not superficially (see “Drainage Errors”). If the cavity is long, the catheter is positioned through the length of the cavity, or two catheters are inserted to prevent regional loculations and enhance complete drainage. We make sure none of the holes in the catheter  are outside the cavity. After the catheter is positioned, we gently evacuate most of the purulent material from the cavity and irrigate it  with saline. If the fluid becomes blood tinged, irrigation should be discontinued. When loculations are present, the catheter should not be overirrigated or overmanipulated to prevent bacteremia. Some authors have recommended the use of a wire at this point to disrupt any septations or adhesions, but this may be imprudent. Using a wire may traumatize the wall and cause a significant bacteremia or hemorrhage from the granulation tissue. We favor use of fibrinolytic irrigations to treat such cases (see “Catheter Irrigation and Fibrinolytics”). Catheters can be sutured to the skin with 2.0 or 3.0 Prolene or other nonabsorbable material for external fixation. There are several adhesive retention devices that may be used in lieu of a suture. Large-bore

2591

catheters placed in the chest or abdomen also may require skin sutures to prevent dislodgment.

CLINICAL MANAGEMENT OF CATHETERS When the catheter is in place, a few guidelines must be followed to ensure a successful drainage. In general, catheters are connected to a collection bag for drainage by gravity. This setup is more effective when the fluid drained is nonviscous. Collections with viscous fluid or particulate material need catheter irrigation and may require intermittent low wall suction. High suction may pull debris and tissue into the catheter holes and impair drainage. Active suction should be used for high-output collections (e.g., gastrointestinal [GI] or urinary tract fistula) to keep the cavity dry and promote healing. The catheter is irrigated to maintain patency. Without irrigation, catheters may occlude regardless of their size. Proper catheter irrigation involves the following steps: placement of a syringe in the stopcock and aspiration of residual fluid, injection of 5 to 10 mL of normal saline, aspiration of the irrigant, and reflushing the catheter with 5 mL of normal saline. Patients should be followed on a daily basis for catheter output, skin access site leakage or infection, and laboratory and clinical signs of complications or failure of drainage. When a drainage procedure has gone well, the patient responds quickly and shows dramatic clinical improvement. In such patients, the fever abates and the WBC count decreases. The progress of the drainage also can be assessed by following the sequential changes in the appearance of the fluid. Initially the material is cloudy and turbid; it then changes to a serosanguineous appearance, and finally it changes to a clear serous material. Cultures of the fluid also change sequentially so that when the abscess has resolved, drainage fluid is sterile. After this clinical response and the changes in character of the material, the catheter can be removed. If the patient does not respond as expected, the fluid drainage is scant, purulent fluid does not evolve as expected, or the drainage begins to show bile or enteric material, repeat imaging is indicated. Persistent large volume of fluid output, output of tube feeds, or output of fecal matter may indicate a fistulous connection. Fistulas usually are not apparent on initial postprocedure scans because the small fistulas can be sealed by inflammatory edema in the area, which temporarily occludes the fistula. As healing begins and edema reduces, the fistula may open up and change the appearance of the drainage material or the appearance on CT. A repeat CT scan should be performed to evaluate the abscess to determine if a fistula is present (shows large amount of air), if portions of the cavity are incompletely drained, or if additional cavities have developed. Evaluation of the catheter status is important to detect any displacement or kinking. A sinogram under fluoroscopy can document the presence of a fistula (Fig. 68-3). Repeat puncture and catheter insertion and upsizing or repositioning of the catheter may be necessary.

Catheter Irrigation and Fibrinolytics If the material being drained is thin and no definite septations are noted on imaging studies, we recommend flushing the catheter with saline. Saline in the amount of 10 to 15 mL should be injected every 6 to 8 hours to keep the catheter lumen clear. If the material being drained is thick or there are septations in the abscess, fibrinolytic agents may be used when the material is thick or septations are present within the abscess. Tissue plasminogen activator (tPA) can be injected into the abscess cavity every 8 to 12 hours for 3 to 4 days. Fibrinolytic agents such as tPA change the substrate plasminogen to the active fibrinolytic agent plasmin. There is no reason to suspect that its activity should be any different than urokinase. If one reviews the table of information from the urokinase study (Table 68-1), it can be seen that

2592

PART III  Image-Guided Procedures

D

A

B

E

C FIG 68-3  This 57-year-old man with a history of chronic pancreatitis and pseudocysts presented with recurrent leukocytosis. A, Coronal CT image through the abdomen demonstrates a loculated peritoneal fluid collection containing gas, with fistulous communication to the adjacent transverse colon (arrow). B, Axial CT image confirms the fistulous communication between this fluid collection and adjacent colon (arrow). C, A 5F Yueh needle was advanced into the fluid collection. D, Postprocedure CT image demonstrates satisfactory position of a 10F catheter. E, Follow-up fluoroscopy-guided sonogram 5 days later. Following contrast injection through the drainage catheter, there is redemonstration of a fistulous communication (arrow) to the adjacent colon with contrast opacification.

2593

CHAPTER 68  Computed Tomography-Guided Drainage TABLE 68-1  Results of Student’s Test for Length of Stay, Treatment Cost, Febrile Course,

Days of Elevated WBC Count, and Days of Drainage Test for

Group

Mean

t

df

P*

Length of stay (days)

Saline Urokinase Saline Urokinase Saline Urokinase Saline Urokinase Saline Urokinase

29.00 ± 17.2 13.11 ± 9.5 62,102.11 ± 42,844 24,178.00 ± 17,871 4.00 ± 6.25 1.45 ± 1.73 9.12 ± 8.79 5.62 ± 5.04 14.88 ± 20.5 6.93 ± 4.2

3.56

24

0.0025

3.49

22.3

0.0021

1.63

18.1

0.1202

1.41

25.7

0.1710

1.57

17.5

0.1361

Treatment cost ($) Febrile course (days) WBC elevation (days) Drainage (days)

From Haaga JR, et al: Intracavitary urokinase for enhancement of percutaneous abscess drainage: Phase II trial. AJR Am J Roentgenol 174:1681–1685, 2000. *Separate variance t-test used. df, degrees of freedom; WBC, white blood cell count.

TABLE 68-2  Effect of Loculation on Length of Stay and Hospital Costs Loculation

Test

Group

Mean

t

df

P*

No

Length of stay (days)

11.5

0.0168

Treatment cost ($)

2.81

11.8

0.0169

Yes

Length of stay (days)

2.13

9

0.0617

Yes

Treatment cost ($)

27.09 ± 17.8 11.04 ± 4.72 58,023.33 ± 44,075 21,567.40 ± 7797 32.50 ± 17 13.20 ± 11.81 70,259.66 ± 42,981 17,315.60 ± 6760

2.81

No

Saline Urokinase Saline Urokinase Saline Urokinase Saline Urokinase

2.97

5.3

0.031

From Haaga JR, et al: Intracavitary urokinase for enhancement of percutaneous abscess drainage: Phase II trial. AJR Am J Roentgenol 174:1681–1685, 2000. *Separate variance t-test used. df, degrees of freedom.

urokinase improved the clinical factors of the patient and shortened the drainage time and hospital stay.9,31 The amount of urokinase used depends on the size of the cavity, as follows: 0 to 3 cm, 12,500 IU; 5 to 10 cm, 50,000 IU; and greater than 10 cm, 100,000 IU. After the urokinase has been injected, a small amount of saline should be used to clear the catheter of urokinase. The catheter should be clamped for 5 minutes then left to gravity drainage. If tPA is used, the corresponding amount is administered; one should check with the pharmacy or package insert for the amount relative to urokinase. The use of tPA is not approved by the FDA but is legal according to off-label regulations. The effectiveness and safety of fibrinolytics has been documented by our group in the literature.93,127,192 Park et al. showed that urokinase decreased the viscosity of purulent material more than saline alone did in vitro.192 The safety of urokinase in vivo was proved by Laborra and colleagues, who showed that intracavitary urokinase produced no systemic effects.127 A later article by Haaga’s group showed considerable benefit for using urokinase compared with saline.93 This article demonstrated that its use improved treatment of abscesses. Its use produced earlier improvement of clinical parameters and shortened hospital stay (Table 68-2; also see Table 68-1).

Continuous Irrigation In some cases when especially thick material is present, placement of several catheters for irrigation may be beneficial. One catheter is used for inflow of saline, and one or more catheters may be used for evacuation of the fluid. In such cases, one should ensure that fluid is moving freely between the catheters so there is no fluid buildup.

Sinograms Some authors have performed catheter sinograms to follow abscess cavities. Diluted contrast is injected through the catheter slowly and under low pressure. A CT sinogram can show if cavities are undrained and if additional catheters are necessary. The study can be obtained for collections in which a fistula is suspected. During the sinogram, one evaluates whether the catheter placed is still in the correct position and is of adequate size for the evacuation of any residual purulent material. If there is considerable purulent material left, one may choose to increase the size of the catheter. At this time, sequential dilation may be performed safely without bacteremia because the pathogens are fewer as a result of the treatment. If the drainage material is thick or loculations are present, one might institute catheter irrigation with

2594

PART III  Image-Guided Procedures

fibrinolytic agents at this time if it was not used initially. With the cavity partially evacuated the contour may change and shift the location of the catheter, requiring adjustment of the position. Repositioning, replacement, or adjustment should be performed as appropriate. Fluoroscopic sinograms reveal any communication with the bowel or biliary system that should be treated with suction or additional catheters. When performing such sinograms, one should be careful not to inject the catheter and cavity with too large a volume of contrast. Excessive volume may increase the intracavitary pressure sufficiently to cause bacteremia or to rupture a weakened wall of a cavity and produce spillage. To prevent political conflict with clinical services, it is always helpful to remind clinicians that in many cases a preexisting fistula may be the primary cause of the abscess. The late appearance of a fistula during the drainage process is due to the fact that inflammatory edema in the wall of the abscess occludes a fistula temporarily. At a later time when the inflammation partially resolves, the fistula opens and manifests itself. One should not accept such fistulas as caused by the PAD unless the bowel was clearly crossed during the procedure.

CATHETER REMOVAL The decision to remove the catheter is easy if the course of the drainage (i.e., clinical response including defervescence and resolution of leukocytosis, evolution of fluid changes, and resolution of the fluid cavity on the sonograms) is as expected. Another criterion used for catheter removal is decrease in catheter output to an immeasurable amount (<10 mL/day). When drainage volume decreases to 20 to 30 mL/day, catheter irrigation is discontinued. If catheter output remains below 10 mL/day for more than 24 hours, the catheter is removed unless a pancreatic, intestinal, or urinary fistula is present. Before removal of a catheter, a final scan can be obtained to ensure that the cavity has closed completely. If this has occurred and there is no continued drainage, the catheter can be removed. An abscess that is completely resolved should show virtually complete obliteration of the cavity (but one may see a small amount of residual fluid and air). A liver abscess may be confusing on follow-up scans. An abscess may be completely drained, but an ill-defined low-density area at the site slightly higher in density than fluid may remain. This represents local fatty change or edema or both, which resolves spontaneously. If a considerable amount of fluid remains, it is best to wait until the cavity has more completely resolved and has been reduced in size by granulation tissue. If a patient is doing well clinically, he or she may be discharged and managed as an outpatient during the time it takes for the cavity to granulate closed. The patient’s antibiotics should be continued for several days after removal of the catheter. Long-term follow-up imaging may be required if the patient has a condition that is likely to recur (i.e., pancreatic pseudocysts, lymphocele, cystic tumors).

FISTULAS GI or biliary fistulas commonly occur with intraabdominal abscesses; some fistulas probably cause abscesses and some are sequelae.176 Management depends on the amount of flow, whether it is of low or high output.62 Low-output fistulas (320 mL/day) usually close spontaneously without additional therapy. High-output fistulas require treatment. High suction on the catheter should be maintained and the bowel should be put at rest. This usually means insertion of a nasogastric tube. Hyperalimentation also may be needed for several weeks  in some severe cases. Failure of these methods indicates a need for surgery.

Fluid collections may recur if the fistula has not healed by the time the drainage catheter is removed. Sometimes a fistula is seen on sinogram despite minimal drainage from the catheter, because fluid can internally drain through another route. In this situation the catheter is clamped for 1 to 3 days to allow reaccumulation of fluid. If no collection is identified by imaging, the catheter can be removed. Fluid collections associated with fistulas may require prolonged drainage. Persistent low-output drainage can be managed by downsizing the drainage tube and gradually removing it over 3 to 5 days. Presumably this technique allows collapse and closure of the tract as the catheter is pulled out.

SUCCESS RATES Since the introduction and development of PADs, there has been controversy among various groups about the achievable success rate, which overall has varied between 70% and 90%. Successful diagnostic aspiration is defined as aspiration of material sufficient for diagnosis. Curative PAD is complete resolution of infection requiring no further operative intervention. Partial success is either adequate drainage of the abscess with surgery subsequently performed to repair an underlying problem or as temporizing drainage performed to stabilize the patient prior to surgery.11 Partial success has been reported in 5% to 10% of patients. Failure occurs in 5% to 10% and recurrence in 5% to 10%.14,74,150 These results are similar for both chest and abdominal PAD procedures.123,278 The differences in success rates depend on numerous factors including relative contraindications, complexity of the collections, severity of the underlying medical problems, drainage technique, and catheter selection. Abscesses can be classified into three main groups that are organized by complexity and reflected in success and failure rates.22 The unilocular and discrete abscess is simplest to drain, and successful PAD has been reported in up to 90% to 95% of cases.34,205,275 More complex abscesses are frequently multilocular with very viscous and purulent contents or contain necrotic tissue; these factors may make complete drainage difficult even with the use of large or multiple catheters. The medium-complexity abscess (i.e., abscesses that have a communication to the GI tract) are cured in 80% to 90% of cases.233,253 Associated GI communication may require an operation, but if surgery is performed after PAD, the surgery is simplified because of the noninfected bed. The most complicated collections include intermixed pancreatic abscess/necrosis, infected tumor, and tenacious organized empyemas. The cure rate in these latter situations are no higher than around 80% and may be as low as 30% to 50%.30,238,240 PAD has also been used for postoperative intraabdominal abscesses with high technical success.121 Benoist and colleagues21 investigated the factors prognostic of PAD failure for postoperative intraabdominal abscesses in 73 patients. They found that complex abscesses, including those associated with enteric fistula, were not associated with significantly increased failure rates. Insufficient antibiotic therapy and abscess diameter of less than 5 cm were the only two independent factors predictive of PAD failure in their study. They reported that small abscesses in the failure group required repeat surgery for persistent or recurrent sepsis after removal of the drainage catheter, likely secondary to incomplete drainage. The overall success rate in this study was 85.6% after 1-year follow-up, similar to the previously reported high success rates.83,225,283 Numerous authors have studied the patient factors that affect the outcome of drainage procedures. Several major determinants have been found, including the status of the immune system, the overall severity of the patient’s illness, and underlying medical condition. Lambiase and colleagues reported a large series of patients treated with

CHAPTER 68  Computed Tomography-Guided Drainage PAD and noted a significant difference in the outcomes of their patients.129 They considered the following states as immunocompromised: absolute neutropenia, infection with human immunodeficiency virus, cancer with distant spread, chemotherapy, radiation therapy, lymphoproliferative disorders, diabetes requiring treatment, long-term renal dialysis, splenectomy, and severe alcoholism. They found that with PAD the overall cure rate for immunocompromised patients was 53.4% compared with the normal rate of 72.6% for normal immunocompetent patients in their series. Many studies have demonstrated that an objective severity-of-illness scoring system is predictive of the patient’s prognosis.98 A widely used system is the APACHE II score, which is prognostic of both mortality and morbidity in patients with intraperitoneal abscesses.124,258 Higher APACHE II scores are generally associated with high morbidity regardless of the treatment method.98,140 Levison and Zeigler correlated the APACHE II score with drainage techniques and outcome.141 These authors found that patients with scores less than 15 had a mortality rate of 1.7%, and patients with scores higher than 15 had a mortality rate of 78%. There was no benefit of one method—surgical or percutaneous drainage—over the other. There are many reasons for the failures reported, including premature withdrawal of the catheter, presence of a GI fistula, pleural contamination, hemorrhage, and the errors of selecting cases, as noted earlier. Lang reported a series of 136 patients with a failure of 29 of 119 attempted curative procedures.130 With infected pancreatic processes, Lang and colleagues were successful in 11 of 22 cases. In their series, they had 17 patients with identified GI fistulas; 9 resolved spontaneously, and 8 required surgical management. Failure to document the complete obliteration of abscesses by CT resulted in failure in 7 cases. They had 2 cases of hemorrhage associated with inappropriate placement of catheters in areas with phlegmons. The most common misdiagnosis was failure to identify loculated or residual cavities. Premature withdrawal was made in many cases because of the cessation of drainage fluid. The authors recommended that a repeat CT scan be performed before removal of the catheter.

DRAINAGE ERRORS Errors associated with abscess drainages have been documented by numerous authors. Perhaps the most common error made is the improper choice of trajectory, which can cause either a portion of the pleura to be crossed when draining a subphrenic abscess or a solid organ to be traversed. Although with today’s potent antibiotics, dissemination of an infection is infrequent, it can occur and produce significant difficulty. If an abscess is very large, one can be overly confident and attempt to hasten a procedure by “skipping” some steps, such as using CT localization. In one such patient, one of our experienced physicians used a one-stick catheter without guidance and penetrated the femoral vein. The patient did well after insertion of a vena cava filter to prevent emboli, insertion of a proper drainage catheter, and removal of the offending catheter from the vein. Removal of some purulent material is important after the initial puncture to relieve the tension of a large abscess. If this is not done, there is a chance that leakage of infected material may spill into the peritoneum during the insertion and manipulation of the wire and catheter. Relative to placement of the catheter, it is important that the largest catheter possible be inserted, not only to permit rapid drainage of the fluid but also to provide ample diameter should purulent material become inspissated on the inner wall of the catheter. This does not mean a massive catheter is required, but one at least 10F to 12F should be used. Also, the catheter should be placed in the most

2595

remote site of the cavity so that loculations do not occur distal to the location of the catheter. Inadvertent or early removal of the catheter can result in recurrence of the abscess. Prior to removing a catheter, we often obtain a CT scan to verify complete resolution of the abscess. Alternatively, if a catheter has been inserted into a sterile fluid collection and it is not removed promptly, a secondary infection of the fluid space can occur. Finally, if an abscess does not resolve with drainage or enlarges, one should suspect a fungal infection or spontaneous bleeding. It is frequently observed that the infectious agent in an abscess may change based on elimination of sensitive organisms and selection of resistant organisms. Candida species commonly becomes the dominant organism in a patient with multiorganism infection who is treated with broad-spectrum antibiotics selected for bacterial pyogens. In several cases, we have encountered an abscess that appeared to increase in size and developed an elevated density. In such cases, localized bleeding may occur because of the invasion of the local blood vessels; effective percutaneous drainage is typically impossible and surgery is preferred.

COMPLICATIONS Even the most talented and compulsive interventional radiologist will experience complications of abscess drainage, including complications of sedation, drug-related allergic reactions, cardiopulmonary complications, infectious complications, bleeding, and nontarget access. The overall procedure threshold for all major complications resulting from adult percutaneous abscess and fluid drainage is 10% according to the Society of Interventional Radiology Standards of Practice Committee.11 Published rates of complications vary widely depending on technique, types of abscess being drained, and various authors’ experience. Reported rate of major complications of abscess drainage has been under 5% in most series of patients.22 Sepsis and bleeding occur in less than 3% of cases. Death is extremely rare. Minor complications, such as secondary wound infection, pneumothorax, discomfort from the catheter, and dislodgment, occur in approximately 5% to 10% of patients.151,165,172 The most common complication in all series is that of sepsis following catheter insertion. Despite all efforts to localize collections by imaging guidance, transient bacteremia results from spillage of infected contents into the bloodstream in up to 5% of cases.96 During primary catheter placement, inherent risks include spread of abscess contents to the adjacent spaces or organs, transient bacteremia, and frank sepsis. Spread to adjacent structures or the bloodstream can result from nontarget puncture or prolonged serial fascial dilation.99 We often use IV contrast injection to improve visualization of vessels, bowel, and other structures before placing the catheter. Another measure to prevent inadvertent spread of infection during fluid drainage is avoidance of nonsterile routes of drainage into potentially sterile collections if possible.

MANAGEMENT OF INTRATHORACIC FLUID COLLECTIONS In the thorax, abnormal fluid can accumulate in the pleural space, lung parenchyma, or mediastinum from a variety of etiologies, including infection, malignancy, surgery, and trauma. Each of these entities has a unique set of associated considerations and potential complications. Percutaneous diagnostic aspiration can be performed safely and accurately, and drainage catheters can be placed for treatment of infected or otherwise symptomatic collections. US is the preferred guidance modality for the more superficial and simple fluid collections. CT is

2596

PART III  Image-Guided Procedures

opted for deeper, loculated, multiloculated, and more complex collections.123,270

Abnormal fluid in the pleural space is easily accessible by percutaneous methods. The four percutaneous procedures most commonly performed are diagnostic thoracentesis, therapeutic thoracentesis, pleural sclerosis, and empyema drainage. Complicated pleural effusions are fluid collections that require drainage to resolve. Exudate, empyema, and hemothorax are examples of complicated effusions. They are the most common indications for drainage catheter placement.174 Other indications for drain placement include malignant effusion, recurrent effusions, chylothorax, pneumothorax, hemopneumothorax, and leakage into the pleural space from esophageal or gastric rupture.

is the fibropurulent stage in which the sterile pleural fluid has become infected with bacteria and the cellularity and protein content in the pleural fluid increase. Fibrin is deposited on the visceral and parietal pleura. As this stage progresses, loculations form, the pleural fluid pH and glucose decrease, and the lactic dehydrogenase (LDH) level and the number of PMNs increase. In the third or organization stage, fibroblasts and capillaries grow into the exudate and form a pleural peel. If untreated, this stage may result in entrapment of adjacent lung or drainage of fluid through the chest wall (empyema necessitatis) or into the lung (bronchopleural fistula). Although the best method of managing an empyema is controversial, early intervention is the key to preventing pleural peel formation and development of an entrapped lung. Once a peel forms, decortication is almost always necessary, although there is evidence that some pleural peels can resolve spontaneously if the underlying empyema is treated satisfactorily.5,100

Empyema

Diagnosis.  Decubitus radiographs can be a sensitive and inexpensive

Empyema is an accumulation of purulent material within the pleural space. Despite the development of new antibiotics, empyema remains a common problem and is associated with mortality rates as high as 40%.45,224 Along with antibiotic therapy and treatment of underlying disease, early and complete drainage of infected fluid is essential in the successful management of empyema.123 Traditionally empyema has been managed by surgical placement of large (22F-34F) catheters in the pleural space. However, small drainage catheters have successfully been used in the management of empyema.27 Shankar and colleagues reported a study of 103 patients with empyema, in which 80 patients were successfully treated by placing small (<14F) drainage catheters under US or CT guidance.229 This result is comparable to those of previous studies in which larger catheters were placed for empyema treatment. The most recent prospective study by Rahman and colleagues also demonstrated that small (<14F) drainage catheters were as effective as large (>14F) catheters in the management of infected pleural fluid collections.204 There was no significant difference in clinical outcome in patients with different-sized drainage catheters. In this study, patients with smaller drainage catheters experienced much less pain than those with large catheters, which were mainly placed by blunt dissection.

method to determine if the effusion is free flowing or loculated. This will determine whether intervention is needed and by what modality it should be performed. The visualization of septations on US suggests an exudative process. However, CT allows a more thorough evaluation of the chest. This includes the ability to detect and localize pleural fluid, determine whether it is free flowing or loculated, assess the number of collections, and evaluate the underlying lung parenchyma. Analysis of the pleural fluid by routine thoracentesis determines the stage of empyema and the most appropriate management. If there is aspiration of gross pus or if the Gram stain is positive, the pleural space should be drained immediately. A positive Gram stain, even if nonpurulent fluid is obtained, is indicative of advanced disease and drainage is recommended. If the pleural fluid is not grossly purulent and does not have microorganisms on the Gram stain, then assessment of the fluid composition can be used to direct management. The most important pleural fluid components to measure are pH, glucose, and LDH levels. When the pleural space becomes infected, the pleural fluid glucose level decreases secondary to glycolysis from PMN phagocytosis and bacterial metabolism. This increase in glucose metabolism results in accumulation of lactic acid and CO2, which decreases the pleural fluid pH. The pleural fluid LDH level rise is due to cell lysis. Of these three components, the pleural fluid pH will usually decrease before the glucose level decreases and the LDH level increases. These three parameters are important predictors of whether an effusion will resolve with or without drainage. Many authors use criteria based on the combined pleural fluid data of Sahn and Light.143,215 For pleural fluid with a pH below 7.10, glucose below 40 mg/dL, and LDH above 1000 U/L, early drainage is recommended to avoid subsequent development of a complicated multiloculated effusion or a pleural peel. For free-flowing fluid with a pH above 7.30, glucose above 60 mg/dL, and LDH below 1000 U/L, the fluid will probably resolve spontaneously if appropriate antibiotics are given. For fluid with a pH between 7.10 and 7.30, it is recommended that antibiotics be given and the patient be observed closely. If needed, repeat thoracentesis can be performed.

PLEURAL FLUID COLLECTIONS

Pathogenesis and Etiologies.  Infected pleural fluid has developed in association with pulmonary infections, surgery, trauma, foreign bodies, esophageal perforation, or inflammatory processes below the diaphragm. Primary lung infection and postoperative complications are the most common etiologies of empyemas and account for up to 65% of reported cases.138 Anaerobic bacteria have replaced gram-positive organisms as the most common cause of infected pleural fluid collections in adults.133 More than one anaerobic organism is usually cultured from an empyema. Because anaerobes are fastidious and so common, culture specimens should be handled appropriately to ensure positive results. The specimens should be cultured quickly, and they should be exposed to as little air as possible because air may kill the anaerobic organisms. Although there has been a recognizable change in the organisms associated with empyema, management and clinical outcome with treatment have not been altered.45,49 Staphylococcus aureus remains the most common organism found in empyemas in children. There are three stages in the evolution of an empyema,142,215 and the stage determines management. The first stage is the exudative stage, in which there is accumulation of sterile free-flowing pleural fluid. The pleural fluid at this time contains rare polymorphonuclear leukocytes (PMNs), a normal glucose level, and a normal pH. The second stage

Indications for Chest Tube.  As stated earlier, the primary indication for chest tube placement is to drain an empyema and prevent progression of the pleural process to the organized stage. Although recent radiologic series have demonstrated fairly high success rates with percutaneous drainage, there is still controversy as to whether radiologic or surgical tubes should be used.41,288 Davies and coworkers, however, presented cases with chronically infected pleural effusions that were successfully treated by small (12F) ambulatory indwelling pigtail catheters for 8 to 21 months.50 The authors also stated that small

CHAPTER 68  Computed Tomography-Guided Drainage long-term drainage catheters were especially beneficial for patients who were not surgical candidates. Pierrepoint and colleagues compared the clinical outcome in 24 pediatric patients with empyema who were treated by surgical open thoracotomy with pleural débridement, conventional stiff chest drain catheter, or pigtail drain catheter placement.197 Pigtail catheters were placed under image guidance and were considerably smaller than conventional stiff drains. Patients who were treated by pigtail catheters or surgical thoracotomy showed better outcomes compared to those with conventional stiff drains. The outcome included shorter drain time, earlier clinical improvement, earlier resolution of fever, and earlier discharge. Pigtail catheter placement was also associated with less pain and discomfort; the authors concluded that pigtail drainage catheters were preferable to thoracotomy. Indications for imaging-guided pleural catheter placement include failure of surgically placed chest tubes, collection too small for safe surgical tube placement, debilitated patients who are too ill to undergo an operative procedure, and multiple loculated fluid collections.251 In general, recurrent chronic pleural infection is considered a contraindication for long-term indwelling catheter placement. The treatment of choice therefore has been open surgical drainage of infected fluid.

Percutaneous Technique.  Prior to the procedure, patients should receive appropriate antibiotics. If the pleural fluid is free flowing, a diagnostic thoracentesis can be performed with US guidance. If there is gross pus or a positive Gram stain, a drain can be placed. However, many pleural fluid collections are already loculated when the patients are referred for percutaneous drainage. We prefer to use CT guidance both to perform diagnostic aspiration and for placement of drainage catheters in loculated collections. CT provides optimal visualization of all loculations, differentiates between collapsed lung and pleural fluid, and allows visualization of the needle, guidewire, and catheter during the drainage procedure (Fig. 68-4). Once the loculated collection is localized, the patient is placed in a position that allows optimal access. A diagnostic thoracentesis is performed. Because CT provides complete visualization of the collection, the trajectory of the needle can be planned such that the entire collection can be traversed. If the collection is located posteriorly, we attempt to use an approach that is as lateral as possible. Using a lateral approach for drainage of a posteriorly located empyema allows the patient to later lie supine without lying on the catheter and potentially kinking it. A direct posterior approach can be used, but the catheter may become kinked when the patient lies down and limit the drainage. Drainage performed with US may be limited to a posterior approach for posteriorly located collections. An important point is to use a more lateral approach in preference to a posterior approach. For CT-guided drainages, we use a 5F Yueh catheter or equivalent catheter-needle because even very thick purulent fluid can be aspirated with little difficulty. If drainage is desired, a catheter can be placed using the Seldinger technique. We generally use the Seldinger technique to introduce the catheter into the pleural space. Single-step catheters with trocars can be used, but the Seldinger technique provides better control and reduces the chance of injury to the underlying lung. Inadvertent entry of air during dilation of the tract is not a problem, because the collections are usually loculated and the catheter will be connected to a Pleur-evac (Teleflex Medical, Research Triangle Park, NC) and wall suction. Viscosity of the fluid determines size of the catheter. Most collections can be drained with 8F to 10F catheters. It is sometimes technically difficult to introduce a catheter into the pleural space because the catheter or the guidewire may buckle. Adequate dissection of the subcutaneous tissue along the needle tract and use of a stiff guidewire

2597

(Amplatz guidewire, Boston Scientific, Marlborogh, MA) can help prevent this buckling. Occasionally a 12F to 14F catheter will be needed to drain viscous fluid. Insertion of a large catheter can be difficult without fluoroscopic guidance. We rarely use a catheter larger than 14F; instead we instill tPA into the pleural space (see later discussion). Apparent unilocular collections frequently contain multiple thin septations that cannot be visualized on CT images. These septations become apparent only when a small amount of fluid is aspirated from a catheter positioned in the middle of the collection. Despite being fine and delicate, these thin septations are probably just as responsible for failure of percutaneous drainage as the viscosity of the fluid. If multiple thin septations are present within a fluid collection, it is helpful to have the needle traverse as much of the collection as possible before removing the sharp stylet. The sharp needle point punctures the septations, subsequently allowing a single catheter to drain many of the loculations. Although additional mechanical disruption of the septations can be performed, we prefer to instill tPA to dissolve these septations. To achieve optimal drainage, it is helpful to have the pigtail portion of the catheter in a dependent position within the fluid. Nephrostomytype catheters with side holes in the distal pigtail are the type of drainage catheters most frequently used. Catheters with additional side holes more proximal to the pigtail (biliary type) can be used, but care must be taken to ensure that all of the side holes remain within the pleural collection. Once the catheter is placed, it can be sutured to the skin or attached with adhesive fixation devices (Percu-Stay, Derma Sciences, Inc., Princeton, NJ). Continuous low wall suction is used for thoracic collections. Chest drainage catheters are connected to a Pleur-evac connected to wall suction at −20 cm of water. The Pleur-evac is a waterseal device with a fluid collection chamber and a safety mechanism to prevent excessive suction. The patient is continued on antibiotics tailored to the culture and sensitivity results of the fluid sample. The pleural fluid output is monitored daily. Assuming there is good drainage, the tube position can be checked every 2 to 4 days with chest radiographs. If there are any problems, a repeat CT can be performed to check tube position. Additional drainage tubes can be placed as needed for any separate collections not being drained. The drainage catheter can be removed when the fluid becomes serous, output decreases to less than 20 mL per 24 hours, the patient is defervesced, and the WBC count is normal. A repeat CT should be performed prior to catheter removal to ensure there are no undrained collections. When empyema drainage is performed, it is important to work closely with the thoracic surgeon. Not all collections are amenable to percutaneous drainage. Patients who develop a pleural peel or have persistent fevers and elevated WBC counts despite 48 to 72 hours of adequate drainage and appropriate antibiotics may require surgical drainage and decortication. Surgical treatment for such patients should not be delayed.

Results.  A successful outcome is defined by drainage of the abscess without need for a more invasive procedure such as a video-assisted thorascopic surgery, surgical thoracostomy, or thoracotomy.280 Success rates for image-guided drainage of empyema or complicated para­ pneumonic effusions range from 70% to 94% in retrospective studies.70,102,168,252 With the use of intracavitary fibrinolytics in more complex cases with empyema or hemothorax, a success rate up to 92% has been reported.111 These success rates compare favorably with success rates of 35% to 75% for surgically placed chest tubes, although it is difficult to compare the radiologic and surgical series directly because the patient populations were so variable in these studies.

2598

PART III  Image-Guided Procedures

A

B

C

FIG 68-4  This 67-year-old woman with lung cancer presented with worsening shortness of breath and leftsided pleuritic chest pain. A, Axial CT image demonstrates a large amount of left-sided pleural effusion. Diagnostic thoracentesis prior to this procedure resulted in findings consistent with malignant pleural effusion. B, Axial CT image demonstrating 5F Yueh needle advanced through the lateral chest wall toward the dependent aspect of the pleural fluid. C, A 10F catheter (arrow) was placed and resulted in complete drainage of the left-sided pleural effusion.

Most authors agree that early drainage is important. A high percentage of early-stage empyemas with free-flowing fluid can be treated with either surgically or percutaneously placed chest tubes. Failure of chest tube drainage is usually due to thick viscous fluid and loculations. Such collections resist even large-bore chest tubes, but it appears that greater success can be achieved with adjunctive use of intrapleural fibrinolytics. Once a pleural peel develops, decortication is usually required, although there are some authors who suggest that some pleural peels may resolve spontaneously after treatment of underlying empyema (Fig. 68-5).137 In debilitated patients who develop pleural peels, conservative management can be tried, and in some of these patients decortication may be avoided. However, for otherwise healthy patients, the surgical literature suggests that decortication should not be delayed when there is persistent pleural sepsis in the presence of a pleural peel.9,200 In the surgical literature, rib resection, limited thoracotomy, and thorascopic débridement have all been advocated as the treatment of choice for empyema.207 Currently there is a trend toward early decortication. Ashbaugh reviewed the effects of delayed surgical treatment

and the choice of operation on the morbidity rate in patients with empyema.8 He concluded that decortication is the procedure of choice when tube drainage fails. Decortication had the lowest morbidity and mortality rates when compared to tube or open surgical drainage. The timing of decortication is important. The fibroblastic pleural peel usually organizes by 8 weeks, at which time a decortication can be performed. There is also an early window during the first 2 to 3 weeks before the fibroblastic reaction begins when decortication can be performed easily. However, between 3 and 6 weeks, the fibroblastic membrane is poorly developed and adherent to the lung. Attempts to perform decortication at this time can lead to tearing of lung tissue and development of a bronchopleural fistula.36

Complications.  Serious complications from “blind” insertion of surgical chest tubes have been reported. These include lung laceration, diaphragmatic and intraabdominal organ injury, and neurovascular injury.163,164 None of these complications has occurred in the radiologic series. Complications associated with percutaneous drainage include pneumothorax, hemothorax, intercostal neurovascular bundle injury,

CHAPTER 68  Computed Tomography-Guided Drainage

2599

to treat any undrained loculated collections. Even if an empyema does not completely resolve after percutaneous placement of a drainage catheter, a critically ill patient may be stabilized enough to permit a more elective surgical procedure.

Malignant Pleural Effusion

Pathogenesis.  Malignant pleural effusion is the second most

A

B

common cause of an exudative pleural effusion in patients older than 60 years, with an annual incidence of 1.5 million cases in the United States.144 Pleural involvement may occur with any type of cancer, but the most common etiologies are lung and breast cancer, which account for up to 75% of cases.20 Other malignancies associated with malignant pleural effusion include lymphoma, ovarian cancer, mesothelioma, and GI cancer. The primary site of malignancy is unknown at the time of initial diagnosis in about 7% of patients.247 Approximately 50% of patients with disseminated malignancy eventually develop malignant effusion in the course of the disease.20 Presence of malignant effusion signifies advanced malignancy from the aspects of cancer staging. The 30-day mortality is approximately 29% to 50%, with a dismal median survival of approximately 3 to12 months, with the exception of breast cancer.208 The etiology of malignant pleural effusion is not fully understood, but it is thought to be secondary to obstruction of lymphatic channels draining the parietal pleura, atelectasis of adjacent lung, or overproduction from diseased pleura. Malignant effusions may also result from systemic effects of tumor.216 It is associated with significant morbidity and mortality. In patients with malignant pleural effusion, dyspnea from ventilator restriction is the most common symptom that compromises their quality of life. Other clinical symptoms include cough and chest discomfort.182

Management Options.  With the diagnosis of malignant pleural

C FIG 68-5  This 38-year-old man with a recent history of pneumonia presented with worsening leukocytosis and pleuritic chest pain. A, Pre­ procedure CT reveals a loculated fluid collection with thick wall, concerning for an empyema. Note enhancing areas of irregularly thickening pleura suggestive of pleural peel (arrows). B, A 10F pigtail drainage catheter was placed. C, Follow-up CT in 16 days for decreased catheter output demonstrates satisfactory position of the catheter and residual empyema, consistent with failed percutaneous drainage. Patient underwent video-assisted thoracoscopic (VATS) drainage of empyema with decortication and pleurectomy.

perforation of major organs and arteries, subcutaneous emphysema, and pulmonary edema due to reexpansion of the underlying lungs.51 Other reported complications including bronchopleural fistula17 and transgression across the diaphragm or into the lung parenchyma have been described.210

Conclusions.  Early drainage is very important in the management of an empyema. Imaging-guided chest tube placement is becoming the method of choice because of its accuracy, safety, and patient acceptance. It can also be used as an adjunct to surgically placed chest tubes

effusion, palliative therapy should always be considered because of the poor prognosis of a vast majority of these patients. The main goal of management is rapid and effective relief of symptoms, with minimal discomfort or inconvenience and minimal disruption of daily activity. Traditional treatment options for malignant effusion include repeated thoracentesis, chest tube drainage with pleurodesis, pleuroperitoneal shunt, pleurectomy, and thoracoscopy.198 Because each approach has unique benefits, risks, and burdens, effective therapy decisions depend on patient preference and a multidisciplinary approach to meet each patient’s specific needs. Therapeutic thoracentesis is the initial approach for immediate symptomatic improvement. Caution should be taken if removing more than 1.5 L on a single occasion, owing to risk of reexpansion pulmonary edema. Reexpansion of the lung against the chest wall after thoracentesis improves the outlook for a successful pleurodesis. Thoracentesis only provides short-term relief of symptoms and requires repeated visits to the hospital for the procedure. Effusion and respiratory symptoms recur in most patients within 30 days after therapeutic thoracentesis.7 Repeated thoracentesis is not an optimal option for palliative management of rapidly reaccumulating malignant effusion, because it can result in pneumothorax, empyema, and intravascular volume and protein depletion. Therefore thoracentesis may be an option only for patients with expected response to systemic therapy, a very short life expectancy, or slow reaccumulation of effusion. Medical thoracoscopy and surgical thoracoscopy with videoassisted thoracic surgery (VATS) allow obtaining a pleural biopsy sample, draining the effusion, and pleurodesis.125 Pleurodesis is a procedure during which a sclerosing agent is instilled into the pleural space to cause chemical irritation, leading to pleuritis and ultimately fibrosis and obliteration of the pleural space.209 Shaw and Agarwal’s meta-analysis of 36 randomized controlled studies that included 1499

2600

PART III  Image-Guided Procedures

patients reported that pleurodesis with different sclerosing agents was more efficacious compared with instillation of isotonic saline or drainage of pleural fluid alone.230 Multiple agents have been used for pleurodesis, including tetracycline compounds, doxycycline, bleomycin, and talc slurry. However, no statistical difference between these agents has been found, and the average success rates for pleurodesis ranges between 60% and 90%.6,247,295 The main limitation of this management option is long-term hospitalization (average, 5-7 days), which is not desirable or cost-effective as palliative management in patients with a short life expectancy.242,257 As an alternative treatment option, an indwelling tunneled drainage catheter has been successfully used for long-term management of malignant pleural effusion. Tunneled drainage catheters have several advantages over other treatment options. They are safe, comfortable, more cost-effective, and make long-term outpatient management of effusion and symptoms possible.256,257 In 1997, the FDA approved the PleurX catheter (Denver Biomedical, Golden, CO) for management of malignant pleural effusion. The tunneled catheter is 66 cm long, 15.5F in diameter, and made of silicone rubber that is soft and comfortable for long-term use. There are fenestrations along the distal 24 cm of the catheter that allow easy drainage of the fluid and low risk of obstruction. The polyester cuff at the proximal portion of the catheter helps prevent catheter dislodgment, tunnel or pleural infection, and pericatheter fluid. Many studies show that the PleurX can provide immediate relief of dyspnea in 81% to 100% of patients, with relief lasting for at least 1 month in 90%.182,199 Approximately 40% to 75% of patients managed with tunneled catheter experience spontaneous pleurodesis within 4 to 6 weeks, and resolution of malignant pleural effusion allowed removal of the catheter before death in 60% of patients.181,183,281 Chemical pleurodesis remains a viable option with indwelling pleural catheters. The complication rate is low and includes catheter infection, obstruction, tension pleural effusion, empyema, and tumor spread along the catheter track.109,201,276

Long-Term Tunneled Pleural Drainage Catheter Procedure.  Tunneled pleural drainage catheter placement can be done in an ambulatory setting under conscious sedation. The main advantages of the PleurX are easy placement, long-term control of the effusion, low cost, and potential simultaneous treatment of bilateral malignant pleural effusions. The procedure is similar to that of nontunneled catheter placement for the most part. After sterile preparation and draping, local anesthesia is achieved using lidocaine. Through a small skin incision, an 18-gauge needle is inserted into the pleural space along the superior margin of the low rib slightly above the diaphragm under US guidance. A lateral approach with slight posterior direction of the needle is usually recommended to position the catheter tip in the posterior costophrenic angle for free-flowing fluid. From 5 to 8 cm anterior and inferior to the initial entry site, the catheter is tunneled through the subcutaneous space. The needle is then exchanged over a 0.035-inch Newton J or 0.035-inch Bentson (AngioDynamics, Queensbury, NY) guidewire for a dilator and the tract is serially dilated to 16F. The dilator is again exchanged for a 16F peel-away sheath through which a tunneled catheter is advanced into the pleural space. The initial pleural puncture site and tunnel exit site are closed using a suture and covered with sterile dressing. A final fluoroscopic spot image or routine chest radiograph should be obtained to evaluate the catheter position, residual fluid, pneumothorax, and underlying lung parenchyma. After the procedure, up to 1.5 L of fluid can be safely removed using wall suction to achieve immediate symptomatic relief without reexpansion pulmonary edema. The catheter is subsequently connected to the special vacuum bottle system.

LUNG ABSCESSES Pathogenesis A lung abscess develops when a localized area of parenchymal infection progresses to necrosis and cavitation. Primary lung abscesses that occur in healthy patients without underlying lung abnormalities are commonly caused by oropharyngeal aspiration due to alcoholic stupor, general anesthesia, seizure disorders, or cerebrovascular accidents.170,184 Other predisposing factors for aspiration include a hiatal hernia, esophageal dysmotility, and gastroesophageal reflux. Pneumonia is the other common cause of a primary lung abscess. Nosocomial pneumonias tend to be mixed with anaerobic bacteria and aerobic gramnegative bacteria and frequently develop in debilitated patients. Nosocomial pneumonias also have higher morbidity and mortality rates compared with community-acquired pneumonias. Secondary lung abscesses result from predisposing conditions such as congenital lung abnormalities, obstructing malignancy, septic emboli and infarction, foreign bodies, and structural defects such as bronchiectasis and abnormal mucociliary mechanisms.101,248 Although lung abscess formation is more commonly associated with anaerobic pneumonias, some aerobic organisms can also cause lung necrosis, including Klebsiella pneumoniae, S. aureus, and various Pseudomonas spp. Regardless of etiology, lung abscesses often arise within the periphery of the lung.16 The inflammatory process surrounding an abscess frequently extends to the visceral pleura and then to the parietal pleura, where it can cause a focal pleurodesis and a lesion-pleural symphysis.290 This pleural symphysis is free of aerated lung and permits percutaneous access to a lung abscess, with minimal risk of pneumothorax or contamination of the pleural cavity.264,282

Diagnosis Clinical symptoms include fevers, chills, fatigue, cough, putrid sputum, and pleuritic chest pain. Diagnosis of lung abscess is made based on clinical symptoms, physical examination, and radiographic findings. Imaging plays an important role in the diagnosis and management of a lung abscess. Chest radiographs can be used to monitor areas of consolidation for possible thick-walled cavity formation with air-fluid level. However, this can be seen with pulmonary infarction, necrotic tumor, infected bulla, pulmonary sequestration, and vasculitis. Therefore contrast-enhanced CT is the imaging procedure of choice once a cavity is identified or if there is strong suspicion for the presence of small subtle cavities. CT can accurately delineate the lung parenchyma and pleura, thus differentiating between empyema and lung abscess. Typical CT findings of lung abscess include a round thick-walled cavity with irregular inner margins and air-fluid levels in areas of lung parenchyma. In contrast, empyema is usually a lenticular-shaped lesion with smooth walls of uniform thickness. Additionally a lung abscess forms a sharp acute angle with the pleura, in contrast to the obtuse angle formed by an empyema. Finally, a lung abscess destroys lung parenchyma, resulting in abrupt termination of bronchial trees and pulmonary vessels leading up to the outer margin of the lesion. Conversely an empyema compresses adjacent lung and displaces lung markings. The difference is vital to directing appropriate management. Although early drainage is important for treating empyemas, medical management is preferred for lung abscess. CT is also used to determine the number of collections, identify the abscess’s position relative to the visceral pleura, vasculature, and airways, and evaluate the pleural space for concomitant empyema. Finally, a careful review of the airways for an obstructing mass such as a lung carcinoma is essential when evaluating a CT scan in a patient with a lung abscess. Appropriate samples from lung abscesses are required to obtain accurate culture results and determine the correct antibiotic regimen.

CHAPTER 68  Computed Tomography-Guided Drainage Deep sputum expectoration can be used, but transtracheal aspirates have been shown to be more accurate.86 Although bronchoscopy may have a role in making an initial diagnosis, particularly in young children unable to cough, it has a more limited role in adults. Because it can open occluded bronchi, bronchoscopy may be most helpful for those patients who cannot expectorate sputum or in patients with an infected cavitary neoplasm.184 The usefulness of bronchoscopy in adults is limited because the samples obtained may still be contaminated.86 Contamination can be a problem in patients with chronic bronchitis, bronchiectasis, or bronchial neoplasm. These patients can have relatively large numbers of bacteria in the lower respiratory tract although they are not clinically infected.86 Percutaneous transthoracic aspiration has the advantage of obtaining culture samples with almost no risk of specimen contamination.155,173

Medical Management Although empyemas usually require some form of drainage to resolve, 80% to 90% of lung abscesses will resolve with broad-spectrum antibiotic therapy.195,249 The antibiotic regimen usually covers S. aureus and anaerobes and should be tailored to the organisms present. In cases  of aspiration-associated lung abscess or if the patient is immuno­ compromised, the antibiotic regimen should include anaerobic and gram-negative organism coverage.65 The typical duration of medical treatment ranges from 1 to 3 months. In addition to antibiotics, drainage of an abscess via the airways by deep sputum expectoration is important for cure, particularly in children. Kosloske et al.126 recommended that lung abscesses be treated with appropriate antibiotics and chest physiotherapy in those children who are able to cough up deep sputum. In those patients unable to cough sufficiently (usually children < 7 years) or in those patients who are deteriorating, percutaneous drainage or surgical resection is recommended. From 11% to 20% of patients with lung abscesses eventually require surgical or percutaneous drainage owing to failure of conservative therapy.60 Signs of treatment failure include persistent fevers and leukocytosis, an absence of improvement on imaging, or development of complications such as an empyema, hemorrhage, or hemoptysis. Several poor prognostic factors are immunocompromised state, advanced age, large abscess cavity (>4-6 cm), underlying malignancy, reduced level of consciousness, or infection with certain aerobic pathogens (K. pneumoniae, Pseudomonas aeruginosa, S. aureus).184

Surgical Management Surgical alternatives for lung abscesses include lobectomy, wedge resection, and surgically placed chest tubes. In the preantibiotic era, lung abscesses were drained by thoracotomy or percutaneous chest tubes.290 With the use of modern antibiotics, most lung abscesses are now successfully treated medically. Surgical methods are required only in the 10% to 20% of patients in whom medical therapy has failed. Indications for surgical management include significant hemoptysis, suspected underlying malignancy, massive necrosis, and delayed closure of an abscess (i.e., >6-8 weeks of adequate antibiotic therapy).47,231 A more limited segmentectomy may be performed; however, this may violate the infected lung parenchyma, thus increasing the chance of pleural contamination. Possible complications of lobectomy include hemorrhage, bronchopleural fistula, and contamination of the contralateral lung.40,100 Surgery is usually associated with high mortality and morbidity.46

Image-Guided Percutaneous Management Surgical and percutaneous drainage are both effective treatment options in patients who fail medical therapy. However, image-guided

2601

percutaneous drainage is a less invasive procedure and is now considered the first therapeutic choice, especially in those with multiple medical comorbidities. The success of percutaneous chest tubes in the treatment of critically ill patients with lung abscess unresponsive to conservative management dating back to the 1980s has been the catalyst for a trend toward managing lung abscess with image-guided percutaneous chest tube drainage. The success rate of CT-guided drainage of a lung abscess is up to 90%.60,103,157,282 Persistent sepsis can be rapidly eliminated by catheter drainage and evacuation of the pus. Cross-sectional imaging defines the presence, location, and size of an abscess. In addition the abscess-pleural symphysis can usually be well delineated, identifying a percutaneous access site into the abscess. Use of imaging guidance also facilitates drainage of small abscesses, even in pediatric patients. The indications for placement of a drainage tube into a lung abscess under imaging guidance are essentially the same as for surgical resection: persistent sepsis after 5 to 7 days of appropriate antibiotic therapy, abscess 4 cm or greater in diameter  and under tension or enlarging, or failure to wean from a ventilator owing to a large abscess.206 Other indications include the need for obtaining a specimen in infants and children who are either too young or too debilitated to produce adequate sputum.134,147 For infants and small children, percutaneous aspiration of lung abscesses has been advocated as both diagnostic and therapeutic.47 In children younger than 8 years, early intervention is recommended because the airway may be too small to adequately drain the abscess even with an intact cough mechanism.37 Contraindications to percutaneous drainage include the inability of a patient to cooperate, an irreversible and severe coagulopathy, a major blood vessel or nerve that impedes proper placement of a drainage catheter, and lack of an abscess-pleural symphysis. Most patients with intact cough mechanisms and patent bronchi should be cured with conservative medical management. As with an empyema, it is important to work closely with the thoracic surgeons.

Technique.  We prefer to use CT guidance for all lung abscess drainages (Fig. 68-6). The abscess is first localized and measured, and then the abscess-pleural symphysis is delineated. If no pleural symphysis is visualized, it is preferable to defer percutaneous drainage. Central abscesses are usually best managed surgically. Some authors also have recommended positioning the patient such that the contralateral noninfected lung is in a nondependent position to avoid endobronchial spread of purulent material during the drainage procedure.123,265 However, lung abscesses requiring percutaneous drainage often do not have adequate drainage through the bronchial tree, which is why the medical therapy failed. Therefore the normal lung can occasionally be placed in a dependent position as long as the patient is not coughing up sputum, since the chance of soiling the normal lung through the bronchial tree is minimal. The procedure is usually performed under conscious sedation using fentanyl and midazolam. In pediatric patients, general anesthesia is required. Once the trajectory of the needle is determined, the patient is prepared and draped in standard fashion. The Seldinger technique is then used to place a 10F or 12F locking pigtail catheter. For very large abscesses, a single-step catheter on a trocar can be used, but Seldinger technique over a guidewire is usually recommended because it generally entails less pain, less discomfort, and has a lower rate of complications. We prefer the control of the Seldinger technique for all smaller abscesses or for abscesses located near vital organs. One should not be too vigorous with catheter or guidewire manipulations within the abscess cavity while performing catheter insertions or exchanges, since lung abscesses may have fine strands of residual normal parenchyma that can bleed.

2602

PART III  Image-Guided Procedures

A

followed by guidewire insertion and tract dilation under fluoroscopy can be performed. The drainage catheter should be secured to the skin using suture and covered using a sterile dressing. Once the catheter is within an abscess, a postprocedural CT scan should be obtained to confirm optimal position of the drainage catheter and evaluate possible complications. The drainage catheter can be placed to gravity drainage, a self-contained suction system such as a Jackson-Pratt (JP) drain, or  the standard water seal (Pleur-evac, Teleflex Medical, Research  Triangle Park, NC) with wall suction at − 20 cm of water. Follow-up should be performed with chest radiographs either daily or every other day during the first week and less frequently thereafter until the cavity is closed.

Postprocedure Management.  The drainage catheter should be flushed with 10 mL of sterile saline every 8 hours. Twenty-four-hour drainage outputs are recorded, and WBC counts and vital signs are monitored daily. The catheters can be removed either blindly or under fluoroscopic visualization when the output decreases to less than 20 mL per day, the fluid becomes serous, the cavity closes, and the patient remains afebrile with normalization of the WBC count. Failures may occur with multiloculated abscesses, infected tumor cavities, or chronic thick-walled abscesses.184,195,279 Fibrinolytic agents such as tPA can be instilled into the abscess if needed.

Results.  The average time to abscess resolution has been reported as B

C FIG 68-6  This 17-year-old young woman with cystic fibrosis presented with leukocytosis and chest pain. A, CT of the chest demonstrates right upper lobe lung abscess. A 5F Yueh needle was introduced using an anterior intercostal approach through the abscess-pleural symphysis. B, Axial CT demonstrating a drainage catheter in satisfactory position without evidence of pneumothorax. C, Follow-up CT demonstrated resolution of the abscess. Now revealed is underlying congenital pulmonary airway malformation (arrows).

US and fluoroscopy can also be used to guide aspiration and drainage of lung abscesses. Yang et al.290 used US to guide aspiration of lung abscesses in 31 of 33 patients. The abscess-pleural symphysis was visualized in 25 patients, and only 2 small pneumothoraces developed. Nonvisualization of some abscesses because of gas and the inability to image an entire abscess and the surrounding structures are limitations of US. Fluoroscopy is limited in its ability to visualize the abscesspleural symphysis. However, a combination of US-guided aspiration

10 to 15 days, with improvement in fever and leukocytosis within 48 hours. Success rates for imaging-guided percutaneous lung abscess drainage range from 67% to 100%.13,87,195,264 PAD failure may occur with multiloculation and organized tissue. Various host (immunosuppression, malignancy, and malnutrition), microbiological, and anatomic factors can also influence treatment outcomes.184 Aerobic organisms are usually found in hospital-acquired lung abscesses and are associated with poor outcome. This may be attributed to more virulent organisms in hospital settings, as well as to underlying patient comorbidities.101 Limited data are available for percutaneous drainage of fungal abscesses, but they often result in lower clinical success rates despite technical success. Poor outcome prognosis includes complex collection, underlying malignancy, and admission to the intensive care unit (ICU).273 Reported complications include bacteremia, pneumothorax, pyopneumothorax, empyema, partial collapse of the lung parenchyma, and bronchopleural fistula.60 Although critics of percutaneous chest tube drainage refer to the possible complications of pleural cavity contamination and bronchopleural fistula formation, these complications have been very rare, ranging from 0% to 4% in most series.175,264,291 This is probably due to the use of the abscess-pleural symphysis to minimize leakage of abscess fluid into the pleural space. Additionally, cardinal technical rules of PAD are not to violate normal lung en route to the abscess and to  use the shortest trajectory to the collection that crosses the least number of pleural spaces to minimize the risk of pneumothorax and cross-contamination.

Conclusion.  Despite current medical management techniques, lung abscesses are associated with high mortality rates. Surgical resection has been the standard alternative treatment for patients who have failed conservative therapy in the past. However, percutaneous transthoracic tube drainage has become the intervention of choice for lung abscesses in recent years. The use of CT guidance to direct drainage tube insertion minimizes complications such as bronchopleural fistula, pneumothorax, and contamination of the pleural space.

CHAPTER 68  Computed Tomography-Guided Drainage

PERICARDIAL FLUID COLLECTION Pathogenesis and Diagnosis Pericardial effusion accumulation may be due to malignancy, postsurgical or posttraumatic leakage, infections, autoimmune pericarditis, or cardiac and renal failure. Effusions may be acute, subacute, or chronic. The leading cause of pericardial effusion is a sequela of cardiac surgery, occurring in as many as 85% of reported cases.117 The incidence varies depending on the type of operative procedure, heart transplantation having the highest incidence.259 Malignancy-related effusions are the second most common cause of pericardial effusions.156,260 Metastases from lung, breast, and hematologic malignancies account for 75% of all malignant pericardial effusions. However, oncology patients can have nonmalignancy-related pericardial effusions associated with either prior radiation treatment or idiopathic etiology. Establishing the etiology of a pericardial effusion in a cancer patient has prognostic and therapeutic implications because it defines advanced-stage disease, poor prognosis, and decreased survival.152,158

Management In up to 60% of cases, the pericardial effusion is associated with a known cause, and the essential treatment is that of the underlying disease. Viral or idiopathic forms often respond to empirical therapy with aspirin or a nonsteroidal antiinflammatory drug (NSAID). For management of recurrent inflammatory cases, the combination of aspirin or an NSAID plus colchicine is considered.56,104,105 Although most cases of pericardial effusion are asymptomatic and resolve spontaneously or with conservative therapy, 1% to 6% of cases become clinically significant and require intervention.48,106,260 Indications for drainage include diagnostic aspiration, large effusions that cause clinical deterioration, hemodynamic compromise (cardiac tamponade), or suspicion of purulent, tuberculous, or neoplastic pericarditis.148,153 The volume known to affect cardiac hemodynamics ranges from 100 to 200 mL in rapidly developing effusions and from 1000 to 2000 mL in slowly accumulating pericardial effusions.227,262 The contraindications to pericardiocentesis include aortic dissection, uncorrected coagulopathy, and thrombocytopenia (<50,000/mm3). Both surgical and percutaneous options are available for management of pericardial effusion. Percutaneous procedures are less invasive and associated with lower postprocedure morbidity compared with surgical drainage and therefore considered to be the initial management choice.158 Prior to the use of imaging-guided methods, pericardiocentesis was associated with significant complications, including perforation of the heart, laceration of the coronary vessels, arrhythmias, bleeding, and death.32, 285, 286 The development of two-dimensional echocardiography in the 1970s allowed better localization of the pericardial effusion and improved safety of percutaneous pericardiocentesis.33 However, it was not considered as definitive treatment because of high rates of recurrence. Therefore concomitant echo-guided catheter drainage placement with pericardiocentesis was introduced in an attempt to reduce recurrence.260,261 However, posterior and loculated effusions may be inadequately visualized using an echocardiogram. Echocardiographic guidance can be limited by the presence of artifacts and difficulty visualizing posterior and loculated effusions.145,259 Interventional radiologists have a technical advantage because of our familiarity with and access to multiple imaging modalities for image guidance. Fluoroscopic and/or sonographic guidance for pericardiocentesis have been advocated.55,69,244 Those authors using US or a combination of US and fluoroscopy have had fewer complications and no procedure-related deaths when compared with echocardiogram guidance. However, use of US guidance is limited in the postoperative period because of mediastinal gas, surgical gas, or posterior

2603

loculations.55 We have found CT to be a safe and effective modality to guide aspiration and drainage of pericardial effusions.28,29 CT allows visualization of an effusion and its relationship to the heart, lungs, and other vital structures such as the liver and internal mammary arteries. It may be particularly helpful in visualizing infected collections containing air bubbles and loculated effusions.26,55,237 Surgical pericardial window formation is usually performed in patients whose effusions do not respond to percutaneous drainage. Indications for surgical intervention include cytologic confirmation of suspected malignancy in recurrent effusions, management of malignant effusions that do not respond to routine therapy in patients with otherwise limited tumor extent and favorable survival time, and benign effusions that recur rapidly following needle drainage.139 Additionally, surgical drainage is preferred in acute traumatic hemopericardium and purulent pericarditis. Surgical drainage is commonly performed with a subxiphoid or transthoracic pericardial window technique. Other techniques include video-assisted thoracoscopic (VATS) fenestration and balloon pericardiotomy.171 In patients who undergo surgery, success can be expected in more than 90% of cases; however, postoperative complications may also occur in as many as 65% of these patients.71 Some surgeons prefer to perform a total pericardiectomy, particularly when the estimated survival time of the patient is greater than 1 year.226

CT-Guided Percutaneous Drainage

Technique.  The decision of treatment and its timing is predominantly based on the clinical situation of the patient.46,265 The density of pericardial effusion as measured by CT may become an indicator  of the feasibility of a successful drainage. Once a decision is made  to pursue a CT-guided PAD, preprocedural diagnostic images are reviewed to assess the size and location of the effusion, determine needle access path, and plan final catheter position. The patient is positioned in the CT gantry so that the trajectory of the needle avoids the lung, internal mammary arteries, and heart. An intercostal approach similar to that used in sonographically guided pericardiocentesis is preferred. An approach perpendicular to the pericardial effusion increases the chances of the needle point striking the heart, especially if the effusion is small. Therefore an obtuse angle of approach to the fluid is safer. After skin disinfection, 1% lidocaine is administered subcutaneously and along the planned needle trajectory for local anesthesia. Vital signs including blood pressure, heart rate, and blood oxygen saturation are monitored by a patient monitoring system. CT is performed to confirm the direction and depth of the approach with the anesthetizing needle kept in place. A small skin incision is then made. We place a 5F Yueh sheathed needle into the pericardial space, with intermittent CT acquisition to confirm needle depth and trajectory. With gentle aspiration, the sheathed needle is advanced through the pericardium. The position of the needle tip within the pericardial cavity is confirmed by aspiration of a small amount of fluid and if necessary CT. The sheath is advanced into the pericardial fluid over the core needle. Fluid samples are sent for microbiological culture and cytologic analysis. If drainage is desired, a 0.035-inch Bentson wire is advanced into the pericardial space through the sheath under CT and/or fluoroscopy guidance. The sheath is removed and the track is dilated with standard dilators (6F-10F) advanced over the wire. An all-purpose drainage catheter (6F-12F) is advanced over the wire into the pericardial sac, and the wire is removed. After confirmation of the catheter position and aspiration of the pericardial effusion fluid, the pigtail is deployed. The effusion is completely aspirated, in most cases by hand or by gravity. If the effusion is large, the initial drainage should be limited  to 200 to 400 mL during the first 20 minutes to decrease the risk of

2604

PART III  Image-Guided Procedures

compromising cardiac hemodynamics.69 The remainder of the fluid can be removed gradually by gravity drainage. The catheter is sutured to the skin. Once the patient is transferred to the ward, the catheter is placed on continuous gravity drainage or intermittent suction, as per cardiothoracic surgery service. Patients with clinical and imaging signs of tamponade usually require complete aspiration of effusion with continuous catheter drainage for 24 to 48 hours.119 Some argue creating of a pericardial window or sclerosing agent instillation.79 The principle techniques of CT-guided pericardiocentesis and drain placement are similar to those used in other body areas. However, it is important to monitor the patient closely using continuous ECG monitoring because the guidewire can stimulate cardiac arrhythmias. During the procedure, it is important to monitor the patient appropriately and to have ICU support staff present if the patient is unstable. The patient should have a good IV access in case pharmacologic intervention is required.

Regardless of the technique chosen for needle guidance, delayed postprocedure arrhythmia has been reported in up to 9% of patients.4,69,246 Given the high rate of delayed postprocedure arrhythmia, patients should be managed jointly by interventional radiology and cardiology services on a telemetry unit.

Results.  Percutaneous drainage of pericardial effusions is associated

Mediastinal abscesses are a rare but devastating condition with mortality rates as high as 40% secondary to overwhelming sepsis, especially when the diagnosis and treatment are delayed.107 Mediastinal fluid collections may result from head and neck infections that descend via fascial planes into the mediastinum (40%-60%), cardiothoracic surgery, or posttraumatic esophageal injury.12,38,44,194 Following esophagogastric surgery, mediastinal abscess have a reported frequency of 8% to 21% and are the leading cause of postoperative morbidity, stricture, and repeat operation.61,245 Clinical presentation varies depending on the location of the infection and includes high fever, cough, dyspnea, tachycardia, and chills. CT scans typically demonstrate localized fluid collections with or without gas bubbles, tissue necrosis, infiltration of soft tissue, and subcutaneous emphysema. As the infection spreads, sepsis, extensive cellulitis, and abscess formation may occur.

with less morbidity and mortality than surgical therapy. It is also costcompetitive. Palmer and colleagues reported that cost for CT- and US-guided procedures have similar cost savings over surgical drainage with general anesthesia.191 CT and US have similar costs, complications, and results. Therefore modality choice can be based on the availability and skill of the consulting interventionalist. In our institution, we prefer CT guidance, given the wide field of view inherent with CT and ability to readily distinguish adjacent structures. The largest reported experience with this technique is by Klein et al., who showed a lower rate of complications using CT guidance when compared with transthoracic echocardiography (TTE)-guided procedures in more than 290 patients.122 Palmer and colleagues reported 92% (33/39) success in resolution of the pericardial effusion without additional procedures.191 In this study, 8% (3/36) patients had recurrence of pericardial effusion. Eichler and colleagues reported their experience in CT-guided pericardiocentesis, showing successful drainage of pericardial effusion in 100% (20/20) of patients under CT guidance.57 Similar success rates were reported by other groups.48

Complications.  Despite imaging guidance, there are significant risks to pericardiocentesis, mostly associated with poor visualization of the anatomic structures or needle tip. Most of the reported complications include atrial and ventricular wall puncture, epicardial laceration, coronary artery injury, arterial injury along the puncture path, arrhythmia, vasovagal reaction, pneumothorax, and infection. Cardiac penetration may occur; however, this is usually not a major complication as long as small-caliber needles are used and only a ventricle is penetrated.55,69 The thick highly contractile walls of the ventricles presumably cause the needle tract to seal. However, penetration into an atrium or laceration of a coronary artery may cause very serious bleeding. Fluoroscopically guided pericardiocentesis has a 5% to 10% incidence of cardiac penetration leading to significant morbidity.292 The reported incidence of cardiac penetration with sonographic guidance is 0% to 2%, with minimal associated morbidity. Other complications encountered during pericardial drainages include arrhythmias, bleeding, pneumopericardium, and injury to the diaphragm or liver. The improved visualization afforded by CT theoretically gives it an advantage over other imaging techniques, especially in the postoperative period. A complication rate between 2% and 42% has been reported in larger series of CT-guided pericardiocentesis and drainages.48,119 Most of the reported complications were pneumothoraces, which could be monitored by consecutive CT and drained within the same session.31

Conclusion.  CT-guided pericardiocentesis and drainage is a safe, reliable, and effective alternative to surgical drainage in patients with clinically significant pericardial effusion. Percutaneous drainage avoids the additional risks of surgical drainage, including those of general anesthesia, prolonged hospitalization, and increased postoperative pain. In critically ill patients who eventually will require a surgical procedure, an imaging-guided drainage may allow the patient’s condition to be stabilized until surgery can be performed.

MEDIASTINAL ABSCESS Pathogenesis and Diagnosis

Management Historically, IV antibiotics combined with surgical débridement had been the mainstay of therapy.39 Conservative therapy includes broadspectrum antibiotics, parenteral nutrition with elimination of oral intake, and nasogastric drainage. Surgical management includes trans­ cervical approaches, median sternotomy, and video-assisted or mediastinoscopically assisted drainage.39,84 There is still general agreement that surgery is indicated for a mediastinal abscess secondary to esophageal perforation.1,222 Misthos et al. published a report on 11 patients with mediastinitis, which demonstrated that early aggressive combined cervical and transthoracic débridement is more effective than less aggressive surgical approaches.166 However, all patients in this series required multiple repeat operations to control infection. Other issues with surgical management include the inherent perioperative risks associated with an invasive surgery and prolonged recovery time. Surgical risk may be prohibitive in many patients, even for a potentially life-saving operation. Imaging-guided drainage is ideal in such patients because a catheter can be placed within the abscess to stabilize and possibly cure critically ill patients. Endoscopic treatment of uncomplicated mediastinal collections has been described.114 Endoscopic techniques include placement of drains,1 clips,232,263 and stents.234,235 However, this technique requires long-term intubation with gradual withdrawal of the catheter as the collection slowly resolves.193 There is little discussion in the literature regarding the role of interventional radiology in the treatment of mediastinal collections since early work published in the late 1980s and early 1990s.12,38,155 The willingness to attempt percutaneous drainage of the mediastinum may be limited because of the proximity of major vascular structures, the

CHAPTER 68  Computed Tomography-Guided Drainage esophagus, and pulmonary parenchyma. Therefore careful review of preprocedural imaging to assess potential access routes and CT guidance to identify critical anatomic structures are essential for percutaneous mediastinal drainages.

Percutaneous Drainage

Technique.  CT is the guiding modality of choice.154 Preprocedural images are analyzed to determine the safest percutaneous route to the abscess that avoids the esophagus, lung, and internal mammary or intercostal arteries. Patients receive IV antibiotics with broad-spectrum coverage before percutaneous drainage procedure. All drainages are performed with the use of IV conscious sedation. Once the decision of percutaneous drainage has been made, the principles behind the techniques of aspiration and drainage are identical to those detailed for other percutaneous drainage. Although single-step catheters on trocars may be used for large abscesses in the anterior or posterior mediastinum, we usually use a 19-gauge Yueh needle and the Seldinger technique to enter and drain middle mediastinal abscesses. Typically an 8F to 10F catheter is placed. The needle and catheter should remain extrapleural. Posterior mediastinal abscess can usually be drained from a direct posterior approach without crossing the pleura or other vital structures. Anterior mediastinal abscesses can be drained from a parasternal approach. It is important to avoid the internal mammary artery and all of the mediastinal great vessels (Fig. 68-7). Middle mediastinal abscesses are more difficult to drain. A posterolateral approach is usually required, either through the right or left paraspinal space (Fig. 68-8). It is also important to use adequate local anesthesia. The needle should enter the skin at the level of the costotransverse process junction. Once the needle reaches the vertebral body, the needle bevel is turned toward the vertebral body and the needle is advanced slowly. The needle path must be carefully planned before the mediastinum is entered because once the needle is against the vertebral body it is difficult to change its trajectory. Once the needle passes medial to the descending aorta, the bevel can be adjusted to facilitate entry into the abscess (Fig. 68-9). If the abscess is paratracheal or subcarinal, it is important to avoid the esophagus, trachea, and main bronchi. An anterior approach can be used for a middle mediastinal abscess; however, it may be difficult to avoid the lung or major vessels unless there is a large amount of mediastinal fat.

Results.  Gobien and colleagues81 treated 5 of 6 patients with mediastinal abscesses and stabilized the sixth patient for later elective surgery. They also excluded the diagnosis of infection in 3 patients by proving that the mediastinal fluid collections were sterile. Wittich et al. reported successful CT-guided catheter drainage of 24 of 26 (92%) mediastinal abscesses.284 Failures were secondary to phlegmonous collections. No significant complications occurred. Tanaka and colleagues reported on two patients with mediastinal abscess after esophageal surgeries that were successfully treated with CT/fluoroscopy-guided drainage.249 Arellano and colleagues reported 100% (25/25) technical success in CT-guided mediastinal fluid collection drainage.7 One patient required an additional drainage procedure for a loculated posterior mediastinal abscess, but no one in this series required a subsequent surgical débridement. These authors’ high success rates revealed two important criteria in determining whether drainages of the mediastinum are potentially feasible: that the patient can be positioned appropriately for a safe access and that the abscess is sufficiently separated from vital mediastinal structures to minimize the complications.

Complications.  The major complications associated with this procedure include inadvertent puncture of cardiovascular or tracheobron-

2605

chial structure, esophagus, or the pleural and lung.7 If vascular structures are punctured, the bleeding should stop quickly unless the vessel is lacerated. Although erosion into vessel or other structures by an indwelling catheter is a possible risk, this has not yet been reported.

INTRAABDOMINAL ABSCESS AND COLLECTIONS There are many etiologies for intraabdominal abscess. They may manifest after penetrating abdominal injury, an operation, or be secondary to conditions such as perforated viscus, diverticulitis, appendicitis, cholangitis, or pancreatitis. Intraabdominal abscesses typically develop as areas of localized peritonitis where infection has been walled off by the omentum, adjacent organs, and the peritoneum. Although the location of abscess may determine its presenting symptoms, abdominal pain, leukocytosis, and fever are most commonly encountered clinical presentation. CT examinations are most helpful when the clinical suspicion of intraabdominal abscess is high. The principle of all intraabdominal abscess management is source control in the treatment algorithm.67 The mechanical barrier created by the fibrous abscess wall results in limited antibiotic penetration, and antibiotics alone may not be ineffective.277 Therefore definitive therapy requires abscess drainage. The mortality in undrained abdominal abscesses is high, with a mortality rate ranging between 10% and 70%.169,235 The traditional therapy of intraabdominal abdominal abscess was operative drainage. In the last few decades, abscesses within deeper structures within the abdomen and pelvis became safely accessible with image guidance. Since that time, percutaneous drainage has emerged as the first-line therapy over surgical intervention in the absence of immediate surgical indication, primarily because of lower complication rates and shorter length of hospital stay.189,220,241 The American Society of Colon and Rectal Surgery, Expert Panels of the Surgical Infection Society, the Infectious Disease Society of America, and the World Society of Emergency Surgery recommend PAD over surgical drainage where feasible.113 Aeder et al.2 recommend the following criteria for CT-assisted percutaneous drainage: (1) no more than two abscess cavities or loculations, (2) the drainage route does not transverse the bowel or any uncontaminated organs, (3) the viscosity of the abscess material allows for complete aspiration at the initial procedure, and (4) the cause of abscess is not definite by radiology. However, the most important consideration in choosing a drainage technique, independent of the technique itself, is the severity of illness of the patient. When an abscess is identified and safe access for drainage is available, percutaneous catheter drainage under CT or US is recommended as the initial mode of therapy.2,80,188

APPROACHES TO SPECIFIC ANATOMIC SITES The various anatomic areas involved by abscesses have unique characteristics. Information unique to the technique approach and outcomes in each area are discussed in the following sections.

Subphrenic Abscesses Subphrenic abscess involves the subdiaphragmatic and subhepatic spaces and the lesser sac. Subphrenic abscesses offer special challenges in regard to access and occasionally determining the origin of the abscess. The cephalad nature of the collections suggest considerations similar to those with abscesses in the dome of the liver. The large majority of these abscesses are presently postoperative, typically following gastric or biliary surgery.159 Because the subphrenic space is in communication potentially with all other peritoneal compartments,

2606

PART III  Image-Guided Procedures

A

C

B

FIG 68-7  This 67-year-old man with a history of coronary artery bypass grafting 2 months prior presented with leukocytosis and chest pain. A, Noncontrast CT demonstrates fluid collection in the anterior mediastinum to the right of the ascending aorta. Postcontrast image demonstrates arterial enhancement of the saphenous vein graft within this fluid collection (arrow). B, This area was carefully accessed using a parasternal approach under CT guidance to avoid the saphenous vein graft. C, Postprocedure CT image demonstrates successful drainage catheter placement and intact saphenous vein graft (arrow).

abscesses that accumulate in the subphrenic spaces may have remote origins or may spread to other spaces. Therefore one must always look carefully at the contiguous anatomic areas to see if there are any extensions of the abscess into these areas. When planning the trajectory and performing such a procedure, the operator should always visualize and avoid the pleural space, gallbladder, and hepatic flexure of the colon. If there is any problem with localizing a subphrenic abscess or separating hepatic flexure of the colon from the liver, the patient should be placed in the lateral or decubitus position. The key elements of the trajectory approach were described earlier in the general technique of catheter drainage. The rate of cure for subphrenic collection is 79% to 85%. Complications include empyema and sepsis, the incidence of which can be up to 2%.92,178 FIG 68-8  Illustration of posterolateral transcostal approach through the

Right Subhepatic Abscess.  An abscess in the right subhepatic

left paraspinal space.

space can be approached easily either laterally or posterolaterally.

CHAPTER 68  Computed Tomography-Guided Drainage

A

2607

B FIG 68-9  This 50-year-old man presented with severe chest pain and leukocytosis 1 month following gastric pull-through surgery. A, CT of the chest demonstrates a middle mediastinal fluid collection within the surgical bed, concerning for an abscess. A 5F Yueh needle was advanced under CT guidance to the middle mediastinal abscess through the left paraspinal space. B, Postprocedure CT image demonstrates pigtail catheter in the middle mediastinal fluid collection.

Penetration of an adjacent structure, such as the colon, gallbladder, or duodenum, should be avoided. In some cases, a left decubitus position also can be helpful in moving the hepatic flexure of the colon to simplify a puncture (Fig. 68-10). In cases in which clear access is unavailable, it is reasonable to traverse a portion of the liver. Because the liver in such cases is clearly exposed to the pyogens owing to close proximity, there is likely no difference in risk between the two approaches.

Left Subhepatic Abscess.  A left subhepatic abscess located in the anterior subphrenic space below the left lobe of the liver appears between the liver and the stomach on axial CT scan. Administration of oral contrast material is important to confirm the stomach location in certain cases. The optimal approach for this area is a directly anterior approach that avoids liver and stomach (Fig. 68-11).

Omental Bursa Drainage of the omental bursa can be approached from several directions. Fluid collections in the inferior recess of the lesser sac can be approached anteriorly beneath the lower margin of the stomach if a fluid collection is large enough. If the fluid collection is in the splenic portion of the inferior recess, the best approach is through the gastrosplenic ligament. This approach is best accomplished when the patient is in a right decubitus position to permit the stomach to shift anteriorly. In such cases, one must be careful to avoid the splenic flexure of the colon, looking carefully for varices in the gastrosplenic ligament to avoid penetration of any such vessels. If the approach through the gastrosplenic ligament is unavailable, two other approaches are possible. For a second choice, a transgastric approach can be used, ideally through a percutaneous endoscopic gastrostomy (PEG). Another approach is a transhepatic approach through the left lobe of the liver (other authors recommend this as a primary approach).177 We do not frequently use this approach because of

possible injury or pyogenic dissemination into the left biliary or portal system.

Inferior Recess of Lesser Sac and Porta Region With mastery of these techniques, it is now possible to approach drainage of deep areas, which can provide remarkable benefit. These deep areas are difficult to drain by inexperienced operators, but with experience and knowledge of the techniques described in this chapter, drainage of very deep areas is possible (Fig. 68-12). To accomplish these procedures successfully, one must take advantage of patient positioning and needle deflection as described in earlier sections.

Pericolonic Spaces The pericolonic spaces are easily approached laterally. When draining an abscess in these spaces, it is optimal to have the catheter traverse the entire length of the cavity. To do this, it is always best to enter either the upper or the lower end of the cavity. Another option is to inset two catheters at the midportion of the cavity, with one extending cephalad and one extending caudad.

Liver Abscesses Pyogenic liver abscesses are a common indication for PAD. Clinically, virtually all pyogenic liver abscesses require drainage. Typical causes for pyogenic liver abscess include extension of gallbladder and gallstone disease, prior trauma or surgery, and seeding from sepsis. Abscesses in the liver should be approached by the shortest trajectory available, with an entrance point on the anterior or lateral surface. When a lateral entrance site is used, the operator should always try to avoid traversing the pleura. When puncturing and draining a hepatic abscess, the operator should avoid any visible hepatic vessel, the gallbladder, and pleura. There should always be a cuff of normal liver to prevent spillage of material into the peritoneum.

2608

A

PART III  Image-Guided Procedures

B

C D

E

F FIG 68-10  This 58-year-old woman presented with sepsis and right upper quadrant pain. A, CT demonstrates a right perihepatic subphrenic abscess containing a small amount of gas. Note the proximity of this abscess to the diaphragm. B, CT image demonstrating the inferior aspect of the fluid collection, caudal to diaphragm. C, The patient was placed in left decubitus position and a 5F Yueh needle was introduced into the inferior aspect of the fluid collection. D, The needle was advanced cephalad. E, A10F pigtail catheter was placed into the right subphrenic fluid collection. F, Follow-up CT in 4 days shows significant decrease in the amount of subphrenic fluid collection, with the drainage catheter in satisfactory position.

CHAPTER 68  Computed Tomography-Guided Drainage

A

2609

ally, multiple hepatic abscesses can be treated with multiple catheters with side holes (Fig. 68-13). Even fairly extensive loculations usually resolve with percutaneous drainage because the septae are often incomplete (Fig. 68-14). Success of percutaneous drainage in pyogenic hepatic abscesses approaches 90% in the cumulated series.9,112 Liver abscess can be drained with up to 4% risk of major complications. Major complications have included hepatocolic fistula, sepsis, and biliary peritonitis.43,221 The most common minor complications is treatment failure resulting in repeat drainage or surgery, which can occur in 14% to 18% of cases.10,23 Presence of biliary communication may prolong catheterization, particularly in liver transplants with diminished hepatic arterial perfusion.18,53,243 Rare complications include bleeding, sepsis, hemobilia, and arterioportal fistula formation.

Pancreatic Abscesses

B

C FIG 68-11  This 43-year-old woman presented with leukocytosis and left upper quadrant pain. A, CT images demonstrate a left subphrenic abscess. B, A 5F Yueh needle (arrow) was advanced into the fluid collection below the level. When draining a subphrenic abscess, it is optimal to have the catheter traverse the entire length of the cavity. To do this, it is always best to enter the lower end of the cavity, below the level of diaphragm. C, Satisfactory positioning of an 8.5F pigtail catheter in the left subphrenic abscess.

Many hepatic abscesses may appear to be multiseptated, but experience has shown that many clear completely with one drainage tube.54,112 Various explanations have been proposed, including that the abscesses intercommunicate, the abscesses drain spontaneously into the drained cavity, and antibiotics clear the different cavities. A repeat scan is important to ensure that all such cavities eventually clear because separate cavities that persist occasionally require drainage. Occasion-

Drainage of pancreatic abscess by PAD is a complex issue that has not yet been resolved. There is a discrepancy in the literature about the objective results, recommendations, and the experience of various authors. Despite isolated reports of spontaneous internal drainage  of pancreatic abscesses, there is now a consensus that pancreatic abscesses do not spontaneously resolve, and early intervention is  warranted.25 Owing to the high mortality associated with operative management of pancreatic abscess and advances in percutaneous drainage methodology, PAD has been proposed as an alternative. Although early reports demonstrated a success rate up to 80%, many of those favorable reports included only a few cases and used a carefully selected patient subgroup with only unilocular abscess formation.116,268 It is now believed that pancreatic abscesses are multilocular in up  to 40% of cases and that infected pseudocysts are a different clinical entity from infected ill-defined necrotic pancreatitis, which is more difficult to treat.25,75 Subsequent larger studies in which the distinction between infected pseudocysts and pancreatic abscess was more sharply defined found effective drainage of pancreatic abscess in only 14% of cases.212 The major challenge with pancreatitis is the diagnosis and management of pancreatic necrosis. Most authors differentiate necrotic illdefined pancreatitis with superimposed infection from well-delineated low-density fluid collections, which more closely resemble infected pseudocysts. When patients have pseudocysts associated with pancreatic necrosis, alternating areas of strictured and dilated pancreatic duct, a suspected neoplasm, or when complications such as perforation or hemorrhage arise, surgery is still a first-line treatment option.64 Pancreatic necrosis is a classic surgical case because it involves nonliquefactive tissue that typically does not respond to drain placement and suction.64 There is a high incidence of treatment failure for necrotic pancreatitis when drained by external radiologic drainage.73 Especially in cases involving ductal injury with direct communication to a pseudocyst, many surgeons are adamant about performing surgery as a first-line treatment to avoid delays and complications associated with percutaneous drainage, such as persistent enteric or cutaneous fistula formation. Nevertheless, literature precedent exists for careful selection of patients to provide the most appropriate treatment option. Patients with mature homogenous collections without communication to the duct by endoscopic or MR pancreatography and patients with infected pseudocysts or frank abscesses typically benefit from percutaneous drainage as a first-line treatment.269 In our experience the technique and approach used depend on the appearance and location of the pancreatic fluid collections. For well-defined collections, we are willing to perform percutaneous drainage because of the high success rate

2610

PART III  Image-Guided Procedures

A

B

C

D FIG 68-12  This 52-year-old woman status post Roux-en-Y gastric bypass presented with worsening abdominal pain and leukocytosis. A, CT study demonstrates a left subphrenic abscess within the lesser sac. B, A 5F Yueh needle was advanced lateral to the left hepatic lobe and stomach and medial to adjacent bowel loops via a subcostal approach. C, Guidewire was advanced into the fluid collection under CT guidance. D, An 8.5F pigtail catheter was advanced over the guidewire into the fluid collection.

(Fig. 68-15). Collections representing infected necrotic tissue or poorly defined multiple collections are best drained surgically as described earlier. However, if a patient has had a surgical drainage and develops a collection postoperatively, we attempt percutaneous drainage regardless of the appearance, knowing that the infection is most likely well confined and that a second surgical procedure would be difficult (Fig. 68-16). We use either an anatomic approach that avoids all adjacent organs or a transgastric approach. If one is attempting to avoid bowel or solid organs, the approach often must be retroperitoneal, anterior to the kidney, or directly anterior. We use such a direct anterior approach only if there is a “window” between the various loops of bowel. If there is any problem with access, we choose a transgastric approach using the same method we have described for pseudocyst drainage, with gastrostomy as a conduit for entry. One must be prepared for multiple catheter insertions because pancreatic abscesses may be difficult to drain effectively.

An occasional source of confusion with pancreatic collection is differentiation of a pancreatic pseudocyst from a cystic pancreatic tumor. The preferred choice for cystic tumors is surgical extirpation. However, occasionally needle aspiration is performed because of  an uncertain diagnosis or because of compromised clinical status of the patient precluding a surgical intervention. Fluid from a cystic malignant pancreatic tumor may contain elevated carcinoembryonic antigen (CEA). In one series of pancreatic abscesses with no coexisting pseudocysts or necrosis, 59 patients underwent percutaneous drainage and antibiotic therapy.269 A total of 51 patients were cured and discharged from the hospital without the need for surgery. Eight patients (14%) treated with PAD failed initial therapy; five of those later required surgery for cure. The reported mean duration with a catheter in place was 33 days. Other authors cited variable success in using PAD as primary therapy for pancreatic abscess, and they report problems such as drains clogged

CHAPTER 68  Computed Tomography-Guided Drainage

A

B

C

D FIG 68-13  This 67-year-old patient on chemotherapy with multiple comorbidities presented with sepsis. A, CT through the upper abdomen demonstrates multiple hepatic abscesses. B, Two 8.5F pigtail catheters traversing multiple abscesses were placed under CT guidance. C, Scout image of the upper abdomen demonstrates two catheters. D, Follow-up CT demonstrates complete resolution of the previously identified hepatic abscesses.

2611

A

E

B F

G C

D FIG 68-14  This 61-year-old woman presented with a hepatic abscess that occurred following diverticulitis status post hemicolectomy and diversion colostomy. Axial (A) and coronal (B) CT images demonstrate a complex multiloculated abscess in the right hepatic lobe. C, A 5F Yueh needle was advanced into the abscess through a cuff of normal liver to prevent spillage of material into the peritoneum. D, A guidewire was advanced into the abscess via the Yueh catheter. E, Subsequently an 8F catheter was advanced over the guidewire. F, Coronal CT image demonstrates the catheter in satisfactory position and interval decompression of the fluid pocket following aspiration of 60 mL of purulent fluid via the catheter. G, Follow-up CT obtained 10 days later demonstrated near-complete resolution of the hepatic abscess.

CHAPTER 68  Computed Tomography-Guided Drainage

A

D

B

E

C FIG 68-15  This 44-year-old man with a history of pancreatitis secondary to a traumatic motor vehicle accident was transferred from an outside hospital with pancreatitis complicated by peripancreatic pseudocysts. A, CT through the upper abdomen demonstrates two pancreatic pseudocysts. A nasobiliary pancreatic tube and transpapillary stent were placed endoscopically into the pseudocyst in the head of pancreas (black arrow). However, the second infected pseudocyst in the tale of pancreas (white arrow) was not accessible endoscopically. B, A 5F Yueh needle was advanced into the pseudocyst at the pancreatic tail from an anterolateral approach. C, Axial CT image demonstrating a guidewire being advanced into the pseudocyst under CT guidance. D and E, A 10F pigtail catheter was successfully placed into the pseudocyst over the guidewire.

2613

2614

PART III  Image-Guided Procedures

A

B

C

D FIG 68-16  A, Axial CT image of a 39-year-old woman with necrotizing pancreatitis status post surgical débridement and surgically placed drain (arrow). B, Follow-up CT was obtained after the surgical drain was inadvertently pulled out. Residual fluid collection is seen in the surgical bed. C, A fistulous tract from the skin to fluid collection created by the removed surgical drain was identified. A red rubber catheter was advanced through the tract (arrow) into the fluid collection. D, Using the Seldinger technique, a 14F drainage catheter was placed.

by necrotic tissue.168 Lee et al. described 30 patients who were initially given PAD, of which 16 later required surgery, and the mean duration with catheter in place was 5 weeks.135 Complications have been short term and long term. Short-term complications have been related to hemorrhage, either for pseudoaneurysms or for inappropriately attempted drainage of poorly defined phlegmons. Initial attempts to drain pancreatic abscesses were nonselective, and the methods were not clearly defined. Regardless of the primary approach taken for drainage of pancreatic abscesses, successful resolution of the problem depends on diligent repetitive clinical and imaging evaluation of the patient. Close cooperation between the surgeon and radiologist is essential because of the recurring and unpredictable nature of the disease. A radiologic procedure often is needed after an operative drainage, or a radiologic procedure can be performed to stabilize a critically ill patient before a surgical procedure.

Renal and Perirenal Abscesses Renal and perirenal abscesses often are associated with nephrolithiasis. Septicemia is another cause of renal abscess; in this situation abscesses occasionally are multiple, bilateral, or both. Renal abscesses may rupture through the renal capsule into the perirenal space. As with intrarenal abscess, perirenal abscesses usually are amenable to

percutaneous catheter drainage. If renal obstruction is associated with a renal or perirenal abscess, percutaneous nephrostomy may be needed for access to stone removal or to relieve a malignant or benign obstruction. Renal and perirenal fluid collections are a medical emergency and despite aggressive surgical drainage may result in high mortality in up to 56% of these patients.218 Despite antibiotics and drainage, the mortality rate in more recent studies has remained over 12%.161 Therefore early recognition, percutaneous drainage of larger collections (>3 cm), and vigorous follow-up is critical to patient survival. Catheter manipulations and changes to maintain drainage and aggressive surgical intervention can improve the results if such efforts fail to produce an effective clinical response.132 Renal abscesses are best approached posteriorly through a path avoiding the spinal erector muscle, colon, liver, and spleen or from a posterior axillary line location. One should traverse the least amount of normal parenchyma to access the abscess, since the renal parenchyma is very vascular. Puncture and drainage of such abscesses can be performed with US or CT guidance. When the upper pole is involved, CT is indicated to avoid inadvertent trauma to the spleen, splenic pedicle, or pancreas (Fig. 68-17). The anterior and posterior pararenal spaces are frequent sites of abscess and can be approached safely131; however, certain facts must be remembered. These spaces extend from the pelvis to the diaphragm

CHAPTER 68  Computed Tomography-Guided Drainage

A

B

C

D FIG 68-17  This 21-year-old man presented with left-sided pyelonephritis and a suspected renal abscess. A, Axial CT image demonstrates loculated fluid collection within the upper pole of the left kidney. The patient was placed in left decubitus position to minimize the risk of traversing pleura and lung tissue. B, A 5F Yueh needle was placed into the fluid collection, avoiding the spinal erector muscle. C, A catheter was advanced over the guidewire. D, Postprocedure CT image demonstrates satisfactory positioning of an 8F catheter.

2615

2616

PART III  Image-Guided Procedures

and are long. Unless a catheter or catheters are positioned throughout the entire length of the cavity, drainage may be unsuccessful. It is especially important to use CT for follow-up of such cases because the abscess may spread toward the diaphragm or pelvis. Examination of these areas may be difficult with US. Success rates for these abscesses range from 68% to 81%. The complication rate is 3% to 6%, with sepsis and hemorrhage being the most common.130,233

Splenic Abscesses Splenic abscesses have a host of interesting etiologies.115 Perhaps the most frequent nowadays are microabscesses, often from Candida; they are seen typically in immunocompromised patients. Posttraumatic superinfection of a hematoma is another cause of splenic abscess. Similarly, infected infarcts are seen in septicemic patients. Splenic cysts may become infected Drainage of splenic abscess is controversial for several reasons. Some early investigators reported favorably on drainage of splenic abscess, but if one carefully reviews their reports, there was a high incidence of emergent splenectomy. In the past, a high rate of treatment failure had been reported for percutaneous treatment of splenic abscesses, possibly resulting from the use of smaller-bore catheters or needle aspiration alone to avoid a theoretically high risk of bleeding complications.202 More recent information has remained inconsistent. Liu and colleagues146 reported a 90% success rate, whereas Green85 reported a 25% success rate. Thanos and Diliana255 reported favorable results with no significant complications. Additional limited series demonstrate that this procedure can be performed with a low or absent risk of bleeding.42,68,149 These reports come from different areas of the world, and there may be differences in selection of the patients and geographic and ethnic differences among the patients. Because the incidence of splenic abscess worldwide is increasing owing to the increasing number of immunocompromised patients, we believe that the PAD procedure should be used but only in selected patients. Because the spleen is so vascular, one probably should drain only abscesses that are very close to the surface of the spleen. When draining a splenic abscess, one should try to traverse the least amount of normal spleen to drain the abscess, owing to the vascularity of splenic tissue (Fig. 68-18). Unlike the liver, where a “cuff ” of normal hepatic parenchyma is desired, a “cuff ” of normal spleen is not needed and should be avoided or made as small as possible. Realizing that this procedure has not proven to be safe in controlled studies, one must always have close communication with surgeons to provide “expectant observation” and make sure an emergent splenectomy can be performed expeditiously if needed.

A

B

Pelvic Abscesses Abscesses that are deep in the pelvis are often technically challenging and require careful planning. The sacrum, ischia, iliac crests, urinary bladder, and multiple bowel loops may be in the direct path of the abscesses, and dense presence of pelvic vasculature further complicates the choice of route of access. The procedure for drainage of pelvis abscesses is similar to that of other areas. The most significant controversy involves the choice of drainage pathways. In our opinions the best pathways, from most to least preferred, are anterior, transrectal, transvaginal, transperineal, and transsciatic approaches. When abscess cavities are close to the rectum, a transrectal approach is indicated. To perform this procedure, a series of baseline scans are taken over the pelvis to show the collection. The following method can be used to safely place the 5F Yueh catheter into the rectum. The plastic sheath covering the Yueh needle that comes in the package can be used to facilitate the procedure (Fig. 68-19). First, the sheath should be cut

C FIG 68-18  This 5-year-old girl with a history of cerebral palsy and developmental delay was status post Nissen fundoplication and gastrostomy tube to treat gastroesophageal reflux. She presented with peritonitis and fungal sepsis secondary to leaks at the gastrostomy tube site and gastroesophageal junction. A, Axial CT image demonstrates a subcapsular splenic fluid collection. B, Anterior access approach was selected to avoid traversing the normal splenic tissue, and a 19-gauge Yueh needle was advanced into the fluid collection in between adjacent bowel loops. C, Postprocedure CT image demonstrates a 6F Skater catheter in a satisfactory position.

CHAPTER 68  Computed Tomography-Guided Drainage

2617

wall,213 transrectally as with a surgical approach,196 and transvaginally.265 The latter is an approach well suited to US guidance. Using trocar technique, a drainage catheter can be advanced through a sheath attached to the US probe. Complications are rare, described in under 3% of cases in many published series.19,187,265 The most common complication associated with this approach is catheter dislodgment because there is usually no ideal location to secure the catheter. The final option would be posteriorly via a transgluteal approach.8

Enteric Abscesses

FIG 68-19  When using a transrectal approach, the plastic sheath covering for the Yueh needle that comes in the package can be used to safely place the catheter into the rectum. First the sheath should be cut so that it is 5 to 6 cm shorter than the Yueh needle. The needle is then placed in the sheath such that its tip is still within the sheath. The operator then holds the needle and sheath and places the index fingertip next to the sheath-covered Yueh needle point.

so that it is 5 to 6 cm shorter than the Yueh needle. The Yueh needle is then placed in the sheath such that the tip of the needle is still within the sheath. The operator then holds the needle and sheath and places his or her index fingertip next to the sheath-covered Yueh needle point. The finger and sheath-needle combination are then carefully inserted into the patient’s rectum. Placing the needle this way protects the operator’s fingertip and the patient’s rectum from the needle point. Once the sheath and needle are in the rectum, the operator removes his/her finger and can carefully advance the sheath-needle combination further by first advancing the sheath and then advancing the Yueh catheter. When the needle is aligned properly, puncture into the fluid collection is made. The needle is removed and a relatively stiff guidewire is advanced into the abscess, but the sheath is left in place to serve as a conduit for guidewire and dilator insertion. The sheath prevents buckling of the wire or catheter in the rectum during the insertion. The sheath is removed from the guidewire and rectum before final catheter placement (Fig. 68-20). Alternatively a transperineal approach can be used, particularly for patients who have had a proctectomy  (Fig. 68-21). The transvaginal approach can be used with US guidance; either the single-step trocar method or Seldinger technique can be used. The second choice for draining pelvic abscess is through the sciatic notch (Fig. 68-22). This pathway is satisfactory for biopsy procedures, but long-term drainage catheters are not recommended for several reasons. Use of this pathway can produce a lot of pain, and there have been reports of sciatic neuritis, abscess tracking into the gluteal muscles, and pseudoaneurysm formation in the iliac vessels. Because of the proximity of the sacral margin and passage through the muscles, catheter kinking is common. Pseudoaneurysm of the inferior gluteal artery has been described as a potential complication, and use of the Seldinger rather than the trocar technique may minimize this risk by maximizing accuracy.287 Aside from enteric and postoperative abscesses, tuboovarian abscess is a frequently encountered lesion that can be treated successfully by PAD.34 The options are many; directly through the anterior abdominal

Abdominal fluid collection resulting from perforated bowel due to such entities as Crohn’s disease, previous operation, diverticulitis, and appendicitis can be drained with a less than 5% incidence of complications, as evidenced by large published series.35,66,186 The general complications of bleeding and infection described earlier may occur infrequently (incidence < 3%). High rates of sepsis were reported in older studies,76 likely reduced secondary to improved antibiotic coverage, catheter design, and techniques for imaging guidance. Percutaneous aspiration and catheter drainage permit numerous options for the surgeon and interventional radiologists when caring for patients with enteric abscesses. In particular, periappendiceal  and diverticular abscesses can be drained preoperatively (Fig. 68-23). Perioperative drainage permits the surgeon to remove the appendix electively at a later date or resect the portion of the colon with diverticulitis when the patient is stabilized and nonseptic. Alternatively, percutaneous drainage may be used in selected cases to obviate all operations or used postoperatively for persistent or new abscesses. In patients who are poor operative candidates, drainage of an abscess may allow all symptoms to resolve; this is particularly helpful in elderly patients with severe cardiopulmonary disease in whom surgery is undesirable.22,213,214 Percutaneous drainage of diverticular abscess is a technically simple procedure for which the indications are limited because in most patients with diverticular abscesses, a colonic diverticular fistula results after successful drainage of the abscess. We concur with the many authors who reported that the primary purpose in performing this procedure is to change a potential two-stage procedure into a one-stage procedure.185,217,239 Specifically, most cases of small diverticular abscesses can be treated with a single surgical procedure consisting of an en bloc resection of a segment of the sigmoid colon and a confined abscess. The entire problem is resolved with a single procedure. In cases in which the abscess is so large that this cannot be accomplished, surgical treatment would require two procedures: one to drain the abscess and divert the bowel, and a second to reanastomose the colon. In such cases, percutaneous drainage can remove the infectious material, and a single surgical procedure can be performed to remove the offending segment of sigmoid colon.

CONTROVERSIAL ISSUES Antibiotics We believe that all patients who have a percutaneous procedure should have systemic antibiotics before and during the course of abscess drainage. With antibiotics, patients have quicker resolution of the abscess and fewer local complications. Some patients not receiving antibiotics may develop a localized cellulitis around the site of catheter insertion; administration of antibiotics resolves or prevents this. If antibiotics are not given, septicemia is more likely to occur after or during instrument manipulation. In our opinion, the only valid controversy is whether antibiotics should be started before or after the diagnostic aspiration. Starting antibiotics when a patient initially develops septic symptoms can be

2618

PART III  Image-Guided Procedures

R

A

D

F

A

B

C

B

E

G FIG 68-20  A 53-year-old man with a history of diverticulitis presented with sepsis. A, Axial CT image demonstrates thick-walled fluid collection (A) within the rectovesical pouch, concerning for an abscess. Rectum (R) and bladder (B) are identified. A decision was made to use a transrectal approach instead of a transsciatic approach because the patient was on Plavix. B-D, A 5F Yueh needle (arrow) is advanced into the fluid collection via a transrectal approach under CT guidance. F, Postprocedure axial CT image demonstrates an 8F catheter in a satisfactory position. G, Follow-up CT 4 weeks later demonstrates the transrectal drainage catheter in a stable position within the decompressed fluid collection.

CHAPTER 68  Computed Tomography-Guided Drainage

A

B

C

D

E

FIG 68-21  This 66-year-old woman who was status post extensive laparoscopic lysis of adhesions, takedown of enterocutaneous fistula, and ileal conduit presented with sepsis and pelvic pain. A, CT through the pelvis demonstrates a presacral fluid collection (arrows) associated with a thick enhancing wall and surrounding inflammatory changes, suggestive of an abscess. The colorectal surgeon requested avoidance of the transvaginal approach owing to risk of fistulization. B-D, A decision was made to use a low posterior transperineal approach. A 5F Yueh needle (arrow) was introduced inferiorly and advanced cephalad into the fluid collection. E, Postprocedure CT demonstrates satisfactory position of a 10F catheter within the fluid collection.

2619

2620

PART III  Image-Guided Procedures

A

B

C

D FIG 68-22  This 60-year-old man with history of rectal cancer, status post lower anterior resection surgery and surgically placed transrectal drain, presented with leukocytosis and pelvic pain. A, There is a focal fluid collection within the posterior aspect of the lower right pelvis (white arrow), posterior to the surgically placed transrectal drain (black arrow). B, A transsciatic approach was used to access the fluid collection. A 5F Yueh catheter was introduced below the level of the piriformis muscle and medial to the neurovascular bundle (dashed arrow). C, The Yueh catheter was advanced into the fluid collection (dashed arrow). D, Postprocedure image demonstrates satisfactory location of a 10F catheter.

justified to prevent further deterioration. It can be argued, however, that a valid culture cannot be obtained if the patient is receiving antibiotics. Either philosophy seems sound; we usually accept the preference expressed by clinical colleagues attending the patient.

colleagues proposed the use of acetylcysteine (Mucomyst) to decrease the viscosity of purulent material.272 However, most interventionalists do not accept this technique as being effective, and no additional reports have appeared.

Catheters

Fibrinolytic Agents

Abscess drainage has become so ubiquitous that there are innumerable drainage catheters available on the market. Each has its own unique quality advertised by the various vendors. The list is too exhaustive to recommend any specific catheters; readers can decide which products meet their own needs. Controversy exists about whether gravity (siphon) drainage, sump drainage, or irrigation drainage (multiple catheters) is the most effective. A report by Park and coworkers showed that there was little difference in drainage success rates regardless of whether small or large catheters were used.192 This is probably the result of selection of catheters by operators (i.e., small catheters for thin material and large catheters for thick material).

Extensive work has been performed by our group using urokinase initially, and now other fibrinolytics have been studied by other groups.93,127 Fibrinolytic agents all work by the same pathway, converting plasminogen to plasmin. Based on recent reports, fibrinolytics, including streptokinase, urokinase, and tPA, can be used as noted earlier.58,127,150,167

Mucolytic Agents There have been several suggestions in the literature about using materials to try to enhance drainage from abscesses. Van Waes and

Infected Neoplasms Although experience has shown that percutaneous drainage of infected neoplasms is not curative,179 there has been some more recent interest in using percutaneous drainage as a palliative procedure. Our experience indicates that this is not a worthwhile procedure; such abscesses never heal because antibiotics are not delivered to the site through the tumor. A chronic drainage fistula typically develops, making skin care difficult.

CHAPTER 68  Computed Tomography-Guided Drainage

A

B RT TRANS T

0

2

4

C

D RT TRANS T

0

S/P

2

4

E

F FIG 68-23  This 6-year-old boy with a perforated appendix and sepsis presented for preoperative percutaneous drainage. A, Contrast-enhanced CT through the lower abdomen demonstrates a multiloculated abscess within the right lower quadrant. B, Axial CT image caudal to (A) demonstrates a narrow isthmus (arrow) of communication between the two largest fluid pockets. C, For continuous visualization of an access needle tract to achieve optimal drainage of multiple abscess pockets, US was used to guide a 5F Yueh catheter (arrow) through the narrow isthmus. D, CT image demonstrates a guidewire advancing into the medial fluid pocket. E, Postprocedure CT showing satisfactory position of an 8.5F catheter. F, US image confirms a catheter traversing the narrow isthmus, allowing drainage of multiple abscess pockets.

2621

2622

PART III  Image-Guided Procedures

Echinococcal Drainage

General Technique

Echinococcal abscesses are endemic in certain areas worldwide but are extremely uncommon in the United States. Although there is concern about the development of anaphylaxis, substantial experience has accumulated in the Middle East suggesting that these abscesses usually can be treated safely by percutaneous drainage.63 Numerous authors have reported on the percutaneous drainage of echinococcal cysts.24,77,190,223 The method is simple and consists of typical percutaneous puncture and aspiration using US or CT. With the Seldinger technique, catheters between 12F and 20F are inserted. The cavity is irrigated with alcohol and povidone-iodine and left in the site for 30 minutes. The catheter is left to drain until fluid ceases. Concurrent treatment with albendazole also is important in appropriate doses and duration. Anaphylactic reactions occurred in several cases but were not fatal. The most definitive paper by Khuroo and colleagues reported 50 patients with hydatidosis; 25 were treated with surgery, 35 were treated by catheter drainage, and both groups were treated with albendazole.120 Percutaneous treatment was effective in 88%, compared with 72% by surgery. Complications occurred in 32% of the percutaneous group and 84% of the surgical group.

The technique is similar to that described for abscess punctures. The skin is prepared with the standard antiseptic preparation and a local anesthetic is administered. Depending on the interventionalist’s preference, either a 21-gauge micropuncture needle or a standard 5F Yueh needle can be used to puncture the renal pelvis. The micropuncture needle is advantageous if the patient is coagulopathic or if the collecting system is nondilated. Once the collecting system is accessed, a 0.018-inch guidewire is advanced into the collecting system. The tract is sequentially dilated such that a standard 0.035-inch guidewire can be placed. If the 5F Yueh catheter is used, the inner trocar needle is removed after it enters the collecting system, and a standard 0.035-inch angiographic wire can be placed. When the nephrostomy is performed under CT, several technical maneuvers are important. After the collecting system has been penetrated, a small sample of fluid should be aspirated for culture, and then a small amount of contrast material should be injected immediately. This injected contrast material permits direct visualization under fluoroscopy when the catheter is being inserted. If one has access to a fluoroscopy unit in the CT interventional suite, this step is simplified. A 0.035-inch guidewire can then be placed, the tract is then dilated, and the appropriate-sized nephrostomy tube is placed. If the patient must be transported to a fluoroscopy suite, we recommend that a 0.035-inch guidewire be inserted into the collecting system before transporting the patient to the fluoroscopy area, since it is less likely to be dislodged.

Amebic Abscess Although amebic abscesses are preferably treated with antibiotics such as metronidazole, certain situations prompt percutaneous drainage.118,228 These cases include poor response to antibiotic therapy, uncertainty as to the diagnosis, imminent or even actual leakage or rupture, size larger than 6 to 8 cm, and left lobe abscesses that have a proclivity to rupture into the thorax.250 Akgun,3 Tandon,250 and Hanna95 and their coworkers reported on the use of aspiration and drainage in conjunction with amebicidal therapy. It was the consensus that percutaneous aspiration and drainage may expedite therapy and provide a cure when there is resistance to medical therapy. For amebic abscesses, complications are uncommon and minor.219

PERCUTANEOUS NEPHROSTOMY The same advantages for CT procedures are applicable to nephrostomy placement, but several different points should be emphasized. CT has been successful in a high percentage of cases; in one early report, we successfully completed 12 of 13 attempts.90 Barbaric and associates reported a series of 148 procedures using CT and fluoroscopy.15 Their success rate was 97%, and there were no major complications. Authors such as Hagspiel94 and Merkle162 and their colleagues have used MRI for placement of nephrostomy tubes.

Indications The indication for CT guidance is limited, since most nephrostomies can be performed easily using either fluoroscopic or US guidance. The cases for which we use CT guidance include obese patients  or patients where the kidney is not seen well with US, unusual anatomic variation, or a kidney with an urinoma associated with the hydronephrosis. When a collecting system with hydronephrosis has ruptured and produced an adjacent urinoma, CT is better than other methods for several reasons. First, if the urinoma is displacing the axis of the kidney, the orientation can be appreciated. Second, if there is active excretion of contrast material, the perinephric space may fill with contrast material, making it difficult to visualize the collection system under fluoroscopy.

PERCUTANEOUS CHOLECYSTOSTOMY Percutaneous cholecystostomy is widely accepted as a viable palliative alternative to surgical cholecystostomy in cases of acute sepsis from cholecystitis in a critically ill patient, particularly in the ICU setting. Either US or CT may be used for performing the procedure.

Indications An acutely inflamed gallbladder producing sepsis is the sole indication for this procedure. Abnormal gallbladders have the expected finding of a very thick wall and may have associated inflammatory stranding in the adjacent tissues. Percutaneous cholecystostomy is useful in draining fluid in acute calculous or acalculous cholecystitis or gallbladder empyema in patients who will be placed at high risk by surgery or anesthesia and whose medical condition is unstable.254,274 In most patients, cholecystectomy is undertaken at a later date. Percutaneous cholecystostomy may also be used to drain an obstructed biliary system when the obstruction is below the cystic duct.

Contraindications There are two contraindications to the procedure. Rarely the gallbladder appears so inflamed that one might be concerned about rupture and spillage of bile into the peritoneum. The interventionalist must consider this possibility, although it is very unlikely because of the adherence of tissues from inflammation around the gallbladder. Unmanageable coagulopathy may preclude the procedure.

General Technique for Percutaneous Cholecystostomy The chosen modality can be either US or CT, depending on the anatomic approach desired. One can choose a transhepatic route or a peritoneal pathway.

Anatomic Pathway.  Either a transhepatic or a peritoneal approach may be used. During the early development of this technique, most

CHAPTER 68  Computed Tomography-Guided Drainage

A

B

C

D

2623

FIG 68-24  This 54-year-old woman in a surgical intensive care unit following exploratory laparotomy and bowel resection for Crohn’s disease presented with increasing bilirubin and transaminitis. A, Contrastenhanced CT through the upper abdomen demonstrates distended gallbladder with sludge. The clinical diagnosis was acalculous cholecystitis. B, Using a transhepatic approach, a 5F Yueh catheter was advanced into the gallbladder under CT guidance. C, A guidewire was advanced into the gallbladder through the Yueh catheter. D, Postprocedure CT image demonstrates satisfactory positioning of an 8.5F catheter within the gallbladder.

authors preferred a transhepatic approach through the liver. The premise was that penetrating the gallbladder through the so-called bare area between the gallbladder and liver would prevent spillage of bile into the peritoneum. As the technique matured, it was determined that when necessary a transperitoneal approach could be used without adverse effects. If the transhepatic approach is used, the middle third of the gallbladder is targeted (Fig. 68-24). The region near the gallbladder neck may be too narrow to allow catheter and guidewire manipulations. The region near the fundus may be intraperitoneal and not within the bare area of the gallbladder. If the transperitoneal pathway is chosen, uninvolved structures such as the colon should be avoided. The target should be the fundus of the gallbladder and should parallel the long axis of the gallbladder. The transperitoneal approach is best performed with single-stick trocar catheters. Typically a small 6F or 8F catheter is used. Since the gallbladder is directly punctured with the catheter, there is less chance of bile spilling into the peritoneum compared to the Seldinger technique, where sequential dilations are performed over a guidewire. The transperitoneal approach is well suited

for US guidance, since the trocar-based catheter can be inserted directly in the long axis of the gallbladder.

Preprocedure Preparation.  The key element in preprocedure preparation is to ensure that the patient is receiving appropriate broadspectrum antibiotics. Just as with abscess, bacteremia can occur during instrumentation of an infected gallbladder.

Insertion of Needle and Catheter.  Using the transhepatic approach and the Seldinger technique, the middle third of the gallbladder lumen is targeted with the 5F Yueh needle. The fluid recovered may be thin or thick; sometimes it is purulent, but in most cases the fluid is dark green and thick. In contrast to other catheter insertions, we prefer a softer more “floppy” guidewire, which can be coiled easily within the gallbladder. Because the wall of the gallbladder may be fragile, we seldom use a stiff wire. The standard dilation and insertion of the catheter is performed; usually a 6F to 8F catheter is used. If one uses the single-stick trocar technique, incremental adjustments are made to

2624

PART III  Image-Guided Procedures

A

B

D

C

E

G

F

H FIG 68-25  This 49-year-old man presented with acute back pain. A, Initial contrast-enhanced CT exam demonstrates a penetrating ulcer with pseudoaneurysm arising from the aorta. Also seen is associated focal fluid collection (arrow). The patient underwent aortic stent placement. B, Follow-up CT angiogram 10 days status post surgery demonstrates reaccumulation of fluid collection adjacent to the aorta. D and E, Preprocedure dynamic contrast-enhanced CT images in soft tissue (D) and lung window (E) demonstrate extensive anterior and posterior abdominal wall collaterals and an enlarged left lumbar artery (arrow). E, A 19-gauge Yueh needle was advanced into the retroperitoneal fluid collection cephalad to the previously identified prominent lumbar artery under CT guidance. F, Using the Seldinger technique, an 8F percutaneous drainage catheter was placed without any complications. G, Ten days later, repeat CT angiogram demonstrates interval resolution of the periaortic fluid collection, with the drainage catheter in place. H, Following percutaneous catheter removal, repeat CT angiogram was obtained 14 days after initial drain placement. No residual periaortic retroperitoneal fluid collection is identified.

position the catheter. Because of the weight of the device, it is best to support the weight of the catheter with folded towels during the movement of the patient in and out of the CT scanner.

Postprocedure Care.  Postprocedure care is straightforward. The catheter is left to drainage, and the patient is given the proper course of antibiotics. When the patient has improved sufficiently, an elective cholecystectomy can be performed if the patient has gallstones. Patients who are unsuitable for surgery, such as patients in hospice, may need to have the catheter in place long term. In patients with

acalculous cholecystitis, the catheter can usually be removed in 2 to 3 weeks. Prior to removing the catheter, a sinogram is performed to verify patency of the cystic duct. If patent, the catheter can be removed. If the transperitoneal approach is used, some authors advocate performing a sinogram along the catheter tract prior to removing the catheter to make sure the tract has matured and that there is no chance for a leak once the catheter is removed. To do this, a guidewire is placed into the gallbladder through the catheter. The catheter is then removed over the guidewire. The guidewire is left in place and contrast is instilled into the tract along the guidewire. If the contrast

CHAPTER 68  Computed Tomography-Guided Drainage stays within the tract and enters the gallbladder, the wire is removed. If contrast enters the peritoneum, a new catheter is placed over the guidewire into the gallbladder and the tract can be reevaluated in another week.

Results and Complications Among the authors reporting, there is consistent agreement that these procedures are effective and provide effective palliation for acute cholecystitis.59,97,266 Complications vary and include bile peritonitis, bleeding, vaginal reactions, hypotension, and acute respiratory distress. Rates are 4% to 10%.267

MYCOTIC ANEURYSMS OR INFECTED VASCULAR PROSTHESES Because of the difficult treatment required for infected vessels or synthetic aortic prostheses, any patient with such a suspected infection should have a diagnostic aspiration as early as possible. In the infected prosthesis and primary mycotic infections we have encountered, cultures have been positive and reliable. In each of these cases, the culture results were crucial and helped expedite treatment of the problem. With prompt treatment, most of these patients survived the surgical therapy. In another group of 10 patients with uninfected grafts, the results were negative. No complication has been associated with the needle procedures. Such cases should be approached from the flank, and a pathway should be chosen to avoid the kidney and major vessels. In most instances a clearly defined fluid collection is seen either adjacent to the graft or between the wall of the aneurysm and the graft. As with other procedures performed from this approach, we try to avoid the psoas muscle, the spinal erector muscles, and the spinal elements to prevent patient discomfort. If there is any question about the location of the vessel or prosthetic wall, a dynamic bolus CT scan should be performed. Dynamic CT is highly effective to show a vascular problem (Fig. 68-25). In one patient with a clotted mycotic aneurysm, only CT correctly showed the vascular nature of the lesion. Some operators have raised concerns about possible injury to an infected vessel, but in our experience this is not a problem. In the more than 20 suspected cases we have aspirated, there have been no complications. The need for aggressive procedures for expedited patient care is clear. The mortality rate of untreated prosthesis infection is close to 100%. Corrective treatment and surgery can be effective only if initiated early. In some instances, percutaneous drainage of infected prostheses may be effective if the infection involves only the midportion and not the anastomotic ends. Other authors such as Rossi,211 Lambiase,128 and their colleagues also have proposed an aggressive approach to the diagnosis and management of these difficult infections. These authors also have advocated percutaneous aspiration for diagnosis of infectious agents and palliating the infection with preoperative catheter drainage.

REFERENCES 1. Abe N, Sugiyama M, Hashimoto Y, et al: Endoscopic nasomediastinal drainage followed by clip application for treatment of delayed esophageal perforation with mediastinitis. Gastrointest Endosc 54:646–648, 2001. 2. Aeder MI, Wellman JL, Haaga JR, et al: Role of surgical and percutaneous drainage in the treatment of abdominal abscesses. Arch Surg 118:273–280, 1984. 3. Akgun Y, Tacyildiz IH, Celik Y: Amebic liver abscess: Changing trends over 20 years. World J Surg 23:102–106, 1999.

2625

4. Allen KB, Faber LP, Warren WH, et al: Pericardial effusion: Subxiphoid pericardiostomy versus percutaneous catheter drainage. Ann Thorac Surg 67(2):437–440, 1999. 5. Almerindo S, Offner PJ, Moore EE, et al: Optimal management of complicated empyema. Am J Surg 180(6):507–511, 2000. 6. Antunes G, Neville E, Duffy J, et al: Pleural Diseases Group, Standards of Care Committee, British Thoracic Society. BTS guidelines for the management of malignant pleural effusions. Thorax 58(Suppl 2):ii29–ii38, 2003. 7. Arellano RS, Gervais DA, Mueller PR: Computed tomography-guided drainage of mediastinal abscesses: Clinical experience with 23 patients. J Vasc Interv Radiol 22:673–677, 2011. 8. Ashbaugh DG: Empyema thoracis: Factors influencing morbidity and mortality. Chest 99:1162–1165, 1991. 9. Attar B, Levendoglu H, Cuasay NS, et al: CT-guided percutaneous aspiration and catheter drainage of pyogenic liver abscesses. Am J Gastroenterol 81:550–555, 1986. 10. Back SY, Lee MG, Cho KS, et al: Therapeutic percutaneous aspiration of hepatic abscesses: Effectiveness in 25 patients. AJR Am J Roentgenol 160:799–802, 1993. 11. Bakal CW, Sacks D, Burke DR, et al, Society of Interventional Radiology Standards of Practice Committee: Quality improvement guidelines for adult percutaneous abscess and fluid drainage. J Vasc Interv Radiol 14:S223–S225, 2003. 12. Bakshi R, Grover G: Retropharyngeal abscess with mediastinal extension in an infant–Still existing? Pediatr Emerg Care 25:181–183, 2009. 13. Ball WS, Bisset GS, Towbin RB: Percutaneous drainage of chest abscess in children. Radiology 171:431–434, 1989. 14. Bamberger DM: Outcome of medical treatment of bacterial abscesses without therapeutic drainage: A review of cases reported in the literature. Clin Infect Dis 23:604–607, 1996. 15. Barbaric ZL, Hall T, Cochran ST, et al: Percutaneous nephrostomy: Placement under CT and fluoroscopy guidance. AJR Am J Roentgenol 169:151–155, 1997. 16. Bartlett JG: The role of anaerobic bacteria in lung abscess. Clin Infect Dis 40:923–925, 2005. 17. Basile A, Boullosa-Seoane E, Dominguez Viguera L, et al: Intrapleural fibrinolysis in the management of empyemas and hemothoraces: Our experience. Radiol Med (Torino) 105:12–16, 2003. 18. Bayraktar Y, Arslan S, Sivri B, et al: Percutaneous drainage of hepatic abscesses: Therapy does not differ for those with identifiable biliary fistula. Hepatogastroenterology 43:620–626, 1996. 19. Bennett JD, Kozak JI, Taylor BM, et al: Deep pelvic abscess: Transrectal drainage with radiologic guidance. Radiology 185:825–828, 1992. 20. Bennett R, Maskell N: Management of malignant pleural effusions. Curr Opin Pulm Med 11(4):296–300, 2005. 21. Benoist S, Panis Y, Pannegeon V, et al: Can failure of percutaneous drainage of postoperative abdominal abscess be predicted? Am J Surg 184:148–153, 2002. 22. Bernini A, Spncer MP, Wong WD, et al: Computed tomography-guided percutaneous abscess drainage in intestinal disease: Factors associated with outcome. Dis Colon Rectum 40:1009–1013, 1997. 23. Bissada AA, Bateman J: Pyogenic liver abscess: A 7-year experience in a large community hospital. Hepatogastroenterology 38:317–320, 1991. 24. Bosanac ZB, Lisanin L: Percutaneous drainage of hydatid cyst in the liver as a primary treatment: Review of 52 consecutive cases with long-term follow-up. Clin Radiol 55:839–848, 2000. 25. Bradley EL, Warshaw AL: Pancreatic abscess. In Go VLW, DiMagno EP, Gardner JD, et al, editors: The Pancreas. Biology, Pathobiology, and Disease, ed 2, New York, 1993, Raven Press, pp 649–662. 26. Breen JF: Imaging of the pericardium. J Thorac Imaging 16:47–54, 2001. 27. Broaddus VC, Light RW: Pleural effusion. In Mason RJ, Broaddus VC, Martic TR, et al, editors: Textbook of Respiratory Medicine, ed 5, Philadelphia, 2010, Saunders/Elsevier, pp 1719–1763. 28. Bruning R, Muehlstaedt M, Becker C, et al: Computed tomographyfluoroscopy guided drainage of pericardial effusions: Experience in 11 cases. Invest Radiol 37:328–332, 2002.

2626

PART III  Image-Guided Procedures

29. Budoff MJ, Lu B, Mao S, et al: Evaluation of fluid collection in the pericardial sinuses and recesses: Noncontrast-enhanced electron beam tomography. Invest Radiol 35:359–365, 2000. 30. Bufalari A, Giustozzi G, Moggi L: Postoperative intraabdominal abscesses: Percutaneous versus surgical treatment. Acta Chir Belg 62:218–222, 1996. 31. Bungay HK, Berger J, Traill ZC: Pneumothorax post CT-guided lung biopsy: A comparison between detection on chest radiographs and CT. Br J Radiol 72:1160–1163, 1999. 32. Callahan JA, Seward JB, Nishimura RA, et al: Two-dimensional echocardiography guided pericardiocentesis, experience in 117 consecutive patients. Am J Cardiol 58:633–635, 1986. 33. Callahan JA, Seward JB, Tajik AJ: Cardiac tamponade: Pericardiocentesis directed by two-dimensional echocardiography. Mayo Clin Proc 60:344–347, 1985. 34. Casola G, vanSonnenberg E, D’Agostino H, et al: Percutaneous drainage of tuboovarian abscesses. Radiology 182:399–402, 1992. 35. Casola G, vanSonnenberg E, Neff CC, et al: Abscesses in Crohn’s disease: Percutaneous drainage. Radiology 163:19–22, 1987. 36. Cham CW, Haq SM, Rahamim J: Empyema thoracis : A problem with late referral? Thorax 48:925–927, 1993. 37. Chan PC, Huang LM, Wu PS, et al: Clinical management and outcome of childhood lung abscess: A 16 year experience. J Microbiol Immunol Infect 38:183–188, 2005. 38. Chang C, Huang J, Lai PC, et al: Posterior mediastinal abscess in a hemodialysis patient– A rare but life-threatening complication of Staphylococcus bacteremia. Clin Nephrol 71:92–95, 2009. 39. Chen K, Chen J, Kuo SW, et al: Descending necrotizing mediastinitis: A 10-year surgical experience in a single institution. J Thorac Cardiovasc Surg 136:191–198, 2008. 40. Chen CH, Huang WC, Chen TY, et al: Massive necrotizing pneumonia with pulmonary gangrene. Ann Thorac Surg 87:310–311, 2009. 41. Chen W, Lin YC, Liang SJ, et al: Hospital acquired thoracic empyema in adults: A 5-year study. South Med J 102:909–914, 2009. 42. Chou YH, Hsu CC, Tiu CM, et al: Splenic abscess: Sonographic diagnosis and percutaneous drainage or aspiration. Gastrointest Radiol 17:262–266, 1992. 43. Chung YF, Tay KH, Stan B, et al: Percutaneous drainage of liver abscess complicated by hepatovenous fistula. Singapore Med J 44:299–301, 2003. 44. Church A, Sivasothy P, Parmer J, et al: Mediastinal abscess after lung transplantation secondary to Burkholderia gladiolo infection. J Heart Lung Transplant 28:511–514, 2009. 45. Colice GL, Curtis A, Deslauriers J, et al: Medical and surgical treatment of parapneumonic effusions: An evidence based guideline. Chest 118(4):1158–1171, 2000. 46. Cooper JP, Oliver RM, Currie P, et al: How do the clinical findings in patients with pericardial effusions influence the success of aspiration? Br Heart J 73:351–354, 1995. 47. Cowles R, Lelli J, Jr, Takayasu J, et al: Lung resection in infants and children with pulmonary infections refractory to medical therapy. J Pediatr Surg 37:643–647, 2002. 48. Cox JE, Chiles C, McManus CM, et al: Transthoracic needle aspiration biopsy: Variables that affect risk of pneumothorax. Radiology 212:165–168, 1999. 49. Davies R, Davies H, Davies C: British Thoracic Society: management of pleural infection-draft guidelines. BTS 29:1–39, 2009. 50. Davies HE, Rahman NM, Parker RJ, et al: Use of indwelling pleural catheters for chronic pleural infection. Chest 133(2):546–549, 2008. 51. Dev SP, Nascimiento B, Jr, Simone C, et al: Videos in clinical medicine. Chest tube insertion. N Engl J Med 357(15):e15, 2007. 52. Deveney CW, Lurie K, Deveney KE: Improved treatment of intraabdominal abscess: A result of improved localization, drainage,   and patient care, not technique. Arch Surg 123:1126–1130, 1988. 53. Do H, Lambiase RE, Deyoe L, et al: Percutaneous drainage of hepatic abscesses with and without intrahepatic biliary communication. AJR Am J Roentgenol 157:1209–1212, 1991.

54. Dondelinger RF, Kurdziel JC, Gathy C: Percutaneous treatment of pyogenic liver abscess: A critical analysis of results. Cardiovasc Intervent Radiol 13:174–182, 1990. 55. Duvernoy O, Borowiec J, Helmius G, et al: Complications of percutaneous pericardiocentesis under fluoroscopic guidance. Acta Radiol 33:309–313, 1992. 56. Duvernoy O, Larsson SG, Persson K, et al: Pericardial effusion and pericardial compartments after open heart surgery: An analysis by computed tomography and echocardiography. Acta Radiol 31:41–46, 1990. 57. Eichler K, Zangos S, Thalhammer A, et al: CT-guided pericardiocentesis: Clinical profile, practice patterns and clinical outcome. Eur J Radiol 75(1):28–31, 2010. 58. Ekingen G, Guvenc B, Sozubir S, et al: Fibrinolytic treatment of complicated pediatric thoracic empyemas with intrapleural streptokinase. Eur J Cardiothorac Surg 26:503–507, 2004. 59. England RE, McDermott VG, Smith TP, et al: Percutaneous cholecystostomy: Who responds? AJR Am J Roentgenol 168:1247–1251, 1997. 60. Erasmus JJ, Goodman PC, Patz EF, Jr: Management of malignant pleural effusions and pneumothorax. Radiol Clin North Am 38(2):375– 383, 2000. 61. Escofet X, Manjunath A, Twine C, et al: Prevalence and outcome of esophagogastric anastomotic leak after esophagectomy in a UK regional cancer network. Dis Esophagus 23:112–116, 2010. 62. Feagins LA, Holubar SD, Kane SV, et al: Current strategies in the management of intra-abdominal abscesses in Crohn’s disease. Clin Gastroenterol Hepatol 9(10):842–850, 2011. 63. Felice C, Pirola F, Brunnetti E, et al: A new therapeutic approach for hydatid liver cysts: Aspiration and alcohol injection under sonographic guidance. Gastroenterology 98:1366–1368, 1990. 64. Fernández-Cruz L, Sáenz A, Astudillo E, et al: Laparoscopic pancreatic surgery in patients with chronic pancreatitis. Surg Endosc 16(6):996– 1003, 2002. 65. Fernandez-Sabe N, Carratala J, Dorca J, et al: Efficacy and safety   of sequential amoxicillin-clavulanate in the treatment of anaerobic   lung infections. Eur J Clin Microbiol Infect Dis 22:185–187, 2003. 66. Flancbaum L, Nosher JL, Brolin RE: Percutaneous catheter drainage of abdominal abscesses associated with perforated viscus. Am Surg 56:52–56, 1990. 67. Gang GI, Moulton JF, Solomkin JS: Drainage. In Schein M, Marshall JC, editors: Source Control: A Guide to the Management of Surgical Infections, Berlin, Germany, 2002, Springer, pp 25–35. 68. Gasparini D, Basadonna PT, DiDonna A: Splenic abscesses: Their percutaneous treatment and the role of the interventional radiologists. Radiol Med (Torino) 87:803–807, 1994. 69. Gatenby RA, Hartz WH, Kessler HB: Percutaneous catheter drainage for malignant pericardial effusion. J Vasc Interv Radiol 2(1):151–155, 2005. 70. Gates RL, Hogan M, Weinstein S, et al: Drainage, fibrinolytics, or surgery: A comparison of treatment options in pediatric empyema. J Pediatr Surg 39:1638–1642, 2004. 71. Geissbuhler K, Keiser A, Fuhrer J, et al: Video-assisted thoracoscopic pericardial fenestration for loculated or recurrent effusions. Eur J Cardiothorac Surg 14:403–408, 1998. 72. Gervais DA, Brown SD, Connolly SA, et al: Percutaneous imagingguided abdominal and pelvic abscess drainage in children. Radiographics 24:737–754, 2004. 73. Gervais DA, Hahn PF, O’Neill MJ, et al: Percutaneous abscess drainage in Crohn disease: Technical success and short- and long-term outcomes during 14 years. Radiology 222:645–651, 2002. 74. Gervais DA, Ho CH, O’Neill MJ, et al: Recurrent abdominal and pelvic abscesses: Incidence, results of repeated percutaneous drainage, and underlying causes of 956 drainage. AJR Am J Roentgenol 182:463–466, 2004. 75. Gerzof SG, Johnson WC, Robbins AH, et al: Percutaneous drainage of infected pancreatic pseudocysts. Arch Surg 119:888–893, 1984.

CHAPTER 68  Computed Tomography-Guided Drainage 76. Gerzof SG, Robbin AH, Johnson WC, et al: Percutaneous catheter drainage of abdominal abscess: A five-year experience. N Engl J Med 305:653–657, 1981. 77. Giorgio A, Tarantino L, de Stefano G, et al: Hydatid liver cyst: An 11-year experience of treatment with percutaneous aspiration and ethanol injection. J Ultrasound Med 20:729–738, 2001. 78. Giorgio A, Tarantino L, Marmniello N, et al: Pyogenic liver abscesses: 13 years of experience in percutaneous needle aspiration with US guidance. Radiology 195:122–124, 1995. 79. Girardi LN, Ginsberg RJ, Burt ME: Pericardiocentesis and intrapericardial sclerosis: Effective therapy for malignant pericardial effusions. Ann Thorac Surg 64:1422–1427, 1997. 80. Glass CA, Cohn IJ: Drainage of intraabdominal abscesses: A comparison of surgical and computerized tomography guided catheter drainage. Am J Surg 147:315–317, 1984. 81. Gobien RP, Stanley JH, Gobien BS: Percutaneous catheter aspiration and drainage of suspected mediastinal abscesses. Radiology 151:69–71, 1984. 82. Golfieri R, Cappelli A: Computed tomography-guided percutaneous abscess drainage in coloproctology: Review of the literature. Tech Coloproctol 197–208, 2007. 83. Gonzalez-Pinto I, Gonzalez EM: Optimising the treatment of upper gastrointestinal fistulae. Gut 49(Suppl 4):iv22–iv31, 2001. 84. Gorlitzer M, Grabenwoeger M, Meinhart J, et al: Descending necrotizing mediastinitis treated with rapid sternotomy followed by vacuum-assisted therapy. Ann Thorac Surg 83:393–396, 2007. 85. Green BT: Splenic abscess: Report of six cases and review of the literature. Am Surg 67:80–85, 2001. 86. Grinan NP, Lucena FM, Romero JV, et al: Yield of percutaneous needle lung aspiration in lung abscess. Chest 97:69–74, 1990. 87. Ha HK, Kang MW, Park JM, et al: Lung abscess: Percutaneous catheter therapy. Acta Radiol 34:362–365, 1993. 88. Haaga J: Definitive treatment of a large pyogenic liver abscess with CT guidance. Cleve Clin Q 43:85–88, 1976. 89. Haaga JR: Imaging intraabdominal abscesses and nonoperative drainage procedure. World J Surg 14:204–209, 1990. 90. Haaga JR, Alfidi RJ, Havrilla TR, et al: CT detection and aspiration of abdominal abscess. AJR Am J Roentgenol 128:465–474, 1977. 91. Haaga JR, George C, Weinstein AJ, et al: New interventional techniques in the diagnosis and management of inflammatory disease within the abdomen. Radiol Clin North Am 17:485–513, 1979. 92. Haaga JR, Nakamoto D: Computed tomography-guided drainage of intra-abdominal infections. Curr Infect Dis Rep 6:105–114, 2004. 93. Haaga JR, Nakamoto D, Stellato T, et al: Intracavitary urokinase for enhancement of percutaneous abscess drainage: Phase II trial. AJR Am J Roentgenol 174:1681–1685, 2000. 94. Hagspiel KD, Kandarpa K, Silverman SG: Interactive MR-guided percutaneous nephrostomy. J Magn Reson Imaging 8:1319–1322, 1998. 95. Hanna RM, Dahniya MH, Badr SS, et al: Percutaneous catheter drainage in drug-resistant amoebic liver abscess. Trop Med Int Health 5:578–581, 2000. 96. Harisinghani MG, Gervais DA, Hahn PF, et al: CT-guided transgluteal drainage of deep pelvic abscesses: Indications, technique, procedurerelated complications, and clinical outcome. Radiographics 22(6):1353– 1367, 2002. 97. Hatjidakis AA, Karampekios S, Prassopoulos P, et al: Maturation of the tract after percutaneous cholecystostomy with regard to the access route. Cardiovasc Intervent Radiol 21:36–40, 1998. 98. Hemming A, Davis NL, Robins RE: Surgical versus percutaneous drainage of intra-abdominal abscesses. Am J Surg 161:593–595, 1991. 99. Heneghan JP, Everts RJ, Nelson RC: Multiple fluid collections: CT- or US-guided aspiration–Evaluation of microbiologic results and implications for clinical practice. Radiology 212:669–672, 1999. 100. Herth F, Ernst A, Becker H: Endoscopic drainage of lung abscesses. Chest 127:1378–1381, 2005. 101. Hirshberg B, Sklair-Levi M, Nir-Paz R, et al: Factors predicting mortality of patients with lung abscess. Chest 115:746–750, 1999.

2627

102. Huang H-C, Chang H-Y, Chen C-W, et al: Predicting factors for outcome of tube thoracostomy in complicated parapneumonic effusion or empyema. Chest 115:751–756, 1999. 103. Imamoglu M, Cay A, Kosucu P, et al: Pneumatoceles in postpneumonic empyema: An algorithmic approach. J Pediatr Surg 40:1111–1117, 2005. 104. Imazio M: Pericardial involvement in systemic inflammatory diseases. Heart 97:1882–1892, 2011. 105. Imazio M, Brucato A, Trinchero R, et al: Colchicine for pericarditis: Hype or hope? Eur Heart J 30:532–539, 2009. 106. Imazio M, Mayosi BM, Brucato A, et al: Triage and management of pericardial effusion. J Cardiovasc Med (Hagerstown) 11:928–935, 2010. 107. Issacs L, Kotton B, Peralta MJ, et al: Fatal mediastinal abscess from upper respiratory infection. Ear Nose Throat J 72:620–622, 1993. 108. Jamieson DH, Chait PG, Filler R: Interventional drainage of appendiceal abscess in children. AJR Am J Roentgenol 169:1619–1622, 1997. 109. Janes SM, Rahman NM, Davies RJ, et al: Catheter-tract metastases associated with chronic indwelling pleural catheters. Chest 131:1232– 1234, 2007. 110. Jeffrey RB, Jr, Federle MP, Tolentino CS: Periappendiceal inflammatory masses: CT-directed management and clinical outcome of 70 patients. Radiology 167:13–16, 1988. 111. Jerjes-Sanchez C, Ramirez-Rivera A, Elizalde JJ: Intrapleural fibrinolysis with streptokinase as an adjunctive treatment in hemothorax and empyema: A multicenter trial. Chest 109:1514–1519, 1996. 112. Johnson RD, Mueller PR, Ferrucci JT, Jr, et al: Percutaneous drainage of pyogenic liver abscess. AJR Am J Roentgenol 144:463–467, 1985. 113. Jong P, Hearnes CW: Intra-abdominal abscess drainage: Interval to surgery. Semin Intervent Radiol 4(29):311–313, 2012. 114. Jorgensen J, Hunt D: Endoscopic drainage of esophageal suture line leaks. Am J Surg 165:362–364, 1993. 115. Kapur A, Vasudeva R, Howden CW: Candida splenic abscess in the absence of obvious immunodeficiency. Am J Gastroenterol 92:509–512, 1997. 116. Karlson KB, Martin EC, Fankuchen EI, et al: Percutaneous drainage of pancreatic pseudocysts and abscess. Radiology 142:619–624, 1982. 117. Katara AN, Samra SS, Bhandarkar DS: Thoracoscopic window for a post-coronary artery bypass grafting pericardial effusion. Indian Heart J 55:180–181, 2003. 118. Ken JG, vanSonnenberg E, Casola G, et al: Perforated amebic liver abscesses: Successful percutaneous treatment. Radiology 179:195–197, 1989. 119. Khan MF, Straub R, Moghaddam SR, et al: Variables affecting the risk of pneumothorax and intrapulmonal hemorrhage in CT-guided transthoracic biopsy. Eur J Radiol 18:1356–1363, 2008. 120. Khuroo MS, Wani NA, Javid G, et al: Percutaneous drainage compared with surgery for hepatic hydatid cysts. N Engl J Med 337:881–887, 1997. 121. Khurrum Baig M, Hua Zhao R, Batista O, et al: Percutaneous postoperative intra-abdominal abscess drainage after elective colorectal surgery. Tech Coloproctol 6:159–164, 2003. 122. Klein SV, Afridi H, Agarwal D, et al: CT directed diagnostic and therapeutic pericardiocentesis: 8-year experience at a single institution. Emerg Radiol 11:353–363, 2005. 123. Klein JS, Schultz S, Heffner JE: Interventional radiology of the chest: Image-guided percutaneous drainage of pleural effusions, lung abscess, and pneumothorax. AJR Am J Roentgenol 164(3):581–588, 1995. 124. Knaus WA, Draper EA, Wagner DP: APACHE II: A severity of disease classification system. Crit Care Med 13:818–829, 1984. 125. Kolschmann S, Ballin A, Gillissen A: Clinical efficacy and safety of thoracoscopic talc pleurodesis in malignant pleural effusions. Chest 128(3):1431–1435, 2005. 126. Kosloske AM, Ball WS, Butler C, et al: Drainage of pediatric lung abscess by cough, catheter, or complete resection. J Pediatr Surg 21:596–600, 1986. 127. Laborra JM, Haaga JR, Stellato T, et al: Safety of intracavitary urokinase with percutaneous abscess drainage. AJR Am J Roentgenol 160:171–174, 1993.

2628

PART III  Image-Guided Procedures

128. Lambiase RE, Cronan JJ, Dorfman GS, et al: Postoperative abscesses with enteric communication: Percutaneous treatment. Radiology 171:497–500, 1989. 129. Lambiase RE, Deyoe L, Cronan JJ, et al: Percutaneous drainage of 335 consecutive abscesses: Results of primary drainage with 1-year follow-up. Radiology 184:167–179, 1992. 130. Lang EK: Abdominal abscess drainage under radiologic guidance: Causes of failure. Radiology 174:329–336, 1986. 131. Lang EK: Renal, perirenal, and pararenal abscesses: Percutaneous drainage. Radiology 174:109–113, 1990. 132. Lang EK, Paolini RM, Pottmeyer A: The efficacy of palliative and definitive percutaneous versus surgical drainage of pancreatic abscesses and pseudocysts: A prospective study of 85 patients. South Med J 84:55–64, 1991. 133. Lee SF, Lawrence D, Booth H: Thoracic empyema: Current opinions in medical and surgical management. Curr Opin Pulm Med 16:194–200, 2010. 134. Lee SK, Morris RF, Cramer B: Percutaneous needle aspiration of neonatal lung abscesses. Pediatr Radiol 21:254–257, 1991. 135. Lee MJ, Rattner DW, Legemate DA, et al: Acute complicated pancreatitis: Redefining the role of interventional radiology. Radiology 183:171–174, 1992. 136. Lee SH, Tomlinson C, Temple M, et al: Imaging-guided percutaneous needle aspiration or catheter drainage of neonatal liver abscesses: 14 year experience. AJR Am J Roentgenol 190:616–622, 2008. 137. LeMense GP, Strange C, Sahn SA: Empyema thoracis. Therapeutic management and outcome. Chest 107:1532–1537, 1995. 138. Lemmer JH, Botham MJ, Orringer MB: Modern management of adult thoracic empyema. J Thorac Cardiovasc Surg 90:849–855, 1985. 139. Lestuzzi C: Subxiphoid pericardial drainage for pericardial tamponade. World J Cardiol 2(9):270–279, 2010. 140. Levison MA: Percutaneous versus open operative drainage of intra-abdominal abscesses. Infect Dis Clin North Am 6:525–544, 1992. 141. Levison MA, Zeigler D: Correlation of APACHE II score, drainage technique and outcome in postoperative intra-abdominal abscess. Surg Gynecol Obstet 172:89–94, 1991. 142. Light RW: Parapneumonic effusions and empyema. Clin Chest Med 6:55–61, 1985. 143. Light RW: A new classification of parapneumonic effusions and empyema. Chest 108:299–301, 1995. 144. Light RW: Pleural effusions. Med Clin North Am 95(6):1055–1070, 2011. 145. Lindenberger M, Kjellberg M, Karlsson E, et al: Pericardiocentesis guided by 2-D echocardiography: The method of choice for treatment of pericardial effusion. J Intern Med 253:411–417, 2003. 146. Liu KY, Shyr YM, Su CH, et al: Splenic abscesses–A changing trend in treatment. S Afr J Surg 38:55–57, 2000. 147. Lorenzo RL, Bradford BF, Black J: Lung abscesses in children: Diagnostic and therapeutic needle aspiration. Radiology 157:79–80, 1985. 148. Lotrionte M, Biondi-Zoccai G, Imazio M, et al: International collaborative systemic review of controlled clinic trials on pharmacologic treatments for acute pericarditis and its recurrences. Am Heart J 160(4):662–670, 2010. 149. Lucey BC, Boland GW, Maher MM, et al: Percutaneous nonvascular splenic intervention: A 10-year review. AJR Am J Roentgenol 179:1591– 1596, 2002. 150. Maher MM, Gervais DA, Kalra MK, et al: The inaccessible or undrainable abscess: How to drain it. Radiographics 717–735, 2004. 151. Maher MM, Kealey S, McNamara A, et al: Management of visceral interventional radiology catheters: A troubleshooting guide for interventional radiologists. Radiographics 22:305–322, 2002. 152. Maher EA, Shepherd FA, Todd TJ: Pericardial sclerosis as the primary management of malignant pericardial effusion and cardiac tamponade. J Thorac Cardiovasc Surg 112(3):637–643, 1996. 153. Maisch B, Seferovic PM, Ristic AD, et al: Guidelines on the diagnosis and management of pericardial diseases. Executive summary. Eur Heart J 25:587–610, 2004.

154. Mandel S, Boyd D, Jaques P, et al: Drainage of hepatic, intraabdominal, and mediastinal abscesses guided by computerized axial tomography. Successful alternative to open drainage. Am J Surg 145:120–125, 1983. 155. Mansharamani N, Koziel H: Chronic lung sepsis: Lung abscess, bronchiectasis and empyema. Curr Opin Pulm Med 9:181–185, 2003. 156. Marcy PY, Bondiau PY, Brunner P: Percutaneous treatment in patients presenting with malignant cardiac tamponade. Eur Radiol 15:2000– 2009, 2005. 157. Mattei P, Allen JL: Treatment of empyema in children. Am J Respir Crit Care Med 174:110–111, 2006. 158. McDonald JM, Meyers BF, Guthrie TJ, et al: Comparison of open subxiphoid pericardial drainage with percutaneous catheter drainage for symptomatic pericardial effusion. Ann Thorac Surg 76(3):811–815, discussion 816, 2003. 159. McNicholas MM, Mueller PR, Lee MJ, et al: Percutaneous drainage of subphrenic fluid collections that occur after splenectomy: Efficacy and safety of transpleural versus extrapleural approach. AJR Am J Roentgenol 165:355–359, 1995. 160. Men S, Akhan O, Koroglu M: Percutaneous drainage of abdominal abscess. Eur J Radiol 43:204–218, 2003. 161. Meng MV, Mario LA, McAninch JW: Current treatment and outcomes of perinephric abscesses. J Urol 168:1337–1340, 2002. 162. Merkle EM, Hashim M, Wendt M, et al: MR-guided percutaneous nephrostomy of the nondilated upper urinary tract in a porcine model. AJR Am J Roentgenol 172:1221–1225, 1999. 163. Miller KS, Sahn SA: Chest tubes: Indications, technique, management and complications. Chest 91:258–264, 1987. 164. Millikan JJ, Moore EE, Steiner E, et al: Complications of tube thoracostomy for acute trauma. Am J Surg 140:738–741, 1980. 165. Minton J, Stanley P: Intra-abdominal infection. Clin Med 6:519–523, 2004. 166. Misthos P, Katsaragakis S, Kakaris S, et al: Descending necrotizing anterior mediastinitis: Analysis of survival and surgical treatment modalities. J Oral Maxillofac Surg 65:635–639, 2007. 167. Misthos P, Sepsas E, Konstantinou M, et al: Early use of intrapleural fibrinolytics in the management of postpneumonic empyema: A prospective study. Eur J Cardiothorac Surg 28:599–603, 2005. 168. Mithofer K, Mueller PR, Warshaw AL: Interventional and surgical treatment of pancreatic abscess. Word J Surg 21(2):162–168, 1997. 169. Montgomery RS, Wilson SE: Intra-abdominal infection. Clin Med 6:519–523, 2004. 170. Moore AV, Zuger JH, Kelley MJ: Lung abscess: An interventional radiology perspective. Semin Intervent Radiol 8(1):36–43, 1991. 171. Moores DW, Allen KB, Faber LP, et al: Subxiphoid pericardial drainage for pericardial tamponade. J Thorac Cardiovasc Surg 109:546–552, 1995. 172. Moreira AL, Stocchi L, Tan E, et al: Outcomes of Crohn’s disease presenting with abdominopelvic abscess. Dis Colon Rectum 52:906–912, 2009. 173. Morrow LE, Jarrett J: Primary lung abscess. In Bope ET, Rakel RE, Kellerman RD, editors: Conn’s Current Therapy 2010, ed 1, Philadelphia, 2010, Saunders/Elsevier, pp 266–267. 174. Moulton JS: Image-guided management of complicated pleural fluid collections. Radiol Clin North Am 38(2):345–374, 2000. 175. Mueller PR, Berlin L: Complications of lung abscess aspiration and drainage. AJR Am J Roentgenol 178:1083–1086, 2002. 176. Mueller PR, Ferrucci JT, Jr, Butch RJ, et al: Inadvertent percutaneous catheter gastroenterostomy during abscess drainage: Significance and management. AJR Am J Roentgenol 145:387–391, 1985. 177. Mueller PR, Ferrucci JT, Jr, Simeone JF, et al: Lesser sac abscesses and fluid collections: Drainage by transhepatic approach. Radiology 155:615–618, 1985. 178. Mueller PR, Simeone JF, Butch RJ, et al: Percutaneous drainage of subphrenic abscess: A review of 62 patients. AJR Am J Roentgenol 147:1237–1240, 1986. 179. Mueller PR, White EM, Glass-Royal M, et al: Infected abdominal tumors: Percutaneous catheter drainage. Radiology 173:627–629, 1989.

CHAPTER 68  Computed Tomography-Guided Drainage 180. Muller-Willie R, Iesalnieks I, Dornia C, et al: Influence of percutaneous abscess drainage on severe postoperative septic complication in patients with Crohn’s disease. Int J Colorectal Dis 26:769–774, 2011. 181. Murthy SC, Okereke I, Mason DP, et al: A simple solution for complicated pleural effusions. J Thorac Oncol 1:697–700, 2006. 182. Musani AI: Treatment options for malignant pleural effusion. Curr Opin Pulm Med 15(4):380–387, 2009. 183. Musani AI, Haas AR, Seijo L, et al: Outpatient management of malignant pleural effusions with small-bore, tunneled pleural catheters. Respiration 71:559–566, 2004. 184. Mwandumba HC, Beeching NJ: Pyogenic lung infections: factors for predicting clinical outcome of lung abscess and thoracic empyema. Curr Opin Pulm Med 6(3):234–239, 2000. 185. Neff CC, vanSonnenberg E, Casola G, et al: Diverticular abscesses: Percutaneous drainage. Radiology 163:15–18, 1987. 186. Neff CC, vanSonnenberg E, Lawson DW, et al: CT follow-up of empyemas: Pleural peels resolve after percutaneous drainage. Radiology 176:195–197, 1992. 187. O’Neill MJ, Rafferty EA, Lee SI, et al: Transvaginal interventional procedures: Aspiration, biopsy, and catheter drainage. Radiographics 21:657–672, 2001. 188. Olak J, Christou NV, Stein LA: Operative vs percutaneous drainage of intraabdominal abscesses; comparison of morbidity and mortality. Arch Surg 121:141–146, 1986. 189. Oliak D, Yamini D, Udani VM, et al: Initial nonoperative management for periappendiceal abscess. Dis Colon Rectum 44(7):936–941, 2001. 190. Ormeci N, Soykan I, Bektas A, et al: A new percutaneous approach for the treatment of hydatid cysts of the liver. Am J Gastroenterol 96:2225–2230, 2001. 191. Palmer SL, Kelly PD, Schenkel FA, et al: CT-guided tube pericardiostomy: A safe and effective technique in the management of postsurgical pericardial effusion. AJR Am J Roentgenol 193:W314– W320, 2009. 192. Park JK, Kraus FC, Haaga JR: Fluid flow during percutaneous drainage procedures: An in vitro study of the effects of fluid viscosity, catheter size, and adjunctive urokinase. AJR Am J Roentgenol 160:165–169, 1993. 193. Pasricha PJ, Fleischer DE, Kalloo AN: Endoscopic perforations of the upper digestive tract: A review of their pathogenesis, prevention, and management. Gastroenterology 106:787–802, 1994. 194. Patel NN, Murphy G, Hamilton M: Traumatic sternal abscess with mediastinal involvement. Ann Thorac Surg 86:1997, 2008. 195. Patradoon-Ho P, Fitzgerald DA: Lung abscess in children. Paediatr Respir Rev 8:77–84, 2007. 196. Pereira JK, Chait PG, Miller SF: Deep pelvic abscess in children: Transrectal drainage under radiologic guidance. Radiology 198:393–396, 1996. 197. Pierrepoint MJ, Evans A, Morris SJ, et al: Pigtail catheter drain in the treatment of empyema thoracis. Arch Dis Child 87(4):331–332, 2002. 198. Pollak JS: Malignant pleural effusions: treatment with tunneled long-term drainage catheters. Curr Opin Pulm Med 8(4):302–307, 2002. 199. Pollak JS, Burdge CM, Rosenblatt M, et al: Treatment of malignant pleural effusions with tunneled long-term drainage catheters. J Vasc Interv Radiol 12:201–208, 2001. 200. Pothula V, Krellenstein DJ: Early aggressive surgical management of parapneumonic empyemas. Chest 105:832–836, 1994. 201. Putnam JB, Jr, Walsh GL, Swisher SG, et al: Outpatient management of malignant pleural effusion by a chronic indwelling pleural catheter. Ann Thorac Surg 69:369–375, 2000. 202. Quinn SF, van Sonnenberg E, Casola G, et al: Interventional radiology in the spleen. Radiology 161:289–291, 1986. 203. Rae E, Aroztegui O, Garcia SE, et al: Percutaneous drainage of pyogenic hepatic abscesses. Medicina (B Aires) 55:665–669, 1995. 204. Rahman NM, Maskell NA, Davies CW, et al: The relationship between chest tube size and clinical outcome in pleural infection. Chest 137(3):536–543, 2010. 205. Rajak CL, Gupta S, Jain S, et al: percutaneous treatment of liver abscesses: Needle aspiration versus catheter drainage. AJR Am J Roentgenol 170:1035–1039, 1998.

2629

206. Rice TW, Ginsberg RJ, Tomas RJ, et al: Tube drainage of lung abscess. Ann Thorac Surg 44:356–359, 1989. 207. Ridley PD, Brainbridge MV: Thorascopic debridement and pleural irrigation in the management of empyema thoracis. Ann Thorac Surg 51:461–464, 1991. 208. Roberts ME, Neville E, Berrisford RG, et al: Management of a malignant pleural effusion: British Thoracic Society Pleural Disease Guideline 2010. Thorax 65(Suppl 2):ii32–ii40, 2010. 209. Rodriguez-Panadero F, Montes-Worboys A: Mechanisms of pleurodesis. Respiration 83(2):91–98, 2012. 210. Romijn S, Sturm M, van der Schelling G: Transphrenic fistulization of a subphrenic abscess to lung parenchyma. J Gastrointest Surg 9:716–717, 2005. 211. Rossi P, Arata FM, Salvatori FM, et al: Prosthetic graft infection: Diagnostic and therapeutic role of interventional radiology. J Vasc Interv Radiol 8:271–277, 1997. 212. Rotman N, Matieu D, Anglade MC, et al: Failure of percutaneous drainage of pancreatic abscesses complicating severe acute pancreatitis. Surg Gynecol Obstet 174:141–144, 1992. 213. Ryan JM, Murphy BL, Boland GW, et al: Use of the transgluteal route for percutaneous abscess drainage in acute diverticulitis to facilitate delayed surgical repair. AJR Am J Roentgenol 170:1189–1193, 1998. 214. Sahai A, Belair M, Gianfelice D, et al: Percutaneous drainage of intra-abdominal abscess in Crohn’s disease: Short- and long-term outcome. Am J Gastroenterol 92:275–278, 1997. 215. Sahn SA: Management of complicated parapneumonic effusions. Am Rev Respir Dis 148:813, 1993. 216. Sahn SA: Pleural diseases related to metastatic malignancies. Eur Respir J 10:1907–1913, 1997. 217. Saini S, Mueller PR, Wittenberg J, et al: Percutaneous drainage of diverticular abscesses: An adjunct to surgical therapy. Arch Surg 121:475–478, 1986. 218. Salvatierra O, Jr, Bucklew WB, Morrow JW: Perinephric abscesses: A report of 71 cases. J Urol 98:296–302, 1967. 219. Saraswat VA, Agarwal DK, Baijal SS, et al: Percutaneous catheter drainage of amoebic liver abscess. Clin Radiol 45:187–189, 1992. 220. Sartelli M, Viale P, Koike K, et al: WSES consensus conference: Guidelines for first-line management of intra-abdominal infections. World J Emerg Surg 6:2, 2011. 221. Satoh H, Matsuyama S, Mashima H, et al: A case of hepatocolic fistula after percutaneous drainage for a gas-containing pyogenic liver abscess. J Gastroenterol 29:782–785, 1994. 222. Sawyer R, Phillips C, Vakil N: Short- and long-term outcome of esophageal perforation. Gastrointest Endosc 41:130–134, 1995. 223. Sayek I, Onat D: Diagnosis and treatment of uncomplicated hydatid cyst of the liver. World J Surg 25:21–27, 2001. 224. Schiza S, Siafakas NM: Clinical presentation and management of empyema, lung abscess and pleural effusion. Curr Opin Pulm Med 12(3):205–211, 2006. 225. Schuster MR, Crummy AB, Wojtowycz MM, et al: Abdominal abscesses associated with enteric fistulas: Percutaneous management. J Vasc Interv Radiol 359–363, 1992. 226. Selig MB: Percutaneous transcatheter pericardial interventions: Aspiration, biopsy, and pericardioplasty. Am Heart J 125:269–271, 1993. 227. Shabetai R: Pericardial effusion: Haemodynamic spectrum. Heart 90:255–256, 2004. 228. Shandera WX, Bollam P, Harshmey RH, et al: Hepatic amebiasis among patients in a public teaching hospital. South Med J 91:829–837, 1998. 229. Shankar S, Gulati M, Kang M, et al: Image-guided percutaneous drainage of thoracic empyema: Can sonography predict the outcome? Eur Radiol 10(3):495–499, 2000. 230. Shaw P, Agarwal R: Pleurodesis for malignant pleural effusions. Cochrane Database Syst Rev (1):CD002916, 2005. 231. Shim C, Santas GH, Zelefsky M: Percutaneous drainage of lung abscess. Lung 168:201–207, 1990. 232. Shimamoto C, Hirata I, Umegaki E, et al: Closure of an esophageal perforation due to fish bone ingestion by endoscopic clip application. Gastrointest Endosc 51:736–739, 2000.

2630

PART III  Image-Guided Procedures

233. Siegel JF, Smith A, Moldwin R: Minimally invasive treatment of renal abscess. J Urol 155:52–55, 1996. 234. Siersema PD, Homs MY, Haringsma J, et al: Use of large-diameter metallic stents to seal traumatic nonmalignant perforations of the esophagus. Gastrointest Endosc 58:356–361, 2003. 235. Solomkin JS, Mazuski JE, Bradley JS, et al: Diagnosis and management of complicated intra-abdominal infection in adults and children: Guidelines by the Surgical Infection Society and the Infectious   Diseases Society of America. Surg Infect (Larchmt) 11:79–109, 2010. 236. Sones PJ: Percutaneous drainage of abdominal collections. AJR Am J Roentgenol 142:35–59, 1984. 237. Song H, Choi YW, Jang JS, et al: Pericardium: Anatomy and spectrum of disease on computed tomography. Curr Probl Diagn Radiol 31:198–209, 2002. 238. Spivak H, Galloway JR, Amerson JR, et al: Management of pancreatic pseudocysts. J Am Coll Surg 186:507–511, 1998. 239. Stabile BE, Puccio E, vanSonnenberg E, et al: Preoperative percutaneous drainage of diverticular abscesses. Am J Surg 159:99–104, 1990. 240. Stamatakos M, Stefanaki C, Kontzoglou K, et al: Walled-off pancreatic necrosis. World J Gastroenterol 16(14):1707–1712, 2010. 241. Stocchi L: Current indications and role of surgery in the management of sigmoid diverticulitis. World J Gastroenterol 16(7):804–817, 2010. 242. Stretton F, Edmonds P, Marrinan M: Malignant pleural effusions. Eur J Palliative Care 6:5–9, 1999. 243. Sugiyama M, Atomi Y: Pyogenic hepatic abscesses with biliary communication. Am J Surg 183:205–208, 2002. 244. Taavitsainen M, Bondestam S, Mankinen P, et al: Ultrasound guidance for pericardiocentesis. Acta Radiol 32:9, 1991. 245. Tabatabai A, Hashemi M, Mohajeri G, et al: Incidence and risk factors predisposing anastomotic leak after transhiatal esophagectomy. Ann Thorac Med 4:197–200, 2009. 246. Tam A, Ensor JE, Snyder H, et al: Image-guided drainage of pericardial effusions in oncology patients. Cardiovasc Intervent Radiol 32:1217– 1226, 2009. 247. Tan C, Sedrakyan A, Browne J, et al: The evidence on the effectiveness of management for malignant pleural effusion: A systematic review. Eur J Cardiothorac Surg 29:829–838, 2006. 248. Tan TQ, Seilheimer DK: Pediatric lung abscess: Clinical management and outcome. Pediatr Infect Dis J 14:51–55, 1995. 249. Tanaka T, Inaba Y, Matsueda K, et al: Mediastinal abscess successfully treated by percutaneous drainage using a unified CT and fluoroscopy system. Br J Radiol 75:470–473, 2002. 250. Tandon A, Jain AK, Dixit VK, et al: Needle aspiration in large amoebic liver abscess. Trop Gastroenterol 18:19–21, 1997. 251. Tarver Rd, Dewey DJ: Interventional chest radiology. Radiol Clin North Am 32:689–709, 1994. 252. Tasci S, Burghard A, Schafer H, et al: Long-term outcome of intrapleural fibrinolytic therapy via a small bore catheter drainage in the management of complicated parapneumonic effusion and empyema: A case series. Med Klin (Munich) 100:181–185, 2005. 253. Tazawa J, Sakai Y, Maekawa S, et al: Solitary and multiple pyogenic liver abscesses: Characteristics of the patients and efficacy of percutaneous drainage. Am J Gastroenterol 92:271–274, 1997. 254. Teplick SK, Haskin PH, Pvalides CA, et al: Percutaneous transcholecystic cholangiography: Experimental study. AJR Am J Roentgenol 144:1059–1063, 1985. 255. Thanos I, Diliana T: Percutaneous CT guided drainage of splenic abscess. AJR Am J Roentgenol 179:629–632, 2002. 256. Thornton RH, Miller Z, Covey AM, et al: Tunneled pleural catheters for treatment of recurrent malignant pleural effusion following failed pleurodesis. J Vasc Interv Radiol 21(5):696–700, 2010. 257. Tremblay A, Michaud G: Single-center experience with 250 tunneled pleural catheter insertions for malignant pleural effusion. Chest 129(2):362–368, 2006. 258. Tsai HL, Chen CY, Chen CW, et al: Computed tomography-guided percutaneous drainage in the management of intraperitoneal abscesses. Foolyin J Health Sci 2(2):53–57, 2010.

259. Tsang TS, Barnes ME, Hayes SN, et al: Clinical and echocardiographic characteristics of significant pericardial effusions following cardiothoracic surgery and outcomes of echo-guided pericardiocentesis for management: Mayo Clinic experience, 1979-1998. Chest 116:322– 331, 1999. 260. Tsang TS, Enriquez-Sarano M, Freeman WK, et al: Consecutive 1127 therapeutic echocardiographically guided pericardiocentesis: Clinical profile, practice patterns, and outcomes spanning 21 years. Mayo Clin Proc 77:429–436, 2002. 261. Tsang TS, Freeman WK, Sinak LJ, et al: Echocardiographically guided pericardiocentesis: Evolution and state-of-the-art technique. Mayo Clin Proc 73:647–652, 1998. 262. Tsang TS, Oh J, Seward JB: Diagnosis and management of cardiac tamponade in the era of echocardiography. Clin Cardiol 2:446–452, 1999. 263. Tsunada S, Ogata S, Ohyama T, et al: Endoscopic closure of perforations caused by EMR in the stomach by application of metallic clips. Gastrointest Endosc 57:948–951, 2003. 264. van Sonnenberg E, D’Agostino HB, Casola G, et al: Lung abscess: CT-guided drainage. Radiology 178:347–351, 1991. 265. Van Sonnenberg E, D’Agostino H, Casola G, et al: US-guided transvaginal drainage of pelvic abscesses and fluid collections. Radiology 181:53–56, 1991. 266. Van Sonnenberg E, Goodacre BW: Percutaneous gallbladder puncture and cholecystostomy: Results, complications, and caveats for safety. Radiology 183:167–170, 1992. 267. Van Sonnenberg E, Wing VW, Pollard JW, et al: Life-threatening vagal reactions associated with percutaneous cholecystotomy. Radiology 151:377–380, 1984. 268. Van Sonnenberg E, Wittich GR, Casola G: Complicated pancreatic inflammatory disease: Diagnostic and therapeutic role of interventional radiology. Radiology 142:619–624, 1982. 269. Van Sonnenberg E, Wittich GR, Chon KS, et al: Percutaneous radiologic drainage of pancreatic abscesses. AJR Am J Roentgenol 168:979–984, 1997. 270. Van Sonnenberg E, Wittich GR, Goodacre BW, et al: Percutaneous drainage of thoracic collections. J Thorac Imaging 13:74–82, 1998. 271. van Sonnenberg E, Wittich GR, Goodacre BW, et al: Percutaneous abscess drainage: Update. World J Surg 25:362–369, discussion 70–72, 2001. 272. Van Waes PF, Feldberg MA, Mali WP, et al: Management of loculated abscesses that are difficult to drain: A new approach. Radiology 147:57–63, 1983. 273. Varghese JC, Hahn PF, Harisinghani MG, et al: Fungus-infected fluid collections in thorax or abdomen: Effectiveness of percutaneous catheter drainage. Radiology 236:730–738, 2005. 274. Vogelzang RL, Nemcek AA, Jr: Percutaneous cholecystotomy: Diagnostic and therapeutic efficacy. Radiology 168:29–34, 1988. 275. Voros D, Gouliamos A, Kotoulas G, et al: Percutaneous drainage of intraabdominal abscess using large lumen tubes under computed tomographic control. Eur J Surg 162:895–898, 1996. 276. Wachsman AM, Hoffer EK, Forauer AR, et al: Tension pneumothorax after placement of a tunneled pleural drainage catheter in a patient with recurrent malignant pleural effusions. Cardiovasc Intervent Radiol 30:531–533, 2007. 277. Wagner C, Sauermann R, Joukhadar C: Principles of antibiotic penetration into abscess fluid. Pharmacology 78:1–10, 2006. 278. Wali SO: An update on the drainage of pyogenic lung abscesses. Ann Thorac Med 7(1):3–7, 2012. 279. Wali S, Shugaeri A, Samman Y, et al: Percutaneous drainage of pyogenic lung abscess. Scand J Infect Dis 34:673–679, 2002. 280. Waller DA, Rengarajan A: Thoracoscopic decortication: A role for video-assisted surgery in chronic postpneumonic pleural empyema. Ann Thorac Surg 71:1813–1816, 2001. 281. Warren WH, Kim AW, Liptay MJ: Identification of clinical factors predicting PleurX catheter removal in patients treated for   malignant pleural effusion. Eur J Cardiothorac Surg 33:89–94, 2008.

CHAPTER 68  Computed Tomography-Guided Drainage 282. Weissberg D: Percutaneous drainage of lung abscess. J Thorac Cardiovasc Surg 87:308–312, 1984. 283. Wittich GR: Radiologic treatment of abdominal abscesses with fistulous communications. Curr Opin Radiol 4:110–115, 1992. 284. Wittich GR, vanSonnenberg E: Percutaneous catheter aspiration and drainage of suspected mediastinal abscesses. Radiology 189(P):156, 1993. 285. Wong B, Murphy J, Chang C, et al: The risk of pericardiocentesis. Am J Cardiol 44:1110, 1970. 286. Wong WD, Wexner SD, Lowry A, et al: The American Society of Colon and Rectal Surgeons. Practice parameters for the treatment of sigmoid diverticulitis–Supporting documentation. The Standard Task Force. Dis Colon Rectum l43(3):290–297, 2000. 287. Woo JKH, Millward SF, Harisinghani M: Transgluteal approach for percutaneous drainage of deep pelvic abscesses: How to avoid injury to vital structures. Radiology 233:300–302, 2004. 288. Wozniak CJ, Paull DE, Moezzi JE, et al: Choice of first intervention is related to outcomes in the management of empyema. Ann Thorac Surg 87:1525–1530, 2009.

2631

289. Wroblicka J, Kuligowska E: One-step needle aspiration and lavage for the treatment of abdominal and pelvic abscess. AJR Am J Roentgenol 170:1197–1203, 1998. 290. Yang P, Luk K, Lee Y, et al: Lung abscesses: US examination and US-guided transthoracic aspiration. Radiology 180:171–175, 1991. 291. Yellin A, Yellin EO, Liebermann Y: Percutaneous tube drainage: The treatment of choice for refractory lung abscess. Ann Thorac Surg 39:266–270, 1985. 292. Zahn EM, Houde C, Benson L, et al: Percutaneous pericardial catheter drainage in childhood. Am J Cardiol 70:678–680, 1992. 293. Zarem E, Hadzic A: Sonographically guided percutaneous catheter drainage versus needle aspiration in the management of pyogenic liver abscess. AJR Am J Roentgenol 189:W138–W142, 2007. 294. Zibari GB, Maguire S, Aultman DF, et al: Pyogenic liver abscess. Surg Infect (Larchmt) 1:15–21, 2000. 295. Zimmer PW, Hill M, Casey K, et al: Prospective randomized trial of talc slurry vs bleomycin in pleurodesis for symptomatic malignant pleural effusions. Chest 112:430–434, 1997.

69  Image-Guided Ablation of Parenchymal Organs Erik M. Velez and Muneeb Ahmed

INTRODUCTION Image-guided tumor ablation is a minimally invasive strategy to treat a range of focal tumors by inducing irreversible cellular injury through the application of thermal and more recently nonthermal energy or chemical injection. This approach is being used to treat a range of focal tumors, most commonly primary and secondary malignancies of the liver, lung, kidney, adrenal glands, and bone. Specific advantages of percutaneous tumor ablation therapies include the ability to treat patients with limited disease who are not surgical candidates owing to medical comorbidities, lower associated morbidity and mortality compared to conventional surgery or systemic chemotherapy, relative costeffectiveness, and equivalent or near-equivalent long-term outcomes in well-selected patient populations. Here, concepts on image-guided tumor ablation will be presented, including the principles and goals of tumor ablation and a brief review of different ablation modalities. Key steps involved in tumor ablation, including initial patient evaluation, intraprocedural imaging guidance and adjunctive techniques, and postprocedure imaging and care will be reviewed. Finally, current applications in common parenchymal organs such as the liver and kidney will be reviewed.

PRINCIPLES OF TUMOR ABLATION Goals of Tumor Ablation Minimally invasive focal tumor ablation encompasses several specific objectives. First, the main goal of tumor ablation (regardless of organ or primary tumor type) is inducing focal cytotoxic injury throughout the entire target tumor by placing needle-like applicators at the center of the tumor to generate high temperatures (>60°C) or an equivalent injury.2 The second objective is to treat an adequate ablative margin of tissue surrounding the target tumor. An appropriate ablative margin is defined as adequately treating a rim of apparently normal tissue around the tumor, which often contains microscopic invasion of malignant cells at the tumor periphery (Fig. 69-1). In most cases, this rim should be 5 to 10 mm thick, extending circumferentially beyond the edge of visible tumor,36 although less (e.g., for renal tumors)35 or more (e.g., for colorectal liver metastases)22,29 of a margin may be needed on a tumor-specific basis. This means that for 3- to 5-cm tumors, a single ablation treatment may not be sufficient to entirely encompass the target volume.8 In these cases, multiple overlapping ablations or simultaneous use of multiple applicators may be required to successfully treat the entire tumor and achieve an ablative margin.10 Additionally, if the target tumor (e.g., hepatic colorectal metastasis) has decreased in size from chemotherapy before ablation is being

2632

performed, the ablation zone should be sufficiently large to include the area originally involved by the tumor. Finally, although complete treatment of the target tumor is of primary importance, specificity and accuracy are also highly preferred, with a secondary goal of incurring as little injury as possible to surrounding nontarget normal tissue. This ability to minimize damage to normal organ parenchyma is one of the significant advantages of minimally invasive percutaneous thermal ablation and can be critical in patients who have focal tumors in the setting of limited functional organ reserve. Examples of clinical situations where this is relevant include focal hepatic tumors in patients with underlying cirrhosis and limited hepatic reserve, patients with Von Hippel-Lindau syndrome who have limited renal function and require treatment of multiple renal tumors, and patients with primary lung tumors with extensive underlying emphysema and limited lung function. Many of these patients are not surgical candidates owing to limited native organ functional reserve placing them at a higher risk for postoperative complications or organ failure. Careful planning is required to ensure that each of these three objectives is met.2

Choice of Ablation Modality Several types of ablation modalities with differing mechanisms of cellular injury are used in current clinical practice. These can be divided into technologies that induce a thermal injury through hightemperature heating (most commonly using radiofrequency [RF] and microwave energy and to a lesser degree ultrasound [US] or laser), extreme cooling (i.e., cryoablation), exposure to cytotoxic chemicals (e.g., ethanol or acetic acid instillation), or nonthermal energy with irreversible electroporation. All of these modalities are based upon a similar premise—delivery occurs through percutaneous placement of needles directly into the tumor using imaging guidance. Finally, operator technique likely has as much influence on performance as device technology; for example, multiple applicators can be used to increase heating rather than switching to a different energy source. A detailed discussion of different types of ablation modalities available for use falls beyond the scope of this chapter. However, it is important to be familiar with modalities that are commonly used in clinical practice, including RF- and microwave-based thermal ablation, cryoablation, and ethanol instillation. For each modality, mechanisms of cellular injury, advantages and disadvantages, and common clinical uses are presented in Table 69-1. Of these available technologies,  RF ablation (RFA) and microwave ablation are more ubiquitous in current clinical practice. Both types of ablation elevate tissue temperatures enough to create zones of irreversible cellular damage. RF energy is easy to generate but suffers in areas of high blood flow or high  tissue impedance and requires use of electrode switching or multiple

CHAPTER 69  Image-Guided Ablation of Parenchymal Organs

A

B

C

D

2633

FIG 69-1  Achieving an ablative margin. Images of a 63-year-old woman with cirrhosis and lobulated right hepatic lobe tumor with arterial enhancement (A) and contrast washout on delayed-phase MRI (B). C, Microwave probe positioned within the tumor. D, Contrast-enhanced CT at 1-month follow-up demonstrates hypoenhancing well-marginated ablation zone encompassing the tumor and a 1-cm ablative margin. An adequate ablation zone should always be larger than the original tumor on follow-up imaging.

applicator use to achieve large ablation zones in a time-efficient manner. Microwave heating is fast and efficient and thus appears better equipped to overcome heat sinks and treat large tumor volumes. In current clinical practice, RF-based platforms are most commonly used and have the largest set of relevant long-term clinical data (i.e., studies were performed with systems still used in clinical practice). However, there are now a number of commercially available microwave-based systems with increasing data becoming available, demonstrating at least equivalency to RFA and potentially superiority for larger tumors. This is particularly relevant in situations (i.e., hepatic colorectal metastases) where the need to achieve a 10-mm ablative margin necessitates larger ablation zones (e.g., 5-cm ablations for 3-cm tumors). Cryoablation and ethanol instillation also offer specific clinical advantages. Cryoablation procedures are largely painless, and the ablation zone is readily visible on noncontrast computed tomography (CT), allowing careful monitoring of the ablation zone during treatment in situations where greater accuracy is required. Ethanol instillation is one of the most cost-effective treatments available (requiring only a 21-gauge needle and dehydrated alcohol), though this is offset by poorer

long-term efficacy compared to thermal ablation. Finally, there are several less commonly used techniques with platforms that are not as well studied and with less overall clinical experience. These include thermal ablation using laser energy or US (either through focused energy platforms or with percutaneous applicators), which are cytotoxic using thermal injury similar to RFA or microwave ablation. Irreversible electroporation is an emerging technology that employs administration of electrical pulses to disrupt cellular membranes  (and is therefore often considered less of a thermal modality), though additional clinical and experimental studies using this modality are ongoing.

PERFORMING IMAGE-GUIDED TUMOR ABLATION Patient Evaluation Development of an outpatient clinical practice to evaluate all patients eligible for image-guided tumor ablation is strongly encouraged. A thorough clinical evaluation should be performed and recent crosssectional imaging of the patient should be reviewed to determine the

2634

PART III  Image-Guided Procedures

TABLE 69-1  Comparison of Ablation Modalities in Common Clinical Use Commonly Used Sites Within the Abdomen

Modality

Mechanism

Advantages

Disadvantages

RFA

High-temperature thermal injury (60-100°C) Friction induced by current oscillation around electrode Tissue injury dependent on diffusion of heat through tissues High-temperature thermal injury (60-150°C) Friction from water molecule agitation within tissues Active heat generation occurs farther/deeper in tissues; less dependence on thermal transmission Alternating freeze/thaw cycles induce injury with intracellular ice formation and vascular thrombosis Protein denaturation and cellular homeostasis changes

Established long-term clinical experience Device platforms with significant clinical experience Multiple electrodes can be used.

Susceptible to heat sink effects from blood flow cooling Requires two poles (electrode and grounding site)

Liver, kidney, adrenal, bone

Less susceptible to heat sink effects Larger ablation zones compared to similarly sized RFA platforms Multiple antennae can be used.

Greater individual variation between available platforms (less long-term clinical experience) Higher temperatures; potentially greater risk of nontarget thermal injury

Liver, kidney, adrenal, bone (similar to RFA but with overall fewer long-term studies)

Ablation zone easily visible on noncontrast CT; permits close intraprocedural monitoring Multiple probes can be used. Extensive clinical experience in some situations Very cost-effective Low-profile needles, minimally invasive delivery

Has been associated with systemic inflammation/side effects during hepatic ablation Limited to use for HCC Long-term studies demonstrate higher recurrence compared to RFA Limited uniformity of injury owing to differences in intratumoral diffusion

Kidney Liver Bone

Microwave ablation

Cryoablation

Ethanol instillation

tumor burden and evaluate adjacent structures. Optimally, baseline imaging should be performed within 1 month prior to ablation to ensure that treatment planning is accurate. Preprocedural laboratory studies should always include a complete blood cell count (CBC), platelet count, creatinine, and coagulation profile. Additional studies such as a blood chemistry panel, baseline tumor markers, electrocardiogram, and blood type and match may also be appropriate depending on the organ system and age of the patient. All anticoagulants and antiplatelet agents should be stopped or appropriately substituted with short-acting agents prior to treatment, with sufficient time for coagulation status to normalize. Contraindications to percutaneous ablation include excessive tumor burden, uncorrectable coagulopathy, active signs of infection, and untreatable diffuse or distant disease.15,28 Overall the primary goal for the initial patient consultation is to determine whether the patient is a good candidate for image-guided ablation. Important factors include performing a detailed clinical evaluation of the patient, determining if the tumor is amenable to treatment, and evaluating whether the primary tumor will be appropriately treated with ablation (or should be considered for alternative therapies). In this regard the combined expertise of an interdisciplinary oncology team including medical oncology, oncologic surgery, hepatology, radiation therapy, transplant surgery, pathology, radiology, and interventional oncology is critical in the development of treatment plans for cancer patients. Given the wide range of healthcare members involved, multidisciplinary meetings are useful to triage complex patients.

Definitions of Ablation Success Familiarity with key standardized definitions of ablation success and failure is essential to prospectively determining treatment plans and

Liver (HCC)

follow-up and adequately interpreting and applying published clinical data.3 Technical success is achieved when a tumor is treated according to the prospective plan/protocol and covered completely (i.e., ablation zone completely overlaps or encompasses target tumor plus an ablative margin) as determined at the time of the procedure. A predefined course of treatment may include several ablation procedures spaced out over time. Primary technical success should be determined on the first follow-up imaging study after completion of the predetermined course of treatment. Distinction between “technical success” and “technique efficacy” must be made for each treated tumor. Efficacy can only be demonstrated with appropriate clinical follow-up. Technique efficacy should therefore refer to a prospectively defined time point (e.g., 1 month after treatment) at which complete ablation of macroscopic tumor, as evidenced by imaging follow-up (or another specified end point), is achieved. Primary efficacy is achieved for those target tumors successfully eradicated following the initial procedure or a defined course of treatment, whereas secondary or assisted efficacy refers to those tumors that have undergone successful repeat ablation following identification of local tumor progression. The term retreatment should be reserved for describing ablation of locally progressive tumor in cases where complete ablation was initially thought to have been achieved based upon imaging demonstrating adequate ablation of the tumor. When initial follow-up imaging demonstrates residual tumor at the ablative margin, this is referred to as residual unablated tumor. Local tumor progression describes the appearance of tumor foci at the edge of the ablation zone after at least one contrast-enhanced follow-up study has documented adequate ablation and an absence of viable tissue in the target tumor and surrounding ablation margin by imaging criteria.

CHAPTER 69  Image-Guided Ablation of Parenchymal Organs This term applies regardless of when tumor foci were discovered, either early or late in the course of imaging follow-up.

Intraprocedural Image Guidance Percutaneous tumor ablation procedures are commonly performed under US, CT (with or without fluoroscopy), or combination US and CT guidance. Magnetic resonance imaging (MRI) guidance has also been described but is not currently in widespread use because of limitations in access to such facilities beyond specialized centers. Successful ablations are contingent upon the operator’s ability to visualize the tumor, position the electrode within the target, and accurately evaluate the treatment zone upon ablation completion, with several significant challenges at each step. For instance, previous diagnostic imaging is often obtained with modalities that are different from the modality being used for treatment (e.g., diagnostic MRI and CT-guided ablation) and cannot be used as an exact overlay owing to patient positional variability. Additionally, target tumors often have different conspicuity with or without contrast on US, CT, and MRI. Ultimately, selection of the appropriate imaging modality depends on optimal tumor visibility. Real-time US guidance is very useful in electrode placement and positioning during the procedure but may have limited visibility of tumors in certain locations such as the hepatic dome. Additionally, US provides limited information on monitoring success of the procedure as the area of treatment becomes obscured by echogenic foci of gas generated by the ablation. CT with or without fluoroscopy can be helpful in targeting deep tumors or in cases where a difficult access trajectory is required. However, target tumors are often difficult to visualize on noncontrast CT images. Intravenous contrast can also be administered at the time of the procedure to confirm target location. Careful planning of probe positioning is required to ensure that the end ablation zone sufficiently encompasses the target tumor. This includes planning initial placement and/or repositioning of the probe or placement of multiple probes simultaneously, with appropriate consideration of the size of ablation zone that will be achieved around each probe. It is important to remember that a number of overlapping ablations will be required to adequately cover the three-dimensional (3D) volume of larger tumors.8 When multiple probes are used, attention to parallel probe orientation and appropriate spacing is required (and is specific to device type). In cases where a biopsy is necessary, a guiding needle can be used through which both biopsy and ablation can be performed, especially beneficial in deep and difficult-to-access tumors while minimizing the number of puncture/passes that are needed. For RF devices, care should be taken to ensure that the guiding needle is either electrically insulated or short enough that the active RF electrode tip does not touch the guiding needle (which would result in electrical short circuiting).

Adjunctive Techniques Structures near the ablation zone, such as the colon, stomach, gallbladder, diaphragm, pancreas, and abdominal wall, can be injured from adjacent heating. Often, creating a sufficient separation between the anticipated edge of the ablation zone and the nearby critical structure is required to complete the ablation safely. Ideally a 5- to 10-mm distance is required, though this may vary when fat or air insulation is present around the ablation zone (e.g., for exophytic tumors surrounded by mesenteric or retroperitoneal fat). There are several techniques that can be used to successfully create a margin of safety around the ablation zone, referred to as either displacement or dissection. This can be done through placement of a separate 18- to 19-gauge needle into the gap between the two structures, followed by injection of air (either room air or CO2) or fluid until an appropriate distance is

2635

achieved. When RFA is being used, only nonionic fluids should be used (sterile water or 5% dextrose in water) to prevent asymmetric transmission of electric current. Contrast (usually 2%-5%) can be added to improve fluid visibility and demonstrate adequate tissue separation7 (Fig. 69-2). Artificial ascites can be created using a Yueh needle placed under US guidance into the perihepatic space. Careful monitoring during the procedure is required, with reinjection as necessary to maintain the distance throughout the procedure. Finally, balloon  catheters can also be placed over a wire to assist with mechanical displacement when fluid or air dissection is insufficient.17 Use of 18- to 19-gauge needles for initial air/fluid injection allows placement of 0.035-inch wires and over-the-wire balloon catheter placement if needed.

Postprocedure Imaging and Follow-up Immediate postablation imaging serves two major roles: to evaluate  for adequacy of treatment (or conversely, identify areas of incomplete treatment) and to evaluate for procedural complications. At the author’s institution, an on-table multiphase contrast-enhanced CT of the liver is performed immediately after ablation. Contrast-enhanced ultrasound (CEUS), where available, can also be used to assess immediate efficacy of ablation. Use of contrast allows for rapid evaluation of the ablation zone and helps identify the presence of any residual unablated tumor, active hemorrhage, pneumothorax, or adjacent organ injury. If any enhancing tumor is identified, electrodes are repositioned and the remaining tumor is ablated. As noted earlier, the ablation zone should encompass the treated tumor as well as a 5- to 10-mm ablative margin.31 The shape of the ablation zone can vary with the number and type of ablation electrodes, as well as the number and size of surrounding blood vessels producing a “heat sink” effect. Thus it is important to evaluate the ablative margin using multiplanar reformats. Recent studies also suggest that more accurate assessments of the ablative margin can be performed using 3D registration of postablation CT and MRI, which may offer a new imaging paradigm for post­ ablation assessment.33 Imaging findings include a hypoattenuating or heterogeneously hyperattenuating ablation zone due to coagulative necrosis and hemorrhagic products.31 When cryoablation is performed, the treated area appears hypodense relative to surrounding soft tissue. Because of the tissue necrosis, it is normal to find tiny air bubbles immediately after ablation on CT, and this usually resolves within 24 hours. However, the formation of an abscess should be considered if the air bubbles persist or increase on subsequent imaging and the patient has suggestive symptomatology, such as persistent abdominal pain, fever, or leukocytosis.20,32 Over time, the ablation zone becomes more homogenously hypoattenuating on noncontrast CT as the ablation-induced coagulation forms fibrous tissue, which gradually reduces in size over a 6- to 12-month interval.11 Periablational edema manifests as a sharply marginated low-attenuation rim on unenhanced CT and as a uniform thin rim of hyperenhancement on contrast-enhanced CT images, representing reactive inflammation that can last up to 4 weeks after ablation18,21 (Fig. 69-3). Long-term follow-up imaging is required at regular intervals to determine completeness of ablation and screen for local tumor progression or new sites of tumor. Specific protocols often depend on the type of tumor being treated. A commonly used follow-up regimen includes imaging at 1 month and then every 3 months after the procedure indefinitely, given the potential for local tumor progression or new tumor development. Either a CT or MRI with multiphasic contrast enhancement (including noncontrast, arterial, and venous phases) can be sufficient. Finally, detecting residual tumor with conventional contrast-enhanced imaging can be challenging and dependent on

2636

PART III  Image-Guided Procedures

A

B

C

D

FIG 69-2  Adjunctive techniques to separate nontarget structures

E

from the ablation zone. A-C, A 54-year-old man with left renal cell carcinoma for cryoablation. A, Contrast-enhanced axial CT image demonstrating descending colon immediately adjacent to the renal tumor. B, A 19-gauge needle was placed in the perirenal fat between the renal tumor and colon, with injection of 5% iodinated contrast/sterile water to create adequate separation. C, Axial CT image after 10 minutes of cryoablation, demonstrating an appropriate safety margin between the ablation zone and colon. D and E, A 67-year-old man with cirrhosis and a subcapsular hepatocellular carcinoma. D, Axial arterial-phase T1-weighted contrast-enhanced MRI demonstrates an arterial enhancing 2-cm hepatic tumor adjacent to the right adrenal gland. E, A 19-gauge needle was placed between the liver capsule and the right adrenal gland, and air was injected for pneumodissection during RFA.

CHAPTER 69  Image-Guided Ablation of Parenchymal Organs

2637

tion of all known tumor cells within the index tumor[s] and without any other known tumor foci in the body) or palliative intent (i.e., complete ablation of the index tumor[s] with other known nontarget tumor foci within the body or complete or partial ablation to  treat sufficient portions of the index tumor[s] to achieve symptom relief).6,23,30,34 Tumor ablation for curative intent is commonly performed for isolated focal primary malignancies in the liver and kidney. Complete ablation of symptomatic benign tumors (e.g., hormonally active benign adrenal aldosteronomas) to achieve symptomatic relief can also be considered curative. In contrast, palliative ablation has been performed for pain relief and to treat hormonally active metastatic tumors.

A

B

C FIG 69-3  Benign periablational enhancement. A, Arterial-phase axial MRI demonstrating an arterial enhancing tumor in the setting of hepatic cirrhosis. B, Delayed-phase axial image demonstrating contrast wash­ out, consistent with hepatocellular carcinoma. C, Immediate postablation arterial-phase contrast-enhanced CT image demonstrates uniform hyperenhancement circumferentially around the ablation zone, consistent with benign periablational enhancement rather than residual tumor.

changes that occur over several follow-up imaging studies, which can delay reintervention. Fluorine-18 fluorodeoxyglucose (18F FDG) positron emission tomography (PET)/CT imaging can be helpful for early confirmation of suspected recurrence on conventional contrastenhanced imaging.

CLINICAL APPLICATIONS FOR TUMOR ABLATION Clinical indications for tumor ablation are divided into ablations performed for curative intent (i.e., achieving the goal of complete eradica-

Liver Image-guided ablation has been used extensively for primary and secondary hepatic malignancies. For hepatocellular carcinoma (HCC), thermal ablation (most commonly using RF) has been used in a substantial number of clinical series. The two largest factors in that affect RFA are tumor size and the presence of large (>3 mm) abutting vessels. Complete tumor necrosis was shown in 83% of tumors less than 3 cm and 88% of tumors in nonperivascular regions.26 A randomized controlled trial comparing the effectiveness of RFA and surgical resection as initial treatment of solitary HCC less than 5 cm demonstrated  1-, 2-, and 3-year overall survival rates of 95.8%, 82.1%, and 71.4%, respectively, in the ablation group and 93.3%, 82.3%, and 73.4%, respectively, in the resection group.9 Tumor ablation is more costeffective and has a lower associated morbidity and mortality compared to conventional surgery or systemic chemotherapy.2,12 Given the success of ablation for early and very early HCC (<2.5 cm in diameter), several recent societal guidelines now recommend it as first-line curative  treatment. In addition, in many patients with cirrhosis awaiting liver transplantation, thermal ablation is used as a “bridge-to-transplant” temporizing therapy (particularly in those parts of the world where wait times for transplantation are longer). Several studies have shown a decrease in dropout rate using this technique, with one demonstrating a dropout rate of 5.7% after 1 year compared to the 31.8% dropout rate in untreated patients.13,25 Tumor ablation can also be combined with other locoregional treatments such as transarterial chemoembolization (TACE) or yttrium-90 transarterial radioembolization. Such therapies offer differing but complementary options that can be used separately, sequentially, or in combination to optimally treat liver tumors. TACE and radioembolization are typically chosen for patients with higher tumor burdens and advanced disease or as a temporary therapy in patients awaiting liver transplant. Percutaneous ablation is preferred for smaller tumors (<4-5 cm in diameter) with fewer than three foci of discrete tumors. However, multimodality treatments combining several approaches tailored to patient and tumor-specific characteristics are increasingly being offered, with respectable long-term outcomes. In particular, combination treatment of larger (3-7 cm) HCCs with TACE followed by thermal ablation have proven to be effective compared to ablation alone (Fig. 69-4). Tumor ablation has also been used to treat isolated or oligometastatic disease of the liver, with the largest experience in patients with colorectal liver-only disease who (1) do not meet criteria for surgical resection but would still benefit from liver-directed therapy, (2) have developed new sites of tumor after having undergone resection, or (3) in whom RFA is performed as an adjunct to staged hepatectomy. Patients with unresectable but limited disease who undergo RFA have better outcomes compared to chemotherapy alone.1 Recent ablation studies have reported improved 5-year survival outcomes more closely approximating surgical resection, in part due to improvements in

2638

PART III  Image-Guided Procedures

A

B

C FIG 69-4  Combined transarterial lipiodol chemoembolization followed by CT-guided RFA. A, A 3.5-cm arterial enhancing hepatocellular carcinoma on axial arterial-phase MRI. B, Contrast-enhanced axial CT image immediately after TACE with RFA demonstrates central hyperdense material (intratumoral uptake of lipiodol-doxorubicin emulsion) with a hypodense rim consistent with the ablation zone after superimposed RFA. C, Three-month follow-up subtracted arterial-phase axial MRI demonstrates complete treatment with no residual tumor enhancement.

patient selection, more accurate and improved follow-up, advances  in chemotherapy regimens, and immediate retreatment of persistent or untreated tumor.34 A review of the existing literature on RFA for hepatic colorectal metastases remains mixed; many studies comparing percutaneous therapy to surgical resection contain inherent selection biases (RFA is often performed in nonsurgical patients). Nevertheless, Gillams et al. report their long-term outcomes in a subset of wellselected patients (tumors < 5 cm, tumor number < 5, no extrahepatic disease) where 5-year survival was between 24% and 33%, approaching that of surgical resection.16 Additionally, although higher rates (≈15%-20%) of local tumor progression can occur after RFA compared to surgical resection, this is compensated for with the ability to easily re-treat residual tumor, resulting in equivalent 5-year survival rates (48%-51%). Finally, Solbiati et al. recently reported their 10-year survival data in 99 patients who underwent RFA for small colorectal liver metastases (<3 cm, <5 in number), with excellent results (48% at 5 years, 25% at 7 years, 18% at 10 years) easily comparable to resection.34 In a classic study in the ablation literature, Livraghi et al. applied a “test-of-time” approach using RFA as first-line treatment in patients who were otherwise eligible for surgery.24 Patients with solitary hepatic metastases smaller than 4 cm in diameter underwent initial RFA rather than surgical resection. They found a significant portion of their patients developed new intra- and extrahepatic disease, and as such, 50% were spared noncurative surgery, 26% of patients were spared surgery because of curative RFA, and no patients lost eligibility for hepatic resection. More recently, Gillams et al. demonstrated 3- and 5-year survival rates of 63% and 34%, respectively, in a series of 309 patients who underwent RFA for hepatic colorectal metastases (tumors < 5 cm and no extrahepatic disease).16 In a smaller series of 40 patients by the same group, treating smaller tumors (mean, 2.3 cm) had even better outcomes (1-, 3-, and 5-year survival rates of 97%, 84%, and 40%, respectively).

Kidney Although nephron-sparing surgery is commonly used to treat small primary renal tumors, tumor ablation has become an attractive alternative to surgical resection. Many patients are deemed unsuitable for surgical resection because of limited renal function, transplanted kidney, significant medical comorbidities, and the desire to avoid renal replacement therapy, therefore becoming ideal ablation candidates. Confirmed pathology of renal cell carcinoma by needle biopsy is strongly recommended prior to ablation because up to 25% of renal masses less than 3 cm are benign. RFA and cryoablation have shown high technical success rates of 97% to 100% in a wide range of tumor sizes; however, multiple ablation sessions are required in larger tumors.4,5 In one series the metastasis-free survival rate and overall 2-year survival rate of RFA for renal cell carcinoma was 97% and 82.5%, respectively.14 Cryoablation demonstrated similar survival rates in a patient series with a wide range of tumor sizes and a mean follow-up of less than 2 years.27 Major complications for percutaneous ablation of renal tumors include hemorrhage, pneumothorax, and bowel and nerve injury, which occur at a rate approximating 2%.19

Adrenal Gland Preliminary studies have also reported use of thermal ablation in the treatment of isolated metastases in the adrenal gland. Studies on surgical resection of isolated adrenal metastases have reported benefits on long-term patient outcomes and have formed the basis for use of thermal ablation in such circumstances as well. In the largest series to date, Welch et al. treated 37 adrenal metastases in 32 patients with cryoablation, with local tumor control achieved in 91.2% of patients.

CHAPTER 69  Image-Guided Ablation of Parenchymal Organs

A

B

C

D

2639

FIG 69-5  Ablation of an isolated lung cancer metastasis to the left adrenal gland. A 69-year-old woman with a resected primary non–small cell lung cancer with a new FDG-avid 12-mm left adrenal nodule on axial noncontrast CT (A) and corresponding PET (B) images. C, Axial intraprocedural CT image demonstrating RFA electrode placement through the nodule while the patient is in a prone position. D, Follow-up 3-month subtracted contrast-enhanced axial MRI demonstrates complete treatment without evidence of residual arterial enhancing tumor.

However, 32% of patients ultimately died from extraadrenal metastases.37 This suggests that although early data are promising, additional studies are required to refine patient selection to maximize treatment outcomes. There are also several studies reporting experience using thermal ablation to treat benign hormone-producing adrenal tumors (e.g., aldosterone- or corticosteroid-producing adenomas). With careful selection of patients, including appropriate laterality studies with adrenal venous sampling, ablation of functional adrenal adenomas is a promising developing clinical indication. Several series have reported success rates of over 90% (using clinical and/or biochemical outcome measures) (Fig. 69-5). Even when compared to minimally invasive surgery (e.g., laparoscopic adrenalectomy), percutaneous imageguided ablation offers advantages of reduced complication rates and short hospital stays.38

CONCLUSION In conclusion, the purpose of this chapter has been to review key concepts regarding image-guided tumor ablation. Key objectives of tumor ablation and different types of commonly used ablation modali-

ties have been discussed. An overview of components to successful ablation, including patient assessment and intra- and postprocedural imaging evaluation, is provided. Finally, common parenchymal applications of tumor ablation for liver, kidney, and adrenal tumors are reviewed.

REFERENCES 1. Abdalla EK, et al: Recurrence and outcomes following hepatic   resection, radiofrequency ablation, and combined resection/ablation   for colorectal liver metastases. Ann Surg 239:818–825, discussion 825–827, 2004. 2. Ahmed M, et al: Principles of and advances in percutaneous ablation. Radiology 258:351–369, 2011. 3. Ahmed M, et al: Image-guided tumor ablation: Standardization of terminology and reporting criteria–A 10-year update. Radiology 273:241–260, 2014. 4. Atwell TD, et al: Percutaneous renal cryoablation: Experience treating 115 tumors. J Urol 179:2136–2140, discussion 2140–2141, 2008. 5. Boss A, et al: Magnetic resonance-guided percutaneous radiofrequency ablation of renal cell carcinomas: A pilot clinical study. Invest Radiol 40:583–590, 2005.

2640

PART III  Image-Guided Procedures

6. Callstrom MR, et al: Percutaneous image-guided cryoablation of painful metastases involving bone: Multicenter trial. Cancer 119:1033–1041, 2013. 7. Campbell C, et al: Contrast media-doped hydrodissection during thermal ablation: Optimizing contrast media concentration for improved visibility on CT images. AJR Am J Roentgenol 199:677–682, 2012. 8. Chen MH, et al: Large liver tumors: Protocol for radiofrequency ablation and its clinical application in 110 patients—Mathematic model, overlapping mode, and electrode placement process. Radiology 232:260–271, 2004. 9. Chen MS, et al: A prospective randomized trial comparing percutaneous local ablative therapy and partial hepatectomy for small hepatocellular carcinoma. Ann Surg 243:321–328, 2006. 10. Dodd GD, 3rd, et al: Radiofrequency thermal ablation: Computer analysis of the size of the thermal injury created by overlapping ablations. AJR Am J Roentgenol 177:777–782, 2001. 11. Dromain C, et al: Hepatic tumors treated with percutaneous radiofrequency ablation: CT and MR imaging follow-up. Radiology 223:255– 262, 2002. 12. El-Serag HB: Hepatocellular carcinoma. N Engl J Med 365:1118–1127, 2011. 13. Freeman RB, Edwards EB, Harper AM: Waiting list removal rates among patients with chronic and malignant liver diseases. Am J Transplant 6:1416–1421, 2006. 14. Gervais DA, et al: Radio-frequency ablation of renal cell carcinoma: Early clinical experience. Radiology 217:665–672, 2000. 15. Geschwind J-FH, Michael C: Soulen e. Interventional Oncology, 2008, Cambridge University Press. 16. Gillams AR, Lees WR: Five-year survival in 309 patients with colorectal liver metastases treated with radiofrequency ablation. Eur Radiol 19:1206–1213, 2009. 17. Ginat DT, Saad WE: Bowel displacement and protection techniques during percutaneous renal tumor thermal ablation. Tech Vasc Interv Radiol 13:66–74, 2010. 18. Goldberg SN, Gazelle GS, Mueller PR: Thermal ablation therapy for focal malignancy: A unified approach to underlying principles, techniques, and diagnostic imaging guidance. AJR Am J Roentgenol 174:323–331, 2000. 19. Johnson DB, et al: Defining the complications of cryoablation and radiofrequency ablation of small renal tumors: A multi-institutional review. J Urol 172:874–877, 2004. 20. Lim HK, et al: Hepatocellular carcinoma treated with percutaneous radio-frequency ablation: Evaluation with follow-up multiphase helical CT. Radiology 221:447–454, 2001. 21. Lim HS, et al: Imaging features of hepatocellular carcinoma after transcatheter arterial chemoembolization and radiofrequency ablation. AJR Am J Roentgenol 187:W341–W349, 2006.

22. Liu CH, et al: Radiofrequency ablation of hepatic tumours: Effect of post-ablation margin on local tumour progression. Eur Radiol 20:877–885, 2010. 23. Livraghi T, et al: Hepatocellular carcinoma and cirrhosis in 746 patients: Long-term results of percutaneous ethanol injection. Radiology 197:101–108, 1995. 24. Livraghi T, et al: Percutaneous radiofrequency ablation of liver metastases in potential candidates for resection: The “test-of-time approach”. Cancer 97:3027–3035, 2003. 25. Lu DS, et al: Percutaneous radiofrequency ablation of hepatocellular carcinoma as a bridge to liver transplantation. Hepatology 41:1130–1137, 2005. 26. Lu DS, et al: Radiofrequency ablation of hepatocellular carcinoma: Treatment success as defined by histologic examination of the explanted liver. Radiology 234:954–960, 2005. 27. Maybody M: An overview of image-guided percutaneous ablation of renal tumors. Semin Intervent Radiol 27:261–267, 2010. 28. McCarley JR, Soulen MC: Percutaneous ablation of hepatic tumors. Semin Intervent Radiol 27:255–260, 2010. 29. Mulier S, et al: Local recurrence after hepatic radiofrequency coagulation: Multivariate meta-analysis and review of contributing factors. Ann Surg 242:158–171, 2005. 30. Psutka SP, et al: Long-term oncologic outcomes after radiofrequency ablation for T1 renal cell carcinoma. Eur Urol 63:486–492, 2013. 31. Sainani NI, et al: Imaging after percutaneous radiofrequency ablation of hepatic tumors: Part 1, Normal findings. AJR Am J Roentgenol 200:184– 193, 2013. 32. Sainani NI, et al: Imaging after percutaneous radiofrequency ablation of hepatic tumors: Part 2, Abnormal findings. AJR Am J Roentgenol 200:194–204, 2013. 33. Sakakibara M, et al: Three-dimensional registration of images obtained before and after radiofrequency ablation of hepatocellular carcinoma to assess treatment adequacy. AJR Am J Roentgenol 202:W487–W495, 2014. 34. Solbiati L, et al: Small liver colorectal metastases treated with percutaneous radiofrequency ablation: Local response rate and long-term survival with up to 10-year follow-up. Radiology 265:958–968, 2012. 35. Stone MJ, et al: Radiofrequency ablation of renal tumors. Tech Vasc Interv Radiol 10:132–139, 2007. 36. Wang X, et al: Margin size is an independent predictor of local tumor progression after ablation of colon cancer liver metastases. Cardiovasc Intervent Radiol 36:166–175, 2013. 37. Welch BT, et al: A single-institution experience in image-guided thermal ablation of adrenal gland metastases. J Vasc Interv Radiol 25:593–598, 2014. 38. Wolf FJ, et al: Adrenal neoplasms: Effectiveness and safety of CT-guided ablation of 23 tumors in 22 patients. Eur J Radiol 81:1717–1723, 2012.

70  Vasculogenesis Nicholas L. Fulton and John R. Haaga

INTRODUCTION The traditional concept of vasculogenesis in cancer maintains that malignant tumors prefer aerobic metabolism for growth. This theory posits that after continued proliferation, the tumor will “outgrow” its arterial vascular supply; as the resulting hypoxia ensues, the tumor will become dormant. The traditional vasculogenesis theory therefore states that tumors will reinitiate growth only after arteriogenesis occurs, allowing adequate blood flow to restore normoxia.42 Unfortunately, tumor pathogenesis is not nearly this straightforward. Were the traditional vasculogenesis theory completely accurate, antiarterial tumor therapies would have profound success rates. Scientific research seems to support a newly proposed theory for tumor vasculogenesis, the ALPHA hypothesis (Acidic Lactate sequentially induces Lymphangiogenesis, PHlebogenesis, and Arteriogenesis). Because the published data exist in diverse subspecialties throughout all of the scientific literature, complementary findings that support the ALPHA hypothesis went overlooked. With the new ALPHA hypothesis for tumor vasculogenesis, many imaging findings associated with malignancies can now be explained. Using this new knowledge, there is the potential for new and diverse oncologic treatments.

Traditional Vasculogenesis Theory and Its Inconsistencies The traditional vasculogenesis theory for tumor development dates back to the 1970s. In a historical paper published by Gimbrone in 1972, it was concluded that implanted rabbit iris tumors increased in size in conjunction with increased vasculogenesis and subsequent perfusion.55 Because of these findings, Gimbrone et al. concluded that “neovascularization is a necessary condition for malignant growth of a solid tumor.” However, their analysis did not consider the presence of lymphatics or veins. Additionally, reevaluation of Gimbrone’s statistics suggests that the tumor growth likely preceded arterial development by up to a day (Vera Chankung, PhD, Professor, and Norman Tien, PhD, Dean and Professor, Department of Engineering, Case Western Reserve University, personal communication, 2009). If this is true, it is impossible for arteriogenesis to be the catalyst for tumor growth, and therefore tumor growth must be dependent on other factors (Table 70-1, Fig. 70-1). The traditional vasculogenesis theory also holds that cancers prefer aerobic metabolism. After a certain level of growth, typically described as beyond 1 to 2 mm or 105 cells, hypoxia occurs as the tumor outgrows its arterial supply. Arteriogenesis then follows in order to restore  normoxia for continued tumor growth. Were this true, it would be expected that antiarterial therapies for cancer would have prolific success rates. Unfortunately, this has not been the case. Surgical arterial ligature has repeatedly failed to result in improved survival out-

2642

comes.60,136 Similarly, bland particle embolization has been shown to have an effect on slowing tumor growth but without improving patient survival.16 Drugs (e.g., bevacizumab) targeting vascular endothelial growth factor (VEGF) have had only limited success.76 Even more harrowing than treatment failure, studies have shown that anti-VEGF treatment failure may result in increased tumor aggressiveness related to increase in lactate levels (likely a byproduct of glycolysis) despite a significant reduction in blood volume and reduction in blood vessels89 (Figs. 70-2 and 70-3). Other studies of anti-VEGF therapies have shown that although primary tumor growth may be inhibited, antiVEGF therapies were associated with an increase in satellite tumors surrounding the initial mass.102 If arteriogenesis is not the catalyst behind tumor growth and development, clearly something else must be responsible. The traditional vasculogenesis theory held that cancers were aerobic, using oxidative phosphorylation to convert a molecule each of glucose and oxygen to create 38 molecules of adenosine triphosphate (ATP). Instead it has been theorized that cancer prefers to use glycolysis for energy production. For the same glucose cost, glycolysis only results in 2 ATP (<10% of that produced by aerobic metabolism) as well as the waste product lactic acid. Although the glycolytic pathway appears to be less efficient compared to aerobic metabolism, cancer cells preferentially use glycolysis for energy production.183 The hypothesis that cancer prefers glycolysis is by no means new—in fact it was proposed by Warburg in the 1950s—but the glycolytic basis of tumor development has continued to confound scientists and researchers.190 Hypoxia has been shown to exist in multiple tumors.94 Even with oxygen readily available, cancer prefers a glycolytic method of metabolism.34 With these results in mind, the ALPHA hypothesis was proposed to help explain hypoxia and its relation to tumor growth and metastasis, the multiple roles of the glycolytic waste product lactate, and vasculogenesis, including the formation of lymphatics and veins.

Principles of ALPHA Vasculogenesis #1: Complete vasculogenesis includes lymphangiogenesis, phlebogenesis, and arteriogenesis. Venous flow must precede arterial flow, but lymphangiogenesis can precede arterial flow. Though its significance has previously been underappreciated, the development of lymphatics is now known to play a vital role in cancer development. Tumors are associated with a large volume of free interstitial fluid.184 Buildup of this fluid may result in a higher pressure gradient between the arterial and venous systems, ultimately leading to diminished or functionally obstructed arterial flow. Cancer glycolysis results in accumulation of the waste product lactic acid within the extracellular spaces. This results in an increase in the concentration of the interstitial space. Simultaneously the expenditure of glucose during glycolysis decreases the concentration of the

2643

CHAPTER 70  Vasculogenesis TABLE 70-1  Reanalysis of the Gimbrone Data Experiment Number

Prevascular Growth Curve Slope (a)

Vascular Growth Curve Slope (b)

Late Growth Curve Slope (c)

Primary Inflection (day)

Vascularization (day)

0.04 0.05 0.08 0.03 0.04 0.16 0.08 0.19 0.06 0.12 0.085 ± 0.055

0.49 0.68 0.58 0.96 0.87 0.54 0.51 0.70 0.64 0.56 0.65 ± 0.155

0.23 0.14 — 0.14 0.25 0.13 0.21 0.15 0.12 0.12 0.16 ± 0.051

5-7 5-6 5-6 6-7 6-7 4-6 6-7 5-6 5-6 5-7

7 6 6 7 7 6 7 6 6 7

1 2 3 4 5 6 7 8 9 10 Mean ± SD

This table demonstrates rates of growth and time of vascularization of 10 iris tumors. The growth curve slopes were estimated from semilogarithmic plots. Vascularization was defined as positive fluorescein tests or gross appearance of tumor vessels. (Data courtesy Vera Chankung, PhD, Professor, and Norman Tien, PhD, Dean and Professor, Department of Engineering, Case Western Reserve University, 2009.)

1000 slope c

Tumor volume (mm 3)

100

10.0

slope b

1.0

Positive fluorescein test

0.1

slope a

Days 0

2

4

Phase I Prevascular

6

8 Phase II Vascular

10

12

14

Phase III Late

FIG 70-1  Reanalysis of the Gimbrone data suggests with 95% certainty that tumor growth preceded the development of arteries. (Data from Vera Chankung, PhD, Professor, and Norman Tien, PhD, Dean and Professor, Department of Engineering, Case Western Reserve University, personal communication, 2009.)

intracellular compartment.154 The resulting concentration difference between the intracellular and extracellular compartments is managed through the formation of aquaporins, a channel in the cellular membrane through which water can freely diffuse.128 Diffusion of water into the extracellular space restores the osmotic balance. However, it is not the osmotic disturbances themselves that stimulate the production of aquaporins.65 Lactic acid has been shown to be a catalyst for the development of aquaporins, and a higher number of aquaporins are detected within tumors, which we will see are highly correlated with lactate.125,156 Additionally the presence of aquaporins may be responsible for the cerebral edema associated with brain tumors.188 Therefore glycolysis results in a rise in interstitial fluid pressure due to its waste product lactate and eventual aquaporin formation. Aquaporins may also help relieve intracellular lactic acid buildup.65 As the interstitial fluid (and therefore interstitial fluid pressure) increase, the interstitium requires tumoral lymphangiogenesis for adequate drainage for homeostatic control.161 Lymphatic proliferation has been identified in tumors even at very small sizes.153 It has also been demonstrated that lymphatic formation precedes metastasis in melanoma, reflecting the integral nature of early homeostasis in cancer development.63 #2: Vasculogenesis occurs in small normoxic tumors before they outgrow their blood supply because of the angiogenic effects of lactate, similar to normal wound healing. In the very early stages of cancer development, tumors may grow independently from the aid of neovasculature. When still at a very small size, the intercellular distances are short enough that oxygen can travel freely between cells. However, upon reaching a size of approximately 1 to 2 mm, or 105 to 106 cells, additional support is required for continued cancer growth. This support comes in the form of newly developed vasculature.40 Tumoral vasculogenesis occurs in two phases: an early incipient normoxic phase and a later hypoxic phase. The initial angiogenic burst occurring in normoxic conditions is mediated by the release of VEGF and fibroblast growth factor (FGF). Angiogenesis and tumor growth have been shown to be induced in dormant MOLT-3 cells (a T-lymphoblast cell line) after pellets embedded with VEGF and FGF were injected, indicating a key role for angiogenesis in continued tumor growth.75 Additionally, previously dormant

2644

PART IV  Leading Edge Imaging Concepts

AI A

AI

G G

A

A

B FIG 70-2  A, Prior to embolization therapy, there were two hepatic tumors (A). B, Months after embolization, the infarcted masses (AI) have reduced in size. Unfortunately, two new large masses have developed (G), fueled by glycolysis.

tumors, upon stimulation to reintroduce angiogenesis, return in a much more aggressive form. Following this initial phase, there is the hypoxic phase, which is multifactorial with various causes ranging from increased distances between tumor cells and arteries, elevated interstitial and capillary pressure, temporal and spatial variation in flow, blood viscosity, thrombosis, and vascular shunts. Hypoxia actually has very complex interplay with cancers, possibly resulting in growth inhibition and cell death while conversely also being a key component in tumor growth, local invasion, and metastatic progression. Phenotypic alterations as a consequence of hypoxia within the microenvironment are closely linked to metastasis.151,185 Irradiation of the tumor bed in animal models of melanoma has resulted in increased metastases, which reflects what is often seen clinically: patients with recurrent malignancy after radiation therapy tend to have increased local disease and distant metastases. The extent of metastasis is also proportional to the radiation dose. Not surprisingly, the pre-irradiated tumor beds also showed significant hypoxia compared with the unradiated beds. As described by Vaupel et al., tumor angiogenesis first begins with vein formation, largely due to the need for waste removal (Fig. 70-4). Although initially host venules may be incorporated into the tumor, continued growth of the tumor and lack of associated growth within the native vessels exhausts the host vasculature. Eventually the incorporated host veins become obliterated or otherwise inadequate for continued tumor growth. Neovascularization then begins, with veins being produced prior to arteries to maintain adequate outflow. Early on, multiple anastomoses between veins may actually result in veins being able to contribute blood inflow in the manner usually conducted by arteries.184 Visualizing tumor cells directly labeled with green fluorescence protein (GFP) and dorsal skin folds in a rodent model. Li et al. described the early phases of angiogenesis, beginning with malignant cells migrating toward the normal host vasculature112 (Fig. 70-5). The next angiogenic phase seen involved dilatation and tortuosity of veins. This finding occurred after 6 days, as the tumor cells had proliferated from 20 to 50

cells to approximately 60 to 80 cells. No change in veins were noted in the nontumor cell line. Neovasculature within the tumor cell line was noted at day 8, approximately the 100- to 300-cell stage of growth. At this early stage of cell proliferation (and far below the 105 cell threshold previously alluded to), it can be assumed that this local microenvironment maintains normoxia at the onset of vasculogenesis. The low extracellular pH seen in early tumor development, mediated by lactic acid, results in activation of phospholipase D and the subsequent expression of matrix metalloproteinase 9 (MMP-9).86 MMPs are important enzymes in tumor development, resulting in breakdown of the extracellular matrix, tumor cell attachment by cleavage of integrins, and migration by removal of E-cadherin.168 Lactate and MMPs are also essential for tumor migration. The acidic extracellular environment results in expression of transforming growth factor (TGF)-β2, which ultimately promotes MMP-2 activity. MMP-2 activity is directly correlated to the cell migration rate, with a significantly greater total distance migrated when compared to tumor cells undergoing MMP-2 inhibition. Conversely, small interfering RNA directed against lactate dehydrogenase type A (siLDHA) downregulates LDHA, lowering lactate levels, altering the extracellular environment, and ultimately reducing the ability of cancer cells to migrate.8 Cell locomotion is initiated by glycosaminoglycan hyaluronan and its interactions with its receptor for hyaluronan-mediated motility (RHAMM). This results in a tyrosine phosphorylation pathway, resulting in focal adhesion turnover and allowing for cell motility.62 High levels of RHAMM mRNA in breast cancer have been shown to be associated with worse patient prognosis.122 Using the chemotherapeutic drug tirapazamine to selectively kill hypoxic cells, Cao et al. demonstrated no evidence of delayed angiogenesis, further supporting the notion that early angiogenesis occurs independently of a hypoxic environment.18 Other studies have similarly shown evidence of normoxia in early angiogenesis owing to the presence of VEGF.66 In the aforementioned study by Li, implanted cells with fibroblastlike morphology were seen early on, prior to the dilatation of host

Nestin

CHAPTER 70  Vasculogenesis

FIG 70-3  Top left, Before anti-VEGF therapy, cancer cells are pleomorphic with mitochondria. Top right, After treatment, cells become spindle shaped with few mitochondria. Bottom left, There is a sharp interface between glioblastoma and normal brain tissue. Bottom right, Following anti-VEGF treatment, cells become glycolytic and therefore motile and invasive. The arrow shows the “tumor wave” invading normal brain. (From Keunen O et al: Anti-VEGF treatment reduces blood supply and increases tumor cell invasion in glioblastoma. Proc Natl Acad Sci U S A 108:3749–3754, 2011.)

A

B

C

FIG 70-4  A, Host veins dilate and develop thin walls. B, After continued development, the veins proliferate into “mother” veins. C, The mother veins then subdivide into other veins. (From Pettersson A et al: Heterogeneity of the angiogenic response induced in different normal adult tissues by vascular permeability factor/ vascular endothelial growth factor. Lab Invest 80:99–115, 2000.)

2645

2646

PART IV  Leading Edge Imaging Concepts

Day 1

Day 2

Day 6

Day 8

FIG 70-5  Transparent window shows the growth of tumor cells and vasculogenesis. Li stated that both of these processes (growth and angiogenesis) were visible at the 20- to 50-cell normoxic stage of tumor size. The explanation for angiogenesis before hypoxia can be understood by noting the cell motility on days 2 to 6. Such cell motility is due to hyaluronan, which is induced by lactate. This confirms that lactate is present prior to the development of hypoxia. (From Li CY et al: Initial stages of tumor cell-induced angiogenesis: Evaluation via skin window chambers in rodent models. J Natl Cancer Inst 92:143–147, 2000.)

venules or neoangiogenesis. This was likely due to what is now referred to as epithelial mesenchymal transition (EMT) in which epithelial cells gain properties of mesenchymal and fibroblastic cells, reducing cellular adhesion and allowing for cell locomotion.107 These changes actually precede those of neovascularity.112 Although EMT may be seen in normal development (e.g., neural crest cell formation), it also plays a key role in oncogenesis. The cell-cell adhesion protein E-cadherin permits cellular adhesion, thus inhibiting metastatic spread, and acts as a tumor inhibitor. In fact, reexpression of this protein in tumors has been shown to decrease their aggressiveness. TGF-β stimulates phosphatidylinositol 3-kinase (PI3K) phosphorylation of AKT, which downregulates E-cadherin expression. This allows EMT changes of poor cellular adhesion and subsequent cellular movement. EMT has also been shown to be decreased in cells lacking JNK1.3 Clinically, upregulation of AKT2 in ovarian carcinoma has been linked to more aggressive disease.34 #3: Bioenergetic chemical reactions of oxidative phosphorylation and glycolysis indicate different vessel requirements for each. From only one molecule of glucose and oxygen, the process of oxidative phosphorylation can produce 38 ATP and 2 CO2. This is a much more efficient means of energy production when compared to glycolysis, which would require nearly 20 times the number of glucose

molecules for an equal amount of ATP production. Given the large amount of lactic acid also produced by glycolysis, it would seem on the surface that cancer would naturally prefer oxidative phosphorylation to glycolysis. A historical point of contention regarding cancer metabolism is the question of why cancer would prefer the less energyefficient glycolysis. However, as noted by Warburg, cancer uses glycolysis even in the presence of ample oxygen; this has since been referred to as the Warburg effect.183 This of course contradicts the edicts of the traditional vasculogenesis concept, which held that neoangiogenesis is required in tumors to allow for and maintain oxygenation, as previously described.55 The question of why cancer prefers glycolysis may have a few answers. A simple reason may simply be that glycolysis can produce energy very rapidly, roughly 100-fold faster compared to oxidative phosphorylation.91 Another potential reason is that too much glucose would be devoted solely to ATP production if oxidative phosphorylation were dominant, at the expense of other macromolecules from being developed. In addition to ATP, proliferating cells have multiple requirements including acetyl-CoA, amino acids, and ribose.183 A key difference between the two metabolic mechanisms is that oxidative phosphorylation requires oxygen, necessitating arterial inflow. Oxygen diffuses quite poorly, so a distance of greater than

CHAPTER 70  Vasculogenesis 1 0.9

Metabolite level

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0

5

10 15 20 25 Distance from basement membrane

30

FIG 70-6  Diagram demonstrates glucose and oxygen concentrations compared to intercellular distance. The horizontal arrow denotes oxygen, and the vertical arrow denotes glucose. Glucose availability remains high regardless of the distance from vessels. (Redrawn from Gillies RJ, Gatenby RA: Adaptive landscapes and emergent phenotypes: Why do cancers have high glycolysis? J Bioenerg Biomembr 39:251–257, 2007.)

approximately 100 micrometers from any arterial source will result in cellular hypoxia. In contradistinction, glycolysis functions without the use of oxygen and therefore an arterial inflow is much less important than venous and lymphatic outflow (Fig. 70-6). Especially as a consequence of the significant buildup of lactic acid, lymphangiogenesis is a vital part of glycolysis given its ability for waste removal. Another more complex and intriguing answer as to why glycolysis is preferred is that perhaps ATP is not the primary goal of metabolism as is often thought; rather, the key outcome may be production of the support waste product lactic acid. Lactic acid has multiple effects in oncogenesis. As already mentioned, local changes in the microenvironment pH are a necessary component for TGF-β expression. Other possible roles of lactate include hyaluronidase and hyaluronan expression, as well as an integral role in hypoxia inducible factor (HIF)-1α expression (which will be discussed in more detail later).53 However, the amount of lactate is tightly regulated by cancers because its overabundance may end up in tumor dormancy. Therefore early employment of an elegant outflow system for waste removal is a necessity for cancer development. The most important factor in lactate removal is lymphatics. Carbonic anhydrase IX and XII can also help regulate the extracellular pH by conversion of acidic CO2 into bicarbonate.171,172 MCT1/MCT2 are used by cancers for lactate absorption from the extracellular matrix. LDH1 is an enzyme used to convert lactate to pyruvate and has been noted to be upregulated in colon carcinomas.95 Pyruvate can then subsequently be used to fuel cancer cells or contribute to fibroblast proliferation. #4: Although cancer uses both aerobic metabolism and glycolysis under various circumstances, glycolysis is the “preferred” pathway in normoxia or hypoxia because glycolysis and its byproduct lactate induce and support many procancer processes. As previously mentioned, cancer preferentially uses a glycolytic method of metabolism, largely because of its production of lactate, which has been shown to have multiple subsequent effects in procancer processes. Even in the presence of sufficient oxygen, glycolysis is the preferred pathway and accounts for approximately 50% of all ATP

2647

production.140 Of course in the absence of arteries, glycolysis is responsible for 100% of the energy production. This reliance on glycolysis confers a tremendous survival benefit to cancer cells, which tend to experience inconsistent levels of oxygenation. Intratumoral hypoxia is multifactorial. It is in part related to inadequate oxygen supply by arteries, related to a disorganized network of vasculature. Tortuous and dilated blood vessels in malignant tumors also decrease the efficiency of oxygen delivery and may result in stasis. As tumors grow and the distance from a vascular supply increases, cancer cells may experience hypoxia due to limitations related to the distance oxygen can freely diffuse between tissues.185 With the use of vascular mapping of tumor oxygenation, it has been demonstrated that the amount of oxygenation within the tumor can be quite heterogeneous as well as highly variable within a specific region over a given time period.19 Cancer cells may experience increased interstitial fluid pressure. This may be a consequence of the higher permeability of the tumor vessels, or it may be related to osmotic effects from the buildup of lactate from glycolysis.154,161 Prior beliefs held that vessel permeability occurred primarily via capillaries, but research now shows that veins play a large part in permeability, with evidence of leakage of molecules as small as 17 nm29,93 (Fig. 70-7). This leakage is most significant in the tumor periphery. Whatever the underlying cause, the resulting pressure changes have been shown to correlate with poor efficacy of chemotherapeutic regimens. Interstitial changes may also play a role in lymphangiogenesis and alterations in fibroblasts, including fibroblast motility. The use of glycolysis—and therefore independence from oxygen— permits cancer cells to survive and even metastasize in hypoxic environments. Glycolysis is therefore not merely a survival mechanism but a key contributor to oncogenic processes including cell motility and invasion.54 Lactic acid production from glycolysis renders the extracellular microenvironment of tumor cells acidic.51 This is advantageous for tumor cells because it results in apoptosis of normal nonmalignant cells via the p53 pathway.138,193 This low-pH environment also promotes release of interleukin (IL)-8 and VEGF, which result in angiogenesis.45,160 The extracellular matrix may also be degraded by enzyme release from fibroblasts and macrophages owing to the acidic environment.56,152 Lastly, acidosis can inhibit the host’s immune response.106 Lactate itself has been shown to adversely affect the host’s natural killer cells, macrophages, and T-lymphocytes.39 Lactate and ketones result in the migration of MDA-MB-231 cells, which act as chemoattractants.53 Additionally, lactate and ketones may cause mutations resulting in cancer “stemness,” which corresponds clinically to poor patient survival because this trait is likely related to both tumor growth and metastasis117 (Fig. 70-8). As briefly alluded to already, lactate has a potential role in the expression of hyaluronan. In this way, lactate may have an indirect effect on metastasis. The hyaluronan receptor RHAMM mediates the motility of fibroblasts. When exposed to hypoxia or tumor growth factor, RHAMM becomes highly expressed. RHAMM is related to the invasive functions of cluster of differentiation (CD)44 by increasing its gene expression122 (Fig. 70-9). CD44 can result in proliferation of tumor cells as well as promote cellular motility by acting on actin networks.180 RHAMM is also related to BRCA1 and affects centrosome and mitotic spindle alterations. Lactate also plays an important role in the nuclear factor (NF)κB pathway. Lactate prompts IL-1 stimulation, which results in activation of NFκB.127,195 The NFκB pathway contributes to multiple procancer processes. Diffuse large B-cell lymphoma results in overexpression of the NFκB pathway. Suppressing this pathway indirectly with an IκB inhibitor results in apoptosis of lymphoma cells, demonstrating  the indispensable nature of NFκB in oncogenic processes.167 This

2648

PART IV  Leading Edge Imaging Concepts

L P

P

R

P P b

a P

L

L

P

L P

A

c

P

d Mother vessels

Normal control

a

b

c

GMP

B

d

e

VM, Capillaries

f

FIG 70-7  A, There is dilatation with thinning of the vein wall (arrow), facilitating diffusion. B, Normal controls (a-c) do not show vascular leakage. Diffusion (d-f) through leaky tumor veins. Nagy et al. noted that all tracers leaked through veins at the tumor-host interface. (From Nagy JA et al: Permeability properties of tumor surrogate blood vessels induced by VEGF-A. Lab Invest 86:767–780, 2006.)

CHAPTER 70  Vasculogenesis

GOLPH3 PRKRIR PSMC1 GYG1 STK39 KIAA1524 YAP1 PSMA7 NDUFC1 BET1 PTPN9 ROBO1 PIAS2 SLC1A4 DNAJC13 SUCLG2 SAP30 SUZ12 ANP32A DDR1 ZNF213 DR1 PRKD1 RYK EIF4G1 MCCC1 CCNB1 MAP3K7 SNRPC UBA2 ZZZ3 SLC30A5 CDC42SE1 AGL IBTK FIGNL1 BUB1 GNPAT PRPF38A EDEM3 PHTF2 YTHDF3 PRKD3 NIT2 TOMM70A CENPK PPP1R2 CPNE3 PLOD3 C8orf33 DAP CDC5L DNAJB6 DNAJC10 CCT3 UTP18 RAD17 RRBP1 QARS PTPRJ USP24 ELAVL1 ACTB NCAPH ORMDL1 GALNT1

Ctl 1 Ctl 2 Ctl 3 Lac 1 Lac 2 Lac 3

Ctl 1 Ctl 2 Ctl 3 Lac 1 Lac 2 Lac 3

Ctl 1 Ctl 2 Ctl 3 Lac 1 Lac 2 Lac 3

Lactate / Stem Cells

TMEM30A OXCT1 LTA4H DDX41 TM9SF2 DLL1 MFN1 LSG1 RFC5 LAP3 SLC4A7 GGCX CEL INTS7 MTMR10 APAF1 E2F5 BACH1 PTPN2 DGAT1 CPT1A ZNF35 MCM3 SRR PPA2 ST13 RGS7 PCSK7 WDR77 CASK LAS1L INCENP COL11A1 ANGPTL4 GUSB MYO10 MID2 EFTUD2 RNASEH2B NEDD1 HTRA2 TNKS2 ITFG1 CDH2 SASH1 TYSND1 PLA2G6 SLC15A2 TM7SF3 RBM8A PLD1 SBF2 WEE1 BRD7 TMEM14C TRIM27 FAM50A DDHD1 SEC24C ABCA1 EPAS1 RNASEH2C NR1H4 NOC4L FHL1 HSDL2

DBT UBE2Z VAC14 VPS53 ENTPD5 CKMT2 NUDT21 EGFR MEIS1 RDH11 SNAP23 TIMM44 ABCF2 EIF2AK3 PAK1IP1 BLM BAG2 SYTL4 PNKD USP38 INPP5D CASP6 ADAM28 NAB2 NPC1 ADAMDEC1 PELI1 DAP3 LIG1 CDCA8 ROD1 E2F8 CDKN2C INPPL1 TSPAN12 NSUN4 HTATIP2 SERPINH1 ABHD5 MDFIC SMO C14orf147 RNF26 C3orf58 A2M FBXW2 SFRS16 EEF1D D1DO1 WDR46 CSTF2 TPMT ATP6V1H ORC1L PTPRA PAFAH1B3 SFXN1 SEC23A CRKL IMPDH2 SIN3B INHBB TIMM8A THOC4 ACIN1

FIG 70-8  Mutation of cancer cells affords a significant advantage for cancer because it can adapt with the environment or change its genome. By mutating, cells become more aggressive, enhance survival pathways, or develop resistance to treatments. This image shows part of a genetic “heatmap,” showing hundreds of genetic changes produced by incubating in lactate. Cancer therefore becomes like a stem cell. (From Martinez-Outschoorn UE et al: Ketones and lactate increase cancer cell “stemness,” driving recurrence, metastasis and poor clinical outcome in breast cancer: Achieving personalized medicine via MetaboloGenomics. Cell Cycle 10:1271–1286, 2011.)

2649

2650

PART IV  Leading Edge Imaging Concepts RHAMM Hyaluronan CD44

PO4

Plasma membrane

ERK 1/2

ERK

PO4

Transcriptional machinery

ERK 1/2

Transcriptional regulation of motogenic genes

Nucleus

FIG 70-9  An advantage provided to cancer is cell motility, which aids in invasion. Lactic acid starts the processes that initiates cell movement. Lactate also induces the hyaluronan molecule, which combines with CD44. CD44 then acts in concert with RHAMM, which activates expression of genes that are required for cancer motility. (Redrawn from Maxwell CA et al: Cell-surface and mitotic-spindle RHAMM: Moonlighting or dual oncogenic functions? J Cell Sci 121:925–932, 2008.)

A

B FIG 70-10  A and B, A rabbit model was implanted with VX2 and grown for 28 days. The tumor size and CT arterial flow measured Q7 days (A = day 7, B = day 21). The tumor continues to grow despite decreasing arterial flow. (Fig. 70-11 shows that flow in the enhancing rim has no correlation with tumor size.) (From Wu H et al: Dynamic evolutionary changes in blood flow measured by MDCT in a hepatic VX2 tumor implant over an extended 28-day growth period: Time-density curve analysis. Acad Radiol 16:1483–1492, 2009.)

importance been shown not just in lymphoma but in most solid tumors.84 More specifically, NFκB has been shown to be integral in inhibiting apoptosis, which can help contribute to cell survival.11 Cell proliferation is also dependent on NFκB through interactions with cyclin D1, which is a necessary component of DNA synthesis.81 Tumor invasion and metastasis are also closely linked to NFκB and are inhibited with NFκB blockade.71 #5: Considering the environmental factors of hypoxia and excessive lactate that might cause quiescence/dormancy, data indicate that excess lactate is more likely than hypoxia to cause dormancy. Like tumor development and progression, tumor quiescence (also referred to as dormancy) can be multifactorial, including genetic and

cellular causes as well as environmental causes, such as hypoxia and elevated lactic acid. This dormancy is temporally variable but may last up to decades. By definition, tumor dormancy means that although asymptomatic, residual cancer cells remain present. This is a very common occurrence in cancer, occurring in 20% to 45% of breast and prostate cancers.2 Based on the assumptions of the traditional vasculogenesis concept, it was originally believed that tumor dormancy occurred as a result of tumor outgrowing its arterial supply. However, as has already been mentioned, tumors are less dependent on a robust arterial supply because they are predominantly fueled via the glycolytic pathway, which works independently of oxygen197 (Figs. 70-10 and 70-11). The

CHAPTER 70  Vasculogenesis 100 90

y = –1.2536x + 53.301

ALP (ml/100ml/min)

80 70 60 50 40 30 20 10 0

0

A

5 10 15 20 25 30 Tumor maximum diameter (mm)

35

100 90

y = –16.072x + 47.886

ALP (ml/100ml/min)

80 70 60 50 40 30 20 10 0

B

0

0.5

1

1.5

2

Tumor rim area (cm2)

FIG 70-11  Pearson’s R correlation of unpublished data from Wu et al. For tumor diameter versus arterial perfusion, r = −0.461 P = .01. For tumor rim nonnecrotic area, r = 0.373, P = .42. These data reflect the fact that small cancers use both aerobic and glycolytic metabolism, but they outgrow their arterial supply and switch to glycolysis completely. (Data from Wu H et al: Dynamic evolutionary changes in blood flow measured by MDCT in a hepatic VX2 tumor implant over an extended 28-day growth period: Time-density curve analysis. Acad Radiol 16: 1483–1492, 2009.)

ALPHA hypothesis posits that lymphatics, not arterial inflow and normoxia, are more important. Lymphatic outflow of acidic lactate decreases the concentration to a moderate level, which is the key factor in tumors emerging from quiescence. One piece of evidence that contradicts the notion that tumor dormancy is secondary to hypoxia is the response of hypoxic tumors to hyperbaric oxygen. If dormant tumors were in a quiescent state because of cellular hypoxia, the return to normoxia as a result of hyperbaric oxygen should correspond to tumor progression and vasculogenesis. However, this has not been shown to be the case.159 Unlike in tumors, hyperbaric oxygen has been shown to have effects on increased rates of neoangiogenesis in wound healing. Curiously the converse is true for tumors, which tend to undergo neovascularization under hypoxic conditions.25,72,92 This may be related to hypoxia stimulating IL-8, which promotes angiogenesis.37 When tumors are quiescent, they are resistant to radiation therapy. Previous attempts have been made to expose dormant tumors to hyperbaric oxygen under the assumption that tumor activity would

2651

resume after hypoxia resolved, and the tumor would therefore be radiosensitive. However, because hypoxia is not the driving factor for tumor dormancy, hyperbaric oxygen has no effect on tumor radiosensitivity.37,159 On the other hand, tumors affiliated with high levels of lactate show radioresistance, whereas those with more modest levels of lactic acid do not.145,158 Therefore although lactate has already been shown to be a key factor in tumor growth, invasion, and metastasis, it is only protumorigenic to a certain point. At very high concentrations, lactic acid pushes the tumor into dormancy through interference of the normal cell cycle.155 It is not until lymphangiogenesis occurs that the concentration of lactate is regulated to a more manageable level. Based on the evidence, the authors postulate that a “bioenergetic quiescence/dormancy” is produced by increased levels of lactate, which causes: 1. End-product inhibition of glycolysis, which decreases energy production 2. Reduced availability of building substrates for cell growth and mitosis (i.e., carbohydrates, lipids, nucleotides, etc.) 3. Impaired growth of “mother cells” that precede mitosis 4. Prevention of apoptosis due to NFκB, which is modulated by lactate 5. Impaired response to mitogens because cells lapse into the G0 dormant state Lactic acid has been shown to be a potent and versatile inhibitor of the enzyme phosphofructokinase, which is the rate-limiting step in glycolysis.111 By inhibiting glycolysis, the negative feedback mechanism perpetrated by lactate accomplishes what varying oxygen levels cannot: a method of preventing energy production in tumors. In addition to glycolysis, lactate also has an inhibitory effect on glutaminolysis. With high lactate levels, ammonia and alanine concentrations have been proven to dramatically decrease.104 Other cell building blocks such as nucleotide precursors, a necessary component of DNA and RNA, are diminished as lactate levels rise.23 In vivo tumor cell cultures taken during the plateau phase of cancer growth have been shown to respond poorly to the addition of nutritional supplements such as glucose and glutamine, as well as to alterations in pH, indicating another etiology for tumor dormancy. Ultimately the effect of tumor inhibition was blamed on elevated lactate levels within the cellular media.119 Cell growth is a key preliminary step prior to mitosis.192 For a mother cell to be able to divide in mitosis, it must first create enough building blocks before dividing into two daughter cells. It is theorized that glycolysis is the pathway by which the macromolecular constituents such as nucleotides, amino acids, and lipids are formed in cancer; therefore if glycolysis is blocked, mitosis cannot continue and cell division will not occur, ultimately resulting in tumor dormancy.183 Lactic acid results in stimulation of the NFκB pathway.157,195 NFκB can help promote cancer cell survival by inducing Bcl-XL, which prevents apoptosis of premalignant cells.83 NFκB also acts on oncogenes called inhibitors of apoptosis, whose key function is self-explanatory.15 Many viral-induced cancers employ the NFκB pathway for tumor growth. Conversely if NFκB is suppressed with inhibition of IκB kinase β (IKKβ), apoptosis in cancer cells resumes.167 TGF-β has been shown to be a potent inhibitor of growth arrest at the G1 step of the cell cycle.120 This corresponds to findings that TGF-β levels are elevated in quiescent tumors, contributing to antimitogenic effects, including inhibition of cyclin/cyclin-dependent kinase, which control the cell cycle.43,48 #6: Acidic lactate releases and induces vascular growth factors and HIF-1α through pathways that function in normoxia and/or hypoxia. Although a well-known and seemingly omnipresent compound, lactate has garnered very little attention in the research community. However, a careful analysis of available data suggests that the ubiquity

2652

PART IV  Leading Edge Imaging Concepts

of lactic acid in regard to cancer extends just beyond a mere curiosity; it is in fact a very important component in a number of oncogenic processes. When compared to oxygen, lactate is present and available in a relatively steady supply because glycolysis is being performed in both hypoxia and normoxia (see earlier discussion). Lactate, either directly or indirectly, has multiple influences on vascular growth factors by: 1. Releasing dormantly stored growth factors 2. Directly inducing factors from different cells 3. Attracting cells as active participants 4. Combining with other factors to stimulate vasculogenesis 5. Starting/participating in multistep signaling pathways such as HIF1α and NFκB As noted earlier, tumor dormancy is interrupted by a so-called angiogenic burst, mostly as a result of two factors, FGF-β (also known as basic [b]FGF and FGF-2) and VEGF-A. Although it is a key component of early tumor growth, VEGF has been shown to be unnecessary for continued growth later on in cancer development. It is theorized that the role of VEGF is assumed by other factors such as FGF-β and TGF-α.201 VEGF and FGF are stored in the heparin sulfate matrix and are not released until stimulated by the release of lactic acid and subsequent decrease in pH.24 This alteration in the extracellular microenvironment allows even cells with weak angiogenic capabilities to cease quiescence and develop into cancers.75 As the initial normoxic angio-

genic phase gives way to the hypoxic phase, HIF-1α is activated with additional support of proangiogenic processes. HIF-1α results in the activation of nearly three dozen genes, many of which are directly related to angiogenesis, tumor cell survival, and metastasis.24,185 SDF-1 and its receptor CXCR4 are regulated by HIF-1α and play a role in the chemotaxis of malignant cells. Cells expressing CXCR4 receptors are also attracted to the cancer site as a result of actions of SDF-1.105,124 Of the cells that are attracted by SDF-1, some are involved in vasculogenesis and include angioblastic mesenchymal stem cells. Not only can blood vessels be developed, but their ultimate fate of developing into either artery or vein can be determined as well. The Notch signaling pathway results in stem cells forming arteries.114 Conversely the PI3k pathway results in vein formation.68 The pathway for angiogenesis is complex and interconnected. Lactate contributes to expression of SDF-1, which attracts many cells, including macrophages and monocytes. These in turn also produce angiogenic factors such as TGF-β, IL-1, IL-8, VEGF-A, VEGF-C, and VEGF-D.73,82,85,118,181 The latter two VEGFs contribute to lymphatic formation through lymphangiogenesis. These complexities grow more daunting with multistep processes such as the HIF-1α and NFκB pathways. Potentially because of its significance in oncogenesis, the NFκB pathway can be activated by lactate in numerous ways, including IL-1,73 toll-like receptor (TLR)4 signaling activation, and monocarboxylate transporter (MCT) to name a few157,186,195 (Fig. 70-12).

Glycolysis

Lactate

H

p ↓p

1

mm

Feed forward/back (HIF → glyco → Lactate)

Hypoxia

normoxia hypoxia

Microenvironment transformation release dormant FGF VEGF

2 Induction vascular growth factors

HIF

4 3

ROS

1*

In situ cells cancer endothelial macrophages fibroblast

Marrow stem progenitor angioblasts

SDF

Peripheral cells macrophages endothelial fibroblasts cancer

Vascular growth factors

5 STEM CD34

VEGF VEGF C (macrophages) FGF, other factors, IL8, COX, PGE2

Vasculogenesis (Lymphaties, Veins, Arteries)

FIG 70-12  This diagram demonstrates four ways lactate increases vascular factors. (1) Acidic low pH from lactate activates metalloproteases, which release FGF and VEGF. (2) Lactate stimulates factor induction from cells. (3) Lactate increases HIF by blocking HIF degradation. (4) HIF from lactate makes reactive oxygen species (ROS). Hypoxia only uses steps 3 and 4 to induce VEGF.

CHAPTER 70  Vasculogenesis #7: Arteriogenesis is not the first morphologic change of vasculogenesis. The sequence of vasculogenic changes is first lymphatics, then veins, and finally arteries. Before reliable biomarkers were known, investigators had to distinguish arteries from veins on the basis of blood flow direction and pressure of vessel fluid content. With the use of biomarkers, identification of vessels has become much more accurate. The distinction of arteries from veins has shown that contrary to the traditional belief, veins form before arteries in oncogenesis. Even before this, lymphatics form. After injection of VEGF, lymphatics were demonstrated to grow prior to veins.31 Previously the identification of lymphatics was challenging because of their transparency. With the use of LYVE-1, a marker specific for lymphatics that is not present in vasculature, lymphatics are now much more easily accounted for.35 Lymphatics and blood vessels play different roles in cancer. Whereas blood vessels are typically responsible for delivery of blood and oxygen to and from the tumor, the role of lymphatics is more based in homeostasis.33 Their role in maintaining a balanced level of lactate in which cancer can grow necessitates their early development (Fig. 70-13). Multiple lab analyses have shown the effects of FGF-β and VEGF on tumor dormancy, with the angiogenic spike being triggered early on in oncogenesis.75 Again, these effects are more important for early growth, with less influence on the growth of more robust established tumors. Prior to vasculogenesis, FGF-2 and VEGF-A stimulate the subtype VEGF-C,21,31,130 which subsequently results in development and enlargement of lymphatics35 (Figs. 70-14 and 70-15). This correlates with the previously discussed theory that lymphatics are a key regulatory component of oncogenesis and help maintain a modest lactic acid concentration that promotes tumor growth and proliferation while not being so high that it results in tumor dormancy. Other studies have clearly shown that vascular endothelial cells and lymphatic endothelial cells have distinctly different responses in tumor growth, further proving the two are separately promoted.33 After lymphatic development, cancer results in the promotion of veins before arteries (Fig. 70-16). Similar to lymphatics, this is likely a result of the importance of waste outflow compared to arterial inflow, insofar as tumors can function well anaerobically through glycolysis in the absence of arterial vasculature. Early on, arteries and veins arise 5 4

Podopl+

LEC

% ECs

* 3

*

2

*

*

1 0

-LM

HPV16 1mo

from the same group of angioblastic progenitor cells.68 The enzyme PI3K promotes differentiation of these multipotent angioblasts to vein formation. COUP-TFII is also an essential regulator in the determination of a venous fate.114 Ephrin-B4 has been used in laboratory studies as a specific marker for venous structures. Conversely ERK (p42/44 MAP kinase) is a key factor in arterial specification.68 Additionally, zebrafish studies have shown that VEGF has a role in the Notch pathway, which helps promote arterial differentiation from stem cells.173 In research on mouse ears, Nagy et al. demonstrated that the first blood vessels to form are veins.131 It is not until after veins have formed, proliferated, and dilated that arteries begin to form31 (Fig. 70-17). Similar findings after microdissections of thousands of rodent tumors showed an abundance of veins, with a marked paucity of arteries, again demonstrating early venous development being favored over arteriogenesis139 (Fig. 70-18). #8: ALPHA vasculogenesis serves the needs of glycolysis and therefore is complementary, supplemental, and synergistic with the traditional vasculogenesis concept based on oxygen/hypoxia. As we have shown, cancer can utilize both aerobic and glycolytic metabolism, but for aerobic metabolism to occur, oxygen supply must be adequate. This necessitates a role of angiogenesis and arterial development. Conversely glycolysis occurs independently of an oxygen supply. The critical factor to favor continued glycolysis is therefore not the inflow of oxygen or other nutrients, but the need for a sufficient outflow mechanism to remove the glycolytic waste products. This comes in two forms: lymphangiogenesis and vein formation. When an oxygen supply is readily available, aerobic metabolism and glycolysis may work in concert. However, a reduction in arterial supply—be it from tumor growth or physician intervention, including chemotherapy or embolization—will result in cancer shifting to glycolysis. If cancer were solely reliant on aerobic metabolism, arterial interruption would be a relatively elegant cure for cancer. Some studies of transarterial chemoembolization (TACE) have shown early partial response of hepatocellular carcinoma (HCC). However, follow-up at 2 and 4 years failed to show improvement in patient survival rates,16 likely because of the ability of tumors to convert to glycolysis. Similarly the chemotherapeutic drug bevacizumab, which works through VEGF inhibition, has been shown to be ineffective in the treatment of breast cancer, which is likely attributable to the use of glycolysis by cancer.182

THE ALPHA HYPOTHESIS AND IMAGING

CD31H1 Podopl– VEC CD31Lo

2653

HPV16 4mo

HPV16 6mo

HPV16 tumor

FIG 70-13  This figure shows flow cytometry of endothelial cells. There is proliferation of endothelial cells during development of squamous cell carcinoma of the skin, from the premalignant phase to 6 months. The blood endothelial cells VEC (blue) proliferated faster than lymphatic endothelial cells LEC (red). When in situ cancer developed, the LEC developed faster than VEC cells. (Redrawn from Eichten A et al: Distinctive features of angiogenesis and lymphangiogenesis determine their functionality during de novo tumor development. Cancer Res 67:5211– 5220, 2007.)

In the early 1970s, Sir Godfrey Hounsfield pioneered the development of computed tomography (CT). This technology, which made use of ionizing radiation to obtain images of a patient in cross section, earned him the Nobel Prize in 1979.10,69 With CT technology, and later with magnetic resonance imaging (MRI), the development of intravenous contrast agents soon followed, allowing for evaluation of blood vessels and subsequent perfusion of solid organs. These technologic developments coincided with the growing popularity of the traditional vasculogenesis theory as reported by Gimbrone.55 It was expected that perfusion imaging would correlate with increased arterial characteristics on the radiologic imaging of tumors. However, this has not proven to be true. Instead it is the venous outflow properties that relate more to tumor development, reflecting the importance of venule development as detailed earlier. One of the most frequently encountered and perhaps most quintessential examples of this are vascular hepatic lesions. On early arterial-phase imaging, benign lesions such as focal nodular hyperplasia and hemangiomas demonstrate enhancement not unlike HCC. However, on more delayed imaging in the venous phase, HCC demonstrates a very characteristic Text continued on p. 2658

2654

PART IV  Leading Edge Imaging Concepts

P P

A

B

C P

Lymphatics

CD34 PECAM

P

Blood vessels

D

E PECAM

G

F TIE2

H

VEGFR-3

I

FIG 70-14  A-I, Lymphatics (red) develop extending from the pellet to the sclera in response to FGF-2, with blood vessels just beginning to form. (From Chang LK et al: Dose-dependent response of FGF-2 for lymphangiogenesis. Proc Natl Acad Sci U S A 101:11658–11663, 2004.)

CHAPTER 70  Vasculogenesis Control

PIGF 3d

A

B

7d

84 d

C

D

VEGF-A 3d

7d

E

F

84 d

270 d

G

H

VEGF-A, 84d 1 min

I

2 min

3 min

4 min

J

K

L

Control 0 min

M

VEGF-A, 35 d

20 min

0 min

150 min

N

O

P

FIG 70-15  Normal lymphatic development versus development of giant lymphatics that formed 1 to 3 days after VEGF exposure. (From Nagy JA et al: Vascular permeability factor/vascular endothelial growth factor induces lymphangiogenesis as well as angiogenesis. J Exp Med 196:1497–1506, 2002.)

2655

2656

PART IV  Leading Edge Imaging Concepts

d1

d3

d15

d22

FIG 70-16  Anatomic appearance of mouse ears after VEGF over the course of 22 days. Note the large draining vein (arrow). The first vessels that form are from the host venules; the veins then subsequently dilate and divide. (From Nagy JA et al: Permeability properties of tumor surrogate blood vessels induced by VEGF-A. Lab Invest 86:767–780, 2006.)

Precursor venule/capillary

Bridging 3–10 days

Daughter capillaries

Pericyte Basement membrane degradation

Endothelial cell

Basement membrane

Mother 1–5 days Apoptosis

Smooth muscle cell

Glomeruloid bodies 1–4 weeks Vascular malformation 1 week – >1 year

FIG 70-17  Diagram showing the development of host venules into mother veins, capillaries, and arteries. (Redrawn from Dvorak HF: How tumors make bad blood vessels and stroma. Am J Pathol 162:1747–1757, 2003.)

CHAPTER 70  Vasculogenesis

*

* *

*

*

*

FIG 70-18  Cross-sectional images confirm the first blood vessels that form are venous (venous loops seen at arrows), with very few arteries. (From Patan S et al: Vascular morphogenesis and remodeling in a human tumor xenograft: Blood vessel formation and growth after ovariectomy and tumor implantation. Circ Res 89:732–739, 2001.)

A

B FIG 70-19  A, Hepatic lesion demonstrating contrast enhancement in the early arterial phase. This may represent either a benign or malignant lesion. B, Delayed postcontrast image demonstrates persistent vascularity within the lesion, typical of a hemangioma.

2657

2658

PART IV  Leading Edge Imaging Concepts

A

B FIG 70-20  A, There is increased enhancement of a liver hepatocellular carcinoma in the early arterial phase of enhancement. B, More delayed imaging demonstrates washout of contrast (arrow), typical of HCC.

A

B FIG 70-21  A, Early contrast enhancement is seen within liver lesions. B, Delayed-phase imaging demonstrates hypovascularity in a patient with cervical cancer metastases to the liver.

washout of contrast in comparison to benign lesions (Figs. 70-19 to 70-21). A second characteristic appearance of cancer is a hypovascular tumor on arterial-phase images with an enhancing rim, likely attributable to the tumor outgrowing its oxygen supply.163 The malignant washout characteristics in hepatic lesions have been shown to correlate to α-fetoprotein levels, including both objective and subjective washout evaluation.110 In addition to the washout pattern of tumors, other imaging characteristics have been developed that are related to the vascular supply and can reliably be performed to differentiate malignancy from benignity. These include alterations in blood volume, vascular permeability (referred to as Ktrans on MRI), and outflow venous curve analysis (either subjective or objective). These imaging parameters have been shown to accurately represent characteristics of malignant tumors; their heavy reliance on vascular outflow and washout are supportive evidence for the ALPHA hypothesis and its components as outlined at the beginning of this chapter (Figs. 70-22 and 70-23).

Imaging evaluation of tumors is therefore highly dependent on vascular (predominantly venous) characteristics. In the absence of intravenous contrast, even otherwise sophisticated MRI techniques are inadequate to detect and fully evaluate cancer.17 Contrast-enhancement characteristics correlate with histopathologic findings,137 whereas unenhanced studies are significantly limited in their abilities to determine the location, extent of, and vascular characteristics of malignant diseases.88 Dynamic contrast-enhanced (DCE) techniques have not just been limited to the evaluation of breast and liver lesions but have been shown to be incredibly useful in the characterization of many malignancies, including pancreatic, prostate, and even hematologic malignancies.22,47,90

Time-Density Curves Time-density curves provide the basis for most perfusion techniques. With DCE imaging, intravenous contrast is injected into the patient, and a series of images of the region of interest are obtained over time.

CHAPTER 70  Vasculogenesis

2659

Ktrans map (Gd-DTPA)

Anatomy

0.4 SCC tumor

0.3

0.2

0.1

A

B 0.0 Ktrans map (OX63)

pO2 map (mmHg)

0.12

40

0.09

30

0.06

20

0.03

10

C

D 0.00

E

0

F

FIG 70-22  Comparison of Ktrans maps of Gd-DTPA and OX63 in a squamous cell carcinoma (SCC) tumor. A, SCC tumor region can be distinguished in a T2-weighted anatomic image. B, Ktrans map of Gd-DTPA using 7T MRI in SCC tumor-bearing mouse. C, Ktrans map of OX63 using the same tumor. D, A pO2 map from the same MRI images for Ktrans OX63 map. E and F, Based on the anatomic image. Tumor region with low Ktrans OX63 value was oxygenated, and the region with high Ktrans values coincided with hypoxia in the pO2 image. (From Matsumoto S et al: Simultaneous imaging of tumor oxygenation and microvascular permeability using Overhauser enhanced MRI. Proc Natl Acad Sci U S A 106:17898–17903, 2009.)

In doing this the vascularity of the area can be discerned as the amount of enhancement increases during the arterial phase, with subsequent changes in the venous phase that vary based on tissue characteristics.123 On MRI this is most reliably performed on T1-weighted images using gadolinium contrast, owing to the sensitivity of T1-weighted sequences for low doses of contrast and the ability of gadolinium to easily cross capillaries. DCE imaging may also be performed by CT or radionuclide studies. From these dynamic images, the contrast within the lesion can be evaluated visually or may be evaluated as a computer-generated curve, a graphic representation of tumor vascularity.

In kinetic curves, there is an initial upslope that represents the arterial inflow phase. In the next phase, the contrast may steadily rise, plateau, or drop precipitously in the venous outflow phase. A downsloping in delayed imaging reflects venous proliferation. Kinetic curves have shown that tumors tend to enhance more strongly in early arterial-phase images, but there is significant overlap between benign malignant lesions, and therefore arterial characteristics cannot be reliably used to categorize a lesion as benign or malignant.99 This supports the ALPHA hypothesis in that venous attributes are more significant in oncogenesis. As previously noted, cancer will preferentially develop

2660

PART IV  Leading Edge Imaging Concepts

1.0 0.8 0.6 0.4 0.2 Baseline

0.0 0

100

200

300

Gd Concentration (mM)

Gd Concentration (mM)

(253,51) Ktrans= 0.24

1.0 0.8 0.6 0.4 0.2

500

400

C1

0.0 0

100

ROI Arrival (LMB)/Peak (RMB)

Arrival (LMB)/Peak (RMB)

1.0 0.8 0.6 0.4 0.2 C4

0.0 300 400 200 Time (sec) ROI Current time Arrival (LMB)/Peak (RMB) Alfn 0

B

ROI

Current time Alfn

Gd Concentration (mM)

Gd Concentration (mM)

A

100

200

300

400

500

Time (sec)

Time (sec)

Current time Alfn

1.0 0.8 0.6 0.4 0.2 C7

0.0 500

0

100

200 300 400 Time (sec) ROI Current time Arrival (LMB)/Peak (RMB) Alfn

FIG 70-23  A and B, Tumor encasement in the involved breast can be seen in the following colors: red and green indicate high enhancement, blue indicates low enhancement. Gadolinium concentration-time curves (dark lines on graphs) show the rate of gadolinium-based contrast material perfusion through the tumor. The blue lines represent arterial input. (From Thukral A et al: Inflammatory breast cancer: Dynamic contrastenhanced MR in patients receiving bevacizumab–Initial experience. Radiology 244:727–735, 2007.)

500

2661

CHAPTER 70  Vasculogenesis venous structures before arterial ones in order to maintain appropriate levels of lactate for continued growth. It is therefore appropriate that time-density curves with a prominent downslope in the venous phase (referred to as a type 3 curve) are highly suggestive of a malignant process. This is in contradistinction to a continuously increasing curve (type 1 curve), which is typically seen in benign processes. A curve that plateaus in the venous phase of imaging (called a type 2 curve) is generally considered an indeterminate finding (Figs. 70-24 and 70-25).

Early postcontrast phase

Intermediate and late postcontrast phase

(SIc - SI) / SI

Ia

Ib

II III

FIG 70-24  Differences in the intensity curves derived from gadoliniumenhanced DCE. Curves that increase or plateau (Ia, Ib, and II) are most likely to be benign. Curves showing decrease (III) are more likely to be malignant. Note that curve II must be correlated closely with anatomic appearances on mammography. (Redrawn from Kuhl CK et al: Dynamic breast MR imaging: Are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 211:101–110, 1999.)

Enhancement curves have been shown to reliably differentiate between malignant and benign lesions. For example, in enhancing benign breast lesions and fibrocystic changes, the predominant pattern was the type 1 curve, which was seen in 83% of cases. The prevailing pattern in malignant lesions was a type 3 curve; type 2 was the next most common pattern seen in cancers. Less than 10% of malignancies demonstrate a type 1 enhancement curve.99 The types 1, 2, and 3 enhancement curves corresponded to a 6%, 64%, and 87% chance of malignancy, respectively. In breast cancer, type 2 kinetic curves can be seen in both malignant and benign diseases but are slightly more common in malignancies by approximately a 3 : 2 margin (Fig. 70-26). Evaluation of kinetic curves shows a significantly high rate of concordance among different readers, with one study demonstrating no discrepancies from type 1 to type 3. Additionally, in MRI of breast cancer, a common pitfall resulting in a false-negative diagnosis is prominent background enhancement of fibroglandular tissue. Therefore the differing DCE characteristics of cancer may help increase the sensitivity of MRI in the diagnosis of breast cancer.97,176 Type 1 and type 3 curves have been shown to reliably differentiate benign from malignant adnexal tumors.178 A solid adnexal mass identified on MRI can be characterized as either probably benign or probably malignant solely based on its DCE curve characteristics. In fact, type 3 curves were equaled only by the presence of peritoneal implants in terms of the ability of imaging features to characterize a lesion as benign or malignant. Other studies have shown similar results, with DCE MRI offering up to 100% accuracy at categorizing a complex cystic and solid ovarian mass as benign or malignant.27 When evaluating the washout characteristics of kinetic curves, it is important to remember that although the predominant cause tends to be venous proliferation, the formation of arteriovenous shunts may also lead to rapid venous outflow. These shunts may result in up to 30% of arterial blood flow passing directly into the venous system and bypassing the microcirculation.184 Although differing on the microscopic level, the kinetic curve of an arteriovenous shunt will result in 300

250 V1 Rapid

% Change

200

150

100 80 60

0

A

B

6 0

7[2] 90

8[3] 150

9[4] 210

10[5] 271

FIG 70-25  A, Red arrow indicates the region of interest. B, Corresponding time–signal intensity curve of the region of interest shows initial enhancement with early washout. The mass was a BI-RADS category 6 related to the patient’s known invasive ductal carcinoma.

11[6] 331

2662

PART IV  Leading Edge Imaging Concepts 250

V1 Rapid

200

% Change

150

100 80 60

0

A

B

4 0

5[2] 90

6[3] 150

7[4] 210

8[5] 269

9[6] 329

FIG 70-26  A, Red arrow indicates the region of interest. B, Corresponding time–signal intensity curve of the region of interest. The mass demonstrates fast initial enhancement followed by persistent enhancement. The mass was determined to be BI-RADS category 2. This was subsequently excited and found to be an inarticulate papilloma.

a similar washout characteristic to a lesion with a prominent venous outflow. Therefore this characteristic of tumor vascularity may not only be pro-oncogenic, given the necessity for adequate washout of glycolytic byproducts, but may also be reflected in the imaging characteristics of contrast washout and type 3 kinetic curves. Kinetic curve analysis is not just useful in the diagnosis of cancer itself but in tumor follow-up. In breast cancer imaging, a change in time-density curve parameters, including loss of the typical washout pattern, signifies response to tumor treatment. In fact the enhancement kinetic characteristics are apparent even prior to the changes in tumor morphology.98 DCE studies can also be used adjunctively for breast-conserving surgical planning to reduce the need for mastectomy without an increased risk of subsequent recurrence. DCE MRI outperforms estrogen receptor (ER) and progesterone receptor (PR) status and the presence of metastatic lymph nodes in predicting recurrent disease.49 Using DCE ultrasound, vascular changes have been depicted within 1 week of antiangiogenic treatment initiation.108

Blood Volume, Veins, and Venous Biomarkers When an enhancement curve is created, the area under the curve (AUC) is equal to the blood volume within the region of interest. This number denotes the entire intravascular blood pool, including the sum of both the arteries and veins. However, owing to early venous recruitment with regard to arteries, it may be anticipated that venous volume would outweigh arterial volume. Indeed the arterial volume typically makes up only approximately 29% of the vascular blood volume,28,133 which is a percentage that has been replicated in multiple studies (Fig. 70-27). Therefore the major contributor to oncogenic blood volume is veins. The majority of research that correlates blood volume to tumor development and aggressiveness has been performed in the field of neuroradiology, although other subspecialties have noted similar findings. In breast cancer, extensive vascular proliferation may predict

AUC

FIG 70-27  This diagram shows the contrast time curve for the aorta and tissue. The area under the curve (AUC) represents the opacified blood in the arterial and venous spaces. This corresponds to the total blood volume. In normal tissues, venous blood accounts for about 70% of the AUC. Higher-grade tumors have greater blood volumes.

metastatic spread at both local and distant metastatic sites.191 Tumor vascularity expectedly corresponds to the amount of blood volume, which has been proven using both histopathologic and angiographic means of evaluation. The imaging assessment of tumor vascularity has been used to discern between low-grade tumors, which demonstrate lower vascularity, and the more highly vascular high-grade malignancies.20 These evaluations include measurements not only of the cerebral blood volume (CBV) itself but also cerebral blood flow (CBF)78,165 (Figs. 70-28 and 70-29). In looking at both high-grade and low-grade gliomas, a higher CBV correlates with more rapid clinical progression

CHAPTER 70  Vasculogenesis

Maximal rCBV ratio

8

6

4

2

0

Low-grade

High-grade

FIG 70-28  Scatterplot of relative cerebral blood volume ratios for each tumor shows significant differences between low-grade and high-grade oligodendroglial tumors (P < .05).

of disease.109 The peripheral enhancement pattern that may be seen in tumors corresponds to a high density of neovascularity at the margins of the tumor, with a central hypovascular region related to either necrosis or desmoplastic changes.17 A common diagnostic dilemma in the radiologic evaluation of cancer is being able to distinguish residual or recurrent malignancy from posttreatment changes, such as with radiation therapy. While other imaging techniques may have significant difficulty in determining whether a focus represents radiation changes or recurrent malignancy, blood volume analyses have shown diagnostic benefit in this arena. Whereas it has already been established that malignancies result in higher blood volumes due to venogenesis, radiation therapy results in diminished venous density.46 Therefore malignancies are expected to correspond to higher CBV, with radiation necrosis producing a decreased CBV. These findings have been verified in the scientific literature.6 Using relative (r)CBV, Hu et al. were able to very reliably categorize a focus as either tumor growth or radiation change. With a cutoff rCBV of 0.71 or less, there was a 96% accuracy in diagnosing postradiation changes, whereas rCBV levels greater than 0.71 corresponded to tumor growth.70 Vascular development can be assessed in vitro by the expression of a certain protein family called ephrins. Ephrin-B2, which is seen on arterial endothelial cells, results in arterial proliferation.64 As noted before, ephrin-B4 conversely works to stimulate venous development. It has been shown that tumors expressing ephrin-B4 have a larger blood volume.134 Therefore the overall blood volume of a tumor, which can be detected on radiologic imaging studies, can be used as an indirect measurement of venous proliferation and therefore tumor development. In contradistinction, although the traditional vasculogenesis theory would posit that overexpression of the ephrin-B2 protein would result in significant tumor growth, this has not been shown to be the case. In fact overexpression of ephrin-B2 has been shown to decrease tumor growth in vivo115 despite the arterial proliferation. This is further supporting evidence that venous development is the key vasculature required for cancer growth.

MRI Permeability Measurements by Ktrans and Kep Newer MR technologies are able to evaluate vessel permeability, indicated by the exchange of fluids between the intracellular and extracellular spaces. More specifically, permeability is measured using the parameters Kep (defined as the rate constant for the backflux from the

2663

extracellular extravascular space to the vasculature) and Ktrans (defined as the rate of flux of contrast agent into the extracellular extravascular space within a given volume).123,196 These parameters are most commonly calculated through the use of DCE MRI. Permeability in tumors was initially believed to be predominantly arterial, but newer data suggest that vessel leakiness is more pronounced in the veins.29 These findings are supported by a lack of correlation between oxygen levels and permeability, as demonstrated by Matsumoto et al.121 This means increased permeability must be associated with veins rather than arteries. Permeability is in part mediated by VEGF-A; in fact, prior to being known as vascular endothelial growth factor, VEGF was known as vascular permeability factor. Additional indirect supporting evidence of the venous vasculature being the key component of permeability is that VEGF is more prominent on veins.30 Using the MRI parameters Ktrans and Kep as markers of the leakiness of tumor-related vasculature, the measurement of tumor permeability was able to differentiate high-grade from low-grade tumors.78,144 In prostate cancers, Ktrans and Kep levels were significantly higher in regions corresponding to tumor when compared to uninvolved areas within the peripheral zone of the prostate.135 Some studies have shown that permeability characteristics can outperform CBV, blood flow, and mean transit time in tumor categorization.78 In breast cancer studies, Kep and Ktrans were shown to be significantly elevated in grade 3 tumors when compared to lower-grade tumors.146 These permeability parameters can therefore be used to monitor the treatment response of cancer. Specifically, Ktrans and Kep have been shown to strongly correlate to tumor response to anti-VEGF medications, including bevacizumab.179 Ktrans and Kep have also been shown to be significantly different among patients classified as nonresponders, partial responders, and complete responders to anticancer therapies.26 The characteristics of tumor permeability may also be used to guide tumor treatment. Studies have shown that rectal cancers with elevated Ktrans levels—and therefore greater tumor permeability—tend to respond better to chemoradiotherapy.52 These findings suggest that patients with lower permeability may benefit from alternative treatments. Additionally, successful treatment itself resulted in lower Ktrans values on subsequent evaluations, allowing the permeability characteristics of the tumors to be a surrogate for tumor response.

DWI and FDG PET Although not directly related to perfusion characteristics, diffusionweighted imaging (DWI) and fluorine-18 fluorodeoxyglucose (FDG) positron emission tomography (PET) are often used in tandem with perfusion imaging. These techniques provide molecular information about glucose transport and metabolic activity, as well as lactate removal when glycolysis is present. FDG PET imaging reflects ATP production from fluorodeoxyglucose via glycolysis. This has been used to accurately detect a large number of malignant diseases and their spread. If the traditional vasculogenesis theory was correct in its assumption that cancer is predominantly fueled by arterial development, FDG PET imaging would correlate with the amount of vascular proliferation in cancer. However, this is not the case. The lack of correlation between FDG avidity and dynamic kinetic data suggests that PET imaging provides information of cancer development independent of arterial factors.32 Other studies have shown that hypoxia and angiogenesis are not directly related to each other141 (Fig. 70-30). DWI provides information about the movement of water molecules in tissue. It has been shown that excessive buildup of lactate secondary to glycolytic breakdown of glucose will result in increased concentrations of lactate in the extracellular space. Although the normal interstitial fluid pressure should be essentially zero,77 the increased oncotic

2664

PART IV  Leading Edge Imaging Concepts

A

B

C

D FIG 70-29  A, Cerebral blood volume (CBV) map of a low-grade oligoastrocytoma. Note the low tumoral vascularity. B, T2 of the low-grade oligoastrocytoma. C, Blood volume map of a high-grade anaplastic oligodendroglioma. Relative CBV map shows elevated vascularization of the tumor. D, Corresponding T2 of the high-grade anaplastic oligodendroglioma.

pressure from extracellular lactate subsequently causes a gradient, resulting in diffusion of water molecules across the cell membrane.154 This finding has already been shown to possibly have the consequence of decreased arterial inflow,184 and it can also result in a gradient that restricts the subsequent diffusion of water molecules. Therefore in addition to other causes of diffusion restriction in tumors, such as hypercellularity (as seen in tumors such as medulloblastoma), the increased interstitial fluid pressure secondary to lactic acid can

contribute to the imaging appearance of malignancies on MRI.79,147 When antiangiogenic agents are administered, the interstitial fluid pressure may decrease toward normal levels.72 Other studies have shown that as tumor size increases, there is increased interstitial fluid pressure, supporting the evidence of multiple roles of lactate in malignant growth.12 The parameters of interstitial fluid pressure and tumor size both inversely correlate with partial pressure of oxygen, again refuting the traditional concept of

CHAPTER 70  Vasculogenesis

125I-Gluco-RGD

tumor-to-muscle ratio

18 16 14 12 10 8 6 4 2 0 0

125I-Gluco-RGD

tumor-to-muscle ratio

A

2

4

6

8

18F-FAZA tumor-to-muscle

10

12

ratio

12 10 8 6 4 2 0 0

B

1

2

3

18F-FAZA tumor-to-muscle

4

5

ratio

FIG 70-30  Top, Pixel distribution of 18F-FAZA and

125 I-Gluco-RGD T/M ratios of the entire EMT6 tumor population under room air breathing conditions. Bottom, No obvious relationship exists between blood vessels and glucose. (Redrawn from Picchio M et al: Intratumoral spatial distribution of hypoxia and angiogenesis assessed by 18F-FAZA and 125I-Gluco-RGD autoradiography. J Nucl Med 49:597–605, 2008.)

vasculogenesis being dependent on arterial inflow. The high interstitial fluid pressure may also limit the transport of chemotherapeutic agents to the site of malignancy.13 New developments using whole-body DWI have been made regarding cancer detection. When compared to CT, whole-body MRI with DWI has a similar ability for the initial detection of lymphoma. MRI evaluation has the added benefit of having superior accuracy in its staging of lymphoma.101 Whole-body MRI with DWI also demonstrates similar accuracy for the detection of bone metastases in non– small cell lung cancer when compared to bone scintigraphy and FDG PET/CT.174 This technique has also been used to identify occult primary malignancies in patients presenting with metastases.57

ALPHA AND IMPLICATIONS IN CANCER TREATMENT Cancer is a notoriously difficult disease entity to treat, and despite extensive research efforts, interventional breakthroughs, and pharmaceutical developments, it remains a leading cause of death. These immense efforts have not been in vain, with overall cancer mortality decreasing in recent years.80 However, cancer is still the leading cause

2665

of death in people between the ages of 40 and 74 years and is a significant cause of death in all age groups. Part of the limitations of the success in cancer treatment is due to the nature of the disease and potential downsides to the therapy itself. Many cancer treatments come with a host of side effects, including significant immunosuppression, which can be quite harmful to patients. In addition, many patients are unable to tolerate the adverse effects of various cancer treatments and may therefore end up being undertreated. In addition to these reasons, however, is some misunderstanding in the nature of the disease and the physiology behind its development. We therefore propose some potential new treatments in cancer therapy that are supported by the ALPHA hypothesis on oncogenesis and its basis in lymphangiogenesis and venous proliferation. The traditional vasculogenesis theory held that cancers were highly dependent on arterial inflow. Were this the case, antiarterial cancer treatments would show immense benefits in overall survival of patients. This unfortunately has not proven to be the case. Although it is true that certain antiarterial treatments may show certain benefits, these are often short lived. For example, in HCC the tumor is solely supplied via the hepatic artery. On the basis of the traditional vasculogenesis theory, many treatments were performed targeting the hepatic artery. Hepatic artery ligation has been attempted, with a goal of interrupting tumor growth; however, these trials did not demonstrate increased patient survival time, and this therapy has been essentially entirely abandoned.60 Similar treatments have been attempted using transarterial bland embolization of the hepatic artery in HCC, using the same rationale used for hepatic artery ligation. This therapy did demonstrate diminished tumor growth, but at 24-month follow-up there was no survival benefit in patients treated with transarterial embolization when compared to those who did not have the procedure, after accounting for Childs-Pugh class.16 Chemotherapeutic treatments directed toward VEGF have also been proposed. Again, on the basis of the traditional vasculogenesis concept, the thought that using antiangiogenic treatments directed toward arterial development would ultimately result in a normoxic tumor environment would suggest that these treatments would be ineffective. This has indeed been repeatedly shown to be the case.41 Bevacizumab (trade name, Avastin) is a monoclonal antibody directed toward VEGF-A.74,179 Given its antiarterial properties, bevacizumab has been shown to result in reduced tumor blood flow and volume.98 However, the radiologic signs of bevacizumab treatment do not necessarily correspond to improved clinical outcomes. In fact, lab analyses have shown that as a consequence of the diminished arterial supply to tumors treated with bevacizumab, the subsequent hypoxia creates an environment in which the tumor must revert to a nearly entirely glycolytic-based form of metabolism. Instead of hampering tumor growth, antiarterial therapies may therefore promote an environment in which the tumor will become even more invasive than it initially was.98 An additional unintended consequence of antiarterial therapies is that they may compound difficulties in the delivery mechanisms of other treatments, including radiation therapy and systemic chemotherapies. In 2011 the U.S. Food and Drug Administration (FDA) revoked its approval for the use of bevacizumab in the treatment of breast cancer, because it was “clear that women who take Avastin for metastatic breast cancer risk potentially life-threatening side effects without proof that the use of Avastin will provide a benefit… that would justify those risks.”182 Although unfortunately antiarterial therapies have not proven to be successful in cancer treatment, data from the literature regarding these treatments can be useful. If antiarterial therapies have been largely unsuccessful, other targets can be focused on to prevent tumor growth and spread and hopefully lead to a cure for the disease. Many

2666

PART IV  Leading Edge Imaging Concepts

new and promising anticancer therapies exist that complement the ALPHA hypothesis and its description of tumor development. The importance of lactate in oncogenesis has already been stressed. It is therefore not surprising that the amount of tumor response to fractionated irradiation therapy correlates to the lactate concentration,158 suggesting that elevated lactate contributes to tumor radioresistance. Tumor lactate concentrations can be evaluated noninvasively by MR spectroscopy.145 Lactate levels ultimately correlate with overall patient survival; high-lactate tumors tend to have a worse prognosis compared to those with lower concentrations of lactate.158 Lactic acid, as we have shown, is a ubiquitous product of tumor glycolysis. It has been theorized that a potential avenue of augmenting the effect of radiation therapy would be to manipulate tumor glycolysis itself to diminish lactate levels and thus promote an environment more responsive to fractionated irradiation. It has already been shown that lactic acid provides a low pH to the extracellular microenvironment. This acidity is a key component in many pro-oncogenic processes, including tumor growth factor activation and vasculogenesis. Radiation resistance related to high lactate concentrations may therefore be related to the decrease in extracellular pH.175 This is further supportive evidence that optimal lactate concentrations may enhance the effects of therapeutic radiation. Lactate’s effects on the microenvironment pH may be extended to the intracellular environment. It is theorized that whereas a low extracellular pH is protumoral, intracellular acidity can result in cancer cell death. Therefore a mechanism that would diminish extracellular lactate and simultaneously increase intracellular lactate may be able to have antioncogenic effects at multiple levels. A lactate transporter known as monocarboxylate transporter 1 (MCT1) has been identified and used as a target for potential further therapies.164 α-Cyano-4hydroxycinnamic acid (CHCA) is a phenylpropanoid derivative of cinnamic acid9 that has been found to be a competitive inhibitor of lactate transport61 at MCT1 and therefore extracellular lactate.164 In vitro analyses using mouse models have shown that treatment with CHCA has resulted in diminished tumor growth and cancer cell apoptosis.100 Additionally, CHCA therapy acted synergistically with radiation therapy, which is again attributed to alterations in lactate transportation across the cell membrane. Because of its prominence among many tumor subtypes, CHCA therapy therefore has promising applications in the treatments of multiple forms of cancer. These include primary central nervous system tumors, head and neck tumors, breast cancer, lung cancer, colon cancer, and neuroblastoma.36,44,95,142,164 An additional benefit of CHCA is its benign side-effect profile and high levels of patient tolerance.143 SDF-1 and its receptor CXCR have already been described as a key component of chemotaxis in cancer development. CXCR4 receptor overexpression is directly linked to malignancies varying from breast and prostate cancer to pancreatic adenocarcinoma and glioblastoma multiforme.181 Inhibition of CXCR4 may therefore be an effective treatment for cancer, particularly in treating and preventing metastatic spread. The CXCR4 inhibitor AMD3100 has been shown to be useful in combination with other cytotoxic chemotherapeutic agents in the treatment of glioblastoma multiforme.149 Similar studies in acute myeloid leukemia have shown that CXCR4 inhibition by AMD3465 complements sorafenib treatment, making leukemic cells more sensitive to chemotherapy.202 Other research has shown that CXCR4 inhibitors were effective in limiting breast cancer metastasis by inhibiting migration.113 The NFκB pathway has already been described as an important pathway in mediating procancer processes, including inhibition of cancer cell apoptosis. Modulating this pathway therefore is another potential target for cancer treatment. IκB proteins regulate and inhibit

the NFκB pathway; IκB kinases of various subtypes can act by degrading IκB proteins, disinhibiting NFκB and allowing cancer cells to proliferate.7 Inhibiting these kinases would therefore be expected to allow continued IκB action and continued inhibition of the NFκB pathway. Inhibition of IKKβ has been shown to directly produce antitumorigenic effects on cancer cells to induce cell death.38,167 Additionally, IKKβ inhibition may work indirectly, resulting in prevention of antiapoptotic actions in cancer cells. IKKβ inhibition could have antioncogenic effects in multiple malignancies, including lymphoma and multiple myeloma. This pathway can be blocked via inhibition of HSP90, therefore blocking induction of the IKKβ pathway.5,189 HSP90 targeted therapy has shown efficacy in killing Hodgkin’s lymphoma cells and diffuse large B-cell lymphoma cells.14 Bortezomib is a proteasome inhibitor that may also inactivate  the NFκB pathway in cancer. This has been used to treat diffuse large B-cell lymphoma when combined with standard chemotherapy. Bortezomib has also been shown to slow the progression of multiple myeloma.150 This drug works more effectively in patients who have low levels of TRAF3 expression.87 Multiple other types of cancer have been shown to be affected by bortezomib, including head and neck, prostate, gastrointestinal, and ovarian cancers.1,4,116,132,170,194,203 Bort­ ezomib works by ultimately inhibiting cancer cell proliferation and inducing apoptosis.15 The protein kinase CK2 also plays a role in activation of the NFκB pathway, including IKKβ activation. This can be inhibited by the plant product apigenin, which subsequently inhibits NFκB activity. Apigenin has been shown to be effective against breast cancer67,103,199 and prostate cancer.58,162 Apigenin has also been shown to be a potential promising new treatment for the notoriously difficult-to-treat anaplastic thyroid cancer.198 Inhibition of CK2 results in decreased cancer cell proliferation and arrest of the cell cycle, as well as inhibiting cancer cell migration.96 Based on the ALPHA hypothesis that malignant tumors are highly dependent on lymphatics, which develop even prior to the development of veins and arteries, antilymphatic-directed therapies are very promising for potentially treating cancer. To briefly summarize, lymphatic vessels grow in response to actions by VEGF-C and VEGF-D.129 VEGF-C is the ligand for the VEGFR-3 receptor.177 It has been theorized that VEGF-C and VEGFR-3 could therefore be targeted to prevent lymphangiogenesis and result in suppression of tumor development. Controlling lymphangiogenesis could also play a prominent role in preventing metastases, because lymphatic development plays a key role in metastatic spread.166 Additional studies will be required to determine the exact role these treatments would play. Additionally the analysis of potential antilymphatic properties of current chemotherapies have been largely ignored or otherwise underrepresented in the clinical data.129 Still, some medications have been studied that play antilymphatic roles. Sunitinib and sorafenib play antioncogenic roles through inhibition of VEGFR. Sunitinib has improved outcomes regarding patient survival in cases of pancreatic neuroendocrine tumors.148 It has also been shown to improve survival in patients with advanced gastrointestinal stromal tumor who have failed imatinib. Sunitinib also demonstrates better outcomes than interferon-alfa in treating metastatic renal cell carcinoma.126 Since lymphatics develop early in cancer, even prior to vasculogenesis, antilymphangiogenic treatments may show even greater benefit to early stages of cancer.129,169 Sodium-potassium-chloride cotransporters (NKCC1) result in efflux of KCl and help regulate cellular volume. NKCC1 transporters have been shown to have an important role in the migration of tumor cells, although the exact mechanism by which this occurs is not fully understood.200 Based on this, attempts have been made to inhibit NKCC1 transport and therefore limit cancer proliferation. The loop

CHAPTER 70  Vasculogenesis diuretic bumetanide is typically used to treat hypertension and congestive heart failure; however, because of its mechanism of action as a potent NKCC1 inhibitor, it may also have function in treating malignancy. Not surprisingly given this information, treatment with bumetanide results in decreased glioma cell invasion, making it a promising new potential treatment for these malignancies.59 The number of invasive cells remaining is inversely correlated to the dose of bumetanide applied.50 Bumetanide may reduce the amount of cell invasion by over half.187 Ultimately the role of future treatments in cancer is promising. Although a large amount of effort is still required by the scientific research community, gigantic strides have been made in recent years to help combat this ubiquitous and complex disease. It is our hope that the information provided in this chapter and the work of the innumerable researchers who directly and indirectly contributed to it with their extensive efforts will help play some role in the continuing battle to understand cancer pathogenesis more fully and to improve future treatments.

REFERENCES 1. Adachi M, et al: Synergistic effect of histone deacetylase inhibitors FK228 and m-carboxycinnamic acid bis-hydroxamide with proteasome inhibitors PSI and PS-341 against gastrointestinal adenocarcinoma cells. Clin Cancer Res 10(11):3853–3862, 2004. 2. Aguirre-Ghiso JA: Models, mechanisms and clinical evidence for cancer dormancy. Nat Rev Cancer 7(11):834–846, 2007. 3. Alcorn JF, et al: Jun N-terminal kinase 1 regulates epithelial-tomesenchymal transition induced by TGF-β1. J Cell Sci 121(7):1036– 1045, 2008. 4. Allen C, et al: Bortezomib-induced apoptosis with limited clinical response is accompanied by inhibition of canonical but not alternative nuclear factor-κB subunits in head and neck cancer. Clin Cancer Res 14(13):4175–4185, 2008. 5. Baldwin AS: Control of oncogenesis and cancer therapy resistance by the transcription factor NF-κB. J Clin Invest 107(3):241–246, 2001. 6. Barajas RF, et al: Distinguishing recurrent intra-axial metastatic tumor from radiation necrosis following gamma knife radiosurgery using dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging. AJNR Am J Neuroradiol 30(2):367–372, 2009. 7. Basak S, et al: A fourth IκB protein within the NF-κB signaling module. Cell 128(2):369–381, 2007. 8. Baumann F, et al: Lactate promotes glioma migration by TGF-β2– dependent regulation of matrix metalloproteinase-2. Neuro Oncol 11(4):368–380, 2009. 9. Beavis RC, Chaudhary T, Chait BT: α-Cyano-4-hydroxycinnamic acid as a matrix for matrix assisted laser desorption mass spectrometry. Organic Mass Spectrometry 27(2):156–158, 1992. 10. Beckmann EC: CT scanning the early days. Br J Radiol 79(937):5–8, 2006. 11. Beg AA, Baltimore D: An essential role for NF-κB in preventing TNF-α-induced cell death. Science 274(5288):782–784, 1996. 12. Boucher Y, Lee I, Jain RK: Lack of general correlation between interstitial fluid pressure and oxygen partial pressure in solid tumors. Microvasc Res 50(2):175–182, 1995. 13. Boucher Y, Jain RK: Microvascular pressure is the principal driving force for interstitial hypertension in solid tumors: Implications for vascular collapse. Cancer Res 52(18):5110–5114, 1992. 14. Broemer M, Krappmann D, Scheidereit C: Requirement of Hsp90 activity for IκB kinase (IKK) biosynthesis and for constitutive and inducible IKK and NF-κB activation. Oncogene 23(31):5378–5386, 2004. 15. Brown M, et al “NF-κB in carcinoma therapy and prevention.” (2008): 1109-1122. 16. Bruix J, et al: Transarterial embolization versus symptomatic treatment in patients with advanced hepatocellular carcinoma: Results of a

17.

18.

19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31. 32.

33.

34. 35.

36.

37. 38.

2667

randomized, controlled trial in a single institution. Hepatology 27(6):1578–1583, 1998. Buadu LD, et al: Breast lesions: Correlation of contrast medium enhancement patterns on MR images with histopathologic findings and tumor angiogenesis. Radiology 200(3):639–649, 1996. Cao Y, et al: Observation of incipient tumor angiogenesis that is independent of hypoxia and hypoxia inducible factor-1 activation. Cancer Res 65(13):5498–5505, 2005. Cárdenas-Navia LI, et al: The pervasive presence of fluctuating oxygenation in tumors. Cancer Res 68(14):5812–5819, 2008. Caseiras GB, et al: Inclusion or exclusion of intratumoral vessels in relative cerebral blood volume characterization in low-grade gliomas: Does it make a difference? AJNR Am J Neuroradiol 29(6):1140–1141, 2008. Chang LK, et al: Dose-dependent response of FGF-2 for lymphangiogenesis. Proc Natl Acad Sci U S A 101(32):11658–11663, 2004. Chen B-B, et al: Dynamic contrast-enhanced MR imaging measurement of vertebral bone marrow perfusion may be indicator of outcome of acute myeloid leukemia patients in remission. Radiology 258(3):821– 831, 2011. Cruz HJ, et al: Effects of ammonia and lactate on growth, metabolism, and productivity of BHK cells. Enzyme Microb Technol 27(1):43–52, 2000. D’Arcangelo D, et al: Acidosis inhibits endothelial cell apoptosis and function and induces basic fibroblast growth factor and vascular endothelial growth factor expression. Circ Res 86(3):312–318, 2000. Davis JC, Buckley CJ, Barr P-O: Compromised soft tissue wounds: Correction of wound hypoxia. In Problem Wounds: The Role of Oxygen, New York, 1988, Elsevier, pp 143–152. de Bazelaire C, et al: Accuracy of perfusion MRI with high spatial but low temporal resolution to assess invasive breast cancer response to neoadjuvant chemotherapy: A retrospective study. BMC Cancer 11(1):361, 2011. Dilks P, et al: Can quantitative dynamic contrast-enhanced MRI independently characterize an ovarian mass? Eur Radiol 20(9):2176– 2183, 2010. Duong TQ, Kim S-G: In vivo MR measurements of regional arterial and venous blood volume fractions in intact rat brain. Magn Reson Med 43(3):393–402, 2000. Dvorak HF, et al: Identification and characterization of the blood vessels of solid tumors that are leaky to circulating macromolecules. Am J Pathol 133(1):95, 1988. Dvorak HF, et al: Distribution of vascular permeability factor (vascular endothelial growth factor) in tumors: Çoncentration in tumor blood vessels. J Exp Med 174(5):1275–1278, 1991. Dvorak HF: How tumors make bad blood vessels and stroma. Am J Pathol 162(6):1747, 2003. Eby PR, et al: Metabolic and vascular features of dynamic contrastenhanced breast magnetic resonance imaging and 15O-water positron emission tomography blood flow in breast cancer. Acad Radiol 15(10):1246–1254, 2008. Eichten A, Hyun WC, Coussens LM: Distinctive features of angiogenesis and lymphangiogenesis determine their functionality during de novo tumor development. Cancer Res 67(11):5211–5220, 2007. Elstrom RL, et al: Akt stimulates aerobic glycolysis in cancer cells. Cancer Res 64(11):3892–3899, 2004. Enholm B, et al: Adenoviral expression of vascular endothelial growth factor-C induces lymphangiogenesis in the skin. Circ Res 88(6):623– 629, 2001. Fang J, et al: The H+-linked monocarboxylate transporter (MCT1/ SLC16A1): A potential therapeutic target for high-risk neuroblastoma. Mol Pharmacol 70(6):2108–2115, 2006. Feldmeier JJ, et al: Hyperbaric oxygen: Does it promote growth or recurrence of malignancy? Undersea Hyperb Med 30(1):1–18, 2003. Feuerhake F, et al: NFκB activity, function, and target-gene signatures in primary mediastinal large B-cell lymphoma and diffuse large B-cell lymphoma subtypes. Blood 106(4):1392–1399, 2005.

2668

PART IV  Leading Edge Imaging Concepts

39. Fischer K, et al: Inhibitory effect of tumor cell–derived lactic acid on human T cells. Blood 109(9):3812–3819, 2007. 40. Folkman J: Tumor angiogenesis: Therapeutic implications. N Engl J Med 285:1182–1186, 1971. 41. Folkman J: Anti-angiogenesis: New concept for therapy of solid tumors. Ann Surg 175(3):409, 1972. 42. Folkman J: Role of angiogenesis in tumor growth and metastasis. Semin Oncol 29(6):2002. WB Saunders. 43. Francis SM, et al: A functional connection between pRB and transforming growth factor β in growth inhibition and mammary gland development. Mol Cell Biol 29(16):4455–4466, 2009. 44. Froberg MK, et al: Expression of monocarboxylate transporter MCT1 in normal and neoplastic human CNS tissues. Neuroreport 12(4):761– 765, 2001. 45. Fukumura D, et al: Hypoxia and acidosis independently up-regulate vascular endothelial growth factor transcription in brain tumors in vivo. Cancer Res 61(16):6020–6024, 2001. 46. Fuss M, et al: Radiation-induced regional cerebral blood volume (rCBV) changes in normal brain and low-grade astrocytomas: Quantification and time and dose-dependent occurrence. Int J Radiat Oncol Biol Phys 48(1):53–58, 2000. 47. Futterer JJ, et al: Prostate cancer localization with dynamic contrastenhanced MR imaging and proton MR spectroscopic imaging 1. Radiology 241(2):449–458, 2006. 48. Galderisi U, Jori FP, Giordano A: Cell cycle regulation and neural differentiation. Oncogene 22(33):5208–5219, 2003. 49. Garimella V, et al: Recurrence rates after DCE-MRI image guided planning for breast-conserving surgery following neoadjuvant chemotherapy for locally advanced breast cancer patients. Eur J Surg Oncol 33(2):157–161, 2007. 50. Garzon-Muvdi T, et al: Regulation of brain tumor dispersal by NKCC1 through a novel role in focal adhesion regulation. PLoS Biol 10(5):e1001320, 2012. 51. Gatenby RA, et al: Acid-mediated tumor invasion: A multidisciplinary study. Cancer Res 66(10):5216–5223, 2006. 52. George ML, et al: Non-invasive methods of assessing angiogenesis and their value in predicting response to treatment in colorectal cancer. Br J Surg 88(12):1628–1636, 2001. 53. Gillies RJ, Robey I, Gatenby RA: Causes and consequences of increased glucose metabolism of cancers. J Nucl Med 49(Suppl 2):24S–42S, 2008. 54. Gillies RJ, Gatenby RA: Adaptive landscapes and emergent phenotypes: Why do cancers have high glycolysis? J Bioenerg Biomembr 39(3):251– 257, 2007. 55. Gimbrone MA, et al: Tumor dormancy in vivo by prevention of neovascularization. J Exp Med 136(2):261–276, 1972. 56. Glunde K, et al: Extracellular acidification alters lysosomal trafficking in human breast cancer cells. Neoplasia 5(6):533–545, 2003. 57. Gu T-F, et al: Diagnostic value of whole body diffusion weighted imaging for screening primary tumors of patients with metastases. Chin Med Sci J 23(3):145–150, 2008. 58. Gupta S, Afaq F, Mukhtar H: Selective growth-inhibitory, cell-cycle deregulatory and apoptotic response of apigenin in normal versus human prostate carcinoma cells. Biochem Biophys Res Commun 287(4):914–920, 2001. 59. Haas BR, Sontheimer H: Inhibition of the sodium-potassium-chloride cotransporter isoform-1 reduces glioma invasion. Cancer Res 70(13):5597–5606, 2010. 60. Hafström L, et al: Hepatic artery occlusion for liver cancer. Ann Chir Gynaecol 72(5):1982. 61. Halestrap AP: Transport of pyruvate and lactate into human erythrocytes. Evidence for the involvement of the chloride carrier and a chloride-independent carrier. Biochem J 156:193–207, 1976. 62. Hall CL, et al: Hyaluronan and the hyaluronan receptor RHAMM promote focal adhesion turnover and transient tyrosine kinase activity. J Cell Biol 126(2):575–588, 1994. 63. Harrell MI, Iritani BM, Ruddell A: Tumor-induced sentinel lymph node lymphangiogenesis and increased lymph flow precede melanoma metastasis. Am J Pathol 170(2):774–786, 2007.

64. Hayashi S-I, et al: Functional ephrin-B2 expression for promotive interaction between arterial and venous vessels in postnatal neovascularization. Circulation 111(17):2210–2218, 2005. 65. Hayashi Y, et al: Regulation and function of aquaporin-1 in glioma cells. Neoplasia 9(9):777–787, 2007. 66. Hendriksen EM, et al: Angiogenesis, hypoxia and VEGF expression during tumour growth in a human xenograft tumour model. Microvasc Res 77(2):96–103, 2009. 67. Hessenauer A, Montenarh M, Gotz C: Inhibition of CK2 activity provokes different responses in hormone-sensitive and hormonerefractory prostate cancer cells. Int J Oncol 22(6):1263–1270, 2003. 68. Hong CC, et al: Artery/vein specification is governed by opposing phosphatidylinositol-3 kinase and MAP kinase/ERK signaling. Curr Biol 16(13):1366–1372, 2006. 69. Hounsfield GN: Computerized transverse axial scanning (tomography): Part 1. Description of system. Br J Radiol 46(552):1016–1022, 1973. 70. Hu LS, et al: Relative cerebral blood volume values to differentiate high-grade glioma recurrence from posttreatment radiation effect: Direct correlation between image-guided tissue histopathology and localized dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging measurements. AJNR Am J Neuroradiol 30(3):552–558, 2009. 71. Huang S, et al: Blockade of NF-kappaB activity in human prostate cancer cells is associated with suppression of angiogenesis, invasion, and metastasis. Oncogene 20(31):4188–4197, 2001. 72. Hunt TK: The physiology of wound healing. Ann Emerg Med 17(12):1265–1273, 1988. 73. Hunt TK, et al: Aerobically derived lactate stimulates revascularization and tissue repair via redox mechanisms. Antioxid Redox Signal 9(8):1115–1124, 2007. 74. Hurwitz H: Integrating the anti–VEGF-A humanized monoclonal antibody bevacizumab with chemotherapy in advanced colorectal cancer. Clin Colorectal Cancer 4:S62–S68, 2004. 75. Indraccolo S, et al: Interruption of tumor dormancy by a transient angiogenic burst within the tumor microenvironment. Proc Natl Acad Sci U S A 103(11):4216–4221, 2006. 76. Jain RK, et al: Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin Pract Oncol 3(1):24–40, 2006. 77. Jain RK, Tong RT, Munn LL: Effect of vascular normalization by antiangiogenic therapy on interstitial hypertension, peritumor edema, and lymphatic metastasis: Insights from a mathematical model. Cancer Res 67(6):2729–2735, 2007. 78. Jain R, et al: Quantitative estimation of permeability surface-area product in astroglial brain tumors using perfusion CT and correlation with histopathologic grade. AJNR Am J Neuroradiol 29(4):694–700, 2008. 79. Jaremko JL, et al: Value and limitations of diffusion-weighted imaging in grading and diagnosis of pediatric posterior fossa tumors. AJNR Am J Neuroradiol 31(9):1613–1616, 2010. 80. Jemal A, et al: Trends in the leading causes of death in the United States, 1970-2002. JAMA 294(10):1255–1259, 2005. 81. Joyce D, et al: NF-κB and cell-cycle regulation: The cyclin connection. Cytokine Growth Factor Rev 12(1):73–90, 2001. 82. Kang KA, Harrison DK, Bruley DF: Oxygen Transport to Tissue XXIX, vol 614, 2008, Springer. 83. Karin M: NF-κB as a critical link between inflammation and cancer. Cold Spring Harb Perspect Biol 1(5):a000141, 2009. 84. Karin M, et al: NF-κB in cancer: From innocent bystander to major culprit. Nat Rev Cancer 2(4):301–310, 2002. 85. Kataru RP, et al: Critical role of CD11b+ macrophages and VEGF in inflammatory lymphangiogenesis, antigen clearance, and inflammation resolution. Blood 113(22):5650–5659, 2009. 86. Kato Y, et al: Acidic extracellular pH induces matrix metalloproteinase-9 expression in mouse metastatic melanoma cells through the phospholipase D-mitogen-activated protein kinase signaling. J Biol Chem 280(12):10938–10944, 2005. 87. Keats JJ, et al: Promiscuous mutations activate the noncanonical NF-κB pathway in multiple myeloma. Cancer Cell 12(2):131–144, 2007.

CHAPTER 70  Vasculogenesis 88. Kerslake RW, et al: Dynamic gradient-echo and fat-suppressed spin-echo contrast-enhanced MRI of the breast. Clin Radiol 50(7):440– 454, 1995. 89. Keunen O, et al: Anti-VEGF treatment reduces blood supply and increases tumor cell invasion in glioblastoma. PNAS 108(9):3749–3754, 2011. 90. Kim JH, et al: Solid pancreatic lesions: Characterization by using timing bolus dynamic contrast-enhanced MR imaging assessment—A preliminary study. Radiology 266(1):185–196, 2013. 91. King M. “The Medical Biochemistry Page.” The Medical Biochemistry Page. N.p., 2013. Web. 16 Aug. 2014. 92. Knighton DR: Mechanisms of wound healing. Hyperbaric Medicine Practice 119–140, 1994. 93. Kohn S, et al: Pathways of macromolecular tracer transport across venules and small veins. Structural basis for the hyperpermeability of tumor blood vessels. Lab Invest 67(5):596–607, 1992. 94. Koong AC, et al: Pancreatic tumors show high levels of hypoxia. Int J Radiat Oncol Biol Phys 48(4):919–922, 2000. 95. Koukourakis MI, et al: Comparison of metabolic pathways between cancer cells and stromal cells in colorectal carcinomas: A metabolic survival role for tumor-associated stroma. Cancer Res 66(2):632–637, 2006. 96. Kramerov AA, et al: Inhibition of protein kinase CK2 suppresses angiogenesis and hematopoietic stem cell recruitment to retinal neovascularization sites. Mol Cell Biochem 316(1–2):177–186, 2008. 97. Kuhl C: The current status of breast MR imaging. Part I. Choice of technique, image interpretation, diagnostic accuracy, and transfer to clinical practice 1. Radiology 244(2):356–378, 2007. 98. Kuhl CK: Current status of breast MR imaging. Part 2. Clinical applications 1. Radiology 244(3):672–691, 2007. 99. Kuhl CK, et al: Dynamic breast MR imaging: Are signal intensity time course data useful for differential diagnosis of enhancing lesions? 1. Radiology 211(1):101–110, 1999. 100. Kumar A, Kant S, Singh SM: α-Cyano-4-hydroxycinnamate induces apoptosis in Dalton’s lymphoma cells: Role of altered cell survival-regulatory mechanisms. Anticancer Drugs 24(2):158–171, 2013. 101. Kwee TC, et al: Whole-body MRI, including diffusion-weighted imaging, for the initial staging of malignant lymphoma: Comparison to computed tomography. Invest Radiol 44(10):683–690, 2009. 102. Lamszus K, Kunkel P, Westphal M: Invasion as limitation to antiangiogenic glioma therapy, Vienna, 2003, Springer. 103. Landesman-Bollag E, et al: Protein kinase CK2: Signaling and tumorigenesis in the mammary gland. Mol Cell Biochem 227(1–2):153– 165, 2001. 104. Lao M-S, Toth D: Effects of ammonium and lactate on growth and metabolism of a recombinant Chinese hamster ovary cell culture. Biotechnol Prog 13(5):688–691, 1997. 105. Lapidot T, Dar A, Kollet O: How do stem cells find their way home? Blood 106(6):1901–1910, 2005. 106. Lardner A: The effects of extracellular pH on immune function.   J Leukoc Biol 69(4):522–530, 2001. 107. Larue L, Bellacosa A: Epithelial-mesenchymal transition in development and cancer: Role of phosphatidylinositol 3′ kinase/AKT pathways. Oncogene 24(50):7443–7454, 2005. 108. Lassau N, et al: Dynamic contrast-enhanced ultrasonography (DCE-US) with quantification of tumor perfusion: A new diagnostic tool to evaluate the early effects of antiangiogenic treatment. Eur Radiol 17(6):89–98, 2007. 109. Law M, et al: Gliomas: Predicting time to progression or survival with cerebral blood volume measurements at dynamic susceptibilityweighted contrast-enhanced perfusion MR imaging 1. Radiology 247(2):490–498, 2008. 110. Lee KHY, et al: Triple-phase MDCT of hepatocellular carcinoma. AJR Am J Roentgenol 182(3):643–649, 2004. 111. Leite TC, et al: Lactate downregulates the glycolytic enzymes hexokinase and phosphofructokinase in diverse tissues from mice. FEBS Lett 585(1):92–98, 2011.

2669

112. Li C-Y, et al: Initial stages of tumor cell-induced angiogenesis: Evaluation via skin window chambers in rodent models. J Natl Cancer Inst 92(2):143–147, 2000. 113. Liang Z, et al: Inhibition of breast cancer metastasis by selective synthetic polypeptide against CXCR4. Cancer Res 64(12):4302–4308, 2004. 114. Lin F-J, Tsai M-J, Tsai SY: Artery and vein formation: A tug of war between different forces. EMBO Rep 8(10):920–924, 2007. 115. Liu W, et al: Effects of overexpression of ephrin-B2 on tumour growth in human colorectal cancer. Br J Cancer 90(8):1620–1626, 2004. 116. Lun M, et al: Nuclear factor-kappaB pathway as a therapeutic target in head and neck squamous cell carcinoma: Pharmaceutical and molecular validation in human cell lines using Velcade and siRNA/ NF-κB. Ann Clin Lab Sci 35(3):251–258, 2005. 117. Martinez-Outschoorn UE, et al: Ketones and lactate increase cancer cell “stemness”, driving recurrence, metastasis and poor clinical outcome in breast cancer: Achieving personalized medicine via MetaboloGenomics. Cell Cycle 10(8):1271, 2011. 118. Maruyama K, et al: Inflammation-induced lymphangiogenesis in the cornea arises from CD11b-positive macrophages. J Clin Invest 115(9):2363–2372, 2005. 119. Marx E, Mueller-Klieser W, Vaupel P: Lactate-induced inhibition of tumor cell proliferation. Int J Radiat Oncol Biol Phys 14(5):947–955, 1988. 120. Massagué J: How cells read TGF-β signals. Nat Rev Mol Cell Biol 1(3):169–178, 2000. 121. Matsumoto S, et al: Simultaneous imaging of tumor oxygenation and microvascular permeability using Overhauser enhanced MRI. PNAS 106(42):17898–17903, 2009. 122. Maxwell CA, McCarthy J, Turley E: Cell-surface and mitotic-spindle RHAMM: Moonlighting or dual oncogenic functions? J Cell Sci 121(7):925–932, 2008. 123. Miller JC, et al: Imaging angiogenesis: Applications and potential for drug development. J Natl Cancer Inst 97(3):172–187, 2005. 124. Milovanova TN, et al: Lactate stimulates vasculogenic stem cells via the thioredoxin system and engages an autocrine activation loop involving hypoxia-inducible factor 1. Mol Cell Biol 28(20):6248–6261, 2008. 125. Morishima T, et al: Lactic acid increases aquaporin 4 expression on the cell membrane of cultured rat astrocytes. Neurosci Res 61(1):18–26, 2008. 126. Motzer RJ, et al: Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. NEJM 356(2):115–124, 2007. 127. Moynagh PN, Williams DC, O’Neill LA: Interleukin-1 activates transcription factor NF kappa B in glial cells. Biochem J 294:343–347, 1993. 128. Murata K, et al: Structural determinants of water permeation through aquaporin-1. Nature 407(6804):599–605, 2000. 129. Nagahashi M, et al: Lymphangiogenesis: A new player in cancer progression. World J Gastroenterol 16(32):4003, 2010. 130. Nagy JA, et al: Vascular permeability factor/vascular endothelial growth factor induces lymphangiogenesis as well as angiogenesis. J Exp Med 196(11):1497–1506, 2002. 131. Nagy JA, et al: Permeability properties of tumor surrogate blood vessels induced by VEGF-A. Lab Invest 86(8):767–780, 2006. 132. Nawrocki ST, et al: Aggresome disruption: A novel strategy to enhance bortezomib-induced apoptosis in pancreatic cancer cells. Cancer Res 66(7):3773–3781, 2006. 133. Neil JJ, Ackerman JJH: Detection of pseudodiffusion in rat brain following blood substitution with perfluorocarbon. J Magn Reson (1969) 97(1):194–201, 1992. 134. Noren NK, et al: Interplay between EphB4 on tumor cells and vascular ephrin-B2 regulates tumor growth. Proc Natl Acad Sci U S A 101(15):5583–5588, 2004. 135. Ocak I, et al: Dynamic contrast-enhanced MRI of prostate cancer at 3 T: A study of pharmacokinetic parameters. AJR Am J Roentgenol 189(4):W192–W201, 2007. 136. Ohlsson B, et al: Dearterialization of colorectal liver cancer: Institutional experience. Dig Surg 16(3):229–235, 1999.

2670

PART IV  Leading Edge Imaging Concepts

137. Orel SG, et al: Suspicious breast lesions: MR imaging with radiologicpathologic correlation. Radiology 190(2):485–493, 1994. 138. Park HJ, et al: Acidic environment causes apoptosis by increasing caspase activity. Br J Cancer 80(12):1892, 1999. 139. Patan S, et al: Vascular morphogenesis and remodeling in a human tumor xenograft blood vessel formation and growth after ovariectomy and tumor implantation. Circ Res 89(8):732–739, 2001. 140. Pedersen PL: Warburg, me and Hexokinase 2: Multiple discoveries of key molecular events underlying one of cancers’ most common phenotypes, the “Warburg Effect”, ie, elevated glycolysis in the presence of oxygen. J Bioenerg Biomembr 39(3):211–222, 2007. 141. Picchio M, et al: Intratumoral spatial distribution of hypoxia and angiogenesis assessed by 18F-FAZA and 125I-Gluco-RGD autoradiography. J Nucl Med 49(4):597–605, 2008. 142. Pinheiro C, et al: Increased expression of monocarboxylate transporters 1, 2, and 4 in colorectal carcinomas. Virchows Arch 452(2):139–146, 2008. 143. Porporato PE, et al: Anticancer targets in the glycolytic metabolism of tumors: A comprehensive review. Front Pharmacol 2:2011. 144. Provenzale JM, et al: Correlation of relative permeability and relative cerebral blood volume in high-grade cerebral neoplasms. AJR Am J Roentgenol 187(4):1036–1042, 2006. 145. Quennet Verena, et al: Tumor lactate content predicts for response to fractionated irradiation of human squamous cell carcinomas in nude mice. Radiother Oncol 81(2):130–135, 2006. 146. Radjenovic A, et al: Measurement of pharmacokinetic parameters in histologically graded invasive breast tumours using dynamic contrastenhanced MRI. Br J Radiol 81(962):2008. 147. Rajeshkannan R, et al: Clinical applications of diffusion weighted MR imaging: A review. Indian J Radiol Imaging 16(4):705, 2006. 148. Raymond E, et al: Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. NEJM 364(6):501–513, 2011. 149. Redjal N, et al: CXCR4 inhibition synergizes with cytotoxic chemotherapy in gliomas. Clin Cancer Res 12(22):6765–6771, 2006. 150. Richardson PG, et al: Bortezomib or high-dose dexamethasone for relapsed multiple myeloma. NEJM 352(24):2487–2498, 2005. 151. Rofstad EK, et al: The tumor bed effect: Increased metastatic dissemination from hypoxia-induced up-regulation of metastasispromoting gene products. Cancer Res 65(6):2387–2396, 2005. 152. Rozhin J, et al: Pericellular pH affects distribution and secretion of cathepsin B in malignant cells. Cancer Res 54(24):6517–6525, 1994. 153. Ruddell A, et al: Dynamic contrast-enhanced magnetic resonance imaging of tumor-induced lymph flow. Neoplasia 10(7):706–713, 2008. 154. Rutz HP: A biophysical basis of enhanced interstitial fluid pressure in tumors. Med Hypotheses 53(6):526–529, 1999. 155. Rutz HP, Little JB: Exogenous lactate interferes with cell-cycle control in Balb3T3 mouse fibroblasts. Int J Radiat Oncol Biol Phys 31(3):525– 528, 1995. 156. Saadoun S, et al: Increased aquaporin 1 water channel expression inhuman brain tumours. Br J Cancer 87(6):621–623, 2002. 157. Samuvel DJ, et al: Lactate boosts TLR4 signaling and NF-κB pathwaymediated gene transcription in macrophages via monocarboxylate transporters and MD-2 up-regulation. J Immunol 182(4):2476–2484, 2009. 158. Sattler UGA, et al: Glycolytic metabolism and tumour response to fractionated irradiation. Radiother Oncol 94(1):102–109, 2010. 159. Schönmeyr BH, et al: The effect of hyperbaric oxygen treatment on squamous cell cancer growth and tumor hypoxia. Ann Plast Surg 60(1):81–88, 2008. 160. Shi Q, et al: Regulation of vascular endothelial growth factor expression by acidosis in human cancer cells. Oncogene 20(28):3751–3756, 2001. 161. Shieh AC, Swartz MA: Regulation of tumor invasion by interstitial fluid flow. Phys Biol 8(1):015012, 2011. 162. Shukla S, Gupta S: Molecular targets for apigenin-induced cell cycle arrest and apoptosis in prostate cancer cell xenograft. Mol Cancer Ther 5(4):843–852, 2006. 163. Sica GT, Ji H, Ros PR: CT and MR imaging of hepatic metastases. AJR Am J Roentgenol 174(3):691–698, 2000.

164. Sonveaux P, et al: Targeting lactate-fueled respiration selectively   kills hypoxic tumor cells in mice. J Clin Invest 118(12):3930–3942, 2008. 165. Spampinato MV, et al: Cerebral blood volume measurements and proton MR spectroscopy in grading of oligodendroglial tumors. AJR Am J Roentgenol 188(1):204–212, 2007. 166. Stacker SA, Achen MG: From anti-angiogenesis to antilymphangiogenesis: Emerging trends in cancer therapy. Lymphat Res Biol 6(3–4):165–172, 2008. 167. Staudt LM: Oncogenic activation of NF-κB. Cold Spring Harb Perspect Biol 2(6):a000109, 2010. 168. Stetler-Stevenson WG: The role of matrix metalloproteinases in tumor invasion, metastasis, and angiogenesis. Surg Oncol Clin N Am 10(2):383–392, 2001. 169. Sundlisaeter E, et al: Increased lymphatic vascular density is seen   before colorectal cancers reach stage II and growth factor FGF-2 is downregulated in tumor tissue compared with normal mucosa. APMIS 117(3):212–221, 2009. 170. Sunwoo JB, et al: Novel proteasome inhibitor PS-341 inhibits activation of nuclear factor-κB, cell survival, tumor growth, and angiogenesis in squamous cell carcinoma. Clin Cancer Res 7(5):1419–1428, 2001. 171. Švastová E, et al: Hypoxia activates the capacity of tumor-associated carbonic anhydrase IX to acidify extracellular pH. FEBS Lett 577(3):439–445, 2004. 172. Swietach P, Vaughan-Jones RD, Harris AL: Regulation of tumor pH and the role of carbonic anhydrase 9. Cancer Metastasis Rev 26(2):299–310, 2007. 173. Swift MR, Weinstein BM: Arterial-venous specification during development. Circ Res 104(5):576–588, 2009. 174. Takenaka D, et al: Detection of bone metastases in non-small cell lung cancer patients: Comparison of whole-body diffusion-weighted imaging (DWI), whole-body MR imaging without and with DWI, whole-body FDG-PET/CT, and bone scintigraphy. J Magn Reson Imaging 30(2):298–308, 2009. 175. Tannock IF, Rotin D: Acid pH in tumors and its potential for therapeutic exploitation. Cancer Res 49(16):4373–4384, 1989. 176. Teifke A, et al: Undetected malignancies of the breast: Dynamic contrast-enhanced MR imaging at 1.0 T 1. Radiology 224(3):881–888, 2002. 177. Thiele W, Sleeman JP: Tumor-induced lymphangiogenesis: A target for cancer therapy? J Biotechnol 124(1):224–241, 2006. 178. Thomassin-Naggara I, et al: Adnexal masses: Development and preliminary validation of an MR imaging scoring system. Radiology 267(2):432–443, 2013. 179. Thukral A, et al: Inflammatory breast cancer: Dynamic contrastenhanced MR in patients receiving bevacizumab—Initial experience 1. Radiology 244(3):727–735, 2007. 180. Toole BP: Hyaluronan-CD44 interactions in cancer: Paradoxes and possibilities. Clin Cancer Res 15(24):7462–7468, 2009. 181. Tseng D, Vasquez-Medrano DA, Brown JM: Targeting SDF-1/CXCR4 to inhibit tumour vasculature for treatment of glioblastomas. Br J Cancer 104(12):1805–1809, 2011. 182. US Food and Drug Administration. (2011). FDA commissioner announces Avastin decision: Drug not shown to be safe and effective in breast cancer patients. Beckmann, Elizabeth C. CT scanning the early days. (2006). 183. Vander Heiden MG, Cantley LC, Thompson CB: Understanding the Warburg effect: The metabolic requirements of cell proliferation. Science 324(5930):1029–1033, 2009. 184. Vaupel P, Kallinowski F, Okunieff P: Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: A review. Cancer Res 49(23):6449–6465, 1989. 185. Vaupel P, Harrison L: Tumor hypoxia: Causative factors, compensatory mechanisms, and cellular response. Oncologist 9(Suppl 5):4–9, 2004. 186. Végran F, et al: Lactate influx through the endothelial cell monocarboxylate transporter MCT1 supports an NF-κB/IL-8 pathway that drives tumor angiogenesis. Cancer Res 71(7):2550–2560, 2011.

CHAPTER 70  Vasculogenesis 187. Veiseh O, et al: Cancer cell invasion: Treatment and monitoring opportunities in nanomedicine. Adv Drug Deliv Rev 63(8):582–596, 2011. 188. Venero JL, Machado A, Cano J: Importance of aquaporins in the physiopathology of brain edema. Curr Pharm Des 10(18):2153–2161, 2004. 189. Wang C-Y, et al: NF-κB antiapoptosis: Induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase-8 activation. Science 281(5383):1680–1683, 1998. 190. Warburg O: On the origin of cancer cells. Science 123(3191):309–314, 1956. 191. Weidner N, et al: Tumor angiogenesis and metastasis—Correlation in invasive breast carcinoma. NEJM 324(1):1–8, 1991. 192. Weinberg RA: The retinoblastoma protein and cell cycle control. Cell 81(3):323–330, 1995. 193. Williams AC, Collard TJ, Paraskeva C: An acidic environment leads to p53 dependent induction of apoptosis in human adenoma and carcinoma cell lines: Implications for clonal selection during colorectal carcinogenesis. Oncogene 18(21):3199–3204, 1999. 194. Williams S, et al: Differential effects of the proteasome inhibitor bortezomib on apoptosis and angiogenesis in human prostate tumor xenografts. Mol Cancer Ther 2(9):835–843, 2003. 195. Wolf JS, et al: IL (interleukin)-1α promotes nuclear factor-κB and AP-1-induced IL-8 expression, cell survival, and proliferation in head and neck squamous cell carcinomas. Clin Cancer Res 7(6):1812–1820, 2001.

2671

196. Workman P, et al: Minimally invasive pharmacokinetic and pharmacodynamic technologies in hypothesis-testing clinical trials of innovative therapies. J Natl Cancer Inst 98(9):580–598, 2006. 197. Wu H, Exner AA, Shi H, et al: Dynamic evolutionary changes in blood flow measured by MDCT in a hepatic VX2 tumor implant over an extended 28-day growth period: Time-density curve analysis. Acad Radiol 16(12):1483–1492, 2009. 198. Yin F, Giuliano AE, Van Herle AJ: Signal pathways involved in apigenin inhibition of growth and induction of apoptosis of human anaplastic thyroid cancer cells (ARO). Anticancer Res 19(5B):4297–4303, 1998. 199. Yin F, et al: Apigenin inhibits growth and induces G2/M arrest by modulating cyclin-CDK regulators and ERK MAP kinase activation in breast carcinoma cells. Anticancer Res 21(1A):413–420, 2000. 200. Yoon SH, et al: Multiphasic MDCT enhancement pattern of hepatocellular carcinoma smaller than 3 cm in diameter: Tumor size and cellular differentiation. AJR Am J Roentgenol 193(6):W482–W489, 2009. 201. Yoshiji H, Harris SR, Thorgeirsson UP: Vascular endothelial growth factor is essential for initial but not continued in vivo growth of human breast carcinoma cells. Cancer Res 57(18):3924–3928, 1997. 202. Zeng Z, et al: Targeting the leukemia microenvironment by CXCR4 inhibition overcomes resistance to kinase inhibitors and chemotherapy in AML. Blood 113(24):6215–6224, 2009. 203. Zhu H, et al: Bik/NBK accumulation correlates with apoptosisinduction by bortezomib (PS-341, Velcade) and other proteasome inhibitors. Oncogene 24(31):4993–4999, 2005.